In the Linux kernel, the following vulnerability has been resolved:
erofs: fix pcluster use-after-free on UP platforms
During stress testing with CONFIG_SMP disabled, KASAN reports as below:
================================================================== BUG: KASAN: use-after-free in _mutexlock+0xe5/0xc30 Read of size 8 at addr ffff8881094223f8 by task stress/7789
CPU: 0 PID: 7789 Comm: stress Not tainted 6.0.0-rc1-00002-g0d53d2e882f9 #3 Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011 Call Trace: <TASK> .. _mutexlock+0xe5/0xc30 .. zerofsdoreadpage+0x8ce/0x1560 .. zerofsreadahead+0x31c/0x580 .. Freed by task 7787 kasansavestack+0x1e/0x40 kasansettrack+0x20/0x30 kasansetfreeinfo+0x20/0x40 _kasanslabfree+0x10c/0x190 kmemcachefree+0xed/0x380 rcucore+0x3d5/0xc90 _do_softirq+0x12d/0x389
Last potentially related work creation: kasansavestack+0x1e/0x40 _kasanrecordauxstack+0x97/0xb0 callrcu+0x3d/0x3f0 erofsshrinkworkstation+0x11f/0x210 erofsshrinkscan+0xdc/0x170 shrinkslab.constprop.0+0x296/0x530 dropslab+0x1c/0x70 dropcachessysctlhandler+0x70/0x80 procsyscallhandler+0x20a/0x2f0 vfswrite+0x555/0x6c0 ksyswrite+0xbe/0x160 dosyscall_64+0x3b/0x90
The root cause is that erofsworkgroupunfreeze() doesn't reset to orig_val thus it causes a race that the pcluster reuses unexpectedly before freeing.
Since UP platforms are quite rare now, such path becomes unnecessary. Let's drop such specific-designed path directly instead.