From 1cff94c6fecdc54d6f022ae5a22888f8272804a1 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 21 Apr 2005 16:20:35 -0700 Subject: [PATCH] fix subarch breakage in amd dual core updates The patch to arch/i386/kernel/cpu/amd.c relies on the variable cpu_core_id which is defined in i386/kernel/smpboot.c. This means it is only present if CONFIG_X86_SMP is defined, not CONFIG_SMP (alternative SMP harnesses won't have it, which is why it breaks voyager). Signed-off-by: James Bottomley Signed-off-by: Linus Torvalds --- arch/i386/kernel/cpu/amd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/cpu/amd.c b/arch/i386/kernel/cpu/amd.c index 8d182e875cd..16dbc4151be 100644 --- a/arch/i386/kernel/cpu/amd.c +++ b/arch/i386/kernel/cpu/amd.c @@ -24,7 +24,7 @@ __asm__(".align 4\nvide: ret"); static void __init init_amd(struct cpuinfo_x86 *c) { -#ifdef CONFIG_SMP +#ifdef CONFIG_X86_SMP int cpu = c == &boot_cpu_data ? 0 : c - cpu_data; #endif u32 l, h; @@ -198,7 +198,7 @@ static void __init init_amd(struct cpuinfo_x86 *c) c->x86_num_cores = 1; } -#ifdef CONFIG_SMP +#ifdef CONFIG_X86_SMP /* * On a AMD dual core setup the lower bits of the APIC id * distingush the cores. Assumes number of cores is a power -- cgit v1.2.3-70-g09d2 From b4bca26c0160f48b4eb04f21d31a229832732013 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 21 Apr 2005 21:42:34 -0700 Subject: [SPARC]: Provide generic ioctls in Sparc RTC driver. Provide support for drivers/char/rtc.c ioctls in the Mostek rtc driver as well as the Sparc specific RTCGET and RTCSET. This allows userspace to be much less messy. Currently util-linux and other spots jump through hoops trying various ioctl variants until it hits the right one whatever driver actually being used supports. Eventually all of this should move over to the genrtc.c driver, but not today... While we are here, fix up the register types for sparse. Thanks to Frans Pop for helping point out this issue. Signed-off-by: David S. Miller --- arch/sparc64/kernel/time.c | 37 +++++++++------- drivers/sbus/char/rtc.c | 103 +++++++++++++++++++++++++++++++++++++++++-- include/asm-sparc64/mostek.h | 6 +-- 3 files changed, 123 insertions(+), 23 deletions(-) (limited to 'arch') diff --git a/arch/sparc64/kernel/time.c b/arch/sparc64/kernel/time.c index 6a717d4d2bc..c60785c046b 100644 --- a/arch/sparc64/kernel/time.c +++ b/arch/sparc64/kernel/time.c @@ -48,7 +48,7 @@ DEFINE_SPINLOCK(mostek_lock); DEFINE_SPINLOCK(rtc_lock); -unsigned long mstk48t02_regs = 0UL; +void * __iomem mstk48t02_regs = 0UL; #ifdef CONFIG_PCI unsigned long ds1287_regs = 0UL; #endif @@ -59,8 +59,8 @@ u64 jiffies_64 = INITIAL_JIFFIES; EXPORT_SYMBOL(jiffies_64); -static unsigned long mstk48t08_regs = 0UL; -static unsigned long mstk48t59_regs = 0UL; +static void * __iomem mstk48t08_regs; +static void * __iomem mstk48t59_regs; static int set_rtc_mmss(unsigned long); @@ -520,7 +520,7 @@ void timer_tick_interrupt(struct pt_regs *regs) /* Kick start a stopped clock (procedure from the Sun NVRAM/hostid FAQ). */ static void __init kick_start_clock(void) { - unsigned long regs = mstk48t02_regs; + void * __iomem regs = mstk48t02_regs; u8 sec, tmp; int i, count; @@ -604,7 +604,7 @@ static void __init kick_start_clock(void) /* Return nonzero if the clock chip battery is low. */ static int __init has_low_battery(void) { - unsigned long regs = mstk48t02_regs; + void * __iomem regs = mstk48t02_regs; u8 data1, data2; spin_lock_irq(&mostek_lock); @@ -623,7 +623,7 @@ static int __init has_low_battery(void) static void __init set_system_time(void) { unsigned int year, mon, day, hour, min, sec; - unsigned long mregs = mstk48t02_regs; + void * __iomem mregs = mstk48t02_regs; #ifdef CONFIG_PCI unsigned long dregs = ds1287_regs; #else @@ -843,7 +843,8 @@ void __init clock_probe(void) !strcmp(model, "m5823")) { ds1287_regs = edev->resource[0].start; } else { - mstk48t59_regs = edev->resource[0].start; + mstk48t59_regs = (void * __iomem) + edev->resource[0].start; mstk48t02_regs = mstk48t59_regs + MOSTEK_48T59_48T02; } break; @@ -865,7 +866,8 @@ try_isa_clock: !strcmp(model, "m5823")) { ds1287_regs = isadev->resource.start; } else { - mstk48t59_regs = isadev->resource.start; + mstk48t59_regs = (void * __iomem) + isadev->resource.start; mstk48t02_regs = mstk48t59_regs + MOSTEK_48T59_48T02; } break; @@ -893,21 +895,24 @@ try_isa_clock: } if(model[5] == '0' && model[6] == '2') { - mstk48t02_regs = (((u64)clk_reg[0].phys_addr) | - (((u64)clk_reg[0].which_io)<<32UL)); + mstk48t02_regs = (void * __iomem) + (((u64)clk_reg[0].phys_addr) | + (((u64)clk_reg[0].which_io)<<32UL)); } else if(model[5] == '0' && model[6] == '8') { - mstk48t08_regs = (((u64)clk_reg[0].phys_addr) | - (((u64)clk_reg[0].which_io)<<32UL)); + mstk48t08_regs = (void * __iomem) + (((u64)clk_reg[0].phys_addr) | + (((u64)clk_reg[0].which_io)<<32UL)); mstk48t02_regs = mstk48t08_regs + MOSTEK_48T08_48T02; } else { - mstk48t59_regs = (((u64)clk_reg[0].phys_addr) | - (((u64)clk_reg[0].which_io)<<32UL)); + mstk48t59_regs = (void * __iomem) + (((u64)clk_reg[0].phys_addr) | + (((u64)clk_reg[0].which_io)<<32UL)); mstk48t02_regs = mstk48t59_regs + MOSTEK_48T59_48T02; } break; } - if (mstk48t02_regs != 0UL) { + if (mstk48t02_regs != NULL) { /* Report a low battery voltage condition. */ if (has_low_battery()) prom_printf("NVRAM: Low battery voltage!\n"); @@ -1087,7 +1092,7 @@ unsigned long long sched_clock(void) static int set_rtc_mmss(unsigned long nowtime) { int real_seconds, real_minutes, chip_minutes; - unsigned long mregs = mstk48t02_regs; + void * __iomem mregs = mstk48t02_regs; #ifdef CONFIG_PCI unsigned long dregs = ds1287_regs; #else diff --git a/drivers/sbus/char/rtc.c b/drivers/sbus/char/rtc.c index bf3273eb1c8..49d1cd99d5a 100644 --- a/drivers/sbus/char/rtc.c +++ b/drivers/sbus/char/rtc.c @@ -28,6 +28,42 @@ static int rtc_busy = 0; +/* This is the structure layout used by drivers/char/rtc.c, we + * support that driver's ioctls so that things are less messy in + * userspace. + */ +struct rtc_time_generic { + int tm_sec; + int tm_min; + int tm_hour; + int tm_mday; + int tm_mon; + int tm_year; + int tm_wday; + int tm_yday; + int tm_isdst; +}; +#define RTC_AIE_ON _IO('p', 0x01) /* Alarm int. enable on */ +#define RTC_AIE_OFF _IO('p', 0x02) /* ... off */ +#define RTC_UIE_ON _IO('p', 0x03) /* Update int. enable on */ +#define RTC_UIE_OFF _IO('p', 0x04) /* ... off */ +#define RTC_PIE_ON _IO('p', 0x05) /* Periodic int. enable on */ +#define RTC_PIE_OFF _IO('p', 0x06) /* ... off */ +#define RTC_WIE_ON _IO('p', 0x0f) /* Watchdog int. enable on */ +#define RTC_WIE_OFF _IO('p', 0x10) /* ... off */ +#define RTC_RD_TIME _IOR('p', 0x09, struct rtc_time_generic) /* Read RTC time */ +#define RTC_SET_TIME _IOW('p', 0x0a, struct rtc_time_generic) /* Set RTC time */ +#define RTC_ALM_SET _IOW('p', 0x07, struct rtc_time) /* Set alarm time */ +#define RTC_ALM_READ _IOR('p', 0x08, struct rtc_time) /* Read alarm time */ +#define RTC_IRQP_READ _IOR('p', 0x0b, unsigned long) /* Read IRQ rate */ +#define RTC_IRQP_SET _IOW('p', 0x0c, unsigned long) /* Set IRQ rate */ +#define RTC_EPOCH_READ _IOR('p', 0x0d, unsigned long) /* Read epoch */ +#define RTC_EPOCH_SET _IOW('p', 0x0e, unsigned long) /* Set epoch */ +#define RTC_WKALM_SET _IOW('p', 0x0f, struct rtc_wkalrm)/* Set wakeup alarm*/ +#define RTC_WKALM_RD _IOR('p', 0x10, struct rtc_wkalrm)/* Get wakeup alarm*/ +#define RTC_PLL_GET _IOR('p', 0x11, struct rtc_pll_info) /* Get PLL correction */ +#define RTC_PLL_SET _IOW('p', 0x12, struct rtc_pll_info) /* Set PLL correction */ + /* Retrieve the current date and time from the real time clock. */ static void get_rtc_time(struct rtc_time *t) { @@ -82,29 +118,87 @@ void set_rtc_time(struct rtc_time *t) spin_unlock_irq(&mostek_lock); } +static int put_rtc_time_generic(void __user *argp, struct rtc_time *tm) +{ + struct rtc_time_generic __user *utm = argp; + + if (__put_user(tm->sec, &utm->tm_sec) || + __put_user(tm->min, &utm->tm_min) || + __put_user(tm->hour, &utm->tm_hour) || + __put_user(tm->dom, &utm->tm_mday) || + __put_user(tm->month, &utm->tm_mon) || + __put_user(tm->year, &utm->tm_year) || + __put_user(tm->dow, &utm->tm_wday) || + __put_user(0, &utm->tm_yday) || + __put_user(0, &utm->tm_isdst)) + return -EFAULT; + + return 0; +} + +static int get_rtc_time_generic(struct rtc_time *tm, void __user *argp) +{ + struct rtc_time_generic __user *utm = argp; + + if (__get_user(tm->sec, &utm->tm_sec) || + __get_user(tm->min, &utm->tm_min) || + __get_user(tm->hour, &utm->tm_hour) || + __get_user(tm->dom, &utm->tm_mday) || + __get_user(tm->month, &utm->tm_mon) || + __get_user(tm->year, &utm->tm_year) || + __get_user(tm->dow, &utm->tm_wday)) + return -EFAULT; + + return 0; +} + static int rtc_ioctl(struct inode *inode, struct file *file, unsigned int cmd, unsigned long arg) { struct rtc_time rtc_tm; void __user *argp = (void __user *)arg; - switch (cmd) - { + switch (cmd) { + /* No interrupt support, return an error + * compatible with drivers/char/rtc.c + */ + case RTC_AIE_OFF: + case RTC_AIE_ON: + case RTC_PIE_OFF: + case RTC_PIE_ON: + case RTC_UIE_OFF: + case RTC_UIE_ON: + case RTC_IRQP_READ: + case RTC_IRQP_SET: + case RTC_EPOCH_SET: + case RTC_EPOCH_READ: + return -EINVAL; + case RTCGET: + case RTC_RD_TIME: memset(&rtc_tm, 0, sizeof(struct rtc_time)); get_rtc_time(&rtc_tm); - if (copy_to_user(argp, &rtc_tm, sizeof(struct rtc_time))) + if (cmd == RTCGET) { + if (copy_to_user(argp, &rtc_tm, + sizeof(struct rtc_time))) + return -EFAULT; + } else if (put_rtc_time_generic(argp, &rtc_tm)) return -EFAULT; return 0; case RTCSET: + case RTC_SET_TIME: if (!capable(CAP_SYS_TIME)) return -EPERM; - if (copy_from_user(&rtc_tm, argp, sizeof(struct rtc_time))) + if (cmd == RTCSET) { + if (copy_from_user(&rtc_tm, argp, + sizeof(struct rtc_time))) + return -EFAULT; + } else if (get_rtc_time_generic(&rtc_tm, argp)) return -EFAULT; set_rtc_time(&rtc_tm); @@ -164,6 +258,7 @@ static int __init rtc_sun_init(void) printk(KERN_ERR "rtc: unable to get misc minor for Mostek\n"); return error; } + printk("rtc_sun_init: Registered Mostek RTC driver.\n"); return 0; } diff --git a/include/asm-sparc64/mostek.h b/include/asm-sparc64/mostek.h index ccf2f5f82d7..1f9b1356a48 100644 --- a/include/asm-sparc64/mostek.h +++ b/include/asm-sparc64/mostek.h @@ -38,7 +38,7 @@ * * We now deal with physical addresses for I/O to the chip. -DaveM */ -static __inline__ u8 mostek_read(unsigned long addr) +static __inline__ u8 mostek_read(void * __iomem addr) { u8 ret; @@ -48,7 +48,7 @@ static __inline__ u8 mostek_read(unsigned long addr) return ret; } -static __inline__ void mostek_write(unsigned long addr, u8 val) +static __inline__ void mostek_write(void * __iomem addr, u8 val) { __asm__ __volatile__("stba %0, [%1] %2" : /* no outputs */ @@ -67,7 +67,7 @@ static __inline__ void mostek_write(unsigned long addr, u8 val) #define MOSTEK_YEAR 0x07ffUL extern spinlock_t mostek_lock; -extern unsigned long mstk48t02_regs; +extern void *__iomem mstk48t02_regs; /* Control register values. */ #define MSTK_CREG_WRITE 0x80 /* Must set this before placing values. */ -- cgit v1.2.3-70-g09d2 From efab7739d99eae948971140b2aa3dddf7f72c900 Mon Sep 17 00:00:00 2001 From: Alexander Nyberg Date: Fri, 22 Apr 2005 10:22:07 -0700 Subject: [PATCH] x86_64: fix new out of line put_user() The labels after the last put_user patch were misplaced so exceptions on the real mov instructions would not be handled. Noted by Brian Gerst --- arch/x86_64/lib/putuser.S | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/x86_64/lib/putuser.S b/arch/x86_64/lib/putuser.S index d9d3e5ec9ad..5828b819166 100644 --- a/arch/x86_64/lib/putuser.S +++ b/arch/x86_64/lib/putuser.S @@ -49,8 +49,8 @@ __put_user_2: jc 20f cmpq threadinfo_addr_limit(%r8),%rcx jae 20f -2: decq %rcx - movw %dx,(%rcx) + decq %rcx +2: movw %dx,(%rcx) xorl %eax,%eax ret 20: decq %rcx @@ -64,8 +64,8 @@ __put_user_4: jc 30f cmpq threadinfo_addr_limit(%r8),%rcx jae 30f -3: subq $3,%rcx - movl %edx,(%rcx) + subq $3,%rcx +3: movl %edx,(%rcx) xorl %eax,%eax ret 30: subq $3,%rcx @@ -79,8 +79,8 @@ __put_user_8: jc 40f cmpq threadinfo_addr_limit(%r8),%rcx jae 40f -4: subq $7,%rcx - movq %rdx,(%rcx) + subq $7,%rcx +4: movq %rdx,(%rcx) xorl %eax,%eax ret 40: subq $7,%rcx -- cgit v1.2.3-70-g09d2 From 7130667107cd3ab9d6802b69bab63c7d22f20bd4 Mon Sep 17 00:00:00 2001 From: Arun Sharma Date: Fri, 22 Apr 2005 13:06:47 -0700 Subject: [IA64] ia32_signal.c: erroneous use of memset/memcpy Found by Alexander Nyberg, improved by Bjorn Helgaas. - Fix the incorrect argument to sizeof() - looks like memcpy() code pass was dervived from code that used copy_from_user(). But in this case we are doing to kernel space to kernel space copy, so memcpy is the right routine, but it doesn't return an error code. Signed-off-by: Arun Sharma Signed-off-by: Tony Luck --- arch/ia64/ia32/ia32_signal.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/ia64/ia32/ia32_signal.c b/arch/ia64/ia32/ia32_signal.c index 19b02adce68..ebb89be2aa2 100644 --- a/arch/ia64/ia32/ia32_signal.c +++ b/arch/ia64/ia32/ia32_signal.c @@ -460,10 +460,9 @@ __ia32_rt_sigsuspend (compat_sigset_t *sset, unsigned int sigsetsize, struct sig sigset_t oldset, set; scr->scratch_unat = 0; /* avoid leaking kernel bits to user level */ - memset(&set, 0, sizeof(&set)); + memset(&set, 0, sizeof(set)); - if (memcpy(&set.sig, &sset->sig, sigsetsize)) - return -EFAULT; + memcpy(&set.sig, &sset->sig, sigsetsize); sigdelsetmask(&set, ~_BLOCKABLE); -- cgit v1.2.3-70-g09d2 From b8d8b883e6f029e99c35c88f853501740e322131 Mon Sep 17 00:00:00 2001 From: Ashok Raj Date: Fri, 22 Apr 2005 14:44:40 -0700 Subject: [IA64] cpu hotplug: return offlined cpus to SAL This patch is required to support cpu removal for IPF systems. Existing code just fakes the real offline by keeping it run the idle thread, and polling for the bit to re-appear in the cpu_state to get out of the idle loop. For the cpu-offline to work correctly, we need to pass control of this CPU back to SAL so it can continue in the boot-rendez mode. This gives the SAL control to not pick this cpu as the monarch processor for global MCA events, and addition does not wait for this cpu to checkin with SAL for global MCA events as well. The handoff is implemented as documented in SAL specification section 3.2.5.1 "OS_BOOT_RENDEZ to SAL return State" Signed-off-by: Ashok Raj Signed-off-by: Tony Luck --- arch/ia64/kernel/head.S | 280 ++++++++++++++++++++++++++++++++++++++++----- arch/ia64/kernel/mca_asm.S | 88 ++++++++------ arch/ia64/kernel/process.c | 22 ++-- arch/ia64/kernel/smpboot.c | 81 ++++++++----- include/asm-ia64/sal.h | 38 ++++++ 5 files changed, 399 insertions(+), 110 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/head.S b/arch/ia64/kernel/head.S index 105c7fec8c6..0d535d65eea 100644 --- a/arch/ia64/kernel/head.S +++ b/arch/ia64/kernel/head.S @@ -15,6 +15,8 @@ * Copyright (C) 1999 Don Dugger * Copyright (C) 2002 Fenghua Yu * -Optimize __ia64_save_fpu() and __ia64_load_fpu() for Itanium 2. + * Copyright (C) 2004 Ashok Raj + * Support for CPU Hotplug */ #include @@ -29,6 +31,134 @@ #include #include #include +#include + +#ifdef CONFIG_HOTPLUG_CPU +#define SAL_PSR_BITS_TO_SET \ + (IA64_PSR_AC | IA64_PSR_BN | IA64_PSR_MFH | IA64_PSR_MFL) + +#define SAVE_FROM_REG(src, ptr, dest) \ + mov dest=src;; \ + st8 [ptr]=dest,0x08 + +#define RESTORE_REG(reg, ptr, _tmp) \ + ld8 _tmp=[ptr],0x08;; \ + mov reg=_tmp + +#define SAVE_BREAK_REGS(ptr, _idx, _breg, _dest)\ + mov ar.lc=IA64_NUM_DBG_REGS-1;; \ + mov _idx=0;; \ +1: \ + SAVE_FROM_REG(_breg[_idx], ptr, _dest);; \ + add _idx=1,_idx;; \ + br.cloop.sptk.many 1b + +#define RESTORE_BREAK_REGS(ptr, _idx, _breg, _tmp, _lbl)\ + mov ar.lc=IA64_NUM_DBG_REGS-1;; \ + mov _idx=0;; \ +_lbl: RESTORE_REG(_breg[_idx], ptr, _tmp);; \ + add _idx=1, _idx;; \ + br.cloop.sptk.many _lbl + +#define SAVE_ONE_RR(num, _reg, _tmp) \ + movl _tmp=(num<<61);; \ + mov _reg=rr[_tmp] + +#define SAVE_REGION_REGS(_tmp, _r0, _r1, _r2, _r3, _r4, _r5, _r6, _r7) \ + SAVE_ONE_RR(0,_r0, _tmp);; \ + SAVE_ONE_RR(1,_r1, _tmp);; \ + SAVE_ONE_RR(2,_r2, _tmp);; \ + SAVE_ONE_RR(3,_r3, _tmp);; \ + SAVE_ONE_RR(4,_r4, _tmp);; \ + SAVE_ONE_RR(5,_r5, _tmp);; \ + SAVE_ONE_RR(6,_r6, _tmp);; \ + SAVE_ONE_RR(7,_r7, _tmp);; + +#define STORE_REGION_REGS(ptr, _r0, _r1, _r2, _r3, _r4, _r5, _r6, _r7) \ + st8 [ptr]=_r0, 8;; \ + st8 [ptr]=_r1, 8;; \ + st8 [ptr]=_r2, 8;; \ + st8 [ptr]=_r3, 8;; \ + st8 [ptr]=_r4, 8;; \ + st8 [ptr]=_r5, 8;; \ + st8 [ptr]=_r6, 8;; \ + st8 [ptr]=_r7, 8;; + +#define RESTORE_REGION_REGS(ptr, _idx1, _idx2, _tmp) \ + mov ar.lc=0x08-1;; \ + movl _idx1=0x00;; \ +RestRR: \ + dep.z _idx2=_idx1,61,3;; \ + ld8 _tmp=[ptr],8;; \ + mov rr[_idx2]=_tmp;; \ + srlz.d;; \ + add _idx1=1,_idx1;; \ + br.cloop.sptk.few RestRR + +/* + * Adjust region registers saved before starting to save + * break regs and rest of the states that need to be preserved. + */ +#define SAL_TO_OS_BOOT_HANDOFF_STATE_SAVE(_reg1,_reg2,_pred) \ + SAVE_FROM_REG(b0,_reg1,_reg2);; \ + SAVE_FROM_REG(b1,_reg1,_reg2);; \ + SAVE_FROM_REG(b2,_reg1,_reg2);; \ + SAVE_FROM_REG(b3,_reg1,_reg2);; \ + SAVE_FROM_REG(b4,_reg1,_reg2);; \ + SAVE_FROM_REG(b5,_reg1,_reg2);; \ + st8 [_reg1]=r1,0x08;; \ + st8 [_reg1]=r12,0x08;; \ + st8 [_reg1]=r13,0x08;; \ + SAVE_FROM_REG(ar.fpsr,_reg1,_reg2);; \ + SAVE_FROM_REG(ar.pfs,_reg1,_reg2);; \ + SAVE_FROM_REG(ar.rnat,_reg1,_reg2);; \ + SAVE_FROM_REG(ar.unat,_reg1,_reg2);; \ + SAVE_FROM_REG(ar.bspstore,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.dcr,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.iva,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.pta,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.itv,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.pmv,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.cmcv,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.lrr0,_reg1,_reg2);; \ + SAVE_FROM_REG(cr.lrr1,_reg1,_reg2);; \ + st8 [_reg1]=r4,0x08;; \ + st8 [_reg1]=r5,0x08;; \ + st8 [_reg1]=r6,0x08;; \ + st8 [_reg1]=r7,0x08;; \ + st8 [_reg1]=_pred,0x08;; \ + SAVE_FROM_REG(ar.lc, _reg1, _reg2);; \ + stf.spill.nta [_reg1]=f2,16;; \ + stf.spill.nta [_reg1]=f3,16;; \ + stf.spill.nta [_reg1]=f4,16;; \ + stf.spill.nta [_reg1]=f5,16;; \ + stf.spill.nta [_reg1]=f16,16;; \ + stf.spill.nta [_reg1]=f17,16;; \ + stf.spill.nta [_reg1]=f18,16;; \ + stf.spill.nta [_reg1]=f19,16;; \ + stf.spill.nta [_reg1]=f20,16;; \ + stf.spill.nta [_reg1]=f21,16;; \ + stf.spill.nta [_reg1]=f22,16;; \ + stf.spill.nta [_reg1]=f23,16;; \ + stf.spill.nta [_reg1]=f24,16;; \ + stf.spill.nta [_reg1]=f25,16;; \ + stf.spill.nta [_reg1]=f26,16;; \ + stf.spill.nta [_reg1]=f27,16;; \ + stf.spill.nta [_reg1]=f28,16;; \ + stf.spill.nta [_reg1]=f29,16;; \ + stf.spill.nta [_reg1]=f30,16;; \ + stf.spill.nta [_reg1]=f31,16;; + +#else +#define SAL_TO_OS_BOOT_HANDOFF_STATE_SAVE(a1,a2) +#define SAVE_REGION_REGS(_tmp, _r0, _r1, _r2, _r3, _r4, _r5, _r6, _r7) +#define STORE_REGION_REGS(ptr, _r0, _r1, _r2, _r3, _r4, _r5, _r6, _r7) +#endif + +#define SET_ONE_RR(num, pgsize, _tmp1, _tmp2, vhpt) \ + movl _tmp1=(num << 61);; \ + mov _tmp2=((ia64_rid(IA64_REGION_ID_KERNEL, (num<<61)) << 8) | (pgsize << 2) | vhpt);; \ + mov rr[_tmp1]=_tmp2 .section __special_page_section,"ax" @@ -63,6 +193,12 @@ start_ap: ;; srlz.i ;; + /* + * Save the region registers, predicate before they get clobbered + */ + SAVE_REGION_REGS(r2, r8,r9,r10,r11,r12,r13,r14,r15); + mov r25=pr;; + /* * Initialize kernel region registers: * rr[0]: VHPT enabled, page size = PAGE_SHIFT @@ -76,32 +212,14 @@ start_ap: * We initialize all of them to prevent inadvertently assuming * something about the state of address translation early in boot. */ - mov r6=((ia64_rid(IA64_REGION_ID_KERNEL, (0<<61)) << 8) | (PAGE_SHIFT << 2) | 1) - movl r7=(0<<61) - mov r8=((ia64_rid(IA64_REGION_ID_KERNEL, (1<<61)) << 8) | (PAGE_SHIFT << 2) | 1) - movl r9=(1<<61) - mov r10=((ia64_rid(IA64_REGION_ID_KERNEL, (2<<61)) << 8) | (PAGE_SHIFT << 2) | 1) - movl r11=(2<<61) - mov r12=((ia64_rid(IA64_REGION_ID_KERNEL, (3<<61)) << 8) | (PAGE_SHIFT << 2) | 1) - movl r13=(3<<61) - mov r14=((ia64_rid(IA64_REGION_ID_KERNEL, (4<<61)) << 8) | (PAGE_SHIFT << 2) | 1) - movl r15=(4<<61) - mov r16=((ia64_rid(IA64_REGION_ID_KERNEL, (5<<61)) << 8) | (PAGE_SHIFT << 2) | 1) - movl r17=(5<<61) - mov r18=((ia64_rid(IA64_REGION_ID_KERNEL, (6<<61)) << 8) | (IA64_GRANULE_SHIFT << 2)) - movl r19=(6<<61) - mov r20=((ia64_rid(IA64_REGION_ID_KERNEL, (7<<61)) << 8) | (IA64_GRANULE_SHIFT << 2)) - movl r21=(7<<61) - ;; - mov rr[r7]=r6 - mov rr[r9]=r8 - mov rr[r11]=r10 - mov rr[r13]=r12 - mov rr[r15]=r14 - mov rr[r17]=r16 - mov rr[r19]=r18 - mov rr[r21]=r20 - ;; + SET_ONE_RR(0, PAGE_SHIFT, r2, r16, 1);; + SET_ONE_RR(1, PAGE_SHIFT, r2, r16, 1);; + SET_ONE_RR(2, PAGE_SHIFT, r2, r16, 1);; + SET_ONE_RR(3, PAGE_SHIFT, r2, r16, 1);; + SET_ONE_RR(4, PAGE_SHIFT, r2, r16, 1);; + SET_ONE_RR(5, PAGE_SHIFT, r2, r16, 1);; + SET_ONE_RR(6, IA64_GRANULE_SHIFT, r2, r16, 0);; + SET_ONE_RR(7, IA64_GRANULE_SHIFT, r2, r16, 0);; /* * Now pin mappings into the TLB for kernel text and data */ @@ -142,6 +260,13 @@ start_ap: ;; 1: // now we are in virtual mode + movl r2=sal_state_for_booting_cpu;; + ld8 r16=[r2];; + + STORE_REGION_REGS(r16, r8,r9,r10,r11,r12,r13,r14,r15); + SAL_TO_OS_BOOT_HANDOFF_STATE_SAVE(r16,r17,r25) + ;; + // set IVT entry point---can't access I/O ports without it movl r3=ia64_ivt ;; @@ -211,12 +336,13 @@ start_ap: mov IA64_KR(CURRENT_STACK)=r16 mov r13=r2 /* - * Reserve space at the top of the stack for "struct pt_regs". Kernel threads - * don't store interesting values in that structure, but the space still needs - * to be there because time-critical stuff such as the context switching can - * be implemented more efficiently (for example, __switch_to() + * Reserve space at the top of the stack for "struct pt_regs". Kernel + * threads don't store interesting values in that structure, but the space + * still needs to be there because time-critical stuff such as the context + * switching can be implemented more efficiently (for example, __switch_to() * always sets the psr.dfh bit of the task it is switching to). */ + addl r12=IA64_STK_OFFSET-IA64_PT_REGS_SIZE-16,r2 addl r2=IA64_RBS_OFFSET,r2 // initialize the RSE mov ar.rsc=0 // place RSE in enforced lazy mode @@ -993,4 +1119,98 @@ END(ia64_spinlock_contention) #endif +#ifdef CONFIG_HOTPLUG_CPU +GLOBAL_ENTRY(ia64_jump_to_sal) + alloc r16=ar.pfs,1,0,0,0;; + rsm psr.i | psr.ic +{ + flushrs + srlz.i +} + tpa r25=in0 + movl r18=tlb_purge_done;; + DATA_VA_TO_PA(r18);; + mov b1=r18 // Return location + movl r18=ia64_do_tlb_purge;; + DATA_VA_TO_PA(r18);; + mov b2=r18 // doing tlb_flush work + mov ar.rsc=0 // Put RSE in enforced lazy, LE mode + movl r17=1f;; + DATA_VA_TO_PA(r17);; + mov cr.iip=r17 + movl r16=SAL_PSR_BITS_TO_SET;; + mov cr.ipsr=r16 + mov cr.ifs=r0;; + rfi;; +1: + /* + * Invalidate all TLB data/inst + */ + br.sptk.many b2;; // jump to tlb purge code + +tlb_purge_done: + RESTORE_REGION_REGS(r25, r17,r18,r19);; + RESTORE_REG(b0, r25, r17);; + RESTORE_REG(b1, r25, r17);; + RESTORE_REG(b2, r25, r17);; + RESTORE_REG(b3, r25, r17);; + RESTORE_REG(b4, r25, r17);; + RESTORE_REG(b5, r25, r17);; + ld8 r1=[r25],0x08;; + ld8 r12=[r25],0x08;; + ld8 r13=[r25],0x08;; + RESTORE_REG(ar.fpsr, r25, r17);; + RESTORE_REG(ar.pfs, r25, r17);; + RESTORE_REG(ar.rnat, r25, r17);; + RESTORE_REG(ar.unat, r25, r17);; + RESTORE_REG(ar.bspstore, r25, r17);; + RESTORE_REG(cr.dcr, r25, r17);; + RESTORE_REG(cr.iva, r25, r17);; + RESTORE_REG(cr.pta, r25, r17);; + RESTORE_REG(cr.itv, r25, r17);; + RESTORE_REG(cr.pmv, r25, r17);; + RESTORE_REG(cr.cmcv, r25, r17);; + RESTORE_REG(cr.lrr0, r25, r17);; + RESTORE_REG(cr.lrr1, r25, r17);; + ld8 r4=[r25],0x08;; + ld8 r5=[r25],0x08;; + ld8 r6=[r25],0x08;; + ld8 r7=[r25],0x08;; + ld8 r17=[r25],0x08;; + mov pr=r17,-1;; + RESTORE_REG(ar.lc, r25, r17);; + /* + * Now Restore floating point regs + */ + ldf.fill.nta f2=[r25],16;; + ldf.fill.nta f3=[r25],16;; + ldf.fill.nta f4=[r25],16;; + ldf.fill.nta f5=[r25],16;; + ldf.fill.nta f16=[r25],16;; + ldf.fill.nta f17=[r25],16;; + ldf.fill.nta f18=[r25],16;; + ldf.fill.nta f19=[r25],16;; + ldf.fill.nta f20=[r25],16;; + ldf.fill.nta f21=[r25],16;; + ldf.fill.nta f22=[r25],16;; + ldf.fill.nta f23=[r25],16;; + ldf.fill.nta f24=[r25],16;; + ldf.fill.nta f25=[r25],16;; + ldf.fill.nta f26=[r25],16;; + ldf.fill.nta f27=[r25],16;; + ldf.fill.nta f28=[r25],16;; + ldf.fill.nta f29=[r25],16;; + ldf.fill.nta f30=[r25],16;; + ldf.fill.nta f31=[r25],16;; + + /* + * Now that we have done all the register restores + * we are now ready for the big DIVE to SAL Land + */ + ssm psr.ic;; + srlz.d;; + br.ret.sptk.many b0;; +END(ia64_jump_to_sal) +#endif /* CONFIG_HOTPLUG_CPU */ + #endif /* CONFIG_SMP */ diff --git a/arch/ia64/kernel/mca_asm.S b/arch/ia64/kernel/mca_asm.S index cf3f8014f9a..ef3fd7265b6 100644 --- a/arch/ia64/kernel/mca_asm.S +++ b/arch/ia64/kernel/mca_asm.S @@ -110,46 +110,19 @@ .global ia64_os_mca_dispatch_end .global ia64_sal_to_os_handoff_state .global ia64_os_to_sal_handoff_state + .global ia64_do_tlb_purge .text .align 16 -ia64_os_mca_dispatch: - - // Serialize all MCA processing - mov r3=1;; - LOAD_PHYSICAL(p0,r2,ia64_mca_serialize);; -ia64_os_mca_spin: - xchg8 r4=[r2],r3;; - cmp.ne p6,p0=r4,r0 -(p6) br ia64_os_mca_spin - - // Save the SAL to OS MCA handoff state as defined - // by SAL SPEC 3.0 - // NOTE : The order in which the state gets saved - // is dependent on the way the C-structure - // for ia64_mca_sal_to_os_state_t has been - // defined in include/asm/mca.h - SAL_TO_OS_MCA_HANDOFF_STATE_SAVE(r2) - ;; - - // LOG PROCESSOR STATE INFO FROM HERE ON.. -begin_os_mca_dump: - br ia64_os_mca_proc_state_dump;; - -ia64_os_mca_done_dump: - - LOAD_PHYSICAL(p0,r16,ia64_sal_to_os_handoff_state+56) - ;; - ld8 r18=[r16] // Get processor state parameter on existing PALE_CHECK. - ;; - tbit.nz p6,p7=r18,60 -(p7) br.spnt done_tlb_purge_and_reload - - // The following code purges TC and TR entries. Then reload all TC entries. - // Purge percpu data TC entries. -begin_tlb_purge_and_reload: +/* + * Just the TLB purge part is moved to a separate function + * so we can re-use the code for cpu hotplug code as well + * Caller should now setup b1, so we can branch once the + * tlb flush is complete. + */ +ia64_do_tlb_purge: #define O(member) IA64_CPUINFO_##member##_OFFSET GET_THIS_PADDR(r2, cpu_info) // load phys addr of cpu_info into r2 @@ -230,6 +203,51 @@ begin_tlb_purge_and_reload: ;; srlz.i ;; + // Now branch away to caller. + br.sptk.many b1 + ;; + +ia64_os_mca_dispatch: + + // Serialize all MCA processing + mov r3=1;; + LOAD_PHYSICAL(p0,r2,ia64_mca_serialize);; +ia64_os_mca_spin: + xchg8 r4=[r2],r3;; + cmp.ne p6,p0=r4,r0 +(p6) br ia64_os_mca_spin + + // Save the SAL to OS MCA handoff state as defined + // by SAL SPEC 3.0 + // NOTE : The order in which the state gets saved + // is dependent on the way the C-structure + // for ia64_mca_sal_to_os_state_t has been + // defined in include/asm/mca.h + SAL_TO_OS_MCA_HANDOFF_STATE_SAVE(r2) + ;; + + // LOG PROCESSOR STATE INFO FROM HERE ON.. +begin_os_mca_dump: + br ia64_os_mca_proc_state_dump;; + +ia64_os_mca_done_dump: + + LOAD_PHYSICAL(p0,r16,ia64_sal_to_os_handoff_state+56) + ;; + ld8 r18=[r16] // Get processor state parameter on existing PALE_CHECK. + ;; + tbit.nz p6,p7=r18,60 +(p7) br.spnt done_tlb_purge_and_reload + + // The following code purges TC and TR entries. Then reload all TC entries. + // Purge percpu data TC entries. +begin_tlb_purge_and_reload: + movl r18=ia64_reload_tr;; + LOAD_PHYSICAL(p0,r18,ia64_reload_tr);; + mov b1=r18;; + br.sptk.many ia64_do_tlb_purge;; + +ia64_reload_tr: // Finally reload the TR registers. // 1. Reload DTR/ITR registers for kernel. mov r18=KERNEL_TR_PAGE_SHIFT<<2 diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c index 91293388dd2..7c43aea5f7f 100644 --- a/arch/ia64/kernel/process.c +++ b/arch/ia64/kernel/process.c @@ -3,6 +3,7 @@ * * Copyright (C) 1998-2003 Hewlett-Packard Co * David Mosberger-Tang + * 04/11/17 Ashok Raj Added CPU Hotplug Support */ #define __KERNEL_SYSCALLS__ /* see */ #include @@ -200,27 +201,20 @@ default_idle (void) static inline void play_dead(void) { extern void ia64_cpu_local_tick (void); + unsigned int this_cpu = smp_processor_id(); + /* Ack it */ __get_cpu_var(cpu_state) = CPU_DEAD; - /* We shouldn't have to disable interrupts while dead, but - * some interrupts just don't seem to go away, and this makes - * it "work" for testing purposes. */ max_xtp(); local_irq_disable(); - /* Death loop */ - while (__get_cpu_var(cpu_state) != CPU_UP_PREPARE) - cpu_relax(); - + idle_task_exit(); + ia64_jump_to_sal(&sal_boot_rendez_state[this_cpu]); /* - * Enable timer interrupts from now on - * Not required if we put processor in SAL_BOOT_RENDEZ mode. + * The above is a point of no-return, the processor is + * expected to be in SAL loop now. */ - local_flush_tlb_all(); - cpu_set(smp_processor_id(), cpu_online_map); - wmb(); - ia64_cpu_local_tick (); - local_irq_enable(); + BUG(); } #else static inline void play_dead(void) diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c index 5318f0cbfc2..ca1536db339 100644 --- a/arch/ia64/kernel/smpboot.c +++ b/arch/ia64/kernel/smpboot.c @@ -9,6 +9,7 @@ * 02/07/31 David Mosberger Switch over to hotplug-CPU boot-sequence. * smp_boot_cpus()/smp_commence() is replaced by * smp_prepare_cpus()/__cpu_up()/smp_cpus_done(). + * 04/06/21 Ashok Raj Added CPU Hotplug Support */ #include @@ -58,6 +59,37 @@ #define Dprintk(x...) #endif +#ifdef CONFIG_HOTPLUG_CPU +/* + * Store all idle threads, this can be reused instead of creating + * a new thread. Also avoids complicated thread destroy functionality + * for idle threads. + */ +struct task_struct *idle_thread_array[NR_CPUS]; + +/* + * Global array allocated for NR_CPUS at boot time + */ +struct sal_to_os_boot sal_boot_rendez_state[NR_CPUS]; + +/* + * start_ap in head.S uses this to store current booting cpu + * info. + */ +struct sal_to_os_boot *sal_state_for_booting_cpu = &sal_boot_rendez_state[0]; + +#define set_brendez_area(x) (sal_state_for_booting_cpu = &sal_boot_rendez_state[(x)]); + +#define get_idle_for_cpu(x) (idle_thread_array[(x)]) +#define set_idle_for_cpu(x,p) (idle_thread_array[(x)] = (p)) + +#else + +#define get_idle_for_cpu(x) (NULL) +#define set_idle_for_cpu(x,p) +#define set_brendez_area(x) +#endif + /* * ITC synchronization related stuff: @@ -345,7 +377,6 @@ start_secondary (void *unused) { /* Early console may use I/O ports */ ia64_set_kr(IA64_KR_IO_BASE, __pa(ia64_iobase)); - Dprintk("start_secondary: starting CPU 0x%x\n", hard_smp_processor_id()); efi_map_pal_code(); cpu_init(); @@ -384,6 +415,13 @@ do_boot_cpu (int sapicid, int cpu) .done = COMPLETION_INITIALIZER(c_idle.done), }; DECLARE_WORK(work, do_fork_idle, &c_idle); + + c_idle.idle = get_idle_for_cpu(cpu); + if (c_idle.idle) { + init_idle(c_idle.idle, cpu); + goto do_rest; + } + /* * We can't use kernel_thread since we must avoid to reschedule the child. */ @@ -396,10 +434,15 @@ do_boot_cpu (int sapicid, int cpu) if (IS_ERR(c_idle.idle)) panic("failed fork for CPU %d", cpu); + + set_idle_for_cpu(cpu, c_idle.idle); + +do_rest: task_for_booting_cpu = c_idle.idle; Dprintk("Sending wakeup vector %lu to AP 0x%x/0x%x.\n", ap_wakeup_vector, cpu, sapicid); + set_brendez_area(cpu); platform_send_ipi(cpu, ap_wakeup_vector, IA64_IPI_DM_INT, 0); /* @@ -555,16 +598,6 @@ void __devinit smp_prepare_boot_cpu(void) #ifdef CONFIG_HOTPLUG_CPU extern void fixup_irqs(void); /* must be called with cpucontrol mutex held */ -static int __devinit cpu_enable(unsigned int cpu) -{ - per_cpu(cpu_state,cpu) = CPU_UP_PREPARE; - wmb(); - - while (!cpu_online(cpu)) - cpu_relax(); - return 0; -} - int __cpu_disable(void) { int cpu = smp_processor_id(); @@ -577,7 +610,7 @@ int __cpu_disable(void) fixup_irqs(); local_flush_tlb_all(); - printk ("Disabled cpu %u\n", smp_processor_id()); + cpu_clear(cpu, cpu_callin_map); return 0; } @@ -589,12 +622,7 @@ void __cpu_die(unsigned int cpu) /* They ack this in play_dead by setting CPU_DEAD */ if (per_cpu(cpu_state, cpu) == CPU_DEAD) { - /* - * TBD: Enable this when physical removal - * or when we put the processor is put in - * SAL_BOOT_RENDEZ mode - * cpu_clear(cpu, cpu_callin_map); - */ + printk ("CPU %d is now offline\n", cpu); return; } msleep(100); @@ -602,11 +630,6 @@ void __cpu_die(unsigned int cpu) printk(KERN_ERR "CPU %u didn't die...\n", cpu); } #else /* !CONFIG_HOTPLUG_CPU */ -static int __devinit cpu_enable(unsigned int cpu) -{ - return 0; -} - int __cpu_disable(void) { return -ENOSYS; @@ -648,16 +671,12 @@ __cpu_up (unsigned int cpu) return -EINVAL; /* - * Already booted.. just enable and get outa idle lool + * Already booted cpu? not valid anymore since we dont + * do idle loop tightspin anymore. */ if (cpu_isset(cpu, cpu_callin_map)) - { - cpu_enable(cpu); - local_irq_enable(); - while (!cpu_isset(cpu, cpu_online_map)) - mb(); - return 0; - } + return -EINVAL; + /* Processor goes to start_secondary(), sets online flag */ ret = do_boot_cpu(sapicid, cpu); if (ret < 0) diff --git a/include/asm-ia64/sal.h b/include/asm-ia64/sal.h index ea1ed377de4..240676f7539 100644 --- a/include/asm-ia64/sal.h +++ b/include/asm-ia64/sal.h @@ -832,6 +832,44 @@ extern int ia64_sal_oemcall_nolock(struct ia64_sal_retval *, u64, u64, u64, u64, u64, u64, u64, u64); extern int ia64_sal_oemcall_reentrant(struct ia64_sal_retval *, u64, u64, u64, u64, u64, u64, u64, u64); +#ifdef CONFIG_HOTPLUG_CPU +/* + * System Abstraction Layer Specification + * Section 3.2.5.1: OS_BOOT_RENDEZ to SAL return State. + * Note: region regs are stored first in head.S _start. Hence they must + * stay up front. + */ +struct sal_to_os_boot { + u64 rr[8]; /* Region Registers */ + u64 br[6]; /* br0: return addr into SAL boot rendez routine */ + u64 gr1; /* SAL:GP */ + u64 gr12; /* SAL:SP */ + u64 gr13; /* SAL: Task Pointer */ + u64 fpsr; + u64 pfs; + u64 rnat; + u64 unat; + u64 bspstore; + u64 dcr; /* Default Control Register */ + u64 iva; + u64 pta; + u64 itv; + u64 pmv; + u64 cmcv; + u64 lrr[2]; + u64 gr[4]; + u64 pr; /* Predicate registers */ + u64 lc; /* Loop Count */ + struct ia64_fpreg fp[20]; +}; + +/* + * Global array allocated for NR_CPUS at boot time + */ +extern struct sal_to_os_boot sal_boot_rendez_state[NR_CPUS]; + +extern void ia64_jump_to_sal(struct sal_to_os_boot *); +#endif extern void ia64_sal_handler_init(void *entry_point, void *gpval); -- cgit v1.2.3-70-g09d2 From df6c6804ce1d383c878ab071bce8125125d96abc Mon Sep 17 00:00:00 2001 From: Ashok Raj Date: Fri, 22 Apr 2005 14:46:24 -0700 Subject: [IA64] Fix build errors for !HOTPLUG case. Signed-off-by: Ashok Raj Signed-off-by: Tony Luck --- arch/ia64/kernel/head.S | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/head.S b/arch/ia64/kernel/head.S index 0d535d65eea..8d3a9291b47 100644 --- a/arch/ia64/kernel/head.S +++ b/arch/ia64/kernel/head.S @@ -95,6 +95,10 @@ RestRR: \ add _idx1=1,_idx1;; \ br.cloop.sptk.few RestRR +#define SET_AREA_FOR_BOOTING_CPU(reg1, reg2) \ + movl reg1=sal_state_for_booting_cpu;; \ + ld8 reg2=[reg1];; + /* * Adjust region registers saved before starting to save * break regs and rest of the states that need to be preserved. @@ -150,7 +154,8 @@ RestRR: \ stf.spill.nta [_reg1]=f31,16;; #else -#define SAL_TO_OS_BOOT_HANDOFF_STATE_SAVE(a1,a2) +#define SET_AREA_FOR_BOOTING_CPU(a1, a2) +#define SAL_TO_OS_BOOT_HANDOFF_STATE_SAVE(a1,a2, a3) #define SAVE_REGION_REGS(_tmp, _r0, _r1, _r2, _r3, _r4, _r5, _r6, _r7) #define STORE_REGION_REGS(ptr, _r0, _r1, _r2, _r3, _r4, _r5, _r6, _r7) #endif @@ -260,8 +265,7 @@ start_ap: ;; 1: // now we are in virtual mode - movl r2=sal_state_for_booting_cpu;; - ld8 r16=[r2];; + SET_AREA_FOR_BOOTING_CPU(r2, r16); STORE_REGION_REGS(r16, r8,r9,r10,r11,r12,r13,r14,r15); SAL_TO_OS_BOOT_HANDOFF_STATE_SAVE(r16,r17,r25) -- cgit v1.2.3-70-g09d2 From e3b9ab1a6da339312bb23747693805fa63e2ffd0 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Apr 2005 12:28:35 -0700 Subject: [PATCH] missing dependency on sparc64 CONFIG_HW_CONSOLE selects vt.c; without the stuff pulled by CONFIG_VT it will not build. Normally we get both in drivers/char/Kconfig and there HW_CONSOLE depends on VT. sparc64 does not pull drivers/char/Kconfig and has that sutff in arch/sparc64/Kconfig instead. However, it forgets to add the same dependency. As the result, turning VT off [which is possible] will end up with broken build. For no good reason... Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/sparc64/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/sparc64/Kconfig b/arch/sparc64/Kconfig index fb1189641c7..a72fd15d5ea 100644 --- a/arch/sparc64/Kconfig +++ b/arch/sparc64/Kconfig @@ -118,6 +118,7 @@ config VT_CONSOLE config HW_CONSOLE bool + depends on VT default y config SMP -- cgit v1.2.3-70-g09d2 From ef0299bf8e868c09bff13e6f3c4fc9b835f60471 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Apr 2005 12:28:36 -0700 Subject: [PATCH] mostek bogus sparse annotations fixed void * __iomem foo is not a pointer to iomem - it's an iomem variable containing void *. A pile of such guys in arch/sparc64/kernel/time.c, drivers/sbus/char/rtc.c and include/asm-sparc64/mostek.h turned into intended void __iomem *. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/sparc64/kernel/time.c | 24 ++++++++++++------------ drivers/sbus/char/rtc.c | 6 +++--- include/asm-sparc64/mostek.h | 6 +++--- 3 files changed, 18 insertions(+), 18 deletions(-) (limited to 'arch') diff --git a/arch/sparc64/kernel/time.c b/arch/sparc64/kernel/time.c index c60785c046b..71b4e380769 100644 --- a/arch/sparc64/kernel/time.c +++ b/arch/sparc64/kernel/time.c @@ -48,7 +48,7 @@ DEFINE_SPINLOCK(mostek_lock); DEFINE_SPINLOCK(rtc_lock); -void * __iomem mstk48t02_regs = 0UL; +void __iomem *mstk48t02_regs = NULL; #ifdef CONFIG_PCI unsigned long ds1287_regs = 0UL; #endif @@ -59,8 +59,8 @@ u64 jiffies_64 = INITIAL_JIFFIES; EXPORT_SYMBOL(jiffies_64); -static void * __iomem mstk48t08_regs; -static void * __iomem mstk48t59_regs; +static void __iomem *mstk48t08_regs; +static void __iomem *mstk48t59_regs; static int set_rtc_mmss(unsigned long); @@ -520,7 +520,7 @@ void timer_tick_interrupt(struct pt_regs *regs) /* Kick start a stopped clock (procedure from the Sun NVRAM/hostid FAQ). */ static void __init kick_start_clock(void) { - void * __iomem regs = mstk48t02_regs; + void __iomem *regs = mstk48t02_regs; u8 sec, tmp; int i, count; @@ -604,7 +604,7 @@ static void __init kick_start_clock(void) /* Return nonzero if the clock chip battery is low. */ static int __init has_low_battery(void) { - void * __iomem regs = mstk48t02_regs; + void __iomem *regs = mstk48t02_regs; u8 data1, data2; spin_lock_irq(&mostek_lock); @@ -623,7 +623,7 @@ static int __init has_low_battery(void) static void __init set_system_time(void) { unsigned int year, mon, day, hour, min, sec; - void * __iomem mregs = mstk48t02_regs; + void __iomem *mregs = mstk48t02_regs; #ifdef CONFIG_PCI unsigned long dregs = ds1287_regs; #else @@ -843,7 +843,7 @@ void __init clock_probe(void) !strcmp(model, "m5823")) { ds1287_regs = edev->resource[0].start; } else { - mstk48t59_regs = (void * __iomem) + mstk48t59_regs = (void __iomem *) edev->resource[0].start; mstk48t02_regs = mstk48t59_regs + MOSTEK_48T59_48T02; } @@ -866,7 +866,7 @@ try_isa_clock: !strcmp(model, "m5823")) { ds1287_regs = isadev->resource.start; } else { - mstk48t59_regs = (void * __iomem) + mstk48t59_regs = (void __iomem *) isadev->resource.start; mstk48t02_regs = mstk48t59_regs + MOSTEK_48T59_48T02; } @@ -895,16 +895,16 @@ try_isa_clock: } if(model[5] == '0' && model[6] == '2') { - mstk48t02_regs = (void * __iomem) + mstk48t02_regs = (void __iomem *) (((u64)clk_reg[0].phys_addr) | (((u64)clk_reg[0].which_io)<<32UL)); } else if(model[5] == '0' && model[6] == '8') { - mstk48t08_regs = (void * __iomem) + mstk48t08_regs = (void __iomem *) (((u64)clk_reg[0].phys_addr) | (((u64)clk_reg[0].which_io)<<32UL)); mstk48t02_regs = mstk48t08_regs + MOSTEK_48T08_48T02; } else { - mstk48t59_regs = (void * __iomem) + mstk48t59_regs = (void __iomem *) (((u64)clk_reg[0].phys_addr) | (((u64)clk_reg[0].which_io)<<32UL)); mstk48t02_regs = mstk48t59_regs + MOSTEK_48T59_48T02; @@ -1092,7 +1092,7 @@ unsigned long long sched_clock(void) static int set_rtc_mmss(unsigned long nowtime) { int real_seconds, real_minutes, chip_minutes; - void * __iomem mregs = mstk48t02_regs; + void __iomem *mregs = mstk48t02_regs; #ifdef CONFIG_PCI unsigned long dregs = ds1287_regs; #else diff --git a/drivers/sbus/char/rtc.c b/drivers/sbus/char/rtc.c index 49d1cd99d5a..9b988baf0b5 100644 --- a/drivers/sbus/char/rtc.c +++ b/drivers/sbus/char/rtc.c @@ -67,7 +67,7 @@ struct rtc_time_generic { /* Retrieve the current date and time from the real time clock. */ static void get_rtc_time(struct rtc_time *t) { - void * __iomem regs = mstk48t02_regs; + void __iomem *regs = mstk48t02_regs; u8 tmp; spin_lock_irq(&mostek_lock); @@ -94,7 +94,7 @@ static void get_rtc_time(struct rtc_time *t) /* Set the current date and time inthe real time clock. */ void set_rtc_time(struct rtc_time *t) { - void * __iomem regs = mstk48t02_regs; + void __iomem *regs = mstk48t02_regs; u8 tmp; spin_lock_irq(&mostek_lock); @@ -250,7 +250,7 @@ static int __init rtc_sun_init(void) /* It is possible we are being driven by some other RTC chip * and thus another RTC driver is handling things. */ - if (mstk48t02_regs == 0) + if (!mstk48t02_regs) return -ENODEV; error = misc_register(&rtc_dev); diff --git a/include/asm-sparc64/mostek.h b/include/asm-sparc64/mostek.h index 1f9b1356a48..09b5aba6678 100644 --- a/include/asm-sparc64/mostek.h +++ b/include/asm-sparc64/mostek.h @@ -38,7 +38,7 @@ * * We now deal with physical addresses for I/O to the chip. -DaveM */ -static __inline__ u8 mostek_read(void * __iomem addr) +static __inline__ u8 mostek_read(void __iomem *addr) { u8 ret; @@ -48,7 +48,7 @@ static __inline__ u8 mostek_read(void * __iomem addr) return ret; } -static __inline__ void mostek_write(void * __iomem addr, u8 val) +static __inline__ void mostek_write(void __iomem *addr, u8 val) { __asm__ __volatile__("stba %0, [%1] %2" : /* no outputs */ @@ -67,7 +67,7 @@ static __inline__ void mostek_write(void * __iomem addr, u8 val) #define MOSTEK_YEAR 0x07ffUL extern spinlock_t mostek_lock; -extern void *__iomem mstk48t02_regs; +extern void __iomem *mstk48t02_regs; /* Control register values. */ #define MSTK_CREG_WRITE 0x80 /* Must set this before placing values. */ -- cgit v1.2.3-70-g09d2 From 28a6815979b4eff29956549d62f72582a81f041e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Apr 2005 14:58:08 -0700 Subject: [PATCH] ppc trivial iomem annotations: pmac_time.c Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/platforms/pmac_time.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/ppc/platforms/pmac_time.c b/arch/ppc/platforms/pmac_time.c index 09636546f44..de60ccc7db9 100644 --- a/arch/ppc/platforms/pmac_time.c +++ b/arch/ppc/platforms/pmac_time.c @@ -165,7 +165,7 @@ int __init via_calibrate_decr(void) { struct device_node *vias; - volatile unsigned char *via; + volatile unsigned char __iomem *via; int count = VIA_TIMER_FREQ_6 / 100; unsigned int dstart, dend; @@ -176,8 +176,7 @@ via_calibrate_decr(void) vias = find_devices("via"); if (vias == 0 || vias->n_addrs == 0) return 0; - via = (volatile unsigned char *) - ioremap(vias->addrs[0].address, vias->addrs[0].size); + via = ioremap(vias->addrs[0].address, vias->addrs[0].size); /* set timer 1 for continuous interrupts */ out_8(&via[ACR], (via[ACR] & ~T1MODE) | T1MODE_CONT); @@ -202,7 +201,7 @@ via_calibrate_decr(void) printk(KERN_INFO "via_calibrate_decr: ticks per jiffy = %u (%u ticks)\n", tb_ticks_per_jiffy, dstart - dend); - iounmap((void*)via); + iounmap(via); return 1; } -- cgit v1.2.3-70-g09d2 From e2178f199b341cc3e3ac86be50a54b0d8d8a132f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Apr 2005 14:58:08 -0700 Subject: [PATCH] ppc trivial iomem annotations: chrp Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/platforms/chrp_pci.c | 7 ++++--- arch/ppc/platforms/chrp_setup.c | 2 +- include/asm-ppc/hydra.h | 2 +- 3 files changed, 6 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/ppc/platforms/chrp_pci.c b/arch/ppc/platforms/chrp_pci.c index 5bb6492ecf8..7d0ee308f66 100644 --- a/arch/ppc/platforms/chrp_pci.c +++ b/arch/ppc/platforms/chrp_pci.c @@ -129,7 +129,7 @@ static struct pci_ops rtas_pci_ops = rtas_write_config }; -volatile struct Hydra *Hydra = NULL; +volatile struct Hydra __iomem *Hydra = NULL; int __init hydra_init(void) @@ -175,13 +175,14 @@ chrp_pcibios_fixup(void) static void __init setup_python(struct pci_controller *hose, struct device_node *dev) { - u32 *reg, val; + u32 __iomem *reg; + u32 val; unsigned long addr = dev->addrs[0].address; setup_indirect_pci(hose, addr + 0xf8000, addr + 0xf8010); /* Clear the magic go-slow bit */ - reg = (u32 *) ioremap(dev->addrs[0].address + 0xf6000, 0x40); + reg = ioremap(dev->addrs[0].address + 0xf6000, 0x40); val = in_be32(®[12]); if (val & PRG_CL_RESET_VALID) { out_be32(®[12], val & ~PRG_CL_RESET_VALID); diff --git a/arch/ppc/platforms/chrp_setup.c b/arch/ppc/platforms/chrp_setup.c index f23c4f32076..57f29ab29bd 100644 --- a/arch/ppc/platforms/chrp_setup.c +++ b/arch/ppc/platforms/chrp_setup.c @@ -356,7 +356,7 @@ static void __init chrp_find_openpic(void) struct device_node *np; int len, i; unsigned int *iranges; - void *isu; + void __iomem *isu; np = find_type_devices("open-pic"); if (np == NULL || np->n_addrs == 0) diff --git a/include/asm-ppc/hydra.h b/include/asm-ppc/hydra.h index 1134431431d..833a8aff2a8 100644 --- a/include/asm-ppc/hydra.h +++ b/include/asm-ppc/hydra.h @@ -51,7 +51,7 @@ struct Hydra { char OpenPIC[0x40000]; }; -extern volatile struct Hydra *Hydra; +extern volatile struct Hydra __iomem *Hydra; /* -- cgit v1.2.3-70-g09d2 From 8e293ada7d6aaee43dd56a8077b83577dd108667 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Apr 2005 14:58:08 -0700 Subject: [PATCH] ppc trivial iomem annotations: pmac_smp.c Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/platforms/pmac_smp.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/ppc/platforms/pmac_smp.c b/arch/ppc/platforms/pmac_smp.c index 731841f9a5b..8e049dab4e6 100644 --- a/arch/ppc/platforms/pmac_smp.c +++ b/arch/ppc/platforms/pmac_smp.c @@ -91,11 +91,11 @@ extern void __secondary_start_psurge3(void); /* Temporary horrible hack */ #define PSURGE_QUAD_BIC(r, v) (PSURGE_QUAD_OUT((r), PSURGE_QUAD_IN(r) & ~(v))) /* virtual addresses for the above */ -static volatile u8 *hhead_base; -static volatile u8 *quad_base; -static volatile u32 *psurge_pri_intr; -static volatile u8 *psurge_sec_intr; -static volatile u32 *psurge_start; +static volatile u8 __iomem *hhead_base; +static volatile u8 __iomem *quad_base; +static volatile u32 __iomem *psurge_pri_intr; +static volatile u8 __iomem *psurge_sec_intr; +static volatile u32 __iomem *psurge_start; /* values for psurge_type */ #define PSURGE_NONE -1 @@ -322,10 +322,10 @@ static int __init smp_psurge_probe(void) /* All released cards using this HW design have 4 CPUs */ ncpus = 4; } else { - iounmap((void *) quad_base); + iounmap(quad_base); if ((in_8(hhead_base + HHEAD_CONFIG) & 0x02) == 0) { /* not a dual-cpu card */ - iounmap((void *) hhead_base); + iounmap(hhead_base); psurge_type = PSURGE_NONE; return 1; } -- cgit v1.2.3-70-g09d2 From 77a6a22f697bbc59cd434ff999dc34162e240f16 Mon Sep 17 00:00:00 2001 From: Tom 'spot' Callaway Date: Sun, 24 Apr 2005 20:41:45 -0700 Subject: [SPARC]: Missing sparc32 ksyms This patch adds some missing sparc32 ksyms that are needed. Specifically, ___rw_read_enter, ___rw_read_exit, ___rw_write_enter, and sys_close. Signed-off-by: Tom 'spot' Callaway Signed-off-by: David S. Miller --- arch/sparc/kernel/sparc_ksyms.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/sparc/kernel/sparc_ksyms.c b/arch/sparc/kernel/sparc_ksyms.c index f91b0e8d0dc..bfc04ca560d 100644 --- a/arch/sparc/kernel/sparc_ksyms.c +++ b/arch/sparc/kernel/sparc_ksyms.c @@ -20,6 +20,7 @@ #include #include #include +#include #ifdef CONFIG_PCI #include #endif @@ -89,6 +90,9 @@ extern void ___atomic24_sub(void); extern void ___set_bit(void); extern void ___clear_bit(void); extern void ___change_bit(void); +extern void ___rw_read_enter(void); +extern void ___rw_read_exit(void); +extern void ___rw_write_enter(void); /* Alias functions whose names begin with "." and export the aliases. * The module references will be fixed up by module_frob_arch_sections. @@ -121,9 +125,9 @@ EXPORT_SYMBOL(_do_write_unlock); #endif #else // XXX find what uses (or used) these. -// EXPORT_SYMBOL_PRIVATE(_rw_read_enter); -// EXPORT_SYMBOL_PRIVATE(_rw_read_exit); -// EXPORT_SYMBOL_PRIVATE(_rw_write_enter); +EXPORT_SYMBOL(___rw_read_enter); +EXPORT_SYMBOL(___rw_read_exit); +EXPORT_SYMBOL(___rw_write_enter); #endif /* semaphores */ EXPORT_SYMBOL(__up); @@ -332,3 +336,6 @@ EXPORT_SYMBOL(do_BUG); /* Sun Power Management Idle Handler */ EXPORT_SYMBOL(pm_idle); + +/* Binfmt_misc needs this */ +EXPORT_SYMBOL(sys_close); -- cgit v1.2.3-70-g09d2 From 962bd5604bf8a4d31d9204e7daf260ea93084bc0 Mon Sep 17 00:00:00 2001 From: Tom 'spot' Callaway Date: Sun, 24 Apr 2005 20:45:06 -0700 Subject: [SPARC]: More sparc32 ksyms cleanups The sparc32 ksyms is missing a few more symbols, these are primarily related to SMP, and will be needed as SMP gets beaten back into functionality. Specifically, add __cpu_data (PER_CPU), cpu_online_map, and phys_cpu_present_map. This patch assumes that the earlier "linux-2.6.11-sparc-fixksyms.patch" is applied, otherwise, it will apply with fuzz. Signed-off-by: Tom 'spot' Callaway Signed-off-by: David S. Miller --- arch/sparc/kernel/sparc_ksyms.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'arch') diff --git a/arch/sparc/kernel/sparc_ksyms.c b/arch/sparc/kernel/sparc_ksyms.c index bfc04ca560d..1bd430d0ca0 100644 --- a/arch/sparc/kernel/sparc_ksyms.c +++ b/arch/sparc/kernel/sparc_ksyms.c @@ -148,6 +148,9 @@ EXPORT_SYMBOL(___set_bit); EXPORT_SYMBOL(___clear_bit); EXPORT_SYMBOL(___change_bit); +/* Per-CPU information table */ +EXPORT_PER_CPU_SYMBOL(__cpu_data); + #ifdef CONFIG_SMP /* IRQ implementation. */ EXPORT_SYMBOL(synchronize_irq); @@ -155,6 +158,10 @@ EXPORT_SYMBOL(synchronize_irq); /* Misc SMP information */ EXPORT_SYMBOL(__cpu_number_map); EXPORT_SYMBOL(__cpu_logical_map); + +/* CPU online map and active count. */ +EXPORT_SYMBOL(cpu_online_map); +EXPORT_SYMBOL(phys_cpu_present_map); #endif EXPORT_SYMBOL(__udelay); -- cgit v1.2.3-70-g09d2 From 24dc6ead53f8fcae4b1908c4ea3fea75ee844a6d Mon Sep 17 00:00:00 2001 From: Tom 'spot' Callaway Date: Sun, 24 Apr 2005 20:46:49 -0700 Subject: [SPARC]: dump_stack for sparc Bob Breuer wrote a patch to add dump_stack for sparc. Supposedly, this was applied, but it doesn't exist in 2.6.11. This is the same patch, rediffed against 2.6.11. Signed-off-by: Tom 'spot' Callaway Signed-off-by: David S. Miller --- arch/sparc/kernel/process.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'arch') diff --git a/arch/sparc/kernel/process.c b/arch/sparc/kernel/process.c index 143fe2f3c1c..066e253f9c1 100644 --- a/arch/sparc/kernel/process.c +++ b/arch/sparc/kernel/process.c @@ -333,6 +333,17 @@ void show_stack(struct task_struct *tsk, unsigned long *_ksp) printk("\n"); } +void dump_stack(void) +{ + unsigned long *ksp; + + __asm__ __volatile__("mov %%fp, %0" + : "=r" (ksp)); + show_stack(current, ksp); +} + +EXPORT_SYMBOL(dump_stack); + /* * Note: sparc64 has a pretty intricated thread_saved_pc, check it out. */ -- cgit v1.2.3-70-g09d2 From 9090e001f28de4bccae232f1b78e78fad5132ecb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 25 Apr 2005 07:55:57 -0700 Subject: [PATCH] ppc user annotations: sigcontext sigcontext.regs is a userland pointer Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/kernel/signal.c | 2 +- include/asm-ppc/sigcontext.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/ppc/kernel/signal.c b/arch/ppc/kernel/signal.c index 645eae19805..d500f5ebc29 100644 --- a/arch/ppc/kernel/signal.c +++ b/arch/ppc/kernel/signal.c @@ -632,7 +632,7 @@ handle_signal(unsigned long sig, struct k_sigaction *ka, if (__put_user((unsigned long) ka->sa.sa_handler, &sc->handler) || __put_user(oldset->sig[0], &sc->oldmask) || __put_user(oldset->sig[1], &sc->_unused[3]) - || __put_user((struct pt_regs *)frame, &sc->regs) + || __put_user((struct pt_regs __user *)frame, &sc->regs) || __put_user(sig, &sc->signal)) goto badframe; diff --git a/include/asm-ppc/sigcontext.h b/include/asm-ppc/sigcontext.h index fc5e358c65f..f82dcccdee1 100644 --- a/include/asm-ppc/sigcontext.h +++ b/include/asm-ppc/sigcontext.h @@ -9,7 +9,7 @@ struct sigcontext { int signal; unsigned long handler; unsigned long oldmask; - struct pt_regs *regs; + struct pt_regs __user *regs; }; #endif -- cgit v1.2.3-70-g09d2 From 92a11f9e7cb876e6cf9080e4a1642142cb9221b2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 25 Apr 2005 07:55:57 -0700 Subject: [PATCH] ppc iomem annotations: ->io_base_virt * ->io_base_virt in struct pci_controller is iomem pointer. Marked as such. Most of the places that used it are already annotated to expect iomem. * places that did gratitious (and wrong) casts a-la isa_io_base = (unsigned long)ioremap(...); hose->io_base_virt = (void *)isa_io_base; turned into hose->io_base_virt = ioremap(...); isa_io_base = (unsigned long)hose->io_base_virt; * pci_bus_io_base() annotated as returning iomem pointer. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/kernel/pci.c | 2 +- arch/ppc/platforms/4xx/ebony.c | 5 ++--- arch/ppc/platforms/4xx/luan.c | 5 ++--- arch/ppc/platforms/4xx/ocotea.c | 5 ++--- arch/ppc/syslib/m8260_pci.c | 5 ++--- arch/ppc/syslib/mpc52xx_pci.c | 8 +++----- arch/ppc/syslib/ppc85xx_setup.c | 12 +++++------- include/asm-ppc/pci-bridge.h | 4 ++-- 8 files changed, 19 insertions(+), 27 deletions(-) (limited to 'arch') diff --git a/arch/ppc/kernel/pci.c b/arch/ppc/kernel/pci.c index 98f94b60204..47a15306823 100644 --- a/arch/ppc/kernel/pci.c +++ b/arch/ppc/kernel/pci.c @@ -1432,7 +1432,7 @@ pci_bus_to_hose(int bus) return NULL; } -void* +void __iomem * pci_bus_io_base(unsigned int bus) { struct pci_controller *hose; diff --git a/arch/ppc/platforms/4xx/ebony.c b/arch/ppc/platforms/4xx/ebony.c index f63bca83e75..55120c317f7 100644 --- a/arch/ppc/platforms/4xx/ebony.c +++ b/arch/ppc/platforms/4xx/ebony.c @@ -210,9 +210,8 @@ ebony_setup_hose(void) hose->io_space.end = EBONY_PCI_UPPER_IO; hose->mem_space.start = EBONY_PCI_LOWER_MEM; hose->mem_space.end = EBONY_PCI_UPPER_MEM; - isa_io_base = - (unsigned long)ioremap64(EBONY_PCI_IO_BASE, EBONY_PCI_IO_SIZE); - hose->io_base_virt = (void *)isa_io_base; + hose->io_base_virt = ioremap64(EBONY_PCI_IO_BASE, EBONY_PCI_IO_SIZE); + isa_io_base = (unsigned long)hose->io_base_virt; setup_indirect_pci(hose, EBONY_PCI_CFGA_PLB32, diff --git a/arch/ppc/platforms/4xx/luan.c b/arch/ppc/platforms/4xx/luan.c index 1df2339f1f6..95359f748e7 100644 --- a/arch/ppc/platforms/4xx/luan.c +++ b/arch/ppc/platforms/4xx/luan.c @@ -223,9 +223,8 @@ luan_setup_hose(struct pci_controller *hose, hose->io_space.end = LUAN_PCIX_UPPER_IO; hose->mem_space.start = lower_mem; hose->mem_space.end = upper_mem; - isa_io_base = - (unsigned long)ioremap64(pcix_io_base, PCIX_IO_SIZE); - hose->io_base_virt = (void *)isa_io_base; + hose->io_base_virt = ioremap64(pcix_io_base, PCIX_IO_SIZE); + isa_io_base = (unsigned long) hose->io_base_virt; setup_indirect_pci(hose, cfga, cfgd); hose->set_cfg_type = 1; diff --git a/arch/ppc/platforms/4xx/ocotea.c b/arch/ppc/platforms/4xx/ocotea.c index 28de707434f..5f82a6bc704 100644 --- a/arch/ppc/platforms/4xx/ocotea.c +++ b/arch/ppc/platforms/4xx/ocotea.c @@ -227,9 +227,8 @@ ocotea_setup_hose(void) hose->io_space.end = OCOTEA_PCI_UPPER_IO; hose->mem_space.start = OCOTEA_PCI_LOWER_MEM; hose->mem_space.end = OCOTEA_PCI_UPPER_MEM; - isa_io_base = - (unsigned long)ioremap64(OCOTEA_PCI_IO_BASE, OCOTEA_PCI_IO_SIZE); - hose->io_base_virt = (void *)isa_io_base; + hose->io_base_virt = ioremap64(OCOTEA_PCI_IO_BASE, OCOTEA_PCI_IO_SIZE); + isa_io_base = (unsigned long) hose->io_base_virt; setup_indirect_pci(hose, OCOTEA_PCI_CFGA_PLB32, diff --git a/arch/ppc/syslib/m8260_pci.c b/arch/ppc/syslib/m8260_pci.c index bd564fb35ab..057cc3f8ff3 100644 --- a/arch/ppc/syslib/m8260_pci.c +++ b/arch/ppc/syslib/m8260_pci.c @@ -171,10 +171,9 @@ void __init m8260_find_bridges(void) m8260_setup_pci(hose); hose->pci_mem_offset = MPC826x_PCI_MEM_OFFSET; - isa_io_base = - (unsigned long) ioremap(MPC826x_PCI_IO_BASE, + hose->io_base_virt = ioremap(MPC826x_PCI_IO_BASE, MPC826x_PCI_IO_SIZE); - hose->io_base_virt = (void *) isa_io_base; + isa_io_base = (unsigned long) hose->io_base_virt; /* setup resources */ pci_init_resource(&hose->mem_resources[0], diff --git a/arch/ppc/syslib/mpc52xx_pci.c b/arch/ppc/syslib/mpc52xx_pci.c index c723efd954a..59cf3e8bd1a 100644 --- a/arch/ppc/syslib/mpc52xx_pci.c +++ b/arch/ppc/syslib/mpc52xx_pci.c @@ -205,13 +205,11 @@ mpc52xx_find_bridges(void) hose->pci_mem_offset = MPC52xx_PCI_MEM_OFFSET; - isa_io_base = - (unsigned long) ioremap(MPC52xx_PCI_IO_BASE, - MPC52xx_PCI_IO_SIZE); - hose->io_base_virt = (void *) isa_io_base; + hose->io_base_virt = ioremap(MPC52xx_PCI_IO_BASE, MPC52xx_PCI_IO_SIZE); + isa_io_base = (unsigned long) hose->io_base_virt; hose->cfg_addr = &pci_regs->car; - hose->cfg_data = (void __iomem *) isa_io_base; + hose->cfg_data = hose->io_base_virt; /* Setup resources */ pci_init_resource(&hose->mem_resources[0], diff --git a/arch/ppc/syslib/ppc85xx_setup.c b/arch/ppc/syslib/ppc85xx_setup.c index 81f1968c326..152c3ef1312 100644 --- a/arch/ppc/syslib/ppc85xx_setup.c +++ b/arch/ppc/syslib/ppc85xx_setup.c @@ -280,16 +280,14 @@ mpc85xx_setup_hose(void) hose_a->io_space.end = MPC85XX_PCI1_UPPER_IO; hose_a->io_base_phys = MPC85XX_PCI1_IO_BASE; #ifdef CONFIG_85xx_PCI2 - isa_io_base = - (unsigned long) ioremap(MPC85XX_PCI1_IO_BASE, + hose_a->io_base_virt = ioremap(MPC85XX_PCI1_IO_BASE, MPC85XX_PCI1_IO_SIZE + MPC85XX_PCI2_IO_SIZE); #else - isa_io_base = - (unsigned long) ioremap(MPC85XX_PCI1_IO_BASE, + hose_a->io_base_virt = ioremap(MPC85XX_PCI1_IO_BASE, MPC85XX_PCI1_IO_SIZE); #endif - hose_a->io_base_virt = (void *) isa_io_base; + isa_io_base = (unsigned long)hose_a->io_base_virt; /* setup resources */ pci_init_resource(&hose_a->mem_resources[0], @@ -329,8 +327,8 @@ mpc85xx_setup_hose(void) hose_b->io_space.start = MPC85XX_PCI2_LOWER_IO; hose_b->io_space.end = MPC85XX_PCI2_UPPER_IO; hose_b->io_base_phys = MPC85XX_PCI2_IO_BASE; - hose_b->io_base_virt = (void *) isa_io_base + MPC85XX_PCI1_IO_SIZE; - + hose_b->io_base_virt = hose_a->io_base_virt + MPC85XX_PCI1_IO_SIZE; + /* setup resources */ pci_init_resource(&hose_b->mem_resources[0], MPC85XX_PCI2_LOWER_MEM, diff --git a/include/asm-ppc/pci-bridge.h b/include/asm-ppc/pci-bridge.h index 78e9be619f1..ffa423456c2 100644 --- a/include/asm-ppc/pci-bridge.h +++ b/include/asm-ppc/pci-bridge.h @@ -12,7 +12,7 @@ struct pci_controller; * pci_io_base returns the memory address at which you can access * the I/O space for PCI bus number `bus' (or NULL on error). */ -extern void *pci_bus_io_base(unsigned int bus); +extern void __iomem *pci_bus_io_base(unsigned int bus); extern unsigned long pci_bus_io_base_phys(unsigned int bus); extern unsigned long pci_bus_mem_base_phys(unsigned int bus); @@ -48,7 +48,7 @@ struct pci_controller { int last_busno; int bus_offset; - void *io_base_virt; + void __iomem *io_base_virt; unsigned long io_base_phys; /* Some machines (PReP) have a non 1:1 mapping of -- cgit v1.2.3-70-g09d2 From ba724a373a7184dce869e8adb79a0a03d91fbdd8 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 25 Apr 2005 07:55:58 -0700 Subject: [PATCH] ppc iomem annotations: ebony_setup_pcix() Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/platforms/4xx/ebony.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ppc/platforms/4xx/ebony.c b/arch/ppc/platforms/4xx/ebony.c index 55120c317f7..cd11734ef7c 100644 --- a/arch/ppc/platforms/4xx/ebony.c +++ b/arch/ppc/platforms/4xx/ebony.c @@ -149,7 +149,7 @@ ebony_map_irq(struct pci_dev *dev, unsigned char idsel, unsigned char pin) static void __init ebony_setup_pcix(void) { - void *pcix_reg_base; + void __iomem *pcix_reg_base; pcix_reg_base = ioremap64(PCIX0_REG_BASE, PCIX_REG_SIZE); -- cgit v1.2.3-70-g09d2 From d01daf72f0a238c5ccfba7bf82999e384927412e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 25 Apr 2005 07:55:58 -0700 Subject: [PATCH] ppc iomem annotations: pmac_low_i2c Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/platforms/pmac_low_i2c.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/ppc/platforms/pmac_low_i2c.c b/arch/ppc/platforms/pmac_low_i2c.c index d07579f2b8b..08583fce169 100644 --- a/arch/ppc/platforms/pmac_low_i2c.c +++ b/arch/ppc/platforms/pmac_low_i2c.c @@ -54,7 +54,7 @@ struct low_i2c_host int mode; /* Current mode */ int channel; /* Current channel */ int num_channels; /* Number of channels */ - unsigned long base; /* For keywest-i2c, base address */ + void __iomem * base; /* For keywest-i2c, base address */ int bsteps; /* And register stepping */ int speed; /* And speed */ }; @@ -154,14 +154,12 @@ static const char *__kw_state_names[] = { static inline u8 __kw_read_reg(struct low_i2c_host *host, reg_t reg) { - return in_8(((volatile u8 *)host->base) - + (((unsigned)reg) << host->bsteps)); + return in_8(host->base + (((unsigned)reg) << host->bsteps)); } static inline void __kw_write_reg(struct low_i2c_host *host, reg_t reg, u8 val) { - out_8(((volatile u8 *)host->base) - + (((unsigned)reg) << host->bsteps), val); + out_8(host->base + (((unsigned)reg) << host->bsteps), val); (void)__kw_read_reg(host, reg_subaddr); } @@ -370,7 +368,7 @@ static void keywest_low_i2c_add(struct device_node *np) break; } host->mode = pmac_low_i2c_mode_std; - host->base = (unsigned long)ioremap(np->addrs[0].address + aoffset, + host->base = ioremap(np->addrs[0].address + aoffset, np->addrs[0].size); host->func = keywest_low_i2c_func; } -- cgit v1.2.3-70-g09d2 From 91de1fff2d431fec1162333a1e4fad9a53cd5b9e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 25 Apr 2005 07:55:58 -0700 Subject: [PATCH] ppc sparse annotations: emulate_string_inst() replaced declaration of EA from u32 to unsigned long - this beast is used only to cast it to (userland) pointer and proper integer type for that is unsigned long. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/kernel/traps.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ppc/kernel/traps.c b/arch/ppc/kernel/traps.c index 002322a1f3c..361865c4bc8 100644 --- a/arch/ppc/kernel/traps.c +++ b/arch/ppc/kernel/traps.c @@ -403,7 +403,7 @@ static int emulate_string_inst(struct pt_regs *regs, u32 instword) u8 rA = (instword >> 16) & 0x1f; u8 NB_RB = (instword >> 11) & 0x1f; u32 num_bytes; - u32 EA; + unsigned long EA; int pos = 0; /* Early out if we are an invalid form of lswx */ -- cgit v1.2.3-70-g09d2 From 056de2fa12febe02597f971eb6ea8f2cc9c9b06e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 25 Apr 2005 07:55:59 -0700 Subject: [PATCH] ppc user annotations: debug_setconetext(2) 3rd argument of sys_debug_setcontext() is also a userland pointer. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc/kernel/signal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ppc/kernel/signal.c b/arch/ppc/kernel/signal.c index d500f5ebc29..7c8437da09d 100644 --- a/arch/ppc/kernel/signal.c +++ b/arch/ppc/kernel/signal.c @@ -511,7 +511,7 @@ int sys_rt_sigreturn(int r3, int r4, int r5, int r6, int r7, int r8, } int sys_debug_setcontext(struct ucontext __user *ctx, - int ndbg, struct sig_dbg_op *dbg, + int ndbg, struct sig_dbg_op __user *dbg, int r6, int r7, int r8, struct pt_regs *regs) { -- cgit v1.2.3-70-g09d2 From e955d82543fea76b02aa243b182e782f71bda82c Mon Sep 17 00:00:00 2001 From: Mark Maule Date: Mon, 25 Apr 2005 11:26:03 -0700 Subject: [IA64-SGI] sn2-pci-dma-abstraction.patch Provide an abstraction of the altix pci dma runtime layer so that multiple pci-based bridges can be supported. Signed-off-by: Mark Maule Signed-off-by: Tony Luck --- arch/ia64/sn/include/pci/pcibr_provider.h | 6 +- arch/ia64/sn/include/pci/pcibus_provider_defs.h | 12 ++- arch/ia64/sn/include/pci/pcidev.h | 4 + arch/ia64/sn/kernel/io_init.c | 72 +++++++++++++++-- arch/ia64/sn/pci/pci_dma.c | 35 ++++---- arch/ia64/sn/pci/pcibr/pcibr_dma.c | 103 ++++++++++++------------ arch/ia64/sn/pci/pcibr/pcibr_provider.c | 20 +++++ 7 files changed, 173 insertions(+), 79 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/include/pci/pcibr_provider.h b/arch/ia64/sn/include/pci/pcibr_provider.h index b1f05ffec70..1cd291d8bad 100644 --- a/arch/ia64/sn/include/pci/pcibr_provider.h +++ b/arch/ia64/sn/include/pci/pcibr_provider.h @@ -123,9 +123,11 @@ pcibr_lock(struct pcibus_info *pcibus_info) } #define pcibr_unlock(pcibus_info, flag) spin_unlock_irqrestore(&pcibus_info->pbi_lock, flag) +extern int pcibr_init_provider(void); extern void *pcibr_bus_fixup(struct pcibus_bussoft *); -extern uint64_t pcibr_dma_map(struct pcidev_info *, unsigned long, size_t, unsigned int); -extern void pcibr_dma_unmap(struct pcidev_info *, dma_addr_t, int); +extern dma_addr_t pcibr_dma_map(struct pci_dev *, unsigned long, size_t); +extern dma_addr_t pcibr_dma_map_consistent(struct pci_dev *, unsigned long, size_t); +extern void pcibr_dma_unmap(struct pci_dev *, dma_addr_t, int); /* * prototypes for the bridge asic register access routines in pcibr_reg.c diff --git a/arch/ia64/sn/include/pci/pcibus_provider_defs.h b/arch/ia64/sn/include/pci/pcibus_provider_defs.h index 07065615bbe..f546b4ece33 100644 --- a/arch/ia64/sn/include/pci/pcibus_provider_defs.h +++ b/arch/ia64/sn/include/pci/pcibus_provider_defs.h @@ -18,6 +18,8 @@ #define PCIIO_ASIC_TYPE_PIC 2 #define PCIIO_ASIC_TYPE_TIOCP 3 +#define PCIIO_ASIC_MAX_TYPES 4 + /* * Common pciio bus provider data. There should be one of these as the * first field in any pciio based provider soft structure (e.g. pcibr_soft @@ -35,9 +37,15 @@ struct pcibus_bussoft { }; /* - * DMA mapping flags + * SN pci bus indirection */ -#define SN_PCIDMA_CONSISTENT 0x0001 +struct sn_pcibus_provider { + dma_addr_t (*dma_map)(struct pci_dev *, unsigned long, size_t); + dma_addr_t (*dma_map_consistent)(struct pci_dev *, unsigned long, size_t); + void (*dma_unmap)(struct pci_dev *, dma_addr_t, int); + void * (*bus_fixup)(struct pcibus_bussoft *); +}; +extern struct sn_pcibus_provider *sn_pci_provider[]; #endif /* _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H */ diff --git a/arch/ia64/sn/include/pci/pcidev.h b/arch/ia64/sn/include/pci/pcidev.h index 81eb95d3bf4..ed4031d8081 100644 --- a/arch/ia64/sn/include/pci/pcidev.h +++ b/arch/ia64/sn/include/pci/pcidev.h @@ -32,6 +32,9 @@ extern struct sn_irq_info **sn_irq; #define SN_PCIDEV_BUSSOFT(pci_dev) \ (SN_PCIDEV_INFO(pci_dev)->pdi_host_pcidev_info->pdi_pcibus_info) +#define SN_PCIDEV_BUSPROVIDER(pci_dev) \ + (SN_PCIDEV_INFO(pci_dev)->pdi_provider) + #define PCIIO_BUS_NONE 255 /* bus 255 reserved */ #define PCIIO_SLOT_NONE 255 #define PCIIO_FUNC_NONE 255 @@ -46,6 +49,7 @@ struct pcidev_info { struct pci_dev *pdi_linux_pcidev; /* Kernel pci_dev */ struct sn_irq_info *pdi_sn_irq_info; + struct sn_pcibus_provider *pdi_provider; /* sn pci ops */ }; extern void sn_irq_fixup(struct pci_dev *pci_dev, diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c index 001880812b7..3e5e4a90130 100644 --- a/arch/ia64/sn/kernel/io_init.c +++ b/arch/ia64/sn/kernel/io_init.c @@ -34,6 +34,37 @@ struct brick { int sn_ioif_inited = 0; /* SN I/O infrastructure initialized? */ +struct sn_pcibus_provider *sn_pci_provider[PCIIO_ASIC_MAX_TYPES]; /* indexed by asic type */ + +/* + * Hooks and struct for unsupported pci providers + */ + +static dma_addr_t +sn_default_pci_map(struct pci_dev *pdev, unsigned long paddr, size_t size) +{ + return 0; +} + +static void +sn_default_pci_unmap(struct pci_dev *pdev, dma_addr_t addr, int direction) +{ + return; +} + +static void * +sn_default_pci_bus_fixup(struct pcibus_bussoft *soft) +{ + return NULL; +} + +static struct sn_pcibus_provider sn_pci_default_provider = { + .dma_map = sn_default_pci_map, + .dma_map_consistent = sn_default_pci_map, + .dma_unmap = sn_default_pci_unmap, + .bus_fixup = sn_default_pci_bus_fixup, +}; + /* * Retrieve the DMA Flush List given nasid. This list is needed * to implement the WAR - Flush DMA data on PIO Reads. @@ -201,6 +232,7 @@ static void sn_pci_fixup_slot(struct pci_dev *dev) struct sn_irq_info *sn_irq_info; struct pci_dev *host_pci_dev; int status = 0; + struct pcibus_bussoft *bs; dev->sysdata = kmalloc(sizeof(struct pcidev_info), GFP_KERNEL); if (SN_PCIDEV_INFO(dev) <= 0) @@ -241,6 +273,7 @@ static void sn_pci_fixup_slot(struct pci_dev *dev) } /* set up host bus linkages */ + bs = SN_PCIBUS_BUSSOFT(dev->bus); host_pci_dev = pci_find_slot(SN_PCIDEV_INFO(dev)->pdi_slot_host_handle >> 32, SN_PCIDEV_INFO(dev)-> @@ -248,10 +281,16 @@ static void sn_pci_fixup_slot(struct pci_dev *dev) SN_PCIDEV_INFO(dev)->pdi_host_pcidev_info = SN_PCIDEV_INFO(host_pci_dev); SN_PCIDEV_INFO(dev)->pdi_linux_pcidev = dev; - SN_PCIDEV_INFO(dev)->pdi_pcibus_info = SN_PCIBUS_BUSSOFT(dev->bus); + SN_PCIDEV_INFO(dev)->pdi_pcibus_info = bs; + + if (bs && bs->bs_asic_type < PCIIO_ASIC_MAX_TYPES) { + SN_PCIDEV_BUSPROVIDER(dev) = sn_pci_provider[bs->bs_asic_type]; + } else { + SN_PCIDEV_BUSPROVIDER(dev) = &sn_pci_default_provider; + } /* Only set up IRQ stuff if this device has a host bus context */ - if (SN_PCIDEV_BUSSOFT(dev) && sn_irq_info->irq_irq) { + if (bs && sn_irq_info->irq_irq) { SN_PCIDEV_INFO(dev)->pdi_sn_irq_info = sn_irq_info; dev->irq = SN_PCIDEV_INFO(dev)->pdi_sn_irq_info->irq_irq; sn_irq_fixup(dev, sn_irq_info); @@ -271,6 +310,7 @@ static void sn_pci_controller_fixup(int segment, int busnum) struct pcibus_bussoft *prom_bussoft_ptr; struct hubdev_info *hubdev_info; void *provider_soft; + struct sn_pcibus_provider *provider; status = sal_get_pcibus_info((u64) segment, (u64) busnum, @@ -291,16 +331,22 @@ static void sn_pci_controller_fixup(int segment, int busnum) /* * Per-provider fixup. Copies the contents from prom to local * area and links SN_PCIBUS_BUSSOFT(). - * - * Note: Provider is responsible for ensuring that prom_bussoft_ptr - * represents an asic-type that it can handle. */ - if (prom_bussoft_ptr->bs_asic_type == PCIIO_ASIC_TYPE_PPB) { - return; /* no further fixup necessary */ + if (prom_bussoft_ptr->bs_asic_type >= PCIIO_ASIC_MAX_TYPES) { + return; /* unsupported asic type */ + } + + provider = sn_pci_provider[prom_bussoft_ptr->bs_asic_type]; + if (provider == NULL) { + return; /* no provider registerd for this asic */ + } + + provider_soft = NULL; + if (provider->bus_fixup) { + provider_soft = (*provider->bus_fixup) (prom_bussoft_ptr); } - provider_soft = pcibr_bus_fixup(prom_bussoft_ptr); if (provider_soft == NULL) { return; /* fixup failed or not applicable */ } @@ -338,6 +384,16 @@ static int __init sn_pci_init(void) if (!ia64_platform_is("sn2") || IS_RUNNING_ON_SIMULATOR()) return 0; + /* + * prime sn_pci_provider[]. Individial provider init routines will + * override their respective default entries. + */ + + for (i = 0; i < PCIIO_ASIC_MAX_TYPES; i++) + sn_pci_provider[i] = &sn_pci_default_provider; + + pcibr_init_provider(); + /* * This is needed to avoid bounce limit checks in the blk layer */ diff --git a/arch/ia64/sn/pci/pci_dma.c b/arch/ia64/sn/pci/pci_dma.c index f680824f819..c2b92b94c56 100644 --- a/arch/ia64/sn/pci/pci_dma.c +++ b/arch/ia64/sn/pci/pci_dma.c @@ -14,7 +14,6 @@ #include #include "pci/pcibus_provider_defs.h" #include "pci/pcidev.h" -#include "pci/pcibr_provider.h" #define SG_ENT_VIRT_ADDRESS(sg) (page_address((sg)->page) + (sg)->offset) #define SG_ENT_PHYS_ADDRESS(SG) virt_to_phys(SG_ENT_VIRT_ADDRESS(SG)) @@ -79,7 +78,8 @@ void *sn_dma_alloc_coherent(struct device *dev, size_t size, { void *cpuaddr; unsigned long phys_addr; - struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(to_pci_dev(dev)); + struct pci_dev *pdev = to_pci_dev(dev); + struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); BUG_ON(dev->bus != &pci_bus_type); @@ -102,8 +102,7 @@ void *sn_dma_alloc_coherent(struct device *dev, size_t size, * resources. */ - *dma_handle = pcibr_dma_map(pcidev_info, phys_addr, size, - SN_PCIDMA_CONSISTENT); + *dma_handle = provider->dma_map_consistent(pdev, phys_addr, size); if (!*dma_handle) { printk(KERN_ERR "%s: out of ATEs\n", __FUNCTION__); free_pages((unsigned long)cpuaddr, get_order(size)); @@ -127,11 +126,12 @@ EXPORT_SYMBOL(sn_dma_alloc_coherent); void sn_dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, dma_addr_t dma_handle) { - struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(to_pci_dev(dev)); + struct pci_dev *pdev = to_pci_dev(dev); + struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); BUG_ON(dev->bus != &pci_bus_type); - pcibr_dma_unmap(pcidev_info, dma_handle, 0); + provider->dma_unmap(pdev, dma_handle, 0); free_pages((unsigned long)cpu_addr, get_order(size)); } EXPORT_SYMBOL(sn_dma_free_coherent); @@ -159,12 +159,13 @@ dma_addr_t sn_dma_map_single(struct device *dev, void *cpu_addr, size_t size, { dma_addr_t dma_addr; unsigned long phys_addr; - struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(to_pci_dev(dev)); + struct pci_dev *pdev = to_pci_dev(dev); + struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); BUG_ON(dev->bus != &pci_bus_type); phys_addr = __pa(cpu_addr); - dma_addr = pcibr_dma_map(pcidev_info, phys_addr, size, 0); + dma_addr = provider->dma_map(pdev, phys_addr, size); if (!dma_addr) { printk(KERN_ERR "%s: out of ATEs\n", __FUNCTION__); return 0; @@ -187,10 +188,12 @@ EXPORT_SYMBOL(sn_dma_map_single); void sn_dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size, int direction) { - struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(to_pci_dev(dev)); + struct pci_dev *pdev = to_pci_dev(dev); + struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); BUG_ON(dev->bus != &pci_bus_type); - pcibr_dma_unmap(pcidev_info, dma_addr, direction); + + provider->dma_unmap(pdev, dma_addr, direction); } EXPORT_SYMBOL(sn_dma_unmap_single); @@ -207,12 +210,13 @@ void sn_dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nhwentries, int direction) { int i; - struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(to_pci_dev(dev)); + struct pci_dev *pdev = to_pci_dev(dev); + struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); BUG_ON(dev->bus != &pci_bus_type); for (i = 0; i < nhwentries; i++, sg++) { - pcibr_dma_unmap(pcidev_info, sg->dma_address, direction); + provider->dma_unmap(pdev, sg->dma_address, direction); sg->dma_address = (dma_addr_t) NULL; sg->dma_length = 0; } @@ -233,7 +237,8 @@ int sn_dma_map_sg(struct device *dev, struct scatterlist *sg, int nhwentries, { unsigned long phys_addr; struct scatterlist *saved_sg = sg; - struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(to_pci_dev(dev)); + struct pci_dev *pdev = to_pci_dev(dev); + struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); int i; BUG_ON(dev->bus != &pci_bus_type); @@ -243,8 +248,8 @@ int sn_dma_map_sg(struct device *dev, struct scatterlist *sg, int nhwentries, */ for (i = 0; i < nhwentries; i++, sg++) { phys_addr = SG_ENT_PHYS_ADDRESS(sg); - sg->dma_address = pcibr_dma_map(pcidev_info, phys_addr, - sg->length, 0); + sg->dma_address = provider->dma_map(pdev, + phys_addr, sg->length); if (!sg->dma_address) { printk(KERN_ERR "%s: out of ATEs\n", __FUNCTION__); diff --git a/arch/ia64/sn/pci/pcibr/pcibr_dma.c b/arch/ia64/sn/pci/pcibr/pcibr_dma.c index b1d66ac065c..3c305f46417 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_dma.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_dma.c @@ -40,7 +40,7 @@ extern int sn_ioif_inited; * we do not have to allocate entries in the PMU. */ -static uint64_t +static dma_addr_t pcibr_dmamap_ate32(struct pcidev_info *info, uint64_t paddr, size_t req_size, uint64_t flags) { @@ -109,7 +109,7 @@ pcibr_dmamap_ate32(struct pcidev_info *info, return pci_addr; } -static uint64_t +static dma_addr_t pcibr_dmatrans_direct64(struct pcidev_info * info, uint64_t paddr, uint64_t dma_attributes) { @@ -141,7 +141,7 @@ pcibr_dmatrans_direct64(struct pcidev_info * info, uint64_t paddr, } -static uint64_t +static dma_addr_t pcibr_dmatrans_direct32(struct pcidev_info * info, uint64_t paddr, size_t req_size, uint64_t flags) { @@ -180,11 +180,11 @@ pcibr_dmatrans_direct32(struct pcidev_info * info, * DMA mappings for Direct 64 and 32 do not have any DMA maps. */ void -pcibr_dma_unmap(struct pcidev_info *pcidev_info, dma_addr_t dma_handle, - int direction) +pcibr_dma_unmap(struct pci_dev *hwdev, dma_addr_t dma_handle, int direction) { - struct pcibus_info *pcibus_info = (struct pcibus_info *)pcidev_info-> - pdi_pcibus_info; + struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(hwdev); + struct pcibus_info *pcibus_info = + (struct pcibus_info *)pcidev_info->pdi_pcibus_info; if (IS_PCI32_MAPPED(dma_handle)) { int ate_index; @@ -316,64 +316,63 @@ void sn_dma_flush(uint64_t addr) } /* - * Wrapper DMA interface. Called from pci_dma.c routines. + * DMA interfaces. Called from pci_dma.c routines. */ -uint64_t -pcibr_dma_map(struct pcidev_info * pcidev_info, unsigned long phys_addr, - size_t size, unsigned int flags) +dma_addr_t +pcibr_dma_map(struct pci_dev * hwdev, unsigned long phys_addr, size_t size) { dma_addr_t dma_handle; - struct pci_dev *pcidev = pcidev_info->pdi_linux_pcidev; - - if (flags & SN_PCIDMA_CONSISTENT) { - /* sn_pci_alloc_consistent interfaces */ - if (pcidev->dev.coherent_dma_mask == ~0UL) { - dma_handle = - pcibr_dmatrans_direct64(pcidev_info, phys_addr, - PCI64_ATTR_BAR); - } else { - dma_handle = - (dma_addr_t) pcibr_dmamap_ate32(pcidev_info, - phys_addr, size, - PCI32_ATE_BAR); - } - } else { - /* map_sg/map_single interfaces */ + struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(hwdev); - /* SN cannot support DMA addresses smaller than 32 bits. */ - if (pcidev->dma_mask < 0x7fffffff) { - return 0; - } + /* SN cannot support DMA addresses smaller than 32 bits. */ + if (hwdev->dma_mask < 0x7fffffff) { + return 0; + } - if (pcidev->dma_mask == ~0UL) { + if (hwdev->dma_mask == ~0UL) { + /* + * Handle the most common case: 64 bit cards. This + * call should always succeed. + */ + + dma_handle = pcibr_dmatrans_direct64(pcidev_info, phys_addr, + PCI64_ATTR_PREF); + } else { + /* Handle 32-63 bit cards via direct mapping */ + dma_handle = pcibr_dmatrans_direct32(pcidev_info, phys_addr, + size, 0); + if (!dma_handle) { /* - * Handle the most common case: 64 bit cards. This - * call should always succeed. + * It is a 32 bit card and we cannot do direct mapping, + * so we use an ATE. */ - dma_handle = - pcibr_dmatrans_direct64(pcidev_info, phys_addr, - PCI64_ATTR_PREF); - } else { - /* Handle 32-63 bit cards via direct mapping */ - dma_handle = - pcibr_dmatrans_direct32(pcidev_info, phys_addr, - size, 0); - if (!dma_handle) { - /* - * It is a 32 bit card and we cannot do direct mapping, - * so we use an ATE. - */ - - dma_handle = - pcibr_dmamap_ate32(pcidev_info, phys_addr, - size, PCI32_ATE_PREF); - } + dma_handle = pcibr_dmamap_ate32(pcidev_info, phys_addr, + size, PCI32_ATE_PREF); } } return dma_handle; } +dma_addr_t +pcibr_dma_map_consistent(struct pci_dev * hwdev, unsigned long phys_addr, + size_t size) +{ + dma_addr_t dma_handle; + struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(hwdev); + + if (hwdev->dev.coherent_dma_mask == ~0UL) { + dma_handle = pcibr_dmatrans_direct64(pcidev_info, phys_addr, + PCI64_ATTR_BAR); + } else { + dma_handle = (dma_addr_t) pcibr_dmamap_ate32(pcidev_info, + phys_addr, size, + PCI32_ATE_BAR); + } + + return dma_handle; +} + EXPORT_SYMBOL(sn_dma_flush); diff --git a/arch/ia64/sn/pci/pcibr/pcibr_provider.c b/arch/ia64/sn/pci/pcibr/pcibr_provider.c index 92bd278cf7f..539ab1fdab2 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_provider.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_provider.c @@ -168,3 +168,23 @@ void pcibr_change_devices_irq(struct sn_irq_info *sn_irq_info) pcibr_force_interrupt(sn_irq_info); } } + +/* + * Provider entries for PIC/CP + */ + +struct sn_pcibus_provider pcibr_provider = { + .dma_map = pcibr_dma_map, + .dma_map_consistent = pcibr_dma_map_consistent, + .dma_unmap = pcibr_dma_unmap, + .bus_fixup = pcibr_bus_fixup, +}; + +int +pcibr_init_provider(void) +{ + sn_pci_provider[PCIIO_ASIC_TYPE_PIC] = &pcibr_provider; + sn_pci_provider[PCIIO_ASIC_TYPE_TIOCP] = &pcibr_provider; + + return 0; +} -- cgit v1.2.3-70-g09d2 From 9b08ebd167a3f3812131a54512f92c3c6738ad03 Mon Sep 17 00:00:00 2001 From: Mark Maule Date: Mon, 25 Apr 2005 11:32:16 -0700 Subject: [IA64-SGI] sn2-move-pci-headers.patch Move a couple of headers out of arch/ia64/sn/include/pci and into include/asm-ia64/sn. Signed-off-by: Mark Maule Signed-off-by: Tony Luck --- arch/ia64/sn/include/pci/pcibus_provider_defs.h | 51 ---------------------- arch/ia64/sn/include/pci/pcidev.h | 58 ------------------------- arch/ia64/sn/kernel/io_init.c | 4 +- arch/ia64/sn/kernel/irq.c | 4 +- arch/ia64/sn/pci/pci_dma.c | 4 +- arch/ia64/sn/pci/pcibr/pcibr_ate.c | 4 +- arch/ia64/sn/pci/pcibr/pcibr_dma.c | 4 +- arch/ia64/sn/pci/pcibr/pcibr_provider.c | 4 +- arch/ia64/sn/pci/pcibr/pcibr_reg.c | 4 +- include/asm-ia64/sn/pcibus_provider_defs.h | 51 ++++++++++++++++++++++ include/asm-ia64/sn/pcidev.h | 58 +++++++++++++++++++++++++ 11 files changed, 123 insertions(+), 123 deletions(-) delete mode 100644 arch/ia64/sn/include/pci/pcibus_provider_defs.h delete mode 100644 arch/ia64/sn/include/pci/pcidev.h create mode 100644 include/asm-ia64/sn/pcibus_provider_defs.h create mode 100644 include/asm-ia64/sn/pcidev.h (limited to 'arch') diff --git a/arch/ia64/sn/include/pci/pcibus_provider_defs.h b/arch/ia64/sn/include/pci/pcibus_provider_defs.h deleted file mode 100644 index f546b4ece33..00000000000 --- a/arch/ia64/sn/include/pci/pcibus_provider_defs.h +++ /dev/null @@ -1,51 +0,0 @@ -/* - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 1992 - 1997, 2000-2004 Silicon Graphics, Inc. All rights reserved. - */ -#ifndef _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H -#define _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H - -/* - * SN pci asic types. Do not ever renumber these or reuse values. The - * values must agree with what prom thinks they are. - */ - -#define PCIIO_ASIC_TYPE_UNKNOWN 0 -#define PCIIO_ASIC_TYPE_PPB 1 -#define PCIIO_ASIC_TYPE_PIC 2 -#define PCIIO_ASIC_TYPE_TIOCP 3 - -#define PCIIO_ASIC_MAX_TYPES 4 - -/* - * Common pciio bus provider data. There should be one of these as the - * first field in any pciio based provider soft structure (e.g. pcibr_soft - * tioca_soft, etc). - */ - -struct pcibus_bussoft { - uint32_t bs_asic_type; /* chipset type */ - uint32_t bs_xid; /* xwidget id */ - uint64_t bs_persist_busnum; /* Persistent Bus Number */ - uint64_t bs_legacy_io; /* legacy io pio addr */ - uint64_t bs_legacy_mem; /* legacy mem pio addr */ - uint64_t bs_base; /* widget base */ - struct xwidget_info *bs_xwidget_info; -}; - -/* - * SN pci bus indirection - */ - -struct sn_pcibus_provider { - dma_addr_t (*dma_map)(struct pci_dev *, unsigned long, size_t); - dma_addr_t (*dma_map_consistent)(struct pci_dev *, unsigned long, size_t); - void (*dma_unmap)(struct pci_dev *, dma_addr_t, int); - void * (*bus_fixup)(struct pcibus_bussoft *); -}; - -extern struct sn_pcibus_provider *sn_pci_provider[]; -#endif /* _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H */ diff --git a/arch/ia64/sn/include/pci/pcidev.h b/arch/ia64/sn/include/pci/pcidev.h deleted file mode 100644 index ed4031d8081..00000000000 --- a/arch/ia64/sn/include/pci/pcidev.h +++ /dev/null @@ -1,58 +0,0 @@ -/* - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 1992 - 1997, 2000-2004 Silicon Graphics, Inc. All rights reserved. - */ -#ifndef _ASM_IA64_SN_PCI_PCIDEV_H -#define _ASM_IA64_SN_PCI_PCIDEV_H - -#include - -extern struct sn_irq_info **sn_irq; - -#define SN_PCIDEV_INFO(pci_dev) \ - ((struct pcidev_info *)(pci_dev)->sysdata) - -/* - * Given a pci_bus, return the sn pcibus_bussoft struct. Note that - * this only works for root busses, not for busses represented by PPB's. - */ - -#define SN_PCIBUS_BUSSOFT(pci_bus) \ - ((struct pcibus_bussoft *)(PCI_CONTROLLER((pci_bus))->platform_data)) - -/* - * Given a struct pci_dev, return the sn pcibus_bussoft struct. Note - * that this is not equivalent to SN_PCIBUS_BUSSOFT(pci_dev->bus) due - * due to possible PPB's in the path. - */ - -#define SN_PCIDEV_BUSSOFT(pci_dev) \ - (SN_PCIDEV_INFO(pci_dev)->pdi_host_pcidev_info->pdi_pcibus_info) - -#define SN_PCIDEV_BUSPROVIDER(pci_dev) \ - (SN_PCIDEV_INFO(pci_dev)->pdi_provider) - -#define PCIIO_BUS_NONE 255 /* bus 255 reserved */ -#define PCIIO_SLOT_NONE 255 -#define PCIIO_FUNC_NONE 255 -#define PCIIO_VENDOR_ID_NONE (-1) - -struct pcidev_info { - uint64_t pdi_pio_mapped_addr[7]; /* 6 BARs PLUS 1 ROM */ - uint64_t pdi_slot_host_handle; /* Bus and devfn Host pci_dev */ - - struct pcibus_bussoft *pdi_pcibus_info; /* Kernel common bus soft */ - struct pcidev_info *pdi_host_pcidev_info; /* Kernel Host pci_dev */ - struct pci_dev *pdi_linux_pcidev; /* Kernel pci_dev */ - - struct sn_irq_info *pdi_sn_irq_info; - struct sn_pcibus_provider *pdi_provider; /* sn pci ops */ -}; - -extern void sn_irq_fixup(struct pci_dev *pci_dev, - struct sn_irq_info *sn_irq_info); - -#endif /* _ASM_IA64_SN_PCI_PCIDEV_H */ diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c index 3e5e4a90130..9f9d0464f72 100644 --- a/arch/ia64/sn/kernel/io_init.c +++ b/arch/ia64/sn/kernel/io_init.c @@ -11,8 +11,8 @@ #include #include #include -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #include "pci/pcibr_provider.h" #include "xtalk/xwidgetdev.h" #include diff --git a/arch/ia64/sn/kernel/irq.c b/arch/ia64/sn/kernel/irq.c index 3be44724f6c..b52d32975f2 100644 --- a/arch/ia64/sn/kernel/irq.c +++ b/arch/ia64/sn/kernel/irq.c @@ -13,8 +13,8 @@ #include #include #include "xtalk/xwidgetdev.h" -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #include "pci/pcibr_provider.h" #include #include diff --git a/arch/ia64/sn/pci/pci_dma.c b/arch/ia64/sn/pci/pci_dma.c index c2b92b94c56..5da9bdbde7c 100644 --- a/arch/ia64/sn/pci/pci_dma.c +++ b/arch/ia64/sn/pci/pci_dma.c @@ -12,8 +12,8 @@ #include #include #include -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #define SG_ENT_VIRT_ADDRESS(sg) (page_address((sg)->page) + (sg)->offset) #define SG_ENT_PHYS_ADDRESS(SG) virt_to_phys(SG_ENT_VIRT_ADDRESS(SG)) diff --git a/arch/ia64/sn/pci/pcibr/pcibr_ate.c b/arch/ia64/sn/pci/pcibr/pcibr_ate.c index 9d6854666f9..0e47bce85f2 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_ate.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_ate.c @@ -8,8 +8,8 @@ #include #include -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #include "pci/pcibr_provider.h" int pcibr_invalidate_ate = 0; /* by default don't invalidate ATE on free */ diff --git a/arch/ia64/sn/pci/pcibr/pcibr_dma.c b/arch/ia64/sn/pci/pcibr/pcibr_dma.c index 3c305f46417..c90685985d8 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_dma.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_dma.c @@ -12,8 +12,8 @@ #include #include "xtalk/xwidgetdev.h" #include "xtalk/hubdev.h" -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #include "pci/tiocp.h" #include "pci/pic.h" #include "pci/pcibr_provider.h" diff --git a/arch/ia64/sn/pci/pcibr/pcibr_provider.c b/arch/ia64/sn/pci/pcibr/pcibr_provider.c index 539ab1fdab2..3893999d23d 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_provider.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_provider.c @@ -13,8 +13,8 @@ #include "xtalk/xwidgetdev.h" #include #include "xtalk/hubdev.h" -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #include "pci/pcibr_provider.h" #include diff --git a/arch/ia64/sn/pci/pcibr/pcibr_reg.c b/arch/ia64/sn/pci/pcibr/pcibr_reg.c index 74a74a7d2a1..865c11c3b50 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_reg.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_reg.c @@ -8,8 +8,8 @@ #include #include -#include "pci/pcibus_provider_defs.h" -#include "pci/pcidev.h" +#include +#include #include "pci/tiocp.h" #include "pci/pic.h" #include "pci/pcibr_provider.h" diff --git a/include/asm-ia64/sn/pcibus_provider_defs.h b/include/asm-ia64/sn/pcibus_provider_defs.h new file mode 100644 index 00000000000..f546b4ece33 --- /dev/null +++ b/include/asm-ia64/sn/pcibus_provider_defs.h @@ -0,0 +1,51 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 1992 - 1997, 2000-2004 Silicon Graphics, Inc. All rights reserved. + */ +#ifndef _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H +#define _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H + +/* + * SN pci asic types. Do not ever renumber these or reuse values. The + * values must agree with what prom thinks they are. + */ + +#define PCIIO_ASIC_TYPE_UNKNOWN 0 +#define PCIIO_ASIC_TYPE_PPB 1 +#define PCIIO_ASIC_TYPE_PIC 2 +#define PCIIO_ASIC_TYPE_TIOCP 3 + +#define PCIIO_ASIC_MAX_TYPES 4 + +/* + * Common pciio bus provider data. There should be one of these as the + * first field in any pciio based provider soft structure (e.g. pcibr_soft + * tioca_soft, etc). + */ + +struct pcibus_bussoft { + uint32_t bs_asic_type; /* chipset type */ + uint32_t bs_xid; /* xwidget id */ + uint64_t bs_persist_busnum; /* Persistent Bus Number */ + uint64_t bs_legacy_io; /* legacy io pio addr */ + uint64_t bs_legacy_mem; /* legacy mem pio addr */ + uint64_t bs_base; /* widget base */ + struct xwidget_info *bs_xwidget_info; +}; + +/* + * SN pci bus indirection + */ + +struct sn_pcibus_provider { + dma_addr_t (*dma_map)(struct pci_dev *, unsigned long, size_t); + dma_addr_t (*dma_map_consistent)(struct pci_dev *, unsigned long, size_t); + void (*dma_unmap)(struct pci_dev *, dma_addr_t, int); + void * (*bus_fixup)(struct pcibus_bussoft *); +}; + +extern struct sn_pcibus_provider *sn_pci_provider[]; +#endif /* _ASM_IA64_SN_PCI_PCIBUS_PROVIDER_H */ diff --git a/include/asm-ia64/sn/pcidev.h b/include/asm-ia64/sn/pcidev.h new file mode 100644 index 00000000000..ed4031d8081 --- /dev/null +++ b/include/asm-ia64/sn/pcidev.h @@ -0,0 +1,58 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 1992 - 1997, 2000-2004 Silicon Graphics, Inc. All rights reserved. + */ +#ifndef _ASM_IA64_SN_PCI_PCIDEV_H +#define _ASM_IA64_SN_PCI_PCIDEV_H + +#include + +extern struct sn_irq_info **sn_irq; + +#define SN_PCIDEV_INFO(pci_dev) \ + ((struct pcidev_info *)(pci_dev)->sysdata) + +/* + * Given a pci_bus, return the sn pcibus_bussoft struct. Note that + * this only works for root busses, not for busses represented by PPB's. + */ + +#define SN_PCIBUS_BUSSOFT(pci_bus) \ + ((struct pcibus_bussoft *)(PCI_CONTROLLER((pci_bus))->platform_data)) + +/* + * Given a struct pci_dev, return the sn pcibus_bussoft struct. Note + * that this is not equivalent to SN_PCIBUS_BUSSOFT(pci_dev->bus) due + * due to possible PPB's in the path. + */ + +#define SN_PCIDEV_BUSSOFT(pci_dev) \ + (SN_PCIDEV_INFO(pci_dev)->pdi_host_pcidev_info->pdi_pcibus_info) + +#define SN_PCIDEV_BUSPROVIDER(pci_dev) \ + (SN_PCIDEV_INFO(pci_dev)->pdi_provider) + +#define PCIIO_BUS_NONE 255 /* bus 255 reserved */ +#define PCIIO_SLOT_NONE 255 +#define PCIIO_FUNC_NONE 255 +#define PCIIO_VENDOR_ID_NONE (-1) + +struct pcidev_info { + uint64_t pdi_pio_mapped_addr[7]; /* 6 BARs PLUS 1 ROM */ + uint64_t pdi_slot_host_handle; /* Bus and devfn Host pci_dev */ + + struct pcibus_bussoft *pdi_pcibus_info; /* Kernel common bus soft */ + struct pcidev_info *pdi_host_pcidev_info; /* Kernel Host pci_dev */ + struct pci_dev *pdi_linux_pcidev; /* Kernel pci_dev */ + + struct sn_irq_info *pdi_sn_irq_info; + struct sn_pcibus_provider *pdi_provider; /* sn pci ops */ +}; + +extern void sn_irq_fixup(struct pci_dev *pci_dev, + struct sn_irq_info *sn_irq_info); + +#endif /* _ASM_IA64_SN_PCI_PCIDEV_H */ -- cgit v1.2.3-70-g09d2 From 9c90bdde77f7b7a42f7ebb900275d459ce2bac05 Mon Sep 17 00:00:00 2001 From: Mark Maule Date: Mon, 25 Apr 2005 11:35:54 -0700 Subject: [IA64-SGI] altix: tioca chip driver (agp) Provide a driver for the altix TIOCA AGP chipset. An agpgart backend will be provided as a separate patch. Signed-off-by: Mark Maule Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/io_init.c | 2 + arch/ia64/sn/pci/Makefile | 2 +- arch/ia64/sn/pci/tioca_provider.c | 668 +++++++++++++++++++++++++++++ include/asm-ia64/sn/pcibus_provider_defs.h | 3 +- include/asm-ia64/sn/tioca.h | 596 +++++++++++++++++++++++++ include/asm-ia64/sn/tioca_provider.h | 206 +++++++++ 6 files changed, 1475 insertions(+), 2 deletions(-) create mode 100644 arch/ia64/sn/pci/tioca_provider.c create mode 100644 include/asm-ia64/sn/tioca.h create mode 100644 include/asm-ia64/sn/tioca_provider.h (limited to 'arch') diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c index 9f9d0464f72..18160a06a8c 100644 --- a/arch/ia64/sn/kernel/io_init.c +++ b/arch/ia64/sn/kernel/io_init.c @@ -19,6 +19,7 @@ #include "xtalk/hubdev.h" #include #include +#include char master_baseio_wid; nasid_t master_nasid = INVALID_NASID; /* Partition Master */ @@ -393,6 +394,7 @@ static int __init sn_pci_init(void) sn_pci_provider[i] = &sn_pci_default_provider; pcibr_init_provider(); + tioca_init_provider(); /* * This is needed to avoid bounce limit checks in the blk layer diff --git a/arch/ia64/sn/pci/Makefile b/arch/ia64/sn/pci/Makefile index b5dca0097a8..2f915bce25f 100644 --- a/arch/ia64/sn/pci/Makefile +++ b/arch/ia64/sn/pci/Makefile @@ -7,4 +7,4 @@ # # Makefile for the sn pci general routines. -obj-y := pci_dma.o pcibr/ +obj-y := pci_dma.o tioca_provider.o pcibr/ diff --git a/arch/ia64/sn/pci/tioca_provider.c b/arch/ia64/sn/pci/tioca_provider.c new file mode 100644 index 00000000000..2234d61cdd4 --- /dev/null +++ b/arch/ia64/sn/pci/tioca_provider.c @@ -0,0 +1,668 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2003-2005 Silicon Graphics, Inc. All Rights Reserved. + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +uint32_t tioca_gart_found; +EXPORT_SYMBOL(tioca_gart_found); /* used by agp-sgi */ + +LIST_HEAD(tioca_list); +EXPORT_SYMBOL(tioca_list); /* used by agp-sgi */ + +static int tioca_gart_init(struct tioca_kernel *); + +/** + * tioca_gart_init - Initialize SGI TIOCA GART + * @tioca_common: ptr to common prom/kernel struct identifying the + * + * If the indicated tioca has devices present, initialize its associated + * GART MMR's and kernel memory. + */ +static int +tioca_gart_init(struct tioca_kernel *tioca_kern) +{ + uint64_t ap_reg; + uint64_t offset; + struct page *tmp; + struct tioca_common *tioca_common; + volatile struct tioca *ca_base; + + tioca_common = tioca_kern->ca_common; + ca_base = (struct tioca *)tioca_common->ca_common.bs_base; + + if (list_empty(tioca_kern->ca_devices)) + return 0; + + ap_reg = 0; + + /* + * Validate aperature size + */ + + switch (CA_APERATURE_SIZE >> 20) { + case 4: + ap_reg |= (0x3ff << CA_GART_AP_SIZE_SHFT); /* 4MB */ + break; + case 8: + ap_reg |= (0x3fe << CA_GART_AP_SIZE_SHFT); /* 8MB */ + break; + case 16: + ap_reg |= (0x3fc << CA_GART_AP_SIZE_SHFT); /* 16MB */ + break; + case 32: + ap_reg |= (0x3f8 << CA_GART_AP_SIZE_SHFT); /* 32 MB */ + break; + case 64: + ap_reg |= (0x3f0 << CA_GART_AP_SIZE_SHFT); /* 64 MB */ + break; + case 128: + ap_reg |= (0x3e0 << CA_GART_AP_SIZE_SHFT); /* 128 MB */ + break; + case 256: + ap_reg |= (0x3c0 << CA_GART_AP_SIZE_SHFT); /* 256 MB */ + break; + case 512: + ap_reg |= (0x380 << CA_GART_AP_SIZE_SHFT); /* 512 MB */ + break; + case 1024: + ap_reg |= (0x300 << CA_GART_AP_SIZE_SHFT); /* 1GB */ + break; + case 2048: + ap_reg |= (0x200 << CA_GART_AP_SIZE_SHFT); /* 2GB */ + break; + case 4096: + ap_reg |= (0x000 << CA_GART_AP_SIZE_SHFT); /* 4 GB */ + break; + default: + printk(KERN_ERR "%s: Invalid CA_APERATURE_SIZE " + "0x%lx\n", __FUNCTION__, (ulong) CA_APERATURE_SIZE); + return -1; + } + + /* + * Set up other aperature parameters + */ + + if (PAGE_SIZE >= 16384) { + tioca_kern->ca_ap_pagesize = 16384; + ap_reg |= CA_GART_PAGE_SIZE; + } else { + tioca_kern->ca_ap_pagesize = 4096; + } + + tioca_kern->ca_ap_size = CA_APERATURE_SIZE; + tioca_kern->ca_ap_bus_base = CA_APERATURE_BASE; + tioca_kern->ca_gart_entries = + tioca_kern->ca_ap_size / tioca_kern->ca_ap_pagesize; + + ap_reg |= (CA_GART_AP_ENB_AGP | CA_GART_AP_ENB_PCI); + ap_reg |= tioca_kern->ca_ap_bus_base; + + /* + * Allocate and set up the GART + */ + + tioca_kern->ca_gart_size = tioca_kern->ca_gart_entries * sizeof(u64); + tmp = + alloc_pages_node(tioca_kern->ca_closest_node, + GFP_KERNEL | __GFP_ZERO, + get_order(tioca_kern->ca_gart_size)); + + if (!tmp) { + printk(KERN_ERR "%s: Could not allocate " + "%lu bytes (order %d) for GART\n", + __FUNCTION__, + tioca_kern->ca_gart_size, + get_order(tioca_kern->ca_gart_size)); + return -ENOMEM; + } + + tioca_kern->ca_gart = page_address(tmp); + tioca_kern->ca_gart_coretalk_addr = + PHYS_TO_TIODMA(virt_to_phys(tioca_kern->ca_gart)); + + /* + * Compute PCI/AGP convenience fields + */ + + offset = CA_PCI32_MAPPED_BASE - CA_APERATURE_BASE; + tioca_kern->ca_pciap_base = CA_PCI32_MAPPED_BASE; + tioca_kern->ca_pciap_size = CA_PCI32_MAPPED_SIZE; + tioca_kern->ca_pcigart_start = offset / tioca_kern->ca_ap_pagesize; + tioca_kern->ca_pcigart_base = + tioca_kern->ca_gart_coretalk_addr + offset; + tioca_kern->ca_pcigart = + &tioca_kern->ca_gart[tioca_kern->ca_pcigart_start]; + tioca_kern->ca_pcigart_entries = + tioca_kern->ca_pciap_size / tioca_kern->ca_ap_pagesize; + tioca_kern->ca_pcigart_pagemap = + kcalloc(1, tioca_kern->ca_pcigart_entries / 8, GFP_KERNEL); + if (!tioca_kern->ca_pcigart_pagemap) { + free_pages((unsigned long)tioca_kern->ca_gart, + get_order(tioca_kern->ca_gart_size)); + return -1; + } + + offset = CA_AGP_MAPPED_BASE - CA_APERATURE_BASE; + tioca_kern->ca_gfxap_base = CA_AGP_MAPPED_BASE; + tioca_kern->ca_gfxap_size = CA_AGP_MAPPED_SIZE; + tioca_kern->ca_gfxgart_start = offset / tioca_kern->ca_ap_pagesize; + tioca_kern->ca_gfxgart_base = + tioca_kern->ca_gart_coretalk_addr + offset; + tioca_kern->ca_gfxgart = + &tioca_kern->ca_gart[tioca_kern->ca_gfxgart_start]; + tioca_kern->ca_gfxgart_entries = + tioca_kern->ca_gfxap_size / tioca_kern->ca_ap_pagesize; + + /* + * various control settings: + * use agp op-combining + * use GET semantics to fetch memory + * participate in coherency domain + * prefetch TLB entries + */ + + ca_base->ca_control1 |= CA_AGPDMA_OP_ENB_COMBDELAY; /* PV895469 ? */ + ca_base->ca_control2 &= ~(CA_GART_MEM_PARAM); + ca_base->ca_control2 |= (0x2ull << CA_GART_MEM_PARAM_SHFT); + tioca_kern->ca_gart_iscoherent = 1; + ca_base->ca_control2 |= + (CA_GART_WR_PREFETCH_ENB | CA_GART_RD_PREFETCH_ENB); + + /* + * Unmask GART fetch error interrupts. Clear residual errors first. + */ + + ca_base->ca_int_status_alias = CA_GART_FETCH_ERR; + ca_base->ca_mult_error_alias = CA_GART_FETCH_ERR; + ca_base->ca_int_mask &= ~CA_GART_FETCH_ERR; + + /* + * Program the aperature and gart registers in TIOCA + */ + + ca_base->ca_gart_aperature = ap_reg; + ca_base->ca_gart_ptr_table = tioca_kern->ca_gart_coretalk_addr | 1; + + return 0; +} + +/** + * tioca_fastwrite_enable - enable AGP FW for a tioca and its functions + * @tioca_kernel: structure representing the CA + * + * Given a CA, scan all attached functions making sure they all support + * FastWrite. If so, enable FastWrite for all functions and the CA itself. + */ + +void +tioca_fastwrite_enable(struct tioca_kernel *tioca_kern) +{ + int cap_ptr; + uint64_t ca_control1; + uint32_t reg; + struct tioca *tioca_base; + struct pci_dev *pdev; + struct tioca_common *common; + + common = tioca_kern->ca_common; + + /* + * Scan all vga controllers on this bus making sure they all + * suport FW. If not, return. + */ + + list_for_each_entry(pdev, tioca_kern->ca_devices, bus_list) { + if (pdev->class != (PCI_CLASS_DISPLAY_VGA << 8)) + continue; + + cap_ptr = pci_find_capability(pdev, PCI_CAP_ID_AGP); + if (!cap_ptr) + return; /* no AGP CAP means no FW */ + + pci_read_config_dword(pdev, cap_ptr + PCI_AGP_STATUS, ®); + if (!