// x86_64.cc -- x86_64 target support for gold.
-// Copyright (C) 2006-2020 Free Software Foundation, Inc.
+// Copyright (C) 2006-2021 Free Software Foundation, Inc.
// Written by Ian Lance Taylor <iant@google.com>.
// This file is part of gold.
rela_irelative_(NULL), copy_relocs_(elfcpp::R_X86_64_COPY),
got_mod_index_offset_(-1U), tlsdesc_reloc_info_(),
tls_base_symbol_defined_(false), isa_1_used_(0), isa_1_needed_(0),
- feature_1_(0), object_isa_1_used_(0), object_feature_1_(0),
- seen_first_object_(false)
+ feature_1_(0), feature_2_used_(0), feature_2_needed_(0),
+ object_isa_1_used_(0), object_feature_1_(0),
+ object_feature_2_used_(0), seen_first_object_(false)
{ }
// Hook for a new output section.
uint32_t isa_1_used_;
uint32_t isa_1_needed_;
uint32_t feature_1_;
+ uint32_t feature_2_used_;
+ uint32_t feature_2_needed_;
// Target-specific properties from the current object.
// These bits get ORed into ISA_1_USED_ after all properties for the object
// have been processed. But if either is all zeroes (as when the property
// These bits get ANDed into FEATURE_1_ after all properties for the object
// have been processed.
uint32_t object_feature_1_;
+ uint32_t object_feature_2_used_;
// Whether we have seen our first object, for use in initializing FEATURE_1_.
bool seen_first_object_;
};
switch (pr_type)
{
+ case elfcpp::GNU_PROPERTY_X86_COMPAT_ISA_1_USED:
+ case elfcpp::GNU_PROPERTY_X86_COMPAT_ISA_1_NEEDED:
+ case elfcpp::GNU_PROPERTY_X86_COMPAT_2_ISA_1_USED:
+ case elfcpp::GNU_PROPERTY_X86_COMPAT_2_ISA_1_NEEDED:
case elfcpp::GNU_PROPERTY_X86_ISA_1_USED:
case elfcpp::GNU_PROPERTY_X86_ISA_1_NEEDED:
case elfcpp::GNU_PROPERTY_X86_FEATURE_1_AND:
+ case elfcpp::GNU_PROPERTY_X86_FEATURE_2_USED:
+ case elfcpp::GNU_PROPERTY_X86_FEATURE_2_NEEDED:
if (pr_datasz != 4)
{
gold_warning(_("%s: corrupt .note.gnu.property section "
// If we see multiple feature props in one object, OR them together.
this->object_feature_1_ |= val;
break;
+ case elfcpp::GNU_PROPERTY_X86_FEATURE_2_USED:
+ this->object_feature_2_used_ |= val;
+ break;
+ case elfcpp::GNU_PROPERTY_X86_FEATURE_2_NEEDED:
+ this->feature_2_needed_ |= val;
+ break;
}
}
else if (this->isa_1_used_ != 0)
this->isa_1_used_ |= this->object_isa_1_used_;
this->feature_1_ &= this->object_feature_1_;
+ // If any object is missing the FEATURE_2_USED property, we must
+ // omit it from the output file.
+ if (this->object_feature_2_used_ == 0)
+ this->feature_2_used_ = 0;
+ else if (this->feature_2_used_ != 0)
+ this->feature_2_used_ |= this->object_feature_2_used_;
}
else
{
this->isa_1_used_ = this->object_isa_1_used_;
this->feature_1_ = this->object_feature_1_;
+ this->feature_2_used_ = this->object_feature_2_used_;
this->seen_first_object_ = true;
}
this->object_isa_1_used_ = 0;
this->object_feature_1_ = 0;
+ this->object_feature_2_used_ = 0;
}
static inline void
if (this->feature_1_ != 0)
add_property(layout, elfcpp::GNU_PROPERTY_X86_FEATURE_1_AND,
this->feature_1_);
+ if (this->feature_2_used_ != 0)
+ add_property(layout, elfcpp::GNU_PROPERTY_X86_FEATURE_2_USED,
+ this->feature_2_used_);
+ if (this->feature_2_needed_ != 0)
+ add_property(layout, elfcpp::GNU_PROPERTY_X86_FEATURE_2_NEEDED,
+ this->feature_2_needed_);
}
// Write the first three reserved words of the .got.plt section.
