[llvm] [LV] Enable auto-vectorisation of loops with uncountable exits (PR #133099)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-llvm-transforms

Author: David Sherwood (david-arm)

<details>
<summary>Changes</summary>

Until now the feature to enable vectorisation of some early exit
loops with uncountable exits was controlled under a flag, off by
default. Now that we have efficient code generation for
vectorising such loops (see PR #<!-- -->130766) and we are still far
enough away from the next LLVM release it seems like a good time
to enable the feature by default. If any issues arise
post-commit it can be easily reverted.

Using this patch I built and ran the LLVM test suite successfully,
which on neoverse-v1 led to the vectorisation of 114 additional
early exit loops. I then performed a bootstrap build of clang,
which built cleanly with 64 extra early exit loops vectorising.

I also built and ran SPEC2017 successfully with no change in
performance for both neoverse-v1 and neoverse-v2.

---
Full diff: https://github.com/llvm/llvm-project/pull/133099.diff


7 Files Affected:

- (modified) llvm/lib/Transforms/Vectorize/LoopVectorize.cpp (+1-1) 
- (modified) llvm/test/Transforms/LoopVectorize/AArch64/simple_early_exit.ll (+1-1) 
- (modified) llvm/test/Transforms/LoopVectorize/multi_early_exit.ll (+1-1) 
- (modified) llvm/test/Transforms/LoopVectorize/multi_early_exit_live_outs.ll (+1-1) 
- (modified) llvm/test/Transforms/LoopVectorize/single_early_exit.ll (+1-1) 
- (modified) llvm/test/Transforms/LoopVectorize/single_early_exit_live_outs.ll (+1-1) 
- (modified) llvm/test/Transforms/LoopVectorize/single_early_exit_with_outer_loop.ll (+1-1) 


``````````diff
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index 5244a5e7b1c41..7010715947a54 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -396,7 +396,7 @@ static cl::opt<bool> UseWiderVFIfCallVariantsPresent(
     cl::desc("Try wider VFs if they enable the use of vector variants"));
 
 static cl::opt<bool> EnableEarlyExitVectorization(
-    "enable-early-exit-vectorization", cl::init(false), cl::Hidden,
+    "enable-early-exit-vectorization", cl::init(true), cl::Hidden,
     cl::desc(
         "Enable vectorization of early exit loops with uncountable exits."));
 
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/simple_early_exit.ll b/llvm/test/Transforms/LoopVectorize/AArch64/simple_early_exit.ll
index 7d5b73477f6ed..b30f46195a326 100644
--- a/llvm/test/Transforms/LoopVectorize/AArch64/simple_early_exit.ll
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/simple_early_exit.ll
@@ -1,5 +1,5 @@
 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
-; RUN: opt -S < %s -p loop-vectorize -enable-early-exit-vectorization | FileCheck %s --check-prefixes=CHECK
+; RUN: opt -S < %s -p loop-vectorize | FileCheck %s --check-prefixes=CHECK
 
 target triple = "aarch64-unknown-linux-gnu"
 
diff --git a/llvm/test/Transforms/LoopVectorize/multi_early_exit.ll b/llvm/test/Transforms/LoopVectorize/multi_early_exit.ll
index 0e753a535cd2d..94af5b7c7607d 100644
--- a/llvm/test/Transforms/LoopVectorize/multi_early_exit.ll
+++ b/llvm/test/Transforms/LoopVectorize/multi_early_exit.ll
@@ -1,5 +1,5 @@
 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
-; RUN: opt -S < %s -p loop-vectorize -enable-early-exit-vectorization | FileCheck %s
+; RUN: opt -S < %s -p loop-vectorize | FileCheck %s
 
 declare void @init_mem(ptr, i64);
 
diff --git a/llvm/test/Transforms/LoopVectorize/multi_early_exit_live_outs.ll b/llvm/test/Transforms/LoopVectorize/multi_early_exit_live_outs.ll
index 4027f6a0f5dfd..7759c10032e9b 100644
--- a/llvm/test/Transforms/LoopVectorize/multi_early_exit_live_outs.ll
+++ b/llvm/test/Transforms/LoopVectorize/multi_early_exit_live_outs.ll
@@ -1,5 +1,5 @@
 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
-; RUN: opt -S < %s -p loop-vectorize -enable-early-exit-vectorization | FileCheck %s
+; RUN: opt -S < %s -p loop-vectorize | FileCheck %s
 
 declare void @init_mem(ptr, i64);
 
diff --git a/llvm/test/Transforms/LoopVectorize/single_early_exit.ll b/llvm/test/Transforms/LoopVectorize/single_early_exit.ll
index dedf5f0be624e..4b580e42f009e 100644
--- a/llvm/test/Transforms/LoopVectorize/single_early_exit.ll
+++ b/llvm/test/Transforms/LoopVectorize/single_early_exit.ll
@@ -1,5 +1,5 @@
 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
-; RUN: opt -S < %s -p loop-vectorize -enable-early-exit-vectorization -force-vector-width=4 | FileCheck %s
+; RUN: opt -S < %s -p loop-vectorize -force-vector-width=4 | FileCheck %s
 
 declare void @init_mem(ptr, i64);
 
diff --git a/llvm/test/Transforms/LoopVectorize/single_early_exit_live_outs.ll b/llvm/test/Transforms/LoopVectorize/single_early_exit_live_outs.ll
index 14651d60e1532..df9d2b477ec68 100644
--- a/llvm/test/Transforms/LoopVectorize/single_early_exit_live_outs.ll
+++ b/llvm/test/Transforms/LoopVectorize/single_early_exit_live_outs.ll
@@ -1,5 +1,5 @@
 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
-; RUN: opt -S < %s -p loop-vectorize -enable-early-exit-vectorization -force-vector-width=4 | FileCheck %s
+; RUN: opt -S < %s -p loop-vectorize -force-vector-width=4 | FileCheck %s
 
 declare void @init_mem(ptr, i64);
 
diff --git a/llvm/test/Transforms/LoopVectorize/single_early_exit_with_outer_loop.ll b/llvm/test/Transforms/LoopVectorize/single_early_exit_with_outer_loop.ll
index 51cfc72752014..da26c962c7d2b 100644
--- a/llvm/test/Transforms/LoopVectorize/single_early_exit_with_outer_loop.ll
+++ b/llvm/test/Transforms/LoopVectorize/single_early_exit_with_outer_loop.ll
@@ -1,4 +1,4 @@
-; RUN: opt -S < %s -p loop-vectorize,'print<loops>' -disable-output -enable-early-exit-vectorization 2>&1 | FileCheck %s
+; RUN: opt -S < %s -p loop-vectorize,'print<loops>' -disable-output 2>&1 | FileCheck %s
 
 declare void @init_mem(ptr, i64);
 

``````````

</details>


https://github.com/llvm/llvm-project/pull/133099