diff options
Diffstat (limited to 'arch/x86')
32 files changed, 729 insertions, 234 deletions
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index dfabfef..299fbc8 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -347,6 +347,7 @@ endif config X86_VSMP bool "ScaleMP vSMP" + select PARAVIRT_GUEST select PARAVIRT depends on X86_64 && PCI depends on X86_EXTENDED_PLATFORM diff --git a/arch/x86/include/asm/highmem.h b/arch/x86/include/asm/highmem.h index 8caac76..3bd0402 100644 --- a/arch/x86/include/asm/highmem.h +++ b/arch/x86/include/asm/highmem.h @@ -59,11 +59,12 @@ extern void kunmap_high(struct page *page); void *kmap(struct page *page); void kunmap(struct page *page); -void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot); -void *kmap_atomic(struct page *page, enum km_type type); -void kunmap_atomic_notypecheck(void *kvaddr, enum km_type type); -void *kmap_atomic_pfn(unsigned long pfn, enum km_type type); -void *kmap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot); + +void *kmap_atomic_prot(struct page *page, pgprot_t prot); +void *__kmap_atomic(struct page *page); +void __kunmap_atomic(void *kvaddr); +void *kmap_atomic_pfn(unsigned long pfn); +void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot); struct page *kmap_atomic_to_page(void *ptr); #define flush_cache_kmaps() do { } while (0) diff --git a/arch/x86/include/asm/iomap.h b/arch/x86/include/asm/iomap.h index c4191b3..363e33e 100644 --- a/arch/x86/include/asm/iomap.h +++ b/arch/x86/include/asm/iomap.h @@ -27,10 +27,10 @@ #include <asm/tlbflush.h> void __iomem * -iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot); +iomap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot); void -iounmap_atomic(void __iomem *kvaddr, enum km_type type); +iounmap_atomic(void __iomem *kvaddr); int iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot); diff --git a/arch/x86/include/asm/irq.h b/arch/x86/include/asm/irq.h index 0bf5b00..13b0eba 100644 --- a/arch/x86/include/asm/irq.h +++ b/arch/x86/include/asm/irq.h @@ -21,10 +21,8 @@ static inline int irq_canonicalize(int irq) #ifdef CONFIG_X86_32 extern void irq_ctx_init(int cpu); -extern void irq_ctx_exit(int cpu); #else # define irq_ctx_init(cpu) do { } while (0) -# define irq_ctx_exit(cpu) do { } while (0) #endif #define __ARCH_HAS_DO_SOFTIRQ diff --git a/arch/x86/include/asm/pgtable_32.h b/arch/x86/include/asm/pgtable_32.h index 8abde9e..0c92113 100644 --- a/arch/x86/include/asm/pgtable_32.h +++ b/arch/x86/include/asm/pgtable_32.h @@ -49,24 +49,14 @@ extern void set_pmd_pfn(unsigned long, unsigned long, pgprot_t); #endif #if defined(CONFIG_HIGHPTE) -#define __KM_PTE \ - (in_nmi() ? KM_NMI_PTE : \ - in_irq() ? KM_IRQ_PTE : \ - KM_PTE0) #define pte_offset_map(dir, address) \ - ((pte_t *)kmap_atomic(pmd_page(*(dir)), __KM_PTE) + \ + ((pte_t *)kmap_atomic(pmd_page(*(dir))) + \ pte_index((address))) -#define pte_offset_map_nested(dir, address) \ - ((pte_t *)kmap_atomic(pmd_page(*(dir)), KM_PTE1) + \ - pte_index((address))) -#define pte_unmap(pte) kunmap_atomic((pte), __KM_PTE) -#define pte_unmap_nested(pte) kunmap_atomic((pte), KM_PTE1) +#define pte_unmap(pte) kunmap_atomic((pte)) #else #define pte_offset_map(dir, address) \ ((pte_t *)page_address(pmd_page(*(dir))) + pte_index((address))) -#define pte_offset_map_nested(dir, address) pte_offset_map((dir), (address)) #define pte_unmap(pte) do { } while (0) -#define pte_unmap_nested(pte) do { } while (0) #endif /* Clear a kernel PTE and flush it from the TLB */ diff --git a/arch/x86/include/asm/pgtable_64.h b/arch/x86/include/asm/pgtable_64.h index f96ac9b..f86da20 100644 --- a/arch/x86/include/asm/pgtable_64.h +++ b/arch/x86/include/asm/pgtable_64.h @@ -127,9 +127,7 @@ static inline int pgd_large(pgd_t pgd) { return 0; } /* x86-64 always has all page tables mapped. */ #define pte_offset_map(dir, address) pte_offset_kernel((dir), (address)) -#define pte_offset_map_nested(dir, address) pte_offset_kernel((dir), (address)) #define pte_unmap(pte) ((void)(pte))/* NOP */ -#define pte_unmap_nested(pte) ((void)(pte)) /* NOP */ #define update_mmu_cache(vma, address, ptep) do { } while (0) diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h index 4cfc908..4c2f63c 100644 --- a/arch/x86/include/asm/smp.h +++ b/arch/x86/include/asm/smp.h @@ -50,7 +50,7 @@ struct smp_ops { void (*smp_prepare_cpus)(unsigned max_cpus); void (*smp_cpus_done)(unsigned max_cpus); - void (*smp_send_stop)(void); + void (*stop_other_cpus)(int wait); void (*smp_send_reschedule)(int cpu); int (*cpu_up)(unsigned cpu); @@ -73,7 +73,12 @@ extern struct smp_ops smp_ops; static inline void smp_send_stop(void) { - smp_ops.smp_send_stop(); + smp_ops.stop_other_cpus(0); +} + +static inline void stop_other_cpus(void) +{ + smp_ops.stop_other_cpus(1); } static inline void smp_prepare_boot_cpu(void) diff --git a/arch/x86/include/asm/xen/hypercall.h b/arch/x86/include/asm/xen/hypercall.h index 7fda040..a3c28ae 100644 --- a/arch/x86/include/asm/xen/hypercall.h +++ b/arch/x86/include/asm/xen/hypercall.h @@ -200,6 +200,23 @@ extern struct { char _entry[32]; } hypercall_page[]; (type)__res; \ }) +static inline long +privcmd_call(unsigned call, + unsigned long a1, unsigned long a2, + unsigned long a3, unsigned long a4, + unsigned long a5) +{ + __HYPERCALL_DECLS; + __HYPERCALL_5ARG(a1, a2, a3, a4, a5); + + asm volatile("call *%[call]" + : __HYPERCALL_5PARAM + : [call] "a" (&hypercall_page[call]) + : __HYPERCALL_CLOBBER5); + + return (long)__res; +} + static inline int HYPERVISOR_set_trap_table(struct trap_info *table) { diff --git a/arch/x86/include/asm/xen/page.h b/arch/x86/include/asm/xen/page.h index bf5f7d3..dd8c141 100644 --- a/arch/x86/include/asm/xen/page.h +++ b/arch/x86/include/asm/xen/page.h @@ -37,14 +37,21 @@ typedef struct xpaddr { extern unsigned long get_phys_to_machine(unsigned long pfn); -extern void set_phys_to_machine(unsigned long pfn, unsigned long mfn); +extern bool set_phys_to_machine(unsigned long pfn, unsigned long mfn); static inline unsigned long pfn_to_mfn(unsigned long pfn) { + unsigned long mfn; + if (xen_feature(XENFEAT_auto_translated_physmap)) return pfn; - return get_phys_to_machine(pfn) & ~FOREIGN_FRAME_BIT; + mfn = get_phys_to_machine(pfn); + + if (mfn != INVALID_P2M_ENTRY) + mfn &= ~FOREIGN_FRAME_BIT; + + return mfn; } static inline int phys_to_machine_mapping_valid(unsigned long pfn) @@ -159,6 +166,7 @@ static inline pte_t __pte_ma(pteval_t x) #define pgd_val_ma(x) ((x).pgd) +void xen_set_domain_pte(pte_t *ptep, pte_t pteval, unsigned domid); xmaddr_t arbitrary_virt_to_machine(void *address); unsigned long arbitrary_virt_to_mfn(void *vaddr); diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index cd8da24..a2baafb 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -701,6 +701,7 @@ static int acpi_cpufreq_cpu_exit(struct cpufreq_policy *policy) per_cpu(acfreq_data, policy->cpu) = NULL; acpi_processor_unregister_performance(data->acpi_data, policy->cpu); + kfree(data->freq_table); kfree(data); } diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 12cd823..17ad033 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -327,6 +327,7 @@ static void __cpuinit amd_calc_l3_indices(struct amd_l3_cache *l3) l3->subcaches[3] = sc3 = !