Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1670320 > unrolled thread

[PATCH 0/7] userfaultfd: enable zeropage support for shmem

Back to article view | Back to linux.kernel

Contents

  [PATCH 0/7] userfaultfd: enable zeropage support for shmem  Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 7/7] userfaultfd: selftest: enable testing of UFFDIO_ZEROPAGE for shmem Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 3/7] userfaultfd: shmem: add shmem_mfill_zeropage_pte for userfaultfd support Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 6/7] userfaultfd: report UFFDIO_ZEROPAGE as available for shmem VMAs Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 2/7] shmem: introduce shmem_inode_acct_block Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 5/7] userfaultfd: shmem: wire up shmem_mfill_zeropage_pte Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 4/7] userfaultfd: mcopy_atomic: introduce mfill_atomic_pte helper Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    [PATCH 1/7] shmem: shmem_charge: verify max_block is not exceeded before inode update Mike Rapoport <rppt@linux.vnet.ibm.com> - 2017-06-20 08:30 +0200
    Re: [PATCH 0/7] userfaultfd: enable zeropage support for shmem Mike Rapoprt <rppt@linux.vnet.ibm.com> - 2017-06-30 11:30 +0200

#1670320 — [PATCH 0/7] userfaultfd: enable zeropage support for shmem

Hi,

These patches enable support for UFFDIO_ZEROPAGE for shared memory.

The first two patches are not strictly related to userfaultfd, they are
just minor refactoring to reduce amount of code duplication.

Mike Rapoport (7):
  shmem: shmem_charge: verify max_block is not exceeded before inode update
  shmem: introduce shmem_inode_acct_block
  userfaultfd: shmem: add shmem_mfill_zeropage_pte for userfaultfd support
  userfaultfd: mcopy_atomic: introduce mfill_atomic_pte helper
  userfaultfd: shmem: wire up shmem_mfill_zeropage_pte
  userfaultfd: report UFFDIO_ZEROPAGE as available for shmem VMAs
  userfaultfd: selftest: enable testing of UFFDIO_ZEROPAGE for shmem

 fs/userfaultfd.c                         |  10 +-
 include/linux/shmem_fs.h                 |   6 ++
 mm/shmem.c                               | 167 +++++++++++++++++--------------
 mm/userfaultfd.c                         |  48 ++++++---
 tools/testing/selftests/vm/userfaultfd.c |   2 +-
 5 files changed, 136 insertions(+), 97 deletions(-)

-- 
2.7.4

[toc] | [next] | [standalone]

#1670322 — [PATCH 7/7] userfaultfd: selftest: enable testing of UFFDIO_ZEROPAGE for shmem

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 tools/testing/selftests/vm/userfaultfd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/vm/userfaultfd.c b/tools/testing/selftests/vm/userfaultfd.c
index 1eae79a..b0c9263 100644
--- a/tools/testing/selftests/vm/userfaultfd.c
+++ b/tools/testing/selftests/vm/userfaultfd.c
@@ -198,7 +198,7 @@ static struct uffd_test_ops anon_uffd_test_ops = {
 };
 
 static struct uffd_test_ops shmem_uffd_test_ops = {
-	.expected_ioctls = UFFD_API_RANGE_IOCTLS_BASIC,
+	.expected_ioctls = ANON_EXPECTED_IOCTLS,
 	.allocate_area	= shmem_allocate_area,
 	.release_pages	= shmem_release_pages,
 };
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1670324 — [PATCH 3/7] userfaultfd: shmem: add shmem_mfill_zeropage_pte for userfaultfd support

shmem_mfill_zeropage_pte is the low level routine that implements the
userfaultfd UFFDIO_ZEROPAGE command. Since for shmem mappings zero pages are
always allocated and accounted, the new method is a slight extension of the
existing shmem_mcopy_atomic_pte.

