[00:17] apw: problem fixed and it wasn't in the kernel [00:34] I === Guest16781 is now known as albrigha [08:15] jjohansen, excellent thanks for your help [08:16] apw: np, sorry for the problems, and wasting your time [08:16] jjohansen, no need to appologise, there are always issues, and them seeing we can resolved them in a timely fashion is good [08:18] yeah I suppose, still feel bad when I caused the issue, but its not the only one this cycle, /me is on a roll [08:19] heh, nothing compared to the upstream issues we have consumed [08:19] at least someone is testing and finding them before release [08:21] :) [08:21] jjohansen, was it luck that it was working for everyone else ? [08:21] apw: yes [08:21] jjohansen, those are the best ones to find [08:22] non terminated string, so it just depended on whether the string was dumped into zeroed memory [08:23] this is when the compiler was deciding what features the kernel supported [08:23] the precise kernel has a new interface for that [08:23] heh, and it should have been randome for everyone probabally, and yet 100% reproducible for him and not for anyone else in the company [08:23] so the old kernel path would always work, the new for most people [08:23] how wack is that [08:23] very [08:24] a good find, you know it would have worked perfectly until release week otherwise [08:25] hehe, yeah [08:31] * smb cries [08:31] smb, that bad? break your coffee maker? run out of coffee ? [08:31] So some launcher icons having strange colours is because of my desktop background [08:32] and they change colour everytime you change your background [08:32] !?! is this a bug or a feature? [08:32] apw: Error: I am only a bot, please don't think I'm intelligent :) [08:32] apparently a feature [08:32] smb: oh I hate that, is it as bad as the spilled cool-aid on my desktop bug [08:32] ubot2 i didn't even consider you present [08:33] jjohansen, Heh, I am not sure I know that [08:34] and apparently working scroll wheel for your terminals is a game of luck [08:34] * ppisati -> reboot [08:34] smb: I'll see if I can't dig it up, but it was related to choosing colors based off your background [08:35] jjohansen, Ah yeah, Paolo said they do that now as a feature. Maybe I get used to it [08:36] Just that I had been wondering all the time why they changed the colors into that particular one [08:37] ok, missing scrollwheel for terminals seems to be fixed by todays updates... I had not closed the one from before [08:37] yeah, some time its a real mess, /me is particularly annoyed with the alt-tab switch where I can't tell which icon is being highlighted [08:37] That one gets me a lot of times [08:38] and the inconsistency of what moves between subselections of the same app [08:39] when you drop in the by time its tab. when you get there by cursor up its cursor left/right [08:39] oh, actually that seems to have changed [08:40] now its only that going in there by cursor will give you round robin within that new selection while dropping in by time gets you out when you go past the last entry [08:42] yeah, /me is going to try the compiz switch again, its working again [08:46] jjohansen, compiz switch ? [08:47] apw: compiz alt-tab switchers if you use ccsm you could still use them instead of the unity one at least until a few weeks ago when they where broken by a compiz update [08:48] * jjohansen decided to give the unity one a good run but it hasn't grown on me at all [08:48] ahh ... jjohansen before you switch away from it, it also now has different settings, like switching within apps by default and stuff on the experimental tab [08:48] apw: yeah [08:48] it also has whether its limited to the same viewport or not etc as tickies [08:48] depends what your issue is of course [08:49] but i was supprised at the slew of options it had grown [08:50] yep, still don't like it. I have several gripes [08:50] Its inconsistent, some time it pulls all an apps windows forward, some times it doesn't [08:50] I can't tell which app I am switching too [08:50] I loose spatial information on the window so I can be looking at the window I want but have to try picking it out of an icon, or wait for it to give a shrunken view of the window, which I still can't distiguish between when its between 3 or 4 terminals [08:51] the spatial info loss is my biggest gripe, assuming that the bugs are worked out [08:55] sigh, no the compiz