[llvm] 79971d0 - [llvm-profdata] Add option to cap profile output size
William Huang via llvm-commits
llvm-commits at lists.llvm.org
Wed Feb 8 14:22:29 PST 2023
Author: William Huang
Date: 2023-02-08T22:21:33Z
New Revision: 79971d0d771a273eaf13697c3a2e55fc0194bc09
URL: https://github.com/llvm/llvm-project/commit/79971d0d771a273eaf13697c3a2e55fc0194bc09
DIFF: https://github.com/llvm/llvm-project/commit/79971d0d771a273eaf13697c3a2e55fc0194bc09.diff
LOG: [llvm-profdata] Add option to cap profile output size
D139603 (add option to llvm-profdata to reduce output profile size) contains test cases that are not cross-platform. Moving those tests to unit test and making sure the feature is callable from llvm library
Reviewed By: snehasish
Differential Revision: https://reviews.llvm.org/D141446
Added:
llvm/unittests/tools/llvm-profdata/CMakeLists.txt
llvm/unittests/tools/llvm-profdata/OutputSizeLimitTest.cpp
Modified:
llvm/include/llvm/ProfileData/SampleProf.h
llvm/include/llvm/ProfileData/SampleProfWriter.h
llvm/lib/ProfileData/SampleProfWriter.cpp
llvm/tools/llvm-profdata/llvm-profdata.cpp
llvm/unittests/tools/CMakeLists.txt
Removed:
################################################################################
diff --git a/llvm/include/llvm/ProfileData/SampleProf.h b/llvm/include/llvm/ProfileData/SampleProf.h
index 13f0157222eca..faee9639ea860 100644
--- a/llvm/include/llvm/ProfileData/SampleProf.h
+++ b/llvm/include/llvm/ProfileData/SampleProf.h
@@ -427,6 +427,14 @@ class SampleRecord {
void print(raw_ostream &OS, unsigned Indent) const;
void dump() const;
+ bool operator==(const SampleRecord &Other) const {
+ return NumSamples == Other.NumSamples && CallTargets == Other.CallTargets;
+ }
+
+ bool operator!=(const SampleRecord &Other) const {
+ return !(*this == Other);
+ }
+
private:
uint64_t NumSamples = 0;
CallTargetMap CallTargets;
@@ -1149,6 +1157,21 @@ class FunctionSamples {
// all the inline instances and names of call targets.
void findAllNames(DenseSet<StringRef> &NameSet) const;
+ bool operator==(const FunctionSamples &Other) const {
+ return (GUIDToFuncNameMap == Other.GUIDToFuncNameMap ||
+ (GUIDToFuncNameMap && Other.GUIDToFuncNameMap &&
+ *GUIDToFuncNameMap == *Other.GUIDToFuncNameMap)) &&
+ FunctionHash == Other.FunctionHash && Context == Other.Context &&
+ TotalSamples == Other.TotalSamples &&
+ TotalHeadSamples == Other.TotalHeadSamples &&
+ BodySamples == Other.BodySamples &&
+ CallsiteSamples == Other.CallsiteSamples;
+ }
+
+ bool operator!=(const FunctionSamples &Other) const {
+ return !(*this == Other);
+ }
+
private:
/// CFG hash value for the function.
uint64_t FunctionHash = 0;
diff --git a/llvm/include/llvm/ProfileData/SampleProfWriter.h b/llvm/include/llvm/ProfileData/SampleProfWriter.h
index b1ed0335e9c93..4edb3b049e704 100644
--- a/llvm/include/llvm/ProfileData/SampleProfWriter.h
+++ b/llvm/include/llvm/ProfileData/SampleProfWriter.h
@@ -35,6 +35,56 @@ enum SectionLayout {
NumOfLayout,
};
+/// When writing a profile with size limit, user may want to use a
diff erent
+/// strategy to reduce function count other than dropping functions with fewest
+/// samples first. In this case a class implementing the same interfaces should
+/// be provided to SampleProfileWriter::writeWithSizeLimit().
