[PATCH v2 0/3] iomap/xfs: fix stale data exposure when truncating realtime inodes

[PATCH v2 0/3] iomap/xfs: fix stale data exposure when truncating realtime inodes

[Zhang Yi]

Changes since v1:
 - In iomap_truncate_page() and dax_truncate_page(), for the case of
   truncate blocksize is not power of 2, use do_dive() to calculate the
   offset.

This series fix a stale data exposure issue reported by Chandan when
running fstests generic/561 on xfs with realtime device[1]. The real
problem is xfs_setattr_size() doesn't zero out enough range when
truncating a realtime inode, please see the third patch or [1] for
details.

The first two patches modify iomap_truncate_page() and
dax_truncate_page() to pass filesystem identified blocksize, and drop
the assumption of i_blocksize() as Dave suggested. The third patch fix
the issue by modifying xfs_truncate_page() to pass the correct
blocksize, and make sure zeroed range have been flushed to disk before
updating i_size.

[1] https://lore.kernel.org/linux-xfs/87ttj8ircu.fsf@debian-BULLSEYE-live-builder-AMD64/

Thanks,
Yi.

Zhang Yi (3):
  iomap: pass blocksize to iomap_truncate_page()
  fsdax: pass blocksize to dax_truncate_page()
  xfs: correct the zeroing truncate range

 fs/dax.c               | 11 ++++++-----
 fs/ext2/inode.c        |  4 ++--
 fs/iomap/buffered-io.c | 11 ++++++-----
 fs/xfs/xfs_iomap.c     | 36 ++++++++++++++++++++++++++++++++----
 fs/xfs/xfs_iops.c      | 10 ----------
 include/linux/dax.h    |  4 ++--
 include/linux/iomap.h  |  4 ++--
 7 files changed, 50 insertions(+), 30 deletions(-)

-- 
2.39.2

[PATCH v2 1/3] iomap: pass blocksize to iomap_truncate_page()

[Zhang Yi]

From: Zhang Yi <yi.zhang@huawei.com>

iomap_truncate_page() always assumes the block size of the truncating
inode is i_blocksize(), this is not always true for some filesystems,
e.g. XFS does extent size alignment for realtime inodes. Drop this
assumption and pass the block size for zeroing into
iomap_truncate_page(), allow filesystems to indicate the correct block
size.

Suggested-by: Dave Chinner <david@fromorbit.com>
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/iomap/buffered-io.c | 11 ++++++-----
 fs/xfs/xfs_iomap.c     |  3 ++-
 include/linux/iomap.h  |  4 ++--
 3 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 0926d216a5af..229dfa3c4906 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -1445,16 +1445,17 @@ iomap_zero_range(struct inode *inode, loff_t pos, loff_t len, bool *did_zero,
 EXPORT_SYMBOL_GPL(iomap_zero_range);
 
 int
-iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
-		const struct iomap_ops *ops)
+iomap_truncate_page(struct inode *inode, loff_t pos, unsigned int blocksize,
+		bool *did_zero, const struct iomap_ops *ops)
 {
-	unsigned int blocksize = i_blocksize(inode);
-	unsigned int off = pos & (blocksize - 1);
+	loff_t start = pos;
+	unsigned int off = is_power_of_2(blocksize) ? (pos & (blocksize - 1)) :
+			   do_div(pos, blocksize);
 
 	/* Block boundary? Nothing to do */
 	if (!off)
 		return 0;
-	return iomap_zero_range(inode, pos, blocksize - off, did_zero, ops);
+	return iomap_zero_range(inode, start, blocksize - off, did_zero, ops);
 }
 EXPORT_SYMBOL_GPL(iomap_truncate_page);
 
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 2857ef1b0272..31ac07bb8425 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -1467,10 +1467,11 @@ xfs_truncate_page(
 	bool			*did_zero)
 {
 	struct inode		*inode = VFS_I(ip);
+	unsigned int		blocksize = i_blocksize(inode);
 
 	if (IS_DAX(inode))
 		return dax_truncate_page(inode, pos, did_zero,
 					&xfs_dax_write_iomap_ops);
-	return iomap_truncate_page(inode, pos, did_zero,
+	return iomap_truncate_page(inode, pos, blocksize, did_zero,
 				   &xfs_buffered_write_iomap_ops);
 }
diff --git a/include/linux/iomap.h b/include/linux/iomap.h
index 6fc1c858013d..d67bf86ec582 100644
--- a/include/linux/iomap.h
+++ b/include/linux/iomap.h
@@ -273,8 +273,8 @@ int iomap_file_unshare(struct inode *inode, loff_t pos, loff_t len,
 		const struct iomap_ops *ops);
 int iomap_zero_range(struct inode *inode, loff_t pos, loff_t len,
 		bool *did_zero, const struct iomap_ops *ops);
-int iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
-		const struct iomap_ops *ops);
+int iomap_truncate_page(struct inode *inode, loff_t pos, unsigned int blocksize,
+		bool *did_zero, const struct iomap_ops *ops);
 vm_fault_t iomap_page_mkwrite(struct vm_fault *vmf,
 			const struct iomap_ops *ops);
 int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
-- 
2.39.2

