2010/04/27/#ubuntu-arm.txt

=== ericm-afk is now known as ericm|ubuntu
=== JaMa|Zzz is now known as JaMa
[07:17] <ericm|ubuntu> saeed, ping
[07:18] <saeed> ericm|ubuntu: hi
[08:13] <saeed> ericm|ubuntu: please check the u-boot under https://marvell.wiki.canonical.com/UploadedFiles?action=AttachFile&do=view&target=Dove_UBoot_4_4_2_Full_Release_NQ.zip
=== ericm_ is now known as ericm|ubuntu
=== hrw|gone is now known as hrw
=== JamieBen1ett is now known as JamieBennett
[13:23] <hrw> http://hrw.pastebin.com/GdyhNPbt - beagleboard c3 with hub on otg
[13:23] <persia> Is that really a 7 minute boot?
[13:24] <hrw> [    0.000000] Registering NAND on CS0
[13:24] <hrw> [  345.080596] OMAP DMA hardware revision 4.0
[13:24] <hrw> nope
[13:25] <persia> Where's the first 5 minutes of time come from?
[13:26] <hrw> waiting in uboot
[13:26] <hrw> probably
[13:26] <persia> ouch!
[13:26] <persia> Or did you have to do something manually in uboot (it's less painful if waiting on us slow humans)
[13:27] <hrw> I took card out, wrote lucid on it, pasted content of boot.scr+serial into uboot and booted
[13:29] <persia> Ah, OK.  Yeah, I can see how that takes 300 seconds :)
[13:30] <hrw> writing card was 242s
=== ogra_ is now known as ogra
=== jmcgee|gone is now known as jmcgee
[15:12] <ogra> persia our shepherd :)
[15:13]  * persia has accepted certain responsibilities as a side effect of being allowed in the #ubuntu-ops channel
[15:13] <ogra> anyway /me needs a break ... back soon
[15:16] <hrw> shepherd... for me it is sharp zaurus c750 ;D
[15:25] <persia> Indeed.
[15:26] <hrw> but c7x0 are not ubuntu targets - armv5te
[15:28] <persia> Omegamoon got Jaunty working, but yeah, these days I recommend installing Debian on those.
[15:28] <persia> Well, maybe not on a 750 :)
[15:36] <hrw> 750 is enough as long as you have rootfs on sd
[15:36] <hrw> hi prpplague
[15:38] <prpplague> hrw: hey bud
[15:38] <prpplague> davidm: ping
[15:39] <davidm> prpplague, hello, I'm sort of here
[15:39] <davidm> traveling again
[15:39] <prpplague> davidm: ahh np, i wanted to get some feedback from you when you have some time
[15:40] <davidm> if I can, also I have a question for you, the TinCan Tools Beagle board parts.  Are the Ethernet parts stable?
[15:40] <prpplague> davidm: give me a shout when you'd have about 20 minutes or so
[15:41] <prpplague> davidm: stable in what way? software, hardware, availibility?
[15:41] <davidm> stable in terms of the 10/100 part working solidly and not causing kernel panics or losing connectivity?
[15:42] <davidm> Might have some folks buying some
[15:42] <prpplague> davidm: i have not heard of any issues, the micrel folks have run a number of tests on the zippy2 and we've run iperf and burning tests
[15:43] <prpplague> davidm: we've sold about 400 zippy2 boards and i've not had any negative feedback
[15:43] <davidm> prpplague, thanks, I'll pass that on
[15:43] <prpplague> davidm: you can purchase the zippy2's via digikey as well
[15:43] <prpplague> davidm: you were at the Nice meeting right?
[15:43] <davidm> prpplague, cool, thanks
[15:43] <davidm> Yes
[15:43] <davidm> I was indeed
[15:44] <cwillu_at_work> what does the rtc with battery do?  is it just supplying the battery to the tps, or is it a different device?
[15:45] <cwillu_at_work> on the zipp2
[15:45] <prpplague> cwillu_at_work: seperate RTC device connected via i2c
[15:46] <cwillu_at_work> ... why?
[15:46] <persia> prpplague: Does it have a battery?
[15:47]  * persia reads again
[15:47] <cwillu_at_work> I guess it's a replaceable battery as opposed to the one you have to solder onto the beagle
[15:47] <persia> Hurray!  It doesn't fail to boot with the current filesystems :)
[15:47] <prpplague> persia: yes
[15:47] <cwillu_at_work> but if you're soldering things on anyway
[15:48] <prpplague> cwillu_at_work: ??
[15:48] <cwillu_at_work> prpplague, to put the expansion header onto the beagle
[15:48] <prpplague> cwillu_at_work: ahh indeed
[15:53] <cwillu_at_work> main difference between zippy and zippy2 is the 100mbit?
[15:57] <prpplague> cwillu_at_work: yea
[15:58] <prpplague> cwillu_at_work: the zippy uses the enc28j60 from microchip and the zippy2 uses a micrel part
=== JamieBen1ett is now known as JamieBennett
[16:01]  * cwillu_at_work considers possibilities
[16:01] <cwillu_at_work> possibilities such as btrfs raid across two mmc cards :p
[16:03] <prpplague> sata seems to be an item i am getting alot of requests for
[16:04] <prpplague> cwillu_at_work: we also have the trainer board if you are into hacking hardware - http://www.elinux.org/BeagleBoard_Trainer
[16:23] <cwillu_at_work> sata would be nice, although I don't have much use for it myself
[16:27] <cwillu_at_work> prpplague, is it sensitive to the length of the connection?  I.e., could I run a two inch ribbon cable between beagle and zippy2?
