diff options
Diffstat (limited to 'arch/ia64')
66 files changed, 1609 insertions, 2143 deletions
diff --git a/arch/ia64/Makefile b/arch/ia64/Makefile index 67932ad5308..f722e1a2594 100644 --- a/arch/ia64/Makefile +++ b/arch/ia64/Makefile @@ -25,7 +25,6 @@ cflags-y := -pipe $(EXTRA) -ffixed-r13 -mfixed-range=f12-f15,f32-f127 \ -falign-functions=32 -frename-registers -fno-optimize-sibling-calls CFLAGS_KERNEL := -mconstant-gp -GCC_VERSION := $(call cc-version) GAS_STATUS = $(shell $(srctree)/arch/ia64/scripts/check-gas "$(CC)" "$(OBJDUMP)") CPPFLAGS += $(shell $(srctree)/arch/ia64/scripts/toolchain-flags "$(CC)" "$(OBJDUMP)" "$(READELF)") @@ -37,11 +36,7 @@ $(error Sorry, you need a newer version of the assember, one that is built from ftp://ftp.hpl.hp.com/pub/linux-ia64/gas-030124.tar.gz) endif -ifneq ($(shell if [ $(GCC_VERSION) -lt 0300 ] ; then echo "bad"; fi ;),) -$(error Sorry, your compiler is too old. GCC v2.96 is known to generate bad code.) -endif - -ifeq ($(GCC_VERSION),0304) +ifeq ($(call cc-version),0304) cflags-$(CONFIG_ITANIUM) += -mtune=merced cflags-$(CONFIG_MCKINLEY) += -mtune=mckinley endif diff --git a/arch/ia64/configs/gensparse_defconfig b/arch/ia64/configs/gensparse_defconfig index 80f8663bc6d..991c07b57c2 100644 --- a/arch/ia64/configs/gensparse_defconfig +++ b/arch/ia64/configs/gensparse_defconfig @@ -557,6 +557,7 @@ CONFIG_E100=m # CONFIG_DL2K is not set CONFIG_E1000=y # CONFIG_E1000_NAPI is not set +# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set # CONFIG_NS83820 is not set # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set @@ -701,6 +702,7 @@ CONFIG_SERIAL_CORE_CONSOLE=y CONFIG_SERIAL_SGI_L1_CONSOLE=y # CONFIG_SERIAL_JSM is not set CONFIG_SERIAL_SGI_IOC4=y +CONFIG_SERIAL_SGI_IOC3=y CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -1046,6 +1048,7 @@ CONFIG_INFINIBAND_IPOIB=m # SN Devices # CONFIG_SGI_IOC4=y +CONFIG_SGI_IOC3=y # # File systems diff --git a/arch/ia64/configs/sn2_defconfig b/arch/ia64/configs/sn2_defconfig index e1924cc9687..3cb503b659e 100644 --- a/arch/ia64/configs/sn2_defconfig +++ b/arch/ia64/configs/sn2_defconfig @@ -113,7 +113,7 @@ CONFIG_IOSAPIC=y CONFIG_IA64_SGI_SN_XP=m CONFIG_FORCE_MAX_ZONEORDER=17 CONFIG_SMP=y -CONFIG_NR_CPUS=512 +CONFIG_NR_CPUS=1024 # CONFIG_HOTPLUG_CPU is not set CONFIG_SCHED_SMT=y CONFIG_PREEMPT=y @@ -659,6 +659,7 @@ CONFIG_SERIAL_CORE_CONSOLE=y CONFIG_SERIAL_SGI_L1_CONSOLE=y # CONFIG_SERIAL_JSM is not set CONFIG_SERIAL_SGI_IOC4=y +CONFIG_SERIAL_SGI_IOC3=y CONFIG_UNIX98_PTYS=y CONFIG_LEGACY_PTYS=y CONFIG_LEGACY_PTY_COUNT=256 @@ -899,6 +900,7 @@ CONFIG_INFINIBAND_SRP=m # SN Devices # CONFIG_SGI_IOC4=y +CONFIG_SGI_IOC3=y # # File systems diff --git a/arch/ia64/configs/tiger_defconfig b/arch/ia64/configs/tiger_defconfig index b1e8f09e9fd..6859119bc9d 100644 --- a/arch/ia64/configs/tiger_defconfig +++ b/arch/ia64/configs/tiger_defconfig @@ -565,6 +565,7 @@ CONFIG_E100=m # CONFIG_DL2K is not set CONFIG_E1000=y # CONFIG_E1000_NAPI is not set +# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set # CONFIG_NS83820 is not set # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set diff --git a/arch/ia64/configs/zx1_defconfig b/arch/ia64/configs/zx1_defconfig index 0856ca67dd5..53899dc8eb5 100644 --- a/arch/ia64/configs/zx1_defconfig +++ b/arch/ia64/configs/zx1_defconfig @@ -548,6 +548,7 @@ CONFIG_E100=y # CONFIG_DL2K is not set CONFIG_E1000=y # CONFIG_E1000_NAPI is not set +# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set # CONFIG_NS83820 is not set # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set diff --git a/arch/ia64/defconfig b/arch/ia64/defconfig index 275a26c6e5a..dcbc78a4cfa 100644 --- a/arch/ia64/defconfig +++ b/arch/ia64/defconfig @@ -565,6 +565,7 @@ CONFIG_E100=m # CONFIG_DL2K is not set CONFIG_E1000=y # CONFIG_E1000_NAPI is not set +# CONFIG_E1000_DISABLE_PACKET_SPLIT is not set # CONFIG_NS83820 is not set # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set diff --git a/arch/ia64/hp/sim/simserial.c b/arch/ia64/hp/sim/simserial.c index 19ee635eeb7..626cdc83668 100644 --- a/arch/ia64/hp/sim/simserial.c +++ b/arch/ia64/hp/sim/simserial.c @@ -26,6 +26,7 @@ #include <linux/fcntl.h> #include <linux/mm.h> #include <linux/slab.h> +#include <linux/capability.h> #include <linux/console.h> #include <linux/module.h> #include <linux/serial.h> @@ -107,7 +108,6 @@ static struct async_struct *IRQ_ports[NR_IRQS]; static struct console *console; static unsigned char *tmp_buf; -static DECLARE_MUTEX(tmp_buf_sem); extern struct console *console_drivers; /* from kernel/printk.c */ @@ -166,15 +166,9 @@ static void receive_chars(struct tty_struct *tty, struct pt_regs *regs) } } seen_esc = 0; - if (tty->flip.count >= TTY_FLIPBUF_SIZE) break; - *tty->flip.char_buf_ptr = ch; - - *tty->flip.flag_buf_ptr = 0; - - tty->flip.flag_buf_ptr++; - tty->flip.char_buf_ptr++; - tty->flip.count++; + if (tty_insert_flip_char(tty, ch, TTY_NORMAL) == 0) + break; } tty_flip_buffer_push(tty); } diff --git a/arch/ia64/ia32/Makefile b/arch/ia64/ia32/Makefile index 2ed90da8116..61cb60affd9 100644 --- a/arch/ia64/ia32/Makefile +++ b/arch/ia64/ia32/Makefile @@ -2,11 +2,9 @@ # Makefile for the ia32 kernel emulation subsystem. # -obj-y := ia32_entry.o sys_ia32.o ia32_ioctl.o ia32_signal.o \ +obj-y := ia32_entry.o sys_ia32.o ia32_signal.o \ ia32_support.o ia32_traps.o binfmt_elf32.o ia32_ldt.o -CFLAGS_ia32_ioctl.o += -Ifs/ - # Don't let GCC uses f16-f31 so that save_ia32_fpstate_live() and # restore_ia32_fpstate_live() can be sure the live register contain user-level state. CFLAGS_ia32_signal.o += -mfixed-range=f16-f31 diff --git a/arch/ia64/ia32/elfcore32.h b/arch/ia64/ia32/elfcore32.h index b73b8b6b10c..a47f63b204f 100644 --- a/arch/ia64/ia32/elfcore32.h +++ b/arch/ia64/ia32/elfcore32.h @@ -95,8 +95,7 @@ static inline void elf_core_copy_regs(elf_gregset_t *elfregs, static inline int elf_core_copy_task_regs(struct task_struct *t, elf_gregset_t* elfregs) { - struct pt_regs *pp = ia64_task_regs(t); - ELF_CORE_COPY_REGS((*elfregs), pp); + ELF_CORE_COPY_REGS((*elfregs), task_pt_regs(t)); return 1; } diff --git a/arch/ia64/ia32/ia32_entry.S b/arch/ia64/ia32/ia32_entry.S index 494fad6bf37..95fe04400f6 100644 --- a/arch/ia64/ia32/ia32_entry.S +++ b/arch/ia64/ia32/ia32_entry.S @@ -469,7 +469,7 @@ ia32_syscall_table: data8 sys32_epoll_wait data8 sys_remap_file_pages data8 sys_set_tid_address - data8 sys32_timer_create + data8 compat_sys_timer_create data8 compat_sys_timer_settime /* 260 */ data8 compat_sys_timer_gettime data8 sys_timer_getoverrun diff --git a/arch/ia64/ia32/ia32_ioctl.c b/arch/ia64/ia32/ia32_ioctl.c deleted file mode 100644 index 88739394f6d..00000000000 --- a/arch/ia64/ia32/ia32_ioctl.c +++ /dev/null @@ -1,45 +0,0 @@ -/* - * IA32 Architecture-specific ioctl shim code - * - * Copyright (C) 2000 VA Linux Co - * Copyright (C) 2000 Don Dugger <n0ano@valinux.com> - * Copyright (C) 2001-2003 Hewlett-Packard Co - * David Mosberger-Tang <davidm@hpl.hp.com> - */ - -#include <linux/signal.h> /* argh, msdos_fs.h isn't self-contained... */ -#include <linux/syscalls.h> -#include "ia32priv.h" - -#define INCLUDES -#include "compat_ioctl.c" - -#define IOCTL_NR(a) ((a) & ~(_IOC_SIZEMASK << _IOC_SIZESHIFT)) - -#define DO_IOCTL(fd, cmd, arg) ({ \ - int _ret; \ - mm_segment_t _old_fs = get_fs(); \ - \ - set_fs(KERNEL_DS); \ - _ret = sys_ioctl(fd, cmd, (unsigned long)arg); \ - set_fs(_old_fs); \ - _ret; \ -}) - -#define CODE -#include "compat_ioctl.c" - -#define COMPATIBLE_IOCTL(cmd) HANDLE_IOCTL((cmd),sys_ioctl) -#define HANDLE_IOCTL(cmd,handler) { (cmd), (ioctl_trans_handler_t)(handler), NULL }, -#define IOCTL_TABLE_START \ - struct ioctl_trans ioctl_start[] = { -#define IOCTL_TABLE_END \ - }; - -IOCTL_TABLE_START -#define DECLARES -#include "compat_ioctl.c" -#include <linux/compat_ioctl.h> -IOCTL_TABLE_END - -int ioctl_table_size = ARRAY_SIZE(ioctl_start); diff --git a/arch/ia64/ia32/ia32_signal.c b/arch/ia64/ia32/ia32_signal.c index aa891c9bc9b..5856510210f 100644 --- a/arch/ia64/ia32/ia32_signal.c +++ b/arch/ia64/ia32/ia32_signal.c @@ -255,7 +255,7 @@ save_ia32_fpstate_live (struct _fpstate_ia32 __user *save) */ fp_tos = (fsr>>11)&0x7; fr8_st_map = (8-fp_tos)&0x7; - ptp = ia64_task_regs(tsk); + ptp = task_pt_regs(tsk); fpregp = (struct _fpreg_ia32 *)(((unsigned long)buf + 15) & ~15); ia64f2ia32f(fpregp, &ptp->f8); copy_to_user(&save->_st[(0+fr8_st_map)&0x7], fpregp, sizeof(struct _fpreg_ia32)); @@ -389,7 +389,7 @@ restore_ia32_fpstate_live (struct _fpstate_ia32 __user *save) fr8_st_map = (8-fp_tos)&0x7; fpregp = (struct _fpreg_ia32 *)(((unsigned long)buf + 15) & ~15); - ptp = ia64_task_regs(tsk); + ptp = task_pt_regs(tsk); copy_from_user(fpregp, &save->_st[(0+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); ia32f2ia64f(&ptp->f8, fpregp); copy_from_user(fpregp, &save->_st[(1+fr8_st_map)&0x7], sizeof(struct _fpreg_ia32)); diff --git a/arch/ia64/ia32/ia32_support.c b/arch/ia64/ia32/ia32_support.c index 4f630043b3a..c187743965a 100644 --- a/arch/ia64/ia32/ia32_support.c +++ b/arch/ia64/ia32/ia32_support.c @@ -58,7 +58,7 @@ load_desc (u16 selector) void ia32_load_segment_descriptors (struct task_struct *task) { - struct pt_regs *regs = ia64_task_regs(task); + struct pt_regs *regs = task_pt_regs(task); /* Setup the segment descriptors */ regs->r24 = load_desc(regs->r16 >> 16); /* ESD */ @@ -113,7 +113,7 @@ void ia32_load_state (struct task_struct *t) { unsigned long eflag, fsr, fcr, fir, fdr, tssd; - struct pt_regs *regs = ia64_task_regs(t); + struct pt_regs *regs = task_pt_regs(t); eflag = t->thread.eflag; fsr = t->thread.fsr; diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c index dc282710421..70dba1f0e2e 100644 --- a/arch/ia64/ia32/sys_ia32.c +++ b/arch/ia64/ia32/sys_ia32.c @@ -48,12 +48,13 @@ #include <linux/ptrace.h> #include <linux/stat.h> #include <linux/ipc.h> +#include <linux/capability.h> #include <linux/compat.h> #include <linux/vfs.h> #include <linux/mman.h> +#include <linux/mutex.h> #include <asm/intrinsics.h> -#include <asm/semaphore.h> #include <asm/types.h> #include <asm/uaccess.h> #include <asm/unistd.h> @@ -85,7 +86,7 @@ * while doing so. */ /* XXX make per-mm: */ -static DECLARE_MUTEX(ia32_mmap_sem); +static DEFINE_MUTEX(ia32_mmap_mutex); asmlinkage long sys32_execve (char __user *name, compat_uptr_t __user *argv, compat_uptr_t __user *envp, @@ -894,11 +895,11 @@ ia32_do_mmap (struct file *file, unsigned long addr, unsigned long len, int prot prot = get_prot32(prot); #if PAGE_SHIFT > IA32_PAGE_SHIFT - down(&ia32_mmap_sem); + mutex_lock(&ia32_mmap_mutex); { addr = emulate_mmap(file, addr, len, prot, flags, offset); } - up(&ia32_mmap_sem); + mutex_unlock(&ia32_mmap_mutex); #else down_write(¤t->mm->mmap_sem); { @@ -999,11 +1000,9 @@ sys32_munmap (unsigned int start, unsigned int len) if (start >= end) return 0; - down(&ia32_mmap_sem); - { - ret = sys_munmap(start, end - start); - } - up(&ia32_mmap_sem); + mutex_lock(&ia32_mmap_mutex); + ret = sys_munmap(start, end - start); + mutex_unlock(&ia32_mmap_mutex); #endif return ret; } @@ -1055,7 +1054,7 @@ sys32_mprotect (unsigned int start, unsigned int len, int prot) if (retval < 0) return retval; - down(&ia32_mmap_sem); + mutex_lock(&ia32_mmap_mutex); { if (offset_in_page(start)) { /* start address is 4KB aligned but not page aligned. */ @@ -1079,7 +1078,7 @@ sys32_mprotect (unsigned int start, unsigned int len, int prot) retval = sys_mprotect(start, end - start, prot); } out: - up(&ia32_mmap_sem); + mutex_unlock(&ia32_mmap_mutex); return retval; #endif } @@ -1123,11 +1122,9 @@ sys32_mremap (unsigned int addr, unsigned int old_len, unsigned int new_len, old_len = PAGE_ALIGN(old_end) - addr; new_len = PAGE_ALIGN(new_end) - addr; - down(&ia32_mmap_sem); - { - ret = sys_mremap(addr, old_len, new_len, flags, new_addr); - } - up(&ia32_mmap_sem); + mutex_lock(&ia32_mmap_mutex); + ret = sys_mremap(addr, old_len, new_len, flags, new_addr); + mutex_unlock(&ia32_mmap_mutex); if ((ret >= 0) && (old_len < new_len)) { /* mremap expanded successfully */ @@ -1481,7 +1478,7 @@ getreg (struct task_struct *child, int regno) { struct pt_regs *child_regs; - child_regs = ia64_task_regs(child); + child_regs = task_pt_regs(child); switch (regno / sizeof(int)) { case PT_EBX: return child_regs->r11; case PT_ECX: return child_regs->r9; @@ -1509,7 +1506,7 @@ putreg (struct task_struct *child, int regno, unsigned int value) { struct pt_regs *child_regs; - child_regs = ia64_task_regs(child); + child_regs = task_pt_regs(child); switch (regno / sizeof(int)) { case PT_EBX: child_regs->r11 = value; break; case PT_ECX: child_regs->r9 = value; break; @@ -1625,7 +1622,7 @@ save_ia32_fpstate (struct task_struct *tsk, struct ia32_user_i387_struct __user * Stack frames start with 16-bytes of temp space */ swp = (struct switch_stack *)(tsk->thread.ksp + 16); - ptp = ia64_task_regs(tsk); + ptp = task_pt_regs(tsk); tos = (tsk->thread.fsr >> 11) & 7; for (i = 0; i < 8; i++) put_fpreg(i, &save->st_space[i], ptp, swp, tos); @@ -1658,7 +1655,7 @@ restore_ia32_fpstate (struct task_struct *tsk, struct ia32_user_i387_struct __us * Stack frames start with 16-bytes of temp space */ swp = (struct switch_stack *)(tsk->thread.ksp + 16); - ptp = ia64_task_regs(tsk); + ptp = task_pt_regs(tsk); tos = (tsk->thread.fsr >> 11) & 7; for (i = 0; i < 8; i++) get_fpreg(i, &save->st_space[i], ptp, swp, tos); @@ -1689,7 +1686,7 @@ save_ia32_fpxstate (struct task_struct *tsk, struct ia32_user_fxsr_struct __user * Stack frames start with 16-bytes of temp space */ swp = (struct switch_stack *)(tsk->thread.ksp + 16); - ptp = ia64_task_regs(tsk); + ptp = task_pt_regs(tsk); tos = (tsk->thread.fsr >> 11) & 7; for (i = 0; i < 8; i++) put_fpreg(i, (struct _fpreg_ia32 __user *)&save->st_space[4*i], ptp, swp, tos); @@ -1733,7 +1730,7 @@ restore_ia32_fpxstate (struct task_struct *tsk, struct ia32_user_fxsr_struct __u * Stack frames start with 16-bytes of temp space */ swp = (struct switch_stack *)(tsk->thread.ksp + 16); - ptp = ia64_task_regs(tsk); + ptp = task_pt_regs(tsk); tos = (tsk->thread.fsr >> 11) & 7; for (i = 0; i < 8; i++) get_fpreg(i, (struct _fpreg_ia32 __user *)&save->st_space[4*i], ptp, swp, tos); @@ -1761,21 +1758,15 @@ sys32_ptrace (int request, pid_t pid, unsigned int addr, unsigned int data) lock_kernel(); if (request == PTRACE_TRACEME) { - ret = sys_ptrace(request, pid, addr, data); + ret = ptrace_traceme(); goto out; } - ret = -ESRCH; - read_lock(&tasklist_lock); - child = find_task_by_pid(pid); - if (child) - get_task_struct(child); - read_unlock(&tasklist_lock); - if (!child) + child = ptrace_get_task_struct(pid); + if (IS_ERR(child)) { + ret = PTR_ERR(child); goto out; - ret = -EPERM; - if (pid == 1) /* no messing around with init! */ - goto out_tsk; + } if (request == PTRACE_ATTACH) { ret = sys_ptrace(request, pid, addr, data); @@ -2559,34 +2550,6 @@ sys32_get_thread_area (struct ia32_user_desc __user *u_info) return 0; } -asmlinkage long -sys32_timer_create(u32 clock, struct compat_sigevent __user *se32, timer_t __user *timer_id) -{ - struct sigevent se; - mm_segment_t oldfs; - timer_t t; - long err; - - if (se32 == NULL) - return sys_timer_create(clock, NULL, timer_id); - - if (get_compat_sigevent(&se, se32)) - return -EFAULT; - - if (!access_ok(VERIFY_WRITE,timer_id,sizeof(timer_t))) - return -EFAULT; - - oldfs = get_fs(); - set_fs(KERNEL_DS); - err = sys_timer_create(clock, (struct sigevent __user *) &se, (timer_t __user *) &t); - set_fs(oldfs); - - if (!err) - err = __put_user (t, timer_id); - - return err; -} - long sys32_fadvise64_64(int fd, __u32 offset_low, __u32 offset_high, __u32 len_low, __u32 len_high, int advice) { diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile index 307514f7a28..09a0dbc17fb 100644 --- a/arch/ia64/kernel/Makefile +++ b/arch/ia64/kernel/Makefile @@ -13,6 +13,11 @@ obj-$(CONFIG_IA64_BRL_EMU) += brl_emu.o obj-$(CONFIG_IA64_GENERIC) += acpi-ext.o obj-$(CONFIG_IA64_HP_ZX1) += acpi-ext.o obj-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += acpi-ext.o + +ifneq ($(CONFIG_ACPI_PROCESSOR),) +obj-y += acpi-processor.o +endif + obj-$(CONFIG_IA64_PALINFO) += palinfo.o obj-$(CONFIG_IOSAPIC) += iosapic.o obj-$(CONFIG_MODULES) += module.o diff --git a/arch/ia64/kernel/acpi-ext.c b/arch/ia64/kernel/acpi-ext.c index 13a5b3b49bf..4a5574ff007 100644 --- a/arch/ia64/kernel/acpi-ext.c +++ b/arch/ia64/kernel/acpi-ext.c @@ -33,33 +33,33 @@ acpi_vendor_resource_match(struct acpi_resource *resource, void *context) struct acpi_vendor_info *info = (struct acpi_vendor_info *)context; struct acpi_resource_vendor *vendor; struct acpi_vendor_descriptor *descriptor; - u32 length; + u32 byte_length; - if (resource->id != ACPI_RSTYPE_VENDOR) + if (resource->type != ACPI_RESOURCE_TYPE_VENDOR) return AE_OK; vendor = (struct acpi_resource_vendor *)&resource->data; - descriptor = (struct acpi_vendor_descriptor *)vendor->reserved; - if (vendor->length <= sizeof(*info->descriptor) || + descriptor = (struct acpi_vendor_descriptor *)vendor->byte_data; + if (vendor->byte_length <= sizeof(*info->descriptor) || descriptor->guid_id != info->descriptor->guid_id || efi_guidcmp(descriptor->guid, info->descriptor->guid)) return AE_OK; - length = vendor->length - sizeof(struct acpi_vendor_descriptor); - info->data = acpi_os_allocate(length); + byte_length = vendor->byte_length - sizeof(struct acpi_vendor_descriptor); + info->data = acpi_os_allocate(byte_length); if (!info->data) return AE_NO_MEMORY; memcpy(info->data, - vendor->reserved + sizeof(struct acpi_vendor_descriptor), - length); - info->length = length; + vendor->byte_data + sizeof(struct acpi_vendor_descriptor), + byte_length); + info->length = byte_length; return AE_CTRL_TERMINATE; } acpi_status acpi_find_vendor_resource(acpi_handle obj, struct acpi_vendor_descriptor * id, - u8 ** data, u32 * length) + u8 ** data, u32 * byte_length) { struct acpi_vendor_info info; @@ -72,7 +72,7 @@ acpi_find_vendor_resource(acpi_handle obj, struct acpi_vendor_descriptor * id, return AE_NOT_FOUND; *data = info.data; - *length = info.length; + *byte_length = info.length; return AE_OK; } diff --git a/arch/ia64/kernel/acpi-processor.c b/arch/ia64/kernel/acpi-processor.c new file mode 100644 index 00000000000..e683630c8ce --- /dev/null +++ b/arch/ia64/kernel/acpi-processor.c @@ -0,0 +1,67 @@ +/* + * arch/ia64/kernel/cpufreq/processor.c + * + * Copyright (C) 2005 Intel Corporation + * Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> + * - Added _PDC for platforms with Intel CPUs + */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/acpi.h> + +#include <acpi/processor.h> +#include <asm/acpi.h> + +static void init_intel_pdc(struct acpi_processor *pr) +{ + struct acpi_object_list *obj_list; + union acpi_object *obj; + u32 *buf; + + /* allocate and initialize pdc. It will be used later. */ + obj_list = kmalloc(sizeof(struct acpi_object_list), GFP_KERNEL); + if (!obj_list) { + printk(KERN_ERR "Memory allocation error\n"); + return; + } + + obj = kmalloc(sizeof(union acpi_object), GFP_KERNEL); + if (!obj) { + printk(KERN_ERR "Memory allocation error\n"); + kfree(obj_list); + return; + } + + buf = kmalloc(12, GFP_KERNEL); + if (!buf) { + printk(KERN_ERR "Memory allocation error\n"); + kfree(obj); + kfree(obj_list); + return; + } + + buf[0] = ACPI_PDC_REVISION_ID; + buf[1] = 1; + buf[2] |= ACPI_PDC_EST_CAPABILITY_SMP; + + obj->type = ACPI_TYPE_BUFFER; + obj->buffer.length = 12; + obj->buffer.pointer = (u8 *) buf; + obj_list->count = 1; + obj_list->pointer = obj; + pr->pdc = obj_list; + + return; +} + +/* Initialize _PDC data based on the CPU vendor */ +void arch_acpi_processor_init_pdc(struct acpi_processor *pr) +{ + pr->pdc = NULL; + init_intel_pdc(pr); + return; +} + +EXPORT_SYMBOL(arch_acpi_processor_init_pdc); diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c index 9ad94ddf668..d2702c419cf 100644 --- a/arch/ia64/kernel/acpi.c +++ b/arch/ia64/kernel/acpi.c @@ -567,16 +567,16 @@ void __init acpi_numa_arch_fixup(void) * success: return IRQ number (>=0) * failure: return < 0 */ -int acpi_register_gsi(u32 gsi, int edge_level, int active_high_low) +int acpi_register_gsi(u32 gsi, int triggering, int polarity) { if (has_8259 && gsi < 16) return isa_irq_to_vector(gsi); return iosapic_register_intr(gsi, - (active_high_low == + (polarity == ACPI_ACTIVE_HIGH) ? IOSAPIC_POL_HIGH : IOSAPIC_POL_LOW, - (edge_level == + (triggering == ACPI_EDGE_SENSITIVE) ? IOSAPIC_EDGE : IOSAPIC_LEVEL); } diff --git a/arch/ia64/kernel/cpufreq/Makefile b/arch/ia64/kernel/cpufreq/Makefile index f748d34c02f..4838f2a57c7 100644 --- a/arch/ia64/kernel/cpufreq/Makefile +++ b/arch/ia64/kernel/cpufreq/Makefile @@ -1 +1,2 @@ obj-$(CONFIG_IA64_ACPI_CPUFREQ) += acpi-cpufreq.o + diff --git a/arch/ia64/kernel/cpufreq/acpi-cpufreq.c b/arch/ia64/kernel/cpufreq/acpi-cpufreq.c index da4d5cf80a4..5a1bf815282 100644 --- a/arch/ia64/kernel/cpufreq/acpi-cpufreq.c +++ b/arch/ia64/kernel/cpufreq/acpi-cpufreq.c @@ -269,48 +269,6 @@ acpi_cpufreq_verify ( } -/* - * processor_init_pdc - let BIOS know about the SMP capabilities - * of this driver - * @perf: processor-specific acpi_io_data struct - * @cpu: CPU being initialized - * - * To avoid issues with legacy OSes, some BIOSes require to be informed of - * the SMP capabilities of OS P-state driver. Here we set the bits in _PDC - * accordingly. Actual call to _PDC is done in driver/acpi/processor.c - */ -static void -processor_init_pdc ( - struct acpi_processor_performance *perf, - unsigned int cpu, - struct acpi_object_list *obj_list - ) -{ - union acpi_object *obj; - u32 *buf; - - dprintk("processor_init_pdc\n"); - - perf->pdc = NULL; - /* Initialize pdc. It will be used later. */ - if (!obj_list) - return; - - if (!(obj_list->count && obj_list->pointer)) - return; - - obj = obj_list->pointer; - if ((obj->buffer.length == 12) && obj->buffer.pointer) { - buf = (u32 *)obj->buffer.pointer; - buf[0] = ACPI_PDC_REVISION_ID; - buf[1] = 1; - buf[2] = ACPI_PDC_EST_CAPABILITY_SMP; - perf->pdc = obj_list; - } - return; -} - - static int acpi_cpufreq_cpu_init ( struct cpufreq_policy *policy) @@ -320,14 +278,7 @@ acpi_cpufreq_cpu_init ( struct cpufreq_acpi_io *data; unsigned int result = 0; - union acpi_object arg0 = {ACPI_TYPE_BUFFER}; - u32 arg0_buf[3]; - struct acpi_object_list arg_list = {1, &arg0}; - dprintk("acpi_cpufreq_cpu_init\n"); - /* setup arg_list for _PDC settings */ - arg0.buffer.length = 12; - arg0.buffer.pointer = (u8 *) arg0_buf; data = kmalloc(sizeof(struct cpufreq_acpi_io), GFP_KERNEL); if (!data) @@ -337,9 +288,7 @@ acpi_cpufreq_cpu_init ( acpi_io_data[cpu] = data; - processor_init_pdc(&data->acpi_data, cpu, &arg_list); result = acpi_processor_register_performance(&data->acpi_data, cpu); - data->acpi_data.pdc = NULL; if (result) goto err_free; diff --git a/arch/ia64/kernel/efi.c b/arch/ia64/kernel/efi.c index a3aa45cbcfa..c485a3b32ba 100644 --- a/arch/ia64/kernel/efi.c +++ b/arch/ia64/kernel/efi.c @@ -247,6 +247,32 @@ typedef struct kern_memdesc { static kern_memdesc_t *kern_memmap; +#define efi_md_size(md) (md->num_pages << EFI_PAGE_SHIFT) + +static inline u64 +kmd_end(kern_memdesc_t *kmd) +{ + return (kmd->start + (kmd->num_pages << EFI_PAGE_SHIFT)); +} + +static inline u64 +efi_md_end(efi_memory_desc_t *md) +{ + return (md->phys_addr + efi_md_size(md)); +} + +static inline int +efi_wb(efi_memory_desc_t *md) +{ + return (md->attribute & EFI_MEMORY_WB); +} + +static inline int +efi_uc(efi_memory_desc_t *md) +{ + return (md->attribute & EFI_MEMORY_UC); +} + static void walk (efi_freemem_callback_t callback, void *arg, u64 attr) { @@ -595,8 +621,8 @@ efi_get_iobase (void) return 0; } -u32 -efi_mem_type (unsigned long phys_addr) +static efi_memory_desc_t * +efi_memory_descriptor (unsigned long phys_addr) { void *efi_map_start, *efi_map_end, *p; efi_memory_desc_t *md; @@ -610,13 +636,13 @@ efi_mem_type (unsigned long phys_addr) md = p; if (phys_addr - md->phys_addr < (md->num_pages << EFI_PAGE_SHIFT)) - return md->type; + return md; } return 0; } -u64 -efi_mem_attributes (unsigned long phys_addr) +static int +efi_memmap_has_mmio (void) { void *efi_map_start, *efi_map_end, *p; efi_memory_desc_t *md; @@ -629,36 +655,98 @@ efi_mem_attributes (unsigned long phys_addr) for (p = efi_map_start; p < efi_map_end; p += efi_desc_size) { md = p; - if (phys_addr - md->phys_addr < (md->num_pages << EFI_PAGE_SHIFT)) - return md->attribute; + if (md->type == EFI_MEMORY_MAPPED_IO) + return 1; } return 0; } + +u32 +efi_mem_type (unsigned long phys_addr) +{ + efi_memory_desc_t *md = efi_memory_descriptor(phys_addr); + + if (md) + return md->type; + return 0; +} + +u64 +efi_mem_attributes (unsigned long phys_addr) +{ + efi_memory_desc_t *md = efi_memory_descriptor(phys_addr); + + if (md) + return md->attribute; + return 0; +} EXPORT_SYMBOL(efi_mem_attributes); +/* + * Determines whether the memory at phys_addr supports the desired + * attribute (WB, UC, etc). If this returns 1, the caller can safely + * access *size bytes at phys_addr with the specified attribute. + */ +static int +efi_mem_attribute_range (unsigned long phys_addr, unsigned long *size, u64 attr) +{ + efi_memory_desc_t *md = efi_memory_descriptor(phys_addr); + unsigned long md_end; + + if (!md || (md->attribute & attr) != attr) + return 0; + + do { + md_end = efi_md_end(md); + if (phys_addr + *size <= md_end) + return 1; + + md = efi_memory_descriptor(md_end); + if (!md || (md->attribute & attr) != attr) { + *size = md_end - phys_addr; + return 1; + } + } while (md); + return 0; +} + +/* + * For /dev/mem, we only allow read & write system calls to access + * write-back memory, because read & write don't allow the user to + * control access size. + */ int valid_phys_addr_range (unsigned long phys_addr, unsigned long *size) { - void *efi_map_start, *efi_map_end, *p; - efi_memory_desc_t *md; - u64 efi_desc_size; + return efi_mem_attribute_range(phys_addr, size, EFI_MEMORY_WB); +} - efi_map_start = __va(ia64_boot_param->efi_memmap); - efi_map_end = efi_map_start + ia64_boot_param->efi_memmap_size; - efi_desc_size = ia64_boot_param->efi_memdesc_size; +/* + * We allow mmap of anything in the EFI memory map that supports + * either write-back or uncacheable access. For uncacheable regions, + * the supported access sizes are system-dependent, and the user is + * responsible for using the correct size. + * + * Note that this doesn't currently allow access to hot-added memory, + * because that doesn't appear in the boot-time EFI memory map. + */ +int +valid_mmap_phys_addr_range (unsigned long phys_addr, unsigned long *size) +{ + if (efi_mem_attribute_range(phys_addr, size, EFI_MEMORY_WB)) + return 1; - for (p = efi_map_start; p < efi_map_end; p += efi_desc_size) { - md = p; + if (efi_mem_attribute_range(phys_addr, size, EFI_MEMORY_UC)) + return 1; - if (phys_addr - md->phys_addr < (md->num_pages << EFI_PAGE_SHIFT)) { - if (!