mirror_zfs/module/os/linux
Chunwei Chen c79d5e4f33 Define sops->free_inode() to prevent use-after-free during lookup
On Linux, when doing path lookup with LOOKUP_RCU, dentry and inode can
be dereferenced without refcounts and locks. For this reason, dentry and
inode must only be freed after RCU grace period.

However, zfs currently frees inode in zfs_inode_destroy synchronously
and we can't use GPL-only call_rcu() in zfs directly. Fortunately, on
Linux 5.2 and after, if we define sops->free_inode(), the kernel will do
call_rcu() for us.

This issue may be triggered more easily with init_on_free=1 boot
parameter:

BUG: kernel NULL pointer dereference, address: 0000000000000020
RIP: 0010:selinux_inode_permission+0x10e/0x1c0
Call Trace:
 ? show_trace_log_lvl+0x1be/0x2d9
 ? show_trace_log_lvl+0x1be/0x2d9
 ? show_trace_log_lvl+0x1be/0x2d9
 ? security_inode_permission+0x37/0x60
 ? __die_body.cold+0x8/0xd
 ? no_context+0x113/0x220
 ? exc_page_fault+0x6d/0x130
 ? asm_exc_page_fault+0x1e/0x30
 ? selinux_inode_permission+0x10e/0x1c0
 security_inode_permission+0x37/0x60
 link_path_walk.part.0.constprop.0+0xb5/0x360
 ? path_init+0x27d/0x3c0
 path_lookupat+0x3e/0x1a0
 filename_lookup+0xc0/0x1d0
 ? __check_object_size.part.0+0x123/0x150
 ? strncpy_from_user+0x4e/0x130
 ? getname_flags.part.0+0x4b/0x1c0
 vfs_statx+0x72/0x120
 ? ioctl_has_perm.constprop.0.isra.0+0xbd/0x120
 __do_sys_newlstat+0x39/0x70
 ? __x64_sys_ioctl+0x8d/0xd0
 do_syscall_64+0x30/0x40
 entry_SYSCALL_64_after_hwframe+0x62/0xc7

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: Rob Norris <rob.norris@klarasystems.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Co-authored-by: Chunwei Chen <david.chen@nutanix.com>
Closes #17546
2025-08-05 12:30:23 -04:00
..
spl tunables: use Linux ullong param ops for u64 2025-06-17 10:50:26 -07:00
zfs Define sops->free_inode() to prevent use-after-free during lookup 2025-08-05 12:30:23 -04:00