- Arch-general - lists.archlinux.org

[arch-general] [HELP] Upgrading kernel results in unmounted partitions
by Thilina Mallawa Arachchi 08 Jun '14

08 Jun '14

Hi Arch Community, Today I upgraded the kernel from a fresh install of 3.14 to 3.4 via the yaourt install of linux-lts34 from AUR. Booting results logging into the recovery console as root. My / partition is being booted but my /var and /home are not. Very similar to this post https://bugs.archlinux.org/task/40706 and also this discussion on the list https://mailman.archlinux.org/pipermail/arch-general/2014-June/036488.html. The above links refer to using UTC time as a solution. I am using UTC time as per the beginners installation guide for the first fresh install of 3.14 but still run into the issue after the kernel upgrade. Would someone guide me (in detail) for fixing the issue? I need to run the 3.4 kernel for the particular piece of hardware I'm working with. Best, Thilina

2 2

[arch-general] Anybody unable to boot Arch after installing latest updates?
by Ralf Mardorf 08 Jun '14

08 Jun '14

Hi, likely an update I run after 2014-06-03 19:39 +0200 has broken something. I only can boot Arch after I commented out everything in /etc/fstab, excepted of /. I have no time for journalctl etc. right now. Does anybody experience the same or another issue? [root@archlinux rocketmouse]# cat /etc/fstab # # /etc/fstab: static file system information # # <file system> <dir> <type> <options> <dump> <pass> # UUID=f386221b-fc9c-434b-b4e7-b3b1b97be8aa LABEL=archlinux /dev/sda9 / ext3 rw,relatime,data=ordered 0 1 # UUID=3e19b702-30ed-4574-91d7-1594fad71842 #/dev/sdb7 none swap defaults 0 0 # UUID=4b65ccbf-2219-4734-98fa-16c58d4f60fa #/dev/sda10 none swap defaults,pri=-2 0 0 #/dev/sda11 /mnt/music ext3 noatime,defaults 0 2 #/dev/sdb12 /home/music ext4 noatime,defaults 0 2 #dev/sda14 #LABEL=debbsd /mnt/debbsd ext4 noatime,defaults 0 2 [snip] Regards, Ralf

2 2

Re: [arch-general] Unwanted forced fsck at each startup - Was: How do I _really_ fix the superblock?
by Ralf Mardorf 08 Jun '14

08 Jun '14

On Sun, 2014-06-08 at 09:38 +0200, Ralf Mardorf wrote: > On Sun, 2014-06-08 at 08:23 +0200, Ralf Mardorf wrote: > > On Sat, 2014-06-07 at 23:24 +0200, Ralf Mardorf wrote: > > > On Sat, 2014-06-07 at 17:13 -0400, Leonid Isaev wrote: > > > > I'd remove the fsck hook to let systemd check the filesystems and see > > > > if that helps. > > > > > > Done, but I won't reboot now. I will report back next time I (re)boot > > > Arch Linux. > > > > fsck still is forced to run :( > > # downgrade systemd libsystemd systemd-sysvcompat > # pacman -Q systemd libsystemd systemd-sysvcompat > systemd 212-3 > libsystemd 212-3 > systemd-sysvcompat 212-3 This doesn't solve the issue and that is very, very bad :(. What package should I downgrade too, to get rid of the brand new "run fsck at each startup" issue? I could restore Arch from a working backup I made a few days ago, but what packages shouldn't I upgrade, after I restored my Arch install?

