Files
clang-p2996/llvm/lib/Transforms/Utils/CallGraphUpdater.cpp
Arthur Eubanks 71497cc7a4 [CGSCC] Fix compile time blowup with large RefSCCs (#94815)
In some modules, e.g. Kotlin-generated IR, we end up with a huge RefSCC
and the call graph updates done as a result of the inliner take a long
time. This is due to RefSCC::removeInternalRefEdges() getting called
many times, each time removing one function from the RefSCC, but each
call to removeInternalRefEdges() is proportional to the size of the
RefSCC.

There are two places that call removeInternalRefEdges(), in
updateCGAndAnalysisManagerForPass() and
LazyCallGraph::removeDeadFunction().

1) Since LazyCallGraph can deal with spurious (edges that exist in the
graph but not in the IR) ref edges, we can simply not call
removeInternalRefEdges() in updateCGAndAnalysisManagerForPass().

2) LazyCallGraph::removeDeadFunction() still ends up taking the brunt of
compile time with the above change for the original reason. So instead
we batch all the dead function removals so we can call
removeInternalRefEdges() just once. This requires some changes to
callers of removeDeadFunction() to not actually erase the function from
the module, but defer it to when we batch delete dead functions at the
end of the CGSCC run, leaving the function body as "unreachable" in the
meantime. We still need to ensure that call edges are accurate. I had
also tried deleting dead functions after visiting a RefSCC, but deleting
them all at once at the end was simpler.

Many test changes are due to not performing unnecessary revisits of an
SCC (the CGSCC infrastructure deems ref edge refinements as unimportant
when it comes to revisiting SCCs, although that seems to not be
consistently true given these changes) because we don't remove some ref
edges. Specifically for devirt-invalidated.ll this seems to expose an
inlining order issue with the inliner. Probably unimportant for this
type of intentionally weird call graph.

Compile time:
https://llvm-compile-time-tracker.com/compare.php?from=6f2c61071c274a1b5e212e6ad4114641ec7c7fc3&to=b08c90d05e290dd065755ea776ceaf1420680224&stat=instructions:u
2024-06-11 09:50:13 -07:00

