[01:20] <rtg> mjg59: Well, I suspect you know who gets to manage those expectations.
[01:20] <mjg59> Heh
[01:48] <bdmurray> uvcvideo is blocking my laptop from suspending
[01:51] <bdmurray> I receive an error message when that kernel module is loaded and I try to suspend with 2.6.22-11.
[02:22] <bdmurray> Actually, I don't understand why that module is being loaded on my system.
[02:23] <bdmurray> Whay does this "alias:          usb:v*p*d*dc*dsc*dp*ic0Eisc01ip00*" alias match?
[02:23] <bdmurray> s/Whay/What/
[02:30] <kylem> any device that claims to be usb video, subclass video control
[02:32] <bdmurray> kylem: So uvcvideo will loaded for any unknown webcam?
[02:32] <kylem> it's a match based on the capability of the usb device, not it's specific vendor/device ids.
[02:34] <mjg59> bdmurray: If it claims to be a USB video device, which most webcams don't
[02:34] <bdmurray> okay, but I don't understand why the module would be loaded for devices that are supported by the driver
[02:34] <bdmurray> maybe webcam was the wrong word, the camera in my laptop then
[02:34] <kylem> lsusb -v | egrep 'bInterface.*Class'
[02:35] <mjg59> Your camera presumably claims to implement the USB video class
[02:37] <bdmurray> Okay, so it might work with uvcvideo then?
[02:39] <mjg59> bdmurray: It claims that it should do. It might be lying, or uvcvideo might be broken.
[02:40] <bdmurray> mjg59: Okay, that module is preventing me from suspending though and if it matches a lot of devices it seems like it might be a problem for a lot of people then.
[02:40] <mjg59> bdmurray: Sounds like we need to fix the module. What are the symptoms?
[02:42] <bdmurray> When trying to suspend vi "echo -n mem > /sys/power/state" I receive a kernel erro message regarding "uvcvideo 5-4:1.1: suspend error -22"
[02:43] <mjg59> bdmurray: Uh. Why are you trying to suspend like that?
[02:43] <bdmurray> line 39 of /etc/acpi/sleep.sh is what was failing and it wasn't very informative
[02:44] <bdmurray> So I was originally using /etc/acpi/sleep.sh
[02:45] <mjg59> dmesg should give you the same information
[02:45] <mjg59> Hm. Interesting. I can't see any obvious reason why that would fail.
[02:46] <bdmurray> I think it worked before this commit - 616d7243b60bb3eb2abed6cf99fe4d3edfc6a874
[02:46] <mjg59> to lum or the main kernel?
[02:46] <bdmurray> lum
[02:47] <mjg59> Interesting.
[02:48] <bdmurray> my git knowledge isn't the greatest but I think that was the last commit changing uvcvideo
[02:49] <mjg59> If you revert that, does it work?
[02:49] <bdmurray> How could I do that?
[02:49] <mjg59> git revert 616d7243b60bb3eb2abed6cf99fe4d3edfc6a874
[02:50] <mjg59> Then try building l-u-m
[02:50] <mjg59> Are you sure it says "suspend error"?
[02:50] <mjg59> I can't find that string in the obvious bits of the kernel
[02:52] <bdmurray> [   83.816850]  usb_endpoint usbdev5.2_ep82: PM: suspend 0->2, parent 5-4:1.0 already 2
[02:52] <bdmurray> [   83.816863]  uvcvideo 5-4:1.1: suspend error -22
[02:52] <mjg59> Ah, ok, thanks
[02:53] <mjg59> Hm. Still not entirely obvious.
[02:59] <bdmurray> mjg59: 2.6.22-10 suspends if that is any help
[03:01] <mjg59> Is that before that commit?
[03:03] <bdmurray> Yeah it looks like
[03:04] <mjg59> Ah, hm. That diff added a suspend method.
[03:04] <mjg59> Oh, I see
[03:05] <mjg59> Right. Can you rmmod uvcvideo, then modprobe uvcvideo trace=255
[03:05] <mjg59> Then try a suspend and paste the full demsg?
[03:06] <bdmurray> sure, just a sec
[03:13] <bdmurray> Did I do something wrong or should have suspending worked?
[03:13] <lamont> kylem: you around?
[03:13] <lamont> or BenC?
[03:15] <mjg59> bdmurray: No, suspending probably shouldn't have worked
[03:17] <lamont> sigh
[03:17] <lamont> kylem: -12.36 will ftbfs on hppa without the patch I mailed to kernel-team.  sorry for not mailing it earlier when it was being discussed in channel.