{
// From Alexandre Oliva, "Thread-Local Storage Descriptors for IA32
// and AMD64/EM64T", Version 0.9.4 (2005-10-10).
+ 0xf3, 0x0f, 0x1e, 0xfa, // endbr64
0xff, 0x35, // pushq x(%rip)
0, 0, 0, 0, // replaced with address of linkmap GOT entry (at PLTGOT + 8)
- 0xf2, 0xff, 0x25, // jmpq *y(%rip)
+ 0xff, 0x25, // jmpq *y(%rip)
0, 0, 0, 0, // replaced with offset of reserved TLSDESC_GOT entry
- 0x0f, 0x1f, 0 // nop
};
template<int size>
unsigned int plt_offset)
{
memcpy(pov, tlsdesc_plt_entry, plt_entry_size);
- elfcpp::Swap_unaligned<32, false>::writeval(pov + 2,
+ elfcpp::Swap_unaligned<32, false>::writeval(pov + 6,
(got_address + 8
- (plt_address + plt_offset
- + 6)));
- elfcpp::Swap_unaligned<32, false>::writeval(pov + 9,
+ + 10)));
+ elfcpp::Swap_unaligned<32, false>::writeval(pov + 12,
(got_base
+ tlsdesc_got_offset
- (plt_address + plt_offset
- + 13)));
+ + 16)));
}
// The .eh_frame unwind information for the PLT.
this->fill_tlsdesc_entry(pov, got_address, plt_address, got_base,
tlsdesc_got_offset, plt_offset);
pov += this->get_plt_entry_size();
+ plt_offset += plt_entry_size;
}
// Write the additional PLT.
this->fill_tlsdesc_entry(pov, got_address, plt_address, got_base,
tlsdesc_got_offset, plt_offset);
pov += this->get_plt_entry_size();
+ plt_offset += plt_entry_size;
}
// Write the additional PLT.
&& (r_type == elfcpp::R_X86_64_GOTPCREL
|| r_type == elfcpp::R_X86_64_GOTPCRELX
|| r_type == elfcpp::R_X86_64_REX_GOTPCRELX)
+ && reloc.get_r_addend() == -4
&& reloc.get_r_offset() >= 2
&& !is_ifunc)
{
unsigned int r_type,
const elfcpp::Sym<size, false>&)
{
- // When building a shared library, do not fold any local symbols as it is
- // not possible to distinguish pointer taken versus a call by looking at
- // the relocation types.
- if (parameters->options().shared())
- return true;
-
return possible_function_pointer_reloc(src_obj, src_indx,
reloc.get_r_offset(), r_type);
}
Output_section* ,
const elfcpp::Rela<size, false>& reloc,
unsigned int r_type,
- Symbol* gsym)
-{
- // When building a shared library, do not fold symbols whose visibility
- // is hidden, internal or protected.
- if (parameters->options().shared()
- && (gsym->visibility() == elfcpp::STV_INTERNAL
- || gsym->visibility() == elfcpp::STV_PROTECTED
- || gsym->visibility() == elfcpp::STV_HIDDEN))
- return true;
-
+ Symbol*)
+{
return possible_function_pointer_reloc(src_obj, src_indx,
reloc.get_r_offset(), r_type);
}
Lazy_view<size> view(object, data_shndx);
size_t r_offset = reloc.get_r_offset();
if (!parameters->incremental()
+ && reloc.get_r_addend() == -4
&& r_offset >= 2
&& Target_x86_64<size>::can_convert_mov_to_lea(gsym, r_type,
r_offset, &view))
case elfcpp::R_X86_64_GOTPCRELX:
case elfcpp::R_X86_64_REX_GOTPCRELX:
{
- // Convert
- // mov foo@GOTPCREL(%rip), %reg
- // to lea foo(%rip), %reg.
- // if possible.
- if (!parameters->incremental()
- && ((gsym == NULL
- && rela.get_r_offset() >= 2
- && view[-2] == 0x8b
- && !psymval->is_ifunc_symbol())
- || (gsym != NULL
- && rela.get_r_offset() >= 2
- && Target_x86_64<size>::can_convert_mov_to_lea(gsym, r_type,
- 0, &view))))
- {
- view[-2] = 0x8d;
- Reloc_funcs::pcrela32(view, object, psymval, addend, address);
- }
- // Convert
- // callq *foo@GOTPCRELX(%rip) to
- // addr32 callq foo
- // and jmpq *foo@GOTPCRELX(%rip) to
- // jmpq foo
- // nop
- else if (!parameters->incremental()
- && gsym != NULL
- && rela.get_r_offset() >= 2
- && Target_x86_64<size>::can_convert_callq_to_direct(gsym,
- r_type,
- 0, &view))
+ bool converted_p = false;
+
+ if (rela.get_r_addend() == -4)
{
- if (view[-1] == 0x15)
+ // Convert
+ // mov foo@GOTPCREL(%rip), %reg
+ // to lea foo(%rip), %reg.
