Quoting Mika Kuoppala (2018-07-11 12:27:49) > Chris Wilson <[email protected]> writes: > > > Since live_workarounds poke around the w/a registers and checks to see > > if they survive across a reset, we are prone to fouling the machine and > > leaving it in a non-recoverable state. Wrap the probe inside a timeout > > to abort the test if the reset fails. > > > > Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107188 > > How can we tell it is not about just reset flakyness but > associated whitelist poking?
By reading the dmesg. All we are doing here is breaking the indefinite wait (as we've disabled hangcheck so as to not interfere with out own resets) before owatch declares INCOMPLETE. -Chris _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
