1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-25 04:02:41 +01:00
llvm-mirror/lib/MC/MCFragment.cpp
Hongtao Yu 85e4f6f241 [CSSPGO] Pseudo probe encoding and emission.
This change implements pseudo probe encoding and emission for CSSPGO. Please see RFC here for more context: https://groups.google.com/g/llvm-dev/c/1p1rdYbL93s

Pseudo probes are in the form of intrinsic calls on IR/MIR but they do not turn into any machine instructions. Instead they are emitted into the binary as a piece of data in standalone sections.  The probe-specific sections are not needed to be loaded into memory at execution time, thus they do not incur a runtime overhead. 

**ELF object emission**

The binary data to emit are organized as two ELF sections, i.e, the `.pseudo_probe_desc` section and the `.pseudo_probe` section. The `.pseudo_probe_desc` section stores a function descriptor for each function and the `.pseudo_probe` section stores the actual probes, each fo which corresponds to an IR basic block or an IR function callsite. A function descriptor is stored as a module-level metadata during the compilation and is serialized into the object file during object emission.

Both the probe descriptors and pseudo probes can be emitted into a separate ELF section per function to leverage the linker for deduplication.  A `.pseudo_probe` section shares the same COMDAT group with the function code so that when the function is dead, the probes are dead and disposed too. On the contrary, a `.pseudo_probe_desc` section has its own COMDAT group. This is because even if a function is dead, its probes may be inlined into other functions and its descriptor is still needed by the profile generation tool.

The format of `.pseudo_probe_desc` section looks like:

```
.section   .pseudo_probe_desc,"",@progbits
.quad   6309742469962978389  // Func GUID
.quad   4294967295           // Func Hash
.byte   9                    // Length of func name
.ascii  "_Z5funcAi"          // Func name
.quad   7102633082150537521
.quad   138828622701
.byte   12
.ascii  "_Z8funcLeafi"
.quad   446061515086924981
.quad   4294967295
.byte   9
.ascii  "_Z5funcBi"
.quad   -2016976694713209516
.quad   72617220756
.byte   7
.ascii  "_Z3fibi"
```

For each `.pseudoprobe` section, the encoded binary data consists of a single function record corresponding to an outlined function (i.e, a function with a code entry in the `.text` section). A function record has the following format :

```
FUNCTION BODY (one for each outlined function present in the text section)
    GUID (uint64)
        GUID of the function
    NPROBES (ULEB128)
        Number of probes originating from this function.
    NUM_INLINED_FUNCTIONS (ULEB128)
        Number of callees inlined into this function, aka number of
        first-level inlinees
    PROBE RECORDS
        A list of NPROBES entries. Each entry contains:
          INDEX (ULEB128)
          TYPE (uint4)
            0 - block probe, 1 - indirect call, 2 - direct call
          ATTRIBUTE (uint3)
            reserved
          ADDRESS_TYPE (uint1)
            0 - code address, 1 - address delta
          CODE_ADDRESS (uint64 or ULEB128)
            code address or address delta, depending on ADDRESS_TYPE
    INLINED FUNCTION RECORDS
        A list of NUM_INLINED_FUNCTIONS entries describing each of the inlined
        callees.  Each record contains:
          INLINE SITE
            GUID of the inlinee (uint64)
            ID of the callsite probe (ULEB128)
          FUNCTION BODY
            A FUNCTION BODY entry describing the inlined function.
```

To support building a context-sensitive profile, probes from inlinees are grouped by their inline contexts. An inline context is logically a call path through which a callee function lands in a caller function. The probe emitter builds an inline tree based on the debug metadata for each outlined function in the form of a trie tree. A tree root is the outlined function. Each tree edge stands for a callsite where inlining happens. Pseudo probes originating from an inlinee function are stored in a tree node and the tree path starting from the root all the way down to the tree node is the inline context of the probes. The emission happens on the whole tree top-down recursively. Probes of a tree node will be emitted altogether with their direct parent edge. Since a pseudo probe corresponds to a real code address, for size savings, the address is encoded as a delta from the previous probe except for the first probe. Variant-sized integer encoding, aka LEB128, is used for address delta and probe index.

