This stack of changes introduces `llvm-profgen` utility which generates a profile data file from given perf script data files for sample-based PGO. It’s part of(not only) the CSSPGO work. Specifically to support context-sensitive with/without pseudo probe profile, it implements a series of functionalities including perf trace parsing, instruction symbolization, LBR stack/call frame stack unwinding, pseudo probe decoding, etc. Also high throughput is achieved by multiple levels of sample aggregation and compatible format with one stop is generated at the end. Please refer to: https://groups.google.com/g/llvm-dev/c/1p1rdYbL93s for the CSSPGO RFC.
As a starter, this change sets up an entry point by introducing PerfReader to load profiled binaries and perf traces(including perf events and perf samples). For the event, here it parses the mmap2 events from perf script to build the loader snaps, which is used to retrieve the image load address in the subsequent perf tracing parsing.
As described in llvm-profgen.rst, the tool being built aims to support multiple input perf data (preprocessed by perf script) as well as multiple input binary images. It should also support dynamic reload/unload shared objects by leveraging the loader snaps being built by this change
Reviewed By: wenlei, wmi
Differential Revision: https://reviews.llvm.org/D89707
--- /dev/null
+llvm-profgen - LLVM SPGO profile generation tool
+=================================
+
+.. program:: llvm-profgen
+
+SYNOPSIS
+--------
+
+:program:`llvm-profgen` [*commands*] [*options*]
+
+DESCRIPTION
+-----------
+
+The :program:`llvm-profgen` utility generates a profile data file
+from given perf script data files for sample-based profile guided
+optimization(SPGO).
+
+COMMANDS
+--------
+At least one of the following commands are required:
+
+.. option:: --perfscript=<string[,string,...]>
+
+ Path of perf-script trace created by Linux perf tool with `script`
+ command(the raw perf.data should be profiled with -b).
+
+.. option:: --output=<string>
+
+ Path of the output profile file.
+
+OPTIONS
+-------
+:program:`llvm-profgen` supports the following options:
+
+.. option:: --binary=<string[,string,...]>
+
+ Path of the input profiled binary files. If no file path is specified, the
+ path of the actual profiled binaries will be used instead.
+
+.. option:: --show-mmap-events
+
+ Print mmap events.
--- /dev/null
+import subprocess
+import lit.util
+
+config.suffixes = ['.test', '.ll', '.s', '.yaml']
+if not lit.util.which("llvm-profgen", config.llvm_tools_dir):
+ config.unsupported = True
--- /dev/null
+; RUN: llvm-profgen --perfscript=%s --output=%t --show-mmap-events | FileCheck %s
+
+PERF_RECORD_MMAP2 2580483/2580483: [0x400000(0x1000) @ 0 103:01 539973862 1972407324]: r-xp /home/a.out
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f2505b40000(0x224000) @ 0 08:04 19532214 4169021329]: r-xp /usr/lib64/ld-2.17.so
+PERF_RECORD_MMAP2 2580483/2580483: [0x7ffe88097000(0x1000) @ 0 00:00 0 0]: r-xp [vdso]
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f2505d56000(0xa000) @ 0 08:04 19530021 4190740662]: r-xp /usr/lib64/perf_fopen_hook.so
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f250593c000(0x204000) @ 0 08:04 19532229 3585508847]: r-xp /usr/lib64/libdl-2.17.so
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f250556e000(0x3ce000) @ 0 08:04 19532221 4003737677]: r-xp /usr/lib64/libc-2.17.so
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f2505358000(0x216000) @ 0 08:04 19534595 2609212015]: r-xp /usr/lib64/libz.so.1.2.7