(reg & PCI_AGP_STATUS_FW)) + return; /* function doesn't support FW */ + } + + /* + * Set fw for all vga fn's + */ + + list_for_each_entry(pdev, tioca_kern->ca_devices, bus_list) { + if (pdev->class != (PCI_CLASS_DISPLAY_VGA << 8)) + continue; + + cap_ptr = pci_find_capability(pdev, PCI_CAP_ID_AGP); + pci_read_config_dword(pdev, cap_ptr + PCI_AGP_COMMAND, ®); + reg |= PCI_AGP_COMMAND_FW; + pci_write_config_dword(pdev, cap_ptr + PCI_AGP_COMMAND, reg); + } + + /* + * Set ca's fw to match + */ + + tioca_base = (struct tioca *)common->ca_common.bs_base; + ca_control1 = tioca_base->ca_control1; + ca_control1 |= CA_AGP_FW_ENABLE; + tioca_base->ca_control1 = ca_control1; +} + +EXPORT_SYMBOL(tioca_fastwrite_enable); /* used by agp-sgi */ + +/** + * tioca_dma_d64 - create a DMA mapping using 64-bit direct mode + * @paddr: system physical address + * + * Map @paddr into 64-bit CA bus space. No device context is necessary. + * Bits 53:0 come from the coretalk address. We just need to mask in the + * following optional bits of the 64-bit pci address: + * + * 63:60 - Coretalk Packet Type - 0x1 for Mem Get/Put (coherent) + * 0x2 for PIO (non-coherent) + * We will always use 0x1 + * 55:55 - Swap bytes Currently unused + */ +static uint64_t +tioca_dma_d64(unsigned long paddr) +{ + dma_addr_t bus_addr; + + bus_addr = PHYS_TO_TIODMA(paddr); + + BUG_ON(!bus_addr); + BUG_ON(bus_addr >> 54); + + /* Set upper nibble to Cache Coherent Memory op */ + bus_addr |= (1UL << 60); + + return bus_addr; +} + +/** + * tioca_dma_d48 - create a DMA mapping using 48-bit direct mode + * @pdev: linux pci_dev representing the function + * @paddr: system physical address + * + * Map @paddr into 64-bit bus space of the CA associated with @pcidev_info. + * + * The CA agp 48 bit direct address falls out as follows: + * + * When direct mapping AGP addresses, the 48 bit AGP address is + * constructed as follows: + * + * [47:40] - Low 8 bits of the page Node ID extracted from coretalk + * address [47:40]. The upper 8 node bits are fixed + * and come from the xxx register bits [5:0] + * [39:38] - Chiplet ID extracted from coretalk address [39:38] + * [37:00] - node offset extracted from coretalk address [37:00] + * + * Since the node id in general will be non-zero, and the chiplet id + * will always be non-zero, it follows that the device must support + * a dma mask of at least 0xffffffffff (40 bits) to target node 0 + * and in general should be 0xffffffffffff (48 bits) to target nodes + * up to 255. Nodes above 255 need the support of the xxx register, + * and so a given CA can only directly target nodes in the range + * xxx - xxx+255. + */ +static uint64_t +tioca_dma_d48(struct pci_dev *pdev, uint64_t paddr) +{ + struct tioca_common *tioca_common; + struct tioca *ca_base; + uint64_t ct_addr; + dma_addr_t bus_addr; + uint32_t node_upper; + uint64_t agp_dma_extn; + struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(pdev); + + tioca_common = (struct tioca_common *)pcidev_info->pdi_pcibus_info; + ca_base = (struct tioca *)tioca_common->ca_common.bs_base; + + ct_addr = PHYS_TO_TIODMA(paddr); + if (!ct_addr) + return 0; + + bus_addr = (dma_addr_t) (ct_addr & 0xffffffffffff); + node_upper = ct_addr >> 48; + + if (node_upper > 64) { + printk(KERN_ERR "%s: coretalk addr 0x%p node id out " + "of range\n", __FUNCTION__, (void *)ct_addr); + return 0; + } + + agp_dma_extn = ca_base->ca_agp_dma_addr_extn; + if (node_upper != (agp_dma_extn >> CA_AGP_DMA_NODE_ID_SHFT)) { + printk(KERN_ERR "%s: coretalk upper node (%u) " + "mismatch with ca_agp_dma_addr_extn (%lu)\n", + __FUNCTION__, + node_upper, (agp_dma_extn >> CA_AGP_DMA_NODE_ID_SHFT)); + return 0; + } + + return bus_addr; +} + +/** + * tioca_dma_mapped - create a DMA mapping using a CA GART + * @pdev: linux pci_dev representing the function + * @paddr: host physical address to map + * @req_size: len (bytes) to map + * + * Map @paddr into CA address space using the GART mechanism. The mapped + * dma_addr_t is guarenteed to be contiguous in CA bus space. + */ +static dma_addr_t +tioca_dma_mapped(struct pci_dev *pdev, uint64_t paddr, size_t req_size) +{ + int i, ps, ps_shift, entry, entries, mapsize, last_entry; + uint64_t xio_addr, end_xio_addr; + struct tioca_common *tioca_common; + struct tioca_kernel *tioca_kern; + dma_addr_t bus_addr = 0; + struct tioca_dmamap *ca_dmamap; + void *map; + unsigned long flags; + struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(pdev);; + + tioca_common = (struct tioca_common *)pcidev_info->pdi_pcibus_info; + tioca_kern = (struct tioca_kernel *)tioca_common->ca_kernel_private; + + xio_addr = PHYS_TO_TIODMA(paddr); + if (!xio_addr) + return 0; + + spin_lock_irqsave(&tioca_kern->ca_lock, flags); + + /* + * allocate a map struct + */ + + ca_dmamap = kcalloc(1, sizeof(struct tioca_dmamap), GFP_ATOMIC); + if (!ca_dmamap) + goto map_return; + + /* + * Locate free entries that can hold req_size. Account for + * unaligned start/length when allocating. + */ + + ps = tioca_kern->ca_ap_pagesize; /* will be power of 2 */ + ps_shift = ffs(ps) - 1; + end_xio_addr = xio_addr + req_size - 1; + + entries = (end_xio_addr >> ps_shift) - (xio_addr >> ps_shift) + 1; + + map = tioca_kern->ca_pcigart_pagemap; + mapsize = tioca_kern->ca_pcigart_entries; + + entry = find_first_zero_bit(map, mapsize); + while (entry < mapsize) { + last_entry = find_next_bit(map, mapsize, entry); + + if (last_entry - entry >= entries) + break; + + entry = find_next_zero_bit(map, mapsize, last_entry); + } + + if (entry > mapsize) + goto map_return; + + for (i = 0; i < entries; i++) + set_bit(entry + i, map); + + bus_addr = tioca_kern->ca_pciap_base + (entry * ps); + + ca_dmamap->cad_dma_addr = bus_addr; + ca_dmamap->cad_gart_size = entries; + ca_dmamap->cad_gart_entry = entry; + list_add(&ca_dmamap->cad_list, &tioca_kern->ca_list); + + if (xio_addr % ps) { + tioca_kern->ca_pcigart[entry] = tioca_paddr_to_gart(xio_addr); + bus_addr += xio_addr & (ps - 1); + xio_addr &= ~(ps - 1); + xio_addr += ps; + entry++; + } + + while (xio_addr < end_xio_addr) { + tioca_kern->ca_pcigart[entry] = tioca_paddr_to_gart(xio_addr); + xio_addr += ps; + entry++; + } + + tioca_tlbflush(tioca_kern); + +map_return: + spin_unlock_irqrestore(&tioca_kern->ca_lock, flags); + return bus_addr; +} + +/** + * tioca_dma_unmap - release CA mapping resources + * @pdev: linux pci_dev representing the function + * @bus_addr: bus address returned by an earlier tioca_dma_map + * @dir: mapping direction (unused) + * + * Locate mapping resources associated with @bus_addr and release them. + * For mappings created using the direct modes (64 or 48) there are no + * resources to release. + */ +void +tioca_dma_unmap(struct pci_dev *pdev, dma_addr_t bus_addr, int dir) +{ + int i, entry; + struct tioca_common *tioca_common; + struct tioca_kernel *tioca_kern; + struct tioca_dmamap *map; + struct pcidev_info *pcidev_info = SN_PCIDEV_INFO(pdev); + unsigned long flags; + + tioca_common = (struct tioca_common *)pcidev_info->pdi_pcibus_info; + tioca_kern = (struct tioca_kernel *)tioca_common->ca_kernel_private; + + /* return straight away if this isn't be a mapped address */ + + if (bus_addr < tioca_kern->ca_pciap_base || + bus_addr >= (tioca_kern->ca_pciap_base + tioca_kern->ca_pciap_size)) + return; + + spin_lock_irqsave(&tioca_kern->ca_lock, flags); + + list_for_each_entry(map, &tioca_kern->ca_dmamaps, cad_list) + if (map->cad_dma_addr == bus_addr) + break; + + BUG_ON(map == NULL); + + entry = map->cad_gart_entry; + + for (i = 0; i < map->cad_gart_size; i++, entry++) { + clear_bit(entry, tioca_kern->ca_pcigart_pagemap); + tioca_kern->ca_pcigart[entry] = 0; + } + tioca_tlbflush(tioca_kern); + + list_del(&map->cad_list); + spin_unlock_irqrestore(&tioca_kern->ca_lock, flags); + kfree(map); +} + +/** + * tioca_dma_map - map pages for PCI DMA + * @pdev: linux pci_dev representing the function + * @paddr: host physical address to map + * @byte_count: bytes to map + * + * This is the main wrapper for mapping host physical pages to CA PCI space. + * The mapping mode used is based on the devices dma_mask. As a last resort + * use the GART mapped mode. + */ +uint64_t +tioca_dma_map(struct pci_dev *pdev, uint64_t paddr, size_t byte_count) +{ + uint64_t mapaddr; + + /* + * If card is 64 or 48 bit addresable, use a direct mapping. 32 + * bit direct is so restrictive w.r.t. where the memory resides that + * we don't use it even though CA has some support. + */ + + if (pdev->dma_mask == ~0UL) + mapaddr = tioca_dma_d64(paddr); + else if (pdev->dma_mask == 0xffffffffffffUL) + mapaddr = tioca_dma_d48(pdev, paddr); + else + mapaddr = 0; + + /* Last resort ... use PCI portion of CA GART */ + + if (mapaddr == 0) + mapaddr = tioca_dma_mapped(pdev, paddr, byte_count); + + return mapaddr; +} + +/** + * tioca_error_intr_handler - SGI TIO CA error interrupt handler + * @irq: unused + * @arg: pointer to tioca_common struct for the given CA + * @pt: unused + * + * Handle a CA error interrupt. Simply a wrapper around a SAL call which + * defers processing to the SGI prom. + */ +static irqreturn_t +tioca_error_intr_handler(int irq, void *arg, struct pt_regs *pt) +{ + struct tioca_common *soft = arg; + struct ia64_sal_retval ret_stuff; + uint64_t segment; + uint64_t busnum; + ret_stuff.status = 0; + ret_stuff.v0 = 0; + + segment = 0; + busnum = soft->ca_common.bs_persist_busnum; + + SAL_CALL_NOLOCK(ret_stuff, + (u64) SN_SAL_IOIF_ERROR_INTERRUPT, + segment, busnum, 0, 0, 0, 0, 0); + + return IRQ_HANDLED; +} + +/** + * tioca_bus_fixup - perform final PCI fixup for a TIO CA bus + * @prom_bussoft: Common prom/kernel struct representing the bus + * + * Replicates the tioca_common pointed to by @prom_bussoft in kernel + * space. Allocates and initializes a kernel-only area for a given CA, + * and sets up an irq for handling CA error interrupts. + * + * On successful setup, returns the kernel version of tioca_common back to + * the caller. + */ +void * +tioca_bus_fixup(struct pcibus_bussoft *prom_bussoft) +{ + struct tioca_common *tioca_common; + struct tioca_kernel *tioca_kern; + struct pci_bus *bus; + + /* sanity check prom rev */ + + if (sn_sal_rev_major() < 4 || + (sn_sal_rev_major() == 4 && sn_sal_rev_minor() < 6)) { + printk + (KERN_ERR "%s: SGI prom rev 4.06 or greater required " + "for tioca support\n", __FUNCTION__); + return NULL; + } + + /* + * Allocate kernel bus soft and copy from prom. + */ + + tioca_common = kcalloc(1, sizeof(struct tioca_common), GFP_KERNEL); + if (!tioca_common) + return NULL; + + memcpy(tioca_common, prom_bussoft, sizeof(struct tioca_common)); + tioca_common->ca_common.bs_base |= __IA64_UNCACHED_OFFSET; + + /* init kernel-private area */ + + tioca_kern = kcalloc(1, sizeof(struct tioca_kernel), GFP_KERNEL); + if (!tioca_kern) { + kfree(tioca_common); + return NULL; + } + + tioca_kern->ca_common = tioca_common; + spin_lock_init(&tioca_kern->ca_lock); + INIT_LIST_HEAD(&tioca_kern->ca_dmamaps); + tioca_kern->ca_closest_node = + nasid_to_cnodeid(tioca_common->ca_closest_nasid); + tioca_common->ca_kernel_private = (uint64_t) tioca_kern; + + bus = pci_find_bus(0, tioca_common->ca_common.bs_persist_busnum); + BUG_ON(!bus); + tioca_kern->ca_devices = &bus->devices; + + /* init GART */ + + if (tioca_gart_init(tioca_kern) < 0) { + kfree(tioca_kern); + kfree(tioca_common); + return NULL; + } + + tioca_gart_found++; + list_add(&tioca_kern->ca_list, &tioca_list); + + if (request_irq(SGI_TIOCA_ERROR, + tioca_error_intr_handler, + SA_SHIRQ, "TIOCA error", (void *)tioca_common)) + printk(KERN_WARNING + "%s: Unable to get irq %d. " + "Error interrupts won't be routed for TIOCA bus %d\n", + __FUNCTION__, SGI_TIOCA_ERROR, + (int)tioca_common->ca_common.bs_persist_busnum); + + return tioca_common; +} + +static struct sn_pcibus_provider tioca_pci_interfaces = { + .dma_map = tioca_dma_map, + .dma_map_consistent = tioca_dma_map, + .dma_unmap = tioca_dma_unmap, + .bus_fixup = tioca_bus_fixup, +}; + +/** + * tioca_init_provider - init SN PCI provider ops for TIO CA + */ +int +tioca_init_provider(void) +{ + sn_pci_provider[PCIIO_ASIC_TYPE_TIOCA] = &tioca_pci_interfaces; + return 0; +} diff --git a/include/asm-ia64/sn/pcibus_provider_defs.h b/include/asm-ia64/sn/pcibus_provider_defs.h index f546b4ece33..04e27d5b382 100644 --- a/include/asm-ia64/sn/pcibus_provider_defs.h +++ b/include/asm-ia64/sn/pcibus_provider_defs.h @@ -17,8 +17,9 @@ #define PCIIO_ASIC_TYPE_PPB 1 #define PCIIO_ASIC_TYPE_PIC 2 #define PCIIO_ASIC_TYPE_TIOCP 3 +#define PCIIO_ASIC_TYPE_TIOCA 4 -#define PCIIO_ASIC_MAX_TYPES 4 +#define PCIIO_ASIC_MAX_TYPES 5 /* * Common pciio bus provider data. There should be one of these as the diff --git a/include/asm-ia64/sn/tioca.h b/include/asm-ia64/sn/tioca.h new file mode 100644 index 00000000000..bc1aacfb948 --- /dev/null +++ b/include/asm-ia64/sn/tioca.h @@ -0,0 +1,596 @@ +#ifndef _ASM_IA64_SN_TIO_TIOCA_H +#define _ASM_IA64_SN_TIO_TIOCA_H + +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2003-2005 Silicon Graphics, Inc. All rights reserved. + */ + + +#define TIOCA_PART_NUM 0xE020 +#define TIOCA_MFGR_NUM 0x24 +#define TIOCA_REV_A 0x1 + +/* + * Register layout for TIO:CA. See below for bitmasks for each register. + */ + +struct tioca { + uint64_t ca_id; /* 0x000000 */ + uint64_t ca_control1; /* 0x000008 */ + uint64_t ca_control2; /* 0x000010 */ + uint64_t ca_status1; /* 0x000018 */ + uint64_t ca_status2; /* 0x000020 */ + uint64_t ca_gart_aperature; /* 0x000028 */ + uint64_t ca_gfx_detach; /* 0x000030 */ + uint64_t ca_inta_dest_addr; /* 0x000038 */ + uint64_t ca_intb_dest_addr; /* 0x000040 */ + uint64_t ca_err_int_dest_addr; /* 0x000048 */ + uint64_t ca_int_status; /* 0x000050 */ + uint64_t ca_int_status_alias; /* 0x000058 */ + uint64_t ca_mult_error; /* 0x000060 */ + uint64_t ca_mult_error_alias; /* 0x000068 */ + uint64_t ca_first_error; /* 0x000070 */ + uint64_t ca_int_mask; /* 0x000078 */ + uint64_t ca_crm_pkterr_type; /* 0x000080 */ + uint64_t ca_crm_pkterr_type_alias; /* 0x000088 */ + uint64_t ca_crm_ct_error_detail_1; /* 0x000090 */ + uint64_t ca_crm_ct_error_detail_2; /* 0x000098 */ + uint64_t ca_crm_tnumto; /* 0x0000A0 */ + uint64_t ca_gart_err; /* 0x0000A8 */ + uint64_t ca_pcierr_type; /* 0x0000B0 */ + uint64_t ca_pcierr_addr; /* 0x0000B8 */ + + uint64_t ca_pad_0000C0[3]; /* 0x0000{C0..D0} */ + + uint64_t ca_pci_rd_buf_flush; /* 0x0000D8 */ + uint64_t ca_pci_dma_addr_extn; /* 0x0000E0 */ + uint64_t ca_agp_dma_addr_extn; /* 0x0000E8 */ + uint64_t ca_force_inta; /* 0x0000F0 */ + uint64_t ca_force_intb; /* 0x0000F8 */ + uint64_t ca_debug_vector_sel; /* 0x000100 */ + uint64_t ca_debug_mux_core_sel; /* 0x000108 */ + uint64_t ca_debug_mux_pci_sel; /* 0x000110 */ + uint64_t ca_debug_domain_sel; /* 0x000118 */ + + uint64_t ca_pad_000120[28]; /* 0x0001{20..F8} */ + + uint64_t ca_gart_ptr_table; /* 0x200 */ + uint64_t ca_gart_tlb_addr[8]; /* 0x2{08..40} */ +}; + +/* + * Mask/shift definitions for TIO:CA registers. The convention here is + * to mainly use the names as they appear in the "TIO AEGIS Programmers' + * Reference" with a CA_ prefix added. Some exceptions were made to fix + * duplicate field names or to generalize fields that are common to + * different registers (ca_debug_mux_core_sel and ca_debug_mux_pci_sel for + * example). + * + * Fields consisting of a single bit have a single #define have a single + * macro declaration to mask the bit. Fields consisting of multiple bits + * have two declarations: one to mask the proper bits in a register, and + * a second with the suffix "_SHFT" to identify how far the mask needs to + * be shifted right to get its base value. + */ + +/* ==== ca_control1 */ +#define CA_SYS_BIG_END (1ull << 0) +#define CA_DMA_AGP_SWAP (1ull << 1) +#define CA_DMA_PCI_SWAP (1ull << 2) +#define CA_PIO_IO_SWAP (1ull << 3) +#define CA_PIO_MEM_SWAP (1ull << 4) +#define CA_GFX_WR_SWAP (1ull << 5) +#define CA_AGP_FW_ENABLE (1ull << 6) +#define CA_AGP_CAL_CYCLE (0x7ull << 7) +#define CA_AGP_CAL_CYCLE_SHFT 7 +#define CA_AGP_CAL_PRSCL_BYP (1ull << 10) +#define CA_AGP_INIT_CAL_ENB (1ull << 11) +#define CA_INJ_ADDR_PERR (1ull << 12) +#define CA_INJ_DATA_PERR (1ull << 13) + /* bits 15:14 unused */ +#define CA_PCIM_IO_NBE_AD (0x7ull << 16) +#define CA_PCIM_IO_NBE_AD_SHFT 16 +#define CA_PCIM_FAST_BTB_ENB (1ull << 19) + /* bits 23:20 unused */ +#define CA_PIO_ADDR_OFFSET (0xffull << 24) +#define CA_PIO_ADDR_OFFSET_SHFT 24 + /* bits 35:32 unused */ +#define CA_AGPDMA_OP_COMBDELAY (0x1full << 36) +#define CA_AGPDMA_OP_COMBDELAY_SHFT 36 + /* bit 41 unused */ +#define CA_AGPDMA_OP_ENB_COMBDELAY (1ull << 42) +#define CA_PCI_INT_LPCNT (0xffull << 44) +#define CA_PCI_INT_LPCNT_SHFT 44 + /* bits 63:52 unused */ + +/* ==== ca_control2 */ +#define CA_AGP_LATENCY_TO (0xffull << 0) +#define CA_AGP_LATENCY_TO_SHFT 0 +#define CA_PCI_LATENCY_TO (0xffull << 8) +#define CA_PCI_LATENCY_TO_SHFT 8 +#define CA_PCI_MAX_RETRY (0x3ffull << 16) +#define CA_PCI_MAX_RETRY_SHFT 16 + /* bits 27:26 unused */ +#define CA_RT_INT_EN (0x3ull << 28) +#define CA_RT_INT_EN_SHFT 28 +#define CA_MSI_INT_ENB (1ull << 30) +#define CA_PCI_ARB_ERR_ENB (1ull << 31) +#define CA_GART_MEM_PARAM (0x3ull << 32) +#define CA_GART_MEM_PARAM_SHFT 32 +#define CA_GART_RD_PREFETCH_ENB (1ull << 34) +#define CA_GART_WR_PREFETCH_ENB (1ull << 35) +#define CA_GART_FLUSH_TLB (1ull << 36) + /* bits 39:37 unused */ +#define CA_CRM_TNUMTO_PERIOD (0x1fffull << 40) +#define CA_CRM_TNUMTO_PERIOD_SHFT 40 + /* bits 55:53 unused */ +#define CA_CRM_TNUMTO_ENB (1ull << 56) +#define CA_CRM_PRESCALER_BYP (1ull << 57) + /* bits 59:58 unused */ +#define CA_CRM_MAX_CREDIT (0x7ull << 60) +#define CA_CRM_MAX_CREDIT_SHFT 60 + /* bit 63 unused */ + +/* ==== ca_status1 */ +#define CA_CORELET_ID (0x3ull << 0) +#define CA_CORELET_ID_SHFT 0 +#define CA_INTA_N (1ull << 2) +#define CA_INTB_N (1ull << 3) +#define CA_CRM_CREDIT_AVAIL (0x7ull << 4) +#define CA_CRM_CREDIT_AVAIL_SHFT 4 + /* bit 7 unused */ +#define CA_CRM_SPACE_AVAIL (0x7full << 8) +#define CA_CRM_SPACE_AVAIL_SHFT 8 + /* bit 15 unused */ +#define CA_GART_TLB_VAL (0xffull << 16) +#define CA_GART_TLB_VAL_SHFT 16 + /* bits 63:24 unused */ + +/* ==== ca_status2 */ +#define CA_GFX_CREDIT_AVAIL (0xffull << 0) +#define CA_GFX_CREDIT_AVAIL_SHFT 0 +#define CA_GFX_OPQ_AVAIL (0xffull << 8) +#define CA_GFX_OPQ_AVAIL_SHFT 8 +#define CA_GFX_WRBUFF_AVAIL (0xffull << 16) +#define CA_GFX_WRBUFF_AVAIL_SHFT 16 +#define CA_ADMA_OPQ_AVAIL (0xffull << 24) +#define CA_ADMA_OPQ_AVAIL_SHFT 24 +#define CA_ADMA_WRBUFF_AVAIL (0xffull << 32) +#define CA_ADMA_WRBUFF_AVAIL_SHFT 32 +#define CA_ADMA_RDBUFF_AVAIL (0x7full << 40) +#define CA_ADMA_RDBUFF_AVAIL_SHFT 40 +#define CA_PCI_PIO_OP_STAT (1ull << 47) +#define CA_PDMA_OPQ_AVAIL (0xfull << 48) +#define CA_PDMA_OPQ_AVAIL_SHFT 48 +#define CA_PDMA_WRBUFF_AVAIL (0xfull << 52) +#define CA_PDMA_WRBUFF_AVAIL_SHFT 52 +#define CA_PDMA_RDBUFF_AVAIL (0x3ull << 56) +#define CA_PDMA_RDBUFF_AVAIL_SHFT 56 + /* bits 63:58 unused */ + +/* ==== ca_gart_aperature */ +#define CA_GART_AP_ENB_AGP (1ull << 0) +#define CA_GART_PAGE_SIZE (1ull << 1) +#define CA_GART_AP_ENB_PCI (1ull << 2) + /* bits 11:3 unused */ +#define CA_GART_AP_SIZE (0x3ffull << 12) +#define CA_GART_AP_SIZE_SHFT 12 +#define CA_GART_AP_BASE (0x3ffffffffffull << 22) +#define CA_GART_AP_BASE_SHFT 22 + +/* ==== ca_inta_dest_addr + ==== ca_intb_dest_addr + ==== ca_err_int_dest_addr */ + /* bits 2:0 unused */ +#define CA_INT_DEST_ADDR (0x7ffffffffffffull << 3) +#define CA_INT_DEST_ADDR_SHFT 3 + /* bits 55:54 unused */ +#define CA_INT_DEST_VECT (0xffull << 56) +#define CA_INT_DEST_VECT_SHFT 56 + +/* ==== ca_int_status */ +/* ==== ca_int_status_alias */ +/* ==== ca_mult_error */ +/* ==== ca_mult_error_alias */ +/* ==== ca_first_error */ +/* ==== ca_int_mask */ +#define CA_PCI_ERR (1ull << 0) + /* bits 3:1 unused */ +#define CA_GART_FETCH_ERR (1ull << 4) +#define CA_GFX_WR_OVFLW (1ull << 5) +#define CA_PIO_REQ_OVFLW (1ull << 6) +#define CA_CRM_PKTERR (1ull << 7) +#define CA_CRM_DVERR (1ull << 8) +#define CA_TNUMTO (1ull << 9) +#define CA_CXM_RSP_CRED_OVFLW (1ull << 10) +#define CA_CXM_REQ_CRED_OVFLW (1ull << 11) +#define CA_PIO_INVALID_ADDR (1ull << 12) +#define CA_PCI_ARB_TO (1ull << 13) +#define CA_AGP_REQ_OFLOW (1ull << 14) +#define CA_SBA_TYPE1_ERR (1ull << 15) + /* bit 16 unused */ +#define CA_INTA (1ull << 17) +#define CA_INTB (1ull << 18) +#define CA_MULT_INTA (1ull << 19) +#define CA_MULT_INTB (1ull << 20) +#define CA_GFX_CREDIT_OVFLW (1ull << 21) + /* bits 63:22 unused */ + +/* ==== ca_crm_pkterr_type */ +/* ==== ca_crm_pkterr_type_alias */ +#define CA_CRM_PKTERR_SBERR_HDR (1ull << 0) +#define CA_CRM_PKTERR_DIDN (1ull << 1) +#define CA_CRM_PKTERR_PACTYPE (1ull << 2) +#define CA_CRM_PKTERR_INV_TNUM (1ull << 3) +#define CA_CRM_PKTERR_ADDR_RNG (1ull << 4) +#define CA_CRM_PKTERR_ADDR_ALGN (1ull << 5) +#define CA_CRM_PKTERR_HDR_PARAM (1ull << 6) +#define CA_CRM_PKTERR_CW_ERR (1ull << 7) +#define CA_CRM_PKTERR_SBERR_NH (1ull << 8) +#define CA_CRM_PKTERR_EARLY_TERM (1ull << 9) +#define CA_CRM_PKTERR_EARLY_TAIL (1ull << 10) +#define CA_CRM_PKTERR_MSSNG_TAIL (1ull << 11) +#define CA_CRM_PKTERR_MSSNG_HDR (1ull << 12) + /* bits 15:13 unused */ +#define CA_FIRST_CRM_PKTERR_SBERR_HDR (1ull << 16) +#define CA_FIRST_CRM_PKTERR_DIDN (1ull << 17) +#define CA_FIRST_CRM_PKTERR_PACTYPE (1ull << 18) +#define CA_FIRST_CRM_PKTERR_INV_TNUM (1ull << 19) +#define CA_FIRST_CRM_PKTERR_ADDR_RNG (1ull << 20) +#define CA_FIRST_CRM_PKTERR_ADDR_ALGN (1ull << 21) +#define CA_FIRST_CRM_PKTERR_HDR_PARAM (1ull << 22) +#define CA_FIRST_CRM_PKTERR_CW_ERR (1ull << 23) +#define CA_FIRST_CRM_PKTERR_SBERR_NH (1ull << 24) +#define CA_FIRST_CRM_PKTERR_EARLY_TERM (1ull << 25) +#define CA_FIRST_CRM_PKTERR_EARLY_TAIL (1ull << 26) +#define CA_FIRST_CRM_PKTERR_MSSNG_TAIL (1ull << 27) +#define CA_FIRST_CRM_PKTERR_MSSNG_HDR (1ull << 28) + /* bits 63:29 unused */ + +/* ==== ca_crm_ct_error_detail_1 */ +#define CA_PKT_TYPE (0xfull << 0) +#define CA_PKT_TYPE_SHFT 0 +#define CA_SRC_ID (0x3ull << 4) +#define CA_SRC_ID_SHFT 4 +#define CA_DATA_SZ (0x3ull << 6) +#define CA_DATA_SZ_SHFT 6 +#define CA_TNUM (0xffull << 8) +#define CA_TNUM_SHFT 8 +#define CA_DW_DATA_EN (0xffull << 16) +#define CA_DW_DATA_EN_SHFT 16 +#define CA_GFX_CRED (0xffull << 24) +#define CA_GFX_CRED_SHFT 24 +#define CA_MEM_RD_PARAM (0x3ull << 32) +#define CA_MEM_RD_PARAM_SHFT 32 +#define CA_PIO_OP (1ull << 34) +#define CA_CW_ERR (1ull << 35) + /* bits 62:36 unused */ +#define CA_VALID (1ull << 63) + +/* ==== ca_crm_ct_error_detail_2 */ + /* bits 2:0 unused */ +#define CA_PKT_ADDR (0x1fffffffffffffull << 3) +#define CA_PKT_ADDR_SHFT 3 + /* bits 63:56 unused */ + +/* ==== ca_crm_tnumto */ +#define CA_CRM_TNUMTO_VAL (0xffull << 0) +#define CA_CRM_TNUMTO_VAL_SHFT 0 +#define CA_CRM_TNUMTO_WR (1ull << 8) + /* bits 63:9 unused */ + +/* ==== ca_gart_err */ +#define CA_GART_ERR_SOURCE (0x3ull << 0) +#define CA_GART_ERR_SOURCE_SHFT 0 + /* bits 3:2 unused */ +#define CA_GART_ERR_ADDR (0xfffffffffull << 4) +#define CA_GART_ERR_ADDR_SHFT 4 + /* bits 63:40 unused */ + +/* ==== ca_pcierr_type */ +#define CA_PCIERR_DATA (0xffffffffull << 0) +#define CA_PCIERR_DATA_SHFT 0 +#define CA_PCIERR_ENB (0xfull << 32) +#define CA_PCIERR_ENB_SHFT 32 +#define CA_PCIERR_CMD (0xfull << 36) +#define CA_PCIERR_CMD_SHFT 36 +#define CA_PCIERR_A64 (1ull << 40) +#define CA_PCIERR_SLV_SERR (1ull << 41) +#define CA_PCIERR_SLV_WR_PERR (1ull << 42) +#define CA_PCIERR_SLV_RD_PERR (1ull << 43) +#define CA_PCIERR_MST_SERR (1ull << 44) +#define CA_PCIERR_MST_WR_PERR (1ull << 45) +#define CA_PCIERR_MST_RD_PERR (1ull << 46) +#define CA_PCIERR_MST_MABT (1ull << 47) +#define CA_PCIERR_MST_TABT (1ull << 48) +#define CA_PCIERR_MST_RETRY_TOUT (1ull << 49) + +#define CA_PCIERR_TYPES \ + (CA_PCIERR_A64|CA_PCIERR_SLV_SERR| \ + CA_PCIERR_SLV_WR_PERR|CA_PCIERR_SLV_RD_PERR| \ + CA_PCIERR_MST_SERR|CA_PCIERR_MST_WR_PERR|CA_PCIERR_MST_RD_PERR| \ + CA_PCIERR_MST_MABT|CA_PCIERR_MST_TABT|CA_PCIERR_MST_RETRY_TOUT) + + /* bits 63:50 unused */ + +/* ==== ca_pci_dma_addr_extn */ +#define CA_UPPER_NODE_OFFSET (0x3full << 0) +#define CA_UPPER_NODE_OFFSET_SHFT 0 + /* bits 7:6 unused */ +#define CA_CHIPLET_ID (0x3ull << 8) +#define CA_CHIPLET_ID_SHFT 8 + /* bits 11:10 unused */ +#define CA_PCI_DMA_NODE_ID (0xffffull << 12) +#define CA_PCI_DMA_NODE_ID_SHFT 12 + /* bits 27:26 unused */ +#define CA_PCI_DMA_PIO_MEM_TYPE (1ull << 28) + /* bits 63:29 unused */ + + +/* ==== ca_agp_dma_addr_extn */ + /* bits 19:0 unused */ +#define CA_AGP_DMA_NODE_ID (0xffffull << 20) +#define CA_AGP_DMA_NODE_ID_SHFT 20 + /* bits 27:26 unused */ +#define CA_AGP_DMA_PIO_MEM_TYPE (1ull << 28) + /* bits 63:29 unused */ + +/* ==== ca_debug_vector_sel */ +#define CA_DEBUG_MN_VSEL (0xfull << 0) +#define CA_DEBUG_MN_VSEL_SHFT 0 +#define CA_DEBUG_PP_VSEL (0xfull << 4) +#define CA_DEBUG_PP_VSEL_SHFT 4 +#define CA_DEBUG_GW_VSEL (0xfull << 8) +#define CA_DEBUG_GW_VSEL_SHFT 8 +#define CA_DEBUG_GT_VSEL (0xfull << 12) +#define CA_DEBUG_GT_VSEL_SHFT 12 +#define CA_DEBUG_PD_VSEL (0xfull << 16) +#define CA_DEBUG_PD_VSEL_SHFT 16 +#define CA_DEBUG_AD_VSEL (0xfull << 20) +#define CA_DEBUG_AD_VSEL_SHFT 20 +#define CA_DEBUG_CX_VSEL (0xfull << 24) +#define CA_DEBUG_CX_VSEL_SHFT 24 +#define CA_DEBUG_CR_VSEL (0xfull << 28) +#define CA_DEBUG_CR_VSEL_SHFT 28 +#define CA_DEBUG_BA_VSEL (0xfull << 32) +#define CA_DEBUG_BA_VSEL_SHFT 32 +#define CA_DEBUG_PE_VSEL (0xfull << 36) +#define CA_DEBUG_PE_VSEL_SHFT 36 +#define CA_DEBUG_BO_VSEL (0xfull << 40) +#define CA_DEBUG_BO_VSEL_SHFT 40 +#define CA_DEBUG_BI_VSEL (0xfull << 44) +#define CA_DEBUG_BI_VSEL_SHFT 44 +#define CA_DEBUG_AS_VSEL (0xfull << 48) +#define CA_DEBUG_AS_VSEL_SHFT 48 +#define CA_DEBUG_PS_VSEL (0xfull << 52) +#define CA_DEBUG_PS_VSEL_SHFT 52 +#define CA_DEBUG_PM_VSEL (0xfull << 56) +#define CA_DEBUG_PM_VSEL_SHFT 56 + /* bits 63:60 unused */ + +/* ==== ca_debug_mux_core_sel */ +/* ==== ca_debug_mux_pci_sel */ +#define CA_DEBUG_MSEL0 (0x7ull << 0) +#define CA_DEBUG_MSEL0_SHFT 0 + /* bit 3 unused */ +#define CA_DEBUG_NSEL0 (0x7ull << 4) +#define CA_DEBUG_NSEL0_SHFT 4 + /* bit 7 unused */ +#define CA_DEBUG_MSEL1 (0x7ull << 8) +#define CA_DEBUG_MSEL1_SHFT 8 + /* bit 11 unused */ +#define CA_DEBUG_NSEL1 (0x7ull << 12) +#define CA_DEBUG_NSEL1_SHFT 12 + /* bit 15 unused */ +#define CA_DEBUG_MSEL2 (0x7ull << 16) +#define CA_DEBUG_MSEL2_SHFT 16 + /* bit 19 unused */ +#define CA_DEBUG_NSEL2 (0x7ull << 20) +#define CA_DEBUG_NSEL2_SHFT 20 + /* bit 23 unused */ +#define CA_DEBUG_MSEL3 (0x7ull << 24) +#define CA_DEBUG_MSEL3_SHFT 24 + /* bit 27 unused */ +#define CA_DEBUG_NSEL3 (0x7ull << 28) +#define CA_DEBUG_NSEL3_SHFT 28 + /* bit 31 unused */ +#define CA_DEBUG_MSEL4 (0x7ull << 32) +#define CA_DEBUG_MSEL4_SHFT 32 + /* bit 35 unused */ +#define CA_DEBUG_NSEL4 (0x7ull << 36) +#define CA_DEBUG_NSEL4_SHFT 36 + /* bit 39 unused */ +#define CA_DEBUG_MSEL5 (0x7ull << 40) +#define CA_DEBUG_MSEL5_SHFT 40 + /* bit 43 unused */ +#define CA_DEBUG_NSEL5 (0x7ull << 44) +#define CA_DEBUG_NSEL5_SHFT 44 + /* bit 47 unused */ +#define CA_DEBUG_MSEL6 (0x7ull << 48) +#define CA_DEBUG_MSEL6_SHFT 48 + /* bit 51 unused */ +#define CA_DEBUG_NSEL6 (0x7ull << 52) +#define CA_DEBUG_NSEL6_SHFT 52 + /* bit 55 unused */ +#define CA_DEBUG_MSEL7 (0x7ull << 56) +#define CA_DEBUG_MSEL7_SHFT 56 + /* bit 59 unused */ +#define CA_DEBUG_NSEL7 (0x7ull << 60) +#define CA_DEBUG_NSEL7_SHFT 60 + /* bit 63 unused */ + + +/* ==== ca_debug_domain_sel */ +#define CA_DEBUG_DOMAIN_L (1ull << 0) +#define CA_DEBUG_DOMAIN_H (1ull << 1) + /* bits 63:2 unused */ + +/* ==== ca_gart_ptr_table */ +#define CA_GART_PTR_VAL (1ull << 0) + /* bits 11:1 unused */ +#define CA_GART_PTR_ADDR (0xfffffffffffull << 12) +#define CA_GART_PTR_ADDR_SHFT 12 + /* bits 63:56 unused */ + +/* ==== ca_gart_tlb_addr[0-7] */ +#define CA_GART_TLB_ADDR (0xffffffffffffffull << 0) +#define CA_GART_TLB_ADDR_SHFT 0 + /* bits 62:56 unused */ +#define CA_GART_TLB_ENTRY_VAL (1ull << 63) + +/* + * PIO address space ranges for TIO:CA + */ + +/* CA internal registers */ +#define CA_PIO_ADMIN 0x00000000 +#define CA_PIO_ADMIN_LEN 0x00010000 + +/* GFX Write Buffer - Diagnostics */ +#define CA_PIO_GFX 0x00010000 +#define CA_PIO_GFX_LEN 0x00010000 + +/* AGP DMA Write Buffer - Diagnostics */ +#define CA_PIO_AGP_DMAWRITE 0x00020000 +#define CA_PIO_AGP_DMAWRITE_LEN 0x00010000 + +/* AGP DMA READ Buffer - Diagnostics */ +#define CA_PIO_AGP_DMAREAD 0x00030000 +#define CA_PIO_AGP_DMAREAD_LEN 0x00010000 + +/* PCI Config Type 0 */ +#define CA_PIO_PCI_TYPE0_CONFIG 0x01000000 +#define CA_PIO_PCI_TYPE0_CONFIG_LEN 0x01000000 + +/* PCI Config Type 1 */ +#define CA_PIO_PCI_TYPE1_CONFIG 0x02000000 +#define CA_PIO_PCI_TYPE1_CONFIG_LEN 0x01000000 + +/* PCI I/O Cycles - mapped to PCI Address 0x00000000-0x04ffffff */ +#define CA_PIO_PCI_IO 0x03000000 +#define CA_PIO_PCI_IO_LEN 0x05000000 + +/* PCI MEM Cycles - mapped to PCI with CA_PIO_ADDR_OFFSET of ca_control1 */ +/* use Fast Write if enabled and coretalk packet type is a GFX request */ +#define CA_PIO_PCI_MEM_OFFSET 0x08000000 +#define CA_PIO_PCI_MEM_OFFSET_LEN 0x08000000 + +/* PCI MEM Cycles - mapped to PCI Address 0x00000000-0xbfffffff */ +/* use Fast Write if enabled and coretalk packet type is a GFX request */ +#define CA_PIO_PCI_MEM 0x40000000 +#define CA_PIO_PCI_MEM_LEN 0xc0000000 + +/* + * DMA space + * + * The CA aperature (ie. bus address range) mapped by the GART is segmented into + * two parts. The lower portion of the aperature is used for mapping 32 bit + * PCI addresses which are managed by the dma interfaces in this file. The + * upper poprtion of the aperature is used for mapping 48 bit AGP addresses. + * The AGP portion of the aperature is managed by the agpgart_be.c driver + * in drivers/linux/agp. There are ca-specific hooks in that driver to + * manipulate the gart, but management of the AGP portion of the aperature + * is the responsibility of that driver. + * + * CA allows three main types of DMA mapping: + * + * PCI 64-bit Managed by this driver + * PCI 32-bit Managed by this driver + * AGP 48-bit Managed by hooks in the /dev/agpgart driver + * + * All of the above can optionally be remapped through the GART. The following + * table lists the combinations of addressing types and GART remapping that + * is currently supported by the driver (h/w supports all, s/w limits this): + * + * PCI64 PCI32 AGP48 + * GART no yes yes + * Direct yes yes no + * + * GART remapping of PCI64 is not done because there is no need to. The + * 64 bit PCI address holds all of the information necessary to target any + * memory in the system. + * + * AGP48 is always mapped through the GART. Management of the AGP48 portion + * of the aperature is the responsibility of code in the agpgart_be driver. + * + * The non-64 bit bus address space will currently be partitioned like this: + * + * 0xffff_ffff_ffff +-------- + * | AGP48 direct + * | Space managed by this driver + * CA_AGP_DIRECT_BASE +-------- + * | AGP GART mapped (gfx aperature) + * | Space managed by /dev/agpgart driver + * | This range is exposed to the agpgart + * | driver as the "graphics aperature" + * CA_AGP_MAPPED_BASE +----- + * | PCI GART mapped + * | Space managed by this driver + * CA_PCI32_MAPPED_BASE +---- + * | PCI32 direct + * | Space managed by this driver + * 0xC000_0000 +-------- + * (CA_PCI32_DIRECT_BASE) + * + * The bus address range CA_PCI32_MAPPED_BASE through CA_AGP_DIRECT_BASE + * is what we call the CA aperature. Addresses falling in this range will + * be remapped using the GART. + * + * The bus address range CA_AGP_MAPPED_BASE through CA_AGP_DIRECT_BASE + * is what we call the graphics aperature. This is a subset of the CA + * aperature and is under the control of the agpgart_be driver. + * + * CA_PCI32_MAPPED_BASE, CA_AGP_MAPPED_BASE, and CA_AGP_DIRECT_BASE are + * somewhat arbitrary values. The known constraints on choosing these is: + * + * 1) CA_AGP_DIRECT_BASE-CA_PCI32_MAPPED_BASE+1 (the CA aperature size) + * must be one of the values supported by the ca_gart_aperature register. + * Currently valid values are: 4MB through 4096MB in powers of 2 increments + * + * 2) CA_AGP_DIRECT_BASE-CA_AGP_MAPPED_BASE+1 (the gfx aperature size) + * must be in MB units since that's what the agpgart driver assumes. + */ + +/* + * Define Bus DMA ranges. These are configurable (see constraints above) + * and will probably need tuning based on experience. + */ + + +/* + * 11/24/03 + * CA has an addressing glitch w.r.t. PCI direct 32 bit DMA that makes it + * generally unusable. The problem is that for PCI direct 32 + * DMA's, all 32 bits of the bus address are used to form the lower 32 bits + * of the coretalk address, and coretalk bits 38:32 come from a register. + * Since only PCI bus addresses 0xC0000000-0xFFFFFFFF (1GB) are available + * for DMA (the rest is allocated to PIO), host node addresses need to be + * such that their lower 32 bits fall in the 0xC0000000-0xffffffff range + * as well. So there can be no PCI32 direct DMA below 3GB!! For this + * reason we set the CA_PCI32_DIRECT_SIZE to 0 which essentially makes + * tioca_dma_direct32() a noop but preserves the code flow should this issue + * be fixed in a respin. + * + * For now, all PCI32 DMA's must be mapped through the GART. + */ + +#define CA_PCI32_DIRECT_BASE 0xC0000000UL /* BASE not configurable */ +#define CA_PCI32_DIRECT_SIZE 0x00000000UL /* 0 MB */ + +#define CA_PCI32_MAPPED_BASE 0xC0000000UL +#define CA_PCI32_MAPPED_SIZE 0x40000000UL /* 2GB */ + +#define CA_AGP_MAPPED_BASE 0x80000000UL +#define CA_AGP_MAPPED_SIZE 0x40000000UL /* 2GB */ + +#define CA_AGP_DIRECT_BASE 0x40000000UL /* 2GB */ +#define CA_AGP_DIRECT_SIZE 0x40000000UL + +#define CA_APERATURE_BASE (CA_AGP_MAPPED_BASE) +#define CA_APERATURE_SIZE (CA_AGP_MAPPED_SIZE+CA_PCI32_MAPPED_SIZE) + +#endif /* _ASM_IA64_SN_TIO_TIOCA_H */ diff --git a/include/asm-ia64/sn/tioca_provider.h b/include/asm-ia64/sn/tioca_provider.h new file mode 100644 index 00000000000..b6acc22ab23 --- /dev/null +++ b/include/asm-ia64/sn/tioca_provider.h @@ -0,0 +1,206 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2003-2005 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef _ASM_IA64_SN_TIO_CA_AGP_PROVIDER_H +#define _ASM_IA64_SN_TIO_CA_AGP_PROVIDER_H + +#include + +/* + * WAR enables + * Defines for individual WARs. Each is a bitmask of applicable + * part revision numbers. (1 << 1) == rev A, (1 << 2) == rev B, + * (3 << 1) == (rev A or rev B), etc + */ + +#define TIOCA_WAR_ENABLED(pv, tioca_common) \ + ((1 << tioca_common->ca_rev) & pv) + + /* TIO:ICE:FRZ:Freezer loses a PIO data ucred on PIO RD RSP with CW error */ +#define PV907908 (1 << 1) + /* ATI config space problems after BIOS execution starts */ +#define PV908234 (1 << 1) + /* CA:AGPDMA write request data mismatch with ABC1CL merge */ +#define PV895469 (1 << 1) + /* TIO:CA TLB invalidate of written GART entries possibly not occuring in CA*/ +#define PV910244 (1 << 1) + +struct tioca_dmamap{ + struct list_head cad_list; /* headed by ca_list */ + + dma_addr_t cad_dma_addr; /* Linux dma handle */ + uint cad_gart_entry; /* start entry in ca_gart_pagemap */ + uint cad_gart_size; /* #entries for this map */ +}; + +/* + * Kernel only fields. Prom may look at this stuff for debugging only. + * Access this structure through the ca_kernel_private ptr. + */ + +struct tioca_common ; + +struct tioca_kernel { + struct tioca_common *ca_common; /* tioca this belongs to */ + struct list_head ca_list; /* list of all ca's */ + struct list_head ca_dmamaps; + spinlock_t ca_lock; /* Kernel lock */ + cnodeid_t ca_closest_node; + struct list_head *ca_devices; /* bus->devices */ + + /* + * General GART stuff + */ + uint64_t ca_ap_size; /* size of aperature in bytes */ + uint32_t ca_gart_entries; /* # uint64_t entries in gart */ + uint32_t ca_ap_pagesize; /* aperature page size in bytes */ + uint64_t ca_ap_bus_base; /* bus address of CA aperature */ + uint64_t ca_gart_size; /* gart size in bytes */ + uint64_t *ca_gart; /* gart table vaddr */ + uint64_t ca_gart_coretalk_addr; /* gart coretalk addr */ + uint8_t ca_gart_iscoherent; /* used in tioca_tlbflush */ + + /* PCI GART convenience values */ + uint64_t ca_pciap_base; /* pci aperature bus base address */ + uint64_t ca_pciap_size; /* pci aperature size (bytes) */ + uint64_t ca_pcigart_base; /* gfx GART bus base address */ + uint64_t *ca_pcigart; /* gfx GART vm address */ + uint32_t ca_pcigart_entries; + uint32_t ca_pcigart_start; /* PCI start index in ca_gart */ + void *ca_pcigart_pagemap; + + /* AGP GART convenience values */ + uint64_t ca_gfxap_base; /* gfx aperature bus base address */ + uint64_t ca_gfxap_size; /* gfx aperature size (bytes) */ + uint64_t ca_gfxgart_base; /* gfx GART bus base address */ + uint64_t *ca_gfxgart; /* gfx GART vm address */ + uint32_t ca_gfxgart_entries; + uint32_t ca_gfxgart_start; /* agpgart start index in ca_gart */ +}; + +/* + * Common tioca info shared between kernel and prom + * + * DO NOT CHANGE THIS STRUCT WITHOUT MAKING CORRESPONDING CHANGES + * TO THE PROM VERSION. + */ + +struct tioca_common { + struct pcibus_bussoft ca_common; /* common pciio header */ + + uint32_t ca_rev; + uint32_t ca_closest_nasid; + + uint64_t ca_prom_private; + uint64_t ca_kernel_private; +}; + +/** + * tioca_paddr_to_gart - Convert an SGI coretalk address to a CA GART entry + * @paddr: page address to convert + * + * Convert a system [coretalk] address to a GART entry. GART entries are + * formed using the following: + * + * data = ( (1<<63) | ( (REMAP_NODE_ID << 40) | (MD_CHIPLET_ID << 38) | + * (REMAP_SYS_ADDR) ) >> 12 ) + * + * DATA written to 1 GART TABLE Entry in system memory is remapped system + * addr for 1 page + * + * The data is for coretalk address format right shifted 12 bits with a + * valid bit. + * + * GART_TABLE_ENTRY [ 25:0 ] -- REMAP_SYS_ADDRESS[37:12]. + * GART_TABLE_ENTRY [ 27:26 ] -- SHUB MD chiplet id. + * GART_TABLE_ENTRY [ 41:28 ] -- REMAP_NODE_ID. + * GART_TABLE_ENTRY [ 63 ] -- Valid Bit + */ +static inline u64 +tioca_paddr_to_gart(unsigned long paddr) +{ + /* + * We are assuming right now that paddr already has the correct + * format since the address from xtalk_dmaXXX should already have + * NODE_ID, CHIPLET_ID, and SYS_ADDR in the correct locations. + */ + + return ((paddr) >> 12) | (1UL << 63); +} + +/** + * tioca_physpage_to_gart - Map a host physical page for SGI CA based DMA + * @page_addr: system page address to map + */ + +static inline unsigned long +tioca_physpage_to_gart(uint64_t page_addr) +{ + uint64_t coretalk_addr; + + coretalk_addr = PHYS_TO_TIODMA(page_addr); + if (!coretalk_addr) { + return 0; + } + + return tioca_paddr_to_gart(coretalk_addr); +} + +/** + * tioca_tlbflush - invalidate cached SGI CA GART TLB entries + * @tioca_kernel: CA context + * + * Invalidate tlb entries for a given CA GART. Main complexity is to account + * for revA bug. + */ +static inline void +tioca_tlbflush(struct tioca_kernel *tioca_kernel) +{ + volatile uint64_t tmp; + volatile struct tioca *ca_base; + struct tioca_common *tioca_common; + + tioca_common = tioca_kernel->ca_common; + ca_base = (struct tioca *)tioca_common->ca_common.bs_base; + + /* + * Explicit flushes not needed if GART is in cached mode + */ + if (tioca_kernel->ca_gart_iscoherent) { + if (TIOCA_WAR_ENABLED(PV910244, tioca_common)) { + /* + * PV910244: RevA CA needs explicit flushes. + * Need to put GART into uncached mode before + * flushing otherwise the explicit flush is ignored. + * + * Alternate WAR would be to leave GART cached and + * touch every CL aligned GART entry. + */ + + ca_base->ca_control2 &= ~(CA_GART_MEM_PARAM); + ca_base->ca_control2 |= CA_GART_FLUSH_TLB; + ca_base->ca_control2 |= + (0x2ull << CA_GART_MEM_PARAM_SHFT); + tmp = ca_base->ca_control2; + } + + return; + } + + /* + * Gart in uncached mode ... need an explicit flush. + */ + + ca_base->ca_control2 |= CA_GART_FLUSH_TLB; + tmp = ca_base->ca_control2; +} + +extern uint32_t tioca_gart_found; +extern int tioca_init_provider(void); +extern void tioca_fastwrite_enable(struct tioca_kernel *tioca_kern); +#endif /* _ASM_IA64_SN_TIO_CA_AGP_PROVIDER_H */ -- cgit v1.2.3-70-g09d2 From 8297511530b0d2a281c796e738683951a59a020c Mon Sep 17 00:00:00 2001 From: David Mosberger-Tang Date: Mon, 25 Apr 2005 11:44:02 -0700 Subject: [IA64] add missing cpu_relax() in ITC syncing code Call cpu_relax() in busy-waiting loops of the ITC-syncing code. Signed-off-by: David Mosberger-Tang Signed-off-by: Tony Luck --- arch/ia64/kernel/smpboot.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c index ca1536db339..dbc6b610cc6 100644 --- a/arch/ia64/kernel/smpboot.c +++ b/arch/ia64/kernel/smpboot.c @@ -1,7 +1,7 @@ /* * SMP boot-related support * - * Copyright (C) 1998-2003 Hewlett-Packard Co + * Copyright (C) 1998-2003, 2005 Hewlett-Packard Co * David Mosberger-Tang * * 01/05/16 Rohit Seth Moved SMP booting functions from smp.c to here. @@ -156,7 +156,8 @@ sync_master (void *arg) local_irq_save(flags); { for (i = 0; i < NUM_ROUNDS*NUM_ITERS; ++i) { - while (!go[MASTER]); + while (!go[MASTER]) + cpu_relax(); go[MASTER] = 0; go[SLAVE] = ia64_get_itc(); } @@ -179,7 +180,8 @@ get_delta (long *rt, long *master) for (i = 0; i < NUM_ITERS; ++i) { t0 = ia64_get_itc(); go[MASTER] = 1; - while (!(tm = go[SLAVE])); + while (!(tm = go[SLAVE])) + cpu_relax(); go[SLAVE] = 0; t1 = ia64_get_itc(); @@ -258,7 +260,8 @@ ia64_sync_itc (unsigned int master) return; } - while (go[MASTER]); /* wait for master to be ready */ + while (go[MASTER]) + cpu_relax(); /* wait for master to be ready */ spin_lock_irqsave(&itc_sync_lock, flags); { -- cgit v1.2.3-70-g09d2 From e8d1cb2f280aa53e1c75c8b5fcbf80b3481d0caa Mon Sep 17 00:00:00 2001 From: Keith Owens Date: Mon, 25 Apr 2005 11:45:26 -0700 Subject: [IA64] Tighten up unw_unwind_to_user check Detect user space by the unwind frame with predicate PRED_USER_STACK set, instead of a user space IP. Tighten up the last ditch check for running off the top of the kernel stack. Based on a suggestion by David Mosberger, reworked to fit the current tree. This survives my stress test which used to break 2.6.9 kernels. Unlike 2.6.11, the stress test now unwinds to the correct point, so gdb can get the user space registers. Signed-off-by: Keith Owens Signed-off-by: Tony Luck --- arch/ia64/kernel/unwind.c | 27 +++++++++++++++++---------- 1 file changed, 17 insertions(+), 10 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/unwind.c b/arch/ia64/kernel/unwind.c index d494ff647ca..2776a074c6f 100644 --- a/arch/ia64/kernel/unwind.c +++ b/arch/ia64/kernel/unwind.c @@ -1943,23 +1943,30 @@ EXPORT_SYMBOL(unw_unwind); int unw_unwind_to_user (struct unw_frame_info *info) { - unsigned long ip, sp; + unsigned long ip, sp, pr = 0; while (unw_unwind(info) >= 0) { - if (unw_get_rp(info, &ip) < 0) { - unw_get_ip(info, &ip); - UNW_DPRINT(0, "unwind.%s: failed to read return pointer (ip=0x%lx)\n", - __FUNCTION__, ip); - return -1; - } unw_get_sp(info, &sp); - if (sp >= (unsigned long)info->task + IA64_STK_OFFSET) + if ((long)((unsigned long)info->task + IA64_STK_OFFSET - sp) + < IA64_PT_REGS_SIZE) { + UNW_DPRINT(0, "unwind.%s: ran off the top of the kernel stack\n", + __FUNCTION__); break; - if (ip < FIXADDR_USER_END) + } + if (unw_is_intr_frame(info) && + (pr & (1UL << PRED_USER_STACK))) return 0; + if (unw_get_pr (info, &pr) < 0) { + unw_get_rp(info, &ip); + UNW_DPRINT(0, "unwind.%s: failed to read " + "predicate register (ip=0x%lx)\n", + __FUNCTION__, ip); + return -1; + } } unw_get_ip(info, &ip); - UNW_DPRINT(0, "unwind.%s: failed to unwind to user-level (ip=0x%lx)\n", __FUNCTION__, ip); + UNW_DPRINT(0, "unwind.%s: failed to unwind to user-level (ip=0x%lx)\n", + __FUNCTION__, ip); return -1; } EXPORT_SYMBOL(unw_unwind_to_user); -- cgit v1.2.3-70-g09d2 From 30325d17715302a60c9afdaacaafaeb056b7e880 Mon Sep 17 00:00:00 2001 From: David Mosberger-Tang Date: Mon, 25 Apr 2005 13:03:16 -0700 Subject: [IA64] speed up syscall path a bit more Recently I noticed that clearing ar.ssd/ar.csd right before srlz.d is causing significant stalling in the syscall path. The patch below fixes that by moving the register-writes after srlz.d. On a Madison, this drops break-based getpid() from 241 to 226 cycles (-15 cycles). Signed-off-by: David Mosberger-Tang Signed-off-by: Tony Luck --- arch/ia64/kernel/entry.S | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S index 0272c010a3b..73e23dafe8e 100644 --- a/arch/ia64/kernel/entry.S +++ b/arch/ia64/kernel/entry.S @@ -728,12 +728,8 @@ ENTRY(ia64_leave_syscall) mov f8=f0 // clear f8 ;; ld8 r30=[r2],16 // M0|1 load cr.ifs - mov.m ar.ssd=r0 // M2 clear ar.ssd - cmp.eq p9,p0=r0,r0 // set p9 to indicate that we should restore cr.ifs - ;; ld8 r25=[r3],16 // M0|1 load ar.unat - mov.m ar.csd=r0 // M2 clear ar.csd - mov r22=r0 // clear r22 + cmp.eq p9,p0=r0,r0 // set p9 to indicate that we should restore cr.ifs ;; ld8 r26=[r2],PT(B0)-PT(AR_PFS) // M0|1 load ar.pfs (pKStk) mov r22=psr // M2 read PSR now that interrupts are disabled @@ -756,11 +752,15 @@ ENTRY(ia64_leave_syscall) mov f7=f0 // clear f7 ;; ld8.fill r12=[r2] // restore r12 (sp) + mov.m ar.ssd=r0 // M2 clear ar.ssd + mov r22=r0 // clear r22 + ld8.fill r15=[r3] // restore r15 +(pUStk) st1 [r14]=r17 addl r3=THIS_CPU(ia64_phys_stacked_size_p8),r0 ;; (pUStk) ld4 r3=[r3] // r3 = cpu_data->phys_stacked_size_p8 -(pUStk) st1 [r14]=r17 + mov.m ar.csd=r0 // M2 clear ar.csd mov b6=r18 // I0 restore b6 ;; mov r14=r0 // clear r14 -- cgit v1.2.3-70-g09d2 From 4a5c13c7eb0d55bfd2cf3100c55f1e3d8df37576 Mon Sep 17 00:00:00 2001 From: Mark Goodwin Date: Mon, 25 Apr 2005 13:04:22 -0700 Subject: [IA64-SGI] Altix SN topology support for new chipsets and pci topology please accept this patch to the Altix SN platform topology export interface to support new chipsets and to export PCI topology. This follows on top of Jack Steiner's patch dated March 1st ("New chipset support for SN platform"). Signed-off-by: Mark Goodwin Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/sn2/sn_hwperf.c | 133 +++++++++++++++++++++++++++++++++--- include/asm-ia64/sn/sn_sal.h | 11 +++ 2 files changed, 135 insertions(+), 9 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/sn2/sn_hwperf.c b/arch/ia64/sn/kernel/sn2/sn_hwperf.c index 197356460ee..3bff99130d5 100644 --- a/arch/ia64/sn/kernel/sn2/sn_hwperf.c +++ b/arch/ia64/sn/kernel/sn2/sn_hwperf.c @@ -28,6 +28,7 @@ #include #include #include +#include #include #include #include @@ -43,6 +44,7 @@ #include #include #include +#include static void *sn_hwperf_salheap = NULL; static int sn_hwperf_obj_cnt = 0; @@ -81,26 +83,45 @@ out: return e; } +static int sn_hwperf_location_to_bpos(char *location, + int *rack, int *bay, int *slot, int *slab) +{ + char type; + + /* first scan for an old style geoid string */ + if (sscanf(location, "%03d%c%02d#%d", + rack, &type, bay, slab) == 4) + *slot = 0; + else /* scan for a new bladed geoid string */ + if (sscanf(location, "%03d%c%02d^%02d#%d", + rack, &type, bay, slot, slab) != 5) + return -1; + /* success */ + return 0; +} + static int sn_hwperf_geoid_to_cnode(char *location) { int cnode; geoid_t geoid; moduleid_t module_id; - char type; - int rack, slot, slab; - int this_rack, this_slot, this_slab; + int rack, bay, slot, slab; + int this_rack, this_bay, this_slot, this_slab; - if (sscanf(location, "%03d%c%02d#%d", &rack, &type, &slot, &slab) != 4) + if (sn_hwperf_location_to_bpos(location, &rack, &bay, &slot, &slab)) return -1; for (cnode = 0; cnode < numionodes; cnode++) { geoid = cnodeid_get_geoid(cnode); module_id = geo_module(geoid); this_rack = MODULE_GET_RACK(module_id); - this_slot = MODULE_GET_BPOS(module_id); + this_bay = MODULE_GET_BPOS(module_id); + this_slot = 0; /* XXX */ this_slab = geo_slab(geoid); - if (rack == this_rack && slot == this_slot && slab == this_slab) + if (rack == this_rack && bay == this_bay && + slot == this_slot && slab == this_slab) { break; + } } return cnode < numionodes ? cnode : -1; @@ -153,11 +174,29 @@ static const char *sn_hwperf_get_slabname(struct sn_hwperf_object_info *obj, return slabname; } +static void print_pci_topology(struct seq_file *s, + struct sn_hwperf_object_info *obj, int *ordinal, + char *pci_topo_buf, int len) +{ + char *p1; + char *p2; + + for (p1=pci_topo_buf; *p1 && p1 < pci_topo_buf + len;) { + if (!(p2 = strchr(p1, '\n'))) + break; + *p2 = '\0'; + seq_printf(s, "pcibus %d %s-%s\n", + *ordinal, obj->location, p1); + (*ordinal)++; + p1 = p2 + 1; + } +} + static int sn_topology_show(struct seq_file *s, void *d) { int sz; int pt; - int e; + int e = 0; int i; int j; const char *slabname; @@ -169,11 +208,46 @@ static int sn_topology_show(struct seq_file *s, void *d) struct sn_hwperf_object_info *p; struct sn_hwperf_object_info *obj = d; /* this object */ struct sn_hwperf_object_info *objs = s->private; /* all objects */ + int rack, bay, slot, slab; + u8 shubtype; + u8 system_size; + u8 sharing_size; + u8 partid; + u8 coher; + u8 nasid_shift; + u8 region_size; + u16 nasid_mask; + int nasid_msb; + char *pci_topo_buf; + int pci_bus_ordinal = 0; + static int pci_topo_buf_len = 256; if (obj == objs) { - seq_printf(s, "# sn_topology version 1\n"); + seq_printf(s, "# sn_topology version 2\n"); seq_printf(s, "# objtype ordinal location partition" " [attribute value [, ...]]\n"); + + if (ia64_sn_get_sn_info(0, + &shubtype, &nasid_mask, &nasid_shift, &system_size, + &sharing_size, &partid, &coher, ®ion_size)) + BUG(); + for (nasid_msb=63; nasid_msb > 0; nasid_msb--) { + if (((u64)nasid_mask << nasid_shift) & (1ULL << nasid_msb)) + break; + } + seq_printf(s, "partition %u %s local " + "shubtype %s, " + "nasid_mask 0x%016lx, " + "nasid_bits %d:%d, " + "system_size %d, " + "sharing_size %d, " + "coherency_domain %d, " + "region_size %d\n", + + partid, system_utsname.nodename, + shubtype ? "shub2" : "shub1", + (u64)nasid_mask << nasid_shift, nasid_msb, nasid_shift, + system_size, sharing_size, coher, region_size); } if (SN_HWPERF_FOREIGN(obj)) { @@ -181,7 +255,7 @@ static int sn_topology_show(struct seq_file *s, void *d) return 0; } - for (i = 0; obj->name[i]; i++) { + for (i = 0; i < SN_HWPERF_MAXSTRING && obj->name[i]; i++) { if (obj->name[i] == ' ') obj->name[i] = '_'; } @@ -221,6 +295,43 @@ static int sn_topology_show(struct seq_file *s, void *d) seq_putc(s, '\n'); } } + + /* + * PCI busses attached to this node, if any + */ + do { + if (!(pci_topo_buf = vmalloc(pci_topo_buf_len))) { + printk("sn_topology_show: kmalloc failed\n"); + break; + } + + if (sn_hwperf_location_to_bpos(obj->location, + &rack, &bay, &slot, &slab) != 0) + continue; + + e = ia64_sn_ioif_get_pci_topology(rack, bay, slot, slab, + pci_topo_buf, pci_topo_buf_len); + + switch (e) { + case SALRET_NOT_IMPLEMENTED: + case SALRET_INVALID_ARG: + /* ignore, don't print anything */ + e = SN_HWPERF_OP_OK; + break; + + case SALRET_ERROR: + /* retry with a bigger buffer */ + pci_topo_buf_len += 256; + break; + + case SN_HWPERF_OP_OK: + /* export pci bus info */ + print_pci_topology(s, obj, &pci_bus_ordinal, + pci_topo_buf, pci_topo_buf_len); + break; + } + vfree(pci_topo_buf); + } while (e != SN_HWPERF_OP_OK && pci_topo_buf_len < 0x200000); } if (obj->ports) { @@ -397,6 +508,9 @@ static int sn_hwperf_map_err(int hwperf_err) break; case SN_HWPERF_OP_BUSY: + e = -EBUSY; + break; + case SN_HWPERF_OP_RECONFIGURE: e = -EAGAIN; break; @@ -549,6 +663,7 @@ sn_hwperf_ioctl(struct inode *in, struct file *fp, u32 op, u64 arg) r = sn_hwperf_op_cpu(&op_info); if (r) { r = sn_hwperf_map_err(r); + a.v0 = v0; goto error; } break; diff --git a/include/asm-ia64/sn/sn_sal.h b/include/asm-ia64/sn/sn_sal.h index 81a1cf1e4f5..410d356b40d 100644 --- a/include/asm-ia64/sn/sn_sal.h +++ b/include/asm-ia64/sn/sn_sal.h @@ -74,6 +74,7 @@ #define SN_SAL_IOIF_GET_PCIBUS_INFO 0x02000056 #define SN_SAL_IOIF_GET_PCIDEV_INFO 0x02000057 #define SN_SAL_IOIF_GET_WIDGET_DMAFLUSH_LIST 0x02000058 +#define SN_SAL_IOIF_GET_PCI_TOPOLOGY 0x02000059 #define SN_SAL_HUB_ERROR_INTERRUPT 0x02000060 @@ -1012,4 +1013,14 @@ ia64_sn_hwperf_op(nasid_t nasid, u64 opcode, u64 a0, u64 a1, u64 a2, return (int) rv.status; } +static inline int +ia64_sn_ioif_get_pci_topology(u64 rack, u64 bay, u64 slot, u64 slab, + char *buf, u64 len) +{ + struct ia64_sal_retval rv; + SAL_CALL_NOLOCK(rv, SN_SAL_IOIF_GET_PCI_TOPOLOGY, + rack, bay, slot, slab, buf, len, 0); + return (int) rv.status; +} + #endif /* _ASM_IA64_SN_SN_SAL_H */ -- cgit v1.2.3-70-g09d2 From f1e2a1c8a1fe16db5f4a7c0c1551d6e1b97dcbb2 Mon Sep 17 00:00:00 2001 From: Mark Goodwin Date: Mon, 25 Apr 2005 13:05:08 -0700 Subject: [IA64-SGI] Altix SN topology fix potential infinite loop Fix infinite loop if sn_hwperf_location_to_bpos() fails. Signed-off-by: Mark Goodwin Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/sn2/sn_hwperf.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/sn2/sn_hwperf.c b/arch/ia64/sn/kernel/sn2/sn_hwperf.c index 3bff99130d5..e731fcb95f9 100644 --- a/arch/ia64/sn/kernel/sn2/sn_hwperf.c +++ b/arch/ia64/sn/kernel/sn2/sn_hwperf.c @@ -300,14 +300,15 @@ static int sn_topology_show(struct seq_file *s, void *d) * PCI busses attached to this node, if any */ do { - if (!(pci_topo_buf = vmalloc(pci_topo_buf_len))) { - printk("sn_topology_show: kmalloc failed\n"); + if (sn_hwperf_location_to_bpos(obj->location, + &rack, &bay, &slot, &slab)) { break; } - if (sn_hwperf_location_to_bpos(obj->location, - &rack, &bay, &slot, &slab) != 0) - continue; + if (!(pci_topo_buf = vmalloc(pci_topo_buf_len))) { + printk("sn_topology_show: vmalloc failed\n"); + break; + } e = ia64_sn_ioif_get_pci_topology(rack, bay, slot, slab, pci_topo_buf, pci_topo_buf_len); @@ -325,6 +326,7 @@ static int sn_topology_show(struct seq_file *s, void *d) break; case SN_HWPERF_OP_OK: + default: /* export pci bus info */ print_pci_topology(s, obj, &pci_bus_ordinal, pci_topo_buf, pci_topo_buf_len); -- cgit v1.2.3-70-g09d2 From be539c73b54dcc9f54fb2c2b70e204c93b616c9b Mon Sep 17 00:00:00 2001 From: Colin Ngam Date: Mon, 25 Apr 2005 13:06:28 -0700 Subject: [IA64-SGI] Shub2 provides an addition of 2 External Interrupt events. Signed-off-by: Colin Ngam Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/irq.c | 15 ++------------- include/asm-ia64/sn/shub_mmr.h | 17 +++++++++++++++++ 2 files changed, 19 insertions(+), 13 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/irq.c b/arch/ia64/sn/kernel/irq.c index b52d32975f2..0f4e8138658 100644 --- a/arch/ia64/sn/kernel/irq.c +++ b/arch/ia64/sn/kernel/irq.c @@ -82,20 +82,9 @@ static void sn_ack_irq(unsigned int irq) nasid = get_nasid(); event_occurred = HUB_L((uint64_t *) GLOBAL_MMR_ADDR(nasid, SH_EVENT_OCCURRED)); - if (event_occurred & SH_EVENT_OCCURRED_UART_INT_MASK) { - mask |= (1 << SH_EVENT_OCCURRED_UART_INT_SHFT); - } - if (event_occurred & SH_EVENT_OCCURRED_IPI_INT_MASK) { - mask |= (1 << SH_EVENT_OCCURRED_IPI_INT_SHFT); - } - if (event_occurred & SH_EVENT_OCCURRED_II_INT0_MASK) { - mask |= (1 << SH_EVENT_OCCURRED_II_INT0_SHFT); - } - if (event_occurred & SH_EVENT_OCCURRED_II_INT1_MASK) { - mask |= (1 << SH_EVENT_OCCURRED_II_INT1_SHFT); - } + mask = event_occurred & SH_ALL_INT_MASK; HUB_S((uint64_t *) GLOBAL_MMR_ADDR(nasid, SH_EVENT_OCCURRED_ALIAS), - mask); + mask); __set_bit(irq, (volatile void *)pda->sn_in_service_ivecs); move_irq(irq); diff --git a/include/asm-ia64/sn/shub_mmr.h b/include/asm-ia64/sn/shub_mmr.h index 5c2fcf13d5c..6ec37e816a9 100644 --- a/include/asm-ia64/sn/shub_mmr.h +++ b/include/asm-ia64/sn/shub_mmr.h @@ -129,6 +129,23 @@ #define SH_EVENT_OCCURRED_II_INT1_SHFT 30 #define SH_EVENT_OCCURRED_II_INT1_MASK 0x0000000040000000 +/* SH2_EVENT_OCCURRED_EXTIO_INT2 */ +/* Description: Pending SHUB 2 EXT IO INT2 */ +#define SH2_EVENT_OCCURRED_EXTIO_INT2_SHFT 33 +#define SH2_EVENT_OCCURRED_EXTIO_INT2_MASK 0x0000000200000000 + +/* SH2_EVENT_OCCURRED_EXTIO_INT3 */ +/* Description: Pending SHUB 2 EXT IO INT3 */ +#define SH2_EVENT_OCCURRED_EXTIO_INT3_SHFT 34 +#define SH2_EVENT_OCCURRED_EXTIO_INT3_MASK 0x0000000400000000 + +#define SH_ALL_INT_MASK \ + (SH_EVENT_OCCURRED_UART_INT_MASK | SH_EVENT_OCCURRED_IPI_INT_MASK | \ + SH_EVENT_OCCURRED_II_INT0_MASK | SH_EVENT_OCCURRED_II_INT1_MASK | \ + SH_EVENT_OCCURRED_II_INT1_MASK | SH2_EVENT_OCCURRED_EXTIO_INT2_MASK | \ + SH2_EVENT_OCCURRED_EXTIO_INT3_MASK) + + /* ==================================================================== */ /* LEDS */ /* ==================================================================== */ -- cgit v1.2.3-70-g09d2 From 4944930ab748942e41ea4dc313fcb0946aee3f17 Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Mon, 25 Apr 2005 13:08:30 -0700 Subject: [IA64] perfmon: make pfm_sysctl a global, and other cleanup - make pfm_sysctl a global such that it is possible to enable/disable debug printk in sampling formats using PFM_DEBUG. - remove unused pfm_debug_var variable - fix a bug in pfm_handle_work where an BUG_ON() could be triggered. There is a path where pfm_handle_work() can be called with interrupts enabled, i.e., when TIF_NEED_RESCHED is set. The fix correct the masking and unmasking of interrupts in pfm_handle_work() such that we restore the interrupt mask as it was upon entry. signed-off-by: stephane eranian Signed-off-by: Tony Luck --- arch/ia64/kernel/perfmon.c | 59 ++++++++++++++++----------------- arch/ia64/kernel/perfmon_default_smpl.c | 13 ++------ include/asm-ia64/perfmon.h | 12 +++++++ 3 files changed, 42 insertions(+), 42 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c index 71147be3279..376fcbc3f8d 100644 --- a/arch/ia64/kernel/perfmon.c +++ b/arch/ia64/kernel/perfmon.c @@ -479,14 +479,6 @@ typedef struct { #define PFM_CMD_ARG_MANY -1 /* cannot be zero */ -typedef struct { - int debug; /* turn on/off debugging via syslog */ - int debug_ovfl; /* turn on/off debug printk in overflow handler */ - int fastctxsw; /* turn on/off fast (unsecure) ctxsw */ - int expert_mode; /* turn on/off value checking */ - int debug_pfm_read; -} pfm_sysctl_t; - typedef struct { unsigned long pfm_spurious_ovfl_intr_count; /* keep track of spurious ovfl interrupts */ unsigned long pfm_replay_ovfl_intr_count; /* keep track of replayed ovfl interrupts */ @@ -514,8 +506,8 @@ static LIST_HEAD(pfm_buffer_fmt_list); static pmu_config_t *pmu_conf; /* sysctl() controls */ -static pfm_sysctl_t pfm_sysctl; -int pfm_debug_var; +pfm_sysctl_t pfm_sysctl; +EXPORT_SYMBOL(pfm_sysctl); static ctl_table pfm_ctl_table[]={ {1, "debug", &pfm_sysctl.debug, sizeof(int), 0666, NULL, &proc_dointvec, NULL,}, @@ -1576,7 +1568,7 @@ pfm_read(struct file *filp, char __user *buf, size_t size, loff_t *ppos) goto abort_locked; } - DPRINT(("[%d] fd=%d type=%d\n", current->pid, msg->pfm_gen_msg.msg_ctx_fd, msg->pfm_gen_msg.msg_type)); + DPRINT(("fd=%d type=%d\n", msg->pfm_gen_msg.msg_ctx_fd, msg->pfm_gen_msg.msg_type)); ret = -EFAULT; if(copy_to_user(buf, msg, sizeof(pfm_msg_t)) == 0) ret = sizeof(pfm_msg_t); @@ -3695,8 +3687,6 @@ pfm_debug(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs) pfm_sysctl.debug = m == 0 ? 0 : 1; - pfm_debug_var = pfm_sysctl.debug; - printk(KERN_INFO "perfmon debugging %s (timing reset)\n", pfm_sysctl.debug ? "on" : "off"); if (m == 0) { @@ -4996,13 +4986,21 @@ pfm_context_force_terminate(pfm_context_t *ctx, struct pt_regs *regs) } static int pfm_ovfl_notify_user(pfm_context_t *ctx, unsigned long ovfl_pmds); - + /* + * pfm_handle_work() can be called with interrupts enabled + * (TIF_NEED_RESCHED) or disabled. The down_interruptible + * call may sleep, therefore we must re-enable interrupts + * to avoid deadlocks. It is safe to do so because this function + * is called ONLY when returning to user level (PUStk=1), in which case + * there is no risk of kernel stack overflow due to deep + * interrupt nesting. + */ void pfm_handle_work(void) { pfm_context_t *ctx; struct pt_regs *regs; - unsigned long flags; + unsigned long flags, dummy_flags; unsigned long ovfl_regs; unsigned int reason; int ret; @@ -5039,18 +5037,15 @@ pfm_handle_work(void) //if (CTX_OVFL_NOBLOCK(ctx)) goto skip_blocking; if (reason == PFM_TRAP_REASON_RESET) goto skip_blocking; + /* + * restore interrupt mask to what it was on entry. + * Could be enabled/diasbled. + */ UNPROTECT_CTX(ctx, flags); - /* - * pfm_handle_work() is currently called with interrupts disabled. - * The down_interruptible call may sleep, therefore we - * must re-enable interrupts to avoid deadlocks. It is - * safe to do so because this function is called ONLY - * when returning to user level (PUStk=1), in which case - * there is no risk of kernel stack overflow due to deep - * interrupt nesting. - */ - BUG_ON(flags & IA64_PSR_I); + /* + * force interrupt enable because of down_interruptible() + */ local_irq_enable(); DPRINT(("before block sleeping\n")); @@ -5064,12 +5059,12 @@ pfm_handle_work(void) DPRINT(("after block sleeping ret=%d\n", ret)); /* - * disable interrupts to restore state we had upon entering - * this function + * lock context and mask interrupts again + * We save flags into a dummy because we may have + * altered interrupts mask compared to entry in this + * function. */ - local_irq_disable(); - - PROTECT_CTX(ctx, flags); + PROTECT_CTX(ctx, dummy_flags); /* * we need to read the ovfl_regs only after wake-up @@ -5095,7 +5090,9 @@ skip_blocking: ctx->ctx_ovfl_regs[0] = 0UL; nothing_to_do: - + /* + * restore flags as they were upon entry + */ UNPROTECT_CTX(ctx, flags); } diff --git a/arch/ia64/kernel/perfmon_default_smpl.c b/arch/ia64/kernel/perfmon_default_smpl.c index 965d2900455..344941db0a9 100644 --- a/arch/ia64/kernel/perfmon_default_smpl.c +++ b/arch/ia64/kernel/perfmon_default_smpl.c @@ -20,24 +20,17 @@ MODULE_AUTHOR("Stephane Eranian "); MODULE_DESCRIPTION("perfmon default sampling format"); MODULE_LICENSE("GPL"); -MODULE_PARM(debug, "i"); -MODULE_PARM_DESC(debug, "debug"); - -MODULE_PARM(debug_ovfl, "i"); -MODULE_PARM_DESC(debug_ovfl, "debug ovfl"); - - #define DEFAULT_DEBUG 1 #ifdef DEFAULT_DEBUG #define DPRINT(a) \ do { \ - if (unlikely(debug >0)) { printk("%s.%d: CPU%d ", __FUNCTION__, __LINE__, smp_processor_id()); printk a; } \ + if (unlikely(pfm_sysctl.debug >0)) { printk("%s.%d: CPU%d ", __FUNCTION__, __LINE__, smp_processor_id()); printk a; } \ } while (0) #define DPRINT_ovfl(a) \ do { \ - if (unlikely(debug_ovfl >0)) { printk("%s.%d: CPU%d ", __FUNCTION__, __LINE__, smp_processor_id()); printk a; } \ + if (unlikely(pfm_sysctl.debug > 0 && pfm_sysctl.debug_ovfl >0)) { printk("%s.%d: CPU%d ", __FUNCTION__, __LINE__, smp_processor_id()); printk a; } \ } while (0) #else @@ -45,8 +38,6 @@ MODULE_PARM_DESC(debug_ovfl, "debug ovfl"); #define DPRINT_ovfl(a) #endif -static int debug, debug_ovfl; - static int default_validate(struct task_struct *task, unsigned int flags, int cpu, void *data) { diff --git a/include/asm-ia64/perfmon.h b/include/asm-ia64/perfmon.h index 136c60e6bfc..ed5416c5b1a 100644 --- a/include/asm-ia64/perfmon.h +++ b/include/asm-ia64/perfmon.h @@ -254,6 +254,18 @@ extern int pfm_mod_write_dbrs(struct task_struct *task, void *req, unsigned int #define PFM_CPUINFO_DCR_PP 0x2 /* if set the system wide session has started */ #define PFM_CPUINFO_EXCL_IDLE 0x4 /* the system wide session excludes the idle task */ +/* + * sysctl control structure. visible to sampling formats + */ +typedef struct { + int debug; /* turn on/off debugging via syslog */ + int debug_ovfl; /* turn on/off debug printk in overflow handler */ + int fastctxsw; /* turn on/off fast (unsecure) ctxsw */ + int expert_mode; /* turn on/off value checking */ +} pfm_sysctl_t; +extern pfm_sysctl_t pfm_sysctl; + + #endif /* __KERNEL__ */ #endif /* _ASM_IA64_PERFMON_H */ -- cgit v1.2.3-70-g09d2 From e1e19747ecce9117610b8f3b57a3e95734230319 Mon Sep 17 00:00:00 2001 From: Bruce Losure Date: Mon, 25 Apr 2005 13:09:41 -0700 Subject: [IA64-SGI] Bus driver for the CX port of SGI's TIO chip. This patch is to provide CX port infrastructure for SGI TIO-based h/w. Also a 'core services' driver for SGI FPGA-based h/w. Signed-off-by: Bruce Losure Signed-off-by: Tony Luck --- arch/ia64/configs/sn2_defconfig | 2 + arch/ia64/sn/kernel/Makefile | 1 + drivers/char/Kconfig | 14 + drivers/char/Makefile | 1 + drivers/char/mbcs.c | 849 ++++++++++++++++++++++++++++++++++++++++ drivers/char/mbcs.h | 553 ++++++++++++++++++++++++++ include/asm-ia64/sn/addrs.h | 3 + include/asm-ia64/sn/tiocx.h | 71 ++++ 8 files changed, 1494 insertions(+) create mode 100644 drivers/char/mbcs.c create mode 100644 drivers/char/mbcs.h create mode 100644 include/asm-ia64/sn/tiocx.h (limited to 'arch') diff --git a/arch/ia64/configs/sn2_defconfig b/arch/ia64/configs/sn2_defconfig index bfeb952fe8e..6ff7107fee4 100644 --- a/arch/ia64/configs/sn2_defconfig +++ b/arch/ia64/configs/sn2_defconfig @@ -574,6 +574,8 @@ CONFIG_SERIAL_NONSTANDARD=y # CONFIG_N_HDLC is not set # CONFIG_STALDRV is not set CONFIG_SGI_SNSC=y +CONFIG_SGI_TIOCX=y +CONFIG_SGI_MBCS=m # # Serial drivers diff --git a/arch/ia64/sn/kernel/Makefile b/arch/ia64/sn/kernel/Makefile index 6c7f4d9e8ea..4f381fb2504 100644 --- a/arch/ia64/sn/kernel/Makefile +++ b/arch/ia64/sn/kernel/Makefile @@ -10,3 +10,4 @@ obj-y += setup.o bte.o bte_error.o irq.o mca.o idle.o \ huberror.o io_init.o iomv.o klconflib.o sn2/ obj-$(CONFIG_IA64_GENERIC) += machvec.o +obj-$(CONFIG_SGI_TIOCX) += tiocx.o diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 096a1202ea0..97ac4edf465 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -399,6 +399,20 @@ config SGI_SNSC controller communication from user space (you want this!), say Y. Otherwise, say N. +config SGI_TIOCX + bool "SGI TIO CX driver support" + depends on (IA64_SGI_SN2 || IA64_GENERIC) + help + If you have an SGI Altix and you have fpga devices attached + to your TIO, say Y here, otherwise say N. + +config SGI_MBCS + tristate "SGI FPGA Core Services driver support" + depends on (IA64_SGI_SN2 || IA64_GENERIC) + help + If you have an SGI Altix with an attached SABrick + say Y or M here, otherwise say N. + source "drivers/serial/Kconfig" config UNIX98_PTYS diff --git a/drivers/char/Makefile b/drivers/char/Makefile index 54ed76af1a4..3ea8cc80ea3 100644 --- a/drivers/char/Makefile +++ b/drivers/char/Makefile @@ -47,6 +47,7 @@ obj-$(CONFIG_MMTIMER) += mmtimer.o obj-$(CONFIG_VIOCONS) += viocons.o obj-$(CONFIG_VIOTAPE) += viotape.o obj-$(CONFIG_HVCS) += hvcs.o +obj-$(CONFIG_SGI_MBCS) += mbcs.o obj-$(CONFIG_PRINTER) += lp.o obj-$(CONFIG_TIPAR) += tipar.o diff --git a/drivers/char/mbcs.c b/drivers/char/mbcs.c new file mode 100644 index 00000000000..ec7100556c5 --- /dev/null +++ b/drivers/char/mbcs.c @@ -0,0 +1,849 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2005 Silicon Graphics, Inc. All rights reserved. + */ + +/* + * MOATB Core Services driver. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "mbcs.h" + +#define MBCS_DEBUG 0 +#if MBCS_DEBUG +#define DBG(fmt...) printk(KERN_ALERT fmt) +#else +#define DBG(fmt...) +#endif +int mbcs_major; + +LIST_HEAD(soft_list); + +/* + * file operations + */ +struct file_operations mbcs_ops = { + .open = mbcs_open, + .llseek = mbcs_sram_llseek, + .read = mbcs_sram_read, + .write = mbcs_sram_write, + .mmap = mbcs_gscr_mmap, +}; + +struct mbcs_callback_arg { + int minor; + struct cx_dev *cx_dev; +}; + +static inline void mbcs_getdma_init(struct getdma *gdma) +{ + memset(gdma, 0, sizeof(struct getdma)); + gdma->DoneIntEnable = 1; +} + +static inline void mbcs_putdma_init(struct putdma *pdma) +{ + memset(pdma, 0, sizeof(struct putdma)); + pdma->DoneIntEnable = 1; +} + +static inline void mbcs_algo_init(struct algoblock *algo_soft) +{ + memset(algo_soft, 0, sizeof(struct algoblock)); +} + +static inline void mbcs_getdma_set(void *mmr, + uint64_t hostAddr, + uint64_t localAddr, + uint64_t localRamSel, + uint64_t numPkts, + uint64_t amoEnable, + uint64_t intrEnable, + uint64_t peerIO, + uint64_t amoHostDest, + uint64_t amoModType, uint64_t intrHostDest, + uint64_t intrVector) +{ + union dma_control rdma_control; + union dma_amo_dest amo_dest; + union intr_dest intr_dest; + union dma_localaddr local_addr; + union dma_hostaddr host_addr; + + rdma_control.dma_control_reg = 0; + amo_dest.dma_amo_dest_reg = 0; + intr_dest.intr_dest_reg = 0; + local_addr.dma_localaddr_reg = 0; + host_addr.dma_hostaddr_reg = 0; + + host_addr.dma_sys_addr = hostAddr; + MBCS_MMR_SET(mmr, MBCS_RD_DMA_SYS_ADDR, host_addr.dma_hostaddr_reg); + + local_addr.dma_ram_addr = localAddr; + local_addr.dma_ram_sel = localRamSel; + MBCS_MMR_SET(mmr, MBCS_RD_DMA_LOC_ADDR, local_addr.dma_localaddr_reg); + + rdma_control.dma_op_length = numPkts; + rdma_control.done_amo_en = amoEnable; + rdma_control.done_int_en = intrEnable; + rdma_control.pio_mem_n = peerIO; + MBCS_MMR_SET(mmr, MBCS_RD_DMA_CTRL, rdma_control.dma_control_reg); + + amo_dest.dma_amo_sys_addr = amoHostDest; + amo_dest.dma_amo_mod_type = amoModType; + MBCS_MMR_SET(mmr, MBCS_RD_DMA_AMO_DEST, amo_dest.dma_amo_dest_reg); + + intr_dest.address = intrHostDest; + intr_dest.int_vector = intrVector; + MBCS_MMR_SET(mmr, MBCS_RD_DMA_INT_DEST, intr_dest.intr_dest_reg); + +} + +static inline void mbcs_putdma_set(void *mmr, + uint64_t hostAddr, + uint64_t localAddr, + uint64_t localRamSel, + uint64_t numPkts, + uint64_t amoEnable, + uint64_t intrEnable, + uint64_t peerIO, + uint64_t amoHostDest, + uint64_t amoModType, + uint64_t intrHostDest, uint64_t intrVector) +{ + union dma_control wdma_control; + union dma_amo_dest amo_dest; + union intr_dest intr_dest; + union dma_localaddr local_addr; + union dma_hostaddr host_addr; + + wdma_control.dma_control_reg = 0; + amo_dest.dma_amo_dest_reg = 0; + intr_dest.intr_dest_reg = 0; + local_addr.dma_localaddr_reg = 0; + host_addr.dma_hostaddr_reg = 0; + + host_addr.dma_sys_addr = hostAddr; + MBCS_MMR_SET(mmr, MBCS_WR_DMA_SYS_ADDR, host_addr.dma_hostaddr_reg); + + local_addr.dma_ram_addr = localAddr; + local_addr.dma_ram_sel = localRamSel; + MBCS_MMR_SET(mmr, MBCS_WR_DMA_LOC_ADDR, local_addr.dma_localaddr_reg); + + wdma_control.dma_op_length = numPkts; + wdma_control.done_amo_en = amoEnable; + wdma_control.done_int_en = intrEnable; + wdma_control.pio_mem_n = peerIO; + MBCS_MMR_SET(mmr, MBCS_WR_DMA_CTRL, wdma_control.dma_control_reg); + + amo_dest.dma_amo_sys_addr = amoHostDest; + amo_dest.dma_amo_mod_type = amoModType; + MBCS_MMR_SET(mmr, MBCS_WR_DMA_AMO_DEST, amo_dest.dma_amo_dest_reg); + + intr_dest.address = intrHostDest; + intr_dest.int_vector = intrVector; + MBCS_MMR_SET(mmr, MBCS_WR_DMA_INT_DEST, intr_dest.intr_dest_reg); + +} + +static inline void mbcs_algo_set(void *mmr, + uint64_t amoHostDest, + uint64_t amoModType, + uint64_t intrHostDest, + uint64_t intrVector, uint64_t algoStepCount) +{ + union dma_amo_dest amo_dest; + union intr_dest intr_dest; + union algo_step step; + + step.algo_step_reg = 0; + intr_dest.intr_dest_reg = 0; + amo_dest.dma_amo_dest_reg = 0; + + amo_dest.dma_amo_sys_addr = amoHostDest; + amo_dest.dma_amo_mod_type = amoModType; + MBCS_MMR_SET(mmr, MBCS_ALG_AMO_DEST, amo_dest.dma_amo_dest_reg); + + intr_dest.address = intrHostDest; + intr_dest.int_vector = intrVector; + MBCS_MMR_SET(mmr, MBCS_ALG_INT_DEST, intr_dest.intr_dest_reg); + + step.alg_step_cnt = algoStepCount; + MBCS_MMR_SET(mmr, MBCS_ALG_STEP, step.algo_step_reg); +} + +static inline int mbcs_getdma_start(struct mbcs_soft *soft) +{ + void *mmr_base; + struct getdma *gdma; + uint64_t numPkts; + union cm_control cm_control; + + mmr_base = soft->mmr_base; + gdma = &soft->getdma; + + /* check that host address got setup */ + if (!gdma->hostAddr) + return -1; + + numPkts = + (gdma->bytes + (MBCS_CACHELINE_SIZE - 1)) / MBCS_CACHELINE_SIZE; + + /* program engine */ + mbcs_getdma_set(mmr_base, tiocx_dma_addr(gdma->hostAddr), + gdma->localAddr, + (gdma->localAddr < MB2) ? 0 : + (gdma->localAddr < MB4) ? 1 : + (gdma->localAddr < MB6) ? 2 : 3, + numPkts, + gdma->DoneAmoEnable, + gdma->DoneIntEnable, + gdma->peerIO, + gdma->amoHostDest, + gdma->amoModType, + gdma->intrHostDest, gdma->intrVector); + + /* start engine */ + cm_control.cm_control_reg = MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.rd_dma_go = 1; + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, cm_control.cm_control_reg); + + return 0; + +} + +static inline int mbcs_putdma_start(struct mbcs_soft *soft) +{ + void *mmr_base; + struct putdma *pdma; + uint64_t numPkts; + union cm_control cm_control; + + mmr_base = soft->mmr_base; + pdma = &soft->putdma; + + /* check that host address got setup */ + if (!pdma->hostAddr) + return -1; + + numPkts = + (pdma->bytes + (MBCS_CACHELINE_SIZE - 1)) / MBCS_CACHELINE_SIZE; + + /* program engine */ + mbcs_putdma_set(mmr_base, tiocx_dma_addr(pdma->hostAddr), + pdma->localAddr, + (pdma->localAddr < MB2) ? 0 : + (pdma->localAddr < MB4) ? 1 : + (pdma->localAddr < MB6) ? 2 : 3, + numPkts, + pdma->DoneAmoEnable, + pdma->DoneIntEnable, + pdma->peerIO, + pdma->amoHostDest, + pdma->amoModType, + pdma->intrHostDest, pdma->intrVector); + + /* start engine */ + cm_control.cm_control_reg = MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.wr_dma_go = 1; + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, cm_control.cm_control_reg); + + return 0; + +} + +static inline int mbcs_algo_start(struct mbcs_soft *soft) +{ + struct algoblock *algo_soft = &soft->algo; + void *mmr_base = soft->mmr_base; + union cm_control cm_control; + + if (down_interruptible(&soft->algolock)) + return -ERESTARTSYS; + + atomic_set(&soft->algo_done, 0); + + mbcs_algo_set(mmr_base, + algo_soft->amoHostDest, + algo_soft->amoModType, + algo_soft->intrHostDest, + algo_soft->intrVector, algo_soft->algoStepCount); + + /* start algorithm */ + cm_control.cm_control_reg = MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.alg_done_int_en = 1; + cm_control.alg_go = 1; + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, cm_control.cm_control_reg); + + up(&soft->algolock); + + return 0; +} + +static inline ssize_t +do_mbcs_sram_dmawrite(struct mbcs_soft *soft, uint64_t hostAddr, + size_t len, loff_t * off) +{ + int rv = 0; + + if (down_interruptible(&soft->dmawritelock)) + return -ERESTARTSYS; + + atomic_set(&soft->dmawrite_done, 0); + + soft->putdma.hostAddr = hostAddr; + soft->putdma.localAddr = *off; + soft->putdma.bytes = len; + + if (mbcs_putdma_start(soft) < 0) { + DBG(KERN_ALERT "do_mbcs_sram_dmawrite: " + "mbcs_putdma_start failed\n"); + rv = -EAGAIN; + goto dmawrite_exit; + } + + if (wait_event_interruptible(soft->dmawrite_queue, + atomic_read(&soft->dmawrite_done))) { + rv = -ERESTARTSYS; + goto dmawrite_exit; + } + + rv = len; + *off += len; + +dmawrite_exit: + up(&soft->dmawritelock); + + return rv; +} + +static inline ssize_t +do_mbcs_sram_dmaread(struct mbcs_soft *soft, uint64_t hostAddr, + size_t len, loff_t * off) +{ + int rv = 0; + + if (down_interruptible(&soft->dmareadlock)) + return -ERESTARTSYS; + + atomic_set(&soft->dmawrite_done, 0); + + soft->getdma.hostAddr = hostAddr; + soft->getdma.localAddr = *off; + soft->getdma.bytes = len; + + if (mbcs_getdma_start(soft) < 0) { + DBG(KERN_ALERT "mbcs_strategy: mbcs_getdma_start failed\n"); + rv = -EAGAIN; + goto dmaread_exit; + } + + if (wait_event_interruptible(soft->dmaread_queue, + atomic_read(&soft->dmaread_done))) { + rv = -ERESTARTSYS; + goto dmaread_exit; + } + + rv = len; + *off += len; + +dmaread_exit: + up(&soft->dmareadlock); + + return rv; +} + +int mbcs_open(struct inode *ip, struct file *fp) +{ + struct mbcs_soft *soft; + int minor; + + minor = iminor(ip); + + list_for_each_entry(soft, &soft_list, list) { + if (soft->nasid == minor) { + fp->private_data = soft->cxdev; + return 0; + } + } + + return -ENODEV; +} + +ssize_t mbcs_sram_read(struct file * fp, char *buf, size_t len, loff_t * off) +{ + struct cx_dev *cx_dev = fp->private_data; + struct mbcs_soft *soft = cx_dev->soft; + uint64_t hostAddr; + int rv = 0; + + hostAddr = __get_dma_pages(GFP_KERNEL, get_order(len)); + if (hostAddr == 0) + return -ENOMEM; + + rv = do_mbcs_sram_dmawrite(soft, hostAddr, len, off); + if (rv < 0) + goto exit; + + if (copy_to_user(buf, (void *)hostAddr, len)) + rv = -EFAULT; + + exit: + free_pages(hostAddr, get_order(len)); + + return rv; +} + +ssize_t +mbcs_sram_write(struct file * fp, const char *buf, size_t len, loff_t * off) +{ + struct cx_dev *cx_dev = fp->private_data; + struct mbcs_soft *soft = cx_dev->soft; + uint64_t hostAddr; + int rv = 0; + + hostAddr = __get_dma_pages(GFP_KERNEL, get_order(len)); + if (hostAddr == 0) + return -ENOMEM; + + if (copy_from_user((void *)hostAddr, buf, len)) { + rv = -EFAULT; + goto exit; + } + + rv = do_mbcs_sram_dmaread(soft, hostAddr, len, off); + + exit: + free_pages(hostAddr, get_order(len)); + + return rv; +} + +loff_t mbcs_sram_llseek(struct file * filp, loff_t off, int whence) +{ + loff_t newpos; + + switch (whence) { + case 0: /* SEEK_SET */ + newpos = off; + break; + + case 1: /* SEEK_CUR */ + newpos = filp->f_pos + off; + break; + + case 2: /* SEEK_END */ + newpos = MBCS_SRAM_SIZE + off; + break; + + default: /* can't happen */ + return -EINVAL; + } + + if (newpos < 0) + return -EINVAL; + + filp->f_pos = newpos; + + return newpos; +} + +static uint64_t mbcs_pioaddr(struct mbcs_soft *soft, uint64_t offset) +{ + uint64_t mmr_base; + + mmr_base = (uint64_t) (soft->mmr_base + offset); + + return mmr_base; +} + +static void mbcs_debug_pioaddr_set(struct mbcs_soft *soft) +{ + soft->debug_addr = mbcs_pioaddr(soft, MBCS_DEBUG_START); +} + +static void mbcs_gscr_pioaddr_set(struct mbcs_soft *soft) +{ + soft->gscr_addr = mbcs_pioaddr(soft, MBCS_GSCR_START); +} + +int mbcs_gscr_mmap(struct file *fp, struct vm_area_struct *vma) +{ + struct cx_dev *cx_dev = fp->private_data; + struct mbcs_soft *soft = cx_dev->soft; + + if (vma->vm_pgoff != 0) + return -EINVAL; + + vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); + + /* Remap-pfn-range will mark the range VM_IO and VM_RESERVED */ + if (remap_pfn_range(vma, + vma->vm_start, + __pa(soft->gscr_addr) >> PAGE_SHIFT, + PAGE_SIZE, + vma->vm_page_prot)) + return -EAGAIN; + + return 0; +} + +/** + * mbcs_completion_intr_handler - Primary completion handler. + * @irq: irq + * @arg: soft struct for device + * @ep: regs + * + */ +static irqreturn_t +mbcs_completion_intr_handler(int irq, void *arg, struct pt_regs *ep) +{ + struct mbcs_soft *soft = (struct mbcs_soft *)arg; + void *mmr_base; + union cm_status cm_status; + union cm_control cm_control; + + mmr_base = soft->mmr_base; + cm_status.cm_status_reg = MBCS_MMR_GET(mmr_base, MBCS_CM_STATUS); + + if (cm_status.rd_dma_done) { + /* stop dma-read engine, clear status */ + cm_control.cm_control_reg = + MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.rd_dma_clr = 1; + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, + cm_control.cm_control_reg); + atomic_set(&soft->dmaread_done, 1); + wake_up(&soft->dmaread_queue); + } + if (cm_status.wr_dma_done) { + /* stop dma-write engine, clear status */ + cm_control.cm_control_reg = + MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.wr_dma_clr = 1; + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, + cm_control.cm_control_reg); + atomic_set(&soft->dmawrite_done, 1); + wake_up(&soft->dmawrite_queue); + } + if (cm_status.alg_done) { + /* clear status */ + cm_control.cm_control_reg = + MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.alg_done_clr = 1; + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, + cm_control.cm_control_reg); + atomic_set(&soft->algo_done, 1); + wake_up(&soft->algo_queue); + } + + return IRQ_HANDLED; +} + +/** + * mbcs_intr_alloc - Allocate interrupts. + * @dev: device pointer + * + */ +static int mbcs_intr_alloc(struct cx_dev *dev) +{ + struct sn_irq_info *sn_irq; + struct mbcs_soft *soft; + struct getdma *getdma; + struct putdma *putdma; + struct algoblock *algo; + + soft = dev->soft; + getdma = &soft->getdma; + putdma = &soft->putdma; + algo = &soft->algo; + + soft->get_sn_irq = NULL; + soft->put_sn_irq = NULL; + soft->algo_sn_irq = NULL; + + sn_irq = tiocx_irq_alloc(dev->cx_id.nasid, TIOCX_CORELET, -1, -1, -1); + if (sn_irq == NULL) + return -EAGAIN; + soft->get_sn_irq = sn_irq; + getdma->intrHostDest = sn_irq->irq_xtalkaddr; + getdma->intrVector = sn_irq->irq_irq; + if (request_irq(sn_irq->irq_irq, + (void *)mbcs_completion_intr_handler, SA_SHIRQ, + "MBCS get intr", (void *)soft)) { + tiocx_irq_free(soft->get_sn_irq); + return -EAGAIN; + } + + sn_irq = tiocx_irq_alloc(dev->cx_id.nasid, TIOCX_CORELET, -1, -1, -1); + if (sn_irq == NULL) { + free_irq(soft->get_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->get_sn_irq); + return -EAGAIN; + } + soft->put_sn_irq = sn_irq; + putdma->intrHostDest = sn_irq->irq_xtalkaddr; + putdma->intrVector = sn_irq->irq_irq; + if (request_irq(sn_irq->irq_irq, + (void *)mbcs_completion_intr_handler, SA_SHIRQ, + "MBCS put intr", (void *)soft)) { + tiocx_irq_free(soft->put_sn_irq); + free_irq(soft->get_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->get_sn_irq); + return -EAGAIN; + } + + sn_irq = tiocx_irq_alloc(dev->cx_id.nasid, TIOCX_CORELET, -1, -1, -1); + if (sn_irq == NULL) { + free_irq(soft->put_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->put_sn_irq); + free_irq(soft->get_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->get_sn_irq); + return -EAGAIN; + } + soft->algo_sn_irq = sn_irq; + algo->intrHostDest = sn_irq->irq_xtalkaddr; + algo->intrVector = sn_irq->irq_irq; + if (request_irq(sn_irq->irq_irq, + (void *)mbcs_completion_intr_handler, SA_SHIRQ, + "MBCS algo intr", (void *)soft)) { + tiocx_irq_free(soft->algo_sn_irq); + free_irq(soft->put_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->put_sn_irq); + free_irq(soft->get_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->get_sn_irq); + return -EAGAIN; + } + + return 0; +} + +/** + * mbcs_intr_dealloc - Remove interrupts. + * @dev: device pointer + * + */ +static void mbcs_intr_dealloc(struct cx_dev *dev) +{ + struct mbcs_soft *soft; + + soft = dev->soft; + + free_irq(soft->get_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->get_sn_irq); + free_irq(soft->put_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->put_sn_irq); + free_irq(soft->algo_sn_irq->irq_irq, soft); + tiocx_irq_free(soft->algo_sn_irq); +} + +static inline int mbcs_hw_init(struct mbcs_soft *soft) +{ + void *mmr_base = soft->mmr_base; + union cm_control cm_control; + union cm_req_timeout cm_req_timeout; + uint64_t err_stat; + + cm_req_timeout.cm_req_timeout_reg = + MBCS_MMR_GET(mmr_base, MBCS_CM_REQ_TOUT); + + cm_req_timeout.time_out = MBCS_CM_CONTROL_REQ_TOUT_MASK; + MBCS_MMR_SET(mmr_base, MBCS_CM_REQ_TOUT, + cm_req_timeout.cm_req_timeout_reg); + + mbcs_gscr_pioaddr_set(soft); + mbcs_debug_pioaddr_set(soft); + + /* clear errors */ + err_stat = MBCS_MMR_GET(mmr_base, MBCS_CM_ERR_STAT); + MBCS_MMR_SET(mmr_base, MBCS_CM_CLR_ERR_STAT, err_stat); + MBCS_MMR_ZERO(mmr_base, MBCS_CM_ERROR_DETAIL1); + + /* enable interrupts */ + /* turn off 2^23 (INT_EN_PIO_REQ_ADDR_INV) */ + MBCS_MMR_SET(mmr_base, MBCS_CM_ERR_INT_EN, 0x3ffffff7e00ffUL); + + /* arm status regs and clear engines */ + cm_control.cm_control_reg = MBCS_MMR_GET(mmr_base, MBCS_CM_CONTROL); + cm_control.rearm_stat_regs = 1; + cm_control.alg_clr = 1; + cm_control.wr_dma_clr = 1; + cm_control.rd_dma_clr = 1; + + MBCS_MMR_SET(mmr_base, MBCS_CM_CONTROL, cm_control.cm_control_reg); + + return 0; +} + +static ssize_t show_algo(struct device *dev, char *buf) +{ + struct cx_dev *cx_dev = to_cx_dev(dev); + struct mbcs_soft *soft = cx_dev->soft; + uint64_t debug0; + + /* + * By convention, the first debug register contains the + * algorithm number and revision. + */ + debug0 = *(uint64_t *) soft->debug_addr; + + return sprintf(buf, "0x%lx 0x%lx\n", + (debug0 >> 32), (debug0 & 0xffffffff)); +} + +static ssize_t store_algo(struct device *dev, const char *buf, size_t count) +{ + int n; + struct cx_dev *cx_dev = to_cx_dev(dev); + struct mbcs_soft *soft = cx_dev->soft; + + if (count <= 0) + return 0; + + n = simple_strtoul(buf, NULL, 0); + + if (n == 1) { + mbcs_algo_start(soft); + if (wait_event_interruptible(soft->algo_queue, + atomic_read(&soft->algo_done))) + return -ERESTARTSYS; + } + + return count; +} + +DEVICE_ATTR(algo, 0644, show_algo, store_algo); + +/** + * mbcs_probe - Initialize for device + * @dev: device pointer + * @device_id: id table pointer + * + */ +static int mbcs_probe(struct cx_dev *dev, const struct cx_device_id *id) +{ + struct mbcs_soft *soft; + + dev->soft = NULL; + + soft = kcalloc(1, sizeof(struct mbcs_soft), GFP_KERNEL); + if (soft == NULL) + return -ENOMEM; + + soft->nasid = dev->cx_id.nasid; + list_add(&soft->list, &soft_list); + soft->mmr_base = (void *)tiocx_swin_base(dev->cx_id.nasid); + dev->soft = soft; + soft->cxdev = dev; + + init_waitqueue_head(&soft->dmawrite_queue); + init_waitqueue_head(&soft->dmaread_queue); + init_waitqueue_head(&soft->algo_queue); + + init_MUTEX(&soft->dmawritelock); + init_MUTEX(&soft->dmareadlock); + init_MUTEX(&soft->algolock); + + mbcs_getdma_init(&soft->getdma); + mbcs_putdma_init(&soft->putdma); + mbcs_algo_init(&soft->algo); + + mbcs_hw_init(soft); + + /* Allocate interrupts */ + mbcs_intr_alloc(dev); + + device_create_file(&dev->dev, &dev_attr_algo); + + return 0; +} + +static int mbcs_remove(struct cx_dev *dev) +{ + if (dev->soft) { + mbcs_intr_dealloc(dev); + kfree(dev->soft); + } + + device_remove_file(&dev->dev, &dev_attr_algo); + + return 0; +} + +const struct cx_device_id __devinitdata mbcs_id_table[] = { + { + .part_num = MBCS_PART_NUM, + .mfg_num = MBCS_MFG_NUM, + }, + { + .part_num = MBCS_PART_NUM_ALG0, + .mfg_num = MBCS_MFG_NUM, + }, + {0, 0} +}; + +MODULE_DEVICE_TABLE(cx, mbcs_id_table); + +struct cx_drv mbcs_driver = { + .name = DEVICE_NAME, + .id_table = mbcs_id_table, + .probe = mbcs_probe, + .remove = mbcs_remove, +}; + +static void __exit mbcs_exit(void) +{ + int rv; + + rv = unregister_chrdev(mbcs_major, DEVICE_NAME); + if (rv < 0) + DBG(KERN_ALERT "Error in unregister_chrdev: %d\n", rv); + + cx_driver_unregister(&mbcs_driver); +} + +static int __init mbcs_init(void) +{ + int rv; + + // Put driver into chrdevs[]. Get major number. + rv = register_chrdev(mbcs_major, DEVICE_NAME, &mbcs_ops); + if (rv < 0) { + DBG(KERN_ALERT "mbcs_init: can't get major number. %d\n", rv); + return rv; + } + mbcs_major = rv; + + return cx_driver_register(&mbcs_driver); +} + +module_init(mbcs_init); +module_exit(mbcs_exit); + +MODULE_AUTHOR("Bruce Losure "); +MODULE_DESCRIPTION("Driver for MOATB Core Services"); +MODULE_LICENSE("GPL"); diff --git a/drivers/char/mbcs.h b/drivers/char/mbcs.h new file mode 100644 index 00000000000..844644d201c --- /dev/null +++ b/drivers/char/mbcs.h @@ -0,0 +1,553 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2005 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef __MBCS_H__ +#define __MBCS_H__ + +/* + * General macros + */ +#define MB (1024*1024) +#define MB2 (2*MB) +#define MB4 (4*MB) +#define MB6 (6*MB) + +/* + * Offsets and masks + */ +#define MBCS_CM_ID 0x0000 /* Identification */ +#define MBCS_CM_STATUS 0x0008 /* Status */ +#define MBCS_CM_ERROR_DETAIL1 0x0010 /* Error Detail1 */ +#define MBCS_CM_ERROR_DETAIL2 0x0018 /* Error Detail2 */ +#define MBCS_CM_CONTROL 0x0020 /* Control */ +#define MBCS_CM_REQ_TOUT 0x0028 /* Request Time-out */ +#define MBCS_CM_ERR_INT_DEST 0x0038 /* Error Interrupt Destination */ +#define MBCS_CM_TARG_FL 0x0050 /* Target Flush */ +#define MBCS_CM_ERR_STAT 0x0060 /* Error Status */ +#define MBCS_CM_CLR_ERR_STAT 0x0068 /* Clear Error Status */ +#define MBCS_CM_ERR_INT_EN 0x0070 /* Error Interrupt Enable */ +#define MBCS_RD_DMA_SYS_ADDR 0x0100 /* Read DMA System Address */ +#define MBCS_RD_DMA_LOC_ADDR 0x0108 /* Read DMA Local Address */ +#define MBCS_RD_DMA_CTRL 0x0110 /* Read DMA Control */ +#define MBCS_RD_DMA_AMO_DEST 0x0118 /* Read DMA AMO Destination */ +#define MBCS_RD_DMA_INT_DEST 0x0120 /* Read DMA Interrupt Destination */ +#define MBCS_RD_DMA_AUX_STAT 0x0130 /* Read DMA Auxillary Status */ +#define MBCS_WR_DMA_SYS_ADDR 0x0200 /* Write DMA System Address */ +#define MBCS_WR_DMA_LOC_ADDR 0x0208 /* Write DMA Local Address */ +#define MBCS_WR_DMA_CTRL 0x0210 /* Write DMA Control */ +#define MBCS_WR_DMA_AMO_DEST 0x0218 /* Write DMA AMO Destination */ +#define MBCS_WR_DMA_INT_DEST 0x0220 /* Write DMA Interrupt Destination */ +#define MBCS_WR_DMA_AUX_STAT 0x0230 /* Write DMA Auxillary Status */ +#define MBCS_ALG_AMO_DEST 0x0300 /* Algorithm AMO Destination */ +#define MBCS_ALG_INT_DEST 0x0308 /* Algorithm Interrupt Destination */ +#define MBCS_ALG_OFFSETS 0x0310 +#define MBCS_ALG_STEP 0x0318 /* Algorithm Step */ + +#define MBCS_GSCR_START 0x0000000 +#define MBCS_DEBUG_START 0x0100000 +#define MBCS_RAM0_START 0x0200000 +#define MBCS_RAM1_START 0x0400000 +#define MBCS_RAM2_START 0x0600000 + +#define MBCS_CM_CONTROL_REQ_TOUT_MASK 0x0000000000ffffffUL +//#define PIO_BASE_ADDR_BASE_OFFSET_MASK 0x00fffffffff00000UL + +#define MBCS_SRAM_SIZE (1024*1024) +#define MBCS_CACHELINE_SIZE 128 + +/* + * MMR get's and put's + */ +#define MBCS_MMR_ADDR(mmr_base, offset)((uint64_t *)(mmr_base + offset)) +#define MBCS_MMR_SET(mmr_base, offset, value) { \ + uint64_t *mbcs_mmr_set_u64p, readback; \ + mbcs_mmr_set_u64p = (uint64_t *)(mmr_base + offset); \ + *mbcs_mmr_set_u64p = value; \ + readback = *mbcs_mmr_set_u64p; \ +} +#define MBCS_MMR_GET(mmr_base, offset) *(uint64_t *)(mmr_base + offset) +#define MBCS_MMR_ZERO(mmr_base, offset) MBCS_MMR_SET(mmr_base, offset, 0) + +/* + * MBCS mmr structures + */ +union cm_id { + uint64_t cm_id_reg; + struct { + uint64_t always_one:1, // 0 + mfg_id:11, // 11:1 + part_num:16, // 27:12 + bitstream_rev:8, // 35:28 + :28; // 63:36 + }; +}; + +union cm_status { + uint64_t cm_status_reg; + struct { + uint64_t pending_reads:8, // 7:0 + pending_writes:8, // 15:8 + ice_rsp_credits:8, // 23:16 + ice_req_credits:8, // 31:24 + cm_req_credits:8, // 39:32 + :1, // 40 + rd_dma_in_progress:1, // 41 + rd_dma_done:1, // 42 + :1, // 43 + wr_dma_in_progress:1, // 44 + wr_dma_done:1, // 45 + alg_waiting:1, // 46 + alg_pipe_running:1, // 47 + alg_done:1, // 48 + :3, // 51:49 + pending_int_reqs:8, // 59:52 + :3, // 62:60 + alg_half_speed_sel:1; // 63 + }; +}; + +union cm_error_detail1 { + uint64_t cm_error_detail1_reg; + struct { + uint64_t packet_type:4, // 3:0 + source_id:2, // 5:4 + data_size:2, // 7:6 + tnum:8, // 15:8 + byte_enable:8, // 23:16 + gfx_cred:8, // 31:24 + read_type:2, // 33:32 + pio_or_memory:1, // 34 + head_cw_error:1, // 35 + :12, // 47:36 + head_error_bit:1, // 48 + data_error_bit:1, // 49 + :13, // 62:50 + valid:1; // 63 + }; +}; + +union cm_error_detail2 { + uint64_t cm_error_detail2_reg; + struct { + uint64_t address:56, // 55:0 + :8; // 63:56 + }; +}; + +union cm_control { + uint64_t cm_control_reg; + struct { + uint64_t cm_id:2, // 1:0 + :2, // 3:2 + max_trans:5, // 8:4 + :3, // 11:9 + address_mode:1, // 12 + :7, // 19:13 + credit_limit:8, // 27:20 + :5, // 32:28 + rearm_stat_regs:1, // 33 + prescalar_byp:1, // 34 + force_gap_war:1, // 35 + rd_dma_go:1, // 36 + wr_dma_go:1, // 37 + alg_go:1, // 38 + rd_dma_clr:1, // 39 + wr_dma_clr:1, // 40 + alg_clr:1, // 41 + :2, // 43:42 + alg_wait_step:1, // 44 + alg_done_amo_en:1, // 45 + alg_done_int_en:1, // 46 + :1, // 47 + alg_sram0_locked:1, // 48 + alg_sram1_locked:1, // 49 + alg_sram2_locked:1, // 50 + alg_done_clr:1, // 51 + :12; // 63:52 + }; +}; + +union cm_req_timeout { + uint64_t cm_req_timeout_reg; + struct { + uint64_t time_out:24, // 23:0 + :40; // 63:24 + }; +}; + +union intr_dest { + uint64_t intr_dest_reg; + struct { + uint64_t address:56, // 55:0 + int_vector:8; // 63:56 + }; +}; + +union cm_error_status { + uint64_t cm_error_status_reg; + struct { + uint64_t ecc_sbe:1, // 0 + ecc_mbe:1, // 1 + unsupported_req:1, // 2 + unexpected_rsp:1, // 3 + bad_length:1, // 4 + bad_datavalid:1, // 5 + buffer_overflow:1, // 6 + request_timeout:1, // 7 + :8, // 15:8 + head_inv_data_size:1, // 16 + rsp_pactype_inv:1, // 17 + head_sb_err:1, // 18 + missing_head:1, // 19 + head_inv_rd_type:1, // 20 + head_cmd_err_bit:1, // 21 + req_addr_align_inv:1, // 22 + pio_req_addr_inv:1, // 23 + req_range_dsize_inv:1, // 24 + early_term:1, // 25 + early_tail:1, // 26 + missing_tail:1, // 27 + data_flit_sb_err:1, // 28 + cm2hcm_req_cred_of:1, // 29 + cm2hcm_rsp_cred_of:1, // 30 + rx_bad_didn:1, // 31 + rd_dma_err_rsp:1, // 32 + rd_dma_tnum_tout:1, // 33 + rd_dma_multi_tnum_tou:1, // 34 + wr_dma_err_rsp:1, // 35 + wr_dma_tnum_tout:1, // 36 + wr_dma_multi_tnum_tou:1, // 37 + alg_data_overflow:1, // 38 + alg_data_underflow:1, // 39 + ram0_access_conflict:1, // 40 + ram1_access_conflict:1, // 41 + ram2_access_conflict:1, // 42 + ram0_perr:1, // 43 + ram1_perr:1, // 44 + ram2_perr:1, // 45 + int_gen_rsp_err:1, // 46 + int_gen_tnum_tout:1, // 47 + rd_dma_prog_err:1, // 48 + wr_dma_prog_err:1, // 49 + :14; // 63:50 + }; +}; + +union cm_clr_error_status { + uint64_t cm_clr_error_status_reg; + struct { + uint64_t clr_ecc_sbe:1, // 0 + clr_ecc_mbe:1, // 1 + clr_unsupported_req:1, // 2 + clr_unexpected_rsp:1, // 3 + clr_bad_length:1, // 4 + clr_bad_datavalid:1, // 5 + clr_buffer_overflow:1, // 6 + clr_request_timeout:1, // 7 + :8, // 15:8 + clr_head_inv_data_siz:1, // 16 + clr_rsp_pactype_inv:1, // 17 + clr_head_sb_err:1, // 18 + clr_missing_head:1, // 19 + clr_head_inv_rd_type:1, // 20 + clr_head_cmd_err_bit:1, // 21 + clr_req_addr_align_in:1, // 22 + clr_pio_req_addr_inv:1, // 23 + clr_req_range_dsize_i:1, // 24 + clr_early_term:1, // 25 + clr_early_tail:1, // 26 + clr_missing_tail:1, // 27 + clr_data_flit_sb_err:1, // 28 + clr_cm2hcm_req_cred_o:1, // 29 + clr_cm2hcm_rsp_cred_o:1, // 30 + clr_rx_bad_didn:1, // 31 + clr_rd_dma_err_rsp:1, // 32 + clr_rd_dma_tnum_tout:1, // 33 + clr_rd_dma_multi_tnum:1, // 34 + clr_wr_dma_err_rsp:1, // 35 + clr_wr_dma_tnum_tout:1, // 36 + clr_wr_dma_multi_tnum:1, // 37 + clr_alg_data_overflow:1, // 38 + clr_alg_data_underflo:1, // 39 + clr_ram0_access_confl:1, // 40 + clr_ram1_access_confl:1, // 41 + clr_ram2_access_confl:1, // 42 + clr_ram0_perr:1, // 43 + clr_ram1_perr:1, // 44 + clr_ram2_perr:1, // 45 + clr_int_gen_rsp_err:1, // 46 + clr_int_gen_tnum_tout:1, // 47 + clr_rd_dma_prog_err:1, // 48 + clr_wr_dma_prog_err:1, // 49 + :14; // 63:50 + }; +}; + +union cm_error_intr_enable { + uint64_t cm_error_intr_enable_reg; + struct { + uint64_t int_en_ecc_sbe:1, // 0 + int_en_ecc_mbe:1, // 1 + int_en_unsupported_re:1, // 2 + int_en_unexpected_rsp:1, // 3 + int_en_bad_length:1, // 4 + int_en_bad_datavalid:1, // 5 + int_en_buffer_overflo:1, // 6 + int_en_request_timeou:1, // 7 + :8, // 15:8 + int_en_head_inv_data_:1, // 16 + int_en_rsp_pactype_in:1, // 17 + int_en_head_sb_err:1, // 18 + int_en_missing_head:1, // 19 + int_en_head_inv_rd_ty:1, // 20 + int_en_head_cmd_err_b:1, // 21 + int_en_req_addr_align:1, // 22 + int_en_pio_req_addr_i:1, // 23 + int_en_req_range_dsiz:1, // 24 + int_en_early_term:1, // 25 + int_en_early_tail:1, // 26 + int_en_missing_tail:1, // 27 + int_en_data_flit_sb_e:1, // 28 + int_en_cm2hcm_req_cre:1, // 29 + int_en_cm2hcm_rsp_cre:1, // 30 + int_en_rx_bad_didn:1, // 31 + int_en_rd_dma_err_rsp:1, // 32 + int_en_rd_dma_tnum_to:1, // 33 + int_en_rd_dma_multi_t:1, // 34 + int_en_wr_dma_err_rsp:1, // 35 + int_en_wr_dma_tnum_to:1, // 36 + int_en_wr_dma_multi_t:1, // 37 + int_en_alg_data_overf:1, // 38 + int_en_alg_data_under:1, // 39 + int_en_ram0_access_co:1, // 40 + int_en_ram1_access_co:1, // 41 + int_en_ram2_access_co:1, // 42 + int_en_ram0_perr:1, // 43 + int_en_ram1_perr:1, // 44 + int_en_ram2_perr:1, // 45 + int_en_int_gen_rsp_er:1, // 46 + int_en_int_gen_tnum_t:1, // 47 + int_en_rd_dma_prog_er:1, // 48 + int_en_wr_dma_prog_er:1, // 49 + :14; // 63:50 + }; +}; + +struct cm_mmr { + union cm_id id; + union cm_status status; + union cm_error_detail1 err_detail1; + union cm_error_detail2 err_detail2; + union cm_control control; + union cm_req_timeout req_timeout; + uint64_t reserved1[1]; + union intr_dest int_dest; + uint64_t reserved2[2]; + uint64_t targ_flush; + uint64_t reserved3[1]; + union cm_error_status err_status; + union cm_clr_error_status clr_err_status; + union cm_error_intr_enable int_enable; +}; + +union dma_hostaddr { + uint64_t dma_hostaddr_reg; + struct { + uint64_t dma_sys_addr:56, // 55:0 + :8; // 63:56 + }; +}; + +union dma_localaddr { + uint64_t dma_localaddr_reg; + struct { + uint64_t dma_ram_addr:21, // 20:0 + dma_ram_sel:2, // 22:21 + :41; // 63:23 + }; +}; + +union dma_control { + uint64_t dma_control_reg; + struct { + uint64_t dma_op_length:16, // 15:0 + :18, // 33:16 + done_amo_en:1, // 34 + done_int_en:1, // 35 + :1, // 36 + pio_mem_n:1, // 37 + :26; // 63:38 + }; +}; + +union dma_amo_dest { + uint64_t dma_amo_dest_reg; + struct { + uint64_t dma_amo_sys_addr:56, // 55:0 + dma_amo_mod_type:3, // 58:56 + :5; // 63:59 + }; +}; + +union rdma_aux_status { + uint64_t rdma_aux_status_reg; + struct { + uint64_t op_num_pacs_left:17, // 16:0 + :5, // 21:17 + lrsp_buff_empty:1, // 22 + :17, // 39:23 + pending_reqs_left:6, // 45:40 + :18; // 63:46 + }; +}; + +struct rdma_mmr { + union dma_hostaddr host_addr; + union dma_localaddr local_addr; + union dma_control control; + union dma_amo_dest amo_dest; + union intr_dest intr_dest; + union rdma_aux_status aux_status; +}; + +union wdma_aux_status { + uint64_t wdma_aux_status_reg; + struct { + uint64_t op_num_pacs_left:17, // 16:0 + :4, // 20:17 + lreq_buff_empty:1, // 21 + :18, // 39:22 + pending_reqs_left:6, // 45:40 + :18; // 63:46 + }; +}; + +struct wdma_mmr { + union dma_hostaddr host_addr; + union dma_localaddr local_addr; + union dma_control control; + union dma_amo_dest amo_dest; + union intr_dest intr_dest; + union wdma_aux_status aux_status; +}; + +union algo_step { + uint64_t algo_step_reg; + struct { + uint64_t alg_step_cnt:16, // 15:0 + :48; // 63:16 + }; +}; + +struct algo_mmr { + union dma_amo_dest amo_dest; + union intr_dest intr_dest; + union { + uint64_t algo_offset_reg; + struct { + uint64_t sram0_offset:7, // 6:0 + reserved0:1, // 7 + sram1_offset:7, // 14:8 + reserved1:1, // 15 + sram2_offset:7, // 22:16 + reserved2:14; // 63:23 + }; + } sram_offset; + union algo_step step; +}; + +struct mbcs_mmr { + struct cm_mmr cm; + uint64_t reserved1[17]; + struct rdma_mmr rdDma; + uint64_t reserved2[25]; + struct wdma_mmr wrDma; + uint64_t reserved3[25]; + struct algo_mmr algo; + uint64_t reserved4[156]; +}; + +/* + * defines + */ +#define DEVICE_NAME "mbcs" +#define MBCS_PART_NUM 0xfff0 +#define MBCS_PART_NUM_ALG0 0xf001 +#define MBCS_MFG_NUM 0x1 + +struct algoblock { + uint64_t amoHostDest; + uint64_t amoModType; + uint64_t intrHostDest; + uint64_t intrVector; + uint64_t algoStepCount; +}; + +struct getdma { + uint64_t hostAddr; + uint64_t localAddr; + uint64_t bytes; + uint64_t DoneAmoEnable; + uint64_t DoneIntEnable; + uint64_t peerIO; + uint64_t amoHostDest; + uint64_t amoModType; + uint64_t intrHostDest; + uint64_t intrVector; +}; + +struct putdma { + uint64_t hostAddr; + uint64_t localAddr; + uint64_t bytes; + uint64_t DoneAmoEnable; + uint64_t DoneIntEnable; + uint64_t peerIO; + uint64_t amoHostDest; + uint64_t amoModType; + uint64_t intrHostDest; + uint64_t intrVector; +}; + +struct mbcs_soft { + struct list_head list; + struct cx_dev *cxdev; + int major; + int nasid; + void *mmr_base; + wait_queue_head_t dmawrite_queue; + wait_queue_head_t dmaread_queue; + wait_queue_head_t algo_queue; + struct sn_irq_info *get_sn_irq; + struct sn_irq_info *put_sn_irq; + struct sn_irq_info *algo_sn_irq; + struct getdma getdma; + struct putdma putdma; + struct algoblock algo; + uint64_t gscr_addr; // pio addr + uint64_t ram0_addr; // pio addr + uint64_t ram1_addr; // pio addr + uint64_t ram2_addr; // pio addr + uint64_t debug_addr; // pio addr + atomic_t dmawrite_done; + atomic_t dmaread_done; + atomic_t algo_done; + struct semaphore dmawritelock; + struct semaphore dmareadlock; + struct semaphore algolock; +}; + +extern int mbcs_open(struct inode *ip, struct file *fp); +extern ssize_t mbcs_sram_read(struct file *fp, char *buf, size_t len, + loff_t * off); +extern ssize_t mbcs_sram_write(struct file *fp, const char *buf, size_t len, + loff_t * off); +extern loff_t mbcs_sram_llseek(struct file *filp, loff_t off, int whence); +extern int mbcs_gscr_mmap(struct file *fp, struct vm_area_struct *vma); + +#endif // __MBCS_H__ diff --git a/include/asm-ia64/sn/addrs.h b/include/asm-ia64/sn/addrs.h index ae0bc99d573..960d626ee58 100644 --- a/include/asm-ia64/sn/addrs.h +++ b/include/asm-ia64/sn/addrs.h @@ -169,7 +169,10 @@ #define TIO_BWIN_SIZE_BITS 30 /* big window size: 1G */ #define NODE_SWIN_BASE(n, w) ((w == 0) ? NODE_BWIN_BASE((n), SWIN0_BIGWIN) \ : RAW_NODE_SWIN_BASE(n, w)) +#define TIO_SWIN_BASE(n, w) (TIO_IO_BASE(n) + \ + ((u64) (w) << TIO_SWIN_SIZE_BITS)) #define NODE_IO_BASE(n) (GLOBAL_MMR_SPACE | NASID_SPACE(n)) +#define TIO_IO_BASE(n) (UNCACHED | NASID_SPACE(n)) #define BWIN_SIZE (1UL << BWIN_SIZE_BITS) #define NODE_BWIN_BASE0(n) (NODE_IO_BASE(n) + BWIN_SIZE) #define NODE_BWIN_BASE(n, w) (NODE_BWIN_BASE0(n) + ((u64) (w) << BWIN_SIZE_BITS)) diff --git a/include/asm-ia64/sn/tiocx.h b/include/asm-ia64/sn/tiocx.h new file mode 100644 index 00000000000..c5447a50450 --- /dev/null +++ b/include/asm-ia64/sn/tiocx.h @@ -0,0 +1,71 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2005 Silicon Graphics, Inc. All rights reserved. + */ + +#ifndef _ASM_IA64_SN_TIO_TIOCX_H +#define _ASM_IA64_SN_TIO_TIOCX_H + +#ifdef __KERNEL__ + +struct cx_id_s { + unsigned int part_num; + unsigned int mfg_num; + int nasid; +}; + +struct cx_dev { + struct cx_id_s cx_id; + void *soft; /* driver specific */ + struct hubdev_info *hubdev; + struct device dev; + struct cx_drv *driver; +}; + +struct cx_device_id { + unsigned int part_num; + unsigned int mfg_num; +}; + +struct cx_drv { + char *name; + const struct cx_device_id *id_table; + struct device_driver driver; + int (*probe) (struct cx_dev * dev, const struct cx_device_id * id); + int (*remove) (struct cx_dev * dev); +}; + +/* create DMA address by stripping AS bits */ +#define TIOCX_DMA_ADDR(a) (uint64_t)((uint64_t)(a) & 0xffffcfffffffffUL) + +#define TIOCX_TO_TIOCX_DMA_ADDR(a) (uint64_t)(((uint64_t)(a) & 0xfffffffff) | \ + ((((uint64_t)(a)) & 0xffffc000000000UL) <<2)) + +#define TIO_CE_ASIC_PARTNUM 0xce00 +#define TIOCX_CORELET 3 + +/* These are taken from tio_mmr_as.h */ +#define TIO_ICE_FRZ_CFG TIO_MMR_ADDR_MOD(0x00000000b0008100UL) +#define TIO_ICE_PMI_TX_CFG TIO_MMR_ADDR_MOD(0x00000000b000b100UL) +#define TIO_ICE_PMI_TX_DYN_CREDIT_STAT_CB3 TIO_MMR_ADDR_MOD(0x00000000b000be18UL) +#define TIO_ICE_PMI_TX_DYN_CREDIT_STAT_CB3_CREDIT_CNT_MASK 0x000000000000000fUL + +#define to_cx_dev(n) container_of(n, struct cx_dev, dev) +#define to_cx_driver(drv) container_of(drv, struct cx_drv, driver) + +extern struct sn_irq_info *tiocx_irq_alloc(nasid_t, int, int, nasid_t, int); +extern void tiocx_irq_free(struct sn_irq_info *); +extern int cx_device_unregister(struct cx_dev *); +extern int cx_device_register(nasid_t, int, int, struct hubdev_info *); +extern int cx_driver_unregister(struct cx_drv *); +extern int cx_driver_register(struct cx_drv *); +extern uint64_t tiocx_dma_addr(uint64_t addr); +extern uint64_t tiocx_swin_base(int nasid); +extern void tiocx_mmr_store(int nasid, uint64_t offset, uint64_t value); +extern uint64_t tiocx_mmr_load(int nasid, uint64_t offset); + +#endif // __KERNEL__ +#endif // _ASM_IA64_SN_TIO_TIOCX__ -- cgit v1.2.3-70-g09d2 From c1298c5c7e5763a3b2fd4a9535d474ff6e54cd53 Mon Sep 17 00:00:00 2001 From: Aaron J Young Date: Mon, 25 Apr 2005 13:11:14 -0700 Subject: [IA64-SGI] Altix: enable poweroff This patch adds the necessary "hook" to allow SGI/SN machines to perform a system power off upon a 'init 0', 'halt -p', 'poweroff' or 'shutdown -h'. The "hook" is to set the pm_power_off callback to ia64_sn_power_down(). pm_power_off is checked in machine_power_off()/do_poweroff() and, if set, is executed. ia64_sn_power_down() is a function already present (but not used currently) in the sn kernel. ia64_sn_power_down() makes a SAL call to execute the power off. Signed-off-by: Aaron J Young Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/setup.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/setup.c b/arch/ia64/sn/kernel/setup.c index f0306b516af..d35f2a6f9c9 100644 --- a/arch/ia64/sn/kernel/setup.c +++ b/arch/ia64/sn/kernel/setup.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include @@ -353,6 +354,14 @@ void __init sn_setup(char **cmdline_p) screen_info = sn_screen_info; sn_timer_init(); + + /* + * set pm_power_off to a SAL call to allow + * sn machines to power off. The SAL call can be replaced + * by an ACPI interface call when ACPI is fully implemented + * for sn. + */ + pm_power_off = ia64_sn_power_down; } /** -- cgit v1.2.3-70-g09d2 From ff3eb55ed97db3f12964beeffe3d34602d295367 Mon Sep 17 00:00:00 2001 From: Bruce Losure Date: Mon, 25 Apr 2005 13:12:02 -0700 Subject: [IA64-SGI] Missed the "bk new" for this file in the last commit. Signed-off-by: Bruce Losure Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/tiocx.c | 548 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 548 insertions(+) create mode 100644 arch/ia64/sn/kernel/tiocx.c (limited to 'arch') diff --git a/arch/ia64/sn/kernel/tiocx.c b/arch/ia64/sn/kernel/tiocx.c new file mode 100644 index 00000000000..66190d7e492 --- /dev/null +++ b/arch/ia64/sn/kernel/tiocx.c @@ -0,0 +1,548 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (c) 2005 Silicon Graphics, Inc. All rights reserved. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "tio.h" +#include "xtalk/xwidgetdev.h" +#include "xtalk/hubdev.h" + +#define CX_DEV_NONE 0 +#define DEVICE_NAME "tiocx" +#define WIDGET_ID 0 +#define TIOCX_DEBUG 0 + +#if TIOCX_DEBUG +#define DBG(fmt...) printk(KERN_ALERT fmt) +#else +#define DBG(fmt...) +#endif + +struct device_attribute dev_attr_cxdev_control; + +/** + * tiocx_match - Try to match driver id list with device. + * @dev: device pointer + * @drv: driver pointer + * + * Returns 1 if match, 0 otherwise. + */ +static int tiocx_match(struct device *dev, struct device_driver *drv) +{ + struct cx_dev *cx_dev = to_cx_dev(dev); + struct cx_drv *cx_drv = to_cx_driver(drv); + const struct cx_device_id *ids = cx_drv->id_table; + + if (!ids) + return 0; + + while (ids->part_num) { + if (ids->part_num == cx_dev->cx_id.part_num) + return 1; + ids++; + } + return 0; + +} + +static int tiocx_hotplug(struct device *dev, char **envp, int num_envp, + char *buffer, int buffer_size) +{ + return -ENODEV; +} + +static void tiocx_bus_release(struct device *dev) +{ + kfree(to_cx_dev(dev)); +} + +struct bus_type tiocx_bus_type = { + .name = "tiocx", + .match = tiocx_match, + .hotplug = tiocx_hotplug, +}; + +/** + * cx_device_match - Find cx_device in the id table. + * @ids: id table from driver + * @cx_device: part/mfg id for the device + * + */ +static const struct cx_device_id *cx_device_match(const struct cx_device_id + *ids, + struct cx_dev *cx_device) +{ + /* + * NOTES: We may want to check for CX_ANY_ID too. + * Do we want to match against nasid too? + * CX_DEV_NONE == 0, if the driver tries to register for + * part/mfg == 0 we should return no-match (NULL) here. + */ + while (ids->part_num && ids->mfg_num) { + if (ids->part_num == cx_device->cx_id.part_num && + ids->mfg_num == cx_device->cx_id.mfg_num) + return ids; + ids++; + } + + return NULL; +} + +/** + * cx_device_probe - Look for matching device. + * Call driver probe routine if found. + * @cx_driver: driver table (cx_drv struct) from driver + * @cx_device: part/mfg id for the device + */ +static int cx_device_probe(struct device *dev) +{ + const struct cx_device_id *id; + struct cx_drv *cx_drv = to_cx_driver(dev->driver); + struct cx_dev *cx_dev = to_cx_dev(dev); + int error = 0; + + if (!cx_dev->driver && cx_drv->probe) { + id = cx_device_match(cx_drv->id_table, cx_dev); + if (id) { + if ((error = cx_drv->probe(cx_dev, id)) < 0) + return error; + else + cx_dev->driver = cx_drv; + } + } + + return error; +} + +/** + * cx_driver_remove - Remove driver from device struct. + * @dev: device + */ +static int cx_driver_remove(struct device *dev) +{ + struct cx_dev *cx_dev = to_cx_dev(dev); + struct cx_drv *cx_drv = cx_dev->driver; + if (cx_drv->remove) + cx_drv->remove(cx_dev); + cx_dev->driver = NULL; + return 0; +} + +/** + * cx_driver_register - Register the driver. + * @cx_driver: driver table (cx_drv struct) from driver + * + * Called from the driver init routine to register a driver. + * The cx_drv struct contains the driver name, a pointer to + * a table of part/mfg numbers and a pointer to the driver's + * probe/attach routine. + */ +int cx_driver_register(struct cx_drv *cx_driver) +{ + cx_driver->driver.name = cx_driver->name; + cx_driver->driver.bus = &tiocx_bus_type; + cx_driver->driver.probe = cx_device_probe; + cx_driver->driver.remove = cx_driver_remove; + + return driver_register(&cx_driver->driver); +} + +/** + * cx_driver_unregister - Unregister the driver. + * @cx_driver: driver table (cx_drv struct) from driver + */ +int cx_driver_unregister(struct cx_drv *cx_driver) +{ + driver_unregister(&cx_driver->driver); + return 0; +} + +/** + * cx_device_register - Register a device. + * @nasid: device's nasid + * @part_num: device's part number + * @mfg_num: device's manufacturer number + * @hubdev: hub info associated with this device + * + */ +int +cx_device_register(nasid_t nasid, int part_num, int mfg_num, + struct hubdev_info *hubdev) +{ + struct cx_dev *cx_dev; + + cx_dev = kcalloc(1, sizeof(struct cx_dev), GFP_KERNEL); + DBG("cx_dev= 0x%p\n", cx_dev); + if (cx_dev == NULL) + return -ENOMEM; + + cx_dev->cx_id.part_num = part_num; + cx_dev->cx_id.mfg_num = mfg_num; + cx_dev->cx_id.nasid = nasid; + cx_dev->hubdev = hubdev; + + cx_dev->dev.parent = NULL; + cx_dev->dev.bus = &tiocx_bus_type; + cx_dev->dev.release = tiocx_bus_release; + snprintf(cx_dev->dev.bus_id, BUS_ID_SIZE, "%d.0x%x", + cx_dev->cx_id.nasid, cx_dev->cx_id.part_num); + device_register(&cx_dev->dev); + get_device(&cx_dev->dev); + + device_create_file(&cx_dev->dev, &dev_attr_cxdev_control); + + return 0; +} + +/** + * cx_device_unregister - Unregister a device. + * @cx_dev: part/mfg id for the device + */ +int cx_device_unregister(struct cx_dev *cx_dev) +{ + put_device(&cx_dev->dev); + device_unregister(&cx_dev->dev); + return 0; +} + +/** + * cx_device_reload - Reload the device. + * @nasid: device's nasid + * @part_num: device's part number + * @mfg_num: device's manufacturer number + * + * Remove the device associated with 'nasid' from device list and then + * call device-register with the given part/mfg numbers. + */ +static int cx_device_reload(struct cx_dev *cx_dev) +{ + device_remove_file(&cx_dev->dev, &dev_attr_cxdev_control); + cx_device_unregister(cx_dev); + return cx_device_register(cx_dev->cx_id.nasid, cx_dev->cx_id.part_num, + cx_dev->cx_id.mfg_num, cx_dev->hubdev); +} + +static inline uint64_t tiocx_intr_alloc(nasid_t nasid, int widget, + u64 sn_irq_info, + int req_irq, nasid_t req_nasid, + int req_slice) +{ + struct ia64_sal_retval rv; + rv.status = 0; + rv.v0 = 0; + + ia64_sal_oemcall_nolock(&rv, SN_SAL_IOIF_INTERRUPT, + SAL_INTR_ALLOC, nasid, + widget, sn_irq_info, req_irq, + req_nasid, req_slice); + return rv.status; +} + +static inline void tiocx_intr_free(nasid_t nasid, int widget, + struct sn_irq_info *sn_irq_info) +{ + struct ia64_sal_retval rv; + rv.status = 0; + rv.v0 = 0; + + ia64_sal_oemcall_nolock(&rv, SN_SAL_IOIF_INTERRUPT, + SAL_INTR_FREE, nasid, + widget, sn_irq_info->irq_irq, + sn_irq_info->irq_cookie, 0, 0); +} + +struct sn_irq_info *tiocx_irq_alloc(nasid_t nasid, int widget, int irq, + nasid_t req_nasid, int slice) +{ + struct sn_irq_info *sn_irq_info; + int status; + int sn_irq_size = sizeof(struct sn_irq_info); + + if ((nasid & 1) == 0) + return NULL; + + sn_irq_info = kmalloc(sn_irq_size, GFP_KERNEL); + if (sn_irq_info == NULL) + return NULL; + + memset(sn_irq_info, 0x0, sn_irq_size); + + status = tiocx_intr_alloc(nasid, widget, __pa(sn_irq_info), irq, + req_nasid, slice); + if (status) { + kfree(sn_irq_info); + return NULL; + } else { + return sn_irq_info; + } +} + +void tiocx_irq_free(struct sn_irq_info *sn_irq_info) +{ + uint64_t bridge = (uint64_t) sn_irq_info->irq_bridge; + nasid_t nasid = NASID_GET(bridge); + int widget; + + if (nasid & 1) { + widget = TIO_SWIN_WIDGETNUM(bridge); + tiocx_intr_free(nasid, widget, sn_irq_info); + kfree(sn_irq_info); + } +} + +uint64_t +tiocx_dma_addr(uint64_t addr) +{ + return PHYS_TO_TIODMA(addr); +} + +uint64_t +tiocx_swin_base(int nasid) +{ + return TIO_SWIN_BASE(nasid, TIOCX_CORELET); +} + +EXPORT_SYMBOL(cx_driver_register); +EXPORT_SYMBOL(cx_driver_unregister); +EXPORT_SYMBOL(cx_device_register); +EXPORT_SYMBOL(cx_device_unregister); +EXPORT_SYMBOL(tiocx_irq_alloc); +EXPORT_SYMBOL(tiocx_irq_free); +EXPORT_SYMBOL(tiocx_bus_type); +EXPORT_SYMBOL(tiocx_dma_addr); +EXPORT_SYMBOL(tiocx_swin_base); + +static uint64_t tiocx_get_hubdev_info(u64 handle, u64 address) +{ + + struct ia64_sal_retval ret_stuff; + ret_stuff.status = 0; + ret_stuff.v0 = 0; + + ia64_sal_oemcall_nolock(&ret_stuff, + SN_SAL_IOIF_GET_HUBDEV_INFO, + handle, address, 0, 0, 0, 0, 0); + return ret_stuff.v0; +} + +static void tio_conveyor_set(nasid_t nasid, int enable_flag) +{ + uint64_t ice_frz; + uint64_t disable_cb = (1ull << 61); + + if (!(nasid & 1)) + return; + + ice_frz = REMOTE_HUB_L(nasid, TIO_ICE_FRZ_CFG); + if (enable_flag) { + if (!(ice_frz & disable_cb)) /* already enabled */ + return; + ice_frz &= ~disable_cb; + } else { + if (ice_frz & disable_cb) /* already disabled */ + return; + ice_frz |= disable_cb; + } + DBG(KERN_ALERT "TIO_ICE_FRZ_CFG= 0x%lx\n", ice_frz); + REMOTE_HUB_S(nasid, TIO_ICE_FRZ_CFG, ice_frz); +} + +#define tio_conveyor_enable(nasid) tio_conveyor_set(nasid, 1) +#define tio_conveyor_disable(nasid) tio_conveyor_set(nasid, 0) + +static void tio_corelet_reset(nasid_t nasid, int corelet) +{ + if (!(nasid & 1)) + return; + + REMOTE_HUB_S(nasid, TIO_ICE_PMI_TX_CFG, 1 << corelet); + udelay(2000); + REMOTE_HUB_S(nasid, TIO_ICE_PMI_TX_CFG, 0); + udelay(2000); +} + +static int fpga_attached(nasid_t nasid) +{ + uint64_t cx_credits; + + cx_credits = REMOTE_HUB_L(nasid, TIO_ICE_PMI_TX_DYN_CREDIT_STAT_CB3); + cx_credits &= TIO_ICE_PMI_TX_DYN_CREDIT_STAT_CB3_CREDIT_CNT_MASK; + DBG("cx_credits= 0x%lx\n", cx_credits); + + return (cx_credits == 0xf) ? 1 : 0; +} + +static int tiocx_reload(struct cx_dev *cx_dev) +{ + int part_num = CX_DEV_NONE; + int mfg_num = CX_DEV_NONE; + nasid_t nasid = cx_dev->cx_id.nasid; + + if (fpga_attached(nasid)) { + uint64_t cx_id; + + cx_id = + *(volatile int32_t *)(TIO_SWIN_BASE(nasid, TIOCX_CORELET) + + WIDGET_ID); + part_num = XWIDGET_PART_NUM(cx_id); + mfg_num = XWIDGET_MFG_NUM(cx_id); + DBG("part= 0x%x, mfg= 0x%x\n", part_num, mfg_num); + /* just ignore it if it's a CE */ + if (part_num == TIO_CE_ASIC_PARTNUM) + return 0; + } + + cx_dev->cx_id.part_num = part_num; + cx_dev->cx_id.mfg_num = mfg_num; + + /* + * Delete old device and register the new one. It's ok if + * part_num/mfg_num == CX_DEV_NONE. We want to register + * devices in the table even if a bitstream isn't loaded. + * That allows use to see that a bitstream isn't loaded via + * TIOCX_IOCTL_DEV_LIST. + */ + return cx_device_reload(cx_dev); +} + +static ssize_t show_cxdev_control(struct device *dev, char *buf) +{ + struct cx_dev *cx_dev = to_cx_dev(dev); + + return sprintf(buf, "0x%x 0x%x 0x%x\n", + cx_dev->cx_id.nasid, + cx_dev->cx_id.part_num, cx_dev->cx_id.mfg_num); +} + +static ssize_t store_cxdev_control(struct device *dev, const char *buf, + size_t count) +{ + int n; + struct cx_dev *cx_dev = to_cx_dev(dev); + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (count <= 0) + return 0; + + n = simple_strtoul(buf, NULL, 0); + + switch (n) { + case 1: + tiocx_reload(cx_dev); + break; + case 3: + tio_corelet_reset(cx_dev->cx_id.nasid, TIOCX_CORELET); + break; + default: + break; + } + + return count; +} + +DEVICE_ATTR(cxdev_control, 0644, show_cxdev_control, store_cxdev_control); + +static int __init tiocx_init(void) +{ + cnodeid_t cnodeid; + int found_tiocx_device = 0; + + bus_register(&tiocx_bus_type); + + for (cnodeid = 0; cnodeid < MAX_COMPACT_NODES; cnodeid++) { + nasid_t nasid; + + if ((nasid = cnodeid_to_nasid(cnodeid)) < 0) + break; /* No more nasids .. bail out of loop */ + + if (nasid & 0x1) { /* TIO's are always odd */ + struct hubdev_info *hubdev; + uint64_t status; + struct xwidget_info *widgetp; + + DBG("Found TIO at nasid 0x%x\n", nasid); + + hubdev = + (struct hubdev_info *)(NODEPDA(cnodeid)->pdinfo); + status = + tiocx_get_hubdev_info(nasid, + (uint64_t) __pa(hubdev)); + if (status) + continue; + + widgetp = &hubdev->hdi_xwidget_info[TIOCX_CORELET]; + + /* The CE hangs off of the CX port but is not an FPGA */ + if (widgetp->xwi_hwid.part_num == TIO_CE_ASIC_PARTNUM) + continue; + + tio_corelet_reset(nasid, TIOCX_CORELET); + tio_conveyor_enable(nasid); + + if (cx_device_register + (nasid, widgetp->xwi_hwid.part_num, + widgetp->xwi_hwid.mfg_num, hubdev) < 0) + return -ENXIO; + else + found_tiocx_device++; + } + } + + /* It's ok if we find zero devices. */ + DBG("found_tiocx_device= %d\n", found_tiocx_device); + + return 0; +} + +static void __exit tiocx_exit(void) +{ + struct device *dev; + struct device *tdev; + + DBG("tiocx_exit\n"); + + /* + * Unregister devices. + */ + list_for_each_entry_safe(dev, tdev, &tiocx_bus_type.devices.list, + bus_list) { + if (dev) { + struct cx_dev *cx_dev = to_cx_dev(dev); + device_remove_file(dev, &dev_attr_cxdev_control); + cx_device_unregister(cx_dev); + } + } + + bus_unregister(&tiocx_bus_type); +} + +module_init(tiocx_init); +module_exit(tiocx_exit); + +/************************************************************************ + * Module licensing and description + ************************************************************************/ +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Bruce Losure "); +MODULE_DESCRIPTION("TIOCX module"); +MODULE_SUPPORTED_DEVICE(DEVICE_NAME); -- cgit v1.2.3-70-g09d2 From fde740e4dd4a05ca8957490d468fa9b2770f5bd6 Mon Sep 17 00:00:00 2001 From: Robin Holt Date: Mon, 25 Apr 2005 13:13:16 -0700 Subject: [IA64] Percpu quicklist for combined allocator for pgd/pmd/pte. This patch introduces using the quicklists for pgd, pmd, and pte levels by combining the alloc and free functions into a common set of routines. This greatly simplifies the reading of this header file. This patch is simple but necessary for large numa configurations. It simply ensures that only pages from the local node are added to a cpus quicklist. This prevents the trapping of pages on a remote nodes quicklist by starting a process, touching a large number of pages to fill pmd and pte entries, migrating to another node, and then unmapping or exiting. With those conditions, the pages get trapped and if the machine has more than 100 nodes of the same size, the calculation of the pgtable high water mark will be larger than any single node so page table cache flushing will never occur. I ran lmbench lat_proc fork and lat_proc exec on a zx1 with and without this patch and did not notice any change. On an sn2 machine, there was a slight improvement which is possibly due to pages from other nodes trapped on the test node before starting the run. I did not investigate further. This patch shrinks the quicklist based upon free memory on the node instead of the high/low water marks. I have written it to enable preemption periodically and recalculate the amount to shrink every time we have freed enough pages that the quicklist size should have grown. I rescan the nodes zones each pass because other processess may be draining node memory at the same time as we are adding. Signed-off-by: Robin Holt Signed-off-by: Tony Luck --- arch/ia64/mm/contig.c | 3 +- arch/ia64/mm/discontig.c | 3 +- arch/ia64/mm/init.c | 74 ++++++++++++++-------- include/asm-ia64/pgalloc.h | 144 ++++++++++++++++++------------------------- include/asm-ia64/processor.h | 3 - 5 files changed, 112 insertions(+), 115 deletions(-) (limited to 'arch') diff --git a/arch/ia64/mm/contig.c b/arch/ia64/mm/contig.c index 6daf15ac894..91a055f5731 100644 --- a/arch/ia64/mm/contig.c +++ b/arch/ia64/mm/contig.c @@ -61,7 +61,8 @@ show_mem (void) printk("%d reserved pages\n", reserved); printk("%d pages shared\n", shared); printk("%d pages swap cached\n", cached); - printk("%ld pages in page table cache\n", pgtable_cache_size); + printk("%ld pages in page table cache\n", + pgtable_quicklist_total_size()); } /* physical address where the bootmem map is located */ diff --git a/arch/ia64/mm/discontig.c b/arch/ia64/mm/discontig.c index 3456a9b6971..c0071092939 100644 --- a/arch/ia64/mm/discontig.c +++ b/arch/ia64/mm/discontig.c @@ -582,7 +582,8 @@ void show_mem(void) printk("%d reserved pages\n", total_reserved); printk("%d pages shared\n", total_shared); printk("%d pages swap cached\n", total_cached); - printk("Total of %ld pages in page table cache\n", pgtable_cache_size); + printk("Total of %ld pages in page table cache\n", + pgtable_quicklist_total_size()); printk("%d free buffer pages\n", nr_free_buffer_pages()); } diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 65cf839573e..4892be53e22 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -39,6 +39,9 @@ DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); +DEFINE_PER_CPU(unsigned long *, __pgtable_quicklist); +DEFINE_PER_CPU(long, __pgtable_quicklist_size); + extern void ia64_tlb_init (void); unsigned long MAX_DMA_ADDRESS = PAGE_OFFSET + 0x100000000UL; @@ -50,27 +53,53 @@ struct page *vmem_map; EXPORT_SYMBOL(vmem_map); #endif -static int pgt_cache_water[2] = { 25, 50 }; - -struct page *zero_page_memmap_ptr; /* map entry for zero page */ +struct page *zero_page_memmap_ptr; /* map entry for zero page */ EXPORT_SYMBOL(zero_page_memmap_ptr); +#define MIN_PGT_PAGES 25UL +#define MAX_PGT_FREES_PER_PASS 16 +#define PGT_FRACTION_OF_NODE_MEM 16 + +static inline long +max_pgt_pages(void) +{ + u64 node_free_pages, max_pgt_pages; + +#ifndef CONFIG_NUMA + node_free_pages = nr_free_pages(); +#else + node_free_pages = nr_free_pages_pgdat(NODE_DATA(numa_node_id())); +#endif + max_pgt_pages = node_free_pages / PGT_FRACTION_OF_NODE_MEM; + max_pgt_pages = max(max_pgt_pages, MIN_PGT_PAGES); + return max_pgt_pages; +} + +static inline long +min_pages_to_free(void) +{ + long pages_to_free; + + pages_to_free = pgtable_quicklist_size - max_pgt_pages(); + pages_to_free = min(pages_to_free, MAX_PGT_FREES_PER_PASS); + return pages_to_free; +} + void -check_pgt_cache (void) +check_pgt_cache(void) { - int low, high; + long pages_to_free; - low = pgt_cache_water[0]; - high = pgt_cache_water[1]; + if (unlikely(pgtable_quicklist_size <= MIN_PGT_PAGES)) + return; preempt_disable(); - if (pgtable_cache_size > (u64) high) { - do { - if (pgd_quicklist) - free_page((unsigned long)pgd_alloc_one_fast(NULL)); - if (pmd_quicklist) - free_page((unsigned long)pmd_alloc_one_fast(NULL, 0)); - } while (pgtable_cache_size > (u64) low); + while (unlikely((pages_to_free = min_pages_to_free()) > 0)) { + while (pages_to_free--) { + free_page((unsigned long)pgtable_quicklist_alloc()); + } + preempt_enable(); + preempt_disable(); } preempt_enable(); } @@ -523,11 +552,14 @@ void mem_init (void) { long reserved_pages, codesize, datasize, initsize; - unsigned long num_pgt_pages; pg_data_t *pgdat; int i; static struct kcore_list kcore_mem, kcore_vmem, kcore_kernel; + BUG_ON(PTRS_PER_PGD * sizeof(pgd_t) != PAGE_SIZE); + BUG_ON(PTRS_PER_PMD * sizeof(pmd_t) != PAGE_SIZE); + BUG_ON(PTRS_PER_PTE * sizeof(pte_t) != PAGE_SIZE); + #ifdef CONFIG_PCI /* * This needs to be called _after_ the command line has been parsed but _before_ @@ -564,18 +596,6 @@ mem_init (void) num_physpages << (PAGE_SHIFT - 10), codesize >> 10, reserved_pages << (PAGE_SHIFT - 10), datasize >> 10, initsize >> 10); - /* - * Allow for enough (cached) page table pages so that we can map the entire memory - * at least once. Each task also needs a couple of page tables pages, so add in a - * fudge factor for that (don't use "threads-max" here; that would be wrong!). - * Don't allow the cache to be more than 10% of total memory, though. - */ -# define NUM_TASKS 500 /* typical number of tasks */ - num_pgt_pages = nr_free_pages() / PTRS_PER_PGD + NUM_TASKS; - if (num_pgt_pages > nr_free_pages() / 10) - num_pgt_pages = nr_free_pages() / 10; - if (num_pgt_pages > (u64) pgt_cache_water[1]) - pgt_cache_water[1] = num_pgt_pages; /* * For fsyscall entrpoints with no light-weight handler, use the ordinary diff --git a/include/asm-ia64/pgalloc.h b/include/asm-ia64/pgalloc.h index 0f05dc8bd46..e86a8c331ee 100644 --- a/include/asm-ia64/pgalloc.h +++ b/include/asm-ia64/pgalloc.h @@ -22,146 +22,124 @@ #include -/* - * Very stupidly, we used to get new pgd's and pmd's, init their contents - * to point to the NULL versions of the next level page table, later on - * completely re-init them the same way, then free them up. This wasted - * a lot of work and caused unnecessary memory traffic. How broken... - * We fix this by caching them. - */ -#define pgd_quicklist (local_cpu_data->pgd_quick) -#define pmd_quicklist (local_cpu_data->pmd_quick) -#define pgtable_cache_size (local_cpu_data->pgtable_cache_sz) +DECLARE_PER_CPU(unsigned long *, __pgtable_quicklist); +#define pgtable_quicklist __ia64_per_cpu_var(__pgtable_quicklist) +DECLARE_PER_CPU(long, __pgtable_quicklist_size); +#define pgtable_quicklist_size __ia64_per_cpu_var(__pgtable_quicklist_size) -static inline pgd_t* -pgd_alloc_one_fast (struct mm_struct *mm) +static inline long pgtable_quicklist_total_size(void) +{ + long ql_size; + int cpuid; + + for_each_online_cpu(cpuid) { + ql_size += per_cpu(__pgtable_quicklist_size, cpuid); + } + return ql_size; +} + +static inline void *pgtable_quicklist_alloc(void) { unsigned long *ret = NULL; preempt_disable(); - ret = pgd_quicklist; + ret = pgtable_quicklist; if (likely(ret != NULL)) { - pgd_quicklist = (unsigned long *)(*ret); + pgtable_quicklist = (unsigned long *)(*ret); ret[0] = 0; - --pgtable_cache_size; - } else - ret = NULL; + --pgtable_quicklist_size; + } else { + ret = (unsigned long *)__get_free_page(GFP_KERNEL | __GFP_ZERO); + } preempt_enable(); - return (pgd_t *) ret; + return ret; } -static inline pgd_t* -pgd_alloc (struct mm_struct *mm) +static inline void pgtable_quicklist_free(void *pgtable_entry) { - /* the VM system never calls pgd_alloc_one_fast(), so we do it here. */ - pgd_t *pgd = pgd_alloc_one_fast(mm); +#ifdef CONFIG_NUMA + unsigned long nid = page_to_nid(virt_to_page(pgtable_entry)); - if (unlikely(pgd == NULL)) { - pgd = (pgd_t *)__get_free_page(GFP_KERNEL|__GFP_ZERO); + if (unlikely(nid != numa_node_id())) { + free_page((unsigned long)pgtable_entry); + return; } - return pgd; -} +#endif -static inline void -pgd_free (pgd_t *pgd) -{ preempt_disable(); - *(unsigned long *)pgd = (unsigned long) pgd_quicklist; - pgd_quicklist = (unsigned long *) pgd; - ++pgtable_cache_size; + *(unsigned long *)pgtable_entry = (unsigned long)pgtable_quicklist; + pgtable_quicklist = (unsigned long *)pgtable_entry; + ++pgtable_quicklist_size; preempt_enable(); } -static inline void -pud_populate (struct mm_struct *mm, pud_t *pud_entry, pmd_t *pmd) +static inline pgd_t *pgd_alloc(struct mm_struct *mm) { - pud_val(*pud_entry) = __pa(pmd); + return pgtable_quicklist_alloc(); } -static inline pmd_t* -pmd_alloc_one_fast (struct mm_struct *mm, unsigned long addr) +static inline void pgd_free(pgd_t * pgd) { - unsigned long *ret = NULL; - - preempt_disable(); - - ret = (unsigned long *)pmd_quicklist; - if (likely(ret != NULL)) { - pmd_quicklist = (unsigned long *)(*ret); - ret[0] = 0; - --pgtable_cache_size; - } - - preempt_enable(); - - return (pmd_t *)ret; + pgtable_quicklist_free(pgd); } -static inline pmd_t* -pmd_alloc_one (struct mm_struct *mm, unsigned long addr) +static inline void +pud_populate(struct mm_struct *mm, pud_t * pud_entry, pmd_t * pmd) { - pmd_t *pmd = (pmd_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO); + pud_val(*pud_entry) = __pa(pmd); +} - return pmd; +static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr) +{ + return pgtable_quicklist_alloc(); } -static inline void -pmd_free (pmd_t *pmd) +static inline void pmd_free(pmd_t * pmd) { - preempt_disable(); - *(unsigned long *)pmd = (unsigned long) pmd_quicklist; - pmd_quicklist = (unsigned long *) pmd; - ++pgtable_cache_size; - preempt_enable(); + pgtable_quicklist_free(pmd); } #define __pmd_free_tlb(tlb, pmd) pmd_free(pmd) static inline void -pmd_populate (struct mm_struct *mm, pmd_t *pmd_entry, struct page *pte) +pmd_populate(struct mm_struct *mm, pmd_t * pmd_entry, struct page *pte) { pmd_val(*pmd_entry) = page_to_phys(pte); } static inline void -pmd_populate_kernel (struct mm_struct *mm, pmd_t *pmd_entry, pte_t *pte) +pmd_populate_kernel(struct mm_struct *mm, pmd_t * pmd_entry, pte_t * pte) { pmd_val(*pmd_entry) = __pa(pte); } -static inline struct page * -pte_alloc_one (struct mm_struct *mm, unsigned long addr) +static inline struct page *pte_alloc_one(struct mm_struct *mm, + unsigned long addr) { - struct page *pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0); - - return pte; + return virt_to_page(pgtable_quicklist_alloc()); } -static inline pte_t * -pte_alloc_one_kernel (struct mm_struct *mm, unsigned long addr) +static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm, + unsigned long addr) { - pte_t *pte = (pte_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO); - - return pte; + return pgtable_quicklist_alloc(); } -static inline void -pte_free (struct page *pte) +static inline void pte_free(struct page *pte) { - __free_page(pte); + pgtable_quicklist_free(page_address(pte)); } -static inline void -pte_free_kernel (pte_t *pte) +static inline void pte_free_kernel(pte_t * pte) { - free_page((unsigned long) pte); + pgtable_quicklist_free(pte); } -#define __pte_free_tlb(tlb, pte) tlb_remove_page((tlb), (pte)) +#define __pte_free_tlb(tlb, pte) pte_free(pte) -extern void check_pgt_cache (void); +extern void check_pgt_cache(void); -#endif /* _ASM_IA64_PGALLOC_H */ +#endif /* _ASM_IA64_PGALLOC_H */ diff --git a/include/asm-ia64/processor.h b/include/asm-ia64/processor.h index 2807f8d766d..983798ec179 100644 --- a/include/asm-ia64/processor.h +++ b/include/asm-ia64/processor.h @@ -137,9 +137,6 @@ struct cpuinfo_ia64 { __u64 nsec_per_cyc; /* (1000000000< Date: Mon, 25 Apr 2005 13:14:36 -0700 Subject: [IA64] sba_iommu bug fixes This fixes a couple of bugs in the zx1/sx1000 sba_iommu. These are all pretty low likelihood of hitting. The first problem is a simple off by one, deep in the sba_alloc_range() error path. Surrounding that was a lock ordering problem that could have potentially deadlocked with the order the locks are grabbed in sba_unmap_single(). I moved the resource locking into sba_search_bitmap() to prevent this. Finally, there's a potential race between unmapping pdir entries and marking incoming DMA pages clean. If you see any oddities, please let me know, but I've tested it pretty thoroughly here. Tony, please apply. Thanks, BTW, many of the options in this driver not on by default are becoming more and more broken. I'll be working on some patches to clean them out, but I wanted to get this bug fix out first. Signed-off-by: Alex Williamson Signed-off-by: Tony Luck --- arch/ia64/hp/common/sba_iommu.c | 96 ++++++++++++++++++++++++----------------- 1 file changed, 56 insertions(+), 40 deletions(-) (limited to 'arch') diff --git a/arch/ia64/hp/common/sba_iommu.c b/arch/ia64/hp/common/sba_iommu.c index 017c9ab5fc1..6a8fcba7a85 100644 --- a/arch/ia64/hp/common/sba_iommu.c +++ b/arch/ia64/hp/common/sba_iommu.c @@ -1,9 +1,9 @@ /* ** IA64 System Bus Adapter (SBA) I/O MMU manager ** -** (c) Copyright 2002-2004 Alex Williamson +** (c) Copyright 2002-2005 Alex Williamson ** (c) Copyright 2002-2003 Grant Grundler -** (c) Copyright 2002-2004 Hewlett-Packard Company +** (c) Copyright 2002-2005 Hewlett-Packard Company ** ** Portions (c) 2000 Grant Grundler (from parisc I/O MMU code) ** Portions (c) 1999 Dave S. Miller (from sparc64 I/O MMU code) @@ -459,21 +459,32 @@ get_iovp_order (unsigned long size) * sba_search_bitmap - find free space in IO PDIR resource bitmap * @ioc: IO MMU structure which owns the pdir we are interested in. * @bits_wanted: number of entries we need. + * @use_hint: use res_hint to indicate where to start looking * * Find consecutive free bits in resource bitmap. * Each bit represents one entry in the IO Pdir. * Cool perf optimization: search for log2(size) bits at a time. */ static SBA_INLINE unsigned long -sba_search_bitmap(struct ioc *ioc, unsigned long bits_wanted) +sba_search_bitmap(struct ioc *ioc, unsigned long bits_wanted, int use_hint) { - unsigned long *res_ptr = ioc->res_hint; + unsigned long *res_ptr; unsigned long *res_end = (unsigned long *) &(ioc->res_map[ioc->res_size]); - unsigned long pide = ~0UL; + unsigned long flags, pide = ~0UL; ASSERT(((unsigned long) ioc->res_hint & (sizeof(unsigned long) - 1UL)) == 0); ASSERT(res_ptr < res_end); + spin_lock_irqsave(&ioc->res_lock, flags); + + /* Allow caller to force a search through the entire resource space */ + if (likely(use_hint)) { + res_ptr = ioc->res_hint; + } else { + res_ptr = (ulong *)ioc->res_map; + ioc->res_bitshift = 0; + } + /* * N.B. REO/Grande defect AR2305 can cause TLB fetch timeouts * if a TLB entry is purged while in use. sba_mark_invalid() @@ -570,10 +581,12 @@ not_found: prefetch(ioc->res_map); ioc->res_hint = (unsigned long *) ioc->res_map; ioc->res_bitshift = 0; + spin_unlock_irqrestore(&ioc->res_lock, flags); return (pide); found_it: ioc->res_hint = res_ptr; + spin_unlock_irqrestore(&ioc->res_lock, flags); return (pide); } @@ -594,36 +607,36 @@ sba_alloc_range(struct ioc *ioc, size_t size) unsigned long itc_start; #endif unsigned long pide; - unsigned long flags; ASSERT(pages_needed); ASSERT(0 == (size & ~iovp_mask)); - spin_lock_irqsave(&ioc->res_lock, flags); - #ifdef PDIR_SEARCH_TIMING itc_start = ia64_get_itc(); #endif /* ** "seek and ye shall find"...praying never hurts either... */ - pide = sba_search_bitmap(ioc, pages_needed); + pide = sba_search_bitmap(ioc, pages_needed, 1); if (unlikely(pide >= (ioc->res_size << 3))) { - pide = sba_search_bitmap(ioc, pages_needed); + pide = sba_search_bitmap(ioc, pages_needed, 0); if (unlikely(pide >= (ioc->res_size << 3))) { #if DELAYED_RESOURCE_CNT > 0 + unsigned long flags; + /* ** With delayed resource freeing, we can give this one more shot. We're ** getting close to being in trouble here, so do what we can to make this ** one count. */ - spin_lock(&ioc->saved_lock); + spin_lock_irqsave(&ioc->saved_lock, flags); if (ioc->saved_cnt > 0) { struct sba_dma_pair *d; int cnt = ioc->saved_cnt; - d = &(ioc->saved[ioc->saved_cnt]); + d = &(ioc->saved[ioc->saved_cnt - 1]); + spin_lock(&ioc->res_lock); while (cnt--) { sba_mark_invalid(ioc, d->iova, d->size); sba_free_range(ioc, d->iova, d->size); @@ -631,10 +644,11 @@ sba_alloc_range(struct ioc *ioc, size_t size) } ioc->saved_cnt = 0; READ_REG(ioc->ioc_hpa+IOC_PCOM); /* flush purges */ + spin_unlock(&ioc->res_lock); } - spin_unlock(&ioc->saved_lock); + spin_unlock_irqrestore(&ioc->saved_lock, flags); - pide = sba_search_bitmap(ioc, pages_needed); + pide = sba_search_bitmap(ioc, pages_needed, 0); if (unlikely(pide >= (ioc->res_size << 3))) panic(__FILE__ ": I/O MMU @ %p is out of mapping resources\n", ioc->ioc_hpa); @@ -664,8 +678,6 @@ sba_alloc_range(struct ioc *ioc, size_t size) (uint) ((unsigned long) ioc->res_hint - (unsigned long) ioc->res_map), ioc->res_bitshift ); - spin_unlock_irqrestore(&ioc->res_lock, flags); - return (pide); } @@ -950,6 +962,30 @@ sba_map_single(struct device *dev, void *addr, size_t size, int dir) return SBA_IOVA(ioc, iovp, offset); } +#ifdef ENABLE_MARK_CLEAN +static SBA_INLINE void +sba_mark_clean(struct ioc *ioc, dma_addr_t iova, size_t size) +{ + u32 iovp = (u32) SBA_IOVP(ioc,iova); + int off = PDIR_INDEX(iovp); + void *addr; + + if (size <= iovp_size) { + addr = phys_to_virt(ioc->pdir_base[off] & + ~0xE000000000000FFFULL); + mark_clean(addr, size); + } else { + do { + addr = phys_to_virt(ioc->pdir_base[off] & + ~0xE000000000000FFFULL); + mark_clean(addr, min(size, iovp_size)); + off++; + size -= iovp_size; + } while (size > 0); + } +} +#endif + /** * sba_unmap_single - unmap one IOVA and free resources * @dev: instance of PCI owned by the driver that's asking. @@ -995,6 +1031,10 @@ void sba_unmap_single(struct device *dev, dma_addr_t iova, size_t size, int dir) size += offset; size = ROUNDUP(size, iovp_size); +#ifdef ENABLE_MARK_CLEAN + if (dir == DMA_FROM_DEVICE) + sba_mark_clean(ioc, iova, size); +#endif #if DELAYED_RESOURCE_CNT > 0 spin_lock_irqsave(&ioc->saved_lock, flags); @@ -1021,30 +1061,6 @@ void sba_unmap_single(struct device *dev, dma_addr_t iova, size_t size, int dir) READ_REG(ioc->ioc_hpa+IOC_PCOM); /* flush purges */ spin_unlock_irqrestore(&ioc->res_lock, flags); #endif /* DELAYED_RESOURCE_CNT == 0 */ -#ifdef ENABLE_MARK_CLEAN - if (dir == DMA_FROM_DEVICE) { - u32 iovp = (u32) SBA_IOVP(ioc,iova); - int off = PDIR_INDEX(iovp); - void *addr; - - if (size <= iovp_size) { - addr = phys_to_virt(ioc->pdir_base[off] & - ~0xE000000000000FFFULL); - mark_clean(addr, size); - } else { - size_t byte_cnt = size; - - do { - addr = phys_to_virt(ioc->pdir_base[off] & - ~0xE000000000000FFFULL); - mark_clean(addr, min(byte_cnt, iovp_size)); - off++; - byte_cnt -= iovp_size; - - } while (byte_cnt > 0); - } - } -#endif } -- cgit v1.2.3-70-g09d2 From e96c9b4779e651a7469bea677be3a08f70be399e Mon Sep 17 00:00:00 2001 From: Tony Luck Date: Mon, 25 Apr 2005 13:16:59 -0700 Subject: [IA64] MAX_PGT_FREES_PER_PASS must be 'L' to avoid warning 'min' is very picky about types of arguments, make it happy Signed-off-by: Tony Luck --- arch/ia64/mm/init.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 4892be53e22..547785e3cba 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -57,7 +57,7 @@ struct page *zero_page_memmap_ptr; /* map entry for zero page */ EXPORT_SYMBOL(zero_page_memmap_ptr); #define MIN_PGT_PAGES 25UL -#define MAX_PGT_FREES_PER_PASS 16 +#define MAX_PGT_FREES_PER_PASS 16L #define PGT_FRACTION_OF_NODE_MEM 16 static inline long -- cgit v1.2.3-70-g09d2 From 4628d7cada7a19166ba8fe57f5ef0f0009694e1e Mon Sep 17 00:00:00 2001 From: Mark Maule Date: Mon, 25 Apr 2005 13:18:02 -0700 Subject: [IA64-SGI] disable TIOCA GART TLB prefetching Patch to disable SGI TIOCA GART TLB prefetching due to hw bug. Signed-off-by: Mark Maule Signed-off-by: Tony Luck --- arch/ia64/sn/pci/tioca_provider.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/pci/tioca_provider.c b/arch/ia64/sn/pci/tioca_provider.c index 2234d61cdd4..54a0dd447e7 100644 --- a/arch/ia64/sn/pci/tioca_provider.c +++ b/arch/ia64/sn/pci/tioca_provider.c @@ -171,15 +171,15 @@ tioca_gart_init(struct tioca_kernel *tioca_kern) * use agp op-combining * use GET semantics to fetch memory * participate in coherency domain - * prefetch TLB entries + * DISABLE GART PREFETCHING due to hw bug tracked in SGI PV930029 */ ca_base->ca_control1 |= CA_AGPDMA_OP_ENB_COMBDELAY; /* PV895469 ? */ ca_base->ca_control2 &= ~(CA_GART_MEM_PARAM); ca_base->ca_control2 |= (0x2ull << CA_GART_MEM_PARAM_SHFT); tioca_kern->ca_gart_iscoherent = 1; - ca_base->ca_control2 |= - (CA_GART_WR_PREFETCH_ENB | CA_GART_RD_PREFETCH_ENB); + ca_base->ca_control2 &= + ~(CA_GART_WR_PREFETCH_ENB | CA_GART_RD_PREFETCH_ENB); /* * Unmask GART fetch error interrupts. Clear residual errors first. -- cgit v1.2.3-70-g09d2 From 95ff439a517835aa2bdf725fafbb025a63984289 Mon Sep 17 00:00:00 2001 From: Russ Anderson Date: Mon, 25 Apr 2005 13:19:11 -0700 Subject: [IA64-SGI] Add new MMR definitions/Modify BTE initialiation©. patch 1: Add new MMR definitions. Modify BTE initialiation. Modify BTE copy. Signed-off-by: Russ Anderson Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/bte.c | 20 +++++++++------- include/asm-ia64/sn/bte.h | 53 +++++++++++++++++++++++++++++++++++++----- include/asm-ia64/sn/nodepda.h | 4 ++-- include/asm-ia64/sn/pda.h | 3 +-- include/asm-ia64/sn/shub_mmr.h | 20 +++++++++++++++- 5 files changed, 80 insertions(+), 20 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/bte.c b/arch/ia64/sn/kernel/bte.c index ce0bc4085ea..647deae9bfc 100644 --- a/arch/ia64/sn/kernel/bte.c +++ b/arch/ia64/sn/kernel/bte.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2000-2003 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) 2000-2005 Silicon Graphics, Inc. All Rights Reserved. */ #include @@ -170,10 +170,6 @@ retry_bteop: /* Initialize the notification to a known value. */ *bte->most_rcnt_na = BTE_WORD_BUSY; - /* Set the status reg busy bit and transfer length */ - BTE_PRINTKV(("IBLS = 0x%lx\n", IBLS_BUSY | transfer_size)); - BTE_LNSTAT_STORE(bte, IBLS_BUSY | transfer_size); - /* Set the source and destination registers */ BTE_PRINTKV(("IBSA = 0x%lx)\n", (TO_PHYS(src)))); BTE_SRC_STORE(bte, TO_PHYS(src)); @@ -188,7 +184,7 @@ retry_bteop: /* Initiate the transfer */ BTE_PRINTK(("IBCT = 0x%lx)\n", BTE_VALID_MODE(mode))); - BTE_CTRL_STORE(bte, BTE_VALID_MODE(mode)); + BTE_START_TRANSFER(bte, transfer_size, BTE_VALID_MODE(mode)); itc_end = ia64_get_itc() + (40000000 * local_cpu_data->cyc_per_usec); @@ -429,10 +425,16 @@ void bte_init_node(nodepda_t * mynodepda, cnodeid_t cnode) mynodepda->bte_recovery_timer.data = (unsigned long)mynodepda; for (i = 0; i < BTES_PER_NODE; i++) { + u64 *base_addr; + /* Which link status register should we use? */ - unsigned long link_status = (i == 0 ? IIO_IBLS0 : IIO_IBLS1); - mynodepda->bte_if[i].bte_base_addr = (u64 *) - REMOTE_HUB_ADDR(cnodeid_to_nasid(cnode), link_status); + base_addr = (u64 *) + REMOTE_HUB_ADDR(cnodeid_to_nasid(cnode), BTE_BASE_ADDR(i)); + mynodepda->bte_if[i].bte_base_addr = base_addr; + mynodepda->bte_if[i].bte_source_addr = BTE_SOURCE_ADDR(base_addr); + mynodepda->bte_if[i].bte_destination_addr = BTE_DEST_ADDR(base_addr); + mynodepda->bte_if[i].bte_control_addr = BTE_CTRL_ADDR(base_addr); + mynodepda->bte_if[i].bte_notify_addr = BTE_NOTIF_ADDR(base_addr); /* * Initialize the notification and spinlock diff --git a/include/asm-ia64/sn/bte.h b/include/asm-ia64/sn/bte.h index 0ec27f99c18..f50da3d91d0 100644 --- a/include/asm-ia64/sn/bte.h +++ b/include/asm-ia64/sn/bte.h @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2000-2004 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) 2000-2005 Silicon Graphics, Inc. All Rights Reserved. */ @@ -13,8 +13,12 @@ #include #include #include +#include #include +#include +#define IBCT_NOTIFY (0x1UL << 4) +#define IBCT_ZFIL_MODE (0x1UL << 0) /* #define BTE_DEBUG */ /* #define BTE_DEBUG_VERBOSE */ @@ -39,8 +43,36 @@ /* Define hardware */ -#define BTES_PER_NODE 2 +#define BTES_PER_NODE (is_shub2() ? 4 : 2) +#define MAX_BTES_PER_NODE 4 +#define BTE2OFF_CTRL (0) +#define BTE2OFF_SRC (SH2_BT_ENG_SRC_ADDR_0 - SH2_BT_ENG_CSR_0) +#define BTE2OFF_DEST (SH2_BT_ENG_DEST_ADDR_0 - SH2_BT_ENG_CSR_0) +#define BTE2OFF_NOTIFY (SH2_BT_ENG_NOTIF_ADDR_0 - SH2_BT_ENG_CSR_0) + +#define BTE_BASE_ADDR(interface) \ + (is_shub2() ? (interface == 0) ? SH2_BT_ENG_CSR_0 : \ + (interface == 1) ? SH2_BT_ENG_CSR_1 : \ + (interface == 2) ? SH2_BT_ENG_CSR_2 : \ + SH2_BT_ENG_CSR_3 \ + : (interface == 0) ? IIO_IBLS0 : IIO_IBLS1) + +#define BTE_SOURCE_ADDR(base) \ + (is_shub2() ? base + (BTE2OFF_SRC/8) \ + : base + (BTEOFF_SRC/8)) + +#define BTE_DEST_ADDR(base) \ + (is_shub2() ? base + (BTE2OFF_DEST/8) \ + : base + (BTEOFF_DEST/8)) + +#define BTE_CTRL_ADDR(base) \ + (is_shub2() ? base + (BTE2OFF_CTRL/8) \ + : base + (BTEOFF_CTRL/8)) + +#define BTE_NOTIF_ADDR(base) \ + (is_shub2() ? base + (BTE2OFF_NOTIFY/8) \ + : base + (BTEOFF_NOTIFY/8)) /* Define hardware modes */ #define BTE_NOTIFY (IBCT_NOTIFY) @@ -68,14 +100,18 @@ #define BTE_LNSTAT_STORE(_bte, _x) \ HUB_S(_bte->bte_base_addr, (_x)) #define BTE_SRC_STORE(_bte, _x) \ - HUB_S(_bte->bte_base_addr + (BTEOFF_SRC/8), (_x)) + HUB_S(_bte->bte_source_addr, (_x)) #define BTE_DEST_STORE(_bte, _x) \ - HUB_S(_bte->bte_base_addr + (BTEOFF_DEST/8), (_x)) + HUB_S(_bte->bte_destination_addr, (_x)) #define BTE_CTRL_STORE(_bte, _x) \ - HUB_S(_bte->bte_base_addr + (BTEOFF_CTRL/8), (_x)) + HUB_S(_bte->bte_control_addr, (_x)) #define BTE_NOTIF_STORE(_bte, _x) \ - HUB_S(_bte->bte_base_addr + (BTEOFF_NOTIFY/8), (_x)) + HUB_S(_bte->bte_notify_addr, (_x)) +#define BTE_START_TRANSFER(_bte, _len, _mode) \ + is_shub2() ? BTE_CTRL_STORE(_bte, IBLS_BUSY | (_mode << 24) | _len) \ + : BTE_LNSTAT_STORE(_bte, _len); \ + BTE_CTRL_STORE(_bte, _mode) /* Possible results from bte_copy and bte_unaligned_copy */ /* The following error codes map into the BTE hardware codes @@ -110,6 +146,10 @@ typedef enum { struct bteinfo_s { volatile u64 notify ____cacheline_aligned; u64 *bte_base_addr ____cacheline_aligned; + u64 *bte_source_addr; + u64 *bte_destination_addr; + u64 *bte_control_addr; + u64 *bte_notify_addr; spinlock_t spinlock; cnodeid_t bte_cnode; /* cnode */ int bte_error_count; /* Number of errors encountered */ @@ -117,6 +157,7 @@ struct bteinfo_s { int cleanup_active; /* Interface is locked for cleanup */ volatile bte_result_t bh_error; /* error while processing */ volatile u64 *most_rcnt_na; + struct bteinfo_s *btes_to_try[MAX_BTES_PER_NODE]; }; diff --git a/include/asm-ia64/sn/nodepda.h b/include/asm-ia64/sn/nodepda.h index 2fbde33656e..13cc1002b29 100644 --- a/include/asm-ia64/sn/nodepda.h +++ b/include/asm-ia64/sn/nodepda.h @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (C) 1992 - 1997, 2000-2004 Silicon Graphics, Inc. All rights reserved. + * Copyright (C) 1992 - 1997, 2000-2005 Silicon Graphics, Inc. All rights reserved. */ #ifndef _ASM_IA64_SN_NODEPDA_H #define _ASM_IA64_SN_NODEPDA_H @@ -43,7 +43,7 @@ struct nodepda_s { /* * The BTEs on this node are shared by the local cpus */ - struct bteinfo_s bte_if[BTES_PER_NODE]; /* Virtual Interface */ + struct bteinfo_s bte_if[MAX_BTES_PER_NODE]; /* Virtual Interface */ struct timer_list bte_recovery_timer; spinlock_t bte_recovery_lock; diff --git a/include/asm-ia64/sn/pda.h b/include/asm-ia64/sn/pda.h index e940d3647c8..cd19f17bf91 100644 --- a/include/asm-ia64/sn/pda.h +++ b/include/asm-ia64/sn/pda.h @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (C) 1992 - 1997, 2000-2004 Silicon Graphics, Inc. All rights reserved. + * Copyright (C) 1992 - 1997, 2000-2005 Silicon Graphics, Inc. All rights reserved. */ #ifndef _ASM_IA64_SN_PDA_H #define _ASM_IA64_SN_PDA_H @@ -11,7 +11,6 @@ #include #include #include -#include /* diff --git a/include/asm-ia64/sn/shub_mmr.h b/include/asm-ia64/sn/shub_mmr.h index 6ec37e816a9..2f885088e09 100644 --- a/include/asm-ia64/sn/shub_mmr.h +++ b/include/asm-ia64/sn/shub_mmr.h @@ -4,7 +4,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2001-2004 Silicon Graphics, Inc. All rights reserved. + * Copyright (c) 2001-2005 Silicon Graphics, Inc. All rights reserved. */ #ifndef _ASM_IA64_SN_SHUB_MMR_H @@ -455,4 +455,22 @@ #define SH_INT_CMPC shubmmr(SH, INT_CMPC) #define SH_INT_CMPD shubmmr(SH, INT_CMPD) +/* ========================================================================== */ +/* Register "SH2_BT_ENG_CSR_0" */ +/* Engine 0 Control and Status Register */ +/* ========================================================================== */ + +#define SH2_BT_ENG_CSR_0 0x0000000030040000 +#define SH2_BT_ENG_SRC_ADDR_0 0x0000000030040080 +#define SH2_BT_ENG_DEST_ADDR_0 0x0000000030040100 +#define SH2_BT_ENG_NOTIF_ADDR_0 0x0000000030040180 + +/* ========================================================================== */ +/* BTE interfaces 1-3 */ +/* ========================================================================== */ + +#define SH2_BT_ENG_CSR_1 0x0000000030050000 +#define SH2_BT_ENG_CSR_2 0x0000000030060000 +#define SH2_BT_ENG_CSR_3 0x0000000030070000 + #endif /* _ASM_IA64_SN_SHUB_MMR_H */ -- cgit v1.2.3-70-g09d2 From 93a07d0a0e7b013ee73fb39d4edb07b47288912e Mon Sep 17 00:00:00 2001 From: Russ Anderson Date: Mon, 25 Apr 2005 13:19:52 -0700 Subject: [IA64-SGI] Shub2 BTE support - BTE recovery code patch 2: Shub2 BTE recovery code will be implemented in SAL. Define the SAL interface. Modify bte_error to call SAL for shub2. Signed-off-by: Russ Anderson Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/bte_error.c | 76 ++++++++++++++++++++++++++--------------- arch/ia64/sn/kernel/huberror.c | 9 +++-- include/asm-ia64/sn/sn_sal.h | 19 +++++++++-- 3 files changed, 71 insertions(+), 33 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/bte_error.c b/arch/ia64/sn/kernel/bte_error.c index fd104312c6b..fcbc748ae43 100644 --- a/arch/ia64/sn/kernel/bte_error.c +++ b/arch/ia64/sn/kernel/bte_error.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2000-2004 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) 2000-2005 Silicon Graphics, Inc. All Rights Reserved. */ #include @@ -33,48 +33,28 @@ void bte_error_handler(unsigned long); * Wait until all BTE related CRBs are completed * and then reset the interfaces. */ -void bte_error_handler(unsigned long _nodepda) +void shub1_bte_error_handler(unsigned long _nodepda) { struct nodepda_s *err_nodepda = (struct nodepda_s *)_nodepda; - spinlock_t *recovery_lock = &err_nodepda->bte_recovery_lock; struct timer_list *recovery_timer = &err_nodepda->bte_recovery_timer; nasid_t nasid; int i; int valid_crbs; - unsigned long irq_flags; - volatile u64 *notify; - bte_result_t bh_error; ii_imem_u_t imem; /* II IMEM Register */ ii_icrb0_d_u_t icrbd; /* II CRB Register D */ ii_ibcr_u_t ibcr; ii_icmr_u_t icmr; ii_ieclr_u_t ieclr; - BTE_PRINTK(("bte_error_handler(%p) - %d\n", err_nodepda, + BTE_PRINTK(("shub1_bte_error_handler(%p) - %d\n", err_nodepda, smp_processor_id())); - spin_lock_irqsave(recovery_lock, irq_flags); - if ((err_nodepda->bte_if[0].bh_error == BTE_SUCCESS) && (err_nodepda->bte_if[1].bh_error == BTE_SUCCESS)) { BTE_PRINTK(("eh:%p:%d Nothing to do.\n", err_nodepda, smp_processor_id())); - spin_unlock_irqrestore(recovery_lock, irq_flags); return; } - /* - * Lock all interfaces on this node to prevent new transfers - * from being queued. - */ - for (i = 0; i < BTES_PER_NODE; i++) { - if (err_nodepda->bte_if[i].cleanup_active) { - continue; - } - spin_lock(&err_nodepda->bte_if[i].spinlock); - BTE_PRINTK(("eh:%p:%d locked %d\n", err_nodepda, - smp_processor_id(), i)); - err_nodepda->bte_if[i].cleanup_active = 1; - } /* Determine information about our hub */ nasid = cnodeid_to_nasid(err_nodepda->bte_if[0].bte_cnode); @@ -101,7 +81,6 @@ void bte_error_handler(unsigned long _nodepda) mod_timer(recovery_timer, HZ * 5); BTE_PRINTK(("eh:%p:%d Marked Giving up\n", err_nodepda, smp_processor_id())); - spin_unlock_irqrestore(recovery_lock, irq_flags); return; } if (icmr.ii_icmr_fld_s.i_crb_vld != 0) { @@ -120,8 +99,6 @@ void bte_error_handler(unsigned long _nodepda) BTE_PRINTK(("eh:%p:%d Valid %d, Giving up\n", err_nodepda, smp_processor_id(), i)); - spin_unlock_irqrestore(recovery_lock, - irq_flags); return; } } @@ -146,6 +123,51 @@ void bte_error_handler(unsigned long _nodepda) ibcr.ii_ibcr_fld_s.i_soft_reset = 1; REMOTE_HUB_S(nasid, IIO_IBCR, ibcr.ii_ibcr_regval); + del_timer(recovery_timer); +} + +/* + * Wait until all BTE related CRBs are completed + * and then reset the interfaces. + */ +void bte_error_handler(unsigned long _nodepda) +{ + struct nodepda_s *err_nodepda = (struct nodepda_s *)_nodepda; + spinlock_t *recovery_lock = &err_nodepda->bte_recovery_lock; + int i; + nasid_t nasid; + unsigned long irq_flags; + volatile u64 *notify; + bte_result_t bh_error; + + BTE_PRINTK(("bte_error_handler(%p) - %d\n", err_nodepda, + smp_processor_id())); + + spin_lock_irqsave(recovery_lock, irq_flags); + + /* + * Lock all interfaces on this node to prevent new transfers + * from being queued. + */ + for (i = 0; i < BTES_PER_NODE; i++) { + if (err_nodepda->bte_if[i].cleanup_active) { + continue; + } + spin_lock(&err_nodepda->bte_if[i].spinlock); + BTE_PRINTK(("eh:%p:%d locked %d\n", err_nodepda, + smp_processor_id(), i)); + err_nodepda->bte_if[i].cleanup_active = 1; + } + + if (is_shub1()) { + shub1_bte_error_handler(_nodepda); + } else { + nasid = cnodeid_to_nasid(err_nodepda->bte_if[0].bte_cnode); + + if (ia64_sn_bte_recovery(nasid)) + panic("bte_error_handler(): Fatal BTE Error"); + } + for (i = 0; i < BTES_PER_NODE; i++) { bh_error = err_nodepda->bte_if[i].bh_error; if (bh_error != BTE_SUCCESS) { @@ -165,8 +187,6 @@ void bte_error_handler(unsigned long _nodepda) spin_unlock(&err_nodepda->bte_if[i].spinlock); } - del_timer(recovery_timer); - spin_unlock_irqrestore(recovery_lock, irq_flags); } diff --git a/arch/ia64/sn/kernel/huberror.c b/arch/ia64/sn/kernel/huberror.c index 2bdf684c506..5c39b43ba3c 100644 --- a/arch/ia64/sn/kernel/huberror.c +++ b/arch/ia64/sn/kernel/huberror.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (C) 1992 - 1997, 2000,2002-2004 Silicon Graphics, Inc. All rights reserved. + * Copyright (C) 1992 - 1997, 2000,2002-2005 Silicon Graphics, Inc. All rights reserved. */ #include @@ -38,8 +38,11 @@ static irqreturn_t hub_eint_handler(int irq, void *arg, struct pt_regs *ep) if ((int)ret_stuff.v0) panic("hubii_eint_handler(): Fatal TIO Error"); - if (!(nasid & 1)) /* Not a TIO, handle CRB errors */ - (void)hubiio_crb_error_handler(hubdev_info); + if (is_shub1()) { + if (!(nasid & 1)) /* Not a TIO, handle CRB errors */ + (void)hubiio_crb_error_handler(hubdev_info); + } else + bte_error_handler((unsigned long)NODEPDA(nasid_to_cnodeid(nasid))); return IRQ_HANDLED; } diff --git a/include/asm-ia64/sn/sn_sal.h b/include/asm-ia64/sn/sn_sal.h index 410d356b40d..581f9a78304 100644 --- a/include/asm-ia64/sn/sn_sal.h +++ b/include/asm-ia64/sn/sn_sal.h @@ -8,7 +8,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2000-2004 Silicon Graphics, Inc. All rights reserved. + * Copyright (c) 2000-2005 Silicon Graphics, Inc. All rights reserved. */ @@ -77,7 +77,7 @@ #define SN_SAL_IOIF_GET_PCI_TOPOLOGY 0x02000059 #define SN_SAL_HUB_ERROR_INTERRUPT 0x02000060 - +#define SN_SAL_BTE_RECOVER 0x02000061 /* * Service-specific constants @@ -1023,4 +1023,19 @@ ia64_sn_ioif_get_pci_topology(u64 rack, u64 bay, u64 slot, u64 slab, return (int) rv.status; } +/* + * BTE error recovery is implemented in SAL + */ +static inline int +ia64_sn_bte_recovery(nasid_t nasid) +{ + struct ia64_sal_retval rv; + + rv.status = 0; + SAL_CALL_NOLOCK(rv, SN_SAL_BTE_RECOVER, 0, 0, 0, 0, 0, 0, 0); + if (rv.status == SALRET_NOT_IMPLEMENTED) + return 0; + return (int) rv.status; +} + #endif /* _ASM_IA64_SN_SN_SAL_H */ -- cgit v1.2.3-70-g09d2 From a37d98f6a98254c05315e0bbf45c4602942d14b1 Mon Sep 17 00:00:00 2001 From: David Mosberger-Tang Date: Mon, 25 Apr 2005 13:20:38 -0700 Subject: [IA64] fix syscall-optimization goof Sadly, I goofed in this syscall-tuning patch: ChangeSet 1.1966.1.40 2005/01/22 13:31:05 davidm@hpl.hp.com [IA64] Improve ia64_leave_syscall() for McKinley-type cores. Optimize ia64_leave_syscall() a bit better for McKinley-type cores. The patch looks big, but that's mostly due to renaming r16/r17 to r2/r3. Good for a 13 cycle improvement. The problem is that the size of the physical stacked registers was loaded into the wrong register (r3 instead of r17). Since r17 by coincidence always had the value 1, this had the effect of turning rse_clear_invalid into a no-op. That poses the risk of leaking kernel state back to user-land and is hence not acceptable. The fix below is simple, but unfortunately it costs us about 28 cycles in syscall overhead. ;-( Unfortunately, there isn't much we can do about that since those registers have to be cleared one way or another. --david Signed-off-by: Tony Luck --- arch/ia64/kernel/entry.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S index 73e23dafe8e..bd86fea49a0 100644 --- a/arch/ia64/kernel/entry.S +++ b/arch/ia64/kernel/entry.S @@ -759,7 +759,7 @@ ENTRY(ia64_leave_syscall) (pUStk) st1 [r14]=r17 addl r3=THIS_CPU(ia64_phys_stacked_size_p8),r0 ;; -(pUStk) ld4 r3=[r3] // r3 = cpu_data->phys_stacked_size_p8 +(pUStk) ld4 r17=[r3] // r17 = cpu_data->phys_stacked_size_p8 mov.m ar.csd=r0 // M2 clear ar.csd mov b6=r18 // I0 restore b6 ;; -- cgit v1.2.3-70-g09d2 From 0985ea8f2db87d32b0b750229889e55fed7458ef Mon Sep 17 00:00:00 2001 From: Mark Goodwin Date: Mon, 25 Apr 2005 13:21:54 -0700 Subject: [IA64-SGI] Altix SN add support for slots in geoid_t locator This patch against ia64-test-2.6.12 is needed for forthcoming Altix chipsets. It renames geoid_any_t to geoid_common_t and splits the 8bit 'slab' field into two 4bit fields for 'slab' and 'slot'. Similar changes in the Altix SAL will retain backward compatibility for old kernels. Signed-off-by: Mark Goodwin Signed-off-by: Tony Luck --- arch/ia64/sn/kernel/sn2/sn_hwperf.c | 71 +++++++++++++------------------------ include/asm-ia64/sn/geo.h | 45 +++++++++++++---------- include/asm-ia64/sn/sn_sal.h | 4 +-- include/asm-ia64/sn/types.h | 3 +- 4 files changed, 55 insertions(+), 68 deletions(-) (limited to 'arch') diff --git a/arch/ia64/sn/kernel/sn2/sn_hwperf.c b/arch/ia64/sn/kernel/sn2/sn_hwperf.c index e731fcb95f9..833e700fdac 100644 --- a/arch/ia64/sn/kernel/sn2/sn_hwperf.c +++ b/arch/ia64/sn/kernel/sn2/sn_hwperf.c @@ -116,7 +116,7 @@ static int sn_hwperf_geoid_to_cnode(char *location) module_id = geo_module(geoid); this_rack = MODULE_GET_RACK(module_id); this_bay = MODULE_GET_BPOS(module_id); - this_slot = 0; /* XXX */ + this_slot = geo_slot(geoid); this_slab = geo_slab(geoid); if (rack == this_rack && bay == this_bay && slot == this_slot && slab == this_slab) { @@ -176,20 +176,27 @@ static const char *sn_hwperf_get_slabname(struct sn_hwperf_object_info *obj, static void print_pci_topology(struct seq_file *s, struct sn_hwperf_object_info *obj, int *ordinal, - char *pci_topo_buf, int len) + u64 rack, u64 bay, u64 slot, u64 slab) { char *p1; char *p2; - - for (p1=pci_topo_buf; *p1 && p1 < pci_topo_buf + len;) { - if (!(p2 = strchr(p1, '\n'))) - break; - *p2 = '\0'; - seq_printf(s, "pcibus %d %s-%s\n", - *ordinal, obj->location, p1); - (*ordinal)++; - p1 = p2 + 1; + char *pg; + + if (!(pg = (char *)get_zeroed_page(GFP_KERNEL))) + return; /* ignore */ + if (ia64_sn_ioif_get_pci_topology(rack, bay, slot, slab, + __pa(pg), PAGE_SIZE) == SN_HWPERF_OP_OK) { + for (p1=pg; *p1 && p1 < pg + PAGE_SIZE;) { + if (!(p2 = strchr(p1, '\n'))) + break; + *p2 = '\0'; + seq_printf(s, "pcibus %d %s-%s\n", + *ordinal, obj->location, p1); + (*ordinal)++; + p1 = p2 + 1; + } } + free_page((unsigned long)pg); } static int sn_topology_show(struct seq_file *s, void *d) @@ -218,9 +225,7 @@ static int sn_topology_show(struct seq_file *s, void *d) u8 region_size; u16 nasid_mask; int nasid_msb; - char *pci_topo_buf; int pci_bus_ordinal = 0; - static int pci_topo_buf_len = 256; if (obj == objs) { seq_printf(s, "# sn_topology version 2\n"); @@ -299,41 +304,13 @@ static int sn_topology_show(struct seq_file *s, void *d) /* * PCI busses attached to this node, if any */ - do { - if (sn_hwperf_location_to_bpos(obj->location, - &rack, &bay, &slot, &slab)) { - break; - } - - if (!(pci_topo_buf = vmalloc(pci_topo_buf_len))) { - printk("sn_topology_show: vmalloc failed\n"); - break; - } + if (sn_hwperf_location_to_bpos(obj->location, + &rack, &bay, &slot, &slab)) { + /* export pci bus info */ + print_pci_topology(s, obj, &pci_bus_ordinal, + rack, bay, slot, slab); - e = ia64_sn_ioif_get_pci_topology(rack, bay, slot, slab, - pci_topo_buf, pci_topo_buf_len); - - switch (e) { - case SALRET_NOT_IMPLEMENTED: - case SALRET_INVALID_ARG: - /* ignore, don't print anything */ - e = SN_HWPERF_OP_OK; - break; - - case SALRET_ERROR: - /* retry with a bigger buffer */ - pci_topo_buf_len += 256; - break; - - case SN_HWPERF_OP_OK: - default: - /* export pci bus info */ - print_pci_topology(s, obj, &pci_bus_ordinal, - pci_topo_buf, pci_topo_buf_len); - break; - } - vfree(pci_topo_buf); - } while (e != SN_HWPERF_OP_OK && pci_topo_buf_len < 0x200000); + } } if (obj->ports) { diff --git a/include/asm-ia64/sn/geo.h b/include/asm-ia64/sn/geo.h index f566343d25f..84b254603b8 100644 --- a/include/asm-ia64/sn/geo.h +++ b/include/asm-ia64/sn/geo.h @@ -18,32 +18,34 @@ #define GEOID_SIZE 8 /* Would 16 be better? The size can be different on different platforms. */ -#define MAX_SLABS 0xe /* slabs per module */ +#define MAX_SLOTS 0xf /* slots per module */ +#define MAX_SLABS 0xf /* slabs per slot */ typedef unsigned char geo_type_t; /* Fields common to all substructures */ -typedef struct geo_any_s { +typedef struct geo_common_s { moduleid_t module; /* The module (box) this h/w lives in */ geo_type_t type; /* What type of h/w is named by this geoid_t */ - slabid_t slab; /* The logical assembly within the module */ -} geo_any_t; + slabid_t slab:4; /* slab (ASIC), 0 .. 15 within slot */ + slotid_t slot:4; /* slot (Blade), 0 .. 15 within module */ +} geo_common_t; /* Additional fields for particular types of hardware */ typedef struct geo_node_s { - geo_any_t any; /* No additional fields needed */ + geo_common_t common; /* No additional fields needed */ } geo_node_t; typedef struct geo_rtr_s { - geo_any_t any; /* No additional fields needed */ + geo_common_t common; /* No additional fields needed */ } geo_rtr_t; typedef struct geo_iocntl_s { - geo_any_t any; /* No additional fields needed */ + geo_common_t common; /* No additional fields needed */ } geo_iocntl_t; typedef struct geo_pcicard_s { - geo_iocntl_t any; + geo_iocntl_t common; char bus; /* Bus/widget number */ char slot; /* PCI slot number */ } geo_pcicard_t; @@ -62,14 +64,14 @@ typedef struct geo_mem_s { typedef union geoid_u { - geo_any_t any; - geo_node_t node; + geo_common_t common; + geo_node_t node; geo_iocntl_t iocntl; geo_pcicard_t pcicard; - geo_rtr_t rtr; - geo_cpu_t cpu; - geo_mem_t mem; - char padsize[GEOID_SIZE]; + geo_rtr_t rtr; + geo_cpu_t cpu; + geo_mem_t mem; + char padsize[GEOID_SIZE]; } geoid_t; @@ -104,19 +106,26 @@ typedef union geoid_u { #define INVALID_CNODEID ((cnodeid_t)-1) #define INVALID_PNODEID ((pnodeid_t)-1) #define INVALID_SLAB (slabid_t)-1 +#define INVALID_SLOT (slotid_t)-1 #define INVALID_MODULE ((moduleid_t)-1) #define INVALID_PARTID ((partid_t)-1) static inline slabid_t geo_slab(geoid_t g) { - return (g.any.type == GEO_TYPE_INVALID) ? - INVALID_SLAB : g.any.slab; + return (g.common.type == GEO_TYPE_INVALID) ? + INVALID_SLAB : g.common.slab; +} + +static inline slotid_t geo_slot(geoid_t g) +{ + return (g.common.type == GEO_TYPE_INVALID) ? + INVALID_SLOT : g.common.slot; } static inline moduleid_t geo_module(geoid_t g) { - return (g.any.type == GEO_TYPE_INVALID) ? - INVALID_MODULE : g.any.module; + return (g.common.type == GEO_TYPE_INVALID) ? + INVALID_MODULE : g.common.module; } extern geoid_t cnodeid_get_geoid(cnodeid_t cnode); diff --git a/include/asm-ia64/sn/sn_sal.h b/include/asm-ia64/sn/sn_sal.h index 581f9a78304..123c1a50a9d 100644 --- a/include/asm-ia64/sn/sn_sal.h +++ b/include/asm-ia64/sn/sn_sal.h @@ -74,10 +74,10 @@ #define SN_SAL_IOIF_GET_PCIBUS_INFO 0x02000056 #define SN_SAL_IOIF_GET_PCIDEV_INFO 0x02000057 #define SN_SAL_IOIF_GET_WIDGET_DMAFLUSH_LIST 0x02000058 -#define SN_SAL_IOIF_GET_PCI_TOPOLOGY 0x02000059 #define SN_SAL_HUB_ERROR_INTERRUPT 0x02000060 #define SN_SAL_BTE_RECOVER 0x02000061 +#define SN_SAL_IOIF_GET_PCI_TOPOLOGY 0x02000062 /* * Service-specific constants @@ -1015,7 +1015,7 @@ ia64_sn_hwperf_op(nasid_t nasid, u64 opcode, u64 a0, u64 a1, u64 a2, static inline int ia64_sn_ioif_get_pci_topology(u64 rack, u64 bay, u64 slot, u64 slab, - char *buf, u64 len) + u64 buf, u64 len) { struct ia64_sal_retval rv; SAL_CALL_NOLOCK(rv, SN_SAL_IOIF_GET_PCI_TOPOLOGY, diff --git a/include/asm-ia64/sn/types.h b/include/asm-ia64/sn/types.h index 586ed47cae9..8e04ee211e5 100644 --- a/include/asm-ia64/sn/types.h +++ b/include/asm-ia64/sn/types.h @@ -16,7 +16,8 @@ typedef signed short nasid_t; /* node id in numa-as-id space */ typedef signed char partid_t; /* partition ID type */ typedef unsigned int moduleid_t; /* user-visible module number type */ typedef unsigned int cmoduleid_t; /* kernel compact module id type */ -typedef signed char slabid_t; +typedef unsigned char slotid_t; /* slot (blade) within module */ +typedef unsigned char slabid_t; /* slab (asic) within slot */ typedef u64 nic_t; typedef unsigned long iopaddr_t; typedef unsigned long paddr_t; -- cgit v1.2.3-70-g09d2 From f0a8d3c9ec1f82d2a41faa6c46b8db7bd5b1eb8d Mon Sep 17 00:00:00 2001 From: Tony Luck Date: Mon, 25 Apr 2005 13:22:44 -0700 Subject: [IA64] Need to handle lfetch in "no_context" case. Thanks to Mark for tracking down this one. Users of __copy_from_user_inatomic() will be sad if we don't handle lfetch faults for the "no_context" case. Signed-off-by: Tony Luck --- arch/ia64/mm/fault.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c index da859125aae..4174ec999dd 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -209,10 +209,13 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re } no_context: - if (isr & IA64_ISR_SP) { + if ((isr & IA64_ISR_SP) + || ((isr & IA64_ISR_NA) && (isr & IA64_ISR_CODE_MASK) == IA64_ISR_CODE_LFETCH)) + { /* - * This fault was due to a speculative load set the "ed" bit in the psr to - * ensure forward progress (target register will get a NaT). + * This fault was due to a speculative load or lfetch.fault, set the "ed" + * bit in the psr to ensure forward progress. (Target register will get a + * NaT for ld.s, lfetch will be canceled.) */ ia64_psr(regs)->ed = 1; return; -- cgit v1.2.3-70-g09d2 From 6118ec847e8e35393efc0f88394c2f5dd48c3313 Mon Sep 17 00:00:00 2001 From: Keith Owens Date: Mon, 25 Apr 2005 13:23:47 -0700 Subject: [IA64] __copy_user breaks on unaligned src memcpy_mck.S::__copy_user breaks in the prefetch code under these conditions :- * src is unaligned and * dst is near the end of a page and * the page after dst is unmapped. Signed-off-by: Keith Owens Signed-off-by: Tony Luck --- arch/ia64/lib/memcpy_mck.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ia64/lib/memcpy_mck.S b/arch/ia64/lib/memcpy_mck.S index 6f26ef7cc23..3c2cd2f04db 100644 --- a/arch/ia64/lib/memcpy_mck.S +++ b/arch/ia64/lib/memcpy_mck.S @@ -300,7 +300,7 @@ EK(.ex_handler, (p[D]) st8 [dst1] = t15, 4*8) add src_pre_mem=0,src0 // prefetch src pointer add dst_pre_mem=0,dst0 // prefetch dest pointer and src0=-8,src0 // 1st src pointer -(p7) mov ar.lc = r21 +(p7) mov ar.lc = cnt (p8) mov ar.lc = r0 ;; TEXT_ALIGN(32) -- cgit v1.2.3-70-g09d2 From e927ecb05e1ce4bbb1e10f57008c94994e2160f5 Mon Sep 17 00:00:00 2001 From: Suresh Siddha Date: Mon, 25 Apr 2005 13:25:06 -0700 Subject: [IA64] multi-core/multi-thread identification Version 3 - rediffed to apply on top of Ashok's hotplug cpu patch. /proc/cpuinfo output in step with x86. This is an updated MC/MT identification patch based on the previous discussions on list. Add the Multi-core and Multi-threading detection for IPF. - Add new core and threading related fields in /proc/cpuinfo. Physical id Core id Thread id Siblings - setup the cpu_core_map and cpu_sibling_map appropriately - Handles Hot plug CPU Signed-off-by: Suresh Siddha Signed-off-by: Gordon Jin Signed-off-by: Rohit Seth Signed-off-by: Tony Luck --- arch/ia64/kernel/setup.c | 69 ++++++++++++++- arch/ia64/kernel/smpboot.c | 206 +++++++++++++++++++++++++++++++++++++++++++ include/asm-ia64/pal.h | 68 ++++++++++++++ include/asm-ia64/processor.h | 7 ++ include/asm-ia64/sal.h | 12 +++ include/asm-ia64/smp.h | 5 ++ 6 files changed, 365 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c index f05650c801d..88043841fb8 100644 --- a/arch/ia64/kernel/setup.c +++ b/arch/ia64/kernel/setup.c @@ -4,10 +4,15 @@ * Copyright (C) 1998-2001, 2003-2004 Hewlett-Packard Co * David Mosberger-Tang * Stephane Eranian - * Copyright (C) 2000, Rohit Seth + * Copyright (C) 2000, 2004 Intel Corp + * Rohit Seth + * Suresh Siddha + * Gordon Jin * Copyright (C) 1999 VA Linux Systems * Copyright (C) 1999 Walt Drummond * + * 12/26/04 S.Siddha, G.Jin, R.Seth + * Add multi-threading and multi-core detection * 11/12/01 D.Mosberger Convert get_cpuinfo() to seq_file based show_cpuinfo(). * 04/04/00 D.Mosberger renamed cpu_initialized to cpu_online_map * 03/31/00 R.Seth cpu_initialized and current->processor fixes @@ -296,6 +301,34 @@ mark_bsp_online (void) #endif } +#ifdef CONFIG_SMP +static void +check_for_logical_procs (void) +{ + pal_logical_to_physical_t info; + s64 status; + + status = ia64_pal_logical_to_phys(0, &info); + if (status == -1) { + printk(KERN_INFO "No logical to physical processor mapping " + "available\n"); + return; + } + if (status) { + printk(KERN_ERR "ia64_pal_logical_to_phys failed with %ld\n", + status); + return; + } + /* + * Total number of siblings that BSP has. Though not all of them + * may have booted successfully. The correct number of siblings + * booted is in info.overview_num_log. + */ + smp_num_siblings = info.overview_tpc; + smp_num_cpucores = info.overview_cpp; +} +#endif + void __init setup_arch (char **cmdline_p) { @@ -356,6 +389,19 @@ setup_arch (char **cmdline_p) #ifdef CONFIG_SMP cpu_physical_id(0) = hard_smp_processor_id(); + + cpu_set(0, cpu_sibling_map[0]); + cpu_set(0, cpu_core_map[0]); + + check_for_logical_procs(); + if (smp_num_cpucores > 1) + printk(KERN_INFO + "cpu package is Multi-Core capable: number of cores=%d\n", + smp_num_cpucores); + if (smp_num_siblings > 1) + printk(KERN_INFO + "cpu package is Multi-Threading capable: number of siblings=%d\n", + smp_num_siblings); #endif cpu_init(); /* initialize the bootstrap CPU */ @@ -459,12 +505,23 @@ show_cpuinfo (struct seq_file *m, void *v) "cpu regs : %u\n" "cpu MHz : %lu.%06lu\n" "itc MHz : %lu.%06lu\n" - "BogoMIPS : %lu.%02lu\n\n", + "BogoMIPS : %lu.%02lu\n", cpunum, c->vendor, family, c->model, c->revision, c->archrev, features, c->ppn, c->number, c->proc_freq / 1000000, c->proc_freq % 1000000, c->itc_freq / 1000000, c->itc_freq % 1000000, lpj*HZ/500000, (lpj*HZ/5000) % 100); +#ifdef CONFIG_SMP + if (c->threads_per_core > 1 || c->cores_per_socket > 1) + seq_printf(m, + "physical id: %u\n" + "core id : %u\n" + "thread id : %u\n", + c->socket_id, c->core_id, c->thread_id); + seq_printf(m, "siblings : %u\n", c->num_log); +#endif + seq_printf(m,"\n"); + return 0; } @@ -533,6 +590,14 @@ identify_cpu (struct cpuinfo_ia64 *c) memcpy(c->vendor, cpuid.field.vendor, 16); #ifdef CONFIG_SMP c->cpu = smp_processor_id(); + + /* below default values will be overwritten by identify_siblings() + * for Multi-Threading/Multi-Core capable cpu's + */ + c->threads_per_core = c->cores_per_socket = c->num_log = 1; + c->socket_id = -1; + + identify_siblings(c); #endif c->ppn = cpuid.field.ppn; c->number = cpuid.field.number; diff --git a/arch/ia64/kernel/smpboot.c b/arch/ia64/kernel/smpboot.c index dbc6b610cc6..0d5ee57c986 100644 --- a/arch/ia64/kernel/smpboot.c +++ b/arch/ia64/kernel/smpboot.c @@ -3,6 +3,11 @@ * * Copyright (C) 1998-2003, 2005 Hewlett-Packard Co * David Mosberger-Tang + * Copyright (C) 2001, 2004-2005 Intel Corp + * Rohit Seth + * Suresh Siddha + * Gordon Jin + * Ashok Raj * * 01/05/16 Rohit Seth Moved SMP booting functions from smp.c to here. * 01/04/27 David Mosberger Added ITC synching code. @@ -10,6 +15,11 @@ * smp_boot_cpus()/smp_commence() is replaced by * smp_prepare_cpus()/__cpu_up()/smp_cpus_done(). * 04/06/21 Ashok Raj Added CPU Hotplug Support + * 04/12/26 Jin Gordon + * 04/12/26 Rohit Seth + * Add multi-threading and multi-core detection + * 05/01/30 Suresh Siddha + * Setup cpu_sibling_map and cpu_core_map */ #include @@ -122,6 +132,11 @@ EXPORT_SYMBOL(cpu_online_map); cpumask_t cpu_possible_map; EXPORT_SYMBOL(cpu_possible_map); +cpumask_t cpu_core_map[NR_CPUS] __cacheline_aligned; +cpumask_t cpu_sibling_map[NR_CPUS] __cacheline_aligned; +int smp_num_siblings = 1; +int smp_num_cpucores = 1; + /* which logical CPU number maps to which CPU (physical APIC ID) */ volatile int ia64_cpu_to_sapicid[NR_CPUS]; EXPORT_SYMBOL(ia64_cpu_to_sapicid); @@ -598,7 +613,68 @@ void __devinit smp_prepare_boot_cpu(void) cpu_set(smp_processor_id(), cpu_callin_map); } +/* + * mt_info[] is a temporary store for all info returned by + * PAL_LOGICAL_TO_PHYSICAL, to be copied into cpuinfo_ia64 when the + * specific cpu comes. + */ +static struct { + __u32 socket_id; + __u16 core_id; + __u16 thread_id; + __u16 proc_fixed_addr; + __u8 valid; +}mt_info[NR_CPUS] __devinit; + #ifdef CONFIG_HOTPLUG_CPU +static inline void +remove_from_mtinfo(int cpu) +{ + int i; + + for_each_cpu(i) + if (mt_info[i].valid && mt_info[i].socket_id == + cpu_data(cpu)->socket_id) + mt_info[i].valid = 0; +} + +static inline void +clear_cpu_sibling_map(int cpu) +{ + int i; + + for_each_cpu_mask(i, cpu_sibling_map[cpu]) + cpu_clear(cpu, cpu_sibling_map[i]); + for_each_cpu_mask(i, cpu_core_map[cpu]) + cpu_clear(cpu, cpu_core_map[i]); + + cpu_sibling_map[cpu] = cpu_core_map[cpu] = CPU_MASK_NONE; +} + +static void +remove_siblinginfo(int cpu) +{ + int last = 0; + + if (cpu_data(cpu)->threads_per_core == 1 && + cpu_data(cpu)->cores_per_socket == 1) { + cpu_clear(cpu, cpu_core_map[cpu]); + cpu_clear(cpu, cpu_sibling_map[cpu]); + return; + } + + last = (cpus_weight(cpu_core_map[cpu]) == 1 ? 1 : 0); + + /* remove it from all sibling map's */ + clear_cpu_sibling_map(cpu); + + /* if this cpu is the last in the core group, remove all its info + * from mt_info structure + */ + if (last) + remove_from_mtinfo(cpu); +} + extern void fixup_irqs(void); /* must be called with cpucontrol mutex held */ int __cpu_disable(void) @@ -611,6 +687,7 @@ int __cpu_disable(void) if (cpu == 0) return -EBUSY; + remove_siblinginfo(cpu); fixup_irqs(); local_flush_tlb_all(); cpu_clear(cpu, cpu_callin_map); @@ -663,6 +740,23 @@ smp_cpus_done (unsigned int dummy) (int)num_online_cpus(), bogosum/(500000/HZ), (bogosum/(5000/HZ))%100); } +static inline void __devinit +set_cpu_sibling_map(int cpu) +{ + int i; + + for_each_online_cpu(i) { + if ((cpu_data(cpu)->socket_id == cpu_data(i)->socket_id)) { + cpu_set(i, cpu_core_map[cpu]); + cpu_set(cpu, cpu_core_map[i]); + if (cpu_data(cpu)->core_id == cpu_data(i)->core_id) { + cpu_set(i, cpu_sibling_map[cpu]); + cpu_set(cpu, cpu_sibling_map[i]); + } + } + } +} + int __devinit __cpu_up (unsigned int cpu) { @@ -685,6 +779,15 @@ __cpu_up (unsigned int cpu) if (ret < 0) return ret; + if (cpu_data(cpu)->threads_per_core == 1 && + cpu_data(cpu)->cores_per_socket == 1) { + cpu_set(cpu, cpu_sibling_map[cpu]); + cpu_set(cpu, cpu_core_map[cpu]); + return 0; + } + + set_cpu_sibling_map(cpu); + return 0; } @@ -712,3 +815,106 @@ init_smp_config(void) ia64_sal_strerror(sal_ret)); } +static inline int __devinit +check_for_mtinfo_index(void) +{ + int i; + + for_each_cpu(i) + if (!mt_info[i].valid) + return i; + + return -1; +} + +/* + * Search the mt_info to find out if this socket's cid/tid information is + * cached or not. If the socket exists, fill in the core_id and thread_id + * in cpuinfo + */ +static int __devinit +check_for_new_socket(__u16 logical_address, struct cpuinfo_ia64 *c) +{ + int i; + __u32 sid = c->socket_id; + + for_each_cpu(i) { + if (mt_info[i].valid && mt_info[i].proc_fixed_addr == logical_address + && mt_info[i].socket_id == sid) { + c->core_id = mt_info[i].core_id; + c->thread_id = mt_info[i].thread_id; + return 1; /* not a new socket */ + } + } + return 0; +} + +/* + * identify_siblings(cpu) gets called from identify_cpu. This populates the + * information related to logical execution units in per_cpu_data structure. + */ +void __devinit +identify_siblings(struct cpuinfo_ia64 *c) +{ + s64 status; + u16 pltid; + u64 proc_fixed_addr; + int count, i; + pal_logical_to_physical_t info; + + if (smp_num_cpucores == 1 && smp_num_siblings == 1) + return; + + if ((status = ia64_pal_logical_to_phys(0, &info)) != PAL_STATUS_SUCCESS) { + printk(KERN_ERR "ia64_pal_logical_to_phys failed with %ld\n", + status); + return; + } + if ((status = ia64_sal_physical_id_info(&pltid)) != PAL_STATUS_SUCCESS) { + printk(KERN_ERR "ia64_sal_pltid failed with %ld\n", status); + return; + } + if ((status = ia64_pal_fixed_addr(&proc_fixed_addr)) != PAL_STATUS_SUCCESS) { + printk(KERN_ERR "ia64_pal_fixed_addr failed with %ld\n", status); + return; + } + + c->socket_id = (pltid << 8) | info.overview_ppid; + c->cores_per_socket = info.overview_cpp; + c->threads_per_core = info.overview_tpc; + count = c->num_log = info.overview_num_log; + + /* If the thread and core id information is already cached, then + * we will simply update cpu_info and return. Otherwise, we will + * do the PAL calls and cache core and thread id's of all the siblings. + */ + if (check_for_new_socket(proc_fixed_addr, c)) + return; + + for (i = 0; i < count; i++) { + int index; + + if (i && (status = ia64_pal_logical_to_phys(i, &info)) + != PAL_STATUS_SUCCESS) { + printk(KERN_ERR "ia64_pal_logical_to_phys failed" + " with %ld\n", status); + return; + } + if (info.log2_la == proc_fixed_addr) { + c->core_id = info.log1_cid; + c->thread_id = info.log1_tid; + } + + index = check_for_mtinfo_index(); + /* We will not do the mt_info caching optimization in this case. + */ + if (index < 0) + continue; + + mt_info[index].valid = 1; + mt_info[index].socket_id = c->socket_id; + mt_info[index].core_id = info.log1_cid; + mt_info[index].thread_id = info.log1_tid; + mt_info[index].proc_fixed_addr = info.log2_la; + } +} diff --git a/include/asm-ia64/pal.h b/include/asm-ia64/pal.h index 5dd477ffb88..2303a10ee59 100644 --- a/include/asm-ia64/pal.h +++ b/include/asm-ia64/pal.h @@ -67,6 +67,7 @@ #define PAL_REGISTER_INFO 39 /* return AR and CR register information*/ #define PAL_SHUTDOWN 40 /* enter processor shutdown state */ #define PAL_PREFETCH_VISIBILITY 41 /* Make Processor Prefetches Visible */ +#define PAL_LOGICAL_TO_PHYSICAL 42 /* returns information on logical to physical processor mapping */ #define PAL_COPY_PAL 256 /* relocate PAL procedures and PAL PMI */ #define PAL_HALT_INFO 257 /* return the low power capabilities of processor */ @@ -1559,6 +1560,73 @@ ia64_pal_prefetch_visibility (s64 trans_type) return iprv.status; } +/* data structure for getting information on logical to physical mappings */ +typedef union pal_log_overview_u { + struct { + u64 num_log :16, /* Total number of logical + * processors on this die + */ + tpc :8, /* Threads per core */ + reserved3 :8, /* Reserved */ + cpp :8, /* Cores per processor */ + reserved2 :8, /* Reserved */ + ppid :8, /* Physical processor ID */ + reserved1 :8; /* Reserved */ + } overview_bits; + u64 overview_data; +} pal_log_overview_t; + +typedef union pal_proc_n_log_info1_u{ + struct { + u64 tid :16, /* Thread id */ + reserved2 :16, /* Reserved */ + cid :16, /* Core id */ + reserved1 :16; /* Reserved */ + } ppli1_bits; + u64 ppli1_data; +} pal_proc_n_log_info1_t; + +typedef union pal_proc_n_log_info2_u { + struct { + u64 la :16, /* Logical address */ + reserved :48; /* Reserved */ + } ppli2_bits; + u64 ppli2_data; +} pal_proc_n_log_info2_t; + +typedef struct pal_logical_to_physical_s +{ + pal_log_overview_t overview; + pal_proc_n_log_info1_t ppli1; + pal_proc_n_log_info2_t ppli2; +} pal_logical_to_physical_t; + +#define overview_num_log overview.overview_bits.num_log +#define overview_tpc overview.overview_bits.tpc +#define overview_cpp overview.overview_bits.cpp +#define overview_ppid overview.overview_bits.ppid +#define log1_tid ppli1.ppli1_bits.tid +#define log1_cid ppli1.ppli1_bits.cid +#define log2_la ppli2.ppli2_bits.la + +/* Get information on logical to physical processor mappings. */ +static inline s64 +ia64_pal_logical_to_phys(u64 proc_number, pal_logical_to_physical_t *mapping) +{ + struct ia64_pal_retval iprv; + + PAL_CALL(iprv, PAL_LOGICAL_TO_PHYSICAL, proc_number, 0, 0); + + if (iprv.status == PAL_STATUS_SUCCESS) + { + if (proc_number == 0) + mapping->overview.overview_data = iprv.v0; + mapping->ppli1.ppli1_data = iprv.v1; + mapping->ppli2.ppli2_data = iprv.v2; + } + + return iprv.status; +} #endif /* __ASSEMBLY__ */ #endif /* _ASM_IA64_PAL_H */ diff --git a/include/asm-ia64/processor.h b/include/asm-ia64/processor.h index 983798ec179..9e1ba8b7fb6 100644 --- a/include/asm-ia64/processor.h +++ b/include/asm-ia64/processor.h @@ -148,6 +148,13 @@ struct cpuinfo_ia64 { #ifdef CONFIG_SMP __u64 loops_per_jiffy; int cpu; + __u32 socket_id; /* physical processor socket id */ + __u16 core_id; /* core id */ + __u16 thread_id; /* thread id */ + __u16 num_log; /* Total number of logical processors on + * this socket that were successfully booted */ + __u8 cores_per_socket; /* Cores per processor socket */ + __u8 threads_per_core; /* Threads per core */ #endif /* CPUID-derived information: */ diff --git a/include/asm-ia64/sal.h b/include/asm-ia64/sal.h index 240676f7539..29df88bdd2b 100644 --- a/include/asm-ia64/sal.h +++ b/include/asm-ia64/sal.h @@ -91,6 +91,7 @@ extern spinlock_t sal_lock; #define SAL_PCI_CONFIG_READ 0x01000010 #define SAL_PCI_CONFIG_WRITE 0x01000011 #define SAL_FREQ_BASE 0x01000012 +#define SAL_PHYSICAL_ID_INFO 0x01000013 #define SAL_UPDATE_PAL 0x01000020 @@ -815,6 +816,17 @@ ia64_sal_update_pal (u64 param_buf, u64 scratch_buf, u64 scratch_buf_size, return isrv.status; } +/* Get physical processor die mapping in the platform. */ +static inline s64 +ia64_sal_physical_id_info(u16 *splid) +{ + struct ia64_sal_retval isrv; + SAL_CALL(isrv, SAL_PHYSICAL_ID_INFO, 0, 0, 0, 0, 0, 0, 0); + if (splid) + *splid = isrv.v0; + return isrv.status; +} + extern unsigned long sal_platform_features; extern int (*salinfo_platform_oemdata)(const u8 *, u8 **, u64 *); diff --git a/include/asm-ia64/smp.h b/include/asm-ia64/smp.h index c4a227acfeb..3ba1a061e4a 100644 --- a/include/asm-ia64/smp.h +++ b/include/asm-ia64/smp.h @@ -56,6 +56,10 @@ extern struct smp_boot_data { extern char no_int_routing __devinitdata; extern cpumask_t cpu_online_map; +extern cpumask_t cpu_core_map[NR_CPUS]; +extern cpumask_t cpu_sibling_map[NR_CPUS]; +extern int smp_num_siblings; +extern int smp_num_cpucores; extern void __iomem *ipi_base_addr; extern unsigned char smp_int_redirect; @@ -124,6 +128,7 @@ extern int smp_call_function_single (int cpuid, void (*func) (void *info), void extern void smp_send_reschedule (int cpu); extern void lock_ipi_calllock(void); extern void unlock_ipi_calllock(void); +extern void identify_siblings (struct cpuinfo_ia64 *); #else -- cgit v1.2.3-70-g09d2 From 24eeb568aeeaee771b9f0a6fd6f5d01040a887da Mon Sep 17 00:00:00 2001 From: Kenji Kaneshige Date: Mon, 25 Apr 2005 13:26:23 -0700 Subject: [IA64] vector sharing (Large I/O system support) Current ia64 linux cannot handle greater than 184 interrupt sources because of the lack of vectors. The following patch enables ia64 linux to handle greater than 184 interrupt sources by allowing the same vector number to be shared by multiple IOSAPIC's RTEs. The design of this patch is besed on "Intel(R) Itanium(R) Processor Family Interrupt Architecture Guide". Even if you don't have a large I/O system, you can see the behavior of vector sharing by changing IOSAPIC_LAST_DEVICE_VECTOR to fewer value. Signed-off-by: Kenji Kaneshige Signed-off-by: Tony Luck --- arch/ia64/kernel/iosapic.c | 358 +++++++++++++++++++++++++++++++++----------- arch/ia64/kernel/irq_ia64.c | 16 +- include/asm-ia64/hw_irq.h | 1 + 3 files changed, 286 insertions(+), 89 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c index c15be5c38f5..11a221cc8dc 100644 --- a/arch/ia64/kernel/iosapic.c +++ b/arch/ia64/kernel/iosapic.c @@ -79,6 +79,7 @@ #include #include #include +#include #include #include @@ -98,19 +99,30 @@ #define DBG(fmt...) #endif +#define NR_PREALLOCATE_RTE_ENTRIES (PAGE_SIZE / sizeof(struct iosapic_rte_info)) +#define RTE_PREALLOCATED (1) + static DEFINE_SPINLOCK(iosapic_lock); /* These tables map IA-64 vectors to the IOSAPIC pin that generates this vector. */ -static struct iosapic_intr_info { +struct iosapic_rte_info { + struct list_head rte_list; /* node in list of RTEs sharing the same vector */ char __iomem *addr; /* base address of IOSAPIC */ - u32 low32; /* current value of low word of Redirection table entry */ unsigned int gsi_base; /* first GSI assigned to this IOSAPIC */ - char rte_index; /* IOSAPIC RTE index (-1 => not an IOSAPIC interrupt) */ + char rte_index; /* IOSAPIC RTE index */ + int refcnt; /* reference counter */ + unsigned int flags; /* flags */ +} ____cacheline_aligned; + +static struct iosapic_intr_info { + struct list_head rtes; /* RTEs using this vector (empty => not an IOSAPIC interrupt) */ + int count; /* # of RTEs that shares this vector */ + u32 low32; /* current value of low word of Redirection table entry */ + unsigned int dest; /* destination CPU physical ID */ unsigned char dmode : 3; /* delivery mode (see iosapic.h) */ unsigned char polarity: 1; /* interrupt polarity (see iosapic.h) */ unsigned char trigger : 1; /* trigger mode (see iosapic.h) */ - int refcnt; /* reference counter */ } iosapic_intr_info[IA64_NUM_VECTORS]; static struct iosapic { @@ -126,6 +138,8 @@ static int num_iosapic; static unsigned char pcat_compat __initdata; /* 8259 compatibility flag */ +static int iosapic_kmalloc_ok; +static LIST_HEAD(free_rte_list); /* * Find an IOSAPIC associated with a GSI @@ -147,10 +161,12 @@ static inline int _gsi_to_vector (unsigned int gsi) { struct iosapic_intr_info *info; + struct iosapic_rte_info *rte; for (info = iosapic_intr_info; info < iosapic_intr_info + IA64_NUM_VECTORS; ++info) - if (info->gsi_base + info->rte_index == gsi) - return info - iosapic_intr_info; + list_for_each_entry(rte, &info->rtes, rte_list) + if (rte->gsi_base + rte->rte_index == gsi) + return info - iosapic_intr_info; return -1; } @@ -167,33 +183,52 @@ gsi_to_vector (unsigned int gsi) int gsi_to_irq (unsigned int gsi) { + unsigned long flags; + int irq; /* * XXX fix me: this assumes an identity mapping vetween IA-64 vector and Linux irq * numbers... */ - return _gsi_to_vector(gsi); + spin_lock_irqsave(&iosapic_lock, flags); + { + irq = _gsi_to_vector(gsi); + } + spin_unlock_irqrestore(&iosapic_lock, flags); + + return irq; +} + +static struct iosapic_rte_info *gsi_vector_to_rte(unsigned int gsi, unsigned int vec) +{ + struct iosapic_rte_info *rte; + + list_for_each_entry(rte, &iosapic_intr_info[vec].rtes, rte_list) + if (rte->gsi_base + rte->rte_index == gsi) + return rte; + return NULL; } static void -set_rte (unsigned int vector, unsigned int dest, int mask) +set_rte (unsigned int gsi, unsigned int vector, unsigned int dest, int mask) { unsigned long pol, trigger, dmode; u32 low32, high32; char __iomem *addr; int rte_index; char redir; + struct iosapic_rte_info *rte; DBG(KERN_DEBUG"IOSAPIC: routing vector %d to 0x%x\n", vector, dest); - rte_index = iosapic_intr_info[vector].rte_index; - if (rte_index < 0) + rte = gsi_vector_to_rte(gsi, vector); + if (!rte) return; /* not an IOSAPIC interrupt */ - addr = iosapic_intr_info[vector].addr; + rte_index = rte->rte_index; + addr = rte->addr; pol = iosapic_intr_info[vector].polarity; trigger = iosapic_intr_info[vector].trigger; dmode = iosapic_intr_info[vector].dmode; - vector &= (~IA64_IRQ_REDIRECTED); redir = (dmode == IOSAPIC_LOWEST_PRIORITY) ? 