Fix forward mode ldg (#2065)

* Fix forward mode ldg

* fix

enzyme/Enzyme/GradientUtils.cpp

@@ -6078,7 +6078,7 @@ Value *GradientUtils::invertPointerM(Value *const oval, IRBuilder<> &BuilderM,
       return applyChainRule(
           II->getType(), bb,
           [&](Value *ptr) {
-            Value *args[] = {ptr};
+            Value *args[] = {ptr, getNewFromOriginal(II->getArgOperand(1))};
             auto li = bb.CreateCall(II->getCalledFunction(), args);
             llvm::SmallVector<unsigned int, 9> ToCopy2(MD_ToCopy);
             ToCopy2.push_back(LLVMContext::MD_noalias);

enzyme/test/Enzyme/ForwardMode/nvvm_ldg.ll

@@ -0,0 +1,29 @@
+; RUN: if [ %llvmver -lt 16 ]; then %opt < %s %loadEnzyme -enzyme-preopt=false -enzyme -S | FileCheck %s; fi
+; RUN: %opt < %s %newLoadEnzyme -enzyme-preopt=false -passes="enzyme" -S | FileCheck %s
+
+target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v16:16:16-v32:32:32-v64:64:64-v128:128:128-n16:32:64-ni:10:11:12:13"
+target triple = "nvptx64-nvidia-cuda"
+
+declare float @llvm.nvvm.ldg.global.f.f32.p1f32(float addrspace(1)* nocapture, i32)
+
+define float @vmul(float addrspace(1)* %inp) {
+top:
+  %ld = call float @llvm.nvvm.ldg.global.f.f32.p1f32(float addrspace(1)* %inp, i32 4)
+  ret float %ld
+}
+
+
+define float @test_derivative(float addrspace(1)* %inp, float addrspace(1)* %dinp) {
+entry:
+  %0 = tail call float (float (float addrspace(1)*)*, ...) @__enzyme_fwddiff(float (float addrspace(1)*)* nonnull @vmul, float addrspace(1)* %inp, float addrspace(1)* %dinp)
+  ret float %0
+}
+
+; Function Attrs: nounwind
+declare float @__enzyme_fwddiff(float (float addrspace(1)*)*, ...)
+
+; CHECK: define internal float @fwddiffevmul(float addrspace(1)* %inp, float addrspace(1)* %"inp'")
+; CHECK-NEXT: top:
+; CHECK-NEXT:   %[[res:.+]] = call{{( fast)?}} float @llvm.nvvm.ldg.global.f.f32.p1f32(float addrspace(1)* %"inp'", i32 4)
+; CHECK-NEXT:   ret float %[[res]]
+; CHECK-NEXT: }