[11:02] <apw> ericm, i've asked on #ubuntu-x about the caps light thing
[11:40] <matti> I have a question: -pae should enable 32bit machine to be able to access whole of 4GB by introducing another level of page tables and utilizing PAE.
[11:40] <matti> But I cannot see a difference between -generic and -generic-pae on Karmic.
[11:52] <apw> matti, it actually should allow the processor to access more than 4GB of physical address space, depending how your bios maps you memory you may or may not get all of your 4GB.  if you can paste the e820 output (from dmesg) from the PAE kernel we can tell where it is
[11:53] <matti> e820?
[11:55] <matti> apw: Hold on.
[11:56] <apw> [    0.000000] BIOS-provided physical RAM map:
[11:56] <apw> [    0.000000]  BIOS-e820: 0000000000000000 - 000000000009c400 (usable)
[11:56] <apw> [...]
[11:57] <apw> that bit of the dmesg output following boot ... shows us wherre the BIOS thinks your memory is
[11:57] <matti> Yes, sorry.
[11:57] <matti> I had to put my glasses back :)
[11:57] <matti> http://pastebin.com/i3LEEepW
[11:58] <apw> matti, ok so your ram is all below 4GB, so PAE will not help you
[11:58] <apw> with the current memory layout
[11:58] <matti> :(
[11:58] <apw> now sometimes its changable (i am told) in the bios, but i've never seen it personally
[11:59] <apw> in my 4GB ram machine i have this
[11:59] <matti> BIOS is very limited and rather simple.
[11:59] <apw> [    0.000000]  modified: 0000000100000000 - 0000000140000000 (usable)
[11:59] <apw> the last GB of my ram is from 4-5GB physical, therefore pae would help me
[11:59] <apw> you are getting less thant 4GB cause your ram is being shadowed by the PCI space, AGP etc etc
[12:00] <apw> what machine is this?
[12:00] <apw> same laptops particually are limited in maximum memory, even if vendors sell you more than they can really utilise
[12:01] <matti> HP Desktop
[12:02] <matti> Give a second, I'll give you the exact model.
[12:02] <matti> HP Compaq dc7700p
[12:02] <matti> From dmidecode
[12:04] <apw> matti, did you add the ram yourself, do you know how many slots the mobo has?
[12:04] <matti> apw: Thank you for your help. I am rather disappointed in this machine ;(
[12:04] <matti> apw: Yes, I did.
[12:04] <matti> apw: 4 slots, each 1 GB.
[12:05] <apw> how much memory does the os report?
[12:05]  * apw suspects 3.5GB
[12:06] <matti> Yes.
[12:06] <matti> 3482884 kB to be precise.
[12:06] <matti> I reckon I have -512 MB for the PCI/AGP space.
[12:07] <matti> Which means that the chipset is rather limited.
[12:07] <apw> crucial tends to concur with that expectation
[12:09] <matti> apw: Hm?
[12:30] <apw> ram seller who tend to tell you how much ram to expect
[12:30] <matti> apw: I know :)
[13:46] <tgardner> apw, you'll be happy to know that I can crater Karmic with stress (the program), but not Lucid. 4 proc, 4 cores, 64 threads.
[13:46] <apw> i am very happy to hear that!
[13:51] <apw> smb, i've just resynced lucid with karmic hopefully for the last time, keep that in mind if you apply anything to lucid ... sync tag is updated
[13:51] <apw> s/lucid/karmic
[13:52] <smb> apw, Ok, I don't think that from now on there should be anything going into Karmic that is not in Lucid, but I keep it in mind
[13:53] <apw> yeah i am pretty sure we have almost no skew now, i did find three quirks marked as upstream which are not so in lucid, so i pulled them in
[13:53] <smb> Just watch out for those keyboard quirks which are supposed to be solved differently
[13:54] <apw> those definatly not needed?  we have the userspace support for those?  i read them as being not necessary if you have the userspace but ... 
[13:55] <smb> Ok, in theory probably we should, but I cannot say for sure. Maybe asking on #ubuntu-devel to know for sure?
[14:24]  * matti hugs apw 
[14:24] <matti> apw: Thanks for your help :)
[14:24] <apw> np
[14:29] <akgraner> sconklin, you around?