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 include/linux/shmem_fs.h |  6 +++++
 mm/shmem.c               | 62 +++++++++++++++++++++++++++++++++++-------------
 2 files changed, 51 insertions(+), 17 deletions(-)

diff --git a/include/linux/shmem_fs.h b/include/linux/shmem_fs.h
index a7d6bd2..b6c3540 100644
--- a/include/linux/shmem_fs.h
+++ b/include/linux/shmem_fs.h
@@ -137,9 +137,15 @@ extern int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm, pmd_t *dst_pmd,
 				  unsigned long dst_addr,
 				  unsigned long src_addr,
 				  struct page **pagep);
+extern int shmem_mfill_zeropage_pte(struct mm_struct *dst_mm,
+				    pmd_t *dst_pmd,
+				    struct vm_area_struct *dst_vma,
+				    unsigned long dst_addr);
 #else
 #define shmem_mcopy_atomic_pte(dst_mm, dst_pte, dst_vma, dst_addr, \
 			       src_addr, pagep)        ({ BUG(); 0; })
+#define shmem_mfill_zeropage_pte(dst_mm, dst_pmd, dst_vma, \
+				 dst_addr)      ({ BUG(); 0; })
 #endif
 
 #endif
diff --git a/mm/shmem.c b/mm/shmem.c
index a92e3d7..e775a49 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2197,12 +2197,13 @@ bool shmem_mapping(struct address_space *mapping)
 	return mapping->a_ops == &shmem_aops;
 }
 
-int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
-			   pmd_t *dst_pmd,
-			   struct vm_area_struct *dst_vma,
-			   unsigned long dst_addr,
-			   unsigned long src_addr,
-			   struct page **pagep)
+static int shmem_mfill_atomic_pte(struct mm_struct *dst_mm,
+				  pmd_t *dst_pmd,
+				  struct vm_area_struct *dst_vma,
+				  unsigned long dst_addr,
+				  unsigned long src_addr,
+				  bool zeropage,
+				  struct page **pagep)
 {
 	struct inode *inode = file_inode(dst_vma->vm_file);
 	struct shmem_inode_info *info = SHMEM_I(inode);
@@ -2225,17 +2226,22 @@ int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
 		if (!page)
 			goto out_unacct_blocks;
 
-		page_kaddr = kmap_atomic(page);
-		ret = copy_from_user(page_kaddr, (const void __user *)src_addr,
-				     PAGE_SIZE);
-		kunmap_atomic(page_kaddr);
-
-		/* fallback to copy_from_user outside mmap_sem */
-		if (unlikely(ret)) {
-			*pagep = page;
-			shmem_inode_unacct_blocks(inode, 1);
-			/* don't free the page */
-			return -EFAULT;
+		if (!zeropage) {	/* mcopy_atomic */
+			page_kaddr = kmap_atomic(page);
+			ret = copy_from_user(page_kaddr,
+					     (const void __user *)src_addr,
+					     PAGE_SIZE);
+			kunmap_atomic(page_kaddr);
+
+			/* fallback to copy_from_user outside mmap_sem */
+			if (unlikely(ret)) {
+				*pagep = page;
+				shmem_inode_unacct_blocks(inode, 1);
+				/* don't free the page */
+				return -EFAULT;
+			}
+		} else {		/* mfill_zeropage_atomic */
+			clear_highpage(page);
 		}
 	} else {
 		page = *pagep;
@@ -2301,6 +2307,28 @@ int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
 	goto out;
 }
 
