[PATCH 1/4] ext4: fix out of bounds punch offset

[PATCH 1/4] ext4: fix out of bounds punch offset

[Zhang Yi]

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

Punching a hole with a start offset that exceeds max_end is not
permitted and will result in a negative length in the
truncate_inode_partial_folio() function while truncating the page cache,
potentially leading to undesirable consequences.

A simple reproducer:

  truncate -s 9895604649994 /mnt/foo
  xfs_io -c "pwrite 8796093022208 4096" /mnt/foo
  xfs_io -c "fpunch 8796093022213 25769803777" /mnt/foo

  kernel BUG at include/linux/highmem.h:275!
  Oops: invalid opcode: 0000 [#1] SMP PTI
  CPU: 3 UID: 0 PID: 710 Comm: xfs_io Not tainted 6.15.0-rc3
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014
  RIP: 0010:zero_user_segments.constprop.0+0xd7/0x110
  RSP: 0018:ffffc90001cf3b38 EFLAGS: 00010287
  RAX: 0000000000000005 RBX: ffffea0001485e40 RCX: 0000000000001000
  RDX: 000000000040b000 RSI: 0000000000000005 RDI: 000000000040b000
  RBP: 000000000040affb R08: ffff888000000000 R09: ffffea0000000000
  R10: 0000000000000003 R11: 00000000fffc7fc5 R12: 0000000000000005
  R13: 000000000040affb R14: ffffea0001485e40 R15: ffff888031cd3000
  FS:  00007f4f63d0b780(0000) GS:ffff8880d337d000(0000)
  knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 000000001ae0b038 CR3: 00000000536aa000 CR4: 00000000000006f0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Call Trace:
   <TASK>
   truncate_inode_partial_folio+0x3dd/0x620
   truncate_inode_pages_range+0x226/0x720
   ? bdev_getblk+0x52/0x3e0
   ? ext4_get_group_desc+0x78/0x150
   ? crc32c_arch+0xfd/0x180
   ? __ext4_get_inode_loc+0x18c/0x840
   ? ext4_inode_csum+0x117/0x160
   ? jbd2_journal_dirty_metadata+0x61/0x390
   ? __ext4_handle_dirty_metadata+0xa0/0x2b0
   ? kmem_cache_free+0x90/0x5a0
   ? jbd2_journal_stop+0x1d5/0x550
   ? __ext4_journal_stop+0x49/0x100
   truncate_pagecache_range+0x50/0x80
   ext4_truncate_page_cache_block_range+0x57/0x3a0
   ext4_punch_hole+0x1fe/0x670
   ext4_fallocate+0x792/0x17d0
   ? __count_memcg_events+0x175/0x2a0
   vfs_fallocate+0x121/0x560
   ksys_fallocate+0x51/0xc0
   __x64_sys_fallocate+0x24/0x40
   x64_sys_call+0x18d2/0x4170
   do_syscall_64+0xa7/0x220
   entry_SYSCALL_64_after_hwframe+0x76/0x7e

Fix this by filtering out cases where the punching start offset exceeds
max_end.

Fixes: 982bf37da09d ("ext4: refactor ext4_punch_hole()")
Reported-by: Liebes Wang <wanghaichi0403@gmail.com>
Closes: https://lore.kernel.org/linux-ext4/ac3a58f6-e686-488b-a9ee-fc041024e43d@huawei.com/
Tested-by: Liebes Wang <wanghaichi0403@gmail.com>
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/ext4/inode.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 94c7d2d828a6..4ec4a80b6879 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -4016,7 +4016,7 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
 	WARN_ON_ONCE(!inode_is_locked(inode));
 
 	/* No need to punch hole beyond i_size */
-	if (offset >= inode->i_size)
+	if (offset >= inode->i_size || offset >= max_end)
 		return 0;
 
 	/*
-- 
2.46.1

[PATCH 2/4] ext4: fix incorrect punch max_end

[Zhang Yi]

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

For the extents inodes, the maxbytes should be sb->s_maxbytes instead of
sbi->s_bitmap_maxbytes. Correct the maxbytes value to correct the
behavior of punch hole.

