[17:30] <Martyn> ogra : I have an alpha-quality u-boot now
[17:30] <ogra> nice
[17:30] <Martyn> ogra : Should I commit a bug and patch?
[17:30] <ogra> sure, but that wont enter jaunty anymore
[17:30] <Martyn> frig
[17:30] <Martyn> we're locked on Redboot for now then, eh?
[17:31] <ogra> yes, and it works just fine, is configurable from the running system etc
[17:31] <Martyn> thanks to a -lot- of work by lool, you, etc...
[17:31] <ogra> http://people.ubuntu.com/~ogra/arm/babbage/ has todays image
[17:31] <Martyn> It's just a pity we have to do all those contortions :)
[17:32] <ogra> lool is nearly done integrating the build script into the autobuilders, so std images will show up this week on cdimage.u.c
[17:32] <Martyn> ogra : Are you experiencing any board lockups or instability surrounding using the USB bus?
[17:32] <Martyn> ogra : I keep getting hard lockups when using a network USB adapter
[17:32] <ogra> we were told to not use the normal USB ports but  the mini
[17:33] <ogra> at least for devices with some payload
[17:33] <lool> Martyn: Same here, but we moved to the builtin ethernet again now
[17:33] <lool> Martyn: It's better with the latest 2.6.28 we ahve
[17:33] <ogra> i have kbd and mouse in normal USB ... HDD on mini with powered HUB
[17:33] <Martyn> lool : Someone found the slow performance issue for the built in ethernet?
[17:33] <ogra> we got patches
[17:33] <Martyn> ogra : using one of those weird usb A-B cables?
[17:34] <Martyn> mini a/ minib
[17:34] <ogra> they are in the current kernels but cause other probs
[17:34] <Martyn> *sigh*
[17:34]  * ogra points to bug 356517 and bug 356975
[17:34] <ubot4> Launchpad bug 356517 in network-manager "udev does not detect eth0 on armel" [High,In progress] https://launchpad.net/bugs/356517
[17:34] <ubot4> Launchpad bug 356975 in linux "imx51 oops after watchdog call" [High,New] https://launchpad.net/bugs/356975
[17:35] <Martyn> So many issues to solve, and Pegatron is already starting tooling up for ARM based netbooks with the i.mx515 in them
[17:35] <Martyn> Yeah, I just saw 356975
[17:35] <ogra> it doesnt stop the NIC from working
[17:35]  * Martyn lets the 3 hour download from ogra's image repo start
[17:36] <ogra> but i dont see any plug events anymore for the cable
[17:39] <Martyn> Is redboot entirely frozen for jaunty?
[17:40] <Martyn> Or is it possible that I might be able to get a usb-boot patch in?
[17:40] <ogra> i dont think we can add new features at this point
[17:44] <Martyn> Oh well.
[17:45] <ogra> we're close to RC
[17:45] <ogra> bugfixing is the main focus now
[17:45]  * ogra would be very very thankful for any idea or help on bug 328167 btw
[17:45] <ubot4> Launchpad bug 328167 in gnome-keyring "[arm] gnome-keyring-daemon eating 100% CPU at login in Jaunty" [High,Triaged] https://launchpad.net/bugs/328167
[17:46] <ogra> its our biggest blocker atm
[18:00] <Martyn> ogra : I'll take a look at it
[18:00] <ogra> that would be cool, we need as many eyes as possible on it
[18:01] <ogra> (i know NCommander planned to look as well and i'll donate the the rest of my week to it too)
[18:03] <NCommander> ogra, as far as I can tell its a Babbage specific issue
[18:03] <NCommander> ogra, I was unable to reproduce it on anything else :-/
[18:03] <ogra> it is
[18:03] <NCommander> ogra, oh, I didn't see that in the bug report :-/
[18:03] <ogra> but we still need to solve it
[18:04] <Martyn> To debug the issue, I'm going to have to do a debug build
[18:04] <Martyn> there's not trace information
[18:04] <Martyn> not/no
[18:05] <ogra> there is the upstream bug
[18:05] <Martyn> NCommander : it happens on the EVM as well as babbage
[18:05] <NCommander> ogra, the upstream bug doesn't have a full backtrace
[18:05] <ogra> but what i added there is apparently not enough
[18:05]  * NCommander is going to have to trace it through the entire GNOME stack until it calls into libc
[18:05] <NCommander> ddebs are your friend for this
[18:05] <ogra> Martyn, you have an EVM that survives gnome ?
[18:05] <Martyn> yeah
[18:05] <Martyn> but I cheat :)
[18:06] <Martyn> usb->video adapter
[18:06] <ogra> soldered RAM on it youreself ?
[18:06] <ogra> -e
[18:06] <Martyn> no, I just live with an extremely swappy system
[18:06] <Martyn> it's slow as tar
[18:06] <ogra> heh
[18:06] <Martyn> no, slower
[18:06] <NCommander> Well, we do know its a regression, we didn't start seeing it until about February
[18:06] <ogra> yeah
[18:06] <NCommander> Martyn, sounds like my Babbage running with the pendisk
[18:06] <Martyn> slow as pure glass flowing down a window
[18:06]  * NCommander is setting up the other one with the USB HDD
[18:07] <Martyn> hey, I have a BeagleBoard tucked away in the drawer
[18:07] <Martyn> this might be a good time to drag it out, and see if I can get the gnome-keyring issue to show up there
[18:07] <NCommander> Martyn, +1
[18:08] <NCommander> I couldn't reproduce on something running a 2.6.25 kernel, but I can't upgrade to the latest kernel either
[18:08] <NCommander> (newest userland)
[18:08] <Martyn> My only issue there, is the crappy usbOTG on the board
[18:08] <NCommander> and my slugs don't have enough RAM
[18:08] <NCommander> Martyn, ow. Thats going to hurt
[18:08] <Martyn> but I think it's possible to solder two pins together on the mini USB port, and not have to deal with getting a special OTG cable
[18:09] <Martyn> I should have done the mod ages ago
[18:09] <ogra> on mine a plain mini->-std USB adapter works
[18:10] <Martyn> ogra : I need keyboard, mouse, ec..
