[llvm-profgen] Deduplicate and improve warning for truncated context

[wenlei]

Summary

This change improves the warning for truncated context by: 1) deduplicate them as one call without probe can appear in many different context leading to duplicated warnings , 2) rephrase the message to make it easier to understand. The term "untracked frame" can be confusing.

[wenlei]

fix typo

[hoy]

LGTM. Thanks for the fix.

llvm/tools/llvm-profgen/PerfReader.h

 //===-- PerfReader.h - perfscript reader -----------------------*- C++ -*-===//
 //
 // 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
 //
 //===----------------------------------------------------------------------===//
 
 #ifndef LLVM_TOOLS_LLVM_PROFGEN_PERFREADER_H
 #define LLVM_TOOLS_LLVM_PROFGEN_PERFREADER_H
 #include "ErrorHandling.h"
 #include "ProfiledBinary.h"
 #include "llvm/Support/Casting.h"
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Support/Regex.h"
+#include <cstdint>
 #include <fstream>
 #include <list>
 #include <map>
 #include <vector>
 
 using namespace llvm;
 using namespace sampleprof;
 
struct ProbeStack {
   ProbeStack(const ProfiledBinary *B) : Binary(B) {}
   bool pushFrame(UnwindState::ProfiledFrame *Cur) {
     const MCDecodedPseudoProbe *CallProbe =
         Binary->getCallProbeForAddr(Cur->Address);
     // We may not find a probe for a merged or external callsite.
     // Callsite merging may cause the loss of original probe IDs.
     // Cutting off the context from here since the inliner will
     // not know how to consume a context with unknown callsites.
-    if (!CallProbe) {
-      if (!Cur->isLeafFrame())
-        WithColor::warning()
-            << "Untracked frame at " << format("%" PRIx64, Cur->Address)
-            << " due to missing call probe\n";
+    if (!CallProbe)
       return false;
-    }
     Stack.push_back(CallProbe);
     return true;
   }
 
   void popFrame() {
     if (!Stack.empty())
       Stack.pop_back();
   }

 unwinding process we will record each call stack as context id and LBR/linear
 range as sample counter for further CS profile generation.
 */
 class VirtualUnwinder {
 public:
   VirtualUnwinder(ContextSampleCounterMap *Counter, const ProfiledBinary *B)
       : CtxCounterMap(Counter), Binary(B) {}
   bool unwind(const PerfSample *Sample, uint64_t Repeat);
+  std::set<uint64_t> &getUntrackedCallsites() { return UntrackedCallsites; }
 
 private:
   bool isCallState(UnwindState &State) const {
     // The tail call frame is always missing here in stack sample, we will
     // use a specific tail call tracker to infer it.
     return Binary->addressIsCall(State.getCurrentLBRSource());
   }
 
private:
   void recordRangeCount(uint64_t Start, uint64_t End, UnwindState &State,
                         uint64_t Repeat);
   void recordBranchCount(const LBREntry &Branch, UnwindState &State,
                          uint64_t Repeat);
 
   ContextSampleCounterMap *CtxCounterMap;
   // Profiled binary that current frame address belongs to
   const ProfiledBinary *Binary;
+  // Keep track of all untracked callsites
+  std::set<uint64_t> UntrackedCallsites;
 };
 
 // Read perf trace to parse the events and samples.
 class PerfReaderBase {
 public:
   PerfReaderBase(ProfiledBinary *B) : Binary(B) {
     // Initialize the base address to preferred address.
     Binary->setBaseAddress(Binary->getPreferredBaseAddress());

llvm/tools/llvm-profgen/PerfReader.cpp

if (!Cur->isDummyRoot()) {
     if (!Stack.pushFrame(Cur)) {
       // Process truncated context
       // Start a new traversal ignoring its bottom context
       T EmptyStack(Binary);
       collectSamplesFromFrame(Cur, EmptyStack);
       for (const auto &Item : Cur->Children) {
         collectSamplesFromFrameTrie(Item.second.get(), EmptyStack);
       }
+
+      // Keep note of untracked call site and deduplicate them
+      // for warning later.
+      if (!Cur->isLeafFrame())
+        UntrackedCallsites.insert(Cur->Address);
+
       return;
     }
   }
 
   collectSamplesFromFrame(Cur, Stack);
   // Process children frame
   for (const auto &Item : Cur->Children) {
     collectSamplesFromFrameTrie(Item.second.get(), Stack);

 void HybridPerfReader::writeRawProfile(raw_fd_ostream &OS) {
   OS << "Binary(" << Binary->getName().str() << ")'s Range Counter:\n";
   printRangeCounter(SampleCounters, Binary, OS);
   OS << "\nBinary(" << Binary->getName().str() << ")'s Branch Counter:\n";
   printBranchCounter(SampleCounters, Binary, OS);
 }
 
 void HybridPerfReader::unwindSamples() {
+  std::set<uint64_t> AllUntrackedCallsites;
   for (const auto &Item : AggregatedSamples) {
     const PerfSample *Sample = Item.first.getPtr();
     VirtualUnwinder Unwinder(&SampleCounters, Binary);
     Unwinder.unwind(Sample, Item.second);
+    auto &CurrUntrackedCallsites = Unwinder.getUntrackedCallsites();
+    AllUntrackedCallsites.insert(CurrUntrackedCallsites.begin(),
+                                 CurrUntrackedCallsites.end());
   }
 
+  // Warn about untracked frames due to missing probes.
+  for (auto Address : AllUntrackedCallsites)
+    WithColor::warning() << "Profile context truncated due to missing probe "
+                         << "for call instructions at "
+                         << format("%" PRIx64, Address) << "\n";
+
   if (SkipSymbolization)
     PerfReaderBase::writeRawProfile(OutputFilename);
 }
 
 bool PerfReaderBase::extractLBRStack(TraceStream &TraceIt,
                                      SmallVectorImpl<LBREntry> &LBRStack) {
   // The raw format of LBR stack is like:
   // 0x4005c8/0x4005dc/P/-/-/0 0x40062f/0x4005b0/P/-/-/0 ...