diff options
Diffstat (limited to 'contrib/llvm/lib/DebugInfo')
20 files changed, 3299 insertions, 0 deletions
diff --git a/contrib/llvm/lib/DebugInfo/DIContext.cpp b/contrib/llvm/lib/DebugInfo/DIContext.cpp new file mode 100644 index 0000000..e2fd55f --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DIContext.cpp @@ -0,0 +1,24 @@ +//===-- DIContext.cpp -----------------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "llvm/DebugInfo/DIContext.h" +#include "DWARFContext.h" +using namespace llvm; + +DIContext::~DIContext() {} + +DIContext *DIContext::getDWARFContext(bool isLittleEndian, + StringRef infoSection, + StringRef abbrevSection, + StringRef aRangeSection, + StringRef lineSection, + StringRef stringSection) { + return new DWARFContextInMemory(isLittleEndian, infoSection, abbrevSection, + aRangeSection, lineSection, stringSection); +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFAbbreviationDeclaration.cpp b/contrib/llvm/lib/DebugInfo/DWARFAbbreviationDeclaration.cpp new file mode 100644 index 0000000..0df692c --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFAbbreviationDeclaration.cpp @@ -0,0 +1,83 @@ +//===-- DWARFAbbreviationDeclaration.cpp ----------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFAbbreviationDeclaration.h" +#include "llvm/Support/Dwarf.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +using namespace llvm; +using namespace dwarf; + +bool +DWARFAbbreviationDeclaration::extract(DataExtractor data, uint32_t* offset_ptr){ + return extract(data, offset_ptr, data.getULEB128(offset_ptr)); +} + +bool +DWARFAbbreviationDeclaration::extract(DataExtractor data, uint32_t* offset_ptr, + uint32_t code) { + Code = code; + Attributes.clear(); + if (Code) { + Tag = data.getULEB128(offset_ptr); + HasChildren = data.getU8(offset_ptr); + + while (data.isValidOffset(*offset_ptr)) { + uint16_t attr = data.getULEB128(offset_ptr); + uint16_t form = data.getULEB128(offset_ptr); + + if (attr && form) + Attributes.push_back(DWARFAttribute(attr, form)); + else + break; + } + + return Tag != 0; + } else { + Tag = 0; + HasChildren = false; + } + + return false; +} + +void DWARFAbbreviationDeclaration::dump(raw_ostream &OS) const { + const char *tagString = TagString(getTag()); + OS << '[' << getCode() << "] "; + if (tagString) + OS << tagString; + else + OS << format("DW_TAG_Unknown_%x", getTag()); + OS << "\tDW_CHILDREN_" << (hasChildren() ? "yes" : "no") << '\n'; + for (unsigned i = 0, e = Attributes.size(); i != e; ++i) { + OS << '\t'; + const char *attrString = AttributeString(Attributes[i].getAttribute()); + if (attrString) + OS << attrString; + else + OS << format("DW_AT_Unknown_%x", Attributes[i].getAttribute()); + OS << '\t'; + const char *formString = FormEncodingString(Attributes[i].getForm()); + if (formString) + OS << formString; + else + OS << format("DW_FORM_Unknown_%x", Attributes[i].getForm()); + OS << '\n'; + } + OS << '\n'; +} + +uint32_t +DWARFAbbreviationDeclaration::findAttributeIndex(uint16_t attr) const { + for (uint32_t i = 0, e = Attributes.size(); i != e; ++i) { + if (Attributes[i].getAttribute() == attr) + return i; + } + return -1U; +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFAbbreviationDeclaration.h b/contrib/llvm/lib/DebugInfo/DWARFAbbreviationDeclaration.h new file mode 100644 index 0000000..2463a3c --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFAbbreviationDeclaration.h @@ -0,0 +1,54 @@ +//===-- DWARFAbbreviationDeclaration.h --------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFABBREVIATIONDECLARATION_H +#define LLVM_DEBUGINFO_DWARFABBREVIATIONDECLARATION_H + +#include "DWARFAttribute.h" +#include "llvm/ADT/SmallVector.h" +#include "llvm/Support/DataExtractor.h" + +namespace llvm { + +class raw_ostream; + +class DWARFAbbreviationDeclaration { + uint32_t Code; + uint32_t Tag; + bool HasChildren; + SmallVector<DWARFAttribute, 8> Attributes; +public: + enum { InvalidCode = 0 }; + DWARFAbbreviationDeclaration() + : Code(InvalidCode), Tag(0), HasChildren(0) {} + + uint32_t getCode() const { return Code; } + uint32_t getTag() const { return Tag; } + bool hasChildren() const { return HasChildren; } + uint32_t getNumAttributes() const { return Attributes.size(); } + uint16_t getAttrByIndex(uint32_t idx) const { + return Attributes.size() > idx ? Attributes[idx].getAttribute() : 0; + } + uint16_t getFormByIndex(uint32_t idx) const { + return Attributes.size() > idx ? Attributes[idx].getForm() : 0; + } + + uint32_t findAttributeIndex(uint16_t attr) const; + bool extract(DataExtractor data, uint32_t* offset_ptr); + bool extract(DataExtractor data, uint32_t* offset_ptr, uint32_t code); + bool isValid() const { return Code != 0 && Tag != 0; } + void dump(raw_ostream &OS) const; + const SmallVectorImpl<DWARFAttribute> &getAttributes() const { + return Attributes; + } +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFAttribute.h b/contrib/llvm/lib/DebugInfo/DWARFAttribute.h new file mode 100644 index 0000000..6f49b63 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFAttribute.h @@ -0,0 +1,30 @@ +//===-- DWARFAttribute.h ----------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFATTRIBUTE_H +#define LLVM_DEBUGINFO_DWARFATTRIBUTE_H + +#include "llvm/Support/DataTypes.h" + +namespace llvm { + +class DWARFAttribute { + uint16_t Attribute; + uint16_t Form; + public: + DWARFAttribute(uint16_t attr, uint16_t form) + : Attribute(attr), Form(form) {} + + uint16_t getAttribute() const { return Attribute; } + uint16_t getForm() const { return Form; } +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFCompileUnit.cpp b/contrib/llvm/lib/DebugInfo/DWARFCompileUnit.cpp new file mode 100644 index 0000000..24bf97f --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFCompileUnit.cpp @@ -0,0 +1,238 @@ +//===-- DWARFCompileUnit.cpp ----------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFCompileUnit.h" +#include "DWARFContext.h" +#include "DWARFFormValue.h" +#include "llvm/Support/Dwarf.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +using namespace llvm; +using namespace dwarf; + +DataExtractor DWARFCompileUnit::getDebugInfoExtractor() const { + return DataExtractor(Context.getInfoSection(), + Context.isLittleEndian(), getAddressByteSize()); +} + +bool DWARFCompileUnit::extract(DataExtractor debug_info, uint32_t *offset_ptr) { + clear(); + + Offset = *offset_ptr; + + if (debug_info.isValidOffset(*offset_ptr)) { + uint64_t abbrOffset; + const DWARFDebugAbbrev *abbr = Context.getDebugAbbrev(); + Length = debug_info.getU32(offset_ptr); + Version = debug_info.getU16(offset_ptr); + abbrOffset = debug_info.getU32(offset_ptr); + AddrSize = debug_info.getU8(offset_ptr); + + bool lengthOK = debug_info.isValidOffset(getNextCompileUnitOffset()-1); + bool versionOK = DWARFContext::isSupportedVersion(Version); + bool abbrOffsetOK = Context.getAbbrevSection().size() > abbrOffset; + bool addrSizeOK = AddrSize == 4 || AddrSize == 8; + + if (lengthOK && versionOK && addrSizeOK && abbrOffsetOK && abbr != NULL) { + Abbrevs = abbr->getAbbreviationDeclarationSet(abbrOffset); + return true; + } + + // reset the offset to where we tried to parse from if anything went wrong + *offset_ptr = Offset; + } + + return false; +} + +uint32_t +DWARFCompileUnit::extract(uint32_t offset, DataExtractor debug_info_data, + const DWARFAbbreviationDeclarationSet *abbrevs) { + clear(); + + Offset = offset; + + if (debug_info_data.isValidOffset(offset)) { + Length = debug_info_data.getU32(&offset); + Version = debug_info_data.getU16(&offset); + bool abbrevsOK = debug_info_data.getU32(&offset) == abbrevs->getOffset(); + Abbrevs = abbrevs; + AddrSize = debug_info_data.getU8 (&offset); + + bool versionOK = DWARFContext::isSupportedVersion(Version); + bool addrSizeOK = AddrSize == 4 || AddrSize == 8; + + if (versionOK && addrSizeOK && abbrevsOK && + debug_info_data.isValidOffset(offset)) + return offset; + } + return 0; +} + +void DWARFCompileUnit::clear() { + Offset = 0; + Length = 0; + Version = 0; + Abbrevs = 0; + AddrSize = 0; + BaseAddr = 0; + DieArray.clear(); +} + +void DWARFCompileUnit::dump(raw_ostream &OS) { + OS << format("0x%08x", Offset) << ": Compile Unit:" + << " length = " << format("0x%08x", Length) + << " version = " << format("0x%04x", Version) + << " abbr_offset = " << format("0x%04x", Abbrevs->getOffset()) + << " addr_size = " << format("0x%02x", AddrSize) + << " (next CU at " << format("0x%08x", getNextCompileUnitOffset()) + << ")\n"; + + getCompileUnitDIE(false)->dump(OS, this, -1U); +} + +void DWARFCompileUnit::setDIERelations() { + if (DieArray.empty()) + return; + DWARFDebugInfoEntryMinimal *die_array_begin = &DieArray.front(); + DWARFDebugInfoEntryMinimal *die_array_end = &DieArray.back(); + DWARFDebugInfoEntryMinimal *curr_die; + // We purposely are skipping the last element in the array in the loop below + // so that we can always have a valid next item + for (curr_die = die_array_begin; curr_die < die_array_end; ++curr_die) { + // Since our loop doesn't include the last element, we can always + // safely access the next die in the array. + DWARFDebugInfoEntryMinimal *next_die = curr_die + 1; + + const DWARFAbbreviationDeclaration *curr_die_abbrev = + curr_die->getAbbreviationDeclarationPtr(); + + if (curr_die_abbrev) { + // Normal DIE + if (curr_die_abbrev->hasChildren()) + next_die->setParent(curr_die); + else + curr_die->setSibling(next_die); + } else { + // NULL DIE that terminates a sibling chain + DWARFDebugInfoEntryMinimal *parent = curr_die->getParent(); + if (parent) + parent->setSibling(next_die); + } + } + + // Since we skipped the last element, we need to fix it up! + if (die_array_begin < die_array_end) + curr_die->setParent(die_array_begin); +} + +size_t DWARFCompileUnit::extractDIEsIfNeeded(bool cu_die_only) { + const size_t initial_die_array_size = DieArray.size(); + if ((cu_die_only && initial_die_array_size > 0) || + initial_die_array_size > 1) + return 0; // Already parsed + + // Set the offset to that of the first DIE and calculate the start of the + // next compilation unit header. + uint32_t offset = getFirstDIEOffset(); + uint32_t next_cu_offset = getNextCompileUnitOffset(); + + DWARFDebugInfoEntryMinimal die; + // Keep a flat array of the DIE for binary lookup by DIE offset + uint32_t depth = 0; + // We are in our compile unit, parse starting at the offset + // we were told to parse + + const uint8_t *fixed_form_sizes = + DWARFFormValue::getFixedFormSizesForAddressSize(getAddressByteSize()); + + while (offset < next_cu_offset && + die.extractFast(this, fixed_form_sizes, &offset)) { + + if (depth == 0) { + uint64_t base_addr = + die.getAttributeValueAsUnsigned(this, DW_AT_low_pc, -1U); + if (base_addr == -1U) + base_addr = die.getAttributeValueAsUnsigned(this, DW_AT_entry_pc, 0); + setBaseAddress(base_addr); + } + + if (cu_die_only) { + addDIE(die); + return 1; + } + else if (depth == 0 && initial_die_array_size == 1) { + // Don't append the CU die as we already did that + } else { + addDIE (die); + } + + const DWARFAbbreviationDeclaration *abbrDecl = + die.getAbbreviationDeclarationPtr(); + if (abbrDecl) { + // Normal DIE + if (abbrDecl->hasChildren()) + ++depth; + } else { + // NULL DIE. + if (depth > 0) + --depth; + if (depth == 0) + break; // We are done with this compile unit! + } + + } + + // Give a little bit of info if we encounter corrupt DWARF (our offset + // should always terminate at or before the start of the next compilation + // unit header). + if (offset > next_cu_offset) { + fprintf (stderr, "warning: DWARF compile unit extends beyond its bounds cu 0x%8.8x at 0x%8.8x'\n", getOffset(), offset); + } + + setDIERelations(); + return DieArray.size(); +} + +void DWARFCompileUnit::clearDIEs(bool keep_compile_unit_die) { + if (DieArray.size() > 1) { + // std::vectors never get any smaller when resized to a smaller size, + // or when clear() or erase() are called, the size will report that it + // is smaller, but the memory allocated remains intact (call capacity() + // to see this). So we need to create a temporary vector and swap the + // contents which will cause just the internal pointers to be swapped + // so that when "tmp_array" goes out of scope, it will destroy the + // contents. + + // Save at least the compile unit DIE + std::vector<DWARFDebugInfoEntryMinimal> tmpArray; + DieArray.swap(tmpArray); + if (keep_compile_unit_die) + DieArray.push_back(tmpArray.front()); + } +} + +void +DWARFCompileUnit::buildAddressRangeTable(DWARFDebugAranges *debug_aranges, + bool