================
@@ -898,15 +899,21 @@ static void defineHLSLInterlockedFunc(Sema &S,
NamespaceDecl *NS,
for (QualType ElemTy : Elems)
for (LangAS AS : AddrSpaces)
- for (bool ThreeArg : {false, true})
+ for (bool ThreeArg : {false, true}) {
+ if (RequiresOriginalValue && !ThreeArg)
+ continue;
buildAtomicOverload(S, NS, FuncName, BuiltinName, ElemTy, AS,
ThreeArg);
+ }
----------------
hekota wrote:
```suggestion
if (!RequiresOriginalValue)
buildAtomicOverload(S, NS, FuncName, BuiltinName, ElemTy, AS,
/*ThreeArg*/ false);
buildAtomicOverload(S, NS, FuncName, BuiltinName, ElemTy, AS,
/*ThreeArg*/ true);
```
Suggestion. Looping over {false, true} and then conditionally skipping over one
iteration feels a little clumsy.
https://github.com/llvm/llvm-project/pull/222161
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
