diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2009-03-30 11:38:31 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2009-03-30 11:38:31 -0700 |
| commit | 019abbc87025a030fd25008612afd4eff8a375f7 (patch) | |
| tree | 6d745dedcf90ceff8f5b7b996a17f666b7c574e3 /arch/x86/kernel | |
| parent | 2d25ee36c84d5b2d6be8bfaf80256ecad69a06ca (diff) | |
| parent | 5a3c8fe7353f78b73b9636353c6f7b881f19ebea (diff) | |
| download | lwn-019abbc87025a030fd25008612afd4eff8a375f7.tar.gz lwn-019abbc87025a030fd25008612afd4eff8a375f7.zip | |
Merge branch 'x86-stage-3-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'x86-stage-3-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (190 commits)
Revert "cpuacct: reduce one NULL check in fast-path"
Revert "x86: don't compile vsmp_64 for 32bit"
x86: Correct behaviour of irq affinity
x86: early_ioremap_init(), use __fix_to_virt(), because we are sure it's safe
x86: use default_cpu_mask_to_apicid for 64bit
x86: fix set_extra_move_desc calling
x86, PAT, PCI: Change vma prot in pci_mmap to reflect inherited prot
x86/dmi: fix dmi_alloc() section mismatches
x86: e820 fix various signedness issues in setup.c and e820.c
x86: apic/io_apic.c define msi_ir_chip and ir_ioapic_chip all the time
x86: irq.c keep CONFIG_X86_LOCAL_APIC interrupts together
x86: irq.c use same path for show_interrupts
x86: cpu/cpu.h cleanup
x86: Fix a couple of sparse warnings in arch/x86/kernel/apic/io_apic.c
Revert "x86: create a non-zero sized bm_pte only when needed"
x86: pci-nommu.c cleanup
x86: io_delay.c cleanup
x86: rtc.c cleanup
x86: i8253 cleanup
x86: kdebugfs.c cleanup
...
Diffstat (limited to 'arch/x86/kernel')
71 files changed, 4735 insertions, 2447 deletions
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 95f216bbfaf1..6e9c1f320acf 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -70,7 +70,6 @@ obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o obj-$(CONFIG_KEXEC) += machine_kexec_$(BITS).o obj-$(CONFIG_KEXEC) += relocate_kernel_$(BITS).o crash.o obj-$(CONFIG_CRASH_DUMP) += crash_dump_$(BITS).o -obj-$(CONFIG_X86_VSMP) += vsmp_64.o obj-$(CONFIG_KPROBES) += kprobes.o obj-$(CONFIG_MODULES) += module_$(BITS).o obj-$(CONFIG_EFI) += efi.o efi_$(BITS).o efi_stub_$(BITS).o @@ -111,7 +110,7 @@ obj-$(CONFIG_SWIOTLB) += pci-swiotlb_64.o # NB rename without _64 ### # 64 bit specific files ifeq ($(CONFIG_X86_64),y) - obj-$(CONFIG_X86_UV) += tlb_uv.o bios_uv.o uv_irq.o uv_sysfs.o + obj-$(CONFIG_X86_UV) += tlb_uv.o bios_uv.o uv_irq.o uv_sysfs.o uv_time.o obj-$(CONFIG_X86_PM_TIMER) += pmtimer_64.o obj-$(CONFIG_AUDIT) += audit_64.o @@ -120,4 +119,5 @@ ifeq ($(CONFIG_X86_64),y) obj-$(CONFIG_AMD_IOMMU) += amd_iommu_init.o amd_iommu.o obj-$(CONFIG_PCI_MMCONFIG) += mmconf-fam10h_64.o + obj-y += vsmp_64.o endif diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 6907b8e85d52..4c80f1557433 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -414,9 +414,17 @@ void __init alternative_instructions(void) that might execute the to be patched code. Other CPUs are not running. */ stop_nmi(); -#ifdef CONFIG_X86_MCE - stop_mce(); -#endif + + /* + * Don't stop machine check exceptions while patching. + * MCEs only happen when something got corrupted and in this + * case we must do something about the corruption. + * Ignoring it is worse than a unlikely patching race. + * Also machine checks tend to be broadcast and if one CPU + * goes into machine check the others follow quickly, so we don't + * expect a machine check to cause undue problems during to code + * patching. + */ apply_alternatives(__alt_instructions, __alt_instructions_end); @@ -456,9 +464,6 @@ void __init alternative_instructions(void) (unsigned long)__smp_locks_end); restart_nmi(); -#ifdef CONFIG_X86_MCE - restart_mce(); -#endif } /** diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index f9cecdfd05c5..85eb8e100818 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -46,6 +46,7 @@ #include <asm/idle.h> #include <asm/mtrr.h> #include <asm/smp.h> +#include <asm/mce.h> unsigned int num_processors; @@ -808,7 +809,7 @@ void clear_local_APIC(void) u32 v; /* APIC hasn't been mapped yet */ - if (!apic_phys) + if (!x2apic && !apic_phys) return; maxlvt = lapic_get_maxlvt(); @@ -842,6 +843,14 @@ void clear_local_APIC(void) apic_write(APIC_LVTTHMR, v | APIC_LVT_MASKED); } #endif +#ifdef CONFIG_X86_MCE_INTEL + if (maxlvt >= 6) { + v = apic_read(APIC_LVTCMCI); + if (!(v & APIC_LVT_MASKED)) + apic_write(APIC_LVTCMCI, v | APIC_LVT_MASKED); + } +#endif + /* * Clean APIC state for other OSs: */ @@ -1241,6 +1250,12 @@ void __cpuinit setup_local_APIC(void) apic_write(APIC_LVT1, value); preempt_enable(); + +#ifdef CONFIG_X86_MCE_INTEL + /* Recheck CMCI information after local APIC is up on CPU #0 */ + if (smp_processor_id() == 0) + cmci_recheck(); +#endif } void __cpuinit end_local_APIC_setup(void) @@ -1319,15 +1334,16 @@ void __init enable_IR_x2apic(void) return; } - local_irq_save(flags); - mask_8259A(); - - ret = save_mask_IO_APIC_setup(); + ret = save_IO_APIC_setup(); if (ret) { pr_info("Saving IO-APIC state failed: %d\n", ret); goto end; } + local_irq_save(flags); + mask_IO_APIC_setup(); + mask_8259A(); + ret = enable_intr_remapping(1); if (ret && x2apic_preenabled) { @@ -1352,10 +1368,10 @@ end_restore: else reinit_intr_remapped_IO_APIC(x2apic_preenabled); -end: unmask_8259A(); local_irq_restore(flags); +end: if (!ret) { if (!x2apic_preenabled) pr_info("Enabled x2apic and interrupt-remapping\n"); @@ -1508,12 +1524,10 @@ void __init early_init_lapic_mapping(void) */ void __init init_apic_mappings(void) { -#ifdef CONFIG_X86_X2APIC if (x2apic) { boot_cpu_physical_apicid = read_apic_id(); return; } -#endif /* * If no local APIC can be found then set up a fake all @@ -1957,12 +1971,9 @@ static int lapic_resume(struct sys_device *dev) local_irq_save(flags); -#ifdef CONFIG_X86_X2APIC if (x2apic) enable_x2apic(); - else -#endif - { + else { /* * Make sure the APICBASE points to the right address * diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index f933822dba18..0014714ea97b 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -159,20 +159,6 @@ static int flat_apic_id_registered(void) return physid_isset(read_xapic_id(), phys_cpu_present_map); } -static unsigned int flat_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - return cpumask_bits(cpumask)[0] & APIC_ALL_CPUS; -} - -static unsigned int flat_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - unsigned long mask1 = cpumask_bits(cpumask)[0] & APIC_ALL_CPUS; - unsigned long mask2 = cpumask_bits(andmask)[0] & APIC_ALL_CPUS; - - return mask1 & mask2; -} - static int flat_phys_pkg_id(int initial_apic_id, int index_msb) { return hard_smp_processor_id() >> index_msb; @@ -213,8 +199,8 @@ struct apic apic_flat = { .set_apic_id = set_apic_id, .apic_id_mask = 0xFFu << 24, - .cpu_mask_to_apicid = flat_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = flat_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid = default_cpu_mask_to_apicid, + .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, .send_IPI_mask = flat_send_IPI_mask, .send_IPI_mask_allbutself = flat_send_IPI_mask_allbutself, diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 00e6071cefc4..da99ffcdfde6 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -389,6 +389,8 @@ struct io_apic { unsigned int index; unsigned int unused[3]; unsigned int data; + unsigned int unused2[11]; + unsigned int eoi; }; static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) @@ -397,6 +399,12 @@ static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) + (mp_ioapics[idx].apicaddr & ~PAGE_MASK); } +static inline void io_apic_eoi(unsigned int apic, unsigned int vector) +{ + struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(vector, &io_apic->eoi); +} + static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg) { struct io_apic __iomem *io_apic = io_apic_base(apic); @@ -546,16 +554,12 @@ static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq apic = entry->apic; pin = entry->pin; -#ifdef CONFIG_INTR_REMAP /* * With interrupt-remapping, destination information comes * from interrupt-remapping table entry. */ if (!irq_remapped(irq)) io_apic_write(apic, 0x11 + pin*2, dest); -#else - io_apic_write(apic, 0x11 + pin*2, dest); -#endif reg = io_apic_read(apic, 0x10 + pin*2); reg &= ~IO_APIC_REDIR_VECTOR_MASK; reg |= vector; @@ -588,10 +592,12 @@ set_desc_affinity(struct irq_desc *desc, const struct cpumask *mask) if (assign_irq_vector(irq, cfg, mask)) return BAD_APICID; - cpumask_and(desc->affinity, cfg->domain, mask); + /* check that before desc->addinity get updated */ set_extra_move_desc(desc, mask); - return apic->cpu_mask_to_apicid_and(desc->affinity, cpu_online_mask); + cpumask_copy(desc->affinity, mask); + + return apic->cpu_mask_to_apicid_and(desc->affinity, cfg->domain); } static void @@ -849,9 +855,9 @@ __setup("pirq=", ioapic_pirq_setup); static struct IO_APIC_route_entry *early_ioapic_entries[MAX_IO_APICS]; /* - * Saves and masks all the unmasked IO-APIC RTE's + * Saves all the IO-APIC RTE's */ -int save_mask_IO_APIC_setup(void) +int save_IO_APIC_setup(void) { union IO_APIC_reg_01 reg_01; unsigned long flags; @@ -876,16 +882,9 @@ int save_mask_IO_APIC_setup(void) } for (apic = 0; apic < nr_ioapics; apic++) - for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { - struct IO_APIC_route_entry entry; - - entry = early_ioapic_entries[apic][pin] = + for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) + early_ioapic_entries[apic][pin] = ioapic_read_entry(apic, pin); - if (!entry.mask) { - entry.mask = 1; - ioapic_write_entry(apic, pin, entry); - } - } return 0; @@ -898,6 +897,25 @@ nomem: return -ENOMEM; } +void mask_IO_APIC_setup(void) +{ + int apic, pin; + + for (apic = 0; apic < nr_ioapics; apic++) { + if (!early_ioapic_entries[apic]) + break; + for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { + struct IO_APIC_route_entry entry; + + entry = early_ioapic_entries[apic][pin]; + if (!entry.mask) { + entry.mask = 1; + ioapic_write_entry(apic, pin, entry); + } + } + } +} + void restore_IO_APIC_setup(void) { int apic, pin; @@ -1411,9 +1429,7 @@ void __setup_vector_irq(int cpu) } static struct irq_chip ioapic_chip; -#ifdef CONFIG_INTR_REMAP static struct irq_chip ir_ioapic_chip; -#endif #define IOAPIC_AUTO -1 #define IOAPIC_EDGE 0 @@ -1452,7 +1468,6 @@ static void ioapic_register_intr(int irq, struct irq_desc *desc, unsigned long t else desc->status &= ~IRQ_LEVEL; -#ifdef CONFIG_INTR_REMAP if (irq_remapped(irq)) { desc->status |= IRQ_MOVE_PCNTXT; if (trigger) @@ -1464,7 +1479,7 @@ static void ioapic_register_intr(int irq, struct irq_desc *desc, unsigned long t handle_edge_irq, "edge"); return; } -#endif + if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) || trigger == IOAPIC_LEVEL) set_irq_chip_and_handler_name(irq, &ioapic_chip, @@ -1478,14 +1493,13 @@ static void ioapic_register_intr(int irq, struct irq_desc *desc, unsigned long t int setup_ioapic_entry(int apic_id, int irq, struct IO_APIC_route_entry *entry, unsigned int destination, int trigger, - int polarity, int vector) + int polarity, int vector, int pin) { /* * add it to the IO-APIC irq-routing table: */ memset(entry,0,sizeof(*entry)); -#ifdef CONFIG_INTR_REMAP if (intr_remapping_enabled) { struct intel_iommu *iommu = map_ioapic_to_ir(apic_id); struct irte irte; @@ -1504,7 +1518,14 @@ int setup_ioapic_entry(int apic_id, int irq, irte.present = 1; irte.dst_mode = apic->irq_dest_mode; - irte.trigger_mode = trigger; + /* + * Trigger mode in the IRTE will always be edge, and the + * actual level or edge trigger will be setup in the IO-APIC + * RTE. This will help simplify level triggered irq migration. + * For more details, see the comments above explainig IO-APIC + * irq migration in the presence of interrupt-remapping. + */ + irte.trigger_mode = 0; irte.dlvry_mode = apic->irq_delivery_mode; irte.vector = vector; irte.dest_id = IRTE_DEST(destination); @@ -1515,18 +1536,21 @@ int setup_ioapic_entry(int apic_id, int irq, ir_entry->zero = 0; ir_entry->format = 1; ir_entry->index = (index & 0x7fff); - } else -#endif - { + /* + * IO-APIC RTE will be configured with virtual vector. + * irq handler will do the explicit EOI to the io-apic. + */ + ir_entry->vector = pin; + } else { entry->delivery_mode = apic->irq_delivery_mode; entry->dest_mode = apic->irq_dest_mode; entry->dest = destination; + entry->vector = vector; } entry->mask = 0; /* enable IRQ */ entry->trigger = trigger; entry->polarity = polarity; - entry->vector = vector; /* Mask level triggered irqs. * Use IRQ_DELAYED_DISABLE for edge triggered irqs. @@ -1561,7 +1585,7 @@ static void setup_IO_APIC_irq(int apic_id, int pin, unsigned int irq, struct irq if (setup_ioapic_entry(mp_ioapics[apic_id].apicid, irq, &entry, - dest, trigger, polarity, cfg->vector)) { + dest, trigger, polarity, cfg->vector, pin)) { printk("Failed to setup ioapic entry for ioapic %d, pin %d\n", mp_ioapics[apic_id].apicid, pin); __clear_irq_vector(irq, cfg); @@ -1642,10 +1666,8 @@ static void __init setup_timer_IRQ0_pin(unsigned int apic_id, unsigned int pin, { struct IO_APIC_route_entry entry; -#ifdef CONFIG_INTR_REMAP if (intr_remapping_enabled) return; -#endif memset(&entry, 0, sizeof(entry)); @@ -2040,8 +2062,13 @@ void disable_IO_APIC(void) * If the i8259 is routed through an IOAPIC * Put that IOAPIC in virtual wire mode * so legacy interrupts can be delivered. + * + * With interrupt-remapping, for now we will use virtual wire A mode, + * as virtual wire B is little complex (need to configure both + * IOAPIC RTE aswell as interrupt-remapping table entry). + * As this gets called during crash dump, keep this simple for now. */ - if (ioapic_i8259.pin != -1) { + if (ioapic_i8259.pin != -1 && !intr_remapping_enabled) { struct IO_APIC_route_entry entry; memset(&entry, 0, sizeof(entry)); @@ -2061,7 +2088,10 @@ void disable_IO_APIC(void) ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry); } - disconnect_bsp_APIC(ioapic_i8259.pin != -1); + /* + * Use virtual wire A mode when interrupt remapping is enabled. + */ + disconnect_bsp_APIC(!intr_remapping_enabled && ioapic_i8259.pin != -1); } #ifdef CONFIG_X86_32 @@ -2303,37 +2333,24 @@ static int ioapic_retrigger_irq(unsigned int irq) #ifdef CONFIG_SMP #ifdef CONFIG_INTR_REMAP -static void ir_irq_migration(struct work_struct *work); - -static DECLARE_DELAYED_WORK(ir_migration_work, ir_irq_migration); /* * Migrate the IO-APIC irq in the presence of intr-remapping. * - * For edge triggered, irq migration is a simple atomic update(of vector - * and cpu destination) of IRTE and flush the hardware cache. - * - * For level triggered, we need to modify the io-apic RTE aswell with the update - * vector information, along with modifying IRTE with vector and destination. - * So irq migration for level triggered is little bit more complex compared to - * edge triggered migration. But the good news is, we use the same algorithm - * for level triggered migration as we have today, only difference being, - * we now initiate the irq migration from process context instead of the - * interrupt context. + * For both level and edge triggered, irq migration is a simple atomic + * update(of vector and cpu destination) of IRTE and flush the hardware cache. * - * In future, when we do a directed EOI (combined with cpu EOI broadcast - * suppression) to the IO-APIC, level triggered irq migration will also be - * as simple as edge triggered migration and we can do the irq migration - * with a simple atomic update to IO-APIC RTE. + * For level triggered, we eliminate the io-apic RTE modification (with the + * updated vector information), by using a virtual vector (io-apic pin number). + * Real vector that is used for interrupting cpu will be coming from + * the interrupt-remapping table entry. */ static void migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask) { struct irq_cfg *cfg; struct irte irte; - int modify_ioapic_rte; unsigned int dest; - unsigned long flags; unsigned int irq; if (!cpumask_intersects(mask, cpu_online_mask)) @@ -2351,13 +2368,6 @@ migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask) dest = apic->cpu_mask_to_apicid_and(cfg->domain, mask); - modify_ioapic_rte = desc->status & IRQ_LEVEL; - if (modify_ioapic_rte) { - spin_lock_irqsave(&ioapic_lock, flags); - __target_IO_APIC_irq(irq, dest, cfg); - spin_unlock_irqrestore(&ioapic_lock, flags); - } - irte.vector = cfg->vector; irte.dest_id = IRTE_DEST(dest); @@ -2372,73 +2382,12 @@ migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask) cpumask_copy(desc->affinity, mask); } -static int migrate_irq_remapped_level_desc(struct irq_desc *desc) -{ - int ret = -1; - struct irq_cfg *cfg = desc->chip_data; - - mask_IO_APIC_irq_desc(desc); - - if (io_apic_level_ack_pending(cfg)) { - /* - * Interrupt in progress. Migrating irq now will change the - * vector information in the IO-APIC RTE and that will confuse - * the EOI broadcast performed by cpu. - * So, delay the irq migration to the next instance. - */ - schedule_delayed_work(&ir_migration_work, 1); - goto unmask; - } - - /* everthing is clear. we have right of way */ - migrate_ioapic_irq_desc(desc, desc->pending_mask); - - ret = 0; - desc->status &= ~IRQ_MOVE_PENDING; - cpumask_clear(desc->pending_mask); - -unmask: - unmask_IO_APIC_irq_desc(desc); - - return ret; -} - -static void ir_irq_migration(struct work_struct *work) -{ - unsigned int irq; - struct irq_desc *desc; - - for_each_irq_desc(irq, desc) { - if (desc->status & IRQ_MOVE_PENDING) { - unsigned long flags; - - spin_lock_irqsave(&desc->lock, flags); - if (!desc->chip->set_affinity || - !(desc->status & IRQ_MOVE_PENDING)) { - desc->status &= ~IRQ_MOVE_PENDING; - spin_unlock_irqrestore(&desc->lock, flags); - continue; - } - - desc->chip->set_affinity(irq, desc->pending_mask); - spin_unlock_irqrestore(&desc->lock, flags); - } - } -} - /* * Migrates the IRQ destination in the process context. */ static void set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask) { - if (desc->status & IRQ_LEVEL) { - desc->status |= IRQ_MOVE_PENDING; - cpumask_copy(desc->pending_mask, mask); - migrate_irq_remapped_level_desc(desc); - return; - } - migrate_ioapic_irq_desc(desc, mask); } static void set_ir_ioapic_affinity_irq(unsigned int irq, @@ -2448,6 +2397,11 @@ static void set_ir_ioapic_affinity_irq(unsigned int irq, set_ir_ioapic_affinity_irq_desc(desc, mask); } +#else +static inline void set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc, + const struct cpumask *mask) +{ +} #endif asmlinkage void smp_irq_move_cleanup_interrupt(void) @@ -2461,6 +2415,7 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void) me = smp_processor_id(); for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) { unsigned int irq; + unsigned int irr; struct irq_desc *desc; struct irq_cfg *cfg; irq = __get_cpu_var(vector_irq)[vector]; @@ -2480,6 +2435,18 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void) if (vector == cfg->vector && cpumask_test_cpu(me, cfg->domain)) goto unlock; + irr = apic_read(APIC_IRR + (vector / 32 * 0x10)); + /* + * Check if the vector that needs to be cleanedup is + * registered at the cpu's IRR. If so, then this is not + * the best time to clean it up. Lets clean it up in the + * next attempt by sending another IRQ_MOVE_CLEANUP_VECTOR + * to myself. + */ + if (irr & (1 << (vector % 32))) { + apic->send_IPI_self(IRQ_MOVE_CLEANUP_VECTOR); + goto unlock; + } __get_cpu_var(vector_irq)[vector] = -1; cfg->move_cleanup_count--; unlock: @@ -2529,9 +2496,44 @@ static inline void irq_complete_move(struct irq_desc **descp) {} #endif #ifdef CONFIG_INTR_REMAP +static void __eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg) +{ + int apic, pin; + struct irq_pin_list *entry; + + entry = cfg->irq_2_pin; + for (;;) { + + if (!entry) + break; + + apic = entry->apic; + pin = entry->pin; + io_apic_eoi(apic, pin); + entry = entry->next; + } +} + +static void +eoi_ioapic_irq(struct irq_desc *desc) +{ + struct irq_cfg *cfg; + unsigned long flags; + unsigned int irq; + + irq = desc->irq; + cfg = desc->chip_data; + + spin_lock_irqsave(&ioapic_lock, flags); + __eoi_ioapic_irq(irq, cfg); + spin_unlock_irqrestore(&ioapic_lock, flags); +} + static void ack_x2apic_level(unsigned int irq) { + struct irq_desc *desc = irq_to_desc(irq); ack_x2APIC_irq(); + eoi_ioapic_irq(desc); } static void ack_x2apic_edge(unsigned int irq) @@ -2662,20 +2664,20 @@ static struct irq_chip ioapic_chip __read_mostly = { .retrigger = ioapic_retrigger_irq, }; -#ifdef CONFIG_INTR_REMAP static struct irq_chip ir_ioapic_chip __read_mostly = { .name = "IR-IO-APIC", .startup = startup_ioapic_irq, .mask = mask_IO_APIC_irq, .unmask = unmask_IO_APIC_irq, +#ifdef CONFIG_INTR_REMAP .ack = ack_x2apic_edge, .eoi = ack_x2apic_level, #ifdef CONFIG_SMP .set_affinity = set_ir_ioapic_affinity_irq, #endif +#endif .retrigger = ioapic_retrigger_irq, }; -#endif static inline void init_IO_APIC_traps(void) { @@ -2901,10 +2903,8 @@ static inline void __init check_timer(void) * 8259A. */ if (pin1 == -1) { -#ifdef CONFIG_INTR_REMAP if (intr_remapping_enabled) panic("BIOS bug: timer not connected to IO-APIC"); -#endif pin1 = pin2; apic1 = apic2; no_pin1 = 1; @@ -2940,10 +2940,8 @@ static inline void __init check_timer(void) clear_IO_APIC_pin(0, pin1); goto out; } -#ifdef CONFIG_INTR_REMAP if (intr_remapping_enabled) panic("timer doesn't work through Interrupt-remapped IO-APIC"); -#endif local_irq_disable(); clear_IO_APIC_pin(apic1, pin1); if (!no_pin1) @@ -3237,9 +3235,7 @@ void destroy_irq(unsigned int irq) if (desc) desc->chip_data = cfg; -#ifdef CONFIG_INTR_REMAP free_irte(irq); -#endif spin_lock_irqsave(&vector_lock, flags); __clear_irq_vector(irq, cfg); spin_unlock_irqrestore(&vector_lock, flags); @@ -3265,7 +3261,6 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms dest = apic->cpu_mask_to_apicid_and(cfg->domain, apic->target_cpus()); -#ifdef CONFIG_INTR_REMAP if (irq_remapped(irq)) { struct irte irte; int ir_index; @@ -3291,10 +3286,13 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms MSI_ADDR_IR_SHV | MSI_ADDR_IR_INDEX1(ir_index) | MSI_ADDR_IR_INDEX2(ir_index); - } else -#endif - { - msg->address_hi = MSI_ADDR_BASE_HI; + } else { + if (x2apic_enabled()) + msg->address_hi = MSI_ADDR_BASE_HI | + MSI_ADDR_EXT_DEST_ID(dest); + else + msg->address_hi = MSI_ADDR_BASE_HI; + msg->address_lo = MSI_ADDR_BASE_LO | ((apic->irq_dest_mode == 0) ? @@ -3394,15 +3392,16 @@ static struct irq_chip msi_chip = { .retrigger = ioapic_retrigger_irq, }; -#ifdef CONFIG_INTR_REMAP static struct irq_chip msi_ir_chip = { .name = "IR-PCI-MSI", .unmask = unmask_msi_irq, .mask = mask_msi_irq, +#ifdef CONFIG_INTR_REMAP .ack = ack_x2apic_edge, #ifdef CONFIG_SMP .set_affinity = ir_set_msi_irq_affinity, #endif +#endif .retrigger = ioapic_retrigger_irq, }; @@ -3432,7 +3431,6 @@ static int msi_alloc_irte(struct pci_dev *dev, int irq, int nvec) } return index; } -#endif static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int irq) { @@ -3446,7 +3444,6 @@ static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int irq) set_irq_msi(irq, msidesc); write_msi_msg(irq, &msg); -#ifdef CONFIG_INTR_REMAP if (irq_remapped(irq)) { struct irq_desc *desc = irq_to_desc(irq); /* @@ -3455,7 +3452,6 @@ static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int irq) desc->status |= IRQ_MOVE_PCNTXT; set_irq_chip_and_handler_name(irq, &msi_ir_chip, handle_edge_irq, "edge"); } else -#endif set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge"); dev_printk(KERN_DEBUG, &dev->dev, "irq %d for MSI/MSI-X\n", irq); @@ -3469,11 +3465,8 @@ int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) int ret, sub_handle; struct msi_desc *msidesc; unsigned int irq_want; - -#ifdef CONFIG_INTR_REMAP - struct intel_iommu *iommu = 0; + struct intel_iommu *iommu = NULL; int index = 0; -#endif irq_want = nr_irqs_gsi; sub_handle = 0; @@ -3482,7 +3475,6 @@ int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) if (irq == 0) return -1; irq_want = irq + 1; -#ifdef CONFIG_INTR_REMAP if (!intr_remapping_enabled) goto no_ir; @@ -3510,7 +3502,6 @@ int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) set_irte_irq(irq, iommu, index, sub_handle); } no_ir: -#endif ret = setup_msi_irq(dev, msidesc, irq); if (ret < 0) goto error; @@ -3528,7 +3519,7 @@ void arch_teardown_msi_irq(unsigned int irq) destroy_irq(irq); } -#ifdef CONFIG_DMAR +#if defined (CONFIG_DMAR) || defined (CONFIG_INTR_REMAP) #ifdef CONFIG_SMP static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) { @@ -3609,7 +3600,7 @@ static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask) #endif /* CONFIG_SMP */ -struct irq_chip hpet_msi_type = { +static struct irq_chip hpet_msi_type = { .name = "HPET_MSI", .unmask = hpet_msi_unmask, .mask = hpet_msi_mask, @@ -4045,11 +4036,9 @@ void __init setup_ioapic_dest(void) else mask = apic->target_cpus(); -#ifdef CONFIG_INTR_REMAP if (intr_remapping_enabled) set_ir_ioapic_affinity_irq_desc(desc, mask); else -#endif set_ioapic_affinity_irq_desc(desc, mask); } @@ -4142,9 +4131,12 @@ static int __init ioapic_insert_resources(void) struct resource *r = ioapic_resources; if (!r) { - printk(KERN_ERR - "IO APIC resources could be not be allocated.\n"); - return -1; + if (nr_ioapics > 0) { + printk(KERN_ERR + "IO APIC resources couldn't be allocated.\n"); + return -1; + } + return 0; } for (i = 0; i < nr_ioapics; i++) { diff --git a/arch/x86/kernel/apic/probe_64.c b/arch/x86/kernel/apic/probe_64.c index 8d7748efe6a8..1783652bb0e5 100644 --- a/arch/x86/kernel/apic/probe_64.c +++ b/arch/x86/kernel/apic/probe_64.c @@ -68,6 +68,13 @@ void __init default_setup_apic_routing(void) apic = &apic_physflat; printk(KERN_INFO "Setting APIC routing to %s\n", apic->name); } + + /* + * Now that apic routing model is selected, configure the + * fault handling for intr remapping. + */ + if (intr_remapping_enabled) + enable_drhd_fault_handling(); } /* Same for both flat and physical. */ diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index 8fb87b6dd633..4a903e2f0d17 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -57,6 +57,8 @@ static void x2apic_send_IPI_mask(const struct cpumask *mask, int vector) unsigned long query_cpu; unsigned long flags; + x2apic_wrmsr_fence(); + local_irq_save(flags); for_each_cpu(query_cpu, mask) { __x2apic_send_IPI_dest( @@ -73,6 +75,8 @@ static void unsigned long query_cpu; unsigned long flags; + x2apic_wrmsr_fence(); + local_irq_save(flags); for_each_cpu(query_cpu, mask) { if (query_cpu == this_cpu) @@ -90,6 +94,8 @@ static void x2apic_send_IPI_allbutself(int vector) unsigned long query_cpu; unsigned long flags; + x2apic_wrmsr_fence(); + local_irq_save(flags); for_each_online_cpu(query_cpu) { if (query_cpu == this_cpu) diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index 23625b9f98b2..a284359627e7 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -58,6 +58,8 @@ static void x2apic_send_IPI_mask(const struct cpumask *mask, int vector) unsigned long query_cpu; unsigned long flags; + x2apic_wrmsr_fence(); + local_irq_save(flags); for_each_cpu(query_cpu, mask) { __x2apic_send_IPI_dest(per_cpu(x86_cpu_to_apicid, query_cpu), @@ -73,6 +75,8 @@ static void unsigned long query_cpu; unsigned long flags; + x2apic_wrmsr_fence(); + local_irq_save(flags); for_each_cpu(query_cpu, mask) { if (query_cpu != this_cpu) @@ -89,6 +93,8 @@ static void x2apic_send_IPI_allbutself(int vector) unsigned long query_cpu; unsigned long flags; + x2apic_wrmsr_fence(); + local_irq_save(flags); for_each_online_cpu(query_cpu) { if (query_cpu == this_cpu) diff --git a/arch/x86/kernel/check.c b/arch/x86/kernel/check.c index 2ac0ab71412a..fc999e6fc46a 100644 --- a/arch/x86/kernel/check.c +++ b/arch/x86/kernel/check.c @@ -83,15 +83,15 @@ void __init setup_bios_corruption_check(void) u64 size; addr = find_e820_area_size(addr, &size, PAGE_SIZE); - if (addr == 0) + if (!(addr + 1)) + break; + + if (addr >= corruption_check_size) break; if ((addr + size) > corruption_check_size) size = corruption_check_size - addr; - if (size == 0) - break; - e820_update_range(addr, size, E820_RAM, E820_RESERVED); scan_areas[num_scan_areas].addr = addr; scan_areas[num_scan_areas].size = size; diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 82db7f45e2de..4e242f9a06e4 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -14,11 +14,12 @@ obj-y += vmware.o hypervisor.o obj-$(CONFIG_X86_32) += bugs.o cmpxchg.o obj-$(CONFIG_X86_64) += bugs_64.o +obj-$(CONFIG_X86_CPU_DEBUG) += cpu_debug.o + obj-$(CONFIG_CPU_SUP_INTEL) += intel.o obj-$(CONFIG_CPU_SUP_AMD) += amd.o obj-$(CONFIG_CPU_SUP_CYRIX_32) += cyrix.o -obj-$(CONFIG_CPU_SUP_CENTAUR_32) += centaur.o -obj-$(CONFIG_CPU_SUP_CENTAUR_64) += centaur_64.o +obj-$(CONFIG_CPU_SUP_CENTAUR) += centaur.o obj-$(CONFIG_CPU_SUP_TRANSMETA_32) += transmeta.o obj-$(CONFIG_CPU_SUP_UMC_32) += umc.o diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c index 6882a735d9c0..8220ae69849d 100644 --- a/arch/x86/kernel/cpu/addon_cpuid_features.c +++ b/arch/x86/kernel/cpu/addon_cpuid_features.c @@ -29,7 +29,7 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c) u32 regs[4]; const struct cpuid_bit *cb; - static const struct cpuid_bit cpuid_bits[] = { + static const struct cpuid_bit __cpuinitconst cpuid_bits[] = { { X86_FEATURE_IDA, CR_EAX, 1, 0x00000006 }, { 0, 0, 0, 0 } }; diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 25423a5b80ed..7e4a459daa64 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -5,6 +5,7 @@ #include <asm/io.h> #include <asm/processor.h> #include <asm/apic.h> +#include <asm/cpu.h> #ifdef CONFIG_X86_64 # include <asm/numa_64.h> @@ -141,6 +142,55 @@ static void __cpuinit init_amd_k6(struct cpuinfo_x86 *c) } } +static void __cpuinit amd_k7_smp_check(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + /* calling is from identify_secondary_cpu() ? */ + if (c->cpu_index == boot_cpu_id) + return; + + /* + * Certain Athlons might work (for various values of 'work') in SMP + * but they are not certified as MP capable. + */ + /* Athlon 660/661 is valid. */ + if ((c->x86_model == 6) && ((c->x86_mask == 0) || + (c->x86_mask == 1))) + goto valid_k7; + + /* Duron 670 is valid */ + if ((c->x86_model == 7) && (c->x86_mask == 0)) + goto valid_k7; + + /* + * Athlon 662, Duron 671, and Athlon >model 7 have capability + * bit. It's worth noting that the A5 stepping (662) of some + * Athlon XP's have the MP bit set. + * See http://www.heise.de/newsticker/data/jow-18.10.01-000 for + * more. + */ + if (((c->x86_model == 6) && (c->x86_mask >= 2)) || + ((c->x86_model == 7) && (c->x86_mask >= 1)) || + (c->x86_model > 7)) + if (cpu_has_mp) + goto valid_k7; + + /* If we get here, not a certified SMP capable AMD system. */ + + /* + * Don't taint if we are running SMP kernel on a single non-MP + * approved Athlon + */ + WARN_ONCE(1, "WARNING: This combination of AMD" + "processors is not suitable for SMP.