[llvm] [AMDGPU] Don't cluster lds instrs together (PR #180908)

via llvm-commits llvm-commits at lists.llvm.org
Sun Feb 15 06:42:52 PST 2026


================
@@ -1,6 +1,8 @@
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx942 -O3 -debug-only=machine-scheduler 2>&1 < %s | FileCheck -check-prefix=DEBUG %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx942 -O3 -debug-only=machine-scheduler 2>&1 < %s
----------------
lijinpei-amd wrote:

Hello~ @lucas-rami  Can I change this lit test to only check it compiles successfully? After my commit, `DAG.MinOccupancy` increase from 3 to 4(which I think benefits performance), so the testcase won't pass anymore.  The files used to cause compiler core dump before your commit. As the testcase is sensitive to scheduler results, can I just check it compiles successfully?

https://github.com/llvm/llvm-project/pull/180908


More information about the llvm-commits mailing list