[08:09] <apw> RAOF_, was that another i915.powersave i saw you handing out?
[08:09] <apw> (on #ubuntu-devel ?)
[08:44] <RAOF> apw: Frustratingly slowly, yes.
[08:44] <apw> heh ... i did see some teeth being pulled
[08:44] <apw> RAOF, what was it on ...
[08:45] <RAOF> A 4 series card.  Let me grab the bug.  I'm pretty sure it'll be covered by our existing stuff.
[08:45]  * apw bounces
[08:45] <RAOF> I can still get a nouveau noaccel quirk into the kernel, can't I?
[08:46] <RAOF> The final kernel upload is tomorrow morning?
[08:50] <RAOF> apw: Yeah, there it is.  GM45.  https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/555503
[08:50] <ubot3> Malone bug 555503 in xserver-xorg-video-intel "screen flickers at least once every few minutes" [Medium,Confirmed] 
[08:50] <apw> RAOF, and that is already covered ... ?  why didn't the user have it
[08:52] <RAOF> They were on the -19 kernel.  Have we pushed the powersave quirk to the -20 kernel?
[08:56]  * apw actually has lost track after the kitten killer yesterday
[08:56]  * apw checks
[08:56] <RAOF> I can't see anything in the changelog.
[08:57] <apw> RAOF, no the change was to disable FBC for those sets not powersave
[08:58] <RAOF> Ah, right.
[08:58] <RAOF> bug #538648 is what I'm thinking of.
[08:58] <ubot3> Malone bug 538648 in xserver-xorg-video-intel "[gm45] Irregular sync flashes with compiz on (Lenovo T500)" [Medium,Triaged] https://launchpad.net/bugs/538648
[09:01] <apw> the bug there does not have the upstream infor on powersave in it does it?
[09:02] <apw> hrm thats oddd, its longer now, much longer ... launchpad what are you up to?
[09:03] <apw> RAOF, so in short if they need powersave = 0 we arn't expecting to need to fix them
[09:04] <RAOF> Because the symptoms are insufficiently bad?
[09:04] <apw> RAOF, no simply because we are unaware the issues still exist and have no bugs open on linux (that i am aware of) to change it, and no patch out there for testing
[09:05] <RAOF> Ok.
[09:05] <apw> final freeze kernels close today ... so i don't know if we can get it in before release
[09:05] <apw> as its not stopping them booting it can easily be an SRU after release of course
[09:05] <RAOF> Yeah.
[09:06] <apw> and indeed has a simple work around, which we could releasenote
[09:20] <RAOF> apw: Will the nouveau acceleration disabling patch make it into lucid?  Are you waiting for anything from me for that?  Splitting it into the components looks like an excelent idea.
[09:20] <RAOF> https://lists.ubuntu.com/archives/kernel-team/2010-April/009884.html and its follow ups.
[09:21] <apw> RAOF, i already split it up and pushed it back for review.  i hope it has all the required ack and 
[09:21] <apw> am intending on merging it today ... for the final kernel
[09:21] <apw> its all a bit behind due to yesterdays thinkpad debackle
[09:21] <RAOF> Yeah, that was mad.
[09:22] <apw> we tend to get at least one a cycle ... its hard to avoid
[09:22] <apw> i am very glad it was this week and not next
[09:22] <RAOF> OOoooh, yes.
[09:23] <apw> and in part was why i pushed half the update last friday instead ..
[09:23] <RAOF> Can I add an extra quirk - bug #542950 would like GeForce 6100 cards (pciid 0242) quirked.
[09:23] <ubot3> Malone bug 542950 in linux "[Needs noaccel quirk] Screen corruption on startup" [High,Confirmed] https://launchpad.net/bugs/542950
[09:30] <apw> RAOF, do we have a patch for that?
[09:30] <RAOF> We don't; I can generate one easily though.
[09:31] <apw> can you test the final patch ?
[09:31] <RAOF> I cannot.  However, setting nouveau.noaccel=0 has been reported to work by multiple people.
