Files
llvm/lld/ELF/LTO.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

418 lines
16 KiB
C++
Raw Normal View History

//===- LTO.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 "LTO.h"
#include "Config.h"
#include "InputFiles.h"
#include "SymbolTable.h"
#include "Symbols.h"
#include "lld/Common/Args.h"
#include "lld/Common/CommonLinkerContext.h"
#include "lld/Common/ErrorHandler.h"
#include "lld/Common/Filesystem.h"
2022-02-07 21:53:34 -08:00
#include "lld/Common/Strings.h"
#include "lld/Common/TargetOptionsCommandFlags.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/ADT/Twine.h"
#include "llvm/BinaryFormat/ELF.h"
#include "llvm/Bitcode/BitcodeWriter.h"
#include "llvm/LTO/Config.h"
#include "llvm/LTO/LTO.h"
#include "llvm/Support/Caching.h"
#include "llvm/Support/CodeGen.h"
#include "llvm/Support/Error.h"
#include "llvm/Support/FileSystem.h"
#include "llvm/Support/MemoryBuffer.h"
#include "llvm/Support/Path.h"
#include <algorithm>
#include <cstddef>
#include <memory>
#include <string>
#include <system_error>
#include <vector>
using namespace llvm;
using namespace llvm::object;
using namespace llvm::ELF;
using namespace lld;
using namespace lld::elf;
static std::string getThinLTOOutputFile(StringRef modulePath) {
return lto::getThinLTOOutputFile(modulePath, config->thinLTOPrefixReplaceOld,
config->thinLTOPrefixReplaceNew);
}
static lto::Config createConfig() {
lto::Config c;
// LLD supports the new relocations and address-significance tables.
c.Options = initTargetOptionsFromCodeGenFlags();
c.Options.EmitAddrsig = true;
for (StringRef C : config->mllvmOpts)
c.MllvmArgs.emplace_back(C.str());
// Always emit a section per function/datum with LTO.
c.Options.FunctionSections = true;
c.Options.DataSections = true;
[SHT_LLVM_BB_ADDR_MAP] Allow basic-block-sections and labels be used together by decoupling the handling of the two features. (#74128) Today `-split-machine-functions` and `-fbasic-block-sections={all,list}` cannot be combined with `-basic-block-sections=labels` (the labels option will be ignored). The inconsistency comes from the way basic block address map -- the underlying mechanism for basic block labels -- encodes basic block addresses (https://lists.llvm.org/pipermail/llvm-dev/2020-July/143512.html). Specifically, basic block offsets are computed relative to the function begin symbol. This relies on functions being contiguous which is not the case for MFS and basic block section binaries. This means Propeller cannot use binary profiles collected from these binaries, which limits the applicability of Propeller for iterative optimization. To make the `SHT_LLVM_BB_ADDR_MAP` feature work with basic block section binaries, we propose modifying the encoding of this section as follows. First let us review the current encoding which emits the address of each function and its number of basic blocks, followed by basic block entries for each basic block. | | | |--|--| | Address of the function | Function Address | | Number of basic blocks in this function | NumBlocks | | BB entry 1 | BB entry 2 | ... | BB entry #NumBlocks To make this work for basic block sections, we treat each basic block section similar to a function, except that basic block sections of the same function must be encapsulated in the same structure so we can map all of them to their single function. We modify the encoding to first emit the number of basic block sections (BB ranges) in the function. Then we emit the address map of each basic block section section as before: the base address of the section, its number of blocks, and BB entries for its basic block. The first section in the BB address map is always the function entry section. | | | |--|--| | Number of sections for this function | NumBBRanges | | Section 1 begin address | BaseAddress[1] | | Number of basic blocks in section 1 | NumBlocks[1] | | BB entries for Section 1 |..................| | Section #NumBBRanges begin address | BaseAddress[NumBBRanges] | | Number of basic blocks in section #NumBBRanges | NumBlocks[NumBBRanges] | | BB entries for Section #NumBBRanges The encoding of basic block entries remains as before with the minor change that each basic block offset is now computed relative to the begin symbol of its containing BB section. This patch adds a new boolean codegen option `-basic-block-address-map`. Correspondingly, the front-end flag `-fbasic-block-address-map` and LLD flag `--lto-basic-block-address-map` are introduced. Analogously, we add a new TargetOption field `BBAddrMap`. This means BB address maps are either generated for all functions in the compiling unit, or for none (depending on `TargetOptions::BBAddrMap`). This patch keeps the functionality of the old `-fbasic-block-sections=labels` option but does not remove it. A subsequent patch will remove the obsolete option. We refactor the `BasicBlockSections` pass by separating the BB address map and BB sections handing to their own functions (named `handleBBAddrMap` and `handleBBSections`). `handleBBSections` renumbers basic blocks and places them in their assigned sections. `handleBBAddrMap` is invoked after `handleBBSections` (if requested) and only renumbers the blocks. - New tests added: - Two tests basic-block-address-map-with-basic-block-sections.ll and basic-block-address-map-with-mfs.ll to exercise the combination of `-basic-block-address-map` with `-basic-block-sections=list` and '-split-machine-functions`. - A driver sanity test for the `-fbasic-block-address-map` option (basic-block-address-map.c). - An LLD test for testing the `--lto-basic-block-address-map` option. This reuses the LLVM IR from `lld/test/ELF/lto/basic-block-sections.ll`. - Renamed and modified the two existing codegen tests for basic block address map (`basic-block-sections-labels-functions-sections.ll` and `basic-block-sections-labels.ll`) - Removed `SHT_LLVM_BB_ADDR_MAP_V0` tests. Full deprecation of `SHT_LLVM_BB_ADDR_MAP_V0` and `SHT_LLVM_BB_ADDR_MAP` version less than 2 will happen in a separate PR in a few months.
