Don't use INSERT_SUBREG to model anyext operations on x86-64, as it
authorDan Gohman <gohman@apple.com>
Wed, 26 Aug 2009 14:59:13 +0000 (14:59 +0000)
committerDan Gohman <gohman@apple.com>
Wed, 26 Aug 2009 14:59:13 +0000 (14:59 +0000)
commitaf70e5c676140a55ec619f552c8fd76b7edf3c6d
treedaaa20890ab20b911913f65f5b0f98e309c53883
parent2385e0e22ca482f2896dfa975a08db2f54926c09
Don't use INSERT_SUBREG to model anyext operations on x86-64, as it
leads to partial-register definitions. To help avoid redundant
zero-extensions, also teach the h-register matching patterns that
use movzbl to match anyext as well as zext.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80099 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86Instr64bit.td
lib/Target/X86/X86InstrInfo.td
test/CodeGen/X86/anyext.ll [new file with mode: 0644]
test/CodeGen/X86/ins_subreg_coalesce-1.ll