diff options
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | 2016-07-15 09:49:12 +0000 |
|---|---|---|
| committer | Simon Pilgrim <llvm-dev@redking.me.uk> | 2016-07-15 09:49:12 +0000 |
| commit | 2683ad54ad0cd5b12db99395a72498f8af92fd45 (patch) | |
| tree | 3e6d77818564be3f4781e15c8ade92c2d2415e30 /lldb/source/Plugins/Process/Linux/NativeProcessLinux.cpp | |
| parent | 39a718c48bed8b45caac909400fc9ea52f564bc4 (diff) | |
| download | llvm-2683ad54ad0cd5b12db99395a72498f8af92fd45.zip llvm-2683ad54ad0cd5b12db99395a72498f8af92fd45.tar.gz llvm-2683ad54ad0cd5b12db99395a72498f8af92fd45.tar.bz2 | |
[X86][AVX2] Improve lowerShuffleAsRepeatedMaskAndLanePermute permutation of 64-bit sub-lanes
As discussed on PR28136, lowerShuffleAsRepeatedMaskAndLanePermute was attempting to match repeated masks at the 128-bit level and then permute the resultant lanes at the 128-bit (AVX1) or 64-bit (AVX2) sub-lane level.
This change allows us to create the repeated masks at the sub-lane level (and then concat them together to create a 128-bit repeated mask) and then select which sub-lane to permute. This has no effect on the AVX1 codegen.
Fixes PR28136.
llvm-svn: 275543
Diffstat (limited to 'lldb/source/Plugins/Process/Linux/NativeProcessLinux.cpp')
0 files changed, 0 insertions, 0 deletions