https://sourceware.org/bugzilla/show_bug.cgi?id=27280
--- Comment #1 from Fangrui Song ---
The idea of .file 0 auto-upgrade is that the compiler should emit .file 0
before .file 1, so the assembler does not need to deal with DWARF v5 md5
syntax.
I compiled a top-of-tree GCC. It emits `.file 1` i
https://sourceware.org/bugzilla/show_bug.cgi?id=26626
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=26065
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=24942
--- Comment #7 from Fangrui Song ---
In addition, have you thought about its interaction with --rename-section?
Which command should I use:
objcopy --rename-section=.foo=.foo2 --set-section-alignment .foo=8 a.o b.o
or
objcopy --rename-secti
https://sourceware.org/bugzilla/show_bug.cgi?id=24942
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
objcopy -I binary -O elf64-x86-64 a.txt b # e_machine is
EM_NONE
objcopy -I binary -B i386:x86-64 -O elf64-x86-64 a.txt c # e_machine is
EM_X86_64
cmp -l b c
https://sourceware.org/bugzilla/show_bug.cgi?id=19823
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=22843
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=24505
--- Comment #2 from Fangrui Song ---
Thanks for the reply.
I've tried hard to find who are using both CONSTANT(COMMONPAGESIZE) and -n/-N
in the wild.
It looks only Tianocore edk2 (ab)uses -n -z common-page-size=0x20 to use
CONSTANT(COMMONPAG
https://sourceware.org/bugzilla/show_bug.cgi?id=24896
--- Comment #3 from Fangrui Song ---
Thanks for the information.
* Aligned R_PPC64_UADDR64 -> R_PPC64_ADDR64 => the code is there and it
doesn't hurt anyone, so let's keep it
* Misaligned R_PPC64_ADDR64 -> R_PPC64_UADDR64 => this matters on
https://sourceware.org/bugzilla/show_bug.cgi?id=24896
--- Comment #1 from Fangrui Song ---
Thinking more. ADDR/UADDR distinction probably makes little sense.
I withdraw my proposal that teachs gas to emit UADDR.
Some ld conversion rules can probably be deleted:
Aligned R_PPC64_UADDR64 -> R_PPC6
https://sourceware.org/bugzilla/show_bug.cgi?id=24896
Fangrui Song changed:
What|Removed |Added
Component|gas |ld
Summary|[powerpc] gas sh
Version: 2.33 (HEAD)
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: gas
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
Related to both gas and ld.
gas emits
https://sourceware.org/bugzilla/show_bug.cgi?id=24896
Fangrui Song changed:
What|Removed |Added
Target||powerpc*
--
You are receiving this ma
https://sourceware.org/bugzilla/show_bug.cgi?id=18815
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
iority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
% readelf -n core
...
Owner Data size Description
CORE 0x0150 N
https://sourceware.org/bugzilla/show_bug.cgi?id=24815
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=24784
--- Comment #7 from Fangrui Song ---
> For the occurrence of name@GOTPCREL in the following assembler instructions:
> call *name@GOTPCREL(%rip)
I agree, yet it also says:
> (PIC) the base of GOT is within 2GB an indirect call to the GOT ent
https://sourceware.org/bugzilla/show_bug.cgi?id=24784
--- Comment #5 from Fangrui Song ---
I don't see how this bug is "invalid". I've already worked around the bug in
clang, but I hope this hack can be deleted in a few years.
> g++ -fno-plt -fpic a.cc -Wa,-mrelax-relocations=no -fuse-ld=bfd
I
https://sourceware.org/bugzilla/show_bug.cgi?id=24784
--- Comment #3 from Fangrui Song ---
This was discovered by a Rust user. They use clang integrated assembler (not
GNU as) with GNU ld. The integrated assembler isn't configured with
-Wa,-mrelax-relocations=yes on by default. I worked around th
https://sourceware.org/bugzilla/show_bug.cgi?id=24784
--- Comment #2 from Fangrui Song ---
(In reply to H.J. Lu from comment #1)
> Since the new linker is required to support no-PLT TLS sequence,
> R_X86_64_GOTPCRELX should be used here. R_X86_64_GOTPCREL won't
> work with the old linker anyway.
https://sourceware.org/bugzilla/show_bug.cgi?id=16794
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=20216
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
thread_local int a;
int main() { return a; }
% g++ -fno-plt -fpic a.cc -Wa,-mrelax-relocations=no -fuse-ld=bfd
https://sourceware.org/bugzilla/show_bug.cgi?id=24769
Fangrui Song changed:
What|Removed |Added
Target||riscv*-*-*
--
You are receiving this
erity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
% riscv64-linux-gnu-gcc -fuse-ld=bfd a.c -Wl,-z,relro -Wl,-Map,a.map -o a
% readelf -l a
07 .preinit_
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
STT_SECTION symbols are not used to symbolize addresses in objdump -d output,
but that will be useful when the relevant symbols are missing.
