[PATCH 0/5] mmu gather changes

[PATCH 0/5] mmu gather changes

[Aneesh Kumar K.V]

Hi,

With commit e77b0852b551 ("mm/mmu_gather: track page size with mmu gather and force flush if page size change")
we added the ability to force a tlb flush when the page size change in
a mmu_gather loop. We did that by checking for a page size change
every time we added a page to mmu_gather for lazy flush/remove. This breaks when
invalidating a range that contain dax mapping. Dax pte entries are
considered special pte entries and the mmu_gather loop doesn't call
tlb_remove_page for them. Hence the page size change check doesn't work when
the invalidate range covers dax mappings.

We can also improve the current code by moving the page size change check
early and not doing it every time we add a page.

This series is not a bug fix series, because only arch that use the page
size change detail is ppc64 and we don't have dax enabled on that arch yet.
The series also include code changes that improve code comments and add new helpers
to makes it easy to follow.


Aneesh Kumar K.V (5):
  mm: Use the correct page size when removing the page
  mm: Update mmu_gather range correctly
  mm/hugetlb: add tlb_remove_hugetlb_entry for handling hugetlb pages
  mm: Add tlb_remove_check_page_size_change to track page size change
  mm: Remove the page size change check in tlb_remove_page

 arch/arm/include/asm/tlb.h     | 21 ++++++-----
 arch/ia64/include/asm/tlb.h    | 25 +++++++------
 arch/powerpc/include/asm/tlb.h | 16 ++++++++
 arch/s390/include/asm/tlb.h    | 14 ++++---
 arch/sh/include/asm/tlb.h      | 15 +++++---
 arch/um/include/asm/tlb.h      | 15 +++++---
 include/asm-generic/tlb.h      | 83 +++++++++++++++++++++++++-----------------
 mm/huge_memory.c               |  8 +++-
 mm/hugetlb.c                   |  7 +++-
 mm/madvise.c                   |  1 +
 mm/memory.c                    | 28 ++++++--------
 11 files changed, 141 insertions(+), 92 deletions(-)

-- 
2.10.1

[PATCH 0/5] mmu gather changes

[Aneesh Kumar K.V]

Hi,

With commit e77b0852b551 ("mm/mmu_gather: track page size with mmu gather and force flush if page size change")
we added the ability to force a tlb flush when the page size change in
a mmu_gather loop. We did that by checking for a page size change
every time we added a page to mmu_gather for lazy flush/remove. This breaks when
invalidating a range that contain dax mapping. Dax pte entries are
considered special pte entries and the mmu_gather loop doesn't call
tlb_remove_page for them. Hence the page size change check doesn't work when
the invalidate range covers dax mappings.

We can also improve the current code by moving the page size change check
early and not doing it every time we add a page.

This series is not a bug fix series, because only arch that use the page
size change detail is ppc64 and we don't have dax enabled on that arch yet.
The series also include code changes that improve code comments and add new helpers
to makes it easy to follow.


Aneesh Kumar K.V (5):
  mm: Use the correct page size when removing the page
  mm: Update mmu_gather range correctly
  mm/hugetlb: add tlb_remove_hugetlb_entry for handling hugetlb pages
  mm: Add tlb_remove_check_page_size_change to track page size change
  mm: Remove the page size change check in tlb_remove_page

 arch/arm/include/asm/tlb.h     | 21 ++++++-----
 arch/ia64/include/asm/tlb.h    | 25 +++++++------
 arch/powerpc/include/asm/tlb.h | 16 ++++++++
 arch/s390/include/asm/tlb.h    | 14 ++++---
 arch/sh/include/asm/tlb.h      | 15 +++++---
 arch/um/include/asm/tlb.h      | 15 +++++---
 include/asm-generic/tlb.h      | 83 +++++++++++++++++++++++++-----------------
 mm/huge_memory.c               |  8 +++-
 mm/hugetlb.c                   |  7 +++-
 mm/madvise.c                   |  1 +
 mm/memory.c                    | 28 ++++++--------
 11 files changed, 141 insertions(+), 92 deletions(-)

-- 
2.10.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 1/5] mm: Use the correct page size when removing the page

[Aneesh Kumar K.V]

We are removing a pmd hugepage here. Use the correct page size.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 mm/huge_memory.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index cdcd25cb30fe..12c6685d4e97 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1398,12 +1398,12 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	if (vma_is_dax(vma)) {
 		spin_unlock(ptl);
 		if (is_huge_zero_pmd(orig_pmd))
-			tlb_remove_page(tlb, pmd_page(orig_pmd));
+			tlb_remove_page_size(tlb, pmd_page(orig_pmd), HPAGE_PMD_SIZE);
 	} else if (is_huge_zero_pmd(orig_pmd)) {
 		pte_free(tlb->mm, pgtable_trans_huge_withdraw(tlb->mm, pmd));
 		atomic_long_dec(&tlb->mm->nr_ptes);
 		spin_unlock(ptl);
-		tlb_remove_page(tlb, pmd_page(orig_pmd));
+		tlb_remove_page_size(tlb, pmd_page(orig_pmd), HPAGE_PMD_SIZE);
 	} else {
 		struct page *page = pmd_page(orig_pmd);
 		page_remove_rmap(page, true);
-- 
2.10.1

[PATCH 1/5] mm: Use the correct page size when removing the page

[Aneesh Kumar K.V]

We are removing a pmd hugepage here. Use the correct page size.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 mm/huge_memory.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index cdcd25cb30fe..12c6685d4e97 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1398,12 +1398,12 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	if (vma_is_dax(vma)) {
 		spin_unlock(ptl);
 		if (is_huge_zero_pmd(orig_pmd))
-			tlb_remove_page(tlb, pmd_page(orig_pmd));
+			tlb_remove_page_size(tlb, pmd_page(orig_pmd), HPAGE_PMD_SIZE);
 	} else if (is_huge_zero_pmd(orig_pmd)) {
 		pte_free(tlb->mm, pgtable_trans_huge_withdraw(tlb->mm, pmd));
 		atomic_long_dec(&tlb->mm->nr_ptes);
 		spin_unlock(ptl);
-		tlb_remove_page(tlb, pmd_page(orig_pmd));
+		tlb_remove_page_size(tlb, pmd_page(orig_pmd), HPAGE_PMD_SIZE);
 	} else {
 		struct page *page = pmd_page(orig_pmd);
 		page_remove_rmap(page, true);
-- 
2.10.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 2/5] mm: Update mmu_gather range correctly

[Aneesh Kumar K.V]

We use __tlb_adjust_range to update range convered by mmu_gather struct.
We later use the 'start' and 'end' to do a mmu_notifier_invalidate_range
in tlb_flush_mmu_tlbonly(). Update the 'end' correctly in
__tlb_adjust_range so that we call mmu_notifier_invalidate_range with the
correct range values.

