mirror of
https://github.com/capstone-engine/llvm-capstone.git
synced 2024-11-23 05:40:09 +00:00
[BOLT] Fix typos (#68121)
Closes https://github.com/llvm/llvm-project/issues/63097 Before merging please make sure the change to bolt/include/bolt/Passes/StokeInfo.h is correct. bolt/include/bolt/Passes/StokeInfo.h ```diff // This Pass solves the two major problems to use the Stoke program without - // proting its code: + // probing its code: ``` I'm still not happy about the awkward wording in this comment. bolt/include/bolt/Passes/FixRelaxationPass.h ``` $ ed -s bolt/include/bolt/Passes/FixRelaxationPass.h <<<'9,12p' // This file declares the FixRelaxations class, which locates instructions with // wrong targets and fixes them. Such problems usually occures when linker // relaxes (changes) instructions, but doesn't fix relocations types properly // for them. $ ``` bolt/docs/doxygen.cfg.in bolt/include/bolt/Core/BinaryContext.h bolt/include/bolt/Core/BinaryFunction.h bolt/include/bolt/Core/BinarySection.h bolt/include/bolt/Core/DebugData.h bolt/include/bolt/Core/DynoStats.h bolt/include/bolt/Core/Exceptions.h bolt/include/bolt/Core/MCPlusBuilder.h bolt/include/bolt/Core/Relocation.h bolt/include/bolt/Passes/FixRelaxationPass.h bolt/include/bolt/Passes/InstrumentationSummary.h bolt/include/bolt/Passes/ReorderAlgorithm.h bolt/include/bolt/Passes/StackReachingUses.h bolt/include/bolt/Passes/StokeInfo.h bolt/include/bolt/Passes/TailDuplication.h bolt/include/bolt/Profile/DataAggregator.h bolt/include/bolt/Profile/DataReader.h bolt/lib/Core/BinaryContext.cpp bolt/lib/Core/BinarySection.cpp bolt/lib/Core/DebugData.cpp bolt/lib/Core/DynoStats.cpp bolt/lib/Core/Relocation.cpp bolt/lib/Passes/Instrumentation.cpp bolt/lib/Passes/JTFootprintReduction.cpp bolt/lib/Passes/ReorderData.cpp bolt/lib/Passes/RetpolineInsertion.cpp bolt/lib/Passes/ShrinkWrapping.cpp bolt/lib/Passes/TailDuplication.cpp bolt/lib/Rewrite/BoltDiff.cpp bolt/lib/Rewrite/DWARFRewriter.cpp bolt/lib/Rewrite/RewriteInstance.cpp bolt/lib/Utils/CommandLineOpts.cpp bolt/runtime/instr.cpp bolt/test/AArch64/got-ld64-relaxation.test bolt/test/AArch64/unmarked-data.test bolt/test/X86/Inputs/dwarf5-cu-no-debug-addr-helper.s bolt/test/X86/Inputs/linenumber.cpp bolt/test/X86/double-jump.test bolt/test/X86/dwarf5-call-pc-function-null-check.test bolt/test/X86/dwarf5-split-dwarf4-monolithic.test bolt/test/X86/dynrelocs.s bolt/test/X86/fallthrough-to-noop.test bolt/test/X86/tail-duplication-cache.s bolt/test/runtime/X86/instrumentation-ind-calls.s
This commit is contained in:
spaette
GitHub
parent
30e4b09d49
commit
1a2f83366b
@ -1954,7 +1954,7 @@ PREDEFINED =
|
||||
EXPAND_AS_DEFINED =
|
||||
|
||||
# If the SKIP_FUNCTION_MACROS tag is set to YES then doxygen's preprocessor will
|
||||
# remove all refrences to function-like macros that are alone on a line, have an
|
||||
# remove all references to function-like macros that are alone on a line, have an
|
||||
# all uppercase name, and do not end with a semicolon. Such function macros are
|
||||
# typically used for boiler-plate code, and will confuse the parser if not
|
||||
# removed.
|
||||
|
||||
@ -950,7 +950,7 @@ public:
|
||||
bool registerFragment(BinaryFunction &TargetFunction,
|
||||
BinaryFunction &Function) const;
|
||||
|
||||
/// Add unterprocedural reference for \p Function to \p Address
|
||||
/// Add interprocedural reference for \p Function to \p Address
|
||||
void addInterproceduralReference(BinaryFunction *Function, uint64_t Address) {
|
||||
InterproceduralReferences.push_back({Function, Address});
|
||||
}
|
||||
|
||||
@ -2172,7 +2172,7 @@ public:
|
||||
/// is corrupted. If it is unable to fix it, it returns false.