(md->attribute & EFI_MEMORY_WB)) - return 0; + /* + * Some firmware doesn't report MMIO regions in the EFI memory map. + * The Intel BigSur (a.k.a. HP i2000) has this problem. In this + * case, we can't use the EFI memory map to validate mmap requests. + */ + if (!efi_memmap_has_mmio()) + return 1; - if (*size > md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT) - phys_addr) - *size = md->phys_addr + (md->num_pages << EFI_PAGE_SHIFT) - phys_addr; - return 1; - } - } return 0; } @@ -707,32 +795,6 @@ efi_uart_console_only(void) return 0; } -#define efi_md_size(md) (md->num_pages << EFI_PAGE_SHIFT) - -static inline u64 -kmd_end(kern_memdesc_t *kmd) -{ - return (kmd->start + (kmd->num_pages << EFI_PAGE_SHIFT)); -} - -static inline u64 -efi_md_end(efi_memory_desc_t *md) -{ - return (md->phys_addr + efi_md_size(md)); -} - -static inline int -efi_wb(efi_memory_desc_t *md) -{ - return (md->attribute & EFI_MEMORY_WB); -} - -static inline int -efi_uc(efi_memory_desc_t *md) -{ - return (md->attribute & EFI_MEMORY_UC); -} - /* * Look for the first granule aligned memory descriptor memory * that is big enough to hold EFI memory map. Make sure this diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S index 0741b066b98..7a6ffd61378 100644 --- a/arch/ia64/kernel/entry.S +++ b/arch/ia64/kernel/entry.S @@ -1600,5 +1600,6 @@ sys_call_table: data8 sys_inotify_init data8 sys_inotify_add_watch data8 sys_inotify_rm_watch + data8 sys_migrate_pages // 1280 .org sys_call_table + 8*NR_syscalls // guard against failures to increase NR_syscalls diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S index 2ddbac6f499..ce423910ca9 100644 --- a/arch/ia64/kernel/fsys.S +++ b/arch/ia64/kernel/fsys.S @@ -903,5 +903,6 @@ fsyscall_table: data8 0 data8 0 data8 0 + data8 0 // 1280 .org fsyscall_table + 8*NR_syscalls // guard against failures to increase NR_syscalls diff --git a/arch/ia64/kernel/head.S b/arch/ia64/kernel/head.S index bfe65b2e862..fbc7ea35dd5 100644 --- a/arch/ia64/kernel/head.S +++ b/arch/ia64/kernel/head.S @@ -1060,7 +1060,7 @@ SET_REG(b5); * the clobber lists for spin_lock() in include/asm-ia64/spinlock.h. */ -#if __GNUC__ < 3 || (__GNUC__ == 3 && __GNUC_MINOR__ < 3) +#if (__GNUC__ == 3 && __GNUC_MINOR__ < 3) GLOBAL_ENTRY(ia64_spinlock_contention_pre3_4) .prologue diff --git a/arch/ia64/kernel/ia64_ksyms.c b/arch/ia64/kernel/ia64_ksyms.c index 5db9d3bcbbc..e72de580ebb 100644 --- a/arch/ia64/kernel/ia64_ksyms.c +++ b/arch/ia64/kernel/ia64_ksyms.c @@ -103,7 +103,7 @@ EXPORT_SYMBOL(unw_init_running); #ifdef ASM_SUPPORTED # ifdef CONFIG_SMP -# if __GNUC__ < 3 || (__GNUC__ == 3 && __GNUC_MINOR__ < 3) +# if (__GNUC__ == 3 && __GNUC_MINOR__ < 3) /* * This is not a normal routine and we don't want a function descriptor for it, so we use * a fake declaration here. diff --git a/arch/ia64/kernel/jprobes.S b/arch/ia64/kernel/jprobes.S index 2323377e369..5cd6226f44f 100644 --- a/arch/ia64/kernel/jprobes.S +++ b/arch/ia64/kernel/jprobes.S @@ -60,3 +60,30 @@ END(jprobe_break) GLOBAL_ENTRY(jprobe_inst_return) br.call.sptk.many b0=jprobe_break END(jprobe_inst_return) + +GLOBAL_ENTRY(invalidate_stacked_regs) + movl r16=invalidate_restore_cfm + ;; + mov b6=r16 + ;; + br.ret.sptk.many b6 + ;; +invalidate_restore_cfm: + mov r16=ar.rsc + ;; + mov ar.rsc=r0 + ;; + loadrs + ;; + mov ar.rsc=r16 + ;; + br.cond.sptk.many rp +END(invalidate_stacked_regs) + +GLOBAL_ENTRY(flush_register_stack) + // flush dirty regs to backing store (must be first in insn group) + flushrs + ;; + br.ret.sptk.many rp +END(flush_register_stack) + diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c index 89a70400c4f..50ae8c7d453 100644 --- a/arch/ia64/kernel/kprobes.c +++ b/arch/ia64/kernel/kprobes.c @@ -467,10 +467,6 @@ void __kprobes arch_disarm_kprobe(struct kprobe *p) flush_icache_range(arm_addr, arm_addr + sizeof(bundle_t)); } -void __kprobes arch_remove_kprobe(struct kprobe *p) -{ -} - /* * We are resuming execution after a single step fault, so the pt_regs * structure reflects the register state after we executed the instruction @@ -642,6 +638,13 @@ static int __kprobes pre_kprobes_handler(struct die_args *args) if (p->break_handler && p->break_handler(p, regs)) { goto ss_probe; } + } else if (!is_ia64_break_inst(regs)) { + /* The breakpoint instruction was removed by + * another cpu right after we hit, no further + * handling of this interrupt is appropriate + */ + ret = 1; + goto no_kprobe; } else { /* Not our break */ goto no_kprobe; @@ -763,11 +766,56 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self, return ret; } +struct param_bsp_cfm { + unsigned long ip; + unsigned long *bsp; + unsigned long cfm; +}; + +static void ia64_get_bsp_cfm(struct unw_frame_info *info, void *arg) +{ + unsigned long ip; + struct param_bsp_cfm *lp = arg; + + do { + unw_get_ip(info, &ip); + if (ip == 0) + break; + if (ip == lp->ip) { + unw_get_bsp(info, (unsigned long*)&lp->bsp); + unw_get_cfm(info, (unsigned long*)&lp->cfm); + return; + } + } while (unw_unwind(info) >= 0); + lp->bsp = 0; + lp->cfm = 0; + return; +} + int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) { struct jprobe *jp = container_of(p, struct jprobe, kp); unsigned long addr = ((struct fnptr *)(jp->entry))->ip; struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + struct param_bsp_cfm pa; + int bytes; + + /* + * Callee owns the argument space and could overwrite it, eg + * tail call optimization. So to be absolutely safe + * we save the argument space before transfering the control + * to instrumented jprobe function which runs in + * the process context + */ + pa.ip = regs->cr_iip; + unw_init_running(ia64_get_bsp_cfm, &pa); + bytes = (char *)ia64_rse_skip_regs(pa.bsp, pa.cfm & 0x3f) + - (char *)pa.bsp; + memcpy( kcb->jprobes_saved_stacked_regs, + pa.bsp, + bytes ); + kcb->bsp = pa.bsp; + kcb->cfm = pa.cfm; /* save architectural state */ kcb->jprobe_saved_regs = *regs; @@ -789,8 +837,20 @@ int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) { struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + int bytes; + /* restoring architectural state */ *regs = kcb->jprobe_saved_regs; + + /* restoring the original argument space */ + flush_register_stack(); + bytes = (char *)ia64_rse_skip_regs(kcb->bsp, kcb->cfm & 0x3f) + - (char *)kcb->bsp; + memcpy( kcb->bsp, + kcb->jprobes_saved_stacked_regs, + bytes ); + invalidate_stacked_regs(); + preempt_enable_no_resched(); return 1; } diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c index 355af15287c..ee7eec9ee57 100644 --- a/arch/ia64/kernel/mca.c +++ b/arch/ia64/kernel/mca.c @@ -766,7 +766,7 @@ ia64_mca_modify_original_stack(struct pt_regs *regs, l = strlen(previous_current->comm); snprintf(comm, sizeof(comm), "%s %*s %d", current->comm, l, previous_current->comm, - previous_current->thread_info->cpu); + task_thread_info(previous_current)->cpu); } memcpy(current->comm, comm, sizeof(current->comm)); @@ -1423,7 +1423,7 @@ format_mca_init_stack(void *mca_data, unsigned long offset, struct task_struct *p = (struct task_struct *)((char *)mca_data + offset); struct thread_info *ti; memset(p, 0, KERNEL_STACK_SIZE); - ti = (struct thread_info *)((char *)p + IA64_TASK_SIZE); + ti = task_thread_info(p); ti->flags = _TIF_MCA_INIT; ti->preempt_count = 1; ti->task = p; diff --git a/arch/ia64/kernel/mca_asm.S b/arch/ia64/kernel/mca_asm.S index db32fc1d393..403a80a58c1 100644 --- a/arch/ia64/kernel/mca_asm.S +++ b/arch/ia64/kernel/mca_asm.S @@ -847,7 +847,7 @@ ia64_state_restore: ;; mov cr.iim=temp3 mov cr.iha=temp4 - dep r22=0,r22,62,2 // pal_min_state, physical, uncached + dep r22=0,r22,62,1 // pal_min_state, physical, uncached mov IA64_KR(CURRENT)=r21 ld8 r8=[temp1] // os_status ld8 r10=[temp2] // context diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c index 410d4804fa6..9c5194b385d 100644 --- a/arch/ia64/kernel/perfmon.c +++ b/arch/ia64/kernel/perfmon.c @@ -38,7 +38,9 @@ #include <linux/pagemap.h> #include <linux/mount.h> #include <linux/bitops.h> +#include <linux/capability.h> #include <linux/rcupdate.h> +#include <linux/completion.h> #include <asm/errno.h> #include <asm/intrinsics.h> @@ -285,7 +287,7 @@ typedef struct pfm_context { unsigned long ctx_ovfl_regs[4]; /* which registers overflowed (notification) */ - struct semaphore ctx_restart_sem; /* use for blocking notification mode */ + struct completion ctx_restart_done; /* use for blocking notification mode */ unsigned long ctx_used_pmds[4]; /* bitmask of PMD used */ unsigned long ctx_all_pmds[4]; /* bitmask of all accessible PMDs */ @@ -627,9 +629,11 @@ static int pfm_write_ibr_dbr(int mode, pfm_context_t *ctx, void *arg, int count, #include "perfmon_itanium.h" #include "perfmon_mckinley.h" +#include "perfmon_montecito.h" #include "perfmon_generic.h" static pmu_config_t *pmu_confs[]={ + &pmu_conf_mont, &pmu_conf_mck, &pmu_conf_ita, &pmu_conf_gen, /* must be last */ @@ -1709,7 +1713,7 @@ static void pfm_syswide_force_stop(void *info) { pfm_context_t *ctx = (pfm_context_t *)info; - struct pt_regs *regs = ia64_task_regs(current); + struct pt_regs *regs = task_pt_regs(current); struct task_struct *owner; unsigned long flags; int ret; @@ -1814,7 +1818,7 @@ pfm_flush(struct file *filp) is_system = ctx->ctx_fl_system; task = PFM_CTX_TASK(ctx); - regs = ia64_task_regs(task); + regs = task_pt_regs(task); DPRINT(("ctx_state=%d is_current=%d\n", state, @@ -1944,7 +1948,7 @@ pfm_close(struct inode *inode, struct file *filp) is_system = ctx->ctx_fl_system; task = PFM_CTX_TASK(ctx); - regs = ia64_task_regs(task); + regs = task_pt_regs(task); DPRINT(("ctx_state=%d is_current=%d\n", state, @@ -1988,7 +1992,7 @@ pfm_close(struct inode *inode, struct file *filp) /* * force task to wake up from MASKED state */ - up(&ctx->ctx_restart_sem); + complete(&ctx->ctx_restart_done); DPRINT(("waking up ctx_state=%d\n", state)); @@ -2703,7 +2707,7 @@ pfm_context_create(pfm_context_t *ctx, void *arg, int count, struct pt_regs *reg /* * init restart semaphore to locked */ - sema_init(&ctx->ctx_restart_sem, 0); + init_completion(&ctx->ctx_restart_done); /* * activation is used in SMP only @@ -3684,7 +3688,7 @@ pfm_restart(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs) */ if (CTX_OVFL_NOBLOCK(ctx) == 0 && state == PFM_CTX_MASKED) { DPRINT(("unblocking [%d] \n", task->pid)); - up(&ctx->ctx_restart_sem); + complete(&ctx->ctx_restart_done); } else { DPRINT(("[%d] armed exit trap\n", task->pid)); @@ -4051,7 +4055,7 @@ pfm_stop(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs) */ ia64_psr(regs)->up = 0; } else { - tregs = ia64_task_regs(task); + tregs = task_pt_regs(task); /* * stop monitoring at the user level @@ -4133,7 +4137,7 @@ pfm_start(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs) ia64_psr(regs)->up = 1; } else { - tregs = ia64_task_regs(ctx->ctx_task); + tregs = task_pt_regs(ctx->ctx_task); /* * start monitoring at the kernel level the next @@ -4403,7 +4407,7 @@ pfm_context_load(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs) /* * when not current, task MUST be stopped, so this is safe */ - regs = ia64_task_regs(task); + regs = task_pt_regs(task); /* force a full reload */ ctx->ctx_last_activation = PFM_INVALID_ACTIVATION; @@ -4529,7 +4533,7 @@ pfm_context_unload(pfm_context_t *ctx, void *arg, int count, struct pt_regs *reg /* * per-task mode */ - tregs = task == current ? regs : ia64_task_regs(task); + tregs = task == current ? regs : task_pt_regs(task); if (task == current) { /* @@ -4592,7 +4596,7 @@ pfm_exit_thread(struct task_struct *task) { pfm_context_t *ctx; unsigned long flags; - struct pt_regs *regs = ia64_task_regs(task); + struct pt_regs *regs = task_pt_regs(task); int ret, state; int free_ok = 0; @@ -4925,7 +4929,7 @@ restart_args: if (unlikely(ret)) goto abort_locked; skip_fd: - ret = (*func)(ctx, args_k, count, ia64_task_regs(current)); + ret = (*func)(ctx, args_k, count, task_pt_regs(current)); call_made = 1; @@ -5049,7 +5053,7 @@ pfm_handle_work(void) pfm_clear_task_notify(); - regs = ia64_task_regs(current); + regs = task_pt_regs(current); /* * extract reason for being here and clear @@ -5086,7 +5090,7 @@ pfm_handle_work(void) * may go through without blocking on SMP systems * if restart has been received already by the time we call down() */ - ret = down_interruptible(&ctx->ctx_restart_sem); + ret = wait_for_completion_interruptible(&ctx->ctx_restart_done); DPRINT(("after block sleeping ret=%d\n", ret)); @@ -5793,7 +5797,7 @@ pfm_syst_wide_update_task(struct task_struct *task, unsigned long info, int is_c * on every CPU, so we can rely on the pid to identify the idle task. */ if ((info & PFM_CPUINFO_EXCL_IDLE) == 0 || task->pid) { - regs = ia64_task_regs(task); + regs = task_pt_regs(task); ia64_psr(regs)->pp = is_ctxswin ? dcr_pp : 0; return; } @@ -5876,7 +5880,7 @@ pfm_save_regs(struct task_struct *task) flags = pfm_protect_ctx_ctxsw(ctx); if (ctx->ctx_state == PFM_CTX_ZOMBIE) { - struct pt_regs *regs = ia64_task_regs(task); + struct pt_regs *regs = task_pt_regs(task); pfm_clear_psr_up(); @@ -6076,7 +6080,7 @@ pfm_load_regs (struct task_struct *task) BUG_ON(psr & IA64_PSR_I); if (unlikely(ctx->ctx_state == PFM_CTX_ZOMBIE)) { - struct pt_regs *regs = ia64_task_regs(task); + struct pt_regs *regs = task_pt_regs(task); BUG_ON(ctx->ctx_smpl_hdr); @@ -6445,7 +6449,7 @@ pfm_alt_save_pmu_state(void *data) { struct pt_regs *regs; - regs = ia64_task_regs(current); + regs = task_pt_regs(current); DPRINT(("called\n")); @@ -6471,7 +6475,7 @@ pfm_alt_restore_pmu_state(void *data) { struct pt_regs *regs; - regs = ia64_task_regs(current); + regs = task_pt_regs(current); DPRINT(("called\n")); @@ -6753,7 +6757,7 @@ dump_pmu_state(const char *from) local_irq_save(flags); this_cpu = smp_processor_id(); - regs = ia64_task_regs(current); + regs = task_pt_regs(current); info = PFM_CPUINFO_GET(); dcr = ia64_getreg(_IA64_REG_CR_DCR); diff --git a/arch/ia64/kernel/perfmon_montecito.h b/arch/ia64/kernel/perfmon_montecito.h new file mode 100644 index 00000000000..cd06ac6a686 --- /dev/null +++ b/arch/ia64/kernel/perfmon_montecito.h @@ -0,0 +1,269 @@ +/* + * This file contains the Montecito PMU register description tables + * and pmc checker used by perfmon.c. + * + * Copyright (c) 2005-2006 Hewlett-Packard Development Company, L.P. + * Contributed by Stephane Eranian <eranian@hpl.hp.com> + */ +static int pfm_mont_pmc_check(struct task_struct *task, pfm_context_t *ctx, unsigned int cnum, unsigned long *val, struct pt_regs *regs); + +#define RDEP_MONT_ETB (RDEP(38)|RDEP(39)|RDEP(48)|RDEP(49)|RDEP(50)|RDEP(51)|RDEP(52)|RDEP(53)|RDEP(54)|\ + RDEP(55)|RDEP(56)|RDEP(57)|RDEP(58)|RDEP(59)|RDEP(60)|RDEP(61)|RDEP(62)|RDEP(63)) +#define RDEP_MONT_DEAR (RDEP(32)|RDEP(33)|RDEP(36)) +#define RDEP_MONT_IEAR (RDEP(34)|RDEP(35)) + +static pfm_reg_desc_t pfm_mont_pmc_desc[PMU_MAX_PMCS]={ +/* pmc0 */ { PFM_REG_CONTROL , 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc1 */ { PFM_REG_CONTROL , 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc2 */ { PFM_REG_CONTROL , 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc3 */ { PFM_REG_CONTROL , 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc4 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(4),0, 0, 0}, {0,0, 0, 0}}, +/* pmc5 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(5),0, 0, 0}, {0,0, 0, 0}}, +/* pmc6 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(6),0, 0, 0}, {0,0, 0, 0}}, +/* pmc7 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(7),0, 0, 0}, {0,0, 0, 0}}, +/* pmc8 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(8),0, 0, 0}, {0,0, 0, 0}}, +/* pmc9 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(9),0, 0, 0}, {0,0, 0, 0}}, +/* pmc10 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(10),0, 0, 0}, {0,0, 0, 0}}, +/* pmc11 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(11),0, 0, 0}, {0,0, 0, 0}}, +/* pmc12 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(12),0, 0, 0}, {0,0, 0, 0}}, +/* pmc13 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(13),0, 0, 0}, {0,0, 0, 0}}, +/* pmc14 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(14),0, 0, 0}, {0,0, 0, 0}}, +/* pmc15 */ { PFM_REG_COUNTING, 6, 0x2000000, 0x7c7fff7f, NULL, pfm_mont_pmc_check, {RDEP(15),0, 0, 0}, {0,0, 0, 0}}, +/* pmc16 */ { PFM_REG_NOTIMPL, }, +/* pmc17 */ { PFM_REG_NOTIMPL, }, +/* pmc18 */ { PFM_REG_NOTIMPL, }, +/* pmc19 */ { PFM_REG_NOTIMPL, }, +/* pmc20 */ { PFM_REG_NOTIMPL, }, +/* pmc21 */ { PFM_REG_NOTIMPL, }, +/* pmc22 */ { PFM_REG_NOTIMPL, }, +/* pmc23 */ { PFM_REG_NOTIMPL, }, +/* pmc24 */ { PFM_REG_NOTIMPL, }, +/* pmc25 */ { PFM_REG_NOTIMPL, }, +/* pmc26 */ { PFM_REG_NOTIMPL, }, +/* pmc27 */ { PFM_REG_NOTIMPL, }, +/* pmc28 */ { PFM_REG_NOTIMPL, }, +/* pmc29 */ { PFM_REG_NOTIMPL, }, +/* pmc30 */ { PFM_REG_NOTIMPL, }, +/* pmc31 */ { PFM_REG_NOTIMPL, }, +/* pmc32 */ { PFM_REG_CONFIG, 0, 0x30f01ffffffffff, 0x30f01ffffffffff, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc33 */ { PFM_REG_CONFIG, 0, 0x0, 0x1ffffffffff, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc34 */ { PFM_REG_CONFIG, 0, 0xf01ffffffffff, 0xf01ffffffffff, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc35 */ { PFM_REG_CONFIG, 0, 0x0, 0x1ffffffffff, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc36 */ { PFM_REG_CONFIG, 0, 0xfffffff0, 0xf, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc37 */ { PFM_REG_MONITOR, 4, 0x0, 0x3fff, NULL, pfm_mont_pmc_check, {RDEP_MONT_IEAR, 0, 0, 0}, {0, 0, 0, 0}}, +/* pmc38 */ { PFM_REG_CONFIG, 0, 0xdb6, 0x2492, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc39 */ { PFM_REG_MONITOR, 6, 0x0, 0xffcf, NULL, pfm_mont_pmc_check, {RDEP_MONT_ETB,0, 0, 0}, {0,0, 0, 0}}, +/* pmc40 */ { PFM_REG_MONITOR, 6, 0x2000000, 0xf01cf, NULL, pfm_mont_pmc_check, {RDEP_MONT_DEAR,0, 0, 0}, {0,0, 0, 0}}, +/* pmc41 */ { PFM_REG_CONFIG, 0, 0x00002078fefefefe, 0x1e00018181818, NULL, pfm_mont_pmc_check, {0,0, 0, 0}, {0,0, 0, 0}}, +/* pmc42 */ { PFM_REG_MONITOR, 6, 0x0, 0x7ff4f, NULL, pfm_mont_pmc_check, {RDEP_MONT_ETB,0, 0, 0}, {0,0, 0, 0}}, + { PFM_REG_END , 0, 0x0, -1, NULL, NULL, {0,}, {0,}}, /* end marker */ +}; + +static pfm_reg_desc_t pfm_mont_pmd_desc[PMU_MAX_PMDS]={ +/* pmd0 */ { PFM_REG_NOTIMPL, }, +/* pmd1 */ { PFM_REG_NOTIMPL, }, +/* pmd2 */ { PFM_REG_NOTIMPL, }, +/* pmd3 */ { PFM_REG_NOTIMPL, }, +/* pmd4 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(4),0, 0, 0}}, +/* pmd5 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(5),0, 0, 0}}, +/* pmd6 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(6),0, 0, 0}}, +/* pmd7 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(7),0, 0, 0}}, +/* pmd8 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(8),0, 0, 0}}, +/* pmd9 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(9),0, 0, 0}}, +/* pmd10 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(10),0, 0, 0}}, +/* pmd11 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(11),0, 0, 0}}, +/* pmd12 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(12),0, 0, 0}}, +/* pmd13 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(13),0, 0, 0}}, +/* pmd14 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(14),0, 0, 0}}, +/* pmd15 */ { PFM_REG_COUNTING, 0, 0x0, -1, NULL, NULL, {0,0, 0, 0}, {RDEP(15),0, 0, 0}}, +/* pmd16 */ { PFM_REG_NOTIMPL, }, +/* pmd17 */ { PFM_REG_NOTIMPL, }, +/* pmd18 */ { PFM_REG_NOTIMPL, }, +/* pmd19 */ { PFM_REG_NOTIMPL, }, +/* pmd20 */ { PFM_REG_NOTIMPL, }, +/* pmd21 */ { PFM_REG_NOTIMPL, }, +/* pmd22 */ { PFM_REG_NOTIMPL, }, +/* pmd23 */ { PFM_REG_NOTIMPL, }, +/* pmd24 */ { PFM_REG_NOTIMPL, }, +/* pmd25 */ { PFM_REG_NOTIMPL, }, +/* pmd26 */ { PFM_REG_NOTIMPL, }, +/* pmd27 */ { PFM_REG_NOTIMPL, }, +/* pmd28 */ { PFM_REG_NOTIMPL, }, +/* pmd29 */ { PFM_REG_NOTIMPL, }, +/* pmd30 */ { PFM_REG_NOTIMPL, }, +/* pmd31 */ { PFM_REG_NOTIMPL, }, +/* pmd32 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP(33)|RDEP(36),0, 0, 0}, {RDEP(40),0, 0, 0}}, +/* pmd33 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP(32)|RDEP(36),0, 0, 0}, {RDEP(40),0, 0, 0}}, +/* pmd34 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP(35),0, 0, 0}, {RDEP(37),0, 0, 0}}, +/* pmd35 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP(34),0, 0, 0}, {RDEP(37),0, 0, 0}}, +/* pmd36 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP(32)|RDEP(33),0, 0, 0}, {RDEP(40),0, 0, 0}}, +/* pmd37 */ { PFM_REG_NOTIMPL, }, +/* pmd38 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd39 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd40 */ { PFM_REG_NOTIMPL, }, +/* pmd41 */ { PFM_REG_NOTIMPL, }, +/* pmd42 */ { PFM_REG_NOTIMPL, }, +/* pmd43 */ { PFM_REG_NOTIMPL, }, +/* pmd44 */ { PFM_REG_NOTIMPL, }, +/* pmd45 */ { PFM_REG_NOTIMPL, }, +/* pmd46 */ { PFM_REG_NOTIMPL, }, +/* pmd47 */ { PFM_REG_NOTIMPL, }, +/* pmd48 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd49 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd50 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd51 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd52 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd53 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd54 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd55 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd56 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd57 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd58 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd59 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd60 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd61 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd62 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, +/* pmd63 */ { PFM_REG_BUFFER, 0, 0x0, -1, NULL, NULL, {RDEP_MONT_ETB,0, 0, 0}, {RDEP(39),0, 0, 0}}, + { PFM_REG_END , 0, 0x0, -1, NULL, NULL, {0,}, {0,}}, /* end marker */ +}; + +/* + * PMC reserved fields must have their power-up values preserved + */ +static int +pfm_mont_reserved(unsigned int cnum, unsigned long *val, struct pt_regs *regs) +{ + unsigned long tmp1, tmp2, ival = *val; + + /* remove reserved areas from user value */ + tmp1 = ival & PMC_RSVD_MASK(cnum); + + /* get reserved fields values */ + tmp2 = PMC_DFL_VAL(cnum) & ~PMC_RSVD_MASK(cnum); + + *val = tmp1 | tmp2; + + DPRINT(("pmc[%d]=0x%lx, mask=0x%lx, reset=0x%lx, val=0x%lx\n", + cnum, ival, PMC_RSVD_MASK(cnum), PMC_DFL_VAL(cnum), *val)); + return 0; +} + +/* + * task can be NULL if the context is unloaded + */ +static int +pfm_mont_pmc_check(struct task_struct *task, pfm_context_t *ctx, unsigned int cnum, unsigned long *val, struct pt_regs *regs) +{ + int ret = 0; + unsigned long val32 = 0, val38 = 0, val41 = 0; + unsigned long tmpval; + int check_case1 = 0; + int is_loaded; + + /* first preserve the reserved fields */ + pfm_mont_reserved(cnum, val, regs); + + tmpval = *val; + + /* sanity check */ + if (ctx == NULL) return -EINVAL; + + is_loaded = ctx->ctx_state == PFM_CTX_LOADED || ctx->ctx_state == PFM_CTX_MASKED; + + /* + * we must clear the debug registers if pmc41 has a value which enable + * memory pipeline event constraints. In this case we need to clear the + * the debug registers if they have not yet been accessed. This is required + * to avoid picking stale state. + * PMC41 is "active" if: + * one of the pmc41.cfg_dtagXX field is different from 0x3 + * AND + * at the corresponding pmc41.en_dbrpXX is set. + * AND + * ctx_fl_using_dbreg == 0 (i.e., dbr not yet used) + */ + DPRINT(("cnum=%u val=0x%lx, using_dbreg=%d loaded=%d\n", cnum, tmpval, ctx->ctx_fl_using_dbreg, is_loaded)); + + if (cnum == 41 && is_loaded + && (tmpval & 0x1e00000000000) && (tmpval & 0x18181818UL) != 0x18181818UL && ctx->ctx_fl_using_dbreg == 