+int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
+			   pmd_t *dst_pmd,
+			   struct vm_area_struct *dst_vma,
+			   unsigned long dst_addr,
+			   unsigned long src_addr,
+			   struct page **pagep)
+{
+	return shmem_mfill_atomic_pte(dst_mm, dst_pmd, dst_vma,
+				      dst_addr, src_addr, false, pagep);
+}
+
+int shmem_mfill_zeropage_pte(struct mm_struct *dst_mm,
+			     pmd_t *dst_pmd,
+			     struct vm_area_struct *dst_vma,
+			     unsigned long dst_addr)
+{
+	struct page *page = NULL;
+
+	return shmem_mfill_atomic_pte(dst_mm, dst_pmd, dst_vma,
+				      dst_addr, 0, true, &page);
+}
+
 #ifdef CONFIG_TMPFS
 static const struct inode_operations shmem_symlink_inode_operations;
 static const struct inode_operations shmem_short_symlink_operations;
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1670326 — [PATCH 6/7] userfaultfd: report UFFDIO_ZEROPAGE as available for shmem VMAs

Now when shmem VMAs can be filled with zero page via userfaultfd we can
report that UFFDIO_ZEROPAGE is available for those VMAs

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 fs/userfaultfd.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
index f7555fc..57794c2 100644
--- a/fs/userfaultfd.c
+++ b/fs/userfaultfd.c
@@ -1183,7 +1183,7 @@ static int userfaultfd_register(struct userfaultfd_ctx *ctx,
 	struct uffdio_register __user *user_uffdio_register;
 	unsigned long vm_flags, new_flags;
 	bool found;
-	bool non_anon_pages;
+	bool basic_ioctls;
 	unsigned long start, end, vma_end;
 
 	user_uffdio_register = (struct uffdio_register __user *) arg;
@@ -1249,7 +1249,7 @@ static int userfaultfd_register(struct userfaultfd_ctx *ctx,
 	 * Search for not compatible vmas.
 	 */
 	found = false;
-	non_anon_pages = false;
+	basic_ioctls = false;
 	for (cur = vma; cur && cur->vm_start < end; cur = cur->vm_next) {
 		cond_resched();
 
@@ -1288,8 +1288,8 @@ static int userfaultfd_register(struct userfaultfd_ctx *ctx,
 		/*
 		 * Note vmas containing huge pages
 		 */
-		if (is_vm_hugetlb_page(cur) || vma_is_shmem(cur))
-			non_anon_pages = true;
+		if (is_vm_hugetlb_page(cur))
+			basic_ioctls = true;
 
 		found = true;
 	}
@@ -1360,7 +1360,7 @@ static int userfaultfd_register(struct userfaultfd_ctx *ctx,
 		 * userland which ioctls methods are guaranteed to
 		 * succeed on this range.
 		 */
-		if (put_user(non_anon_pages ? UFFD_API_RANGE_IOCTLS_BASIC :
+		if (put_user(basic_ioctls ? UFFD_API_RANGE_IOCTLS_BASIC :
 			     UFFD_API_RANGE_IOCTLS,
 			     &user_uffdio_register->ioctls))
 			ret = -EFAULT;
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1670327 — [PATCH 2/7] shmem: introduce shmem_inode_acct_block

The shmem_acct_block and the update of used_blocks are following one
another in all the places they are used. Combine these two into a helper
function.

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 mm/shmem.c | 102 ++++++++++++++++++++++++++++---------------------------------
 1 file changed, 46 insertions(+), 56 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index 40a43ae..a92e3d7 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -187,6 +187,38 @@ static inline void shmem_unacct_blocks(unsigned long flags, long pages)
 		vm_unacct_memory(pages * VM_ACCT(PAGE_SIZE));
 }
 
+static inline bool shmem_inode_acct_block(struct inode *inode, long pages)
+{
+	struct shmem_inode_info *info = SHMEM_I(inode);
+	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
+
+	if (shmem_acct_block(info->flags, pages))
+		return false;
+
+	if (sbinfo->max_blocks) {
+		if (percpu_counter_compare(&sbinfo->used_blocks,
+					   sbinfo->max_blocks - pages) > 0)
+			goto unacct;
+		percpu_counter_add(&sbinfo->used_blocks, pages);
+	}
+
+	return true;
+
+unacct:
+	shmem_unacct_blocks(info->flags, pages);
+	return false;
+}
+
+static inline void shmem_inode_unacct_blocks(struct inode *inode, long pages)
+{
+	struct shmem_inode_info *info = SHMEM_I(inode);
+	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
+
+	if (sbinfo->max_blocks)
+		percpu_counter_sub(&sbinfo->used_blocks, pages);
+	shmem_unacct_blocks(info->flags, pages);
+}
+
 static const struct super_operations shmem_ops;
 static const struct address_space_operations shmem_aops;
 static const struct file_operations shmem_file_operations;
@@ -248,31 +280,20 @@ static void shmem_recalc_inode(struct inode *inode)
 