2024-02-01 17:50:46 -08:00
c.Options.BBAddrMap = config->ltoBBAddrMap;
LLD Support for Basic Block Sections This is part of the Propeller framework to do post link code layout optimizations. Please see the RFC here: https://groups.google.com/forum/#!msg/llvm-dev/ef3mKzAdJ7U/1shV64BYBAAJ and the detailed RFC doc here: https://github.com/google/llvm-propeller/blob/plo-dev/Propeller_RFC.pdf This patch adds lld support for basic block sections and performs relaxations after the basic blocks have been reordered. After the linker has reordered the basic block sections according to the desired sequence, it runs a relaxation pass to optimize jump instructions. Currently, the compiler emits the long form of all jump instructions. AMD64 ISA supports variants of jump instructions with one byte offset or a four byte offset. The compiler generates jump instructions with R_X86_64 32-bit PC relative relocations. We would like to use a new relocation type for these jump instructions as it makes it easy and accurate while relaxing these instructions. The relaxation pass does two things: First, it deletes all explicit fall-through direct jump instructions between adjacent basic blocks. This is done by discarding the tail of the basic block section. Second, If there are consecutive jump instructions, it checks if the first conditional jump can be inverted to convert the second into a fall through and delete the second. The jump instructions are relaxed by using jump instruction mods, something like relocations. These are used to modify the opcode of the jump instruction. Jump instruction mods contain three values, instruction offset, jump type and size. While writing this jump instruction out to the final binary, the linker uses the jump instruction mod to determine the opcode and the size of the modified jump instruction. These mods are required because the input object files are memory-mapped without write permissions and directly modifying the object files requires copying these sections. Copying a large number of basic block sections significantly bloats memory. Differential Revision: https://reviews.llvm.org/D68065
2020-04-07 06:48:18 -07:00
// Check if basic block sections must be used.
// Allowed values for --lto-basic-block-sections are "all", "labels",
LLD Support for Basic Block Sections This is part of the Propeller framework to do post link code layout optimizations. Please see the RFC here: https://groups.google.com/forum/#!msg/llvm-dev/ef3mKzAdJ7U/1shV64BYBAAJ and the detailed RFC doc here: https://github.com/google/llvm-propeller/blob/plo-dev/Propeller_RFC.pdf This patch adds lld support for basic block sections and performs relaxations after the basic blocks have been reordered. After the linker has reordered the basic block sections according to the desired sequence, it runs a relaxation pass to optimize jump instructions. Currently, the compiler emits the long form of all jump instructions. AMD64 ISA supports variants of jump instructions with one byte offset or a four byte offset. The compiler generates jump instructions with R_X86_64 32-bit PC relative relocations. We would like to use a new relocation type for these jump instructions as it makes it easy and accurate while relaxing these instructions. The relaxation pass does two things: First, it deletes all explicit fall-through direct jump instructions between adjacent basic blocks. This is done by discarding the tail of the basic block section. Second, If there are consecutive jump instructions, it checks if the first conditional jump can be inverted to convert the second into a fall through and delete the second. The jump instructions are relaxed by using jump instruction mods, something like relocations. These are used to modify the opcode of the jump instruction. Jump instruction mods contain three values, instruction offset, jump type and size. While writing this jump instruction out to the final binary, the linker uses the jump instruction mod to determine the opcode and the size of the modified jump instruction. These mods are required because the input object files are memory-mapped without write permissions and directly modifying the object files requires copying these sections. Copying a large number of basic block sections significantly bloats memory. Differential Revision: https://reviews.llvm.org/D68065
2020-04-07 06:48:18 -07:00
// "<file name specifying basic block ids>", or none. This is the equivalent
// of -fbasic-block-sections= flag in clang.
