Merge commit 'v2.6.26-rc9' into cpus4096

6 files changed, 80 insertions, 231 deletions

diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig
index 2e641be2737e..6c388e593bc8 100644
--- a/arch/x86/xen/Kconfig
+++ b/arch/x86/xen/Kconfig
@@ -5,8 +5,9 @@
 config XEN
         bool "Xen guest support"
         select PARAVIRT
+        select PARAVIRT_CLOCK
         depends on X86_32
-        depends on X86_CMPXCHG && X86_TSC && !(X86_VISWS || X86_VOYAGER)
+        depends on X86_CMPXCHG && X86_TSC && X86_PAE && !(X86_VISWS || X86_VOYAGER)
         help
           This is the Linux Xen port.  Enabling this will allow the
           kernel to boot in a paravirtualized environment under the
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index c8a56e457d61..f09c1c69c37a 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -785,38 +785,35 @@ static __init void xen_set_pte_init(pte_t *ptep, pte_t pte)
 static __init void xen_pagetable_setup_start(pgd_t *base)
 {
         pgd_t *xen_pgd = (pgd_t *)xen_start_info->pt_base;
+        int i;
 
         /* special set_pte for pagetable initialization */
         pv_mmu_ops.set_pte = xen_set_pte_init;
 
         init_mm.pgd = base;
         /*
-         * copy top-level of Xen-supplied pagetable into place.  For
-         * !PAE we can use this as-is, but for PAE it is a stand-in
-         * while we copy the pmd pages.
+         * copy top-level of Xen-supplied pagetable into place.  This
+         * is a stand-in while we copy the pmd pages.
          */
         memcpy(base, xen_pgd, PTRS_PER_PGD * sizeof(pgd_t));
 
-        if (PTRS_PER_PMD > 1) {
-                int i;
-                /*
-                 * For PAE, need to allocate new pmds, rather than
-                 * share Xen's, since Xen doesn't like pmd's being
-                 * shared between address spaces.
-                 */
-                for (i = 0; i < PTRS_PER_PGD; i++) {
-                        if (pgd_val_ma(xen_pgd[i]) & _PAGE_PRESENT) {
-                                pmd_t *pmd = (pmd_t *)alloc_bootmem_low_pages(PAGE_SIZE);
+        /*
+         * For PAE, need to allocate new pmds, rather than
+         * share Xen's, since Xen doesn't like pmd's being
+         * shared between address spaces.
+         */
+        for (i = 0; i < PTRS_PER_PGD; i++) {
+                if (pgd_val_ma(xen_pgd[i]) & _PAGE_PRESENT) {
+                        pmd_t *pmd = (pmd_t *)alloc_bootmem_low_pages(PAGE_SIZE);
 
-                                memcpy(pmd, (void *)pgd_page_vaddr(xen_pgd[i]),
-                                       PAGE_SIZE);
+                        memcpy(pmd, (void *)pgd_page_vaddr(xen_pgd[i]),
+                               PAGE_SIZE);
 
-                                make_lowmem_page_readonly(pmd);
+                        make_lowmem_page_readonly(pmd);
 
-                                set_pgd(&base[i], __pgd(1 + __pa(pmd)));
-                        } else
-                                pgd_clear(&base[i]);
-                }
+                        set_pgd(&base[i], __pgd(1 + __pa(pmd)));
+                } else
+                        pgd_clear(&base[i]);
         }
 
         /* make sure zero_page is mapped RO so we can use it in pagetables */
@@ -873,17 +870,7 @@ static __init void xen_pagetable_setup_done(pgd_t *base)
 
         /* Actually pin the pagetable down, but we can't set PG_pinned
            yet because the page structures don't exist yet. */
-        {
-                unsigned level;
-
-#ifdef CONFIG_X86_PAE
-                level = MMUEXT_PIN_L3_TABLE;
-#else
-                level = MMUEXT_PIN_L2_TABLE;
-#endif
-
-                pin_pagetable_pfn(level, PFN_DOWN(__pa(base)));
-        }
+        pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(base)));
 }
 
 /* This is called once we have the cpu_possible_map */
@@ -1093,7 +1080,6 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
         .make_pte = xen_make_pte,
         .make_pgd = xen_make_pgd,
 
