> It is not just static glibc. PR 19161 has another usage.
David Li is a main reviewer of LLVM PGO. He should be happy if GNU ld and gold
can now garbage collect PGO sections
https://reviews.llvm.org/D97649#inline-915909 (he asked me to add some tests
for GNU ld and gold).
PR19161 was f
https://sourceware.org/bugzilla/show_bug.cgi?id=27441
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #16
https://sourceware.org/bugzilla/show_bug.cgi?id=26105
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #1
https://sourceware.org/bugzilla/show_bug.cgi?id=26404
--- Comment #8 from Fangrui Song ---
(In reply to Nick Clifton from comment #7)
> (In reply to Fangrui Song from comment #6)
>
> > scriptfile defines multiple SECTIONS commands. Each SECTIONS defines exactly
> > one output section (for the fi
: gas
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
R_*_NONE relocation types do not modify the location but can be used to
indicate a dependency between two sections. For example, ARM EHABI uses
R_ARM_NONE from an exception
https://sourceware.org/bugzilla/show_bug.cgi?id=27530
--- Comment #1 from Fangrui Song ---
OK, I see that
https://sourceware.org/git/?p=binutils-gdb.git;a=blobdiff;f=gas/read.c;h=2224c0ed241845eb2e73f59b807dedd617f2a0e0;hp=b2d50272d7fde8ba2af87a29c905c8dd245e08ae;hb=740bdc67c057ee8012327420848eb1
https://sourceware.org/bugzilla/show_bug.cgi?id=26706
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #1
https://sourceware.org/bugzilla/show_bug.cgi?id=27441
--- Comment #21 from Fangrui Song ---
I made a mistake in #comment 16.
In GNU ld, the rule is probably:
* it is linked at least once in `--no-as-needed` mode (i.e. `--as-needed a.so
--no-as-needed a.so` => needed)
* or it has a definition re
https://sourceware.org/bugzilla/show_bug.cgi?id=27296
Fangrui Song changed:
What|Removed |Added
Resolution|--- |NOTABUG
Status|UNCONFIRMED
https://sourceware.org/bugzilla/show_bug.cgi?id=27282
--- Comment #1 from Fangrui Song ---
You can close this one.
For SHF_GNU_RETAIN, I've changed clang 13 to set ELFOSABI_GNU.
I am happy with the current GNU ld behavior.
For STT_GNU_IFUNC, clang has always been using ELFOSABI_NONE.
So please
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
REVERSE complements SORT, e.g.
SECTIONS { .aaa : { *(REVERSE(SORT_BY_NAME(.aaa.*))) } }
If the user wants to detect some static initialization order
https://sourceware.org/bugzilla/show_bug.cgi?id=27487
--- Comment #3 from Fangrui Song ---
Hi Nick, thanks for the patch! --format=just-symbols and -j look good to me. I
wonder whether it is necessary to reserve -J. -J can probably be used for other
purposes.
--
You are receiving this mail
https://sourceware.org/bugzilla/show_bug.cgi?id=27497
--- Comment #1 from Fangrui Song ---
https://reviews.llvm.org/D96883 it may soon be accepted :)
--
You are receiving this mail because:
You are on the CC list for the bug.
: normal
Priority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
On ARM, mapping symbols are suppressed by default.
% arm-linux-gnueabi-gcc-nm test.o
https://sourceware.org/bugzilla/show_bug.cgi?id=27584
Fangrui Song changed:
What|Removed |Added
Target||riscv*-*-*
CC|
https://sourceware.org/bugzilla/show_bug.cgi?id=27584
--- Comment #2 from Fangrui Song ---
You can try any C file. Due to label differences, there are always lots of
STB_LOCAL STT_NOTYPE symbols. It seems that GCC uses .L0 while clang uses an
empty name.
The question is whether such symbols shou
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
Extracted from https://reviews.llvm.org/D95916#2627640 This may be related to
PR27584.
echo 'void _start() {}' > a.c
# Make sure riscv64-linux-gnu-b
https://sourceware.org/bugzilla/show_bug.cgi?id=27585
Fangrui Song changed:
What|Removed |Added
Target||riscv*-*-*
CC|
Priority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: mtvec at pm dot me
Target Milestone: ---
Target: RISC-V
Created attachment 13418
--> https://sourceware.org/bugzilla/attachment.cgi?id=13418&action=e
https://sourceware.org/bugzilla/show_bug.cgi?id=27814
--- Comment #2 from mtvec at pm dot me ---
I don't have a copyright assignment yet but I would be interested in getting
one. What is the procedure for this?
