[opensuse-factory] factory-tested status
Hi folks, Factory iso build numbers have finally passed the 1000. Of course that is completely meaningless, so now to something completely different: the important stuff. I had started to include many important development projects in the automated testing, so that there is a chance to find bugs before they even hit Factory... This has the advantage that it might be easier to tell what is responsible for breakage. As for the random kernel crashes observed since three weeks https://bugzilla.novell.com/show_bug.cgi?id=660464 there is hope, that linux-2.6.37 already has a fix. Then I had added to openQA a regression test for sntp https://bugzilla.novell.com/show_bug.cgi?id=657626 and found that there is already a newer version of ntp waiting in network:time which is fixing this problem. The extra-repo testing so far had turned up a segfault in KDE-Playground's amarok that already got fixed and a more complicated problem in GNOME-devel's gdm concerning VT-allocation and Xorg. tested extra-repos now include: Kernel:/HEAD X11:/XOrg Base:/System KDE:/Distro:/Factory KDE:/Unstable:/Playground GNOME:/Factory X11:/xfce Then libreoffice and kontact are missing since a new build(0889). And GNOME started to look rather different and complained about firefox not being its default browser http://openqa.opensuse.org/results/openSUSE-GNOME-LiveCD-x86_64-Build1001 And finally... but this is getting into minor details... there is "amarok /usr/share/sounds/alsa/test.wav" not queuing this file (working on KDF and KDE-playground) Oh and akonadi is wasting 128MB of RAM on KDE-LiveCDs for mysql binlogs du -s ~/.local/share/akonadi/db_data/ 143652 https://bugs.kde.org/show_bug.cgi?id=260630 has workarounds I hope we can get fixes for many of the above issues into Factory next week (which is btw. your last chance to get (well-tested) code into Milestone 6 of 6). This should make Milestone 6 of 6 a nice one. Ciao Bernhard M. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Sun, 09 Jan 2011 23:08:11 +0100
"Bernhard M. Wiedemann"
Oh and akonadi is wasting 128MB of RAM on KDE-LiveCDs for mysql binlogs du -s ~/.local/share/akonadi/db_data/ 143652 https://bugs.kde.org/show_bug.cgi?id=260630 has workarounds
This was a good hint. Changing the DB file size in the config file makes akonadi crash immediately and prevents it from starting - something I was not able to achieve otherwise. No matter what I did, it always started up and complained after a few minutes that nepomuk was not runnign - no wonder, as I had disabled it in systemsettings :-) -- Stefan Seyfried "Dispatch war rocket Ajax to bring back his body!" -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le dimanche 09 janvier 2011, à 23:08 +0100, Bernhard M. Wiedemann a écrit :
And GNOME started to look rather different and complained about firefox not being its default browser http://openqa.opensuse.org/results/openSUSE-GNOME-LiveCD-x86_64-Build1001
That's because gconf2-branding-openSUSE is not installed. I don't know why, though. Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Am Montag, 10. Januar 2011 schrieb Vincent Untz:
Le dimanche 09 janvier 2011, à 23:08 +0100, Bernhard M. Wiedemann a écrit :
And GNOME started to look rather different and complained about firefox not being its default browser http://openqa.opensuse.org/results/openSUSE-GNOME-LiveCD-x86_64-Build1001
That's because gconf2-branding-openSUSE is not installed. I don't know why, though.
nothing provides desktop-data-openSUSE = 11.3 needed by gconf2-branding- openSUSE-11.3-3.5.noarch Greetings, Stephan -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le lundi 10 janvier 2011, à 11:10 +0100, Stephan Kulow a écrit :
Am Montag, 10. Januar 2011 schrieb Vincent Untz:
Le dimanche 09 janvier 2011, à 23:08 +0100, Bernhard M. Wiedemann a écrit :
And GNOME started to look rather different and complained about firefox not being its default browser http://openqa.opensuse.org/results/openSUSE-GNOME-LiveCD-x86_64-Build1001
That's because gconf2-branding-openSUSE is not installed. I don't know why, though.
nothing provides desktop-data-openSUSE = 11.3 needed by gconf2-branding- openSUSE-11.3-3.5.noarch
Ah right, I fixed this on Saturday. Let me push it now to Factory. Thanks, Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Bernhard M. Wiedemann write:
Hi folks,
Factory iso build numbers have finally passed the 1000. Of course that is completely meaningless, so now to something completely different: the important stuff.
I had started to include many important development projects in the automated testing, so that there is a chance to find bugs before they even hit Factory...
Hi, nice improvement. How often do you test such development projects? ( So I know how long should I keep newer version in Base:System before I submit it to factory ) Josef -- Josef Reidinger Appliance Toolkit team maintainer of perl-Bootloader, yast2-bootloader and parts of webyast and SLMS -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Am 10.01.2011 09:58, schrieb Josef Reidinger:
Bernhard M. Wiedemann write:
Hi folks,
Factory iso build numbers have finally passed the 1000. Of course that is completely meaningless, so now to something completely different: the important stuff.
I had started to include many important development projects in the automated testing, so that there is a chance to find bugs before they even hit Factory...
Hi, nice improvement. How often do you test such development projects? ( So I know how long should I keep newer version in Base:System before I submit it to factory ) Josef
At the moment, it might take 12-24h until a certain project is tested. And it is done at most once for each Buildnr+arch (I can always manually re-trigger, though). The test scheduler http://openqa.opensuse.org/opensuse/tools/testloop first does the main tests with no extra repos and then picks a random target, to prevent targets from never being tested when there are frequent new Builds. It is using a shuffled list of targets to more evenly spread the testing. I am also working to increase the testing throughput, so wait-times might get shorter later. Ciao Bernhard M. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le dimanche 09 janvier 2011, à 23:08 +0100, Bernhard M. Wiedemann a écrit :
The extra-repo testing so far had turned up a segfault in KDE-Playground's amarok that already got fixed and a more complicated problem in GNOME-devel's gdm concerning VT-allocation and Xorg.
Part of this is a bug in Xorg that was eating 100% of the CPU, and after some more investigation, the fix is now submitted (sr#57731). But that's not the full fix. Since gdm lets Xorg decides on which vt to start, Xorg chooses the first free one. And since Xorg is started before the end of the boot process, and before init starts the mingetty processes, we end up with Xorg on vt2. And a bit later, a mingetty is started on tty2 (following what's in /etc/inittab). The issue is that whatever you type in X is then also sent to mingetty, and you can see failed attempts to login in /var/log/messages. (Note: it works fine with systemd) I'm not sure what's the proper fix here: should init check there's nothing on tty2 before starting mingetty there? Or should we just force Xorg to start on vt7 anyway? Cheers, Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Mon, 2011-01-10 at 16:55 +0100, Vincent Untz wrote:
Le dimanche 09 janvier 2011, à 23:08 +0100, Bernhard M. Wiedemann a écrit :
The extra-repo testing so far had turned up a segfault in KDE-Playground's amarok that already got fixed and a more complicated problem in GNOME-devel's gdm concerning VT-allocation and Xorg.
Part of this is a bug in Xorg that was eating 100% of the CPU, and after some more investigation, the fix is now submitted (sr#57731).
But that's not the full fix. Since gdm lets Xorg decides on which vt to start, Xorg chooses the first free one. And since Xorg is started before the end of the boot process, and before init starts the mingetty processes, we end up with Xorg on vt2. And a bit later, a mingetty is started on tty2 (following what's in /etc/inittab).
The issue is that whatever you type in X is then also sent to mingetty, and you can see failed attempts to login in /var/log/messages.
(Note: it works fine with systemd)
I'm not sure what's the proper fix here: should init check there's nothing on tty2 before starting mingetty there? Or should we just force Xorg to start on vt7 anyway?
I guess we should always force X/gdm to be on the active tty, regardless of any splash logic, and along with that just never start getty1 in runlevel 5. Fedora officially does not support booting without plymouth, hence they have such logic always applied. Kay -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le lundi 10 janvier 2011, à 17:34 +0100, Kay Sievers a écrit :
On Mon, 2011-01-10 at 16:55 +0100, Vincent Untz wrote:
I'm not sure what's the proper fix here: should init check there's nothing on tty2 before starting mingetty there? Or should we just force Xorg to start on vt7 anyway?
I guess we should always force X/gdm to be on the active tty, regardless of any splash logic, and along with that just never start getty1 in runlevel 5.
I'm fine if we decide to always start X on tty1. But does that mean we should fix the Xorg code (since it currently starts on tty2 because tty1 is busy), or should we fix xdm, gdm, kdm & friends? Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Vincent Untz wrote:
I'm not sure what's the proper fix here: should init check there's nothing on tty2 before starting mingetty there? Or should we just force Xorg to start on vt7 anyway?
Keep X on vt7+. That's what everyone is used to. Also keep in mind bug 302010. cu Ludwig -- (o_ Ludwig Nussel //\ V_/_ http://www.suse.de/ SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nuernberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le mercredi 12 janvier 2011, à 11:14 +0100, Ludwig Nussel a écrit :
Vincent Untz wrote:
I'm not sure what's the proper fix here: should init check there's nothing on tty2 before starting mingetty there? Or should we just force Xorg to start on vt7 anyway?
Keep X on vt7+. That's what everyone is used to. Also keep in mind bug 302010.
So between what Kay suggests, what you suggest and what Xorg does by default, we have three different things. Can the relevant people (sysvinit maintainers, Kay for systemd, xorg maintainers) have a quick meeting on irc and agree on one direction? Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
participants (7)
-
Bernhard M. Wiedemann -
Josef Reidinger -
Kay Sievers -
Ludwig Nussel -
Stefan Seyfried -
Stephan Kulow -
Vincent Untz