Details | Last modification | View Log | RSS feed
Rev | Author | Line No. | Line |
---|---|---|---|
14 | pmbaty | 1 | //===- SampleProfWriter.h - Write LLVM sample profile data ------*- C++ -*-===// |
2 | // |
||
3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
||
4 | // See https://llvm.org/LICENSE.txt for license information. |
||
5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
||
6 | // |
||
7 | //===----------------------------------------------------------------------===// |
||
8 | // |
||
9 | // This file contains definitions needed for writing sample profiles. |
||
10 | // |
||
11 | //===----------------------------------------------------------------------===// |
||
12 | #ifndef LLVM_PROFILEDATA_SAMPLEPROFWRITER_H |
||
13 | #define LLVM_PROFILEDATA_SAMPLEPROFWRITER_H |
||
14 | |||
15 | #include "llvm/ADT/MapVector.h" |
||
16 | #include "llvm/ADT/StringRef.h" |
||
17 | #include "llvm/IR/ProfileSummary.h" |
||
18 | #include "llvm/ProfileData/SampleProf.h" |
||
19 | #include "llvm/Support/ErrorOr.h" |
||
20 | #include "llvm/Support/raw_ostream.h" |
||
21 | #include <cstdint> |
||
22 | #include <memory> |
||
23 | #include <set> |
||
24 | #include <system_error> |
||
25 | |||
26 | namespace llvm { |
||
27 | namespace sampleprof { |
||
28 | |||
29 | enum SectionLayout { |
||
30 | DefaultLayout, |
||
31 | // The layout splits profile with context information from profile without |
||
32 | // context information. When Thinlto is enabled, ThinLTO postlink phase only |
||
33 | // has to load profile with context information and can skip the other part. |
||
34 | CtxSplitLayout, |
||
35 | NumOfLayout, |
||
36 | }; |
||
37 | |||
38 | /// Sample-based profile writer. Base class. |
||
39 | class SampleProfileWriter { |
||
40 | public: |
||
41 | virtual ~SampleProfileWriter() = default; |
||
42 | |||
43 | /// Write sample profiles in \p S. |
||
44 | /// |
||
45 | /// \returns status code of the file update operation. |
||
46 | virtual std::error_code writeSample(const FunctionSamples &S) = 0; |
||
47 | |||
48 | /// Write all the sample profiles in the given map of samples. |
||
49 | /// |
||
50 | /// \returns status code of the file update operation. |
||
51 | virtual std::error_code write(const SampleProfileMap &ProfileMap); |
||
52 | |||
53 | raw_ostream &getOutputStream() { return *OutputStream; } |
||
54 | |||
55 | /// Profile writer factory. |
||
56 | /// |
||
57 | /// Create a new file writer based on the value of \p Format. |
||
58 | static ErrorOr<std::unique_ptr<SampleProfileWriter>> |
||
59 | create(StringRef Filename, SampleProfileFormat Format); |
||
60 | |||
61 | /// Create a new stream writer based on the value of \p Format. |
||
62 | /// For testing. |
||
63 | static ErrorOr<std::unique_ptr<SampleProfileWriter>> |
||
64 | create(std::unique_ptr<raw_ostream> &OS, SampleProfileFormat Format); |
||
65 | |||
66 | virtual void setProfileSymbolList(ProfileSymbolList *PSL) {} |
||
67 | virtual void setToCompressAllSections() {} |
||
68 | virtual void setUseMD5() {} |
||
69 | virtual void setPartialProfile() {} |
||
70 | virtual void resetSecLayout(SectionLayout SL) {} |
||
71 | |||
72 | protected: |
||
73 | SampleProfileWriter(std::unique_ptr<raw_ostream> &OS) |
||
74 | : OutputStream(std::move(OS)) {} |
||
75 | |||
76 | /// Write a file header for the profile file. |
||
77 | virtual std::error_code writeHeader(const SampleProfileMap &ProfileMap) = 0; |
||
78 | |||
79 | // Write function profiles to the profile file. |
||
80 | virtual std::error_code writeFuncProfiles(const SampleProfileMap &ProfileMap); |
||
81 | |||
82 | /// Output stream where to emit the profile to. |
||
83 | std::unique_ptr<raw_ostream> OutputStream; |
||
84 | |||
