Wind River Support Network

Description

In the Linux kernel, the following vulnerability has been resolved:EOL][EOL]f2fs: fix to do sanity check on ino and xnid[EOL][EOL]syzbot reported a f2fs bug as below:[EOL][EOL]INFO: task syz-executor140:5308 blocked for more than 143 seconds.[EOL]      Not tainted 6.14.0-rc7-syzkaller-00069-g81e4f8d68c66 #0[EOL]"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.[EOL]task:syz-executor140 state:D stack:24016 pid:5308  tgid:5308  ppid:5306   task_flags:0x400140 flags:0x00000006[EOL]Call Trace:[EOL] <TASK>[EOL] context_switch kernel/sched/core.c:5378 [inline][EOL] __schedule+0x190e/0x4c90 kernel/sched/core.c:6765[EOL] __schedule_loop kernel/sched/core.c:6842 [inline][EOL] schedule+0x14b/0x320 kernel/sched/core.c:6857[EOL] io_schedule+0x8d/0x110 kernel/sched/core.c:7690[EOL] folio_wait_bit_common+0x839/0xee0 mm/filemap.c:1317[EOL] __folio_lock mm/filemap.c:1664 [inline][EOL] folio_lock include/linux/pagemap.h:1163 [inline][EOL] __filemap_get_folio+0x147/0xb40 mm/filemap.c:1917[EOL] pagecache_get_page+0x2c/0x130 mm/folio-compat.c:87[EOL] find_get_page_flags include/linux/pagemap.h:842 [inline][EOL] f2fs_grab_cache_page+0x2b/0x320 fs/f2fs/f2fs.h:2776[EOL] __get_node_page+0x131/0x11b0 fs/f2fs/node.c:1463[EOL] read_xattr_block+0xfb/0x190 fs/f2fs/xattr.c:306[EOL] lookup_all_xattrs fs/f2fs/xattr.c:355 [inline][EOL] f2fs_getxattr+0x676/0xf70 fs/f2fs/xattr.c:533[EOL] __f2fs_get_acl+0x52/0x870 fs/f2fs/acl.c:179[EOL] f2fs_acl_create fs/f2fs/acl.c:375 [inline][EOL] f2fs_init_acl+0xd7/0x9b0 fs/f2fs/acl.c:418[EOL] f2fs_init_inode_metadata+0xa0f/0x1050 fs/f2fs/dir.c:539[EOL] f2fs_add_inline_entry+0x448/0x860 fs/f2fs/inline.c:666[EOL] f2fs_add_dentry+0xba/0x1e0 fs/f2fs/dir.c:765[EOL] f2fs_do_add_link+0x28c/0x3a0 fs/f2fs/dir.c:808[EOL] f2fs_add_link fs/f2fs/f2fs.h:3616 [inline][EOL] f2fs_mknod+0x2e8/0x5b0 fs/f2fs/namei.c:766[EOL] vfs_mknod+0x36d/0x3b0 fs/namei.c:4191[EOL] unix_bind_bsd net/unix/af_unix.c:1286 [inline][EOL] unix_bind+0x563/0xe30 net/unix/af_unix.c:1379[EOL] __sys_bind_socket net/socket.c:1817 [inline][EOL] __sys_bind+0x1e4/0x290 net/socket.c:1848[EOL] __do_sys_bind net/socket.c:1853 [inline][EOL] __se_sys_bind net/socket.c:1851 [inline][EOL] __x64_sys_bind+0x7a/0x90 net/socket.c:1851[EOL] do_syscall_x64 arch/x86/entry/common.c:52 [inline][EOL] do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83[EOL] entry_SYSCALL_64_after_hwframe+0x77/0x7f[EOL][EOL]Let's dump and check metadata of corrupted inode, it shows its xattr_nid[EOL]is the same to its i_ino.[EOL][EOL]dump.f2fs -i 3 chaseyu.img.raw[EOL]i_xattr_nid                             [0x       3 : 3][EOL][EOL]So that, during mknod in the corrupted directory, it tries to get and[EOL]lock inode page twice, result in deadlock.[EOL][EOL]- f2fs_mknod[EOL] - f2fs_add_inline_entry[EOL]  - f2fs_get_inode_page --- lock dir's inode page[EOL]   - f2fs_init_acl[EOL]    - f2fs_acl_create(dir,..)[EOL]     - __f2fs_get_acl[EOL]      - f2fs_getxattr[EOL]       - lookup_all_xattrs[EOL]        - __get_node_page --- try to lock dir's inode page[EOL][EOL]In order to fix this, let's add sanity check on ino and xnid.

CREATE(Triage):(User=admin) [CVE-2025-38347 (https://nvd.nist.gov/vuln/detail/CVE-2025-38347)