mmu_gather: move minimal range calculations into generic code

On architectures with hardware broadcasting of TLB invalidation messages
, it makes sense to reduce the range of the mmu_gather structure when
unmapping page ranges based on the dirty address information passed to
tlb_remove_tlb_entry.

arm64 already does this by directly manipulating the start/end fields
of the gather structure, but this confuses the generic code which
does not expect these fields to change and can end up calculating
invalid, negative ranges when forcing a flush in zap_pte_range.

This patch moves the minimal range calculation out of the arm64 code
and into the generic implementation, simplifying zap_pte_range in the
process (which no longer needs to care about start/end, since they will
point to the appropriate ranges already). With the range being tracked
by core code, the need_flush flag is dropped in favour of checking that
the end of the range has actually been set.

Cc: Benjamin Herrenschmidt <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Russell King - ARM Linux <[email protected]>
Cc: Michal Simek <[email protected]>
Acked-by: Linus Torvalds <[email protected]>
Signed-off-by: Will Deacon <[email protected]>

arch/arm64/include/asm/tlb.h

@@ -19,10 +19,6 @@
 #ifndef __ASM_TLB_H
 #define __ASM_TLB_H
 
-#define  __tlb_remove_pmd_tlb_entry __tlb_remove_pmd_tlb_entry
-
-#include <asm-generic/tlb.h>
-
 #include <linux/pagemap.h>
 #include <linux/swap.h>
 
@@ -37,79 +33,29 @@ static inline void __tlb_remove_table(void *_table)
 #define tlb_remove_entry(tlb, entry)	tlb_remove_page(tlb, entry)
 #endif /* CONFIG_HAVE_RCU_TABLE_FREE */
 
-/*
- * There's three ways the TLB shootdown code is used:
- *  1. Unmapping a range of vmas.  See zap_page_range(), unmap_region().
- *     tlb->fullmm = 0, and tlb_start_vma/tlb_end_vma will be called.
- *  2. Unmapping all vmas.  See exit_mmap().
- *     tlb->fullmm = 1, and tlb_start_vma/tlb_end_vma will be called.
- *     Page tables will be freed.
- *  3. Unmapping argument pages.  See shift_arg_pages().
- *     tlb->fullmm = 0, but tlb_start_vma/tlb_end_vma will not be called.
- */
+#include <asm-generic/tlb.h>
+
 static inline void tlb_flush(struct mmu_gather *tlb)
 {
 	if (tlb->fullmm) {
 		flush_tlb_mm(tlb->mm);
-	} else if (tlb->end > 0) {
+	} else {
 		struct vm_area_struct vma = { .vm_mm = tlb->mm, };
 		flush_tlb_range(&vma, tlb->start, tlb->end);
-		tlb->start = TASK_SIZE;
-		tlb->end = 0;
-	}
-}
-
-static inline void tlb_add_flush(struct mmu_gather *tlb, unsigned long addr)
-{
-	if (!tlb->fullmm) {
-		tlb->start = min(tlb->start, addr);
-		tlb->end = max(tlb->end, addr + PAGE_SIZE);
-	}
-}
-
-/*
- * Memorize the range for the TLB flush.
- */
-static inline void __tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep,
-					  unsigned long addr)
-{
-	tlb_add_flush(tlb, addr);
-}
-
-/*
- * In the case of tlb vma handling, we can optimise these away in the
- * case where we're doing a full MM flush.  When we're doing a munmap,
- * the vmas are adjusted to only cover the region to be torn down.
- */
-static inline void tlb_start_vma(struct mmu_gather *tlb,
-				 struct vm_area_struct *vma)
-{
-	if (!tlb->fullmm) {
-		tlb->start = TASK_SIZE;
-		tlb->end = 0;
 	}
 }
 
-static inline void tlb_end_vma(struct mmu_gather *tlb,
-			       struct vm_area_struct *vma)
-{
-	if (!tlb->fullmm)
-		tlb_flush(tlb);
-}
-
 static inline void __pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte,
 				  unsigned long addr)
 {
 	pgtable_page_dtor(pte);
-	tlb_add_flush(tlb, addr);
 	tlb_remove_entry(tlb, pte);
 }
 
 #if CONFIG_ARM64_PGTABLE_LEVELS > 2
 static inline void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmdp,
 				  unsigned long addr)
 {
-	tlb_add_flush(tlb, addr);
 	tlb_remove_entry(tlb, virt_to_page(pmdp));
 }
 #endif
@@ -118,15 +64,8 @@ static inline void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmdp,
 static inline void __pud_free_tlb(struct mmu_gather *tlb, pud_t *pudp,
 				  unsigned long addr)
 {
-	tlb_add_flush(tlb, addr);
 	tlb_remove_entry(tlb, virt_to_page(pudp));
 }
 #endif
 
