diff options
| author | Sanjay Patel <spatel@rotateright.com> | 2018-10-09 14:04:14 +0000 |
|---|---|---|
| committer | Sanjay Patel <spatel@rotateright.com> | 2018-10-09 14:04:14 +0000 |
| commit | f5fac1826a867d5e65310456670cc1d5a495ed02 (patch) | |
| tree | 8b91cb196a284c34941c1b8346d3b386adeecf10 /clang/lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp | |
| parent | 0b24a86f6322d1fa3e458b8114fa9a88f39ea360 (diff) | |
| download | bcm5719-llvm-f5fac1826a867d5e65310456670cc1d5a495ed02.tar.gz bcm5719-llvm-f5fac1826a867d5e65310456670cc1d5a495ed02.zip | |
[x86] use demanded bits to simplify masked store codegen
As noted in D52747, if we prefer IR to use trunc for bool vectors rather
than and+icmp, we can expose codegen shortcomings as seen here with masked store.
Replace a hard-coded PCMPGT simplification with the more general demanded bits call
to improve things.
Differential Revision: https://reviews.llvm.org/D52964
llvm-svn: 344048
Diffstat (limited to 'clang/lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp')
0 files changed, 0 insertions, 0 deletions