LLD Support for Basic Block Sections This is part of the Propeller framework to do post link code layout optimizations. Please see the RFC here: https://groups.google.com/forum/#!msg/llvm-dev/ef3mKzAdJ7U/1shV64BYBAAJ and the detailed RFC doc here: https://github.com/google/llvm-propeller/blob/plo-dev/Propeller_RFC.pdf This patch adds lld support for basic block sections and performs relaxations after the basic blocks have been reordered. After the linker has reordered the basic block sections according to the desired sequence, it runs a relaxation pass to optimize jump instructions. Currently, the compiler emits the long form of all jump instructions. AMD64 ISA supports variants of jump instructions with one byte offset or a four byte offset. The compiler generates jump instructions with R_X86_64 32-bit PC relative relocations. We would like to use a new relocation type for these jump instructions as it makes it easy and accurate while relaxing these instructions. The relaxation pass does two things: First, it deletes all explicit fall-through direct jump instructions between adjacent basic blocks. This is done by discarding the tail of the basic block section. Second, If there are consecutive jump instructions, it checks if the first conditional jump can be inverted to convert the second into a fall through and delete the second. The jump instructions are relaxed by using jump instruction mods, something like relocations. These are used to modify the opcode of the jump instruction. Jump instruction mods contain three values, instruction offset, jump type and size. While writing this jump instruction out to the final binary, the linker uses the jump instruction mod to determine the opcode and the size of the modified jump instruction. These mods are required because the input object files are memory-mapped without write permissions and directly modifying the object files requires copying these sections. Copying a large number of basic block sections significantly bloats memory. Differential Revision: https://reviews.llvm.org/D68065
2020-04-07 06:48:18 -07:00
if (!config->ltoBasicBlockSections.empty()) {
if (config->ltoBasicBlockSections == "all") {
c.Options.BBSections = BasicBlockSection::All;
} else if (config->ltoBasicBlockSections == "labels") {
c.Options.BBSections = BasicBlockSection::Labels;
} else if (config->ltoBasicBlockSections == "none") {
c.Options.BBSections = BasicBlockSection::None;
} else {
ErrorOr<std::unique_ptr<MemoryBuffer>> MBOrErr =
MemoryBuffer::getFile(config->ltoBasicBlockSections.str());
if (!MBOrErr) {
error("cannot open " + config->ltoBasicBlockSections + ":" +
MBOrErr.getError().message());
} else {
c.Options.BBSectionsFuncListBuf = std::move(*MBOrErr);
}
c.Options.BBSections = BasicBlockSection::List;
}
}
c.Options.UniqueBasicBlockSectionNames =
config->ltoUniqueBasicBlockSectionNames;
LLD Support for Basic Block Sections This is part of the Propeller framework to do post link code layout optimizations. Please see the RFC here: https://groups.google.com/forum/#!msg/llvm-dev/ef3mKzAdJ7U/1shV64BYBAAJ and the detailed RFC doc here: https://github.com/google/llvm-propeller/blob/plo-dev/Propeller_RFC.pdf This patch adds lld support for basic block sections and performs relaxations after the basic blocks have been reordered. After the linker has reordered the basic block sections according to the desired sequence, it runs a relaxation pass to optimize jump instructions. Currently, the compiler emits the long form of all jump instructions. AMD64 ISA supports variants of jump instructions with one byte offset or a four byte offset. The compiler generates jump instructions with R_X86_64 32-bit PC relative relocations. We would like to use a new relocation type for these jump instructions as it makes it easy and accurate while relaxing these instructions. The relaxation pass does two things: First, it deletes all explicit