[llvm-branch-commits] [llvm] [SelectionDAG] Fold nonzero extract-of-extract indices (PR #200935)

[Matt Arsenault via llvm-branch-commits]

================
@@ -0,0 +1,43 @@
+; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 6
+; RUN: llc -O2 -global-isel=0 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx950 \
+; RUN:   -verify-machineinstrs -stop-after=amdgpu-isel < %s | FileCheck %s \
+; RUN:   --check-prefix=COMBINE --implicit-check-not=REG_SEQUENCE
+; RUN: llc -O2 -global-isel=0 -mtriple=amdgcn-amd-amdhsa -mcpu=gfx950 \
+; RUN:   -verify-machineinstrs -combiner-disabled -stop-after=amdgpu-isel < %s \
+; RUN:   | FileCheck %s --check-prefix=NOCOMBINE
----------------
arsenm wrote:

-O2 is the default. Also -global-isel=0 should only ever be used in tests that are testing both selector. Also this should just run to the end 

https://github.com/llvm/llvm-project/pull/200935