aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/lib')
-rw-r--r--llvm/lib/Target/AArch64/AArch64.h2
-rw-r--r--llvm/lib/Target/AArch64/AArch64TargetMachine.cpp6
-rw-r--r--llvm/lib/Target/AArch64/MachineSMEABIPass.cpp31
3 files changed, 31 insertions, 8 deletions
diff --git a/llvm/lib/Target/AArch64/AArch64.h b/llvm/lib/Target/AArch64/AArch64.h
index 8d0ff41..1396841 100644
--- a/llvm/lib/Target/AArch64/AArch64.h
+++ b/llvm/lib/Target/AArch64/AArch64.h
@@ -60,7 +60,7 @@ FunctionPass *createAArch64CleanupLocalDynamicTLSPass();
FunctionPass *createAArch64CollectLOHPass();
FunctionPass *createSMEABIPass();
FunctionPass *createSMEPeepholeOptPass();
-FunctionPass *createMachineSMEABIPass();
+FunctionPass *createMachineSMEABIPass(CodeGenOptLevel);
ModulePass *createSVEIntrinsicOptsPass();
InstructionSelector *
createAArch64InstructionSelector(const AArch64TargetMachine &,
diff --git a/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp b/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp
index dde1d88..6606a73 100644
--- a/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp
+++ b/llvm/lib/Target/AArch64/AArch64TargetMachine.cpp
@@ -770,8 +770,8 @@ bool AArch64PassConfig::addGlobalInstructionSelect() {
}
void AArch64PassConfig::addMachineSSAOptimization() {
- if (EnableNewSMEABILowering && TM->getOptLevel() != CodeGenOptLevel::None)
- addPass(createMachineSMEABIPass());
+ if (TM->getOptLevel() != CodeGenOptLevel::None && EnableNewSMEABILowering)
+ addPass(createMachineSMEABIPass(TM->getOptLevel()));
if (TM->getOptLevel() != CodeGenOptLevel::None && EnableSMEPeepholeOpt)
addPass(createSMEPeepholeOptPass());
@@ -804,7 +804,7 @@ bool AArch64PassConfig::addILPOpts() {
void AArch64PassConfig::addPreRegAlloc() {
if (TM->getOptLevel() == CodeGenOptLevel::None && EnableNewSMEABILowering)
- addPass(createMachineSMEABIPass());
+ addPass(createMachineSMEABIPass(CodeGenOptLevel::None));
// Change dead register definitions to refer to the zero register.
if (TM->getOptLevel() != CodeGenOptLevel::None &&
diff --git a/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp b/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp
index cced0fa..6e0cebb 100644
--- a/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp
+++ b/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp
@@ -63,6 +63,7 @@
#include "llvm/CodeGen/LivePhysRegs.h"
#include "llvm/CodeGen/MachineBasicBlock.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
+#include "llvm/CodeGen/MachineLoopInfo.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/TargetRegisterInfo.h"
@@ -70,6 +71,12 @@ using namespace llvm;
#define DEBUG_TYPE "aarch64-machine-sme-abi"
+static cl::opt<int>
+ LoopEdgeWeight("aarch64-sme-abi-loop-edge-weight", cl::ReallyHidden,
+ cl::init(10),
+ cl::desc("Edge weight for basic blocks witin loops (used "
+ "for placing ZA saves/restores)"));
+
namespace {
enum ZAState {
@@ -238,7 +245,8 @@ getZAStateBeforeInst(const TargetRegisterInfo &TRI, MachineInstr &MI,
struct MachineSMEABI : public MachineFunctionPass {
inline static char ID = 0;
- MachineSMEABI() : MachineFunctionPass(ID) {}
+ MachineSMEABI(CodeGenOptLevel OptLevel = CodeGenOptLevel::Default)
+ : MachineFunctionPass(ID), OptLevel(OptLevel) {}
bool runOnMachineFunction(MachineFunction &MF) override;
@@ -247,6 +255,9 @@ struct MachineSMEABI : public MachineFunctionPass {
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
AU.addRequired<EdgeBundlesWrapperLegacy>();
+ // Only analyse loops at -01 and above.
