[llvm] [RFC] implement convergence control in MIR using SelectionDAG (PR #71785)
Matt Arsenault via llvm-commits
llvm-commits at lists.llvm.org
Mon Nov 13 01:11:58 PST 2023
================
@@ -901,6 +901,20 @@ bool AMDGPUTargetLowering::aggressivelyPreferBuildVectorSources(EVT VecVT) const
return true;
}
+void AMDGPUTargetLowering::CollectTargetIntrinsicOperands(
+ const CallInst &I, SmallVectorImpl<SDValue> &Ops, SelectionDAG &DAG,
+ function_ref<SDValue(const Value *)> getValue) const {
+ if (auto Bundle = I.getOperandBundle(LLVMContext::OB_convergencectrl)) {
----------------
arsenm wrote:
There's nothing AMDGPU specific here. This is general infrastructure so I would expect the target callback to be unnecessary
https://github.com/llvm/llvm-project/pull/71785
More information about the llvm-commits
mailing list