Fixes: 2da376228a24 ("ext4: limit length to bitmap_maxbytes - blocksize in punch_hole")
Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/ext4/inode.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 4ec4a80b6879..f9725e6347c7 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -4006,7 +4006,7 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
 	struct inode *inode = file_inode(file);
 	struct super_block *sb = inode->i_sb;
 	ext4_lblk_t start_lblk, end_lblk;
-	loff_t max_end = EXT4_SB(sb)->s_bitmap_maxbytes - sb->s_blocksize;
+	loff_t max_end;
 	loff_t end = offset + length;
 	handle_t *handle;
 	unsigned int credits;
@@ -4015,6 +4015,12 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
 	trace_ext4_punch_hole(inode, offset, length, 0);
 	WARN_ON_ONCE(!inode_is_locked(inode));
 
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
+		max_end = sb->s_maxbytes;
+	else
+		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
+	max_end -= sb->s_blocksize;
+
 	/* No need to punch hole beyond i_size */
 	if (offset >= inode->i_size || offset >= max_end)
 		return 0;
-- 
2.46.1

[PATCH 3/4] ext4: factor out ext4_get_maxbytes()

[Zhang Yi]

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

There are several locations that get the correct maxbytes value based on
the inode's block type. It would be beneficial to extract a common
helper function to make the code more clear.

Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/ext4/ext4.h    | 7 +++++++
 fs/ext4/extents.c | 7 +------
 fs/ext4/file.c    | 7 +------
 fs/ext4/inode.c   | 8 +-------
 4 files changed, 10 insertions(+), 19 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 5a20e9cd7184..8664bb5367c5 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -3378,6 +3378,13 @@ static inline unsigned int ext4_flex_bg_size(struct ext4_sb_info *sbi)
 	return 1 << sbi->s_log_groups_per_flex;
 }
 
+static inline loff_t ext4_get_maxbytes(struct inode *inode)
+{
+	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
+		return inode->i_sb->s_maxbytes;
+	return EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
+}
+
 #define ext4_std_error(sb, errno)				\
 do {								\
 	if ((errno))						\
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index c616a16a9f36..b294d2f35a26 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -4931,12 +4931,7 @@ static const struct iomap_ops ext4_iomap_xattr_ops = {
 
 static int ext4_fiemap_check_ranges(struct inode *inode, u64 start, u64 *len)
 {
-	u64 maxbytes;
-
-	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
-		maxbytes = inode->i_sb->s_maxbytes;
-	else
-		maxbytes = EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
+	u64 maxbytes = ext4_get_maxbytes(inode);
 
 	if (*len == 0)
 		return -EINVAL;
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index beb078ee4811..b845a25f7932 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -929,12 +929,7 @@ static int ext4_file_open(struct inode *inode, struct file *filp)
 loff_t ext4_llseek(struct file *file, loff_t offset, int whence)
 {
 	struct inode *inode = file->f_mapping->host;
-	loff_t maxbytes;
-
-	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
-		maxbytes = EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
-	else
-		maxbytes = inode->i_sb->s_maxbytes;
+	loff_t maxbytes = ext4_get_maxbytes(inode);
 
 	switch (whence) {
 	default:
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index f9725e6347c7..9f32af1241ff 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -4006,7 +4006,7 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
 	struct inode *inode = file_inode(file);
 	struct super_block *sb = inode->i_sb;
 	ext4_lblk_t start_lblk, end_lblk;
-	loff_t max_end;
+	loff_t max_end = ext4_get_maxbytes(inode) - sb->s_blocksize;
 	loff_t end = offset + length;
 	handle_t *handle;
 	unsigned int credits;
@@ -4015,12 +4015,6 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
 	trace_ext4_punch_hole(inode, offset, length, 0);
 	WARN_ON_ONCE(!inode_is_locked(inode));
 
-	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
-		max_end = sb->s_maxbytes;
-	else
-		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
-	max_end -= sb->s_blocksize;
-
 	/* No need to punch hole beyond i_size */
 	if (offset >= inode->i_size || offset >= max_end)
 		return 0;
-- 
2.46.1

[PATCH 4/4] ext4: ensure i_size is smaller than maxbytes

[Zhang Yi]

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

The inode i_size cannot be larger than maxbytes, check it while loading
inode from the disk.

Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
---
 fs/ext4/inode.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 9f32af1241ff..2bd2a4f4d9d7 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -4916,7 +4916,8 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
 		ei->i_file_acl |=
 			((__u64)le16_to_cpu(raw_inode->i_file_acl_high)) << 32;
 	inode->i_size = ext4_isize(sb, raw_inode);
-	if ((size = i_size_read(inode)) < 0) {
+	size = i_size_read(inode);
+	if (size < 0 || size > ext4_get_maxbytes(inode)) {
 		ext4_error_inode(inode, function, line, 0,
 				 "iget: bad i_size value: %lld", size);
 		ret = -EFSCORRUPTED;
-- 
2.46.1

Re: [PATCH 4/4] ext4: ensure i_size is smaller than maxbytes

[Matthew Wilcox]

On Wed, Apr 30, 2025 at 09:13:01AM +0800, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
> 
> The inode i_size cannot be larger than maxbytes, check it while loading
> inode from the disk.

With this patch, why do we need patch 1?

Re: [PATCH 4/4] ext4: ensure i_size is smaller than maxbytes

[Zhang Yi]

On 2025/4/30 11:20, Matthew Wilcox wrote:
> On Wed, Apr 30, 2025 at 09:13:01AM +0800, Zhang Yi wrote:
>> From: Zhang Yi <yi.zhang@huawei.com>
>>
>> The inode i_size cannot be larger than maxbytes, check it while loading
>> inode from the disk.
> 
> With this patch, why do we need patch 1?

The max_end in ext4_punch_hole() is maxbytes - blocksize. Therefore,
without the patch 1, it is still possible to encounter a negative
length if the offset exceeds max_end but is less than an i_size that
also exceeds max_end.

                 max_end    i_size  max_bytes
                    |         |        |
   0------------------------------------
                            |
                         offset

Thanks,
Yi.

Re: [PATCH 1/4] ext4: fix out of bounds punch offset

[Jan Kara]

