Quoting Michal Wajdeczko (2017-10-20 13:50:23) > On Fri, 20 Oct 2017 14:43:46 +0200, Chris Wilson > <chris@xxxxxxxxxxxxxxxxxx> wrote: > > Heh, I would just go back and make sure we had a clear well of deducing > > when fw was being utilised. Maybe > > > > if (uc_fw.load_status != INTEL_UC_FIRMWARE_NONE) > > > > If we don't care about earlier fw fetch attempts/errors, then ok. Just since the last reset is about all we aim for. It's a game of trying to work out what information will be required to debug future hangs. Too much and it's a needle-in-a-haystack-problem, too little and we have the round-trip latency in getting a new kernel to where the problem is seen. And then for the next few years being frustrated at seeing hangs with the old error state. I defer what information is required for debugging in the future to those already attempting it; but I do know that recording what firmware and what status we left the uc in will always be useful. :) -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
