[PATCH v3 0/4] Fix two issue about ext4 extended attribute

[PATCH v3 0/4] Fix two issue about ext4 extended attribute

[Ye Bin]

From: Ye Bin <yebin10@huawei.com>

Diff v3 vs v2:
1. Remove patch [2-3].
2. Modify change log according to Bagas Sanjaya's suggestions.

Diff v2 vs v1:
1.Modify commit message about "ext4: fix WARNING in ext4_expand_extra_isize_ea"
2.Modify the indentation of arguments about "ext4: rename xattr_find_entry()
and __xattr_check_inode()"

This patchset fix two issues:
1. Patch [1]-[4] fix WARNING in ext4_expand_extra_isize_ea.
2. Patch [6] fix inode leak in 'ext4_xattr_inode_create()'.
3. Patch [5] is cleanup.

Ye Bin (4):
  ext4: fix WARNING in ext4_expand_extra_isize_ea
  ext4: allocate extended attribute value in vmalloc area
  ext4: rename xattr_find_entry() and __xattr_check_inode()
  ext4: fix inode leak in 'ext4_xattr_inode_create()'

 fs/ext4/xattr.c | 42 +++++++++++++++++++++++-------------------
 fs/ext4/xattr.h | 11 +++--------
 2 files changed, 26 insertions(+), 27 deletions(-)

-- 
2.31.1

[PATCH v3 1/4] ext4: fix WARNING in ext4_expand_extra_isize_ea

[Ye Bin]

From: Ye Bin <yebin10@huawei.com>

Syzbot found the following issue:
------------[ cut here ]------------
WARNING: CPU: 1 PID: 3631 at mm/page_alloc.c:5534 __alloc_pages+0x30a/0x560 mm/page_alloc.c:5534
Modules linked in:
CPU: 1 PID: 3631 Comm: syz-executor261 Not tainted 6.1.0-rc6-syzkaller-00308-g644e9524388a #0
Hardware name: Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022
RIP: 0010:__alloc_pages+0x30a/0x560 mm/page_alloc.c:5534
RSP: 0018:ffffc90003ccf080 EFLAGS: 00010246
RAX: ffffc90003ccf0e0 RBX: 000000000000000c RCX: 0000000000000000
RDX: 0000000000000028 RSI: 0000000000000000 RDI: ffffc90003ccf108
RBP: ffffc90003ccf198 R08: dffffc0000000000 R09: ffffc90003ccf0e0
R10: fffff52000799e21 R11: 1ffff92000799e1c R12: 0000000000040c40
R13: 1ffff92000799e18 R14: dffffc0000000000 R15: 1ffff92000799e14
FS:  0000555555c10300(0000) GS:ffff8880b9900000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007ffc36f70000 CR3: 00000000744ad000 CR4: 00000000003506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 __alloc_pages_node include/linux/gfp.h:223 [inline]
 alloc_pages_node include/linux/gfp.h:246 [inline]
 __kmalloc_large_node+0x8a/0x1a0 mm/slab_common.c:1096
 __do_kmalloc_node mm/slab_common.c:943 [inline]
 __kmalloc+0xfe/0x1a0 mm/slab_common.c:968
 kmalloc include/linux/slab.h:558 [inline]
 ext4_xattr_move_to_block fs/ext4/xattr.c:2558 [inline]
 ext4_xattr_make_inode_space fs/ext4/xattr.c:2673 [inline]
 ext4_expand_extra_isize_ea+0xe3f/0x1cd0 fs/ext4/xattr.c:2765
 __ext4_expand_extra_isize+0x2b8/0x3f0 fs/ext4/inode.c:5857
 ext4_try_to_expand_extra_isize fs/ext4/inode.c:5900 [inline]
 __ext4_mark_inode_dirty+0x51a/0x670 fs/ext4/inode.c:5978
 ext4_inline_data_truncate+0x548/0xd00 fs/ext4/inline.c:2021
 ext4_truncate+0x341/0xeb0 fs/ext4/inode.c:4221
 ext4_process_orphan+0x1aa/0x2d0 fs/ext4/orphan.c:339
 ext4_orphan_cleanup+0xb60/0x1340 fs/ext4/orphan.c:474
 __ext4_fill_super fs/ext4/super.c:5515 [inline]
 ext4_fill_super+0x80ed/0x8610 fs/ext4/super.c:5643
 get_tree_bdev+0x400/0x620 fs/super.c:1324
 vfs_get_tree+0x88/0x270 fs/super.c:1531
 do_new_mount+0x289/0xad0 fs/namespace.c:3040
 do_mount fs/namespace.c:3383 [inline]
 __do_sys_mount fs/namespace.c:3591 [inline]
 __se_sys_mount+0x2d3/0x3c0 fs/namespace.c:3568
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x3d/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x63/0xcd
 </TASK>