[PATCH v2 2/3] fsdax: pass blocksize to dax_truncate_page()

[Zhang Yi]

From: Zhang Yi <yi.zhang@huawei.com>

dax_truncate_page() always assumes the block size of the truncating
inode is i_blocksize(), this is not always true for some filesystems,
e.g. XFS does extent size alignment for realtime inodes. Drop this
assumption and pass the block size for zeroing into dax_truncate_page(),
allow filesystems to indicate the correct block size.

Suggested-by: Dave Chinner <david@fromorbit.com>
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/dax.c            | 11 ++++++-----
 fs/ext2/inode.c     |  4 ++--
 fs/xfs/xfs_iomap.c  |  2 +-
 include/linux/dax.h |  4 ++--
 4 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/fs/dax.c b/fs/dax.c
index 423fc1607dfa..f672c9a663c1 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -1403,16 +1403,17 @@ int dax_zero_range(struct inode *inode, loff_t pos, loff_t len, bool *did_zero,
 }
 EXPORT_SYMBOL_GPL(dax_zero_range);
 
-int dax_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
-		const struct iomap_ops *ops)
+int dax_truncate_page(struct inode *inode, loff_t pos, unsigned int blocksize,
+		bool *did_zero, const struct iomap_ops *ops)
 {
-	unsigned int blocksize = i_blocksize(inode);
-	unsigned int off = pos & (blocksize - 1);
+	loff_t start = pos;
+	unsigned int off = is_power_of_2(blocksize) ? (pos & (blocksize - 1)) :
+			   do_div(pos, blocksize);
 
 	/* Block boundary? Nothing to do */
 	if (!off)
 		return 0;
-	return dax_zero_range(inode, pos, blocksize - off, did_zero, ops);
+	return dax_zero_range(inode, start, blocksize - off, did_zero, ops);
 }
 EXPORT_SYMBOL_GPL(dax_truncate_page);
 
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index f3d570a9302b..fbbd479f3c4e 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -1278,8 +1278,8 @@ static int ext2_setsize(struct inode *inode, loff_t newsize)
 	inode_dio_wait(inode);
 
 	if (IS_DAX(inode))
-		error = dax_truncate_page(inode, newsize, NULL,
-					  &ext2_iomap_ops);
+		error = dax_truncate_page(inode, newsize, i_blocksize(inode),
+					  NULL, &ext2_iomap_ops);
 	else
 		error = block_truncate_page(inode->i_mapping,
 				newsize, ext2_get_block);
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 31ac07bb8425..4958cc3337bc 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -1470,7 +1470,7 @@ xfs_truncate_page(
 	unsigned int		blocksize = i_blocksize(inode);
 
 	if (IS_DAX(inode))
-		return dax_truncate_page(inode, pos, did_zero,
+		return dax_truncate_page(inode, pos, blocksize, did_zero,
 					&xfs_dax_write_iomap_ops);
 	return iomap_truncate_page(inode, pos, blocksize, did_zero,
 				   &xfs_buffered_write_iomap_ops);
diff --git a/include/linux/dax.h b/include/linux/dax.h
index 9d3e3327af4c..4aa8ef7c8fd4 100644
--- a/include/linux/dax.h
+++ b/include/linux/dax.h
@@ -210,8 +210,8 @@ int dax_file_unshare(struct inode *inode, loff_t pos, loff_t len,
 		const struct iomap_ops *ops);
 int dax_zero_range(struct inode *inode, loff_t pos, loff_t len, bool *did_zero,
 		const struct iomap_ops *ops);
-int dax_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
-		const struct iomap_ops *ops);
+int dax_truncate_page(struct inode *inode, loff_t pos, unsigned int blocksize,
+		bool *did_zero, const struct iomap_ops *ops);
 
 #if IS_ENABLED(CONFIG_DAX)
 int dax_read_lock(void);
-- 
2.39.2

[PATCH v2 3/3] xfs: correct the zeroing truncate range

[Zhang Yi]

From: Zhang Yi <yi.zhang@huawei.com>

When truncating a realtime file unaligned to a shorter size,
xfs_setattr_size() only flush the EOF page before zeroing out, and
xfs_truncate_page() also only zeros the EOF block. This could expose
stale data since 943bc0882ceb ("iomap: don't increase i_size if it's not
a write operation").

