// output.cc -- manage the output file for gold
-// Copyright 2006, 2007, 2008 Free Software Foundation, Inc.
+// Copyright 2006, 2007, 2008, 2009, 2010 Free Software Foundation, Inc.
// Written by Ian Lance Taylor <iant@google.com>.
// This file is part of gold.
#include <sys/mman.h>
#include <sys/stat.h>
#include <algorithm>
-#include "libiberty.h" // for unlink_if_ordinary()
+#include "libiberty.h"
#include "parameters.h"
#include "object.h"
# define MAP_ANONYMOUS MAP_ANON
#endif
+#ifndef HAVE_POSIX_FALLOCATE
+// A dummy, non general, version of posix_fallocate. Here we just set
+// the file size and hope that there is enough disk space. FIXME: We
+// could allocate disk space by walking block by block and writing a
+// zero byte into each block.
+static int
+posix_fallocate(int o, off_t offset, off_t len)
+{
+ return ftruncate(o, offset + len);
+}
+#endif // !defined(HAVE_POSIX_FALLOCATE)
+
namespace gold
{
unattached_section_list_(unattached_section_list),
secnamepool_(secnamepool),
shstrtab_section_(shstrtab_section)
+{
+}
+
+// Compute the current data size.
+
+off_t
+Output_section_headers::do_size() const
{
// Count all the sections. Start with 1 for the null section.
off_t count = 1;
if (!parameters->options().relocatable())
{
- for (Layout::Segment_list::const_iterator p = segment_list->begin();
- p != segment_list->end();
+ for (Layout::Segment_list::const_iterator p =
+ this->segment_list_->begin();
+ p != this->segment_list_->end();
++p)
if ((*p)->type() == elfcpp::PT_LOAD)
count += (*p)->output_section_count();
}
else
{
- for (Layout::Section_list::const_iterator p = section_list->begin();
- p != section_list->end();
+ for (Layout::Section_list::const_iterator p =
+ this->section_list_->begin();
+ p != this->section_list_->end();
++p)
if (((*p)->flags() & elfcpp::SHF_ALLOC) != 0)
++count;
}
- count += unattached_section_list->size();
+ count += this->unattached_section_list_->size();
const int size = parameters->target().get_size();
int shdr_size;
else
gold_unreachable();
- this->set_data_size(count * shdr_size);
+ return count * shdr_size;
}
// Write out the section headers.
else
oshdr.put_sh_link(shstrndx);
- oshdr.put_sh_info(0);
+ size_t segment_count = this->segment_list_->size();
+ oshdr.put_sh_info(segment_count >= elfcpp::PN_XNUM ? segment_count : 0);
+
oshdr.put_sh_addralign(0);
oshdr.put_sh_entsize(0);
}
const Layout::Segment_list& segment_list)
: segment_list_(segment_list)
{
- const int size = parameters->target().get_size();
- int phdr_size;
- if (size == 32)
- phdr_size = elfcpp::Elf_sizes<32>::phdr_size;
- else if (size == 64)
- phdr_size = elfcpp::Elf_sizes<64>::phdr_size;
- else
- gold_unreachable();
-
- this->set_data_size(segment_list.size() * phdr_size);
}
void
of->write_output_view(this->offset(), all_phdrs_size, view);
}
+off_t
+Output_segment_headers::do_size() const
+{
+ const int size = parameters->target().get_size();
+ int phdr_size;
+ if (size == 32)
+ phdr_size = elfcpp::Elf_sizes<32>::phdr_size;
+ else if (size == 64)
+ phdr_size = elfcpp::Elf_sizes<64>::phdr_size;
+ else
+ gold_unreachable();
+
+ return this->segment_list_.size() * phdr_size;
+}
+
// Output_file_header methods.
Output_file_header::Output_file_header(const Target* target,
shstrtab_(NULL),
entry_(entry)
{
- const int size = parameters->target().get_size();
- int ehdr_size;
- if (size == 32)
- ehdr_size = elfcpp::Elf_sizes<32>::ehdr_size;
- else if (size == 64)
- ehdr_size = elfcpp::Elf_sizes<64>::ehdr_size;
- else
- gold_unreachable();
-
- this->set_data_size(ehdr_size);
+ this->set_data_size(this->do_size());
}
// Set the section table information for a file header.
? elfcpp::ELFDATA2MSB
: elfcpp::ELFDATA2LSB);
e_ident[elfcpp::EI_VERSION] = elfcpp::EV_CURRENT;
- // FIXME: Some targets may need to set EI_OSABI and EI_ABIVERSION.
oehdr.put_e_ident(e_ident);
elfcpp::ET e_type;
if (parameters->options().relocatable())
e_type = elfcpp::ET_REL;
- else if (parameters->options().shared())
+ else if (parameters->options().output_is_position_independent())
e_type = elfcpp::ET_DYN;
else
e_type = elfcpp::ET_EXEC;
oehdr.put_e_phoff(this->segment_header_->offset());
oehdr.put_e_shoff(this->section_header_->offset());
-
- // FIXME: The target needs to set the flags.
- oehdr.put_e_flags(0);
-
+ oehdr.put_e_flags(this->target_->processor_specific_flags());
oehdr.put_e_ehsize(elfcpp::Elf_sizes<size>::ehdr_size);
if (this->segment_header_ == NULL)
else
{
oehdr.put_e_phentsize(elfcpp::Elf_sizes<size>::phdr_size);
- oehdr.put_e_phnum(this->segment_header_->data_size()
- / elfcpp::Elf_sizes<size>::phdr_size);
+ size_t phnum = (this->segment_header_->data_size()
+ / elfcpp::Elf_sizes<size>::phdr_size);
+ if (phnum > elfcpp::PN_XNUM)
+ phnum = elfcpp::PN_XNUM;
+ oehdr.put_e_phnum(phnum);
}
oehdr.put_e_shentsize(elfcpp::Elf_sizes<size>::shdr_size);
else
oehdr.put_e_shstrndx(elfcpp::SHN_XINDEX);
+ // Let the target adjust the ELF header, e.g., to set EI_OSABI in
+ // the e_ident field.
+ parameters->target().adjust_elf_header(view, ehdr_size);
+
of->write_output_view(0, ehdr_size, view);
}
return v;
}
+// Compute the current data size.
+
+off_t
+Output_file_header::do_size() const
+{
+ const int size = parameters->target().get_size();
+ if (size == 32)
+ return elfcpp::Elf_sizes<32>::ehdr_size;
+ else if (size == 64)
+ return elfcpp::Elf_sizes<64>::ehdr_size;
+ else
+ gold_unreachable();
+}
+
// Output_data_const methods.
void
unsigned int type,
Output_data* od,
Address address,
- bool is_relative)
+ bool is_relative,
+ bool is_symbolless)
: address_(address), local_sym_index_(GSYM_CODE), type_(type),
- is_relative_(is_relative), is_section_symbol_(false), shndx_(INVALID_CODE)
+ is_relative_(is_relative), is_symbolless_(is_symbolless),
+ is_section_symbol_(false), shndx_(INVALID_CODE)
{
// this->type_ is a bitfield; make sure TYPE fits.
gold_assert(this->type_ == type);
Sized_relobj<size, big_endian>* relobj,
unsigned int shndx,
Address address,
- bool is_relative)
+ bool is_relative,
+ bool is_symbolless)
: address_(address), local_sym_index_(GSYM_CODE), type_(type),
- is_relative_(is_relative), is_section_symbol_(false), shndx_(shndx)
+ is_relative_(is_relative), is_symbolless_(is_symbolless),
+ is_section_symbol_(false), shndx_(shndx)
{
gold_assert(shndx != INVALID_CODE);
// this->type_ is a bitfield; make sure TYPE fits.
Output_data* od,
Address address,
bool is_relative,
+ bool is_symbolless,
bool is_section_symbol)
: address_(address), local_sym_index_(local_sym_index), type_(type),
- is_relative_(is_relative), is_section_symbol_(is_section_symbol),
- shndx_(INVALID_CODE)
+ is_relative_(is_relative), is_symbolless_(is_symbolless),
+ is_section_symbol_(is_section_symbol), shndx_(INVALID_CODE)
{
gold_assert(local_sym_index != GSYM_CODE
&& local_sym_index != INVALID_CODE);
unsigned int shndx,
Address address,
bool is_relative,
+ bool is_symbolless,
bool is_section_symbol)
: address_(address), local_sym_index_(local_sym_index), type_(type),
- is_relative_(is_relative), is_section_symbol_(is_section_symbol),
- shndx_(shndx)
+ is_relative_(is_relative), is_symbolless_(is_symbolless),
+ is_section_symbol_(is_section_symbol), shndx_(shndx)
{
gold_assert(local_sym_index != GSYM_CODE
&& local_sym_index != INVALID_CODE);
Output_data* od,
Address address)
: address_(address), local_sym_index_(SECTION_CODE), type_(type),
- is_relative_(false), is_section_symbol_(true), shndx_(INVALID_CODE)
+ is_relative_(false), is_symbolless_(false),
+ is_section_symbol_(true), shndx_(INVALID_CODE)
{
// this->type_ is a bitfield; make sure TYPE fits.
gold_assert(this->type_ == type);
unsigned int shndx,
Address address)
: address_(address), local_sym_index_(SECTION_CODE), type_(type),
- is_relative_(false), is_section_symbol_(true), shndx_(shndx)
+ is_relative_(false), is_symbolless_(false),
+ is_section_symbol_(true), shndx_(shndx)
{
gold_assert(shndx != INVALID_CODE);
// this->type_ is a bitfield; make sure TYPE fits.
os->set_needs_symtab_index();
}
+// An absolute relocation.
+
+template<bool dynamic, int size, bool big_endian>
+Output_reloc<elfcpp::SHT_REL, dynamic, size, big_endian>::Output_reloc(
+ unsigned int type,
+ Output_data* od,
+ Address address)
+ : address_(address), local_sym_index_(0), type_(type),
+ is_relative_(false), is_symbolless_(false),
+ is_section_symbol_(false), shndx_(INVALID_CODE)
+{
+ // this->type_ is a bitfield; make sure TYPE fits.
+ gold_assert(this->type_ == type);
+ this->u1_.relobj = NULL;
+ this->u2_.od = od;
+}
+
+template<bool dynamic, int size, bool big_endian>
+Output_reloc<elfcpp::SHT_REL, dynamic, size, big_endian>::Output_reloc(
+ unsigned int type,
+ Sized_relobj<size, big_endian>* relobj,
+ unsigned int shndx,
+ Address address)
+ : address_(address), local_sym_index_(0), type_(type),
+ is_relative_(false), is_symbolless_(false),
+ is_section_symbol_(false), shndx_(shndx)
+{
+ gold_assert(shndx != INVALID_CODE);
+ // this->type_ is a bitfield; make sure TYPE fits.
+ gold_assert(this->type_ == type);
+ this->u1_.relobj = NULL;
+ this->u2_.relobj = relobj;
+}
+
+// A target specific relocation.
+
+template<bool dynamic, int size, bool big_endian>
+Output_reloc<elfcpp::SHT_REL, dynamic, size, big_endian>::Output_reloc(
+ unsigned int type,
+ void* arg,
+ Output_data* od,
+ Address address)
+ : address_(address), local_sym_index_(TARGET_CODE), type_(type),
+ is_relative_(false), is_symbolless_(false),
+ is_section_symbol_(false), shndx_(INVALID_CODE)
+{
+ // this->type_ is a bitfield; make sure TYPE fits.
+ gold_assert(this->type_ == type);
+ this->u1_.arg = arg;
+ this->u2_.od = od;
+}
+
+template<bool dynamic, int size, bool big_endian>
+Output_reloc<elfcpp::SHT_REL, dynamic, size, big_endian>::Output_reloc(
+ unsigned int type,
+ void* arg,
+ Sized_relobj<size, big_endian>* relobj,
+ unsigned int shndx,
+ Address address)
+ : address_(address), local_sym_index_(TARGET_CODE), type_(type),
+ is_relative_(false), is_symbolless_(false),
+ is_section_symbol_(false), shndx_(shndx)
+{
+ gold_assert(shndx != INVALID_CODE);
+ // this->type_ is a bitfield; make sure TYPE fits.