-#ifdef CONFIG_X86_PAE
         .set_pte_atomic = xen_set_pte_atomic,
         .set_pte_present = xen_set_pte_at,
         .set_pud = xen_set_pud,
@@ -1102,7 +1088,6 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
 
         .make_pmd = xen_make_pmd,
         .pmd_val = xen_pmd_val,
-#endif  /* PAE */
 
         .activate_mm = xen_activate_mm,
         .dup_mmap = xen_dup_mmap,
@@ -1228,6 +1213,11 @@ asmlinkage void __init xen_start_kernel(void)
         if (xen_feature(XENFEAT_supervisor_mode_kernel))
                 pv_info.kernel_rpl = 0;
 
+        /* Prevent unwanted bits from being set in PTEs. */
+        __supported_pte_mask &= ~_PAGE_GLOBAL;
+        if (!is_initial_xendomain())
+                __supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
+
         /* set the limit of our address space */
         xen_reserve_top();
 
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index 126766d43aea..4e527e7893a8 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -60,7 +60,7 @@ xmaddr_t arbitrary_virt_to_machine(unsigned long address)
 {
         unsigned int level;
         pte_t *pte = lookup_address(address, &level);
-        unsigned offset = address & PAGE_MASK;
+        unsigned offset = address & ~PAGE_MASK;
 
         BUG_ON(pte == NULL);
 
@@ -179,50 +179,56 @@ out:
                 preempt_enable();
 }
 
-pteval_t xen_pte_val(pte_t pte)
+/* Assume pteval_t is equivalent to all the other *val_t types. */
+static pteval_t pte_mfn_to_pfn(pteval_t val)
+{
+        if (val & _PAGE_PRESENT) {
+                unsigned long mfn = (val & PTE_MASK) >> PAGE_SHIFT;
+                pteval_t flags = val & ~PTE_MASK;
+                val = ((pteval_t)mfn_to_pfn(mfn) << PAGE_SHIFT) | flags;
+        }
+
+        return val;
+}
+
+static pteval_t pte_pfn_to_mfn(pteval_t val)
 {
-        pteval_t ret = pte.pte;
+        if (val & _PAGE_PRESENT) {
+                unsigned long pfn = (val & PTE_MASK) >> PAGE_SHIFT;
+                pteval_t flags = val & ~PTE_MASK;
+                val = ((pteval_t)pfn_to_mfn(pfn) << PAGE_SHIFT) | flags;
+        }
 
-        if (ret & _PAGE_PRESENT)
-                ret = machine_to_phys(XMADDR(ret)).paddr | _PAGE_PRESENT;
+        return val;
+}
 
-        return ret;
+pteval_t xen_pte_val(pte_t pte)
+{
+        return pte_mfn_to_pfn(pte.pte);
 }
 
 pgdval_t xen_pgd_val(pgd_t pgd)
 {
-        pgdval_t ret = pgd.pgd;
-        if (ret & _PAGE_PRESENT)
-                ret = machine_to_phys(XMADDR(ret)).paddr | _PAGE_PRESENT;
-        return ret;
+        return pte_mfn_to_pfn(pgd.pgd);
 }
 
 pte_t xen_make_pte(pteval_t pte)
 {
-        if (pte & _PAGE_PRESENT) {
-                pte = phys_to_machine(XPADDR(pte)).maddr;
-                pte &= ~(_PAGE_PCD | _PAGE_PWT);
-        }
-
-        return (pte_t){ .pte = pte };
+        pte = pte_pfn_to_mfn(pte);
+        return native_make_pte(pte);
 }
 
 pgd_t xen_make_pgd(pgdval_t pgd)
 {
-        if (pgd & _PAGE_PRESENT)
-                pgd = phys_to_machine(XPADDR(pgd)).maddr;
-
-        return (pgd_t){ pgd };
+        pgd = pte_pfn_to_mfn(pgd);
+        return native_make_pgd(pgd);
 }
 
