| === bladernr_ is now known as bladernr_afk | ||
| === kengyu_ is now known as kengyu | ||
| bullgard4 | [Linux T61 3.0.0-16 server #29-Ubuntu SMP] Why does '~$ nmon' > m state "Slab=119,1 MB" but /proc/meminfo is empty? | 06:48 |
|---|---|---|
| infinity | cooloney: In fact, the only thing that's saved jani's ac100 kernels from not overlapping due to ABI conflicts is that he was versioning as 3.0.19 instead of 3.0.0. ;) | 07:40 |
| infinity | cooloney: I've already asked him to switch to 3.2.0-foo when he revs to 3.2.0, so he'll need a unique ABI range too. | 07:40 |
| infinity | cooloney: So, yeah. A table in the wiki (or maybe even in the base git packaging templates?) that defines the out-of-tree-flavour to ABI-range mappings would be nice. | 07:41 |
| cooloney | oh, that's interesting | 07:41 |
| infinity | cooloney: Grandfathering in the ones we already have (omap4, armadaxp, highbank, and some linaro kernels) | 07:42 |
| cooloney | apw: hey, andy, do we have a wiki page about ABIs like infinity said | 07:42 |
| infinity | cooloney: And making it clear to people that those ranges belong to the out-of-tree flavours, and they shouldn't CHANGE (like they did for omap4...) | 07:42 |
| cooloney | infinity: yeah, we need document it | 07:43 |
| infinity | (Not that changing is a problem in itself, just that it eats more namespace, and soon all the flavours will be using ABI versions in the tens of thousands) :P | 07:43 |
| cooloney | infinity: if we enable ubuntu kernel for every ARM stuff, it will explode soon | 07:44 |
| infinity | Heh. | 07:45 |
| infinity | A bit, but such is life. | 07:45 |
| === smb` is now known as smb | ||
| smb | Life... don't tell me about life... :-P | 08:18 |
| cooloney | smb and ppisati do you guys know whether do you have a wiki page about abi numbers.like ti-omap4 is 1400 | 08:18 |
| smb | cooloney, yes | 08:18 |
| cooloney | since we are going to use 1600 for armadaxp | 08:18 |
| cooloney | and 1800 for highbank | 08:18 |
| smb | Now to find it (again) is a different issue | 08:19 |
| cooloney | this 2 kernel will maintained by pes for a while | 08:19 |
| cooloney | smb: heh, that's a big issue | 08:19 |
| smb | cooloney, https://wiki.ubuntu.com/Kernel/Dev/TopicBranches (I believe) | 08:20 |
| cooloney | smb: thanks a lot, so if we are going to use 1600 for armadaxp and 1800 for highbank | 08:25 |
| cooloney | smb: is there any process i need go through to apply such numbers, | 08:25 |
| smb | cooloney, You should talk to Brad or Herton as they do stable | 08:28 |
| apw | infinity, we change the numbers where some fool is going to use the same version of a krnel in two releases, as often happens with ARM | 08:35 |
| apw | infinity, and as smb says they are documented in https://wiki.ubuntu.com/Kernel/Dev/TopicBranches and allocated by the stable team | 08:35 |
| cooloney | apw: infinity probablly sleep now. | 08:36 |
| cooloney | apw: just sent out an email about that, oh, forget you. | 08:36 |
| cooloney | apw: let put you in the loop | 08:36 |
| * smb is afraid to start the update check | 08:45 | |
| cking | smb, go for it, see what happens, what could go wrong? | 08:48 |
| smb | cking, brzzz buzzz | 08:49 |
| apw | can anyone hear me | 08:49 |
| cking | nope not in irc | 08:49 |
| apw | *slap* | 08:50 |
| apw | diwic, hey ... i have a set of usb speakers and would like to construct a pactl command line to swtich between internal speakers and usb speakers ... is that even possible | 09:12 |
| diwic | apw, try "pacmd set-default-sink <name of usb speakers>" | 09:12 |
| diwic | apw, "pacmd list-sinks" will give you the name to choose | 09:12 |
| apw | diwic, doesn't seem to change anything | 09:14 |
| apw | alsa_output.usb-Logitech_Logitech_Speaker-00-ker.analog-stereo | 09:14 |
| apw | assuming that is the right sort of name | 09:15 |
| diwic | apw, hmm, it could be that setting the default sink does not move the streams, only puts new ones on the USB speakers, can you verify? | 09:16 |
| apw | diwic, nope, a new rhythmbox is still on internal | 09:17 |
| diwic | apw, can you check with "pactl stat" if the default sink has been changed? | 09:19 |
| apw | Default Sink: alsa_output.usb-Logitech_Logitech_Speaker-00-Speaker.analog-stereo | 09:19 |
| apw | yep is indeed, does | 09:20 |
| apw | doesn't seem to mean anything :/ | 09:20 |
| apw | diwic, is it me or is pa designed to make your head hurt | 09:22 |
| diwic | apw, then maybe there is an application level override that makes rhythmbox always go the internal speakers | 09:23 |
| apw | and how would i have made that, and indeed see that, so i can change it | 09:24 |
| diwic | apw, have you used the playback tab in pavucontrol, chances are you have one | 09:24 |
| * apw hankers for the old days of ... sound cards | 09:24 | |
| apw | diwic, how can i find out if i do have one | 09:25 |
| diwic | apw, you can remove all your application overrides by deleting ~/.pulse/ (in particular the device-volume and stream-volume files) | 09:26 |
| apw | and that is the only way i can even see i have them? | 09:26 |
| diwic | apw, analysing a pulseaudio verbose log would also reveal that information I believe | 09:28 |
| apw | diwic, oh well i think its reached "too much effort" level even for me ... i feel sorry for users | 09:28 |
| apw | they have no chance what so ever to understand this spagetti | 09:28 |
| diwic | apw, well, that's why gnome-volume-control never exposes that functionality | 09:29 |
| apw | diwic, that reminds me ... many mornings i have to kill pulseaudio to get mumble to work, and i'd say that more than one of us has this issue | 09:30 |
| apw | diwic, anything you have heard of ? | 09:30 |
| diwic | apw, nope. Kill permanently or just once and let it respawn? | 09:30 |
| apw | diwic, killall pulseaudio and let it respawn then start mumble again | 09:30 |
| apw | it seems to get stuck when trying to 'send' the first time, so i guess when it opens the mic channel, but not idea how to be sure | 09:31 |
| diwic | apw, on Lucid I always had to start the "Audio wizard" to make audio work. | 09:31 |
| diwic | apw, let me start mumble and check | 09:32 |
| apw | the visual symptom is the red lips don't work on push-to-talk | 09:32 |
| smb | To me it seems to get locked playing something (which would also look like stuck lips) | 09:32 |
| diwic | apw, smb, edit /etc/pulse/default.pa and add "set-log-level 4" last in that file | 09:39 |
| cking | apw, pie: http://www.geekologie.com/2010/04/omg-omg-omg-314-in-a-mirror-sp.php | 09:44 |
| Laibsch | cking: thank you for taking a stab at bug 745836. You are not in a position to accept the SRU process for the ticket? | 10:33 |
| ubot2` | Launchpad bug 745836 in linux "encrypted swap corrupts application stack/heap [was: soffice.bin SIGSEGV cppu::throwException()]" [High,Confirmed] https://launchpad.net/bugs/745836 | 10:33 |
| Laibsch | IOW, accept the taks | 10:33 |
| Laibsch | task, even | 10:33 |
| cking | Laibsch, it needs to be ACK'd by kernel devs and it will work through the process in due course | 10:34 |
| Laibsch | yeah, that was kind of my question, if you are an "official" kernel dev | 10:34 |
| Laibsch | I think apw is ;-) Would you mind having a look at https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bug/745836/comments/111 and accepting the lucid task? Would be much appreciated. | 10:35 |
| cking | Laibsch, I can't and won't do that - there is due process ;-) | 10:35 |