+ gold_assert(this->type_ == type);
+ this->u1_.arg = arg;
+ this->u2_.relobj = relobj;
+}
+
// Record that we need a dynamic symbol index for this relocation.
template<bool dynamic, int size, bool big_endian>
Output_reloc<elfcpp::SHT_REL, dynamic, size, big_endian>::
set_needs_dynsym_index()
{
- if (this->is_relative_)
+ if (this->is_symbolless_)
return;
switch (this->local_sym_index_)
{
this->u1_.os->set_needs_dynsym_index();
break;
+ case TARGET_CODE:
+ // The target must take care of this if necessary.
+ break;
+
case 0:
break;
const
{
unsigned int index;
+ if (this->is_symbolless_)
+ return 0;
switch (this->local_sym_index_)
{
case INVALID_CODE:
index = this->u1_.os->symtab_index();
break;
+ case TARGET_CODE:
+ index = parameters->target().reloc_symbol_index(this->u1_.arg,
+ this->type_);
+ break;
+
case 0:
// Relocations without symbols use a symbol index of 0.
index = 0;
{
gold_assert(this->local_sym_index_ != GSYM_CODE
&& this->local_sym_index_ != SECTION_CODE
+ && this->local_sym_index_ != TARGET_CODE
&& this->local_sym_index_ != INVALID_CODE
+ && this->local_sym_index_ != 0
&& this->is_section_symbol_);
const unsigned int lsi = this->local_sym_index_;
Output_section* os = this->u1_.relobj->output_section(lsi);
Write_rel* wr) const
{
wr->put_r_offset(this->get_address());
- unsigned int sym_index = this->is_relative_ ? 0 : this->get_symbol_index();
+ unsigned int sym_index = this->get_symbol_index();
wr->put_r_info(elfcpp::elf_r_info<size>(sym_index, this->type_));
}
return sym->value() + addend;
}
gold_assert(this->local_sym_index_ != SECTION_CODE
+ && this->local_sym_index_ != TARGET_CODE
&& this->local_sym_index_ != INVALID_CODE
+ && this->local_sym_index_ != 0
&& !this->is_section_symbol_);
const unsigned int lsi = this->local_sym_index_;
const Symbol_value<size>* symval = this->u1_.relobj->local_symbol(lsi);
elfcpp::Rela_write<size, big_endian> orel(pov);
this->rel_.write_rel(&orel);
Addend addend = this->addend_;
- if (this->rel_.is_relative())
+ if (this->rel_.is_target_specific())
+ addend = parameters->target().reloc_addend(this->rel_.target_arg(),
+ this->rel_.type(), addend);
+ else if (this->rel_.is_symbolless())
addend = this->rel_.symbol_value(addend);
else if (this->rel_.is_local_section_symbol())
addend = this->rel_.local_section_offset(addend);
os->set_entsize(elfcpp::Elf_sizes<size>::rela_size);
else
gold_unreachable();
- if (dynamic)
- os->set_should_link_to_dynsym();
- else
+
+ // A STT_GNU_IFUNC symbol may require a IRELATIVE reloc when doing a
+ // static link. The backends will generate a dynamic reloc section
+ // to hold this. In that case we don't want to link to the dynsym
+ // section, because there isn't one.
+ if (!dynamic)
os->set_should_link_to_symtab();
+ else if (parameters->doing_static_link())
+ ;
+ else
+ os->set_should_link_to_dynsym();
}
// Write out relocation data.
const off_t oview_size = this->data_size();
unsigned char* const oview = of->get_output_view(off, oview_size);
- if (this->sort_relocs_)
+ if (this->sort_relocs())
{
gold_assert(dynamic);
std::sort(this->relocs_.begin(), this->relocs_.end(),
section_size_type entry_count,
elfcpp::Elf_Word flags,
std::vector<unsigned int>* input_shndxes)
- : Output_section_data(entry_count * 4, 4),
+ : Output_section_data(entry_count * 4, 4, false),
relobj_(relobj),
flags_(flags)
{
// link-time value, which will be relocated dynamically by a
// RELATIVE relocation.
Symbol* gsym = this->u_.gsym;
- Sized_symbol<size>* sgsym;
- // This cast is a bit ugly. We don't want to put a
- // virtual method in Symbol, because we want Symbol to be
- // as small as possible.
- sgsym = static_cast<Sized_symbol<size>*>(gsym);
- val = sgsym->value();
+ if (this->use_plt_offset_ && gsym->has_plt_offset())
+ val = (parameters->target().plt_section_for_global(gsym)->address()
+ + gsym->plt_offset());
+ else
+ {
+ Sized_symbol<size>* sgsym;
+ // This cast is a bit ugly. We don't want to put a
+ // virtual method in Symbol, because we want Symbol to be
+ // as small as possible.
+ sgsym = static_cast<Sized_symbol<size>*>(gsym);
+ val = sgsym->value();
+ }
}
break;
default:
{
+ const Sized_relobj<size, big_endian>* object = this->u_.object;
const unsigned int lsi = this->local_sym_index_;
- const Symbol_value<size>* symval = this->u_.object->local_symbol(lsi);
- val = symval->value(this->u_.object, 0);
+ const Symbol_value<size>* symval = object->local_symbol(lsi);
+ if (!this->use_plt_offset_)
+ val = symval->value(this->u_.object, 0);
+ else
+ {
+ const Output_data* plt =
+ parameters->target().plt_section_for_local(object, lsi);
+ val = plt->address() + object->local_plt_offset(lsi);
+ }
}
break;
}
if (gsym->has_got_offset(got_type))
return false;
- this->entries_.push_back(Got_entry(gsym));
+ this->entries_.push_back(Got_entry(gsym, false));
+ this->set_got_size();
+ gsym->set_got_offset(got_type, this->last_got_offset());
+ return true;
+}
+
+// Like add_global, but use the PLT offset.
+
+template<int size, bool big_endian>
+bool
+Output_data_got<size, big_endian>::add_global_plt(Symbol* gsym,
+ unsigned int got_type)
+{
+ if (gsym->has_got_offset(got_type))
+ return false;
+
+ this->entries_.push_back(Got_entry(gsym, true));
this->set_got_size();
gsym->set_got_offset(got_type, this->last_got_offset());
return true;
// Add an entry for a global symbol to the GOT, and add a dynamic
// relocation of type R_TYPE for the GOT entry.
+
template<int size, bool big_endian>
void
Output_data_got<size, big_endian>::add_global_with_rel(
if (object->local_has_got_offset(symndx, got_type))
return false;
- this->entries_.push_back(Got_entry(object, symndx));
+ this->entries_.push_back(Got_entry(object, symndx, false));
+ this->set_got_size();
+ object->set_local_got_offset(symndx, got_type, this->last_got_offset());
+ return true;
+}
+
+// Like add_local, but use the PLT offset.
+
+template<int size, bool big_endian>
+bool
+Output_data_got<size, big_endian>::add_local_plt(
+ Sized_relobj<size, big_endian>* object,
+ unsigned int symndx,
+ unsigned int got_type)
+{
+ if (object->local_has_got_offset(symndx, got_type))
+ return false;
+
+ this->entries_.push_back(Got_entry(object, symndx, true));
this->set_got_size();
object->set_local_got_offset(symndx, got_type, this->last_got_offset());
return true;
// Add an entry for a local symbol to the GOT, and add a dynamic
// relocation of type R_TYPE for the GOT entry.
+
template<int size, bool big_endian>
void
Output_data_got<size, big_endian>::add_local_with_rel(
Output_section* os = object->output_section(shndx);
rel_dyn->add_output_section(os, r_type_1, this, got_offset);
- this->entries_.push_back(Got_entry(object, symndx));
+ this->entries_.push_back(Got_entry(object, symndx, false));
if (r_type_2 != 0)
{
got_offset = this->last_got_offset();
Output_section* os = object->output_section(shndx);
rela_dyn->add_output_section(os, r_type_1, this, got_offset, 0);
- this->entries_.push_back(Got_entry(object, symndx));
+ this->entries_.push_back(Got_entry(object, symndx, false));
if (r_type_2 != 0)
{
got_offset = this->last_got_offset();
case DYNAMIC_SECTION_SIZE:
val = this->u_.od->data_size();
+ if (this->od2 != NULL)
+ val += this->od2->data_size();
break;
case DYNAMIC_SYMBOL:
void
Output_data_dynamic::set_final_data_size()
{
- // Add the terminating entry.
- this->add_constant(elfcpp::DT_NULL, 0);
+ // Add the terminating entry if it hasn't been added.
+ // Because of relaxation, we can run this multiple times.
+ if (this->entries_.empty() || this->entries_.back().tag() != elfcpp::DT_NULL)
+ {
+ int extra = parameters->options().spare_dynamic_tags();
+ for (int i = 0; i < extra; ++i)
+ this->add_constant(elfcpp::DT_NULL, 0);
+ this->add_constant(elfcpp::DT_NULL, 0);
+ }
int dyn_size;
if (parameters->target().get_size() == 32)
for (Xindex_entries::const_iterator p = this->entries_.begin();
p != this->entries_.end();
++p)
- elfcpp::Swap<32, big_endian>::writeval(oview + p->first * 4, p->second);
+ {
+ unsigned int symndx = p->first;
+ gold_assert(symndx * 4 < this->data_size());
+ elfcpp::Swap<32, big_endian>::writeval(oview + symndx * 4, p->second);
+ }
}
// Output_section::Input_section methods.
return this->u2_.posd->data_size();
}
+// Return the object for an input section.
+
+Relobj*
+Output_section::Input_section::relobj() const
+{
+ if (this->is_input_section())
+ return this->u2_.object;
+ else if (this->is_merge_section())
+ {
+ gold_assert(this->u2_.pomb->first_relobj() != NULL);
+ return this->u2_.pomb->first_relobj();
+ }
+ else if (this->is_relaxed_input_section())
+ return this->u2_.poris->relobj();
+ else
+ gold_unreachable();
+}
+
+// Return the input section index for an input section.
+
+unsigned int
+Output_section::Input_section::shndx() const
+{
+ if (this->is_input_section())
+ return this->shndx_;
+ else if (this->is_merge_section())
+ {
+ gold_assert(this->u2_.pomb->first_relobj() != NULL);
+ return this->u2_.pomb->first_shndx();
+ }
+ else if (this->is_relaxed_input_section())
+ return this->u2_.poris->shndx();
+ else
+ gold_unreachable();
+}
+
// Set the address and file offset.
void
const Relobj* object,
unsigned int shndx,
section_offset_type offset,
- section_offset_type *poutput) const
+ section_offset_type* poutput) const
{
if (!this->is_input_section())
return this->u2_.posd->output_offset(object, shndx, offset, poutput);
this->u2_.posd->print_to_mapfile(mapfile);
break;
+ case RELAXED_INPUT_SECTION_CODE:
+ {
+ Output_relaxed_input_section* relaxed_section =
+ this->relaxed_input_section();
+ mapfile->print_input_section(relaxed_section->relobj(),
+ relaxed_section->shndx());
+ }
+ break;
default:
mapfile->print_input_section(this->u2_.object, this->shndx_);
break;
info_(0),
type_(type),
flags_(flags),
+ order_(ORDER_INVALID),
out_shndx_(-1U),
symtab_index_(0),
dynsym_index_(0),
found_in_sections_clause_(false),
has_load_address_(false),
info_uses_section_index_(false),
+ input_section_order_specified_(false),
may_sort_attached_input_sections_(false),
must_sort_attached_input_sections_(false),
attached_input_sections_are_sorted_(false),
is_relro_(false),
- is_relro_local_(false),
- tls_offset_(0)
+ is_small_section_(false),
+ is_large_section_(false),
+ generate_code_fills_at_write_(false),
+ is_entsize_zero_(false),
+ section_offsets_need_adjustment_(false),
+ is_noload_(false),
+ always_keeps_input_sections_(false),
+ tls_offset_(0),
+ checkpoint_(NULL),
+ lookup_maps_(new Output_section_lookup_maps)
{
// An unallocated section has no address. Forcing this means that
// we don't need special treatment for symbols defined in debug
Output_section::~Output_section()
{
+ delete this->checkpoint_;
}
// Set the entry size.
