[llvm] [AMDGPU] Calc IsVALU correctly during UADDO/USUBO selection (PR #159814)

[Jay Foad via llvm-commits]

================
@@ -0,0 +1,26 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
+; RUN: llc -mtriple=amdgcn-amd-amdpal -stop-after=amdgpu-isel < %s | FileCheck -check-prefixes=GCN-ISEL %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx900 < %s | FileCheck %s
+; Ensure that S_UADDO_PSEUDO is selected when carryout user is S_ADD_CO_PSEUDO
+
+; GCN-ISEL-LABEL: name:   s_uaddo_pseudo
+; GCN-ISEL-LABEL: body:
+; GCN-ISEL: S_UADDO_PSEUDO
+; GCN-ISEL: S_ADD_CO_PSEUDO
+
+define amdgpu_ps i32 @s_uaddo_pseudo(i32 inreg %val0) {
+; CHECK-LABEL: s_uaddo_pseudo:
+; CHECK:       ; %bb.0:
+; CHECK-NEXT:    s_add_u32 s0, s0, 1
+; CHECK-NEXT:    s_cselect_b64 s[0:1], -1, 0
+; CHECK-NEXT:    s_cmp_lg_u64 s[0:1], 0
+; CHECK-NEXT:    s_addc_u32 s0, 1, 0
----------------
jayfoad wrote:

Objectively this code is still terrible. It should just keep the carry in scc and feed it directly from s_add -> s_addc.

https://github.com/llvm/llvm-project/pull/159814