 pmdval_t xen_pmd_val(pmd_t pmd)
 {
-        pmdval_t ret = native_pmd_val(pmd);
-        if (ret & _PAGE_PRESENT)
-                ret = machine_to_phys(XMADDR(ret)).paddr | _PAGE_PRESENT;
-        return ret;
+        return pte_mfn_to_pfn(pmd.pmd);
 }
-#ifdef CONFIG_X86_PAE
+
 void xen_set_pud(pud_t *ptr, pud_t val)
 {
         struct multicall_space mcs;
@@ -267,17 +273,9 @@ void xen_pmd_clear(pmd_t *pmdp)
 
 pmd_t xen_make_pmd(pmdval_t pmd)
 {
-        if (pmd & _PAGE_PRESENT)
-                pmd = phys_to_machine(XPADDR(pmd)).maddr;
-
+        pmd = pte_pfn_to_mfn(pmd);
         return native_make_pmd(pmd);
 }
-#else  /* !PAE */
-void xen_set_pte(pte_t *ptep, pte_t pte)
-{
-        *ptep = pte;
-}
-#endif  /* CONFIG_X86_PAE */
 
 /*
   (Yet another) pagetable walker.  This one is intended for pinning a
@@ -430,8 +428,6 @@ static int pin_page(struct page *page, enum pt_level level)
    read-only, and can be pinned. */
 void xen_pgd_pin(pgd_t *pgd)
 {
-        unsigned level;
-
         xen_mc_batch();
 
         if (pgd_walk(pgd, pin_page, TASK_SIZE)) {
@@ -441,14 +437,7 @@ void xen_pgd_pin(pgd_t *pgd)
                 xen_mc_batch();
         }
 
-#ifdef CONFIG_X86_PAE
-        level = MMUEXT_PIN_L3_TABLE;
-#else
-        level = MMUEXT_PIN_L2_TABLE;
-#endif
-
-        xen_do_pin(level, PFN_DOWN(__pa(pgd)));
-
+        xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd)));
         xen_mc_issue(0);
 }
 
diff --git a/arch/x86/xen/mmu.h b/arch/x86/xen/mmu.h
index b5e189b1519d..5fe961caffd4 100644
--- a/arch/x86/xen/mmu.h
+++ b/arch/x86/xen/mmu.h
@@ -37,14 +37,13 @@ void xen_exit_mmap(struct mm_struct *mm);
 void xen_pgd_pin(pgd_t *pgd);
 //void xen_pgd_unpin(pgd_t *pgd);
 
-#ifdef CONFIG_X86_PAE
-unsigned long long xen_pte_val(pte_t);
-unsigned long long xen_pmd_val(pmd_t);
-unsigned long long xen_pgd_val(pgd_t);
+pteval_t xen_pte_val(pte_t);
+pmdval_t xen_pmd_val(pmd_t);
+pgdval_t xen_pgd_val(pgd_t);
 
-pte_t xen_make_pte(unsigned long long);
-pmd_t xen_make_pmd(unsigned long long);
-pgd_t xen_make_pgd(unsigned long long);
+pte_t xen_make_pte(pteval_t);
+pmd_t xen_make_pmd(pmdval_t);
+pgd_t xen_make_pgd(pgdval_t);
 
 void xen_set_pte_at(struct mm_struct *mm, unsigned long addr,
                     pte_t *ptep, pte_t pteval);
@@ -53,15 +52,4 @@ void xen_set_pud(pud_t *ptr, pud_t val);
 void xen_pte_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep);
 void xen_pmd_clear(pmd_t *pmdp);
 
-
-#else
-unsigned long xen_pte_val(pte_t);
-unsigned long xen_pmd_val(pmd_t);
-unsigned long xen_pgd_val(pgd_t);
-
-pte_t xen_make_pte(unsigned long);
-pmd_t xen_make_pmd(unsigned long);
-pgd_t xen_make_pgd(unsigned long);
-#endif
-
 #endif  /* _XEN_MMU_H */
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index c39e1a5aa241..41e217503c96 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -12,7 +12,9 @@
 #include <linux/clocksource.h>
 #include <linux/clockchips.h>
 #include <linux/kernel_stat.h>
+#include <linux/math64.h>
 
+#include <asm/pvclock.h>
 #include <asm/xen/hypervisor.h>
 #include <asm/xen/hypercall.h>
 
