On Tue, Dec 5, 2017 at 7:05 PM, Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> wrote: > Quoting Chris Wilson (2017-12-05 18:00:00) >> When we see an e1000e HW lockup in CI, it is typically fatal with the >> hang repeating until the host is forcibly rebooted. Speed up that >> process by tainting the kernel, which CI can trivially detect (and is >> being used to detect similarly fatal CI conditions) and reboot soon >> after. >> >> Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> >> Cc: Daniel Vetter <daniel.vetter@xxxxxxxx> >> Cc: Tomi Sarvela <tomi.p.sarvela@xxxxxxxxx> > > I'm not concerned on selling this to e1000e, but if it helps improving > CI robustness, then topic/core-for-CI. Or maybe we should create a new > topic, Daniel? topic/taints-for-CI? Sounds like a usable idea for CI. Would be especially interesting because despite applying the suggested w/a, we still hit lockups. Before we do that though I think we should get an ack from the e1000e team. Jani S. maybe something you can driver? Adding more folks to cc. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
