[X86] Improve the AVX512 bailout in combineTruncateWithSat to allow pack instructions...
authorCraig Topper <craig.topper@intel.com>
Fri, 11 Oct 2019 00:38:51 +0000 (00:38 +0000)
committerCraig Topper <craig.topper@intel.com>
Fri, 11 Oct 2019 00:38:51 +0000 (00:38 +0000)
commitb560fd6c52eb5ce52315fff97f65952b53298728
tree93fa1896440b32a496676a35624c16eacdf689cb
parent4dc27c69b6d13d497b2bd9f430d91ce3dab8847b
[X86] Improve the AVX512 bailout in combineTruncateWithSat to allow pack instructions in more situations.

If we don't have VLX we won't end up selecting a saturating
truncate for 256-bit or smaller vectors so we should just use
the pack lowering.

llvm-svn: 374487
llvm/lib/Target/X86/X86ISelLowering.cpp
llvm/test/CodeGen/X86/masked_store_trunc_ssat.ll
llvm/test/CodeGen/X86/pmaddubsw.ll
llvm/test/CodeGen/X86/vector-trunc-packus.ll
llvm/test/CodeGen/X86/vector-trunc-ssat.ll