[04:10] <MrNOKIA> hello again :)
[04:14] <MrNOKIA> i'm a little bit puzzled: could anynone please explain the "your processor does not support long mode, use a 32bit version" error ? i got it twice, once on gentoo 2007 and also on feisty fawn, when trying to boot a dell inspiron 6400 centrino dual core box
[04:16] <MrNOKIA> what is more strange, is that on a dell inspiron 6400 intel dual core 2 the error doesn;t appear
[04:17] <MrNOKIA> both distros, however, are puzzled by the ATi x1400 vga, which is very strange, at least from ubuntu point of view: both dapper and edgy detected it flawlessly
[04:17] <MrNOKIA> but not feisty...
[05:01] <crimsun> MrNOKIA: tell me the output from `grep ^flags /proc/cpuinfo |uniq` in #ubuntu, please.
[05:45] <johanbr> crimsun: Don't know if you remember, but I asked you about my USB phone a few days ago. I couldn't get audio over USB to work, so I tried bluetooth instead and I can reliably get the Feisty kernel to panic.
[05:46] <crimsun> johanbr: snd-bt-sco or a2dpd plugin?
[05:46] <johanbr> crimsun: snd-bt-sco
[05:46] <crimsun> johanbr: the latter is preferred.
[05:47] <johanbr> I didn't realize they were functionally equivalent. I'll give a2dpd a try, thank you.
[05:49] <BenC> crimsun: where is a2dpd?
[05:50] <crimsun> BenC: alsa-lib plugin distributed by bluetooth-alsa.sf.net.
[05:51] <Amaranth> that's an odd name for a module
[05:51] <BenC> crimsun: should I ditch snd-bt-sco?
[05:51] <crimsun> BenC: I can't say at this point (I don't have enough feedback to provide an answer).
[05:52] <johanbr> I assume that then you'd have to package bluetooth-alsa?
[05:53] <crimsun> johanbr: adding the plugin to our alsa-lib source is not difficult.
[05:55] <crimsun> A lot of people are having problems with snd-bt-sco, but that's with older code.  I don't know offhand (no local equipment) if updating it resolves the issues.
[05:56] <crimsun> johanbr: I'm trying to hold off updating the alsa-* source until 1.0.14 final is released, since there are so many fixes.
[05:59] <johanbr> crimsun: I see. I should have some more time in a week or so, then I'll do some experimenting with a2dpd and maybe a newer snd-bt-sco. I'll let you know if I find anything interesting.
[06:00] <crimsun> johanbr: many thanks.
[06:00] <johanbr> I'm happy to contribute what little I can. :)
[06:00] <johanbr> Well, gotta go...
[09:31] <MrNOKIA> thanks crimsun, meanwhile i found out with stupefaction that intel's centrino duo hasn't x64 support
[09:34] <crimsun> correct.
[09:34] <Mithrandir> duo 2 has
[09:34] <Mithrandir> duo 1 does not.
[09:35] <crimsun> gotta love core duos.
[09:39] <MrNOKIA> only trouble is, the dell with centrino is mine, the dell with core 2 duo is not :(
[09:40] <MrNOKIA> speaking of core 2 duo inspiron, it has 2 problems, one being very strange
[09:41] <MrNOKIA> first of all, it seems to lack support for the BCM... bluetooth card
[09:41] <crimsun> it would be useful for the kernel team if you can point to a detailed wiki page with the components, dmesg, and lspci -nvv .
[09:41] <MrNOKIA> on the other hand, the ATI x1400 doesn't work with official ati drivers
[09:42] <MrNOKIA> pastebin is ok ?
[09:43] <MrNOKIA> it would be best to have a comparison between dell inspiron 6400 centrino duo and dell inspiron 6400 core 2 duo
[09:43] <MrNOKIA> they are 95% similar
[09:44] <MrNOKIA> and i have them both right now on feisty and feisty_x64
[09:48] <crimsun> MrNOKIA: a more ... permanent page on the Ubuntu wiki under the LaptopTestingTeam hierarchy is arguably more helpful.
[09:51] <MrNOKIA> ok, let me see what i can do. never done that before :)
[10:43] <zdzichuBG> is it safe to install gutsy kernel on feisty?