On Wed 30-04-25 09:12:58, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
> 
> Punching a hole with a start offset that exceeds max_end is not
> permitted and will result in a negative length in the
> truncate_inode_partial_folio() function while truncating the page cache,
> potentially leading to undesirable consequences.
> 
> A simple reproducer:
> 
>   truncate -s 9895604649994 /mnt/foo
>   xfs_io -c "pwrite 8796093022208 4096" /mnt/foo
>   xfs_io -c "fpunch 8796093022213 25769803777" /mnt/foo
> 
>   kernel BUG at include/linux/highmem.h:275!
>   Oops: invalid opcode: 0000 [#1] SMP PTI
>   CPU: 3 UID: 0 PID: 710 Comm: xfs_io Not tainted 6.15.0-rc3
>   Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-2.fc40 04/01/2014
>   RIP: 0010:zero_user_segments.constprop.0+0xd7/0x110
>   RSP: 0018:ffffc90001cf3b38 EFLAGS: 00010287
>   RAX: 0000000000000005 RBX: ffffea0001485e40 RCX: 0000000000001000
>   RDX: 000000000040b000 RSI: 0000000000000005 RDI: 000000000040b000
>   RBP: 000000000040affb R08: ffff888000000000 R09: ffffea0000000000
>   R10: 0000000000000003 R11: 00000000fffc7fc5 R12: 0000000000000005
>   R13: 000000000040affb R14: ffffea0001485e40 R15: ffff888031cd3000
>   FS:  00007f4f63d0b780(0000) GS:ffff8880d337d000(0000)
>   knlGS:0000000000000000
>   CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>   CR2: 000000001ae0b038 CR3: 00000000536aa000 CR4: 00000000000006f0
>   DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
>   DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
>   Call Trace:
>    <TASK>
>    truncate_inode_partial_folio+0x3dd/0x620
>    truncate_inode_pages_range+0x226/0x720
>    ? bdev_getblk+0x52/0x3e0
>    ? ext4_get_group_desc+0x78/0x150
>    ? crc32c_arch+0xfd/0x180
>    ? __ext4_get_inode_loc+0x18c/0x840
>    ? ext4_inode_csum+0x117/0x160
>    ? jbd2_journal_dirty_metadata+0x61/0x390
>    ? __ext4_handle_dirty_metadata+0xa0/0x2b0
>    ? kmem_cache_free+0x90/0x5a0
>    ? jbd2_journal_stop+0x1d5/0x550
>    ? __ext4_journal_stop+0x49/0x100
>    truncate_pagecache_range+0x50/0x80
>    ext4_truncate_page_cache_block_range+0x57/0x3a0
>    ext4_punch_hole+0x1fe/0x670
>    ext4_fallocate+0x792/0x17d0
>    ? __count_memcg_events+0x175/0x2a0
>    vfs_fallocate+0x121/0x560
>    ksys_fallocate+0x51/0xc0
>    __x64_sys_fallocate+0x24/0x40
>    x64_sys_call+0x18d2/0x4170
>    do_syscall_64+0xa7/0x220
>    entry_SYSCALL_64_after_hwframe+0x76/0x7e
> 
> Fix this by filtering out cases where the punching start offset exceeds
> max_end.
> 
> Fixes: 982bf37da09d ("ext4: refactor ext4_punch_hole()")
> Reported-by: Liebes Wang <wanghaichi0403@gmail.com>
> Closes: https://lore.kernel.org/linux-ext4/ac3a58f6-e686-488b-a9ee-fc041024e43d@huawei.com/
> Tested-by: Liebes Wang <wanghaichi0403@gmail.com>
> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>

Looks good. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
>  fs/ext4/inode.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 94c7d2d828a6..4ec4a80b6879 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -4016,7 +4016,7 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>  	WARN_ON_ONCE(!inode_is_locked(inode));
>  
>  	/* No need to punch hole beyond i_size */
> -	if (offset >= inode->i_size)
> +	if (offset >= inode->i_size || offset >= max_end)
>  		return 0;
>  
>  	/*
> -- 
> 2.46.1
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH 2/4] ext4: fix incorrect punch max_end

[Jan Kara]

On Wed 30-04-25 09:12:59, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
> 
> For the extents inodes, the maxbytes should be sb->s_maxbytes instead of
> sbi->s_bitmap_maxbytes. Correct the maxbytes value to correct the
> behavior of punch hole.
> 
> Fixes: 2da376228a24 ("ext4: limit length to bitmap_maxbytes - blocksize in punch_hole")
> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>

Looks good. Thanks! Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
>  fs/ext4/inode.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 4ec4a80b6879..f9725e6347c7 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -4006,7 +4006,7 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>  	struct inode *inode = file_inode(file);
>  	struct super_block *sb = inode->i_sb;
>  	ext4_lblk_t start_lblk, end_lblk;
> -	loff_t max_end = EXT4_SB(sb)->s_bitmap_maxbytes - sb->s_blocksize;
> +	loff_t max_end;
>  	loff_t end = offset + length;
>  	handle_t *handle;
>  	unsigned int credits;
> @@ -4015,6 +4015,12 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>  	trace_ext4_punch_hole(inode, offset, length, 0);
>  	WARN_ON_ONCE(!inode_is_locked(inode));
>  
> +	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
> +		max_end = sb->s_maxbytes;
> +	else
> +		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
> +	max_end -= sb->s_blocksize;
> +
>  	/* No need to punch hole beyond i_size */
>  	if (offset >= inode->i_size || offset >= max_end)
>  		return 0;
> -- 
> 2.46.1
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH 3/4] ext4: factor out ext4_get_maxbytes()

[Jan Kara]

On Wed 30-04-25 09:13:00, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
> 
> There are several locations that get the correct maxbytes value based on
> the inode's block type. It would be beneficial to extract a common
> helper function to make the code more clear.
> 
> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>