85 | /// Profile summary. |
||
86 | std::unique_ptr<ProfileSummary> Summary; |
||
87 | |||
88 | /// Compute summary for this profile. |
||
89 | void computeSummary(const SampleProfileMap &ProfileMap); |
||
90 | |||
91 | /// Profile format. |
||
92 | SampleProfileFormat Format = SPF_None; |
||
93 | }; |
||
94 | |||
95 | /// Sample-based profile writer (text format). |
||
96 | class SampleProfileWriterText : public SampleProfileWriter { |
||
97 | public: |
||
98 | std::error_code writeSample(const FunctionSamples &S) override; |
||
99 | |||
100 | protected: |
||
101 | SampleProfileWriterText(std::unique_ptr<raw_ostream> &OS) |
||
102 | : SampleProfileWriter(OS), Indent(0) {} |
||
103 | |||
104 | std::error_code writeHeader(const SampleProfileMap &ProfileMap) override { |
||
105 | return sampleprof_error::success; |
||
106 | } |
||
107 | |||
108 | private: |
||
109 | /// Indent level to use when writing. |
||
110 | /// |
||
111 | /// This is used when printing inlined callees. |
||
112 | unsigned Indent; |
||
113 | |||
114 | friend ErrorOr<std::unique_ptr<SampleProfileWriter>> |
||
115 | SampleProfileWriter::create(std::unique_ptr<raw_ostream> &OS, |
||
116 | SampleProfileFormat Format); |
||
117 | }; |
||
118 | |||
119 | /// Sample-based profile writer (binary format). |
||
120 | class SampleProfileWriterBinary : public SampleProfileWriter { |
||
121 | public: |
||
122 | SampleProfileWriterBinary(std::unique_ptr<raw_ostream> &OS) |
||
123 | : SampleProfileWriter(OS) {} |
||
124 | |||
125 | std::error_code writeSample(const FunctionSamples &S) override; |
||
126 | |||
127 | protected: |
||
128 | virtual MapVector<StringRef, uint32_t> &getNameTable() { return NameTable; } |
||
129 | virtual std::error_code writeMagicIdent(SampleProfileFormat Format); |
||
130 | virtual std::error_code writeNameTable(); |
||
131 | std::error_code writeHeader(const SampleProfileMap &ProfileMap) override; |
||
132 | std::error_code writeSummary(); |
||
133 | virtual std::error_code writeContextIdx(const SampleContext &Context); |
||
134 | std::error_code writeNameIdx(StringRef FName); |
||
135 | std::error_code writeBody(const FunctionSamples &S); |
||
136 | inline void stablizeNameTable(MapVector<StringRef, uint32_t> &NameTable, |
||
137 | std::set<StringRef> &V); |
||
138 | |||
139 | MapVector<StringRef, uint32_t> NameTable; |
||
140 | |||
141 | void addName(StringRef FName); |
||
142 | virtual void addContext(const SampleContext &Context); |
||
143 | void addNames(const FunctionSamples &S); |
||
144 | |||
145 | private: |
||
146 | friend ErrorOr<std::unique_ptr<SampleProfileWriter>> |
||
147 | SampleProfileWriter::create(std::unique_ptr<raw_ostream> &OS, |
||
148 | SampleProfileFormat Format); |
||
149 | }; |
||
150 | |||
151 | class SampleProfileWriterRawBinary : public SampleProfileWriterBinary { |
||
152 | using SampleProfileWriterBinary::SampleProfileWriterBinary; |
||
153 | }; |
||
154 | |||
155 | const std::array<SmallVector<SecHdrTableEntry, 8>, NumOfLayout> |
||
156 | ExtBinaryHdrLayoutTable = { |
||
157 | // Note that SecFuncOffsetTable section is written after SecLBRProfile |
||
158 | // in the profile, but is put before SecLBRProfile in SectionHdrLayout. |
||
159 | // This is because sample reader follows the order in SectionHdrLayout |
||
160 | // to read each section. To read function profiles on demand, sample |
||
161 | // reader need to get the offset of each function profile first. |
||
162 | // |
||
163 | // DefaultLayout |
||
164 | SmallVector<SecHdrTableEntry, 8>({{SecProfSummary, 0, 0, 0, 0}, |
||
165 | {SecNameTable, 0, 0, 0, 0}, |
||
166 | {SecCSNameTable, 0, 0, 0, 0}, |
||
167 | {SecFuncOffsetTable, 0, 0, 0, 0}, |
||
168 | {SecLBRProfile, 0, 0, 0, 0}, |
||
