[llvm] Swap UnrollAndJam Pass to before the SLP Vectorizer Pass (PR #97029)

[via llvm-commits]

github-actions[bot] wrote:

<!--LLVM CODE FORMAT COMMENT: {clang-format}-->


:warning: C/C++ code formatter, clang-format found issues in your code. :warning:

<details>
<summary>
You can test this locally with the following command:
</summary>

``````````bash
git-clang-format --diff bfa0cae31c1db7627c41dcebd660f5eaea796778 5088bacf52f667ed327018e958a01da9bf546d70 --extensions cpp -- llvm/lib/Passes/PassBuilderPipelines.cpp
``````````

</details>

<details>
<summary>
View the diff from clang-format here.
</summary>

``````````diff
diff --git a/llvm/lib/Passes/PassBuilderPipelines.cpp b/llvm/lib/Passes/PassBuilderPipelines.cpp
index 77743f275b..a8f9a1590c 100644
--- a/llvm/lib/Passes/PassBuilderPipelines.cpp
+++ b/llvm/lib/Passes/PassBuilderPipelines.cpp
@@ -1238,14 +1238,17 @@ void PassBuilder::addVectorPasses(OptimizationLevel Level,
     FPM.addPass(InferAlignmentPass());
 
   // We do UnrollAndJam in a separate LPM to Unroll to ensure it happens first.
-  // In order for outer loop vectorization to be done, UnrollAndJam must occur before the SLPVectorizerPass.
-  // Placing UnrollAndJam immediately after the LoopVectorizePass when !IsFullLTO leads to improved compile times versus
-  // placing it immediately before the SLPVectorizerPass, due to analysis re-use.
+  // In order for outer loop vectorization to be done, UnrollAndJam must occur
+  // before the SLPVectorizerPass. Placing UnrollAndJam immediately after the
+  // LoopVectorizePass when !IsFullLTO leads to improved compile times versus
+  // placing it immediately before the SLPVectorizerPass, due to analysis
+  // re-use.
   if (EnableUnrollAndJam && PTO.LoopUnrolling) {
     // Cleanup after loop vectorization. Simplification passes like CVP and
     // GVN, loop transforms, and others have already run, so it's now better to
     // convert to more optimized IR using more aggressive simplify CFG options.
-    // SimplifyCFGPass must be run before UnrollAndJam for UnrollAndJam-SLP outer loop vectorization to happen.
+    // SimplifyCFGPass must be run before UnrollAndJam for UnrollAndJam-SLP
+    // outer loop vectorization to happen.
     FPM.addPass(SimplifyCFGPass(SimplifyCFGOptions()
                                     .forwardSwitchCondToPhi(true)
                                     .convertSwitchRangeToICmp(true)

``````````

</details>


https://github.com/llvm/llvm-project/pull/97029