|
||||
bool finalizeCFIState();
|
||||
|
||||
/// Return true if this function needs an address-transaltion table after
|
||||
/// Return true if this function needs an address-translation table after
|
||||
/// its code emission.
|
||||
bool requiresAddressTranslation() const;
|
||||
|
||||
|
||||
@ -127,7 +127,7 @@ class BinarySection {
|
||||
return *ContentsOrErr;
|
||||
}
|
||||
|
||||
/// Get the set of relocations refering to data in this section that
|
||||
/// Get the set of relocations referring to data in this section that
|
||||
/// has been reordered. The relocation offsets will be modified to
|
||||
/// reflect the new data locations.
|
||||
RelocationSetType reorderRelocations(bool Inplace) const;
|
||||
|
||||
@ -64,7 +64,7 @@ std::optional<AttrInfo> findAttributeInfo(const DWARFDie DIE,
|
||||
dwarf::Attribute Attr);
|
||||
|
||||
// DWARF5 Header in order of encoding.
|
||||
// Types represent encodnig sizes.
|
||||
// Types represent encoding sizes.
|
||||
using UnitLengthType = uint32_t;
|
||||
using VersionType = uint16_t;
|
||||
using AddressSizeType = uint8_t;
|
||||
@ -372,13 +372,13 @@ protected:
|
||||
}
|
||||
|
||||
/// Updates AddressToIndex Map.
|
||||
/// Follows the same symantics as unordered map [].
|
||||
/// Follows the same semantics as unordered map [].
|
||||
void updateAddressToIndex(uint64_t Address, uint32_t Index) {
|
||||
AddressToIndex[Address] = Index;
|
||||
}
|
||||
|
||||
/// Updates IndexToAddress Map.
|
||||
/// Follows the same symantics as unordered map [].
|
||||
/// Follows the same semantics as unordered map [].
|
||||
void updateIndexToAddrss(uint64_t Address, uint32_t Index) {
|
||||
IndexToAddress[Index] = Address;
|
||||
}
|
||||
@ -474,7 +474,7 @@ public:
|
||||
}
|
||||
|
||||
/// Adds string to .debug_str.
|
||||
/// On first invokation it initializes internal data stractures.
|
||||
/// On first invocation it initializes internal data structures.
|
||||
uint32_t addString(StringRef Str);
|
||||
|
||||
/// Returns False if no strings were added to .debug_str.
|
||||
@ -485,7 +485,7 @@ private:
|
||||
std::mutex WriterMutex;
|
||||
/// Initializes Buffer and Stream.
|
||||
void initialize();
|
||||
/// Creats internal data stractures.
|
||||
/// Creates internal data structures.
|
||||
void create();
|
||||
std::unique_ptr<DebugStrBufferVector> StrBuffer;
|
||||
std::unique_ptr<raw_svector_ostream> StrStream;
|
||||
@ -585,7 +585,7 @@ public:
|
||||
void finalize(DIEBuilder &DIEBldr, DIE &Die) override;
|
||||
|
||||
/// Returns CU ID.
|
||||
/// For Skelton CU it is a CU Offset.
|
||||
/// For Skeleton CU it is a CU Offset.
|
||||
/// For DWO CU it is a DWO ID.
|
||||
uint64_t getCUID() const {
|
||||
return CU.isDWOUnit() ? *CU.getDWOId() : CU.getOffset();
|
||||
|
||||
@ -121,7 +121,7 @@ public:
|
||||
/// Maps instruction opcodes to:
|
||||
/// 1. Accumulated executed instruction counts.
|
||||
/// 2. a multimap that records highest execution counts, function names,
|
||||
/// and BB offsets where intructions of these opcodes occur.
|
||||
/// and BB offsets where instructions of these opcodes occur.
|
||||
using MaxOpcodeHistogramTy =
|
||||
std::multimap<uint64_t, std::pair<StringRef, uint32_t>>;
|
||||
using OpcodeHistogramTy =
|
||||
|
||||
@ -77,7 +77,7 @@ public:
|
||||
/// void PatcherCallback(uint64_t Value, uint64_t Offset, uint64_t Type);
|
||||
///
|
||||
/// where Value is a value of the reference, Offset - is an offset into the
|
||||
/// frame data at which the reference occured, and Type is a DWARF encoding
|
||||
/// frame data at which the reference occurred, and Type is a DWARF encoding
|
||||
/// type of the reference.
|
||||
static Error parse(DWARFDataExtractor Data, uint64_t EHFrameAddress,
|
||||
PatcherCallbackTy PatcherCallback);
|
||||
|
||||
@ -1082,7 +1082,7 @@ public:
|
||||
|
||||
/// Replace the compound memory operand of Inst with an immediate operand.
|
||||
/// The value of the immediate operand is computed by reading the \p
|
||||
/// ConstantData array starting from \p offset and assuming little-endianess.