[10:47] <crimsun> I don't recommend it, but some people report it works dandy.  And I presume you mean 2.6.22-2, because the "gutsy kernel" still is feisty's 2.6.20-15.
[11:01] <zdzichuBG> crimsun: thanks. And yes, I mean latest  upload, with dyntics, timerstats and other good stuff.
[02:04] <zul> hey
[02:50] <elmarco> FATAL: modpost: GPL-incompatible module fglrx.ko uses GPL-only symbol 'pagefault_disable'
[02:50] <elmarco> any idea what I can do to override this for private use?
[02:54] <mjg59> Change EXPORT_SYMBOL_GPL(pagefault_disable) to EXPORT_SYMBOL?
[02:59] <zul> BenC: is there a meeting tomorrow?
[05:14] <DBO> is there anyone on the kernel team that wouldn't mind answering a couple quick questions?
[05:16] <mjg59> Depends how quick
[05:16] <mjg59> But go ahead :)
[05:17] <DBO> are you guys aware of the considerable stability issues many people are having with the feisty kernel?  (i dont mean to be offensive)
[05:17] <DBO> also, thank you =)
[05:18] <DBO> and if so can we expect a fix?
[05:18] <mjg59> If there are bugs filed, then we're aware of them
[05:18] <DBO> the problem is nobody knows how to file this one
[05:18] <DBO> its just... "it hangs while in X"
[05:18] <mjg59> Well, it's clearly not generic
[05:18] <mjg59> So there'll be some degree of commonality
[05:19] <DBO> there is a large thread on the forums about it
[05:19] <DBO> nobody seems to be able to find it...
[05:19] <DBO> nvidia, ati, intel all find it (so not video card)
[05:19] <mjg59> If there's no commonality across all the reports, then it's unlikely to be the same bug
[05:19] <mjg59> To a first approximation, computers are deterministic
[05:20] <DBO> some people claim its the ralink, but I dont have that chipset... others find the -386 kernel
[05:20] <DBO> thats the other weird thing
[05:20] <DBO> there doesnt seem to be an obvious trigger for it
[05:20] <DBO> I cant trigger it with any degree of consistency, nor can anyone else.  It just happens everywhere from 5 minutes to 6 hours after boot
[05:21] <DBO> =/
[05:21] <DBO> Im just worried it wont get fixed, so I figured I would come and ask
[05:21] <mjg59> The only thing I can suggest is hooking up a serial console
[05:22] <DBO> *looking to see if I got one*
[05:23] <DBO> =/
[05:23] <DBO> nope
[05:23] <zdzichuBG> BTW, when is feisty kernel update planned?
[05:23] <mjg59> No clue. Ben should know.
[05:24] <zdzichuBG> thanks
[05:24] <DBO> mjg59, is there any other way I can get the message from the kernel panic?
[05:24] <DBO> I want to help... I'd mail you the machine if I could afford to lose it
[05:24] <mjg59> DBO: Netconsole may be another option
[05:25] <DBO> mjg59, would you like a link to the ubuntu forums thread on the matter so you can see what people are saying?
[05:26] <mjg59> DBO: If you've summarised it recently, I'm afraid it's unlikely to be much help
[05:27] <DBO> =/
[05:28] <DBO> is there anyone else I should contact, or will filing an admittedly generic launchpad bug report be helpful?
[05:30] <mjg59> Without any diagnostic information, without any indication as to what's triggering the problem and without any common drivers, then I'm afraid it's unlikely to be helpful
[05:31] <DBO> last one then, would the gutsy kernel be worth a try? (provided I can swing getting the packaging to work without doing a full upgrade to gutsy)
[05:33] <mjg59> It's certainly worth a go
[05:33] <mariah>  I'm debugging a problem I've had with my Sony Laptop and the generic kernel. The machine runs really slow just after finishing initrd.
[05:34] <mariah>  The i386 kernel doesn't do this.
[05:37] <maks_> wow colors
[05:39] <mariah>  Anyone familiar with initrd process in Ubuntu?
[05:40] <maks_> maybe 
[05:40] <maks_> mariah: this a dev channel the support is in #ubuntu afaik
[05:41] <maks_> unless you are running latest g thing you should ask there
[05:41] <maks_> also colors put people off usually :P
[05:41] <mariah>  for kernel problems as well
[05:41] <maks_> _support_
[05:42] <mariah>  only if your english lol
[05:43] <racarr> I keep on thinking you are pinging me. And wonder why I would be pinged in #ubuntu-kernel. Red color == bad choice.
