[PATCH] D33170: [X86] Adding avx512_vpopcntdq feature set and its intrinsics

[Craig Topper via Phabricator via cfe-commits]

craig.topper added inline comments.


================
Comment at: lib/CodeGen/CGBuiltin.cpp:7526
+    llvm::Type *ResultType = ConvertType(E->getType());
+    Value *X = EmitScalarExpr(E->getArg(0));
+    llvm::Function *F = CGM.getIntrinsic(Intrinsic::ctpop, ResultType);
----------------
I'm not sure what EmitScalarExpr does, but I got think its not right for a vector argument.


================
Comment at: lib/CodeGen/CGBuiltin.cpp:7527
+    Value *X = EmitScalarExpr(E->getArg(0));
+    llvm::Function *F = CGM.getIntrinsic(Intrinsic::ctpop, ResultType);
+    return Builder.CreateCall(F, X);
----------------
I think you can use Ops[0]->getType() and avoid the call to ConvertType. See the code for BI__builtin_ia32_vplzcntq_512_mask.


Repository:
  rL LLVM

https://reviews.llvm.org/D33170