+ 7f2505b49811
+ 0x7f2505b49811/0x7f2505b509f0/P/-/-/0 0x7f2505b4974c/0x7f2505b4975b/P/-/-/0 0x7f2505b49837/0x7f2505b49720/P/-/-/0 0x7f2505b50a5a/0x7f2505b49816/P/-/-/0 0x7f2505b50a27/0x7f2505b50a50/P/-/-/0 0x7f2505b50a36/0x7f2505b50a20/P/-/-/0 0x7f2505b59dd0/0x7f2505b50a34/P/-/-/0 0x7f2505b59db4/0x7f2505b59dc3/P/-/-/0 0x7f2505b50a2f/0x7f2505b59db0/P/-/-/0 0x7f2505b50a15/0x7f2505b50a29/P/-/-/0 0x7f2505b59dd0/0x7f2505b50a05/P/-/-/0 0x7f2505b59db4/0x7f2505b59dc3/P/-/-/0 0x7f2505b50a00/0x7f2505b59db0/P/-/-/0 0x7f2505b49811/0x7f2505b509f0/P/-/-/0 0x7f2505b4974c/0x7f2505b4975b/P/-/-/0 0x7f2505b4a08a/0x7f2505b496a0/P/-/-/0
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f2505d56000(0x8000) @ 0 08:04 19530021 4190740662]: r-xp /usr/lib64/perf_fopen_hook.so
+ 4006b1
+ 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f2505156000(0x202000) @ 0 103:01 539962022 734061270]: r-xp /home/hoy/test/dlopen/helper.so
+ 4006b1
+ 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0 0x4006b1/0x4006a0/P/-/-/0
+PERF_RECORD_MMAP2 2580483/2580483: [0x7f2505156000(0x202000) @ 0 103:01 539962022 734061270]: r-xp /home/hoy/test/dlopen/helper.so
+
+
+; CHECK: Mmap: Binary /home/a.out loaded at 0x400000
+; CHECK: Mmap: Binary /usr/lib64/ld-2.17.so loaded at 0x7f2505b40000
+; CHECK: Mmap: Binary [vdso] loaded at 0x7ffe88097000
+; CHECK: Mmap: Binary /usr/lib64/perf_fopen_hook.so loaded at 0x7f2505d56000
+; CHECK: Mmap: Binary /usr/lib64/libdl-2.17.so loaded at 0x7f250593c000
+; CHECK: Mmap: Binary /usr/lib64/libc-2.17.so loaded at 0x7f250556e000
+; CHECK: Mmap: Binary /usr/lib64/libz.so.1.2.7 loaded at 0x7f2505358000
+; CHECK: Mmap: Binary /usr/lib64/perf_fopen_hook.so loaded at 0x7f2505d56000
+; CHECK: Mmap: Binary /home/hoy/test/dlopen/helper.so loaded at 0x7f2505156000
+; CHECK: Mmap: Binary /home/hoy/test/dlopen/helper.so loaded at 0x7f2505156000
--- /dev/null
+set(LLVM_LINK_COMPONENTS
+ Core
+ ProfileData
+ Support
+ Symbolize
+ )
+
+add_llvm_tool(llvm-profgen
+ llvm-profgen.cpp
+ PerfReader.cpp
+ )
--- /dev/null
+//===-- ErrorHandling.h - Error handler -------------------------*- 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_ERRORHANDLING_H
+#define LLVM_TOOLS_LLVM_PROFGEN_ERRORHANDLING_H
+
+#include "llvm/ADT/Twine.h"
+#include "llvm/Support/Errc.h"
+#include "llvm/Support/Error.h"
+#include "llvm/Support/ErrorOr.h"
+#include "llvm/Support/WithColor.h"
+#include <system_error>
+
+using namespace llvm;
+
+LLVM_ATTRIBUTE_NORETURN inline void
+exitWithError(const Twine &Message, StringRef Whence = StringRef(),
+ StringRef Hint = StringRef()) {
+ WithColor::error(errs(), "llvm-profgen");
+ if (!Whence.empty())
+ errs() << Whence.str() << ": ";
+ errs() << Message << "\n";
+ if (!Hint.empty())
+ WithColor::note() << Hint.str() << "\n";
+ ::exit(EXIT_FAILURE);
+}
+
+LLVM_ATTRIBUTE_NORETURN inline void
+exitWithError(std::error_code EC, StringRef Whence = StringRef()) {
+ exitWithError(EC.message(), Whence);
+}
+
+LLVM_ATTRIBUTE_NORETURN inline void exitWithError(Error E, StringRef Whence) {
+ exitWithError(errorToErrorCode(std::move(E)), Whence);
+}
+#endif
--- /dev/null
+;===- ./tools/llvm-profgen/LLVMBuild.txt ----------------------*- Conf -*--===;
+;
+; 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
+;
+;===------------------------------------------------------------------------===;
+;
+; This is an LLVMBuild description file for the components in this subdirectory.