Nice. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
>  fs/ext4/ext4.h    | 7 +++++++
>  fs/ext4/extents.c | 7 +------
>  fs/ext4/file.c    | 7 +------
>  fs/ext4/inode.c   | 8 +-------
>  4 files changed, 10 insertions(+), 19 deletions(-)
> 
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 5a20e9cd7184..8664bb5367c5 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -3378,6 +3378,13 @@ static inline unsigned int ext4_flex_bg_size(struct ext4_sb_info *sbi)
>  	return 1 << sbi->s_log_groups_per_flex;
>  }
>  
> +static inline loff_t ext4_get_maxbytes(struct inode *inode)
> +{
> +	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
> +		return inode->i_sb->s_maxbytes;
> +	return EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
> +}
> +
>  #define ext4_std_error(sb, errno)				\
>  do {								\
>  	if ((errno))						\
> diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
> index c616a16a9f36..b294d2f35a26 100644
> --- a/fs/ext4/extents.c
> +++ b/fs/ext4/extents.c
> @@ -4931,12 +4931,7 @@ static const struct iomap_ops ext4_iomap_xattr_ops = {
>  
>  static int ext4_fiemap_check_ranges(struct inode *inode, u64 start, u64 *len)
>  {
> -	u64 maxbytes;
> -
> -	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
> -		maxbytes = inode->i_sb->s_maxbytes;
> -	else
> -		maxbytes = EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
> +	u64 maxbytes = ext4_get_maxbytes(inode);
>  
>  	if (*len == 0)
>  		return -EINVAL;
> diff --git a/fs/ext4/file.c b/fs/ext4/file.c
> index beb078ee4811..b845a25f7932 100644
> --- a/fs/ext4/file.c
> +++ b/fs/ext4/file.c
> @@ -929,12 +929,7 @@ static int ext4_file_open(struct inode *inode, struct file *filp)
>  loff_t ext4_llseek(struct file *file, loff_t offset, int whence)
>  {
>  	struct inode *inode = file->f_mapping->host;
> -	loff_t maxbytes;
> -
> -	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
> -		maxbytes = EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
> -	else
> -		maxbytes = inode->i_sb->s_maxbytes;
> +	loff_t maxbytes = ext4_get_maxbytes(inode);
>  
>  	switch (whence) {
>  	default:
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index f9725e6347c7..9f32af1241ff 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -4006,7 +4006,7 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>  	struct inode *inode = file_inode(file);
>  	struct super_block *sb = inode->i_sb;
>  	ext4_lblk_t start_lblk, end_lblk;
> -	loff_t max_end;
> +	loff_t max_end = ext4_get_maxbytes(inode) - sb->s_blocksize;
>  	loff_t end = offset + length;
>  	handle_t *handle;
>  	unsigned int credits;
> @@ -4015,12 +4015,6 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>  	trace_ext4_punch_hole(inode, offset, length, 0);
>  	WARN_ON_ONCE(!inode_is_locked(inode));
>  
> -	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
> -		max_end = sb->s_maxbytes;
> -	else
> -		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
> -	max_end -= sb->s_blocksize;
> -
>  	/* No need to punch hole beyond i_size */
>  	if (offset >= inode->i_size || offset >= max_end)
>  		return 0;
> -- 
> 2.46.1
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH 4/4] ext4: ensure i_size is smaller than maxbytes

[Jan Kara]

On Wed 30-04-25 09:13:01, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
> 
> The inode i_size cannot be larger than maxbytes, check it while loading
> inode from the disk.
> 
> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>

Looks good. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
>  fs/ext4/inode.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index 9f32af1241ff..2bd2a4f4d9d7 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -4916,7 +4916,8 @@ struct inode *__ext4_iget(struct super_block *sb, unsigned long ino,
>  		ei->i_file_acl |=
>  			((__u64)le16_to_cpu(raw_inode->i_file_acl_high)) << 32;
>  	inode->i_size = ext4_isize(sb, raw_inode);
> -	if ((size = i_size_read(inode)) < 0) {
> +	size = i_size_read(inode);
> +	if (size < 0 || size > ext4_get_maxbytes(inode)) {
>  		ext4_error_inode(inode, function, line, 0,
>  				 "iget: bad i_size value: %lld", size);
>  		ret = -EFSCORRUPTED;
> -- 
> 2.46.1
> 
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH 2/4] ext4: fix incorrect punch max_end

