[x86] Delete still more piles of complex code now that we have a good
authorChandler Carruth <chandlerc@gmail.com>
Thu, 19 Feb 2015 15:21:57 +0000 (15:21 +0000)
committerChandler Carruth <chandlerc@gmail.com>
Thu, 19 Feb 2015 15:21:57 +0000 (15:21 +0000)
commitb7012af85fc7bd67ad9035f650ac1d802dfd0ec4
tree8e85145b5d4f198e409690076ffd59022084e172
parentc57e90422fe64d89c527a8a7886e0f7d875fef3f
[x86] Delete still more piles of complex code now that we have a good
systematic lowering of v8i16.

This required a slight strategy shift to prefer unpack lowerings in more
places. While this isn't a cut-and-dry win in every case, it is in the
overwhelming majority. There are only a few places where the old
lowering would probably be a touch faster, and then only by a small
margin.

In some cases, this is yet another significant improvement.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229859 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86ISelLowering.cpp
test/CodeGen/X86/vector-shuffle-128-v16.ll
test/CodeGen/X86/vector-shuffle-128-v8.ll
test/CodeGen/X86/vector-shuffle-256-v16.ll
test/CodeGen/X86/vector-trunc.ll
test/CodeGen/X86/widen_shuffle-1.ll