/srv/irclogs.ubuntu.com/2009/10/16/#ubuntu-devel.txt

directhexanyone else have broken colorspace issues on video files using nvidia on karmic?00:32
chrisccoulsonhow broken? i know there is an issue with totem and nvidia users having to adjust the hue when they first run it00:33
TheMusodirecthex: I used to have weird colour issues when viewing flv/youtube videos in mplayer/totem/xine earlier in the cycle, although after a fresh install, that went away.00:40
directhexchrisccoulson, adjust it where? the weird thing is it's fine for the duration of having nvidia-settings running00:40
chrisccoulsondirecthex - that's wierd. i had to adjust the hue slider to the center in totem preferences when i did a fresh install00:41
directhexhm, not anymore. worked last week00:41
directhexwow, wait, what? chrisccoulson, well spotted! why is that randomly so low?00:42
chrisccoulsondirecthex - that's likely the same issue as bug 39547600:42
ubottuLaunchpad bug 395476 in nvidia-graphics-drivers-180 "nvidia sets HUE to -1000" [Undecided,Confirmed] https://launchpad.net/bugs/39547600:42
directhexlow priority?00:43
chrisccoulsonheh, i'm not sure who set that00:44
directhexYokoZar i guess00:44
LaserJockdailies up in ~ 1 hr?00:55
keesKeybuk: bug 452503 -- both "file" and older "blkid" worked correctly.01:00
ubottuLaunchpad bug 452503 in util-linux "blkid fails to recognize swap partition / file" [Undecided,Confirmed] https://launchpad.net/bugs/45250301:00
=== freeflyi1g is now known as freeflying
Keybukkees: did older vol_id work correctly?01:01
Keybukactually, the argument is that file is working *in*correctly01:01
keesor blkid should say "LUKS" but it doesn't01:01
Keybukblkid should say neither01:01
Keybukit has BOTH metadata01:01
keesthe main point is that it _is_ a valid swap partition.  the kernel can use it, etc01:01
Keybukit could be either, you can't pick one without risking corrupting the other01:02
Keybukyes, but blkid doesn't know it's *not* a valid LUKS partition01:02
Keybukthat's the point01:02
keesok, that's reasonable.01:02
keesit'd be nice to have a --verbose :)01:02
Keybukif blkid tells you it's swap, and you really meant LUKS, you just trashed your LUKS01:02
Keybukif blkid tells you it's LUKS, and you really meant swap, you just trashed your swap01:02
Keybuknow01:02
Keybukreplace LUKS and swap in those with01:02
Keybuksay01:02
Keybukext4 and vfat01:02
* kees nods01:02
Keybuk(which are the other common pairs)01:02
Keybukthat's why this is a Won't Fix01:02
keesalmost seems like a bug in mkswap then...01:02
Keybukindeed, I'm surprised mkswap didn't wipe01:03
Keybukthat is certainly a bug01:03
Keybukespecially given ff3bed806863d1c2075d0efda70b39ea6af9ecba01:03
Keybukwhich is the code in mkswap to explicitly wipe things out01:03
* kees changes the bug title's name01:04
keeser, bug's title01:04
AdamSchackartquick question, what VCS are you guys using?01:05
LaserJockAdamSchackart: all kinds but mostly bzr is what you'll find in Ubuntu01:06
KeybukAdamSchackart: bzr and git01:06
AdamSchackartcool01:06
mathiazKeybuk: hey - upstart job question01:08
mathiazKeybuk: http://people.canonical.com/~mathiaz/eucalyptus.conf01:08
mathiazKeybuk: ^^ will the opts variable be passed to the exec?01:08
Keybukmathiaz: no.01:08
mathiazKeybuk: ah!01:08
mathiazKeybuk: how should this be done?01:08
ionscript01:08
Keybukmdz has some weird things elsewhere to output that stuff to a file, then source the file in the exec line01:08
ion  opts=foo01:08
ion  exec $opts01:09
ionend script01:09
Keybukalternatively just put the whole lot of pre-start and exec into one script01:09
Keybukas ion just demo'd :p01:09
mathiazKeybuk: great - thanks01:09
mathiazion: thanks as well01:09
Keybukholy crap, I just read the moblin sreadahead pack01:10
Keybukand really wish I hadn't01:10
Keybukthis thing really is SSD optimised ;)01:10
Keybukfind / | make-pack01:10
ionHehe01:11
Keybukoh, I see01:11
Keybukthen that mincore()s it01:11
Keybukso it's not quite as crazy01:11
ionAh01:11
slangasekKeybuk: so if you call 'stop' in a script instead of a pre-start script, does it still work?01:12
=== TheMuso` is now known as TheMuso
slangasek(i.e., the only thing you don't want to do in a script is fork another process that's not the one to be supervised?)01:14
mathiazKeybuk: exec . /var/run/eucalyptus/eucalyptus.options && eucalyptus-cloud $opts01:17
mathiazKeybuk: would something like this ^^ work?01:17
mathiazKeybuk: (provided that the opts are dumped correctly in /var/run/eucalyptus/eucalyptus.options in the pre-start part of the job)01:17
chrisccoulsonis kees likely to be about this evening?01:18
keeschrisccoulson: he is! :)01:19
slangasekmathiaz: a) should be in a script instead; b) really would be better to not need the tmp file01:20
mathiazslangasek: ok01:20
chrisccoulsonkees - i'm confused - i don't see you on the list of users ;)01:21
slangasekmathiaz: I think the current script has a bug anyway, because calling "exit 0" from pre-start doesn't stop the job from running, which I think is what the goal was there?01:21
mathiazslangasek: should I exec in the script?01:21
chrisccoulsonmust be empathy playing up!01:21
slangasekmathiaz: yes01:21
chrisccoulsonkees - i have a patch for the screensaver issue. not sure if you want to take a quick look at it (or test it)?01:22
chrisccoulsonit seems to be working ok here01:22
keeschrisccoulson: sure!01:22
chrisccoulsonkees - the patch is here: http://pastebin.com/m239703a201:23
slangasekmathiaz: do you want the job to fail, or just not stop, if eucalyptus is not configured yet?01:23
mathiazslangasek: hm - what does "just not stop" mean?01:23
slangasekmathiaz: it means "just not start" :-)01:24
slangasekmathiaz: modulo the answer to the above question, I think what you really want is this: http://people.canonical.com/~vorlon/eucalyptus.conf01:24
=== TheMuso` is now known as TheMuso
mathiazslangasek: exit 1 will make the job fail?01:25
slangasekmathiaz: if you want the job to stop instead of failing, you can replace 'exit 1' with 'stop'; and then you can also mark the job as 'respawn'...01:25
slangasekmathiaz: yes01:25
mathiazslangasek: exit 0 will make the job not start ?01:25
slangasekno01:25
keeschrisccoulson: nice!01:25
slangasekmathiaz: exit 0 called from the pre-start script just says "the script is done, exited successfully" and the job will still start01:26
keeschrisccoulson: can you attach that to the bug, and I can get it built and get a freeze exception.01:26
mathiazslangasek: what does script+respawn do then?01:26
mathiazslangasek: it reruns the script until it's successfull?01:26
slangasekmathiaz: 'exec foo' is shorthand for 'script exec foo end script'01:27
=== norly is now known as ejat
slangasekmathiaz: for a normal job, the definition of "successful" is "running"01:27
slangasekmathiaz: "respawn" says "if the job dies for any reason, respawn it"01:28
chrisccoulsonkees - yeah, no worries. i can push it to ubuntu-desktop bzr if you're happy with the change01:28
mathiazslangasek: ok01:28
slangasekmathiaz: updated the script a little, to use the Keybuk-approved 'stop' runes01:28
keeschrisccoulson: if it works, yeah!  :)01:28
mathiazslangasek: { stop; exit 0; }01:29
mathiazslangasek: ^^ that will mark the job as stopped01:29
slangasekmathiaz: yes01:29
mathiazslangasek: if euca_conf is not there01:29
mathiazslangasek: however exit 1 will make the job fail if it's not configured01:30
slangasekmathiaz: right - hence, asking what you want the behavior to be :)01:30
mathiazslangasek: ok - just making sure I understand the semantics :)01:30
mathiazslangasek: hmmm... so what's the difference between a failed job and a stopped job?01:33
mathiazslangasek: the default configuration after package installation should be a working one01:33
slangasekmathiaz: whether or not 'respawn' will try to start it again :)01:33
slangasek(and what gets logged, etc)01:33
slangasekI would expect eucalyptus-cloud to be a normal service, where you want init to be the process supervisor and respawn it if needed01:33
mathiazslangasek: yes01:34
slangasekso I think you want to change that 'exit 1' to a 'stop; exit 0' and then add 'respawn'01:34
mathiazslangasek: even if there isn't any respawn option?01:34
mathiazslangasek: ok01:34
slangasekmathiaz: updated again, have a look there01:34
mathiazslangasek: what would 'exit 1' + no respawn option do?01:34
mathiazslangasek: in the case of a non-configured system01:35
slangasekmathiaz: result in the job being marked failed, with upstart not trying to restart it01:35
slangasekbut, if you want upstart to *ever* restart your job for you when it dies, you need 'respawn', and then you want to write your job to avoid upstart getting into a busy loop01:36
mathiazslangasek: busy loop meaning that the job would always fail and upstart would always restart it?01:37
ionUpstart will stop trying if it’s respawning too fast for a while.01:40
Keybukthe difference between exit 1 and { stop ; exit 0; } is, basically, in the return code ;)01:41
Keybuka job that just does exit 1 is marked failed01:41
Keybukthe event tells everybody that you failed01:42
Keybukand that it was your pre-start that failed01:42
Keybukthe user who runs "start" gets told it failed, and gets a "1" return code01:42
Keybuketc.01:42
Keybuk{ stop; exit 0; } means you're not marked failed01:42
Keybukyou just stopped again01:42
Keybukthe event says you stopped normally01:42
chrisccoulsonkees - i've pushed the change to bzr now. do you want me to subscribe u-m-s or will you handle this?01:42
keeschrisccoulson: I can handle that, thanks!01:43
Keybukthe user gets told (basically) that the job they wanted started, stopped again01:43
Keybuketc.01:43
chrisccoulsonkees - thank you too:)01:43
Keybukprobably the biggest difference is that you don't get a console message ;)01:43
mathiazKeybuk: so for the use case of: 'service not configured' what is the option to tell the user to first configuring the service before retrying to start?01:44
Keybukmathiaz: there's no method to do that right now01:45
mathiazKeybuk: failed job or start/stopped job?01:45
mathiazKeybuk: which are the existing options for now01:45
Keybukas slangasek says, we've settled on stop01:46
mathiazKeybuk: ok01:46
Keybukin the next version of Upstart, you won't need that at all01:46
slangasekmathiaz: are you assuming that the output of the pre-start script will be passed to the user calling 'start'?01:46
Keybukyou can just ship your package in manual mode01:46
Keybuk(then it won't be automatically started - but will be started if the sysadmin starts it by hand)01:47
ionKeybuk is planning to implement output logging for jobs. With it, it would be easy to print what the script said when e.g. the start command fails.01:47
mathiazslangasek: hm not really.01:47
slangasekmathiaz: ok good ;)01:47
mathiazslangasek: I'm just asking questions about how upstart works :D01:47
Keybukmathiaz: the problem is I tend to answer such questions with the question "how do you think it should work?"01:48
Keybukion: yes, but Keybuk is also planning to implement manual mode - which will do away with the need to say "aha! you didn't enable it"01:49
mathiazKeybuk: ah ok. Than I'd like to be able to tell the user to configure the service by editing this file before attempting to restart the job01:49
Keybukmathiaz: you can't do that01:49
LaserJockso ... how bad is it to do a Pre-Depends if you can't see a reasonable way without it?01:50
mathiazKeybuk: right - I understand that. I'll use the stop+exit0 for now01:50
Keybukpersonally I strongly dislike that kind of message anyway01:50
Keybuk"start foo" ... "but you didn't enable foo"01:50
Keybuk"yes, but I just told you to start it, stop being stupid and do what I told you"01:51
slangasekKeybuk: the use case here isn't /enabling/ foo, it's /configuring/ foo01:52
Keybukslangasek: well, that's different ;)01:52
mathiazKeybuk: so you don't consider that upstart could help the sysadmin in diagnosing wrong configuration?01:52
slangasekyes, it is01:52
slangasek:)01:52
mathiazKeybuk: a similar use case is to check for incompatible options01:52
Keybukin that case, you probably *do* want to exit 101:53
* Keybuk misunderstood, sorry01:53
mathiazKeybuk: right - that's a good use of failed job01:53
mathiazKeybuk: IIUC there isn't for now good logging?01:54
slangasekKeybuk: "respawn" only affects the main script, not pre-start, yes?01:54
Keybukslangasek: we don't respawn pre-start scripts ;)01:54
* slangasek nods01:54
mathiazKeybuk: like being able to say: your job has failed because foo option is incompitble with bar01:54
Keybukbut a pre-start script failed stops the job in a way that won't be respawned01:54
slangasekwho here knows how to grab translations out of rosetta for, e.g., ubiquity?01:55
mathiazslangasek: seems like your eucalyptus job is doing what it's supposed to then01:56
slangasekmathiaz: except Keybuk suggests 'exit 1' is better for the unconfigured case01:56
slangasekinstead of stop ; exit 001:56
mathiazhuh?01:56
LaserJockslangasek: mind if I upload this: http://launchpadlibrarian.net/33762653/moodle_1.9.4.dfsg-0ubuntu3.debdiff ?01:56
Keybukif it's unconfigured, or the configuration is wrong, that's a failure01:56
Keybukyou want that communicated01:57
Keybukit's not the same as "you didn't put splash on the kernel command-line" (which is a normal condition meaning the job is disabled)01:57
mathiazKeybuk: ok - gotcha now.01:57
slangasekLaserJock: <blink> what's the libc6 change that makes this necessary?01:59
slangasekLaserJock: oh, it's doing IPv601:59
LaserJockslangasek: yep02:00
LaserJockslangasek: you get a 403 when trying to configure moodle without that patch02:00
slangasekLaserJock: I think you probably want to list both values, then?  Otherwise, you'll also be denying connections to <hostname> on systems that have dynamic IPs, which is arguably a regression?02:00
LaserJockslangasek: ok, makes sense02:00
slangasek(since /etc/hosts is set up with 127.0.1.1 <hostname>, and users may expect that to work)02:00
mathiazkees: has eucalyptus 1.6~bzr931-0ubuntu4 been already uploaded?02:01
mathiazkees: the bzr branch still show UNRELEASED02:01
LaserJockslangasek: but good to go with that change?02:01
Keybukdamnit, fingers, remember how GRUB1 works02:01
slangasekLaserJock: I think so, yes02:02
=== lfaraone_ is now known as lfaraone
Keybukoh02:20
Keybukthat's neate02:20
Keybukecho 0 > /sys/devices/virtual/graphics/fbcon/cursor_blink02:21
Keybukslangasek: would you have any objection if we did something like02:27
Keybuktty1 - boot messages, then a getty02:27
Keybuktty2 - splash, then X02:27
Keybuktty3+ - more gettys, if configuration allows02:27
Keybuk(lynx obviously)02:29
AmaranthWhere do the other Xes go?02:30
Keybukthat's what I mean02:30
Keybukie. we hardcode something to own tty102:31
Keybukwe make usplash appear on tty2, then move it out of the way for X02:31
KeybukX will claim the first free tty (taking out the tty7 hardcode), which will be tty202:31
Keybukthen tty3+ will be gettys and Xs depending on configuration02:31
Keybukeach taking the first free02:31
Keybukif we don't boot with splash, you stay on tty1 until X chvts02:32
Keybukand then you see useful messages02:32
Amaranthah, neat02:32
Keybukif boot appears to hang while usplash is up Ctrl-Alt-F1 will take you to useful messages02:33
Keybuk(Alt-F1 even)02:33
KeybukCtrl-Alt-F1 will still work in X02:33
Keybukwe could use tty7 for splash/X but then we need to keep the ugly hack that makes it go there02:35
Keybukthe hack that I keep breaking by accident :p02:35
Keybukand then if you had only one getty configured, you'd have the weird thing where the second X login would be on tty2 ;P02:36
=== TheMuso` is now known as TheMuso
docroHI! I was wondering if anyone has noticed the default netcat behavior for waiting on STDIN has been changed. It appears that it indefintely waits on stdin, whereas the default use to be to terminated immediately upon EOF02:48
docrothis seems to be a result of the addition of the -q flag, but I can't be entirely sure02:49
Keybukdocro: I don't think that's changed02:52
Keybuknetcat has always been a bit weird about EOF02:52
docroI know at least in suse distros it termiantes immediately on stdin closing02:53
docroecho "foo" | netcat server.com02:53
docrowill close once foo is written02:53
docrolet me grab the version #s real quickl02:53
docroubuntu: [v1.10-38] ... SLES 11: [v1.10]02:55
docroSLES11 does not offer the -q option in its version of netcat02:55
Keybukdocro: suse could have patched it02:56
docroKeybuk are you running ubuntu right now?02:56
docroI doubt they would patch functionality OUT of netcat02:56
Keybukdocro: yes.02:56
docrowhat version? if not 9.04 can you test the behavior02:57
Keybuk9.1002:57
docrowhat version of netcat? netcat -h will show02:57
Keybukas you say, v1.10-3802:57
docrohmmm I guess I would need to troll through commits to netcat in both distro to figure it out02:58
Keybukthe documentation says without -q, it will not quit on EOF on STDIN02:58
docrofor now I have juas aliased netcat to netcat -q 002:58
docroKeybuk: the ubuntu documentation, yes02:58
Keybukdocro: just about every tool has minor differences between distros03:00
docroKeybuk: agreed, it's just causing me grief :) I was wondering if ubuntu ever worked the other way03:00
KeybukI don't think so, we inherit from Debian03:00
docronot closing on EOF is not really a minor difference03:01
Keybukaccording to the changelog, -q isn't even *in* the upstream netcat ;)03:01
Keybukit looks like netcat is one of those "maintained by patches" tools03:01
docroI have lots of code that depends on netcat closing upon EOF to gather results03:02
docrobecause that is how it works in red hat and suse03:02
Keybukit looks like there's netcat-bsd and netcat-traditional in Ubuntu as well03:02
docrohmm03:02
docroMaybe I will try those03:02
docrothanks :)03:02
Keybukdocro: and just as many people (more in fact, we have more users <g>) have code that depends on netcat not closing on EOF03:02
docroKeybuk: you have more desktop users maybe :P03:03
docroenterprise is dominated with RH and SUSE03:03
Keybukso? :)03:03
docrowho pays the bills :)03:03
Keybukthey don't pay our bills03:03
docronot yet hehehe03:04
docroThanks for your help though, I will try the traditional package. Otherwise I will just leave my alias in03:05
sladendocro: think OEMs, ODMs, SoC chip/CPU designers (... companies that actually have *serious* amounts of money)03:05
docrosladen i work for a company with serious amounts of money03:06
docronot that my nagging would make them pay up!03:06
sladendocro: I'd approach Canonical for a support contract :)03:06
docrosladen: currently one of the lone wolves moving towards desktop ubuntu for dev box and trying to get our tools working :) Doubtful our servers will ever go that way :(03:07
docroalright good night, thanks for the help again03:08
Keybukhe was pleasant03:10
Keybuknormally there's more depramming of toys03:10
LaserJockheh, depramming of toys, I like that03:11
mathiazkirkland: have you seen https://code.launchpad.net/~mathiaz/eucalyptus/karmic-fix-cloud-upstart-job/+merge/13452?03:39
mathiazkirkland: for https://bugs.launchpad.net/ubuntu/+source/eucalyptus/+bug/45266503:39
ubottuLaunchpad bug 452665 in eucalyptus "eucalyptus-cloud runs without any option set" [High,In progress]03:39
kirklandmathiaz: the upstart fixes?03:39
kirklandmathiaz: yeah03:39
kirklandmathiaz: why the respawn?03:39
mathiazkirkland: if eucalytpus-cloud dies, you wanna get it restarted no?03:41
kirklandmathiaz: i think so; just didn't see a note about that in the changelog03:41
kirklandmathiaz: ifconfig eth0:1 169.254.169.254 netmask 255.255.255.25503:43
kirklandmathiaz: that looks to be working for me03:43
mathiazkirkland: you may wanna call it differently03:44
mathiazkirkland: like eth0:eucalocal03:44
kirklandmathiaz: i couldn't get it to work with a string03:44
kirklandmathiaz: only with an int03:44
mathiazkirkland: avahi creates an eth0.avahi03:44
mathiazkirkland: IIRC03:44
mathiazkirkland: hm - may be you don't wanna have a virtual interface03:45
kirklandmathiaz: i couldn't get it to work with the ip command either03:45
kirklandmathiaz: had to use ifconfig03:45
sladenkirkland: ip ... label xyz03:46
kirklandsladen: hmmm03:48
kirklandsladen: sudo ip addr add 169.254.169.254/32 scope link dev eth0 label eth0:1 ?03:48
sladen ... dev eth0:myfoo label eth0:myfoo03:49
j1mcScottK: ping03:50
kirklandsladen: ubuntu@cluster:~$ sudo ip addr add 169.254.169.254/32 scope link dev eth0:9 label eth0:903:50
kirklandRTNETLINK answers: File exists03:50
sladenkirkland:03:50
sladenj1mc:03:50
mathiazkirkland: have you removed the existing 169.254.169.254/32 ?03:51
sladenkirkland: try adding an address that isn't in-use, eg   .169.25303:51
kirklandsladen: mathiaz; bingo03:51
kirklandmathiaz: would you like me to merge and upload your upstart changes?03:54
kirklandmathiaz: i have no objections03:54
mathiazkirkland: yes please03:54
slangasekKeybuk: I would object to us changing that in karmic ... :)03:55
kirklandmathiaz: how much longer are you online04:06
mathiazkirkland: don't know04:06
mathiazkirkland: why?04:06
mathiazkirkland: are you off tomorrow?04:07
kirklandmathiaz: i was thinking about it, but i didn't hear from alice04:07
kirklandmathiaz: i wanted you to review my eth device labeling04:07
mathiazkirkland: why - do you have a fix to test?04:07
kirklandmathiaz: i'm testing it now04:08
mathiazkirkland: sure - send a branch04:08
kirklandmathiaz: k04:08
kirklandmathiaz: the lintian output for eucalyptus is a mess04:09
kirklandmathiaz: brilliant!04:16
kirklandsladen: thanks for the pointer!04:16
mathiazkirkland: I know - thanks :)04:16
kirklandmathiaz: lp:~kirkland/eucalyptus/eth-label04:28
kirklandmathiaz: looks like it's working to me04:35
mathiazkirkland: I'll build a local version here and run through my tests04:36
Teddy_Hi there.  My package needs updating from Debian testing.  Is this the right place?04:38
kirklandkirkland: 1, scorpion: 004:51
kirklandhttp://rookery.canonical.com/~kirkland/CIMG0097.jpg04:52
kirklandback to eucalyptus04:52
mathiazkirkland: hm - build your branch04:53
mathiazkirkland: upgraded packages - still 169.X on eth004:53
kirklandmathiaz: i had to reboot04:53
kirklandmathiaz: restarting networking and eucalyptus might do it, though04:54
mathiazkirkland: does this mean that on shutdown eucalyptus-cloud doesn't remove its ip address?04:54
kirklandmathiaz: probably04:54
mathiazkirkland: well - the package upgrade should restart eucalyptus-cloud04:54
mathiazkirkland: well - I've just rebooted my CC04:55
kirklandmathiaz: i'm fixing it to handle the pub and priv addresses too, with a label04:56
kirklandmathiaz: eth0:pub and eth0:priv04:56
kirklandmathiaz: for the guest ip's04:56
kirklandmathiaz: which are also stacked on eth004:56
mathiazkirkland: hm04:56
mathiazkirkland: are you using multiple addresses?04:56
kirklandmathiaz: ?04:56
mathiazkirkland: there can be more than one public IP assigned to the CC04:57
mathiazkirkland: like .2 .3 .4 .504:57
mathiazkirkland: all assigned to the CC04:57
kirklandmathiaz: i'm not using an enumerator right now04:57
kirklandmathiaz: that's more complex04:57
kirklandmathiaz: mainly, i'm getting the superflous addresses off of eth004:57
mathiazkirkland: right - just saying - multiple public IPs are common04:57
kirklandmathiaz: and onto their own label04:57
mathiazkirkland: right - so eth0:pub would have multiple public IPS04:58
kirklandmathiaz: i'm just fixing the ones that eucalyptus creates for vm guests04:58
kirklandmathiaz: right04:58
mathiazkirkland: ok - if that works well that's great04:58
kirklandmathiaz: i'm pushing an update now04:59
kirklandmathiaz: did your CC come back up?04:59
mathiazkirkland: yop04:59
kirklandmathiaz: actually, pushing to lp:~kirkland/eucalyptus/eth-label204:59
kirklandmathiaz: and?  ifconfig says?  and ip addr says?04:59
mathiazkirkland: the eht0:metadata is there :)05:00
kirklandmathiaz: nice05:00
mathiazkirkland: let me check if instances are working correclty05:00
kirklandmathiaz: please05:00
mathiazkirkland: http://paste.ubuntu.com/294367/05:00
kirklandmathiaz: looks good05:01
kirklandmathiaz: and ifconfig?05:01
mathiazkirkland: well - there is still 169.X in there05:01
mathiazkirkland: ifconfig properly shows eth0 and eth0:metadata05:01
kirklandmathiaz: pastebin that05:01
mathiazkirkland: http://paste.ubuntu.com/294368/05:02
kirklandmathiaz: good05:02
kirklandmathiaz:           inet addr:192.168.12.118  Bcast:192.168.12.255  Mask:255.255.255.005:02
kirklandmathiaz: that's the key one05:02
mathiazkirkland: yes05:02
kirklandmathiaz: i'm please with that05:02
* mathiaz nods05:03
kirklandmathiaz: okay, branch lp:~kirkland/eucalyptus/eth-label205:03
kirklandmathiaz: sorry, had to push a new branch05:03
kirklandmathiaz: i uncommitted something, and things diverged05:03
mathiazkirkland: now eucalyptus-ipaddr.conf uses ip addr05:03
mathiazkirkland: ok05:03
kirklandmathiaz: yeah, i think it should use ifconfig05:03
kirklandmathiaz: but that's just my preference05:03
kirklandmathiaz: i'm not going to change it, but i would support someone else in doing so05:03
mathiazkirkland: so I was poking around the configuration and found something interesting in /etc/eucalyptus/eucalyptus.conf05:03
kirklandmathiaz: ?05:04
mathiazkirkland: VNET_PUBINTERFACE="eth0"05:04
mathiazVNET_PRIVINTERFACE="eth0"05:04
mathiazkirkland: seems that we already know which interface should be used for private /public interface05:04
kirklandmathiaz: i suspect that's intended for servers with multiple eth cards05:04
kirklandmathiaz: so you could put eth0 in one, and eth1 in the other05:04
mathiazkirkland: so we wouldn't need to do the ip route default thingy05:05
kirklandmathiaz: oh!05:05
mathiazkirkland: right - one of my setup is like that05:05
kirklandmathiaz: you mean to get the default interface05:05
mathiazkirkland: yes05:05
kirklandmathiaz: yeah, good idea05:05
mathiazkirkland: and in that setup the ip was wrong05:05
mathiazkirkland: ie the avahi_publish process was using the wrong ip05:06
mathiazkirkland: that being said things weren't working afterwards05:07
mathiazkirkland: the NC was using the wrong IP address to access walrus05:08
kirklandmathiaz: ip addr show label eth005:09
kirklandmathiaz: ;-)05:09
kirklandmathiaz: that drops the metadata labeled interfaces05:09
mathiazkirkland: nice05:09
mathiazkirkland: instances seem to be running correclty05:11
kirklandmathiaz: http://paste.ubuntu.com/294373/05:11
mathiazkirkland: http://paste.ubuntu.com/294374/05:12
mathiazkirkland: ^^ this is with a running instance05:12
mathiazkirkland: with a public IP05:13
kirklandmathiaz: right, that's without my latest patch05:13
mathiazkirkland: yop - I'm building your label2 branch05:13
=== troy_s1 is now known as troy_s
fabrice_spHi. I've found a problem with binutils libs: the soname does no match the shlibs file. Is it still possible to get it fixed for Karmic, or it has to go though SRU, and will be fixed after release?05:21
kirklandmathiaz: http://pastebin.ubuntu.com/294379/05:26
mathiazkirkland: 2x nice05:27
kirklandmathiaz: i'm quite happy with this05:27
kirklandmathiaz: perhaps would be better if there was an interator on :pub05:27
kirklandmathiaz: but, meh, this is easy, clean05:27
mathiazkirkland: did you try to boot two instances?05:27
mathiazkirkland: and see if two IPs can be added to the labeled interfaces?05:27
kirklandmathiaz: this is 2 instances05:28
kirklandmathiaz: see lines 49 and 5005:28
mathiazkirkland: ah right05:28
mathiazkirkland: I was looking at the ifconfig output05:29
kirklandmathiaz: and both can communicate05:29
kirklandmathiaz: right, that's what i mean by an iterator on :pub05:29
mathiazkirkland: which only supports one IP05:29
kirklandmathiaz: yeah05:29
kirklandmathiaz: we could add the last octet of the ip address to that05:29
kirklandmathiaz: so mine would be eth0:pub30 and eth0:pub3105:29
mathiazkirkland: can you log on both guests via their public IP address?05:29
kirklandmathiaz: yup05:29
mathiazkirkland: well - you can necessarly use the last octet only05:30
mathiazkirkland: what happens if you have a /16 network?05:30
kirklandmathiaz: eucalyptus only supports /24 networks05:31
kirklandmathiaz: dan told me that last week, while you were dozing in the car :-)05:31
mathiazkirkland: :o05:31
kirklandmathiaz: for the public ip's05:31
mathiazkirkland: hm well. I think for the time being the current solution is enough05:31
kirklandmathiaz: i agree05:32
mathiazkirkland: we'll have to get upstream opinion on this05:32
kirklandmathiaz: we could also add a random string05:32
mathiazkirkland: it's cleaner - is it worth putting in the release?05:32
kirklandmathiaz: i think it is05:32
kirklandmathiaz: i was going to roll all of this05:32
kirklandmathiaz: i could put it into a ppa right now05:32
kirklandmathiaz: until we get approval05:32
mathiazkirkland: let me do more testing of label205:32
kirklandmathiaz: agreed05:33
mathiazkirkland: I don't think putting in a PPA is helpfull at this tage05:33
mathiazkirkland: stage05:33
kirklandmathiaz: what more testing do you suggest?05:33
kirklandmathiaz: do you have it running yet?05:33
mathiazkirkland: well - it's more that noone is using a PPA05:33
mathiazkirkland: I'm upgrading05:33
kirklandmathiaz: eucalyptus has been using my ppa every day05:33
kirklandmathiaz: i meant ppa it for us + euca to test05:34
mathiazkirkland: oh ok.05:34
mathiazkirkland: I guess it could help then05:34
kirklandmathiaz: for the ~12 hours it would take to get an upload approved05:34
kirklandmathiaz: i think dan will gladly take this change05:35
kirklandmathiaz: it doesn't affect the actual networking, which was dan's concern05:35
mathiazkirkland: so IIUC the 169.X address is not deleted when eucalyptus is stopped?05:35
kirklandmathiaz: the diff only adds a label argument05:35
mathiazkirkland: right - which is a good thing05:36
kirklandmathiaz: that's probably true; and a bug05:36
mathiazkirkland: right - priv and pub addresses are removed correctly05:36
mathiazkirkland: what does happen to the :pub and :priv when there are no more instances running?05:37
kirklandmathiaz: let's find out!05:37
kirklandmathiaz: http://pastebin.ubuntu.com/294382/05:38
kirklandmathiaz: pub removed, priv not05:39
mathiazkirkland: right - probably make sense05:39
mathiazkirkland: the remaining address is the private gw IIRC05:40
kirklandmathiaz: ack05:40
kirklandmathiaz: did you see the fun you missed here tonight?  http://rookery.canonical.com/~kirkland/CIMG0097.jpg05:41
kirklandmathiaz: i had to take a 10 minute catch-the-scorpion-without-getting-stung-break05:41
mathiazkirkland: well - fun - like you life was at stake?05:41
mathiazkirkland: sounds like fun05:41
kirklandmathiaz: :-)05:41
mathiazkirkland: who won?05:42
kirklandmathiaz: i'll save him in a jar for you05:42
kirklandmathiaz: it was a very close battle05:42
kirklandmathiaz: but i bested him with my wit, and stunning looks05:42
mathiazkirkland: so you caught him alive?05:42
kirklandmathiaz: yessir.  in a jar now05:42
kirklandmathiaz: i'll save him for you05:42
mathiazkirkland: tya05:43
mathiazkirkland: will it survive 3/4 weeks?05:43
kirklandmathiaz: absolutely05:43
kirklandmathiaz: they can go a month without water05:43
kirklandmathiaz: i'll treat him better than that, though05:43
kirklandmathiaz: okay, what else should we do tonight?05:44
kirklandmathiaz: after the scorpion battle, i'm ready to turn in05:44
kirklandmathiaz: i think i'm going to commit this to the core-dev branch05:44
mathiazkirkland: right05:44
kirklandmathiaz: push to my ppa05:44
mathiazkirkland: I'd still ask upstream for their opinion05:44
kirklandmathiaz: email the relevant parties05:44
kirklandmathiaz: we need to get a release team member's approval to upload, right?05:45
mathiazkirkland: well - you can upload - they will review it anyway05:45
kirklandmathiaz: ah05:45
mathiazkirkland: it's actually easier for them05:45
mathiazkirkland: as they will see the diff directly05:46
mathiazkirkland: it just won't get in05:46
kirklandmathiaz: okay, should we wait to do that, until after we get upstream's opinion?05:46
mathiazkirkland: yes - I'd wait for upstream's opinion05:46
mathiazkirkland: if you push it to your PPA05:46
kirklandmathiaz: ack05:46
mathiazkirkland: upstream can easily test it and give feedback05:46
kirklandmathiaz: should i upload the rest, minus my network change?05:47
mathiazkirkland: if they're happy with it and accept the change, we can push into the archive05:47
kirklandmathiaz: okay, cool05:47
mathiazkirkland: I wouldn't do two uploads05:47
kirklandmathiaz: cool05:47
mathiazkirkland: I'll ask in the release meeting tomorrow morning05:47
mathiazkirkland: the diff is not that big05:48
kirklandmathiaz: cool05:49
kirklandmathiaz: i think i have it05:49
kirklandmathiaz: pushed to bzr+ssh://bazaar.launchpad.net/~ubuntu-core-dev/eucalyptus/ubuntu/05:50
kirklandmathiaz: uploading to PPA05:50
kirklandmathiaz: uploaded05:51
kirklandmathiaz: emailing upstream05:51
kirklandmathiaz: mail sent, you're on cc06:03
kirklandmathiaz: nurmi likes it :-)06:21
LaibschArneGoetje: The information you asked for in bug 338217 has been given.  What now?06:22
ubottuLaunchpad bug 338217 in scim-bridge "scim-bridge crashed with SIGSEGV in scim::Module::unload()" [Medium,Incomplete] https://launchpad.net/bugs/33821706:22
LaibschThat ticket is receiving dupes almost daily now.06:22
Laibsch1@all, can we please see a sync to fix bug 444883?06:33
Laibsch1bug 44488306:33
ubottuLaunchpad bug 444883 in whereami "please sync whereami 0.3.34-0.2 from Debian testing" [Undecided,New] https://launchpad.net/bugs/44488306:33
=== bjf is now known as bjf-afk
pittiGood morning06:47
ScottKGood morning pitti.06:47
pittihey ScottK, how are you ?06:47
ScottKTired.06:48
ScottKI should be asleep.06:48
ScottKpitti: Do you still have your Latitude D430?06:48
pittiScottK: yes, I do; you as well?06:48
ScottKYes.06:48
ScottKI've been having trouble the last couple of days with lid opening not being recognized.06:49
pittiweird06:49
pittiScottK: works quite well her06:49
pittie06:49
ScottKSo if I close the lid, it never notices it opens and so never turns the display back on.06:49
ScottKOK.06:49
ScottKI tried the previous intel driver, just to make sure and it's not that.06:50
pittiScottK: could you do the following: install input-utils, then "sudo lsinput" and find the event number of the lid switch, then do "sudo input-events N" with that number and close/open the lid?06:50
ScottKI guess I'd have to ssh into the machine06:50
liwI had trouble once with the open/close sensor being wonky (it was a pin that got pushed in to indicate close, and the spring that pushed it back out went bad and didn't push it out enough)06:50
ScottKBecause I won't have a working X06:50
ScottKIt's not hardware06:50
ScottKComing back from suspend on lid open works fine.06:51
Laibsch1I think ubuntu-laptop-mode should be removed: bug 45000406:52
ubottuLaunchpad bug 450004 in ubuntu-laptop-mode "time for removal of ubuntu-laptop-mode package?" [Undecided,New] https://launchpad.net/bugs/45000406:52
=== Laibsch1 is now known as Laibsch
ArneGoetjeLaibsch: I can't fix the bug. In related (if not the same) bug #199592 Zhengpeng Hou has asked for testing his PPA package, so far no reply on that. As I cannot reproduce this issue at all anyways I'm afraid I'm unable to help on this issue. From the comments it seems to appear at random for some users.06:54
ubottuLaunchpad bug 199592 in scim-bridge "scim-bridge crashed with SIGSEGV in scim::Module::unload()" [Medium,Confirmed] https://launchpad.net/bugs/19959206:54
LaibschOK06:54
LaibschWhy did you ask for removing the config files and reinstallation of the package?06:55
ArneGoetjeLaibsch: sometimes old config data can cause such things... it happened in the past.06:55
LaibschCan you read stack traces?06:59
ArneGoetjeLaibsch: no. I'm not a programmer.06:59
LaibschIf those two are indeed dupes with a certain degree of certainty, it would be nice to mark them as such06:59
LaibschOK06:59
LaibschBummer06:59
LaibschNeither am I06:59
ScottKpitti: I'm going to wait until my kdelibs build finishes to close the lid (I have a theory about the Kubuntu netbook RC bug you assigned to Riddell today), so that'll probably be after I sleep.07:02
dholbachgood morning07:23
=== tkamppeter_ is now known as tkamppeter
LaibschArneGoetje: I'll ask for help in #ubuntu-bugs with reading the stacktraces07:28
mvohey dholbach07:29
dholbachhi mvo07:29
ArneGoetjeLaibsch: thanks07:29
=== YDdraigGoch is now known as Richie
slangasekKeybuk: no bug numbers in this usplash upload?  What's broken if we don't do this?07:47
=== TheMuso` is now known as TheMuso
=== Laibsch1 is now known as Laibsch
=== dholbach_ is now known as dholbach
Laibschslangasek: Are you aware that the latest change to the emacs package now forces the installation of emacs22 again?09:12
LaibschFrom what I can see it's a step back, not a step forward09:12
Laibschtowards the migration to emacs2309:13
pittiLaibsch: emacs22 is in main, 23 is in universe, so the default should be 2209:16
pittiit wasn't considered for main early enough before FF09:16
Laibschic09:16
dholbachlool, paulliu: what happened to mutter-moblin now?09:20
paulliudholbach: Can I upload the new package to Debian and then request sync again?09:24
paulliudholbach: Just fix the things you mentioned yesterday.09:24
dholbachsure, but the earlier the better09:24
paulliudholbach: OK. I do it now.09:24
dholbachrock on09:24
dpmhi slangasek09:26
dpmre: bug 447383, and since the string freeze break was granted after NonLanguagePackDeadline, thus not leaving translators time to do their work, would it be possible to have an exception on ubiquity-slideshow-ubuntu so that translations are exported by LanguagePackDeadline on the 22nd and still used in the LiveCD?09:26
ubottuLaunchpad bug 447383 in ubiquity-slideshow-ubuntu "UI freeze exception for ubiquity-slideshow-ubuntu 9 (Ubuntu One and FF 3.5)" [Undecided,Fix released] https://launchpad.net/bugs/44738309:26
dpmIf so, how could I request this exception?09:27
pittidpm: we can't upload it _on_ the 22nd, since that's the day when we release RC09:30
pittidpm: but we can do/accept an upload next Friday, after RC, so that they'll be in final09:30
=== cking_ is now known as cking
dpmpitti, that would be perfect. How do I go about requesting this?09:33
pittidpm: best would be to open a bug against it and subscribe ubuntu-release, then nominate it for karmic09:33
pittidpm: please point me at it, then I'll approve/milestone09:34
pittidpm: then it'll be on the release radar and we won't forget about it09:34
dpmpitti, ok, I'll do that now, thanks a lot!09:34
pittiI'm off for ~ 2 hours for a doctor appt.09:38
cjwatsonkees: re that mkswap/luks thing you were discussing with Keybuk, I think it would be worth filing a bug on busybox; I don't think its mkswap clears out the bootbits09:48
cjwatsonkees: (but unfortunately I also don't think it's safe to try to do that for release; util-linux mkswap has several checks in there that aren't trivial to implement in busybox)09:48
dpmpitti, (when you're back), here it is -> bug 452889, please feel free to approve/milestone if appropriate10:12
ubottuLaunchpad bug 452889 in ubiquity-slideshow-ubuntu "NonLanguagePackTranslationDeadline exception request" [Undecided,New] https://launchpad.net/bugs/45288910:12
geserpitti: do you have time to sponsor the diff in http://paste.ubuntu.com/294525/? it fixes the upload failure for your openthesaurus upload10:12
* ogra is looking at the gltk FTBFS ... if i refresh the .symbols file with the 1.1.9 symbols added, it works error free with fakeroot debian/rules binary ... as soon as i build it in pbuilder it fails with differences in the .symbols file ... 10:17
ogra*the fltk FTBFS10:17
mdzmathiaz, what's wrong with the script as written? the weird temporary file stuff that Keybuk was referring to was fixed ages ago, and it now has the logic in a start script as he suggested10:41
mdz(the eucalyptus upstart job)10:41
pittidpm: done11:16
pittigeser: oh, that was still on my list, thanks!11:16
pittigeser: any idea why it does that hackery at all?11:17
geserpitti: no idea11:18
=== elmo_ is now known as elmo
=== agateau is now known as agateau|lunch
=== MacSlow is now known as MacSlow|lunch
=== agateau|lunch is now known as agateau
=== pedro__ is now known as pedro_
=== MacSlow|lunch is now known as MacSlow
=== marjomercado is now known as marjo
jdstrandmdeslaur: is 'Low' the right priority for bug #446524? as I read it, aa-logprof is busted for everything. is that an incorrect assessment?14:01
ubottuLaunchpad bug 446524 in apparmor "aa-logprof: doesn't parse new null profile syntax" [Low,In progress] https://launchpad.net/bugs/44652414:01
mdeslaurjdstrand: no, I'll change it to high14:04
jdstrandmdeslaur: can you add a regression-potential tag as well?14:05
mdeslaurjdstrand: If I can ever figure out how to add a tag, sure!14:06
mdeslaurjdstrand: oh, found it14:06
mdeslaurd'uh14:06
jdstrand:)14:06
ograpitti, what exactly does devicekit disks do with usb disks when i fire up gnome-session14:07
ograpitti, trying to debug bug 431963 i found a weird issue ... it doesnt boot into graphical mode at all. i can set text on the cmdline and properly boot into a text console though ... creating ~/.xsession that only fires up xterm i can even startx ...14:09
ubottuLaunchpad bug 431963 in linux-fsl-imx51 "io/fs errors when launching gdm on imx51 with sata" [High,Confirmed] https://launchpad.net/bugs/43196314:09
ograas soon as i fire up gnome-session from that xterm, the system hangs and i see the following in dmesg ... http://paste.ubuntu.com/294674/14:10
ograi suspect one of the gdu monitors or devicekit-disks to be the issue here, the disk is sitting on a special SATA->USB controller that lives on the board14:11
=== robbiew_ is now known as robbiew
Chipzz~..14:24
mnemonikkChipzz: flaky ssh connection? 8-)14:28
pitticjwatson: do you happen to know if usplash uses bogl insead of svgalib when running under KMS? (if you don't know, nevermind)14:31
cjwatsonyes, it does14:31
pittiaaaah14:31
cjwatsonthis is a feature :)14:31
pittiok, then I know why bug 448988 happens14:31
ubottuLaunchpad bug 448988 in usplash "--pulse-logo doesn't actually pulse the logo" [Medium,Triaged] https://launchpad.net/bugs/44898814:31
pittiand usplash sets an entirely black palette14:32
pitticjwatson: thanks14:32
cjwatsonbeware, Scott's last upload doesn't seem to be in bzr ...14:32
ionmnemonikk: Whenever i have to do ~.enter and irssi is active, i try to create a new screen window first before typing that, in case the input actually reaches the box. :-)14:32
ionenter~., that is.14:32
* cjwatson tends to hit C-u first14:32
pitticjwatson: I know, Steve had a question about it; thanks for the warning14:33
pittiogra: highvoltage14:33
pittioops, meant to say "hi"14:33
ogra?14:33
ograheh14:33
mnemonikkion: 8;-)14:33
* ogra feels shocked14:33
cjwatsonDANGER DANGER HIGH VOLTAGE14:33
ograhehe14:33
pittibzzzzzt14:33
* cjwatson puts on his stovepipe hat14:34
pittiogra: dk-disks itself doesn't do anything by itself, but vfs currently has a bug that it (sometimes) automounts all internal hard disks at startup14:34
* sistpoty|work tried tab-completion for his password once :P14:34
pittiogra: bug 45161314:34
ubottuLaunchpad bug 451613 in gvfs "all disks and partitions are automounted in the live session" [High,In progress] https://launchpad.net/bugs/45161314:34
cjwatsonoh wait, that was a different Electric Six video :)14:35
ograpitti, hmm, any way to disable that to prove it's actually at fault ?14:35
Chipzzmnemonikk: flakey wireess connection :S14:35
Chipzzthink someone is trying to hack my wireless14:36
ograpitti, mounting shouldnt trigger a device reset though14:36
paulliudholbach: I've pushed new mutter-moblin into sid few hours ago. It still needs some time to appeared for requestsync.14:40
pittiogra: uninstall dk-disks temporarily? or move aside devkit-disks-daemon14:43
dholbachpaulliu: just file the bug manually then?14:43
dholbachif necessary we can sync from incoming14:43
paulliudholbach: ok. I'll do that now.14:43
paulliudholbach: sorry. Which package/target should I report to?14:44
dholbachubuntu14:44
paulliuok14:44
dholbachhttps://launchpad.net/ubuntu/+filebug?no-redirect I think14:44
paulliudholbach: thanks.14:44
ograpitti, will try ...14:44
ograi removed everything in /etc/xdg/autostart already but that doesnt seem to help14:45
=== jamie is now known as Guest6805
paulliudholbach: oops. Launchpad timeout error.14:47
sivanganybody know where the group photo from UDS paris is ?14:47
* sivang admits a rather odd question at this time of year14:47
dholbachpaulliu: without edge. maybe?14:47
paulliudholbach: I disabled beta tester for 2 hours but still Timeout.14:52
paulliudholbach: strange.14:52
sivanghmm, close to release time, I'm being ignored, granted :)14:53
=== greg_g is now known as greg-g
=== bjf-afk is now known as bjf
ograseb128, is there any code in gnome-session that could trigger something like that: http://paste.ubuntu.com/294674/15:14
ograi have cut down all preipherial apps to nearly zero, starting gnome-session or gdm (which spawns gnome-session) makes the root disk disappear15:15
seb128not that I know about no15:16
seb128gnome-session is mainly starting other things15:16
ograif i use ~/.xsession, install xfce and put startxfce4 in there it works fine with startx15:16
ograi have removed everything from /etc/xdg/autostart15:16
ionThose look like USB problems.15:16
seb128it's probably devicekit-disks15:16
ograi wouldnt know what else to remove15:16
ograion, its a SATA drive :P15:17
seb128ther are quite some things dbus activated15:17
seb128ie gvfs, devicekit-disks15:17
seb128you can move the binaries to try15:17
ograhmm, how do i disable them15:17
seb128move the binaries away15:17
ograi removed devkit-disks-daemon, that didnt help yet15:17
seb128or dbus .service15:18
seb128dunno then15:18
LaserJockcjwatson: do you think you'll have time to look at my cdimage merge request today?15:19
james_wNg: I think there's a chance that you could get the latest bash-completion synced from Debian. You'd have to talk to a release manager though15:21
Ngjames_w: I've had it installed since lunch and I'm pretty much on the verge of retracting what I said. It's nice not having the regression, but the new avahi stuff is driving me crazy ;)15:22
Ngavahi-browse takes 5 seconds to complete15:22
james_wheh15:22
cjwatsonLaserJock: I don't know, I'll try15:24
LaserJockcjwatson: k, thanks, I know you're busy15:25
ccheneymvo: ping15:38
mvoccheney: pong15:47
keescjwatson: yeah, the checks util-linux's mkswap does depend on the libraries, etc.  are we using the busybox mkswap for installs?16:01
keeskirkland: oooh, I'd never seen "label" from ip before.  nice.16:03
kirklandkees: it was new to me too16:04
kirklandkees: sladen showed it to me16:04
pittielmo, james_w: btw, bug 395302 is another report of this16:04
ubottuLaunchpad bug 395302 in gdm "kills X session during upgrade" [Critical,Fix released] https://launchpad.net/bugs/39530216:04
pitti(at the bottom)16:04
cjwatsonkees: yes16:05
cjwatsonalways have16:05
keescjwatson: oh, I thought the liveCD and alt have util-linux?16:07
cjwatsonkees: live CD uses util-linux mkswap16:08
cjwatsonkees: alternate CD *installs* util-linux, but doesn't have it in the installation environment16:08
keesah-ha.16:08
mathiazmdz: re eucalyptus job start: the pre-start script was setting up the correct opts variable - however eucalyptus-cloud was using an exec statement with an opts variable16:27
mathiazmdz: and variables are *not* propagated from the pre-start block to the exec block16:28
mathiazmdz: the fix was to move all the code that deals with updating the opts variable within a script statement instead of an exec statement16:30
=== The_Company is now known as Company
ccheneydoko: ping16:42
ccheneydoko: are you doing that new binutils upload?16:42
=== The_Company is now known as Company
bdrungTheMuso, dtchen: is bug #453158 a alsa or pulseaudio bug?16:43
kirklandpitti: looks like richard hughes has a fix for the encrypted-swap/hibernate problem16:43
ubottuLaunchpad bug 453158 in xmms2 "problem with xmms2 toggleplay" [Undecided,New] https://launchpad.net/bugs/45315816:43
kirklandpitti: you think this is SRU-able?16:43
pittikirkland: I saw the bug mail16:43
ccheneyslangasek: or do you happen to know if he is doing it? I'm trying to determine when I should do the OOo upload16:44
pittikirkland: looked quite complex, so I wouldn't like to upload it now; but if we tested it a bit in lucid, we can consider that16:44
kirklandpitti: okay, cool16:44
kirklandpitti: i'll leave this with you16:44
kirklandpitti: cheers16:44
pittithanks!16:44
kirklandpitti: there's a dozen or so bugs in launchpad related to this16:45
kirklandpitti: filed against linux, hal, acpi, gnome, cryptsetup, pm-utils16:45
mathiazwhich component is responsible for the fsck on boot?16:46
mathiazI've run into a fsck check this morning but couldn't cancel it16:46
jcastrohttp://www.fewt.com/2009/10/i-give-up.html <-- anyone know anything about this?16:47
pittimathiaz: known bug, Keybuk has a branch ready16:47
jcastrospecifically the part about being ignored by ubuntu developers?16:47
mathiazpitti: great thanks16:47
dokoccheney: will reply on email. are all planned changes now really included in the upload? ,P16:47
ccheneydoko: the ppc fix and the kde4 fixes will be in, not sure if there were any other planned ones?16:49
ccheneythe kde4 fixes didn't even exist when i did the previous upload16:50
=== beuno is now known as beuno-lunch
xan_Hi, sorry for the inconvenience but I can't get information via web. Anyone know if it's planning to improve the UNR with the "speed" of the ubuntu moblin remix?17:02
=== cdE|Woozy_ is now known as cdE|Woozy
=== The_Company is now known as Company
ccheneydoko: ok will get the upload staged for you today17:18
EvanCarrollThis very big perl-xml-sax bug is set to plague the jaunty upgrade17:18
EvanCarrollhttps://bugs.launchpad.net/ubuntu/+source/libxml-sax-perl/+bug/1391717:19
ubottuLaunchpad bug 13917 in libxml-sax-perl "CPAN's XML::SAX update conflicts with libxml-sax-perl's XML::SAX" [Medium,Confirmed]17:19
dokoccheney: thanks17:19
EvanCarrolllike 20 dupes of the bug now, with 2 of them being posted since karmic beta17:20
EvanCarrollI wish i could change the importance17:20
EvanCarrollI talked to the debian guys about the bug and the changes need to fix it, they're pretty perl retarded though17:21
EvanCarrollI'd say the best of them has 2 credit hours and no real life experience with the language.17:21
james_wEvanCarroll: insulting our colleagues isn't likely to be a good strategy to get us to help you17:22
james_wplease try and state your concern without the unnecessary editorialising17:23
pittijames_w, mvo: with u-m switched back to synaptics, is bug 445303 still an issue for karmic then?17:23
ubottuLaunchpad bug 445303 in update-manager "update-manager stucked on polkit password dialog" [High,Invalid] https://launchpad.net/bugs/44530317:23
james_wpitti: it's seen in other polkit using things, so I would say yes17:24
EvanCarrollI'm not needing help -- this was more of an informative message if anyone cares to do some work on the issue moving away from debian. They simply do not undersatnd perl, I've written letters filed bugs, and fixed both launchpad and their tracker marking dupes of this issue as appropriate -- they don't get it.17:24
pittijames_w: but the crashes in NM for example were fixed in today's uploada17:24
EvanCarrollI'm thinking about requesting permissions from XML::SAX author to send a request to cease and decist.17:24
pittijames_w: right now I'm a bit puzzled what to do with this bug, if it's not even reproducible in karmic any more? (since u-m switched backend)17:25
james_wpitti: I don't think they are the same bug are they?17:25
pittijames_w: not originally, I just asked to test the new PK again because 445303 is obviously also a crash17:25
pittid-bus timeouts usually mean that the backend has crashes17:25
EvanCarrollthey hack code into the module in an unsupported fashion, without subclassing, and it litterally breaks everything when you upgrade that module outside of dpkg.17:25
pittis/crashes/crashed/17:25
james_wpitti: oh, I see17:26
EvanCarrolltheir answer is, "CPAN is not supported on debian systems"17:26
EvanCarrollwhich is laughable.17:26
ScottKEvanCarroll: It's not.17:26
pittijames_w: it was worth a shot, anyway17:26
james_wEvanCarroll: please desist17:26
ScottKWe have a package management system and people should use it.17:26
james_wpitti: I suspect that fix will not have made this go away, but it would be good to know.17:27
cjwatsonwe should do a better job of it, and I don't think CPAN should be declared as unsupported since it is very useful for admins using Perl17:27
james_wpitti: as for tracking the bug, I suggest we still do for software-center etc.17:27
EvanCarrollNothing should /have/ to use it for modules, Firefox (add-ons) doesn't use it, ruby (gems), and even python (pypi), php (pear) don't have to use it -- why should perl.17:27
cjwatsonbut I don't see the point of discussing it here - I believe I commented on the patronising attitude last time too!17:28
james_wjdong: did you have a chance to test the policykit changes I made?17:28
pittijames_w: ah, right; that uses aptdaemon, too17:28
cjwatsonit needs to be fixed in Debian17:28
jdongjames_w: no; sorry, got caught up in back-to-back midterms...17:43
james_wjdong: no problem17:44
james_wjdong: don't worry if you have more important things to be doing17:44
jdongI'll try to make some time tonight to get that info back to you17:46
jdongjames_w: haha admittedly I wasted some time earlier in the week converting to a btrfs root... ended in tragedy ;-)17:47
james_wjdong: sounds like a sensible thing to be doing ;-)17:47
jdong:)17:47
* jdong views "EXPERIMENTAL" as an empty threat ;-)17:48
james_wheh17:48
ScottKpitti: I got the event information for lid close/open on my D430 that we discussed early this morning.17:48
ScottKhttp://paste.ubuntu.com/294833/17:50
=== zul_ is now known as zul
jcolefyi guys, seems to be another problem with flash on 64 bit firefox again... reddit users are starting to complain (they suggest to just get the 64 bits from adobe directly) -> http://www.reddit.com/r/linux/comments/9tr3d/the_ubuntu_flash_flu/18:12
jcoleive also got my own strange "white box" problem happening lately18:13
JanCjcole: AFAIK Adobe forbids Canonical to distribute the 64-bit alpha plugin...  :-(18:20
jcoleJanC: maybe "someone" should throw some "support" towards the swfdec developers ;)18:24
JanCboth swfdec & gnash work for some sites and not for others, maybe they should start working together first...18:25
=== RainCT_ is now known as RainCT
=== beuno-lunch is now known as beuno
keeschrisccoulson: so, I'm trying to force the g-s-s crash by adding delays, etc, so I can "prove" that your patch fixes the problem (also so I can test jaunty and earlier).  I'm not having much luck.19:10
chrisccoulsonkees - i tried to force it by adding delays too, but i don't think delays make it any worse19:11
chrisccoulsonyou just need to be unlucky with the timing19:11
chrisccoulsoni've not had a look at older versions of g-s-s yet19:11
keeschrisccoulson: but where is the actual problem...19:11
keeschrisccoulson: I'm trying to figure out at which line of code having the dialog vanish causes the problem.19:12
keeschrisccoulson: the shake_dialog even checks window->priv->lock_box before the alignment_set_padding call..19:13
keeschrisccoulson: I assume it's during the gtk_main_iteration(), but it's not clear what19:13
keesshake_dialog is called via lock_command_watch.19:14
chrisccoulsonkees - yeah, it will occur once gtk_main_iteration is called. gtk_alignment_set_padding registers some sources on the event loop19:14
chrisccoulsonand then those events are dispatched with gtk_main_iteration19:14
keesso at what point are those sources unmapped?19:15
chrisccoulsonthe issue will occur if the dialog exits after dispatching an event that accesses the resource at some point19:15
keeslock_command_watch is called on pipe input19:15
chrisccoulsonbecause once the event has been dispatched, the main process cannot notice the dialog has disappeared until execution has returned back to the main loop19:16
keeschrisccoulson: right, but doesn't that mean everything stays allocated?  I'm not clear what's tearing down the allocation.19:16
chrisccoulsoni'm not 100% sure how XEmbed works, but i'm fairly sure it is the gnome-screensaver-dialog process which allocated the resources on the server19:17
keeschrisccoulson: so, some event happens between gtk_alignment_set_padding and gtk_main_iteration in shake_dialog that causes gtk_container_idle_sizer to die?19:19
chrisccoulsonkees - yeah, that's likely. it could also happen at any point in the while loop too19:21
* kees tries adding more delays...19:23
jdstrandkirkland: hi! can you explain to me the issue with pulseaudio and qemu-kvm?19:32
kirklandjdstrand: "the issue"?19:32
jdstrandkirkland: yes-- debian/patches/10_qemu-allow-pulseaudio-to-be-the-default.patch19:33
kirklandjdstrand: i reverted that one19:33
kirklandjdstrand: i had cherry-picked it from fedora, thinking that might be a good thing19:34
jdstrandkirkland: something is not right. specifically, on a brand new vm I hit bug #45332919:34
ubottuLaunchpad bug 453329 in libvirt "libvirt apparmor profile denies access to pulseaudio" [Medium,Triaged] https://launchpad.net/bugs/45332919:34
kirklandjdstrand: the initial bug reporter just wanted kvm to "prefer" pa over oss and alsa19:34
jdstrandkirkland: (fine). so I go to fix the apparmor profile, and now when I start the vm, it hangs19:35
jdstrandkirkland: in other words, now that the apparmor profile isn't denying access to pulseaudio, the vm won't start19:35
chrisccoulsonkees - you might have more luck recreating the issue by ignoring the "destroy" and "plug_removed" signals from the GtkSocket in gs-window-x11.c, but i can't predict what other wierd effects that would have19:35
chrisccoulsonperhaps delay executing the callbacks?19:35
chrisccoulsonyou could maybe use those signals to trigger a callback which just registers the real callback to be executed after a delay with g_timeout_add19:37
chrisccoulsoni need to go for dinner shortly, but i may try that afterwards19:38
keeschrisccoulson: ok, cool.19:38
jdstrandkirkland: let me rephrase-- it starts just fine, but during boot it hangs19:38
kirklandjdstrand: hmm, i'm not sure what to say ...  i added that patch, and then reverted it19:38
kirklandjdstrand: guest kernel hangs?19:39
jdstrandkirkland: I don't know19:39
kirklandslangasek: i just uploaded eucalyptus for exception approval; please let me know if you need anything of me19:39
jdstrandkirkland: but if I do:19:39
jdstrandecho autospawn = no|tee -a ~/.pulse/client.conf19:39
kirklandjdstrand: okay, let me install a new deskto vm19:39
jdstrandkillall pulseaudio19:39
jdstrandactually, strike that19:40
jdstrandkirkland: if I remove the sound device from the libvirt xml, it starts fine19:40
jdstrandkirkland: it showed up by default as an es1370 iirc19:41
jdstrandkirkland: yeah, es137019:41
jdstrandkirkland: you won't reproduce it without a patch to the apparmor profile19:41
jcolebig windows 7 releases and promotions happening all over this weekend and next few weeks... so, hold onto your hats... apparently, windows 7 can cure cancer and create world peace19:42
* jdstrand goes to get it19:42
kirklandjdstrand: ?19:42
kirklandjdstrand: you're seeing this on a package that's not in karmic right now?19:42
jdstrandkirkland: no19:42
jdstrandkirkland: like I said, the apparmor profile denies access to pulseaudio19:43
jdstrandkirkland: that is a bug I was trying to fix19:43
kirklandjdstrand: oh19:43
kirklandjdstrand: okay, and in trying to fix that, you're getting a hanging vm19:43