This patch is quite trivial by the way so we might be able to skip it in this
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
[Extracted from my reply to [PATCH v4 00/10] Function Granular KASLR in
January]
mkdir -p ./a
echo 'a: a.1: a.2: nop' > ./a/a.s
echo 'a: nop'
: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
Depending on the workload, -Bsymbolic-functions can be a very effective
optimization. -Bno-symbolic makes it easy to exclude
Priority: P2
Component: gold
Assignee: ccoutant at gmail dot com
Reporter: chen.yunxing at me dot com
CC: ian at airs dot com
Target Milestone: ---
I use gold with gcc 9.3 do lto link;
and int lto-trans phase; failed with message:
collect2
: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
As a poor man's direct binding feature, -Bsymbolic-functions is incompatible
with two things
https://sourceware.org/bugzilla/show_bug.cgi?id=27871
Fangrui Song changed:
What|Removed |Added
Summary|ld: Add |ld: Add
|-Bsymbolic-
t
> happens, but what on earth are you linking that has so many output sections?
> That just shouldn't happen in any practical application. Solving that may work
> around this bug.
>
> Please add "-v -Wl,--debug=plugin" options, and send me the verbose gcc ou
ound this bug.
>
> Please add "-v -Wl,--debug=plugin" options, and send me the verbose gcc output
> and a tar file of the resulting debug directory.
>
> --
> You are receiving this mail because:
> You reported the bug.
--
You are receiving this mail because:
You are on the CC list for the bug.
> that happens, but what on earth are you linking that has so many output
> sections? That just shouldn't happen in any practical application. Solving
> that may work around this bug.
>
> Please add "-v -Wl,--debug=plugin" options, and send me the verbose gcc
> outpu
https://sourceware.org/bugzilla/show_bug.cgi?id=27871
Fangrui Song changed:
What|Removed |Added
Summary|ld: Add |ld: Add
|-Bsymbolic-
https://sourceware.org/bugzilla/show_bug.cgi?id=27871
Fangrui Song changed:
What|Removed |Added
CC||hjl.tools at gmail dot com
--
You are
https://sourceware.org/bugzilla/show_bug.cgi?id=27871
--- Comment #1 from Fangrui Song ---
ld patch: https://sourceware.org/pipermail/binutils/2021-May/116703.html
gold patch: https://sourceware.org/pipermail/binutils/2021-May/116683.html
(I was considering -Bsymbolic-global-functions but Peter
: unspecified
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: gas
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
% cat a.s
ldr q24, [x23, :dtprel_lo12_nc:sym]
% ~/Dev/binutils-gdb
https://sourceware.org/bugzilla/show_bug.cgi?id=27904
Fangrui Song changed:
What|Removed |Added
Target||aarch64-*
--
You are receiving this m
https://sourceware.org/bugzilla/show_bug.cgi?id=27904
Fangrui Song changed:
What|Removed |Added
CC||nsz at port70 dot net
--
You are rece
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
The latter part of
https://sourceware.org/pipermail/binutils/2020-May/111086.html
Position independent executables and shared objects with
-Bsymbolic/-Bsymbolic
https://sourceware.org/bugzilla/show_bug.cgi?id=27923
Fangrui Song changed:
What|Removed |Added
See Also||https://sourceware.org/bugz
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
cat > ./a.s <https://groups.google.com/g/generic-abi/c/2X6mR-s2zoc "GRP_C
https://sourceware.org/bugzilla/show_bug.cgi?id=27931
Fangrui Song changed:
What|Removed |Added
CC||jh7370 at my dot bristol.ac.uk
--
You
https://sourceware.org/bugzilla/show_bug.cgi?id=26404
--- Comment #9 from Fangrui Song ---
I created a ld.lld patch for the --overwrite-script idea
https://reviews.llvm.org/D103303
--
You are receiving this mail because:
You are on the CC list for the bug.
