diff options
| author | Cameron Zwarich <zwarich@apple.com> | 2011-06-18 06:17:51 +0000 |
|---|---|---|
| committer | Cameron Zwarich <zwarich@apple.com> | 2011-06-18 06:17:51 +0000 |
| commit | 9601ddb2f306b6fad7aa1e34ccd3973fe7145275 (patch) | |
| tree | d5049ea6eb0d258744dd55ff18a74dde99c8fea6 /clang/lib/CodeGen/CodeGenModule.cpp | |
| parent | 80ed9dc9e58f356139785cfb6a9f24cb64b76986 (diff) | |
| download | llvm-9601ddb2f306b6fad7aa1e34ccd3973fe7145275.zip llvm-9601ddb2f306b6fad7aa1e34ccd3973fe7145275.tar.gz llvm-9601ddb2f306b6fad7aa1e34ccd3973fe7145275.tar.bz2 | |
When scalar replacement returns a vector type, only accept it if the vector
type's bitwidth matches the (allocated) size of the alloca. This severely
pessimizes vector scalar replacement when the only vector type being used is
something like <3 x float> on x86 or ARM whose allocated size matches a
<4 x float>.
I hope to fix some of the flawed assumptions about allocated size throughout
scalar replacement and reenable this in most cases.
llvm-svn: 133338
Diffstat (limited to 'clang/lib/CodeGen/CodeGenModule.cpp')
0 files changed, 0 insertions, 0 deletions