Reason is allocate 16M memory by kmalloc, but MAX_ORDER is 11, kmalloc
can allocate maximum size memory is 4M.
XATTR_SIZE_MAX is currently 64K, but EXT4_XATTR_SIZE_MAX is 16M, so
'ext4_xattr_check_entries()' regards this length as legal. Then trigger
warning in 'ext4_xattr_move_to_block()'.
To solve above issue, change EXT4_XATTR_SIZE_MAX from 16M to 64K.
As VFS limit extended attribute maximum size to 64K. So we can assume
that there will be no extended attribute with a length greater than 64K.

Reported-by: syzbot+4d99a966fd74bdeeec36@syzkaller.appspotmail.com
Fixes: 54dd0e0a1b25 ("ext4: add extra checks to ext4_xattr_block_get()")
Signed-off-by: Ye Bin <yebin10@huawei.com>
---
 fs/ext4/xattr.h | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

diff --git a/fs/ext4/xattr.h b/fs/ext4/xattr.h
index 824faf0b15a8..c71e582b1007 100644
--- a/fs/ext4/xattr.h
+++ b/fs/ext4/xattr.h
@@ -71,15 +71,10 @@ struct ext4_xattr_entry {
 #define IFIRST(hdr) ((struct ext4_xattr_entry *)((hdr)+1))
 
 /*
- * XATTR_SIZE_MAX is currently 64k, but for the purposes of checking
- * for file system consistency errors, we use a somewhat bigger value.
- * This allows XATTR_SIZE_MAX to grow in the future, but by using this
- * instead of INT_MAX for certain consistency checks, we don't need to
- * worry about arithmetic overflows.  (Actually XATTR_SIZE_MAX is
- * defined in include/uapi/linux/limits.h, so changing it is going
- * not going to be trivial....)
+ * Use XATTR_SIZE_MAX to checking for file system consistency errors. Extended
+ * attribute length exceed XATTR_SIZE_MAX is illegal.
  */
-#define EXT4_XATTR_SIZE_MAX (1 << 24)
+#define EXT4_XATTR_SIZE_MAX XATTR_SIZE_MAX
 
 /*
  * The minimum size of EA value when you start storing it in an external inode
-- 
2.31.1

Re: [PATCH v3 1/4] ext4: fix WARNING in ext4_expand_extra_isize_ea

[Theodore Ts'o]

In general, please don't use commit summaries like:

	ext4: fix WARNING in ext4_expand_extra_isize_ea
	
The commit summary should be a summary of what the change *does*.  A
description of *why* the change is needed is properly placed in the
body of the description.  So something like "ext4: clamp
EXT4_XATTR_SIZE_MAX to XATTR_SIZE_MAX" would be a better one-line
commit summary.

That being said, I've decided to not apply this patch.  A 16 MB
extended attribute size is not _that_ insane, although admittedly it's
quite bit larger than any other file system or OS supports (unless you
include Solaris and NTFS's alternate data forks as "extended
attributes").

I personally don't think the mm/page_alloc warning is that big of a
deal (even if it does cause a Syzbot report), and if you really do
care about suppressing the warning, using the kvmalloc() instead of
kmalloc() will do that --- and it does make ext4 friendly in the case
of where we are storing, say, a 30KB Windows ACL as an extended
attribute, since it avoided needing to make a order 3 page allocation.
(Which again, as far as I'm concerned, is more important than the
suppressing the page_alloc warning.)

