On Tue, Nov 21, 2017 at 3:52 AM, Fengguang Wu <fengguang.wu@intel.com> wrote:
> Here is a different call trace related to rhashtable.
>
> It occurs in 17 out of 90 boots.
>
> [  342.666234] Writes:  Total: 4  Max/Min: 0/0   Fail: 0
> [  381.918047] INFO: task swapper/0:1 blocked for more than 120 seconds.

This looks very similar to the other lockup case: lock torture testing
and rhashtable testing at the same time, causing long long delays..

          Linus