0) { + + DPRINT(("pmc[%d]=0x%lx has active pmc41 settings, clearing dbr\n", cnum, tmpval)); + + /* don't mix debug with perfmon */ + if (task && (task->thread.flags & IA64_THREAD_DBG_VALID) != 0) return -EINVAL; + + /* + * a count of 0 will mark the debug registers if: + * AND + */ + ret = pfm_write_ibr_dbr(PFM_DATA_RR, ctx, NULL, 0, regs); + if (ret) return ret; + } + /* + * we must clear the (instruction) debug registers if: + * pmc38.ig_ibrpX is 0 (enabled) + * AND + * ctx_fl_using_dbreg == 0 (i.e., dbr not yet used) + */ + if (cnum == 38 && is_loaded && ((tmpval & 0x492UL) != 0x492UL) && ctx->ctx_fl_using_dbreg == 0) { + + DPRINT(("pmc38=0x%lx has active pmc38 settings, clearing ibr\n", tmpval)); + + /* don't mix debug with perfmon */ + if (task && (task->thread.flags & IA64_THREAD_DBG_VALID) != 0) return -EINVAL; + + /* + * a count of 0 will mark the debug registers as in use and also + * ensure that they are properly cleared. + */ + ret = pfm_write_ibr_dbr(PFM_CODE_RR, ctx, NULL, 0, regs); + if (ret) return ret; + + } + switch(cnum) { + case 32: val32 = *val; + val38 = ctx->ctx_pmcs[38]; + val41 = ctx->ctx_pmcs[41]; + check_case1 = 1; + break; + case 38: val38 = *val; + val32 = ctx->ctx_pmcs[32]; + val41 = ctx->ctx_pmcs[41]; + check_case1 = 1; + break; + case 41: val41 = *val; + val32 = ctx->ctx_pmcs[32]; + val38 = ctx->ctx_pmcs[38]; + check_case1 = 1; + break; + } + /* check illegal configuration which can produce inconsistencies in tagging + * i-side events in L1D and L2 caches + */ + if (check_case1) { + ret = (((val41 >> 45) & 0xf) == 0 && ((val32>>57) & 0x1) == 0) + && ((((val38>>1) & 0x3) == 0x2 || ((val38>>1) & 0x3) == 0) + || (((val38>>4) & 0x3) == 0x2 || ((val38>>4) & 0x3) == 0)); + if (ret) { + DPRINT(("invalid config pmc38=0x%lx pmc41=0x%lx pmc32=0x%lx\n", val38, val41, val32)); + return -EINVAL; + } + } + *val = tmpval; + return 0; +} + +/* + * impl_pmcs, impl_pmds are computed at runtime to minimize errors! + */ +static pmu_config_t pmu_conf_mont={ + .pmu_name = "Montecito", + .pmu_family = 0x20, + .flags = PFM_PMU_IRQ_RESEND, + .ovfl_val = (1UL << 47) - 1, + .pmd_desc = pfm_mont_pmd_desc, + .pmc_desc = pfm_mont_pmc_desc, + .num_ibrs = 8, + .num_dbrs = 8, + .use_rr_dbregs = 1 /* debug register are use for range retrictions */ +}; diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c index e9904c74d2b..309d59658e5 100644 --- a/arch/ia64/kernel/process.c +++ b/arch/ia64/kernel/process.c @@ -328,7 +328,7 @@ ia64_save_extra (struct task_struct *task) #endif #ifdef CONFIG_IA32_SUPPORT - if (IS_IA32_PROCESS(ia64_task_regs(task))) + if (IS_IA32_PROCESS(task_pt_regs(task))) ia32_save_state(task); #endif } @@ -353,7 +353,7 @@ ia64_load_extra (struct task_struct *task) #endif #ifdef CONFIG_IA32_SUPPORT - if (IS_IA32_PROCESS(ia64_task_regs(task))) + if (IS_IA32_PROCESS(task_pt_regs(task))) ia32_load_state(task); #endif } @@ -488,7 +488,7 @@ copy_thread (int nr, unsigned long clone_flags, * If we're cloning an IA32 task then save the IA32 extra * state from the current task to the new task */ - if (IS_IA32_PROCESS(ia64_task_regs(current))) { + if (IS_IA32_PROCESS(task_pt_regs(current))) { ia32_save_state(p); if (clone_flags & CLONE_SETTLS) retval = ia32_clone_tls(p, child_ptregs); @@ -701,7 +701,7 @@ int kernel_thread_helper (int (*fn)(void *), void *arg) { #ifdef CONFIG_IA32_SUPPORT - if (IS_IA32_PROCESS(ia64_task_regs(current))) { + if (IS_IA32_PROCESS(task_pt_regs(current))) { /* A kernel thread is always a 64-bit process. */ current->thread.map_base = DEFAULT_MAP_BASE; current->thread.task_size = DEFAULT_TASK_SIZE; @@ -722,7 +722,7 @@ flush_thread (void) current->thread.flags &= ~(IA64_THREAD_FPH_VALID | IA64_THREAD_DBG_VALID); ia64_drop_fpu(current); #ifdef CONFIG_IA32_SUPPORT - if (IS_IA32_PROCESS(ia64_task_regs(current))) { + if (IS_IA32_PROCESS(task_pt_regs(current))) { ia32_drop_partial_page_list(current); current->thread.task_size = IA32_PAGE_OFFSET; set_fs(USER_DS); @@ -755,7 +755,7 @@ exit_thread (void) if (current->thread.flags & IA64_THREAD_DBG_VALID) pfm_release_debug_registers(current); #endif - if (IS_IA32_PROCESS(ia64_task_regs(current))) + if (IS_IA32_PROCESS(task_pt_regs(current))) ia32_drop_partial_page_list(current); } diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 4b19d041063..eaed14aac6a 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -254,7 +254,7 @@ get_rnat (struct task_struct *task, struct switch_stack *sw, long num_regs, nbits; struct pt_regs *pt; - pt = ia64_task_regs(task); + pt = task_pt_regs(task); kbsp = (unsigned long *) sw->ar_bspstore; ubspstore = (unsigned long *) pt->ar_bspstore; @@ -314,7 +314,7 @@ put_rnat (struct task_struct *task, struct switch_stack *sw, struct pt_regs *pt; unsigned long cfm, *urbs_kargs; - pt = ia64_task_regs(task); + pt = task_pt_regs(task); kbsp = (unsigned long *) sw->ar_bspstore; ubspstore = (unsigned long *) pt->ar_bspstore; @@ -407,7 +407,7 @@ ia64_peek (struct task_struct *child, struct switch_stack *child_stack, urbs_end = (long *) user_rbs_end; laddr = (unsigned long *) addr; - child_regs = ia64_task_regs(child); + child_regs = task_pt_regs(child); bspstore = (unsigned long *) child_regs->ar_bspstore; krbs = (unsigned long *) child + IA64_RBS_OFFSET/8; if (on_kernel_rbs(addr, (unsigned long) bspstore, @@ -467,7 +467,7 @@ ia64_poke (struct task_struct *child, struct switch_stack *child_stack, struct pt_regs *child_regs; laddr = (unsigned long *) addr; - child_regs = ia64_task_regs(child); + child_regs = task_pt_regs(child); bspstore = (unsigned long *) child_regs->ar_bspstore; krbs = (unsigned long *) child + IA64_RBS_OFFSET/8; if (on_kernel_rbs(addr, (unsigned long) bspstore, @@ -567,7 +567,7 @@ thread_matches (struct task_struct *thread, unsigned long addr) */ return 0; - thread_regs = ia64_task_regs(thread); + thread_regs = task_pt_regs(thread); thread_rbs_end = ia64_get_user_rbs_end(thread, thread_regs, NULL); if (!on_kernel_rbs(addr, thread_regs->ar_bspstore, thread_rbs_end)) return 0; @@ -627,7 +627,7 @@ find_thread_for_addr (struct task_struct *child, unsigned long addr) inline void ia64_flush_fph (struct task_struct *task) { - struct ia64_psr *psr = ia64_psr(ia64_task_regs(task)); + struct ia64_psr *psr = ia64_psr(task_pt_regs(task)); /* * Prevent migrating this task while @@ -653,7 +653,7 @@ ia64_flush_fph (struct task_struct *task) void ia64_sync_fph (struct task_struct *task) { - struct ia64_psr *psr = ia64_psr(ia64_task_regs(task)); + struct ia64_psr *psr = ia64_psr(task_pt_regs(task)); ia64_flush_fph(task); if (!(task->thread.flags & IA64_THREAD_FPH_VALID)) { @@ -794,7 +794,7 @@ access_uarea (struct task_struct *child, unsigned long addr, + offsetof(struct pt_regs, reg))) - pt = ia64_task_regs(child); + pt = task_pt_regs(child); sw = (struct switch_stack *) (child->thread.ksp + 16); if ((addr & 0x7) != 0) { @@ -1120,7 +1120,7 @@ ptrace_getregs (struct task_struct *child, struct pt_all_user_regs __user *ppr) if (!access_ok(VERIFY_WRITE, ppr, sizeof(struct pt_all_user_regs))) return -EIO; - pt = ia64_task_regs(child); + pt = task_pt_regs(child); sw = (struct switch_stack *) (child->thread.ksp + 16); unw_init_from_blocked_task(&info, child); if (unw_unwind_to_user(&info) < 0) { @@ -1265,7 +1265,7 @@ ptrace_setregs (struct task_struct *child, struct pt_all_user_regs __user *ppr) if (!access_ok(VERIFY_READ, ppr, sizeof(struct pt_all_user_regs))) return -EIO; - pt = ia64_task_regs(child); + pt = task_pt_regs(child); sw = (struct switch_stack *) (child->thread.ksp + 16); unw_init_from_blocked_task(&info, child); if (unw_unwind_to_user(&info) < 0) { @@ -1403,7 +1403,7 @@ ptrace_setregs (struct task_struct *child, struct pt_all_user_regs __user *ppr) void ptrace_disable (struct task_struct *child) { - struct ia64_psr *child_psr = ia64_psr(ia64_task_regs(child)); + struct ia64_psr *child_psr = ia64_psr(task_pt_regs(child)); /* make sure the single step/taken-branch trap bits are not set: */ child_psr->ss = 0; @@ -1422,14 +1422,7 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) lock_kernel(); ret = -EPERM; if (request == PTRACE_TRACEME) { - /* are we already being traced? */ - if (current->ptrace & PT_PTRACED) - goto out; - ret = security_ptrace(current->parent, current); - if (ret) - goto out; - current->ptrace |= PT_PTRACED; - ret = 0; + ret = ptrace_traceme(); goto out; } @@ -1463,7 +1456,7 @@ sys_ptrace (long request, pid_t pid, unsigned long addr, unsigned long data) if (ret < 0) goto out_tsk; - pt = ia64_task_regs(child); + pt = task_pt_regs(child); sw = (struct switch_stack *) (child->thread.ksp + 16); switch (request) { diff --git a/arch/ia64/kernel/salinfo.c b/arch/ia64/kernel/salinfo.c index 1461dc660b4..9d5a823479a 100644 --- a/arch/ia64/kernel/salinfo.c +++ b/arch/ia64/kernel/salinfo.c @@ -3,7 +3,7 @@ * * Creates entries in /proc/sal for various system features. * - * Copyright (c) 2003 Silicon Graphics, Inc. All rights reserved. + * Copyright (c) 2003, 2006 Silicon Graphics, Inc. All rights reserved. * Copyright (c) 2003 Hewlett-Packard Co * Bjorn Helgaas <bjorn.helgaas@hp.com> * @@ -27,8 +27,17 @@ * mca.c may not pass a buffer, a NULL buffer just indicates that a new * record is available in SAL. * Replace some NR_CPUS by cpus_online, for hotplug cpu. + * + * Jan 5 2006 kaos@sgi.com + * Handle hotplug cpus coming online. + * Handle hotplug cpus going offline while they still have outstanding records. + * Use the cpu_* macros consistently. + * Replace the counting semaphore with a mutex and a test if the cpumask is non-empty. + * Modify the locking to make the test for "work to do" an atomic operation. */ +#include <linux/capability.h> +#include <linux/cpu.h> #include <linux/types.h> #include <linux/proc_fs.h> #include <linux/module.h> @@ -131,8 +140,8 @@ enum salinfo_state { }; struct salinfo_data { - volatile cpumask_t cpu_event; /* which cpus have outstanding events */ - struct semaphore sem; /* count of cpus with outstanding events (bits set in cpu_event) */ + cpumask_t cpu_event; /* which cpus have outstanding events */ + struct semaphore mutex; u8 *log_buffer; u64 log_size; u8 *oemdata; /* decoded oem data */ @@ -173,6 +182,21 @@ struct salinfo_platform_oemdata_parms { int ret; }; +/* Kick the mutex that tells user space that there is work to do. Instead of + * trying to track the state of the mutex across multiple cpus, in user + * context, interrupt context, non-maskable interrupt context and hotplug cpu, + * it is far easier just to grab the mutex if it is free then release it. + * + * This routine must be called with data_saved_lock held, to make the down/up + * operation atomic. + */ +static void +salinfo_work_to_do(struct salinfo_data *data) +{ + down_trylock(&data->mutex); + up(&data->mutex); +} + static void salinfo_platform_oemdata_cpu(void *context) { @@ -211,9 +235,9 @@ salinfo_log_wakeup(int type, u8 *buffer, u64 size, int irqsafe) BUG_ON(type >= ARRAY_SIZE(salinfo_log_name)); + if (irqsafe) + spin_lock_irqsave(&data_saved_lock, flags); if (buffer) { - if (irqsafe) - spin_lock_irqsave(&data_saved_lock, flags); for (i = 0, data_saved = data->data_saved; i < saved_size; ++i, ++data_saved) { if (!data_saved->buffer) break; @@ -231,13 +255,11 @@ salinfo_log_wakeup(int type, u8 *buffer, u64 size, int irqsafe) data_saved->size = size; data_saved->buffer = buffer; } - if (irqsafe) - spin_unlock_irqrestore(&data_saved_lock, flags); } - - if (!test_and_set_bit(smp_processor_id(), &data->cpu_event)) { - if (irqsafe) - up(&data->sem); + cpu_set(smp_processor_id(), data->cpu_event); + if (irqsafe) { + salinfo_work_to_do(data); + spin_unlock_irqrestore(&data_saved_lock, flags); } } @@ -248,20 +270,17 @@ static struct timer_list salinfo_timer; static void salinfo_timeout_check(struct salinfo_data *data) { - int i; + unsigned long flags; if (!data->open) return; - for_each_online_cpu(i) { - if (test_bit(i, &data->cpu_event)) { - /* double up() is not a problem, user space will see no - * records for the additional "events". - */ - up(&data->sem); - } + if (!cpus_empty(data->cpu_event)) { + spin_lock_irqsave(&data_saved_lock, flags); + salinfo_work_to_do(data); + spin_unlock_irqrestore(&data_saved_lock, flags); } } -static void +static void salinfo_timeout (unsigned long arg) { salinfo_timeout_check(salinfo_data + SAL_INFO_TYPE_MCA); @@ -289,16 +308,20 @@ salinfo_event_read(struct file *file, char __user *buffer, size_t count, loff_t int i, n, cpu = -1; retry: - if (down_trylock(&data->sem)) { + if (cpus_empty(data->cpu_event) && down_trylock(&data->mutex)) { if (file->f_flags & O_NONBLOCK) return -EAGAIN; - if (down_interruptible(&data->sem)) + if (down_interruptible(&data->mutex)) return -EINTR; } n = data->cpu_check; for (i = 0; i < NR_CPUS; i++) { - if (test_bit(n, &data->cpu_event) && cpu_online(n)) { + if (cpu_isset(n, data->cpu_event)) { + if (!cpu_online(n)) { + cpu_clear(n, data->cpu_event); + continue; + } cpu = n; break; } @@ -309,9 +332,6 @@ retry: if (cpu == -1) goto retry; - /* events are sticky until the user says "clear" */ - up(&data->sem); - /* for next read, start checking at next CPU */ data->cpu_check = cpu; if (++data->cpu_check == NR_CPUS) @@ -380,10 +400,8 @@ salinfo_log_release(struct inode *inode, struct file *file) static void call_on_cpu(int cpu, void (*fn)(void *), void *arg) { - cpumask_t save_cpus_allowed, new_cpus_allowed; - memcpy(&save_cpus_allowed, ¤t->cpus_allowed, sizeof(save_cpus_allowed)); - memset(&new_cpus_allowed, 0, sizeof(new_cpus_allowed)); - set_bit(cpu, &new_cpus_allowed); + cpumask_t save_cpus_allowed = current->cpus_allowed; + cpumask_t new_cpus_allowed = cpumask_of_cpu(cpu); set_cpus_allowed(current, new_cpus_allowed); (*fn)(arg); set_cpus_allowed(current, save_cpus_allowed); @@ -432,10 +450,10 @@ retry: if (!data->saved_num) call_on_cpu(cpu, salinfo_log_read_cpu, data); if (!data->log_size) { - data->state = STATE_NO_DATA; - clear_bit(cpu, &data->cpu_event); + data->state = STATE_NO_DATA; + cpu_clear(cpu, data->cpu_event); } else { - data->state = STATE_LOG_RECORD; + data->state = STATE_LOG_RECORD; } } @@ -472,27 +490,31 @@ static int salinfo_log_clear(struct salinfo_data *data, int cpu) { sal_log_record_header_t *rh; + unsigned long flags; + spin_lock_irqsave(&data_saved_lock, flags); data->state = STATE_NO_DATA; - if (!test_bit(cpu, &data->cpu_event)) + if (!cpu_isset(cpu, data->cpu_event)) { + spin_unlock_irqrestore(&data_saved_lock, flags); return 0; - down(&data->sem); - clear_bit(cpu, &data->cpu_event); + } + cpu_clear(cpu, data->cpu_event); if (data->saved_num) { - unsigned long flags; - spin_lock_irqsave(&data_saved_lock, flags); - shift1_data_saved(data, data->saved_num - 1 ); + shift1_data_saved(data, data->saved_num - 1); data->saved_num = 0; - spin_unlock_irqrestore(&data_saved_lock, flags); } + spin_unlock_irqrestore(&data_saved_lock, flags); rh = (sal_log_record_header_t *)(data->log_buffer); /* Corrected errors have already been cleared from SAL */ if (rh->severity != sal_log_severity_corrected) call_on_cpu(cpu, salinfo_log_clear_cpu, data); /* clearing a record may make a new record visible */ salinfo_log_new_read(cpu, data); - if (data->state == STATE_LOG_RECORD && - !test_and_set_bit(cpu, &data->cpu_event)) - up(&data->sem); + if (data->state == STATE_LOG_RECORD) { + spin_lock_irqsave(&data_saved_lock, flags); + cpu_set(cpu, data->cpu_event); + salinfo_work_to_do(data); + spin_unlock_irqrestore(&data_saved_lock, flags); + } return 0; } @@ -549,6 +571,53 @@ static struct file_operations salinfo_data_fops = { .write = salinfo_log_write, }; +#ifdef CONFIG_HOTPLUG_CPU +static int __devinit +salinfo_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) +{ + unsigned int i, cpu = (unsigned long)hcpu; + unsigned long flags; + struct salinfo_data *data; + switch (action) { + case CPU_ONLINE: + spin_lock_irqsave(&data_saved_lock, flags); + for (i = 0, data = salinfo_data; + i < ARRAY_SIZE(salinfo_data); + ++i, ++data) { + cpu_set(cpu, data->cpu_event); + salinfo_work_to_do(data); + } + spin_unlock_irqrestore(&data_saved_lock, flags); + break; + case CPU_DEAD: + spin_lock_irqsave(&data_saved_lock, flags); + for (i = 0, data = salinfo_data; + i < ARRAY_SIZE(salinfo_data); + ++i, ++data) { + struct salinfo_data_saved *data_saved; + int j; + for (j = ARRAY_SIZE(data->data_saved) - 1, data_saved = data->data_saved + j; + j >= 0; + --j, --data_saved) { + if (data_saved->buffer && data_saved->cpu == cpu) { + shift1_data_saved(data, j); + } + } + cpu_clear(cpu, data->cpu_event); + } + spin_unlock_irqrestore(&data_saved_lock, flags); + break; + } + return NOTIFY_OK; +} + +static struct notifier_block salinfo_cpu_notifier = +{ + .notifier_call = salinfo_cpu_callback, + .priority = 0, +}; +#endif /* CONFIG_HOTPLUG_CPU */ + static int __init salinfo_init(void) { @@ -556,7 +625,7 @@ salinfo_init(void) struct proc_dir_entry **sdir = salinfo_proc_entries; /* keeps track of every entry */ struct proc_dir_entry *dir, *entry; struct salinfo_data *data; - int i, j, online; + int i, j; salinfo_dir = proc_mkdir("sal", NULL); if (!salinfo_dir) @@ -571,7 +640,7 @@ salinfo_init(void) for (i = 0; i < ARRAY_SIZE(salinfo_log_name); i++) { data = salinfo_data + i; data->type = i; - sema_init(&data->sem, 0); + init_MUTEX(&data->mutex); dir = proc_mkdir(salinfo_log_name[i], salinfo_dir); if (!dir) continue; @@ -591,12 +660,8 @@ salinfo_init(void) *sdir++ = entry; /* we missed any events before now */ - online = 0; - for_each_online_cpu(j) { - set_bit(j, &data->cpu_event); - ++online; - } - sema_init(&data->sem, online); + for_each_online_cpu(j) + cpu_set(j, data->cpu_event); *sdir++ = dir; } @@ -608,6 +673,10 @@ salinfo_init(void) salinfo_timer.function = &salinfo_timeout; add_timer(&salinfo_timer); +#ifdef CONFIG_HOTPLUG_CPU + register_cpu_notifier(&salinfo_cpu_notifier); +#endif + return 0; } diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c index 5add0bcf87a..c0766575a3a 100644 --- a/arch/ia64/kernel/setup.c +++ b/arch/ia64/kernel/setup.c @@ -43,6 +43,7 @@ #include <linux/initrd.h> #include <linux/platform.h> #include <linux/pm.h> +#include <linux/cpufreq.h> #include <asm/ia32.h> #include <asm/machvec.h> @@ -59,6 +60,7 @@ #include <asm/smp.h> #include <asm/system.h> #include <asm/unistd.h> +#include <asm/system.h> #if defined(CONFIG_SMP) && (IA64_CPU_SIZE > PAGE_SIZE) # error "struct cpuinfo_ia64 too big!" @@ -517,6 +519,7 @@ show_cpuinfo (struct seq_file *m, void *v) char family[32], features[128], *cp, sep; struct cpuinfo_ia64 *c = v; unsigned long mask; + unsigned long proc_freq; int i; mask = c->features; @@ -549,6 +552,10 @@ show_cpuinfo (struct seq_file *m, void *v) sprintf(cp, " 0x%lx", mask); } + proc_freq = cpufreq_quick_get(cpunum); + if (!proc_freq) + proc_freq = c->proc_freq / 1000; + seq_printf(m, "processor : %d\n" "vendor : %s\n" @@ -565,7 +572,7 @@ show_cpuinfo (struct seq_file *m, void *v) "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, + proc_freq / 1000, proc_freq % 1000, c->itc_freq / 1000000, c->itc_freq % 1000000, lpj*HZ/500000, (lpj*HZ/5000) % 100); #ifdef CONFIG_SMP @@ -689,6 +696,7 @@ static void get_max_cacheline_size (void) { unsigned long line_size, max = 1; + unsigned int cache_size = 0; u64 l, levels, unique_caches; pal_cache_config_info_t cci; s64 status; @@ -718,6 +726,8 @@ get_max_cacheline_size (void) line_size = 1 << cci.pcci_line_size; if (line_size > max) max = line_size; + if (cache_size < cci.pcci_cache_size) + cache_size = cci.pcci_cache_size; if (!cci.pcci_unified) { status = ia64_pal_cache_config_info(l, /* cache_type (instruction)= */ 1, @@ -734,6 +744,9 @@ get_max_cacheline_size (void) ia64_i_cache_stride_shift = cci.pcci_stride; } out: +#ifdef CONFIG_SMP + max_cache_size = max(max_cache_size, cache_size); +#endif if (max > ia64_max_cacheline_size) ia64_max_cacheline_size = max; } @@ -788,7 +801,7 @@ cpu_init (void) #endif /* Clear the stack memory reserved for pt_regs: */ - memset(ia64_task_regs(current), 0, sizeof(struct pt_regs)); + memset(task_pt_regs(current), 0, sizeof(struct pt_regs)); ia64_set_kr(IA64_KR_FPU_OWNER, 0); @@ -864,6 +877,15 @@ cpu_init (void) pm_idle = default_idle; } +/* + * On SMP systems, when the scheduler does migration-cost autodetection, + * it needs a way to flush as much of the CPU's caches as possible. + */ +void sched_cacheflush(void) +{ + ia64_sal_cache_flush(3); +} + void check_bugs (void) { diff --git a/arch/ia64/kernel/signal.c b/arch/ia64/kernel/signal.c index 58ce07efc56..463f6bb44d0 100644 --- a/arch/ia64/kernel/signal.c +++ b/arch/ia64/kernel/signal.c @@ -655,11 +655,11 @@ set_sigdelayed(pid_t pid, int signo, int code, void __user *addr) if (!t) return; - t->thread_info->sigdelayed.signo = signo; - t->thread_info->sigdelayed.code = code; - t->thread_info->sigdelayed.addr = addr; - t->thread_info->sigdelayed.start_time = start_time; - t->thread_info->sigdelayed.pid = pid; + task_thread_info(t)->sigdelayed.signo = signo; + task_thread_info(t)->sigdelayed.code = code; + task_thread_info(t)->sigdelayed.addr = addr; + task_thread_info(t)->sigdelayed.start_time = start_time; + task_thread_info(t)->sigdelayed.pid = pid; wmb(); set_tsk_thread_flag(t, TIF_SIGDELAYED); } diff --git a/arch/ia64/kernel/sys_ia64.c b/arch/ia64/kernel/sys_ia64.c index f2dbcd1db0d..c7b943f1019 100644 --- a/arch/ia64/kernel/sys_ia64.c +++ b/arch/ia64/kernel/sys_ia64.c @@ -151,7 +151,7 @@ out: asmlinkage long sys_pipe (void) { - struct pt_regs *regs = ia64_task_regs(current); + struct pt_regs *regs = task_pt_regs(current); int fd[2]; int retval; diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c index 5b7e736f3b4..028a2b95936 100644 --- a/arch/ia64/kernel/time.c +++ b/arch/ia64/kernel/time.c @@ -249,3 +249,32 @@ time_init (void) */ set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); } + +#define SMALLUSECS 100 + +void +udelay (unsigned long usecs) +{ + unsigned long start; + unsigned long cycles; + unsigned long smallusecs; + + /* + * Execute the non-preemptible delay loop (because the ITC might + * not be synchronized between CPUS) in relatively short time + * chunks, allowing preemption between the chunks. + */ + while (usecs > 0) { + smallusecs = (usecs > SMALLUSECS) ? SMALLUSECS : usecs; + preempt_disable(); + cycles = smallusecs*local_cpu_data->cyc_per_usec; + start = ia64_get_itc(); + + while (ia64_get_itc() - start < cycles) + cpu_relax(); + + preempt_enable(); + usecs -= smallusecs; + } +} +EXPORT_SYMBOL(udelay); diff --git a/arch/ia64/kernel/traps.c b/arch/ia64/kernel/traps.c index d3e0ecb56d6..55391901b01 100644 --- a/arch/ia64/kernel/traps.c +++ b/arch/ia64/kernel/traps.c @@ -530,12 +530,15 @@ ia64_fault (unsigned long vector, unsigned long isr, unsigned long ifa, if (fsys_mode(current, ®s)) { extern char __kernel_syscall_via_break[]; /* - * Got a trap in fsys-mode: Taken Branch Trap and Single Step trap - * need special handling; Debug trap is not supposed to happen. + * Got a trap in fsys-mode: Taken Branch Trap + * and Single Step trap need special handling; + * Debug trap is ignored (we disable it here + * and re-enable it in the lower-privilege trap). */ if (unlikely(vector == 29)) { - die("Got debug trap in fsys-mode---not supposed to happen!", - ®s, 0); + set_thread_flag(TIF_DB_DISABLED); + ia64_psr(®s)->db = 0; + ia64_psr(®s)->lp = 1; return; } /* re-do the system call via break 0x100000: */ @@ -589,10 +592,19 @@ ia64_fault (unsigned long vector, unsigned long isr, unsigned long ifa, case 34: if (isr & 0x2) { /* Lower-Privilege Transfer Trap */ + + /* If we disabled debug traps during an fsyscall, + * re-enable them here. + */ + if (test_thread_flag(TIF_DB_DISABLED)) { + clear_thread_flag(TIF_DB_DISABLED); + ia64_psr(®s)->db = 1; + } + /* - * Just clear PSR.lp and then return immediately: all the - * interesting work (e.g., signal delivery is done in the kernel - * exit path). + * Just clear PSR.lp and then return immediately: + * all the interesting work (e.g., signal delivery) + * is done in the kernel exit path. */ ia64_psr(®s)->lp = 0; return; diff --git a/arch/ia64/kernel/uncached.c b/arch/ia64/kernel/uncached.c index c6d40446c2c..fcd2bad0286 100644 --- a/arch/ia64/kernel/uncached.c +++ b/arch/ia64/kernel/uncached.c @@ -53,7 +53,7 @@ static void uncached_ipi_visibility(void *data) if ((status != PAL_VISIBILITY_OK) && (status != PAL_VISIBILITY_OK_REMOTE_NEEDED)) printk(KERN_DEBUG "pal_prefetch_visibility() returns %i on " - "CPU %i\n", status, get_cpu()); + "CPU %i\n", status, raw_smp_processor_id()); } @@ -63,7 +63,7 @@ static void uncached_ipi_mc_drain(void *data) status = ia64_pal_mc_drain(); if (status) printk(KERN_WARNING "ia64_pal_mc_drain() failed with %i on " - "CPU %i\n", status, get_cpu()); + "CPU %i\n", status, raw_smp_processor_id()); } @@ -105,7 +105,7 @@ uncached_get_new_chunk(struct gen_pool *poolp) status = ia64_pal_prefetch_visibility(PAL_VISIBILITY_PHYSICAL); dprintk(KERN_INFO "pal_prefetch_visibility() returns %i on cpu %i\n", - status, get_cpu()); + status, raw_smp_processor_id()); if (!status) { status = smp_call_function(uncached_ipi_visibility, NULL, 0, 1); @@ -210,6 +210,7 @@ uncached_build_memmap(unsigned long start, unsigned long end, void *arg) dprintk(KERN_ERR "uncached_build_memmap(%lx %lx)\n", start, end); + touch_softlockup_watchdog(); memset((char *)start, 0, length); node = paddr_to_nid(start - __IA64_UNCACHED_OFFSET); diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S index 30d8564e960..73af6267d2e 100644 --- a/arch/ia64/kernel/vmlinux.lds.S +++ b/arch/ia64/kernel/vmlinux.lds.S @@ -177,6 +177,9 @@ SECTIONS } . = ALIGN(PAGE_SIZE); /* make sure the gate page doesn't expose kernel data */ + .