AMDGPU: Fix a comment

[nhaehnle]

This verifier check will complain if there aren't enough implicit
operands -- so it doesn't allow those operands, it requires them.

[nhaehnle]

(no review expected)

[llvmbot]

@llvm/pr-subscribers-backend-amdgpu

Author: Nicolai Hähnle (nhaehnle)

Changes

This verifier check will complain if there aren't enough implicit
operands -- so it doesn't allow those operands, it requires them.

---

Full diff: https://github.com/llvm/llvm-project/pull/169403.diff

1 Files Affected:

• (modified) llvm/lib/Target/AMDGPU/SIInstrInfo.cpp (+4-3)

diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
index a7333e3373f38..7504452dc777f 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -5518,9 +5518,10 @@ bool SIInstrInfo::verifyInstruction(const MachineInstr &MI,
         Desc.getNumOperands() + Desc.implicit_uses().size();
     const unsigned NumImplicitOps = IsDst ? 2 : 1;
 
-    // Allow additional implicit operands. This allows a fixup done by the post
-    // RA scheduler where the main implicit operand is killed and implicit-defs
-    // are added for sub-registers that remain live after this instruction.
+    // Require additional implicit operands. This allows a fixup done by the
+    // post RA scheduler where the main implicit operand is killed and
+    // implicit-defs are added for sub-registers that remain live after this
+    // instruction.
     if (MI.getNumOperands() < StaticNumOps + NumImplicitOps) {
       ErrInfo = "missing implicit register operands";
       return false;