W.r.t to tlbflush, this should not have any impact, because a flush with
correct start address will flush tlb mapping for the range.

Also add comment w.r.t updating the range when we free pagetable pages.
For now we don't support a range based page table cache flush.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 include/asm-generic/tlb.h | 43 +++++++++++++++++++++++++++++++------------
 1 file changed, 31 insertions(+), 12 deletions(-)

diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index c6d667187608..dba727becd5f 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -125,10 +125,11 @@ extern bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page,
 				   int page_size);
 
 static inline void __tlb_adjust_range(struct mmu_gather *tlb,
-				      unsigned long address)
+				      unsigned long address,
+				      unsigned int range_size)
 {
 	tlb->start = min(tlb->start, address);
-	tlb->end = max(tlb->end, address + PAGE_SIZE);
+	tlb->end = max(tlb->end, address + range_size);
 	/*
 	 * Track the last address with which we adjusted the range. This
 	 * will be used later to adjust again after a mmu_flush due to
@@ -153,7 +154,7 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 	if (__tlb_remove_page_size(tlb, page, page_size)) {
 		tlb_flush_mmu(tlb);
 		tlb->page_size = page_size;
-		__tlb_adjust_range(tlb, tlb->addr);
+		__tlb_adjust_range(tlb, tlb->addr, page_size);
 		__tlb_remove_page_size(tlb, page, page_size);
 	}
 }
@@ -177,7 +178,7 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 	/* active->nr should be zero when we call this */
 	VM_BUG_ON_PAGE(tlb->active->nr, page);
 	tlb->page_size = PAGE_SIZE;
-	__tlb_adjust_range(tlb, tlb->addr);
+	__tlb_adjust_range(tlb, tlb->addr, PAGE_SIZE);
 	return __tlb_remove_page(tlb, page);
 }
 
@@ -215,7 +216,7 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
  */
 #define tlb_remove_tlb_entry(tlb, ptep, address)		\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
@@ -227,29 +228,47 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 #define __tlb_remove_pmd_tlb_entry(tlb, pmdp, address) do {} while (0)
 #endif
 
-#define tlb_remove_pmd_tlb_entry(tlb, pmdp, address)		\
-	do {							\
-		__tlb_adjust_range(tlb, address);		\
-		__tlb_remove_pmd_tlb_entry(tlb, pmdp, address);	\
+#define tlb_remove_pmd_tlb_entry(tlb, pmdp, address)			\
+	do {								\
+		__tlb_adjust_range(tlb, address, HPAGE_PMD_SIZE);	\
+		__tlb_remove_pmd_tlb_entry(tlb, pmdp, address);		\
 	} while (0)
 
+/*
+ * For things like page tables caches (ie caching addresses "inside" the
+ * page tables, like x86 does), for legacy reasons, flushing an
+ * individual page had better flush the page table caches behind it. This
+ * is definitely how x86 works, for example. And if you have an
+ * architected non-legacy page table cache (which I'm not aware of
+ * anybody actually doing), you're going to have some architecturally
+ * explicit flushing for that, likely *separate* from a regular TLB entry
+ * flush, and thus you'd need more than just some range expansion..
+ *
+ * So if we ever find an architecture
+ * that would want something that odd, I think it is up to that
+ * architecture to do its own odd thing, not cause pain for others
+ * http://lkml.kernel.org/r/CA+55aFzBggoXtNXQeng5d_mRoDnaMBE5Y+URs+PHR67nUpMtaw@mail.gmail.com
+ *
+ * For now w.r.t page table cache, mark the range_size as PAGE_SIZE
+ */
+
 #define pte_free_tlb(tlb, ptep, address)			\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__pte_free_tlb(tlb, ptep, address);		\
 	} while (0)
 
 #ifndef __ARCH_HAS_4LEVEL_HACK
 #define pud_free_tlb(tlb, pudp, address)			\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__pud_free_tlb(tlb, pudp, address);		\
 	} while (0)
 #endif
 
 #define pmd_free_tlb(tlb, pmdp, address)			\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__pmd_free_tlb(tlb, pmdp, address);		\
 	} while (0)
 
-- 
2.10.1

[PATCH 2/5] mm: Update mmu_gather range correctly

[Aneesh Kumar K.V]

We use __tlb_adjust_range to update range convered by mmu_gather struct.
We later use the 'start' and 'end' to do a mmu_notifier_invalidate_range
in tlb_flush_mmu_tlbonly(). Update the 'end' correctly in
__tlb_adjust_range so that we call mmu_notifier_invalidate_range with the
correct range values.

W.r.t to tlbflush, this should not have any impact, because a flush with
correct start address will flush tlb mapping for the range.

Also add comment w.r.t updating the range when we free pagetable pages.
For now we don't support a range based page table cache flush.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 include/asm-generic/tlb.h | 43 +++++++++++++++++++++++++++++++------------
 1 file changed, 31 insertions(+), 12 deletions(-)

diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index c6d667187608..dba727becd5f 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -125,10 +125,11 @@ extern bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page,
 				   int page_size);
 
 static inline void __tlb_adjust_range(struct mmu_gather *tlb,
-				      unsigned long address)
+				      unsigned long address,
+				      unsigned int range_size)
 {
 	tlb->start = min(tlb->start, address);
-	tlb->end = max(tlb->end, address + PAGE_SIZE);
+	tlb->end = max(tlb->end, address + range_size);
 	/*
 	 * Track the last address with which we adjusted the range. This
 	 * will be used later to adjust again after a mmu_flush due to
@@ -153,7 +154,7 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 	if (__tlb_remove_page_size(tlb, page, page_size)) {
 		tlb_flush_mmu(tlb);
 		tlb->page_size = page_size;
-		__tlb_adjust_range(tlb, tlb->addr);
+		__tlb_adjust_range(tlb, tlb->addr, page_size);
 		__tlb_remove_page_size(tlb, page, page_size);
 	}
 }
@@ -177,7 +178,7 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 	/* active->nr should be zero when we call this */
 	VM_BUG_ON_PAGE(tlb->active->nr, page);
 	tlb->page_size = PAGE_SIZE;
-	__tlb_adjust_range(tlb, tlb->addr);
+	__tlb_adjust_range(tlb, tlb->addr, PAGE_SIZE);
 	return __tlb_remove_page(tlb, page);
 }
 
