[Bug target/126454] [16/17 Regression] RISCV: cm.popretz pass deletes "li a0,0" and gives wrong return value

"law at gcc dot gnu.org via Gcc-bugs" <[email protected]> Tue, 04 Aug 2026 01:04:12 +0000

Newsgroups	gmane.comp.gcc.bugs
Message-ID	<[email protected]/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D126454

Jeffrey A. Law <law at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kito at gcc dot gnu.org
             Blocks|120763                      |

--- Comment #8 from Jeffrey A. Law <law at gcc dot gnu.org> ---
What's weird is I know we fixed at least one bug in this space not terribly
long ago.   There was a pretty significant revamp of that code and clearly =
it's
still not right.  I don't know if the patch is correct as I don't really de=
al
with this kind of embedded stuff for RISC-V.  Kito knows much more about th=
is
code than I do, I've added him on CC.


Referenced Bugs:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D120763
[Bug 120763] [meta-bug] Tracker for bugs to visit during weekly RISC-V meet=
ing=