|
||||
/// ConstantData array starting from \p offset and assuming little-endianness.
|
||||
/// Return true on success. The given instruction is modified in place.
|
||||
virtual bool replaceMemOperandWithImm(MCInst &Inst, StringRef ConstantData,
|
||||
uint64_t Offset) const {
|
||||
|
||||
@ -61,7 +61,7 @@ struct Relocation {
|
||||
|
||||
/// Handle special cases when relocation should not be processed by BOLT or
|
||||
/// change relocation \p Type to proper one before continuing if \p Contents
|
||||
/// and \P Type mismatch occured.
|
||||
/// and \P Type mismatch occurred.
|
||||
static bool skipRelocationProcess(uint64_t &Type, uint64_t Contents);
|
||||
|
||||
// Adjust value depending on relocation type (make it PC relative or not)
|
||||
|
||||
@ -7,7 +7,7 @@
|
||||
//===----------------------------------------------------------------------===//
|
||||
//
|
||||
// This file declares the FixRelaxations class, which locates instructions with
|
||||
// wrong targets and fixes them. Such problems usually occures when linker
|
||||
// wrong targets and fixes them. Such problems usually occurs when linker
|
||||
// relaxes (changes) instructions, but doesn't fix relocations types properly
|
||||
// for them.
|
||||
//
|
||||
|
||||
@ -35,7 +35,7 @@ class BinaryFunction;
|
||||
// counters. It also provides function names and offsets used when writing the
|
||||
// fdata file.
|
||||
|
||||
// Location information -- analoguous to the concept of the same name in fdata
|
||||
// Location information -- analogous to the concept of the same name in fdata
|
||||
// writing/reading. The difference is that the name is stored as an index to a
|
||||
// string table written separately.
|
||||
struct LocDescription {
|
||||
|
||||
@ -42,7 +42,7 @@ public:
|
||||
virtual void clusterBasicBlocks(BinaryFunction &BF,
|
||||
bool ComputeEdges = false) = 0;
|
||||
|
||||
/// Compute for each cluster its averagae execution frequency, that is
|
||||
/// Compute for each cluster its average execution frequency, that is
|
||||
/// the sum of average frequencies of its blocks (execution count / # instrs).
|
||||
/// The average frequencies are stored in the AvgFreq vector, index by the
|
||||
/// cluster indices in the Clusters vector.
|
||||
@ -128,7 +128,7 @@ private:
|
||||
// A -> B with execution count 500,
|
||||
// A -> C with execution count 100, and
|
||||
// D -> B with execution count 150
|
||||
// wher B, C are the only successors of A and A, D are thr only predessecors
|
||||
// where B, C are the only successors of A and A, D are the only predecessors
|
||||
// of B. Then if we choose to layout edge A -> B as a fallthrough, the win in
|
||||
// branches would be 500 - 100 - 150 = 250. That is the weight of edge A->B.
|
||||
int64_t calculateWeight(const EdgeTy &E, const BinaryFunction &BF) const;
|
||||
@ -144,8 +144,8 @@ public:
|
||||
void reset() override;
|
||||
};
|
||||
|
||||
/// Objects of this class implement various basic block reordering alogrithms.
|
||||
/// Most of these algorithms depend on a clustering alogrithm.
|
||||
/// Objects of this class implement various basic block reordering algorithms.
|
||||
/// Most of these algorithms depend on a clustering algorithm.
|
||||
/// Here we have 3 conflicting goals as to how to layout clusters. If we want
|
||||
/// to minimize jump offsets, we should put clusters with heavy inter-cluster
|
||||
/// dependence as close as possible. If we want to maximize the probability
|
||||
|
||||
@ -41,7 +41,7 @@ public:
|
||||
/// Answer whether the stack position written by the store represented in
|
||||
/// \p StoreFIE is loaded from or consumed in any way. The set of all
|
||||
/// relevant expressions reaching this store should be in \p Candidates.
|
||||
/// If \p IncludelocalAccesses is false, we only consider wheter there is
|
||||
/// If \p IncludelocalAccesses is false, we only consider whether there is
|
||||
/// a callee that consumes this stack position.
|
||||
bool isStoreUsed(const FrameIndexEntry &StoreFIE, ExprIterator Candidates,
|
||||
bool IncludeLocalAccesses = true) const;
|
||||
|
||||
@ -9,7 +9,7 @@
|
||||
// Pass to get information for functions for the Stoke Optimization
|
||||
// To use the Stoke optimization technique to optimize the HHVM.
|
||||
// This Pass solves the two major problems to use the Stoke program without
|
||||
// proting its code:
|
||||
// probing its code:
|
||||
//
|
||||
// 1. Stoke works on function level, but it is only limited to relative
|
||||
// small functions which are loop-free, call-free, exception-free, etc.