[14:29] <sconklin> akgraner: yeah, I'm here
[15:23] <manjo> smb, does you lenovo S12 take a long time to boot and shutdown ? 
[15:24] <smb> manjo, no. but it could happen that it needed a keypress to trigger it to go on. (like shift)
[15:24] <manjo> smb, yes or mouse movements ? 
[15:24] <smb> manjo, Strangely it seemed to be solved in that case by mem=nopentium
[15:24] <manjo> smb, I was trying to track down why this weekend ... but failed 
[15:24] <manjo> thought you might have some insight 
[15:25] <smb> manjo, Normally it something which is described in debugging voodoo (solved by acpi_skip_timer_override)
[15:25] <smb> manjo, But it this case the mem= option seems to solve quite a lot of strangenesses
[15:26] <manjo> smb, you mean its already doc in some wiki ? 
[15:26] <smb> manjo, https://wiki.ubuntu.com/KernelTeam/DebuggingVoodoo
[15:28] <manjo> smb, thanks did not see that 
[15:28] <apw> smb, where might i find that wacom change?  i don't see it in patchworks
[15:29] <smb> manjo, np, but please try mem=nopentium on your S12 to see it also helps you in that case
[15:29] <manjo> yep doing it now 
[15:29] <smb> apw, Hm, might have been a pull req?
[15:29] <apw> smb, doesn't look like it either
[15:29] <apw> cnd, was it you who had that?
[15:30] <smb> apw, yep
[15:30] <apw> i am running seriously short of uploads so i want them to count
[15:30] <smb> https://bugs.launchpad.net/bugs/516777
[15:30] <apw> cnd, did you ever send that wacome thing out for review? i know smb tested it
[15:30] <ubot3> Malone bug 516777 in linux "HP Touchsmart tm2 requires newer wacom driver" [Low,In progress] 
[15:30] <cnd> apw, a decision hadn't been made yet as to what we should do with it
[15:31] <cnd> I was going to get to it sometime today
[15:31] <cnd> should I send the patches out for review to kernel-team?
[15:31] <cnd> apw, oh yeah, I'm remembering now
[15:31] <cnd> I sent an email (you should have been cc'd) to the wacom driver maintainer
[15:31] <cnd> he's interested in helping to push it to -stable
[15:31] <apw> if we want it in lucid before it gets hard to get things in, ie before smb has a hand ... then it needs to be very very soon
[15:32] <cnd> apw, should we put it in as a pre-stable?
[15:32] <apw> smb you seemed to be keen
[15:32] <apw> cnd i was assuming it was a pre-stable from your conv with them, but ... not 100% sure
[15:32] <cnd> conversation with who? the wacom driver maintainer?
[15:32] <smb> apw, Well for the same reasoning. I think, yes we were thinking of how to progress, the problem is that its hard to predict wether the change as is gets into stable
[15:33] <smb> apw, So pre-stable might be wrong
[15:33] <apw> smb, yep ... ok so perhaps i'll punt that one to you :)
[15:33] <smb> apw, On the other hand our (limited) testing seemed to show no regression
[15:34] <smb> apw, So it might be a goody to have at that point of time (at least some users of this new hw will be thankful)
[15:34] <apw> yep, it seemed like we were already getting people hitting issues with new wacome stuff
[15:35] <smb> apw, So I would tend to include it (maybe still pre-stable as thats our intend)
[15:35] <apw> i probabally am doing just two uploads of distro before freeze.  one today with everything but drm33 and then if thats a go, the other will be just that
[15:35] <smb> If we get people screaming for regressions with some wacom stuff we still could revert before release...
[15:36] <apw> so ... i need it sooner rather than later
[15:36] <cnd> apw, there's another issue I want to talk to you about, re intel graphics
[15:36] <apw> smb, yep as thats a 'fix'
[15:36] <cnd> but I'll come back to it
[15:36] <apw> cnd i wouldn't wait, whats up
[15:36] <smb> cnd, So forward the wacom patches today to kernel
[15:36] <cnd> first, what are we doing with the wacom stuff? you want me to submit patches as pre-stable to k-t?
[15:36] <smb> (our kernel list)
[15:36] <cnd> smb, k
[15:36] <smb> apw, agree?