void
Output_section::set_entsize(uint64_t v)
{
- if (this->entsize_ == 0)
+ if (this->is_entsize_zero_)
+ ;
+ else if (this->entsize_ == 0)
this->entsize_ = v;
- else
- gold_assert(this->entsize_ == v);
+ else if (this->entsize_ != v)
+ {
+ this->entsize_ = 0;
+ this->is_entsize_zero_ = 1;
+ }
}
// Add the input section SHNDX, with header SHDR, named SECNAME, in
template<int size, bool big_endian>
off_t
-Output_section::add_input_section(Sized_relobj<size, big_endian>* object,
+Output_section::add_input_section(Layout* layout,
+ Sized_relobj<size, big_endian>* object,
unsigned int shndx,
const char* secname,
const elfcpp::Shdr<size, big_endian>& shdr,
this->addralign_ = addralign;
typename elfcpp::Elf_types<size>::Elf_WXword sh_flags = shdr.get_sh_flags();
- this->update_flags_for_input_section(sh_flags);
-
uint64_t entsize = shdr.get_sh_entsize();
// .debug_str is a mergeable string section, but is not always so
entsize = 1;
}
+ this->update_flags_for_input_section(sh_flags);
+ this->set_entsize(entsize);
+
// If this is a SHF_MERGE section, we pass all the input sections to
// a Output_data_merge. We don't try to handle relocations for such
// a section. We don't try to handle empty merge sections--they
&& reloc_shndx == 0
&& shdr.get_sh_size() > 0)
{
- if (this->add_merge_input_section(object, shndx, sh_flags,
- entsize, addralign))
+ // Keep information about merged input sections for rebuilding fast
+ // lookup maps if we have sections-script or we do relaxation.
+ bool keeps_input_sections = (this->always_keeps_input_sections_
+ || have_sections_script
+ || parameters->target().may_relax());
+
+ if (this->add_merge_input_section(object, shndx, sh_flags, entsize,
+ addralign, keeps_input_sections))
{
// Tell the relocation routines that they need to call the
// output_offset method to determine the final address.
off_t aligned_offset_in_section = align_address(offset_in_section,
addralign);
+ // Determine if we want to delay code-fill generation until the output
+ // section is written. When the target is relaxing, we want to delay fill
+ // generating to avoid adjusting them during relaxation. Also, if we are
+ // sorting input sections we must delay fill generation.
+ if (!this->generate_code_fills_at_write_
+ && !have_sections_script
+ && (sh_flags & elfcpp::SHF_EXECINSTR) != 0
+ && parameters->target().has_code_fill()
+ && (parameters->target().may_relax()
+ || parameters->options().section_ordering_file()))
+ {
+ gold_assert(this->fills_.empty());
+ this->generate_code_fills_at_write_ = true;
+ }
+
if (aligned_offset_in_section > offset_in_section
+ && !this->generate_code_fills_at_write_
&& !have_sections_script
&& (sh_flags & elfcpp::SHF_EXECINSTR) != 0
- && object->target()->has_code_fill())
+ && parameters->target().has_code_fill())
{
// We need to add some fill data. Using fill_list_ when
// possible is an optimization, since we will often have fill
this->fills_.push_back(Fill(offset_in_section, fill_len));
else
{
- // FIXME: When relaxing, the size needs to adjust to
- // maintain a constant alignment.
- std::string fill_data(object->target()->code_fill(fill_len));
+ std::string fill_data(parameters->target().code_fill(fill_len));
Output_data_const* odc = new Output_data_const(fill_data, 1);
this->input_sections_.push_back(Input_section(odc));
}
}
+ section_size_type input_section_size = shdr.get_sh_size();
+ section_size_type uncompressed_size;
+ if (object->section_is_compressed(shndx, &uncompressed_size))
+ input_section_size = uncompressed_size;
+
this->set_current_data_size_for_child(aligned_offset_in_section
- + shdr.get_sh_size());
+ + input_section_size);
// We need to keep track of this section if we are already keeping
// track of sections, or if we are relaxing. Also, if this is a
// section which requires sorting, or which may require sorting in
- // the future, we keep track of the sections. FIXME: Add test for
- // relaxing.
- if (have_sections_script
+ // the future, we keep track of the sections. If the
+ // --section-ordering-file option is used to specify the order of
+ // sections, we need to keep track of sections.
+ if (this->always_keeps_input_sections_
+ || have_sections_script
|| !this->input_sections_.empty()
|| this->may_sort_attached_input_sections()
|| this->must_sort_attached_input_sections()
- || parameters->options().user_set_Map())
- this->input_sections_.push_back(Input_section(object, shndx,
- shdr.get_sh_size(),
- addralign));
+ || parameters->options().user_set_Map()
+ || parameters->target().may_relax()
+ || parameters->options().section_ordering_file())
+ {
+ Input_section isecn(object, shndx, input_section_size, addralign);
+ if (parameters->options().section_ordering_file())
+ {
+ unsigned int section_order_index =
+ layout->find_section_order_index(std::string(secname));
+ if (section_order_index != 0)
+ {
+ isecn.set_section_order_index(section_order_index);
+ this->set_input_section_order_specified();
+ }
+ }
+ this->input_sections_.push_back(isecn);
+ }
return aligned_offset_in_section;
}
}
}
+// Add a relaxed input section.
+
+void
+Output_section::add_relaxed_input_section(Layout* layout,
+ Output_relaxed_input_section* poris,
+ const std::string& name)
+{
+ Input_section inp(poris);
+
+ // If the --section-ordering-file option is used to specify the order of
+ // sections, we need to keep track of sections.
+ if (parameters->options().section_ordering_file())
+ {
+ unsigned int section_order_index =
+ layout->find_section_order_index(name);
+ if (section_order_index != 0)
+ {
+ inp.set_section_order_index(section_order_index);
+ this->set_input_section_order_specified();
+ }
+ }
+
+ this->add_output_section_data(&inp);
+ if (this->lookup_maps_->is_valid())
+ this->lookup_maps_->add_relaxed_input_section(poris->relobj(),
+ poris->shndx(), poris);
+
+ // For a relaxed section, we use the current data size. Linker scripts
+ // get all the input sections, including relaxed one from an output
+ // section and add them back to them same output section to compute the
+ // output section size. If we do not account for sizes of relaxed input
+ // sections, an output section would be incorrectly sized.
+ off_t offset_in_section = this->current_data_size_for_child();
+ off_t aligned_offset_in_section = align_address(offset_in_section,
+ poris->addralign());
+ this->set_current_data_size_for_child(aligned_offset_in_section
+ + poris->current_data_size());
+}
+
// Add arbitrary data to an output section by Input_section.
void
bool
Output_section::add_merge_input_section(Relobj* object, unsigned int shndx,
uint64_t flags, uint64_t entsize,
- uint64_t addralign)
+ uint64_t addralign,
+ bool keeps_input_sections)
{
bool is_string = (flags & elfcpp::SHF_STRINGS) != 0;
if (is_string && addralign > entsize)
return false;
- Input_section_list::iterator p;
- for (p = this->input_sections_.begin();
- p != this->input_sections_.end();
- ++p)
- if (p->is_merge_section(is_string, entsize, addralign))
- {
- p->add_input_section(object, shndx);
- return true;
- }
-
- // We handle the actual constant merging in Output_merge_data or
- // Output_merge_string_data.
- Output_section_data* posd;
- if (!is_string)
- posd = new Output_merge_data(entsize, addralign);
+ // We cannot restore merged input section states.
+ gold_assert(this->checkpoint_ == NULL);
+
+ // Look up merge sections by required properties.
+ // Currently, we only invalidate the lookup maps in script processing
+ // and relaxation. We should not have done either when we reach here.
+ // So we assume that the lookup maps are valid to simply code.
+ gold_assert(this->lookup_maps_->is_valid());
+ Merge_section_properties msp(is_string, entsize, addralign);
+ Output_merge_base* pomb = this->lookup_maps_->find_merge_section(msp);
+ bool is_new = false;
+ if (pomb != NULL)
+ {
+ gold_assert(pomb->is_string() == is_string
+ && pomb->entsize() == entsize
+ && pomb->addralign() == addralign);
+ }
else
{
- switch (entsize)
+ // Create a new Output_merge_data or Output_merge_string_data.
+ if (!is_string)
+ pomb = new Output_merge_data(entsize, addralign);
+ else
{
- case 1:
- posd = new Output_merge_string<char>(addralign);
- break;
- case 2:
- posd = new Output_merge_string<uint16_t>(addralign);
- break;
- case 4:
- posd = new Output_merge_string<uint32_t>(addralign);
- break;
- default:
- return false;
+ switch (entsize)
+ {
+ case 1:
+ pomb = new Output_merge_string<char>(addralign);
+ break;
+ case 2:
+ pomb = new Output_merge_string<uint16_t>(addralign);
+ break;
+ case 4:
+ pomb = new Output_merge_string<uint32_t>(addralign);
+ break;
+ default:
+ return false;
+ }
}
+ // If we need to do script processing or relaxation, we need to keep
+ // the original input sections to rebuild the fast lookup maps.
+ if (keeps_input_sections)
+ pomb->set_keeps_input_sections();
+ is_new = true;
}
- this->add_output_merge_section(posd, is_string, entsize);
- posd->add_input_section(object, shndx);
+ if (pomb->add_input_section(object, shndx))
+ {
+ // Add new merge section to this output section and link merge
+ // section properties to new merge section in map.
+ if (is_new)
+ {
+ this->add_output_merge_section(pomb, is_string, entsize);
+ this->lookup_maps_->add_merge_section(msp, pomb);
+ }
- return true;
+ // Add input section to new merge section and link input section to new
+ // merge section in map.
+ this->lookup_maps_->add_merge_input_section(object, shndx, pomb);
+ return true;
+ }
+ else
+ {
+ // If add_input_section failed, delete new merge section to avoid
+ // exporting empty merge sections in Output_section::get_input_section.
+ if (is_new)
+ delete pomb;
+ return false;
+ }
}
-// Given an address OFFSET relative to the start of input section
-// SHNDX in OBJECT, return whether this address is being included in
-// the final link. This should only be called if SHNDX in OBJECT has
-// a special mapping.
+// Build a relaxation map to speed up relaxation of existing input sections.
+// Look up to the first LIMIT elements in INPUT_SECTIONS.
-bool
-Output_section::is_input_address_mapped(const Relobj* object,
- unsigned int shndx,
- off_t offset) const
+void
+Output_section::build_relaxation_map(
+ const Input_section_list& input_sections,
+ size_t limit,
+ Relaxation_map* relaxation_map) const
{
- for (Input_section_list::const_iterator p = this->input_sections_.begin();
- p != this->input_sections_.end();
- ++p)
+ for (size_t i = 0; i < limit; ++i)
{
- section_offset_type output_offset;
- if (p->output_offset(object, shndx, offset, &output_offset))
- return output_offset != -1;
+ const Input_section& is(input_sections[i]);
+ if (is.is_input_section() || is.is_relaxed_input_section())
+ {
+ Section_id sid(is.relobj(), is.shndx());
+ (*relaxation_map)[sid] = i;
+ }
}
-
- // By default we assume that the address is mapped. This should
- // only be called after we have passed all sections to Layout. At
- // that point we should know what we are discarding.
- return true;
}
-// Given an address OFFSET relative to the start of input section
-// SHNDX in object OBJECT, return the output offset relative to the
-// start of the input section in the output section. This should only
-// be called if SHNDX in OBJECT has a special mapping.
+// Convert regular input sections in INPUT_SECTIONS into relaxed input
+// sections in RELAXED_SECTIONS. MAP is a prebuilt map from section id
+// indices of INPUT_SECTIONS.
