From fc4ac7a5f5996712d9123ae4850948c640edb315 Mon Sep 17 00:00:00 2001 From: Andres Salomon Date: Sat, 5 Jun 2010 11:40:42 -0400 Subject: x86: use __ASSEMBLY__ rather than __ASSEMBLER__ As Ingo pointed out in a separate patch, we should be using __ASSEMBLY__. Make that the case in pgtable headers. Signed-off-by: Andres Salomon LKML-Reference: <20100605114042.35ac69c1@dev.queued.net> Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/pgtable_32_types.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/pgtable_32_types.h b/arch/x86/include/asm/pgtable_32_types.h index 5e67c1532314..ed5903be26fe 100644 --- a/arch/x86/include/asm/pgtable_32_types.h +++ b/arch/x86/include/asm/pgtable_32_types.h @@ -26,7 +26,7 @@ */ #define VMALLOC_OFFSET (8 * 1024 * 1024) -#ifndef __ASSEMBLER__ +#ifndef __ASSEMBLY__ extern bool __vmalloc_start_set; /* set once high_memory is set */ #endif -- cgit v1.2.3 From a4384df3e24579d6292a1b3b41d500349948f30b Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 8 Jun 2010 11:44:32 -0700 Subject: x86, irq: Rename gsi_end gsi_top, and fix off by one errors When I introduced the global variable gsi_end I thought gsi_end on io_apics was one past the end of the gsi range for the io_apic. After it was pointed out the the range on io_apics was inclusive I changed my global variable to match. That was a big mistake. Inclusive semantics without a range start cannot describe the case when no gsi's are allocated. Describing the case where no gsi's are allocated is important in sfi.c and mpparse.c so that we can assign gsi numbers instead of blindly copying the gsi assignments the BIOS has done as we do in the acpi case. To keep from getting the global variable confused with the gsi range end rename it gsi_top. To allow describing the case where no gsi's are allocated have gsi_top be one place the highest gsi number seen in the system. This fixes an off by one bug in sfi.c: Reported-by: jacob pan This fixes the same off by one bug in mpparse.c: This fixes an off unreachable by one bug in acpi/boot.c:irq_to_gsi Reported-by: Yinghai Signed-off-by: Eric W. Biederman LKML-Reference: Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/io_apic.h | 4 ++-- arch/x86/kernel/acpi/boot.c | 8 ++++---- arch/x86/kernel/apic/io_apic.c | 12 ++++++------ arch/x86/kernel/mpparse.c | 2 +- arch/x86/kernel/sfi.c | 2 +- 5 files changed, 14 insertions(+), 14 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/io_apic.h b/arch/x86/include/asm/io_apic.h index 63cb4096c3dc..9cb2edb87c2f 100644 --- a/arch/x86/include/asm/io_apic.h +++ b/arch/x86/include/asm/io_apic.h @@ -183,7 +183,7 @@ struct mp_ioapic_gsi{ u32 gsi_end; }; extern struct mp_ioapic_gsi mp_gsi_routing[]; -extern u32 gsi_end; +extern u32 gsi_top; int mp_find_ioapic(u32 gsi); int mp_find_ioapic_pin(int ioapic, u32 gsi); void __init mp_register_ioapic(int id, u32 address, u32 gsi_base); @@ -197,7 +197,7 @@ static const int timer_through_8259 = 0; static inline void ioapic_init_mappings(void) { } static inline void ioapic_insert_resources(void) { } static inline void probe_nr_irqs_gsi(void) { } -#define gsi_end (NR_IRQS_LEGACY - 1) +#define gsi_top (NR_IRQS_LEGACY) static inline int mp_find_ioapic(u32 gsi) { return 0; } struct io_apic_irq_attr; diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 60cc4058ed5f..c05872aa3ce0 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -118,7 +118,7 @@ static unsigned int gsi_to_irq(unsigned int gsi) if (gsi >= NR_IRQS_LEGACY) irq = gsi; else - irq = gsi_end + 1 + gsi; + irq = gsi_top + gsi; return irq; } @@ -129,10 +129,10 @@ static u32 irq_to_gsi(int irq) if (irq < NR_IRQS_LEGACY) gsi = isa_irq_to_gsi[irq]; - else if (irq <= gsi_end) + else if (irq < gsi_top) gsi = irq; - else if (irq <= (gsi_end + NR_IRQS_LEGACY)) - gsi = irq - gsi_end; + else if (irq < (gsi_top + NR_IRQS_LEGACY)) + gsi = irq - gsi_top; else gsi = 0xffffffff; diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 33f3563a2a52..e41ed24ab26d 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -89,8 +89,8 @@ int nr_ioapics; /* IO APIC gsi routing info */ struct mp_ioapic_gsi mp_gsi_routing[MAX_IO_APICS]; -/* The last gsi number used */ -u32 gsi_end; +/* The one past the highest gsi number used */ +u32 gsi_top; /* MP IRQ source entries */ struct mpc_intsrc mp_irqs[MAX_IRQ_SOURCES]; @@ -1035,7 +1035,7 @@ static int pin_2_irq(int idx, int apic, int pin) if (gsi >= NR_IRQS_LEGACY) irq = gsi; else - irq = gsi_end + 1 + gsi; + irq = gsi_top + gsi; } #ifdef CONFIG_X86_32 @@ -3853,7 +3853,7 @@ void __init probe_nr_irqs_gsi(void) { int nr; - nr = gsi_end + 1 + NR_IRQS_LEGACY; + nr = gsi_top + NR_IRQS_LEGACY; if (nr > nr_irqs_gsi) nr_irqs_gsi = nr; @@ -4294,8 +4294,8 @@ void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) */ nr_ioapic_registers[idx] = entries; - if (mp_gsi_routing[idx].gsi_end > gsi_end) - gsi_end = mp_gsi_routing[idx].gsi_end; + if (mp_gsi_routing[idx].gsi_end >= gsi_top) + gsi_top = mp_gsi_routing[idx].gsi_end + 1; printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, " "GSI %d-%d\n", idx, mp_ioapics[idx].apicid, diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index 5ae5d2426edf..d86dbf7e54be 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -123,7 +123,7 @@ static void __init MP_ioapic_info(struct mpc_ioapic *m) printk(KERN_INFO "I/O APIC #%d Version %d at 0x%X.\n", m->apicid, m->apicver, m->apicaddr); - mp_register_ioapic(m->apicid, m->apicaddr, gsi_end + 1); + mp_register_ioapic(m->apicid, m->apicaddr, gsi_top); } static void print_MP_intsrc_info(struct mpc_intsrc *m) diff --git a/arch/x86/kernel/sfi.c b/arch/x86/kernel/sfi.c index 7ded57896c0a..cb22acf3ed09 100644 --- a/arch/x86/kernel/sfi.c +++ b/arch/x86/kernel/sfi.c @@ -93,7 +93,7 @@ static int __init sfi_parse_ioapic(struct sfi_table_header *table) pentry = (struct sfi_apic_table_entry *)sb->pentry; for (i = 0; i < num; i++) { - mp_register_ioapic(i, pentry->phys_addr, gsi_end + 1); + mp_register_ioapic(i, pentry->phys_addr, gsi_top); pentry++; } -- cgit v1.2.3 From d11007703c31db534674ebeeb9eb047bbbe758bd Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Thu, 10 Jun 2010 13:25:01 +0200 Subject: perf_events: Fix Intel Westmere event constraints Based on Intel Vol3b (March 2010), the event SNOOPQ_REQUEST_OUTSTANDING is restricted to counters 0,1 so update the event table for Intel Westmere accordingly. Signed-off-by: Stephane Eranian Cc: peterz@infradead.org Cc: paulus@samba.org Cc: davem@davemloft.net Cc: fweisbec@gmail.com Cc: perfmon2-devel@lists.sf.net Cc: eranian@gmail.com Cc: # .34.x LKML-Reference: <4c10cb56.5120e30a.2eb4.ffffc3de@mx.google.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_intel.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index fdbc652d3feb..214ac860ebe0 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -72,6 +72,7 @@ static struct event_constraint intel_westmere_event_constraints[] = INTEL_EVENT_CONSTRAINT(0x51, 0x3), /* L1D */ INTEL_EVENT_CONSTRAINT(0x60, 0x1), /* OFFCORE_REQUESTS_OUTSTANDING */ INTEL_EVENT_CONSTRAINT(0x63, 0x3), /* CACHE_LOCK_CYCLES */ + INTEL_EVENT_CONSTRAINT(0xb3, 0x1), /* SNOOPQ_REQUEST_OUTSTANDING */ EVENT_CONSTRAINT_END }; -- cgit v1.2.3 From dd4c4f17d722ffeb2515bf781400675a30fcead7 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Fri, 28 May 2010 16:32:14 -0400 Subject: suspend: Move NVS save/restore code to generic suspend functionality Saving platform non-volatile state may be required for suspend to RAM as well as hibernation. Move it to more generic code. Signed-off-by: Matthew Garrett Acked-by: Rafael J. Wysocki Tested-by: Maxim Levitsky Signed-off-by: Len Brown --- arch/x86/kernel/e820.c | 2 +- drivers/acpi/sleep.c | 12 ++-- include/linux/suspend.h | 26 ++++----- kernel/power/Kconfig | 9 +-- kernel/power/Makefile | 2 +- kernel/power/hibernate_nvs.c | 136 ------------------------------------------- kernel/power/nvs.c | 136 +++++++++++++++++++++++++++++++++++++++++++ kernel/power/suspend.c | 6 ++ 8 files changed, 168 insertions(+), 161 deletions(-) delete mode 100644 kernel/power/hibernate_nvs.c create mode 100644 kernel/power/nvs.c (limited to 'arch/x86') diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 7bca3c6a02fb..0d6fc71bedb1 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -729,7 +729,7 @@ static int __init e820_mark_nvs_memory(void) struct e820entry *ei = &e820.map[i]; if (ei->type == E820_NVS) - hibernate_nvs_register(ei->addr, ei->size); + suspend_nvs_register(ei->addr, ei->size); } return 0; diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index 4ab2275b4461..bcaa6efa8136 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -393,7 +393,7 @@ static int acpi_hibernation_begin(void) { int error; - error = s4_no_nvs ? 0 : hibernate_nvs_alloc(); + error = s4_no_nvs ? 0 : suspend_nvs_alloc(); if (!error) { acpi_target_sleep_state = ACPI_STATE_S4; acpi_sleep_tts_switch(acpi_target_sleep_state); @@ -407,7 +407,7 @@ static int acpi_hibernation_pre_snapshot(void) int error = acpi_pm_prepare(); if (!error) - hibernate_nvs_save(); + suspend_nvs_save(); return error; } @@ -432,7 +432,7 @@ static int acpi_hibernation_enter(void) static void acpi_hibernation_finish(void) { - hibernate_nvs_free(); + suspend_nvs_free(); acpi_pm_finish(); } @@ -452,7 +452,7 @@ static void acpi_hibernation_leave(void) panic("ACPI S4 hardware signature mismatch"); } /* Restore the NVS memory area */ - hibernate_nvs_restore(); + suspend_nvs_restore(); } static int acpi_pm_pre_restore(void) @@ -501,7 +501,7 @@ static int acpi_hibernation_begin_old(void) if (!error) { if (!s4_no_nvs) - error = hibernate_nvs_alloc(); + error = suspend_nvs_alloc(); if (!error) acpi_target_sleep_state = ACPI_STATE_S4; } @@ -513,7 +513,7 @@ static int acpi_hibernation_pre_snapshot_old(void) int error = acpi_pm_disable_gpes(); if (!error) - hibernate_nvs_save(); + suspend_nvs_save(); return error; } diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 5e781d824e6d..bc7d6bb4cd8e 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -256,22 +256,22 @@ static inline int hibernate(void) { return -ENOSYS; } static inline bool system_entering_hibernation(void) { return false; } #endif /* CONFIG_HIBERNATION */ -#ifdef CONFIG_HIBERNATION_NVS -extern int hibernate_nvs_register(unsigned long start, unsigned long size); -extern int hibernate_nvs_alloc(void); -extern void hibernate_nvs_free(void); -extern void hibernate_nvs_save(void); -extern void hibernate_nvs_restore(void); -#else /* CONFIG_HIBERNATION_NVS */ -static inline int hibernate_nvs_register(unsigned long a, unsigned long b) +#ifdef CONFIG_SUSPEND_NVS +extern int suspend_nvs_register(unsigned long start, unsigned long size); +extern int suspend_nvs_alloc(void); +extern void suspend_nvs_free(void); +extern void suspend_nvs_save(void); +extern void suspend_nvs_restore(void); +#else /* CONFIG_SUSPEND_NVS */ +static inline int suspend_nvs_register(unsigned long a, unsigned long b) { return 0; } -static inline int hibernate_nvs_alloc(void) { return 0; } -static inline void hibernate_nvs_free(void) {} -static inline void hibernate_nvs_save(void) {} -static inline void hibernate_nvs_restore(void) {} -#endif /* CONFIG_HIBERNATION_NVS */ +static inline int suspend_nvs_alloc(void) { return 0; } +static inline void suspend_nvs_free(void) {} +static inline void suspend_nvs_save(void) {} +static inline void suspend_nvs_restore(void) {} +#endif /* CONFIG_SUSPEND_NVS */ #ifdef CONFIG_PM_SLEEP void save_processor_state(void); diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 5c36ea9d55d2..ca6066a6952e 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -99,9 +99,13 @@ config PM_SLEEP_ADVANCED_DEBUG depends on PM_ADVANCED_DEBUG default n +config SUSPEND_NVS + bool + config SUSPEND bool "Suspend to RAM and standby" depends on PM && ARCH_SUSPEND_POSSIBLE + select SUSPEND_NVS if HAS_IOMEM default y ---help--- Allow the system to enter sleep states in which main memory is @@ -130,13 +134,10 @@ config SUSPEND_FREEZER Turning OFF this setting is NOT recommended! If in doubt, say Y. -config HIBERNATION_NVS - bool - config HIBERNATION bool "Hibernation (aka 'suspend to disk')" depends on PM && SWAP && ARCH_HIBERNATION_POSSIBLE - select HIBERNATION_NVS if HAS_IOMEM + select SUSPEND_NVS if HAS_IOMEM ---help--- Enable the suspend to disk (STD) functionality, which is usually called "hibernation" in user interfaces. STD checkpoints the diff --git a/kernel/power/Makefile b/kernel/power/Makefile index 524e058dcf06..f9063c6b185d 100644 --- a/kernel/power/Makefile +++ b/kernel/power/Makefile @@ -10,6 +10,6 @@ obj-$(CONFIG_SUSPEND) += suspend.o obj-$(CONFIG_PM_TEST_SUSPEND) += suspend_test.o obj-$(CONFIG_HIBERNATION) += hibernate.o snapshot.o swap.o user.o \ block_io.o -obj-$(CONFIG_HIBERNATION_NVS) += hibernate_nvs.o +obj-$(CONFIG_SUSPEND_NVS) += nvs.o obj-$(CONFIG_MAGIC_SYSRQ) += poweroff.o diff --git a/kernel/power/hibernate_nvs.c b/kernel/power/hibernate_nvs.c deleted file mode 100644 index fdcad9ed5a7b..000000000000 --- a/kernel/power/hibernate_nvs.c +++ /dev/null @@ -1,136 +0,0 @@ -/* - * linux/kernel/power/hibernate_nvs.c - Routines for handling NVS memory - * - * Copyright (C) 2008,2009 Rafael J. Wysocki , Novell Inc. - * - * This file is released under the GPLv2. - */ - -#include -#include -#include -#include -#include -#include - -/* - * Platforms, like ACPI, may want us to save some memory used by them during - * hibernation and to restore the contents of this memory during the subsequent - * resume. The code below implements a mechanism allowing us to do that. - */ - -struct nvs_page { - unsigned long phys_start; - unsigned int size; - void *kaddr; - void *data; - struct list_head node; -}; - -static LIST_HEAD(nvs_list); - -/** - * hibernate_nvs_register - register platform NVS memory region to save - * @start - physical address of the region - * @size - size of the region - * - * The NVS region need not be page-aligned (both ends) and we arrange - * things so that the data from page-aligned addresses in this region will - * be copied into separate RAM pages. - */ -int hibernate_nvs_register(unsigned long start, unsigned long size) -{ - struct nvs_page *entry, *next; - - while (size > 0) { - unsigned int nr_bytes; - - entry = kzalloc(sizeof(struct nvs_page), GFP_KERNEL); - if (!entry) - goto Error; - - list_add_tail(&entry->node, &nvs_list); - entry->phys_start = start; - nr_bytes = PAGE_SIZE - (start & ~PAGE_MASK); - entry->size = (size < nr_bytes) ? size : nr_bytes; - - start += entry->size; - size -= entry->size; - } - return 0; - - Error: - list_for_each_entry_safe(entry, next, &nvs_list, node) { - list_del(&entry->node); - kfree(entry); - } - return -ENOMEM; -} - -/** - * hibernate_nvs_free - free data pages allocated for saving NVS regions - */ -void hibernate_nvs_free(void) -{ - struct nvs_page *entry; - - list_for_each_entry(entry, &nvs_list, node) - if (entry->data) { - free_page((unsigned long)entry->data); - entry->data = NULL; - if (entry->kaddr) { - iounmap(entry->kaddr); - entry->kaddr = NULL; - } - } -} - -/** - * hibernate_nvs_alloc - allocate memory necessary for saving NVS regions - */ -int hibernate_nvs_alloc(void) -{ - struct nvs_page *entry; - - list_for_each_entry(entry, &nvs_list, node) { - entry->data = (void *)__get_free_page(GFP_KERNEL); - if (!entry->data) { - hibernate_nvs_free(); - return -ENOMEM; - } - } - return 0; -} - -/** - * hibernate_nvs_save - save NVS memory regions - */ -void hibernate_nvs_save(void) -{ - struct nvs_page *entry; - - printk(KERN_INFO "PM: Saving platform NVS memory\n"); - - list_for_each_entry(entry, &nvs_list, node) - if (entry->data) { - entry->kaddr = ioremap(entry->phys_start, entry->size); - memcpy(entry->data, entry->kaddr, entry->size); - } -} - -/** - * hibernate_nvs_restore - restore NVS memory regions - * - * This function is going to be called with interrupts disabled, so it - * cannot iounmap the virtual addresses used to access the NVS region. - */ -void hibernate_nvs_restore(void) -{ - struct nvs_page *entry; - - printk(KERN_INFO "PM: Restoring platform NVS memory\n"); - - list_for_each_entry(entry, &nvs_list, node) - if (entry->data) - memcpy(entry->kaddr, entry->data, entry->size); -} diff --git a/kernel/power/nvs.c b/kernel/power/nvs.c new file mode 100644 index 000000000000..1836db60bbb6 --- /dev/null +++ b/kernel/power/nvs.c @@ -0,0 +1,136 @@ +/* + * linux/kernel/power/hibernate_nvs.c - Routines for handling NVS memory + * + * Copyright (C) 2008,2009 Rafael J. Wysocki , Novell Inc. + * + * This file is released under the GPLv2. + */ + +#include +#include +#include +#include +#include +#include + +/* + * Platforms, like ACPI, may want us to save some memory used by them during + * suspend and to restore the contents of this memory during the subsequent + * resume. The code below implements a mechanism allowing us to do that. + */ + +struct nvs_page { + unsigned long phys_start; + unsigned int size; + void *kaddr; + void *data; + struct list_head node; +}; + +static LIST_HEAD(nvs_list); + +/** + * suspend_nvs_register - register platform NVS memory region to save + * @start - physical address of the region + * @size - size of the region + * + * The NVS region need not be page-aligned (both ends) and we arrange + * things so that the data from page-aligned addresses in this region will + * be copied into separate RAM pages. + */ +int suspend_nvs_register(unsigned long start, unsigned long size) +{ + struct nvs_page *entry, *next; + + while (size > 0) { + unsigned int nr_bytes; + + entry = kzalloc(sizeof(struct nvs_page), GFP_KERNEL); + if (!entry) + goto Error; + + list_add_tail(&entry->node, &nvs_list); + entry->phys_start = start; + nr_bytes = PAGE_SIZE - (start & ~PAGE_MASK); + entry->size = (size < nr_bytes) ? size : nr_bytes; + + start += entry->size; + size -= entry->size; + } + return 0; + + Error: + list_for_each_entry_safe(entry, next, &nvs_list, node) { + list_del(&entry->node); + kfree(entry); + } + return -ENOMEM; +} + +/** + * suspend_nvs_free - free data pages allocated for saving NVS regions + */ +void suspend_nvs_free(void) +{ + struct nvs_page *entry; + + list_for_each_entry(entry, &nvs_list, node) + if (entry->data) { + free_page((unsigned long)entry->data); + entry->data = NULL; + if (entry->kaddr) { + iounmap(entry->kaddr); + entry->kaddr = NULL; + } + } +} + +/** + * suspend_nvs_alloc - allocate memory necessary for saving NVS regions + */ +int suspend_nvs_alloc(void) +{ + struct nvs_page *entry; + + list_for_each_entry(entry, &nvs_list, node) { + entry->data = (void *)__get_free_page(GFP_KERNEL); + if (!entry->data) { + suspend_nvs_free(); + return -ENOMEM; + } + } + return 0; +} + +/** + * suspend_nvs_save - save NVS memory regions + */ +void suspend_nvs_save(void) +{ + struct nvs_page *entry; + + printk(KERN_INFO "PM: Saving platform NVS memory\n"); + + list_for_each_entry(entry, &nvs_list, node) + if (entry->data) { + entry->kaddr = ioremap(entry->phys_start, entry->size); + memcpy(entry->data, entry->kaddr, entry->size); + } +} + +/** + * suspend_nvs_restore - restore NVS memory regions + * + * This function is going to be called with interrupts disabled, so it + * cannot iounmap the virtual addresses used to access the NVS region. + */ +void suspend_nvs_restore(void) +{ + struct nvs_page *entry; + + printk(KERN_INFO "PM: Restoring platform NVS memory\n"); + + list_for_each_entry(entry, &nvs_list, node) + if (entry->data) + memcpy(entry->kaddr, entry->data, entry->size); +} diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index 56e7dbb8b996..f37cb7dd4402 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -16,6 +16,12 @@ #include #include #include +#include +#include +#include +#include +#include +#include #include "power.h" -- cgit v1.2.3 From 23b764d056bfec0a212a67229074ac281e86e021 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 10 Jun 2010 13:10:36 +0200 Subject: percpu, x86: Avoid warnings of unused variables in per cpu Avoid hundreds of warnings with a gcc 4.6 -Wall build. Signed-off-by: Andi Kleen Acked-by: Tejun Heo Signed-off-by: Ingo Molnar --- arch/x86/include/asm/percpu.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/percpu.h b/arch/x86/include/asm/percpu.h index 0797e748d280..cd28f9ad910d 100644 --- a/arch/x86/include/asm/percpu.h +++ b/arch/x86/include/asm/percpu.h @@ -77,6 +77,7 @@ do { \ if (0) { \ pto_T__ pto_tmp__; \ pto_tmp__ = (val); \ + (void)pto_tmp__; \ } \ switch (sizeof(var)) { \ case 1: \ @@ -115,6 +116,7 @@ do { \ if (0) { \ pao_T__ pao_tmp__; \ pao_tmp__ = (val); \ + (void)pao_tmp__; \ } \ switch (sizeof(var)) { \ case 1: \ -- cgit v1.2.3 From cf3bdc29fcbf2cb4cfa238591021d41ea8f8431f Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 10 Jun 2010 13:10:40 +0200 Subject: x86, setup: Set ax register in boot vga query Catch missing conversion to the register structure "glove box" scheme. Found by gcc 4.6's new warnings. Signed-off-by: Andi Kleen LKML-Reference: <20100610111040.F1781B1A2B@basil.firstfloor.org> Signed-off-by: H. Peter Anvin --- arch/x86/boot/video-vga.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/boot/video-vga.c b/arch/x86/boot/video-vga.c index ed7aeff786b2..45bc9402aa49 100644 --- a/arch/x86/boot/video-vga.c +++ b/arch/x86/boot/video-vga.c @@ -41,13 +41,12 @@ static __videocard video_vga; static u8 vga_set_basic_mode(void) { struct biosregs ireg, oreg; - u16 ax; u8 mode; initregs(&ireg); /* Query current mode */ - ax = 0x0f00; + ireg.ax = 0x0f00; intcall(0x10, &ireg, &oreg); mode = oreg.al; -- cgit v1.2.3 From 6a4f3b523779b67e7d560ed42652f8a59f2f9782 Mon Sep 17 00:00:00 2001 From: Venkatesh Pallipadi Date: Thu, 10 Jun 2010 17:45:01 -0700 Subject: x86, pat: Proper init of memtype subtree_max_end subtree_max_end that was recently added to struct memtype was not getting properly initialized resulting in WARNING: kmemcheck: Caught 64-bit read from uninitialized memory in memtype_rb_augment_cb() reported here https://bugzilla.kernel.org/show_bug.cgi?id=16092 This change fixes the problem. Reported-by: Christian Casteyde Tested-by: Christian Casteyde Signed-off-by: Venkatesh Pallipadi LKML-Reference: <1276217101-11515-1-git-send-email-venki@google.com> Signed-off-by: H. Peter Anvin Cc: Suresh Siddha --- arch/x86/mm/pat.c | 2 +- arch/x86/mm/pat_rbtree.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index acc15b23b743..64121a18b8cb 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -302,7 +302,7 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type, return -EINVAL; } - new = kmalloc(sizeof(struct memtype), GFP_KERNEL); + new = kzalloc(sizeof(struct memtype), GFP_KERNEL); if (!new) return -ENOMEM; diff --git a/arch/x86/mm/pat_rbtree.c b/arch/x86/mm/pat_rbtree.c index f537087bb740..f20eeec85a86 100644 --- a/arch/x86/mm/pat_rbtree.c +++ b/arch/x86/mm/pat_rbtree.c @@ -226,6 +226,7 @@ int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) if (ret_type) new->type = *ret_type; + new->subtree_max_end = new->end; memtype_rb_insert(&memtype_rbroot, new); } return err; -- cgit v1.2.3 From 124482935fb7fb9303c8a8ab930149c6a93d9910 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Fri, 18 Jun 2010 23:09:00 +0200 Subject: x86: Fix vsyscall on gcc 4.5 with -Os This fixes the -Os breaks with gcc 4.5 bug. rdtsc_barrier needs to be force inlined, otherwise user space will jump into kernel space and kill init. This also addresses http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44129 I believe. Signed-off-by: Andi Kleen LKML-Reference: <20100618210859.GA10913@basil.fritz.box> Signed-off-by: H. Peter Anvin Cc: --- arch/x86/include/asm/system.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/system.h b/arch/x86/include/asm/system.h index b8fe48ee2ed9..e7f4d33c55ed 100644 --- a/arch/x86/include/asm/system.h +++ b/arch/x86/include/asm/system.h @@ -451,7 +451,7 @@ void stop_this_cpu(void *dummy); * * (Could use an alternative three way for this if there was one.) */ -static inline void rdtsc_barrier(void) +static __always_inline void rdtsc_barrier(void) { alternative(ASM_NOP3, "mfence", X86_FEATURE_MFENCE_RDTSC); alternative(ASM_NOP3, "lfence", X86_FEATURE_LFENCE_RDTSC); -- cgit v1.2.3 From 890ffedc7c3e95595926379e28ad2e16e7d7c613 Mon Sep 17 00:00:00 2001 From: Thomas Backlund Date: Sat, 19 Jun 2010 16:32:25 +0300 Subject: x86: Fix rebooting on Dell Precision WorkStation T7400 Dell Precision WorkStation T7400 freezes on reboot unless reboot=b is used. Reference: https://qa.mandriva.com/show_bug.cgi?id=58017 Signed-off-by: Thomas Backlund LKML-Reference: <4C1CC6E9.6000701@mandriva.org> Signed-off-by: Ingo Molnar --- arch/x86/kernel/reboot.