[Jan Kara]

On Wed 30-04-25 09:12:59, Zhang Yi wrote:
> From: Zhang Yi <yi.zhang@huawei.com>
> 
> For the extents inodes, the maxbytes should be sb->s_maxbytes instead of
> sbi->s_bitmap_maxbytes. Correct the maxbytes value to correct the
> behavior of punch hole.
> 
> Fixes: 2da376228a24 ("ext4: limit length to bitmap_maxbytes - blocksize in punch_hole")
> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>

Thinking about this some more...

> @@ -4015,6 +4015,12 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>  	trace_ext4_punch_hole(inode, offset, length, 0);
>  	WARN_ON_ONCE(!inode_is_locked(inode));
>  
> +	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
> +		max_end = sb->s_maxbytes;
> +	else
> +		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
> +	max_end -= sb->s_blocksize;

I think the -= sb->s_blocksize is needed only for indirect-block based
scheme (due to an implementation quirk in ext4_ind_remove_space()). But
ext4_ext_remove_space() should be fine with punch hole ending right at
sb->s_maxbytes. And since I find it somewhat odd that you can create file
upto s_maxbytes but cannot punch hole to the end, it'd limit that behavior
as much as possible. Ideally we'd fix ext4_ind_remove_space() but I can't
be really bothered for the ancient format...

								Honza

-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH 2/4] ext4: fix incorrect punch max_end

[Zhang Yi]

On 2025/4/30 16:18, Jan Kara wrote:
> On Wed 30-04-25 09:12:59, Zhang Yi wrote:
>> From: Zhang Yi <yi.zhang@huawei.com>
>>
>> For the extents inodes, the maxbytes should be sb->s_maxbytes instead of
>> sbi->s_bitmap_maxbytes. Correct the maxbytes value to correct the
>> behavior of punch hole.
>>
>> Fixes: 2da376228a24 ("ext4: limit length to bitmap_maxbytes - blocksize in punch_hole")
>> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
> 
> Thinking about this some more...
> 
>> @@ -4015,6 +4015,12 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>>  	trace_ext4_punch_hole(inode, offset, length, 0);
>>  	WARN_ON_ONCE(!inode_is_locked(inode));
>>  
>> +	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
>> +		max_end = sb->s_maxbytes;
>> +	else
>> +		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
>> +	max_end -= sb->s_blocksize;
> 
> I think the -= sb->s_blocksize is needed only for indirect-block based
> scheme (due to an implementation quirk in ext4_ind_remove_space()). But
> ext4_ext_remove_space() should be fine with punch hole ending right at
> sb->s_maxbytes. And since I find it somewhat odd that you can create file
> upto s_maxbytes but cannot punch hole to the end, it'd limit that behavior
> as much as possible. Ideally we'd fix ext4_ind_remove_space() but I can't
> be really bothered for the ancient format...
> 

Yes, I share your feelings. Currently, we do not seem to have any
practical issues. To maintain consistent behavior between the two inode
types and to keep the code simple, I retained the -= sb->s_blocksize
operation. Would you suggest that we should at least address the extents
inodes by removing the -=sb->s_blocksize now?

Thanks,
Yi.