We may ultimately decide to clamp EXT4_XATTR_SIZE_MAX to
XATTR_SIZE_MAX, but one could imagine some system attribute that in
the future might want to be larger than 32k, that wouldn't be accessed
via the standard xattr API.  That seems unlikely, but I want to keep
our options open.

Cheers,

					- Ted

[PATCH v3 2/4] ext4: allocate extended attribute value in vmalloc area

[Ye Bin]

From: Ye Bin <yebin10@huawei.com>

Now, extended attribute value maximum length is 64K. The memory requested
here does not need continuous physical addresses, so it is appropriate to
use kvmalloc to request memory. At the same time, it can also cope with
the situation that the extended attribute will become longer in the future.

Signed-off-by: Ye Bin <yebin10@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/xattr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 718ef3987f94..23cfefc6d262 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -2556,7 +2556,7 @@ static int ext4_xattr_move_to_block(handle_t *handle, struct inode *inode,
 
 	is = kzalloc(sizeof(struct ext4_xattr_ibody_find), GFP_NOFS);
 	bs = kzalloc(sizeof(struct ext4_xattr_block_find), GFP_NOFS);
-	buffer = kmalloc(value_size, GFP_NOFS);
+	buffer = kvmalloc(value_size, GFP_NOFS);
 	b_entry_name = kmalloc(entry->e_name_len + 1, GFP_NOFS);
 	if (!is || !bs || !buffer || !b_entry_name) {
 		error = -ENOMEM;
@@ -2608,7 +2608,7 @@ static int ext4_xattr_move_to_block(handle_t *handle, struct inode *inode,
 	error = 0;
 out:
 	kfree(b_entry_name);
-	kfree(buffer);
+	kvfree(buffer);
 	if (is)
 		brelse(is->iloc.bh);
 	if (bs)
-- 
2.31.1

Re: [PATCH v3 2/4] ext4: allocate extended attribute value in vmalloc area

[Theodore Ts'o]

On Thu, Dec 08, 2022 at 10:32:31AM +0800, Ye Bin wrote:
> From: Ye Bin <yebin10@huawei.com>
> 
> Now, extended attribute value maximum length is 64K. The memory requested
> here does not need continuous physical addresses, so it is appropriate to
> use kvmalloc to request memory. At the same time, it can also cope with
> the situation that the extended attribute will become longer in the future.
> 
> Signed-off-by: Ye Bin <yebin10@huawei.com>
> Reviewed-by: Jan Kara <jack@suse.cz>

Applied, thanks.

					- Ted

[PATCH v3 3/4] ext4: rename xattr_find_entry() and __xattr_check_inode()

[Ye Bin]

From: Ye Bin <yebin10@huawei.com>

Prepend ext4 prefix to function names of xattr_find_entry() and
__xattr_check_inode() for consistency with other functions in ext4
xattr module.

Signed-off-by: Ye Bin <yebin10@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/xattr.c | 35 ++++++++++++++++++-----------------
 1 file changed, 18 insertions(+), 17 deletions(-)

diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 23cfefc6d262..99dacb0393fa 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -260,8 +260,9 @@ __ext4_xattr_check_block(struct inode *inode, struct buffer_head *bh,
 
 
 static int
-__xattr_check_inode(struct inode *inode, struct ext4_xattr_ibody_header *header,
-			 void *end, const char *function, unsigned int line)
+__ext4_xattr_check_inode(struct inode *inode,
+			 struct ext4_xattr_ibody_header *header, void *end,
+			 const char *function, unsigned int line)
 {
 	int error = -EFSCORRUPTED;
 
@@ -276,12 +277,12 @@ __xattr_check_inode(struct inode *inode, struct ext4_xattr_ibody_header *header,
 	return error;
 }
 
