https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111071
Andrew Pinski <pinskia at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Keywords| |testsuite-fail Version|unknown |14.0 Ever confirmed|0 |1 Target Milestone|--- |14.0 Status|UNCONFIRMED |NEW Summary|gcc.target/aarch64/sve/acle |[14 Regression] |/aarch64-sve-acle-asm.exp |gcc.target/aarch64/sve/acle |has new failures since |/aarch64-sve-acle-asm.exp |commit e7a36e4715c7 |has new failures since | |commit e7a36e4715c7 Component|target |testsuite Last reconfirmed| |2023-08-19 --- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> --- Confirmed. The code is better: For subr_m1_s8_m, GCC now produces: not z0.b, p0/m, z0.b ret Rather than what was originally expected: /* ** subr_m1_s8_m: ** mov (z[0-9]+\.b), #-1 ** subr z0\.b, p0/m, z0\.b, \1 ** ret */ So just a testsuite update is needed here.