[03:20] <bdmurray> mjg59: well, here is dmesg, I'll make sure it fails again too
[03:20] <bdmurray> http://pastebin.osuosl.org/2372
[03:24] <mjg59> bdmurray: Ah, I think I know what the issue is
[03:24] <mjg59> bdmurray: It's failing to initialise, but not releasing the device
[03:25] <mjg59> Hm. No, that /should/ release things correctly
[03:34] <bdmurray> Hrm, ocassionally the system becomes nonresponsive when suspending too.
[04:04] <lamont> bdmurray: mine turns almost completely off when I suspend it. :)
[08:25] <kraut> moin
[01:26] <verwilst> hi guys
[01:26] <verwilst> i was wondering, shouldn't there be a linux-image-xen or something?
[01:27] <verwilst> i apt-get install'ed linux-image-2.6.22-11-xen
[01:27] <verwilst> but apparently, when 2.6.22-12-xen is out, it doesn't automatically upgrade
[01:27] <verwilst> could it be because be installing it like that, it pinned itself to -11 ?
[01:30] <Chorus> there is a package named "linux-image-xen"
[01:30] <Chorus> http://packages.ubuntu.com/gutsy/metapackages/linux-image-xen
[01:41] <verwilst> aaah
[01:41] <verwilst> but not for amd64
[01:41] <verwilst> bugreport it?
[01:47] <zul> verwilst: its coming soon
[01:49] <verwilst> zul: coolnes :)
[01:50] <verwilst> just trying out xenman
[01:50] <verwilst> 0.5 on feisty doesn't seem to let me login to my server
[01:50] <verwilst> 0.6 doesn't seem to work on feisty, it cant find some modules :)
[01:50] <verwilst> python modules that is
[01:50] <zul> because xen-3.1 changed stuff
[01:50] <zul> try the one on gutsy
[01:51] <verwilst> yeah, i don't have a gutsy yet
[01:51] <verwilst> oh! my laptop!
[01:51] <verwilst> it scans for /usr/lib/xen-*, maybe if i just change the location, it'll work
[02:20] <verwilst> zul: apparmor seems to be disabled for the xen kernel
[02:20] <verwilst> any idea why?
[02:20] <zul> verwilst: because its not, i dont use it for one
[02:20] <verwilst> its not disabled you mean?
[02:21] <zul> verwilst: but but in a bug request and ill enable it this weekend
[02:21] <verwilst> thanks man :)
[02:26] <verwilst> zul: hm, it seems to be available in linux-ubuntu-modules-xen
[02:26] <verwilst> E: Couldn't find package linux-ubuntu-modules-2.6.22-12-xen
[02:26] <verwilst> but my mirror is acting up so it seems :p
[02:26] <zul> for amd64?
[02:28] <verwilst> in the changelog entry for ",linux-ubuntu-modules-2.6.22 (2.6.22-12.29)"  it says "Ben Collins: ubuntu: Add amd64 to xen target"
[02:28] <zul> could be your mirro 
[02:29] <verwilst> ftp://ftp.ubuntu.com/ubuntu/pool/universe/l/linux-ubuntu-modules-2.6.22
[02:29] <verwilst> idd
[02:29] <verwilst> the kernel has been sycned
[02:29] <verwilst> but the modules aren't yet
[02:30] <verwilst> xenman looks pretty nice btw
[02:44] <jetsaredim> did we ever get an answer on the vmware modules question?
[03:13] <verwilst> zul: hm, do you know of a bug that makes the bridge not being creates?
[03:13] <verwilst> created*
[03:13] <verwilst> i have bridge-utils installed
[03:14] <verwilst> and my xend-config.sxp has the network-bridge thing correct
[03:14] <verwilst> i have lots of vif's, but no bridge
[03:14] <zul> nope
[03:15] <verwilst> had that problem before with -10 i think
[03:15] <verwilst> booting with 2.6.16-xen worked fine
[03:15] <verwilst> so must be a kernel issue
[03:15] <verwilst> 2.6.18-xen sorry, the official one
[03:16] <zul> you might want to check #xen
[04:12] <Nafallo> Lenovo ThinkPad R61 (NA01FUK)
[04:12] <Nafallo> I hope its a good one, cause I have it early next week :-)
[04:43] <verwilst> zul: read everything on #xen? :)
[04:43] <zul> uh no
[04:43] <verwilst> zul: only the last lines are important :)
[05:01] <bdmurray> BenC: I'm having some issues with suspend and uvcvideo on my laptop.
[05:01] <BenC> bdmurray: yeah, I've heard the same from others
[05:01] <mjg59> bdmurray: Did you generate the full trace I asked for?