If the sb_rextsize is bigger than one block, and we have a realtime
inode that contains a long enough written extent. If we unaligned
truncate into the middle of this extent, xfs_itruncate_extents() could
split the extent and align the it's tail to sb_rextsize, there maybe
have more than one blocks more between the end of the file. Since
xfs_truncate_page() only zeros the trailing portion of the i_blocksize()
value, so it may leftover some blocks contains stale data that could be
exposed if we append write it over a long enough distance later.

xfs_truncate_page() should flush, zeros out the entire rtextsize range,
and make sure the entire zeroed range have been flushed to disk before
updating the inode size.

Fixes: 943bc0882ceb ("iomap: don't increase i_size if it's not a write operation")
Reported-by: Chandan Babu R <chandanbabu@kernel.org>
Link: https://lore.kernel.org/linux-xfs/0b92a215-9d9b-3788-4504-a520778953c2@huaweicloud.com
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/xfs/xfs_iomap.c | 35 +++++++++++++++++++++++++++++++----
 fs/xfs/xfs_iops.c  | 10 ----------
 2 files changed, 31 insertions(+), 14 deletions(-)

diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 4958cc3337bc..fc379450fe74 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -1466,12 +1466,39 @@ xfs_truncate_page(
 	loff_t			pos,
 	bool			*did_zero)
 {
+	struct xfs_mount	*mp = ip->i_mount;
 	struct inode		*inode = VFS_I(ip);
 	unsigned int		blocksize = i_blocksize(inode);
+	int			error;
+
+	if (XFS_IS_REALTIME_INODE(ip))
+		blocksize = XFS_FSB_TO_B(mp, mp->m_sb.sb_rextsize);
+
+	/*
+	 * iomap won't detect a dirty page over an unwritten block (or a
+	 * cow block over a hole) and subsequently skips zeroing the
+	 * newly post-EOF portion of the page. Flush the new EOF to
+	 * convert the block before the pagecache truncate.
+	 */
+	error = filemap_write_and_wait_range(inode->i_mapping, pos,
+					     roundup_64(pos, blocksize));
+	if (error)
+		return error;
 
 	if (IS_DAX(inode))
-		return dax_truncate_page(inode, pos, blocksize, did_zero,
-					&xfs_dax_write_iomap_ops);
-	return iomap_truncate_page(inode, pos, blocksize, did_zero,
-				   &xfs_buffered_write_iomap_ops);
+		error = dax_truncate_page(inode, pos, blocksize, did_zero,
+					  &xfs_dax_write_iomap_ops);
+	else
+		error = iomap_truncate_page(inode, pos, blocksize, did_zero,
+					    &xfs_buffered_write_iomap_ops);
+	if (error)
+		return error;
+
+	/*
+	 * Write back path won't write dirty blocks post EOF folio,
+	 * flush the entire zeroed range before updating the inode
+	 * size.
+	 */
+	return filemap_write_and_wait_range(inode->i_mapping, pos,
+					    roundup_64(pos, blocksize));
 }
diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
index 66f8c47642e8..baeeddf4a6bb 100644
--- a/fs/xfs/xfs_iops.c
+++ b/fs/xfs/xfs_iops.c
@@ -845,16 +845,6 @@ xfs_setattr_size(
 		error = xfs_zero_range(ip, oldsize, newsize - oldsize,
 				&did_zeroing);
 	} else {
-		/*
-		 * iomap won't detect a dirty page over an unwritten block (or a
-		 * cow block over a hole) and subsequently skips zeroing the
-		 * newly post-EOF portion of the page. Flush the new EOF to
-		 * convert the block before the pagecache truncate.
-		 */
-		error = filemap_write_and_wait_range(inode->i_mapping, newsize,
-						     newsize);
-		if (error)
-			return error;
 		error = xfs_truncate_page(ip, newsize, &did_zeroing);
 	}
 
-- 
2.39.2

Re: [PATCH v2 1/3] iomap: pass blocksize to iomap_truncate_page()

[kernel test robot]

Hi Zhang,

kernel test robot noticed the following build warnings:

[auto build test WARNING on brauner-vfs/vfs.all]
[also build test WARNING on linus/master v6.9 next-20240516]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Zhang-Yi/iomap-pass-blocksize-to-iomap_truncate_page/20240516-154238
base:   https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git vfs.all
patch link:    https://lore.kernel.org/r/20240516073001.1066373-2-yi.zhang%40huaweicloud.com
patch subject: [PATCH v2 1/3] iomap: pass blocksize to iomap_truncate_page()
config: arm-davinci_all_defconfig (https://download.01.org/0day-ci/archive/20240517/202405170356.gLrF2NDl-lkp@intel.com/config)
compiler: clang version 19.0.0git (https://github.com/llvm/llvm-project d3455f4ddd16811401fa153298fadd2f59f6914e)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240517/202405170356.gLrF2NDl-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202405170356.gLrF2NDl-lkp@intel.com/