[09:32] <apw> thats tricky ... after yesterday i am nuvous of breaking people ... get the patch out ... and we can look it over.  i am sure it'll be 'ovvious'
[09:32] <apw> if you get it out i presume on top of the other one (that was nouveau right)
[09:32] <apw> then i can make a kernel and ask for testing ... and we can make a decision later today
[09:34] <RAOF> It'll be an extra patch on top of your existing 3 in the series.  Is there an easy way to apply them on ubuntu-lucid?
[09:42] <apw> save them into a mailbox and git am the mailbox
[10:47]  * apw wonders if akgraner tested that kernel
[10:48] <apw> i'd like to get that patch into the kernel and some testing feedback would help
[10:55] <zimbatm> hi there
[11:11] <apw> hi
[11:15] <jjohansen> apw: is it to late for include a driver in the initrd
[11:15] <jjohansen> Bug #530361
[11:15] <ubot3> Malone bug 530361 in linux "Support for DELL H200/H700/H800 SAS cards is missing in the CD/DVD installer for Lucid (mpt2sas)" [Undecided,New] https://launchpad.net/bugs/530361
[11:17] <apw> jjohansen, you know just how late it is in the cycle!
[11:17] <apw> as its a CD driver for a dell server we might be able to do something
[11:17] <apw> as they are likely just hosed without it
[11:17] <apw> however, that is predicated on it being a simple fix
[11:17] <apw> if its 'whole replacement driver' its not going to make the release images imo
[11:18] <jjohansen> apw: yes, and I expect no but I just got the Q in my inbox and figured I'd ask
[11:18] <apw> looking at the bug i don't see anything other than 'its broken' so its hard to evaluate whether there is a possibility or not
[11:19] <apw> finally it may just be that the driver is not in the udebs
[11:19] <apw> jjohansen, do we have someone who has one who can answer some of these questions?
[11:19] <jjohansen> from the bug that we actually ship the driver its just not in the initrd
[11:20] <jjohansen> not that I know of but I'll poke and see
[11:22] <apw> it does appear that we have the driver building
[11:22] <apw> debian.master/config/config.common.ubuntu:CONFIG_SCSI_MPT2SAS=m
[11:22] <apw> obj-$(CONFIG_SCSI_MPT2SAS) += mpt2sas.o
[11:23] <apw> it appears to be called mpt2sas
[11:23] <apw> and we don't appear to put it in the installer stuff
[11:23] <jjohansen> yep, there are reports of using it to "fix" the problem, its just not in the initrd
[11:23] <jjohansen> right, or the installer
[11:23] <apw> so if that the only issue then it is fixable i'd say ... but why the hell can't people ask last week not this
[11:24] <apw> jjohansen, so who asked
[11:25] <jjohansen> ttx
[11:25] <apw> ttx should know better
[11:25] <apw> i even sent out an email last week saying the deadline was here
[11:26] <jjohansen> right, but everyone else is focused on a thursday deadline, so that gets forgotten
[11:28] <apw> the thurday deadline doesn't mean thursday for _anyone_ really
[11:28]  * apw needs more brown brain fuel
[13:16] <mirsal> Hello there
[13:21] <apw> hello
[13:30] <mirsal> It's probably not the right place to ask for help but I'm trying to build NetworkManager from the upstream source tree and I'm having trouble with the wireless-tools
[13:30] <mirsal> http://pastebin.com/rmUedvqs
[13:36] <apw> mirsal, that normall implies there is a -dev package
[13:36] <apw> you probabally want to ask about that on #nm on this server
[13:37] <mirsal> apw, yeah, usually there is, but I can't find it
[13:37] <mirsal> apw, I just asked there. No reply so far :)
[13:37]  * apw swears ... that updated EC fix works for dells and doesn't break thinkpads ... but still breaks damn macs
[13:38] <mirsal> (your last two words summarize an awful lot of issues)
[13:39] <JFo> apw, very odd
[13:39] <mirsal> apw, ah ! found it, thanks
[13:39] <mirsal> libiw-dev - Wireless tools - development files
[13:40] <mirsal> go figure ;x
[13:40] <apw> mirsal, 'so' obvious
[13:46] <mirsal> hmm, why I did not think about apt-get build-dep is another story...
