[X86] Fix the cost of v16i1->v16i16 sext/zext on avx targets.
authorCraig Topper <craig.topper@gmail.com>
Sun, 26 Apr 2020 06:00:57 +0000 (23:00 -0700)
committerCraig Topper <craig.topper@gmail.com>
Sun, 26 Apr 2020 06:16:20 +0000 (23:16 -0700)
commitb9de62c2b69e3b6342045ae82e1532d4f70c6d34
tree92811c6813667511b78ffefb8d931feeb6482b72
parent19cb26f51760ec17be5f7aefbf4160127f575da1
[X86] Fix the cost of v16i1->v16i16 sext/zext on avx targets.

Previously we were hitting the scalarization case in the default
implementation.
llvm/lib/Target/X86/X86TargetTransformInfo.cpp
llvm/test/Analysis/CostModel/X86/extend.ll
llvm/test/Analysis/CostModel/X86/min-legal-vector-width.ll