[Bug target/119629] mismatch between [power9-64] builtins and their instructions

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119629 --- Comment #10 from Alexandre Oliva --- >> - instructions and expanders for these builtins don't have their conditions >> tested, so they must necessarily follow from the builtin conditions, and >> this case clearly isn't > "They don't have th

[Bug target/119629] mismatch between [power9-64] builtins and their instructions

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119629 --- Comment #11 from Alexandre Oliva --- >> On a powerpc-elf standard build, TARGET_POWERPC64 is enabled, but >> TARGET_64BIT isn't, and so gcc.target/powerpc/byte-in-set-2.c fails >> to compile with an ICE (instruction not recognized) instead o

[Bug target/119629] mismatch between [power9-64] builtins and their instructions

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119629 --- Comment #12 from Alexandre Oliva --- Peter, Segher, thanks for the patches and the feedback. I'll give them a try and report back.

[Bug target/118182] RISC-V: Miscompile for 410.bwaves, 416.gamess and 465.tonto from spec2006

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118182 Alexandre Oliva changed: What|Removed |Added CC||aoliva at gcc dot gnu.org

[Bug target/118182] RISC-V: Miscompile for 410.bwaves, 416.gamess and 465.tonto from spec2006

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118182 --- Comment #7 from Alexandre Oliva --- bisection with this PR's patch led me to the patch that added the late-combine pass as the one that enables the intended result. That's all I know so far.

[Bug target/118182] RISC-V: Miscompile for 410.bwaves, 416.gamess and 465.tonto from spec2006

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118182 --- Comment #8 from Alexandre Oliva --- The problem is that the @pred_broadcast pattern expands to _zvfh insns even when _zero or _imm would do. The scalar constant gets allocated to a register, and vec_duplicated in the pred_broadcast insn, on

[Bug target/118182] RISC-V: Miscompile for 410.bwaves, 416.gamess and 465.tonto from spec2006

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118182 --- Comment #9 from Alexandre Oliva --- https://gcc.gnu.org/pipermail/gcc-patches/2025-April/680824.html

[Bug tree-optimization/118805] [15 Regression] wrong code at -O1 with "-fno-tree-ccp -fno-tree-copy-prop -fno-tree-forwprop -fno-tree-fre" on x86_64-linux-gnu since r15-6173

|RESOLVED Assignee|unassigned at gcc dot gnu.org |aoliva at gcc dot gnu.org --- Comment #7 from Alexandre Oliva --- Fixed

[Bug target/119629] New: mismatch between [power9-64] builtins and their instructions

Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: aoliva at gcc dot gnu.org Target Milestone: --- cmpeqb's conditions are TARGET_P9_MISC && TARGET_64BIT, but the conditions that enable the expansion of __builtin_scalar_by

[Bug target/119609] New: [powerpc-elf] load_toc_v4_pic_si may clobber r12 and crt

Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: aoliva at gcc dot gnu.org Target Milestone: --- A buggy linker script placed .got too far from .text and triggered a latent bug. A small leaf function that needed the got register to

<    13   14   15   16   17   18