clear_dies_if_already_not_parsed){ + // This function is usually called if there in no .debug_aranges section + // in order to produce a compile unit level set of address ranges that + // is accurate. If the DIEs weren't parsed, then we don't want all dies for + // all compile units to stay loaded when they weren't needed. So we can end + // up parsing the DWARF and then throwing them all away to keep memory usage + // down. + const bool clear_dies = extractDIEsIfNeeded(false) > 1; + + DieArray[0].buildAddressRangeTable(this, debug_aranges); + + // Keep memory down by clearing DIEs if this generate function + // caused them to be parsed. + if (clear_dies) + clearDIEs(true); +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFCompileUnit.h b/contrib/llvm/lib/DebugInfo/DWARFCompileUnit.h new file mode 100644 index 0000000..d916729 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFCompileUnit.h @@ -0,0 +1,111 @@ +//===-- DWARFCompileUnit.h --------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFCOMPILEUNIT_H +#define LLVM_DEBUGINFO_DWARFCOMPILEUNIT_H + +#include "DWARFDebugAbbrev.h" +#include "DWARFDebugInfoEntry.h" +#include <vector> + +namespace llvm { + +class DWARFContext; +class raw_ostream; + +class DWARFCompileUnit { + DWARFContext &Context; + + uint32_t Offset; + uint32_t Length; + uint16_t Version; + const DWARFAbbreviationDeclarationSet *Abbrevs; + uint8_t AddrSize; + uint64_t BaseAddr; + // The compile unit debug information entry item. + std::vector<DWARFDebugInfoEntryMinimal> DieArray; +public: + DWARFCompileUnit(DWARFContext &context) : Context(context) { + clear(); + } + + DWARFContext &getContext() const { return Context; } + DataExtractor getDebugInfoExtractor() const; + + bool extract(DataExtractor debug_info, uint32_t* offset_ptr); + uint32_t extract(uint32_t offset, DataExtractor debug_info_data, + const DWARFAbbreviationDeclarationSet *abbrevs); + + /// extractDIEsIfNeeded - Parses a compile unit and indexes its DIEs if it + /// hasn't already been done. + size_t extractDIEsIfNeeded(bool cu_die_only); + void clear(); + void dump(raw_ostream &OS); + uint32_t getOffset() const { return Offset; } + /// Size in bytes of the compile unit header. + uint32_t getSize() const { return 11; } + bool containsDIEOffset(uint32_t die_offset) const { + return die_offset >= getFirstDIEOffset() && + die_offset < getNextCompileUnitOffset(); + } + uint32_t getFirstDIEOffset() const { return Offset + getSize(); } + uint32_t getNextCompileUnitOffset() const { return Offset + Length + 4; } + /// Size in bytes of the .debug_info data associated with this compile unit. + size_t getDebugInfoSize() const { return Length + 4 - getSize(); } + uint32_t getLength() const { return Length; } + uint16_t getVersion() const { return Version; } + const DWARFAbbreviationDeclarationSet *getAbbreviations() const { + return Abbrevs; + } + uint8_t getAddressByteSize() const { return AddrSize; } + uint64_t getBaseAddress() const { return BaseAddr; } + + void setBaseAddress(uint64_t base_addr) { + BaseAddr = base_addr; + } + + const DWARFDebugInfoEntryMinimal * + getCompileUnitDIE(bool extract_cu_die_only = true) { + extractDIEsIfNeeded(extract_cu_die_only); + if (DieArray.empty()) + return NULL; + return &DieArray[0]; + } + + /// setDIERelations - We read in all of the DIE entries into our flat list + /// of DIE entries and now we need to go back through all of them and set the + /// parent, sibling and child pointers for quick DIE navigation. + void setDIERelations(); + + void addDIE(DWARFDebugInfoEntryMinimal &die) { + // The average bytes per DIE entry has been seen to be + // around 14-20 so lets pre-reserve the needed memory for + // our DIE entries accordingly. Search forward for "Compute + // average bytes per DIE" to see #if'ed out code that does + // that determination. + + // Only reserve the memory if we are adding children of + // the main compile unit DIE. The compile unit DIE is always + // the first entry, so if our size is 1, then we are adding + // the first compile unit child DIE and should reserve + // the memory. + if (DieArray.empty()) + DieArray.reserve(getDebugInfoSize() / 14); + DieArray.push_back(die); + } + + void clearDIEs(bool keep_compile_unit_die); + + void buildAddressRangeTable(DWARFDebugAranges *debug_aranges, + bool clear_dies_if_already_not_parsed); +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFContext.cpp b/contrib/llvm/lib/DebugInfo/DWARFContext.cpp new file mode 100644 index 0000000..e1ac398 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFContext.cpp @@ -0,0 +1,167 @@ +//===-- DWARFContext.cpp --------------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFContext.h" +#include "llvm/Support/Dwarf.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +#include <algorithm> +using namespace llvm; +using namespace dwarf; + +void DWARFContext::dump(raw_ostream &OS) { + OS << ".debug_abbrev contents:\n"; + getDebugAbbrev()->dump(OS); + + OS << "\n.debug_info contents:\n"; + for (unsigned i = 0, e = getNumCompileUnits(); i != e; ++i) + getCompileUnitAtIndex(i)->dump(OS); + + OS << "\n.debug_aranges contents:\n"; + DataExtractor arangesData(getARangeSection(), isLittleEndian(), 0); + uint32_t offset = 0; + DWARFDebugArangeSet set; + while (set.extract(arangesData, &offset)) + set.dump(OS); + + OS << "\n.debug_lines contents:\n"; + for (unsigned i = 0, e = getNumCompileUnits(); i != e; ++i) { + DWARFCompileUnit *cu = getCompileUnitAtIndex(i); + unsigned stmtOffset = + cu->getCompileUnitDIE()->getAttributeValueAsUnsigned(cu, DW_AT_stmt_list, + -1U); + if (stmtOffset != -1U) { + DataExtractor lineData(getLineSection(), isLittleEndian(), + cu->getAddressByteSize()); + DWARFDebugLine::DumpingState state(OS); + DWARFDebugLine::parseStatementTable(lineData, &stmtOffset, state); + } + } + + OS << "\n.debug_str contents:\n"; + DataExtractor strData(getStringSection(), isLittleEndian(), 0); + offset = 0; + uint32_t lastOffset = 0; + while (const char *s = strData.getCStr(&offset)) { + OS << format("0x%8.8x: \"%s\"\n", lastOffset, s); + lastOffset = offset; + } +} + +const DWARFDebugAbbrev *DWARFContext::getDebugAbbrev() { + if (Abbrev) + return Abbrev.get(); + + DataExtractor abbrData(getAbbrevSection(), isLittleEndian(), 0); + + Abbrev.reset(new DWARFDebugAbbrev()); + Abbrev->parse(abbrData); + return Abbrev.get(); +} + +const DWARFDebugAranges *DWARFContext::getDebugAranges() { + if (Aranges) + return Aranges.get(); + + DataExtractor arangesData(getARangeSection(), isLittleEndian(), 0); + + Aranges.reset(new DWARFDebugAranges()); + Aranges->extract(arangesData); + if (Aranges->isEmpty()) // No aranges in file, generate them from the DIEs. + Aranges->generate(this); + return Aranges.get(); +} + +const DWARFDebugLine::LineTable * +DWARFContext::getLineTableForCompileUnit(DWARFCompileUnit *cu) { + if (!Line) + Line.reset(new DWARFDebugLine()); + + unsigned stmtOffset = + cu->getCompileUnitDIE()->getAttributeValueAsUnsigned(cu, DW_AT_stmt_list, + -1U); + if (stmtOffset == -1U) + return 0; // No line table for this compile unit. + + // See if the line table is cached. + if (const DWARFDebugLine::LineTable *lt = Line->getLineTable(stmtOffset)) + return lt; + + // We have to parse it first. + DataExtractor lineData(getLineSection(), isLittleEndian(), + cu->getAddressByteSize()); + return Line->getOrParseLineTable(lineData, stmtOffset); +} + +void DWARFContext::parseCompileUnits() { + uint32_t offset = 0; + const DataExtractor &debug_info_data = DataExtractor(getInfoSection(), + isLittleEndian(), 0); + while (debug_info_data.isValidOffset(offset)) { + CUs.push_back(DWARFCompileUnit(*this)); + if (!CUs.back().extract(debug_info_data, &offset)) { + CUs.pop_back(); + break; + } + + offset = CUs.back().getNextCompileUnitOffset(); + } +} + +namespace { + struct OffsetComparator { + bool operator()(const DWARFCompileUnit &LHS, + const DWARFCompileUnit &RHS) const { + return LHS.getOffset() < RHS.getOffset(); + } + bool operator()(const DWARFCompileUnit &LHS, uint32_t RHS) const { + return LHS.getOffset() < RHS; + } + bool operator()(uint32_t LHS, const DWARFCompileUnit &RHS) const { + return LHS < RHS.getOffset(); + } + }; +} + +DWARFCompileUnit *DWARFContext::getCompileUnitForOffset(uint32_t offset) { + if (CUs.empty()) + parseCompileUnits(); + + DWARFCompileUnit *i = std::lower_bound(CUs.begin(), CUs.end(), offset, + OffsetComparator()); + if (i != CUs.end()) + return &*i; + return 0; +} + +DILineInfo DWARFContext::getLineInfoForAddress(uint64_t address) { + // First, get the offset of the compile unit. + uint32_t cuOffset = getDebugAranges()->findAddress(address); + // Retrieve the compile unit. + DWARFCompileUnit *cu = getCompileUnitForOffset(cuOffset); + if (!cu) + return DILineInfo("<invalid>", 0, 0); + // Get the line table for this compile unit. + const DWARFDebugLine::LineTable *lineTable = getLineTableForCompileUnit(cu); + if (!lineTable) + return DILineInfo("<invalid>", 0, 0); + // Get the index of the row we're looking for in the line table. + uint64_t hiPC = + cu->getCompileUnitDIE()->getAttributeValueAsUnsigned(cu, DW_AT_high_pc, + -1ULL); + uint32_t rowIndex = lineTable->lookupAddress(address, hiPC); + if (rowIndex == -1U) + return DILineInfo("<invalid>", 0, 0); + + // From here, contruct the DILineInfo. + const DWARFDebugLine::Row &row = lineTable->Rows[rowIndex]; + const std::string &fileName = lineTable->Prologue.FileNames[row.File-1].Name; + + return DILineInfo(fileName.c_str(), row.Line, row.Column); +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFContext.h b/contrib/llvm/lib/DebugInfo/DWARFContext.h new file mode 100644 index 0000000..746a463 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFContext.h @@ -0,0 +1,118 @@ +//===-- DWARFContext.h ------------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===/ + +#ifndef LLVM_DEBUGINFO_DWARFCONTEXT_H +#define LLVM_DEBUGINFO_DWARFCONTEXT_H + +#include "DWARFCompileUnit.h" +#include "DWARFDebugAranges.h" +#include "DWARFDebugLine.h" +#include "llvm/DebugInfo/DIContext.h" +#include "llvm/ADT/OwningPtr.h" +#include "llvm/ADT/SmallVector.h" + +namespace llvm { + +/// DWARFContext +/// This data structure is the top level entity that deals with dwarf debug +/// information parsing. The actual data is supplied through pure virtual +/// methods that a concrete implementation provides. +class DWARFContext : public DIContext { + bool IsLittleEndian; + + SmallVector<DWARFCompileUnit, 1> CUs; + OwningPtr<DWARFDebugAbbrev> Abbrev; + OwningPtr<DWARFDebugAranges> Aranges; + OwningPtr<DWARFDebugLine> Line; + + DWARFContext(DWARFContext &); // = delete + DWARFContext &operator=(DWARFContext &); // = delete + + /// Read compile units from the debug_info section and store them in CUs. + void parseCompileUnits(); +protected: + DWARFContext(bool isLittleEndian) : IsLittleEndian(isLittleEndian) {} +public: + virtual void dump(raw_ostream &OS); + /// Get the number of compile units in this context. + unsigned getNumCompileUnits() { + if (CUs.empty()) + parseCompileUnits(); + return CUs.size(); + } + /// Get the compile unit at the specified index for this compile unit. + DWARFCompileUnit *getCompileUnitAtIndex(unsigned index) { + if (CUs.empty()) + parseCompileUnits(); + return &CUs[index]; + } + + /// Return the compile unit that includes an offset (relative to .debug_info). + DWARFCompileUnit *getCompileUnitForOffset(uint32_t offset); + + /// Get a pointer to the parsed DebugAbbrev object. + const DWARFDebugAbbrev *getDebugAbbrev(); + + /// Get a pointer to the parsed DebugAranges object. + const DWARFDebugAranges *getDebugAranges(); + + /// Get a pointer to a parsed line table corresponding to a compile unit. + const DWARFDebugLine::LineTable * + getLineTableForCompileUnit(DWARFCompileUnit *cu); + + virtual DILineInfo getLineInfoForAddress(uint64_t address); + + bool isLittleEndian() const { return IsLittleEndian; } + + virtual StringRef getInfoSection() = 0; + virtual StringRef getAbbrevSection() = 0; + virtual StringRef getARangeSection() = 0; + virtual StringRef getLineSection() = 0; + virtual StringRef getStringSection() = 0; + + static bool isSupportedVersion(unsigned version) { + return version == 2 || version == 3; + } +}; + + +/// DWARFContextInMemory is the simplest possible implementation of a +/// DWARFContext. It assumes all content is available in memory and stores +/// pointers to it. +class DWARFContextInMemory : public DWARFContext { + StringRef InfoSection; + StringRef AbbrevSection; + StringRef ARangeSection; + StringRef LineSection; + StringRef StringSection; +public: + DWARFContextInMemory(bool isLittleEndian, + StringRef infoSection, + StringRef abbrevSection, + StringRef aRangeSection, + StringRef lineSection, + StringRef stringSection) + : DWARFContext(isLittleEndian), + InfoSection(infoSection), + AbbrevSection(abbrevSection), + ARangeSection(aRangeSection), + LineSection(lineSection), + StringSection(stringSection) + {} + + virtual StringRef getInfoSection() { return InfoSection; } + virtual StringRef getAbbrevSection() { return AbbrevSection; } + virtual StringRef getARangeSection() { return ARangeSection; } + virtual StringRef getLineSection() { return LineSection; } + virtual StringRef getStringSection() { return StringSection; } +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugAbbrev.cpp b/contrib/llvm/lib/DebugInfo/DWARFDebugAbbrev.cpp new file mode 100644 index 0000000..a11ae3f --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugAbbrev.cpp @@ -0,0 +1,106 @@ +//===-- DWARFDebugAbbrev.cpp ----------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFDebugAbbrev.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +using namespace llvm; + +bool DWARFAbbreviationDeclarationSet::extract(DataExtractor data, + uint32_t* offset_ptr) { + const uint32_t beginOffset = *offset_ptr; + Offset = beginOffset; + clear(); + DWARFAbbreviationDeclaration abbrevDeclaration; + uint32_t prevAbbrAode = 0; + while (abbrevDeclaration.extract(data, offset_ptr)) { + Decls.push_back(abbrevDeclaration); + if (IdxOffset == 0) { + IdxOffset = abbrevDeclaration.getCode(); + } else { + if (prevAbbrAode + 1 != abbrevDeclaration.getCode()) + IdxOffset = UINT32_MAX;// Out of order indexes, we can't do O(1) lookups + } + prevAbbrAode = abbrevDeclaration.getCode(); + } + return beginOffset != *offset_ptr; +} + +void DWARFAbbreviationDeclarationSet::dump(raw_ostream &OS) const { + for (unsigned i = 0, e = Decls.size(); i != e; ++i) + Decls[i].dump(OS); +} + +const DWARFAbbreviationDeclaration* +DWARFAbbreviationDeclarationSet::getAbbreviationDeclaration(uint32_t abbrCode) + const { + if (IdxOffset == UINT32_MAX) { + DWARFAbbreviationDeclarationCollConstIter pos; + DWARFAbbreviationDeclarationCollConstIter end = Decls.end(); + for (pos = Decls.begin(); pos != end; ++pos) { + if (pos->getCode() == abbrCode) + return &(*pos); + } + } else { + uint32_t idx = abbrCode - IdxOffset; + if (idx < Decls.size()) + return &Decls[idx]; + } + return NULL; +} + +DWARFDebugAbbrev::DWARFDebugAbbrev() : + AbbrevCollMap(), + PrevAbbrOffsetPos(AbbrevCollMap.end()) {} + + +void DWARFDebugAbbrev::parse(DataExtractor data) { + uint32_t offset = 0; + + while (data.isValidOffset(offset)) { + uint32_t initial_cu_offset = offset; + DWARFAbbreviationDeclarationSet abbrevDeclSet; + + if (abbrevDeclSet.extract(data, &offset)) + AbbrevCollMap[initial_cu_offset] = abbrevDeclSet; + else + break; + } + PrevAbbrOffsetPos = AbbrevCollMap.end(); +} + +void DWARFDebugAbbrev::dump(raw_ostream &OS) const { + if (AbbrevCollMap.empty()) { + OS << "< EMPTY >\n"; + return; + } + + DWARFAbbreviationDeclarationCollMapConstIter pos; + for (pos = AbbrevCollMap.begin(); pos != AbbrevCollMap.end(); ++pos) { + OS << format("Abbrev table for offset: 0x%8.8x\n", pos->first); + pos->second.dump(OS); + } +} + +const DWARFAbbreviationDeclarationSet* +DWARFDebugAbbrev::getAbbreviationDeclarationSet(uint64_t cu_abbr_offset) const { + DWARFAbbreviationDeclarationCollMapConstIter end = AbbrevCollMap.end(); + DWARFAbbreviationDeclarationCollMapConstIter pos; + if (PrevAbbrOffsetPos != end && + PrevAbbrOffsetPos->first == cu_abbr_offset) { + return &(PrevAbbrOffsetPos->second); + } else { + pos = AbbrevCollMap.find(cu_abbr_offset); + PrevAbbrOffsetPos = pos; + } + + if (pos != AbbrevCollMap.end()) + return &(pos->second); + return NULL; +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugAbbrev.h b/contrib/llvm/lib/DebugInfo/DWARFDebugAbbrev.h new file mode 100644 index 0000000..03189b1 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugAbbrev.h @@ -0,0 +1,73 @@ +//===-- DWARFDebugAbbrev.h --------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFDEBUGABBREV_H +#define LLVM_DEBUGINFO_DWARFDEBUGABBREV_H + +#include "DWARFAbbreviationDeclaration.h" +#include <list> +#include <map> +#include <vector> + +namespace llvm { + +typedef std::vector<DWARFAbbreviationDeclaration> + DWARFAbbreviationDeclarationColl; +typedef DWARFAbbreviationDeclarationColl::iterator + DWARFAbbreviationDeclarationCollIter; +typedef DWARFAbbreviationDeclarationColl::const_iterator + DWARFAbbreviationDeclarationCollConstIter; + +class DWARFAbbreviationDeclarationSet { + uint64_t Offset; + uint32_t IdxOffset; + std::vector<DWARFAbbreviationDeclaration> Decls; + public: + DWARFAbbreviationDeclarationSet() + : Offset(0), IdxOffset(0) {} + + DWARFAbbreviationDeclarationSet(uint64_t offset, uint32_t idxOffset) + : Offset(offset), IdxOffset(idxOffset) {} + + void clear() { + IdxOffset = 0; + Decls.clear(); + } + uint64_t getOffset() const { return Offset; } + void dump(raw_ostream &OS) const; + bool extract(DataExtractor data, uint32_t* offset_ptr); + + const DWARFAbbreviationDeclaration * + getAbbreviationDeclaration(uint32_t abbrCode) const; +}; + +class DWARFDebugAbbrev { +public: + typedef std::map<uint64_t, DWARFAbbreviationDeclarationSet> + DWARFAbbreviationDeclarationCollMap; + typedef DWARFAbbreviationDeclarationCollMap::iterator + DWARFAbbreviationDeclarationCollMapIter; + typedef DWARFAbbreviationDeclarationCollMap::const_iterator + DWARFAbbreviationDeclarationCollMapConstIter; + +private: + DWARFAbbreviationDeclarationCollMap AbbrevCollMap; + mutable DWARFAbbreviationDeclarationCollMapConstIter PrevAbbrOffsetPos; + +public: + DWARFDebugAbbrev(); + const DWARFAbbreviationDeclarationSet * + getAbbreviationDeclarationSet(uint64_t cu_abbr_offset) const; + void dump(raw_ostream &OS) const; + void parse(DataExtractor data); +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugArangeSet.cpp b/contrib/llvm/lib/DebugInfo/DWARFDebugArangeSet.cpp new file mode 100644 index 0000000..b0c0354 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugArangeSet.cpp @@ -0,0 +1,150 @@ +//===-- DWARFDebugArangeSet.cpp -------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFDebugArangeSet.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +#include <algorithm> +#include <cassert> +using namespace llvm; + +void DWARFDebugArangeSet::clear() { + Offset = -1U; + std::memset(&Header, 0, sizeof(Header)); + ArangeDescriptors.clear(); +} + +void DWARFDebugArangeSet::compact() { + if (ArangeDescriptors.empty()) + return; + + // Iterate through all arange descriptors and combine any ranges that + // overlap or have matching boundaries. The ArangeDescriptors are assumed + // to be in ascending order. + uint32_t i = 0; + while (i + 1 < ArangeDescriptors.size()) { + if (ArangeDescriptors[i].getEndAddress() >= ArangeDescriptors[i+1].Address){ + // The current range ends at or exceeds the start of the next address + // range. Compute the max end address between the two and use that to + // make the new length. + const uint64_t max_end_addr = + std::max(ArangeDescriptors[i].getEndAddress(), + ArangeDescriptors[i+1].getEndAddress()); + ArangeDescriptors[i].Length = max_end_addr - ArangeDescriptors[i].Address; + // Now remove the next entry as it was just combined with the previous one + ArangeDescriptors.erase(ArangeDescriptors.begin()+i+1); + } else { + // Discontiguous address range, just proceed to the next one. + ++i; + } + } +} + +bool +DWARFDebugArangeSet::extract(DataExtractor data, uint32_t *offset_ptr) { + if (data.isValidOffset(*offset_ptr)) { + ArangeDescriptors.clear(); + Offset = *offset_ptr; + + // 7.20 Address Range Table + // + // Each set of entries in the table of address ranges contained in + // the .debug_aranges section begins with a header consisting of: a + // 4-byte length containing the length of the set of entries for this + // compilation unit, not including the length field itself; a 2-byte + // version identifier containing the value 2 for DWARF Version 2; a + // 4-byte offset into the.debug_infosection; a 1-byte unsigned integer + // containing the size in bytes of an address (or the offset portion of + // an address for segmented addressing) on the target system; and a + // 1-byte unsigned integer containing the size in bytes of a segment + // descriptor on the target system. This header is followed by a series + // of tuples. Each tuple consists of an address and a length, each in + // the size appropriate for an address on the target architecture. + Header.Length = data.getU32(offset_ptr); + Header.Version = data.getU16(offset_ptr); + Header.CuOffset = data.getU32(offset_ptr); + Header.AddrSize = data.getU8(offset_ptr); + Header.SegSize = data.getU8(offset_ptr); + + // Perform basic validation of the header fields. + if (!data.isValidOffsetForDataOfSize(Offset, Header.Length) || + (Header.AddrSize != 4 && Header.AddrSize != 8)) { + clear(); + return false; + } + + // The first tuple following the header in each set begins at an offset + // that is a multiple of the size of a single tuple (that is, twice the + // size of an address). The header is padded, if necessary, to the + // appropriate boundary. + const uint32_t header_size = *offset_ptr - Offset; + const uint32_t tuple_size = Header.AddrSize * 2; + uint32_t first_tuple_offset = 0; + while (first_tuple_offset < header_size) + first_tuple_offset += tuple_size; + + *offset_ptr = Offset + first_tuple_offset; + + Descriptor arangeDescriptor; + + assert(sizeof(arangeDescriptor.Address) == sizeof(arangeDescriptor.Length)); + assert(sizeof(arangeDescriptor.Address) >= Header.AddrSize); + + while (data.isValidOffset(*offset_ptr)) { + arangeDescriptor.Address = data.getUnsigned(offset_ptr, Header.AddrSize); + arangeDescriptor.Length = data.getUnsigned(offset_ptr, Header.AddrSize); + + // Each set of tuples is terminated by a 0 for the address and 0 + // for the length. + if (arangeDescriptor.Address || arangeDescriptor.Length) + ArangeDescriptors.push_back(arangeDescriptor); + else + break; // We are done if we get a zero address and length + } + + return !ArangeDescriptors.empty(); + } + return false; +} + +void DWARFDebugArangeSet::dump(raw_ostream &OS) const { + OS << format("Address Range Header: length = 0x%8.8x, version = 0x%4.4x, ", + Header.Length, Header.Version) + << format("cu_offset = 0x%8.8x, addr_size = 0x%2.2x, seg_size = 0x%2.2x\n", + Header.CuOffset, Header.AddrSize, Header.SegSize); + + const uint32_t hex_width = Header.AddrSize * 2; + for (DescriptorConstIter pos = ArangeDescriptors.begin(), + end = ArangeDescriptors.end(); pos != end; ++pos) + OS << format("[0x%*.*llx -", hex_width, hex_width, pos->Address) + << format(" 0x%*.*llx)\n", hex_width, hex_width, pos->getEndAddress()); +} + + +namespace { + class DescriptorContainsAddress { + const uint64_t Address; + public: + DescriptorContainsAddress(uint64_t address) : Address(address) {} + bool operator()(const DWARFDebugArangeSet::Descriptor &desc) const { + return Address >= desc.Address && Address < (desc.Address + desc.Length); + } + }; +} + +uint32_t DWARFDebugArangeSet::findAddress(uint64_t address) const { + DescriptorConstIter end = ArangeDescriptors.end(); + DescriptorConstIter pos = + std::find_if(ArangeDescriptors.begin(), end, // Range + DescriptorContainsAddress(address)); // Predicate + if (pos != end) + return Header.CuOffset; + + return -1U; +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugArangeSet.h b/contrib/llvm/lib/DebugInfo/DWARFDebugArangeSet.h new file mode 100644 index 0000000..9a2a6d0 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugArangeSet.h @@ -0,0 +1,75 @@ +//===-- DWARFDebugArangeSet.h -----------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFDEBUGARANGESET_H +#define LLVM_DEBUGINFO_DWARFDEBUGARANGESET_H + +#include "llvm/Support/DataExtractor.h" +#include <vector> + +namespace llvm { + +class raw_ostream; + +class DWARFDebugArangeSet { +public: + struct Header { + // The total length of the entries for that set, not including the length + // field itself. + uint32_t Length; + // The offset from the beginning of the .debug_info section of the + // compilation unit entry referenced by the table. + uint32_t CuOffset; + // The DWARF version number. + uint16_t Version; + // The size in bytes of an address on the target architecture. For segmented + // addressing, this is the size of the offset portion of the address. + uint8_t AddrSize; + // The size in bytes of a segment descriptor on the target architecture. + // If the target system uses a flat address space, this value is 0. + uint8_t SegSize; + }; + + struct Descriptor { + uint64_t Address; + uint64_t Length; + uint64_t getEndAddress() const { return Address + Length; } + }; + +private: + typedef std::vector<Descriptor> DescriptorColl; + typedef DescriptorColl::iterator DescriptorIter; + typedef DescriptorColl::const_iterator DescriptorConstIter; + + uint32_t Offset; + Header Header; + DescriptorColl ArangeDescriptors; + +public: + DWARFDebugArangeSet() { clear(); } + void clear(); + void compact(); + bool extract(DataExtractor data, uint32_t *offset_ptr); + void dump(raw_ostream &OS) const; + + uint32_t getCompileUnitDIEOffset() const { return Header.CuOffset; } + uint32_t getOffsetOfNextEntry() const { return Offset + Header.Length + 4; } + uint32_t findAddress(uint64_t address) const; + uint32_t getNumDescriptors() const { return ArangeDescriptors.size(); } + const struct Header &getHeader() const { return Header; } + const Descriptor *getDescriptor(uint32_t i) const { + if (i < ArangeDescriptors.size()) + return &ArangeDescriptors[i]; + return NULL; + } +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugAranges.cpp b/contrib/llvm/lib/DebugInfo/DWARFDebugAranges.cpp new file mode 100644 index 0000000..576d37d --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugAranges.cpp @@ -0,0 +1,223 @@ +//===-- DWARFDebugAranges.cpp -----------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFDebugAranges.h" +#include "DWARFCompileUnit.h" +#include "DWARFContext.