Files
clang-p2996/lld/MachO/ConcatOutputSection.cpp
Jez Ng bcaf57cae8 [lld-macho] Parse relocations quickly by assuming sorted order
clang and gcc both seem to emit relocations in reverse order of
address. That means we can match relocations to their containing
subsections in `O(relocs + subsections)` rather than the `O(relocs *
log(subsections))` that our previous binary search implementation
required.

Unfortunately, `ld -r` can still emit unsorted relocations, so we have a
fallback code path for that (less common) case.

Numbers for linking chromium_framework on my 3.2 GHz 16-Core Intel Xeon W:

      N           Min           Max        Median           Avg        Stddev
  x  20          4.04          4.11         4.075        4.0775   0.018027756
  +  20          3.95          4.02          3.98         3.985   0.020900768
  Difference at 95.0% confidence
          -0.0925 +/- 0.0124919
          -2.26855% +/- 0.306361%
          (Student's t, pooled s = 0.0195172)

Reviewed By: #lld-macho, thakis

Differential Revision: https://reviews.llvm.org/D105410
2021-07-05 01:13:44 -04:00

360 lines
15 KiB
C++

//===- ConcatOutputSection.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 "ConcatOutputSection.h"
#include "Config.h"
#include "OutputSegment.h"
#include "SymbolTable.h"
#include "Symbols.h"
#include "SyntheticSections.h"
#include "Target.h"
#include "lld/Common/ErrorHandler.h"
#include "lld/Common/Memory.h"
#include "llvm/BinaryFormat/MachO.h"
#include "llvm/Support/ScopedPrinter.h"
#include "llvm/Support/TimeProfiler.h"
using namespace llvm;
using namespace llvm::MachO;
using namespace lld;
using namespace lld::macho;
void ConcatOutputSection::addInput(ConcatInputSection *input) {
if (inputs.empty()) {
align = input->align;
flags = input->getFlags();
} else {
align = std::max(align, input->align);
finalizeFlags(input);
}
inputs.push_back(input);
input->parent = this;
}
// Branch-range extension can be implemented in two ways, either through ...
//
// (1) Branch islands: Single branch instructions (also of limited range),
// that might be chained in multiple hops to reach the desired
// destination. On ARM64, as 16 branch islands are needed to hop between
// opposite ends of a 2 GiB program. LD64 uses branch islands exclusively,
// even when it needs excessive hops.
//
// (2) Thunks: Instruction(s) to load the destination address into a scratch
// register, followed by a register-indirect branch. Thunks are
// constructed to reach any arbitrary address, so need not be
// chained. Although thunks need not be chained, a program might need
// multiple thunks to the same destination distributed throughout a large
// program so that all call sites can have one within range.
//
// The optimal approach is to mix islands for distinations within two hops,
// and use thunks for destinations at greater distance. For now, we only
// implement thunks. TODO: Adding support for branch islands!
//
// Internally -- as expressed in LLD's data structures -- a
// branch-range-extension thunk comprises ...
//
// (1) new Defined privateExtern symbol for the thunk named
// <FUNCTION>.thunk.<SEQUENCE>, which references ...
// (2) new InputSection, which contains ...
// (3.1) new data for the instructions to load & branch to the far address +
// (3.2) new Relocs on instructions to load the far address, which reference ...
// (4.1) existing Defined extern symbol for the real function in __text, or
// (4.2) existing DylibSymbol for the real function in a dylib
//
// Nearly-optimal thunk-placement algorithm features:
//
// * Single pass: O(n) on the number of call sites.
//
// * Accounts for the exact space overhead of thunks - no heuristics
//
// * Exploits the full range of call instructions - forward & backward
//
// Data:
//
// * DenseMap<Symbol *, ThunkInfo> thunkMap: Maps the function symbol
// to its thunk bookkeeper.
//
// * struct ThunkInfo (bookkeeper): Call instructions have limited range, and
// distant call sites might be unable to reach the same thunk, so multiple
// thunks are necessary to serve all call sites in a very large program. A
// thunkInfo stores state for all thunks associated with a particular
// function: (a) thunk symbol, (b) input section containing stub code, and
// (c) sequence number for the active thunk incarnation. When an old thunk
// goes out of range, we increment the sequence number and create a new
// thunk named <FUNCTION>.thunk.<SEQUENCE>.
//
// * A thunk incarnation comprises (a) private-extern Defined symbol pointing
// to (b) an InputSection holding machine instructions (similar to a MachO
// stub), and (c) Reloc(s) that reference the real function for fixing-up
// the stub code.
//
// * std::vector<InputSection *> MergedInputSection::thunks: A vector parallel
// to the inputs vector. We store new thunks via cheap vector append, rather
// than costly insertion into the inputs vector.
//
// Control Flow:
//
// * During address assignment, MergedInputSection::finalize() examines call
// sites by ascending address and creates thunks. When a function is beyond
// the range of a call site, we need a thunk. Place it at the largest
// available forward address from the call site. Call sites increase
// monotonically and thunks are always placed as far forward as possible;
// thus, we place thunks at monotonically increasing addresses. Once a thunk
// is placed, it and all previous input-section addresses are final.
//
// * MergedInputSection::finalize() and MergedInputSection::writeTo() merge
// the inputs and thunks vectors (both ordered by ascending address), which
// is simple and cheap.
DenseMap<Symbol *, ThunkInfo> lld::macho::thunkMap;
// Determine whether we need thunks, which depends on the target arch -- RISC
// (i.e., ARM) generally does because it has limited-range branch/call
// instructions, whereas CISC (i.e., x86) generally doesn't. RISC only needs
// thunks for programs so large that branch source & destination addresses
// might differ more than the range of branch instruction(s).
bool ConcatOutputSection::needsThunks() const {
if (!target->usesThunks())
return false;
uint64_t isecAddr = addr;
for (InputSection *isec : inputs)
isecAddr = alignTo(isecAddr, isec->align) + isec->getSize();
if (isecAddr - addr + in.stubs->getSize() <= target->branchRange)
return false;
// Yes, this program is large enough to need thunks.
for (InputSection *isec : inputs) {
for (Reloc &r : isec->relocs) {
if (!target->hasAttr(r.type, RelocAttrBits::BRANCH))
continue;
auto *sym = r.referent.get<Symbol *>();
// Pre-populate the thunkMap and memoize call site counts for every
// InputSection and ThunkInfo. We do this for the benefit of
// ConcatOutputSection::estimateStubsInRangeVA()
ThunkInfo &thunkInfo = thunkMap[sym];
// Knowing ThunkInfo call site count will help us know whether or not we
// might need to create more for this referent at the time we are
// estimating distance to __stubs in .
++thunkInfo.callSiteCount;
// Knowing InputSection call site count will help us avoid work on those
// that have no BRANCH relocs.
++isec->callSiteCount;
}
}
return true;
}
// Since __stubs is placed after __text, we must estimate the address
// beyond which stubs are within range of a simple forward branch.
uint64_t ConcatOutputSection::estimateStubsInRangeVA(size_t callIdx) const {
