[Bug ld/26869] New: ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu
https://sourceware.org/bugzilla/show_bug.cgi?id=26869
Bug ID: 26869
Summary: ld: assertion fail ../../bfd/elflink.c:14806 on
i686-linux-gnu
Product: binutils
Version: 2.35.1
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: mliska at suse dot cz
Target Milestone: ---
I hit the following assert:
void
elf_append_rel (bfd *abfd, asection *s, Elf_Internal_Rela *rel)
{
const struct elf_backend_data *bed = get_elf_backend_data (abfd);
bfd_byte *loc = s->contents + (s->reloc_count++ * bed->s->sizeof_rel);
BFD_ASSERT (loc + bed->s->sizeof_rel <= s->contents + s->size);
bed->s->swap_reloc_out (abfd, rel, loc);
}
I use tip of 2.35-branch and it happens when I link
../../lib/libQt5WebEngineCore.so.5.12.10 on i686 with LTO enabled.
It's related to .rel.got section and it contains ~2000 reloc_count and it
really writes one after s->reloc_count * bed->s->sizeof_rel.
Any hint on how to debug that?
--
You are receiving this mail because:
You are on the CC list for the bug.
[Bug ld/26869] ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu
https://sourceware.org/bugzilla/show_bug.cgi?id=26869 --- Comment #1 from Martin Liska --- ld.gold links that fine -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/26869] ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu
https://sourceware.org/bugzilla/show_bug.cgi?id=26869 Fabian Vogt changed: What|Removed |Added CC||fab...@ritter-vogt.de -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/26869] ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu
https://sourceware.org/bugzilla/show_bug.cgi?id=26869 --- Comment #2 from Martin Liska --- I'm reducing list of object files needed: $ g++ -flto=160 @files.txt -shared -o libQt5WebEngineCore.so.5.12.10 /usr/lib/gcc/i586-suse-linux/10/../../../../i586-suse-linux/bin/ld: /home/abuild/rpmbuild/BUILD/qtwebengine-everywhere-src-5.12.10/src/core/release/obj/ui/base/base/os_exchange_data.o: in function `ui::OSExchangeData::OSExchangeData()': /home/abuild/rpmbuild/BUILD/qtwebengine-everywhere-src-5.12.10/src/core/release/../../3rdparty/chromium/ui/base/dragdrop/os_exchange_data.cc:24: undefined reference to `ui::OSExchangeDataProviderFactory::CreateProvider()' /usr/lib/gcc/i586-suse-linux/10/../../../../i586-suse-linux/bin/ld: BFD (GNU Binutils; home:marxin:branches:home:Vogtinator:qt5.12 / openSUSE_Tumbleweed) 2.35.1.20201112-399 assertion fail ../../bfd/elflink.c:14806 collect2: error: ld returned 1 exit status -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/26869] ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu since 382aae06322799a25ea52fe61b243cbca4db8d66
https://sourceware.org/bugzilla/show_bug.cgi?id=26869
Martin Liska changed:
What|Removed |Added
Summary|ld: assertion fail |ld: assertion fail
|../../bfd/elflink.c:14806 |../../bfd/elflink.c:14806
|on i686-linux-gnu |on i686-linux-gnu since
||382aae06322799a25ea52fe61b2
||43cbca4db8d66
CC||hjl.tools at gmail dot com
--- Comment #3 from Martin Liska ---
I have a nice reproducer:
$ cat ~/Programming/testcases/metrics.ii
void ReadElfBuildId(void *);
extern char __ehdr_start;
void ParseSmapsHeader() { ReadElfBuildId(&__ehdr_start); }
$ g++ ~/Programming/testcases/metrics.ii -fPIC -shared -O -m32 -o
/tmp/metrics.o -c
$ ./ld/ld-new /tmp/metrics.o -shared -melf_i386
./ld/ld-new: BFD (GNU Binutils) 2.34.50.20200401 assertion fail
../../bfd/elflink.c:14687
Segmentation fault (core dumped)
For completeness there's .s file:
$ cat metrics.s
.file "metrics.ii"
.text
.globl _Z16ParseSmapsHeaderv
.type _Z16ParseSmapsHeaderv, @function
_Z16ParseSmapsHeaderv:
.LFB0:
.cfi_startproc
pushl %ebx
.cfi_def_cfa_offset 8
.cfi_offset 3, -8
subl$20, %esp
.cfi_def_cfa_offset 28
call__x86.get_pc_thunk.bx
addl$_GLOBAL_OFFSET_TABLE_, %ebx
pushl __ehdr_start@GOT(%ebx)
.cfi_def_cfa_offset 32
call_Z14ReadElfBuildIdPv@PLT
addl$24, %esp
.cfi_def_cfa_offset 8
popl%ebx
.cfi_restore 3
.cfi_def_cfa_offset 4
ret
.cfi_endproc
.LFE0:
.size _Z16ParseSmapsHeaderv, .-_Z16ParseSmapsHeaderv
.section
.text.__x86.get_pc_thunk.bx,"axG",@progbits,__x86.get_pc_thunk.bx,comdat
.globl __x86.get_pc_thunk.bx
.hidden __x86.get_pc_thunk.bx
.type __x86.get_pc_thunk.bx, @function
__x86.get_pc_thunk.bx:
.LFB1:
.cfi_startproc
movl(%esp), %ebx
ret
.cfi_endproc
.LFE1:
.ident "GCC: (SUSE Linux) 10.2.1 20201028 [revision
a78cd759754c92cecbf235ac9b447dcdff6c6e2f]"
.section.note.GNU-stack,"",@progbits
Started with H.J.'s revision 382aae06322799a25ea52fe61b243cbca4db8d66.