+;
+; For more information on the LLVMBuild system, please see:
+;
+; http://llvm.org/docs/LLVMBuild.html
+;
+;===------------------------------------------------------------------------===;
+
+[component_0]
+type = Tool
+name = llvm-profgen
+parent = Tools
+required_libraries = Support
--- /dev/null
+//===-- PerfReader.cpp - 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
+//
+//===----------------------------------------------------------------------===//
+#include "PerfReader.h"
+
+static cl::opt<bool> ShowMmapEvents("show-mmap-events", cl::ReallyHidden,
+ cl::init(false), cl::ZeroOrMore,
+ cl::desc("Print binary load events."));
+
+namespace llvm {
+namespace sampleprof {
+
+PerfReader::PerfReader(cl::list<std::string> &BinaryFilenames) {
+ // Load the binaries.
+ for (auto Filename : BinaryFilenames)
+ loadBinary(Filename, /*AllowNameConflict*/ false);
+}
+
+ProfiledBinary &PerfReader::loadBinary(const StringRef BinaryPath,
+ bool AllowNameConflict) {
+ // The binary table is currently indexed by the binary name not the full
+ // binary path. This is because the user-given path may not match the one
+ // that was actually executed.
+ StringRef BinaryName = llvm::sys::path::filename(BinaryPath);
+
+ // Call to load the binary in the ctor of ProfiledBinary.
+ auto Ret = BinaryTable.insert({BinaryName, ProfiledBinary(BinaryPath)});
+
+ if (!Ret.second && !AllowNameConflict) {
+ std::string ErrorMsg = "Binary name conflict: " + BinaryPath.str() +
+ " and " + Ret.first->second.getPath().str() + " \n";
+ exitWithError(ErrorMsg);
+ }
+
+ return Ret.first->second;
+}
+
+void PerfReader::updateBinaryAddress(const MMapEvent &Event) {
+ // Load the binary.
+ StringRef BinaryPath = Event.BinaryPath;
+ StringRef BinaryName = llvm::sys::path::filename(BinaryPath);
+
+ auto I = BinaryTable.find(BinaryName);
+ // Drop the event which doesn't belong to user-provided binaries
+ // or if its image is loaded at the same address
+ if (I == BinaryTable.end() || Event.BaseAddress == I->second.getBaseAddress())
+ return;
+
+ ProfiledBinary &Binary = I->second;
+
+ // A binary image could be uploaded and then reloaded at different
+ // place, so update the address map here
+ AddrToBinaryMap.erase(Binary.getBaseAddress());
+ AddrToBinaryMap[Event.BaseAddress] = &Binary;
+
+ // Update binary load address.