[15:37] <smb> Its still your tree. :)
[15:37] <cnd> apw, bug 488328
[15:37] <ubot3> Malone bug 488328 in linux "screen brightness won't come back when I open the lid" [Unknown,Fix released] https://launchpad.net/bugs/488328
[15:37] <apw> cnd get them to kernel-team now yes, i'll put them in as U: S: for now
[15:37] <cnd> it makes suspend/resume (or even just backlight turning off on screen closure, impossible to resume from on some intel hw
[15:37] <cnd> there's a patch that seems to fix things
[15:38] <apw> whats the patch
[15:38] <cnd> but it seems to have gotten lost upstream when one of the other intel gfx maintainers questioned it
[15:38] <cnd> apw: http://lists.freedesktop.org/archives/intel-gfx/2010-February/005902.html
[15:38] <cnd> you can see the reservation noted in the reply email
[15:39] <cnd> I'm not sure if the reply email is referring to issues that should be fixed in the future (making this moot for them, but still necessary for us)
[15:39] <cnd> or if we shouldn't apply it at all
[15:40] <cnd> I'm thinking the former, but wanted to know what you think should be done about it
[15:40] <apw> it sounds like a 'better' way to fix it, in that we are fixing the wrong bug
[15:40] <apw> but, if its broke and the fix fixes it ... then it may be much better than nothing
[15:41] <apw> cnd, do you have affected h/w ?
[15:41] <apw> wondering if the drm33 kernel also does this?
[15:41] <cnd> apw, I may in the form of this montevina box rtg sent me, but it seems to have other issues in that it hangs when I flip the lid switch
[15:41] <cnd> so even though I may have the hw, I can't seem to effectively test it
[15:41] <apw> heh great
[15:42] <cnd> apw, I think someone testing against one of the .33-rc kernels, and it wasn't fixed
[15:43] <cnd> apw, upstream bug report shows someone testing 2.6.33-rc6
[15:43] <cnd> and it was still broken
[15:44] <apw> ok get that one out on the list too so we can see how horrific it is
[15:44] <apw> and make a decision there ... thanks
[15:45] <cnd> apw, will do, thanks
[16:42] <cnd> smb, I'm looking into sending the wacom patches upstream to -stable, but the commits don't apply perfectly cleanly to the 2.6.32.y tree
[16:42] <cnd> is it still reasonable to say, please pick these two changes?
[16:42] <cnd> or should I point to my own kernel tree hosted at kernel.ubuntu.com with a branch for the changes?
[16:44] <smb> cnd, In this case generate patches that apply to current 2.6.32.y and send the patches as whole
[16:44] <cnd> smb, ok
[16:45] <tgardner> cnd, be sure to reference the upstream commit IDs from which your patch is generated
[16:45] <cnd> tgardner: ok
[16:57] <apw> yeah we use the upstream ids when stripping the commits we won't need in a rebase for M
[17:28] <manjo> have you guys seen this : [    0.562289] ACPI Error (psparse-0537): Method parse/execution failed [\_SB_.PCI0._OSC] (Node f7016a68), AE_ALREADY_EXISTS
[17:29] <manjo> apw, ^^ smb, 
[17:29] <manjo> [    0.562269] ACPI Error (dsfield-0143): [CAPB] Namespace lookup failure, AE_ALREADY_EXISTS
[17:29] <smb> manjo, Yes
[17:30] <smb> manjo, I think there is an acpi option to make execution serialized, maybe try that
[17:36] <manjo> https://bugs.edge.launchpad.net/ubuntu/lucid/+source/linux/+bug/507148
[17:36] <ubot3> Malone bug 507148 in linux "[lucid] desktop runs out of video memory on ATI Radeon Mobility 7500" [High,Confirmed] 
[17:37] <smb> manjo, Would "acpi_serialize" help with you error messages?
[17:50] <manjo> smb, will check thst 
[18:07] <manjo> apw, I need to focus on what I am doing .. so dropping off... 
[18:07] <apw> manjo, i hope i said that you could all do that!
[19:20] <manjo> another pulseaudio problem, this time on lenovo s12 bug #530234
[19:20] <ubot3> Malone bug 530234 in pulseaudio "[LUCID] unable to record sound though external mic Lenovo SL12" [Undecided,New] https://launchpad.net/bugs/530234
[19:20] <manjo> smb, If you are on karmic can you confirm that this works/fails on karmic ? 