--
You are receiving this mail because:
You are on the CC list for the bug.
[Bug gas/26850] as -g: remove non-executable sections' contribution from DW_AT_ranges
https://sourceware.org/bugzilla/show_bug.cgi?id=26850 --- Comment #1 from cvs-commit at gcc dot gnu.org --- The master branch has been updated by Nick Clifton : https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=bcb78b47614ed252226da1a688558a17d8a99ce8 commit bcb78b47614ed252226da1a688558a17d8a99ce8 Author: Nick Clifton Date: Thu Nov 12 11:43:20 2020 + Stop Gas from generating line info or address ranges for sections that do not contain code or are not loaded. PR 26850 * dwarf2dbg.c (dwarf2_gen_line_info_1): Do not record lines in sections that are not executable or not loadable. (out_debug_line): Move warning message into dwarf2_gen_line_info_1. * testsuite/gas/elf/dwarf2-20.s: New test. * testsuite/gas/elf/dwarf2-20.d: New test driver. * testsuite/gas/elf/elf.exp: Run the new test. * testsuite/gas/elf/warn-2.s: Use the .nop directive. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug gas/26850] as -g: remove non-executable sections' contribution from DW_AT_ranges
https://sourceware.org/bugzilla/show_bug.cgi?id=26850 Nick Clifton changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #2 from Nick Clifton --- Hi Fangrui, Agreed. I have applied a patch to implement this suggestion. I also extended it to cover sections that are not loaded as well, since I do not see how debug information for them would be useful either. Cheers Nick -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/26869] ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu since 382aae06322799a25ea52fe61b243cbca4db8d66
https://sourceware.org/bugzilla/show_bug.cgi?id=26869
H.J. Lu changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
Last reconfirmed||2020-11-12
Ever confirmed|0 |1
Target Milestone|--- |2.36
--- Comment #4 from H.J. Lu ---
linker.c has
.{* Return TRUE if the symbol described by a linker hash entry H
. is going to be absolute. Linker-script defined symbols can be
. converted from absolute to section-relative ones late in the
. link. Use this macro to correctly determine whether the symbol
. will actually end up absolute in output. *}
.#define bfd_is_abs_symbol(H) \
. (((H)->type == bfd_link_hash_defined \
.|| (H)->type == bfd_link_hash_defweak) \
. && bfd_is_abs_section ((H)->u.def.section) \
. && !(H)->rel_from_abs)
.
elfxx-x86.h has
/* Return TRUE if the symbol described by a linker hash entry H is
going to be absolute. Similar to bfd_is_abs_symbol, but excluding
all linker-script defined symbols. */
#define ABS_SYMBOL_P(H) \
(bfd_is_abs_symbol (&(H)->root) && !(H)->root.ldscript_def)
But ldscript_def is set for __ehdr_start which is defined by linker
and isn't a absolute symbol.
--
You are receiving this mail because:
You are on the CC list for the bug.
[Bug ld/26869] ld: assertion fail ../../bfd/elflink.c:14806 on i686-linux-gnu since 382aae06322799a25ea52fe61b243cbca4db8d66
https://sourceware.org/bugzilla/show_bug.cgi?id=26869 --- Comment #5 from H.J. Lu --- A patch is posted at https://sourceware.org/pipermail/binutils/2020-November/114107.html -- You are receiving this mail because: You are on the CC list for the bug.
[Bug gas/26858] [arm] unsupported vmul instruction for vfp5-sp-d16 + armv8-r
https://sourceware.org/bugzilla/show_bug.cgi?id=26858 Andre Vieira changed: What|Removed |Added CC||avieira at gcc dot gnu.org Last reconfirmed||2020-11-12 Ever confirmed|0 |1 Assignee|unassigned at sourceware dot org |avieira at gcc dot gnu.org Status|UNCONFIRMED |NEW --- Comment #2 from Andre Vieira --- Confirmed this, typo in the ARM_VARIANT minimum requirement used the double precision feature bit instead of the single precision one. Testing a patch. -- You are receiving this mail because: You are on the CC list for the bug.
Issue 24906 in oss-fuzz: binutils:fuzz_readelf: Abrt with empty stacktrace
Updates: Labels: -restrict-view-commit -deadline-approaching Comment #4 on issue 24906 by sheriffbot: binutils:fuzz_readelf: Abrt with empty stacktrace https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=24906#c4 This bug has exceeded our disclosure deadline. It has been opened to the public. - Your friendly Sheriffbot -- You received this message because: 1. You were specifically CC'd on the issue You may adjust your notification preferences at: https://bugs.chromium.org/hosting/settings Reply to this email to add a comment.