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 8e1aac86b50c..e3af342fe83a 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -228,6 +228,14 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = { DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T5400"), }, }, + { /* Handle problems with rebooting on Dell T7400's */ + .callback = set_bios_reboot, + .ident = "Dell Precision T7400", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T7400"), + }, + }, { /* Handle problems with rebooting on HP laptops */ .callback = set_bios_reboot, .ident = "HP Compaq Laptop", -- cgit v1.2.3 From 499a00e92dd9a75395081f595e681629eb1eebad Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 24 Jun 2010 14:26:47 -0700 Subject: x86, Calgary: Increase max PHB number Newer systems (x3950M2) can have 48 PHBs per chassis and 8 chassis, so bump the limits up and provide an explanation of the requirements for each class. Signed-off-by: Darrick J. Wong Acked-by: Muli Ben-Yehuda Cc: Corinna Schultz Cc: LKML-Reference: <20100624212647.GI15515@tux1.beaverton.ibm.com> [ v2: Fixed build bug, added back PHBS_PER_CALGARY == 4 ] Signed-off-by: Ingo Molnar --- arch/x86/kernel/pci-calgary_64.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index fb99f7edb341..0b96b5589f08 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -103,11 +103,16 @@ int use_calgary __read_mostly = 0; #define PMR_SOFTSTOPFAULT 0x40000000 #define PMR_HARDSTOP 0x20000000 -#define MAX_NUM_OF_PHBS 8 /* how many PHBs in total? */ -#define MAX_NUM_CHASSIS 8 /* max number of chassis */ -/* MAX_PHB_BUS_NUM is the maximal possible dev->bus->number */ -#define MAX_PHB_BUS_NUM (MAX_NUM_OF_PHBS * MAX_NUM_CHASSIS * 2) -#define PHBS_PER_CALGARY 4 +/* + * The maximum PHB bus number. + * x3950M2 (rare): 8 chassis, 48 PHBs per chassis = 384 + * x3950M2: 4 chassis, 48 PHBs per chassis = 192 + * x3950 (PCIE): 8 chassis, 32 PHBs per chassis = 256 + * x3950 (PCIX): 8 chassis, 16 PHBs per chassis = 128 + */ +#define MAX_PHB_BUS_NUM 384 + +#define PHBS_PER_CALGARY 4 /* register offsets in Calgary's internal register space */ static const unsigned long tar_offsets[] = { -- cgit v1.2.3 From a1e80fafc9f0742a1776a0490258cb64912411b0 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Wed, 30 Jun 2010 15:09:06 +0200 Subject: x86: Send a SIGTRAP for user icebp traps Before we had a generic breakpoint layer, x86 used to send a sigtrap for any debug event that happened in userspace, except if it was caused by lazy dr7 switches. Currently we only send such signal for single step or breakpoint events. However, there are three other kind of debug exceptions: - debug register access detected: trigger an exception if the next instruction touches the debug registers. We don't use it. - task switch, but we don't use tss. - icebp/int01 trap. This instruction (0xf1) is undocumented and generates an int 1 exception. Unlike single step through TF flag, it doesn't set the single step origin of the exception in dr6. icebp then used to be reported in userspace using trap signals but this have been incidentally broken with the new breakpoint code. Reenable this. Since this is the only debug event that doesn't set anything in dr6, this is all we have to check. This fixes a regression in Wine where World Of Warcraft got broken as it uses this for software protection checks purposes. And probably other apps do. Reported-and-tested-by: Alexandre Julliard Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: H. Peter Anvin Cc: Thomas Gleixner Cc: Prasad Cc: 2.6.33.x 2.6.34.x --- arch/x86/kernel/traps.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 142d70c74b02..725ef4d17cd5 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -526,6 +526,7 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs) dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) { struct task_struct *tsk = current; + int user_icebp = 0; unsigned long dr6; int si_code; @@ -534,6 +535,14 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) /* Filter out all the reserved bits which are preset to 1 */ dr6 &= ~DR6_RESERVED; + /* + * If dr6 has no reason to give us about the origin of this trap, + * then it's very likely the result of an icebp/int01 trap. + * User wants a sigtrap for that. + */ + if (!dr6 && user_mode(regs)) + user_icebp = 1; + /* Catch kmemcheck conditions first of all! */ if ((dr6 & DR_STEP) && kmemcheck_trap(regs)) return; @@ -575,7 +584,7 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) regs->flags &= ~X86_EFLAGS_TF; } si_code = get_si_code(tsk->thread.debugreg6); - if (tsk->thread.debugreg6 & (DR_STEP | DR_TRAP_BITS)) + if (tsk->thread.debugreg6 & (DR_STEP | DR_TRAP_BITS) || user_icebp) send_sigtrap(tsk, regs, error_code, si_code); preempt_conditional_cli(regs); -- cgit v1.2.3 From d596043d71ff0d7b3d0bead19b1d68c55f003093 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Wed, 30 Jun 2010 17:45:19 -0700 Subject: x86, Calgary: Limit the max PHB number to 256 The x3950 family can have as many as 256 PCI buses in a single system, so change the limits to the maximum. Since there can only be 256 PCI buses in one domain, we no longer need the BUG_ON check. Signed-off-by: Darrick J. Wong LKML-Reference: <20100701004519.GQ15515@tux1.beaverton.ibm.com> Signed-off-by: H. Peter Anvin --- arch/x86/kernel/pci-calgary_64.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 0b96b5589f08..078d4ec1a9d9 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -110,7 +110,7 @@ int use_calgary __read_mostly = 0; * x3950 (PCIE): 8 chassis, 32 PHBs per chassis = 256 * x3950 (PCIX): 8 chassis, 16 PHBs per chassis = 128 */ -#define MAX_PHB_BUS_NUM 384 +#define MAX_PHB_BUS_NUM 256 #define PHBS_PER_CALGARY 4 @@ -1056,8 +1056,6 @@ static int __init calgary_init_one(struct pci_dev *dev) struct iommu_table *tbl; int ret; - BUG_ON(dev->bus->number >= MAX_PHB_BUS_NUM); - bbar = busno_to_bbar(dev->bus->number); ret = calgary_setup_tar(dev, bbar); if (ret) -- cgit v1.2.3 From f287d332ce835f77a4f5077d2c0ef1e3f9ea42d2 Mon Sep 17 00:00:00 2001 From: Vince Weaver Date: Thu, 1 Jul 2010 15:30:16 -0400 Subject: perf, x86: Fix incorrect branches event on AMD CPUs While doing some performance counter validation tests on some assembly language programs I noticed that the "branches:u" count was very wrong on AMD machines. It looks like the wrong event was selected. Signed-off-by: Vince Weaver Acked-by: Peter Zijlstra Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Robert Richter Cc: Borislav Petkov Cc: Frederic Weisbecker Cc: LKML-Reference: Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_amd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 611df11ba15e..c2897b7b4a3b 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -102,8 +102,8 @@ static const u64 amd_perfmon_event_map[] = [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, [PERF_COUNT_HW_CACHE_REFERENCES] = 0x0080, [PERF_COUNT_HW_CACHE_MISSES] = 0x0081, - [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, - [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c2, + [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c3, }; static u64 amd_pmu_event_map(int hw_event) -- cgit v1.2.3 From b945d6b2554d550fe95caadc61e521c0ad71fb9c Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 29 May 2010 15:31:43 +0200 Subject: rbtree: Undo augmented trees performance damage and regression Reimplement augmented RB-trees without sprinkling extra branches all over the RB-tree code (which lives in the scheduler hot path). This approach is 'borrowed' from Fabio's BFQ implementation and relies on traversing the rebalance path after the RB-tree-op to correct the heap property for insertion/removal and make up for the damage done by the tree rotations. For insertion the rebalance path is trivially that from the new node upwards to the root, for removal it is that from the deepest node in the path from the to be removed node that will still be around after the removal. [ This patch also fixes a video driver regression reported by Ali Gholami Rudi - the memtype->subtree_max_end was updated incorrectly. ] Acked-by: Suresh Siddha Acked-by: Venkatesh Pallipadi Signed-off-by: Peter Zijlstra Tested-by: Ali Gholami Rudi Cc: Fabio Checconi Cc: "H. Peter Anvin" Cc: Andrew Morton Cc: Linus Torvalds LKML-Reference: <1275414172.27810.27961.camel@twins> Signed-off-by: Ingo Molnar --- arch/x86/mm/pat_rbtree.c | 34 +++----------- include/linux/rbtree.h | 13 ++++-- lib/rbtree.c | 116 +++++++++++++++++++++++++++++------------------ 3 files changed, 87 insertions(+), 76 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/pat_rbtree.c b/arch/x86/mm/pat_rbtree.c index f20eeec85a86..8acaddd0fb21 100644 --- a/arch/x86/mm/pat_rbtree.c +++ b/arch/x86/mm/pat_rbtree.c @@ -34,8 +34,7 @@ * memtype_lock protects the rbtree. */ -static void memtype_rb_augment_cb(struct rb_node *node); -static struct rb_root memtype_rbroot = RB_AUGMENT_ROOT(&memtype_rb_augment_cb); +static struct rb_root memtype_rbroot = RB_ROOT; static int is_node_overlap(struct memtype *node, u64 start, u64 end) { @@ -56,7 +55,7 @@ static u64 get_subtree_max_end(struct rb_node *node) } /* Update 'subtree_max_end' for a node, based on node and its children */ -static void update_node_max_end(struct rb_node *node) +static void memtype_rb_augment_cb(struct rb_node *node, void *__unused) { struct memtype *data; u64 max_end, child_max_end; @@ -78,25 +77,6 @@ static void update_node_max_end(struct rb_node *node) data->subtree_max_end = max_end; } -/* Update 'subtree_max_end' for a node and all its ancestors */ -static void update_path_max_end(struct rb_node *node) -{ - u64 old_max_end, new_max_end; - - while (node) { - struct memtype *data = container_of(node, struct memtype, rb); - - old_max_end = data->subtree_max_end; - update_node_max_end(node); - new_max_end = data->subtree_max_end; - - if (new_max_end == old_max_end) - break; - - node = rb_parent(node); - } -} - /* Find the first (lowest start addr) overlapping range from rb tree */ static struct memtype *memtype_rb_lowest_match(struct rb_root *root, u64 start, u64 end) @@ -190,12 +170,6 @@ failure: return -EBUSY; } -static void memtype_rb_augment_cb(struct rb_node *node) -{ - if (node) - update_path_max_end(node); -} - static void memtype_rb_insert(struct rb_root *root, struct memtype *newdata) { struct rb_node **node = &(root->rb_node); @@ -213,6 +187,7 @@ static void memtype_rb_insert(struct rb_root *root, struct memtype *newdata) rb_link_node(&newdata->rb, parent, node); rb_insert_color(&newdata->rb, root); + rb_augment_insert(&newdata->rb, memtype_rb_augment_cb, NULL); } int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) @@ -234,13 +209,16 @@ int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) struct memtype *rbt_memtype_erase(u64 start, u64 end) { + struct rb_node *deepest; struct memtype *data; data = memtype_rb_exact_match(&memtype_rbroot, start, end); if (!data) goto out; + deepest = rb_augment_erase_begin(&data->rb); rb_erase(&data->rb, &memtype_rbroot); + rb_augment_erase_end(deepest, memtype_rb_augment_cb, NULL); out: return data; } diff --git a/include/linux/rbtree.h b/include/linux/rbtree.h index fe1872e5b37e..7066acb2c530 100644 --- a/include/linux/rbtree.h +++ b/include/linux/rbtree.h @@ -110,7 +110,6 @@ struct rb_node struct rb_root { struct rb_node *rb_node; - void (*augment_cb)(struct rb_node *node); }; @@ -130,9 +129,7 @@ static inline void rb_set_color(struct rb_node *rb, int color) rb->rb_parent_color = (rb->rb_parent_color & ~1) | color; } -#define RB_ROOT (struct rb_root) { NULL, NULL, } -#define RB_AUGMENT_ROOT(x) (struct rb_root) { NULL, x} - +#define RB_ROOT (struct rb_root) { NULL, } #define rb_entry(ptr, type, member) container_of(ptr, type, member) #define RB_EMPTY_ROOT(root) ((root)->rb_node == NULL) @@ -142,6 +139,14 @@ static inline void rb_set_color(struct rb_node *rb, int color) extern void rb_insert_color(struct rb_node *, struct rb_root *); extern void rb_erase(struct rb_node *, struct rb_root *); +typedef void (*rb_augment_f)(struct rb_node *node, void *data); + +extern void rb_augment_insert(struct rb_node *node, + rb_augment_f func, void *data); +extern struct rb_node *rb_augment_erase_begin(struct rb_node *node); +extern void rb_augment_erase_end(struct rb_node *node, + rb_augment_f func, void *data); + /* Find logical next and previous nodes in a tree */ extern struct rb_node *rb_next(const struct rb_node *); extern struct rb_node *rb_prev(const struct rb_node *); diff --git a/lib/rbtree.c b/lib/rbtree.c index 15e10b1afdd2..4693f79195d3 100644 --- a/lib/rbtree.c +++ b/lib/rbtree.c @@ -44,11 +44,6 @@ static void __rb_rotate_left(struct rb_node *node, struct rb_root *root) else root->rb_node = right; rb_set_parent(node, right); - - if (root->augment_cb) { - root->augment_cb(node); - root->augment_cb(right); - } } static void __rb_rotate_right(struct rb_node *node, struct rb_root *root) @@ -72,20 +67,12 @@ static void __rb_rotate_right(struct rb_node *node, struct rb_root *root) else root->rb_node = left; rb_set_parent(node, left); - - if (root->augment_cb) { - root->augment_cb(node); - root->augment_cb(left); - } } void rb_insert_color(struct rb_node *node, struct rb_root *root) { struct rb_node *parent, *gparent; - if (root->augment_cb) - root->augment_cb(node); - while ((parent = rb_parent(node)) && rb_is_red(parent)) { gparent = rb_parent(parent); @@ -240,15 +227,12 @@ void rb_erase(struct rb_node *node, struct rb_root *root) else { struct rb_node *old = node, *left; - int old_parent_cb = 0; - int successor_parent_cb = 0; node = node->rb_right; while ((left = node->rb_left) != NULL) node = left; if (rb_parent(old)) { - old_parent_cb = 1; if (rb_parent(old)->rb_left == old) rb_parent(old)->rb_left = node; else @@ -263,10 +247,8 @@ void rb_erase(struct rb_node *node, struct rb_root *root) if (parent == old) { parent = node; } else { - successor_parent_cb = 1; if (child) rb_set_parent(child, parent); - parent->rb_left = child; node->rb_right = old->rb_right; @@ -277,24 +259,6 @@ void rb_erase(struct rb_node *node, struct rb_root *root) node->rb_left = old->rb_left; rb_set_parent(old->rb_left, node); - if (root->augment_cb) { - /* - * Here, three different nodes can have new children. - * The parent of the successor node that was selected - * to replace the node to be erased. - * The node that is getting erased and is now replaced - * by its successor. - * The parent of the node getting erased-replaced. - */ - if (successor_parent_cb) - root->augment_cb(parent); - - root->augment_cb(node); - - if (old_parent_cb) - root->augment_cb(rb_parent(old)); - } - goto color; } @@ -303,19 +267,15 @@ void rb_erase(struct rb_node *node, struct rb_root *root) if (child) rb_set_parent(child, parent); - - if (parent) { + if (parent) + { if (parent->rb_left == node) parent->rb_left = child; else parent->rb_right = child; - - if (root->augment_cb) - root->augment_cb(parent); - - } else { - root->rb_node = child; } + else + root->rb_node = child; color: if (color == RB_BLACK) @@ -323,6 +283,74 @@ void rb_erase(struct rb_node *node, struct rb_root *root) } EXPORT_SYMBOL(rb_erase); +static void rb_augment_path(struct rb_node *node, rb_augment_f func, void *data) +{ + struct rb_node *parent; + +up: + func(node, data); + parent = rb_parent(node); + if (!parent) + return; + + if (node == parent->rb_left && parent->rb_right) + func(parent->rb_right, data); + else if (parent->rb_left) + func(parent->rb_left, data); + + node = parent; + goto up; +} + +/* + * after inserting @node into the tree, update the tree to account for + * both the new entry and any damage done by rebalance + */ +void rb_augment_insert(struct rb_node *node, rb_augment_f func, void *data) +{ + if (node->rb_left) + node = node->rb_left; + else if (node->rb_right) + node = node->rb_right; + + rb_augment_path(node, func, data); +} + +/* + * before removing the node, find the deepest node on the rebalance path + * that will still be there after @node gets removed + */ +struct rb_node *rb_augment_erase_begin(struct rb_node *node) +{ + struct rb_node *deepest; + + if (!node->rb_right && !node->rb_left) + deepest = rb_parent(node); + else if (!node->rb_right) + deepest = node->rb_left; + else if (!node->rb_left) + deepest = node->rb_right; + else { + deepest = rb_next(node); + if (deepest->rb_right) + deepest = deepest->rb_right; + else if (rb_parent(deepest) != node) + deepest = rb_parent(deepest); + } + + return deepest; +} + +/* + * after removal, update the tree to account for the removed entry + * and any rebalance damage. + */ +void rb_augment_erase_end(struct rb_node *node, rb_augment_f func, void *data) +{ + if (node) + rb_augment_path(node, func, data); +} + /* * This function returns the first node (in sort order) of the tree. */ -- cgit v1.2.3 From da38f43859467a8048365b9e1cce99ccbc62b6e2 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Tue, 6 Jul 2010 11:30:49 +0300 Subject: KVM: VMX: Fix host MSR_KERNEL_GS_BASE corruption enter_lmode() and exit_lmode() modify the guest's EFER.LMA before calling vmx_set_efer(). However, the latter function depends on the value of EFER.LMA to determine whether MSR_KERNEL_GS_BASE needs reloading, via vmx_load_host_state(). With EFER.LMA changing under its feet, it took the wrong choice and corrupted userspace's %gs. This causes 32-on-64 host userspace to fault. Fix not touching EFER.LMA; instead ask vmx_set_efer() to change it. Signed-off-by: Avi Kivity --- arch/x86/kvm/vmx.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index 859a01a07dbf..ee03679efe78 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -1744,18 +1744,15 @@ static void enter_lmode(struct kvm_vcpu *vcpu) (guest_tr_ar & ~AR_TYPE_MASK) | AR_TYPE_BUSY_64_TSS); } - vcpu->arch.efer |= EFER_LMA; - vmx_set_efer(vcpu, vcpu->arch.efer); + vmx_set_efer(vcpu, vcpu->arch.efer | EFER_LMA); } static void exit_lmode(struct kvm_vcpu *vcpu) { - vcpu->arch.efer &= ~EFER_LMA; - vmcs_write32(VM_ENTRY_CONTROLS, vmcs_read32(VM_ENTRY_CONTROLS) & ~VM_ENTRY_IA32E_MODE); - vmx_set_efer(vcpu, vcpu->arch.efer); + vmx_set_efer(vcpu, vcpu->arch.efer & ~EFER_LMA); } #endif -- cgit v1.2.3 From 72550b3ae545c75897c769d43d62d4be3f3d48fe Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Wed, 7 Jul 2010 16:57:46 -0700 Subject: x86, platform: Export x86_platform to modules Export x86_platform to modules in preparation of using it for i8042 discovery control. Signed-off-by: H. Peter Anvin LKML-Reference: <1278342202-10973-1-git-send-email-feng.tang@intel.com> Cc: Thomas Gleixner Cc: Feng Tang Cc: Dmitry Torokhov --- arch/x86/kernel/x86_init.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c index 61a1e8c7e19f..ebfb8e4c9f22 100644 --- a/arch/x86/kernel/x86_init.c +++ b/arch/x86/kernel/x86_init.c @@ -5,6 +5,7 @@ */ #include #include +#include #include #include @@ -94,3 +95,5 @@ struct x86_platform_ops x86_platform = { .is_untracked_pat_range = is_ISA_range, .nmi_init = default_nmi_init }; + +EXPORT_SYMBOL_GPL(x86_platform); -- cgit v1.2.3 From c516ac583973196162b1ba7e4d597d6f6892dac0 Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Mon, 5 Jul 2010 23:03:18 +0800 Subject: x86: Add i8042 pre-detection hook to x86_platform_ops Some x86 platforms like Intel MID platforms don't have i8042 controllers, and i8042 driver's probe to some legacy IO ports may hang the MID processor. With this hook, i8042 driver can runtime check and skip the probe when the pretection fail which also saves some probe time [ hpa note: this is currently a compile-time check, which breaks the i386 allyesconfig build. This patch series thus does fix a regression. ] Signed-off-by: Feng Tang LKML-Reference: <1278342202-10973-2-git-send-email-feng.tang@intel.com> Acked-by: Dmitry Torokhov Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/x86_init.h | 2 ++ arch/x86/kernel/x86_init.c | 4 +++- 2 files changed, 5 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/x86_init.h b/arch/x86/include/asm/x86_init.h index 519b54327d75..baa579c8e038 100644 --- a/arch/x86/include/asm/x86_init.h +++ b/arch/x86/include/asm/x86_init.h @@ -142,6 +142,7 @@ struct x86_cpuinit_ops { * @set_wallclock: set time back to HW clock * @is_untracked_pat_range exclude from PAT logic * @nmi_init enable NMI on cpus + * @i8042_detect pre-detect if i8042 controller exists */ struct x86_platform_ops { unsigned long (*calibrate_tsc)(void); @@ -150,6 +151,7 @@ struct x86_platform_ops { void (*iommu_shutdown)(void); bool (*is_untracked_pat_range)(u64 start, u64 end); void (*nmi_init)(void); + int (*i8042_detect)(void); }; extern struct x86_init_ops x86_init; diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c index ebfb8e4c9f22..cd6da6bf3eca 100644 --- a/arch/x86/kernel/x86_init.c +++ b/arch/x86/kernel/x86_init.c @@ -86,6 +86,7 @@ struct x86_cpuinit_ops x86_cpuinit __cpuinitdata = { }; static void default_nmi_init(void) { }; +static int default_i8042_detect(void) { return 1; }; struct x86_platform_ops x86_platform = { .calibrate_tsc = native_calibrate_tsc, @@ -93,7 +94,8 @@ struct x86_platform_ops x86_platform = { .set_wallclock = mach_set_rtc_mmss, .iommu_shutdown = iommu_shutdown_noop, .is_untracked_pat_range = is_ISA_range, - .nmi_init = default_nmi_init + .nmi_init = default_nmi_init, + .i8042_detect = default_i8042_detect }; EXPORT_SYMBOL_GPL(x86_platform); -- cgit v1.2.3 From 6d2cce62017efe957e34cfcbba23861b7671980b Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Mon, 5 Jul 2010 23:03:19 +0800 Subject: x86, mrst: Add i8042_detect API for Moorestwon platform It will just return 0 as there is no i8042 controller Signed-off-by: Feng Tang LKML-Reference: <1278342202-10973-3-git-send-email-feng.tang@intel.com> Acked-by: Dmitry Torokhov Signed-off-by: H. Peter Anvin --- arch/x86/kernel/mrst.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/mrst.c b/arch/x86/kernel/mrst.c index e796448f0eb5..5915e0b33303 100644 --- a/arch/x86/kernel/mrst.c +++ b/arch/x86/kernel/mrst.c @@ -216,6 +216,12 @@ static void __init mrst_setup_boot_clock(void) setup_boot_APIC_clock(); }; +/* MID systems don't have i8042 controller */ +static int mrst_i8042_detect(void) +{ + return 0; +} + /* * Moorestown specific x86_init function overrides and early setup * calls. @@ -233,6 +239,7 @@ void __init x86_mrst_early_setup(void) x86_cpuinit.setup_percpu_clockev = mrst_setup_secondary_clock; x86_platform.calibrate_tsc = mrst_calibrate_tsc; + x86_platform.i8042_detect = mrst_i8042_detect; x86_init.pci.init = pci_mrst_init; x86_init.pci.fixup_irqs = x86_init_noop; -- cgit v1.2.3 From 91546356d0e550fa23abf7f4b04a903c2855761f Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Wed, 30 Jun 2010 16:04:06 +0800 Subject: KVM: MMU: flush remote tlbs when overwriting spte with different pfn After remove a rmap, we should flush all vcpu's tlb Signed-off-by: Xiao Guangrong Signed-off-by: Marcelo Tosatti --- arch/x86/kvm/mmu.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index a6f695d76928..3699613e8830 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -1879,6 +1879,8 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *sptep, pgprintk("hfn old %lx new %lx\n", spte_to_pfn(*sptep), pfn); rmap_remove(vcpu->kvm, sptep); + __set_spte(sptep, shadow_trap_nonpresent_pte); + kvm_flush_remote_tlbs(vcpu->kvm); } else was_rmapped = 1; } -- cgit v1.2.3 From 08be97962bf338161325d4901642f956ce8c1adb Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Wed, 14 Jul 2010 21:36:27 +0200 Subject: x86: Force HPET readback_cmp for all ATI chipsets commit 30a564be (x86, hpet: Restrict read back to affected ATI chipset) restricted the workaround for the HPET bug to SMX00 chipsets. This was reasonable as those were the only ones against which we ever got a bug report. Stephan Wolf reported now that this patch breaks his IXP400 based machine. Though it's confirmed to work on other IXP400 based systems. To error out on the safe side, we force the HPET readback workaround for all ATI SMbus class chipsets. Reported-by: Stephan Wolf LKML-Reference: Signed-off-by: Thomas Gleixner Tested-by: Stephan Wolf Acked-by: Borislav Petkov --- arch/x86/kernel/early-quirks.c | 18 ++++++++++++++++++ arch/x86/kernel/quirks.c | 5 ----- 2 files changed, 18 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c index ebdb85cf2686..e5cc7e82e60d 100644 --- a/arch/x86/kernel/early-quirks.c +++ b/arch/x86/kernel/early-quirks.c @@ -18,6 +18,7 @@ #include #include #include +#include static void __init fix_hypertransport_config(int num, int slot, int func) { @@ -191,6 +192,21 @@ static void __init ati_bugs_contd(int num, int slot, int func) } #endif +/* + * Force the read back of the CMP register in hpet_next_event() + * to work around the problem that the CMP register write seems to be + * delayed. See hpet_next_event() for