UbuntuIRC / 2012 /04 /20 /#ubuntu-arm.txt
niansa
Initial commit
4aa5fce
=== NekoXPP is now known as Neko
=== plars-away is now known as plars
[05:41] <scientes> anybody have a static armel coreutils? i only need chroot
[05:42] <infinity> scientes: busybox-static should have chroot built in.
[05:43] <scientes> infinity, i've got a guy in #debian-arm where debootstrap from lenny arm fails
[05:43] <scientes> with invalid instruction
[05:44] <twb> scientes: is he using qemu arm cpu emu?
[05:45] <twb> There was a memory initialization bug in pre-2.0 qemu
[05:45] <scientes> twb, no he is using the NSLU2
[05:45] <scientes> with 2.6.32
[05:45] <scientes> *2.6.26
[05:45] <twb> pre 1.0 I meant, but OK
[05:45] <scientes> he is doing a armel chroot
[05:46] <scientes> does he need a newer kernel for EABI support?
[05:46] <twb> NFI
[05:46] <twb> Is it a recent ubuntu armel chroot? You know recent ubuntu is v7 only right?
[05:46] <scientes> no its squeeze
[05:47] <scientes> ubuntu wouldn't work
[05:47] <infinity> It's been a LONG time since I played with OABI, but aren't there kernel config differences between OABI and EABI-supporting kernels?
[05:47] <scientes> infinity, every syscall is differn't
[05:47] <scientes> but the new kernels support the old arch
[05:47] <infinity> That would be a "yes". :P
[05:47] <scientes> i just dont know which version
[05:48] <infinity> Sure, but does he have a new kernel?
[05:48] <infinity> I'd assume not.
[05:48] <scientes> the EABI syscalls are 0x90000000something+
[05:48] <scientes> 2.6.26
[05:48] <scientes> guess he needs to upgrade that first
[05:48] <scientes> i mean "new" relatively
=== mckoan|away is now known as mckoan
=== doko_ is now known as doko
[09:03] <ogra_> hmm http://www.phoronix.com/scan.php?page=news_item&px=MTA4NjA
[09:06] <lilstevie> ogra_: it doesn't do a great deal yet
[09:06] <ogra_> sure sure
[09:06] <lilstevie> but a great start
[09:06] <ogra_> but it seems to do more than just xfbdev
[09:07] <lilstevie> yeah
[09:09] <marvin24> needs xf86-video-modeset driver
[09:11] <marvin24> I'll hackup an upstream kernel after the next iteration
[09:11] <marvin24> let's see how far it will come ;-)
[09:14] <lilstevie> we have an almost working nv-tegra branched kernel running on the transformer prime :D
[09:19] <twb> "if it compiles it's great if it boots ship it" ?
[09:22] <lilstevie> lol
[09:23] <twb> linus said that
[09:23] <lilstevie> it boots
[09:33] <lilstevie> well, unless I build it with the cross compiler in the ubuntu repo
[09:34] <twb> A mere detail
[09:34] * twb waves hand dismissively
[09:34] <lilstevie> :p
[09:35] <LetoThe2nd> jedi hand move? ;)
[09:35] <twb> more like politician
[09:35] <LetoThe2nd> hrhr
[09:56] <nimesh_accenture> what does this error mean? sh: cannot create /dev/null: Permission denied ? i get it on the chrooted ubuntu-core on android
=== mckoan is now known as mckoan|away
[10:38] <vstehle> nimesh_accenture: looks like you have an empty /dev directory. Try to 'mount -o bind /dev chroot/dev' before you chroot.
[10:40] <nimesh_accenture> vstehle: :-) did that and it worked
[10:42] <nimesh_accenture> Also, I wanted to check the stability of my chroot, so is are there a set of tests that I can run on ubuntu shell to check if all the basic things like audio/framebuffer/wifi/ethernet etc. are in order?
[10:49] <vstehle> nimesh_accenture: for fb, you can always 'cat /dev/urandom > /dev/fb0' :)
[10:49] <vstehle> for audio, aplay comes to mind
[10:50] <vstehle> for wifi, try an iwconfig first
[10:50] <vstehle> for ethernet, a ping is a good start
[10:53] <ogra_> infinity, i guess you dont need to wait for the compiz stuff, the branch and the package have diverged (unuploaded changes in the branch), so i cant produce a sane patch that wouldnt cause havoc between package and branch (once again) ... unless i upload whats in the branch ... it seems that didrocks is on vacation and i dont want to blindly upload bits i dont know about...