| ubot2` | Launchpad bug 745836 in linux "encrypted swap corrupts application stack/heap [was: soffice.bin SIGSEGV cppu::throwException()]" [High,Confirmed] | 10:35 |
| cking | Laibsch, it will get looked at, don't worry | 10:36 |
| Laibsch | well, I do worry. This is fairly troublesome. And my previous experiences with kernel bugs was less than stellar. | 10:36 |
| cking | the process can only go so fast | 10:37 |
| Laibsch | basically, it already fallen through the cracks until I revisited the issue yesterday. It's been my experience that until a task for a release is accepted it's off the radar if ubuntu+1 is not affected. | 10:38 |
| cking | and I did get up at 7am to do this ;-) | 10:38 |
| Laibsch | much appreciated! | 10:38 |
| apw | Laibsch, in the kernel process accepting the lucid task has no real meaning, it does not imply we will accept the sru if the patch is rejected | 10:38 |
| apw | anyhow approved the nom | 10:38 |
| Laibsch | apw: I know that. But acceptance puts it on the radar. The issue shows up in searches, etc. | 10:38 |
| Laibsch | apw: cool, thanks. | 10:38 |
| apw | if cking has sent stuff the kernel-team@ list then its on our highest radar | 10:39 |
| cking | :-) | 10:39 |
| Laibsch | good bye | 10:51 |
| ppisati | next time i pull request something not_mandatory from upstream, beat me with a club | 11:23 |
| cking | ppisati, will do | 11:49 |
| ppisati | smb: which flight do you take from AMS to SFO? | 12:22 |
| * ppisati is evaluating the options and is gonna flight on Sat too | 12:23 | |
| smb | ppisati, 9:50 kl605 | 12:23 |
| ppisati | smb: ok, it's part of one of my options but i was wondering: i've 1hr when i arrive there... uhmm... | 12:24 |
| ppisati | smb: since we are entering the US, don't they have to check us again? | 12:24 |
| smb | ppisati, Yeah, 1 seems tight with AMS | 12:24 |
| tgardner | ppisati, thats not really enough time, is it ? | 12:24 |
| ppisati | tgardner: right | 12:24 |
| smb | At least to my experience there is that huge passport portal you have to go through, which takes time | 12:25 |
| ppisati | tgardner: problem is that almost all my options have ~1hr as buffer between coindidences | 12:25 |
| ppisati | grrr... | 12:25 |
| smb | ppisati, IMO it helps to get back and state this | 12:25 |
| ppisati | smb: will do | 12:26 |
| tgardner | ppisati, did you see the ti-omap4 email from ricardo salveti ? | 12:38 |
| ppisati | tgardner: yep, i'm preparing the new kernel | 12:38 |
| tgardner | ppisati, ack | 12:38 |
| ppisati | tgardner: i'll answer to the email later | 12:38 |
| jsalisbury | cking, I'll respond to Peter with my system data. | 12:54 |
| * apw lunches ... | 12:58 | |
| * ppisati lunches too | 13:04 | |
| === bladernr_afk is now known as bladernr_ | ||
| rsalveti | ppisati: cool, thanks, will you send the pull request today still? | 13:28 |
| rsalveti | once the new kernel lands I'll try to also push the sgx driver to the archive | 13:29 |
| rsalveti | ppisati: can you also check bug 956693 after the kernel update/ | 13:35 |
| ubot2` | Launchpad bug 956693 in linux-ti-omap4 "Broken perf with Ubuntu-3.2.0-1408.11 at Pandaboard" [Medium,Confirmed] https://launchpad.net/bugs/956693 | 13:35 |
| ppisati | rsalveti: i'll send the pull req in a minute | 13:37 |
| rsalveti | ppisati: awesome, thanks | 13:37 |
| === bladernr_ is now known as bladernr_afk | ||
| === bladernr_afk is now known as bladernr_ | ||
| * apw is tired | 13:58 | |
| cking | apw, been swimming then? | 14:04 |
| smb | oh dear... /me forgot about the abi changing for lucid-natty... | 14:05 |
| cking | jsalisbury, thanks - I think we may get some response by middle of next week - that's the kind of timescale they work at | 14:08 |
| jsalisbury | cking, ok. I thought I had to wait to reboot and go into the BIOS, but it looks like I can get the data with dmidecode | 14:10 |
| smb | herton, that kvm change had been causing hashes to change. are you already looking into it or do you want me to insert abi bumps? | 14:11 |
| cking | jsalisbury, yep, the guys an lenovo don't realise that there are tools to do this ;-) | 14:11 |
| jsalisbury | cking, heh | 14:11 |
| herton | smb, I fixed natty already, lucid I think we already had bumped the abi before some days ago | 14:12 |
| smb | herton, Ah ok, cool. Just saw that build failure | 14:12 |
| smb | then remembered I had been forcing some of my test builds | 14:13 |
| herton | smb, hmm, lucid hadn't an ABI bump, I'm going to check and do one | 14:14 |
| smb | herton, OK, unfortunately I had been more concentrating on the testing. So I would not remember where exactly I had used ignoreabi. But since the x86_emulate_ops struct changes in all releases it requires a bump too (if not already havin one) | 14:16 |
| apw | is there any way we can get the gnome volume control to not "shrink back" when an application goes away ... as for me its full then every time pidgin makes a noise it connects and disconnects from PA which makes the gnome volue thingy get bigger <bing bong> and then smaller | 14:19 |
| herton | smb, pushed the abi bump to lucid now | 14:21 |
| jcastro | jsalisbury, nuts, I missed the bugmail where you prod me for the new kernel for the x220 wireless, testing now, sorry about that! | 14:22 |
| apw | cking, yeah went swimming ... lest i end up chair shaped | 14:22 |
| jsalisbury | jcastro, np, thanks again for helping out! | 14:22 |
| * cking goes and does some paper work - yawn | 14:27 | |
| smb | apw, If we would be as slim as our chairs there might be some good in that | 14:27 |
| smb | cking, Oregami? :) | 14:28 |
| cking | smb, nah, if only | 14:28 |
| smb | *sigh* that keybinding for semi-maximise would not be as bad if I had more than a pixel to hit in order to undo it when accidentally done... | 14:31 |
| apw | heheh | 14:46 |
| apw | anyone know if any of the thinkpads have hybrid graphics ? | 14:55 |
| mjg59 | apw: Some do | 14:56 |
| apw | mjg59, hmm then that might explain some reports of "port X works but not port Y" | 14:56 |
| apw | thanks | 14:57 |
| hrw | apw: thx for reply to my patch mail. suihkulokki filled bug with a bit more expanded version: bug 957028 | 15:10 |
| ubot2` | Launchpad bug 957028 in linux "Fix cross-compiling kernel" [Undecided,Incomplete] https://launchpad.net/bugs/957028 | 15:10 |
| ogasawara | apw, tgardner: could one of you do me a favor and do a quick boot test the i386 3.2.0-19.30 kernel I've built on gomeisa (in my home dir under precise-i386/) | 15:12 |
| tgardner | ogasawara, your wish is my demand | 15:12 |
| apw | ogasawara, can i check that the upload you plan includes the d-i fix for hyper-v | 15:12 |
| ogasawara | tgardner: thanks | 15:12 |
| apw | ogasawara, ps the meeting is now, so i'll be busy for a bit | 15:13 |
| ogasawara | apw: yep, I've pushed to master-next what will be in the upload | 15:13 |
| ogasawara | apw: ack | 15:13 |
| ogasawara | apw: ah, I thought the meeting would be an hour from now. in that case I can sit in if you like. | 15:17 |
| apw | ogasawara, they got all confused over the time change in teh US of course, i have been listening so don't bother | 15:17 |
| herton | smb, I'm gonna revert the ABI bump on lucid, the kvm changes didn't trigger abi-check complaints | 15:24 |
| smb | herton, Oh.. a bit surprising (and I though I had to force abi there). But if there were not any compaints... | 15:25 |
