Bah, I just noticed that the recent revert should have been marked for stable for 6.5, but I pushed it out and it's too late now. It may be that you end up auto-flagging reverts the same way you do for "Fixes:" tags, so maybe it would end up on your radar, but just to make sure, I thought I'd mention the commit here explicitly: fbe1bf1e5ff1 Revert "x86/smp: Put CPUs into INIT on shutdown if possible" just so that it doesn't get missed. Linus