uint64_t branchRange = target->branchRange;
size_t endIdx = inputs.size();
ConcatInputSection *isec = inputs[callIdx];
uint64_t isecVA = isec->getVA();
// Tally the non-stub functions which still have call sites
// remaining to process, which yields the maximum number
// of thunks we might yet place.
size_t maxPotentialThunks = 0;
for (auto &tp : thunkMap) {
ThunkInfo &ti = tp.second;
maxPotentialThunks +=
!tp.first->isInStubs() && ti.callSitesUsed < ti.callSiteCount;
}
// Tally the total size of input sections remaining to process.
uint64_t isecEnd = isec->getVA();
for (size_t i = callIdx; i < endIdx; i++) {
InputSection *isec = inputs[i];
isecEnd = alignTo(isecEnd, isec->align) + isec->getSize();
}
// Estimate the address after which call sites can safely call stubs
// directly rather than through intermediary thunks.
uint64_t stubsInRangeVA = isecEnd + maxPotentialThunks * target->thunkSize +
in.stubs->getSize() - branchRange;
log("thunks = " + std::to_string(thunkMap.size()) +
", potential = " + std::to_string(maxPotentialThunks) +
", stubs = " + std::to_string(in.stubs->getSize()) + ", isecVA = " +
to_hexString(isecVA) + ", threshold = " + to_hexString(stubsInRangeVA) +
", isecEnd = " + to_hexString(isecEnd) +
", tail = " + to_hexString(isecEnd - isecVA) +
", slop = " + to_hexString(branchRange - (isecEnd - isecVA)));
return stubsInRangeVA;
}
void ConcatOutputSection::finalize() {
uint64_t isecAddr = addr;
uint64_t isecFileOff = fileOff;
auto finalizeOne = [&](ConcatInputSection *isec) {
isecAddr = alignTo(isecAddr, isec->align);
isecFileOff = alignTo(isecFileOff, isec->align);
isec->outSecOff = isecAddr - addr;
isec->isFinal = true;
isecAddr += isec->getSize();
isecFileOff += isec->getFileSize();
};
if (!needsThunks()) {
for (ConcatInputSection *isec : inputs)
finalizeOne(isec);
size = isecAddr - addr;
fileSize = isecFileOff - fileOff;
return;
}
uint64_t branchRange = target->branchRange;
uint64_t stubsInRangeVA = TargetInfo::outOfRangeVA;
size_t thunkSize = target->thunkSize;
size_t relocCount = 0;
size_t callSiteCount = 0;
size_t thunkCallCount = 0;
size_t thunkCount = 0;
// inputs[finalIdx] is for finalization (address-assignment)
size_t finalIdx = 0;
// Kick-off by ensuring that the first input section has an address
for (size_t callIdx = 0, endIdx = inputs.size(); callIdx < endIdx;
++callIdx) {
if (finalIdx == callIdx)
finalizeOne(inputs[finalIdx++]);
ConcatInputSection *isec = inputs[callIdx];
assert(isec->isFinal);
uint64_t isecVA = isec->getVA();
// Assign addresses up-to the forward branch-range limit
while (finalIdx < endIdx &&
isecAddr + inputs[finalIdx]->getSize() < isecVA + branchRange)
finalizeOne(inputs[finalIdx++]);
if (isec->callSiteCount == 0)
continue;
if (finalIdx == endIdx && stubsInRangeVA == TargetInfo::outOfRangeVA) {
// When we have finalized all input sections, __stubs (destined
// to follow __text) comes within range of forward branches and
// we can estimate the threshold address after which we can
// reach any stub with a forward branch. Note that although it
// sits in the middle of a loop, this code executes only once.
// It is in the loop because we need to call it at the proper
// time: the earliest call site from which the end of __text
// (and start of __stubs) comes within range of a forward branch.
stubsInRangeVA = estimateStubsInRangeVA(callIdx);
}
// Process relocs by ascending address, i.e., ascending offset within isec
std::vector<Reloc> &relocs = isec->relocs;
// FIXME: This property does not hold for object files produced by ld64's
// `-r` mode.
assert(is_sorted(relocs,
[](Reloc &a, Reloc &b) { return a.offset > b.offset; }));
for (Reloc &r : reverse(relocs)) {
++relocCount;
