projects / gcc.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 85ec4fe) | patch
x86: don't use AVX512BW vmovdqu variants without -mavx512bw
authorJan Beulich <jbeulich@suse.com>
Wed, 3 Jan 2018 10:42:08 +0000 (10:42 +0000)
committerJan Beulich <jbeulich@gcc.gnu.org>
Wed, 3 Jan 2018 10:42:08 +0000 (10:42 +0000)
commitb853346e2c2ea98e35c6043712303a5fc8b87dd8
tree26c96624b0d0ddcb91294085689f0404c1ebfba1tree
parent85ec4feb11167c9e4489361bf2399a20afbe52c8commit | diff
x86: don't use AVX512BW vmovdqu variants without -mavx512bw

Simply mirror the MODE_XI logic of handling unaligned operands in
mov<mode>_internal into MODE_TI / MODE_OI handling.

gcc/
2018-01-03  Jan Beulich  <jbeulich@suse.com>

* sse.md (mov<mode>_internal): Tighten condition for when to use
vmovdqu<ssescalarsize> for TI and OI modes.

gcc/testsuite/
2018-01-03  Jan Beulich  <jbeulich@suse.com>

* gcc.target/i386/avx512vl-no-vmovdqu8.c,
gcc.target/i386/avx512vl-no-vmovdqu16.c: New.

From-SVN: r256170
gcc/ChangeLog diff | blob | history
gcc/config/i386/sse.md diff | blob | history
gcc/testsuite/ChangeLog diff | blob | history
gcc/testsuite/gcc.target/i386/avx512vl-no-vmovdqu16.c [new file with mode: 0644] blob
gcc/testsuite/gcc.target/i386/avx512vl-no-vmovdqu8.c [new file with mode: 0644] blob