 	freed = info->alloced - info->swapped - inode->i_mapping->nrpages;
 	if (freed > 0) {
-		struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
-		if (sbinfo->max_blocks)
-			percpu_counter_add(&sbinfo->used_blocks, -freed);
 		info->alloced -= freed;
 		inode->i_blocks -= freed * BLOCKS_PER_PAGE;
-		shmem_unacct_blocks(info->flags, freed);
+		shmem_inode_unacct_blocks(inode, freed);
 	}
 }
 
 bool shmem_charge(struct inode *inode, long pages)
 {
 	struct shmem_inode_info *info = SHMEM_I(inode);
-	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
 	unsigned long flags;
 
-	if (shmem_acct_block(info->flags, pages))
+	if (!shmem_inode_acct_block(inode, pages))
 		return false;
 
-	if (sbinfo->max_blocks) {
-		if (percpu_counter_compare(&sbinfo->used_blocks,
-					   sbinfo->max_blocks - pages) > 0)
-			goto unacct;
-		percpu_counter_add(&sbinfo->used_blocks, pages);
-	}
-
 	spin_lock_irqsave(&info->lock, flags);
 	info->alloced += pages;
 	inode->i_blocks += pages * BLOCKS_PER_PAGE;
@@ -281,16 +302,11 @@ bool shmem_charge(struct inode *inode, long pages)
 	inode->i_mapping->nrpages += pages;
 
 	return true;
-
-unacct:
-	shmem_unacct_blocks(info->flags, pages);
-	return false;
 }
 
 void shmem_uncharge(struct inode *inode, long pages)
 {
 	struct shmem_inode_info *info = SHMEM_I(inode);
-	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
 	unsigned long flags;
 
 	spin_lock_irqsave(&info->lock, flags);
@@ -299,9 +315,7 @@ void shmem_uncharge(struct inode *inode, long pages)
 	shmem_recalc_inode(inode);
 	spin_unlock_irqrestore(&info->lock, flags);
 
-	if (sbinfo->max_blocks)
-		percpu_counter_sub(&sbinfo->used_blocks, pages);
-	shmem_unacct_blocks(info->flags, pages);
+	shmem_inode_unacct_blocks(inode, pages);
 }
 
 /*
@@ -1446,9 +1460,10 @@ static struct page *shmem_alloc_page(gfp_t gfp,
 }
 
 static struct page *shmem_alloc_and_acct_page(gfp_t gfp,
-		struct shmem_inode_info *info, struct shmem_sb_info *sbinfo,
+		struct inode *inode,
 		pgoff_t index, bool huge)
 {
+	struct shmem_inode_info *info = SHMEM_I(inode);
 	struct page *page;
 	int nr;
 	int err = -ENOSPC;
@@ -1457,14 +1472,8 @@ static struct page *shmem_alloc_and_acct_page(gfp_t gfp,
 		huge = false;
 	nr = huge ? HPAGE_PMD_NR : 1;
 