[10:54] <infinity> ogra_: Gah.
[10:54] <ogra_> yeah
[10:54] <infinity> ogra_: Tell me there's a hope we can fix it for release?
[10:54] <infinity> (But I guess 0-ish day SRUs will be "okay"...)
[10:54] <ogra_> i dont think so, we would have to update the patch after each commit
[10:54] <infinity> ogra_: Mon/Tue would still be fine, if we can manage.
[10:54] <ogra_> k
[10:55] <ogra_> this will be a mess for five years though ... i dont see a clean solution (apart from the above which surely is a fulltime job)
[10:57] <nimesh_accenture> vstehle: thx
[10:58] <infinity> ogra_: Well, I would hope unity isn't getting major updates post-release.
[10:58] <nimesh_accenture> vstehle: but is there a test suit that does this for us? something similar to cts in android
[10:58] <ogra_> infinity, hahaha
[10:58] <infinity> ogra_: LET ME HAVE MY FANTASY.
[10:59] <vstehle> nimesh_accenture: a test suite? There is a "Linux test project" and a Phoronix test suite, I think. But I did not run them.
[10:59] * ogra_ grins
[11:00] <nimesh_accenture> vstehle: ok...cool! i'll take a look thx again!
[12:20] <rsalveti> ogra_: do you have the new package or patch done locally already? I can also help testing it if you want
[12:26] <ogra_> rsalveti, alf_, http://people.canonical.com/~ogra/compiz/
[12:37] <vstehle> ogra_: thanks! it works on omap5 :)
[12:38] <ogra_> heh
[12:41] <nimesh_accenture> guys i'm getting this error on ubuntu-core 11.10 chrooted on Android ICS : Temporary failure resolving 'ports.ubuntu.com' . is this a DNS issue? i changed /etc/resolve.conf with my local DNS ip add... the ubuntu-server version 11.10 works fine...
[13:01] <ogra_> rsalveti, alf_, any news ?
[13:03] <rsalveti> ogra_: updating/upgrading
[13:04] <rsalveti> want to make sure it works with the latest stuff
[13:04] <ogra_> right
[13:38] <infinity> nimesh_accenture: Either you got resolv.conf wrong, or your DNS server is/was upset. There's not a whole lot else that can go wrong there.
[13:38] <infinity> nimesh_accenture: (Well, or your system isn't connected to the internets)
[13:42] <nimesh_accenture> infinity: hmmm the same DNS works for my desktop ubuntu ... is there a possibility that some ports are blocked? because i'm behind a corporate network , which has some restrictions
[13:43] <infinity> nimesh_accenture: I have no idea how your network or your android system are configured.
[13:43] <infinity> nimesh_accenture: All I can tell you is that setting resolv.conf in the chroot is all you need to do. If that doesn't work, it's the host system or the network.
[13:45] <nimesh_accenture> hmmm ...ok...thx infinity...i'll try and talk to the network team to find out whats going on
[13:45] <infinity> nimesh_accenture: It's more likely to be your android setup than the network team's fault. But they may have some handy tips.
[13:47] <nimesh_accenture> infinity: The README of pandaboard says "Ethernet networking is hardcoded to use 8.8.8.8 and 8.8.4.4 DNS." , but should that affect the chrooted Ubuntu?
[13:48] <infinity> No.
[13:57] <ndec> nimesh_accenture: if you are behind firewall, you need to configure your proxy in the chroot as well. as your config outside of the chroot isn't visible
[14:04] <nimesh_accenture> ndec: but I didn't have to configure http_proxy on my desktop ubuntu, however /cat /etc/resolv.config on my desktop shows me the DNS ip's
[14:06] <ndec> nimesh_accenture: well, if you are behind proxy, you must have some configuration somewhere, no?
[14:08] <nimesh_accenture> I dont know what the networking team has done , but we dont have to configure any proxy... our ip addresses are like 10.211.xx.xx
[14:09] <nimesh_accenture> ndec: thx man...gtg ...have a happy weekend guys
[14:37] <rsalveti> ogra_: working nicely!
[14:37] <rsalveti> alf_: and it's actually usable!
[14:38] <ndec> cool, from the archive?
