[llvm] [AMDGPU] Infer amdgpu-no-flat-scratch-init attribute in AMDGPUAttributor (PR #94647)
Jun Wang via llvm-commits
llvm-commits at lists.llvm.org
Sat Aug 17 17:44:59 PDT 2024
================
@@ -677,6 +693,33 @@ struct AAAMDAttributesFunction : public AAAMDAttributes {
return !A.checkForAllCallLikeInstructions(DoesNotRetrieve, *this,
UsedAssumedInformation);
}
+
+ // Returns true if FlatScratchInit is needed, i.e., no-flat-scratch-init is
+ // not to be set.
+ bool needFlatScratchInit(Attributor &A) {
+ // This is called on each callee; false means callee shouldn't have
+ // no-flat-scratch-init.
+ auto CheckForNoFlatScratchInit = [&](Instruction &I) {
+ const auto &CB = cast<CallBase>(I);
+ const Value *CalleeOp = CB.getCalledOperand();
+ const Function *Callee = dyn_cast<Function>(CalleeOp);
+ if (!Callee) // indirect call
+ return CB.isInlineAsm();
----------------
jwanggit86 wrote:
On the other hand, a patch where `checkForAllCallees()` is used is as follows:
```diff
bool needFlatScratchInit(Attributor &A) {
assert(isAssumed(FLAT_SCRATCH_INIT)); // only called if the bit is still set
auto CheckForNoFlatScratchInit = [&](Instruction &I) {
const auto &CB = cast<CallBase>(I);
- const Function *Callee = CB.getCalledFunction();
-
- if (Callee && Callee->isIntrinsic())
- return Callee->getIntrinsicID() !=
- Intrinsic::amdgcn_addrspacecast_nonnull;
-
+ auto PredOnCallees = [&](ArrayRef<const Function *> Callees) {
+ for (const Function *Callee : Callees) {
+ if (Callee->isIntrinsic())
+ if (Callee->getIntrinsicID() == Intrinsic::amdgcn_addrspacecast_nonnull)
+ return false;
+ }
+ return true;
+ };
+ if(!A.checkForAllCallees(PredOnCallees, *this, CB))
+ return CB.isInlineAsm();
return true;
};
bool UsedAssumedInformation = false;
return !A.checkForAllCallLikeInstructions(CheckForNoFlatScratchInit, *this,
```
https://github.com/llvm/llvm-project/pull/94647
More information about the llvm-commits
mailing list