application switcher is still buggy [09:14] jjohansen, i do find working out which one is highlighted difficult for some reason [09:15] apw: it used to be a little easier, but they changed the highlighting color and now I find it really hard === amitk is now known as amitk-afk [10:08] apw, the sound card name of your snowball mic? [10:13] diwic, hi [10:14] diwic, how do i find that out [10:14] apw, so I got upstream's blessing for my solution to blacklist USB devices which don't have digital in/out, just need the card names to put in the blacklist [10:14] apw, pastebin /proc/asound/cards [10:14] with the mic connected [10:15] 1 [Snowball ]: USB-Audio - Blue Snowball [10:15] BLUE MICROPHONE Blue Snowball at usb-0000:00:1d.7-4.5.4, full speed [10:15] thanks! [10:15] that is showing two entries, one is 'iec958-stereo-input' the other 'Microphone' [10:16] though for all i know it really does have stereo support [10:16] it was expensive enoug [10:16] h [10:16] the normal "microphone" input should enable stereo recording by default. [10:17] ahh ok, then yeah i don't need two :) [10:17] why do we get two? [10:17] oh interesting they are linked, the mute on one mutes both anyhow [10:18] the iec958 shouldn't be there. [10:18] it's a bug. [10:18] good enough [10:19] apw, feel free to read bug 940145 if you're interested in more details. [10:19] Launchpad bug 940145 in pulseaudio "USB analog only devices sometimes show up with digital profile as well" [Undecided,In progress] https://launchpad.net/bugs/940145 === amitk-afk is now known as amitk [13:24] * apw prepares to shift location ... [15:04] * ogasawara back in 20 [15:14] * apw returns [15:20] apw -> pop stack into kent [15:20] * apw smells coffee ... yumm [15:20] apw, that didn't take long. [15:21] apparently 1:45 which is a lot really [15:25] apw, M25/M20 slow then? [15:25] cking, no it was a pretty steady run, no expalnation other than faffing [15:35] ogasawara, closed a couple of WIs for the hyper-v stuff ... [15:36] apw: sweet, thanks [15:38] apw, tgardner: will likely upload 3.2.0-18.29 later today, gcc was uploaded yesterday so am going to wait for it to finish up its builds [15:39] ogasawara, sounds good [15:39] ogasawara, I updated tangerine/gomeisa chroots this AM with new gcc and binutils [15:39] tgardner: cool, thanks. yah, the only build still going is powerpc [15:49] ogasawara, any idea how long it takes on powerpc normally ? === JanC_ is now known as JanC [15:49] apw: previous build was ~9hrs [15:49] ogasawara, so another 5 then [15:50] ogasawara, if you want me to upload it in the morning, drop me a note [15:50] apw: no worries, I'm hoping it'll just be finishing up before I usually bail. If not, I'll check it this evening and upload. [15:51] if not, you know where i am :) [16:12] apw, ... in the basement right hand side room? [16:35] apw, there is a hyperv patch that you should cherry-pick after David Miller's latest pull request is honored; '10) hyperv network driver uses the wrong driver name string, fix from Haiyang Zhang.' [16:36] tgardner, ack [16:36] apw, its real subject line is "net/hyperv: Use the built-in macro KBUILD_MODNAME for this driver" [16:42] apw, huh, its already committed: d31b20fcc89efa8c5d3f5ea2720e08a286b69a36 [16:42] tgardner, yay thanks ... will pull it over [16:43] tgardner, that must have been recent ... as tis not in my pull from this morning [16:43] apw, no, I think it was in the last few minutes [16:43] tgardner, and there it is ... will get it now [16:47] yawn [16:55] tgardner, ok pushed [16:55] apw, ack [17:14] * apw reboots to test a kernel [17:37] * tgardner has an appt, back in awhile. === tgardner is now known as tgardner-afk [18:02] apw ping? [18:04] akgraner, hi [18:04] apw - guess what is happening again - I'm getting heat issues but I don't know if it's chromium or something else [18:05] my machine has hit 90+ C temps 5 times in less than 5 hours [18:05] and shuts down hard [18:05] akgraner, same old machine that did it before ? [18:06] akgraner, this issue went away for the longest time didn't it [18:06] different machine [18:06] thinkpad now [18:06] akgraner, i assume it is on precise now? [18:06] the dell finally died a heat ridden death [18:06] apw, yep upgraded today [18:07] akgraner, and it was fine on oneiric ? [18:07] yep [18:07] just started today [18:07] ok so then ... we need to get a bug filed for this [18:08] akgraner: is this after a few suspend/resumes or from cold boot and after some time running ? [18:08] and your first debugging task is to boot back into your oneiric kernel [18:08] It's docked [18:08] and confirm it is ok in that combination, if so then we need to try the kernels in between [18:08] to see where it is coming from [18:08] I have have dual monitors as well - so first I lose a monitor then the heat pegs then it dies [18:12] apw, ok - let me put it through the paces again and see if I can reproduce on purpose [18:12] the drop back the oneiric kernel and go from there [18:12] akgraner, sounds good ... [18:13] akgraner, bjf had a question, which might imply he has an idea what the issue is ^^ [18:13] thanks - just thought I'd give you a heads up [18:13] akgraner, and what thinkpad model is it [18:13] bjf - after it's been running - longest time was a little over an houe [18:14] akgraner, i think he is asking is it from cold boot, use, break [18:14] akgraner, or is it cold boot, s/r a few times, use, break [18:14] bjf T410 [18:14] no s/r [18:15] happen once from cold boot, the rest of the times during use [18:16] akgraner, when you have a bug filed, get the number to jsalisbury so he can do a bisect once you confirm that O is good and P is bad with the same userspace [18:17] akgraner: i ask because i have a x301 and after some number of suspend/resumes it seems to forget to regulate the fans [18:17] apw - will do [18:18] akgraner: i can "fix" it by just rebooting or I have a small script which i can manually adjust the fans [18:20] bjf - ahh ok...I'll pay closer attention and verify its not the fans if it is I'll let ya know :-) [18:25] bjf, i'm looking at thinkpad fan controls at the moment, I'm doing some precise measurements with specific loads and comparing kernels etc [18:26] bjf, can you send me that script too? [18:27] cking, sent [18:27] ta! [18:30] apw - so I just killed deja dupe and that looks like it was part of the issue [18:30] bjf, so the default "auto" seems not be OK for a load of users, I'm seeing how auto performs compared to different level settings [18:32] akgraner, the machine should not overheat under load really, something is wrong [18:32] akgraner, how old is the machine? My lenovo kept on rebooting because of thermal issues because the fan outlet was full of fluff [18:34] cking, I know - b/c of your post that's the first thing I check now :-) [18:34] fluff 'n' fans don't mix [18:35] its tempting to set the fan level to "full-speed" and see how hot it gets [18:35] apw - 2 years old [18:36] apw - Its an arrendale chipset [18:36] heh, !quality [18:39] * cking shudders [18:40] cking, I hope it's because you a cold :-) === tgardner-afk is now known as tgardner [18:40] ok - I'll get this filed and let you all know more soon thanks! [18:42] apw - if would help and save time for all of us - I can share my screen and open a term for you to have a look [18:43] you'll just have to tell me what you want to see etc [18:45] akgraner, i think what needs doing is something that'll take longish testing, so it'll make more sense for you to do it [18:45] no worries [18:45] you know I don't mind :-) [18:49] akgraner, have you got a bug number I can put some comments into for this issue? [18:50] I will in a minute :-) [18:50] cool [19:02] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/951049 [19:02] Launchpad bug 951049 in linux "Lenovo T410 is overheating and shutting down" [Undecided,New] [19:03] cking, apw- I wasn't sure what all to say - let me know if I need to add anything else right now === tgardner is now known as tgardner-lunch [19:09] I will put some comments.. /me needs to reboot first [19:13] popey: do you have that viewsonic monitor handy, and would you be willing to try some magic to possibly fix it? :) [19:15] akgraner, I've updated the bug with something to try for the next 24 or so hours. [19:15] ok will do :-) thank you! === tgardner-lunch is now known as tgardner [19:30] sconklin: Got a sec? I was thinking about our power test measurement, and looking at http://www.amazon.com/Electronic-Specialties-695P-Current-Multimeter/dp/B000I17RAM/ref=sr_1_18?ie=UTF8&qid=1331320671&sr=8-18 as a possible better solution to tapping inline. [19:31] Wanted to know your thoughts. [19:39] GrueMaster: my first impressions are 1) 10 mA resolution really isn't good enough. and 2) you'd have to either