[14:39] <rsalveti> ndec: yup
[14:39] <rsalveti> ndec: but with the latest package from ogra, that just starts gtw-window-decorator automatically
[14:39] <rsalveti> that should hit the archive soon
[14:39] <ndec> rsalveti: nice. i am running a minimal system atm, what's the simplest way to install unity3d? ubuntu-desktop?
[14:40] <rsalveti> alf_: the only thing I noticed that's a bit slow, is when you're using dash in full screen
[14:40] <rsalveti> ndec: if you have to install ubuntu-desktop I'd probably recommend you to install the pre-installed image from scratch
[14:40] <rsalveti> it takes so much time to update/install the packages
[14:40] <rsalveti> that it's just faster to flash a new one :-)
[14:41] <ndec> rsalveti: from usb it's usually not to bad...
[14:41] <alf_> rsalveti: ogra_: Great! Actually I missed the ping... Ricardo can you please ping me as the second name to test something
[14:41] <ndec> i have lots of things on my /, i don't really want to flash something else ;-)
[14:42] <rsalveti> ndec: so I believe installing ubuntu-desktop will probably do the work
[14:42] <rsalveti> alf_: the dash with the default size sounds ok, but with full screen it feels quite slow
[14:57] <alf_> rsalveti: I think the full-screen active blur is too much for the sgx, when I disable the dash blur it's much better
[14:57] <rsalveti> alf_: makes sense
=== prpplague^2 is now known as prpplague
[15:07] <ppisati> guys, rootstock is not supported anymore, right?
[15:07] <ppisati> what should i use instead of it?
[15:09] <ogra_> rsalveti, alf_ awesome, committed and pushed, the rest has to happen once didrocks is back from vacation
[15:12] <ndec> ppisati: ubuntu-core
[15:14] <ogra_> ppisati, depends, there is no 1:1 replacement
[15:17] <kenws> Hi there, Does anyone know what eglibc options are used on ARM (in terms of option-groups.config)?
[15:19] * kenws is looking at https://launchpad.net/ubuntu/+source/eglibc/2.15-0ubuntu2/+build/3199667
[15:20] <kenws> hm, probably just the defaults (everything enabled)
[15:21] <ppisati> ndec: ogra_ so, what do i use to roll an ubuntu filesystem in a tarball?
[15:22] <ndec> ppisati: i don't understand the question...
[15:22] <ogra_> you can use -core but need to do all configuration
[15:22] <ndec> and i wrote that a while ago : http://omapedia.org/wiki/OMAP_Ubuntu_Core
[15:22] <ogra_> yeah, i guess i will steal from it for the ubuntu wiki :)
[15:23] * ppisati goes to read
[15:24] <ndec> ogra_: omappedia, says this "Content is available under Creative Commons Attribution-Share Alike 3.0 license.." so you should be fine ;-)
[15:24] <ogra_> heh
[15:24] <ndec> just say my name, and how nice i am..
[15:25] <ogra_> and bring some bribing money ?
[15:25] <ndec> that could do it
[15:26] <ndec> rsalveti: apt-get install ubuntu-desktop took less than 15 mins ;-)
[15:27] <rsalveti> ndec: nice :-)
[15:35] <Adebar> Anyone knows a ubuntu compatible notebook (ARM) with bigger screnn than Toshiba AC100 or Genesi Efika MX Smartbook?
[15:36] <steev> there isn't one
[15:36] <steev> well, maybe the macbook air
[15:36] <steev> oh wait, you said ARM
[15:36] <steev> there still isn't one
[15:37] <steev> that said, you could get either the AC100 or the Smartbook and replace the screen, there is a 10" 1280x720 screen that is compatible (disclaimer: i work for Genesi)
[15:39] <Adebar> mhm... perhaps another one knows such a device. Sry if I asked three times in ten minutes, looked like it didn't work.
[15:42] <alf_> vstehle: How does unity feel on OMAP4 performance-wise?
[15:42] <alf_> vstehle: oops, OMAP5
[15:43] <vstehle> alf_: it does not work so well...
[15:45] <alf_> vstehle: Do you think (or guess!) this is a driver issue or a compiz/unity issue?
[15:45] <ndec> alf_: i am trying to setup a panda with the same driver.
[15:46] <vstehle> alf_: issue is probably on my side, between keyboard and chair :)
[15:46] <ndec> in fact we are now using the new xf86-video-omap driver, which rsalveti is also trying to bring up for next LEB
[15:46] <ndec> for now we have driver issues.
