[lld] r329117 - Inline initOffsetMap.
Rafael Espindola via llvm-commits
llvm-commits at lists.llvm.org
Tue Apr 3 14:38:18 PDT 2018
Author: rafael
Date: Tue Apr 3 14:38:18 2018
New Revision: 329117
URL: http://llvm.org/viewvc/llvm-project?rev=329117&view=rev
Log:
Inline initOffsetMap.
In the lld perf builder r328686 had a negative impact in
stalled-cycles-frontend. Somehow that stat is not showing on my
machine, but the attached patch shows an improvement on cache-misses,
which is probably a reasonable proxy.
My working theory is that given a large input the pieces vector is out
of cache by the time initOffsetMap runs.
Both finalizeContents implementation have a convenient location for
initializing the OffsetMap, so this seems the best solution.
Modified:
lld/trunk/ELF/InputSection.cpp
lld/trunk/ELF/InputSection.h
lld/trunk/ELF/SyntheticSections.cpp
Modified: lld/trunk/ELF/InputSection.cpp
URL: http://llvm.org/viewvc/llvm-project/lld/trunk/ELF/InputSection.cpp?rev=329117&r1=329116&r2=329117&view=diff
==============================================================================
--- lld/trunk/ELF/InputSection.cpp (original)
+++ lld/trunk/ELF/InputSection.cpp Tue Apr 3 14:38:18 2018
@@ -985,12 +985,6 @@ uint64_t MergeInputSection::getOffset(ui
return Piece.OutputOff + Addend;
}
-void MergeInputSection::initOffsetMap() {
- OffsetMap.reserve(Pieces.size());
- for (size_t I = 0; I < Pieces.size(); ++I)
- OffsetMap[Pieces[I].InputOff] = I;
-}
-
template InputSection::InputSection(ObjFile<ELF32LE> &, const ELF32LE::Shdr &,
StringRef);
template InputSection::InputSection(ObjFile<ELF32BE> &, const ELF32BE::Shdr &,
Modified: lld/trunk/ELF/InputSection.h
URL: http://llvm.org/viewvc/llvm-project/lld/trunk/ELF/InputSection.h?rev=329117&r1=329116&r2=329117&view=diff
==============================================================================
--- lld/trunk/ELF/InputSection.h (original)
+++ lld/trunk/ELF/InputSection.h Tue Apr 3 14:38:18 2018
@@ -236,6 +236,7 @@ public:
// Splittable sections are handled as a sequence of data
// rather than a single large blob of data.
std::vector<SectionPiece> Pieces;
+ llvm::DenseMap<uint32_t, uint32_t> OffsetMap;
// Returns I'th piece's data. This function is very hot when
// string merging is enabled, so we want to inline.
@@ -254,14 +255,11 @@ public:
}
SyntheticSection *getParent() const;
- void initOffsetMap();
private:
void splitStrings(ArrayRef<uint8_t> A, size_t Size);
void splitNonStrings(ArrayRef<uint8_t> A, size_t Size);
- llvm::DenseMap<uint32_t, uint32_t> OffsetMap;
-
llvm::DenseSet<uint32_t> LiveOffsets;
};
Modified: lld/trunk/ELF/SyntheticSections.cpp
URL: http://llvm.org/viewvc/llvm-project/lld/trunk/ELF/SyntheticSections.cpp?rev=329117&r1=329116&r2=329117&view=diff
==============================================================================
--- lld/trunk/ELF/SyntheticSections.cpp (original)
+++ lld/trunk/ELF/SyntheticSections.cpp Tue Apr 3 14:38:18 2018
@@ -2438,10 +2438,15 @@ void MergeTailSection::finalizeContents(
// finalize() fixed tail-optimized strings, so we can now get
// offsets of strings. Get an offset for each string and save it
// to a corresponding StringPiece for easy access.
- for (MergeInputSection *Sec : Sections)
- for (size_t I = 0, E = Sec->Pieces.size(); I != E; ++I)
- if (Sec->Pieces[I].Live)
- Sec->Pieces[I].OutputOff = Builder.getOffset(Sec->getData(I));
+ for (MergeInputSection *Sec : Sections) {
+ Sec->OffsetMap.reserve(Sec->Pieces.size());
+ for (size_t I = 0, E = Sec->Pieces.size(); I != E; ++I) {
+ SectionPiece &P = Sec->Pieces[I];
+ Sec->OffsetMap[P.InputOff] = I;
+ if (P.Live)
+ P.OutputOff = Builder.getOffset(Sec->getData(I));
+ }
+ }
}
void MergeNoTailSection::writeTo(uint8_t *Buf) {
@@ -2494,10 +2499,13 @@ void MergeNoTailSection::finalizeContent
// So far, section pieces have offsets from beginning of shards, but
// we want offsets from beginning of the whole section. Fix them.
parallelForEach(Sections, [&](MergeInputSection *Sec) {
- for (size_t I = 0, E = Sec->Pieces.size(); I != E; ++I)
- if (Sec->Pieces[I].Live)
- Sec->Pieces[I].OutputOff +=
- ShardOffsets[getShardId(Sec->Pieces[I].Hash)];
+ Sec->OffsetMap.reserve(Sec->Pieces.size());
+ for (size_t I = 0, E = Sec->Pieces.size(); I != E; ++I) {
+ SectionPiece &P = Sec->Pieces[I];
+ Sec->OffsetMap[P.InputOff] = I;
+ if (P.Live)
+ P.OutputOff += ShardOffsets[getShardId(P.Hash)];
+ }
});
}
@@ -2573,11 +2581,8 @@ void elf::mergeSections() {
}
(*I)->addSection(MS);
}
- for (auto *MS : MergeSections) {
+ for (auto *MS : MergeSections)
MS->finalizeContents();
- parallelForEach(MS->Sections,
- [](MergeInputSection *Sec) { Sec->initOffsetMap(); });
- }
std::vector<InputSectionBase *> &V = InputSections;
V.erase(std::remove(V.begin(), V.end(), nullptr), V.end());
More information about the llvm-commits
mailing list