+ Binary.setBaseAddress(Event.BaseAddress);
+}
+
+void PerfReader::parseMMap2Event(TraceStream &TraceIt) {
+ // Parse a line like:
+ // PERF_RECORD_MMAP2 2113428/2113428: [0x7fd4efb57000(0x204000) @ 0
+ // 08:04 19532229 3585508847]: r-xp /usr/lib64/libdl-2.17.so
+ constexpr static const char *const Pattern =
+ "PERF_RECORD_MMAP2 ([0-9]+)/[0-9]+: "
+ "\\[(0x[a-f0-9]+)\\((0x[a-f0-9]+)\\) @ "
+ "(0x[a-f0-9]+|0) .*\\]: [-a-z]+ (.*)";
+ // Field 0 - whole line
+ // Field 1 - PID
+ // Field 2 - base address
+ // Field 3 - mmapped size
+ // Field 4 - page offset
+ // Field 5 - binary path
+ enum EventIndex {
+ WHOLE_LINE = 0,
+ PID = 1,
+ BASE_ADDRESS = 2,
+ MMAPPED_SIZE = 3,
+ PAGE_OFFSET = 4,
+ BINARY_PATH = 5
+ };
+
+ Regex RegMmap2(Pattern);
+ SmallVector<StringRef, 6> Fields;
+ bool R = RegMmap2.match(TraceIt.getCurrentLine(), &Fields);
+ if (!R) {
+ std::string ErrorMsg = "Cannot parse mmap event: Line" +
+ Twine(TraceIt.getLineNumber()).str() + ": " +
+ TraceIt.getCurrentLine().str() + " \n";
+ exitWithError(ErrorMsg);
+ }
+ MMapEvent Event;
+ Fields[PID].getAsInteger(10, Event.PID);
+ Fields[BASE_ADDRESS].getAsInteger(0, Event.BaseAddress);
+ Fields[MMAPPED_SIZE].getAsInteger(0, Event.Size);
+ Fields[PAGE_OFFSET].getAsInteger(0, Event.Offset);
+ Event.BinaryPath = Fields[BINARY_PATH];
+ updateBinaryAddress(Event);
+ if (ShowMmapEvents) {
+ outs() << "Mmap: Binary " << Event.BinaryPath << " loaded at "
+ << format("0x%" PRIx64 ":", Event.BaseAddress) << " \n";
+ }
+}
+
+void PerfReader::parseEvent(TraceStream &TraceIt) {
+ if (TraceIt.getCurrentLine().startswith("PERF_RECORD_MMAP2"))
+ parseMMap2Event(TraceIt);
+
+ TraceIt.advance();
+}
+
+void PerfReader::parseTrace(StringRef Filename) {
+ // Trace line iterator
+ TraceStream TraceIt(Filename);
+ while (!TraceIt.isAtEoF()) {
+ parseEvent(TraceIt);
+ }
+}
+
+void PerfReader::parsePerfTraces(cl::list<std::string> &PerfTraceFilenames) {
+ // Parse perf traces.
+ for (auto Filename : PerfTraceFilenames)
+ parseTrace(Filename);
+}
+
+} // namespace sampleprof
+} // namespace llvm
--- /dev/null
+//===-- 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/CommandLine.h"
+#include "llvm/Support/Regex.h"
+#include <fstream>
+#include <list>
+#include <map>
+#include <vector>
+
+using namespace llvm;
+using namespace sampleprof;
+
+namespace llvm {
+namespace sampleprof {
+
+// Stream based trace line iterator
+class TraceStream {
+ std::string CurrentLine;
+ std::ifstream Fin;
+ bool IsAtEoF = false;
+ uint64_t LineNumber = 0;
+
+public:
+ TraceStream(StringRef Filename) : Fin(Filename.str()) {
+ if (!Fin.good())
+ exitWithError("Error read input perf script file", Filename);
+ advance();
+ }
+
+ StringRef getCurrentLine() {
+ assert(!IsAtEoF && "Line iterator reaches the End-of-File!");
+ return CurrentLine;
+ }
+
+ uint64_t getLineNumber() { return LineNumber; }
+
+ bool isAtEoF() { return IsAtEoF; }
+
+ // Read the next line
+ void advance() {
+ if (!std::getline(Fin, CurrentLine)) {
+ IsAtEoF = true;
+ return;
+ }
+ LineNumber++;
+ }
+};
+
+// Filename to binary map
+using BinaryMap = StringMap<ProfiledBinary>;
+// Address to binary map for fast look-up
+using AddressBinaryMap = std::map<uint64_t, ProfiledBinary *>;
+
+// Load binaries and read perf trace to parse the events and samples
+class PerfReader {
+
+ BinaryMap BinaryTable;
+ AddressBinaryMap AddrToBinaryMap; // Used by address-based lookup.