-#define xattr_check_inode(inode, header, end) \
-	__xattr_check_inode((inode), (header), (end), __func__, __LINE__)
+#define ext4_xattr_check_inode(inode, header, end) \
+	__ext4_xattr_check_inode((inode), (header), (end), __func__, __LINE__)
 
 static int
-xattr_find_entry(struct inode *inode, struct ext4_xattr_entry **pentry,
-		 void *end, int name_index, const char *name, int sorted)
+ext4_xattr_find_entry(struct inode *inode, struct ext4_xattr_entry **pentry,
+		      void *end, int name_index, const char *name, int sorted)
 {
 	struct ext4_xattr_entry *entry, *next;
 	size_t name_len;
@@ -554,7 +555,7 @@ ext4_xattr_block_get(struct inode *inode, int name_index, const char *name,
 	ext4_xattr_block_cache_insert(ea_block_cache, bh);
 	entry = BFIRST(bh);
 	end = bh->b_data + bh->b_size;
-	error = xattr_find_entry(inode, &entry, end, name_index, name, 1);
+	error = ext4_xattr_find_entry(inode, &entry, end, name_index, name, 1);
 	if (error)
 		goto cleanup;
 	size = le32_to_cpu(entry->e_value_size);
@@ -605,11 +606,11 @@ ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name,
 	raw_inode = ext4_raw_inode(&iloc);
 	header = IHDR(inode, raw_inode);
 	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
-	error = xattr_check_inode(inode, header, end);
+	error = ext4_xattr_check_inode(inode, header, end);
 	if (error)
 		goto cleanup;
 	entry = IFIRST(header);
-	error = xattr_find_entry(inode, &entry, end, name_index, name, 0);
+	error = ext4_xattr_find_entry(inode, &entry, end, name_index, name, 0);
 	if (error)
 		goto cleanup;
 	size = le32_to_cpu(entry->e_value_size);
@@ -750,7 +751,7 @@ ext4_xattr_ibody_list(struct dentry *dentry, char *buffer, size_t buffer_size)
 	raw_inode = ext4_raw_inode(&iloc);
 	header = IHDR(inode, raw_inode);
 	end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
-	error = xattr_check_inode(inode, header, end);
+	error = ext4_xattr_check_inode(inode, header, end);
 	if (error)
 		goto cleanup;
 	error = ext4_xattr_list_entries(dentry, IFIRST(header),
@@ -836,7 +837,7 @@ int ext4_get_inode_usage(struct inode *inode, qsize_t *usage)
 		raw_inode = ext4_raw_inode(&iloc);
 		header = IHDR(inode, raw_inode);
 		end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
-		ret = xattr_check_inode(inode, header, end);
+		ret = ext4_xattr_check_inode(inode, header, end);
 		if (ret)
 			goto out;
 
@@ -1839,8 +1840,8 @@ ext4_xattr_block_find(struct inode *inode, struct ext4_xattr_info *i,
 		bs->s.first = BFIRST(bs->bh);
 		bs->s.end = bs->bh->b_data + bs->bh->b_size;
 		bs->s.here = bs->s.first;
-		error = xattr_find_entry(inode, &bs->s.here, bs->s.end,
-					 i->name_index, i->name, 1);
+		error = ext4_xattr_find_entry(inode, &bs->s.here, bs->s.end,
+					      i->name_index, i->name, 1);
 		if (error && error != -ENODATA)
 			return error;
 		bs->s.not_found = error;
@@ -2199,12 +2200,12 @@ int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i,
 	is->s.here = is->s.first;
 	is->s.end = (void *)raw_inode + EXT4_SB(inode->i_sb)->s_inode_size;
 	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
-		error = xattr_check_inode(inode, header, is->s.end);
+		error = ext4_xattr_check_inode(inode, header, is->s.end);
 		if (error)
 			return error;
 		/* Find the named attribute. */
-		error = xattr_find_entry(inode, &is->s.here, is->s.end,
-					 i->name_index, i->name, 0);
+		error = ext4_xattr_find_entry(inode, &is->s.here, is->s.end,
+					      i->name_index, i->name, 0);
 		if (error && error != -ENODATA)
 			return error;
 		is->s.not_found = error;
@@ -2719,7 +2720,7 @@ int ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize,
 	min_offs = end - base;
 	total_ino = sizeof(struct ext4_xattr_ibody_header) + sizeof(u32);
 