(val & BIT(12)) + !(val & BIT(13)); l3->indices = (max(max(max(sc0, sc1), sc2), sc3) << 10) - 1; + l3->indices = (max(max3(sc0, sc1, sc2), sc3) << 10) - 1; } static struct amd_l3_cache * __cpuinit amd_init_l3_cache(int node) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index fe73c18..c1e8c7a 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -49,7 +49,6 @@ static unsigned long copy_from_user_nmi(void *to, const void __user *from, unsigned long n) { unsigned long offset, addr = (unsigned long)from; - int type = in_nmi() ? KM_NMI : KM_IRQ0; unsigned long size, len = 0; struct page *page; void *map; @@ -63,9 +62,9 @@ copy_from_user_nmi(void *to, const void __user *from, unsigned long n) offset = addr & (PAGE_SIZE - 1); size = min(PAGE_SIZE - offset, n - len); - map = kmap_atomic(page, type); + map = kmap_atomic(page); memcpy(to, map+offset, size); - kunmap_atomic(map, type); + kunmap_atomic(map); put_page(page); len += size; diff --git a/arch/x86/kernel/crash_dump_32.c b/arch/x86/kernel/crash_dump_32.c index 6741455..d5cd139 100644 --- a/arch/x86/kernel/crash_dump_32.c +++ b/arch/x86/kernel/crash_dump_32.c @@ -61,7 +61,7 @@ ssize_t copy_oldmem_page(unsigned long pfn, char *buf, if (!is_crashed_pfn_valid(pfn)) return -EFAULT; - vaddr = kmap_atomic_pfn(pfn, KM_PTE0); + vaddr = kmap_atomic_pfn(pfn); if (!userbuf) { memcpy(buf, (vaddr + offset), csize); diff --git a/arch/x86/kernel/dumpstack_32.c b/arch/x86/kernel/dumpstack_32.c index 0f6376f..1bc7f75 100644 --- a/arch/x86/kernel/dumpstack_32.c +++ b/arch/x86/kernel/dumpstack_32.c @@ -82,11 +82,11 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, if (kstack_end(stack)) break; if (i && ((i % STACKSLOTS_PER_LINE) == 0)) - printk("\n%s", log_lvl); - printk(" %08lx", *stack++); + printk(KERN_CONT "\n"); + printk(KERN_CONT " %08lx", *stack++); touch_nmi_watchdog(); } - printk("\n"); + printk(KERN_CONT "\n"); show_trace_log_lvl(task, regs, sp, bp, log_lvl); } diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c index 57a21f1..6a34048 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -265,20 +265,20 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, if (stack >= irq_stack && stack <= irq_stack_end) { if (stack == irq_stack_end) { stack = (unsigned long *) (irq_stack_end[-1]); - printk(" <EOI> "); + printk(KERN_CONT " <EOI> "); } } else { if (((long) stack & (THREAD_SIZE-1)) == 0) break; } if (i && ((i % STACKSLOTS_PER_LINE) == 0)) - printk("\n%s", log_lvl); - printk(" %016lx", *stack++); + printk(KERN_CONT "\n"); + printk(KERN_CONT " %016lx", *stack++); touch_nmi_watchdog(); } preempt_enable(); - printk("\n"); + printk(KERN_CONT "\n"); show_trace_log_lvl(task, regs, sp, bp, log_lvl); } diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index aff0b3c..ae03cab 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -713,7 +713,7 @@ static int hpet_cpuhp_notify(struct notifier_block *n, switch (action & 0xf) { case CPU_ONLINE: - INIT_DELAYED_WORK_ON_STACK(&work.work, hpet_work); + INIT_DELAYED_WORK_ONSTACK(&work.work, hpet_work); init_completion(&work.complete); /* FIXME: add schedule_work_on() */ schedule_delayed_work_on(cpu, &work.work, 0); diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index 50fbbe6..64668db 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -60,9 +60,6 @@ union irq_ctx { static DEFINE_PER_CPU(union irq_ctx *, hardirq_ctx); static DEFINE_PER_CPU(union irq_ctx *, softirq_ctx); -static DEFINE_PER_CPU_MULTIPAGE_ALIGNED(union irq_ctx, hardirq_stack, THREAD_SIZE); -static DEFINE_PER_CPU_MULTIPAGE_ALIGNED(union irq_ctx, softirq_stack, THREAD_SIZE); - static void call_on_stack(void *func, void *stack) { asm volatile("xchgl %%ebx,%%esp \n" @@ -128,7 +125,7 @@ void __cpuinit irq_ctx_init(int cpu) if (per_cpu(hardirq_ctx, cpu)) return; - irqctx = &per_cpu(hardirq_stack, cpu); + irqctx = (union irq_ctx *)__get_free_pages(THREAD_FLAGS, THREAD_ORDER); irqctx->tinfo.task = NULL; irqctx->tinfo.exec_domain = NULL; irqctx->tinfo.cpu = cpu; @@ -137,7 +134,7 @@ void __cpuinit irq_ctx_init(int cpu) per_cpu(hardirq_ctx, cpu) = irqctx; - irqctx = &per_cpu(softirq_stack, cpu); + irqctx = (union irq_ctx *)__get_free_pages(THREAD_FLAGS, THREAD_ORDER); irqctx->tinfo.task = NULL; irqctx->tinfo.exec_domain = NULL; irqctx->tinfo.cpu = cpu; @@ -150,11 +147,6 @@ void __cpuinit irq_ctx_init(int cpu) cpu, per_cpu(hardirq_ctx, cpu), per_cpu(softirq_ctx, cpu)); } -void irq_ctx_exit(int cpu) -{ - per_cpu(hardirq_ctx, cpu) = NULL; -} - asmlinkage void do_softirq(void) { unsigned long flags; diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 70c4872..45892dc 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -801,7 +801,8 @@ void ptrace_disable(struct task_struct *child) static const struct user_regset_view user_x86_32_view; /* Initialized below. */ #endif -long arch_ptrace(struct task_struct *child, long request, long addr, long data) +long arch_ptrace(struct task_struct *child, long request, + unsigned long addr, unsigned long data) { int ret; unsigned long __user *datap = (unsigned long __user *)data; @@ -812,8 +813,7 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data) unsigned long tmp; ret = -EIO; - if ((addr & (sizeof(data) - 1)) || addr < 0 || - addr >= sizeof(struct user)) + if ((addr & (sizeof(data) - 1)) || addr >= sizeof(struct user)) break; tmp = 0; /* Default return condition */ @@ -830,8 +830,7 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data) case PTRACE_POKEUSR: /* write the word at location addr in the USER area */ ret = -EIO; - if ((addr & (sizeof(data) - 1)) || addr < 0 || - addr >= sizeof(struct user)) + if ((addr & (sizeof(data) - 1)) || addr >= sizeof(struct user)) break; if (addr < sizeof(struct user_regs_struct)) @@ -888,17 +887,17 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data) #if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION case PTRACE_GET_THREAD_AREA: - if (addr < 0) + if ((int) addr < 0) return -EIO; ret = do_get_thread_area(child, addr, - (struct user_desc __user *) data); + (struct user_desc __user *)data); break; case