diff --git a/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp b/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp
index 8e97da6..bbfcc8b 100644
--- a/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp
+++ b/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp
@@ -1635,12 +1635,8 @@
     case Intrinsic::amdgcn_struct_buffer_load_format:
       return simplifyAMDGCNMemoryIntrinsicDemanded(II, DemandedElts);
     default: {
-      if (getAMDGPUImageDMaskIntrinsic(II->getIntrinsicID())) {
-        assert(cast<ConstantInt>(
-                 II->getArgOperand(
-                   II->getNumOperands() - 2))->getZExtValue() == 0);
+      if (getAMDGPUImageDMaskIntrinsic(II->getIntrinsicID()))
         return simplifyAMDGCNMemoryIntrinsicDemanded(II, DemandedElts, 0);
-      }
 
       break;
     }
diff --git a/test/Transforms/InstCombine/AMDGPU/amdgcn-demanded-vector-elts.ll b/test/Transforms/InstCombine/AMDGPU/amdgcn-demanded-vector-elts.ll
index e0b8fab..6c03ed2 100644
--- a/test/Transforms/InstCombine/AMDGPU/amdgcn-demanded-vector-elts.ll
+++ b/test/Transforms/InstCombine/AMDGPU/amdgcn-demanded-vector-elts.ll
@@ -2402,6 +2402,17 @@
 
 declare {<4 x float>, i32} @llvm.amdgcn.image.load.1d.sl_v4f32i32s.i32(i32, i32, <8 x i32>, i32, i32) #1
 
+; CHECK: @tfe_check_assert(
+; CHECK: %data = call float @llvm.amdgcn.image.load.2d.f32.i32(i32 1, i32 undef, i32 undef, <8 x i32> undef, i32 0, i32 1)
+; CHECK-NEXT: ret float %data
+define amdgpu_hs float @tfe_check_assert() #0 {
+  %data = call nsz <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32 15, i32 undef, i32 undef, <8 x i32> undef, i32 0, i32 1) #2
+  %elt0 = extractelement <4 x float> %data, i32 0
+  ret float %elt0
+}
+
+declare <4 x float> @llvm.amdgcn.image.load.2d.v4f32.i32(i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #1
+
 attributes #0 = { nounwind }
 attributes #1 = { nounwind readonly }