diff options
43 files changed, 1399 insertions, 636 deletions
diff --git a/arch/frv/Kconfig b/arch/frv/Kconfig index 8a5bd7a9c6f..b86e19c9b5b 100644 --- a/arch/frv/Kconfig +++ b/arch/frv/Kconfig @@ -7,6 +7,7 @@ config FRV default y select HAVE_IDE select HAVE_ARCH_TRACEHOOK + select HAVE_PERF_COUNTERS config ZONE_DMA bool diff --git a/arch/frv/include/asm/atomic.h b/arch/frv/include/asm/atomic.h index 0409d981fd3..00a57af79af 100644 --- a/arch/frv/include/asm/atomic.h +++ b/arch/frv/include/asm/atomic.h @@ -121,10 +121,72 @@ static inline void atomic_dec(atomic_t *v) #define atomic_dec_and_test(v) (atomic_sub_return(1, (v)) == 0) #define atomic_inc_and_test(v) (atomic_add_return(1, (v)) == 0) +/* + * 64-bit atomic ops + */ +typedef struct { + volatile long long counter; +} atomic64_t; + +#define ATOMIC64_INIT(i) { (i) } + +static inline long long atomic64_read(atomic64_t *v) +{ + long long counter; + + asm("ldd%I1 %M1,%0" + : "=e"(counter) + : "m"(v->counter)); + return counter; +} + +static inline void atomic64_set(atomic64_t *v, long long i) +{ + asm volatile("std%I0 %1,%M0" + : "=m"(v->counter) + : "e"(i)); +} + +extern long long atomic64_inc_return(atomic64_t *v); +extern long long atomic64_dec_return(atomic64_t *v); +extern long long atomic64_add_return(long long i, atomic64_t *v); +extern long long atomic64_sub_return(long long i, atomic64_t *v); + +static inline long long atomic64_add_negative(long long i, atomic64_t *v) +{ + return atomic64_add_return(i, v) < 0; +} + +static inline void atomic64_add(long long i, atomic64_t *v) +{ + atomic64_add_return(i, v); +} + +static inline void atomic64_sub(long long i, atomic64_t *v) +{ + atomic64_sub_return(i, v); +} + +static inline void atomic64_inc(atomic64_t *v) +{ + atomic64_inc_return(v); +} + +static inline void atomic64_dec(atomic64_t *v) +{ + atomic64_dec_return(v); +} + +#define atomic64_sub_and_test(i,v) (atomic64_sub_return((i), (v)) == 0) +#define atomic64_dec_and_test(v) (atomic64_dec_return((v)) == 0) +#define atomic64_inc_and_test(v) (atomic64_inc_return((v)) == 0) + /*****************************************************************************/ /* * exchange value with memory */ +extern uint64_t __xchg_64(uint64_t i, volatile void *v); + #ifndef CONFIG_FRV_OUTOFLINE_ATOMIC_OPS #define xchg(ptr, x) \ @@ -174,8 +236,10 @@ extern uint32_t __xchg_32(uint32_t i, volatile void *v); #define tas(ptr) (xchg((ptr), 1)) -#define atomic_cmpxchg(v, old, new) (cmpxchg(&((v)->counter), old, new)) -#define atomic_xchg(v, new) (xchg(&((v)->counter), new)) +#define atomic_cmpxchg(v, old, new) (cmpxchg(&(v)->counter, old, new)) +#define atomic_xchg(v, new) (xchg(&(v)->counter, new)) +#define atomic64_cmpxchg(v, old, new) (__cmpxchg_64(old, new, &(v)->counter)) +#define atomic64_xchg(v, new) (__xchg_64(new, &(v)->counter)) static __inline__ int atomic_add_unless(atomic_t *v, int a, int u) { diff --git a/arch/frv/include/asm/perf_counter.h b/arch/frv/include/asm/perf_counter.h new file mode 100644 index 00000000000..ccf726e61b2 --- /dev/null +++ b/arch/frv/include/asm/perf_counter.h @@ -0,0 +1,17 @@ +/* FRV performance counter support + * + * Copyright (C) 2009 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ + +#ifndef _ASM_PERF_COUNTER_H +#define _ASM_PERF_COUNTER_H + +#define PERF_COUNTER_INDEX_OFFSET 0 + +#endif /* _ASM_PERF_COUNTER_H */ diff --git a/arch/frv/include/asm/system.h b/arch/frv/include/asm/system.h index 7742ec000cc..efd22d9077a 100644 --- a/arch/frv/include/asm/system.h +++ b/arch/frv/include/asm/system.h @@ -208,6 +208,8 @@ extern void free_initmem(void); * - if (*ptr == test) then orig = *ptr; *ptr = test; * - if (*ptr != test) then orig = *ptr; */ +extern uint64_t __cmpxchg_64(uint64_t test, uint64_t new, volatile uint64_t *v); + #ifndef CONFIG_FRV_OUTOFLINE_ATOMIC_OPS #define cmpxchg(ptr, test, new) \ diff --git a/arch/frv/kernel/frv_ksyms.c b/arch/frv/kernel/frv_ksyms.c index 0316b3c50ef..a89803b58b9 100644 --- a/arch/frv/kernel/frv_ksyms.c +++ b/arch/frv/kernel/frv_ksyms.c @@ -67,6 +67,10 @@ EXPORT_SYMBOL(atomic_sub_return); EXPORT_SYMBOL(__xchg_32); EXPORT_SYMBOL(__cmpxchg_32); #endif +EXPORT_SYMBOL(atomic64_add_return); +EXPORT_SYMBOL(atomic64_sub_return); +EXPORT_SYMBOL(__xchg_64); +EXPORT_SYMBOL(__cmpxchg_64); EXPORT_SYMBOL(__debug_bug_printk); EXPORT_SYMBOL(__delay_loops_MHz); diff --git a/arch/frv/lib/Makefile b/arch/frv/lib/Makefile index 08be305c9f4..0a377210c89 100644 --- a/arch/frv/lib/Makefile +++ b/arch/frv/lib/Makefile @@ -4,5 +4,5 @@ lib-y := \ __ashldi3.o __lshrdi3.o __muldi3.o __ashrdi3.o __negdi2.o __ucmpdi2.o \ - checksum.o memcpy.o memset.o atomic-ops.o \ - outsl_ns.o outsl_sw.o insl_ns.o insl_sw.o cache.o + checksum.o memcpy.o memset.o atomic-ops.o atomic64-ops.o \ + outsl_ns.o outsl_sw.o insl_ns.o insl_sw.o cache.o perf_counter.o diff --git a/arch/frv/lib/atomic-ops.S b/arch/frv/lib/atomic-ops.S index ee0ac905fb0..5e9e6ab5dd0 100644 --- a/arch/frv/lib/atomic-ops.S +++ b/arch/frv/lib/atomic-ops.S @@ -163,11 +163,10 @@ __cmpxchg_32: ld.p @(gr11,gr0),gr8 orcr cc7,cc7,cc3 subcc gr8,gr9,gr7,icc0 - bne icc0,#0,1f + bnelr icc0,#0 cst.p gr10,@(gr11,gr0) ,cc3,#1 corcc gr29,gr29,gr0 ,cc3,#1 beq icc3,#0,0b -1: bralr .size __cmpxchg_32, .-__cmpxchg_32 diff --git a/arch/frv/lib/atomic64-ops.S b/arch/frv/lib/atomic64-ops.S new file mode 100644 index 00000000000..b6194eeac12 --- /dev/null +++ b/arch/frv/lib/atomic64-ops.S @@ -0,0 +1,162 @@ +/* kernel atomic64 operations + * + * For an explanation of how atomic ops work in this arch, see: + * Documentation/frv/atomic-ops.txt + * + * Copyright (C) 2009 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include <asm/spr-regs.h> + + .text + .balign 4 + + +############################################################################### +# +# long long atomic64_inc_return(atomic64_t *v) +# +############################################################################### + .globl atomic64_inc_return + .type atomic64_inc_return,@function +atomic64_inc_return: + or.p gr8,gr8,gr10 +0: + orcc gr0,gr0,gr0,icc3 /* set ICC3.Z */ + ckeq icc3,cc7 + ldd.p @(gr10,gr0),gr8 /* LDD.P/ORCR must be atomic */ + orcr cc7,cc7,cc3 /* set CC3 to true */ + addicc gr9,#1,gr9,icc0 + addxi gr8,#0,gr8,icc0 + cstd.p gr8,@(gr10,gr0) ,cc3,#1 + corcc gr29,gr29,gr0 ,cc3,#1 /* clear ICC3.Z if store happens */ + beq icc3,#0,0b + bralr + + .size atomic64_inc_return, .-atomic64_inc_return + +############################################################################### +# +# long long atomic64_dec_return(atomic64_t *v) +# +############################################################################### + .globl atomic64_dec_return + .type atomic64_dec_return,@function +atomic64_dec_return: + or.p gr8,gr8,gr10 +0: + orcc gr0,gr0,gr0,icc3 /* set ICC3.Z */ + ckeq icc3,cc7 + ldd.p @(gr10,gr0),gr8 /* LDD.P/ORCR must be atomic */ + orcr cc7,cc7,cc3 /* set CC3 to true */ + subicc gr9,#1,gr9,icc0 + subxi gr8,#0,gr8,icc0 + cstd.p gr8,@(gr10,gr0) ,cc3,#1 + corcc gr29,gr29,gr0 ,cc3,#1 /* clear ICC3.Z if store happens */ + beq icc3,#0,0b + bralr + + .size atomic64_dec_return, .-atomic64_dec_return + +############################################################################### +# +# long long atomic64_add_return(long long i, atomic64_t *v) +# +############################################################################### + .globl atomic64_add_return + .type atomic64_add_return,@function +atomic64_add_return: + or.p gr8,gr8,gr4 + or gr9,gr9,gr5 +0: + orcc gr0,gr0,gr0,icc3 /* set ICC3.Z */ + ckeq icc3,cc7 + ldd.p @(gr10,gr0),gr8 /* LDD.P/ORCR must be atomic */ + orcr cc7,cc7,cc3 /* set CC3 to true */ + addcc gr9,gr5,gr9,icc0 + addx gr8,gr4,gr8,icc0 + cstd.p gr8,@(gr10,gr0) ,cc3,#1 + corcc gr29,gr29,gr0 ,cc3,#1 /* clear ICC3.Z if store happens */ + beq icc3,#0,0b + bralr + + .size atomic64_add_return, .-atomic64_add_return + +############################################################################### +# +# long long atomic64_sub_return(long long i, atomic64_t *v) +# +############################################################################### + .globl atomic64_sub_return + .type atomic64_sub_return,@function +atomic64_sub_return: + or.p gr8,gr8,gr4 + or gr9,gr9,gr5 +0: + orcc gr0,gr0,gr0,icc3 /* set ICC3.Z */ + ckeq icc3,cc7 + ldd.p @(gr10,gr0),gr8 /* LDD.P/ORCR must be atomic */ + orcr cc7,cc7,cc3 /* set CC3 to true */ + subcc gr9,gr5,gr9,icc0 + subx gr8,gr4,gr8,icc0 + cstd.p gr8,@(gr10,gr0) ,cc3,#1 + corcc gr29,gr29,gr0 ,cc3,#1 /* clear ICC3.Z if store happens */ + beq icc3,#0,0b + bralr + + .size atomic64_sub_return, .-atomic64_sub_return + +############################################################################### +# +# uint64_t __xchg_64(uint64_t i, uint64_t *v) +# +############################################################################### + .globl __xchg_64 + .type __xchg_64,@function +__xchg_64: + or.p gr8,gr8,gr4 + or gr9,gr9,gr5 +0: + orcc gr0,gr0,gr0,icc3 /* set ICC3.Z */ + ckeq icc3,cc7 + ldd.p @(gr10,gr0),gr8 /* LDD.P/ORCR must be atomic */ + orcr cc7,cc7,cc3 /* set CC3 to true */ + cstd.p gr4,@(gr10,gr0) ,cc3,#1 + corcc gr29,gr29,gr0 ,cc3,#1 /* clear ICC3.Z if store happens */ + beq icc3,#0,0b + bralr + + .size __xchg_64, .-__xchg_64 + +############################################################################### +# +# uint64_t __cmpxchg_64(uint64_t test, uint64_t new, uint64_t *v) +# +############################################################################### + .globl __cmpxchg_64 + .type __cmpxchg_64,@function +__cmpxchg_64: + or.p gr8,gr8,gr4 + or gr9,gr9,gr5 +0: + orcc gr0,gr0,gr0,icc3 /* set ICC3.Z */ + ckeq icc3,cc7 + ldd.p @(gr12,gr0),gr8 /* LDD.P/ORCR must be atomic */ + orcr cc7,cc7,cc3 + subcc gr8,gr4,gr0,icc0 + subcc.p gr9,gr5,gr0,icc1 + bnelr icc0,#0 + bnelr icc1,#0 + cstd.p gr10,@(gr12,gr0) ,cc3,#1 + corcc gr29,gr29,gr0 ,cc3,#1 /* clear ICC3.Z if store happens */ + beq icc3,#0,0b + bralr + + .size __cmpxchg_64, .-__cmpxchg_64 + diff --git a/arch/frv/lib/perf_counter.c b/arch/frv/lib/perf_counter.c new file mode 100644 index 00000000000..2000feecd57 --- /dev/null +++ b/arch/frv/lib/perf_counter.c @@ -0,0 +1,19 @@ +/* Performance counter handling + * + * Copyright (C) 2009 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ + +#include <linux/perf_counter.h> + +/* + * mark the performance counter as pending + */ +void set_perf_counter_pending(void) +{ +} diff --git a/arch/sh/Kconfig.debug b/arch/sh/Kconfig.debug index 8ece0b5bd02..39224b57c6e 100644 --- a/arch/sh/Kconfig.debug +++ b/arch/sh/Kconfig.debug @@ -61,10 +61,6 @@ config EARLY_PRINTK select both the EARLY_SCIF_CONSOLE and SH_STANDARD_BIOS, using the kernel command line option to toggle back and forth. -config DEBUG_BOOTMEM - depends on DEBUG_KERNEL - bool "Debug BOOTMEM initialization" - config DEBUG_STACKOVERFLOW bool "Check for stack overflows" depends on DEBUG_KERNEL && SUPERH32 diff --git a/arch/sh/boards/mach-se/7206/io.c b/arch/sh/boards/mach-se/7206/io.c index 9c3a33210d6..180455642a4 100644 --- a/arch/sh/boards/mach-se/7206/io.c +++ b/arch/sh/boards/mach-se/7206/io.c @@ -50,7 +50,7 @@ unsigned char se7206_inb_p(unsigned long port) unsigned short se7206_inw(unsigned long port) { - return *port2adr(port);; + return *port2adr(port); } void se7206_outb(unsigned char value, unsigned long port) diff --git a/arch/sh/boards/mach-se/7724/setup.c b/arch/sh/boards/mach-se/7724/setup.c index 9cd04bd558b..c050a8d76df 100644 --- a/arch/sh/boards/mach-se/7724/setup.c +++ b/arch/sh/boards/mach-se/7724/setup.c @@ -23,6 +23,8 @@ #include <media/sh_mobile_ceu.h> #include <asm/io.h> #include <asm/heartbeat.h> +#include <asm/sh_eth.h> +#include <asm/clock.h> #include <asm/sh_keysc.h> #include <cpu/sh7724.h> #include <mach-se/mach/se7724.h> @@ -272,6 +274,34 @@ static struct platform_device keysc_device = { }, }; +/* SH Eth */ +static struct resource sh_eth_resources[] = { + [0] = { + .start = SH_ETH_ADDR, + .end = SH_ETH_ADDR + 0x1FC, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = 91, + .flags = IORESOURCE_IRQ | IORESOURCE_IRQ_HIGHLEVEL, + }, +}; + +struct sh_eth_plat_data sh_eth_plat = { + .phy = 0x1f, /* SMSC LAN8187 */ + .edmac_endian = EDMAC_LITTLE_ENDIAN, +}; + +static struct platform_device sh_eth_device = { + .name = "sh-eth", + .id = 0, + .dev = { + .platform_data = &sh_eth_plat, + }, + .num_resources = ARRAY_SIZE(sh_eth_resources), + .resource = sh_eth_resources, +}; + static struct platform_device *ms7724se_devices[] __initdata = { &heartbeat_device, &smc91x_eth_device, @@ -280,8 +310,57 @@ static struct platform_device *ms7724se_devices[] __initdata = { &ceu0_device, &ceu1_device, &keysc_device, + &sh_eth_device, }; +#define EEPROM_OP 0xBA206000 +#define EEPROM_ADR 0xBA206004 +#define EEPROM_DATA 0xBA20600C +#define EEPROM_STAT 0xBA206010 +#define EEPROM_STRT 0xBA206014 +static int __init sh_eth_is_eeprom_ready(void) +{ + int t = 10000; + + while (t--) { + if (!ctrl_inw(EEPROM_STAT)) + return 1; + cpu_relax(); + } + + printk(KERN_ERR "ms7724se can not access to eeprom\n"); + return 0; +} + +static void __init sh_eth_init(void) +{ + int i; + u16 mac[3]; + + /* check EEPROM status */ + if (!sh_eth_is_eeprom_ready()) + return; + + /* read MAC addr from EEPROM */ + for (i = 0 ; i < 3 ; i++) { + ctrl_outw(0x0, EEPROM_OP); /* read */ + ctrl_outw(i*2, EEPROM_ADR); + ctrl_outw(0x1, EEPROM_STRT); + if (!sh_eth_is_eeprom_ready()) + return; + + mac[i] = ctrl_inw(EEPROM_DATA); + mac[i] = ((mac[i] & 0xFF) << 8) | (mac[i] >> 8); /* swap */ + } + + /* reset sh-eth */ + ctrl_outl(0x1, SH_ETH_ADDR + 0x0); + + /* set MAC addr */ + ctrl_outl(((mac[0] << 16) | (mac[1])), SH_ETH_MAHR); + ctrl_outl((mac[2]), SH_ETH_MALR); +} + #define SW4140 0xBA201000 #define FPGA_OUT 0xBA200400 #define PORT_HIZA 0xA4050158 @@ -302,7 +381,8 @@ static int __init devices_setup(void) ctrl_outw(ctrl_inw(FPGA_OUT) & ~((1 << 1) | /* LAN */ (1 << 6) | /* VIDEO DAC */ - (1 << 12)), /* USB0 */ + (1 << 12) | /* USB0 */ + (1 << 14)), /* RMII */ FPGA_OUT); /* enable IRQ 0,1,2 */ @@ -374,7 +454,7 @@ static int __init devices_setup(void) gpio_request(GPIO_FN_VIO0_CLK, NULL); gpio_request(GPIO_FN_VIO0_FLD, NULL); gpio_request(GPIO_FN_VIO0_HD, NULL); - platform_resource_setup_memory(&ceu0_device, "ceu", 4 << 20); + platform_resource_setup_memory(&ceu0_device, "ceu0", 4 << 20); /* enable CEU1 */ gpio_request(GPIO_FN_VIO1_D7, NULL); @@ -389,7 +469,7 @@ static int __init devices_setup(void) gpio_request(GPIO_FN_VIO1_HD, NULL); gpio_request(GPIO_FN_VIO1_VD, NULL); gpio_request(GPIO_FN_VIO1_CLK, NULL); - platform_resource_setup_memory(&ceu1_device, "ceu", 4 << 20); + platform_resource_setup_memory(&ceu1_device, "ceu1", 4 << 20); /* KEYSC */ gpio_request(GPIO_FN_KEYOUT5_IN5, NULL); @@ -404,6 +484,28 @@ static int __init devices_setup(void) gpio_request(GPIO_FN_KEYOUT1, NULL); gpio_request(GPIO_FN_KEYOUT0, NULL); + /* + * enable SH-Eth + * + * please remove J33 pin from your board !! + * + * ms7724 board should not use GPIO_FN_LNKSTA pin + * So, This time PTX5 is set to input pin + */ + gpio_request(GPIO_FN_RMII_RXD0, NULL); + gpio_request(GPIO_FN_RMII_RXD1, NULL); + gpio_request(GPIO_FN_RMII_TXD0, NULL); + gpio_request(GPIO_FN_RMII_TXD1, NULL); + gpio_request(GPIO_FN_RMII_REF_CLK, NULL); + gpio_request(GPIO_FN_RMII_TX_EN, NULL); + gpio_request(GPIO_FN_RMII_RX_ER, NULL); + gpio_request(GPIO_FN_RMII_CRS_DV, NULL); + gpio_request(GPIO_FN_MDIO, NULL); + gpio_request(GPIO_FN_MDC, NULL); + gpio_request(GPIO_PTX5, NULL); + gpio_direction_input(GPIO_PTX5); + sh_eth_init(); + if (sw & SW41_B) { /* SVGA */ lcdc_info.ch[0].lcd_cfg.xres = 800; @@ -437,7 +539,7 @@ static int __init devices_setup(void) } return platform_add_devices(ms7724se_devices, - ARRAY_SIZE(ms7724se_devices)); + ARRAY_SIZE(ms7724se_devices)); } device_initcall(devices_setup); diff --git a/arch/sh/configs/migor_defconfig b/arch/sh/configs/migor_defconfig index da627d22c00..b18cfd39cac 100644 --- a/arch/sh/configs/migor_defconfig +++ b/arch/sh/configs/migor_defconfig @@ -309,7 +309,7 @@ CONFIG_ZERO_PAGE_OFFSET=0x00001000 CONFIG_BOOT_LINK_OFFSET=0x00800000 CONFIG_ENTRY_OFFSET=0x00001000 CONFIG_CMDLINE_BOOL=y -CONFIG_CMDLINE="console=ttySC0,115200 earlyprintk=serial ip=on root=/dev/nfs ip=dhcp" +CONFIG_CMDLINE="console=tty0 console=ttySC0,115200 earlyprintk=serial ip=on root=/dev/nfs ip=dhcp" # # Bus options @@ -858,7 +858,35 @@ CONFIG_VIDEO_SH_MOBILE_CEU=y # # CONFIG_VGASTATE is not set # CONFIG_VIDEO_OUTPUT_CONTROL is not set -# CONFIG_FB is not set +CONFIG_FB=y +# CONFIG_FIRMWARE_EDID is not set +# CONFIG_FB_DDC is not set +# CONFIG_FB_BOOT_VESA_SUPPORT is not set +# CONFIG_FB_CFB_FILLRECT is not set +# CONFIG_FB_CFB_COPYAREA is not set +# CONFIG_FB_CFB_IMAGEBLIT is not set +# CONFIG_FB_CFB_REV_PIXELS_IN_BYTE is not set +CONFIG_FB_SYS_FILLRECT=y +CONFIG_FB_SYS_COPYAREA=y +CONFIG_FB_SYS_IMAGEBLIT=y +# CONFIG_FB_FOREIGN_ENDIAN is not set +CONFIG_FB_SYS_FOPS=y +CONFIG_FB_DEFERRED_IO=y +# CONFIG_FB_SVGALIB is not set +# CONFIG_FB_MACMODES is not set +# CONFIG_FB_BACKLIGHT is not set +# CONFIG_FB_MODE_HELPERS is not set +# CONFIG_FB_TILEBLITTING is not set + +# +# Frame buffer hardware drivers +# +# CONFIG_FB_S1D13XXX is not set +CONFIG_FB_SH_MOBILE_LCDC=y +# CONFIG_FB_VIRTUAL is not set +# CONFIG_FB_METRONOME is not set +# CONFIG_FB_MB862XX is not set +# CONFIG_FB_BROADSHEET is not set # CONFIG_BACKLIGHT_LCD_SUPPORT is not set # @@ -870,6 +898,27 @@ CONFIG_VIDEO_SH_MOBILE_CEU=y # Console display driver support # CONFIG_DUMMY_CONSOLE=y +CONFIG_FRAMEBUFFER_CONSOLE=y +CONFIG_FRAMEBUFFER_CONSOLE_DETECT_PRIMARY=y +# CONFIG_FRAMEBUFFER_CONSOLE_ROTATION is not set +CONFIG_FONTS=y +# CONFIG_FONT_8x8 is not set +# CONFIG_FONT_8x16 is not set +# CONFIG_FONT_6x11 is not set +# CONFIG_FONT_7x14 is not set +# CONFIG_FONT_PEARL_8x8 is not set +# CONFIG_FONT_ACORN_8x8 is not set +CONFIG_FONT_MINI_4x6=y +# CONFIG_FONT_SUN8x16 is not set +# CONFIG_FONT_SUN12x22 is not set +# CONFIG_FONT_10x18 is not set +CONFIG_LOGO=y +# CONFIG_LOGO_LINUX_MONO is not set +# CONFIG_LOGO_LINUX_VGA16 is not set +# CONFIG_LOGO_LINUX_CLUT224 is not set +# CONFIG_LOGO_SUPERH_MONO is not set +CONFIG_LOGO_SUPERH_VGA16=y +# CONFIG_LOGO_SUPERH_CLUT224 is not set # CONFIG_SOUND is not set CONFIG_HID_SUPPORT=y CONFIG_HID=y diff --git a/arch/sh/configs/se7724_defconfig b/arch/sh/configs/se7724_defconfig index 3840270283e..3ee783a0a07 100644 --- a/arch/sh/configs/se7724_defconfig +++ b/arch/sh/configs/se7724_defconfig @@ -1,7 +1,7 @@ # # Automatically generated make config: don't edit # Linux kernel version: 2.6.30 -# Thu Jun 18 16:09:05 2009 +# Mon Jun 29 16:28:43 2009 # CONFIG_SUPERH=y CONFIG_SUPERH32=y @@ -14,6 +14,7 @@ CONFIG_GENERIC_HWEIGHT=y CONFIG_GENERIC_HARDIRQS=y CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ=y CONFIG_GENERIC_IRQ_PROBE=y +CONFIG_IRQ_PER_CPU=y CONFIG_GENERIC_GPIO=y CONFIG_GENERIC_TIME=y CONFIG_GENERIC_CLOCKEVENTS=y @@ -28,7 +29,9 @@ CONFIG_HAVE_LATENCYTOP_SUPPORT=y # CONFIG_ARCH_HAS_ILOG2_U64 is not set CONFIG_ARCH_NO_VIRT_TO_BUS=y CONFIG_ARCH_HAS_DEFAULT_IDLE=y +CONFIG_ARCH_HAS_CPU_IDLE_WAIT=y CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config" +CONFIG_CONSTRUCTORS=y # # General setup @@ -88,10 +91,12 @@ CONFIG_TIMERFD=y CONFIG_EVENTFD=y CONFIG_SHMEM=y CONFIG_AIO=y +CONFIG_HAVE_PERF_COUNTERS=y # # Performance Counters # +# CONFIG_PERF_COUNTERS is not set CONFIG_VM_EVENT_COUNTERS=y # CONFIG_STRIP_ASM_SYMS is not set CONFIG_COMPAT_BRK=y @@ -107,6 +112,10 @@ CONFIG_HAVE_KRETPROBES=y CONFIG_HAVE_ARCH_TRACEHOOK=y CONFIG_HAVE_CLK=y CONFIG_HAVE_DMA_API_DEBUG=y + +# +# GCOV-based kernel profiling +# # CONFIG_SLOW_WORK is not set CONFIG_HAVE_GENERIC_DMA_COHERENT=y CONFIG_SLABINFO=y @@ -119,7 +128,7 @@ CONFIG_MODULE_UNLOAD=y # CONFIG_MODVERSIONS is not set # CONFIG_MODULE_SRCVERSION_ALL is not set CONFIG_BLOCK=y -# CONFIG_LBD is not set +CONFIG_LBDAF=y # CONFIG_BLK_DEV_BSG is not set # CONFIG_BLK_DEV_INTEGRITY is not set @@ -584,7 +593,6 @@ CONFIG_SCSI_WAIT_SCAN=m # CONFIG_SCSI_SRP_ATTRS is not set CONFIG_SCSI_LOWLEVEL=y # CONFIG_ISCSI_TCP is not set -# CONFIG_SCSI_BNX2_ISCSI is not set # CONFIG_LIBFC is not set # CONFIG_LIBFCOE is not set # CONFIG_SCSI_DEBUG is not set @@ -624,7 +632,7 @@ CONFIG_NET_ETHERNET=y CONFIG_MII=y # CONFIG_AX88796 is not set # CONFIG_STNIC is not set -# CONFIG_SH_ETH is not set +CONFIG_SH_ETH=y CONFIG_SMC91X=y # CONFIG_ENC28J60 is not set # CONFIG_ETHOC is not set @@ -801,6 +809,11 @@ CONFIG_SPI_BITBANG=y # # CONFIG_SPI_SPIDEV is not set # CONFIG_SPI_TLE62X0 is not set + +# +# PPS support +# +# CONFIG_PPS is not set CONFIG_ARCH_REQUIRE_GPIOLIB=y CONFIG_GPIOLIB=y # CONFIG_GPIO_SYSFS is not set @@ -851,6 +864,8 @@ CONFIG_SSB_POSSIBLE=y # CONFIG_MFD_WM8400 is not set # CONFIG_MFD_WM8350_I2C is not set # CONFIG_MFD_PCF50633 is not set +# CONFIG_AB3100_CORE is not set +# CONFIG_EZX_PCAP is not set # CONFIG_REGULATOR is not set CONFIG_MEDIA_SUPPORT=y @@ -1196,6 +1211,7 @@ CONFIG_RTC_DRV_PCF8563=y # CONFIG_RTC_DRV_S35390A is not set # CONFIG_RTC_DRV_FM3130 is not set # CONFIG_RTC_DRV_RX8581 is not set +# CONFIG_RTC_DRV_RX8025 is not set # # SPI RTC drivers @@ -1260,6 +1276,7 @@ CONFIG_FS_MBCACHE=y # CONFIG_JFS_FS is not set CONFIG_FS_POSIX_ACL=y # CONFIG_XFS_FS is not set +# CONFIG_GFS2_FS is not set # CONFIG_OCFS2_FS is not set # CONFIG_BTRFS_FS is not set CONFIG_FILE_LOCKING=y diff --git a/arch/sh/include/asm/perf_counter.h b/arch/sh/include/asm/perf_counter.h index a8153c2aa6f..61c2b40c802 100644 --- a/arch/sh/include/asm/perf_counter.h +++ b/arch/sh/include/asm/perf_counter.h @@ -2,6 +2,6 @@ #define __ASM_SH_PERF_COUNTER_H /* SH only supports software counters through this interface. */ -#define set_perf_counter_pending() do { } while (0) +static inline void set_perf_counter_pending(void) {} #endif /* __ASM_SH_PERF_COUNTER_H */ diff --git a/arch/sh/include/asm/syscall_32.h b/arch/sh/include/asm/syscall_32.h index 5bc34681d99..6f83f2cc45c 100644 --- a/arch/sh/include/asm/syscall_32.h +++ b/arch/sh/include/asm/syscall_32.h @@ -3,6 +3,7 @@ #include <linux/kernel.h> #include <linux/sched.h> +#include <linux/err.h> #include <asm/ptrace.h> /* The system call number is given by the user in R3 */ diff --git a/arch/sh/include/mach-se/mach/se7724.h b/arch/sh/include/mach-se/mach/se7724.h index 74164b60d0d..29514a39d0f 100644 --- a/arch/sh/include/mach-se/mach/se7724.h +++ b/arch/sh/include/mach-se/mach/se7724.h @@ -20,6 +20,11 @@ */ #include <asm/addrspace.h> +/* SH Eth */ +#define SH_ETH_ADDR (0xA4600000) +#define SH_ETH_MAHR (SH_ETH_ADDR + 0x1C0) +#define SH_ETH_MALR (SH_ETH_ADDR + 0x1C8) + #define PA_LED (0xba203000) /* 8bit LED */ #define IRQ_MODE (0xba200010) #define IRQ0_SR (0xba200014) diff --git a/arch/sh/mm/fault_32.c b/arch/sh/mm/fault_32.c index cc8ddbdf3d7..71925946f1e 100644 --- a/arch/sh/mm/fault_32.c +++ b/arch/sh/mm/fault_32.c @@ -15,12 +15,28 @@ #include <linux/mm.h> #include <linux/hardirq.h> #include <linux/kprobes.h> -#include <linux/marker.h> +#include <linux/perf_counter.h> #include <asm/io_trapped.h> #include <asm/system.h> #include <asm/mmu_context.h> #include <asm/tlbflush.h> +static inline int notify_page_fault(struct pt_regs *regs, int trap) +{ + int ret = 0; + +#ifdef CONFIG_KPROBES + if (!user_mode(regs)) { + preempt_disable(); + if (kprobe_running() && kprobe_fault_handler(regs, trap)) + ret = 1; + preempt_enable(); + } +#endif + + return ret; +} + /* * This routine handles page faults. It determines the address, * and the problem, and then passes it off to one of the appropriate @@ -87,13 +103,16 @@ asmlinkage void __kprobes do_page_fault(struct pt_regs *regs, return; } + mm = tsk->mm; + + if (unlikely(notify_page_fault(regs, lookup_exception_vector()))) + return; + /* Only enable interrupts if they were on before the fault */ - if ((regs->sr & SR_IMASK) != SR_IMASK) { - trace_hardirqs_on(); + if ((regs->sr & SR_IMASK) != SR_IMASK) local_irq_enable(); - } - mm = tsk->mm; + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address); /* * If we're in an interrupt or have no user @@ -141,10 +160,15 @@ survive: goto do_sigbus; BUG(); } - if (fault & VM_FAULT_MAJOR) + if (fault & VM_FAULT_MAJOR) { tsk->maj_flt++; - else + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0, + regs, address); + } else { tsk->min_flt++; + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0, + regs, address); + } up_read(&mm->mmap_sem); return; @@ -245,22 +269,6 @@ do_sigbus: goto no_context; } -static inline int notify_page_fault(struct pt_regs *regs, int trap) -{ - int ret = 0; - -#ifdef CONFIG_KPROBES - if (!user_mode(regs)) { - preempt_disable(); - if (kprobe_running() && kprobe_fault_handler(regs, trap)) - ret = 1; - preempt_enable(); - } -#endif - - return ret; -} - /* * Called with interrupts disabled. */ @@ -273,12 +281,7 @@ asmlinkage int __kprobes __do_page_fault(struct pt_regs *regs, pmd_t *pmd; pte_t *pte; pte_t entry; - int ret = 0; - - if (notify_page_fault(regs, lookup_exception_vector())) - goto out; - - ret = 1; + int ret = 1; /* * We don't take page faults for P1, P2, and parts of P4, these diff --git a/arch/sh/mm/tlbflush_64.c b/arch/sh/mm/tlbflush_64.c index fcbb6e135ce..3ce40ea3482 100644 --- a/arch/sh/mm/tlbflush_64.c +++ b/arch/sh/mm/tlbflush_64.c @@ -3,7 +3,7 @@ * * Copyright (C) 2000, 2001 Paolo Alberelli * Copyright (C) 2003 Richard Curnow (/proc/tlb, bug fixes) - * Copyright (C) 2003 Paul Mundt + * Copyright (C) 2003 - 2009 Paul Mundt * * 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 @@ -20,6 +20,7 @@ #include <linux/mman.h> #include <linux/mm.h> #include <linux/smp.h> +#include <linux/perf_counter.h> #include <linux/interrupt.h> #include <asm/system.h> #include <asm/io.h> @@ -115,6 +116,8 @@ asmlinkage void do_page_fault(struct pt_regs *regs, unsigned long writeaccess, /* Not an IO address, so reenable interrupts */ local_irq_enable(); + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address); + /* * If we're in an interrupt or have no user * context, we must not take the fault.. @@ -195,10 +198,16 @@ survive: goto do_sigbus; BUG(); } - if (fault & VM_FAULT_MAJOR) + + if (fault & VM_FAULT_MAJOR) { tsk->maj_flt++; - else + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0, + regs, address); + } else { tsk->min_flt++; + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0, + regs, address); + } /* If we get here, the page fault has been handled. Do the TLB refill now from the newly-setup PTE, to avoid having to fault again right diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index d1430ef6b4f..c07f7220590 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1913,25 +1913,14 @@ config DMAR_DEFAULT_ON recommended you say N here while the DMAR code remains experimental. -config DMAR_GFX_WA - def_bool y - prompt "Support for Graphics workaround" - depends on DMAR - ---help--- - Current Graphics drivers tend to use physical address - for DMA and avoid using DMA APIs. Setting this config - option permits the IOMMU driver to set a unity map for - all the OS-visible memory. Hence the driver can continue - to use physical addresses for DMA. - config DMAR_FLOPPY_WA def_bool y depends on DMAR ---help--- - Floppy disk drivers are know to bypass DMA API calls + Floppy disk drivers are known to bypass DMA API calls thereby failing to work when IOMMU is enabled. This workaround will setup a 1:1 mapping for the first - 16M to make floppy (an ISA device) work. + 16MiB to make floppy (an ISA device) work. config INTR_REMAP bool "Support for Interrupt Remapping (EXPERIMENTAL)" diff --git a/arch/x86/include/asm/proto.h b/arch/x86/include/asm/proto.h index 49fb3ecf3bb..621f56d7312 100644 --- a/arch/x86/include/asm/proto.h +++ b/arch/x86/include/asm/proto.h @@ -22,7 +22,14 @@ extern int reboot_force; long do_arch_prctl(struct task_struct *task, int code, unsigned long addr); -#define round_up(x, y) (((x) + (y) - 1) & ~((y) - 1)) -#define round_down(x, y) ((x) & ~((y) - 1)) +/* + * This looks more complex than it should be. But we need to + * get the type for the ~ right in round_down (it needs to be + * as wide as the result!), and we want to evaluate the macro + * arguments just once each. + */ +#define __round_mask(x,y) ((__typeof__(x))((y)-1)) +#define round_up(x,y) ((((x)-1) | __round_mask(x,y))+1) +#define round_down(x,y) ((x) & ~__round_mask(x,y)) #endif /* _ASM_X86_PROTO_H */ diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 7271fa33d79..c4ca89d9aaf 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -1383,6 +1383,8 @@ static unsigned long ram_alignment(resource_size_t pos) return 32*1024*1024; } +#define MAX_RESOURCE_SIZE ((resource_size_t)-1) + void __init e820_reserve_resources_late(void) { int i; @@ -1400,17 +1402,19 @@ void __init e820_reserve_resources_late(void) * avoid stolen RAM: */ for (i = 0; i < e820.nr_map; i++) { - struct e820entry *entry = &e820_saved.map[i]; - resource_size_t start, end; + struct e820entry *entry = &e820.map[i]; + u64 start, end; if (entry->type != E820_RAM) continue; start = entry->addr + entry->size; - end = round_up(start, ram_alignment(start)); - if (start == end) + end = round_up(start, ram_alignment(start)) - 1; + if (end > MAX_RESOURCE_SIZE) + end = MAX_RESOURCE_SIZE; + if (start >= end) continue; - reserve_region_with_split(&iomem_resource, start, - end - 1, "RAM buffer"); + reserve_region_with_split(&iomem_resource, start, end, + "RAM buffer"); } } diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index 47630479b06..1a041bcf506 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -211,11 +211,11 @@ static __init int iommu_setup(char *p) #ifdef CONFIG_SWIOTLB if (!strncmp(p, "soft", 4)) swiotlb = 1; +#endif if (!strncmp(p, "pt", 2)) { iommu_pass_through = 1; return 1; } -#endif gart_parse_options(p); diff --git a/drivers/clocksource/sh_tmu.c b/drivers/clocksource/sh_tmu.c index 9ffb05f4095..93c2322feab 100644 --- a/drivers/clocksource/sh_tmu.c +++ b/drivers/clocksource/sh_tmu.c @@ -161,7 +161,7 @@ static void sh_tmu_set_next(struct sh_tmu_priv *p, unsigned long delta, if (periodic) sh_tmu_write(p, TCOR, delta); else - sh_tmu_write(p, TCOR, 0); + sh_tmu_write(p, TCOR, 0xffffffff); sh_tmu_write(p, TCNT, delta); diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 420afa88728..53075424a43 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -56,14 +56,32 @@ #define MAX_AGAW_WIDTH 64 #define DOMAIN_MAX_ADDR(gaw) ((((u64)1) << gaw) - 1) +#define DOMAIN_MAX_PFN(gaw) ((((u64)1) << (gaw-VTD_PAGE_SHIFT)) - 1) #define IOVA_PFN(addr) ((addr) >> PAGE_SHIFT) #define DMA_32BIT_PFN IOVA_PFN(DMA_BIT_MASK(32)) #define DMA_64BIT_PFN IOVA_PFN(DMA_BIT_MASK(64)) -#ifndef PHYSICAL_PAGE_MASK -#define PHYSICAL_PAGE_MASK PAGE_MASK -#endif + +/* VT-d pages must always be _smaller_ than MM pages. Otherwise things + are never going to work. */ +static inline unsigned long dma_to_mm_pfn(unsigned long dma_pfn) +{ + return dma_pfn >> (PAGE_SHIFT - VTD_PAGE_SHIFT); +} + +static inline unsigned long mm_to_dma_pfn(unsigned long mm_pfn) +{ + return mm_pfn << (PAGE_SHIFT - VTD_PAGE_SHIFT); +} +static inline unsigned long page_to_dma_pfn(struct page *pg) +{ + return mm_to_dma_pfn(page_to_pfn(pg)); +} +static inline unsigned long virt_to_dma_pfn(void *p) +{ + return page_to_dma_pfn(virt_to_page(p)); +} /* global iommu list, set NULL for ignored DMAR units */ static struct intel_iommu **g_iommus; @@ -204,12 +222,17 @@ static inline void dma_set_pte_prot(struct dma_pte *pte, unsigned long prot) static inline u64 dma_pte_addr(struct dma_pte *pte) { - return (pte->val & VTD_PAGE_MASK); +#ifdef CONFIG_64BIT + return pte->val & VTD_PAGE_MASK; +#else + /* Must have a full atomic 64-bit read */ + return __cmpxchg64(pte, 0ULL, 0ULL) & VTD_PAGE_MASK; +#endif } -static inline void dma_set_pte_addr(struct dma_pte *pte, u64 addr) +static inline void dma_set_pte_pfn(struct dma_pte *pte, unsigned long pfn) { - pte->val |= (addr & VTD_PAGE_MASK); + pte->val |= (uint64_t)pfn << VTD_PAGE_SHIFT; } static inline bool dma_pte_present(struct dma_pte *pte) @@ -217,6 +240,11 @@ static inline bool dma_pte_present(struct dma_pte *pte) return (pte->val & 3) != 0; } +static inline int first_pte_in_page(struct dma_pte *pte) +{ + return !((unsigned long)pte & ~VTD_PAGE_MASK); +} + /* * This domain is a statically identity mapping domain. * 1. This domain creats a static 1:1 mapping to all usable memory. @@ -244,7 +272,6 @@ struct dmar_domain { struct iova_domain iovad; /* iova's that belong to this domain */ struct dma_pte *pgd; /* virtual address */ - spinlock_t mapping_lock; /* page table lock */ int gaw; /* max guest address width */ /* adjusted guest address width, 0 is level 2 30-bit */ @@ -648,80 +675,78 @@ static inline int width_to_agaw(int width) static inline unsigned int level_to_offset_bits(int level) { - return (12 + (level - 1) * LEVEL_STRIDE); + return (level - 1) * LEVEL_STRIDE; } -static inline int address_level_offset(u64 addr, int level) +static inline int pfn_level_offset(unsigned long pfn, int level) { - return ((addr >> level_to_offset_bits(level)) & LEVEL_MASK); + return (pfn >> level_to_offset_bits(level)) & LEVEL_MASK; } -static inline u64 level_mask(int level) +static inline unsigned long level_mask(int level) { - return ((u64)-1 << level_to_offset_bits(level)); + return -1UL << level_to_offset_bits(level); } -static inline u64 level_size(int level) +static inline unsigned long level_size(int level) { - return ((u64)1 << level_to_offset_bits(level)); + return 1UL << level_to_offset_bits(level); } -static inline u64 align_to_level(u64 addr, int level) +static inline unsigned long align_to_level(unsigned long pfn, int level) { - return ((addr + level_size(level) - 1) & level_mask(level)); + return (pfn + level_size(level) - 1) & level_mask(level); } -static struct dma_pte * addr_to_dma_pte(struct dmar_domain *domain, u64 addr) +static struct dma_pte *pfn_to_dma_pte(struct dmar_domain *domain, + unsigned long pfn) { - int addr_width = agaw_to_width(domain->agaw); + int addr_width = agaw_to_width(domain->agaw) - VTD_PAGE_SHIFT; struct dma_pte *parent, *pte = NULL; int level = agaw_to_level(domain->agaw); int offset; - unsigned long flags; BUG_ON(!domain->pgd); - - addr &= (((u64)1) << addr_width) - 1; + BUG_ON(addr_width < BITS_PER_LONG && pfn >> addr_width); parent = domain->pgd; - spin_lock_irqsave(&domain->mapping_lock, flags); while (level > 0) { void *tmp_page; - offset = address_level_offset(addr, level); + offset = pfn_level_offset(pfn, level); pte = &parent[offset]; if (level == 1) break; if (!dma_pte_present(pte)) { + uint64_t pteval; + tmp_page = alloc_pgtable_page(); - if (!tmp_page) { - spin_unlock_irqrestore(&domain->mapping_lock, - flags); + if (!tmp_page) return NULL; + + domain_flush_cache(domain, tmp_page, VTD_PAGE_SIZE); + pteval = (virt_to_dma_pfn(tmp_page) << VTD_PAGE_SHIFT) | DMA_PTE_READ | DMA_PTE_WRITE; + if (cmpxchg64(&pte->val, 0ULL, pteval)) { + /* Someone else set it while we were thinking; use theirs. */ + free_pgtable_page(tmp_page); + } else { + dma_pte_addr(pte); + domain_flush_cache(domain, pte, sizeof(*pte)); } - domain_flush_cache(domain, tmp_page, PAGE_SIZE); - dma_set_pte_addr(pte, virt_to_phys(tmp_page)); - /* - * high level table always sets r/w, last level page - * table control read/write - */ - dma_set_pte_readable(pte); - dma_set_pte_writable(pte); - domain_flush_cache(domain, pte, sizeof(*pte)); } parent = phys_to_virt(dma_pte_addr(pte)); level--; } - spin_unlock_irqrestore(&domain->mapping_lock, flags); return pte; } /* return address's pte at specific level */ -static struct dma_pte *dma_addr_level_pte(struct dmar_domain *domain, u64 addr, - int level) +static struct dma_pte *dma_pfn_level_pte(struct dmar_domain *domain, + unsigned long pfn, + int level) { struct dma_pte *parent, *pte = NULL; int total = agaw_to_level(domain->agaw); @@ -729,7 +754,7 @@ static struct dma_pte *dma_addr_level_pte(struct dmar_domain *domain, u64 addr, parent = domain->pgd; while (level <= total) { - offset = address_level_offset(addr, total); + offset = pfn_level_offset(pfn, total); pte = &parent[offset]; if (level == total) return pte; @@ -742,74 +767,82 @@ static struct dma_pte *dma_addr_level_pte(struct dmar_domain *domain, u64 addr, return NULL; } -/* clear one page's page table */ -static void dma_pte_clear_one(struct dmar_domain *domain, u64 addr) -{ - struct dma_pte *pte = NULL; - - /* get last level pte */ - pte = dma_addr_level_pte(domain, addr, 1); - - if (pte) { - dma_clear_pte(pte); - domain_flush_cache(domain, pte, sizeof(*pte)); - } -} - /* clear last level pte, a tlb flush should be followed */ -static void dma_pte_clear_range(struct dmar_domain *domain, u64 start, u64 end) +static void dma_pte_clear_range(struct dmar_domain *domain, + unsigned long start_pfn, + unsigned long last_pfn) { - int addr_width = agaw_to_width(domain->agaw); - int npages; + int addr_width = agaw_to_width(domain->agaw) - VTD_PAGE_SHIFT; + struct dma_pte *first_pte, *pte; + + BUG_ON(addr_width < BITS_PER_LONG && start_pfn >> addr_width); + BUG_ON(addr_width < BITS_PER_LONG && last_pfn >> addr_width); - start &= (((u64)1) << addr_width) - 1; - end &= (((u64)1) << addr_width) - 1; - /* in case it's partial page */ - start &= PAGE_MASK; - end = PAGE_ALIGN(end); - npages = (end - start) / VTD_PAGE_SIZE; + /* we don't need lock here; nobody else touches the iova range */ + while (start_pfn <= last_pfn) { + first_pte = pte = dma_pfn_level_pte(domain, start_pfn, 1); + if (!pte) { + start_pfn = align_to_level(start_pfn + 1, 2); + continue; + } + do { + dma_clear_pte(pte); + start_pfn++; + pte++; + } while (start_pfn <= last_pfn && !first_pte_in_page(pte)); - /* we don't need lock here, nobody else touches the iova range */ - while (npages--) { - dma_pte_clear_one(domain, start); - start += VTD_PAGE_SIZE; + domain_flush_cache(domain, first_pte, + (void *)pte - (void *)first_pte); } } /* free page table pages. last level pte should already be cleared */ static void dma_pte_free_pagetable(struct dmar_domain *domain, - u64 start, u64 end) + unsigned long start_pfn, + unsigned long last_pfn) { - int addr_width = agaw_to_width(domain->agaw); - struct dma_pte *pte; + int addr_width = agaw_to_width(domain->agaw) - VTD_PAGE_SHIFT; + struct dma_pte *first_pte, *pte; int total = agaw_to_level(domain->agaw); int level; - u64 tmp; + unsigned long tmp; - start &= (((u64)1) << addr_width) - 1; - end &= (((u64)1) << addr_width) - 1; + BUG_ON(addr_width < BITS_PER_LONG && start_pfn >> addr_width); + BUG_ON(addr_width < BITS_PER_LONG && last_pfn >> addr_width); - /* we don't need lock here, nobody else touches the iova range */ + /* We don't need lock here; nobody else touches the iova range */ level = 2; while (level <= total) { - tmp = align_to_level(start, level); - if (tmp >= end || (tmp + level_size(level) > end)) + tmp = align_to_level(start_pfn, level); + + /* If we can't even clear one PTE at this level, we're done */ + if (tmp + level_size(level) - 1 > last_pfn) return; - while (tmp < end) { - pte = dma_addr_level_pte(domain, tmp, level); - if (pte) { - free_pgtable_page( - phys_to_virt(dma_pte_addr(pte))); - dma_clear_pte(pte); - domain_flush_cache(domain, pte, sizeof(*pte)); + while (tmp + level_size(level) - 1 <= last_pfn) { + first_pte = pte = dma_pfn_level_pte(domain, tmp, level); + if (!pte) { + tmp = align_to_level(tmp + 1, level + 1); + continue; } - tmp += level_size(level); + do { + if (dma_pte_present(pte)) { + free_pgtable_page(phys_to_virt(dma_pte_addr(pte))); + dma_clear_pte(pte); + } + pte++; + tmp += level_size(level); + } while (!first_pte_in_page(pte) && + tmp + level_size(level) - 1 <= last_pfn); + + domain_flush_cache(domain, first_pte, + (void *)pte - (void *)first_pte); + } level++; } /* free pgd */ - if (start == 0 && end >= ((((u64)1) << addr_width) - 1)) { + if (start_pfn == 0 && last_pfn == DOMAIN_MAX_PFN(domain->gaw)) { free_pgtable_page(domain->pgd); domain->pgd = NULL; } @@ -1035,11 +1068,11 @@ static void iommu_flush_dev_iotlb(struct dmar_domain *domain, } static void iommu_flush_iotlb_psi(struct intel_iommu *iommu, u16 did, - u64 addr, unsigned int pages) + unsigned long pfn, unsigned int pages) { unsigned int mask = ilog2(__roundup_pow_of_two(pages)); + uint64_t addr = (uint64_t)pfn << VTD_PAGE_SHIFT; - BUG_ON(addr & (~VTD_PAGE_MASK)); BUG_ON(pages == 0); /* @@ -1054,7 +1087,12 @@ static void iommu_flush_iotlb_psi(struct intel_iommu *iommu, u16 did, else iommu->flush.flush_iotlb(iommu, did, addr, mask, DMA_TLB_PSI_FLUSH); - if (did) + + /* + * In caching mode, domain ID 0 is reserved for non-present to present + * mapping flush. Device IOTLB doesn't need to be flushed in this case. + */ + if (!cap_caching_mode(iommu->cap) || did) iommu_flush_dev_iotlb(iommu->domains[did], addr, mask); } @@ -1279,7 +1317,6 @@ static void dmar_init_reserved_ranges(void) struct pci_dev *pdev = NULL; struct iova *iova; int i; - u64 addr, size; init_iova_domain(&reserved_iova_list, DMA_32BIT_PFN); @@ -1302,12 +1339,9 @@ static void dmar_init_reserved_ranges(void) r = &pdev->resource[i]; if (!r->flags || !(r->flags & IORESOURCE_MEM)) continue; - addr = r->start; - addr &= PHYSICAL_PAGE_MASK; - size = r->end - addr; - size = PAGE_ALIGN(size); - iova = reserve_iova(&reserved_iova_list, IOVA_PFN(addr), - IOVA_PFN(size + addr) - 1); + iova = reserve_iova(&reserved_iova_list, + IOVA_PFN(r->start), + IOVA_PFN(r->end)); if (!iova) printk(KERN_ERR "Reserve iova failed\n"); } @@ -1341,7 +1375,6 @@ static int domain_init(struct dmar_domain *domain, int guest_width) unsigned long sagaw; init_iova_domain(&domain->iovad, DMA_32BIT_PFN); - spin_lock_init(&domain->mapping_lock); spin_lock_init(&domain->iommu_lock); domain_reserve_special_ranges(domain); @@ -1388,7 +1421,6 @@ static void domain_exit(struct dmar_domain *domain) { struct dmar_drhd_unit *drhd; struct intel_iommu *iommu; - u64 end; /* Domain 0 is reserved, so dont process it */ if (!domain) @@ -1397,14 +1429,12 @@ static void domain_exit(struct dmar_domain *domain) domain_remove_dev_info(domain); /* destroy iovas */ put_iova_domain(&domain->iovad); - end = DOMAIN_MAX_ADDR(domain->gaw); - end = end & (~PAGE_MASK); /* clear ptes */ - dma_pte_clear_range(domain, 0, end); + dma_pte_clear_range(domain, 0, DOMAIN_MAX_PFN(domain->gaw)); /* free page tables */ - dma_pte_free_pagetable(domain, 0, end); + dma_pte_free_pagetable(domain, 0, DOMAIN_MAX_PFN(domain->gaw)); for_each_active_iommu(iommu, drhd) if (test_bit(iommu->seq_id, &domain->iommu_bmp)) @@ -1618,42 +1648,86 @@ static int domain_context_mapped(struct pci_dev *pdev) tmp->devfn); } -static int -domain_page_mapping(struct dmar_domain *domain, dma_addr_t iova, - u64 hpa, size_t size, int prot) +static int __domain_mapping(struct dmar_domain *domain, unsigned long iov_pfn, + struct scatterlist *sg, unsigned long phys_pfn, + unsigned long nr_pages, int prot) { - u64 start_pfn, end_pfn; - struct dma_pte *pte; - int index; - int addr_width = agaw_to_width(domain->agaw); + struct dma_pte *first_pte = NULL, *pte = NULL; + phys_addr_t uninitialized_var(pteval); + int addr_width = agaw_to_width(domain->agaw) - VTD_PAGE_SHIFT; + unsigned long sg_res; - hpa &= (((u64)1) << addr_width) - 1; + BUG_ON(addr_width < BITS_PER_LONG && (iov_pfn + nr_pages - 1) >> addr_width); if ((prot & (DMA_PTE_READ|DMA_PTE_WRITE)) == 0) return -EINVAL; - iova &= PAGE_MASK; - start_pfn = ((u64)hpa) >> VTD_PAGE_SHIFT; - end_pfn = (VTD_PAGE_ALIGN(((u64)hpa) + size)) >> VTD_PAGE_SHIFT; - index = 0; - while (start_pfn < end_pfn) { - pte = addr_to_dma_pte(domain, iova + VTD_PAGE_SIZE * index); - if (!pte) - return -ENOMEM; + + prot &= DMA_PTE_READ | DMA_PTE_WRITE | DMA_PTE_SNP; + + if (sg) + sg_res = 0; + else { + sg_res = nr_pages + 1; + pteval = ((phys_addr_t)phys_pfn << VTD_PAGE_SHIFT) | prot; + } + + while (nr_pages--) { + uint64_t tmp; + + if (!sg_res) { + sg_res = (sg->offset + sg->length + VTD_PAGE_SIZE - 1) >> VTD_PAGE_SHIFT; + sg->dma_address = ((dma_addr_t)iov_pfn << VTD_PAGE_SHIFT) + sg->offset; + sg->dma_length = sg->length; + pteval = page_to_phys(sg_page(sg)) | prot; + } + if (!pte) { + first_pte = pte = pfn_to_dma_pte(domain, iov_pfn); + if (!pte) + return -ENOMEM; + } /* We don't need lock here, nobody else * touches the iova range */ - BUG_ON(dma_pte_addr(pte)); - dma_set_pte_addr(pte, start_pfn << VTD_PAGE_SHIFT); - dma_set_pte_prot(pte, prot); - if (prot & DMA_PTE_SNP) - dma_set_pte_snp(pte); - domain_flush_cache(domain, pte, sizeof(*pte)); - start_pfn++; - index++; + tmp = cmpxchg64_local(&pte->val, 0ULL, pteval); + if (tmp) { + static int dumps = 5; + printk(KERN_CRIT "ERROR: DMA PTE for vPFN 0x%lx already set (to %llx not %llx)\n", + iov_pfn, tmp, (unsigned long long)pteval); + if (dumps) { + dumps--; + debug_dma_dump_mappings(NULL); + } + WARN_ON(1); + } + pte++; + if (!nr_pages || first_pte_in_page(pte)) { + domain_flush_cache(domain, first_pte, + (void *)pte - (void *)first_pte); + pte = NULL; + } + iov_pfn++; + pteval += VTD_PAGE_SIZE; + sg_res--; + if (!sg_res) + sg = sg_next(sg); } return 0; } +static inline int domain_sg_mapping(struct dmar_domain *domain, unsigned long iov_pfn, + struct scatterlist *sg, unsigned long nr_pages, + int prot) +{ + return __domain_mapping(domain, iov_pfn, sg, 0, nr_pages, prot); +} + +static inline int domain_pfn_mapping(struct dmar_domain *domain, unsigned long iov_pfn, + unsigned long phys_pfn, unsigned long nr_pages, + int prot) +{ + return __domain_mapping(domain, iov_pfn, NULL, phys_pfn, nr_pages, prot); +} + static void iommu_detach_dev(struct intel_iommu *iommu, u8 bus, u8 devfn) { if (!iommu) @@ -1844,58 +1918,61 @@ error: static int iommu_identity_mapping; +static int iommu_domain_identity_map(struct dmar_domain *domain, + unsigned long long start, + unsigned long long end) +{ + unsigned long first_vpfn = start >> VTD_PAGE_SHIFT; + unsigned long last_vpfn = end >> VTD_PAGE_SHIFT; + + if (!reserve_iova(&domain->iovad, dma_to_mm_pfn(first_vpfn), + dma_to_mm_pfn(last_vpfn))) { + printk(KERN_ERR "IOMMU: reserve iova failed\n"); + return -ENOMEM; + } + + pr_debug("Mapping reserved region %llx-%llx for domain %d\n", + start, end, domain->id); + /* + * RMRR range might have overlap with physical memory range, + * clear it first + */ + dma_pte_clear_range(domain, first_vpfn, last_vpfn); + + return domain_pfn_mapping(domain, first_vpfn, first_vpfn, + last_vpfn - first_vpfn + 1, + DMA_PTE_READ|DMA_PTE_WRITE); +} + static int iommu_prepare_identity_map(struct pci_dev *pdev, unsigned long long start, unsigned long long end) { struct dmar_domain *domain; - unsigned long size; - unsigned long long base; int ret; printk(KERN_INFO - "IOMMU: Setting identity map for device %s [0x%Lx - 0x%Lx]\n", - pci_name(pdev), start, end); - if (iommu_identity_mapping) - domain = si_domain; - else - /* page table init */ - domain = get_domain_for_dev(pdev, DEFAULT_DOMAIN_ADDRESS_WIDTH); + "IOMMU: Setting identity map for device %s [0x%Lx - 0x%Lx]\n", + pci_name(pdev), start, end); + + domain = get_domain_for_dev(pdev, DEFAULT_DOMAIN_ADDRESS_WIDTH); if (!domain) return -ENOMEM; - /* The address might not be aligned */ - base = start & PAGE_MASK; - size = end - base; - size = PAGE_ALIGN(size); - if (!reserve_iova(&domain->iovad, IOVA_PFN(base), - IOVA_PFN(base + size) - 1)) { - printk(KERN_ERR "IOMMU: reserve iova failed\n"); - ret = -ENOMEM; - goto error; - } - - pr_debug("Mapping reserved region %lx@%llx for %s\n", - size, base, pci_name(pdev)); - /* - * RMRR range might have overlap with physical memory range, - * clear it first - */ - dma_pte_clear_range(domain, base, base + size); - - ret = domain_page_mapping(domain, base, base, size, - DMA_PTE_READ|DMA_PTE_WRITE); + ret = iommu_domain_identity_map(domain, start, end); if (ret) goto error; /* context entry init */ ret = domain_context_mapping(domain, pdev, CONTEXT_TT_MULTI_LEVEL); - if (!ret) - return 0; -error: + if (ret) + goto error; + + return 0; + + error: domain_exit(domain); return ret; - } static inline int iommu_prepare_rmrr_dev(struct dmar_rmrr_unit *rmrr, @@ -1907,64 +1984,6 @@ static inline int iommu_prepare_rmrr_dev(struct dmar_rmrr_unit *rmrr, rmrr->end_address + 1); } -struct iommu_prepare_data { - struct pci_dev *pdev; - int ret; -}; - -static int __init iommu_prepare_work_fn(unsigned long start_pfn, - unsigned long end_pfn, void *datax) -{ - struct iommu_prepare_data *data; - - data = (struct iommu_prepare_data *)datax; - - data->ret = iommu_prepare_identity_map(data->pdev, - start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT); - return data->ret; - -} - -static int __init iommu_prepare_with_active_regions(struct pci_dev *pdev) -{ - int nid; - struct iommu_prepare_data data; - - data.pdev = pdev; - data.ret = 0; - - for_each_online_node(nid) { - work_with_active_regions(nid, iommu_prepare_work_fn, &data); - if (data.ret) - return data.ret; - } - return data.ret; -} - -#ifdef CONFIG_DMAR_GFX_WA -static void __init iommu_prepare_gfx_mapping(void) -{ - struct pci_dev *pdev = NULL; - int ret; - - for_each_pci_dev(pdev) { - if (pdev->dev.archdata.iommu == DUMMY_DEVICE_DOMAIN_INFO || - !IS_GFX_DEVICE(pdev)) - continue; - printk(KERN_INFO "IOMMU: gfx device %s 1-1 