[13:53] <abogani> smb: Hi Stefan! I'm wondering if you could sponsor me (for PPU right on linux-rt package https://wiki.ubuntu.com/AlessioIgorBogani/linux-rtPPUApplication) adding your comments about Feisty-Gutsy-Hardy period...
[13:55] <smb> abogani, I guess I can, but you better mail me a reminder to look at it. I am currently at the airport and I'd rather do it with a bit more time later
[13:55] <abogani> smb: Of course, thanks!
[15:34] <achiang> hm, is there a way to get dev_dbg() statements in usb to display on console?
[15:35] <achiang> maybe just pass 'debug' on the kernel command line?
[15:36] <apw> achiang, do they not appear in dmesg output?
[15:37] <apw> if they are in there you probabally can get them on the console direct via loglevel, if i've understood
[15:37] <achiang> apw: haven't confirmed, i'm just doing some drive-by triaging on https://bugs.launchpad.net/ubuntu/lucid/+source/linux/+bug/433438
[15:37] <ubot3> Malone bug 433438 in linux "device descriptor read/64, error -110" [Medium,Confirmed] 
[15:38] <achiang> apw: in my 5 minutes of reading, that message is due to an urb timeout of some sort
[15:38] <achiang> so i was just going to suggest something that the bug reporters could do to get some more info before someone who's really on the kernel team can take a look
[15:39] <apw> soz can't be more authorative on that 
[15:41] <achiang> "soz" ?
[15:54] <apw> short for sorry
[15:54] <achiang> heh, thanks
[16:48] <kaouthia> Hello all. Amitk suggested it would be a good idea to say hello to you all and introduce myself. My name is Lee Jones. I have recently accepted a role within Canonical as an Arm Kernel Engineer, with a view to start at the beginning of May.
[16:53] <ogra> kaouthia, awesome ! i see you are also in #ubuntu-arm already :)
[16:54] <kaouthia> Yes, and ubuntu-meeting (a bit too keen me thinks) ;)
[16:58] <kaouthia> I would like to pick your brains if I may. I have to purchase a laptop and I’m finding it difficult to choose.  I would like something which compiles Kernels quickly and is going to last. Do any of you have any suggestions you’d like to make? I hear a lot of you like ThinkPads - anything in particular?
[16:59]  * ogra doesnt lkie thinkpads ... i tend to buy whitebox systems to be able to replace parts if they get outdated ... but then i'm not a kernel guy i just happen to hang around here to annoy apw for ARM stuff :)
[16:59] <ogra> *like
[17:00] <achiang> kaouthia: building kernels quickly requires lots of cores and fast I/O, so something with Core i7 and an SSD would be good
[17:02] <apw> generally i don't build on my laptop, the disk is just too slow
[17:02] <achiang> your best bet is probably getting a fast desktop and a cheap laptop for travel, actually
[17:02] <apw> yeah i have a "desktop replacement" latop i use as my 'home machine' a 10" laptop for travel, and use remote machines for build generally
[17:03] <abogani> apw: distcc?
[17:03] <achiang> fast local CPUs + fast I/O + ccache generally outperforms distcc, at least in my experience
[17:04] <apw> abogani, nothing so sophisticated, just porters really
[17:04] <abogani> apw: right
[17:05] <abogani> achiang: Obviously ccache and distcc have different goals so I don't think that you can compare these.
[17:05] <apw> i am more of a ccache user myself
[17:06] <apw> not that i see a whole heap of difference to be fran
[17:06] <apw> frank
[17:09] <kaouthia> I already have a Sony laptop, which I really like, but I can't put Linux on it. It is the only Windows based machine in the house now. If I put Linux on it, the girlfriend will BSOD.