fall-through direct jump instructions between adjacent basic blocks. This is done by discarding the tail of the basic block section. Second, If there are consecutive jump instructions, it checks if the first conditional jump can be inverted to convert the second into a fall through and delete the second. The jump instructions are relaxed by using jump instruction mods, something like relocations. These are used to modify the opcode of the jump instruction. Jump instruction mods contain three values, instruction offset, jump type and size. While writing this jump instruction out to the final binary, the linker uses the jump instruction mod to determine the opcode and the size of the modified jump instruction. These mods are required because the input object files are memory-mapped without write permissions and directly modifying the object files requires copying these sections. Copying a large number of basic block sections significantly bloats memory. Differential Revision: https://reviews.llvm.org/D68065
2020-04-07 06:48:18 -07:00
if (auto relocModel = getRelocModelFromCMModel())
c.RelocModel = *relocModel;
else if (config->relocatable)
c.RelocModel = std::nullopt;
else if (config->isPic)
c.RelocModel = Reloc::PIC_;
else
c.RelocModel = Reloc::Static;
c.CodeModel = getCodeModelFromCMModel();
c.DisableVerify = config->disableVerify;
c.DiagHandler = diagnosticHandler;
c.OptLevel = config->ltoo;
c.CPU = getCPUStr();
c.MAttrs = getMAttrs();
c.CGOptLevel = config->ltoCgo;
c.PTO.LoopVectorization = c.OptLevel > 1;
c.PTO.SLPVectorization = c.OptLevel > 1;
// Set up a custom pipeline if we've been asked to.
c.OptPipeline = std::string(config->ltoNewPmPasses);
c.AAPipeline = std::string(config->ltoAAPipeline);
// Set up optimization remarks if we've been asked to.
c.RemarksFilename = std::string(config->optRemarksFilename);
c.RemarksPasses = std::string(config->optRemarksPasses);
c.RemarksWithHotness = config->optRemarksWithHotness;
c.RemarksHotnessThreshold = config->optRemarksHotnessThreshold;
c.RemarksFormat = std::string(config->optRemarksFormat);
[Coding style change] Rename variables so that they start with a lowercase letter This patch is mechanically generated by clang-llvm-rename tool that I wrote using Clang Refactoring Engine just for creating this patch. You can see the source code of the tool at https://reviews.llvm.org/D64123. There's no manual post-processing; you can generate the same patch by re-running the tool against lld's code base. Here is the main discussion thread to change the LLVM coding style: https://lists.llvm.org/pipermail/llvm-dev/2019-February/130083.html In the discussion thread, I proposed we use lld as a testbed for variable naming scheme change, and this patch does that. I chose to rename variables so that they are in camelCase, just because that is a minimal change to make variables to start with a lowercase letter. Note to downstream patch maintainers: if you are maintaining a downstream lld repo, just rebasing ahead of this commit would cause massive merge conflicts because this patch essentially changes every line in the lld subdirectory. But there's a remedy. clang-llvm-rename tool is a batch tool, so you can rename variables in your downstream repo with the tool. Given that, here is how to rebase your repo to a commit after the mass renaming: 1. rebase to the commit just before the mass variable renaming, 2. apply the tool to your downstream repo to mass-rename variables locally, and 3. rebase again to the head. Most changes made by the tool should be identical for a downstream repo and for the head, so at the step 3, almost all changes should be merged and disappear. I'd expect that there would be some lines that you need to merge by hand, but that shouldn't be too many. Differential Revision: https://reviews.llvm.org/D64121 llvm-svn: 365595
2019-07-10 05:00:37 +00:00
// Set up output file to emit statistics.