e.g. In a
https://sourceware.org/bugzilla/show_bug.cgi?id=17369
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=18289
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=24490
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=24683
Fangrui Song changed:
What|Removed |Added
CC||maskray at google dot com
--- Comment
https://sourceware.org/bugzilla/show_bug.cgi?id=24685
Fangrui Song changed:
What|Removed |Added
Target||riscv*-*-*
--
You are receiving this
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
A canonical PLT (the term is not widely used. I just saw it used by a few
people) in an executable has a section
https://sourceware.org/bugzilla/show_bug.cgi?id=24676
Fangrui Song changed:
What|Removed |Added
Summary|[RISCV] Redundant |[RISCV] Redundant
|R
https://sourceware.org/bugzilla/show_bug.cgi?id=24676
Fangrui Song changed:
What|Removed |Added
Target||riscv*-*-*
--
You are receiving this
EAD)
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
RISC-V psABI doesn't specify TLS relaxation. Though the code sequences
https://sourceware.org/bugzilla/show_bug.cgi?id=24673
Fangrui Song changed:
What|Removed |Added
Target||riscv64-*-*
--
You are receiving this
: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
static __thread int a;
int foo() {return a;}
int main() {}
riscv-gcc -fuse-ld=bfd -fPIC
https://sourceware.org/bugzilla/show_bug.cgi?id=24627
--- Comment #5 from Fangrui Song ---
One remaining issue (after ae9ac79e5e291a2bc5c6620a10c37a3558d55d36). Does
`00: Rev: 1` below need a similar %# fix?
Version definition section '.gnu.version_d' contains 26 entries:
Addr: 0x00
https://sourceware.org/bugzilla/show_bug.cgi?id=24627
Fangrui Song changed:
What|Removed |Added
Summary|readelf -V output is ugly |readelf -V output is
https://sourceware.org/bugzilla/show_bug.cgi?id=24627
--- Comment #1 from Fangrui Song ---
Sorry, in .gnu.version_d, `Addr:` is indented by 2.
--
You are receiving this mail because:
You are on the CC list for the bug.
___
bug-binutils mailing list
bu
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
% readelf -V /lib/x86_64-linux-gnu/libc.so.6
Version symbols section '.gnu.version' contains 2247 entries:
Addr: 00016a3a Offset: 0x016a3a Link: 4 (.dyn
MED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
% cat a.c
__attribute__((weak,tls_model("local-exec")))
extern __thread char tls;
char *g
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
--start-lib gives following object files (before --end-lib) archive semantics.
They are handy if you don't want to create archives for the semantics.
Some build system make
Component: gold
Assignee: ccoutant at gmail dot com
Reporter: maskray at google dot com
CC: ian at airs dot com
Target Milestone: ---
There was a discussion
https://groups.google.com/d/msg/generic-abi/_CbBM6T6WeM/9nNnwRNHAQAJ to extend
the semantics of
iority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
// bfd/syms.c
static const struct section_to_type stt[] =
{
{".bss", 'b'},
{"code", 't
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
-n/--nmagic disables page alignment of sections. Should it affect the target
specific constants MAXPAGESIZE and COMMONPAGESIZE by changing them to 1?
% cat
https://sourceware.org/bugzilla/show_bug.cgi?id=23850
Fangrui Song changed:
What|Removed |Added
Status|ASSIGNED|RESOLVED
Resolution|---
https://sourceware.org/bugzilla/show_bug.cgi?id=23850
--- Comment #3 from Fangrui Song ---
(In reply to Alan Modra from comment #2)
> .rela.plt in a static executable can only contain IRELATIVE relocations that
> must have a zero symbol index. Thus it is not necessary for that particular
> SHT_R
https://sourceware.org/bugzilla/show_bug.cgi?id=23850
--- Comment #1 from Fangrui Song ---
Is this because .rela.plt (SHF_ALLOC) should not have sh_link pointing to
.symtab (non-SHF_ALLOC)? If it is the case, perhaps `readelf -S a` should not
warn on such executable.
--
You are receiving this m
Severity: normal
Priority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: maskray at google dot com
Target Milestone: ---
This is similar to Bug 10337 but is a different incarnation.
When linking glibc statically, .rel
51 matches
Mail list logo