@@ -30,17 +32,6 @@
 
 static cycle_t xen_clocksource_read(void);
 
-/* These are perodically updated in shared_info, and then copied here. */
-struct shadow_time_info {
-        u64 tsc_timestamp;     /* TSC at last update of time vals.  */
-        u64 system_timestamp;  /* Time, in nanosecs, since boot.    */
-        u32 tsc_to_nsec_mul;
-        int tsc_shift;
-        u32 version;
-};
-
-static DEFINE_PER_CPU(struct shadow_time_info, shadow_time);
-
 /* runstate info updated by Xen */
 static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
 
@@ -150,11 +141,7 @@ static void do_stolen_accounting(void)
         if (stolen < 0)
                 stolen = 0;
 
-        ticks = 0;
-        while (stolen >= NS_PER_TICK) {
-                ticks++;
-                stolen -= NS_PER_TICK;
-        }
+        ticks = iter_div_u64_rem(stolen, NS_PER_TICK, &stolen);
         __get_cpu_var(residual_stolen) = stolen;
         account_steal_time(NULL, ticks);
 
@@ -166,11 +153,7 @@ static void do_stolen_accounting(void)
         if (blocked < 0)
                 blocked = 0;
 
-        ticks = 0;
-        while (blocked >= NS_PER_TICK) {
-                ticks++;
-                blocked -= NS_PER_TICK;
-        }
+        ticks = iter_div_u64_rem(blocked, NS_PER_TICK, &blocked);
         __get_cpu_var(residual_blocked) = blocked;
         account_steal_time(idle_task(smp_processor_id()), ticks);
 }
@@ -218,7 +201,7 @@ unsigned long long xen_sched_clock(void)
 unsigned long xen_cpu_khz(void)
 {
         u64 xen_khz = 1000000ULL << 32;
-        const struct vcpu_time_info *info =
+        const struct pvclock_vcpu_time_info *info =
                 &HYPERVISOR_shared_info->vcpu_info[0].time;
 
         do_div(xen_khz, info->tsc_to_system_mul);
@@ -230,121 +213,26 @@ unsigned long xen_cpu_khz(void)
         return xen_khz;
 }
 
-/*
- * Reads a consistent set of time-base values from Xen, into a shadow data
- * area.
- */
-static unsigned get_time_values_from_xen(void)
-{
-        struct vcpu_time_info   *src;
-        struct shadow_time_info *dst;
-
-        /* src is shared memory with the hypervisor, so we need to
-           make sure we get a consistent snapshot, even in the face of
-           being preempted. */
-        src = &__get_cpu_var(xen_vcpu)->time;
-        dst = &__get_cpu_var(shadow_time);
-
-        do {
-                dst->version = src->version;
-                rmb();          /* fetch version before data */
-                dst->tsc_timestamp     = src->tsc_timestamp;
-                dst->system_timestamp  = src->system_time;
-                dst->tsc_to_nsec_mul   = src->tsc_to_system_mul;
-                dst->tsc_shift         = src->tsc_shift;
-                rmb();          /* test version after fetching data */
-        } while ((src->version & 1) | (dst->version ^ src->version));
-
-        return dst->version;
-}
-
-/*
- * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
- * yielding a 64-bit result.
- */
-static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
-{
-        u64 product;
-#ifdef __i386__
-        u32 tmp1, tmp2;
-#endif
-
-        if (shift < 0)
-                delta >>= -shift;
-        else
-                delta <<= shift;
-
-#ifdef __i386__
-        __asm__ (
-                "mul  %5       ; "
-                "mov  %4,%%eax ; "
-                "mov  %%edx,%4 ; "
-                "mul  %5       ; "
-                "xor  %5,%5    ; "
-                "add  %4,%%eax ; "
-                "adc  %5,%%edx ; "
-                : "=A" (product), "=r" (tmp1), "=r" (tmp2)
-                : "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
-#elif __x86_64__
-        __asm__ (
-                "mul %%rdx ; shrd $32,%%rdx,%%rax"
-                : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
-#else
-#error implement me!
-#endif
-
-        return product;
-}
-
-static u64 get_nsec_offset(struct shadow_time_info *shadow)
-{
-        u64 now, delta;
-        now = native_read_tsc();
-        delta = now - shadow->tsc_timestamp;
-        return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
-}
-
 static cycle_t xen_clocksource_read(void)
 {
-        struct shadow_time_info *shadow = &get_cpu_var(shadow_time);
+        struct pvclock_vcpu_time_info *src;
         cycle_t ret;
-        unsigned version;
-
-        do {
-                version = get_time_values_from_xen();
-                barrier();
-                ret = shadow->system_timestamp + get_nsec_offset(shadow);
-                barrier();
-        } while (version != __get_cpu_var(xen_vcpu)->time.version);
-
-        put_cpu_var(shadow_time);
 
