[Profile] Add binary profile correlation for code coverage. (#69493)

## Motivation
Since we don't need the metadata sections at runtime, we can somehow
offload them from memory at runtime. Initially, I explored [debug info
correlation](https://discourse.llvm.org/t/instrprofiling-lightweight-instrumentation/59113),
which is used for PGO with value profiling disabled. However, it
currently only works with DWARF and it's be hard to add such artificial
debug info for every function in to CodeView which is used on Windows.
So, offloading profile metadata sections at runtime seems to be a
platform independent option.

## Design
The idea is to use new section names for profile name and data sections
and mark them as metadata sections. Under this mode, the new sections
are non-SHF_ALLOC in ELF. So, they are not loaded into memory at runtime
and can be stripped away as a post-linking step. After the process
exits, the generated raw profiles will contains only headers + counters.
llvm-profdata can be used correlate raw profiles with the unstripped
binary to generate indexed profile.

## Data
For chromium base_unittests with code coverage on linux, the binary size
overhead due to instrumentation reduced from 64M to 38.8M (39.4%) and
the raw profile files size reduce from 128M to 68M (46.9%)
```
$ bloaty out/cov/base_unittests.stripped -- out/no-cov/base_unittests.stripped
    FILE SIZE        VM SIZE
 --------------  --------------
  +121% +30.4Mi  +121% +30.4Mi    .text
  [NEW] +14.6Mi  [NEW] +14.6Mi    __llvm_prf_data
  [NEW] +10.6Mi  [NEW] +10.6Mi    __llvm_prf_names
  [NEW] +5.86Mi  [NEW] +5.86Mi    __llvm_prf_cnts
   +95% +1.75Mi   +95% +1.75Mi    .eh_frame
  +108%  +400Ki  +108%  +400Ki    .eh_frame_hdr
  +9.5%  +211Ki  +9.5%  +211Ki    .rela.dyn
  +9.2% +95.0Ki  +9.2% +95.0Ki    .data.rel.ro
  +5.0% +87.3Ki  +5.0% +87.3Ki    .rodata
  [ = ]       0   +13% +47.0Ki    .bss
   +40% +1.78Ki   +40% +1.78Ki    .got
   +12% +1.49Ki   +12% +1.49Ki    .gcc_except_table
  [ = ]       0   +65% +1.23Ki    .relro_padding
   +62% +1.20Ki  [ = ]       0    [Unmapped]
   +13%    +448   +19%    +448    .init_array
  +8.8%    +192  [ = ]       0    [ELF Section Headers]
  +0.0%    +136  +0.0%     +80    [7 Others]
  +0.1%     +96  +0.1%     +96    .dynsym
  +1.2%     +96  +1.2%     +96    .rela.plt
  +1.5%     +80  +1.2%     +64    .plt
  [ = ]       0 -99.2% -3.68Ki    [LOAD #5 [RW]]
  +195% +64.0Mi  +194% +64.0Mi    TOTAL
$ bloaty out/cov-cor/base_unittests.stripped -- out/no-cov/base_unittests.stripped
    FILE SIZE        VM SIZE
 --------------  --------------
  +121% +30.4Mi  +121% +30.4Mi    .text
  [NEW] +5.86Mi  [NEW] +5.86Mi    __llvm_prf_cnts
   +95% +1.75Mi   +95% +1.75Mi    .eh_frame
  +108%  +400Ki  +108%  +400Ki    .eh_frame_hdr
  +9.5%  +211Ki  +9.5%  +211Ki    .rela.dyn
  +9.2% +95.0Ki  +9.2% +95.0Ki    .data.rel.ro
  +5.0% +87.3Ki  +5.0% +87.3Ki    .rodata
  [ = ]       0   +13% +47.0Ki    .bss
   +40% +1.78Ki   +40% +1.78Ki    .got
   +12% +1.49Ki   +12% +1.49Ki    .gcc_except_table
   +13%    +448   +19%    +448    .init_array
  +0.1%     +96  +0.1%     +96    .dynsym
  +1.2%     +96  +1.2%     +96    .rela.plt
  +1.2%     +64  +1.2%     +64    .plt
  +2.9%     +64  [ = ]       0    [ELF Section Headers]
  +0.0%     +40  +0.0%     +40    .data
  +1.2%     +32  +1.2%     +32    .got.plt
  +0.0%     +24  +0.0%      +8    [5 Others]
  [ = ]       0 -22.9%    -872    [LOAD #5 [RW]]
 -74.5% -1.44Ki  [ = ]       0    [Unmapped]
  [ = ]       0 -76.5% -1.45Ki    .relro_padding
  +118% +38.8Mi  +117% +38.8Mi    TOTAL
```

