[PATCH] D153169: [AMDGPU] Fix operand class of v_ldexp_f16 src1

[Joe Nash via Phabricator via llvm-commits]

Joe_Nash created this revision.
Joe_Nash added reviewers: arsenm, foad.
Herald added subscribers: StephenFan, kerbowa, hiraditya, tpr, dstuttard, yaxunl, jvesely, kzhuravl.
Herald added a project: All.
Joe_Nash requested review of this revision.
Herald added subscribers: llvm-commits, wdng.
Herald added a project: LLVM.

Patch eece6ba283bd <https://reviews.llvm.org/rGeece6ba283bd763e6d7109ae9e155e81cfee0651> changed the src1 type of v_ldexp_f16 from i32 to
i16. Though semantically src1 is an i16, the hardware reads this operand as an
f16 type, which primarily enables floating point inline constants.
Therefore this patch changes the operand type to f16. It maintains the
current behavior where floating point source modifiers are not allowed
on src1. SDWA sext modifier continues to be allowed.
The test asm and disasm test changes in eece6ba283bd <https://reviews.llvm.org/rGeece6ba283bd763e6d7109ae9e155e81cfee0651> are reverted,
because the floating point inline constants are allowed.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D153169

Files:
  llvm/lib/Target/AMDGPU/VOP2Instructions.td
  llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
  llvm/test/MC/AMDGPU/gfx11_asm_vop3_from_vop2.s
  llvm/test/MC/AMDGPU/gfx8_asm_vop3.s
  llvm/test/MC/AMDGPU/gfx9_asm_vop3.s
  llvm/test/MC/Disassembler/AMDGPU/gfx10_vop3.txt
  llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_from_vop2.txt
  llvm/test/MC/Disassembler/AMDGPU/gfx8_vop3.txt
  llvm/test/MC/Disassembler/AMDGPU/gfx9_vop3.txt

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D153169.532289.patch
Type: text/x-patch
Size: 10364 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230616/0aefd22b/attachment.bin>