+class FunctionPruningStrategy {
+protected:
+ SampleProfileMap &ProfileMap;
+ size_t OutputSizeLimit;
+
+public:
+ /// \p ProfileMap A reference to the original profile map. It will be modified
+ /// by Erase().
+ /// \p OutputSizeLimit Size limit in bytes of the output profile. This is
+ /// necessary to estimate how many functions to remove.
+ FunctionPruningStrategy(SampleProfileMap &ProfileMap, size_t OutputSizeLimit)
+ : ProfileMap(ProfileMap), OutputSizeLimit(OutputSizeLimit) {}
+
+ virtual ~FunctionPruningStrategy() = default;
+
+ /// SampleProfileWriter::writeWithSizeLimit() calls this after every write
+ /// iteration if the output size still exceeds the limit. This function
+ /// should erase some functions from the profile map so that the writer tries
+ /// to write the profile again with fewer functions. At least 1 entry from the
+ /// profile map must be erased.
+ ///
+ /// \p CurrentOutputSize Number of bytes in the output if current profile map
+ /// is written.
+ virtual void Erase(size_t CurrentOutputSize) = 0;
+};
+
+class DefaultFunctionPruningStrategy : public FunctionPruningStrategy {
+ std::vector<NameFunctionSamples> SortedFunctions;
+
+public:
+ DefaultFunctionPruningStrategy(SampleProfileMap &ProfileMap,
+ size_t OutputSizeLimit);
+
+ /// In this default implementation, functions with fewest samples are dropped
+ /// first. Since the exact size of the output cannot be easily calculated due
+ /// to compression, we use a heuristic to remove as many functions as
+ /// necessary but not too many, aiming to minimize the number of write
+ /// iterations.
+ /// Empirically, functions with larger total sample count contain linearly
+ /// more sample entries, meaning it takes linearly more space to write them.
+ /// The cumulative length is therefore quadratic if all functions are sorted
+ /// by total sample count.
+ /// TODO: Find better heuristic.
+ void Erase(size_t CurrentOutputSize) override;
+};
+
/// Sample-based profile writer. Base class.
class SampleProfileWriter {
public:
@@ -50,6 +100,17 @@ class SampleProfileWriter {
/// \returns status code of the file update operation.
virtual std::error_code write(const SampleProfileMap &ProfileMap);
+ /// Write sample profiles up to given size limit, using the pruning strategy
+ /// to drop some functions if necessary.
+ ///
+ /// \returns status code of the file update operation.
+ template <typename FunctionPruningStrategy = DefaultFunctionPruningStrategy>
+ std::error_code writeWithSizeLimit(SampleProfileMap &ProfileMap,
+ size_t OutputSizeLimit) {
+ FunctionPruningStrategy Strategy(ProfileMap, OutputSizeLimit);
+ return writeWithSizeLimitInternal(ProfileMap, OutputSizeLimit, &Strategy);
+ }
+
raw_ostream &getOutputStream() { return *OutputStream; }
/// Profile writer factory.
@@ -79,6 +140,15 @@ class SampleProfileWriter {
// Write function profiles to the profile file.
virtual std::error_code writeFuncProfiles(const SampleProfileMap &ProfileMap);
+ std::error_code writeWithSizeLimitInternal(SampleProfileMap &ProfileMap,
+ size_t OutputSizeLimit,
+ FunctionPruningStrategy *Strategy);
+
+ /// For writeWithSizeLimit in text mode, each newline takes 1 additional byte
+ /// on Windows when actually written to the file, but not written to a memory
+ /// buffer. This needs to be accounted for when rewriting the profile.
