On Sun, Apr 02, 2023 at 06:19:20AM +0800, Rongwei Wang wrote:
Without this modification, a core will wait (mostly)"Currently, a core will ..."
But, a worse consequence, panic also can be caused by"However, in the worst case, ..."
In this patch, we lock p->lock before calling"Lock p->lock before calling ..."
We also find this problem exists in stable 5.10.So, you claim that 5.15.y and 6.1.y aren't affected, right?
Thanks, next version will to do.
Also, Cc: stable@xxxxxxxxxxxxxxx on the SoB area (as pointed by kernel
test robot [1].
Thanks.
[1]: https://lore.kernel.org/stable/ZCiuGEkyk%2F1Afisk@ec83ac1404bb/