1 : 0; @@ -221,6 +256,7 @@ set_rte (unsigned int vector, unsigned int dest, int mask) iosapic_write(addr, IOSAPIC_RTE_HIGH(rte_index), high32); iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); iosapic_intr_info[vector].low32 = low32; + iosapic_intr_info[vector].dest = dest; } static void @@ -237,18 +273,20 @@ mask_irq (unsigned int irq) u32 low32; int rte_index; ia64_vector vec = irq_to_vector(irq); + struct iosapic_rte_info *rte; - addr = iosapic_intr_info[vec].addr; - rte_index = iosapic_intr_info[vec].rte_index; - - if (rte_index < 0) + if (list_empty(&iosapic_intr_info[vec].rtes)) return; /* not an IOSAPIC interrupt! */ spin_lock_irqsave(&iosapic_lock, flags); { /* set only the mask bit */ low32 = iosapic_intr_info[vec].low32 |= IOSAPIC_MASK; - iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); + list_for_each_entry(rte, &iosapic_intr_info[vec].rtes, rte_list) { + addr = rte->addr; + rte_index = rte->rte_index; + iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); + } } spin_unlock_irqrestore(&iosapic_lock, flags); } @@ -261,16 +299,19 @@ unmask_irq (unsigned int irq) u32 low32; int rte_index; ia64_vector vec = irq_to_vector(irq); + struct iosapic_rte_info *rte; - addr = iosapic_intr_info[vec].addr; - rte_index = iosapic_intr_info[vec].rte_index; - if (rte_index < 0) + if (list_empty(&iosapic_intr_info[vec].rtes)) return; /* not an IOSAPIC interrupt! */ spin_lock_irqsave(&iosapic_lock, flags); { low32 = iosapic_intr_info[vec].low32 &= ~IOSAPIC_MASK; - iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); + list_for_each_entry(rte, &iosapic_intr_info[vec].rtes, rte_list) { + addr = rte->addr; + rte_index = rte->rte_index; + iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); + } } spin_unlock_irqrestore(&iosapic_lock, flags); } @@ -286,6 +327,7 @@ iosapic_set_affinity (unsigned int irq, cpumask_t mask) char __iomem *addr; int redir = (irq & IA64_IRQ_REDIRECTED) ? 1 : 0; ia64_vector vec; + struct iosapic_rte_info *rte; irq &= (~IA64_IRQ_REDIRECTED); vec = irq_to_vector(irq); @@ -295,10 +337,7 @@ iosapic_set_affinity (unsigned int irq, cpumask_t mask) dest = cpu_physical_id(first_cpu(mask)); - rte_index = iosapic_intr_info[vec].rte_index; - addr = iosapic_intr_info[vec].addr; - - if (rte_index < 0) + if (list_empty(&iosapic_intr_info[vec].rtes)) return; /* not an IOSAPIC interrupt */ set_irq_affinity_info(irq, dest, redir); @@ -318,8 +357,13 @@ iosapic_set_affinity (unsigned int irq, cpumask_t mask) low32 |= (IOSAPIC_FIXED << IOSAPIC_DELIVERY_SHIFT); iosapic_intr_info[vec].low32 = low32; - iosapic_write(addr, IOSAPIC_RTE_HIGH(rte_index), high32); - iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); + iosapic_intr_info[vec].dest = dest; + list_for_each_entry(rte, &iosapic_intr_info[vec].rtes, rte_list) { + addr = rte->addr; + rte_index = rte->rte_index; + iosapic_write(addr, IOSAPIC_RTE_HIGH(rte_index), high32); + iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32); + } } spin_unlock_irqrestore(&iosapic_lock, flags); #endif @@ -340,9 +384,11 @@ static void iosapic_end_level_irq (unsigned int irq) { ia64_vector vec = irq_to_vector(irq); + struct iosapic_rte_info *rte; move_irq(irq); - iosapic_eoi(iosapic_intr_info[vec].addr, vec); + list_for_each_entry(rte, &iosapic_intr_info[vec].rtes, rte_list) + iosapic_eoi(rte->addr, vec); } #define iosapic_shutdown_level_irq mask_irq @@ -422,6 +468,34 @@ iosapic_version (char __iomem *addr) return iosapic_read(addr, IOSAPIC_VERSION); } +static int iosapic_find_sharable_vector (unsigned long trigger, unsigned long pol) +{ + int i, vector = -1, min_count = -1; + struct iosapic_intr_info *info; + + /* + * shared vectors for edge-triggered interrupts are not + * supported yet + */ + if (trigger == IOSAPIC_EDGE) + return -1; + + for (i = IA64_FIRST_DEVICE_VECTOR; i <= IA64_LAST_DEVICE_VECTOR; i++) { + info = &iosapic_intr_info[i]; + if (info->trigger == trigger && info->polarity == pol && + (info->dmode == IOSAPIC_FIXED || info->dmode == IOSAPIC_LOWEST_PRIORITY)) { + if (min_count == -1 || info->count < min_count) { + vector = i; + min_count = info->count; + } + } + } + if (vector < 0) + panic("%s: out of interrupt vectors!\n", __FUNCTION__); + + return vector; +} + /* * if the given vector is already owned by other, * assign a new vector for the other and make the vector available @@ -431,19 +505,63 @@ iosapic_reassign_vector (int vector) { int new_vector; - if (iosapic_intr_info[vector].rte_index >= 0 || iosapic_intr_info[vector].addr - || iosapic_intr_info[vector].gsi_base || iosapic_intr_info[vector].dmode - || iosapic_intr_info[vector].polarity || iosapic_intr_info[vector].trigger) - { + if (!list_empty(&iosapic_intr_info[vector].rtes)) { new_vector = assign_irq_vector(AUTO_ASSIGN); printk(KERN_INFO "Reassigning vector %d to %d\n", vector, new_vector); memcpy(&iosapic_intr_info[new_vector], &iosapic_intr_info[vector], sizeof(struct iosapic_intr_info)); + INIT_LIST_HEAD(&iosapic_intr_info[new_vector].rtes); + list_move(iosapic_intr_info[vector].rtes.next, &iosapic_intr_info[new_vector].rtes); memset(&iosapic_intr_info[vector], 0, sizeof(struct iosapic_intr_info)); - iosapic_intr_info[vector].rte_index = -1; + iosapic_intr_info[vector].low32 = IOSAPIC_MASK; + INIT_LIST_HEAD(&iosapic_intr_info[vector].rtes); } } +static struct iosapic_rte_info *iosapic_alloc_rte (void) +{ + int i; + struct iosapic_rte_info *rte; + int preallocated = 0; + + if (!iosapic_kmalloc_ok && list_empty(&free_rte_list)) { + rte = alloc_bootmem(sizeof(struct iosapic_rte_info) * NR_PREALLOCATE_RTE_ENTRIES); + if (!rte) + return NULL; + for (i = 0; i < NR_PREALLOCATE_RTE_ENTRIES; i++, rte++) + list_add(&rte->rte_list, &free_rte_list); + } + + if (!list_empty(&free_rte_list)) { + rte = list_entry(free_rte_list.next, struct iosapic_rte_info, rte_list); + list_del(&rte->rte_list); + preallocated++; + } else { + rte = kmalloc(sizeof(struct iosapic_rte_info), GFP_ATOMIC); + if (!rte) + return NULL; + } + + memset(rte, 0, sizeof(struct iosapic_rte_info)); + if (preallocated) + rte->flags |= RTE_PREALLOCATED; + + return rte; +} + +static void iosapic_free_rte (struct iosapic_rte_info *rte) +{ + if (rte->flags & RTE_PREALLOCATED) + list_add_tail(&rte->rte_list, &free_rte_list); + else + kfree(rte); +} + +static inline int vector_is_shared (int vector) +{ + return (iosapic_intr_info[vector].count > 1); +} + static void register_intr (unsigned int gsi, int vector, unsigned char delivery, unsigned long polarity, unsigned long trigger) @@ -454,6 +572,7 @@ register_intr (unsigned int gsi, int vector, unsigned char delivery, int index; unsigned long gsi_base; void __iomem *iosapic_address; + struct iosapic_rte_info *rte; index = find_iosapic(gsi); if (index < 0) { @@ -464,14 +583,33 @@ register_intr (unsigned int gsi, int vector, unsigned char delivery, iosapic_address = iosapic_lists[index].addr; gsi_base = iosapic_lists[index].gsi_base; - rte_index = gsi - gsi_base; - iosapic_intr_info[vector].rte_index = rte_index; + rte = gsi_vector_to_rte(gsi, vector); + if (!rte) { + rte = iosapic_alloc_rte(); + if (!rte) { + printk(KERN_WARNING "%s: cannot allocate memory\n", __FUNCTION__); + return; + } + + rte_index = gsi - gsi_base; + rte->rte_index = rte_index; + rte->addr = iosapic_address; + rte->gsi_base = gsi_base; + rte->refcnt++; + list_add_tail(&rte->rte_list, &iosapic_intr_info[vector].rtes); + iosapic_intr_info[vector].count++; + } + else if (vector_is_shared(vector)) { + struct iosapic_intr_info *info = &iosapic_intr_info[vector]; + if (info->trigger != trigger || info->polarity != polarity) { + printk (KERN_WARNING "%s: cannot override the interrupt\n", __FUNCTION__); + return; + } + } + iosapic_intr_info[vector].polarity = polarity; iosapic_intr_info[vector].dmode = delivery; - iosapic_intr_info[vector].addr = iosapic_address; - iosapic_intr_info[vector].gsi_base = gsi_base; iosapic_intr_info[vector].trigger = trigger; - iosapic_intr_info[vector].refcnt++; if (trigger == IOSAPIC_EDGE) irq_type = &irq_type_iosapic_edge; @@ -493,6 +631,13 @@ get_target_cpu (unsigned int gsi, int vector) #ifdef CONFIG_SMP static int cpu = -1; + /* + * In case of vector shared by multiple RTEs, all RTEs that + * share the vector need to use the same destination CPU. + */ + if (!list_empty(&iosapic_intr_info[vector].rtes)) + return iosapic_intr_info[vector].dest; + /* * If the platform supports redirection via XTP, let it * distribute interrupts. @@ -565,10 +710,12 @@ int iosapic_register_intr (unsigned int gsi, unsigned long polarity, unsigned long trigger) { - int vector; + int vector, mask = 1; unsigned int dest; unsigned long flags; - + struct iosapic_rte_info *rte; + u32 low32; +again: /* * If this GSI has already been registered (i.e., it's a * shared interrupt, or we lost a race to register it), @@ -578,19 +725,45 @@ iosapic_register_intr (unsigned int gsi, { vector = gsi_to_vector(gsi); if (vector > 0) { - iosapic_intr_info[vector].refcnt++; + rte = gsi_vector_to_rte(gsi, vector); + rte->refcnt++; spin_unlock_irqrestore(&iosapic_lock, flags); return vector; } + } + spin_unlock_irqrestore(&iosapic_lock, flags); + + /* If vector is running out, we try to find a sharable vector */ + vector = assign_irq_vector_nopanic(AUTO_ASSIGN); + if (vector < 0) + vector = iosapic_find_sharable_vector(trigger, polarity); + + spin_lock_irqsave(&irq_descp(vector)->lock, flags); + spin_lock(&iosapic_lock); + { + if (gsi_to_vector(gsi) > 0) { + if (list_empty(&iosapic_intr_info[vector].rtes)) + free_irq_vector(vector); + spin_unlock(&iosapic_lock); + spin_unlock_irqrestore(&irq_descp(vector)->lock, flags); + goto again; + } - vector = assign_irq_vector(AUTO_ASSIGN); dest = get_target_cpu(gsi, vector); register_intr(gsi, vector, IOSAPIC_LOWEST_PRIORITY, - polarity, trigger); + polarity, trigger); - set_rte(vector, dest, 1); + /* + * If the vector is shared and already unmasked for + * other interrupt sources, don't mask it. + */ + low32 = iosapic_intr_info[vector].low32; + if (vector_is_shared(vector) && !(low32 & IOSAPIC_MASK)) + mask = 0; + set_rte(gsi, vector, dest, mask); } - spin_unlock_irqrestore(&iosapic_lock, flags); + spin_unlock_irq(&iosapic_lock); + spin_unlock_irqrestore(&irq_descp(vector)->lock, flags); printk(KERN_INFO "GSI %u (%s, %s) -> CPU %d (0x%04x) vector %d\n", gsi, (trigger == IOSAPIC_EDGE ? "edge" : "level"), @@ -607,8 +780,10 @@ iosapic_unregister_intr (unsigned int gsi) unsigned long flags; int irq, vector; irq_desc_t *idesc; - int rte_index; + u32 low32; unsigned long trigger, polarity; + unsigned int dest; + struct iosapic_rte_info *rte; /* * If the irq associated with the gsi is not found, @@ -627,54 +802,56 @@ iosapic_unregister_intr (unsigned int gsi) spin_lock_irqsave(&idesc->lock, flags); spin_lock(&iosapic_lock); { - rte_index = iosapic_intr_info[vector].rte_index; - if (rte_index < 0) { - spin_unlock(&iosapic_lock); - spin_unlock_irqrestore(&idesc->lock, flags); + if ((rte = gsi_vector_to_rte(gsi, vector)) == NULL) { printk(KERN_ERR "iosapic_unregister_intr(%u) unbalanced\n", gsi); WARN_ON(1); - return; + goto out; } - if (--iosapic_intr_info[vector].refcnt > 0) { - spin_unlock(&iosapic_lock); - spin_unlock_irqrestore(&idesc->lock, flags); - return; - } + if (--rte->refcnt > 0) + goto out; - /* - * If interrupt handlers still exist on the irq - * associated with the gsi, don't unregister the - * interrupt. - */ - if (idesc->action) { - iosapic_intr_info[vector].refcnt++; - spin_unlock(&iosapic_lock); - spin_unlock_irqrestore(&idesc->lock, flags); - printk(KERN_WARNING "Cannot unregister GSI. IRQ %u is still in use.\n", irq); - return; - } + /* Mask the interrupt */ + low32 = iosapic_intr_info[vector].low32 | IOSAPIC_MASK; + iosapic_write(rte->addr, IOSAPIC_RTE_LOW(rte->rte_index), low32); - /* Clear the interrupt controller descriptor. */ - idesc->handler = &no_irq_type; + /* Remove the rte entry from the list */ + list_del(&rte->rte_list); + iosapic_intr_info[vector].count--; + iosapic_free_rte(rte); - trigger = iosapic_intr_info[vector].trigger; + trigger = iosapic_intr_info[vector].trigger; polarity = iosapic_intr_info[vector].polarity; + dest = iosapic_intr_info[vector].dest; + printk(KERN_INFO "GSI %u (%s, %s) -> CPU %d (0x%04x) vector %d unregistered\n", + gsi, (trigger == IOSAPIC_EDGE ? "edge" : "level"), + (polarity == IOSAPIC_POL_HIGH ? "high" : "low"), + cpu_logical_id(dest), dest, vector); + + if (list_empty(&iosapic_intr_info[vector].rtes)) { + /* Sanity check */ + BUG_ON(iosapic_intr_info[vector].count); + + /* Clear the interrupt controller descriptor */ + idesc->handler = &no_irq_type; + + /* Clear the interrupt information */ + memset(&iosapic_intr_info[vector], 0, sizeof(struct iosapic_intr_info)); + iosapic_intr_info[vector].low32 |= IOSAPIC_MASK; + INIT_LIST_HEAD(&iosapic_intr_info[vector].rtes); + + if (idesc->action) { + printk(KERN_ERR "interrupt handlers still exist on IRQ %u\n", irq); + WARN_ON(1); + } - /* Clear the interrupt information. */ - memset(&iosapic_intr_info[vector], 0, sizeof(struct iosapic_intr_info)); - iosapic_intr_info[vector].rte_index = -1; /* mark as unused */ + /* Free the interrupt vector */ + free_irq_vector(vector); + } } + out: spin_unlock(&iosapic_lock); spin_unlock_irqrestore(&idesc->lock, flags); - - /* Free the interrupt vector */ - free_irq_vector(vector); - - printk(KERN_INFO "GSI %u (%s, %s) -> vector %d unregisterd.\n", - gsi, (trigger == IOSAPIC_EDGE ? "edge" : "level"), - (polarity == IOSAPIC_POL_HIGH ? "high" : "low"), - vector); } #endif /* CONFIG_ACPI_DEALLOCATE_IRQ */ @@ -724,7 +901,7 @@ iosapic_register_platform_intr (u32 int_type, unsigned int gsi, (polarity == IOSAPIC_POL_HIGH ? "high" : "low"), cpu_logical_id(dest), dest, vector); - set_rte(vector, dest, mask); + set_rte(gsi, vector, dest, mask); return vector; } @@ -750,7 +927,7 @@ iosapic_override_isa_irq (unsigned int isa_irq, unsigned int gsi, polarity == IOSAPIC_POL_HIGH ? "high" : "low", cpu_logical_id(dest), dest, vector); - set_rte(vector, dest, 1); + set_rte(gsi, vector, dest, 1); } void __init @@ -758,8 +935,10 @@ iosapic_system_init (int system_pcat_compat) { int vector; - for (vector = 0; vector < IA64_NUM_VECTORS; ++vector) - iosapic_intr_info[vector].rte_index = -1; /* mark as unused */ + for (vector = 0; vector < IA64_NUM_VECTORS; ++vector) { + iosapic_intr_info[vector].low32 = IOSAPIC_MASK; + INIT_LIST_HEAD(&iosapic_intr_info[vector].rtes); /* mark as unused */ + } pcat_compat = system_pcat_compat; if (pcat_compat) { @@ -825,3 +1004,10 @@ map_iosapic_to_node(unsigned int gsi_base, int node) return; } #endif + +static int __init iosapic_enable_kmalloc (void) +{ + iosapic_kmalloc_ok = 1; + return 0; +} +core_initcall (iosapic_enable_kmalloc); diff --git a/arch/ia64/kernel/irq_ia64.c b/arch/ia64/kernel/irq_ia64.c index 5ba06ebe355..4fe60c7a2e9 100644 --- a/arch/ia64/kernel/irq_ia64.c +++ b/arch/ia64/kernel/irq_ia64.c @@ -63,20 +63,30 @@ EXPORT_SYMBOL(isa_irq_to_vector_map); static unsigned long ia64_vector_mask[BITS_TO_LONGS(IA64_NUM_DEVICE_VECTORS)]; int -assign_irq_vector (int irq) +assign_irq_vector_nopanic (int irq) { int pos, vector; again: pos = find_first_zero_bit(ia64_vector_mask, IA64_NUM_DEVICE_VECTORS); vector = IA64_FIRST_DEVICE_VECTOR + pos; if (vector > IA64_LAST_DEVICE_VECTOR) - /* XXX could look for sharable vectors instead of panic'ing... */ - panic("assign_irq_vector: out of interrupt vectors!"); + return -1; if (test_and_set_bit(pos, ia64_vector_mask)) goto again; return vector; } +int +assign_irq_vector (int irq) +{ + int vector = assign_irq_vector_nopanic(irq); + + if (vector < 0) + panic("assign_irq_vector: out of interrupt vectors!"); + + return vector; +} + void free_irq_vector (int vector) { diff --git a/include/asm-ia64/hw_irq.h b/include/asm-ia64/hw_irq.h index 041ab8c51a6..cd4e06b74ab 100644 --- a/include/asm-ia64/hw_irq.h +++ b/include/asm-ia64/hw_irq.h @@ -81,6 +81,7 @@ extern __u8 isa_irq_to_vector_map[16]; extern struct hw_interrupt_type irq_type_ia64_lsapic; /* CPU-internal interrupt controller */ +extern int assign_irq_vector_nopanic (int irq); /* allocate a free vector without panic */ extern int assign_irq_vector (int irq); /* allocate a free vector */ extern void free_irq_vector (int vector); extern void ia64_send_ipi (int cpu, int vector, int delivery_mode, int redirect); -- cgit v1.2.3-70-g09d2 From e1ed81ab7a34fc0f92f2e200825bdb6d86d6c8ef Mon Sep 17 00:00:00 2001 From: Tony Luck Date: Mon, 25 Apr 2005 13:27:12 -0700 Subject: [IA64] print "siblings" before {physical,core,thread} id Rohit and Suresh changed their mind about the order to print things in /proc/cpuinfo, but didn't include the change in the version of the patch they sent to me. Signed-off-by: Tony Luck --- arch/ia64/kernel/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c index 88043841fb8..b7e6b4cb374 100644 --- a/arch/ia64/kernel/setup.c +++ b/arch/ia64/kernel/setup.c @@ -512,13 +512,13 @@ show_cpuinfo (struct seq_file *m, void *v) c->itc_freq / 1000000, c->itc_freq % 1000000, lpj*HZ/500000, (lpj*HZ/5000) % 100); #ifdef CONFIG_SMP + seq_printf(m, "siblings : %u\n", c->num_log); if (c->threads_per_core > 1 || c->cores_per_socket > 1) seq_printf(m, "physical id: %u\n" "core id : %u\n" "thread id : %u\n", c->socket_id, c->core_id, c->thread_id); - seq_printf(m, "siblings : %u\n", c->num_log); #endif seq_printf(m,"\n"); -- cgit v1.2.3-70-g09d2 From b9e41d7fb62ae26adee84c18048037214ce5d866 Mon Sep 17 00:00:00 2001 From: Kenji Kaneshige Date: Mon, 25 Apr 2005 13:27:48 -0700 Subject: [IA64] iosapic.c: typo ... s/spin_unlock_irq/spin_unlock/ vector sharing patch had a typo ... mismatched spin_lock() with a spin_unlock_irq(). Fix from Kenji Kaneshige. Signed-off-by: Tony Luck --- arch/ia64/kernel/iosapic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c index 11a221cc8dc..88b014381df 100644 --- a/arch/ia64/kernel/iosapic.c +++ b/arch/ia64/kernel/iosapic.c @@ -762,7 +762,7 @@ again: mask = 0; set_rte(gsi, vector, dest, mask); } - spin_unlock_irq(&iosapic_lock); + spin_unlock(&iosapic_lock); spin_unlock_irqrestore(&irq_descp(vector)->lock, flags); printk(KERN_INFO "GSI %u (%s, %s) -> CPU %d (0x%04x) vector %d\n", -- cgit v1.2.3-70-g09d2 From 605036cf8443b9172b24954dc1fd218e2049f91b Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 25 Apr 2005 13:31:04 -0700 Subject: From: jbarnes@sgi.com [IA64] fix ia64 Kconfig to allow CONFIG_PM on sn2 This probably should have been fixed when I fixed up the generic build for discontig+numa machines, but oh well. CONFIG_PM is allowable for generic builds but not for sn2 builds, which doesn't make much sense, and in fact breaks the build if recent ACPI bits are added to the tree. It looks like the only arch that needs to prevent CONFIG_PM stuff is the ski simulator (though those options could probably use some cleanup as well), so remove the big conditional and replace it with a simple test for IA64_HP_SIM instead. Signed-off-by: Jesse Barnes Signed-off-by: Tony Luck --- arch/ia64/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 33fcb205fcb..468dbe8a6b9 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -329,7 +329,7 @@ menu "Power management and ACPI" config PM bool "Power Management support" - depends on IA64_GENERIC || IA64_DIG || IA64_HP_ZX1 || IA64_HP_ZX1_SWIOTLB + depends on !IA64_HP_SIM default y help "Power Management" means that parts of your computer are shut -- cgit v1.2.3-70-g09d2 From 41130d37a449dbff3593c8585a102d5e9173eea7 Mon Sep 17 00:00:00 2001 From: Jeff Lackey Date: Mon, 25 Apr 2005 23:38:55 +0100 Subject: [PATCH] ARM: 2650/1: PXA27x sleep - workaround Errata 39 & 50 (Patch 2667) Patch from Jeff Lackey This patch updates arch/arm/mach-pxa/sleep.S to support the PXA270 CPU. It works around Errata 39 & 50 from the Intel(R) PXA27x Processor Family Specification Update. Signed-off-by: Jeff Lackey Signed-off-by: Russell King --- arch/arm/mach-pxa/sleep.S | 56 ++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 50 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-pxa/sleep.S b/arch/arm/mach-pxa/sleep.S index 16cad2c2497..5786ccad938 100644 --- a/arch/arm/mach-pxa/sleep.S +++ b/arch/arm/mach-pxa/sleep.S @@ -18,6 +18,11 @@ #include +#ifdef CONFIG_PXA27x // workaround for Errata 50 +#define MDREFR_KDIV 0x200a4000 // all banks +#define CCCR_SLEEP 0x00000107 // L=7 2N=2 A=0 PPDIS=0 CPDIS=0 +#endif + .text /* @@ -28,7 +33,9 @@ ENTRY(pxa_cpu_suspend) +#ifndef CONFIG_IWMMXT mra r2, r3, acc0 +#endif stmfd sp!, {r2 - r12, lr} @ save registers on stack @ get coprocessor registers @@ -61,14 +68,23 @@ ENTRY(pxa_cpu_suspend) @ prepare value for sleep mode mov r1, #3 @ sleep mode - @ prepare to put SDRAM into self-refresh manually + @ prepare pointer to physical address 0 (virtual mapping in generic.c) + mov r2, #UNCACHED_PHYS_0 + + @ prepare SDRAM refresh settings ldr r4, =MDREFR ldr r5, [r4] + + @ enable SDRAM self-refresh mode orr r5, r5, #MDREFR_SLFRSH - @ prepare pointer to physical address 0 (virtual mapping in generic.c) - mov r2, #UNCACHED_PHYS_0 +#ifdef CONFIG_PXA27x + @ set SDCLKx divide-by-2 bits (this is part of a workaround for Errata 50) + ldr r6, =MDREFR_KDIV + orr r5, r5, r6 +#endif +#ifdef CONFIG_PXA25x @ Intel PXA255 Specification Update notes problems @ about suspending with PXBus operating above 133MHz @ (see Errata 31, GPIO output signals, ... unpredictable in sleep @@ -100,6 +116,18 @@ ENTRY(pxa_cpu_suspend) mov r0, #0 mcr p14, 0, r0, c6, c0, 0 orr r0, r0, #2 @ initiate change bit +#endif +#ifdef CONFIG_PXA27x + @ Intel PXA270 Specification Update notes problems sleeping + @ with core operating above 91 MHz + @ (see Errata 50, ...processor does not exit from sleep...) + + ldr r6, =CCCR + ldr r8, [r6] @ keep original value for resume + + ldr r7, =CCCR_SLEEP @ prepare CCCR sleep value + mov r0, #0x2 @ prepare value for CLKCFG +#endif @ align execution to a cache line b 1f @@ -111,6 +139,7 @@ ENTRY(pxa_cpu_suspend) @ All needed values are now in registers. @ These last instructions should be in cache +#if defined(CONFIG_PXA25x) || defined(CONFIG_PXA27x) @ initiate the frequency change... str r7, [r6] mcr p14, 0, r0, c6, c0, 0 @@ -118,14 +147,27 @@ ENTRY(pxa_cpu_suspend) @ restore the original cpu speed value for resume str r8, [r6] - @ put SDRAM into self-refresh - str r5, [r4] + @ need 6 13-MHz cycles before changing PWRMODE + @ just set frequency to 91-MHz... 6*91/13 = 42 + + mov r0, #42 +10: subs r0, r0, #1 + bne 10b +#endif + + @ Do not reorder... + @ Intel PXA270 Specification Update notes problems performing + @ external accesses after SDRAM is put in self-refresh mode + @ (see Errata 39 ...hangs when entering self-refresh mode) @ force address lines low by reading at physical address 0 ldr r3, [r2] + @ put SDRAM into self-refresh + str r5, [r4] + @ enter sleep mode - mcr p14, 0, r1, c7, c0, 0 + mcr p14, 0, r1, c7, c0, 0 @ PWRMODE 20: b 20b @ loop waiting for sleep @@ -188,7 +230,9 @@ resume_after_mmu: bl cpu_xscale_proc_init #endif ldmfd sp!, {r2, r3} +#ifndef CONFIG_IWMMXT mar acc0, r2, r3 +#endif ldmfd sp!, {r4 - r12, pc} @ return to caller -- cgit v1.2.3-70-g09d2 From 7933523dc75823342dbd70a27760a82a5302baf4 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:17:42 +0100 Subject: [PATCH] ARM: remove some entry initialisation asm code Convert the trivial vector entry initialisation code to C code. Signed-off-by: Russell King --- arch/arm/kernel/entry-armv.S | 47 +++++++++++++++++--------------------------- arch/arm/kernel/traps.c | 11 +++++++++-- 2 files changed, 27 insertions(+), 31 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index bb27c317d94..37723bfb96a 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -522,8 +522,9 @@ ENTRY(__switch_to) /* * Vector stubs. * - * This code is copied to 0x200 or 0xffff0200 so we can use branches in the - * vectors, rather than ldr's. + * This code is copied to 0xffff0200 so we can use branches in the + * vectors, rather than ldr's. Note that this code must not + * exceed 0x300 bytes. * * Common stub entry macro: * Enter in IRQ mode, spsr = SVC/USR CPSR, lr = SVC/USR PC @@ -552,6 +553,7 @@ vector_\name: movs pc, lr @ Changes mode and branches .endm + .globl __stubs_start __stubs_start: /* * Interrupt dispatcher @@ -686,37 +688,24 @@ vector_addrexcptn: .LCsabt: .word __temp_abt + .globl __stubs_end __stubs_end: - .equ __real_stubs_start, .LCvectors + 0x200 + .equ stubs_offset, __vectors_start + 0x200 - __stubs_start -.LCvectors: + .globl __vectors_start +__vectors_start: swi SYS_ERROR0 - b __real_stubs_start + (vector_und - __stubs_start) - ldr pc, __real_stubs_start + (.LCvswi - __stubs_start) - b __real_stubs_start + (vector_pabt - __stubs_start) - b __real_stubs_start + (vector_dabt - __stubs_start) - b __real_stubs_start + (vector_addrexcptn - __stubs_start) - b __real_stubs_start + (vector_irq - __stubs_start) - b __real_stubs_start + (vector_fiq - __stubs_start) - -ENTRY(__trap_init) - stmfd sp!, {r4 - r6, lr} - - mov r0, #0xff000000 - orr r0, r0, #0x00ff0000 @ high vectors position - adr r1, .LCvectors @ set up the vectors - ldmia r1, {r1, r2, r3, r4, r5, r6, ip, lr} - stmia r0, {r1, r2, r3, r4, r5, r6, ip, lr} - - add r2, r0, #0x200 - adr r0, __stubs_start @ copy stubs to 0x200 - adr r1, __stubs_end -1: ldr r3, [r0], #4 - str r3, [r2], #4 - cmp r0, r1 - blt 1b - LOADREGS(fd, sp!, {r4 - r6, pc}) + b vector_und + stubs_offset + ldr pc, .LCvswi + stubs_offset + b vector_pabt + stubs_offset + b vector_dabt + stubs_offset + b vector_addrexcptn + stubs_offset + b vector_irq + stubs_offset + b vector_fiq + stubs_offset + + .globl __vectors_end +__vectors_end: .data diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 6e31718f600..0078aeb8573 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -578,9 +578,16 @@ EXPORT_SYMBOL(abort); void __init trap_init(void) { - extern void __trap_init(void); + extern char __stubs_start[], __stubs_end[]; + extern char __vectors_start[], __vectors_end[]; - __trap_init(); + /* + * Copy the vectors and stubs (in entry-armv.S) into the + * vector page, mapped at 0xffff0000, and ensure these are + * visible to the instruction stream. + */ + memcpy((void *)0xffff0000, __vectors_start, __vectors_end - __vectors_start); + memcpy((void *)0xffff0200, __stubs_start, __stubs_end - __stubs_start); flush_icache_range(0xffff0000, 0xffff0000 + PAGE_SIZE); modify_domain(DOMAIN_USER, DOMAIN_CLIENT); } -- cgit v1.2.3-70-g09d2 From 1ec42c0c97186fadc48810ccaf2dc573cd957ea1 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:18:26 +0100 Subject: [PATCH] ARM: Remove argument for disable_irq/enable_irq Since we do not require a register for these operations, we can remove this unnecessary argument. Signed-off-by: Russell King --- arch/arm/kernel/entry-armv.S | 12 ++++++------ arch/arm/kernel/entry-common.S | 8 ++++---- arch/arm/kernel/entry-header.S | 16 ++++++---------- 3 files changed, 16 insertions(+), 20 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 37723bfb96a..fc142890547 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -118,7 +118,7 @@ __dabt_svc: @ @ IRQs off again before pulling preserved data off the stack @ - disable_irq r0 + disable_irq @ @ restore SPSR and restart the instruction @@ -198,7 +198,7 @@ __und_svc: @ @ IRQs off again before pulling preserved data off the stack @ -1: disable_irq r0 +1: disable_irq @ @ restore SPSR and restart the instruction @@ -232,7 +232,7 @@ __pabt_svc: @ @ IRQs off again before pulling preserved data off the stack @ - disable_irq r0 + disable_irq @ @ restore SPSR and restart the instruction @@ -316,7 +316,7 @@ __dabt_usr: @ @ IRQs on, then call the main handler @ - enable_irq r2 + enable_irq mov r2, sp adr lr, ret_from_exception b do_DataAbort @@ -418,7 +418,7 @@ call_fpe: movcss r7, r5, lsr #(TIF_USING_IWMMXT + 1) bcs iwmmxt_task_enable #endif - enable_irq r7 + enable_irq add pc, pc, r8, lsr #6 mov r0, r0 @@ -472,7 +472,7 @@ fpundefinstr: __pabt_usr: usr_entry abt - enable_irq r0 @ Enable interrupts + enable_irq @ Enable interrupts mov r0, r2 @ address (pc) mov r1, sp @ regs bl do_PrefetchAbort @ call abort handler diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 53a7e0dea44..0cd9f7bcaad 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -30,7 +30,7 @@ * stack. */ ret_fast_syscall: - disable_irq r1 @ disable interrupts + disable_irq @ disable interrupts ldr r1, [tsk, #TI_FLAGS] tst r1, #_TIF_WORK_MASK bne fast_work_pending @@ -49,7 +49,7 @@ work_pending: mov r0, sp @ 'regs' mov r2, why @ 'syscall' bl do_notify_resume - disable_irq r1 @ disable interrupts + disable_irq @ disable interrupts b no_work_pending work_resched: @@ -59,7 +59,7 @@ work_resched: */ ENTRY(ret_to_user) ret_slow_syscall: - disable_irq r1 @ disable interrupts + disable_irq @ disable interrupts ldr r1, [tsk, #TI_FLAGS] tst r1, #_TIF_WORK_MASK bne work_pending @@ -126,7 +126,7 @@ ENTRY(vector_swi) ldr ip, [ip] mcr p15, 0, ip, c1, c0 @ update control register #endif - enable_irq ip + enable_irq str r4, [sp, #-S_OFF]! @ push fifth arg diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 4039d8c120b..109f4f10d31 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -63,25 +63,21 @@ #define S_R0 0 #define S_OFF 8 - .macro set_cpsr_c, reg, mode - msr cpsr_c, \mode - .endm - #if __LINUX_ARM_ARCH__ >= 6 - .macro disable_irq, temp + .macro disable_irq cpsid i .endm - .macro enable_irq, temp + .macro enable_irq cpsie i .endm #else - .macro disable_irq, temp - set_cpsr_c \temp, #PSR_I_BIT | MODE_SVC + .macro disable_irq + msr cpsr_c, #PSR_I_BIT | SVC_MODE .endm - .macro enable_irq, temp - set_cpsr_c \temp, #MODE_SVC + .macro enable_irq + msr cpsr_c, #SVC_MODE .endm #endif -- cgit v1.2.3-70-g09d2 From 925c8a1a8cb9d7a33a8e39516d7fb679030553fc Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:18:59 +0100 Subject: [PATCH] ARM: pt_regs offsets Generate pt_regs S_xx offsets from the structure itself instead of #defining them. Signed-off-by: Russell King --- arch/arm/kernel/asm-offsets.c | 20 ++++++++++++++++++++ arch/arm/kernel/entry-common.S | 7 ------- arch/arm/kernel/entry-header.S | 31 ++++++++++--------------------- 3 files changed, 30 insertions(+), 28 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/asm-offsets.c b/arch/arm/kernel/asm-offsets.c index 99d43259ff8..c1ff4d1f1bf 100644 --- a/arch/arm/kernel/asm-offsets.c +++ b/arch/arm/kernel/asm-offsets.c @@ -64,6 +64,26 @@ int main(void) DEFINE(TI_VFPSTATE, offsetof(struct thread_info, vfpstate)); DEFINE(TI_IWMMXT_STATE, (offsetof(struct thread_info, fpstate)+4)&~7); BLANK(); + DEFINE(S_R0, offsetof(struct pt_regs, ARM_r0)); + DEFINE(S_R1, offsetof(struct pt_regs, ARM_r1)); + DEFINE(S_R2, offsetof(struct pt_regs, ARM_r2)); + DEFINE(S_R3, offsetof(struct pt_regs, ARM_r3)); + DEFINE(S_R4, offsetof(struct pt_regs, ARM_r4)); + DEFINE(S_R5, offsetof(struct pt_regs, ARM_r5)); + DEFINE(S_R6, offsetof(struct pt_regs, ARM_r6)); + DEFINE(S_R7, offsetof(struct pt_regs, ARM_r7)); + DEFINE(S_R8, offsetof(struct pt_regs, ARM_r8)); + DEFINE(S_R9, offsetof(struct pt_regs, ARM_r9)); + DEFINE(S_R10, offsetof(struct pt_regs, ARM_r10)); + DEFINE(S_FP, offsetof(struct pt_regs, ARM_fp)); + DEFINE(S_IP, offsetof(struct pt_regs, ARM_ip)); + DEFINE(S_SP, offsetof(struct pt_regs, ARM_sp)); + DEFINE(S_LR, offsetof(struct pt_regs, ARM_lr)); + DEFINE(S_PC, offsetof(struct pt_regs, ARM_pc)); + DEFINE(S_PSR, offsetof(struct pt_regs, ARM_cpsr)); + DEFINE(S_OLD_R0, offsetof(struct pt_regs, ARM_ORIG_r0)); + DEFINE(S_FRAME_SIZE, sizeof(struct pt_regs)); + BLANK(); #if __LINUX_ARM_ARCH__ >= 6 DEFINE(MM_CONTEXT_ID, offsetof(struct mm_struct, context.id)); BLANK(); diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 0cd9f7bcaad..55201db0f14 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -15,13 +15,6 @@ #include "entry-header.S" -/* - * We rely on the fact that R0 is at the bottom of the stack (due to - * slow/fast restore user regs). - */ -#if S_R0 != 0 -#error "Please fix" -#endif .align 5 /* diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 109f4f10d31..9d4d286d380 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -39,30 +39,19 @@ #define ARMSWI_OFFSET 0x000f0000 @ -@ Stack format (ensured by USER_* and SVC_*) +@ Most of the stack format comes from struct pt_regs, but with +@ the addition of 8 bytes for storing syscall args 5 and 6. @ -#define S_FRAME_SIZE 72 -#define S_OLD_R0 68 -#define S_PSR 64 - -#define S_PC 60 -#define S_LR 56 -#define S_SP 52 -#define S_IP 48 -#define S_FP 44 -#define S_R10 40 -#define S_R9 36 -#define S_R8 32 -#define S_R7 28 -#define S_R6 24 -#define S_R5 20 -#define S_R4 16 -#define S_R3 12 -#define S_R2 8 -#define S_R1 4 -#define S_R0 0 #define S_OFF 8 +/* + * The SWI code relies on the fact that R0 is at the bottom of the stack + * (due to slow/fast restore user regs). + */ +#if S_R0 != 0 +#error "Please fix" +#endif + #if __LINUX_ARM_ARCH__ >= 6 .macro disable_irq cpsid i -- cgit v1.2.3-70-g09d2 From e0f9f4a622cec66d09a5d9339f048877c665cec3 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:19:24 +0100 Subject: [PATCH] ARM: Use __NR_SYSCALL_BASE and __ARM_NR_BASE in asm code Don't define our own local constants, but use those already defined in asm/unistd.h instead. Signed-off-by: Russell King --- arch/arm/kernel/entry-common.S | 20 +++++++++++++++----- arch/arm/kernel/entry-header.S | 22 ---------------------- 2 files changed, 15 insertions(+), 27 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 55201db0f14..65c58b34db9 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -111,7 +111,17 @@ ENTRY(ret_from_fork) ENTRY(vector_swi) save_user_regs zero_fp - get_scno + + /* + * Get the system call number. + */ +#ifdef CONFIG_ARM_THUMB + tst r8, #PSR_T_BIT @ this is SPSR from save_user_regs + addne scno, r7, #__NR_SYSCALL_BASE @ put OS number in + ldreq scno, [lr, #-4] +#else + ldr scno, [lr, #-4] @ get SWI instruction +#endif arm710_bug_check scno, ip #ifdef CONFIG_ALIGNMENT_TRAP @@ -126,7 +136,7 @@ ENTRY(vector_swi) get_thread_info tsk ldr ip, [tsk, #TI_FLAGS] @ check for syscall tracing bic scno, scno, #0xff000000 @ mask off SWI op-code - eor scno, scno, #OS_NUMBER << 20 @ check OS number + eor scno, scno, #__NR_SYSCALL_BASE @ check OS number adr tbl, sys_call_table @ load syscall table pointer tst ip, #_TIF_SYSCALL_TRACE @ are we tracing syscalls? bne __sys_trace @@ -137,8 +147,8 @@ ENTRY(vector_swi) add r1, sp, #S_OFF 2: mov why, #0 @ no longer a real syscall - cmp scno, #ARMSWI_OFFSET - eor r0, scno, #OS_NUMBER << 20 @ put OS number back + cmp scno, #(__ARM_NR_BASE - __NR_SYSCALL_BASE) + eor r0, scno, #__NR_SYSCALL_BASE @ put OS number back bcs arm_syscall b sys_ni_syscall @ not private func @@ -183,7 +193,7 @@ ENTRY(sys_call_table) @ r5 = syscall table .type sys_syscall, #function sys_syscall: - eor scno, r0, #OS_NUMBER << 20 + eor scno, r0, #__NR_SYSCALL_BASE cmp scno, #__NR_syscall - __NR_SYSCALL_BASE cmpne scno, #NR_syscalls @ check range stmloia sp, {r5, r6} @ shuffle args diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 9d4d286d380..0eb7fc935e2 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -31,13 +31,6 @@ #define PT_TRACESYS 0x00000002 -@ OS version number used in SWIs -@ RISC OS is 0 -@ RISC iX is 8 -@ -#define OS_NUMBER 9 -#define ARMSWI_OFFSET 0x000f0000 - @ @ Most of the stack format comes from struct pt_regs, but with @ the addition of 8 bytes for storing syscall args 5 and 6. @@ -150,18 +143,3 @@ scno .req r7 @ syscall number tbl .req r8 @ syscall table pointer why .req r8 @ Linux syscall (!= 0) tsk .req r9 @ current thread_info - -/* - * Get the system call number. - */ - .macro get_scno -#ifdef CONFIG_ARM_THUMB - tst r8, #PSR_T_BIT @ this is SPSR from save_user_regs - addne scno, r7, #OS_NUMBER << 20 @ put OS number in - ldreq scno, [lr, #-4] - -#else - mask_pc lr, lr - ldr scno, [lr, #-4] @ get SWI instruction -#endif - .endm -- cgit v1.2.3-70-g09d2 From acaca3c91536491119e97d428c1da935f411cc9a Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:19:48 +0100 Subject: [PATCH] ARM: Remove SVC_MODE definition SVC_MODE reflects the MODE_SVC definition in asm/ptrace.h. Use the asm/ptrace.h definition instead, and remove SVC_MODE. Signed-off-by: Russell King --- arch/arm/kernel/entry-armv.S | 2 +- arch/arm/kernel/entry-header.S | 4 ---- 2 files changed, 1 insertion(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index fc142890547..41470567186 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -545,7 +545,7 @@ vector_\name: @ mrs r13, cpsr bic r13, r13, #MODE_MASK - orr r13, r13, #MODE_SVC + orr r13, r13, #SVC_MODE msr spsr_cxsf, r13 @ switch to SVC_32 mode and lr, lr, #15 diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 0eb7fc935e2..5d8a8451629 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -8,10 +8,6 @@ #include #include -#ifndef MODE_SVC -#define MODE_SVC 0x13 -#endif - .macro zero_fp #ifdef CONFIG_FRAME_POINTER mov fp, #0 -- cgit v1.2.3-70-g09d2 From cf88b417f9b1365c5aeeffa5c41208be68adfd76 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:20:12 +0100 Subject: [PATCH] ARM: remove PT_TRACESYS PT_TRACESYS is unused, remove it. Signed-off-by: Russell King --- arch/arm/kernel/entry-header.S | 2 -- 1 file changed, 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 5d8a8451629..b7d54a4f98f 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -25,8 +25,6 @@ #define BAD_IRQ 3 #define BAD_UNDEFINSTR 4 -#define PT_TRACESYS 0x00000002 - @ @ Most of the stack format comes from struct pt_regs, but with @ the addition of 8 bytes for storing syscall args 5 and 6. -- cgit v1.2.3-70-g09d2 From f4dc9a4cf26278f5b608d6e4bd16e7b27ddcf1a5 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:20:34 +0100 Subject: [PATCH] ARM: Remove single-use user save/restore macros Assembly macros are pointless if they're only used once. Move them inline. Signed-off-by: Russell King --- arch/arm/kernel/entry-common.S | 28 ++++++++++++++++++++--- arch/arm/kernel/entry-header.S | 51 ------------------------------------------ 2 files changed, 25 insertions(+), 54 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 65c58b34db9..07da010b67d 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -27,7 +27,15 @@ ret_fast_syscall: ldr r1, [tsk, #TI_FLAGS] tst r1, #_TIF_WORK_MASK bne fast_work_pending - fast_restore_user_regs + + @ fast_restore_user_regs + ldr r1, [sp, #S_OFF + S_PSR] @ get calling cpsr + ldr lr, [sp, #S_OFF + S_PC]! @ get pc + msr spsr_cxsf, r1 @ save in spsr_svc + ldmdb sp, {r1 - lr}^ @ get calling r1 - lr + mov r0, r0 + add sp, sp, #S_FRAME_SIZE - S_PC + movs pc, lr @ return & move spsr_svc into cpsr /* * Ok, we need to do extra processing, enter the slow path. @@ -57,7 +65,14 @@ ret_slow_syscall: tst r1, #_TIF_WORK_MASK bne work_pending no_work_pending: - slow_restore_user_regs + @ slow_restore_user_regs + ldr r1, [sp, #S_PSR] @ get calling cpsr + ldr lr, [sp, #S_PC]! @ get pc + msr spsr_cxsf, r1 @ save in spsr_svc + ldmdb sp, {r0 - lr}^ @ get calling r1 - lr + mov r0, r0 + add sp, sp, #S_FRAME_SIZE - S_PC + movs pc, lr @ return & move spsr_svc into cpsr /* * This is how we return from a fork. @@ -109,7 +124,14 @@ ENTRY(ret_from_fork) .align 5 ENTRY(vector_swi) - save_user_regs + sub sp, sp, #S_FRAME_SIZE + stmia sp, {r0 - r12} @ Calling r0 - r12 + add r8, sp, #S_PC + stmdb r8, {sp, lr}^ @ Calling sp, lr + mrs r8, spsr @ called from non-FIQ mode, so ok. + str lr, [sp, #S_PC] @ Save calling PC + str r8, [sp, #S_PSR] @ Save CPSR + str r0, [sp, #S_OLD_R0] @ Save OLD_R0 zero_fp /* diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index b7d54a4f98f..956af0bff3f 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -57,57 +57,6 @@ .endm #endif - .macro save_user_regs - sub sp, sp, #S_FRAME_SIZE - stmia sp, {r0 - r12} @ Calling r0 - r12 - add r8, sp, #S_PC - stmdb r8, {sp, lr}^ @ Calling sp, lr - mrs r8, spsr @ called from non-FIQ mode, so ok. - str lr, [sp, #S_PC] @ Save calling PC - str r8, [sp, #S_PSR] @ Save CPSR - str r0, [sp, #S_OLD_R0] @ Save OLD_R0 - .endm - - .macro restore_user_regs - ldr r1, [sp, #S_PSR] @ Get calling cpsr - disable_irq ip @ disable IRQs - ldr lr, [sp, #S_PC]! @ Get PC - msr spsr_cxsf, r1 @ save in spsr_svc - ldmdb sp, {r0 - lr}^ @ Get calling r0 - lr - mov r0, r0 - add sp, sp, #S_FRAME_SIZE - S_PC - movs pc, lr @ return & move spsr_svc into cpsr - .endm - -/* - * Must be called with IRQs already disabled. - */ - .macro fast_restore_user_regs - ldr r1, [sp, #S_OFF + S_PSR] @ get calling cpsr - ldr lr, [sp, #S_OFF + S_PC]! @ get pc - msr spsr_cxsf, r1 @ save in spsr_svc - ldmdb sp, {r1 - lr}^ @ get calling r1 - lr - mov r0, r0 - add sp, sp, #S_FRAME_SIZE - S_PC - movs pc, lr @ return & move spsr_svc into cpsr - .endm - -/* - * Must be called with IRQs already disabled. - */ - .macro slow_restore_user_regs - ldr r1, [sp, #S_PSR] @ get calling cpsr - ldr lr, [sp, #S_PC]! @ get pc - msr spsr_cxsf, r1 @ save in spsr_svc - ldmdb sp, {r0 - lr}^ @ get calling r1 - lr - mov r0, r0 - add sp, sp, #S_FRAME_SIZE - S_PC - movs pc, lr @ return & move spsr_svc into cpsr - .endm - - .macro mask_pc, rd, rm - .endm - .macro get_thread_info, rd mov \rd, sp, lsr #13 mov \rd, \rd, lsl #13 -- cgit v1.2.3-70-g09d2 From bce495d865829d7a1d8102a834d3e3af32616567 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 26 Apr 2005 15:21:02 +0100 Subject: [PATCH] ARM: make entry*.S includes more logical Move common includes to entry-header, and file specific includes to the relevant file. Signed-off-by: Russell King --- arch/arm/kernel/Makefile | 5 ----- arch/arm/kernel/entry-armv.S | 6 +++--- arch/arm/kernel/entry-common.S | 2 -- arch/arm/kernel/entry-header.S | 21 +++++++++------------ 4 files changed, 12 insertions(+), 22 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile index 07a56ff6149..4a2af55e134 100644 --- a/arch/arm/kernel/Makefile +++ b/arch/arm/kernel/Makefile @@ -31,8 +31,3 @@ head-y := head.o obj-$(CONFIG_DEBUG_LL) += debug.o extra-y := $(head-y) init_task.o vmlinux.lds - -# Spell out some dependencies that aren't automatically figured out -$(obj)/entry-armv.o: $(obj)/entry-header.S include/asm-arm/constants.h -$(obj)/entry-common.o: $(obj)/entry-header.S include/asm-arm/constants.h \ - $(obj)/calls.S diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 41470567186..2a5c3fe09a9 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -14,12 +14,12 @@ * it to save wrong values... Be aware! */ #include -#include -#include #include -#include #include +#include @ should be moved into entry-macro.S +#include @ should be moved into entry-macro.S +#include #include "entry-header.S" diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 07da010b67d..3f8d0e3aefa 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -9,8 +9,6 @@ */ #include -#include -#include #include #include "entry-header.S" diff --git a/arch/arm/kernel/entry-header.S b/arch/arm/kernel/entry-header.S index 956af0bff3f..a3d40a0e2b0 100644 --- a/arch/arm/kernel/entry-header.S +++ b/arch/arm/kernel/entry-header.S @@ -1,20 +1,11 @@ -#include /* for CONFIG_ARCH_xxxx */ +#include +#include #include #include #include #include -#include -#include -#include - - .macro zero_fp -#ifdef CONFIG_FRAME_POINTER - mov fp, #0 -#endif - .endm - - .text +#include @ Bad Abort numbers @ ----------------- @@ -39,6 +30,12 @@ #error "Please fix" #endif + .macro zero_fp +#ifdef CONFIG_FRAME_POINTER + mov fp, #0 +#endif + .endm + #if __LINUX_ARM_ARCH__ >= 6 .macro disable_irq cpsid i -- cgit v1.2.3-70-g09d2 From 66768eb26c471091dc4ea1ed59b4bab7edc3d7bd Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 26 Apr 2005 07:43:41 -0700 Subject: [PATCH] ppc-opc NULL noise removal Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc64/xmon/ppc-opc.c | 119 +++++++++++++++++++++++----------------------- 1 file changed, 60 insertions(+), 59 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/xmon/ppc-opc.c b/arch/ppc64/xmon/ppc-opc.c index 1e4e7e31997..5ee8fc32f82 100644 --- a/arch/ppc64/xmon/ppc-opc.c +++ b/arch/ppc64/xmon/ppc-opc.c @@ -20,6 +20,7 @@ Software Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ +#include #include "nonstdio.h" #include "ppc.h" @@ -110,12 +111,12 @@ const struct powerpc_operand powerpc_operands[] = /* The zero index is used to indicate the end of the list of operands. */ #define UNUSED 0 - { 0, 0, 0, 0, 0 }, + { 0, 0, NULL, NULL, 0 }, /* The BA field in an XL form instruction. */ #define BA UNUSED + 1 #define BA_MASK (0x1f << 16) - { 5, 16, 0, 0, PPC_OPERAND_CR }, + { 5, 16, NULL, NULL, PPC_OPERAND_CR }, /* The BA field in an XL form instruction when it must be the same as the BT field in the same instruction. */ @@ -125,7 +126,7 @@ const struct powerpc_operand powerpc_operands[] = /* The BB field in an XL form instruction. */ #define BB BAT + 1 #define BB_MASK (0x1f << 11) - { 5, 11, 0, 0, PPC_OPERAND_CR }, + { 5, 11, NULL, NULL, PPC_OPERAND_CR }, /* The BB field in an XL form instruction when it must be the same as the BA field in the same instruction. */ @@ -168,21 +169,21 @@ const struct powerpc_operand powerpc_operands[] = /* The BF field in an X or XL form instruction. */ #define BF BDPA + 1 - { 3, 23, 0, 0, PPC_OPERAND_CR }, + { 3, 23, NULL, NULL, PPC_OPERAND_CR }, /* An optional BF field. This is used for comparison instructions, in which an omitted BF field is taken as zero. */ #define OBF BF + 1 - { 3, 23, 0, 0, PPC_OPERAND_CR | PPC_OPERAND_OPTIONAL }, + { 3, 23, NULL, NULL, PPC_OPERAND_CR | PPC_OPERAND_OPTIONAL }, /* The BFA field in an X or XL form instruction. */ #define BFA OBF + 1 - { 3, 18, 0, 0, PPC_OPERAND_CR }, + { 3, 18, NULL, NULL, PPC_OPERAND_CR }, /* The BI field in a B form or XL form instruction. */ #define BI BFA + 1 #define BI_MASK (0x1f << 16) - { 5, 16, 0, 0, PPC_OPERAND_CR }, + { 5, 16, NULL, NULL, PPC_OPERAND_CR }, /* The BO field in a B form instruction. Certain values are illegal. */ @@ -197,36 +198,36 @@ const struct powerpc_operand powerpc_operands[] = /* The BT field in an X or XL form instruction. */ #define BT BOE + 1 - { 5, 21, 0, 0, PPC_OPERAND_CR }, + { 5, 21, NULL, NULL, PPC_OPERAND_CR }, /* The condition register number portion of the BI field in a B form or XL form instruction. This is used for the extended conditional branch mnemonics, which set the lower two bits of the BI field. This field is optional. */ #define CR BT + 1 - { 3, 18, 0, 0, PPC_OPERAND_CR | PPC_OPERAND_OPTIONAL }, + { 3, 18, NULL, NULL, PPC_OPERAND_CR | PPC_OPERAND_OPTIONAL }, /* The CRB field in an X form instruction. */ #define CRB CR + 1 - { 5, 6, 0, 0, 0 }, + { 5, 6, NULL, NULL, 0 }, /* The CRFD field in an X form instruction. */ #define CRFD CRB + 1 - { 3, 23, 0, 0, PPC_OPERAND_CR }, + { 3, 23, NULL, NULL, PPC_OPERAND_CR }, /* The CRFS field in an X form instruction. */ #define CRFS CRFD + 1 - { 3, 0, 0, 0, PPC_OPERAND_CR }, + { 3, 0, NULL, NULL, PPC_OPERAND_CR }, /* The CT field in an X form instruction. */ #define CT CRFS + 1 - { 5, 21, 0, 0, PPC_OPERAND_OPTIONAL }, + { 5, 21, NULL, NULL, PPC_OPERAND_OPTIONAL }, /* The D field in a D form instruction. This is a displacement off a register, and implies that the next operand is a register in parentheses. */ #define D CT + 1 - { 16, 0, 0, 0, PPC_OPERAND_PARENS | PPC_OPERAND_SIGNED }, + { 16, 0, NULL, NULL, PPC_OPERAND_PARENS | PPC_OPERAND_SIGNED }, /* The DE field in a DE form instruction. This is like D, but is 12 bits only. */ @@ -252,40 +253,40 @@ const struct powerpc_operand powerpc_operands[] = /* The E field in a wrteei instruction. */ #define E DS + 1 - { 1, 15, 0, 0, 0 }, + { 1, 15, NULL, NULL, 0 }, /* The FL1 field in a POWER SC form instruction. */ #define FL1 E + 1 - { 4, 12, 0, 0, 0 }, + { 4, 12, NULL, NULL, 0 }, /* The FL2 field in a POWER SC form instruction. */ #define FL2 FL1 + 1 - { 3, 2, 0, 0, 0 }, + { 3, 2, NULL, NULL, 0 }, /* The FLM field in an XFL form instruction. */ #define FLM FL2 + 1 - { 8, 17, 0, 0, 0 }, + { 8, 17, NULL, NULL, 0 }, /* The FRA field in an X or A form instruction. */ #define FRA FLM + 1 #define FRA_MASK (0x1f << 16) - { 5, 16, 0, 0, PPC_OPERAND_FPR }, + { 5, 16, NULL, NULL, PPC_OPERAND_FPR }, /* The FRB field in an X or A form instruction. */ #define FRB FRA + 1 #define FRB_MASK (0x1f << 11) - { 5, 11, 0, 0, PPC_OPERAND_FPR }, + { 5, 11, NULL, NULL, PPC_OPERAND_FPR }, /* The FRC field in an A form instruction. */ #define FRC FRB + 1 #define FRC_MASK (0x1f << 6) - { 5, 6, 0, 0, PPC_OPERAND_FPR }, + { 5, 6, NULL, NULL, PPC_OPERAND_FPR }, /* The FRS field in an X form instruction or the FRT field in a D, X or A form instruction. */ #define FRS FRC + 1 #define FRT FRS - { 5, 21, 0, 0, PPC_OPERAND_FPR }, + { 5, 21, NULL, NULL, PPC_OPERAND_FPR }, /* The FXM field in an XFX instruction. */ #define FXM FRS + 1 @@ -298,11 +299,11 @@ const struct powerpc_operand powerpc_operands[] = /* The L field in a D or X form instruction. */ #define L FXM4 + 1 - { 1, 21, 0, 0, PPC_OPERAND_OPTIONAL }, + { 1, 21, NULL, NULL, PPC_OPERAND_OPTIONAL }, /* The LEV field in a POWER SC form instruction. */ #define LEV L + 1 - { 7, 5, 0, 0, 0 }, + { 7, 5, NULL, NULL, 0 }, /* The LI field in an I form instruction. The lower two bits are forced to zero. */ @@ -316,24 +317,24 @@ const struct powerpc_operand powerpc_operands[] = /* The LS field in an X (sync) form instruction. */ #define LS LIA + 1 - { 2, 21, 0, 0, PPC_OPERAND_OPTIONAL }, + { 2, 21, NULL, NULL, PPC_OPERAND_OPTIONAL }, /* The MB field in an M form instruction. */ #define MB LS + 1 #define MB_MASK (0x1f << 6) - { 5, 6, 0, 0, 0 }, + { 5, 6, NULL, NULL, 0 }, /* The ME field in an M form instruction. */ #define ME MB + 1 #define ME_MASK (0x1f << 1) - { 5, 1, 0, 0, 0 }, + { 5, 1, NULL, NULL, 0 }, /* The MB and ME fields in an M form instruction expressed a single operand which is a bitmask indicating which bits to select. This is a two operand form using PPC_OPERAND_NEXT. See the description in opcode/ppc.h for what this means. */ #define MBE ME + 1 - { 5, 6, 0, 0, PPC_OPERAND_OPTIONAL | PPC_OPERAND_NEXT }, + { 5, 6, NULL, NULL, PPC_OPERAND_OPTIONAL | PPC_OPERAND_NEXT }, { 32, 0, insert_mbe, extract_mbe, 0 }, /* The MB or ME field in an MD or MDS form instruction. The high @@ -345,7 +346,7 @@ const struct powerpc_operand powerpc_operands[] = /* The MO field in an mbar instruction. */ #define MO MB6 + 1 - { 5, 21, 0, 0, 0 }, + { 5, 21, NULL, NULL, 0 }, /* The NB field in an X form instruction. The value 32 is stored as 0. */ @@ -361,34 +362,34 @@ const struct powerpc_operand powerpc_operands[] = /* The RA field in an D, DS, DQ, X, XO, M, or MDS form instruction. */ #define RA NSI + 1 #define RA_MASK (0x1f << 16) - { 5, 16, 0, 0, PPC_OPERAND_GPR }, + { 5, 16, NULL, NULL, PPC_OPERAND_GPR }, /* The RA field in the DQ form lq instruction, which has special value restrictions. */ #define RAQ RA + 1 - { 5, 16, insert_raq, 0, PPC_OPERAND_GPR }, + { 5, 16, insert_raq, NULL, PPC_OPERAND_GPR }, /* The RA field in a D or X form instruction which is an updating load, which means that the RA field may not be zero and may not equal the RT field. */ #define RAL RAQ + 1 - { 5, 16, insert_ral, 0, PPC_OPERAND_GPR }, + { 5, 16, insert_ral, NULL, PPC_OPERAND_GPR }, /* The RA field in an lmw instruction, which has special value restrictions. */ #define RAM RAL + 1 - { 5, 16, insert_ram, 0, PPC_OPERAND_GPR }, + { 5, 16, insert_ram, NULL, PPC_OPERAND_GPR }, /* The RA field in a D or X form instruction which is an updating store or an updating floating point load, which means that the RA field may not be zero. */ #define RAS RAM + 1 - { 5, 16, insert_ras, 0, PPC_OPERAND_GPR }, + { 5, 16, insert_ras, NULL, PPC_OPERAND_GPR }, /* The RB field in an X, XO, M, or MDS form instruction. */ #define RB RAS + 1 #define RB_MASK (0x1f << 11) - { 5, 11, 0, 0, PPC_OPERAND_GPR }, + { 5, 11, NULL, NULL, PPC_OPERAND_GPR }, /* The RB field in an X form instruction when it must be the same as the RS field in the instruction. This is used for extended @@ -402,22 +403,22 @@ const struct powerpc_operand powerpc_operands[] = #define RS RBS + 1 #define RT RS #define RT_MASK (0x1f << 21) - { 5, 21, 0, 0, PPC_OPERAND_GPR }, + { 5, 21, NULL, NULL, PPC_OPERAND_GPR }, /* The RS field of the DS form stq instruction, which has special value restrictions. */ #define RSQ RS + 1 - { 5, 21, insert_rsq, 0, PPC_OPERAND_GPR }, + { 5, 21, insert_rsq, NULL, PPC_OPERAND_GPR }, /* The RT field of the DQ form lq instruction, which has special value restrictions. */ #define RTQ RSQ + 1 - { 5, 21, insert_rtq, 0, PPC_OPERAND_GPR }, + { 5, 21, insert_rtq, NULL, PPC_OPERAND_GPR }, /* The SH field in an X or M form instruction. */ #define SH RTQ + 1 #define SH_MASK (0x1f << 11) - { 5, 11, 0, 0, 0 }, + { 5, 11, NULL, NULL, 0 }, /* The SH field in an MD form instruction. This is split. */ #define SH6 SH + 1 @@ -426,12 +427,12 @@ const struct powerpc_operand powerpc_operands[] = /* The SI field in a D form instruction. */ #define SI SH6 + 1 - { 16, 0, 0, 0, PPC_OPERAND_SIGNED }, + { 16, 0, NULL, NULL, PPC_OPERAND_SIGNED }, /* The SI field in a D form instruction when we accept a wide range of positive values. */ #define SISIGNOPT SI + 1 - { 16, 0, 0, 0, PPC_OPERAND_SIGNED | PPC_OPERAND_SIGNOPT }, + { 16, 0, NULL, NULL, PPC_OPERAND_SIGNED | PPC_OPERAND_SIGNOPT }, /* The SPR field in an XFX form instruction. This is flipped--the lower 5 bits are stored in the upper 5 and vice- versa. */ @@ -443,25 +444,25 @@ const struct powerpc_operand powerpc_operands[] = /* The BAT index number in an XFX form m[ft]ibat[lu] instruction. */ #define SPRBAT SPR + 1 #define SPRBAT_MASK (0x3 << 17) - { 2, 17, 0, 0, 0 }, + { 2, 17, NULL, NULL, 0 }, /* The SPRG register number in an XFX form m[ft]sprg instruction. */ #define SPRG SPRBAT + 1 #define SPRG_MASK (0x3 << 16) - { 2, 16, 0, 0, 0 }, + { 2, 16, NULL, NULL, 0 }, /* The SR field in an X form instruction. */ #define SR SPRG + 1 - { 4, 16, 0, 0, 0 }, + { 4, 16, NULL, NULL, 0 }, /* The STRM field in an X AltiVec form instruction. */ #define STRM SR + 1 #define STRM_MASK (0x3 << 21) - { 2, 21, 0, 0, 0 }, + { 2, 21, NULL, NULL, 0 }, /* The SV field in a POWER SC form instruction. */ #define SV STRM + 1 - { 14, 2, 0, 0, 0 }, + { 14, 2, NULL, NULL, 0 }, /* The TBR field in an XFX form instruction. This is like the SPR field, but it is optional. */ @@ -471,52 +472,52 @@ const struct powerpc_operand powerpc_operands[] = /* The TO field in a D or X form instruction. */ #define TO TBR + 1 #define TO_MASK (0x1f << 21) - { 5, 21, 0, 0, 0 }, + { 5, 21, NULL, NULL, 0 }, /* The U field in an X form instruction. */ #define U TO + 1 - { 4, 12, 0, 0, 0 }, + { 4, 12, NULL, NULL, 0 }, /* The UI field in a D form instruction. */ #define UI U + 1 - { 16, 0, 0, 0, 0 }, + { 16, 0, NULL, NULL, 0 }, /* The VA field in a VA, VX or VXR form instruction. */ #define VA UI + 1 #define VA_MASK (0x1f << 16) - { 5, 16, 0, 0, PPC_OPERAND_VR }, + { 5, 16, NULL, NULL, PPC_OPERAND_VR }, /* The VB field in a VA, VX or VXR form instruction. */ #define VB VA + 1 #define VB_MASK (0x1f << 11) - { 5, 11, 0, 0, PPC_OPERAND_VR }, + { 5, 11, NULL, NULL, PPC_OPERAND_VR }, /* The VC field in a VA form instruction. */ #define VC VB + 1 #define VC_MASK (0x1f << 6) - { 5, 6, 0, 0, PPC_OPERAND_VR }, + { 5, 6, NULL, NULL, PPC_OPERAND_VR }, /* The VD or VS field in a VA, VX, VXR or X form instruction. */ #define VD VC + 1 #define VS VD #define VD_MASK (0x1f << 21) - { 5, 21, 0, 0, PPC_OPERAND_VR }, + { 5, 21, NULL, NULL, PPC_OPERAND_VR }, /* The SIMM field in a VX form instruction. */ #define SIMM VD + 1 - { 5, 16, 0, 0, PPC_OPERAND_SIGNED}, + { 5, 16, NULL, NULL, PPC_OPERAND_SIGNED}, /* The UIMM field in a VX form instruction. */ #define UIMM SIMM + 1 - { 5, 16, 0, 0, 0 }, + { 5, 16, NULL, NULL, 0 }, /* The SHB field in a VA form instruction. */ #define SHB UIMM + 1 - { 4, 6, 0, 0, 0 }, + { 4, 6, NULL, NULL, 0 }, /* The other UIMM field in a EVX form instruction. */ #define EVUIMM SHB + 1 - { 5, 11, 0, 0, 0 }, + { 5, 11, NULL, NULL, 0 }, /* The other UIMM field in a half word EVX form instruction. */ #define EVUIMM_2 EVUIMM + 1 @@ -533,11 +534,11 @@ const struct powerpc_operand powerpc_operands[] = /* The WS field. */ #define WS EVUIMM_8 + 1 #define WS_MASK (0x7 << 11) - { 3, 11, 0, 0, 0 }, + { 3, 11, NULL, NULL, 0 }, /* The L field in an mtmsrd instruction */ #define MTMSRD_L WS + 1 - { 1, 16, 0, 0, PPC_OPERAND_OPTIONAL }, + { 1, 16, NULL, NULL, PPC_OPERAND_OPTIONAL }, }; -- cgit v1.2.3-70-g09d2 From 993fb38b1c0cd1a56d15aa99b337c780a538baa8 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 26 Apr 2005 07:43:42 -0700 Subject: [PATCH] amd64 rt_sigframe user annotation ->pretcode in struct rt_sigframe is a userland pointer (and already treated as such by code using that field). Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/signal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86_64/kernel/signal.c b/arch/x86_64/kernel/signal.c index 7760224cdfe..d439ced150c 100644 --- a/arch/x86_64/kernel/signal.c +++ b/arch/x86_64/kernel/signal.c @@ -83,7 +83,7 @@ sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss, struct rt_sigframe { - char *pretcode; + char __user *pretcode; struct ucontext uc; struct siginfo info; }; -- cgit v1.2.3-70-g09d2 From efa545791ff439d778f79d02255d59ff1f7ea0fc Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 26 Apr 2005 11:26:53 -0700 Subject: [PATCH] ppc64: trivial user annotations Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/rtas_flash.c | 12 ++++++------ arch/ppc64/kernel/scanlog.c | 4 ++-- 2 files changed, 8 insertions(+), 8 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/kernel/rtas_flash.c b/arch/ppc64/kernel/rtas_flash.c index 3213837282c..923e2e201a7 100644 --- a/arch/ppc64/kernel/rtas_flash.c +++ b/arch/ppc64/kernel/rtas_flash.c @@ -218,7 +218,7 @@ static void get_flash_status_msg(int status, char *buf) } /* Reading the proc file will show status (not the firmware contents) */ -static ssize_t rtas_flash_read(struct file *file, char *buf, +static ssize_t rtas_flash_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct proc_dir_entry *dp = PDE(file->f_dentry->d_inode); @@ -256,7 +256,7 @@ static ssize_t rtas_flash_read(struct file *file, char *buf, * count is. If the system is low on memory it will be just as well * that we fail.... */ -static ssize_t rtas_flash_write(struct file *file, const char *buffer, +static ssize_t rtas_flash_write(struct file *file, const char __user *buffer, size_t count, loff_t *off) { struct proc_dir_entry *dp = PDE(file->f_dentry->d_inode); @@ -356,7 +356,7 @@ static void manage_flash(struct rtas_manage_flash_t *args_buf) args_buf->status = rc; } -static ssize_t manage_flash_read(struct file *file, char *buf, +static ssize_t manage_flash_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct proc_dir_entry *dp = PDE(file->f_dentry->d_inode); @@ -386,7 +386,7 @@ static ssize_t manage_flash_read(struct file *file, char *buf, return msglen; } -static ssize_t manage_flash_write(struct file *file, const char *buf, +static ssize_t manage_flash_write(struct file *file, const char __user *buf, size_t count, loff_t *off) { struct proc_dir_entry *dp = PDE(file->f_dentry->d_inode); @@ -466,7 +466,7 @@ static int get_validate_flash_msg(struct rtas_validate_flash_t *args_buf, return n; } -static ssize_t validate_flash_read(struct file *file, char *buf, +static ssize_t validate_flash_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct proc_dir_entry *dp = PDE(file->f_dentry->d_inode); @@ -494,7 +494,7 @@ static ssize_t validate_flash_read(struct file *file, char *buf, return msglen; } -static ssize_t validate_flash_write(struct file *file, const char *buf, +static ssize_t validate_flash_write(struct file *file, const char __user *buf, size_t count, loff_t *off) { struct proc_dir_entry *dp = PDE(file->f_dentry->d_inode); diff --git a/arch/ppc64/kernel/scanlog.c b/arch/ppc64/kernel/scanlog.c index 189b81a4198..4d70736619c 100644 --- a/arch/ppc64/kernel/scanlog.c +++ b/arch/ppc64/kernel/scanlog.c @@ -43,7 +43,7 @@ static int scanlog_debug; static unsigned int ibm_scan_log_dump; /* RTAS token */ static struct proc_dir_entry *proc_ppc64_scan_log_dump; /* The proc file */ -static ssize_t scanlog_read(struct file *file, char *buf, +static ssize_t scanlog_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { struct inode * inode = file->f_dentry->d_inode; @@ -129,7 +129,7 @@ static ssize_t scanlog_read(struct file *file, char *buf, /*NOTREACHED*/ } -static ssize_t scanlog_write(struct file * file, const char * buf, +static ssize_t scanlog_write(struct file * file, const char __user * buf, size_t count, loff_t *ppos) { char stkbuf[20]; -- cgit v1.2.3-70-g09d2 From bdceb6a0162274934386f19f3ea5a9d44feb0b20 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 27 Apr 2005 18:04:45 -0700 Subject: [PATCH] ppc64: Fix return value of some vDSO calls The ppc vDSO would not properly clear the return value for some calls, which will be a problem when interfacing those calls with glibc. This should be fixed before 2.6.12 is released (as it is the first kernel with the ppc vDSO) so that we don't have to play with symbol versioning and ugly workarounds. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/vdso32/cacheflush.S | 2 ++ arch/ppc64/kernel/vdso32/gettimeofday.S | 1 + arch/ppc64/kernel/vdso64/cacheflush.S | 2 ++ 3 files changed, 5 insertions(+) (limited to 'arch') diff --git a/arch/ppc64/kernel/vdso32/cacheflush.S b/arch/ppc64/kernel/vdso32/cacheflush.S index c74fddb6afd..0ed7ea72171 100644 --- a/arch/ppc64/kernel/vdso32/cacheflush.S +++ b/arch/ppc64/kernel/vdso32/cacheflush.S @@ -47,6 +47,7 @@ V_FUNCTION_BEGIN(__kernel_sync_dicache) addi r6,r6,128 bdnz 1b isync + li r3,0 blr .cfi_endproc V_FUNCTION_END(__kernel_sync_dicache) @@ -59,6 +60,7 @@ V_FUNCTION_BEGIN(__kernel_sync_dicache_p5) .cfi_startproc sync isync + li r3,0 blr .cfi_endproc V_FUNCTION_END(__kernel_sync_dicache_p5) diff --git a/arch/ppc64/kernel/vdso32/gettimeofday.S b/arch/ppc64/kernel/vdso32/gettimeofday.S index ca7f415195c..2b48bf1fb10 100644 --- a/arch/ppc64/kernel/vdso32/gettimeofday.S +++ b/arch/ppc64/kernel/vdso32/gettimeofday.S @@ -58,6 +58,7 @@ V_FUNCTION_BEGIN(__kernel_gettimeofday) stw r5,TZONE_TZ_DSTTIME(r11) 1: mtlr r12 + li r3,0 blr 2: mr r3,r10 diff --git a/arch/ppc64/kernel/vdso64/cacheflush.S b/arch/ppc64/kernel/vdso64/cacheflush.S index d9696ffcf33..e0725b7b700 100644 --- a/arch/ppc64/kernel/vdso64/cacheflush.S +++ b/arch/ppc64/kernel/vdso64/cacheflush.S @@ -47,6 +47,7 @@ V_FUNCTION_BEGIN(__kernel_sync_dicache) addi r6,r6,128 bdnz 1b isync + li r3,0 blr .cfi_endproc V_FUNCTION_END(__kernel_sync_dicache) @@ -59,6 +60,7 @@ V_FUNCTION_BEGIN(__kernel_sync_dicache_p5) .cfi_startproc sync isync + li r3,0 blr .cfi_endproc V_FUNCTION_END(__kernel_sync_dicache_p5) -- cgit v1.2.3-70-g09d2 From c60c390620e0abb60d4ae8c43583714bda27763f Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Thu, 28 Apr 2005 22:47:29 -0700 Subject: [PATCH] x86_64: fix PT_NOTE addition to IA32 vDSO The addition of the PT_NOTE didn't take in the x86_64 version of the i386 vDSO, because I forgot the linker script bit in that copy. Signed-off-by: Roland McGrath Signed-off-by: Linus Torvalds --- arch/x86_64/ia32/vsyscall.lds | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch') diff --git a/arch/x86_64/ia32/vsyscall.lds b/arch/x86_64/ia32/vsyscall.lds index fa4b4dd4a9f..f2e75ed4c6c 100644 --- a/arch/x86_64/ia32/vsyscall.lds +++ b/arch/x86_64/ia32/vsyscall.lds @@ -36,6 +36,7 @@ SECTIONS .text.rtsigreturn : { *(.text.rtsigreturn) } :text =0x90909090 + .note : { *(.note.*) } :text :note .eh_frame_hdr : { *(.eh_frame_hdr) } :text :eh_frame_hdr .eh_frame : { KEEP (*(.eh_frame)) } :text .dynamic : { *(.dynamic) } :text :dynamic @@ -55,6 +56,7 @@ PHDRS { text PT_LOAD FILEHDR PHDRS FLAGS(5); /* PF_R|PF_X */ dynamic PT_DYNAMIC FLAGS(4); /* PF_R */ + note PT_NOTE FLAGS(4); /* PF_R */ eh_frame_hdr 0x6474e550; /* PT_GNU_EH_FRAME, but ld doesn't match the name */ } -- cgit v1.2.3-70-g09d2 From 2fd6f58ba6efc82ea2c9c2630f7ff5ed9eeaf34a Mon Sep 17 00:00:00 2001 From: Date: Fri, 29 Apr 2005 16:08:28 +0100 Subject: [AUDIT] Don't allow ptrace to fool auditing, log arch of audited syscalls. We were calling ptrace_notify() after auditing the syscall and arguments, but the debugger could have _changed_ them before the syscall was actually invoked. Reorder the calls to fix that. While we're touching ever call to audit_syscall_entry(), we also make it take an extra argument: the architecture of the syscall which was made, because some architectures allow more than one type of syscall. Also add an explicit success/failure flag to audit_syscall_exit(), for the benefit of architectures which return that in a condition register rather than only returning a single register. Change type of syscall return value to 'long' not 'int'. Signed-off-by: David Woodhouse --- arch/i386/kernel/ptrace.c | 19 +++++++++--------- arch/ia64/kernel/ptrace.c | 21 ++++++++++++-------- arch/mips/kernel/ptrace.c | 38 +++++++++++++++++++++++++---------- arch/ppc64/kernel/ptrace.c | 15 +++++++++----- arch/s390/kernel/ptrace.c | 21 ++++++++++---------- arch/x86_64/kernel/ptrace.c | 13 +++++++----- include/linux/audit.h | 48 +++++++++++++++++++++++++++++++++++++++++---- kernel/auditsc.c | 22 +++++++++++++++------ 8 files changed, 139 insertions(+), 58 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/ptrace.c b/arch/i386/kernel/ptrace.c index b2f17640cef..5606ec7a5c2 100644 --- a/arch/i386/kernel/ptrace.c +++ b/arch/i386/kernel/ptrace.c @@ -682,24 +682,18 @@ void do_syscall_trace(struct pt_regs *regs, int entryexit) /* do the secure computing check first */ secure_computing(regs->orig_eax); - if (unlikely(current->audit_context)) { - if (!entryexit) - audit_syscall_entry(current, regs->orig_eax, - regs->ebx, regs->ecx, - regs->edx, regs->esi); - else - audit_syscall_exit(current, regs->eax); - } + if (unlikely(current->audit_context) && entryexit) + audit_syscall_exit(current, AUDITSC_RESULT(regs->eax), regs->eax); if (!(current->ptrace & PT_PTRACED)) - return; + goto out; /* Fake a debug trap */ if (test_thread_flag(TIF_SINGLESTEP)) send_sigtrap(current, regs, 0); if (!test_thread_flag(TIF_SYSCALL_TRACE)) - return; + goto out; /* the 0x80 provides a way for the tracing parent to distinguish between a syscall stop and SIGTRAP delivery */ @@ -714,4 +708,9 @@ void do_syscall_trace(struct pt_regs *regs, int entryexit) send_sig(current->exit_code, current, 1); current->exit_code = 0; } + out: + if (unlikely(current->audit_context) && !entryexit) + audit_syscall_entry(current, AUDIT_ARCH_I386, regs->orig_eax, + regs->ebx, regs->ecx, regs->edx, regs->esi); + } diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 55789fcd721..8dde0b16d4c 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -1595,20 +1595,25 @@ syscall_trace_enter (long arg0, long arg1, long arg2, long arg3, long arg4, long arg5, long arg6, long arg7, struct pt_regs regs) { - long syscall; + if (test_thread_flag(TIF_SYSCALL_TRACE) + && (current->ptrace & PT_PTRACED)) + syscall_trace(); if (unlikely(current->audit_context)) { - if (IS_IA32_PROCESS(®s)) + long syscall; + int arch; + + if (IS_IA32_PROCESS(®s)) { syscall = regs.r1; - else + arch = AUDIT_ARCH_I386; + } else { syscall = regs.r15; + arch = AUDIT_ARCH_IA64; + } - audit_syscall_entry(current, syscall, arg0, arg1, arg2, arg3); + audit_syscall_entry(current, arch, syscall, arg0, arg1, arg2, arg3); } - if (test_thread_flag(TIF_SYSCALL_TRACE) - && (current->ptrace & PT_PTRACED)) - syscall_trace(); } /* "asmlinkage" so the input arguments are preserved... */ @@ -1619,7 +1624,7 @@ syscall_trace_leave (long arg0, long arg1, long arg2, long arg3, struct pt_regs regs) { if (unlikely(current->audit_context)) - audit_syscall_exit(current, regs.r8); + audit_syscall_exit(current, AUDITSC_RESULT(regs.r10), regs.r8); if (test_thread_flag(TIF_SYSCALL_TRACE) && (current->ptrace & PT_PTRACED)) diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index 92f2c39afe2..eaf7be9d0b0 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -300,25 +300,38 @@ out: return ret; } +static inline int audit_arch() +{ +#ifdef CONFIG_CPU_LITTLE_ENDIAN +#ifdef CONFIG_MIPS64 + if (!(current->thread.mflags & MF_32BIT_REGS)) + return AUDIT_ARCH_MIPSEL64; +#endif /* MIPS64 */ + return AUDIT_ARCH_MIPSEL; + +#else /* big endian... */ +#ifdef CONFIG_MIPS64 + if (!(current->thread.mflags & MF_32BIT_REGS)) + return AUDIT_ARCH_MIPS64; +#endif /* MIPS64 */ + return AUDIT_ARCH_MIPS; + +#endif /* endian */ +} + /* * Notification of system call entry/exit * - triggered by current->work.syscall_trace */ asmlinkage void do_syscall_trace(struct pt_regs *regs, int entryexit) { - if (unlikely(current->audit_context)) { - if (!entryexit) - audit_syscall_entry(current, regs->regs[2], - regs->regs[4], regs->regs[5], - regs->regs[6], regs->regs[7]); - else - audit_syscall_exit(current, regs->regs[2]); - } + if (unlikely(current->audit_context) && entryexit) + audit_syscall_exit(current, AUDITSC_RESULT(regs->regs[2]), regs->regs[2]); if (!test_thread_flag(TIF_SYSCALL_TRACE)) - return; + goto out; if (!(current->ptrace & PT_PTRACED)) - return; + goto out; /* The 0x80 provides a way for the tracing parent to distinguish between a syscall stop and SIGTRAP delivery */ @@ -334,4 +347,9 @@ asmlinkage void do_syscall_trace(struct pt_regs *regs, int entryexit) send_sig(current->exit_code, current, 1); current->exit_code = 0; } + out: + if (unlikely(current->audit_context) && !entryexit) + audit_syscall_entry(current, audit_arch(), regs->regs[2], + regs->regs[4], regs->regs[5], + regs->regs[6], regs->regs[7]); } diff --git a/arch/ppc64/kernel/ptrace.c b/arch/ppc64/kernel/ptrace.c index 354a287c67e..3c76333ec3a 100644 --- a/arch/ppc64/kernel/ptrace.c +++ b/arch/ppc64/kernel/ptrace.c @@ -304,14 +304,17 @@ static void do_syscall_trace(void) void do_syscall_trace_enter(struct pt_regs *regs) { + if (test_thread_flag(TIF_SYSCALL_TRACE) + && (current->ptrace & PT_PTRACED)) + do_syscall_trace(); + if (unlikely(current->audit_context)) - audit_syscall_entry(current, regs->gpr[0], + audit_syscall_entry(current, + test_thread_flag(TIF_32BIT)?AUDIT_ARCH_PPC:AUDIT_ARCH_PPC64, + regs->gpr[0], regs->gpr[3], regs->gpr[4], regs->gpr[5], regs->gpr[6]); - if (test_thread_flag(TIF_SYSCALL_TRACE) - && (current->ptrace & PT_PTRACED)) - do_syscall_trace(); } void do_syscall_trace_leave(struct pt_regs *regs) @@ -319,7 +322,9 @@ void do_syscall_trace_leave(struct pt_regs *regs) secure_computing(regs->gpr[0]); if (unlikely(current->audit_context)) - audit_syscall_exit(current, regs->result); + audit_syscall_exit(current, + (regs->ccr&0x1000)?AUDITSC_FAILURE:AUDITSC_SUCCESS, + regs->result); if ((test_thread_flag(TIF_SYSCALL_TRACE) || test_thread_flag(TIF_SINGLESTEP)) diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 647233c02fc..2d546c67f7c 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -711,18 +711,13 @@ out: asmlinkage void syscall_trace(struct pt_regs *regs, int entryexit) { - if (unlikely(current->audit_context)) { - if (!entryexit) - audit_syscall_entry(current, regs->gprs[2], - regs->orig_gpr2, regs->gprs[3], - regs->gprs[4], regs->gprs[5]); - else - audit_syscall_exit(current, regs->gprs[2]); - } + if (unlikely(current->audit_context) && entryexit) + audit_syscall_exit(current, AUDITSC_RESULT(regs->gprs[2]), regs->gprs[2]); + if (!test_thread_flag(TIF_SYSCALL_TRACE)) - return; + goto out; if (!(current->ptrace & PT_PTRACED)) - return; + goto out; ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD) ? 0x80 : 0)); @@ -735,4 +730,10 @@ syscall_trace(struct pt_regs *regs, int entryexit) send_sig(current->exit_code, current, 1); current->exit_code = 0; } + out: + if (unlikely(current->audit_context) && !entryexit) + audit_syscall_entry(current, + test_thread_flag(TIF_31BIT)?AUDIT_ARCH_S390:AUDIT_ARCH_S390X, + regs->gprs[2], regs->orig_gpr2, regs->gprs[3], + regs->gprs[4], regs->gprs[5]); } diff --git a/arch/x86_64/kernel/ptrace.c b/arch/x86_64/kernel/ptrace.c index c7011675007..ecbccbbf5c2 100644 --- a/arch/x86_64/kernel/ptrace.c +++ b/arch/x86_64/kernel/ptrace.c @@ -629,25 +629,28 @@ static void syscall_trace(struct pt_regs *regs) } } +#define audit_arch() (test_thread_flag(TIF_IA32) ? AUDIT_ARCH_I386 : AUDIT_ARCH_X86_64) + asmlinkage void syscall_trace_enter(struct pt_regs *regs) { /* do the secure computing check first */ secure_computing(regs->orig_rax); + if (test_thread_flag(TIF_SYSCALL_TRACE) + && (current->ptrace & PT_PTRACED)) + syscall_trace(regs); + if (unlikely(current->audit_context)) - audit_syscall_entry(current, regs->orig_rax, + audit_syscall_entry(current, audit_arch(), regs->orig_rax, regs->rdi, regs->rsi, regs->rdx, regs->r10); - if (test_thread_flag(TIF_SYSCALL_TRACE) - && (current->ptrace & PT_PTRACED)) - syscall_trace(regs); } asmlinkage void syscall_trace_leave(struct pt_regs *regs) { if (unlikely(current->audit_context)) - audit_syscall_exit(current, regs->rax); + audit_syscall_exit(current, AUDITSC_RESULT(regs->rax), regs->rax); if ((test_thread_flag(TIF_SYSCALL_TRACE) || test_thread_flag(TIF_SINGLESTEP)) diff --git a/include/linux/audit.h b/include/linux/audit.h index 9b77992c488..fad0c1dc21a 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -24,6 +24,9 @@ #ifndef _LINUX_AUDIT_H_ #define _LINUX_AUDIT_H_ +#include +#include + /* Request and reply types */ #define AUDIT_GET 1000 /* Get status */ #define AUDIT_SET 1001 /* Set status (enable/disable/auditd) */ @@ -67,6 +70,7 @@ #define AUDIT_FSGID 8 #define AUDIT_LOGINUID 9 #define AUDIT_PERS 10 +#define AUDIT_ARCH 11 /* These are ONLY useful when checking * at syscall exit time (AUDIT_AT_EXIT). */ @@ -96,6 +100,38 @@ #define AUDIT_FAIL_PRINTK 1 #define AUDIT_FAIL_PANIC 2 +/* distinguish syscall tables */ +#define __AUDIT_ARCH_64BIT 0x80000000 +#define __AUDIT_ARCH_LE 0x40000000 +#define AUDIT_ARCH_ALPHA (EM_ALPHA|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_ARM (EM_ARM|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_ARMEB (EM_ARM) +#define AUDIT_ARCH_CRIS (EM_CRIS|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_FRV (EM_FRV) +#define AUDIT_ARCH_H8300 (EM_H8_300) +#define AUDIT_ARCH_I386 (EM_386|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_IA64 (EM_IA_64|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_M32R (EM_M32R) +#define AUDIT_ARCH_M68K (EM_68K) +#define AUDIT_ARCH_MIPS (EM_MIPS) +#define AUDIT_ARCH_MIPSEL (EM_MIPS|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_MIPS64 (EM_MIPS|__AUDIT_ARCH_64BIT) +#define AUDIT_ARCH_MIPSEL64 (EM_MIPS|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_PARISC (EM_PARISC) +#define AUDIT_ARCH_PARISC64 (EM_PARISC|__AUDIT_ARCH_64BIT) +#define AUDIT_ARCH_PPC (EM_PPC) +#define AUDIT_ARCH_PPC64 (EM_PPC64|__AUDIT_ARCH_64BIT) +#define AUDIT_ARCH_S390 (EM_S390) +#define AUDIT_ARCH_S390X (EM_S390|__AUDIT_ARCH_64BIT) +#define AUDIT_ARCH_SH (EM_SH) +#define AUDIT_ARCH_SHEL (EM_SH|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_SH64 (EM_SH|__AUDIT_ARCH_64BIT) +#define AUDIT_ARCH_SHEL64 (EM_SH|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_SPARC (EM_SPARC) +#define AUDIT_ARCH_SPARC64 (EM_SPARC64|__AUDIT_ARCH_64BIT) +#define AUDIT_ARCH_V850 (EM_V850|__AUDIT_ARCH_LE) +#define AUDIT_ARCH_X86_64 (EM_X86_64|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE) + #ifndef __KERNEL__ struct audit_message { struct nlmsghdr nlh; @@ -129,15 +165,19 @@ struct audit_buffer; struct audit_context; struct inode; +#define AUDITSC_INVALID 0 +#define AUDITSC_SUCCESS 1 +#define AUDITSC_FAILURE 2 +#define AUDITSC_RESULT(x) ( ((long)(x))<0?AUDITSC_FAILURE:AUDITSC_SUCCESS ) #ifdef CONFIG_AUDITSYSCALL /* These are defined in auditsc.c */ /* Public API */ extern int audit_alloc(struct task_struct *task); extern void audit_free(struct task_struct *task); -extern void audit_syscall_entry(struct task_struct *task, +extern void audit_syscall_entry(struct task_struct *task, int arch, int major, unsigned long a0, unsigned long a1, unsigned long a2, unsigned long a3); -extern void audit_syscall_exit(struct task_struct *task, int return_code); +extern void audit_syscall_exit(struct task_struct *task, int failed, long return_code); extern void audit_getname(const char *name); extern void audit_putname(const char *name); extern void audit_inode(const char *name, const struct inode *inode); @@ -153,8 +193,8 @@ extern int audit_ipc_perms(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mo #else #define audit_alloc(t) ({ 0; }) #define audit_free(t) do { ; } while (0) -#define audit_syscall_entry(t,a,b,c,d,e) do { ; } while (0) -#define audit_syscall_exit(t,r) do { ; } while (0) +#define audit_syscall_entry(t,ta,a,b,c,d,e) do { ; } while (0) +#define audit_syscall_exit(t,f,r) do { ; } while (0) #define audit_getname(n) do { ; } while (0) #define audit_putname(n) do { ; } while (0) #define audit_inode(n,i) do { ; } while (0) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 00e87ffff13..77e92592de5 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -123,7 +123,7 @@ struct audit_context { int major; /* syscall number */ unsigned long argv[4]; /* syscall arguments */ int return_valid; /* return code is valid */ - int return_code;/* syscall return code */ + long return_code;/* syscall return code */ int auditable; /* 1 if record should be written */ int name_count; struct audit_names names[AUDIT_NAMES]; @@ -135,6 +135,7 @@ struct audit_context { uid_t uid, euid, suid, fsuid; gid_t gid, egid, sgid, fsgid; unsigned long personality; + int arch; #if AUDIT_DEBUG int put_count; @@ -348,6 +349,10 @@ static int audit_filter_rules(struct task_struct *tsk, case AUDIT_PERS: result = (tsk->personality == value); break; + case AUDIT_ARCH: + if (ctx) + result = (ctx->arch == value); + break; case AUDIT_EXIT: if (ctx && ctx->return_valid) @@ -355,7 +360,7 @@ static int audit_filter_rules(struct task_struct *tsk, break; case AUDIT_SUCCESS: if (ctx && ctx->return_valid) - result = (ctx->return_code >= 0); + result = (ctx->return_valid == AUDITSC_SUCCESS); break; case AUDIT_DEVMAJOR: if (ctx) { @@ -648,8 +653,11 @@ static void audit_log_exit(struct audit_context *context) audit_log_format(ab, "syscall=%d", context->major); if (context->personality != PER_LINUX) audit_log_format(ab, " per=%lx", context->personality); + audit_log_format(ab, " arch=%x", context->arch); if (context->return_valid) - audit_log_format(ab, " exit=%d", context->return_code); + audit_log_format(ab, " success=%s exit=%ld", + (context->return_valid==AUDITSC_SUCCESS)?"yes":"no", + context->return_code); audit_log_format(ab, " a0=%lx a1=%lx a2=%lx a3=%lx items=%d" " pid=%d loginuid=%d uid=%d gid=%d" @@ -773,7 +781,7 @@ static inline unsigned int audit_serial(void) * then the record will be written at syscall exit time (otherwise, it * will only be written if another part of the kernel requests that it * be written). */ -void audit_syscall_entry(struct task_struct *tsk, int major, +void audit_syscall_entry(struct task_struct *tsk, int arch, int major, unsigned long a1, unsigned long a2, unsigned long a3, unsigned long a4) { @@ -827,6 +835,7 @@ void audit_syscall_entry(struct task_struct *tsk, int major, if (!audit_enabled) return; + context->arch = arch; context->major = major; context->argv[0] = a1; context->argv[1] = a2; @@ -850,13 +859,13 @@ void audit_syscall_entry(struct task_struct *tsk, int major, * filtering, or because some other part of the kernel write an audit * message), then write out the syscall information. In call cases, * free the names stored from getname(). */ -void audit_syscall_exit(struct task_struct *tsk, int return_code) +void audit_syscall_exit(struct task_struct *tsk, int valid, long return_code) { struct audit_context *context; get_task_struct(tsk); task_lock(tsk); - context = audit_get_context(tsk, 1, return_code); + context = audit_get_context(tsk, valid, return_code); task_unlock(tsk); /* Not having a context here is ok, since the parent may have @@ -869,6 +878,7 @@ void audit_syscall_exit(struct task_struct *tsk, int return_code) context->in_syscall = 0; context->auditable = 0; + if (context->previous) { struct audit_context *new_context = context->previous; context->previous = NULL; -- cgit v1.2.3-70-g09d2 From 3ac3ed555bec5b1f92bb22cb94823a0e99d0f320 Mon Sep 17 00:00:00 2001 From: Amy Griffis Date: Fri, 29 Apr 2005 16:12:55 +0100 Subject: [PATCH] fix ia64 syscall auditing Attached is a patch against David's audit.17 kernel that adds checks for the TIF_SYSCALL_AUDIT thread flag to the ia64 system call and signal handling code paths.The patch enables auditing of system calls set up via fsys_bubble_down, as well as ensuring that audit_syscall_exit() is called on return from sigreturn. Neglecting to check for TIF_SYSCALL_AUDIT at these points results in incorrect information in audit_context, causing frequent system panics when system call auditing is enabled on an ia64 system. Signed-off-by: Amy Griffis Signed-off-by: David Woodhouse --- arch/ia64/kernel/fsys.S | 4 +++- arch/ia64/kernel/signal.c | 3 ++- 2 files changed, 5 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S index 0d8650f7fce..4f3cdef7579 100644 --- a/arch/ia64/kernel/fsys.S +++ b/arch/ia64/kernel/fsys.S @@ -611,8 +611,10 @@ GLOBAL_ENTRY(fsys_bubble_down) movl r2=ia64_ret_from_syscall ;; mov rp=r2 // set the real return addr - tbit.z p8,p0=r3,TIF_SYSCALL_TRACE + and r3=_TIF_SYSCALL_TRACEAUDIT,r3 ;; + cmp.eq p8,p0=r3,r0 + (p10) br.cond.spnt.many ia64_ret_from_syscall // p10==true means out registers are more than 8 (p8) br.call.sptk.many b6=b6 // ignore this return addr br.cond.sptk ia64_trace_syscall diff --git a/arch/ia64/kernel/signal.c b/arch/ia64/kernel/signal.c index 6891d86937d..64555741872 100644 --- a/arch/ia64/kernel/signal.c +++ b/arch/ia64/kernel/signal.c @@ -224,7 +224,8 @@ ia64_rt_sigreturn (struct sigscratch *scr) * could be corrupted. */ retval = (long) &ia64_leave_kernel; - if (test_thread_flag(TIF_SYSCALL_TRACE)) + if (test_thread_flag(TIF_SYSCALL_TRACE) + || test_thread_flag(TIF_SYSCALL_AUDIT)) /* * strace expects to be notified after sigreturn returns even though the * context to which we return may not be in the middle of a syscall. -- cgit v1.2.3-70-g09d2 From 67eb81e1686f44bcf8f005b296213fd2c21b4719 Mon Sep 17 00:00:00 2001 From: Yoichi Yuasa Date: Fri, 29 Apr 2005 16:13:35 +0100 Subject: mips: warning fix audit_arch() arch/mips/kernel/ptrace.c:305: warning: function declaration isn't a prototype Signed-off-by: Yoichi Yuasa Signed-off-by: Andrew Morton Signed-off-by: David Woodhouse --- arch/mips/kernel/ptrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index eaf7be9d0b0..4abc2ee53b4 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -300,7 +300,7 @@ out: return ret; } -static inline int audit_arch() +static inline int audit_arch(void) { #ifdef CONFIG_CPU_LITTLE_ENDIAN #ifdef CONFIG_MIPS64 -- cgit v1.2.3-70-g09d2 From a879cbbb34cbecfa9707fbb6e5a00c503ac1ecb9 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 29 Apr 2005 09:38:44 -0700 Subject: x86: make traps on 'iret' be debuggable in user space This makes a trap on the 'iret' that returns us to user space cause a nice clean SIGSEGV, instead of just a hard (and silent) exit. That way a debugger can actually try to see what happened, and we also properly notify everybody who might be interested about us being gone. This loses the error code, but tells the debugger what happened with ILL_BADSTK in the siginfo. --- arch/i386/kernel/entry.S | 8 +++----- arch/i386/kernel/traps.c | 1 + 2 files changed, 4 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S index 3c73dc865ea..fe1918cc68d 100644 --- a/arch/i386/kernel/entry.S +++ b/arch/i386/kernel/entry.S @@ -260,11 +260,9 @@ restore_nocheck: .section .fixup,"ax" iret_exc: sti - movl $__USER_DS, %edx - movl %edx, %ds - movl %edx, %es - movl $11,%eax - call do_exit + pushl $0 # no error code + pushl $do_iret_error + jmp error_code .previous .section __ex_table,"a" .align 4 diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index 6c0e383915b..d70819481f6 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -451,6 +451,7 @@ DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS) DO_ERROR(11, SIGBUS, "segment not present", segment_not_present) DO_ERROR(12, SIGBUS, "stack segment", stack_segment) DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0) +DO_ERROR_INFO(32, SIGSEGV, "iret exception", iret_error, ILL_BADSTK, 0) fastcall void do_general_protection(struct pt_regs * regs, long error_code) { -- cgit v1.2.3-70-g09d2 From 8443b165f13d21214e5d5495eee7c3bf7f2456bf Mon Sep 17 00:00:00 2001 From: Lennert Buytenhek Date: Fri, 29 Apr 2005 21:58:15 +0100 Subject: [PATCH] ARM: 2657/1: export ixp2000_pci_config_addr Patch from Lennert Buytenhek Export ixp2000_pci_config_addr, to be used by the IXDP2800 platform setup code to coordinate booting the master and slave NPU. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King --- arch/arm/mach-ixp2000/pci.c | 2 +- include/asm-arm/arch-ixp2000/platform.h | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-ixp2000/pci.c b/arch/arm/mach-ixp2000/pci.c index 831f8ffb6b6..36c9a94298c 100644 --- a/arch/arm/mach-ixp2000/pci.c +++ b/arch/arm/mach-ixp2000/pci.c @@ -37,7 +37,7 @@ static int pci_master_aborts = 0; static int clear_master_aborts(void); -static u32 * +u32 * ixp2000_pci_config_addr(unsigned int bus_nr, unsigned int devfn, int where) { u32 *paddress; diff --git a/include/asm-arm/arch-ixp2000/platform.h b/include/asm-arm/arch-ixp2000/platform.h index 509e44d528d..901bba6d02b 100644 --- a/include/asm-arm/arch-ixp2000/platform.h +++ b/include/asm-arm/arch-ixp2000/platform.h @@ -121,6 +121,7 @@ unsigned long ixp2000_gettimeoffset(void); struct pci_sys_data; +u32 *ixp2000_pci_config_addr(unsigned int bus, unsigned int devfn, int where); void ixp2000_pci_preinit(void); int ixp2000_pci_setup(int, struct pci_sys_data*); struct pci_bus* ixp2000_pci_scan_bus(int, struct pci_sys_data*); -- cgit v1.2.3-70-g09d2 From ae36bf5861e1091dd337f0b475e043ab07d4a937 Mon Sep 17 00:00:00 2001 From: Lennert Buytenhek Date: Fri, 29 Apr 2005 21:58:15 +0100 Subject: [PATCH] ARM: 2658/1: start ixp2000 pci memory resource at 0xe0000000 Patch from Lennert Buytenhek On the IXDP2800, the bootloader does an awful job of configuring the PCI bus, so we make linux reconfigure everything. Having a 1:1 pci:phys address mapping generally simplifies everything, so try to allocate PCI addresses from the [e0000000..ffffffff] range, which is the physical address range of the outbound PCI window on the IXP2000. This does not affect any of the other IXP2000 platforms since they all use their bootloader's PCI resource assignment. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King --- arch/arm/mach-ixp2000/pci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-ixp2000/pci.c b/arch/arm/mach-ixp2000/pci.c index 36c9a94298c..3844d5c7cf8 100644 --- a/arch/arm/mach-ixp2000/pci.c +++ b/arch/arm/mach-ixp2000/pci.c @@ -208,7 +208,7 @@ ixp2000_pci_preinit(void) * use our own resource space. */ static struct resource ixp2000_pci_mem_space = { - .start = 0x00000000, + .start = 0xe0000000, .end = 0xffffffff, .flags = IORESOURCE_MEM, .name = "PCI Mem Space" -- cgit v1.2.3-70-g09d2 From 458a83fa43e83505f9401783ce9ed41b5a8b5591 Mon Sep 17 00:00:00 2001 From: Lennert Buytenhek Date: Fri, 29 Apr 2005 21:58:16 +0100 Subject: [PATCH] ARM: 2659/1: do not assign PCI I/O address zero on IXP2000 Patch from Lennert Buytenhek Assigning the address zero to a PCI device BAR causes some part of the PCI subsystem to believe that resource allocation for that BAR failed due to resource conflicts, which will make attempts to enable the device fail. Work around this by assigning I/O addresses starting from 00010000. While we're at it, make the PCI I/O resource end at 0001ffff, since we only have 64k of outbound I/O window on the IXP2000, and we don't do bank switching. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King --- arch/arm/mach-ixp2000/pci.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-ixp2000/pci.c b/arch/arm/mach-ixp2000/pci.c index 3844d5c7cf8..5ff2f2718c5 100644 --- a/arch/arm/mach-ixp2000/pci.c +++ b/arch/arm/mach-ixp2000/pci.c @@ -215,8 +215,8 @@ static struct resource ixp2000_pci_mem_space = { }; static struct resource ixp2000_pci_io_space = { - .start = 0x00000000, - .end = 0xffffffff, + .start = 0x00010000, + .end = 0x0001ffff, .flags = IORESOURCE_IO, .name = "PCI I/O Space" }; -- cgit v1.2.3-70-g09d2 From 3a1e501511a1e2c665c566939047794dcf86466b Mon Sep 17 00:00:00 2001 From: "George G. Davis" Date: Fri, 29 Apr 2005 22:08:33 +0100 Subject: [PATCH] ARM: 2655/1: ARM1136 SWP instruction abort handler fix Patch from George G. Davis As noted in http://www.arm.com/linux/patch-2.6.9-arm1.gz, the "Faulty SWP instruction on 1136 doesn't set bit 11 in DFSR." So the v6_early_abort handler does not report the correct rd/wr direction for the SWP instruction which may result in SEGVS or hangs. In order to work around this problem, this patch merely updates the fix contained in the ARM Ltd. patch to use the macroised abort handler fixups. Signed-off-by: George G. Davis Signed-off-by: Russell King --- arch/arm/mm/abort-ev6.S | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'arch') diff --git a/arch/arm/mm/abort-ev6.S b/arch/arm/mm/abort-ev6.S index 38b2cbb89be..8f76f3df7b4 100644 --- a/arch/arm/mm/abort-ev6.S +++ b/arch/arm/mm/abort-ev6.S @@ -1,5 +1,6 @@ #include #include +#include "abort-macro.S" /* * Function: v6_early_abort * @@ -13,11 +14,26 @@ * : sp = pointer to registers * * Purpose : obtain information about current aborted instruction. + * Note: we read user space. This means we might cause a data + * abort here if the I-TLB and D-TLB aren't seeing the same + * picture. Unfortunately, this does happen. We live with it. */ .align 5 ENTRY(v6_early_abort) mrc p15, 0, r1, c5, c0, 0 @ get FSR mrc p15, 0, r0, c6, c0, 0 @ get FAR +/* + * Faulty SWP instruction on 1136 doesn't set bit 11 in DFSR. + * The test below covers all the write situations, including Java bytecodes + */ + bic r1, r1, #1 << 11 | 1 << 10 @ clear bits 11 and 10 of FSR + tst r3, #PSR_J_BIT @ Java? + movne pc, lr + do_thumb_abort + ldreq r3, [r2] @ read aborted ARM instruction + do_ldrd_abort + tst r3, #1 << 20 @ L = 0 -> write + orreq r1, r1, #1 << 11 @ yes. mov pc, lr -- cgit v1.2.3-70-g09d2 From 2d2669b62984b8d76b05a6a045390a3250317d21 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Fri, 29 Apr 2005 22:08:33 +0100 Subject: [PATCH] ARM: 2651/3: kernel helpers for NPTL support Patch from Nicolas Pitre This patch entirely reworks the kernel assistance for NPTL on ARM. In particular this provides an efficient way to retrieve the TLS value and perform atomic operations without any instruction emulation nor special system call. This even allows for pre ARMv6 binaries to be forward compatible with SMP systems without any penalty. The problematic and performance critical operations are performed through segment of kernel provided user code reachable from user space at a fixed address in kernel memory. Those fixed entry points are within the vector page so we basically get it for free as no extra memory page is required and nothing else may be mapped at that location anyway. This is different from (but doesn't preclude) a full blown VDSO implementation, however a VDSO would prevent some assembly tricks with constants that allows for efficient branching to those code segments. And since those code segments only use a few cycles before returning to user code, the overhead of a VDSO far call would add a significant overhead to such minimalistic operations. The ARM_NR_set_tls syscall also changed number. This is done for two reasons: 1) this patch changes the way the TLS value was previously meant to be retrieved, therefore we ensure whatever library using the old way gets fixed (they only exist in private tree at the moment since the NPTL work is still progressing). 