[llvm] 6654578 - [LLVM][AArch64] Enable SVEIntrinsicOpts at all optimisation levels.

[Paul Walker via llvm-commits]

Author: Paul Walker
Date: 2024-10-09T14:17:39Z
New Revision: 665457815f11118f7e755a471f33606c8562a4be

URL: https://github.com/llvm/llvm-project/commit/665457815f11118f7e755a471f33606c8562a4be
DIFF: https://github.com/llvm/llvm-project/commit/665457815f11118f7e755a471f33606c8562a4be.diff

LOG: [LLVM][AArch64] Enable SVEIntrinsicOpts at all optimisation levels.

Added: 
    

Modified: 
    flang/test/Driver/default-backend-pipelines.f90
    llvm/lib/Target/AArch64/AArch64TargetMachine.cpp

Removed: 
    


################################################################################
diff  --git a/flang/test/Driver/default-backend-pipelines.f90 b/flang/test/Driver/default-backend-pipelines.f90
index e6910c170f1179..12b929e9fa40f6 100644
--- a/flang/test/Driver/default-backend-pipelines.f90
+++ b/flang/test/Driver/default-backend-pipelines.f90
@@ -5,9 +5,9 @@
 ! RUN: %flang_fc1 -S -O2 %s -triple aarch64-unknown-linux-gnu -mllvm -debug-pass=Structure -o /dev/null 2>&1 | FileCheck %s --check-prefix=CHECK-O2
 ! RUN: %flang_fc1 -S -O3 %s -triple aarch64-unknown-linux-gnu -mllvm -debug-pass=Structure -o /dev/null 2>&1 | FileCheck %s --check-prefix=CHECK-O3
 
-! CHECK-O2-NOT: SVE intrinsics optimizations
+! CHECK-O2-NOT: Optimize selects
 
-! CHECK-O3: SVE intrinsics optimizations
+! CHECK-O3: Optimize selects
 
 subroutine simple_loop
   integer :: i

diff  --git a/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp b/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp
index 9f96f6c5e83ec4..7b0ae23358673e 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp
+++ b/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp
@@ -588,7 +588,7 @@ void AArch64PassConfig::addIRPasses() {
 
   // Expand any SVE vector library calls that we can't code generate directly.
   if (EnableSVEIntrinsicOpts &&
-      TM->getOptLevel() == CodeGenOptLevel::Aggressive)
+      TM->getOptLevel() != CodeGenOptLevel::None)
     addPass(createSVEIntrinsicOptsPass());
 
   // Cmpxchg instructions are often used with a subsequent comparison to