Tinygrad AMD Linux Driver Crash Hang Tracker and Updates

I totally understand why George is upset. Troubleshooting crashes within a vendor&#x27;s closed source product is one of the most painful things I&#x27;ve had to do professionally. In my case, it took months of patience and collecting pcaps to finally get it resolved. George doesn&#x27;t have months.<p>I&#x27;ve been fascinated by his live streams where he has been digging into the bug himself: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;@geohotarchive&#x2F;videos" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;@geohotarchive&#x2F;videos</a><p>As crappy as the situation is, I must say it&#x27;s been great to see George&#x27;s passion to solve these types of technical puzzles reignited! It seems like he&#x27;s tried his hardest to work with AMD in an nonabrasive manor but he has become very frustrated by the lack of progress.<p>AMD would be wise to just pair one of their best, whom has access to the source, with George for just a couple days.

Relevant Tweets by George Hotz,<p>[1] Spoke with @LisaSu. She was very polite, but said nothing of substance and indirectly rejected my offer.<p>Here’s a firmware crash, run .&#x2F;loop.sh to trigger. Tested on tinybox and 1x7900XTX machine, ROCm 6.0.2 and 6.0.3 preview. Must reboot to bring back.<p>[2] No need to hire me, just open source the 7900XTX firmware+docs and remove the signature check. We&#x27;d treat it like bring up for our own chip, get builds in CI, HITL testing, fuzzing, etc...<p>Deadline, end of the week? Otherwise I&#x27;m not spending more time thinking about this.<p>Source <a href="https:&#x2F;&#x2F;twitter.com&#x2F;realGeorgeHotz&#x2F;status&#x2F;1771666351653781862" rel="nofollow">https:&#x2F;&#x2F;twitter.com&#x2F;realGeorgeHotz&#x2F;status&#x2F;177166635165378186...</a>