[PATCH] D157857: [AMDGPU] Make VOPD insts with a FMAMK or FMAMK component have a fixed size

[Mirko Brkusanin via Phabricator via llvm-commits]

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

This fixes a failure from the expensive-checks buildbot


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D157857

Files:
  llvm/lib/Target/AMDGPU/VOPDInstructions.td


Index: llvm/lib/Target/AMDGPU/VOPDInstructions.td
===================================================================
--- llvm/lib/Target/AMDGPU/VOPDInstructions.td
+++ llvm/lib/Target/AMDGPU/VOPDInstructions.td
@@ -111,6 +111,7 @@
   let Inst{16-9} = !if (!eq(VDX.Mnemonic, "v_mov_b32"), 0x0, vsrc1X);
   let Inst{48-41} = !if (!eq(VDY.Mnemonic, "v_mov_b32"), 0x0, vsrc1Y);
   let Size = 12;
+  let FixedSize = 1;
 }
 
 // V_DUAL_DOT2ACC_F32_BF16 is a legal instruction, but V_DOT2ACC_F32_BF16 is


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D157857.549867.patch
Type: text/x-patch
Size: 498 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230814/4d04741a/attachment.bin>