https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128058
Bug ID: 128058
Summary: [17 Regression] AOP changes cause minor AArch64
testsuite regressions
Product: gcc
Version: 17.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: law at gcc dot gnu.org
Target Milestone: ---
commit 53c3316a8c79bd873a884bf89c11144d42af7f7c
Author: Roger Sayle <[email protected]>
Date: Wed Oct 7 09:50:53 2026 +0100
PR tree-optimization/108477: Use ranger to select PLUS vs IOR in expand.
[ ... ]
Is causing these regressions in the aarch64 backend:
gcc: gcc.target/aarch64/combine_bfi_1.c scan-assembler-times \\tbfi\\t 5
gcc: gcc.target/aarch64/combine_bfi_2.c scan-assembler-times \\tbfi\\t 2
gcc: gcc.target/aarch64/combine_bfxil.c scan-assembler-times bfi\\t 15
gcc: gcc.target/aarch64/combine_bfxil.c scan-assembler-times bfxil\\t 3
Well, I really just biseced the combine_bfi_1 test, but I strongly suspec the
other three would bisect to the same change. Given what I've seen from AOP, I
suspect this is either a target or RTL optimization issue.
I haven't debugged this beyond bisection.