[05:43] <mariah>  especially from the north!
[05:43] <Mithrandir> mariah: Please turn off colours.
[05:47] <stgraber> racarr: switch back to a classical hilight color :), say yellow :)
[05:48] <racarr> Well. It actually is yellow. But I used to use Konversation, which was red.
[05:52] <mariah> No help there just a lot of noise
[05:52] <mariah> So anyone here wish to assist me. This is a dev level issue.
[05:53] <mariah> I'm probably going to need to modify the initrc.img
[05:53] <mariah> Plus I turned of colors
[05:54] <mjg59> mariah: Is there a bug number?
[05:54] <mariah> I need to identify it as a bug 1st.
[05:54] <mariah> You wish me to report something unvalidated?
[05:55] <mjg59> If you install Ubuntu, does it work properly? If yes, it's not a bug. If it doesn't, it's a bug.
[05:55] <BenC> part of a bug is the validation process
[05:55] <mariah> I would like to be sure first rather then wasting others/my own time.
[05:56] <mjg59> mariah: Dealing with a bug takes exactly the same amount of time as asking people here
[05:56] <mariah> It isn't live
[05:56] <mariah> Or is there another irc ch I'm not familiar with?
[05:57] <mariah> normal ubuntu channel is very noisy right now. Lots of flooding going on.
[05:57] <BenC> mariah: bug report
[05:57] <mjg59> mariah: If the distribution doesn't work, then just file a bug report
[05:59] <mariah> I can see this is going nowhere.
[05:59] <mariah> Ok. Enjoy your talk about how you used Konversation which was red. I'm sure that is a dev issue.
[06:00] <mjg59> mariah: If you haven't filed a bug, then I'll agree
[06:02] <maks_> she left
[06:03] <maks_> i guess ide-generic does nothing good for her
[06:03] <maks_> but that is a wild guess
[06:16] <BenC> rtg_, mjg59: I'm about to upload a -rc1 based kernel for gutsy with timer stats enabled
[06:20] <maks_> BenC hmm i just wanted to check if fedora has timer stats enabled
[06:20] <BenC> unless they have 2.6.21+, they wont
[06:20] <BenC> have to have NO_HZ enabled too
[06:22] <maks_> git-iwlwifi.patch
[06:23] <maks_> sure they are 2.6.21 based, they always almost track latest
[06:25] <maks_> yup enabled on -generic
[06:30] <mjg59> BenC: Thanks
[06:30] <mjg59> BenC: They do
[06:31] <mjg59> Have NO_HZ, that is
[06:31] <BenC> just NO_HZ alone improved my battery life noticably
[06:31] <BenC> mjg59: is that ipw2200 patch upstream?
[06:32] <mjg59> BenC: No
[06:32] <mjg59> Also entirely untested
[06:32] <BenC> ah, the powertop site seemed to make it sound golden :)
[06:32] <mjg59> Haha
[06:32] <mjg59> The ipw2100 one is tested
[06:32] <mjg59> The 2200 one should be equivalent - Arjan was going to test it today
[06:32] <BenC> ah, it is 2100
[06:33] <mjg59> Oh, right
[06:33] <mjg59> Ok
[06:33] <mjg59> The only thing that should be done with the 2100 one is to release the IRQ on close and allocate it on open
[06:33] <mjg59> Shouldn't make any difference
[06:33] <mjg59> But slightly tidier
[06:34] <mjg59> Also, there's a tidier version of the appletouch patch
[06:35] <mjg59> http://www.codon.org.uk/~mjg59/tmp/appletouch.diff
[06:35] <BenC> mjg59: would make my life a lot easier if these got sent to linus
[06:35] <mjg59> No point in sending the ipw stuff until the Intel people ack them
[06:36] <mjg59> Should get some progress with that today
[06:36] <mjg59> appletouch has been sent to -input
[06:36] <mjg59> Oh, and lkml
[06:36] <BenC> I just today tainted my tree with source patches, but only to get rc1 out :)
[06:36] <BenC> couple of one-liners for sparc
[06:37] <mjg59> There's no point in me sending these direct to Linus, as far as I can tell - they're all going to have to go through the maintainer trees
[06:38] <BenC> which probably means we've missed the window
[06:38] <mjg59> Yeah
[06:38] <rtg_> BenC: 2.6.22 is closed.