A few things to note:
1. llvm-profdata doesn't support filter raw profiles by binary id yet,
so when a raw profile doesn't belongs to the binary being digested by
llvm-profdata, merging will fail. Once this is implemented,
llvm-profdata should be able to only merge raw profiles with the same
binary id as the binary and discard the rest (with mismatched/missing
binary id). The workflow I have in mind is to have scripts invoke
llvm-profdata to get all binary ids for all raw profiles, and
selectively choose the raw pnrofiles with matching binary id and the
binary to llvm-profdata for merging.
2. Note: In COFF, currently they are still loaded into memory but not
used. I didn't do it in this patch because I noticed that `.lcovmap` and
`.lcovfunc` are loaded into memory. A separate patch will address it.
3. This should works with PGO when value profiling is disabled as debug
info correlation currently doing, though I haven't tested this yet.

1 files changed, 39 insertions, 26 deletions

diff --git a/llvm/lib/ProfileData/Coverage/CoverageMappingReader.cpp b/llvm/lib/ProfileData/Coverage/CoverageMappingReader.cpp
index 56a7ab2..ac8e6b5 100644
--- a/llvm/lib/ProfileData/Coverage/CoverageMappingReader.cpp
+++ b/llvm/lib/ProfileData/Coverage/CoverageMappingReader.cpp
@@ -493,9 +493,13 @@ Error InstrProfSymtab::create(SectionRef &Section) {
 
   // If this is a linked PE/COFF file, then we have to skip over the null byte
   // that is allocated in the .lprfn$A section in the LLVM profiling runtime.
+  // If the name section is .lprfcovnames, it doesn't have the null byte at the
+  // beginning.
   const ObjectFile *Obj = Section.getObject();
   if (isa<COFFObjectFile>(Obj) && !Obj->isRelocatableObject())
-    Data = Data.drop_front(1);
+    if (Expected<StringRef> NameOrErr = Section.getName())
+      if (*NameOrErr != getInstrProfSectionName(IPSK_covname, Triple::COFF))
+        Data = Data.drop_front(1);
 
   return Error::success();
 }
@@ -1024,10 +1028,13 @@ loadTestingFormat(StringRef Data, StringRef CompilationDir) {
       BytesInAddress, Endian, CompilationDir);
 }
 