@@ -215,7 +216,7 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
  */
 #define tlb_remove_tlb_entry(tlb, ptep, address)		\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
@@ -227,29 +228,47 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 #define __tlb_remove_pmd_tlb_entry(tlb, pmdp, address) do {} while (0)
 #endif
 
-#define tlb_remove_pmd_tlb_entry(tlb, pmdp, address)		\
-	do {							\
-		__tlb_adjust_range(tlb, address);		\
-		__tlb_remove_pmd_tlb_entry(tlb, pmdp, address);	\
+#define tlb_remove_pmd_tlb_entry(tlb, pmdp, address)			\
+	do {								\
+		__tlb_adjust_range(tlb, address, HPAGE_PMD_SIZE);	\
+		__tlb_remove_pmd_tlb_entry(tlb, pmdp, address);		\
 	} while (0)
 
+/*
+ * For things like page tables caches (ie caching addresses "inside" the
+ * page tables, like x86 does), for legacy reasons, flushing an
+ * individual page had better flush the page table caches behind it. This
+ * is definitely how x86 works, for example. And if you have an
+ * architected non-legacy page table cache (which I'm not aware of
+ * anybody actually doing), you're going to have some architecturally
+ * explicit flushing for that, likely *separate* from a regular TLB entry
+ * flush, and thus you'd need more than just some range expansion..
+ *
+ * So if we ever find an architecture
+ * that would want something that odd, I think it is up to that
+ * architecture to do its own odd thing, not cause pain for others
+ * http://lkml.kernel.org/r/CA+55aFzBggoXtNXQeng5d_mRoDnaMBE5Y+URs+PHR67nUpMtaw@mail.gmail.com
+ *
+ * For now w.r.t page table cache, mark the range_size as PAGE_SIZE
+ */
+
 #define pte_free_tlb(tlb, ptep, address)			\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__pte_free_tlb(tlb, ptep, address);		\
 	} while (0)
 
 #ifndef __ARCH_HAS_4LEVEL_HACK
 #define pud_free_tlb(tlb, pudp, address)			\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__pud_free_tlb(tlb, pudp, address);		\
 	} while (0)
 #endif
 
 #define pmd_free_tlb(tlb, pmdp, address)			\
 	do {							\
-		__tlb_adjust_range(tlb, address);		\
+		__tlb_adjust_range(tlb, address, PAGE_SIZE);	\
 		__pmd_free_tlb(tlb, pmdp, address);		\
 	} while (0)
 
-- 
2.10.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 3/5] mm/hugetlb: add tlb_remove_hugetlb_entry for handling hugetlb pages

[Aneesh Kumar K.V]

This add tlb_remove_hugetlb_entry similar to tlb_remove_pmd_tlb_entry.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 arch/arm/include/asm/tlb.h  | 2 ++
 arch/ia64/include/asm/tlb.h | 3 +++
 arch/s390/include/asm/tlb.h | 2 ++
 arch/sh/include/asm/tlb.h   | 3 +++
 arch/um/include/asm/tlb.h   | 3 +++
 include/asm-generic/tlb.h   | 6 ++++++
 mm/hugetlb.c                | 2 +-
 7 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h
index 1e25cd80589e..82841ba1f51f 100644
--- a/arch/arm/include/asm/tlb.h
+++ b/arch/arm/include/asm/tlb.h
@@ -186,6 +186,8 @@ tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep, unsigned long addr)
 	tlb_add_flush(tlb, addr);
 }
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
 /*
  * 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,
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index 77e541cf0e5d..b3f369ab844d 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -283,6 +283,9 @@ do {							\
 	__tlb_remove_tlb_entry(tlb, ptep, addr);	\
 } while (0)
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
+
 #define pte_free_tlb(tlb, ptep, address)		\
 do {							\
 	tlb->need_flush = 1;				\
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 15711de10403..094440b59f9e 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -162,5 +162,7 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud,
 #define tlb_remove_tlb_entry(tlb, ptep, addr)	do { } while (0)
 #define tlb_remove_pmd_tlb_entry(tlb, pmdp, addr)	do { } while (0)
 #define tlb_migrate_finish(mm)			do { } while (0)
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
 
 #endif /* _S390_TLB_H */
diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h
index 025cdb1032f6..e7d15e8c75c1 100644
--- a/arch/sh/include/asm/tlb.h
+++ b/arch/sh/include/asm/tlb.h
@@ -65,6 +65,9 @@ tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep, unsigned long address)
 		tlb->end = address + PAGE_SIZE;
 }
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
+
 /*
  * 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,
diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h
index 821ff0acfe17..a4427029c3c8 100644
--- a/arch/um/include/asm/tlb.h
+++ b/arch/um/include/asm/tlb.h
@@ -141,6 +141,9 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
+
 #define pte_free_tlb(tlb, ptep, addr) __pte_free_tlb(tlb, ptep, addr)
 
 #define pud_free_tlb(tlb, pudp, addr) __pud_free_tlb(tlb, pudp, addr)
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index dba727becd5f..38c2b708df6e 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -220,6 +220,12 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	     \
+	do {							     \
+		__tlb_adjust_range(tlb, address, huge_page_size(h)); \
+		__tlb_remove_tlb_entry(tlb, ptep, address);	     \
+	} while (0)
+
 /**
  * tlb_remove_pmd_tlb_entry - remember a pmd mapping for later tlb invalidation
  * This is a nop so far, because only x86 needs it.
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index ec49d9ef1eef..655d3cf05551 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3272,7 +3272,7 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 		}
 
 		pte = huge_ptep_get_and_clear(mm, address, ptep);
-		tlb_remove_tlb_entry(tlb, ptep, address);
+		tlb_remove_huge_tlb_entry(h, tlb, ptep, address);
 		if (huge_pte_dirty(pte))
 			set_page_dirty(page);
 
-- 
2.10.1

[PATCH 3/5] mm/hugetlb: add tlb_remove_hugetlb_entry for handling hugetlb pages

[Aneesh Kumar K.V]