[15:46] <alf_> vstehle: lol
[15:46] <alf_> ndec: ok, thanks
[15:47] <ndec> vstehle: on panda, i have the black screen issue... i guess our PPA does not have the latest fixes from robclark?
[15:48] <vstehle> ndec: I don't know... I have not tested on OMAP4 since a week or two.
[16:55] <prpplague> rsalveti: http://tincantools.com/product.php?productid=16158&cat=0&page=1&featured
[16:55] <prpplague> rsalveti: kits will be ready on monday
[17:01] <ogra_> alf_, rsalveti, compiz is now in precise-proposed, would be good to give the packages in the archives another test once they built to make sure everything is fine and i didnt make any mistakes
[17:23] <alf_> ogra_: great, thanks!
[18:17] <rsalveti> prpplague: awesome!
[18:17] <rsalveti> ogra_: will do, thanks
=== W1N9Zr4549 is now known as W1N9Zr0
[19:11] <gdane> hello
[19:12] <gdane> i recompile my kernel and i need ipkg-utils 1.6 r21
[19:13] <gdane> can some one say me wherr i can find it?
[19:14] <gdane> all the time bitbake try to compile it, it says that it cant go to cvs handheld.com
[19:14] <gdane> hendheld isnt work now
[19:14] <gdane> so i thought may be some one help me
[19:15] <gdane> i asked google before, but it shows only misstakes and not solving of this problem
[22:13] <jeinor> hi! I was redirect here from #ubuntu-devel because I had arm-related questions to getting Ubuntu Core up and running on a ARM-system.
[22:15] <jeinor> I got an error message trying to boot a 3.3.1 kernel about "VFS: Cannot open root device "mmcblk0p2" or unknown-block(0,0)"
[22:16] <jeinor> the kernel command line is from a older kernel (2.6), I'm not really sure what command line to use with the 3.3.1 kernel
[22:16] <jeinor> grateful for any assistance
[22:27] <rcn-ee> jeinor, it's probally either, your bootargs don't match your partition layout, or... you don't have the necessary drives to access the drive..
[22:34] <jeinor> thanks for answering . I think the drivers are there, I got the kernel configuration from a forum thread where this user (which does not respond back) said he got it up and running with that configuration.
[22:34] <jeinor> it's probably the bootargs, and me not knowing how to set them in u-boot
[22:35] <jeinor> I tried editing the script /boot/boot.scr by hand (vim), but whenever I do it seems to break the boot process completely (even if I just add the --verbose flag)
[22:52] <jeinor> rcn-ee, ok I figured out how to change kernel command line args. Apparently I can't just change the args directly in boot.scr, I had to regenerate it using mkimage :)
[22:55] <jeinor> rcn-ee, did you just reenter?
[22:56] <jeinor> ah, I missed your quit notice there
[22:56] <jeinor> I'll repost my last message: rcn-ee, ok I figured out how to change kernel command line args. Apparently I can't just change the args directly in boot.scr, I had to regenerate it using mkimage :)
[22:57] <rcn-ee> yeah, you can't edit that directly, take a look into your u-boot printenv, it might also support "uEnv.txt" files, which are raw text files.. so nice to edit..
[22:58] <jeinor> ok, thanks. I think I'll go with the method I found for now, just running the mkimage-command
[22:58] <jeinor> still no luck in finding that root partition though
[22:59] <rcn-ee> then it might be either a missing driver for the interface, or even misssing support for that file system.. and assuming you've ran 'fdisk -l' on it so you know it's correct..
[23:01] <jeinor> it's a dd copy to a new sd card from the stock one sent with the cubox, so I think it's correct. I'll check the filesystem though, as I replaced a 2.6 kernel with 3.3.1 perhaps the default support for ext2 is dropped
[23:01] <jeinor> thanks for the tips
[23:03] <rcn-ee> humm, if it's ext2, i've never seen anyone disable that by default..
[23:04] <jeinor> you're right, checked with "make menuconfig" and it's enabled (you can see I'm pretty new with this)
[23:08] <jeinor> this makes me believe the driver is there: "mmc0: SDHCI controller on sdhci-dove.0 [sdhci-dove.0] using DMA"
[23:08] <jeinor> (I'm trying to boot from a sdcard)