+
+ // The parsed MMap event
+ struct MMapEvent {
+ pid_t PID = 0;
+ uint64_t BaseAddress = 0;
+ uint64_t Size = 0;
+ uint64_t Offset = 0;
+ StringRef BinaryPath;
+ };
+
+ /// Load symbols and disassemble the code of a give binary.
+ /// Also register the binary in the binary table.
+ ///
+ ProfiledBinary &loadBinary(const StringRef BinaryPath,
+ bool AllowNameConflict = true);
+ void updateBinaryAddress(const MMapEvent &Event);
+
+public:
+ PerfReader(cl::list<std::string> &BinaryFilenames);
+
+ /// Parse a single line of a PERF_RECORD_MMAP2 event looking for a
+ /// mapping between the binary name and its memory layout.
+ ///
+ void parseMMap2Event(TraceStream &TraceIt);
+ void parseEvent(TraceStream &TraceIt);
+ // Parse perf events and samples
+ void parseTrace(StringRef Filename);
+ void parsePerfTraces(cl::list<std::string> &PerfTraceFilenames);
+};
+
+} // end namespace sampleprof
+} // end namespace llvm
+
+#endif
--- /dev/null
+//===-- ProfiledBinary.h - Binary decoder -----------------------*- 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_PROFILEDBINARY_H
+#define LLVM_TOOLS_LLVM_PROFGEN_PROFILEDBINARY_H
+#include "llvm/ADT/StringRef.h"
+#include "llvm/Support/Path.h"
+
+namespace llvm {
+namespace sampleprof {
+
+class ProfiledBinary {
+ std::string Path;
+ mutable uint64_t BaseAddress = 0;
+
+public:
+ ProfiledBinary(StringRef Path) : Path(Path) { load(); }
+
+ const StringRef getPath() const { return Path; }
+ const StringRef getName() const { return llvm::sys::path::filename(Path); }
+ uint64_t getBaseAddress() const { return BaseAddress; }
+ void setBaseAddress(uint64_t Address) { BaseAddress = Address; }
+
+private:
+ void load() {
+ // TODO:
+ }
+};
+
+} // end namespace sampleprof
+} // end namespace llvm
+
+#endif
--- /dev/null
+//===- llvm-profgen.cpp - LLVM SPGO profile generation tool ---------------===//
+//
+// 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
+//
+//===----------------------------------------------------------------------===//
+//
+// llvm-profgen generates SPGO profiles from perf script ouput.
+//
+//===----------------------------------------------------------------------===//
+
+#include "ErrorHandling.h"
+#include "PerfReader.h"
+#include "ProfiledBinary.h"
+#include "llvm/Support/CommandLine.h"
+#include "llvm/Support/InitLLVM.h"
+
+static cl::list<std::string> PerfTraceFilenames(
+ "perfscript", cl::value_desc("perfscript"), cl::OneOrMore,
+ llvm::cl::MiscFlags::CommaSeparated,
+ cl::desc("Path of perf-script trace created by Linux perf tool with "
+ "`script` command(the raw perf.data should be profiled with -b)"));
+
+static cl::list<std::string>
+ BinaryFilenames("binary", cl::value_desc("binary"), cl::ZeroOrMore,
+ llvm::cl::MiscFlags::CommaSeparated,
+ cl::desc("Path of profiled binary files"));
+
+static cl::opt<std::string> OutputFilename("output", cl::value_desc("output"),
+ cl::Required,
+ cl::desc("Output profile file"));
+
+using namespace llvm;
+using namespace sampleprof;
+
+int main(int argc, const char *argv[]) {
+ InitLLVM X(argc, argv);
+
+ cl::ParseCommandLineOptions(argc, argv, "llvm SPGO profile generator\n");
+
+ // Load binaries and parse perf events and samples
+ PerfReader Reader(BinaryFilenames);
+ Reader.parsePerfTraces(PerfTraceFilenames);
+
+ return EXIT_SUCCESS;
+}
projects / platform / upstream / llvm.git / commitdiff