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) + { *(.data.read_mostly) } + .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { *(.data.cacheline_aligned) } diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index e3215ba64ff..b38b6d213c1 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -635,3 +635,39 @@ mem_init (void) ia32_mem_init(); #endif } + +#ifdef CONFIG_MEMORY_HOTPLUG +void online_page(struct page *page) +{ + ClearPageReserved(page); + set_page_count(page, 1); + __free_page(page); + totalram_pages++; + num_physpages++; +} + +int add_memory(u64 start, u64 size) +{ + pg_data_t *pgdat; + struct zone *zone; + unsigned long start_pfn = start >> PAGE_SHIFT; + unsigned long nr_pages = size >> PAGE_SHIFT; + int ret; + + pgdat = NODE_DATA(0); + + zone = pgdat->node_zones + ZONE_NORMAL; + ret = __add_pages(zone, start_pfn, nr_pages); + + if (ret) + printk("%s: Problem encountered in __add_pages() as ret=%d\n", + __FUNCTION__, ret); + + return ret; +} + +int remove_memory(u64 start, u64 size) +{ + return -EINVAL; +} +#endif diff --git a/arch/ia64/mm/tlb.c b/arch/ia64/mm/tlb.c index 41105d45442..6a4eec9113e 100644 --- a/arch/ia64/mm/tlb.c +++ b/arch/ia64/mm/tlb.c @@ -90,7 +90,7 @@ ia64_global_tlb_purge (struct mm_struct *mm, unsigned long start, { static DEFINE_SPINLOCK(ptcg_lock); - if (mm != current->active_mm) { + if (mm != current->active_mm || !current->mm) { flush_tlb_all(); return; } diff --git a/arch/ia64/oprofile/backtrace.c b/arch/ia64/oprofile/backtrace.c index b7dabbfb0d6..adb01566bd5 100644 --- a/arch/ia64/oprofile/backtrace.c +++ b/arch/ia64/oprofile/backtrace.c @@ -32,7 +32,7 @@ typedef struct u64 *prev_pfs_loc; /* state for WAR for old spinlock ool code */ } ia64_backtrace_t; -#if __GNUC__ < 3 || (__GNUC__ == 3 && __GNUC_MINOR__ < 3) +#if (__GNUC__ == 3 && __GNUC_MINOR__ < 3) /* * Returns non-zero if the PC is in the spinlock contention out-of-line code * with non-standard calling sequence (on older compilers). diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c index 20d76fae24e..0b30ca00628 100644 --- a/arch/ia64/pci/pci.c +++ b/arch/ia64/pci/pci.c @@ -193,12 +193,12 @@ add_io_space (struct pci_root_info *info, struct acpi_resource_address64 *addr) goto free_resource; } - min = addr->min_address_range; + min = addr->minimum; max = min + addr->address_length - 1; - if (addr->attribute.io.translation_attribute == ACPI_SPARSE_TRANSLATION) + if (addr->info.io.translation_type == ACPI_SPARSE_TRANSLATION) sparse = 1; - space_nr = new_space(addr->address_translation_offset, sparse); + space_nr = new_space(addr->translation_offset, sparse); if (space_nr == ~0) goto free_name; @@ -285,7 +285,7 @@ static __devinit acpi_status add_window(struct acpi_resource *res, void *data) if (addr.resource_type == ACPI_MEMORY_RANGE) { flags = IORESOURCE_MEM; root = &iomem_resource; - offset = addr.address_translation_offset; + offset = addr.translation_offset; } else if (addr.resource_type == ACPI_IO_RANGE) { flags = IORESOURCE_IO; root = &ioport_resource; @@ -298,7 +298,7 @@ static __devinit acpi_status add_window(struct acpi_resource *res, void *data) window = &info->controller->window[info->controller->windows++]; window->resource.name = info->name; window->resource.flags = flags; - window->resource.start = addr.min_address_range + offset; + window->resource.start = addr.minimum + offset; window->resource.end = window->resource.start + addr.address_length - 1; window->resource.child = NULL; window->offset = offset; @@ -454,14 +454,13 @@ static int __devinit is_valid_resource(struct pci_dev *dev, int idx) return 0; } -static void __devinit pcibios_fixup_device_resources(struct pci_dev *dev) +static void __devinit +pcibios_fixup_resources(struct pci_dev *dev, int start, int limit) { struct pci_bus_region region; int i; - int limit = (dev->hdr_type == PCI_HEADER_TYPE_NORMAL) ? \ - PCI_BRIDGE_RESOURCES : PCI_NUM_RESOURCES; - for (i = 0; i < limit; i++) { + for (i = start; i < limit; i++) { if (!dev->resource[i].flags) continue; region.start = dev->resource[i].start; @@ -472,6 +471,16 @@ static void __devinit pcibios_fixup_device_resources(struct pci_dev *dev) } } +static void __devinit pcibios_fixup_device_resources(struct pci_dev *dev) +{ + pcibios_fixup_resources(dev, 0, PCI_BRIDGE_RESOURCES); +} + +static void __devinit pcibios_fixup_bridge_resources(struct pci_dev *dev) +{ + pcibios_fixup_resources(dev, PCI_BRIDGE_RESOURCES, PCI_NUM_RESOURCES); +} + /* * Called after each bus is probed, but before its children are examined. */ @@ -482,7 +491,7 @@ pcibios_fixup_bus (struct pci_bus *b) if (b->self) { pci_read_bridge_bases(b); - pcibios_fixup_device_resources(b->self); + pcibios_fixup_bridge_resources(b->self); } list_for_each_entry(dev, &b->devices, bus_list) pcibios_fixup_device_resources(dev); @@ -700,7 +709,7 @@ int ia64_pci_legacy_read(struct pci_bus *bus, u16 port, u32 *val, u8 size) */ int ia64_pci_legacy_write(struct pci_dev *bus, u16 port, u32 val, u8 size) { - int ret = 0; + int ret = size; switch (size) { case 1: diff --git a/arch/ia64/sn/include/xtalk/hubdev.h b/arch/ia64/sn/include/xtalk/hubdev.h index 71c2b271b4c..8182583c762 100644 --- a/arch/ia64/sn/include/xtalk/hubdev.h +++ b/arch/ia64/sn/include/xtalk/hubdev.h @@ -26,29 +26,46 @@ #define IIO_NUM_ITTES 7 #define HUB_NUM_BIG_WINDOW (IIO_NUM_ITTES - 1) -struct sn_flush_device_list { +/* This struct is shared between the PROM and the kernel. + * Changes to this struct will require corresponding changes to the kernel. + */ +struct sn_flush_device_common { int sfdl_bus; int sfdl_slot; int sfdl_pin; - struct bar_list { + struct common_bar_list { unsigned long start; unsigned long end; } sfdl_bar_list[6]; unsigned long sfdl_force_int_addr; unsigned long sfdl_flush_value; volatile unsigned long *sfdl_flush_addr; - uint32_t sfdl_persistent_busnum; - uint32_t sfdl_persistent_segment; + u32 sfdl_persistent_busnum; + u32 sfdl_persistent_segment; struct pcibus_info *sfdl_pcibus_info; +}; + +/* This struct is kernel only and is not used by the PROM */ +struct sn_flush_device_kernel { spinlock_t sfdl_flush_lock; + struct sn_flush_device_common *common; +}; + +/* 01/16/06 This struct is the old PROM/kernel struct and needs to be included + * for older official PROMs to function on the new kernel base. This struct + * will be removed when the next official PROM release occurs. */ + +struct sn_flush_device_war { + struct sn_flush_device_common common; + u32 filler; /* older PROMs expect the default size of a spinlock_t */ }; /* - * **widget_p - Used as an array[wid_num][device] of sn_flush_device_list. + * **widget_p - Used as an array[wid_num][device] of sn_flush_device_kernel. */ struct sn_flush_nasid_entry { - struct sn_flush_device_list **widget_p; /* Used as a array of wid_num */ - uint64_t iio_itte[8]; + struct sn_flush_device_kernel **widget_p; // Used as an array of wid_num + u64 iio_itte[8]; }; struct hubdev_info { @@ -62,8 +79,8 @@ struct hubdev_info { void *hdi_nodepda; void *hdi_node_vertex; - uint32_t max_segment_number; - uint32_t max_pcibus_number; + u32 max_segment_number; + u32 max_pcibus_number; }; extern void hubdev_init_node(nodepda_t *, cnodeid_t); diff --git a/arch/ia64/sn/include/xtalk/xbow.h b/arch/ia64/sn/include/xtalk/xbow.h index ec56b3432f1..90f37a4133d 100644 --- a/arch/ia64/sn/include/xtalk/xbow.h +++ b/arch/ia64/sn/include/xtalk/xbow.h @@ -3,7 +3,8 @@ * 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) 1992-1997,2000-2006 Silicon Graphics, Inc. All Rights + * Reserved. */ #ifndef _ASM_IA64_SN_XTALK_XBOW_H #define _ASM_IA64_SN_XTALK_XBOW_H @@ -21,94 +22,94 @@ /* Register set for each xbow link */ typedef volatile struct xb_linkregs_s { -/* +/* * we access these through synergy unswizzled space, so the address * gets twiddled (i.e. references to 0x4 actually go to 0x0 and vv.) * That's why we put the register first and filler second. */ - uint32_t link_ibf; - uint32_t filler0; /* filler for proper alignment */ - uint32_t link_control; - uint32_t filler1; - uint32_t link_status; - uint32_t filler2; - uint32_t link_arb_upper; - uint32_t filler3; - uint32_t link_arb_lower; - uint32_t filler4; - uint32_t link_status_clr; - uint32_t filler5; - uint32_t link_reset; - uint32_t filler6; - uint32_t link_aux_status; - uint32_t filler7; + u32 link_ibf; + u32 filler0; /* filler for proper alignment */ + u32 link_control; + u32 filler1; + u32 link_status; + u32 filler2; + u32 link_arb_upper; + u32 filler3; + u32 link_arb_lower; + u32 filler4; + u32 link_status_clr; + u32 filler5; + u32 link_reset; + u32 filler6; + u32 link_aux_status; + u32 filler7; } xb_linkregs_t; typedef volatile struct xbow_s { - /* standard widget configuration 0x000000-0x000057 */ - struct widget_cfg xb_widget; /* 0x000000 */ - - /* helper fieldnames for accessing bridge widget */ - -#define xb_wid_id xb_widget.w_id -#define xb_wid_stat xb_widget.w_status -#define xb_wid_err_upper xb_widget.w_err_upper_addr -#define xb_wid_err_lower xb_widget.w_err_lower_addr -#define xb_wid_control xb_widget.w_control -#define xb_wid_req_timeout xb_widget.w_req_timeout -#define xb_wid_int_upper xb_widget.w_intdest_upper_addr -#define xb_wid_int_lower xb_widget.w_intdest_lower_addr -#define xb_wid_err_cmdword xb_widget.w_err_cmd_word -#define xb_wid_llp xb_widget.w_llp_cfg -#define xb_wid_stat_clr xb_widget.w_tflush - -/* + /* standard widget configuration 0x000000-0x000057 */ + struct widget_cfg xb_widget; /* 0x000000 */ + + /* helper fieldnames for accessing bridge widget */ + +#define xb_wid_id xb_widget.w_id +#define xb_wid_stat xb_widget.w_status +#define xb_wid_err_upper xb_widget.w_err_upper_addr +#define xb_wid_err_lower xb_widget.w_err_lower_addr +#define xb_wid_control xb_widget.w_control +#define xb_wid_req_timeout xb_widget.w_req_timeout +#define xb_wid_int_upper xb_widget.w_intdest_upper_addr +#define xb_wid_int_lower xb_widget.w_intdest_lower_addr +#define xb_wid_err_cmdword xb_widget.w_err_cmd_word +#define xb_wid_llp xb_widget.w_llp_cfg +#define xb_wid_stat_clr xb_widget.w_tflush + +/* * we access these through synergy unswizzled space, so the address * gets twiddled (i.e. references to 0x4 actually go to 0x0 and vv.) * That's why we put the register first and filler second. */ - /* xbow-specific widget configuration 0x000058-0x0000FF */ - uint32_t xb_wid_arb_reload; /* 0x00005C */ - uint32_t _pad_000058; - uint32_t xb_perf_ctr_a; /* 0x000064 */ - uint32_t _pad_000060; - uint32_t xb_perf_ctr_b; /* 0x00006c */ - uint32_t _pad_000068; - uint32_t xb_nic; /* 0x000074 */ - uint32_t _pad_000070; - - /* Xbridge only */ - uint32_t xb_w0_rst_fnc; /* 0x00007C */ - uint32_t _pad_000078; - uint32_t xb_l8_rst_fnc; /* 0x000084 */ - uint32_t _pad_000080; - uint32_t xb_l9_rst_fnc; /* 0x00008c */ - uint32_t _pad_000088; - uint32_t xb_la_rst_fnc; /* 0x000094 */ - uint32_t _pad_000090; - uint32_t xb_lb_rst_fnc; /* 0x00009c */ - uint32_t _pad_000098; - uint32_t xb_lc_rst_fnc; /* 0x0000a4 */ - uint32_t _pad_0000a0; - uint32_t xb_ld_rst_fnc; /* 0x0000ac */ - uint32_t _pad_0000a8; - uint32_t xb_le_rst_fnc; /* 0x0000b4 */ - uint32_t _pad_0000b0; - uint32_t xb_lf_rst_fnc; /* 0x0000bc */ - uint32_t _pad_0000b8; - uint32_t xb_lock; /* 0x0000c4 */ - uint32_t _pad_0000c0; - uint32_t xb_lock_clr; /* 0x0000cc */ - uint32_t _pad_0000c8; - /* end of Xbridge only */ - uint32_t _pad_0000d0[12]; - - /* Link Specific Registers, port 8..15 0x000100-0x000300 */ - xb_linkregs_t xb_link_raw[MAX_XBOW_PORTS]; -#define xb_link(p) xb_link_raw[(p) & (MAX_XBOW_PORTS - 1)] - + /* xbow-specific widget configuration 0x000058-0x0000FF */ + u32 xb_wid_arb_reload; /* 0x00005C */ + u32 _pad_000058; + u32 xb_perf_ctr_a; /* 0x000064 */ + u32 _pad_000060; + u32 xb_perf_ctr_b; /* 0x00006c */ + u32 _pad_000068; + u32 xb_nic; /* 0x000074 */ + u32 _pad_000070; + + /* Xbridge only */ + u32 xb_w0_rst_fnc; /* 0x00007C */ + u32 _pad_000078; + u32 xb_l8_rst_fnc; /* 0x000084 */ + u32 _pad_000080; + u32 xb_l9_rst_fnc; /* 0x00008c */ + u32 _pad_000088; + u32 xb_la_rst_fnc; /* 0x000094 */ + u32 _pad_000090; + u32 xb_lb_rst_fnc; /* 0x00009c */ + u32 _pad_000098; + u32 xb_lc_rst_fnc; /* 0x0000a4 */ + u32 _pad_0000a0; + u32 xb_ld_rst_fnc; /* 0x0000ac */ + u32 _pad_0000a8; + u32 xb_le_rst_fnc; /* 0x0000b4 */ + u32 _pad_0000b0; + u32 xb_lf_rst_fnc; /* 0x0000bc */ + u32 _pad_0000b8; + u32 xb_lock; /* 0x0000c4 */ + u32 _pad_0000c0; + u32 xb_lock_clr; /* 0x0000cc */ + u32 _pad_0000c8; + /* end of Xbridge only */ + u32 _pad_0000d0[12]; + + /* Link Specific Registers, port 8..15 0x000100-0x000300 */ + xb_linkregs_t xb_link_raw[MAX_XBOW_PORTS]; } xbow_t; +#define xb_link(p) xb_link_raw[(p) & (MAX_XBOW_PORTS - 1)] + #define XB_FLAGS_EXISTS 0x1 /* device exists */ #define XB_FLAGS_MASTER 0x2 #define XB_FLAGS_SLAVE 0x0 @@ -160,7 +161,7 @@ typedef volatile struct xbow_s { /* End of Xbridge only */ /* used only in ide, but defined here within the reserved portion */ -/* of the widget0 address space (before 0xf4) */ +/* of the widget0 address space (before 0xf4) */ #define XBOW_WID_UNDEF 0xe4 /* xbow link register set base, legal value for x is 0x8..0xf */ @@ -179,29 +180,37 @@ typedef volatile struct xbow_s { /* link_control(x) */ #define XB_CTRL_LINKALIVE_IE 0x80000000 /* link comes alive */ - /* reserved: 0x40000000 */ +/* reserved: 0x40000000 */ #define XB_CTRL_PERF_CTR_MODE_MSK 0x30000000 /* perf counter mode */ -#define XB_CTRL_IBUF_LEVEL_MSK 0x0e000000 /* input packet buffer level */ -#define XB_CTRL_8BIT_MODE 0x01000000 /* force link into 8 bit mode */ -#define XB_CTRL_BAD_LLP_PKT 0x00800000 /* force bad LLP packet */ -#define XB_CTRL_WIDGET_CR_MSK 0x007c0000 /* LLP widget credit mask */ -#define XB_CTRL_WIDGET_CR_SHFT 18 /* LLP widget credit shift */ -#define XB_CTRL_ILLEGAL_DST_IE 0x00020000 /* illegal destination */ -#define XB_CTRL_OALLOC_IBUF_IE 0x00010000 /* overallocated input buffer */ - /* reserved: 0x0000fe00 */ +#define XB_CTRL_IBUF_LEVEL_MSK 0x0e000000 /* input packet buffer + level */ +#define XB_CTRL_8BIT_MODE 0x01000000 /* force link into 8 + bit mode */ +#define XB_CTRL_BAD_LLP_PKT 0x00800000 /* force bad LLP + packet */ +#define XB_CTRL_WIDGET_CR_MSK 0x007c0000 /* LLP widget credit + mask */ +#define XB_CTRL_WIDGET_CR_SHFT 18 /* LLP widget credit + shift */ +#define XB_CTRL_ILLEGAL_DST_IE 0x00020000 /* illegal destination + */ +#define XB_CTRL_OALLOC_IBUF_IE 0x00010000 /* overallocated input + buffer */ +/* reserved: 0x0000fe00 */ #define XB_CTRL_BNDWDTH_ALLOC_IE 0x00000100 /* bandwidth alloc */ #define XB_CTRL_RCV_CNT_OFLOW_IE 0x00000080 /* rcv retry overflow */ #define XB_CTRL_XMT_CNT_OFLOW_IE 0x00000040 /* xmt retry overflow */ #define XB_CTRL_XMT_MAX_RTRY_IE 0x00000020 /* max transmit retry */ #define XB_CTRL_RCV_IE 0x00000010 /* receive */ #define XB_CTRL_XMT_RTRY_IE 0x00000008 /* transmit retry */ - /* reserved: 0x00000004 */ -#define XB_CTRL_MAXREQ_TOUT_IE 0x00000002 /* maximum request timeout */ +/* reserved: 0x00000004 */ +#define XB_CTRL_MAXREQ_TOUT_IE 0x00000002 /* maximum request + timeout */ #define XB_CTRL_SRC_TOUT_IE 0x00000001 /* source timeout */ /* link_status(x) */ #define XB_STAT_LINKALIVE XB_CTRL_LINKALIVE_IE - /* reserved: 0x7ff80000 */ +/* reserved: 0x7ff80000 */ #define XB_STAT_MULTI_ERR 0x00040000 /* multi error */ #define XB_STAT_ILLEGAL_DST_ERR XB_CTRL_ILLEGAL_DST_IE #define XB_STAT_OALLOC_IBUF_ERR XB_CTRL_OALLOC_IBUF_IE @@ -211,7 +220,7 @@ typedef volatile struct xbow_s { #define XB_STAT_XMT_MAX_RTRY_ERR XB_CTRL_XMT_MAX_RTRY_IE #define XB_STAT_RCV_ERR XB_CTRL_RCV_IE #define XB_STAT_XMT_RTRY_ERR XB_CTRL_XMT_RTRY_IE - /* reserved: 0x00000004 */ +/* reserved: 0x00000004 */ #define XB_STAT_MAXREQ_TOUT_ERR XB_CTRL_MAXREQ_TOUT_IE #define XB_STAT_SRC_TOUT_ERR XB_CTRL_SRC_TOUT_IE @@ -222,7 +231,7 @@ typedef volatile struct xbow_s { #define XB_AUX_LINKFAIL_RST_BAD 0x00000040 #define XB_AUX_STAT_PRESENT 0x00000020 #define XB_AUX_STAT_PORT_WIDTH 0x00000010 - /* reserved: 0x0000000f */ +/* reserved: 0x0000000f */ /* * link_arb_upper/link_arb_lower(x), (reg) should be the link_arb_upper @@ -238,7 +247,8 @@ typedef volatile struct xbow_s { /* XBOW_WID_STAT */ #define XB_WID_STAT_LINK_INTR_SHFT (24) #define XB_WID_STAT_LINK_INTR_MASK (0xFF << XB_WID_STAT_LINK_INTR_SHFT) -#define XB_WID_STAT_LINK_INTR(x) (0x1 << (((x)&7) + XB_WID_STAT_LINK_INTR_SHFT)) +#define XB_WID_STAT_LINK_INTR(x) \ + (0x1 << (((x)&7) + XB_WID_STAT_LINK_INTR_SHFT)) #define XB_WID_STAT_WIDGET0_INTR 0x00800000 #define XB_WID_STAT_SRCID_MASK 0x000003c0 /* Xbridge only */ #define XB_WID_STAT_REG_ACC_ERR 0x00000020 @@ -264,7 +274,7 @@ typedef volatile struct xbow_s { #define XXBOW_WIDGET_PART_NUM 0xd000 /* Xbridge */ #define XBOW_WIDGET_MFGR_NUM 0x0 #define XXBOW_WIDGET_MFGR_NUM 0x0 -#define PXBOW_WIDGET_PART_NUM 0xd100 /* PIC */ +#define PXBOW_WIDGET_PART_NUM 0xd100 /* PIC */ #define XBOW_REV_1_0 0x1 /* xbow rev 1.0 is "1" */ #define XBOW_REV_1_1 0x2 /* xbow rev 1.1 is "2" */ @@ -279,13 +289,13 @@ typedef volatile struct xbow_s { #define XBOW_WID_ARB_RELOAD_INT 0x3f /* GBR reload interval */ #define IS_XBRIDGE_XBOW(wid) \ - (XWIDGET_PART_NUM(wid) == XXBOW_WIDGET_PART_NUM && \ - XWIDGET_MFG_NUM(wid) == XXBOW_WIDGET_MFGR_NUM) + (XWIDGET_PART_NUM(wid) == XXBOW_WIDGET_PART_NUM && \ + XWIDGET_MFG_NUM(wid) == XXBOW_WIDGET_MFGR_NUM) #define IS_PIC_XBOW(wid) \ - (XWIDGET_PART_NUM(wid) == PXBOW_WIDGET_PART_NUM && \ - XWIDGET_MFG_NUM(wid) == XXBOW_WIDGET_MFGR_NUM) + (XWIDGET_PART_NUM(wid) == PXBOW_WIDGET_PART_NUM && \ + XWIDGET_MFG_NUM(wid) == XXBOW_WIDGET_MFGR_NUM) #define XBOW_WAR_ENABLED(pv, widid) ((1 << XWIDGET_REV_NUM(widid)) & pv) -#endif /* _ASM_IA64_SN_XTALK_XBOW_H */ +#endif /* _ASM_IA64_SN_XTALK_XBOW_H */ diff --git a/arch/ia64/sn/include/xtalk/xwidgetdev.h b/arch/ia64/sn/include/xtalk/xwidgetdev.h index c5f4bc5cc03..2800eda0fd6 100644 --- a/arch/ia64/sn/include/xtalk/xwidgetdev.h +++ b/arch/ia64/sn/include/xtalk/xwidgetdev.h @@ -25,28 +25,28 @@ /* widget configuration registers */ struct widget_cfg{ - uint32_t w_id; /* 0x04 */ - uint32_t w_pad_0; /* 0x00 */ - uint32_t w_status; /* 0x0c */ - uint32_t w_pad_1; /* 0x08 */ - uint32_t w_err_upper_addr; /* 0x14 */ - uint32_t w_pad_2; /* 0x10 */ - uint32_t w_err_lower_addr; /* 0x1c */ - uint32_t w_pad_3; /* 0x18 */ - uint32_t w_control; /* 0x24 */ - uint32_t w_pad_4; /* 0x20 */ - uint32_t w_req_timeout; /* 0x2c */ - uint32_t w_pad_5; /* 0x28 */ - uint32_t w_intdest_upper_addr; /* 0x34 */ - uint32_t w_pad_6; /* 0x30 */ - uint32_t w_intdest_lower_addr; /* 0x3c */ - uint32_t w_pad_7; /* 0x38 */ - uint32_t w_err_cmd_word; /* 0x44 */ - uint32_t w_pad_8; /* 0x40 */ - uint32_t w_llp_cfg; /* 0x4c */ - uint32_t w_pad_9; /* 0x48 */ - uint32_t w_tflush; /* 0x54 */ - uint32_t w_pad_10; /* 0x50 */ + u32 w_id; /* 0x04 */ + u32 w_pad_0; /* 0x00 */ + u32 w_status; /* 0x0c */ + u32 w_pad_1; /* 0x08 */ + u32 w_err_upper_addr; /* 0x14 */ + u32 w_pad_2; /* 0x10 */ + u32 w_err_lower_addr; /* 0x1c */ + u32 w_pad_3; /* 0x18 */ + u32 w_control; /* 0x24 */ + u32 w_pad_4; /* 0x20 */ + u32 w_req_timeout; /* 0x2c */ + u32 w_pad_5; /* 0x28 */ + u32 w_intdest_upper_addr; /* 0x34 */ + u32 w_pad_6; /* 0x30 */ + u32 w_intdest_lower_addr; /* 0x3c */ + u32 w_pad_7; /* 0x38 */ + u32 w_err_cmd_word; /* 0x44 */ + u32 w_pad_8; /* 0x40 */ + u32 w_llp_cfg; /* 0x4c */ + u32 w_pad_9; /* 0x48 */ + u32 w_tflush; /* 0x54 */ + u32 w_pad_10; /* 0x50 */ }; /* @@ -63,7 +63,7 @@ struct xwidget_info{ struct xwidget_hwid xwi_hwid; /* Widget Identification */ char xwi_masterxid; /* Hub's Widget Port Number */ void *xwi_hubinfo; /* Hub's provider private info */ - uint64_t *xwi_hub_provider; /* prom provider functions */ + u64 *xwi_hub_provider; /* prom provider functions */ void *xwi_vertex; }; diff --git a/arch/ia64/sn/kernel/bte_error.c b/arch/ia64/sn/kernel/bte_error.c index fcbc748ae43..f1ec1370b3e 100644 --- a/arch/ia64/sn/kernel/bte_error.c +++ b/arch/ia64/sn/kernel/bte_error.c @@ -33,7 +33,7 @@ void bte_error_handler(unsigned long); * Wait until all BTE related CRBs are completed * and then reset the interfaces. */ -void shub1_bte_error_handler(unsigned long _nodepda) +int shub1_bte_error_handler(unsigned long _nodepda) { struct nodepda_s *err_nodepda = (struct nodepda_s *)_nodepda; struct timer_list *recovery_timer = &err_nodepda->bte_recovery_timer; @@ -53,7 +53,7 @@ void shub1_bte_error_handler(unsigned long _nodepda) (err_nodepda->bte_if[1].bh_error == BTE_SUCCESS)) { BTE_PRINTK(("eh:%p:%d Nothing to do.\n", err_nodepda, smp_processor_id())); - return; + return 1; } /* Determine information about our hub */ @@ -81,7 +81,7 @@ void shub1_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())); - return; + return 1; } if (icmr.ii_icmr_fld_s.i_crb_vld != 0) { @@ -99,7 +99,7 @@ void shub1_bte_error_handler(unsigned long _nodepda) BTE_PRINTK(("eh:%p:%d Valid %d, Giving up\n", err_nodepda, smp_processor_id(), i)); - return; + return 1; } } } @@ -124,6 +124,42 @@ void shub1_bte_error_handler(unsigned long _nodepda) REMOTE_HUB_S(nasid, IIO_IBCR, ibcr.ii_ibcr_regval); del_timer(recovery_timer); + return 0; +} + +/* + * Wait until all BTE related CRBs are completed + * and then reset the interfaces. + */ +int shub2_bte_error_handler(unsigned long _nodepda) +{ + struct nodepda_s *err_nodepda = (struct nodepda_s *)_nodepda; + struct timer_list *recovery_timer = &err_nodepda->bte_recovery_timer; + struct bteinfo_s *bte; + nasid_t nasid; + u64 status; + int i; + + nasid = cnodeid_to_nasid(err_nodepda->bte_if[0].bte_cnode); + + /* + * Verify that all the BTEs are complete + */ + for (i = 0; i < BTES_PER_NODE; i++) { + bte = &err_nodepda->bte_if[i]; + status = BTE_LNSTAT_LOAD(bte); + if ((status & IBLS_ERROR) || !(status & IBLS_BUSY)) + continue; + mod_timer(recovery_timer, HZ * 5); + BTE_PRINTK(("eh:%p:%d Marked Giving up\n", err_nodepda, + smp_processor_id())); + return 1; + } + if (ia64_sn_bte_recovery(nasid)) + panic("bte_error_handler(): Fatal BTE Error"); + + del_timer(recovery_timer); + return 0; } /* @@ -135,7 +171,6 @@ 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; @@ -160,12 +195,15 @@ void bte_error_handler(unsigned long _nodepda) } if (is_shub1()) { - shub1_bte_error_handler(_nodepda); + if (shub1_bte_error_handler(_nodepda)) { + spin_unlock_irqrestore(recovery_lock, irq_flags); + return; + } } 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"); + if (shub2_bte_error_handler(_nodepda)) { + spin_unlock_irqrestore(recovery_lock, irq_flags); + return; + } } for (i = 0; i < BTES_PER_NODE; i++) { diff --git a/arch/ia64/sn/kernel/huberror.c b/arch/ia64/sn/kernel/huberror.c index 5c5eb01c50f..56ab6bae00e 100644 --- a/arch/ia64/sn/kernel/huberror.c +++ b/arch/ia64/sn/kernel/huberror.c @@ -32,13 +32,14 @@ static irqreturn_t hub_eint_handler(int irq, void *arg, struct pt_regs *ep) ret_stuff.v0 = 0; hubdev_info = (struct hubdev_info *)arg; nasid = hubdev_info->hdi_nasid; - SAL_CALL_NOLOCK(ret_stuff, SN_SAL_HUB_ERROR_INTERRUPT, + + if (is_shub1()) { + SAL_CALL_NOLOCK(ret_stuff, SN_SAL_HUB_ERROR_INTERRUPT, (u64) nasid, 0, 0, 0, 0, 0, 0); - if ((int)ret_stuff.v0) - panic("hubii_eint_handler(): Fatal TIO Error"); + if ((int)ret_stuff.v0) + panic("hubii_eint_handler(): Fatal TIO Error"); - if (is_shub1()) { if (!