-	error = xattr_check_inode(inode, header, end);
+	error = ext4_xattr_check_inode(inode, header, end);
 	if (error)
 		goto cleanup;
 
-- 
2.31.1

Re: [PATCH v3 3/4] ext4: rename xattr_find_entry() and __xattr_check_inode()

[Theodore Ts'o]

On Thu, Dec 08, 2022 at 10:32:32AM +0800, Ye Bin wrote:
> From: Ye Bin <yebin10@huawei.com>
> 
> Prepend ext4 prefix to function names of xattr_find_entry() and
> __xattr_check_inode() for consistency with other functions in ext4
> xattr module.

The reason why these function names don't have the ext4 prefix is
because functions are static, so there is need to worry about
namespace pollution.

For static functions, there is no real need to add the ext4 prefix, so
why not keep the functions shorter?  Yes, there are places where
static functions have the ext4 prefix.  The main reason why we have
changed it is that it's not worth the code churn.  Similarly, adding
the ext4 prefix is *definitely* not worth the code churn.

Cheers,

					- Ted

[PATCH v3 4/4] ext4: fix inode leak in 'ext4_xattr_inode_create()'

[Ye Bin]

From: Ye Bin <yebin10@huawei.com>

There is issue as follows when do setxattr with inject fault:
[localhost]#fsck.ext4  -fn  /dev/sda
e2fsck 1.46.6-rc1 (12-Sep-2022)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Unattached zero-length inode 15.  Clear? no

Unattached inode 15
Connect to /lost+found? no

Pass 5: Checking group summary information

/dev/sda: ********** WARNING: Filesystem still has errors **********

/dev/sda: 15/655360 files (0.0% non-contiguous), 66755/2621440 blocks

This occurs in 'ext4_xattr_inode_create()'. If 'ext4_mark_inode_dirty()'
fails, dropping i_nlink of the inode is needed. Or will lead to inode leak.

Signed-off-by: Ye Bin <yebin10@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/xattr.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 99dacb0393fa..aad7de2e366a 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -1442,6 +1442,9 @@ static struct inode *ext4_xattr_inode_create(handle_t *handle,
 		if (!err)
 			err = ext4_inode_attach_jinode(ea_inode);
 		if (err) {
+			if (ext4_xattr_inode_dec_ref(handle, ea_inode))
+				ext4_warning_inode(ea_inode,
+					"cleanup dec ref error %d", err);
 			iput(ea_inode);
 			return ERR_PTR(err);
 		}
-- 
2.31.1

Re: [PATCH v3 4/4] ext4: fix inode leak in 'ext4_xattr_inode_create()'

[Theodore Ts'o]

On Thu, Dec 08, 2022 at 10:32:33AM +0800, Ye Bin wrote:
> From: Ye Bin <yebin10@huawei.com>
> 
> There is issue as follows when do setxattr with inject fault:
> [localhost]#fsck.ext4  -fn  /dev/sda
> e2fsck 1.46.6-rc1 (12-Sep-2022)
> Pass 1: Checking inodes, blocks, and sizes
> Pass 2: Checking directory structure
> Pass 3: Checking directory connectivity
> Pass 4: Checking reference counts
> Unattached zero-length inode 15.  Clear? no
> 
> Unattached inode 15
> Connect to /lost+found? no
> 
> Pass 5: Checking group summary information
> 
> /dev/sda: ********** WARNING: Filesystem still has errors **********
> 
> /dev/sda: 15/655360 files (0.0% non-contiguous), 66755/2621440 blocks
> 
> This occurs in 'ext4_xattr_inode_create()'. If 'ext4_mark_inode_dirty()'
> fails, dropping i_nlink of the inode is needed. Or will lead to inode leak.
> 
> Signed-off-by: Ye Bin <yebin10@huawei.com>
> Reviewed-by: Jan Kara <jack@suse.cz>

Applied, thanks.

					- Ted