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
An input ET_EXEC will cause
% as /dev/null -o a.o
% ld.bfd a.o -o a; ld.bfd a
...
ld.bfd: Using an executable file (a) as input to a link is deprecated - support
is
https://sourceware.org/bugzilla/show_bug.cgi?id=26404
--- Comment #11 from Fangrui Song ---
My latest revision of https://reviews.llvm.org/D103303 switched to a new
keyword "OVERWRITE_SECTIONS". I dropped the command line option.One can place
more than descriptions, though there is no ordering re
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
// a.c
__attribute__((visibility("protected"))) void *foo () {
return (void *)foo;
}
// This roughly maps to:
.
https://sourceware.org/bugzilla/show_bug.cgi?id=27973
Fangrui Song changed:
What|Removed |Added
Target||x86_64-*
--
You are receiving this ma
https://sourceware.org/bugzilla/show_bug.cgi?id=27973
Fangrui Song changed:
What|Removed |Added
CC||fweimer at redhat dot com,
: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
% gcc a.c -static -fuse-ld=bfd -Wl,-t
/usr/lib/gcc/x86_64-linux-gnu/10/../../../x86_64-linux-gnu/crt1.o
/usr/lib/gcc/x86_64-linux
https://sourceware.org/bugzilla/show_bug.cgi?id=27990
Fangrui Song changed:
What|Removed |Added
CC||ccoutant at gmail dot com,
https://sourceware.org/bugzilla/show_bug.cgi?id=27990
--- Comment #1 from Fangrui Song ---
I'll add a comment that on Mach-O, ld64 -why_live gives you clue why a symbol
is defined (via a potentially long archive member extraction chain).
The --trace feature can make some other debugging tasks co
https://sourceware.org/bugzilla/show_bug.cgi?id=27753
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #4
https://sourceware.org/bugzilla/show_bug.cgi?id=27753
--- Comment #6 from Fangrui Song ---
A dynamic loader needs PT_GNU_PROPERTY to know the ISA usage of a component.
.note.gnu.property makes up PT_GNU_PROPERTY. Therefore, the section needs to
SHF_ALLOC to be part of the program image.
% cat a.
y: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
With 'clang' and 'llvm-profdata' built from the main branch of llvm-project
(the future clang 13.0.0 release):
echo -e '#in
https://sourceware.org/bugzilla/show_bug.cgi?id=26406
Niklas Hambüchen changed:
What|Removed |Added
CC||mail at nh2 dot me
--
You are
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: mail at nh2 dot me
Target Milestone: ---
Dear maintainer,
I found a performance regression between binutils 2.34 and 2.35.1 that makes
stripping a 50 MB `.a` take 300 seconds instead of 1 second.
Original
https://sourceware.org/bugzilla/show_bug.cgi?id=28058
Niklas Hambüchen changed:
What|Removed |Added
CC||mail at nh2 dot me
--
You are
https://sourceware.org/bugzilla/show_bug.cgi?id=28058
Niklas Hambüchen changed:
What|Removed |Added
CC||nickc at redhat dot com
--
You ar
https://sourceware.org/bugzilla/show_bug.cgi?id=28058
Niklas Hambüchen changed:
What|Removed |Added
CC||matz at suse dot de
--- Comment #1
https://sourceware.org/bugzilla/show_bug.cgi?id=28058
Niklas Hambüchen changed:
What|Removed |Added
Status|UNCONFIRMED |RESOLVED
Resolution|---
https://sourceware.org/bugzilla/show_bug.cgi?id=21782
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #6
https://sourceware.org/bugzilla/show_bug.cgi?id=28058
--- Comment #6 from Niklas Hambüchen ---
Thank you!
--
You are receiving this mail because:
You are on the CC list for the bug.
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
% cat a.s
.globl _start, bbb
_start:
.section .mytext.bbb,"axG",@progbits,bbb,comdat
bbb:
.byte 0
.section .init_array.bbb,"awG",@init_arra
https://sourceware.org/bugzilla/show_bug.cgi?id=28124
--- Comment #1 from Fangrui Song ---
For SHF_LINK_ORDER:
% cat a.s
.globl _start; _start: nop
.section .text.a,"ax",@progbits; ret
.section .init_array.a,"awo",@init_array,.text.a; ret
% gcc -fuse-ld=bfd a.s -nostdlib -Wl,--gc-sections
/usr/b
https://sourceware.org/bugzilla/show_bug.cgi?id=27931
--- Comment #1 from Fangrui Song ---
Ping:)
I am updating llvm-objcopy to drop GRP_COMDAT
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=1
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #5
Priority: P2
Component: gas
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
PR 25295 added the new syntax `, remove` to `.symver`.