(nasid & 1)) /* Not a TIO, handle CRB errors */ (void)hubiio_crb_error_handler(hubdev_info); } else diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c index 318087e35b6..00700f7e683 100644 --- a/arch/ia64/sn/kernel/io_init.c +++ b/arch/ia64/sn/kernel/io_init.c @@ -76,11 +76,12 @@ static struct sn_pcibus_provider sn_pci_default_provider = { }; /* - * Retrieve the DMA Flush List given nasid. This list is needed - * to implement the WAR - Flush DMA data on PIO Reads. + * Retrieve the DMA Flush List given nasid, widget, and device. + * This list is needed to implement the WAR - Flush DMA data on PIO Reads. */ -static inline uint64_t -sal_get_widget_dmaflush_list(u64 nasid, u64 widget_num, u64 address) +static inline u64 +sal_get_device_dmaflush_list(u64 nasid, u64 widget_num, u64 device_num, + u64 address) { struct ia64_sal_retval ret_stuff; @@ -88,17 +89,17 @@ sal_get_widget_dmaflush_list(u64 nasid, u64 widget_num, u64 address) ret_stuff.v0 = 0; SAL_CALL_NOLOCK(ret_stuff, - (u64) SN_SAL_IOIF_GET_WIDGET_DMAFLUSH_LIST, - (u64) nasid, (u64) widget_num, (u64) address, 0, 0, 0, - 0); - return ret_stuff.v0; + (u64) SN_SAL_IOIF_GET_DEVICE_DMAFLUSH_LIST, + (u64) nasid, (u64) widget_num, + (u64) device_num, (u64) address, 0, 0, 0); + return ret_stuff.status; } /* * Retrieve the hub device info structure for the given nasid. */ -static inline uint64_t sal_get_hubdev_info(u64 handle, u64 address) +static inline u64 sal_get_hubdev_info(u64 handle, u64 address) { struct ia64_sal_retval ret_stuff; @@ -114,7 +115,7 @@ static inline uint64_t sal_get_hubdev_info(u64 handle, u64 address) /* * Retrieve the pci bus information given the bus number. */ -static inline uint64_t sal_get_pcibus_info(u64 segment, u64 busnum, u64 address) +static inline u64 sal_get_pcibus_info(u64 segment, u64 busnum, u64 address) { struct ia64_sal_retval ret_stuff; @@ -130,9 +131,9 @@ static inline uint64_t sal_get_pcibus_info(u64 segment, u64 busnum, u64 address) /* * Retrieve the pci device information given the bus and device|function number. */ -static inline uint64_t -sal_get_pcidev_info(u64 segment, u64 bus_number, u64 devfn, u64 pci_dev, - u64 sn_irq_info) +static inline u64 +sal_get_pcidev_info(u64 segment, u64 bus_number, u64 devfn, u64 pci_dev, + u64 sn_irq_info) { struct ia64_sal_retval ret_stuff; ret_stuff.status = 0; @@ -140,7 +141,7 @@ sal_get_pcidev_info(u64 segment, u64 bus_number, u64 devfn, u64 pci_dev, SAL_CALL_NOLOCK(ret_stuff, (u64) SN_SAL_IOIF_GET_PCIDEV_INFO, - (u64) segment, (u64) bus_number, (u64) devfn, + (u64) segment, (u64) bus_number, (u64) devfn, (u64) pci_dev, sn_irq_info, 0, 0); return ret_stuff.v0; @@ -164,18 +165,55 @@ sn_pcidev_info_get(struct pci_dev *dev) return NULL; } +/* Older PROM flush WAR + * + * 01/16/06 -- This war will be in place until a new official PROM is released. + * Additionally note that the struct sn_flush_device_war also has to be + * removed from arch/ia64/sn/include/xtalk/hubdev.h + */ +static u8 war_implemented = 0; + +static void sn_device_fixup_war(u64 nasid, u64 widget, int device, + struct sn_flush_device_common *common) +{ + struct sn_flush_device_war *war_list; + struct sn_flush_device_war *dev_entry; + struct ia64_sal_retval isrv = {0,0,0,0}; + + if (!war_implemented) { + printk(KERN_WARNING "PROM version < 4.50 -- implementing old " + "PROM flush WAR\n"); + war_implemented = 1; + } + + war_list = kzalloc(DEV_PER_WIDGET * sizeof(*war_list), GFP_KERNEL); + if (!war_list) + BUG(); + + SAL_CALL_NOLOCK(isrv, SN_SAL_IOIF_GET_WIDGET_DMAFLUSH_LIST, + nasid, widget, __pa(war_list), 0, 0, 0 ,0); + if (isrv.status) + panic("sn_device_fixup_war failed: %s\n", + ia64_sal_strerror(isrv.status)); + + dev_entry = war_list + device; + memcpy(common,dev_entry, sizeof(*common)); + + kfree(war_list); +} + /* - * sn_fixup_ionodes() - This routine initializes the HUB data strcuture for + * sn_fixup_ionodes() - This routine initializes the HUB data strcuture for * each node in the system. */ static void sn_fixup_ionodes(void) { - - struct sn_flush_device_list *sn_flush_device_list; + struct sn_flush_device_kernel *sn_flush_device_kernel; + struct sn_flush_device_kernel *dev_entry; struct hubdev_info *hubdev; - uint64_t status; - uint64_t nasid; - int i, widget; + u64 status; + u64 nasid; + int i, widget, device; /* * Get SGI Specific HUB chipset information. @@ -186,7 +224,7 @@ static void sn_fixup_ionodes(void) nasid = cnodeid_to_nasid(i); hubdev->max_segment_number = 0xffffffff; hubdev->max_pcibus_number = 0xff; - status = sal_get_hubdev_info(nasid, (uint64_t) __pa(hubdev)); + status = sal_get_hubdev_info(nasid, (u64) __pa(hubdev)); if (status) continue; @@ -213,38 +251,60 @@ static void sn_fixup_ionodes(void) hubdev->hdi_flush_nasid_list.widget_p = kmalloc((HUB_WIDGET_ID_MAX + 1) * - sizeof(struct sn_flush_device_list *), GFP_KERNEL); - + sizeof(struct sn_flush_device_kernel *), + GFP_KERNEL); memset(hubdev->hdi_flush_nasid_list.widget_p, 0x0, (HUB_WIDGET_ID_MAX + 1) * - sizeof(struct sn_flush_device_list *)); + sizeof(struct sn_flush_device_kernel *)); for (widget = 0; widget <= HUB_WIDGET_ID_MAX; widget++) { - sn_flush_device_list = kmalloc(DEV_PER_WIDGET * - sizeof(struct - sn_flush_device_list), - GFP_KERNEL); - memset(sn_flush_device_list, 0x0, + sn_flush_device_kernel = kmalloc(DEV_PER_WIDGET * + sizeof(struct + sn_flush_device_kernel), + GFP_KERNEL); + if (!sn_flush_device_kernel) + BUG(); + memset(sn_flush_device_kernel, 0x0, DEV_PER_WIDGET * - sizeof(struct sn_flush_device_list)); - - status = - sal_get_widget_dmaflush_list(nasid, widget, - (uint64_t) - __pa - (sn_flush_device_list)); - if (status) { - kfree(sn_flush_device_list); - continue; + sizeof(struct sn_flush_device_kernel)); + + dev_entry = sn_flush_device_kernel; + for (device = 0; device < DEV_PER_WIDGET; + device++,dev_entry++) { + dev_entry->common = kmalloc(sizeof(struct + sn_flush_device_common), + GFP_KERNEL); + if (!dev_entry->common) + BUG(); + memset(dev_entry->common, 0x0, sizeof(struct + sn_flush_device_common)); + + status = sal_get_device_dmaflush_list(nasid, + widget, + device, + (u64)(dev_entry->common)); + if (status) { + if (sn_sal_rev() < 0x0450) { + /* shortlived WAR for older + * PROM images + */ + sn_device_fixup_war(nasid, + widget, + device, + dev_entry->common); + } + else + BUG(); + } + + spin_lock_init(&dev_entry->sfdl_flush_lock); } - spin_lock_init(&sn_flush_device_list->sfdl_flush_lock); - hubdev->hdi_flush_nasid_list.widget_p[widget] = - sn_flush_device_list; - } - + if (sn_flush_device_kernel) + hubdev->hdi_flush_nasid_list.widget_p[widget] = + sn_flush_device_kernel; + } } - } /* @@ -256,7 +316,7 @@ static void sn_fixup_ionodes(void) */ static void sn_pci_window_fixup(struct pci_dev *dev, unsigned int count, - int64_t * pci_addrs) + s64 * pci_addrs) { struct pci_controller *controller = PCI_CONTROLLER(dev->bus); unsigned int i; @@ -316,7 +376,7 @@ void sn_pci_fixup_slot(struct pci_dev *dev) struct pci_bus *host_pci_bus; struct pci_dev *host_pci_dev; struct pcidev_info *pcidev_info; - int64_t pci_addrs[PCI_ROM_RESOURCE + 1]; + s64 pci_addrs[PCI_ROM_RESOURCE + 1]; struct sn_irq_info *sn_irq_info; unsigned long size; unsigned int bus_no, devfn; diff --git a/arch/ia64/sn/kernel/irq.c b/arch/ia64/sn/kernel/irq.c index 01d18b7b5bb..ec37084bdc1 100644 --- a/arch/ia64/sn/kernel/irq.c +++ b/arch/ia64/sn/kernel/irq.c @@ -28,7 +28,7 @@ extern int sn_ioif_inited; static struct list_head **sn_irq_lh; static spinlock_t sn_irq_info_lock = SPIN_LOCK_UNLOCKED; /* non-IRQ lock */ -static inline uint64_t sn_intr_alloc(nasid_t local_nasid, int local_widget, +static inline u64 sn_intr_alloc(nasid_t local_nasid, int local_widget, u64 sn_irq_info, int req_irq, nasid_t req_nasid, int req_slice) @@ -123,7 +123,7 @@ static void sn_set_affinity_irq(unsigned int irq, cpumask_t mask) list_for_each_entry_safe(sn_irq_info, sn_irq_info_safe, sn_irq_lh[irq], list) { - uint64_t bridge; + u64 bridge; int local_widget, status; nasid_t local_nasid; struct sn_irq_info *new_irq_info; @@ -134,7 +134,7 @@ static void sn_set_affinity_irq(unsigned int irq, cpumask_t mask) break; memcpy(new_irq_info, sn_irq_info, sizeof(struct sn_irq_info)); - bridge = (uint64_t) new_irq_info->irq_bridge; + bridge = (u64) new_irq_info->irq_bridge; if (!bridge) { kfree(new_irq_info); break; /* irq is not a device interrupt */ @@ -349,10 +349,10 @@ static void force_interrupt(int irq) */ static void sn_check_intr(int irq, struct sn_irq_info *sn_irq_info) { - uint64_t regval; + u64 regval; int irr_reg_num; int irr_bit; - uint64_t irr_reg; + u64 irr_reg; struct pcidev_info *pcidev_info; struct pcibus_info *pcibus_info; diff --git a/arch/ia64/sn/kernel/mca.c b/arch/ia64/sn/kernel/mca.c index 6546db6abdb..9ab684d1bb5 100644 --- a/arch/ia64/sn/kernel/mca.c +++ b/arch/ia64/sn/kernel/mca.c @@ -10,6 +10,7 @@ #include <linux/kernel.h> #include <linux/timer.h> #include <linux/vmalloc.h> +#include <linux/mutex.h> #include <asm/mca.h> #include <asm/sal.h> #include <asm/sn/sn_sal.h> @@ -27,7 +28,7 @@ void sn_init_cpei_timer(void); /* Printing oemdata from mca uses data that is not passed through SAL, it is * global. Only one user at a time. */ -static DECLARE_MUTEX(sn_oemdata_mutex); +static DEFINE_MUTEX(sn_oemdata_mutex); static u8 **sn_oemdata; static u64 *sn_oemdata_size, sn_oemdata_bufsize; @@ -89,7 +90,7 @@ static int sn_platform_plat_specific_err_print(const u8 * sect_header, u8 ** oemdata, u64 * oemdata_size) { - down(&sn_oemdata_mutex); + mutex_lock(&sn_oemdata_mutex); sn_oemdata = oemdata; sn_oemdata_size = oemdata_size; sn_oemdata_bufsize = 0; @@ -107,7 +108,7 @@ sn_platform_plat_specific_err_print(const u8 * sect_header, u8 ** oemdata, *sn_oemdata_size = 0; ia64_sn_plat_specific_err_print(print_hook, (char *)sect_header); } - up(&sn_oemdata_mutex); + mutex_unlock(&sn_oemdata_mutex); return 0; } diff --git a/arch/ia64/sn/kernel/sn2/sn2_smp.c b/arch/ia64/sn/kernel/sn2/sn2_smp.c index 5d54f5f4e92..471bbaa65d1 100644 --- a/arch/ia64/sn/kernel/sn2/sn2_smp.c +++ b/arch/ia64/sn/kernel/sn2/sn2_smp.c @@ -202,7 +202,7 @@ sn2_global_tlb_purge(struct mm_struct *mm, unsigned long start, unsigned long end, unsigned long nbits) { int i, opt, shub1, cnode, mynasid, cpu, lcpu = 0, nasid, flushed = 0; - int mymm = (mm == current->active_mm); + int mymm = (mm == current->active_mm && current->mm); volatile unsigned long *ptc0, *ptc1; unsigned long itc, itc2, flags, data0 = 0, data1 = 0, rr_value; short nasids[MAX_NUMNODES], nix; diff --git a/arch/ia64/sn/kernel/tiocx.c b/arch/ia64/sn/kernel/tiocx.c index 0d8592a745a..d263d3e8fbb 100644 --- a/arch/ia64/sn/kernel/tiocx.c +++ b/arch/ia64/sn/kernel/tiocx.c @@ -11,6 +11,7 @@ #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/proc_fs.h> +#include <linux/capability.h> #include <linux/device.h> #include <linux/delay.h> #include <asm/system.h> @@ -65,7 +66,7 @@ static int tiocx_match(struct device *dev, struct device_driver *drv) } -static int tiocx_hotplug(struct device *dev, char **envp, int num_envp, +static int tiocx_uevent(struct device *dev, char **envp, int num_envp, char *buffer, int buffer_size) { return -ENODEV; @@ -76,12 +77,6 @@ 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 @@ -148,6 +143,14 @@ static int cx_driver_remove(struct device *dev) return 0; } +struct bus_type tiocx_bus_type = { + .name = "tiocx", + .match = tiocx_match, + .uevent = tiocx_uevent, + .probe = cx_device_probe, + .remove = cx_driver_remove, +}; + /** * cx_driver_register - Register the driver. * @cx_driver: driver table (cx_drv struct) from driver @@ -161,8 +164,6 @@ 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); } @@ -244,7 +245,7 @@ static int cx_device_reload(struct cx_dev *cx_dev) cx_dev->bt); } -static inline uint64_t tiocx_intr_alloc(nasid_t nasid, int widget, +static inline u64 tiocx_intr_alloc(nasid_t nasid, int widget, u64 sn_irq_info, int req_irq, nasid_t req_nasid, int req_slice) @@ -301,7 +302,7 @@ struct sn_irq_info *tiocx_irq_alloc(nasid_t nasid, int widget, int irq, void tiocx_irq_free(struct sn_irq_info *sn_irq_info) { - uint64_t bridge = (uint64_t) sn_irq_info->irq_bridge; + u64 bridge = (u64) sn_irq_info->irq_bridge; nasid_t nasid = NASID_GET(bridge); int widget; @@ -312,12 +313,12 @@ void tiocx_irq_free(struct sn_irq_info *sn_irq_info) } } -uint64_t tiocx_dma_addr(uint64_t addr) +u64 tiocx_dma_addr(u64 addr) { return PHYS_TO_TIODMA(addr); } -uint64_t tiocx_swin_base(int nasid) +u64 tiocx_swin_base(int nasid) { return TIO_SWIN_BASE(nasid, TIOCX_CORELET); } @@ -334,8 +335,8 @@ EXPORT_SYMBOL(tiocx_swin_base); static void tio_conveyor_set(nasid_t nasid, int enable_flag) { - uint64_t ice_frz; - uint64_t disable_cb = (1ull << 61); + u64 ice_frz; + u64 disable_cb = (1ull << 61); if (!(nasid & 1)) return; @@ -387,7 +388,7 @@ static int is_fpga_tio(int nasid, int *bt) static int bitstream_loaded(nasid_t nasid) { - uint64_t cx_credits; + u64 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; @@ -403,14 +404,14 @@ static int tiocx_reload(struct cx_dev *cx_dev) nasid_t nasid = cx_dev->cx_id.nasid; if (bitstream_loaded(nasid)) { - uint64_t cx_id; + u64 cx_id; int rv; rv = ia64_sn_sysctl_tio_clock_reset(nasid); if (rv) { printk(KERN_ALERT "CX port JTAG reset failed.\n"); } else { - cx_id = *(volatile uint64_t *) + cx_id = *(volatile u64 *) (TIO_SWIN_BASE(nasid, TIOCX_CORELET) + WIDGET_ID); part_num = XWIDGET_PART_NUM(cx_id); diff --git a/arch/ia64/sn/kernel/xp_main.c b/arch/ia64/sn/kernel/xp_main.c index 3be52a34c80..b7ea46645e1 100644 --- a/arch/ia64/sn/kernel/xp_main.c +++ b/arch/ia64/sn/kernel/xp_main.c @@ -19,6 +19,7 @@ #include <linux/kernel.h> #include <linux/interrupt.h> #include <linux/module.h> +#include <linux/mutex.h> #include <asm/sn/intr.h> #include <asm/sn/sn_sal.h> #include <asm/sn/xp.h> @@ -136,13 +137,13 @@ xpc_connect(int ch_number, xpc_channel_func func, void *key, u16 payload_size, registration = &xpc_registrations[ch_number]; - if (down_interruptible(®istration->sema) != 0) { + if (mutex_lock_interruptible(®istration->mutex) != 0) { return xpcInterrupted; } /* if XPC_CHANNEL_REGISTERED(ch_number) */ if (registration->func != NULL) { - up(®istration->sema); + mutex_unlock(®istration->mutex); return xpcAlreadyRegistered; } @@ -154,7 +155,7 @@ xpc_connect(int ch_number, xpc_channel_func func, void *key, u16 payload_size, registration->key = key; registration->func = func; - up(®istration->sema); + mutex_unlock(®istration->mutex); xpc_interface.connect(ch_number); @@ -190,11 +191,11 @@ xpc_disconnect(int ch_number) * figured XPC's users will just turn around and call xpc_disconnect() * again anyways, so we might as well wait, if need be. */ - down(®istration->sema); + mutex_lock(®istration->mutex); /* if !XPC_CHANNEL_REGISTERED(ch_number) */ if (registration->func == NULL) { - up(®istration->sema); + mutex_unlock(®istration->mutex); return; } @@ -208,7 +209,7 @@ xpc_disconnect(int ch_number) xpc_interface.disconnect(ch_number); - up(®istration->sema); + mutex_unlock(®istration->mutex); return; } @@ -250,9 +251,9 @@ xp_init(void) xp_nofault_PIOR_target = SH1_IPI_ACCESS; } - /* initialize the connection registration semaphores */ + /* initialize the connection registration mutex */ for (ch_number = 0; ch_number < XPC_NCHANNELS; ch_number++) { - sema_init(&xpc_registrations[ch_number].sema, 1); /* mutex */ + mutex_init(&xpc_registrations[ch_number].mutex); } return 0; diff --git a/arch/ia64/sn/kernel/xpc.h b/arch/ia64/sn/kernel/xpc.h deleted file mode 100644 index 5483a9f227d..00000000000 --- a/arch/ia64/sn/kernel/xpc.h +++ /dev/null @@ -1,1273 +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) 2004-2005 Silicon Graphics, Inc. All Rights Reserved. - */ - - -/* - * Cross Partition Communication (XPC) structures and macros. - */ - -#ifndef _IA64_SN_KERNEL_XPC_H -#define _IA64_SN_KERNEL_XPC_H - - -#include <linux/config.h> -#include <linux/interrupt.h> -#include <linux/sysctl.h> -#include <linux/device.h> -#include <asm/pgtable.h> -#include <asm/processor.h> -#include <asm/sn/bte.h> -#include <asm/sn/clksupport.h> -#include <asm/sn/addrs.h> -#include <asm/sn/mspec.h> -#include <asm/sn/shub_mmr.h> -#include <asm/sn/xp.h> - - -/* - * XPC Version numbers consist of a major and minor number. XPC can always - * talk to versions with same major #, and never talk to versions with a - * different major #. - */ -#define _XPC_VERSION(_maj, _min) (((_maj) << 4) | ((_min) & 0xf)) -#define XPC_VERSION_MAJOR(_v) ((_v) >> 4) -#define XPC_VERSION_MINOR(_v) ((_v) & 0xf) - - -/* - * The next macros define word or bit representations for given - * C-brick nasid in either the SAL provided bit array representing - * nasids in the partition/machine or the AMO_t array used for - * inter-partition initiation communications. - * - * For SN2 machines, C-Bricks are alway even numbered NASIDs. As - * such, some space will be saved by insisting that nasid information - * passed from SAL always be packed for C-Bricks and the - * cross-partition interrupts use the same packing scheme. - */ -#define XPC_NASID_W_INDEX(_n) (((_n) / 64) / 2) -#define XPC_NASID_B_INDEX(_n) (((_n) / 2) & (64 - 1)) -#define XPC_NASID_IN_ARRAY(_n, _p) ((_p)[XPC_NASID_W_INDEX(_n)] & \ - (1UL << XPC_NASID_B_INDEX(_n))) -#define XPC_NASID_FROM_W_B(_w, _b) (((_w) * 64 + (_b)) * 2) - -#define XPC_HB_DEFAULT_INTERVAL 5 /* incr HB every x secs */ -#define XPC_HB_CHECK_DEFAULT_INTERVAL 20 /* check HB every x secs */ - -/* define the process name of HB checker and the CPU it is pinned to */ -#define XPC_HB_CHECK_THREAD_NAME "xpc_hb" -#define XPC_HB_CHECK_CPU 0 - -/* define the process name of the discovery thread */ -#define XPC_DISCOVERY_THREAD_NAME "xpc_discovery" - - -/* - * the reserved page - * - * SAL reserves one page of memory per partition for XPC. Though a full page - * in length (16384 bytes), its starting address is not page aligned, but it - * is cacheline aligned. The reserved page consists of the following: - * - * reserved page header - * - * The first cacheline of the reserved page contains the header - * (struct xpc_rsvd_page). Before SAL initialization has completed, - * SAL has set up the following fields of the reserved page header: - * SAL_signature, SAL_version, partid, and nasids_size. The other - * fields are set up by XPC. (xpc_rsvd_page points to the local - * partition's reserved page.) - * - * part_nasids mask - * mach_nasids mask - * - * SAL also sets up two bitmaps (or masks), one that reflects the actual - * nasids in this partition (part_nasids), and the other that reflects - * the actual nasids in the entire machine (mach_nasids). We're only - * interested in the even numbered nasids (which contain the processors - * and/or memory), so we only need half as many bits to represent the - * nasids. The part_nasids mask is located starting at the first cacheline - * following the reserved page header. The mach_nasids mask follows right - * after the part_nasids mask. The size in bytes of each mask is reflected - * by the reserved page header field 'nasids_size'. (Local partition's - * mask pointers are xpc_part_nasids and xpc_mach_nasids.) - * - * vars - * vars part - * - * Immediately following the mach_nasids mask are the XPC variables - * required by other partitions. First are those that are generic to all - * partitions (vars), followed on the next available cacheline by those - * which are partition specific (vars part). These are setup by XPC. - * (Local partition's vars pointers are xpc_vars and xpc_vars_part.) - * - * Note: Until vars_pa is set, the partition XPC code has not been initialized. - */ -struct xpc_rsvd_page { - u64 SAL_signature; /* SAL: unique signature */ - u64 SAL_version; /* SAL: version */ - u8 partid; /* SAL: partition ID */ - u8 version; - u8 pad1[6]; /* align to next u64 in cacheline */ - volatile u64 vars_pa; - struct timespec stamp; /* time when reserved page was setup by XPC */ - u64 pad2[9]; /* align to last u64 in cacheline */ - u64 nasids_size; /* SAL: size of each nasid mask in bytes */ -}; - -#define XPC_RP_VERSION _XPC_VERSION(1,1) /* version 1.1 of the reserved page */ - -#define XPC_SUPPORTS_RP_STAMP(_version) \ - (_version >= _XPC_VERSION(1,1)) - -/* - * compare stamps - the return value is: - * - * < 0, if stamp1 < stamp2 - * = 0, if stamp1 == stamp2 - * > 0, if stamp1 > stamp2 - */ -static inline int -xpc_compare_stamps(struct timespec *stamp1, struct timespec *stamp2) -{ - int ret; - - - if ((ret = stamp1->tv_sec - stamp2->tv_sec) == 0) { - ret = stamp1->tv_nsec - stamp2->tv_nsec; - } - return ret; -} - - -/* - * Define the structures by which XPC variables can be exported to other - * partitions. (There are two: struct xpc_vars and struct xpc_vars_part) - */ - -/* - * The following structure describes the partition generic variables - * needed by other partitions in order to properly initialize. - * - * struct xpc_vars version number also applies to struct xpc_vars_part. - * Changes to either structure and/or related functionality should be - * reflected by incrementing either the major or minor version numbers - * of struct xpc_vars. - */ -struct xpc_vars { - u8 version; - u64 heartbeat; - u64 heartbeating_to_mask; - u64 heartbeat_offline; /* if 0, heartbeat should be changing */ - int act_nasid; - int act_phys_cpuid; - u64 vars_part_pa; - u64 amos_page_pa; /* paddr of page of AMOs from MSPEC driver */ - AMO_t *amos_page; /* vaddr of page of AMOs from MSPEC driver */ -}; - -#define XPC_V_VERSION _XPC_VERSION(3,1) /* version 3.1 of the cross vars */ - -#define XPC_SUPPORTS_DISENGAGE_REQUEST(_version) \ - (_version >= _XPC_VERSION(3,1)) - - -static inline int -xpc_hb_allowed(partid_t partid, struct xpc_vars *vars) -{ - return ((vars->heartbeating_to_mask & (1UL << partid)) != 0); -} - -static inline void -xpc_allow_hb(partid_t partid, struct xpc_vars *vars) -{ - u64 old_mask, new_mask; - - do { - old_mask = vars->heartbeating_to_mask; - new_mask = (old_mask | (1UL << partid)); - } while (cmpxchg(&vars->heartbeating_to_mask, old_mask, new_mask) != - old_mask); -} - -static inline void -xpc_disallow_hb(partid_t partid, struct xpc_vars *vars) -{ - u64 old_mask, new_mask; - - do { - old_mask = vars->heartbeating_to_mask; - new_mask = (old_mask & ~(1UL << partid)); - } while (cmpxchg(&vars->heartbeating_to_mask, old_mask, new_mask) != - old_mask); -} - - -/* - * The AMOs page consists of a number of AMO variables which are divided into - * four groups, The first two groups are used to identify an IRQ's sender. - * These two groups consist of 64 and 128 AMO variables respectively. The last - * two groups, consisting of just one AMO variable each, are used to identify - * the remote partitions that are currently engaged (from the viewpoint of - * the XPC running on the remote partition). - */ -#define XPC_NOTIFY_IRQ_AMOS 0 -#define XPC_ACTIVATE_IRQ_AMOS (XPC_NOTIFY_IRQ_AMOS + XP_MAX_PARTITIONS) -#define XPC_ENGAGED_PARTITIONS_AMO (XPC_ACTIVATE_IRQ_AMOS + XP_NASID_MASK_WORDS) -#define XPC_DISENGAGE_REQUEST_AMO (XPC_ENGAGED_PARTITIONS_AMO + 1) - - -/* - * The following structure describes the per partition specific variables. - * - * An array of these structures, one per partition, will be defined. As a - * partition becomes active XPC will copy the array entry corresponding to - * itself from that partition. It is desirable that the size of this - * structure evenly divide into a cacheline, such that none of the entries - * in this array crosses a cacheline boundary. As it is now, each entry - * occupies half a cacheline. - */ -struct xpc_vars_part { - volatile u64 magic; - - u64 openclose_args_pa; /* physical address of open and close args */ - u64 GPs_pa; /* physical address of Get/Put values */ - - u64 IPI_amo_pa; /* physical address of IPI AMO_t structure */ - int IPI_nasid; /* nasid of where to send IPIs */ - int IPI_phys_cpuid; /* physical CPU ID of where to send IPIs */ - - u8 nchannels; /* #of defined channels supported */ - - u8 reserved[23]; /* pad to a full 64 bytes */ -}; - -/* - * The vars_part MAGIC numbers play a part in the first contact protocol. - * - * MAGIC1 indicates that the per partition specific variables for a remote - * partition have been initialized by this partition. - * - * MAGIC2 indicates that this partition has pulled the remote partititions - * per partition variables that pertain to this partition. - */ -#define XPC_VP_MAGIC1 0x0053524156435058L /* 'XPCVARS\0'L (little endian) */ -#define XPC_VP_MAGIC2 0x0073726176435058L /* 'XPCvars\0'L (little endian) */ - - -/* the reserved page sizes and offsets */ - -#define XPC_RP_HEADER_SIZE L1_CACHE_ALIGN(sizeof(struct xpc_rsvd_page)) -#define XPC_RP_VARS_SIZE L1_CACHE_ALIGN(sizeof(struct xpc_vars)) - -#define XPC_RP_PART_NASIDS(_rp) (u64 *) ((u8 *) _rp + XPC_RP_HEADER_SIZE) -#define XPC_RP_MACH_NASIDS(_rp) (XPC_RP_PART_NASIDS(_rp) + xp_nasid_mask_words) -#define XPC_RP_VARS(_rp) ((struct xpc_vars *) XPC_RP_MACH_NASIDS(_rp) + xp_nasid_mask_words) -#define XPC_RP_VARS_PART(_rp) (struct xpc_vars_part *) ((u8 *) XPC_RP_VARS(rp) + XPC_RP_VARS_SIZE) - - -/* - * Functions registered by add_timer() or called by kernel_thread() only - * allow for a single 64-bit argument. The following macros can be used to - * pack and unpack two (32-bit, 16-bit or 8-bit) arguments into or out from - * the passed argument. - */ -#define XPC_PACK_ARGS(_arg1, _arg2) \ - ((((u64) _arg1) & 0xffffffff) | \ - ((((u64) _arg2) & 0xffffffff) << 32)) - -#define XPC_UNPACK_ARG1(_args) (((u64) _args) & 0xffffffff) -#define XPC_UNPACK_ARG2(_args) ((((u64) _args) >> 32) & 0xffffffff) - - - -/* - * Define a Get/Put value pair (pointers) used with a message queue. - */ -struct xpc_gp { - volatile s64 get; /* Get value */ - volatile s64 put; /* Put value */ -}; - -#define XPC_GP_SIZE \ - L1_CACHE_ALIGN(sizeof(struct xpc_gp) * XPC_NCHANNELS) - - - -/* - * Define a structure that contains arguments associated with opening and - * closing a channel. - */ -struct xpc_openclose_args { - u16 reason; /* reason why channel is closing */ - u16 msg_size; /* sizeof each message entry */ - u16 remote_nentries; /* #of message entries in remote msg queue */ - u16 local_nentries; /* #of message entries in local msg queue */ - u64 local_msgqueue_pa; /* physical address of local message queue */ -}; - -#define XPC_OPENCLOSE_ARGS_SIZE \ - L1_CACHE_ALIGN(sizeof(struct xpc_openclose_args) * XPC_NCHANNELS) - - - -/* struct xpc_msg flags */ - -#define XPC_M_DONE 0x01 /* msg has been received/consumed */ -#define XPC_M_READY 0x02 /* msg is ready to be sent */ -#define XPC_M_INTERRUPT 0x04 /* send interrupt when msg consumed */ - - -#define XPC_MSG_ADDRESS(_payload) \ - ((struct xpc_msg *)((u8 *)(_payload) - XPC_MSG_PAYLOAD_OFFSET)) - - - -/* - * Defines notify entry. - * - * This is used to notify a message's sender that their message was received - * and consumed by the intended recipient. - */ -struct xpc_notify { - struct semaphore sema; /* notify semaphore */ - volatile u8 type; /* type of notification */ - - /* the following two fields are only used if type == XPC_N_CALL */ - xpc_notify_func func; /* user's notify function */ - void *key; /* pointer to user's key */ -}; - -/* struct xpc_notify type of notification */ - -#define XPC_N_CALL 0x01 /* notify function provided by user */ - - - -/* - * Define the structure that manages all the stuff required by a channel. In - * particular, they are used to manage the messages sent across the channel. - * - * This structure is private to a partition, and is NOT shared across the - * partition boundary. - * - * There is an array of these structures for each remote partition. It is - * allocated at the time a partition becomes active. The array contains one - * of these structures for each potential channel connection to that partition. - * - * Each of these structures manages two message queues (circular buffers). - * They are allocated at the time a channel connection is made. One of - * these message queues (local_msgqueue) holds the locally created messages - * that are destined for the remote partition. The other of these message - * queues (remote_msgqueue) is a locally cached copy of the remote partition's - * own local_msgqueue. - * - * The following is a description of the Get/Put pointers used to manage these - * two message queues. Consider the local_msgqueue to be on one partition - * and the remote_msgqueue to be its cached copy on another partition. A - * description of what each of the lettered areas contains is included. - * - * - * local_msgqueue remote_msgqueue - * - * |/////////| |/////////| - * w_remote_GP.get --> +---------+ |/////////| - * | F | |/////////| - * remote_GP.get --> +---------+ +---------+ <-- local_GP->get - * | | | | - * | | | E | - * | | | | - * | | +---------+ <-- w_local_GP.get - * | B | |/////////| - * | | |////D////| - * | | |/////////| - * | | +---------+ <-- w_remote_GP.put - * | | |////C////| - * local_GP->put --> +---------+ +---------+ <-- remote_GP.put - * | | |/////////| - * | A | |/////////| - * | | |/////////| - * w_local_GP.put --> +---------+ |/////////| - * |/////////| |/////////| - * - * - * ( remote_GP.