[PATCH] Proper fix for highmem kmap_atomic functions for VMI for 2.6.21

Since lazy MMU batching mode still allows interrupts to enter, it is
possible for interrupt handlers to try to use kmap_atomic, which fails when
lazy mode is active, since the PTE update to highmem will be delayed.  The
best workaround is to issue an explicit flush in kmap_atomic_functions
case; this is the only way nested PTE updates can happen in the interrupt
handler.

Thanks to Jeremy Fitzhardinge for noting the bug and suggestions on a fix.

This patch gets reverted again when we start 2.6.22 and the bug gets fixed
differently.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Andi Kleen <ak@muc.de>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

4 files changed, 29 insertions, 1 deletions

diff --git a/arch/i386/kernel/vmi.c b/arch/i386/kernel/vmi.c
index fb07a1aad225..edc339fa5038 100644
--- a/arch/i386/kernel/vmi.c
+++ b/arch/i386/kernel/vmi.c
@@ -69,6 +69,7 @@ struct {
         void (*flush_tlb)(int);
         void (*set_initial_ap_state)(int, int);
         void (*halt)(void);
+        void (*set_lazy_mode)(int mode);
 } vmi_ops;
 
 /* XXX move this to alternative.h */
@@ -574,6 +575,26 @@ vmi_startup_ipi_hook(int phys_apicid, unsigned long start_eip,
 }
 #endif
 
+static void vmi_set_lazy_mode(int mode)
+{
+        static DEFINE_PER_CPU(int, lazy_mode);
+
+        if (!vmi_ops.set_lazy_mode)
+                return;
+
+        /* Modes should never nest or overlap */
+        BUG_ON(__get_cpu_var(lazy_mode) && !(mode == PARAVIRT_LAZY_NONE ||
+                                             mode == PARAVIRT_LAZY_FLUSH));
+
+        if (mode == PARAVIRT_LAZY_FLUSH) {
+                vmi_ops.set_lazy_mode(0);
+                vmi_ops.set_lazy_mode(__get_cpu_var(lazy_mode));
+        } else {
+                vmi_ops.set_lazy_mode(mode);
+                __get_cpu_var(lazy_mode) = mode;
+        }
+}
+
 static inline int __init check_vmi_rom(struct vrom_header *rom)
 {
         struct pci_header *pci;
@@ -804,7 +825,7 @@ static inline int __init activate_vmi(void)
         para_wrap(load_esp0, vmi_load_esp0, set_kernel_stack, UpdateKernelStack);
         para_fill(set_iopl_mask, SetIOPLMask);
         para_fill(io_delay, IODelay);
-        para_fill(set_lazy_mode, SetLazyMode);
+        para_wrap(set_lazy_mode, vmi_set_lazy_mode, set_lazy_mode, SetLazyMode);
 
         /* user and kernel flush are just handled with different flags to FlushTLB */
         para_wrap(flush_tlb_user, vmi_flush_tlb_user, flush_tlb, FlushTLB);
diff --git a/arch/i386/mm/highmem.c b/arch/i386/mm/highmem.c
index bb2de1089add..ac70d09df7ee 100644
--- a/arch/i386/mm/highmem.c
+++ b/arch/i386/mm/highmem.c
@@ -42,6 +42,7 @@ void *kmap_atomic(struct page *page, enum km_type type)
 
         vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
         set_pte(kmap_pte-idx, mk_pte(page, kmap_prot));
+        arch_flush_lazy_mmu_mode();
 
         return (void*) vaddr;
 }
@@ -82,6 +83,7 @@ void *kmap_atomic_pfn(unsigned long pfn, enum km_type type)
         idx = type + KM_TYPE_NR*smp_processor_id();
         vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
         set_pte(kmap_pte-idx, pfn_pte(pfn, kmap_prot));
+        arch_flush_lazy_mmu_mode();
 
         return (void*) vaddr;
 }
diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h
index 00c23433b39f..6d7e279b1490 100644
--- a/include/asm-generic/pgtable.h
+++ b/include/asm-generic/pgtable.h
@@ -180,6 +180,7 @@ static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addres
 #ifndef __HAVE_ARCH_ENTER_LAZY_MMU_MODE
 #define arch_enter_lazy_mmu_mode()      do {} while (0)
 #define arch_leave_lazy_mmu_mode()      do {} while (0)
+#define arch_flush_lazy_mmu_mode()      do {} while (0)
 #endif
 
 /*
@@ -193,6 +194,7 @@ static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addres
 #ifndef __HAVE_ARCH_ENTER_LAZY_CPU_MODE
 #define arch_enter_lazy_cpu_mode()      do {} while (0)
 #define arch_leave_lazy_cpu_mode()      do {} while (0)
+#define arch_flush_lazy_cpu_mode()      do {} while (0)
 #endif
 
 /*
diff --git a/include/asm-i386/paravirt.h b/include/asm-i386/paravirt.h
index 46dc34ca887a..e63f1e444fcf 100644
--- a/include/asm-i386/paravirt.h
+++ b/include/asm-i386/paravirt.h
@@ -421,14 +421,17 @@ static inline void pmd_clear(pmd_t *pmdp)
 #define PARAVIRT_LAZY_NONE 0
 #define PARAVIRT_LAZY_MMU  1
 #define PARAVIRT_LAZY_CPU  2
+#define PARAVIRT_LAZY_FLUSH 3
 
 #define  __HAVE_ARCH_ENTER_LAZY_CPU_MODE
 #define arch_enter_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_CPU)
 #define arch_leave_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_NONE)
+#define arch_flush_lazy_cpu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_FLUSH)
 
 #define  __HAVE_ARCH_ENTER_LAZY_MMU_MODE
 #define arch_enter_lazy_mmu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_MMU)
 #define arch_leave_lazy_mmu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_NONE)
+#define arch_flush_lazy_mmu_mode() paravirt_ops.set_lazy_mode(PARAVIRT_LAZY_FLUSH)
 
 /* These all sit in the .parainstructions section to tell us what to patch. */
 struct paravirt_patch {