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +#include <algorithm> +#include <cassert> +using namespace llvm; + +// Compare function DWARFDebugAranges::Range structures +static bool RangeLessThan(const DWARFDebugAranges::Range &range1, + const DWARFDebugAranges::Range &range2) { + return range1.LoPC < range2.LoPC; +} + +namespace { + class CountArangeDescriptors { + public: + CountArangeDescriptors(uint32_t &count_ref) : Count(count_ref) {} + void operator()(const DWARFDebugArangeSet &set) { + Count += set.getNumDescriptors(); + } + uint32_t &Count; + }; + + class AddArangeDescriptors { + public: + AddArangeDescriptors(DWARFDebugAranges::RangeColl &ranges) + : RangeCollection(ranges) {} + void operator()(const DWARFDebugArangeSet& set) { + const DWARFDebugArangeSet::Descriptor* arange_desc_ptr; + DWARFDebugAranges::Range range; + range.Offset = set.getCompileUnitDIEOffset(); + + for (uint32_t i=0; (arange_desc_ptr = set.getDescriptor(i)) != NULL; ++i){ + range.LoPC = arange_desc_ptr->Address; + range.Length = arange_desc_ptr->Length; + + // Insert each item in increasing address order so binary searching + // can later be done! + DWARFDebugAranges::RangeColl::iterator insert_pos = + std::lower_bound(RangeCollection.begin(), RangeCollection.end(), + range, RangeLessThan); + RangeCollection.insert(insert_pos, range); + } + } + DWARFDebugAranges::RangeColl& RangeCollection; + }; +} + +bool DWARFDebugAranges::extract(DataExtractor debug_aranges_data) { + if (debug_aranges_data.isValidOffset(0)) { + uint32_t offset = 0; + + typedef std::vector<DWARFDebugArangeSet> SetCollection; + typedef SetCollection::const_iterator SetCollectionIter; + SetCollection sets; + + DWARFDebugArangeSet set; + Range range; + while (set.extract(debug_aranges_data, &offset)) + sets.push_back(set); + + uint32_t count = 0; + + std::for_each(sets.begin(), sets.end(), CountArangeDescriptors(count)); + + if (count > 0) { + Aranges.reserve(count); + AddArangeDescriptors range_adder(Aranges); + std::for_each(sets.begin(), sets.end(), range_adder); + } + } + return false; +} + +bool DWARFDebugAranges::generate(DWARFContext *ctx) { + clear(); + if (ctx) { + const uint32_t num_compile_units = ctx->getNumCompileUnits(); + for (uint32_t cu_idx = 0; cu_idx < num_compile_units; ++cu_idx) { + DWARFCompileUnit *cu = ctx->getCompileUnitAtIndex(cu_idx); + if (cu) + cu->buildAddressRangeTable(this, true); + } + } + return !isEmpty(); +} + +void DWARFDebugAranges::dump(raw_ostream &OS) const { + const uint32_t num_ranges = getNumRanges(); + for (uint32_t i = 0; i < num_ranges; ++i) { + const Range &range = Aranges[i]; + OS << format("0x%8.8x: [0x%8.8llx - 0x%8.8llx)\n", range.Offset, + (uint64_t)range.LoPC, (uint64_t)range.HiPC()); + } +} + +void DWARFDebugAranges::Range::dump(raw_ostream &OS) const { + OS << format("{0x%8.8x}: [0x%8.8llx - 0x%8.8llx)\n", Offset, LoPC, HiPC()); +} + +void DWARFDebugAranges::appendRange(uint32_t offset, uint64_t low_pc, + uint64_t high_pc) { + if (!Aranges.empty()) { + if (Aranges.back().Offset == offset && Aranges.back().HiPC() == low_pc) { + Aranges.back().setHiPC(high_pc); + return; + } + } + Aranges.push_back(Range(low_pc, high_pc, offset)); +} + +void DWARFDebugAranges::sort(bool minimize, uint32_t n) { + const size_t orig_arange_size = Aranges.size(); + // Size of one? If so, no sorting is needed + if (orig_arange_size <= 1) + return; + // Sort our address range entries + std::stable_sort(Aranges.begin(), Aranges.end(), RangeLessThan); + + if (!minimize) + return; + + // Most address ranges are contiguous from function to function + // so our new ranges will likely be smaller. We calculate the size + // of the new ranges since although std::vector objects can be resized, + // the will never reduce their allocated block size and free any excesss + // memory, so we might as well start a brand new collection so it is as + // small as possible. + + // First calculate the size of the new minimal arange vector + // so we don't have to do a bunch of re-allocations as we + // copy the new minimal stuff over to the new collection. + size_t minimal_size = 1; + for (size_t i = 1; i < orig_arange_size; ++i) { + if (!Range::SortedOverlapCheck(Aranges[i-1], Aranges[i], n)) + ++minimal_size; + } + + // If the sizes are the same, then no consecutive aranges can be + // combined, we are done. + if (minimal_size == orig_arange_size) + return; + + // Else, make a new RangeColl that _only_ contains what we need. + RangeColl minimal_aranges; + minimal_aranges.resize(minimal_size); + uint32_t j = 0; + minimal_aranges[j] = Aranges[0]; + for (size_t i = 1; i < orig_arange_size; ++i) { + if(Range::SortedOverlapCheck (minimal_aranges[j], Aranges[i], n)) { + minimal_aranges[j].setHiPC (Aranges[i].HiPC()); + } else { + // Only increment j if we aren't merging + minimal_aranges[++j] = Aranges[i]; + } + } + assert (j+1 == minimal_size); + + // Now swap our new minimal aranges into place. The local + // minimal_aranges will then contian the old big collection + // which will get freed. + minimal_aranges.swap(Aranges); +} + +uint32_t DWARFDebugAranges::findAddress(uint64_t address) const { + if (!Aranges.empty()) { + Range range(address); + RangeCollIterator begin = Aranges.begin(); + RangeCollIterator end = Aranges.end(); + RangeCollIterator pos = lower_bound(begin, end, range, RangeLessThan); + + if (pos != end && pos->LoPC <= address && address < pos->HiPC()) { + return pos->Offset; + } else if (pos != begin) { + --pos; + if (pos->LoPC <= address && address < pos->HiPC()) + return (*pos).Offset; + } + } + return -1U; +} + +bool +DWARFDebugAranges::allRangesAreContiguous(uint64_t &LoPC, uint64_t &HiPC) const{ + if (Aranges.empty()) + return false; + + uint64_t next_addr = 0; + RangeCollIterator begin = Aranges.begin(); + for (RangeCollIterator pos = begin, end = Aranges.end(); pos != end; + ++pos) { + if (pos != begin && pos->LoPC != next_addr) + return false; + next_addr = pos->HiPC(); + } + // We checked for empty at the start of function so front() will be valid. + LoPC = Aranges.front().LoPC; + // We checked for empty at the start of function so back() will be valid. + HiPC = Aranges.back().HiPC(); + return true; +} + +bool DWARFDebugAranges::getMaxRange(uint64_t &LoPC, uint64_t &HiPC) const { + if (Aranges.empty()) + return false; + // We checked for empty at the start of function so front() will be valid. + LoPC = Aranges.front().LoPC; + // We checked for empty at the start of function so back() will be valid. + HiPC = Aranges.back().HiPC(); + return true; +} + diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugAranges.h b/contrib/llvm/lib/DebugInfo/DWARFDebugAranges.h new file mode 100644 index 0000000..12afb60 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugAranges.h @@ -0,0 +1,98 @@ +//===-- DWARFDebugAranges.h -------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFDEBUGARANGES_H +#define LLVM_DEBUGINFO_DWARFDEBUGARANGES_H + +#include "DWARFDebugArangeSet.h" +#include <list> + +namespace llvm { + +class DWARFContext; + +class DWARFDebugAranges { +public: + struct Range { + explicit Range(uint64_t lo = -1ULL, uint64_t hi = -1ULL, + uint32_t off = -1U) + : LoPC(lo), Length(hi-lo), Offset(off) {} + + void clear() { + LoPC = -1ULL; + Length = 0; + Offset = -1U; + } + + void setHiPC(uint64_t HiPC) { + if (HiPC == -1ULL || HiPC <= LoPC) + Length = 0; + else + Length = HiPC - LoPC; + } + uint64_t HiPC() const { + if (Length) + return LoPC + Length; + return -1ULL; + } + bool isValidRange() const { return Length > 0; } + + static bool SortedOverlapCheck(const Range &curr_range, + const Range &next_range, uint32_t n) { + if (curr_range.Offset != next_range.Offset) + return false; + return curr_range.HiPC() + n >= next_range.LoPC; + } + + bool contains(const Range &range) const { + return LoPC <= range.LoPC && range.HiPC() <= HiPC(); + } + + void dump(raw_ostream &OS) const; + uint64_t LoPC; // Start of address range + uint32_t Length; // End of address range (not including this address) + uint32_t Offset; // Offset of the compile unit or die + }; + + void clear() { Aranges.clear(); } + bool allRangesAreContiguous(uint64_t& LoPC, uint64_t& HiPC) const; + bool getMaxRange(uint64_t& LoPC, uint64_t& HiPC) const; + bool extract(DataExtractor debug_aranges_data); + bool generate(DWARFContext *ctx); + + // Use append range multiple times and then call sort + void appendRange(uint32_t cu_offset, uint64_t low_pc, uint64_t high_pc); + void sort(bool minimize, uint32_t n); + + const Range *rangeAtIndex(uint32_t idx) const { + if (idx < Aranges.size()) + return &Aranges[idx]; + return NULL; + } + void dump(raw_ostream &OS) const; + uint32_t findAddress(uint64_t address) const; + bool isEmpty() const { return Aranges.empty(); } + uint32_t getNumRanges() const { return Aranges.size(); } + + uint32_t offsetAtIndex(uint32_t idx) const { + if (idx < Aranges.size()) + return Aranges[idx].Offset; + return -1U; + } + + typedef std::vector<Range> RangeColl; + typedef RangeColl::const_iterator RangeCollIterator; + +private: + RangeColl Aranges; +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugInfoEntry.cpp b/contrib/llvm/lib/DebugInfo/DWARFDebugInfoEntry.cpp new file mode 100644 index 0000000..1b089ad --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugInfoEntry.cpp @@ -0,0 +1,444 @@ +//===-- DWARFDebugInfoEntry.cpp --------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFDebugInfoEntry.h" +#include "DWARFCompileUnit.h" +#include "DWARFContext.h" +#include "DWARFDebugAbbrev.h" +#include "DWARFFormValue.h" +#include "llvm/Support/Dwarf.