**Assembling**

Pseudo probes can be printed as assembly directives alternatively. This allows for good assembly code readability and also provides a view of how optimizations and pseudo probes affect each other, especially helpful for diff time assembly analysis.

A pseudo probe directive has the following operands in order: function GUID, probe index, probe type, probe attributes and inline context. The directive is generated by the compiler and can be parsed by the assembler to form an encoded `.pseudoprobe` section in the object file.

A example assembly looks like:

```
foo2: # @foo2
# %bb.0: # %bb0
pushq %rax
testl %edi, %edi
.pseudoprobe 837061429793323041 1 0 0
je .LBB1_1
# %bb.2: # %bb2
.pseudoprobe 837061429793323041 6 2 0
callq foo
.pseudoprobe 837061429793323041 3 0 0
.pseudoprobe 837061429793323041 4 0 0
popq %rax
retq
.LBB1_1: # %bb1
.pseudoprobe 837061429793323041 5 1 0
callq *%rsi
.pseudoprobe 837061429793323041 2 0 0
.pseudoprobe 837061429793323041 4 0 0
popq %rax
retq
# -- End function
.section .pseudo_probe_desc,"",@progbits
.quad 6699318081062747564
.quad 72617220756
.byte 3
.ascii "foo"
.quad 837061429793323041
.quad 281547593931412
.byte 4
.ascii "foo2"
```

With inlining turned on, the assembly may look different around %bb2 with an inlined probe:

```
# %bb.2:                                # %bb2
.pseudoprobe    837061429793323041 3 0
.pseudoprobe    6699318081062747564 1 0 @ 837061429793323041:6
.pseudoprobe    837061429793323041 4 0
popq    %rax
retq
```

**Disassembling**

We have a disassembling tool (llvm-profgen) that can display disassembly alongside with pseudo probes. So far it only supports ELF executable file.

An example disassembly looks like:

```
00000000002011a0 <foo2>:
  2011a0: 50                    push   rax
  2011a1: 85 ff                 test   edi,edi
  [Probe]:  FUNC: foo2  Index: 1  Type: Block
  2011a3: 74 02                 je     2011a7 <foo2+0x7>
  [Probe]:  FUNC: foo2  Index: 3  Type: Block
  [Probe]:  FUNC: foo2  Index: 4  Type: Block
  [Probe]:  FUNC: foo   Index: 1  Type: Block  Inlined: @ foo2:6
  2011a5: 58                    pop    rax
  2011a6: c3                    ret
  [Probe]:  FUNC: foo2  Index: 2  Type: Block
  2011a7: bf 01 00 00 00        mov    edi,0x1
  [Probe]:  FUNC: foo2  Index: 5  Type: IndirectCall
  2011ac: ff d6                 call   rsi
  [Probe]:  FUNC: foo2  Index: 4  Type: Block
  2011ae: 58                    pop    rax
  2011af: c3                    ret
```

Reviewed By: wmi

Differential Revision: https://reviews.llvm.org/D91878
2020-12-10 17:29:28 -08:00