-/// Find all sections that match \p Name. There may be more than one if comdats
-/// are in use, e.g. for the __llvm_covfun section on ELF.
-static Expected<std::vector<SectionRef>> lookupSections(ObjectFile &OF,
-                                                        StringRef Name) {
+/// Find all sections that match \p IPSK name. There may be more than one if
+/// comdats are in use, e.g. for the __llvm_covfun section on ELF.
+static Expected<std::vector<SectionRef>>
+lookupSections(ObjectFile &OF, InstrProfSectKind IPSK) {
+  auto ObjFormat = OF.getTripleObjectFormat();
+  auto Name =
+      getInstrProfSectionName(IPSK, ObjFormat, /*AddSegmentInfo=*/false);
   // On COFF, the object file section name may end in "$M". This tells the
   // linker to sort these sections between "$A" and "$Z". The linker removes the
   // dollar and everything after it in the final binary. Do the same to match.
@@ -1042,8 +1049,13 @@ static Expected<std::vector<SectionRef>> lookupSections(ObjectFile &OF,
     Expected<StringRef> NameOrErr = Section.getName();
     if (!NameOrErr)
       return NameOrErr.takeError();
-    if (stripSuffix(*NameOrErr) == Name)
+    if (stripSuffix(*NameOrErr) == Name) {
+      // COFF profile name section contains two null bytes indicating the
+      // start/end of the section. If its size is 2 bytes, it's empty.
+      if (IsCOFF && IPSK == IPSK_name && Section.getSize() == 2)
+        continue;
       Sections.push_back(Section);
+    }
   }
   if (Sections.empty())
     return make_error<CoverageMapError>(coveragemap_error::no_data_found);
@@ -1079,15 +1091,27 @@ loadBinaryFormat(std::unique_ptr<Binary> Bin, StringRef Arch,
       OF->isLittleEndian() ? llvm::endianness::little : llvm::endianness::big;
 
   // Look for the sections that we are interested in.
-  auto ObjFormat = OF->getTripleObjectFormat();
-  auto NamesSection =
-      lookupSections(*OF, getInstrProfSectionName(IPSK_name, ObjFormat,
-                                                 /*AddSegmentInfo=*/false));
-  if (auto E = NamesSection.takeError())
+  InstrProfSymtab ProfileNames;
+  std::vector<SectionRef> NamesSectionRefs;
+  // If IPSK_name is not found, fallback to search for IPK_covname, which is
+  // used when binary correlation is enabled.
+  auto NamesSection = lookupSections(*OF, IPSK_name);
+  if (auto E = NamesSection.takeError()) {
+    consumeError(std::move(E));
+    NamesSection = lookupSections(*OF, IPSK_covname);
+    if (auto E = NamesSection.takeError())
+      return std::move(E);
+  }
+  NamesSectionRefs = *NamesSection;
+
+  if (NamesSectionRefs.size() != 1)
+    return make_error<CoverageMapError>(
+        coveragemap_error::malformed,
+        "the size of coverage mapping section is not one");
+  if (Error E = ProfileNames.create(NamesSectionRefs.back()))
     return std::move(E);
-  auto CoverageSection =
-      lookupSections(*OF, getInstrProfSectionName(IPSK_covmap, ObjFormat,
-                                                  /*AddSegmentInfo=*/false));
+
+  auto CoverageSection = lookupSections(*OF, IPSK_covmap);
   if (auto E = CoverageSection.takeError())
     return std::move(E);
   std::vector<SectionRef> CoverageSectionRefs = *CoverageSection;
@@ -1099,19 +1123,8 @@ loadBinaryFormat(std::unique_ptr<Binary> Bin, StringRef Arch,
     return CoverageMappingOrErr.takeError();
   StringRef CoverageMapping = CoverageMappingOrErr.get();
 
-  InstrProfSymtab ProfileNames;
-  std::vector<SectionRef> NamesSectionRefs = *NamesSection;
-  if (NamesSectionRefs.size() != 1)
-    return make_error<CoverageMapError>(
-        coveragemap_error::malformed,
-        "the size of coverage mapping section is not one");
-  if (Error E = ProfileNames.create(NamesSectionRefs.back()))
-    return std::move(E);
-
   // Look for the coverage records section (Version4 only).
-  auto CoverageRecordsSections =
-      lookupSections(*OF, getInstrProfSectionName(IPSK_covfun, ObjFormat,
-                                                  /*AddSegmentInfo=*/false));
+  auto CoverageRecordsSections = lookupSections(*OF, IPSK_covfun);
 
   BinaryCoverageReader::FuncRecordsStorage FuncRecords;
   if (auto E = CoverageRecordsSections.takeError()) {