All warnings (new ones prefixed by >>):

   In file included from fs/iomap/buffered-io.c:9:
   In file included from include/linux/iomap.h:7:
   In file included from include/linux/blk_types.h:10:
   In file included from include/linux/bvec.h:10:
   In file included from include/linux/highmem.h:8:
   In file included from include/linux/cacheflush.h:5:
   In file included from arch/arm/include/asm/cacheflush.h:10:
   In file included from include/linux/mm.h:2210:
   include/linux/vmstat.h:522:36: warning: arithmetic between different enumeration types ('enum node_stat_item' and 'enum lru_list') [-Wenum-enum-conversion]
     522 |         return node_stat_name(NR_LRU_BASE + lru) + 3; // skip "nr_"
         |                               ~~~~~~~~~~~ ^ ~~~
>> fs/iomap/buffered-io.c:1453:7: warning: comparison of distinct pointer types ('typeof ((pos)) *' (aka 'long long *') and 'uint64_t *' (aka 'unsigned long long *')) [-Wcompare-distinct-pointer-types]
    1453 |                            do_div(pos, blocksize);
         |                            ^~~~~~~~~~~~~~~~~~~~~~
   include/asm-generic/div64.h:222:28: note: expanded from macro 'do_div'
     222 |         (void)(((typeof((n)) *)0) == ((uint64_t *)0));  \
         |                ~~~~~~~~~~~~~~~~~~ ^  ~~~~~~~~~~~~~~~
   2 warnings generated.


vim +1453 fs/iomap/buffered-io.c

  1446	
  1447	int
  1448	iomap_truncate_page(struct inode *inode, loff_t pos, unsigned int blocksize,
  1449			bool *did_zero, const struct iomap_ops *ops)
  1450	{
  1451		loff_t start = pos;
  1452		unsigned int off = is_power_of_2(blocksize) ? (pos & (blocksize - 1)) :
> 1453				   do_div(pos, blocksize);
  1454	
  1455		/* Block boundary? Nothing to do */
  1456		if (!off)
  1457			return 0;
  1458		return iomap_zero_range(inode, start, blocksize - off, did_zero, ops);
  1459	}
  1460	EXPORT_SYMBOL_GPL(iomap_truncate_page);
  1461	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: [PATCH v2 1/3] iomap: pass blocksize to iomap_truncate_page()

[kernel test robot]

Hi Zhang,

kernel test robot noticed the following build warnings:

[auto build test WARNING on brauner-vfs/vfs.all]
[also build test WARNING on linus/master v6.9 next-20240516]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Zhang-Yi/iomap-pass-blocksize-to-iomap_truncate_page/20240516-154238
base:   https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git vfs.all
patch link:    https://lore.kernel.org/r/20240516073001.1066373-2-yi.zhang%40huaweicloud.com
patch subject: [PATCH v2 1/3] iomap: pass blocksize to iomap_truncate_page()
config: arm-randconfig-r111-20240517 (https://download.01.org/0day-ci/archive/20240517/202405170624.liC4qYj3-lkp@intel.com/config)
compiler: arm-linux-gnueabi-gcc (GCC) 13.2.0
reproduce: (https://download.01.org/0day-ci/archive/20240517/202405170624.liC4qYj3-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202405170624.liC4qYj3-lkp@intel.com/

sparse warnings: (new ones prefixed by >>)
>> fs/iomap/buffered-io.c:1453:28: sparse: sparse: incompatible types in comparison expression (different signedness):
   fs/iomap/buffered-io.c:1453:28: sparse:    long long *
   fs/iomap/buffered-io.c:1453:28: sparse:    unsigned long long [usertype] *

vim +1453 fs/iomap/buffered-io.c

  1446	
  1447	int
  1448	iomap_truncate_page(struct inode *inode, loff_t pos, unsigned int blocksize,
  1449			bool *did_zero, const struct iomap_ops *ops)
  1450	{
  1451		loff_t start = pos;
  1452		unsigned int off = is_power_of_2(blocksize) ? (pos & (blocksize - 1)) :
> 1453				   do_div(pos, blocksize);
  1454	
  1455		/* Block boundary? Nothing to do */
  1456		if (!off)
  1457			return 0;
  1458		return iomap_zero_range(inode, start, blocksize - off, did_zero, ops);
  1459	}
  1460	EXPORT_SYMBOL_GPL(iomap_truncate_page);
  1461	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki