| commit | 19792fb2704439766356e9d86d171ea8b7d815bd | [log] [tgz] |
|---|---|---|
| author | Sanjay Patel <spatel@rotateright.com> | Tue Mar 10 16:08:36 2015 +0000 |
| committer | Sanjay Patel <spatel@rotateright.com> | Tue Mar 10 16:08:36 2015 +0000 |
| tree | 8017f6aa7d50ef4c9e32349079478b3aa8fbf040 | |
| parent | c98950671d826aea2ad53a5951cc0061b3c761bc [diff] |
[X86, AVX] replace vinsertf128 intrinsics with generic shuffles We want to replace as much custom x86 shuffling via intrinsics as possible because pushing the code down the generic shuffle optimization path allows for better codegen and less complexity in LLVM. This is the sibling patch for the Clang half of this change: http://reviews.llvm.org/D8088 Differential Revision: http://reviews.llvm.org/D8086 llvm-svn: 231794