However, the new syntax cannot be used in relocations. This makes it difficult
https://sourceware.org/bugzilla/show_bug.cgi?id=28157
Fangrui Song changed:
What|Removed |Added
CC||hjl.tools at gmail dot com
--
You are
https://sourceware.org/bugzilla/show_bug.cgi?id=25295
--- Comment #23 from Fangrui Song ---
.symver *, *, remove
The original symbol cannot be used in relocations. This limits `, remove`'s
usage. I filed PR gas/28157
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=28158
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--
You are
Severity: normal
Priority: P2
Component: gold
Assignee: ccoutant at gmail dot com
Reporter: i at maskray dot me
CC: ian at airs dot com
Target Milestone: ---
cat > ./a.s < ./a.ver <
: binutils
Version: unspecified
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: gold
Assignee: ccoutant at gmail dot com
Reporter: i at maskray dot me
CC: ian at airs dot com
Target Milestone: ---
cat
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
Looks like # is supported in version scripts and the VERSION command in a
linker script, but is unsupported outside of VERSION.
VERSION {
# meow meow
GLIBC_2.2.5
https://sourceware.org/bugzilla/show_bug.cgi?id=28157
--- Comment #2 from Fangrui Song ---
(In reply to H.J. Lu from comment #1)
> Created attachment 13571 [details]
> A patch
>
> Please try this.
I applied this patch.
% cat /tmp/c/y.s
call __free_hook
.symver __free_hook, __free_hook@GLIBC.2.
https://sourceware.org/bugzilla/show_bug.cgi?id=28157
--- Comment #3 from Fangrui Song ---
For defined non-`remove` single-@, you may want to change the relocations to
reference the versioned symbol as well.
Then the only difference between defined `remove` single-@ and defined
non-`remove` sing
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
If an ifunc resolver calls a PLT with lazy JUMP_SLOT, should it work?
My impression is that this does not necessarily need to work
https://sourceware.org/bugzilla/show_bug.cgi?id=27490
--- Comment #1 from Fangrui Song ---
LLVM integrated assembler now sets ELFOSABI_GNU.
That said, the gold behavior creating multiple output sections may continue
trapping future new section flags.
--
You are receiving this mail because:
You
https://sourceware.org/bugzilla/show_bug.cgi?id=27408
Fangrui Song changed:
What|Removed |Added
Resolution|--- |FIXED
Version|unspecified
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
cat > ./a.s < ./ver <
https://sourceware.org/bugzilla/show_bug.cgi?id=28149
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #5
Product: binutils
Version: unspecified
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
GCC -fpic may gener
https://sourceware.org/bugzilla/show_bug.cgi?id=28370
Fangrui Song changed:
What|Removed |Added
Target||aarch64-*
--
You are receiving this m
: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
Lazy R_*_TLSDESC resolving has several problems (e.g. data race) and glibc has
eagerly resolved R_*_TLSDESC for all(?) supported ports: arm (PR18572)/aarch64
(PR18034)/x86
https://sourceware.org/bugzilla/show_bug.cgi?id=28387
Fangrui Song changed:
What|Removed |Added
CC||nsz at gcc dot gnu.org
--
You are rec
erity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
PT_GNU_RELRO / 1685382482 does not work in the PHDRS command.