169 | {SecProfileSymbolList, 0, 0, 0, 0}, |
||
170 | {SecFuncMetadata, 0, 0, 0, 0}}), |
||
171 | // CtxSplitLayout |
||
172 | SmallVector<SecHdrTableEntry, 8>({{SecProfSummary, 0, 0, 0, 0}, |
||
173 | {SecNameTable, 0, 0, 0, 0}, |
||
174 | // profile with context |
||
175 | // for next two sections |
||
176 | {SecFuncOffsetTable, 0, 0, 0, 0}, |
||
177 | {SecLBRProfile, 0, 0, 0, 0}, |
||
178 | // profile without context |
||
179 | // for next two sections |
||
180 | {SecFuncOffsetTable, 0, 0, 0, 0}, |
||
181 | {SecLBRProfile, 0, 0, 0, 0}, |
||
182 | {SecProfileSymbolList, 0, 0, 0, 0}, |
||
183 | {SecFuncMetadata, 0, 0, 0, 0}}), |
||
184 | }; |
||
185 | |||
186 | class SampleProfileWriterExtBinaryBase : public SampleProfileWriterBinary { |
||
187 | using SampleProfileWriterBinary::SampleProfileWriterBinary; |
||
188 | public: |
||
189 | std::error_code write(const SampleProfileMap &ProfileMap) override; |
||
190 | |||
191 | void setToCompressAllSections() override; |
||
192 | void setToCompressSection(SecType Type); |
||
193 | std::error_code writeSample(const FunctionSamples &S) override; |
||
194 | |||
195 | // Set to use MD5 to represent string in NameTable. |
||
196 | void setUseMD5() override { |
||
197 | UseMD5 = true; |
||
198 | addSectionFlag(SecNameTable, SecNameTableFlags::SecFlagMD5Name); |
||
199 | // MD5 will be stored as plain uint64_t instead of variable-length |
||
200 | // quantity format in NameTable section. |
||
201 | addSectionFlag(SecNameTable, SecNameTableFlags::SecFlagFixedLengthMD5); |
||
202 | } |
||
203 | |||
204 | // Set the profile to be partial. It means the profile is for |
||
205 | // common/shared code. The common profile is usually merged from |
||
206 | // profiles collected from running other targets. |
||
207 | void setPartialProfile() override { |
||
208 | addSectionFlag(SecProfSummary, SecProfSummaryFlags::SecFlagPartial); |
||
209 | } |
||
210 | |||
211 | void setProfileSymbolList(ProfileSymbolList *PSL) override { |
||
212 | ProfSymList = PSL; |
||
213 | }; |
||
214 | |||
215 | void resetSecLayout(SectionLayout SL) override { |
||
216 | verifySecLayout(SL); |
||
217 | #ifndef NDEBUG |
||
218 | // Make sure resetSecLayout is called before any flag setting. |
||
219 | for (auto &Entry : SectionHdrLayout) { |
||
220 | assert(Entry.Flags == 0 && |
||
221 | "resetSecLayout has to be called before any flag setting"); |
||
222 | } |
||
223 | #endif |
||
224 | SecLayout = SL; |
||
225 | SectionHdrLayout = ExtBinaryHdrLayoutTable[SL]; |
||
226 | } |
||
227 | |||
228 | protected: |
||
229 | uint64_t markSectionStart(SecType Type, uint32_t LayoutIdx); |
||
230 | std::error_code addNewSection(SecType Sec, uint32_t LayoutIdx, |
||
231 | uint64_t SectionStart); |
||
232 | template <class SecFlagType> |
||
233 | void addSectionFlag(SecType Type, SecFlagType Flag) { |
||
234 | for (auto &Entry : SectionHdrLayout) { |
||
235 | if (Entry.Type == Type) |
||
236 | addSecFlag(Entry, Flag); |
||
237 | } |
||
238 | } |
||
239 | template <class SecFlagType> |
||
240 | void addSectionFlag(uint32_t SectionIdx, SecFlagType Flag) { |
||
241 | addSecFlag(SectionHdrLayout[SectionIdx], Flag); |
||
242 | } |
||
243 | |||
244 | void addContext(const SampleContext &Context) override; |
||
245 | |||
246 | // placeholder for subclasses to dispatch their own section writers. |
||
247 | virtual std::error_code writeCustomSection(SecType Type) = 0; |
||
248 | // Verify the SecLayout is supported by the format. |
||
249 | virtual void verifySecLayout(SectionLayout SL) = 0; |
||
250 | |||
251 | // specify the order to write sections. |
||
252 | virtual std::error_code writeSections(const SampleProfileMap &ProfileMap) = 0; |
||
253 | |||