+
+void
+Output_section::convert_input_sections_in_list_to_relaxed_sections(
+ const std::vector<Output_relaxed_input_section*>& relaxed_sections,
+ const Relaxation_map& map,
+ Input_section_list* input_sections)
+{
+ for (size_t i = 0; i < relaxed_sections.size(); ++i)
+ {
+ Output_relaxed_input_section* poris = relaxed_sections[i];
+ Section_id sid(poris->relobj(), poris->shndx());
+ Relaxation_map::const_iterator p = map.find(sid);
+ gold_assert(p != map.end());
+ gold_assert((*input_sections)[p->second].is_input_section());
+
+ // Remember section order index of original input section
+ // if it is set. Copy it to the relaxed input section.
+ unsigned int soi =
+ (*input_sections)[p->second].section_order_index();
+ (*input_sections)[p->second] = Input_section(poris);
+ (*input_sections)[p->second].set_section_order_index(soi);
+ }
+}
+
+// Convert regular input sections into relaxed input sections. RELAXED_SECTIONS
+// is a vector of pointers to Output_relaxed_input_section or its derived
+// classes. The relaxed sections must correspond to existing input sections.
+
+void
+Output_section::convert_input_sections_to_relaxed_sections(
+ const std::vector<Output_relaxed_input_section*>& relaxed_sections)
+{
+ gold_assert(parameters->target().may_relax());
+
+ // We want to make sure that restore_states does not undo the effect of
+ // this. If there is no checkpoint active, just search the current
+ // input section list and replace the sections there. If there is
+ // a checkpoint, also replace the sections there.
+
+ // By default, we look at the whole list.
+ size_t limit = this->input_sections_.size();
+
+ if (this->checkpoint_ != NULL)
+ {
+ // Replace input sections with relaxed input section in the saved
+ // copy of the input section list.
+ if (this->checkpoint_->input_sections_saved())
+ {
+ Relaxation_map map;
+ this->build_relaxation_map(
+ *(this->checkpoint_->input_sections()),
+ this->checkpoint_->input_sections()->size(),
+ &map);
+ this->convert_input_sections_in_list_to_relaxed_sections(
+ relaxed_sections,
+ map,
+ this->checkpoint_->input_sections());
+ }
+ else
+ {
+ // We have not copied the input section list yet. Instead, just
+ // look at the portion that would be saved.
+ limit = this->checkpoint_->input_sections_size();
+ }
+ }
+
+ // Convert input sections in input_section_list.
+ Relaxation_map map;
+ this->build_relaxation_map(this->input_sections_, limit, &map);
+ this->convert_input_sections_in_list_to_relaxed_sections(
+ relaxed_sections,
+ map,
+ &this->input_sections_);
+
+ // Update fast look-up map.
+ if (this->lookup_maps_->is_valid())
+ for (size_t i = 0; i < relaxed_sections.size(); ++i)
+ {
+ Output_relaxed_input_section* poris = relaxed_sections[i];
+ this->lookup_maps_->add_relaxed_input_section(poris->relobj(),
+ poris->shndx(), poris);
+ }
+}
+
+// Update the output section flags based on input section flags.
+
+void
+Output_section::update_flags_for_input_section(elfcpp::Elf_Xword flags)
+{
+ // If we created the section with SHF_ALLOC clear, we set the
+ // address. If we are now setting the SHF_ALLOC flag, we need to
+ // undo that.
+ if ((this->flags_ & elfcpp::SHF_ALLOC) == 0
+ && (flags & elfcpp::SHF_ALLOC) != 0)
+ this->mark_address_invalid();
+
+ this->flags_ |= (flags
+ & (elfcpp::SHF_WRITE
+ | elfcpp::SHF_ALLOC
+ | elfcpp::SHF_EXECINSTR));
+
+ if ((flags & elfcpp::SHF_MERGE) == 0)
+ this->flags_ &=~ elfcpp::SHF_MERGE;
+ else
+ {
+ if (this->current_data_size_for_child() == 0)
+ this->flags_ |= elfcpp::SHF_MERGE;
+ }
+
+ if ((flags & elfcpp::SHF_STRINGS) == 0)
+ this->flags_ &=~ elfcpp::SHF_STRINGS;
+ else
+ {
+ if (this->current_data_size_for_child() == 0)
+ this->flags_ |= elfcpp::SHF_STRINGS;
+ }
+}
+
+// Find the merge section into which an input section with index SHNDX in
+// OBJECT has been added. Return NULL if none found.
+
+Output_section_data*
+Output_section::find_merge_section(const Relobj* object,
+ unsigned int shndx) const
+{
+ if (!this->lookup_maps_->is_valid())
+ this->build_lookup_maps();
+ return this->lookup_maps_->find_merge_section(object, shndx);
+}
+
+// Build the lookup maps for merge and relaxed sections. This is needs
+// to be declared as a const methods so that it is callable with a const
+// Output_section pointer. The method only updates states of the maps.
+
+void
+Output_section::build_lookup_maps() const
+{
+ this->lookup_maps_->clear();
+ for (Input_section_list::const_iterator p = this->input_sections_.begin();
+ p != this->input_sections_.end();
+ ++p)
+ {
+ if (p->is_merge_section())
+ {
+ Output_merge_base* pomb = p->output_merge_base();
+ Merge_section_properties msp(pomb->is_string(), pomb->entsize(),
+ pomb->addralign());
+ this->lookup_maps_->add_merge_section(msp, pomb);
+ for (Output_merge_base::Input_sections::const_iterator is =
+ pomb->input_sections_begin();
+ is != pomb->input_sections_end();
+ ++is)
+ {
+ const Const_section_id& csid = *is;
+ this->lookup_maps_->add_merge_input_section(csid.first,
+ csid.second, pomb);
+ }
+
+ }
+ else if (p->is_relaxed_input_section())
+ {
+ Output_relaxed_input_section* poris = p->relaxed_input_section();
+ this->lookup_maps_->add_relaxed_input_section(poris->relobj(),
+ poris->shndx(), poris);
+ }
+ }
+}
+
+// Find an relaxed input section corresponding to an input section
+// in OBJECT with index SHNDX.
+
+const Output_relaxed_input_section*
+Output_section::find_relaxed_input_section(const Relobj* object,
+ unsigned int shndx) const
+{
+ if (!this->lookup_maps_->is_valid())
+ this->build_lookup_maps();
+ return this->lookup_maps_->find_relaxed_input_section(object, shndx);
+}
+
+// Given an address OFFSET relative to the start of input section
+// SHNDX in OBJECT, return whether this address is being included in
+// the final link. This should only be called if SHNDX in OBJECT has
+// a special mapping.
+
+bool
+Output_section::is_input_address_mapped(const Relobj* object,
+ unsigned int shndx,
+ off_t offset) const
+{
+ // Look at the Output_section_data_maps first.
+ const Output_section_data* posd = this->find_merge_section(object, shndx);
+ if (posd == NULL)
+ posd = this->find_relaxed_input_section(object, shndx);
+
+ if (posd != NULL)
+ {
+ section_offset_type output_offset;
+ bool found = posd->output_offset(object, shndx, offset, &output_offset);
+ gold_assert(found);
+ return output_offset != -1;
+ }
+
+ // Fall back to the slow look-up.
+ for (Input_section_list::const_iterator p = this->input_sections_.begin();
+ p != this->input_sections_.end();
+ ++p)
+ {
+ section_offset_type output_offset;
+ if (p->output_offset(object, shndx, offset, &output_offset))
+ return output_offset != -1;
+ }
+
+ // By default we assume that the address is mapped. This should
+ // only be called after we have passed all sections to Layout. At
+ // that point we should know what we are discarding.
+ return true;
+}
+
+// Given an address OFFSET relative to the start of input section
+// SHNDX in object OBJECT, return the output offset relative to the
+// start of the input section in the output section. This should only
+// be called if SHNDX in OBJECT has a special mapping.
section_offset_type
Output_section::output_offset(const Relobj* object, unsigned int shndx,
section_offset_type offset) const
{
- // This can only be called meaningfully when layout is complete.
- gold_assert(Output_data::is_layout_complete());
+ // This can only be called meaningfully when we know the data size
+ // of this.
+ gold_assert(this->is_data_size_valid());
+
+ // Look at the Output_section_data_maps first.
+ const Output_section_data* posd = this->find_merge_section(object, shndx);
+ if (posd == NULL)
+ posd = this->find_relaxed_input_section(object, shndx);
+ if (posd != NULL)
+ {
+ section_offset_type output_offset;
+ bool found = posd->output_offset(object, shndx, offset, &output_offset);
+ gold_assert(found);
+ return output_offset;
+ }
+ // Fall back to the slow look-up.
for (Input_section_list::const_iterator p = this->input_sections_.begin();
p != this->input_sections_.end();
++p)
off_t offset) const
{
uint64_t addr = this->address() + this->first_input_offset_;
+
+ // Look at the Output_section_data_maps first.
+ const Output_section_data* posd = this->find_merge_section(object, shndx);
+ if (posd == NULL)
+ posd = this->find_relaxed_input_section(object, shndx);
+ if (posd != NULL && posd->is_address_valid())
+ {
+ section_offset_type output_offset;
+ bool found = posd->output_offset(object, shndx, offset, &output_offset);
+ gold_assert(found);
+ return posd->address() + output_offset;
+ }
+
+ // Fall back to the slow look-up.
for (Input_section_list::const_iterator p = this->input_sections_.begin();
p != this->input_sections_.end();
++p)
gold_unreachable();
}
-// Return the output address of the start of the merged section for
+// Find the output address of the start of the merged section for
// input section SHNDX in object OBJECT.
-uint64_t
-Output_section::starting_output_address(const Relobj* object,
- unsigned int shndx) const
+bool
+Output_section::find_starting_output_address(const Relobj* object,
+ unsigned int shndx,
+ uint64_t* paddr) const
{
+ // FIXME: This becomes a bottle-neck if we have many relaxed sections.
+ // Looking up the merge section map does not always work as we sometimes
+ // find a merge section without its address set.
uint64_t addr = this->address() + this->first_input_offset_;
for (Input_section_list::const_iterator p = this->input_sections_.begin();
p != this->input_sections_.end();
// Unfortunately we don't know for sure that input offset 0 is
// mapped at all.
if (p->is_merge_section_for(object, shndx))
- return addr;
+ {
+ *paddr = addr;
+ return true;
+ }
addr += p->data_size();
}
- gold_unreachable();
+
+ // We couldn't find a merge output section for this input section.
+ return false;
}
// Set the data size of an Output_section. This is where we handle
return;
}
- if (this->must_sort_attached_input_sections())
+ if (this->must_sort_attached_input_sections()
+ || this->input_section_order_specified())
this->sort_attached_input_sections();
uint64_t address = this->address();
void
Output_section::do_reset_address_and_file_offset()
{
+ // An unallocated section has no address. Forcing this means that
+ // we don't need special treatment for symbols defined in debug
+ // sections. We do the same in the constructor. This does not
+ // apply to NOLOAD sections though.
+ if (((this->flags_ & elfcpp::SHF_ALLOC) == 0) && !this->is_noload_)
+ this->set_address(0);
+
for (Input_section_list::iterator p = this->input_sections_.begin();
p != this->input_sections_.end();
++p)
p->reset_address_and_file_offset();
}
+
+// Return true if address and file offset have the values after reset.
+
+bool
+Output_section::do_address_and_file_offset_have_reset_values() const
+{
+ if (this->is_offset_valid())
+ return false;
+
+ // An unallocated section has address 0 after its construction or a reset.
+ if ((this->flags_ & elfcpp::SHF_ALLOC) == 0)
+ return this->is_address_valid() && this->address() == 0;
+ else
+ return !this->is_address_valid();
+}
// Set the TLS offset. Called only for SHT_TLS sections.
