| commit | 6456d2f1ca95fbcd09e441c317dce15ead0d10a2 | |
| author | Erik Lindahl <erik@kth.se> | |
| Wed, 29 Nov 2017 21:58:38 +0000 (29 22:58 +0100) | ||
| committer | Mark Abraham <mark.j.abraham@gmail.com> | |
| Mon, 4 Dec 2017 07:21:27 +0000 (4 08:21 +0100) | ||
| tree | 50c0ce166a66da80e648d406cc3b799141b6a302 | treesnapshot (tar.gz zip) |
| parent | cf5e082b9791081e31b750f0635bcddf12f7e35e | commitdiff |
Fix compilation issues for AVX-512
- gcc-5.4.0 incorrectly requires the second argument of
_mm512_i32gather_pd() to be a double pointer instead
of void, but this should fix compilation for both
cases.
- Work around double precision permute instruction
only available with AVX512VL instructions.
Fixes #2312.
Change-Id: I31420e71064b1c5c25c8af29a1d41c7f372375c1
- gcc-5.4.0 incorrectly requires the second argument of
_mm512_i32gather_pd() to be a double pointer instead
of void, but this should fix compilation for both
cases.
- Work around double precision permute instruction
only available with AVX512VL instructions.
Fixes #2312.
Change-Id: I31420e71064b1c5c25c8af29a1d41c7f372375c1
| src/gromacs/simd/impl_x86_avx_512/impl_x86_avx_512_util_double.h | diffblobblamehistory | |
| src/gromacs/simd/impl_x86_avx_512/impl_x86_avx_512_util_float.h | diffblobblamehistory |