505 lines
16 KiB
C++

//===- lib/MC/MCFragment.cpp - Assembler Fragment Implementation ----------===//
//
// 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 "llvm/MC/MCFragment.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/ADT/Twine.h"
#include "llvm/Config/llvm-config.h"
#include "llvm/MC/MCAsmLayout.h"
#include "llvm/MC/MCAssembler.h"
#include "llvm/MC/MCContext.h"
#include "llvm/MC/MCExpr.h"
#include "llvm/MC/MCFixup.h"
#include "llvm/MC/MCSection.h"
#include "llvm/MC/MCSymbol.h"
#include "llvm/MC/MCValue.h"
#include "llvm/Support/Casting.h"
#include "llvm/Support/Compiler.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/raw_ostream.h"
#include <cassert>
#include <cstdint>
#include <utility>
using namespace llvm;
MCAsmLayout::MCAsmLayout(MCAssembler &Asm) : Assembler(Asm) {
// Compute the section layout order. Virtual sections must go last.
for (MCSection &Sec : Asm)
if (!Sec.isVirtualSection())
SectionOrder.push_back(&Sec);
for (MCSection &Sec : Asm)
if (Sec.isVirtualSection())
SectionOrder.push_back(&Sec);
}
bool MCAsmLayout::isFragmentValid(const MCFragment *F) const {
const MCSection *Sec = F->getParent();
const MCFragment *LastValid = LastValidFragment.lookup(Sec);
if (!LastValid)
return false;
assert(LastValid->getParent() == Sec);
return F->getLayoutOrder() <= LastValid->getLayoutOrder();
}
bool MCAsmLayout::canGetFragmentOffset(const MCFragment *F) const {
MCSection *Sec = F->getParent();
MCSection::iterator I;
if (MCFragment *LastValid = LastValidFragment[Sec]) {
// Fragment already valid, offset is available.
if (F->getLayoutOrder() <= LastValid->getLayoutOrder())
return true;
I = ++MCSection::iterator(LastValid);
} else
I = Sec->begin();
// A fragment ordered before F is currently being laid out.
const MCFragment *FirstInvalidFragment = &*I;
if (FirstInvalidFragment->IsBeingLaidOut)
return false;
return true;
}
void MCAsmLayout::invalidateFragmentsFrom(MCFragment *F) {
// If this fragment wasn't already valid, we don't need to do anything.
if (!isFragmentValid(F))
return;
// Otherwise, reset the last valid fragment to the previous fragment
// (if this is the first fragment, it will be NULL).
LastValidFragment[F->getParent()] = F->getPrevNode();
}
void MCAsmLayout::ensureValid(const MCFragment *F) const {
MCSection *Sec = F->getParent();
MCSection::iterator I;
if (MCFragment *Cur = LastValidFragment[Sec])
I = ++MCSection::iterator(Cur);
else
I = Sec->begin();
// Advance the layout position until the fragment is valid.
while (!isFragmentValid(F)) {
assert(I != Sec->end() && "Layout bookkeeping error");
const_cast<MCAsmLayout *>(this)->layoutFragment(&*I);
++I;
}
}
uint64_t MCAsmLayout::getFragmentOffset(const MCFragment *F) const {
ensureValid(F);
assert(F->Offset != ~UINT64_C(0) && "Address not set!");
return F->Offset;
}
// Simple getSymbolOffset helper for the non-variable case.
static bool getLabelOffset(const MCAsmLayout &Layout, const MCSymbol &S,
bool ReportError, uint64_t &Val) {
if (!S.getFragment()) {
if (ReportError)
report_fatal_error("unable to evaluate offset to undefined symbol '" +
S.getName() + "'");
return false;
}
Val = Layout.getFragmentOffset(S.getFragment()) + S.getOffset();
return true;
}
static bool getSymbolOffsetImpl(const MCAsmLayout &Layout, const MCSymbol &S,
bool ReportError, uint64_t &Val) {
if (!S.isVariable())
return getLabelOffset(Layout, S, ReportError, Val);
// If SD is a variable, evaluate it.
MCValue Target;
if (!S.getVariableValue()->evaluateAsValue(Target, Layout))
report_fatal_error("unable to evaluate offset for variable '" +
