diff options
| author | Craig Topper <craig.topper@intel.com> | 2019-09-08 00:43:52 +0000 |
|---|---|---|
| committer | Craig Topper <craig.topper@intel.com> | 2019-09-08 00:43:52 +0000 |
| commit | 37dd59298fd46e28ae2b2569465c5195d5708a0a (patch) | |
| tree | bbcaaaeff89f42ad2c185432251944116dc9e576 /clang/lib/CodeGen/BackendUtil.cpp | |
| parent | 5bd4a4806aafc31c67a55c7d8b2993c879dc5bc2 (diff) | |
| download | bcm5719-llvm-37dd59298fd46e28ae2b2569465c5195d5708a0a.tar.gz bcm5719-llvm-37dd59298fd46e28ae2b2569465c5195d5708a0a.zip | |
[X86] Make getZeroVector return floating point vectors in their native type on SSE2 and later.
isel used to require zero vectors to be canonicalized to a single
type to minimize the number of patterns needed to match. This is
no longer required.
I plan to do this to integers too, but floating point was simpler
to start with. Integer has a complication where v32i16/v64i8 aren't
legal when the other 512-bit integer types are.
llvm-svn: 371325
Diffstat (limited to 'clang/lib/CodeGen/BackendUtil.cpp')
0 files changed, 0 insertions, 0 deletions