This add tlb_remove_hugetlb_entry similar to tlb_remove_pmd_tlb_entry.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 arch/arm/include/asm/tlb.h  | 2 ++
 arch/ia64/include/asm/tlb.h | 3 +++
 arch/s390/include/asm/tlb.h | 2 ++
 arch/sh/include/asm/tlb.h   | 3 +++
 arch/um/include/asm/tlb.h   | 3 +++
 include/asm-generic/tlb.h   | 6 ++++++
 mm/hugetlb.c                | 2 +-
 7 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h
index 1e25cd80589e..82841ba1f51f 100644
--- a/arch/arm/include/asm/tlb.h
+++ b/arch/arm/include/asm/tlb.h
@@ -186,6 +186,8 @@ tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep, unsigned long addr)
 	tlb_add_flush(tlb, addr);
 }
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
 /*
  * 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,
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index 77e541cf0e5d..b3f369ab844d 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -283,6 +283,9 @@ do {							\
 	__tlb_remove_tlb_entry(tlb, ptep, addr);	\
 } while (0)
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
+
 #define pte_free_tlb(tlb, ptep, address)		\
 do {							\
 	tlb->need_flush = 1;				\
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 15711de10403..094440b59f9e 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -162,5 +162,7 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud,
 #define tlb_remove_tlb_entry(tlb, ptep, addr)	do { } while (0)
 #define tlb_remove_pmd_tlb_entry(tlb, pmdp, addr)	do { } while (0)
 #define tlb_migrate_finish(mm)			do { } while (0)
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
 
 #endif /* _S390_TLB_H */
diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h
index 025cdb1032f6..e7d15e8c75c1 100644
--- a/arch/sh/include/asm/tlb.h
+++ b/arch/sh/include/asm/tlb.h
@@ -65,6 +65,9 @@ tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep, unsigned long address)
 		tlb->end = address + PAGE_SIZE;
 }
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
+
 /*
  * 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,
diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h
index 821ff0acfe17..a4427029c3c8 100644
--- a/arch/um/include/asm/tlb.h
+++ b/arch/um/include/asm/tlb.h
@@ -141,6 +141,9 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
+	tlb_remove_tlb_entry(tlb, ptep, address)
+
 #define pte_free_tlb(tlb, ptep, addr) __pte_free_tlb(tlb, ptep, addr)
 
 #define pud_free_tlb(tlb, pudp, addr) __pud_free_tlb(tlb, pudp, addr)
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index dba727becd5f..38c2b708df6e 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -220,6 +220,12 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 		__tlb_remove_tlb_entry(tlb, ptep, address);	\
 	} while (0)
 
+#define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	     \
+	do {							     \
+		__tlb_adjust_range(tlb, address, huge_page_size(h)); \
+		__tlb_remove_tlb_entry(tlb, ptep, address);	     \
+	} while (0)
+
 /**
  * tlb_remove_pmd_tlb_entry - remember a pmd mapping for later tlb invalidation
  * This is a nop so far, because only x86 needs it.
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index ec49d9ef1eef..655d3cf05551 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3272,7 +3272,7 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 		}
 
 		pte = huge_ptep_get_and_clear(mm, address, ptep);
-		tlb_remove_tlb_entry(tlb, ptep, address);
+		tlb_remove_huge_tlb_entry(h, tlb, ptep, address);
 		if (huge_pte_dirty(pte))
 			set_page_dirty(page);
 
-- 
2.10.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 4/5] mm: Add tlb_remove_check_page_size_change to track page size change

[Aneesh Kumar K.V]

With commit e77b0852b551 ("mm/mmu_gather: track page size with mmu gather and force flush if page size change")
we added the ability to force a tlb flush when the page size change in
a mmu_gather loop. We did that by checking for a page size change
every time we added a page to mmu_gather for lazy flush/remove. We can
improve that by moving the page size change check early and not doing
it every time we add a page.

This also help us to do tlb flush when invalidating a range covering
dax mapping. W.r.t dax mapping we don't have a backing struct page and
hence we don't call tlb_remove_page, which earlier forced the tlb flush
on page size change. Moving the page size change check earlier means
we will do the same even for dax mapping.

We also avoid doing this check on architecture other than powerpc.

In the later patch we will remove page size check from tlb_remove_page().

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 arch/arm/include/asm/tlb.h     |  6 ++++++
 arch/ia64/include/asm/tlb.h    |  6 ++++++
 arch/powerpc/include/asm/tlb.h | 16 ++++++++++++++++
 arch/s390/include/asm/tlb.h    |  6 ++++++
 arch/sh/include/asm/tlb.h      |  6 ++++++
 arch/um/include/asm/tlb.h      |  6 ++++++
 include/asm-generic/tlb.h      | 16 ++++++++++++++++
 mm/huge_memory.c               |  4 ++++
 mm/hugetlb.c                   |  5 +++++
 mm/madvise.c                   |  1 +
 mm/memory.c                    |  7 ++++++-
 11 files changed, 78 insertions(+), 1 deletion(-)

diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h
index 82841ba1f51f..a9d6de4746ea 100644
--- a/arch/arm/include/asm/tlb.h
+++ b/arch/arm/include/asm/tlb.h
@@ -286,5 +286,11 @@ tlb_remove_pmd_tlb_entry(struct mmu_gather *tlb, pmd_t *pmdp, unsigned long addr
 
 #define tlb_migrate_finish(mm)		do { } while (0)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #endif /* CONFIG_MMU */
 #endif
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index b3f369ab844d..bfe6295aa746 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -286,6 +286,12 @@ do {							\
 #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
 	tlb_remove_tlb_entry(tlb, ptep, address)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #define pte_free_tlb(tlb, ptep, address)		\
 do {							\
 	tlb->need_flush = 1;				\
diff --git a/arch/powerpc/include/asm/tlb.h b/arch/powerpc/include/asm/tlb.h
index f6f68f73e858..c54e9f2cec87 100644
--- a/arch/powerpc/include/asm/tlb.h
+++ b/arch/powerpc/include/asm/tlb.h
@@ -28,6 +28,7 @@
 #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
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
 
 extern void tlb_flush(struct mmu_gather *tlb);
 
@@ -46,6 +47,21 @@ static inline void __tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep,
 #endif
 }
 
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+	if (!tlb->page_size)
+		tlb->page_size = page_size;
+	else if (tlb->page_size != page_size) {
+		tlb_flush_mmu(tlb);
+		/*
+		 * update the page size after flush for the new
+		 * mmu_gather.
+		 */
+		tlb->page_size = page_size;
+	}
+}
+
 #ifdef CONFIG_SMP
 static inline int mm_is_core_local(struct mm_struct *mm)
 {
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 094440b59f9e..28b159c87c38 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -165,4 +165,10 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud,
 #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
 	tlb_remove_tlb_entry(tlb, ptep, address)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #endif /* _S390_TLB_H */
diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h
index e7d15e8c75c1..0f988b3e484b 100644
--- a/arch/sh/include/asm/tlb.h
+++ b/arch/sh/include/asm/tlb.h
@@ -130,6 +130,12 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 	return tlb_remove_page(tlb, page);
 }
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #define pte_free_tlb(tlb, ptep, addr)	pte_free((tlb)->mm, ptep)
 #define pmd_free_tlb(tlb, pmdp, addr)	pmd_free((tlb)->mm, pmdp)
 #define pud_free_tlb(tlb, pudp, addr)	pud_free((tlb)->mm, pudp)
diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h
index a4427029c3c8..8258dd4bb13c 100644
--- a/arch/um/include/asm/tlb.h
+++ b/arch/um/include/asm/tlb.h
@@ -144,6 +144,12 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
 	tlb_remove_tlb_entry(tlb, ptep, address)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #define pte_free_tlb(tlb, ptep, addr) __pte_free_tlb(tlb, ptep, addr)
 
