[PATCH] D19132: AMDGPU: Correct reported code size
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Thu Apr 14 14:20:58 PDT 2016
arsenm created this revision.
arsenm added reviewers: tstellarAMD, vpykhtin, SamWot.
arsenm added a subscriber: llvm-commits.
Herald added a subscriber: arsenm.
For a while this has been counting many instructions as 0,
so the printed code size comment hasn't been very meaningful.
This sets the instruction size to 8 for all 64-bit instructions.
For instructions which may include an additional 4 byte literal,
it is set to 0 because the size must be determined from the operands.
This currently breaks the disassembler for all of the 4-byte
instructions which may have a literal.
http://reviews.llvm.org/D19132
Files:
lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
lib/Target/AMDGPU/CIInstructions.td
lib/Target/AMDGPU/MCTargetDesc/SIMCCodeEmitter.cpp
lib/Target/AMDGPU/SIInstrFormats.td
lib/Target/AMDGPU/SIInstrInfo.cpp
lib/Target/AMDGPU/SIInstrInfo.h
lib/Target/AMDGPU/SIInstrInfo.td
test/CodeGen/AMDGPU/hsa.ll
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D19132.53785.patch
Type: text/x-patch
Size: 10935 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20160414/6b91999f/attachment.bin>
More information about the llvm-commits
mailing list