2) the previous number was allocated in a range causing an undefined instruction trap on kernels not supporting that syscall and it was determined that allocating it in a range returning -ENOSYS would be much nicer for libraries trying to determine if the feature is present or not. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King --- arch/arm/kernel/entry-armv.S | 213 ++++++++++++++++++++++++++++++++++++++++++- arch/arm/kernel/traps.c | 58 ++++++++++-- arch/arm/mm/Kconfig | 14 +++ include/asm-arm/unistd.h | 3 +- 4 files changed, 277 insertions(+), 11 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 2a5c3fe09a9..080df907f24 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -269,6 +269,12 @@ __pabt_svc: add r5, sp, #S_PC ldmia r7, {r2 - r4} @ Get USR pc, cpsr +#if __LINUX_ARM_ARCH__ < 6 + @ make sure our user space atomic helper is aborted + cmp r2, #VIRT_OFFSET + bichs r3, r3, #PSR_Z_BIT +#endif + @ @ We are now ready to fill in the remaining blanks on the stack: @ @@ -499,8 +505,12 @@ ENTRY(__switch_to) mra r4, r5, acc0 stmia ip, {r4, r5} #endif +#ifdef CONFIG_HAS_TLS_REG + mcr p15, 0, r3, c13, c0, 3 @ set TLS register +#else mov r4, #0xffff0fff - str r3, [r4, #-3] @ Set TLS ptr + str r3, [r4, #-15] @ TLS val at 0xffff0ff0 +#endif mcr p15, 0, r6, c3, c0, 0 @ Set domain register #ifdef CONFIG_VFP @ Always disable VFP so we can lazily save/restore the old @@ -519,6 +529,207 @@ ENTRY(__switch_to) ldmib r2, {r4 - sl, fp, sp, pc} @ Load all regs saved previously __INIT + +/* + * User helpers. + * + * These are segment of kernel provided user code reachable from user space + * at a fixed address in kernel memory. This is used to provide user space + * with some operations which require kernel help because of unimplemented + * native feature and/or instructions in many ARM CPUs. The idea is for + * this code to be executed directly in user mode for best efficiency but + * which is too intimate with the kernel counter part to be left to user + * libraries. In fact this code might even differ from one CPU to another + * depending on the available instruction set and restrictions like on + * SMP systems. In other words, the kernel reserves the right to change + * this code as needed without warning. Only the entry points and their + * results are guaranteed to be stable. + * + * Each segment is 32-byte aligned and will be moved to the top of the high + * vector page. New segments (if ever needed) must be added in front of + * existing ones. This mechanism should be used only for things that are + * really small and justified, and not be abused freely. + * + * User space is expected to implement those things inline when optimizing + * for a processor that has the necessary native support, but only if such + * resulting binaries are already to be incompatible with earlier ARM + * processors due to the use of unsupported instructions other than what + * is provided here. In other words don't make binaries unable to run on + * earlier processors just for the sake of not using these kernel helpers + * if your compiled code is not going to use the new instructions for other + * purpose. + */ + + .align 5 + .globl __kuser_helper_start +__kuser_helper_start: + +/* + * Reference prototype: + * + * int __kernel_cmpxchg(int oldval, int newval, int *ptr) + * + * Input: + * + * r0 = oldval + * r1 = newval + * r2 = ptr + * lr = return address + * + * Output: + * + * r0 = returned value (zero or non-zero) + * C flag = set if r0 == 0, clear if r0 != 0 + * + * Clobbered: + * + * r3, ip, flags + * + * Definition and user space usage example: + * + * typedef int (__kernel_cmpxchg_t)(int oldval, int newval, int *ptr); + * #define __kernel_cmpxchg (*(__kernel_cmpxchg_t *)0xffff0fc0) + * + * Atomically store newval in *ptr if *ptr is equal to oldval for user space. + * Return zero if *ptr was changed or non-zero if no exchange happened. + * The C flag is also set if *ptr was changed to allow for assembly + * optimization in the calling code. + * + * For example, a user space atomic_add implementation could look like this: + * + * #define atomic_add(ptr, val) \ + * ({ register unsigned int *__ptr asm("r2") = (ptr); \ + * register unsigned int __result asm("r1"); \ + * asm volatile ( \ + * "1: @ atomic_add\n\t" \ + * "ldr r0, [r2]\n\t" \ + * "mov r3, #0xffff0fff\n\t" \ + * "add lr, pc, #4\n\t" \ + * "add r1, r0, %2\n\t" \ + * "add pc, r3, #(0xffff0fc0 - 0xffff0fff)\n\t" \ + * "bcc 1b" \ + * : "=&r" (__result) \ + * : "r" (__ptr), "rIL" (val) \ + * : "r0","r3","ip","lr","cc","memory" ); \ + * __result; }) + */ + +__kuser_cmpxchg: @ 0xffff0fc0 + +#if __LINUX_ARM_ARCH__ < 6 + +#ifdef CONFIG_SMP /* sanity check */ +#error "CONFIG_SMP on a machine supporting pre-ARMv6 processors?" +#endif + + /* + * Theory of operation: + * + * We set the Z flag before loading oldval. If ever an exception + * occurs we can not be sure the loaded value will still be the same + * when the exception returns, therefore the user exception handler + * will clear the Z flag whenever the interrupted user code was + * actually from the kernel address space (see the usr_entry macro). + * + * The post-increment on the str is used to prevent a race with an + * exception happening just after the str instruction which would + * clear the Z flag although the exchange was done. + */ + teq ip, ip @ set Z flag + ldr ip, [r2] @ load current val + add r3, r2, #1 @ prepare store ptr + teqeq ip, r0 @ compare with oldval if still allowed + streq r1, [r3, #-1]! @ store newval if still allowed + subs r0, r2, r3 @ if r2 == r3 the str occured + mov pc, lr + +#else + + ldrex r3, [r2] + subs r3, r3, r0 + strexeq r3, r1, [r2] + rsbs r0, r3, #0 + mov pc, lr + +#endif + + .align 5 + +/* + * Reference prototype: + * + * int __kernel_get_tls(void) + * + * Input: + * + * lr = return address + * + * Output: + * + * r0 = TLS value + * + * Clobbered: + * + * the Z flag might be lost + * + * Definition and user space usage example: + * + * typedef int (__kernel_get_tls_t)(void); + * #define __kernel_get_tls (*(__kernel_get_tls_t *)0xffff0fe0) + * + * Get the TLS value as previously set via the __ARM_NR_set_tls syscall. + * + * This could be used as follows: + * + * #define __kernel_get_tls() \ + * ({ register unsigned int __val asm("r0"); \ + * asm( "mov r0, #0xffff0fff; mov lr, pc; sub pc, r0, #31" \ + * : "=r" (__val) : : "lr","cc" ); \ + * __val; }) + */ + +__kuser_get_tls: @ 0xffff0fe0 + +#ifndef CONFIG_HAS_TLS_REG + +#ifdef CONFIG_SMP /* sanity check */ +#error "CONFIG_SMP without CONFIG_HAS_TLS_REG is wrong" +#endif + + ldr r0, [pc, #(16 - 8)] @ TLS stored at 0xffff0ff0 + mov pc, lr + +#else + + mrc p15, 0, r0, c13, c0, 3 @ read TLS register + mov pc, lr + +#endif + + .rep 5 + .word 0 @ pad up to __kuser_helper_version + .endr + +/* + * Reference declaration: + * + * extern unsigned int __kernel_helper_version; + * + * Definition and user space usage example: + * + * #define __kernel_helper_version (*(unsigned int *)0xffff0ffc) + * + * User space may read this to determine the curent number of helpers + * available. + */ + +__kuser_helper_version: @ 0xffff0ffc + .word ((__kuser_helper_end - __kuser_helper_start) >> 5) + + .globl __kuser_helper_end +__kuser_helper_end: + + /* * Vector stubs. * diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 0078aeb8573..3a001fe5540 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -450,13 +450,17 @@ asmlinkage int arm_syscall(int no, struct pt_regs *regs) case NR(set_tls): thread->tp_value = regs->ARM_r0; +#ifdef CONFIG_HAS_TLS_REG + asm ("mcr p15, 0, %0, c13, c0, 3" : : "r" (regs->ARM_r0) ); +#else /* - * Our user accessible TLS ptr is located at 0xffff0ffc. - * On SMP read access to this address must raise a fault - * and be emulated from the data abort handler. - * m + * User space must never try to access this directly. + * Expect your app to break eventually if you do so. + * The user helper at 0xffff0fe0 must be used instead. + * (see entry-armv.S for details) */ - *((unsigned long *)0xffff0ffc) = thread->tp_value; + *((unsigned int *)0xffff0ff0) = regs->ARM_r0; +#endif return 0; default: @@ -493,6 +497,41 @@ asmlinkage int arm_syscall(int no, struct pt_regs *regs) return 0; } +#if defined(CONFIG_CPU_32v6) && !defined(CONFIG_HAS_TLS_REG) + +/* + * We might be running on an ARMv6+ processor which should have the TLS + * register, but for some reason we can't use it and have to emulate it. + */ + +static int get_tp_trap(struct pt_regs *regs, unsigned int instr) +{ + int reg = (instr >> 12) & 15; + if (reg == 15) + return 1; + regs->uregs[reg] = current_thread_info()->tp_value; + regs->ARM_pc += 4; + return 0; +} + +static struct undef_hook arm_mrc_hook = { + .instr_mask = 0x0fff0fff, + .instr_val = 0x0e1d0f70, + .cpsr_mask = PSR_T_BIT, + .cpsr_val = 0, + .fn = get_tp_trap, +}; + +static int __init arm_mrc_hook_init(void) +{ + register_undef_hook(&arm_mrc_hook); + return 0; +} + +late_initcall(arm_mrc_hook_init); + +#endif + void __bad_xchg(volatile void *ptr, int size) { printk("xchg: bad data size: pc 0x%p, ptr 0x%p, size %d\n", @@ -580,14 +619,17 @@ void __init trap_init(void) { extern char __stubs_start[], __stubs_end[]; extern char __vectors_start[], __vectors_end[]; + extern char __kuser_helper_start[], __kuser_helper_end[]; + int kuser_sz = __kuser_helper_end - __kuser_helper_start; /* - * Copy the vectors and stubs (in entry-armv.S) into the - * vector page, mapped at 0xffff0000, and ensure these are - * visible to the instruction stream. + * Copy the vectors, stubs and kuser helpers (in entry-armv.S) + * into the vector page, mapped at 0xffff0000, and ensure these + * are visible to the instruction stream. */ memcpy((void *)0xffff0000, __vectors_start, __vectors_end - __vectors_start); memcpy((void *)0xffff0200, __stubs_start, __stubs_end - __stubs_start); + memcpy((void *)0xffff1000 - kuser_sz, __kuser_helper_start, kuser_sz); flush_icache_range(0xffff0000, 0xffff0000 + PAGE_SIZE); modify_domain(DOMAIN_USER, DOMAIN_CLIENT); } diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index 5b670c9ac5e..007766a0644 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -409,3 +409,17 @@ config CPU_BPREDICT_DISABLE depends on CPU_ARM1020 help Say Y here to disable branch prediction. If unsure, say N. + +config HAS_TLS_REG + bool + depends on CPU_32v6 && !CPU_32v5 && !CPU_32v4 && !CPU_32v3 + help + This selects support for the CP15 thread register. + It is defined to be available on ARMv6 or later. However + if the kernel is configured to support multiple CPUs including + a pre-ARMv6 processors, or if a given ARMv6 processor doesn't + implement the thread register for some reason, then access to + this register from user space must be trapped and emulated. + If user space is relying on the __kuser_get_tls code then + there should not be any impact. + diff --git a/include/asm-arm/unistd.h b/include/asm-arm/unistd.h index a19ec09eaa0..ace27480886 100644 --- a/include/asm-arm/unistd.h +++ b/include/asm-arm/unistd.h @@ -359,8 +359,7 @@ #define __ARM_NR_cacheflush (__ARM_NR_BASE+2) #define __ARM_NR_usr26 (__ARM_NR_BASE+3) #define __ARM_NR_usr32 (__ARM_NR_BASE+4) - -#define __ARM_NR_set_tls (__ARM_NR_BASE+0x800) +#define __ARM_NR_set_tls (__ARM_NR_BASE+5) #define __sys2(x) #x #define __sys1(x) __sys2(x) -- cgit v1.2.3-70-g09d2 From ca315159dfa80a2bfc7d917a717a7ee8d771bdf9 Mon Sep 17 00:00:00 2001 From: "George G. Davis" Date: Fri, 29 Apr 2005 22:08:35 +0100 Subject: [PATCH] ARM: 2656/1: Access permission bits are wrong for kernel XIP sections on ARMv6 Patch from George G. Davis This patch is required for kernel XIP support on ARMv6 machines. It ensures that the access permission bits for kernel XIP section descriptors are APX=1 and AP[1:0]=01, which is Kernel read-only/User no access permissions. Prior to this change, kernel XIP section descriptor access permissions were set to Kernel no access/User no access on ARMv6 machines and the kernel would therefore hang upon entry to userspace when set_fs(USER_DS) was executed. Signed-off-by: Steve Longerbeam Signed-off-by: George G. Davis Signed-off-by: Russell King --- arch/arm/mm/mm-armv.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/mm/mm-armv.c b/arch/arm/mm/mm-armv.c index f5a87db8b49..585dfb8e20b 100644 --- a/arch/arm/mm/mm-armv.c +++ b/arch/arm/mm/mm-armv.c @@ -411,9 +411,10 @@ static void __init build_mem_type_table(void) mem_types[MT_MEMORY].prot_sect &= ~PMD_BIT4; mem_types[MT_ROM].prot_sect &= ~PMD_BIT4; /* - * Mark cache clean areas read only from SVC mode - * and no access from userspace. + * Mark cache clean areas and XIP ROM read only + * from SVC mode and no access from userspace. */ + mem_types[MT_ROM].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; mem_types[MT_MINICLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; } -- cgit v1.2.3-70-g09d2 From 53e173f62c318e65e6ae13524b04c5cf38c1bc3c Mon Sep 17 00:00:00 2001 From: Lennert Buytenhek Date: Fri, 29 Apr 2005 22:13:57 +0100 Subject: [PATCH] ARM: 2660/2: fix ixdp2800 boot and pci init Patch from Lennert Buytenhek The IXDP2800 is an evalution platform for the IXP2800 processor that has two IXP2800s connected to the same PCI bus. This is problematic as both CPUs will try to configure the PCI bus as they boot linux. Contrary to on the other IXP2000 platforms, the boot loader on the IXDP2800 doesn't configure the PCI bus properly, so we do want the linux instance on one of the CPUs to do that. Making one of the CPUs ignore the PCI bus (and thus act like a pure PCI slave device) is not an option because there is a 82559 NIC on the PCI bus for each of the CPUs. The chosen solution is to have the master CPU configure the PCI bus while the slave is kept in a quiescent state, and then to have the slave CPU scan the PCI bus (without assigning resources) while the master is kept in a quiescent state. After this ritual, the master deletes the slave NIC from its PCI device list, the slave deletes the master NIC from its device list, and (almost) all is well. There's still one little problem: each of the CPUs has a 1G SDRAM BAR, but the IXP2000 only has 512M of outbound PCI memory window. We solve this by hand-assigning the master and slave SDRAM BARs to a location outside each of the IXP's outbound PCI windows, and by having the rest of the BARs autoconfigured in the outbound PCI windows, in the range [e0000000..ffffffff], so that there is a 1:1 pci:phys mapping between them. Even with this patch, a number of issues still remain -- just imagine what happens if one of the CPUs is rebooted, by watchdog or by hand, but the other one isn't. But those issues are not easily fixable given the strange PCI layout of this board and the behavior of the boot loader shipped with the platform. Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King --- arch/arm/configs/ixdp2800_defconfig | 2 +- arch/arm/mach-ixp2000/ixdp2800.c | 147 ++++++++++++++++++++++++++++++++---- 2 files changed, 135 insertions(+), 14 deletions(-) (limited to 'arch') diff --git a/arch/arm/configs/ixdp2800_defconfig b/arch/arm/configs/ixdp2800_defconfig index d36f9919296..7be3521f91f 100644 --- a/arch/arm/configs/ixdp2800_defconfig +++ b/arch/arm/configs/ixdp2800_defconfig @@ -133,7 +133,7 @@ CONFIG_ALIGNMENT_TRAP=y # CONFIG_ZBOOT_ROM_TEXT=0x0 CONFIG_ZBOOT_ROM_BSS=0x0 -CONFIG_CMDLINE="console=ttyS0,9600 root=/dev/nfs ip=bootp mem=64M@0x0 pci=firmware" +CONFIG_CMDLINE="console=ttyS0,9600 root=/dev/nfs ip=bootp mem=64M@0x0" # CONFIG_XIP_KERNEL is not set # diff --git a/arch/arm/mach-ixp2000/ixdp2800.c b/arch/arm/mach-ixp2000/ixdp2800.c index c4683aaff84..aec13c7108a 100644 --- a/arch/arm/mach-ixp2000/ixdp2800.c +++ b/arch/arm/mach-ixp2000/ixdp2800.c @@ -65,19 +65,102 @@ static struct sys_timer ixdp2800_timer = { /************************************************************************* * IXDP2800 PCI *************************************************************************/ +static void __init ixdp2800_slave_disable_pci_master(void) +{ + *IXP2000_PCI_CMDSTAT &= ~(PCI_COMMAND_MASTER | PCI_COMMAND_MEMORY); +} + +static void __init ixdp2800_master_wait_for_slave(void) +{ + volatile u32 *addr; + + printk(KERN_INFO "IXDP2800: waiting for slave NPU to configure " + "its BAR sizes\n"); + + addr = ixp2000_pci_config_addr(0, IXDP2X00_SLAVE_NPU_DEVFN, + PCI_BASE_ADDRESS_1); + do { + *addr = 0xffffffff; + cpu_relax(); + } while (*addr != 0xfe000008); + + addr = ixp2000_pci_config_addr(0, IXDP2X00_SLAVE_NPU_DEVFN, + PCI_BASE_ADDRESS_2); + do { + *addr = 0xffffffff; + cpu_relax(); + } while (*addr != 0xc0000008); + + /* + * Configure the slave's SDRAM BAR by hand. + */ + *addr = 0x40000008; +} + +static void __init ixdp2800_slave_wait_for_master_enable(void) +{ + printk(KERN_INFO "IXDP2800: waiting for master NPU to enable us\n"); + + while ((*IXP2000_PCI_CMDSTAT & PCI_COMMAND_MASTER) == 0) + cpu_relax(); +} + void __init ixdp2800_pci_preinit(void) { printk("ixdp2x00_pci_preinit called\n"); - *IXP2000_PCI_ADDR_EXT = 0x0000e000; + *IXP2000_PCI_ADDR_EXT = 0x0001e000; + + if (!ixdp2x00_master_npu()) + ixdp2800_slave_disable_pci_master(); - *IXP2000_PCI_DRAM_BASE_ADDR_MASK = (0x40000000 - 1) & ~0xfffff; *IXP2000_PCI_SRAM_BASE_ADDR_MASK = (0x2000000 - 1) & ~0x3ffff; + *IXP2000_PCI_DRAM_BASE_ADDR_MASK = (0x40000000 - 1) & ~0xfffff; ixp2000_pci_preinit(); + + if (ixdp2x00_master_npu()) { + /* + * Wait until the slave set its SRAM/SDRAM BAR sizes + * correctly before we proceed to scan and enumerate + * the bus. + */ + ixdp2800_master_wait_for_slave(); + + /* + * We configure the SDRAM BARs by hand because they + * are 1G and fall outside of the regular allocated + * PCI address space. + */ + *IXP2000_PCI_SDRAM_BAR = 0x00000008; + } else { + /* + * Wait for the master to complete scanning the bus + * and assigning resources before we proceed to scan + * the bus ourselves. Set pci=firmware to honor the + * master's resource assignment. + */ + ixdp2800_slave_wait_for_master_enable(); + pcibios_setup("firmware"); + } } -int ixdp2800_pci_setup(int nr, struct pci_sys_data *sys) +/* + * We assign the SDRAM BARs for the two IXP2800 CPUs by hand, outside + * of the regular PCI window, because there's only 512M of outbound PCI + * memory window on each IXP, while we need 1G for each of the BARs. + */ +static void __devinit ixp2800_pci_fixup(struct pci_dev *dev) +{ + if (machine_is_ixdp2800()) { + dev->resource[2].start = 0; + dev->resource[2].end = 0; + dev->resource[2].flags = 0; + } +} +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IXP2800, ixp2800_pci_fixup); + +static int __init ixdp2800_pci_setup(int nr, struct pci_sys_data *sys) { sys->mem_offset = 0x00000000; @@ -129,22 +212,47 @@ static int __init ixdp2800_pci_map_irq(struct pci_dev *dev, u8 slot, u8 pin) } else return IRQ_IXP2000_PCIB; /* Slave NIC interrupt */ } -static void ixdp2800_pci_postinit(void) +static void __init ixdp2800_master_enable_slave(void) { - struct pci_dev *dev; + volatile u32 *addr; - if (ixdp2x00_master_npu()) { - dev = pci_find_slot(1, IXDP2800_SLAVE_ENET_DEVFN); - pci_remove_bus_device(dev); - } else { - dev = pci_find_slot(1, IXDP2800_MASTER_ENET_DEVFN); - pci_remove_bus_device(dev); + printk(KERN_INFO "IXDP2800: enabling slave NPU\n"); + + addr = (volatile u32 *)ixp2000_pci_config_addr(0, + IXDP2X00_SLAVE_NPU_DEVFN, + PCI_COMMAND); + + *addr |= PCI_COMMAND_MASTER; +} +static void __init ixdp2800_master_wait_for_slave_bus_scan(void) +{ + volatile u32 *addr; + + printk(KERN_INFO "IXDP2800: waiting for slave to finish bus scan\n"); + + addr = (volatile u32 *)ixp2000_pci_config_addr(0, + IXDP2X00_SLAVE_NPU_DEVFN, + PCI_COMMAND); + while ((*addr & PCI_COMMAND_MEMORY) == 0) + cpu_relax(); +} + +static void __init ixdp2800_slave_signal_bus_scan_completion(void) +{ + printk(KERN_INFO "IXDP2800: bus scan done, signaling master\n"); + *IXP2000_PCI_CMDSTAT |= PCI_COMMAND_MEMORY; +} + +static void __init ixdp2800_pci_postinit(void) +{ + if (!ixdp2x00_master_npu()) { ixdp2x00_slave_pci_postinit(); + ixdp2800_slave_signal_bus_scan_completion(); } } -struct hw_pci ixdp2800_pci __initdata = { +struct __initdata hw_pci ixdp2800_pci __initdata = { .nr_controllers = 1, .setup = ixdp2800_pci_setup, .preinit = ixdp2800_pci_preinit, @@ -155,8 +263,21 @@ struct hw_pci ixdp2800_pci __initdata = { int __init ixdp2800_pci_init(void) { - if (machine_is_ixdp2800()) + if (machine_is_ixdp2800()) { + struct pci_dev *dev; + pci_common_init(&ixdp2800_pci); + if (ixdp2x00_master_npu()) { + dev = pci_find_slot(1, IXDP2800_SLAVE_ENET_DEVFN); + pci_remove_bus_device(dev); + + ixdp2800_master_enable_slave(); + ixdp2800_master_wait_for_slave_bus_scan(); + } else { + dev = pci_find_slot(1, IXDP2800_MASTER_ENET_DEVFN); + pci_remove_bus_device(dev); + } + } return 0; } -- cgit v1.2.3-70-g09d2 From d5aa207e46ff7ee838683a7d95ecf46fe42a9a56 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 30 Apr 2005 12:19:28 +0100 Subject: [PATCH] ARM: RTC: allow driver methods to return error Allow RTC drivers to return error codes from their read_time or read_alarm methods. Signed-off-by: Russell King --- arch/arm/common/rtctime.c | 29 +++++++++++++++-------------- arch/arm/mach-integrator/time.c | 17 ++++++++++++----- drivers/char/s3c2410-rtc.c | 8 ++++++-- include/asm-arm/rtc.h | 4 ++-- 4 files changed, 35 insertions(+), 23 deletions(-) (limited to 'arch') diff --git a/arch/arm/common/rtctime.c b/arch/arm/common/rtctime.c index c397e71f938..72b03f201eb 100644 --- a/arch/arm/common/rtctime.c +++ b/arch/arm/common/rtctime.c @@ -141,10 +141,10 @@ void rtc_next_alarm_time(struct rtc_time *next, struct rtc_time *now, struct rtc next->tm_sec = alrm->tm_sec; } -static inline void rtc_read_time(struct rtc_ops *ops, struct rtc_time *tm) +static inline int rtc_read_time(struct rtc_ops *ops, struct rtc_time *tm) { memset(tm, 0, sizeof(struct rtc_time)); - ops->read_time(tm); + return ops->read_time(tm); } static inline int rtc_set_time(struct rtc_ops *ops, struct rtc_time *tm) @@ -163,8 +163,7 @@ static inline int rtc_read_alarm(struct rtc_ops *ops, struct rtc_wkalrm *alrm) int ret = -EINVAL; if (ops->read_alarm) { memset(alrm, 0, sizeof(struct rtc_wkalrm)); - ops->read_alarm(alrm); - ret = 0; + ret = ops->read_alarm(alrm); } return ret; } @@ -283,7 +282,9 @@ static int rtc_ioctl(struct inode *inode, struct file *file, unsigned int cmd, break; case RTC_RD_TIME: - rtc_read_time(ops, &tm); + ret = rtc_read_time(ops, &tm); + if (ret) + break; ret = copy_to_user(uarg, &tm, sizeof(tm)); if (ret) ret = -EFAULT; @@ -424,15 +425,15 @@ static int rtc_read_proc(char *page, char **start, off_t off, int count, int *eo struct rtc_time tm; char *p = page; - rtc_read_time(ops, &tm); - - p += sprintf(p, - "rtc_time\t: %02d:%02d:%02d\n" - "rtc_date\t: %04d-%02d-%02d\n" - "rtc_epoch\t: %04lu\n", - tm.tm_hour, tm.tm_min, tm.tm_sec, - tm.tm_year + 1900, tm.tm_mon + 1, tm.tm_mday, - rtc_epoch); + if (rtc_read_time(ops, &tm) == 0) { + p += sprintf(p, + "rtc_time\t: %02d:%02d:%02d\n" + "rtc_date\t: %04d-%02d-%02d\n" + "rtc_epoch\t: %04lu\n", + tm.tm_hour, tm.tm_min, tm.tm_sec, + tm.tm_year + 1900, tm.tm_mon + 1, tm.tm_mday, + rtc_epoch); + } if (rtc_read_alarm(ops, &alrm) == 0) { p += sprintf(p, "alrm_time\t: "); diff --git a/arch/arm/mach-integrator/time.c b/arch/arm/mach-integrator/time.c index 20729de2af2..1a844ca139e 100644 --- a/arch/arm/mach-integrator/time.c +++ b/arch/arm/mach-integrator/time.c @@ -40,25 +40,32 @@ static int integrator_set_rtc(void) return 1; } -static void rtc_read_alarm(struct rtc_wkalrm *alrm) +static int rtc_read_alarm(struct rtc_wkalrm *alrm) { rtc_time_to_tm(readl(rtc_base + RTC_MR), &alrm->time); + return 0; } -static int rtc_set_alarm(struct rtc_wkalrm *alrm) +static inline int rtc_set_alarm(struct rtc_wkalrm *alrm) { unsigned long time; int ret; - ret = rtc_tm_to_time(&alrm->time, &time); + /* + * At the moment, we can only deal with non-wildcarded alarm times. + */ + ret = rtc_valid_tm(&alrm->time); + if (ret == 0) + ret = rtc_tm_to_time(&alrm->time, &time); if (ret == 0) writel(time, rtc_base + RTC_MR); return ret; } -static void rtc_read_time(struct rtc_time *tm) +static int rtc_read_time(struct rtc_time *tm) { rtc_time_to_tm(readl(rtc_base + RTC_DR), tm); + return 0; } /* @@ -69,7 +76,7 @@ static void rtc_read_time(struct rtc_time *tm) * edge of the 1Hz clock, we must write the time one second * in advance. */ -static int rtc_set_time(struct rtc_time *tm) +static inline int rtc_set_time(struct rtc_time *tm) { unsigned long time; int ret; diff --git a/drivers/char/s3c2410-rtc.c b/drivers/char/s3c2410-rtc.c index 8e61be34a1d..ed867db550a 100644 --- a/drivers/char/s3c2410-rtc.c +++ b/drivers/char/s3c2410-rtc.c @@ -116,7 +116,7 @@ static void s3c2410_rtc_setfreq(int freq) /* Time read/write */ -static void s3c2410_rtc_gettime(struct rtc_time *rtc_tm) +static int s3c2410_rtc_gettime(struct rtc_time *rtc_tm) { unsigned int have_retried = 0; @@ -151,6 +151,8 @@ static void s3c2410_rtc_gettime(struct rtc_time *rtc_tm) rtc_tm->tm_year += 100; rtc_tm->tm_mon -= 1; + + return 0; } @@ -171,7 +173,7 @@ static int s3c2410_rtc_settime(struct rtc_time *tm) return 0; } -static void s3c2410_rtc_getalarm(struct rtc_wkalrm *alrm) +static int s3c2410_rtc_getalarm(struct rtc_wkalrm *alrm) { struct rtc_time *alm_tm = &alrm->time; unsigned int alm_en; @@ -231,6 +233,8 @@ static void s3c2410_rtc_getalarm(struct rtc_wkalrm *alrm) } /* todo - set alrm->enabled ? */ + + return 0; } static int s3c2410_rtc_setalarm(struct rtc_wkalrm *alrm) diff --git a/include/asm-arm/rtc.h b/include/asm-arm/rtc.h index aa7e16b2e22..370dfe77589 100644 --- a/include/asm-arm/rtc.h +++ b/include/asm-arm/rtc.h @@ -18,9 +18,9 @@ struct rtc_ops { void (*release)(void); int (*ioctl)(unsigned int, unsigned long); - void (*read_time)(struct rtc_time *); + int (*read_time)(struct rtc_time *); int (*set_time)(struct rtc_time *); - void (*read_alarm)(struct rtc_wkalrm *); + int (*read_alarm)(struct rtc_wkalrm *); int (*set_alarm)(struct rtc_wkalrm *); int (*proc)(char *buf); }; -- cgit v1.2.3-70-g09d2 From bb9bffcbef6166cf03385fbcde97c27bc1a5e689 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 30 Apr 2005 13:26:06 +0100 Subject: [PATCH] ARM: PXA I2C: add platform device Add the PXA I2C platform device. Signed-off-by: Russell King --- arch/arm/mach-pxa/generic.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) (limited to 'arch') diff --git a/arch/arm/mach-pxa/generic.c b/arch/arm/mach-pxa/generic.c index b1575b8dc1c..a45aaa115a7 100644 --- a/arch/arm/mach-pxa/generic.c +++ b/arch/arm/mach-pxa/generic.c @@ -220,6 +220,30 @@ static struct platform_device stuart_device = { .id = 2, }; +static struct resource i2c_resources[] = { + { + .start = 0x40301680, + .end = 0x403016a3, + .flags = IORESOURCE_MEM, + }, { + .start = IRQ_I2C, + .end = IRQ_I2C, + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device i2c_device = { + .name = "pxa2xx-i2c", + .id = 0, + .resource = i2c_resources, + .num_resources = ARRAY_SIZE(i2c_resources), +}; + +void __init pxa_set_i2c_info(struct i2c_pxa_platform_data *info) +{ + i2c_device.dev.platform_data = info; +} + static struct platform_device *devices[] __initdata = { &pxamci_device, &udc_device, @@ -227,6 +251,7 @@ static struct platform_device *devices[] __initdata = { &ffuart_device, &btuart_device, &stuart_device, + &i2c_device, }; static int __init pxa_init(void) -- cgit v1.2.3-70-g09d2 From 9747dd6fa98f2983f4dd09cd6dad1fa3d2a4c5f2 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Sat, 30 Apr 2005 10:01:40 -0700 Subject: [PATCH] ppc64: fix 32-bit signal frame back link When the kernel creates a signal frame on the user stack, it puts the old stack pointer value at the beginning so that the signal frame is linked into the chain of stack frames like any other frame. Unfortunately, for 32-bit processes we are writing the old stack pointer as a 64-bit value rather than a 32-bit value, and the process sees that as a null pointer, since it only looks at the first 32 bits, which are zero since ppc is bigendian and the stack pointer is below 4GB. This bug is in SLES9 and RHEL4 too, hence the ccs. This patch fixes the bug by making the signal code write the old stack pointer as a u32 instead of an unsigned long. Signed-off-by: Paul Mackerras Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/signal32.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/kernel/signal32.c b/arch/ppc64/kernel/signal32.c index b0e167db6af..3c2fa5c284c 100644 --- a/arch/ppc64/kernel/signal32.c +++ b/arch/ppc64/kernel/signal32.c @@ -657,7 +657,7 @@ static int handle_rt_signal32(unsigned long sig, struct k_sigaction *ka, /* Save user registers on the stack */ frame = &rt_sf->uc.uc_mcontext; - if (put_user(regs->gpr[1], (unsigned long __user *)newsp)) + if (put_user(regs->gpr[1], (u32 __user *)newsp)) goto badframe; if (vdso32_rt_sigtramp && current->thread.vdso_base) { @@ -842,7 +842,7 @@ static int handle_signal32(unsigned long sig, struct k_sigaction *ka, regs->link = (unsigned long) frame->mctx.tramp; } - if (put_user(regs->gpr[1], (unsigned long __user *)newsp)) + if (put_user(regs->gpr[1], (u32 __user *)newsp)) goto badframe; regs->gpr[1] = (unsigned long) newsp; regs->gpr[3] = sig; -- cgit v1.2.3-70-g09d2 From 4774e2260cf25c54f2188dd0407676e3af6f1f23 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 30 Apr 2005 23:32:38 +0100 Subject: [PATCH] ARM: IntegratorCP: Fix CLCD MUX selection values The documentation on these values seems to be rather wrong. These values have been determined by mere trial and error. Signed-off-by: Russell King --- arch/arm/mach-integrator/integrator_cp.c | 17 ++++++++++++++++- include/asm-arm/arch-integrator/cm.h | 6 +++--- 2 files changed, 19 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-integrator/integrator_cp.c b/arch/arm/mach-integrator/integrator_cp.c index 68e15c36e33..3b948e8c275 100644 --- a/arch/arm/mach-integrator/integrator_cp.c +++ b/arch/arm/mach-integrator/integrator_cp.c @@ -420,7 +420,22 @@ static struct clcd_panel vga = { */ static void cp_clcd_enable(struct clcd_fb *fb) { - cm_control(CM_CTRL_LCDMUXSEL_MASK, CM_CTRL_LCDMUXSEL_VGA); + u32 val; + + if (fb->fb.var.bits_per_pixel <= 8) + val = CM_CTRL_LCDMUXSEL_VGA_8421BPP; + else if (fb->fb.var.bits_per_pixel <= 16) + val = CM_CTRL_LCDMUXSEL_VGA_16BPP; + else + val = 0; /* no idea for this, don't trust the docs */ + + cm_control(CM_CTRL_LCDMUXSEL_MASK| + CM_CTRL_LCDEN0| + CM_CTRL_LCDEN1| + CM_CTRL_STATIC1| + CM_CTRL_STATIC2| + CM_CTRL_STATIC| + CM_CTRL_n24BITEN, val); } static unsigned long framesize = SZ_1M; diff --git a/include/asm-arm/arch-integrator/cm.h b/include/asm-arm/arch-integrator/cm.h index d31c1a71f78..1ab353e2359 100644 --- a/include/asm-arm/arch-integrator/cm.h +++ b/include/asm-arm/arch-integrator/cm.h @@ -24,9 +24,9 @@ void cm_control(u32, u32); #define CM_CTRL_LCDBIASDN (1 << 10) #define CM_CTRL_LCDMUXSEL_MASK (7 << 11) #define CM_CTRL_LCDMUXSEL_GENLCD (1 << 11) -#define CM_CTRL_LCDMUXSEL_SHARPLCD1 (3 << 11) -#define CM_CTRL_LCDMUXSEL_SHARPLCD2 (4 << 11) -#define CM_CTRL_LCDMUXSEL_VGA (7 << 11) +#define CM_CTRL_LCDMUXSEL_VGA_16BPP (2 << 11) +#define CM_CTRL_LCDMUXSEL_SHARPLCD (3 << 11) +#define CM_CTRL_LCDMUXSEL_VGA_8421BPP (4 << 11) #define CM_CTRL_LCDEN0 (1 << 14) #define CM_CTRL_LCDEN1 (1 << 15) #define CM_CTRL_STATIC1 (1 << 16) -- cgit v1.2.3-70-g09d2 From 2cacb3da620a4a93f3a77e1d2c8c06bb3c74bcb0 Mon Sep 17 00:00:00 2001 From: Sam Ravnborg Date: Sat, 30 Apr 2005 16:51:42 -0700 Subject: [PATCH] kbuild/i386: re-introduce dependency on vmlinux for install target, and add kernel_install Removing the dependency on vmlinux for the install target raised a few complaints, so instead a new target i added: kernel_install. kernel_install will install the kernel just like the ordinary install target. The only difference is that install has a dependency on vmlinux, kernel_install does not. Therefore kernel_install is the best choice when accessing the kernel over a NFS mount or as another user. kernel_install is similar to modules_install in the fact that neither does a full kernel compile before performing the install. In this way they are good for root use. Also added back the dependency on vmlinux for the install target so peoples scripts are no longer broken. Signed-off-by: Sam Ravnborg Signed-off-by: Linus Torvalds --- arch/i386/Makefile | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/i386/Makefile b/arch/i386/Makefile index 314c7146e9b..04783ceb050 100644 --- a/arch/i386/Makefile +++ b/arch/i386/Makefile @@ -123,7 +123,7 @@ AFLAGS += $(mflags-y) boot := arch/i386/boot .PHONY: zImage bzImage compressed zlilo bzlilo \ - zdisk bzdisk fdimage fdimage144 fdimage288 install + zdisk bzdisk fdimage fdimage144 fdimage288 install kernel_install all: bzImage @@ -145,8 +145,9 @@ zdisk bzdisk: vmlinux fdimage fdimage144 fdimage288: vmlinux $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) $@ -install: - $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) $@ +install: vmlinux +install kernel_install: + $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(KBUILD_IMAGE) install prepare: include/asm-$(ARCH)/asm_offsets.h CLEAN_FILES += include/asm-$(ARCH)/asm_offsets.h -- cgit v1.2.3-70-g09d2 From b3d9ae4b98f6d28481d4d4b768d860a2cfb5805d Mon Sep 17 00:00:00 2001 From: Sam Ravnborg Date: Sat, 30 Apr 2005 16:51:42 -0700 Subject: [PATCH] kbuild/ppc: tell when uimage was not built Tom Rini said: Note that there is still a trivial'ish change to make. When mkimage doesn't exist on the host we should say "uImage not made" or something similar. So I did like Tom asked. Signed-off-by: Sam Ravnborg Signed-off-by: Linus Torvalds --- arch/ppc/boot/images/Makefile | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ppc/boot/images/Makefile b/arch/ppc/boot/images/Makefile index 774de8e2387..f850fb0fb51 100644 --- a/arch/ppc/boot/images/Makefile +++ b/arch/ppc/boot/images/Makefile @@ -20,8 +20,9 @@ quiet_cmd_uimage = UIMAGE $@ targets += uImage $(obj)/uImage: $(obj)/vmlinux.gz + $(Q)rm -f $@ $(call if_changed,uimage) - @echo ' Image $@ is ready' + @echo ' Image: $@' $(if $(wildcard $@),'is ready','not made') # Files generated that shall be removed upon make clean clean-files := sImage vmapus vmlinux* miboot* zImage* uImage -- cgit v1.2.3-70-g09d2 From f1c55dea0bb2df94aa2b01b0871cb02f2e206676 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Sun, 1 May 2005 08:58:40 -0700 Subject: [PATCH] ppc32: Fix errata for some G3 CPUs Some G3 CPUs can crash in funny way if a store from an FPU register instruction is executed on a register that has never been initialized since power on. This patch fixes it by making sure all FP registers have been properly initialized at kernel boot and when waking from sleep. It also makes the code that decides wether HID0_BTIC and HID0_DPM are allowed on a given CPU smarter (it can actually _clear_ them now if they are not allowed instead of just setting them when they are allowed in case the firmware got them wrong) Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/kernel/cpu_setup_6xx.S | 42 +++++++++++++++++++++++++++++++++++------ arch/ppc/platforms/pmac_sleep.S | 4 ++++ 2 files changed, 40 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/ppc/kernel/cpu_setup_6xx.S b/arch/ppc/kernel/cpu_setup_6xx.S index 74f781b486a..468721d9ebd 100644 --- a/arch/ppc/kernel/cpu_setup_6xx.S +++ b/arch/ppc/kernel/cpu_setup_6xx.S @@ -30,12 +30,14 @@ _GLOBAL(__setup_cpu_604) blr _GLOBAL(__setup_cpu_750) mflr r4 + bl __init_fpu_registers bl setup_common_caches bl setup_750_7400_hid0 mtlr r4 blr _GLOBAL(__setup_cpu_750cx) mflr r4 + bl __init_fpu_registers bl setup_common_caches bl setup_750_7400_hid0 bl setup_750cx @@ -43,6 +45,7 @@ _GLOBAL(__setup_cpu_750cx) blr _GLOBAL(__setup_cpu_750fx) mflr r4 + bl __init_fpu_registers bl setup_common_caches bl setup_750_7400_hid0 bl setup_750fx @@ -50,6 +53,7 @@ _GLOBAL(__setup_cpu_750fx) blr _GLOBAL(__setup_cpu_7400) mflr r4 + bl __init_fpu_registers bl setup_7400_workarounds bl setup_common_caches bl setup_750_7400_hid0 @@ -57,6 +61,7 @@ _GLOBAL(__setup_cpu_7400) blr _GLOBAL(__setup_cpu_7410) mflr r4 + bl __init_fpu_registers bl setup_7410_workarounds bl setup_common_caches bl setup_750_7400_hid0 @@ -80,7 +85,7 @@ setup_common_caches: bne 1f /* don't invalidate the D-cache */ ori r8,r8,HID0_DCI /* unless it wasn't enabled */ 1: sync - mtspr SPRN_HID0,r8 /* enable and invalidate caches */ + mtspr SPRN_HID0,r8 /* enable and invalidate caches */ sync mtspr SPRN_HID0,r11 /* enable caches */ sync @@ -152,9 +157,13 @@ setup_7410_workarounds: setup_750_7400_hid0: mfspr r11,SPRN_HID0 ori r11,r11,HID0_SGE | HID0_ABE | HID0_BHTE | HID0_BTIC + oris r11,r11,HID0_DPM@h BEGIN_FTR_SECTION - oris r11,r11,HID0_DPM@h /* enable dynamic power mgmt */ -END_FTR_SECTION_IFCLR(CPU_FTR_NO_DPM) + xori r11,r11,HID0_BTIC +END_FTR_SECTION_IFSET(CPU_FTR_NO_BTIC) +BEGIN_FTR_SECTION + xoris r11,r11,HID0_DPM@h /* disable dynamic power mgmt */ +END_FTR_SECTION_IFSET(CPU_FTR_NO_DPM) li r3,HID0_SPD andc r11,r11,r3 /* clear SPD: enable speculative */ li r3,0 @@ -218,13 +227,15 @@ setup_745x_specifics: /* All of the bits we have to set..... */ - ori r11,r11,HID0_SGE | HID0_FOLD | HID0_BHTE | HID0_LRSTK | HID0_BTIC + ori r11,r11,HID0_SGE | HID0_FOLD | HID0_BHTE + ori r11,r11,HID0_LRSTK | HID0_BTIC + oris r11,r11,HID0_DPM@h BEGIN_FTR_SECTION xori r11,r11,HID0_BTIC END_FTR_SECTION_IFSET(CPU_FTR_NO_BTIC) BEGIN_FTR_SECTION - oris r11,r11,HID0_DPM@h /* enable dynamic power mgmt */ -END_FTR_SECTION_IFCLR(CPU_FTR_NO_DPM) + xoris r11,r11,HID0_DPM@h /* disable dynamic power mgmt */ +END_FTR_SECTION_IFSET(CPU_FTR_NO_DPM) /* All of the bits we have to clear.... */ @@ -248,6 +259,25 @@ END_FTR_SECTION_IFCLR(CPU_FTR_NO_DPM) isync blr +/* + * Initialize the FPU registers. This is needed to work around an errata + * in some 750 cpus where using a not yet initialized FPU register after + * power on reset may hang the CPU + */ +_GLOBAL(__init_fpu_registers) + mfmsr r10 + ori r11,r10,MSR_FP + mtmsr r11 + isync + addis r9,r3,empty_zero_page@ha + addi r9,r9,empty_zero_page@l + REST_32FPRS(0,r9) + sync + mtmsr r10 + isync + blr + + /* Definitions for the table use to save CPU states */ #define CS_HID0 0 #define CS_HID1 4 diff --git a/arch/ppc/platforms/pmac_sleep.S b/arch/ppc/platforms/pmac_sleep.S index 3139b6766ad..f459ade1bd6 100644 --- a/arch/ppc/platforms/pmac_sleep.S +++ b/arch/ppc/platforms/pmac_sleep.S @@ -267,6 +267,10 @@ grackle_wake_up: /* Restore various CPU config stuffs */ bl __restore_cpu_setup + /* Make sure all FPRs have been initialized */ + bl reloc_offset + bl __init_fpu_registers + /* Invalidate & enable L1 cache, we don't care about * whatever the ROM may have tried to write to memory */ -- cgit v1.2.3-70-g09d2 From 443a848cd30eb5bb5c1038e6371d83404775dcfc Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Sun, 1 May 2005 08:58:40 -0700 Subject: [PATCH] ppc32: refactor FPU exception handling Moved common FPU exception handling code out of head.S so it can be used by several of the sub-architectures that might of a full PowerPC FPU. Also, uses new CONFIG_PPC_FPU define to fix alignment exception handling for floating point load/store instructions to only occur if we have a hardware FPU. Signed-off-by: Jason McMullan Signed-off-by: Kumar Gala Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/Kconfig | 4 + arch/ppc/Makefile | 1 + arch/ppc/kernel/Makefile | 1 + arch/ppc/kernel/align.c | 8 ++ arch/ppc/kernel/entry.S | 59 ++++++++++++++ arch/ppc/kernel/fpu.S | 133 ++++++++++++++++++++++++++++++++ arch/ppc/kernel/head.S | 163 --------------------------------------- arch/ppc/kernel/head_44x.S | 6 ++ arch/ppc/kernel/head_booke.h | 7 ++ arch/ppc/kernel/head_fsl_booke.S | 8 +- arch/ppc/kernel/misc.S | 12 +-- arch/ppc/kernel/traps.c | 2 +- include/asm-ppc/reg_booke.h | 1 + 13 files changed, 229 insertions(+), 176 deletions(-) create mode 100644 arch/ppc/kernel/fpu.S (limited to 'arch') diff --git a/arch/ppc/Kconfig b/arch/ppc/Kconfig index 74aa1e92a39..c3d941345e3 100644 --- a/arch/ppc/Kconfig +++ b/arch/ppc/Kconfig @@ -53,6 +53,7 @@ choice config 6xx bool "6xx/7xx/74xx/52xx/82xx/83xx" + select PPC_FPU help There are four types of PowerPC chips supported. The more common types (601, 603, 604, 740, 750, 7400), the Motorola embedded @@ -86,6 +87,9 @@ config E500 endchoice +config PPC_FPU + bool + config BOOKE bool depends on E500 diff --git a/arch/ppc/Makefile b/arch/ppc/Makefile index 73cbdda5b59..0432a25b473 100644 --- a/arch/ppc/Makefile +++ b/arch/ppc/Makefile @@ -53,6 +53,7 @@ head-$(CONFIG_FSL_BOOKE) := arch/ppc/kernel/head_fsl_booke.o head-$(CONFIG_6xx) += arch/ppc/kernel/idle_6xx.o head-$(CONFIG_POWER4) += arch/ppc/kernel/idle_power4.o +head-$(CONFIG_PPC_FPU) += arch/ppc/kernel/fpu.o core-y += arch/ppc/kernel/ arch/ppc/platforms/ \ arch/ppc/mm/ arch/ppc/lib/ arch/ppc/syslib/ diff --git a/arch/ppc/kernel/Makefile b/arch/ppc/kernel/Makefile index 86bc878cb3e..b284451802c 100644 --- a/arch/ppc/kernel/Makefile +++ b/arch/ppc/kernel/Makefile @@ -9,6 +9,7 @@ extra-$(CONFIG_FSL_BOOKE) := head_fsl_booke.o extra-$(CONFIG_8xx) := head_8xx.o extra-$(CONFIG_6xx) += idle_6xx.o extra-$(CONFIG_POWER4) += idle_power4.o +extra-$(CONFIG_PPC_FPU) += fpu.o extra-y += vmlinux.lds obj-y := entry.o traps.o irq.o idle.o time.o misc.o \ diff --git a/arch/ppc/kernel/align.c b/arch/ppc/kernel/align.c index 79c92947503..40d356c66c4 100644 --- a/arch/ppc/kernel/align.c +++ b/arch/ppc/kernel/align.c @@ -368,16 +368,24 @@ fix_alignment(struct pt_regs *regs) /* Single-precision FP load and store require conversions... */ case LD+F+S: +#ifdef CONFIG_PPC_FPU preempt_disable(); enable_kernel_fp(); cvt_fd(&data.f, &data.d, ¤t->thread.fpscr); preempt_enable(); +#else + return 0; +#endif break; case ST+F+S: +#ifdef CONFIG_PPC_FPU preempt_disable(); enable_kernel_fp(); cvt_df(&data.d, &data.f, ¤t->thread.fpscr); preempt_enable(); +#else + return 0; +#endif break; } diff --git a/arch/ppc/kernel/entry.S b/arch/ppc/kernel/entry.S index 035217d6c0f..5f075dbc4ee 100644 --- a/arch/ppc/kernel/entry.S +++ b/arch/ppc/kernel/entry.S @@ -563,6 +563,65 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC) addi r1,r1,INT_FRAME_SIZE blr + .globl fast_exception_return +fast_exception_return: +#if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE)) + andi. r10,r9,MSR_RI /* check for recoverable interrupt */ + beq 1f /* if not, we've got problems */ +#endif + +2: REST_4GPRS(3, r11) + lwz r10,_CCR(r11) + REST_GPR(1, r11) + mtcr r10 + lwz r10,_LINK(r11) + mtlr r10 + REST_GPR(10, r11) + mtspr SPRN_SRR1,r9 + mtspr SPRN_SRR0,r12 + REST_GPR(9, r11) + REST_GPR(12, r11) + lwz r11,GPR11(r11) + SYNC + RFI + +#if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE)) +/* check if the exception happened in a restartable section */ +1: lis r3,exc_exit_restart_end@ha + addi r3,r3,exc_exit_restart_end@l + cmplw r12,r3 + bge 3f + lis r4,exc_exit_restart@ha + addi r4,r4,exc_exit_restart@l + cmplw r12,r4 + blt 3f + lis r3,fee_restarts@ha + tophys(r3,r3) + lwz r5,fee_restarts@l(r3) + addi r5,r5,1 + stw r5,fee_restarts@l(r3) + mr r12,r4 /* restart at exc_exit_restart */ + b 2b + + .comm fee_restarts,4 + +/* aargh, a nonrecoverable interrupt, panic */ +/* aargh, we don't know which trap this is */ +/* but the 601 doesn't implement the RI bit, so assume it's OK */ +3: +BEGIN_FTR_SECTION + b 2b +END_FTR_SECTION_IFSET(CPU_FTR_601) + li r10,-1 + stw r10,TRAP(r11) + addi r3,r1,STACK_FRAME_OVERHEAD + lis r10,MSR_KERNEL@h + ori r10,r10,MSR_KERNEL@l + bl transfer_to_handler_full + .long nonrecoverable_exception + .long ret_from_except +#endif + .globl sigreturn_exit sigreturn_exit: subi r1,r3,STACK_FRAME_OVERHEAD diff --git a/arch/ppc/kernel/fpu.S b/arch/ppc/kernel/fpu.S new file mode 100644 index 00000000000..6189b26f640 --- /dev/null +++ b/arch/ppc/kernel/fpu.S @@ -0,0 +1,133 @@ +/* + * FPU support code, moved here from head.S so that it can be used + * by chips which use other head-whatever.S files. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* + * This task wants to use the FPU now. + * On UP, disable FP for the task which had the FPU previously, + * and save its floating-point registers in its thread_struct. + * Load up this task's FP registers from its thread_struct, + * enable the FPU for the current task and return to the task. + */ + .globl load_up_fpu +load_up_fpu: + mfmsr r5 + ori r5,r5,MSR_FP +#ifdef CONFIG_PPC64BRIDGE + clrldi r5,r5,1 /* turn off 64-bit mode */ +#endif /* CONFIG_PPC64BRIDGE */ + SYNC + MTMSRD(r5) /* enable use of fpu now */ + isync +/* + * For SMP, we don't do lazy FPU switching because it just gets too + * horrendously complex, especially when a task switches from one CPU + * to another. Instead we call giveup_fpu in switch_to. + */ +#ifndef CONFIG_SMP + tophys(r6,0) /* get __pa constant */ + addis r3,r6,last_task_used_math@ha + lwz r4,last_task_used_math@l(r3) + cmpwi 0,r4,0 + beq 1f + add r4,r4,r6 + addi r4,r4,THREAD /* want last_task_used_math->thread */ + SAVE_32FPRS(0, r4) + mffs fr0 + stfd fr0,THREAD_FPSCR-4(r4) + lwz r5,PT_REGS(r4) + add r5,r5,r6 + lwz r4,_MSR-STACK_FRAME_OVERHEAD(r5) + li r10,MSR_FP|MSR_FE0|MSR_FE1 + andc r4,r4,r10 /* disable FP for previous task */ + stw r4,_MSR-STACK_FRAME_OVERHEAD(r5) +1: +#endif /* CONFIG_SMP */ + /* enable use of FP after return */ + mfspr r5,SPRN_SPRG3 /* current task's THREAD (phys) */ + lwz r4,THREAD_FPEXC_MODE(r5) + ori r9,r9,MSR_FP /* enable FP for current */ + or r9,r9,r4 + lfd fr0,THREAD_FPSCR-4(r5) + mtfsf 0xff,fr0 + REST_32FPRS(0, r5) +#ifndef CONFIG_SMP + subi r4,r5,THREAD + sub r4,r4,r6 + stw r4,last_task_used_math@l(r3) +#endif /* CONFIG_SMP */ + /* restore registers and return */ + /* we haven't used ctr or xer or lr */ + b fast_exception_return + +/* + * FP unavailable trap from kernel - print a message, but let + * the task use FP in the kernel until it returns to user mode. + */ + .globl KernelFP +KernelFP: + lwz r3,_MSR(r1) + ori r3,r3,MSR_FP + stw r3,_MSR(r1) /* enable use of FP after return */ + lis r3,86f@h + ori r3,r3,86f@l + mr r4,r2 /* current */ + lwz r5,_NIP(r1) + bl printk + b ret_from_except +86: .string "floating point used in kernel (task=%p, pc=%x)\n" + .align 4,0 + +/* + * giveup_fpu(tsk) + * Disable FP for the task given as the argument, + * and save the floating-point registers in its thread_struct. + * Enables the FPU for use in the kernel on return. + */ + .globl giveup_fpu +giveup_fpu: + mfmsr r5 + ori r5,r5,MSR_FP + SYNC_601 + ISYNC_601 + MTMSRD(r5) /* enable use of fpu now */ + SYNC_601 + isync + cmpwi 0,r3,0 + beqlr- /* if no previous owner, done */ + addi r3,r3,THREAD /* want THREAD of task */ + lwz r5,PT_REGS(r3) + cmpwi 0,r5,0 + SAVE_32FPRS(0, r3) + mffs fr0 + stfd fr0,THREAD_FPSCR-4(r3) + beq 1f + lwz r4,_MSR-STACK_FRAME_OVERHEAD(r5) + li r3,MSR_FP|MSR_FE0|MSR_FE1 + andc r4,r4,r3 /* disable FP for previous task */ + stw r4,_MSR-STACK_FRAME_OVERHEAD(r5) +1: +#ifndef CONFIG_SMP + li r5,0 + lis r4,last_task_used_math@ha + stw r5,last_task_used_math@l(r4) +#endif /* CONFIG_SMP */ + blr diff --git a/arch/ppc/kernel/head.S b/arch/ppc/kernel/head.S index 1a89a71e0ac..a931d773715 100644 --- a/arch/ppc/kernel/head.S +++ b/arch/ppc/kernel/head.S @@ -775,133 +775,6 @@ InstructionSegment: EXC_XFER_STD(0x480, UnknownException) #endif /* CONFIG_PPC64BRIDGE */ -/* - * This task wants to use the FPU now. - * On UP, disable FP for the task which had the FPU previously, - * and save its floating-point registers in its thread_struct. - * Load up this task's FP registers from its thread_struct, - * enable the FPU for the current task and return to the task. - */ -load_up_fpu: - mfmsr r5 - ori r5,r5,MSR_FP -#ifdef CONFIG_PPC64BRIDGE - clrldi r5,r5,1 /* turn off 64-bit mode */ -#endif /* CONFIG_PPC64BRIDGE */ - SYNC - MTMSRD(r5) /* enable use of fpu now */ - isync -/* - * For SMP, we don't do lazy FPU switching because it just gets too - * horrendously complex, especially when a task switches from one CPU - * to another. Instead we call giveup_fpu in switch_to. - */ -#ifndef CONFIG_SMP - tophys(r6,0) /* get __pa constant */ - addis r3,r6,last_task_used_math@ha - lwz r4,last_task_used_math@l(r3) - cmpwi 0,r4,0 - beq 1f - add r4,r4,r6 - addi r4,r4,THREAD /* want last_task_used_math->thread */ - SAVE_32FPRS(0, r4) - mffs fr0 - stfd fr0,THREAD_FPSCR-4(r4) - lwz r5,PT_REGS(r4) - add r5,r5,r6 - lwz r4,_MSR-STACK_FRAME_OVERHEAD(r5) - li r10,MSR_FP|MSR_FE0|MSR_FE1 - andc r4,r4,r10 /* disable FP for previous task */ - stw r4,_MSR-STACK_FRAME_OVERHEAD(r5) -1: -#endif /* CONFIG_SMP */ - /* enable use of FP after return */ - mfspr r5,SPRN_SPRG3 /* current task's THREAD (phys) */ - lwz r4,THREAD_FPEXC_MODE(r5) - ori r9,r9,MSR_FP /* enable FP for current */ - or r9,r9,r4 - lfd fr0,THREAD_FPSCR-4(r5) - mtfsf 0xff,fr0 - REST_32FPRS(0, r5) -#ifndef CONFIG_SMP - subi r4,r5,THREAD - sub r4,r4,r6 - stw r4,last_task_used_math@l(r3) -#endif /* CONFIG_SMP */ - /* restore registers and return */ - /* we haven't used ctr or xer or lr */ - /* fall through to fast_exception_return */ - - .globl fast_exception_return -fast_exception_return: - andi. r10,r9,MSR_RI /* check for recoverable interrupt */ - beq 1f /* if not, we've got problems */ -2: REST_4GPRS(3, r11) - lwz r10,_CCR(r11) - REST_GPR(1, r11) - mtcr r10 - lwz r10,_LINK(r11) - mtlr r10 - REST_GPR(10, r11) - mtspr SPRN_SRR1,r9 - mtspr SPRN_SRR0,r12 - REST_GPR(9, r11) - REST_GPR(12, r11) - lwz r11,GPR11(r11) - SYNC - RFI - -/* check if the exception happened in a restartable section */ -1: lis r3,exc_exit_restart_end@ha - addi r3,r3,exc_exit_restart_end@l - cmplw r12,r3 - bge 3f - lis r4,exc_exit_restart@ha - addi r4,r4,exc_exit_restart@l - cmplw r12,r4 - blt 3f - lis r3,fee_restarts@ha - tophys(r3,r3) - lwz r5,fee_restarts@l(r3) - addi r5,r5,1 - stw r5,fee_restarts@l(r3) - mr r12,r4 /* restart at exc_exit_restart */ - b 2b - - .comm fee_restarts,4 - -/* aargh, a nonrecoverable interrupt, panic */ -/* aargh, we don't know which trap this is */ -/* but the 601 doesn't implement the RI bit, so assume it's OK */ -3: -BEGIN_FTR_SECTION - b 2b -END_FTR_SECTION_IFSET(CPU_FTR_601) - li r10,-1 - stw r10,TRAP(r11) - addi r3,r1,STACK_FRAME_OVERHEAD - li r10,MSR_KERNEL - bl transfer_to_handler_full - .long nonrecoverable_exception - .long ret_from_except - -/* - * FP unavailable trap from kernel - print a message, but let - * the task use FP in the kernel until it returns to user mode. - */ -KernelFP: - lwz r3,_MSR(r1) - ori r3,r3,MSR_FP - stw r3,_MSR(r1) /* enable use of FP after return */ - lis r3,86f@h - ori r3,r3,86f@l - mr r4,r2 /* current */ - lwz r5,_NIP(r1) - bl printk - b ret_from_except -86: .string "floating point used in kernel (task=%p, pc=%x)\n" - .align 4,0 - #ifdef CONFIG_ALTIVEC /* Note that the AltiVec support is closely modeled after the FP * support. Changes to one are likely to be applicable to the @@ -1015,42 +888,6 @@ giveup_altivec: blr #endif /* CONFIG_ALTIVEC */ -/* - * giveup_fpu(tsk) - * Disable FP for the task given as the argument, - * and save the floating-point registers in its thread_struct. - * Enables the FPU for use in the kernel on return. - */ - .globl giveup_fpu -giveup_fpu: - mfmsr r5 - ori r5,r5,MSR_FP - SYNC_601 - ISYNC_601 - MTMSRD(r5) /* enable use of fpu now */ - SYNC_601 - isync - cmpwi 0,r3,0 - beqlr- /* if no previous owner, done */ - addi r3,r3,THREAD /* want THREAD of task */ - lwz r5,PT_REGS(r3) - cmpwi 0,r5,0 - SAVE_32FPRS(0, r3) - mffs fr0 - stfd fr0,THREAD_FPSCR-4(r3) - beq 1f - lwz r4,_MSR-STACK_FRAME_OVERHEAD(r5) - li r3,MSR_FP|MSR_FE0|MSR_FE1 - andc r4,r4,r3 /* disable FP for previous task */ - stw r4,_MSR-STACK_FRAME_OVERHEAD(r5) -1: -#ifndef CONFIG_SMP - li r5,0 - lis r4,last_task_used_math@ha - stw r5,last_task_used_math@l(r4) -#endif /* CONFIG_SMP */ - blr - /* * This code is jumped to from the startup code to copy * the kernel image to physical address 0. diff --git a/arch/ppc/kernel/head_44x.S b/arch/ppc/kernel/head_44x.S index 9ed8165a3d6..9b6a8e51365 100644 --- a/arch/ppc/kernel/head_44x.S +++ b/arch/ppc/kernel/head_44x.S @@ -426,7 +426,11 @@ interrupt_base: PROGRAM_EXCEPTION /* Floating Point Unavailable Interrupt */ +#ifdef CONFIG_PPC_FPU + FP_UNAVAILABLE_EXCEPTION +#else EXCEPTION(0x2010, FloatingPointUnavailable, UnknownException, EXC_XFER_EE) +#endif /* System Call Interrupt */ START_EXCEPTION(SystemCall) @@ -686,8 +690,10 @@ _GLOBAL(giveup_altivec) * * The 44x core does not have an FPU. */ +#ifndef CONFIG_PPC_FPU _GLOBAL(giveup_fpu) blr +#endif /* * extern void abort(void) diff --git a/arch/ppc/kernel/head_booke.h b/arch/ppc/kernel/head_booke.h index 884dac916bc..f213d12eec0 100644 --- a/arch/ppc/kernel/head_booke.h +++ b/arch/ppc/kernel/head_booke.h @@ -337,4 +337,11 @@ label: addi r3,r1,STACK_FRAME_OVERHEAD; \ EXC_XFER_LITE(0x0900, timer_interrupt) +#define FP_UNAVAILABLE_EXCEPTION \ + START_EXCEPTION(FloatingPointUnavailable) \ + NORMAL_EXCEPTION_PROLOG; \ + bne load_up_fpu; /* if from user, just load it up */ \ + addi r3,r1,STACK_FRAME_OVERHEAD; \ + EXC_XFER_EE_LITE(0x800, KernelFP) + #endif /* __HEAD_BOOKE_H__ */ diff --git a/arch/ppc/kernel/head_fsl_booke.S b/arch/ppc/kernel/head_fsl_booke.S index d64bf61d2b1..f22ddce3613 100644 --- a/arch/ppc/kernel/head_fsl_booke.S +++ b/arch/ppc/kernel/head_fsl_booke.S @@ -504,7 +504,11 @@ interrupt_base: PROGRAM_EXCEPTION /* Floating Point Unavailable Interrupt */ +#ifdef CONFIG_PPC_FPU + FP_UNAVAILABLE_EXCEPTION +#else EXCEPTION(0x0800, FloatingPointUnavailable, UnknownException, EXC_XFER_EE) +#endif /* System Call Interrupt */ START_EXCEPTION(SystemCall) @@ -916,10 +920,12 @@ _GLOBAL(giveup_spe) /* * extern void giveup_fpu(struct task_struct *prev) * - * The e500 core does not have an FPU. + * Not all FSL Book-E cores have an FPU */ +#ifndef CONFIG_PPC_FPU _GLOBAL(giveup_fpu) blr +#endif /* * extern void abort(void) diff --git a/arch/ppc/kernel/misc.S b/arch/ppc/kernel/misc.S index 73f7c23b0dd..e4f1615ec13 100644 --- a/arch/ppc/kernel/misc.S +++ b/arch/ppc/kernel/misc.S @@ -1096,17 +1096,7 @@ _GLOBAL(_get_SP) * and exceptions as if the cpu had performed the load or store. */ -#if defined(CONFIG_4xx) || defined(CONFIG_E500) -_GLOBAL(cvt_fd) - lfs 0,0(r3) - stfd 0,0(r4) - blr - -_GLOBAL(cvt_df) - lfd 0,0(r3) - stfs 0,0(r4) - blr -#else +#ifdef CONFIG_PPC_FPU _GLOBAL(cvt_fd) lfd 0,-4(r5) /* load up fpscr value */ mtfsf 0xff,0 diff --git a/arch/ppc/kernel/traps.c b/arch/ppc/kernel/traps.c index 361865c4bc8..f8e7e324a17 100644 --- a/arch/ppc/kernel/traps.c +++ b/arch/ppc/kernel/traps.c @@ -176,7 +176,7 @@ static inline int check_io_access(struct pt_regs *regs) #else #define get_mc_reason(regs) (mfspr(SPRN_MCSR)) #endif -#define REASON_FP 0 +#define REASON_FP ESR_FP #define REASON_ILLEGAL ESR_PIL #define REASON_PRIVILEGED ESR_PPR #define REASON_TRAP ESR_PTR diff --git a/include/asm-ppc/reg_booke.h b/include/asm-ppc/reg_booke.h index e70c25f3c33..45c5e6f2b7a 100644 --- a/include/asm-ppc/reg_booke.h +++ b/include/asm-ppc/reg_booke.h @@ -305,6 +305,7 @@ do { \ #define ESR_PIL 0x08000000 /* Program Exception - Illegal */ #define ESR_PPR 0x04000000 /* Program Exception - Priveleged */ #define ESR_PTR 0x02000000 /* Program Exception - Trap */ +#define ESR_FP 0x01000000 /* Floating Point Operation */ #define ESR_DST 0x00800000 /* Storage Exception - Data miss */ #define ESR_DIZ 0x00400000 /* Storage Exception - Zone fault */ #define ESR_ST 0x00800000 /* Store Operation */ -- cgit v1.2.3-70-g09d2 From 630710e3f72b6a324a525f6e5db3f7985d0fc4a2 Mon Sep 17 00:00:00 2001 From: Chris Elston Date: Sun, 1 May 2005 08:58:40 -0700 Subject: [PATCH] ppc32: fix for misreported SDRAM size on Radstone PPC7D platform This patch fixes the SDRAM output from /proc/cpuinfo. The previous code assumed that there was only one bank of SDRAM, and that the size in the memory configuration register was the total size. Signed-off-by: Chris Elston Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/platforms/radstone_ppc7d.c | 18 ++++++++++++------ arch/ppc/platforms/radstone_ppc7d.h | 1 + 2 files changed, 13 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/ppc/platforms/radstone_ppc7d.c b/arch/ppc/platforms/radstone_ppc7d.c index 2a99b43737a..df2ea051380 100644 --- a/arch/ppc/platforms/radstone_ppc7d.c +++ b/arch/ppc/platforms/radstone_ppc7d.c @@ -253,6 +253,8 @@ static int ppc7d_show_cpuinfo(struct seq_file *m) u8 val1, val2; static int flash_sizes[4] = { 64, 32, 0, 16 }; static int flash_banks[4] = { 4, 3, 2, 1 }; + static int sdram_bank_sizes[4] = { 128, 256, 512, 1 }; + int sdram_num_banks = 2; static char *pci_modes[] = { "PCI33", "PCI66", "Unknown", "Unknown", "PCIX33", "PCIX66", @@ -279,13 +281,17 @@ static int ppc7d_show_cpuinfo(struct seq_file *m) (val1 == PPC7D_CPLD_MB_TYPE_PLL_100) ? 100 : (val1 == PPC7D_CPLD_MB_TYPE_PLL_64) ? 64 : 0); + val = inb(PPC7D_CPLD_MEM_CONFIG); + if (val & PPC7D_CPLD_SDRAM_BANK_NUM_MASK) sdram_num_banks--; + val = inb(PPC7D_CPLD_MEM_CONFIG_EXTEND); - val1 = val & PPC7D_CPLD_SDRAM_BANK_SIZE_MASK; - seq_printf(m, "SDRAM\t\t: %d%c", - (val1 == PPC7D_CPLD_SDRAM_BANK_SIZE_128M) ? 128 : - (val1 == PPC7D_CPLD_SDRAM_BANK_SIZE_256M) ? 256 : - (val1 == PPC7D_CPLD_SDRAM_BANK_SIZE_512M) ? 512 : 1, - (val1 == PPC7D_CPLD_SDRAM_BANK_SIZE_1G) ? 'G' : 'M'); + val1 = (val & PPC7D_CPLD_SDRAM_BANK_SIZE_MASK) >> 6; + seq_printf(m, "SDRAM\t\t: %d banks of %d%c, total %d%c", + sdram_num_banks, + sdram_bank_sizes[val1], + (sdram_bank_sizes[val1] < 128) ? 'G' : 'M', + sdram_num_banks * sdram_bank_sizes[val1], + (sdram_bank_sizes[val1] < 128) ? 'G' : 'M'); if (val2 & PPC7D_CPLD_MB_TYPE_ECC_FITTED_MASK) { seq_printf(m, " [ECC %sabled]", (val2 & PPC7D_CPLD_MB_TYPE_ECC_ENABLE_MASK) ? "en" : diff --git a/arch/ppc/platforms/radstone_ppc7d.h b/arch/ppc/platforms/radstone_ppc7d.h index 4546fff2b0c..938375510be 100644 --- a/arch/ppc/platforms/radstone_ppc7d.h +++ b/arch/ppc/platforms/radstone_ppc7d.h @@ -240,6 +240,7 @@ #define PPC7D_CPLD_FLASH_CNTL 0x086E /* MEMORY_CONFIG_EXTEND */ +#define PPC7D_CPLD_SDRAM_BANK_NUM_MASK 0x02 #define PPC7D_CPLD_SDRAM_BANK_SIZE_MASK 0xc0 #define PPC7D_CPLD_SDRAM_BANK_SIZE_128M 0 #define PPC7D_CPLD_SDRAM_BANK_SIZE_256M 0x40 -- cgit v1.2.3-70-g09d2 From a497aa20e5ea54fdee474192d6dc138b4832fc9e Mon Sep 17 00:00:00 2001 From: Chris Elston Date: Sun, 1 May 2005 08:58:41 -0700 Subject: [PATCH] ppc32: add rtc hooks in PPC7D platform file This patch adds the hooks into the PPC7D platforms file to support the DS1337 RTC device as the clock device for the PPC7D board. Signed-off-by: Chris Elston Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/platforms/radstone_ppc7d.c | 42 +++++++++++++++++++++++++++++++++++++ 1 file changed, 42 insertions(+) (limited to 'arch') diff --git a/arch/ppc/platforms/radstone_ppc7d.c b/arch/ppc/platforms/radstone_ppc7d.c index df2ea051380..c30607a972d 100644 --- a/arch/ppc/platforms/radstone_ppc7d.c +++ b/arch/ppc/platforms/radstone_ppc7d.c @@ -68,6 +68,7 @@ #define PPC7D_RST_PIN 17 /* GPP17 */ extern u32 mv64360_irq_base; +extern spinlock_t rtc_lock; static struct mv64x60_handle bh; static int ppc7d_has_alma; @@ -75,6 +76,11 @@ static int ppc7d_has_alma; extern void gen550_progress(char *, unsigned short); extern void gen550_init(int, struct uart_port *); +/* FIXME - move to h file */ +extern int ds1337_do_command(int id, int cmd, void *arg); +#define DS1337_GET_DATE 0 +#define DS1337_SET_DATE 1 + /* residual data */ unsigned char __res[sizeof(bd_t)]; @@ -1242,6 +1248,38 @@ static void __init ppc7d_setup_arch(void) printk(KERN_INFO "Radstone Technology PPC7D\n"); if (ppc_md.progress) ppc_md.progress("ppc7d_setup_arch: exit", 0); + +} + +/* Real Time Clock support. + * PPC7D has a DS1337 accessed by I2C. + */ +static ulong ppc7d_get_rtc_time(void) +{ + struct rtc_time tm; + int result; + + spin_lock(&rtc_lock); + result = ds1337_do_command(0, DS1337_GET_DATE, &tm); + spin_unlock(&rtc_lock); + + if (result == 0) + result = mktime(tm.tm_year, tm.tm_mon, tm.tm_mday, tm.tm_hour, tm.tm_min, tm.tm_sec); + + return result; +} + +static int ppc7d_set_rtc_time(unsigned long nowtime) +{ + struct rtc_time tm; + int result; + + spin_lock(&rtc_lock); + to_tm(nowtime, &tm); + result = ds1337_do_command(0, DS1337_SET_DATE, &tm); + spin_unlock(&rtc_lock); + + return result; } /* This kernel command line parameter can be used to have the target @@ -1299,6 +1337,10 @@ static void ppc7d_init2(void) data8 |= 0x07; outb(data8, PPC7D_CPLD_LEDS); + /* Hook up RTC. We couldn't do this earlier because we need the I2C subsystem */ + ppc_md.set_rtc_time = ppc7d_set_rtc_time; + ppc_md.get_rtc_time = ppc7d_get_rtc_time; + pr_debug("%s: exit\n", __FUNCTION__); } -- cgit v1.2.3-70-g09d2 From b20cc8aff258eea5a2339107605ebea949fa6ecf Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Sun, 1 May 2005 08:58:42 -0700 Subject: [PATCH] ppc32: Fix a sleep issues on some laptops Some earlier models of aluminium powerbooks and ibook G4s have a clock chip that requires some tweaking before and after sleep. It seems that without that magic incantation to disable and re-enable clock spreading, RAM isn't properly refreshed during sleep. This fixes it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/platforms/pmac_feature.c | 214 ++++++++++++++++++++------------------ 1 file changed, 114 insertions(+), 100 deletions(-) (limited to 'arch') diff --git a/arch/ppc/platforms/pmac_feature.c b/arch/ppc/platforms/pmac_feature.c index 46cbf36722d..be41f6fa66a 100644 --- a/arch/ppc/platforms/pmac_feature.c +++ b/arch/ppc/platforms/pmac_feature.c @@ -1590,6 +1590,112 @@ intrepid_shutdown(struct macio_chip* macio, int sleep_mode) mdelay(10); } + +static void __pmac pmac_tweak_clock_spreading(struct macio_chip* macio, int enable) +{ + /* Hack for doing clock spreading on some machines PowerBooks and + * iBooks. This implements the "platform-do-clockspreading" OF + * property as decoded manually on various models. For safety, we also + * check the product ID in the device-tree in cases we'll whack the i2c + * chip to make reasonably sure we won't set wrong values in there + * + * Of course, ultimately, we have to implement a real parser for + * the platform-do-* stuff... + */ + + if (macio->type == macio_intrepid) { + if (enable) + UN_OUT(UNI_N_CLOCK_SPREADING, 2); + else + UN_OUT(UNI_N_CLOCK_SPREADING, 0); + mdelay(40); + } + + while (machine_is_compatible("PowerBook5,2") || + machine_is_compatible("PowerBook5,3") || + machine_is_compatible("PowerBook6,2") || + machine_is_compatible("PowerBook6,3")) { + struct device_node *ui2c = of_find_node_by_type(NULL, "i2c"); + struct device_node *dt = of_find_node_by_name(NULL, "device-tree"); + u8 buffer[9]; + u32 *productID; + int i, rc, changed = 0; + + if (dt == NULL) + break; + productID = (u32 *)get_property(dt, "pid#", NULL); + if (productID == NULL) + break; + while(ui2c) { + struct device_node *p = of_get_parent(ui2c); + if (p && !strcmp(p->name, "uni-n")) + break; + ui2c = of_find_node_by_type(ui2c, "i2c"); + } + if (ui2c == NULL) + break; + DBG("Trying to bump clock speed for PID: %08x...\n", *productID); + rc = pmac_low_i2c_open(ui2c, 1); + if (rc != 0) + break; + pmac_low_i2c_setmode(ui2c, pmac_low_i2c_mode_combined); + rc = pmac_low_i2c_xfer(ui2c, 0xd2 | pmac_low_i2c_read, 0x80, buffer, 9); + DBG("read result: %d,", rc); + if (rc != 0) { + pmac_low_i2c_close(ui2c); + break; + } + for (i=0; i<9; i++) + DBG(" %02x", buffer[i]); + DBG("\n"); + + switch(*productID) { + case 0x1182: /* AlBook 12" rev 2 */ + case 0x1183: /* iBook G4 12" */ + buffer[0] = (buffer[0] & 0x8f) | 0x70; + buffer[2] = (buffer[2] & 0x7f) | 0x00; + buffer[5] = (buffer[5] & 0x80) | 0x31; + buffer[6] = (buffer[6] & 0x40) | 0xb0; + buffer[7] = (buffer[7] & 0x00) | (enable ? 0xc0 : 0xba); + buffer[8] = (buffer[8] & 0x00) | 0x30; + changed = 1; + break; + case 0x3142: /* AlBook 15" (ATI M10) */ + case 0x3143: /* AlBook 17" (ATI M10) */ + buffer[0] = (buffer[0] & 0xaf) | 0x50; + buffer[2] = (buffer[2] & 0x7f) | 0x00; + buffer[5] = (buffer[5] & 0x80) | 0x31; + buffer[6] = (buffer[6] & 0x40) | 0xb0; + buffer[7] = (buffer[7] & 0x00) | (enable ? 0xd0 : 0xc0); + buffer[8] = (buffer[8] & 0x00) | 0x30; + changed = 1; + break; + default: + DBG("i2c-hwclock: Machine model not handled\n"); + break; + } + if (!changed) { + pmac_low_i2c_close(ui2c); + break; + } + pmac_low_i2c_setmode(ui2c, pmac_low_i2c_mode_stdsub); + rc = pmac_low_i2c_xfer(ui2c, 0xd2 | pmac_low_i2c_write, 0x80, buffer, 9); + DBG("write result: %d,", rc); + pmac_low_i2c_setmode(ui2c, pmac_low_i2c_mode_combined); + rc = pmac_low_i2c_xfer(ui2c, 0xd2 | pmac_low_i2c_read, 0x80, buffer, 9); + DBG("read result: %d,", rc); + if (rc != 0) { + pmac_low_i2c_close(ui2c); + break; + } + for (i=0; i<9; i++) + DBG(" %02x", buffer[i]); + pmac_low_i2c_close(ui2c); + break; + } +} + + static int __pmac core99_sleep(void) { @@ -1601,11 +1707,8 @@ core99_sleep(void) macio->type != macio_intrepid) return -ENODEV; - /* The device-tree contains that in the hwclock node */ - if (macio->type == macio_intrepid) { - UN_OUT(UNI_N_CLOCK_SPREADING, 0); - mdelay(40); - } + /* Disable clock spreading */ + pmac_tweak_clock_spreading(macio, 0); /* We power off the wireless slot in case it was not done * by the driver. We don't power it on automatically however @@ -1749,11 +1852,8 @@ core99_wake_up(void) UN_OUT(UNI_N_CLOCK_CNTL, save_unin_clock_ctl); udelay(100); - /* Restore clock spreading */ - if (macio->type == macio_intrepid) { - UN_OUT(UNI_N_CLOCK_SPREADING, 2); - mdelay(40); - } + /* Enable clock spreading */ + pmac_tweak_clock_spreading(macio, 1); return 0; } @@ -2718,97 +2818,11 @@ set_initial_features(void) MACIO_BIC(HEATHROW_FCR, HRW_SOUND_POWER_N); } - /* Hack for bumping clock speed on the new PowerBooks and the - * iBook G4. This implements the "platform-do-clockspreading" OF - * property. For safety, we also check the product ID in the - * device-tree to make reasonably sure we won't set wrong values - * in the clock chip. - * - * Of course, ultimately, we have to implement a real parser for - * the platform-do-* stuff... + /* Some machine models need the clock chip to be properly setup for + * clock spreading now. This should be a platform function but we + * don't do these at the moment */ - while (machine_is_compatible("PowerBook5,2") || - machine_is_compatible("PowerBook5,3") || - machine_is_compatible("PowerBook6,2") || - machine_is_compatible("PowerBook6,3")) { - struct device_node *ui2c = of_find_node_by_type(NULL, "i2c"); - struct device_node *dt = of_find_node_by_name(NULL, "device-tree"); - u8 buffer[9]; - u32 *productID; - int i, rc, changed = 0; - - if (dt == NULL) - break; - productID = (u32 *)get_property(dt, "pid#", NULL); - if (productID == NULL) - break; - while(ui2c) { - struct device_node *p = of_get_parent(ui2c); - if (p && !strcmp(p->name, "uni-n")) - break; - ui2c = of_find_node_by_type(ui2c, "i2c"); - } - if (ui2c == NULL) - break; - DBG("Trying to bump clock speed for PID: %08x...\n", *productID); - rc = pmac_low_i2c_open(ui2c, 1); - if (rc != 0) - break; - pmac_low_i2c_setmode(ui2c, pmac_low_i2c_mode_combined); - rc = pmac_low_i2c_xfer(ui2c, 0xd2 | pmac_low_i2c_read, 0x80, buffer, 9); - DBG("read result: %d,", rc); - if (rc != 0) { - pmac_low_i2c_close(ui2c); - break; - } - for (i=0; i<9; i++) - DBG(" %02x", buffer[i]); - DBG("\n"); - - switch(*productID) { - case 0x1182: /* AlBook 12" rev 2 */ - case 0x1183: /* iBook G4 12" */ - buffer[0] = (buffer[0] & 0x8f) | 0x70; - buffer[2] = (buffer[2] & 0x7f) | 0x00; - buffer[5] = (buffer[5] & 0x80) | 0x31; - buffer[6] = (buffer[6] & 0x40) | 0xb0; - buffer[7] = (buffer[7] & 0x00) | 0xc0; - buffer[8] = (buffer[8] & 0x00) | 0x30; - changed = 1; - break; - case 0x3142: /* AlBook 15" (ATI M10) */ - case 0x3143: /* AlBook 17" (ATI M10) */ - buffer[0] = (buffer[0] & 0xaf) | 0x50; - buffer[2] = (buffer[2] & 0x7f) | 0x00; - buffer[5] = (buffer[5] & 0x80) | 0x31; - buffer[6] = (buffer[6] & 0x40) | 0xb0; - buffer[7] = (buffer[7] & 0x00) | 0xd0; - buffer[8] = (buffer[8] & 0x00) | 0x30; - changed = 1; - break; - default: - DBG("i2c-hwclock: Machine model not handled\n"); - break; - } - if (!changed) { - pmac_low_i2c_close(ui2c); - break; - } - pmac_low_i2c_setmode(ui2c, pmac_low_i2c_mode_stdsub); - rc = pmac_low_i2c_xfer(ui2c, 0xd2 | pmac_low_i2c_write, 0x80, buffer, 9); - DBG("write result: %d,", rc); - pmac_low_i2c_setmode(ui2c, pmac_low_i2c_mode_combined); - rc = pmac_low_i2c_xfer(ui2c, 0xd2 | pmac_low_i2c_read, 0x80, buffer, 9); - DBG("read result: %d,", rc); - if (rc != 0) { - pmac_low_i2c_close(ui2c); - break; - } - for (i=0; i<9; i++) - DBG(" %02x", buffer[i]); - pmac_low_i2c_close(ui2c); - break; - } + pmac_tweak_clock_spreading(&macio_chips[0], 1); #endif /* CONFIG_POWER4 */ -- cgit v1.2.3-70-g09d2 From d5812a77e5803468a5033be91af978be0f7a17d9 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Sun, 1 May 2005 08:58:42 -0700 Subject: [PATCH] ppc32: Fix address checking on lmw/stmw align exception The handling of misaligned load/store multiple instructions did not check to see if the address was ok to access before using __{get,put}_user(). Signed-off-by: Kumar Gala Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/kernel/align.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch') diff --git a/arch/ppc/kernel/align.c b/arch/ppc/kernel/align.c index 40d356c66c4..ff81da9598d 100644 --- a/arch/ppc/kernel/align.c +++ b/arch/ppc/kernel/align.c @@ -290,6 +290,10 @@ fix_alignment(struct pt_regs *regs) /* lwm, stmw */ nb = (32 - reg) * 4; } + + if (!access_ok((flags & ST? VERIFY_WRITE: VERIFY_READ), addr, nb+nb0)) + return -EFAULT; /* bad address */ + rptr = (unsigned char *) ®s->gpr[reg]; if (flags & LD) { for (i = 0; i < nb; ++i) -- cgit v1.2.3-70-g09d2 From 1bdacf88ebd7969fecbbf4c5b388cc094871222e Mon Sep 17 00:00:00 2001 From: Dan Malek Date: Sun, 1 May 2005 08:58:42 -0700 Subject: [PATCH] ppc32: workaround for spurious IRQs on PQ2 There is a problem with large amounts of spurious IRQs on PowerPC 82xx systems. The problem is corrected by adding sync at the end of cpm2_mask_and_ack. This may be needed on 8xx as well but has not yet been confirmed. Signed-off-by: Eugene Surovegin Signed-off-by: Dan Malek Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/syslib/cpm2_pic.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'arch') diff --git a/arch/ppc/syslib/cpm2_pic.c b/arch/ppc/syslib/cpm2_pic.c index 954b07fc1df..c867be6981c 100644 --- a/arch/ppc/syslib/cpm2_pic.c +++ b/arch/ppc/syslib/cpm2_pic.c @@ -107,6 +107,11 @@ static void cpm2_end_irq(unsigned int irq_nr) simr = &(cpm2_immr->im_intctl.ic_simrh); ppc_cached_irq_mask[word] |= 1 << bit; simr[word] = ppc_cached_irq_mask[word]; + /* + * Work around large numbers of spurious IRQs on PowerPC 82xx + * systems. + */ + mb(); } } -- cgit v1.2.3-70-g09d2 From 1b29f9d13e3cf0fe86bf7f82a3399c9e3caf58e5 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Sun, 1 May 2005 08:58:43 -0700 Subject: [PATCH] ppc64: add PT_NOTE section to vDSO This patch from Roland adds a PT_NOTE section to both 32 and 64 bits vDSOs to expose the kernel version to glibc, thus avoiding a uname syscall on every launch. This is equivalent to the patches Roland posted already for x86 and x86-64. Note: the 64 bits .note is actually using the 32 bits format. This is normal. The ELF spec specifies a different format for 64 bits .note, but for some reason, this was never properly implemented, the core dumps for example are all using 32 bits format .note, and binutils cannot even read a 64 bits format .note. Talking to our toolchain folks, they think we'd rather stick to 32 bits format .note everywhere and get the spec fixed some day ... Signed-off-by: Roland McGrath Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/vdso32/Makefile | 2 +- arch/ppc64/kernel/vdso32/note.S | 25 +++++++++++++++++++++++++ arch/ppc64/kernel/vdso32/vdso32.lds.S | 3 +++ arch/ppc64/kernel/vdso64/Makefile | 2 +- arch/ppc64/kernel/vdso64/note.S | 1 + arch/ppc64/kernel/vdso64/vdso64.lds.S | 5 ++++- 6 files changed, 35 insertions(+), 3 deletions(-) create mode 100644 arch/ppc64/kernel/vdso32/note.S create mode 100644 arch/ppc64/kernel/vdso64/note.S (limited to 'arch') diff --git a/arch/ppc64/kernel/vdso32/Makefile b/arch/ppc64/kernel/vdso32/Makefile index ede2f7e477c..0b1b0df973e 100644 --- a/arch/ppc64/kernel/vdso32/Makefile +++ b/arch/ppc64/kernel/vdso32/Makefile @@ -1,7 +1,7 @@ # List of files in the vdso, has to be asm only for now -obj-vdso32 = sigtramp.o gettimeofday.o datapage.o cacheflush.o +obj-vdso32 = sigtramp.o gettimeofday.o datapage.o cacheflush.o note.o # Build rules diff --git a/arch/ppc64/kernel/vdso32/note.S b/arch/ppc64/kernel/vdso32/note.S new file mode 100644 index 00000000000..d4b5be4f3d5 --- /dev/null +++ b/arch/ppc64/kernel/vdso32/note.S @@ -0,0 +1,25 @@ +/* + * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text. + * Here we can supply some information useful to userland. + */ + +#include +#include + +#define ASM_ELF_NOTE_BEGIN(name, flags, vendor, type) \ + .section name, flags; \ + .balign 4; \ + .long 1f - 0f; /* name length */ \ + .long 3f - 2f; /* data length */ \ + .long type; /* note type */ \ +0: .asciz vendor; /* vendor name */ \ +1: .balign 4; \ +2: + +#define ASM_ELF_NOTE_END \ +3: .balign 4; /* pad out section */ \ + .previous + + ASM_ELF_NOTE_BEGIN(".note.kernel-version", "a", UTS_SYSNAME, 0) + .long LINUX_VERSION_CODE + ASM_ELF_NOTE_END diff --git a/arch/ppc64/kernel/vdso32/vdso32.lds.S b/arch/ppc64/kernel/vdso32/vdso32.lds.S index cca27bd03a5..11290c902ba 100644 --- a/arch/ppc64/kernel/vdso32/vdso32.lds.S +++ b/arch/ppc64/kernel/vdso32/vdso32.lds.S @@ -20,6 +20,8 @@ SECTIONS .gnu.version_d : { *(.gnu.version_d) } .gnu.version_r : { *(.gnu.version_r) } + .note : { *(.note.*) } :text :note + . = ALIGN (16); .text : { @@ -87,6 +89,7 @@ SECTIONS PHDRS { text PT_LOAD FILEHDR PHDRS FLAGS(5); /* PF_R|PF_X */ + note PT_NOTE FLAGS(4); /* PF_R */ dynamic PT_DYNAMIC FLAGS(4); /* PF_R */ eh_frame_hdr 0x6474e550; /* PT_GNU_EH_FRAME, but ld doesn't match the name */ } diff --git a/arch/ppc64/kernel/vdso64/Makefile b/arch/ppc64/kernel/vdso64/Makefile index bd3f70b1a38..ab39988452c 100644 --- a/arch/ppc64/kernel/vdso64/Makefile +++ b/arch/ppc64/kernel/vdso64/Makefile @@ -1,6 +1,6 @@ # List of files in the vdso, has to be asm only for now -obj-vdso64 = sigtramp.o gettimeofday.o datapage.o cacheflush.o +obj-vdso64 = sigtramp.o gettimeofday.o datapage.o cacheflush.o note.o # Build rules diff --git a/arch/ppc64/kernel/vdso64/note.S b/arch/ppc64/kernel/vdso64/note.S new file mode 100644 index 00000000000..dc2a509f7e8 --- /dev/null +++ b/arch/ppc64/kernel/vdso64/note.S @@ -0,0 +1 @@ +#include "../vdso32/note.S" diff --git a/arch/ppc64/kernel/vdso64/vdso64.lds.S b/arch/ppc64/kernel/vdso64/vdso64.lds.S index 942c815c7bc..9cb28181da8 100644 --- a/arch/ppc64/kernel/vdso64/vdso64.lds.S +++ b/arch/ppc64/kernel/vdso64/vdso64.lds.S @@ -18,12 +18,14 @@ SECTIONS .gnu.version_d : { *(.gnu.version_d) } .gnu.version_r : { *(.gnu.version_r) } + .note : { *(.note.*) } :text :note + . = ALIGN (16); .text : { *(.text .stub .text.* .gnu.linkonce.t.*) *(.sfpr .glink) - } + } :text PROVIDE (__etext = .); PROVIDE (_etext = .); PROVIDE (etext = .); @@ -88,6 +90,7 @@ SECTIONS PHDRS { text PT_LOAD FILEHDR PHDRS FLAGS(5); /* PF_R|PF_X */ + note PT_NOTE FLAGS(4); /* PF_R */ dynamic PT_DYNAMIC FLAGS(4); /* PF_R */ eh_frame_hdr 0x6474e550; /* PT_GNU_EH_FRAME, but ld doesn't match the name */ } -- cgit v1.2.3-70-g09d2 From bb78cb72201985ae9269b723c82ea0f892048b9e Mon Sep 17 00:00:00 2001 From: Olof Johansson Date: Sun, 1 May 2005 08:58:44 -0700 Subject: [PATCH] ppc64: remove unused argument to create_slbe Remove vsid argument to create_slbe, since it's no longer used. Spotted by R Sharada. Signed-off-by: Olof Johansson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/mm/slb.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/mm/slb.c b/arch/ppc64/mm/slb.c index 6a20773f695..244150a0bc1 100644 --- a/arch/ppc64/mm/slb.c +++ b/arch/ppc64/mm/slb.c @@ -33,8 +33,8 @@ static inline unsigned long mk_vsid_data(unsigned long ea, unsigned long flags) return (get_kernel_vsid(ea) << SLB_VSID_SHIFT) | flags; } -static inline void create_slbe(unsigned long ea, unsigned long vsid, - unsigned long flags, unsigned long entry) +static inline void create_slbe(unsigned long ea, unsigned long flags, + unsigned long entry) { asm volatile("slbmte %0,%1" : : "r" (mk_vsid_data(ea, flags)), @@ -145,9 +145,8 @@ void slb_initialize(void) asm volatile("isync":::"memory"); asm volatile("slbmte %0,%0"::"r" (0) : "memory"); asm volatile("isync; slbia; isync":::"memory"); - create_slbe(KERNELBASE, get_kernel_vsid(KERNELBASE), flags, 0); - create_slbe(VMALLOCBASE, get_kernel_vsid(KERNELBASE), - SLB_VSID_KERNEL, 1); + create_slbe(KERNELBASE, flags, 0); + create_slbe(VMALLOCBASE, SLB_VSID_KERNEL, 1); /* We don't bolt the stack for the time being - we're in boot, * so the stack is in the bolted segment. By the time it goes * elsewhere, we'll call _switch() which will bolt in the new -- cgit v1.2.3-70-g09d2 From dc3ec7503e693e05c01d85b664482d0f694429ab Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Sun, 1 May 2005 08:58:44 -0700 Subject: [PATCH] ppc64: Fix irq parsing on powermac When I tried Ben's patches to the powermac sound driver on my G5, I found that it was taking enormous numbers of sound DMA transmit interrupts. This turned out to be because it was incorrectly configured as level-sensitive instead of edge-sensitive, which in turn was because the code that parses the interrupt tree that Open Firmware gives us was incorrectly assigning another device the same irq number as the sound DMA transmit interrupt (i.e. 1). This patch fixes the problem, in a somewhat quick and dirty way for now, but one which will work for all the machines we currently run on. Ultimately Ben and I want to do something more general and robust, but this should go in for 2.6.12. Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/prom.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch') diff --git a/arch/ppc64/kernel/prom.c b/arch/ppc64/kernel/prom.c index 45a4ad08fbc..fe2946c5831 100644 --- a/arch/ppc64/kernel/prom.c +++ b/arch/ppc64/kernel/prom.c @@ -321,6 +321,10 @@ static int __devinit finish_node_interrupts(struct device_node *np, char *name = get_property(ic->parent, "name", NULL); if (name && !strcmp(name, "u3")) np->intrs[intrcount].line += 128; + else if (!(name && !strcmp(name, "mac-io"))) + /* ignore other cascaded controllers, such as + the k2-sata-root */ + break; } np->intrs[intrcount].sense = 1; if (n > 1) -- cgit v1.2.3-70-g09d2 From 0339ad77c4a06fa8529db17c91f790058e18b65b Mon Sep 17 00:00:00 2001 From: "akpm@osdl.org" Date: Sun, 1 May 2005 08:58:44 -0700 Subject: [PATCH] ppc64: nvram cleanups - Fix arch/ppc64/kernel/nvram.c:342: warning: `part' might be used uninitialized in this function - Various codingstyle tweaks. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/nvram.c | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/kernel/nvram.c b/arch/ppc64/kernel/nvram.c index b9069c2d193..4e71781a441 100644 --- a/arch/ppc64/kernel/nvram.c +++ b/arch/ppc64/kernel/nvram.c @@ -339,9 +339,9 @@ static int nvram_remove_os_partition(void) static int nvram_create_os_partition(void) { struct list_head * p; - struct nvram_partition * part; - struct nvram_partition * new_part = NULL; - struct nvram_partition * free_part = NULL; + struct nvram_partition *part = NULL; + struct nvram_partition *new_part = NULL; + struct nvram_partition *free_part = NULL; int seq_init[2] = { 0, 0 }; loff_t tmp_index; long size = 0; @@ -364,13 +364,11 @@ static int nvram_create_os_partition(void) free_part = part; } } - if (!size) { + if (!size) return -ENOSPC; - } /* Create our OS partition */ - new_part = (struct nvram_partition *) - kmalloc(sizeof(struct nvram_partition), GFP_KERNEL); + new_part = kmalloc(sizeof(*new_part), GFP_KERNEL); if (!new_part) { printk(KERN_ERR "nvram_create_os_partition: kmalloc failed\n"); return -ENOMEM; @@ -379,7 +377,7 @@ static int nvram_create_os_partition(void) new_part->index = free_part->index; new_part->header.signature = NVRAM_SIG_OS; new_part->header.length = size; - sprintf(new_part->header.name, "ppc64,linux"); + strcpy(new_part->header.name, "ppc64,linux"); new_part->header.checksum = nvram_checksum(&new_part->header); rc = nvram_write_header(new_part); @@ -394,7 +392,8 @@ static int nvram_create_os_partition(void) tmp_index = new_part->index + NVRAM_HEADER_LEN; rc = ppc_md.nvram_write((char *)&seq_init, sizeof(seq_init), &tmp_index); if (rc <= 0) { - printk(KERN_ERR "nvram_create_os_partition: nvram_write failed (%d)\n", rc); + printk(KERN_ERR "nvram_create_os_partition: nvram_write " + "failed (%d)\n", rc); return rc; } -- cgit v1.2.3-70-g09d2 From 58366af5861eee1479426380e3c91ecb334c301d Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Sun, 1 May 2005 08:58:44 -0700 Subject: [PATCH] ppc64: update to use the new 4L headers This patch converts ppc64 to use the generic pgtable-nopud.h instead of the "fixup" header. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/mm/hugetlbpage.c | 45 +++++----- arch/ppc64/mm/init.c | 198 ++++++++++++++++++++++---------------------- include/asm-ppc64/pgalloc.h | 2 +- include/asm-ppc64/pgtable.h | 41 +++++---- 4 files changed, 147 insertions(+), 139 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/mm/hugetlbpage.c b/arch/ppc64/mm/hugetlbpage.c index 390296efe3e..d3bf86a5c1a 100644 --- a/arch/ppc64/mm/hugetlbpage.c +++ b/arch/ppc64/mm/hugetlbpage.c @@ -42,7 +42,7 @@ static inline int hugepgd_index(unsigned long addr) return (addr & ~REGION_MASK) >> HUGEPGDIR_SHIFT; } -static pgd_t *hugepgd_offset(struct mm_struct *mm, unsigned long addr) +static pud_t *hugepgd_offset(struct mm_struct *mm, unsigned long addr) { int index; @@ -52,21 +52,21 @@ static pgd_t *hugepgd_offset(struct mm_struct *mm, unsigned long addr) index = hugepgd_index(addr); BUG_ON(index >= PTRS_PER_HUGEPGD); - return mm->context.huge_pgdir + index; + return (pud_t *)(mm->context.huge_pgdir + index); } -static inline pte_t *hugepte_offset(pgd_t *dir, unsigned long addr) +static inline pte_t *hugepte_offset(pud_t *dir, unsigned long addr) { int index; - if (pgd_none(*dir)) + if (pud_none(*dir)) return NULL; index = (addr >> HPAGE_SHIFT) % PTRS_PER_HUGEPTE; - return (pte_t *)pgd_page(*dir) + index; + return (pte_t *)pud_page(*dir) + index; } -static pgd_t *hugepgd_alloc(struct mm_struct *mm, unsigned long addr) +static pud_t *hugepgd_alloc(struct mm_struct *mm, unsigned long addr) { BUG_ON(! in_hugepage_area(mm->context, addr)); @@ -90,10 +90,9 @@ static pgd_t *hugepgd_alloc(struct mm_struct *mm, unsigned long addr) return hugepgd_offset(mm, addr); } -static pte_t *hugepte_alloc(struct mm_struct *mm, pgd_t *dir, - unsigned long addr) +static pte_t *hugepte_alloc(struct mm_struct *mm, pud_t *dir, unsigned long addr) { - if (! pgd_present(*dir)) { + if (! pud_present(*dir)) { pte_t *new; spin_unlock(&mm->page_table_lock); @@ -104,7 +103,7 @@ static pte_t *hugepte_alloc(struct mm_struct *mm, pgd_t *dir, * Because we dropped the lock, we should re-check the * entry, as somebody else could have populated it.. */ - if (pgd_present(*dir)) { + if (pud_present(*dir)) { if (new) kmem_cache_free(zero_cache, new); } else { @@ -115,7 +114,7 @@ static pte_t *hugepte_alloc(struct mm_struct *mm, pgd_t *dir, ptepage = virt_to_page(new); ptepage->mapping = (void *) mm; ptepage->index = addr & HUGEPGDIR_MASK; - pgd_populate(mm, dir, new); + pud_populate(mm, dir, new); } } @@ -124,28 +123,28 @@ static pte_t *hugepte_alloc(struct mm_struct *mm, pgd_t *dir, static pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr) { - pgd_t *pgd; + pud_t *pud; BUG_ON(! in_hugepage_area(mm->context, addr)); - pgd = hugepgd_offset(mm, addr); - if (! pgd) + pud = hugepgd_offset(mm, addr); + if (! pud) return NULL; - return hugepte_offset(pgd, addr); + return hugepte_offset(pud, addr); } static pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr) { - pgd_t *pgd; + pud_t *pud; BUG_ON(! in_hugepage_area(mm->context, addr)); - pgd = hugepgd_alloc(mm, addr); - if (! pgd) + pud = hugepgd_alloc(mm, addr); + if (! pud) return NULL; - return hugepte_alloc(mm, pgd, addr); + return hugepte_alloc(mm, pud, addr); } static void set_huge_pte(struct mm_struct *mm, struct vm_area_struct *vma, @@ -709,10 +708,10 @@ void hugetlb_mm_free_pgd(struct mm_struct *mm) /* cleanup any hugepte pages leftover */ for (i = 0; i < PTRS_PER_HUGEPGD; i++) { - pgd_t *pgd = pgdir + i; + pud_t *pud = (pud_t *)(pgdir + i); - if (! pgd_none(*pgd)) { - pte_t *pte = (pte_t *)pgd_page(*pgd); + if (! pud_none(*pud)) { + pte_t *pte = (pte_t *)pud_page(*pud); struct page *ptepage = virt_to_page(pte); ptepage->mapping = NULL; @@ -720,7 +719,7 @@ void hugetlb_mm_free_pgd(struct mm_struct *mm) BUG_ON(memcmp(pte, empty_zero_page, PAGE_SIZE)); kmem_cache_free(zero_cache, pte); } - pgd_clear(pgd); + pud_clear(pud); } BUG_ON(memcmp(pgdir, empty_zero_page, PAGE_SIZE)); diff --git a/arch/ppc64/mm/init.c b/arch/ppc64/mm/init.c index a7149b9fc35..cf33d7ec2e2 100644 --- a/arch/ppc64/mm/init.c +++ b/arch/ppc64/mm/init.c @@ -136,14 +136,78 @@ void iounmap(volatile void __iomem *addr) #else +static void unmap_im_area_pte(pmd_t *pmd, unsigned long addr, + unsigned long end) +{ + pte_t *pte; + + pte = pte_offset_kernel(pmd, addr); + do { + pte_t ptent = ptep_get_and_clear(&ioremap_mm, addr, pte); + WARN_ON(!pte_none(ptent) && !pte_present(ptent)); + } while (pte++, addr += PAGE_SIZE, addr != end); +} + +static inline void unmap_im_area_pmd(pud_t *pud, unsigned long addr, + unsigned long end) +{ + pmd_t *pmd; + unsigned long next; + + pmd = pmd_offset(pud, addr); + do { + next = pmd_addr_end(addr, end); + if (pmd_none_or_clear_bad(pmd)) + continue; + unmap_im_area_pte(pmd, addr, next); + } while (pmd++, addr = next, addr != end); +} + +static inline void unmap_im_area_pud(pgd_t *pgd, unsigned long addr, + unsigned long end) +{ + pud_t *pud; + unsigned long next; + + pud = pud_offset(pgd, addr); + do { + next = pud_addr_end(addr, end); + if (pud_none_or_clear_bad(pud)) + continue; + unmap_im_area_pmd(pud, addr, next); + } while (pud++, addr = next, addr != end); +} + +static void unmap_im_area(unsigned long addr, unsigned long end) +{ + struct mm_struct *mm = &ioremap_mm; + unsigned long next; + pgd_t *pgd; + + spin_lock(&mm->page_table_lock); + + pgd = pgd_offset_i(addr); + flush_cache_vunmap(addr, end); + do { + next = pgd_addr_end(addr, end); + if (pgd_none_or_clear_bad(pgd)) + continue; + unmap_im_area_pud(pgd, addr, next); + } while (pgd++, addr = next, addr != end); + flush_tlb_kernel_range(start, end); + + spin_unlock(&mm->page_table_lock); +} + /* * map_io_page currently only called by __ioremap * map_io_page adds an entry to the ioremap page table * and adds an entry to the HPT, possibly bolting it */ -static void map_io_page(unsigned long ea, unsigned long pa, int flags) +static int map_io_page(unsigned long ea, unsigned long pa, int flags) { pgd_t *pgdp; + pud_t *pudp; pmd_t *pmdp; pte_t *ptep; unsigned long vsid; @@ -151,9 +215,15 @@ static void map_io_page(unsigned long ea, unsigned long pa, int flags) if (mem_init_done) { spin_lock(&ioremap_mm.page_table_lock); pgdp = pgd_offset_i(ea); - pmdp = pmd_alloc(&ioremap_mm, pgdp, ea); + pudp = pud_alloc(&ioremap_mm, pgdp, ea); + if (!pudp) + return -ENOMEM; + pmdp = pmd_alloc(&ioremap_mm, pudp, ea); + if (!pmdp) + return -ENOMEM; ptep = pte_alloc_kernel(&ioremap_mm, pmdp, ea); - + if (!ptep) + return -ENOMEM; pa = abs_to_phys(pa); set_pte_at(&ioremap_mm, ea, ptep, pfn_pte(pa >> PAGE_SHIFT, __pgprot(flags))); @@ -181,6 +251,7 @@ static void map_io_page(unsigned long ea, unsigned long pa, int flags) panic("map_io_page: could not insert mapping"); } } + return 0; } @@ -194,9 +265,14 @@ static void __iomem * __ioremap_com(unsigned long addr, unsigned long pa, flags |= pgprot_val(PAGE_KERNEL); for (i = 0; i < size; i += PAGE_SIZE) - map_io_page(ea+i, pa+i, flags); + if (map_io_page(ea+i, pa+i, flags)) + goto failure; return (void __iomem *) (ea + (addr & ~PAGE_MASK)); + failure: + if (mem_init_done) + unmap_im_area(ea, ea + size); + return NULL; } @@ -206,10 +282,11 @@ ioremap(unsigned long addr, unsigned long size) return __ioremap(addr, size, _PAGE_NO_CACHE | _PAGE_GUARDED); } -void __iomem * -__ioremap(unsigned long addr, unsigned long size, unsigned long flags) +void __iomem * __ioremap(unsigned long addr, unsigned long size, + unsigned long flags) { unsigned long pa, ea; + void __iomem *ret; /* * Choose an address to map it to. @@ -232,12 +309,16 @@ __ioremap(unsigned long addr, unsigned long size, unsigned long flags) if (area == NULL) return NULL; ea = (unsigned long)(area->addr); + ret = __ioremap_com(addr, pa, ea, size, flags); + if (!ret) + im_free(area->addr); } else { ea = ioremap_bot; - ioremap_bot += size; + ret = __ioremap_com(addr, pa, ea, size, flags); + if (ret) + ioremap_bot += size; } - - return __ioremap_com(addr, pa, ea, size, flags); + return ret; } #define IS_PAGE_ALIGNED(_val) ((_val) == ((_val) & PAGE_MASK)) @@ -246,6 +327,7 @@ int __ioremap_explicit(unsigned long pa, unsigned long ea, unsigned long size, unsigned long flags) { struct vm_struct *area; + void __iomem *ret; /* For now, require page-aligned values for pa, ea, and size */ if (!IS_PAGE_ALIGNED(pa) || !IS_PAGE_ALIGNED(ea) || @@ -276,7 +358,12 @@ int __ioremap_explicit(unsigned long pa, unsigned long ea, } } - if (__ioremap_com(pa, pa, ea, size, flags) != (void *) ea) { + ret = __ioremap_com(pa, pa, ea, size, flags); + if (ret == NULL) { + printk(KERN_ERR "ioremap_explicit() allocation failure !