{ }
Input_section_sort_entry(const Input_section& input_section,
- unsigned int index)
+ unsigned int index,
+ bool must_sort_attached_input_sections)
: input_section_(input_section), index_(index),
- section_has_name_(input_section.is_input_section())
+ section_has_name_(input_section.is_input_section()
+ || input_section.is_relaxed_input_section())
{
- if (this->section_has_name_)
+ if (this->section_has_name_
+ && must_sort_attached_input_sections)
{
// This is only called single-threaded from Layout::finalize,
// so it is OK to lock. Unfortunately we have no way to pass
// in a Task token.
const Task* dummy_task = reinterpret_cast<const Task*>(-1);
- Object* obj = input_section.relobj();
+ Object* obj = (input_section.is_input_section()
+ ? input_section.relobj()
+ : input_section.relaxed_input_section()->relobj());
Task_lock_obj<Object> tl(dummy_task, obj);
// This is a slow operation, which should be cached in
has_priority() const
{
gold_assert(this->section_has_name_);
- return this->section_name_.find('.', 1);
+ return this->section_name_.find('.', 1) != std::string::npos;
}
// Return true if this an input file whose base name matches
return memcmp(base_name + base_len - 2, ".o", 2) == 0;
}
+ // Returns 1 if THIS should appear before S in section order, -1 if S
+ // appears before THIS and 0 if they are not comparable.
+ int
+ compare_section_ordering(const Input_section_sort_entry& s) const
+ {
+ unsigned int this_secn_index = this->input_section_.section_order_index();
+ unsigned int s_secn_index = s.input_section().section_order_index();
+ if (this_secn_index > 0 && s_secn_index > 0)
+ {
+ if (this_secn_index < s_secn_index)
+ return 1;
+ else if (this_secn_index > s_secn_index)
+ return -1;
+ }
+ return 0;
+ }
+
private:
// The Input_section we are sorting.
Input_section input_section_;
}
// A section with a priority follows a section without a priority.
- // The GNU linker does this for all but .init_array sections; until
- // further notice we'll assume that that is an mistake.
bool s1_has_priority = s1.has_priority();
bool s2_has_priority = s2.has_priority();
if (s1_has_priority && !s2_has_priority)
if (!s1_has_priority && s2_has_priority)
return true;
+ // Check if a section order exists for these sections through a section
+ // ordering file. If sequence_num is 0, an order does not exist.
+ int sequence_num = s1.compare_section_ordering(s2);
+ if (sequence_num != 0)
+ return sequence_num == 1;
+
+ // Otherwise we sort by name.
+ int compare = s1.section_name().compare(s2.section_name());
+ if (compare != 0)
+ return compare < 0;
+
+ // Otherwise we keep the input order.
+ return s1.index() < s2.index();
+}
+
+// Return true if S1 should come before S2 in an .init_array or .fini_array
+// output section.
+
+bool
+Output_section::Input_section_sort_init_fini_compare::operator()(
+ const Output_section::Input_section_sort_entry& s1,
+ const Output_section::Input_section_sort_entry& s2) const
+{
+ // We sort all the sections with no names to the end.
+ if (!s1.section_has_name() || !s2.section_has_name())
+ {
+ if (s1.section_has_name())
+ return true;
+ if (s2.section_has_name())
+ return false;
+ return s1.index() < s2.index();
+ }
+
+ // A section without a priority follows a section with a priority.
+ // This is the reverse of .ctors and .dtors sections.
+ bool s1_has_priority = s1.has_priority();
+ bool s2_has_priority = s2.has_priority();
+ if (s1_has_priority && !s2_has_priority)
+ return true;
+ if (!s1_has_priority && s2_has_priority)
+ return false;
+
+ // Check if a section order exists for these sections through a section
+ // ordering file. If sequence_num is 0, an order does not exist.
+ int sequence_num = s1.compare_section_ordering(s2);
+ if (sequence_num != 0)
+ return sequence_num == 1;
+
// Otherwise we sort by name.
int compare = s1.section_name().compare(s2.section_name());
if (compare != 0)
return s1.index() < s2.index();
}
+// Return true if S1 should come before S2. Sections that do not match
+// any pattern in the section ordering file are placed ahead of the sections
+// that match some pattern.
+
+bool
+Output_section::Input_section_sort_section_order_index_compare::operator()(
+ const Output_section::Input_section_sort_entry& s1,
+ const Output_section::Input_section_sort_entry& s2) const
+{
+ unsigned int s1_secn_index = s1.input_section().section_order_index();
+ unsigned int s2_secn_index = s2.input_section().section_order_index();
+
+ // Keep input order if section ordering cannot determine order.
+ if (s1_secn_index == s2_secn_index)
+ return s1.index() < s2.index();
+
+ return s1_secn_index < s2_secn_index;
+}
+
// Sort the input sections attached to an output section.
void
if (this->attached_input_sections_are_sorted_)
return;
+ if (this->checkpoint_ != NULL
+ && !this->checkpoint_->input_sections_saved())
+ this->checkpoint_->save_input_sections();
+
// The only thing we know about an input section is the object and
// the section index. We need the section name. Recomputing this
// is slow but this is an unusual case. If this becomes a speed
for (Input_section_list::iterator p = this->input_sections_.begin();
p != this->input_sections_.end();
++p, ++i)
- sort_list.push_back(Input_section_sort_entry(*p, i));
+ sort_list.push_back(Input_section_sort_entry(*p, i,
+ this->must_sort_attached_input_sections()));
// Sort the input sections.
- std::sort(sort_list.begin(), sort_list.end(), Input_section_sort_compare());
+ if (this->must_sort_attached_input_sections())
+ {
+ if (this->type() == elfcpp::SHT_PREINIT_ARRAY
+ || this->type() == elfcpp::SHT_INIT_ARRAY
+ || this->type() == elfcpp::SHT_FINI_ARRAY)
+ std::sort(sort_list.begin(), sort_list.end(),
+ Input_section_sort_init_fini_compare());
+ else
+ std::sort(sort_list.begin(), sort_list.end(),
+ Input_section_sort_compare());
+ }
+ else
+ {
+ gold_assert(parameters->options().section_ordering_file());
+ std::sort(sort_list.begin(), sort_list.end(),
+ Input_section_sort_section_order_index_compare());
+ }
// Copy the sorted input sections back to our list.
this->input_sections_.clear();
p != sort_list.end();
++p)
this->input_sections_.push_back(p->input_section());
+ sort_list.clear();
// Remember that we sorted the input sections, since we might get
// called again.
{
gold_assert(!this->requires_postprocessing());
+ // If the target performs relaxation, we delay filler generation until now.
+ gold_assert(!this->generate_code_fills_at_write_ || this->fills_.empty());
+
off_t output_section_file_offset = this->offset();
for (Fill_list::iterator p = this->fills_.begin();
p != this->fills_.end();
fill_data.data(), fill_data.size());
}
+ off_t off = this->offset() + this->first_input_offset_;
for (Input_section_list::iterator p = this->input_sections_.begin();
p != this->input_sections_.end();
++p)
- p->write(of);
+ {
+ off_t aligned_off = align_address(off, p->addralign());
+ if (this->generate_code_fills_at_write_ && (off != aligned_off))
+ {
+ size_t fill_len = aligned_off - off;
+ std::string fill_data(parameters->target().code_fill(fill_len));
+ of->write(off, fill_data.data(), fill_data.size());
+ }
+
+ p->write(of);
+ off = aligned_off + p->data_size();
+ }
}
// If a section requires postprocessing, create the buffer to use.
{
gold_assert(this->requires_postprocessing());
+ // If the target performs relaxation, we delay filler generation until now.
+ gold_assert(!this->generate_code_fills_at_write_ || this->fills_.empty());
+
unsigned char* buffer = this->postprocessing_buffer();
for (Fill_list::iterator p = this->fills_.begin();
p != this->fills_.end();
p != this->input_sections_.end();
++p)
{
- off = align_address(off, p->addralign());
- p->write_to_buffer(buffer + off);
- off += p->data_size();
+ off_t aligned_off = align_address(off, p->addralign());
+ if (this->generate_code_fills_at_write_ && (off != aligned_off))
+ {
+ size_t fill_len = aligned_off - off;
+ std::string fill_data(parameters->target().code_fill(fill_len));
+ memcpy(buffer + off, fill_data.data(), fill_data.size());
+ }
+
+ p->write_to_buffer(buffer + aligned_off);
+ off = aligned_off + p->data_size();
}
}
Output_section::get_input_sections(
uint64_t address,
const std::string& fill,
- std::list<std::pair<Relobj*, unsigned int> >* input_sections)
+ std::list<Input_section>* input_sections)
{
+ if (this->checkpoint_ != NULL
+ && !this->checkpoint_->input_sections_saved())
+ this->checkpoint_->save_input_sections();
+
+ // Invalidate fast look-up maps.
+ this->lookup_maps_->invalidate();
+
uint64_t orig_address = address;
address = align_address(address, this->addralign());
p != this->input_sections_.end();
++p)
{
- if (p->is_input_section())
- input_sections->push_back(std::make_pair(p->relobj(), p->shndx()));
+ if (p->is_input_section()
+ || p->is_relaxed_input_section()
+ || p->is_merge_section())
+ input_sections->push_back(*p);
else
{
uint64_t aligned_address = align_address(address, p->addralign());
return data_size;
}
-// Add an input section from a script.
+// Add a script input section. SIS is an Output_section::Input_section,
+// which can be either a plain input section or a special input section like
+// a relaxed input section. For a special input section, its size must be
+// finalized.
void
-Output_section::add_input_section_for_script(Relobj* object,
- unsigned int shndx,
- off_t data_size,
- uint64_t addralign)
+Output_section::add_script_input_section(const Input_section& sis)
{
+ uint64_t data_size = sis.data_size();
+ uint64_t addralign = sis.addralign();
if (addralign > this->addralign_)
this->addralign_ = addralign;
this->set_current_data_size_for_child(aligned_offset_in_section
+ data_size);
- this->input_sections_.push_back(Input_section(object, shndx,
- data_size, addralign));
+ this->input_sections_.push_back(sis);
+
+ // Update fast lookup maps if necessary.
+ if (this->lookup_maps_->is_valid())
+ {
+ if (sis.is_merge_section())
+ {
+ Output_merge_base* pomb = sis.output_merge_base();
+ Merge_section_properties msp(pomb->is_string(), pomb->entsize(),
+ pomb->addralign());
+ this->lookup_maps_->add_merge_section(msp, pomb);
+ for (Output_merge_base::Input_sections::const_iterator p =
+ pomb->input_sections_begin();
+ p != pomb->input_sections_end();
+ ++p)
+ this->lookup_maps_->add_merge_input_section(p->first, p->second,
+ pomb);
+ }
+ else if (sis.is_relaxed_input_section())
+ {
+ Output_relaxed_input_section* poris = sis.relaxed_input_section();
+ this->lookup_maps_->add_relaxed_input_section(poris->relobj(),
+ poris->shndx(), poris);
+ }
+ }
+}
+
+// Save states for relaxation.
+
+void
+Output_section::save_states()
+{
+ gold_assert(this->checkpoint_ == NULL);
+ Checkpoint_output_section* checkpoint =
+ new Checkpoint_output_section(this->addralign_, this->flags_,
+ this->input_sections_,
+ this->first_input_offset_,
+ this->attached_input_sections_are_sorted_);
+ this->checkpoint_ = checkpoint;
+ gold_assert(this->fills_.empty());
+}
+
+void
+Output_section::discard_states()
+{
+ gold_assert(this->checkpoint_ != NULL);
+ delete this->checkpoint_;
+ this->checkpoint_ = NULL;
+ gold_assert(this->fills_.empty());
+
+ // Simply invalidate the fast lookup maps since we do not keep
+ // track of them.
+ this->lookup_maps_->invalidate();
+}
+
+void
+Output_section::restore_states()
+{
+ gold_assert(this->checkpoint_ != NULL);
+ Checkpoint_output_section* checkpoint = this->checkpoint_;
+
+ this->addralign_ = checkpoint->addralign();
+ this->flags_ = checkpoint->flags();
+ this->first_input_offset_ = checkpoint->first_input_offset();
+
+ if (!checkpoint->input_sections_saved())
+ {
+ // If we have not copied the input sections, just resize it.
