So Firefox 106 triggered a bug on 1st gen Ryzen processors. We found this by noticing a bunch of nonsensical crashes in the painting code that only happened on these processors.
Conversation
Replying to
How did we fix it? By rebuilding the exact same code as 106. The timing differences caused by using a new profile when doing PGO were enough to cause the generated code to be different enough to avoid triggering the processor bug.
5
13
116
Sigh.
2
69
Replying to
Ryzen 1 is/was a mess! I guess it’s what you expect from a big architectural overhaul but “RDRAND always generates 1s” and “synchronized reads can get stale data” are pretty big misses. support.amd.com/TechDocs/55449
1
18
In their defense the stale data stuff is very timing-dependent and the logic doing the store-to-load forwarding is stupidly complex. Given it was their first iteration it wasn't so bad, all CPUs have erratas.
1
18
Show replies
Show additional replies, including those that may contain offensive content
Show