[18:10] <Martyn> so I need to connect it to a usb hub first :)
[18:10] <Martyn> brb
[18:15] <Martyn> okay, BeagleBoard pulled out of drawer
[18:15] <Martyn> ogra, any reason to think the kernel that was built for your image wouldn't boot it?
[18:16] <ogra> ??
[18:16] <ogra> the imx51 kernel on the beagle ?
[18:16] <ogra> no, that surely wont boot
[18:16] <Martyn> ogra : Sorry, brain not working .. I meant "any reason your image wouldn't boot with a proper Beagle kernel..."
[18:16] <Martyn> error existed between user and keyboard
[18:17] <ogra> no, should work
[18:18] <Martyn> good, that gives me a consistent platform to test
[18:19] <Martyn> Of course, it helps if the clock is set
[18:19] <Martyn> There has -got- to be a way to kill that @#$@$ annoying bug
[18:19] <NCommander> Martyn, we know its a regression, it popped up around Feb 20th
[18:19] <Martyn> I mean, sure, we know it's not really 1970 ...
[18:19] <NCommander> Oh, THAT bug
[18:19] <Martyn> yep, the pam bug
[18:20] <NCommander> That one got fixed
[18:20] <NCommander> Its not pam, its shadow
[18:20] <NCommander> Martyn, is that using a daily vs. a beta image?
[18:20] <NCommander> (it was fixed after beta)
[18:22] <Martyn> daily, but I'm downloading Oscar's latest one now
[18:23] <Martyn> 24% complete .. it's a slow download today
[18:23] <NCommander> Martyn, Oscar?
[18:24] <NCommander> Martyn, if your still seeing the password bug on a fresh install thats a problem
[18:24] <NCommander> :-/
[18:29] <Martyn> ogra
[18:29] <Martyn> Well, i've got another 30 minutes before the download is complete
[18:29] <Martyn> NCommander : Then we'll see if the password bug remains
[18:30] <Martyn> but if it's reported fixed, I shouldn't see it
[19:07]  * ogra points to http://cdimage.ubuntu.com/ports/daily-live/20090407.2/ look, there is a babbage image ;)
[19:09] <lool> http://cdimage.ubuntu.com/ports/daily-live/current/ is the stable URL
[19:09] <ogra> right
[19:11] <NCommander> ogra, I don't see a babbge image at that URL ...
[19:11] <NCommander> Oh
[19:11] <NCommander> bah, you need to mark it armel+imx51 :-P
[19:12] <ogra> cjwatson didnt want to
[19:12] <ogra> given that its the only arm live image we build anyway
[19:12] <ogra> the html text will be updated though
[19:13] <ogra> anyway, out now ... bye
[19:24] <NCommander> Martyn, how goes your install?
[19:24]  * NCommander is currently grabbing all the ddebs
[19:53] <NCommander> ogra, you around?
[19:57] <lool> Did someone try usplash on babbage?
[19:58] <lool> The initramfs is really slow, that surprizes me
[20:11] <NCommander> lool, it works on PowerPC AFAIK, so it *should* work. How well, I dunno
[20:13]  * NCommander finally is getting some useful backtraces :-)
[20:21] <Martyn> re
[20:22] <Martyn> NCommander : I'm recompiling the kernel for the Babbage board
[20:22] <Martyn> NCommander : Turns out there wasn't one handily available for some reason
[20:22] <Martyn> er, Beagle, not babbage
[20:22] <NCommander> Martyn, there are plenty available, why are you recompiling for Babbage?
[20:22] <Martyn> Brainfrarting continues
[20:22] <NCommander> Oh, Ubuntu doesn't officially support Beagle
[20:23] <Martyn> Right
[20:23] <Martyn> so I'm compiling Beagle
[20:23] <Martyn> but >only< changing the parameters needed to get it working on Beagle, so as to hopefully see if I see the same issues
[20:23] <Martyn> because I'll be using the Babbage userland
[20:24] <NCommander> heh
[20:24] <Martyn> time to compilation completion -- 1hr 8min
[20:24] <Martyn> Ah, the joys of native compilation
[20:37] <pwnguin> how significant is uclib to the arm port?
[20:39] <Martyn> we use uclib?
[20:39] <pwnguin> looks like no
[20:40] <Martyn> I use uclibc to produce busybox images early on during a bring up of a new arm board...
[20:40] <Martyn> but AFIK (YMMV) the jaunty arm port is full blown glibc
[20:40] <pwnguin> yea, i was thinking it'd be part of busybox or something but i guess not
[20:42] <pwnguin> intersting. uclibc is in hardy
[20:42] <pwnguin> dropped from intrepid and later i guess.
[20:44] <pwnguin> only reason i ask is because i was revisiting a patch dump from Garmin, trying to figure out why a patch wasn't applied upstream. it seems tracked down to uclibc, but if debian / ubuntu aren't using uclibc there's probably no interest in the patch or upstreaming it.
[21:19]  * NCommander headthunks 
[21:49] <pwnguin> NCommander: ?
[22:25] <NCommander> pwnguin, working on g-keyring-daemon
[22:40] <pwnguin> fun