 #define pud_free_tlb(tlb, pudp, addr) __pud_free_tlb(tlb, pudp, addr)
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index 38c2b708df6e..256c9de71fdb 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -182,6 +182,22 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 	return __tlb_remove_page(tlb, page);
 }
 
+#ifndef tlb_remove_check_page_size_change
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+	/*
+	 * We don't care about page size change, just update
+	 * mmu_gather page size here so that debug checks
+	 * doesn't throw false warning.
+	 */
+#ifdef CONFIG_DEBUG_VM
+	tlb->page_size = page_size;
+#endif
+}
+#endif
+
 /*
  * 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,
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 12c6685d4e97..2d1d6bb9534b 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1322,6 +1322,8 @@ bool madvise_free_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	struct mm_struct *mm = tlb->mm;
 	bool ret = false;
 
+	tlb_remove_check_page_size_change(tlb, HPAGE_PMD_SIZE);
+
 	ptl = pmd_trans_huge_lock(pmd, vma);
 	if (!ptl)
 		goto out_unlocked;
@@ -1383,6 +1385,8 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	pmd_t orig_pmd;
 	spinlock_t *ptl;
 
+	tlb_remove_check_page_size_change(tlb, HPAGE_PMD_SIZE);
+
 	ptl = __pmd_trans_huge_lock(pmd, vma);
 	if (!ptl)
 		return 0;
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 655d3cf05551..958ba1d87d20 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3222,6 +3222,11 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	BUG_ON(start & ~huge_page_mask(h));
 	BUG_ON(end & ~huge_page_mask(h));
 
+	/*
+	 * This is a hugetlb vma, all the pte entries should point
+	 * to huge page.
+	 */
+	tlb_remove_check_page_size_change(tlb, sz);
 	tlb_start_vma(tlb, vma);
 	mmu_notifier_invalidate_range_start(mm, mmun_start, mmun_end);
 	address = start;
diff --git a/mm/madvise.c b/mm/madvise.c
index 93fb63e88b5e..0e3828eae9f8 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -281,6 +281,7 @@ static int madvise_free_pte_range(pmd_t *pmd, unsigned long addr,
 	if (pmd_trans_unstable(pmd))
 		return 0;
 
+	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 	orig_pte = pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
 	arch_enter_lazy_mmu_mode();
 	for (; addr != end; pte++, addr += PAGE_SIZE) {
diff --git a/mm/memory.c b/mm/memory.c
index e18c57bdc75c..40752dc7750f 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -528,7 +528,11 @@ void free_pgd_range(struct mmu_gather *tlb,
 		end -= PMD_SIZE;
 	if (addr > end - 1)
 		return;
-
+	/*
+	 * We add page table cache pages with PAGE_SIZE,
+	 * (see pte_free_tlb()), flush the tlb if we need
+	 */
+	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 	pgd = pgd_offset(tlb->mm, addr);
 	do {
 		next = pgd_addr_end(addr, end);
@@ -1120,6 +1124,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	swp_entry_t entry;
 	struct page *pending_page = NULL;
 
+	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 again:
 	init_rss_vec(rss);
 	start_pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
-- 
2.10.1

[PATCH 4/5] mm: Add tlb_remove_check_page_size_change to track page size change

[Aneesh Kumar K.V]

With commit e77b0852b551 ("mm/mmu_gather: track page size with mmu gather and force flush if page size change")
we added the ability to force a tlb flush when the page size change in
a mmu_gather loop. We did that by checking for a page size change
every time we added a page to mmu_gather for lazy flush/remove. We can
improve that by moving the page size change check early and not doing
it every time we add a page.

This also help us to do tlb flush when invalidating a range covering
dax mapping. W.r.t dax mapping we don't have a backing struct page and
hence we don't call tlb_remove_page, which earlier forced the tlb flush
on page size change. Moving the page size change check earlier means
we will do the same even for dax mapping.

We also avoid doing this check on architecture other than powerpc.

In the later patch we will remove page size check from tlb_remove_page().

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 arch/arm/include/asm/tlb.h     |  6 ++++++
 arch/ia64/include/asm/tlb.h    |  6 ++++++
 arch/powerpc/include/asm/tlb.h | 16 ++++++++++++++++
 arch/s390/include/asm/tlb.h    |  6 ++++++
 arch/sh/include/asm/tlb.h      |  6 ++++++
 arch/um/include/asm/tlb.h      |  6 ++++++
 include/asm-generic/tlb.h      | 16 ++++++++++++++++
 mm/huge_memory.c               |  4 ++++
 mm/hugetlb.c                   |  5 +++++
 mm/madvise.c                   |  1 +
 mm/memory.c                    |  7 ++++++-
 11 files changed, 78 insertions(+), 1 deletion(-)

diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h
index 82841ba1f51f..a9d6de4746ea 100644
--- a/arch/arm/include/asm/tlb.h
+++ b/arch/arm/include/asm/tlb.h
@@ -286,5 +286,11 @@ tlb_remove_pmd_tlb_entry(struct mmu_gather *tlb, pmd_t *pmdp, unsigned long addr
 
 #define tlb_migrate_finish(mm)		do { } while (0)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #endif /* CONFIG_MMU */
 #endif
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index b3f369ab844d..bfe6295aa746 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -286,6 +286,12 @@ do {							\
 #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
 	tlb_remove_tlb_entry(tlb, ptep, address)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #define pte_free_tlb(tlb, ptep, address)		\
 do {							\
 	tlb->need_flush = 1;				\
diff --git a/arch/powerpc/include/asm/tlb.h b/arch/powerpc/include/asm/tlb.h
index f6f68f73e858..c54e9f2cec87 100644
--- a/arch/powerpc/include/asm/tlb.h
+++ b/arch/powerpc/include/asm/tlb.h
@@ -28,6 +28,7 @@
 #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
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
 
 extern void tlb_flush(struct mmu_gather *tlb);
 