\n"); + return 1; + } + if (ret != (void *) ea) { printk(KERN_ERR "__ioremap_com() returned unexpected addr\n"); return 1; } @@ -284,69 +371,6 @@ int __ioremap_explicit(unsigned long pa, unsigned long ea, return 0; } -static void unmap_im_area_pte(pmd_t *pmd, unsigned long address, - unsigned long size) -{ - unsigned long base, end; - pte_t *pte; - - if (pmd_none(*pmd)) - return; - if (pmd_bad(*pmd)) { - pmd_ERROR(*pmd); - pmd_clear(pmd); - return; - } - - pte = pte_offset_kernel(pmd, address); - base = address & PMD_MASK; - address &= ~PMD_MASK; - end = address + size; - if (end > PMD_SIZE) - end = PMD_SIZE; - - do { - pte_t page; - page = ptep_get_and_clear(&ioremap_mm, base + address, pte); - address += PAGE_SIZE; - pte++; - if (pte_none(page)) - continue; - if (pte_present(page)) - continue; - printk(KERN_CRIT "Whee.. Swapped out page in kernel page" - " table\n"); - } while (address < end); -} - -static void unmap_im_area_pmd(pgd_t *dir, unsigned long address, - unsigned long size) -{ - unsigned long base, end; - pmd_t *pmd; - - if (pgd_none(*dir)) - return; - if (pgd_bad(*dir)) { - pgd_ERROR(*dir); - pgd_clear(dir); - return; - } - - pmd = pmd_offset(dir, address); - base = address & PGDIR_MASK; - address &= ~PGDIR_MASK; - end = address + size; - if (end > PGDIR_SIZE) - end = PGDIR_SIZE; - - do { - unmap_im_area_pte(pmd, base + address, end - address); - address = (address + PMD_SIZE) & PMD_MASK; - pmd++; - } while (address < end); -} - /* * Unmap an IO region and remove it from imalloc'd list. * Access to IO memory should be serialized by driver. @@ -356,39 +380,19 @@ static void unmap_im_area_pmd(pgd_t *dir, unsigned long address, */ void iounmap(volatile void __iomem *token) { - unsigned long address, start, end, size; - struct mm_struct *mm; - pgd_t *dir; + unsigned long address, size; void *addr; - if (!mem_init_done) { + if (!mem_init_done) return; - } addr = (void *) ((unsigned long __force) token & PAGE_MASK); - if ((size = im_free(addr)) == 0) { + if ((size = im_free(addr)) == 0) return; - } address = (unsigned long)addr; - start = address; - end = address + size; - - mm = &ioremap_mm; - spin_lock(&mm->page_table_lock); - - dir = pgd_offset_i(address); - flush_cache_vunmap(address, end); - do { - unmap_im_area_pmd(dir, address, end - address); - address = (address + PGDIR_SIZE) & PGDIR_MASK; - dir++; - } while (address && (address < end)); - flush_tlb_kernel_range(start, end); - - spin_unlock(&mm->page_table_lock); - return; + unmap_im_area(address, address + size); } static int iounmap_subset_regions(unsigned long addr, unsigned long size) diff --git a/include/asm-ppc64/pgalloc.h b/include/asm-ppc64/pgalloc.h index 16232d74017..4fc4b739b38 100644 --- a/include/asm-ppc64/pgalloc.h +++ b/include/asm-ppc64/pgalloc.h @@ -27,7 +27,7 @@ pgd_free(pgd_t *pgd) kmem_cache_free(zero_cache, pgd); } -#define pgd_populate(MM, PGD, PMD) pgd_set(PGD, PMD) +#define pud_populate(MM, PUD, PMD) pud_set(PUD, PMD) static inline pmd_t * pmd_alloc_one(struct mm_struct *mm, unsigned long addr) diff --git a/include/asm-ppc64/pgtable.h b/include/asm-ppc64/pgtable.h index a26120517c5..b984e2747e0 100644 --- a/include/asm-ppc64/pgtable.h +++ b/include/asm-ppc64/pgtable.h @@ -1,8 +1,6 @@ #ifndef _PPC64_PGTABLE_H #define _PPC64_PGTABLE_H -#include - /* * This file contains the functions and defines necessary to modify and use * the ppc64 hashed page table. @@ -17,6 +15,8 @@ #include #endif /* __ASSEMBLY__ */ +#include + /* PMD_SHIFT determines what a second-level page table entry can map */ #define PMD_SHIFT (PAGE_SHIFT + PAGE_SHIFT - 3) #define PMD_SIZE (1UL << PMD_SHIFT) @@ -228,12 +228,13 @@ void hugetlb_mm_free_pgd(struct mm_struct *mm); #define pmd_page_kernel(pmd) \ (__bpn_to_ba(pmd_val(pmd) >> PMD_TO_PTEPAGE_SHIFT)) #define pmd_page(pmd) virt_to_page(pmd_page_kernel(pmd)) -#define pgd_set(pgdp, pmdp) (pgd_val(*(pgdp)) = (__ba_to_bpn(pmdp))) -#define pgd_none(pgd) (!pgd_val(pgd)) -#define pgd_bad(pgd) ((pgd_val(pgd)) == 0) -#define pgd_present(pgd) (pgd_val(pgd) != 0UL) -#define pgd_clear(pgdp) (pgd_val(*(pgdp)) = 0UL) -#define pgd_page(pgd) (__bpn_to_ba(pgd_val(pgd))) + +#define pud_set(pudp, pmdp) (pud_val(*(pudp)) = (__ba_to_bpn(pmdp))) +#define pud_none(pud) (!pud_val(pud)) +#define pud_bad(pud) ((pud_val(pud)) == 0UL) +#define pud_present(pud) (pud_val(pud) != 0UL) +#define pud_clear(pudp) (pud_val(*(pudp)) = 0UL) +#define pud_page(pud) (__bpn_to_ba(pud_val(pud))) /* * Find an entry in a page-table-directory. We combine the address region @@ -245,12 +246,13 @@ void hugetlb_mm_free_pgd(struct mm_struct *mm); #define pgd_offset(mm, address) ((mm)->pgd + pgd_index(address)) /* Find an entry in the second-level page table.. */ -#define pmd_offset(dir,addr) \ - ((pmd_t *) pgd_page(*(dir)) + (((addr) >> PMD_SHIFT) & (PTRS_PER_PMD - 1))) +#define pmd_offset(pudp,addr) \ + ((pmd_t *) pud_page(*(pudp)) + (((addr) >> PMD_SHIFT) & (PTRS_PER_PMD - 1))) /* Find an entry in the third-level page table.. */ #define pte_offset_kernel(dir,addr) \ - ((pte_t *) pmd_page_kernel(*(dir)) + (((addr) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))) + ((pte_t *) pmd_page_kernel(*(dir)) \ + + (((addr) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))) #define pte_offset_map(dir,addr) pte_offset_kernel((dir), (addr)) #define pte_offset_map_nested(dir,addr) pte_offset_kernel((dir), (addr)) @@ -582,19 +584,22 @@ extern long native_hpte_insert(unsigned long hpte_group, unsigned long va, static inline pte_t *find_linux_pte(pgd_t *pgdir, unsigned long ea) { pgd_t *pg; + pud_t *pu; pmd_t *pm; pte_t *pt = NULL; pte_t pte; pg = pgdir + pgd_index(ea); if (!pgd_none(*pg)) { - - pm = pmd_offset(pg, ea); - if (pmd_present(*pm)) { - pt = pte_offset_kernel(pm, ea); - pte = *pt; - if (!pte_present(pte)) - pt = NULL; + pu = pud_offset(pg, ea); + if (!pud_none(*pu)) { + pm = pmd_offset(pu, ea); + if (pmd_present(*pm)) { + pt = pte_offset_kernel(pm, ea); + pte = *pt; + if (!pte_present(pte)) + pt = NULL; + } } } -- cgit v1.2.3-70-g09d2 From 66faf9845a05905d75da380767e93455f3e6d620 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Sun, 1 May 2005 08:58:45 -0700 Subject: [PATCH] ppc64: tell firmware about kernel capabilities On pSeries systems, according to the platform architecture specs, we are supposed to be supplying a structure to firmware that tells firmware about our capabilities, such as which version of the data structures that describe available memory we are expecting to see. The way we end up having to supply this data structure is a bit gross, since it was designed for AIX and doesn't suit us very well. This patch adds the code to supply this data structure to the firmware. Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/boot/addnote.c | 60 +++++++++++++++++++---- arch/ppc64/kernel/prom_init.c | 107 ++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 157 insertions(+), 10 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/boot/addnote.c b/arch/ppc64/boot/addnote.c index 66ff8103bf4..719663a694b 100644 --- a/arch/ppc64/boot/addnote.c +++ b/arch/ppc64/boot/addnote.c @@ -19,6 +19,7 @@ #include #include +/* CHRP note section */ char arch[] = "PowerPC"; #define N_DESCR 6 @@ -31,6 +32,29 @@ unsigned int descr[N_DESCR] = { 0x4000, /* load-base */ }; +/* RPA note section */ +char rpaname[] = "IBM,RPA-Client-Config"; + +/* + * Note: setting ignore_my_client_config *should* mean that OF ignores + * all the other fields, but there is a firmware bug which means that + * it looks at the splpar field at least. So these values need to be + * reasonable. + */ +#define N_RPA_DESCR 8 +unsigned int rpanote[N_RPA_DESCR] = { + 0, /* lparaffinity */ + 64, /* min_rmo_size */ + 0, /* min_rmo_percent */ + 40, /* max_pft_size */ + 1, /* splpar */ + -1, /* min_load */ + 0, /* new_mem_def */ + 1, /* ignore_my_client_config */ +}; + +#define ROUNDUP(len) (((len) + 3) & ~3) + unsigned char buf[512]; #define GET_16BE(off) ((buf[off] << 8) + (buf[(off)+1])) @@ -69,7 +93,7 @@ main(int ac, char **av) { int fd, n, i; int ph, ps, np; - int nnote, ns; + int nnote, nnote2, ns; if (ac != 2) { fprintf(stderr, "Usage: %s elf-file\n", av[0]); @@ -81,7 +105,8 @@ main(int ac, char **av) exit(1); } - nnote = strlen(arch) + 1 + (N_DESCR + 3) * 4; + nnote = 12 + ROUNDUP(strlen(arch) + 1) + sizeof(descr); + nnote2 = 12 + ROUNDUP(strlen(rpaname) + 1) + sizeof(rpanote); n = read(fd, buf, sizeof(buf)); if (n < 0) { @@ -104,7 +129,7 @@ main(int ac, char **av) np = GET_16BE(E_PHNUM); if (ph < E_HSIZE || ps < PH_HSIZE || np < 1) goto notelf; - if (ph + (np + 1) * ps + nnote > n) + if (ph + (np + 2) * ps + nnote + nnote2 > n) goto nospace; for (i = 0; i < np; ++i) { @@ -117,12 +142,12 @@ main(int ac, char **av) } /* XXX check that the area we want to use is all zeroes */ - for (i = 0; i < ps + nnote; ++i) + for (i = 0; i < 2 * ps + nnote + nnote2; ++i) if (buf[ph + i] != 0) goto nospace; /* fill in the program header entry */ - ns = ph + ps; + ns = ph + 2 * ps; PUT_32BE(ph + PH_TYPE, PT_NOTE); PUT_32BE(ph + PH_OFFSET, ns); PUT_32BE(ph + PH_FILESZ, nnote); @@ -134,11 +159,26 @@ main(int ac, char **av) PUT_32BE(ns + 8, 0x1275); strcpy(&buf[ns + 12], arch); ns += 12 + strlen(arch) + 1; - for (i = 0; i < N_DESCR; ++i) - PUT_32BE(ns + i * 4, descr[i]); + for (i = 0; i < N_DESCR; ++i, ns += 4) + PUT_32BE(ns, descr[i]); + + /* fill in the second program header entry and the RPA note area */ + ph += ps; + PUT_32BE(ph + PH_TYPE, PT_NOTE); + PUT_32BE(ph + PH_OFFSET, ns); + PUT_32BE(ph + PH_FILESZ, nnote2); + + /* fill in the note area we point to */ + PUT_32BE(ns, strlen(rpaname) + 1); + PUT_32BE(ns + 4, sizeof(rpanote)); + PUT_32BE(ns + 8, 0x12759999); + strcpy(&buf[ns + 12], rpaname); + ns += 12 + ROUNDUP(strlen(rpaname) + 1); + for (i = 0; i < N_RPA_DESCR; ++i, ns += 4) + PUT_32BE(ns, rpanote[i]); /* Update the number of program headers */ - PUT_16BE(E_PHNUM, np + 1); + PUT_16BE(E_PHNUM, np + 2); /* write back */ lseek(fd, (long) 0, SEEK_SET); @@ -155,11 +195,11 @@ main(int ac, char **av) exit(0); notelf: - fprintf(stderr, "%s does not appear to be an ELF file\n", av[0]); + fprintf(stderr, "%s does not appear to be an ELF file\n", av[1]); exit(1); nospace: fprintf(stderr, "sorry, I can't find space in %s to put the note\n", - av[0]); + av[1]); exit(1); } diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c index 8dffa9ae262..b0b784f9a4e 100644 --- a/arch/ppc64/kernel/prom_init.c +++ b/arch/ppc64/kernel/prom_init.c @@ -492,6 +492,113 @@ static void __init early_cmdline_parse(void) } } +/* + * To tell the firmware what our capabilities are, we have to pass + * it a fake 32-bit ELF header containing a couple of PT_NOTE sections + * that contain structures that contain the actual values. + */ +static struct fake_elf { + Elf32_Ehdr elfhdr; + Elf32_Phdr phdr[2]; + struct chrpnote { + u32 namesz; + u32 descsz; + u32 type; + char name[8]; /* "PowerPC" */ + struct chrpdesc { + u32 real_mode; + u32 real_base; + u32 real_size; + u32 virt_base; + u32 virt_size; + u32 load_base; + } chrpdesc; + } chrpnote; + struct rpanote { + u32 namesz; + u32 descsz; + u32 type; + char name[24]; /* "IBM,RPA-Client-Config" */ + struct rpadesc { + u32 lpar_affinity; + u32 min_rmo_size; + u32 min_rmo_percent; + u32 max_pft_size; + u32 splpar; + u32 min_load; + u32 new_mem_def; + u32 ignore_me; + } rpadesc; + } rpanote; +} fake_elf = { + .elfhdr = { + .e_ident = { 0x7f, 'E', 'L', 'F', + ELFCLASS32, ELFDATA2MSB, EV_CURRENT }, + .e_type = ET_EXEC, /* yeah right */ + .e_machine = EM_PPC, + .e_version = EV_CURRENT, + .e_phoff = offsetof(struct fake_elf, phdr), + .e_phentsize = sizeof(Elf32_Phdr), + .e_phnum = 2 + }, + .phdr = { + [0] = { + .p_type = PT_NOTE, + .p_offset = offsetof(struct fake_elf, chrpnote), + .p_filesz = sizeof(struct chrpnote) + }, [1] = { + .p_type = PT_NOTE, + .p_offset = offsetof(struct fake_elf, rpanote), + .p_filesz = sizeof(struct rpanote) + } + }, + .chrpnote = { + .namesz = sizeof("PowerPC"), + .descsz = sizeof(struct chrpdesc), + .type = 0x1275, + .name = "PowerPC", + .chrpdesc = { + .real_mode = ~0U, /* ~0 means "don't care" */ + .real_base = ~0U, + .real_size = ~0U, + .virt_base = ~0U, + .virt_size = ~0U, + .load_base = ~0U + }, + }, + .rpanote = { + .namesz = sizeof("IBM,RPA-Client-Config"), + .descsz = sizeof(struct rpadesc), + .type = 0x12759999, + .name = "IBM,RPA-Client-Config", + .rpadesc = { + .lpar_affinity = 0, + .min_rmo_size = 64, /* in megabytes */ + .min_rmo_percent = 0, + .max_pft_size = 48, /* 2^48 bytes max PFT size */ + .splpar = 1, + .min_load = ~0U, + .new_mem_def = 0 + } + } +}; + +static void __init prom_send_capabilities(void) +{ + unsigned long offset = reloc_offset(); + ihandle elfloader; + int ret; + + elfloader = call_prom("open", 1, 1, ADDR("/packages/elf-loader")); + if (elfloader == 0) { + prom_printf("couldn't open /packages/elf-loader\n"); + return; + } + ret = call_prom("call-method", 3, 1, ADDR("process-elf-header"), + elfloader, ADDR(&fake_elf)); + call_prom("close", 1, 0, elfloader); +} + /* * Memory allocation strategy... our layout is normally: * -- cgit v1.2.3-70-g09d2 From d03853d566fb32c6bb8cab4bf2ecf53e692f001c Mon Sep 17 00:00:00 2001 From: Olof Johansson Date: Sun, 1 May 2005 08:58:45 -0700 Subject: [PATCH] PPC64: Remove hot busy-wait loop in __hash_page It turns out that our current __hash_page code will do a very hot busy-wait loop waiting on _PAGE_BUSY to be cleared. It even does ldarx/stdcx in the loop, which will bounce reservations around like crazy if there's more than one CPU spinning on the same PTE (or even another PTE in the same reservation granule). The end result is that each fault takes longer when there's contention, which in turn increases the chance of another thread hitting the same fault and also piling up. Not pretty. There's two options here: 1. Do an out-of-line busy loop a'la spinlocks with just loads (no reserves) 2. Just bail and refault if needed. (2) makes sense here: If the PTE is busy, chances are it's in flux anyway and the other code path making a change might just be ready to hash it. This fixes a stampede seen on a large-ish system where a multithreaded HPC app faults in the same text pages on several cpus at the same time. Signed-off-by: Olof Johansson Cc: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/mm/hash_low.S | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ppc64/mm/hash_low.S b/arch/ppc64/mm/hash_low.S index 8c0156a3700..c23d46956dd 100644 --- a/arch/ppc64/mm/hash_low.S +++ b/arch/ppc64/mm/hash_low.S @@ -85,7 +85,10 @@ _GLOBAL(__hash_page) bne- htab_wrong_access /* Check if PTE is busy */ andi. r0,r31,_PAGE_BUSY - bne- 1b + /* If so, just bail out and refault if needed. Someone else + * is changing this PTE anyway and might hash it. + */ + bne- bail_ok /* Prepare new PTE value (turn access RW into DIRTY, then * add BUSY,HASHPTE and ACCESSED) */ @@ -215,6 +218,10 @@ _GLOBAL(htab_call_hpte_remove) /* Try all again */ b htab_insert_pte +bail_ok: + li r3,0 + b bail + htab_pte_insert_ok: /* Insert slot number & secondary bit in PTE */ rldimi r30,r3,12,63-15 -- cgit v1.2.3-70-g09d2 From c4005e4f661a9ec8d91720a6b570865d060afcaf Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 1 May 2005 08:58:46 -0700 Subject: [PATCH] ppc64: firmware workaround Recent gcc 4.0 testing uncovered a firmware issue. Some properties are larger than 31 bytes and due to gcc 4.0s better stack allocation this overflow ran over non volatile register storage. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/prom_init.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c index b0b784f9a4e..35ec42de962 100644 --- a/arch/ppc64/kernel/prom_init.c +++ b/arch/ppc64/kernel/prom_init.c @@ -1555,6 +1555,12 @@ static void __init scan_dt_build_strings(phandle node, unsigned long *mem_start, } } +/* + * The Open Firmware 1275 specification states properties must be 31 bytes or + * less, however not all firmwares obey this. Make it 64 bytes to be safe. + */ +#define MAX_PROPERTY_NAME 64 + static void __init scan_dt_build_struct(phandle node, unsigned long *mem_start, unsigned long *mem_end) { @@ -1564,7 +1570,7 @@ static void __init scan_dt_build_struct(phandle node, unsigned long *mem_start, unsigned long soff; unsigned char *valp; unsigned long offset = reloc_offset(); - char pname[32]; + char pname[MAX_PROPERTY_NAME]; char *path; path = RELOC(prom_scratch); -- cgit v1.2.3-70-g09d2 From eeb24de431ac8c80fd13a2c479cd0eb51b70484e Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 1 May 2005 08:58:46 -0700 Subject: [PATCH] ppc64: enforce medium thread priority in hypervisor calls Calls into the hypervisor do not raise the thread priority. Ensure we are running at medium priority upon entry to the hypervisor. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/pSeries_hvCall.S | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'arch') diff --git a/arch/ppc64/kernel/pSeries_hvCall.S b/arch/ppc64/kernel/pSeries_hvCall.S index 0715d303801..176e8da7646 100644 --- a/arch/ppc64/kernel/pSeries_hvCall.S +++ b/arch/ppc64/kernel/pSeries_hvCall.S @@ -28,6 +28,8 @@ unsigned long *out3); R10 */ _GLOBAL(plpar_hcall) + HMT_MEDIUM + mfcr r0 std r8,STK_PARM(r8)(r1) /* Save out ptrs */ @@ -53,6 +55,8 @@ _GLOBAL(plpar_hcall) /* Simple interface with no output values (other than status) */ _GLOBAL(plpar_hcall_norets) + HMT_MEDIUM + mfcr r0 stw r0,8(r1) @@ -75,6 +79,8 @@ _GLOBAL(plpar_hcall_norets) unsigned long *out1); 120(R1) */ _GLOBAL(plpar_hcall_8arg_2ret) + HMT_MEDIUM + mfcr r0 ld r11,STK_PARM(r11)(r1) /* put arg8 in R11 */ stw r0,8(r1) @@ -99,6 +105,8 @@ _GLOBAL(plpar_hcall_8arg_2ret) unsigned long *out4); 112(R1) */ _GLOBAL(plpar_hcall_4out) + HMT_MEDIUM + mfcr r0 stw r0,8(r1) -- cgit v1.2.3-70-g09d2 From 0d8d4d42f2d00eb65262b49f4edd4cf7ef4eb6fc Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sun, 1 May 2005 08:58:47 -0700 Subject: [PATCH] ppc64: use smp_mb and smp_wmb Use smp_mb and smp_wmb. In particular smp_wmb is lighter weight than wmb. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/smp.c | 12 ++++++------ arch/ppc64/kernel/time.c | 12 ++++++------ 2 files changed, 12 insertions(+), 12 deletions(-) (limited to 'arch') diff --git a/arch/ppc64/kernel/smp.c b/arch/ppc64/kernel/smp.c index 1c92da3e452..3b906cd9403 100644 --- a/arch/ppc64/kernel/smp.c +++ b/arch/ppc64/kernel/smp.c @@ -125,7 +125,7 @@ void __devinit smp_generic_kick_cpu(int nr) * the processor will continue on to secondary_start */ paca[nr].cpu_start = 1; - mb(); + smp_mb(); } #endif /* CONFIG_PPC_MULTIPLATFORM */ @@ -256,7 +256,7 @@ int smp_call_function (void (*func) (void *info), void *info, int nonatomic, } call_data = &data; - wmb(); + smp_wmb(); /* Send a message to all other CPUs and wait for them to respond */ smp_ops->message_pass(MSG_ALL_BUT_SELF, PPC_MSG_CALL_FUNCTION); @@ -431,7 +431,7 @@ int generic_cpu_enable(unsigned int cpu) /* get the target out of it's holding state */ per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; - wmb(); + smp_wmb(); while (!cpu_online(cpu)) cpu_relax(); @@ -447,7 +447,7 @@ void generic_cpu_die(unsigned int cpu) int i; for (i = 0; i < 100; i++) { - rmb(); + smp_rmb(); if (per_cpu(cpu_state, cpu) == CPU_DEAD) return; msleep(100); @@ -463,7 +463,7 @@ void generic_mach_cpu_die(void) cpu = smp_processor_id(); printk(KERN_DEBUG "CPU%d offline\n", cpu); __get_cpu_var(cpu_state) = CPU_DEAD; - wmb(); + smp_wmb(); while (__get_cpu_var(cpu_state) != CPU_UP_PREPARE) cpu_relax(); @@ -515,7 +515,7 @@ int __devinit __cpu_up(unsigned int cpu) * be written out to main store before we release * the processor. */ - mb(); + smp_mb(); /* wake up cpus */ DBG("smp: kicking cpu %d\n", cpu); diff --git a/arch/ppc64/kernel/time.c b/arch/ppc64/kernel/time.c index 77ded5a363b..772a465b49f 100644 --- a/arch/ppc64/kernel/time.c +++ b/arch/ppc64/kernel/time.c @@ -221,15 +221,15 @@ static __inline__ void timer_recalc_offset(unsigned long cur_tb) temp_varp->tb_to_xs = do_gtod.varp->tb_to_xs; temp_varp->tb_orig_stamp = new_tb_orig_stamp; temp_varp->stamp_xsec = new_stamp_xsec; - mb(); + smp_mb(); do_gtod.varp = temp_varp; do_gtod.var_idx = temp_idx; ++(systemcfg->tb_update_count); - wmb(); + smp_wmb(); systemcfg->tb_orig_stamp = new_tb_orig_stamp; systemcfg->stamp_xsec = new_stamp_xsec; - wmb(); + smp_wmb(); ++(systemcfg->tb_update_count); } @@ -648,7 +648,7 @@ void ppc_adjtimex(void) temp_varp->tb_to_xs = new_tb_to_xs; temp_varp->stamp_xsec = new_stamp_xsec; temp_varp->tb_orig_stamp = do_gtod.varp->tb_orig_stamp; - mb(); + smp_mb(); do_gtod.varp = temp_varp; do_gtod.var_idx = temp_idx; @@ -662,10 +662,10 @@ void ppc_adjtimex(void) * loops back and reads them again until this criteria is met. */ ++(systemcfg->tb_update_count); - wmb(); + smp_wmb(); systemcfg->tb_to_xs = new_tb_to_xs; systemcfg->stamp_xsec = new_stamp_xsec; - wmb(); + smp_wmb(); ++(systemcfg->tb_update_count); write_sequnlock_irqrestore( &xtime_lock, flags ); -- cgit v1.2.3-70-g09d2 From fd51f666fa591294bd7462447512666e61c56ea0 Mon Sep 17 00:00:00 2001 From: "H. J. Lu" Date: Sun, 1 May 2005 08:58:48 -0700 Subject: [PATCH] i386/x86_64 segment register access update The new i386/x86_64 assemblers no longer accept instructions for moving between a segment register and a 32bit memory location, i.e., movl (%eax),%ds movl %ds,(%eax) To generate instructions for moving between a segment register and a 16bit memory location without the 16bit operand size prefix, 0x66, mov (%eax),%ds mov %ds,(%eax) should be used. It will work with both new and old assemblers. The assembler starting from 2.16.90.0.1 will also support movw (%eax),%ds movw %ds,(%eax) without the 0x66 prefix. I am enclosing patches for 2.4 and 2.6 kernels here. The resulting kernel binaries should be unchanged as before, with old and new assemblers, if gcc never generates memory access for unsigned gsindex; asm volatile("movl %%gs,%0" : "=g" (gsindex)); If gcc does generate memory access for the code above, the upper bits in gsindex are undefined and the new assembler doesn't allow it. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/process.c | 4 ++-- arch/i386/kernel/vm86.c | 4 ++-- arch/x86_64/kernel/process.c | 12 ++++++------ include/asm-i386/system.h | 6 +++--- 4 files changed, 13 insertions(+), 13 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/process.c b/arch/i386/kernel/process.c index b2203e21acb..85bd56d4431 100644 --- a/arch/i386/kernel/process.c +++ b/arch/i386/kernel/process.c @@ -611,8 +611,8 @@ struct task_struct fastcall * __switch_to(struct task_struct *prev_p, struct tas * Save away %fs and %gs. No need to save %es and %ds, as * those are always kernel segments while inside the kernel. */ - asm volatile("movl %%fs,%0":"=m" (*(int *)&prev->fs)); - asm volatile("movl %%gs,%0":"=m" (*(int *)&prev->gs)); + asm volatile("mov %%fs,%0":"=m" (prev->fs)); + asm volatile("mov %%gs,%0":"=m" (prev->gs)); /* * Restore %fs and %gs if needed. diff --git a/arch/i386/kernel/vm86.c b/arch/i386/kernel/vm86.c index 2f3d52dacff..d16cd3738a4 100644 --- a/arch/i386/kernel/vm86.c +++ b/arch/i386/kernel/vm86.c @@ -294,8 +294,8 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk */ info->regs32->eax = 0; tsk->thread.saved_esp0 = tsk->thread.esp0; - asm volatile("movl %%fs,%0":"=m" (tsk->thread.saved_fs)); - asm volatile("movl %%gs,%0":"=m" (tsk->thread.saved_gs)); + asm volatile("mov %%fs,%0":"=m" (tsk->thread.saved_fs)); + asm volatile("mov %%gs,%0":"=m" (tsk->thread.saved_gs)); tss = &per_cpu(init_tss, get_cpu()); tsk->thread.esp0 = (unsigned long) &info->VM86_TSS_ESP0; diff --git a/arch/x86_64/kernel/process.c b/arch/x86_64/kernel/process.c index 9922d2ba24a..761b6d35e33 100644 --- a/arch/x86_64/kernel/process.c +++ b/arch/x86_64/kernel/process.c @@ -402,10 +402,10 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long rsp, p->thread.fs = me->thread.fs; p->thread.gs = me->thread.gs; - asm("movl %%gs,%0" : "=m" (p->thread.gsindex)); - asm("movl %%fs,%0" : "=m" (p->thread.fsindex)); - asm("movl %%es,%0" : "=m" (p->thread.es)); - asm("movl %%ds,%0" : "=m" (p->thread.ds)); + asm("mov %%gs,%0" : "=m" (p->thread.gsindex)); + asm("mov %%fs,%0" : "=m" (p->thread.fsindex)); + asm("mov %%es,%0" : "=m" (p->thread.es)); + asm("mov %%ds,%0" : "=m" (p->thread.ds)); if (unlikely(me->thread.io_bitmap_ptr != NULL)) { p->thread.io_bitmap_ptr = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL); @@ -468,11 +468,11 @@ struct task_struct *__switch_to(struct task_struct *prev_p, struct task_struct * * Switch DS and ES. * This won't pick up thread selector changes, but I guess that is ok. */ - asm volatile("movl %%es,%0" : "=m" (prev->es)); + asm volatile("mov %%es,%0" : "=m" (prev->es)); if (unlikely(next->es | prev->es)) loadsegment(es, next->es); - asm volatile ("movl %%ds,%0" : "=m" (prev->ds)); + asm volatile ("mov %%ds,%0" : "=m" (prev->ds)); if (unlikely(next->ds | prev->ds)) loadsegment(ds, next->ds); diff --git a/include/asm-i386/system.h b/include/asm-i386/system.h index 6f74d4c44a0..3db717a244f 100644 --- a/include/asm-i386/system.h +++ b/include/asm-i386/system.h @@ -81,7 +81,7 @@ static inline unsigned long _get_base(char * addr) #define loadsegment(seg,value) \ asm volatile("\n" \ "1:\t" \ - "movl %0,%%" #seg "\n" \ + "mov %0,%%" #seg "\n" \ "2:\n" \ ".section .fixup,\"ax\"\n" \ "3:\t" \ @@ -93,13 +93,13 @@ static inline unsigned long _get_base(char * addr) ".align 4\n\t" \ ".long 1b,3b\n" \ ".previous" \ - : :"m" (*(unsigned int *)&(value))) + : :"m" (value)) /* * Save a segment register away */ #define savesegment(seg, value) \ - asm volatile("movl %%" #seg ",%0":"=m" (*(int *)&(value))) + asm volatile("mov %%" #seg ",%0":"=m" (value)) /* * Clear and set 'TS' bit respectively -- cgit v1.2.3-70-g09d2 From 67701ae9767534534d3710664037dfde2cc04935 Mon Sep 17 00:00:00 2001 From: Jack F Vogel Date: Sun, 1 May 2005 08:58:48 -0700 Subject: [PATCH] check nmi watchdog is broken A bug against an xSeries system showed up recently noting that the check_nmi_watchdog() test was failing. I have been investigating it and discovered in both i386 and x86_64 the recent change to the routine to use the cpu_callin_map has uncovered a problem. Prior to that change, on an SMP box, the test was trivally passing because all cpu's were found to not yet be online, but now with the callin_map they are discovered, it goes on to test the counter and they have not yet begun to increment, so it announces a CPU is stuck and bails out. On all the systems I have access to test, the announcement of failure is also bougs... by the time you can login and check /proc/interrupts, the NMI count is happily incrementing on all CPUs. Its just that the test is being done too early. I have tried moving the call to the test around a bit, and it was always too early. I finally hit on this proposed solution, it delays the routine via a late_initcall(), seems like the right solution to me. Signed-off-by: Adrian Bunk Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/apic.c | 2 -- arch/i386/kernel/io_apic.c | 2 -- arch/i386/kernel/nmi.c | 11 +++++++---- arch/i386/kernel/smpboot.c | 3 --- arch/x86_64/kernel/io_apic.c | 2 -- arch/x86_64/kernel/nmi.c | 9 +++++++-- include/asm-i386/apic.h | 1 - include/asm-x86_64/apic.h | 1 - 8 files changed, 14 insertions(+), 17 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/apic.c b/arch/i386/kernel/apic.c index e3879f7625c..d509836b70c 100644 --- a/arch/i386/kernel/apic.c +++ b/arch/i386/kernel/apic.c @@ -1265,8 +1265,6 @@ int __init APIC_init_uniprocessor (void) setup_local_APIC(); - if (nmi_watchdog == NMI_LOCAL_APIC) - check_nmi_watchdog(); #ifdef CONFIG_X86_IO_APIC if (smp_found_config) if (!skip_ioapic_setup && nr_ioapics) diff --git a/arch/i386/kernel/io_apic.c b/arch/i386/kernel/io_apic.c index 5e0d55be543..7a324e8b86f 100644 --- a/arch/i386/kernel/io_apic.c +++ b/arch/i386/kernel/io_apic.c @@ -2175,7 +2175,6 @@ static inline void check_timer(void) disable_8259A_irq(0); setup_nmi(); enable_8259A_irq(0); - check_nmi_watchdog(); } return; } @@ -2198,7 +2197,6 @@ static inline void check_timer(void) add_pin_to_irq(0, 0, pin2); if (nmi_watchdog == NMI_IO_APIC) { setup_nmi(); - check_nmi_watchdog(); } return; } diff --git a/arch/i386/kernel/nmi.c b/arch/i386/kernel/nmi.c index 2f89d000f95..2c0ee9c2d02 100644 --- a/arch/i386/kernel/nmi.c +++ b/arch/i386/kernel/nmi.c @@ -102,20 +102,21 @@ int nmi_active; (P4_CCCR_OVF_PMI0|P4_CCCR_THRESHOLD(15)|P4_CCCR_COMPLEMENT| \ P4_CCCR_COMPARE|P4_CCCR_REQUIRED|P4_CCCR_ESCR_SELECT(4)|P4_CCCR_ENABLE) -int __init check_nmi_watchdog (void) +static int __init check_nmi_watchdog(void) { unsigned int prev_nmi_count[NR_CPUS]; int cpu; - printk(KERN_INFO "testing NMI watchdog ... "); + if (nmi_watchdog == NMI_NONE) + return 0; + + printk(KERN_INFO "Testing NMI watchdog ... "); for (cpu = 0; cpu < NR_CPUS; cpu++) prev_nmi_count[cpu] = per_cpu(irq_stat, cpu).__nmi_count; local_irq_enable(); mdelay((10*1000)/nmi_hz); // wait 10 ticks - /* FIXME: Only boot CPU is online at this stage. Check CPUs - as they come up. */ for (cpu = 0; cpu < NR_CPUS; cpu++) { #ifdef CONFIG_SMP /* Check cpu_callin_map here because that is set @@ -139,6 +140,8 @@ int __init check_nmi_watchdog (void) return 0; } +/* This needs to happen later in boot so counters are working */ +late_initcall(check_nmi_watchdog); static int __init setup_nmi_watchdog(char *str) { diff --git a/arch/i386/kernel/smpboot.c b/arch/i386/kernel/smpboot.c index fd36d2f65f8..cbea7ac582e 100644 --- a/arch/i386/kernel/smpboot.c +++ b/arch/i386/kernel/smpboot.c @@ -1089,9 +1089,6 @@ static void __init smp_boot_cpus(unsigned int max_cpus) } } - if (nmi_watchdog == NMI_LOCAL_APIC) - check_nmi_watchdog(); - smpboot_setup_io_apic(); setup_boot_APIC_clock(); diff --git a/arch/x86_64/kernel/io_apic.c b/arch/x86_64/kernel/io_apic.c index 29a25729548..60be58617eb 100644 --- a/arch/x86_64/kernel/io_apic.c +++ b/arch/x86_64/kernel/io_apic.c @@ -1607,7 +1607,6 @@ static inline void check_timer(void) disable_8259A_irq(0); setup_nmi(); enable_8259A_irq(0); - check_nmi_watchdog(); } return; } @@ -1627,7 +1626,6 @@ static inline void check_timer(void) nmi_watchdog_default(); if (nmi_watchdog == NMI_IO_APIC) { setup_nmi(); - check_nmi_watchdog(); } return; } diff --git a/arch/x86_64/kernel/nmi.c b/arch/x86_64/kernel/nmi.c index e00d4adec36..61de0b34a01 100644 --- a/arch/x86_64/kernel/nmi.c +++ b/arch/x86_64/kernel/nmi.c @@ -112,17 +112,20 @@ static __init int cpu_has_lapic(void) } } -int __init check_nmi_watchdog (void) +static int __init check_nmi_watchdog (void) { int counts[NR_CPUS]; int cpu; + if (nmi_watchdog == NMI_NONE) + return 0; + if (nmi_watchdog == NMI_LOCAL_APIC && !cpu_has_lapic()) { nmi_watchdog = NMI_NONE; return -1; } - printk(KERN_INFO "testing NMI watchdog ... "); + printk(KERN_INFO "Testing NMI watchdog ... "); for (cpu = 0; cpu < NR_CPUS; cpu++) counts[cpu] = cpu_pda[cpu].__nmi_count; @@ -148,6 +151,8 @@ int __init check_nmi_watchdog (void) return 0; } +/* Have this called later during boot so counters are updating */ +late_initcall(check_nmi_watchdog); int __init setup_nmi_watchdog(char *str) { diff --git a/include/asm-i386/apic.h b/include/asm-i386/apic.h index e1de67483f3..a5810cf7b57 100644 --- a/include/asm-i386/apic.h +++ b/include/asm-i386/apic.h @@ -109,7 +109,6 @@ extern int APIC_init_uniprocessor (void); extern void disable_APIC_timer(void); extern void enable_APIC_timer(void); -extern int check_nmi_watchdog (void); extern void enable_NMI_through_LVT0 (void * dummy); extern unsigned int nmi_watchdog; diff --git a/include/asm-x86_64/apic.h b/include/asm-x86_64/apic.h index c025cc3ef78..e4b1017b8b2 100644 --- a/include/asm-x86_64/apic.h +++ b/include/asm-x86_64/apic.h @@ -99,7 +99,6 @@ extern void disable_APIC_timer(void); extern void enable_APIC_timer(void); extern void clustered_apic_check(void); -extern int check_nmi_watchdog(void); extern void nmi_watchdog_default(void); extern int setup_nmi_watchdog(char *); -- cgit v1.2.3-70-g09d2 From a2f7c354159b87dfbd9900f597d48d18755a9d16 Mon Sep 17 00:00:00 2001 From: Jaya Kumar Date: Sun, 1 May 2005 08:58:49 -0700 Subject: [PATCH] x86 reboot: Add reboot fixup for gx1/cs5530a This patch by Jaya Kumar introduces a generic infrastructure to deal with x86 chipsets with nonstandard reset sequences, and adds support for the Geode gx1/cs5530a chipset. Signed-off-by: Jaya Kumar Signed-off-by: H. Peter Anvin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/Kconfig | 18 +++++++++++++ arch/i386/kernel/Makefile | 1 + arch/i386/kernel/reboot.c | 2 ++ arch/i386/kernel/reboot_fixups.c | 56 ++++++++++++++++++++++++++++++++++++++++ include/linux/reboot_fixups.h | 10 +++++++ 5 files changed, 87 insertions(+) create mode 100644 arch/i386/kernel/reboot_fixups.c create mode 100644 include/linux/reboot_fixups.h (limited to 'arch') diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig index 17a0cbce6f3..99b4f294a52 100644 --- a/arch/i386/Kconfig +++ b/arch/i386/Kconfig @@ -653,6 +653,24 @@ config I8K Say Y if you intend to run this kernel on a Dell Inspiron 8000. Say N otherwise. +config X86_REBOOTFIXUPS + bool "Enable X86 board specific fixups for reboot" + depends on X86 + default n + ---help--- + This enables chipset and/or board specific fixups to be done + in order to get reboot to work correctly. This is only needed on + some combinations of hardware and BIOS. The symptom, for which + this config is intended, is when reboot ends with a stalled/hung + system. + + Currently, the only fixup is for the Geode GX1/CS5530A/TROM2.1. + combination. + + Say Y if you want to enable the fixup. Currently, it's safe to + enable this option even if you don't need it. + Say N otherwise. + config MICROCODE tristate "/dev/cpu/microcode - Intel IA32 CPU microcode support" ---help--- diff --git a/arch/i386/kernel/Makefile b/arch/i386/kernel/Makefile index aacdae6f372..0fbcfe00dd8 100644 --- a/arch/i386/kernel/Makefile +++ b/arch/i386/kernel/Makefile @@ -23,6 +23,7 @@ obj-$(CONFIG_X86_TRAMPOLINE) += trampoline.o obj-$(CONFIG_X86_MPPARSE) += mpparse.o obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o obj-$(CONFIG_X86_IO_APIC) += io_apic.o +obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups.o obj-$(CONFIG_X86_NUMAQ) += numaq.o obj-$(CONFIG_X86_SUMMIT_NUMA) += summit.o obj-$(CONFIG_KPROBES) += kprobes.o diff --git a/arch/i386/kernel/reboot.c b/arch/i386/kernel/reboot.c index 3d7e994563d..6dc27eb70ee 100644 --- a/arch/i386/kernel/reboot.c +++ b/arch/i386/kernel/reboot.c @@ -13,6 +13,7 @@ #include #include #include "mach_reboot.h" +#include /* * Power off function, if any @@ -348,6 +349,7 @@ void machine_restart(char * __unused) /* rebooting needs to touch the page at absolute addr 0 */ *((unsigned short *)__va(0x472)) = reboot_mode; for (;;) { + mach_reboot_fixups(); /* for board specific fixups */ mach_reboot(); /* That didn't work - force a triple fault.. */ __asm__ __volatile__("lidt %0": :"m" (no_idt)); diff --git a/arch/i386/kernel/reboot_fixups.c b/arch/i386/kernel/reboot_fixups.c new file mode 100644 index 00000000000..1b183b378c2 --- /dev/null +++ b/arch/i386/kernel/reboot_fixups.c @@ -0,0 +1,56 @@ +/* + * linux/arch/i386/kernel/reboot_fixups.c + * + * This is a good place to put board specific reboot fixups. + * + * List of supported fixups: + * geode-gx1/cs5530a - Jaya Kumar + * + */ + +#include +#include + +static void cs5530a_warm_reset(struct pci_dev *dev) +{ + /* writing 1 to the reset control register, 0x44 causes the + cs5530a to perform a system warm reset */ + pci_write_config_byte(dev, 0x44, 0x1); + udelay(50); /* shouldn't get here but be safe and spin-a-while */ + return; +} + +struct device_fixup { + unsigned int vendor; + unsigned int device; + void (*reboot_fixup)(struct pci_dev *); +}; + +static struct device_fixup fixups_table[] = { +{ PCI_VENDOR_ID_CYRIX, PCI_DEVICE_ID_CYRIX_5530_LEGACY, cs5530a_warm_reset }, +}; + +/* + * we see if any fixup is available for our current hardware. if there + * is a fixup, we call it and we expect to never return from it. if we + * do return, we keep looking and then eventually fall back to the + * standard mach_reboot on return. + */ +void mach_reboot_fixups(void) +{ + struct device_fixup *cur; + struct pci_dev *dev; + int i; + + for (i=0; i < (sizeof(fixups_table)/sizeof(fixups_table[0])); i++) { + cur = &(fixups_table[i]); + dev = pci_get_device(cur->vendor, cur->device, 0); + if (!dev) + continue; + + cur->reboot_fixup(dev); + } + + printk(KERN_WARNING "No reboot fixup found for your hardware\n"); +} + diff --git a/include/linux/reboot_fixups.h b/include/linux/reboot_fixups.h new file mode 100644 index 00000000000..480ea2d489d --- /dev/null +++ b/include/linux/reboot_fixups.h @@ -0,0 +1,10 @@ +#ifndef _LINUX_REBOOT_FIXUPS_H +#define _LINUX_REBOOT_FIXUPS_H + +#ifdef CONFIG_X86_REBOOTFIXUPS +extern void mach_reboot_fixups(void); +#else +#define mach_reboot_fixups() ((void)(0)) +#endif + +#endif /* _LINUX_REBOOT_FIXUPS_H */ -- cgit v1.2.3-70-g09d2 From 48c88211a65bc30f030fe228bda30fda870deb90 Mon Sep 17 00:00:00 2001 From: Stas Sergeev Date: Sun, 1 May 2005 08:58:49 -0700 Subject: [PATCH] x86: entry.S trap return fixes do_debug() and do_int3() return void. This patch fixes the CONFIG_KPROBES variant of do_int3() to return void too and adjusts entry.S accordingly. Signed-off-by: Stas Sergeev Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/entry.S | 4 ---- arch/i386/kernel/traps.c | 5 ++--- 2 files changed, 2 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S index fe1918cc68d..25bf7589e0c 100644 --- a/arch/i386/kernel/entry.S +++ b/arch/i386/kernel/entry.S @@ -514,8 +514,6 @@ debug_stack_correct: xorl %edx,%edx # error code 0 movl %esp,%eax # pt_regs pointer call do_debug - testl %eax,%eax - jnz restore_all jmp ret_from_exception /* @@ -596,8 +594,6 @@ ENTRY(int3) xorl %edx,%edx # zero error code movl %esp,%eax # pt_regs pointer call do_int3 - testl %eax,%eax - jnz restore_all jmp ret_from_exception ENTRY(overflow) diff --git a/arch/i386/kernel/traps.c b/arch/i386/kernel/traps.c index d70819481f6..00c63419c06 100644 --- a/arch/i386/kernel/traps.c +++ b/arch/i386/kernel/traps.c @@ -643,16 +643,15 @@ void unset_nmi_callback(void) } #ifdef CONFIG_KPROBES -fastcall int do_int3(struct pt_regs *regs, long error_code) +fastcall void do_int3(struct pt_regs *regs, long error_code) { if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) - return 1; + return; /* This is an interrupt gate, because kprobes wants interrupts disabled. Normal trap handlers don't. */ restore_interrupts(regs); do_trap(3, SIGTRAP, "int3", 1, regs, error_code, NULL); - return 0; } #endif -- cgit v1.2.3-70-g09d2 From a6954ba2e8d344a07e066737827116eb7bc0fdcd Mon Sep 17 00:00:00 2001 From: Lee Revell Date: Sun, 1 May 2005 08:58:49 -0700 Subject: [PATCH] Enable write combining for server works LE rev > 6 Enable write combining for server works LE rev > 6 per http://www.ussg.iu.edu/hypermail/linux/kernel/0104.3/1007.html Signed-Off-By: Lee Revell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/cpu/mtrr/main.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/cpu/mtrr/main.c b/arch/i386/kernel/cpu/mtrr/main.c index 54999e4c55f..e1c2042b9b7 100644 --- a/arch/i386/kernel/cpu/mtrr/main.c +++ b/arch/i386/kernel/cpu/mtrr/main.c @@ -72,17 +72,21 @@ void set_mtrr_ops(struct mtrr_ops * ops) static int have_wrcomb(void) { struct pci_dev *dev; + u8 rev; if ((dev = pci_get_class(PCI_CLASS_BRIDGE_HOST << 8, NULL)) != NULL) { - /* ServerWorks LE chipsets have problems with write-combining + /* ServerWorks LE chipsets < rev 6 have problems with write-combining Don't allow it and leave room for other chipsets to be tagged */ if (dev->vendor == PCI_VENDOR_ID_SERVERWORKS && dev->device == PCI_DEVICE_ID_SERVERWORKS_LE) { - printk(KERN_INFO "mtrr: Serverworks LE detected. Write-combining disabled.\n"); - pci_dev_put(dev); - return 0; + pci_read_config_byte(dev, PCI_CLASS_REVISION, &rev); + if (rev <= 5) { + printk(KERN_INFO "mtrr: Serverworks LE rev < 6 detected. Write-combining disabled.\n"); + pci_dev_put(dev); + return 0; + } } - /* Intel 450NX errata # 23. Non ascending cachline evictions to + /* Intel 450NX errata # 23. Non ascending cacheline evictions to write combining memory may resulting in data corruption */ if (dev->vendor == PCI_VENDOR_ID_INTEL && dev->device == PCI_DEVICE_ID_INTEL_82451NX) { -- cgit v1.2.3-70-g09d2 From 5b7abc6fdcaf103f15e06c518ef0aec02a9c00e7 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Sun, 1 May 2005 08:58:49 -0700 Subject: [PATCH] CPUID bug and inconsistency fix The recent support for K8 multicore was misported from x86-64 to i386, due to an unnecessary inconsistency between the CPUID code. Sure, there is are no x86-64 VIA chips yet, but it should happen eventually. This patch fixes the i386 bug as well as makes x86-64 match i386 in the handing of the CPUID array. Signed-off-by: H. Peter Anvin Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/setup.c | 10 ++++++++-- include/asm-i386/cpufeature.h | 4 ++-- include/asm-x86_64/cpufeature.h | 14 ++++++++++---- 3 files changed, 20 insertions(+), 8 deletions(-) (limited to 'arch') diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c index b18c114c764..2959ce703df 100644 --- a/arch/x86_64/kernel/setup.c +++ b/arch/x86_64/kernel/setup.c @@ -977,7 +977,7 @@ void __init identify_cpu(struct cpuinfo_x86 *c) if ((xlvl & 0xffff0000) == 0x80000000) { if (xlvl >= 0x80000001) { c->x86_capability[1] = cpuid_edx(0x80000001); - c->x86_capability[5] = cpuid_ecx(0x80000001); + c->x86_capability[6] = cpuid_ecx(0x80000001); } if (xlvl >= 0x80000004) get_model_name(c); /* Default name */ @@ -1100,11 +1100,17 @@ static int show_cpuinfo(struct seq_file *m, void *v) NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, + /* VIA/Cyrix/Centaur-defined */ + NULL, NULL, "rng", "rng_en", NULL, NULL, "ace", "ace_en", + NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, + NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, + NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, + /* AMD-defined (#2) */ "lahf_lm", "cmp_legacy", NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, - NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL + NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, }; static char *x86_power_flags[] = { "ts", /* temperature sensor */ diff --git a/include/asm-i386/cpufeature.h b/include/asm-i386/cpufeature.h index e147cabd3bf..ff1187e80c3 100644 --- a/include/asm-i386/cpufeature.h +++ b/include/asm-i386/cpufeature.h @@ -87,8 +87,8 @@ #define X86_FEATURE_XCRYPT_EN (5*32+ 7) /* on-CPU crypto enabled */ /* More extended AMD flags: CPUID level 0x80000001, ecx, word 6 */ -#define X86_FEATURE_LAHF_LM (5*32+ 0) /* LAHF/SAHF in long mode */ -#define X86_FEATURE_CMP_LEGACY (5*32+ 1) /* If yes HyperThreading not valid */ +#define X86_FEATURE_LAHF_LM (6*32+ 0) /* LAHF/SAHF in long mode */ +#define X86_FEATURE_CMP_LEGACY (6*32+ 1) /* If yes HyperThreading not valid */ #define cpu_has(c, bit) test_bit(bit, (c)->x86_capability) #define boot_cpu_has(bit) test_bit(bit, boot_cpu_data.x86_capability) diff --git a/include/asm-x86_64/cpufeature.h b/include/asm-x86_64/cpufeature.h index e68ad97a631..aea308c6570 100644 --- a/include/asm-x86_64/cpufeature.h +++ b/include/asm-x86_64/cpufeature.h @@ -7,7 +7,7 @@ #ifndef __ASM_X8664_CPUFEATURE_H #define __ASM_X8664_CPUFEATURE_H -#define NCAPINTS 6 +#define NCAPINTS 7 /* N 32-bit words worth of info */ /* Intel-defined CPU features, CPUID level 0x00000001, word 0 */ #define X86_FEATURE_FPU (0*32+ 0) /* Onboard FPU */ @@ -74,9 +74,15 @@ #define X86_FEATURE_CX16 (4*32+13) /* CMPXCHG16B */ #define X86_FEATURE_XTPR (4*32+14) /* Send Task Priority Messages */ -/* More extended AMD flags: CPUID level 0x80000001, ecx, word 5 */ -#define X86_FEATURE_LAHF_LM (5*32+ 0) /* LAHF/SAHF in long mode */ -#define X86_FEATURE_CMP_LEGACY (5*32+ 1) /* If yes HyperThreading not valid */ +/* VIA/Cyrix/Centaur-defined CPU features, CPUID level 0xC0000001, word 5 */ +#define X86_FEATURE_XSTORE (5*32+ 2) /* on-CPU RNG present (xstore insn) */ +#define X86_FEATURE_XSTORE_EN (5*32+ 3) /* on-CPU RNG enabled */ +#define X86_FEATURE_XCRYPT (5*32+ 6) /* on-CPU crypto (xcrypt insn) */ +#define X86_FEATURE_XCRYPT_EN (5*32+ 7) /* on-CPU crypto enabled */ + +/* More extended AMD flags: CPUID level 0x80000001, ecx, word 6 */ +#define X86_FEATURE_LAHF_LM (6*32+ 0) /* LAHF/SAHF in long mode */ +#define X86_FEATURE_CMP_LEGACY (6*32+ 1) /* If yes HyperThreading not valid */ #define cpu_has(c, bit) test_bit(bit, (c)->x86_capability) #define boot_cpu_has(bit) test_bit(bit, boot_cpu_data.x86_capability) -- cgit v1.2.3-70-g09d2 From 35492df5ae0f36f717448b2aea908d3a8891d1c4 Mon Sep 17 00:00:00 2001 From: john stultz Date: Sun, 1 May 2005 08:58:50 -0700 Subject: [PATCH] i386: fix hpet for systems that don't support legacy replacement Currently the i386 HPET code assumes the entire HPET implementation from the spec is present. This breaks on boxes that do not implement the optional legacy timer replacement functionality portion of the spec. This patch, which is very similar to my x86-64 patch for the same issue, fixes the problem allowing i386 systems that cannot use the HPET for the timer interrupt and RTC to still use the HPET as a time source. I've tested this patch on a system systems without HPET, with HPET but without legacy timer replacement, as well as HPET with legacy timer replacement. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/time.c | 2 +- arch/i386/kernel/time_hpet.c | 48 ++++++++++++++++++++---------------- arch/i386/kernel/timers/timer_hpet.c | 11 ++++++--- arch/i386/kernel/timers/timer_tsc.c | 2 +- include/asm-i386/hpet.h | 1 + 5 files changed, 37 insertions(+), 27 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/time.c b/arch/i386/kernel/time.c index 4d75b373f90..a0dcb7c87c3 100644 --- a/arch/i386/kernel/time.c +++ b/arch/i386/kernel/time.c @@ -441,7 +441,7 @@ static void __init hpet_time_init(void) set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); - if (hpet_enable() >= 0) { + if ((hpet_enable() >= 0) && hpet_use_timer) { printk("Using HPET for base-timer\n"); } diff --git a/arch/i386/kernel/time_hpet.c b/arch/i386/kernel/time_hpet.c index 244a31b04be..10a0cbb88e7 100644 --- a/arch/i386/kernel/time_hpet.c +++ b/arch/i386/kernel/time_hpet.c @@ -26,6 +26,7 @@ static unsigned long hpet_period; /* fsecs / HPET clock */ unsigned long hpet_tick; /* hpet clks count per tick */ unsigned long hpet_address; /* hpet memory map physical address */ +int hpet_use_timer; static int use_hpet; /* can be used for runtime check of hpet */ static int boot_hpet_disable; /* boottime override for HPET timer */ @@ -73,27 +74,30 @@ static int hpet_timer_stop_set_go(unsigned long tick) hpet_writel(0, HPET_COUNTER); hpet_writel(0, HPET_COUNTER + 4); - /* - * Set up timer 0, as periodic with first interrupt to happen at - * hpet_tick, and period also hpet_tick. - */ - cfg = hpet_readl(HPET_T0_CFG); - cfg |= HPET_TN_ENABLE | HPET_TN_PERIODIC | - HPET_TN_SETVAL | HPET_TN_32BIT; - hpet_writel(cfg, HPET_T0_CFG); - - /* - * The first write after writing TN_SETVAL to the config register sets - * the counter value, the second write sets the threshold. - */ - hpet_writel(tick, HPET_T0_CMP); - hpet_writel(tick, HPET_T0_CMP); + if (hpet_use_timer) { + /* + * Set up timer 0, as periodic with first interrupt to happen at + * hpet_tick, and period also hpet_tick. + */ + cfg = hpet_readl(HPET_T0_CFG); + cfg |= HPET_TN_ENABLE | HPET_TN_PERIODIC | + HPET_TN_SETVAL | HPET_TN_32BIT; + hpet_writel(cfg, HPET_T0_CFG); + /* + * The first write after writing TN_SETVAL to the config register sets + * the counter value, the second write sets the threshold. + */ + hpet_writel(tick, HPET_T0_CMP); + hpet_writel(tick, HPET_T0_CMP); + } /* * Go! */ cfg = hpet_readl(HPET_CFG); - cfg |= HPET_CFG_ENABLE | HPET_CFG_LEGACY; + if (hpet_use_timer) + cfg |= HPET_CFG_LEGACY; + cfg |= HPET_CFG_ENABLE; hpet_writel(cfg, HPET_CFG); return 0; @@ -128,12 +132,11 @@ int __init hpet_enable(void) * However, we can do with one timer otherwise using the * the single HPET timer for system time. */ - if ( #ifdef CONFIG_HPET_EMULATE_RTC - !(id & HPET_ID_NUMBER) || -#endif - !(id & HPET_ID_LEGSUP)) + if (!(id & HPET_ID_NUMBER)) return -1; +#endif + hpet_period = hpet_readl(HPET_PERIOD); if ((hpet_period < HPET_MIN_PERIOD) || (hpet_period > HPET_MAX_PERIOD)) @@ -152,6 +155,8 @@ int __init hpet_enable(void) if (hpet_tick_rem > (hpet_period >> 1)) hpet_tick++; /* rounding the result */ + hpet_use_timer = id & HPET_ID_LEGSUP; + if (hpet_timer_stop_set_go(hpet_tick)) return -1; @@ -202,7 +207,8 @@ int __init hpet_enable(void) #endif #ifdef CONFIG_X86_LOCAL_APIC - wait_timer_tick = wait_hpet_tick; + if (hpet_use_timer) + wait_timer_tick = wait_hpet_tick; #endif return 0; } diff --git a/arch/i386/kernel/timers/timer_hpet.c b/arch/i386/kernel/timers/timer_hpet.c index 713134e7184..f778f471a09 100644 --- a/arch/i386/kernel/timers/timer_hpet.c +++ b/arch/i386/kernel/timers/timer_hpet.c @@ -79,7 +79,7 @@ static unsigned long get_offset_hpet(void) eax = hpet_readl(HPET_COUNTER); eax -= hpet_last; /* hpet delta */ - + eax = min(hpet_tick, eax); /* * Time offset = (hpet delta) * ( usecs per HPET clock ) * = (hpet delta) * ( usecs per tick / HPET clocks per tick) @@ -105,9 +105,12 @@ static void mark_offset_hpet(void) last_offset = ((unsigned long long)last_tsc_high<<32)|last_tsc_low; rdtsc(last_tsc_low, last_tsc_high); - offset = hpet_readl(HPET_T0_CMP) - hpet_tick; - if (unlikely(((offset - hpet_last) > hpet_tick) && (hpet_last != 0))) { - int lost_ticks = (offset - hpet_last) / hpet_tick; + if (hpet_use_timer) + offset = hpet_readl(HPET_T0_CMP) - hpet_tick; + else + offset = hpet_readl(HPET_COUNTER); + if (unlikely(((offset - hpet_last) >= (2*hpet_tick)) && (hpet_last != 0))) { + int lost_ticks = ((offset - hpet_last) / hpet_tick) - 1; jiffies_64 += lost_ticks; } hpet_last = offset; diff --git a/arch/i386/kernel/timers/timer_tsc.c b/arch/i386/kernel/timers/timer_tsc.c index a685994e5c8..7926d967be0 100644 --- a/arch/i386/kernel/timers/timer_tsc.c +++ b/arch/i386/kernel/timers/timer_tsc.c @@ -477,7 +477,7 @@ static int __init init_tsc(char* override) if (cpu_has_tsc) { unsigned long tsc_quotient; #ifdef CONFIG_HPET_TIMER - if (is_hpet_enabled()){ + if (is_hpet_enabled() && hpet_use_timer) { unsigned long result, remain; printk("Using TSC for gettimeofday\n"); tsc_quotient = calibrate_tsc_hpet(NULL); diff --git a/include/asm-i386/hpet.h b/include/asm-i386/hpet.h index 6e20b079f1d..16ef9f996e3 100644 --- a/include/asm-i386/hpet.h +++ b/include/asm-i386/hpet.h @@ -92,6 +92,7 @@ extern unsigned long hpet_tick; /* hpet clks count per tick */ extern unsigned long hpet_address; /* hpet memory map physical address */ +extern int hpet_use_timer; extern int hpet_rtc_timer_init(void); extern int hpet_enable(void); -- cgit v1.2.3-70-g09d2 From 4d24a439a6b2280357d62fb30a73350cf253bdb7 Mon Sep 17 00:00:00 2001 From: Jason Gaston Date: Sun, 1 May 2005 08:58:50 -0700 Subject: [PATCH] irq and pci_ids for Intel ICH7DH & ICH7-M DH MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch adds the Intel ICH7DH and ICH7-M DH DID's to the irq.c and pci_ids.h files. Signed-off-by:  Jason Gaston Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/pci/irq.c | 2 ++ include/linux/pci_ids.h | 2 ++ 2 files changed, 4 insertions(+) (limited to 'arch') diff --git a/arch/i386/pci/irq.c b/arch/i386/pci/irq.c index e07589d04f6..d6598da4b67 100644 --- a/arch/i386/pci/irq.c +++ b/arch/i386/pci/irq.c @@ -495,6 +495,8 @@ static __init int intel_router_probe(struct irq_router *r, struct pci_dev *route case PCI_DEVICE_ID_INTEL_ICH6_1: case PCI_DEVICE_ID_INTEL_ICH7_0: case PCI_DEVICE_ID_INTEL_ICH7_1: + case PCI_DEVICE_ID_INTEL_ICH7_30: + case PCI_DEVICE_ID_INTEL_ICH7_31: case PCI_DEVICE_ID_INTEL_ESB2_0: r->name = "PIIX/ICH"; r->get = pirq_piix_get; diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 6a189748194..5d5820a4cf1 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -2415,6 +2415,8 @@ #define PCI_DEVICE_ID_INTEL_ICH7_1 0x27b9 #define PCI_DEVICE_ID_INTEL_ICH7_2 0x27c0 #define PCI_DEVICE_ID_INTEL_ICH7_3 0x27c1 +#define PCI_DEVICE_ID_INTEL_ICH7_30 0x27b0 +#define PCI_DEVICE_ID_INTEL_ICH7_31 0x27bd #define PCI_DEVICE_ID_INTEL_ICH7_5 0x27c4 #define PCI_DEVICE_ID_INTEL_ICH7_6 0x27c5 #define PCI_DEVICE_ID_INTEL_ICH7_7 0x27c8 -- cgit v1.2.3-70-g09d2 From 3c3b73b6f50a47601b1897deb9c6b4b5ecadf591 Mon Sep 17 00:00:00 2001 From: Zwane Mwaikambo Date: Sun, 1 May 2005 08:58:51 -0700 Subject: [PATCH] cpuid x87 bit on AMD falsely marked as PNI http://bugme.osdl.org/show_bug.cgi?id=4426 vendor_id : AuthenticAMD cpu family : 6 model : 10 model name : AMD Athlon(tm) XP stepping : 0 cpu MHz : 2204.807 cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse pni syscall mmxext 3dnowext 3dnow bogomips : 4358.14 We're marking bit 0 of extended function 0x80000001 cpuid as PNI support on AMD processors, when it actually denotes x87 FPU present. Patch for i386 and x86_64 below. Signed-off-by: Zwane Mwaikambo Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/cpu/proc.c | 2 +- arch/x86_64/kernel/setup.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/cpu/proc.c b/arch/i386/kernel/cpu/proc.c index 4f28eba7fb8..7323c19f354 100644 --- a/arch/i386/kernel/cpu/proc.c +++ b/arch/i386/kernel/cpu/proc.c @@ -25,7 +25,7 @@ static int show_cpuinfo(struct seq_file *m, void *v) "fxsr", "sse", "sse2", "ss", "ht", "tm", "ia64", "pbe", /* AMD-defined */ - "pni", NULL, NULL, NULL, NULL, NULL, NULL, NULL, + NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, "syscall", NULL, NULL, NULL, NULL, NULL, NULL, NULL, "mp", "nx", NULL, "mmxext", NULL, NULL, "fxsr_opt", NULL, NULL, NULL, "lm", "3dnowext", "3dnow", diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c index 2959ce703df..2129cf9ba6b 100644 --- a/arch/x86_64/kernel/setup.c +++ b/arch/x86_64/kernel/setup.c @@ -1076,7 +1076,7 @@ static int show_cpuinfo(struct seq_file *m, void *v) "fxsr", "sse", "sse2", "ss", "ht", "tm", "ia64", NULL, /* AMD-defined */ - "pni", NULL, NULL, NULL, NULL, NULL, NULL, NULL, + NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, "syscall", NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, "nx", NULL, "mmxext", NULL, NULL, "fxsr_opt", NULL, NULL, NULL, "lm", "3dnowext", "3dnow", -- cgit v1.2.3-70-g09d2 From be9e68703c53fece406bc2e40976a5cfe2a8e77f Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Sun, 1 May 2005 08:58:51 -0700 Subject: [PATCH] x86_64: interrupt handling fix - Initialize workmask correctly on interrupt signal handling - Readd missing cli's in the interrupt return path. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/entry.S | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86_64/kernel/entry.S b/arch/x86_64/kernel/entry.S index 3233a15cc4e..1086b5fcac2 100644 --- a/arch/x86_64/kernel/entry.S +++ b/arch/x86_64/kernel/entry.S @@ -296,6 +296,7 @@ int_very_careful: call syscall_trace_leave popq %rdi andl $~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SINGLESTEP),%edi + cli jmp int_restore_rest int_signal: @@ -307,6 +308,7 @@ int_signal: 1: movl $_TIF_NEED_RESCHED,%edi int_restore_rest: RESTORE_REST + cli jmp int_with_check CFI_ENDPROC @@ -490,7 +492,8 @@ retint_signal: call do_notify_resume RESTORE_REST cli - GET_THREAD_INFO(%rcx) + GET_THREAD_INFO(%rcx) + movl $_TIF_WORK_MASK,%edi jmp retint_check #ifdef CONFIG_PREEMPT -- cgit v1.2.3-70-g09d2 From f9ba70535dc12d9eb57d466a2ecd749e16eca866 Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Sun, 1 May 2005 08:58:51 -0700 Subject: [PATCH] Increase number of e820 entries hard limit from 32 to 128 The specifications that talk about E820 map doesn't have an upper limit on the number of e820 entries. But, today's kernel has a hard limit of 32. With increase in memory size, we are seeing the number of E820 entries reaching close to 32. Patch below bumps the number upto 128. The patch changes the location of EDDBUF in zero-page (as it comes after E820). As, EDDBUF is not used by boot loaders, this patch should not have any effect on bootloader-setup code interface. Patch covers both i386 and x86-64. Tested on: * grub booting bzImage * lilo booting bzImage with EDID info enabled * pxeboot of bzImage Side-effect: bss increases by ~ 2K and init.data increases by ~7.5K on all systems, due to increase in size of static arrays. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/i386/zero-page.txt | 6 +++--- arch/i386/boot/setup.S | 6 +++--- arch/x86_64/boot/setup.S | 6 +++--- arch/x86_64/kernel/head64.c | 4 +--- arch/x86_64/kernel/setup64.c | 3 ++- include/asm-i386/e820.h | 2 +- include/asm-i386/setup.h | 2 +- include/asm-x86_64/bootsetup.h | 3 ++- include/asm-x86_64/e820.h | 2 +- include/linux/edd.h | 2 +- 10 files changed, 18 insertions(+), 18 deletions(-) (limited to 'arch') diff --git a/Documentation/i386/zero-page.txt b/Documentation/i386/zero-page.txt index 67c053a099e..df28c741678 100644 --- a/Documentation/i386/zero-page.txt +++ b/Documentation/i386/zero-page.txt @@ -79,6 +79,6 @@ Offset Type Description 0x22c unsigned long ramdisk_max 0x230 16 bytes trampoline 0x290 - 0x2cf EDD_MBR_SIG_BUFFER (edd.S) -0x2d0 - 0x600 E820MAP -0x600 - 0x7ff EDDBUF (edd.S) for disk signature read sector -0x600 - 0x7eb EDDBUF (edd.S) for edd data +0x2d0 - 0xd00 E820MAP +0xd00 - 0xeff EDDBUF (edd.S) for disk signature read sector +0xd00 - 0xeeb EDDBUF (edd.S) for edd data diff --git a/arch/i386/boot/setup.S b/arch/i386/boot/setup.S index a934ab32bf8..caa1fde6904 100644 --- a/arch/i386/boot/setup.S +++ b/arch/i386/boot/setup.S @@ -164,7 +164,7 @@ ramdisk_max: .long (-__PAGE_OFFSET-(512 << 20)-1) & 0x7fffffff trampoline: call start_of_setup .align 16 # The offset at this point is 0x240 - .space (0x7ff-0x240+1) # E820 & EDD space (ending at 0x7ff) + .space (0xeff-0x240+1) # E820 & EDD space (ending at 0xeff) # End of setup header ##################################################### start_of_setup: @@ -333,9 +333,9 @@ jmpe820: # sizeof(e820rec). # good820: - movb (E820NR), %al # up to 32 entries + movb (E820NR), %al # up to 128 entries cmpb $E820MAX, %al - jnl bail820 + jae bail820 incb (E820NR) movw %di, %ax diff --git a/arch/x86_64/boot/setup.S b/arch/x86_64/boot/setup.S index 3e838be9dbe..75d4d2ad93b 100644 --- a/arch/x86_64/boot/setup.S +++ b/arch/x86_64/boot/setup.S @@ -160,7 +160,7 @@ ramdisk_max: .long 0xffffffff trampoline: call start_of_setup .align 16 # The offset at this point is 0x240 - .space (0x7ff-0x240+1) # E820 & EDD space (ending at 0x7ff) + .space (0xeff-0x240+1) # E820 & EDD space (ending at 0xeff) # End of setup header ##################################################### start_of_setup: @@ -412,9 +412,9 @@ jmpe820: # sizeof(e820rec). # good820: - movb (E820NR), %al # up to 32 entries + movb (E820NR), %al # up to 128 entries cmpb $E820MAX, %al - jnl bail820 + jae bail820 incb (E820NR) movw %di, %ax diff --git a/arch/x86_64/kernel/head64.c b/arch/x86_64/kernel/head64.c index 6cad46c98a2..77df2ccb08f 100644 --- a/arch/x86_64/kernel/head64.c +++ b/arch/x86_64/kernel/head64.c @@ -29,8 +29,6 @@ static void __init clear_bss(void) (unsigned long) __bss_end - (unsigned long) __bss_start); } -extern char x86_boot_params[2048]; - #define NEW_CL_POINTER 0x228 /* Relative to real mode data */ #define OLD_CL_MAGIC_ADDR 0x90020 #define OLD_CL_MAGIC 0xA33F @@ -44,7 +42,7 @@ static void __init copy_bootdata(char *real_mode_data) int new_data; char * command_line; - memcpy(x86_boot_params, real_mode_data, 2048); + memcpy(x86_boot_params, real_mode_data, BOOT_PARAM_SIZE); new_data = *(int *) (x86_boot_params + NEW_CL_POINTER); if (!new_data) { if (OLD_CL_MAGIC != * (u16 *) OLD_CL_MAGIC_ADDR) { diff --git a/arch/x86_64/kernel/setup64.c b/arch/x86_64/kernel/setup64.c index b5305b04bc4..678b7ac33b8 100644 --- a/arch/x86_64/kernel/setup64.c +++ b/arch/x86_64/kernel/setup64.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include @@ -26,7 +27,7 @@ #include #include -char x86_boot_params[2048] __initdata = {0,}; +char x86_boot_params[BOOT_PARAM_SIZE] __initdata = {0,}; cpumask_t cpu_initialized __initdata = CPU_MASK_NONE; diff --git a/include/asm-i386/e820.h b/include/asm-i386/e820.h index 5c285aee729..edf65be21a9 100644 --- a/include/asm-i386/e820.h +++ b/include/asm-i386/e820.h @@ -13,7 +13,7 @@ #define __E820_HEADER #define E820MAP 0x2d0 /* our map */ -#define E820MAX 32 /* number of entries in E820MAP */ +#define E820MAX 128 /* number of entries in E820MAP */ #define E820NR 0x1e8 /* # entries in E820MAP */ #define E820_RAM 1 diff --git a/include/asm-i386/setup.h b/include/asm-i386/setup.h index 8814b54c75d..7a32184d54b 100644 --- a/include/asm-i386/setup.h +++ b/include/asm-i386/setup.h @@ -16,7 +16,7 @@ #define MAXMEM_PFN PFN_DOWN(MAXMEM) #define MAX_NONPAE_PFN (1 << 20) -#define PARAM_SIZE 2048 +#define PARAM_SIZE 4096 #define COMMAND_LINE_SIZE 256 #define OLD_CL_MAGIC_ADDR 0x90020 diff --git a/include/asm-x86_64/bootsetup.h b/include/asm-x86_64/bootsetup.h index b570a484dc5..b829f7b534b 100644 --- a/include/asm-x86_64/bootsetup.h +++ b/include/asm-x86_64/bootsetup.h @@ -2,7 +2,8 @@ #ifndef _X86_64_BOOTSETUP_H #define _X86_64_BOOTSETUP_H 1 -extern char x86_boot_params[2048]; +#define BOOT_PARAM_SIZE 4096 +extern char x86_boot_params[BOOT_PARAM_SIZE]; /* * This is set up by the setup-routine at boot-time diff --git a/include/asm-x86_64/e820.h b/include/asm-x86_64/e820.h index 08f83a4b4f4..8e94edf0b98 100644 --- a/include/asm-x86_64/e820.h +++ b/include/asm-x86_64/e820.h @@ -14,7 +14,7 @@ #include #define E820MAP 0x2d0 /* our map */ -#define E820MAX 32 /* number of entries in E820MAP */ +#define E820MAX 128 /* number of entries in E820MAP */ #define E820NR 0x1e8 /* # entries in E820MAP */ #define E820_RAM 1 diff --git a/include/linux/edd.h b/include/linux/edd.h index c6e6747a401..162512b886f 100644 --- a/include/linux/edd.h +++ b/include/linux/edd.h @@ -32,7 +32,7 @@ #define EDDNR 0x1e9 /* addr of number of edd_info structs at EDDBUF in boot_params - treat this as 1 byte */ -#define EDDBUF 0x600 /* addr of edd_info structs in boot_params */ +#define EDDBUF 0xd00 /* addr of edd_info structs in boot_params */ #define EDDMAXNR 6 /* number of edd_info structs starting at EDDBUF */ #define EDDEXTSIZE 8 /* change these if you muck with the structures */ #define EDDPARMSIZE 74 -- cgit v1.2.3-70-g09d2 From ad6714230f2269d5d7db2cd1900fe7bfc7aa76dc Mon Sep 17 00:00:00 2001 From: Pavel Pisa Date: Sun, 1 May 2005 08:58:52 -0700 Subject: [PATCH] Linux 2.6.x VM86 interrupt emulation fixes Patch solves VM86 interrupt emulation deadlock on SMP systems. The VM86 interrupt emulation has been heavily tested and works well on UP systems after last update, but it seems to deadlock when we have used it on SMP/HT boxes now. It seems, that disable_irq() cannot be called from interrupts, because it waits until disabled interrupt handler finishes (/kernel/irq/manage.c:synchronize_irq():while(IRQ_INPROGRESS);). This blocks one CPU after another. Solved by use disable_irq_nosync. There is the second problem. If IRQ source is fast, it is possible, that interrupt is sometimes processed and re-enabled by the second CPU, before it is disabled by the first one, but negative IRQ disable depths are not allowed. The spinlocking and disabling IRQs over call to disable_irq_nosync/enable_irq is the only solution found reliable till now. Signed-off-by: Michal Sojka Signed-off-by: Pavel Pisa Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/vm86.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/vm86.c b/arch/i386/kernel/vm86.c index d16cd3738a4..d3b4c540eb6 100644 --- a/arch/i386/kernel/vm86.c +++ b/arch/i386/kernel/vm86.c @@ -717,12 +717,12 @@ static irqreturn_t irq_handler(int intno, void *dev_id, struct pt_regs * regs) irqbits |= irq_bit; if (vm86_irqs[intno].sig) send_sig(vm86_irqs[intno].sig, vm86_irqs[intno].tsk, 1); - spin_unlock_irqrestore(&irqbits_lock, flags); /* * IRQ will be re-enabled when user asks for the irq (whether * polling or as a result of the signal) */ - disable_irq(intno); + disable_irq_nosync(intno); + spin_unlock_irqrestore(&irqbits_lock, flags); return IRQ_HANDLED; out: @@ -754,17 +754,20 @@ static inline int get_and_reset_irq(int irqnumber) { int bit; unsigned long flags; + int ret = 0; if (invalid_vm86_irq(irqnumber)) return 0; if (vm86_irqs[irqnumber].tsk != current) return 0; spin_lock_irqsave(&irqbits_lock, flags); bit = irqbits & (1 << irqnumber); irqbits &= ~bit; + if (bit) { + enable_irq(irqnumber); + ret = 1; + } + spin_unlock_irqrestore(&irqbits_lock, flags); - if (!bit) - return 0; - enable_irq(irqnumber); - return 1; + return ret; } -- cgit v1.2.3-70-g09d2 From 8059b2a292d59b6d94f9e3af93bf711f161c9476 Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Sun, 1 May 2005 08:58:52 -0700 Subject: [PATCH] x86-64: Handle empty E820 regions correctly Brings sanitize_e820_map() in x86-64 in sync with that of i386. x86_64 version was missing the changes from this patch. http://linux.bkbits.net:8080/linux-2.6/cset@3e5e4083Y3HevldZl5KCy94V4DcZww?nav=index.html|src/|src/arch|src/arch/i386|src/arch/i386/kernel|related/arch/i386/kernel/setup.c Signed-off-by: Venkatesh Pallipadi Acked-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/e820.c | 26 ++++++++++++++++++-------- 1 file changed, 18 insertions(+), 8 deletions(-) (limited to 'arch') diff --git a/arch/x86_64/kernel/e820.c b/arch/x86_64/kernel/e820.c index 56516ac92e5..7c154dfff64 100644 --- a/arch/x86_64/kernel/e820.c +++ b/arch/x86_64/kernel/e820.c @@ -2,6 +2,12 @@ * Handle the memory map. * The functions here do the job until bootmem takes over. * $Id: e820.c,v 1.4 2002/09/19 19:25:32 ak Exp $ + * + * Getting sanitize_e820_map() in sync with i386 version by applying change: + * - Provisions for empty E820 memory regions (reported by certain BIOSes). + * Alex Achenbach , December 2002. + * Venkatesh Pallipadi + * */ #include #include @@ -277,7 +283,7 @@ static int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map) int chgidx, still_changing; int overlap_entries; int new_bios_entry; - int old_nr, new_nr; + int old_nr, new_nr, chg_nr; int i; /* @@ -331,20 +337,24 @@ static int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map) for (i=0; i < 2*old_nr; i++) change_point[i] = &change_point_list[i]; - /* record all known change-points (starting and ending addresses) */ + /* record all known change-points (starting and ending addresses), + omitting those that are for empty memory regions */ chgidx = 0; for (i=0; i < old_nr; i++) { - change_point[chgidx]->addr = biosmap[i].addr; - change_point[chgidx++]->pbios = &biosmap[i]; - change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size; - change_point[chgidx++]->pbios = &biosmap[i]; + if (biosmap[i].size != 0) { + change_point[chgidx]->addr = biosmap[i].addr; + change_point[chgidx++]->pbios = &biosmap[i]; + change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size; + change_point[chgidx++]->pbios = &biosmap[i]; + } } + chg_nr = chgidx; /* sort change-point list by memory addresses (low -> high) */ still_changing = 1; while (still_changing) { still_changing = 0; - for (i=1; i < 2*old_nr; i++) { + for (i=1; i < chg_nr; i++) { /* if > , swap */ /* or, if current= & last=, swap */ if ((change_point[i]->addr < change_point[i-1]->addr) || @@ -367,7 +377,7 @@ static int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map) last_type = 0; /* start with undefined memory type */ last_addr = 0; /* start with 0 as last starting address */ /* loop through change-points, determining affect on the new bios map */ - for (chgidx=0; chgidx < 2*old_nr; chgidx++) + for (chgidx=0; chgidx < chg_nr; chgidx++) { /* keep track of all overlapping bios entries */ if (change_point[chgidx]->addr == change_point[chgidx]->pbios->addr) -- cgit v1.2.3-70-g09d2 From 429e9c5eeb66fc272c15568849ad0b23689a4a06 Mon Sep 17 00:00:00 2001 From: Alexander Nyberg Date: Sun, 1 May 2005 08:58:52 -0700 Subject: [PATCH] x86_64: saved_command_line overflow fix This strcpy can run off the end of saved_command_line, and we don't need it any more anyway. Signed-off-by: Alexander Nyberg Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/head64.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'arch') diff --git a/arch/x86_64/kernel/head64.c b/arch/x86_64/kernel/head64.c index 77df2ccb08f..0f8c78dcd38 100644 --- a/arch/x86_64/kernel/head64.c +++ b/arch/x86_64/kernel/head64.c @@ -91,9 +91,6 @@ void __init x86_64_start_kernel(char * real_mode_data) #ifdef CONFIG_SMP cpu_set(0, cpu_online_map); #endif - /* default console: */ - if (!strstr(saved_command_line, "console=")) - strcat(saved_command_line, " console=tty0"); s = strstr(saved_command_line, "earlyprintk="); if (s != NULL) setup_early_printk(s); -- cgit v1.2.3-70-g09d2 From 8e7a4f69e4ed328aa09ce48601073052557fa85d Mon Sep 17 00:00:00 2001 From: "blaisorblade@yahoo.it" Date: Sun, 1 May 2005 08:58:53 -0700 Subject: [PATCH] uml: add nfsd syscall when nfsd is modular This trick is useless, because sys_ni.c will handle this problem by itself, like it does even on UML for other syscalls. Also, it does not provide the NFSD syscall when NFSD is compiled as a module, which is a big problem. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/sys_call_table.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/um/kernel/sys_call_table.c b/arch/um/kernel/sys_call_table.c index 7fc06c85b29..b671a31a006 100644 --- a/arch/um/kernel/sys_call_table.c +++ b/arch/um/kernel/sys_call_table.c @@ -14,12 +14,6 @@ #include "sysdep/syscalls.h" #include "kern_util.h" -#ifdef CONFIG_NFSD -#define NFSSERVCTL sys_nfsservctl -#else -#define NFSSERVCTL sys_ni_syscall -#endif - #define LAST_GENERIC_SYSCALL __NR_keyctl #if LAST_GENERIC_SYSCALL > LAST_ARCH_SYSCALL @@ -189,7 +183,7 @@ syscall_handler_t *sys_call_table[] = { [ __NR_getresuid ] = (syscall_handler_t *) sys_getresuid16, [ __NR_query_module ] = (syscall_handler_t *) sys_ni_syscall, [ __NR_poll ] = (syscall_handler_t *) sys_poll, - [ __NR_nfsservctl ] = (syscall_handler_t *) NFSSERVCTL, + [ __NR_nfsservctl ] = (syscall_handler_t *) sys_nfsservctl, [ __NR_setresgid ] = (syscall_handler_t *) sys_setresgid16, [ __NR_getresgid ] = (syscall_handler_t *) sys_getresgid16, [ __NR_prctl ] = (syscall_handler_t *) sys_prctl, -- cgit v1.2.3-70-g09d2 From 6aec3072833283a5e2ecf07318bd383a67321657 Mon Sep 17 00:00:00 2001 From: Andree Leidenfrost Date: Sun, 1 May 2005 08:58:53 -0700 Subject: [PATCH] uml: fix handling of no fpx_regs Paolo 'Blaisorblade' Giarrusso Fix the error path, which is triggered when the processor misses the fpx regs (i.e. the "fxsr" cpuinfo feature). For instance by VIA C3 Samuel2. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/os-Linux/sys-i386/registers.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/um/os-Linux/sys-i386/registers.c b/arch/um/os-Linux/sys-i386/registers.c index 148645b1448..9a0ad094d92 100644 --- a/arch/um/os-Linux/sys-i386/registers.c +++ b/arch/um/os-Linux/sys-i386/registers.c @@ -105,14 +105,15 @@ void init_registers(int pid) panic("check_ptrace : PTRACE_GETREGS failed, errno = %d", err); + errno = 0; err = ptrace(PTRACE_GETFPXREGS, pid, 0, exec_fpx_regs); if(!err) return; + if(errno != EIO) + panic("check_ptrace : PTRACE_GETFPXREGS failed, errno = %d", + errno); have_fpx_regs = 0; - if(err != EIO) - panic("check_ptrace : PTRACE_GETFPXREGS failed, errno = %d", - err); err = ptrace(PTRACE_GETFPREGS, pid, 0, exec_fp_regs); if(err) -- cgit v1.2.3-70-g09d2 From b05d85a87d9711f5f5f2eb05c79038d5d5ff1f44 Mon Sep 17 00:00:00 2001 From: Rob Landley Date: Sun, 1 May 2005 08:58:54 -0700 Subject: [PATCH] uml: workaround old problematic sed behaviour Old versions of sed from 1998 (predating the first release of gcc 2.95, but still in use by debian stable) don't understand the single-line version of the sed append command. Since newer versions of sed still understand the... ahem, "vintage" form of the command, change our code to use that. Signed-off-by: Rob Landley Acked-by: Ian McDonald Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/Makefile | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/um/kernel/Makefile b/arch/um/kernel/Makefile index dc796c1bf39..280d3000111 100644 --- a/arch/um/kernel/Makefile +++ b/arch/um/kernel/Makefile @@ -53,6 +53,7 @@ quiet_cmd_quote2 = QUOTE $@ cmd_quote2 = sed -e '/CONFIG/{' \ -e 's/"CONFIG"\;/""/' \ -e 'r $(obj)/config.tmp' \ - -e 'a""\;' \ + -e 'a \' \ + -e '""\;' \ -e '}' \ $< > $@ -- cgit v1.2.3-70-g09d2 From c45166be3cc666ce88fe623ad79276c943e74eff Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:54 -0700 Subject: [PATCH] uml: support AES i586 crypto driver We want to make possible, for the user, to enable the i586 AES implementation. This requires a restructure. - Add a CONFIG_UML_X86 to notify that we are building a UML for i386. - Rename CONFIG_64_BIT to CONFIG_64BIT as is used for all other archs - Tell crypto/Kconfig that UML_X86 is as good as X86 - Tell it that it must exclude not X86_64 but 64BIT, which will give the same results. - Tell kbuild to descend down into arch/i386/crypto/ to build what's needed. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/Kconfig | 1 + arch/um/Kconfig_i386 | 6 +++++- arch/um/Kconfig_x86_64 | 6 +++++- arch/um/Makefile-i386 | 2 +- arch/um/defconfig | 10 +++++----- crypto/Kconfig | 4 ++-- include/asm-um/elf.h | 2 +- include/asm-um/page.h | 2 +- include/asm-um/pgtable-3level.h | 2 +- 9 files changed, 22 insertions(+), 13 deletions(-) (limited to 'arch') diff --git a/arch/um/Kconfig b/arch/um/Kconfig index 9a23df18212..c5292181a66 100644 --- a/arch/um/Kconfig +++ b/arch/um/Kconfig @@ -244,6 +244,7 @@ config KERNEL_HALF_GIGS config HIGHMEM bool "Highmem support" + depends on !64BIT config KERNEL_STACK_ORDER int "Kernel stack size order" diff --git a/arch/um/Kconfig_i386 b/arch/um/Kconfig_i386 index 203c242201b..e41f3748d30 100644 --- a/arch/um/Kconfig_i386 +++ b/arch/um/Kconfig_i386 @@ -1,4 +1,8 @@ -config 64_BIT +config UML_X86 + bool + default y + +config 64BIT bool default n diff --git a/arch/um/Kconfig_x86_64 b/arch/um/Kconfig_x86_64 index 768dc6626a8..fd8d7e8982b 100644 --- a/arch/um/Kconfig_x86_64 +++ b/arch/um/Kconfig_x86_64 @@ -1,4 +1,8 @@ -config 64_BIT +config UML_X86 + bool + default y + +config 64BIT bool default y diff --git a/arch/um/Makefile-i386 b/arch/um/Makefile-i386 index 97b223bfa78..f9e3c0f0654 100644 --- a/arch/um/Makefile-i386 +++ b/arch/um/Makefile-i386 @@ -1,4 +1,4 @@ -SUBARCH_CORE := arch/um/sys-i386/ +SUBARCH_CORE := arch/um/sys-i386/ arch/i386/crypto/ TOP_ADDR := $(CONFIG_TOP_ADDR) diff --git a/arch/um/defconfig b/arch/um/defconfig index fc3075c589d..4067c3aa5b6 100644 --- a/arch/um/defconfig +++ b/arch/um/defconfig @@ -1,7 +1,7 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.12-rc1-bk1 -# Sun Mar 20 16:53:00 2005 +# Linux kernel version: 2.6.12-rc3-skas3-v9-pre2 +# Sun Apr 24 19:46:10 2005 # CONFIG_GENERIC_HARDIRQS=y CONFIG_UML=y @@ -15,7 +15,8 @@ CONFIG_GENERIC_CALIBRATE_DELAY=y # CONFIG_MODE_TT=y CONFIG_MODE_SKAS=y -# CONFIG_64_BIT is not set +CONFIG_UML_X86=y +# CONFIG_64BIT is not set CONFIG_TOP_ADDR=0xc0000000 # CONFIG_3_LEVEL_PGTABLES is not set CONFIG_ARCH_HAS_SC_SIGNALS=y @@ -41,6 +42,7 @@ CONFIG_UML_REAL_TIME_CLOCK=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_BROKEN_ON_SMP=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -158,7 +160,6 @@ CONFIG_UML_NET_SLIRP=y # CONFIG_PACKET=y CONFIG_PACKET_MMAP=y -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y # CONFIG_NET_KEY is not set CONFIG_INET=y @@ -412,6 +413,5 @@ CONFIG_DEBUG_INFO=y # CONFIG_DEBUG_FS is not set CONFIG_FRAME_POINTER=y CONFIG_PT_PROXY=y -# CONFIG_GPROF is not set # CONFIG_GCOV is not set # CONFIG_SYSCALL_DEBUG is not set diff --git a/crypto/Kconfig b/crypto/Kconfig index 536754faf4d..90d6089d60e 100644 --- a/crypto/Kconfig +++ b/crypto/Kconfig @@ -146,7 +146,7 @@ config CRYPTO_SERPENT config CRYPTO_AES tristate "AES cipher algorithms" - depends on CRYPTO && !(X86 && !X86_64) + depends on CRYPTO && !((X86 || UML_X86) && !64BIT) help AES cipher algorithms (FIPS-197). AES uses the Rijndael algorithm. @@ -166,7 +166,7 @@ config CRYPTO_AES config CRYPTO_AES_586 tristate "AES cipher algorithms (i586)" - depends on CRYPTO && (X86 && !X86_64) + depends on CRYPTO && ((X86 || UML_X86) && !64BIT) help AES cipher algorithms (FIPS-197). AES uses the Rijndael algorithm. diff --git a/include/asm-um/elf.h b/include/asm-um/elf.h index b3a7258f997..7908f8fe823 100644 --- a/include/asm-um/elf.h +++ b/include/asm-um/elf.h @@ -13,7 +13,7 @@ extern long elf_aux_hwcap; #define elf_check_arch(x) (1) -#ifdef CONFIG_64_BIT +#ifdef CONFIG_64BIT #define ELF_CLASS ELFCLASS64 #else #define ELF_CLASS ELFCLASS32 diff --git a/include/asm-um/page.h b/include/asm-um/page.h index 3620a08dc9f..102eb3df1aa 100644 --- a/include/asm-um/page.h +++ b/include/asm-um/page.h @@ -27,7 +27,7 @@ struct page; #define clear_user_page(page, vaddr, pg) clear_page(page) #define copy_user_page(to, from, vaddr, pg) copy_page(to, from) -#if defined(CONFIG_3_LEVEL_PGTABLES) && !defined(CONFIG_64_BIT) +#if defined(CONFIG_3_LEVEL_PGTABLES) && !defined(CONFIG_64BIT) typedef struct { unsigned long pte_low, pte_high; } pte_t; typedef struct { unsigned long long pmd; } pmd_t; diff --git a/include/asm-um/pgtable-3level.h b/include/asm-um/pgtable-3level.h index bdbc3f97e20..d309f3a9e6f 100644 --- a/include/asm-um/pgtable-3level.h +++ b/include/asm-um/pgtable-3level.h @@ -145,7 +145,7 @@ static inline pmd_t pfn_pmd(pfn_t page_nr, pgprot_t pgprot) */ #define PTE_FILE_MAX_BITS 32 -#ifdef CONFIG_64_BIT +#ifdef CONFIG_64BIT #define pte_to_pgoff(p) ((p).pte >> 32) -- cgit v1.2.3-70-g09d2 From c16993d9009b4311f0e6088af38844eabc8b5e5b Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:54 -0700 Subject: [PATCH] uml: inline empty proc Cleanup: make an inline of this empty proc. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/process_kern.c | 4 ---- include/asm-um/processor-generic.h | 6 +++++- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/um/kernel/process_kern.c b/arch/um/kernel/process_kern.c index 1d719d5b4bb..7a943696f95 100644 --- a/arch/um/kernel/process_kern.c +++ b/arch/um/kernel/process_kern.c @@ -161,10 +161,6 @@ void *get_current(void) return(current); } -void prepare_to_copy(struct task_struct *tsk) -{ -} - int copy_thread(int nr, unsigned long clone_flags, unsigned long sp, unsigned long stack_top, struct task_struct * p, struct pt_regs *regs) diff --git a/include/asm-um/processor-generic.h b/include/asm-um/processor-generic.h index 038ba6fc88b..4d9404989b5 100644 --- a/include/asm-um/processor-generic.h +++ b/include/asm-um/processor-generic.h @@ -89,7 +89,11 @@ extern struct task_struct *alloc_task_struct(void); extern void release_thread(struct task_struct *); extern int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags); extern void dump_thread(struct pt_regs *regs, struct user *u); -extern void prepare_to_copy(struct task_struct *tsk); + +static inline void prepare_to_copy(struct task_struct *tsk) +{ +} + extern unsigned long thread_saved_pc(struct task_struct *t); -- cgit v1.2.3-70-g09d2 From e9c527163d31da9f616e989a90429729525c5233 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:54 -0700 Subject: [PATCH] uml: move va_copy conditional def GCC 2.95 uses __va_copy instead of va_copy. Handle it inside compiler.h instead of in a casual file, and avoid the risk that this breaks with a newer compiler (which it could do). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/skas/uaccess.c | 4 ++-- include/linux/compiler-gcc2.h | 5 +++++ 2 files changed, 7 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/um/kernel/skas/uaccess.c b/arch/um/kernel/skas/uaccess.c index 7575ec489b6..f7da9d02767 100644 --- a/arch/um/kernel/skas/uaccess.c +++ b/arch/um/kernel/skas/uaccess.c @@ -3,6 +3,7 @@ * Licensed under the GPL */ +#include "linux/compiler.h" #include "linux/stddef.h" #include "linux/kernel.h" #include "linux/string.h" @@ -61,8 +62,7 @@ static void do_buffer_op(void *jmpbuf, void *arg_ptr) void *arg; int *res; - /* Some old gccs recognize __va_copy, but not va_copy */ - __va_copy(args, *(va_list *)arg_ptr); + va_copy(args, *(va_list *)arg_ptr); addr = va_arg(args, unsigned long); len = va_arg(args, int); is_write = va_arg(args, int); diff --git a/include/linux/compiler-gcc2.h b/include/linux/compiler-gcc2.h index 5a359153ffd..ebed17660c5 100644 --- a/include/linux/compiler-gcc2.h +++ b/include/linux/compiler-gcc2.h @@ -22,3 +22,8 @@ # define __attribute_pure__ __attribute__((pure)) # define __attribute_const__ __attribute__((__const__)) #endif + +/* GCC 2.95.x/2.96 recognize __va_copy, but not va_copy. Actually later GCC's + * define both va_copy and __va_copy, but the latter may go away, so limit this + * to this header */ +#define va_copy __va_copy -- cgit v1.2.3-70-g09d2 From 5e7b83ffc67e15791d9bf8b2a18e4f5fd0eb69b8 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:55 -0700 Subject: [PATCH] uml: fix syscall table by including $(SUBARCH)'s one, for i386 Split the i386 entry.S files into entry.S and syscall_table.S which is included in the previous one (so actually there is no difference between them) and use the syscall_table.S in the UML build, instead of tracking by hand the syscall table changes (which is inherently error-prone). We must only insert the right #defines to inject the changes we need from the i386 syscall table (for instance some different function names); also, we don't implement some i386 syscalls, as ioperm(), nor some TLS-related ones (yet to provide). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/entry.S | 292 +------------------------------ arch/i386/kernel/syscall_table.S | 291 ++++++++++++++++++++++++++++++ arch/um/include/sysdep-i386/syscalls.h | 99 ----------- arch/um/include/sysdep-x86_64/syscalls.h | 11 -- arch/um/kernel/Makefile | 2 +- arch/um/kernel/sys_call_table.c | 270 ---------------------------- arch/um/sys-i386/Makefile | 3 +- arch/um/sys-i386/sys_call_table.S | 16 ++ arch/um/sys-x86_64/Makefile | 2 +- arch/um/sys-x86_64/sys_call_table.c | 276 +++++++++++++++++++++++++++++ 10 files changed, 588 insertions(+), 674 deletions(-) create mode 100644 arch/i386/kernel/syscall_table.S delete mode 100644 arch/um/kernel/sys_call_table.c create mode 100644 arch/um/sys-i386/sys_call_table.S create mode 100644 arch/um/sys-x86_64/sys_call_table.c (limited to 'arch') diff --git a/arch/i386/kernel/entry.S b/arch/i386/kernel/entry.S index 25bf7589e0c..a991d4e5edd 100644 --- a/arch/i386/kernel/entry.S +++ b/arch/i386/kernel/entry.S @@ -652,296 +652,6 @@ ENTRY(spurious_interrupt_bug) pushl $do_spurious_interrupt_bug jmp error_code -.data -ENTRY(sys_call_table) - .long sys_restart_syscall /* 0 - old "setup()" system call, used for restarting */ - .long sys_exit - .long sys_fork - .long sys_read - .long sys_write - .long sys_open /* 5 */ - .long sys_close - .long sys_waitpid - .long sys_creat - .long sys_link - .long sys_unlink /* 10 */ - .long sys_execve - .long sys_chdir - .long sys_time - .long sys_mknod - .long sys_chmod /* 15 */ - .long sys_lchown16 - .long sys_ni_syscall /* old break syscall holder */ - .long sys_stat - .long sys_lseek - .long sys_getpid /* 20 */ - .long sys_mount - .long sys_oldumount - .long sys_setuid16 - .long sys_getuid16 - .long sys_stime /* 25 */ - .long sys_ptrace - .long sys_alarm - .long sys_fstat - .long sys_pause - .long sys_utime /* 30 */ - .long sys_ni_syscall /* old stty syscall holder */ - .long sys_ni_syscall /* old gtty syscall holder */ - .long sys_access - .long sys_nice - .long sys_ni_syscall /* 35 - old ftime syscall holder */ - .long sys_sync - .long sys_kill - .long sys_rename - .long sys_mkdir - .long sys_rmdir /* 40 */ - .long sys_dup - .long sys_pipe - .long sys_times - .long sys_ni_syscall /* old prof syscall holder */ - .long sys_brk /* 45 */ - .long sys_setgid16 - .long sys_getgid16 - .long sys_signal - .long sys_geteuid16 - .long sys_getegid16 /* 50 */ - .long sys_acct - .long sys_umount /* recycled never used phys() */ - .long sys_ni_syscall /* old lock syscall holder */ - .long sys_ioctl - .long sys_fcntl /* 55 */ - .long sys_ni_syscall /* old mpx syscall holder */ - .long sys_setpgid - .long sys_ni_syscall /* old ulimit syscall holder */ - .long sys_olduname - .long sys_umask /* 60 */ - .long sys_chroot - .long sys_ustat - .long sys_dup2 - .long sys_getppid - .long sys_getpgrp /* 65 */ - .long sys_setsid - .long sys_sigaction - .long sys_sgetmask - .long sys_ssetmask - .long sys_setreuid16 /* 70 */ - .long sys_setregid16 - .long sys_sigsuspend - .long sys_sigpending - .long sys_sethostname - .long sys_setrlimit /* 75 */ - .long sys_old_getrlimit - .long sys_getrusage - .long sys_gettimeofday - .long sys_settimeofday - .long sys_getgroups16 /* 80 */ - .long sys_setgroups16 - .long old_select - .long sys_symlink - .long sys_lstat - .long sys_readlink /* 85 */ - .long sys_uselib - .long sys_swapon - .long sys_reboot - .long old_readdir - .long old_mmap /* 90 */ - .long sys_munmap - .long sys_truncate - .long sys_ftruncate - .long sys_fchmod - .long sys_fchown16 /* 95 */ - .long sys_getpriority - .long sys_setpriority - .long sys_ni_syscall /* old profil syscall holder */ - .long sys_statfs - .long sys_fstatfs /* 100 */ - .long sys_ioperm - .long sys_socketcall - .long sys_syslog - .long sys_setitimer - .long sys_getitimer /* 105 */ - .long sys_newstat - .long sys_newlstat - .long sys_newfstat - .long sys_uname - .long sys_iopl /* 110 */ - .long sys_vhangup - .long sys_ni_syscall /* old "idle" system call */ - .long sys_vm86old - .long sys_wait4 - .long sys_swapoff /* 115 */ - .long sys_sysinfo - .long sys_ipc - .long sys_fsync - .long sys_sigreturn - .long sys_clone /* 120 */ - .long sys_setdomainname - .long sys_newuname - .long sys_modify_ldt - .long sys_adjtimex - .long sys_mprotect /* 125 */ - .long sys_sigprocmask - .long sys_ni_syscall /* old "create_module" */ - .long sys_init_module - .long sys_delete_module - .long sys_ni_syscall /* 130: old "get_kernel_syms" */ - .long sys_quotactl - .long sys_getpgid - .long sys_fchdir - .long sys_bdflush - .long sys_sysfs /* 135 */ - .long sys_personality - .long sys_ni_syscall /* reserved for afs_syscall */ - .long sys_setfsuid16 - .long sys_setfsgid16 - .long sys_llseek /* 140 */ - .long sys_getdents - .long sys_select - .long sys_flock - .long sys_msync - .long sys_readv /* 145 */ - .long sys_writev - .long sys_getsid - .long sys_fdatasync - .long sys_sysctl - .long sys_mlock /* 150 */ - .long sys_munlock - .long sys_mlockall - .long sys_munlockall - .long sys_sched_setparam - .long sys_sched_getparam /* 155 */ - .long sys_sched_setscheduler - .long sys_sched_getscheduler - .long sys_sched_yield - .long sys_sched_get_priority_max - .long sys_sched_get_priority_min /* 160 */ - .long sys_sched_rr_get_interval - .long sys_nanosleep - .long sys_mremap - .long sys_setresuid16 - .long sys_getresuid16 /* 165 */ - .long sys_vm86 - .long sys_ni_syscall /* Old sys_query_module */ - .long sys_poll - .long sys_nfsservctl - .long sys_setresgid16 /* 170 */ - .long sys_getresgid16 - .long sys_prctl - .long sys_rt_sigreturn - .long sys_rt_sigaction - .long sys_rt_sigprocmask /* 175 */ - .long sys_rt_sigpending - .long sys_rt_sigtimedwait - .long sys_rt_sigqueueinfo - .long sys_rt_sigsuspend - .long sys_pread64 /* 180 */ - .long sys_pwrite64 - .long sys_chown16 - .long sys_getcwd - .long sys_capget - .long sys_capset /* 185 */ - .long sys_sigaltstack - .long sys_sendfile - .long sys_ni_syscall /* reserved for streams1 */ - .long sys_ni_syscall /* reserved for streams2 */ - .long sys_vfork /* 190 */ - .long sys_getrlimit - .long sys_mmap2 - .long sys_truncate64 - .long sys_ftruncate64 - .long sys_stat64 /* 195 */ - .long sys_lstat64 - .long sys_fstat64 - .long sys_lchown - .long sys_getuid - .long sys_getgid /* 200 */ - .long sys_geteuid - .long sys_getegid - .long sys_setreuid - .long sys_setregid - .long sys_getgroups /* 205 */ - .long sys_setgroups - .long sys_fchown - .long sys_setresuid - .long sys_getresuid - .long sys_setresgid /* 210 */ - .long sys_getresgid - .long sys_chown - .long sys_setuid - .long sys_setgid - .long sys_setfsuid /* 215 */ - .long sys_setfsgid - .long sys_pivot_root - .long sys_mincore - .long sys_madvise - .long sys_getdents64 /* 220 */ - .long sys_fcntl64 - .long sys_ni_syscall /* reserved for TUX */ - .long sys_ni_syscall - .long sys_gettid - .long sys_readahead /* 225 */ - .long sys_setxattr - .long sys_lsetxattr - .long sys_fsetxattr - .long sys_getxattr - .long sys_lgetxattr /* 230 */ - .long sys_fgetxattr - .long sys_listxattr - .long sys_llistxattr - .long sys_flistxattr - .long sys_removexattr /* 235 */ - .long sys_lremovexattr - .long sys_fremovexattr - .long sys_tkill - .long sys_sendfile64 - .long sys_futex /* 240 */ - .long sys_sched_setaffinity - .long sys_sched_getaffinity - .long sys_set_thread_area - .long sys_get_thread_area - .long sys_io_setup /* 245 */ - .long sys_io_destroy - .long sys_io_getevents - .long sys_io_submit - .long sys_io_cancel - .long sys_fadvise64 /* 250 */ - .long sys_ni_syscall - .long sys_exit_group - .long sys_lookup_dcookie - .long sys_epoll_create - .long sys_epoll_ctl /* 255 */ - .long sys_epoll_wait - .long sys_remap_file_pages - .long sys_set_tid_address - .long sys_timer_create - .long sys_timer_settime /* 260 */ - .long sys_timer_gettime - .long sys_timer_getoverrun - .long sys_timer_delete - .long sys_clock_settime - .long sys_clock_gettime /* 265 */ - .long sys_clock_getres - .long sys_clock_nanosleep - .long sys_statfs64 - .long sys_fstatfs64 - .long sys_tgkill /* 270 */ - .long sys_utimes - .long sys_fadvise64_64 - .long sys_ni_syscall /* sys_vserver */ - .long sys_mbind - .long sys_get_mempolicy - .long sys_set_mempolicy - .long sys_mq_open - .long sys_mq_unlink - .long sys_mq_timedsend - .long sys_mq_timedreceive /* 280 */ - .long sys_mq_notify - .long sys_mq_getsetattr - .long sys_ni_syscall /* reserved for kexec */ - .long sys_waitid - .long sys_ni_syscall /* 285 */ /* available */ - .long sys_add_key - .long sys_request_key - .long sys_keyctl +#include "syscall_table.S" syscall_table_size=(.-sys_call_table) diff --git a/arch/i386/kernel/syscall_table.S b/arch/i386/kernel/syscall_table.S new file mode 100644 index 00000000000..6cd1ed311f0 --- /dev/null +++ b/arch/i386/kernel/syscall_table.S @@ -0,0 +1,291 @@ +.data +ENTRY(sys_call_table) + .long sys_restart_syscall /* 0 - old "setup()" system call, used for restarting */ + .long sys_exit + .long sys_fork + .long sys_read + .long sys_write + .long sys_open /* 5 */ + .long sys_close + .long sys_waitpid + .long sys_creat + .long sys_link + .long sys_unlink /* 10 */ + .long sys_execve + .long sys_chdir + .long sys_time + .long sys_mknod + .long sys_chmod /* 15 */ + .long sys_lchown16 + .long sys_ni_syscall /* old break syscall holder */ + .long sys_stat + .long sys_lseek + .long sys_getpid /* 20 */ + .long sys_mount + .long sys_oldumount + .long sys_setuid16 + .long sys_getuid16 + .long sys_stime /* 25 */ + .long sys_ptrace + .long sys_alarm + .long sys_fstat + .long sys_pause + .long sys_utime /* 30 */ + .long sys_ni_syscall /* old stty syscall holder */ + .long sys_ni_syscall /* old gtty syscall holder */ + .long sys_access + .long sys_nice + .long sys_ni_syscall /* 35 - old ftime syscall holder */ + .long sys_sync + .long sys_kill + .long sys_rename + .long sys_mkdir + .long sys_rmdir /* 40 */ + .long sys_dup + .long sys_pipe + .long sys_times + .long sys_ni_syscall /* old prof syscall holder */ + .long sys_brk /* 45 */ + .long sys_setgid16 + .long sys_getgid16 + .long sys_signal + .long sys_geteuid16 + .long sys_getegid16 /* 50 */ + .long sys_acct + .long sys_umount /* recycled never used phys() */ + .long sys_ni_syscall /* old lock syscall holder */ + .long sys_ioctl + .long sys_fcntl /* 55 */ + .long sys_ni_syscall /* old mpx syscall holder */ + .long sys_setpgid + .long sys_ni_syscall /* old ulimit syscall holder */ + .long sys_olduname + .long sys_umask /* 60 */ + .long sys_chroot + .long sys_ustat + .long sys_dup2 + .long sys_getppid + .long sys_getpgrp /* 65 */ + .long sys_setsid + .long sys_sigaction + .long sys_sgetmask + .long sys_ssetmask + .long sys_setreuid16 /* 70 */ + .long sys_setregid16 + .long sys_sigsuspend + .long sys_sigpending + .long sys_sethostname + .long sys_setrlimit /* 75 */ + .long sys_old_getrlimit + .long sys_getrusage + .long sys_gettimeofday + .long sys_settimeofday + .long sys_getgroups16 /* 80 */ + .long sys_setgroups16 + .long old_select + .long sys_symlink + .long sys_lstat + .long sys_readlink /* 85 */ + .long sys_uselib + .long sys_swapon + .long sys_reboot + .long old_readdir + .long old_mmap /* 90 */ + .long sys_munmap + .long sys_truncate + .long sys_ftruncate + .long sys_fchmod + .long sys_fchown16 /* 95 */ + .long sys_getpriority + .long sys_setpriority + .long sys_ni_syscall /* old profil syscall holder */ + .long sys_statfs + .long sys_fstatfs /* 100 */ + .long sys_ioperm + .long sys_socketcall + .long sys_syslog + .long sys_setitimer + .long sys_getitimer /* 105 */ + .long sys_newstat + .long sys_newlstat + .long sys_newfstat + .long sys_uname + .long sys_iopl /* 110 */ + .long sys_vhangup + .long sys_ni_syscall /* old "idle" system call */ + .long sys_vm86old + .long sys_wait4 + .long sys_swapoff /* 115 */ + .long sys_sysinfo + .long sys_ipc + .long sys_fsync + .long sys_sigreturn + .long sys_clone /* 120 */ + .long sys_setdomainname + .long sys_newuname + .long sys_modify_ldt + .long sys_adjtimex + .long sys_mprotect /* 125 */ + .long sys_sigprocmask + .long sys_ni_syscall /* old "create_module" */ + .long sys_init_module + .long sys_delete_module + .long sys_ni_syscall /* 130: old "get_kernel_syms" */ + .long sys_quotactl + .long sys_getpgid + .long sys_fchdir + .long sys_bdflush + .long sys_sysfs /* 135 */ + .long sys_personality + .long sys_ni_syscall /* reserved for afs_syscall */ + .long sys_setfsuid16 + .long sys_setfsgid16 + .long sys_llseek /* 140 */ + .long sys_getdents + .long sys_select + .long sys_flock + .long sys_msync + .long sys_readv /* 145 */ + .long sys_writev + .long sys_getsid + .long sys_fdatasync + .long sys_sysctl + .long sys_mlock /* 150 */ + .long sys_munlock + .long sys_mlockall + .long sys_munlockall + .long sys_sched_setparam + .long sys_sched_getparam /* 155 */ + .long sys_sched_setscheduler + .long sys_sched_getscheduler + .long sys_sched_yield + .long sys_sched_get_priority_max + .long sys_sched_get_priority_min /* 160 */ + .long sys_sched_rr_get_interval + .long sys_nanosleep + .long sys_mremap + .long sys_setresuid16 + .long sys_getresuid16 /* 165 */ + .long sys_vm86 + .long sys_ni_syscall /* Old sys_query_module */ + .long sys_poll + .long sys_nfsservctl + .long sys_setresgid16 /* 170 */ + .long sys_getresgid16 + .long sys_prctl + .long sys_rt_sigreturn + .long sys_rt_sigaction + .long sys_rt_sigprocmask /* 175 */ + .long sys_rt_sigpending + .long sys_rt_sigtimedwait + .long sys_rt_sigqueueinfo + .long sys_rt_sigsuspend + .long sys_pread64 /* 180 */ + .long sys_pwrite64 + .long sys_chown16 + .long sys_getcwd + .long sys_capget + .long sys_capset /* 185 */ + .long sys_sigaltstack + .long sys_sendfile + .long sys_ni_syscall /* reserved for streams1 */ + .long sys_ni_syscall /* reserved for streams2 */ + .long sys_vfork /* 190 */ + .long sys_getrlimit + .long sys_mmap2 + .long sys_truncate64 + .long sys_ftruncate64 + .long sys_stat64 /* 195 */ + .long sys_lstat64 + .long sys_fstat64 + .long sys_lchown + .long sys_getuid + .long sys_getgid /* 200 */ + .long sys_geteuid + .long sys_getegid + .long sys_setreuid + .long sys_setregid + .long sys_getgroups /* 205 */ + .long sys_setgroups + .long sys_fchown + .long sys_setresuid + .long sys_getresuid + .long sys_setresgid /* 210 */ + .long sys_getresgid + .long sys_chown + .long sys_setuid + .long sys_setgid + .long sys_setfsuid /* 215 */ + .long sys_setfsgid + .long sys_pivot_root + .long sys_mincore + .long sys_madvise + .long sys_getdents64 /* 220 */ + .long sys_fcntl64 + .long sys_ni_syscall /* reserved for TUX */ + .long sys_ni_syscall + .long sys_gettid + .long sys_readahead /* 225 */ + .long sys_setxattr + .long sys_lsetxattr + .long sys_fsetxattr + .long sys_getxattr + .long sys_lgetxattr /* 230 */ + .long sys_fgetxattr + .long sys_listxattr + .long sys_llistxattr + .long sys_flistxattr + .long sys_removexattr /* 235 */ + .long sys_lremovexattr + .long sys_fremovexattr + .long sys_tkill + .long sys_sendfile64 + .long sys_futex /* 240 */ + .long sys_sched_setaffinity + .long sys_sched_getaffinity + .long sys_set_thread_area + .long sys_get_thread_area + .long sys_io_setup /* 245 */ + .long sys_io_destroy + .long sys_io_getevents + .long sys_io_submit + .long sys_io_cancel + .long sys_fadvise64 /* 250 */ + .long sys_ni_syscall + .long sys_exit_group + .long sys_lookup_dcookie + .long sys_epoll_create + .long sys_epoll_ctl /* 255 */ + .long sys_epoll_wait + .long sys_remap_file_pages + .long sys_set_tid_address + .long sys_timer_create + .long sys_timer_settime /* 260 */ + .long sys_timer_gettime + .long sys_timer_getoverrun + .long sys_timer_delete + .long sys_clock_settime + .long sys_clock_gettime /* 265 */ + .long sys_clock_getres + .long sys_clock_nanosleep + .long sys_statfs64 + .long sys_fstatfs64 + .long sys_tgkill /* 270 */ + .long sys_utimes + .long sys_fadvise64_64 + .long sys_ni_syscall /* sys_vserver */ + .long sys_mbind + .long sys_get_mempolicy + .long sys_set_mempolicy + .long sys_mq_open + .long sys_mq_unlink + .long sys_mq_timedsend + .long sys_mq_timedreceive /* 280 */ + .long sys_mq_notify + .long sys_mq_getsetattr + .long sys_ni_syscall /* reserved for kexec */ + .long sys_waitid + .long sys_ni_syscall /* 285 */ /* available */ + .long sys_add_key + .long sys_request_key + .long sys_keyctl diff --git a/arch/um/include/sysdep-i386/syscalls.h b/arch/um/include/sysdep-i386/syscalls.h index 5db81ec9087..be0a3e3469e 100644 --- a/arch/um/include/sysdep-i386/syscalls.h +++ b/arch/um/include/sysdep-i386/syscalls.h @@ -22,102 +22,3 @@ extern syscall_handler_t old_mmap_i386; extern long sys_mmap2(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags, unsigned long fd, unsigned long pgoff); - -/* On i386 they choose a meaningless naming.