|
||||
|
||||
@ -81,7 +81,7 @@ class TailDuplication : public BinaryFunctionPass {
|
||||
/// Returns true if Reg is used by Inst
|
||||
bool regIsUsed(const MCInst &Inst, unsigned Reg, BinaryContext &BC) const;
|
||||
|
||||
/// Returns true if Reg is overwritten before its used by StartBB's sucessors
|
||||
/// Returns true if Reg is overwritten before its used by StartBB's successors
|
||||
bool isOverwrittenBeforeUsed(BinaryBasicBlock &StartBB, unsigned Reg) const;
|
||||
|
||||
/// Constant and Copy Propagate for the block formed by OriginalBB and
|
||||
|
||||
@ -301,7 +301,7 @@ private:
|
||||
ErrorOr<AggregatedLBREntry> parseAggregatedLBREntry();
|
||||
|
||||
/// Parse either buildid:offset or just offset, representing a location in the
|
||||
/// binary. Used exclusevely for pre-aggregated LBR samples.
|
||||
/// binary. Used exclusively for pre-aggregated LBR samples.
|
||||
ErrorOr<Location> parseLocationOrOffset();
|
||||
|
||||
/// Check if a field separator is the next char to parse and, if yes, consume
|
||||
|
||||
@ -350,9 +350,9 @@ protected:
|
||||
///
|
||||
/// The examples records branches from local symbol func (from t2.c), offset
|
||||
/// 11, to global symbol globalfunc, offset 1d, with 1775 branches, no
|
||||
/// mispreds. Of these branches, 1002 were preceeded by a sequence of
|
||||
/// mispreds. Of these branches, 1002 were preceded by a sequence of
|
||||
/// branches from func, offset 18 to offset 20 and then from offset 31 to
|
||||
/// offset d. The rest 773 branches were preceeded by a different sequence
|
||||
/// offset d. The rest 773 branches were preceded by a different sequence
|
||||
/// of branches, from func, offset 18 to offset 60 and then from offset 71 to
|
||||
/// offset d.
|
||||
std::error_code parse();
|
||||
|
||||
@ -1402,7 +1402,7 @@ void BinaryContext::foldFunction(BinaryFunction &ChildBF,
|
||||
}
|
||||
|
||||
void BinaryContext::fixBinaryDataHoles() {
|
||||
assert(validateObjectNesting() && "object nesting inconsitency detected");
|
||||
assert(validateObjectNesting() && "object nesting inconsistency detected");
|
||||
|
||||
for (BinarySection &Section : allocatableSections()) {
|
||||
std::vector<std::pair<uint64_t, uint64_t>> Holes;
|
||||
@ -1451,7 +1451,7 @@ void BinaryContext::fixBinaryDataHoles() {
|
||||
}
|
||||
}
|
||||
|
||||
assert(validateObjectNesting() && "object nesting inconsitency detected");
|
||||
assert(validateObjectNesting() && "object nesting inconsistency detected");
|
||||
assert(validateHoles() && "top level hole detected in object map");
|
||||
}
|
||||
|
||||
@ -1796,7 +1796,7 @@ void BinaryContext::printCFI(raw_ostream &OS, const MCCFIInstruction &Inst) {
|
||||
MarkerSymType BinaryContext::getMarkerType(const SymbolRef &Symbol) const {
|
||||
// For aarch64 and riscv, the ABI defines mapping symbols so we identify data
|
||||
// in the code section (see IHI0056B). $x identifies a symbol starting code or
|
||||
// the end of a data chunk inside code, $d indentifies start of data.
|
||||
// the end of a data chunk inside code, $d identifies start of data.
|
||||
if ((!isAArch64() && !isRISCV()) || ELFSymbolRef(Symbol).getSize())
|
||||
return MarkerSymType::NONE;
|
||||
|
||||
|
||||
@ -227,7 +227,7 @@ void BinarySection::print(raw_ostream &OS) const {
|
||||
BinarySection::RelocationSetType
|
||||
BinarySection::reorderRelocations(bool Inplace) const {
|
||||
assert(PendingRelocations.empty() &&
|
||||
"reodering pending relocations not supported");
|
||||
"reordering pending relocations not supported");
|
||||
RelocationSetType NewRelocations;
|
||||
for (const Relocation &Rel : relocations()) {
|
||||
uint64_t RelAddr = Rel.Offset + getAddress();
|
||||
|
||||
@ -188,7 +188,7 @@ uint64_t DebugRangeListsSectionWriter::addRanges(
|
||||
}
|
||||
|
||||
struct LocListsRangelistsHeader {
|
||||
UnitLengthType UnitLength; // Size of loclist entris section, not including
|
||||
UnitLengthType UnitLength; // Size of loclist entries section, not including
|
||||
// size of header.
