[AArch64] Turn on FP Load Balancing by Default for Cortex-A53

[zzheng]

Summary

We havd found that the FP Load Balancing pass designed for Cortex-A57 does not hurt Cortex-A53 performance. It would be helpful in the scenario where "-mcpu=cortex-a53" is used for Cortex-A57 hardware.

Will make this pass and it's test case target agnostic in the next patch.

[t.p.northover]

This seems obvious enough, as you've done the testing for it. Go ahead.

Tim.

[zzheng]

Not a committer yet... Can someone do me the favor?

Thanks a lot.

[mcrosier]

Committed in r220744.

lib/Target/AArch64/AArch64TargetMachine.cpp

   if (TM->getOptLevel() != CodeGenOpt::None && EnableDeadRegisterElimination)
     addPass(createAArch64DeadRegisterDefinitions());
   if (TM->getOptLevel() != CodeGenOpt::None &&
-      TM->getSubtarget<AArch64Subtarget>().isCortexA57() &&
+      (TM->getSubtarget<AArch64Subtarget>().isCortexA53() ||
+       TM->getSubtarget<AArch64Subtarget>().isCortexA57()) &&
       usingDefaultRegAlloc())
     // Improve performance for some FP/SIMD code for A57.
     addPass(createAArch64A57FPLoadBalancing());

test/CodeGen/AArch64/aarch64-a57-fp-load-balancing.ll

-; RUN: llc < %s -mcpu=cortex-a57 -aarch64-a57-fp-load-balancing-override=1 -aarch64-a57-fp-load-balancing-force-all | FileCheck %s --check-prefix CHECK --check-prefix CHECK-EVEN
-; RUN: llc < %s -mcpu=cortex-a57 -aarch64-a57-fp-load-balancing-override=2 -aarch64-a57-fp-load-balancing-force-all | FileCheck %s --check-prefix CHECK --check-prefix CHECK-ODD
+; RUN: llc < %s -mcpu=cortex-a57 -aarch64-a57-fp-load-balancing-override=1 -aarch64-a57-fp-load-balancing-force-all | FileCheck %s --check-prefix CHECK --check-prefix CHECK-A57 --check-prefix CHECK-EVEN
+; RUN: llc < %s -mcpu=cortex-a57 -aarch64-a57-fp-load-balancing-override=2 -aarch64-a57-fp-load-balancing-force-all | FileCheck %s --check-prefix CHECK --check-prefix CHECK-A57 --check-prefix CHECK-ODD
+; RUN: llc < %s -mcpu=cortex-a53 -aarch64-a57-fp-load-balancing-override=1 -aarch64-a57-fp-load-balancing-force-all | FileCheck %s --check-prefix CHECK --check-prefix CHECK-A53 --check-prefix CHECK-EVEN
+; RUN: llc < %s -mcpu=cortex-a53 -aarch64-a57-fp-load-balancing-override=2 -aarch64-a57-fp-load-balancing-force-all | FileCheck %s --check-prefix CHECK --check-prefix CHECK-A53 --check-prefix CHECK-ODD
 
 ; Test the AArch64A57FPLoadBalancing pass. This pass relies heavily on register allocation, so
 ; our test strategy is to:

 ; CHECK: fmsub [[x]]
 ; CHECK: fmadd [[y]]
 ; CHECK: fmadd [[x]]
-; CHECK: stp [[x]], [[y]]
+; CHECK-A57: stp [[x]], [[y]]
+; CHECK-A53-DAG: str [[x]]
+; CHECK-A53-DAG: str [[y]]
 
 define void @f2(double* nocapture readonly %p, double* nocapture %q) #0 {
 entry:

 ; CHECK: fmsub [[x]]
 ; CHECK: fmadd [[y]]
 ; CHECK: fmadd [[x]]
-; CHECK: stp [[x]], [[y]]
+; CHECK-A57: stp [[x]], [[y]]
+; CHECK-A53-DAG: str [[x]]
+; CHECK-A53-DAG: str [[y]]
 
 define void @f4(float* nocapture readonly %p, float* nocapture %q) #0 {
 entry: