Stale package in Rawhide
Hi, I maintain automatic weekly builds of DoubleCmd in COPR. Recently the project started to fail due to fpc upgrade without a lazarus rebuild against it. The maintainer make a new build 4 days ago: https://koji.fedoraproject.org/koji/taskinfo?taskID=25282137 however, that package is not visible in Rawhide still braking my COPR build: https://copr-be.cloud.fedoraproject.org/results/vondruch/doublecmd/fedora-rawhide-x86_64/00721408-doublecmd-gtk/builder-live.log And in fact, there is no new package on the server: https://dl.fedoraproject.org/pub/fedora/linux/development/rawhide/Everything/x86_64/os/Packages/l/ > lazarus-1.6.4-3.fc27.x86_64.rpm 2017-08-20 13:22 98M (the same is for fc28) Regarding the schedule it's not a freeze time. Is there anything else that need to be done to push the new package to rawhide (in addition to a successful Koji build)? Marcin -- http://blog.solidsoft.info/ - Working code is not enough ___ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-le...@lists.fedoraproject.org
Re: Duplicated conflicting packages after failed dnf system-upgrade from 26 to 27
On 2017-10-18 23:28, Samuel Sieb wrote: > On 10/18/2017 12:59 PM, Marcin Zajączkowski wrote: >> How can I solve that? > > What happens if you add the suggested "--best --allowerasing" to the > command? I was too obvious :). As predicted it removed libcrypt-nss and installed libcrypt. Also chromium-libs-media was removed. Thanks. Marcin ___ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-le...@lists.fedoraproject.org
Re: How to change that list subsciption?
Marcin Zajączkowski composed on 2017-10-18 01:56 (UTC+0200): > Always hard for me to figure out too. :-( Try: > https://lists.fedoraproject.org/admin/accounts/per-subscription-preferences/ Thanks. The thing that midled me was the fact that I cannot create an account directly. However, it was possible to use a 3rd-party provider to do that. Marcin ___ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-le...@lists.fedoraproject.org
Re: Duplicated conflicting packages after failed dnf system-upgrade from 26 to 27
On Wed, 2017-10-18 at 10:02 +0200, Kamil Paral wrote:
> I'd also like to see the logs. If the system is broken enough that
> journalctl doesn't work, you can copy the files out of /var/log/journal
> and examine them later with journalctl from another system.
I put them in the previous message.
> My standard strategy for dealing with this situation is to remove the
> older side of all the duplicate pairs with:
>
> rpm -e --justdb --noscripts
>
> and then 'dnf reinstall' the newer side. You can hack up a script to
> help with this, of course (I don't keep one around, I just reinvent it
> every time).
Thanks, "--justdb" is something that I missed to do cope with the files. "dnf
reinstall" is also useful - I didn't know that and I had to reinstall packages
to "fix" missing/changed files with "rpm -U --force" which requires more effort.
For the reference for people with similar problems reading that list in the
future, in the meantime I used the more bumpy way and removed all fc27 packages
using "rpm -e $(cat list-of-fc27-packages)" to perform an update once again.
However, it resulted in the fact that fc27 packages were removed (only fc26
packages remained), but applications started to complain that there is no
gllibc in F26 version (only F27). "rpm -V" reports that files are corrupted. I
suspect that "rpm -e" removed newer packages, but as the files were conflicted
(and already upgraded to F27 at the filesystem) they remained. I needed to
manually reinstall (rpm -U --force) the key packages to bring dnf to life again
(e.g. issue with SSL certificated in curl). Later on I redownloaded f27
packages ("sudo dnf system-upgrade download --releasever=27"), forced proper
system state (symlink, etc.) with "sudo dnf system-upgrade reboot" (system
didn't reboot as I was chrooted) and manually call "sudo dnf --releasever=27 s
ystem-upgrade upgrade" (taken from "dnf-system-upgrade.service" file) to
observe the output. Everything went fine and after restart I was able to boot
my Fedora 27!
Most things seem to work at the first look (I had to downgrade Firefox to 56 to
bring all legacy packages to life), but I still have some conflicted packages.
$ sudo dnf update
Last metadata expiration check: 1:23:03 ago on 2017-10-18T20:16:58 CEST.
Dependencies resolved.
Problem 1: cannot install both
chromium-libs-media-freeworld-59.0.3071.109-5.fc27.x86_64 and
chromium-libs-media-freeworld-61.0.3163.100-1.fc27.x86_64
- cannot install the best update candidate for package
chromium-libs-media-freeworld-61.0.3163.100-1.fc27.x86_64
- cannot install the best update candidate for package
chromium-libs-media-61.0.3163.100-1.fc27.x86_64
Problem 2: package libcrypt-nss-2.26-8.fc27.i686 requires glibc(x86-32) =
2.26-8.fc27, but none of the providers can be installed
- cannot install both glibc-2.26-14.fc27.i686 and glibc-2.26-8.fc27.i686
- cannot install both glibc-2.26-8.fc27.i686 and glibc-2.26-14.fc27.i686
- cannot install the best update candidate for package
libcrypt-nss-2.26-8.fc27.i686
- cannot install the best update candidate for package glibc-2.26-8.fc27.i686
Problem 3: package pam-1.3.0-6.fc27.i686 requires libcrypt.so.1, but none of
the providers can be installed
- libcrypt-nss-2.26-8.fc27.i686 has inferior architecture
- cannot install both libcrypt-nss-2.26-14.fc27.x86_64 and
libcrypt-nss-2.26-8.fc27.x86_64
- package libcrypt-nss-2.26-14.fc27.x86_64 conflicts with libcrypt provided
by libcrypt-2.26-14.fc27.i686
- package libcrypt-2.26-8.fc27.i686 conflicts with libcrypt-nss provided by
libcrypt-nss-2.26-14.fc27.x86_64
- cannot install both libcrypt-nss-2.26-8.fc27.x86_64 and
libcrypt-nss-2.26-14.fc27.x86_64
- cannot install the best update candidate for package pam-1.3.0-6.fc27.i686
- cannot install the best update candidate for package
libcrypt-nss-2.26-8.fc27.x86_64
Problem 4: package cups-libs-1:2.2.4-6.fc27.i686 requires libcrypt.so.1, but
none of the providers can be installed
- package libcrypt-nss-2.26-14.fc27.x86_64 conflicts with libcrypt provided
by libcrypt-2.26-14.fc27.i686
- package libcrypt-nss-2.26-8.fc27.i686 requires glibc(x86-32) = 2.26-8.fc27,
but none of the providers can be installed
- problem with installed package libcrypt-nss-2.26-8.fc27.x86_64
- libcrypt-2.26-8.fc27.i686 has inferior architecture
- glibc-2.26-8.fc27.i686 has inferior architecture
- package libcrypt-nss-2.26-8.fc27.x86_64 requires glibc(x86-64) =
2.26-8.fc27, but none of the providers can be installed
- package libcrypt-2.26-8.fc27.x86_64 requires glibc(x86-64) = 2.26-8.fc27,
but none of the providers can be installed
- cannot install both glibc-2.26-14.fc27.x86_64 and glibc-2.26-8.fc27.x86_64
- cannot install both glibc-2.26-8.fc27.x86_64 and glibc-2.26-14.fc27.x86_64
- cannot install the best update candidate for package
glibc-2.26-8.fc27.x86_64
- cannot install the best update candidate for package
cups-libs-1:2.2.4-6.fc27.i686
Problem 5:
Re: Duplicated conflicting packages after failed dnf system-upgrade from 26 to 27
> I will try to get it in the evening.
Unfortunately dnf.log ends at the moment when the packages are started to be
updated (doing a separate "dnf update" session I don't see anything during the
update process itself, only before and after):
Transaction Summary
Install 146 Packages
Upgrade3172 Packages
Remove6 Packages
Downgrade 9 Packages
2017-10-17T18:28:37Z INFO Total size: 3.0 G
2017-10-17T18:28:37Z INFO Downloading Packages:
2017-10-17T18:28:37Z SUBDEBUG Call: RPMPayload._end_cb: (, 1, 'Already downloaded'), {}
(...)
2017-10-17T18:28:38Z SUBDEBUG Call: RPMPayload._end_cb: (, 1, 'Already downloaded'), {}
2017-10-17T18:28:38Z SUBDEBUG Call: RPMPayload._end_cb: (, 1, 'Already downloaded'), {}
2017-10-17T18:28:47Z INFO Running transaction check
2017-10-17T18:28:51Z INFO Transaction check succeeded.
2017-10-17T18:28:51Z INFO Running transaction test
2017-10-17T18:28:59Z INFO Transaction test succeeded.
2017-10-17T18:28:59Z DDEBUG timer: transaction test: 7137 ms
2017-10-17T18:28:59Z INFO Running transaction
2017-10-17T18:29:01Z DDEBUG RPM transaction start.
(here stars another session started manually from chroot)
2017-10-18T01:31:01Z INFO --- logging initialized ---
2017-10-18T01:31:01Z DDEBUG timer: config: 12 ms
2017-10-18T01:31:01Z DEBUG Loaded plugins: builddep, config-manager, copr,
debug, debuginfo-install, download, generate_completion_cache, needs-restar
ting, playground, repoclosure, repograph, repomanage, reposync, system-upgrade
2017-10-18T01:31:01Z DEBUG DNF version: 2.7.3
(...)
The system journal from the session with dnf upgrade after reboot terminates
unexpectedly:
Oct 17 20:27:55 foobar systemd[1]: Reached target System Initialization.
Oct 17 20:27:55 foobar kernel: audit: type=1130 audit(1508264875.449:92): pid=1
uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0
msg='unit=systemd-update-utmp comm="systemd" exe="/usr/lib/s
Oct 17 20:27:55 foobar systemd[1]: Starting System Upgrade using DNF...
Oct 17 20:27:55 foobar systemd[1]: Started Updates device firmware whilst
offline.
Oct 17 20:27:55 foobar audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295
ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=fwupd-offline-update
comm="systemd" exe="/usr/lib/systemd/systemd" hostname=?
Oct 17 20:27:55 foobar systemd[1]: Listening on D-Bus System Message Bus Socket.
Oct 17 20:27:55 foobar systemd[1]: Starting Update the operating system whilst
offline...
Oct 17 20:27:55 foobar pk-offline-update[807]: another framework set up the
trigger
Oct 17 20:27:55 foobar systemd[1]: Started Update the operating system whilst
offline.
Oct 17 20:27:55 foobar audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295
ses=4294967295 subj=system_u:system_r:init_t:s0
msg='unit=packagekit-offline-update comm="systemd"
exe="/usr/lib/systemd/systemd" hostn
Oct 17 20:27:55 foobar audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295
ses=4294967295 subj=system_u:system_r:init_t:s0
msg='unit=packagekit-offline-update comm="systemd"
exe="/usr/lib/systemd/systemd" hostna
Oct 17 20:27:55 foobar audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295
ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=fwupd-offline-update
comm="systemd" exe="/usr/lib/systemd/systemd" hostname=?
Oct 17 20:27:56 foobar systemd[1]: Received SIGRTMIN+20 from PID 362
(plymouthd).
Oct 17 20:27:58 foobar audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295
ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-rfkill
comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=?
Oct 17 20:27:58 foobar systemd[1]: Received SIGRTMIN+20 from PID 362
(plymouthd).
Oct 17 20:27:59 foobar dnf[805]: Last metadata expiration check: 0:18:43 ago on
Tue 17 Oct 2017 08:09:16 PM CEST.
Oct 17 20:27:59 foobar system-python[805]: Starting system upgrade. This will
take a while.
Oct 17 20:28:19 foobar dnf[805]: Dependencies resolved.
(...)
Oct 17 20:37:00 foobar dnf[805]: Upgrading:
pygtk2-codegen-2.24.0-18.fc27.x86_64 1368/6548
Oct 17 20:37:00 foobar dnf[805]: Running scriptlet:
glibc-headers-2.26-8.fc27.x86_64 1369/6548
Oct 17 20:37:01 foobar dnf[805]: Upgrading:
glibc-headers-2.26-8.fc27.x86_64 1369/6548
Oct 17 20:37:01 foobar dnf[805]: Upgrading:
glibc-devel-2.26-8.fc27.x86_64 1370/6548
(... end ...)
I left the computer alone during the operation and it was shut down when I came
back. My system suffers from 100% CPU used by kworker:
https://bugzilla.redhat.com/show_bug.cgi?id=1115658#c4 (there is open bug, but
I cannot find it right now). It might got overheat and did emergency shutdown
(I haven't seen that before, thus). It was very hot and I see in that log:
Oct 17 20:33:34 foobar kernel: CPU1: Core temperature above threshold, cpu
clock throttled (total events = 1)
Therefore, I suspect it wasn't a
Re: Duplicated conflicting packages after failed dnf system-upgrade from 26 to 27
On Wed, Oct 18, 2017 at 1:50 AM, Marcin Zajączkowski <mszpak(a)wp.pl wrote:
> Honestly if it crashed during system upgrade, it might be easiest to
> perform a clean F27 installation (keep your existing /home partition, or
> move the user data to an external drive and back).
That's something I would prefer to avoid (it would be my personal failure ;) ).
But definitely I consider that as a last resort.
> However, I'd be very interested to see a log of the upgrade process, to
> determine why it failed. Could you file a bug against
> dnf-plugin-system-upgrade, link it here, and attach journalctl output of
> the upgrade process ("journalctl -b $number" when chrooted, where $number
> can be discovered from "journalctl --list-boots") and /var/log/dnf*.log in
> it? Thanks.
I will try to get it in the evening.
Marcin
___
test mailing list -- test@lists.fedoraproject.org
To unsubscribe send an email to test-le...@lists.fedoraproject.org
Re: Duplicated conflicting packages after failed dnf system-upgrade from 26 to 27
I will try `dnf distro-sync`, I would prefer to do not install it from scratch.
In the meantime I removed all F27 packages with `rpm` do get rid of conflicts.
However, it brought me to another issue with CA and curl ("curl error: Problem
with the SSL CA cert (path access rights?)") and the fact that
`update-ca-trust` doesn't see glibc in the proper version...
Anyway, I wonder if it is safe to install packages with dnf being chrooted?
Marcin
___
test mailing list -- test@lists.fedoraproject.org
To unsubscribe send an email to test-le...@lists.fedoraproject.org
How to change that list subsciption?
Hello again, That may be not the best place, however, I have a problem with changing the delivery options for that mailing list. I've been long subscribed to that list, but I had mail delivery disabled. I cannot find old mailman interface and in the new one (https://lists.fedoraproject.org/admin/lists/test.lists.fedoraproject.org/) I cannot create an account as "We are sorry, but the sign up is currently closed.". I would like to enable delivery to mail mailbox. How can I do that? Marcin ___ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-le...@lists.fedoraproject.org
Duplicated conflicting packages after failed dnf system-upgrade from 26 to 27
Hello guys. I was upgrading Fedora with the system-upgrade plugin from 26 to 27. The upgrade process failed for unknown reasons (during the packages installing), my computer rebooted and I'm not able to use my Fedora :). I booted up with LiveCD, chrooted to my broken Fedora and I tried to resume the upgrade process. Unfortunately there is a lot (>200) conflicting packages (for Fedora 26 and 27). I tried to use "dnf remove --duplicated", however it would like to remove also all dependent packages. Q. What can I resume update/upgrade process having duplicated packages with DNF (install new packages and remove all)? Marcin ___ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-le...@lists.fedoraproject.org
Bug in uvcvideo kernel module makes Fedora 16 LiveCD unusable for part of users
Hi, Due to a kernel bug in uvcvideo module [1] Fedora 16 doesn't work properly with specific web camera model(s) (at least available in Asus N43S series). It's required to blacklist that module in /etc/modprobe.d/blacklist.conf (boot parameter rd.blacklist doesn't help) which is impossible with LiveCD (without ISO modification). Maybe it would a good workaroud to temporarily blacklist that module in Fedora 16 (+ make a note in common bugs how to enable that module to use web camera) to allow those people to test further Fedora 16 version from LiveCD)? [1] - https://bugzilla.redhat.com/show_bug.cgi?id=734963 Regards Marcin -- http://solidsoft.wordpress.com/ - Working code is not enough -- test mailing list test@lists.fedoraproject.org To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test