PTRACE_SET_THREAD_AREA: - if (addr < 0) + if ((int) addr < 0) return -EIO; ret = do_set_thread_area(child, addr, - (struct user_desc __user *) data, 0); + (struct user_desc __user *)data, 0); break; #endif diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index f7f53dc..c495aa8d 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -635,7 +635,7 @@ void native_machine_shutdown(void) /* O.K Now that I'm on the appropriate processor, * stop all of the others. */ - smp_send_stop(); + stop_other_cpus(); #endif lapic_shutdown(); diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c index d801210..513deac 100644 --- a/arch/x86/kernel/smp.c +++ b/arch/x86/kernel/smp.c @@ -159,10 +159,10 @@ asmlinkage void smp_reboot_interrupt(void) irq_exit(); } -static void native_smp_send_stop(void) +static void native_stop_other_cpus(int wait) { unsigned long flags; - unsigned long wait; + unsigned long timeout; if (reboot_force) return; @@ -179,9 +179,12 @@ static void native_smp_send_stop(void) if (num_online_cpus() > 1) { apic->send_IPI_allbutself(REBOOT_VECTOR); - /* Don't wait longer than a second */ - wait = USEC_PER_SEC; - while (num_online_cpus() > 1 && wait--) + /* + * Don't wait longer than a second if the caller + * didn't ask us to wait. + */ + timeout = USEC_PER_SEC; + while (num_online_cpus() > 1 && (wait || timeout--)) udelay(1); } @@ -227,7 +230,7 @@ struct smp_ops smp_ops = { .smp_prepare_cpus = native_smp_prepare_cpus, .smp_cpus_done = native_smp_cpus_done, - .smp_send_stop = native_smp_send_stop, + .stop_other_cpus = native_stop_other_cpus, .smp_send_reschedule = native_smp_send_reschedule, .cpu_up = native_cpu_up, diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 6af1185..083e99d 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -747,7 +747,7 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) .done = COMPLETION_INITIALIZER_ONSTACK(c_idle.done), }; - INIT_WORK_ON_STACK(&c_idle.work, do_fork_idle); + INIT_WORK_ONSTACK(&c_idle.work, do_fork_idle); alternatives_smp_switch(1); @@ -1373,7 +1373,6 @@ void play_dead_common(void) { idle_task_exit(); reset_lazy_tlbstate(); - irq_ctx_exit(raw_smp_processor_id()); c1e_remove_cpu(raw_smp_processor_id()); mb(); diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index 852b319..7d90ceb 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -919,9 +919,9 @@ spurious_fault(unsigned long error_code, unsigned long address) int show_unhandled_signals = 1; static inline int -access_error(unsigned long error_code, int write, struct vm_area_struct *vma) +access_error(unsigned long error_code, struct vm_area_struct *vma) { - if (write) { + if (error_code & PF_WRITE) { /* write, present and write, not present: */ if (unlikely(!(vma->vm_flags & VM_WRITE))) return 1; @@ -956,8 +956,10 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code) struct task_struct *tsk; unsigned long address; struct mm_struct *mm; - int write; int fault; + int write = error_code & PF_WRITE; + unsigned int flags = FAULT_FLAG_ALLOW_RETRY | + (write ? FAULT_FLAG_WRITE : 0); tsk = current; mm = tsk->mm; @@ -1068,6 +1070,7 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code) bad_area_nosemaphore(regs, error_code, address); return; } +retry: down_read(&mm->mmap_sem); } else { /* @@ -1111,9 +1114,7 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code) * we can handle it.. */ good_area: - write = error_code & PF_WRITE; - - if (unlikely(access_error(error_code, write, vma))) { + if (unlikely(access_error(error_code, vma))) { bad_area_access_error(regs, error_code, address); return; } @@ -1123,21 +1124,34 @@ good_area: * make sure we exit gracefully rather than endlessly redo * the fault: */ - fault = handle_mm_fault(mm, vma, address, write ? FAULT_FLAG_WRITE : 0); + fault = handle_mm_fault(mm, vma, address, flags); if (unlikely(fault & VM_FAULT_ERROR)) { mm_fault_error(regs, error_code, address, fault); return; } - if (fault & VM_FAULT_MAJOR) { - tsk->maj_flt++; - perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0, - regs, address); - } else { - tsk->min_flt++; - perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0, - regs, address); + /* + * Major/minor page fault accounting is only done on the + * initial attempt. If we go through a retry, it is extremely + * likely that the page will be found in page cache at that point. + */ + if (flags & FAULT_FLAG_ALLOW_RETRY) { + if (fault & VM_FAULT_MAJOR) { + tsk->maj_flt++; + perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0, + regs, address); + } else { + tsk->min_flt++; + perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0, + regs, address); + } + if (fault & VM_FAULT_RETRY) { + /* Clear FAULT_FLAG_ALLOW_RETRY to avoid any risk + * of starvation. */ + flags &= ~FAULT_FLAG_ALLOW_RETRY; + goto retry; + } } check_v8086_mode(regs, address, tsk); diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c index 5e8fa12..b499626 100644 --- a/arch/x86/mm/highmem_32.c +++ b/arch/x86/mm/highmem_32.c @@ -9,6 +9,7 @@ void *kmap(struct page *page) return page_address(page); return kmap_high(page); } +EXPORT_SYMBOL(kmap); void kunmap(struct page *page) { @@ -18,6 +19,7 @@ void kunmap(struct page *page) return; kunmap_high(page); } +EXPORT_SYMBOL(kunmap); /* * kmap_atomic/kunmap_atomic is significantly faster than kmap/kunmap because @@ -27,10 +29,10 @@ void kunmap(struct page *page) * However when holding an atomic kmap it is not legal to sleep, so atomic * kmaps are appropriate for short, tight code paths only. */ -void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot) +void *kmap_atomic_prot(struct page *page, pgprot_t prot) { - enum fixed_addresses idx; unsigned long vaddr; + int idx, type; /* even !CONFIG_PREEMPT needs this, for in_atomic in do_page_fault */ pagefault_disable(); @@ -38,8 +40,7 @@ void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot) if (!PageHighMem(page)) return page_address(page); - debug_kmap_atomic(type); - + type = kmap_atomic_idx_push(); idx = type + KM_TYPE_NR*smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); BUG_ON(!pte_none(*(kmap_pte-idx))); @@ -47,44 +48,57 @@ void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot) return (void *)vaddr; } +EXPORT_SYMBOL(kmap_atomic_prot); + +void *__kmap_atomic(struct page *page) +{ + return kmap_atomic_prot(page, kmap_prot); +} +EXPORT_SYMBOL(__kmap_atomic); -void *kmap_atomic(struct page *page, enum km_type type) +/* + * This is the same as kmap_atomic() but can map memory that doesn't + * have a struct page associated with it. + */ +void *kmap_atomic_pfn(unsigned long pfn) { - return kmap_atomic_prot(page, type, kmap_prot); + return