177 lines
6.2 KiB
C++

//===- CallGraphUpdater.cpp - A (lazy) call graph update helper -----------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
/// \file
///
/// This file provides interfaces used to manipulate a call graph, regardless
/// if it is a "old style" CallGraph or an "new style" LazyCallGraph.
///
//===----------------------------------------------------------------------===//
#include "llvm/Transforms/Utils/CallGraphUpdater.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/CallGraphSCCPass.h"
#include "llvm/IR/Constants.h"
#include "llvm/Transforms/Utils/ModuleUtils.h"
using namespace llvm;
bool CallGraphUpdater::finalize() {
if (!DeadFunctionsInComdats.empty()) {
filterDeadComdatFunctions(DeadFunctionsInComdats);
DeadFunctions.append(DeadFunctionsInComdats.begin(),
DeadFunctionsInComdats.end());
}
if (CG) {
// First remove all references, e.g., outgoing via called functions. This is
// necessary as we can delete functions that have circular references.
for (Function *DeadFn : DeadFunctions) {
DeadFn->removeDeadConstantUsers();
CallGraphNode *DeadCGN = (*CG)[DeadFn];
DeadCGN->removeAllCalledFunctions();
CG->getExternalCallingNode()->removeAnyCallEdgeTo(DeadCGN);
DeadFn->replaceAllUsesWith(PoisonValue::get(DeadFn->getType()));
}
// Then remove the node and function from the module.
for (Function *DeadFn : DeadFunctions) {
CallGraphNode *DeadCGN = CG->getOrInsertFunction(DeadFn);
assert(DeadCGN->getNumReferences() == 0 &&
"References should have been handled by now");
delete CG->removeFunctionFromModule(DeadCGN);
}
} else {
// This is the code path for the new lazy call graph and for the case were
// no call graph was provided.
for (Function *DeadFn : DeadFunctions) {
DeadFn->removeDeadConstantUsers();
DeadFn->replaceAllUsesWith(PoisonValue::get(DeadFn->getType()));
if (LCG && !ReplacedFunctions.count(DeadFn)) {
// Taken mostly from the inliner:
LazyCallGraph::Node &N = LCG->get(*DeadFn);
auto *DeadSCC = LCG->lookupSCC(N);
assert(DeadSCC && DeadSCC->size() == 1 &&
&DeadSCC->begin()->getFunction() == DeadFn);
auto &DeadRC = DeadSCC->getOuterRefSCC();
FunctionAnalysisManager &FAM =
AM->getResult<FunctionAnalysisManagerCGSCCProxy>(*DeadSCC, *LCG)
.getManager();
FAM.clear(*DeadFn, DeadFn->getName());
AM->clear(*DeadSCC, DeadSCC->getName());
LCG->markDeadFunction(*DeadFn);
// Mark the relevant parts of the call graph as invalid so we don't
// visit them.
UR->InvalidatedSCCs.insert(DeadSCC);
UR->InvalidatedRefSCCs.insert(&DeadRC);
UR->DeadFunctions.push_back(DeadFn);
} else {
// The CGSCC infrastructure batch deletes functions at the end of the
// call graph walk, so only erase the function if we're not using that
// infrastructure.
// The function is now really dead and de-attached from everything.
DeadFn->eraseFromParent();
}
}
}
bool Changed = !DeadFunctions.empty();
DeadFunctionsInComdats.clear();
DeadFunctions.clear();
return Changed;
}
void CallGraphUpdater::reanalyzeFunction(Function &Fn) {
if (CG) {
CallGraphNode *OldCGN = CG->getOrInsertFunction(&Fn);
OldCGN->removeAllCalledFunctions();
CG->populateCallGraphNode(OldCGN);
} else if (LCG) {
LazyCallGraph::Node &N = LCG->get(Fn);
LazyCallGraph::SCC *C = LCG->lookupSCC(N);
updateCGAndAnalysisManagerForCGSCCPass(*LCG, *C, N, *AM, *UR, *FAM);
}
}
void CallGraphUpdater::registerOutlinedFunction(Function &OriginalFn,
Function &NewFn) {
if (CG)
CG->addToCallGraph(&NewFn);
else if (LCG)
LCG->addSplitFunction(OriginalFn, NewFn);
}
void CallGraphUpdater::removeFunction(Function &DeadFn) {
DeadFn.deleteBody();
DeadFn.setLinkage(GlobalValue::ExternalLinkage);
if (DeadFn.hasComdat())
DeadFunctionsInComdats.push_back(&DeadFn);
else
DeadFunctions.push_back(&DeadFn);
// For the old call graph we remove the function from the SCC right away.
if (CG && !ReplacedFunctions.count(&DeadFn)) {
CallGraphNode *DeadCGN = (*CG)[&DeadFn];
DeadCGN->removeAllCalledFunctions();
CGSCC->DeleteNode(DeadCGN);
}
if (FAM)
FAM->clear(DeadFn, DeadFn.getName());
}
void CallGraphUpdater::replaceFunctionWith(Function &OldFn, Function &NewFn) {
OldFn.removeDeadConstantUsers();
ReplacedFunctions.insert(&OldFn);
if (CG) {
// Update the call graph for the newly promoted function.
CallGraphNode *OldCGN = (*CG)[&OldFn];
CallGraphNode *NewCGN = CG->getOrInsertFunction(&NewFn);
NewCGN->stealCalledFunctionsFrom(OldCGN);
CG->ReplaceExternalCallEdge(OldCGN, NewCGN);
// And update the SCC we're iterating as well.
CGSCC->ReplaceNode(OldCGN, NewCGN);
} else if (LCG) {
// Directly substitute the functions in the call graph.
LazyCallGraph::Node &OldLCGN = LCG->get(OldFn);
SCC->getOuterRefSCC().replaceNodeFunction(OldLCGN, NewFn);
}
removeFunction(OldFn);
}
bool CallGraphUpdater::replaceCallSite(CallBase &OldCS, CallBase &NewCS) {
// This is only necessary in the (old) CG.
if (!CG)
return true;
Function *Caller = OldCS.getCaller();
CallGraphNode *NewCalleeNode =
CG->getOrInsertFunction(NewCS.getCalledFunction());
CallGraphNode *CallerNode = (*CG)[Caller];
if (llvm::none_of(*CallerNode, [&OldCS](const CallGraphNode::CallRecord &CR) {
return CR.first && *CR.first == &OldCS;
}))
return false;
CallerNode->replaceCallEdge(OldCS, NewCS, NewCalleeNode);
return true;
}
void CallGraphUpdater::removeCallSite(CallBase &CS) {
// This is only necessary in the (old) CG.
if (!CG)
return;
Function *Caller = CS.getCaller();
CallGraphNode *CallerNode = (*CG)[Caller];
CallerNode->removeCallEdgeFor(CS);
}