jdstrandkirkland: when I fixed the bug (ie, kvm has access to pa), then I saw this issue19:43
jdstrandkirkland: simply put. pulseaudio doesn't work with libvirt/kvm due to apparmor. however, when I fix that, you will find that kvm and pulseaudio aren't getting along19:44
jdstrandkirkland: http://paste.ubuntu.com/294893/19:45
kirklandjdstrand: http://cvs.fedoraproject.org/viewvc/F-12/qemu/qemu-allow-pulseaudio-to-be-the-default.patch?revision=1.3&view=markup19:45
kirklandjdstrand: that *was* the patch I had cherry picked19:45
kirklandjdstrand: i periodically do a sweep over fedora's patches, and pick ones that look tasty19:45
kirklandjdstrand: i thought that one looked good, but as soon as I applied it and uploaded, i got a complaint from a user whose bug I had previously fixed, and he said my upload regressed his fixed19:46
kirklandjdstrand: tbh, I didn't really dig too deeply into his issue19:46
jdstrandkirkland: apply my diff to /etc/apparmor.d/abstractions/libvirt-qemu, then try to install karmic off the livecd in a new vm19:46
jdstrandkirkland: using virt-manager19:47
jdstrandkirkland: assuming pulseaudio is running on your machine, the vm will hang somewhere during boot19:47
kirklandjdstrand: what's your command line say?19:50
kirklandjdstrand: -soundhw ?19:50
jdstrandkirkland: -soundhw es137019:51
jdstrandkirkland: http://paste.ubuntu.com/294897/19:51
kirklandjdstrand: hmm19:53
jdstrandkirkland: are you seeing it with straight kvm?19:53
kirklandjdstrand: i was able to get sound okay, with straight kvm19:57
kirklandjdstrand: i'm syncing the daily desktop iso now19:57
kirklandjdstrand: i should be able to reproduce it with just that, right?19:57
jdstrandkirkland: well, my desktop-iso was old. I used virt-manager with generic/generic, kvm/i686 and an i386 desktop livecd19:58
=== asac_ is now known as asac
kirklandjdstrand: hmm...  there could be any one of a number of kernel problems with an old iso ....19:59
kirklandjdstrand: i'm rsyncing now19:59
jdstrandkirkland: related to sound specifically? like I said, if I don't use -soundhw es1370 it works fine20:00
=== robbiew is now known as robbiew_
kirklandjdstrand: so no -soundhw at all, or a different -soundhw ?20:00
kirklandjdstrand: when it works ^20:00
jdstrandkirkland: no -soundhw at all (I removed the sound device from the xml)20:01
kirklandjdstrand: and sound works?20:01
kirklandjdstrand: or it's just able to boot?20:02
jdstrandkirkland: no. the vm boots to the desktop20:02
* kirkland would not expect sound to work20:02
jdstrandkirkland: with -soundhw es1370 it hangs during the boot20:02
kirklandjdstrand: okay ....20:02
kirklandjdstrand: i'm applying your patch20:02
kirklandjdstrand: actually, i'm going to first boot today's desktop iso with sound, without your patch20:04
kirkland/usr/bin/kvm -m 512 -smp 2 -usb -usbdevice tablet -net nic,model=virtio -net tap,script=/home/kirkland/bin/bridge.sh -soundhw es1370 -cdrom karmic-desktop-amd64.iso20:04
jdstrandkirkland: you should then see apparmor denied messages in kern.log (or /var/log/autit/audit.log if you use auditd)20:04
kirklandjdstrand: okay, here's what I have ...20:05
kirklandjdstrand: my desktop hardware is running pulse audio20:05
kirklandjdstrand: i have not applied your patch yet20:05
kirklandjdstrand: i booted today's desktop iso using bare qemu-kvm20:06
kirklandjdstrand: with the command line above (with es1370)20:06
kirklandjdstrand: i did get the startup sound, though it was a little garbled20:06
kirklandjdstrand: okay, so i assume applying your patch won't have any affect, unless i run through libvirt, right?20:07
kirklandjdstrand: i'll try virt-manager20:07
jdstrandkirkland: the apparmor protection does not apply to kvm on its own, no. you must use libvirt20:08
jdstrandkirkland: use generic/generic with kvm/i686 (that is what I did)20:08
jdstrandI'm rsyncing the image now too20:08
kirklandjdstrand: okay, i'm using amd64, ubuntu/karmic20:09
kirklandjdstrand: and i'm setting a placebo using virt-manager, without your patch20:09
kirklandjdstrand: will apply it in a moment20:09
jdstrandI too use amd64/karmic20:09
kirklandjdstrand: http://pastebin.ubuntu.com/294909/20:10
kirklandjdstrand: okay, no startup sound20:10
jdstrandkirkland: what version of libvirt do you have?20:10
kirklandii  libvirt-bin                              0.7.0-1ubuntu11                            the programs for the libvirt library20:10
kirklandjdstrand: ^20:10
jdstrandok20:11
kirklandjdstrand: okay, so i didn't have sound working ... is this expected at this point?20:11
jdstrandI wonder why it is looking in /home/kirkland/.pulse-cookie20:11
jdstrandkirkland: yes, but I have a feeling you'll get another denied message after the patch, but let's see what happens20:12
kirklandjdstrand: http://pastebin.ubuntu.com/294913/20:15
jdstrandkirkland: a blank line may have crept in. can you eyeball the result to make sure it applied right20:16
kirklandjdstrand: okay, looks right20:17
kirklandjdstrand: do i need to restart libvirt?20:17
jdstrandkirkland: no20:17
jdstrandkirkland: just try a new VM20:18
kirklandjdstrand: http://pastebin.ubuntu.com/294916/20:19
kirklandjdstrand: it booted to the desktop20:19
jdstrandkirkland: ok good-- apparmor is not denying access anymore20:19
jdstrandkirkland: do you have sound?20:19
kirklandjdstrand: no20:20
=== Keybuk_ is now known as Keybuk
jdstrandkirkland: do you have the sound device?20:21
kirklandjdstrand: yes20:21
kirklandjdstrand: http://pastebin.ubuntu.com/294918/20:22
kirklandjdstrand: that's what i get when i try to run the same command by hand20:22
kirklandjdstrand: no libvirt20:22
jdstrandkirkland: well, at least it booted. I'll try with a new vm. sounds like there may be a bug somewhere between how you start it with kvm and how libvirt starts it20:22
kirklandjdstrand: not sure it's related20:22
kirklandjdstrand: have you tried with the latest iso yet?20:22
jdstrandkirkland: still downloading20:23
jdstrandkirkland: is 'sound' the name of your vm?20:24
kirklandjdstrand: yeah20:24
jdstrandkirkland: that is unrelated20:24
kirklandjdstrand: okay, i've simplified the command line20:25
jdstrandkirkland: 0.7.0 started using a monitor socket20:25
jdstrandkirkland: Switch to using a unix socket for the qemu monitor (Mark McLoughlin)20:25
jdstrand(from the NEWS file)20:25
ChipzzMark McLoughlin, now there's a name that sounds familiar :)20:26
jdstrandkirkland: are you running the barebones kvm as root? that is how libvirt starts it when using qemu:///system20:26
Chipzzbe it from GNOME and not qemu :P20:26
kirklandjdstrand: no, as non-priv user20:27
kirklandjdstrand: i'm trying it with sudo now20:27
jdstrandkirkland: dude, it is taking *forever* to get the iso...20:32
kirklandjdstrand: where are you pulling it from?20:32
kirklandjdstrand: and at what rate?20:32
jdstrandcdimage.ubuntu.com20:32
kirklandjdstrand: i have pretty good upload20:32
jdstrand61.06kB/s20:32
kirklandjdstrand: private link coming20:32
waltersChipzz: he's on the Red Hat virt team now20:33
kirklandjdstrand: sounds works if i don't use libvirt20:34
jdstrandkirkland: let's try this20:36
jdstrandkirkland: can you do:20:36
jdstrandsudo apparmor_parser -R /etc/apparmor.d/usr.sbin.libvirtd && sudo /etc/init.d/libvirt-bin restart20:37
jdstrandkirkland: then do 'virsh capabilites | grep apparmor'20:37
jdstrandkirkland: what this does is start libvirt without the apparmor security driver. if you can't get sound to work there, it is clearly a problem between libvirt and kvm20:38
jdstrandkirkland: it is also temporary, so on the next reboot (or profile load and restart of libvirt) you'll have the standard, default libvirt configuration20:39
kirklandjdstrand: how urgent is all of this sound business?20:39
mdkeRiddell: something seems to have gone wrong with https://help.ubuntu.com/community/KarmicUpgrades - did you overwrite the page somehow? I can't see the old page history to revert it20:39
kirklandjdstrand: i don't particularly care about it, tbh ... i think i have some higher priorities20:40
jdstrandkirkland: well, I only brought it up to you because I plan to fix the apparmor issue. if this then exposes a pulseaudio/kvm/libvirt bug that causes a massive regression, I thought it be best to let you know about it first :)20:40
kirklandjdstrand: i have working sound when i use qemu-kvm; non-working sound when i use libvirt20:40
jdstrandkirkland: I don't care about sound working20:40
jdstrandkirkland: I care about the vm not starting20:41
kirklandjdstrand: right; i haven't reproduced that20:41
jdstrandkirkland: err, not booting to the desktop20:41
kirklandjdstrand: i suspect you won't either, when you get an up-to-date iso20:41
jdstrandkirkland: feel free move on. If I continue to have problems with the newer iso, I'll let you iknow20:41
kirklandjdstrand: and if the vm is still hanging, i'd start looking at the guest's kernel's dmesg20:41
kirklandjdstrand: okay, thanks;  sorry to be abrupt :-)20:42
kirklandjdstrand: i don't mean to be rude20:42
jdstrandkirkland: it isn't rude. I just don't want to catch you off guard with my upload20:42
kirklandjdstrand: right; vm not booting, that would be a regression.  vm without sound...well, i don't have sound right now through libvirt20:45
kirklandjdstrand: which i'm sure people will complain about20:45
kirklandjdstrand: but i'm not that bothered right now20:45
Riddellmdke: there was nothing there before20:45
chrisccoulsonkees - any luck trying to trigger the g-s-s crash reliably?20:48
chrisccoulsoni had a look through my xtrace log again, and that shows that it is the gnome-screensaver-dialog process which creates the resource which gnome-screensaver tries to access after it is destroyed20:49
keeschrisccoulson: no.  :(20:49
keeschrisccoulson: I can see the child defunct while I wait.20:49
keeschrisccoulson: it must be a very specific order of events20:49
chrisccoulsonkees - i agree. i could only recreate it occasionally before20:49
chrisccoulsonbut i've tried and tried to trigger it with the patch, and i don't get it any more20:50