[05:01] <BenC> some people have added it to the list of modules to load/unload for suspend
[05:02] <mjg59> There's a trivial one line workaround
[05:03] <mjg59> Just change the return -EINVAL to return 0 in uvc_suspend()
[05:03] <mjg59> It's a sanity check that's failing
[05:03] <mjg59> I suspect because the driver never bound, so dev->video.streaming->intf never got set
[05:03] <mjg59> Quite /why/ the driver is still attached to the device when it failed to set up the video interface, I've no idea
[05:04] <zul> mjg59: http://svn.berlios.de/viewcvs/linux-uvc/linux-uvc/trunk/uvc_driver.c?rev=122&r1=120&r2=122
[05:04] <mjg59> zul: Yes, that's the code we have
[05:04] <zul> ah ok
[05:05] <bdmurray> When the module is initially loaded I do see a message about it failing to initialize the device
[05:05] <mjg59> bdmurray: The logs from an attempted suspend with the tracing turned on would be helpful
[05:06] <bdmurray> mjg59: okay, I'll keep trying to get one
[06:04] <natenewz> submitted bug #129972 this last summer, if somebody wants to take a look, I can test and run commands and things
[06:04] <ubotu> Launchpad bug 129972 in alsa-driver "no sound on hp dv6500 pavilion laptop" [Undecided,Confirmed]  https://launchpad.net/bugs/129972
[06:06] <bdmurray> natenewz: is that with gutsy or feisty?
[06:09] <natenewz> bdmurray: gutsy
[06:12] <bdmurray> natenewz: installed or a live cd?
[06:13] <natenewz> bdmurray: installed
[06:44] <jdong> what architectures are "tickless"?
[06:44] <jdong> just i386, or i386 and amd64?
[06:45] <jdong> sorry if I, err, ticked... anyone off by asking that in -devel
[06:45] <mjg59> Just 9386
[06:45] <mjg59> i386
[06:45] <jdong> ah, ok
[06:45] <jdong> thanks
[07:00] <bdmurray> Is there a reason PM_TRACE isn't available on amd64?
[07:02] <mjg59> bdmurray: It shouldn't make any difference in this case?
[07:03] <bdmurray> mjg59: The architecture shouldn't make any difference?
[07:04] <mjg59> bdmurray: Oh, sorry, are we not on uvcvideo any more?
[07:05] <mjg59> I believe rtg added PM_TRACE support to x86_64
[07:05] <mjg59> Back in Feisty
[07:05] <bdmurray> Right I see it in 2.6.20 but not 2.6.22
[07:05] <rtg> mjg59: I don't think we carried that forward.
[07:06] <mjg59> sigh.
[07:06] <mjg59> BenC: Please, please, please could you generate a list of patches that weren't carried forward?
[07:07] <kylem> eh?
[07:07] <kylem> it's upstream.
[07:07] <BenC> those got merged upstream
[07:07] <mjg59> BenC: It's still something of a general problem, though
[07:07] <rtg> mjg59: PM_TRACE was used to find out which drivers were causing suspend problems. I think most of them have been fixed. The remaining video issues are not solvable using PM_TRACE.
[07:08] <mjg59> rtg: No, there will still be a large number of drivers causing issues.
[07:08] <BenC> mjg59: at this point, suspend tracing doesn't much matter, does it?
[07:08] <BenC> but i was sure all of our patches got merged upstream, including the amd64 one
[07:09] <mjg59> BenC: We still seem to be missing the MSI and MMCONFIG stuff
[07:09] <BenC> Now that I was sure was merged at the start of gutsy work
[07:09] <mjg59> Nope
[07:10] <mjg59> static int pci_msi_enable = 1;
[07:11] <mjg59> BenC: This is why I keep asking for a list - nobody seems to have any clue which patches have been dropped by accident :)
[07:18] <bdmurray> I think there should be some consistency with pm_trace.  Having it available on i386 but not on x86_64 seems odd.  Also the wiki page DebuggingKernelSuspend points people to it.
[07:18] <mjg59> Yeah. It's in-kernel, it should be switched on.
[07:23] <BenC> yeah, the +6 is still there, so no amd64 support...hmm
[07:24] <rtg> BenC: Cherry-pick from Feisty?
[07:25] <BenC> rtg: yeah
[07:25] <rtg> Feisty c63e3d917a84214a98c81ae497bcd4a72ead62ed
[07:25] <rtg> I'll get it after kylem sorts out the ABI brokeness.
[09:28] <Kano> hi, are those vfs changes from 2.6.23?
[09:28] <Kano> they make that aufs does not compile again
[09:40] <BenC> Kano: no, they are caused by apparmor needing vfs changes for proper hooks
[09:41] <BenC> Kano: search for an aufs patch to make it work with latest apparmor
[10:01] <Kano> BenC: any idea where to find that?
[11:20] <okaratas> hello
[11:29] <jetsaredim> did we ever get an answer on the vmware modules question?