mapping\n", - pci_name(pdev)); - ret = iommu_prepare_with_active_regions(pdev); - if (ret) - printk(KERN_ERR "IOMMU: mapping reserved region failed\n"); - } -} -#else /* !CONFIG_DMAR_GFX_WA */ -static inline void iommu_prepare_gfx_mapping(void) -{ - return; -} -#endif - #ifdef CONFIG_DMAR_FLOPPY_WA static inline void iommu_prepare_isa(void) { @@ -1975,12 +1994,12 @@ static inline void iommu_prepare_isa(void) if (!pdev) return; - printk(KERN_INFO "IOMMU: Prepare 0-16M unity mapping for LPC\n"); + printk(KERN_INFO "IOMMU: Prepare 0-16MiB unity mapping for LPC\n"); ret = iommu_prepare_identity_map(pdev, 0, 16*1024*1024); if (ret) - printk(KERN_ERR "IOMMU: Failed to create 0-64M identity map, " - "floppy might not work\n"); + printk(KERN_ERR "IOMMU: Failed to create 0-16MiB identity map; " + "floppy might not work\n"); } #else @@ -2008,16 +2027,30 @@ static int __init init_context_pass_through(void) } static int md_domain_init(struct dmar_domain *domain, int guest_width); + +static int __init si_domain_work_fn(unsigned long start_pfn, + unsigned long end_pfn, void *datax) +{ + int *ret = datax; + + *ret = iommu_domain_identity_map(si_domain, + (uint64_t)start_pfn << PAGE_SHIFT, + (uint64_t)end_pfn << PAGE_SHIFT); + return *ret; + +} + static int si_domain_init(void) { struct dmar_drhd_unit *drhd; struct intel_iommu *iommu; - int ret = 0; + int nid, ret = 0; si_domain = alloc_domain(); if (!si_domain) return -EFAULT; + pr_debug("Identity mapping domain is domain %d\n", si_domain->id); for_each_active_iommu(iommu, drhd) { ret = iommu_attach_domain(si_domain, iommu); @@ -2034,6 +2067,12 @@ static int si_domain_init(void) si_domain->flags = DOMAIN_FLAG_STATIC_IDENTITY; + for_each_online_node(nid) { + work_with_active_regions(nid, si_domain_work_fn, &ret); + if (ret) + return ret; + } + return 0; } @@ -2087,13 +2126,14 @@ static int iommu_prepare_static_identity_mapping(void) if (ret) return -EFAULT; - printk(KERN_INFO "IOMMU: Setting identity map:\n"); for_each_pci_dev(pdev) { - ret = iommu_prepare_with_active_regions(pdev); - if (ret) { - printk(KERN_INFO "1:1 mapping to one domain failed.\n"); - return -EFAULT; - } + printk(KERN_INFO "IOMMU: identity mapping for device %s\n", + pci_name(pdev)); + + ret = domain_context_mapping(si_domain, pdev, + CONTEXT_TT_MULTI_LEVEL); + if (ret) + return ret; ret = domain_add_dev_info(si_domain, pdev); if (ret) return ret; @@ -2284,8 +2324,6 @@ int __init init_dmars(void) } } - iommu_prepare_gfx_mapping(); - iommu_prepare_isa(); } @@ -2330,50 +2368,40 @@ error: return ret; } -static inline u64 aligned_size(u64 host_addr, size_t size) -{ - u64 addr; - addr = (host_addr & (~PAGE_MASK)) + size; - return PAGE_ALIGN(addr); -} - -struct iova * -iommu_alloc_iova(struct dmar_domain *domain, size_t size, u64 end) +static inline unsigned long aligned_nrpages(unsigned long host_addr, + size_t size) { - struct iova *piova; + host_addr &= ~PAGE_MASK; + host_addr += size + PAGE_SIZE - 1; - /* Make sure it's in range */ - end = min_t(u64, DOMAIN_MAX_ADDR(domain->gaw), end); - if (!size || (IOVA_START_ADDR + size > end)) - return NULL; - - piova = alloc_iova(&domain->iovad, - size >> PAGE_SHIFT, IOVA_PFN(end), 1); - return piova; + return host_addr >> VTD_PAGE_SHIFT; } -static struct iova * -__intel_alloc_iova(struct device *dev, struct dmar_domain *domain, - size_t size, u64 dma_mask) +static struct iova *intel_alloc_iova(struct device *dev, + struct dmar_domain *domain, + unsigned long nrpages, uint64_t dma_mask) { struct pci_dev *pdev = to_pci_dev(dev); struct iova *iova = NULL; - if (dma_mask <= DMA_BIT_MASK(32) || dmar_forcedac) - iova = iommu_alloc_iova(domain, size, dma_mask); - else { + /* Restrict dma_mask to the width that the iommu can handle */ + dma_mask = min_t(uint64_t, DOMAIN_MAX_ADDR(domain->gaw), dma_mask); + + if (!dmar_forcedac && dma_mask > DMA_BIT_MASK(32)) { /* * First try to allocate an io virtual address in * DMA_BIT_MASK(32) and if that fails then try allocating * from higher range */ - iova = iommu_alloc_iova(domain, size, DMA_BIT_MASK(32)); - if (!iova) - iova = iommu_alloc_iova(domain, size, dma_mask); - } - - if (!iova) { - printk(KERN_ERR"Allocating iova for %s failed", pci_name(pdev)); + iova = alloc_iova(&domain->iovad, nrpages, + IOVA_PFN(DMA_BIT_MASK(32)), 1); + if (iova) + return iova; + } + iova = alloc_iova(&domain->iovad, nrpages, IOVA_PFN(dma_mask), 1); + if (unlikely(!iova)) { + printk(KERN_ERR "Allocating %ld-page iova for %s failed", + nrpages, pci_name(pdev)); return NULL; } @@ -2476,14 +2504,12 @@ static dma_addr_t __intel_map_single(struct device *hwdev, phys_addr_t paddr, return 0; iommu = domain_get_iommu(domain); - size = aligned_size((u64)paddr, size); + size = aligned_nrpages(paddr, size); - iova = __intel_alloc_iova(hwdev, domain, size, pdev->dma_mask); + iova = intel_alloc_iova(hwdev, domain, size, pdev->dma_mask); if (!iova) goto error; - start_paddr = (phys_addr_t)iova->pfn_lo << PAGE_SHIFT; - /* * Check if DMAR supports zero-length reads on write only * mappings.. @@ -2499,20 +2525,20 @@ static dma_addr_t __intel_map_single(struct device *hwdev, phys_addr_t paddr, * might have two guest_addr mapping to the same host paddr, but this * is not a big problem */ - ret = domain_page_mapping(domain, start_paddr, - ((u64)paddr) & PHYSICAL_PAGE_MASK, - size, prot); + ret = domain_pfn_mapping(domain, mm_to_dma_pfn(iova->pfn_lo), + paddr >> VTD_PAGE_SHIFT, size, prot); if (ret) goto error; /* it's a non-present to present mapping. Only flush if caching mode */ if (cap_caching_mode(iommu->cap)) - iommu_flush_iotlb_psi(iommu, 0, start_paddr, - size >> VTD_PAGE_SHIFT); + iommu_flush_iotlb_psi(iommu, 0, mm_to_dma_pfn(iova->pfn_lo), size); else iommu_flush_write_buffer(iommu); - return start_paddr + ((u64)paddr & (~PAGE_MASK)); + start_paddr = (phys_addr_t)iova->pfn_lo << PAGE_SHIFT; + start_paddr += paddr & ~PAGE_MASK; + return start_paddr; error: if (iova) @@ -2605,7 +2631,7 @@ static void intel_unmap_page(struct device *dev, dma_addr_t dev_addr, { struct pci_dev *pdev = to_pci_dev(dev); struct dmar_domain *domain; - unsigned long start_addr; + unsigned long start_pfn, last_pfn; struct iova *iova; struct intel_iommu *iommu; @@ -2618,22 +2644,25 @@ static void intel_unmap_page(struct device *dev, dma_addr_t dev_addr, iommu = domain_get_iommu(domain); iova = find_iova(&domain->iovad, IOVA_PFN(dev_addr)); - if (!iova) + if (WARN_ONCE(!iova, "Driver unmaps unmatched page at PFN %llx\n", + (unsigned long long)dev_addr)) return; - start_addr = iova->pfn_lo << PAGE_SHIFT; - size = aligned_size((u64)dev_addr, size); + start_pfn = mm_to_dma_pfn(iova->pfn_lo); + last_pfn = mm_to_dma_pfn(iova->pfn_hi + 1) - 1; - pr_debug("Device %s unmapping: %zx@%llx\n", - pci_name(pdev), size, (unsigned long long)start_addr); + pr_debug("Device %s unmapping: pfn %lx-%lx\n", + pci_name(pdev), start_pfn, last_pfn); /* clear the whole page */ - dma_pte_clear_range(domain, start_addr, start_addr + size); + dma_pte_clear_range(domain, start_pfn, last_pfn); + /* free page tables */ - dma_pte_free_pagetable(domain, start_addr, start_addr + size); + dma_pte_free_pagetable(domain, start_pfn, last_pfn); + if (intel_iommu_strict) { - iommu_flush_iotlb_psi(iommu, domain->id, start_addr, - size >> VTD_PAGE_SHIFT); + iommu_flush_iotlb_psi(iommu, domain->id, start_pfn, + last_pfn - start_pfn + 1); /* free iova */ __free_iova(&domain->iovad, iova); } else { @@ -2691,14 +2720,10 @@ static void intel_unmap_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, enum dma_data_direction dir, struct dma_attrs *attrs) { - int i; struct pci_dev *pdev = to_pci_dev(hwdev); struct dmar_domain *domain; - unsigned long start_addr; + unsigned long start_pfn, last_pfn; struct iova *iova; - size_t size = 0; - phys_addr_t addr; - struct scatterlist *sg; struct intel_iommu *iommu; if (iommu_no_mapping(pdev)) @@ -2710,22 +2735,21 @@ static void intel_unmap_sg(struct device *hwdev, struct scatterlist *sglist, iommu = domain_get_iommu(domain); iova = find_iova(&domain->iovad, IOVA_PFN(sglist[0].dma_address)); - if (!iova) + if (WARN_ONCE(!iova, "Driver unmaps unmatched sglist at PFN %llx\n", + (unsigned long long)sglist[0].dma_address)) return; - for_each_sg(sglist, sg, nelems, i) { - addr = page_to_phys(sg_page(sg)) + sg->offset; - size += aligned_size((u64)addr, sg->length); - } - start_addr = iova->pfn_lo << PAGE_SHIFT; + start_pfn = mm_to_dma_pfn(iova->pfn_lo); + last_pfn = mm_to_dma_pfn(iova->pfn_hi + 1) - 1; /* clear the whole page */ - dma_pte_clear_range(domain, start_addr, start_addr + size); + dma_pte_clear_range(domain, start_pfn, last_pfn); + /* free page tables */ - dma_pte_free_pagetable(domain, start_addr, start_addr + size); + dma_pte_free_pagetable(domain, start_pfn, last_pfn); - iommu_flush_iotlb_psi(iommu, domain->id, start_addr, - size >> VTD_PAGE_SHIFT); + iommu_flush_iotlb_psi(iommu, domain->id, start_pfn, + (last_pfn - start_pfn + 1)); /* free iova */ __free_iova(&domain->iovad, iova); @@ -2748,17 +2772,16 @@ static int intel_nontranslate_map_sg(struct device *hddev, static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, enum dma_data_direction dir, struct dma_attrs *attrs) { - phys_addr_t addr; int i; struct pci_dev *pdev = to_pci_dev(hwdev); struct dmar_domain *domain; size_t size = 0; int prot = 0; - size_t offset = 0; + size_t offset_pfn = 0; struct iova *iova = NULL; int ret; struct scatterlist *sg; - unsigned long start_addr; + unsigned long start_vpfn; struct intel_iommu *iommu; BUG_ON(dir == DMA_NONE); @@ -2771,12 +2794,10 @@ static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int ne iommu = domain_get_iommu(domain); - for_each_sg(sglist, sg, nelems, i) { - addr = page_to_phys(sg_page(sg)) + sg->offset; - size += aligned_size((u64)addr, sg->length); - } + for_each_sg(sglist, sg, nelems, i) + size += aligned_nrpages(sg->offset, sg->length); - iova = __intel_alloc_iova(hwdev, domain, size, pdev->dma_mask); + iova = intel_alloc_iova(hwdev, domain, size, pdev->dma_mask); if (!iova) { sglist->dma_length = 0; return 0; @@ -2792,35 +2813,24 @@ static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int ne if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL) prot |= DMA_PTE_WRITE; - start_addr = iova->pfn_lo << PAGE_SHIFT; - offset = 0; - for_each_sg(sglist, sg, nelems, i) { - addr = page_to_phys(sg_page(sg)) + sg->offset; - size = aligned_size((u64)addr, sg->length); - ret = domain_page_mapping(domain, start_addr + offset, - ((u64)addr) & PHYSICAL_PAGE_MASK, - size, prot); - if (ret) { - /* clear the page */ - dma_pte_clear_range(domain, start_addr, - start_addr + offset); - /* free page tables */ - dma_pte_free_pagetable(domain, start_addr, - start_addr + offset); - /* free iova */ - __free_iova(&domain->iovad, iova); - return 0; - } - sg->dma_address = start_addr + offset + - ((u64)addr & (~PAGE_MASK)); - sg->dma_length = sg->length; - offset += size; + start_vpfn = mm_to_dma_pfn(iova->pfn_lo); + + ret = domain_sg_mapping(domain, start_vpfn, sglist, mm_to_dma_pfn(size), prot); + if (unlikely(ret)) { + /* clear the page */ + dma_pte_clear_range(domain, start_vpfn, + start_vpfn + size - 1); + /* free page tables */ + dma_pte_free_pagetable(domain, start_vpfn, + start_vpfn + size - 1); + /* free iova */ + __free_iova(&domain->iovad, iova); + return 0; } /* it's a non-present to present mapping. Only flush if caching mode */ if (cap_caching_mode(iommu->cap)) - iommu_flush_iotlb_psi(iommu, 0, start_addr, - offset >> VTD_PAGE_SHIFT); + iommu_flush_iotlb_psi(iommu, 0, start_vpfn, offset_pfn); else iommu_flush_write_buffer(iommu); @@ -3325,7 +3335,6 @@ static int md_domain_init(struct dmar_domain *domain, int guest_width) int adjust_width; init_iova_domain(&domain->iovad, DMA_32BIT_PFN); - spin_lock_init(&domain->mapping_lock); spin_lock_init(&domain->iommu_lock); domain_reserve_special_ranges(domain); @@ -3379,8 +3388,6 @@ static void iommu_free_vm_domain(struct dmar_domain *domain) static void vm_domain_exit(struct dmar_domain *domain) { - u64 end; - /* Domain 0 is reserved, so dont process it */ if (!domain) return; @@ -3388,14 +3395,12 @@ static void vm_domain_exit(struct dmar_domain *domain) vm_domain_remove_all_dev_info(domain); /* destroy iovas */ put_iova_domain(&domain->iovad); - end = DOMAIN_MAX_ADDR(domain->gaw); - end = end & (~VTD_PAGE_MASK); /* clear ptes */ - dma_pte_clear_range(domain, 0, end); + dma_pte_clear_range(domain, 0, DOMAIN_MAX_PFN(domain->gaw)); /* free page tables */ - dma_pte_free_pagetable(domain, 0, end); + dma_pte_free_pagetable(domain, 0, DOMAIN_MAX_PFN(domain->gaw)); iommu_free_vm_domain(domain); free_domain_mem(domain); @@ -3504,7 +3509,7 @@ static int intel_iommu_map_range(struct iommu_domain *domain, if ((iommu_prot & IOMMU_CACHE) && dmar_domain->iommu_snooping) prot |= DMA_PTE_SNP; - max_addr = (iova & VTD_PAGE_MASK) + VTD_PAGE_ALIGN(size); + max_addr = iova + size; if (dmar_domain->max_addr < max_addr) { int min_agaw; u64 end; @@ -3522,8 +3527,11 @@ static int intel_iommu_map_range(struct iommu_domain *domain, } dmar_domain->max_addr = max_addr; } - - ret = domain_page_mapping(dmar_domain, iova, hpa, size, prot); + /* Round up size to next multiple of PAGE_SIZE, if it and + the low bits of hpa would take us onto the next page */ + size = aligned_nrpages(hpa, size); + ret = domain_pfn_mapping(dmar_domain, iova >> VTD_PAGE_SHIFT, + hpa >> VTD_PAGE_SHIFT, size, prot); return ret; } @@ -3531,15 +3539,12 @@ static void intel_iommu_unmap_range(struct iommu_domain *domain, unsigned long iova, size_t size) { struct dmar_domain *dmar_domain = domain->priv; - dma_addr_t base; - /* The address might not be aligned */ - base = iova & VTD_PAGE_MASK; - size = VTD_PAGE_ALIGN(size); - dma_pte_clear_range(dmar_domain, base, base + size); + dma_pte_clear_range(dmar_domain, iova >> VTD_PAGE_SHIFT, + (iova + size - 1) >> VTD_PAGE_SHIFT); - if (dmar_domain->max_addr == base + size) - dmar_domain->max_addr = base; + if (dmar_domain->max_addr == iova + size) + dmar_domain->max_addr = iova; } static phys_addr_t intel_iommu_iova_to_phys(struct iommu_domain *domain, @@ -3549,7 +3554,7 @@ static phys_addr_t intel_iommu_iova_to_phys(struct iommu_domain *domain, struct dma_pte *pte; u64 phys = 0; - pte = addr_to_dma_pte(dmar_domain, iova); + pte = pfn_to_dma_pte(dmar_domain, iova >> VTD_PAGE_SHIFT); if (pte) phys = dma_pte_addr(pte); diff --git a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c index 74369a3f963..c399f485aa7 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c +++ b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c @@ -13,6 +13,7 @@ #include <linux/inet.h> #include <linux/crypto.h> +#include <linux/if_vlan.h> #include <net/dst.h> #include <net/tcp.h> #include <scsi/scsi_cmnd.h> @@ -184,6 +185,9 @@ static struct cxgb3i_hba *cxgb3i_hba_find_by_netdev(struct net_device *ndev) struct cxgb3i_adapter *snic; int i; + if (ndev->priv_flags & IFF_802_1Q_VLAN) + ndev = vlan_dev_real_dev(ndev); + read_lock(&cxgb3i_snic_rwlock); list_for_each_entry(snic, &cxgb3i_snic_list, list_head) { for (i = 0; i < snic->hba_cnt; i++) { diff --git a/drivers/scsi/fnic/fnic_main.c b/drivers/scsi/fnic/fnic_main.c index a84072865fc..2c266c01dc5 100644 --- a/drivers/scsi/fnic/fnic_main.c +++ b/drivers/scsi/fnic/fnic_main.c @@ -473,16 +473,16 @@ static int __devinit fnic_probe(struct pci_dev *pdev, * limitation for the device. Try 40-bit first, and * fail to 32-bit. */ - err = pci_set_dma_mask(pdev, DMA_40BIT_MASK); + err = pci_set_dma_mask(pdev, DMA_BIT_MASK(40)); if (err) { - err = pci_set_dma_mask(pdev, DMA_32BIT_MASK); + err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); if (err) { shost_printk(KERN_ERR, fnic->lport->host, "No usable DMA configuration " "aborting\n"); goto err_out_release_regions; } - err = pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK); + err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32)); if (err) { shost_printk(KERN_ERR, fnic->lport->host, "Unable to obtain 32-bit DMA " @@ -490,7 +490,7 @@ static int __devinit fnic_probe(struct pci_dev *pdev, goto err_out_release_regions; } } else { - err = pci_set_consistent_dma_mask(pdev, DMA_40BIT_MASK); + err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(40)); if (err) { shost_printk(KERN_ERR, fnic->lport->host, "Unable to obtain 40-bit DMA " diff --git a/drivers/scsi/fnic/fnic_scsi.c b/drivers/scsi/fnic/fnic_scsi.c index eabf3650285..bfc996971b8 100644 --- a/drivers/scsi/fnic/fnic_scsi.c +++ b/drivers/scsi/fnic/fnic_scsi.c @@ -245,7 +245,7 @@ static inline int fnic_queue_wq_copy_desc(struct fnic *fnic, struct vnic_wq_copy *wq, struct fnic_io_req *io_req, struct scsi_cmnd *sc, - u32 sg_count) + int sg_count) { struct scatterlist *sg; struct fc_rport *rport = starget_to_rport(scsi_target(sc->device)); @@ -260,9 +260,6 @@ static inline int fnic_queue_wq_copy_desc(struct fnic *fnic, char msg[2]; if (sg_count) { - BUG_ON(sg_count < 0); - BUG_ON(sg_count > FNIC_MAX_SG_DESC_CNT); - /* For each SGE, create a device desc entry */ desc = io_req->sgl_list; for_each_sg(scsi_sglist(sc), sg, sg_count, i) { @@ -344,7 +341,7 @@ int fnic_queuecommand(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) struct fnic *fnic; struct vnic_wq_copy *wq; int ret; - u32 sg_count; + int sg_count; unsigned long flags; unsigned long ptr; diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c index 869a11bdccb..9928704e235 100644 --- a/drivers/scsi/ibmvscsi/ibmvscsi.c +++ b/drivers/scsi/ibmvscsi/ibmvscsi.c @@ -1095,9 +1095,14 @@ static void adapter_info_rsp(struct srp_event_struct *evt_struct) MAX_INDIRECT_BUFS); hostdata->host->sg_tablesize = MAX_INDIRECT_BUFS; } + + if (hostdata->madapter_info.os_type == 3) { + enable_fast_fail(hostdata); + return; + } } - enable_fast_fail(hostdata); + send_srp_login(hostdata); } /** diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c index 2eee9e6e4fe..292c02f810d 100644 --- a/drivers/scsi/scsi_transport_fc.c +++ b/drivers/scsi/scsi_transport_fc.c @@ -3670,13 +3670,14 @@ static void fc_bsg_goose_queue(struct fc_rport *rport) { int flagset; + unsigned long flags; if (!rport->rqst_q) return; get_device(&rport->dev); - spin_lock(rport->rqst_q->queue_lock); + spin_lock_irqsave(rport->rqst_q->queue_lock, flags); flagset = test_bit(QUEUE_FLAG_REENTER, &rport->rqst_q->queue_flags) && !test_bit(QUEUE_FLAG_REENTER, &rport->rqst_q->queue_flags); if (flagset) @@ -3684,7 +3685,7 @@ fc_bsg_goose_queue(struct fc_rport *rport) __blk_run_queue(rport->rqst_q); if (flagset) queue_flag_clear(QUEUE_FLAG_REENTER, rport->rqst_q); - spin_unlock(rport->rqst_q->queue_lock); + spin_unlock_irqrestore(rport->rqst_q->queue_lock, flags); put_device(&rport->dev); } diff --git a/drivers/scsi/zalon.c b/drivers/scsi/zalon.c index 97f3158fa7b..27e84e4b1fa 100644 --- a/drivers/scsi/zalon.c +++ b/drivers/scsi/zalon.c @@ -134,7 +134,7 @@ zalon_probe(struct parisc_device *dev) host = ncr_attach(&zalon7xx_template, unit, &device); if (!host) - goto fail; + return -ENODEV; if (request_irq(dev->irq, ncr53c8xx_intr, IRQF_SHARED, "zalon", host)) { dev_printk(KERN_ERR, &dev->dev, "irq problem with %d, detaching\n ", diff --git a/drivers/video/sh_mobile_lcdcfb.c b/drivers/video/sh_mobile_lcdcfb.c index f10d2fbeda0..da983b720f0 100644 --- a/drivers/video/sh_mobile_lcdcfb.c +++ b/drivers/video/sh_mobile_lcdcfb.c @@ -17,6 +17,7 @@ #include <linux/platform_device.h> #include <linux/dma-mapping.h> #include <linux/interrupt.h> +#include <linux/vmalloc.h> #include <video/sh_mobile_lcdc.h> #include <asm/atomic.h> @@ -33,6 +34,7 @@ struct sh_mobile_lcdc_chan { struct fb_info info; dma_addr_t dma_handle; struct fb_deferred_io defio; + struct scatterlist *sglist; unsigned long frame_end; wait_queue_head_t frame_end_wait; }; @@ -206,16 +208,38 @@ static void sh_mobile_lcdc_clk_on(struct sh_mobile_lcdc_priv *priv) {} static void sh_mobile_lcdc_clk_off(struct sh_mobile_lcdc_priv *priv) {} #endif +static int sh_mobile_lcdc_sginit(struct fb_info *info, + struct list_head *pagelist) +{ + struct sh_mobile_lcdc_chan *ch = info->par; + unsigned int nr_pages_max = info->fix.smem_len >> PAGE_SHIFT; + struct page *page; + int nr_pages = 0; + + sg_init_table(ch->sglist, nr_pages_max); + + list_for_each_entry(page, pagelist, lru) + sg_set_page(&ch->sglist[nr_pages++], page, PAGE_SIZE, 0); + + return nr_pages; +} + static void sh_mobile_lcdc_deferred_io(struct fb_info *info, struct list_head *pagelist) { struct sh_mobile_lcdc_chan *ch = info->par; + unsigned int nr_pages; /* enable clocks before accessing hardware */ sh_mobile_lcdc_clk_on(ch->lcdc); + nr_pages = sh_mobile_lcdc_sginit(info, pagelist); + dma_map_sg(info->dev, ch->sglist, nr_pages, DMA_TO_DEVICE); + /* trigger panel update */ lcdc_write_chan(ch, LDSM2R, 1); + + dma_unmap_sg(info->dev, ch->sglist, nr_pages, DMA_TO_DEVICE); } static void sh_mobile_lcdc_deferred_io_touch(struct fb_info *info) @@ -846,21 +870,31 @@ static int __init sh_mobile_lcdc_probe(struct platform_device *pdev) } for (i = 0; i < j; i++) { - error = register_framebuffer(&priv->ch[i].info); + struct sh_mobile_lcdc_chan *ch = priv->ch + i; + + info = &ch->info; + + if (info->fbdefio) { + priv->ch->sglist = vmalloc(sizeof(struct scatterlist) * + info->fix.smem_len >> PAGE_SHIFT); + if (!priv->ch->sglist) { + dev_err(&pdev->dev, "cannot allocate sglist\n"); + goto err1; + } + } + + error = register_framebuffer(info); if (error < 0) goto err1; - } - for (i = 0; i < j; i++) { - info = &priv->ch[i].info; dev_info(info->dev, "registered %s/%s as %dx%d %dbpp.\n", pdev->name, - (priv->ch[i].cfg.chan == LCDC_CHAN_MAINLCD) ? + (ch->cfg.chan == LCDC_CHAN_MAINLCD) ? "mainlcd" : "sublcd", - (int) priv->ch[i].cfg.lcd_cfg.xres, - (int) priv->ch[i].cfg.lcd_cfg.yres, - priv->ch[i].cfg.bpp); + (int) ch->cfg.lcd_cfg.xres, + (int) ch->cfg.lcd_cfg.yres, + ch->cfg.bpp); /* deferred io mode: disable clock to save power */ if (info->fbdefio) @@ -892,6 +926,9 @@ static int sh_mobile_lcdc_remove(struct platform_device *pdev) if (!info->device) continue; + if (priv->ch[i].sglist) + vfree(priv->ch[i].sglist); + dma_free_coherent(&pdev->dev, info->fix.smem_len, info->screen_base, priv->ch[i].dma_handle); fb_dealloc_cmap(&info->cmap); diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c index 7f88628a1a7..6e4f6c50a12 100644 --- a/fs/btrfs/async-thread.c +++ b/fs/btrfs/async-thread.c @@ -299,8 +299,8 @@ int btrfs_start_workers(struct btrfs_workers *workers, int num_workers) "btrfs-%s-%d", workers->name, workers->num_workers + i); if (IS_ERR(worker->task)) { - kfree(worker); ret = PTR_ERR(worker->task); + kfree(worker); goto fail; } diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 2779c2f5360..98a87383871 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2074,8 +2074,7 @@ static inline int btrfs_insert_empty_item(struct btrfs_trans_handle *trans, int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path); int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path); int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf); -int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root - *root); +int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref); int btrfs_drop_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *node, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index edc7d208c5c..a5aca3997d4 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -990,15 +990,13 @@ static inline int extent_ref_type(u64 parent, u64 owner) return type; } -static int find_next_key(struct btrfs_path *path, struct btrfs_key *key) +static int find_next_key(struct btrfs_path *path, int level, + struct btrfs_key *key) { - int level; - BUG_ON(!path->keep_locks); - for (level = 0; level < BTRFS_MAX_LEVEL; level++) { + for (; level < BTRFS_MAX_LEVEL; level++) { if (!path->nodes[level]) break; - btrfs_assert_tree_locked(path->nodes[level]); if (path->slots[level] + 1 >= btrfs_header_nritems(path->nodes[level])) continue; @@ -1158,7 +1156,8 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans, * For simplicity, we just do not add new inline back * ref if there is any kind of item for this block */ - if (find_next_key(path, &key) == 0 && key.objectid == bytenr && + if (find_next_key(path, 0, &key) == 0 && + key.objectid == bytenr && key.type < BTRFS_BLOCK_GROUP_ITEM_KEY) { err = -EAGAIN; goto out; @@ -2697,7 +2696,7 @@ again: printk(KERN_ERR "no space left, need %llu, %llu delalloc bytes" ", %llu bytes_used, %llu bytes_reserved, " - "%llu bytes_pinned, %llu bytes_readonly, %llu may use" + "%llu bytes_pinned, %llu bytes_readonly, %llu may use " "%llu total\n", (unsigned long