still get down to a single supply line or be able to clip it over all the wires for a given voltage. [19:39] GrueMaster: but it should just plug into what we have, and for the price, if 10 mA is good enough resolution for you it might be worth a try [19:40] cking, neat - poof magic - now I hear the fans [19:40] Clipping over the wires is easier than splicing in between. That was my thought at least. [19:40] akgraner, that's good - lets see if it overheats with them on full tilt [19:42] cking, I'll let ya know [19:42] pop the details in the bug and we can work on finding out where it's going wrong [19:42] ta! [19:42] * cking calls it a day [20:15] sforshee, re: bug #950320 - you should put the workaround to enable passive DFS channels in the bug. [20:15] Launchpad bug 950320 in linux "[MacBook Air 4,1] Wireless scan fails to see APs on DFS channels" [Medium,Confirmed] https://launchpad.net/bugs/950320 [20:16] tgardner, it's there in the bug description [20:17] sforshee, ah, its just damn hard to see. [20:17] I guess I could add ***WORKAROUND*** to make it easier to find ;) [20:20] tgardner, I made it a little easier to find [20:21] sforshee, yeah, that works [20:22] tgardner, sounds like broadcom is going to fix that problem by adding the custom domain to brcmsmac, but I don't know when we can expect it [20:24] sforshee, likely 3.4 timeframe [20:24] if we're lucky [20:24] the merge window will be pretty soon now [20:24] sforshee, perhaps you could accelerate the process by sending a patch to Linville as a bug fix. [20:25] sforshee, and Cc stable [20:25] tgardner, a patch to do what? add the custom domain? [20:26] I don't think I actually know what's valid for this broadcom-custom regulatory domain [20:26] although likely it's the same as the world domain but with passive scanning on the DFS channels [20:27] sforshee, is there a domain the allows passive scanning on DFS channels ? [20:27] tgardner, US does for sure [20:27] I guess I could detect X0 and override it with US and see what they think of that [20:28] seems like a hack. does the X0 come from EEPROM or something ? [20:28] tgardner, yes, from eeprom [20:28] it's definitely a hack, although it sounds like they already do something similar [20:29] sforshee, I thought you said X0 is bogus, right? Is that what causes mac80211 to use the world domain ? [20:30] tgardner, yes. X0 is something broadcom made up. Since CRDA doesn't know about it it results in using the default, i.e. the world domain [20:30] sforshee, well then, perhaps defaulting to a US domain isn't such a hack. [20:31] tgardner, it's still a hack, but not an unreasonable one [20:31] it might not be that hard to do the custom domain though, I'll look into it [20:31] isn't the world one the subset of all the regulatory domains [20:31] it's easy, i've done it before [20:31] when I send the patch I suppose broadcom can correct what I get wrong [20:31] you add it to crda, regen; sign it yourself and it's automatically used if put in the right place [20:32] ohsix, yes, but we want it to work by default [20:32] the one the kernel uses is hard-coded in the wireless core code [20:32] right, so ubuntu would sign it instead of having the linux-wireless guys key on it :] [20:32] really? i thought that went away a long time ago, or are you referring just to the broadcom driver? [20:32] nope, look at net/wireless/reg.c [20:33] world_regdom specifically [20:33] hm, i see [20:34] oh, world is in there; the rest are in crda [20:34] ohsix, there's a reasonable explanation for the DFS channels being disabled by default. See http://www.spinics.net/lists/linux-wireless/msg86540.html [20:36] DFS is the radar thing isn't it? afaik it was almost always off; so i can't really provide any input, i thought having added my own regdom to crda was useful :] [20:37] yes, it's related to ceasing transmission when radar is detected on the channel [20:38] * apw gives it up [20:38] * tgardner is about ready to give it up [20:39] quitters [20:40] * apw sends sforshee 10 machines to debug, by monday [20:40] apw, par for the course ;) [20:40] :) [20:40] have a good 'en [20:41] you too [20:51] sforshee: awesome, thanks for the workaround! was wondering why i could never get on the 5ghz band on this thing [20:55] Sarvatt, I just noticed it earlier this week after I set up my new router and couldn't see my 5 GHz AP === yofel_ is now known as yofel