|
||||
VersionType Version;
|
||||
AddressSizeType AddressSize;
|
||||
@ -449,8 +449,8 @@ static void updateAddressBase(DIEBuilder &DIEBlder, DebugAddrWriter &AddrWriter,
|
||||
}
|
||||
|
||||
void DebugAddrWriter::update(DIEBuilder &DIEBlder, DWARFUnit &CU) {
|
||||
// Handling the case wehre debug information is a mix of Debug fission and
|
||||
// monolitic.
|
||||
// Handling the case where debug information is a mix of Debug fission and
|
||||
// monolithic.
|
||||
if (!CU.getDWOId())
|
||||
return;
|
||||
const uint64_t CUID = getCUID(CU);
|
||||
|
||||
@ -136,7 +136,7 @@ void DynoStats::operator+=(const DynoStats &Other) {
|
||||
if (I == OpcodeHistogram.end()) {
|
||||
OpcodeHistogram.emplace(Stat);
|
||||
} else {
|
||||
// Merge Other Historgrams, log only the opts::PrintDynoOpcodeStat'th
|
||||
// Merge other histograms, log only the opts::PrintDynoOpcodeStat'th
|
||||
// maximum counts.
|
||||
I->second.first += Stat.second.first;
|
||||
auto &MMap = I->second.second;
|
||||
|
||||
@ -295,7 +295,7 @@ static bool skipRelocationProcessAArch64(uint64_t &Type, uint64_t Contents) {
|
||||
// changed TLS access model (e.g. changed global dynamic model
|
||||
// to initial exec), thus changing the instructions. The static
|
||||
// relocations might be invalid at this point and we might no
|
||||
// need to proccess these relocations anymore.
|
||||
// need to process these relocations anymore.
|
||||
// More information could be found by searching
|
||||
// elfNN_aarch64_tls_relax in bfd
|
||||
switch (Type) {
|
||||
@ -503,7 +503,7 @@ static uint64_t extractValueAArch64(uint64_t Type, uint64_t Contents,
|
||||
case ELF::R_AARCH64_MOVW_UABS_G1:
|
||||
case ELF::R_AARCH64_MOVW_UABS_G0_NC:
|
||||
case ELF::R_AARCH64_MOVW_UABS_G0:
|
||||
// The shift goest in bits 22:21 of MOV* instructions
|
||||
// The shift goes in bits 22:21 of MOV* instructions
|
||||
uint8_t Shift = (Contents >> 21) & 0x3;
|
||||
// Immediate goes in bits 20:5
|
||||
Contents = (Contents >> 5) & 0xffff;
|
||||
|
||||
@ -34,7 +34,7 @@ cl::opt<std::string> InstrumentationFilename(
|
||||
|
||||
cl::opt<std::string> InstrumentationBinpath(
|
||||
"instrumentation-binpath",
|
||||
cl::desc("path to instumented binary in case if /proc/self/map_files "
|
||||
cl::desc("path to instrumented binary in case if /proc/self/map_files "
|
||||
"is not accessible due to access restriction issues"),
|
||||
cl::Optional, cl::cat(BoltInstrCategory));
|
||||
|
||||
|
||||
@ -232,7 +232,7 @@ void JTFootprintReduction::optimizeFunction(BinaryFunction &Function,
|
||||
continue;
|
||||
}
|
||||
|
||||
llvm_unreachable("Should either optimize PIC or NonPIC successfuly");
|
||||
llvm_unreachable("Should either optimize PIC or NonPIC successfully");
|
||||
}
|
||||
|
||||
if (!Modified.count(&Function))
|
||||
|
||||
@ -11,7 +11,7 @@
|
||||
//===----------------------------------------------------------------------===//
|
||||
|
||||
// TODO:
|
||||
// - make sure writeable data isn't put on same cache line unless temporally
|
||||
// - make sure writable data isn't put on same cache line unless temporally
|
||||
// local
|
||||
// - estimate temporal locality by looking at CFG?
|
||||
|
||||
@ -328,7 +328,7 @@ ReorderData::sortedByCount(BinaryContext &BC,
|
||||
|
||||
// TODO
|
||||
// add option for cache-line alignment (or just use cache-line when section
|
||||
// is writeable)?
|
||||
// is writable)?