S.getName() + "'");
uint64_t Offset = Target.getConstant();
const MCSymbolRefExpr *A = Target.getSymA();
if (A) {
uint64_t ValA;
if (!getLabelOffset(Layout, A->getSymbol(), ReportError, ValA))
return false;
Offset += ValA;
}
const MCSymbolRefExpr *B = Target.getSymB();
if (B) {
uint64_t ValB;
if (!getLabelOffset(Layout, B->getSymbol(), ReportError, ValB))
return false;
Offset -= ValB;
}
Val = Offset;
return true;
}
bool MCAsmLayout::getSymbolOffset(const MCSymbol &S, uint64_t &Val) const {
return getSymbolOffsetImpl(*this, S, false, Val);
}
uint64_t MCAsmLayout::getSymbolOffset(const MCSymbol &S) const {
uint64_t Val;
getSymbolOffsetImpl(*this, S, true, Val);
return Val;
}
const MCSymbol *MCAsmLayout::getBaseSymbol(const MCSymbol &Symbol) const {
if (!Symbol.isVariable())
return &Symbol;
const MCExpr *Expr = Symbol.getVariableValue();
MCValue Value;
if (!Expr->evaluateAsValue(Value, *this)) {
Assembler.getContext().reportError(
Expr->getLoc(), "expression could not be evaluated");
return nullptr;
}
const MCSymbolRefExpr *RefB = Value.getSymB();
if (RefB) {
Assembler.getContext().reportError(
Expr->getLoc(), Twine("symbol '") + RefB->getSymbol().getName() +
"' could not be evaluated in a subtraction expression");
return nullptr;
}
const MCSymbolRefExpr *A = Value.getSymA();
if (!A)
return nullptr;
const MCSymbol &ASym = A->getSymbol();
const MCAssembler &Asm = getAssembler();
if (ASym.isCommon()) {
Asm.getContext().reportError(Expr->getLoc(),
"Common symbol '" + ASym.getName() +
"' cannot be used in assignment expr");
return nullptr;
}
return &ASym;
}
uint64_t MCAsmLayout::getSectionAddressSize(const MCSection *Sec) const {
// The size is the last fragment's end offset.
const MCFragment &F = Sec->getFragmentList().back();
return getFragmentOffset(&F) + getAssembler().computeFragmentSize(*this, F);
}
uint64_t MCAsmLayout::getSectionFileSize(const MCSection *Sec) const {
// Virtual sections have no file size.
if (Sec->isVirtualSection())
return 0;
// Otherwise, the file size is the same as the address space size.
return getSectionAddressSize(Sec);
}
uint64_t llvm::computeBundlePadding(const MCAssembler &Assembler,
const MCEncodedFragment *F,
uint64_t FOffset, uint64_t FSize) {
uint64_t BundleSize = Assembler.getBundleAlignSize();
assert(BundleSize > 0 &&
"computeBundlePadding should only be called if bundling is enabled");
uint64_t BundleMask = BundleSize - 1;
uint64_t OffsetInBundle = FOffset & BundleMask;
uint64_t EndOfFragment = OffsetInBundle + FSize;
// There are two kinds of bundling restrictions:
//
// 1) For alignToBundleEnd(), add padding to ensure that the fragment will
// *end* on a bundle boundary.
// 2) Otherwise, check if the fragment would cross a bundle boundary. If it
// would, add padding until the end of the bundle so that the fragment
// will start in a new one.
if (F->alignToBundleEnd()) {
// Three possibilities here:
//
// A) The fragment just happens to end at a bundle boundary, so we're good.
// B) The fragment ends before the current bundle boundary: pad it just
// enough to reach the boundary.
// C) The fragment ends after the current bundle boundary: pad it until it
// reaches the end of the next bundle boundary.
//
// Note: this code could be made shorter with some modulo trickery, but it's
// intentionally kept in its more explicit form for simplicity.
if (EndOfFragment == BundleSize)
return 0;
else if (EndOfFragment < BundleSize)
return BundleSize - EndOfFragment;
else { // EndOfFragment > BundleSize
return 2 * BundleSize - EndOfFragment;
}