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +using namespace llvm; +using namespace dwarf; + +void DWARFDebugInfoEntryMinimal::dump(raw_ostream &OS, + const DWARFCompileUnit *cu, + unsigned recurseDepth, + unsigned indent) const { + DataExtractor debug_info_data = cu->getDebugInfoExtractor(); + uint32_t offset = Offset; + + if (debug_info_data.isValidOffset(offset)) { + uint64_t abbrCode = debug_info_data.getULEB128(&offset); + + OS << format("\n0x%8.8x: ", Offset); + if (abbrCode) { + if (AbbrevDecl) { + const char *tagString = TagString(getTag()); + if (tagString) + OS.indent(indent) << tagString; + else + OS.indent(indent) << format("DW_TAG_Unknown_%x", getTag()); + OS << format(" [%u] %c\n", abbrCode, + AbbrevDecl->hasChildren() ? '*' : ' '); + + // Dump all data in the .debug_info for the attributes + const uint32_t numAttributes = AbbrevDecl->getNumAttributes(); + for (uint32_t i = 0; i != numAttributes; ++i) { + uint16_t attr = AbbrevDecl->getAttrByIndex(i); + uint16_t form = AbbrevDecl->getFormByIndex(i); + dumpAttribute(OS, cu, &offset, attr, form, indent); + } + + const DWARFDebugInfoEntryMinimal *child = getFirstChild(); + if (recurseDepth > 0 && child) { + while (child) { + child->dump(OS, cu, recurseDepth-1, indent+2); + child = child->getSibling(); + } + } + } else { + OS << "Abbreviation code not found in 'debug_abbrev' class for code: " + << abbrCode << '\n'; + } + } else { + OS.indent(indent) << "NULL\n"; + } + } +} + +void DWARFDebugInfoEntryMinimal::dumpAttribute(raw_ostream &OS, + const DWARFCompileUnit *cu, + uint32_t* offset_ptr, + uint16_t attr, + uint16_t form, + unsigned indent) const { + OS << format("0x%8.8x: ", *offset_ptr); + OS.indent(indent+2); + const char *attrString = AttributeString(attr); + if (attrString) + OS << attrString; + else + OS << format("DW_AT_Unknown_%x", attr); + const char *formString = FormEncodingString(form); + if (formString) + OS << " [" << formString << ']'; + else + OS << format(" [DW_FORM_Unknown_%x]", form); + + DWARFFormValue formValue(form); + + if (!formValue.extractValue(cu->getDebugInfoExtractor(), offset_ptr, cu)) + return; + + OS << "\t("; + formValue.dump(OS, cu); + OS << ")\n"; +} + +bool DWARFDebugInfoEntryMinimal::extractFast(const DWARFCompileUnit *cu, + const uint8_t *fixed_form_sizes, + uint32_t *offset_ptr) { + Offset = *offset_ptr; + + DataExtractor debug_info_data = cu->getDebugInfoExtractor(); + uint64_t abbrCode = debug_info_data.getULEB128(offset_ptr); + + assert (fixed_form_sizes); // For best performance this should be specified! + + if (abbrCode) { + uint32_t offset = *offset_ptr; + + AbbrevDecl = cu->getAbbreviations()->getAbbreviationDeclaration(abbrCode); + + // Skip all data in the .debug_info for the attributes + const uint32_t numAttributes = AbbrevDecl->getNumAttributes(); + uint32_t i; + uint16_t form; + for (i=0; i<numAttributes; ++i) { + form = AbbrevDecl->getFormByIndex(i); + + const uint8_t fixed_skip_size = fixed_form_sizes[form]; + if (fixed_skip_size) + offset += fixed_skip_size; + else { + bool form_is_indirect = false; + do { + form_is_indirect = false; + uint32_t form_size = 0; + switch (form) { + // Blocks if inlined data that have a length field and the data bytes + // inlined in the .debug_info. + case DW_FORM_block: + form_size = debug_info_data.getULEB128(&offset); + break; + case DW_FORM_block1: + form_size = debug_info_data.getU8(&offset); + break; + case DW_FORM_block2: + form_size = debug_info_data.getU16(&offset); + break; + case DW_FORM_block4: + form_size = debug_info_data.getU32(&offset); + break; + + // Inlined NULL terminated C-strings + case DW_FORM_string: + debug_info_data.getCStr(&offset); + break; + + // Compile unit address sized values + case DW_FORM_addr: + case DW_FORM_ref_addr: + form_size = cu->getAddressByteSize(); + break; + + // 1 byte values + case DW_FORM_data1: + case DW_FORM_flag: + case DW_FORM_ref1: + form_size = 1; + break; + + // 2 byte values + case DW_FORM_data2: + case DW_FORM_ref2: + form_size = 2; + break; + + // 4 byte values + case DW_FORM_strp: + case DW_FORM_data4: + case DW_FORM_ref4: + form_size = 4; + break; + + // 8 byte values + case DW_FORM_data8: + case DW_FORM_ref8: + form_size = 8; + break; + + // signed or unsigned LEB 128 values + case DW_FORM_sdata: + case DW_FORM_udata: + case DW_FORM_ref_udata: + debug_info_data.getULEB128(&offset); + break; + + case DW_FORM_indirect: + form_is_indirect = true; + form = debug_info_data.getULEB128(&offset); + break; + + default: + *offset_ptr = Offset; + return false; + } + offset += form_size; + + } while (form_is_indirect); + } + } + *offset_ptr = offset; + return true; + } else { + AbbrevDecl = NULL; + return true; // NULL debug tag entry + } + + return false; +} + +bool +DWARFDebugInfoEntryMinimal::extract(const DWARFCompileUnit *cu, + uint32_t *offset_ptr) { + DataExtractor debug_info_data = cu->getDebugInfoExtractor(); + const uint32_t cu_end_offset = cu->getNextCompileUnitOffset(); + const uint8_t cu_addr_size = cu->getAddressByteSize(); + uint32_t offset = *offset_ptr; + if ((offset < cu_end_offset) && debug_info_data.isValidOffset(offset)) { + Offset = offset; + + uint64_t abbrCode = debug_info_data.getULEB128(&offset); + + if (abbrCode) { + AbbrevDecl = cu->getAbbreviations()->getAbbreviationDeclaration(abbrCode); + + if (AbbrevDecl) { + uint16_t tag = AbbrevDecl->getTag(); + + bool isCompileUnitTag = tag == DW_TAG_compile_unit; + if(cu && isCompileUnitTag) + const_cast<DWARFCompileUnit*>(cu)->setBaseAddress(0); + + // Skip all data in the .debug_info for the attributes + const uint32_t numAttributes = AbbrevDecl->getNumAttributes(); + for (uint32_t i = 0; i != numAttributes; ++i) { + uint16_t attr = AbbrevDecl->getAttrByIndex(i); + uint16_t form = AbbrevDecl->getFormByIndex(i); + + if (isCompileUnitTag && + ((attr == DW_AT_entry_pc) || (attr == DW_AT_low_pc))) { + DWARFFormValue form_value(form); + if (form_value.extractValue(debug_info_data, &offset, cu)) { + if (attr == DW_AT_low_pc || attr == DW_AT_entry_pc) + const_cast<DWARFCompileUnit*>(cu) + ->setBaseAddress(form_value.getUnsigned()); + } + } else { + bool form_is_indirect = false; + do { + form_is_indirect = false; + register uint32_t form_size = 0; + switch (form) { + // Blocks if inlined data that have a length field and the data + // bytes // inlined in the .debug_info + case DW_FORM_block: + form_size = debug_info_data.getULEB128(&offset); + break; + case DW_FORM_block1: + form_size = debug_info_data.getU8(&offset); + break; + case DW_FORM_block2: + form_size = debug_info_data.getU16(&offset); + break; + case DW_FORM_block4: + form_size = debug_info_data.getU32(&offset); + break; + + // Inlined NULL terminated C-strings + case DW_FORM_string: + debug_info_data.getCStr(&offset); + break; + + // Compile unit address sized values + case DW_FORM_addr: + case DW_FORM_ref_addr: + form_size = cu_addr_size; + break; + + // 1 byte values + case DW_FORM_data1: + case DW_FORM_flag: + case DW_FORM_ref1: + form_size = 1; + break; + + // 2 byte values + case DW_FORM_data2: + case DW_FORM_ref2: + form_size = 2; + break; + + // 4 byte values + case DW_FORM_strp: + form_size = 4; + break; + + case DW_FORM_data4: + case DW_FORM_ref4: + form_size = 4; + break; + + // 8 byte values + case DW_FORM_data8: + case DW_FORM_ref8: + form_size = 8; + break; + + // signed or unsigned LEB 128 values + case DW_FORM_sdata: + case DW_FORM_udata: + case DW_FORM_ref_udata: + debug_info_data.getULEB128(&offset); + break; + + case DW_FORM_indirect: + form = debug_info_data.getULEB128(&offset); + form_is_indirect = true; + break; + + default: + *offset_ptr = offset; + return false; + } + + offset += form_size; + } while (form_is_indirect); + } + } + *offset_ptr = offset; + return true; + } + } else { + AbbrevDecl = NULL; + *offset_ptr = offset; + return true; // NULL debug tag entry + } + } + + return false; +} + +uint32_t +DWARFDebugInfoEntryMinimal::getAttributeValue(const DWARFCompileUnit *cu, + const uint16_t attr, + DWARFFormValue &form_value, + uint32_t *end_attr_offset_ptr) + const { + if (AbbrevDecl) { + uint32_t attr_idx = AbbrevDecl->findAttributeIndex(attr); + + if (attr_idx != -1U) { + uint32_t offset = getOffset(); + + DataExtractor debug_info_data = cu->getDebugInfoExtractor(); + + // Skip the abbreviation code so we are at the data for the attributes + debug_info_data.getULEB128(&offset); + + uint32_t idx = 0; + while (idx < attr_idx) + DWARFFormValue::skipValue(AbbrevDecl->getFormByIndex(idx++), + debug_info_data, &offset, cu); + + const uint32_t attr_offset = offset; + form_value = DWARFFormValue(AbbrevDecl->getFormByIndex(idx)); + if (form_value.extractValue(debug_info_data, &offset, cu)) { + if (end_attr_offset_ptr) + *end_attr_offset_ptr = offset; + return attr_offset; + } + } + } + + return 0; +} + +const char* +DWARFDebugInfoEntryMinimal::getAttributeValueAsString( + const DWARFCompileUnit* cu, + const uint16_t attr, + const char* fail_value) const { + DWARFFormValue form_value; + if (getAttributeValue(cu, attr, form_value)) { + DataExtractor stringExtractor(cu->getContext().getStringSection(), + false, 0); + return form_value.getAsCString(&stringExtractor); + } + return fail_value; +} + +uint64_t +DWARFDebugInfoEntryMinimal::getAttributeValueAsUnsigned( + const DWARFCompileUnit* cu, + const uint16_t attr, + uint64_t fail_value) const { + DWARFFormValue form_value; + if (getAttributeValue(cu, attr, form_value)) + return form_value.getUnsigned(); + return fail_value; +} + +int64_t +DWARFDebugInfoEntryMinimal::getAttributeValueAsSigned( + const DWARFCompileUnit* cu, + const uint16_t attr, + int64_t fail_value) const { + DWARFFormValue form_value; + if (getAttributeValue(cu, attr, form_value)) + return form_value.getSigned(); + return fail_value; +} + +uint64_t +DWARFDebugInfoEntryMinimal::getAttributeValueAsReference( + const DWARFCompileUnit* cu, + const uint16_t attr, + uint64_t fail_value) const { + DWARFFormValue form_value; + if (getAttributeValue(cu, attr, form_value)) + return form_value.getReference(cu); + return fail_value; +} + +void +DWARFDebugInfoEntryMinimal::buildAddressRangeTable(const DWARFCompileUnit *cu, + DWARFDebugAranges *debug_aranges) + const { + if (AbbrevDecl) { + uint16_t tag = AbbrevDecl->getTag(); + if (tag == DW_TAG_subprogram) { + uint64_t hi_pc = -1ULL; + uint64_t lo_pc = getAttributeValueAsUnsigned(cu, DW_AT_low_pc, -1ULL); + if (lo_pc != -1ULL) + hi_pc = getAttributeValueAsUnsigned(cu, DW_AT_high_pc, -1ULL); + if (hi_pc != -1ULL) + debug_aranges->appendRange(cu->getOffset(), lo_pc, hi_pc); + } + + const DWARFDebugInfoEntryMinimal *child = getFirstChild(); + while (child) { + child->buildAddressRangeTable(cu, debug_aranges); + child = child->getSibling(); + } + } +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugInfoEntry.h b/contrib/llvm/lib/DebugInfo/DWARFDebugInfoEntry.h new file mode 100644 index 0000000..aff2e85 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugInfoEntry.h @@ -0,0 +1,135 @@ +//===-- DWARFDebugInfoEntry.h -----------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFDEBUGINFOENTRY_H +#define LLVM_DEBUGINFO_DWARFDEBUGINFOENTRY_H + +#include "DWARFAbbreviationDeclaration.h" +#include "llvm/Support/DataTypes.h" + +namespace llvm { + +class DWARFDebugAranges; +class DWARFCompileUnit; +class DWARFContext; +class DWARFFormValue; + +/// DWARFDebugInfoEntryMinimal - A DIE with only the minimum required data. +class DWARFDebugInfoEntryMinimal { + /// Offset within the .debug_info of the start of this entry. + uint64_t Offset; + + /// How many to subtract from "this" to get the parent. + /// If zero this die has no parent. + uint32_t ParentIdx; + + /// How many to add to "this" to get the sibling. + uint32_t SiblingIdx; + + const DWARFAbbreviationDeclaration *AbbrevDecl; +public: + DWARFDebugInfoEntryMinimal() + : Offset(0), ParentIdx(0), SiblingIdx(0), AbbrevDecl(0) {} + + void dump(raw_ostream &OS, const DWARFCompileUnit *cu, + unsigned recurseDepth, unsigned indent = 0) const; + void dumpAttribute(raw_ostream &OS, const DWARFCompileUnit *cu, + uint32_t *offset_ptr, uint16_t attr, uint16_t form, + unsigned indent = 0) const; + + bool extractFast(const DWARFCompileUnit *cu, const uint8_t *fixed_form_sizes, + uint32_t *offset_ptr); + + /// Extract a debug info entry for a given compile unit from the + /// .debug_info and .debug_abbrev data starting at the given offset. + bool extract(const DWARFCompileUnit *cu, uint32_t *offset_ptr); + + uint32_t getTag() const { return AbbrevDecl ? AbbrevDecl->getTag() : 0; } + bool isNULL() const { return AbbrevDecl == 0; } + uint64_t getOffset() const { return Offset; } + uint32_t getNumAttributes() const { + return !isNULL() ? AbbrevDecl->getNumAttributes() : 0; + } + bool hasChildren() const { return !isNULL() && AbbrevDecl->hasChildren(); } + + // We know we are kept in a vector of contiguous entries, so we know + // our parent will be some index behind "this". + DWARFDebugInfoEntryMinimal *getParent() { + return ParentIdx > 0 ? this - ParentIdx : 0; + } + const DWARFDebugInfoEntryMinimal *getParent() const { + return ParentIdx > 0 ? this - ParentIdx : 0; + } + // We know we are kept in a vector of contiguous entries, so we know + // our sibling will be some index after "this". + DWARFDebugInfoEntryMinimal *getSibling() { + return SiblingIdx > 0 ? this + SiblingIdx : 0; + } + const DWARFDebugInfoEntryMinimal *getSibling() const { + return SiblingIdx > 0 ? this + SiblingIdx : 0; + } + // We know we are kept in a vector of contiguous entries, so we know + // we don't need to store our child pointer, if we have a child it will + // be the next entry in the list... + DWARFDebugInfoEntryMinimal *getFirstChild() { + return hasChildren() ? this + 1 : 0; + } + const DWARFDebugInfoEntryMinimal *getFirstChild() const { + return hasChildren() ? this + 1 : 0; + } + + void setParent(DWARFDebugInfoEntryMinimal *parent) { + if (parent) { + // We know we are kept in a vector of contiguous entries, so we know + // our parent will be some index behind "this". + ParentIdx = this - parent; + } else + ParentIdx = 0; + } + void setSibling(DWARFDebugInfoEntryMinimal *sibling) { + if (sibling) { + // We know we are kept in a vector of contiguous entries, so we know + // our sibling will be some index after "this". + SiblingIdx = sibling - this; + sibling->setParent(getParent()); + } else + SiblingIdx = 0; + } + + const DWARFAbbreviationDeclaration *getAbbreviationDeclarationPtr() const { + return AbbrevDecl; + } + + uint32_t getAttributeValue(const DWARFCompileUnit *cu, + const uint16_t attr, DWARFFormValue &formValue, + uint32_t *end_attr_offset_ptr = 0) const; + + const char* getAttributeValueAsString(const DWARFCompileUnit* cu, + const uint16_t attr, + const char *fail_value) const; + + uint64_t getAttributeValueAsUnsigned(const DWARFCompileUnit *cu, + const uint16_t attr, + uint64_t fail_value) const; + + uint64_t getAttributeValueAsReference(const DWARFCompileUnit *cu, + const uint16_t attr, + uint64_t fail_value) const; + + int64_t getAttributeValueAsSigned(const DWARFCompileUnit* cu, + const uint16_t attr, + int64_t fail_value) const; + + void buildAddressRangeTable(const DWARFCompileUnit *cu, + DWARFDebugAranges *debug_aranges) const; +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugLine.cpp b/contrib/llvm/lib/DebugInfo/DWARFDebugLine.cpp new file mode 100644 index 0000000..fe1ef78 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugLine.cpp @@ -0,0 +1,475 @@ +//===-- DWARFDebugLine.cpp ------------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFDebugLine.h" +#include "llvm/Support/Dwarf.