-static inline void __tlb_remove_pmd_tlb_entry(struct mmu_gather *tlb, pmd_t *pmdp,
-						unsigned long address)
-{
-	tlb_add_flush(tlb, address);
-}
-
 #endif

arch/microblaze/include/asm/tlb.h

@@ -14,12 +14,13 @@
 #define tlb_flush(tlb)	flush_tlb_mm((tlb)->mm)
 
 #include <linux/pagemap.h>
-#include <asm-generic/tlb.h>
 
 #ifdef CONFIG_MMU
 #define tlb_start_vma(tlb, vma)		do { } while (0)
 #define tlb_end_vma(tlb, vma)		do { } while (0)
 #define __tlb_remove_tlb_entry(tlb, pte, address) do { } while (0)
 #endif
 
+#include <asm-generic/tlb.h>
+
 #endif /* _ASM_MICROBLAZE_TLB_H */

arch/powerpc/include/asm/pgalloc.h

@@ -3,7 +3,6 @@
 #ifdef __KERNEL__
 
 #include <linux/mm.h>
-#include <asm-generic/tlb.h>
 
 #ifdef CONFIG_PPC_BOOK3E
 extern void tlb_flush_pgtable(struct mmu_gather *tlb, unsigned long address);
@@ -14,6 +13,8 @@ static inline void tlb_flush_pgtable(struct mmu_gather *tlb,
 }
 #endif /* !CONFIG_PPC_BOOK3E */
 
+extern void tlb_remove_table(struct mmu_gather *tlb, void *table);
+
 #ifdef CONFIG_PPC64
 #include <asm/pgalloc-64.h>
 #else

arch/powerpc/include/asm/tlb.h

@@ -27,6 +27,7 @@
 
 #define tlb_start_vma(tlb, vma)	do { } while (0)
 #define tlb_end_vma(tlb, vma)	do { } while (0)
+#define __tlb_remove_tlb_entry	__tlb_remove_tlb_entry
 
 extern void tlb_flush(struct mmu_gather *tlb);
 

arch/powerpc/mm/hugetlbpage.c

@@ -517,8 +517,6 @@ static void free_hugepd_range(struct mmu_gather *tlb, hugepd_t *hpdp, int pdshif
 	for (i = 0; i < num_hugepd; i++, hpdp++)
 		hpdp->pd = 0;
 
-	tlb->need_flush = 1;
-
 #ifdef CONFIG_PPC_FSL_BOOK3E
 	hugepd_free(tlb, hugepte);
 #else

include/asm-generic/tlb.h

@@ -96,10 +96,9 @@ struct mmu_gather {
 #endif
 	unsigned long		start;
 	unsigned long		end;
-	unsigned int		need_flush : 1,	/* Did free PTEs */
 	/* we are in the middle of an operation to clear
 	 * a full mm and can make some optimizations */
-				fullmm : 1,
+	unsigned int		fullmm : 1,
 	/* we have performed an operation which
 	 * requires a complete flush of the tlb */
 				need_flush_all : 1;
@@ -128,16 +127,54 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 		tlb_flush_mmu(tlb);
 }
 
+static inline void __tlb_adjust_range(struct mmu_gather *tlb,
+				      unsigned long address)
+{
+	tlb->start = min(tlb->start, address);
+	tlb->end = max(tlb->end, address + PAGE_SIZE);
+}
+
+static inline void __tlb_reset_range(struct mmu_gather *tlb)
+{
+	tlb->start = TASK_SIZE;
+	tlb->end = 0;
+}
+
+/*
+ * In the case of tlb vma handling, we can optimise these away in the
+ * case where we're doing a full MM flush.  When we're doing a munmap,
+ * the vmas are adjusted to only cover the region to be torn down.
+ */
+#ifndef tlb_start_vma
+#define tlb_start_vma(tlb, vma) do { } while (0)
+#endif
+
+#define __tlb_end_vma(tlb, vma)					\
+	do {							\
+		if (!tlb->fullmm && tlb->end) {			\
+			tlb_flush(tlb);				\
+			__tlb_reset_range(tlb);			\
+		}						\
+	} while (0)
+
+#ifndef tlb_end_vma
+#define tlb_end_vma	__tlb_end_vma
+#endif
+
+#ifndef __tlb_remove_tlb_entry
+#define __tlb_remove_tlb_entry(tlb, ptep, address) do { } while (0)
+#endif
+
 /**
  * tlb_remove_tlb_entry - remember a pte unmapping for later tlb invalidation.
  *
- * Record the fact that pte's were really umapped in ->need_flush, so we can
- * later optimise away the tlb invalidate.   This helps when userspace is
- * unmapping already-unmapped pages, which happens quite a lot.
+ * Record the fact that pte's were really unmapped by updating the range,
+ * so we can later optimise away the tlb invalidate.   This helps when
+ * userspace is unmapping already-unmapped pages, which happens quite a lot.
  */
 #define tlb_remove_tlb_entry(tlb, ptep, address)		\
 	do {							\
-		tlb->need_flush = 1;				\
+		__tlb_adjust_range(tlb, address);		\
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
@@ -151,27 +188,27 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 
 #define tlb_remove_pmd_tlb_entry(tlb, pmdp, address)		\
 	do {							\
-		tlb->need_flush = 1;				\
+		__tlb_adjust_range(tlb, address);		\
 		__tlb_remove_pmd_tlb_entry(tlb, pmdp, address);	\
 	} while (0)
 