-	if (shmem_acct_block(info->flags, nr))
+	if (!shmem_inode_acct_block(inode, nr))
 		goto failed;
-	if (sbinfo->max_blocks) {
-		if (percpu_counter_compare(&sbinfo->used_blocks,
-					sbinfo->max_blocks - nr) > 0)
-			goto unacct;
-		percpu_counter_add(&sbinfo->used_blocks, nr);
-	}
 
 	if (huge)
 		page = shmem_alloc_hugepage(gfp, info, index);
@@ -1477,10 +1486,7 @@ static struct page *shmem_alloc_and_acct_page(gfp_t gfp,
 	}
 
 	err = -ENOMEM;
-	if (sbinfo->max_blocks)
-		percpu_counter_add(&sbinfo->used_blocks, -nr);
-unacct:
-	shmem_unacct_blocks(info->flags, nr);
+	shmem_inode_unacct_blocks(inode, nr);
 failed:
 	return ERR_PTR(err);
 }
@@ -1746,10 +1752,9 @@ static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
 		}
 
 alloc_huge:
-		page = shmem_alloc_and_acct_page(gfp, info, sbinfo,
-				index, true);
+		page = shmem_alloc_and_acct_page(gfp, inode, index, true);
 		if (IS_ERR(page)) {
-alloc_nohuge:		page = shmem_alloc_and_acct_page(gfp, info, sbinfo,
+alloc_nohuge:		page = shmem_alloc_and_acct_page(gfp, inode,
 					index, false);
 		}
 		if (IS_ERR(page)) {
@@ -1867,10 +1872,7 @@ alloc_nohuge:		page = shmem_alloc_and_acct_page(gfp, info, sbinfo,
 	 * Error recovery.
 	 */
 unacct:
-	if (sbinfo->max_blocks)
-		percpu_counter_sub(&sbinfo->used_blocks,
-				1 << compound_order(page));
-	shmem_unacct_blocks(info->flags, 1 << compound_order(page));
+	shmem_inode_unacct_blocks(inode, 1 << compound_order(page));
 
 	if (PageTransHuge(page)) {
 		unlock_page(page);
@@ -2204,7 +2206,6 @@ int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
 {
 	struct inode *inode = file_inode(dst_vma->vm_file);
 	struct shmem_inode_info *info = SHMEM_I(inode);
-	struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
 	struct address_space *mapping = inode->i_mapping;
 	gfp_t gfp = mapping_gfp_mask(mapping);
 	pgoff_t pgoff = linear_page_index(dst_vma, dst_addr);
@@ -2216,19 +2217,13 @@ int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
 	int ret;
 
 	ret = -ENOMEM;
-	if (shmem_acct_block(info->flags, 1))
+	if (!shmem_inode_acct_block(inode, 1))
 		goto out;
-	if (sbinfo->max_blocks) {
-		if (percpu_counter_compare(&sbinfo->used_blocks,
-					   sbinfo->max_blocks) >= 0)
-			goto out_unacct_blocks;
-		percpu_counter_inc(&sbinfo->used_blocks);
-	}
 
 	if (!*pagep) {
 		page = shmem_alloc_page(gfp, info, pgoff);
 		if (!page)
-			goto out_dec_used_blocks;
+			goto out_unacct_blocks;
 
 		page_kaddr = kmap_atomic(page);
 		ret = copy_from_user(page_kaddr, (const void __user *)src_addr,
@@ -2238,9 +2233,7 @@ int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
 		/* fallback to copy_from_user outside mmap_sem */
 		if (unlikely(ret)) {
 			*pagep = page;
-			if (sbinfo->max_blocks)
-				percpu_counter_add(&sbinfo->used_blocks, -1);
-			shmem_unacct_blocks(info->flags, 1);
+			shmem_inode_unacct_blocks(inode, 1);
 			/* don't free the page */
 			return -EFAULT;
 		}
@@ -2303,11 +2296,8 @@ int shmem_mcopy_atomic_pte(struct mm_struct *dst_mm,
 out_release:
 	unlock_page(page);
 	put_page(page);
-out_dec_used_blocks:
-	if (sbinfo->max_blocks)
-		percpu_counter_add(&sbinfo->used_blocks, -1);
 out_unacct_blocks:
-	shmem_unacct_blocks(info->flags, 1);
+	shmem_inode_unacct_blocks(inode, 1);
 	goto out;
 }
 