+        src = &get_cpu_var(xen_vcpu)->time;
+        ret = pvclock_clocksource_read(src);
+        put_cpu_var(xen_vcpu);
         return ret;
 }
 
 static void xen_read_wallclock(struct timespec *ts)
 {
-        const struct shared_info *s = HYPERVISOR_shared_info;
-        u32 version;
-        u64 delta;
-        struct timespec now;
-
-        /* get wallclock at system boot */
-        do {
-                version = s->wc_version;
-                rmb();          /* fetch version before time */
-                now.tv_sec  = s->wc_sec;
-                now.tv_nsec = s->wc_nsec;
-                rmb();          /* fetch time before checking version */
-        } while ((s->wc_version & 1) | (version ^ s->wc_version));
+        struct shared_info *s = HYPERVISOR_shared_info;
+        struct pvclock_wall_clock *wall_clock = &(s->wc);
+        struct pvclock_vcpu_time_info *vcpu_time;
 
-        delta = xen_clocksource_read(); /* time since system boot */
-        delta += now.tv_sec * (u64)NSEC_PER_SEC + now.tv_nsec;
-
-        now.tv_nsec = do_div(delta, NSEC_PER_SEC);
-        now.tv_sec = delta;
-
-        set_normalized_timespec(ts, now.tv_sec, now.tv_nsec);
+        vcpu_time = &get_cpu_var(xen_vcpu)->time;
+        pvclock_read_wallclock(wall_clock, vcpu_time, ts);
+        put_cpu_var(xen_vcpu);
 }
 
 unsigned long xen_get_wallclock(void)
@@ -352,7 +240,6 @@ unsigned long xen_get_wallclock(void)
         struct timespec ts;
 
         xen_read_wallclock(&ts);
-
         return ts.tv_sec;
 }
 
@@ -576,8 +463,6 @@ __init void xen_time_init(void)
 {
         int cpu = smp_processor_id();
 
-        get_time_values_from_xen();
-
         clocksource_register(&xen_clocksource);
 
         if (HYPERVISOR_vcpu_op(VCPUOP_stop_periodic_timer, cpu, NULL) == 0) {
diff --git a/arch/x86/xen/xen-head.S b/arch/x86/xen/xen-head.S
index 288d587ce73c..6ec3b4f7719b 100644
--- a/arch/x86/xen/xen-head.S
+++ b/arch/x86/xen/xen-head.S
@@ -17,7 +17,7 @@ ENTRY(startup_xen)
 
         __FINIT
 
-.pushsection .bss.page_aligned
+.pushsection .text
         .align PAGE_SIZE_asm
 ENTRY(hypercall_page)
         .skip 0x1000
@@ -30,11 +30,7 @@ ENTRY(hypercall_page)
         ELFNOTE(Xen, XEN_ELFNOTE_ENTRY,          .long  startup_xen)
         ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, .long  hypercall_page)
         ELFNOTE(Xen, XEN_ELFNOTE_FEATURES,       .asciz "!writable_page_tables|pae_pgdir_above_4gb")
-#ifdef CONFIG_X86_PAE
         ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE,       .asciz "yes")
-#else
-        ELFNOTE(Xen, XEN_ELFNOTE_PAE_MODE,       .asciz "no")
-#endif
         ELFNOTE(Xen, XEN_ELFNOTE_LOADER,         .asciz "generic")
 
 #endif /*CONFIG_XEN */