[06:38] <BenC> well closed for major syncs, but not for fixes
[06:38] <rtg_> right.
[06:38] <BenC> IMO, some of these can be considered fixes
[06:39] <rtg_> Since when does our opinion count :)
[06:39] <BenC> saving my Li lifetime is a fix, IMO :)
[06:39] <mjg59> Anyone here have a late-generation G4 Powerbook?
[06:39] <mjg59> The ones that have USB touchpads
[06:39] <Mithrandir> mjg59: cjwatson maybe?
[06:39] <BenC> I have a 17: pb
[06:39] <BenC> 17" I mean
[06:39] <mjg59> BenC: USB input devices?
[06:40] <BenC> I think, but I'd have to check
[06:40] <mjg59> BenC: If so, I could do with you testing something :)
[06:40] <mjg59> Basically, I need to know whether the first touch on the touchpad sends a button down event
[06:40] <BenC> my wife has to go to a doctor appointment later, so there's a small point of opportunity to get it and install a small feisty partition :)
[06:41] <BenC> fabbione: ping
[06:41] <mjg59> Oh, and also spend some time figuring out whether it has the same behaviour as the geyser 3 does
[06:41] <BenC> fabbione has the same PB I have
[06:42] <mjg59> I'll sort a quick diff
[06:48] <zul_> BenC: evil..my wife would kill me
[06:49] <maks_> zul_ did you ever try?
[06:50] <BenC> mjg59: are you going to take care of the hit list on linuxpowertop.org for the apps we have in gutsy?
[06:50] <BenC> lot of patches we could probably user, I need to do the one I did for gaim against pidgin
[06:53] <BenC> time to reboot
[07:27] <BenC> uhci/ehci and ipw3945 are killing my battery :/
[07:28] <zdzichuBG> usb is bad, but ipw?
[07:28] <BenC> that's on an idle system
[07:29] <BenC> probably network-manager doing scans
[07:29] <zdzichuBG> or regulatory daemon doing voodoo
[07:29] <zdzichuBG> maybe iwlwifi will be nicer
[07:29] <mjg59> If you have a wireless interface up, it'll spew stuff for an extended period of time
[07:30] <BenC> it's not reg daemon, it's definitely network-manager making it do bad stuff
[07:30] <rtg_> How about the signal strength monitor?
[07:30] <BenC> rtg_: that's network-daemon
[07:30] <BenC> err, network-manager
[07:31] <BenC>   42.4% (100.0)       <interrupt> : uhci_hcd:usb1, ehci_hcd:usb5 
[07:31] <BenC>   12.5% (29.4)       <interrupt> : ipw3945 
[07:31] <BenC>    5.3% (12.6)      S20powernowd : queue_delayed_work_on (delayed_work_timer_fn) 
[07:31] <BenC> I have no idea what the S20powernowd thing is
[07:31] <mjg59> BenC: What hardware is this?
[07:31] <mjg59> BenC: powernowd is due to ondemand waking up too often
[07:31] <Mithrandir> S20powernowd is the cpufreq daemon
[07:31] <mjg59> That's /supposed/ to be fixed in git
[07:31] <BenC> mjg59: Dell XPS 1710, core2duo
[07:31] <mjg59> BenC: Bluetooth?
[07:32] <BenC> mjg59: yeah
[07:32] <mjg59> If so, hciconfig hci0 down
[07:32] <BenC> mjg59: wow, that killed half the wakeups, so most of uhci/ehci
[07:33] <BenC> it's not even in my top 10 list anymore
[07:33] <mjg59> We need to improve bluetooth in that respect
[07:33] <BenC> definitely
[07:35] <BenC> 40% of my wakeups come from ipw3945 now
[07:35] <BenC> ondemand, mixer_applet2 and firefox seem to be the worst offenders now
[07:35] <BenC> besides 3945 that is
[07:36] <mjg59> I've uploaded fixes for mixer_applet2
[07:36] <maks_>  20.8%        <interrupt> : i8042
[07:36] <maks_> what's that biest?
