diff mbox series

[v2] cachefiles: Fix KASAN slab-out-of-bounds in cachefiles_set_volume_xattr

Message ID 20220405142810.8208-1-dwysocha@redhat.com (mailing list archive)
State New, archived
Headers show
Series [v2] cachefiles: Fix KASAN slab-out-of-bounds in cachefiles_set_volume_xattr | expand

Commit Message

David Wysochanski April 5, 2022, 2:28 p.m. UTC 
Use the actual length of volume coherency data when setting the
xattr to avoid the following KASAN report.

 BUG: KASAN: slab-out-of-bounds in cachefiles_set_volume_xattr+0xa0/0x350 [cachefiles]
 Write of size 4 at addr ffff888101e02af4 by task kworker/6:0/1347

 CPU: 6 PID: 1347 Comm: kworker/6:0 Kdump: loaded Not tainted 5.18.0-rc1-nfs-fscache-netfs+ #13
 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-4.fc34 04/01/2014
 Workqueue: events fscache_create_volume_work [fscache]
 Call Trace:
  <TASK>
  dump_stack_lvl+0x45/0x5a
  print_report.cold+0x5e/0x5db
  ? __lock_text_start+0x8/0x8
  ? cachefiles_set_volume_xattr+0xa0/0x350 [cachefiles]
  kasan_report+0xab/0x120
  ? cachefiles_set_volume_xattr+0xa0/0x350 [cachefiles]
  kasan_check_range+0xf5/0x1d0
  memcpy+0x39/0x60
  cachefiles_set_volume_xattr+0xa0/0x350 [cachefiles]
  cachefiles_acquire_volume+0x2be/0x500 [cachefiles]
  ? __cachefiles_free_volume+0x90/0x90 [cachefiles]
  fscache_create_volume_work+0x68/0x160 [fscache]
  process_one_work+0x3b7/0x6a0
  worker_thread+0x2c4/0x650
  ? process_one_work+0x6a0/0x6a0
  kthread+0x16c/0x1a0
  ? kthread_complete_and_exit+0x20/0x20
  ret_from_fork+0x22/0x30
  </TASK>

 Allocated by task 1347:
  kasan_save_stack+0x1e/0x40
  __kasan_kmalloc+0x81/0xa0
  cachefiles_set_volume_xattr+0x76/0x350 [cachefiles]
  cachefiles_acquire_volume+0x2be/0x500 [cachefiles]
  fscache_create_volume_work+0x68/0x160 [fscache]
  process_one_work+0x3b7/0x6a0
  worker_thread+0x2c4/0x650
  kthread+0x16c/0x1a0
  ret_from_fork+0x22/0x30

 The buggy address belongs to the object at ffff888101e02af0
 which belongs to the cache kmalloc-8 of size 8
 The buggy address is located 4 bytes inside of
 8-byte region [ffff888101e02af0, ffff888101e02af8)

 The buggy address belongs to the physical page:
 page:00000000a2292d70 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x101e02
 flags: 0x17ffffc0000200(slab|node=0|zone=2|lastcpupid=0x1fffff)
 raw: 0017ffffc0000200 0000000000000000 dead000000000001 ffff888100042280
 raw: 0000000000000000 0000000080660066 00000001ffffffff 0000000000000000
 page dumped because: kasan: bad access detected

 Memory state around the buggy address:
 ffff888101e02980: fc 00 fc fc fc fc 00 fc fc fc fc 00 fc fc fc fc
 ffff888101e02a00: 00 fc fc fc fc 00 fc fc fc fc 00 fc fc fc fc 00
 >ffff888101e02a80: fc fc fc fc 00 fc fc fc fc 00 fc fc fc fc 04 fc
                                                            ^
 ffff888101e02b00: fc fc fc 00 fc fc fc fc 00 fc fc fc fc 00 fc fc
 ffff888101e02b80: fc fc 00 fc fc fc fc 00 fc fc fc fc 00 fc fc fc
 ==================================================================

Fixes: 413a4a6b0b55 "cachefiles: Fix volume coherency attribute"
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
---
 fs/cachefiles/xattr.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

Comments

David Howells April 8, 2022, 10:13 p.m. UTC | #1 
Dave Wysochanski <dwysocha@redhat.com> wrote:

> -	len += sizeof(*buf);
> -	buf = kmalloc(len, GFP_KERNEL);
> +	buf = kmalloc(sizeof(*buf) + len, GFP_KERNEL);

Okay, your V2 is wrong and your V1 is correct (len must include the reserverd
word so that it gets saved onto disk).

David
diff mbox series

Patch

diff --git a/fs/cachefiles/xattr.c b/fs/cachefiles/xattr.c
index 35465109d9c4..f13b642c1d14 100644
--- a/fs/cachefiles/xattr.c
+++ b/fs/cachefiles/xattr.c
@@ -198,8 +198,7 @@  bool cachefiles_set_volume_xattr(struct cachefiles_volume *volume)
 
 	_enter("%x,#%d", volume->vcookie->debug_id, len);
 
-	len += sizeof(*buf);
-	buf = kmalloc(len, GFP_KERNEL);
+	buf = kmalloc(sizeof(*buf) + len, GFP_KERNEL);
 	if (!buf)
 		return false;
 	buf->reserved = cpu_to_be32(0);