details. + * + * We do this on all SMBUS incarnations for now until we have more + * information about the affected chipsets. + */ +static void __init ati_hpet_bugs(int num, int slot, int func) +{ +#ifdef CONFIG_HPET_TIMER + hpet_readback_cmp = 1; +#endif +} + #define QFLAG_APPLY_ONCE 0x1 #define QFLAG_APPLIED 0x2 #define QFLAG_DONE (QFLAG_APPLY_ONCE|QFLAG_APPLIED) @@ -220,6 +236,8 @@ static struct chipset early_qrk[] __initdata = { PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_bugs }, { PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_SBX00_SMBUS, PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_bugs_contd }, + { PCI_VENDOR_ID_ATI, PCI_ANY_ID, + PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_hpet_bugs }, {} }; diff --git a/arch/x86/kernel/quirks.c b/arch/x86/kernel/quirks.c index e72d3fc6547d..939b9e98245f 100644 --- a/arch/x86/kernel/quirks.c +++ b/arch/x86/kernel/quirks.c @@ -498,15 +498,10 @@ void force_hpet_resume(void) * See erratum #27 (Misinterpreted MSI Requests May Result in * Corrupted LPC DMA Data) in AMD Publication #46837, * "SB700 Family Product Errata", Rev. 1.0, March 2010. - * - * Also force the read back of the CMP register in hpet_next_event() - * to work around the problem that the CMP register write seems to be - * delayed. See hpet_next_event() for details. */ static void force_disable_hpet_msi(struct pci_dev *unused) { hpet_msi_disable = 1; - hpet_readback_cmp = 1; } DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_SBX00_SMBUS, -- cgit v1.2.3 From 58c84eda07560a6b75b03e8d3b26d6eddfc14011 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 15 Jul 2010 09:41:42 -0600 Subject: PCI: fall back to original BIOS BAR addresses If we fail to assign resources to a PCI BAR, this patch makes us try the original address from BIOS rather than leaving it disabled. Linux tries to make sure all PCI device BARs are inside the upstream PCI host bridge or P2P bridge apertures, reassigning BARs if necessary. Windows does similar reassignment. Before this patch, if we could not move a BAR into an aperture, we left the resource unassigned, i.e., at address zero. Windows leaves such BARs at the original BIOS addresses, and this patch makes Linux do the same. This is a bit ugly because we disable the resource long before we try to reassign it, so we have to keep track of the BIOS BAR address somewhere. For lack of a better place, I put it in the struct pci_dev. I think it would be cleaner to attempt the assignment immediately when the claim fails, so we could easily remember the original address. But we currently claim motherboard resources in the middle, after attempting to claim PCI resources and before assigning new PCI resources, and changing that is a fairly big job. Addresses https://bugzilla.kernel.org/show_bug.cgi?id=16263 Reported-by: Andrew Tested-by: Andrew Signed-off-by: Bjorn Helgaas Signed-off-by: Jesse Barnes --- arch/x86/pci/i386.c | 1 + drivers/pci/setup-res.c | 32 ++++++++++++++++++++++++++++++++ include/linux/pci.h | 1 + 3 files changed, 34 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c index 6fdb3ec30c31..55253095be84 100644 --- a/arch/x86/pci/i386.c +++ b/arch/x86/pci/i386.c @@ -184,6 +184,7 @@ static void __init pcibios_allocate_resources(int pass) idx, r, disabled, pass); if (pci_claim_resource(dev, idx) < 0) { /* We'll assign a new address later */ + dev->fw_addr[idx] = r->start; r->end -= r->start; r->start = 0; } diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c index 92379e2d37e7..2aaa13150de3 100644 --- a/drivers/pci/setup-res.c +++ b/drivers/pci/setup-res.c @@ -156,6 +156,38 @@ static int __pci_assign_resource(struct pci_bus *bus, struct pci_dev *dev, pcibios_align_resource, dev); } + if (ret < 0 && dev->fw_addr[resno]) { + struct resource *root, *conflict; + resource_size_t start, end; + + /* + * If we failed to assign anything, let's try the address + * where firmware left it. That at least has a chance of + * working, which is better than just leaving it disabled. + */ + + if (res->flags & IORESOURCE_IO) + root = &ioport_resource; + else + root = &iomem_resource; + + start = res->start; + end = res->end; + res->start = dev->fw_addr[resno]; + res->end = res->start + size - 1; + dev_info(&dev->dev, "BAR %d: trying firmware assignment %pR\n", + resno, res); + conflict = request_resource_conflict(root, res); + if (conflict) { + dev_info(&dev->dev, + "BAR %d: %pR conflicts with %s %pR\n", resno, + res, conflict->name, conflict); + res->start = start; + res->end = end; + } else + ret = 0; + } + if (!ret) { res->flags &= ~IORESOURCE_STARTALIGN; dev_info(&dev->dev, "BAR %d: assigned %pR\n", resno, res); diff --git a/include/linux/pci.h b/include/linux/pci.h index 7cb00845f150..f26fda76b87f 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -288,6 +288,7 @@ struct pci_dev { */ unsigned int irq; struct resource resource[DEVICE_COUNT_RESOURCE]; /* I/O and memory regions + expansion ROMs */ + resource_size_t fw_addr[DEVICE_COUNT_RESOURCE]; /* FW-assigned addr */ /* These fields are used by common fixups */ unsigned int transparent:1; /* Transparent PCI bridge */ -- cgit v1.2.3 From fd19dce7ac07973f700b0f13fb7f94b951414a4c Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Thu, 15 Jul 2010 00:00:59 -0700 Subject: x86: Fix x2apic preenabled system with kexec Found one x2apic system kexec loop test failed when CONFIG_NMI_WATCHDOG=y (old) or CONFIG_LOCKUP_DETECTOR=y (current tip) first kernel can kexec second kernel, but second kernel can not kexec third one. it can be duplicated on another system with BIOS preenabled x2apic. First kernel can not kexec second kernel. It turns out, when kernel boot with pre-enabled x2apic, it will not execute disable_local_APIC on shutdown path. when init_apic_mappings() is called in setup_arch, it will skip setting of apic_phys when x2apic_mode is set. ( x2apic_mode is much early check_x2apic()) Then later, disable_local_APIC() will bail out early because !apic_phys. So check !x2apic_mode in x2apic_mode in disable_local_APIC with !apic_phys. another solution could be updating init_apic_mappings() to set apic_phys even for preenabled x2apic system. Actually even for x2apic system, that lapic address is mapped already in early stage. BTW: is there any x2apic preenabled system with apicid of boot cpu > 255? Signed-off-by: Yinghai Lu LKML-Reference: <4C3EB22B.3000701@kernel.org> Acked-by: Suresh Siddha Cc: stable@kernel.org Signed-off-by: H. Peter Anvin --- arch/x86/kernel/apic/apic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index c02cc692985c..a96489ee6cab 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -921,7 +921,7 @@ void disable_local_APIC(void) unsigned int value; /* APIC hasn't been mapped yet */ - if (!apic_phys) + if (!x2apic_mode && !apic_phys) return; clear_local_APIC(); -- cgit v1.2.3 From f82c3d71d6fd2e6a3e3416f09099e29087e39abf Mon Sep 17 00:00:00 2001 From: Jacob Pan Date: Fri, 16 Jul 2010 11:58:26 -0700 Subject: x86, pci, mrst: Add extra sanity check in walking the PCI extended cap chain The fixed bar capability structure is searched in PCI extended configuration space. We need to make sure there is a valid capability ID to begin with otherwise, the search code may stuck in a infinite loop which results in boot hang. This patch adds additional check for cap ID 0, which is also invalid, and indicates end of chain. End of chain is supposed to have all fields zero, but that doesn't seem to always be the case in the field. Suggested-by: "H. Peter Anvin" Signed-off-by: Jacob Pan Reviewed-by: Jesse Barnes LKML-Reference: <1279306706-27087-1-git-send-email-jacob.jun.pan@linux.intel.com> Signed-off-by: H. Peter Anvin --- arch/x86/pci/mrst.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/pci/mrst.c b/arch/x86/pci/mrst.c index 7ef3a2735df3..cb29191cee58 100644 --- a/arch/x86/pci/mrst.c +++ b/arch/x86/pci/mrst.c @@ -66,8 +66,9 @@ static int fixed_bar_cap(struct pci_bus *bus, unsigned int devfn) devfn, pos, 4, &pcie_cap)) return 0; - if (pcie_cap == 0xffffffff) - return 0; + if (PCI_EXT_CAP_ID(pcie_cap) == 0x0000 || + PCI_EXT_CAP_ID(pcie_cap) == 0xffff) + break; if (PCI_EXT_CAP_ID(pcie_cap) == PCI_EXT_CAP_ID_VNDR) { raw_pci_ext_ops->read(pci_domain_nr(bus), bus->number, @@ -76,7 +77,7 @@ static int fixed_bar_cap(struct pci_bus *bus, unsigned int devfn) return pos; } - pos = pcie_cap >> 20; + pos = PCI_EXT_CAP_NEXT(pcie_cap); } return 0; -- cgit v1.2.3 From a197479848a2f1a2a5c07cffa6c31ab5e8c82797 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Fri, 16 Jul 2010 18:17:12 -0700 Subject: x86: kprobes: fix swapped segment registers in kretprobe In commit f007ea26, the order of the %es and %ds segment registers got accidentally swapped, so synthesized 'struct pt_regs' frames have the two values inverted. It's almost sure that these values never matter, and that they also never differ. But wrong is wrong. Signed-off-by: Roland McGrath --- arch/x86/kernel/kprobes.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index 345a4b1fe144..675879b65ce6 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -640,8 +640,8 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) /* Skip cs, ip, orig_ax and gs. */ \ " subl $16, %esp\n" \ " pushl %fs\n" \ - " pushl %ds\n" \ " pushl %es\n" \ + " pushl %ds\n" \ " pushl %eax\n" \ " pushl %ebp\n" \ " pushl %edi\n" \ -- cgit v1.2.3 From 7f8275d0d660c146de6ee3017e1e2e594c49e820 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 19 Jul 2010 14:56:17 +1000 Subject: mm: add context argument to shrinker callback The current shrinker implementation requires the registered callback to have global state to work from. This makes it difficult to shrink caches that are not global (e.g. per-filesystem caches). Pass the shrinker structure to the callback so that users can embed the shrinker structure in the context the shrinker needs to operate on and get back to it in the callback via container_of(). Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig --- arch/x86/kvm/mmu.c | 2 +- drivers/gpu/drm/i915/i915_gem.c | 2 +- fs/dcache.c | 2 +- fs/gfs2/glock.c | 2 +- fs/gfs2/quota.c | 2 +- fs/gfs2/quota.h | 2 +- fs/inode.c | 2 +- fs/mbcache.c | 5 +++-- fs/nfs/dir.c | 2 +- fs/nfs/internal.h | 3 ++- fs/quota/dquot.c | 2 +- fs/ubifs/shrinker.c | 2 +- fs/ubifs/ubifs.h | 2 +- fs/xfs/linux-2.6/xfs_buf.c | 5 +++-- fs/xfs/linux-2.6/xfs_sync.c | 1 + fs/xfs/quota/xfs_qm.c | 7 +++++-- include/linux/mm.h | 2 +- mm/vmscan.c | 8 +++++--- 18 files changed, 31 insertions(+), 22 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 3699613e8830..b1ed0a1a5913 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -2926,7 +2926,7 @@ static int kvm_mmu_remove_some_alloc_mmu_pages(struct kvm *kvm) return kvm_mmu_zap_page(kvm, page) + 1; } -static int mmu_shrink(int nr_to_scan, gfp_t gfp_mask) +static int mmu_shrink(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) { struct kvm *kvm; struct kvm *kvm_freed = NULL; diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 8757ecf6e96b..e7018708cc31 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -4978,7 +4978,7 @@ i915_gpu_is_active(struct drm_device *dev) } static int -i915_gem_shrink(int nr_to_scan, gfp_t gfp_mask) +i915_gem_shrink(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) { drm_i915_private_t *dev_priv, *next_dev; struct drm_i915_gem_object *obj_priv, *next_obj; diff --git a/fs/dcache.c b/fs/dcache.c index c8c78ba07827..86d4db15473e 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -896,7 +896,7 @@ EXPORT_SYMBOL(shrink_dcache_parent); * * In this case we return -1 to tell the caller that we baled. */ -static int shrink_dcache_memory(int nr, gfp_t gfp_mask) +static int shrink_dcache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) { if (nr) { if (!