c.StatsFile = std::string(config->optStatsFilename);
c.SampleProfile = std::string(config->ltoSampleProfile);
for (StringRef pluginFn : config->passPlugins)
c.PassPlugins.push_back(std::string(pluginFn));
c.DebugPassManager = config->ltoDebugPassManager;
c.DwoDir = std::string(config->dwoDir);
[Coding style change] Rename variables so that they start with a lowercase letter This patch is mechanically generated by clang-llvm-rename tool that I wrote using Clang Refactoring Engine just for creating this patch. You can see the source code of the tool at https://reviews.llvm.org/D64123. There's no manual post-processing; you can generate the same patch by re-running the tool against lld's code base. Here is the main discussion thread to change the LLVM coding style: https://lists.llvm.org/pipermail/llvm-dev/2019-February/130083.html In the discussion thread, I proposed we use lld as a testbed for variable naming scheme change, and this patch does that. I chose to rename variables so that they are in camelCase, just because that is a minimal change to make variables to start with a lowercase letter. Note to downstream patch maintainers: if you are maintaining a downstream lld repo, just rebasing ahead of this commit would cause massive merge conflicts because this patch essentially changes every line in the lld subdirectory. But there's a remedy. clang-llvm-rename tool is a batch tool, so you can rename variables in your downstream repo with the tool. Given that, here is how to rebase your repo to a commit after the mass renaming: 1. rebase to the commit just before the mass variable renaming, 2. apply the tool to your downstream repo to mass-rename variables locally, and 3. rebase again to the head. Most changes made by the tool should be identical for a downstream repo and for the head, so at the step 3, almost all changes should be merged and disappear. I'd expect that there would be some lines that you need to merge by hand, but that shouldn't be too many. Differential Revision: https://reviews.llvm.org/D64121 llvm-svn: 365595
2019-07-10 05:00:37 +00:00
c.HasWholeProgramVisibility = config->ltoWholeProgramVisibility;
[WPD][LLD] Add option to validate RTTI is enabled on all native types and prevent devirtualization on types with native RTTI Discussion about this approach: https://discourse.llvm.org/t/rfc-safer-whole-program-class-hierarchy-analysis/65144/18 When enabling WPD in an environment where native binaries are present, types we want to optimize can be derived from inside these native files and devirtualizing them can lead to correctness issues. RTTI can be used as a way to determine all such types in native files and exclude them from WPD providing a safe checked way to enable WPD. The approach is: 1. In the linker, identify if RTTI is available for all native types. If not, under `--lto-validate-all-vtables-have-type-infos` `--lto-whole-program-visibility` is automatically disabled. This is done by examining all .symtab symbols in object files and .dynsym symbols in DSOs for vtable (_ZTV) and typeinfo (_ZTI) symbols and ensuring there's always a match for every vtable symbol. 2. During thinlink, if `--lto-validate-all-vtables-have-type-infos` is set and RTTI is available for all native types, identify all typename (_ZTS) symbols via their corresponding typeinfo (_ZTI) symbols that are used natively or outside of our summary and exclude them from WPD. Testing: ninja check-all large Meta service that uses boost, glog and libstdc++.so runs successfully with WPD via --lto-whole-program-visibility. Previously, native types in boost caused incorrect devirtualization that led to crashes. Reviewed By: MaskRay, tejohnson Differential Revision: https://reviews.llvm.org/D155659
2023-07-13 19:02:52 -07:00
c.ValidateAllVtablesHaveTypeInfos =
config->ltoValidateAllVtablesHaveTypeInfos;
c.AllVtablesHaveTypeInfos = ctx.ltoAllVtablesHaveTypeInfos;
c.AlwaysEmitRegularLTOObj = !config->ltoObjPath.empty();
for (const llvm::StringRef &name : config->thinLTOModulesToCompile)
c.ThinLTOModulesToCompile.emplace_back(name);
c.TimeTraceEnabled = config->timeTraceEnabled;
c.TimeTraceGranularity = config->timeTraceGranularity;
c.CSIRProfile = std::string(config->ltoCSProfileFile);
c.RunCSIRInstr = config->ltoCSProfileGenerate;
c.PGOWarnMismatch = config->ltoPGOWarnMismatch;
[Coding style change] Rename variables so that they start with a lowercase letter This patch is mechanically generated by clang-llvm-rename tool that I wrote using Clang Refactoring Engine just for creating this patch. You can see the source code of the tool at https://reviews.llvm.org/D64123. There's no manual post-processing; you can generate the same patch by re-running the tool against lld's code base. Here is the main discussion thread to change the LLVM coding style: https://lists.llvm.org/pipermail/llvm-dev/2019-February/130083.html In the discussion thread, I proposed we use lld as a testbed for variable naming scheme change, and this patch does that. I chose to rename variables so that they are in camelCase, just because that is a minimal change to make variables to start with a lowercase letter. Note to downstream patch maintainers: if you are maintaining a downstream lld repo, just rebasing ahead of this commit would cause massive merge conflicts because this patch essentially changes every line in the lld subdirectory. But there's a remedy. clang-llvm-rename tool is a batch tool, so you can rename variables in your downstream repo with the tool. Given that, here is how to rebase your repo to a commit after the mass renaming: 1. rebase to the commit just before the mass variable renaming, 2. apply the tool to your downstream repo to mass-rename variables locally, and 3. rebase again to the head. Most changes made by the tool should be identical for a downstream repo and for the head, so at the step 3, almost all changes should be merged and disappear. I'd expect that there would be some lines that you need to merge by hand, but that shouldn't be too many. Differential Revision: https://reviews.llvm.org/D64121 llvm-svn: 365595
2019-07-10 05:00:37 +00:00
if (config->emitLLVM) {
c.PostInternalizeModuleHook = [](size_t task, const Module &m) {
if (std::unique_ptr<raw_fd_ostream> os =
openLTOOutputFile(config->outputFile))
WriteBitcodeToFile(m, *os, false);
return false;
};
}
if (config->ltoEmitAsm) {
c.CGFileType = CodeGenFileType::AssemblyFile;
c.Options.MCOptions.AsmVerbose = true;
}
if (!config->saveTempsArgs.empty())
checkError(c.addSaveTemps(config->outputFile.str() + ".",
/*UseInputModulePath*/ true,
config->saveTempsArgs));
return c;
}
BitcodeCompiler::BitcodeCompiler() {
// Initialize indexFile.
if (!config->thinLTOIndexOnlyArg.empty())
indexFile = openFile(config->thinLTOIndexOnlyArg);
// Initialize ltoObj.
lto::ThinBackend backend;
auto onIndexWrite = [&](StringRef s) { thinIndices.erase(s); };
if (config->thinLTOIndexOnly) {
backend = lto::createWriteIndexesThinBackend(
[lld] Support separate native object file path in --thinlto-prefix-replace Currently, the --thinlto-prefix-replace="oldpath;newpath" option is used during distributed ThinLTO thin links to specify the mapping of the input bitcode object files' directory tree (oldpath) to the directory tree (newpath) used for both: 1) the output files of the thin link itself (the .thinlto.bc index files and the optional .imports files) 2) the specified object file paths written to the response file given in the --thinlto-index-only=${response} option, which is used by the final native link and must match the paths of the native object files that will be produced by ThinLTO backend compiles. This patch expands the --thinlto-prefix-replace option to allow a separate directory tree mapping to be specified for the object file paths written to the response file (number 2 above). This is important to support builds and build systems where the same output directory may not be written by multiple build actions (e.g. the thin link and the ThinLTO backend compiles). The new format is: --thinlto-prefix-replace="origpath;outpath[;objpath]" This replaces the origpath directory tree of the thin link input files with outpath when writing the thin link index and imports outputs (number 1 above). If objpath is specified it replaces origpath of the input files with objpath when writing the response file (number 2 above), otherwise it falls back to the old behavior of using outpath for this as well. Reviewed By: tejohnson, MaskRay Differential Revision: https://reviews.llvm.org/D144596
2023-04-04 09:57:53 -07:00
std::string(config->thinLTOPrefixReplaceOld),
std::string(config->thinLTOPrefixReplaceNew),
std::string(config->thinLTOPrefixReplaceNativeObject),
config->thinLTOEmitImportsFiles, indexFile.get(), onIndexWrite);
} else {
backend = lto::createInProcessThinBackend(
llvm::heavyweight_hardware_concurrency(config->thinLTOJobs),
onIndexWrite, config->thinLTOEmitIndexFiles,
config->thinLTOEmitImportsFiles);
}
constexpr llvm::lto::LTO::LTOKind ltoModes[3] =
{llvm::lto::LTO::LTOKind::LTOK_UnifiedThin,
llvm::lto::LTO::LTOKind::LTOK_UnifiedRegular,
llvm::lto::LTO::LTOKind::LTOK_Default};
ltoObj = std::make_unique<lto::LTO>(
createConfig(), backend, config->ltoPartitions,
ltoModes[config->ltoKind]);
// Initialize usedStartStop.