} else if (OffsetInBundle > 0 && EndOfFragment > BundleSize)
return BundleSize - OffsetInBundle;
else
return 0;
}
/* *** */
void ilist_alloc_traits<MCFragment>::deleteNode(MCFragment *V) { V->destroy(); }
MCFragment::MCFragment(FragmentType Kind, bool HasInstructions,
MCSection *Parent)
: Parent(Parent), Atom(nullptr), Offset(~UINT64_C(0)), LayoutOrder(0),
Kind(Kind), IsBeingLaidOut(false), HasInstructions(HasInstructions) {
if (Parent && !isa<MCDummyFragment>(*this))
Parent->getFragmentList().push_back(this);
}
void MCFragment::destroy() {
// First check if we are the sentinal.
if (Kind == FragmentType(~0)) {
delete this;
return;
}
switch (Kind) {
case FT_Align:
delete cast<MCAlignFragment>(this);
return;
case FT_Data:
delete cast<MCDataFragment>(this);
return;
case FT_CompactEncodedInst:
delete cast<MCCompactEncodedInstFragment>(this);
return;
case FT_Fill:
delete cast<MCFillFragment>(this);
return;
case FT_Nops:
delete cast<MCNopsFragment>(this);
return;
case FT_Relaxable:
delete cast<MCRelaxableFragment>(this);
return;
case FT_Org:
delete cast<MCOrgFragment>(this);
return;
case FT_Dwarf:
delete cast<MCDwarfLineAddrFragment>(this);
return;
case FT_DwarfFrame:
delete cast<MCDwarfCallFrameFragment>(this);
return;
case FT_LEB:
delete cast<MCLEBFragment>(this);
return;
case FT_BoundaryAlign:
delete cast<MCBoundaryAlignFragment>(this);
return;
case FT_SymbolId:
delete cast<MCSymbolIdFragment>(this);
return;
case FT_CVInlineLines:
delete cast<MCCVInlineLineTableFragment>(this);
return;
case FT_CVDefRange:
delete cast<MCCVDefRangeFragment>(this);
return;
case FT_PseudoProbe:
delete cast<MCPseudoProbeAddrFragment>(this);
return;
case FT_Dummy:
delete cast<MCDummyFragment>(this);
return;
}
}
// Debugging methods
namespace llvm {
raw_ostream &operator<<(raw_ostream &OS, const MCFixup &AF) {
OS << "<MCFixup" << " Offset:" << AF.getOffset()
<< " Value:" << *AF.getValue()
<< " Kind:" << AF.getKind() << ">";
return OS;
}
} // end namespace llvm
#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
LLVM_DUMP_METHOD void MCFragment::dump() const {
raw_ostream &OS = errs();
OS << "<";
switch (getKind()) {
case MCFragment::FT_Align: OS << "MCAlignFragment"; break;
case MCFragment::FT_Data: OS << "MCDataFragment"; break;
case MCFragment::FT_CompactEncodedInst:
OS << "MCCompactEncodedInstFragment"; break;
case MCFragment::FT_Fill: OS << "MCFillFragment"; break;
case MCFragment::FT_Nops:
OS << "MCFNopsFragment";
break;
case MCFragment::FT_Relaxable: OS << "MCRelaxableFragment"; break;
case MCFragment::FT_Org: OS << "MCOrgFragment"; break;
case MCFragment::FT_Dwarf: OS << "MCDwarfFragment"; break;
case MCFragment::FT_DwarfFrame: OS << "MCDwarfCallFrameFragment"; break;
case MCFragment::FT_LEB: OS << "MCLEBFragment"; break;
case MCFragment::FT_BoundaryAlign: OS<<"MCBoundaryAlignFragment"; break;
case MCFragment::FT_SymbolId: OS << "MCSymbolIdFragment"; break;
case MCFragment::FT_CVInlineLines: OS << "MCCVInlineLineTableFragment"; break;
case MCFragment::FT_CVDefRange: OS << "MCCVDefRangeTableFragment"; break;
case MCFragment::FT_PseudoProbe:
OS << "MCPseudoProbe";
break;
case MCFragment::FT_Dummy: OS << "MCDummyFragment"; break;
}
OS << "<MCFragment " << (const void *)this << " LayoutOrder:" << LayoutOrder
<< " Offset:" << Offset << " HasInstructions:" << hasInstructions();
if (const auto *EF = dyn_cast<MCEncodedFragment>(this))
OS << " BundlePadding:" << static_cast<unsigned>(EF->getBundlePadding());
OS << ">";
switch (getKind()) {
case MCFragment::FT_Align: {
const auto *AF = cast<MCAlignFragment>(this);
if (AF->hasEmitNops())