| tgardner | ogasawara, ubuntu@precise-i386:~$ uname -a | 15:26 |
| tgardner | Linux precise-i386 3.2.0-19-generic #30 SMP Fri Mar 16 05:39:35 UTC 2012 i686 i686 i386 GNU/Linux | 15:26 |
| ogasawara | tgardner: sweet, thanks | 15:26 |
| tgardner | ogasawara, no complaints with AA in dmesg | 15:26 |
| herton | smb, yeah, I was trying to hunting down how changing struct x86_emulate_ops affected struct kvm_vcpu, as this seems to have triggered the hash changes, but there is a lot of underlying structs to check | 15:26 |
| herton | on natty | 15:27 |
| tgardner | herton, smb: don't you guys do full builds to check ABI ? | 15:27 |
| herton | tgardner, usually I do, just we expected this time to lucid have gone of the way of natty | 15:28 |
| smb | tgardner, well I did full builds to get test kernels, but there were plenty of those and I usually use skipabi to get things done | 15:29 |
| smb | And then forget details.. :( | 15:29 |
| * ppisati disapperas for ~30mins | 15:29 | |
| * ogasawara reboots | 15:30 | |
| === yofel_ is now known as yofel | ||
| apw | ppisati, where are we at with pending fixes for ARM for 3d ? | 15:37 |
| * apw pops out to get some supplies | 15:51 | |
| ogasawara | tgardner: non-smp powerpc linux-meta package, I assume we'll want to just point that to the powerpc-smp flavor for a smooth update/upgrade path | 16:01 |
| tgardner | ogasawara, that seems right to me | 16:01 |
| tgardner | ogasawara, I've got a test build that is _still_ running on davis | 16:02 |
| ogasawara | tgardner: I did a full test build last night | 16:02 |
| tgardner | ogasawara, on master-next ? | 16:02 |
| ogasawara | tgardner: yep | 16:02 |
| tgardner | ogasawara, cool, then I'll kill mine | 16:02 |
| tgardner | ogasawara, did I actually get it right the first time ? | 16:03 |
| ogasawara | tgardner: it looked right to me, powerpc-smp and powerpc64-smp udebs were built and powerpc went away. | 16:03 |
| tgardner | ogasawara, woot! | 16:04 |
| ogasawara | tgardner: I've still got it on davis if you wanted to take a quick peek | 16:04 |
| ogasawara | tgardner: in my home dir under precise-powerpc | 16:04 |
| tgardner | ogasawara, nah, I'll take your word for it :) | 16:04 |
| ppisati | apw: the new kernel has all the needed bits | 16:08 |
| ppisati | apw: ricardo will roll a new sgx driver when this kernel hits the archive i guess | 16:08 |
| ppisati | rsalveti: ^^ | 16:09 |
| cyphermox | hey; just to make sure you are aware of this issue with WPA on ad-hoc: https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/905748 | 16:15 |
| ubot2` | Launchpad bug 905748 in network-manager "Create WPA2 adhoc is Open, not encrypted" [High,In progress] | 16:15 |
| cyphermox | there's a linux task as well ^ | 16:15 |
| tgardner | cyphermox, is there a reason for the linux task ? seems like its nm or wpa-supplicant at fault, though I have not read the bug report. | 16:24 |
| cyphermox | tgardner: yes, it's reproducible on various drivers, with NM or just wpasupplicant, and the kernel displays the network as WPA 1 in iw dev wlan0 scan; but on a different system (or on an android phone) it gets seen as open. | 16:26 |
| cyphermox | tgardner: any way I can help debugging this and ensuring it's really not wpasupplicant? | 16:28 |
| tgardner | cyphermox, I've not used wpa_supplicant from the command line, but I believe sforshee has. | 16:29 |
| cyphermox | tgardner: that's how I was testing it | 16:29 |
| * smb thinks this Friday has went far enough... | 16:30 | |
| tgardner | cyphermox, are there any clues in syslog that indicate encryption is or isn't in use ? | 16:30 |
| tgardner | cyphermox, you might also ask how to debug this on linux-wireless@vger.kernel.org | 16:32 |