kmap_atomic_prot_pfn(pfn, kmap_prot); } +EXPORT_SYMBOL_GPL(kmap_atomic_pfn); -void kunmap_atomic_notypecheck(void *kvaddr, enum km_type type) +void __kunmap_atomic(void *kvaddr) { unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK; - enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id(); - - /* - * Force other mappings to Oops if they'll try to access this pte - * without first remap it. Keeping stale mappings around is a bad idea - * also, in case the page changes cacheability attributes or becomes - * a protected page in a hypervisor. - */ - if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) + + if (vaddr >= __fix_to_virt(FIX_KMAP_END) && + vaddr <= __fix_to_virt(FIX_KMAP_BEGIN)) { + int idx, type; + + type = kmap_atomic_idx(); + idx = type + KM_TYPE_NR * smp_processor_id(); + +#ifdef CONFIG_DEBUG_HIGHMEM + WARN_ON_ONCE(vaddr != __fix_to_virt(FIX_KMAP_BEGIN + idx)); +#endif + /* + * Force other mappings to Oops if they'll try to access this + * pte without first remap it. Keeping stale mappings around + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ kpte_clear_flush(kmap_pte-idx, vaddr); - else { + kmap_atomic_idx_pop(); + } #ifdef CONFIG_DEBUG_HIGHMEM + else { BUG_ON(vaddr < PAGE_OFFSET); BUG_ON(vaddr >= (unsigned long)high_memory); -#endif } +#endif pagefault_enable(); } - -/* - * This is the same as kmap_atomic() but can map memory that doesn't - * have a struct page associated with it. - */ -void *kmap_atomic_pfn(unsigned long pfn, enum km_type type) -{ - return kmap_atomic_prot_pfn(pfn, type, kmap_prot); -} -EXPORT_SYMBOL_GPL(kmap_atomic_pfn); /* temporarily in use by i915 GEM until vmap */ +EXPORT_SYMBOL(__kunmap_atomic); struct page *kmap_atomic_to_page(void *ptr) { @@ -98,12 +112,6 @@ struct page *kmap_atomic_to_page(void *ptr) pte = kmap_pte - (idx - FIX_KMAP_BEGIN); return pte_page(*pte); } - -EXPORT_SYMBOL(kmap); -EXPORT_SYMBOL(kunmap); -EXPORT_SYMBOL(kmap_atomic); -EXPORT_SYMBOL(kunmap_atomic_notypecheck); -EXPORT_SYMBOL(kmap_atomic_prot); EXPORT_SYMBOL(kmap_atomic_to_page); void __init set_highmem_pages_init(void) diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index 8434620..71a5929 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -51,7 +51,6 @@ #include <asm/numa.h> #include <asm/cacheflush.h> #include <asm/init.h> -#include <linux/bootmem.h> static int __init parse_direct_gbpages_off(char *arg) { diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 72fc70c..7b179b49 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -48,21 +48,20 @@ int iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot) } EXPORT_SYMBOL_GPL(iomap_create_wc); -void -iomap_free(resource_size_t base, unsigned long size) +void iomap_free(resource_size_t base, unsigned long size) { io_free_memtype(base, base + size); } EXPORT_SYMBOL_GPL(iomap_free); -void *kmap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) +void *kmap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) { - enum fixed_addresses idx; unsigned long vaddr; + int idx, type; pagefault_disable(); - debug_kmap_atomic(type); + type = kmap_atomic_idx_push(); idx = type + KM_TYPE_NR * smp_processor_id(); vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); set_pte(kmap_pte - idx, pfn_pte(pfn, prot)); @@ -72,10 +71,10 @@ void *kmap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) } /* - * Map 'pfn' using fixed map 'type' and protections 'prot' + * Map 'pfn' using protections 'prot' */ void __iomem * -iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) +iomap_atomic_prot_pfn(unsigned long pfn, pgprot_t prot) { /* * For non-PAT systems, promote PAGE_KERNEL_WC to PAGE_KERNEL_UC_MINUS. @@ -86,24 +85,34 @@ iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot) if (!pat_enabled && pgprot_val(prot) == pgprot_val(PAGE_KERNEL_WC)) prot = PAGE_KERNEL_UC_MINUS; - return (void __force __iomem *) kmap_atomic_prot_pfn(pfn, type, prot); + return (void __force __iomem *) kmap_atomic_prot_pfn(pfn, prot); } EXPORT_SYMBOL_GPL(iomap_atomic_prot_pfn); void -iounmap_atomic(void __iomem *kvaddr, enum km_type type) +iounmap_atomic(void __iomem *kvaddr) { unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK; - enum fixed_addresses idx = type + KM_TYPE_NR*smp_processor_id(); - /* - * Force other mappings to Oops if they'll try to access this pte - * without first remap it. Keeping stale mappings around is a bad idea - * also, in case the page changes cacheability attributes or becomes - * a protected page in a hypervisor. - */ - if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) + if (vaddr >= __fix_to_virt(FIX_KMAP_END) && + vaddr <= __fix_to_virt(FIX_KMAP_BEGIN)) { + int idx, type; + + type = kmap_atomic_idx(); + idx = type + KM_TYPE_NR * smp_processor_id(); + +#ifdef CONFIG_DEBUG_HIGHMEM + WARN_ON_ONCE(vaddr != __fix_to_virt(FIX_KMAP_BEGIN + idx)); +#endif + /* + * Force other mappings to Oops if they'll try to access this + * pte without first remap it. Keeping stale mappings around + * is a bad idea also, in case the page changes cacheability + * attributes or becomes a protected page in a hypervisor. + */ kpte_clear_flush(kmap_pte-idx, vaddr); + kmap_atomic_idx_pop(); + } pagefault_enable(); } diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig index 68128a1..90a7f5a 100644 --- a/arch/x86/xen/Kconfig +++ b/arch/x86/xen/Kconfig @@ -19,15 +19,12 @@ config XEN_PVHVM depends on X86_LOCAL_APIC config XEN_MAX_DOMAIN_MEMORY - int "Maximum allowed size of a domain in gigabytes" - default 8 if X86_32 - default 32 if X86_64 + int + default 128 depends on XEN help - The pseudo-physical to machine address array is sized - according to the maximum possible memory size of a Xen - domain. This array uses 1 page per gigabyte, so there's no - need to be too stingy here. + This only affects the sizing of some bss arrays, the unused + portions of which are freed. config XEN_SAVE_RESTORE bool diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 63b83ce..70ddeae 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -59,7 +59,6 @@ #include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/reboot.h> -#include <asm/setup.h> #include <asm/stackprotector.h> #include <asm/hypervisor.h> @@ -136,9 +135,6 @@ static void xen_vcpu_setup(int cpu) info.mfn = arbitrary_virt_to_mfn(vcpup); info.offset = offset_in_page(vcpup); - printk(KERN_DEBUG "trying to map vcpu_info %d at %p, mfn %llx, offset %d\n", - cpu, vcpup, info.mfn, info.offset); - /* Check to see if the hypervisor will put the vcpu_info structure where we want it, which allows direct access via a percpu-variable. */ @@ -152,9 +148,6 @@ static void xen_vcpu_setup(int cpu) /* This cpu is