[17:10] <apw> kaouthia, heh this is the time of equal rights, she can buy her own windows laptop
[17:12] <kaouthia> That could be a cheap alt - I'll buy her a cheap laptop to surf the web and I'll have the Sony back - great suggestion!
[17:13] <apw> kaouthia, now you are thinking
[17:13] <apw> if you have a desktop and the sony u are likely sorted
[17:13] <apw> though i suspect none of your buttons will work ... you can fix that :)
[17:19] <kaouthia> Buttons?
[17:19] <apw> brightness up/down volume up/down eject, that sort of button
[17:19] <apw> hotkeys is prolly a better description
[17:19] <kaouthia> Couldn't possibly, it doesn't have an Arm Core ;)
[17:21] <ogra> nothing a hammer and soldering iron couldnt fix :)
[17:22] <kaouthia> What would you want to solder a hammer to? ;)
[17:22] <ogra> the old CPU to get it ripped out more easily :)
[17:23] <mjg59> Hotkeys ought to work by default on all Sonys now
[17:24] <apw> mjg59, heh that'd be nice
[17:24] <kaouthia> Nothing a few extra volts and a cup of liquid nitrogen wouldn't fix
[17:25] <mjg59> Support for the new-style systems went in a year ago
[17:28] <apw> mjg59, yeah i admit i've not heard any whining about them for a while :)
[17:29] <mjg59> apw: The fun bit was writing it without any of the actual hardware
[17:29] <apw> qudos for that for sure
[18:39] <lamont> I used to be able to see my SD card (karmic, inspiron 1520).  Now it's not there...  wtf?
[18:40] <jjohansen> lamont: what kernel
[18:41] <lamont> 2.6.32-17-generic
[18:41] <lamont> though I guess I could reboot into -19
[18:42] <jjohansen> lamont: -20 is the current kernel
[18:43] <lamont> meh
=== yofel_ is now known as yofel
[19:21] <smoser> jjohansen, or anyone, can i disable the function-f2 (disable wireless) button easily ?
[19:21] <jjohansen> smoser: depends
[19:22] <jjohansen> smoser: on some machines its a key press the kernel sees and triggers events, on others the kernel doesn't even see it
[19:22] <smoser> well this one sees it
[19:23] <smoser> it probably gets routed to gnome-power-manager, right ?
[19:24] <jjohansen> smoser: well does it see the keypress or just propagate an event after the fact
[19:24] <smoser> the event goes through and my wirelss drops
[19:26] <jjohansen> smoser: right, but there are a couple ways it can go through, and that will determine whether the kernel see it early enough to stop it
[19:28] <jjohansen> smoser: basic key debugging is start with xev, see it shows up there, then move to show keys and then acpi_listen
[19:28] <jjohansen> but it also could be handled by the ec
[19:30] <smoser> ok. i'll poke at it sometime
[19:30] <smoser> but thanks. i was hoping something easy. i keep accidently hitting it
[19:38] <anoteng> Anybody wanna take a look at Bug #183461, it's not fully triaged yet, but I think it should have importance high as it makes a default installation unusable for some users. This is a very old bug, and I think it could use a pair of kernel-eyes by now.
[19:38] <ubot3> Malone bug 183461 in opensuse "ksoftirqd/1 using 50% CPU or above" [Unknown,In progress] https://launchpad.net/bugs/183461
[19:40] <mjg59> anoteng: I think that's several dozen different bugs
[19:42] <anoteng> I was thinking the same to, should I make the reporters file separate bugs?
[19:46] <JFo> anoteng, there are several thousand old bugs that need a look. it is just not currently possible to look at them all
[19:46] <JFo> but yes, the reporter needs to file separate bugs 
[19:46] <JFo> as much as it pains me to say it :)
[19:47] <JFo> I will put this one on my list for review
[19:53] <anoteng> thanks. I asked everybody to file their own bugs with apport-bug -p linux.
=== amitk is now known as amitk-afk