[19:20] <manjo> and update the bug ? 
[19:21] <smb> manjo, I can't tell much about the boot speed
[19:21] <manjo> smb, regarding the bug above 
[19:22] <manjo> its about pulse audio I am asking to you see if this is a regression from karmic 
[19:27] <smb> manjo, On Karmic it works partially (I have to switch to microphone 1 or 2 manually to make internal or external microphone work) It does not seem to switch automatically
[19:27] <smb> Will add comments
[19:27] <manjo> :) thanks 
[19:56] <manjo> apw, I need RTL8192SE driver to get wireless working on EEEPC, currently the driver is in realtek website and not even in staging, I can add that to ubuntu-delta ? 
[19:57] <apw> you sure? that sounds familiar number?
[19:57] <manjo> yep I looked 
[19:57] <manjo> rtl8192e is available
[19:57] <manjo> which is pci 
[19:57] <apw> how bad is the delta.  and is it in wireless-testing?
[19:57] <manjo> let me look at wireless-testing 
[19:58]  * apw reminds manjo just how close kernel freeze is
[19:58] <manjo> apw, yes I know 
[19:59] <apw> i likely will be closing my distro tree by friday ... so ... hurry
[19:59] <manjo> apw, I will build and test the module 
[19:59] <manjo> today
[20:00] <manjo> apw, they should have thought about these netbooks at the start of the cycle
[20:00] <manjo> apw, I will do my best then its your call 
[20:00] <apw> manjo, indeed
[20:02] <manjo> apw, do you mind if I open a bug and do that delta that way ? 
[20:02] <manjo> apw, that way I have a number I can give krafty regarding enablement etc 
[20:02] <apw> manjo, you can use bugs for it if it helps you
[20:03] <apw> we just reserve the right not to use bugs in developement where it doesn't help us
[20:03] <manjo> apw, ok
[20:04] <manjo> apw, wireless-testing does not carry that module 
[20:04] <manjo> apw, there is e, su, 
[20:04] <manjo> one is pci the other is usb
[20:06] <apw> manjo, deep joy
[20:06] <manjo> apw, will start work on it now ... should have test results for you by this evening / request-pull
[20:07] <apw> i guess it will miss this upload
[20:45] <apw> manjo, this fix for bug #507148 ... how did you choose this fix for it?
[20:45] <ubot3> Malone bug 507148 in linux "[lucid] desktop runs out of video memory on ATI Radeon Mobility 7500" [High,Confirmed] https://launchpad.net/bugs/507148
[20:45] <manjo> apw, look at the upstream but on this one 
[20:45] <apw> manjo, as that fix fixes an oops
[20:46] <apw> and the error is not an oops
[20:46] <manjo> oh that last cherry pick I posted ? 
[20:48] <apw> manjo, yes that one, it seems to be an Oops avoider and to my eye not likely to fix the drm error he has
[20:49] <manjo> apw, you are correct... 
[20:50] <manjo> apw, sorry dude, I thought I had the fix 
[20:55] <apw> manjo, happens ... no worries
[23:16] <manjo> apw, is there a way to build just a subdir in kernel ? 
[23:26] <jk-> RAOF: welcome :)
[23:26] <RAOF> jk-: Thank you!
[23:26] <jk-> RAOF: you're in .au? whereabouts?
[23:27] <RAOF> Sydney.
[23:27] <RAOF> Where the glorious non-hot weather continues!
[23:27] <jk-> :D
[23:27] <jk-> plenty of colleagues around then!
[23:27] <RAOF> Yeah.
[23:28] <RAOF> I'll be moving to Hobart soonish, though.  Fewer collegues there, but it's a nicer place.  Also, it's a place where buying a house doesn't require your firstborn.
[23:28] <RAOF> Maybe I can persuade jml back there, too :)
[23:34]  * jk- is moving to Perth next week
[23:46] <jjohansen> jk-: wow, I had missed that.  Hope the move goes well for you
[23:46] <jk-> jjohansen: thanks! should be OK, I think we have most things organised
[23:47] <jjohansen> cool, I would offer to come help but there is a little ticket issue standing in the way :)
[23:47] <jk-> :D