diff options
| author | Sjoerd Meijer <sjoerd.meijer@arm.com> | 2021-06-24 19:29:13 +0100 |
|---|---|---|
| committer | Sjoerd Meijer <sjoerd.meijer@arm.com> | 2021-06-25 09:53:51 +0100 |
| commit | 51e434fc2590d1d3ffa6545cd07290a238db2b88 (patch) | |
| tree | b59314389377f4db12f56603922991d73b59aeea /llvm/lib/ProfileData/Coverage/CoverageMappingReader.cpp | |
| parent | 18d7e822ab22e6747f01b0409ace5044733be162 (diff) | |
| download | llvm-51e434fc2590d1d3ffa6545cd07290a238db2b88.zip llvm-51e434fc2590d1d3ffa6545cd07290a238db2b88.tar.gz llvm-51e434fc2590d1d3ffa6545cd07290a238db2b88.tar.bz2 | |
[AArch64] Custom lower <4 x i8> loads
This custom lowers <4 x i8> vector loads using a 32-bit load, followed by 2
SSHLL instructions to extend it to e.g. a <4 x i32> vector. Before, it was
really inefficient and expensive to construct a <4 x i32> for this as 4 byte
loads and 4 moves were used. With this improvement SLP vectorisation might for
example become profitable, see D103629.
Differential Revision: https://reviews.llvm.org/D104782
Diffstat (limited to 'llvm/lib/ProfileData/Coverage/CoverageMappingReader.cpp')
0 files changed, 0 insertions, 0 deletions