1 0

[arch-general] How do I _really_ fix the superblock?
by Ralf Mardorf 08 Jun '14

08 Jun '14

Hi, I still need to comment out everything in fstab, excepted of / (it does include home) to boot Arch Linux, so at least I'm able to boot. Anyway, each time I boot, I get a message that the superblock's last write time was in the future, it always gets fixed and than always the / partition is checked. The Arch Linux partition is an ext3 FS, so I run fsck.ext3 -p from a live media and it pretended too, that the issue is fixe, but it isn't. JFTR the hardware clock does work correctly. I suspect it isn't related to the perl pain, that makes a lot of important software from AUR unusable on my machine. Could systemd be the culprit? [rocketmouse@archlinux ~]$ grep systemd /var/log/pacman.log | grep 2014-06-04 [2014-06-04 08:34] [PACMAN] upgraded libsystemd (212-3 -> 213-5) [2014-06-04 08:34] [PACMAN] upgraded systemd (212-3 -> 213-5) [2014-06-04 08:34] [PACMAN] upgraded systemd-sysvcompat (212-3 -> 213-5) I could restore my Arch from a backup to fix it, but than I wouldn't know what packages to upgrade and what package upgrades I need to ignore. JFTR after I booted Arch Linux I can remove the #'s from fstab and mount everything in fstab. Regards, Ralf

6 22

[arch-general] misleading pacman errors when upgrade fails
by Carl Schaefer 08 Jun '14

08 Jun '14

I just tried to upgrade three packages, including core/filesystem, and saw these errors: filesystem-2014.05-... 8.7 KiB 4.23M/s 00:00 [######################] 100% e2fsprogs-1.42.10-1... 691.8 KiB 1305K/s 00:01 [######################] 100% cups-filters-1.0.54... 648.6 KiB 2033K/s 00:00 [######################] 100% (3/3) checking keys in keyring [######################] 100% (3/3) checking package integrity [######################] 100% (3/3) loading package files [######################] 100% (3/3) checking for file conflicts [######################] 100% (3/3) checking available disk space [######################] 100% (1/3) upgrading cups-filters [######################] 100% (2/3) upgrading filesystem [######################] 100% warning: /etc/passwd installed as /etc/passwd.pacnew warning: /etc/group installed as /etc/group.pacnew error: extract: not overwriting dir with file /usr/local/share/man error: problem occurred while upgrading filesystem error: could not commit transaction error: failed to commit transaction (transaction aborted) Errors occurred, no packages were upgraded. The failure was apparently caused by /usr/local/share/man being a real directory on my system, when the "filesystem" package wants it to be a symlink to "../man". I imagine pacman could check in advance for this sort of extraction problem for all packages before upgrading or installing any packages, which could be part of the existing "checking for file conflicts" phase. Also, the error messages about the transaction being aborted and no packages being upgraded are wrong. The upgrade of "cups-filters" was successful, and "filesystem" appears to have been upgraded too. The only upgrade that was aborted was "e2fsprogs" that pacman hadn't started. Is any of this worth reporting as bugs or enhancements? Carl

1 0

[arch-general] Forgot username and password
by Bhushan Shah 06 Jun '14

06 Jun '14

Hello I registered to Arch Linux bug tracking using this email address long ago but I forgot my username and password. Can anyone send me my credentials? Thanks! -- Bhushan Shah http://bhush9.github.io IRC Nick : bshah on Freenode

2 2

[arch-general] spamassassin and perl-net-dns fail to find nameserver
by Florian Pritz 06 Jun '14

06 Jun '14

Hi, I've worked around the error caused by net-dns being a developer release, but SA still fails to find nameservers so DNSBL checks will not be used. If you have the same problem add 'dns_server 127.0.0.1' or similar to your /etc/mail/spamassassin/local.cf or better yet figure out why it fails and contact the right people. I currently don't have enough time to take a good look at the issue, but I figured I should probably write this so you can at least properly block spam. Happy mailing and sorry for all the perl breakage. Ideas on how to prevent them (library soname? sadly that seems unsupported by upstream afaict) are very welcome.

2 1

[arch-general] 'grub_term_highlight_color' not found
by Dimitris Zervas 05 Jun '14