if (ctx.bitcodeFiles.empty())
return;
for (Symbol *sym : symtab.getSymbols()) {
if (sym->isPlaceholder())
continue;
StringRef s = sym->getName();
for (StringRef prefix : {"__start_", "__stop_"})
if (s.starts_with(prefix))
usedStartStop.insert(s.substr(prefix.size()));
}
}
BitcodeCompiler::~BitcodeCompiler() = default;
void BitcodeCompiler::add(BitcodeFile &f) {
lto::InputFile &obj = *f.obj;
bool isExec = !config->shared && !config->relocatable;
if (config->thinLTOEmitIndexFiles)
thinIndices.insert(obj.getName());
ArrayRef<Symbol *> syms = f.getSymbols();
ArrayRef<lto::InputFile::Symbol> objSyms = obj.symbols();
std::vector<lto::SymbolResolution> resols(syms.size());
// Provide a resolution to the LTO API for each symbol.
for (size_t i = 0, e = syms.size(); i != e; ++i) {
Symbol *sym = syms[i];
const lto::InputFile::Symbol &objSym = objSyms[i];
lto::SymbolResolution &r = resols[i];
// Ideally we shouldn't check for SF_Undefined but currently IRObjectFile
// reports two symbols for module ASM defined. Without this check, lld
// flags an undefined in IR with a definition in ASM as prevailing.
// Once IRObjectFile is fixed to report only one symbol this hack can
// be removed.
r.Prevailing = !objSym.isUndefined() && sym->file == &f;
// We ask LTO to preserve following global symbols:
// 1) All symbols when doing relocatable link, so that them can be used
// for doing final link.
// 2) Symbols that are used in regular objects.
// 3) C named sections if we have corresponding __start_/__stop_ symbol.
// 4) Symbols that are defined in bitcode files and used for dynamic
// linking.
// 5) Symbols that will be referenced after linker wrapping is performed.
r.VisibleToRegularObj = config->relocatable || sym->isUsedInRegularObj ||
sym->referencedAfterWrap ||
(r.Prevailing && sym->includeInDynsym()) ||
usedStartStop.count(objSym.getSectionName());
// Identify symbols exported dynamically, and that therefore could be
// referenced by a shared library not visible to the linker.
r.ExportDynamic =
sym->computeBinding() != STB_LOCAL &&
(config->exportDynamic || sym->exportDynamic || sym->inDynamicList);
const auto *dr = dyn_cast<Defined>(sym);
r.FinalDefinitionInLinkageUnit =
(isExec || sym->visibility() != STV_DEFAULT) && dr &&
// Skip absolute symbols from ELF objects, otherwise PC-rel relocations
// will be generated by for them, triggering linker errors.
// Symbol section is always null for bitcode symbols, hence the check
// for isElf(). Skip linker script defined symbols as well: they have
// no File defined.
!(dr->section == nullptr &&
(sym->file->isInternal() || sym->file->isElf()));
if (r.Prevailing)
Undefined(ctx.internalFile, StringRef(), STB_GLOBAL, STV_DEFAULT,
sym->type)
.overwrite(*sym);
// We tell LTO to not apply interprocedural optimization for wrapped
// (with --wrap) symbols because otherwise LTO would inline them while
// their values are still not final.
r.LinkerRedefined = sym->scriptDefined;
}
checkError(ltoObj->add(std::move(f.obj), resols));
}
// If LazyObjFile has not been added to link, emit empty index files.
// This is needed because this is what GNU gold plugin does and we have a
// distributed build system that depends on that behavior.
static void thinLTOCreateEmptyIndexFiles() {
DenseSet<StringRef> linkedBitCodeFiles;
for (BitcodeFile *f : ctx.bitcodeFiles)
linkedBitCodeFiles.insert(f->getName());
for (BitcodeFile *f : ctx.lazyBitcodeFiles) {
if (!f->lazy)
continue;
if (linkedBitCodeFiles.contains(f->getName()))
continue;
std::string path =
replaceThinLTOSuffix(getThinLTOOutputFile(f->obj->getName()));
std::unique_ptr<raw_fd_ostream> os = openFile(path + ".thinlto.bc");
if (!os)
continue;
ModuleSummaryIndex m(/*HaveGVs*/ false);
m.setSkipModuleByDistributedBackend();
writeIndexToFile(m, *os);
if (config->thinLTOEmitImportsFiles)
openFile(path + ".imports");
}
}
// Merge all the bitcode files we have seen, codegen the result
// and return the resulting ObjectFile(s).
std::vector<InputFile *> BitcodeCompiler::compile() {
unsigned maxTasks = ltoObj->getMaxTasks();
buf.resize(maxTasks);
files.resize(maxTasks);
filenames.resize(maxTasks);
// The --thinlto-cache-dir option specifies the path to a directory in which
// to cache native object files for ThinLTO incremental builds. If a path was
// specified, configure LTO to use it as the cache directory.