254 | // Dispatch section writer for each section. \p LayoutIdx is the sequence |
||
255 | // number indicating where the section is located in SectionHdrLayout. |
||
256 | virtual std::error_code writeOneSection(SecType Type, uint32_t LayoutIdx, |
||
257 | const SampleProfileMap &ProfileMap); |
||
258 | |||
259 | // Helper function to write name table. |
||
260 | std::error_code writeNameTable() override; |
||
261 | std::error_code writeContextIdx(const SampleContext &Context) override; |
||
262 | std::error_code writeCSNameIdx(const SampleContext &Context); |
||
263 | std::error_code writeCSNameTableSection(); |
||
264 | |||
265 | std::error_code writeFuncMetadata(const SampleProfileMap &Profiles); |
||
266 | std::error_code writeFuncMetadata(const FunctionSamples &Profile); |
||
267 | |||
268 | // Functions to write various kinds of sections. |
||
269 | std::error_code writeNameTableSection(const SampleProfileMap &ProfileMap); |
||
270 | std::error_code writeFuncOffsetTable(); |
||
271 | std::error_code writeProfileSymbolListSection(); |
||
272 | |||
273 | SectionLayout SecLayout = DefaultLayout; |
||
274 | // Specifiy the order of sections in section header table. Note |
||
275 | // the order of sections in SecHdrTable may be different that the |
||
276 | // order in SectionHdrLayout. sample Reader will follow the order |
||
277 | // in SectionHdrLayout to read each section. |
||
278 | SmallVector<SecHdrTableEntry, 8> SectionHdrLayout = |
||
279 | ExtBinaryHdrLayoutTable[DefaultLayout]; |
||
280 | |||
281 | // Save the start of SecLBRProfile so we can compute the offset to the |
||
282 | // start of SecLBRProfile for each Function's Profile and will keep it |
||
283 | // in FuncOffsetTable. |
||
284 | uint64_t SecLBRProfileStart = 0; |
||
285 | |||
286 | private: |
||
287 | void allocSecHdrTable(); |
||
288 | std::error_code writeSecHdrTable(); |
||
289 | std::error_code writeHeader(const SampleProfileMap &ProfileMap) override; |
||
290 | std::error_code compressAndOutput(); |
||
291 | |||
292 | // We will swap the raw_ostream held by LocalBufStream and that |
||
293 | // held by OutputStream if we try to add a section which needs |
||
294 | // compression. After the swap, all the data written to output |
||
295 | // will be temporarily buffered into the underlying raw_string_ostream |
||
296 | // originally held by LocalBufStream. After the data writing for the |
||
297 | // section is completed, compress the data in the local buffer, |
||
298 | // swap the raw_ostream back and write the compressed data to the |
||
299 | // real output. |
||
300 | std::unique_ptr<raw_ostream> LocalBufStream; |
||
301 | // The location where the output stream starts. |
||
302 | uint64_t FileStart; |
||
303 | // The location in the output stream where the SecHdrTable should be |
||
304 | // written to. |
||
305 | uint64_t SecHdrTableOffset; |
||
306 | // The table contains SecHdrTableEntry entries in order of how they are |
||
307 | // populated in the writer. It may be different from the order in |
||
308 | // SectionHdrLayout which specifies the sequence in which sections will |
||
309 | // be read. |
||
310 | std::vector<SecHdrTableEntry> SecHdrTable; |
||
311 | |||
312 | // FuncOffsetTable maps function context to its profile offset in |
||
313 | // SecLBRProfile section. It is used to load function profile on demand. |
||
314 | MapVector<SampleContext, uint64_t> FuncOffsetTable; |
||
315 | // Whether to use MD5 to represent string. |
||
316 | bool UseMD5 = false; |
||
317 | |||
318 | /// CSNameTable maps function context to its offset in SecCSNameTable section. |
||
319 | /// The offset will be used everywhere where the context is referenced. |
||
320 | MapVector<SampleContext, uint32_t> CSNameTable; |
