[Bug target/64669] [5 Regression] aarch64-linux profiledbootstrap failure

["jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64669

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jiwang at gcc dot gnu.org,
                   |                            |zhenqiang.chen at arm dot com

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Seems this is because of the gen_ccmp_{first,next} stuff that is being
constantly added and reverted again.
How can it work reliably?  I mean, if expansion of something returns you
something in completely different mode, how do you ensure that the side that
consumes it will know it has to handle it specially?  Some new RTX code
signalizing that would at least make it more clear, but you'd need to ensure
that if you don't handle it specially, that it will be automatically turned
back to the expansion that didn't do anything special.