FileCache cache;
if (!config->thinLTOCacheDir.empty())
cache = check(localCache("ThinLTO", "Thin", config->thinLTOCacheDir,
[&](size_t task, const Twine &moduleName,
std::unique_ptr<MemoryBuffer> mb) {
files[task] = std::move(mb);
filenames[task] = moduleName.str();
}));
if (!ctx.bitcodeFiles.empty())
checkError(ltoObj->run(
[&](size_t task, const Twine &moduleName) {
buf[task].first = moduleName.str();
return std::make_unique<CachedFileStream>(
std::make_unique<raw_svector_ostream>(buf[task].second));
},
cache));
// Emit empty index files for non-indexed files but not in single-module mode.
if (config->thinLTOModulesToCompile.empty()) {
for (StringRef s : thinIndices) {
std::string path = getThinLTOOutputFile(s);
openFile(path + ".thinlto.bc");
if (config->thinLTOEmitImportsFiles)
openFile(path + ".imports");
}
}
if (config->thinLTOEmitIndexFiles)
thinLTOCreateEmptyIndexFiles();
if (config->thinLTOIndexOnly) {
if (!config->ltoObjPath.empty())
saveBuffer(buf[0].second, config->ltoObjPath);
// ThinLTO with index only option is required to generate only the index
// files. After that, we exit from linker and ThinLTO backend runs in a
// distributed environment.
if (indexFile)
indexFile->close();
return {};
}
if (!config->thinLTOCacheDir.empty())
pruneCache(config->thinLTOCacheDir, config->thinLTOCachePolicy, files);
if (!config->ltoObjPath.empty()) {
saveBuffer(buf[0].second, config->ltoObjPath);
for (unsigned i = 1; i != maxTasks; ++i)
saveBuffer(buf[i].second, config->ltoObjPath + Twine(i));
}
bool savePrelink = config->saveTempsArgs.contains("prelink");
std::vector<InputFile *> ret;
const char *ext = config->ltoEmitAsm ? ".s" : ".o";
for (unsigned i = 0; i != maxTasks; ++i) {
StringRef bitcodeFilePath;
StringRef objBuf;
if (files[i]) {
// When files[i] is not null, we get the native relocatable file from the
// cache. filenames[i] contains the original BitcodeFile's identifier.
objBuf = files[i]->getBuffer();
bitcodeFilePath = filenames[i];
} else {
// Get the native relocatable file after in-process LTO compilation.
objBuf = buf[i].second;
bitcodeFilePath = buf[i].first;
}
if (objBuf.empty())
continue;
// If the input bitcode file is path/to/x.o and -o specifies a.out, the
// corresponding native relocatable file path will look like:
// path/to/a.out.lto.x.o.
StringRef ltoObjName;
if (bitcodeFilePath == "ld-temp.o") {
ltoObjName =
saver().save(Twine(config->outputFile) + ".lto" +
(i == 0 ? Twine("") : Twine('.') + Twine(i)) + ext);
} else {
StringRef directory = sys::path::parent_path(bitcodeFilePath);
// For an archive member, which has an identifier like "d/a.a(coll.o at
// 8)" (see BitcodeFile::BitcodeFile), use the filename; otherwise, use
// the stem (d/a.o => a).
StringRef baseName = bitcodeFilePath.ends_with(")")
? sys::path::filename(bitcodeFilePath)
: sys::path::stem(bitcodeFilePath);
StringRef outputFileBaseName = sys::path::filename(config->outputFile);
SmallString<256> path;
sys::path::append(path, directory,
outputFileBaseName + ".lto." + baseName + ext);
sys::path::remove_dots(path, true);
ltoObjName = saver().save(path.str());
}
if (savePrelink || config->ltoEmitAsm)
saveBuffer(buf[i].second, ltoObjName);
if (!config->ltoEmitAsm)
ret.push_back(createObjFile(MemoryBufferRef(objBuf, ltoObjName)));
}
return ret;
}