+ size_t LineCount;
+
/// Output stream where to emit the profile to.
std::unique_ptr<raw_ostream> OutputStream;
@@ -102,6 +172,7 @@ class SampleProfileWriterText : public SampleProfileWriter {
: SampleProfileWriter(OS), Indent(0) {}
std::error_code writeHeader(const SampleProfileMap &ProfileMap) override {
+ LineCount = 0;
return sampleprof_error::success;
}
diff --git a/llvm/lib/ProfileData/SampleProfWriter.cpp b/llvm/lib/ProfileData/SampleProfWriter.cpp
index 093790afe2d66..bce858a99a819 100644
--- a/llvm/lib/ProfileData/SampleProfWriter.cpp
+++ b/llvm/lib/ProfileData/SampleProfWriter.cpp
@@ -30,6 +30,7 @@
#include "llvm/Support/MD5.h"
#include "llvm/Support/raw_ostream.h"
#include <algorithm>
+#include <cmath>
#include <cstdint>
#include <memory>
#include <set>
@@ -37,9 +38,109 @@
#include <utility>
#include <vector>
+#define DEBUG_TYPE "llvm-profdata"
+
using namespace llvm;
using namespace sampleprof;
+namespace llvm {
+namespace support {
+namespace endian {
+namespace {
+
+// Adapter class to llvm::support::endian::Writer for pwrite().
+struct SeekableWriter {
+ raw_pwrite_stream &OS;
+ endianness Endian;
+ SeekableWriter(raw_pwrite_stream &OS, endianness Endian)
+ : OS(OS), Endian(Endian) {}
+
+ template <typename ValueType>
+ void pwrite(ValueType Val, size_t Offset) {
+ std::string StringBuf;
+ raw_string_ostream SStream(StringBuf);
+ Writer(SStream, Endian).write(Val);
+ OS.pwrite(StringBuf.data(), StringBuf.size(), Offset);
+ }
+};
+
+} // namespace
+} // namespace endian
+} // namespace support
+} // namespace llvm
+
+DefaultFunctionPruningStrategy::DefaultFunctionPruningStrategy(
+ SampleProfileMap &ProfileMap, size_t OutputSizeLimit)
+ : FunctionPruningStrategy(ProfileMap, OutputSizeLimit) {
+ sortFuncProfiles(ProfileMap, SortedFunctions);
+}
+
+void DefaultFunctionPruningStrategy::Erase(size_t CurrentOutputSize) {
+ double D = (double)OutputSizeLimit / CurrentOutputSize;
+ size_t NewSize = (size_t)round(ProfileMap.size() * D * D);
+ size_t NumToRemove = ProfileMap.size() - NewSize;
+ if (NumToRemove < 1)
+ NumToRemove = 1;
+
+ assert(NumToRemove <= SortedFunctions.size());
+ llvm::for_each(
+ llvm::make_range(SortedFunctions.begin() + SortedFunctions.size() -
+ NumToRemove,
+ SortedFunctions.end()),
+ [&](const NameFunctionSamples &E) { ProfileMap.erase(E.first); });
+ SortedFunctions.resize(SortedFunctions.size() - NumToRemove);
+}
+
+std::error_code SampleProfileWriter::writeWithSizeLimitInternal(
+ SampleProfileMap &ProfileMap, size_t OutputSizeLimit,
+ FunctionPruningStrategy *Strategy) {
+ if (OutputSizeLimit == 0)
+ return write(ProfileMap);
+
+ size_t OriginalFunctionCount = ProfileMap.size();
+
+ std::unique_ptr<raw_ostream> OriginalOutputStream;
+ OutputStream.swap(OriginalOutputStream);
+
+ size_t IterationCount = 0;
+ size_t TotalSize;
+
+ SmallVector<char> StringBuffer;
+ do {
+ StringBuffer.clear();
+ OutputStream.reset(new raw_svector_ostream(StringBuffer));
+ if (std::error_code EC = write(ProfileMap))
+ return EC;
+
+ TotalSize = StringBuffer.size();
+ // On Windows every "\n" is actually written as "\r\n" to disk but not to
+ // memory buffer, this
diff erence should be added when considering the total
+ // output size.