[get|put] are cached copies of the remote - * partition's local_GP->[get|put], and thus their values can - * lag behind their counterparts on the remote partition. ) - * - * - * A - Messages that have been allocated, but have not yet been sent to the - * remote partition. - * - * B - Messages that have been sent, but have not yet been acknowledged by the - * remote partition as having been received. - * - * C - Area that needs to be prepared for the copying of sent messages, by - * the clearing of the message flags of any previously received messages. - * - * D - Area into which sent messages are to be copied from the remote - * partition's local_msgqueue and then delivered to their intended - * recipients. [ To allow for a multi-message copy, another pointer - * (next_msg_to_pull) has been added to keep track of the next message - * number needing to be copied (pulled). It chases after w_remote_GP.put. - * Any messages lying between w_local_GP.get and next_msg_to_pull have - * been copied and are ready to be delivered. ] - * - * E - Messages that have been copied and delivered, but have not yet been - * acknowledged by the recipient as having been received. - * - * F - Messages that have been acknowledged, but XPC has not yet notified the - * sender that the message was received by its intended recipient. - * This is also an area that needs to be prepared for the allocating of - * new messages, by the clearing of the message flags of the acknowledged - * messages. - */ -struct xpc_channel { - partid_t partid; /* ID of remote partition connected */ - spinlock_t lock; /* lock for updating this structure */ - u32 flags; /* general flags */ - - enum xpc_retval reason; /* reason why channel is disconnect'g */ - int reason_line; /* line# disconnect initiated from */ - - u16 number; /* channel # */ - - u16 msg_size; /* sizeof each msg entry */ - u16 local_nentries; /* #of msg entries in local msg queue */ - u16 remote_nentries; /* #of msg entries in remote msg queue*/ - - void *local_msgqueue_base; /* base address of kmalloc'd space */ - struct xpc_msg *local_msgqueue; /* local message queue */ - void *remote_msgqueue_base; /* base address of kmalloc'd space */ - struct xpc_msg *remote_msgqueue;/* cached copy of remote partition's */ - /* local message queue */ - u64 remote_msgqueue_pa; /* phys addr of remote partition's */ - /* local message queue */ - - atomic_t references; /* #of external references to queues */ - - atomic_t n_on_msg_allocate_wq; /* #on msg allocation wait queue */ - wait_queue_head_t msg_allocate_wq; /* msg allocation wait queue */ - - u8 delayed_IPI_flags; /* IPI flags received, but delayed */ - /* action until channel disconnected */ - - /* queue of msg senders who want to be notified when msg received */ - - atomic_t n_to_notify; /* #of msg senders to notify */ - struct xpc_notify *notify_queue;/* notify queue for messages sent */ - - xpc_channel_func func; /* user's channel function */ - void *key; /* pointer to user's key */ - - struct semaphore msg_to_pull_sema; /* next msg to pull serialization */ - struct semaphore wdisconnect_sema; /* wait for channel disconnect */ - - struct xpc_openclose_args *local_openclose_args; /* args passed on */ - /* opening or closing of channel */ - - /* various flavors of local and remote Get/Put values */ - - struct xpc_gp *local_GP; /* local Get/Put values */ - struct xpc_gp remote_GP; /* remote Get/Put values */ - struct xpc_gp w_local_GP; /* working local Get/Put values */ - struct xpc_gp w_remote_GP; /* working remote Get/Put values */ - s64 next_msg_to_pull; /* Put value of next msg to pull */ - - /* kthread management related fields */ - -// >>> rethink having kthreads_assigned_limit and kthreads_idle_limit; perhaps -// >>> allow the assigned limit be unbounded and let the idle limit be dynamic -// >>> dependent on activity over the last interval of time - atomic_t kthreads_assigned; /* #of kthreads assigned to channel */ - u32 kthreads_assigned_limit; /* limit on #of kthreads assigned */ - atomic_t kthreads_idle; /* #of kthreads idle waiting for work */ - u32 kthreads_idle_limit; /* limit on #of kthreads idle */ - atomic_t kthreads_active; /* #of kthreads actively working */ - // >>> following field is temporary - u32 kthreads_created; /* total #of kthreads created */ - - wait_queue_head_t idle_wq; /* idle kthread wait queue */ - -} ____cacheline_aligned; - - -/* struct xpc_channel flags */ - -#define XPC_C_WASCONNECTED 0x00000001 /* channel was connected */ - -#define XPC_C_ROPENREPLY 0x00000002 /* remote open channel reply */ -#define XPC_C_OPENREPLY 0x00000004 /* local open channel reply */ -#define XPC_C_ROPENREQUEST 0x00000008 /* remote open channel request */ -#define XPC_C_OPENREQUEST 0x00000010 /* local open channel request */ - -#define XPC_C_SETUP 0x00000020 /* channel's msgqueues are alloc'd */ -#define XPC_C_CONNECTCALLOUT 0x00000040 /* channel connected callout made */ -#define XPC_C_CONNECTED 0x00000080 /* local channel is connected */ -#define XPC_C_CONNECTING 0x00000100 /* channel is being connected */ - -#define XPC_C_RCLOSEREPLY 0x00000200 /* remote close channel reply */ -#define XPC_C_CLOSEREPLY 0x00000400 /* local close channel reply */ -#define XPC_C_RCLOSEREQUEST 0x00000800 /* remote close channel request */ -#define XPC_C_CLOSEREQUEST 0x00001000 /* local close channel request */ - -#define XPC_C_DISCONNECTED 0x00002000 /* channel is disconnected */ -#define XPC_C_DISCONNECTING 0x00004000 /* channel is being disconnected */ -#define XPC_C_DISCONNECTCALLOUT 0x00008000 /* chan disconnected callout made */ -#define XPC_C_WDISCONNECT 0x00010000 /* waiting for channel disconnect */ - - - -/* - * Manages channels on a partition basis. There is one of these structures - * for each partition (a partition will never utilize the structure that - * represents itself). - */ -struct xpc_partition { - - /* XPC HB infrastructure */ - - u8 remote_rp_version; /* version# of partition's rsvd pg */ - struct timespec remote_rp_stamp;/* time when rsvd pg was initialized */ - u64 remote_rp_pa; /* phys addr of partition's rsvd pg */ - u64 remote_vars_pa; /* phys addr of partition's vars */ - u64 remote_vars_part_pa; /* phys addr of partition's vars part */ - u64 last_heartbeat; /* HB at last read */ - u64 remote_amos_page_pa; /* phys addr of partition's amos page */ - int remote_act_nasid; /* active part's act/deact nasid */ - int remote_act_phys_cpuid; /* active part's act/deact phys cpuid */ - u32 act_IRQ_rcvd; /* IRQs since activation */ - spinlock_t act_lock; /* protect updating of act_state */ - u8 act_state; /* from XPC HB viewpoint */ - u8 remote_vars_version; /* version# of partition's vars */ - enum xpc_retval reason; /* reason partition is deactivating */ - int reason_line; /* line# deactivation initiated from */ - int reactivate_nasid; /* nasid in partition to reactivate */ - - unsigned long disengage_request_timeout; /* timeout in jiffies */ - struct timer_list disengage_request_timer; - - - /* XPC infrastructure referencing and teardown control */ - - volatile u8 setup_state; /* infrastructure setup state */ - wait_queue_head_t teardown_wq; /* kthread waiting to teardown infra */ - atomic_t references; /* #of references to infrastructure */ - - - /* - * NONE OF THE PRECEDING FIELDS OF THIS STRUCTURE WILL BE CLEARED WHEN - * XPC SETS UP THE NECESSARY INFRASTRUCTURE TO SUPPORT CROSS PARTITION - * COMMUNICATION. ALL OF THE FOLLOWING FIELDS WILL BE CLEARED. (THE - * 'nchannels' FIELD MUST BE THE FIRST OF THE FIELDS TO BE CLEARED.) - */ - - - u8 nchannels; /* #of defined channels supported */ - atomic_t nchannels_active; /* #of channels that are not DISCONNECTED */ - atomic_t nchannels_engaged;/* #of channels engaged with remote part */ - struct xpc_channel *channels;/* array of channel structures */ - - void *local_GPs_base; /* base address of kmalloc'd space */ - struct xpc_gp *local_GPs; /* local Get/Put values */ - void *remote_GPs_base; /* base address of kmalloc'd space */ - struct xpc_gp *remote_GPs;/* copy of remote partition's local Get/Put */ - /* values */ - u64 remote_GPs_pa; /* phys address of remote partition's local */ - /* Get/Put values */ - - - /* fields used to pass args when opening or closing a channel */ - - void *local_openclose_args_base; /* base address of kmalloc'd space */ - struct xpc_openclose_args *local_openclose_args; /* local's args */ - void *remote_openclose_args_base; /* base address of kmalloc'd space */ - struct xpc_openclose_args *remote_openclose_args; /* copy of remote's */ - /* args */ - u64 remote_openclose_args_pa; /* phys addr of remote's args */ - - - /* IPI sending, receiving and handling related fields */ - - int remote_IPI_nasid; /* nasid of where to send IPIs */ - int remote_IPI_phys_cpuid; /* phys CPU ID of where to send IPIs */ - AMO_t *remote_IPI_amo_va; /* address of remote IPI AMO_t structure */ - - AMO_t *local_IPI_amo_va; /* address of IPI AMO_t structure */ - u64 local_IPI_amo; /* IPI amo flags yet to be handled */ - char IPI_owner[8]; /* IPI owner's name */ - struct timer_list dropped_IPI_timer; /* dropped IPI timer */ - - spinlock_t IPI_lock; /* IPI handler lock */ - - - /* channel manager related fields */ - - atomic_t channel_mgr_requests; /* #of requests to activate chan mgr */ - wait_queue_head_t channel_mgr_wq; /* channel mgr's wait queue */ - -} ____cacheline_aligned; - - -/* struct xpc_partition act_state values (for XPC HB) */ - -#define XPC_P_INACTIVE 0x00 /* partition is not active */ -#define XPC_P_ACTIVATION_REQ 0x01 /* created thread to activate */ -#define XPC_P_ACTIVATING 0x02 /* activation thread started */ -#define XPC_P_ACTIVE 0x03 /* xpc_partition_up() was called */ -#define XPC_P_DEACTIVATING 0x04 /* partition deactivation initiated */ - - -#define XPC_DEACTIVATE_PARTITION(_p, _reason) \ - xpc_deactivate_partition(__LINE__, (_p), (_reason)) - - -/* struct xpc_partition setup_state values */ - -#define XPC_P_UNSET 0x00 /* infrastructure was never setup */ -#define XPC_P_SETUP 0x01 /* infrastructure is setup */ -#define XPC_P_WTEARDOWN 0x02 /* waiting to teardown infrastructure */ -#define XPC_P_TORNDOWN 0x03 /* infrastructure is torndown */ - - - -/* - * struct xpc_partition IPI_timer #of seconds to wait before checking for - * dropped IPIs. These occur whenever an IPI amo write doesn't complete until - * after the IPI was received. - */ -#define XPC_P_DROPPED_IPI_WAIT (0.25 * HZ) - - -/* number of seconds to wait for other partitions to disengage */ -#define XPC_DISENGAGE_REQUEST_DEFAULT_TIMELIMIT 90 - -/* interval in seconds to print 'waiting disengagement' messages */ -#define XPC_DISENGAGE_PRINTMSG_INTERVAL 10 - - -#define XPC_PARTID(_p) ((partid_t) ((_p) - &xpc_partitions[0])) - - - -/* found in xp_main.c */ -extern struct xpc_registration xpc_registrations[]; - - -/* found in xpc_main.c */ -extern struct device *xpc_part; -extern struct device *xpc_chan; -extern int xpc_disengage_request_timelimit; -extern irqreturn_t xpc_notify_IRQ_handler(int, void *, struct pt_regs *); -extern void xpc_dropped_IPI_check(struct xpc_partition *); -extern void xpc_activate_partition(struct xpc_partition *); -extern void xpc_activate_kthreads(struct xpc_channel *, int); -extern void xpc_create_kthreads(struct xpc_channel *, int); -extern void xpc_disconnect_wait(int); - - -/* found in xpc_partition.c */ -extern int xpc_exiting; -extern struct xpc_vars *xpc_vars; -extern struct xpc_rsvd_page *xpc_rsvd_page; -extern struct xpc_vars_part *xpc_vars_part; -extern struct xpc_partition xpc_partitions[XP_MAX_PARTITIONS + 1]; -extern char xpc_remote_copy_buffer[]; -extern struct xpc_rsvd_page *xpc_rsvd_page_init(void); -extern void xpc_allow_IPI_ops(void); -extern void xpc_restrict_IPI_ops(void); -extern int xpc_identify_act_IRQ_sender(void); -extern int xpc_partition_disengaged(struct xpc_partition *); -extern enum xpc_retval xpc_mark_partition_active(struct xpc_partition *); -extern void xpc_mark_partition_inactive(struct xpc_partition *); -extern void xpc_discovery(void); -extern void xpc_check_remote_hb(void); -extern void xpc_deactivate_partition(const int, struct xpc_partition *, - enum xpc_retval); -extern enum xpc_retval xpc_initiate_partid_to_nasids(partid_t, void *); - - -/* found in xpc_channel.c */ -extern void xpc_initiate_connect(int); -extern void xpc_initiate_disconnect(int); -extern enum xpc_retval xpc_initiate_allocate(partid_t, int, u32, void **); -extern enum xpc_retval xpc_initiate_send(partid_t, int, void *); -extern enum xpc_retval xpc_initiate_send_notify(partid_t, int, void *, - xpc_notify_func, void *); -extern void xpc_initiate_received(partid_t, int, void *); -extern enum xpc_retval xpc_setup_infrastructure(struct xpc_partition *); -extern enum xpc_retval xpc_pull_remote_vars_part(struct xpc_partition *); -extern void xpc_process_channel_activity(struct xpc_partition *); -extern void xpc_connected_callout(struct xpc_channel *); -extern void xpc_deliver_msg(struct xpc_channel *); -extern void xpc_disconnect_channel(const int, struct xpc_channel *, - enum xpc_retval, unsigned long *); -extern void xpc_disconnecting_callout(struct xpc_channel *); -extern void xpc_partition_going_down(struct xpc_partition *, enum xpc_retval); -extern void xpc_teardown_infrastructure(struct xpc_partition *); - - - -static inline void -xpc_wakeup_channel_mgr(struct xpc_partition *part) -{ - if (atomic_inc_return(&part->channel_mgr_requests) == 1) { - wake_up(&part->channel_mgr_wq); - } -} - - - -/* - * These next two inlines are used to keep us from tearing down a channel's - * msg queues while a thread may be referencing them. - */ -static inline void -xpc_msgqueue_ref(struct xpc_channel *ch) -{ - atomic_inc(&ch->references); -} - -static inline void -xpc_msgqueue_deref(struct xpc_channel *ch) -{ - s32 refs = atomic_dec_return(&ch->references); - - DBUG_ON(refs < 0); - if (refs == 0) { - xpc_wakeup_channel_mgr(&xpc_partitions[ch->partid]); - } -} - - - -#define XPC_DISCONNECT_CHANNEL(_ch, _reason, _irqflgs) \ - xpc_disconnect_channel(__LINE__, _ch, _reason, _irqflgs) - - -/* - * These two inlines are used to keep us from tearing down a partition's - * setup infrastructure while a thread may be referencing it. - */ -static inline void -xpc_part_deref(struct xpc_partition *part) -{ - s32 refs = atomic_dec_return(&part->references); - - - DBUG_ON(refs < 0); - if (refs == 0 && part->setup_state == XPC_P_WTEARDOWN) { - wake_up(&part->teardown_wq); - } -} - -static inline int -xpc_part_ref(struct xpc_partition *part) -{ - int setup; - - - atomic_inc(&part->references); - setup = (part->setup_state == XPC_P_SETUP); - if (!setup) { - xpc_part_deref(part); - } - return setup; -} - - - -/* - * The following macro is to be used for the setting of the reason and - * reason_line fields in both the struct xpc_channel and struct xpc_partition - * structures. - */ -#define XPC_SET_REASON(_p, _reason, _line) \ - { \ - (_p)->reason = _reason; \ - (_p)->reason_line = _line; \ - } - - - -/* - * This next set of inlines are used to keep track of when a partition is - * potentially engaged in accessing memory belonging to another partition. - */ - -static inline void -xpc_mark_partition_engaged(struct xpc_partition *part) -{ - unsigned long irq_flags; - AMO_t *amo = (AMO_t *) __va(part->remote_amos_page_pa + - (XPC_ENGAGED_PARTITIONS_AMO * sizeof(AMO_t))); - - - local_irq_save(irq_flags); - - /* set bit corresponding to our partid in remote partition's AMO */ - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_OR, - (1UL << sn_partition_id)); - /* - * We must always use the nofault function regardless of whether we - * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we - * didn't, we'd never know that the other partition is down and would - * keep sending IPIs and AMOs to it until the heartbeat times out. - */ - (void) xp_nofault_PIOR((u64 *) GLOBAL_MMR_ADDR(NASID_GET(&amo-> - variable), xp_nofault_PIOR_target)); - - local_irq_restore(irq_flags); -} - -static inline void -xpc_mark_partition_disengaged(struct xpc_partition *part) -{ - unsigned long irq_flags; - AMO_t *amo = (AMO_t *) __va(part->remote_amos_page_pa + - (XPC_ENGAGED_PARTITIONS_AMO * sizeof(AMO_t))); - - - local_irq_save(irq_flags); - - /* clear bit corresponding to our partid in remote partition's AMO */ - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_AND, - ~(1UL << sn_partition_id)); - /* - * We must always use the nofault function regardless of whether we - * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we - * didn't, we'd never know that the other partition is down and would - * keep sending IPIs and AMOs to it until the heartbeat times out. - */ - (void) xp_nofault_PIOR((u64 *) GLOBAL_MMR_ADDR(NASID_GET(&amo-> - variable), xp_nofault_PIOR_target)); - - local_irq_restore(irq_flags); -} - -static inline void -xpc_request_partition_disengage(struct xpc_partition *part) -{ - unsigned long irq_flags; - AMO_t *amo = (AMO_t *) __va(part->remote_amos_page_pa + - (XPC_DISENGAGE_REQUEST_AMO * sizeof(AMO_t))); - - - local_irq_save(irq_flags); - - /* set bit corresponding to our partid in remote partition's AMO */ - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_OR, - (1UL << sn_partition_id)); - /* - * We must always use the nofault function regardless of whether we - * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we - * didn't, we'd never know that the other partition is down and would - * keep sending IPIs and AMOs to it until the heartbeat times out. - */ - (void) xp_nofault_PIOR((u64 *) GLOBAL_MMR_ADDR(NASID_GET(&amo-> - variable), xp_nofault_PIOR_target)); - - local_irq_restore(irq_flags); -} - -static inline void -xpc_cancel_partition_disengage_request(struct xpc_partition *part) -{ - unsigned long irq_flags; - AMO_t *amo = (AMO_t *) __va(part->remote_amos_page_pa + - (XPC_DISENGAGE_REQUEST_AMO * sizeof(AMO_t))); - - - local_irq_save(irq_flags); - - /* clear bit corresponding to our partid in remote partition's AMO */ - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_AND, - ~(1UL << sn_partition_id)); - /* - * We must always use the nofault function regardless of whether we - * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we - * didn't, we'd never know that the other partition is down and would - * keep sending IPIs and AMOs to it until the heartbeat times out. - */ - (void) xp_nofault_PIOR((u64 *) GLOBAL_MMR_ADDR(NASID_GET(&amo-> - variable), xp_nofault_PIOR_target)); - - local_irq_restore(irq_flags); -} - -static inline u64 -xpc_partition_engaged(u64 partid_mask) -{ - AMO_t *amo = xpc_vars->amos_page + XPC_ENGAGED_PARTITIONS_AMO; - - - /* return our partition's AMO variable ANDed with partid_mask */ - return (FETCHOP_LOAD_OP(TO_AMO((u64) &amo->variable), FETCHOP_LOAD) & - partid_mask); -} - -static inline u64 -xpc_partition_disengage_requested(u64 partid_mask) -{ - AMO_t *amo = xpc_vars->amos_page + XPC_DISENGAGE_REQUEST_AMO; - - - /* return our partition's AMO variable ANDed with partid_mask */ - return (FETCHOP_LOAD_OP(TO_AMO((u64) &amo->variable), FETCHOP_LOAD) & - partid_mask); -} - -static inline void -xpc_clear_partition_engaged(u64 partid_mask) -{ - AMO_t *amo = xpc_vars->amos_page + XPC_ENGAGED_PARTITIONS_AMO; - - - /* clear bit(s) based on partid_mask in our partition's AMO */ - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_AND, - ~partid_mask); -} - -static inline void -xpc_clear_partition_disengage_request(u64 partid_mask) -{ - AMO_t *amo = xpc_vars->amos_page + XPC_DISENGAGE_REQUEST_AMO; - - - /* clear bit(s) based on partid_mask in our partition's AMO */ - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_AND, - ~partid_mask); -} - - - -/* - * The following set of macros and inlines are used for the sending and - * receiving of IPIs (also known as IRQs). There are two flavors of IPIs, - * one that is associated with partition activity (SGI_XPC_ACTIVATE) and - * the other that is associated with channel activity (SGI_XPC_NOTIFY). - */ - -static inline u64 -xpc_IPI_receive(AMO_t *amo) -{ - return FETCHOP_LOAD_OP(TO_AMO((u64) &amo->variable), FETCHOP_CLEAR); -} - - -static inline enum xpc_retval -xpc_IPI_send(AMO_t *amo, u64 flag, int nasid, int phys_cpuid, int vector) -{ - int ret = 0; - unsigned long irq_flags; - - - local_irq_save(irq_flags); - - FETCHOP_STORE_OP(TO_AMO((u64) &amo->variable), FETCHOP_OR, flag); - sn_send_IPI_phys(nasid, phys_cpuid, vector, 0); - - /* - * We must always use the nofault function regardless of whether we - * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we - * didn't, we'd never know that the other partition is down and would - * keep sending IPIs and AMOs to it until the heartbeat times out. - */ - ret = xp_nofault_PIOR((u64 *) GLOBAL_MMR_ADDR(NASID_GET(&amo->variable), - xp_nofault_PIOR_target)); - - local_irq_restore(irq_flags); - - return ((ret == 0) ? xpcSuccess : xpcPioReadError); -} - - -/* - * IPIs associated with SGI_XPC_ACTIVATE IRQ. - */ - -/* - * Flag the appropriate AMO variable and send an IPI to the specified node. - */ -static inline void -xpc_activate_IRQ_send(u64 amos_page_pa, int from_nasid, int to_nasid, - int to_phys_cpuid) -{ - int w_index = XPC_NASID_W_INDEX(from_nasid); - int b_index = XPC_NASID_B_INDEX(from_nasid); - AMO_t *amos = (AMO_t *) __va(amos_page_pa + - (XPC_ACTIVATE_IRQ_AMOS * sizeof(AMO_t))); - - - (void) xpc_IPI_send(&amos[w_index], (1UL << b_index), to_nasid, - to_phys_cpuid, SGI_XPC_ACTIVATE); -} - -static inline void -xpc_IPI_send_activate(struct xpc_vars *vars) -{ - xpc_activate_IRQ_send(vars->amos_page_pa, cnodeid_to_nasid(0), - vars->act_nasid, vars->act_phys_cpuid); -} - -static inline void -xpc_IPI_send_activated(struct xpc_partition *part) -{ - xpc_activate_IRQ_send(part->remote_amos_page_pa, cnodeid_to_nasid(0), - part->remote_act_nasid, part->remote_act_phys_cpuid); -} - -static inline void -xpc_IPI_send_reactivate(struct xpc_partition *part) -{ - xpc_activate_IRQ_send(xpc_vars->amos_page_pa, part->reactivate_nasid, - xpc_vars->act_nasid, xpc_vars->act_phys_cpuid); -} - -static inline void -xpc_IPI_send_disengage(struct xpc_partition *part) -{ - xpc_activate_IRQ_send(part->remote_amos_page_pa, cnodeid_to_nasid(0), - part->remote_act_nasid, part->remote_act_phys_cpuid); -} - - -/* - * IPIs associated with SGI_XPC_NOTIFY IRQ. - */ - -/* - * Send an IPI to the remote partition that is associated with the - * specified channel. - */ -#define XPC_NOTIFY_IRQ_SEND(_ch, _ipi_f, _irq_f) \ - xpc_notify_IRQ_send(_ch, _ipi_f, #_ipi_f, _irq_f) - -static inline void -xpc_notify_IRQ_send(struct xpc_channel *ch, u8 ipi_flag, char *ipi_flag_string, - unsigned long *irq_flags) -{ - struct xpc_partition *part = &xpc_partitions[ch->partid]; - enum xpc_retval ret; - - - if (likely(part->act_state != XPC_P_DEACTIVATING)) { - ret = xpc_IPI_send(part->remote_IPI_amo_va, - (u64) ipi_flag << (ch->number * 8), - part->remote_IPI_nasid, - part->remote_IPI_phys_cpuid, - SGI_XPC_NOTIFY); - dev_dbg(xpc_chan, "%s sent to partid=%d, channel=%d, ret=%d\n", - ipi_flag_string, ch->partid, ch->number, ret); - if (unlikely(ret != xpcSuccess)) { - if (irq_flags != NULL) { - spin_unlock_irqrestore(&ch->lock, *irq_flags); - } - XPC_DEACTIVATE_PARTITION(part, ret); - if (irq_flags != NULL) { - spin_lock_irqsave(&ch->lock, *irq_flags); - } - } - } -} - - -/* - * Make it look like the remote partition, which is associated with the - * specified channel, sent us an IPI. This faked IPI will be handled - * by xpc_dropped_IPI_check(). - */ -#define XPC_NOTIFY_IRQ_SEND_LOCAL(_ch, _ipi_f) \ - xpc_notify_IRQ_send_local(_ch, _ipi_f, #_ipi_f) - -static inline void -xpc_notify_IRQ_send_local(struct xpc_channel *ch, u8 ipi_flag, - char *ipi_flag_string) -{ - struct xpc_partition *part = &xpc_partitions[ch->partid]; - - - FETCHOP_STORE_OP(TO_AMO((u64) &part->local_IPI_amo_va->variable), - FETCHOP_OR, ((u64) ipi_flag << (ch->number * 8))); - dev_dbg(xpc_chan, "%s sent local from partid=%d, channel=%d\n", - ipi_flag_string, ch->partid, ch->number); -} - - -/* - * The sending and receiving of IPIs includes the setting of an AMO variable - * to indicate the reason the IPI was sent. The 64-bit variable is divided - * up into eight bytes, ordered from right to left. Byte zero pertains to - * channel 0, byte one to channel 1, and so on. Each byte is described by - * the following IPI flags. - */ - -#define XPC_IPI_CLOSEREQUEST 0x01 -#define XPC_IPI_CLOSEREPLY 0x02 -#define XPC_IPI_OPENREQUEST 0x04 -#define XPC_IPI_OPENREPLY 0x08 -#define XPC_IPI_MSGREQUEST 0x10 - - -/* given an AMO variable and a channel#, get its associated IPI flags */ -#define XPC_GET_IPI_FLAGS(_amo, _c) ((u8) (((_amo) >> ((_c) * 8)) & 0xff)) -#define XPC_SET_IPI_FLAGS(_amo, _c, _f) (_amo) |= ((u64) (_f) << ((_c) * 8)) - -#define XPC_ANY_OPENCLOSE_IPI_FLAGS_SET(_amo) ((_amo) & 0x0f0f0f0f0f0f0f0f) -#define XPC_ANY_MSG_IPI_FLAGS_SET(_amo) ((_amo) & 0x1010101010101010) - - -static inline void -xpc_IPI_send_closerequest(struct xpc_channel *ch, unsigned long *irq_flags) -{ - struct xpc_openclose_args *args = ch->local_openclose_args; - - - args->reason = ch->reason; - - XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_CLOSEREQUEST, irq_flags); -} - -static inline void -xpc_IPI_send_closereply(struct xpc_channel *ch, unsigned long *irq_flags) -{ - XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_CLOSEREPLY, irq_flags); -} - -static inline void -xpc_IPI_send_openrequest(struct xpc_channel *ch, unsigned long *irq_flags) -{ - struct xpc_openclose_args *args = ch->local_openclose_args; - - - args->msg_size = ch->msg_size; - args->local_nentries = ch->local_nentries; - - XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_OPENREQUEST, irq_flags); -} - -static inline void -xpc_IPI_send_openreply(struct xpc_channel *ch, unsigned long *irq_flags) -{ - struct xpc_openclose_args *args = ch->local_openclose_args; - - - args->remote_nentries = ch->remote_nentries; - args->local_nentries = ch->local_nentries; - args->local_msgqueue_pa = __pa(ch->local_msgqueue); - - XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_OPENREPLY, irq_flags); -} - -static inline void -xpc_IPI_send_msgrequest(struct xpc_channel *ch) -{ - XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_MSGREQUEST, NULL); -} - -static inline void -xpc_IPI_send_local_msgrequest(struct xpc_channel *ch) -{ - XPC_NOTIFY_IRQ_SEND_LOCAL(ch, XPC_IPI_MSGREQUEST); -} - - -/* - * Memory for XPC's AMO variables is allocated by the MSPEC driver. These - * pages are located in the lowest granule. The lowest granule uses 4k pages - * for cached references and an alternate TLB handler to never provide a - * cacheable mapping for the entire region. This will prevent speculative - * reading of cached copies of our lines from being issued which will cause - * a PI FSB Protocol error to be generated by the SHUB. For XPC, we need 64 - * AMO variables (based on XP_MAX_PARTITIONS) for message notification and an - * additional 128 AMO variables (based on XP_NASID_MASK_WORDS) for partition - * activation and 2 AMO variables for partition deactivation. - */ -static inline AMO_t * -xpc_IPI_init(int index) -{ - AMO_t *amo = xpc_vars->amos_page + index; - - - (void) xpc_IPI_receive(amo); /* clear AMO variable */ - return amo; -} - - - -static inline enum xpc_retval -xpc_map_bte_errors(bte_result_t error) -{ - switch (error) { - case BTE_SUCCESS: return xpcSuccess; - case BTEFAIL_DIR: return xpcBteDirectoryError; - case BTEFAIL_POISON: return xpcBtePoisonError; - case BTEFAIL_WERR: return xpcBteWriteError; - case BTEFAIL_ACCESS: return xpcBteAccessError; - case BTEFAIL_PWERR: return xpcBtePWriteError; - case BTEFAIL_PRERR: return xpcBtePReadError; - case BTEFAIL_TOUT: return xpcBteTimeOutError; - case BTEFAIL_XTERR: return xpcBteXtalkError; - case BTEFAIL_NOTAVAIL: return xpcBteNotAvailable; - default: return xpcBteUnmappedError; - } -} - - - -static inline void * -xpc_kmalloc_cacheline_aligned(size_t size, gfp_t flags, void **base) -{ - /* see if kmalloc will give us cachline aligned memory by default */ - *base = kmalloc(size, flags); - if (*base == NULL) { - return NULL; - } - if ((u64) *base == L1_CACHE_ALIGN((u64) *base)) { - return *base; - } - kfree(*base); - - /* nope, we'll have to do it ourselves */ - *base = kmalloc(size + L1_CACHE_BYTES, flags); - if (*base == NULL) { - return NULL; - } - return (void *) L1_CACHE_ALIGN((u64) *base); -} - - -/* - * Check to see if there is any channel activity to/from the specified - * partition. - */ -static inline void -xpc_check_for_channel_activity(struct xpc_partition *part) -{ - u64 IPI_amo; - unsigned long irq_flags; - - - IPI_amo = xpc_IPI_receive(part->local_IPI_amo_va); - if (IPI_amo == 0) { - return; - } - - spin_lock_irqsave(&part->IPI_lock, irq_flags); - part->local_IPI_amo |= IPI_amo; - spin_unlock_irqrestore(&part->IPI_lock, irq_flags); - - dev_dbg(xpc_chan, "received IPI from partid=%d, IPI_amo=0x%lx\n", - XPC_PARTID(part), IPI_amo); - - xpc_wakeup_channel_mgr(part); -} - - -#endif /* _IA64_SN_KERNEL_XPC_H */ - diff --git a/arch/ia64/sn/kernel/xpc_channel.c b/arch/ia64/sn/kernel/xpc_channel.c index abf4fc2a87b..8d950c778bb 100644 --- a/arch/ia64/sn/kernel/xpc_channel.c +++ b/arch/ia64/sn/kernel/xpc_channel.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2004-2005 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) 2004-2006 Silicon Graphics, Inc. All Rights Reserved. */ @@ -22,9 +22,11 @@ #include <linux/cache.h> #include <linux/interrupt.h> #include <linux/slab.h> +#include <linux/mutex.h> +#include <linux/completion.h> #include <asm/sn/bte.h> #include <asm/sn/sn_sal.h> -#include "xpc.h" +#include <asm/sn/xpc.h> /* @@ -56,8 +58,8 @@ xpc_initialize_channels(struct xpc_partition *part, partid_t partid) atomic_set(&ch->n_to_notify, 0); spin_lock_init(&ch->lock); - sema_init(&ch->msg_to_pull_sema, 1); /* mutex */ - sema_init(&ch->wdisconnect_sema, 0); /* event wait */ + mutex_init(&ch->msg_to_pull_mutex); + init_completion(&ch->wdisconnect_wait); atomic_set(&ch->n_on_msg_allocate_wq, 0); init_waitqueue_head(&ch->msg_allocate_wq); @@ -534,7 +536,6 @@ static enum xpc_retval xpc_allocate_msgqueues(struct xpc_channel *ch) { unsigned long irq_flags; - int i; enum xpc_retval ret; @@ -552,11 +553,6 @@ xpc_allocate_msgqueues(struct xpc_channel *ch) return ret; } - for (i = 0; i < ch->local_nentries; i++) { - /* use a semaphore as an event wait queue */ - sema_init(&ch->notify_queue[i].sema, 0); - } - spin_lock_irqsave(&ch->lock, irq_flags); ch->flags |= XPC_C_SETUP; spin_unlock_irqrestore(&ch->lock, irq_flags); @@ -779,6 +775,12 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags) /* both sides are disconnected now */ + if (ch->flags & XPC_C_CONNECTCALLOUT) { + spin_unlock_irqrestore(&ch->lock, *irq_flags); + xpc_disconnect_callout(ch, xpcDisconnected); + spin_lock_irqsave(&ch->lock, *irq_flags); + } + /* it's now safe to free the channel's message queues */ xpc_free_msgqueues(ch); @@ -793,10 +795,8 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags) } if (ch->flags & XPC_C_WDISCONNECT) { - spin_unlock_irqrestore(&ch->lock, *irq_flags); - up(&ch->wdisconnect_sema); - spin_lock_irqsave(&ch->lock, *irq_flags); - + /* we won't lose the CPU since we're holding ch->lock */ + complete(&ch->wdisconnect_wait); } else if (ch->delayed_IPI_flags) { if (part->act_state != XPC_P_DEACTIVATING) { /* time to take action on any delayed IPI flags */ @@ -1086,12 +1086,12 @@ xpc_connect_channel(struct xpc_channel *ch) struct xpc_registration *registration = &xpc_registrations[ch->number]; - if (down_trylock(®istration->sema) != 0) { + if (mutex_trylock(®istration->mutex) == 0) { return xpcRetry; } if (!XPC_CHANNEL_REGISTERED(ch->number)) { - up(®istration->sema); + mutex_unlock(®istration->mutex); return xpcUnregistered; } @@ -1102,7 +1102,7 @@ xpc_connect_channel(struct xpc_channel *ch) if (ch->flags & XPC_C_DISCONNECTING) { spin_unlock_irqrestore(&ch->lock, irq_flags); - up(®istration->sema); + mutex_unlock(®istration->mutex); return ch->reason; } @@ -1134,7 +1134,7 @@ xpc_connect_channel(struct xpc_channel *ch) * channel lock be locked and will unlock and relock * the channel lock as needed. */ - up(®istration->sema); + mutex_unlock(®istration->mutex); XPC_DISCONNECT_CHANNEL(ch, xpcUnequalMsgSizes, &irq_flags); spin_unlock_irqrestore(&ch->lock, irq_flags); @@ -1149,7 +1149,7 @@ xpc_connect_channel(struct xpc_channel *ch) atomic_inc(&xpc_partitions[ch->partid].nchannels_active); } - up(®istration->sema); + mutex_unlock(®istration->mutex); /* initiate the connection */ @@ -1645,7 +1645,7 @@ xpc_disconnect_channel(const int line, struct xpc_channel *ch, void -xpc_disconnecting_callout(struct xpc_channel *ch) +xpc_disconnect_callout(struct xpc_channel *ch, enum xpc_retval reason) { /* * Let the channel's registerer know that the channel is being @@ -1654,15 +1654,13 @@ xpc_disconnecting_callout(struct xpc_channel *ch) */ if (ch->func != NULL) { - dev_dbg(xpc_chan, "ch->func() called, reason=xpcDisconnecting," - " partid=%d, channel=%d\n", ch->partid, ch->number); + dev_dbg(xpc_chan, "ch->func() called, reason=%d, partid=%d, " + "channel=%d\n", reason, ch->partid, ch->number); - ch->func(xpcDisconnecting, ch->partid, ch->number, NULL, - ch->key); + ch->func(reason, ch->partid, ch->number, NULL, ch->key); - dev_dbg(xpc_chan, "ch->func() returned, reason=" - "xpcDisconnecting, partid=%d, channel=%d\n", - ch->partid, ch->number); + dev_dbg(xpc_chan, "ch->func() returned, reason=%d, partid=%d, " + "channel=%d\n", reason, ch->partid, ch->number); } } @@ -2085,7 +2083,7 @@ xpc_pull_remote_msg(struct xpc_channel *ch, s64 get) enum xpc_retval ret; - if (down_interruptible(&ch->msg_to_pull_sema) != 0) { + if (mutex_lock_interruptible(&ch->msg_to_pull_mutex) != 0) { /* we were interrupted by a signal */ return NULL; } @@ -2121,7 +2119,7 @@ xpc_pull_remote_msg(struct xpc_channel *ch, s64 get) XPC_DEACTIVATE_PARTITION(part, ret); - up(&ch->msg_to_pull_sema); + mutex_unlock(&ch->msg_to_pull_mutex); return NULL; } @@ -2130,7 +2128,7 @@ xpc_pull_remote_msg(struct xpc_channel *ch, s64 get) ch->next_msg_to_pull += nmsgs; } - up(&ch->msg_to_pull_sema); + mutex_unlock(&ch->msg_to_pull_mutex); /* return the message we were looking for */ msg_offset = (get % ch->remote_nentries) * ch->msg_size; diff --git a/arch/ia64/sn/kernel/xpc_main.c b/arch/ia64/sn/kernel/xpc_main.c index b617236524c..c75f8aeefc2 100644 --- a/arch/ia64/sn/kernel/xpc_main.c +++ b/arch/ia64/sn/kernel/xpc_main.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2004-2005 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) 2004-2006 Silicon Graphics, Inc. All Rights Reserved. */ @@ -55,11 +55,12 @@ #include <linux/slab.h> #include <linux/delay.h> #include <linux/reboot.h> +#include <linux/completion.h> #include <asm/sn/intr.h> #include <asm/sn/sn_sal.h> #include <asm/kdebug.h> #include <asm/uaccess.h> -#include "xpc.h" +#include <asm/sn/xpc.h> /* define two XPC debug device structures to be used with dev_dbg() et al */ @@ -82,6 +83,9 @@ struct device *xpc_part = &xpc_part_dbg_subname; struct device *xpc_chan = &xpc_chan_dbg_subname; +static int xpc_kdebug_ignore; + + /* systune related variables for /proc/sys directories */ static int xpc_hb_interval = XPC_HB_DEFAULT_INTERVAL; @@ -162,6 +166,8 @@ static ctl_table xpc_sys_dir[] = { }; static struct ctl_table_header *xpc_sysctl; +/* non-zero if any remote partition disengage request was timed out */ +int xpc_disengage_request_timedout; /* #of IRQs received */ static atomic_t xpc_act_IRQ_rcvd; @@ -172,10 +178,10 @@ static DECLARE_WAIT_QUEUE_HEAD(xpc_act_IRQ_wq); static unsigned long xpc_hb_check_timeout; /* notification that the xpc_hb_checker thread has exited */ -static DECLARE_MUTEX_LOCKED(xpc_hb_checker_exited); +static DECLARE_COMPLETION(xpc_hb_checker_exited); /* notification that the xpc_discovery thread has exited */ -static DECLARE_MUTEX_LOCKED(xpc_discovery_exited); +static DECLARE_COMPLETION(xpc_discovery_exited); static struct timer_list xpc_hb_timer; @@ -316,7 +322,7 @@ xpc_hb_checker(void *ignore) /* mark this thread as having exited */ - up(&xpc_hb_checker_exited); + complete(&xpc_hb_checker_exited); return 0; } @@ -336,7 +342,7 @@ xpc_initiate_discovery(void *ignore) dev_dbg(xpc_part, "discovery thread is exiting\n"); /* mark this thread as having exited */ - up(&xpc_discovery_exited); + complete(&xpc_discovery_exited); return 0; } @@ -773,7 +779,7 @@ xpc_daemonize_kthread(void *args) ch->flags |= XPC_C_DISCONNECTCALLOUT; spin_unlock_irqrestore(&ch->lock, irq_flags); - xpc_disconnecting_callout(ch); + xpc_disconnect_callout(ch, xpcDisconnecting); } else { spin_unlock_irqrestore(&ch->lock, irq_flags); } @@ -888,7 +894,7 @@ xpc_disconnect_wait(int ch_number) continue; } - (void) down(&ch->wdisconnect_sema); + wait_for_completion(&ch->wdisconnect_wait); spin_lock_irqsave(&ch->lock, irq_flags); DBUG_ON(!(ch->flags & XPC_C_DISCONNECTED)); @@ -921,9 +927,9 @@ static void xpc_do_exit(enum xpc_retval reason) { partid_t partid; - int active_part_count; + int active_part_count, printed_waiting_msg = 0; struct xpc_partition *part; - unsigned long printmsg_time; + unsigned long printmsg_time, disengage_request_timeout = 0; /* a 'rmmod XPC' and a 'reboot' cannot both end up here together */ @@ -941,10 +947,10 @@ xpc_do_exit(enum xpc_retval reason) free_irq(SGI_XPC_ACTIVATE, NULL); /* wait for the discovery thread to exit */ - down(&xpc_discovery_exited); + wait_for_completion(&xpc_discovery_exited); /* wait for the heartbeat checker thread to exit */ - down(&xpc_hb_checker_exited); + wait_for_completion(&xpc_hb_checker_exited); /* sleep for a 1/3 of a second or so */ @@ -953,7 +959,8 @@ xpc_do_exit(enum xpc_retval reason) /* wait for all partitions to become inactive */ - printmsg_time = jiffies; + printmsg_time = jiffies + (XPC_DISENGAGE_PRINTMSG_INTERVAL * HZ); + xpc_disengage_request_timedout = 0; do { active_part_count = 0; @@ -969,20 +976,39 @@ xpc_do_exit(enum xpc_retval reason) active_part_count++; XPC_DEACTIVATE_PARTITION(part, reason); - } - if (active_part_count == 0) { - break; + if (part->disengage_request_timeout > + disengage_request_timeout) { + disengage_request_timeout = + part->disengage_request_timeout; + } } - if (jiffies >= printmsg_time) { - dev_info(xpc_part, "waiting for partitions to " - "deactivate/disengage, active count=%d, remote " - "engaged=0x%lx\n", active_part_count, - xpc_partition_engaged(1UL << partid)); - - printmsg_time = jiffies + + if (xpc_partition_engaged(-1UL)) { + if (time_after(jiffies, printmsg_time)) { + dev_info(xpc_part, "waiting for remote " + "partitions to disengage, timeout in " + "%ld seconds\n", + (disengage_request_timeout - jiffies) + / HZ); + printmsg_time = jiffies + (XPC_DISENGAGE_PRINTMSG_INTERVAL * HZ); + printed_waiting_msg = 1; + } + + } else if (active_part_count > 0) { + if (printed_waiting_msg) { + dev_info(xpc_part, "waiting for local partition" + " to disengage\n"); + printed_waiting_msg = 0; + } + + } else { + if (!xpc_disengage_request_timedout) { + dev_info(xpc_part, "all partitions have " + "disengaged\n"); + } + break; } /* sleep for a 1/3 of a second or so */ @@ -1000,11 +1026,13 @@ xpc_do_exit(enum xpc_retval reason) del_timer_sync(&xpc_hb_timer); DBUG_ON(xpc_vars->heartbeating_to_mask != 0); - /* take ourselves off of the reboot_notifier_list */ - (void) unregister_reboot_notifier(&xpc_reboot_notifier); + if (reason == xpcUnloading) { + /* take ourselves off of the reboot_notifier_list */ + (void) unregister_reboot_notifier(&xpc_reboot_notifier); - /* take ourselves off of the die_notifier list */ - (void) unregister_die_notifier(&xpc_die_notifier); + /* take ourselves off of the die_notifier list */ + (void) unregister_die_notifier(&xpc_die_notifier); + } /* close down protections for IPI operations */ xpc_restrict_IPI_ops(); @@ -1020,7 +1048,35 @@ xpc_do_exit(enum xpc_retval reason) /* - * Called when the system is about to be either restarted or halted. + * This function is called when the system is being rebooted. + */ +static int +xpc_system_reboot(struct notifier_block *nb, unsigned long event, void *unused) +{ + enum xpc_retval reason; + + + switch (event) { + case SYS_RESTART: + reason = xpcSystemReboot; + break; + case SYS_HALT: + reason = xpcSystemHalt; + break; + case SYS_POWER_OFF: + reason = xpcSystemPoweroff; + break; + default: + reason = xpcSystemGoingDown; + } + + xpc_do_exit(reason); + return NOTIFY_DONE; +} + + +/* + * Notify other partitions to disengage from all references to our memory. */ static void xpc_die_disengage(void) @@ -1028,7 +1084,7 @@ xpc_die_disengage(void) struct xpc_partition *part; partid_t partid; unsigned long engaged; - long time, print_time, disengage_request_timeout; + long time, printmsg_time, disengage_request_timeout; /* keep xpc_hb_checker thread from doing anything (just in case) */ @@ -1055,57 +1111,53 @@ xpc_die_disengage(void) } } - print_time = rtc_time(); - disengage_request_timeout = print_time + + time = rtc_time(); + printmsg_time = time + + (XPC_DISENGAGE_PRINTMSG_INTERVAL * sn_rtc_cycles_per_second); + disengage_request_timeout = time + (xpc_disengage_request_timelimit * sn_rtc_cycles_per_second); /* wait for all other partitions to disengage from us */ - while ((engaged = xpc_partition_engaged(-1UL)) && - (time = rtc_time()) < disengage_request_timeout) { + while (1) { + engaged = xpc_partition_engaged(-1UL); + if (!engaged) { + dev_info(xpc_part, "all partitions have disengaged\n"); + break; + } - if (time >= print_time) { + time = rtc_time(); + if (time >= disengage_request_timeout) { + for (partid = 1; partid < XP_MAX_PARTITIONS; partid++) { + if (engaged & (1UL << partid)) { + dev_info(xpc_part, "disengage from " + "remote partition %d timed " + "out\n", partid); + } + } + break; + } + + if (time >= printmsg_time) { dev_info(xpc_part, "waiting for remote partitions to " - "disengage, engaged=0x%lx\n", engaged); - print_time = time + (XPC_DISENGAGE_PRINTMSG_INTERVAL * + "disengage, timeout in %ld seconds\n", + (disengage_request_timeout - time) / + sn_rtc_cycles_per_second); + printmsg_time = time + + (XPC_DISENGAGE_PRINTMSG_INTERVAL * sn_rtc_cycles_per_second); } } - dev_info(xpc_part, "finished waiting for remote partitions to " - "disengage, engaged=0x%lx\n", engaged); -} - - -/* - * This function is called when the system is being rebooted. - */ -static int -xpc_system_reboot(struct notifier_block *nb, unsigned long event, void *unused) -{ - enum xpc_retval reason; - - - switch (event) { - case SYS_RESTART: - reason = xpcSystemReboot; - break; - case SYS_HALT: - reason = xpcSystemHalt; - break; - case SYS_POWER_OFF: - reason = xpcSystemPoweroff; - break; - default: - reason = xpcSystemGoingDown; - } - - xpc_do_exit(reason); - return NOTIFY_DONE; } /* - * This function is called when the system is being rebooted. + * This function is called when the system is being restarted or halted due + * to some sort of system failure. If this is the case we need to notify the + * other partitions to disengage from all references to our memory. + * This function can also be called when our heartbeater could be offlined + * for a time. In this case we need to notify other partitions to not worry + * about the lack of a heartbeat. */ static int xpc_system_die(struct notifier_block *nb, unsigned long event, void *unused) @@ -1115,11 +1167,25 @@ xpc_system_die(struct notifier_block *nb, unsigned long event, void *unused) case DIE_MACHINE_HALT: xpc_die_disengage(); break; + + case DIE_KDEBUG_ENTER: + /* Should lack of heartbeat be ignored by other partitions? */ + if (!xpc_kdebug_ignore) { + break; + } + /* fall through */ case DIE_MCA_MONARCH_ENTER: case DIE_INIT_MONARCH_ENTER: xpc_vars->heartbeat++; xpc_vars->heartbeat_offline = 1; break; + + case DIE_KDEBUG_LEAVE: + /* Is lack of heartbeat being ignored by other partitions? */ + if (!xpc_kdebug_ignore) { + break; + } + /* fall through */ case DIE_MCA_MONARCH_LEAVE: case DIE_INIT_MONARCH_LEAVE: xpc_vars->heartbeat++; @@ -1302,7 +1368,7 @@ xpc_init(void) dev_err(xpc_part, "failed while forking discovery thread\n"); /* mark this new thread as a non-starter */ - up(&xpc_discovery_exited); + complete(&xpc_discovery_exited); xpc_do_exit(xpcUnloading); return -EBUSY; @@ -1344,3 +1410,7 @@ module_param(xpc_disengage_request_timelimit, int, 0); MODULE_PARM_DESC(xpc_disengage_request_timelimit, "Number of seconds to wait " "for disengage request to complete."); +module_param(xpc_kdebug_ignore, int, 0); +MODULE_PARM_DESC(xpc_kdebug_ignore, "Should lack of heartbeat be ignored by " + "other partitions when dropping into kdebug."); + diff --git a/arch/ia64/sn/kernel/xpc_partition.c b/arch/ia64/sn/kernel/xpc_partition.c index cdd6431853a..88a730e6cfd 100644 --- a/arch/ia64/sn/kernel/xpc_partition.c +++ b/arch/ia64/sn/kernel/xpc_partition.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (c) 2004-2005 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) 2004-2006 Silicon Graphics, Inc. All Rights Reserved. */ @@ -28,7 +28,7 @@ #include <asm/sn/sn_sal.h> #include <asm/sn/nodepda.h> #include <asm/sn/addrs.h> -#include "xpc.h" +#include <asm/sn/xpc.h> /* XPC is exiting flag */ @@ -771,7 +771,8 @@ xpc_identify_act_IRQ_req(int nasid) } } - if (!xpc_partition_disengaged(part)) { + if (part->disengage_request_timeout > 0 && + !xpc_partition_disengaged(part)) { /* still waiting on other side to disengage from us */ return; } @@ -873,6 +874,9 @@ xpc_partition_disengaged(struct xpc_partition *part) * request in a timely fashion, so assume it's dead. */ + dev_info(xpc_part, "disengage from remote partition %d " + "timed out\n", partid); + xpc_disengage_request_timedout = 1; xpc_clear_partition_engaged(1UL << partid); disengaged = 1; } diff --git a/arch/ia64/sn/pci/pcibr/pcibr_ate.c b/arch/ia64/sn/pci/pcibr/pcibr_ate.c index d1647b863e6..aa3fa5152a3 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_ate.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_ate.c @@ -18,10 +18,10 @@ int pcibr_invalidate_ate = 0; /* by default don't invalidate ATE on free */ * mark_ate: Mark the ate as either free or inuse. */ static void mark_ate(struct ate_resource *ate_resource, int start, int number, - uint64_t value) + u64 value) { - uint64_t *ate = ate_resource->ate; + u64 *ate = ate_resource->ate; int index; int length = 0; @@ -38,7 +38,7 @@ static int find_free_ate(struct ate_resource *ate_resource, int start, int count) { - uint64_t *ate = ate_resource->ate; + u64 *ate = ate_resource->ate; int index; int start_free; @@ -119,7 +119,7 @@ static inline int alloc_ate_resource(struct ate_resource *ate_resource, int pcibr_ate_alloc(struct pcibus_info *pcibus_info, int count) { int status = 0; - uint64_t flag; + u64 flag; flag = pcibr_lock(pcibus_info); status = alloc_ate_resource(&pcibus_info->pbi_int_ate_resource, count); @@ -139,7 +139,7 @@ int pcibr_ate_alloc(struct pcibus_info *pcibus_info, int count) * Setup an Address Translation Entry as specified. Use either the Bridge * internal maps or the external map RAM, as appropriate. */ -static inline uint64_t *pcibr_ate_addr(struct pcibus_info *pcibus_info, +static inline u64 *pcibr_ate_addr(struct pcibus_info *pcibus_info, int ate_index) { if (ate_index < pcibus_info->pbi_int_ate_size) { @@ -153,7 +153,7 @@ static inline uint64_t *pcibr_ate_addr(struct pcibus_info *pcibus_info, */ void inline ate_write(struct pcibus_info *pcibus_info, int ate_index, int count, - volatile uint64_t ate) + volatile u64 ate) { while (count-- > 0) { if (ate_index < pcibus_info->pbi_int_ate_size) { @@ -171,9 +171,9 @@ ate_write(struct pcibus_info *pcibus_info, int ate_index, int count, void pcibr_ate_free(struct pcibus_info *pcibus_info, int index) { - volatile uint64_t ate; + volatile u64 ate; int count; - uint64_t flags; + u64 flags; if (pcibr_invalidate_ate) { /* For debugging purposes, clear the valid bit in the ATE */ diff --git a/arch/ia64/sn/pci/pcibr/pcibr_dma.c b/arch/ia64/sn/pci/pcibr/pcibr_dma.c index 34093476e96..54ce5b7ceed 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_dma.