chrisccoulsoni tried for 20 minutes continuously entering the wrong password last night, with the change in, and i didn't make it crash20:50
jdstrandkirkland: fyi, I tried with an intrepid desktop livecd. I disables 'splash and quiet' and it booted all the way20:50
chrisccoulsonand i tried again today too20:51
kirklandjdstrand: excellent20:51
jdstrandkirkland: there may be a race, I don't know20:51
jdstrandkirkland: oh, I left out part of it20:51
kirklandjdstrand: for your iso woes ...  i just realize that we're not building qemu-kvm against libcurl20:51
jdstrandkirkland: I tried with an intrepid desktop livecd. the first time, it hung. the second time I disabled 'splash and quiet' and it booted all the way20:52
kirklandjdstrand: if we did, you could -cdrom http://...iso and stream over the internet :-)20:52
chrisccoulsonkees - want me to put it in my PPA and ask some of the reporters to test it?20:53
keeschrisccoulson: I'm nearly 100% sure your patch is correct, but I'm mostly after trying to prove that Jaunty and earlier are/aren't affected.20:53
mdkeRiddell: hmm, that used to be quite an important page. not sure what happened to it then20:53
mdkeRiddell: but I hear the release managers are working on getting it back20:54
keeschrisccoulson: PPA> sure, yeah, can't hurt20:54
ccheneygrr is the a way to make chroots work now that we use upstart?20:54
ccheneyi'm getting stuff like: start: Unable to connect to Upstart: Failed to connect to socket /com/ubuntu/upstart: Connection refused20:54
ccheneycausing package installation in a chroot to fail20:54
chrisccoulsonkees - i had a look at the code for the jaunty version, and not much has changed in that area, so i can't see why jaunty wouldn't have the same issue. but then, there might be a GTK change in karmic that exposes the underlying issue20:55
keeschrisccoulson: yeah, that's what I was thinking20:55
chrisccoulsonkees - if i get some time, i will test my jaunty VM and see if i can trigger it there20:56
chrisccoulsonbut i could do with a more automated way of testing it really:(20:56
Riddellmdke: well it's never existed for karmic20:56
Riddellmdke: I thought ubuntu desktop used a page on ubuntu.com now20:56
keeschrisccoulson: does dogtail or the other thing let someone automate hitting enter on that dialog?20:57
chrisccoulsonkees - i'm not sure. i was thinking that i could maybe do a special jaunty build which bypasses all the pam stuff in the dialog process, and instead just loops and sends auth failure message back to gnome-screensaver, to simulate a user repeatedly entering the wrong password20:59
chrisccoulsonand then just leave that running for a while20:59
keeschrisccoulson: oh, heh20:59
chrisccoulsonbut perhaps there is a way of faking the enter keypress event21:00
jdstrandkirkland: well, 1 out of 10 times in hung. I think I'm done poking at this for now.21:01
kirklandjdstrand: ugh, that does sound racy21:01
kirklandjdstrand: i think we actually do need qemu-allow-pulseaudio-to-be-the-default.patch21:01
kirklandjdstrand: i thing i was too quick to revert that patch21:01
kirklandjdstrand: and i think the user/bugreporter may have been wrong21:01
sebnermighty Riddell, any plans to update k3b?21:02
jdstrandkirkland: it *really* feels like a too many resources accessing the sound card issue21:02
kirklandjdstrand: i have had working sound at various times duing karmic21:02
jdstrandkirkland: eg, one time I was listening to rhythmbox (ie pa was using the soundcard) and no sound in the vm21:02
jdstrandkirkland: other times sound in the vm21:02
jdstrandkirkland: and then the one time it hung21:02
mdkeRiddell: ah, possibly I may be out of touch21:03
mdkeRiddell: false alarm maybe21:03
Riddellsebner: update to what?21:03
jdstrandkirkland: can I hand this off to you for investigation whenever you see fit?21:04
sebnerRiddell: alpha321:04
kirklandjdstrand: you can hand off the non-libvirt, non-apparmor piece to me, yes21:05
sebnerRiddell: 1.68.0 alpha3 > 1.66.0~alpha2-0ubuntu721:05
kirklandjdstrand: i really won't be able to efficiently debug the libvirt or the apparmor aspects21:05
jdstrandkirkland: this was all with apparmor disabled21:05
jdstrandkirkland: it is not an apparmor issue. it hung without it21:05
kirklandjdstrand: i'll own the getting bare qemu-kvm to work with audio21:06
jdstrandkirkland: I'll just file a bug then21:06
kirklandjdstrand: oh, the hang?21:06
kirklandjdstrand: file that against the linux package21:06
Riddellsebner: guess I'll add it to my evening's todo list :)21:06
kirklandjdstrand: if the guest kernel is hanging, that's not a qemu-kvm issue, i don't think21:06
jdstrandkirkland: I don't think it is a 'linux' issue21:06
sebnerRiddell: cool! (I'm not using kde or k3b though, just let you know because of your users sake ;-D)21:06
jdstrandkirkland: I think the kvm process hung cause it was trying to access the soundcard21:06
kirklandjdstrand: okay, sure file the bug, i probably won't look at it until after karmic21:07
jdstrandkirkland: I believe this to be true because all of libvirt hung and I could use 'virsh list' until I killed off the kvm process trying to access the soundcard21:07
kirklandjdstrand: did you strace -p the kvm process?21:08
kirklandjdstrand: attach that to the bug21:08
jdstrandno21:08
kirklandjdstrand: that would help a lot21:08
jdstrandI can21:08
* jdstrand tries to get another hang...21:08
keesslangasek: I've just uploaded m2crypto for bug 451998.  no changes to the installed package -- just enables the testsuite and works around a weird issue on the builders.21:12
ubottuLaunchpad bug 451998 in m2crypto "Please run the test suite on build by default" [Medium,Fix committed] https://launchpad.net/bugs/45199821:12
keeschrisccoulson: heh, I've gotten as far as getting these, but no crashes:21:13
kees(gnome-screensaver:20207): Gtk-CRITICAL **: gtk_alignment_set_padding: assertion `GTK_IS_ALIGNMENT (alignment)' failed21:13
chrisccoulsonkees - those would suggest that the dialog has exitted at a point where gnome-screensaver has noticed it, and freed some client-side resources21:15
dobeyanyone around that can sponsor an upload for karmic main?21:15
keeschrisccoulson: yeah.  I'm more and more convinced this is an internal gtk change.  anyway, I'm going to upload your fix, as I think it is good for karmic, and I want to keep the late churn for RC down.  :)21:16
chrisccoulsonkees - thanks. i'll try and come up with a way of automatically testing this on a jaunty machine too21:16
chrisccoulsoni don't want to sit here for hours hitting my enter key ;)21:16
keeshehe21:17
kirklandjdstrand: question for you about your hang ....21:23
kirklandjdstrand: what cpu do you have?21:23
kirklandjdstrand: and do you have cpu freq scaling enabled?21:23
kirklandjdstrand: there are some known, non-deterministic hangs related to some AMD rev-F or old CPUs, when freq scaling is enabled21:24
jdstrandkirkland: dual core Intel. model name: Genuine Intel(R) CPU 3.00GHz21:24
kirklandjdstrand: okay, should not be affected21:25
jdstrandkirkland: it also hung in the same place in usplash each time21:25
jdstrandkirkland: 3 'ticks' in21:25
kirklandjdstrand: k21:25
jdstrandkirkland: fyi, bug #45345321:25
ubottuLaunchpad bug 453453 in libvirt "libvirt sometimes hangs when using pulseaudio" [Undecided,New] https://launchpad.net/bugs/45345321:25
jdstrandkirkland: I don't have the strace, but will attach when I see another hang21:26
kirklandjdstrand: thanks21:38
kirklandkees: could you throw some build super powers at https://edge.launchpad.net/~kirkland/+archive/ppa/+build/1295329 ?21:57
keessure, one sec21:57
keeskirkland: "Start in 1 minute"21:58
kirklandkees: beautiful21:58
elleuca hi, could someone try to check this bug? https://bugs.launchpad.net/bugs/45197422:00
ubottuLaunchpad bug 451974 in compiz "Black video minimizing Totem window" [Undecided,New]22:00
=== rmcbride_ is now known as rmcbride
jdstrandkirkland: fyi-- bug #45349522:21
ubottuLaunchpad bug 453495 in virt-manager "virt-manager does not allow selecting other architectures when using qemu" [Undecided,New] https://launchpad.net/bugs/45349522:21
jdstrandkirkland: I haven't looked to the why, but thought it might have had something to do with the qemu-kvm packaging22:21
jdstrandkirkland: that title was misleading. this is more accurate (bug #453495)22:23
ubottuLaunchpad bug 453495 in virt-manager "virt-manager does not allow selecting other architectures when using qemu" [Undecided,New] https://launchpad.net/bugs/45349522:23
jdstrandmeh22:23
jdstrandvirt-manager does not honor other architectures when using qemu22:23
kirklandjdstrand: perhaps... other arches are in qemu-kvm-extras22:23
jdstrandkirkland: see the bug. it's installed22:23
mathiazmvo: hi tried a mysql cluster bug upgrade - run into bug 45351322:36
ubottuLaunchpad bug 453513 in update-manager "mysql-server-5.0 is removed if mysql-server is removed on jaunty->karmic upgrade" [High,Triaged] https://launchpad.net/bugs/45351322:36
mathiazmvo: mysql-server-5.0 is removed on upgrade because it became an unused dependency (mysql-server being removed)22:37
=== ryu2 is now known as ryu
cjwatsonccheney: link /sbin/initctl to /bin/true; this will be in the release notes23:09
cjwatson(you'll probably want to move it aside rather than just blatting it)23:09
cjwatson(or maybe dpkg-divert)23:09
kirklandjdstrand: okay, i think i'm seeing your virt-manager hang23:34
kirklandjdstrand: so when i run kvm by hand, i see in my sound mixer qemu under the processes that pulse audio is attached to23:35
kirklandjdstrand: that's when sound works23:35
chrisccoulsonkees - i have an automated screensaver test up and running on jaunty now23:39
chrisccoulsonso i'll leave that going overnight and see what happens23:39
keeschrisccoulson: oh, cool; does it work in karmic?23:55
chrisccoulsonkees - i've not tried it in karmic yet, as it's implemented as a patch in gnome-screensaver23:58
chrisccoulsonit just repeatedly fakes keypresses23:59

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!