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +#include <algorithm> +using namespace llvm; +using namespace dwarf; + +void DWARFDebugLine::Prologue::dump(raw_ostream &OS) const { + OS << "Line table prologue:\n" + << format(" total_length: 0x%8.8x\n", TotalLength) + << format(" version: %u\n", Version) + << format("prologue_length: 0x%8.8x\n", PrologueLength) + << format("min_inst_length: %u\n", MinInstLength) + << format("default_is_stmt: %u\n", DefaultIsStmt) + << format(" line_base: %i\n", LineBase) + << format(" line_range: %u\n", LineRange) + << format(" opcode_base: %u\n", OpcodeBase); + + for (uint32_t i = 0; i < StandardOpcodeLengths.size(); ++i) + OS << format("standard_opcode_lengths[%s] = %u\n", LNStandardString(i+1), + StandardOpcodeLengths[i]); + + if (!IncludeDirectories.empty()) + for (uint32_t i = 0; i < IncludeDirectories.size(); ++i) + OS << format("include_directories[%3u] = '", i+1) + << IncludeDirectories[i] << "'\n"; + + if (!FileNames.empty()) { + OS << " Dir Mod Time File Len File Name\n" + << " ---- ---------- ---------- -----------" + "----------------\n"; + for (uint32_t i = 0; i < FileNames.size(); ++i) { + const FileNameEntry& fileEntry = FileNames[i]; + OS << format("file_names[%3u] %4u ", i+1, fileEntry.DirIdx) + << format("0x%8.8x 0x%8.8x ", fileEntry.ModTime, fileEntry.Length) + << fileEntry.Name << '\n'; + } + } +} + +void DWARFDebugLine::Row::postAppend() { + BasicBlock = false; + PrologueEnd = false; + EpilogueBegin = false; +} + +void DWARFDebugLine::Row::reset(bool default_is_stmt) { + Address = 0; + Line = 1; + Column = 0; + File = 1; + Isa = 0; + IsStmt = default_is_stmt; + BasicBlock = false; + EndSequence = false; + PrologueEnd = false; + EpilogueBegin = false; +} + +void DWARFDebugLine::Row::dump(raw_ostream &OS) const { + OS << format("0x%16.16llx %6u %6u", Address, Line, Column) + << format(" %6u %3u ", File, Isa) + << (IsStmt ? " is_stmt" : "") + << (BasicBlock ? " basic_block" : "") + << (PrologueEnd ? " prologue_end" : "") + << (EpilogueBegin ? " epilogue_begin" : "") + << (EndSequence ? " end_sequence" : "") + << '\n'; +} + +void DWARFDebugLine::LineTable::dump(raw_ostream &OS) const { + Prologue.dump(OS); + OS << '\n'; + + if (!Rows.empty()) { + OS << "Address Line Column File ISA Flags\n" + << "------------------ ------ ------ ------ --- -------------\n"; + for (std::vector<Row>::const_iterator pos = Rows.begin(), + end = Rows.end(); pos != end; ++pos) + pos->dump(OS); + } +} + +DWARFDebugLine::State::~State() {} + +void DWARFDebugLine::State::appendRowToMatrix(uint32_t offset) { + ++row; // Increase the row number. + LineTable::appendRow(*this); + Row::postAppend(); +} + +DWARFDebugLine::DumpingState::~DumpingState() {} + +void DWARFDebugLine::DumpingState::finalize(uint32_t offset) { + LineTable::dump(OS); +} + +const DWARFDebugLine::LineTable * +DWARFDebugLine::getLineTable(uint32_t offset) const { + LineTableConstIter pos = LineTableMap.find(offset); + if (pos != LineTableMap.end()) + return &pos->second; + return 0; +} + +const DWARFDebugLine::LineTable * +DWARFDebugLine::getOrParseLineTable(DataExtractor debug_line_data, + uint32_t offset) { + std::pair<LineTableIter, bool> pos = + LineTableMap.insert(LineTableMapTy::value_type(offset, LineTable())); + if (pos.second) { + // Parse and cache the line table for at this offset. + State state; + if (!parseStatementTable(debug_line_data, &offset, state)) + return 0; + pos.first->second = state; + } + return &pos.first->second; +} + +bool +DWARFDebugLine::parsePrologue(DataExtractor debug_line_data, + uint32_t *offset_ptr, Prologue *prologue) { + const uint32_t prologue_offset = *offset_ptr; + + prologue->clear(); + prologue->TotalLength = debug_line_data.getU32(offset_ptr); + prologue->Version = debug_line_data.getU16(offset_ptr); + if (prologue->Version != 2) + return false; + + prologue->PrologueLength = debug_line_data.getU32(offset_ptr); + const uint32_t end_prologue_offset = prologue->PrologueLength + *offset_ptr; + prologue->MinInstLength = debug_line_data.getU8(offset_ptr); + prologue->DefaultIsStmt = debug_line_data.getU8(offset_ptr); + prologue->LineBase = debug_line_data.getU8(offset_ptr); + prologue->LineRange = debug_line_data.getU8(offset_ptr); + prologue->OpcodeBase = debug_line_data.getU8(offset_ptr); + + prologue->StandardOpcodeLengths.reserve(prologue->OpcodeBase-1); + for (uint32_t i = 1; i < prologue->OpcodeBase; ++i) { + uint8_t op_len = debug_line_data.getU8(offset_ptr); + prologue->StandardOpcodeLengths.push_back(op_len); + } + + while (*offset_ptr < end_prologue_offset) { + const char *s = debug_line_data.getCStr(offset_ptr); + if (s && s[0]) + prologue->IncludeDirectories.push_back(s); + else + break; + } + + while (*offset_ptr < end_prologue_offset) { + const char *name = debug_line_data.getCStr(offset_ptr); + if (name && name[0]) { + FileNameEntry fileEntry; + fileEntry.Name = name; + fileEntry.DirIdx = debug_line_data.getULEB128(offset_ptr); + fileEntry.ModTime = debug_line_data.getULEB128(offset_ptr); + fileEntry.Length = debug_line_data.getULEB128(offset_ptr); + prologue->FileNames.push_back(fileEntry); + } else { + break; + } + } + + if (*offset_ptr != end_prologue_offset) { + fprintf(stderr, "warning: parsing line table prologue at 0x%8.8x should" + " have ended at 0x%8.8x but it ended ad 0x%8.8x\n", + prologue_offset, end_prologue_offset, *offset_ptr); + } + return end_prologue_offset; +} + +bool +DWARFDebugLine::parseStatementTable(DataExtractor debug_line_data, + uint32_t *offset_ptr, State &state) { + const uint32_t debug_line_offset = *offset_ptr; + + Prologue *prologue = &state.Prologue; + + if (!parsePrologue(debug_line_data, offset_ptr, prologue)) { + // Restore our offset and return false to indicate failure! + *offset_ptr = debug_line_offset; + return false; + } + + const uint32_t end_offset = debug_line_offset + prologue->TotalLength + + sizeof(prologue->TotalLength); + + state.reset(); + + while (*offset_ptr < end_offset) { + uint8_t opcode = debug_line_data.getU8(offset_ptr); + + if (opcode == 0) { + // Extended Opcodes always start with a zero opcode followed by + // a uleb128 length so you can skip ones you don't know about + uint32_t ext_offset = *offset_ptr; + uint64_t len = debug_line_data.getULEB128(offset_ptr); + uint32_t arg_size = len - (*offset_ptr - ext_offset); + + uint8_t sub_opcode = debug_line_data.getU8(offset_ptr); + switch (sub_opcode) { + case DW_LNE_end_sequence: + // Set the end_sequence register of the state machine to true and + // append a row to the matrix using the current values of the + // state-machine registers. Then reset the registers to the initial + // values specified above. Every statement program sequence must end + // with a DW_LNE_end_sequence instruction which creates a row whose + // address is that of the byte after the last target machine instruction + // of the sequence. + state.EndSequence = true; + state.appendRowToMatrix(*offset_ptr); + state.reset(); + break; + + case DW_LNE_set_address: + // Takes a single relocatable address as an operand. The size of the + // operand is the size appropriate to hold an address on the target + // machine. Set the address register to the value given by the + // relocatable address. All of the other statement program opcodes + // that affect the address register add a delta to it. This instruction + // stores a relocatable value into it instead. + state.Address = debug_line_data.getAddress(offset_ptr); + break; + + case DW_LNE_define_file: + // Takes 4 arguments. The first is a null terminated string containing + // a source file name. The second is an unsigned LEB128 number + // representing the directory index of the directory in which the file + // was found. The third is an unsigned LEB128 number representing the + // time of last modification of the file. The fourth is an unsigned + // LEB128 number representing the length in bytes of the file. The time + // and length fields may contain LEB128(0) if the information is not + // available. + // + // The directory index represents an entry in the include_directories + // section of the statement program prologue. The index is LEB128(0) + // if the file was found in the current directory of the compilation, + // LEB128(1) if it was found in the first directory in the + // include_directories section, and so on. The directory index is + // ignored for file names that represent full path names. + // + // The files are numbered, starting at 1, in the order in which they + // appear; the names in the prologue come before names defined by + // the DW_LNE_define_file instruction. These numbers are used in the + // the file register of the state machine. + { + FileNameEntry fileEntry; + fileEntry.Name = debug_line_data.getCStr(offset_ptr); + fileEntry.DirIdx = debug_line_data.getULEB128(offset_ptr); + fileEntry.ModTime = debug_line_data.getULEB128(offset_ptr); + fileEntry.Length = debug_line_data.getULEB128(offset_ptr); + prologue->FileNames.push_back(fileEntry); + } + break; + + default: + // Length doesn't include the zero opcode byte or the length itself, but + // it does include the sub_opcode, so we have to adjust for that below + (*offset_ptr) += arg_size; + break; + } + } else if (opcode < prologue->OpcodeBase) { + switch (opcode) { + // Standard Opcodes + case DW_LNS_copy: + // Takes no arguments. Append a row to the matrix using the + // current values of the state-machine registers. Then set + // the basic_block register to false. + state.appendRowToMatrix(*offset_ptr); + break; + + case DW_LNS_advance_pc: + // Takes a single unsigned LEB128 operand, multiplies it by the + // min_inst_length field of the prologue, and adds the + // result to the address register of the state machine. + state.Address += debug_line_data.getULEB128(offset_ptr) * + prologue->MinInstLength; + break; + + case DW_LNS_advance_line: + // Takes a single signed LEB128 operand and adds that value to + // the line register of the state machine. + state.Line += debug_line_data.getSLEB128(offset_ptr); + break; + + case DW_LNS_set_file: + // Takes a single unsigned LEB128 operand and stores it in the file + // register of the state machine. + state.File = debug_line_data.getULEB128(offset_ptr); + break; + + case DW_LNS_set_column: + // Takes a single unsigned LEB128 operand and stores it in the + // column register of the state machine. + state.Column = debug_line_data.getULEB128(offset_ptr); + break; + + case DW_LNS_negate_stmt: + // Takes no arguments. Set the is_stmt register of the state + // machine to the logical negation of its current value. + state.IsStmt = !state.IsStmt; + break; + + case DW_LNS_set_basic_block: + // Takes no arguments. Set