[07:36] <mjg59> maks_: You're moving the mouse or typing
[07:36] <maks_> thanks mjg59
[07:37] <mjg59> i8042 is the keyboard/mouse controller for ps/2 systems
[07:37] <mjg59> BenC: How recently did you pull from Linus?
[07:37] <maks_> update-notifier is also always shown
[07:37] <BenC> this is -rc1
[07:37] <maks_> 4.8%    update-notifier : schedule_timeout (process_timeout)
[07:37] <mjg59> BenC: Hm.
[07:38] <BenC> maks_: on a fresh boot, that probably means it is downloading updates
[07:38] <BenC> or checking for them
[07:38] <maks_> fresh boot since 30 min
[07:38] <maks_> 0.42.12-3
[07:40] <fabbione> BenC: pong
[07:41] <BenC> fabbione: is your touchpad on your PB connected via USB?
[07:41] <fabbione> BenC: yes like your
[07:43] <mjg59> fabbione: Any chance you can test something for me at some point?
[07:43] <fabbione> mjg59: can it be done on a LiveCD?
[07:43] <mjg59> fabbione: No
[07:43] <mjg59> (Well, yes, but I'll need you to build a kernel as well, so not really)
[07:44] <fabbione> mjg59: then no.. sorry.. i removed Linux from this machine.
[07:44] <mjg59> No problem
[07:44] <BenC> guess I can test it later
[07:44] <mjg59> Ok
[07:44] <mjg59> I'll generate a diff shortly
[07:47] <BenC> fabbione: thanks anyway fabio
[07:49] <BenC> mjg59: have you done the mixer_applet2 10-HZ patch already?
[07:50] <mjg59> BenC: 10-HZ?
[07:50] <BenC> it has a 10hz timer
[07:50] <mjg59> Oh, yes, that's gone
[07:50] <BenC> guess I should selectively pull in some gutsy stuff
[07:51] <mjg59> 2.18.0-3ubuntu2 and the latest gstreamer0.10-plugins-base
[07:52] <fabbione> BenC: no problem at all
[08:12] <BenC> -3ubuntu2 doesn't appear yet...but I did update ipw3945, so I'll check later if that helps
[08:13] <mjg59> BenC: Yeah, gucharmap is fucked and has broken the build
[08:13] <BenC> libnss appears kind of fucked too
[08:14] <BenC> yay for merge weeks :)
[08:14] <BenC> powertop is my new best friend though
[08:15] <BenC> mjg59: Have you met our new guy Amit yet?
[08:15] <mjg59> Nope
[08:15] <BenC> he was at UDS, wont be starting for another month though
[08:15] <rtg_> How do you get S20powernowd to slow down? Was that supposed to be fixed in -rc1?
[08:15] <mjg59> rtg_: Just chatted to the cpufreq people
[08:15] <BenC> he's got crap loads of power savings experience
[08:15] <mjg59> rtg_: It's effectively a false positive
[08:15] <mjg59> BenC: Surname?
[08:16] <mjg59> rtg_: ondemand will opportunistically schedule whenever another interrupt appears
[08:16] <BenC> Kucheria, but I may have mispelled it
[08:16] <Amaranth> does the -2.7 kernel have the options on powertop needs or do i have to build myself?
[08:16] <mjg59> Looks like he's done wireless hacking as well?
[08:17] <BenC> mjg59: appears so
[08:17] <mjg59> Sweet
[08:18] <BenC> your search probably shows where he came from :)
[08:18] <mjg59> Haha
[08:20] <BenC> he met Keith(?) at UDS, and they are working on a power savings spec for gutsy
[08:21] <BenC> revolves around this powertop thing, and getting patches in for stuff like gnome to aggregate timers (there's a kde patch that does the same there)
[08:21] <mjg59> I'm fixing stuff that pops up here
[08:22] <BenC> basically does the same thing for gnome that round_jiffies() does for the kernel
[08:22] <mjg59> Yeah
[08:22] <mjg59> I think glib support for that exists
[08:23] <BenC>    3.3% ( 8.0)          modprobe : usb_hcd_poll_rh_status (rh_timer_func) 
[08:23] <BenC> any idea what the hell that is?
[08:23] <mjg59> USB host controller
[08:23] <mjg59> Got any USB devices?