using the registered vcpu info, even if later ones fail to. */ per_cpu(xen_vcpu, cpu) = vcpup; - - printk(KERN_DEBUG "cpu %d using vcpu_info at %p\n", - cpu, vcpup); } } @@ -836,6 +829,11 @@ static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high) Xen console noise. */ break; + case MSR_IA32_CR_PAT: + if (smp_processor_id() == 0) + xen_set_pat(((u64)high << 32) | low); + break; + default: ret = native_write_msr_safe(msr, low, high); } @@ -874,8 +872,6 @@ void xen_setup_vcpu_info_placement(void) /* xen_vcpu_setup managed to place the vcpu_info within the percpu area for all cpus, so make use of it */ if (have_vcpu_info_placement) { - printk(KERN_INFO "Xen: using vcpu_info placement\n"); - pv_irq_ops.save_fl = __PV_IS_CALLEE_SAVE(xen_save_fl_direct); pv_irq_ops.restore_fl = __PV_IS_CALLEE_SAVE(xen_restore_fl_direct); pv_irq_ops.irq_disable = __PV_IS_CALLEE_SAVE(xen_irq_disable_direct); @@ -1019,7 +1015,7 @@ static void xen_reboot(int reason) struct sched_shutdown r = { .reason = reason }; #ifdef CONFIG_SMP - smp_send_stop(); + stop_other_cpus(); #endif if (HYPERVISOR_sched_op(SCHEDOP_shutdown, &r)) @@ -1189,6 +1185,9 @@ asmlinkage void __init xen_start_kernel(void) xen_raw_console_write("mapping kernel into physical memory\n"); pgd = xen_setup_kernel_pagetable(pgd, xen_start_info->nr_pages); + /* Allocate and initialize top and mid mfn levels for p2m structure */ + xen_build_mfn_list_list(); + init_mm.pgd = pgd; /* keep using Xen gdt for now; no urgent need to change it */ diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index f72d18c..9631c90 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -57,6 +57,7 @@ #include <asm/linkage.h> #include <asm/page.h> #include <asm/init.h> +#include <asm/pat.h> #include <asm/xen/hypercall.h> #include <asm/xen/hypervisor.h> @@ -140,7 +141,8 @@ static inline void check_zero(void) * large enough to allocate page table pages to allocate the rest. * Each page can map 2MB. */ -static pte_t level1_ident_pgt[PTRS_PER_PTE * 4] __page_aligned_bss; +#define LEVEL1_IDENT_ENTRIES (PTRS_PER_PTE * 4) +static RESERVE_BRK_ARRAY(pte_t, level1_ident_pgt, LEVEL1_IDENT_ENTRIES); #ifdef CONFIG_X86_64 /* l3 pud for userspace vsyscall mapping */ @@ -171,49 +173,182 @@ DEFINE_PER_CPU(unsigned long, xen_current_cr3); /* actual vcpu cr3 */ */ #define USER_LIMIT ((STACK_TOP_MAX + PGDIR_SIZE - 1) & PGDIR_MASK) +/* + * Xen leaves the responsibility for maintaining p2m mappings to the + * guests themselves, but it must also access and update the p2m array + * during suspend/resume when all the pages are reallocated. + * + * The p2m table is logically a flat array, but we implement it as a + * three-level tree to allow the address space to be sparse. + * + * Xen + * | + * p2m_top p2m_top_mfn + * / \ / \ + * p2m_mid p2m_mid p2m_mid_mfn p2m_mid_mfn + * / \ / \ / / + * p2m p2m p2m p2m p2m p2m p2m ... + * + * The p2m_mid_mfn pages are mapped by p2m_top_mfn_p. + * + * The p2m_top and p2m_top_mfn levels are limited to 1 page, so the + * maximum representable pseudo-physical address space is: + * P2M_TOP_PER_PAGE * P2M_MID_PER_PAGE * P2M_PER_PAGE pages + * + * P2M_PER_PAGE depends on the architecture, as a mfn is always + * unsigned long (8 bytes on 64-bit, 4 bytes on 32), leading to + * 512 and 1024 entries respectively. + */ + +unsigned long xen_max_p2m_pfn __read_mostly; -#define P2M_ENTRIES_PER_PAGE (PAGE_SIZE / sizeof(unsigned long)) -#define TOP_ENTRIES (MAX_DOMAIN_PAGES / P2M_ENTRIES_PER_PAGE) +#define P2M_PER_PAGE (PAGE_SIZE / sizeof(unsigned long)) +#define P2M_MID_PER_PAGE (PAGE_SIZE / sizeof(unsigned long *)) +#define P2M_TOP_PER_PAGE (PAGE_SIZE / sizeof(unsigned long **)) -/* Placeholder for holes in the address space */ -static unsigned long p2m_missing[P2M_ENTRIES_PER_PAGE] __page_aligned_data = - { [ 0 ... P2M_ENTRIES_PER_PAGE-1 ] = ~0UL }; +#define MAX_P2M_PFN (P2M_TOP_PER_PAGE * P2M_MID_PER_PAGE * P2M_PER_PAGE) - /* Array of pointers to pages containing p2m entries */ -static unsigned long *p2m_top[TOP_ENTRIES] __page_aligned_data = - { [ 0 ... TOP_ENTRIES - 1] = &p2m_missing[0] }; +/* Placeholders for holes in the address space */ +static RESERVE_BRK_ARRAY(unsigned long, p2m_missing, P2M_PER_PAGE); +static RESERVE_BRK_ARRAY(unsigned long *, p2m_mid_missing, P2M_MID_PER_PAGE); +static RESERVE_BRK_ARRAY(unsigned long, p2m_mid_missing_mfn, P2M_MID_PER_PAGE); -/* Arrays of p2m arrays expressed in mfns used for save/restore */ -static unsigned long p2m_top_mfn[TOP_ENTRIES] __page_aligned_bss; +static RESERVE_BRK_ARRAY(unsigned long **, p2m_top, P2M_TOP_PER_PAGE); +static RESERVE_BRK_ARRAY(unsigned long, p2m_top_mfn, P2M_TOP_PER_PAGE); +static RESERVE_BRK_ARRAY(unsigned long *, p2m_top_mfn_p, P2M_TOP_PER_PAGE); -static unsigned long p2m_top_mfn_list[TOP_ENTRIES / P2M_ENTRIES_PER_PAGE] - __page_aligned_bss; +RESERVE_BRK(p2m_mid, PAGE_SIZE * (MAX_DOMAIN_PAGES / (P2M_PER_PAGE * P2M_MID_PER_PAGE))); +RESERVE_BRK(p2m_mid_mfn, PAGE_SIZE * (MAX_DOMAIN_PAGES / (P2M_PER_PAGE * P2M_MID_PER_PAGE))); static inline unsigned p2m_top_index(unsigned long pfn) { - BUG_ON(pfn >= MAX_DOMAIN_PAGES); - return pfn / P2M_ENTRIES_PER_PAGE; + BUG_ON(pfn >= MAX_P2M_PFN); + return pfn / (P2M_MID_PER_PAGE * P2M_PER_PAGE); +} + +static inline unsigned p2m_mid_index(unsigned long pfn) +{ + return (pfn / P2M_PER_PAGE) % P2M_MID_PER_PAGE; } static inline unsigned p2m_index(unsigned long pfn) { - return pfn % P2M_ENTRIES_PER_PAGE; + return pfn % P2M_PER_PAGE; +} + +static void p2m_top_init(unsigned long ***top) +{ + unsigned i; + + for (i = 0; i < P2M_TOP_PER_PAGE; i++) + top[i] = p2m_mid_missing; +} + +static void p2m_top_mfn_init(unsigned long *top) +{ + unsigned i; + + for (i = 0; i < P2M_TOP_PER_PAGE; i++) + top[i] = virt_to_mfn(p2m_mid_missing_mfn); +} + +static void p2m_top_mfn_p_init(unsigned long **top) +{ + unsigned i; + + for (i = 0; i < P2M_TOP_PER_PAGE; i++) + top[i] = p2m_mid_missing_mfn; +} + +static void p2m_mid_init(unsigned long **mid) +{ + unsigned i; + + for (i = 0; i < P2M_MID_PER_PAGE; i++) + mid[i] = p2m_missing; +} + +static void p2m_mid_mfn_init(unsigned long *mid) +{ + unsigned i; + + for (i = 0; i < P2M_MID_PER_PAGE; i++) + mid[i] = virt_to_mfn(p2m_missing); } -/* Build the parallel p2m_top_mfn structures */ +static void p2m_init(unsigned long *p2m) +{ + unsigned i; + + for (i = 0; i < P2M_MID_PER_PAGE; i++) + p2m[i] = INVALID_P2M_ENTRY; +} + +/* + * Build the parallel p2m_top_mfn and p2m_mid_mfn structures + * + * This is called both at boot time, and after resuming from suspend: + * - At boot time we're called very early, and must use extend_brk() + * to allocate memory. + * + * - After resume we're