https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84547
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Summary|Suboptimal code for masked |Suboptimal code for int128
|shifts (ARM64) |masked shifts (ARM64)
Last reconfirmed| |2021-04-17
Ever confirmed|0 |1
Status|UNCONFIRMED |NEW
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Yes int128 (or rather double wide register) shifts are not optimized that well.
They are not used by many people even. I wonder if there is a way to use
vector instructions to do them.