[PATCH] D62134: MC: Allow getMaxInstLength to depend on the subtarget

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
arsenm added reviewers: echristo, rampitec.
Herald added subscribers: tpr, nhaehnle, wdng, jvesely.

Keep it optional in cases this is ever needed in some global
context. Currently it's only used for getting an upper bound inline
asm code size.

      

For AMDGPU, gfx10 increases the maximum instruction size to
12-bytes. This avoids penalizing older subtargets when estimating code
size, and making some annoying branch relaxation test adjustments.


https://reviews.llvm.org/D62134

Files:
  include/llvm/CodeGen/TargetInstrInfo.h
  include/llvm/MC/MCAsmInfo.h
  lib/CodeGen/TargetInstrInfo.cpp
  lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.cpp
  lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.h
  lib/Target/AMDGPU/SIInstrInfo.cpp
  lib/Target/Hexagon/HexagonInstrInfo.cpp
  lib/Target/Hexagon/HexagonInstrInfo.h
  test/CodeGen/AMDGPU/branch-relaxation-inst-size-gfx10.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D62134.200245.patch
Type: text/x-patch
Size: 8880 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190520/d6a44ad9/attachment.bin>