Re: [PATCH 2/4] ext4: fix incorrect punch max_end

[Jan Kara]

On Wed 30-04-25 16:44:25, Zhang Yi wrote:
> On 2025/4/30 16:18, Jan Kara wrote:
> > On Wed 30-04-25 09:12:59, Zhang Yi wrote:
> >> From: Zhang Yi <yi.zhang@huawei.com>
> >>
> >> For the extents inodes, the maxbytes should be sb->s_maxbytes instead of
> >> sbi->s_bitmap_maxbytes. Correct the maxbytes value to correct the
> >> behavior of punch hole.
> >>
> >> Fixes: 2da376228a24 ("ext4: limit length to bitmap_maxbytes - blocksize in punch_hole")
> >> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
> > 
> > Thinking about this some more...
> > 
> >> @@ -4015,6 +4015,12 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
> >>  	trace_ext4_punch_hole(inode, offset, length, 0);
> >>  	WARN_ON_ONCE(!inode_is_locked(inode));
> >>  
> >> +	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
> >> +		max_end = sb->s_maxbytes;
> >> +	else
> >> +		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
> >> +	max_end -= sb->s_blocksize;
> > 
> > I think the -= sb->s_blocksize is needed only for indirect-block based
> > scheme (due to an implementation quirk in ext4_ind_remove_space()). But
> > ext4_ext_remove_space() should be fine with punch hole ending right at
> > sb->s_maxbytes. And since I find it somewhat odd that you can create file
> > upto s_maxbytes but cannot punch hole to the end, it'd limit that behavior
> > as much as possible. Ideally we'd fix ext4_ind_remove_space() but I can't
> > be really bothered for the ancient format...
> > 
> 
> Yes, I share your feelings. Currently, we do not seem to have any
> practical issues. To maintain consistent behavior between the two inode
> types and to keep the code simple, I retained the -= sb->s_blocksize
> operation. Would you suggest that we should at least address the extents
> inodes by removing the -=sb->s_blocksize now?

Yes, what I'm suggesting is that we keep -=sb->s_blocksize specific for the
case !ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS).

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH 2/4] ext4: fix incorrect punch max_end

[Zhang Yi]

On 2025/4/30 18:09, Jan Kara wrote:
> On Wed 30-04-25 16:44:25, Zhang Yi wrote:
>> On 2025/4/30 16:18, Jan Kara wrote:
>>> On Wed 30-04-25 09:12:59, Zhang Yi wrote:
>>>> From: Zhang Yi <yi.zhang@huawei.com>
>>>>
>>>> For the extents inodes, the maxbytes should be sb->s_maxbytes instead of
>>>> sbi->s_bitmap_maxbytes. Correct the maxbytes value to correct the
>>>> behavior of punch hole.
>>>>
>>>> Fixes: 2da376228a24 ("ext4: limit length to bitmap_maxbytes - blocksize in punch_hole")
>>>> Signed-off-by: Zhang Yi <yi.zhang@huawei.com>
>>>
>>> Thinking about this some more...
>>>
>>>> @@ -4015,6 +4015,12 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length)
>>>>  	trace_ext4_punch_hole(inode, offset, length, 0);
>>>>  	WARN_ON_ONCE(!inode_is_locked(inode));
>>>>  
>>>> +	if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
>>>> +		max_end = sb->s_maxbytes;
>>>> +	else
>>>> +		max_end = EXT4_SB(sb)->s_bitmap_maxbytes;
>>>> +	max_end -= sb->s_blocksize;
>>>
>>> I think the -= sb->s_blocksize is needed only for indirect-block based
>>> scheme (due to an implementation quirk in ext4_ind_remove_space()). But
>>> ext4_ext_remove_space() should be fine with punch hole ending right at
>>> sb->s_maxbytes. And since I find it somewhat odd that you can create file
>>> upto s_maxbytes but cannot punch hole to the end, it'd limit that behavior
>>> as much as possible. Ideally we'd fix ext4_ind_remove_space() but I can't
>>> be really bothered for the ancient format...
>>>
>>
>> Yes, I share your feelings. Currently, we do not seem to have any
>> practical issues. To maintain consistent behavior between the two inode
>> types and to keep the code simple, I retained the -= sb->s_blocksize
>> operation. Would you suggest that we should at least address the extents
>> inodes by removing the -=sb->s_blocksize now?
> 
> Yes, what I'm suggesting is that we keep -=sb->s_blocksize specific for the
> case !ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS).
> 

Sure. Let's do it.

Thanks,
Yi.