\n"); + if (!test_taint(TAINT_UNSAFE_SMP)) + add_taint(TAINT_UNSAFE_SMP); + +valid_k7: + ; +#endif +} + static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c) { u32 l, h; @@ -175,6 +225,8 @@ static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c) } set_cpu_cap(c, X86_FEATURE_K7); + + amd_k7_smp_check(c); } #endif @@ -450,7 +502,7 @@ static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 *c, unsigned int } #endif -static struct cpu_dev amd_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst amd_cpu_dev = { .c_vendor = "AMD", .c_ident = { "AuthenticAMD" }, #ifdef CONFIG_X86_32 diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c index 89bfdd9cacc6..c95e831bb095 100644 --- a/arch/x86/kernel/cpu/centaur.c +++ b/arch/x86/kernel/cpu/centaur.c @@ -1,11 +1,11 @@ +#include <linux/bitops.h> #include <linux/kernel.h> #include <linux/init.h> -#include <linux/bitops.h> #include <asm/processor.h> -#include <asm/msr.h> #include <asm/e820.h> #include <asm/mtrr.h> +#include <asm/msr.h> #include "cpu.h" @@ -276,7 +276,7 @@ static void __cpuinit init_c3(struct cpuinfo_x86 *c) */ c->x86_capability[5] = cpuid_edx(0xC0000001); } - +#ifdef CONFIG_X86_32 /* Cyrix III family needs CX8 & PGE explicitly enabled. */ if (c->x86_model >= 6 && c->x86_model <= 9) { rdmsr(MSR_VIA_FCR, lo, hi); @@ -288,6 +288,11 @@ static void __cpuinit init_c3(struct cpuinfo_x86 *c) /* Before Nehemiah, the C3's had 3dNOW! */ if (c->x86_model >= 6 && c->x86_model < 9) set_cpu_cap(c, X86_FEATURE_3DNOW); +#endif + if (c->x86 == 0x6 && c->x86_model >= 0xf) { + c->x86_cache_alignment = c->x86_clflush_size * 2; + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + } display_cacheinfo(c); } @@ -316,16 +321,25 @@ enum { static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) { switch (c->x86) { +#ifdef CONFIG_X86_32 case 5: /* Emulate MTRRs using Centaur's MCR. */ set_cpu_cap(c, X86_FEATURE_CENTAUR_MCR); break; +#endif + case 6: + if (c->x86_model >= 0xf) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + break; } +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_SYSENTER32); +#endif } static void __cpuinit init_centaur(struct cpuinfo_x86 *c) { - +#ifdef CONFIG_X86_32 char *name; u32 fcr_set = 0; u32 fcr_clr = 0; @@ -337,8 +351,10 @@ static void __cpuinit init_centaur(struct cpuinfo_x86 *c) * 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ clear_cpu_cap(c, 0*32+31); - +#endif + early_init_centaur(c); switch (c->x86) { +#ifdef CONFIG_X86_32 case 5: switch (c->x86_model) { case 4: @@ -442,16 +458,20 @@ static void __cpuinit init_centaur(struct cpuinfo_x86 *c) } sprintf(c->x86_model_id, "WinChip %s", name); break; - +#endif case 6: init_c3(c); break; } +#ifdef CONFIG_X86_64 + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); +#endif } static unsigned int __cpuinit centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) { +#ifdef CONFIG_X86_32 /* VIA C3 CPUs (670-68F) need further shifting. */ if ((c->x86 == 6) && ((c->x86_model == 7) || (c->x86_model == 8))) size >>= 8; @@ -464,11 +484,11 @@ centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) if ((c->x86 == 6) && (c->x86_model == 9) && (c->x86_mask == 1) && (size == 65)) size -= 1; - +#endif return size; } -static struct cpu_dev centaur_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst centaur_cpu_dev = { .c_vendor = "Centaur", .c_ident = { "CentaurHauls" }, .c_early_init = early_init_centaur, diff --git a/arch/x86/kernel/cpu/centaur_64.c b/arch/x86/kernel/cpu/centaur_64.c deleted file mode 100644 index a1625f5a1e78..000000000000 --- a/arch/x86/kernel/cpu/centaur_64.c +++ /dev/null @@ -1,37 +0,0 @@ -#include <linux/init.h> -#include <linux/smp.h> - -#include <asm/cpufeature.h> -#include <asm/processor.h> - -#include "cpu.h" - -static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) -{ - if (c->x86 == 0x6 && c->x86_model >= 0xf) - set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); - - set_cpu_cap(c, X86_FEATURE_SYSENTER32); -} - -static void __cpuinit init_centaur(struct cpuinfo_x86 *c) -{ - early_init_centaur(c); - - if (c->x86 == 0x6 && c->x86_model >= 0xf) { - c->x86_cache_alignment = c->x86_clflush_size * 2; - set_cpu_cap(c, X86_FEATURE_REP_GOOD); - } - set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); -} - -static struct cpu_dev centaur_cpu_dev __cpuinitdata = { - .c_vendor = "Centaur", - .c_ident = { "CentaurHauls" }, - .c_early_init = early_init_centaur, - .c_init = init_centaur, - .c_x86_vendor = X86_VENDOR_CENTAUR, -}; - -cpu_dev_register(centaur_cpu_dev); - diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 826d5c876278..e2962cc1e27b 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1,52 +1,52 @@ -#include <linux/init.h> -#include <linux/kernel.h> -#include <linux/sched.h> -#include <linux/string.h> #include <linux/bootmem.h> +#include <linux/linkage.h> #include <linux/bitops.h> +#include <linux/kernel.h> #include <linux/module.h> -#include <linux/kgdb.h> -#include <linux/topology.h> +#include <linux/percpu.h> +#include <linux/string.h> #include <linux/delay.h> +#include <linux/sched.h> +#include <linux/init.h> +#include <linux/kgdb.h> #include <linux/smp.h> -#include <linux/percpu.h> -#include <asm/i387.h> -#include <asm/msr.h> -#include <asm/io.h> -#include <asm/linkage.h> +#include <linux/io.h> + +#include <asm/stackprotector.h> #include <asm/mmu_context.h> +#include <asm/hypervisor.h> +#include <asm/processor.h> +#include <asm/sections.h> +#include <asm/topology.h> +#include <asm/cpumask.h> +#include <asm/pgtable.h> +#include <asm/atomic.h> +#include <asm/proto.h> +#include <asm/setup.h> +#include <asm/apic.h> +#include <asm/desc.h> +#include <asm/i387.h> #include <asm/mtrr.h> +#include <asm/numa.h> +#include <asm/asm.h> +#include <asm/cpu.h> #include <asm/mce.h> +#include <asm/msr.h> #include <asm/pat.h> -#include <asm/asm.h> -#include <asm/numa.h> #include <asm/smp.h> -#include <asm/cpu.h> -#include <asm/cpumask.h> -#include <asm/apic.h> #ifdef CONFIG_X86_LOCAL_APIC #include <asm/uv/uv.h> #endif -#include <asm/pgtable.h> -#include <asm/processor.h> -#include <asm/desc.h> -#include <asm/atomic.h> -#include <asm/proto.h> -#include <asm/sections.h> -#include <asm/setup.h> -#include <asm/hypervisor.h> -#include <asm/stackprotector.h> - #include "cpu.h" #ifdef CONFIG_X86_64 /* all of these masks are initialized in setup_cpu_local_masks() */ -cpumask_var_t cpu_callin_mask; -cpumask_var_t cpu_callout_mask; cpumask_var_t cpu_initialized_mask; +cpumask_var_t cpu_callout_mask; +cpumask_var_t cpu_callin_mask; /* representing cpus for which sibling maps can be computed */ cpumask_var_t cpu_sibling_setup_mask; @@ -62,15 +62,15 @@ void __init setup_cpu_local_masks(void) #else /* CONFIG_X86_32 */ -cpumask_t cpu_callin_map; +cpumask_t cpu_sibling_setup_map; cpumask_t cpu_callout_map; cpumask_t cpu_initialized; -cpumask_t cpu_sibling_setup_map; +cpumask_t cpu_callin_map; #endif /* CONFIG_X86_32 */ -static struct cpu_dev *this_cpu __cpuinitdata; +static const struct cpu_dev *this_cpu __cpuinitdata; DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { #ifdef CONFIG_X86_64 @@ -79,48 +79,48 @@ DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { * IRET will check the segment types kkeil 2000/10/28 * Also sysret mandates a special GDT layout * - * The TLS descriptors are currently at a different place compared to i386. + * TLS descriptors are currently at a different place compared to i386. * Hopefully nobody expects them at a fixed place (Wine?) */ - [GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } }, - [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } }, - [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } }, - [GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } }, - [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } }, - [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } }, + [GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } }, + [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } }, + [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } }, + [GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } }, + [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } }, + [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } }, #else - [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } }, - [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } }, - [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } }, - [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff200 } } }, + [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00cf9a00 } } }, + [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9200 } } }, + [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00cffa00 } } }, + [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff200 } } }, /* * Segments used for calling PnP BIOS have byte granularity. * They code segments and data segments have fixed 64k limits, * the transfer segment sizes are set at run time. */ /* 32-bit code */ - [GDT_ENTRY_PNPBIOS_CS32] = { { { 0x0000ffff, 0x00409a00 } } }, + [GDT_ENTRY_PNPBIOS_CS32] = { { { 0x0000ffff, 0x00409a00 } } }, /* 16-bit code */ - [GDT_ENTRY_PNPBIOS_CS16] = { { { 0x0000ffff, 0x00009a00 } } }, + [GDT_ENTRY_PNPBIOS_CS16] = { { { 0x0000ffff, 0x00009a00 } } }, /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_DS] = { { { 0x0000ffff, 0x00009200 } } }, + [GDT_ENTRY_PNPBIOS_DS] = { { { 0x0000ffff, 0x00009200 } } }, /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_TS1] = { { { 0x00000000, 0x00009200 } } }, + [GDT_ENTRY_PNPBIOS_TS1] = { { { 0x00000000, 0x00009200 } } }, /* 16-bit data */ - [GDT_ENTRY_PNPBIOS_TS2] = { { { 0x00000000, 0x00009200 } } }, + [GDT_ENTRY_PNPBIOS_TS2] = { { { 0x00000000, 0x00009200 } } }, /* * The APM segments have byte granularity and their bases * are set at run time. All have 64k limits. */ /* 32-bit code */ - [GDT_ENTRY_APMBIOS_BASE] = { { { 0x0000ffff, 0x00409a00 } } }, + [GDT_ENTRY_APMBIOS_BASE] = { { { 0x0000ffff, 0x00409a00 } } }, /* 16-bit code */ - [GDT_ENTRY_APMBIOS_BASE+1] = { { { 0x0000ffff, 0x00009a00 } } }, + [GDT_ENTRY_APMBIOS_BASE+1] = { { { 0x0000ffff, 0x00009a00 } } }, /* data */ - [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } }, + [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } }, - [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } }, - [GDT_ENTRY_PERCPU] = { { { 0x0000ffff, 0x00cf9200 } } }, + [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } }, + [GDT_ENTRY_PERCPU] = { { { 0x0000ffff, 0x00cf9200 } } }, GDT_STACK_CANARY_INIT #endif } }; @@ -164,16 +164,17 @@ static inline int flag_is_changeable_p(u32 flag) * the CPUID. Add "volatile" to not allow gcc to * optimize the subsequent calls to this function. */ - asm volatile ("pushfl\n\t" - "pushfl\n\t" - "popl %0\n\t" - "movl %0,%1\n\t" - "xorl %2,%0\n\t" - "pushl %0\n\t" - "popfl\n\t" - "pushfl\n\t" - "popl %0\n\t" - "popfl\n\t" + asm volatile ("pushfl \n\t" + "pushfl \n\t" + "popl %0 \n\t" + "movl %0, %1 \n\t" + "xorl %2, %0 \n\t" + "pushl %0 \n\t" + "popfl \n\t" + "pushfl \n\t" + "popl %0 \n\t" + "popfl \n\t" + : "=&r" (f1), "=&r" (f2) : "ir" (flag)); @@ -188,18 +189,22 @@ static int __cpuinit have_cpuid_p(void) static void __cpuinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c) { - if (cpu_has(c, X86_FEATURE_PN) && disable_x86_serial_nr) { - /* Disable processor serial number */ - unsigned long lo, hi; - rdmsr(MSR_IA32_BBL_CR_CTL, lo, hi); - lo |= 0x200000; - wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi); - printk(KERN_NOTICE "CPU serial number disabled.\n"); - clear_cpu_cap(c, X86_FEATURE_PN); - - /* Disabling the serial number may affect the cpuid level */ - c->cpuid_level = cpuid_eax(0); - } + unsigned long lo, hi; + + if (!cpu_has(c, X86_FEATURE_PN) || !disable_x86_serial_nr) + return; + + /* Disable processor serial number: */ + + rdmsr(MSR_IA32_BBL_CR_CTL, lo, hi); + lo |= 0x200000; + wrmsr(MSR_IA32_BBL_CR_CTL, lo, hi); + + printk(KERN_NOTICE "CPU serial number disabled.\n"); + clear_cpu_cap(c, X86_FEATURE_PN); + + /* Disabling the serial number may affect the cpuid level */ + c->cpuid_level = cpuid_eax(0); } static int __init x86_serial_nr_setup(char *s) @@ -232,6 +237,7 @@ struct cpuid_dependent_feature { u32 feature; u32 level; }; + static const struct cpuid_dependent_feature __cpuinitconst cpuid_dependent_features[] = { { X86_FEATURE_MWAIT, 0x00000005 }, @@ -243,7 +249,11 @@ cpuid_dependent_features[] = { static void __cpuinit filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) { const struct cpuid_dependent_feature *df; + for (df = cpuid_dependent_features; df->feature; df++) { + + if (!cpu_has(c, df->feature)) + continue; /* * Note: cpuid_level is set to -1 if unavailable, but * extended_extended_level is set to 0 if unavailable @@ -251,32 +261,32 @@ static void __cpuinit filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) * when signed; hence the weird messing around with * signs here... */ - if (cpu_has(c, df->feature) && - ((s32)df->level < 0 ? + if (!((s32)df->level < 0 ? (u32)df->level > (u32)c->extended_cpuid_level : - (s32)df->level > (s32)c->cpuid_level)) { - clear_cpu_cap(c, df->feature); - if (warn) - printk(KERN_WARNING - "CPU: CPU feature %s disabled " - "due to lack of CPUID level 0x%x\n", - x86_cap_flags[df->feature], - df->level); - } + (s32)df->level > (s32)c->cpuid_level)) + continue; + + clear_cpu_cap(c, df->feature); + if (!warn) + continue; + + printk(KERN_WARNING + "CPU: CPU feature %s disabled, no CPUID level 0x%x\n", + x86_cap_flags[df->feature], df->level); } } /* * Naming convention should be: <Name> [(<Codename>)] * This table only is used unless init_<vendor>() below doesn't set it; - * in particular, if CPUID levels 0x80000002..4 are supported, this isn't used - * + * in particular, if CPUID levels 0x80000002..4 are supported, this + * isn't used */ /* Look up CPU names by table lookup. */ -static char __cpuinit *table_lookup_model(struct cpuinfo_x86 *c) +static const char *__cpuinit table_lookup_model(struct cpuinfo_x86 *c) { - struct cpu_model_info *info; + const struct cpu_model_info *info; if (c->x86_model >= 16) return NULL; /* Range check */ @@ -307,8 +317,10 @@ void load_percpu_segment(int cpu) load_stack_canary_segment(); } -/* Current gdt points %fs at the "master" per-cpu area: after this, - * it's on the real one. */ +/* + * Current gdt points %fs at the "master" per-cpu area: after this, + * it's on the real one. + */ void switch_to_new_gdt(int cpu) { struct desc_ptr gdt_descr; @@ -321,7 +333,7 @@ void switch_to_new_gdt(int cpu) load_percpu_segment(cpu); } -static struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {}; +static const struct cpu_dev *__cpuinitdata cpu_devs[X86_VENDOR_NUM] = {}; static void __cpuinit default_init(struct cpuinfo_x86 *c) { @@ -340,7 +352,7 @@ static void __cpuinit default_init(struct cpuinfo_x86 *c) #endif } -static struct cpu_dev __cpuinitdata default_cpu = { +static const struct cpu_dev __cpuinitconst default_cpu = { .c_init = default_init, .c_vendor = "Unknown", .c_x86_vendor = X86_VENDOR_UNKNOWN, @@ -354,22 +366,24 @@ static void __cpuinit get_model_name(struct cpuinfo_x86 *c) if (c->extended_cpuid_level < 0x80000004) return; - v = (unsigned int *) c->x86_model_id; + v = (unsigned int *)c->x86_model_id; cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]); cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]); cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]); c->x86_model_id[48] = 0; - /* Intel chips right-justify this string for some dumb reason; - undo that brain damage */ + /* + * Intel chips right-justify this string for some dumb reason; + * undo that brain damage: + */ p = q = &c->x86_model_id[0]; while (*p == ' ') - p++; + p++; if (p != q) { - while (*p) - *q++ = *p++; - while (q <= &c->x86_model_id[48]) - *q++ = '\0'; /* Zero-pad the rest */ + while (*p) + *q++ = *p++; + while (q <= &c->x86_model_id[48]) + *q++ = '\0'; /* Zero-pad the rest */ } } @@ -438,27 +452,30 @@ void __cpuinit detect_ht(struct cpuinfo_x86 *c) if (smp_num_siblings == 1) { printk(KERN_INFO "CPU: Hyper-Threading is disabled\n"); - } else if (smp_num_siblings > 1) { + goto out; + } - if (smp_num_siblings > nr_cpu_ids) { - printk(KERN_WARNING "CPU: Unsupported number of siblings %d", - smp_num_siblings); - smp_num_siblings = 1; - return; - } + if (smp_num_siblings <= 1) + goto out; + + if (smp_num_siblings > nr_cpu_ids) { + pr_warning("CPU: Unsupported number of siblings %d", + smp_num_siblings); + smp_num_siblings = 1; + return; + } - index_msb = get_count_order(smp_num_siblings); - c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, index_msb); + index_msb = get_count_order(smp_num_siblings); + c->phys_proc_id = apic->phys_pkg_id(c->initial_apicid, index_msb); - smp_num_siblings = smp_num_siblings / c->x86_max_cores; + smp_num_siblings = smp_num_siblings / c->x86_max_cores; - index_msb = get_count_order(smp_num_siblings); + index_msb = get_count_order(smp_num_siblings); - core_bits = get_count_order(c->x86_max_cores); + core_bits = get_count_order(c->x86_max_cores); - c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, index_msb) & - ((1 << core_bits) - 1); - } + c->cpu_core_id = apic->phys_pkg_id(c->initial_apicid, index_msb) & + ((1 << core_bits) - 1); out: if ((c->x86_max_cores * smp_num_siblings) > 1) { @@ -473,8 +490,8 @@ out: static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) { char *v = c->x86_vendor_id; - int i; static int printed; + int i; for (i = 0; i < X86_VENDOR_NUM; i++) { if (!cpu_devs[i]) @@ -483,6 +500,7 @@ static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) if (!strcmp(v, cpu_devs[i]->c_ident[0]) || (cpu_devs[i]->c_ident[1] && !strcmp(v, cpu_devs[i]->c_ident[1]))) { + this_cpu = cpu_devs[i]; c->x86_vendor = this_cpu->c_x86_vendor; return; @@ -491,7 +509,9 @@ static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) if (!printed) { printed++; - printk(KERN_ERR "CPU: vendor_id '%s' unknown, using generic init.\n", v); + printk(KERN_ERR + "CPU: vendor_id '%s' unknown, using generic init.\n", v); + printk(KERN_ERR "CPU: Your system may be unstable.\n"); } @@ -511,14 +531,17 @@ void __cpuinit cpu_detect(struct cpuinfo_x86 *c) /* Intel-defined flags: level 0x00000001 */ if (c->cpuid_level >= 0x00000001) { u32 junk, tfms, cap0, misc; + cpuid(0x00000001, &tfms, &misc, &junk, &cap0); c->x86 = (tfms >> 8) & 0xf; c->x86_model = (tfms >> 4) & 0xf; c->x86_mask = tfms & 0xf; + if (c->x86 == 0xf) c->x86 += (tfms >> 20) & 0xff; if (c->x86 >= 0x6) c->x86_model += ((tfms >> 16) & 0xf) << 4; + if (cap0 & (1<<19)) { c->x86_clflush_size = ((misc >> 8) & 0xff) * 8; c->x86_cache_alignment = c->x86_clflush_size; @@ -534,6 +557,7 @@ static void __cpuinit get_cpu_cap(struct cpuinfo_x86 *c) /* Intel-defined flags: level 0x00000001 */ if (c->cpuid_level >= 0x00000001) { u32 capability, excap; + cpuid(0x00000001, &tfms, &ebx, &excap, &capability); c->x86_capability[0] = capability; c->x86_capability[4] = excap; @@ -542,6 +566,7 @@ static void __cpuinit get_cpu_cap(struct cpuinfo_x86 *c) /* AMD-defined flags: level 0x80000001 */ xlvl = cpuid_eax(0x80000000); c->extended_cpuid_level = xlvl; + if ((xlvl & 0xffff0000) == 0x80000000) { if (xlvl >= 0x80000001) { c->x86_capability[1] = cpuid_edx(0x80000001); @@ -549,13 +574,15 @@ static void __cpuinit get_cpu_cap(struct cpuinfo_x86 *c) } } -#ifdef CONFIG_X86_64 if (c->extended_cpuid_level >= 0x80000008) { u32 eax = cpuid_eax(0x80000008); c->x86_virt_bits = (eax >> 8) & 0xff; c->x86_phys_bits = eax & 0xff; } +#ifdef CONFIG_X86_32 + else if (cpu_has(c, X86_FEATURE_PAE) || cpu_has(c, X86_FEATURE_PSE36)) + c->x86_phys_bits = 36; #endif if (c->extended_cpuid_level >= 0x80000007) @@ -602,8 +629,12 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_64 c->x86_clflush_size = 64; + c->x86_phys_bits = 36; + c->x86_virt_bits = 48; #else c->x86_clflush_size = 32; + c->x86_phys_bits = 32; + c->x86_virt_bits = 32; #endif c->x86_cache_alignment = c->x86_clflush_size; @@ -634,12 +665,12 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c) void __init early_cpu_init(void) { - struct cpu_dev **cdev; + const struct cpu_dev *const *cdev; int count = 0; - printk("KERNEL supported cpus:\n"); + printk(KERN_INFO "KERNEL supported cpus:\n"); for (cdev = __x86_cpu_dev_start; cdev < __x86_cpu_dev_end; cdev++) { - struct cpu_dev *cpudev = *cdev; + const struct cpu_dev *cpudev = *cdev; unsigned int j; if (count >= X86_VENDOR_NUM) @@ -650,7 +681,7 @@ void __init early_cpu_init(void) for (j = 0; j < 2; j++) { if (!cpudev->c_ident[j]) continue; - printk(" %s %s\n", cpudev->c_vendor, + printk(KERN_INFO " %s %s\n", cpudev->c_vendor, cpudev->c_ident[j]); } } @@ -726,9 +757,13 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) c->x86_coreid_bits = 0; #ifdef CONFIG_X86_64 c->x86_clflush_size = 64; + c->x86_phys_bits = 36; + c->x86_virt_bits = 48; #else c->cpuid_level = -1; /* CPUID not detected */ c->x86_clflush_size = 32; + c->x86_phys_bits = 32; + c->x86_virt_bits = 32; #endif c->x86_cache_alignment = c->x86_clflush_size; memset(&c->x86_capability, 0, sizeof c->x86_capability); @@ -759,8 +794,8 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) squash_the_stupid_serial_number(c); /* - * The vendor-specific functions might have changed features. Now - * we do "generic changes." + * The vendor-specific functions might have changed features. + * Now we do "generic changes." */ /* Filter out anything that depends on CPUID levels we don't have */ @@ -768,7 +803,7 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) /* If the model name is still unset, do table lookup. */ if (!c->x86_model_id[0]) { - char *p; + const char *p; p = table_lookup_model(c); if (p) strcpy(c->x86_model_id, p); @@ -843,11 +878,11 @@ void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c) } struct msr_range { - unsigned min; - unsigned max; + unsigned min; + unsigned max; }; -static struct msr_range msr_range_array[] __cpuinitdata = { +static const struct msr_range msr_range_array[] __cpuinitconst = { { 0x00000000, 0x00000418}, { 0xc0000000, 0xc000040b}, { 0xc0010000, 0xc0010142}, @@ -856,14 +891,15 @@ static struct msr_range msr_range_array[] __cpuinitdata = { static void __cpuinit print_cpu_msr(void) { + unsigned index_min, index_max; unsigned index; u64 val; int i; - unsigned index_min, index_max; for (i = 0; i < ARRAY_SIZE(msr_range_array); i++) { index_min = msr_range_array[i].min; index_max = msr_range_array[i].max; + for (index = index_min; index < index_max; index++) { if (rdmsrl_amd_safe(index, &val)) continue; @@ -873,6 +909,7 @@ static void __cpuinit print_cpu_msr(void) } static int show_msr __cpuinitdata; + static __init int setup_show_msr(char *arg) { int num; @@ -894,12 +931,14 @@ __setup("noclflush", setup_noclflush); void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) { - char *vendor = NULL; + const char *vendor = NULL; - if (c->x86_vendor < X86_VENDOR_NUM) + if (c->x86_vendor < X86_VENDOR_NUM) { vendor = this_cpu->c_vendor; - else if (c->cpuid_level >= 0) - vendor = c->x86_vendor_id; + } else { + if (c->cpuid_level >= 0) + vendor = c->x86_vendor_id; + } if (vendor && !strstr(c->x86_model_id, vendor)) printk(KERN_CONT "%s ", vendor); @@ -926,10 +965,12 @@ void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) static __init int setup_disablecpuid(char *arg) { int bit; + if (get_option(&arg, &bit) && bit < NCAPINTS*32) setup_clear_cpu_cap(bit); else return 0; + return 1; } __setup("clearcpuid=", setup_disablecpuid); @@ -939,6 +980,7 @@ struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table }; DEFINE_PER_CPU_FIRST(union irq_stack_union, irq_stack_union) __aligned(PAGE_SIZE); + DEFINE_PER_CPU(char *, irq_stack_ptr) = init_per_cpu_var(irq_stack_union.irq_stack) + IRQ_STACK_SIZE - 64; @@ -948,12 +990,21 @@ EXPORT_PER_CPU_SYMBOL(kernel_stack); DEFINE_PER_CPU(unsigned int, irq_count) = -1; +/* + * Special IST stacks which the CPU switches to when it calls + * an IST-marked descriptor entry. Up to 7 stacks (hardware + * limit), all of them are 4K, except the debug stack which + * is 8K. + */ +static const unsigned int exception_stack_sizes[N_EXCEPTION_STACKS] = { + [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ, + [DEBUG_STACK - 1] = DEBUG_STKSZ +}; + static DEFINE_PER_CPU_PAGE_ALIGNED(char, exception_stacks [(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ]) __aligned(PAGE_SIZE); -extern asmlinkage void ignore_sysret(void); - /* May not be marked __init: used by software suspend */ void syscall_init(void) { @@ -983,7 +1034,7 @@ unsigned long kernel_eflags; */ DEFINE_PER_CPU(struct orig_ist, orig_ist); -#else /* x86_64 */ +#else /* CONFIG_X86_64 */ #ifdef CONFIG_CC_STACKPROTECTOR DEFINE_PER_CPU(unsigned long, stack_canary); @@ -995,9 +1046,26 @@ struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs) memset(regs, 0, sizeof(struct pt_regs)); regs->fs = __KERNEL_PERCPU; regs->gs = __KERNEL_STACK_CANARY; + return regs; } -#endif /* x86_64 */ +#endif /* CONFIG_X86_64 */ + +/* + * Clear all 6 debug registers: + */ +static void clear_all_debug_regs(void) +{ + int i; + + for (i = 0; i < 8; i++) { + /* Ignore db4, db5 */ + if ((i == 4) || (i == 5)) + continue; + + set_debugreg(0, i); + } +} /* * cpu_init() initializes state that is per-CPU. Some data is already @@ -1007,15 +1075,20 @@ struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs) * A lot of state is already set up in PDA init for 64 bit */ #ifdef CONFIG_X86_64 + void __cpuinit cpu_init(void) { - int cpu = stack_smp_processor_id(); - struct tss_struct *t = &per_cpu(init_tss, cpu); - struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu); - unsigned long v; + struct orig_ist *orig_ist; struct task_struct *me; + struct tss_struct *t; + unsigned long v; + int cpu; int i; + cpu = stack_smp_processor_id(); + t = &per_cpu(init_tss, cpu); + orig_ist = &per_cpu(orig_ist, cpu); + #ifdef CONFIG_NUMA if (cpu != 0 && percpu_read(node_number) == 0 && cpu_to_node(cpu) != NUMA_NO_NODE) @@ -1056,19 +1129,17 @@ void __cpuinit cpu_init(void) * set up and load the per-CPU TSS */ if (!orig_ist->ist[0]) { - static const unsigned int sizes[N_EXCEPTION_STACKS] = { - [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ, - [DEBUG_STACK - 1] = DEBUG_STKSZ - }; char *estacks = per_cpu(exception_stacks, cpu); + for (v = 0; v < N_EXCEPTION_STACKS; v++) { - estacks += sizes[v]; + estacks += exception_stack_sizes[v]; orig_ist->ist[v] = t->x86_tss.ist[v] = (unsigned long)estacks; } } t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap); + /* * <= is required because the CPU will access up to * 8 bits beyond the end of the IO permission bitmap. @@ -1078,8 +1149,7 @@ void __cpuinit cpu_init(void) atomic_inc(&init_mm.mm_count); me->active_mm = &init_mm; - if (me->mm) - BUG(); + BUG_ON(me->mm); enter_lazy_tlb(&init_mm, me); load_sp0(t, ¤t->thread); @@ -1098,17 +1168,7 @@ void __cpuinit cpu_init(void) arch_kgdb_ops.correct_hw_break(); else #endif - { - /* - * Clear all 6 debug registers: - */ - set_debugreg(0UL, 0); - set_debugreg(0UL, 1); - set_debugreg(0UL, 2); - set_debugreg(0UL, 3); - set_debugreg(0UL, 6); - set_debugreg(0UL, 7); - } + clear_all_debug_regs(); fpu_init(); @@ -1129,7 +1189,8 @@ void __cpuinit cpu_init(void) if (cpumask_test_and_set_cpu(cpu, cpu_initialized_mask)) { printk(KERN_WARNING "CPU#%d already initialized!\n", cpu); - for (;;) local_irq_enable(); + for (;;) + local_irq_enable(); } printk(KERN_INFO "Initializing CPU#%d\n", cpu); @@ -1145,8 +1206,7 @@ void __cpuinit cpu_init(void) */ atomic_inc(&init_mm.mm_count); curr->active_mm = &init_mm; - if (curr->mm) - BUG(); + BUG_ON(curr->mm); enter_lazy_tlb(&init_mm, curr); load_sp0(t, thread); @@ -1159,13 +1219,7 @@ void __cpuinit cpu_init(void) __set_tss_desc(cpu, GDT_ENTRY_DOUBLEFAULT_TSS, &doublefault_tss); #endif - /* Clear all 6 debug registers: */ - set_debugreg(0, 0); - set_debugreg(0, 1); - set_debugreg(0, 2); - set_debugreg(0, 3); - set_debugreg(0, 6); - set_debugreg(0, 7); + clear_all_debug_regs(); /* * Force FPU initialization: @@ -1185,6 +1239,4 @@ void __cpuinit cpu_init(void) xsave_init(); } - - #endif diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index de4094a39210..6de9a908e400 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -3,33 +3,34 @@ #define ARCH_X86_CPU_H struct cpu_model_info { - int vendor; - int family; - char *model_names[16]; + int vendor; + int family; + const char *model_names[16]; }; /* attempt to consolidate cpu attributes */ struct cpu_dev { - char * c_vendor; + const char *c_vendor; /* some have two possibilities for cpuid string */ - char * c_ident[2]; + const char *c_ident[2]; struct cpu_model_info c_models[4]; - void (*c_early_init)(struct cpuinfo_x86 *c); - void (*c_init)(struct cpuinfo_x86 * c); - void (*c_identify)(struct cpuinfo_x86 * c); - unsigned int (*c_size_cache)(struct cpuinfo_x86 * c, unsigned int size); - int c_x86_vendor; + void (*c_early_init)(struct cpuinfo_x86 *); + void (*c_init)(struct cpuinfo_x86 *); + void (*c_identify)(struct cpuinfo_x86 *); + unsigned int (*c_size_cache)(struct cpuinfo_x86 *, unsigned int); + int c_x86_vendor; }; #define cpu_dev_register(cpu_devX) \ - static struct cpu_dev *__cpu_dev_##cpu_devX __used \ + static const struct cpu_dev *const __cpu_dev_##cpu_devX __used \ __attribute__((__section__(".x86_cpu_dev.init"))) = \ &cpu_devX; -extern struct cpu_dev *__x86_cpu_dev_start[], *__x86_cpu_dev_end[]; +extern const struct cpu_dev *const __x86_cpu_dev_start[], + *const __x86_cpu_dev_end[]; extern void display_cacheinfo(struct cpuinfo_x86 *c); diff --git a/arch/x86/kernel/cpu/cpu_debug.c b/arch/x86/kernel/cpu/cpu_debug.c new file mode 100755 index 000000000000..46e29ab96c6a --- /dev/null +++ b/arch/x86/kernel/cpu/cpu_debug.c @@ -0,0 +1,901 @@ +/* + * CPU x86 architecture debug code + * + * Copyright(C) 2009 Jaswinder Singh Rajput + * + * For licencing details see kernel-base/COPYING + */ + +#include <linux/interrupt.h> +#include <linux/compiler.h> +#include <linux/seq_file.h> +#include <linux/debugfs.h> +#include <linux/kprobes.h> +#include <linux/uaccess.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/percpu.h> +#include <linux/signal.h> +#include <linux/errno.h> +#include <linux/sched.h> +#include <linux/types.h> +#include <linux/init.h> +#include <linux/slab.h> +#include <linux/smp.h> + +#include <asm/cpu_debug.h> +#include <asm/paravirt.h> +#include <asm/system.h> +#include <asm/traps.h> +#include <asm/apic.h> +#include <asm/desc.h> + +static DEFINE_PER_CPU(struct cpu_cpuX_base, cpu_arr[CPU_REG_ALL_BIT]); +static DEFINE_PER_CPU(struct cpu_private *, priv_arr[MAX_CPU_FILES]); +static DEFINE_PER_CPU(unsigned, cpu_modelflag); +static DEFINE_PER_CPU(int, cpu_priv_count); +static DEFINE_PER_CPU(unsigned, cpu_model); + +static DEFINE_MUTEX(cpu_debug_lock); + +static struct dentry *cpu_debugfs_dir; + +static struct cpu_debug_base cpu_base[] = { + { "mc", CPU_MC, 0 }, + { "monitor", CPU_MONITOR, 0 }, + { "time", CPU_TIME, 0 }, + { "pmc", CPU_PMC, 1 }, + { "platform", CPU_PLATFORM, 0 }, + { "apic", CPU_APIC, 0 }, + { "poweron", CPU_POWERON, 0 }, + { "control", CPU_CONTROL, 0 }, + { "features", CPU_FEATURES, 0 }, + { "lastbranch", CPU_LBRANCH, 0 }, + { "bios", CPU_BIOS, 0 }, + { "freq", CPU_FREQ, 0 }, + { "mtrr", CPU_MTRR, 0 }, + { "perf", CPU_PERF, 0 }, + { "cache", CPU_CACHE, 0 }, + { "sysenter", CPU_SYSENTER, 0 }, + { "therm", CPU_THERM, 0 }, + { "misc", CPU_MISC, 0 }, + { "debug", CPU_DEBUG, 0 }, + { "pat", CPU_PAT, 0 }, + { "vmx", CPU_VMX, 0 }, + { "call", CPU_CALL, 0 }, + { "base", CPU_BASE, 0 }, + { "ver", CPU_VER, 0 }, + { "conf", CPU_CONF, 0 }, + { "smm", CPU_SMM, 0 }, + { "svm", CPU_SVM, 0 }, + { "osvm", CPU_OSVM, 0 }, + { "tss", CPU_TSS, 0 }, + { "cr", CPU_CR, 0 }, + { "dt", CPU_DT, 0 }, + { "registers", CPU_REG_ALL, 0 }, +}; + +static struct cpu_file_base cpu_file[] = { + { "index", CPU_REG_ALL, 0 }, + { "value", CPU_REG_ALL, 1 }, +}; + +/* Intel Registers Range */ +static struct cpu_debug_range cpu_intel_range[] = { + { 0x00000000, 0x00000001, CPU_MC, CPU_INTEL_ALL }, + { 0x00000006, 0x00000007, CPU_MONITOR, CPU_CX_AT_XE }, + { 0x00000010, 0x00000010, CPU_TIME, CPU_INTEL_ALL }, + { 0x00000011, 0x00000013, CPU_PMC, CPU_INTEL_PENTIUM }, + { 0x00000017, 0x00000017, CPU_PLATFORM, CPU_PX_CX_AT_XE }, + { 0x0000001B, 0x0000001B, CPU_APIC, CPU_P6_CX_AT_XE }, + + { 0x0000002A, 0x0000002A, CPU_POWERON, CPU_PX_CX_AT_XE }, + { 0x0000002B, 0x0000002B, CPU_POWERON, CPU_INTEL_XEON }, + { 0x0000002C, 0x0000002C, CPU_FREQ, CPU_INTEL_XEON }, + { 0x0000003A, 0x0000003A, CPU_CONTROL, CPU_CX_AT_XE }, + + { 0x00000040, 0x00000043, CPU_LBRANCH, CPU_PM_CX_AT_XE }, + { 0x00000044, 0x00000047, CPU_LBRANCH, CPU_PM_CO_AT }, + { 0x00000060, 0x00000063, CPU_LBRANCH, CPU_C2_AT }, + { 0x00000064, 0x00000067, CPU_LBRANCH, CPU_INTEL_ATOM }, + + { 0x00000079, 0x00000079, CPU_BIOS, CPU_P6_CX_AT_XE }, + { 0x00000088, 0x0000008A, CPU_CACHE, CPU_INTEL_P6 }, + { 0x0000008B, 0x0000008B, CPU_BIOS, CPU_P6_CX_AT_XE }, + { 0x0000009B, 0x0000009B, CPU_MONITOR, CPU_INTEL_XEON }, + + { 0x000000C1, 0x000000C2, CPU_PMC, CPU_P6_CX_AT }, + { 0x000000CD, 0x000000CD, CPU_FREQ, CPU_CX_AT }, + { 0x000000E7, 0x000000E8, CPU_PERF, CPU_CX_AT }, + { 0x000000FE, 0x000000FE, CPU_MTRR, CPU_P6_CX_XE }, + + { 0x00000116, 0x00000116, CPU_CACHE, CPU_INTEL_P6 }, + { 0x00000118, 0x00000118, CPU_CACHE, CPU_INTEL_P6 }, + { 0x00000119, 0x00000119, CPU_CACHE, CPU_INTEL_PX }, + { 0x0000011A, 0x0000011B, CPU_CACHE, CPU_INTEL_P6 }, + { 0x0000011E, 0x0000011E, CPU_CACHE, CPU_PX_CX_AT }, + + { 0x00000174, 0x00000176, CPU_SYSENTER, CPU_P6_CX_AT_XE }, + { 0x00000179, 0x0000017A, CPU_MC, CPU_PX_CX_AT_XE }, + { 0x0000017B, 0x0000017B, CPU_MC, CPU_P6_XE }, + { 0x00000186, 0x00000187, CPU_PMC, CPU_P6_CX_AT }, + { 0x00000198, 0x00000199, CPU_PERF, CPU_PM_CX_AT_XE }, + { 0x0000019A, 0x0000019A, CPU_TIME, CPU_PM_CX_AT_XE }, + { 0x0000019B, 0x0000019D, CPU_THERM, CPU_PM_CX_AT_XE }, + { 0x000001A0, 0x000001A0, CPU_MISC, CPU_PM_CX_AT_XE }, + + { 0x000001C9, 0x000001C9, CPU_LBRANCH, CPU_PM_CX_AT }, + { 0x000001D7, 0x000001D8, CPU_LBRANCH, CPU_INTEL_XEON }, + { 0x000001D9, 0x000001D9, CPU_DEBUG, CPU_CX_AT_XE }, + { 0x000001DA, 0x000001DA, CPU_LBRANCH, CPU_INTEL_XEON }, + { 0x000001DB, 0x000001DB, CPU_LBRANCH, CPU_P6_XE }, + { 0x000001DC, 0x000001DC, CPU_LBRANCH, CPU_INTEL_P6 }, + { 0x000001DD, 0x000001DE, CPU_LBRANCH, CPU_PX_CX_AT_XE }, + { 0x000001E0, 0x000001E0, CPU_LBRANCH, CPU_INTEL_P6 }, + + { 0x00000200, 0x0000020F, CPU_MTRR, CPU_P6_CX_XE }, + { 0x00000250, 0x00000250, CPU_MTRR, CPU_P6_CX_XE }, + { 0x00000258, 0x00000259, CPU_MTRR, CPU_P6_CX_XE }, + { 0x00000268, 0x0000026F, CPU_MTRR, CPU_P6_CX_XE }, + { 0x00000277, 0x00000277, CPU_PAT, CPU_C2_AT_XE }, + { 0x000002FF, 0x000002FF, CPU_MTRR, CPU_P6_CX_XE }, + + { 0x00000300, 0x00000308, CPU_PMC, CPU_INTEL_XEON }, + { 0x00000309, 0x0000030B, CPU_PMC, CPU_C2_AT_XE }, + { 0x0000030C, 0x00000311, CPU_PMC, CPU_INTEL_XEON }, + { 0x00000345, 0x00000345, CPU_PMC, CPU_C2_AT }, + { 0x00000360, 0x00000371, CPU_PMC, CPU_INTEL_XEON }, + { 0x0000038D, 0x00000390, CPU_PMC, CPU_C2_AT }, + { 0x000003A0, 0x000003BE, CPU_PMC, CPU_INTEL_XEON }, + { 0x000003C0, 0x000003CD, CPU_PMC, CPU_INTEL_XEON }, + { 0x000003E0, 0x000003E1, CPU_PMC, CPU_INTEL_XEON }, + { 0x000003F0, 0x000003F0, CPU_PMC, CPU_INTEL_XEON }, + { 0x000003F1, 0x000003F1, CPU_PMC, CPU_C2_AT_XE }, + { 0x000003F2, 0x000003F2, CPU_PMC, CPU_INTEL_XEON }, + + { 0x00000400, 0x00000402, CPU_MC, CPU_PM_CX_AT_XE }, + { 0x00000403, 0x00000403, CPU_MC, CPU_INTEL_XEON }, + { 0x00000404, 0x00000406, CPU_MC, CPU_PM_CX_AT_XE }, + { 0x00000407, 0x00000407, CPU_MC, CPU_INTEL_XEON }, + { 0x00000408, 0x0000040A, CPU_MC, CPU_PM_CX_AT_XE }, + { 0x0000040B, 0x0000040B, CPU_MC, CPU_INTEL_XEON }, + { 0x0000040C, 0x0000040E, CPU_MC, CPU_PM_CX_XE }, + { 0x0000040F, 0x0000040F, CPU_MC, CPU_INTEL_XEON }, + { 0x00000410, 0x00000412, CPU_MC, CPU_PM_CX_AT_XE }, + { 0x00000413, 0x00000417, CPU_MC, CPU_CX_AT_XE }, + { 0x00000480, 0x0000048B, CPU_VMX, CPU_CX_AT_XE }, + + { 0x00000600, 0x00000600, CPU_DEBUG, CPU_PM_CX_AT_XE }, + { 0x00000680, 0x0000068F, CPU_LBRANCH, CPU_INTEL_XEON }, + { 0x000006C0, 0x000006CF, CPU_LBRANCH, CPU_INTEL_XEON }, + + { 0x000107CC, 0x000107D3, CPU_PMC, CPU_INTEL_XEON_MP }, + + { 0xC0000080, 0xC0000080, CPU_FEATURES, CPU_INTEL_XEON }, + { 0xC0000081, 0xC0000082, CPU_CALL, CPU_INTEL_XEON }, + { 0xC0000084, 0xC0000084, CPU_CALL, CPU_INTEL_XEON }, + { 0xC0000100, 0xC0000102, CPU_BASE, CPU_INTEL_XEON }, +}; + +/* AMD Registers Range */ +static struct cpu_debug_range cpu_amd_range[] = { + { 0x00000000, 0x00000001, CPU_MC, CPU_K10_PLUS, }, + { 0x00000010, 0x00000010, CPU_TIME, CPU_K8_PLUS, }, + { 0x0000001B, 0x0000001B, CPU_APIC, CPU_K8_PLUS, }, + { 0x0000002A, 0x0000002A, CPU_POWERON, CPU_K7_PLUS }, + { 0x0000008B, 0x0000008B, CPU_VER, CPU_K8_PLUS }, + { 0x000000FE, 0x000000FE, CPU_MTRR, CPU_K8_PLUS, }, + + { 0x00000174, 0x00000176, CPU_SYSENTER, CPU_K8_PLUS, }, + { 0x00000179, 0x0000017B, CPU_MC, CPU_K8_PLUS, }, + { 0x000001D9, 0x000001D9, CPU_DEBUG, CPU_K8_PLUS, }, + { 0x000001DB, 0x000001DE, CPU_LBRANCH, CPU_K8_PLUS, }, + + { 0x00000200, 0x0000020F, CPU_MTRR, CPU_K8_PLUS, }, + { 0x00000250, 0x00000250, CPU_MTRR, CPU_K8_PLUS, }, + { 0x00000258, 0x00000259, CPU_MTRR, CPU_K8_PLUS, }, + { 0x00000268, 0x0000026F, CPU_MTRR, CPU_K8_PLUS, }, + { 0x00000277, 0x00000277, CPU_PAT, CPU_K8_PLUS, }, + { 0x000002FF, 0x000002FF, CPU_MTRR, CPU_K8_PLUS, }, + + { 0x00000400, 0x00000413, CPU_MC, CPU_K8_PLUS, }, + + { 0xC0000080, 0xC0000080, CPU_FEATURES, CPU_AMD_ALL, }, + { 0xC0000081, 0xC0000084, CPU_CALL, CPU_K8_PLUS, }, + { 0xC0000100, 0xC0000102, CPU_BASE, CPU_K8_PLUS, }, + { 0xC0000103, 0xC0000103, CPU_TIME, CPU_K10_PLUS, }, + + { 0xC0010000, 0xC0010007, CPU_PMC, CPU_K8_PLUS, }, + { 0xC0010010, 0xC0010010, CPU_CONF, CPU_K7_PLUS, }, + { 0xC0010015, 0xC0010015, CPU_CONF, CPU_K7_PLUS, }, + { 0xC0010016, 0xC001001A, CPU_MTRR, CPU_K8_PLUS, }, + { 0xC001001D, 0xC001001D, CPU_MTRR, CPU_K8_PLUS, }, + { 0xC001001F, 0xC001001F, CPU_CONF, CPU_K8_PLUS, }, + { 0xC0010030, 0xC0010035, CPU_BIOS, CPU_K8_PLUS, }, + { 0xC0010044, 0xC0010048, CPU_MC, CPU_K8_PLUS, }, + { 0xC0010050, 0xC0010056, CPU_SMM, CPU_K0F_PLUS, }, + { 0xC0010058, 0xC0010058, CPU_CONF, CPU_K10_PLUS, }, + { 0xC0010060, 0xC0010060, CPU_CACHE, CPU_AMD_11, }, + { 0xC0010061, 0xC0010068, CPU_SMM, CPU_K10_PLUS, }, + { 0xC0010069, 0xC001006B, CPU_SMM, CPU_AMD_11, }, + { 0xC0010070, 0xC0010071, CPU_SMM, CPU_K10_PLUS, }, + { 0xC0010111, 0xC0010113, CPU_SMM, CPU_K8_PLUS, }, + { 0xC0010114, 0xC0010118, CPU_SVM, CPU_K10_PLUS, }, + { 0xC0010140, 0xC0010141, CPU_OSVM, CPU_K10_PLUS, }, + { 0xC0011022, 0xC0011023, CPU_CONF, CPU_K10_PLUS, }, +}; + + +/* Intel */ +static int get_intel_modelflag(unsigned model) +{ + int flag; + + switch (model) { + case 0x0501: + case 0x0502: + case 0x0504: + flag = CPU_INTEL_PENTIUM; + break; + case 0x0601: + case 0x0603: + case 0x0605: + case 0x0607: + case 0x0608: + case 0x060A: + case 0x060B: + flag = CPU_INTEL_P6; + break; + case 0x0609: + case 0x060D: + flag = CPU_INTEL_PENTIUM_M; + break; + case 0x060E: + flag = CPU_INTEL_CORE; + break; + case 0x060F: + case 0x0617: + flag = CPU_INTEL_CORE2; + break; + case 0x061C: + flag = CPU_INTEL_ATOM; + break; + case 0x0F00: + case 0x0F01: + case 0x0F02: + case 0x0F03: + case 0x0F04: + flag = CPU_INTEL_XEON_P4; + break; + case 0x0F06: + flag = CPU_INTEL_XEON_MP; + break; + default: + flag = CPU_NONE; + break; + } + + return flag; +} + +/* AMD */ +static int get_amd_modelflag(unsigned model) +{ + int flag; + + switch (model >> 8) { + case 0x6: + flag = CPU_AMD_K6; + break; + case 0x7: + flag = CPU_AMD_K7; + break; + case 0x8: + flag = CPU_AMD_K8; + break; + case 0xf: + flag = CPU_AMD_0F; + break; + case 0x10: + flag = CPU_AMD_10; + break; + case 0x11: + flag = CPU_AMD_11; + break; + default: + flag = CPU_NONE; + break; + } + + return flag; +} + +static int get_cpu_modelflag(unsigned cpu) +{ + int flag; + + flag = per_cpu(cpu_model, cpu); + + switch (flag >> 16) { + case X86_VENDOR_INTEL: + flag = get_intel_modelflag(flag); + break; + case X86_VENDOR_AMD: + flag = get_amd_modelflag(flag & 0xffff); + break; + default: + flag = CPU_NONE; + break; + } + + return flag; +} + +static int get_cpu_range_count(unsigned cpu) +{ + int index; + + switch (per_cpu(cpu_model, cpu) >> 16) { + case X86_VENDOR_INTEL: + index = ARRAY_SIZE(cpu_intel_range); + break; + case X86_VENDOR_AMD: + index = ARRAY_SIZE(cpu_amd_range); + break; + default: + index = 0; + break; + } + + return index; +} + +static int is_typeflag_valid(unsigned cpu, unsigned flag) +{ + unsigned vendor, modelflag; + int i, index; + + /* Standard Registers should be always valid */ + if (flag >= CPU_TSS) + return 1; + + modelflag = per_cpu(cpu_modelflag, cpu); + vendor = per_cpu(cpu_model, cpu) >> 16; + index = get_cpu_range_count(cpu); + + for (i = 0; i < index; i++) { + switch (vendor) { + case X86_VENDOR_INTEL: + if ((cpu_intel_range[i].model & modelflag) && + (cpu_intel_range[i].flag & flag)) + return 1; + break; + case X86_VENDOR_AMD: + if ((cpu_amd_range[i].model & modelflag) && + (cpu_amd_range[i].flag & flag)) + return 1; + break; + } + } + + /* Invalid */ + return 0; +} + +static unsigned get_cpu_range(unsigned cpu, unsigned *min, unsigned *max, + int index, unsigned flag) +{ + unsigned modelflag; + + modelflag = per_cpu(cpu_modelflag, cpu); + *max = 0; + switch (per_cpu(cpu_model, cpu) >> 16) { + case X86_VENDOR_INTEL: + if ((cpu_intel_range[index].model & modelflag) && + (cpu_intel_range[index].flag & flag)) { + *min = cpu_intel_range[index].min; + *max = cpu_intel_range[index].max; + } + break; + case X86_VENDOR_AMD: + if ((cpu_amd_range[index].model & modelflag) && + (cpu_amd_range[index].flag & flag)) { + *min = cpu_amd_range[index].min; + *max = cpu_amd_range[index].max; + } + break; + } + + return *max; +} + +/* This function can also be called with seq = NULL for printk */ +static void print_cpu_data(struct seq_file *seq, unsigned type, + u32 low, u32 high) +{ + struct cpu_private *priv; + u64 val = high; + + if (seq) { + priv = seq->private; + if (priv->file) { + val = (val << 32) | low; + seq_printf(seq, "0x%llx\n", val); + } else + seq_printf(seq, " %08x: %08x_%08x\n", + type, high, low); + } else + printk(KERN_INFO " %08x: %08x_%08x\n", type, high, low); +} + +/* This function can also be called with seq = NULL for printk */ +static void print_msr(struct seq_file *seq, unsigned cpu, unsigned flag) +{ + unsigned msr, msr_min, msr_max; + struct cpu_private *priv; + u32 low, high; + int i, range; + + if (seq) { + priv = seq->private; + if (priv->file) { + if (!rdmsr_safe_on_cpu(priv->cpu, priv->reg, + &low, &high)) + print_cpu_data(seq, priv->reg, low, high); + return; + } + } + + range = get_cpu_range_count(cpu); + + for (i = 0; i < range; i++) { + if (!get_cpu_range(cpu, &msr_min, &msr_max, i, flag)) + continue; + + for (msr = msr_min; msr <= msr_max; msr++) { + if (rdmsr_safe_on_cpu(cpu, msr, &low, &high)) + continue; + print_cpu_data(seq, msr, low, high); + } + } +} + +static void print_tss(void *arg) +{ + struct pt_regs *regs = task_pt_regs(current); + struct seq_file *seq = arg; + unsigned int seg; + + seq_printf(seq, " RAX\t: %016lx\n", regs->ax); + seq_printf(seq, " RBX\t: %016lx\n", regs->bx); + seq_printf(seq, " RCX\t: %016lx\n", regs->cx); + seq_printf(seq, " RDX\t: %016lx\n", regs->dx); + + seq_printf(seq, " RSI\t: %016lx\n", regs->si); + seq_printf(seq, " RDI\t: %016lx\n", regs->di); + seq_printf(seq, " RBP\t: %016lx\n", regs->bp); + seq_printf(seq, " ESP\t: %016lx\n", regs->sp); + +#ifdef CONFIG_X86_64 + seq_printf(seq, " R08\t: %016lx\n", regs->r8); + seq_printf(seq, " R09\t: %016lx\n", regs->r9); + seq_printf(seq, " R10\t: %016lx\n", regs->r10); + seq_printf(seq, " R11\t: %016lx\n", regs->r11); + seq_printf(seq, " R12\t: %016lx\n", regs->r12); + seq_printf(seq, " R13\t: %016lx\n", regs->r13); + seq_printf(seq, " R14\t: %016lx\n", regs->r14); + seq_printf(seq, " R15\t: %016lx\n", regs->r15); +#endif + + asm("movl %%cs,%0" : "=r" (seg)); + seq_printf(seq, " CS\t: %04x\n", seg); + asm("movl %%ds,%0" : "=r" (seg)); + seq_printf(seq, " DS\t: %04x\n", seg); + seq_printf(seq, " SS\t: %04lx\n", regs->ss & 0xffff); + asm("movl %%es,%0" : "=r" (seg)); + seq_printf(seq, " ES\t: %04x\n", seg); + asm("movl %%fs,%0" : "=r" (seg)); + seq_printf(seq, " FS\t: %04x\n", seg); + asm("movl %%gs,%0" : "=r" (seg)); + seq_printf(seq, " GS\t: %04x\n", seg); + + seq_printf(seq, " EFLAGS\t: %016lx\n", regs->flags); + + seq_printf(seq, " EIP\t: %016lx\n", regs->ip); +} + +static void print_cr(void *arg) +{ + struct seq_file *seq = arg; + + seq_printf(seq, " cr0\t: %016lx\n", read_cr0()); + seq_printf(seq, " cr2\t: %016lx\n", read_cr2()); + seq_printf(seq, " cr3\t: %016lx\n", read_cr3()); + seq_printf(seq, " cr4\t: %016lx\n", read_cr4_safe()); +#ifdef CONFIG_X86_64 + seq_printf(seq, " cr8\t: %016lx\n", read_cr8()); +#endif +} + +static void print_desc_ptr(char *str, struct seq_file *seq, struct desc_ptr dt) +{ + seq_printf(seq, " %s\t: %016llx\n", str, (u64)(dt.address | dt.size)); +} + +static void print_dt(void *seq) +{ + struct desc_ptr dt; + unsigned long ldt; + + /* IDT */ + store_idt((struct desc_ptr *)&dt); + print_desc_ptr("IDT", seq, dt); + + /* GDT */ + store_gdt((struct desc_ptr *)&dt); + print_desc_ptr("GDT", seq, dt); + + /* LDT */ + store_ldt(ldt); + seq_printf(seq, " LDT\t: %016lx\n", ldt); + + /* TR */ + store_tr(ldt); + seq_printf(seq, " TR\t: %016lx\n", ldt); +} + +static void print_dr(void *arg) +{ + struct seq_file *seq = arg; + unsigned long dr; + int i; + + for (i = 0; i < 8; i++) { + /* Ignore db4, db5 */ + if ((i == 4) || (i == 5)) + continue; + get_debugreg(dr, i); + seq_printf(seq, " dr%d\t: %016lx\n", i, dr); + } + + seq_printf(seq, "\n MSR\t:\n"); +} + +static void print_apic(void *arg) +{ + struct seq_file *seq = arg; + +#ifdef CONFIG_X86_LOCAL_APIC + seq_printf(seq, " LAPIC\t:\n"); + seq_printf(seq, " ID\t\t: %08x\n", apic_read(APIC_ID) >> 24); + seq_printf(seq, " LVR\t\t: %08x\n", apic_read(APIC_LVR)); + seq_printf(seq, " TASKPRI\t: %08x\n", apic_read(APIC_TASKPRI)); + seq_printf(seq, " ARBPRI\t\t: %08x\n", apic_read(APIC_ARBPRI)); + seq_printf(seq, " PROCPRI\t: %08x\n", apic_read(APIC_PROCPRI)); + seq_printf(seq, " LDR\t\t: %08x\n", apic_read(APIC_LDR)); + seq_printf(seq, " DFR\t\t: %08x\n", apic_read(APIC_DFR)); + seq_printf(seq, " SPIV\t\t: %08x\n", apic_read(APIC_SPIV)); + seq_printf(seq, " ISR\t\t: %08x\n", apic_read(APIC_ISR)); + seq_printf(seq, " ESR\t\t: %08x\n", apic_read(APIC_ESR)); + seq_printf(seq, " ICR\t\t: %08x\n", apic_read(APIC_ICR)); + seq_printf(seq, " ICR2\t\t: %08x\n", apic_read(APIC_ICR2)); + seq_printf(seq, " LVTT\t\t: %08x\n", apic_read(APIC_LVTT)); + seq_printf(seq, " LVTTHMR\t: %08x\n", apic_read(APIC_LVTTHMR)); + seq_printf(seq, " LVTPC\t\t: %08x\n", apic_read(APIC_LVTPC)); + seq_printf(seq, " LVT0\t\t: %08x\n", apic_read(APIC_LVT0)); + seq_printf(seq, " LVT1\t\t: %08x\n", apic_read(APIC_LVT1)); + seq_printf(seq, " LVTERR\t\t: %08x\n", apic_read(APIC_LVTERR)); + seq_printf(seq, " TMICT\t\t: %08x\n", apic_read(APIC_TMICT)); + seq_printf(seq, " TMCCT\t\t: %08x\n", apic_read(APIC_TMCCT)); + seq_printf(seq, " TDCR\t\t: %08x\n", apic_read(APIC_TDCR)); +#endif /* CONFIG_X86_LOCAL_APIC */ + + seq_printf(seq, "\n MSR\t:\n"); +} + +static int cpu_seq_show(struct seq_file *seq, void *v) +{ + struct cpu_private *priv = seq->private; + + if (priv == NULL) + return -EINVAL; + + switch (cpu_base[priv->type].flag) { + case CPU_TSS: + smp_call_function_single(priv->cpu, print_tss, seq, 1); + break; + case CPU_CR: + smp_call_function_single(priv->cpu, print_cr, seq, 1); + break; + case CPU_DT: + smp_call_function_single(priv->cpu, print_dt, seq, 1); + break; + case CPU_DEBUG: + if (priv->file == CPU_INDEX_BIT) + smp_call_function_single(priv->cpu, print_dr, seq, 1); + print_msr(seq, priv->cpu, cpu_base[priv->type].flag); + break; + case CPU_APIC: + if (priv->file == CPU_INDEX_BIT) + smp_call_function_single(priv->cpu, print_apic, seq, 1); + print_msr(seq, priv->cpu, cpu_base[priv->type].flag); + break; + + default: + print_msr(seq, priv->cpu, cpu_base[priv->type].flag); + break; + } + seq_printf(seq, "\n"); + + return 0; +} + +static void *cpu_seq_start(struct seq_file *seq, loff_t *pos) +{ + if (*pos == 0) /* One time is enough ;-) */ + return seq; + + return NULL; +} + +static void *cpu_seq_next(struct seq_file *seq, void *v, loff_t *pos) +{ + (*pos)++; + + return cpu_seq_start(seq, pos); +} + +static void cpu_seq_stop(struct seq_file *seq, void *v) +{ +} + +static const struct seq_operations cpu_seq_ops = { + .start = cpu_seq_start, + .next = cpu_seq_next, + .stop = cpu_seq_stop, + .show = cpu_seq_show, +}; + +static int cpu_seq_open(struct inode *inode, struct file *file) +{ + struct cpu_private *priv = inode->i_private; + struct seq_file *seq; + int err; + + err = seq_open(file, &cpu_seq_ops); + if (!err) { + seq = file->private_data; + seq->private = priv; + } + + return err; +} + +static int write_msr(struct cpu_private *priv, u64 val) +{ + u32 low, high; + + high = (val >> 32) & 0xffffffff; + low = val & 0xffffffff; + + if (!wrmsr_safe_on_cpu(priv->cpu, priv->reg, low, high)) + return 0; + + return -EPERM; +} + +static int write_cpu_register(struct cpu_private *priv, const char *buf) +{ + int ret = -EPERM; + u64 val; + + ret = strict_strtoull(buf, 0, &val); + if (ret < 0) + return ret; + + /* Supporting only MSRs */ + if (priv->type < CPU_TSS_BIT) + return write_msr(priv, val); + + return ret; +} + +static ssize_t cpu_write(struct file *file, const char __user *ubuf, + size_t count, loff_t *off) +{ + struct seq_file *seq = file->private_data; + struct cpu_private *priv = seq->private; + char buf[19]; + + if ((priv == NULL) || (count >= sizeof(buf))) + return -EINVAL; + + if (copy_from_user(&buf, ubuf, count)) + return -EFAULT; + + buf[count] = 0; + + if ((cpu_base[priv->type].write) && (cpu_file[priv->file].write)) + if (!write_cpu_register(priv, buf)) + return count; + + return -EACCES; +} + +static const struct file_operations cpu_fops = { + .owner = THIS_MODULE, + .open = cpu_seq_open, + .read = seq_read, + .write = cpu_write, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int cpu_create_file(unsigned cpu, unsigned type, unsigned reg, + unsigned file, struct dentry *dentry) +{ + struct cpu_private *priv = NULL; + + /* Already intialized */ + if (file == CPU_INDEX_BIT) + if (per_cpu(cpu_arr[type].init, cpu)) + return 0; + + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (priv == NULL) + return -ENOMEM; + + priv->cpu = cpu; + priv->type = type; + priv->reg = reg; + priv->file = file; + mutex_lock(&cpu_debug_lock); + per_cpu(priv_arr[type], cpu) = priv; + per_cpu(cpu_priv_count, cpu)++; + mutex_unlock(&cpu_debug_lock); + + if (file) + debugfs_create_file(cpu_file[file].name, S_IRUGO, + dentry, (void *)priv, &cpu_fops); + else { + debugfs_create_file(cpu_base[type].name, S_IRUGO, + per_cpu(cpu_arr[type].dentry, cpu), + (void *)priv, &cpu_fops); + mutex_lock(&cpu_debug_lock); + per_cpu(cpu_arr[type].init, cpu) = 1; + mutex_unlock(&cpu_debug_lock); + } + + return 0; +} + +static int cpu_init_regfiles(unsigned cpu, unsigned int type, unsigned reg, + struct dentry *dentry) +{ + unsigned file; + int err = 0; + + for (file = 0; file < ARRAY_SIZE(cpu_file); file++) { + err = cpu_create_file(cpu, type, reg, file, dentry); + if (err) + return err; + } + + return err; +} + +static int cpu_init_msr(unsigned cpu, unsigned type, struct dentry *dentry) +{ + struct dentry *cpu_dentry = NULL; + unsigned reg, reg_min, reg_max; + int i, range, err = 0; + char reg_dir[12]; + u32 low, high; + + range = get_cpu_range_count(cpu); + + for (i = 0; i < range; i++) { + if (!get_cpu_range(cpu, ®_min, ®_max, i, + cpu_base[type].flag)) + continue; + + for (reg = reg_min; reg <= reg_max; reg++) { + if (rdmsr_safe_on_cpu(cpu, reg, &low, &high)) + continue; + + sprintf(reg_dir, "0x%x", reg); + cpu_dentry = debugfs_create_dir(reg_dir, dentry); + err = cpu_init_regfiles(cpu, type, reg, cpu_dentry); + if (err) + return err; + } + } + + return err; +} + +static int cpu_init_allreg(unsigned cpu, struct dentry *dentry) +{ + struct dentry *cpu_dentry = NULL; + unsigned type; + int err = 0; + + for (type = 0; type < ARRAY_SIZE(cpu_base) - 1; type++) { + if (!is_typeflag_valid(cpu, cpu_base[type].flag)) + continue; + cpu_dentry = debugfs_create_dir(cpu_base[type].name, dentry); + per_cpu(cpu_arr[type].dentry, cpu) = cpu_dentry; + + if (type < CPU_TSS_BIT) + err = cpu_init_msr(cpu, type, cpu_dentry); + else + err = cpu_create_file(cpu, type, 0, CPU_INDEX_BIT, + cpu_dentry); + if (err) + return err; + } + + return err; +} + +static int cpu_init_cpu(void) +{ + struct dentry *cpu_dentry = NULL; + struct cpuinfo_x86 *cpui; + char cpu_dir[12]; + unsigned cpu; + int err = 0; + + for (cpu = 0; cpu < nr_cpu_ids; cpu++) { + cpui = &cpu_data(cpu); + if (!cpu_has(cpui, X86_FEATURE_MSR)) + continue; + per_cpu(cpu_model, cpu) = ((cpui->x86_vendor << 16) | + (cpui->x86 << 8) | + (cpui->x86_model)); + per_cpu(cpu_modelflag, cpu) = get_cpu_modelflag(cpu); + + sprintf(cpu_dir, "cpu%d", cpu); + cpu_dentry = debugfs_create_dir(cpu_dir, cpu_debugfs_dir); + err = cpu_init_allreg(cpu, cpu_dentry); + + pr_info("cpu%d(%d) debug files %d\n", + cpu, nr_cpu_ids, per_cpu(cpu_priv_count, cpu)); + if (per_cpu(cpu_priv_count, cpu) > MAX_CPU_FILES) { + pr_err("Register files count %d exceeds limit %d\n", + per_cpu(cpu_priv_count, cpu), MAX_CPU_FILES); + per_cpu(cpu_priv_count, cpu) = MAX_CPU_FILES; + err = -ENFILE; + } + if (err) + return err; + } + + return err; +} + +static int __init cpu_debug_init(void) +{ + cpu_debugfs_dir = debugfs_create_dir("cpu", arch_debugfs_dir); + + return cpu_init_cpu(); +} + +static void __exit cpu_debug_exit(void) +{ + int i, cpu; + + if (cpu_debugfs_dir) + debugfs_remove_recursive(cpu_debugfs_dir); + + for (cpu = 0; cpu < nr_cpu_ids; cpu++) + for (i = 0; i < per_cpu(cpu_priv_count, cpu); i++) + kfree(per_cpu(priv_arr[i], cpu)); +} + +module_init(cpu_debug_init); +module_exit(cpu_debug_exit); + +MODULE_AUTHOR("Jaswinder Singh Rajput"); +MODULE_DESCRIPTION("CPU Debug module"); +MODULE_LICENSE("GPL"); diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c index ffd0f5ed071a..593171e967ef 100644 --- a/arch/x86/kernel/cpu/cyrix.c +++ b/arch/x86/kernel/cpu/cyrix.c @@ -61,23 +61,23 @@ static void __cpuinit do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) */ static unsigned char Cx86_dir0_msb __cpuinitdata = 0; -static char Cx86_model[][9] __cpuinitdata = { +static const char __cpuinitconst Cx86_model[][9] = { "Cx486", "Cx486", "5x86 ", "6x86", "MediaGX ", "6x86MX ", "M II ", "Unknown" }; -static char Cx486_name[][5] __cpuinitdata = { +static const char __cpuinitconst Cx486_name[][5] = { "SLC", "DLC", "SLC2", "DLC2", "SRx", "DRx", "SRx2", "DRx2" }; -static char Cx486S_name[][4] __cpuinitdata = { +static const char __cpuinitconst Cx486S_name[][4] = { "S", "S2", "Se", "S2e" }; -static char Cx486D_name[][4] __cpuinitdata = { +static const char __cpuinitconst Cx486D_name[][4] = { "DX", "DX2", "?", "?", "?", "DX4" }; static char Cx86_cb[] __cpuinitdata = "?.5x Core/Bus Clock"; -static char cyrix_model_mult1[] __cpuinitdata = "12??43"; -static char cyrix_model_mult2[] __cpuinitdata = "12233445"; +static const char __cpuinitconst cyrix_model_mult1[] = "12??43"; +static const char __cpuinitconst cyrix_model_mult2[] = "12233445"; /* * Reset the slow-loop (SLOP) bit on the 686(L) which is set by some old @@ -435,7 +435,7 @@ static void __cpuinit cyrix_identify(struct cpuinfo_x86 *c) } } -static struct cpu_dev cyrix_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst cyrix_cpu_dev = { .c_vendor = "Cyrix", .c_ident = { "CyrixInstead" }, .c_early_init = early_init_cyrix, @@ -446,7 +446,7 @@ static struct cpu_dev cyrix_cpu_dev __cpuinitdata = { cpu_dev_register(cyrix_cpu_dev); -static struct cpu_dev nsc_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst nsc_cpu_dev = { .c_vendor = "NSC", .c_ident = { "Geode by NSC" }, .c_init = init_nsc, diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 1a89a2b68d15..7437fa133c02 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -14,6 +14,7 @@ #include <asm/uaccess.h> #include <asm/ds.h> #include <asm/bugs.h> +#include <asm/cpu.h> #ifdef CONFIG_X86_64 #include <asm/topology.h> @@ -54,6 +55,11 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) c->x86_cache_alignment = 128; #endif + /* CPUID workaround for 0F33/0F34 CPU */ + if (c->x86 == 0xF && c->x86_model == 0x3 + && (c->x86_mask == 0x3 || c->x86_mask == 0x4)) + c->x86_phys_bits = 36; + /* * c->x86_power is 8000_0007 edx. Bit 8 is TSC runs at constant rate * with P/T states and does not stop in deep C-states. @@ -116,6 +122,28 @@ static void __cpuinit trap_init_f00f_bug(void) } #endif +static void __cpuinit intel_smp_check(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + /* calling is from identify_secondary_cpu() ? */ + if (c->cpu_index == boot_cpu_id) + return; + + /* + * Mask B, Pentium, but not Pentium MMX + */ + if (c->x86 == 5 && + c->x86_mask >= 1 && c->x86_mask <= 4 && + c->x86_model <= 3) { + /* + * Remember we have B step Pentia with bugs + */ + WARN_ONCE(1, "WARNING: SMP operation may be unreliable" + "with B stepping processors.\n"); + } +#endif +} + static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) { unsigned long lo, hi; @@ -192,6 +220,8 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) #ifdef CONFIG_X86_NUMAQ numaq_tsc_disable(); #endif + + intel_smp_check(c); } #else static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) @@ -391,7 +421,7 @@ static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 *c, unsigned i } #endif -static struct cpu_dev intel_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst intel_cpu_dev = { .c_vendor = "Intel", .c_ident = { "GenuineIntel" }, #ifdef CONFIG_X86_32 diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 7293508d8f5c..c471eb1a389c 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -32,7 +32,7 @@ struct _cache_table }; /* all the cache descriptor types we care about (no TLB or trace cache entries) */ -static struct _cache_table cache_table[] __cpuinitdata = +static const struct _cache_table __cpuinitconst cache_table[] = { { 0x06, LVL_1_INST, 8 }, /* 4-way set assoc, 32 byte line size */ { 0x08, LVL_1_INST, 16 }, /* 4-way set assoc, 32 byte line size */ @@ -206,15 +206,15 @@ union l3_cache { unsigned val; }; -static unsigned short assocs[] __cpuinitdata = { +static const unsigned short __cpuinitconst assocs[] = { [1] = 1, [2] = 2, [4] = 4, [6] = 8, [8] = 16, [0xa] = 32, [0xb] = 48, [0xc] = 64, [0xf] = 0xffff // ?? }; -static unsigned char levels[] __cpuinitdata = { 1, 1, 2, 3 }; -static unsigned char types[] __cpuinitdata = { 1, 2, 3, 3 }; +static const unsigned char __cpuinitconst levels[] = { 1, 1, 2, 3 }; +static const unsigned char __cpuinitconst types[] = { 1, 2, 3, 3 }; static void __cpuinit amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax, diff --git a/arch/x86/kernel/cpu/mcheck/Makefile b/arch/x86/kernel/cpu/mcheck/Makefile index d7d2323bbb69..b2f89829bbe8 100644 --- a/arch/x86/kernel/cpu/mcheck/Makefile +++ b/arch/x86/kernel/cpu/mcheck/Makefile @@ -4,3 +4,4 @@ obj-$(CONFIG_X86_32) += k7.o p4.o p5.o p6.o winchip.o obj-$(CONFIG_X86_MCE_INTEL) += mce_intel_64.o obj-$(CONFIG_X86_MCE_AMD) += mce_amd_64.o obj-$(CONFIG_X86_MCE_NONFATAL) += non-fatal.o +obj-$(CONFIG_X86_MCE_THRESHOLD) += threshold.o diff --git a/arch/x86/kernel/cpu/mcheck/mce_32.c b/arch/x86/kernel/cpu/mcheck/mce_32.c index dfaebce3633e..3552119b091d 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_32.c +++ b/arch/x86/kernel/cpu/mcheck/mce_32.c @@ -60,20 +60,6 @@ void mcheck_init(struct cpuinfo_x86 *c) } } -static unsigned long old_cr4 __initdata; - -void __init stop_mce(void) -{ - old_cr4 = read_cr4(); - clear_in_cr4(X86_CR4_MCE); -} - -void __init restart_mce(void) -{ - if (old_cr4 & X86_CR4_MCE) - set_in_cr4(X86_CR4_MCE); -} - static int __init mcheck_disable(char *str) { mce_disabled = 1; diff --git a/arch/x86/kernel/cpu/mcheck/mce_64.c b/arch/x86/kernel/cpu/mcheck/mce_64.c index fe79985ce0f2..ca14604611ec 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_64.c @@ -3,6 +3,8 @@ * K8 parts Copyright 2002,2003 Andi Kleen, SuSE Labs. * Rest from unknown author(s). * 2004 Andi Kleen. Rewrote most of it. + * Copyright 2008 Intel Corporation + * Author: Andi Kleen */ #include <linux/init.h> @@ -24,6 +26,9 @@ #include <linux/ctype.h> #include <linux/kmod.h> #include <linux/kdebug.h> +#include <linux/kobject.h> +#include <linux/sysfs.h> +#include <linux/ratelimit.h> #include <asm/processor.h> #include <asm/msr.h> #include <asm/mce.h> @@ -32,7 +37,6 @@ #include <asm/idle.h> #define MISC_MCELOG_MINOR 227 -#define NR_SYSFS_BANKS 6 atomic_t mce_entry; @@ -47,7 +51,7 @@ static int mce_dont_init; */ static int tolerant = 1; static int banks; -static unsigned long bank[NR_SYSFS_BANKS] = { [0 ... NR_SYSFS_BANKS-1] = ~0UL }; +static u64 *bank; static unsigned long notify_user; static int rip_msr; static int mce_bootlog = -1; @@ -58,6 +62,19 @@ static char *trigger_argv[2] = { trigger, NULL }; static DECLARE_WAIT_QUEUE_HEAD(mce_wait); +/* MCA banks polled by the period polling timer for corrected events */ +DEFINE_PER_CPU(mce_banks_t, mce_poll_banks) = { + [0 ... BITS_TO_LONGS(MAX_NR_BANKS)-1] = ~0UL +}; + +/* Do initial initialization of a struct mce */ +void mce_setup(struct mce *m) +{ + memset(m, 0, sizeof(struct mce)); + m->cpu = smp_processor_id(); + rdtscll(m->tsc); +} + /* * Lockless MCE logging infrastructure. * This avoids deadlocks on printk locks without having to break locks. Also @@ -119,11 +136,11 @@ static void print_mce(struct mce *m) print_symbol("{%s}", m->ip); printk("\n"); } - printk(KERN_EMERG "TSC %Lx ", m->tsc); + printk(KERN_EMERG "TSC %llx ", m->tsc); if (m->addr) - printk("ADDR %Lx ", m->addr); + printk("ADDR %llx ", m->addr); if (m->misc) - printk("MISC %Lx ", m->misc); + printk("MISC %llx ", m->misc); printk("\n"); printk(KERN_EMERG "This is not a software problem!\n"); printk(KERN_EMERG "Run through mcelog --ascii to decode " @@ -149,8 +166,10 @@ static void mce_panic(char *msg, struct mce *backup, unsigned long start) panic(msg); } -static int mce_available(struct cpuinfo_x86 *c) +int mce_available(struct cpuinfo_x86 *c) { + if (mce_dont_init) + return 0; return cpu_has(c, X86_FEATURE_MCE) && cpu_has(c, X86_FEATURE_MCA); } @@ -172,7 +191,77 @@ static inline void mce_get_rip(struct mce *m, struct pt_regs *regs) } /* - * The actual machine check handler + * Poll for corrected events or events that happened before reset. + * Those are just logged through /dev/mcelog. + * + * This is executed in standard interrupt context. + */ +void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) +{ + struct mce m; + int i; + + mce_setup(&m); + + rdmsrl(MSR_IA32_MCG_STATUS, m.mcgstatus); + for (i = 0; i < banks; i++) { + if (!bank[i] || !test_bit(i, *b)) + continue; + + m.misc = 0; + m.addr = 0; + m.bank = i; + m.tsc = 0; + + barrier(); + rdmsrl(MSR_IA32_MC0_STATUS + i*4, m.status); + if (!(m.status & MCI_STATUS_VAL)) + continue; + + /* + * Uncorrected events are handled by the exception handler + * when it is enabled. But when the exception is disabled log + * everything. + * + * TBD do the same check for MCI_STATUS_EN here? + */ + if ((m.status & MCI_STATUS_UC) && !(flags & MCP_UC)) + continue; + + if (m.status & MCI_STATUS_MISCV) + rdmsrl(MSR_IA32_MC0_MISC + i*4, m.misc); + if (m.status & MCI_STATUS_ADDRV) + rdmsrl(MSR_IA32_MC0_ADDR + i*4, m.addr); + + if (!(flags & MCP_TIMESTAMP)) + m.tsc = 0; + /* + * Don't get the IP here because it's unlikely to + * have anything to do with the actual error location. + */ + + mce_log(&m); + add_taint(TAINT_MACHINE_CHECK); + + /* + * Clear state for this bank. + */ + wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0); + } + + /* + * Don't clear MCG_STATUS here because it's only defined for + * exceptions. + */ +} + +/* + * The actual machine check handler. This only handles real + * exceptions when something got corrupted coming in through int 18. + * + * This is executed in NMI context not subject to normal locking rules. This + * implies that most kernel services cannot be safely used. Don't even + * think about putting a printk in there! */ void do_machine_check(struct pt_regs * regs, long error_code) { @@ -190,17 +279,18 @@ void do_machine_check(struct pt_regs * regs, long error_code) * error. */ int kill_it = 0; + DECLARE_BITMAP(toclear, MAX_NR_BANKS); atomic_inc(&mce_entry); - if ((regs - && notify_die(DIE_NMI, "machine check", regs, error_code, + if (notify_die(DIE_NMI, "machine check", regs, error_code, 18, SIGKILL) == NOTIFY_STOP) - || !banks) + goto out2; + if (!banks) goto out2; - memset(&m, 0, sizeof(struct mce)); - m.cpu = smp_processor_id(); + mce_setup(&m); + rdmsrl(MSR_IA32_MCG_STATUS, m.mcgstatus); /* if the restart IP is not valid, we're done for */ if (!(m.mcgstatus & MCG_STATUS_RIPV)) @@ -210,18 +300,32 @@ void do_machine_check(struct pt_regs * regs, long error_code) barrier(); for (i = 0; i < banks; i++) { - if (i < NR_SYSFS_BANKS && !bank[i]) + __clear_bit(i, toclear); + if (!bank[i]) continue; m.misc = 0; m.addr = 0; m.bank = i; - m.tsc = 0; rdmsrl(MSR_IA32_MC0_STATUS + i*4, m.status); if ((m.status & MCI_STATUS_VAL) == 0) continue; + /* + * Non uncorrected errors are handled by machine_check_poll + * Leave them alone. + */ + if ((m.status & MCI_STATUS_UC) == 0) + continue; + + /* + * Set taint even when machine check was not enabled. + */ + add_taint(TAINT_MACHINE_CHECK); + + __set_bit(i, toclear); + if (m.status & MCI_STATUS_EN) { /* if PCC was set, there's no way out */ no_way_out |= !!(m.status & MCI_STATUS_PCC); @@ -235,6 +339,12 @@ void do_machine_check(struct pt_regs * regs, long error_code) no_way_out = 1; kill_it = 1; } + } else { + /* + * Machine check event was not enabled. Clear, but + * ignore. + */ + continue; } if (m.status & MCI_STATUS_MISCV) @@ -243,10 +353,7 @@ void do_machine_check(struct pt_regs * regs, long error_code) rdmsrl(MSR_IA32_MC0_ADDR + i*4, m.addr); mce_get_rip(&m, regs); - if (error_code >= 0) - rdtscll(m.tsc); - if (error_code != -2) - mce_log(&m); + mce_log(&m); /* Did this bank cause the exception? */ /* Assume that the bank with uncorrectable errors did it, @@ -255,14 +362,8 @@ void do_machine_check(struct pt_regs * regs, long error_code) panicm = m; panicm_found = 1; } - - add_taint(TAINT_MACHINE_CHECK); } - /* Never do anything final in the polling timer */ - if (!regs) - goto out; - /* If we didn't find an uncorrectable error, pick the last one (shouldn't happen, just being safe). */ if (!panicm_found) @@ -309,10 +410,11 @@ void do_machine_check(struct pt_regs * regs, long error_code) /* notify userspace ASAP */ set_thread_flag(TIF_MCE_NOTIFY); - out: /* the last thing we do is clear state */ - for (i = 0; i < banks; i++) - wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0); + for (i = 0; i < banks; i++) { + if (test_bit(i, toclear)) + wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0); + } wrmsrl(MSR_IA32_MCG_STATUS, 0); out2: atomic_dec(&mce_entry); @@ -332,15 +434,13 @@ void do_machine_check(struct pt_regs * regs, long error_code) * and historically has been the register value of the * MSR_IA32_THERMAL_STATUS (Intel) msr. */ -void mce_log_therm_throt_event(unsigned int cpu, __u64 status) +void mce_log_therm_throt_event(__u64 status) { struct mce m; - memset(&m, 0, sizeof(m)); - m.cpu = cpu; + mce_setup(&m); m.bank = MCE_THERMAL_BANK; m.status = status; - rdtscll(m.tsc); mce_log(&m); } #endif /* CONFIG_X86_MCE_INTEL */ @@ -353,18 +453,18 @@ void mce_log_therm_throt_event(unsigned int cpu, __u64 status) static int check_interval = 5 * 60; /* 5 minutes */ static int next_interval; /* in jiffies */ -static void mcheck_timer(struct work_struct *work); -static DECLARE_DELAYED_WORK(mcheck_work, mcheck_timer); +static void mcheck_timer(unsigned long); +static DEFINE_PER_CPU(struct timer_list, mce_timer); -static void mcheck_check_cpu(void *info) +static void mcheck_timer(unsigned long data) { - if (mce_available(¤t_cpu_data)) - do_machine_check(NULL, 0); -} + struct timer_list *t = &per_cpu(mce_timer, data); -static void mcheck_timer(struct work_struct *work) -{ - on_each_cpu(mcheck_check_cpu, NULL, 1); + WARN_ON(smp_processor_id() != data); + + if (mce_available(¤t_cpu_data)) + machine_check_poll(MCP_TIMESTAMP, + &__get_cpu_var(mce_poll_banks)); /* * Alert userspace if needed. If we logged an MCE, reduce the @@ -377,31 +477,41 @@ static void mcheck_timer(struct work_struct *work) (int)round_jiffies_relative(check_interval*HZ)); } - schedule_delayed_work(&mcheck_work, next_interval); + t->expires = jiffies + next_interval; + add_timer(t); +} + +static void mce_do_trigger(struct work_struct *work) +{ + call_usermodehelper(trigger, trigger_argv, NULL, UMH_NO_WAIT); } +static DECLARE_WORK(mce_trigger_work, mce_do_trigger); + /* - * This is only called from process context. This is where we do - * anything we need to alert userspace about new MCEs. This is called - * directly from the poller and also from entry.S and idle, thanks to - * TIF_MCE_NOTIFY. + * Notify the user(s) about new machine check events. + * Can be called from interrupt context, but not from machine check/NMI + * context. */ int mce_notify_user(void) { + /* Not more than two messages every minute */ + static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); + clear_thread_flag(TIF_MCE_NOTIFY); if (test_and_clear_bit(0, ¬ify_user)) { - static unsigned long last_print; - unsigned long now = jiffies; - wake_up_interruptible(&mce_wait); - if (trigger[0]) - call_usermodehelper(trigger, trigger_argv, NULL, - UMH_NO_WAIT); - if (time_after_eq(now, last_print + (check_interval*HZ))) { - last_print = now; + /* + * There is no risk of missing notifications because + * work_pending is always cleared before the function is + * executed. + */ + if (trigger[0] && !work_pending(&mce_trigger_work)) + schedule_work(&mce_trigger_work); + + if (__ratelimit(&ratelimit)) printk(KERN_INFO "Machine check events logged\n"); - } return 1; } @@ -425,63 +535,78 @@ static struct notifier_block mce_idle_notifier = { static __init int periodic_mcheck_init(void) { - next_interval = check_interval * HZ; - if (next_interval) - schedule_delayed_work(&mcheck_work, - round_jiffies_relative(next_interval)); - idle_notifier_register(&mce_idle_notifier); - return 0; + idle_notifier_register(&mce_idle_notifier); + return 0; } __initcall(periodic_mcheck_init); - /* * Initialize Machine Checks for a CPU. */ -static void mce_init(void *dummy) +static int mce_cap_init(void) { u64 cap; - int i; + unsigned b; rdmsrl(MSR_IA32_MCG_CAP, cap); - banks = cap & 0xff; - if (banks > MCE_EXTENDED_BANK) { - banks = MCE_EXTENDED_BANK; - printk(KERN_INFO "MCE: warning: using only %d banks\n", - MCE_EXTENDED_BANK); + b = cap & 0xff; + if (b > MAX_NR_BANKS) { + printk(KERN_WARNING + "MCE: Using only %u machine check banks out of %u\n", + MAX_NR_BANKS, b); + b = MAX_NR_BANKS; } + + /* Don't support asymmetric configurations today */ + WARN_ON(banks != 0 && b != banks); + banks = b; + if (!bank) { + bank = kmalloc(banks * sizeof(u64), GFP_KERNEL); + if (!bank) + return -ENOMEM; + memset(bank, 0xff, banks * sizeof(u64)); + } + /* Use accurate RIP reporting if available. */ if ((cap & (1<<9)) && ((cap >> 16) & 0xff) >= 9) rip_msr = MSR_IA32_MCG_EIP; - /* Log the machine checks left over from the previous reset. - This also clears all registers */ - do_machine_check(NULL, mce_bootlog ? -1 : -2); + return 0; +} + +static void mce_init(void *dummy) +{ + u64 cap; + int i; + mce_banks_t all_banks; + + /* + * Log the machine checks left over from the previous reset. + */ + bitmap_fill(all_banks, MAX_NR_BANKS); + machine_check_poll(MCP_UC, &all_banks); set_in_cr4(X86_CR4_MCE); + rdmsrl(MSR_IA32_MCG_CAP, cap); if (cap & MCG_CTL_P) wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); for (i = 0; i < banks; i++) { - if (i < NR_SYSFS_BANKS) - wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]); - else - wrmsrl(MSR_IA32_MC0_CTL+4*i, ~0UL); - + wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]); wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0); } } /* Add per CPU specific workarounds here */ -static void __cpuinit mce_cpu_quirks(struct cpuinfo_x86 *c) +static void mce_cpu_quirks(struct cpuinfo_x86 *c) { /* This should be disabled by the BIOS, but isn't always */ if (c->x86_vendor == X86_VENDOR_AMD) { - if(c->x86 == 15) + if (c->x86 == 15 && banks > 4) /* disable GART TBL walk error reporting, which trips off incorrectly with the IOMMU & 3ware & Cerberus. */ - clear_bit(10, &bank[4]); + clear_bit(10, (unsigned long *)&bank[4]); if(c->x86 <= 17 && mce_bootlog < 0) /* Lots of broken BIOS around that don't clear them by default and leave crap in there. Don't log. */ @@ -504,20 +629,38 @@ static void mce_cpu_features(struct cpuinfo_x86 *c) } } +static void mce_init_timer(void) +{ + struct timer_list *t = &__get_cpu_var(mce_timer); + + /* data race harmless because everyone sets to the same value */ + if (!next_interval) + next_interval = check_interval * HZ; + if (!next_interval) + return; + setup_timer(t, mcheck_timer, smp_processor_id()); + t->expires = round_jiffies(jiffies + next_interval); + add_timer(t); +} + /* * Called for each booted CPU to set up machine checks. * Must be called with preempt off. */ void __cpuinit mcheck_init(struct cpuinfo_x86 *c) { - mce_cpu_quirks(c); + if (!mce_available(c)) + return; - if (mce_dont_init || - !mce_available(c)) + if (mce_cap_init() < 0) { + mce_dont_init = 1; return; + } + mce_cpu_quirks(c); mce_init(NULL); mce_cpu_features(c); + mce_init_timer(); } /* @@ -573,7 +716,7 @@ static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize, { unsigned long *cpu_tsc; static DEFINE_MUTEX(mce_read_mutex); - unsigned next; + unsigned prev, next; char __user *buf = ubuf; int i, err; @@ -592,25 +735,32 @@ static ssize_t mce_read(struct file *filp, char __user *ubuf, size_t usize, } err = 0; - for (i = 0; i < next; i++) { - unsigned long start = jiffies; - - while (!mcelog.entry[i].finished) { - if (time_after_eq(jiffies, start + 2)) { - memset(mcelog.entry + i,0, sizeof(struct mce)); - goto timeout; + prev = 0; + do { + for (i = prev; i < next; i++) { + unsigned long start = jiffies; + + while (!mcelog.entry[i].finished) { + if (time_after_eq(jiffies, start + 2)) { + memset(mcelog.entry + i, 0, + sizeof(struct mce)); + goto timeout; + } + cpu_relax(); } - cpu_relax(); + smp_rmb(); + err |= copy_to_user(buf, mcelog.entry + i, + sizeof(struct mce)); + buf += sizeof(struct mce); +timeout: + ; } - smp_rmb(); - err |= copy_to_user(buf, mcelog.entry + i, sizeof(struct mce)); - buf += sizeof(struct mce); - timeout: - ; - } - memset(mcelog.entry, 0, next * sizeof(struct mce)); - mcelog.next = 0; + memset(mcelog.entry + prev, 0, + (next - prev) * sizeof(struct mce)); + prev = next; + next = cmpxchg(&mcelog.next, prev, 0); + } while (next != prev); synchronize_sched(); @@ -680,20 +830,6 @@ static struct miscdevice mce_log_device = { &mce_chrdev_ops, }; -static unsigned long old_cr4 __initdata; - -void __init stop_mce(void) -{ - old_cr4 = read_cr4(); - clear_in_cr4(X86_CR4_MCE); -} - -void __init restart_mce(void) -{ - if (old_cr4 & X86_CR4_MCE) - set_in_cr4(X86_CR4_MCE); -} - /* * Old style boot options parsing. Only for compatibility. */ @@ -703,8 +839,7 @@ static int __init mcheck_disable(char *str) return 1; } -/* mce=off disables machine check. Note you can re-enable it later - using sysfs. +/* mce=off disables machine check. mce=TOLERANCELEVEL (number, see above) mce=bootlog Log MCEs from before booting. Disabled by default on AMD. mce=nobootlog Don't log MCEs from before booting. */ @@ -728,6 +863,29 @@ __setup("mce=", mcheck_enable); * Sysfs support */ +/* + * Disable machine checks on suspend and shutdown. We can't really handle + * them later. + */ +static int mce_disable(void) +{ + int i; + + for (i = 0; i < banks; i++) + wrmsrl(MSR_IA32_MC0_CTL + i*4, 0); + return 0; +} + +static int mce_suspend(struct sys_device *dev, pm_message_t state) +{ + return mce_disable(); +} + +static int mce_shutdown(struct sys_device *dev) +{ + return mce_disable(); +} + /* On resume clear all MCE state. Don't want to see leftovers from the BIOS. Only one CPU is active at this time, the others get readded later using CPU hotplug. */ @@ -738,20 +896,24 @@ static int mce_resume(struct sys_device *dev) return 0; } +static void mce_cpu_restart(void *data) +{ + del_timer_sync(&__get_cpu_var(mce_timer)); + if (mce_available(¤t_cpu_data)) + mce_init(NULL); + mce_init_timer(); +} + /* Reinit MCEs after user configuration changes */ static void mce_restart(void) { - if (next_interval) - cancel_delayed_work(&mcheck_work); - /* Timer race is harmless here */ - on_each_cpu(mce_init, NULL, 1); next_interval = check_interval * HZ; - if (next_interval) - schedule_delayed_work(&mcheck_work, - round_jiffies_relative(next_interval)); + on_each_cpu(mce_cpu_restart, NULL, 1); } static struct sysdev_class mce_sysclass = { + .suspend = mce_suspend, + .shutdown = mce_shutdown, .resume = mce_resume, .name = "machinecheck", }; @@ -778,16 +940,26 @@ void (*threshold_cpu_callback)(unsigned long action, unsigned int cpu) __cpuinit } \ static SYSDEV_ATTR(name, 0644, show_ ## name, set_ ## name); -/* - * TBD should generate these dynamically based on number of available banks. - * Have only 6 contol banks in /sysfs until then. - */ -ACCESSOR(bank0ctl,bank[0],mce_restart()) -ACCESSOR(bank1ctl,bank[1],mce_restart()) -ACCESSOR(bank2ctl,bank[2],mce_restart()) -ACCESSOR(bank3ctl,bank[3],mce_restart()) -ACCESSOR(bank4ctl,bank[4],mce_restart()) -ACCESSOR(bank5ctl,bank[5],mce_restart()) +static struct sysdev_attribute *bank_attrs; + +static ssize_t show_bank(struct sys_device *s, struct sysdev_attribute *attr, + char *buf) +{ + u64 b = bank[attr - bank_attrs]; + return sprintf(buf, "%llx\n", b); +} + +static ssize_t set_bank(struct sys_device *s, struct sysdev_attribute *attr, + const char *buf, size_t siz) +{ + char *end; + u64 new = simple_strtoull(buf, &end, 0); + if (end == buf) + return -EINVAL; + bank[attr - bank_attrs] = new; + mce_restart(); + return end-buf; +} static ssize_t show_trigger(struct sys_device *s, struct sysdev_attribute *attr, char *buf) @@ -814,8 +986,6 @@ static SYSDEV_ATTR(trigger, 0644, show_trigger, set_trigger); static SYSDEV_INT_ATTR(tolerant, 0644, tolerant); ACCESSOR(check_interval,check_interval,mce_restart()) static struct sysdev_attribute *mce_attributes[] = { - &attr_bank0ctl, &attr_bank1ctl, &attr_bank2ctl, - &attr_bank3ctl, &attr_bank4ctl, &attr_bank5ctl, &attr_tolerant.attr, &attr_check_interval, &attr_trigger, NULL }; @@ -845,11 +1015,22 @@ static __cpuinit int mce_create_device(unsigned int cpu) if (err) goto error; } + for (i = 0; i < banks; i++) { + err = sysdev_create_file(&per_cpu(device_mce, cpu), + &bank_attrs[i]); + if (err) + goto error2; + } cpu_set(cpu, mce_device_initialized); return 0; +error2: + while (--i >= 0) { + sysdev_remove_file(&per_cpu(device_mce, cpu), + &bank_attrs[i]); + } error: - while (i--) { + while (--i >= 0) { sysdev_remove_file(&per_cpu(device_mce,cpu), mce_attributes[i]); } @@ -868,15 +1049,46 @@ static __cpuinit void mce_remove_device(unsigned int cpu) for (i = 0; mce_attributes[i]; i++) sysdev_remove_file(&per_cpu(device_mce,cpu), mce_attributes[i]); + for (i = 0; i < banks; i++) + sysdev_remove_file(&per_cpu(device_mce, cpu), + &bank_attrs[i]); sysdev_unregister(&per_cpu(device_mce,cpu)); cpu_clear(cpu, mce_device_initialized); } +/* Make sure there are no machine checks on offlined CPUs. */ +static void mce_disable_cpu(void *h) +{ + int i; + unsigned long action = *(unsigned long *)h; + + if (!mce_available(¤t_cpu_data)) + return; + if (!