Bug #1234
[librechroot] `librechroot -A armv7h make` aborts during `locale-gen`
100%
Description
$ sudo librechroot -A armv7h -n armv7h make [...] :: Running post-transaction hooks... [...] Generating locales... en_US.UTF-8...localedef: ../sysdeps/unix/sysv/linux/spawni.c:360: __spawnix: Assertion `ec >= 0' failed. qemu: uncaught target signal 6 (Aborted) - core dumped /usr/bin/locale-gen: line 41: 27 Aborted (core dumped) localedef -i $input -c -f $charset -A /usr/share/locale/locale.alias $locale
The problem isn't specific to librechroot, I have confirmed it in 2 machines by running locale-gen inside a chroot not made by librechroot. However it prevents everything after locale-gen from occurring, with awful consequences for libremakepkg:
$ sudo libremakepkg -n armv7h [...] ==> Starting to build the package... | /usr/bin/cp: cannot stat '/repo/repo.db': No such file or directory | ==> Checking runtime dependencies... | warning: database file for 'repo' does not exist | ==> Installing missing dependencies... | warning: database file for 'repo' does not exist | error: failed to prepare transaction (could not find database) | ==> ERROR: 'pacman' failed to install missing dependencies. ==> Copying log and package files out of the chroot...
History
Updated by isacdaavid about 6 years ago
upstream report https://bugs.launchpad.net/qemu/+bug/1673976
Updated by oaken-source about 6 years ago
A workaround for the failing locale-gen call is running in the chroot:
# gunzip --keep /usr/share/i18n/charmaps/UTF-8.gz
afterwards, locale-gen succeeds:
# locale-gen Generating locales... en_US.UTF-8... done de_DE.UTF-8... done Generation complete.
Now, is there any way that I can take this to a working chroot? what other steps are required?
Updated by isacdaavid about 6 years ago
oaken-source wrote:
Now, is there any way that I can take this to a working chroot? what other steps are required?
Last time I checked just touching an empty `/repo/repo.db` was able to do the trick.
Updated by lukeshu over 5 years ago
I wonder if this is "caused" by librechroot not calling setarch before calling mkarchroot; in contrast with Arch's archbuild, which does call setarch on mkarchroot.
Updated by lukeshu over 5 years ago
Upon further reflection, that was a dumb thought. We can't call setarch to ARM from x86; how could not calling it be the issue?
Updated by lukeshu about 5 years ago
- % Done changed from 0 to 100
- Status changed from open to fixed
This is fixed in current versions of things. I don't know what was causing it.