[08:23] <mjg59> If so, and if they don't support runtime suspending, there'll be some housekeeping interrupts
[08:24] <BenC> hard to say...I thought my keyboard and touchpad were...then I also have a usb mouse connected
[08:25] <BenC> weird that it shows as modprobe
[08:25] <mjg59> If it's a module, it'll show as whatever userspace task started it
[08:25] <BenC> ah, gotcha
[08:25] <mjg59> And yeah, if you have a USB mouse that'll do it
[08:26] <mjg59> There should be some patches for runtime HID suspend appearing soon
[08:26] <BenC> dhcdbd still has a 5hz timer
[08:26] <mjg59> Yeah, need to chase Keith about a patch for that
[08:53] <BenC> brb
[09:15] <BenC> ok, ipw3945 is just horrible
[09:16] <BenC> just loading the driver causes 16 wakeups/sec
[09:16] <BenC> up/down leaves it doing 50-70/sec
[09:16] <mjg59> I've just uploaded dhcdpd
[09:17] <BenC> wonder why 3945 would need to do 16 interrupts/second when the iface hasn't even been brought up
[09:17] <mjg59> Wouldn't surprise me if it's chatting to ipw3945d
[09:18] <BenC> which would make it unfixable at this point :/
[09:18] <BenC> should test iwlwifi see if that has the same issue
[09:19] <BenC> I actually had it working at UDS, but always showed 100% for strength, and broke suspend
[09:19] <mjg59> BenC: If you could test http://www.codon.org.uk/~mjg59/tmp/appletouch.diag.diff, that would be great
[09:20] <mjg59> Two things I want to know: (1) does the first time you touch (not tap) the trackpad result in it printing an "Appletouch: button" message, and (2) when not touching it, does it generate "Received a packet" messages constantly?
[09:20] <mjg59> The first time you touch being the very first time after loading the driver, that is
[09:30] <BenC> mjg59: Ok, I'll let you know how it goes
[09:40] <mjg59> BenC: Where does the ubuntu-modules tree live?
[09:40] <BenC> until we get kernel.ubuntu.com, it's on rookery
[09:41] <mjg59> Is it a public URL/
[09:41] <mjg59> ?
[09:41] <BenC> no, but I can tar it up real quick for you
[09:41] <mjg59> That'd be good
[09:42] <BenC> uploading to http://people.ubuntu.com/~bcollins/linux-ubuntu-modules-2.6.22.tar.gz
[09:42] <BenC> give it about 2 minutes
[09:43] <BenC> let me push ubuntu-gutsy.git while I'm at it
[09:44] <BenC> all ready
[09:46] <mjg59> Ta
[09:46] <mjg59> Just updating the 945 machine
[09:57] <mjg59> BenC: Have any of the 2.6.22-2 builds actually hit the archive?
[09:57] <BenC> doesn't look like it
[09:57] <mjg59> Ok, not just me then :)
[09:57] <BenC> but it's about to be obsolete
[09:58] <BenC> I'm about to upload -3 in less than an hour
[09:58] <BenC> this should get us successful builds and lum/lrm/linux-meta close behind
[09:58] <mjg59> Hey, I may finally have got gnome-applets uploaded successfully
[10:52] <BenC> cool, got my first real vmcore
[11:03] <mjg59> BenC: What's the right incantation to invoke the build with -j3?
[11:03] <BenC> CONCURRENCY_LEVEL=3 fakeroot debian/rules binary-generic
[11:03] <BenC> you can just export that too
[11:04] <mjg59> Ah, cool
[11:10] <Gadi> does anyone here know:  if a NIC module creates the NIC as "eth1_ifre" what is the problem?
[11:11] <Gadi> the module in question here is e1000, but I'm not sure if the module is changing the name or a udev rule
[11:11] <Gadi> as it appears that the device is unconfigurable
[11:17] <BenC> Gadi: udev and ifrename
[11:17] <BenC> Gadi: probably just want to edit /etc/iftab
[11:18] <Gadi> can I edit iftab without specifying a dev by MAC - I want to use this OS image cloned on multiple machines
[11:19] <stgraber> Gadi: you can keep it empty, this way the interface names will be the order of detection
[11:19] <Gadi> great
[11:19] <Gadi> thx