(action & CPU_TASKS_FROZEN)) + cmci_clear(); + for (i = 0; i < banks; i++) + wrmsrl(MSR_IA32_MC0_CTL + i*4, 0); +} + +static void mce_reenable_cpu(void *h) +{ + int i; + unsigned long action = *(unsigned long *)h; + + if (!mce_available(¤t_cpu_data)) + return; + if (!(action & CPU_TASKS_FROZEN)) + cmci_reenable(); + for (i = 0; i < banks; i++) + wrmsrl(MSR_IA32_MC0_CTL + i*4, bank[i]); +} + /* Get notified when a cpu comes on/off. Be hotplug friendly. */ static int __cpuinit mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; + struct timer_list *t = &per_cpu(mce_timer, cpu); switch (action) { case CPU_ONLINE: @@ -891,6 +1103,21 @@ static int __cpuinit mce_cpu_callback(struct notifier_block *nfb, threshold_cpu_callback(action, cpu); mce_remove_device(cpu); break; + case CPU_DOWN_PREPARE: + case CPU_DOWN_PREPARE_FROZEN: + del_timer_sync(t); + smp_call_function_single(cpu, mce_disable_cpu, &action, 1); + break; + case CPU_DOWN_FAILED: + case CPU_DOWN_FAILED_FROZEN: + t->expires = round_jiffies(jiffies + next_interval); + add_timer_on(t, cpu); + smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); + break; + case CPU_POST_DEAD: + /* intentionally ignoring frozen here */ + cmci_rediscover(cpu); + break; } return NOTIFY_OK; } @@ -899,6 +1126,34 @@ static struct notifier_block mce_cpu_notifier __cpuinitdata = { .notifier_call = mce_cpu_callback, }; +static __init int mce_init_banks(void) +{ + int i; + + bank_attrs = kzalloc(sizeof(struct sysdev_attribute) * banks, + GFP_KERNEL); + if (!bank_attrs) + return -ENOMEM; + + for (i = 0; i < banks; i++) { + struct sysdev_attribute *a = &bank_attrs[i]; + a->attr.name = kasprintf(GFP_KERNEL, "bank%d", i); + if (!a->attr.name) + goto nomem; + a->attr.mode = 0644; + a->show = show_bank; + a->store = set_bank; + } + return 0; + +nomem: + while (--i >= 0) + kfree(bank_attrs[i].attr.name); + kfree(bank_attrs); + bank_attrs = NULL; + return -ENOMEM; +} + static __init int mce_init_device(void) { int err; @@ -906,6 +1161,11 @@ static __init int mce_init_device(void) if (!mce_available(&boot_cpu_data)) return -EIO; + + err = mce_init_banks(); + if (err) + return err; + err = sysdev_class_register(&mce_sysclass); if (err) return err; diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c index 9817506dd469..7d01be868870 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd_64.c @@ -79,6 +79,8 @@ static unsigned char shared_bank[NR_BANKS] = { static DEFINE_PER_CPU(unsigned char, bank_map); /* see which banks are on */ +static void amd_threshold_interrupt(void); + /* * CPU Initialization */ @@ -90,7 +92,8 @@ struct thresh_restart { }; /* must be called with correct cpu affinity */ -static long threshold_restart_bank(void *_tr) +/* Called via smp_call_function_single() */ +static void threshold_restart_bank(void *_tr) { struct thresh_restart *tr = _tr; u32 mci_misc_hi, mci_misc_lo; @@ -117,7 +120,6 @@ static long threshold_restart_bank(void *_tr) mci_misc_hi |= MASK_COUNT_EN_HI; wrmsr(tr->b->address, mci_misc_lo, mci_misc_hi); - return 0; } /* cpu init entry point, called from mce.c with preempt off */ @@ -174,6 +176,8 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c) tr.reset = 0; tr.old_limit = 0; threshold_restart_bank(&tr); + + mce_threshold_vector = amd_threshold_interrupt; } } } @@ -187,19 +191,13 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c) * the interrupt goes off when error_count reaches threshold_limit. * the handler will simply log mcelog w/ software defined bank number. */ -asmlinkage void mce_threshold_interrupt(void) +static void amd_threshold_interrupt(void) { unsigned int bank, block; struct mce m; u32 low = 0, high = 0, address = 0; - ack_APIC_irq(); - exit_idle(); - irq_enter(); - - memset(&m, 0, sizeof(m)); - rdtscll(m.tsc); - m.cpu = smp_processor_id(); + mce_setup(&m); /* assume first bank caused it */ for (bank = 0; bank < NR_BANKS; ++bank) { @@ -233,7 +231,8 @@ asmlinkage void mce_threshold_interrupt(void) /* Log the machine check that caused the threshold event. */ - do_machine_check(NULL, 0); + machine_check_poll(MCP_TIMESTAMP, + &__get_cpu_var(mce_poll_banks)); if (high & MASK_OVERFLOW_HI) { rdmsrl(address, m.misc); @@ -243,13 +242,10 @@ asmlinkage void mce_threshold_interrupt(void) + bank * NR_BLOCKS + block; mce_log(&m); - goto out; + return; } } } -out: - inc_irq_stat(irq_threshold_count); - irq_exit(); } /* @@ -283,7 +279,7 @@ static ssize_t store_interrupt_enable(struct threshold_block *b, tr.b = b; tr.reset = 0; tr.old_limit = 0; - work_on_cpu(b->cpu, threshold_restart_bank, &tr); + smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1); return end - buf; } @@ -305,23 +301,32 @@ static ssize_t store_threshold_limit(struct threshold_block *b, tr.b = b; tr.reset = 0; - work_on_cpu(b->cpu, threshold_restart_bank, &tr); + smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1); return end - buf; } -static long local_error_count(void *_b) +struct threshold_block_cross_cpu { + struct threshold_block *tb; + long retval; +}; + +static void local_error_count_handler(void *_tbcc) { - struct threshold_block *b = _b; + struct threshold_block_cross_cpu *tbcc = _tbcc; + struct threshold_block *b = tbcc->tb; u32 low, high; rdmsr(b->address, low, high); - return (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit); + tbcc->retval = (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit); } static ssize_t show_error_count(struct threshold_block *b, char *buf) { - return sprintf(buf, "%lx\n", work_on_cpu(b->cpu, local_error_count, b)); + struct threshold_block_cross_cpu tbcc = { .tb = b, }; + + smp_call_function_single(b->cpu, local_error_count_handler, &tbcc, 1); + return sprintf(buf, "%lx\n", tbcc.retval); } static ssize_t store_error_count(struct threshold_block *b, @@ -329,7 +334,7 @@ static ssize_t store_error_count(struct threshold_block *b, { struct thresh_restart tr = { .b = b, .reset = 1, .old_limit = 0 }; - work_on_cpu(b->cpu, threshold_restart_bank, &tr); + smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1); return 1; } @@ -398,7 +403,7 @@ static __cpuinit int allocate_threshold_blocks(unsigned int cpu, if ((bank >= NR_BANKS) || (block >= NR_BLOCKS)) return 0; - if (rdmsr_safe(address, &low, &high)) + if (rdmsr_safe_on_cpu(cpu, address, &low, &high)) return 0; if (!(high & MASK_VALID_HI)) { @@ -462,12 +467,11 @@ out_free: return err; } -static __cpuinit long local_allocate_threshold_blocks(void *_bank) +static __cpuinit long +local_allocate_threshold_blocks(int cpu, unsigned int bank) { - unsigned int *bank = _bank; - - return allocate_threshold_blocks(smp_processor_id(), *bank, 0, - MSR_IA32_MC0_MISC + *bank * 4); + return allocate_threshold_blocks(cpu, bank, 0, + MSR_IA32_MC0_MISC + bank * 4); } /* symlinks sibling shared banks to first core. first core owns dir/files. */ @@ -530,7 +534,7 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) per_cpu(threshold_banks, cpu)[bank] = b; - err = work_on_cpu(cpu, local_allocate_threshold_blocks, &bank); + err = local_allocate_threshold_blocks(cpu, bank); if (err) goto out_free; diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index aa5e287c98e0..57df3d383470 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -1,6 +1,8 @@ /* * Intel specific MCE features. * Copyright 2004 Zwane Mwaikambo <zwane@linuxpower.ca> + * Copyright (C) 2008, 2009 Intel Corporation + * Author: Andi Kleen */ #include <linux/init.h> @@ -13,6 +15,7 @@ #include <asm/hw_irq.h> #include <asm/idle.h> #include <asm/therm_throt.h> +#include <asm/apic.h> asmlinkage void smp_thermal_interrupt(void) { @@ -25,7 +28,7 @@ asmlinkage void smp_thermal_interrupt(void) rdmsrl(MSR_IA32_THERM_STATUS, msr_val); if (therm_throt_process(msr_val & 1)) - mce_log_therm_throt_event(smp_processor_id(), msr_val); + mce_log_therm_throt_event(msr_val); inc_irq_stat(irq_thermal_count); irq_exit(); @@ -85,7 +88,209 @@ static void intel_init_thermal(struct cpuinfo_x86 *c) return; } +/* + * Support for Intel Correct Machine Check Interrupts. This allows + * the CPU to raise an interrupt when a corrected machine check happened. + * Normally we pick those up using a regular polling timer. + * Also supports reliable discovery of shared banks. + */ + +static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned); + +/* + * cmci_discover_lock protects against parallel discovery attempts + * which could race against each other. + */ +static DEFINE_SPINLOCK(cmci_discover_lock); + +#define CMCI_THRESHOLD 1 + +static int cmci_supported(int *banks) +{ + u64 cap; + + /* + * Vendor check is not strictly needed, but the initial + * initialization is vendor keyed and this + * makes sure none of the backdoors are entered otherwise. + */ + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) + return 0; + if (!cpu_has_apic || lapic_get_maxlvt() < 6) + return 0; + rdmsrl(MSR_IA32_MCG_CAP, cap); + *banks = min_t(unsigned, MAX_NR_BANKS, cap & 0xff); + return !!(cap & MCG_CMCI_P); +} + +/* + * The interrupt handler. This is called on every event. + * Just call the poller directly to log any events. + * This could in theory increase the threshold under high load, + * but doesn't for now. + */ +static void intel_threshold_interrupt(void) +{ + machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); + mce_notify_user(); +} + +static void print_update(char *type, int *hdr, int num) +{ + if (*hdr == 0) + printk(KERN_INFO "CPU %d MCA banks", smp_processor_id()); + *hdr = 1; + printk(KERN_CONT " %s:%d", type, num); +} + +/* + * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks + * on this CPU. Use the algorithm recommended in the SDM to discover shared + * banks. + */ +static void cmci_discover(int banks, int boot) +{ + unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned); + int hdr = 0; + int i; + + spin_lock(&cmci_discover_lock); + for (i = 0; i < banks; i++) { + u64 val; + + if (test_bit(i, owned)) + continue; + + rdmsrl(MSR_IA32_MC0_CTL2 + i, val); + + /* Already owned by someone else? */ + if (val & CMCI_EN) { + if (test_and_clear_bit(i, owned) || boot) + print_update("SHD", &hdr, i); + __clear_bit(i, __get_cpu_var(mce_poll_banks)); + continue; + } + + val |= CMCI_EN | CMCI_THRESHOLD; + wrmsrl(MSR_IA32_MC0_CTL2 + i, val); + rdmsrl(MSR_IA32_MC0_CTL2 + i, val); + + /* Did the enable bit stick? -- the bank supports CMCI */ + if (val & CMCI_EN) { + if (!test_and_set_bit(i, owned) || boot) + print_update("CMCI", &hdr, i); + __clear_bit(i, __get_cpu_var(mce_poll_banks)); + } else { + WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks))); + } + } + spin_unlock(&cmci_discover_lock); + if (hdr) + printk(KERN_CONT "\n"); +} + +/* + * Just in case we missed an event during initialization check + * all the CMCI owned banks. + */ +void cmci_recheck(void) +{ + unsigned long flags; + int banks; + + if (!mce_available(¤t_cpu_data) || !cmci_supported(&banks)) + return; + local_irq_save(flags); + machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); + local_irq_restore(flags); +} + +/* + * Disable CMCI on this CPU for all banks it owns when it goes down. + * This allows other CPUs to claim the banks on rediscovery. + */ +void cmci_clear(void) +{ + int i; + int banks; + u64 val; + + if (!cmci_supported(&banks)) + return; + spin_lock(&cmci_discover_lock); + for (i = 0; i < banks; i++) { + if (!test_bit(i, __get_cpu_var(mce_banks_owned))) + continue; + /* Disable CMCI */ + rdmsrl(MSR_IA32_MC0_CTL2 + i, val); + val &= ~(CMCI_EN|CMCI_THRESHOLD_MASK); + wrmsrl(MSR_IA32_MC0_CTL2 + i, val); + __clear_bit(i, __get_cpu_var(mce_banks_owned)); + } + spin_unlock(&cmci_discover_lock); +} + +/* + * After a CPU went down cycle through all the others and rediscover + * Must run in process context. + */ +void cmci_rediscover(int dying) +{ + int banks; + int cpu; + cpumask_var_t old; + + if (!cmci_supported(&banks)) + return; + if (!alloc_cpumask_var(&old, GFP_KERNEL)) + return; + cpumask_copy(old, ¤t->cpus_allowed); + + for_each_online_cpu (cpu) { + if (cpu == dying) + continue; + if (set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu))) + continue; + /* Recheck banks in case CPUs don't all have the same */ + if (cmci_supported(&banks)) + cmci_discover(banks, 0); + } + + set_cpus_allowed_ptr(current, old); + free_cpumask_var(old); +} + +/* + * Reenable CMCI on this CPU in case a CPU down failed. + */ +void cmci_reenable(void) +{ + int banks; + if (cmci_supported(&banks)) + cmci_discover(banks, 0); +} + +static void intel_init_cmci(void) +{ + int banks; + + if (!cmci_supported(&banks)) + return; + + mce_threshold_vector = intel_threshold_interrupt; + cmci_discover(banks, 1); + /* + * For CPU #0 this runs with still disabled APIC, but that's + * ok because only the vector is set up. We still do another + * check for the banks later for CPU #0 just to make sure + * to not miss any events. + */ + apic_write(APIC_LVTCMCI, THRESHOLD_APIC_VECTOR|APIC_DM_FIXED); + cmci_recheck(); +} + void mce_intel_feature_init(struct cpuinfo_x86 *c) { intel_init_thermal(c); + intel_init_cmci(); } diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c new file mode 100644 index 000000000000..23ee9e730f78 --- /dev/null +++ b/arch/x86/kernel/cpu/mcheck/threshold.c @@ -0,0 +1,29 @@ +/* + * Common corrected MCE threshold handler code: + */ +#include <linux/interrupt.h> +#include <linux/kernel.h> + +#include <asm/irq_vectors.h> +#include <asm/apic.h> +#include <asm/idle.h> +#include <asm/mce.h> + +static void default_threshold_interrupt(void) +{ + printk(KERN_ERR "Unexpected threshold interrupt at vector %x\n", + THRESHOLD_APIC_VECTOR); +} + +void (*mce_threshold_vector)(void) = default_threshold_interrupt; + +asmlinkage void mce_threshold_interrupt(void) +{ + exit_idle(); + irq_enter(); + inc_irq_stat(irq_threshold_count); + mce_threshold_vector(); + irq_exit(); + /* Ack only at the end to avoid potential reentry */ + ack_APIC_irq(); +} diff --git a/arch/x86/kernel/cpu/mtrr/Makefile b/arch/x86/kernel/cpu/mtrr/Makefile index 191fc0533649..f4361b56f8e9 100644 --- a/arch/x86/kernel/cpu/mtrr/Makefile +++ b/arch/x86/kernel/cpu/mtrr/Makefile @@ -1,3 +1,3 @@ -obj-y := main.o if.o generic.o state.o +obj-y := main.o if.o generic.o state.o cleanup.o obj-$(CONFIG_X86_32) += amd.o cyrix.o centaur.o diff --git a/arch/x86/kernel/cpu/mtrr/cleanup.c b/arch/x86/kernel/cpu/mtrr/cleanup.c new file mode 100644 index 000000000000..ce0fe4b5c04f --- /dev/null +++ b/arch/x86/kernel/cpu/mtrr/cleanup.c @@ -0,0 +1,1101 @@ +/* MTRR (Memory Type Range Register) cleanup + + Copyright (C) 2009 Yinghai Lu + + This library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public + License as published by the Free Software Foundation; either + version 2 of the License, or (at your option) any later version. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with this library; if not, write to the Free + Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. +*/ + +#include <linux/module.h> +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/smp.h> +#include <linux/cpu.h> +#include <linux/mutex.h> +#include <linux/sort.h> + +#include <asm/e820.h> +#include <asm/mtrr.h> +#include <asm/uaccess.h> +#include <asm/processor.h> +#include <asm/msr.h> +#include <asm/kvm_para.h> +#include "mtrr.h" + +/* should be related to MTRR_VAR_RANGES nums */ +#define RANGE_NUM 256 + +struct res_range { + unsigned long start; + unsigned long end; +}; + +static int __init +add_range(struct res_range *range, int nr_range, unsigned long start, + unsigned long end) +{ + /* out of slots */ + if (nr_range >= RANGE_NUM) + return nr_range; + + range[nr_range].start = start; + range[nr_range].end = end; + + nr_range++; + + return nr_range; +} + +static int __init +add_range_with_merge(struct res_range *range, int nr_range, unsigned long start, + unsigned long end) +{ + int i; + + /* try to merge it with old one */ + for (i = 0; i < nr_range; i++) { + unsigned long final_start, final_end; + unsigned long common_start, common_end; + + if (!range[i].end) + continue; + + common_start = max(range[i].start, start); + common_end = min(range[i].end, end); + if (common_start > common_end + 1) + continue; + + final_start = min(range[i].start, start); + final_end = max(range[i].end, end); + + range[i].start = final_start; + range[i].end = final_end; + return nr_range; + } + + /* need to add that */ + return add_range(range, nr_range, start, end); +} + +static void __init +subtract_range(struct res_range *range, unsigned long start, unsigned long end) +{ + int i, j; + + for (j = 0; j < RANGE_NUM; j++) { + if (!range[j].end) + continue; + + if (start <= range[j].start && end >= range[j].end) { + range[j].start = 0; + range[j].end = 0; + continue; + } + + if (start <= range[j].start && end < range[j].end && + range[j].start < end + 1) { + range[j].start = end + 1; + continue; + } + + + if (start > range[j].start && end >= range[j].end && + range[j].end > start - 1) { + range[j].end = start - 1; + continue; + } + + if (start > range[j].start && end < range[j].end) { + /* find the new spare */ + for (i = 0; i < RANGE_NUM; i++) { + if (range[i].end == 0) + break; + } + if (i < RANGE_NUM) { + range[i].end = range[j].end; + range[i].start = end + 1; + } else { + printk(KERN_ERR "run of slot in ranges\n"); + } + range[j].end = start - 1; + continue; + } + } +} + +static int __init cmp_range(const void *x1, const void *x2) +{ + const struct res_range *r1 = x1; + const struct res_range *r2 = x2; + long start1, start2; + + start1 = r1->start; + start2 = r2->start; + + return start1 - start2; +} + +struct var_mtrr_range_state { + unsigned long base_pfn; + unsigned long size_pfn; + mtrr_type type; +}; + +static struct var_mtrr_range_state __initdata range_state[RANGE_NUM]; +static int __initdata debug_print; + +static int __init +x86_get_mtrr_mem_range(struct res_range *range, int nr_range, + unsigned long extra_remove_base, + unsigned long extra_remove_size) +{ + unsigned long base, size; + mtrr_type type; + int i; + + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_WRBACK) + continue; + base = range_state[i].base_pfn; + size = range_state[i].size_pfn; + nr_range = add_range_with_merge(range, nr_range, base, + base + size - 1); + } + if (debug_print) { + printk(KERN_DEBUG "After WB checking\n"); + for (i = 0; i < nr_range; i++) + printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", + range[i].start, range[i].end + 1); + } + + /* take out UC ranges */ + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_UNCACHABLE && + type != MTRR_TYPE_WRPROT) + continue; + size = range_state[i].size_pfn; + if (!size) + continue; + base = range_state[i].base_pfn; + if (base < (1<<(20-PAGE_SHIFT)) && mtrr_state.have_fixed && + (mtrr_state.enabled & 1)) { + /* Var MTRR contains UC entry below 1M? Skip it: */ + printk(KERN_WARNING "WARNING: BIOS bug: VAR MTRR %d " + "contains strange UC entry under 1M, check " + "with your system vendor!\n", i); + if (base + size <= (1<<(20-PAGE_SHIFT))) + continue; + size -= (1<<(20-PAGE_SHIFT)) - base; + base = 1<<(20-PAGE_SHIFT); + } + subtract_range(range, base, base + size - 1); + } + if (extra_remove_size) + subtract_range(range, extra_remove_base, + extra_remove_base + extra_remove_size - 1); + + /* get new range num */ + nr_range = 0; + for (i = 0; i < RANGE_NUM; i++) { + if (!range[i].end) + continue; + nr_range++; + } + if (debug_print) { + printk(KERN_DEBUG "After UC checking\n"); + for (i = 0; i < nr_range; i++) + printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", + range[i].start, range[i].end + 1); + } + + /* sort the ranges */ + sort(range, nr_range, sizeof(struct res_range), cmp_range, NULL); + if (debug_print) { + printk(KERN_DEBUG "After sorting\n"); + for (i = 0; i < nr_range; i++) + printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", + range[i].start, range[i].end + 1); + } + + /* clear those is not used */ + for (i = nr_range; i < RANGE_NUM; i++) + memset(&range[i], 0, sizeof(range[i])); + + return nr_range; +} + +static struct res_range __initdata range[RANGE_NUM]; +static int __initdata nr_range; + +#ifdef CONFIG_MTRR_SANITIZER + +static unsigned long __init sum_ranges(struct res_range *range, int nr_range) +{ + unsigned long sum; + int i; + + sum = 0; + for (i = 0; i < nr_range; i++) + sum += range[i].end + 1 - range[i].start; + + return sum; +} + +static int enable_mtrr_cleanup __initdata = + CONFIG_MTRR_SANITIZER_ENABLE_DEFAULT; + +static int __init disable_mtrr_cleanup_setup(char *str) +{ + enable_mtrr_cleanup = 0; + return 0; +} +early_param("disable_mtrr_cleanup", disable_mtrr_cleanup_setup); + +static int __init enable_mtrr_cleanup_setup(char *str) +{ + enable_mtrr_cleanup = 1; + return 0; +} +early_param("enable_mtrr_cleanup", enable_mtrr_cleanup_setup); + +static int __init mtrr_cleanup_debug_setup(char *str) +{ + debug_print = 1; + return 0; +} +early_param("mtrr_cleanup_debug", mtrr_cleanup_debug_setup); + +struct var_mtrr_state { + unsigned long range_startk; + unsigned long range_sizek; + unsigned long chunk_sizek; + unsigned long gran_sizek; + unsigned int reg; +}; + +static void __init +set_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, + unsigned char type, unsigned int address_bits) +{ + u32 base_lo, base_hi, mask_lo, mask_hi; + u64 base, mask; + + if (!sizek) { + fill_mtrr_var_range(reg, 0, 0, 0, 0); + return; + } + + mask = (1ULL << address_bits) - 1; + mask &= ~((((u64)sizek) << 10) - 1); + + base = ((u64)basek) << 10; + + base |= type; + mask |= 0x800; + + base_lo = base & ((1ULL<<32) - 1); + base_hi = base >> 32; + + mask_lo = mask & ((1ULL<<32) - 1); + mask_hi = mask >> 32; + + fill_mtrr_var_range(reg, base_lo, base_hi, mask_lo, mask_hi); +} + +static void __init +save_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, + unsigned char type) +{ + range_state[reg].base_pfn = basek >> (PAGE_SHIFT - 10); + range_state[reg].size_pfn = sizek >> (PAGE_SHIFT - 10); + range_state[reg].type = type; +} + +static void __init +set_var_mtrr_all(unsigned int address_bits) +{ + unsigned long basek, sizek; + unsigned char type; + unsigned int reg; + + for (reg = 0; reg < num_var_ranges; reg++) { + basek = range_state[reg].base_pfn << (PAGE_SHIFT - 10); + sizek = range_state[reg].size_pfn << (PAGE_SHIFT - 10); + type = range_state[reg].type; + + set_var_mtrr(reg, basek, sizek, type, address_bits); + } +} + +static unsigned long to_size_factor(unsigned long sizek, char *factorp) +{ + char factor; + unsigned long base = sizek; + + if (base & ((1<<10) - 1)) { + /* not MB alignment */ + factor = 'K'; + } else if (base & ((1<<20) - 1)) { + factor = 'M'; + base >>= 10; + } else { + factor = 'G'; + base >>= 20; + } + + *factorp = factor; + + return base; +} + +static unsigned int __init +range_to_mtrr(unsigned int reg, unsigned long range_startk, + unsigned long range_sizek, unsigned char type) +{ + if (!range_sizek || (reg >= num_var_ranges)) + return reg; + + while (range_sizek) { + unsigned long max_align, align; + unsigned long sizek; + + /* Compute the maximum size I can make a range */ + if (range_startk) + max_align = ffs(range_startk) - 1; + else + max_align = 32; + align = fls(range_sizek) - 1; + if (align > max_align) + align = max_align; + + sizek = 1 << align; + if (debug_print) { + char start_factor = 'K', size_factor = 'K'; + unsigned long start_base, size_base; + + start_base = to_size_factor(range_startk, + &start_factor), + size_base = to_size_factor(sizek, &size_factor), + + printk(KERN_DEBUG "Setting variable MTRR %d, " + "base: %ld%cB, range: %ld%cB, type %s\n", + reg, start_base, start_factor, + size_base, size_factor, + (type == MTRR_TYPE_UNCACHABLE) ? "UC" : + ((type == MTRR_TYPE_WRBACK) ? "WB" : "Other") + ); + } + save_var_mtrr(reg++, range_startk, sizek, type); + range_startk += sizek; + range_sizek -= sizek; + if (reg >= num_var_ranges) + break; + } + return reg; +} + +static unsigned __init +range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek, + unsigned long sizek) +{ + unsigned long hole_basek, hole_sizek; + unsigned long second_basek, second_sizek; + unsigned long range0_basek, range0_sizek; + unsigned long range_basek, range_sizek; + unsigned long chunk_sizek; + unsigned long gran_sizek; + + hole_basek = 0; + hole_sizek = 0; + second_basek = 0; + second_sizek = 0; + chunk_sizek = state->chunk_sizek; + gran_sizek = state->gran_sizek; + + /* align with gran size, prevent small block used up MTRRs */ + range_basek = ALIGN(state->range_startk, gran_sizek); + if ((range_basek > basek) && basek) + return second_sizek; + state->range_sizek -= (range_basek - state->range_startk); + range_sizek = ALIGN(state->range_sizek, gran_sizek); + + while (range_sizek > state->range_sizek) { + range_sizek -= gran_sizek; + if (!range_sizek) + return 0; + } + state->range_sizek = range_sizek; + + /* try to append some small hole */ + range0_basek = state->range_startk; + range0_sizek = ALIGN(state->range_sizek, chunk_sizek); + + /* no increase */ + if (range0_sizek == state->range_sizek) { + if (debug_print) + printk(KERN_DEBUG "rangeX: %016lx - %016lx\n", + range0_basek<<10, + (range0_basek + state->range_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range0_basek, + state->range_sizek, MTRR_TYPE_WRBACK); + return 0; + } + + /* only cut back, when it is not the last */ + if (sizek) { + while (range0_basek + range0_sizek > (basek + sizek)) { + if (range0_sizek >= chunk_sizek) + range0_sizek -= chunk_sizek; + else + range0_sizek = 0; + + if (!range0_sizek) + break; + } + } + +second_try: + range_basek = range0_basek + range0_sizek; + + /* one hole in the middle */ + if (range_basek > basek && range_basek <= (basek + sizek)) + second_sizek = range_basek - basek; + + if (range0_sizek > state->range_sizek) { + + /* one hole in middle or at end */ + hole_sizek = range0_sizek - state->range_sizek - second_sizek; + + /* hole size should be less than half of range0 size */ + if (hole_sizek >= (range0_sizek >> 1) && + range0_sizek >= chunk_sizek) { + range0_sizek -= chunk_sizek; + second_sizek = 0; + hole_sizek = 0; + + goto second_try; + } + } + + if (range0_sizek) { + if (debug_print) + printk(KERN_DEBUG "range0: %016lx - %016lx\n", + range0_basek<<10, + (range0_basek + range0_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range0_basek, + range0_sizek, MTRR_TYPE_WRBACK); + } + + if (range0_sizek < state->range_sizek) { + /* need to handle left over */ + range_sizek = state->range_sizek - range0_sizek; + + if (debug_print) + printk(KERN_DEBUG "range: %016lx - %016lx\n", + range_basek<<10, + (range_basek + range_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range_basek, + range_sizek, MTRR_TYPE_WRBACK); + } + + if (hole_sizek) { + hole_basek = range_basek - hole_sizek - second_sizek; + if (debug_print) + printk(KERN_DEBUG "hole: %016lx - %016lx\n", + hole_basek<<10, + (hole_basek + hole_sizek)<<10); + state->reg = range_to_mtrr(state->reg, hole_basek, + hole_sizek, MTRR_TYPE_UNCACHABLE); + } + + return second_sizek; +} + +static void __init +set_var_mtrr_range(struct var_mtrr_state *state, unsigned long base_pfn, + unsigned long size_pfn) +{ + unsigned long basek, sizek; + unsigned long second_sizek = 0; + + if (state->reg >= num_var_ranges) + return; + + basek = base_pfn << (PAGE_SHIFT - 10); + sizek = size_pfn << (PAGE_SHIFT - 10); + + /* See if I can merge with the last range */ + if ((basek <= 1024) || + (state->range_startk + state->range_sizek == basek)) { + unsigned long endk = basek + sizek; + state->range_sizek = endk - state->range_startk; + return; + } + /* Write the range mtrrs */ + if (state->range_sizek != 0) + second_sizek = range_to_mtrr_with_hole(state, basek, sizek); + + /* Allocate an msr */ + state->range_startk = basek + second_sizek; + state->range_sizek = sizek - second_sizek; +} + +/* mininum size of mtrr block that can take hole */ +static u64 mtrr_chunk_size __initdata = (256ULL<<20); + +static int __init parse_mtrr_chunk_size_opt(char *p) +{ + if (!p) + return -EINVAL; + mtrr_chunk_size = memparse(p, &p); + return 0; +} +early_param("mtrr_chunk_size", parse_mtrr_chunk_size_opt); + +/* granity of mtrr of block */ +static u64 mtrr_gran_size __initdata; + +static int __init parse_mtrr_gran_size_opt(char *p) +{ + if (!p) + return -EINVAL; + mtrr_gran_size = memparse(p, &p); + return 0; +} +early_param("mtrr_gran_size", parse_mtrr_gran_size_opt); + +static int nr_mtrr_spare_reg __initdata = + CONFIG_MTRR_SANITIZER_SPARE_REG_NR_DEFAULT; + +static int __init parse_mtrr_spare_reg(char *arg) +{ + if (arg) + nr_mtrr_spare_reg = simple_strtoul(arg, NULL, 0); + return 0; +} + +early_param("mtrr_spare_reg_nr", parse_mtrr_spare_reg); + +static int __init +x86_setup_var_mtrrs(struct res_range *range, int nr_range, + u64 chunk_size, u64 gran_size) +{ + struct var_mtrr_state var_state; + int i; + int num_reg; + + var_state.range_startk = 0; + var_state.range_sizek = 0; + var_state.reg = 0; + var_state.chunk_sizek = chunk_size >> 10; + var_state.gran_sizek = gran_size >> 10; + + memset(range_state, 0, sizeof(range_state)); + + /* Write the range etc */ + for (i = 0; i < nr_range; i++) + set_var_mtrr_range(&var_state, range[i].start, + range[i].end - range[i].start + 1); + + /* Write the last range */ + if (var_state.range_sizek != 0) + range_to_mtrr_with_hole(&var_state, 0, 0); + + num_reg = var_state.reg; + /* Clear out the extra MTRR's */ + while (var_state.reg < num_var_ranges) { + save_var_mtrr(var_state.reg, 0, 0, 0); + var_state.reg++; + } + + return num_reg; +} + +struct mtrr_cleanup_result { + unsigned long gran_sizek; + unsigned long chunk_sizek; + unsigned long lose_cover_sizek; + unsigned int num_reg; + int bad; +}; + +/* + * gran_size: 64K, 128K, 256K, 512K, 1M, 2M, ..., 2G + * chunk size: gran_size, ..., 2G + * so we need (1+16)*8 + */ +#define NUM_RESULT 136 +#define PSHIFT (PAGE_SHIFT - 10) + +static struct mtrr_cleanup_result __initdata result[NUM_RESULT]; +static unsigned long __initdata min_loss_pfn[RANGE_NUM]; + +static void __init print_out_mtrr_range_state(void) +{ + int i; + char start_factor = 'K', size_factor = 'K'; + unsigned long start_base, size_base; + mtrr_type type; + + for (i = 0; i < num_var_ranges; i++) { + + size_base = range_state[i].size_pfn << (PAGE_SHIFT - 10); + if (!size_base) + continue; + + size_base = to_size_factor(size_base, &size_factor), + start_base = range_state[i].base_pfn << (PAGE_SHIFT - 10); + start_base = to_size_factor(start_base, &start_factor), + type = range_state[i].type; + + printk(KERN_DEBUG "reg %d, base: %ld%cB, range: %ld%cB, type %s\n", + i, start_base, start_factor, + size_base, size_factor, + (type == MTRR_TYPE_UNCACHABLE) ? "UC" : + ((type == MTRR_TYPE_WRPROT) ? "WP" : + ((type == MTRR_TYPE_WRBACK) ? "WB" : "Other")) + ); + } +} + +static int __init mtrr_need_cleanup(void) +{ + int i; + mtrr_type type; + unsigned long size; + /* extra one for all 0 */ + int num[MTRR_NUM_TYPES + 1]; + + /* check entries number */ + memset(num, 0, sizeof(num)); + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + size = range_state[i].size_pfn; + if (type >= MTRR_NUM_TYPES) + continue; + if (!size) + type = MTRR_NUM_TYPES; + if (type == MTRR_TYPE_WRPROT) + type = MTRR_TYPE_UNCACHABLE; + num[type]++; + } + + /* check if we got UC entries */ + if (!num[MTRR_TYPE_UNCACHABLE]) + return 0; + + /* check if we only had WB and UC */ + if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != + num_var_ranges - num[MTRR_NUM_TYPES]) + return 0; + + return 1; +} + +static unsigned long __initdata range_sums; +static void __init mtrr_calc_range_state(u64 chunk_size, u64 gran_size, + unsigned long extra_remove_base, + unsigned long extra_remove_size, + int i) +{ + int num_reg; + static struct res_range range_new[RANGE_NUM]; + static int nr_range_new; + unsigned long range_sums_new; + + /* convert ranges to var ranges state */ + num_reg = x86_setup_var_mtrrs(range, nr_range, + chunk_size, gran_size); + + /* we got new setting in range_state, check it */ + memset(range_new, 0, sizeof(range_new)); + nr_range_new = x86_get_mtrr_mem_range(range_new, 0, + extra_remove_base, extra_remove_size); + range_sums_new = sum_ranges(range_new, nr_range_new); + + result[i].chunk_sizek = chunk_size >> 10; + result[i].gran_sizek = gran_size >> 10; + result[i].num_reg = num_reg; + if (range_sums < range_sums_new) { + result[i].lose_cover_sizek = + (range_sums_new - range_sums) << PSHIFT; + result[i].bad = 1; + } else + result[i].lose_cover_sizek = + (range_sums - range_sums_new) << PSHIFT; + + /* double check it */ + if (!result[i].bad && !result[i].lose_cover_sizek) { + if (nr_range_new != nr_range || + memcmp(range, range_new, sizeof(range))) + result[i].bad = 1; + } + + if (!result[i].bad && (range_sums - range_sums_new < + min_loss_pfn[num_reg])) { + min_loss_pfn[num_reg] = + range_sums - range_sums_new; + } +} + +static void __init mtrr_print_out_one_result(int i) +{ + char gran_factor, chunk_factor, lose_factor; + unsigned long gran_base, chunk_base, lose_base; + + gran_base = to_size_factor(result[i].gran_sizek, &gran_factor), + chunk_base = to_size_factor(result[i].chunk_sizek, &chunk_factor), + lose_base = to_size_factor(result[i].lose_cover_sizek, &lose_factor), + printk(KERN_INFO "%sgran_size: %ld%c \tchunk_size: %ld%c \t", + result[i].bad ? "*BAD*" : " ", + gran_base, gran_factor, chunk_base, chunk_factor); + printk(KERN_CONT "num_reg: %d \tlose cover RAM: %s%ld%c\n", + result[i].num_reg, result[i].bad ? "-" : "", + lose_base, lose_factor); +} + +static int __init mtrr_search_optimal_index(void) +{ + int i; + int num_reg_good; + int index_good; + + if (nr_mtrr_spare_reg >= num_var_ranges) + nr_mtrr_spare_reg = num_var_ranges - 1; + num_reg_good = -1; + for (i = num_var_ranges - nr_mtrr_spare_reg; i > 0; i--) { + if (!min_loss_pfn[i]) + num_reg_good = i; + } + + index_good = -1; + if (num_reg_good != -1) { + for (i = 0; i < NUM_RESULT; i++) { + if (!result[i].bad && + result[i].num_reg == num_reg_good && + !result[i].lose_cover_sizek) { + index_good = i; + break; + } + } + } + + return index_good; +} + + +int __init mtrr_cleanup(unsigned address_bits) +{ + unsigned long extra_remove_base, extra_remove_size; + unsigned long base, size, def, dummy; + mtrr_type type; + u64 chunk_size, gran_size; + int index_good; + int i; + + if (!is_cpu(INTEL) || enable_mtrr_cleanup < 1) + return 0; + rdmsr(MTRRdefType_MSR, def, dummy); + def &= 0xff; + if (def != MTRR_TYPE_UNCACHABLE) + return 0; + + /* get it and store it aside */ + memset(range_state, 0, sizeof(range_state)); + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &base, &size, &type); + range_state[i].base_pfn = base; + range_state[i].size_pfn = size; + range_state[i].type = type; + } + + /* check if we need handle it and can handle it */ + if (!mtrr_need_cleanup()) + return 0; + + /* print original var MTRRs at first, for debugging: */ + printk(KERN_DEBUG "original variable MTRRs\n"); + print_out_mtrr_range_state(); + + memset(range, 0, sizeof(range)); + extra_remove_size = 0; + extra_remove_base = 1 << (32 - PAGE_SHIFT); + if (mtrr_tom2) + extra_remove_size = + (mtrr_tom2 >> PAGE_SHIFT) - extra_remove_base; + nr_range = x86_get_mtrr_mem_range(range, 0, extra_remove_base, + extra_remove_size); + /* + * [0, 1M) should always be coverred by var mtrr with WB + * and fixed mtrrs should take effective before var mtrr for it + */ + nr_range = add_range_with_merge(range, nr_range, 0, + (1ULL<<(20 - PAGE_SHIFT)) - 1); + /* sort the ranges */ + sort(range, nr_range, sizeof(struct res_range), cmp_range, NULL); + + range_sums = sum_ranges(range, nr_range); + printk(KERN_INFO "total RAM coverred: %ldM\n", + range_sums >> (20 - PAGE_SHIFT)); + + if (mtrr_chunk_size && mtrr_gran_size) { + i = 0; + mtrr_calc_range_state(mtrr_chunk_size, mtrr_gran_size, + extra_remove_base, extra_remove_size, i); + + mtrr_print_out_one_result(i); + + if (!result[i].bad) { + set_var_mtrr_all(address_bits); + printk(KERN_DEBUG "New variable MTRRs\n"); + print_out_mtrr_range_state(); + return 1; + } + printk(KERN_INFO "invalid mtrr_gran_size or mtrr_chunk_size, " + "will find optimal one\n"); + } + + i = 0; + memset(min_loss_pfn, 0xff, sizeof(min_loss_pfn)); + memset(result, 0, sizeof(result)); + for (gran_size = (1ULL<<16); gran_size < (1ULL<<32); gran_size <<= 1) { + + for (chunk_size = gran_size; chunk_size < (1ULL<<32); + chunk_size <<= 1) { + + if (i >= NUM_RESULT) + continue; + + mtrr_calc_range_state(chunk_size, gran_size, + extra_remove_base, extra_remove_size, i); + if (debug_print) { + mtrr_print_out_one_result(i); + printk(KERN_INFO "\n"); + } + + i++; + } + } + + /* try to find the optimal index */ + index_good = mtrr_search_optimal_index(); + + if (index_good != -1) { + printk(KERN_INFO "Found optimal setting for mtrr clean up\n"); + i = index_good; + mtrr_print_out_one_result(i); + + /* convert ranges to var ranges state */ + chunk_size = result[i].chunk_sizek; + chunk_size <<= 10; + gran_size = result[i].gran_sizek; + gran_size <<= 10; + x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size); + set_var_mtrr_all(address_bits); + printk(KERN_DEBUG "New variable MTRRs\n"); + print_out_mtrr_range_state(); + return 1; + } else { + /* print out all */ + for (i = 0; i < NUM_RESULT; i++) + mtrr_print_out_one_result(i); + } + + printk(KERN_INFO "mtrr_cleanup: can not find optimal value\n"); + printk(KERN_INFO "please specify mtrr_gran_size/mtrr_chunk_size\n"); + + return 0; +} +#else +int __init mtrr_cleanup(unsigned address_bits) +{ + return 0; +} +#endif + +static int disable_mtrr_trim; + +static int __init disable_mtrr_trim_setup(char *str) +{ + disable_mtrr_trim = 1; + return 0; +} +early_param("disable_mtrr_trim", disable_mtrr_trim_setup); + +/* + * Newer AMD K8s and later CPUs have a special magic MSR way to force WB + * for memory >4GB. Check for that here. + * Note this won't check if the MTRRs < 4GB where the magic bit doesn't + * apply to are wrong, but so far we don't know of any such case in the wild. + */ +#define Tom2Enabled (1U << 21) +#define Tom2ForceMemTypeWB (1U << 22) + +int __init amd_special_default_mtrr(void) +{ + u32 l, h; + + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) + return 0; + if (boot_cpu_data.x86 < 0xf || boot_cpu_data.x86 > 0x11) + return 0; + /* In case some hypervisor doesn't pass SYSCFG through */ + if (rdmsr_safe(MSR_K8_SYSCFG, &l, &h) < 0) + return 0; + /* + * Memory between 4GB and top of mem is forced WB by this magic bit. + * Reserved before K8RevF, but should be zero there. + */ + if ((l & (Tom2Enabled | Tom2ForceMemTypeWB)) == + (Tom2Enabled | Tom2ForceMemTypeWB)) + return 1; + return 0; +} + +static u64 __init real_trim_memory(unsigned long start_pfn, + unsigned long limit_pfn) +{ + u64 trim_start, trim_size; + trim_start = start_pfn; + trim_start <<= PAGE_SHIFT; + trim_size = limit_pfn; + trim_size <<= PAGE_SHIFT; + trim_size -= trim_start; + + return e820_update_range(trim_start, trim_size, E820_RAM, + E820_RESERVED); +} +/** + * mtrr_trim_uncached_memory - trim RAM not covered by MTRRs + * @end_pfn: ending page frame number + * + * Some buggy BIOSes don't setup the MTRRs properly for systems with certain + * memory configurations. This routine checks that the highest MTRR matches + * the end of memory, to make sure the MTRRs having a write back type cover + * all of the memory the kernel is intending to use. If not, it'll trim any + * memory off the end by adjusting end_pfn, removing it from the kernel's + * allocation pools, warning the user with an obnoxious message. + */ +int __init mtrr_trim_uncached_memory(unsigned long end_pfn) +{ + unsigned long i, base, size, highest_pfn = 0, def, dummy; + mtrr_type type; + u64 total_trim_size; + + /* extra one for all 0 */ + int num[MTRR_NUM_TYPES + 1]; + /* + * Make sure we only trim uncachable memory on machines that + * support the Intel MTRR architecture: + */ + if (!is_cpu(INTEL) || disable_mtrr_trim) + return 0; + rdmsr(MTRRdefType_MSR, def, dummy); + def &= 0xff; + if (def != MTRR_TYPE_UNCACHABLE) + return 0; + + /* get it and store it aside */ + memset(range_state, 0, sizeof(range_state)); + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &base, &size, &type); + range_state[i].base_pfn = base; + range_state[i].size_pfn = size; + range_state[i].type = type; + } + + /* Find highest cached pfn */ + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_WRBACK) + continue; + base = range_state[i].base_pfn; + size = range_state[i].size_pfn; + if (highest_pfn < base + size) + highest_pfn = base + size; + } + + /* kvm/qemu doesn't have mtrr set right, don't trim them all */ + if (!highest_pfn) { + printk(KERN_INFO "CPU MTRRs all blank - virtualized system.\n"); + return 0; + } + + /* check entries number */ + memset(num, 0, sizeof(num)); + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type >= MTRR_NUM_TYPES) + continue; + size = range_state[i].size_pfn; + if (!size) + type = MTRR_NUM_TYPES; + num[type]++; + } + + /* no entry for WB? */ + if (!num[MTRR_TYPE_WRBACK]) + return 0; + + /* check if we only had WB and UC */ + if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != + num_var_ranges - num[MTRR_NUM_TYPES]) + return 0; + + memset(range, 0, sizeof(range)); + nr_range = 0; + if (mtrr_tom2) { + range[nr_range].start = (1ULL<<(32 - PAGE_SHIFT)); + range[nr_range].end = (mtrr_tom2 >> PAGE_SHIFT) - 1; + if (highest_pfn < range[nr_range].end + 1) + highest_pfn = range[nr_range].end + 1; + nr_range++; + } + nr_range = x86_get_mtrr_mem_range(range, nr_range, 0, 0); + + total_trim_size = 0; + /* check the head */ + if (range[0].start) + total_trim_size += real_trim_memory(0, range[0].start); + /* check the holes */ + for (i = 0; i < nr_range - 1; i++) { + if (range[i].end + 1 < range[i+1].start) + total_trim_size += real_trim_memory(range[i].end + 1, + range[i+1].start); + } + /* check the top */ + i = nr_range - 1; + if (range[i].end + 1 < end_pfn) + total_trim_size += real_trim_memory(range[i].end + 1, + end_pfn); + + if (total_trim_size) { + printk(KERN_WARNING "WARNING: BIOS bug: CPU MTRRs don't cover" + " all of memory, losing %lluMB of RAM.\n", + total_trim_size >> 20); + + if (!changed_by_mtrr_cleanup) + WARN_ON(1); + + printk(KERN_INFO "update e820 for mtrr\n"); + update_e820(); + + return 1; + } + + return 0; +} + diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index 0c0a455fe95c..37f28fc7cf95 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -33,13 +33,31 @@ u64 mtrr_tom2; struct mtrr_state_type mtrr_state = {}; EXPORT_SYMBOL_GPL(mtrr_state); -static int __initdata mtrr_show; -static int __init mtrr_debug(char *opt) +/** + * BIOS is expected to clear MtrrFixDramModEn bit, see for example + * "BIOS and Kernel Developer's Guide for the AMD Athlon 64 and AMD + * Opteron Processors" (26094 Rev. 3.30 February 2006), section + * "13.2.1.2 SYSCFG Register": "The MtrrFixDramModEn bit should be set + * to 1 during BIOS initalization of the fixed MTRRs, then cleared to + * 0 for operation." + */ +static inline void k8_check_syscfg_dram_mod_en(void) { - mtrr_show = 1; - return 0; + u32 lo, hi; + + if (!((boot_cpu_data.x86_vendor == X86_VENDOR_AMD) && + (boot_cpu_data.x86 >= 0x0f))) + return; + + rdmsr(MSR_K8_SYSCFG, lo, hi); + if (lo & K8_MTRRFIXRANGE_DRAM_MODIFY) { + printk(KERN_ERR FW_WARN "MTRR: CPU %u: SYSCFG[MtrrFixDramModEn]" + " not cleared by BIOS, clearing this bit\n", + smp_processor_id()); + lo &= ~K8_MTRRFIXRANGE_DRAM_MODIFY; + mtrr_wrmsr(MSR_K8_SYSCFG, lo, hi); + } } -early_param("mtrr.show", mtrr_debug); /* * Returns the effective MTRR type for the region @@ -174,6 +192,8 @@ get_fixed_ranges(mtrr_type * frs) unsigned int *p = (unsigned int *) frs; int i; + k8_check_syscfg_dram_mod_en(); + rdmsr(MTRRfix64K_00000_MSR, p[0], p[1]); for (i = 0; i < 2; i++) @@ -188,18 +208,94 @@ void mtrr_save_fixed_ranges(void *info) get_fixed_ranges(mtrr_state.fixed_ranges); } -static void print_fixed(unsigned base, unsigned step, const mtrr_type*types) +static unsigned __initdata last_fixed_start; +static unsigned __initdata last_fixed_end; +static mtrr_type __initdata last_fixed_type; + +static void __init print_fixed_last(void) +{ + if (!last_fixed_end) + return; + + printk(KERN_DEBUG " %05X-%05X %s\n", last_fixed_start, + last_fixed_end - 1, mtrr_attrib_to_str(last_fixed_type)); + + last_fixed_end = 0; +} + +static void __init update_fixed_last(unsigned base, unsigned end, + mtrr_type type) +{ + last_fixed_start = base; + last_fixed_end = end; + last_fixed_type = type; +} + +static void __init print_fixed(unsigned base, unsigned step, + const mtrr_type *types) { unsigned i; - for (i = 0; i < 8; ++i, ++types, base += step) - printk(KERN_INFO "MTRR %05X-%05X %s\n", - base, base + step - 1, mtrr_attrib_to_str(*types)); + for (i = 0; i < 8; ++i, ++types, base += step) { + if (last_fixed_end == 0) { + update_fixed_last(base, base + step, *types); + continue; + } + if (last_fixed_end == base && last_fixed_type == *types) { + last_fixed_end = base + step; + continue; + } + /* new segments: gap or different type */ + print_fixed_last(); + update_fixed_last(base, base + step, *types); + } } static void prepare_set(void); static void post_set(void); +static void __init print_mtrr_state(void) +{ + unsigned int i; + int high_width; + + printk(KERN_DEBUG "MTRR default type: %s\n", + mtrr_attrib_to_str(mtrr_state.def_type)); + if (mtrr_state.have_fixed) { + printk(KERN_DEBUG "MTRR fixed ranges %sabled:\n", + mtrr_state.enabled & 1 ? "en" : "dis"); + print_fixed(0x00000, 0x10000, mtrr_state.fixed_ranges + 0); + for (i = 0; i < 2; ++i) + print_fixed(0x80000 + i * 0x20000, 0x04000, mtrr_state.fixed_ranges + (i + 1) * 8); + for (i = 0; i < 8; ++i) + print_fixed(0xC0000 + i * 0x08000, 0x01000, mtrr_state.fixed_ranges + (i + 3) * 8); + + /* tail */ + print_fixed_last(); + } + printk(KERN_DEBUG "MTRR variable ranges %sabled:\n", + mtrr_state.enabled & 2 ? "en" : "dis"); + high_width = ((size_or_mask ? ffs(size_or_mask) - 1 : 32) - (32 - PAGE_SHIFT) + 3) / 4; + for (i = 0; i < num_var_ranges; ++i) { + if (mtrr_state.var_ranges[i].mask_lo & (1 << 11)) + printk(KERN_DEBUG " %u base %0*X%05X000 mask %0*X%05X000 %s\n", + i, + high_width, + mtrr_state.var_ranges[i].base_hi, + mtrr_state.var_ranges[i].base_lo >> 12, + high_width, + mtrr_state.var_ranges[i].mask_hi, + mtrr_state.var_ranges[i].mask_lo >> 12, + mtrr_attrib_to_str(mtrr_state.var_ranges[i].base_lo & 0xff)); + else + printk(KERN_DEBUG " %u disabled\n", i); + } + if (mtrr_tom2) { + printk(KERN_DEBUG "TOM2: %016llx aka %lldM\n", + mtrr_tom2, mtrr_tom2>>20); + } +} + /* Grab all of the MTRR state for this CPU into *state */ void __init get_mtrr_state(void) { @@ -231,41 +327,9 @@ void __init get_mtrr_state(void) mtrr_tom2 |= low; mtrr_tom2 &= 0xffffff800000ULL; } - if (mtrr_show) { - int high_width; - - printk(KERN_INFO "MTRR default type: %s\n", mtrr_attrib_to_str(mtrr_state.def_type)); - if (mtrr_state.have_fixed) { - printk(KERN_INFO "MTRR fixed ranges %sabled:\n", - mtrr_state.enabled & 1 ? "en" : "dis"); - print_fixed(0x00000, 0x10000, mtrr_state.fixed_ranges + 0); - for (i = 0; i < 2; ++i) - print_fixed(0x80000 + i * 0x20000, 0x04000, mtrr_state.fixed_ranges + (i + 1) * 8); - for (i = 0; i < 8; ++i) - print_fixed(0xC0000 + i * 0x08000, 0x01000, mtrr_state.fixed_ranges + (i + 3) * 8); - } - printk(KERN_INFO "MTRR variable ranges %sabled:\n", - mtrr_state.enabled & 2 ? "en" : "dis"); - high_width = ((size_or_mask ? ffs(size_or_mask) - 1 : 32) - (32 - PAGE_SHIFT) + 3) / 4; - for (i = 0; i < num_var_ranges; ++i) { - if (mtrr_state.var_ranges[i].mask_lo & (1 << 11)) - printk(KERN_INFO "MTRR %u base %0*X%05X000 mask %0*X%05X000 %s\n", - i, - high_width, - mtrr_state.var_ranges[i].base_hi, - mtrr_state.var_ranges[i].base_lo >> 12, - high_width, - mtrr_state.var_ranges[i].mask_hi, - mtrr_state.var_ranges[i].mask_lo >> 12, - mtrr_attrib_to_str(mtrr_state.var_ranges[i].base_lo & 0xff)); - else - printk(KERN_INFO "MTRR %u disabled\n", i); - } - if (mtrr_tom2) { - printk(KERN_INFO "TOM2: %016llx aka %lldM\n", - mtrr_tom2, mtrr_tom2>>20); - } - } + + print_mtrr_state(); + mtrr_state_set = 1; /* PAT setup for BP. We need to go through sync steps here */ @@ -308,27 +372,10 @@ void mtrr_wrmsr(unsigned msr, unsigned a, unsigned b) } /** - * Enable and allow read/write of extended fixed-range MTRR bits on K8 CPUs - * see AMD publication no. 24593, chapter 3.2.1 for more information - */ -static inline void k8_enable_fixed_iorrs(void) -{ - unsigned lo, hi; - - rdmsr(MSR_K8_SYSCFG, lo, hi); - mtrr_wrmsr(MSR_K8_SYSCFG, lo - | K8_MTRRFIXRANGE_DRAM_ENABLE - | K8_MTRRFIXRANGE_DRAM_MODIFY, hi); -} - -/** * set_fixed_range - checks & updates a fixed-range MTRR if it differs from the value it should have * @msr: MSR address of the MTTR which should be checked and updated * @changed: pointer which indicates whether the MTRR needed to be changed * @msrwords: pointer to the MSR values which the MSR should have - * - * If K8 extentions are wanted, update the K8 SYSCFG MSR also. - * See AMD publication no. 24593, chapter 7.8.1, page 233 for more information. */ static void set_fixed_range(int msr, bool *changed, unsigned int *msrwords) { @@ -337,10 +384,6 @@ static void set_fixed_range(int msr, bool *changed, unsigned int *msrwords) rdmsr(msr, lo, hi); if (lo != msrwords[0] || hi != msrwords[1]) { - if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD && - (boot_cpu_data.x86 >= 0x0f && boot_cpu_data.x86 <= 0x11) && - ((msrwords[0] | msrwords[1]) & K8_MTRR_RDMEM_WRMEM_MASK)) - k8_enable_fixed_iorrs(); mtrr_wrmsr(msr, msrwords[0], msrwords[1]); *changed = true; } @@ -376,22 +419,31 @@ static void generic_get_mtrr(unsigned int reg, unsigned long *base, { unsigned int mask_lo, mask_hi, base_lo, base_hi; unsigned int tmp, hi; + int cpu; + + /* + * get_mtrr doesn't need to update mtrr_state, also it could be called + * from any cpu, so try to print it out directly. + */ + cpu = get_cpu(); rdmsr(MTRRphysMask_MSR(reg), mask_lo, mask_hi); + if ((mask_lo & 0x800) == 0) { /* Invalid (i.e. free) range */ *base = 0; *size = 0; *type = 0; - return; + goto out_put_cpu; } rdmsr(MTRRphysBase_MSR(reg), base_lo, base_hi); - /* Work out the shifted address mask. */ + /* Work out the shifted address mask: */ tmp = mask_hi << (32 - PAGE_SHIFT) | mask_lo >> PAGE_SHIFT; mask_lo = size_or_mask | tmp; - /* Expand tmp with high bits to all 1s*/ + + /* Expand tmp with high bits to all 1s: */ hi = fls(tmp); if (hi > 0) { tmp |= ~((1<<(hi - 1)) - 1); @@ -402,11 +454,19 @@ static void generic_get_mtrr(unsigned int reg, unsigned long *base, } } - /* This works correctly if size is a power of two, i.e. a - contiguous range. */ + /* + * This works correctly if size is a power of two, i.e. a + * contiguous range: + */ *size = -mask_lo; *base = base_hi << (32 - PAGE_SHIFT) | base_lo >> PAGE_SHIFT; *type = base_lo & 0xff; + + printk(KERN_DEBUG " get_mtrr: cpu%d reg%02d base=%010lx size=%010lx %s\n", + cpu, reg, *base, *size, + mtrr_attrib_to_str(*type & 0xff)); +out_put_cpu: + put_cpu(); } /** @@ -419,6 +479,8 @@ static int set_fixed_ranges(mtrr_type * frs) bool changed = false; int block=-1, range; + k8_check_syscfg_dram_mod_en(); + while (fixed_range_blocks[++block].ranges) for (range=0; range < fixed_range_blocks[block].ranges; range++) set_fixed_range(fixed_range_blocks[block].base_msr + range, diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index 236a401b8259..03cda01f57c7 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -574,7 +574,7 @@ struct mtrr_value { unsigned long lsize; }; -static struct mtrr_value mtrr_state[MTRR_MAX_VAR_RANGES]; +static struct mtrr_value mtrr_value[MTRR_MAX_VAR_RANGES]; static int mtrr_save(struct sys_device * sysdev, pm_message_t state) { @@ -582,9 +582,9 @@ static int mtrr_save(struct sys_device * sysdev, pm_message_t state) for (i = 0; i < num_var_ranges; i++) { mtrr_if->get(i, - &mtrr_state[i].lbase, - &mtrr_state[i].lsize, - &mtrr_state[i].ltype); + &mtrr_value[i].lbase, + &mtrr_value[i].lsize, + &mtrr_value[i].ltype); } return 0; } @@ -594,11 +594,11 @@ static int mtrr_restore(struct sys_device * sysdev) int i; for (i = 0; i < num_var_ranges; i++) { - if (mtrr_state[i].lsize) + if (mtrr_value[i].lsize) set_mtrr(i, - mtrr_state[i].lbase, - mtrr_state[i].lsize, - mtrr_state[i].ltype); + mtrr_value[i].lbase, + mtrr_value[i].lsize, + mtrr_value[i].ltype); } return 0; } @@ -610,1058 +610,7 @@ static struct sysdev_driver mtrr_sysdev_driver = { .resume = mtrr_restore, }; -/* should be related to MTRR_VAR_RANGES nums */ -#define RANGE_NUM 256 - -struct res_range { - unsigned long start; - unsigned long end; -}; - -static int __init -add_range(struct res_range *range, int nr_range, unsigned long start, - unsigned long end) -{ - /* out of slots */ - if (nr_range >= RANGE_NUM) - return nr_range; - - range[nr_range].start = start; - range[nr_range].end = end; - - nr_range++; - - return nr_range; -} - -static int __init -add_range_with_merge(struct res_range *range, int nr_range, unsigned long start, - unsigned long end) -{ - int i; - - /* try to merge it with old one */ - for (i = 0; i < nr_range; i++) { - unsigned long final_start, final_end; - unsigned long common_start, common_end; - - if (!range[i].end) - continue; - - common_start = max(range[i].start, start); - common_end = min(range[i].end, end); - if (common_start > common_end + 1) - continue; - - final_start = min(range[i].start, start); - final_end = max(range[i].end, end); - - range[i].start = final_start; - range[i].end = final_end; - return nr_range; - } - - /* need to add that */ - return add_range(range, nr_range, start, end); -} - -static void __init -subtract_range(struct res_range *range, unsigned long start, unsigned long end) -{ - int i, j; - - for (j = 0; j < RANGE_NUM; j++) { - if (!range[j].end) - continue; - - if (start <= range[j].start && end >= range[j].end) { - range[j].start = 0; - range[j].end = 0; - continue; - } - - if (start <= range[j].start && end < range[j].end && - range[j].start < end + 1) { - range[j].start = end + 1; - continue; - } - - - if (start > range[j].start && end >= range[j].end && - range[j].end > start - 1) { - range[j].end = start - 1; - continue; - } - - if (start > range[j].start && end < range[j].end) { - /* find the new spare */ - for (i = 0; i < RANGE_NUM; i++) { - if (range[i].end == 0) - break; - } - if (i < RANGE_NUM) { - range[i].end = range[j].end; - range[i].start = end + 1; - } else { - printk(KERN_ERR "run of slot in ranges\n"); - } - range[j].end = start - 1; - continue; - } - } -} - -static int __init cmp_range(const void *x1, const void *x2) -{ - const struct res_range *r1 = x1; - const struct res_range *r2 = x2; - long start1, start2; - - start1 = r1->start; - start2 = r2->start; - - return start1 - start2; -} - -struct var_mtrr_range_state { - unsigned long base_pfn; - unsigned long size_pfn; - mtrr_type type; -}; - -static struct var_mtrr_range_state __initdata range_state[RANGE_NUM]; -static int __initdata debug_print; - -static int __init -x86_get_mtrr_mem_range(struct res_range *range, int nr_range, - unsigned long extra_remove_base, - unsigned long extra_remove_size) -{ - unsigned long i, base, size; - mtrr_type type; - - for (i = 0; i < num_var_ranges; i++) { - type = range_state[i].type; - if (type != MTRR_TYPE_WRBACK) - continue; - base = range_state[i].base_pfn; - size = range_state[i].size_pfn; - nr_range = add_range_with_merge(range, nr_range, base, - base + size - 1); - } - if (debug_print) { - printk(KERN_DEBUG "After WB checking\n"); - for (i = 0; i < nr_range; i++) - printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", - range[i].start, range[i].end + 1); - } - - /* take out UC ranges */ - for (i = 0; i < num_var_ranges; i++) { - type = range_state[i].type; - if (type != MTRR_TYPE_UNCACHABLE && - type != MTRR_TYPE_WRPROT) - continue; - size = range_state[i].size_pfn; - if (!size) - continue; - base = range_state[i].base_pfn; - subtract_range(range, base, base + size - 1); - } - if (extra_remove_size) - subtract_range(range, extra_remove_base, - extra_remove_base + extra_remove_size - 1); - - /* get new range num */ - nr_range = 0; - for (i = 0; i < RANGE_NUM; i++) { - if (!range[i].end) - continue; - nr_range++; - } - if (debug_print) { - printk(KERN_DEBUG "After UC checking\n"); - for (i = 0; i < nr_range; i++) - printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", - range[i].start, range[i].end + 1); - } - - /* sort the ranges */ - sort(range, nr_range, sizeof(struct res_range), cmp_range, NULL); - if (debug_print) { - printk(KERN_DEBUG "After sorting\n"); - for (i = 0; i < nr_range; i++) - printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", - range[i].start, range[i].end + 1); - } - - /* clear those is not used */ - for (i = nr_range; i < RANGE_NUM; i++) - memset(&range[i], 0, sizeof(range[i])); - - return nr_range; -} - -static struct res_range __initdata range[RANGE_NUM]; -static int __initdata nr_range; - -#ifdef CONFIG_MTRR_SANITIZER - -static unsigned long __init sum_ranges(struct res_range *range, int nr_range) -{ - unsigned long sum; - int i; - - sum = 0; - for (i = 0; i < nr_range; i++) - sum += range[i].end + 1 - range[i].start; - - return sum; -} - -static int enable_mtrr_cleanup __initdata = - CONFIG_MTRR_SANITIZER_ENABLE_DEFAULT; - -static int __init disable_mtrr_cleanup_setup(char *str) -{ - enable_mtrr_cleanup = 0; - return 0; -} -early_param("disable_mtrr_cleanup", disable_mtrr_cleanup_setup); - -static int __init enable_mtrr_cleanup_setup(char *str) -{ - enable_mtrr_cleanup = 1; - return 0; -} -early_param("enable_mtrr_cleanup", enable_mtrr_cleanup_setup); - -static int __init mtrr_cleanup_debug_setup(char *str) -{ - debug_print = 1; - return 0; -} -early_param("mtrr_cleanup_debug", mtrr_cleanup_debug_setup); - -struct var_mtrr_state { - unsigned long range_startk; - unsigned long range_sizek; - unsigned long chunk_sizek; - unsigned long gran_sizek; - unsigned int reg; -}; - -static void __init -set_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, - unsigned char type, unsigned int address_bits) -{ - u32 base_lo, base_hi, mask_lo, mask_hi; - u64 base, mask; - - if (!sizek) { - fill_mtrr_var_range(reg, 0, 0, 0, 0); - return; - } - - mask = (1ULL << address_bits) - 1; - mask &= ~((((u64)sizek) << 10) - 1); - - base = ((u64)basek) << 10; - - base |= type; - mask |= 0x800; - - base_lo = base & ((1ULL<<32) - 1); - base_hi = base >> 32; - - mask_lo = mask & ((1ULL<<32) - 1); - mask_hi = mask >> 32; - - fill_mtrr_var_range(reg, base_lo, base_hi, mask_lo, mask_hi); -} - -static void __init -save_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, - unsigned char type) -{ - range_state[reg].base_pfn = basek >> (PAGE_SHIFT - 10); - range_state[reg].size_pfn = sizek >> (PAGE_SHIFT - 10); - range_state[reg].type = type; -} - -static void __init -set_var_mtrr_all(unsigned int address_bits) -{ - unsigned long basek, sizek; - unsigned char type; - unsigned int reg; - - for (reg = 0; reg < num_var_ranges; reg++) { - basek = range_state[reg].base_pfn << (PAGE_SHIFT - 10); - sizek = range_state[reg].size_pfn << (PAGE_SHIFT - 10); - type = range_state[reg].type; - - set_var_mtrr(reg, basek, sizek, type, address_bits); - } -} - -static unsigned long to_size_factor(unsigned long sizek, char *factorp) -{ - char factor; - unsigned long base = sizek; - - if (base & ((1<<10) - 1)) { - /* not MB alignment */ - factor = 'K'; - } else if (base & ((1<<20) - 1)){ - factor = 'M'; - base >>= 10; - } else { - factor = 'G'; - base >>= 20; - } - - *factorp = factor; - - return base; -} - -static unsigned int __init -range_to_mtrr(unsigned int reg, unsigned long range_startk, - unsigned long range_sizek, unsigned char type) -{ - if (!range_sizek || (reg >= num_var_ranges)) - return reg; - - while (range_sizek) { - unsigned long max_align, align; - unsigned long sizek; - - /* Compute the maximum size I can make a range */ - if (range_startk) - max_align = ffs(range_startk) - 1; - else - max_align = 32; - align = fls(range_sizek) - 1; - if (align > max_align) - align = max_align; - - sizek = 1 << align; - if (debug_print) { - char start_factor = 'K', size_factor = 'K'; - unsigned long start_base, size_base; - - start_base = to_size_factor(range_startk, &start_factor), - size_base = to_size_factor(sizek, &size_factor), - - printk(KERN_DEBUG "Setting variable MTRR %d, " - "base: %ld%cB, range: %ld%cB, type %s\n", - reg, start_base, start_factor, - size_base, size_factor, - (type == MTRR_TYPE_UNCACHABLE)?"UC": - ((type == MTRR_TYPE_WRBACK)?"WB":"Other") - ); - } - save_var_mtrr(reg++, range_startk, sizek, type); - range_startk += sizek; - range_sizek -= sizek; - if (reg >= num_var_ranges) - break; - } - return reg; -} - -static unsigned __init -range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek, - unsigned long sizek) -{ - unsigned long hole_basek, hole_sizek; - unsigned long second_basek, second_sizek; - unsigned long range0_basek, range0_sizek; - unsigned long range_basek, range_sizek; - unsigned long chunk_sizek; - unsigned long gran_sizek; - - hole_basek = 0; - hole_sizek = 0; - second_basek = 0; - second_sizek = 0; - chunk_sizek = state->chunk_sizek; - gran_sizek = state->gran_sizek; - - /* align with gran size, prevent small block used up MTRRs */ - range_basek = ALIGN(state->range_startk, gran_sizek); - if ((range_basek > basek) && basek) - return second_sizek; - state->range_sizek -= (range_basek - state->range_startk); - range_sizek = ALIGN(state->range_sizek, gran_sizek); - - while (range_sizek > state->range_sizek) { - range_sizek -= gran_sizek; - if (!range_sizek) - return 0; - } - state->range_sizek = range_sizek; - - /* try to append some small hole */ - range0_basek = state->range_startk; - range0_sizek = ALIGN(state->range_sizek, chunk_sizek); - - /* no increase */ - if (range0_sizek == state->range_sizek) { - if (debug_print) - printk(KERN_DEBUG "rangeX: %016lx - %016lx\n", - range0_basek<<10, - (range0_basek + state->range_sizek)<<10); - state->reg = range_to_mtrr(state->reg, range0_basek, - state->range_sizek, MTRR_TYPE_WRBACK); - return 0; - } - - /* only cut back, when it is not the last */ - if (sizek) { - while (range0_basek + range0_sizek > (basek + sizek)) { - if (range0_sizek >= chunk_sizek) - range0_sizek -= chunk_sizek; - else - range0_sizek = 0; - - if (!range0_sizek) - break; - } - } - -second_try: - range_basek = range0_basek + range0_sizek; - - /* one hole in the middle */ - if (range_basek > basek && range_basek <= (basek + sizek)) - second_sizek = range_basek - basek; - - if (range0_sizek > state->range_sizek) { - - /* one hole in middle or at end */ - hole_sizek = range0_sizek - state->range_sizek - second_sizek; - - /* hole size should be less than half of range0 size */ - if (hole_sizek >= (range0_sizek >> 1) && - range0_sizek >= chunk_sizek) { - range0_sizek -= chunk_sizek; - second_sizek = 0; - hole_sizek = 0; - - goto second_try; - } - } - - if (range0_sizek) { - if (debug_print) - printk(KERN_DEBUG "range0: %016lx - %016lx\n", - range0_basek<<10, - (range0_basek + range0_sizek)<<10); - state->reg = range_to_mtrr(state->reg, range0_basek, - range0_sizek, MTRR_TYPE_WRBACK); - } - - if (range0_sizek < state->range_sizek) { - /* need to handle left over */ - range_sizek = state->range_sizek - range0_sizek; - - if (debug_print) - printk(KERN_DEBUG "range: %016lx - %016lx\n", - range_basek<<10, - (range_basek + range_sizek)<<10); - state->reg = range_to_mtrr(state->reg, range_basek, - range_sizek, MTRR_TYPE_WRBACK); - } - - if (hole_sizek) { - hole_basek = range_basek - hole_sizek - second_sizek; - if (debug_print) - printk(KERN_DEBUG "hole: %016lx - %016lx\n", - hole_basek<<10, - (hole_basek + hole_sizek)<<10); - state->reg = range_to_mtrr(state->reg, hole_basek, - hole_sizek, MTRR_TYPE_UNCACHABLE); - } - - return second_sizek; -} - -static void __init -set_var_mtrr_range(struct var_mtrr_state *state, unsigned long base_pfn, - unsigned long size_pfn) -{ - unsigned long basek, sizek; - unsigned long second_sizek = 0; - - if (state->reg >= num_var_ranges) - return; - - basek = base_pfn << (PAGE_SHIFT - 10); - sizek = size_pfn << (PAGE_SHIFT - 10); - - /* See if I can merge with the last range */ - if ((basek <= 1024) || - (state->range_startk + state->range_sizek == basek)) { - unsigned long endk = basek + sizek; - state->range_sizek = endk - state->range_startk; - return; - } - /* Write the range mtrrs */ - if (state->range_sizek != 0) - second_sizek = range_to_mtrr_with_hole(state, basek, sizek); - - /* Allocate an msr */ - state->range_startk = basek + second_sizek; - state->range_sizek = sizek - second_sizek; -} - -/* mininum size of mtrr block that can take hole */ -static u64 mtrr_chunk_size __initdata = (256ULL<<20); - -static int __init parse_mtrr_chunk_size_opt(char *p) -{ - if (!p) - return -EINVAL; - mtrr_chunk_size = memparse(p, &p); - return 0; -} -early_param("mtrr_chunk_size", parse_mtrr_chunk_size_opt); - -/* granity of mtrr of block */ -static u64 mtrr_gran_size __initdata; - -static int __init parse_mtrr_gran_size_opt(char *p) -{ - if (!p) - return -EINVAL; - mtrr_gran_size = memparse(p, &p); - return 0; -} -early_param("mtrr_gran_size", parse_mtrr_gran_size_opt); - -static int nr_mtrr_spare_reg __initdata = - CONFIG_MTRR_SANITIZER_SPARE_REG_NR_DEFAULT; - -static int __init parse_mtrr_spare_reg(char *arg) -{ - if (arg) - nr_mtrr_spare_reg = simple_strtoul(arg, NULL, 0); - return 0; -} - -early_param("mtrr_spare_reg_nr", parse_mtrr_spare_reg); - -static int __init -x86_setup_var_mtrrs(struct res_range *range, int nr_range, - u64 chunk_size, u64 gran_size) -{ - struct var_mtrr_state var_state; - int i; - int num_reg; - - var_state.range_startk = 0; - var_state.range_sizek = 0; - var_state.reg = 0; - var_state.chunk_sizek = chunk_size >> 10; - var_state.gran_sizek = gran_size >> 10; - - memset(range_state, 0, sizeof(range_state)); - - /* Write the range etc */ - for (i = 0; i < nr_range; i++) - set_var_mtrr_range(&var_state, range[i].start, - range[i].end - range[i].start + 1); - - /* Write the last range */ - if (var_state.range_sizek != 0) - range_to_mtrr_with_hole(&var_state, 0, 0); - - num_reg = var_state.reg; - /* Clear out the extra MTRR's */ - while (var_state.reg < num_var_ranges) { - save_var_mtrr(var_state.reg, 0, 0, 0); - var_state.reg++; - } - - return num_reg; -} - -struct mtrr_cleanup_result { - unsigned long gran_sizek; - unsigned long chunk_sizek; - unsigned long lose_cover_sizek; - unsigned int num_reg; - int bad; -}; - -/* - * gran_size: 64K, 128K, 256K, 512K, 1M, 2M, ..., 2G - * chunk size: gran_size, ..., 2G - * so we need (1+16)*8 - */ -#define NUM_RESULT 136 -#define PSHIFT (PAGE_SHIFT - 10) - -static struct mtrr_cleanup_result __initdata result[NUM_RESULT]; -static unsigned long __initdata min_loss_pfn[RANGE_NUM]; - -static void __init print_out_mtrr_range_state(void) -{ - int i; - char start_factor = 'K', size_factor = 'K'; - unsigned long start_base, size_base; - mtrr_type type; - - for (i = 0; i < num_var_ranges; i++) { - - size_base = range_state[i].size_pfn << (PAGE_SHIFT - 10); - if (!size_base) - continue; - - size_base = to_size_factor(size_base, &size_factor), - start_base = range_state[i].base_pfn << (PAGE_SHIFT - 10); - start_base = to_size_factor(start_base, &start_factor), - type = range_state[i].type; - - printk(KERN_DEBUG "reg %d, base: %ld%cB, range: %ld%cB, type %s\n", - i, start_base, start_factor, - size_base, size_factor, - (type == MTRR_TYPE_UNCACHABLE) ? "UC" : - ((type == MTRR_TYPE_WRPROT) ? "WP" : - ((type == MTRR_TYPE_WRBACK) ? "WB" : "Other")) - ); - } -} - -static int __init mtrr_need_cleanup(void) -{ - int i; - mtrr_type type; - unsigned long size; - /* extra one for all 0 */ - int num[MTRR_NUM_TYPES + 1]; - - /* check entries number */ - memset(num, 0, sizeof(num)); - for (i = 0; i < num_var_ranges; i++) { - type = range_state[i].type; - size = range_state[i].size_pfn; - if (type >= MTRR_NUM_TYPES) - continue; - if (!size) - type = MTRR_NUM_TYPES; - if (type == MTRR_TYPE_WRPROT) - type = MTRR_TYPE_UNCACHABLE; - num[type]++; - } - - /* check if we got UC entries */ - if (!num[MTRR_TYPE_UNCACHABLE]) - return 0; - - /* check if we only had WB and UC */ - if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != - num_var_ranges - num[MTRR_NUM_TYPES]) - return 0; - - return 1; -} - -static unsigned long __initdata range_sums; -static void __init mtrr_calc_range_state(u64 chunk_size, u64 gran_size, - unsigned long extra_remove_base, - unsigned long extra_remove_size, - int i) -{ - int num_reg; - static struct res_range range_new[RANGE_NUM]; - static int nr_range_new; - unsigned long range_sums_new; - - /* convert ranges to var ranges state */ - num_reg = x86_setup_var_mtrrs(range, nr_range, - chunk_size, gran_size); - - /* we got new setting in range_state, check it */ - memset(range_new, 0, sizeof(range_new)); - nr_range_new = x86_get_mtrr_mem_range(range_new, 0, - extra_remove_base, extra_remove_size); - range_sums_new = sum_ranges(range_new, nr_range_new); - - result[i].chunk_sizek = chunk_size >> 10; - result[i].gran_sizek = gran_size >> 10; - result[i].num_reg = num_reg; - if (range_sums < range_sums_new) { - result[i].lose_cover_sizek = - (range_sums_new - range_sums) << PSHIFT; - result[i].bad = 1; - } else - result[i].lose_cover_sizek = - (range_sums - range_sums_new) << PSHIFT; - - /* double check it */ - if (!result[i].bad && !result[i].lose_cover_sizek) { - if (nr_range_new != nr_range || - memcmp(range, range_new, sizeof(range))) - result[i].bad = 1; - } - - if (!result[i].bad && (range_sums - range_sums_new < - min_loss_pfn[num_reg])) { - min_loss_pfn[num_reg] = - range_sums - range_sums_new; - } -} - -static void __init mtrr_print_out_one_result(int i) -{ - char gran_factor, chunk_factor, lose_factor; - unsigned long gran_base, chunk_base, lose_base; - - gran_base = to_size_factor(result[i].gran_sizek, &gran_factor), - chunk_base = to_size_factor(result[i].chunk_sizek, &chunk_factor), - lose_base = to_size_factor(result[i].lose_cover_sizek, &lose_factor), - printk(KERN_INFO "%sgran_size: %ld%c \tchunk_size: %ld%c \t", - result[i].bad ? "*BAD*" : " ", - gran_base, gran_factor, chunk_base, chunk_factor); - printk(KERN_CONT "num_reg: %d \tlose cover RAM: %s%ld%c\n", - result[i].num_reg, result[i].bad ? "-" : "", - lose_base, lose_factor); -} - -static int __init mtrr_search_optimal_index(void) -{ - int i; - int num_reg_good; - int index_good; - - if (nr_mtrr_spare_reg >= num_var_ranges) - nr_mtrr_spare_reg = num_var_ranges - 1; - num_reg_good = -1; - for (i = num_var_ranges - nr_mtrr_spare_reg; i > 0; i--) { - if (!min_loss_pfn[i]) - num_reg_good = i; - } - - index_good = -1; - if (num_reg_good != -1) { - for (i = 0; i < NUM_RESULT; i++) { - if (!result[i].bad && - result[i].num_reg == num_reg_good && - !result[i].lose_cover_sizek) { - index_good = i; - break; - } - } - } - - return index_good; -} - - -static int __init mtrr_cleanup(unsigned address_bits) -{ - unsigned long extra_remove_base, extra_remove_size; - unsigned long base, size, def, dummy; - mtrr_type type; - u64 chunk_size, gran_size; - int index_good; - int i; - - if (!is_cpu(INTEL) || enable_mtrr_cleanup < 1) - return 0; - rdmsr(MTRRdefType_MSR, def, dummy); - def &= 0xff; - if (def != MTRR_TYPE_UNCACHABLE) - return 0; - - /* get it and store it aside */ - memset(range_state, 0, sizeof(range_state)); - for (i = 0; i < num_var_ranges; i++) { - mtrr_if->get(i, &base, &size, &type); - range_state[i].base_pfn = base; - range_state[i].size_pfn = size; - range_state[i].type = type; - } - - /* check if we need handle it and can handle it */ - if (!mtrr_need_cleanup()) - return 0; - - /* print original var MTRRs at first, for debugging: */ - printk(KERN_DEBUG "original variable MTRRs\n"); - print_out_mtrr_range_state(); - - memset(range, 0, sizeof(range)); - extra_remove_size = 0; - extra_remove_base = 1 << (32 - PAGE_SHIFT); - if (mtrr_tom2) - extra_remove_size = - (mtrr_tom2 >> PAGE_SHIFT) - extra_remove_base; - nr_range = x86_get_mtrr_mem_range(range, 0, extra_remove_base, - extra_remove_size); - /* - * [0, 1M) should always be coverred by var mtrr with WB - * and fixed mtrrs should take effective before var mtrr for it - */ - nr_range = add_range_with_merge(range, nr_range, 0, - (1ULL<<(20 - PAGE_SHIFT)) - 1); - /* sort the ranges */ - sort(range, nr_range, sizeof(struct res_range), cmp_range, NULL); - - range_sums = sum_ranges(range, nr_range); - printk(KERN_INFO "total RAM coverred: %ldM\n", - range_sums >> (20 - PAGE_SHIFT)); - - if (mtrr_chunk_size && mtrr_gran_size) { - i = 0; - mtrr_calc_range_state(mtrr_chunk_size, mtrr_gran_size, - extra_remove_base, extra_remove_size, i); - - mtrr_print_out_one_result(i); - - if (!result[i].bad) { - set_var_mtrr_all(address_bits); - return 1; - } - printk(KERN_INFO "invalid mtrr_gran_size or mtrr_chunk_size, " - "will find optimal one\n"); - } - - i = 0; - memset(min_loss_pfn, 0xff, sizeof(min_loss_pfn)); - memset(result, 0, sizeof(result)); - for (gran_size = (1ULL<<16); gran_size < (1ULL<<32); gran_size <<= 1) { - - for (chunk_size = gran_size; chunk_size < (1ULL<<32); - chunk_size <<= 1) { - - if (i >= NUM_RESULT) - continue; - - mtrr_calc_range_state(chunk_size, gran_size, - extra_remove_base, extra_remove_size, i); - if (debug_print) { - mtrr_print_out_one_result(i); - printk(KERN_INFO "\n"); - } - - i++; - } - } - - /* try to find the optimal index */ - index_good = mtrr_search_optimal_index(); - - if (index_good != -1) { - printk(KERN_INFO "Found optimal setting for mtrr clean up\n"); - i = index_good; - mtrr_print_out_one_result(i); - - /* convert ranges to var ranges state */ - chunk_size = result[i].chunk_sizek; - chunk_size <<= 10; - gran_size = result[i].gran_sizek; - gran_size <<= 10; - x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size); - set_var_mtrr_all(address_bits); - printk(KERN_DEBUG "New variable MTRRs\n"); - print_out_mtrr_range_state(); - return 1; - } else { - /* print out all */ - for (i = 0; i < NUM_RESULT; i++) - mtrr_print_out_one_result(i); - } - - printk(KERN_INFO "mtrr_cleanup: can not find optimal value\n"); - printk(KERN_INFO "please specify mtrr_gran_size/mtrr_chunk_size\n"); - - return 0; -} -#else -static int __init mtrr_cleanup(unsigned address_bits) -{ - return 0; -} -#endif - -static int __initdata changed_by_mtrr_cleanup; - -static int disable_mtrr_trim; - -static int __init disable_mtrr_trim_setup(char *str) -{ - disable_mtrr_trim = 1; - return 0; -} -early_param("disable_mtrr_trim", disable_mtrr_trim_setup); - -/* - * Newer AMD K8s and later CPUs have a special magic MSR way to force WB - * for memory >4GB. Check for that here. - * Note this won't check if the MTRRs < 4GB where the magic bit doesn't - * apply to are wrong, but so far we don't know of any such case in the wild. - */ -#define Tom2Enabled (1U << 21) -#define Tom2ForceMemTypeWB (1U << 22) - -int __init amd_special_default_mtrr(void) -{ - u32 l, h; - - if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) - return 0; - if (boot_cpu_data.x86 < 0xf || boot_cpu_data.x86 > 0x11) - return 0; - /* In case some hypervisor doesn't pass SYSCFG through */ - if (rdmsr_safe(MSR_K8_SYSCFG, &l, &h) < 0) - return 0; - /* - * Memory between 4GB and top of mem is forced WB by this magic bit. - * Reserved before K8RevF, but should be zero there. - */ - if ((l & (Tom2Enabled | Tom2ForceMemTypeWB)) == - (Tom2Enabled | Tom2ForceMemTypeWB)) - return 1; - return 0; -} - -static u64 __init real_trim_memory(unsigned long start_pfn, - unsigned long limit_pfn) -{ - u64 trim_start, trim_size; - trim_start = start_pfn; - trim_start <<= PAGE_SHIFT; - trim_size = limit_pfn; - trim_size <<= PAGE_SHIFT; - trim_size -= trim_start; - - return e820_update_range(trim_start, trim_size, E820_RAM, - E820_RESERVED); -} -/** - * mtrr_trim_uncached_memory - trim RAM not covered by MTRRs - * @end_pfn: ending page frame number - * - * Some buggy BIOSes don't setup the MTRRs properly for systems with certain - * memory configurations. This routine checks that the highest MTRR matches - * the end of memory, to make sure the MTRRs having a write back type cover - * all of the memory the kernel is intending to use. If not, it'll trim any - * memory off the end by adjusting end_pfn, removing it from the kernel's - * allocation pools, warning the user with an obnoxious message. - */ -int __init mtrr_trim_uncached_memory(unsigned long end_pfn) -{ - unsigned long i, base, size, highest_pfn = 0, def, dummy; - mtrr_type type; - u64 total_trim_size; - - /* extra one for all 0 */ - int num[MTRR_NUM_TYPES + 1]; - /* - * Make sure we only trim uncachable memory on machines that - * support the Intel MTRR architecture: - */ - if (!is_cpu(INTEL) || disable_mtrr_trim) - return 0; - rdmsr(MTRRdefType_MSR, def, dummy); - def &= 0xff; - if (def != MTRR_TYPE_UNCACHABLE) - return 0; - - /* get it and store it aside */ - memset(range_state, 0, sizeof(range_state)); - for (i = 0; i < num_var_ranges; i++) { - mtrr_if->get(i, &base, &size, &type); - range_state[i].base_pfn = base; - range_state[i].size_pfn = size; - range_state[i].type = type; - } - - /* Find highest cached pfn */ - for (i = 0; i < num_var_ranges; i++) { - type = range_state[i].type; - if (type != MTRR_TYPE_WRBACK) - continue; - base = range_state[i].base_pfn; - size = range_state[i].size_pfn; - if (highest_pfn < base + size) - highest_pfn = base + size; - } - - /* kvm/qemu doesn't have mtrr set right, don't trim them all */ - if (!highest_pfn) { - printk(KERN_INFO "CPU MTRRs all blank - virtualized system.\n"); - return 0; - } - - /* check entries number */ - memset(num, 0, sizeof(num)); - for (i = 0; i < num_var_ranges; i++) { - type = range_state[i].type; - if (type >= MTRR_NUM_TYPES) - continue; - size = range_state[i].size_pfn; - if (!size) - type = MTRR_NUM_TYPES; - num[type]++; - } - - /* no entry for WB? */ - if (!num[MTRR_TYPE_WRBACK]) - return 0; - - /* check if we only had WB and UC */ - if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != - num_var_ranges - num[MTRR_NUM_TYPES]) - return 0; - - memset(range, 0, sizeof(range)); - nr_range = 0; - if (mtrr_tom2) { - range[nr_range].start = (1ULL<<(32 - PAGE_SHIFT)); - range[nr_range].end = (mtrr_tom2 >> PAGE_SHIFT) - 1; - if (highest_pfn < range[nr_range].end + 1) - highest_pfn = range[nr_range].end + 1; - nr_range++; - } - nr_range = x86_get_mtrr_mem_range(range, nr_range, 0, 0); - - total_trim_size = 0; - /* check the head */ - if (range[0].start) - total_trim_size += real_trim_memory(0, range[0].start); - /* check the holes */ - for (i = 0; i < nr_range - 1; i++) { - if (range[i].end + 1 < range[i+1].start) - total_trim_size += real_trim_memory(range[i].end + 1, - range[i+1].start); - } - /* check the top */ - i = nr_range - 1; - if (range[i].end + 1 < end_pfn) - total_trim_size += real_trim_memory(range[i].end + 1, - end_pfn); - - if (total_trim_size) { - printk(KERN_WARNING "WARNING: BIOS bug: CPU MTRRs don't cover" - " all of memory, losing %lluMB of RAM.\n", - total_trim_size >> 20); - - if (!changed_by_mtrr_cleanup) - WARN_ON(1); - - printk(KERN_INFO "update e820 for mtrr\n"); - update_e820(); - - return 1; - } - - return 0; -} +int __initdata changed_by_mtrr_cleanup; /** * mtrr_bp_init - initialize mtrrs on the boot CPU diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.h b/arch/x86/kernel/cpu/mtrr/mtrr.h index ffd60409cc6d..77f67f7b347a 100644 --- a/arch/x86/kernel/cpu/mtrr/mtrr.h +++ b/arch/x86/kernel/cpu/mtrr/mtrr.h @@ -79,6 +79,7 @@ extern struct mtrr_ops * mtrr_if; extern unsigned int num_var_ranges; extern u64 mtrr_tom2; +extern struct mtrr_state_type mtrr_state; void mtrr_state_warn(void); const char *mtrr_attrib_to_str(int x); @@ -88,3 +89,6 @@ void mtrr_wrmsr(unsigned, unsigned, unsigned); int amd_init_mtrr(void); int cyrix_init_mtrr(void); int centaur_init_mtrr(void); + +extern int changed_by_mtrr_cleanup; +extern int mtrr_cleanup(unsigned address_bits); diff --git a/arch/x86/kernel/cpu/transmeta.c b/arch/x86/kernel/cpu/transmeta.c index 52b3fefbd5af..bb62b3e5caad 100644 --- a/arch/x86/kernel/cpu/transmeta.c +++ b/arch/x86/kernel/cpu/transmeta.c @@ -98,7 +98,7 @@ static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) #endif } -static struct cpu_dev transmeta_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst transmeta_cpu_dev = { .c_vendor = "Transmeta", .c_ident = { "GenuineTMx86", "TransmetaCPU" }, .c_early_init = early_init_transmeta, diff --git a/arch/x86/kernel/cpu/umc.c b/arch/x86/kernel/cpu/umc.c index e777f79e0960..fd2c37bf7acb 100644 --- a/arch/x86/kernel/cpu/umc.c +++ b/arch/x86/kernel/cpu/umc.c @@ -8,7 +8,7 @@ * so no special init takes place. */ -static struct cpu_dev umc_cpu_dev __cpuinitdata = { +static const struct cpu_dev __cpuinitconst umc_cpu_dev = { .c_vendor = "UMC", .c_ident = { "UMC UMC UMC" }, .c_models = { diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 508bec1cee27..ef2c3563357d 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -110,19 +110,50 @@ int __init e820_all_mapped(u64 start, u64 end, unsigned type) /* * Add a memory region to the kernel e820 map. */ -void __init e820_add_region(u64 start, u64 size, int type) +static void __init __e820_add_region(struct e820map *e820x, u64 start, u64 size, + int type) { - int x = e820.nr_map; + int x = e820x->nr_map; - if (x == ARRAY_SIZE(e820.map)) { + if (x == ARRAY_SIZE(e820x->map)) { printk(KERN_ERR "Ooops! Too many entries in the memory map!\n"); return; } - e820.map[x].addr = start; - e820.map[x].size = size; - e820.map[x].type = type; - e820.nr_map++; + e820x->map[x].addr = start; + e820x->map[x].size = size; + e820x->map[x].type = type; + e820x->nr_map++; +} + +void __init e820_add_region(u64 start, u64 size, int type) +{ + __e820_add_region(&e820, start, size, type); +} + +static void __init e820_print_type(u32 type) +{ + switch (type) { + case E820_RAM: + case E820_RESERVED_KERN: + printk(KERN_CONT "(usable)"); + break; + case E820_RESERVED: + printk(KERN_CONT "(reserved)"); + break; + case E820_ACPI: + printk(KERN_CONT "(ACPI data)"); + break; + case E820_NVS: + printk(KERN_CONT "(ACPI NVS)"); + break; + case E820_UNUSABLE: + printk(KERN_CONT "(unusable)"); + break; + default: + printk(KERN_CONT "type %u", type); + break; + } } void __init e820_print_map(char *who) @@ -134,27 +165,8 @@ void __init e820_print_map(char *who) (unsigned long long) e820.map[i].addr, (unsigned long long) (e820.map[i].addr + e820.map[i].size)); - switch (e820.map[i].type) { - case E820_RAM: - case E820_RESERVED_KERN: - printk(KERN_CONT "(usable)\n"); - break; - case E820_RESERVED: - printk(KERN_CONT "(reserved)\n"); - break; - case E820_ACPI: - printk(KERN_CONT "(ACPI data)\n"); - break; - case E820_NVS: - printk(KERN_CONT "(ACPI NVS)\n"); - break; - case E820_UNUSABLE: - printk("(unusable)\n"); - break; - default: - printk(KERN_CONT "type %u\n", e820.map[i].type); - break; - } + e820_print_type(e820.map[i].type); + printk(KERN_CONT "\n"); } } @@ -221,7 +233,7 @@ void __init e820_print_map(char *who) */ int __init sanitize_e820_map(struct e820entry *biosmap, int max_nr_map, - int *pnr_map) + u32 *pnr_map) { struct change_member { struct e820entry *pbios; /* pointer to original bios entry */ @@ -417,11 +429,12 @@ static int __init append_e820_map(struct e820entry *biosmap, int nr_map) return __append_e820_map(biosmap, nr_map); } -static u64 __init e820_update_range_map(struct e820map *e820x, u64 start, +static u64 __init __e820_update_range(struct e820map *e820x, u64 start, u64 size, unsigned old_type, unsigned new_type) { - int i; + u64 end; + unsigned int i; u64 real_updated_size = 0; BUG_ON(old_type == new_type); @@ -429,27 +442,55 @@ static u64 __init e820_update_range_map(struct e820map *e820x, u64 start, if (size > (ULLONG_MAX - start)) size = ULLONG_MAX - start; - for (i = 0; i < e820.nr_map; i++) { + end = start + size; + printk(KERN_DEBUG "e820 update range: %016Lx - %016Lx ", + (unsigned long long) start, + (unsigned long long) end); + e820_print_type(old_type); + printk(KERN_CONT " ==> "); + e820_print_type(new_type); + printk(KERN_CONT "\n"); + + for (i = 0; i < e820x->nr_map; i++) { struct e820entry *ei = &e820x->map[i]; u64 final_start, final_end; + u64 ei_end; + if (ei->type != old_type) continue; - /* totally covered? */ - if (ei->addr >= start && - (ei->addr + ei->size) <= (start + size)) { + + ei_end = ei->addr + ei->size; + /* totally covered by new range? */ + if (ei->addr >= start && ei_end <= end) { ei->type = new_type; real_updated_size += ei->size; continue; } + + /* new range is totally covered? */ + if (ei->addr < start && ei_end > end) { + __e820_add_region(e820x, start, size, new_type); + __e820_add_region(e820x, end, ei_end - end, ei->type); + ei->size = start - ei->addr; + real_updated_size += size; + continue; + } + /* partially covered */ final_start = max(start, ei->addr); - final_end = min(start + size, ei->addr + ei->size); + final_end = min(end, ei_end); if (final_start >= final_end) continue; - e820_add_region(final_start, final_end - final_start, - new_type); + + __e820_add_region(e820x, final_start, final_end - final_start, + new_type); + real_updated_size += final_end - final_start; + /* + * left range could be head or tail, so need to update + * size at first. + */ ei->size -= final_end - final_start; if (ei->addr < final_start) continue; @@ -461,13 +502,13 @@ static u64 __init e820_update_range_map(struct e820map *e820x, u64 start, u64 __init e820_update_range(u64 start, u64 size, unsigned old_type, unsigned new_type) { - return e820_update_range_map(&e820, start, size, old_type, new_type); + return __e820_update_range(&e820, start, size, old_type, new_type); } static u64 __init e820_update_range_saved(u64 start, u64 size, unsigned old_type, unsigned new_type) { - return e820_update_range_map(&e820_saved, start, size, old_type, + return __e820_update_range(&e820_saved, start, size, old_type, new_type); } @@ -511,7 +552,7 @@ u64 __init e820_remove_range(u64 start, u64 size, unsigned old_type, void __init update_e820(void) { - int nr_map; + u32 nr_map; nr_map = e820.nr_map; if (sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &nr_map)) @@ -522,7 +563,7 @@ void __init update_e820(void) } static void __init update_e820_saved(void) { - int nr_map; + u32 nr_map; nr_map = e820_saved.nr_map; if (sanitize_e820_map(e820_saved.map, ARRAY_SIZE(e820_saved.map), &nr_map)) @@ -1020,8 +1061,8 @@ u64 __init find_e820_area_size(u64 start, u64 *sizep, u64 align) continue; return addr; } - return -1UL; + return -1ULL; } /* @@ -1034,13 +1075,22 @@ u64 __init early_reserve_e820(u64 startt, u64 sizet, u64 align) u64 start; start = startt; - while (size < sizet) + while (size < sizet && (start + 1)) start = find_e820_area_size(start, &size, align); if (size < sizet) return 0; +#ifdef CONFIG_X86_32 + if (start >= MAXMEM) + return 0; + if (start + size > MAXMEM) + size = MAXMEM - start; +#endif + addr = round_down(start + size - sizet, align); + if (addr < start) + return 0; e820_update_range(addr, sizet, E820_RAM, E820_RESERVED); e820_update_range_saved(addr, sizet, E820_RAM, E820_RESERVED); printk(KERN_INFO "update e820 for early_reserve_e820\n"); @@ -1253,7 +1303,7 @@ early_param("memmap", parse_memmap_opt); void __init finish_e820_parsing(void) { if (userdef) { - int nr = e820.nr_map; + u32 nr = e820.nr_map; if (sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &nr) < 0) early_panic("Invalid user supplied memory map"); @@ -1336,7 +1386,7 @@ void __init e820_reserve_resources_late(void) char *__init default_machine_specific_memory_setup(void) { char *who = "BIOS-e820"; - int new_nr; + u32 new_nr; /* * Try to copy the BIOS-supplied E820-map. * diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c index 639ad98238a2..335f049d110f 100644 --- a/arch/x86/kernel/early_printk.c +++ b/arch/x86/kernel/early_printk.c @@ -250,7 +250,7 @@ static int dbgp_wait_until_complete(void) return (ctrl & DBGP_ERROR) ? -DBGP_ERRCODE(ctrl) : DBGP_LEN(ctrl); } -static void dbgp_mdelay(int ms) +static void __init dbgp_mdelay(int ms) { int i; @@ -311,7 +311,7 @@ static void dbgp_set_data(const void *buf, int size) writel(hi, &ehci_debug->data47); } -static void dbgp_get_data(void *buf, int size) +static void __init dbgp_get_data(void *buf, int size) { unsigned char *bytes = buf; u32 lo, hi; @@ -355,7 +355,7 @@ static int dbgp_bulk_write(unsigned devnum, unsigned endpoint, return ret; } -static int dbgp_bulk_read(unsigned devnum, unsigned endpoint, void *data, +static int __init dbgp_bulk_read(unsigned devnum, unsigned endpoint, void *data, int size) { u32 pids, addr, ctrl; @@ -386,8 +386,8 @@ static int dbgp_bulk_read(unsigned devnum, unsigned endpoint, void *data, return ret; } -static int dbgp_control_msg(unsigned devnum, int requesttype, int request, - int value, int index, void *data, int size) +static int __init dbgp_control_msg(unsigned devnum, int requesttype, + int request, int value, int index, void *data, int size) { u32 pids, addr, ctrl; struct usb_ctrlrequest req; @@ -489,7 +489,7 @@ static u32 __init find_dbgp(int ehci_num, u32 *rbus, u32 *rslot, u32 *rfunc) return 0; } -static int ehci_reset_port(int port) +static int __init ehci_reset_port(int port) { u32 portsc; u32 delay_time, delay; @@ -532,7 +532,7 @@ static int ehci_reset_port(int port) return -EBUSY; } -static int ehci_wait_for_port(int port) +static int __init ehci_wait_for_port(int port) { u32 status; int ret, reps; @@ -557,13 +557,13 @@ static inline void dbgp_printk(const char *fmt, ...) { } typedef void (*set_debug_port_t)(int port); -static void default_set_debug_port(int port) +static void __init default_set_debug_port(int port) { } -static set_debug_port_t set_debug_port = default_set_debug_port; +static set_debug_port_t __initdata set_debug_port = default_set_debug_port; -static void nvidia_set_debug_port(int port) +static void __init nvidia_set_debug_port(int port) { u32 dword; dword = read_pci_config(ehci_dev.bus, ehci_dev.slot, ehci_dev.func, diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 899e8938e79f..c929add475c9 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -442,8 +442,7 @@ sysenter_past_esp: GET_THREAD_INFO(%ebp) - /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */ - testw $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp) + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp) jnz sysenter_audit sysenter_do_call: cmpl $(nr_syscalls), %eax @@ -454,7 +453,7 @@ sysenter_do_call: DISABLE_INTERRUPTS(CLBR_ANY) TRACE_IRQS_OFF movl TI_flags(%ebp), %ecx - testw $_TIF_ALLWORK_MASK, %cx + testl $_TIF_ALLWORK_MASK, %ecx jne sysexit_audit sysenter_exit: /* if something modifies registers it must also disable sysexit */ @@ -468,7 +467,7 @@ sysenter_exit: #ifdef CONFIG_AUDITSYSCALL sysenter_audit: - testw $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags(%ebp) + testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),TI_flags(%ebp) jnz syscall_trace_entry addl $4,%esp CFI_ADJUST_CFA_OFFSET -4 @@ -485,7 +484,7 @@ sysenter_audit: jmp sysenter_do_call sysexit_audit: - testw $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %cx + testl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %ecx jne syscall_exit_work TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_ANY) @@ -498,7 +497,7 @@ sysexit_audit: DISABLE_INTERRUPTS(CLBR_ANY) TRACE_IRQS_OFF movl TI_flags(%ebp), %ecx - testw $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %cx + testl $(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %ecx jne syscall_exit_work movl PT_EAX(%esp),%eax /* reload syscall return value */ jmp sysenter_exit @@ -523,8 +522,7 @@ ENTRY(system_call) SAVE_ALL GET_THREAD_INFO(%ebp) # system call tracing in operation / emulation - /* Note, _TIF_SECCOMP is bit number 8, and so it needs testw and not testb */ - testw $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp) + testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags(%ebp) jnz syscall_trace_entry cmpl $(nr_syscalls), %eax jae syscall_badsys @@ -538,7 +536,7 @@ syscall_exit: # between sampling and the iret TRACE_IRQS_OFF movl TI_flags(%ebp), %ecx - testw $_TIF_ALLWORK_MASK, %cx # current->work + testl $_TIF_ALLWORK_MASK, %ecx # current->work jne syscall_exit_work restore_all: @@ -673,7 +671,7 @@ END(syscall_trace_entry) # perform syscall exit tracing ALIGN syscall_exit_work: - testb $_TIF_WORK_SYSCALL_EXIT, %cl + testl $_TIF_WORK_SYSCALL_EXIT, %ecx jz work_pending TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_ANY) # could let syscall_trace_leave() call diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 83d1836b9467..a331ec38af9e 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -368,6 +368,7 @@ ENTRY(save_rest) END(save_rest) /* save complete stack frame */ + .pushsection .kprobes.text, "ax" ENTRY(save_paranoid) XCPT_FRAME 1 RDI+8 cld @@ -396,6 +397,7 @@ ENTRY(save_paranoid) 1: ret CFI_ENDPROC END(save_paranoid) + .popsection /* * A newly forked process directly context switches into this address. @@ -416,7 +418,6 @@ ENTRY(ret_from_fork) GET_THREAD_INFO(%rcx) - CFI_REMEMBER_STATE RESTORE_REST testl $3, CS-ARGOFFSET(%rsp) # from kernel_thread? @@ -428,7 +429,6 @@ ENTRY(ret_from_fork) RESTORE_TOP_OF_STACK %rdi, -ARGOFFSET jmp ret_from_sys_call # go to the SYSRET fastpath - CFI_RESTORE_STATE CFI_ENDPROC END(ret_from_fork) @@ -984,6 +984,8 @@ apicinterrupt UV_BAU_MESSAGE \ #endif apicinterrupt LOCAL_TIMER_VECTOR \ apic_timer_interrupt smp_apic_timer_interrupt +apicinterrupt GENERIC_INTERRUPT_VECTOR \ + generic_interrupt smp_generic_interrupt #ifdef CONFIG_SMP apicinterrupt INVALIDATE_TLB_VECTOR_START+0 \ diff --git a/arch/x86/kernel/head32.c b/arch/x86/kernel/head32.c index ac108d1fe182..3f8579f8d42c 100644 --- a/arch/x86/kernel/head32.c +++ b/arch/x86/kernel/head32.c @@ -18,7 +18,7 @@ void __init i386_start_kernel(void) { reserve_trampoline_memory(); - reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS"); + reserve_early(__pa_symbol(&_text), __pa_symbol(&__bss_stop), "TEXT DATA BSS"); #ifdef CONFIG_BLK_DEV_INITRD /* Reserve INITRD */ @@ -29,9 +29,6 @@ void __init i386_start_kernel(void) reserve_early(ramdisk_image, ramdisk_end, "RAMDISK"); } #endif - reserve_early(init_pg_tables_start, init_pg_tables_end, - "INIT_PG_TABLE"); - reserve_ebda_region(); /* diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index f5b272247690..70eaa852c732 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -100,7 +100,7 @@ void __init x86_64_start_reservations(char *real_mode_data) reserve_trampoline_memory(); - reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS"); + reserve_early(__pa_symbol(&_text), __pa_symbol(&__bss_stop), "TEXT DATA BSS"); #ifdef CONFIG_BLK_DEV_INITRD /* Reserve INITRD */ diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index c32ca19d591a..30683883e0cd 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -38,42 +38,40 @@ #define X86_VENDOR_ID new_cpu_data+CPUINFO_x86_vendor_id /* - * This is how much memory *in addition to the memory covered up to - * and including _end* we need mapped initially. + * This is how much memory in addition to the memory covered up to + * and including _end we need mapped initially. * We need: - * - one bit for each possible page, but only in low memory, which means - * 2^32/4096/8 = 128K worst case (4G/4G split.) - * - enough space to map all low memory, which means - * (2^32/4096) / 1024 pages (worst case, non PAE) - * (2^32/4096) / 512 + 4 pages (worst case for PAE) - * - a few pages for allocator use before the kernel pagetable has - * been set up + * (KERNEL_IMAGE_SIZE/4096) / 1024 pages (worst case, non PAE) + * (KERNEL_IMAGE_SIZE/4096) / 512 + 4 pages (worst case for PAE) * * Modulo rounding, each megabyte assigned here requires a kilobyte of * memory, which is currently unreclaimed. * * This should be a multiple of a page. + * + * KERNEL_IMAGE_SIZE should be greater than pa(_end) + * and small than max_low_pfn, otherwise will waste some page table entries */ -LOW_PAGES = 1<<(32-PAGE_SHIFT_asm) - -/* - * To preserve the DMA pool in PAGEALLOC kernels, we'll allocate - * pagetables from above the 16MB DMA limit, so we'll have to set - * up pagetables 16MB more (worst-case): - */ -#ifdef CONFIG_DEBUG_PAGEALLOC -LOW_PAGES = LOW_PAGES + 0x1000000 -#endif #if PTRS_PER_PMD > 1 -PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PMD) + PTRS_PER_PGD +#define PAGE_TABLE_SIZE(pages) (((pages) / PTRS_PER_PMD) + PTRS_PER_PGD) #else -PAGE_TABLE_SIZE = (LOW_PAGES / PTRS_PER_PGD) +#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PGD) #endif -BOOTBITMAP_SIZE = LOW_PAGES / 8 -ALLOCATOR_SLOP = 4 -INIT_MAP_BEYOND_END = BOOTBITMAP_SIZE + (PAGE_TABLE_SIZE + ALLOCATOR_SLOP)*PAGE_SIZE_asm +/* Enough space to fit pagetables for the low memory linear map */ +MAPPING_BEYOND_END = \ + PAGE_TABLE_SIZE(((1<<32) - __PAGE_OFFSET) >> PAGE_SHIFT) << PAGE_SHIFT + +/* + * Worst-case size of the kernel mapping we need to make: + * the worst-case size of the kernel itself, plus the extra we need + * to map for the linear map. + */ +KERNEL_PAGES = (KERNEL_IMAGE_SIZE + MAPPING_BEYOND_END)>>PAGE_SHIFT + +INIT_MAP_SIZE = PAGE_TABLE_SIZE(KERNEL_PAGES) * PAGE_SIZE_asm +RESERVE_BRK(pagetables, INIT_MAP_SIZE) /* * 32-bit kernel entrypoint; only used by the boot CPU. On entry, @@ -166,10 +164,10 @@ num_subarch_entries = (. - subarch_entries) / 4 /* * Initialize page tables. This creates a PDE and a set of page - * tables, which are located immediately beyond _end. The variable - * init_pg_tables_end is set up to point to the first "safe" location. + * tables, which are located immediately beyond __brk_base. The variable + * _brk_end is set up to point to the first "safe" location. * Mappings are created both at virtual address 0 (identity mapping) - * and PAGE_OFFSET for up to _end+sizeof(page tables)+INIT_MAP_BEYOND_END. + * and PAGE_OFFSET for up to _end. * * Note that the stack is not yet set up! */ @@ -190,8 +188,7 @@ default_entry: xorl %ebx,%ebx /* %ebx is kept at zero */ - movl $pa(pg0), %edi - movl %edi, pa(init_pg_tables_start) + movl $pa(__brk_base), %edi movl $pa(swapper_pg_pmd), %edx movl $PTE_IDENT_ATTR, %eax 10: @@ -209,14 +206,14 @@ default_entry: loop 11b /* - * End condition: we must map up to and including INIT_MAP_BEYOND_END - * bytes beyond the end of our own page tables. + * End condition: we must map up to the end + MAPPING_BEYOND_END. */ - leal (INIT_MAP_BEYOND_END+PTE_IDENT_ATTR)(%edi),%ebp + movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp cmpl %ebp,%eax jb 10b 1: - movl %edi,pa(init_pg_tables_end) + addl $__PAGE_OFFSET, %edi + movl %edi, pa(_brk_end) shrl $12, %eax movl %eax, pa(max_pfn_mapped) @@ -227,8 +224,7 @@ default_entry: page_pde_offset = (__PAGE_OFFSET >> 20); - movl $pa(pg0), %edi - movl %edi, pa(init_pg_tables_start) + movl $pa(__brk_base), %edi movl $pa(swapper_pg_dir), %edx movl $PTE_IDENT_ATTR, %eax 10: @@ -242,14 +238,13 @@ page_pde_offset = (__PAGE_OFFSET >> 20); addl $0x1000,%eax loop 11b /* - * End condition: we must map up to and including INIT_MAP_BEYOND_END - * bytes beyond the end of our own page tables; the +0x007 is - * the attribute bits + * End condition: we must map up to the end + MAPPING_BEYOND_END. */ - leal (INIT_MAP_BEYOND_END+PTE_IDENT_ATTR)(%edi),%ebp + movl $pa(_end) + MAPPING_BEYOND_END + PTE_IDENT_ATTR, %ebp cmpl %ebp,%eax jb 10b - movl %edi,pa(init_pg_tables_end) + addl $__PAGE_OFFSET, %edi + movl %edi, pa(_brk_end) shrl $12, %eax movl %eax, pa(max_pfn_mapped) @@ -636,6 +631,7 @@ swapper_pg_fixmap: .fill 1024,4,0 ENTRY(empty_zero_page) .fill 4096,1,0 + /* * This starts the data section. */ diff --git a/arch/x86/kernel/i8253.c b/arch/x86/kernel/i8253.c index 10f92fb532f3..3475440baa54 100644 --- a/arch/x86/kernel/i8253.c +++ b/arch/x86/kernel/i8253.c @@ -3,17 +3,17 @@ * */ #include <linux/clockchips.h> -#include <linux/init.h> #include <linux/interrupt.h> +#include <linux/spinlock.h> #include <linux/jiffies.h> #include <linux/module.h> -#include <linux/spinlock.h> +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/io.h> -#include <asm/smp.h> -#include <asm/delay.h> #include <asm/i8253.h> -#include <asm/io.h> #include <asm/hpet.h> +#include <asm/smp.h> DEFINE_SPINLOCK(i8253_lock); EXPORT_SYMBOL(i8253_lock); @@ -40,7 +40,7 @@ static void init_pit_timer(enum clock_event_mode mode, { spin_lock(&i8253_lock); - switch(mode) { + switch (mode) { case CLOCK_EVT_MODE_PERIODIC: /* binary, mode 2, LSB/MSB, ch 0 */ outb_pit(0x34, PIT_MODE); @@ -95,7 +95,7 @@ static int pit_next_event(unsigned long delta, struct clock_event_device *evt) * registered. This mechanism replaces the previous #ifdef LOCAL_APIC - * !using_apic_timer decisions in do_timer_interrupt_hook() */ -static struct clock_event_device pit_clockevent = { +static struct clock_event_device pit_ce = { .name = "pit", .features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT, .set_mode = init_pit_timer, @@ -114,15 +114,13 @@ void __init setup_pit_timer(void) * Start pit with the boot cpu mask and make it global after the * IO_APIC has been initialized. */ - pit_clockevent.cpumask = cpumask_of(smp_processor_id()); - pit_clockevent.mult = div_sc(CLOCK_TICK_RATE, NSEC_PER_SEC, - pit_clockevent.shift); - pit_clockevent.max_delta_ns = - clockevent_delta2ns(0x7FFF, &pit_clockevent); - pit_clockevent.min_delta_ns = - clockevent_delta2ns(0xF, &pit_clockevent); - clockevents_register_device(&pit_clockevent); - global_clock_event = &pit_clockevent; + pit_ce.cpumask = cpumask_of(smp_processor_id()); + pit_ce.mult = div_sc(CLOCK_TICK_RATE, NSEC_PER_SEC, pit_ce.shift); + pit_ce.max_delta_ns = clockevent_delta2ns(0x7FFF, &pit_ce); + pit_ce.min_delta_ns = clockevent_delta2ns(0xF, &pit_ce); + + clockevents_register_device(&pit_ce); + global_clock_event = &pit_ce; } #ifndef CONFIG_X86_64 @@ -133,11 +131,11 @@ void __init setup_pit_timer(void) */ static cycle_t pit_read(void) { + static int old_count; + static u32 old_jifs; unsigned long flags; int count; u32 jifs; - static int old_count; - static u32 old_jifs; spin_lock_irqsave(&i8253_lock, flags); /* @@ -179,9 +177,9 @@ static cycle_t pit_read(void) * Previous attempts to handle these cases intelligently were * buggy, so we just do the simple thing now. */ - if (count > old_count && jifs == old_jifs) { + if (count > old_count && jifs == old_jifs) count = old_count; - } + old_count = count; old_jifs = jifs; @@ -192,13 +190,13 @@ static cycle_t pit_read(void) return (cycle_t)(jifs * LATCH) + count; } -static struct clocksource clocksource_pit = { - .name = "pit", - .rating = 110, - .read = pit_read, - .mask = CLOCKSOURCE_MASK(32), - .mult = 0, - .shift = 20, +static struct clocksource pit_cs = { + .name = "pit", + .rating = 110, + .read = pit_read, + .mask = CLOCKSOURCE_MASK(32), + .mult = 0, + .shift = 20, }; static void pit_disable_clocksource(void) @@ -206,9 +204,9 @@ static void pit_disable_clocksource(void) /* * Use mult to check whether it is registered or not */ - if (clocksource_pit.mult) { - clocksource_unregister(&clocksource_pit); - clocksource_pit.mult = 0; + if (pit_cs.mult) { + clocksource_unregister(&pit_cs); + pit_cs.mult = 0; } } @@ -222,13 +220,13 @@ static int __init init_pit_clocksource(void) * - when local APIC timer is active (PIT is switched off) */ if (num_possible_cpus() > 1 || is_hpet_enabled() || - pit_clockevent.mode != CLOCK_EVT_MODE_PERIODIC) + pit_ce.mode != CLOCK_EVT_MODE_PERIODIC) return 0; - clocksource_pit.mult = clocksource_hz2mult(CLOCK_TICK_RATE, - clocksource_pit.shift); - return clocksource_register(&clocksource_pit); + pit_cs.mult = clocksource_hz2mult(CLOCK_TICK_RATE, pit_cs.shift); + + return clocksource_register(&pit_cs); } arch_initcall(init_pit_clocksource); -#endif +#endif /* !CONFIG_X86_64 */ diff --git a/arch/x86/kernel/io_delay.c b/arch/x86/kernel/io_delay.c index 720d2607aacb..a979b5bd2fc0 100644 --- a/arch/x86/kernel/io_delay.c +++ b/arch/x86/kernel/io_delay.c @@ -7,10 +7,10 @@ */ #include <linux/kernel.h> #include <linux/module.h> -#include <linux/init.h> #include <linux/delay.h> +#include <linux/init.h> #include <linux/dmi.h> -#include <asm/io.h> +#include <linux/io.h> int io_delay_type __read_mostly = CONFIG_DEFAULT_IO_DELAY_TYPE; @@ -47,8 +47,7 @@ EXPORT_SYMBOL(native_io_delay); static int __init dmi_io_delay_0xed_port(const struct dmi_system_id *id) { if (io_delay_type == CONFIG_IO_DELAY_TYPE_0X80) { - printk(KERN_NOTICE "%s: using 0xed I/O delay port\n", - id->ident); + pr_notice("%s: using 0xed I/O delay port\n", id->ident); io_delay_type = CONFIG_IO_DELAY_TYPE_0XED; } @@ -64,40 +63,40 @@ static struct dmi_system_id __initdata io_delay_0xed_port_dmi_table[] = { .callback = dmi_io_delay_0xed_port, .ident = "Compaq Presario V6000", .