OS << " (emit nops)";
OS << "\n ";
OS << " Alignment:" << AF->getAlignment()
<< " Value:" << AF->getValue() << " ValueSize:" << AF->getValueSize()
<< " MaxBytesToEmit:" << AF->getMaxBytesToEmit() << ">";
break;
}
case MCFragment::FT_Data: {
const auto *DF = cast<MCDataFragment>(this);
OS << "\n ";
OS << " Contents:[";
const SmallVectorImpl<char> &Contents = DF->getContents();
for (unsigned i = 0, e = Contents.size(); i != e; ++i) {
if (i) OS << ",";
OS << hexdigit((Contents[i] >> 4) & 0xF) << hexdigit(Contents[i] & 0xF);
}
OS << "] (" << Contents.size() << " bytes)";
if (DF->fixup_begin() != DF->fixup_end()) {
OS << ",\n ";
OS << " Fixups:[";
for (MCDataFragment::const_fixup_iterator it = DF->fixup_begin(),
ie = DF->fixup_end(); it != ie; ++it) {
if (it != DF->fixup_begin()) OS << ",\n ";
OS << *it;
}
OS << "]";
}
break;
}
case MCFragment::FT_CompactEncodedInst: {
const auto *CEIF =
cast<MCCompactEncodedInstFragment>(this);
OS << "\n ";
OS << " Contents:[";
const SmallVectorImpl<char> &Contents = CEIF->getContents();
for (unsigned i = 0, e = Contents.size(); i != e; ++i) {
if (i) OS << ",";
OS << hexdigit((Contents[i] >> 4) & 0xF) << hexdigit(Contents[i] & 0xF);
}
OS << "] (" << Contents.size() << " bytes)";
break;
}
case MCFragment::FT_Fill: {
const auto *FF = cast<MCFillFragment>(this);
OS << " Value:" << static_cast<unsigned>(FF->getValue())
<< " ValueSize:" << static_cast<unsigned>(FF->getValueSize())
<< " NumValues:" << FF->getNumValues();
break;
}
case MCFragment::FT_Nops: {
const auto *NF = cast<MCNopsFragment>(this);
OS << " NumBytes:" << NF->getNumBytes()
<< " ControlledNopLength:" << NF->getControlledNopLength();
break;
}
case MCFragment::FT_Relaxable: {
const auto *F = cast<MCRelaxableFragment>(this);
OS << "\n ";
OS << " Inst:";
F->getInst().dump_pretty(OS);
OS << " (" << F->getContents().size() << " bytes)";
break;
}
case MCFragment::FT_Org: {
const auto *OF = cast<MCOrgFragment>(this);
OS << "\n ";
OS << " Offset:" << OF->getOffset()
<< " Value:" << static_cast<unsigned>(OF->getValue());
break;
}
case MCFragment::FT_Dwarf: {
const auto *OF = cast<MCDwarfLineAddrFragment>(this);
OS << "\n ";
OS << " AddrDelta:" << OF->getAddrDelta()
<< " LineDelta:" << OF->getLineDelta();
break;
}
case MCFragment::FT_DwarfFrame: {
const auto *CF = cast<MCDwarfCallFrameFragment>(this);
OS << "\n ";
OS << " AddrDelta:" << CF->getAddrDelta();
break;
}
case MCFragment::FT_LEB: {
const auto *LF = cast<MCLEBFragment>(this);
OS << "\n ";
OS << " Value:" << LF->getValue() << " Signed:" << LF->isSigned();
break;
}
case MCFragment::FT_BoundaryAlign: {
const auto *BF = cast<MCBoundaryAlignFragment>(this);
OS << "\n ";
OS << " BoundarySize:" << BF->getAlignment().value()
<< " LastFragment:" << BF->getLastFragment()
<< " Size:" << BF->getSize();
break;
}
case MCFragment::FT_SymbolId: {
const auto *F = cast<MCSymbolIdFragment>(this);
OS << "\n ";
OS << " Sym:" << F->getSymbol();
break;
}
case MCFragment::FT_CVInlineLines: {
const auto *F = cast<MCCVInlineLineTableFragment>(this);
OS << "\n ";
OS << " Sym:" << *F->getFnStartSym();
break;
}
case MCFragment::FT_CVDefRange: {
const auto *F = cast<MCCVDefRangeFragment>(this);
OS << "\n ";
for (std::pair<const MCSymbol *, const MCSymbol *> RangeStartEnd :
F->getRanges()) {
OS << " RangeStart:" << RangeStartEnd.first;
OS << " RangeEnd:" << RangeStartEnd.second;
}
break;
}
case MCFragment::FT_PseudoProbe: {
const auto *OF = cast<MCPseudoProbeAddrFragment>(this);
OS << "\n ";
OS << " AddrDelta:" << OF->getAddrDelta();
break;
}
case MCFragment::FT_Dummy:
break;
}
OS << ">";
}
#endif