(gfp_mask & __GFP_FS)) diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index dbab3fdc2582..0898f3ec8212 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1358,7 +1358,7 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) } -static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) +static int gfs2_shrink_glock_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) { struct gfs2_glock *gl; int may_demote; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index b256d6f24288..8f02d3db8f42 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -77,7 +77,7 @@ static LIST_HEAD(qd_lru_list); static atomic_t qd_lru_count = ATOMIC_INIT(0); static DEFINE_SPINLOCK(qd_lru_lock); -int gfs2_shrink_qd_memory(int nr, gfp_t gfp_mask) +int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) { struct gfs2_quota_data *qd; struct gfs2_sbd *sdp; diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h index 195f60c8bd14..e7d236ca48bd 100644 --- a/fs/gfs2/quota.h +++ b/fs/gfs2/quota.h @@ -51,7 +51,7 @@ static inline int gfs2_quota_lock_check(struct gfs2_inode *ip) return ret; } -extern int gfs2_shrink_qd_memory(int nr, gfp_t gfp_mask); +extern int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask); extern const struct quotactl_ops gfs2_quotactl_ops; #endif /* __QUOTA_DOT_H__ */ diff --git a/fs/inode.c b/fs/inode.c index 2bee20ae3d65..722860b323a9 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -512,7 +512,7 @@ static void prune_icache(int nr_to_scan) * This function is passed the number of inodes to scan, and it returns the * total number of remaining possibly-reclaimable inodes. */ -static int shrink_icache_memory(int nr, gfp_t gfp_mask) +static int shrink_icache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) { if (nr) { /* diff --git a/fs/mbcache.c b/fs/mbcache.c index ec88ff3d04a9..e28f21b95344 100644 --- a/fs/mbcache.c +++ b/fs/mbcache.c @@ -115,7 +115,7 @@ mb_cache_indexes(struct mb_cache *cache) * What the mbcache registers as to get shrunk dynamically. */ -static int mb_cache_shrink_fn(int nr_to_scan, gfp_t gfp_mask); +static int mb_cache_shrink_fn(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask); static struct shrinker mb_cache_shrinker = { .shrink = mb_cache_shrink_fn, @@ -191,13 +191,14 @@ forget: * This function is called by the kernel memory management when memory * gets low. * + * @shrink: (ignored) * @nr_to_scan: Number of objects to scan * @gfp_mask: (ignored) * * Returns the number of objects which are present in the cache. */ static int -mb_cache_shrink_fn(int nr_to_scan, gfp_t gfp_mask) +mb_cache_shrink_fn(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) { LIST_HEAD(free_list); struct list_head *l, *ltmp; diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 782b431ef91c..e60416d3f818 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1710,7 +1710,7 @@ static void nfs_access_free_list(struct list_head *head) } } -int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask) +int nfs_access_cache_shrinker(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) { LIST_HEAD(head); struct nfs_inode *nfsi; diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index d8bd619e386c..e70f44b9b3f4 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -205,7 +205,8 @@ extern struct rpc_procinfo nfs4_procedures[]; void nfs_close_context(struct nfs_open_context *ctx, int is_sync); /* dir.c */ -extern int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask); +extern int nfs_access_cache_shrinker(struct shrinker *shrink, + int nr_to_scan, gfp_t gfp_mask); /* inode.c */ extern struct workqueue_struct *nfsiod_workqueue; diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 12c233da1b6b..437d2ca2de97 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -676,7 +676,7 @@ static void prune_dqcache(int count) * This is called from kswapd when we think we need some * more memory */ -static int shrink_dqcache_memory(int nr, gfp_t gfp_mask) +static int shrink_dqcache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) { if (nr) { spin_lock(&dq_list_lock); diff --git a/fs/ubifs/shrinker.c b/fs/ubifs/shrinker.c index 02feb59cefca..0b201114a5ad 100644 --- a/fs/ubifs/shrinker.c +++ b/fs/ubifs/shrinker.c @@ -277,7 +277,7 @@ static int kick_a_thread(void) return 0; } -int ubifs_shrinker(int nr, gfp_t gfp_mask) +int ubifs_shrinker(struct shrinker *shrink, int nr, gfp_t gfp_mask) { int freed, contention = 0; long clean_zn_cnt = atomic_long_read(&ubifs_clean_zn_cnt); diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 2eef553d50c8..04310878f449 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -1575,7 +1575,7 @@ int ubifs_tnc_start_commit(struct ubifs_info *c, struct ubifs_zbranch *zroot); int ubifs_tnc_end_commit(struct ubifs_info *c); /* shrinker.c */ -int ubifs_shrinker(int nr_to_scan, gfp_t gfp_mask); +int ubifs_shrinker(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask); /* commit.c */ int ubifs_bg_thread(void *info); diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 649ade8ef598..2ee3f7a60163 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -45,7 +45,7 @@ static kmem_zone_t *xfs_buf_zone; STATIC int xfsbufd(void *); -STATIC int xfsbufd_wakeup(int, gfp_t); +STATIC int xfsbufd_wakeup(struct shrinker *, int, gfp_t); STATIC void xfs_buf_delwri_queue(xfs_buf_t *, int); static struct shrinker xfs_buf_shake = { .shrink = xfsbufd_wakeup, @@ -340,7 +340,7 @@ _xfs_buf_lookup_pages( __func__, gfp_mask); XFS_STATS_INC(xb_page_retries); - xfsbufd_wakeup(0, gfp_mask); + xfsbufd_wakeup(NULL, 0, gfp_mask); congestion_wait(BLK_RW_ASYNC, HZ/50); goto retry; } @@ -1762,6 +1762,7 @@ xfs_buf_runall_queues( STATIC int xfsbufd_wakeup( + struct shrinker *shrink, int priority, gfp_t mask) { diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index ef7f0218bccb..be375827af98 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -838,6 +838,7 @@ static struct rw_semaphore xfs_mount_list_lock; static int xfs_reclaim_inode_shrink( + struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) { diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 8c117ff2e3ab..67c018392d62 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -69,7 +69,7 @@ STATIC void xfs_qm_list_destroy(xfs_dqlist_t *); STATIC int xfs_qm_init_quotainos(xfs_mount_t *); STATIC int xfs_qm_init_quotainfo(xfs_mount_t *); -STATIC int xfs_qm_shake(int, gfp_t); +STATIC int xfs_qm_shake(struct shrinker *, int, gfp_t); static struct shrinker xfs_qm_shaker = { .shrink = xfs_qm_shake, @@ -2117,7 +2117,10 @@ xfs_qm_shake_freelist( */ /* ARGSUSED */ STATIC int -xfs_qm_shake(int nr_to_scan, gfp_t gfp_mask) +xfs_qm_shake( + struct shrinker *shrink, + int nr_to_scan, + gfp_t gfp_mask) { int ndqused, nfree, n; diff --git a/include/linux/mm.h b/include/linux/mm.h index b969efb03787..a2b48041b910 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -999,7 +999,7 @@ static inline void sync_mm_rss(struct task_struct *task, struct mm_struct *mm) * querying the cache size, so a fastpath for that case is appropriate. */ struct shrinker { - int (*shrink)(int nr_to_scan, gfp_t gfp_mask); + int (*shrink)(struct shrinker *, int nr_to_scan, gfp_t gfp_mask); int seeks; /* seeks to recreate an obj */ /* These are for internal use */ diff --git a/mm/vmscan.c b/mm/vmscan.c index 9c7e57cc63a3..199fa436c0dd 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -213,8 +213,9 @@ unsigned long shrink_slab(unsigned long scanned, gfp_t gfp_mask, list_for_each_entry(shrinker, &shrinker_list, list) { unsigned long long delta; unsigned long total_scan; - unsigned long max_pass = (*shrinker->shrink)(0, gfp_mask); + unsigned long max_pass; + max_pass = (*shrinker->shrink)(shrinker, 0, gfp_mask); delta = (4 * scanned) / shrinker->seeks; delta *= max_pass; do_div(delta, lru_pages + 1); @@ -242,8 +243,9 @@ unsigned long shrink_slab(unsigned long scanned, gfp_t gfp_mask, int shrink_ret; int nr_before; - nr_before = (*shrinker->shrink)(0, gfp_mask); - shrink_ret = (*shrinker->shrink)(this_scan, gfp_mask); + nr_before = (*shrinker->shrink)(shrinker, 0, gfp_mask); + shrink_ret = (*shrinker->shrink)(shrinker, this_scan, + gfp_mask); if (shrink_ret == -1) break; if (shrink_ret < nr_before) -- cgit v1.2.3 From 087b255a2b43f417af83cb44e0bb02507f36b7fe Mon Sep 17 00:00:00 2001 From: Adam Lackorzynski Date: Tue, 20 Jul 2010 15:18:19 -0700 Subject: x86, i8259: Only register sysdev if we have a real 8259 PIC My platform makes use of the null_legacy_pic choice and oopses when doing a shutdown as the shutdown code goes through all the registered sysdevs and calls their shutdown method which in my case poke on a non-existing i8259. Imho the i8259 specific sysdev should only be registered if the i8259 is actually there. Do not register the sysdev function when the null_legacy_pic is used so that the i8259 resume, suspend and shutdown functions are not called. Signed-off-by: Adam Lackorzynski LKML-Reference: <201007202218.o6KMIJ3m020955@imap1.linux-foundation.org> Cc: Jacob Pan Cc: 2.6.34 Signed-off-by: Andrew Morton Signed-off-by: H. Peter Anvin --- arch/x86/kernel/i8259.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/i8259.c b/arch/x86/kernel/i8259.c index 7c9f02c130f3..cafa7c80ac95 100644 --- a/arch/x86/kernel/i8259.c +++ b/arch/x86/kernel/i8259.c @@ -276,16 +276,6 @@ static struct sys_device device_i8259A = { .cls = &i8259_sysdev_class, }; -static int __init i8259A_init_sysfs(void) -{ - int error = sysdev_class_register(&i8259_sysdev_class); - if (!error) - error = sysdev_register(&device_i8259A); - return error; -} - -device_initcall(i8259A_init_sysfs); - static void mask_8259A(void) { unsigned long flags; @@ -407,3 +397,18 @@ struct legacy_pic default_legacy_pic = { }; struct legacy_pic *legacy_pic = &default_legacy_pic; + +static int __init i8259A_init_sysfs(void) +{ + int error; + + if (legacy_pic != &default_legacy_pic) + return 0; + + error = sysdev_class_register(&i8259_sysdev_class); + if (!error) + error = sysdev_register(&device_i8259A); + return error; +} + +device_initcall(i8259A_init_sysfs); -- cgit v1.2.3 From 9aebbdb637a73a6092e1456ebb4a2df32cc1f611 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 20 Jul 2010 13:24:30 -0700 Subject: x86, numa: fix boot without RAM on node0 again Commit e534c7c5f8d6 ("numa: x86_64: use generic percpu var numa_node_id() implementation") broke numa systems that don't have ram on node0 when MEMORY_HOTPLUG is enabled, because cpu_up() will call cpu_to_node() before per_cpu(numa_node) is setup for APs. When Node0 doesn't have RAM, on x86, cpus already round it to nearest node with RAM in x86_cpu_to_node_map. and per_cpu(numa_node) is not set up until in c_init for APs. When later cpu_up() calling cpu_to_node() will get 0 again, and make it online even there is no RAM on node0. so later all APs can not booted up, and later will have panic. [ 1.611101] On node 0 totalpages: 0 ......... [ 2.608558] On node 0 totalpages: 0 [ 2.612065] Brought up 1 CPUs [ 2.615199] Total of 1 processors activated (3990.31 BogoMIPS). ... 93.225341] calling loop_init+0x0/0x1a4 @ 1 [ 93.229314] PERCPU: allocation failed, size=80 align=8, failed to populate [ 93.246539] Pid: 1, comm: swapper Tainted: G W 2.6.35-rc4-tip-yh-04371-gd64e6c4-dirty #354 [ 93.264621] Call Trace: [ 93.266533] [] pcpu_alloc+0x83a/0x8e7 [ 93.270710] [] __alloc_percpu+0x10/0x12 [ 93.285849] [] alloc_disk_node+0x94/0x16d [ 93.291811] [] alloc_disk+0x11/0x13 [ 93.306157] [] loop_alloc+0xa7/0x180 [ 93.310538] [] loop_init+0x9b/0x1a4 [ 93.324909] [] ? loop_init+0x0/0x1a4 [ 93.329650] [] do_one_initcall+0x57/0x136 [ 93.345197] [] kernel_init+0x184/0x20e [ 93.348146] [] kernel_thread_helper+0x4/0x10 [ 93.365194] [] ? restore_args+0x0/0x30 [ 93.369305] [] ? kernel_init+0x0/0x20e [ 93.386011] [] ? kernel_thread_helper+0x0/0x10 [ 93.392047] loop: out of memory ... Try to assign per_cpu(numa_node) early [akpm@linux-foundation.org: tidy up code comment] Signed-off-by: Yinghai Cc: Thomas Gleixner Cc: Ingo Molnar Cc: "H. Peter Anvin" Cc: Tejun Heo Cc: Denys Vlasenko Acked-by: Lee Schermerhorn Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86/kernel/setup_percpu.