@@ -46,6 +47,21 @@ static inline void __tlb_remove_tlb_entry(struct mmu_gather *tlb, pte_t *ptep,
 #endif
 }
 
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+	if (!tlb->page_size)
+		tlb->page_size = page_size;
+	else if (tlb->page_size != page_size) {
+		tlb_flush_mmu(tlb);
+		/*
+		 * update the page size after flush for the new
+		 * mmu_gather.
+		 */
+		tlb->page_size = page_size;
+	}
+}
+
 #ifdef CONFIG_SMP
 static inline int mm_is_core_local(struct mm_struct *mm)
 {
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 094440b59f9e..28b159c87c38 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -165,4 +165,10 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud,
 #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
 	tlb_remove_tlb_entry(tlb, ptep, address)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #endif /* _S390_TLB_H */
diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h
index e7d15e8c75c1..0f988b3e484b 100644
--- a/arch/sh/include/asm/tlb.h
+++ b/arch/sh/include/asm/tlb.h
@@ -130,6 +130,12 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 	return tlb_remove_page(tlb, page);
 }
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #define pte_free_tlb(tlb, ptep, addr)	pte_free((tlb)->mm, ptep)
 #define pmd_free_tlb(tlb, pmdp, addr)	pmd_free((tlb)->mm, pmdp)
 #define pud_free_tlb(tlb, pudp, addr)	pud_free((tlb)->mm, pudp)
diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h
index a4427029c3c8..8258dd4bb13c 100644
--- a/arch/um/include/asm/tlb.h
+++ b/arch/um/include/asm/tlb.h
@@ -144,6 +144,12 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address)	\
 	tlb_remove_tlb_entry(tlb, ptep, address)
 
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+}
+
 #define pte_free_tlb(tlb, ptep, addr) __pte_free_tlb(tlb, ptep, addr)
 
 #define pud_free_tlb(tlb, pudp, addr) __pud_free_tlb(tlb, pudp, addr)
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index 38c2b708df6e..256c9de71fdb 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -182,6 +182,22 @@ static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *pa
 	return __tlb_remove_page(tlb, page);
 }
 
+#ifndef tlb_remove_check_page_size_change
+#define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
+static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
+						     unsigned int page_size)
+{
+	/*
+	 * We don't care about page size change, just update
+	 * mmu_gather page size here so that debug checks
+	 * doesn't throw false warning.
+	 */
+#ifdef CONFIG_DEBUG_VM
+	tlb->page_size = page_size;
+#endif
+}
+#endif
+
 /*
  * 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,
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 12c6685d4e97..2d1d6bb9534b 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1322,6 +1322,8 @@ bool madvise_free_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	struct mm_struct *mm = tlb->mm;
 	bool ret = false;
 
+	tlb_remove_check_page_size_change(tlb, HPAGE_PMD_SIZE);
+
 	ptl = pmd_trans_huge_lock(pmd, vma);
 	if (!ptl)
 		goto out_unlocked;
@@ -1383,6 +1385,8 @@ int zap_huge_pmd(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	pmd_t orig_pmd;
 	spinlock_t *ptl;
 
+	tlb_remove_check_page_size_change(tlb, HPAGE_PMD_SIZE);
+
 	ptl = __pmd_trans_huge_lock(pmd, vma);
 	if (!ptl)
 		return 0;
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 655d3cf05551..958ba1d87d20 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3222,6 +3222,11 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	BUG_ON(start & ~huge_page_mask(h));
 	BUG_ON(end & ~huge_page_mask(h));
 
+	/*
+	 * This is a hugetlb vma, all the pte entries should point
+	 * to huge page.
+	 */
+	tlb_remove_check_page_size_change(tlb, sz);
 	tlb_start_vma(tlb, vma);
 	mmu_notifier_invalidate_range_start(mm, mmun_start, mmun_end);
 	address = start;
diff --git a/mm/madvise.c b/mm/madvise.c
index 93fb63e88b5e..0e3828eae9f8 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -281,6 +281,7 @@ static int madvise_free_pte_range(pmd_t *pmd, unsigned long addr,
 	if (pmd_trans_unstable(pmd))
 		return 0;
 
+	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 	orig_pte = pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
 	arch_enter_lazy_mmu_mode();
 	for (; addr != end; pte++, addr += PAGE_SIZE) {
diff --git a/mm/memory.c b/mm/memory.c
index e18c57bdc75c..40752dc7750f 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -528,7 +528,11 @@ void free_pgd_range(struct mmu_gather *tlb,
 		end -= PMD_SIZE;
 	if (addr > end - 1)
 		return;
-
+	/*
+	 * We add page table cache pages with PAGE_SIZE,
+	 * (see pte_free_tlb()), flush the tlb if we need
+	 */
+	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 	pgd = pgd_offset(tlb->mm, addr);
 	do {
 		next = pgd_addr_end(addr, end);
@@ -1120,6 +1124,7 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	swp_entry_t entry;
 	struct page *pending_page = NULL;
 
+	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 again:
 	init_rss_vec(rss);
 	start_pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
-- 
2.10.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[PATCH 5/5] mm: Remove the page size change check in tlb_remove_page

[Aneesh Kumar K.V]

Now that we check for page size change early in the loop, we can
partially revert e9d55e157034a9efd99405c99c1565d64619d82b
("mm: change the interface for __tlb_remove_page").

This simplies the code much, by removing the need to track the last
address with which we adjusted the range. We also go back to the older
way of filling the mmu_gather array, ie, we add an entry and then
check whether the gather batch is full.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 arch/arm/include/asm/tlb.h  | 13 +++----------
 arch/ia64/include/asm/tlb.h | 16 ++++------------
 arch/s390/include/asm/tlb.h |  6 ------
 arch/sh/include/asm/tlb.h   |  6 ------
 arch/um/include/asm/tlb.h   |  6 ------
 include/asm-generic/tlb.h   | 28 ++--------------------------
 mm/memory.c                 | 21 ++++++---------------
 7 files changed, 15 insertions(+), 81 deletions(-)

diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h
index a9d6de4746ea..3f2eb76243e3 100644
--- a/arch/arm/include/asm/tlb.h
+++ b/arch/arm/include/asm/tlb.h
@@ -213,18 +213,17 @@ tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma)
 
 static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
+	tlb->pages[tlb->nr++] = page;
+	VM_WARN_ON(tlb->nr > tlb->max);
 	if (tlb->nr == tlb->max)
 		return true;
-	tlb->pages[tlb->nr++] = page;
 	return false;
 }
 
 static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
-	if (__tlb_remove_page(tlb, page)) {
+	if (__tlb_remove_page(tlb, page))
 		tlb_flush_mmu(tlb);
-		__tlb_remove_page(tlb, page);
-	}
 }
 