+#ifdef _WIN32
+ if (Format == SPF_Text)
+ TotalSize += LineCount;
+#endif
+ if (TotalSize <= OutputSizeLimit)
+ break;
+
+ Strategy->Erase(TotalSize);
+ IterationCount++;
+ } while (ProfileMap.size() != 0);
+
+ if (ProfileMap.size() == 0)
+ return sampleprof_error::too_large;
+
+ OutputStream.swap(OriginalOutputStream);
+ OutputStream->write(StringBuffer.data(), StringBuffer.size());
+ LLVM_DEBUG(dbgs() << "Profile originally has " << OriginalFunctionCount
+ << " functions, reduced to " << ProfileMap.size() << " in "
+ << IterationCount << " iterations\n");
+ // Silence warning on Release build.
+ (void)OriginalFunctionCount;
+ (void)IterationCount;
+ return sampleprof_error::success;
+}
+
std::error_code
SampleProfileWriter::writeFuncProfiles(const SampleProfileMap &ProfileMap) {
std::vector<NameFunctionSamples> V;
@@ -116,6 +217,12 @@ std::error_code SampleProfileWriterExtBinaryBase::addNewSection(
std::error_code
SampleProfileWriterExtBinaryBase::write(const SampleProfileMap &ProfileMap) {
+ // When calling write on a
diff erent profile map, existing states should be
+ // cleared.
+ NameTable.clear();
+ CSNameTable.clear();
+ SecHdrTable.clear();
+
if (std::error_code EC = writeHeader(ProfileMap))
return EC;
@@ -477,6 +584,7 @@ std::error_code SampleProfileWriterText::writeSample(const FunctionSamples &S) {
if (Indent == 0)
OS << ":" << S.getHeadSamples();
OS << "\n";
+ LineCount++;
SampleSorter<LineLocation, SampleRecord> SortedSamples(S.getBodySamples());
for (const auto &I : SortedSamples.get()) {
@@ -493,6 +601,7 @@ std::error_code SampleProfileWriterText::writeSample(const FunctionSamples &S) {
for (const auto &J : Sample.getSortedCallTargets())
OS << " " << J.first << ":" << J.second;
OS << "\n";
+ LineCount++;
}
SampleSorter<LineLocation, FunctionSamplesMap> SortedCallsiteSamples(
@@ -515,11 +624,13 @@ std::error_code SampleProfileWriterText::writeSample(const FunctionSamples &S) {
if (FunctionSamples::ProfileIsProbeBased) {
OS.indent(Indent + 1);
OS << "!CFGChecksum: " << S.getFunctionHash() << "\n";
+ LineCount++;
}
if (S.getContext().getAllAttributes()) {
OS.indent(Indent + 1);
OS << "!Attributes: " << S.getContext().getAllAttributes() << "\n";
+ LineCount++;
}
return sampleprof_error::success;
@@ -605,14 +716,10 @@ std::error_code SampleProfileWriterCompactBinary::writeFuncOffsetTable() {
auto &OS = *OutputStream;
// Fill the slot remembered by TableOffset with the offset of FuncOffsetTable.
- auto &OFS = static_cast<raw_fd_ostream &>(OS);
uint64_t FuncOffsetTableStart = OS.tell();
- if (OFS.seek(TableOffset) == (uint64_t)-1)
- return sampleprof_error::ostream_seek_unsupported;
- support::endian::Writer Writer(*OutputStream, support::little);
- Writer.write(FuncOffsetTableStart);
- if (OFS.seek(FuncOffsetTableStart) == (uint64_t)-1)
- return sampleprof_error::ostream_seek_unsupported;
+ support::endian::SeekableWriter Writer(static_cast<raw_pwrite_stream &>(OS),
+ support::little);
+ Writer.pwrite(FuncOffsetTableStart, TableOffset);
// Write out the table size.
encodeULEB128(FuncOffsetTable.size(), OS);
@@ -623,6 +730,7 @@ std::error_code SampleProfileWriterCompactBinary::writeFuncOffsetTable() {
return EC;
encodeULEB128(Entry.second, OS);
}
+ FuncOffsetTable.clear();
return sampleprof_error::success;
}
@@ -650,6 +758,10 @@ SampleProfileWriterBinary::writeMagicIdent(SampleProfileFormat Format) {
std::error_code
SampleProfileWriterBinary::writeHeader(const SampleProfileMap &ProfileMap) {
+ // When calling write on a
diff erent profile map, existing names should be
+ // cleared.