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_dma.c @@ -41,21 +41,21 @@ extern int sn_ioif_inited; static dma_addr_t pcibr_dmamap_ate32(struct pcidev_info *info, - uint64_t paddr, size_t req_size, uint64_t flags) + u64 paddr, size_t req_size, u64 flags) { struct pcidev_info *pcidev_info = info->pdi_host_pcidev_info; struct pcibus_info *pcibus_info = (struct pcibus_info *)pcidev_info-> pdi_pcibus_info; - uint8_t internal_device = (PCI_SLOT(pcidev_info->pdi_host_pcidev_info-> + u8 internal_device = (PCI_SLOT(pcidev_info->pdi_host_pcidev_info-> pdi_linux_pcidev->devfn)) - 1; int ate_count; int ate_index; - uint64_t ate_flags = flags | PCI32_ATE_V; - uint64_t ate; - uint64_t pci_addr; - uint64_t xio_addr; - uint64_t offset; + u64 ate_flags = flags | PCI32_ATE_V; + u64 ate; + u64 pci_addr; + u64 xio_addr; + u64 offset; /* PIC in PCI-X mode does not supports 32bit PageMap mode */ if (IS_PIC_SOFT(pcibus_info) && IS_PCIX(pcibus_info)) { @@ -109,12 +109,12 @@ pcibr_dmamap_ate32(struct pcidev_info *info, } static dma_addr_t -pcibr_dmatrans_direct64(struct pcidev_info * info, uint64_t paddr, - uint64_t dma_attributes) +pcibr_dmatrans_direct64(struct pcidev_info * info, u64 paddr, + u64 dma_attributes) { struct pcibus_info *pcibus_info = (struct pcibus_info *) ((info->pdi_host_pcidev_info)->pdi_pcibus_info); - uint64_t pci_addr; + u64 pci_addr; /* Translate to Crosstalk View of Physical Address */ pci_addr = (IS_PIC_SOFT(pcibus_info) ? PHYS_TO_DMA(paddr) : @@ -127,7 +127,7 @@ pcibr_dmatrans_direct64(struct pcidev_info * info, uint64_t paddr, /* Handle Bridge Chipset differences */ if (IS_PIC_SOFT(pcibus_info)) { pci_addr |= - ((uint64_t) pcibus_info-> + ((u64) pcibus_info-> pbi_hub_xid << PIC_PCI64_ATTR_TARG_SHFT); } else pci_addr |= TIOCP_PCI64_CMDTYPE_MEM; @@ -142,17 +142,17 @@ pcibr_dmatrans_direct64(struct pcidev_info * info, uint64_t paddr, static dma_addr_t pcibr_dmatrans_direct32(struct pcidev_info * info, - uint64_t paddr, size_t req_size, uint64_t flags) + u64 paddr, size_t req_size, u64 flags) { struct pcidev_info *pcidev_info = info->pdi_host_pcidev_info; struct pcibus_info *pcibus_info = (struct pcibus_info *)pcidev_info-> pdi_pcibus_info; - uint64_t xio_addr; + u64 xio_addr; - uint64_t xio_base; - uint64_t offset; - uint64_t endoff; + u64 xio_base; + u64 offset; + u64 endoff; if (IS_PCIX(pcibus_info)) { return 0; @@ -209,16 +209,18 @@ pcibr_dma_unmap(struct pci_dev *hwdev, dma_addr_t dma_handle, int direction) * unlike the PIC Device(x) Write Request Buffer Flush register. */ -void sn_dma_flush(uint64_t addr) +void sn_dma_flush(u64 addr) { nasid_t nasid; int is_tio; int wid_num; int i, j; - uint64_t flags; - uint64_t itte; + u64 flags; + u64 itte; struct hubdev_info *hubinfo; - volatile struct sn_flush_device_list *p; + volatile struct sn_flush_device_kernel *p; + volatile struct sn_flush_device_common *common; + struct sn_flush_nasid_entry *flush_nasid_list; if (!sn_ioif_inited) @@ -268,17 +270,17 @@ void sn_dma_flush(uint64_t addr) p = &flush_nasid_list->widget_p[wid_num][0]; /* find a matching BAR */ - for (i = 0; i < DEV_PER_WIDGET; i++) { + for (i = 0; i < DEV_PER_WIDGET; i++,p++) { + common = p->common; for (j = 0; j < PCI_ROM_RESOURCE; j++) { - if (p->sfdl_bar_list[j].start == 0) + if (common->sfdl_bar_list[j].start == 0) break; - if (addr >= p->sfdl_bar_list[j].start - && addr <= p->sfdl_bar_list[j].end) + if (addr >= common->sfdl_bar_list[j].start + && addr <= common->sfdl_bar_list[j].end) break; } - if (j < PCI_ROM_RESOURCE && p->sfdl_bar_list[j].start != 0) + if (j < PCI_ROM_RESOURCE && common->sfdl_bar_list[j].start != 0) break; - p++; } /* if no matching BAR, return without doing anything. */ @@ -297,31 +299,31 @@ void sn_dma_flush(uint64_t addr) * If CE ever needs the sn_dma_flush mechanism, we will have * to account for that here and in tioce_bus_fixup(). */ - uint32_t tio_id = HUB_L(TIO_IOSPACE_ADDR(nasid, TIO_NODE_ID)); - uint32_t revnum = XWIDGET_PART_REV_NUM(tio_id); + u32 tio_id = HUB_L(TIO_IOSPACE_ADDR(nasid, TIO_NODE_ID)); + u32 revnum = XWIDGET_PART_REV_NUM(tio_id); /* TIOCP BRINGUP WAR (PV907516): Don't write buffer flush reg */ if ((1 << XWIDGET_PART_REV_NUM_REV(revnum)) & PV907516) { return; } else { - pcireg_wrb_flush_get(p->sfdl_pcibus_info, - (p->sfdl_slot - 1)); + pcireg_wrb_flush_get(common->sfdl_pcibus_info, + (common->sfdl_slot - 1)); } } else { - spin_lock_irqsave(&((struct sn_flush_device_list *)p)-> - sfdl_flush_lock, flags); - - *p->sfdl_flush_addr = 0; + spin_lock_irqsave((spinlock_t *)&p->sfdl_flush_lock, + flags); + *common->sfdl_flush_addr = 0; /* force an interrupt. */ - *(volatile uint32_t *)(p->sfdl_force_int_addr) = 1; + *(volatile u32 *)(common->sfdl_force_int_addr) = 1; /* wait for the interrupt to come back. */ - while (*(p->sfdl_flush_addr) != 0x10f) + while (*(common->sfdl_flush_addr) != 0x10f) cpu_relax(); /* okay, everything is synched up. */ - spin_unlock_irqrestore((spinlock_t *)&p->sfdl_flush_lock, flags); + spin_unlock_irqrestore((spinlock_t *)&p->sfdl_flush_lock, + flags); } return; } diff --git a/arch/ia64/sn/pci/pcibr/pcibr_provider.c b/arch/ia64/sn/pci/pcibr/pcibr_provider.c index 1f500c81002..2fac27049bf 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_provider.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_provider.c @@ -23,14 +23,16 @@ int sal_pcibr_slot_enable(struct pcibus_info *soft, int device, void *resp) { struct ia64_sal_retval ret_stuff; - uint64_t busnum; + u64 busnum; + u64 segment; ret_stuff.status = 0; ret_stuff.v0 = 0; + segment = soft->pbi_buscommon.bs_persist_segment; busnum = soft->pbi_buscommon.bs_persist_busnum; - SAL_CALL_NOLOCK(ret_stuff, (u64) SN_SAL_IOIF_SLOT_ENABLE, (u64) busnum, - (u64) device, (u64) resp, 0, 0, 0, 0); + SAL_CALL_NOLOCK(ret_stuff, (u64) SN_SAL_IOIF_SLOT_ENABLE, segment, + busnum, (u64) device, (u64) resp, 0, 0, 0); return (int)ret_stuff.v0; } @@ -40,15 +42,17 @@ sal_pcibr_slot_disable(struct pcibus_info *soft, int device, int action, void *resp) { struct ia64_sal_retval ret_stuff; - uint64_t busnum; + u64 busnum; + u64 segment; ret_stuff.status = 0; ret_stuff.v0 = 0; + segment = soft->pbi_buscommon.bs_persist_segment; busnum = soft->pbi_buscommon.bs_persist_busnum; SAL_CALL_NOLOCK(ret_stuff, (u64) SN_SAL_IOIF_SLOT_DISABLE, - (u64) busnum, (u64) device, (u64) action, - (u64) resp, 0, 0, 0); + segment, busnum, (u64) device, (u64) action, + (u64) resp, 0, 0); return (int)ret_stuff.v0; } @@ -56,7 +60,7 @@ sal_pcibr_slot_disable(struct pcibus_info *soft, int device, int action, static int sal_pcibr_error_interrupt(struct pcibus_info *soft) { struct ia64_sal_retval ret_stuff; - uint64_t busnum; + u64 busnum; int segment; ret_stuff.status = 0; ret_stuff.v0 = 0; @@ -92,7 +96,8 @@ pcibr_bus_fixup(struct pcibus_bussoft *prom_bussoft, struct pci_controller *cont cnodeid_t near_cnode; struct hubdev_info *hubdev_info; struct pcibus_info *soft; - struct sn_flush_device_list *sn_flush_device_list; + struct sn_flush_device_kernel *sn_flush_device_kernel; + struct sn_flush_device_common *common; if (! IS_PCI_BRIDGE_ASIC(prom_bussoft->bs_asic_type)) { return NULL; @@ -137,20 +142,19 @@ pcibr_bus_fixup(struct pcibus_bussoft *prom_bussoft, struct pci_controller *cont hubdev_info = (struct hubdev_info *)(NODEPDA(cnode)->pdinfo); if (hubdev_info->hdi_flush_nasid_list.widget_p) { - sn_flush_device_list = hubdev_info->hdi_flush_nasid_list. + sn_flush_device_kernel = hubdev_info->hdi_flush_nasid_list. widget_p[(int)soft->pbi_buscommon.bs_xid]; - if (sn_flush_device_list) { + if (sn_flush_device_kernel) { for (j = 0; j < DEV_PER_WIDGET; - j++, sn_flush_device_list++) { - if (sn_flush_device_list->sfdl_slot == -1) + j++, sn_flush_device_kernel++) { + common = sn_flush_device_kernel->common; + if (common->sfdl_slot == -1) continue; - if ((sn_flush_device_list-> - sfdl_persistent_segment == + if ((common->sfdl_persistent_segment == soft->pbi_buscommon.bs_persist_segment) && - (sn_flush_device_list-> - sfdl_persistent_busnum == + (common->sfdl_persistent_busnum == soft->pbi_buscommon.bs_persist_busnum)) - sn_flush_device_list->sfdl_pcibus_info = + common->sfdl_pcibus_info = soft; } } @@ -159,9 +163,9 @@ pcibr_bus_fixup(struct pcibus_bussoft *prom_bussoft, struct pci_controller *cont /* Setup the PMU ATE map */ soft->pbi_int_ate_resource.lowest_free_index = 0; soft->pbi_int_ate_resource.ate = - kmalloc(soft->pbi_int_ate_size * sizeof(uint64_t), GFP_KERNEL); + kmalloc(soft->pbi_int_ate_size * sizeof(u64), GFP_KERNEL); memset(soft->pbi_int_ate_resource.ate, 0, - (soft->pbi_int_ate_size * sizeof(uint64_t))); + (soft->pbi_int_ate_size * sizeof(u64))); if (prom_bussoft->bs_asic_type == PCIIO_ASIC_TYPE_TIOCP) { /* TIO PCI Bridge: find nearest node with CPUs */ @@ -203,7 +207,7 @@ void pcibr_target_interrupt(struct sn_irq_info *sn_irq_info) struct pcidev_info *pcidev_info; struct pcibus_info *pcibus_info; int bit = sn_irq_info->irq_int_bit; - uint64_t xtalk_addr = sn_irq_info->irq_xtalkaddr; + u64 xtalk_addr = sn_irq_info->irq_xtalkaddr; pcidev_info = (struct pcidev_info *)sn_irq_info->irq_pciioinfo; if (pcidev_info) { diff --git a/arch/ia64/sn/pci/pcibr/pcibr_reg.c b/arch/ia64/sn/pci/pcibr/pcibr_reg.c index 5d534091262..8b8bbd51d43 100644 --- a/arch/ia64/sn/pci/pcibr/pcibr_reg.c +++ b/arch/ia64/sn/pci/pcibr/pcibr_reg.c @@ -23,9 +23,9 @@ union br_ptr { /* * Control Register Access -- Read/Write 0000_0020 */ -void pcireg_control_bit_clr(struct pcibus_info *pcibus_info, uint64_t bits) +void pcireg_control_bit_clr(struct pcibus_info *pcibus_info, u64 bits) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -38,14 +38,14 @@ void pcireg_control_bit_clr(struct pcibus_info *pcibus_info, uint64_t bits) default: panic ("pcireg_control_bit_clr: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } -void pcireg_control_bit_set(struct pcibus_info *pcibus_info, uint64_t bits) +void pcireg_control_bit_set(struct pcibus_info *pcibus_info, u64 bits) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -58,7 +58,7 @@ void pcireg_control_bit_set(struct pcibus_info *pcibus_info, uint64_t bits) default: panic ("pcireg_control_bit_set: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } @@ -66,10 +66,10 @@ void pcireg_control_bit_set(struct pcibus_info *pcibus_info, uint64_t bits) /* * PCI/PCIX Target Flush Register Access -- Read Only 0000_0050 */ -uint64_t pcireg_tflush_get(struct pcibus_info *pcibus_info) +u64 pcireg_tflush_get(struct pcibus_info *pcibus_info) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; - uint64_t ret = 0; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; + u64 ret = 0; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -82,7 +82,7 @@ uint64_t pcireg_tflush_get(struct pcibus_info *pcibus_info) default: panic ("pcireg_tflush_get: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } @@ -96,10 +96,10 @@ uint64_t pcireg_tflush_get(struct pcibus_info *pcibus_info) /* * Interrupt Status Register Access -- Read Only 0000_0100 */ -uint64_t pcireg_intr_status_get(struct pcibus_info * pcibus_info) +u64 pcireg_intr_status_get(struct pcibus_info * pcibus_info) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; - uint64_t ret = 0; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; + u64 ret = 0; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -112,7 +112,7 @@ uint64_t pcireg_intr_status_get(struct pcibus_info * pcibus_info) default: panic ("pcireg_intr_status_get: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } return ret; @@ -121,9 +121,9 @@ uint64_t pcireg_intr_status_get(struct pcibus_info * pcibus_info) /* * Interrupt Enable Register Access -- Read/Write 0000_0108 */ -void pcireg_intr_enable_bit_clr(struct pcibus_info *pcibus_info, uint64_t bits) +void pcireg_intr_enable_bit_clr(struct pcibus_info *pcibus_info, u64 bits) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -136,14 +136,14 @@ void pcireg_intr_enable_bit_clr(struct pcibus_info *pcibus_info, uint64_t bits) default: panic ("pcireg_intr_enable_bit_clr: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } -void pcireg_intr_enable_bit_set(struct pcibus_info *pcibus_info, uint64_t bits) +void pcireg_intr_enable_bit_set(struct pcibus_info *pcibus_info, u64 bits) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -156,7 +156,7 @@ void pcireg_intr_enable_bit_set(struct pcibus_info *pcibus_info, uint64_t bits) default: panic ("pcireg_intr_enable_bit_set: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } @@ -165,9 +165,9 @@ void pcireg_intr_enable_bit_set(struct pcibus_info *pcibus_info, uint64_t bits) * Intr Host Address Register (int_addr) -- Read/Write 0000_0130 - 0000_0168 */ void pcireg_intr_addr_addr_set(struct pcibus_info *pcibus_info, int int_n, - uint64_t addr) + u64 addr) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -186,7 +186,7 @@ void pcireg_intr_addr_addr_set(struct pcibus_info *pcibus_info, int int_n, default: panic ("pcireg_intr_addr_addr_get: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } @@ -196,7 +196,7 @@ void pcireg_intr_addr_addr_set(struct pcibus_info *pcibus_info, int int_n, */ void pcireg_force_intr_set(struct pcibus_info *pcibus_info, int int_n) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -209,7 +209,7 @@ void pcireg_force_intr_set(struct pcibus_info *pcibus_info, int int_n) default: panic ("pcireg_force_intr_set: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } @@ -217,10 +217,10 @@ void pcireg_force_intr_set(struct pcibus_info *pcibus_info, int int_n) /* * Device(x) Write Buffer Flush Reg Access -- Read Only 0000_0240 - 0000_0258 */ -uint64_t pcireg_wrb_flush_get(struct pcibus_info *pcibus_info, int device) +u64 pcireg_wrb_flush_get(struct pcibus_info *pcibus_info, int device) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; - uint64_t ret = 0; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; + u64 ret = 0; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -233,7 +233,7 @@ uint64_t pcireg_wrb_flush_get(struct pcibus_info *pcibus_info, int device) __sn_readq_relaxed(&ptr->pic.p_wr_req_buf[device]); break; default: - panic("pcireg_wrb_flush_get: unknown bridgetype bridge 0x%p", (void *)ptr); + panic("pcireg_wrb_flush_get: unknown bridgetype bridge 0x%p", ptr); } } @@ -242,9 +242,9 @@ uint64_t pcireg_wrb_flush_get(struct pcibus_info *pcibus_info, int device) } void pcireg_int_ate_set(struct pcibus_info *pcibus_info, int ate_index, - uint64_t val) + u64 val) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -257,15 +257,15 @@ void pcireg_int_ate_set(struct pcibus_info *pcibus_info, int ate_index, default: panic ("pcireg_int_ate_set: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } } -uint64_t *pcireg_int_ate_addr(struct pcibus_info *pcibus_info, int ate_index) +u64 __iomem *pcireg_int_ate_addr(struct pcibus_info *pcibus_info, int ate_index) { - union br_ptr *ptr = (union br_ptr *)pcibus_info->pbi_buscommon.bs_base; - uint64_t *ret = (uint64_t *) 0; + union br_ptr __iomem *ptr = (union br_ptr __iomem *)pcibus_info->pbi_buscommon.bs_base; + u64 __iomem *ret = NULL; if (pcibus_info) { switch (pcibus_info->pbi_bridge_type) { @@ -278,7 +278,7 @@ uint64_t *pcireg_int_ate_addr(struct pcibus_info *pcibus_info, int ate_index) default: panic ("pcireg_int_ate_addr: unknown bridgetype bridge 0x%p", - (void *)ptr); + ptr); } } return ret; diff --git a/arch/ia64/sn/pci/tioca_provider.c b/arch/ia64/sn/pci/tioca_provider.c index 46b646a6d34..7571a402552 100644 --- a/arch/ia64/sn/pci/tioca_provider.c +++ b/arch/ia64/sn/pci/tioca_provider.c @@ -16,7 +16,7 @@ #include <asm/sn/pcibus_provider_defs.h> #include <asm/sn/tioca_provider.h> -uint32_t tioca_gart_found; +u32 tioca_gart_found; EXPORT_SYMBOL(tioca_gart_found); /* used by agp-sgi */ LIST_HEAD(tioca_list); @@ -34,14 +34,14 @@ static int tioca_gart_init(struct tioca_kernel *); static int tioca_gart_init(struct tioca_kernel *tioca_kern) { - uint64_t ap_reg; - uint64_t offset; + u64 ap_reg; + u64 offset; struct page *tmp; struct tioca_common *tioca_common; - struct tioca *ca_base; + struct tioca __iomem *ca_base; tioca_common = tioca_kern->ca_common; - ca_base = (struct tioca *)tioca_common->ca_common.bs_base; + ca_base = (struct tioca __iomem *)tioca_common->ca_common.bs_base; if (list_empty(tioca_kern->ca_devices)) return 0; @@ -214,8 +214,8 @@ void tioca_fastwrite_enable(struct tioca_kernel *tioca_kern) { int cap_ptr; - uint32_t reg; - struct tioca *tioca_base; + u32 reg; + struct tioca __iomem *tioca_base; struct pci_dev *pdev; struct tioca_common *common; @@ -257,7 +257,7 @@ tioca_fastwrite_enable(struct tioca_kernel *tioca_kern) * Set ca's fw to match */ - tioca_base = (struct tioca *)common->ca_common.bs_base; + tioca_base = (struct tioca __iomem*)common->ca_common.bs_base; __sn_setq_relaxed(&tioca_base->ca_control1, CA_AGP_FW_ENABLE); } @@ -276,7 +276,7 @@ EXPORT_SYMBOL(tioca_fastwrite_enable); /* used by agp-sgi */ * We will always use 0x1 * 55:55 - Swap bytes Currently unused */ -static uint64_t +static u64 tioca_dma_d64(unsigned long paddr) { dma_addr_t bus_addr; @@ -318,19 +318,19 @@ tioca_dma_d64(unsigned long paddr) * 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) +static u64 +tioca_dma_d48(struct pci_dev *pdev, u64 paddr) { struct tioca_common *tioca_common; - struct tioca *ca_base; - uint64_t ct_addr; + struct tioca __iomem *ca_base; + u64 ct_addr; dma_addr_t bus_addr; - uint32_t node_upper; - uint64_t agp_dma_extn; + u32 node_upper; + u64 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; + ca_base = (struct tioca __iomem *)tioca_common->ca_common.bs_base; ct_addr = PHYS_TO_TIODMA(paddr); if (!ct_addr) @@ -367,10 +367,10 @@ tioca_dma_d48(struct pci_dev *pdev, uint64_t paddr) * 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) +tioca_dma_mapped(struct pci_dev *pdev, u64 paddr, size_t req_size) { int i, ps, ps_shift, entry, entries, mapsize, last_entry; - uint64_t xio_addr, end_xio_addr; + u64 xio_addr, end_xio_addr; struct tioca_common *tioca_common; struct tioca_kernel *tioca_kern; dma_addr_t bus_addr = 0; @@ -514,10 +514,10 @@ tioca_dma_unmap(struct pci_dev *pdev, dma_addr_t bus_addr, int dir) * The mapping mode used is based on the devices dma_mask. As a last resort * use the GART mapped mode. */ -static uint64_t -tioca_dma_map(struct pci_dev *pdev, uint64_t paddr, size_t byte_count) +static u64 +tioca_dma_map(struct pci_dev *pdev, u64 paddr, size_t byte_count) { - uint64_t mapaddr; + u64 mapaddr; /* * If card is 64 or 48 bit addresable, use a direct mapping. 32 @@ -554,8 +554,8 @@ 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; + u64 segment; + u64 busnum; ret_stuff.status = 0; ret_stuff.v0 = 0; @@ -620,7 +620,7 @@ tioca_bus_fixup(struct pcibus_bussoft *prom_bussoft, struct pci_controller *cont 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; + tioca_common->ca_kernel_private = (u64) tioca_kern; bus = pci_find_bus(tioca_common->ca_common.bs_persist_segment, tioca_common->ca_common.bs_persist_busnum); diff --git a/arch/ia64/sn/pci/tioce_provider.c b/arch/ia64/sn/pci/tioce_provider.c index dda196c9e32..e52831ed93e 100644 --- a/arch/ia64/sn/pci/tioce_provider.c +++ b/arch/ia64/sn/pci/tioce_provider.c @@ -81,10 +81,10 @@ * 61 - 0 since this is not an MSI transaction * 60:54 - reserved, MBZ */ -static uint64_t +static u64 tioce_dma_d64(unsigned long ct_addr) { - uint64_t bus_addr; + u64 bus_addr; bus_addr = ct_addr | (1UL << 63); @@ -141,9 +141,9 @@ pcidev_to_tioce(struct pci_dev *pdev, struct tioce **base, * length, and if enough resources exist, fill in the ATE's and construct a * tioce_dmamap struct to track the mapping. */ -static uint64_t +static u64 tioce_alloc_map(struct tioce_kernel *ce_kern, int type, int port, - uint64_t ct_addr, int len) + u64 ct_addr, int len) { int i; int j; @@ -152,11 +152,11 @@ tioce_alloc_map(struct tioce_kernel *ce_kern, int type, int port, int entries; int nates; int pagesize; - uint64_t *ate_shadow; - uint64_t *ate_reg; - uint64_t addr; + u64 *ate_shadow; + u64 *ate_reg; + u64 addr; struct tioce *ce_mmr; - uint64_t bus_base; + u64 bus_base; struct tioce_dmamap *map; ce_mmr = (struct tioce *)ce_kern->ce_common->ce_pcibus.bs_base; @@ -224,7 +224,7 @@ tioce_alloc_map(struct tioce_kernel *ce_kern, int type, int port, addr = ct_addr; for (j = 0; j < nates; j++) { - uint64_t ate; + u64 ate; ate = ATE_MAKE(addr, pagesize); ate_shadow[i + j] = ate; @@ -252,15 +252,15 @@ tioce_alloc_map(struct tioce_kernel *ce_kern, int type, int port, * * Map @paddr into 32-bit bus space of the CE associated with @pcidev_info. */ -static uint64_t -tioce_dma_d32(struct pci_dev *pdev, uint64_t ct_addr) +static u64 +tioce_dma_d32(struct pci_dev *pdev, u64 ct_addr) { int dma_ok; int port; struct tioce *ce_mmr; struct tioce_kernel *ce_kern; - uint64_t ct_upper; - uint64_t ct_lower; + u64 ct_upper; + u64 ct_lower; dma_addr_t bus_addr; ct_upper = ct_addr & ~0x3fffffffUL; @@ -269,7 +269,7 @@ tioce_dma_d32(struct pci_dev *pdev, uint64_t ct_addr) pcidev_to_tioce(pdev, &ce_mmr, &ce_kern, &port); if (ce_kern->ce_port[port].dirmap_refcnt == 0) { - uint64_t tmp; + u64 tmp; ce_kern->ce_port[port].dirmap_shadow = ct_upper; writeq(ct_upper, &ce_mmr->ce_ure_dir_map[port]); @@ -295,10 +295,10 @@ tioce_dma_d32(struct pci_dev *pdev, uint64_t ct_addr) * Given a TIOCE bus address, set the appropriate bit to indicate barrier * attributes. */ -static uint64_t -tioce_dma_barrier(uint64_t bus_addr, int on) +static u64 +tioce_dma_barrier(u64 bus_addr, int on) { - uint64_t barrier_bit; + u64 barrier_bit; /* barrier not supported in M40/M40S mode */ if (TIOCE_M40_ADDR(bus_addr) || TIOCE_M40S_ADDR(bus_addr)) @@ -351,7 +351,7 @@ tioce_dma_unmap(struct pci_dev *pdev, dma_addr_t bus_addr, int dir) list_for_each_entry(map, &ce_kern->ce_dmamap_list, ce_dmamap_list) { - uint64_t last; + u64 last; last = map->pci_start + map->nbytes - 1; if (bus_addr >= map->pci_start && bus_addr <= last) @@ -385,17 +385,17 @@ tioce_dma_unmap(struct pci_dev *pdev, dma_addr_t bus_addr, int dir) * This is the main wrapper for mapping host physical pages to CE PCI space. * The mapping mode used is based on the device's dma_mask. */ -static uint64_t -tioce_do_dma_map(struct pci_dev *pdev, uint64_t paddr, size_t byte_count, +static u64 +tioce_do_dma_map(struct pci_dev *pdev, u64 paddr, size_t byte_count, int barrier) { unsigned long flags; - uint64_t ct_addr; - uint64_t mapaddr = 0; + u64 ct_addr; + u64 mapaddr = 0; struct tioce_kernel *ce_kern; struct tioce_dmamap *map; int port; - uint64_t dma_mask; + u64 dma_mask; dma_mask = (barrier) ? pdev->dev.coherent_dma_mask : pdev->dma_mask; @@ -425,7 +425,7 @@ tioce_do_dma_map(struct pci_dev *pdev, uint64_t paddr, size_t byte_count, * address bits than this device can support. */ list_for_each_entry(map, &ce_kern->ce_dmamap_list, ce_dmamap_list) { - uint64_t last; + u64 last; last = map->ct_start + map->nbytes - 1; if (ct_addr >= map->ct_start && @@ -501,8 +501,8 @@ dma_map_done: * Simply call tioce_do_dma_map() to create a map with the barrier bit clear * in the address. */ -static uint64_t -tioce_dma(struct pci_dev *pdev, uint64_t paddr, size_t byte_count) +static u64 +tioce_dma(struct pci_dev *pdev, u64 paddr, size_t byte_count) { return tioce_do_dma_map(pdev, paddr, byte_count, 0); } @@ -515,8 +515,8 @@ tioce_dma(struct pci_dev *pdev, uint64_t paddr, size_t byte_count) * * Simply call tioce_do_dma_map() to create a map with the barrier bit set * in the address. - */ static uint64_t -tioce_dma_consistent(struct pci_dev *pdev, uint64_t paddr, size_t byte_count) + */ static u64 +tioce_dma_consistent(struct pci_dev *pdev, u64 paddr, size_t byte_count) { return tioce_do_dma_map(pdev, paddr, byte_count, 1); } @@ -551,7 +551,7 @@ tioce_error_intr_handler(int irq, void *arg, struct pt_regs *pt) tioce_kern_init(struct tioce_common *tioce_common) { int i; - uint32_t tmp; + u32 tmp; struct tioce *tioce_mmr; struct tioce_kernel *tioce_kern; @@ -563,7 +563,7 @@ tioce_kern_init(struct tioce_common *tioce_common) tioce_kern->ce_common = tioce_common; spin_lock_init(&tioce_kern->ce_lock); INIT_LIST_HEAD(&tioce_kern->ce_dmamap_list); - tioce_common->ce_kernel_private = (uint64_t) tioce_kern; + tioce_common->ce_kernel_private = (u64) tioce_kern; /* * Determine the secondary bus number of the port2 logical PPB. @@ -575,7 +575,7 @@ tioce_kern_init(struct tioce_common *tioce_common) raw_pci_ops->read(tioce_common->ce_pcibus.bs_persist_segment, tioce_common->ce_pcibus.bs_persist_busnum, PCI_DEVFN(2, 0), PCI_SECONDARY_BUS, 1, &tmp); - tioce_kern->ce_port1_secondary = (uint8_t) tmp; + tioce_kern->ce_port1_secondary = (u8) tmp; /* * Set PMU pagesize to the largest size available, and zero out @@ -615,7 +615,7 @@ tioce_force_interrupt(struct sn_irq_info *sn_irq_info) struct pcidev_info *pcidev_info; struct tioce_common *ce_common; struct tioce *ce_mmr; - uint64_t force_int_val; + u64 force_int_val; if (!sn_irq_info->irq_bridge) return; @@ -687,7 +687,7 @@ tioce_target_interrupt(struct sn_irq_info *sn_irq_info) struct tioce_common *ce_common; struct tioce *ce_mmr; int bit; - uint64_t vector; + u64 vector; pcidev_info = (struct pcidev_info *)sn_irq_info->irq_pciioinfo; if (!pcidev_info) @@ -699,7 +699,7 @@ tioce_target_interrupt(struct sn_irq_info *sn_irq_info) bit = sn_irq_info->irq_int_bit; __sn_setq_relaxed(&ce_mmr->ce_adm_int_mask, (1UL << bit)); - vector = (uint64_t)sn_irq_info->irq_irq << INTR_VECTOR_SHFT; + vector = (u64)sn_irq_info->irq_irq << INTR_VECTOR_SHFT; vector |= sn_irq_info->irq_xtalkaddr; writeq(vector, &ce_mmr->ce_adm_int_dest[bit]); __sn_clrq_relaxed(&ce_mmr->ce_adm_int_mask, (1UL << bit)); |