[clang] [llvm] [X86][AMX] Support AMX-TRANSPOSE (PR #113532)

[Nikita Popov via llvm-commits]

nikic wrote:

Looks like this causes a significant compile-time regression, but only for ReleaseLTO-g: https://llvm-compile-time-tracker.com/compare.php?from=1e19f0f9d92b5e9c43d53893e387341835d3d96b&to=c72a751dabff4260dcc309e48008941d51b31d21&stat=instructions:u

I wouldn't really expect this change to only affect that configuration -- I guess that means that optimized debuginfo processing has become a lot slower? Any idea what is going on here?

https://github.com/llvm/llvm-project/pull/113532