|
||||
void ReorderData::setSectionOrder(BinaryContext &BC,
|
||||
BinarySection &OutputSection,
|
||||
DataOrder::iterator Begin,
|
||||
|
||||
@ -52,9 +52,9 @@ cl::opt<RetpolineInsertion::AvailabilityOptions> R11Availability(
|
||||
cl::values(clEnumValN(RetpolineInsertion::AvailabilityOptions::NEVER,
|
||||
"never", "r11 not available"),
|
||||
clEnumValN(RetpolineInsertion::AvailabilityOptions::ALWAYS,
|
||||
"always", "r11 avaialable before calls and jumps"),
|
||||
"always", "r11 available before calls and jumps"),
|
||||
clEnumValN(RetpolineInsertion::AvailabilityOptions::ABI, "abi",
|
||||
"r11 avaialable before calls but not before jumps")),
|
||||
"r11 available before calls but not before jumps")),
|
||||
cl::ZeroOrMore, cl::cat(BoltCategory));
|
||||
|
||||
} // namespace opts
|
||||
|
||||
@ -903,7 +903,7 @@ bool ShrinkWrapping::isBestSavePosCold(unsigned CSR, MCInst *&BestPosSave,
|
||||
return true;
|
||||
}
|
||||
|
||||
/// Auxiliar function used to create basic blocks for critical edges and update
|
||||
/// Auxiliary function used to create basic blocks for critical edges and update
|
||||
/// the dominance frontier with these new locations
|
||||
void ShrinkWrapping::splitFrontierCritEdges(
|
||||
BinaryFunction *Func, SmallVector<ProgramPoint, 4> &Frontier,
|
||||
|
||||
@ -236,7 +236,7 @@ void TailDuplication::constantAndCopyPropagate(
|
||||
break;
|
||||
}
|
||||
|
||||
// If the register was replaced everwhere and it was overwritten in either
|
||||
// If the register was replaced everywhere and it was overwritten in either
|
||||
// one of the iterated through blocks or one of the successor blocks, delete
|
||||
// the original move instruction
|
||||
if (ReplacedEverywhere &&
|
||||
@ -276,7 +276,7 @@ TailDuplication::moderateDuplicate(BinaryBasicBlock &BB,
|
||||
// The block must be hot
|
||||
if (BB.getKnownExecutionCount() == 0)
|
||||
return BlocksToDuplicate;
|
||||
// and its sucessor is not already in the same cache line
|
||||
// and its successor is not already in the same cache line
|
||||
if (isInCacheLine(BB, Tail))
|
||||
return BlocksToDuplicate;
|
||||
// and its size do not exceed the maximum allowed size
|
||||
@ -299,7 +299,7 @@ TailDuplication::aggressiveDuplicate(BinaryBasicBlock &BB,
|
||||
// The block must be hot
|
||||
if (BB.getKnownExecutionCount() == 0)
|
||||
return BlocksToDuplicate;
|
||||
// and its sucessor is not already in the same cache line
|
||||
// and its successor is not already in the same cache line
|
||||
if (isInCacheLine(BB, Tail))
|
||||
return BlocksToDuplicate;
|
||||
|
||||
|
||||
@ -672,7 +672,7 @@ public:
|
||||
|
||||
};
|
||||
|
||||
} // end nampespace bolt
|
||||
} // end namespace bolt
|
||||
} // end namespace llvm
|
||||
|
||||
void RewriteInstance::compare(RewriteInstance &RI2) {
|
||||
|
||||
@ -1396,7 +1396,7 @@ void DWARFRewriter::updateLineTableOffsets(const MCAsmLayout &Layout) {
|
||||
std::optional<DWARFFormValue> StmtList =
|
||||
Unit->getUnitDIE().find(dwarf::DW_AT_stmt_list);
|
||||
std::optional<uint64_t> Offset = dwarf::toSectionOffset(StmtList);
|
||||
assert(Offset && "Was not able to retreive value of DW_AT_stmt_list.");
|
||||
assert(Offset && "Was not able to retrieve value of DW_AT_stmt_list.");
|
||||
return *Offset;
|
||||
};
|
||||
|
||||
@ -1725,7 +1725,7 @@ std::optional<StringRef> updateDebugData(
|
||||
case DWARFSectionKind::DW_SECT_LOCLISTS: {
|
||||
OutputBuffer = LocWriter.getBuffer();
|
||||
// Creating explicit StringRef here, otherwise
|
||||
// with impicit conversion it will take null byte as end of
|
||||
// with implicit conversion it will take null byte as end of
|
||||
// string.
|
||||
return StringRef(reinterpret_cast<const char *>(OutputBuffer->data()),
|
||||
OutputBuffer->size());
|
||||
@ -1994,7 +1994,7 @@ void DWARFRewriter::writeDWOFiles(
|
||||
"No RangeListsWriter for DWO ID.");
|
||||
RangeListssWriter = RangeListsWritersByCU[DWOId].get();
|
||||
|
||||
// Handling .debug_rnglists.dwo seperatly. The original .o/.dwo might not
|
||||
// Handling .debug_rnglists.dwo separately. The original .o/.dwo might not
|
||||
// have .debug_rnglists so won't be part of the loop below.