+ size_t old_size = checkpoint->input_sections_size();
+ gold_assert(this->input_sections_.size() >= old_size);
+ this->input_sections_.resize(old_size);
+ }
+ else
+ {
+ // We need to copy the whole list. This is not efficient for
+ // extremely large output with hundreads of thousands of input
+ // objects. We may need to re-think how we should pass sections
+ // to scripts.
+ this->input_sections_ = *checkpoint->input_sections();
+ }
+
+ this->attached_input_sections_are_sorted_ =
+ checkpoint->attached_input_sections_are_sorted();
+
+ // Simply invalidate the fast lookup maps since we do not keep
+ // track of them.
+ this->lookup_maps_->invalidate();
+}
+
+// Update the section offsets of input sections in this. This is required if
+// relaxation causes some input sections to change sizes.
+
+void
+Output_section::adjust_section_offsets()
+{
+ if (!this->section_offsets_need_adjustment_)
+ return;
+
+ off_t off = 0;
+ for (Input_section_list::iterator p = this->input_sections_.begin();
+ p != this->input_sections_.end();
+ ++p)
+ {
+ off = align_address(off, p->addralign());
+ if (p->is_input_section())
+ p->relobj()->set_section_offset(p->shndx(), off);
+ off += p->data_size();
+ }
+
+ this->section_offsets_need_adjustment_ = false;
}
// Print to the map file.
// Output segment methods.
Output_segment::Output_segment(elfcpp::Elf_Word type, elfcpp::Elf_Word flags)
- : output_data_(),
- output_bss_(),
- vaddr_(0),
+ : vaddr_(0),
paddr_(0),
memsz_(0),
max_align_(0),
type_(type),
flags_(flags),
is_max_align_known_(false),
- are_addresses_set_(false)
+ are_addresses_set_(false),
+ is_large_data_segment_(false)
{
+ // The ELF ABI specifies that a PT_TLS segment always has PF_R as
+ // the flags.
+ if (type == elfcpp::PT_TLS)
+ this->flags_ = elfcpp::PF_R;
}
-// Add an Output_section to an Output_segment.
+// Add an Output_section to a PT_LOAD Output_segment.
void
-Output_segment::add_output_section(Output_section* os,
- elfcpp::Elf_Word seg_flags)
+Output_segment::add_output_section_to_load(Layout* layout,
+ Output_section* os,
+ elfcpp::Elf_Word seg_flags)
{
+ gold_assert(this->type() == elfcpp::PT_LOAD);
gold_assert((os->flags() & elfcpp::SHF_ALLOC) != 0);
gold_assert(!this->is_max_align_known_);
+ gold_assert(os->is_large_data_section() == this->is_large_data_segment());
- // Update the segment flags.
- this->flags_ |= seg_flags;
+ this->update_flags_for_output_section(seg_flags);
- Output_segment::Output_data_list* pdl;
- if (os->type() == elfcpp::SHT_NOBITS)
- pdl = &this->output_bss_;
+ // We don't want to change the ordering if we have a linker script
+ // with a SECTIONS clause.
+ Output_section_order order = os->order();
+ if (layout->script_options()->saw_sections_clause())
+ order = static_cast<Output_section_order>(0);
else
- pdl = &this->output_data_;
+ gold_assert(order != ORDER_INVALID);
- // So that PT_NOTE segments will work correctly, we need to ensure
- // that all SHT_NOTE sections are adjacent. This will normally
- // happen automatically, because all the SHT_NOTE input sections
- // will wind up in the same output section. However, it is possible
- // for multiple SHT_NOTE input sections to have different section
- // flags, and thus be in different output sections, but for the
- // different section flags to map into the same segment flags and
- // thus the same output segment.
-
- // Note that while there may be many input sections in an output
- // section, there are normally only a few output sections in an
- // output segment. This loop is expected to be fast.
-
- if (os->type() == elfcpp::SHT_NOTE && !pdl->empty())
- {
- Output_segment::Output_data_list::iterator p = pdl->end();
- do
- {
- --p;
- if ((*p)->is_section_type(elfcpp::SHT_NOTE))
- {
- ++p;
- pdl->insert(p, os);
- return;
- }
- }
- while (p != pdl->begin());
- }
-
- // Similarly, so that PT_TLS segments will work, we need to group
- // SHF_TLS sections. An SHF_TLS/SHT_NOBITS section is a special
- // case: we group the SHF_TLS/SHT_NOBITS sections right after the
- // SHF_TLS/SHT_PROGBITS sections. This lets us set up PT_TLS
- // correctly. SHF_TLS sections get added to both a PT_LOAD segment
- // and the PT_TLS segment -- we do this grouping only for the
- // PT_LOAD segment.
- if (this->type_ != elfcpp::PT_TLS
- && (os->flags() & elfcpp::SHF_TLS) != 0)
- {
- pdl = &this->output_data_;
- bool nobits = os->type() == elfcpp::SHT_NOBITS;
- bool sawtls = false;
- Output_segment::Output_data_list::iterator p = pdl->end();
- do
- {
- --p;
- bool insert;
- if ((*p)->is_section_flag_set(elfcpp::SHF_TLS))
- {
- sawtls = true;
- // Put a NOBITS section after the first TLS section.
- // Put a PROGBITS section after the first TLS/PROGBITS
- // section.
- insert = nobits || !(*p)->is_section_type(elfcpp::SHT_NOBITS);
- }
- else
- {
- // If we've gone past the TLS sections, but we've seen a
- // TLS section, then we need to insert this section now.
- insert = sawtls;
- }
-
- if (insert)
- {
- ++p;
- pdl->insert(p, os);
- return;
- }
- }
- while (p != pdl->begin());
-
- // There are no TLS sections yet; put this one at the requested
- // location in the section list.
- }
+ this->output_lists_[order].push_back(os);
+}
- // For the PT_GNU_RELRO segment, we need to group relro sections,
- // and we need to put them before any non-relro sections. Also,
- // relro local sections go before relro non-local sections.
- if (parameters->options().relro() && os->is_relro())
- {
- gold_assert(pdl == &this->output_data_);
- Output_segment::Output_data_list::iterator p;
- for (p = pdl->begin(); p != pdl->end(); ++p)
- {
- if (!(*p)->is_section())
- break;
+// Add an Output_section to a non-PT_LOAD Output_segment.
- Output_section* pos = (*p)->output_section();
- if (!pos->is_relro()
- || (os->is_relro_local() && !pos->is_relro_local()))
- break;
- }
+void
+Output_segment::add_output_section_to_nonload(Output_section* os,
+ elfcpp::Elf_Word seg_flags)
+{
+ gold_assert(this->type() != elfcpp::PT_LOAD);
+ gold_assert((os->flags() & elfcpp::SHF_ALLOC) != 0);
+ gold_assert(!this->is_max_align_known_);
- pdl->insert(p, os);
- return;
- }
+ this->update_flags_for_output_section(seg_flags);
- pdl->push_back(os);
+ this->output_lists_[0].push_back(os);
}
// Remove an Output_section from this segment. It is an error if it
void
Output_segment::remove_output_section(Output_section* os)
{
- // We only need this for SHT_PROGBITS.
- gold_assert(os->type() == elfcpp::SHT_PROGBITS);
- for (Output_data_list::iterator p = this->output_data_.begin();
- p != this->output_data_.end();
- ++p)
- {
- if (*p == os)
- {
- this->output_data_.erase(p);
- return;
- }
- }
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ Output_data_list* pdl = &this->output_lists_[i];
+ for (Output_data_list::iterator p = pdl->begin(); p != pdl->end(); ++p)
+ {
+ if (*p == os)
+ {
+ pdl->erase(p);
+ return;
+ }
+ }
+ }
gold_unreachable();
}
-// Add an Output_data (which is not an Output_section) to the start of
-// a segment.
+// Add an Output_data (which need not be an Output_section) to the
+// start of a segment.
void
Output_segment::add_initial_output_data(Output_data* od)
{
gold_assert(!this->is_max_align_known_);
- this->output_data_.push_front(od);
+ Output_data_list::iterator p = this->output_lists_[0].begin();
+ this->output_lists_[0].insert(p, od);
}
-// Return whether the first data section is a relro section.
+// Return true if this segment has any sections which hold actual
+// data, rather than being a BSS section.
+
+bool
+Output_segment::has_any_data_sections() const
+{
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ const Output_data_list* pdl = &this->output_lists_[i];
+ for (Output_data_list::const_iterator p = pdl->begin();
+ p != pdl->end();
+ ++p)
+ {
+ if (!(*p)->is_section())
+ return true;
+ if ((*p)->output_section()->type() != elfcpp::SHT_NOBITS)
+ return true;
+ }
+ }
+ return false;
+}
+
+// Return whether the first data section (not counting TLS sections)
+// is a relro section.
bool
Output_segment::is_first_section_relro() const
{
- return (!this->output_data_.empty()
- && this->output_data_.front()->is_section()
- && this->output_data_.front()->output_section()->is_relro());
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ if (i == static_cast<int>(ORDER_TLS_DATA)
+ || i == static_cast<int>(ORDER_TLS_BSS))
+ continue;
+ const Output_data_list* pdl = &this->output_lists_[i];
+ if (!pdl->empty())
+ {
+ Output_data* p = pdl->front();
+ return p->is_section() && p->output_section()->is_relro();
+ }
+ }
+ return false;
}
// Return the maximum alignment of the Output_data in Output_segment.
{
if (!this->is_max_align_known_)
{
- uint64_t addralign;
-
- addralign = Output_segment::maximum_alignment_list(&this->output_data_);
- if (addralign > this->max_align_)
- this->max_align_ = addralign;
-
- addralign = Output_segment::maximum_alignment_list(&this->output_bss_);
- if (addralign > this->max_align_)
- this->max_align_ = addralign;
-
- // If -z relro is in effect, and the first section in this
- // segment is a relro section, then the segment must be aligned
- // to at least the common page size. This ensures that the
- // PT_GNU_RELRO segment will start at a page boundary.
- if (this->type_ == elfcpp::PT_LOAD
- && parameters->options().relro()
- && this->is_first_section_relro())
- {
- addralign = parameters->target().common_pagesize();
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ const Output_data_list* pdl = &this->output_lists_[i];
+ uint64_t addralign = Output_segment::maximum_alignment_list(pdl);
if (addralign > this->max_align_)
this->max_align_ = addralign;
}
-
this->is_max_align_known_ = true;
}
return ret;
}
-// Return the number of dynamic relocs applied to this segment.
+// Return whether this segment has any dynamic relocs.
-unsigned int
-Output_segment::dynamic_reloc_count() const
+bool
+Output_segment::has_dynamic_reloc() const
{
- return (this->dynamic_reloc_count_list(&this->output_data_)
- + this->dynamic_reloc_count_list(&this->output_bss_));
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ if (this->has_dynamic_reloc_list(&this->output_lists_[i]))
+ return true;
+ return false;
}
-// Return the number of dynamic relocs applied to an Output_data_list.
+// Return whether this Output_data_list has any dynamic relocs.
-unsigned int
-Output_segment::dynamic_reloc_count_list(const Output_data_list* pdl) const
+bool
+Output_segment::has_dynamic_reloc_list(const Output_data_list* pdl) const
{
- unsigned int count = 0;
for (Output_data_list::const_iterator p = pdl->begin();
p != pdl->end();
++p)
- count += (*p)->dynamic_reloc_count();
- return count;
+ if ((*p)->has_dynamic_reloc())
+ return true;
+ return false;
}
// Set the section addresses for an Output_segment. If RESET is true,
// reset the addresses first. ADDR is the address and *POFF is the
// file offset. Set the section indexes starting with *PSHNDX.
-// Return the address of the immediately following segment. Update
-// *POFF and *PSHNDX.
+// INCREASE_RELRO is the size of the portion of the first non-relro
+// section that should be included in the PT_GNU_RELRO segment.