+ NameTable.clear();
+
writeMagicIdent(Format);
computeSummary(ProfileMap);
@@ -690,14 +802,6 @@ void SampleProfileWriterExtBinaryBase::allocSecHdrTable() {
}
std::error_code SampleProfileWriterExtBinaryBase::writeSecHdrTable() {
- auto &OFS = static_cast<raw_fd_ostream &>(*OutputStream);
- uint64_t Saved = OutputStream->tell();
-
- // Set OutputStream to the location saved in SecHdrTableOffset.
- if (OFS.seek(SecHdrTableOffset) == (uint64_t)-1)
- return sampleprof_error::ostream_seek_unsupported;
- support::endian::Writer Writer(*OutputStream, support::little);
-
assert(SecHdrTable.size() == SectionHdrLayout.size() &&
"SecHdrTable entries doesn't match SectionHdrLayout");
SmallVector<uint32_t, 16> IndexMap(SecHdrTable.size(), -1);
@@ -714,21 +818,23 @@ std::error_code SampleProfileWriterExtBinaryBase::writeSecHdrTable() {
// needs to be computed after SecLBRProfile (the order in SecHdrTable),
// but it needs to be read before SecLBRProfile (the order in
// SectionHdrLayout). So we use IndexMap above to switch the order.
+ support::endian::SeekableWriter Writer(
+ static_cast<raw_pwrite_stream &>(*OutputStream), support::little);
for (uint32_t LayoutIdx = 0; LayoutIdx < SectionHdrLayout.size();
LayoutIdx++) {
assert(IndexMap[LayoutIdx] < SecHdrTable.size() &&
"Incorrect LayoutIdx in SecHdrTable");
auto Entry = SecHdrTable[IndexMap[LayoutIdx]];
- Writer.write(static_cast<uint64_t>(Entry.Type));
- Writer.write(static_cast<uint64_t>(Entry.Flags));
- Writer.write(static_cast<uint64_t>(Entry.Offset));
- Writer.write(static_cast<uint64_t>(Entry.Size));
+ Writer.pwrite(static_cast<uint64_t>(Entry.Type),
+ SecHdrTableOffset + 4 * LayoutIdx * sizeof(uint64_t));
+ Writer.pwrite(static_cast<uint64_t>(Entry.Flags),
+ SecHdrTableOffset + (4 * LayoutIdx + 1) * sizeof(uint64_t));
+ Writer.pwrite(static_cast<uint64_t>(Entry.Offset),
+ SecHdrTableOffset + (4 * LayoutIdx + 2) * sizeof(uint64_t));
+ Writer.pwrite(static_cast<uint64_t>(Entry.Size),
+ SecHdrTableOffset + (4 * LayoutIdx + 3) * sizeof(uint64_t));
}
- // Reset OutputStream.
- if (OFS.seek(Saved) == (uint64_t)-1)
- return sampleprof_error::ostream_seek_unsupported;
-
return sampleprof_error::success;
}
diff --git a/llvm/tools/llvm-profdata/llvm-profdata.cpp b/llvm/tools/llvm-profdata/llvm-profdata.cpp
index 72ca0891c7402..8ebbac5021be5 100644
--- a/llvm/tools/llvm-profdata/llvm-profdata.cpp
+++ b/llvm/tools/llvm-profdata/llvm-profdata.cpp
@@ -970,7 +970,7 @@ mergeSampleProfile(const WeightedFileVector &Inputs, SymbolRemapper *Remapper,
bool UseMD5, bool GenPartialProfile, bool GenCSNestedProfile,
bool SampleMergeColdContext, bool SampleTrimColdContext,
bool SampleColdContextFrameDepth, FailureMode FailMode,
- bool DropProfileSymbolList) {
+ bool DropProfileSymbolList, size_t OutputSizeLimit) {
using namespace sampleprof;
SampleProfileMap ProfileMap;
SmallVector<std::unique_ptr<sampleprof::SampleProfileReader>, 5> Readers;
@@ -1064,7 +1064,10 @@ mergeSampleProfile(const WeightedFileVector &Inputs, SymbolRemapper *Remapper,
auto Buffer = getInputFileBuf(ProfileSymbolListFile);
handleExtBinaryWriter(*Writer, OutputFormat, Buffer.get(), WriterList,
CompressAllSections, UseMD5, GenPartialProfile);
- if (std::error_code EC = Writer->write(ProfileMap))
+
+ // If OutputSizeLimit is 0 (default), it is the same as write().