05 Jun '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 So, yesterday night, I decided to update my system. I had 4-5 hours to deal with all the nuclear blast that an update of 3 months would cause (the good old arch we all love... I don't know which is worse, rm -rm / or pacman -Syu?). After 10-20 everything was updated and decided to restart. Also, I wanted to change the boot order of my UEFI and boot arch by default instead of win (finals are over). I made the change and saved. The scree was blank for ~10s and I decided to hard reboot. BOOM. The uefi entry was lost. Also, if I launched grub2 as a uefi shell (which I have for times like that) I got "error: symbol 'grub_term_highlight_color' not found" and grub gives me a rescue terminal. The first problem (uefi entry) is very painful. I have an Asus P9X79 Deluxe. It does NOT support v2 uefi shell (so, no bcfg) and no matter how hard I try to add an entry with efibootmgr, nothing happens (via SysResCD, uefi mode boot with modprobe efivars) The second problem (grub pukes) I have no clue what's going on. I just updated, even tried grub-mkconfig via chroot and nothing changed. I found 2 threads on arch forums which where not helpful at all. Any help appreciated :) -----BEGIN PGP SIGNATURE----- Version: APG v1.1.1 iQJCBAEBCgAsBQJTkC/tJRxEaW1pdHJpcyBaZXJ2YXMgPGR6ZXJ2YXNAZHplcnZh cy5ncj4ACgkQ+77hPkfJypZSeA/8DcxVPg+WWDva/E2hfjX5e1OFTMJbxcweWmBG hDoNDShtm1sDfJ057fmidYqozud6ytNpug9qu+ixIBCRqmOfETjcSogNkr09Dfs5 zr0Opm15AQDlCmXcnfFWY8cPeEl8I+LqdoJM5RMM3Z6an6QmvVETdxVQ6wxouVgL AvHd7Lk/aNAYLN6uI8JFBOX0H9CzjCB/sviYmOwmVkFZLlVOKkJiYpWHL8p4pUCp Wf+2CUu42wvvvaSdrdiddc1WVHcCknr1HEYK4juiJCCgCUldUwRZGXJiS2j/0ZH/ L+VoufUMqlc+NWjV67PtwhmtVl9jvrcymAA2UOAnrWAvZZjynMyFIspdDiBffWP+ Cw+01oubUDPW2NI/9HiLm2u0QohemlYHPoFoM1a5nhsXG0pC/6qqzT0DEy3K2pl2 a+FkNum8KVl6vmtS3JXvXeH1HRKJ6aIK7qSinATKjK/Ust7saC0TsHZieWSNqZ+v 7OFOd5IYkMJyKxBuwvrpcplJmfbYuSnjRwP+OYf7nsZXNKKhPjvYi5mO1hw1fHOK Bd9GMxOjOaPrQHXSl8iJ3BTUXAXcBGZ/gbXyqyx6mieWwOGuhix/IwkoVsJU2j3i 6/TsGhykK5STdMG6augYugcPkkgdjqCteJx+LHP0By45bY6e7S9WOnmpJdzubWPJ nKoE3JY= =uxTT -----END PGP SIGNATURE-----

3 4

[arch-general] Perl 5.20 in [testing]
by Florian Pritz 05 Jun '14

05 Jun '14

Hi, Perl 5.20, as any other new perl version, requires all modules that are not purely perl code to be rebuilt. We did that for all packages in our repos. For a list of upstream changes please refer to `man perldelta`. Since users probably installed some from AUR or with CPANPLUS::Dist::Arch, I wrote a script[1] that generates a local rebuild list. [1]: https://git.server-speed.net/users/flo/bin/plain/find-broken-perl-packages.… - raw.txt contains a list of files that generated an error - perl-modules.txt contains a list of modules the files belong to - perl-dists.txt contains a list of distributions - pacman.txt contains a list of pacman packages the files belong to Binaries linking with libperl.so will also need to be rebuilt. You can use lddd from devtools to find those. If you get an error that perl-scalar-list-utils can't be found when running -Syu install it with -S perl-scalar-list-utils. Seems to be a bug in pacman. Please report any issue you encounter.

5 12

[arch-general] for raid10 users
by Łukasz Michalski 05 Jun '14

05 Jun '14

Hi, Last year I had a lot of crashes on my linux box. I experienced hard reset or processes hanged in "D" state forever. It was much more likely to happen when raid0 array was resyncing. Yesterday I finally managed to track this problem and it turned out that it was related to raid10 kernel module. After applying this patch: https://lkml.org/lkml/2014/6/4/768 My problems are gone. Today I forced array resync, run kernel kompile and two VirtualBox machines with heavy I/O load. On arch linux kernel my machine hanged in 2-3 minutes. Now everything works ok. Regards, Łukasz

1 0