|
||||
if (!RangeListssWriter->empty()) {
|
||||
std::unique_ptr<DebugBufferVector> OutputData;
|
||||
|
||||
@ -4562,7 +4562,7 @@ void RewriteInstance::updateELFSymbolTable(
|
||||
OutputAddress >= Function->getCodeSection()->getOutputAddress() &&
|
||||
OutputAddress < (Function->getCodeSection()->getOutputAddress() +
|
||||
Function->getCodeSection()->getOutputSize()) &&
|
||||
"Cannot locate fragment containg secondary entrypoint");
|
||||
"Cannot locate fragment containing secondary entrypoint");
|
||||
FF = Function->getLayout().fragment_begin();
|
||||
}
|
||||
|
||||
|
||||
@ -160,7 +160,7 @@ cl::opt<ProfileFormatKind> ProfileFormat(
|
||||
"format to dump profile output in aggregation mode, default is fdata"),
|
||||
cl::init(PF_Fdata),
|
||||
cl::values(clEnumValN(PF_Fdata, "fdata", "offset-based plaintext format"),
|
||||
clEnumValN(PF_YAML, "yaml", "dense YAML reprensentation")),
|
||||
clEnumValN(PF_YAML, "yaml", "dense YAML representation")),
|
||||
cl::ZeroOrMore, cl::Hidden, cl::cat(BoltCategory));
|
||||
|
||||
cl::opt<bool> SplitEH("split-eh", cl::desc("split C++ exception handling code"),
|
||||
|
||||
@ -323,7 +323,7 @@ public:
|
||||
|
||||
/// Basic member accessing interface. Here we pass the allocator explicitly to
|
||||
/// avoid storing a pointer to it as part of this table (remember there is one
|
||||
/// hash for each indirect call site, so we wan't to minimize our footprint).
|
||||
/// hash for each indirect call site, so we want to minimize our footprint).
|
||||
MapEntry &get(uint64_t Key, BumpPtrAllocator &Alloc) {
|
||||
if (!__bolt_instr_conservative) {
|
||||
TryLock L(M);
|
||||
@ -1281,7 +1281,7 @@ void Graph::computeEdgeFrequencies(const uint64_t *Counters,
|
||||
|
||||
/// Write to \p FD all of the edge profiles for function \p FuncDesc. Uses
|
||||
/// \p Alloc to allocate helper dynamic structures used to compute profile for
|
||||
/// edges that we do not explictly instrument.
|
||||
/// edges that we do not explicitly instrument.
|
||||
const uint8_t *writeFunctionProfile(int FD, ProfileWriterContext &Ctx,
|
||||
const uint8_t *FuncDesc,
|
||||
BumpPtrAllocator &Alloc) {
|
||||
@ -1514,7 +1514,7 @@ extern "C" void __bolt_instr_clear_counters() {
|
||||
/// * Program execution ended, finalization methods are running and BOLT
|
||||
/// hooked into FINI from your binary dynamic section;
|
||||
/// * You used the sleep timer option and during initialization we forked
|
||||
/// a separete process that will call this function periodically;
|
||||
/// a separate process that will call this function periodically;
|
||||
/// * BOLT prints this function address so you can attach a debugger and
|
||||
/// call this function directly to get your profile written to disk
|
||||
/// on demand.
|
||||
|
||||
@ -1,5 +1,5 @@
|
||||
// This test checks that ADR+LDR instruction sequence relaxed by the linker
|
||||
// to the ADR+ADD sequence is properly reconized and handled by bolt
|
||||
// to the ADR+ADD sequence is properly recognized and handled by bolt
|
||||
|
||||
// RUN: yaml2obj %p/Inputs/got-ld64-relaxation.yaml &> %t.exe
|
||||
// RUN: llvm-bolt %t.exe -o /dev/null --print-fix-relaxations \
|
||||
|
||||
@ -9,7 +9,7 @@
|
||||
|
||||
// YAML is based in the following assembly:
|
||||
|
||||
.equ val, 0x070b0f03 // we use constant that is not a valid instruction so that it can't be silently dissassembled
|
||||
.equ val, 0x070b0f03 // we use constant that is not a valid instruction so that it can't be silently disassembled
|
||||
.text
|
||||
|
||||
first:
|
||||
|
||||
@ -4,7 +4,7 @@
|
||||
# int i;
|
||||
# } Foo;
|
||||
|
||||
# Manually modifed to remove DW_AT_addr_base and DW_AT_location.
|
||||
# Manually modified to remove DW_AT_addr_base and DW_AT_location.
|
||||
|
||||
.text
|
||||
.file "helper.cpp"
|
||||
@ -104,7 +104,7 @@
|
||||
.long .Lstr_offsets_base0 # DW_AT_str_offsets_base
|
||||
.long .Lline_table_start0 # DW_AT_stmt_list
|
||||
.byte 2 # DW_AT_comp_dir
|
||||
#.long .Laddr_table_base0 # DW_AT_addr_base mannually removed
|
||||
#.long .Laddr_table_base0 # DW_AT_addr_base manually removed
|
||||
.byte 2 # Abbrev [2] 0x1e:0xc DW_TAG_variable
|
||||
.byte 3 # DW_AT_name
|
||||
.long 35 # DW_AT_type
|
||||
|
||||
@ -1,6 +1,6 @@
|
||||
int f() {
|
||||
// This will be removed by BOLT but they make sure we have some extra space
|
||||
// to insert branches and don't run out of space when rewritting the function.
|
||||
// to insert branches and don't run out of space when rewriting the function.
|
||||
asm("nop");
|
||||
asm("nop");
|
||||
asm("nop");
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
# Test the double jump removqal peephole.
|
||||
# Test the double jump removal peephole.
|
||||
|
||||
# This test has commands that rely on shell capabilities that won't execute
|
||||
# correctly on Windows e.g. subshell execution
|
||||
|
||||
@ -8,7 +8,7 @@
|
||||
# RUN: llvm-dwarfdump --show-form --verbose --debug-info %t.exe.bolt >> %t.txt
|
||||
# RUN: cat %t.txt | FileCheck --check-prefix=CHECK %s
|
||||
|
||||
# Test checks we correclty handle nullptr returned by getBinaryFunctionContainingAddress for DW_AT_call_pc.
|
||||
# Test checks we correctly handle nullptr returned by getBinaryFunctionContainingAddress for DW_AT_call_pc.
|
||||
# This happens when address is not contained in any function.
|
||||
|
||||
# CHECK: DW_AT_call_pc [DW_FORM_addrx]
|
||||
|
||||
@ -244,7 +244,7 @@
|
||||
# POSTCHECK-DWO-MAIN: DW_TAG_inlined_subroutine [12]
|
||||
# POSTCHECK-DWO-MAIN-NEXT: DW_AT_abstract_origin [DW_FORM_ref4]
|
||||
# POSTCHECK-DWO-MAIN-NEXT: DW_AT_ranges [DW_FORM_rnglistx] (indexed (0x2) rangelist = 0x00000020
|
||||
# Encded as a pair. So it's offset from base address.
|
||||
# Encoded as a pair. So it's offset from base address.
|
||||
# POSTCHECK-DWO-MAIN-NEXT: [0x0000000000000003, 0x0000000000000007)
|
||||
# POSTCHECK-DWO-MAIN-NEXT: [0x0000000000000013, 0x0000000000000019))
|
||||
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# but it can happen. In these cases we cannot corrupt the
|
||||
# output binary by writing out dynamic relocs incorrectly. The linker
|
||||
# avoids emitting relocs against read-only sections but we override
|
||||
# this behvior with the -z notext flag. During runtime, these pages
|
||||
# this behavior with the -z notext flag. During runtime, these pages
|
||||
# are mapped with write permission and then changed to read-only after
|
||||
# the dynamic linker finishes processing the dynamic relocs.
|
||||
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
# Check that profile data for the fall-through jump is not ingored when there is
|
||||
# Check that profile data for the fall-through jump is not ignored when there is
|
||||
# a conditional jump followed by a no-op.
|
||||
|
||||
RUN: llvm-mc -filetype=obj -triple x86_64-unknown-unknown \
|
||||
|
||||
@ -11,7 +11,7 @@
|
||||
# RUN: --print-finalized --tail-duplication=cache -o %t.out2 \
|
||||
# RUN: | FileCheck --check-prefix="CHECK2" %s
|
||||
|
||||
# A test where the tail is duplicated to eliminate an uncoditional jump
|
||||
# A test where the tail is duplicated to eliminate an unconditional jump
|
||||
# FDATA: 1 main #.BB0_br# 1 main #.BB4# 0 100
|
||||
# FDATA: 1 main #.BB0_br# 1 main #.BB1# 0 100
|
||||
# FDATA: 1 main #.BB1_br# 1 main #.BB3# 0 50
|
||||
|
||||
@ -1,5 +1,5 @@
|
||||
# This reproduces a bug with instrumentation when trying to count calls
|
||||
# when the target address is computed with a referece to the stack pointer.
|
||||
# when the target address is computed with a reference to the stack pointer.
|
||||
# Our instrumentation code uses the stack to save registers to be
|
||||
# transparent with the instrumented code, but we end up updating the stack
|
||||
# pointer while doing so, which affects this target address calculation.
|
||||
|
||||
Loading…
Reference in New Issue
Block a user