+ if (std::error_code EC =
+ Writer->writeWithSizeLimit(ProfileMap, OutputSizeLimit))
exitWithErrorCode(std::move(EC));
}
@@ -1207,6 +1210,11 @@ static int merge_main(int argc, const char *argv[]) {
"sample-frame-depth-for-cold-context", cl::init(1),
cl::desc("Keep the last K frames while merging cold profile. 1 means the "
"context-less base profile"));
+ cl::opt<size_t> OutputSizeLimit(
+ "output-size-limit", cl::init(0), cl::Hidden,
+ cl::desc("Trim cold functions until profile size is below specified "
+ "limit in bytes. This uses a heursitic and functions may be "
+ "excessively trimmed"));
cl::opt<bool> GenPartialProfile(
"gen-partial-profile", cl::init(false), cl::Hidden,
cl::desc("Generate a partial profile (only meaningful for -extbinary)"));
@@ -1293,7 +1301,8 @@ static int merge_main(int argc, const char *argv[]) {
WeightedInputs, Remapper.get(), OutputFilename, OutputFormat,
ProfileSymbolListFile, CompressAllSections, UseMD5, GenPartialProfile,
GenCSNestedProfile, SampleMergeColdContext, SampleTrimColdContext,
- SampleColdContextFrameDepth, FailureMode, DropProfileSymbolList);
+ SampleColdContextFrameDepth, FailureMode, DropProfileSymbolList,
+ OutputSizeLimit);
return 0;
}
diff --git a/llvm/unittests/tools/CMakeLists.txt b/llvm/unittests/tools/CMakeLists.txt
index 7ef64f1180a00..e032113fa77e6 100644
--- a/llvm/unittests/tools/CMakeLists.txt
+++ b/llvm/unittests/tools/CMakeLists.txt
@@ -7,5 +7,6 @@ endif()
add_subdirectory(
llvm-exegesis
)
+add_subdirectory(llvm-profdata)
add_subdirectory(llvm-profgen)
add_subdirectory(llvm-mca)
diff --git a/llvm/unittests/tools/llvm-profdata/CMakeLists.txt b/llvm/unittests/tools/llvm-profdata/CMakeLists.txt
new file mode 100644
index 0000000000000..dab1ac523ed00
--- /dev/null
+++ b/llvm/unittests/tools/llvm-profdata/CMakeLists.txt
@@ -0,0 +1,12 @@
+set(LLVM_LINK_COMPONENTS
+ ProfileData
+ Support
+ )
+
+add_llvm_unittest(LLVMProfdataTests
+ OutputSizeLimitTest.cpp
+ )
+
+target_link_libraries(LLVMProfdataTests PRIVATE LLVMTestingSupport)
+
+set_property(TARGET LLVMProfdataTests PROPERTY FOLDER "Tests/UnitTests/ToolTests")
diff --git a/llvm/unittests/tools/llvm-profdata/OutputSizeLimitTest.cpp b/llvm/unittests/tools/llvm-profdata/OutputSizeLimitTest.cpp
new file mode 100644
index 0000000000000..c4f8248358969
--- /dev/null
+++ b/llvm/unittests/tools/llvm-profdata/OutputSizeLimitTest.cpp
@@ -0,0 +1,182 @@
+//===- llvm/unittests/tools/llvm-profdata/OutputSizeLimitTest.cpp ---------===//
+//
+// 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 "llvm/ProfileData/SampleProfReader.h"
+#include "llvm/ProfileData/SampleProfWriter.h"
+#include "llvm/Support/FileSystem.h"
+#include "llvm/Support/VirtualFileSystem.h"
+#include "llvm/Testing/Support/Error.h"
+#include "gtest/gtest.h"
+
+using namespace llvm;
+using llvm::unittest::TempFile;
+
+std::string Input1 = R"(main:184019:0
+ 4: 534
+ 4.2: 534
+ 5: 1075
+ 5.1: 1075
+ 6: 2080
+ 7: 534
+ 9: 2064 _Z3bari:1471 _Z3fooi:631
+ 10: inline1:1000
+ 1: 1000
+ 10: inline2:2000
+ 1: 2000
+_Z3bari:20301:1437
+ 1: 1437
+_Z3fooi:7711:610
+ 1: 610)";
+
+const char EmptyProfile[18] = "\xff\xe5\xd0\xb1\xf4\xc9\x94\xa8\x53\x67";
+
+/// sys::fs and SampleProf mix Error and error_code, making an adapter class
+/// to keep code elegant.