+ if (OptLevel != CodeGenOptLevel::None)
+ AU.addRequired<MachineLoopInfoWrapperPass>();
AU.addPreservedID(MachineLoopInfoID);
AU.addPreservedID(MachineDominatorsID);
MachineFunctionPass::getAnalysisUsage(AU);
@@ -329,12 +340,15 @@ struct MachineSMEABI : public MachineFunctionPass {
MachineBasicBlock::iterator MBBI, DebugLoc DL);
private:
+ CodeGenOptLevel OptLevel = CodeGenOptLevel::Default;
+
MachineFunction *MF = nullptr;
const AArch64Subtarget *Subtarget = nullptr;
const AArch64RegisterInfo *TRI = nullptr;
const AArch64FunctionInfo *AFI = nullptr;
const TargetInstrInfo *TII = nullptr;
MachineRegisterInfo *MRI = nullptr;
+ MachineLoopInfo *MLI = nullptr;
};
FunctionInfo MachineSMEABI::collectNeededZAStates(SMEAttrs SMEFnAttrs) {
@@ -436,18 +450,23 @@ MachineSMEABI::assignBundleZAStates(const EdgeBundles &Bundles,
LLVM_DEBUG(dbgs() << " (no state preference)\n");
continue;
}
+ bool IsLoop = MLI && MLI->getLoopFor(MF->getBlockNumbered(BlockID));
bool InEdge = Bundles.getBundle(BlockID, /*Out=*/false) == I;
bool OutEdge = Bundles.getBundle(BlockID, /*Out=*/true) == I;
+ int EdgeWeight = IsLoop ? LoopEdgeWeight : 1;
+ if (IsLoop)
+ LLVM_DEBUG(dbgs() << " IsLoop");
+ LLVM_DEBUG(dbgs() << " (EdgeWeight: " << EdgeWeight << ')');
ZAState DesiredIncomingState = Block.Insts.front().NeededState;
if (InEdge && isLegalEdgeBundleZAState(DesiredIncomingState)) {
- EdgeStateCounts[DesiredIncomingState]++;
+ EdgeStateCounts[DesiredIncomingState] += EdgeWeight;
LLVM_DEBUG(dbgs() << " DesiredIncomingState: "
<< getZAStateString(DesiredIncomingState));
}
ZAState DesiredOutgoingState = Block.Insts.back().NeededState;
if (OutEdge && isLegalEdgeBundleZAState(DesiredOutgoingState)) {
- EdgeStateCounts[DesiredOutgoingState]++;
+ EdgeStateCounts[DesiredOutgoingState] += EdgeWeight;
LLVM_DEBUG(dbgs() << " DesiredOutgoingState: "
<< getZAStateString(DesiredOutgoingState));
}
@@ -849,6 +868,8 @@ bool MachineSMEABI::runOnMachineFunction(MachineFunction &MF) {
TII = Subtarget->getInstrInfo();
TRI = Subtarget->getRegisterInfo();
MRI = &MF.getRegInfo();
+ if (OptLevel != CodeGenOptLevel::None)
+ MLI = &getAnalysis<MachineLoopInfoWrapperPass>().getLI();
const EdgeBundles &Bundles =
getAnalysis<EdgeBundlesWrapperLegacy>().getEdgeBundles();
@@ -877,4 +898,6 @@ bool MachineSMEABI::runOnMachineFunction(MachineFunction &MF) {
return true;
}
-FunctionPass *llvm::createMachineSMEABIPass() { return new MachineSMEABI(); }
+FunctionPass *llvm::createMachineSMEABIPass(CodeGenOptLevel OptLevel) {
+ return new MachineSMEABI(OptLevel);
+}
generated by cgit v1.1 at 2025-09-27 11:41:36 +0000