In the Linux kernel, the following vulnerability has been resolved:
f2fs: fix to don't dirty inode for readonly filesystem
syzbot reports f2fs bug as below:
kernel BUG at fs/f2fs/inode.c:933! RIP: 0010:f2fsevictinode+0x1576/0x1590 fs/f2fs/inode.c:933 Call Trace: evict+0x2a4/0x620 fs/inode.c:664 disposelist fs/inode.c:697 [inline] evictinodes+0x5f8/0x690 fs/inode.c:747 genericshutdownsuper+0x9d/0x2c0 fs/super.c:675 killblocksuper+0x44/0x90 fs/super.c:1667 killf2fssuper+0x303/0x3b0 fs/f2fs/super.c:4894 deactivatelockedsuper+0xc1/0x130 fs/super.c:484 cleanupmnt+0x426/0x4c0 fs/namespace.c:1256 taskworkrun+0x24a/0x300 kernel/taskwork.c:180 ptracenotify+0x2cd/0x380 kernel/signal.c:2399 ptracereportsyscall include/linux/ptrace.h:411 [inline] ptracereportsyscallexit include/linux/ptrace.h:473 [inline] syscallexitwork kernel/entry/common.c:251 [inline] syscallexittousermodeprepare kernel/entry/common.c:278 [inline] _syscallexittousermodework kernel/entry/common.c:283 [inline] syscallexittousermode+0x15c/0x280 kernel/entry/common.c:296 dosyscall64+0x50/0x110 arch/x86/entry/common.c:88 entrySYSCALL64after_hwframe+0x63/0x6b
The root cause is: - dosysopen - f2fslookup - _f2fsfindentry - f2fsidepthwrite - f2fsmarkinodedirtysync - f2fsdirtyinode - setinodeflag(inode, FIDIRTY_INODE)
When we try to repair icurrentdepth in readonly filesystem, let's skip dirty inode to avoid panic in later f2fsevictinode().