*/ -#define __NR_kexec_load __NR_sys_kexec_load - -#define ARCH_SYSCALLS \ - [ __NR_waitpid ] = (syscall_handler_t *) sys_waitpid, \ - [ __NR_break ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_oldstat ] = (syscall_handler_t *) sys_stat, \ - [ __NR_umount ] = (syscall_handler_t *) sys_oldumount, \ - [ __NR_stime ] = um_stime, \ - [ __NR_oldfstat ] = (syscall_handler_t *) sys_fstat, \ - [ __NR_stty ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_gtty ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_nice ] = (syscall_handler_t *) sys_nice, \ - [ __NR_ftime ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_prof ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_signal ] = (syscall_handler_t *) sys_signal, \ - [ __NR_lock ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_mpx ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_ulimit ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_oldolduname ] = (syscall_handler_t *) sys_olduname, \ - [ __NR_sigaction ] = (syscall_handler_t *) sys_sigaction, \ - [ __NR_sgetmask ] = (syscall_handler_t *) sys_sgetmask, \ - [ __NR_ssetmask ] = (syscall_handler_t *) sys_ssetmask, \ - [ __NR_sigsuspend ] = (syscall_handler_t *) sys_sigsuspend, \ - [ __NR_sigpending ] = (syscall_handler_t *) sys_sigpending, \ - [ __NR_oldlstat ] = (syscall_handler_t *) sys_lstat, \ - [ __NR_readdir ] = old_readdir, \ - [ __NR_profil ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_socketcall ] = (syscall_handler_t *) sys_socketcall, \ - [ __NR_olduname ] = (syscall_handler_t *) sys_uname, \ - [ __NR_iopl ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_idle ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_ipc ] = (syscall_handler_t *) sys_ipc, \ - [ __NR_sigreturn ] = (syscall_handler_t *) sys_sigreturn, \ - [ __NR_sigprocmask ] = (syscall_handler_t *) sys_sigprocmask, \ - [ __NR_bdflush ] = (syscall_handler_t *) sys_bdflush, \ - [ __NR__llseek ] = (syscall_handler_t *) sys_llseek, \ - [ __NR__newselect ] = (syscall_handler_t *) sys_select, \ - [ __NR_vm86 ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_mmap ] = (syscall_handler_t *) old_mmap_i386, \ - [ __NR_ugetrlimit ] = (syscall_handler_t *) sys_getrlimit, \ - [ __NR_mmap2 ] = (syscall_handler_t *) sys_mmap2, \ - [ __NR_truncate64 ] = (syscall_handler_t *) sys_truncate64, \ - [ __NR_ftruncate64 ] = (syscall_handler_t *) sys_ftruncate64, \ - [ __NR_stat64 ] = (syscall_handler_t *) sys_stat64, \ - [ __NR_lstat64 ] = (syscall_handler_t *) sys_lstat64, \ - [ __NR_fstat64 ] = (syscall_handler_t *) sys_fstat64, \ - [ __NR_fcntl64 ] = (syscall_handler_t *) sys_fcntl64, \ - [ __NR_sendfile64 ] = (syscall_handler_t *) sys_sendfile64, \ - [ __NR_statfs64 ] = (syscall_handler_t *) sys_statfs64, \ - [ __NR_fstatfs64 ] = (syscall_handler_t *) sys_fstatfs64, \ - [ __NR_fadvise64_64 ] = (syscall_handler_t *) sys_fadvise64_64, \ - [ __NR_select ] = (syscall_handler_t *) old_select, \ - [ __NR_vm86old ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_modify_ldt ] = (syscall_handler_t *) sys_modify_ldt, \ - [ __NR_lchown32 ] = (syscall_handler_t *) sys_lchown, \ - [ __NR_getuid32 ] = (syscall_handler_t *) sys_getuid, \ - [ __NR_getgid32 ] = (syscall_handler_t *) sys_getgid, \ - [ __NR_geteuid32 ] = (syscall_handler_t *) sys_geteuid, \ - [ __NR_getegid32 ] = (syscall_handler_t *) sys_getegid, \ - [ __NR_setreuid32 ] = (syscall_handler_t *) sys_setreuid, \ - [ __NR_setregid32 ] = (syscall_handler_t *) sys_setregid, \ - [ __NR_getgroups32 ] = (syscall_handler_t *) sys_getgroups, \ - [ __NR_setgroups32 ] = (syscall_handler_t *) sys_setgroups, \ - [ __NR_fchown32 ] = (syscall_handler_t *) sys_fchown, \ - [ __NR_setresuid32 ] = (syscall_handler_t *) sys_setresuid, \ - [ __NR_getresuid32 ] = (syscall_handler_t *) sys_getresuid, \ - [ __NR_setresgid32 ] = (syscall_handler_t *) sys_setresgid, \ - [ __NR_getresgid32 ] = (syscall_handler_t *) sys_getresgid, \ - [ __NR_chown32 ] = (syscall_handler_t *) sys_chown, \ - [ __NR_setuid32 ] = (syscall_handler_t *) sys_setuid, \ - [ __NR_setgid32 ] = (syscall_handler_t *) sys_setgid, \ - [ __NR_setfsuid32 ] = (syscall_handler_t *) sys_setfsuid, \ - [ __NR_setfsgid32 ] = (syscall_handler_t *) sys_setfsgid, \ - [ __NR_pivot_root ] = (syscall_handler_t *) sys_pivot_root, \ - [ __NR_mincore ] = (syscall_handler_t *) sys_mincore, \ - [ __NR_madvise ] = (syscall_handler_t *) sys_madvise, \ - [ 222 ] = (syscall_handler_t *) sys_ni_syscall, \ - [ 223 ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_set_thread_area ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_get_thread_area ] = (syscall_handler_t *) sys_ni_syscall, \ - [ 251 ] = (syscall_handler_t *) sys_ni_syscall, \ - [ 285 ] = (syscall_handler_t *) sys_ni_syscall, - -/* 222 doesn't yet have a name in include/asm-i386/unistd.h */ - -#define LAST_ARCH_SYSCALL 285 - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/include/sysdep-x86_64/syscalls.h b/arch/um/include/sysdep-x86_64/syscalls.h index b187a4157ff..65a01e3d870 100644 --- a/arch/um/include/sysdep-x86_64/syscalls.h +++ b/arch/um/include/sysdep-x86_64/syscalls.h @@ -78,14 +78,3 @@ extern syscall_handler_t sys_arch_prctl; #define NR_syscalls 1024 #endif - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/kernel/Makefile b/arch/um/kernel/Makefile index 280d3000111..246f0e7fb4c 100644 --- a/arch/um/kernel/Makefile +++ b/arch/um/kernel/Makefile @@ -10,7 +10,7 @@ obj-y = checksum.o config.o exec_kern.o exitcode.o \ helper.o init_task.o irq.o irq_user.o ksyms.o main.o mem.o mem_user.o \ physmem.o process.o process_kern.o ptrace.o reboot.o resource.o \ sigio_user.o sigio_kern.o signal_kern.o signal_user.o smp.o \ - syscall_kern.o sysrq.o sys_call_table.o tempfile.o time.o time_kern.o \ + syscall_kern.o sysrq.o tempfile.o time.o time_kern.o \ tlb.o trap_kern.o trap_user.o uaccess_user.o um_arch.o umid.o \ user_util.o diff --git a/arch/um/kernel/sys_call_table.c b/arch/um/kernel/sys_call_table.c deleted file mode 100644 index b671a31a006..00000000000 --- a/arch/um/kernel/sys_call_table.c +++ /dev/null @@ -1,270 +0,0 @@ -/* - * Copyright (C) 2000 Jeff Dike (jdike@karaya.com) - * Copyright 2003 PathScale, Inc. - * Licensed under the GPL - */ - -#include "linux/config.h" -#include "linux/unistd.h" -#include "linux/sys.h" -#include "linux/swap.h" -#include "linux/syscalls.h" -#include "linux/sysctl.h" -#include "asm/signal.h" -#include "sysdep/syscalls.h" -#include "kern_util.h" - -#define LAST_GENERIC_SYSCALL __NR_keyctl - -#if LAST_GENERIC_SYSCALL > LAST_ARCH_SYSCALL -#define LAST_SYSCALL LAST_GENERIC_SYSCALL -#else -#define LAST_SYSCALL LAST_ARCH_SYSCALL -#endif - -extern syscall_handler_t sys_fork; -extern syscall_handler_t sys_execve; -extern syscall_handler_t um_time; -extern syscall_handler_t um_stime; -extern syscall_handler_t sys_pipe; -extern syscall_handler_t sys_olduname; -extern syscall_handler_t sys_sigaction; -extern syscall_handler_t sys_sigsuspend; -extern syscall_handler_t old_readdir; -extern syscall_handler_t sys_uname; -extern syscall_handler_t sys_ipc; -extern syscall_handler_t sys_sigreturn; -extern syscall_handler_t sys_clone; -extern syscall_handler_t sys_rt_sigreturn; -extern syscall_handler_t sys_sigaltstack; -extern syscall_handler_t sys_vfork; -extern syscall_handler_t old_select; -extern syscall_handler_t sys_modify_ldt; -extern syscall_handler_t sys_rt_sigsuspend; -extern syscall_handler_t sys_mbind; -extern syscall_handler_t sys_get_mempolicy; -extern syscall_handler_t sys_set_mempolicy; -extern syscall_handler_t sys_sys_setaltroot; - -syscall_handler_t *sys_call_table[] = { - [ __NR_restart_syscall ] = (syscall_handler_t *) sys_restart_syscall, - [ __NR_exit ] = (syscall_handler_t *) sys_exit, - [ __NR_fork ] = (syscall_handler_t *) sys_fork, - [ __NR_read ] = (syscall_handler_t *) sys_read, - [ __NR_write ] = (syscall_handler_t *) sys_write, - - /* These three are declared differently in asm/unistd.h */ - [ __NR_open ] = (syscall_handler_t *) sys_open, - [ __NR_close ] = (syscall_handler_t *) sys_close, - [ __NR_creat ] = (syscall_handler_t *) sys_creat, - [ __NR_link ] = (syscall_handler_t *) sys_link, - [ __NR_unlink ] = (syscall_handler_t *) sys_unlink, - [ __NR_execve ] = (syscall_handler_t *) sys_execve, - - /* declared differently in kern_util.h */ - [ __NR_chdir ] = (syscall_handler_t *) sys_chdir, - [ __NR_time ] = um_time, - [ __NR_mknod ] = (syscall_handler_t *) sys_mknod, - [ __NR_chmod ] = (syscall_handler_t *) sys_chmod, - [ __NR_lchown ] = (syscall_handler_t *) sys_lchown16, - [ __NR_lseek ] = (syscall_handler_t *) sys_lseek, - [ __NR_getpid ] = (syscall_handler_t *) sys_getpid, - [ __NR_mount ] = (syscall_handler_t *) sys_mount, - [ __NR_setuid ] = (syscall_handler_t *) sys_setuid16, - [ __NR_getuid ] = (syscall_handler_t *) sys_getuid16, - [ __NR_ptrace ] = (syscall_handler_t *) sys_ptrace, - [ __NR_alarm ] = (syscall_handler_t *) sys_alarm, - [ __NR_pause ] = (syscall_handler_t *) sys_pause, - [ __NR_utime ] = (syscall_handler_t *) sys_utime, - [ __NR_access ] = (syscall_handler_t *) sys_access, - [ __NR_sync ] = (syscall_handler_t *) sys_sync, - [ __NR_kill ] = (syscall_handler_t *) sys_kill, - [ __NR_rename ] = (syscall_handler_t *) sys_rename, - [ __NR_mkdir ] = (syscall_handler_t *) sys_mkdir, - [ __NR_rmdir ] = (syscall_handler_t *) sys_rmdir, - - /* Declared differently in asm/unistd.h */ - [ __NR_dup ] = (syscall_handler_t *) sys_dup, - [ __NR_pipe ] = (syscall_handler_t *) sys_pipe, - [ __NR_times ] = (syscall_handler_t *) sys_times, - [ __NR_brk ] = (syscall_handler_t *) sys_brk, - [ __NR_setgid ] = (syscall_handler_t *) sys_setgid16, - [ __NR_getgid ] = (syscall_handler_t *) sys_getgid16, - [ __NR_geteuid ] = (syscall_handler_t *) sys_geteuid16, - [ __NR_getegid ] = (syscall_handler_t *) sys_getegid16, - [ __NR_acct ] = (syscall_handler_t *) sys_acct, - [ __NR_umount2 ] = (syscall_handler_t *) sys_umount, - [ __NR_ioctl ] = (syscall_handler_t *) sys_ioctl, - [ __NR_fcntl ] = (syscall_handler_t *) sys_fcntl, - [ __NR_setpgid ] = (syscall_handler_t *) sys_setpgid, - [ __NR_umask ] = (syscall_handler_t *) sys_umask, - [ __NR_chroot ] = (syscall_handler_t *) sys_chroot, - [ __NR_ustat ] = (syscall_handler_t *) sys_ustat, - [ __NR_dup2 ] = (syscall_handler_t *) sys_dup2, - [ __NR_getppid ] = (syscall_handler_t *) sys_getppid, - [ __NR_getpgrp ] = (syscall_handler_t *) sys_getpgrp, - [ __NR_setsid ] = (syscall_handler_t *) sys_setsid, - [ __NR_setreuid ] = (syscall_handler_t *) sys_setreuid16, - [ __NR_setregid ] = (syscall_handler_t *) sys_setregid16, - [ __NR_sethostname ] = (syscall_handler_t *) sys_sethostname, - [ __NR_setrlimit ] = (syscall_handler_t *) sys_setrlimit, - [ __NR_getrlimit ] = (syscall_handler_t *) sys_old_getrlimit, - [ __NR_getrusage ] = (syscall_handler_t *) sys_getrusage, - [ __NR_gettimeofday ] = (syscall_handler_t *) sys_gettimeofday, - [ __NR_settimeofday ] = (syscall_handler_t *) sys_settimeofday, - [ __NR_getgroups ] = (syscall_handler_t *) sys_getgroups16, - [ __NR_setgroups ] = (syscall_handler_t *) sys_setgroups16, - [ __NR_symlink ] = (syscall_handler_t *) sys_symlink, - [ __NR_readlink ] = (syscall_handler_t *) sys_readlink, - [ __NR_uselib ] = (syscall_handler_t *) sys_uselib, - [ __NR_swapon ] = (syscall_handler_t *) sys_swapon, - [ __NR_reboot ] = (syscall_handler_t *) sys_reboot, - [ __NR_munmap ] = (syscall_handler_t *) sys_munmap, - [ __NR_truncate ] = (syscall_handler_t *) sys_truncate, - [ __NR_ftruncate ] = (syscall_handler_t *) sys_ftruncate, - [ __NR_fchmod ] = (syscall_handler_t *) sys_fchmod, - [ __NR_fchown ] = (syscall_handler_t *) sys_fchown16, - [ __NR_getpriority ] = (syscall_handler_t *) sys_getpriority, - [ __NR_setpriority ] = (syscall_handler_t *) sys_setpriority, - [ __NR_statfs ] = (syscall_handler_t *) sys_statfs, - [ __NR_fstatfs ] = (syscall_handler_t *) sys_fstatfs, - [ __NR_ioperm ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_syslog ] = (syscall_handler_t *) sys_syslog, - [ __NR_setitimer ] = (syscall_handler_t *) sys_setitimer, - [ __NR_getitimer ] = (syscall_handler_t *) sys_getitimer, - [ __NR_stat ] = (syscall_handler_t *) sys_newstat, - [ __NR_lstat ] = (syscall_handler_t *) sys_newlstat, - [ __NR_fstat ] = (syscall_handler_t *) sys_newfstat, - [ __NR_vhangup ] = (syscall_handler_t *) sys_vhangup, - [ __NR_wait4 ] = (syscall_handler_t *) sys_wait4, - [ __NR_swapoff ] = (syscall_handler_t *) sys_swapoff, - [ __NR_sysinfo ] = (syscall_handler_t *) sys_sysinfo, - [ __NR_fsync ] = (syscall_handler_t *) sys_fsync, - [ __NR_clone ] = (syscall_handler_t *) sys_clone, - [ __NR_setdomainname ] = (syscall_handler_t *) sys_setdomainname, - [ __NR_uname ] = (syscall_handler_t *) sys_newuname, - [ __NR_adjtimex ] = (syscall_handler_t *) sys_adjtimex, - [ __NR_mprotect ] = (syscall_handler_t *) sys_mprotect, - [ __NR_create_module ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_init_module ] = (syscall_handler_t *) sys_init_module, - [ __NR_delete_module ] = (syscall_handler_t *) sys_delete_module, - [ __NR_get_kernel_syms ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_quotactl ] = (syscall_handler_t *) sys_quotactl, - [ __NR_getpgid ] = (syscall_handler_t *) sys_getpgid, - [ __NR_fchdir ] = (syscall_handler_t *) sys_fchdir, - [ __NR_sysfs ] = (syscall_handler_t *) sys_sysfs, - [ __NR_personality ] = (syscall_handler_t *) sys_personality, - [ __NR_afs_syscall ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_setfsuid ] = (syscall_handler_t *) sys_setfsuid16, - [ __NR_setfsgid ] = (syscall_handler_t *) sys_setfsgid16, - [ __NR_getdents ] = (syscall_handler_t *) sys_getdents, - [ __NR_flock ] = (syscall_handler_t *) sys_flock, - [ __NR_msync ] = (syscall_handler_t *) sys_msync, - [ __NR_readv ] = (syscall_handler_t *) sys_readv, - [ __NR_writev ] = (syscall_handler_t *) sys_writev, - [ __NR_getsid ] = (syscall_handler_t *) sys_getsid, - [ __NR_fdatasync ] = (syscall_handler_t *) sys_fdatasync, - [ __NR__sysctl ] = (syscall_handler_t *) sys_sysctl, - [ __NR_mlock ] = (syscall_handler_t *) sys_mlock, - [ __NR_munlock ] = (syscall_handler_t *) sys_munlock, - [ __NR_mlockall ] = (syscall_handler_t *) sys_mlockall, - [ __NR_munlockall ] = (syscall_handler_t *) sys_munlockall, - [ __NR_sched_setparam ] = (syscall_handler_t *) sys_sched_setparam, - [ __NR_sched_getparam ] = (syscall_handler_t *) sys_sched_getparam, - [ __NR_sched_setscheduler ] = (syscall_handler_t *) sys_sched_setscheduler, - [ __NR_sched_getscheduler ] = (syscall_handler_t *) sys_sched_getscheduler, - [ __NR_sched_yield ] = (syscall_handler_t *) yield, - [ __NR_sched_get_priority_max ] = (syscall_handler_t *) sys_sched_get_priority_max, - [ __NR_sched_get_priority_min ] = (syscall_handler_t *) sys_sched_get_priority_min, - [ __NR_sched_rr_get_interval ] = (syscall_handler_t *) sys_sched_rr_get_interval, - [ __NR_nanosleep ] = (syscall_handler_t *) sys_nanosleep, - [ __NR_mremap ] = (syscall_handler_t *) sys_mremap, - [ __NR_setresuid ] = (syscall_handler_t *) sys_setresuid16, - [ __NR_getresuid ] = (syscall_handler_t *) sys_getresuid16, - [ __NR_query_module ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_poll ] = (syscall_handler_t *) sys_poll, - [ __NR_nfsservctl ] = (syscall_handler_t *) sys_nfsservctl, - [ __NR_setresgid ] = (syscall_handler_t *) sys_setresgid16, - [ __NR_getresgid ] = (syscall_handler_t *) sys_getresgid16, - [ __NR_prctl ] = (syscall_handler_t *) sys_prctl, - [ __NR_rt_sigreturn ] = (syscall_handler_t *) sys_rt_sigreturn, - [ __NR_rt_sigaction ] = (syscall_handler_t *) sys_rt_sigaction, - [ __NR_rt_sigprocmask ] = (syscall_handler_t *) sys_rt_sigprocmask, - [ __NR_rt_sigpending ] = (syscall_handler_t *) sys_rt_sigpending, - [ __NR_rt_sigtimedwait ] = (syscall_handler_t *) sys_rt_sigtimedwait, - [ __NR_rt_sigqueueinfo ] = (syscall_handler_t *) sys_rt_sigqueueinfo, - [ __NR_rt_sigsuspend ] = (syscall_handler_t *) sys_rt_sigsuspend, - [ __NR_pread64 ] = (syscall_handler_t *) sys_pread64, - [ __NR_pwrite64 ] = (syscall_handler_t *) sys_pwrite64, - [ __NR_chown ] = (syscall_handler_t *) sys_chown16, - [ __NR_getcwd ] = (syscall_handler_t *) sys_getcwd, - [ __NR_capget ] = (syscall_handler_t *) sys_capget, - [ __NR_capset ] = (syscall_handler_t *) sys_capset, - [ __NR_sigaltstack ] = (syscall_handler_t *) sys_sigaltstack, - [ __NR_sendfile ] = (syscall_handler_t *) sys_sendfile, - [ __NR_getpmsg ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_putpmsg ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_vfork ] = (syscall_handler_t *) sys_vfork, - [ __NR_getdents64 ] = (syscall_handler_t *) sys_getdents64, - [ __NR_gettid ] = (syscall_handler_t *) sys_gettid, - [ __NR_readahead ] = (syscall_handler_t *) sys_readahead, - [ __NR_setxattr ] = (syscall_handler_t *) sys_setxattr, - [ __NR_lsetxattr ] = (syscall_handler_t *) sys_lsetxattr, - [ __NR_fsetxattr ] = (syscall_handler_t *) sys_fsetxattr, - [ __NR_getxattr ] = (syscall_handler_t *) sys_getxattr, - [ __NR_lgetxattr ] = (syscall_handler_t *) sys_lgetxattr, - [ __NR_fgetxattr ] = (syscall_handler_t *) sys_fgetxattr, - [ __NR_listxattr ] = (syscall_handler_t *) sys_listxattr, - [ __NR_llistxattr ] = (syscall_handler_t *) sys_llistxattr, - [ __NR_flistxattr ] = (syscall_handler_t *) sys_flistxattr, - [ __NR_removexattr ] = (syscall_handler_t *) sys_removexattr, - [ __NR_lremovexattr ] = (syscall_handler_t *) sys_lremovexattr, - [ __NR_fremovexattr ] = (syscall_handler_t *) sys_fremovexattr, - [ __NR_tkill ] = (syscall_handler_t *) sys_tkill, - [ __NR_futex ] = (syscall_handler_t *) sys_futex, - [ __NR_sched_setaffinity ] = (syscall_handler_t *) sys_sched_setaffinity, - [ __NR_sched_getaffinity ] = (syscall_handler_t *) sys_sched_getaffinity, - [ __NR_io_setup ] = (syscall_handler_t *) sys_io_setup, - [ __NR_io_destroy ] = (syscall_handler_t *) sys_io_destroy, - [ __NR_io_getevents ] = (syscall_handler_t *) sys_io_getevents, - [ __NR_io_submit ] = (syscall_handler_t *) sys_io_submit, - [ __NR_io_cancel ] = (syscall_handler_t *) sys_io_cancel, - [ __NR_exit_group ] = (syscall_handler_t *) sys_exit_group, - [ __NR_lookup_dcookie ] = (syscall_handler_t *) sys_lookup_dcookie, - [ __NR_epoll_create ] = (syscall_handler_t *) sys_epoll_create, - [ __NR_epoll_ctl ] = (syscall_handler_t *) sys_epoll_ctl, - [ __NR_epoll_wait ] = (syscall_handler_t *) sys_epoll_wait, - [ __NR_remap_file_pages ] = (syscall_handler_t *) sys_remap_file_pages, - [ __NR_set_tid_address ] = (syscall_handler_t *) sys_set_tid_address, - [ __NR_timer_create ] = (syscall_handler_t *) sys_timer_create, - [ __NR_timer_settime ] = (syscall_handler_t *) sys_timer_settime, - [ __NR_timer_gettime ] = (syscall_handler_t *) sys_timer_gettime, - [ __NR_timer_getoverrun ] = (syscall_handler_t *) sys_timer_getoverrun, - [ __NR_timer_delete ] = (syscall_handler_t *) sys_timer_delete, - [ __NR_clock_settime ] = (syscall_handler_t *) sys_clock_settime, - [ __NR_clock_gettime ] = (syscall_handler_t *) sys_clock_gettime, - [ __NR_clock_getres ] = (syscall_handler_t *) sys_clock_getres, - [ __NR_clock_nanosleep ] = (syscall_handler_t *) sys_clock_nanosleep, - [ __NR_tgkill ] = (syscall_handler_t *) sys_tgkill, - [ __NR_utimes ] = (syscall_handler_t *) sys_utimes, - [ __NR_fadvise64 ] = (syscall_handler_t *) sys_fadvise64, - [ __NR_vserver ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_mbind ] = (syscall_handler_t *) sys_mbind, - [ __NR_get_mempolicy ] = (syscall_handler_t *) sys_get_mempolicy, - [ __NR_set_mempolicy ] = (syscall_handler_t *) sys_set_mempolicy, - [ __NR_mq_open ] = (syscall_handler_t *) sys_mq_open, - [ __NR_mq_unlink ] = (syscall_handler_t *) sys_mq_unlink, - [ __NR_mq_timedsend ] = (syscall_handler_t *) sys_mq_timedsend, - [ __NR_mq_timedreceive ] = (syscall_handler_t *) sys_mq_timedreceive, - [ __NR_mq_notify ] = (syscall_handler_t *) sys_mq_notify, - [ __NR_mq_getsetattr ] = (syscall_handler_t *) sys_mq_getsetattr, - [ __NR_kexec_load ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_waitid ] = (syscall_handler_t *) sys_waitid, - [ __NR_add_key ] = (syscall_handler_t *) sys_add_key, - [ __NR_request_key ] = (syscall_handler_t *) sys_request_key, - [ __NR_keyctl ] = (syscall_handler_t *) sys_keyctl, - - ARCH_SYSCALLS - [ LAST_SYSCALL + 1 ... NR_syscalls ] = - (syscall_handler_t *) sys_ni_syscall -}; diff --git a/arch/um/sys-i386/Makefile b/arch/um/sys-i386/Makefile index 71b47e61860..950781e354d 100644 --- a/arch/um/sys-i386/Makefile +++ b/arch/um/sys-i386/Makefile @@ -1,5 +1,6 @@ obj-y = bitops.o bugs.o checksum.o delay.o fault.o ksyms.o ldt.o ptrace.o \ - ptrace_user.o semaphore.o signal.o sigcontext.o syscalls.o sysrq.o + ptrace_user.o semaphore.o signal.o sigcontext.o syscalls.o sysrq.o \ + sys_call_table.o obj-$(CONFIG_HIGHMEM) += highmem.o obj-$(CONFIG_MODULES) += module.o diff --git a/arch/um/sys-i386/sys_call_table.S b/arch/um/sys-i386/sys_call_table.S new file mode 100644 index 00000000000..ad75c27afe3 --- /dev/null +++ b/arch/um/sys-i386/sys_call_table.S @@ -0,0 +1,16 @@ +#include +/* Steal i386 syscall table for our purposes, but with some slight changes.*/ + +#define sys_iopl sys_ni_syscall +#define sys_ioperm sys_ni_syscall + +#define sys_vm86old sys_ni_syscall +#define sys_vm86 sys_ni_syscall +#define sys_set_thread_area sys_ni_syscall +#define sys_get_thread_area sys_ni_syscall + +#define sys_stime um_stime +#define sys_time um_time +#define old_mmap old_mmap_i386 + +#include "../../i386/kernel/syscall_table.S" diff --git a/arch/um/sys-x86_64/Makefile b/arch/um/sys-x86_64/Makefile index 2129e314355..274158e7da3 100644 --- a/arch/um/sys-x86_64/Makefile +++ b/arch/um/sys-x86_64/Makefile @@ -6,7 +6,7 @@ lib-y = bitops.o bugs.o csum-partial.o delay.o fault.o mem.o memcpy.o \ ptrace.o ptrace_user.o semaphore.o sigcontext.o signal.o \ - syscalls.o sysrq.o thunk.o + syscalls.o sysrq.o thunk.o sys_call_table.o USER_OBJS := ptrace_user.o sigcontext.o diff --git a/arch/um/sys-x86_64/sys_call_table.c b/arch/um/sys-x86_64/sys_call_table.c new file mode 100644 index 00000000000..3f5efbfb46d --- /dev/null +++ b/arch/um/sys-x86_64/sys_call_table.c @@ -0,0 +1,276 @@ +/* + * Copyright (C) 2000 Jeff Dike (jdike@karaya.com) + * Copyright 2003 PathScale, Inc. + * Licensed under the GPL + */ + +#include "linux/config.h" +#include "linux/unistd.h" +#include "linux/sys.h" +#include "linux/swap.h" +#include "linux/syscalls.h" +#include "linux/sysctl.h" +#include "asm/signal.h" +#include "sysdep/syscalls.h" +#include "kern_util.h" + +#ifdef CONFIG_NFSD +#define NFSSERVCTL sys_nfsservctl +#else +#define NFSSERVCTL sys_ni_syscall +#endif + +#define LAST_GENERIC_SYSCALL __NR_keyctl + +#if LAST_GENERIC_SYSCALL > LAST_ARCH_SYSCALL +#define LAST_SYSCALL LAST_GENERIC_SYSCALL +#else +#define LAST_SYSCALL LAST_ARCH_SYSCALL +#endif + +extern syscall_handler_t sys_fork; +extern syscall_handler_t sys_execve; +extern syscall_handler_t um_time; +extern syscall_handler_t um_stime; +extern syscall_handler_t sys_pipe; +extern syscall_handler_t sys_olduname; +extern syscall_handler_t sys_sigaction; +extern syscall_handler_t sys_sigsuspend; +extern syscall_handler_t old_readdir; +extern syscall_handler_t sys_uname; +extern syscall_handler_t sys_ipc; +extern syscall_handler_t sys_sigreturn; +extern syscall_handler_t sys_clone; +extern syscall_handler_t sys_rt_sigreturn; +extern syscall_handler_t sys_sigaltstack; +extern syscall_handler_t sys_vfork; +extern syscall_handler_t old_select; +extern syscall_handler_t sys_modify_ldt; +extern syscall_handler_t sys_rt_sigsuspend; +extern syscall_handler_t sys_mbind; +extern syscall_handler_t sys_get_mempolicy; +extern syscall_handler_t sys_set_mempolicy; +extern syscall_handler_t sys_sys_setaltroot; + +syscall_handler_t *sys_call_table[] = { + [ __NR_restart_syscall ] = (syscall_handler_t *) sys_restart_syscall, + [ __NR_exit ] = (syscall_handler_t *) sys_exit, + [ __NR_fork ] = (syscall_handler_t *) sys_fork, + [ __NR_read ] = (syscall_handler_t *) sys_read, + [ __NR_write ] = (syscall_handler_t *) sys_write, + + /* These three are declared differently in asm/unistd.h */ + [ __NR_open ] = (syscall_handler_t *) sys_open, + [ __NR_close ] = (syscall_handler_t *) sys_close, + [ __NR_creat ] = (syscall_handler_t *) sys_creat, + [ __NR_link ] = (syscall_handler_t *) sys_link, + [ __NR_unlink ] = (syscall_handler_t *) sys_unlink, + [ __NR_execve ] = (syscall_handler_t *) sys_execve, + + /* declared differently in kern_util.h */ + [ __NR_chdir ] = (syscall_handler_t *) sys_chdir, + [ __NR_time ] = um_time, + [ __NR_mknod ] = (syscall_handler_t *) sys_mknod, + [ __NR_chmod ] = (syscall_handler_t *) sys_chmod, + [ __NR_lchown ] = (syscall_handler_t *) sys_lchown16, + [ __NR_lseek ] = (syscall_handler_t *) sys_lseek, + [ __NR_getpid ] = (syscall_handler_t *) sys_getpid, + [ __NR_mount ] = (syscall_handler_t *) sys_mount, + [ __NR_setuid ] = (syscall_handler_t *) sys_setuid16, + [ __NR_getuid ] = (syscall_handler_t *) sys_getuid16, + [ __NR_ptrace ] = (syscall_handler_t *) sys_ptrace, + [ __NR_alarm ] = (syscall_handler_t *) sys_alarm, + [ __NR_pause ] = (syscall_handler_t *) sys_pause, + [ __NR_utime ] = (syscall_handler_t *) sys_utime, + [ __NR_access ] = (syscall_handler_t *) sys_access, + [ __NR_sync ] = (syscall_handler_t *) sys_sync, + [ __NR_kill ] = (syscall_handler_t *) sys_kill, + [ __NR_rename ] = (syscall_handler_t *) sys_rename, + [ __NR_mkdir ] = (syscall_handler_t *) sys_mkdir, + [ __NR_rmdir ] = (syscall_handler_t *) sys_rmdir, + + /* Declared differently in asm/unistd.h */ + [ __NR_dup ] = (syscall_handler_t *) sys_dup, + [ __NR_pipe ] = (syscall_handler_t *) sys_pipe, + [ __NR_times ] = (syscall_handler_t *) sys_times, + [ __NR_brk ] = (syscall_handler_t *) sys_brk, + [ __NR_setgid ] = (syscall_handler_t *) sys_setgid16, + [ __NR_getgid ] = (syscall_handler_t *) sys_getgid16, + [ __NR_geteuid ] = (syscall_handler_t *) sys_geteuid16, + [ __NR_getegid ] = (syscall_handler_t *) sys_getegid16, + [ __NR_acct ] = (syscall_handler_t *) sys_acct, + [ __NR_umount2 ] = (syscall_handler_t *) sys_umount, + [ __NR_ioctl ] = (syscall_handler_t *) sys_ioctl, + [ __NR_fcntl ] = (syscall_handler_t *) sys_fcntl, + [ __NR_setpgid ] = (syscall_handler_t *) sys_setpgid, + [ __NR_umask ] = (syscall_handler_t *) sys_umask, + [ __NR_chroot ] = (syscall_handler_t *) sys_chroot, + [ __NR_ustat ] = (syscall_handler_t *) sys_ustat, + [ __NR_dup2 ] = (syscall_handler_t *) sys_dup2, + [ __NR_getppid ] = (syscall_handler_t *) sys_getppid, + [ __NR_getpgrp ] = (syscall_handler_t *) sys_getpgrp, + [ __NR_setsid ] = (syscall_handler_t *) sys_setsid, + [ __NR_setreuid ] = (syscall_handler_t *) sys_setreuid16, + [ __NR_setregid ] = (syscall_handler_t *) sys_setregid16, + [ __NR_sethostname ] = (syscall_handler_t *) sys_sethostname, + [ __NR_setrlimit ] = (syscall_handler_t *) sys_setrlimit, + [ __NR_getrlimit ] = (syscall_handler_t *) sys_old_getrlimit, + [ __NR_getrusage ] = (syscall_handler_t *) sys_getrusage, + [ __NR_gettimeofday ] = (syscall_handler_t *) sys_gettimeofday, + [ __NR_settimeofday ] = (syscall_handler_t *) sys_settimeofday, + [ __NR_getgroups ] = (syscall_handler_t *) sys_getgroups16, + [ __NR_setgroups ] = (syscall_handler_t *) sys_setgroups16, + [ __NR_symlink ] = (syscall_handler_t *) sys_symlink, + [ __NR_readlink ] = (syscall_handler_t *) sys_readlink, + [ __NR_uselib ] = (syscall_handler_t *) sys_uselib, + [ __NR_swapon ] = (syscall_handler_t *) sys_swapon, + [ __NR_reboot ] = (syscall_handler_t *) sys_reboot, + [ __NR_munmap ] = (syscall_handler_t *) sys_munmap, + [ __NR_truncate ] = (syscall_handler_t *) sys_truncate, + [ __NR_ftruncate ] = (syscall_handler_t *) sys_ftruncate, + [ __NR_fchmod ] = (syscall_handler_t *) sys_fchmod, + [ __NR_fchown ] = (syscall_handler_t *) sys_fchown16, + [ __NR_getpriority ] = (syscall_handler_t *) sys_getpriority, + [ __NR_setpriority ] = (syscall_handler_t *) sys_setpriority, + [ __NR_statfs ] = (syscall_handler_t *) sys_statfs, + [ __NR_fstatfs ] = (syscall_handler_t *) sys_fstatfs, + [ __NR_ioperm ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_syslog ] = (syscall_handler_t *) sys_syslog, + [ __NR_setitimer ] = (syscall_handler_t *) sys_setitimer, + [ __NR_getitimer ] = (syscall_handler_t *) sys_getitimer, + [ __NR_stat ] = (syscall_handler_t *) sys_newstat, + [ __NR_lstat ] = (syscall_handler_t *) sys_newlstat, + [ __NR_fstat ] = (syscall_handler_t *) sys_newfstat, + [ __NR_vhangup ] = (syscall_handler_t *) sys_vhangup, + [ __NR_wait4 ] = (syscall_handler_t *) sys_wait4, + [ __NR_swapoff ] = (syscall_handler_t *) sys_swapoff, + [ __NR_sysinfo ] = (syscall_handler_t *) sys_sysinfo, + [ __NR_fsync ] = (syscall_handler_t *) sys_fsync, + [ __NR_clone ] = (syscall_handler_t *) sys_clone, + [ __NR_setdomainname ] = (syscall_handler_t *) sys_setdomainname, + [ __NR_uname ] = (syscall_handler_t *) sys_newuname, + [ __NR_adjtimex ] = (syscall_handler_t *) sys_adjtimex, + [ __NR_mprotect ] = (syscall_handler_t *) sys_mprotect, + [ __NR_create_module ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_init_module ] = (syscall_handler_t *) sys_init_module, + [ __NR_delete_module ] = (syscall_handler_t *) sys_delete_module, + [ __NR_get_kernel_syms ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_quotactl ] = (syscall_handler_t *) sys_quotactl, + [ __NR_getpgid ] = (syscall_handler_t *) sys_getpgid, + [ __NR_fchdir ] = (syscall_handler_t *) sys_fchdir, + [ __NR_sysfs ] = (syscall_handler_t *) sys_sysfs, + [ __NR_personality ] = (syscall_handler_t *) sys_personality, + [ __NR_afs_syscall ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_setfsuid ] = (syscall_handler_t *) sys_setfsuid16, + [ __NR_setfsgid ] = (syscall_handler_t *) sys_setfsgid16, + [ __NR_getdents ] = (syscall_handler_t *) sys_getdents, + [ __NR_flock ] = (syscall_handler_t *) sys_flock, + [ __NR_msync ] = (syscall_handler_t *) sys_msync, + [ __NR_readv ] = (syscall_handler_t *) sys_readv, + [ __NR_writev ] = (syscall_handler_t *) sys_writev, + [ __NR_getsid ] = (syscall_handler_t *) sys_getsid, + [ __NR_fdatasync ] = (syscall_handler_t *) sys_fdatasync, + [ __NR__sysctl ] = (syscall_handler_t *) sys_sysctl, + [ __NR_mlock ] = (syscall_handler_t *) sys_mlock, + [ __NR_munlock ] = (syscall_handler_t *) sys_munlock, + [ __NR_mlockall ] = (syscall_handler_t *) sys_mlockall, + [ __NR_munlockall ] = (syscall_handler_t *) sys_munlockall, + [ __NR_sched_setparam ] = (syscall_handler_t *) sys_sched_setparam, + [ __NR_sched_getparam ] = (syscall_handler_t *) sys_sched_getparam, + [ __NR_sched_setscheduler ] = (syscall_handler_t *) sys_sched_setscheduler, + [ __NR_sched_getscheduler ] = (syscall_handler_t *) sys_sched_getscheduler, + [ __NR_sched_yield ] = (syscall_handler_t *) yield, + [ __NR_sched_get_priority_max ] = (syscall_handler_t *) sys_sched_get_priority_max, + [ __NR_sched_get_priority_min ] = (syscall_handler_t *) sys_sched_get_priority_min, + [ __NR_sched_rr_get_interval ] = (syscall_handler_t *) sys_sched_rr_get_interval, + [ __NR_nanosleep ] = (syscall_handler_t *) sys_nanosleep, + [ __NR_mremap ] = (syscall_handler_t *) sys_mremap, + [ __NR_setresuid ] = (syscall_handler_t *) sys_setresuid16, + [ __NR_getresuid ] = (syscall_handler_t *) sys_getresuid16, + [ __NR_query_module ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_poll ] = (syscall_handler_t *) sys_poll, + [ __NR_nfsservctl ] = (syscall_handler_t *) NFSSERVCTL, + [ __NR_setresgid ] = (syscall_handler_t *) sys_setresgid16, + [ __NR_getresgid ] = (syscall_handler_t *) sys_getresgid16, + [ __NR_prctl ] = (syscall_handler_t *) sys_prctl, + [ __NR_rt_sigreturn ] = (syscall_handler_t *) sys_rt_sigreturn, + [ __NR_rt_sigaction ] = (syscall_handler_t *) sys_rt_sigaction, + [ __NR_rt_sigprocmask ] = (syscall_handler_t *) sys_rt_sigprocmask, + [ __NR_rt_sigpending ] = (syscall_handler_t *) sys_rt_sigpending, + [ __NR_rt_sigtimedwait ] = (syscall_handler_t *) sys_rt_sigtimedwait, + [ __NR_rt_sigqueueinfo ] = (syscall_handler_t *) sys_rt_sigqueueinfo, + [ __NR_rt_sigsuspend ] = (syscall_handler_t *) sys_rt_sigsuspend, + [ __NR_pread64 ] = (syscall_handler_t *) sys_pread64, + [ __NR_pwrite64 ] = (syscall_handler_t *) sys_pwrite64, + [ __NR_chown ] = (syscall_handler_t *) sys_chown16, + [ __NR_getcwd ] = (syscall_handler_t *) sys_getcwd, + [ __NR_capget ] = (syscall_handler_t *) sys_capget, + [ __NR_capset ] = (syscall_handler_t *) sys_capset, + [ __NR_sigaltstack ] = (syscall_handler_t *) sys_sigaltstack, + [ __NR_sendfile ] = (syscall_handler_t *) sys_sendfile, + [ __NR_getpmsg ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_putpmsg ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_vfork ] = (syscall_handler_t *) sys_vfork, + [ __NR_getdents64 ] = (syscall_handler_t *) sys_getdents64, + [ __NR_gettid ] = (syscall_handler_t *) sys_gettid, + [ __NR_readahead ] = (syscall_handler_t *) sys_readahead, + [ __NR_setxattr ] = (syscall_handler_t *) sys_setxattr, + [ __NR_lsetxattr ] = (syscall_handler_t *) sys_lsetxattr, + [ __NR_fsetxattr ] = (syscall_handler_t *) sys_fsetxattr, + [ __NR_getxattr ] = (syscall_handler_t *) sys_getxattr, + [ __NR_lgetxattr ] = (syscall_handler_t *) sys_lgetxattr, + [ __NR_fgetxattr ] = (syscall_handler_t *) sys_fgetxattr, + [ __NR_listxattr ] = (syscall_handler_t *) sys_listxattr, + [ __NR_llistxattr ] = (syscall_handler_t *) sys_llistxattr, + [ __NR_flistxattr ] = (syscall_handler_t *) sys_flistxattr, + [ __NR_removexattr ] = (syscall_handler_t *) sys_removexattr, + [ __NR_lremovexattr ] = (syscall_handler_t *) sys_lremovexattr, + [ __NR_fremovexattr ] = (syscall_handler_t *) sys_fremovexattr, + [ __NR_tkill ] = (syscall_handler_t *) sys_tkill, + [ __NR_futex ] = (syscall_handler_t *) sys_futex, + [ __NR_sched_setaffinity ] = (syscall_handler_t *) sys_sched_setaffinity, + [ __NR_sched_getaffinity ] = (syscall_handler_t *) sys_sched_getaffinity, + [ __NR_io_setup ] = (syscall_handler_t *) sys_io_setup, + [ __NR_io_destroy ] = (syscall_handler_t *) sys_io_destroy, + [ __NR_io_getevents ] = (syscall_handler_t *) sys_io_getevents, + [ __NR_io_submit ] = (syscall_handler_t *) sys_io_submit, + [ __NR_io_cancel ] = (syscall_handler_t *) sys_io_cancel, + [ __NR_exit_group ] = (syscall_handler_t *) sys_exit_group, + [ __NR_lookup_dcookie ] = (syscall_handler_t *) sys_lookup_dcookie, + [ __NR_epoll_create ] = (syscall_handler_t *) sys_epoll_create, + [ __NR_epoll_ctl ] = (syscall_handler_t *) sys_epoll_ctl, + [ __NR_epoll_wait ] = (syscall_handler_t *) sys_epoll_wait, + [ __NR_remap_file_pages ] = (syscall_handler_t *) sys_remap_file_pages, + [ __NR_set_tid_address ] = (syscall_handler_t *) sys_set_tid_address, + [ __NR_timer_create ] = (syscall_handler_t *) sys_timer_create, + [ __NR_timer_settime ] = (syscall_handler_t *) sys_timer_settime, + [ __NR_timer_gettime ] = (syscall_handler_t *) sys_timer_gettime, + [ __NR_timer_getoverrun ] = (syscall_handler_t *) sys_timer_getoverrun, + [ __NR_timer_delete ] = (syscall_handler_t *) sys_timer_delete, + [ __NR_clock_settime ] = (syscall_handler_t *) sys_clock_settime, + [ __NR_clock_gettime ] = (syscall_handler_t *) sys_clock_gettime, + [ __NR_clock_getres ] = (syscall_handler_t *) sys_clock_getres, + [ __NR_clock_nanosleep ] = (syscall_handler_t *) sys_clock_nanosleep, + [ __NR_tgkill ] = (syscall_handler_t *) sys_tgkill, + [ __NR_utimes ] = (syscall_handler_t *) sys_utimes, + [ __NR_fadvise64 ] = (syscall_handler_t *) sys_fadvise64, + [ __NR_vserver ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_mbind ] = (syscall_handler_t *) sys_mbind, + [ __NR_get_mempolicy ] = (syscall_handler_t *) sys_get_mempolicy, + [ __NR_set_mempolicy ] = (syscall_handler_t *) sys_set_mempolicy, + [ __NR_mq_open ] = (syscall_handler_t *) sys_mq_open, + [ __NR_mq_unlink ] = (syscall_handler_t *) sys_mq_unlink, + [ __NR_mq_timedsend ] = (syscall_handler_t *) sys_mq_timedsend, + [ __NR_mq_timedreceive ] = (syscall_handler_t *) sys_mq_timedreceive, + [ __NR_mq_notify ] = (syscall_handler_t *) sys_mq_notify, + [ __NR_mq_getsetattr ] = (syscall_handler_t *) sys_mq_getsetattr, + [ __NR_kexec_load ] = (syscall_handler_t *) sys_ni_syscall, + [ __NR_waitid ] = (syscall_handler_t *) sys_waitid, + [ __NR_add_key ] = (syscall_handler_t *) sys_add_key, + [ __NR_request_key ] = (syscall_handler_t *) sys_request_key, + [ __NR_keyctl ] = (syscall_handler_t *) sys_keyctl, + + ARCH_SYSCALLS + [ LAST_SYSCALL + 1 ... NR_syscalls ] = + (syscall_handler_t *) sys_ni_syscall +}; -- cgit v1.2.3-70-g09d2 From ddcd6b175793839e51fdd55debb670a9f512c448 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:55 -0700 Subject: [PATCH] uml: quick fix syscall table for x86_64 Fix the moved syscall table for the x86_64 SUBARCH: - redirect __NR_chown and such to versions aware of 32-bit UIDs, - avoid the useless hack for sys_nfsservctl, - use sys_sendfile64 in the table rather than sys_sendfile. - __NR_uselib is sys_ni_syscall on x86_64 (which does not support A.OUT). - __NR_getrlimit is sys_getrlimit, not sys_old_getrlimit Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/sys-x86_64/sys_call_table.c | 53 +++++++++++++++++-------------------- 1 file changed, 24 insertions(+), 29 deletions(-) (limited to 'arch') diff --git a/arch/um/sys-x86_64/sys_call_table.c b/arch/um/sys-x86_64/sys_call_table.c index 3f5efbfb46d..a31e0dffb27 100644 --- a/arch/um/sys-x86_64/sys_call_table.c +++ b/arch/um/sys-x86_64/sys_call_table.c @@ -14,12 +14,6 @@ #include "sysdep/syscalls.h" #include "kern_util.h" -#ifdef CONFIG_NFSD -#define NFSSERVCTL sys_nfsservctl -#else -#define NFSSERVCTL sys_ni_syscall -#endif - #define LAST_GENERIC_SYSCALL __NR_keyctl #if LAST_GENERIC_SYSCALL > LAST_ARCH_SYSCALL @@ -52,6 +46,7 @@ extern syscall_handler_t sys_get_mempolicy; extern syscall_handler_t sys_set_mempolicy; extern syscall_handler_t sys_sys_setaltroot; +/* On X86-64 all syscalls are aware of 32-bit [ug]ids, so I had to fix this!*/ syscall_handler_t *sys_call_table[] = { [ __NR_restart_syscall ] = (syscall_handler_t *) sys_restart_syscall, [ __NR_exit ] = (syscall_handler_t *) sys_exit, @@ -72,12 +67,12 @@ syscall_handler_t *sys_call_table[] = { [ __NR_time ] = um_time, [ __NR_mknod ] = (syscall_handler_t *) sys_mknod, [ __NR_chmod ] = (syscall_handler_t *) sys_chmod, - [ __NR_lchown ] = (syscall_handler_t *) sys_lchown16, + [ __NR_lchown ] = (syscall_handler_t *) sys_lchown, [ __NR_lseek ] = (syscall_handler_t *) sys_lseek, [ __NR_getpid ] = (syscall_handler_t *) sys_getpid, [ __NR_mount ] = (syscall_handler_t *) sys_mount, - [ __NR_setuid ] = (syscall_handler_t *) sys_setuid16, - [ __NR_getuid ] = (syscall_handler_t *) sys_getuid16, + [ __NR_setuid ] = (syscall_handler_t *) sys_setuid, + [ __NR_getuid ] = (syscall_handler_t *) sys_getuid, [ __NR_ptrace ] = (syscall_handler_t *) sys_ptrace, [ __NR_alarm ] = (syscall_handler_t *) sys_alarm, [ __NR_pause ] = (syscall_handler_t *) sys_pause, @@ -94,10 +89,10 @@ syscall_handler_t *sys_call_table[] = { [ __NR_pipe ] = (syscall_handler_t *) sys_pipe, [ __NR_times ] = (syscall_handler_t *) sys_times, [ __NR_brk ] = (syscall_handler_t *) sys_brk, - [ __NR_setgid ] = (syscall_handler_t *) sys_setgid16, - [ __NR_getgid ] = (syscall_handler_t *) sys_getgid16, - [ __NR_geteuid ] = (syscall_handler_t *) sys_geteuid16, - [ __NR_getegid ] = (syscall_handler_t *) sys_getegid16, + [ __NR_setgid ] = (syscall_handler_t *) sys_setgid, + [ __NR_getgid ] = (syscall_handler_t *) sys_getgid, + [ __NR_geteuid ] = (syscall_handler_t *) sys_geteuid, + [ __NR_getegid ] = (syscall_handler_t *) sys_getegid, [ __NR_acct ] = (syscall_handler_t *) sys_acct, [ __NR_umount2 ] = (syscall_handler_t *) sys_umount, [ __NR_ioctl ] = (syscall_handler_t *) sys_ioctl, @@ -110,26 +105,26 @@ syscall_handler_t *sys_call_table[] = { [ __NR_getppid ] = (syscall_handler_t *) sys_getppid, [ __NR_getpgrp ] = (syscall_handler_t *) sys_getpgrp, [ __NR_setsid ] = (syscall_handler_t *) sys_setsid, - [ __NR_setreuid ] = (syscall_handler_t *) sys_setreuid16, - [ __NR_setregid ] = (syscall_handler_t *) sys_setregid16, + [ __NR_setreuid ] = (syscall_handler_t *) sys_setreuid, + [ __NR_setregid ] = (syscall_handler_t *) sys_setregid, [ __NR_sethostname ] = (syscall_handler_t *) sys_sethostname, [ __NR_setrlimit ] = (syscall_handler_t *) sys_setrlimit, - [ __NR_getrlimit ] = (syscall_handler_t *) sys_old_getrlimit, + [ __NR_getrlimit ] = (syscall_handler_t *) sys_getrlimit, [ __NR_getrusage ] = (syscall_handler_t *) sys_getrusage, [ __NR_gettimeofday ] = (syscall_handler_t *) sys_gettimeofday, [ __NR_settimeofday ] = (syscall_handler_t *) sys_settimeofday, - [ __NR_getgroups ] = (syscall_handler_t *) sys_getgroups16, - [ __NR_setgroups ] = (syscall_handler_t *) sys_setgroups16, + [ __NR_getgroups ] = (syscall_handler_t *) sys_getgroups, + [ __NR_setgroups ] = (syscall_handler_t *) sys_setgroups, [ __NR_symlink ] = (syscall_handler_t *) sys_symlink, [ __NR_readlink ] = (syscall_handler_t *) sys_readlink, - [ __NR_uselib ] = (syscall_handler_t *) sys_uselib, + [ __NR_uselib ] = (syscall_handler_t *) sys_ni_syscall, [ __NR_swapon ] = (syscall_handler_t *) sys_swapon, [ __NR_reboot ] = (syscall_handler_t *) sys_reboot, [ __NR_munmap ] = (syscall_handler_t *) sys_munmap, [ __NR_truncate ] = (syscall_handler_t *) sys_truncate, [ __NR_ftruncate ] = (syscall_handler_t *) sys_ftruncate, [ __NR_fchmod ] = (syscall_handler_t *) sys_fchmod, - [ __NR_fchown ] = (syscall_handler_t *) sys_fchown16, + [ __NR_fchown ] = (syscall_handler_t *) sys_fchown, [ __NR_getpriority ] = (syscall_handler_t *) sys_getpriority, [ __NR_setpriority ] = (syscall_handler_t *) sys_setpriority, [ __NR_statfs ] = (syscall_handler_t *) sys_statfs, @@ -161,8 +156,8 @@ syscall_handler_t *sys_call_table[] = { [ __NR_sysfs ] = (syscall_handler_t *) sys_sysfs, [ __NR_personality ] = (syscall_handler_t *) sys_personality, [ __NR_afs_syscall ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_setfsuid ] = (syscall_handler_t *) sys_setfsuid16, - [ __NR_setfsgid ] = (syscall_handler_t *) sys_setfsgid16, + [ __NR_setfsuid ] = (syscall_handler_t *) sys_setfsuid, + [ __NR_setfsgid ] = (syscall_handler_t *) sys_setfsgid, [ __NR_getdents ] = (syscall_handler_t *) sys_getdents, [ __NR_flock ] = (syscall_handler_t *) sys_flock, [ __NR_msync ] = (syscall_handler_t *) sys_msync, @@ -185,13 +180,13 @@ syscall_handler_t *sys_call_table[] = { [ __NR_sched_rr_get_interval ] = (syscall_handler_t *) sys_sched_rr_get_interval, [ __NR_nanosleep ] = (syscall_handler_t *) sys_nanosleep, [ __NR_mremap ] = (syscall_handler_t *) sys_mremap, - [ __NR_setresuid ] = (syscall_handler_t *) sys_setresuid16, - [ __NR_getresuid ] = (syscall_handler_t *) sys_getresuid16, + [ __NR_setresuid ] = (syscall_handler_t *) sys_setresuid, + [ __NR_getresuid ] = (syscall_handler_t *) sys_getresuid, [ __NR_query_module ] = (syscall_handler_t *) sys_ni_syscall, [ __NR_poll ] = (syscall_handler_t *) sys_poll, - [ __NR_nfsservctl ] = (syscall_handler_t *) NFSSERVCTL, - [ __NR_setresgid ] = (syscall_handler_t *) sys_setresgid16, - [ __NR_getresgid ] = (syscall_handler_t *) sys_getresgid16, + [ __NR_nfsservctl ] = (syscall_handler_t *) sys_nfsservctl, + [ __NR_setresgid ] = (syscall_handler_t *) sys_setresgid, + [ __NR_getresgid ] = (syscall_handler_t *) sys_getresgid, [ __NR_prctl ] = (syscall_handler_t *) sys_prctl, [ __NR_rt_sigreturn ] = (syscall_handler_t *) sys_rt_sigreturn, [ __NR_rt_sigaction ] = (syscall_handler_t *) sys_rt_sigaction, @@ -202,12 +197,12 @@ syscall_handler_t *sys_call_table[] = { [ __NR_rt_sigsuspend ] = (syscall_handler_t *) sys_rt_sigsuspend, [ __NR_pread64 ] = (syscall_handler_t *) sys_pread64, [ __NR_pwrite64 ] = (syscall_handler_t *) sys_pwrite64, - [ __NR_chown ] = (syscall_handler_t *) sys_chown16, + [ __NR_chown ] = (syscall_handler_t *) sys_chown, [ __NR_getcwd ] = (syscall_handler_t *) sys_getcwd, [ __NR_capget ] = (syscall_handler_t *) sys_capget, [ __NR_capset ] = (syscall_handler_t *) sys_capset, [ __NR_sigaltstack ] = (syscall_handler_t *) sys_sigaltstack, - [ __NR_sendfile ] = (syscall_handler_t *) sys_sendfile, + [ __NR_sendfile ] = (syscall_handler_t *) sys_sendfile64, [ __NR_getpmsg ] = (syscall_handler_t *) sys_ni_syscall, [ __NR_putpmsg ] = (syscall_handler_t *) sys_ni_syscall, [ __NR_vfork ] = (syscall_handler_t *) sys_vfork, -- cgit v1.2.3-70-g09d2 From 80f9507886076de0cadfdf2b87701e68156829c2 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:55 -0700 Subject: [PATCH] uml: fix syscall table by including $(SUBARCH)'s one, for x86-64 Reuse asm-x86-64/unistd.h to build our syscall table, like x86-64 already does. Like for i386, we must add some #defines for all the (right!) changes UML does to x86-64 syscall table. Note: I noted a bogus: [ __NR_sched_yield ] = (syscall_handler_t *) yield, while doing this patch (which could only be a workaround for some strange bug, but I would ignore this possibility). I'm changing this without notice. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/include/sysdep-x86_64/syscalls.h | 47 +----- arch/um/sys-x86_64/Makefile | 2 +- arch/um/sys-x86_64/sys_call_table.c | 271 ------------------------------- arch/um/sys-x86_64/syscall_table.c | 59 +++++++ arch/um/sys-x86_64/syscalls.c | 12 ++ 5 files changed, 73 insertions(+), 318 deletions(-) delete mode 100644 arch/um/sys-x86_64/sys_call_table.c create mode 100644 arch/um/sys-x86_64/syscall_table.c (limited to 'arch') diff --git a/arch/um/include/sysdep-x86_64/syscalls.h b/arch/um/include/sysdep-x86_64/syscalls.h index 65a01e3d870..b56b335c351 100644 --- a/arch/um/include/sysdep-x86_64/syscalls.h +++ b/arch/um/include/sysdep-x86_64/syscalls.h @@ -30,51 +30,6 @@ extern syscall_handler_t wrap_sys_shmat; extern syscall_handler_t sys_modify_ldt; extern syscall_handler_t sys_arch_prctl; -#define ARCH_SYSCALLS \ - [ __NR_mmap ] = (syscall_handler_t *) old_mmap, \ - [ __NR_select ] = (syscall_handler_t *) sys_select, \ - [ __NR_mincore ] = (syscall_handler_t *) sys_mincore, \ - [ __NR_madvise ] = (syscall_handler_t *) sys_madvise, \ - [ __NR_shmget ] = (syscall_handler_t *) sys_shmget, \ - [ __NR_shmat ] = (syscall_handler_t *) wrap_sys_shmat, \ - [ __NR_shmctl ] = (syscall_handler_t *) sys_shmctl, \ - [ __NR_semop ] = (syscall_handler_t *) sys_semop, \ - [ __NR_semget ] = (syscall_handler_t *) sys_semget, \ - [ __NR_semctl ] = (syscall_handler_t *) sys_semctl, \ - [ __NR_shmdt ] = (syscall_handler_t *) sys_shmdt, \ - [ __NR_msgget ] = (syscall_handler_t *) sys_msgget, \ - [ __NR_msgsnd ] = (syscall_handler_t *) sys_msgsnd, \ - [ __NR_msgrcv ] = (syscall_handler_t *) sys_msgrcv, \ - [ __NR_msgctl ] = (syscall_handler_t *) sys_msgctl, \ - [ __NR_pivot_root ] = (syscall_handler_t *) sys_pivot_root, \ - [ __NR_tuxcall ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_security ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_epoll_ctl_old ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_epoll_wait_old ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_modify_ldt ] = (syscall_handler_t *) sys_modify_ldt, \ - [ __NR_arch_prctl ] = (syscall_handler_t *) sys_arch_prctl, \ - [ __NR_socket ] = (syscall_handler_t *) sys_socket, \ - [ __NR_connect ] = (syscall_handler_t *) sys_connect, \ - [ __NR_accept ] = (syscall_handler_t *) sys_accept, \ - [ __NR_recvfrom ] = (syscall_handler_t *) sys_recvfrom, \ - [ __NR_recvmsg ] = (syscall_handler_t *) sys_recvmsg, \ - [ __NR_sendmsg ] = (syscall_handler_t *) sys_sendmsg, \ - [ __NR_bind ] = (syscall_handler_t *) sys_bind, \ - [ __NR_listen ] = (syscall_handler_t *) sys_listen, \ - [ __NR_getsockname ] = (syscall_handler_t *) sys_getsockname, \ - [ __NR_getpeername ] = (syscall_handler_t *) sys_getpeername, \ - [ __NR_socketpair ] = (syscall_handler_t *) sys_socketpair, \ - [ __NR_sendto ] = (syscall_handler_t *) sys_sendto, \ - [ __NR_shutdown ] = (syscall_handler_t *) sys_shutdown, \ - [ __NR_setsockopt ] = (syscall_handler_t *) sys_setsockopt, \ - [ __NR_getsockopt ] = (syscall_handler_t *) sys_getsockopt, \ - [ __NR_iopl ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_set_thread_area ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_get_thread_area ] = (syscall_handler_t *) sys_ni_syscall, \ - [ __NR_semtimedop ] = (syscall_handler_t *) sys_semtimedop, \ - [ 251 ] = (syscall_handler_t *) sys_ni_syscall, - -#define LAST_ARCH_SYSCALL 251 -#define NR_syscalls 1024 +#define NR_syscalls (__NR_syscall_max + 1) #endif diff --git a/arch/um/sys-x86_64/Makefile b/arch/um/sys-x86_64/Makefile index 274158e7da3..d7ed2f7908d 100644 --- a/arch/um/sys-x86_64/Makefile +++ b/arch/um/sys-x86_64/Makefile @@ -6,7 +6,7 @@ lib-y = bitops.o bugs.o csum-partial.o delay.o fault.o mem.o memcpy.o \ ptrace.o ptrace_user.o semaphore.o sigcontext.o signal.o \ - syscalls.o sysrq.o thunk.o sys_call_table.o + syscalls.o sysrq.o thunk.o syscall_table.o USER_OBJS := ptrace_user.o sigcontext.o diff --git a/arch/um/sys-x86_64/sys_call_table.c b/arch/um/sys-x86_64/sys_call_table.c deleted file mode 100644 index a31e0dffb27..00000000000 --- a/arch/um/sys-x86_64/sys_call_table.c +++ /dev/null @@ -1,271 +0,0 @@ -/* - * Copyright (C) 2000 Jeff Dike (jdike@karaya.com) - * Copyright 2003 PathScale, Inc. - * Licensed under the GPL - */ - -#include "linux/config.h" -#include "linux/unistd.h" -#include "linux/sys.h" -#include "linux/swap.h" -#include "linux/syscalls.h" -#include "linux/sysctl.h" -#include "asm/signal.h" -#include "sysdep/syscalls.h" -#include "kern_util.h" - -#define LAST_GENERIC_SYSCALL __NR_keyctl - -#if LAST_GENERIC_SYSCALL > LAST_ARCH_SYSCALL -#define LAST_SYSCALL LAST_GENERIC_SYSCALL -#else -#define LAST_SYSCALL LAST_ARCH_SYSCALL -#endif - -extern syscall_handler_t sys_fork; -extern syscall_handler_t sys_execve; -extern syscall_handler_t um_time; -extern syscall_handler_t um_stime; -extern syscall_handler_t sys_pipe; -extern syscall_handler_t sys_olduname; -extern syscall_handler_t sys_sigaction; -extern syscall_handler_t sys_sigsuspend; -extern syscall_handler_t old_readdir; -extern syscall_handler_t sys_uname; -extern syscall_handler_t sys_ipc; -extern syscall_handler_t sys_sigreturn; -extern syscall_handler_t sys_clone; -extern syscall_handler_t sys_rt_sigreturn; -extern syscall_handler_t sys_sigaltstack; -extern syscall_handler_t sys_vfork; -extern syscall_handler_t old_select; -extern syscall_handler_t sys_modify_ldt; -extern syscall_handler_t sys_rt_sigsuspend; -extern syscall_handler_t sys_mbind; -extern syscall_handler_t sys_get_mempolicy; -extern syscall_handler_t sys_set_mempolicy; -extern syscall_handler_t sys_sys_setaltroot; - -/* On X86-64 all syscalls are aware of 32-bit [ug]ids, so I had to fix this!*/ -syscall_handler_t *sys_call_table[] = { - [ __NR_restart_syscall ] = (syscall_handler_t *) sys_restart_syscall, - [ __NR_exit ] = (syscall_handler_t *) sys_exit, - [ __NR_fork ] = (syscall_handler_t *) sys_fork, - [ __NR_read ] = (syscall_handler_t *) sys_read, - [ __NR_write ] = (syscall_handler_t *) sys_write, - - /* These three are declared differently in asm/unistd.h */ - [ __NR_open ] = (syscall_handler_t *) sys_open, - [ __NR_close ] = (syscall_handler_t *) sys_close, - [ __NR_creat ] = (syscall_handler_t *) sys_creat, - [ __NR_link ] = (syscall_handler_t *) sys_link, - [ __NR_unlink ] = (syscall_handler_t *) sys_unlink, - [ __NR_execve ] = (syscall_handler_t *) sys_execve, - - /* declared differently in kern_util.h */ - [ __NR_chdir ] = (syscall_handler_t *) sys_chdir, - [ __NR_time ] = um_time, - [ __NR_mknod ] = (syscall_handler_t *) sys_mknod, - [ __NR_chmod ] = (syscall_handler_t *) sys_chmod, - [ __NR_lchown ] = (syscall_handler_t *) sys_lchown, - [ __NR_lseek ] = (syscall_handler_t *) sys_lseek, - [ __NR_getpid ] = (syscall_handler_t *) sys_getpid, - [ __NR_mount ] = (syscall_handler_t *) sys_mount, - [ __NR_setuid ] = (syscall_handler_t *) sys_setuid, - [ __NR_getuid ] = (syscall_handler_t *) sys_getuid, - [ __NR_ptrace ] = (syscall_handler_t *) sys_ptrace, - [ __NR_alarm ] = (syscall_handler_t *) sys_alarm, - [ __NR_pause ] = (syscall_handler_t *) sys_pause, - [ __NR_utime ] = (syscall_handler_t *) sys_utime, - [ __NR_access ] = (syscall_handler_t *) sys_access, - [ __NR_sync ] = (syscall_handler_t *) sys_sync, - [ __NR_kill ] = (syscall_handler_t *) sys_kill, - [ __NR_rename ] = (syscall_handler_t *) sys_rename, - [ __NR_mkdir ] = (syscall_handler_t *) sys_mkdir, - [ __NR_rmdir ] = (syscall_handler_t *) sys_rmdir, - - /* Declared differently in asm/unistd.h */ - [ __NR_dup ] = (syscall_handler_t *) sys_dup, - [ __NR_pipe ] = (syscall_handler_t *) sys_pipe, - [ __NR_times ] = (syscall_handler_t *) sys_times, - [ __NR_brk ] = (syscall_handler_t *) sys_brk, - [ __NR_setgid ] = (syscall_handler_t *) sys_setgid, - [ __NR_getgid ] = (syscall_handler_t *) sys_getgid, - [ __NR_geteuid ] = (syscall_handler_t *) sys_geteuid, - [ __NR_getegid ] = (syscall_handler_t *) sys_getegid, - [ __NR_acct ] = (syscall_handler_t *) sys_acct, - [ __NR_umount2 ] = (syscall_handler_t *) sys_umount, - [ __NR_ioctl ] = (syscall_handler_t *) sys_ioctl, - [ __NR_fcntl ] = (syscall_handler_t *) sys_fcntl, - [ __NR_setpgid ] = (syscall_handler_t *) sys_setpgid, - [ __NR_umask ] = (syscall_handler_t *) sys_umask, - [ __NR_chroot ] = (syscall_handler_t *) sys_chroot, - [ __NR_ustat ] = (syscall_handler_t *) sys_ustat, - [ __NR_dup2 ] = (syscall_handler_t *) sys_dup2, - [ __NR_getppid ] = (syscall_handler_t *) sys_getppid, - [ __NR_getpgrp ] = (syscall_handler_t *) sys_getpgrp, - [ __NR_setsid ] = (syscall_handler_t *) sys_setsid, - [ __NR_setreuid ] = (syscall_handler_t *) sys_setreuid, - [ __NR_setregid ] = (syscall_handler_t *) sys_setregid, - [ __NR_sethostname ] = (syscall_handler_t *) sys_sethostname, - [ __NR_setrlimit ] = (syscall_handler_t *) sys_setrlimit, - [ __NR_getrlimit ] = (syscall_handler_t *) sys_getrlimit, - [ __NR_getrusage ] = (syscall_handler_t *) sys_getrusage, - [ __NR_gettimeofday ] = (syscall_handler_t *) sys_gettimeofday, - [ __NR_settimeofday ] = (syscall_handler_t *) sys_settimeofday, - [ __NR_getgroups ] = (syscall_handler_t *) sys_getgroups, - [ __NR_setgroups ] = (syscall_handler_t *) sys_setgroups, - [ __NR_symlink ] = (syscall_handler_t *) sys_symlink, - [ __NR_readlink ] = (syscall_handler_t *) sys_readlink, - [ __NR_uselib ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_swapon ] = (syscall_handler_t *) sys_swapon, - [ __NR_reboot ] = (syscall_handler_t *) sys_reboot, - [ __NR_munmap ] = (syscall_handler_t *) sys_munmap, - [ __NR_truncate ] = (syscall_handler_t *) sys_truncate, - [ __NR_ftruncate ] = (syscall_handler_t *) sys_ftruncate, - [ __NR_fchmod ] = (syscall_handler_t *) sys_fchmod, - [ __NR_fchown ] = (syscall_handler_t *) sys_fchown, - [ __NR_getpriority ] = (syscall_handler_t *) sys_getpriority, - [ __NR_setpriority ] = (syscall_handler_t *) sys_setpriority, - [ __NR_statfs ] = (syscall_handler_t *) sys_statfs, - [ __NR_fstatfs ] = (syscall_handler_t *) sys_fstatfs, - [ __NR_ioperm ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_syslog ] = (syscall_handler_t *) sys_syslog, - [ __NR_setitimer ] = (syscall_handler_t *) sys_setitimer, - [ __NR_getitimer ] = (syscall_handler_t *) sys_getitimer, - [ __NR_stat ] = (syscall_handler_t *) sys_newstat, - [ __NR_lstat ] = (syscall_handler_t *) sys_newlstat, - [ __NR_fstat ] = (syscall_handler_t *) sys_newfstat, - [ __NR_vhangup ] = (syscall_handler_t *) sys_vhangup, - [ __NR_wait4 ] = (syscall_handler_t *) sys_wait4, - [ __NR_swapoff ] = (syscall_handler_t *) sys_swapoff, - [ __NR_sysinfo ] = (syscall_handler_t *) sys_sysinfo, - [ __NR_fsync ] = (syscall_handler_t *) sys_fsync, - [ __NR_clone ] = (syscall_handler_t *) sys_clone, - [ __NR_setdomainname ] = (syscall_handler_t *) sys_setdomainname, - [ __NR_uname ] = (syscall_handler_t *) sys_newuname, - [ __NR_adjtimex ] = (syscall_handler_t *) sys_adjtimex, - [ __NR_mprotect ] = (syscall_handler_t *) sys_mprotect, - [ __NR_create_module ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_init_module ] = (syscall_handler_t *) sys_init_module, - [ __NR_delete_module ] = (syscall_handler_t *) sys_delete_module, - [ __NR_get_kernel_syms ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_quotactl ] = (syscall_handler_t *) sys_quotactl, - [ __NR_getpgid ] = (syscall_handler_t *) sys_getpgid, - [ __NR_fchdir ] = (syscall_handler_t *) sys_fchdir, - [ __NR_sysfs ] = (syscall_handler_t *) sys_sysfs, - [ __NR_personality ] = (syscall_handler_t *) sys_personality, - [ __NR_afs_syscall ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_setfsuid ] = (syscall_handler_t *) sys_setfsuid, - [ __NR_setfsgid ] = (syscall_handler_t *) sys_setfsgid, - [ __NR_getdents ] = (syscall_handler_t *) sys_getdents, - [ __NR_flock ] = (syscall_handler_t *) sys_flock, - [ __NR_msync ] = (syscall_handler_t *) sys_msync, - [ __NR_readv ] = (syscall_handler_t *) sys_readv, - [ __NR_writev ] = (syscall_handler_t *) sys_writev, - [ __NR_getsid ] = (syscall_handler_t *) sys_getsid, - [ __NR_fdatasync ] = (syscall_handler_t *) sys_fdatasync, - [ __NR__sysctl ] = (syscall_handler_t *) sys_sysctl, - [ __NR_mlock ] = (syscall_handler_t *) sys_mlock, - [ __NR_munlock ] = (syscall_handler_t *) sys_munlock, - [ __NR_mlockall ] = (syscall_handler_t *) sys_mlockall, - [ __NR_munlockall ] = (syscall_handler_t *) sys_munlockall, - [ __NR_sched_setparam ] = (syscall_handler_t *) sys_sched_setparam, - [ __NR_sched_getparam ] = (syscall_handler_t *) sys_sched_getparam, - [ __NR_sched_setscheduler ] = (syscall_handler_t *) sys_sched_setscheduler, - [ __NR_sched_getscheduler ] = (syscall_handler_t *) sys_sched_getscheduler, - [ __NR_sched_yield ] = (syscall_handler_t *) yield, - [ __NR_sched_get_priority_max ] = (syscall_handler_t *) sys_sched_get_priority_max, - [ __NR_sched_get_priority_min ] = (syscall_handler_t *) sys_sched_get_priority_min, - [ __NR_sched_rr_get_interval ] = (syscall_handler_t *) sys_sched_rr_get_interval, - [ __NR_nanosleep ] = (syscall_handler_t *) sys_nanosleep, - [ __NR_mremap ] = (syscall_handler_t *) sys_mremap, - [ __NR_setresuid ] = (syscall_handler_t *) sys_setresuid, - [ __NR_getresuid ] = (syscall_handler_t *) sys_getresuid, - [ __NR_query_module ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_poll ] = (syscall_handler_t *) sys_poll, - [ __NR_nfsservctl ] = (syscall_handler_t *) sys_nfsservctl, - [ __NR_setresgid ] = (syscall_handler_t *) sys_setresgid, - [ __NR_getresgid ] = (syscall_handler_t *) sys_getresgid, - [ __NR_prctl ] = (syscall_handler_t *) sys_prctl, - [ __NR_rt_sigreturn ] = (syscall_handler_t *) sys_rt_sigreturn, - [ __NR_rt_sigaction ] = (syscall_handler_t *) sys_rt_sigaction, - [ __NR_rt_sigprocmask ] = (syscall_handler_t *) sys_rt_sigprocmask, - [ __NR_rt_sigpending ] = (syscall_handler_t *) sys_rt_sigpending, - [ __NR_rt_sigtimedwait ] = (syscall_handler_t *) sys_rt_sigtimedwait, - [ __NR_rt_sigqueueinfo ] = (syscall_handler_t *) sys_rt_sigqueueinfo, - [ __NR_rt_sigsuspend ] = (syscall_handler_t *) sys_rt_sigsuspend, - [ __NR_pread64 ] = (syscall_handler_t *) sys_pread64, - [ __NR_pwrite64 ] = (syscall_handler_t *) sys_pwrite64, - [ __NR_chown ] = (syscall_handler_t *) sys_chown, - [ __NR_getcwd ] = (syscall_handler_t *) sys_getcwd, - [ __NR_capget ] = (syscall_handler_t *) sys_capget, - [ __NR_capset ] = (syscall_handler_t *) sys_capset, - [ __NR_sigaltstack ] = (syscall_handler_t *) sys_sigaltstack, - [ __NR_sendfile ] = (syscall_handler_t *) sys_sendfile64, - [ __NR_getpmsg ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_putpmsg ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_vfork ] = (syscall_handler_t *) sys_vfork, - [ __NR_getdents64 ] = (syscall_handler_t *) sys_getdents64, - [ __NR_gettid ] = (syscall_handler_t *) sys_gettid, - [ __NR_readahead ] = (syscall_handler_t *) sys_readahead, - [ __NR_setxattr ] = (syscall_handler_t *) sys_setxattr, - [ __NR_lsetxattr ] = (syscall_handler_t *) sys_lsetxattr, - [ __NR_fsetxattr ] = (syscall_handler_t *) sys_fsetxattr, - [ __NR_getxattr ] = (syscall_handler_t *) sys_getxattr, - [ __NR_lgetxattr ] = (syscall_handler_t *) sys_lgetxattr, - [ __NR_fgetxattr ] = (syscall_handler_t *) sys_fgetxattr, - [ __NR_listxattr ] = (syscall_handler_t *) sys_listxattr, - [ __NR_llistxattr ] = (syscall_handler_t *) sys_llistxattr, - [ __NR_flistxattr ] = (syscall_handler_t *) sys_flistxattr, - [ __NR_removexattr ] = (syscall_handler_t *) sys_removexattr, - [ __NR_lremovexattr ] = (syscall_handler_t *) sys_lremovexattr, - [ __NR_fremovexattr ] = (syscall_handler_t *) sys_fremovexattr, - [ __NR_tkill ] = (syscall_handler_t *) sys_tkill, - [ __NR_futex ] = (syscall_handler_t *) sys_futex, - [ __NR_sched_setaffinity ] = (syscall_handler_t *) sys_sched_setaffinity, - [ __NR_sched_getaffinity ] = (syscall_handler_t *) sys_sched_getaffinity, - [ __NR_io_setup ] = (syscall_handler_t *) sys_io_setup, - [ __NR_io_destroy ] = (syscall_handler_t *) sys_io_destroy, - [ __NR_io_getevents ] = (syscall_handler_t *) sys_io_getevents, - [ __NR_io_submit ] = (syscall_handler_t *) sys_io_submit, - [ __NR_io_cancel ] = (syscall_handler_t *) sys_io_cancel, - [ __NR_exit_group ] = (syscall_handler_t *) sys_exit_group, - [ __NR_lookup_dcookie ] = (syscall_handler_t *) sys_lookup_dcookie, - [ __NR_epoll_create ] = (syscall_handler_t *) sys_epoll_create, - [ __NR_epoll_ctl ] = (syscall_handler_t *) sys_epoll_ctl, - [ __NR_epoll_wait ] = (syscall_handler_t *) sys_epoll_wait, - [ __NR_remap_file_pages ] = (syscall_handler_t *) sys_remap_file_pages, - [ __NR_set_tid_address ] = (syscall_handler_t *) sys_set_tid_address, - [ __NR_timer_create ] = (syscall_handler_t *) sys_timer_create, - [ __NR_timer_settime ] = (syscall_handler_t *) sys_timer_settime, - [ __NR_timer_gettime ] = (syscall_handler_t *) sys_timer_gettime, - [ __NR_timer_getoverrun ] = (syscall_handler_t *) sys_timer_getoverrun, - [ __NR_timer_delete ] = (syscall_handler_t *) sys_timer_delete, - [ __NR_clock_settime ] = (syscall_handler_t *) sys_clock_settime, - [ __NR_clock_gettime ] = (syscall_handler_t *) sys_clock_gettime, - [ __NR_clock_getres ] = (syscall_handler_t *) sys_clock_getres, - [ __NR_clock_nanosleep ] = (syscall_handler_t *) sys_clock_nanosleep, - [ __NR_tgkill ] = (syscall_handler_t *) sys_tgkill, - [ __NR_utimes ] = (syscall_handler_t *) sys_utimes, - [ __NR_fadvise64 ] = (syscall_handler_t *) sys_fadvise64, - [ __NR_vserver ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_mbind ] = (syscall_handler_t *) sys_mbind, - [ __NR_get_mempolicy ] = (syscall_handler_t *) sys_get_mempolicy, - [ __NR_set_mempolicy ] = (syscall_handler_t *) sys_set_mempolicy, - [ __NR_mq_open ] = (syscall_handler_t *) sys_mq_open, - [ __NR_mq_unlink ] = (syscall_handler_t *) sys_mq_unlink, - [ __NR_mq_timedsend ] = (syscall_handler_t *) sys_mq_timedsend, - [ __NR_mq_timedreceive ] = (syscall_handler_t *) sys_mq_timedreceive, - [ __NR_mq_notify ] = (syscall_handler_t *) sys_mq_notify, - [ __NR_mq_getsetattr ] = (syscall_handler_t *) sys_mq_getsetattr, - [ __NR_kexec_load ] = (syscall_handler_t *) sys_ni_syscall, - [ __NR_waitid ] = (syscall_handler_t *) sys_waitid, - [ __NR_add_key ] = (syscall_handler_t *) sys_add_key, - [ __NR_request_key ] = (syscall_handler_t *) sys_request_key, - [ __NR_keyctl ] = (syscall_handler_t *) sys_keyctl, - - ARCH_SYSCALLS - [ LAST_SYSCALL + 1 ... NR_syscalls ] = - (syscall_handler_t *) sys_ni_syscall -}; diff --git a/arch/um/sys-x86_64/syscall_table.c b/arch/um/sys-x86_64/syscall_table.c new file mode 100644 index 00000000000..34b2e842864 --- /dev/null +++ b/arch/um/sys-x86_64/syscall_table.c @@ -0,0 +1,59 @@ +/* System call table for UML/x86-64, copied from arch/x86_64/kernel/syscall.c + * with some changes for UML. */ + +#include +#include +#include +#include + +#define __NO_STUBS + +/* Below you can see, in terms of #define's, the differences between the x86-64 + * and the UML syscall table. */ + +/* Not going to be implemented by UML, since we have no hardware. */ +#define stub_iopl sys_ni_syscall +#define sys_ioperm sys_ni_syscall + +/* The UML TLS problem. Note that x86_64 does not implement this, so the below + * is needed only for the ia32 compatibility. */ +/*#define sys_set_thread_area sys_ni_syscall +#define sys_get_thread_area sys_ni_syscall*/ + +/* For __NR_time. The x86-64 name hopefully will change from sys_time64 to + * sys_time (since the current situation is bogus). I've sent a patch to cleanup + * this. Remove below the obsoleted line. */ +#define sys_time64 um_time +#define sys_time um_time + +/* On UML we call it this way ("old" means it's not mmap2) */ +#define sys_mmap old_mmap +/* On x86-64 sys_uname is actually sys_newuname plus a compatibility trick. + * See arch/x86_64/kernel/sys_x86_64.c */ +#define sys_uname sys_uname64 + +#define stub_clone sys_clone +#define stub_fork sys_fork +#define stub_vfork sys_vfork +#define stub_execve sys_execve +#define stub_rt_sigsuspend sys_rt_sigsuspend +#define stub_sigaltstack sys_sigaltstack +#define stub_rt_sigreturn sys_rt_sigreturn + +#define __SYSCALL(nr, sym) extern asmlinkage void sym(void) ; +#undef _ASM_X86_64_UNISTD_H_ +#include + +#undef __SYSCALL +#define __SYSCALL(nr, sym) [ nr ] = sym, +#undef _ASM_X86_64_UNISTD_H_ + +typedef void (*sys_call_ptr_t)(void); + +extern void sys_ni_syscall(void); + +sys_call_ptr_t sys_call_table[__NR_syscall_max+1] __cacheline_aligned = { + /* Smells like a like a compiler bug -- it doesn't work when the & below is removed. */ + [0 ... __NR_syscall_max] = &sys_ni_syscall, +#include +}; diff --git a/arch/um/sys-x86_64/syscalls.c b/arch/um/sys-x86_64/syscalls.c index 68205a03364..15768c96ceb 100644 --- a/arch/um/sys-x86_64/syscalls.c +++ b/arch/um/sys-x86_64/syscalls.c @@ -14,6 +14,7 @@ #include "asm/prctl.h" /* XXX This should get the constants from libc */ #include "choose-mode.h" +/* XXX: copied from x86-64: arch/x86_64/kernel/sys_x86_64.c */ asmlinkage long wrap_sys_shmat(int shmid, char __user *shmaddr, int shmflg) { unsigned long raddr; @@ -21,6 +22,17 @@ asmlinkage long wrap_sys_shmat(int shmid, char __user *shmaddr, int shmflg) return do_shmat(shmid, shmaddr, shmflg, &raddr) ?: (long) raddr; } +asmlinkage long sys_uname64(struct new_utsname __user * name) +{ + int err; + down_read(&uts_sem); + err = copy_to_user(name, &system_utsname, sizeof (*name)); + up_read(&uts_sem); + if (personality(current->personality) == PER_LINUX32) + err |= copy_to_user(&name->machine, "i686", 5); + return err ? -EFAULT : 0; +} + #ifdef CONFIG_MODE_TT extern int modify_ldt(int func, void *ptr, unsigned long bytecount); -- cgit v1.2.3-70-g09d2 From b97b77cca627b2e3e6d0015e2e14b1d1c12de0c8 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:56 -0700 Subject: [PATCH] uml: redo console locking Fix some console locking problems (including scheduling in atomic) and various reorderings and cleanup in that code. Not yet ready for 2.6.12 probably. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/drivers/chan_kern.c | 16 +- arch/um/drivers/line.c | 318 +++++++++++++++++++++++++--------------- arch/um/drivers/ssl.c | 25 +--- arch/um/drivers/stdio_console.c | 19 +-- arch/um/include/line.h | 36 ++--- 5 files changed, 239 insertions(+), 175 deletions(-) (limited to 'arch') diff --git a/arch/um/drivers/chan_kern.c b/arch/um/drivers/chan_kern.c index 1f77deb3fd2..0150038af79 100644 --- a/arch/um/drivers/chan_kern.c +++ b/arch/um/drivers/chan_kern.c @@ -22,7 +22,7 @@ #ifdef CONFIG_NOCONFIG_CHAN static void *not_configged_init(char *str, int device, struct chan_opts *opts) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); return(NULL); } @@ -30,27 +30,27 @@ static void *not_configged_init(char *str, int device, struct chan_opts *opts) static int not_configged_open(int input, int output, int primary, void *data, char **dev_out) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); return(-ENODEV); } static void not_configged_close(int fd, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); } static int not_configged_read(int fd, char *c_out, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); return(-EIO); } static int not_configged_write(int fd, const char *buf, int len, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); return(-EIO); } @@ -58,7 +58,7 @@ static int not_configged_write(int fd, const char *buf, int len, void *data) static int not_configged_console_write(int fd, const char *buf, int len, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); return(-EIO); } @@ -66,14 +66,14 @@ static int not_configged_console_write(int fd, const char *buf, int len, static int not_configged_window_size(int fd, void *data, unsigned short *rows, unsigned short *cols) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); return(-ENODEV); } static void not_configged_free(void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + printf(KERN_ERR "Using a channel type which is configured out of " "UML\n"); } diff --git a/arch/um/drivers/line.c b/arch/um/drivers/line.c index 6924f273ced..d0f97127adf 100644 --- a/arch/um/drivers/line.c +++ b/arch/um/drivers/line.c @@ -39,19 +39,69 @@ static void line_timer_cb(void *arg) line_interrupt(line->driver->read_irq, arg, NULL); } -static int write_room(struct line *dev) +/* Returns the free space inside the ring buffer of this line. + * + * Should be called while holding line->lock (this does not modify datas). + */ +static int write_room(struct line *line) { int n; - if (dev->buffer == NULL) - return (LINE_BUFSIZE - 1); + if (line->buffer == NULL) + return LINE_BUFSIZE - 1; + + /* This is for the case where the buffer is wrapped! */ + n = line->head - line->tail; - n = dev->head - dev->tail; if (n <= 0) - n = LINE_BUFSIZE + n; - return (n - 1); + n = LINE_BUFSIZE + n; /* The other case */ + return n - 1; +} + +int line_write_room(struct tty_struct *tty) +{ + struct line *line = tty->driver_data; + unsigned long flags; + int room; + + if (tty->stopped) + return 0; + + spin_lock_irqsave(&line->lock, flags); + room = write_room(line); + spin_unlock_irqrestore(&line->lock, flags); + + /*XXX: Warning to remove */ + if (0 == room) + printk(KERN_DEBUG "%s: %s: no room left in buffer\n", + __FUNCTION__,tty->name); + return room; } +int line_chars_in_buffer(struct tty_struct *tty) +{ + struct line *line = tty->driver_data; + unsigned long flags; + int ret; + + spin_lock_irqsave(&line->lock, flags); + + /*write_room subtracts 1 for the needed NULL, so we readd it.*/ + ret = LINE_BUFSIZE - (write_room(line) + 1); + spin_unlock_irqrestore(&line->lock, flags); + + return ret; +} + +/* + * This copies the content of buf into the circular buffer associated with + * this line. + * The return value is the number of characters actually copied, i.e. the ones + * for which there was space: this function is not supposed to ever flush out + * the circular buffer. + * + * Must be called while holding line->lock! + */ static int buffer_data(struct line *line, const char *buf, int len) { int end, room; @@ -70,48 +120,95 @@ static int buffer_data(struct line *line, const char *buf, int len) len = (len > room) ? room : len; end = line->buffer + LINE_BUFSIZE - line->tail; - if(len < end){ + + if (len < end){ memcpy(line->tail, buf, len); line->tail += len; - } - else { + } else { + /* The circular buffer is wrapping */ memcpy(line->tail, buf, end); buf += end; memcpy(line->buffer, buf, len - end); line->tail = line->buffer + len - end; } - return(len); + return len; } +/* + * Flushes the ring buffer to the output channels. That is, write_chan is + * called, passing it line->head as buffer, and an appropriate count. + * + * On exit, returns 1 when the buffer is empty, + * 0 when the buffer is not empty on exit, + * and -errno when an error occurred. + * + * Must be called while holding line->lock!*/ static int flush_buffer(struct line *line) { int n, count; if ((line->buffer == NULL) || (line->head == line->tail)) - return(1); + return 1; if (line->tail < line->head) { + /* line->buffer + LINE_BUFSIZE is the end of the buffer! */ count = line->buffer + LINE_BUFSIZE - line->head; + n = write_chan(&line->chan_list, line->head, count, line->driver->write_irq); if (n < 0) - return(n); - if (n == count) + return n; + if (n == count) { + /* We have flushed from ->head to buffer end, now we + * must flush only from the beginning to ->tail.*/ line->head = line->buffer; - else { + } else { line->head += n; - return(0); + return 0; } } count = line->tail - line->head; n = write_chan(&line->chan_list, line->head, count, line->driver->write_irq); - if(n < 0) return(n); + + if(n < 0) + return n; line->head += n; - return(line->head == line->tail); + return line->head == line->tail; +} + +void line_flush_buffer(struct tty_struct *tty) +{ + struct line *line = tty->driver_data; + unsigned long flags; + int err; + + /*XXX: copied from line_write, verify if it is correct!*/ + if(tty->stopped) + return; + //return 0; + + spin_lock_irqsave(&line->lock, flags); + err = flush_buffer(line); + /*if (err == 1) + err = 0;*/ + spin_unlock_irqrestore(&line->lock, flags); + //return err; +} + +/* We map both ->flush_chars and ->put_char (which go in pair) onto ->flush_buffer + * and ->write. Hope it's not that bad.*/ +void line_flush_chars(struct tty_struct *tty) +{ + line_flush_buffer(tty); +} + +void line_put_char(struct tty_struct *tty, unsigned char ch) +{ + line_write(tty, &ch, sizeof(ch)); } int line_write(struct tty_struct *tty, const unsigned char *buf, int len) @@ -120,38 +217,31 @@ int line_write(struct tty_struct *tty, const unsigned char *buf, int len) unsigned long flags; int n, err, ret = 0; - if(tty->stopped) return 0; + if(tty->stopped) + return 0; - down(&line->sem); - if(line->head != line->tail){ - local_irq_save(flags); + spin_lock_irqsave(&line->lock, flags); + if (line->head != line->tail) { ret = buffer_data(line, buf, len); err = flush_buffer(line); - local_irq_restore(flags); - if(err <= 0 && (err != -EAGAIN || !ret)) + if (err <= 0 && (err != -EAGAIN || !ret)) ret = err; - } - else { + } else { n = write_chan(&line->chan_list, buf, len, line->driver->write_irq); - if(n < 0){ + if (n < 0) { ret = n; goto out_up; } len -= n; ret += n; - if(len > 0) + if (len > 0) ret += buffer_data(line, buf + n, len); } - out_up: - up(&line->sem); - return(ret); -} - -void line_put_char(struct tty_struct *tty, unsigned char ch) -{ - line_write(tty, &ch, sizeof(ch)); +out_up: + spin_unlock_irqrestore(&line->lock, flags); + return ret; } void line_set_termios(struct tty_struct *tty, struct termios * old) @@ -159,11 +249,6 @@ void line_set_termios(struct tty_struct *tty, struct termios * old) /* nothing */ } -int line_chars_in_buffer(struct tty_struct *tty) -{ - return 0; -} - static struct { int cmd; char *level; @@ -250,7 +335,7 @@ int line_ioctl(struct tty_struct *tty, struct file * file, ret = -ENOIOCTLCMD; break; } - return(ret); + return ret; } static irqreturn_t line_write_interrupt(int irq, void *data, @@ -260,18 +345,23 @@ static irqreturn_t line_write_interrupt(int irq, void *data, struct line *line = tty->driver_data; int err; + /* Interrupts are enabled here because we registered the interrupt with + * SA_INTERRUPT (see line_setup_irq).*/ + + spin_lock_irq(&line->lock); err = flush_buffer(line); - if(err == 0) - return(IRQ_NONE); - else if(err < 0){ + if (err == 0) { + return IRQ_NONE; + } else if(err < 0) { line->head = line->buffer; line->tail = line->buffer; } + spin_unlock_irq(&line->lock); if(tty == NULL) - return(IRQ_NONE); + return IRQ_NONE; - if(test_bit(TTY_DO_WRITE_WAKEUP, &tty->flags) && + if (test_bit(TTY_DO_WRITE_WAKEUP, &tty->flags) && (tty->ldisc.write_wakeup != NULL)) (tty->ldisc.write_wakeup)(tty); @@ -281,9 +371,9 @@ static irqreturn_t line_write_interrupt(int irq, void *data, * writes. */ - if(waitqueue_active(&tty->write_wait)) + if (waitqueue_active(&tty->write_wait)) wake_up_interruptible(&tty->write_wait); - return(IRQ_HANDLED); + return IRQ_HANDLED; } int line_setup_irq(int fd, int input, int output, struct tty_struct *tty) @@ -292,15 +382,18 @@ int line_setup_irq(int fd, int input, int output, struct tty_struct *tty) struct line_driver *driver = line->driver; int err = 0, flags = SA_INTERRUPT | SA_SHIRQ | SA_SAMPLE_RANDOM; - if(input) err = um_request_irq(driver->read_irq, fd, IRQ_READ, + if (input) + err = um_request_irq(driver->read_irq, fd, IRQ_READ, line_interrupt, flags, driver->read_irq_name, tty); - if(err) return(err); - if(output) err = um_request_irq(driver->write_irq, fd, IRQ_WRITE, + if (err) + return err; + if (output) + err = um_request_irq(driver->write_irq, fd, IRQ_WRITE, line_write_interrupt, flags, driver->write_irq_name, tty); line->have_irq = 1; - return(err); + return err; } void line_disable(struct tty_struct *tty, int current_irq) @@ -336,7 +429,9 @@ int line_open(struct line *lines, struct tty_struct *tty, line = &lines[tty->index]; tty->driver_data = line; - down(&line->sem); + /* The IRQ which takes this lock is not yet enabled and won't be run + * before the end, so we don't need to use spin_lock_irq.*/ + spin_lock(&line->lock); if (tty->count == 1) { if (!line->valid) { err = -ENODEV; @@ -349,6 +444,7 @@ int line_open(struct line *lines, struct tty_struct *tty, err = open_chan(&line->chan_list); if(err) goto out; } + /* Here the interrupt is registered.*/ enable_chan(&line->chan_list, tty); INIT_WORK(&line->task, line_timer_cb, tty); } @@ -362,21 +458,27 @@ int line_open(struct line *lines, struct tty_struct *tty, line->count++; out: - up(&line->sem); - return(err); + spin_unlock(&line->lock); + return err; } void line_close(struct tty_struct *tty, struct file * filp) { struct line *line = tty->driver_data; - down(&line->sem); + /* XXX: I assume this should be called in process context, not with interrupt + * disabled!*/ + spin_lock_irq(&line->lock); + + /* We ignore the error anyway! */ + flush_buffer(line); + line->count--; if (tty->count == 1) { line_disable(tty, -1); tty->driver_data = NULL; } - up(&line->sem); + spin_unlock_irq(&line->lock); } void close_lines(struct line *lines, int nlines) @@ -387,31 +489,41 @@ void close_lines(struct line *lines, int nlines) close_chan(&lines[i].chan_list); } -int line_setup(struct line *lines, int num, char *init, int all_allowed) +/* Common setup code for both startup command line and mconsole initialization. + * @lines contains the the array (of size @num) to modify; + * @init is the setup string; + * @all_allowed is a boolean saying if we can setup the whole @lines + * at once. For instance, it will be usually true for startup init. (where we + * can use con=xterm) and false for mconsole.*/ + +int line_setup(struct line *lines, unsigned int num, char *init, int all_allowed) { int i, n; char *end; - if(*init == '=') n = -1; - else { + if(*init == '=') { + /* We said con=/ssl= instead of con#=, so we are configuring all + * consoles at once.*/ + n = -1; + } else { n = simple_strtoul(init, &end, 0); if(*end != '='){ printk(KERN_ERR "line_setup failed to parse \"%s\"\n", init); - return(0); + return 0; } init = end; } init++; - if((n >= 0) && (n >= num)){ + + if (n >= (signed int) num) { printk("line_setup - %d out of range ((0 ... %d) allowed)\n", n, num - 1); - return(0); - } - else if (n >= 0){ + return 0; + } else if (n >= 0){ if (lines[n].count > 0) { printk("line_setup - device %d is open\n", n); - return(0); + return 0; } if (lines[n].init_pri <= INIT_ONE){ lines[n].init_pri = INIT_ONE; @@ -422,13 +534,11 @@ int line_setup(struct line *lines, int num, char *init, int all_allowed) lines[n].valid = 1; } } - } - else if(!all_allowed){ + } else if(!all_allowed){ printk("line_setup - can't configure all devices from " "mconsole\n"); - return(0); - } - else { + return 0; + } else { for(i = 0; i < num; i++){ if(lines[i].init_pri <= INIT_ALL){ lines[i].init_pri = INIT_ALL; @@ -440,21 +550,21 @@ int line_setup(struct line *lines, int num, char *init, int all_allowed) } } } - return(1); + return 1; } -int line_config(struct line *lines, int num, char *str) +int line_config(struct line *lines, unsigned int num, char *str) { char *new = uml_strdup(str); if(new == NULL){ printk("line_config - uml_strdup failed\n"); - return(-ENOMEM); + return -ENOMEM; } - return(!line_setup(lines, num, new, 0)); + return !line_setup(lines, num, new, 0); } -int line_get_config(char *name, struct line *lines, int num, char *str, +int line_get_config(char *name, struct line *lines, unsigned int num, char *str, int size, char **error_out) { struct line *line; @@ -464,47 +574,33 @@ int line_get_config(char *name, struct line *lines, int num, char *str, dev = simple_strtoul(name, &end, 0); if((*end != '\0') || (end == name)){ *error_out = "line_get_config failed to parse device number"; - return(0); + return 0; } if((dev < 0) || (dev >= num)){ - *error_out = "device number of of range"; - return(0); + *error_out = "device number out of range"; + return 0; } line = &lines[dev]; - down(&line->sem); + spin_lock(&line->lock); if(!line->valid) CONFIG_CHUNK(str, size, n, "none", 1); else if(line->count == 0) CONFIG_CHUNK(str, size, n, line->init_str, 1); else n = chan_config_string(&line->chan_list, str, size, error_out); - up(&line->sem); + spin_unlock(&line->lock); - return(n); + return n; } -int line_remove(struct line *lines, int num, char *str) +int line_remove(struct line *lines, unsigned int num, char *str) { char config[sizeof("conxxxx=none\0")]; sprintf(config, "%s=none", str); - return(!line_setup(lines, num, config, 0)); -} - -int line_write_room(struct tty_struct *tty) -{ - struct line *dev = tty->driver_data; - int room; - - if (tty->stopped) - return 0; - room = write_room(dev); - if (0 == room) - printk(KERN_DEBUG "%s: %s: no room left in buffer\n", - __FUNCTION__,tty->name); - return room; + return !line_setup(lines, num, config, 0); } struct tty_driver *line_register_devfs(struct lines *set, @@ -553,7 +649,7 @@ void lines_init(struct line *lines, int nlines) for(i = 0; i < nlines; i++){ line = &lines[i]; INIT_LIST_HEAD(&line->chan_list); - sema_init(&line->sem, 1); + spin_lock_init(&line->lock); if(line->init_str != NULL){ line->init_str = uml_strdup(line->init_str); if(line->init_str == NULL) @@ -587,7 +683,7 @@ irqreturn_t winch_interrupt(int irq, void *data, struct pt_regs *unused) "errno = %d\n", -err); printk("fd %d is losing SIGWINCH support\n", winch->tty_fd); - return(IRQ_HANDLED); + return IRQ_HANDLED; } goto out; } @@ -603,7 +699,7 @@ irqreturn_t winch_interrupt(int irq, void *data, struct pt_regs *unused) out: if(winch->fd != -1) reactivate_fd(winch->fd, WINCH_IRQ); - return(IRQ_HANDLED); + return IRQ_HANDLED; } DECLARE_MUTEX(winch_handler_sem); @@ -625,7 +721,7 @@ void register_winch_irq(int fd, int tty_fd, int pid, struct tty_struct *tty) .pid = pid, .tty = tty }); list_add(&winch->list, &winch_handlers); - if(um_request_irq(WINCH_IRQ, fd, IRQ_READ, winch_interrupt, + if(um_request_irq(WINCH_IRQ, fd, IRQ_READ, winch_interrupt, SA_INTERRUPT | SA_SHIRQ | SA_SAMPLE_RANDOM, "winch", winch) < 0) printk("register_winch_irq - failed to register IRQ\n"); @@ -656,26 +752,16 @@ char *add_xterm_umid(char *base) int len; umid = get_umid(1); - if(umid == NULL) return(base); + if(umid == NULL) + return base; len = strlen(base) + strlen(" ()") + strlen(umid) + 1; title = kmalloc(len, GFP_KERNEL); if(title == NULL){ printk("Failed to allocate buffer for xterm title\n"); - return(base); + return base; } snprintf(title, len, "%s (%s)", base, umid); - return(title); + return title; } - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/drivers/ssl.c b/arch/um/drivers/ssl.c index c5839c3141f..a2bac429f3d 100644 --- a/arch/um/drivers/ssl.c +++ b/arch/um/drivers/ssl.c @@ -107,11 +107,6 @@ int ssl_open(struct tty_struct *tty, struct file *filp) } #if 0 -static int ssl_chars_in_buffer(struct tty_struct *tty) -{ - return(0); -} - static void ssl_flush_buffer(struct tty_struct *tty) { return; @@ -149,11 +144,11 @@ static struct tty_operations ssl_ops = { .put_char = line_put_char, .write_room = line_write_room, .chars_in_buffer = line_chars_in_buffer, + .flush_buffer = line_flush_buffer, + .flush_chars = line_flush_chars, .set_termios = line_set_termios, .ioctl = line_ioctl, #if 0 - .flush_chars = ssl_flush_chars, - .flush_buffer = ssl_flush_buffer, .throttle = ssl_throttle, .unthrottle = ssl_unthrottle, .stop = ssl_stop, @@ -171,10 +166,11 @@ static void ssl_console_write(struct console *c, const char *string, unsigned len) { struct line *line = &serial_lines[c->index]; + unsigned long flags; - down(&line->sem); + spin_lock_irqsave(&line->lock, flags); console_write_chan(&line->chan_list, string, len); - up(&line->sem); + spin_unlock_irqrestore(&line->lock, flags); } static struct tty_driver *ssl_console_device(struct console *c, int *index) @@ -238,14 +234,3 @@ static int ssl_chan_setup(char *str) __setup("ssl", ssl_chan_setup); __channel_help(ssl_chan_setup, "ssl"); - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/drivers/stdio_console.c b/arch/um/drivers/stdio_console.c index e604d7c8769..361d0be342b 100644 --- a/arch/um/drivers/stdio_console.c +++ b/arch/um/drivers/stdio_console.c @@ -116,8 +116,11 @@ static struct tty_operations console_ops = { .open = con_open, .close = line_close, .write = line_write, + .put_char = line_put_char, .write_room = line_write_room, .chars_in_buffer = line_chars_in_buffer, + .flush_buffer = line_flush_buffer, + .flush_chars = line_flush_chars, .set_termios = line_set_termios, .ioctl = line_ioctl, }; @@ -126,10 +129,11 @@ static void uml_console_write(struct console *console, const char *string, unsigned len) { struct line *line = &vts[console->index]; + unsigned long flags; - down(&line->sem); + spin_lock_irqsave(&line->lock, flags); console_write_chan(&line->chan_list, string, len); - up(&line->sem); + spin_unlock_irqrestore(&line->lock, flags); } static struct tty_driver *uml_console_device(struct console *c, int *index) @@ -192,14 +196,3 @@ static int console_chan_setup(char *str) } __setup("con", console_chan_setup); __channel_help(console_chan_setup, "con"); - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ diff --git a/arch/um/include/line.h b/arch/um/include/line.h index 6d81ecc17be..4c5e92c04cc 100644 --- a/arch/um/include/line.h +++ b/arch/um/include/line.h @@ -10,7 +10,7 @@ #include "linux/workqueue.h" #include "linux/tty.h" #include "linux/interrupt.h" -#include "asm/semaphore.h" +#include "linux/spinlock.h" #include "chan_user.h" #include "mconsole_kern.h" @@ -37,10 +37,18 @@ struct line { struct list_head chan_list; int valid; int count; - struct semaphore sem; + /*This lock is actually, mostly, local to*/ + spinlock_t lock; + + /* Yes, this is a real circular buffer. + * XXX: And this should become a struct kfifo! + * + * buffer points to a buffer allocated on demand, of length + * LINE_BUFSIZE, head to the start of the ring, tail to the end.*/ char *buffer; char *head; char *tail; + int sigio; struct work_struct task; struct line_driver *driver; @@ -52,7 +60,6 @@ struct line { init_pri : INIT_STATIC, \ chan_list : { }, \ valid : 1, \ - sem : { }, \ buffer : NULL, \ head : NULL, \ tail : NULL, \ @@ -69,15 +76,18 @@ struct lines { extern void line_close(struct tty_struct *tty, struct file * filp); extern int line_open(struct line *lines, struct tty_struct *tty, struct chan_opts *opts); -extern int line_setup(struct line *lines, int num, char *init, +extern int line_setup(struct line *lines, unsigned int sizeof_lines, char *init, int all_allowed); extern int line_write(struct tty_struct *tty, const unsigned char *buf, int len); extern void line_put_char(struct tty_struct *tty, unsigned char ch); extern void line_set_termios(struct tty_struct *tty, struct termios * old); extern int line_chars_in_buffer(struct tty_struct *tty); +extern void line_flush_buffer(struct tty_struct *tty); +extern void line_flush_chars(struct tty_struct *tty); extern int line_write_room(struct tty_struct *tty); extern int line_ioctl(struct tty_struct *tty, struct file * file, unsigned int cmd, unsigned long arg); + extern char *add_xterm_umid(char *base); extern int line_setup_irq(int fd, int input, int output, struct tty_struct *tty); extern void line_close_chan(struct line *line); @@ -89,20 +99,10 @@ extern struct tty_driver * line_register_devfs(struct lines *set, int nlines); extern void lines_init(struct line *lines, int nlines); extern void close_lines(struct line *lines, int nlines); -extern int line_config(struct line *lines, int num, char *str); -extern int line_remove(struct line *lines, int num, char *str); -extern int line_get_config(char *dev, struct line *lines, int num, char *str, + +extern int line_config(struct line *lines, unsigned int sizeof_lines, char *str); +extern int line_remove(struct line *lines, unsigned int sizeof_lines, char *str); +extern int line_get_config(char *dev, struct line *lines, unsigned int sizeof_lines, char *str, int size, char **error_out); #endif - -/* - * Overrides for Emacs so that we follow Linus's tabbing style. - * Emacs will notice this stuff at the end of the file and automatically - * adjust the settings for this buffer only. This must remain at the end - * of the file. - * --------------------------------------------------------------------------- - * Local variables: - * c-file-style: "linux" - * End: - */ -- cgit v1.2.3-70-g09d2 From 2c49be99ede1e3685a9817a564b4b1678d48e1b9 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Sun, 1 May 2005 08:58:57 -0700 Subject: [PATCH] uml ubd: handle readonly status Use the set_disk_ro() API when the backing file is read-only, to mark the disk read-only, during the ->open(). The current hack does not work when doing a mount -o remount. Also, mark explicitly the code paths which should no more be triggerable (I've removed the WARN_ON(1) things). They should actually become BUG()s probably but I'll avoid that since I'm not so sure the change works so well. I gave it only some limited testing. Signed-off-by: Paolo 'Blaisorblade' Giarrusso CC: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/drivers/ubd_kern.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c index 4d8b165bfa4..9a56ff94308 100644 --- a/arch/um/drivers/ubd_kern.c +++ b/arch/um/drivers/ubd_kern.c @@ -156,6 +156,7 @@ static struct gendisk *fake_gendisk[MAX_DEV]; static struct openflags global_openflags = OPEN_FLAGS; struct cow { + /* This is the backing file, actually */ char *file; int fd; unsigned long *bitmap; @@ -927,10 +928,14 @@ static int ubd_open(struct inode *inode, struct file *filp) } } dev->count++; - if((filp->f_mode & FMODE_WRITE) && !dev->openflags.w){ + set_disk_ro(disk, !dev->openflags.w); + + /* This should no more be needed. And it didn't work anyway to exclude + * read-write remounting of filesystems.*/ + /*if((filp->f_mode & FMODE_WRITE) && !dev->openflags.w){ if(--dev->count == 0) ubd_close(dev); err = -EROFS; - } + }*/ out: return(err); } @@ -1096,6 +1101,7 @@ static int prepare_request(struct request *req, struct io_thread_req *io_req) if(req->rq_status == RQ_INACTIVE) return(1); + /* This should be impossible now */ if((rq_data_dir(req) == WRITE) && !dev->openflags.w){ printk("Write attempted on readonly ubd device %s\n", disk->disk_name); @@ -1243,6 +1249,7 @@ static int ubd_check_remapped(int fd, unsigned long address, int is_write, /* It's a write to a ubd device */ + /* This should be impossible now */ if(!dev->openflags.w){ /* It's a write access on a read-only device - probably * shouldn't happen. If the kernel is trying to change @@ -1605,8 +1612,7 @@ void do_io(struct io_thread_req *req) } } while((n < len) && (n != 0)); if (n < len) memset(&buf[n], 0, len - n); - } - else { + } else { n = os_write_file(req->fds[bit], buf, len); if(n != len){ printk("do_io - write failed err = %d " -- cgit v1.2.3-70-g09d2 From b2c6678c858c5e54d95b996754adfb319cf65735 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Sun, 1 May 2005 08:58:57 -0700 Subject: [PATCH] s390: regenerate defconfig Regenerate the default configuration for s390. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/defconfig | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/s390/defconfig b/arch/s390/defconfig index 1358b420170..07fd0414a4b 100644 --- a/arch/s390/defconfig +++ b/arch/s390/defconfig @@ -1,7 +1,7 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.11 -# Wed Mar 2 16:57:55 2005 +# Linux kernel version: 2.6.12-rc3 +# Fri Apr 22 15:30:58 2005 # CONFIG_MMU=y CONFIG_RWSEM_XCHGADD_ALGORITHM=y @@ -15,6 +15,7 @@ CONFIG_UID16=y CONFIG_EXPERIMENTAL=y CONFIG_CLEAN_COMPILE=y CONFIG_LOCK_KERNEL=y +CONFIG_INIT_ENV_ARG_LIMIT=32 # # General setup @@ -26,24 +27,25 @@ CONFIG_SYSVIPC=y # CONFIG_BSD_PROCESS_ACCT is not set CONFIG_SYSCTL=y # CONFIG_AUDIT is not set -CONFIG_LOG_BUF_SHIFT=17 CONFIG_HOTPLUG=y CONFIG_KOBJECT_UEVENT=y CONFIG_IKCONFIG=y CONFIG_IKCONFIG_PROC=y +# CONFIG_CPUSETS is not set # CONFIG_EMBEDDED is not set CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_ALL is not set # CONFIG_KALLSYMS_EXTRA_PASS is not set +CONFIG_BASE_FULL=y CONFIG_FUTEX=y CONFIG_EPOLL=y -# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set CONFIG_SHMEM=y CONFIG_CC_ALIGN_FUNCTIONS=0 CONFIG_CC_ALIGN_LABELS=0 CONFIG_CC_ALIGN_LOOPS=0 CONFIG_CC_ALIGN_JUMPS=0 # CONFIG_TINY_SHMEM is not set +CONFIG_BASE_SMALL=0 # # Loadable module support @@ -261,7 +263,6 @@ CONFIG_NET=y # CONFIG_PACKET=y # CONFIG_PACKET_MMAP is not set -# CONFIG_NETLINK_DEV is not set CONFIG_UNIX=y CONFIG_NET_KEY=y CONFIG_INET=y @@ -329,6 +330,7 @@ CONFIG_NET_SCH_DSMARK=m CONFIG_NET_QOS=y CONFIG_NET_ESTIMATOR=y CONFIG_NET_CLS=y +# CONFIG_NET_CLS_BASIC is not set CONFIG_NET_CLS_TCINDEX=m CONFIG_NET_CLS_ROUTE4=m CONFIG_NET_CLS_ROUTE=y @@ -338,6 +340,7 @@ CONFIG_NET_CLS_U32=m # CONFIG_NET_CLS_IND is not set CONFIG_NET_CLS_RSVP=m CONFIG_NET_CLS_RSVP6=m +# CONFIG_NET_EMATCH is not set # CONFIG_NET_CLS_ACT is not set CONFIG_NET_CLS_POLICE=y @@ -393,6 +396,8 @@ CONFIG_CTC=m CONFIG_IUCV=m # CONFIG_NETIUCV is not set # CONFIG_SMSGIUCV is not set +# CONFIG_CLAW is not set +# CONFIG_MPC is not set CONFIG_QETH=y # @@ -532,10 +537,13 @@ CONFIG_MSDOS_PARTITION=y # # Kernel hacking # +# CONFIG_PRINTK_TIME is not set CONFIG_DEBUG_KERNEL=y CONFIG_MAGIC_SYSRQ=y +CONFIG_LOG_BUF_SHIFT=17 # CONFIG_SCHEDSTATS is not set # CONFIG_DEBUG_SLAB is not set +# CONFIG_DEBUG_SPINLOCK is not set # CONFIG_DEBUG_SPINLOCK_SLEEP is not set # CONFIG_DEBUG_KOBJECT is not set # CONFIG_DEBUG_INFO is not set @@ -560,6 +568,7 @@ CONFIG_CRYPTO=y # CONFIG_CRYPTO_SHA256 is not set # CONFIG_CRYPTO_SHA512 is not set # CONFIG_CRYPTO_WP512 is not set +# CONFIG_CRYPTO_TGR192 is not set # CONFIG_CRYPTO_DES is not set # CONFIG_CRYPTO_DES_Z990 is not set # CONFIG_CRYPTO_BLOWFISH is not set -- cgit v1.2.3-70-g09d2 From 4b7e0706620e3947dc1685dfdbc1413404afb545 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Sun, 1 May 2005 08:58:57 -0700 Subject: [PATCH] s390: idle timer setup Fix overflow in calculation of the new tod value in stop_hz_timer and fix wrong virtual timer list idle time in case the virtual timer is already expired in stop_cpu_timer. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/kernel/time.c | 12 +++++++++--- arch/s390/kernel/vtime.c | 25 +++++++++++++------------ 2 files changed, 22 insertions(+), 15 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index 061e81138dc..8ca48567678 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -244,7 +244,7 @@ int sysctl_hz_timer = 1; */ static inline void stop_hz_timer(void) { - __u64 timer; + __u64 timer, todval; if (sysctl_hz_timer != 0) return; @@ -265,8 +265,14 @@ static inline void stop_hz_timer(void) * for the next event. */ timer = (__u64) (next_timer_interrupt() - jiffies) + jiffies_64; - timer = jiffies_timer_cc + timer * CLK_TICKS_PER_JIFFY; - asm volatile ("SCKC %0" : : "m" (timer)); + todval = -1ULL; + /* Be careful about overflows. */ + if (timer < (-1ULL / CLK_TICKS_PER_JIFFY)) { + timer = jiffies_timer_cc + timer * CLK_TICKS_PER_JIFFY; + if (timer >= jiffies_timer_cc) + todval = timer; + } + asm volatile ("SCKC %0" : : "m" (todval)); } /* diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c index bb6cf02418a..fa0726507b3 100644 --- a/arch/s390/kernel/vtime.c +++ b/arch/s390/kernel/vtime.c @@ -122,12 +122,17 @@ static void start_cpu_timer(void) struct vtimer_queue *vt_list; vt_list = &per_cpu(virt_cpu_timer, smp_processor_id()); - set_vtimer(vt_list->idle); + + /* CPU timer interrupt is pending, don't reprogramm it */ + if (vt_list->idle & 1LL<<63) + return; + + if (!list_empty(&vt_list->list)) + set_vtimer(vt_list->idle); } static void stop_cpu_timer(void) { - __u64 done; struct vtimer_queue *vt_list; vt_list = &per_cpu(virt_cpu_timer, smp_processor_id()); @@ -138,21 +143,17 @@ static void stop_cpu_timer(void) goto fire; } - /* store progress */ - asm volatile ("STPT %0" : "=m" (done)); + /* store the actual expire value */ + asm volatile ("STPT %0" : "=m" (vt_list->idle)); /* - * If done is negative we do not stop the CPU timer - * because we will get instantly an interrupt that - * will start the CPU timer again. + * If the CPU timer is negative we don't reprogramm + * it because we will get instantly an interrupt. */ - if (done & 1LL<<63) + if (vt_list->idle & 1LL<<63) return; - else - vt_list->offset += vt_list->to_expire - done; - /* save the actual expire value */ - vt_list->idle = done; + vt_list->offset += vt_list->to_expire - vt_list->idle; /* * We cannot halt the CPU timer, we just write a value that -- cgit v1.2.3-70-g09d2 From c9e3735359ac2d74ee61c6f1e5724f4a6db570bf Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Sun, 1 May 2005 08:58:57 -0700 Subject: [PATCH] s390: fix memory holes and cleanup setup_arch The memory setup didn't take care of memory holes and this makes the memory management think there would be more memory available than there is in reality. That causes the OOM killer to kill processes even if there is enough memory left that can be written to the swap space. The patch fixes this by using free_area_init_node with an array of memory holes instead of free_area_init. Further the patch cleans up the code in setup.c by splitting setup_arch into smaller pieces. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/kernel/setup.c | 385 ++++++++++++++++++++++++++--------------------- arch/s390/mm/init.c | 15 +- 2 files changed, 224 insertions(+), 176 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index c879c40aa7a..f0679be4f96 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -60,6 +60,8 @@ struct { #define CHUNK_READ_WRITE 0 #define CHUNK_READ_ONLY 1 volatile int __cpu_logical_map[NR_CPUS]; /* logical cpu to cpu address */ +unsigned long __initdata zholes_size[MAX_NR_ZONES]; +static unsigned long __initdata memory_end; /* * Setup options @@ -78,11 +80,15 @@ static char command_line[COMMAND_LINE_SIZE] = { 0, }; static struct resource code_resource = { .name = "Kernel code", + .start = (unsigned long) &_text, + .end = (unsigned long) &_etext - 1, .flags = IORESOURCE_BUSY | IORESOURCE_MEM, }; static struct resource data_resource = { .name = "Kernel data", + .start = (unsigned long) &_etext, + .end = (unsigned long) &_edata - 1, .flags = IORESOURCE_BUSY | IORESOURCE_MEM, }; @@ -310,90 +316,50 @@ void machine_power_off(void) EXPORT_SYMBOL(machine_power_off); -/* - * Setup function called from init/main.c just after the banner - * was printed. - */ -extern char _pstart, _pend, _stext; - -void __init setup_arch(char **cmdline_p) +static void __init +add_memory_hole(unsigned long start, unsigned long end) { - unsigned long bootmap_size; - unsigned long memory_start, memory_end; - char c = ' ', cn, *to = command_line, *from = COMMAND_LINE; - unsigned long start_pfn, end_pfn; - static unsigned int smptrap=0; - unsigned long delay = 0; - struct _lowcore *lc; - int i; + unsigned long dma_pfn = MAX_DMA_ADDRESS >> PAGE_SHIFT; + + if (end <= dma_pfn) + zholes_size[ZONE_DMA] += end - start + 1; + else if (start > dma_pfn) + zholes_size[ZONE_NORMAL] += end - start + 1; + else { + zholes_size[ZONE_DMA] += dma_pfn - start + 1; + zholes_size[ZONE_NORMAL] += end - dma_pfn; + } +} - if (smptrap) - return; - smptrap=1; +static void __init +parse_cmdline_early(char **cmdline_p) +{ + char c = ' ', cn, *to = command_line, *from = COMMAND_LINE; + unsigned long delay = 0; - /* - * print what head.S has found out about the machine - */ -#ifndef CONFIG_ARCH_S390X - printk((MACHINE_IS_VM) ? - "We are running under VM (31 bit mode)\n" : - "We are running native (31 bit mode)\n"); - printk((MACHINE_HAS_IEEE) ? - "This machine has an IEEE fpu\n" : - "This machine has no IEEE fpu\n"); -#else /* CONFIG_ARCH_S390X */ - printk((MACHINE_IS_VM) ? - "We are running under VM (64 bit mode)\n" : - "We are running native (64 bit mode)\n"); -#endif /* CONFIG_ARCH_S390X */ + /* Save unparsed command line copy for /proc/cmdline */ + memcpy(saved_command_line, COMMAND_LINE, COMMAND_LINE_SIZE); + saved_command_line[COMMAND_LINE_SIZE-1] = '\0'; - ROOT_DEV = Root_RAM0; - memory_start = (unsigned long) &_end; /* fixit if use $CODELO etc*/ -#ifndef CONFIG_ARCH_S390X - memory_end = memory_size & ~0x400000UL; /* align memory end to 4MB */ - /* - * We need some free virtual space to be able to do vmalloc. - * On a machine with 2GB memory we make sure that we have at - * least 128 MB free space for vmalloc. - */ - if (memory_end > 1920*1024*1024) - memory_end = 1920*1024*1024; -#else /* CONFIG_ARCH_S390X */ - memory_end = memory_size & ~0x200000UL; /* detected in head.s */ -#endif /* CONFIG_ARCH_S390X */ - init_mm.start_code = PAGE_OFFSET; - init_mm.end_code = (unsigned long) &_etext; - init_mm.end_data = (unsigned long) &_edata; - init_mm.brk = (unsigned long) &_end; - - code_resource.start = (unsigned long) &_text; - code_resource.end = (unsigned long) &_etext - 1; - data_resource.start = (unsigned long) &_etext; - data_resource.end = (unsigned long) &_edata - 1; - - /* Save unparsed command line copy for /proc/cmdline */ - memcpy(saved_command_line, COMMAND_LINE, COMMAND_LINE_SIZE); - saved_command_line[COMMAND_LINE_SIZE-1] = '\0'; - - for (;;) { - /* - * "mem=XXX[kKmM]" sets memsize - */ - if (c == ' ' && strncmp(from, "mem=", 4) == 0) { - memory_end = simple_strtoul(from+4, &from, 0); - if ( *from == 'K' || *from == 'k' ) { - memory_end = memory_end << 10; - from++; - } else if ( *from == 'M' || *from == 'm' ) { - memory_end = memory_end << 20; - from++; - } - } - /* - * "ipldelay=XXX[sm]" sets ipl delay in seconds or minutes - */ - if (c == ' ' && strncmp(from, "ipldelay=", 9) == 0) { - delay = simple_strtoul(from+9, &from, 0); + for (;;) { + /* + * "mem=XXX[kKmM]" sets memsize + */ + if (c == ' ' && strncmp(from, "mem=", 4) == 0) { + memory_end = simple_strtoul(from+4, &from, 0); + if ( *from == 'K' || *from == 'k' ) { + memory_end = memory_end << 10; + from++; + } else if ( *from == 'M' || *from == 'm' ) { + memory_end = memory_end << 20; + from++; + } + } + /* + * "ipldelay=XXX[sm]" sets ipl delay in seconds or minutes + */ + if (c == ' ' && strncmp(from, "ipldelay=", 9) == 0) { + delay = simple_strtoul(from+9, &from, 0); if (*from == 's' || *from == 'S') { delay = delay*1000000; from++; @@ -403,24 +369,110 @@ void __init setup_arch(char **cmdline_p) } /* now wait for the requested amount of time */ udelay(delay); - } - cn = *(from++); - if (!cn) - break; - if (cn == '\n') - cn = ' '; /* replace newlines with space */ + } + cn = *(from++); + if (!cn) + break; + if (cn == '\n') + cn = ' '; /* replace newlines with space */ if (cn == 0x0d) cn = ' '; /* replace 0x0d with space */ - if (cn == ' ' && c == ' ') - continue; /* remove additional spaces */ - c = cn; - if (to - command_line >= COMMAND_LINE_SIZE) - break; - *(to++) = c; - } - if (c == ' ' && to > command_line) to--; - *to = '\0'; - *cmdline_p = command_line; + if (cn == ' ' && c == ' ') + continue; /* remove additional spaces */ + c = cn; + if (to - command_line >= COMMAND_LINE_SIZE) + break; + *(to++) = c; + } + if (c == ' ' && to > command_line) to--; + *to = '\0'; + *cmdline_p = command_line; +} + +static void __init +setup_lowcore(void) +{ + struct _lowcore *lc; + int lc_pages; + + /* + * Setup lowcore for boot cpu + */ + lc_pages = sizeof(void *) == 8 ? 2 : 1; + lc = (struct _lowcore *) + __alloc_bootmem(lc_pages * PAGE_SIZE, lc_pages * PAGE_SIZE, 0); + memset(lc, 0, lc_pages * PAGE_SIZE); + lc->restart_psw.mask = PSW_BASE_BITS; + lc->restart_psw.addr = + PSW_ADDR_AMODE | (unsigned long) restart_int_handler; + lc->external_new_psw.mask = PSW_KERNEL_BITS; + lc->external_new_psw.addr = + PSW_ADDR_AMODE | (unsigned long) ext_int_handler; + lc->svc_new_psw.mask = PSW_KERNEL_BITS | PSW_MASK_IO | PSW_MASK_EXT; + lc->svc_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) system_call; + lc->program_new_psw.mask = PSW_KERNEL_BITS; + lc->program_new_psw.addr = + PSW_ADDR_AMODE | (unsigned long)pgm_check_handler; + lc->mcck_new_psw.mask = PSW_KERNEL_BITS; + lc->mcck_new_psw.addr = + PSW_ADDR_AMODE | (unsigned long) mcck_int_handler; + lc->io_new_psw.mask = PSW_KERNEL_BITS; + lc->io_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) io_int_handler; + lc->ipl_device = S390_lowcore.ipl_device; + lc->jiffy_timer = -1LL; + lc->kernel_stack = ((unsigned long) &init_thread_union) + THREAD_SIZE; + lc->async_stack = (unsigned long) + __alloc_bootmem(ASYNC_SIZE, ASYNC_SIZE, 0) + ASYNC_SIZE; +#ifdef CONFIG_CHECK_STACK + lc->panic_stack = (unsigned long) + __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0) + PAGE_SIZE; +#endif + lc->current_task = (unsigned long) init_thread_union.thread_info.task; + lc->thread_info = (unsigned long) &init_thread_union; +#ifdef CONFIG_ARCH_S390X + if (MACHINE_HAS_DIAG44) + lc->diag44_opcode = 0x83000044; + else + lc->diag44_opcode = 0x07000700; +#endif /* CONFIG_ARCH_S390X */ + set_prefix((u32)(unsigned long) lc); +} + +static void __init +setup_resources(void) +{ + struct resource *res; + int i; + + for (i = 0; i < MEMORY_CHUNKS && memory_chunk[i].size > 0; i++) { + res = alloc_bootmem_low(sizeof(struct resource)); + res->flags = IORESOURCE_BUSY | IORESOURCE_MEM; + switch (memory_chunk[i].type) { + case CHUNK_READ_WRITE: + res->name = "System RAM"; + break; + case CHUNK_READ_ONLY: + res->name = "System ROM"; + res->flags |= IORESOURCE_READONLY; + break; + default: + res->name = "reserved"; + } + res->start = memory_chunk[i].addr; + res->end = memory_chunk[i].addr + memory_chunk[i].size - 1; + request_resource(&iomem_resource, res); + request_resource(res, &code_resource); + request_resource(res, &data_resource); + } +} + +static void __init +setup_memory(void) +{ + unsigned long bootmap_size; + unsigned long start_pfn, end_pfn; + unsigned long last_rw_end; + int i; /* * partially used pages are not usable - thus @@ -437,6 +489,8 @@ void __init setup_arch(char **cmdline_p) /* * Register RAM areas with the bootmem allocator. */ + last_rw_end = start_pfn; + for (i = 0; i < 16 && memory_chunk[i].size > 0; i++) { unsigned long start_chunk, end_chunk; @@ -450,102 +504,91 @@ void __init setup_arch(char **cmdline_p) start_chunk = start_pfn; if (end_chunk > end_pfn) end_chunk = end_pfn; - if (start_chunk < end_chunk) + if (start_chunk < end_chunk) { free_bootmem(start_chunk << PAGE_SHIFT, (end_chunk - start_chunk) << PAGE_SHIFT); + if (last_rw_end < start_chunk) + add_memory_hole(last_rw_end, start_chunk - 1); + last_rw_end = end_chunk; + } } - /* - * Reserve the bootmem bitmap itself as well. We do this in two - * steps (first step was init_bootmem()) because this catches - * the (very unlikely) case of us accidentally initializing the - * bootmem allocator with an invalid RAM area. - */ - reserve_bootmem(start_pfn << PAGE_SHIFT, bootmap_size); + if (last_rw_end < end_pfn - 1) + add_memory_hole(last_rw_end, end_pfn - 1); + + /* + * Reserve the bootmem bitmap itself as well. We do this in two + * steps (first step was init_bootmem()) because this catches + * the (very unlikely) case of us accidentally initializing the + * bootmem allocator with an invalid RAM area. + */ + reserve_bootmem(start_pfn << PAGE_SHIFT, bootmap_size); #ifdef CONFIG_BLK_DEV_INITRD - if (INITRD_START) { + if (INITRD_START) { if (INITRD_START + INITRD_SIZE <= memory_end) { reserve_bootmem(INITRD_START, INITRD_SIZE); initrd_start = INITRD_START; initrd_end = initrd_start + INITRD_SIZE; } else { - printk("initrd extends beyond end of memory " - "(0x%08lx > 0x%08lx)\ndisabling initrd\n", - initrd_start + INITRD_SIZE, memory_end); - initrd_start = initrd_end = 0; + printk("initrd extends beyond end of memory " + "(0x%08lx > 0x%08lx)\ndisabling initrd\n", + initrd_start + INITRD_SIZE, memory_end); + initrd_start = initrd_end = 0; } - } + } #endif +} - for (i = 0; i < 16 && memory_chunk[i].size > 0; i++) { - struct resource *res; - - res = alloc_bootmem_low(sizeof(struct resource)); - res->flags = IORESOURCE_BUSY | IORESOURCE_MEM; - - switch (memory_chunk[i].type) { - case CHUNK_READ_WRITE: - res->name = "System RAM"; - break; - case CHUNK_READ_ONLY: - res->name = "System ROM"; - res->flags |= IORESOURCE_READONLY; - break; - default: - res->name = "reserved"; - } - res->start = memory_chunk[i].addr; - res->end = memory_chunk[i].addr + memory_chunk[i].size - 1; - request_resource(&iomem_resource, res); - request_resource(res, &code_resource); - request_resource(res, &data_resource); - } +/* + * Setup function called from init/main.c just after the banner + * was printed. + */ +void __init +setup_arch(char **cmdline_p) +{ /* - * Setup lowcore for boot cpu + * print what head.S has found out about the machine */ #ifndef CONFIG_ARCH_S390X - lc = (struct _lowcore *) __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0); - memset(lc, 0, PAGE_SIZE); + printk((MACHINE_IS_VM) ? + "We are running under VM (31 bit mode)\n" : + "We are running native (31 bit mode)\n"); + printk((MACHINE_HAS_IEEE) ? + "This machine has an IEEE fpu\n" : + "This machine has no IEEE fpu\n"); #else /* CONFIG_ARCH_S390X */ - lc = (struct _lowcore *) __alloc_bootmem(2*PAGE_SIZE, 2*PAGE_SIZE, 0); - memset(lc, 0, 2*PAGE_SIZE); + printk((MACHINE_IS_VM) ? + "We are running under VM (64 bit mode)\n" : + "We are running native (64 bit mode)\n"); #endif /* CONFIG_ARCH_S390X */ - lc->restart_psw.mask = PSW_BASE_BITS; - lc->restart_psw.addr = - PSW_ADDR_AMODE | (unsigned long) restart_int_handler; - lc->external_new_psw.mask = PSW_KERNEL_BITS; - lc->external_new_psw.addr = - PSW_ADDR_AMODE | (unsigned long) ext_int_handler; - lc->svc_new_psw.mask = PSW_KERNEL_BITS | PSW_MASK_IO | PSW_MASK_EXT; - lc->svc_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) system_call; - lc->program_new_psw.mask = PSW_KERNEL_BITS; - lc->program_new_psw.addr = - PSW_ADDR_AMODE | (unsigned long)pgm_check_handler; - lc->mcck_new_psw.mask = PSW_KERNEL_BITS; - lc->mcck_new_psw.addr = - PSW_ADDR_AMODE | (unsigned long) mcck_int_handler; - lc->io_new_psw.mask = PSW_KERNEL_BITS; - lc->io_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) io_int_handler; - lc->ipl_device = S390_lowcore.ipl_device; - lc->jiffy_timer = -1LL; - lc->kernel_stack = ((unsigned long) &init_thread_union) + THREAD_SIZE; - lc->async_stack = (unsigned long) - __alloc_bootmem(ASYNC_SIZE, ASYNC_SIZE, 0) + ASYNC_SIZE; -#ifdef CONFIG_CHECK_STACK - lc->panic_stack = (unsigned long) - __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0) + PAGE_SIZE; -#endif - lc->current_task = (unsigned long) init_thread_union.thread_info.task; - lc->thread_info = (unsigned long) &init_thread_union; -#ifdef CONFIG_ARCH_S390X - if (MACHINE_HAS_DIAG44) - lc->diag44_opcode = 0x83000044; - else - lc->diag44_opcode = 0x07000700; + + ROOT_DEV = Root_RAM0; +#ifndef CONFIG_ARCH_S390X + memory_end = memory_size & ~0x400000UL; /* align memory end to 4MB */ + /* + * We need some free virtual space to be able to do vmalloc. + * On a machine with 2GB memory we make sure that we have at + * least 128 MB free space for vmalloc. + */ + if (memory_end > 1920*1024*1024) + memory_end = 1920*1024*1024; +#else /* CONFIG_ARCH_S390X */ + memory_end = memory_size & ~0x200000UL; /* detected in head.s */ #endif /* CONFIG_ARCH_S390X */ - set_prefix((u32)(unsigned long) lc); + + init_mm.start_code = PAGE_OFFSET; + init_mm.end_code = (unsigned long) &_etext; + init_mm.end_data = (unsigned long) &_edata; + init_mm.brk = (unsigned long) &_end; + + parse_cmdline_early(cmdline_p); + + setup_memory(); + setup_resources(); + setup_lowcore(); + cpu_init(); __cpu_logical_map[0] = S390_lowcore.cpu_data.cpu_addr; diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index 8e723bc7f79..6ec5cd981e7 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -101,6 +101,7 @@ extern unsigned long _end; extern unsigned long __init_begin; extern unsigned long __init_end; +extern unsigned long __initdata zholes_size[]; /* * paging_init() sets up the page tables */ @@ -163,10 +164,13 @@ void __init paging_init(void) local_flush_tlb(); { - unsigned long zones_size[MAX_NR_ZONES] = { 0, 0, 0}; + unsigned long zones_size[MAX_NR_ZONES]; + memset(zones_size, 0, sizeof(zones_size)); zones_size[ZONE_DMA] = max_low_pfn; - free_area_init(zones_size); + free_area_init_node(0, &contig_page_data, zones_size, + __pa(PAGE_OFFSET) >> PAGE_SHIFT, + zholes_size); } return; } @@ -184,9 +188,10 @@ void __init paging_init(void) _KERN_REGION_TABLE; static const int ssm_mask = 0x04000000L; - unsigned long zones_size[MAX_NR_ZONES] = {0, 0, 0}; + unsigned long zones_size[MAX_NR_ZONES]; unsigned long dma_pfn, high_pfn; + memset(zones_size, 0, sizeof(zones_size)); dma_pfn = MAX_DMA_ADDRESS >> PAGE_SHIFT; high_pfn = max_low_pfn; @@ -198,8 +203,8 @@ void __init paging_init(void) } /* Initialize mem_map[]. */ - free_area_init(zones_size); - + free_area_init_node(0, &contig_page_data, zones_size, + __pa(PAGE_OFFSET) >> PAGE_SHIFT, zholes_size); /* * map whole physical memory to virtual memory (identity mapping) -- cgit v1.2.3-70-g09d2 From 0b642ede47969d4180b0922d982777fe64379228 Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Sun, 1 May 2005 08:58:58 -0700 Subject: [PATCH] s390: default storage key Provide an easy way to define a non-zero storage key at compile time. This is useful for debugging purposes. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/kernel/s390_ksyms.c | 1 - arch/s390/kernel/setup.c | 20 ++++++++++++++++---- drivers/s390/cio/cio.c | 4 ++-- drivers/s390/cio/device_ops.c | 6 +++--- drivers/s390/cio/qdio.h | 6 ++++-- include/asm-s390/page.h | 2 ++ include/asm-s390/processor.h | 2 +- include/asm-s390/ptrace.h | 16 +++++++++++++--- 8 files changed, 41 insertions(+), 16 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/s390_ksyms.c b/arch/s390/kernel/s390_ksyms.c index 11fd6d556d8..bee654abb6d 100644 --- a/arch/s390/kernel/s390_ksyms.c +++ b/arch/s390/kernel/s390_ksyms.c @@ -34,7 +34,6 @@ EXPORT_SYMBOL(__clear_user_asm); EXPORT_SYMBOL(__strncpy_from_user_asm); EXPORT_SYMBOL(__strnlen_user_asm); EXPORT_SYMBOL(diag10); -EXPORT_SYMBOL(default_storage_key); /* * semaphore ops diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index f0679be4f96..df83215beac 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -44,6 +44,8 @@ #include #include #include +#include +#include /* * Machine setup.. @@ -53,7 +55,6 @@ unsigned int console_devno = -1; unsigned int console_irq = -1; unsigned long memory_size = 0; unsigned long machine_flags = 0; -unsigned int default_storage_key = 0; struct { unsigned long addr, size, type; } memory_chunk[MEMORY_CHUNKS] = { { 0 } }; @@ -402,7 +403,7 @@ setup_lowcore(void) lc = (struct _lowcore *) __alloc_bootmem(lc_pages * PAGE_SIZE, lc_pages * PAGE_SIZE, 0); memset(lc, 0, lc_pages * PAGE_SIZE); - lc->restart_psw.mask = PSW_BASE_BITS; + lc->restart_psw.mask = PSW_BASE_BITS | PSW_DEFAULT_KEY; lc->restart_psw.addr = PSW_ADDR_AMODE | (unsigned long) restart_int_handler; lc->external_new_psw.mask = PSW_KERNEL_BITS; @@ -470,7 +471,7 @@ static void __init setup_memory(void) { unsigned long bootmap_size; - unsigned long start_pfn, end_pfn; + unsigned long start_pfn, end_pfn, init_pfn; unsigned long last_rw_end; int i; @@ -481,6 +482,10 @@ setup_memory(void) start_pfn = (__pa(&_end) + PAGE_SIZE - 1) >> PAGE_SHIFT; end_pfn = max_pfn = memory_end >> PAGE_SHIFT; + /* Initialize storage key for kernel pages */ + for (init_pfn = 0 ; init_pfn < start_pfn; init_pfn++) + page_set_storage_key(init_pfn << PAGE_SHIFT, PAGE_DEFAULT_KEY); + /* * Initialize the boot-time allocator (with low memory only): */ @@ -491,7 +496,7 @@ setup_memory(void) */ last_rw_end = start_pfn; - for (i = 0; i < 16 && memory_chunk[i].size > 0; i++) { + for (i = 0; i < MEMORY_CHUNKS && memory_chunk[i].size > 0; i++) { unsigned long start_chunk, end_chunk; if (memory_chunk[i].type != CHUNK_READ_WRITE) @@ -505,6 +510,11 @@ setup_memory(void) if (end_chunk > end_pfn) end_chunk = end_pfn; if (start_chunk < end_chunk) { + /* Initialize storage key for RAM pages */ + for (init_pfn = start_chunk ; init_pfn < end_chunk; + init_pfn++) + page_set_storage_key(init_pfn << PAGE_SHIFT, + PAGE_DEFAULT_KEY); free_bootmem(start_chunk << PAGE_SHIFT, (end_chunk - start_chunk) << PAGE_SHIFT); if (last_rw_end < start_chunk) @@ -513,6 +523,8 @@ setup_memory(void) } } + psw_set_key(PAGE_DEFAULT_KEY); + if (last_rw_end < end_pfn - 1) add_memory_hole(last_rw_end, end_pfn - 1); diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c index 99ce5a56798..1d9b3f18d8d 100644 --- a/drivers/s390/cio/cio.c +++ b/drivers/s390/cio/cio.c @@ -1,7 +1,7 @@ /* * drivers/s390/cio/cio.c * S/390 common I/O routines -- low level i/o calls - * $Revision: 1.131 $ + * $Revision: 1.133 $ * * Copyright (C) 1999-2002 IBM Deutschland Entwicklung GmbH, * IBM Corporation @@ -228,7 +228,7 @@ cio_start_key (struct subchannel *sch, /* subchannel structure */ int cio_start (struct subchannel *sch, struct ccw1 *cpa, __u8 lpm) { - return cio_start_key(sch, cpa, lpm, default_storage_key); + return cio_start_key(sch, cpa, lpm, PAGE_DEFAULT_KEY); } /* diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c index 11e260e0b9c..02d01a0de16 100644 --- a/drivers/s390/cio/device_ops.c +++ b/drivers/s390/cio/device_ops.c @@ -1,7 +1,7 @@ /* * drivers/s390/cio/device_ops.c * - * $Revision: 1.55 $ + * $Revision: 1.56 $ * * Copyright (C) 2002 IBM Deutschland Entwicklung GmbH, * IBM Corporation @@ -128,7 +128,7 @@ ccw_device_start(struct ccw_device *cdev, struct ccw1 *cpa, unsigned long intparm, __u8 lpm, unsigned long flags) { return ccw_device_start_key(cdev, cpa, intparm, lpm, - default_storage_key, flags); + PAGE_DEFAULT_KEY, flags); } int @@ -137,7 +137,7 @@ ccw_device_start_timeout(struct ccw_device *cdev, struct ccw1 *cpa, int expires) { return ccw_device_start_timeout_key(cdev, cpa, intparm, lpm, - default_storage_key, flags, + PAGE_DEFAULT_KEY, flags, expires); } diff --git a/drivers/s390/cio/qdio.h b/drivers/s390/cio/qdio.h index 9ad14db2414..b6daadac4e8 100644 --- a/drivers/s390/cio/qdio.h +++ b/drivers/s390/cio/qdio.h @@ -1,7 +1,9 @@ #ifndef _CIO_QDIO_H #define _CIO_QDIO_H -#define VERSION_CIO_QDIO_H "$Revision: 1.26 $" +#include + +#define VERSION_CIO_QDIO_H "$Revision: 1.32 $" #ifdef CONFIG_QDIO_DEBUG #define QDIO_VERBOSE_LEVEL 9 @@ -42,7 +44,7 @@ #define QDIO_Q_LAPS 5 -#define QDIO_STORAGE_KEY 0 +#define QDIO_STORAGE_KEY PAGE_DEFAULT_KEY #define L2_CACHELINE_SIZE 256 #define INDICATORS_PER_CACHELINE (L2_CACHELINE_SIZE/sizeof(__u32)) diff --git a/include/asm-s390/page.h b/include/asm-s390/page.h index 614e2a93c70..2be287b9df8 100644 --- a/include/asm-s390/page.h +++ b/include/asm-s390/page.h @@ -16,6 +16,8 @@ #define PAGE_SHIFT 12 #define PAGE_SIZE (1UL << PAGE_SHIFT) #define PAGE_MASK (~(PAGE_SIZE-1)) +#define PAGE_DEFAULT_ACC 0 +#define PAGE_DEFAULT_KEY (PAGE_DEFAULT_ACC << 4) #ifdef __KERNEL__ #ifndef __ASSEMBLY__ diff --git a/include/asm-s390/processor.h b/include/asm-s390/processor.h index 88c272ca48b..fb46e9090b5 100644 --- a/include/asm-s390/processor.h +++ b/include/asm-s390/processor.h @@ -245,7 +245,7 @@ static inline void enabled_wait(void) psw_t wait_psw; wait_psw.mask = PSW_BASE_BITS | PSW_MASK_IO | PSW_MASK_EXT | - PSW_MASK_MCHECK | PSW_MASK_WAIT; + PSW_MASK_MCHECK | PSW_MASK_WAIT | PSW_DEFAULT_KEY; #ifndef __s390x__ asm volatile ( " basr %0,0\n" diff --git a/include/asm-s390/ptrace.h b/include/asm-s390/ptrace.h index 1dc80666e97..4eff8f2e3bf 100644 --- a/include/asm-s390/ptrace.h +++ b/include/asm-s390/ptrace.h @@ -185,6 +185,7 @@ #include #include #include +#include typedef union { @@ -235,6 +236,7 @@ typedef struct #define PSW_ADDR_INSN 0x7FFFFFFFUL #define PSW_BASE_BITS 0x00080000UL +#define PSW_DEFAULT_KEY (((unsigned long) PAGE_DEFAULT_ACC) << 20) #define PSW_ASC_PRIMARY 0x00000000UL #define PSW_ASC_ACCREG 0x00004000UL @@ -260,6 +262,7 @@ typedef struct #define PSW_BASE_BITS 0x0000000180000000UL #define PSW_BASE32_BITS 0x0000000080000000UL +#define PSW_DEFAULT_KEY (((unsigned long) PAGE_DEFAULT_ACC) << 52) #define PSW_ASC_PRIMARY 0x0000000000000000UL #define PSW_ASC_ACCREG 0x0000400000000000UL @@ -268,14 +271,15 @@ typedef struct #define PSW_USER32_BITS (PSW_BASE32_BITS | PSW_MASK_DAT | PSW_ASC_HOME | \ PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK | \ - PSW_MASK_PSTATE) + PSW_MASK_PSTATE | PSW_DEFAULT_KEY) #endif /* __s390x__ */ -#define PSW_KERNEL_BITS (PSW_BASE_BITS | PSW_MASK_DAT | PSW_ASC_PRIMARY) +#define PSW_KERNEL_BITS (PSW_BASE_BITS | PSW_MASK_DAT | PSW_ASC_PRIMARY | \ + PSW_DEFAULT_KEY) #define PSW_USER_BITS (PSW_BASE_BITS | PSW_MASK_DAT | PSW_ASC_HOME | \ PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK | \ - PSW_MASK_PSTATE) + PSW_MASK_PSTATE | PSW_DEFAULT_KEY) /* This macro merges a NEW PSW mask specified by the user into the currently active PSW mask CURRENT, modifying only those @@ -470,6 +474,12 @@ struct user_regs_struct extern void show_regs(struct pt_regs * regs); #endif +static inline void +psw_set_key(unsigned int key) +{ + asm volatile ( "spka 0(%0)" : : "d" (key) ); +} + #endif /* __ASSEMBLY__ */ #endif /* _S390_PTRACE_H */ -- cgit v1.2.3-70-g09d2 From 15439d74f621ff8252b53136d259781148e6e768 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Sun, 1 May 2005 08:58:58 -0700 Subject: [PATCH] s390: cmm guest sender id An arbitrary guest must not be allowed to trigger cmm actions. Only one specific guest namely the one that serves as the resource monitor may send cmm messages. Add a parameter that allows to specify the guest that may send messages. z/VMs resource manager has the name 'VMRMSVM' which is the default. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/mm/cmm.c | 9 ++++++++- drivers/s390/net/smsgiucv.c | 19 ++++++++++++++----- drivers/s390/net/smsgiucv.h | 4 ++-- 3 files changed, 24 insertions(+), 8 deletions(-) (limited to 'arch') diff --git a/arch/s390/mm/cmm.c b/arch/s390/mm/cmm.c index d30cdb4248a..f5a5bc09b8f 100644 --- a/arch/s390/mm/cmm.c +++ b/arch/s390/mm/cmm.c @@ -20,6 +20,11 @@ #include #include +static char *sender = "VMRMSVM"; +module_param(sender, charp, 0); +MODULE_PARM_DESC(sender, + "Guest name that may send SMSG messages (default VMRMSVM)"); + #include "../../../drivers/s390/net/smsgiucv.h" #define CMM_NR_PAGES ((PAGE_SIZE / sizeof(unsigned long)) - 2) @@ -367,10 +372,12 @@ static struct ctl_table cmm_dir_table[] = { #ifdef CONFIG_CMM_IUCV #define SMSG_PREFIX "CMM" static void -cmm_smsg_target(char *msg) +cmm_smsg_target(char *from, char *msg) { long pages, seconds; + if (strlen(sender) > 0 && strcmp(from, sender) != 0) + return; if (!cmm_skip_blanks(msg + strlen(SMSG_PREFIX), &msg)) return; if (strncmp(msg, "SHRINK", 6) == 0) { diff --git a/drivers/s390/net/smsgiucv.c b/drivers/s390/net/smsgiucv.c index a3d28585956..1e3f7f3c662 100644 --- a/drivers/s390/net/smsgiucv.c +++ b/drivers/s390/net/smsgiucv.c @@ -32,7 +32,7 @@ struct smsg_callback { struct list_head list; char *prefix; int len; - void (*callback)(char *str); + void (*callback)(char *from, char *str); }; MODULE_AUTHOR @@ -55,8 +55,9 @@ smsg_message_pending(iucv_MessagePending *eib, void *pgm_data) { struct smsg_callback *cb; unsigned char *msg; + unsigned char sender[9]; unsigned short len; - int rc; + int rc, i; len = eib->ln1msg2.ipbfln1f; msg = kmalloc(len + 1, GFP_ATOMIC|GFP_DMA); @@ -69,10 +70,18 @@ smsg_message_pending(iucv_MessagePending *eib, void *pgm_data) if (rc == 0) { msg[len] = 0; EBCASC(msg, len); + memcpy(sender, msg, 8); + sender[8] = 0; + /* Remove trailing whitespace from the sender name. */ + for (i = 7; i >= 0; i--) { + if (sender[i] != ' ' && sender[i] != '\t') + break; + sender[i] = 0; + } spin_lock(&smsg_list_lock); list_for_each_entry(cb, &smsg_list, list) if (strncmp(msg + 8, cb->prefix, cb->len) == 0) { - cb->callback(msg + 8); + cb->callback(sender, msg + 8); break; } spin_unlock(&smsg_list_lock); @@ -91,7 +100,7 @@ static struct device_driver smsg_driver = { }; int -smsg_register_callback(char *prefix, void (*callback)(char *str)) +smsg_register_callback(char *prefix, void (*callback)(char *from, char *str)) { struct smsg_callback *cb; @@ -108,7 +117,7 @@ smsg_register_callback(char *prefix, void (*callback)(char *str)) } void -smsg_unregister_callback(char *prefix, void (*callback)(char *str)) +smsg_unregister_callback(char *prefix, void (*callback)(char *from, char *str)) { struct smsg_callback *cb, *tmp; diff --git a/drivers/s390/net/smsgiucv.h b/drivers/s390/net/smsgiucv.h index 04cd8715296..67f5d4f8378 100644 --- a/drivers/s390/net/smsgiucv.h +++ b/drivers/s390/net/smsgiucv.h @@ -5,6 +5,6 @@ * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com) */ -int smsg_register_callback(char *, void (*)(char *)); -void smsg_unregister_callback(char *, void (*)(char *)); +int smsg_register_callback(char *, void (*)(char *, char *)); +void smsg_unregister_callback(char *, void (*)(char *, char *)); -- cgit v1.2.3-70-g09d2 From 4beb37097b20b61054b15c56848e4ffcef093819 Mon Sep 17 00:00:00 2001 From: Cornelia Huck Date: Sun, 1 May 2005 08:58:59 -0700 Subject: [PATCH] s390: remove ioctl32 from dasdcmb The ioctl32_conversion routines will be deprecated: Remove them from dasd_cmb and handle the three cmb ioctls like all other dasd ioctls. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/kernel/compat_ioctl.c | 7 ++++++- drivers/s390/block/dasd_cmb.c | 19 ++----------------- include/asm-s390/cmb.h | 2 +- 3 files changed, 9 insertions(+), 19 deletions(-) (limited to 'arch') diff --git a/arch/s390/kernel/compat_ioctl.c b/arch/s390/kernel/compat_ioctl.c index 96571ff7115..03d03c6d3cb 100644 --- a/arch/s390/kernel/compat_ioctl.c +++ b/arch/s390/kernel/compat_ioctl.c @@ -16,6 +16,7 @@ #define CODE #include "../../../fs/compat_ioctl.c" #include +#include #include static int do_ioctl32_pointer(unsigned int fd, unsigned int cmd, @@ -58,7 +59,11 @@ COMPATIBLE_IOCTL(BIODASDPRRD) COMPATIBLE_IOCTL(BIODASDPSRD) COMPATIBLE_IOCTL(BIODASDGATTR) COMPATIBLE_IOCTL(BIODASDSATTR) - +#if defined(CONFIG_DASD_CMB) || defined(CONFIG_DASD_CMB_MODULE) +COMPATIBLE_IOCTL(BIODASDCMFENABLE) +COMPATIBLE_IOCTL(BIODASDCMFDISABLE) +COMPATIBLE_IOCTL(BIODASDREADALLCMB) +#endif #endif #if defined(CONFIG_S390_TAPE) || defined(CONFIG_S390_TAPE_MODULE) diff --git a/drivers/s390/block/dasd_cmb.c b/drivers/s390/block/dasd_cmb.c index ed1ab474c0c..4f365bff275 100644 --- a/drivers/s390/block/dasd_cmb.c +++ b/drivers/s390/block/dasd_cmb.c @@ -1,5 +1,5 @@ /* - * linux/drivers/s390/block/dasd_cmb.c ($Revision: 1.6 $) + * linux/drivers/s390/block/dasd_cmb.c ($Revision: 1.9 $) * * Linux on zSeries Channel Measurement Facility support * (dasd device driver interface) @@ -23,7 +23,6 @@ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #include -#include #include #include #include @@ -84,27 +83,13 @@ dasd_ioctl_readall_cmb(struct block_device *bdev, int no, long args) static inline int ioctl_reg(unsigned int no, dasd_ioctl_fn_t handler) { - int ret; - ret = dasd_ioctl_no_register(THIS_MODULE, no, handler); -#ifdef CONFIG_COMPAT - if (ret) - return ret; - - ret = register_ioctl32_conversion(no, NULL); - if (ret) - dasd_ioctl_no_unregister(THIS_MODULE, no, handler); -#endif - return ret; + return dasd_ioctl_no_register(THIS_MODULE, no, handler); } static inline void ioctl_unreg(unsigned int no, dasd_ioctl_fn_t handler) { dasd_ioctl_no_unregister(THIS_MODULE, no, handler); -#ifdef CONFIG_COMPAT - unregister_ioctl32_conversion(no); -#endif - } static void diff --git a/include/asm-s390/cmb.h b/include/asm-s390/cmb.h index 1bfe2bd630b..dae1dd4fb93 100644 --- a/include/asm-s390/cmb.h +++ b/include/asm-s390/cmb.h @@ -52,7 +52,7 @@ struct cmbdata { #define BIODASDREADALLCMB _IOWR(DASD_IOCTL_LETTER,33,struct cmbdata) #ifdef __KERNEL__ - +struct ccw_device; /** * enable_cmf() - switch on the channel measurement for a specific device * @cdev: The ccw device to be enabled -- cgit v1.2.3-70-g09d2 From cd7619d6bf36564cf54ff7218ef54e558a741913 Mon Sep 17 00:00:00 2001 From: Matt Mackall Date: Sun, 1 May 2005 08:59:01 -0700 Subject: [PATCH] Exterminate PAGE_BUG Remove PAGE_BUG - repalce it with BUG and BUG_ON. Signed-off-by: Matt Mackall Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/arm26/mm/small_page.c | 6 ++---- fs/afs/file.c | 3 +-- fs/buffer.c | 3 +-- fs/jffs2/file.c | 3 +-- fs/udf/file.c | 6 ++---- fs/udf/inode.c | 4 ++-- include/asm-cris/page.h | 4 ---- include/asm-generic/bug.h | 11 ----------- include/asm-sh64/bug.h | 4 ---- mm/filemap.c | 3 +-- 10 files changed, 10 insertions(+), 37 deletions(-) (limited to 'arch') diff --git a/arch/arm26/mm/small_page.c b/arch/arm26/mm/small_page.c index 77be86cca78..30447106c25 100644 --- a/arch/arm26/mm/small_page.c +++ b/arch/arm26/mm/small_page.c @@ -92,8 +92,7 @@ static unsigned long __get_small_page(int priority, struct order *order) page = list_entry(order->queue.next, struct page, lru); again: #ifdef PEDANTIC - if (USED_MAP(page) & ~order->all_used) - PAGE_BUG(page); + BUG_ON(USED_MAP(page) & ~order->all_used); #endif offset = ffz(USED_MAP(page)); SET_USED(page, offset); @@ -141,8 +140,7 @@ static void __free_small_page(unsigned long spage, struct order *order) goto non_small; #ifdef PEDANTIC - if (USED_MAP(page) & ~order->all_used) - PAGE_BUG(page); + BUG_ON(USED_MAP(page) & ~order->all_used); #endif spage = spage >> order->shift; diff --git a/fs/afs/file.c b/fs/afs/file.c index 6b6bb7c8abf..23c12512802 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -131,8 +131,7 @@ static int afs_file_readpage(struct file *file, struct page *page) vnode = AFS_FS_I(inode); - if (!PageLocked(page)) - PAGE_BUG(page); + BUG_ON(!PageLocked(page)); ret = -ESTALE; if (vnode->flags & AFS_VNODE_DELETED) diff --git a/fs/buffer.c b/fs/buffer.c index 188365c7920..792cbacbbf4 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -2078,8 +2078,7 @@ int block_read_full_page(struct page *page, get_block_t *get_block) int nr, i; int fully_mapped = 1; - if (!PageLocked(page)) - PAGE_BUG(page); + BUG_ON(!PageLocked(page)); blocksize = 1 << inode->i_blkbits; if (!page_has_buffers(page)) create_empty_buffers(page, blocksize, 0); diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c index 0c607c1388f..771a554701d 100644 --- a/fs/jffs2/file.c +++ b/fs/jffs2/file.c @@ -79,8 +79,7 @@ static int jffs2_do_readpage_nolock (struct inode *inode, struct page *pg) D2(printk(KERN_DEBUG "jffs2_do_readpage_nolock(): ino #%lu, page at offset 0x%lx\n", inode->i_ino, pg->index << PAGE_CACHE_SHIFT)); - if (!PageLocked(pg)) - PAGE_BUG(pg); + BUG_ON(!PageLocked(pg)); pg_buf = kmap(pg); /* FIXME: Can kmap fail? */ diff --git a/fs/udf/file.c b/fs/udf/file.c index 2faa4172b9f..bb40d63f328 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -49,8 +49,7 @@ static int udf_adinicb_readpage(struct file *file, struct page * page) struct inode *inode = page->mapping->host; char *kaddr; - if (!PageLocked(page)) - PAGE_BUG(page); + BUG_ON(!PageLocked(page)); kaddr = kmap(page); memset(kaddr, 0, PAGE_CACHE_SIZE); @@ -67,8 +66,7 @@ static int udf_adinicb_writepage(struct page *page, struct writeback_control *wb struct inode *inode = page->mapping->host; char *kaddr; - if (!PageLocked(page)) - PAGE_BUG(page); + BUG_ON(!PageLocked(page)); kaddr = kmap(page); memcpy(UDF_I_DATA(inode) + UDF_I_LENEATTR(inode), kaddr, inode->i_size); diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 0506e117378..3d68de39fad 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -167,8 +167,8 @@ void udf_expand_file_adinicb(struct inode * inode, int newsize, int * err) } page = grab_cache_page(inode->i_mapping, 0); - if (!PageLocked(page)) - PAGE_BUG(page); + BUG_ON(!PageLocked(page)); + if (!PageUptodate(page)) { kaddr = kmap(page); diff --git a/include/asm-cris/page.h b/include/asm-cris/page.h index ddd8915e41e..c767da1ef8f 100644 --- a/include/asm-cris/page.h +++ b/include/asm-cris/page.h @@ -77,10 +77,6 @@ typedef struct { unsigned long pgprot; } pgprot_t; printk("kernel BUG at %s:%d!\n", __FILE__, __LINE__); \ } while (0) -#define PAGE_BUG(page) do { \ - BUG(); \ -} while (0) - /* Pure 2^n version of get_order */ static inline int get_order(unsigned long size) { diff --git a/include/asm-generic/bug.h b/include/asm-generic/bug.h index 6e5aaaa9a2f..400c2b41896 100644 --- a/include/asm-generic/bug.h +++ b/include/asm-generic/bug.h @@ -12,13 +12,6 @@ } while (0) #endif -#ifndef HAVE_ARCH_PAGE_BUG -#define PAGE_BUG(page) do { \ - printk("page BUG for page at %p\n", page); \ - BUG(); \ -} while (0) -#endif - #ifndef HAVE_ARCH_BUG_ON #define BUG_ON(condition) do { if (unlikely((condition)!=0)) BUG(); } while(0) #endif @@ -37,10 +30,6 @@ #define BUG() #endif -#ifndef HAVE_ARCH_PAGE_BUG -#define PAGE_BUG(page) do { if (page) ; } while (0) -#endif - #ifndef HAVE_ARCH_BUG_ON #define BUG_ON(condition) do { if (condition) ; } while(0) #endif diff --git a/include/asm-sh64/bug.h b/include/asm-sh64/bug.h index 3acd54d5956..5d659ec28e1 100644 --- a/include/asm-sh64/bug.h +++ b/include/asm-sh64/bug.h @@ -17,10 +17,6 @@ BUG(); \ } while(0) -#define PAGE_BUG(page) do { \ - BUG(); \ -} while (0) - #define WARN_ON(condition) do { \ if (unlikely((condition)!=0)) { \ printk("Badness in %s at %s:%d\n", __FUNCTION__, __FILE__, __LINE__); \ diff --git a/mm/filemap.c b/mm/filemap.c index ee79b5d3439..c085af2332d 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -123,8 +123,7 @@ void remove_from_page_cache(struct page *page) { struct address_space *mapping = page->mapping; - if (unlikely(!PageLocked(page))) - PAGE_BUG(page); + BUG_ON(!PageLocked(page)); write_lock_irq(&mapping->tree_lock); __remove_from_page_cache(page); -- cgit v1.2.3-70-g09d2 From d59745ce3e7aa13856bca16d3bcbb95041775ff6 Mon Sep 17 00:00:00 2001 From: Matt Mackall Date: Sun, 1 May 2005 08:59:02 -0700 Subject: [PATCH] clean up kernel messages Arrange for all kernel printks to be no-ops. Only available if CONFIG_EMBEDDED. This patch saves about 375k on my laptop config and nearly 100k on minimal configs. Signed-off-by: Matt Mackall Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/head.S | 2 ++ include/linux/kernel.h | 9 +++++++++ init/Kconfig | 11 +++++++++++ kernel/printk.c | 27 +++++++++++++++++++++------ 4 files changed, 43 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/head.S b/arch/i386/kernel/head.S index d273fd74619..e966fc8c44c 100644 --- a/arch/i386/kernel/head.S +++ b/arch/i386/kernel/head.S @@ -380,6 +380,7 @@ rp_sidt: ALIGN ignore_int: cld +#ifdef CONFIG_PRINTK pushl %eax pushl %ecx pushl %edx @@ -400,6 +401,7 @@ ignore_int: popl %edx popl %ecx popl %eax +#endif iret /* diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 7c1cba4a527..e25b97062ce 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -115,10 +115,19 @@ extern int __kernel_text_address(unsigned long addr); extern int kernel_text_address(unsigned long addr); extern int session_of_pgrp(int pgrp); +#ifdef CONFIG_PRINTK asmlinkage int vprintk(const char *fmt, va_list args) __attribute__ ((format (printf, 1, 0))); asmlinkage int printk(const char * fmt, ...) __attribute__ ((format (printf, 1, 2))); +#else +static inline int vprintk(const char *s, va_list args) + __attribute__ ((format (printf, 1, 0))); +static inline int vprintk(const char *s, va_list args) { return 0; } +static inline int printk(const char *s, ...) + __attribute__ ((format (printf, 1, 2))); +static inline int printk(const char *s, ...) { return 0; } +#endif unsigned long int_sqrt(unsigned long); diff --git a/init/Kconfig b/init/Kconfig index 42dca393b94..40d286d1d11 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -275,6 +275,17 @@ config KALLSYMS_EXTRA_PASS reported. KALLSYMS_EXTRA_PASS is only a temporary workaround while you wait for kallsyms to be fixed. + +config PRINTK + default y + bool "Enable support for printk" if EMBEDDED + help + This option enables normal printk support. Removing it + eliminates most of the message strings from the kernel image + and makes the kernel more or less silent. As this makes it + very difficult to diagnose system problems, saying N here is + strongly discouraged. + config BUG bool "BUG() support" if EMBEDDED default y diff --git a/kernel/printk.c b/kernel/printk.c index 1498689548d..290a07ce2c8 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -85,10 +85,6 @@ static int console_locked; */ static DEFINE_SPINLOCK(logbuf_lock); -static char __log_buf[__LOG_BUF_LEN]; -static char *log_buf = __log_buf; -static int log_buf_len = __LOG_BUF_LEN; - #define LOG_BUF_MASK (log_buf_len-1) #define LOG_BUF(idx) (log_buf[(idx) & LOG_BUF_MASK]) @@ -99,7 +95,6 @@ static int log_buf_len = __LOG_BUF_LEN; static unsigned long log_start; /* Index into log_buf: next char to be read by syslog() */ static unsigned long con_start; /* Index into log_buf: next char to be sent to consoles */ static unsigned long log_end; /* Index into log_buf: most-recently-written-char + 1 */ -static unsigned long logged_chars; /* Number of chars produced since last read+clear operation */ /* * Array of consoles built from command line options (console=) @@ -120,6 +115,13 @@ static int preferred_console = -1; /* Flag: console code may call schedule() */ static int console_may_schedule; +#ifdef CONFIG_PRINTK + +static char __log_buf[__LOG_BUF_LEN]; +static char *log_buf = __log_buf; +static int log_buf_len = __LOG_BUF_LEN; +static unsigned long logged_chars; /* Number of chars produced since last read+clear operation */ + /* * Setup a list of consoles. Called from init/main.c */ @@ -535,6 +537,7 @@ __setup("time", printk_time_setup); * then changes console_loglevel may break. This is because console_loglevel * is inspected when the actual printing occurs. */ + asmlinkage int printk(const char *fmt, ...) { va_list args; @@ -655,6 +658,18 @@ out: EXPORT_SYMBOL(printk); EXPORT_SYMBOL(vprintk); +#else + +asmlinkage long sys_syslog(int type, char __user * buf, int len) +{ + return 0; +} + +int do_syslog(int type, char __user * buf, int len) { return 0; } +static void call_console_drivers(unsigned long start, unsigned long end) {} + +#endif + /** * acquire_console_sem - lock the console system for exclusive use. * @@ -931,7 +946,7 @@ int unregister_console(struct console * console) return res; } EXPORT_SYMBOL(unregister_console); - + /** * tty_write_message - write a message to a certain tty, not just the console. * -- cgit v1.2.3-70-g09d2 From fbd568a3e61a7decb8a754ad952aaa5b5c82e9e5 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sun, 1 May 2005 08:59:04 -0700 Subject: [PATCH] Change synchronize_kernel to _rcu and _sched This patch changes calls to synchronize_kernel(), deprecated in the earlier "Deprecate synchronize_kernel, GPL replacement" patch to instead call the new synchronize_rcu() and synchronize_sched() APIs. Signed-off-by: Paul E. McKenney Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/oprofile/nmi_timer_int.c | 2 +- arch/ppc64/kernel/HvLpEvent.c | 2 +- drivers/acpi/processor_idle.c | 2 +- drivers/char/ipmi/ipmi_si_intf.c | 4 ++-- drivers/input/keyboard/atkbd.c | 2 +- drivers/md/multipath.c | 2 +- drivers/md/raid1.c | 2 +- drivers/md/raid10.c | 2 +- drivers/md/raid5.c | 2 +- drivers/md/raid6main.c | 2 +- drivers/net/r8169.c | 2 +- drivers/s390/cio/airq.c | 4 ++-- kernel/module.c | 2 +- kernel/profile.c | 2 +- mm/slab.c | 2 +- net/core/dev.c | 2 +- 16 files changed, 18 insertions(+), 18 deletions(-) (limited to 'arch') diff --git a/arch/i386/oprofile/nmi_timer_int.c b/arch/i386/oprofile/nmi_timer_int.c index b2e462abf33..c58d0c14f27 100644 --- a/arch/i386/oprofile/nmi_timer_int.c +++ b/arch/i386/oprofile/nmi_timer_int.c @@ -36,7 +36,7 @@ static void timer_stop(void) { enable_timer_nmi_watchdog(); unset_nmi_callback(); - synchronize_kernel(); + synchronize_sched(); /* Allow already-started NMIs to complete. */ } diff --git a/arch/ppc64/kernel/HvLpEvent.c b/arch/ppc64/kernel/HvLpEvent.c index 9802beefa21..f8f19637f73 100644 --- a/arch/ppc64/kernel/HvLpEvent.c +++ b/arch/ppc64/kernel/HvLpEvent.c @@ -45,7 +45,7 @@ int HvLpEvent_unregisterHandler( HvLpEvent_Type eventType ) /* We now sleep until all other CPUs have scheduled. This ensures that * the deletion is seen by all other CPUs, and that the deleted handler * isn't still running on another CPU when we return. */ - synchronize_kernel(); + synchronize_rcu(); } } return rc; diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 05a17812d52..ff64d333e95 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -838,7 +838,7 @@ int acpi_processor_cst_has_changed (struct acpi_processor *pr) /* Fall back to the default idle loop */ pm_idle = pm_idle_save; - synchronize_kernel(); + synchronize_sched(); /* Relies on interrupts forcing exit from idle. */ pr->flags.power = 0; result = acpi_processor_get_power_info(pr); diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 29de259a981..44a7f13c788 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -2199,7 +2199,7 @@ static int init_one_smi(int intf_num, struct smi_info **smi) /* Wait until we know that we are out of any interrupt handlers might have been running before we freed the interrupt. */ - synchronize_kernel(); + synchronize_sched(); if (new_smi->si_sm) { if (new_smi->handlers) @@ -2312,7 +2312,7 @@ static void __exit cleanup_one_si(struct smi_info *to_clean) /* Wait until we know that we are out of any interrupt handlers might have been running before we freed the interrupt. */ - synchronize_kernel(); + synchronize_sched(); /* Wait for the timer to stop. This avoids problems with race conditions removing the timer here. */ diff --git a/drivers/input/keyboard/atkbd.c b/drivers/input/keyboard/atkbd.c index f7304f0ce54..ff66ed4ee2c 100644 --- a/drivers/input/keyboard/atkbd.c +++ b/drivers/input/keyboard/atkbd.c @@ -678,7 +678,7 @@ static void atkbd_disconnect(struct serio *serio) atkbd_disable(atkbd); /* make sure we don't have a command in flight */ - synchronize_kernel(); + synchronize_sched(); /* Allow atkbd_interrupt()s to complete. */ flush_scheduled_work(); device_remove_file(&serio->dev, &atkbd_attr_extra); diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c index c9b134cd153..1891e4930dc 100644 --- a/drivers/md/multipath.c +++ b/drivers/md/multipath.c @@ -355,7 +355,7 @@ static int multipath_remove_disk(mddev_t *mddev, int number) goto abort; } p->rdev = NULL; - synchronize_kernel(); + synchronize_rcu(); if (atomic_read(&rdev->nr_pending)) { /* lost the race, try later */ err = -EBUSY; diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index a389394b52f..83380b5d659 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -797,7 +797,7 @@ static int raid1_remove_disk(mddev_t *mddev, int number) goto abort; } p->rdev = NULL; - synchronize_kernel(); + synchronize_rcu(); if (atomic_read(&rdev->nr_pending)) { /* lost the race, try later */ err = -EBUSY; diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index b100bfe4fdc..e9dc2876a62 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -977,7 +977,7 @@ static int raid10_remove_disk(mddev_t *mddev, int number) goto abort; } p->rdev = NULL; - synchronize_kernel(); + synchronize_rcu(); if (atomic_read(&rdev->nr_pending)) { /* lost the race, try later */ err = -EBUSY; diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 52c3a81c4aa..e96e2a10a9c 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -1873,7 +1873,7 @@ static int raid5_remove_disk(mddev_t *mddev, int number) goto abort; } p->rdev = NULL; - synchronize_kernel(); + synchronize_rcu(); if (atomic_read(&rdev->nr_pending)) { /* lost the race, try later */ err = -EBUSY; diff --git a/drivers/md/raid6main.c b/drivers/md/raid6main.c index 7e30ab29691..8a33f351e09 100644 --- a/drivers/md/raid6main.c +++ b/drivers/md/raid6main.c @@ -2038,7 +2038,7 @@ static int raid6_remove_disk(mddev_t *mddev, int number) goto abort; } p->rdev = NULL; - synchronize_kernel(); + synchronize_rcu(); if (atomic_read(&rdev->nr_pending)) { /* lost the race, try later */ err = -EBUSY; diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index 07e2df09491..c59507f8a76 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -2385,7 +2385,7 @@ core_down: } /* Give a racing hard_start_xmit a few cycles to complete. */ - synchronize_kernel(); + synchronize_sched(); /* FIXME: should this be synchronize_irq()? */ /* * And now for the 50k$ question: are IRQ disabled or not ? diff --git a/drivers/s390/cio/airq.c b/drivers/s390/cio/airq.c index 3720e77b465..83e6a060668 100644 --- a/drivers/s390/cio/airq.c +++ b/drivers/s390/cio/airq.c @@ -45,7 +45,7 @@ s390_register_adapter_interrupt (adapter_int_handler_t handler) else ret = (cmpxchg(&adapter_handler, NULL, handler) ? -EBUSY : 0); if (!ret) - synchronize_kernel(); + synchronize_sched(); /* Allow interrupts to complete. */ sprintf (dbf_txt, "ret:%d", ret); CIO_TRACE_EVENT (4, dbf_txt); @@ -65,7 +65,7 @@ s390_unregister_adapter_interrupt (adapter_int_handler_t handler) ret = -EINVAL; else { adapter_handler = NULL; - synchronize_kernel(); + synchronize_sched(); /* Allow interrupts to complete. */ ret = 0; } sprintf (dbf_txt, "ret:%d", ret); diff --git a/kernel/module.c b/kernel/module.c index 2dbfa0773fa..5734ab09d3f 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -1801,7 +1801,7 @@ sys_init_module(void __user *umod, /* Init routine failed: abort. Try to protect us from buggy refcounters. */ mod->state = MODULE_STATE_GOING; - synchronize_kernel(); + synchronize_sched(); if (mod->unsafe) printk(KERN_ERR "%s: module is now stuck!\n", mod->name); diff --git a/kernel/profile.c b/kernel/profile.c index a66be468c42..0221a50ca86 100644 --- a/kernel/profile.c +++ b/kernel/profile.c @@ -184,7 +184,7 @@ void unregister_timer_hook(int (*hook)(struct pt_regs *)) WARN_ON(hook != timer_hook); timer_hook = NULL; /* make sure all CPUs see the NULL hook */ - synchronize_kernel(); + synchronize_sched(); /* Allow ongoing interrupts to complete. */ } EXPORT_SYMBOL_GPL(register_timer_hook); diff --git a/mm/slab.c b/mm/slab.c index 771cc09f9f1..84074264115 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -1666,7 +1666,7 @@ int kmem_cache_destroy(kmem_cache_t * cachep) } if (unlikely(cachep->flags & SLAB_DESTROY_BY_RCU)) - synchronize_kernel(); + synchronize_rcu(); /* no cpu_online check required here since we clear the percpu * array on cpu offline and set this to NULL. diff --git a/net/core/dev.c b/net/core/dev.c index 7bd4cd4502c..f5f005846fe 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3091,7 +3091,7 @@ void free_netdev(struct net_device *dev) void synchronize_net(void) { might_sleep(); - synchronize_kernel(); + synchronize_rcu(); } /** -- cgit v1.2.3-70-g09d2 From e49332bd12e92da2df6d002f857ec62675ba2648 Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Sun, 1 May 2005 08:59:08 -0700 Subject: [PATCH] misc verify_area cleanups There were still a few comments left refering to verify_area, and two functions, verify_area_skas & verify_area_tt that just wrap corresponding access_ok_skas & access_ok_tt functions, just like verify_area does for access_ok - deprecate those. There was also a few places that still used verify_area in commented-out code, fix those up to use access_ok. After applying this one there should not be anything left but finally removing verify_area completely, which will happen after a kernel release or two. Signed-off-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/vm86.c | 2 +- arch/um/kernel/skas/include/uaccess-skas.h | 2 +- arch/um/kernel/tt/include/uaccess-tt.h | 2 +- drivers/char/dtlk.c | 2 +- drivers/char/specialix.c | 11 ++++------- include/asm-frv/pgtable.h | 4 ++-- include/asm-i386/checksum.h | 2 +- include/asm-i386/pgtable.h | 4 ++-- include/asm-parisc/uaccess.h | 2 +- include/asm-sh/checksum.h | 2 +- include/asm-sh64/checksum.h | 2 +- include/asm-sparc/uaccess.h | 2 +- net/8021q/vlanproc.c | 2 +- net/atm/common.c | 4 ++-- net/core/iovec.c | 2 +- net/wanrouter/wanmain.c | 4 ++-- 16 files changed, 23 insertions(+), 26 deletions(-) (limited to 'arch') diff --git a/arch/i386/kernel/vm86.c b/arch/i386/kernel/vm86.c index d3b4c540eb6..ec0f68ce688 100644 --- a/arch/i386/kernel/vm86.c +++ b/arch/i386/kernel/vm86.c @@ -222,7 +222,7 @@ asmlinkage int sys_vm86(struct pt_regs regs) goto out; case VM86_PLUS_INSTALL_CHECK: /* NOTE: on old vm86 stuff this will return the error - from verify_area(), because the subfunction is + from access_ok(), because the subfunction is interpreted as (invalid) address to vm86_struct. So the installation check works. */ diff --git a/arch/um/kernel/skas/include/uaccess-skas.h b/arch/um/kernel/skas/include/uaccess-skas.h index 11986c9b9dd..c35620385da 100644 --- a/arch/um/kernel/skas/include/uaccess-skas.h +++ b/arch/um/kernel/skas/include/uaccess-skas.h @@ -18,7 +18,7 @@ ((unsigned long) (addr) + (size) <= FIXADDR_USER_END) && \ ((unsigned long) (addr) + (size) >= (unsigned long)(addr)))) -static inline int verify_area_skas(int type, const void * addr, +static inline int __deprecated verify_area_skas(int type, const void * addr, unsigned long size) { return(access_ok_skas(type, addr, size) ? 0 : -EFAULT); diff --git a/arch/um/kernel/tt/include/uaccess-tt.h b/arch/um/kernel/tt/include/uaccess-tt.h index f0bad010ceb..bb69d6b7d02 100644 --- a/arch/um/kernel/tt/include/uaccess-tt.h +++ b/arch/um/kernel/tt/include/uaccess-tt.h @@ -33,7 +33,7 @@ extern unsigned long uml_physmem; (((unsigned long) (addr) <= ((unsigned long) (addr) + (size))) && \ (under_task_size(addr, size) || is_stack(addr, size)))) -static inline int verify_area_tt(int type, const void * addr, +static inline int __deprecated verify_area_tt(int type, const void * addr, unsigned long size) { return(access_ok_tt(type, addr, size) ? 0 : -EFAULT); diff --git a/drivers/char/dtlk.c b/drivers/char/dtlk.c index 903e4c3cc20..a229915ce1b 100644 --- a/drivers/char/dtlk.c +++ b/drivers/char/dtlk.c @@ -52,7 +52,7 @@ #define KERNEL #include #include -#include /* for verify_area */ +#include #include /* for -EBUSY */ #include /* for request_region */ #include /* for loops_per_jiffy */ diff --git a/drivers/char/specialix.c b/drivers/char/specialix.c index c789d5ceac7..50e0b612a8a 100644 --- a/drivers/char/specialix.c +++ b/drivers/char/specialix.c @@ -1987,10 +1987,9 @@ static inline int sx_set_serial_info(struct specialix_port * port, func_enter(); /* - error = verify_area(VERIFY_READ, (void *) newinfo, sizeof(tmp)); - if (error) { + if (!access_ok(VERIFY_READ, (void *) newinfo, sizeof(tmp))) { func_exit(); - return error; + return -EFAULT; } */ if (copy_from_user(&tmp, newinfo, sizeof(tmp))) { @@ -2046,14 +2045,12 @@ static inline int sx_get_serial_info(struct specialix_port * port, { struct serial_struct tmp; struct specialix_board *bp = port_Board(port); - // int error; func_enter(); /* - error = verify_area(VERIFY_WRITE, (void *) retinfo, sizeof(tmp)); - if (error) - return error; + if (!access_ok(VERIFY_WRITE, (void *) retinfo, sizeof(tmp))) + return -EFAULT; */ memset(&tmp, 0, sizeof(tmp)); diff --git a/include/asm-frv/pgtable.h b/include/asm-frv/pgtable.h index 3c6d42a22df..d0a9c2f9c13 100644 --- a/include/asm-frv/pgtable.h +++ b/include/asm-frv/pgtable.h @@ -349,9 +349,9 @@ static inline pmd_t *pmd_offset(pud_t *dir, unsigned long address) /* * Define this to warn about kernel memory accesses that are - * done without a 'verify_area(VERIFY_WRITE,..)' + * done without a 'access_ok(VERIFY_WRITE,..)' */ -#undef TEST_VERIFY_AREA +#undef TEST_ACCESS_OK #define pte_present(x) (pte_val(x) & _PAGE_PRESENT) #define pte_clear(mm,addr,xp) do { set_pte_at(mm, addr, xp, __pte(0)); } while (0) diff --git a/include/asm-i386/checksum.h b/include/asm-i386/checksum.h index d76a5f081c9..641342002bc 100644 --- a/include/asm-i386/checksum.h +++ b/include/asm-i386/checksum.h @@ -33,7 +33,7 @@ asmlinkage unsigned int csum_partial_copy_generic(const unsigned char *src, unsi * passed in an incorrect kernel address to one of these functions. * * If you use these functions directly please don't forget the - * verify_area(). + * access_ok(). */ static __inline__ unsigned int csum_partial_copy_nocheck (const unsigned char *src, unsigned char *dst, diff --git a/include/asm-i386/pgtable.h b/include/asm-i386/pgtable.h index 5c725425d86..8d60c2b4b00 100644 --- a/include/asm-i386/pgtable.h +++ b/include/asm-i386/pgtable.h @@ -193,9 +193,9 @@ extern unsigned long long __PAGE_KERNEL, __PAGE_KERNEL_EXEC; /* * Define this if things work differently on an i386 and an i486: * it will (on an i486) warn about kernel memory accesses that are - * done without a 'verify_area(VERIFY_WRITE,..)' + * done without a 'access_ok(VERIFY_WRITE,..)' */ -#undef TEST_VERIFY_AREA +#undef TEST_ACCESS_OK /* The boot page tables (all created as a single array) */ extern unsigned long pg0[]; diff --git a/include/asm-parisc/uaccess.h b/include/asm-parisc/uaccess.h index 8a08423b757..c1b5bdea53e 100644 --- a/include/asm-parisc/uaccess.h +++ b/include/asm-parisc/uaccess.h @@ -24,7 +24,7 @@ /* * Note that since kernel addresses are in a separate address space on - * parisc, we don't need to do anything for access_ok() or verify_area(). + * parisc, we don't need to do anything for access_ok(). * We just let the page fault handler do the right thing. This also means * that put_user is the same as __put_user, etc. */ diff --git a/include/asm-sh/checksum.h b/include/asm-sh/checksum.h index 5113c7f8a73..5ebd0f24299 100644 --- a/include/asm-sh/checksum.h +++ b/include/asm-sh/checksum.h @@ -42,7 +42,7 @@ asmlinkage unsigned int csum_partial_copy_generic(const unsigned char *src, unsi * passed in an incorrect kernel address to one of these functions. * * If you use these functions directly please don't forget the - * verify_area(). + * access_ok(). */ static __inline__ unsigned int csum_partial_copy_nocheck (const unsigned char *src, unsigned char *dst, diff --git a/include/asm-sh64/checksum.h b/include/asm-sh64/checksum.h index aa3911a9949..fd034e9ae6e 100644 --- a/include/asm-sh64/checksum.h +++ b/include/asm-sh64/checksum.h @@ -34,7 +34,7 @@ asmlinkage unsigned int csum_partial(const unsigned char *buff, int len, * passed in an incorrect kernel address to one of these functions. * * If you use these functions directly please don't forget the - * verify_area(). + * access_ok(). */ diff --git a/include/asm-sparc/uaccess.h b/include/asm-sparc/uaccess.h index 3f47889883b..f461144067e 100644 --- a/include/asm-sparc/uaccess.h +++ b/include/asm-sparc/uaccess.h @@ -18,7 +18,7 @@ #ifndef __ASSEMBLY__ -/* Sparc is not segmented, however we need to be able to fool verify_area() +/* Sparc is not segmented, however we need to be able to fool access_ok() * when doing system calls from kernel mode legitimately. * * "For historical reasons, these macros are grossly misnamed." -Linus diff --git a/net/8021q/vlanproc.c b/net/8021q/vlanproc.c index c32d27af0a3..7b214cffc95 100644 --- a/net/8021q/vlanproc.c +++ b/net/8021q/vlanproc.c @@ -23,7 +23,7 @@ #include /* return codes */ #include #include /* kmalloc(), kfree() */ -#include /* verify_area(), etc. */ +#include #include /* inline mem*, str* functions */ #include /* __initfunc et al. */ #include /* htons(), etc. */ diff --git a/net/atm/common.c b/net/atm/common.c index 6d16be334ea..e93e838069e 100644 --- a/net/atm/common.c +++ b/net/atm/common.c @@ -12,7 +12,7 @@ #include /* SOL_SOCKET */ #include /* error codes */ #include -#include /* verify_area */ +#include #include #include /* struct timeval */ #include @@ -540,7 +540,7 @@ int vcc_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *m, error = -EMSGSIZE; goto out; } - /* verify_area is done by net/socket.c */ + eff = (size+3) & ~3; /* align to word boundary */ prepare_to_wait(sk->sk_sleep, &wait, TASK_INTERRUPTIBLE); error = 0; diff --git a/net/core/iovec.c b/net/core/iovec.c index d57ace949ab..65e4b56fbc7 100644 --- a/net/core/iovec.c +++ b/net/core/iovec.c @@ -33,7 +33,7 @@ * Verify iovec. The caller must ensure that the iovec is big enough * to hold the message iovec. * - * Save time not doing verify_area. copy_*_user will make this work + * Save time not doing access_ok. copy_*_user will make this work * in any case. */ diff --git a/net/wanrouter/wanmain.c b/net/wanrouter/wanmain.c index 956c17f6c54..d6844ac226f 100644 --- a/net/wanrouter/wanmain.c +++ b/net/wanrouter/wanmain.c @@ -48,8 +48,8 @@ #include #include #include /* support for loadable modules */ -#include /* kmalloc(), kfree() */ -#include /* verify_area(), etc. */ +#include /* kmalloc(), kfree() */ +#include #include /* inline mem*, str* functions */ #include /* htons(), etc. */ -- cgit v1.2.3-70-g09d2 From 7d87e14c236d6c4cab66d87cf0bc1e0f0375d308 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Sun, 1 May 2005 08:59:12 -0700 Subject: [PATCH] consolidate sys_shmat Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/kernel/osf_sys.c | 16 ---------------- arch/alpha/kernel/systbls.S | 2 +- arch/arm/kernel/sys_arm.c | 12 ------------ arch/ia64/kernel/entry.S | 2 +- arch/ia64/kernel/sys_ia64.c | 14 -------------- arch/mips/kernel/syscall.c | 16 ---------------- arch/parisc/kernel/sys_parisc.c | 11 ----------- arch/parisc/kernel/syscall_table.S | 2 +- arch/sh64/kernel/sys_sh64.c | 15 --------------- arch/sh64/kernel/syscalls.S | 2 +- arch/um/include/sysdep-x86_64/syscalls.h | 1 - arch/um/sys-x86_64/syscalls.c | 8 -------- arch/x86_64/kernel/sys_x86_64.c | 6 ------ include/asm-x86_64/unistd.h | 2 +- include/linux/syscalls.h | 3 +-- ipc/shm.c | 14 ++++++++++++++ kernel/sys_ni.c | 1 + 17 files changed, 21 insertions(+), 106 deletions(-) (limited to 'arch') diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c index b5d0fd2bb10..64e450dddb4 100644 --- a/arch/alpha/kernel/osf_sys.c +++ b/arch/alpha/kernel/osf_sys.c @@ -457,22 +457,6 @@ osf_getdomainname(char __user *name, int namelen) return 0; } -asmlinkage long -osf_shmat(int shmid, void __user *shmaddr, int shmflg) -{ - unsigned long raddr; - long err; - - err = do_shmat(shmid, shmaddr, shmflg, &raddr); - - /* - * This works because all user-level addresses are - * non-negative longs! - */ - return err ? err : (long)raddr; -} - - /* * The following stuff should move into a header file should it ever * be labeled "officially supported." Right now, there is just enough diff --git a/arch/alpha/kernel/systbls.S b/arch/alpha/kernel/systbls.S index 3864b33562e..05212088287 100644 --- a/arch/alpha/kernel/systbls.S +++ b/arch/alpha/kernel/systbls.S @@ -227,7 +227,7 @@ sys_call_table: .quad sys_semop .quad osf_utsname .quad sys_lchown - .quad osf_shmat + .quad sys_shmat .quad sys_shmctl /* 210 */ .quad sys_shmdt .quad sys_shmget diff --git a/arch/arm/kernel/sys_arm.c b/arch/arm/kernel/sys_arm.c index 7ba6342cf93..ef32577da30 100644 --- a/arch/arm/kernel/sys_arm.c +++ b/arch/arm/kernel/sys_arm.c @@ -227,18 +227,6 @@ asmlinkage int sys_ipc(uint call, int first, int second, int third, } } -asmlinkage long sys_shmat(int shmid, char __user *shmaddr, int shmflg, - unsigned long __user *addr) -{ - unsigned long ret; - long err; - - err = do_shmat(shmid, shmaddr, shmflg, &ret); - if (err == 0) - err = put_user(ret, addr); - return err; -} - /* Fork a new task - this creates a new program thread. * This is called indirectly via a small wrapper */ diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S index bd86fea49a0..d3f093820bc 100644 --- a/arch/ia64/kernel/entry.S +++ b/arch/ia64/kernel/entry.S @@ -1417,7 +1417,7 @@ sys_call_table: data8 sys_msgrcv data8 sys_msgctl data8 sys_shmget - data8 ia64_shmat + data8 sys_shmat data8 sys_shmdt // 1115 data8 sys_shmctl data8 sys_syslog diff --git a/arch/ia64/kernel/sys_ia64.c b/arch/ia64/kernel/sys_ia64.c index 3ac216e1c8b..a8cf6d8a509 100644 --- a/arch/ia64/kernel/sys_ia64.c +++ b/arch/ia64/kernel/sys_ia64.c @@ -92,20 +92,6 @@ sys_getpagesize (void) return PAGE_SIZE; } -asmlinkage unsigned long -ia64_shmat (int shmid, void __user *shmaddr, int shmflg) -{ - unsigned long raddr; - int retval; - - retval = do_shmat(shmid, shmaddr, shmflg, &raddr); - if (retval < 0) - return retval; - - force_successful_syscall_return(); - return raddr; -} - asmlinkage unsigned long ia64_brk (unsigned long brk) { diff --git a/arch/mips/kernel/syscall.c b/arch/mips/kernel/syscall.c index 598bfe7426a..ae2a1312d4e 100644 --- a/arch/mips/kernel/syscall.c +++ b/arch/mips/kernel/syscall.c @@ -373,22 +373,6 @@ asmlinkage int sys_ipc (uint call, int first, int second, } } -/* - * Native ABI that is O32 or N64 version - */ -asmlinkage long sys_shmat(int shmid, char __user *shmaddr, - int shmflg, unsigned long *addr) -{ - unsigned long raddr; - int err; - - err = do_shmat(shmid, shmaddr, shmflg, &raddr); - if (err) - return err; - - return put_user(raddr, addr); -} - /* * No implemented yet ... */ diff --git a/arch/parisc/kernel/sys_parisc.c b/arch/parisc/kernel/sys_parisc.c index 7958cd8c8bf..d15a1d53e10 100644 --- a/arch/parisc/kernel/sys_parisc.c +++ b/arch/parisc/kernel/sys_parisc.c @@ -161,17 +161,6 @@ asmlinkage unsigned long sys_mmap(unsigned long addr, unsigned long len, } } -long sys_shmat_wrapper(int shmid, char __user *shmaddr, int shmflag) -{ - unsigned long raddr; - int r; - - r = do_shmat(shmid, shmaddr, shmflag, &raddr); - if (r < 0) - return r; - return raddr; -} - /* Fucking broken ABI */ #ifdef CONFIG_64BIT diff --git a/arch/parisc/kernel/syscall_table.S b/arch/parisc/kernel/syscall_table.S index 779b537100e..dcfa4d3d0e7 100644 --- a/arch/parisc/kernel/syscall_table.S +++ b/arch/parisc/kernel/syscall_table.S @@ -297,7 +297,7 @@ ENTRY_DIFF(msgrcv) ENTRY_SAME(msgget) /* 190 */ ENTRY_SAME(msgctl) - ENTRY_SAME(shmat_wrapper) + ENTRY_SAME(shmat) ENTRY_SAME(shmdt) ENTRY_SAME(shmget) ENTRY_SAME(shmctl) /* 195 */ diff --git a/arch/sh64/kernel/sys_sh64.c b/arch/sh64/kernel/sys_sh64.c index 4546845b9ca..58ff7d522d8 100644 --- a/arch/sh64/kernel/sys_sh64.c +++ b/arch/sh64/kernel/sys_sh64.c @@ -283,18 +283,3 @@ asmlinkage int sys_uname(struct old_utsname * name) up_read(&uts_sem); return err?-EFAULT:0; } - -/* Copy from mips version */ -asmlinkage long sys_shmatcall(int shmid, char __user *shmaddr, - int shmflg) -{ - unsigned long raddr; - int err; - - err = do_shmat(shmid, shmaddr, shmflg, &raddr); - if (err) - return err; - - err = raddr; - return err; -} diff --git a/arch/sh64/kernel/syscalls.S b/arch/sh64/kernel/syscalls.S index 8ed417df3dc..6aabc63e451 100644 --- a/arch/sh64/kernel/syscalls.S +++ b/arch/sh64/kernel/syscalls.S @@ -268,7 +268,7 @@ sys_call_table: .long sys_msgrcv .long sys_msgget .long sys_msgctl - .long sys_shmatcall + .long sys_shmat .long sys_shmdt /* 245 */ .long sys_shmget .long sys_shmctl diff --git a/arch/um/include/sysdep-x86_64/syscalls.h b/arch/um/include/sysdep-x86_64/syscalls.h index b56b335c351..67923cca569 100644 --- a/arch/um/include/sysdep-x86_64/syscalls.h +++ b/arch/um/include/sysdep-x86_64/syscalls.h @@ -26,7 +26,6 @@ extern syscall_handler_t *ia32_sys_call_table[]; extern long old_mmap(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags, unsigned long fd, unsigned long pgoff); -extern syscall_handler_t wrap_sys_shmat; extern syscall_handler_t sys_modify_ldt; extern syscall_handler_t sys_arch_prctl; diff --git a/arch/um/sys-x86_64/syscalls.c b/arch/um/sys-x86_64/syscalls.c index 15768c96ceb..ab4b0abf8af 100644 --- a/arch/um/sys-x86_64/syscalls.c +++ b/arch/um/sys-x86_64/syscalls.c @@ -14,14 +14,6 @@ #include "asm/prctl.h" /* XXX This should get the constants from libc */ #include "choose-mode.h" -/* XXX: copied from x86-64: arch/x86_64/kernel/sys_x86_64.c */ -asmlinkage long wrap_sys_shmat(int shmid, char __user *shmaddr, int shmflg) -{ - unsigned long raddr; - - return do_shmat(shmid, shmaddr, shmflg, &raddr) ?: (long) raddr; -} - asmlinkage long sys_uname64(struct new_utsname __user * name) { int err; diff --git a/arch/x86_64/kernel/sys_x86_64.c b/arch/x86_64/kernel/sys_x86_64.c index 477d8be57d6..dbebd5ccba6 100644 --- a/arch/x86_64/kernel/sys_x86_64.c +++ b/arch/x86_64/kernel/sys_x86_64.c @@ -152,12 +152,6 @@ asmlinkage long sys_uname(struct new_utsname __user * name) return err ? -EFAULT : 0; } -asmlinkage long wrap_sys_shmat(int shmid, char __user *shmaddr, int shmflg) -{ - unsigned long raddr; - return do_shmat(shmid,shmaddr,shmflg,&raddr) ?: (long)raddr; -} - asmlinkage long sys_time64(long __user * tloc) { struct timeval now; diff --git a/include/asm-x86_64/unistd.h b/include/asm-x86_64/unistd.h index 3d65d240dc9..3c9af6fd433 100644 --- a/include/asm-x86_64/unistd.h +++ b/include/asm-x86_64/unistd.h @@ -76,7 +76,7 @@ __SYSCALL(__NR_madvise, sys_madvise) #define __NR_shmget 29 __SYSCALL(__NR_shmget, sys_shmget) #define __NR_shmat 30 -__SYSCALL(__NR_shmat, wrap_sys_shmat) +__SYSCALL(__NR_shmat, sys_shmat) #define __NR_shmctl 31 __SYSCALL(__NR_shmctl, sys_shmctl) diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 757cd9be774..c39f6f72cbb 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -456,8 +456,7 @@ asmlinkage long sys_semctl(int semid, int semnum, int cmd, union semun arg); asmlinkage long sys_semtimedop(int semid, struct sembuf __user *sops, unsigned nsops, const struct timespec __user *timeout); -asmlinkage long sys_shmat(int shmid, char __user *shmaddr, - int shmflg, unsigned long __user *addr); +asmlinkage long sys_shmat(int shmid, char __user *shmaddr, int shmflg); asmlinkage long sys_shmget(key_t key, size_t size, int flag); asmlinkage long sys_shmdt(char __user *shmaddr); asmlinkage long sys_shmctl(int shmid, int cmd, struct shmid_ds __user *buf); diff --git a/ipc/shm.c b/ipc/shm.c index 06cd5c91056..cce022435db 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -28,6 +28,8 @@ #include #include #include +#include + #include #include "util.h" @@ -771,6 +773,18 @@ out: return err; } +asmlinkage long sys_shmat(int shmid, char __user *shmaddr, int shmflg) +{ + unsigned long ret; + long err; + + err = do_shmat(shmid, shmaddr, shmflg, &ret); + if (err) + return err; + force_successful_syscall_return(); + return (long)ret; +} + /* * detach and kill segment if marked destroyed. * The work is done in shm_close. diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c index 1802a311dd3..0dda70ed1f9 100644 --- a/kernel/sys_ni.c +++ b/kernel/sys_ni.c @@ -52,6 +52,7 @@ cond_syscall(sys_msgsnd); cond_syscall(sys_msgrcv); cond_syscall(sys_msgctl); cond_syscall(sys_shmget); +cond_syscall(sys_shmat); cond_syscall(sys_shmdt); cond_syscall(sys_shmctl); cond_syscall(sys_mq_open); -- cgit v1.2.3-70-g09d2 From 7ed20e1ad521b5f5df61bf6559ae60738e393741 Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Sun, 1 May 2005 08:59:14 -0700 Subject: [PATCH] convert that currently tests _NSIG directly to use valid_signal() Convert most of the current code that uses _NSIG directly to instead use valid_signal(). This avoids gcc -W warnings and off-by-one errors. Signed-off-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/kernel/ptrace.c | 5 +++-- arch/arm/kernel/ptrace.c | 5 +++-- arch/arm26/kernel/ptrace.c | 5 +++-- arch/cris/arch-v10/kernel/ptrace.c | 5 +++-- arch/frv/kernel/ptrace.c | 5 +++-- arch/h8300/kernel/ptrace.c | 5 +++-- arch/i386/kernel/ptrace.c | 5 +++-- arch/ia64/kernel/ptrace.c | 5 +++-- arch/m32r/kernel/ptrace.c | 5 +++-- arch/m68k/kernel/ptrace.c | 5 +++-- arch/m68knommu/kernel/ptrace.c | 5 +++-- arch/mips/kernel/ptrace.c | 3 ++- arch/mips/kernel/ptrace32.c | 3 ++- arch/parisc/kernel/ptrace.c | 7 ++++--- arch/ppc/kernel/ptrace.c | 5 +++-- arch/ppc64/kernel/ptrace.c | 5 +++-- arch/ppc64/kernel/ptrace32.c | 5 +++-- arch/s390/kernel/ptrace.c | 5 +++-- arch/sh/kernel/ptrace.c | 5 +++-- arch/sh64/kernel/ptrace.c | 5 +++-- arch/sparc/kernel/ptrace.c | 3 ++- arch/sparc64/kernel/ptrace.c | 3 ++- arch/um/kernel/ptrace.c | 4 ++-- arch/v850/kernel/ptrace.c | 3 ++- arch/x86_64/kernel/ptrace.c | 5 +++-- drivers/char/vt_ioctl.c | 3 ++- fs/fcntl.c | 3 ++- ipc/mqueue.c | 4 ++-- kernel/exit.c | 5 +++-- kernel/futex.c | 3 ++- kernel/ptrace.c | 3 ++- kernel/signal.c | 9 +++++---- kernel/sys.c | 3 ++- 33 files changed, 90 insertions(+), 59 deletions(-) (limited to 'arch') diff --git a/arch/alpha/kernel/ptrace.c b/arch/alpha/kernel/ptrace.c index d0058316157..bbd37536d14 100644 --- a/arch/alpha/kernel/ptrace.c +++ b/arch/alpha/kernel/ptrace.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -335,7 +336,7 @@ do_sys_ptrace(long request, long pid, long addr, long data, /* continue and stop at next (return from) syscall */ case PTRACE_CONT: /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -365,7 +366,7 @@ do_sys_ptrace(long request, long pid, long addr, long data, case PTRACE_SINGLESTEP: /* execute single instruction. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; /* Mark single stepping. */ child->thread_info->bpt_nsaved = -1; diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index efd7a341614..cd99b83f14c 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include @@ -693,7 +694,7 @@ static int do_ptrace(int request, struct task_struct *child, long addr, long dat case PTRACE_SYSCALL: case PTRACE_CONT: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -728,7 +729,7 @@ static int do_ptrace(int request, struct task_struct *child, long addr, long dat */ case PTRACE_SINGLESTEP: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; child->ptrace |= PT_SINGLESTEP; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/arm26/kernel/ptrace.c b/arch/arm26/kernel/ptrace.c index 2a137146a77..8a52124de0e 100644 --- a/arch/arm26/kernel/ptrace.c +++ b/arch/arm26/kernel/ptrace.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -591,7 +592,7 @@ static int do_ptrace(int request, struct task_struct *child, long addr, long dat case PTRACE_SYSCALL: case PTRACE_CONT: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -626,7 +627,7 @@ static int do_ptrace(int request, struct task_struct *child, long addr, long dat */ case PTRACE_SINGLESTEP: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; child->ptrace |= PT_SINGLESTEP; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/cris/arch-v10/kernel/ptrace.c b/arch/cris/arch-v10/kernel/ptrace.c index da15db8ae48..581ecabaae5 100644 --- a/arch/cris/arch-v10/kernel/ptrace.c +++ b/arch/cris/arch-v10/kernel/ptrace.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include @@ -184,7 +185,7 @@ sys_ptrace(long request, long pid, long addr, long data) case PTRACE_CONT: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) { @@ -219,7 +220,7 @@ sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/frv/kernel/ptrace.c b/arch/frv/kernel/ptrace.c index 2a0efb739ad..cbe03cba9f0 100644 --- a/arch/frv/kernel/ptrace.c +++ b/arch/frv/kernel/ptrace.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include @@ -239,7 +240,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -267,7 +268,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); ptrace_enable(child); diff --git a/arch/h8300/kernel/ptrace.c b/arch/h8300/kernel/ptrace.c index 5f19d774a28..05c15e86977 100644 --- a/arch/h8300/kernel/ptrace.c +++ b/arch/h8300/kernel/ptrace.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include @@ -171,7 +172,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data >= _NSIG) + if (!valid_signal(data)) break ; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -202,7 +203,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: { /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); child->exit_code = data; diff --git a/arch/i386/kernel/ptrace.c b/arch/i386/kernel/ptrace.c index b2f17640cef..e8c965ce86e 100644 --- a/arch/i386/kernel/ptrace.c +++ b/arch/i386/kernel/ptrace.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include @@ -511,7 +512,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) { set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -543,7 +544,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); set_singlestep(child); diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 55789fcd721..c253fd5914f 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include @@ -1481,7 +1482,7 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) case PTRACE_CONT: /* restart after signal. */ ret = -EIO; - if (data > _NSIG) + if (!valid_signal(data)) goto out_tsk; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -1520,7 +1521,7 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) /* let child execute for one instruction */ case PTRACE_SINGLEBLOCK: ret = -EIO; - if (data > _NSIG) + if (!valid_signal(data)) goto out_tsk; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/m32r/kernel/ptrace.c b/arch/m32r/kernel/ptrace.c index 8b40f362dd6..124f7c1b775 100644 --- a/arch/m32r/kernel/ptrace.c +++ b/arch/m32r/kernel/ptrace.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include @@ -665,7 +666,7 @@ do_ptrace(long request, struct task_struct *child, long addr, long data) case PTRACE_SYSCALL: case PTRACE_CONT: ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -700,7 +701,7 @@ do_ptrace(long request, struct task_struct *child, long addr, long data) unsigned long pc, insn; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); if ((child->ptrace & PT_DTRACE) == 0) { diff --git a/arch/m68k/kernel/ptrace.c b/arch/m68k/kernel/ptrace.c index 0beb53333ba..f4e1e5eb8e1 100644 --- a/arch/m68k/kernel/ptrace.c +++ b/arch/m68k/kernel/ptrace.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include @@ -251,7 +252,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) long tmp; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) { child->thread.work.syscall_trace = ~0; @@ -292,7 +293,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) long tmp; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; child->thread.work.syscall_trace = 0; tmp = get_reg(child, PT_SR) | (TRACE_BITS << 16); diff --git a/arch/m68knommu/kernel/ptrace.c b/arch/m68knommu/kernel/ptrace.c index 15cf79080b1..9724e1cd82e 100644 --- a/arch/m68knommu/kernel/ptrace.c +++ b/arch/m68knommu/kernel/ptrace.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include @@ -240,7 +241,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) long tmp; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -278,7 +279,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) long tmp; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); tmp = get_reg(child, PT_SR) | (TRACE_BITS << 16); diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index 92f2c39afe2..a2f899c2f4d 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include @@ -257,7 +258,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) { set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/mips/kernel/ptrace32.c b/arch/mips/kernel/ptrace32.c index 611dee919d5..eee207969c2 100644 --- a/arch/mips/kernel/ptrace32.c +++ b/arch/mips/kernel/ptrace32.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include @@ -241,7 +242,7 @@ asmlinkage int sys32_ptrace(int request, int pid, int addr, int data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned int) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) { set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/parisc/kernel/ptrace.c b/arch/parisc/kernel/ptrace.c index 2937a923638..c07db9dff7c 100644 --- a/arch/parisc/kernel/ptrace.c +++ b/arch/parisc/kernel/ptrace.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include @@ -285,7 +286,7 @@ long sys_ptrace(long request, pid_t pid, long addr, long data) ret = -EIO; DBG("sys_ptrace(%s)\n", request == PTRACE_SYSCALL ? "SYSCALL" : "CONT"); - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) goto out_tsk; child->ptrace &= ~(PT_SINGLESTEP|PT_BLOCKSTEP); if (request == PTRACE_SYSCALL) { @@ -311,7 +312,7 @@ long sys_ptrace(long request, pid_t pid, long addr, long data) case PTRACE_SINGLEBLOCK: DBG("sys_ptrace(SINGLEBLOCK)\n"); ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) goto out_tsk; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); child->ptrace &= ~PT_SINGLESTEP; @@ -328,7 +329,7 @@ long sys_ptrace(long request, pid_t pid, long addr, long data) case PTRACE_SINGLESTEP: DBG("sys_ptrace(SINGLESTEP)\n"); ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) goto out_tsk; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); diff --git a/arch/ppc/kernel/ptrace.c b/arch/ppc/kernel/ptrace.c index 426b6f7d9de..59d59a8dc24 100644 --- a/arch/ppc/kernel/ptrace.c +++ b/arch/ppc/kernel/ptrace.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include @@ -356,7 +357,7 @@ int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) { set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -389,7 +390,7 @@ int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: { /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); set_single_step(child); diff --git a/arch/ppc64/kernel/ptrace.c b/arch/ppc64/kernel/ptrace.c index 354a287c67e..5a846324ca8 100644 --- a/arch/ppc64/kernel/ptrace.c +++ b/arch/ppc64/kernel/ptrace.c @@ -28,6 +28,7 @@ #include #include #include +#include #include #include @@ -162,7 +163,7 @@ int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -194,7 +195,7 @@ int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: { /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); set_single_step(child); diff --git a/arch/ppc64/kernel/ptrace32.c b/arch/ppc64/kernel/ptrace32.c index ee81b1b776c..16436426c7e 100644 --- a/arch/ppc64/kernel/ptrace32.c +++ b/arch/ppc64/kernel/ptrace32.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include @@ -293,7 +294,7 @@ int sys32_ptrace(long request, long pid, unsigned long addr, unsigned long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -325,7 +326,7 @@ int sys32_ptrace(long request, long pid, unsigned long addr, unsigned long data) case PTRACE_SINGLESTEP: { /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); set_single_step(child); diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 647233c02fc..9f0d73e3f5f 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include @@ -609,7 +610,7 @@ do_ptrace(struct task_struct *child, long request, long addr, long data) /* continue and stop at next (return from) syscall */ case PTRACE_CONT: /* restart after signal. */ - if ((unsigned long) data >= _NSIG) + if (!valid_signal(data)) return -EIO; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -637,7 +638,7 @@ do_ptrace(struct task_struct *child, long request, long addr, long data) case PTRACE_SINGLESTEP: /* set the trap flag. */ - if ((unsigned long) data >= _NSIG) + if (!valid_signal(data)) return -EIO; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); child->exit_code = data; diff --git a/arch/sh/kernel/ptrace.c b/arch/sh/kernel/ptrace.c index 1b0dfb4d8ea..b28919b6568 100644 --- a/arch/sh/kernel/ptrace.c +++ b/arch/sh/kernel/ptrace.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include @@ -197,7 +198,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -228,7 +229,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) struct pt_regs *dummy = NULL; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); if ((child->ptrace & PT_DTRACE) == 0) { diff --git a/arch/sh64/kernel/ptrace.c b/arch/sh64/kernel/ptrace.c index 800288c1562..fd2000956da 100644 --- a/arch/sh64/kernel/ptrace.c +++ b/arch/sh64/kernel/ptrace.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include @@ -255,7 +256,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child, TIF_SYSCALL_TRACE); @@ -285,7 +286,7 @@ asmlinkage int sys_ptrace(long request, long pid, long addr, long data) struct pt_regs *regs; ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); if ((child->ptrace & PT_DTRACE) == 0) { diff --git a/arch/sparc/kernel/ptrace.c b/arch/sparc/kernel/ptrace.c index c4f93bd2daf..475c4c13462 100644 --- a/arch/sparc/kernel/ptrace.c +++ b/arch/sparc/kernel/ptrace.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -526,7 +527,7 @@ asmlinkage void do_ptrace(struct pt_regs *regs) addr = 1; case PTRACE_CONT: { /* restart after signal. */ - if (data > _NSIG) { + if (!valid_signal(data)) { pt_error_return(regs, EIO); goto out_tsk; } diff --git a/arch/sparc64/kernel/ptrace.c b/arch/sparc64/kernel/ptrace.c index 5f080cf04b3..80a76e2ad73 100644 --- a/arch/sparc64/kernel/ptrace.c +++ b/arch/sparc64/kernel/ptrace.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include @@ -559,7 +560,7 @@ asmlinkage void do_ptrace(struct pt_regs *regs) addr = 1; case PTRACE_CONT: { /* restart after signal. */ - if (data > _NSIG) { + if (!valid_signal(data)) { pt_error_return(regs, EIO); goto out_tsk; } diff --git a/arch/um/kernel/ptrace.c b/arch/um/kernel/ptrace.c index 3a99ee6d94e..e50e60ff5d2 100644 --- a/arch/um/kernel/ptrace.c +++ b/arch/um/kernel/ptrace.c @@ -143,7 +143,7 @@ long sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */ case PTRACE_CONT: { /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; child->ptrace &= ~PT_DTRACE; @@ -179,7 +179,7 @@ long sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SINGLESTEP: { /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE); child->ptrace |= PT_DTRACE; diff --git a/arch/v850/kernel/ptrace.c b/arch/v850/kernel/ptrace.c index 8fa780757dc..4726b87f5e5 100644 --- a/arch/v850/kernel/ptrace.c +++ b/arch/v850/kernel/ptrace.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include @@ -208,7 +209,7 @@ int sys_ptrace(long request, long pid, long addr, long data) /* Execute a single instruction. */ case PTRACE_SINGLESTEP: rval = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; /* Turn CHILD's single-step flag on or off. */ diff --git a/arch/x86_64/kernel/ptrace.c b/arch/x86_64/kernel/ptrace.c index c7011675007..c64b9c97c74 100644 --- a/arch/x86_64/kernel/ptrace.c +++ b/arch/x86_64/kernel/ptrace.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -467,7 +468,7 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data case PTRACE_CONT: /* restart after signal. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; if (request == PTRACE_SYSCALL) set_tsk_thread_flag(child,TIF_SYSCALL_TRACE); @@ -529,7 +530,7 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data case PTRACE_SINGLESTEP: /* set the trap flag. */ ret = -EIO; - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) break; clear_tsk_thread_flag(child,TIF_SYSCALL_TRACE); set_singlestep(child); diff --git a/drivers/char/vt_ioctl.c b/drivers/char/vt_ioctl.c index 5d386f4bea4..8971484b956 100644 --- a/drivers/char/vt_ioctl.c +++ b/drivers/char/vt_ioctl.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include @@ -641,7 +642,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file, extern int spawnpid, spawnsig; if (!perm || !capable(CAP_KILL)) return -EPERM; - if (arg < 1 || arg > _NSIG || arg == SIGKILL) + if (!valid_signal(arg) || arg < 1 || arg == SIGKILL) return -EINVAL; spawnpid = current->pid; spawnsig = arg; diff --git a/fs/fcntl.c b/fs/fcntl.c index 3e7ab16ed15..286a9f8f3d4 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -308,7 +309,7 @@ static long do_fcntl(int fd, unsigned int cmd, unsigned long arg, break; case F_SETSIG: /* arg == 0 restores default behaviour. */ - if (arg < 0 || arg > _NSIG) { + if (!valid_signal(arg)) { break; } err = 0; diff --git a/ipc/mqueue.c b/ipc/mqueue.c index 33f71520b89..0acf245f441 100644 --- a/ipc/mqueue.c +++ b/ipc/mqueue.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include "util.h" @@ -976,8 +977,7 @@ asmlinkage long sys_mq_notify(mqd_t mqdes, notification.sigev_notify != SIGEV_THREAD)) return -EINVAL; if (notification.sigev_notify == SIGEV_SIGNAL && - (notification.sigev_signo < 0 || - notification.sigev_signo > _NSIG)) { + !valid_signal(notification.sigev_signo)) { return -EINVAL; } if (notification.sigev_notify == SIGEV_THREAD) { diff --git a/kernel/exit.c b/kernel/exit.c index 93851bcd958..eb8da36e13d 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include @@ -277,7 +278,7 @@ void set_special_pids(pid_t session, pid_t pgrp) */ int allow_signal(int sig) { - if (sig < 1 || sig > _NSIG) + if (!valid_signal(sig) || sig < 1) return -EINVAL; spin_lock_irq(¤t->sighand->siglock); @@ -298,7 +299,7 @@ EXPORT_SYMBOL(allow_signal); int disallow_signal(int sig) { - if (sig < 1 || sig > _NSIG) + if (!valid_signal(sig) || sig < 1) return -EINVAL; spin_lock_irq(¤t->sighand->siglock); diff --git a/kernel/futex.c b/kernel/futex.c index 7b54a672d0a..c7130f86106 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -39,6 +39,7 @@ #include #include #include +#include #define FUTEX_HASHBITS (CONFIG_BASE_SMALL ? 4 : 8) @@ -654,7 +655,7 @@ static int futex_fd(unsigned long uaddr, int signal) int ret, err; ret = -EINVAL; - if (signal < 0 || signal > _NSIG) + if (!valid_signal(signal)) goto out; ret = get_unused_fd(); diff --git a/kernel/ptrace.c b/kernel/ptrace.c index f5cc1cec0fb..8dcb8f6288b 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include @@ -166,7 +167,7 @@ bad: int ptrace_detach(struct task_struct *child, unsigned int data) { - if ((unsigned long) data > _NSIG) + if (!valid_signal(data)) return -EIO; /* Architecture-specific hardware disable .. */ diff --git a/kernel/signal.c b/kernel/signal.c index e6567d7f2b6..8f3debc77c5 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include #include @@ -646,7 +647,7 @@ static int check_kill_permission(int sig, struct siginfo *info, struct task_struct *t) { int error = -EINVAL; - if (sig < 0 || sig > _NSIG) + if (!valid_signal(sig)) return error; error = -EPERM; if ((!info || ((unsigned long)info != 1 && @@ -1245,7 +1246,7 @@ send_sig_info(int sig, struct siginfo *info, struct task_struct *p) * Make sure legacy kernel users don't send in bad values * (normal paths check this in check_kill_permission). */ - if (sig < 0 || sig > _NSIG) + if (!valid_signal(sig)) return -EINVAL; /* @@ -1520,7 +1521,7 @@ void do_notify_parent(struct task_struct *tsk, int sig) if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN) sig = 0; } - if (sig > 0 && sig <= _NSIG) + if (valid_signal(sig) && sig > 0) __group_send_sig_info(sig, &info, tsk->parent); __wake_up_parent(tsk, tsk->parent); spin_unlock_irqrestore(&psig->siglock, flags); @@ -2364,7 +2365,7 @@ do_sigaction(int sig, const struct k_sigaction *act, struct k_sigaction *oact) { struct k_sigaction *k; - if (sig < 1 || sig > _NSIG || (act && sig_kernel_only(sig))) + if (!valid_signal(sig) || sig < 1 || (act && sig_kernel_only(sig))) return -EINVAL; k = ¤t->sighand->action[sig-1]; diff --git a/kernel/sys.c b/kernel/sys.c index 7f43d6e62c7..f64e97cabe2 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include @@ -1637,7 +1638,7 @@ asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3, switch (option) { case PR_SET_PDEATHSIG: sig = arg2; - if (sig < 0 || sig > _NSIG) { + if (!valid_signal(sig)) { error = -EINVAL; break; } -- cgit v1.2.3-70-g09d2 From 5f76be80d96f60adfc91f2acf22b146ce0e3072f Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Sun, 1 May 2005 08:59:23 -0700 Subject: [PATCH] fbdev: edid.h cleanups This patch removes some completely unused code. Signed-off-by: Adrian Bunk Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/boot/compressed/misc.c | 1 - drivers/video/fbmon.c | 1 - drivers/video/vesafb.c | 3 --- include/video/edid.h | 9 --------- 4 files changed, 14 deletions(-) (limited to 'arch') diff --git a/arch/i386/boot/compressed/misc.c b/arch/i386/boot/compressed/misc.c index fa67045234a..cedc55cc47d 100644 --- a/arch/i386/boot/compressed/misc.c +++ b/arch/i386/boot/compressed/misc.c @@ -12,7 +12,6 @@ #include #include #include -#include