This issue was fixed with commit 6b6fb7872486, "gnu: glibc: Build with
'--strip-debug' instead of '--strip-all'."
--
Simon South
si...@simonsouth.net
I too have been affected by this bug for a long time and hope it will be fixed
one day.
For end-users, attached is a drop-in workaround you can use to make guix
utilize the old cryptsetup build. It's sloppy, but at least the system can boot
again.
Thanks for the research, Simon.
;; workaround fo
Here's a patch that fixes this problem by modifying the glibc package so the
"__pthread_key_create" symbol in its pthread library is preserved, as opposed
to being stripped off as it is today.
This tests fine for me on both AArch64 and x86-64: Stepping through the code
in gdb I can see libgcc's __
After some testing I've found the regression was introduced in commits
f32a6055a5 and e0f31baacc, "build-system/gnu: strip with
--strip-unneeded", which replace "--strip-debug" with "--strip-unneeded"
for packages that use the GNU build system. It seems this is now
stripping a bit too much.
The s
This seems to be caused by the symbol "__pthread_key_create" being
absent from cryptsetup's symbol table during linking, even though it is
meant to be exported explicitly by glibc.
The fundamental issue is that cryptsetup's Argon2i implementation (and
thus cryptsetup itself) is multithreaded, but
Currently cryptsetup from the "cryptsetup-static" package is unable to
open LUKS2 encrypted volumes that use the Argon2i key-derivation
algorithm, the default for LUKS2. It catches SIGABRT and exits without
opening the volume.
This appears to be a regression following the merge of the
core-update