 #define pte_free_tlb(tlb, ptep, address)			\
 	do {							\
-		tlb->need_flush = 1;				\
+		__tlb_adjust_range(tlb, address);		\
 		__pte_free_tlb(tlb, ptep, address);		\
 	} while (0)
 
 #ifndef __ARCH_HAS_4LEVEL_HACK
 #define pud_free_tlb(tlb, pudp, address)			\
 	do {							\
-		tlb->need_flush = 1;				\
+		__tlb_adjust_range(tlb, address);		\
 		__pud_free_tlb(tlb, pudp, address);		\
 	} while (0)
 #endif
 
 #define pmd_free_tlb(tlb, pmdp, address)			\
 	do {							\
-		tlb->need_flush = 1;				\
+		__tlb_adjust_range(tlb, address);		\
 		__pmd_free_tlb(tlb, pmdp, address);		\
 	} while (0)
 

mm/memory.c

@@ -220,9 +220,6 @@ void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, unsigned long
 	/* Is it from 0 to ~0? */
 	tlb->fullmm     = !(start | (end+1));
 	tlb->need_flush_all = 0;
-	tlb->start	= start;
-	tlb->end	= end;
-	tlb->need_flush = 0;
 	tlb->local.next = NULL;
 	tlb->local.nr   = 0;
 	tlb->local.max  = ARRAY_SIZE(tlb->__pages);
@@ -232,15 +229,20 @@ void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, unsigned long
 #ifdef CONFIG_HAVE_RCU_TABLE_FREE
 	tlb->batch = NULL;
 #endif
+
+	__tlb_reset_range(tlb);
 }
 
 static void tlb_flush_mmu_tlbonly(struct mmu_gather *tlb)
 {
-	tlb->need_flush = 0;
+	if (!tlb->end)
+		return;
+
 	tlb_flush(tlb);
 #ifdef CONFIG_HAVE_RCU_TABLE_FREE
 	tlb_table_flush(tlb);
 #endif
+	__tlb_reset_range(tlb);
 }
 
 static void tlb_flush_mmu_free(struct mmu_gather *tlb)
@@ -256,8 +258,6 @@ static void tlb_flush_mmu_free(struct mmu_gather *tlb)
 
 void tlb_flush_mmu(struct mmu_gather *tlb)
 {
-	if (!tlb->need_flush)
-		return;
 	tlb_flush_mmu_tlbonly(tlb);
 	tlb_flush_mmu_free(tlb);
 }
@@ -292,7 +292,7 @@ int __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
 	struct mmu_gather_batch *batch;
 
-	VM_BUG_ON(!tlb->need_flush);
+	VM_BUG_ON(!tlb->end);
 
 	batch = tlb->active;
 	batch->pages[batch->nr++] = page;
@@ -359,8 +359,6 @@ void tlb_remove_table(struct mmu_gather *tlb, void *table)
 {
 	struct mmu_table_batch **batch = &tlb->batch;
 
-	tlb->need_flush = 1;
-
 	/*
 	 * When there's less then two users of this mm there cannot be a
 	 * concurrent page-table walk.
@@ -1185,20 +1183,8 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	arch_leave_lazy_mmu_mode();
 
 	/* Do the actual TLB flush before dropping ptl */
-	if (force_flush) {
-		unsigned long old_end;
-
-		/*
-		 * Flush the TLB just for the previous segment,
-		 * then update the range to be the remaining
-		 * TLB range.
-		 */
-		old_end = tlb->end;
-		tlb->end = addr;
+	if (force_flush)
 		tlb_flush_mmu_tlbonly(tlb);
-		tlb->start = addr;
-		tlb->end = old_end;
-	}
 	pte_unmap_unlock(start_pte, ptl);
 
 	/*