In the Linux kernel, the following vulnerability has been resolved:
ptp: remove ptp->nvclocks check logic in ptpvclockinuse()
There is no disagreement that we should check both ptp->isvirtualclock and ptp->n_vclocks to check if the ptp virtual clock is in use.
However, when we acquire ptp->nvclocksmux to read ptp->nvclocks in ptpvclockinuse(), we observe a recursive lock in the call trace starting from nvclocksstore().
============================================ WARNING: possible recursive locking detected
syz.0.1540/13807 is trying to acquire lock: ffff888035a24868 (&ptp->nvclocksmux){+.+.}-{4:4}, at: ptpvclockinuse drivers/ptp/ptpprivate.h:103 [inline] ffff888035a24868 (&ptp->nvclocksmux){+.+.}-{4:4}, at: ptpclockunregister+0x21/0x250 drivers/ptp/ptp_clock.c:415
but task is already holding lock: ffff888030704868 (&ptp->nvclocksmux){+.+.}-{4:4}, at: nvclocksstore+0xf1/0x6d0 drivers/ptp/ptp_sysfs.c:215
other info that might help us debug this: Possible unsafe locking scenario:
CPU0
----
lock(&ptp->nvclocksmux); lock(&ptp->nvclocksmux);
* DEADLOCK *
The best way to solve this is to remove the logic that checks ptp->nvclocks in ptpvclockinuse().
The reason why this is appropriate is that any path that uses ptp->nvclocks must unconditionally check if ptp->nvclocks is greater than 0 before unregistering vclocks, and all functions are already written this way. And in the function that uses ptp->nvclocks, we already get ptp->nvclocks_mux before unregistering vclocks.
Therefore, we need to remove the redundant check for ptp->nvclocks in ptpvclockinuse() to prevent recursive locking.