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index de3b63ae3da2..690c2c09faf3 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -238,6 +238,15 @@ void __init setup_per_cpu_areas(void) #ifdef CONFIG_NUMA per_cpu(x86_cpu_to_node_map, cpu) = early_per_cpu_map(x86_cpu_to_node_map, cpu); + /* + * Ensure taht the boot cpu numa_node is correct when the boot + * cpu is on a node that doesn't have memory installed. + * Also cpu_up() will call cpu_to_node() for APs when + * MEMORY_HOTPLUG is defined, before per_cpu(numa_node) is set + * up later with c_init aka intel_init/amd_init. + * So set them all (boot cpu and all APs). + */ + set_cpu_numa_node(cpu, early_cpu_to_node(cpu)); #endif #endif /* @@ -257,14 +266,6 @@ void __init setup_per_cpu_areas(void) early_per_cpu_ptr(x86_cpu_to_node_map) = NULL; #endif -#if defined(CONFIG_X86_64) && defined(CONFIG_NUMA) - /* - * make sure boot cpu numa_node is right, when boot cpu is on the - * node that doesn't have mem installed - */ - set_cpu_numa_node(boot_cpu_id, early_cpu_to_node(boot_cpu_id)); -#endif - /* Setup node to cpumask map */ setup_node_to_cpumask_map(); -- cgit v1.2.3 From a4ce96ac356e7024a7724ade9d18ba1bdf3c5c06 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 21 Jul 2010 09:25:42 -0700 Subject: Fix up trivial spelling errors ('taht' -> 'that') Pointed out by Lucas who found the new one in a comment in setup_percpu.c. And then I fixed the others that I grepped for. Reported-by: Lucas Signed-off-by: Linus Torvalds --- arch/x86/kernel/setup_percpu.c | 2 +- drivers/usb/gadget/f_fs.c | 2 +- drivers/video/aty/radeon_pm.c | 2 +- fs/jffs2/xattr.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 690c2c09faf3..a60df9ae6454 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -239,7 +239,7 @@ void __init setup_per_cpu_areas(void) per_cpu(x86_cpu_to_node_map, cpu) = early_per_cpu_map(x86_cpu_to_node_map, cpu); /* - * Ensure taht the boot cpu numa_node is correct when the boot + * Ensure that the boot cpu numa_node is correct when the boot * cpu is on a node that doesn't have memory installed. * Also cpu_up() will call cpu_to_node() for APs when * MEMORY_HOTPLUG is defined, before per_cpu(numa_node) is set diff --git a/drivers/usb/gadget/f_fs.c b/drivers/usb/gadget/f_fs.c index d69eccf5f197..2aaa0f75c6cf 100644 --- a/drivers/usb/gadget/f_fs.c +++ b/drivers/usb/gadget/f_fs.c @@ -136,7 +136,7 @@ struct ffs_data { * handling setup requests immidiatelly user space may be so * slow that another setup will be sent to the gadget but this * time not to us but another function and then there could be - * a race. Is taht the case? Or maybe we can use cdev->req + * a race. Is that the case? Or maybe we can use cdev->req * after all, maybe we just need some spinlock for that? */ struct usb_request *ep0req; /* P: mutex */ struct completion ep0req_completion; /* P: mutex */ diff --git a/drivers/video/aty/radeon_pm.c b/drivers/video/aty/radeon_pm.c index 515cf1978d19..c4e17642d9c5 100644 --- a/drivers/video/aty/radeon_pm.c +++ b/drivers/video/aty/radeon_pm.c @@ -2872,7 +2872,7 @@ void radeonfb_pm_init(struct radeonfb_info *rinfo, int dynclk, int ignore_devlis } #if 0 - /* Power down TV DAC, taht saves a significant amount of power, + /* Power down TV DAC, that saves a significant amount of power, * we'll have something better once we actually have some TVOut * support */ diff --git a/fs/jffs2/xattr.c b/fs/jffs2/xattr.c index a2d58c96f1b4..d258e261bdc7 100644 --- a/fs/jffs2/xattr.c +++ b/fs/jffs2/xattr.c @@ -626,7 +626,7 @@ void jffs2_xattr_free_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *i static int check_xattr_ref_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *ic) { - /* success of check_xattr_ref_inode() means taht inode (ic) dose not have + /* success of check_xattr_ref_inode() means that inode (ic) dose not have * duplicate name/value pairs. If duplicate name/value pair would be found, * one will be removed. */ -- cgit v1.2.3 From 0327559151c6886814d6d5b373b4bf6de63fb9f6 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Wed, 21 Jul 2010 17:44:12 -0700 Subject: x86: auditsyscall: fix fastpath return value after reschedule In the CONFIG_AUDITSYSCALL fast-path for x86 64-bit system calls, we can pass a bad return value and/or error indication for the system call to audit_syscall_exit(). This happens when TIF_NEED_RESCHED was set as the system call returned, so we went out to schedule() and came back to the exit-audit fast-path. The fix is to reload the user return value register from the pt_regs before using it for audit_syscall_exit(). Both the 32-bit kernel's fast path and the 64-bit kernel's 32-bit system call fast paths work slightly differently, so that they always leave the fast path entirely to reschedule and don't return there, so they don't have the analogous bugs. Reported-by: Alexander Viro Signed-off-by: Roland McGrath --- arch/x86/kernel/entry_64.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 0697ff139837..4db7c4d12ffa 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -571,8 +571,8 @@ auditsys: * masked off. */ sysret_audit: - movq %rax,%rsi /* second arg, syscall return value */ - cmpq $0,%rax /* is it < 0? */ + movq RAX-ARGOFFSET(%rsp),%rsi /* second arg, syscall return value */ + cmpq $0,%rsi /* is it < 0? */ setl %al /* 1 if so, 0 if not */ movzbl %al,%edi /* zero-extend that into %edi */ inc %edi /* first arg, 0->1(AUDITSC_SUCCESS), 1->2(AUDITSC_FAILURE) */ -- cgit v1.2.3 From 718be4aaf3613cf7c2d097f925abc3d3553c0605 Mon Sep 17 00:00:00 2001 From: Len Brown Date: Thu, 22 Jul 2010 16:54:27 -0400 Subject: ACPI: skip checking BM_STS if the BIOS doesn't ask for it It turns out that there is a bit in the _CST for Intel FFH C3 that tells the OS if we should be checking BM_STS or not. Linux has been unconditionally checking BM_STS. If the chip-set is configured to enable BM_STS, it can retard or completely prevent entry into deep C-states -- as illustrated by turbostat: http://userweb.kernel.org/~lenb/acpi/utils/pmtools/turbostat/ ref: Intel Processor Vendor-Specific ACPI Interface Specification table 4 "_CST FFH GAS Field Encoding" Bit 1: Set to 1 if OSPM should use Bus Master avoidance for this C-state https://bugzilla.kernel.org/show_bug.cgi?id=15886 Signed-off-by: Len Brown --- arch/x86/kernel/acpi/cstate.c | 9 +++++++++ drivers/acpi/processor_idle.c | 2 +- include/acpi/processor.h | 3 ++- 3 files changed, 12 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/acpi/cstate.c b/arch/x86/kernel/acpi/cstate.c index 2e837f5080fe..fb7a5f052e2b 100644 --- a/arch/x86/kernel/acpi/cstate.c +++ b/arch/x86/kernel/acpi/cstate.c @@ -145,6 +145,15 @@ int acpi_processor_ffh_cstate_probe(unsigned int cpu, percpu_entry->states[cx->index].eax = cx->address; percpu_entry->states[cx->index].ecx = MWAIT_ECX_INTERRUPT_BREAK; } + + /* + * For _CST FFH on Intel, if GAS.access_size bit 1 is cleared, + * then we should skip checking BM_STS for this C-state. + * ref: "Intel Processor Vendor-Specific ACPI Interface Specification" + */ + if ((c->x86_vendor == X86_VENDOR_INTEL) && !(reg->access_size & 0x2)) + cx->bm_sts_skip = 1; + return retval; } EXPORT_SYMBOL_GPL(acpi_processor_ffh_cstate_probe); diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index b1b385692f46..b351342f1faf 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -947,7 +947,7 @@ static int acpi_idle_enter_bm(struct cpuidle_device *dev, if (acpi_idle_suspend) return(acpi_idle_enter_c1(dev, state)); - if (acpi_idle_bm_check()) { + if (!cx->bm_sts_skip && acpi_idle_bm_check()) { if (dev->safe_state) { dev->last_state = dev->safe_state; return dev->safe_state->enter(dev, dev->safe_state); diff --git a/include/acpi/processor.h b/include/acpi/processor.h index da565a48240e..a68ca8a11a53 100644 --- a/include/acpi/processor.h +++ b/include/acpi/processor.h @@ -48,7 +48,7 @@ struct acpi_power_register { u8 space_id; u8 bit_width; u8 bit_offset; - u8 reserved; + u8 access_size; u64 address; } __attribute__ ((packed)); @@ -63,6 +63,7 @@ struct acpi_processor_cx { u32 power; u32 usage; u64 time; + u8 bm_sts_skip; char desc[ACPI_CX_DESC_LEN]; }; -- cgit v1.2.3 From 6aa0b9dec5d6dde26ea17b0b5be8fccfe19df3c9 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Wed, 30 Jun 2010 16:02:45 +0800 Subject: KVM: MMU: fix conflict access permissions in direct sp In no-direct mapping, we mark sp is 'direct' when we mapping the guest's larger page, but its access is encoded form upper page-struct entire not include the last mapping, it will cause access conflict. For example, have this mapping: [W] / PDE1 -> |---| P[W] | | LPA \ PDE2 -> |---| [R] P have two children, PDE1 and PDE2, both PDE1 and PDE2 mapping the same lage page(LPA). The P's access is WR, PDE1's access is WR, PDE2's access is RO(just consider read-write permissions here) When guest access PDE1, we will create a direct sp for LPA, the sp's access is from P, is W, then we will mark the ptes is W in this sp. Then, guest access PDE2, we will find LPA's shadow page, is the same as PDE's, and mark the ptes is RO. So, if guest access PDE1, the incorrect #PF is occured. Fixed by encode the last mapping access into direct shadow page Signed-off-by: Xiao Guangrong Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- arch/x86/kvm/paging_tmpl.h | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h index 89d66ca4d87c..2331bdc2b549 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -342,6 +342,7 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr, /* advance table_gfn when emulating 1gb pages with 4k */ if (delta == 0) table_gfn += PT_INDEX(addr, level); + access &= gw->pte_access; } else { direct = 0; table_gfn = gw->table_gfn[level - 2]; -- cgit v1.2.3 From 7a73c0283dadf1cf360a79de396ff0962e781b60 Mon Sep 17 00:00:00 2001 From: Avi Kivity Date: Thu, 22 Jul 2010 23:24:52 +0300 Subject: KVM: Use kmalloc() instead of vmalloc() for KVM_[GS]ET_MSR We don't need more than a page, and vmalloc() is slower (much slower recently due to a regression). Signed-off-by: Avi Kivity --- arch/x86/kvm/x86.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 05d571f6f196..7fa89c39c64f 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1562,7 +1562,7 @@ static int msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs __user *user_msrs, r = -ENOMEM; size = sizeof(struct kvm_msr_entry) * msrs.nmsrs; - entries = vmalloc(size); + entries = kmalloc(size, GFP_KERNEL); if (!entries) goto out; @@ -1581,7 +1581,7 @@ static int msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs __user *user_msrs, r = n; out_free: - vfree(entries); + kfree(entries); out: return r; } -- cgit v1.2.3 From ff4878089e1eaeac79d57878ad4ea32910fb4037 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Wed, 21 Jul 2010 18:32:37 +0100 Subject: x86: Do not try to disable hpet if it hasn't been initialized before hpet_disable is called unconditionally on machine reboot if hpet support is compiled in the kernel. hpet_disable only checks if the machine is hpet capable but doesn't make sure that hpet has been initialized. [ tglx: Made it a one liner and removed the redundant hpet_address check ] Signed-off-by: Stefano Stabellini Acked-by: Venkatesh Pallipadi LKML-Reference: Cc: stable@kernel.org Signed-off-by: Thomas Gleixner --- arch/x86/kernel/hpet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index a198b7c87a12..ba390d731175 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -964,7 +964,7 @@ fs_initcall(hpet_late_init); void hpet_disable(void) { - if (is_hpet_capable()) { + if (is_hpet_capable() && hpet_virt_address) { unsigned int cfg = hpet_readl(HPET_CFG); if (hpet_legacy_int_enabled) { -- cgit v1.2.3 From 72ad5d77fb981963edae15eee8196c80238f5ed0 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 23 Jul 2010 22:59:09 +0200 Subject: ACPI / Sleep: Allow the NVS saving to be skipped during suspend to RAM Commit 2a6b69765ad794389f2fc3e14a0afa1a995221c2 (ACPI: Store NVS state even when entering suspend to RAM) caused the ACPI suspend code save the NVS area during suspend and restore it during resume unconditionally, although it is known that some systems need to use acpi_sleep=s4_nonvs for hibernation to work. To allow the affected systems to avoid saving and restoring the NVS area during suspend to RAM and resume, introduce kernel command line option acpi_sleep=nonvs and make acpi_sleep=s4_nonvs work as its alias temporarily (add acpi_sleep=s4_nonvs to the feature removal file). Addresses https://bugzilla.kernel.org/show_bug.cgi?id=16396 . Signed-off-by: Rafael J. Wysocki Reported-and-tested-by: tomas m Signed-off-by: Len Brown --- Documentation/feature-removal-schedule.txt | 7 ++++++ Documentation/kernel-parameters.txt | 4 ++-- arch/x86/kernel/acpi/sleep.c | 9 ++++++-- drivers/acpi/sleep.c | 35 +++++++++++++++--------------- include/linux/acpi.h | 2 +- 5 files changed, 34 insertions(+), 23 deletions(-) (limited to 'arch/x86') diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index c268783bc4e7..1571c0c83dba 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -647,3 +647,10 @@ Who: Stefan Richter ---------------------------- +What: The acpi_sleep=s4_nonvs command line option +When: 2.6.37 +Files: arch/x86/kernel/acpi/sleep.c +Why: superseded by acpi_sleep=nonvs +Who: Rafael J. Wysocki + +---------------------------- diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 4ddb58df081e..2b2407d9a6d0 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -254,8 +254,8 @@ and is between 256 and 4096 characters. It is defined in the file control method, with respect to putting devices into low power states, to be enforced (the ACPI 2.0 ordering of _PTS is used by default). - s4_nonvs prevents the kernel from saving/restoring the - ACPI NVS memory during hibernation. + nonvs prevents the kernel from saving/restoring the + ACPI NVS memory during suspend/hibernation and resume. sci_force_enable causes the kernel to set SCI_EN directly on resume from S1/S3 (which is against the ACPI spec, but some broken systems don't work without it). diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index 82e508677b91..fcc3c61fdecc 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -157,9 +157,14 @@ static int __init acpi_sleep_setup(char *str) #ifdef CONFIG_HIBERNATION if (strncmp(str, "s4_nohwsig", 10) == 0) acpi_no_s4_hw_signature(); - if (strncmp(str, "s4_nonvs", 8) == 0) - acpi_s4_no_nvs(); + if (strncmp(str, "s4_nonvs", 8) == 0) { + pr_warning("ACPI: acpi_sleep=s4_nonvs is deprecated, " + "please use acpi_sleep=nonvs instead"); + acpi_nvs_nosave(); + } #endif + if (strncmp(str, "nonvs", 5) == 0) + acpi_nvs_nosave(); if (strncmp(str, "old_ordering", 12) == 0) acpi_old_suspend_ordering(); str = strchr(str, ','); diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index 5b7c52e4a00f..2862c781b372 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -81,6 +81,20 @@ static int acpi_sleep_prepare(u32 acpi_state) #ifdef CONFIG_ACPI_SLEEP static u32 acpi_target_sleep_state = ACPI_STATE_S0; +/* + * The ACPI specification wants us to save NVS memory regions during hibernation + * and to restore them during the subsequent resume. Windows does that also for + * suspend to RAM. However, it is known that this mechanism does not work on + * all machines, so we allow the user to disable it with the help of the + * 'acpi_sleep=nonvs' kernel command line option. + */ +static bool nvs_nosave; + +void __init acpi_nvs_nosave(void) +{ + nvs_nosave = true; +} + /* * ACPI 1.0 wants us to execute _PTS before suspending devices, so we allow the * user to request that behavior by using the 'acpi_old_suspend_ordering' @@ -197,8 +211,7 @@ static int acpi_suspend_begin(suspend_state_t pm_state) u32 acpi_state = acpi_suspend_states[pm_state]; int error = 0; - error = suspend_nvs_alloc(); - + error = nvs_nosave ? 0 : suspend_nvs_alloc(); if (error) return error; @@ -388,20 +401,6 @@ static struct dmi_system_id __initdata acpisleep_dmi_table[] = { #endif /* CONFIG_SUSPEND */ #ifdef CONFIG_HIBERNATION -/* - * The ACPI specification wants us to save NVS memory regions during hibernation - * and to restore them during the subsequent resume. However, it is not certain - * if this mechanism is going to work on all machines, so we allow the user to - * disable this mechanism using the 'acpi_sleep=s4_nonvs' kernel command line - * option. - */ -static bool s4_no_nvs; - -void __init acpi_s4_no_nvs(void) -{ - s4_no_nvs = true; -} - static unsigned long s4_hardware_signature; static struct acpi_table_facs *facs; static bool nosigcheck; @@ -415,7 +414,7 @@ static int acpi_hibernation_begin(void) { int error; - error = s4_no_nvs ? 0 : suspend_nvs_alloc(); + error = nvs_nosave ? 0 : suspend_nvs_alloc(); if (!error) { acpi_target_sleep_state = ACPI_STATE_S4; acpi_sleep_tts_switch(acpi_target_sleep_state); @@ -510,7 +509,7 @@ static int acpi_hibernation_begin_old(void) error = acpi_sleep_prepare(ACPI_STATE_S4); if (!error) { - if (!s4_no_nvs) + if (!nvs_nosave) error = suspend_nvs_alloc(); if (!error) acpi_target_sleep_state = ACPI_STATE_S4; diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 224a38c960d4..ccf94dc5acdf 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -253,7 +253,7 @@ int acpi_resources_are_enforced(void); #ifdef CONFIG_PM_SLEEP void __init acpi_no_s4_hw_signature(void); void __init acpi_old_suspend_ordering(void); -void __init acpi_s4_no_nvs(void); +void __init acpi_nvs_nosave(void); #endif /* CONFIG_PM_SLEEP */ struct acpi_osc_context { -- cgit v1.2.3 From 47f8bcf362410b631a4d99ff5c79ec6b9dd3ace6 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Tue, 20 Jul 2010 13:52:00 -0400 Subject: [CPUFREQ] pcc driver should check for pcch method before calling _OSC The pcc specification documents an _OSC method that's incompatible with the one defined as part of the ACPI spec. This shouldn't be a problem as both are supposed to be guarded with a UUID. Unfortunately approximately nobody (including HP, who wrote this spec) properly check the UUID on entry to the _OSC call. Right now this could result in surprising behaviour if the pcc driver performs an _OSC call on a machine that doesn't implement the pcc specification. Check whether the PCCH method exists first in order to reduce this probability. Signed-off-by: Matthew Garrett Cc: Naga Chumbalkar Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c index ce7cde713e71..01bd25c3c7ca 100644 --- a/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c @@ -397,13 +397,17 @@ static int __init pcc_cpufreq_probe(void) struct pcc_memory_resource *mem_resource; struct pcc_register_resource *reg_resource; union acpi_object *out_obj, *member; - acpi_handle handle, osc_handle; + acpi_handle handle, osc_handle, pcch_handle; int ret = 0; status = acpi_get_handle(NULL, "\\_SB", &handle); if (ACPI_FAILURE(status)) return -ENODEV; + status = acpi_get_handle(handle, "PCCH", &pcch_handle); + if (ACPI_FAILURE(status)) + return -ENODEV; + status = acpi_get_handle(handle, "_OSC", &osc_handle); if (ACPI_SUCCESS(status)) { ret = pcc_cpufreq_do_osc(&osc_handle); -- cgit v1.2.3 From 3847d223f2e4da5ceb47ea8996618010192f3197 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Fri, 23 Jul 2010 23:06:52 +0100 Subject: [CPUFREQ] fix double freeing in error path of pcc-cpufreq Prevent double freeing on error path. Signed-off-by: Daniel J Blueman Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c index 01bd25c3c7ca..900702888bfb 100644 --- a/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c @@ -368,22 +368,16 @@ static int __init pcc_cpufreq_do_osc(acpi_handle *handle) return -ENODEV; out_obj = output.pointer; - if (out_obj->type != ACPI_TYPE_BUFFER) { - ret = -ENODEV; - goto out_free; - } + if (out_obj->type != ACPI_TYPE_BUFFER) + return -ENODEV; errors = *((u32 *)out_obj->buffer.pointer) & ~(1 << 0); - if (errors) { - ret = -ENODEV; - goto out_free; - } + if (errors) + return -ENODEV; supported = *((u32 *)(out_obj->buffer.pointer + 4)); - if (!(supported & 0x1)) { - ret = -ENODEV; - goto out_free; - } + if (!(supported & 0x1)) + return -ENODEV; out_free: kfree(output.pointer); -- cgit v1.2.3 From 179ee43465343d1f8f2a4af25ead4ae15e43fa6e Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Thu, 15 Jul 2010 11:44:00 -0400 Subject: [CPUFREQ] Fix PCC driver error path The PCC cpufreq driver unmaps the mailbox address range if any CPUs fail to initialise, but doesn't do anything to remove the registered CPUs from the cpufreq core resulting in failures further down the line. We're better off simply returning a failure - the cpufreq core will unregister us cleanly if we end up with no successfully registered CPUs. Tidy up the failure path and also add a sanity check to ensure that the firmware gives us a realistic frequency - the core deals badly with that being set to 0. Signed-off-by: Matthew Garrett Cc: Naga Chumbalkar Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c index 900702888bfb..a36de5bbb622 100644 --- a/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c @@ -541,13 +541,13 @@ static int pcc_cpufreq_cpu_init(struct cpufreq_policy *policy) if (!pcch_virt_addr) { result = -1; - goto pcch_null; + goto out; } result = pcc_get_offset(cpu); if (result) { dprintk("init: PCCP evaluation failed\n"); - goto free; + goto out; } policy->max = policy->cpuinfo.max_freq = @@ -556,14 +556,15 @@ static int pcc_cpufreq_cpu_init(struct cpufreq_policy *policy) ioread32(&pcch_hdr->minimum_frequency) * 1000; policy->cur = pcc_get_freq(cpu); + if (!policy->cur) { + dprintk("init: Unable to get current CPU frequency\n"); + result = -EINVAL; + goto out; + } + dprintk("init: policy->max is %d, policy->min is %d\n", policy->max, policy->min); - - return 0; -free: - pcc_clear_mapping(); - free_percpu(pcc_cpu_info); -pcch_null: +out: return result; } -- cgit v1.2.3 From 3581ced3b6ac289b5cd31663b34914a7347186a6 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Thu, 8 Jul 2010 17:55:30 +0200 Subject: [CPUFREQ] powernow-k8: Limit Pstate transition latency check The Pstate transition latency check was added for broken F10h BIOSen which wrongly contain a value of 0 for transition and bus master latency. Fam11h and later, however, (will) have similar transition latency so extend that behavior for them too. Signed-off-by: Borislav Petkov Signed-off-by: Dave Jones --- arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index 7ec2123838e6..3e90cce3dc8b 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -1023,13 +1023,12 @@ static int get_transition_latency(struct powernow_k8_data *data) } if (max_latency == 0) { /* - * Fam 11h always returns 0 as transition latency. - * This is intended and means "very fast". While cpufreq core - * and governors currently can handle that gracefully, better - * set it to 1 to avoid problems in the future. - * For all others it's a BIOS bug. + * Fam 11h and later may return 0 as transition latency. This + * is intended and means "very fast". While cpufreq core and + * governors currently can handle that gracefully, better set it + * to 1 to avoid problems in the future. */ - if (boot_cpu_data.x86 != 0x11) + if (boot_cpu_data.x86 < 0x11) printk(KERN_ERR FW_WARN PFX "Invalid zero transition " "latency\n"); max_latency = 1; -- cgit v1.2.3 From ba773f7c510c0b252145933926c636c439889207 Mon Sep 17 00:00:00 2001 From: Jason Wessel Date: Wed, 28 Jul 2010 19:10:30 -0500 Subject: x86,kgdb: Fix hw breakpoint regression HW breakpoints events stopped working correctly with kgdb as a result of commit: 018cbffe6819f6f8db20a0a3acd9bab9bfd667e4 (Merge commit 'v2.6.33' into perf/core). The regression occurred because the behavior changed for setting NOTIFY_STOP as the return value to the die notifier if the breakpoint was known to the HW breakpoint API. Because kgdb is using the HW breakpoint API to register HW breakpoints slots, it must also now implement the overflow_handler call back else kgdb does not get to see the events from the die notifier. The kgdb_ll_trap function will be changed to be general purpose code which can allow an easy way to implement the hw_breakpoint API overflow call back. Signed-off-by: Jason Wessel Acked-by: Dongdong Deng Acked-by: Frederic Weisbecker --- arch/x86/kernel/kgdb.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c index 4f4af75b9482..01ab17ae2ae7 100644 --- a/arch/x86/kernel/kgdb.c +++ b/arch/x86/kernel/kgdb.c @@ -572,7 +572,6 @@ static int __kgdb_notify(struct die_args *args, unsigned long cmd) return NOTIFY_STOP; } -#ifdef CONFIG_KGDB_LOW_LEVEL_TRAP int kgdb_ll_trap(int cmd, const char *str, struct pt_regs *regs, long err, int trap, int sig) { @@ -590,7 +589,6 @@ int kgdb_ll_trap(int cmd, const char *str, return __kgdb_notify(&args, cmd); } -#endif /* CONFIG_KGDB_LOW_LEVEL_TRAP */ static int kgdb_notify(struct notifier_block *self, unsigned long cmd, void *ptr) @@ -625,6 +623,12 @@ int kgdb_arch_init(void) return register_die_notifier(&kgdb_notifier); } +static void kgdb_hw_overflow_handler(struct perf_event *event, int nmi, + struct perf_sample_data *data, struct pt_regs *regs) +{ + kgdb_ll_trap(DIE_DEBUG, "debug", regs, 0, 0, SIGTRAP); +} + void kgdb_arch_late(void) { int i, cpu; @@ -655,6 +659,7 @@ void kgdb_arch_late(void) for_each_online_cpu(cpu) { pevent = per_cpu_ptr(breakinfo[i].pev, cpu); pevent[0]->hw.sample_period = 1; + pevent[0]->overflow_handler = kgdb_hw_overflow_handler; if (pevent[0]->destroy != NULL) { pevent[0]->destroy = NULL; release_bp_slot(*pevent); -- cgit v1.2.3