Looks like someone else noticed the same bug

Posted Jul 14, 2025 5:08 UTC (Mon) by neggles (subscriber, #153254)
Parent article: Kernel prepatch 6.16-rc6

if anyone else is curious about what the relevant issue is/was, here's the message where Linus found the culprit, an attempted epoll priority inversion fix. I also found a patch from a few days ago that looks like a fix for the regression and the issue that it was originally trying to fix?