On 7/23/26 23:39, Miao Wang wrote:
Hi,
On Thu, 23 Jul 2026 17:44:30 +0200 Matthias Klose <[email protected]> wrote:
Control: tags -1 - patch
this patch is not yet available on the 2.47 release branch. Please work
with upstream to get this fixed upstream.
On 7/22/26 01:45, Miao Wang wrote:
Package: binutils-loongarch64-linux-gnu
Version: 2.46.90.20260712-1
Severity: serious
Tags: patch, upstream, fixed-upstream
Dear maintainer,
ld in binutils-loongarch64-linux-gnu 2.46.90.20260712-1 introduced a bug
where relocations R_LARCH_ALIGN are inserted to ld -r outputs even if
there is no such relocation in the input. This has broken the compilation
of Linux kernel modules, rendering unloadable modules due to R_LARCH_ALIGN
is not expected to be exist in kernel modules and cannot be handled by the
kernel module loader.
This has been fixed in upstream [1].
Cheers,
Miao Wang
[1]:
https://sourceware.org/cgit/binutils-gdb/patch/?id=a4481c8ff808e7237b99815d84b0d4c06ded6124
I'm a bit confused, since I have just double checked that the commit
a4481c8ff808 ("LoongArch: only insert align section for ld -r if an
input has R_LARCH_ALIGN or R_LARCH_RELAX") is shown in the git history
of the current master branch. Another commit e644656548fc ("[v2]
LoongArch: only insert align section for ld -r if an input has
R_LARCH_ALIGN or R_LARCH_RELAX") with the same changes is also shown
in the git history of the current binutils-2_47-branch branch. In the
refs tab of that repo, I cannot find such a branch or a tag which is
updated recently and contains the word "release". I wonder which is
actually the 2.47 release branch you are referring to.
is the suggested fix on the binutils-2_47-branch branch? I'm just
confused that you are referring to a commit on HEAD in the first place.