if (!target->hasAttr(r.type, RelocAttrBits::BRANCH))
continue;
++callSiteCount;
// Calculate branch reachability boundaries
uint64_t callVA = isecVA + r.offset;
uint64_t lowVA = branchRange < callVA ? callVA - branchRange : 0;
uint64_t highVA = callVA + branchRange;
// Calculate our call referent address
auto *funcSym = r.referent.get<Symbol *>();
ThunkInfo &thunkInfo = thunkMap[funcSym];
// The referent is not reachable, so we need to use a thunk ...
if (funcSym->isInStubs() && callVA >= stubsInRangeVA) {
// ... Oh, wait! We are close enough to the end that __stubs
// are now within range of a simple forward branch.
continue;
}
uint64_t funcVA = funcSym->resolveBranchVA();
++thunkInfo.callSitesUsed;
if (lowVA < funcVA && funcVA < highVA) {
// The referent is reachable with a simple call instruction.
continue;
}
++thunkInfo.thunkCallCount;
++thunkCallCount;
// If an existing thunk is reachable, use it ...
if (thunkInfo.sym) {
uint64_t thunkVA = thunkInfo.isec->getVA();
if (lowVA < thunkVA && thunkVA < highVA) {
r.referent = thunkInfo.sym;
continue;
}
}
// ... otherwise, create a new thunk
if (isecAddr > highVA) {
// When there is small-to-no margin between highVA and
// isecAddr and the distance between subsequent call sites is
// smaller than thunkSize, then a new thunk can go out of
// range. Fix by unfinalizing inputs[finalIdx] to reduce the
// distance between callVA and highVA, then shift some thunks
// to occupy address-space formerly occupied by the
// unfinalized inputs[finalIdx].
fatal(Twine(__FUNCTION__) + ": FIXME: thunk range overrun");
}
thunkInfo.isec =
make<ConcatInputSection>(isec->getSegName(), isec->getName());
thunkInfo.isec->parent = this;
StringRef thunkName = saver.save(funcSym->getName() + ".thunk." +
std::to_string(thunkInfo.sequence++));
r.referent = thunkInfo.sym = symtab->addDefined(
thunkName, /*file=*/nullptr, thunkInfo.isec, /*value=*/0,
/*size=*/thunkSize, /*isWeakDef=*/false, /*isPrivateExtern=*/true,
/*isThumb=*/false, /*isReferencedDynamically=*/false,
/*noDeadStrip=*/false);
target->populateThunk(thunkInfo.isec, funcSym);
finalizeOne(thunkInfo.isec);
thunks.push_back(thunkInfo.isec);
++thunkCount;
}
}
size = isecAddr - addr;
fileSize = isecFileOff - fileOff;
log("thunks for " + parent->name + "," + name +
": funcs = " + std::to_string(thunkMap.size()) +
", relocs = " + std::to_string(relocCount) +
", all calls = " + std::to_string(callSiteCount) +
", thunk calls = " + std::to_string(thunkCallCount) +
", thunks = " + std::to_string(thunkCount));
}
void ConcatOutputSection::writeTo(uint8_t *buf) const {
// Merge input sections from thunk & ordinary vectors
size_t i = 0, ie = inputs.size();
size_t t = 0, te = thunks.size();
while (i < ie || t < te) {
while (i < ie && (t == te || inputs[i]->getSize() == 0 ||
inputs[i]->outSecOff < thunks[t]->outSecOff)) {
inputs[i]->writeTo(buf + inputs[i]->outSecOff);
++i;
}
while (t < te && (i == ie || thunks[t]->outSecOff < inputs[i]->outSecOff)) {
thunks[t]->writeTo(buf + thunks[t]->outSecOff);
++t;
}
}
}
void ConcatOutputSection::finalizeFlags(InputSection *input) {
switch (sectionType(input->getFlags())) {
default /*type-unspec'ed*/:
// FIXME: Add additional logics here when supporting emitting obj files.
break;
case S_4BYTE_LITERALS:
case S_8BYTE_LITERALS:
case S_16BYTE_LITERALS:
case S_CSTRING_LITERALS:
case S_ZEROFILL:
case S_LAZY_SYMBOL_POINTERS:
case S_MOD_TERM_FUNC_POINTERS:
case S_THREAD_LOCAL_REGULAR:
case S_THREAD_LOCAL_ZEROFILL:
case S_THREAD_LOCAL_VARIABLES:
case S_THREAD_LOCAL_INIT_FUNCTION_POINTERS:
case S_THREAD_LOCAL_VARIABLE_POINTERS:
case S_NON_LAZY_SYMBOL_POINTERS:
case S_SYMBOL_STUBS:
flags |= input->getFlags();
break;
}
}