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1670328 — [PATCH 5/7] userfaultfd: shmem: wire up shmem_mfill_zeropage_pte

For shmem VMAs we can use shmem_mfill_zeropage_pte for UFFDIO_ZEROPAGE

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 mm/userfaultfd.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c
index 48c015c..8119270 100644
--- a/mm/userfaultfd.c
+++ b/mm/userfaultfd.c
@@ -389,11 +389,13 @@ static __always_inline ssize_t mfill_atomic_pte(struct mm_struct *dst_mm,
 			err = mfill_zeropage_pte(dst_mm, dst_pmd,
 						 dst_vma, dst_addr);
 	} else {
-		err = -EINVAL; /* if zeropage is true return -EINVAL */
-		if (likely(!zeropage))
+		if (!zeropage)
 			err = shmem_mcopy_atomic_pte(dst_mm, dst_pmd,
 						     dst_vma, dst_addr,
 						     src_addr, page);
+		else
+			err = shmem_mfill_zeropage_pte(dst_mm, dst_pmd,
+						       dst_vma, dst_addr);
 	}
 
 	return err;
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1670329 — [PATCH 4/7] userfaultfd: mcopy_atomic: introduce mfill_atomic_pte helper

Shuffle the code a bit to improve readability.

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 mm/userfaultfd.c | 46 ++++++++++++++++++++++++++++++----------------
 1 file changed, 30 insertions(+), 16 deletions(-)

diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c
index 8bcb501..48c015c 100644
--- a/mm/userfaultfd.c
+++ b/mm/userfaultfd.c
@@ -371,6 +371,34 @@ extern ssize_t __mcopy_atomic_hugetlb(struct mm_struct *dst_mm,
 				      bool zeropage);
 #endif /* CONFIG_HUGETLB_PAGE */
 
+static __always_inline ssize_t mfill_atomic_pte(struct mm_struct *dst_mm,
+						pmd_t *dst_pmd,
+						struct vm_area_struct *dst_vma,
+						unsigned long dst_addr,
+						unsigned long src_addr,
+						struct page **page,
+						bool zeropage)
+{
+	ssize_t err;
+
+	if (vma_is_anonymous(dst_vma)) {
+		if (!zeropage)
+			err = mcopy_atomic_pte(dst_mm, dst_pmd, dst_vma,
+					       dst_addr, src_addr, page);
+		else
+			err = mfill_zeropage_pte(dst_mm, dst_pmd,
+						 dst_vma, dst_addr);
+	} else {
+		err = -EINVAL; /* if zeropage is true return -EINVAL */
+		if (likely(!zeropage))
+			err = shmem_mcopy_atomic_pte(dst_mm, dst_pmd,
+						     dst_vma, dst_addr,
+						     src_addr, page);
+	}
+
+	return err;
+}
+
 static __always_inline ssize_t __mcopy_atomic(struct mm_struct *dst_mm,
 					      unsigned long dst_start,
 					      unsigned long src_start,
@@ -487,22 +515,8 @@ static __always_inline ssize_t __mcopy_atomic(struct mm_struct *dst_mm,
 		BUG_ON(pmd_none(*dst_pmd));
 		BUG_ON(pmd_trans_huge(*dst_pmd));
 
-		if (vma_is_anonymous(dst_vma)) {
-			if (!zeropage)
-				err = mcopy_atomic_pte(dst_mm, dst_pmd, dst_vma,
-						       dst_addr, src_addr,
-						       &page);
-			else
-				err = mfill_zeropage_pte(dst_mm, dst_pmd,
-							 dst_vma, dst_addr);
-		} else {
-			err = -EINVAL; /* if zeropage is true return -EINVAL */
-			if (likely(!zeropage))
-				err = shmem_mcopy_atomic_pte(dst_mm, dst_pmd,
-							     dst_vma, dst_addr,
-							     src_addr, &page);
-		}
-
+		err = mfill_atomic_pte(dst_mm, dst_pmd, dst_vma, dst_addr,
+				       src_addr, &page, zeropage);
 		cond_resched();
 
 		if (unlikely(err == -EFAULT)) {
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1670331 — [PATCH 1/7] shmem: shmem_charge: verify max_block is not exceeded before inode update

