On Sun, Apr 9, 2023 at 9:52 AM Genes Lists <lists@xxxxxxxxxxxx> wrote: > > On 4/9/23 08:06, Jonas Malaco wrote: > > As you say, your case is definitely a hard crash. > > - Do you have any modules that taint the kernel? Hi, On that particular system, no. > - the s2idle was only AMD not intel as far as I know. But good to > prevent any kind of sleep or hibernation just to keep that away. I'm curious to see if disabling s2idle will have any effect on the less common crashes that happen when just the monitor turns off, without suspending to RAM. > - your occasional hangups might also suggest hardware problems of > course, which can be tricky to diagnose - esp since happens in both > windows and linux. What leads you to bios bugs vs hardware? Assume your > filesys are all clean. Honestly, it's been a while since the last one, and I can't find my notes or old logs. IIRC out of nowhere a bunch of PCI AER errors would pop up (and Windows also showed similar warnings from time to time). Initially this happened a lot, but only sometimes a hangup would follow. At some point I found similar reports, one or more of them which pointed to a possible BIOS/firmware issue. As time went on, and new BIOS versions were available, the problem got progressively less common, despite virtually no hardware changes (there was one, but there was no observable decrease or increase in AER errors or hangups). And now I don't have a single AER error logged for the past 2 months (which is as far as the logs on that machine go). But I went too far before when I said that it _was_ a firmware issue. It would be more reasonable to say that those _might_ have been firmware issues. Thanks, Jonas