+ // if possible.
+ if (!parameters->incremental()
+ && ((gsym == NULL
+ && rela.get_r_offset() >= 2
+ && view[-2] == 0x8b
+ && !psymval->is_ifunc_symbol())
+ || (gsym != NULL
+ && rela.get_r_offset() >= 2
+ && Target_x86_64<size>::can_convert_mov_to_lea(gsym,
+ r_type,
+ 0,
+ &view))))
{
- // Convert callq *foo@GOTPCRELX(%rip) to addr32 callq.
- // Opcode of addr32 is 0x67 and opcode of direct callq is 0xe8.
- view[-2] = 0x67;
- view[-1] = 0xe8;
- // Convert GOTPCRELX to 32-bit pc relative reloc.
+ view[-2] = 0x8d;
Reloc_funcs::pcrela32(view, object, psymval, addend, address);
+ converted_p = true;
}
- else
+ // Convert
+ // callq *foo@GOTPCRELX(%rip) to
+ // addr32 callq foo
+ // and jmpq *foo@GOTPCRELX(%rip) to
+ // jmpq foo
+ // nop
+ else if (!parameters->incremental()
+ && gsym != NULL
+ && rela.get_r_offset() >= 2
+ && Target_x86_64<size>::can_convert_callq_to_direct(gsym,
+ r_type,
+ 0,
+ &view))
{
- // Convert jmpq *foo@GOTPCRELX(%rip) to
- // jmpq foo
- // nop
- // The opcode of direct jmpq is 0xe9.
- view[-2] = 0xe9;
- // The opcode of nop is 0x90.
- view[3] = 0x90;
- // Convert GOTPCRELX to 32-bit pc relative reloc. jmpq is rip
- // relative and since the instruction following the jmpq is now
- // the nop, offset the address by 1 byte. The start of the
- // relocation also moves ahead by 1 byte.
- Reloc_funcs::pcrela32(&view[-1], object, psymval, addend,
- address - 1);
+ if (view[-1] == 0x15)
+ {
+ // Convert callq *foo@GOTPCRELX(%rip) to addr32 callq.
+ // Opcode of addr32 is 0x67 and opcode of direct callq
+ // is 0xe8.
+ view[-2] = 0x67;
+ view[-1] = 0xe8;
+ // Convert GOTPCRELX to 32-bit pc relative reloc.
+ Reloc_funcs::pcrela32(view, object, psymval, addend,
+ address);
+ converted_p = true;
+ }
+ else
+ {
+ // Convert jmpq *foo@GOTPCRELX(%rip) to
+ // jmpq foo
+ // nop
+ // The opcode of direct jmpq is 0xe9.
+ view[-2] = 0xe9;
+ // The opcode of nop is 0x90.
+ view[3] = 0x90;
+ // Convert GOTPCRELX to 32-bit pc relative reloc. jmpq
+ // is rip relative and since the instruction following
+ // the jmpq is now the nop, offset the address by 1
+ // byte. The start of the relocation also moves ahead
+ // by 1 byte.