[16:27] <prpplague> cwillu_at_work: yea you could
[16:28] <prpplague> cwillu_at_work: or even use some right angle connectors
[16:28] <cwillu_at_work> hmm, I guess, ya
[16:28] <prpplague> cwillu_at_work: no more than about 6 to 8 inches
[16:30] <hrw> prpplague: can zippy2 and trainer be stacked on one bb?
[16:30] <prpplague> hrw: not currently
[16:31] <prpplague> hrw: we are looking at doing a board so you can mount them side by side
=== asac_ is now known as asac
[16:58] <cwillu_at_work> prpplague, too bad the connector is already soldered onto the zippy2
[16:59] <cwillu_at_work> prpplague, specifically, keeping the right-angle connectors on top would recover significant vertical clearance
[17:38] <prpplague> cwillu_at_work: we have a small right angle pcb we use inhouse for some testing
[17:38] <prpplague> cwillu_at_work: we had considered making it available on the web site
[17:39] <cwillu_at_work> I don't follow
[17:39] <cwillu_at_work> like _||__[]_ ?
[17:41] <prpplague> cwillu_at_work: yea
[17:41] <cwillu_at_work> okay, you missed my point :)
[17:41] <cwillu_at_work> I want the connectors on the populated sides of both boards, so that I can have them sitting lower in their cases :)
[17:43] <prpplague> cwillu_at_work: ahh
[17:43] <prpplague> cwillu_at_work: i follow
[17:44] <prpplague> cwillu_at_work: that could be done with a small board
=== hrw is now known as hrw|gone
[17:46] <Vazz> hello, i've got a question: how to build rootfs of debian squeeze from scratch if I've got fedora 12 on my desktop now?
[17:46] <cwillu_at_work> Vazz, debootstrap is meant to do that
[17:46] <cwillu_at_work> rootstock is basically a wrapper around it to make things easier
[17:47] <ogra> rootstock wont work on fedora
[17:47] <cwillu_at_work> it will with the right things in the right places
[17:47] <persia> debootstrap will.
[17:47] <ogra> it has dependencies fedora wont fulfill
[17:47] <Vazz> so only debootstrap method?
[17:48] <ogra> right, rather take the harder part and use debootstrap
[17:48] <Vazz> yum install debootstrap? : D
[17:48] <ogra> its less work to bring up on fedora but more work setting up the system afterwards
[17:52] <Vazz> hmm i think there's no script for squeeze in debootstrap
[17:55] <Vazz> can i install squeeze with it?
[17:57] <prpplague> orbarron: has nicolas been over by your desk already?
[17:58] <orbarron> prpplague: not yet but should stop by later today...
[17:58] <prpplague> orbarron: let me know if he stops over there
[18:14] <Vazz> i've downloaded new version of debootstrap from webpage, it downloaded packages...then W: Failure trying to run: chroot /home/arek/d_db mount -t proc proc /proc
[18:15] <Vazz> any advices?
[18:20] <cwillu_at_work> you'll have to --foreign --arch-armel
[18:20] <cwillu_at_work> otherwise it'll try to use arm binaries to install the system on your very-much-not-an-arm system
[18:21] <cwillu_at_work> then on first boot you'll do a debootstrap --second-stage
[18:21] <Vazz> thx
[18:23] <Vazz> so now i have to boot it on my arm devce and debootstrap --second-strage?
[18:24] <cwillu_at_work> yes;  probably/may need to boot it with an explicit init=/bin/bash
[18:33] <Vazz> "debug rootdelay=2 init=/bin/bash root=/dev/mmcblk0p2 fbcon=rotate:1 mem=64M@0xa0000000 mem=64M@0xb0000000"
[18:34] <Vazz> thats part of my startup.txt
[18:34] <Vazz> and it's saying that 'unable to open initial console' then kernel panic
[18:41] <cwillu_at_work> use "rootwait" instead of "rootdelay=2"; try dropping the fbcon bit, not sure why you'd need that
[18:41] <cwillu_at_work> might need to specify "console=ttyS2,115200n8" or somesuch
[18:42] <Vazz> fbcon rotates my screen : )
[18:42] <cwillu_at_work> what hardware is this on?
[18:43] <Vazz> toshiba g900
[18:43] <cwillu_at_work> can't help you there then :p
[18:44] <cwillu_at_work> but you'll probably have to muck with the console
[18:44] <Vazz> : /
[18:47] <Vazz> with rootwait same thing like before
[18:47] <cwillu_at_work> yes, that's just a "doing the right thing" thing, not to fix your problem
[18:47] <cwillu_at_work> rootdelay=2 is a hack, rootwait is the fix
[18:50] <Vazz> hmm
[18:51] <cwillu_at_work> and neither is related to your current problem :p
[18:52] <Vazz> thats not good :D
=== XorA is now known as XorA|gone
[21:01] <inhand1> In trying to run Ubuntu ARM NetInstall on a BeagleBoard, USB does not seem to be working. Neither OTG nor HOST. Without mouse or keyboard, it is a little tough to click the OK button. Is this a bug in the kernel?
[21:02]  * sveinse is wondering if Beagleboard has the TI feature in which it requires a HUB to do USB1.1...
[21:02] <sveinse> Try connecting via a HUB and see if you get kbd+mouse then
[21:03] <sveinse> I don't know id that's the case, but I'd give it a go
[21:03] <inhand1> I am using a powered hub.
[21:04] <sveinse> inhand1: Sorry then, I don't know. Best of luck
[21:05] <sveinse> Uhm. Do you know if the HUB is 1.1 or 2.0?
[21:05] <inhand1> it is 2.0
=== nslu2-log_ is now known as nslu2-log
=== jmcgee is now known as jmcgee|gone