+template <typename T> class ExpectedErrorOr : public Expected<T> {
+public:
+ ExpectedErrorOr(T &&Obj) : Expected<T>(Obj) {}
+
+ ExpectedErrorOr(std::error_code EC) : Expected<T>(errorCodeToError(EC)) {}
+
+ ExpectedErrorOr(Error &&E) : Expected<T>(std::move(E)) {}
+
+ template <typename U>
+ ExpectedErrorOr(ErrorOr<U> &&E)
+ : Expected<T>(errorCodeToError(E.getError())) {}
+
+ template <typename U>
+ ExpectedErrorOr(Expected<U> &&E) : Expected<T>(E.takeError()) {}
+};
+
+#define DEF_VAR_RETURN_IF_ERROR(Var, Value) \
+ auto Var##OrErr = Value; \
+ if (!Var##OrErr) \
+ return Var##OrErr; \
+ auto Var = std::move(Var##OrErr.get())
+
+#define VAR_RETURN_IF_ERROR(Var, Value) \
+ Var##OrErr = Value; \
+ if (!Var##OrErr) \
+ return Var##OrErr; \
+ Var = std::move(Var##OrErr.get())
+
+#define RETURN_IF_ERROR(Value) \
+ if (auto E = Value) \
+ return std::move(E)
+
+/// The main testing routine. After rewriting profiles with size limit, check
+/// the following:
+/// 1. The file size of the new profile is within the size limit.
+/// 2. The new profile is a subset of the old profile, and the content of every
+/// sample in the new profile is unchanged.
+/// Note that even though by default samples with fewest total count are dropped
+/// first, this is not a requirement. Samples can be dropped by any order.
+static ExpectedErrorOr<void *> RunTest(StringRef Input, size_t SizeLimit,
+ SampleProfileFormat Format) {
+ // Read Input profile.
+ auto FS = vfs::getRealFileSystem();
+ LLVMContext Context;
+ auto InputBuffer = MemoryBuffer::getMemBuffer(Input);
+ DEF_VAR_RETURN_IF_ERROR(
+ Reader, SampleProfileReader::create(InputBuffer, Context, *FS));
+ RETURN_IF_ERROR(Reader->read());
+ SampleProfileMap OldProfiles = Reader->getProfiles();
+
+ // Rewrite it to a temp file with size limit.
+ TempFile Temp("profile", "afdo", "", true);
+ bool isEmpty = false;
+ {
+ DEF_VAR_RETURN_IF_ERROR(Writer,
+ SampleProfileWriter::create(Temp.path(), Format));
+ std::error_code EC = Writer->writeWithSizeLimit(OldProfiles, SizeLimit);
+ // too_large means no sample could be written because SizeLimit is too
+ // small. Otherwise any other error code indicates unexpected failure.