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
@@ -233,12 +232,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index bfe6295aa746..fced197b9626 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -207,15 +207,15 @@ tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
  */
 static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
-	if (tlb->nr == tlb->max)
-		return true;
-
 	tlb->need_flush = 1;
 
 	if (!tlb->nr && tlb->pages == tlb->local)
 		__tlb_alloc_page(tlb);
 
 	tlb->pages[tlb->nr++] = page;
+	VM_WARN_ON(tlb->nr > tlb->max);
+	if (tlb->nr == tlb->max)
+		return true;
 	return false;
 }
 
@@ -236,10 +236,8 @@ static inline void tlb_flush_mmu(struct mmu_gather *tlb)
 
 static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
-	if (__tlb_remove_page(tlb, page)) {
+	if (__tlb_remove_page(tlb, page))
 		tlb_flush_mmu(tlb);
-		__tlb_remove_page(tlb, page);
-	}
 }
 
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
@@ -248,12 +246,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 28b159c87c38..853b2a3d8dee 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -104,12 +104,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h
index 0f988b3e484b..46e0d635e36f 100644
--- a/arch/sh/include/asm/tlb.h
+++ b/arch/sh/include/asm/tlb.h
@@ -118,12 +118,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h
index 8258dd4bb13c..600a2e9bfee2 100644
--- a/arch/um/include/asm/tlb.h
+++ b/arch/um/include/asm/tlb.h
@@ -116,12 +116,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index 256c9de71fdb..7eed8cf3130a 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -107,11 +107,6 @@ struct mmu_gather {
 	struct mmu_gather_batch	local;
 	struct page		*__pages[MMU_GATHER_BUNDLE];
 	unsigned int		batch_count;
-	/*
-	 * __tlb_adjust_range  will track the new addr here,
-	 * that that we can adjust the range after the flush
-	 */
-	unsigned long addr;
 	int page_size;
 };
 
@@ -130,12 +125,6 @@ static inline void __tlb_adjust_range(struct mmu_gather *tlb,
 {
 	tlb->start = min(tlb->start, address);
 	tlb->end = max(tlb->end, address + range_size);
-	/*
-	 * Track the last address with which we adjusted the range. This
-	 * will be used later to adjust again after a mmu_flush due to
-	 * failed __tlb_remove_page
-	 */
-	tlb->addr = address;
 }
 
 static inline void __tlb_reset_range(struct mmu_gather *tlb)
@@ -151,15 +140,11 @@ static inline void __tlb_reset_range(struct mmu_gather *tlb)
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
-	if (__tlb_remove_page_size(tlb, page, page_size)) {
+	if (__tlb_remove_page_size(tlb, page, page_size))
 		tlb_flush_mmu(tlb);
-		tlb->page_size = page_size;
-		__tlb_adjust_range(tlb, tlb->addr, page_size);
-		__tlb_remove_page_size(tlb, page, page_size);
-	}
 }
 
-static bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
+static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
 	return __tlb_remove_page_size(tlb, page, PAGE_SIZE);
 }
@@ -173,15 +158,6 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 	return tlb_remove_page_size(tlb, page, PAGE_SIZE);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *page)
-{
-	/* active->nr should be zero when we call this */
-	VM_BUG_ON_PAGE(tlb->active->nr, page);
-	tlb->page_size = PAGE_SIZE;
-	__tlb_adjust_range(tlb, tlb->addr, PAGE_SIZE);
-	return __tlb_remove_page(tlb, page);
-}
-
 #ifndef tlb_remove_check_page_size_change
 #define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
 static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
diff --git a/mm/memory.c b/mm/memory.c
index 40752dc7750f..371604f08eb2 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -300,15 +300,14 @@ bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page, int page_
 	struct mmu_gather_batch *batch;
 
 	VM_BUG_ON(!tlb->end);
-
-	if (!tlb->page_size)
-		tlb->page_size = page_size;
-	else {
-		if (page_size != tlb->page_size)
-			return true;
-	}
+	VM_WARN_ON(tlb->page_size != page_size);
 
 	batch = tlb->active;
+	/*
+	 * Add the page and check if we are full. If so
+	 * force a flush.
+	 */
+	batch->pages[batch->nr++] = page;
 	if (batch->nr == batch->max) {
 		if (!tlb_next_batch(tlb))
 			return true;
@@ -316,7 +315,6 @@ bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page, int page_
 	}
 	VM_BUG_ON_PAGE(batch->nr > batch->max, page);
 
-	batch->pages[batch->nr++] = page;
 	return false;
 }
 
@@ -1122,7 +1120,6 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	pte_t *start_pte;
 	pte_t *pte;
 	swp_entry_t entry;
-	struct page *pending_page = NULL;
 
 	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 again:
@@ -1177,7 +1174,6 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				print_bad_pte(vma, addr, ptent, page);
 			if (unlikely(__tlb_remove_page(tlb, page))) {
 				force_flush = 1;
-				pending_page = page;
 				addr += PAGE_SIZE;
 				break;
 			}
@@ -1218,11 +1214,6 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	if (force_flush) {
 		force_flush = 0;
 		tlb_flush_mmu_free(tlb);
-		if (pending_page) {
-			/* remove the page with new size */
-			__tlb_remove_pte_page(tlb, pending_page);
-			pending_page = NULL;
-		}
 		if (addr != end)
 			goto again;
 	}
-- 
2.10.1

[PATCH 5/5] mm: Remove the page size change check in tlb_remove_page

[Aneesh Kumar K.V]

Now that we check for page size change early in the loop, we can
partially revert e9d55e157034a9efd99405c99c1565d64619d82b
("mm: change the interface for __tlb_remove_page").

