subject:"\[Bug target\/106910\] roundss not vectorized"

[Bug target/106910] roundss not vectorized

2022-09-22 Thread crazylht at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #13 from Hongtao.liu --- (In reply to Pilar Latiesa from comment #12) > Wouldn't it make sense for scalar and vector versions to be affected by the > same options? I guess the reason is we don't have __builtin_floorf for vector vers

[Bug target/106910] roundss not vectorized

2022-09-22 Thread pilarlatiesa at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #12 from Pilar Latiesa --- Wouldn't it make sense for scalar and vector versions to be affected by the same options?

[Bug target/106910] roundss not vectorized

2022-09-21 Thread crazylht at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #11 from Hongtao.liu --- (In reply to Hongtao.liu from comment #10) > Because the round insn does not trap on denormals. Note for scalar version, gcc generates roundss $9, %xmm0, %xmm0 which doesn't raise exception because it's allo

[Bug target/106910] roundss not vectorized

2022-09-21 Thread crazylht at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #10 from Hongtao.liu --- Because the round insn does not trap on denormals.

[Bug target/106910] roundss not vectorized

2022-09-21 Thread pilarlatiesa at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #9 from Pilar Latiesa --- Thans so much for the fix. I have tested last night (20220920) snapshot and I can confirm the testcase is vectorized. However, for the vectorization to kick in I had to use -fno-trapping-math. I assume tha

[Bug target/106910] roundss not vectorized

2022-09-20 Thread crazylht at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #8 from Hongtao.liu --- Fixed in GCC13.

[Bug target/106910] roundss not vectorized

2022-09-19 Thread cvs-commit at gcc dot gnu.org via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #7 from CVS Commits --- The master branch has been updated by hongtao Liu : https://gcc.gnu.org/g:d0c73b6c85677e6755b60fa02d79a5c5e1a8eacd commit r13-2730-gd0c73b6c85677e6755b60fa02d79a5c5e1a8eacd Author: liuhongt Date: Fri Sep

[Bug target/106910] roundss not vectorized

2022-09-16 Thread cvs-commit at gcc dot gnu.org via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #6 from CVS Commits --- The master branch has been updated by hongtao Liu : https://gcc.gnu.org/g:3e8c4b925a9825fdb8c81f47b621f63108894362 commit r13-2694-g3e8c4b925a9825fdb8c81f47b621f63108894362 Author: liuhongt Date: Thu Sep

[Bug target/106910] roundss not vectorized

2022-09-15 Thread rguenther at suse dot de via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #5 from rguenther at suse dot de --- On Thu, 15 Sep 2022, crazylht at gmail dot com wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 > > --- Comment #4 from Hongtao.liu --- > > > The vectorizer would go the direct int

[Bug target/106910] roundss not vectorized

2022-09-15 Thread crazylht at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #4 from Hongtao.liu --- > The vectorizer would go the direct internal fn way for those, querying the > floor optab but the x86 backend only has scalar modes supported for the > rounding optabs. For CFN_BUILT_IN_ICEIL, the modifier i

[Bug target/106910] roundss not vectorized

2022-09-13 Thread crazylht at gmail dot com via Gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106910 --- Comment #3 from Hongtao.liu --- > The backend should modernize itself, get rid of the > ix86_builtin_vectorized_function parts for those functions and instead rely > on define_expands with vector modes. Indeed, let me do it.