called from within stop_machine, but the mfn + * tree should alreay be completely allocated. + */ void xen_build_mfn_list_list(void) { - unsigned pfn, idx; + unsigned long pfn; - for (pfn = 0; pfn < MAX_DOMAIN_PAGES; pfn += P2M_ENTRIES_PER_PAGE) { - unsigned topidx = p2m_top_index(pfn); + /* Pre-initialize p2m_top_mfn to be completely missing */ + if (p2m_top_mfn == NULL) { + p2m_mid_missing_mfn = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_mid_mfn_init(p2m_mid_missing_mfn); + + p2m_top_mfn_p = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_top_mfn_p_init(p2m_top_mfn_p); - p2m_top_mfn[topidx] = virt_to_mfn(p2m_top[topidx]); + p2m_top_mfn = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_top_mfn_init(p2m_top_mfn); + } else { + /* Reinitialise, mfn's all change after migration */ + p2m_mid_mfn_init(p2m_mid_missing_mfn); } - for (idx = 0; idx < ARRAY_SIZE(p2m_top_mfn_list); idx++) { - unsigned topidx = idx * P2M_ENTRIES_PER_PAGE; - p2m_top_mfn_list[idx] = virt_to_mfn(&p2m_top_mfn[topidx]); + for (pfn = 0; pfn < xen_max_p2m_pfn; pfn += P2M_PER_PAGE) { + unsigned topidx = p2m_top_index(pfn); + unsigned mididx = p2m_mid_index(pfn); + unsigned long **mid; + unsigned long *mid_mfn_p; + + mid = p2m_top[topidx]; + mid_mfn_p = p2m_top_mfn_p[topidx]; + + /* Don't bother allocating any mfn mid levels if + * they're just missing, just update the stored mfn, + * since all could have changed over a migrate. + */ + if (mid == p2m_mid_missing) { + BUG_ON(mididx); + BUG_ON(mid_mfn_p != p2m_mid_missing_mfn); + p2m_top_mfn[topidx] = virt_to_mfn(p2m_mid_missing_mfn); + pfn += (P2M_MID_PER_PAGE - 1) * P2M_PER_PAGE; + continue; + } + + if (mid_mfn_p == p2m_mid_missing_mfn) { + /* + * XXX boot-time only! We should never find + * missing parts of the mfn tree after + * runtime. extend_brk() will BUG if we call + * it too late. + */ + mid_mfn_p = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_mid_mfn_init(mid_mfn_p); + + p2m_top_mfn_p[topidx] = mid_mfn_p; + } + + p2m_top_mfn[topidx] = virt_to_mfn(mid_mfn_p); + mid_mfn_p[mididx] = virt_to_mfn(mid[mididx]); } } @@ -222,8 +357,8 @@ void xen_setup_mfn_list_list(void) BUG_ON(HYPERVISOR_shared_info == &xen_dummy_shared_info); HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list = - virt_to_mfn(p2m_top_mfn_list); - HYPERVISOR_shared_info->arch.max_pfn = xen_start_info->nr_pages; + virt_to_mfn(p2m_top_mfn); + HYPERVISOR_shared_info->arch.max_pfn = xen_max_p2m_pfn; } /* Set up p2m_top to point to the domain-builder provided p2m pages */ @@ -231,98 +366,176 @@ void __init xen_build_dynamic_phys_to_machine(void) { unsigned long *mfn_list = (unsigned long *)xen_start_info->mfn_list; unsigned long max_pfn = min(MAX_DOMAIN_PAGES, xen_start_info->nr_pages); - unsigned pfn; + unsigned long pfn; + + xen_max_p2m_pfn = max_pfn; - for (pfn = 0; pfn < max_pfn; pfn += P2M_ENTRIES_PER_PAGE) { + p2m_missing = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_init(p2m_missing); + + p2m_mid_missing = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_mid_init(p2m_mid_missing); + + p2m_top = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_top_init(p2m_top); + + /* + * The domain builder gives us a pre-constructed p2m array in + * mfn_list for all the pages initially given to us, so we just + * need to graft that into our tree structure. + */ + for (pfn = 0; pfn < max_pfn; pfn += P2M_PER_PAGE) { unsigned topidx = p2m_top_index(pfn); + unsigned mididx = p2m_mid_index(pfn); - p2m_top[topidx] = &mfn_list[pfn]; - } + if (p2m_top[topidx] == p2m_mid_missing) { + unsigned long **mid = extend_brk(PAGE_SIZE, PAGE_SIZE); + p2m_mid_init(mid); + + p2m_top[topidx] = mid; + } - xen_build_mfn_list_list(); + p2m_top[topidx][mididx] = &mfn_list[pfn]; + } } unsigned long get_phys_to_machine(unsigned long pfn) { - unsigned topidx, idx; + unsigned topidx, mididx, idx; - if (unlikely(pfn >= MAX_DOMAIN_PAGES)) + if (unlikely(pfn >= MAX_P2M_PFN)) return INVALID_P2M_ENTRY; topidx = p2m_top_index(pfn); + mididx = p2m_mid_index(pfn); idx = p2m_index(pfn); - return p2m_top[topidx][idx]; + + return p2m_top[topidx][mididx][idx]; } EXPORT_SYMBOL_GPL(get_phys_to_machine); -/* install a new p2m_top page */ -bool install_p2mtop_page(unsigned long pfn, unsigned long *p) +static void *alloc_p2m_page(void) { - unsigned topidx = p2m_top_index(pfn); - unsigned long **pfnp, *mfnp; - unsigned i; + return (void *)__get_free_page(GFP_KERNEL | __GFP_REPEAT); +} - pfnp = &p2m_top[topidx]; - mfnp = &p2m_top_mfn[topidx]; +static void free_p2m_page(void *p) +{ + free_page((unsigned long)p); +} - for (i = 0; i < P2M_ENTRIES_PER_PAGE; i++) - p[i] = INVALID_P2M_ENTRY; +/* + * Fully allocate the p2m structure for a given pfn. We need to check + * that both the top and mid levels are allocated, and make sure the + * parallel mfn tree is kept in sync. We may race with other cpus, so + * the new pages are installed with cmpxchg; if we lose the race then + * simply free the page we allocated and use the one that's there. + */ +static bool alloc_p2m(unsigned long pfn) +{ + unsigned topidx, mididx; + unsigned long ***top_p, **mid; + unsigned long *top_mfn_p, *mid_mfn; - if (cmpxchg(pfnp, p2m_missing, p) == p2m_missing) { - *mfnp = virt_to_mfn(p); - return true; + topidx = p2m_top_index(pfn); + mididx = p2m_mid_index(pfn); + + top_p = &p2m_top[topidx]; + mid = *top_p; + + if (mid == p2m_mid_missing) { + /* Mid level is missing, allocate a new one */ + mid = alloc_p2m_page(); + if (!mid) + return false; + + p2m_mid_init(mid); + + if (cmpxchg(top_p, p2m_mid_missing, mid) != p2m_mid_missing) + free_p2m_page(mid); } - return false; -} + top_mfn_p = &p2m_top_mfn[topidx]; + mid_mfn = p2m_top_mfn_p[topidx]; -static void alloc_p2m(unsigned long pfn) -{ - unsigned long *p; + BUG_ON(virt_to_mfn(mid_mfn) != *top_mfn_p); + + if (mid_mfn == p2m_mid_missing_mfn) { + /* Separately check the mid mfn level */ + unsigned long missing_mfn; + unsigned long mid_mfn_mfn; + + mid_mfn = alloc_p2m_page(); + if (!mid_mfn) + return false; + + p2m_mid_mfn_init(mid_mfn); + + missing_mfn = virt_to_mfn(p2m_mid_missing_mfn); + mid_mfn_mfn = virt_to_mfn(mid_mfn); + if (cmpxchg(top_mfn_p, missing_mfn, mid_mfn_mfn) != missing_mfn) + free_p2m_page(mid_mfn); + else + p2m_top_mfn_p[topidx] = mid_mfn; + } + + if (p2m_top[topidx][mididx] == p2m_missing) { + /* p2m leaf page is missing */ + unsigned long *p2m; + + p2m = alloc_p2m_page(); + if (!p2m) + return false; - p = (void *)__get_free_page(GFP_KERNEL | __GFP_NOFAIL); - BUG_ON(p == NULL); + p2m_init(p2m); + + if (cmpxchg(&mid[mididx], p2m_missing, p2m) != p2m_missing) + free_p2m_page(p2m); + else + mid_mfn[mididx] = virt_to_mfn(p2m); + } - if (!install_p2mtop_page(pfn, p)) - free_page((unsigned long)p); + return true; } /* Try to install p2m mapping; fail if intermediate bits missing */ bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn) { - unsigned topidx, idx; + unsigned topidx, mididx, idx; - if (unlikely(pfn >= MAX_DOMAIN_PAGES)) { + if (unlikely(pfn >= MAX_P2M_PFN)) { BUG_ON(mfn != INVALID_P2M_ENTRY); return true; } topidx = p2m_top_index(pfn); - if (p2m_top[topidx] == p2m_missing) { - if (mfn == INVALID_P2M_ENTRY) - return true; - return false; - } - + mididx = p2m_mid_index(pfn); idx = p2m_index(pfn); - p2m_top[topidx][idx] = mfn; + + if (p2m_top[topidx][mididx] == p2m_missing) + return mfn == INVALID_P2M_ENTRY; + + p2m_top[topidx][mididx][idx] = mfn; return true; } -void set_phys_to_machine(unsigned long pfn, unsigned long mfn) +bool set_phys_to_machine(unsigned long pfn, unsigned long mfn) { if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) { BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY); - return; + return true; } if (unlikely(!__set_phys_to_machine(pfn, mfn))) { - alloc_p2m(pfn); + if (!alloc_p2m(pfn)) + return false; if (!__set_phys_to_machine(pfn, mfn)) - BUG(); + return false; } + + return true; } unsigned long arbitrary_virt_to_mfn(void *vaddr) @@ -399,7 +612,7 @@ static bool xen_iomap_pte(pte_t pte) return pte_flags(pte) & _PAGE_IOMAP; } -static void xen_set_iomap_pte(pte_t *ptep, pte_t pteval) +void xen_set_domain_pte(pte_t *ptep, pte_t pteval, unsigned domid) { struct multicall_space mcs; struct mmu_update *u; @@ -411,10 +624,16 @@ static void xen_set_iomap_pte(pte_t *ptep, pte_t pteval) u->ptr = arbitrary_virt_to_machine(ptep).maddr; u->val = pte_val_ma(pteval); - MULTI_mmu_update(mcs.mc, mcs.args, 1, NULL, DOMID_IO); + MULTI_mmu_update(mcs.mc, mcs.args, 1, NULL, domid); xen_mc_issue(PARAVIRT_LAZY_MMU); } +EXPORT_SYMBOL_GPL(xen_set_domain_pte); + +static void xen_set_iomap_pte(pte_t *ptep, pte_t pteval) +{ + xen_set_domain_pte(ptep, pteval, DOMID_IO); +} static void xen_extend_mmu_update(const struct mmu_update *update) { @@ -561,7 +780,20 @@ static pteval_t pte_pfn_to_mfn(pteval_t val) if (val & _PAGE_PRESENT) { unsigned long pfn = (val & PTE_PFN_MASK) >> PAGE_SHIFT; pteval_t flags = val & PTE_FLAGS_MASK; - val = ((pteval_t)pfn_to_mfn(pfn) << PAGE_SHIFT) | flags; + unsigned long mfn = pfn_to_mfn(pfn); + + /* + * If there's no mfn for the pfn, then just create an + * empty non-present pte. Unfortunately this loses + * information about the original pfn, so + * pte_mfn_to_pfn is asymmetric. + */ + if (unlikely(mfn == INVALID_P2M_ENTRY)) { + mfn = 0; + flags = 0; + } + + val = ((pteval_t)mfn << PAGE_SHIFT) | flags; } return val; @@ -583,10 +815,18 @@ static pteval_t iomap_pte(pteval_t val) pteval_t xen_pte_val(pte_t pte) { - if (xen_initial_domain() && (pte.pte & _PAGE_IOMAP)) - return pte.pte; + pteval_t pteval = pte.pte; + + /* If this is a WC pte, convert back from Xen WC to Linux WC */ + if ((pteval & (_PAGE_PAT | _PAGE_PCD | _PAGE_PWT)) == _PAGE_PAT) { + WARN_ON(!pat_enabled); + pteval = (pteval & ~_PAGE_PAT) | _PAGE_PWT; + } - return pte_mfn_to_pfn(pte.pte); + if (xen_initial_domain() && (pteval & _PAGE_IOMAP)) + return pteval; + + return pte_mfn_to_pfn(pteval); } PV_CALLEE_SAVE_REGS_THUNK(xen_pte_val); @@ -596,10 +836,48 @@ pgdval_t xen_pgd_val(pgd_t pgd) } PV_CALLEE_SAVE_REGS_THUNK(xen_pgd_val); +/* + * Xen's PAT setup is part of its ABI, though I assume entries 6 & 7 + * are reserved for now, to correspond to the Intel-reserved PAT + * types. + * + * We expect Linux's PAT set as follows: + * + * Idx PTE flags Linux Xen Default + * 0 WB WB WB + * 1 PWT WC WT WT + * 2 PCD UC- UC- UC- + * 3 PCD PWT UC UC UC + * 4 PAT WB WC WB + * 5 PAT PWT WC WP WT + * 6 PAT PCD UC- UC UC- + * 7 PAT PCD PWT UC UC UC + */ + +void xen_set_pat(u64 pat) +{ + /* We expect Linux to use a PAT setting of + * UC UC- WC WB (ignoring the PAT flag) */ + WARN_ON(pat != 0x0007010600070106ull); +} + pte_t xen_make_pte(pteval_t pte) { phys_addr_t addr = (pte & PTE_PFN_MASK); + /* If Linux is trying to set a WC pte, then map to the Xen WC. + * If _PAGE_PAT is set, then it probably means it is really + * _PAGE_PSE, so avoid fiddling with the PAT mapping and hope + * things work out OK... + * + * (We should never see kernel mappings with _PAGE_PSE set, + * but we could see hugetlbfs mappings, I think.). + */ + if (pat_enabled && !WARN_ON(pte & _PAGE_PAT)) { + if ((pte & (_PAGE_PCD | _PAGE_PWT)) == _PAGE_PWT) + pte = (pte & ~(_PAGE_PCD | _PAGE_PWT)) | _PAGE_PAT; + } + /* * Unprivileged domains are allowed to do IOMAPpings for * PCI passthrough, but not map ISA space. The ISA @@ -1712,6 +1990,9 @@ static __init void xen_map_identity_early(pmd_t *pmd, unsigned long max_pfn) unsigned ident_pte; unsigned long pfn; + level1_ident_pgt = extend_brk(sizeof(pte_t) * LEVEL1_IDENT_ENTRIES, + PAGE_SIZE); + ident_pte = 0; pfn = 0; for (pmdidx = 0; pmdidx < PTRS_PER_PMD && pfn < max_pfn; pmdidx++) { @@ -1722,7 +2003,7 @@ static __init void xen_map_identity_early(pmd_t *pmd, unsigned long max_pfn) pte_page = m2v(pmd[pmdidx].pmd); else { /* Check for free pte pages */ - if (ident_pte == ARRAY_SIZE(level1_ident_pgt)) + if (ident_pte == LEVEL1_IDENT_ENTRIES) break; pte_page = &level1_ident_pgt[ident_pte]; @@ -1837,13 +2118,15 @@ __init pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd, return pgd; } #else /* !CONFIG_X86_64 */ -static pmd_t level2_kernel_pgt[PTRS_PER_PMD] __page_aligned_bss; +static RESERVE_BRK_ARRAY(pmd_t, level2_kernel_pgt, PTRS_PER_PMD); __init pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn) { pmd_t *kernel_pmd; + level2_kernel_pgt = extend_brk(sizeof(pmd_t *) * PTRS_PER_PMD, PAGE_SIZE); + max_pfn_mapped = PFN_DOWN(__pa(xen_start_info->pt_base) + xen_start_info->nr_pt_frames * PAGE_SIZE + 512*1024); @@ -2269,6 +2552,72 @@ void __init xen_hvm_init_mmu_ops(void) } #endif +#define REMAP_BATCH_SIZE 16 + +struct remap_data { + unsigned long mfn; + pgprot_t prot; + struct mmu_update *mmu_update; +}; + +static int remap_area_mfn_pte_fn(pte_t *ptep, pgtable_t token, + unsigned long addr, void *data) +{ + struct remap_data *rmd = data; + pte_t pte = pte_mkspecial(pfn_pte(rmd->mfn++, rmd->prot)); + + rmd->mmu_update->ptr = arbitrary_virt_to_machine(ptep).maddr; + rmd->mmu_update->val = pte_val_ma(pte); + rmd->mmu_update++; + + return 0; +} + +int xen_remap_domain_mfn_range(struct vm_area_struct *vma, + unsigned long addr, + unsigned long mfn, int nr, + pgprot_t prot, unsigned domid) +{ + struct remap_data rmd; + struct mmu_update mmu_update[REMAP_BATCH_SIZE]; + int batch; + unsigned long range; + int err = 0; + + prot = __pgprot(pgprot_val(prot) | _PAGE_IOMAP); + + vma->vm_flags |= VM_IO | VM_RESERVED | VM_PFNMAP; + + rmd.mfn = mfn; + rmd.prot = prot; + + while (nr) { + batch = min(REMAP_BATCH_SIZE, nr); + range = (unsigned long)batch << PAGE_SHIFT; + + rmd.mmu_update = mmu_update; + err = apply_to_page_range(vma->vm_mm, addr, range, + remap_area_mfn_pte_fn, &rmd); + if (err) + goto out; + + err = -EFAULT; + if (HYPERVISOR_mmu_update(mmu_update, batch, NULL, domid) < 0) + goto out; + + nr -= batch; + addr += range; + } + + err = 0; +out: + + flush_tlb_all(); + + return err; +} +EXPORT_SYMBOL_GPL(xen_remap_domain_mfn_range); + #ifdef CONFIG_XEN_DEBUG_FS static struct dentry *d_mmu_debug; diff --git a/arch/x86/xen/mmu.h b/arch/x86/xen/mmu.h index fa938c4..537bb9a 100644 --- a/arch/x86/xen/mmu.h +++ b/arch/x86/xen/mmu.h @@ -12,7 +12,6 @@ enum pt_level { bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn); -bool install_p2mtop_page(unsigned long pfn, unsigned long *p); void set_pte_mfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags); diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index 9729c90..105db25 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -18,8 +18,10 @@ #include <asm/xen/hypervisor.h> #include <asm/xen/hypercall.h> +#include <xen/xen.h> #include <xen/page.h> #include <xen/interface/callback.h> +#include <xen/interface/memory.h> #include <xen/interface/physdev.h> #include <xen/interface/memory.h> #include <xen/features.h> @@ -34,6 +36,39 @@ extern void xen_sysenter_target(void); extern void xen_syscall_target(void); extern void xen_syscall32_target(void); +/* Amount of extra memory space we add to the e820 ranges */ +phys_addr_t xen_extra_mem_start, xen_extra_mem_size; + +/* + * The maximum amount of extra memory compared to the base size. The + * main scaling factor is the size of struct page. At extreme ratios + * of base:extra, all the base memory can be filled with page + * structures for the extra memory, leaving no space for anything + * else. + * + * 10x seems like a reasonable balance between scaling flexibility and + * leaving a practically usable system. + */ +#define EXTRA_MEM_RATIO (10) + +static __init void xen_add_extra_mem(unsigned long pages) +{ + u64 size = (u64)pages * PAGE_SIZE; + u64 extra_start = xen_extra_mem_start + xen_extra_mem_size; + + if (!pages) + return; + + e820_add_region(extra_start, size, E820_RAM); + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); + + memblock_x86_reserve_range(extra_start, extra_start + size, "XEN EXTRA"); + + xen_extra_mem_size += size; + + xen_max_p2m_pfn = PFN_DOWN(extra_start + size); +} + static unsigned long __init xen_release_chunk(phys_addr_t start_addr, phys_addr_t end_addr) { @@ -105,16 +140,65 @@ static unsigned long __init xen_return_unused_memory(unsigned long max_pfn, /** * machine_specific_memory_setup - Hook for machine specific memory setup. **/ - char * __init xen_memory_setup(void) { + static struct e820entry map[E820MAX] __initdata; + unsigned long max_pfn = xen_start_info->nr_pages; + unsigned long long mem_end; + int rc; + struct xen_memory_map memmap; + unsigned long extra_pages = 0; + unsigned long extra_limit; + int i; + int op; max_pfn = min(MAX_DOMAIN_PAGES, max_pfn); + mem_end = PFN_PHYS(max_pfn); + + memmap.nr_entries = E820MAX; + set_xen_guest_handle(memmap.buffer, map); + + op = xen_initial_domain() ? + XENMEM_machine_memory_map : + XENMEM_memory_map; + rc = HYPERVISOR_memory_op(op, &memmap); + if (rc == -ENOSYS) { + memmap.nr_entries = 1; + map[0].addr = 0ULL; + map[0].size = mem_end; + /* 8MB slack (to balance backend allocations). */ + map[0].size += 8ULL << 20; + map[0].type = E820_RAM; + rc = 0; + } + BUG_ON(rc); e820.nr_map = 0; + xen_extra_mem_start = mem_end; + for (i = 0; i < memmap.nr_entries; i++) { + unsigned long long end = map[i].addr + map[i].size; + + if (map[i].type == E820_RAM) { + if (map[i].addr < mem_end && end > mem_end) { + /* Truncate region to max_mem. */ + u64 delta = end - mem_end; + + map[i].size -= delta; + extra_pages += PFN_DOWN(delta); + + end = mem_end; + } + } - e820_add_region(0, PFN_PHYS((u64)max_pfn), E820_RAM); + if (end > xen_extra_mem_start) + xen_extra_mem_start = end; + + /* If region is non-RAM or below mem_end, add what remains */ + if ((map[i].type != E820_RAM || map[i].addr < mem_end) && + map[i].size > 0) + e820_add_region(map[i].addr, map[i].size, map[i].type); + } /* * Even though this is normal, usable memory under Xen, reserve @@ -136,7 +220,29 @@ char * __init xen_memory_setup(void) sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); - xen_return_unused_memory(xen_start_info->nr_pages, &e820); + extra_pages += xen_return_unused_memory(xen_start_info->nr_pages, &e820); + + /* + * Clamp the amount of extra memory to a EXTRA_MEM_RATIO + * factor the base size. On non-highmem systems, the base + * size is the full initial memory allocation; on highmem it + * is limited to the max size of lowmem, so that it doesn't + * get completely filled. + * + * In principle there could be a problem in lowmem systems if + * the initial memory is also very large with respect to + * lowmem, but we won't try to deal with that here. + */ + extra_limit = min(EXTRA_MEM_RATIO * min(max_pfn, PFN_DOWN(MAXMEM)), + max_pfn + extra_pages); + + if (extra_limit >= max_pfn) + extra_pages = extra_limit - max_pfn; + else + extra_pages = 0; + + if (!xen_initial_domain()) + xen_add_extra_mem(extra_pages); return "Xen"; } diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c index 25f232b..f4d0100 100644 --- a/arch/x86/xen/smp.c +++ b/arch/x86/xen/smp.c @@ -400,9 +400,9 @@ static void stop_self(void *v) BUG(); } -static void xen_smp_send_stop(void) +static void xen_stop_other_cpus(int wait) { - smp_call_function(stop_self, NULL, 0); + smp_call_function(stop_self, NULL, wait); } static void xen_smp_send_reschedule(int cpu) @@ -470,7 +470,7 @@ static const struct smp_ops xen_smp_ops __initdata = { .cpu_disable = xen_cpu_disable, .play_dead = xen_play_dead, - .smp_send_stop = xen_smp_send_stop, + .stop_other_cpus = xen_stop_other_cpus, .smp_send_reschedule = xen_smp_send_reschedule, .send_call_func_ipi = xen_smp_send_call_function_ipi, diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h index 7c8ab86..6404474 100644 --- a/arch/x86/xen/xen-ops.h +++ b/arch/x86/xen/xen-ops.h @@ -30,6 +30,9 @@ void xen_setup_machphys_mapping(void); pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn); void xen_ident_map_ISA(void); void xen_reserve_top(void); +extern unsigned long xen_max_p2m_pfn; + +void xen_set_pat(u64); char * __init xen_memory_setup(void); void __init xen_arch_setup(void); |