the basic_block register of the + // state machine to true + state.BasicBlock = true; + break; + + case DW_LNS_const_add_pc: + // Takes no arguments. Add to the address register of the state + // machine the address increment value corresponding to special + // opcode 255. The motivation for DW_LNS_const_add_pc is this: + // when the statement program needs to advance the address by a + // small amount, it can use a single special opcode, which occupies + // a single byte. When it needs to advance the address by up to + // twice the range of the last special opcode, it can use + // DW_LNS_const_add_pc followed by a special opcode, for a total + // of two bytes. Only if it needs to advance the address by more + // than twice that range will it need to use both DW_LNS_advance_pc + // and a special opcode, requiring three or more bytes. + { + uint8_t adjust_opcode = 255 - prologue->OpcodeBase; + uint64_t addr_offset = (adjust_opcode / prologue->LineRange) * + prologue->MinInstLength; + state.Address += addr_offset; + } + break; + + case DW_LNS_fixed_advance_pc: + // Takes a single uhalf operand. Add to the address register of + // the state machine the value of the (unencoded) operand. This + // is the only extended opcode that takes an argument that is not + // a variable length number. The motivation for DW_LNS_fixed_advance_pc + // is this: existing assemblers cannot emit DW_LNS_advance_pc or + // special opcodes because they cannot encode LEB128 numbers or + // judge when the computation of a special opcode overflows and + // requires the use of DW_LNS_advance_pc. Such assemblers, however, + // can use DW_LNS_fixed_advance_pc instead, sacrificing compression. + state.Address += debug_line_data.getU16(offset_ptr); + break; + + case DW_LNS_set_prologue_end: + // Takes no arguments. Set the prologue_end register of the + // state machine to true + state.PrologueEnd = true; + break; + + case DW_LNS_set_epilogue_begin: + // Takes no arguments. Set the basic_block register of the + // state machine to true + state.EpilogueBegin = true; + break; + + case DW_LNS_set_isa: + // Takes a single unsigned LEB128 operand and stores it in the + // column register of the state machine. + state.Isa = debug_line_data.getULEB128(offset_ptr); + break; + + default: + // Handle any unknown standard opcodes here. We know the lengths + // of such opcodes because they are specified in the prologue + // as a multiple of LEB128 operands for each opcode. + { + assert(opcode - 1U < prologue->StandardOpcodeLengths.size()); + uint8_t opcode_length = prologue->StandardOpcodeLengths[opcode - 1]; + for (uint8_t i=0; i<opcode_length; ++i) + debug_line_data.getULEB128(offset_ptr); + } + break; + } + } else { + // Special Opcodes + + // A special opcode value is chosen based on the amount that needs + // to be added to the line and address registers. The maximum line + // increment for a special opcode is the value of the line_base + // field in the header, plus the value of the line_range field, + // minus 1 (line base + line range - 1). If the desired line + // increment is greater than the maximum line increment, a standard + // opcode must be used instead of a special opcode. The "address + // advance" is calculated by dividing the desired address increment + // by the minimum_instruction_length field from the header. The + // special opcode is then calculated using the following formula: + // + // opcode = (desired line increment - line_base) + + // (line_range * address advance) + opcode_base + // + // If the resulting opcode is greater than 255, a standard opcode + // must be used instead. + // + // To decode a special opcode, subtract the opcode_base from the + // opcode itself to give the adjusted opcode. The amount to + // increment the address register is the result of the adjusted + // opcode divided by the line_range multiplied by the + // minimum_instruction_length field from the header. That is: + // + // address increment = (adjusted opcode / line_range) * + // minimum_instruction_length + // + // The amount to increment the line register is the line_base plus + // the result of the adjusted opcode modulo the line_range. That is: + // + // line increment = line_base + (adjusted opcode % line_range) + + uint8_t adjust_opcode = opcode - prologue->OpcodeBase; + uint64_t addr_offset = (adjust_opcode / prologue->LineRange) * + prologue->MinInstLength; + int32_t line_offset = prologue->LineBase + + (adjust_opcode % prologue->LineRange); + state.Line += line_offset; + state.Address += addr_offset; + state.appendRowToMatrix(*offset_ptr); + } + } + + state.finalize(*offset_ptr); + + return end_offset; +} + +static bool findMatchingAddress(const DWARFDebugLine::Row& row1, + const DWARFDebugLine::Row& row2) { + return row1.Address < row2.Address; +} + +uint32_t +DWARFDebugLine::LineTable::lookupAddress(uint64_t address, + uint64_t cu_high_pc) const { + uint32_t index = UINT32_MAX; + if (!Rows.empty()) { + // Use the lower_bound algorithm to perform a binary search since we know + // that our line table data is ordered by address. + DWARFDebugLine::Row row; + row.Address = address; + typedef std::vector<Row>::const_iterator iterator; + iterator begin_pos = Rows.begin(); + iterator end_pos = Rows.end(); + iterator pos = std::lower_bound(begin_pos, end_pos, row, + findMatchingAddress); + if (pos == end_pos) { + if (address < cu_high_pc) + return Rows.size()-1; + } else { + // Rely on fact that we are using a std::vector and we can do + // pointer arithmetic to find the row index (which will be one less + // that what we found since it will find the first position after + // the current address) since std::vector iterators are just + // pointers to the container type. + index = pos - begin_pos; + if (pos->Address > address) { + if (index > 0) + --index; + else + index = UINT32_MAX; + } + } + } + return index; // Failed to find address. +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFDebugLine.h b/contrib/llvm/lib/DebugInfo/DWARFDebugLine.h new file mode 100644 index 0000000..bc6a70b --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFDebugLine.h @@ -0,0 +1,190 @@ +//===-- DWARFDebugLine.h ----------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFDEBUGLINE_H +#define LLVM_DEBUGINFO_DWARFDEBUGLINE_H + +#include "llvm/Support/DataExtractor.h" +#include <map> +#include <string> +#include <vector> + +namespace llvm { + +class raw_ostream; + +class DWARFDebugLine { +public: + struct FileNameEntry { + FileNameEntry() : DirIdx(0), ModTime(0), Length(0) {} + + std::string Name; + uint64_t DirIdx; + uint64_t ModTime; + uint64_t Length; + }; + + struct Prologue { + Prologue() + : TotalLength(0), Version(0), PrologueLength(0), MinInstLength(0), + DefaultIsStmt(0), LineBase(0), LineRange(0), OpcodeBase(0) {} + + // The size in bytes of the statement information for this compilation unit + // (not including the total_length field itself). + uint32_t TotalLength; + // Version identifier for the statement information format. + uint16_t Version; + // The number of bytes following the prologue_length field to the beginning + // of the first byte of the statement program itself. + uint32_t PrologueLength; + // The size in bytes of the smallest target machine instruction. Statement + // program opcodes that alter the address register first multiply their + // operands by this value. + uint8_t MinInstLength; + // The initial value of theis_stmtregister. + uint8_t DefaultIsStmt; + // This parameter affects the meaning of the special opcodes. See below. + int8_t LineBase; + // This parameter affects the meaning of the special opcodes. See below. + uint8_t LineRange; + // The number assigned to the first special opcode. + uint8_t OpcodeBase; + std::vector<uint8_t> StandardOpcodeLengths; + std::vector<std::string> IncludeDirectories; + std::vector<FileNameEntry> FileNames; + + // Length of the prologue in bytes. + uint32_t getLength() const { + return PrologueLength + sizeof(TotalLength) + sizeof(Version) + + sizeof(PrologueLength); + } + // Length of the line table data in bytes (not including the prologue). + uint32_t getStatementTableLength() const { + return TotalLength + sizeof(TotalLength) - getLength(); + } + int32_t getMaxLineIncrementForSpecialOpcode() const { + return LineBase + (int8_t)LineRange - 1; + } + void dump(raw_ostream &OS) const; + void clear() { + TotalLength = Version = PrologueLength = 0; + MinInstLength = LineBase = LineRange = OpcodeBase = 0; + StandardOpcodeLengths.clear(); + IncludeDirectories.clear(); + FileNames.clear(); + } + }; + + // Standard .debug_line state machine structure. + struct Row { + Row(bool default_is_stmt = false) { reset(default_is_stmt); } + /// Called after a row is appended to the matrix. + void postAppend(); + void reset(bool default_is_stmt); + void dump(raw_ostream &OS) const; + + // The program-counter value corresponding to a machine instruction + // generated by the compiler. + uint64_t Address; + // An unsigned integer indicating a source line number. Lines are numbered + // beginning at 1. The compiler may emit the value 0 in cases where an + // instruction cannot be attributed to any source line. + uint32_t Line; + // An unsigned integer indicating a column number within a source line. + // Columns are numbered beginning at 1. The value 0 is reserved to indicate + // that a statement begins at the 'left edge' of the line. + uint16_t Column; + // An unsigned integer indicating the identity of the source file + // corresponding to a machine instruction. + uint16_t File; + // An unsigned integer whose value encodes the applicable instruction set + // architecture for the current instruction. + uint8_t Isa; + // A boolean indicating that the current instruction is the beginning of a + // statement. + uint8_t IsStmt:1, + // A boolean indicating that the current instruction is the + // beginning of a basic block. + BasicBlock:1, + // A boolean indicating that the current address is that of the + // first byte after the end of a sequence of target machine + // instructions. + EndSequence:1, + // A boolean indicating that the current address is one (of possibly + // many) where execution should be suspended for an entry breakpoint + // of a function. + PrologueEnd:1, + // A boolean indicating that the current address is one (of possibly + // many) where execution should be suspended for an exit breakpoint + // of a function. + EpilogueBegin:1; + }; + + struct LineTable { + void appendRow(const DWARFDebugLine::Row &state) { Rows.push_back(state); } + void clear() { + Prologue.clear(); + Rows.clear(); + } + + uint32_t lookupAddress(uint64_t address, uint64_t cu_high_pc) const; + void dump(raw_ostream &OS) const; + + struct Prologue Prologue; + std::vector<Row> Rows; + }; + + struct State : public Row, public LineTable { + // Special row codes. + enum { + StartParsingLineTable = 0, + DoneParsingLineTable = -1 + }; + + State() : row(StartParsingLineTable) {} + virtual ~State(); + + virtual void appendRowToMatrix(uint32_t offset); + virtual void finalize(uint32_t offset) { row = DoneParsingLineTable; } + virtual void reset() { Row::reset(Prologue.DefaultIsStmt); } + + // The row number that starts at zero for the prologue, and increases for + // each row added to the matrix. + unsigned row; + }; + + struct DumpingState : public State { + DumpingState(raw_ostream &OS) : OS(OS) {} + virtual ~DumpingState(); + virtual void finalize(uint32_t offset); + private: + raw_ostream &OS; + }; + + static bool parsePrologue(DataExtractor debug_line_data, uint32_t *offset_ptr, + Prologue *prologue); + /// Parse a single line table (prologue and all rows). + static bool parseStatementTable(DataExtractor debug_line_data, + uint32_t *offset_ptr, State &state); + + const LineTable *getLineTable(uint32_t offset) const; + const LineTable *getOrParseLineTable(DataExtractor debug_line_data, + uint32_t offset); + +private: + typedef std::map<uint32_t, LineTable> LineTableMapTy; + typedef LineTableMapTy::iterator LineTableIter; + typedef LineTableMapTy::const_iterator LineTableConstIter; + + LineTableMapTy LineTableMap; +}; + +} + +#endif diff --git a/contrib/llvm/lib/DebugInfo/DWARFFormValue.cpp b/contrib/llvm/lib/DebugInfo/DWARFFormValue.cpp new file mode 100644 index 0000000..705efe5 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFFormValue.cpp @@ -0,0 +1,427 @@ +//===-- DWARFFormValue.cpp ------------------------------------------------===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#include "DWARFFormValue.h" +#include "DWARFCompileUnit.h" +#include "DWARFContext.h" +#include "llvm/Support/Dwarf.h" +#include "llvm/Support/Format.h" +#include "llvm/Support/raw_ostream.h" +#include <cassert> +using namespace llvm; +using namespace dwarf; + +static const uint8_t form_sizes_addr4[] = { + 0, // 0x00 unused + 4, // 0x01 DW_FORM_addr + 0, // 0x02 unused + 0, // 0x03 DW_FORM_block2 + 0, // 0x04 DW_FORM_block4 + 2, // 0x05 DW_FORM_data2 + 4, // 0x06 DW_FORM_data4 + 8, // 0x07 DW_FORM_data8 + 0, // 0x08 DW_FORM_string + 0, // 0x09 DW_FORM_block + 0, // 0x0a DW_FORM_block1 + 1, // 0x0b DW_FORM_data1 + 1, // 0x0c DW_FORM_flag + 0, // 0x0d DW_FORM_sdata + 4, // 0x0e DW_FORM_strp + 0, // 0x0f DW_FORM_udata + 4, // 0x10 DW_FORM_ref_addr + 1, // 0x11 DW_FORM_ref1 + 2, // 0x12 DW_FORM_ref2 + 4, // 0x13 DW_FORM_ref4 + 8, // 0x14 DW_FORM_ref8 + 0, // 0x15 DW_FORM_ref_udata + 0, // 0x16 DW_FORM_indirect +}; + +static const uint8_t form_sizes_addr8[] = { + 0, // 0x00 unused + 8, // 0x01 DW_FORM_addr + 0, // 0x02 unused + 0, // 0x03 DW_FORM_block2 + 0, // 0x04 DW_FORM_block4 + 2, // 0x05 DW_FORM_data2 + 4, // 0x06 DW_FORM_data4 + 8, // 0x07 DW_FORM_data8 + 0, // 0x08 DW_FORM_string + 0, // 0x09 DW_FORM_block + 0, // 0x0a DW_FORM_block1 + 1, // 0x0b DW_FORM_data1 + 1, // 0x0c DW_FORM_flag + 0, // 0x0d DW_FORM_sdata + 4, // 0x0e DW_FORM_strp + 0, // 0x0f DW_FORM_udata + 8, // 0x10 DW_FORM_ref_addr + 1, // 0x11 DW_FORM_ref1 + 2, // 0x12 DW_FORM_ref2 + 4, // 0x13 DW_FORM_ref4 + 8, // 0x14 DW_FORM_ref8 + 0, // 0x15 DW_FORM_ref_udata + 0, // 0x16 DW_FORM_indirect +}; + +const uint8_t * +DWARFFormValue::getFixedFormSizesForAddressSize(uint8_t addr_size) { + switch (addr_size) { + case 4: return form_sizes_addr4; + case 8: return form_sizes_addr8; + } + return NULL; +} + +bool +DWARFFormValue::extractValue(DataExtractor data, uint32_t *offset_ptr, + const DWARFCompileUnit *cu) { + bool indirect = false; + bool is_block = false; + Value.data = NULL; + // Read the value for the form into value and follow and DW_FORM_indirect + // instances we run into + do { + indirect = false; + switch (Form) { + case DW_FORM_addr: + case DW_FORM_ref_addr: + Value.uval = data.getUnsigned(offset_ptr, cu->getAddressByteSize()); + break; + case DW_FORM_block: + Value.uval = data.getULEB128(offset_ptr); + is_block = true; + break; + case DW_FORM_block1: + Value.uval = data.getU8(offset_ptr); + is_block = true; + break; + case DW_FORM_block2: + Value.uval = data.getU16(offset_ptr); + is_block = true; + break; + case DW_FORM_block4: + Value.uval = data.getU32(offset_ptr); + is_block = true; + break; + case DW_FORM_data1: + case DW_FORM_ref1: + case DW_FORM_flag: + Value.uval = data.getU8(offset_ptr); + break; + case DW_FORM_data2: + case DW_FORM_ref2: + Value.uval = data.getU16(offset_ptr); + break; + case DW_FORM_data4: + case DW_FORM_ref4: + Value.uval = data.getU32(offset_ptr); + break; + case DW_FORM_data8: + case DW_FORM_ref8: + Value.uval = data.getU64(offset_ptr); + break; + case DW_FORM_sdata: + Value.sval = data.getSLEB128(offset_ptr); + break; + case DW_FORM_strp: + Value.uval = data.getU32(offset_ptr); + break; + case DW_FORM_udata: + case DW_FORM_ref_udata: + Value.uval = data.getULEB128(offset_ptr); + break; + case DW_FORM_string: + Value.cstr = data.getCStr(offset_ptr); + // Set the string value to also be the data for inlined cstr form + // values only so we can tell the differnence between DW_FORM_string + // and DW_FORM_strp form values + Value.data = (uint8_t*)Value.cstr; + break; + case DW_FORM_indirect: + Form = data.getULEB128(offset_ptr); + indirect = true; + break; + default: + return false; + } + } while (indirect); + + if (is_block) { + StringRef str = data.getData().substr(*offset_ptr, Value.uval); + Value.data = NULL; + if (!str.empty()) { + Value.data = reinterpret_cast<const uint8_t *>(str.data()); + *offset_ptr += Value.uval; + } + } + + return true; +} + +bool +DWARFFormValue::skipValue(DataExtractor debug_info_data, uint32_t* offset_ptr, + const DWARFCompileUnit *cu) const { + return DWARFFormValue::skipValue(Form, debug_info_data, offset_ptr, cu); +} + +bool +DWARFFormValue::skipValue(uint16_t form, DataExtractor debug_info_data, + uint32_t *offset_ptr, const DWARFCompileUnit *cu) { + bool indirect = false; + do { + indirect = false; + switch (form) { + // Blocks if inlined data that have a length field and the data bytes + // inlined in the .debug_info + case DW_FORM_block: { + uint64_t size = debug_info_data.getULEB128(offset_ptr); + *offset_ptr += size; + return true; + } + case DW_FORM_block1: { + uint8_t size = debug_info_data.getU8(offset_ptr); + *offset_ptr += size; + return true; + } + case DW_FORM_block2: { + uint16_t size = debug_info_data.getU16(offset_ptr); + *offset_ptr += size; + return true; + } + case DW_FORM_block4: { + uint32_t size = debug_info_data.getU32(offset_ptr); + *offset_ptr += size; + return true; + } + + // Inlined NULL terminated C-strings + case DW_FORM_string: + debug_info_data.getCStr(offset_ptr); + return true; + + // Compile unit address sized values + case DW_FORM_addr: + case DW_FORM_ref_addr: + *offset_ptr += cu->getAddressByteSize(); + return true; + + // 1 byte values + case DW_FORM_data1: + case DW_FORM_flag: + case DW_FORM_ref1: + *offset_ptr += 1; + return true; + + // 2 byte values + case DW_FORM_data2: + case DW_FORM_ref2: + *offset_ptr += 2; + return true; + + // 4 byte values + case DW_FORM_strp: + case DW_FORM_data4: + case DW_FORM_ref4: + *offset_ptr += 4; + return true; + + // 8 byte values + case DW_FORM_data8: + case DW_FORM_ref8: + *offset_ptr += 8; + return true; + + // signed or unsigned LEB 128 values + // case DW_FORM_APPLE_db_str: + case DW_FORM_sdata: + case DW_FORM_udata: + case DW_FORM_ref_udata: + debug_info_data.getULEB128(offset_ptr); + return true; + + case DW_FORM_indirect: + indirect = true; + form = debug_info_data.getULEB128(offset_ptr); + break; + default: + return false; + } + } while (indirect); + return true; +} + +void +DWARFFormValue::dump(raw_ostream &OS, const DWARFCompileUnit *cu) const { + DataExtractor debug_str_data(cu->getContext().getStringSection(), true, 0); + uint64_t uvalue = getUnsigned(); + bool cu_relative_offset = false; + + switch (Form) { + case DW_FORM_addr: OS << format("0x%016x", uvalue); break; + case DW_FORM_flag: + case DW_FORM_data1: OS << format("0x%02x", uvalue); break; + case DW_FORM_data2: OS << format("0x%04x", uvalue); break; + case DW_FORM_data4: OS << format("0x%08x", uvalue); break; + case DW_FORM_data8: OS << format("0x%016x", uvalue); break; + case DW_FORM_string: + OS << '"'; + OS.write_escaped(getAsCString(NULL)); + OS << '"'; + break; + case DW_FORM_block: + case DW_FORM_block1: + case DW_FORM_block2: + case DW_FORM_block4: + if (uvalue > 0) { + switch (Form) { + case DW_FORM_block: OS << format("<0x%llx> ", uvalue); break; + case DW_FORM_block1: OS << format("<0x%2.2x> ", (uint8_t)uvalue); break; + case DW_FORM_block2: OS << format("<0x%4.4x> ", (uint16_t)uvalue); break; + case DW_FORM_block4: OS << format("<0x%8.8x> ", (uint32_t)uvalue); break; + default: break; + } + + const uint8_t* data_ptr = Value.data; + if (data_ptr) { + // uvalue contains size of block + const uint8_t* end_data_ptr = data_ptr + uvalue; + while (data_ptr < end_data_ptr) { + OS << format("%2.2x ", *data_ptr); + ++data_ptr; + } + } + else + OS << "NULL"; + } + break; + + case DW_FORM_sdata: OS << getSigned(); break; + case DW_FORM_udata: OS << getUnsigned(); break; + case DW_FORM_strp: { + OS << format(" .debug_str[0x%8.8x] = ", (uint32_t)uvalue); + const char* dbg_str = getAsCString(&debug_str_data); + if (dbg_str) { + OS << '"'; + OS.write_escaped(dbg_str); + OS << '"'; + } + break; + } + case DW_FORM_ref_addr: + OS << format("0x%016x", uvalue); + break; + case DW_FORM_ref1: + cu_relative_offset = true; + OS << format("cu + 0x%2.2x", (uint8_t)uvalue); + break; + case DW_FORM_ref2: + cu_relative_offset = true; + OS << format("cu + 0x%4.4x", (uint16_t)uvalue); + break; + case DW_FORM_ref4: + cu_relative_offset = true; + OS << format("cu + 0x%4.4x", (uint32_t)uvalue); + break; + case DW_FORM_ref8: + cu_relative_offset = true; + OS << format("cu + 0x%8.8llx", uvalue); + break; + case DW_FORM_ref_udata: + cu_relative_offset = true; + OS << format("cu + 0x%llx", uvalue); + break; + + // All DW_FORM_indirect attributes should be resolved prior to calling + // this function + case DW_FORM_indirect: + OS << "DW_FORM_indirect"; + break; + default: + OS << format("DW_FORM(0x%4.4x)", Form); + break; + } + + if (cu_relative_offset) + OS << format(" => {0x%8.8x}", (uvalue + (cu ? cu->getOffset() : 0))); +} + +const char* +DWARFFormValue::getAsCString(const DataExtractor *debug_str_data_ptr) const { + if (isInlinedCStr()) { + return Value.cstr; + } else if (debug_str_data_ptr) { + uint32_t offset = Value.uval; + return debug_str_data_ptr->getCStr(&offset); + } + return NULL; +} + +uint64_t DWARFFormValue::getReference(const DWARFCompileUnit *cu) const { + uint64_t die_offset = Value.uval; + switch (Form) { + case DW_FORM_ref1: + case DW_FORM_ref2: + case DW_FORM_ref4: + case DW_FORM_ref8: + case DW_FORM_ref_udata: + die_offset += (cu ? cu->getOffset() : 0); + break; + default: + break; + } + + return die_offset; +} + +bool +DWARFFormValue::resolveCompileUnitReferences(const DWARFCompileUnit *cu) { + switch (Form) { + case DW_FORM_ref1: + case DW_FORM_ref2: + case DW_FORM_ref4: + case DW_FORM_ref8: + case DW_FORM_ref_udata: + Value.uval += cu->getOffset(); + Form = DW_FORM_ref_addr; + return true; + default: + break; + } + return false; +} + +const uint8_t *DWARFFormValue::BlockData() const { + if (!isInlinedCStr()) + return Value.data; + return NULL; +} + +bool DWARFFormValue::isBlockForm(uint16_t form) { + switch (form) { + case DW_FORM_block: + case DW_FORM_block1: + case DW_FORM_block2: + case DW_FORM_block4: + return true; + } + return false; +} + +bool DWARFFormValue::isDataForm(uint16_t form) { + switch (form) { + case DW_FORM_sdata: + case DW_FORM_udata: + case DW_FORM_data1: + case DW_FORM_data2: + case DW_FORM_data4: + case DW_FORM_data8: + return true; + } + return false; +} diff --git a/contrib/llvm/lib/DebugInfo/DWARFFormValue.h b/contrib/llvm/lib/DebugInfo/DWARFFormValue.h new file mode 100644 index 0000000..22ac011 --- /dev/null +++ b/contrib/llvm/lib/DebugInfo/DWARFFormValue.h @@ -0,0 +1,78 @@ +//===-- DWARFFormValue.h ----------------------------------------*- C++ -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//===----------------------------------------------------------------------===// + +#ifndef LLVM_DEBUGINFO_DWARFFORMVALUE_H +#define LLVM_DEBUGINFO_DWARFFORMVALUE_H + +#include "llvm/Support/DataExtractor.h" + +namespace llvm { + +class DWARFCompileUnit; +class raw_ostream; + +class DWARFFormValue { +public: + struct ValueType { + ValueType() : data(NULL) { + uval = 0; + } + + union { + uint64_t uval; + int64_t sval; + const char* cstr; + }; + const uint8_t* data; + }; + + enum { + eValueTypeInvalid = 0, + eValueTypeUnsigned, + eValueTypeSigned, + eValueTypeCStr, + eValueTypeBlock + }; + +private: + uint16_t Form; // Form for this value. + ValueType Value; // Contains all data for the form. + +public: + DWARFFormValue(uint16_t form = 0) : Form(form) {} + uint16_t getForm() const { return Form; } + const ValueType& value() const { return Value; } + void dump(raw_ostream &OS, const DWARFCompileUnit* cu) const; + bool extractValue(DataExtractor data, uint32_t *offset_ptr, + const DWARFCompileUnit *cu); + bool isInlinedCStr() const { + return Value.data != NULL && Value.data == (uint8_t*)Value.cstr; + } + const uint8_t *BlockData() const; + uint64_t getReference(const DWARFCompileUnit* cu) const; + + /// Resolve any compile unit specific references so that we don't need + /// the compile unit at a later time in order to work with the form + /// value. + bool resolveCompileUnitReferences(const DWARFCompileUnit* cu); + uint64_t getUnsigned() const { return Value.uval; } + int64_t getSigned() const { return Value.sval; } + const char *getAsCString(const DataExtractor *debug_str_data_ptr) const; + bool skipValue(DataExtractor debug_info_data, uint32_t *offset_ptr, + const DWARFCompileUnit *cu) const; + static bool skipValue(uint16_t form, DataExtractor debug_info_data, + uint32_t *offset_ptr, const DWARFCompileUnit *cu); + static bool isBlockForm(uint16_t form); + static bool isDataForm(uint16_t form); + static const uint8_t *getFixedFormSizesForAddressSize(uint8_t addr_size); +}; + +} + +#endif |