This simplies the code much, by removing the need to track the last
address with which we adjusted the range. We also go back to the older
way of filling the mmu_gather array, ie, we add an entry and then
check whether the gather batch is full.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
---
 arch/arm/include/asm/tlb.h  | 13 +++----------
 arch/ia64/include/asm/tlb.h | 16 ++++------------
 arch/s390/include/asm/tlb.h |  6 ------
 arch/sh/include/asm/tlb.h   |  6 ------
 arch/um/include/asm/tlb.h   |  6 ------
 include/asm-generic/tlb.h   | 28 ++--------------------------
 mm/memory.c                 | 21 ++++++---------------
 7 files changed, 15 insertions(+), 81 deletions(-)

diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h
index a9d6de4746ea..3f2eb76243e3 100644
--- a/arch/arm/include/asm/tlb.h
+++ b/arch/arm/include/asm/tlb.h
@@ -213,18 +213,17 @@ tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma)
 
 static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
+	tlb->pages[tlb->nr++] = page;
+	VM_WARN_ON(tlb->nr > tlb->max);
 	if (tlb->nr == tlb->max)
 		return true;
-	tlb->pages[tlb->nr++] = page;
 	return false;
 }
 
 static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
-	if (__tlb_remove_page(tlb, page)) {
+	if (__tlb_remove_page(tlb, page))
 		tlb_flush_mmu(tlb);
-		__tlb_remove_page(tlb, page);
-	}
 }
 
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
@@ -233,12 +232,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h
index bfe6295aa746..fced197b9626 100644
--- a/arch/ia64/include/asm/tlb.h
+++ b/arch/ia64/include/asm/tlb.h
@@ -207,15 +207,15 @@ tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
  */
 static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
-	if (tlb->nr == tlb->max)
-		return true;
-
 	tlb->need_flush = 1;
 
 	if (!tlb->nr && tlb->pages == tlb->local)
 		__tlb_alloc_page(tlb);
 
 	tlb->pages[tlb->nr++] = page;
+	VM_WARN_ON(tlb->nr > tlb->max);
+	if (tlb->nr == tlb->max)
+		return true;
 	return false;
 }
 
@@ -236,10 +236,8 @@ static inline void tlb_flush_mmu(struct mmu_gather *tlb)
 
 static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
-	if (__tlb_remove_page(tlb, page)) {
+	if (__tlb_remove_page(tlb, page))
 		tlb_flush_mmu(tlb);
-		__tlb_remove_page(tlb, page);
-	}
 }
 
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
@@ -248,12 +246,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index 28b159c87c38..853b2a3d8dee 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -104,12 +104,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h
index 0f988b3e484b..46e0d635e36f 100644
--- a/arch/sh/include/asm/tlb.h
+++ b/arch/sh/include/asm/tlb.h
@@ -118,12 +118,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h
index 8258dd4bb13c..600a2e9bfee2 100644
--- a/arch/um/include/asm/tlb.h
+++ b/arch/um/include/asm/tlb.h
@@ -116,12 +116,6 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 	return __tlb_remove_page(tlb, page);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb,
-					 struct page *page)
-{
-	return __tlb_remove_page(tlb, page);
-}
-
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index 256c9de71fdb..7eed8cf3130a 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -107,11 +107,6 @@ struct mmu_gather {
 	struct mmu_gather_batch	local;
 	struct page		*__pages[MMU_GATHER_BUNDLE];
 	unsigned int		batch_count;
-	/*
-	 * __tlb_adjust_range  will track the new addr here,
-	 * that that we can adjust the range after the flush
-	 */
-	unsigned long addr;
 	int page_size;
 };
 
@@ -130,12 +125,6 @@ static inline void __tlb_adjust_range(struct mmu_gather *tlb,
 {
 	tlb->start = min(tlb->start, address);
 	tlb->end = max(tlb->end, address + range_size);
-	/*
-	 * Track the last address with which we adjusted the range. This
-	 * will be used later to adjust again after a mmu_flush due to
-	 * failed __tlb_remove_page
-	 */
-	tlb->addr = address;
 }
 
 static inline void __tlb_reset_range(struct mmu_gather *tlb)
@@ -151,15 +140,11 @@ static inline void __tlb_reset_range(struct mmu_gather *tlb)
 static inline void tlb_remove_page_size(struct mmu_gather *tlb,
 					struct page *page, int page_size)
 {
-	if (__tlb_remove_page_size(tlb, page, page_size)) {
+	if (__tlb_remove_page_size(tlb, page, page_size))
 		tlb_flush_mmu(tlb);
-		tlb->page_size = page_size;
-		__tlb_adjust_range(tlb, tlb->addr, page_size);
-		__tlb_remove_page_size(tlb, page, page_size);
-	}
 }
 
-static bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
+static inline bool __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
 	return __tlb_remove_page_size(tlb, page, PAGE_SIZE);
 }
@@ -173,15 +158,6 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 	return tlb_remove_page_size(tlb, page, PAGE_SIZE);
 }
 
-static inline bool __tlb_remove_pte_page(struct mmu_gather *tlb, struct page *page)
-{
-	/* active->nr should be zero when we call this */
-	VM_BUG_ON_PAGE(tlb->active->nr, page);
-	tlb->page_size = PAGE_SIZE;
-	__tlb_adjust_range(tlb, tlb->addr, PAGE_SIZE);
-	return __tlb_remove_page(tlb, page);
-}
-
 #ifndef tlb_remove_check_page_size_change
 #define tlb_remove_check_page_size_change tlb_remove_check_page_size_change
 static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb,
diff --git a/mm/memory.c b/mm/memory.c
index 40752dc7750f..371604f08eb2 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -300,15 +300,14 @@ bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page, int page_
 	struct mmu_gather_batch *batch;
 
 	VM_BUG_ON(!tlb->end);
-
-	if (!tlb->page_size)
-		tlb->page_size = page_size;
-	else {
-		if (page_size != tlb->page_size)
-			return true;
-	}
+	VM_WARN_ON(tlb->page_size != page_size);
 
 	batch = tlb->active;
+	/*
+	 * Add the page and check if we are full. If so
+	 * force a flush.
+	 */
+	batch->pages[batch->nr++] = page;
 	if (batch->nr == batch->max) {
 		if (!tlb_next_batch(tlb))
 			return true;
@@ -316,7 +315,6 @@ bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page, int page_
 	}
 	VM_BUG_ON_PAGE(batch->nr > batch->max, page);
 
-	batch->pages[batch->nr++] = page;
 	return false;
 }
 
@@ -1122,7 +1120,6 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	pte_t *start_pte;
 	pte_t *pte;
 	swp_entry_t entry;
-	struct page *pending_page = NULL;
 
 	tlb_remove_check_page_size_change(tlb, PAGE_SIZE);
 again:
@@ -1177,7 +1174,6 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				print_bad_pte(vma, addr, ptent, page);
 			if (unlikely(__tlb_remove_page(tlb, page))) {
 				force_flush = 1;
-				pending_page = page;
 				addr += PAGE_SIZE;
 				break;
 			}
@@ -1218,11 +1214,6 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	if (force_flush) {
 		force_flush = 0;
 		tlb_flush_mmu_free(tlb);
-		if (pending_page) {
-			/* remove the page with new size */
-			__tlb_remove_pte_page(tlb, pending_page);
-			pending_page = NULL;
-		}
 		if (addr != end)
 			goto again;
 	}
-- 
2.10.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>