long)bytes, (unsigned long long)data_sinfo->bytes_delalloc, (unsigned long long)data_sinfo->bytes_used, @@ -4128,6 +4127,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, return buf; } +#if 0 int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *leaf) { @@ -4171,8 +4171,6 @@ int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, return 0; } -#if 0 - static noinline int cache_drop_leaf_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_leaf_ref *ref) @@ -4553,262 +4551,471 @@ out: } #endif +struct walk_control { + u64 refs[BTRFS_MAX_LEVEL]; + u64 flags[BTRFS_MAX_LEVEL]; + struct btrfs_key update_progress; + int stage; + int level; + int shared_level; + int update_ref; + int keep_locks; +}; + +#define DROP_REFERENCE 1 +#define UPDATE_BACKREF 2 + /* - * helper function for drop_subtree, this function is similar to - * walk_down_tree. The main difference is that it checks reference - * counts while tree blocks are locked. + * hepler to process tree block while walking down the tree. + * + * when wc->stage == DROP_REFERENCE, this function checks + * reference count of the block. if the block is shared and + * we need update back refs for the subtree rooted at the + * block, this function changes wc->stage to UPDATE_BACKREF + * + * when wc->stage == UPDATE_BACKREF, this function updates + * back refs for pointers in the block. + * + * NOTE: return value 1 means we should stop walking down. */ -static noinline int walk_down_tree(struct btrfs_trans_handle *trans, +static noinline int walk_down_proc(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct btrfs_path *path, int *level) + struct btrfs_path *path, + struct walk_control *wc) { - struct extent_buffer *next; - struct extent_buffer *cur; - struct extent_buffer *parent; - u64 bytenr; - u64 ptr_gen; - u64 refs; - u64 flags; - u32 blocksize; + int level = wc->level; + struct extent_buffer *eb = path->nodes[level]; + struct btrfs_key key; + u64 flag = BTRFS_BLOCK_FLAG_FULL_BACKREF; int ret; - cur = path->nodes[*level]; - ret = btrfs_lookup_extent_info(trans, root, cur->start, cur->len, - &refs, &flags); - BUG_ON(ret); - if (refs > 1) - goto out; + if (wc->stage == UPDATE_BACKREF && + btrfs_header_owner(eb) != root->root_key.objectid) + return 1; - BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)); + /* + * when reference count of tree block is 1, it won't increase + * again. once full backref flag is set, we never clear it. + */ + if ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) || + (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag))) { + BUG_ON(!path->locks[level]); + ret = btrfs_lookup_extent_info(trans, root, + eb->start, eb->len, + &wc->refs[level], + &wc->flags[level]); + BUG_ON(ret); + BUG_ON(wc->refs[level] == 0); + } - while (*level >= 0) { - cur = path->nodes[*level]; - if (*level == 0) { - ret = btrfs_drop_leaf_ref(trans, root, cur); - BUG_ON(ret); - clean_tree_block(trans, root, cur); - break; - } - if (path->slots[*level] >= btrfs_header_nritems(cur)) { - clean_tree_block(trans, root, cur); - break; + if (wc->stage == DROP_REFERENCE && + wc->update_ref && wc->refs[level] > 1) { + BUG_ON(eb == root->node); + BUG_ON(path->slots[level] > 0); + if (level == 0) + btrfs_item_key_to_cpu(eb, &key, path->slots[level]); + else + btrfs_node_key_to_cpu(eb, &key, path->slots[level]); + if (btrfs_header_owner(eb) == root->root_key.objectid && + btrfs_comp_cpu_keys(&key, &wc->update_progress) >= 0) { + wc->stage = UPDATE_BACKREF; + wc->shared_level = level; } + } - bytenr = btrfs_node_blockptr(cur, path->slots[*level]); - blocksize = btrfs_level_size(root, *level - 1); - ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]); + if (wc->stage == DROP_REFERENCE) { + if (wc->refs[level] > 1) + return 1; - next = read_tree_block(root, bytenr, blocksize, ptr_gen); - btrfs_tree_lock(next); - btrfs_set_lock_blocking(next); + if (path->locks[level] && !wc->keep_locks) { + btrfs_tree_unlock(eb); + path->locks[level] = 0; + } + return 0; + } - ret = btrfs_lookup_extent_info(trans, root, bytenr, blocksize, - &refs, &flags); + /* wc->stage == UPDATE_BACKREF */ + if (!(wc->flags[level] & flag)) { + BUG_ON(!path->locks[level]); + ret = btrfs_inc_ref(trans, root, eb, 1); BUG_ON(ret); - if (refs > 1) { - parent = path->nodes[*level]; - ret = btrfs_free_extent(trans, root, bytenr, - blocksize, parent->start, - btrfs_header_owner(parent), - *level - 1, 0); + ret = btrfs_dec_ref(trans, root, eb, 0); + BUG_ON(ret); + ret = btrfs_set_disk_extent_flags(trans, root, eb->start, + eb->len, flag, 0); + BUG_ON(ret); + wc->flags[level] |= flag; + } + + /* + * the block is shared by multiple trees, so it's not good to + * keep the tree lock + */ + if (path->locks[level] && level > 0) { + btrfs_tree_unlock(eb); + path->locks[level] = 0; + } + return 0; +} + +/* + * hepler to process tree block while walking up the tree. + * + * when wc->stage == DROP_REFERENCE, this function drops + * reference count on the block. + * + * when wc->stage == UPDATE_BACKREF, this function changes + * wc->stage back to DROP_REFERENCE if we changed wc->stage + * to UPDATE_BACKREF previously while processing the block. + * + * NOTE: return value 1 means we should stop walking up. + */ +static noinline int walk_up_proc(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct walk_control *wc) +{ + int ret = 0; + int level = wc->level; + struct extent_buffer *eb = path->nodes[level]; + u64 parent = 0; + + if (wc->stage == UPDATE_BACKREF) { + BUG_ON(wc->shared_level < level); + if (level < wc->shared_level) + goto out; + + BUG_ON(wc->refs[level] <= 1); + ret = find_next_key(path, level + 1, &wc->update_progress); + if (ret > 0) + wc->update_ref = 0; + + wc->stage = DROP_REFERENCE; + wc->shared_level = -1; + path->slots[level] = 0; + + /* + * check reference count again if the block isn't locked. + * we should start walking down the tree again if reference + * count is one. + */ + if (!path->locks[level]) { + BUG_ON(level == 0); + btrfs_tree_lock(eb); + btrfs_set_lock_blocking(eb); + path->locks[level] = 1; + + ret = btrfs_lookup_extent_info(trans, root, + eb->start, eb->len, + &wc->refs[level], + &wc->flags[level]); BUG_ON(ret); - path->slots[*level]++; - btrfs_tree_unlock(next); - free_extent_buffer(next); - continue; + BUG_ON(wc->refs[level] == 0); + if (wc->refs[level] == 1) { + btrfs_tree_unlock(eb); + path->locks[level] = 0; + return 1; + } + } else { + BUG_ON(level != 0); } + } - BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)); + /* wc->stage == DROP_REFERENCE */ + BUG_ON(wc->refs[level] > 1 && !path->locks[level]); - *level = btrfs_header_level(next); - path->nodes[*level] = next; - path->slots[*level] = 0; - path->locks[*level] = 1; - cond_resched(); + if (wc->refs[level] == 1) { + if (level == 0) { + if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) + ret = btrfs_dec_ref(trans, root, eb, 1); + else + ret = btrfs_dec_ref(trans, root, eb, 0); + BUG_ON(ret); + } + /* make block locked assertion in clean_tree_block happy */ + if (!path->locks[level] && + btrfs_header_generation(eb) == trans->transid) { + btrfs_tree_lock(eb); + btrfs_set_lock_blocking(eb); + path->locks[level] = 1; + } + clean_tree_block(trans, root, eb); + } + + if (eb == root->node) { + if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) + parent = eb->start; + else + BUG_ON(root->root_key.objectid != + btrfs_header_owner(eb)); + } else { + if (wc->flags[level + 1] & BTRFS_BLOCK_FLAG_FULL_BACKREF) + parent = path->nodes[level + 1]->start; + else + BUG_ON(root->root_key.objectid != + btrfs_header_owner(path->nodes[level + 1])); } -out: - if (path->nodes[*level] == root->node) - parent = path->nodes[*level]; - else - parent = path->nodes[*level + 1]; - bytenr = path->nodes[*level]->start; - blocksize = path->nodes[*level]->len; - ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent->start, - btrfs_header_owner(parent), *level, 0); + ret = btrfs_free_extent(trans, root, eb->start, eb->len, parent, + root->root_key.objectid, level, 0); BUG_ON(ret); +out: + wc->refs[level] = 0; + wc->flags[level] = 0; + return ret; +} + +static noinline int walk_down_tree(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct walk_control *wc) +{ + struct extent_buffer *next; + struct extent_buffer *cur; + u64 bytenr; + u64 ptr_gen; + u32 blocksize; + int level = wc->level; + int ret; + + while (level >= 0) { + cur = path->nodes[level]; + BUG_ON(path->slots[level] >= btrfs_header_nritems(cur)); - if (path->locks[*level]) { - btrfs_tree_unlock(path->nodes[*level]); - path->locks[*level] = 0; + ret = walk_down_proc(trans, root, path, wc); + if (ret > 0) + break; + + if (level == 0) + break; + + bytenr = btrfs_node_blockptr(cur, path->slots[level]); + blocksize = btrfs_level_size(root, level - 1); + ptr_gen = btrfs_node_ptr_generation(cur, path->slots[level]); + + next = read_tree_block(root, bytenr, blocksize, ptr_gen); + btrfs_tree_lock(next); + btrfs_set_lock_blocking(next); + + level--; + BUG_ON(level != btrfs_header_level(next)); + path->nodes[level] = next; + path->slots[level] = 0; + path->locks[level] = 1; + wc->level = level; } - free_extent_buffer(path->nodes[*level]); - path->nodes[*level] = NULL; - *level += 1; - cond_resched(); return 0; } -/* - * helper for dropping snapshots. This walks back up the tree in the path - * to find the first node higher up where we haven't yet gone through - * all the slots - */ static noinline int walk_up_tree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, - int *level, int max_level) + struct walk_control *wc, int max_level) { - struct btrfs_root_item *root_item = &root->root_item; - int i; - int slot; + int level = wc->level; int ret; - for (i = *level; i < max_level && path->nodes[i]; i++) { - slot = path->slots[i]; - if (slot + 1 < btrfs_header_nritems(path->nodes[i])) { - /* - * there is more work to do in this level. - * Update the drop_progress marker to reflect - * the work we've done so far, and then bump - * the slot number - */ - path->slots[i]++; - WARN_ON(*level == 0); - if (max_level == BTRFS_MAX_LEVEL) { - btrfs_node_key(path->nodes[i], - &root_item->drop_progress, - path->slots[i]); - root_item->drop_level = i; - } - *level = i; + path->slots[level] = btrfs_header_nritems(path->nodes[level]); + while (level < max_level && path->nodes[level]) { + wc->level = level; + if (path->slots[level] + 1 < + btrfs_header_nritems(path->nodes[level])) { + path->slots[level]++; return 0; } else { - struct extent_buffer *parent; - - /* - * this whole node is done, free our reference - * on it and go up one level - */ - if (path->nodes[*level] == root->node) - parent = path->nodes[*level]; - else - parent = path->nodes[*level + 1]; + ret = walk_up_proc(trans, root, path, wc); + if (ret > 0) + return 0; - clean_tree_block(trans, root, path->nodes[i]); - ret = btrfs_free_extent(trans, root, - path->nodes[i]->start, - path->nodes[i]->len, - parent->start, - btrfs_header_owner(parent), - *level, 0); - BUG_ON(ret); - if (path->locks[*level]) { - btrfs_tree_unlock(path->nodes[i]); - path->locks[i] = 0; + if (path->locks[level]) { + btrfs_tree_unlock(path->nodes[level]); + path->locks[level] = 0; } - free_extent_buffer(path->nodes[i]); - path->nodes[i] = NULL; - *level = i + 1; + free_extent_buffer(path->nodes[level]); + path->nodes[level] = NULL; + level++; } } return 1; } /* - * drop the reference count on the tree rooted at 'snap'. This traverses - * the tree freeing any blocks that have a ref count of zero after being - * decremented. + * drop a subvolume tree. + * + * this function traverses the tree freeing any blocks that only + * referenced by the tree. + * + * when a shared tree block is found. this function decreases its + * reference count by one. if update_ref is true, this function + * also make sure backrefs for the shared block and all lower level + * blocks are properly updated. */ -int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root - *root) +int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref) { - int ret = 0; - int wret; - int level; struct btrfs_path *path; - int update_count; + struct btrfs_trans_handle *trans; + struct btrfs_root *tree_root = root->fs_info->tree_root; struct btrfs_root_item *root_item = &root->root_item; + struct walk_control *wc; + struct btrfs_key key; + int err = 0; + int ret; + int level; path = btrfs_alloc_path(); BUG_ON(!path); - level = btrfs_header_level(root->node); + wc = kzalloc(sizeof(*wc), GFP_NOFS); + BUG_ON(!wc); + + trans = btrfs_start_transaction(tree_root, 1); + if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) { + level = btrfs_header_level(root->node); path->nodes[level] = btrfs_lock_root_node(root); btrfs_set_lock_blocking(path->nodes[level]); path->slots[level] = 0; path->locks[level] = 1; + memset(&wc->update_progress, 0, + sizeof(wc->update_progress)); } else { - struct btrfs_key key; - struct btrfs_disk_key found_key; - struct extent_buffer *node; - btrfs_disk_key_to_cpu(&key, &root_item->drop_progress); + memcpy(&wc->update_progress, &key, + sizeof(wc->update_progress)); + level = root_item->drop_level; + BUG_ON(level == 0); path->lowest_level = level; - wret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - if (wret < 0) { - ret = wret; + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + path->lowest_level = 0; + if (ret < 0) { + err = ret; goto out; } - node = path->nodes[level]; - btrfs_node_key(node, &found_key, path->slots[level]); - WARN_ON(memcmp(&found_key, &root_item->drop_progress, - sizeof(found_key))); + btrfs_node_key_to_cpu(path->nodes[level], &key, + path->slots[level]); + WARN_ON(memcmp(&key, &wc->update_progress, sizeof(key))); + /* * unlock our path, this is safe because only this * function is allowed to delete this snapshot */ btrfs_unlock_up_safe(path, 0); + + level = btrfs_header_level(root->node); + while (1) { + btrfs_tree_lock(path->nodes[level]); + btrfs_set_lock_blocking(path->nodes[level]); + + ret = btrfs_lookup_extent_info(trans, root, + path->nodes[level]->start, + path->nodes[level]->len, + &wc->refs[level], + &wc->flags[level]); + BUG_ON(ret); + BUG_ON(wc->refs[level] == 0); + + if (level == root_item->drop_level) + break; + + btrfs_tree_unlock(path->nodes[level]); + WARN_ON(wc->refs[level] != 1); + level--; + } } + + wc->level = level; + wc->shared_level = -1; + wc->stage = DROP_REFERENCE; + wc->update_ref = update_ref; + wc->keep_locks = 0; + while (1) { - unsigned long update; - wret = walk_down_tree(trans, root, path, &level); - if (wret > 0) + ret = walk_down_tree(trans, root, path, wc); + if (ret < 0) { + err = ret; break; - if (wret < 0) - ret = wret; + } - wret = walk_up_tree(trans, root, path, &level, - BTRFS_MAX_LEVEL); - if (wret > 0) + ret = walk_up_tree(trans, root, path, wc, BTRFS_MAX_LEVEL); + if (ret < 0) { + err = ret; break; - if (wret < 0) - ret = wret; - if (trans->transaction->in_commit || - trans->transaction->delayed_refs.flushing) { - ret = -EAGAIN; + } + + if (ret > 0) { + BUG_ON(wc->stage != DROP_REFERENCE); break; } - for (update_count = 0; update_count < 16; update_count++) { + + if (wc->stage == DROP_REFERENCE) { + level = wc->level; + btrfs_node_key(path->nodes[level], + &root_item->drop_progress, + path->slots[level]); + root_item->drop_level = level; + } + + BUG_ON(wc->level == 0); + if (trans->transaction->in_commit || + trans->transaction->delayed_refs.flushing) { + ret = btrfs_update_root(trans, tree_root, + &root->root_key, + root_item); + BUG_ON(ret); + + btrfs_end_transaction(trans, tree_root); + trans = btrfs_start_transaction(tree_root, 1); + } else { + unsigned long update; update = trans->delayed_ref_updates; trans->delayed_ref_updates = 0; if (update) - btrfs_run_delayed_refs(trans, root, update); - else - break; + btrfs_run_delayed_refs(trans, tree_root, + update); } } + btrfs_release_path(root, path); + BUG_ON(err); + + ret = btrfs_del_root(trans, tree_root, &root->root_key); + BUG_ON(ret); + + free_extent_buffer(root->node); + free_extent_buffer(root->commit_root); + kfree(root); out: + btrfs_end_transaction(trans, tree_root); + kfree(wc); btrfs_free_path(path); - return ret; + return err; } +/* + * drop subtree rooted at tree block 'node'. + * + * NOTE: this function will unlock and release tree block 'node' + */ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *node, struct extent_buffer *parent) { struct btrfs_path *path; + struct walk_control *wc; int level; int parent_level; int ret = 0; int wret; + BUG_ON(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID); + path = btrfs_alloc_path(); BUG_ON(!path); + wc = kzalloc(sizeof(*wc), GFP_NOFS); + BUG_ON(!wc); + btrfs_assert_tree_locked(parent); parent_level = btrfs_header_level(parent); extent_buffer_get(parent); @@ -4817,24 +5024,33 @@ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, btrfs_assert_tree_locked(node); level = btrfs_header_level(node); - extent_buffer_get(node); path->nodes[level] = node; path->slots[level] = 0; + path->locks[level] = 1; + + wc->refs[parent_level] = 1; + wc->flags[parent_level] = BTRFS_BLOCK_FLAG_FULL_BACKREF; + wc->level = level; + wc->shared_level = -1; + wc->stage = DROP_REFERENCE; + wc->update_ref = 0; + wc->keep_locks = 1; while (1) { - wret = walk_down_tree(trans, root, path, &level); - if (wret < 0) + wret = walk_down_tree(trans, root, path, wc); + if (wret < 0) { ret = wret; - if (wret != 0) break; + } - wret = walk_up_tree(trans, root, path, &level, parent_level); + wret = walk_up_tree(trans, root, path, wc, parent_level); if (wret < 0) ret = wret; if (wret != 0) break; } + kfree(wc); btrfs_free_path(path); return ret; } diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 126477eaecf..7c3cd248d8d 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -151,7 +151,10 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans, } if (end_pos > isize) { i_size_write(inode, end_pos); - btrfs_update_inode(trans, root, inode); + /* we've only changed i_size in ram, and we haven't updated + * the disk i_size. There is no need to log the inode + * at this time. + */ } err = btrfs_end_transaction(trans, root); out_unlock: diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index dbe1aabf96c..7ffa3d34ea1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3580,12 +3580,6 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, owner = 1; BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0, alloc_hint, owner); - if ((mode & S_IFREG)) { - if (btrfs_test_opt(root, NODATASUM)) - BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM; - if (btrfs_test_opt(root, NODATACOW)) - BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; - } key[0].objectid = objectid; btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY); @@ -3640,6 +3634,13 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, btrfs_inherit_iflags(inode, dir); + if ((mode & S_IFREG)) { + if (btrfs_test_opt(root, NODATASUM)) + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM; + if (btrfs_test_opt(root, NODATACOW)) + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; + } + insert_inode_hash(inode); inode_tree_add(inode); return inode; @@ -5082,6 +5083,7 @@ static long btrfs_fallocate(struct inode *inode, int mode, u64 mask = BTRFS_I(inode)->root->sectorsize - 1; struct extent_map *em; struct btrfs_trans_handle *trans; + struct btrfs_root *root; int ret; alloc_start = offset & ~mask; @@ -5100,6 +5102,13 @@ static long btrfs_fallocate(struct inode *inode, int mode, goto out; } + root = BTRFS_I(inode)->root; + + ret = btrfs_check_data_free_space(root, inode, + alloc_end - alloc_start); + if (ret) + goto out; + locked_end = alloc_end - 1; while (1) { struct btrfs_ordered_extent *ordered; @@ -5107,7 +5116,7 @@ static long btrfs_fallocate(struct inode *inode, int mode, trans = btrfs_start_transaction(BTRFS_I(inode)->root, 1); if (!trans) { ret = -EIO; - goto out; + goto out_free; } /* the extent lock is ordered inside the running @@ -5168,6 +5177,8 @@ static long btrfs_fallocate(struct inode *inode, int mode, GFP_NOFS); btrfs_end_transaction(trans, BTRFS_I(inode)->root); +out_free: + btrfs_free_reserved_data_space(root, inode, alloc_end - alloc_start); out: mutex_unlock(&inode->i_mutex); return ret; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index eff18f5b536..9f4db848db1 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1028,7 +1028,8 @@ static long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, struct btrfs_file_extent_item); comp = btrfs_file_extent_compression(leaf, extent); type = btrfs_file_extent_type(leaf, extent); - if (type == BTRFS_FILE_EXTENT_REG) { + if (type == BTRFS_FILE_EXTENT_REG || + type == BTRFS_FILE_EXTENT_PREALLOC) { disko = btrfs_file_extent_disk_bytenr(leaf, extent); diskl = btrfs_file_extent_disk_num_bytes(leaf, @@ -1051,7 +1052,8 @@ static long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, new_key.objectid = inode->i_ino; new_key.offset = key.offset + destoff - off; - if (type == BTRFS_FILE_EXTENT_REG) { + if (type == BTRFS_FILE_EXTENT_REG || + type == BTRFS_FILE_EXTENT_PREALLOC) { ret = btrfs_insert_empty_item(trans, root, path, &new_key, size); if (ret) diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index b23dc209ae1..00839793477 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1788,7 +1788,7 @@ static void merge_func(struct btrfs_work *work) btrfs_end_transaction(trans, root); } - btrfs_drop_dead_root(reloc_root); + btrfs_drop_snapshot(reloc_root, 0); if (atomic_dec_and_test(async->num_pending)) complete(async->done); @@ -2075,9 +2075,6 @@ static int do_relocation(struct btrfs_trans_handle *trans, ret = btrfs_drop_subtree(trans, root, eb, upper->eb); BUG_ON(ret); - - btrfs_tree_unlock(eb); - free_extent_buffer(eb); } if (!lowest) { btrfs_tree_unlock(upper->eb); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 4e83457ea25..2dbf1c1f56e 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -593,6 +593,7 @@ int btrfs_defrag_root(struct btrfs_root *root, int cacheonly) return 0; } +#if 0 /* * when dropping snapshots, we generate a ton of delayed refs, and it makes * sense not to join the transaction while it is trying to flush the current @@ -681,6 +682,7 @@ int btrfs_drop_dead_root(struct btrfs_root *root) btrfs_btree_balance_dirty(tree_root, nr); return ret; } +#endif /* * new snapshots need to be created at a very specific time in the @@ -1081,7 +1083,7 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root) while (!list_empty(&list)) { root = list_entry(list.next, struct btrfs_root, root_list); list_del_init(&root->root_list); - btrfs_drop_dead_root(root); + btrfs_drop_snapshot(root, 0); } return 0; } diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index ff231ad2389..ff27a296584 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -296,12 +296,15 @@ static int inotify_fasync(int fd, struct file *file, int on) static int inotify_release(struct inode *ignored, struct file *file) { struct fsnotify_group *group = file->private_data; + struct user_struct *user = group->inotify_data.user; fsnotify_clear_marks_by_group(group); /* free this group, matching get was inotify_init->fsnotify_obtain_group */ fsnotify_put_group(group); + atomic_dec(&user->inotify_devs); + return 0; } diff --git a/mm/nommu.c b/mm/nommu.c index bf0cc762a7d..53cab10fece 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -238,6 +238,27 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, } EXPORT_SYMBOL(get_user_pages); +/** + * follow_pfn - look up PFN at a user virtual address + * @vma: memory mapping + * @address: user virtual address + * @pfn: location to store found PFN + * + * Only IO mappings and raw PFN mappings are allowed. + * + * Returns zero and the pfn at @pfn on success, -ve otherwise. + */ +int follow_pfn(struct vm_area_struct *vma, unsigned long address, + unsigned long *pfn) +{ + if (!(vma->vm_flags & (VM_IO | VM_PFNMAP))) + return -EINVAL; + + *pfn = address >> PAGE_SHIFT; + return 0; +} +EXPORT_SYMBOL(follow_pfn); + DEFINE_RWLOCK(vmlist_lock); struct vm_struct *vmlist; diff --git a/tools/perf/perf.h b/tools/perf/perf.h index ce394192c85..8f729aedc1a 100644 --- a/tools/perf/perf.h +++ b/tools/perf/perf.h @@ -19,6 +19,16 @@ #define cpu_relax() asm volatile("" ::: "memory"); #endif +#ifdef __sh__ +#include "../../arch/sh/include/asm/unistd.h" +#if defined(__SH4A__) || defined(__SH5__) +# define rmb() asm volatile("synco" ::: "memory") +#else +# define rmb() asm volatile("" ::: "memory") +#endif +#define cpu_relax() asm volatile("" ::: "memory") +#endif + #include <time.h> #include <unistd.h> #include <sys/types.h> |