R600/SI: Allow f64 inline immediates in i64 operands
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Fri, 13 Feb 2015 19:05:03 +0000 (19:05 +0000)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Fri, 13 Feb 2015 19:05:03 +0000 (19:05 +0000)
commit1751616522bdb9911f9851fec576514b02360a85
treec935647650d30964c54ce28a10efba17fb7efba1
parentd14e5ec25d9eb2de80cf95ef02efb9c36ba3cfdd
R600/SI: Allow f64 inline immediates in i64 operands

This requires considering the size of the operand when
checking immediate legality.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@229135 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/R600/MCTargetDesc/SIMCCodeEmitter.cpp
lib/Target/R600/SIFoldOperands.cpp
lib/Target/R600/SIISelLowering.cpp
lib/Target/R600/SIInstrInfo.cpp
lib/Target/R600/SIInstrInfo.h
lib/Target/R600/SIShrinkInstructions.cpp
test/CodeGen/R600/and.ll
test/CodeGen/R600/imm.ll
test/CodeGen/R600/select64.ll
test/CodeGen/R600/sint_to_fp.f64.ll
test/CodeGen/R600/uint_to_fp.f64.ll