+ Reloc_funcs::pcrela32(&view[-1], object, psymval, addend,
+ address - 1);
+ converted_p = true;
+ }
}
}
- else
+
+ if (!converted_p)
{
if (gsym != NULL)
{
{
if (r_type == elfcpp::R_X86_64_GOTPC32_TLSDESC)
{
- // leaq foo@tlsdesc(%rip), %rax
- // ==> movq foo@gottpoff(%rip), %rax
+ // LP64: leaq foo@tlsdesc(%rip), %rax
+ // ==> movq foo@gottpoff(%rip), %rax
+ // X32: rex leal foo@tlsdesc(%rip), %eax
+ // ==> rex movl foo@gottpoff(%rip), %eax
tls::check_range(relinfo, relnum, rela.get_r_offset(), view_size, -3);
tls::check_range(relinfo, relnum, rela.get_r_offset(), view_size, 4);
tls::check_tls(relinfo, relnum, rela.get_r_offset(),
- view[-3] == 0x48 && view[-2] == 0x8d && view[-1] == 0x05);
+ (((view[-3] & 0xfb) == 0x48
+ || (size == 32 && (view[-3] & 0xfb) == 0x40))
+ && view[-2] == 0x8d
+ && (view[-1] & 0xc7) == 0x05));
view[-2] = 0x8b;
const elfcpp::Elf_Xword addend = rela.get_r_addend();
Relocate_functions<size, false>::pcrela32(view, value, addend, address);
}
else
{
- // call *foo@tlscall(%rax)
- // ==> nop; nop
+ // LP64: call *foo@tlscall(%rax)
+ // ==> xchg %ax, %ax
+ // X32: call *foo@tlscall(%eax)
+ // ==> nopl (%rax)
gold_assert(r_type == elfcpp::R_X86_64_TLSDESC_CALL);
tls::check_range(relinfo, relnum, rela.get_r_offset(), view_size, 2);
+ int prefix = 0;
+ if (size == 32 && view[0] == 0x67)
+ {
+ tls::check_range(relinfo, relnum, rela.get_r_offset(),
+ view_size, 3);
+ prefix = 1;
+ }
tls::check_tls(relinfo, relnum, rela.get_r_offset(),
- view[0] == 0xff && view[1] == 0x10);
- view[0] = 0x66;
- view[1] = 0x90;
+ view[prefix] == 0xff && view[prefix + 1] == 0x10);
+ if (prefix)
+ {
+ view[0] = 0x0f;
+ view[1] = 0x1f;
+ view[2] = 0x00;
+ }
+ else
+ {
+ view[0] = 0x66;
+ view[1] = 0x90;
+ }
}
}
{
if (r_type == elfcpp::R_X86_64_GOTPC32_TLSDESC)
{
- // leaq foo@tlsdesc(%rip), %rax
- // ==> movq foo@tpoff, %rax
+ // LP64: leaq foo@tlsdesc(%rip), %rax
+ // ==> movq foo@tpoff, %rax
+ // X32: rex leal foo@tlsdesc(%rip), %eax
+ // ==> rex movl foo@tpoff, %eax
tls::check_range(relinfo, relnum, rela.get_r_offset(), view_size, -3);
tls::check_range(relinfo, relnum, rela.get_r_offset(), view_size, 4);
tls::check_tls(relinfo, relnum, rela.get_r_offset(),
- view[-3] == 0x48 && view[-2] == 0x8d && view[-1] == 0x05);
+ (((view[-3] & 0xfb) == 0x48
+ || (size == 32 && (view[-3] & 0xfb) == 0x40))
+ && view[-2] == 0x8d
+ && (view[-1] & 0xc7) == 0x05));
+ view[-3] = (view[-3] & 0x48) | ((view[-3] >> 2) & 1);
view[-2] = 0xc7;
- view[-1] = 0xc0;
+ view[-1] = 0xc0 | ((view[-1] >> 3) & 7);
value -= tls_segment->memsz();
Relocate_functions<size, false>::rela32(view, value, 0);
}
else
{
- // call *foo@tlscall(%rax)
- // ==> nop; nop
+ // LP64: call *foo@tlscall(%rax)
+ // ==> xchg %ax, %ax
+ // X32: call *foo@tlscall(%eax)
+ // ==> nopl (%rax)
gold_assert(r_type == elfcpp::R_X86_64_TLSDESC_CALL);
tls::check_range(relinfo, relnum, rela.get_r_offset(), view_size, 2);
+ int prefix = 0;
+ if (size == 32 && view[0] == 0x67)
+ {
+ tls::check_range(relinfo, relnum, rela.get_r_offset(),
+ view_size, 3);
+ prefix = 1;
+ }
tls::check_tls(relinfo, relnum, rela.get_r_offset(),
- view[0] == 0xff && view[1] == 0x10);
- view[0] = 0x66;
- view[1] = 0x90;
+ view[prefix] == 0xff && view[prefix + 1] == 0x10);
+ if (prefix)
+ {
+ view[0] = 0x0f;
+ view[1] = 0x1f;
+ view[2] = 0x00;
+ }
+ else
+ {
+ view[0] = 0x66;
+ view[1] = 0x90;
+ }
}
}
projects / deliverable / binutils-gdb.git / blobdiff