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), - DMI_MATCH(DMI_BOARD_NAME, "30B7") + DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), + DMI_MATCH(DMI_BOARD_NAME, "30B7") } }, { .callback = dmi_io_delay_0xed_port, .ident = "HP Pavilion dv9000z", .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), - DMI_MATCH(DMI_BOARD_NAME, "30B9") + DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), + DMI_MATCH(DMI_BOARD_NAME, "30B9") } }, { .callback = dmi_io_delay_0xed_port, .ident = "HP Pavilion dv6000", .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), - DMI_MATCH(DMI_BOARD_NAME, "30B8") + DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), + DMI_MATCH(DMI_BOARD_NAME, "30B8") } }, { .callback = dmi_io_delay_0xed_port, .ident = "HP Pavilion tx1000", .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), - DMI_MATCH(DMI_BOARD_NAME, "30BF") + DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), + DMI_MATCH(DMI_BOARD_NAME, "30BF") } }, { .callback = dmi_io_delay_0xed_port, .ident = "Presario F700", .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), - DMI_MATCH(DMI_BOARD_NAME, "30D3") + DMI_MATCH(DMI_BOARD_VENDOR, "Quanta"), + DMI_MATCH(DMI_BOARD_NAME, "30D3") } }, { } diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index f13ca1650aaf..3aaf7b9e3a8b 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -15,6 +15,9 @@ atomic_t irq_err_count; +/* Function pointer for generic interrupt vector handling */ +void (*generic_interrupt_extension)(void) = NULL; + /* * 'what should we do if we get a hw irq event on an illegal vector'. * each architecture has to answer this themselves. @@ -42,55 +45,60 @@ void ack_bad_irq(unsigned int irq) /* * /proc/interrupts printing: */ -static int show_other_interrupts(struct seq_file *p) +static int show_other_interrupts(struct seq_file *p, int prec) { int j; - seq_printf(p, "NMI: "); + seq_printf(p, "%*s: ", prec, "NMI"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->__nmi_count); seq_printf(p, " Non-maskable interrupts\n"); #ifdef CONFIG_X86_LOCAL_APIC - seq_printf(p, "LOC: "); + seq_printf(p, "%*s: ", prec, "LOC"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->apic_timer_irqs); seq_printf(p, " Local timer interrupts\n"); + + seq_printf(p, "%*s: ", prec, "SPU"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->irq_spurious_count); + seq_printf(p, " Spurious interrupts\n"); #endif + if (generic_interrupt_extension) { + seq_printf(p, "PLT: "); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->generic_irqs); + seq_printf(p, " Platform interrupts\n"); + } #ifdef CONFIG_SMP - seq_printf(p, "RES: "); + seq_printf(p, "%*s: ", prec, "RES"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_resched_count); seq_printf(p, " Rescheduling interrupts\n"); - seq_printf(p, "CAL: "); + seq_printf(p, "%*s: ", prec, "CAL"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_call_count); seq_printf(p, " Function call interrupts\n"); - seq_printf(p, "TLB: "); + seq_printf(p, "%*s: ", prec, "TLB"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_tlb_count); seq_printf(p, " TLB shootdowns\n"); #endif #ifdef CONFIG_X86_MCE - seq_printf(p, "TRM: "); + seq_printf(p, "%*s: ", prec, "TRM"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_thermal_count); seq_printf(p, " Thermal event interrupts\n"); # ifdef CONFIG_X86_64 - seq_printf(p, "THR: "); + seq_printf(p, "%*s: ", prec, "THR"); for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_threshold_count); seq_printf(p, " Threshold APIC interrupts\n"); # endif #endif -#ifdef CONFIG_X86_LOCAL_APIC - seq_printf(p, "SPU: "); - for_each_online_cpu(j) - seq_printf(p, "%10u ", irq_stats(j)->irq_spurious_count); - seq_printf(p, " Spurious interrupts\n"); -#endif - seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count)); + seq_printf(p, "%*s: %10u\n", prec, "ERR", atomic_read(&irq_err_count)); #if defined(CONFIG_X86_IO_APIC) - seq_printf(p, "MIS: %10u\n", atomic_read(&irq_mis_count)); + seq_printf(p, "%*s: %10u\n", prec, "MIS", atomic_read(&irq_mis_count)); #endif return 0; } @@ -98,19 +106,22 @@ static int show_other_interrupts(struct seq_file *p) int show_interrupts(struct seq_file *p, void *v) { unsigned long flags, any_count = 0; - int i = *(loff_t *) v, j; + int i = *(loff_t *) v, j, prec; struct irqaction *action; struct irq_desc *desc; if (i > nr_irqs) return 0; + for (prec = 3, j = 1000; prec < 10 && j <= nr_irqs; ++prec) + j *= 10; + if (i == nr_irqs) - return show_other_interrupts(p); + return show_other_interrupts(p, prec); /* print header */ if (i == 0) { - seq_printf(p, " "); + seq_printf(p, "%*s", prec + 8, ""); for_each_online_cpu(j) seq_printf(p, "CPU%-8d", j); seq_putc(p, '\n'); @@ -121,23 +132,15 @@ int show_interrupts(struct seq_file *p, void *v) return 0; spin_lock_irqsave(&desc->lock, flags); -#ifndef CONFIG_SMP - any_count = kstat_irqs(i); -#else for_each_online_cpu(j) any_count |= kstat_irqs_cpu(i, j); -#endif action = desc->action; if (!action && !any_count) goto out; - seq_printf(p, "%3d: ", i); -#ifndef CONFIG_SMP - seq_printf(p, "%10u ", kstat_irqs(i)); -#else + seq_printf(p, "%*d: ", prec, i); for_each_online_cpu(j) seq_printf(p, "%10u ", kstat_irqs_cpu(i, j)); -#endif seq_printf(p, " %8s", desc->chip->name); seq_printf(p, "-%-8s", desc->name); @@ -162,7 +165,10 @@ u64 arch_irq_stat_cpu(unsigned int cpu) #ifdef CONFIG_X86_LOCAL_APIC sum += irq_stats(cpu)->apic_timer_irqs; + sum += irq_stats(cpu)->irq_spurious_count; #endif + if (generic_interrupt_extension) + sum += irq_stats(cpu)->generic_irqs; #ifdef CONFIG_SMP sum += irq_stats(cpu)->irq_resched_count; sum += irq_stats(cpu)->irq_call_count; @@ -174,9 +180,6 @@ u64 arch_irq_stat_cpu(unsigned int cpu) sum += irq_stats(cpu)->irq_threshold_count; #endif #endif -#ifdef CONFIG_X86_LOCAL_APIC - sum += irq_stats(cpu)->irq_spurious_count; -#endif return sum; } @@ -226,4 +229,27 @@ unsigned int __irq_entry do_IRQ(struct pt_regs *regs) return 1; } +/* + * Handler for GENERIC_INTERRUPT_VECTOR. + */ +void smp_generic_interrupt(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + ack_APIC_irq(); + + exit_idle(); + + irq_enter(); + + inc_irq_stat(generic_irqs); + + if (generic_interrupt_extension) + generic_interrupt_extension(); + + irq_exit(); + + set_irq_regs(old_regs); +} + EXPORT_SYMBOL_GPL(vector_used_by_percpu_irq); diff --git a/arch/x86/kernel/irqinit_32.c b/arch/x86/kernel/irqinit_32.c index 50b8c3a3006c..bc1326105448 100644 --- a/arch/x86/kernel/irqinit_32.c +++ b/arch/x86/kernel/irqinit_32.c @@ -175,6 +175,9 @@ void __init native_init_IRQ(void) /* self generated IPI for local APIC timer */ alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); + /* generic IPI for platform specific use */ + alloc_intr_gate(GENERIC_INTERRUPT_VECTOR, generic_interrupt); + /* IPI vectors for APIC spurious and error interrupts */ alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt); diff --git a/arch/x86/kernel/irqinit_64.c b/arch/x86/kernel/irqinit_64.c index da481a1e3f30..c7a49e0ffbfb 100644 --- a/arch/x86/kernel/irqinit_64.c +++ b/arch/x86/kernel/irqinit_64.c @@ -147,6 +147,9 @@ static void __init apic_intr_init(void) /* self generated IPI for local APIC timer */ alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); + /* generic IPI for platform specific use */ + alloc_intr_gate(GENERIC_INTERRUPT_VECTOR, generic_interrupt); + /* IPI vectors for APIC spurious and error interrupts */ alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt); diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c index ff7d3b0124f1..e444357375ce 100644 --- a/arch/x86/kernel/kdebugfs.c +++ b/arch/x86/kernel/kdebugfs.c @@ -8,11 +8,11 @@ */ #include <linux/debugfs.h> #include <linux/uaccess.h> -#include <linux/stat.h> +#include <linux/module.h> #include <linux/init.h> +#include <linux/stat.h> #include <linux/io.h> #include <linux/mm.h> -#include <linux/module.h> #include <asm/setup.h> @@ -26,9 +26,8 @@ struct setup_data_node { u32 len; }; -static ssize_t -setup_data_read(struct file *file, char __user *user_buf, size_t count, - loff_t *ppos) +static ssize_t setup_data_read(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) { struct setup_data_node *node = file->private_data; unsigned long remain; @@ -39,20 +38,21 @@ setup_data_read(struct file *file, char __user *user_buf, size_t count, if (pos < 0) return -EINVAL; + if (pos >= node->len) return 0; if (count > node->len - pos) count = node->len - pos; + pa = node->paddr + sizeof(struct setup_data) + pos; pg = pfn_to_page((pa + count - 1) >> PAGE_SHIFT); if (PageHighMem(pg)) { p = ioremap_cache(pa, count); if (!p) return -ENXIO; - } else { + } else p = __va(pa); - } remain = copy_to_user(user_buf, p, count); @@ -70,12 +70,13 @@ setup_data_read(struct file *file, char __user *user_buf, size_t count, static int setup_data_open(struct inode *inode, struct file *file) { file->private_data = inode->i_private; + return 0; } static const struct file_operations fops_setup_data = { - .read = setup_data_read, - .open = setup_data_open, + .read = setup_data_read, + .open = setup_data_open, }; static int __init @@ -84,57 +85,50 @@ create_setup_data_node(struct dentry *parent, int no, { struct dentry *d, *type, *data; char buf[16]; - int error; sprintf(buf, "%d", no); d = debugfs_create_dir(buf, parent); - if (!d) { - error = -ENOMEM; - goto err_return; - } + if (!d) + return -ENOMEM; + type = debugfs_create_x32("type", S_IRUGO, d, &node->type); - if (!type) { - error = -ENOMEM; + if (!type) goto err_dir; - } + data = debugfs_create_file("data", S_IRUGO, d, node, &fops_setup_data); - if (!data) { - error = -ENOMEM; + if (!data) goto err_type; - } + return 0; err_type: debugfs_remove(type); err_dir: debugfs_remove(d); -err_return: - return error; + return -ENOMEM; } static int __init create_setup_data_nodes(struct dentry *parent) { struct setup_data_node *node; struct setup_data *data; - int error, no = 0; + int error = -ENOMEM; struct dentry *d; struct page *pg; u64 pa_data; + int no = 0; d = debugfs_create_dir("setup_data", parent); - if (!d) { - error = -ENOMEM; - goto err_return; - } + if (!d) + return -ENOMEM; pa_data = boot_params.hdr.setup_data; while (pa_data) { node = kmalloc(sizeof(*node), GFP_KERNEL); - if (!node) { - error = -ENOMEM; + if (!node) goto err_dir; - } + pg = pfn_to_page((pa_data+sizeof(*data)-1) >> PAGE_SHIFT); if (PageHighMem(pg)) { data = ioremap_cache(pa_data, sizeof(*data)); @@ -143,9 +137,8 @@ static int __init create_setup_data_nodes(struct dentry *parent) error = -ENXIO; goto err_dir; } - } else { + } else data = __va(pa_data); - } node->paddr = pa_data; node->type = data->type; @@ -159,11 +152,11 @@ static int __init create_setup_data_nodes(struct dentry *parent) goto err_dir; no++; } + return 0; err_dir: debugfs_remove(d); -err_return: return error; } @@ -175,28 +168,26 @@ static struct debugfs_blob_wrapper boot_params_blob = { static int __init boot_params_kdebugfs_init(void) { struct dentry *dbp, *version, *data; - int error; + int error = -ENOMEM; dbp = debugfs_create_dir("boot_params", NULL); - if (!dbp) { - error = -ENOMEM; - goto err_return; - } + if (!dbp) + return -ENOMEM; + version = debugfs_create_x16("version", S_IRUGO, dbp, &boot_params.hdr.version); - if (!version) { - error = -ENOMEM; + if (!version) goto err_dir; - } + data = debugfs_create_blob("data", S_IRUGO, dbp, &boot_params_blob); - if (!data) { - error = -ENOMEM; + if (!data) goto err_version; - } + error = create_setup_data_nodes(dbp); if (error) goto err_data; + return 0; err_data: @@ -205,10 +196,9 @@ err_version: debugfs_remove(version); err_dir: debugfs_remove(dbp); -err_return: return error; } -#endif +#endif /* CONFIG_DEBUG_BOOT_PARAMS */ static int __init arch_kdebugfs_init(void) { diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index 4558dd3918cf..55b94614e348 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -193,7 +193,7 @@ static int __kprobes can_boost(kprobe_opcode_t *opcodes) kprobe_opcode_t opcode; kprobe_opcode_t *orig_opcodes = opcodes; - if (search_exception_tables(opcodes)) + if (search_exception_tables((unsigned long)opcodes)) return 0; /* Page fault may occur on this address. */ retry: diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index 478bca986eca..33019ddb56b4 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -138,12 +138,6 @@ static void kvm_set_pte_atomic(pte_t *ptep, pte_t pte) kvm_mmu_write(ptep, pte_val(pte)); } -static void kvm_set_pte_present(struct mm_struct *mm, unsigned long addr, - pte_t *ptep, pte_t pte) -{ - kvm_mmu_write(ptep, pte_val(pte)); -} - static void kvm_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) { @@ -220,7 +214,6 @@ static void paravirt_ops_setup(void) #if PAGETABLE_LEVELS >= 3 #ifdef CONFIG_X86_PAE pv_mmu_ops.set_pte_atomic = kvm_set_pte_atomic; - pv_mmu_ops.set_pte_present = kvm_set_pte_present; pv_mmu_ops.pte_clear = kvm_pte_clear; pv_mmu_ops.pmd_clear = kvm_pmd_clear; #endif diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index f5fc8c781a62..e7368c1da01d 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -14,12 +14,12 @@ #include <linux/ftrace.h> #include <linux/suspend.h> #include <linux/gfp.h> +#include <linux/io.h> #include <asm/pgtable.h> #include <asm/pgalloc.h> #include <asm/tlbflush.h> #include <asm/mmu_context.h> -#include <asm/io.h> #include <asm/apic.h> #include <asm/cpufeature.h> #include <asm/desc.h> @@ -63,7 +63,7 @@ static void load_segments(void) "\tmovl %%eax,%%fs\n" "\tmovl %%eax,%%gs\n" "\tmovl %%eax,%%ss\n" - ::: "eax", "memory"); + : : : "eax", "memory"); #undef STR #undef __STR } @@ -205,7 +205,8 @@ void machine_kexec(struct kimage *image) if (image->preserve_context) { #ifdef CONFIG_X86_IO_APIC - /* We need to put APICs in legacy mode so that we can + /* + * We need to put APICs in legacy mode so that we can * get timer interrupts in second kernel. kexec/kdump * paths already have calls to disable_IO_APIC() in * one form or other. kexec jump path also need @@ -227,7 +228,8 @@ void machine_kexec(struct kimage *image) page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page) << PAGE_SHIFT); - /* The segment registers are funny things, they have both a + /* + * The segment registers are funny things, they have both a * visible and an invisible part. Whenever the visible part is * set to a specific selector, the invisible part is loaded * with from a table in memory. At no other time is the @@ -237,11 +239,12 @@ void machine_kexec(struct kimage *image) * segments, before I zap the gdt with an invalid value. */ load_segments(); - /* The gdt & idt are now invalid. + /* + * The gdt & idt are now invalid. * If you want to load them you must set up your own idt & gdt. */ - set_gdt(phys_to_virt(0),0); - set_idt(phys_to_virt(0),0); + set_gdt(phys_to_virt(0), 0); + set_idt(phys_to_virt(0), 0); /* now call it */ image->start = relocate_kernel_ptr((unsigned long)image->head, diff --git a/arch/x86/kernel/machine_kexec_64.c b/arch/x86/kernel/machine_kexec_64.c index 6993d51b7fd8..89cea4d44679 100644 --- a/arch/x86/kernel/machine_kexec_64.c +++ b/arch/x86/kernel/machine_kexec_64.c @@ -12,11 +12,47 @@ #include <linux/reboot.h> #include <linux/numa.h> #include <linux/ftrace.h> +#include <linux/io.h> +#include <linux/suspend.h> #include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/mmu_context.h> -#include <asm/io.h> + +static int init_one_level2_page(struct kimage *image, pgd_t *pgd, + unsigned long addr) +{ + pud_t *pud; + pmd_t *pmd; + struct page *page; + int result = -ENOMEM; + + addr &= PMD_MASK; + pgd += pgd_index(addr); + if (!pgd_present(*pgd)) { + page = kimage_alloc_control_pages(image, 0); + if (!page) + goto out; + pud = (pud_t *)page_address(page); + memset(pud, 0, PAGE_SIZE); + set_pgd(pgd, __pgd(__pa(pud) | _KERNPG_TABLE)); + } + pud = pud_offset(pgd, addr); + if (!pud_present(*pud)) { + page = kimage_alloc_control_pages(image, 0); + if (!page) + goto out; + pmd = (pmd_t *)page_address(page); + memset(pmd, 0, PAGE_SIZE); + set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE)); + } + pmd = pmd_offset(pud, addr); + if (!pmd_present(*pmd)) + set_pmd(pmd, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC)); + result = 0; +out: + return result; +} static void init_level2_page(pmd_t *level2p, unsigned long addr) { @@ -83,9 +119,8 @@ static int init_level4_page(struct kimage *image, pgd_t *level4p, } level3p = (pud_t *)page_address(page); result = init_level3_page(image, level3p, addr, last_addr); - if (result) { + if (result) goto out; - } set_pgd(level4p++, __pgd(__pa(level3p) | _KERNPG_TABLE)); addr += PGDIR_SIZE; } @@ -156,6 +191,13 @@ static int init_pgtable(struct kimage *image, unsigned long start_pgtable) result = init_level4_page(image, level4p, 0, max_pfn << PAGE_SHIFT); if (result) return result; + /* + * image->start may be outside 0 ~ max_pfn, for example when + * jump back to original kernel from kexeced kernel + */ + result = init_one_level2_page(image, level4p, image->start); + if (result) + return result; return init_transition_pgtable(image, level4p); } @@ -229,20 +271,45 @@ void machine_kexec(struct kimage *image) { unsigned long page_list[PAGES_NR]; void *control_page; + int save_ftrace_enabled; - tracer_disable(); +#ifdef CONFIG_KEXEC_JUMP + if (kexec_image->preserve_context) + save_processor_state(); +#endif + + save_ftrace_enabled = __ftrace_enabled_save(); /* Interrupts aren't acceptable while we reboot */ local_irq_disable(); + if (image->preserve_context) { +#ifdef CONFIG_X86_IO_APIC + /* + * We need to put APICs in legacy mode so that we can + * get timer interrupts in second kernel. kexec/kdump + * paths already have calls to disable_IO_APIC() in + * one form or other. kexec jump path also need + * one. + */ + disable_IO_APIC(); +#endif + } + control_page = page_address(image->control_code_page) + PAGE_SIZE; - memcpy(control_page, relocate_kernel, PAGE_SIZE); + memcpy(control_page, relocate_kernel, KEXEC_CONTROL_CODE_MAX_SIZE); page_list[PA_CONTROL_PAGE] = virt_to_phys(control_page); + page_list[VA_CONTROL_PAGE] = (unsigned long)control_page; page_list[PA_TABLE_PAGE] = (unsigned long)__pa(page_address(image->control_code_page)); - /* The segment registers are funny things, they have both a + if (image->type == KEXEC_TYPE_DEFAULT) + page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page) + << PAGE_SHIFT); + + /* + * The segment registers are funny things, they have both a * visible and an invisible part. Whenever the visible part is * set to a specific selector, the invisible part is loaded * with from a table in memory. At no other time is the @@ -252,15 +319,25 @@ void machine_kexec(struct kimage *image) * segments, before I zap the gdt with an invalid value. */ load_segments(); - /* The gdt & idt are now invalid. + /* + * The gdt & idt are now invalid. * If you want to load them you must set up your own idt & gdt. */ - set_gdt(phys_to_virt(0),0); - set_idt(phys_to_virt(0),0); + set_gdt(phys_to_virt(0), 0); + set_idt(phys_to_virt(0), 0); /* now call it */ - relocate_kernel((unsigned long)image->head, (unsigned long)page_list, - image->start); + image->start = relocate_kernel((unsigned long)image->head, + (unsigned long)page_list, + image->start, + image->preserve_context); + +#ifdef CONFIG_KEXEC_JUMP + if (kexec_image->preserve_context) + restore_processor_state(); +#endif + + __ftrace_enabled_restore(save_ftrace_enabled); } void arch_crash_save_vmcoreinfo(void) diff --git a/arch/x86/kernel/mmconf-fam10h_64.c b/arch/x86/kernel/mmconf-fam10h_64.c index 666e43df51f9..712d15fdc416 100644 --- a/arch/x86/kernel/mmconf-fam10h_64.c +++ b/arch/x86/kernel/mmconf-fam10h_64.c @@ -226,7 +226,7 @@ static int __devinit set_check_enable_amd_mmconf(const struct dmi_system_id *d) return 0; } -static struct dmi_system_id __devinitdata mmconf_dmi_table[] = { +static const struct dmi_system_id __cpuinitconst mmconf_dmi_table[] = { { .callback = set_check_enable_amd_mmconf, .ident = "Sun Microsystems Machine", diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index 37cb1bda1baf..dce99dca6cf8 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -109,9 +109,6 @@ static void __init MP_bus_info(struct mpc_bus *m) } else printk(KERN_WARNING "Unknown bustype %s - ignoring\n", str); } -#endif - -#ifdef CONFIG_X86_IO_APIC static int bad_ioapic(unsigned long address) { @@ -224,8 +221,12 @@ static void __init MP_intsrc_info(struct mpc_intsrc *m) if (++mp_irq_entries == MAX_IRQ_SOURCES) panic("Max # of irq sources exceeded!!\n"); } +#else /* CONFIG_X86_IO_APIC */ +static inline void __init MP_bus_info(struct mpc_bus *m) {} +static inline void __init MP_ioapic_info(struct mpc_ioapic *m) {} +static inline void __init MP_intsrc_info(struct mpc_intsrc *m) {} +#endif /* CONFIG_X86_IO_APIC */ -#endif static void __init MP_lintsrc_info(struct mpc_lintsrc *m) { @@ -275,6 +276,20 @@ static int __init smp_check_mpc(struct mpc_table *mpc, char *oem, char *str) return 1; } +static void skip_entry(unsigned char **ptr, int *count, int size) +{ + *ptr += size; + *count += size; +} + +static void __init smp_dump_mptable(struct mpc_table *mpc, unsigned char *mpt) +{ + printk(KERN_ERR "Your mptable is wrong, contact your HW vendor!\n" + "type %x\n", *mpt); + print_hex_dump(KERN_ERR, " ", DUMP_PREFIX_ADDRESS, 16, + 1, mpc, mpc->length, 1); +} + static int __init smp_read_mpc(struct mpc_table *mpc, unsigned early) { char str[16]; @@ -310,61 +325,30 @@ static int __init smp_read_mpc(struct mpc_table *mpc, unsigned early) while (count < mpc->length) { switch (*mpt) { case MP_PROCESSOR: - { - struct mpc_cpu *m = (struct mpc_cpu *)mpt; - /* ACPI may have already provided this data */ - if (!acpi_lapic) - MP_processor_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } + /* ACPI may have already provided this data */ + if (!acpi_lapic) + MP_processor_info((struct mpc_cpu *)mpt); + skip_entry(&mpt, &count, sizeof(struct mpc_cpu)); + break; case MP_BUS: - { - struct mpc_bus *m = (struct mpc_bus *)mpt; -#ifdef CONFIG_X86_IO_APIC - MP_bus_info(m); -#endif - mpt += sizeof(*m); - count += sizeof(*m); - break; - } + MP_bus_info((struct mpc_bus *)mpt); + skip_entry(&mpt, &count, sizeof(struct mpc_bus)); + break; case MP_IOAPIC: - { -#ifdef CONFIG_X86_IO_APIC - struct mpc_ioapic *m = (struct mpc_ioapic *)mpt; - MP_ioapic_info(m); -#endif - mpt += sizeof(struct mpc_ioapic); - count += sizeof(struct mpc_ioapic); - break; - } + MP_ioapic_info((struct mpc_ioapic *)mpt); + skip_entry(&mpt, &count, sizeof(struct mpc_ioapic)); + break; case MP_INTSRC: - { -#ifdef CONFIG_X86_IO_APIC - struct mpc_intsrc *m = (struct mpc_intsrc *)mpt; - - MP_intsrc_info(m); -#endif - mpt += sizeof(struct mpc_intsrc); - count += sizeof(struct mpc_intsrc); - break; - } + MP_intsrc_info((struct mpc_intsrc *)mpt); + skip_entry(&mpt, &count, sizeof(struct mpc_intsrc)); + break; case MP_LINTSRC: - { - struct mpc_lintsrc *m = - (struct mpc_lintsrc *)mpt; - MP_lintsrc_info(m); - mpt += sizeof(*m); - count += sizeof(*m); - break; - } + MP_lintsrc_info((struct mpc_lintsrc *)mpt); + skip_entry(&mpt, &count, sizeof(struct mpc_lintsrc)); + break; default: /* wrong mptable */ - printk(KERN_ERR "Your mptable is wrong, contact your HW vendor!\n"); - printk(KERN_ERR "type %x\n", *mpt); - print_hex_dump(KERN_ERR, " ", DUMP_PREFIX_ADDRESS, 16, - 1, mpc, mpc->length, 1); + smp_dump_mptable(mpc, mpt); count = mpc->length; break; } @@ -558,6 +542,68 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) static struct mpf_intel *mpf_found; +static unsigned long __init get_mpc_size(unsigned long physptr) +{ + struct mpc_table *mpc; + unsigned long size; + + mpc = early_ioremap(physptr, PAGE_SIZE); + size = mpc->length; + early_iounmap(mpc, PAGE_SIZE); + apic_printk(APIC_VERBOSE, " mpc: %lx-%lx\n", physptr, physptr + size); + + return size; +} + +static int __init check_physptr(struct mpf_intel *mpf, unsigned int early) +{ + struct mpc_table *mpc; + unsigned long size; + + size = get_mpc_size(mpf->physptr); + mpc = early_ioremap(mpf->physptr, size); + /* + * Read the physical hardware table. Anything here will + * override the defaults. + */ + if (!smp_read_mpc(mpc, early)) { +#ifdef CONFIG_X86_LOCAL_APIC + smp_found_config = 0; +#endif + printk(KERN_ERR "BIOS bug, MP table errors detected!...\n" + "... disabling SMP support. (tell your hw vendor)\n"); + early_iounmap(mpc, size); + return -1; + } + early_iounmap(mpc, size); + + if (early) + return -1; + +#ifdef CONFIG_X86_IO_APIC + /* + * If there are no explicit MP IRQ entries, then we are + * broken. We set up most of the low 16 IO-APIC pins to + * ISA defaults and hope it will work. + */ + if (!mp_irq_entries) { + struct mpc_bus bus; + + printk(KERN_ERR "BIOS bug, no explicit IRQ entries, " + "using default mptable. (tell your hw vendor)\n"); + + bus.type = MP_BUS; + bus.busid = 0; + memcpy(bus.bustype, "ISA ", 6); + MP_bus_info(&bus); + + construct_default_ioirq_mptable(0); + } +#endif + + return 0; +} + /* * Scan the memory blocks for an SMP configuration block. */ @@ -611,45 +657,8 @@ static void __init __get_smp_config(unsigned int early) construct_default_ISA_mptable(mpf->feature1); } else if (mpf->physptr) { - - /* - * Read the physical hardware table. Anything here will - * override the defaults. - */ - if (!smp_read_mpc(phys_to_virt(mpf->physptr), early)) { -#ifdef CONFIG_X86_LOCAL_APIC - smp_found_config = 0; -#endif - printk(KERN_ERR - "BIOS bug, MP table errors detected!...\n"); - printk(KERN_ERR "... disabling SMP support. " - "(tell your hw vendor)\n"); - return; - } - - if (early) + if (check_physptr(mpf, early)) return; -#ifdef CONFIG_X86_IO_APIC - /* - * If there are no explicit MP IRQ entries, then we are - * broken. We set up most of the low 16 IO-APIC pins to - * ISA defaults and hope it will work. - */ - if (!mp_irq_entries) { - struct mpc_bus bus; - - printk(KERN_ERR "BIOS bug, no explicit IRQ entries, " - "using default mptable. " - "(tell your hw vendor)\n"); - - bus.type = MP_BUS; - bus.busid = 0; - memcpy(bus.bustype, "ISA ", 6); - MP_bus_info(&bus); - - construct_default_ioirq_mptable(0); - } -#endif } else BUG(); @@ -670,6 +679,31 @@ void __init get_smp_config(void) __get_smp_config(0); } +static void smp_reserve_bootmem(struct mpf_intel *mpf) +{ + unsigned long size = get_mpc_size(mpf->physptr); +#ifdef CONFIG_X86_32 + /* + * We cannot access to MPC table to compute table size yet, + * as only few megabytes from the bottom is mapped now. + * PC-9800's MPC table places on the very last of physical + * memory; so that simply reserving PAGE_SIZE from mpf->physptr + * yields BUG() in reserve_bootmem. + * also need to make sure physptr is below than max_low_pfn + * we don't need reserve the area above max_low_pfn + */ + unsigned long end = max_low_pfn * PAGE_SIZE; + + if (mpf->physptr < end) { + if (mpf->physptr + size > end) + size = end - mpf->physptr; + reserve_bootmem_generic(mpf->physptr, size, BOOTMEM_DEFAULT); + } +#else + reserve_bootmem_generic(mpf->physptr, size, BOOTMEM_DEFAULT); +#endif +} + static int __init smp_scan_config(unsigned long base, unsigned long length, unsigned reserve) { @@ -697,36 +731,10 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, if (!reserve) return 1; - reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE, - BOOTMEM_DEFAULT); - if (mpf->physptr) { - unsigned long size = PAGE_SIZE; -#ifdef CONFIG_X86_32 - /* - * We cannot access to MPC table to compute - * table size yet, as only few megabytes from - * the bottom is mapped now. - * PC-9800's MPC table places on the very last - * of physical memory; so that simply reserving - * PAGE_SIZE from mpf->physptr yields BUG() - * in reserve_bootmem. - * also need to make sure physptr is below than - * max_low_pfn - * we don't need reserve the area above max_low_pfn - */ - unsigned long end = max_low_pfn * PAGE_SIZE; - - if (mpf->physptr < end) { - if (mpf->physptr + size > end) - size = end - mpf->physptr; - reserve_bootmem_generic(mpf->physptr, size, - BOOTMEM_DEFAULT); - } -#else - reserve_bootmem_generic(mpf->physptr, size, + reserve_bootmem_generic(virt_to_phys(mpf), sizeof(*mpf), BOOTMEM_DEFAULT); -#endif - } + if (mpf->physptr) + smp_reserve_bootmem(mpf); return 1; } @@ -829,7 +837,57 @@ static int __init get_MP_intsrc_index(struct mpc_intsrc *m) #define SPARE_SLOT_NUM 20 static struct mpc_intsrc __initdata *m_spare[SPARE_SLOT_NUM]; -#endif + +static void check_irq_src(struct mpc_intsrc *m, int *nr_m_spare) +{ + int i; + + apic_printk(APIC_VERBOSE, "OLD "); + print_MP_intsrc_info(m); + + i = get_MP_intsrc_index(m); + if (i > 0) { + assign_to_mpc_intsrc(&mp_irqs[i], m); + apic_printk(APIC_VERBOSE, "NEW "); + print_mp_irq_info(&mp_irqs[i]); + return; + } + if (!i) { + /* legacy, do nothing */ + return; + } + if (*nr_m_spare < SPARE_SLOT_NUM) { + /* + * not found (-1), or duplicated (-2) are invalid entries, + * we need to use the slot later + */ + m_spare[*nr_m_spare] = m; + *nr_m_spare += 1; + } +} +#else /* CONFIG_X86_IO_APIC */ +static inline void check_irq_src(struct mpc_intsrc *m, int *nr_m_spare) {} +#endif /* CONFIG_X86_IO_APIC */ + +static int check_slot(unsigned long mpc_new_phys, unsigned long mpc_new_length, + int count) +{ + if (!mpc_new_phys) { + pr_info("No spare slots, try to append...take your risk, " + "new mpc_length %x\n", count); + } else { + if (count <= mpc_new_length) + pr_info("No spare slots, try to append..., " + "new mpc_length %x\n", count); + else { + pr_err("mpc_new_length %lx is too small\n", + mpc_new_length); + return -1; + } + } + + return 0; +} static int __init replace_intsrc_all(struct mpc_table *mpc, unsigned long mpc_new_phys, @@ -837,77 +895,33 @@ static int __init replace_intsrc_all(struct mpc_table *mpc, { #ifdef CONFIG_X86_IO_APIC int i; - int nr_m_spare = 0; #endif - int count = sizeof(*mpc); + int nr_m_spare = 0; unsigned char *mpt = ((unsigned char *)mpc) + count; printk(KERN_INFO "mpc_length %x\n", mpc->length); while (count < mpc->length) { switch (*mpt) { case MP_PROCESSOR: - { - struct mpc_cpu *m = (struct mpc_cpu *)mpt; - mpt += sizeof(*m); - count += sizeof(*m); - break; - } + skip_entry(&mpt, &count, sizeof(struct mpc_cpu)); + break; case MP_BUS: - { - struct mpc_bus *m = (struct mpc_bus *)mpt; - mpt += sizeof(*m); - count += sizeof(*m); - break; - } + skip_entry(&mpt, &count, sizeof(struct mpc_bus)); + break; case MP_IOAPIC: - { - mpt += sizeof(struct mpc_ioapic); - count += sizeof(struct mpc_ioapic); - break; - } + skip_entry(&mpt, &count, sizeof(struct mpc_ioapic)); + break; case MP_INTSRC: - { -#ifdef CONFIG_X86_IO_APIC - struct mpc_intsrc *m = (struct mpc_intsrc *)mpt; - - printk(KERN_INFO "OLD "); - print_MP_intsrc_info(m); - i = get_MP_intsrc_index(m); - if (i > 0) { - assign_to_mpc_intsrc(&mp_irqs[i], m); - printk(KERN_INFO "NEW "); - print_mp_irq_info(&mp_irqs[i]); - } else if (!i) { - /* legacy, do nothing */ - } else if (nr_m_spare < SPARE_SLOT_NUM) { - /* - * not found (-1), or duplicated (-2) - * are invalid entries, - * we need to use the slot later - */ - m_spare[nr_m_spare] = m; - nr_m_spare++; - } -#endif - mpt += sizeof(struct mpc_intsrc); - count += sizeof(struct mpc_intsrc); - break; - } + check_irq_src((struct mpc_intsrc *)mpt, &nr_m_spare); + skip_entry(&mpt, &count, sizeof(struct mpc_intsrc)); + break; case MP_LINTSRC: - { - struct mpc_lintsrc *m = - (struct mpc_lintsrc *)mpt; - mpt += sizeof(*m); - count += sizeof(*m); - break; - } + skip_entry(&mpt, &count, sizeof(struct mpc_lintsrc)); + break; default: /* wrong mptable */ - printk(KERN_ERR "Your mptable is wrong, contact your HW vendor!