https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94649
--- Comment #10 from andysem at mail dot ru --- (In reply to Florian Weimer from comment #9) > > Is this still an issue today? I thought libatomic would use CMPXCHG16B if > the current CPU supports it. I think the complaint was that the cmpxchg16b instruction is not generated directly and instead there is a library call. I don't think this has changed. We need an option "-mreally-do-use-cx16-pretty-please" that will force emitting cmpxchg16b despite any interop issues. This mode would be useful when one requires a CPU with cmpxchg16b and is willing to compile his entire project with this option. Which, realistically, should be everyone.