% cat a.lds
PHDRS {
exec PT_LOAD;
rw PT_LOAD;
dy
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
cat >a.s <>> defined in arm.o
>>> referenced by arm.o:(.text+0x4)
% ld.lld -shared -Bsymbolic a
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
j __sigsetjmp
.globl __sigsetjmp
__sigsetjmp:
ret
riscv64-linux-gnu-as a.s -o a.o
riscv64-linux-gnu-ld -shared a.o
https://sourceware.org/bugzilla/show_bug.cgi?id=28509
Fangrui Song changed:
What|Removed |Added
CC||nelson.chu at sifive dot com
--
You a
https://sourceware.org/bugzilla/show_bug.cgi?id=28509
--- Comment #1 from Fangrui Song ---
glibc commit
https://sourceware.org/git/?p=glibc.git;a=commit;h=68389203832ab39dd0dbaabbc4059e7fff51c29b
("riscv: Fix incorrect jal with HIDDEN_JUMPTARGET") could have been avoided if
such an error had been
https://sourceware.org/bugzilla/show_bug.cgi?id=18599
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #4
Component: gold
Assignee: ccoutant at gmail dot com
Reporter: i at maskray dot me
CC: ian at airs dot com
Target Milestone: ---
gold version of https://sourceware.org/bugzilla/show_bug.cgi?id=27923 (GNU ld
support).
I have a patch which will be posted soon
https://sourceware.org/bugzilla/show_bug.cgi?id=28601
--- Comment #1 from Fangrui Song ---
Patch: https://sourceware.org/pipermail/binutils/2021-November/118567.html
("[PATCH] gold: Add --pack-dyn-relocs=relr for arm/aarch64/x86-64 [PR 28601]")
--
You are receiving this mail because:
You are on
: UNCONFIRMED
Severity: normal
Priority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
binutils/testsuite/binutils-all/ gas/testsuite/gas ld/testsuite/ mainly use
dump tests
https://sourceware.org/bugzilla/show_bug.cgi?id=28602
--- Comment #1 from Fangrui Song ---
>From https://maskray.me/blog/2021-08-08-toolchain-testing#split-file
In Aug 2020, I added `split-file` to llvm-project which allows you to place
multiple extra files in one file. It'd be nice if binutils
https://sourceware.org/bugzilla/show_bug.cgi?id=27491
--- Comment #5 from Fangrui Song ---
(In reply to H.J. Lu from comment #4)
> (In reply to Fangrui Song from comment #3)
> >
> > I don't know what happened to the 2010-01 patch. It likely did not let
> > __start_ reference retain magic sections
https://sourceware.org/bugzilla/show_bug.cgi?id=27491
Fangrui Song changed:
What|Removed |Added
Status|RESOLVED|REOPENED
Resolution|NOTABUG
https://sourceware.org/bugzilla/show_bug.cgi?id=27491
--- Comment #8 from Fangrui Song ---
(In reply to H.J. Lu from comment #7)
> Created attachment 13811 [details]
> A patch
>
> Try this.
Thanks! I applied it locally and the original example works.
Adding `.section xx,"aR",unique,2; .byte 2`
https://sourceware.org/bugzilla/show_bug.cgi?id=28689
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--- Comment #4
https://sourceware.org/bugzilla/show_bug.cgi?id=28689
--- Comment #5 from Fangrui Song ---
(In reply to Fangrui Song from comment #4)
> (In reply to H.J. Lu from comment #2)
> > One problem:
> > [...]
> > File offsets are changed.
>
> For objcopy, it is important to not transform the file to dec
Severity: normal
Priority: P2
Component: binutils
Assignee: unassigned at sourceware dot org
Reporter: i at maskray dot me
Target Milestone: ---
https://pubs.opengroup.org/onlinepubs/9699919799/utilities/ar.html
-T
[XSI] [Option Start] Allow filename tru
https://sourceware.org/bugzilla/show_bug.cgi?id=28759
Fangrui Song changed:
What|Removed |Added
CC||nickc at redhat dot com
--
You are re
https://sourceware.org/bugzilla/show_bug.cgi?id=28759
--- Comment #1 from Fangrui Song ---
Patch: https://sourceware.org/pipermail/binutils/2022-January/119250.html
--
You are receiving this mail because:
You are on the CC list for the bug.
https://sourceware.org/bugzilla/show_bug.cgi?id=28759
Fangrui Song changed:
What|Removed |Added
Status|UNCONFIRMED |RESOLVED
Resolution|---
https://sourceware.org/bugzilla/show_bug.cgi?id=28772
Fangrui Song changed:
What|Removed |Added
CC||i at maskray dot me
--
You are
401 - 500 of 527 matches
Mail list logo
