Description

In the Linux kernel, the following vulnerability has been resolved: f2fs: use global inline_xattr_slab instead of per-sb slab cache As Hong Yun reported in mailing list: loop7: detected capacity change from 0 to 131072 ------------[ cut here ]------------ kmem_cache of name 'f2fs_xattr_entry-7:7' already exists WARNING: CPU: 0 PID: 24426 at mm/slab_common.c:110 kmem_cache_sanity_check mm/slab_common.c:109 [inline] WARNING: CPU: 0 PID: 24426 at mm/slab_common.c:110 __kmem_cache_create_args+0xa6/0x320 mm/slab_common.c:307 CPU: 0 UID: 0 PID: 24426 Comm: syz.7.1370 Not tainted 6.17.0-rc4 #1 PREEMPT(full) Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014 RIP: 0010:kmem_cache_sanity_check mm/slab_common.c:109 [inline] RIP: 0010:__kmem_cache_create_args+0xa6/0x320 mm/slab_common.c:307 Call Trace:  __kmem_cache_create include/linux/slab.h:353 [inline]  f2fs_kmem_cache_create fs/f2fs/f2fs.h:2943 [inline]  f2fs_init_xattr_caches+0xa5/0xe0 fs/f2fs/xattr.c:843  f2fs_fill_super+0x1645/0x2620 fs/f2fs/super.c:4918  get_tree_bdev_flags+0x1fb/0x260 fs/super.c:1692  vfs_get_tree+0x43/0x140 fs/super.c:1815  do_new_mount+0x201/0x550 fs/namespace.c:3808  do_mount fs/namespace.c:4136 [inline]  __do_sys_mount fs/namespace.c:4347 [inline]  __se_sys_mount+0x298/0x2f0 fs/namespace.c:4324  do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]  do_syscall_64+0x8e/0x3a0 arch/x86/entry/syscall_64.c:94  entry_SYSCALL_64_after_hwframe+0x76/0x7e The bug can be reproduced w/ below scripts: - mount /dev/vdb /mnt1 - mount /dev/vdc /mnt2 - umount /mnt1 - mounnt /dev/vdb /mnt1 The reason is if we created two slab caches, named f2fs_xattr_entry-7:3 and f2fs_xattr_entry-7:7, and they have the same slab size. Actually, slab system will only create one slab cache core structure which has slab name of "f2fs_xattr_entry-7:3", and two slab caches share the same structure and cache address. So, if we destroy f2fs_xattr_entry-7:3 cache w/ cache address, it will decrease reference count of slab cache, rather than release slab cache entirely, since there is one more user has referenced the cache. Then, if we try to create slab cache w/ name "f2fs_xattr_entry-7:3" again, slab system will find that there is existed cache which has the same name and trigger the warning. Let's changes to use global inline_xattr_slab instead of per-sb slab cache for fixing.

Affected Products 15

References 7

• git.kernel.org https://git.kernel.org/stable/c/1eb0b130196bcbc56c5c80c83139fa70c0aa82c5
  Patch
• git.kernel.org https://git.kernel.org/stable/c/1f27ef42bb0b7c0740c5616ec577ec188b8a1d05
  Patch
• git.kernel.org https://git.kernel.org/stable/c/474cc3ed37436ddfd63cac8dbffe3b1e219e9100
  Patch
• git.kernel.org https://git.kernel.org/stable/c/72ce19dfed162da6e430467333b2da70471d08a4
  Patch
• git.kernel.org https://git.kernel.org/stable/c/93d30fe19660dec6bf1bd3d5c186c1c737b21aa5
  Patch
• git.kernel.org https://git.kernel.org/stable/c/be4c3a3c6c2304a8fcd14095d18d26f0cc4e222a
  Patch
• git.kernel.org https://git.kernel.org/stable/c/e6d828eae00ec192e18c2ddaa2fd32050a96048a
  Patch

Remediation

• git.kernel.org https://git.kernel.org/stable/c/1eb0b130196bcbc56c5c80c83139fa70c0aa82c5
  Patch
• git.kernel.org https://git.kernel.org/stable/c/1f27ef42bb0b7c0740c5616ec577ec188b8a1d05
  Patch
• git.kernel.org https://git.kernel.org/stable/c/474cc3ed37436ddfd63cac8dbffe3b1e219e9100
  Patch
• git.kernel.org https://git.kernel.org/stable/c/72ce19dfed162da6e430467333b2da70471d08a4
  Patch
• git.kernel.org https://git.kernel.org/stable/c/93d30fe19660dec6bf1bd3d5c186c1c737b21aa5
  Patch
• git.kernel.org https://git.kernel.org/stable/c/be4c3a3c6c2304a8fcd14095d18d26f0cc4e222a
  Patch
• git.kernel.org https://git.kernel.org/stable/c/e6d828eae00ec192e18c2ddaa2fd32050a96048a
  Patch