\n"); - printk(KERN_ERR "type %x\n", *mpt); - print_hex_dump(KERN_ERR, " ", DUMP_PREFIX_ADDRESS, 16, - 1, mpc, mpc->length, 1); + smp_dump_mptable(mpc, mpt); goto out; } } @@ -924,23 +938,15 @@ static int __init replace_intsrc_all(struct mpc_table *mpc, continue; if (nr_m_spare > 0) { - printk(KERN_INFO "*NEW* found "); + apic_printk(APIC_VERBOSE, "*NEW* found\n"); nr_m_spare--; assign_to_mpc_intsrc(&mp_irqs[i], m_spare[nr_m_spare]); m_spare[nr_m_spare] = NULL; } else { struct mpc_intsrc *m = (struct mpc_intsrc *)mpt; count += sizeof(struct mpc_intsrc); - if (!mpc_new_phys) { - printk(KERN_INFO "No spare slots, try to append...take your risk, new mpc_length %x\n", count); - } else { - if (count <= mpc_new_length) - printk(KERN_INFO "No spare slots, try to append..., new mpc_length %x\n", count); - else { - printk(KERN_ERR "mpc_new_length %lx is too small\n", mpc_new_length); - goto out; - } - } + if (!check_slot(mpc_new_phys, mpc_new_length, count)) + goto out; assign_to_mpc_intsrc(&mp_irqs[i], m); mpc->length = count; mpt += sizeof(struct mpc_intsrc); diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 63dd358d8ee1..8e45f4464880 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -470,7 +470,6 @@ struct pv_mmu_ops pv_mmu_ops = { #if PAGETABLE_LEVELS >= 3 #ifdef CONFIG_X86_PAE .set_pte_atomic = native_set_pte_atomic, - .set_pte_present = native_set_pte_present, .pte_clear = native_pte_clear, .pmd_clear = native_pmd_clear, #endif diff --git a/arch/x86/kernel/pci-nommu.c b/arch/x86/kernel/pci-nommu.c index c70ab5a5d4c8..8b02a3936d42 100644 --- a/arch/x86/kernel/pci-nommu.c +++ b/arch/x86/kernel/pci-nommu.c @@ -1,14 +1,14 @@ /* Fallback functions when the main IOMMU code is not compiled in. This code is roughly equivalent to i386. */ -#include <linux/mm.h> -#include <linux/init.h> -#include <linux/pci.h> -#include <linux/string.h> #include <linux/dma-mapping.h> #include <linux/scatterlist.h> +#include <linux/string.h> +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/mm.h> -#include <asm/iommu.h> #include <asm/processor.h> +#include <asm/iommu.h> #include <asm/dma.h> static int @@ -79,11 +79,11 @@ static void nommu_free_coherent(struct device *dev, size_t size, void *vaddr, } struct dma_mapping_ops nommu_dma_ops = { - .alloc_coherent = dma_generic_alloc_coherent, - .free_coherent = nommu_free_coherent, - .map_single = nommu_map_single, - .map_sg = nommu_map_sg, - .is_phys = 1, + .alloc_coherent = dma_generic_alloc_coherent, + .free_coherent = nommu_free_coherent, + .map_single = nommu_map_single, + .map_sg = nommu_map_sg, + .is_phys = 1, }; void __init no_iommu_init(void) diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 6afa5232dbb7..156f87582c6c 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -65,11 +65,11 @@ void exit_thread(void) { struct task_struct *me = current; struct thread_struct *t = &me->thread; + unsigned long *bp = t->io_bitmap_ptr; - if (me->thread.io_bitmap_ptr) { + if (bp) { struct tss_struct *tss = &per_cpu(init_tss, get_cpu()); - kfree(t->io_bitmap_ptr); t->io_bitmap_ptr = NULL; clear_thread_flag(TIF_IO_BITMAP); /* @@ -78,6 +78,7 @@ void exit_thread(void) memset(tss->io_bitmap, 0xff, t->io_bitmap_max); t->io_bitmap_max = 0; put_cpu(); + kfree(bp); } ds_exit_thread(current); diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 3d9672e59c16..19378715f415 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -685,9 +685,8 @@ static int ptrace_bts_config(struct task_struct *child, if (!cfg.signal) return -EINVAL; - return -EOPNOTSUPP; - child->thread.bts_ovfl_signal = cfg.signal; + return -EOPNOTSUPP; } if ((cfg.flags & PTRACE_BTS_O_ALLOC) && diff --git a/arch/x86/kernel/quirks.c b/arch/x86/kernel/quirks.c index 697d1b78cfbf..e95022e4f5d5 100644 --- a/arch/x86/kernel/quirks.c +++ b/arch/x86/kernel/quirks.c @@ -74,8 +74,7 @@ static void ich_force_hpet_resume(void) if (!force_hpet_address) return; - if (rcba_base == NULL) - BUG(); + BUG_ON(rcba_base == NULL); /* read the Function Disable register, dword mode only */ val = readl(rcba_base + 0x3404); diff --git a/arch/x86/kernel/relocate_kernel_32.S b/arch/x86/kernel/relocate_kernel_32.S index 2064d0aa8d28..41235531b11c 100644 --- a/arch/x86/kernel/relocate_kernel_32.S +++ b/arch/x86/kernel/relocate_kernel_32.S @@ -17,7 +17,8 @@ #define PTR(x) (x << 2) -/* control_page + KEXEC_CONTROL_CODE_MAX_SIZE +/* + * control_page + KEXEC_CONTROL_CODE_MAX_SIZE * ~ control_page + PAGE_SIZE are used as data storage and stack for * jumping back */ @@ -76,8 +77,10 @@ relocate_kernel: movl %eax, CP_PA_SWAP_PAGE(%edi) movl %ebx, CP_PA_BACKUP_PAGES_MAP(%edi) - /* get physical address of control page now */ - /* this is impossible after page table switch */ + /* + * get physical address of control page now + * this is impossible after page table switch + */ movl PTR(PA_CONTROL_PAGE)(%ebp), %edi /* switch to new set of page tables */ @@ -97,7 +100,8 @@ identity_mapped: /* store the start address on the stack */ pushl %edx - /* Set cr0 to a known state: + /* + * Set cr0 to a known state: * - Paging disabled * - Alignment check disabled * - Write protect disabled @@ -113,7 +117,8 @@ identity_mapped: /* clear cr4 if applicable */ testl %ecx, %ecx jz 1f - /* Set cr4 to a known state: + /* + * Set cr4 to a known state: * Setting everything to zero seems safe. */ xorl %eax, %eax @@ -132,15 +137,18 @@ identity_mapped: call swap_pages addl $8, %esp - /* To be certain of avoiding problems with self-modifying code + /* + * To be certain of avoiding problems with self-modifying code * I need to execute a serializing instruction here. * So I flush the TLB, it's handy, and not processor dependent. */ xorl %eax, %eax movl %eax, %cr3 - /* set all of the registers to known values */ - /* leave %esp alone */ + /* + * set all of the registers to known values + * leave %esp alone + */ testl %esi, %esi jnz 1f diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S index d32cfb27a479..4de8f5b3d476 100644 --- a/arch/x86/kernel/relocate_kernel_64.S +++ b/arch/x86/kernel/relocate_kernel_64.S @@ -19,29 +19,77 @@ #define PTR(x) (x << 3) #define PAGE_ATTR (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | _PAGE_DIRTY) +/* + * control_page + KEXEC_CONTROL_CODE_MAX_SIZE + * ~ control_page + PAGE_SIZE are used as data storage and stack for + * jumping back + */ +#define DATA(offset) (KEXEC_CONTROL_CODE_MAX_SIZE+(offset)) + +/* Minimal CPU state */ +#define RSP DATA(0x0) +#define CR0 DATA(0x8) +#define CR3 DATA(0x10) +#define CR4 DATA(0x18) + +/* other data */ +#define CP_PA_TABLE_PAGE DATA(0x20) +#define CP_PA_SWAP_PAGE DATA(0x28) +#define CP_PA_BACKUP_PAGES_MAP DATA(0x30) + .text .align PAGE_SIZE .code64 .globl relocate_kernel relocate_kernel: - /* %rdi indirection_page + /* + * %rdi indirection_page * %rsi page_list * %rdx start address + * %rcx preserve_context */ + /* Save the CPU context, used for jumping back */ + pushq %rbx + pushq %rbp + pushq %r12 + pushq %r13 + pushq %r14 + pushq %r15 + pushf + + movq PTR(VA_CONTROL_PAGE)(%rsi), %r11 + movq %rsp, RSP(%r11) + movq %cr0, %rax + movq %rax, CR0(%r11) + movq %cr3, %rax + movq %rax, CR3(%r11) + movq %cr4, %rax + movq %rax, CR4(%r11) + /* zero out flags, and disable interrupts */ pushq $0 popfq - /* get physical address of control page now */ - /* this is impossible after page table switch */ + /* + * get physical address of control page now + * this is impossible after page table switch + */ movq PTR(PA_CONTROL_PAGE)(%rsi), %r8 /* get physical address of page table now too */ - movq PTR(PA_TABLE_PAGE)(%rsi), %rcx + movq PTR(PA_TABLE_PAGE)(%rsi), %r9 + + /* get physical address of swap page now */ + movq PTR(PA_SWAP_PAGE)(%rsi), %r10 + + /* save some information for jumping back */ + movq %r9, CP_PA_TABLE_PAGE(%r11) + movq %r10, CP_PA_SWAP_PAGE(%r11) + movq %rdi, CP_PA_BACKUP_PAGES_MAP(%r11) /* Switch to the identity mapped page tables */ - movq %rcx, %cr3 + movq %r9, %cr3 /* setup a new stack at the end of the physical control page */ lea PAGE_SIZE(%r8), %rsp @@ -55,7 +103,8 @@ identity_mapped: /* store the start address on the stack */ pushq %rdx - /* Set cr0 to a known state: + /* + * Set cr0 to a known state: * - Paging enabled * - Alignment check disabled * - Write protect disabled @@ -68,7 +117,8 @@ identity_mapped: orl $(X86_CR0_PG | X86_CR0_PE), %eax movq %rax, %cr0 - /* Set cr4 to a known state: + /* + * Set cr4 to a known state: * - physical address extension enabled */ movq $X86_CR4_PAE, %rax @@ -78,9 +128,87 @@ identity_mapped: 1: /* Flush the TLB (needed?) */ - movq %rcx, %cr3 + movq %r9, %cr3 + + movq %rcx, %r11 + call swap_pages + + /* + * To be certain of avoiding problems with self-modifying code + * I need to execute a serializing instruction here. + * So I flush the TLB by reloading %cr3 here, it's handy, + * and not processor dependent. + */ + movq %cr3, %rax + movq %rax, %cr3 + + /* + * set all of the registers to known values + * leave %rsp alone + */ + + testq %r11, %r11 + jnz 1f + xorq %rax, %rax + xorq %rbx, %rbx + xorq %rcx, %rcx + xorq %rdx, %rdx + xorq %rsi, %rsi + xorq %rdi, %rdi + xorq %rbp, %rbp + xorq %r8, %r8 + xorq %r9, %r9 + xorq %r10, %r9 + xorq %r11, %r11 + xorq %r12, %r12 + xorq %r13, %r13 + xorq %r14, %r14 + xorq %r15, %r15 + + ret + +1: + popq %rdx + leaq PAGE_SIZE(%r10), %rsp + call *%rdx + + /* get the re-entry point of the peer system */ + movq 0(%rsp), %rbp + call 1f +1: + popq %r8 + subq $(1b - relocate_kernel), %r8 + movq CP_PA_SWAP_PAGE(%r8), %r10 + movq CP_PA_BACKUP_PAGES_MAP(%r8), %rdi + movq CP_PA_TABLE_PAGE(%r8), %rax + movq %rax, %cr3 + lea PAGE_SIZE(%r8), %rsp + call swap_pages + movq $virtual_mapped, %rax + pushq %rax + ret + +virtual_mapped: + movq RSP(%r8), %rsp + movq CR4(%r8), %rax + movq %rax, %cr4 + movq CR3(%r8), %rax + movq CR0(%r8), %r8 + movq %rax, %cr3 + movq %r8, %cr0 + movq %rbp, %rax + + popf + popq %r15 + popq %r14 + popq %r13 + popq %r12 + popq %rbp + popq %rbx + ret /* Do the copies */ +swap_pages: movq %rdi, %rcx /* Put the page_list in %rcx */ xorq %rdi, %rdi xorq %rsi, %rsi @@ -112,36 +240,27 @@ identity_mapped: movq %rcx, %rsi /* For ever source page do a copy */ andq $0xfffffffffffff000, %rsi + movq %rdi, %rdx + movq %rsi, %rax + + movq %r10, %rdi movq $512, %rcx rep ; movsq - jmp 0b -3: - - /* To be certain of avoiding problems with self-modifying code - * I need to execute a serializing instruction here. - * So I flush the TLB by reloading %cr3 here, it's handy, - * and not processor dependent. - */ - movq %cr3, %rax - movq %rax, %cr3 - /* set all of the registers to known values */ - /* leave %rsp alone */ + movq %rax, %rdi + movq %rdx, %rsi + movq $512, %rcx + rep ; movsq - xorq %rax, %rax - xorq %rbx, %rbx - xorq %rcx, %rcx - xorq %rdx, %rdx - xorq %rsi, %rsi - xorq %rdi, %rdi - xorq %rbp, %rbp - xorq %r8, %r8 - xorq %r9, %r9 - xorq %r10, %r9 - xorq %r11, %r11 - xorq %r12, %r12 - xorq %r13, %r13 - xorq %r14, %r14 - xorq %r15, %r15 + movq %rdx, %rdi + movq %r10, %rsi + movq $512, %rcx + rep ; movsq + lea PAGE_SIZE(%rax), %rsi + jmp 0b +3: ret + + .globl kexec_control_code_size +.set kexec_control_code_size, . - relocate_kernel diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index dd6f2b71561b..5d465b207e72 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -1,14 +1,14 @@ /* * RTC related functions */ +#include <linux/platform_device.h> +#include <linux/mc146818rtc.h> #include <linux/acpi.h> #include <linux/bcd.h> -#include <linux/mc146818rtc.h> -#include <linux/platform_device.h> #include <linux/pnp.h> -#include <asm/time.h> #include <asm/vsyscall.h> +#include <asm/time.h> #ifdef CONFIG_X86_32 /* @@ -16,9 +16,9 @@ * register we are working with. It is required for NMI access to the * CMOS/RTC registers. See include/asm-i386/mc146818rtc.h for details. */ -volatile unsigned long cmos_lock = 0; +volatile unsigned long cmos_lock; EXPORT_SYMBOL(cmos_lock); -#endif +#endif /* CONFIG_X86_32 */ /* For two digit years assume time is always after that */ #define CMOS_YEARS_OFFS 2000 @@ -38,9 +38,9 @@ EXPORT_SYMBOL(rtc_lock); */ int mach_set_rtc_mmss(unsigned long nowtime) { - int retval = 0; int real_seconds, real_minutes, cmos_minutes; unsigned char save_control, save_freq_select; + int retval = 0; /* tell the clock it's being set */ save_control = CMOS_READ(RTC_CONTROL); @@ -72,8 +72,8 @@ int mach_set_rtc_mmss(unsigned long nowtime) real_seconds = bin2bcd(real_seconds); real_minutes = bin2bcd(real_minutes); } - CMOS_WRITE(real_seconds,RTC_SECONDS); - CMOS_WRITE(real_minutes,RTC_MINUTES); + CMOS_WRITE(real_seconds, RTC_SECONDS); + CMOS_WRITE(real_minutes, RTC_MINUTES); } else { printk(KERN_WARNING "set_rtc_mmss: can't update from %d to %d\n", @@ -151,6 +151,7 @@ unsigned char rtc_cmos_read(unsigned char addr) outb(addr, RTC_PORT(0)); val = inb(RTC_PORT(1)); lock_cmos_suffix(addr); + return val; } EXPORT_SYMBOL(rtc_cmos_read); @@ -166,8 +167,8 @@ EXPORT_SYMBOL(rtc_cmos_write); static int set_rtc_mmss(unsigned long nowtime) { - int retval; unsigned long flags; + int retval; spin_lock_irqsave(&rtc_lock, flags); retval = set_wallclock(nowtime); @@ -242,6 +243,7 @@ static __init int add_rtc_cmos(void) platform_device_register(&rtc_device); dev_info(&rtc_device.dev, "registered platform RTC device (no PNP device found)\n"); + return 0; } device_initcall(add_rtc_cmos); diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index b746deb9ebc6..a0d26237d7cf 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -112,8 +112,13 @@ #define ARCH_SETUP #endif +RESERVE_BRK(dmi_alloc, 65536); + unsigned int boot_cpu_id __read_mostly; +static __initdata unsigned long _brk_start = (unsigned long)__brk_base; +unsigned long _brk_end = (unsigned long)__brk_base; + #ifdef CONFIG_X86_64 int default_cpu_present_to_apicid(int mps_cpu) { @@ -158,12 +163,6 @@ static struct resource bss_resource = { #ifdef CONFIG_X86_32 -/* This value is set up by the early boot code to point to the value - immediately after the boot time page tables. It contains a *physical* - address, and must not be in the .bss segment! */ -unsigned long init_pg_tables_start __initdata = ~0UL; -unsigned long init_pg_tables_end __initdata = ~0UL; - static struct resource video_ram_resource = { .name = "Video RAM area", .start = 0xa0000, @@ -202,7 +201,9 @@ struct ist_info ist_info; #endif #else -struct cpuinfo_x86 boot_cpu_data __read_mostly; +struct cpuinfo_x86 boot_cpu_data __read_mostly = { + .x86_phys_bits = MAX_PHYSMEM_BITS, +}; EXPORT_SYMBOL(boot_cpu_data); #endif @@ -217,12 +218,6 @@ unsigned long mmu_cr4_features = X86_CR4_PAE; int bootloader_type; /* - * Early DMI memory - */ -int dmi_alloc_index; -char dmi_alloc_data[DMI_MAX_DATA]; - -/* * Setup options */ struct screen_info screen_info; @@ -267,6 +262,35 @@ static inline void copy_edd(void) } #endif +void * __init extend_brk(size_t size, size_t align) +{ + size_t mask = align - 1; + void *ret; + + BUG_ON(_brk_start == 0); + BUG_ON(align & mask); + + _brk_end = (_brk_end + mask) & ~mask; + BUG_ON((char *)(_brk_end + size) > __brk_limit); + + ret = (void *)_brk_end; + _brk_end += size; + + memset(ret, 0, size); + + return ret; +} + +static void __init reserve_brk(void) +{ + if (_brk_end > _brk_start) + reserve_early(__pa(_brk_start), __pa(_brk_end), "BRK"); + + /* Mark brk area as locked down and no longer taking any + new allocations */ + _brk_start = 0; +} + #ifdef CONFIG_BLK_DEV_INITRD #ifdef CONFIG_X86_32 @@ -715,11 +739,7 @@ void __init setup_arch(char **cmdline_p) init_mm.start_code = (unsigned long) _text; init_mm.end_code = (unsigned long) _etext; init_mm.end_data = (unsigned long) _edata; -#ifdef CONFIG_X86_32 - init_mm.brk = init_pg_tables_end + PAGE_OFFSET; -#else - init_mm.brk = (unsigned long) &_end; -#endif + init_mm.brk = _brk_end; code_resource.start = virt_to_phys(_text); code_resource.end = virt_to_phys(_etext)-1; @@ -840,6 +860,8 @@ void __init setup_arch(char **cmdline_p) setup_bios_corruption_check(); #endif + reserve_brk(); + /* max_pfn_mapped is updated here */ max_low_pfn_mapped = init_memory_mapping(0, max_low_pfn<<PAGE_SHIFT); max_pfn_mapped = max_low_pfn_mapped; diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c index d2cc6428c587..dfcc74ab0ab6 100644 --- a/arch/x86/kernel/signal.c +++ b/arch/x86/kernel/signal.c @@ -211,31 +211,27 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size, { /* Default to using normal stack */ unsigned long sp = regs->sp; + int onsigstack = on_sig_stack(sp); #ifdef CONFIG_X86_64 /* redzone */ sp -= 128; #endif /* CONFIG_X86_64 */ - /* - * If we are on the alternate signal stack and would overflow it, don't. - * Return an always-bogus address instead so we will die with SIGSEGV. - */ - if (on_sig_stack(sp) && !likely(on_sig_stack(sp - frame_size))) - return (void __user *) -1L; - - /* This is the X/Open sanctioned signal stack switching. */ - if (ka->sa.sa_flags & SA_ONSTACK) { - if (sas_ss_flags(sp) == 0) - sp = current->sas_ss_sp + current->sas_ss_size; - } else { + if (!onsigstack) { + /* This is the X/Open sanctioned signal stack switching. */ + if (ka->sa.sa_flags & SA_ONSTACK) { + if (sas_ss_flags(sp) == 0) + sp = current->sas_ss_sp + current->sas_ss_size; + } else { #ifdef CONFIG_X86_32 - /* This is the legacy signal stack switching. */ - if ((regs->ss & 0xffff) != __USER_DS && - !(ka->sa.sa_flags & SA_RESTORER) && - ka->sa.sa_restorer) - sp = (unsigned long) ka->sa.sa_restorer; + /* This is the legacy signal stack switching. */ + if ((regs->ss & 0xffff) != __USER_DS && + !(ka->sa.sa_flags & SA_RESTORER) && + ka->sa.sa_restorer) + sp = (unsigned long) ka->sa.sa_restorer; #endif /* CONFIG_X86_32 */ + } } if (used_math()) { @@ -244,12 +240,22 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size, sp = round_down(sp, 64); #endif /* CONFIG_X86_64 */ *fpstate = (void __user *)sp; - - if (save_i387_xstate(*fpstate) < 0) - return (void __user *)-1L; } - return (void __user *)align_sigframe(sp - frame_size); + sp = align_sigframe(sp - frame_size); + + /* + * If we are on the alternate signal stack and would overflow it, don't. + * Return an always-bogus address instead so we will die with SIGSEGV. + */ + if (onsigstack && !likely(on_sig_stack(sp))) + return (void __user *)-1L; + + /* save i387 state */ + if (used_math() && save_i387_xstate(*fpstate) < 0) + return (void __user *)-1L; + + return (void __user *)sp; } #ifdef CONFIG_X86_32 diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 249334f5080a..ef7d10170c30 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -114,10 +114,6 @@ EXPORT_PER_CPU_SYMBOL(cpu_info); atomic_t init_deasserted; - -/* Set if we find a B stepping CPU */ -static int __cpuinitdata smp_b_stepping; - #if defined(CONFIG_NUMA) && defined(CONFIG_X86_32) /* which logical CPUs are on which nodes */ @@ -271,8 +267,6 @@ static void __cpuinit smp_callin(void) cpumask_set_cpu(cpuid, cpu_callin_mask); } -static int __cpuinitdata unsafe_smp; - /* * Activate a secondary processor. */ @@ -340,76 +334,6 @@ notrace static void __cpuinit start_secondary(void *unused) cpu_idle(); } -static void __cpuinit smp_apply_quirks(struct cpuinfo_x86 *c) -{ - /* - * Mask B, Pentium, but not Pentium MMX - */ - if (c->x86_vendor == X86_VENDOR_INTEL && - c->x86 == 5 && - c->x86_mask >= 1 && c->x86_mask <= 4 && - c->x86_model <= 3) - /* - * Remember we have B step Pentia with bugs - */ - smp_b_stepping = 1; - - /* - * Certain Athlons might work (for various values of 'work') in SMP - * but they are not certified as MP capable. - */ - if ((c->x86_vendor == X86_VENDOR_AMD) && (c->x86 == 6)) { - - if (num_possible_cpus() == 1) - goto valid_k7; - - /* Athlon 660/661 is valid. */ - if ((c->x86_model == 6) && ((c->x86_mask == 0) || - (c->x86_mask == 1))) - goto valid_k7; - - /* Duron 670 is valid */ - if ((c->x86_model == 7) && (c->x86_mask == 0)) - goto valid_k7; - - /* - * Athlon 662, Duron 671, and Athlon >model 7 have capability - * bit. It's worth noting that the A5 stepping (662) of some - * Athlon XP's have the MP bit set. - * See http://www.heise.de/newsticker/data/jow-18.10.01-000 for - * more. - */ - if (((c->x86_model == 6) && (c->x86_mask >= 2)) || - ((c->x86_model == 7) && (c->x86_mask >= 1)) || - (c->x86_model > 7)) - if (cpu_has_mp) - goto valid_k7; - - /* If we get here, not a certified SMP capable AMD system. */ - unsafe_smp = 1; - } - -valid_k7: - ; -} - -static void __cpuinit smp_checks(void) -{ - if (smp_b_stepping) - printk(KERN_WARNING "WARNING: SMP operation may be unreliable" - "with B stepping processors.\n"); - - /* - * Don't taint if we are running SMP kernel on a single non-MP - * approved Athlon - */ - if (unsafe_smp && num_online_cpus() > 1) { - printk(KERN_INFO "WARNING: This combination of AMD" - "processors is not suitable for SMP.\n"); - add_taint(TAINT_UNSAFE_SMP); - } -} - /* * The bootstrap kernel entry code has set these up. Save them for * a given CPU @@ -423,7 +347,6 @@ void __cpuinit smp_store_cpu_info(int id) c->cpu_index = id; if (id != 0) identify_secondary_cpu(c); - smp_apply_quirks(c); } @@ -1193,7 +1116,6 @@ void __init native_smp_cpus_done(unsigned int max_cpus) pr_debug("Boot done.\n"); impress_friends(); - smp_checks(); #ifdef CONFIG_X86_IO_APIC setup_ioapic_dest(); #endif diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c index d038b9c45cf8..79c073247284 100644 --- a/arch/x86/kernel/tlb_uv.c +++ b/arch/x86/kernel/tlb_uv.c @@ -750,7 +750,7 @@ static int __init uv_bau_init(void) int node; int nblades; int last_blade; - int cur_cpu = 0; + int cur_cpu; if (!is_uv_system()) return 0; @@ -760,6 +760,7 @@ static int __init uv_bau_init(void) uv_mmask = (1UL << uv_hub_info->n_val) - 1; nblades = 0; last_blade = -1; + cur_cpu = 0; for_each_online_node(node) { blade = uv_node_to_blade_id(node); if (blade == last_blade) diff --git a/arch/x86/kernel/topology.c b/arch/x86/kernel/topology.c index 0fcc95a354f7..7e4515957a1c 100644 --- a/arch/x86/kernel/topology.c +++ b/arch/x86/kernel/topology.c @@ -25,10 +25,10 @@ * * Send feedback to <colpatch@us.ibm.com> */ -#include <linux/init.h> -#include <linux/smp.h> #include <linux/nodemask.h> #include <linux/mmzone.h> +#include <linux/init.h> +#include <linux/smp.h> #include <asm/cpu.h> static DEFINE_PER_CPU(struct x86_cpu, cpu_devices); @@ -47,6 +47,7 @@ int __ref arch_register_cpu(int num) */ if (num) per_cpu(cpu_devices, num).cpu.hotpluggable = 1; + return register_cpu(&per_cpu(cpu_devices, num).cpu, num); } EXPORT_SYMBOL(arch_register_cpu); @@ -56,12 +57,13 @@ void arch_unregister_cpu(int num) unregister_cpu(&per_cpu(cpu_devices, num).cpu); } EXPORT_SYMBOL(arch_unregister_cpu); -#else +#else /* CONFIG_HOTPLUG_CPU */ + static int __init arch_register_cpu(int num) { return register_cpu(&per_cpu(cpu_devices, num).cpu, num); } -#endif /*CONFIG_HOTPLUG_CPU*/ +#endif /* CONFIG_HOTPLUG_CPU */ static int __init topology_init(void) { @@ -70,11 +72,11 @@ static int __init topology_init(void) #ifdef CONFIG_NUMA for_each_online_node(i) register_one_node(i); -#endif /* CONFIG_NUMA */ +#endif for_each_present_cpu(i) arch_register_cpu(i); + return 0; } - subsys_initcall(topology_init); diff --git a/arch/x86/kernel/uv_time.c b/arch/x86/kernel/uv_time.c new file mode 100644 index 000000000000..2ffb6c53326e --- /dev/null +++ b/arch/x86/kernel/uv_time.c @@ -0,0 +1,393 @@ +/* + * SGI RTC clock/timer routines. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Copyright (c) 2009 Silicon Graphics, Inc. All Rights Reserved. + * Copyright (c) Dimitri Sivanich + */ +#include <linux/clockchips.h> + +#include <asm/uv/uv_mmrs.h> +#include <asm/uv/uv_hub.h> +#include <asm/uv/bios.h> +#include <asm/uv/uv.h> +#include <asm/apic.h> +#include <asm/cpu.h> + +#define RTC_NAME "sgi_rtc" + +static cycle_t uv_read_rtc(void); +static int uv_rtc_next_event(unsigned long, struct clock_event_device *); +static void uv_rtc_timer_setup(enum clock_event_mode, + struct clock_event_device *); + +static struct clocksource clocksource_uv = { + .name = RTC_NAME, + .rating = 400, + .read = uv_read_rtc, + .mask = (cycle_t)UVH_RTC_REAL_TIME_CLOCK_MASK, + .shift = 10, + .flags = CLOCK_SOURCE_IS_CONTINUOUS, +}; + +static struct clock_event_device clock_event_device_uv = { + .name = RTC_NAME, + .features = CLOCK_EVT_FEAT_ONESHOT, + .shift = 20, + .rating = 400, + .irq = -1, + .set_next_event = uv_rtc_next_event, + .set_mode = uv_rtc_timer_setup, + .event_handler = NULL, +}; + +static DEFINE_PER_CPU(struct clock_event_device, cpu_ced); + +/* There is one of these allocated per node */ +struct uv_rtc_timer_head { + spinlock_t lock; + /* next cpu waiting for timer, local node relative: */ + int next_cpu; + /* number of cpus on this node: */ + int ncpus; + struct { + int lcpu; /* systemwide logical cpu number */ + u64 expires; /* next timer expiration for this cpu */ + } cpu[1]; +}; + +/* + * Access to uv_rtc_timer_head via blade id. + */ +static struct uv_rtc_timer_head **blade_info __read_mostly; + +static int uv_rtc_enable; + +/* + * Hardware interface routines + */ + +/* Send IPIs to another node */ +static void uv_rtc_send_IPI(int cpu) +{ + unsigned long apicid, val; + int pnode; + + apicid = cpu_physical_id(cpu); + pnode = uv_apicid_to_pnode(apicid); + val = (1UL << UVH_IPI_INT_SEND_SHFT) | + (apicid << UVH_IPI_INT_APIC_ID_SHFT) | + (GENERIC_INTERRUPT_VECTOR << UVH_IPI_INT_VECTOR_SHFT); + + uv_write_global_mmr64(pnode, UVH_IPI_INT, val); +} + +/* Check for an RTC interrupt pending */ +static int uv_intr_pending(int pnode) +{ + return uv_read_global_mmr64(pnode, UVH_EVENT_OCCURRED0) & + UVH_EVENT_OCCURRED0_RTC1_MASK; +} + +/* Setup interrupt and return non-zero if early expiration occurred. */ +static int uv_setup_intr(int cpu, u64 expires) +{ + u64 val; + int pnode = uv_cpu_to_pnode(cpu); + + uv_write_global_mmr64(pnode, UVH_RTC1_INT_CONFIG, + UVH_RTC1_INT_CONFIG_M_MASK); + uv_write_global_mmr64(pnode, UVH_INT_CMPB, -1L); + + uv_write_global_mmr64(pnode, UVH_EVENT_OCCURRED0_ALIAS, + UVH_EVENT_OCCURRED0_RTC1_MASK); + + val = (GENERIC_INTERRUPT_VECTOR << UVH_RTC1_INT_CONFIG_VECTOR_SHFT) | + ((u64)cpu_physical_id(cpu) << UVH_RTC1_INT_CONFIG_APIC_ID_SHFT); + + /* Set configuration */ + uv_write_global_mmr64(pnode, UVH_RTC1_INT_CONFIG, val); + /* Initialize comparator value */ + uv_write_global_mmr64(pnode, UVH_INT_CMPB, expires); + + return (expires < uv_read_rtc() && !uv_intr_pending(pnode)); +} + +/* + * Per-cpu timer tracking routines + */ + +static __init void uv_rtc_deallocate_timers(void) +{ + int bid; + + for_each_possible_blade(bid) { + kfree(blade_info[bid]); + } + kfree(blade_info); +} + +/* Allocate per-node list of cpu timer expiration times. */ +static __init int uv_rtc_allocate_timers(void) +{ + int cpu; + + blade_info = kmalloc(uv_possible_blades * sizeof(void *), GFP_KERNEL); + if (!blade_info) + return -ENOMEM; + memset(blade_info, 0, uv_possible_blades * sizeof(void *)); + + for_each_present_cpu(cpu) { + int nid = cpu_to_node(cpu); + int bid = uv_cpu_to_blade_id(cpu); + int bcpu = uv_cpu_hub_info(cpu)->blade_processor_id; + struct uv_rtc_timer_head *head = blade_info[bid]; + + if (!head) { + head = kmalloc_node(sizeof(struct uv_rtc_timer_head) + + (uv_blade_nr_possible_cpus(bid) * + 2 * sizeof(u64)), + GFP_KERNEL, nid); + if (!head) { + uv_rtc_deallocate_timers(); + return -ENOMEM; + } + spin_lock_init(&head->lock); + head->ncpus = uv_blade_nr_possible_cpus(bid); + head->next_cpu = -1; + blade_info[bid] = head; + } + + head->cpu[bcpu].lcpu = cpu; + head->cpu[bcpu].expires = ULLONG_MAX; + } + + return 0; +} + +/* Find and set the next expiring timer. */ +static void uv_rtc_find_next_timer(struct uv_rtc_timer_head *head, int pnode) +{ + u64 lowest = ULLONG_MAX; + int c, bcpu = -1; + + head->next_cpu = -1; + for (c = 0; c < head->ncpus; c++) { + u64 exp = head->cpu[c].expires; + if (exp < lowest) { + bcpu = c; + lowest = exp; + } + } + if (bcpu >= 0) { + head->next_cpu = bcpu; + c = head->cpu[bcpu].lcpu; + if (uv_setup_intr(c, lowest)) + /* If we didn't set it up in time, trigger */ + uv_rtc_send_IPI(c); + } else { + uv_write_global_mmr64(pnode, UVH_RTC1_INT_CONFIG, + UVH_RTC1_INT_CONFIG_M_MASK); + } +} + +/* + * Set expiration time for current cpu. + * + * Returns 1 if we missed the expiration time. + */ +static int uv_rtc_set_timer(int cpu, u64 expires) +{ + int pnode = uv_cpu_to_pnode(cpu); + int bid = uv_cpu_to_blade_id(cpu); + struct uv_rtc_timer_head *head = blade_info[bid]; + int bcpu = uv_cpu_hub_info(cpu)->blade_processor_id; + u64 *t = &head->cpu[bcpu].expires; + unsigned long flags; + int next_cpu; + + spin_lock_irqsave(&head->lock, flags); + + next_cpu = head->next_cpu; + *t = expires; + /* Will this one be next to go off? */ + if (next_cpu < 0 || bcpu == next_cpu || + expires < head->cpu[next_cpu].expires) { + head->next_cpu = bcpu; + if (uv_setup_intr(cpu, expires)) { + *t = ULLONG_MAX; + uv_rtc_find_next_timer(head, pnode); + spin_unlock_irqrestore(&head->lock, flags); + return 1; + } + } + + spin_unlock_irqrestore(&head->lock, flags); + return 0; +} + +/* + * Unset expiration time for current cpu. + * + * Returns 1 if this timer was pending. + */ +static int uv_rtc_unset_timer(int cpu) +{ + int pnode = uv_cpu_to_pnode(cpu); + int bid = uv_cpu_to_blade_id(cpu); + struct uv_rtc_timer_head *head = blade_info[bid]; + int bcpu = uv_cpu_hub_info(cpu)->blade_processor_id; + u64 *t = &head->cpu[bcpu].expires; + unsigned long flags; + int rc = 0; + + spin_lock_irqsave(&head->lock, flags); + + if (head->next_cpu == bcpu && uv_read_rtc() >= *t) + rc = 1; + + *t = ULLONG_MAX; + + /* Was the hardware setup for this timer? */ + if (head->next_cpu == bcpu) + uv_rtc_find_next_timer(head, pnode); + + spin_unlock_irqrestore(&head->lock, flags); + + return rc; +} + + +/* + * Kernel interface routines. + */ + +/* + * Read the RTC. + */ +static cycle_t uv_read_rtc(void) +{ + return (cycle_t)uv_read_local_mmr(UVH_RTC); +} + +/* + * Program the next event, relative to now + */ +static int uv_rtc_next_event(unsigned long delta, + struct clock_event_device *ced) +{ + int ced_cpu = cpumask_first(ced->cpumask); + + return uv_rtc_set_timer(ced_cpu, delta + uv_read_rtc()); +} + +/* + * Setup the RTC timer in oneshot mode + */ +static void uv_rtc_timer_setup(enum clock_event_mode mode, + struct clock_event_device *evt) +{ + int ced_cpu = cpumask_first(evt->cpumask); + + switch (mode) { + case CLOCK_EVT_MODE_PERIODIC: + case CLOCK_EVT_MODE_ONESHOT: + case CLOCK_EVT_MODE_RESUME: + /* Nothing to do here yet */ + break; + case CLOCK_EVT_MODE_UNUSED: + case CLOCK_EVT_MODE_SHUTDOWN: + uv_rtc_unset_timer(ced_cpu); + break; + } +} + +static void uv_rtc_interrupt(void) +{ + struct clock_event_device *ced = &__get_cpu_var(cpu_ced); + int cpu = smp_processor_id(); + + if (!ced || !ced->event_handler) + return; + + if (uv_rtc_unset_timer(cpu) != 1) + return; + + ced->event_handler(ced); +} + +static int __init uv_enable_rtc(char *str) +{ + uv_rtc_enable = 1; + + return 1; +} +__setup("uvrtc", uv_enable_rtc); + +static __init void uv_rtc_register_clockevents(struct work_struct *dummy) +{ + struct clock_event_device *ced = &__get_cpu_var(cpu_ced); + + *ced = clock_event_device_uv; + ced->cpumask = cpumask_of(smp_processor_id()); + clockevents_register_device(ced); +} + +static __init int uv_rtc_setup_clock(void) +{ + int rc; + + if (!uv_rtc_enable || !is_uv_system() || generic_interrupt_extension) + return -ENODEV; + + generic_interrupt_extension = uv_rtc_interrupt; + + clocksource_uv.mult = clocksource_hz2mult(sn_rtc_cycles_per_second, + clocksource_uv.shift); + + rc = clocksource_register(&clocksource_uv); + if (rc) { + generic_interrupt_extension = NULL; + return rc; + } + + /* Setup and register clockevents */ + rc = uv_rtc_allocate_timers(); + if (rc) { + clocksource_unregister(&clocksource_uv); + generic_interrupt_extension = NULL; + return rc; + } + + clock_event_device_uv.mult = div_sc(sn_rtc_cycles_per_second, + NSEC_PER_SEC, clock_event_device_uv.shift); + + clock_event_device_uv.min_delta_ns = NSEC_PER_SEC / + sn_rtc_cycles_per_second; + + clock_event_device_uv.max_delta_ns = clocksource_uv.mask * + (NSEC_PER_SEC / sn_rtc_cycles_per_second); + + rc = schedule_on_each_cpu(uv_rtc_register_clockevents); + if (rc) { + clocksource_unregister(&clocksource_uv); + generic_interrupt_extension = NULL; + uv_rtc_deallocate_timers(); + } + + return rc; +} +arch_initcall(uv_rtc_setup_clock); diff --git a/arch/x86/kernel/visws_quirks.c b/arch/x86/kernel/visws_quirks.c index 191a876e9e87..31ffc24eec4d 100644 --- a/arch/x86/kernel/visws_quirks.c +++ b/arch/x86/kernel/visws_quirks.c @@ -578,7 +578,7 @@ static struct irq_chip piix4_virtual_irq_type = { static irqreturn_t piix4_master_intr(int irq, void *dev_id) { int realirq; - irq_desc_t *desc; + struct irq_desc *desc; unsigned long flags; spin_lock_irqsave(&i8259A_lock, flags); diff --git a/arch/x86/kernel/vmi_32.c b/arch/x86/kernel/vmi_32.c index 2cc4a90e2cb3..95deb9f2211e 100644 --- a/arch/x86/kernel/vmi_32.c +++ b/arch/x86/kernel/vmi_32.c @@ -395,11 +395,6 @@ static void vmi_set_pte_atomic(pte_t *ptep, pte_t pteval) vmi_ops.update_pte(ptep, VMI_PAGE_PT); } -static void vmi_set_pte_present(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pte) -{ - vmi_ops.set_pte(pte, ptep, vmi_flags_addr_defer(mm, addr, VMI_PAGE_PT, 1)); -} - static void vmi_set_pud(pud_t *pudp, pud_t pudval) { /* Um, eww */ @@ -750,7 +745,6 @@ static inline int __init activate_vmi(void) pv_mmu_ops.set_pmd = vmi_set_pmd; #ifdef CONFIG_X86_PAE pv_mmu_ops.set_pte_atomic = vmi_set_pte_atomic; - pv_mmu_ops.set_pte_present = vmi_set_pte_present; pv_mmu_ops.set_pud = vmi_set_pud; pv_mmu_ops.pte_clear = vmi_pte_clear; pv_mmu_ops.pmd_clear = vmi_pmd_clear; diff --git a/arch/x86/kernel/vmlinux_32.lds.S b/arch/x86/kernel/vmlinux_32.lds.S index 0d860963f268..62ad500d55f3 100644 --- a/arch/x86/kernel/vmlinux_32.lds.S +++ b/arch/x86/kernel/vmlinux_32.lds.S @@ -189,15 +189,24 @@ SECTIONS *(.bss) . = ALIGN(4); __bss_stop = .; - _end = . ; - /* This is where the kernel creates the early boot page tables */ + } + + .brk : AT(ADDR(.brk) - LOAD_OFFSET) { . = ALIGN(PAGE_SIZE); - pg0 = . ; + __brk_base = . ; + . += 64 * 1024 ; /* 64k alignment slop space */ + *(.brk_reservation) /* areas brk users have reserved */ + __brk_limit = . ; + } + + .end : AT(ADDR(.end) - LOAD_OFFSET) { + _end = . ; } /* Sections to be discarded */ /DISCARD/ : { *(.exitcall.exit) + *(.discard) } STABS_DEBUG @@ -205,6 +214,12 @@ SECTIONS DWARF_DEBUG } +/* + * Build-time check on the image size: + */ +ASSERT((_end - LOAD_OFFSET <= KERNEL_IMAGE_SIZE), + "kernel image bigger than KERNEL_IMAGE_SIZE") + #ifdef CONFIG_KEXEC /* Link time checks */ #include <asm/kexec.h> diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S index fbfced6f6800..c8742507b030 100644 --- a/arch/x86/kernel/vmlinux_64.lds.S +++ b/arch/x86/kernel/vmlinux_64.lds.S @@ -29,8 +29,8 @@ SECTIONS { . = __START_KERNEL; phys_startup_64 = startup_64 - LOAD_OFFSET; - _text = .; /* Text and read-only data */ .text : AT(ADDR(.text) - LOAD_OFFSET) { + _text = .; /* Text and read-only data */ /* First the code that has to be first for bootstrapping */ *(.text.head) _stext = .; @@ -61,13 +61,13 @@ SECTIONS .data : AT(ADDR(.data) - LOAD_OFFSET) { DATA_DATA CONSTRUCTORS + _edata = .; /* End of data section */ } :data - _edata = .; /* End of data section */ - . = ALIGN(PAGE_SIZE); - . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { + . = ALIGN(PAGE_SIZE); + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); *(.data.cacheline_aligned) } . = ALIGN(CONFIG_X86_INTERNODE_CACHE_BYTES); @@ -125,29 +125,29 @@ SECTIONS #undef VVIRT_OFFSET #undef VVIRT - . = ALIGN(THREAD_SIZE); /* init_task */ .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { + . = ALIGN(THREAD_SIZE); /* init_task */ *(.data.init_task) }:data.init - . = ALIGN(PAGE_SIZE); .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { + . = ALIGN(PAGE_SIZE); *(.data.page_aligned) } - /* might get freed after init */ - . = ALIGN(PAGE_SIZE); - __smp_alt_begin = .; - __smp_locks = .; .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { + /* might get freed after init */ + . = ALIGN(PAGE_SIZE); + __smp_alt_begin = .; + __smp_locks = .; *(.smp_locks) + __smp_locks_end = .; + . = ALIGN(PAGE_SIZE); + __smp_alt_end = .; } - __smp_locks_end = .; - . = ALIGN(PAGE_SIZE); - __smp_alt_end = .; . = ALIGN(PAGE_SIZE); /* Init code and data */ - __init_begin = .; + __init_begin = .; /* paired with __init_end */ .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { _sinittext = .; INIT_TEXT @@ -159,40 +159,42 @@ SECTIONS __initdata_end = .; } - . = ALIGN(16); - __setup_start = .; - .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { *(.init.setup) } - __setup_end = .; - __initcall_start = .; + .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { + . = ALIGN(16); + __setup_start = .; + *(.init.setup) + __setup_end = .; + } .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) { + __initcall_start = .; INITCALLS + __initcall_end = .; } - __initcall_end = .; - __con_initcall_start = .; .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { + __con_initcall_start = .; *(.con_initcall.init) + __con_initcall_end = .; } - __con_initcall_end = .; - __x86_cpu_dev_start = .; .x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) { + __x86_cpu_dev_start = .; *(.x86_cpu_dev.init) + __x86_cpu_dev_end = .; } - __x86_cpu_dev_end = .; SECURITY_INIT . = ALIGN(8); .parainstructions : AT(ADDR(.parainstructions) - LOAD_OFFSET) { - __parainstructions = .; + __parainstructions = .; *(.parainstructions) - __parainstructions_end = .; + __parainstructions_end = .; } - . = ALIGN(8); - __alt_instructions = .; .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { + . = ALIGN(8); + __alt_instructions = .; *(.altinstructions) + __alt_instructions_end = .; } - __alt_instructions_end = .; .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) { *(.altinstr_replacement) } @@ -207,9 +209,11 @@ SECTIONS #ifdef CONFIG_BLK_DEV_INITRD . = ALIGN(PAGE_SIZE); - __initramfs_start = .; - .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { *(.init.ramfs) } - __initramfs_end = .; + .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { + __initramfs_start = .; + *(.init.ramfs) + __initramfs_end = .; + } #endif #ifdef CONFIG_SMP @@ -229,20 +233,29 @@ SECTIONS . = ALIGN(PAGE_SIZE); __init_end = .; - . = ALIGN(PAGE_SIZE); - __nosave_begin = .; .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { - *(.data.nosave) + . = ALIGN(PAGE_SIZE); + __nosave_begin = .; + *(.data.nosave) + . = ALIGN(PAGE_SIZE); + __nosave_end = .; } :data.init2 /* use another section data.init2, see PERCPU_VADDR() above */ - . = ALIGN(PAGE_SIZE); - __nosave_end = .; - __bss_start = .; /* BSS */ .bss : AT(ADDR(.bss) - LOAD_OFFSET) { + . = ALIGN(PAGE_SIZE); + __bss_start = .; /* BSS */ *(.bss.page_aligned) *(.bss) - } - __bss_stop = .; + __bss_stop = .; + } + + .brk : AT(ADDR(.brk) - LOAD_OFFSET) { + . = ALIGN(PAGE_SIZE); + __brk_base = . ; + . += 64 * 1024 ; /* 64k alignment slop space */ + *(.brk_reservation) /* areas brk users have reserved */ + __brk_limit = . ; + } _end = . ; @@ -250,6 +263,7 @@ SECTIONS /DISCARD/ : { *(.exitcall.exit) *(.eh_frame) + *(.discard) } STABS_DEBUG @@ -275,3 +289,10 @@ ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), ASSERT((per_cpu__irq_stack_union == 0), "irq_stack_union is not at start of per-cpu area"); #endif + +#ifdef CONFIG_KEXEC +#include <asm/kexec.h> + +ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, + "kexec control code size is too big") +#endif diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index 74de562812cc..a1d804bcd483 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -22,7 +22,7 @@ #include <asm/paravirt.h> #include <asm/setup.h> -#ifdef CONFIG_PARAVIRT +#if defined CONFIG_PCI && defined CONFIG_PARAVIRT /* * Interrupt control on vSMPowered systems: * ~AC is a shadow of IF. If IF is 'on' AC should be 'off' @@ -114,6 +114,7 @@ static void __init set_vsmp_pv_ops(void) } #endif +#ifdef CONFIG_PCI static int is_vsmp = -1; static void __init detect_vsmp_box(void) @@ -139,6 +140,15 @@ int is_vsmp_box(void) } } +#else +static void __init detect_vsmp_box(void) +{ +} +int is_vsmp_box(void) +{ + return 0; +} +#endif void __init vsmp_init(void) { detect_vsmp_box(); |