[clang] [CIR][AMDGPU] Route printf in device code through the OpenCL runtime (PR #226435)

[Steffen Larsen via cfe-commits]

================
@@ -1096,3 +1096,45 @@ CIRGenFunction::emitAMDGPUBuiltinExpr(unsigned builtinId,
     return std::nullopt;
   }
 }
+
+// Emit AMDGPU printf CIR stand-in function call. This stand-in function call is
+// lowered to the appropriate call structure during LLVM IR lowering.
+mlir::Value
+CIRGenFunction::emitAMDGPUDevicePrintfCallExpr(const CallExpr *expr) {
+  assert(cgm.getTriple().isAMDGCN() ||
+         (cgm.getTriple().isSPIRV() &&
+          cgm.getTriple().getVendor() == llvm::Triple::AMD));
+  assert(expr->getBuiltinCallee() == Builtin::BIprintf ||
+         expr->getBuiltinCallee() == Builtin::BI__builtin_printf);
+  assert(expr->getNumArgs() >= 1);
+
+  const FunctionProtoType *funcPrototype =
+      expr->getDirectCallee()->getType()->getAs<FunctionProtoType>();
+  CallArgList args;
+  emitCallArgs(args, funcPrototype, expr->arguments(), expr->getDirectCallee());
+
+  mlir::Location loc = getLoc(expr->getBeginLoc());
+
+  // We don't know how to emit non-scalar varargs.
+  bool hasNonScalar = llvm::any_of(args, [&](const CallArg &a) {
+    return a.hasLValue() || !a.getKnownRValue().isScalar();
+  });
+  if (hasNonScalar) {
+    cgm.errorUnsupported(expr, "non-scalar args to printf");
+    return builder.getConstInt(loc, builder.getSInt32Ty(), 0);
+  }
+
+  llvm::SmallVector<mlir::Value, 8> callArgs;
+  for (const CallArg &a : args)
+    callArgs.push_back(a.getKnownRValue().getValue());
+
+  // int __cir_amdgpu_printf(char *format, ...);
+  auto fnTy = cir::FuncType::get({cir::PointerType::get(builder.getSInt8Ty())},
+                                 builder.getSInt32Ty(),
+                                 /*isVarArg=*/true);
+  cir::FuncOp fn = cgm.createRuntimeFunction(fnTy, "__cir_amdgpu_printf");
----------------
steffenlarsen wrote:

CUDA doesn't seem to need it, in the way they have it implemented. We could generalize it, but it wouldn't change the fact that it would be the first of its kind.

https://github.com/llvm/llvm-project/pull/226435