+// If this segment has relro sections, and has been aligned for
+// that purpose, set *HAS_RELRO to TRUE. Return the address of
+// the immediately following segment. Update *HAS_RELRO, *POFF,
+// and *PSHNDX.
uint64_t
Output_segment::set_section_addresses(const Layout* layout, bool reset,
- uint64_t addr, off_t* poff,
+ uint64_t addr,
+ unsigned int* increase_relro,
+ bool* has_relro,
+ off_t* poff,
unsigned int* pshndx)
{
gold_assert(this->type_ == elfcpp::PT_LOAD);
+ uint64_t last_relro_pad = 0;
+ off_t orig_off = *poff;
+
+ bool in_tls = false;
+
+ // If we have relro sections, we need to pad forward now so that the
+ // relro sections plus INCREASE_RELRO end on a common page boundary.
+ if (parameters->options().relro()
+ && this->is_first_section_relro()
+ && (!this->are_addresses_set_ || reset))
+ {
+ uint64_t relro_size = 0;
+ off_t off = *poff;
+ uint64_t max_align = 0;
+ for (int i = 0; i <= static_cast<int>(ORDER_RELRO_LAST); ++i)
+ {
+ Output_data_list* pdl = &this->output_lists_[i];
+ Output_data_list::iterator p;
+ for (p = pdl->begin(); p != pdl->end(); ++p)
+ {
+ if (!(*p)->is_section())
+ break;
+ uint64_t align = (*p)->addralign();
+ if (align > max_align)
+ max_align = align;
+ if ((*p)->is_section_flag_set(elfcpp::SHF_TLS))
+ in_tls = true;
+ else if (in_tls)
+ {
+ // Align the first non-TLS section to the alignment
+ // of the TLS segment.
+ align = max_align;
+ in_tls = false;
+ }
+ relro_size = align_address(relro_size, align);
+ // Ignore the size of the .tbss section.
+ if ((*p)->is_section_flag_set(elfcpp::SHF_TLS)
+ && (*p)->is_section_type(elfcpp::SHT_NOBITS))
+ continue;
+ if ((*p)->is_address_valid())
+ relro_size += (*p)->data_size();
+ else
+ {
+ // FIXME: This could be faster.
+ (*p)->set_address_and_file_offset(addr + relro_size,
+ off + relro_size);
+ relro_size += (*p)->data_size();
+ (*p)->reset_address_and_file_offset();
+ }
+ }
+ if (p != pdl->end())
+ break;
+ }
+ relro_size += *increase_relro;
+ // Pad the total relro size to a multiple of the maximum
+ // section alignment seen.
+ uint64_t aligned_size = align_address(relro_size, max_align);
+ // Note the amount of padding added after the last relro section.
+ last_relro_pad = aligned_size - relro_size;
+ *has_relro = true;
+
+ uint64_t page_align = parameters->target().common_pagesize();
+
+ // Align to offset N such that (N + RELRO_SIZE) % PAGE_ALIGN == 0.
+ uint64_t desired_align = page_align - (aligned_size % page_align);
+ if (desired_align < *poff % page_align)
+ *poff += page_align - *poff % page_align;
+ *poff += desired_align - *poff % page_align;
+ addr += *poff - orig_off;
+ orig_off = *poff;
+ }
+
if (!reset && this->are_addresses_set_)
{
gold_assert(this->paddr_ == addr);
this->are_addresses_set_ = true;
}
- bool in_tls = false;
-
- bool in_relro = (parameters->options().relro()
- && this->is_first_section_relro());
+ in_tls = false;
- off_t orig_off = *poff;
this->offset_ = orig_off;
- addr = this->set_section_list_addresses(layout, reset, &this->output_data_,
- addr, poff, pshndx, &in_tls,
- &in_relro);
- this->filesz_ = *poff - orig_off;
-
- off_t off = *poff;
+ off_t off = 0;
+ uint64_t ret;
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ if (i == static_cast<int>(ORDER_RELRO_LAST))
+ {
+ *poff += last_relro_pad;
+ addr += last_relro_pad;
+ if (this->output_lists_[i].empty())
+ {
+ // If there is nothing in the ORDER_RELRO_LAST list,
+ // the padding will occur at the end of the relro
+ // segment, and we need to add it to *INCREASE_RELRO.
+ *increase_relro += last_relro_pad;
+ }
+ }
+ addr = this->set_section_list_addresses(layout, reset,
+ &this->output_lists_[i],
+ addr, poff, pshndx, &in_tls);
+ if (i < static_cast<int>(ORDER_SMALL_BSS))
+ {
+ this->filesz_ = *poff - orig_off;
+ off = *poff;
+ }
- uint64_t ret = this->set_section_list_addresses(layout, reset,
- &this->output_bss_,
- addr, poff, pshndx,
- &in_tls, &in_relro);
+ ret = addr;
+ }
// If the last section was a TLS section, align upward to the
// alignment of the TLS segment, so that the overall size of the TLS
*poff = align_address(*poff, segment_align);
}
- // If all the sections were relro sections, align upward to the
- // common page size.
- if (in_relro)
- {
- uint64_t page_align = parameters->target().common_pagesize();
- *poff = align_address(*poff, page_align);
- }
-
this->memsz_ = *poff - orig_off;
// Ignore the file offset adjustments made by the BSS Output_data
Output_data_list* pdl,
uint64_t addr, off_t* poff,
unsigned int* pshndx,
- bool* in_tls, bool* in_relro)
+ bool* in_tls)
{
off_t startoff = *poff;
}
}
- // If this is a non-relro section after a relro section,
- // align it to a common page boundary so that the dynamic
- // linker has a page to mark as read-only.
- if (*in_relro
- && (!(*p)->is_section()
- || !(*p)->output_section()->is_relro()))
- {
- uint64_t page_align = parameters->target().common_pagesize();
- if (page_align > align)
- align = page_align;
- *in_relro = false;
- }
-
off = align_address(off, align);
(*p)->set_address_and_file_offset(addr + (off - startoff), off);
}
{
// The script may have inserted a skip forward, but it
// better not have moved backward.
- gold_assert((*p)->address() >= addr + (off - startoff));
- off += (*p)->address() - (addr + (off - startoff));
+ if ((*p)->address() >= addr + (off - startoff))
+ off += (*p)->address() - (addr + (off - startoff));
+ else
+ {
+ if (!layout->script_options()->saw_sections_clause())
+ gold_unreachable();
+ else
+ {
+ Output_section* os = (*p)->output_section();
+
+ // Cast to unsigned long long to avoid format warnings.
+ unsigned long long previous_dot =
+ static_cast<unsigned long long>(addr + (off - startoff));
+ unsigned long long dot =
+ static_cast<unsigned long long>((*p)->address());
+
+ if (os == NULL)
+ gold_error(_("dot moves backward in linker script "
+ "from 0x%llx to 0x%llx"), previous_dot, dot);
+ else
+ gold_error(_("address of section '%s' moves backward "
+ "from 0x%llx to 0x%llx"),
+ os->name(), previous_dot, dot);
+ }
+ }
(*p)->set_file_offset(off);
(*p)->finalize_data_size();
}
}
// For a non-PT_LOAD segment, set the offset from the sections, if
-// any.
+// any. Add INCREASE to the file size and the memory size.
void
-Output_segment::set_offset()
+Output_segment::set_offset(unsigned int increase)
{
gold_assert(this->type_ != elfcpp::PT_LOAD);
gold_assert(!this->are_addresses_set_);
- if (this->output_data_.empty() && this->output_bss_.empty())
+ // A non-load section only uses output_lists_[0].
+
+ Output_data_list* pdl = &this->output_lists_[0];
+
+ if (pdl->empty())
{
+ gold_assert(increase == 0);
this->vaddr_ = 0;
this->paddr_ = 0;
this->are_addresses_set_ = true;
return;
}
- const Output_data* first;
- if (this->output_data_.empty())
- first = this->output_bss_.front();
- else
- first = this->output_data_.front();
+ // Find the first and last section by address.
+ const Output_data* first = NULL;
+ const Output_data* last_data = NULL;
+ const Output_data* last_bss = NULL;
+ for (Output_data_list::const_iterator p = pdl->begin();
+ p != pdl->end();
+ ++p)
+ {
+ if (first == NULL
+ || (*p)->address() < first->address()
+ || ((*p)->address() == first->address()
+ && (*p)->data_size() < first->data_size()))
+ first = *p;
+ const Output_data** plast;
+ if ((*p)->is_section()
+ && (*p)->output_section()->type() == elfcpp::SHT_NOBITS)
+ plast = &last_bss;
+ else
+ plast = &last_data;
+ if (*plast == NULL
+ || (*p)->address() > (*plast)->address()
+ || ((*p)->address() == (*plast)->address()
+ && (*p)->data_size() > (*plast)->data_size()))
+ *plast = *p;
+ }
+
this->vaddr_ = first->address();
this->paddr_ = (first->has_load_address()
? first->load_address()
this->are_addresses_set_ = true;
this->offset_ = first->offset();
- if (this->output_data_.empty())
+ if (last_data == NULL)
this->filesz_ = 0;
else
- {
- const Output_data* last_data = this->output_data_.back();
- this->filesz_ = (last_data->address()
- + last_data->data_size()
- - this->vaddr_);
- }
+ this->filesz_ = (last_data->address()
+ + last_data->data_size()
+ - this->vaddr_);
- const Output_data* last;
- if (this->output_bss_.empty())
- last = this->output_data_.back();
- else
- last = this->output_bss_.back();
+ const Output_data* last = last_bss != NULL ? last_bss : last_data;
this->memsz_ = (last->address()
+ last->data_size()
- this->vaddr_);
+ this->filesz_ += increase;
+ this->memsz_ += increase;
+
+ // If this is a RELRO segment, verify that the segment ends at a
+ // page boundary.
+ if (this->type_ == elfcpp::PT_GNU_RELRO)
+ {
+ uint64_t page_align = parameters->target().common_pagesize();
+ uint64_t segment_end = this->vaddr_ + this->memsz_;
+ gold_assert(segment_end == align_address(segment_end, page_align));
+ }
+
// If this is a TLS segment, align the memory size. The code in
// set_section_list ensures that the section after the TLS segment
// is aligned to give us room.
gold_assert(this->vaddr_ == align_address(this->vaddr_, segment_align));
this->memsz_ = align_address(this->memsz_, segment_align);
}
-
- // If this is a RELRO segment, align the memory size. The code in
- // set_section_list ensures that the section after the RELRO segment
- // is aligned to give us room.
- if (this->type_ == elfcpp::PT_GNU_RELRO)
- {
- uint64_t page_align = parameters->target().common_pagesize();
- gold_assert(this->vaddr_ == align_address(this->vaddr_, page_align));
- this->memsz_ = align_address(this->memsz_, page_align);
- }
}
// Set the TLS offsets of the sections in the PT_TLS segment.
{
gold_assert(this->type_ == elfcpp::PT_TLS);
- for (Output_data_list::iterator p = this->output_data_.begin();
- p != this->output_data_.end();
- ++p)
- (*p)->set_tls_offset(this->vaddr_);
-
- for (Output_data_list::iterator p = this->output_bss_.begin();
- p != this->output_bss_.end();
+ for (Output_data_list::iterator p = this->output_lists_[0].begin();
+ p != this->output_lists_[0].end();
++p)
(*p)->set_tls_offset(this->vaddr_);
}
-// Return the address of the first section.
+// Return the load address of the first section.
uint64_t
Output_segment::first_section_load_address() const
{
- for (Output_data_list::const_iterator p = this->output_data_.begin();
- p != this->output_data_.end();
- ++p)
- if ((*p)->is_section())
- return (*p)->has_load_address() ? (*p)->load_address() : (*p)->address();
-
- for (Output_data_list::const_iterator p = this->output_bss_.begin();
- p != this->output_bss_.end();
- ++p)
- if ((*p)->is_section())
- return (*p)->has_load_address() ? (*p)->load_address() : (*p)->address();
-
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ const Output_data_list* pdl = &this->output_lists_[i];
+ for (Output_data_list::const_iterator p = pdl->begin();
+ p != pdl->end();
+ ++p)
+ {
+ if ((*p)->is_section())
+ return ((*p)->has_load_address()
+ ? (*p)->load_address()
+ : (*p)->address());
+ }
+ }
gold_unreachable();
}
unsigned int
Output_segment::output_section_count() const
{
- return (this->output_section_count_list(&this->output_data_)
- + this->output_section_count_list(&this->output_bss_));
+ unsigned int ret = 0;
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ ret += this->output_section_count_list(&this->output_lists_[i]);
+ return ret;
}
// Return the number of Output_sections in an Output_data_list.
