[clang] [llvm] [HLSL] [DXIL] Implement the `AddUint64` HLSL function and the `UAddc` DXIL op (PR #125319)

[Farzon Lotfi via cfe-commits]

================
@@ -0,0 +1,30 @@
+; We use llc for this test so that we don't abort after the first error.
+; RUN: not llc %s -o /dev/null 2>&1 | FileCheck %s
+
+target triple = "dxil-pc-shadermodel6.3-library"
+
+; DXIL operation UAddc only supports i32. Other integer types are unsupported.
+; CHECK: error:
+; CHECK-SAME: in function uaddc_i16
+; CHECK-SAME: Cannot create UAddc operation: Invalid overload type
+
+define noundef i16 @uaddc_i16(i16 noundef %a, i16 noundef %b) "hlsl.export" {
+  %uaddc = call { i16, i1 } @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
+  %carry = extractvalue { i16, i1 } %uaddc, 1
+  %sum = extractvalue { i16, i1 } %uaddc, 0
+  %carry_zext = zext i1 %carry to i16
+  %result = add i16 %sum, %carry_zext
+  ret i16 %result
+}
+
+; CHECK: error:
----------------
farzonl wrote:

I'm assuming the `not` in the run line is letting you do multiple error checks in this file? If that is the case there are potentially many *_error.ll cases that could be consolidated.  Maybe worth writing up a ticket for this.

https://github.com/llvm/llvm-project/pull/125319