| cyphermox | tgardner: thanks. so no, not really anything to suggest it is or isn't in used, though in all honesty I wasn't running wpasupplicant in debug more when I tested it; and should have | 16:33 |
| cyphermox | I'll try that again after lunch | 16:33 |
| tgardner | cyphermox, yeah, if wpa_supplicant thinks its starting an encrypted session, but other clients see it as open, then it likely is a kernel issue | 16:33 |
| cyphermox | tgardner: what leads me to believe it's kernel is especially how when it's set up on my system 'iw dev wlan0 scan' lists it as WPA protocol 1; but android sees it open (and so do other systems) | 16:36 |
| tgardner | cyphermox, that makes sense | 16:36 |
| cyphermox | I've seen another ubuntu machine have iw report security, but NM happily connects to it with any passphrase, traffic passes without issues | 16:37 |
| cyphermox | all these fun things | 16:37 |
| * tgardner relocates | 16:48 | |
| cking | hrm downloading a kernel .ddeb is taking ages | 16:54 |
| ogra_ | yeah, we should stop putting all these symbols into them to make the download faster :) | 16:55 |
| apw | cking, used up your whole years allowance i expect | 16:57 |
| cking | apw, too right | 16:57 |
| * cking -> EOD | 17:59 | |
| === tgardner is now known as tgardner-lunch | ||
| rsalveti | ppisati: yup, once it lands I'll try to push the sgx driver in | 18:19 |
| ppisati | rsalveti: k | 18:23 |
| === mahmoh1 is now known as mahmoh | ||
| === tgardner-lunch is now known as tgardner | ||
| cyphermox | tgardner: I went pretty verbose on that one, but I updated the bug (bug 905748) with debug logs from wpasupplicant and other tests I could think of | 19:30 |
| ubot2` | Launchpad bug 905748 in network-manager "Create WPA2 adhoc is Open, not encrypted" [High,In progress] https://launchpad.net/bugs/905748 | 19:30 |
| tgardner | cyphermox, that looks like enough info to send an email to upstream | 19:32 |
| cyphermox | tgardner: should I read this as "please contact upstream" or do you have contacts? :) | 19:34 |
| tgardner | cyphermox, the former. | 19:35 |
| cyphermox | alright :) | 19:35 |
| tgardner | cyphermox, this bug is well beyond my knowledge of the protocol code | 19:36 |
| cyphermox | tgardner: no problem, just trying to be thorough because this is kind of stupidly broken | 19:36 |
| cyphermox | tgardner: as stop-gap I was trying to disable creating ad-hoc WPA via NM, but my patch didn't work as expected, upstream NM says they'll look into it | 19:37 |
| cyphermox | tgardner: what upstream version is the current precise kernel equivalent to? | 19:42 |
| cyphermox | ( I see 3.2.0 but also 3.2.9 in proc/version_signature) | 19:42 |
| tgardner | cyphermox, 3.2.11 (I think). lemme check for sure | 19:42 |
| tgardner | cyphermox, 'UBUNTU: Rebase to v3.2.11' | 19:43 |
| cyphermox | ah, thanks | 19:43 |
| tgardner | cyphermox, that is what was just uploaded a couple of hours ago. your version may perhaps be older | 19:43 |
| cyphermox | on, in that case yeah just a bit older -- Ubuntu 3.2.0-18.29-generic 3.2.9 | 19:44 |
| cyphermox | actually, just noticed something interesting, a link error message from nl80211 | 19:53 |
| cyphermox | tgardner: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/905748/comments/45 ; or just tell me to shut up if you don't care ;) | 19:57 |
| ubot2` | Launchpad bug 905748 in network-manager "Create WPA2 adhoc is Open, not encrypted" [High,In progress] | 19:57 |
| tgardner | cyphermox, that looks like a good clue. | 19:58 |
| tgardner | cyphermox, I'm EOD and gotta blast off. | 19:58 |
| * tgardner -> EOD | 19:58 | |
| === bladernr_ is now known as bladernr_afk | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!