+ if (EC == sampleprof_error::too_large)
+ isEmpty = true;
+ else if (EC)
+ return EC;
+ }
+
+ // Read the temp file to get new profiles. Use the default empty profile if
+ // temp file was not written because size limit is too small.
+ SampleProfileMap NewProfiles;
+ InputBuffer = MemoryBuffer::getMemBuffer(StringRef(EmptyProfile, 17));
+ DEF_VAR_RETURN_IF_ERROR(
+ NewReader, SampleProfileReader::create(InputBuffer, Context, *FS));
+ if (!isEmpty) {
+ VAR_RETURN_IF_ERROR(NewReader, SampleProfileReader::create(
+ Temp.path().str(), Context, *FS));
+ RETURN_IF_ERROR(NewReader->read());
+ NewProfiles = NewReader->getProfiles();
+ }
+
+ // Check temp file is actually within size limit.
+ uint64_t FileSize;
+ RETURN_IF_ERROR(sys::fs::file_size(Temp.path(), FileSize));
+ EXPECT_LE(FileSize, SizeLimit);
+
+ // For compact binary format, function names are stored as MD5, so we cannot
+ // directly match the samples of the new profile with the old profile. A
+ // simple way is to convert the old profile to compact binary format and read
+ // it back
+ if (Format == llvm::sampleprof::SPF_Compact_Binary) {
+ TempFile CompBinary("compbinary", "afdo", "", true);
+ {
+ DEF_VAR_RETURN_IF_ERROR(
+ Writer, SampleProfileWriter::create(
+ CompBinary.path(), llvm::sampleprof::SPF_Compact_Binary));
+ RETURN_IF_ERROR(Writer->write(OldProfiles));
+ }
+ VAR_RETURN_IF_ERROR(Reader, SampleProfileReader::create(
+ CompBinary.path().str(), Context, *FS));
+ RETURN_IF_ERROR(Reader->read());
+ OldProfiles = Reader->getProfiles();
+ }
+
+ // For every sample in the new profile, confirm it is in the old profile and
+ // unchanged.
+ for (auto Sample : NewProfiles) {
+ auto FindResult = OldProfiles.find(Sample.first);
+ EXPECT_NE(FindResult, OldProfiles.end());
+ if (FindResult != OldProfiles.end()) {
+ EXPECT_EQ(Sample.second.getHeadSamples(),
+ FindResult->second.getHeadSamples());
+ EXPECT_EQ(Sample.second, FindResult->second);
+ }
+ }
+ return nullptr;
+}
+
+TEST(TestOutputSizeLimit, TestOutputSizeLimitExtBinary) {
+ for (size_t OutputSizeLimit : {490, 489, 488, 475, 474, 459, 400})
+ ASSERT_THAT_EXPECTED(
+ RunTest(Input1, OutputSizeLimit, llvm::sampleprof::SPF_Ext_Binary),
+ Succeeded());
+}
+
+TEST(TestOutputSizeLimit, TestOutputSizeLimitBinary) {
+ for (size_t OutputSizeLimit : {250, 249, 248, 237, 236, 223, 200})
+ ASSERT_THAT_EXPECTED(
+ RunTest(Input1, OutputSizeLimit, llvm::sampleprof::SPF_Binary),
+ Succeeded());
+}
+
+TEST(TestOutputSizeLimit, TestOutputSizeLimitCompBinary) {
+ for (size_t OutputSizeLimit : {277, 276, 275, 264, 263, 250, 200})
+ ASSERT_THAT_EXPECTED(
+ RunTest(Input1, OutputSizeLimit, llvm::sampleprof::SPF_Compact_Binary),
+ Succeeded());
+}
+
+TEST(TestOutputSizeLimit, TestOutputSizeLimitText) {
+ for (size_t OutputSizeLimit :
+ {229, 228, 227, 213, 212, 211, 189, 188, 187, 186, 150})
+ ASSERT_THAT_EXPECTED(
+ RunTest(Input1, OutputSizeLimit, llvm::sampleprof::SPF_Text),
+ Succeeded());
+}
More information about the llvm-commits
mailing list