Sanjay Patel 11d77223a5 [X86, AVX] use blends instead of insert128 with index 0
Another case of x86-specific shuffle strength reduction:
avoid generating insert*128 instructions with index 0 because
they are slower than their non-lane-changing blend equivalents.

Shuffle lowering already catches most of these cases, but
the zero vector case and some other paths such as in the
modified test in vector-shuffle-256-v32.ll were getting
through.

Differential Revision: http://reviews.llvm.org/D8366


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@232773 91177308-0d34-0410-b5e6-96231b3b80d8
2015-03-19 22:29:40 +00:00
..
2015-03-18 01:34:06 +00:00
2015-03-06 20:51:25 +00:00
2015-03-06 21:50:27 +00:00
2015-03-06 21:30:18 +00:00
2015-03-06 20:38:55 +00:00
2015-03-06 20:46:16 +00:00