{
Output_section* found = NULL;
uint64_t found_lma = 0;
- this->lowest_load_address_in_list(&this->output_data_, &found, &found_lma);
-
- Output_section* found_data = found;
- this->lowest_load_address_in_list(&this->output_bss_, &found, &found_lma);
- if (found != found_data && found_data != NULL)
- {
- gold_error(_("nobits section %s may not precede progbits section %s "
- "in same segment"),
- found->name(), found_data->name());
- return NULL;
- }
-
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ this->lowest_load_address_in_list(&this->output_lists_[i], &found,
+ &found_lma);
return found;
}
Output_segment::write_section_headers(const Layout* layout,
const Stringpool* secnamepool,
unsigned char* v,
- unsigned int *pshndx) const
+ unsigned int* pshndx) const
{
// Every section that is attached to a segment must be attached to a
// PT_LOAD segment, so we only write out section headers for PT_LOAD
if (this->type_ != elfcpp::PT_LOAD)
return v;
- v = this->write_section_headers_list<size, big_endian>(layout, secnamepool,
- &this->output_data_,
- v, pshndx);
- v = this->write_section_headers_list<size, big_endian>(layout, secnamepool,
- &this->output_bss_,
- v, pshndx);
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ {
+ const Output_data_list* pdl = &this->output_lists_[i];
+ v = this->write_section_headers_list<size, big_endian>(layout,
+ secnamepool,
+ pdl,
+ v, pshndx);
+ }
+
return v;
}
{
if (this->type() != elfcpp::PT_LOAD)
return;
- this->print_section_list_to_mapfile(mapfile, &this->output_data_);
- this->print_section_list_to_mapfile(mapfile, &this->output_bss_);
+ for (int i = 0; i < static_cast<int>(ORDER_MAX); ++i)
+ this->print_section_list_to_mapfile(mapfile, &this->output_lists_[i]);
}
// Print an output section list to the map file.
{
}
+// Try to open an existing file. Returns false if the file doesn't
+// exist, has a size of 0 or can't be mmapped.
+
+bool
+Output_file::open_for_modification()
+{
+ // The name "-" means "stdout".
+ if (strcmp(this->name_, "-") == 0)
+ return false;
+
+ // Don't bother opening files with a size of zero.
+ struct stat s;
+ if (::stat(this->name_, &s) != 0 || s.st_size == 0)
+ return false;
+
+ int o = open_descriptor(-1, this->name_, O_RDWR, 0);
+ if (o < 0)
+ gold_fatal(_("%s: open: %s"), this->name_, strerror(errno));
+ this->o_ = o;
+ this->file_size_ = s.st_size;
+
+ // If the file can't be mmapped, copying the content to an anonymous
+ // map will probably negate the performance benefits of incremental
+ // linking. This could be helped by using views and loading only
+ // the necessary parts, but this is not supported as of now.
+ if (!this->map_no_anonymous())
+ {
+ release_descriptor(o, true);
+ this->o_ = -1;
+ this->file_size_ = 0;
+ return false;
+ }
+
+ return true;
+}
+
// Open the output file.
void
else
{
struct stat s;
- if (::stat(this->name_, &s) == 0 && s.st_size != 0)
- unlink_if_ordinary(this->name_);
+ if (::stat(this->name_, &s) == 0
+ && (S_ISREG (s.st_mode) || S_ISLNK (s.st_mode)))
+ {
+ if (s.st_size != 0)
+ ::unlink(this->name_);
+ else if (!parameters->options().relocatable())
+ {
+ // If we don't unlink the existing file, add execute
+ // permission where read permissions already exist
+ // and where the umask permits.
+ int mask = ::umask(0);
+ ::umask(mask);
+ s.st_mode |= (s.st_mode & 0444) >> 2;
+ ::chmod(this->name_, s.st_mode & ~mask);
+ }
+ }
int mode = parameters->options().relocatable() ? 0666 : 0777;
int o = open_descriptor(-1, this->name_, O_RDWR | O_CREAT | O_TRUNC,
{
this->unmap();
this->file_size_ = file_size;
- this->map();
+ if (!this->map_no_anonymous())
+ gold_fatal(_("%s: mmap: %s"), this->name_, strerror(errno));
}
}
-// Map the file into memory.
+// Map an anonymous block of memory which will later be written to the
+// file. Return whether the map succeeded.
-void
-Output_file::map()
+bool
+Output_file::map_anonymous()
+{
+ void* base = ::mmap(NULL, this->file_size_, PROT_READ | PROT_WRITE,
+ MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+ if (base != MAP_FAILED)
+ {
+ this->map_is_anonymous_ = true;
+ this->base_ = static_cast<unsigned char*>(base);
+ return true;
+ }
+ return false;
+}
+
+// Map the file into memory. Return whether the mapping succeeded.
+
+bool
+Output_file::map_no_anonymous()
{
const int o = this->o_;
|| ::fstat(o, &statbuf) != 0
|| !S_ISREG(statbuf.st_mode)
|| this->is_temporary_)
- {
- this->map_is_anonymous_ = true;
- base = ::mmap(NULL, this->file_size_, PROT_READ | PROT_WRITE,
- MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
- }
- else
- {
- // Write out one byte to make the file the right size.
- if (::lseek(o, this->file_size_ - 1, SEEK_SET) < 0)
- gold_fatal(_("%s: lseek: %s"), this->name_, strerror(errno));
- char b = 0;
- if (::write(o, &b, 1) != 1)
- gold_fatal(_("%s: write: %s"), this->name_, strerror(errno));
+ return false;
- // Map the file into memory.
- this->map_is_anonymous_ = false;
- base = ::mmap(NULL, this->file_size_, PROT_READ | PROT_WRITE,
- MAP_SHARED, o, 0);
- }
+ // Ensure that we have disk space available for the file. If we
+ // don't do this, it is possible that we will call munmap, close,
+ // and exit with dirty buffers still in the cache with no assigned
+ // disk blocks. If the disk is out of space at that point, the
+ // output file will wind up incomplete, but we will have already
+ // exited. The alternative to fallocate would be to use fdatasync,
+ // but that would be a more significant performance hit.
+ if (::posix_fallocate(o, 0, this->file_size_) < 0)
+ gold_fatal(_("%s: %s"), this->name_, strerror(errno));
+
+ // Map the file into memory.
+ base = ::mmap(NULL, this->file_size_, PROT_READ | PROT_WRITE,
+ MAP_SHARED, o, 0);
+
+ // The mmap call might fail because of file system issues: the file
+ // system might not support mmap at all, or it might not support
+ // mmap with PROT_WRITE.
if (base == MAP_FAILED)
- gold_fatal(_("%s: mmap: %s"), this->name_, strerror(errno));
+ return false;
+
+ this->map_is_anonymous_ = false;
this->base_ = static_cast<unsigned char*>(base);
+ return true;
+}
+
+// Map the file into memory.
+
+void
+Output_file::map()
+{
+ if (this->map_no_anonymous())
+ return;
+
+ // The mmap call might fail because of file system issues: the file
+ // system might not support mmap at all, or it might not support
+ // mmap with PROT_WRITE. I'm not sure which errno values we will
+ // see in all cases, so if the mmap fails for any reason and we
+ // don't care about file contents, try for an anonymous map.
+ if (this->map_anonymous())
+ return;
+
+ gold_fatal(_("%s: mmap: failed to allocate %lu bytes for output file: %s"),
+ this->name_, static_cast<unsigned long>(this->file_size_),
+ strerror(errno));
}
// Unmap the file from memory.
if (this->map_is_anonymous_ && !this->is_temporary_)
{
size_t bytes_to_write = this->file_size_;
+ size_t offset = 0;
while (bytes_to_write > 0)
{
- ssize_t bytes_written = ::write(this->o_, this->base_, bytes_to_write);
+ ssize_t bytes_written = ::write(this->o_, this->base_ + offset,
+ bytes_to_write);
if (bytes_written == 0)
gold_error(_("%s: write: unexpected 0 return-value"), this->name_);
else if (bytes_written < 0)
gold_error(_("%s: write: %s"), this->name_, strerror(errno));
else
- bytes_to_write -= bytes_written;
+ {
+ bytes_to_write -= bytes_written;
+ offset += bytes_written;
+ }
}
}
this->unmap();
template
off_t
Output_section::add_input_section<32, false>(
+ Layout* layout,
Sized_relobj<32, false>* object,
unsigned int shndx,
const char* secname,
template
off_t
Output_section::add_input_section<32, true>(
+ Layout* layout,
Sized_relobj<32, true>* object,
unsigned int shndx,
const char* secname,
template
off_t
Output_section::add_input_section<64, false>(
+ Layout* layout,
Sized_relobj<64, false>* object,
unsigned int shndx,
const char* secname,
template
off_t
Output_section::add_input_section<64, true>(
+ Layout* layout,
Sized_relobj<64, true>* object,
unsigned int shndx,
const char* secname,
bool have_sections_script);
#endif
+#ifdef HAVE_TARGET_32_LITTLE
+template
+class Output_reloc<elfcpp::SHT_REL, false, 32, false>;
+#endif
+
+#ifdef HAVE_TARGET_32_BIG
+template
+class Output_reloc<elfcpp::SHT_REL, false, 32, true>;
+#endif
+
+#ifdef HAVE_TARGET_64_LITTLE
+template
+class Output_reloc<elfcpp::SHT_REL, false, 64, false>;
+#endif
+
+#ifdef HAVE_TARGET_64_BIG
+template
+class Output_reloc<elfcpp::SHT_REL, false, 64, true>;
+#endif
+
+#ifdef HAVE_TARGET_32_LITTLE
+template
+class Output_reloc<elfcpp::SHT_REL, true, 32, false>;
+#endif
+
+#ifdef HAVE_TARGET_32_BIG
+template
+class Output_reloc<elfcpp::SHT_REL, true, 32, true>;
+#endif
+
+#ifdef HAVE_TARGET_64_LITTLE
+template
+class Output_reloc<elfcpp::SHT_REL, true, 64, false>;
+#endif
+
+#ifdef HAVE_TARGET_64_BIG
+template
+class Output_reloc<elfcpp::SHT_REL, true, 64, true>;
+#endif
+
+#ifdef HAVE_TARGET_32_LITTLE
+template
+class Output_reloc<elfcpp::SHT_RELA, false, 32, false>;
+#endif
+
+#ifdef HAVE_TARGET_32_BIG
+template
+class Output_reloc<elfcpp::SHT_RELA, false, 32, true>;
+#endif
+
+#ifdef HAVE_TARGET_64_LITTLE
+template
+class Output_reloc<elfcpp::SHT_RELA, false, 64, false>;
+#endif
+
+#ifdef HAVE_TARGET_64_BIG
+template
+class Output_reloc<elfcpp::SHT_RELA, false, 64, true>;
+#endif
+
+#ifdef HAVE_TARGET_32_LITTLE
+template
+class Output_reloc<elfcpp::SHT_RELA, true, 32, false>;
+#endif
+
+#ifdef HAVE_TARGET_32_BIG
+template
+class Output_reloc<elfcpp::SHT_RELA, true, 32, true>;
+#endif
+
+#ifdef HAVE_TARGET_64_LITTLE
+template
+class Output_reloc<elfcpp::SHT_RELA, true, 64, false>;
+#endif
+
+#ifdef HAVE_TARGET_64_BIG
+template
+class Output_reloc<elfcpp::SHT_RELA, true, 64, true>;
+#endif
+
#ifdef HAVE_TARGET_32_LITTLE
template
class Output_data_reloc<elfcpp::SHT_REL, false, 32, false>;