Currently we update inode and shmem_inode_info before verifying that
used_blocks will not exceed max_blocks. In case it will, we undo the
update. Let's switch the order and move the verification of the blocks
count before the inode and shmem_inode_info update.

Signed-off-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
---
 mm/shmem.c | 25 ++++++++++++-------------
 1 file changed, 12 insertions(+), 13 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index e67d6ba..40a43ae 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -265,6 +265,14 @@ bool shmem_charge(struct inode *inode, long pages)
 
 	if (shmem_acct_block(info->flags, pages))
 		return false;
+
+	if (sbinfo->max_blocks) {
+		if (percpu_counter_compare(&sbinfo->used_blocks,
+					   sbinfo->max_blocks - pages) > 0)
+			goto unacct;
+		percpu_counter_add(&sbinfo->used_blocks, pages);
+	}
+
 	spin_lock_irqsave(&info->lock, flags);
 	info->alloced += pages;
 	inode->i_blocks += pages * BLOCKS_PER_PAGE;
@@ -272,20 +280,11 @@ bool shmem_charge(struct inode *inode, long pages)
 	spin_unlock_irqrestore(&info->lock, flags);
 	inode->i_mapping->nrpages += pages;
 
-	if (!sbinfo->max_blocks)
-		return true;
-	if (percpu_counter_compare(&sbinfo->used_blocks,
-				sbinfo->max_blocks - pages) > 0) {
-		inode->i_mapping->nrpages -= pages;
-		spin_lock_irqsave(&info->lock, flags);
-		info->alloced -= pages;
-		shmem_recalc_inode(inode);
-		spin_unlock_irqrestore(&info->lock, flags);
-		shmem_unacct_blocks(info->flags, pages);
-		return false;
-	}
-	percpu_counter_add(&sbinfo->used_blocks, pages);
 	return true;
+
+unacct:
+	shmem_unacct_blocks(info->flags, pages);
+	return false;
 }
 
 void shmem_uncharge(struct inode *inode, long pages)
-- 
2.7.4

[toc] | [prev] | [next] | [standalone]

#1678627

Hi,

Any updates/comments?


On June 20, 2017 9:20:45 AM GMT+03:00, Mike Rapoport <rppt@linux.vnet.ibm.com> wrote:
>Hi,
>
>These patches enable support for UFFDIO_ZEROPAGE for shared memory.
>
>The first two patches are not strictly related to userfaultfd, they are
>just minor refactoring to reduce amount of code duplication.
>
>Mike Rapoport (7):
>shmem: shmem_charge: verify max_block is not exceeded before inode
>update
>  shmem: introduce shmem_inode_acct_block
>userfaultfd: shmem: add shmem_mfill_zeropage_pte for userfaultfd
>support
>  userfaultfd: mcopy_atomic: introduce mfill_atomic_pte helper
>  userfaultfd: shmem: wire up shmem_mfill_zeropage_pte
>  userfaultfd: report UFFDIO_ZEROPAGE as available for shmem VMAs
>  userfaultfd: selftest: enable testing of UFFDIO_ZEROPAGE for shmem
>
> fs/userfaultfd.c                         |  10 +-
> include/linux/shmem_fs.h                 |   6 ++
>mm/shmem.c                               | 167
>+++++++++++++++++--------------
> mm/userfaultfd.c                         |  48 ++++++---
> tools/testing/selftests/vm/userfaultfd.c |   2 +-
> 5 files changed, 136 insertions(+), 97 deletions(-)

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web