||
321 | |||
322 | ProfileSymbolList *ProfSymList = nullptr; |
||
323 | }; |
||
324 | |||
325 | class SampleProfileWriterExtBinary : public SampleProfileWriterExtBinaryBase { |
||
326 | public: |
||
327 | SampleProfileWriterExtBinary(std::unique_ptr<raw_ostream> &OS) |
||
328 | : SampleProfileWriterExtBinaryBase(OS) {} |
||
329 | |||
330 | private: |
||
331 | std::error_code writeDefaultLayout(const SampleProfileMap &ProfileMap); |
||
332 | std::error_code writeCtxSplitLayout(const SampleProfileMap &ProfileMap); |
||
333 | |||
334 | std::error_code writeSections(const SampleProfileMap &ProfileMap) override; |
||
335 | |||
336 | std::error_code writeCustomSection(SecType Type) override { |
||
337 | return sampleprof_error::success; |
||
338 | }; |
||
339 | |||
340 | void verifySecLayout(SectionLayout SL) override { |
||
341 | assert((SL == DefaultLayout || SL == CtxSplitLayout) && |
||
342 | "Unsupported layout"); |
||
343 | } |
||
344 | }; |
||
345 | |||
346 | // CompactBinary is a compact format of binary profile which both reduces |
||
347 | // the profile size and the load time needed when compiling. It has two |
||
348 | // major difference with Binary format. |
||
349 | // 1. It represents all the strings in name table using md5 hash. |
||
350 | // 2. It saves a function offset table which maps function name index to |
||
351 | // the offset of its function profile to the start of the binary profile, |
||
352 | // so by using the function offset table, for those function profiles which |
||
353 | // will not be needed when compiling a module, the profile reader does't |
||
354 | // have to read them and it saves compile time if the profile size is huge. |
||
355 | // The layout of the compact format is shown as follows: |
||
356 | // |
||
357 | // Part1: Profile header, the same as binary format, containing magic |
||
358 | // number, version, summary, name table... |
||
359 | // Part2: Function Offset Table Offset, which saves the position of |
||
360 | // Part4. |
||
361 | // Part3: Function profile collection |
||
362 | // function1 profile start |
||
363 | // .... |
||
364 | // function2 profile start |
||
365 | // .... |
||
366 | // function3 profile start |
||
367 | // .... |
||
368 | // ...... |
||
369 | // Part4: Function Offset Table |
||
370 | // function1 name index --> function1 profile start |
||
371 | // function2 name index --> function2 profile start |
||
372 | // function3 name index --> function3 profile start |
||
373 | // |
||
374 | // We need Part2 because profile reader can use it to find out and read |
||
375 | // function offset table without reading Part3 first. |
||
376 | class SampleProfileWriterCompactBinary : public SampleProfileWriterBinary { |
||
377 | using SampleProfileWriterBinary::SampleProfileWriterBinary; |
||
378 | |||
379 | public: |
||
380 | std::error_code writeSample(const FunctionSamples &S) override; |
||
381 | std::error_code write(const SampleProfileMap &ProfileMap) override; |
||
382 | |||
383 | protected: |
||
384 | /// The table mapping from function name to the offset of its FunctionSample |
||
385 | /// towards profile start. |
||
386 | MapVector<StringRef, uint64_t> FuncOffsetTable; |
||
387 | /// The offset of the slot to be filled with the offset of FuncOffsetTable |
||
388 | /// towards profile start. |
||
389 | uint64_t TableOffset; |
||
390 | std::error_code writeNameTable() override; |
||
391 | std::error_code writeHeader(const SampleProfileMap &ProfileMap) override; |
||
392 | std::error_code writeFuncOffsetTable(); |
||
393 | }; |
||
394 | |||
395 | } // end namespace sampleprof |
||
396 | } // end namespace llvm |
||
397 | |||
398 | #endif // LLVM_PROFILEDATA_SAMPLEPROFWRITER_H |