**** BEGIN LOGGING AT Mon Jan 16 02:59:57 2012
Jan 16 05:41:48 <AdamOutler> hello!
Jan 16 05:42:57 <AdamOutler> I was wondering if there was a guide somewhere which can lead me through obtaining Ubuntu/linux sources and compiling it into a boot.img?
Jan 16 05:43:07 <AdamOutler> and a bootable drive.
Jan 16 05:43:10 <twb> For what hardware?
Jan 16 05:43:26 <AdamOutler> I am using a blaze-tabet based piece of equipment.
Jan 16 05:43:41 <AdamOutler> Texas Instruments OMAP 4430, blaze tablet.
Jan 16 05:43:51 <twb> There are prebuilt omap4 images
Jan 16 05:44:03 <AdamOutler> It's a production version though and I have to recompile sources.
Jan 16 05:44:14 <twb> Why?
Jan 16 05:44:21 <AdamOutler> because the drivers are different.
Jan 16 05:44:34 <twb> Uh, so just compile the drivers
Jan 16 05:45:33 <AdamOutler> well, there's a caviet...  it's a device based on blaze, it's actually a Nook Tablet...  I located a security hole and it will only work for boot.img files.  I need to use my own kernel on it.
Jan 16 05:46:24 <AdamOutler> so, I'd like to find a guide on compiling a kernel from scratch and making a boot.img.
Jan 16 05:46:36 <twb> I don't know what you mean by boot.img
Jan 16 05:47:05 <AdamOutler> boot.img is Android standard..   basically it has to authenticate with the bootloader before it can be loaded.
Jan 16 05:47:16 <twb> OK, a blobtools boot.img
Jan 16 05:47:24 <twb> Er, I mean abootimg
Jan 16 05:47:30 <AdamOutler> kinda.
Jan 16 05:47:32 <AdamOutler> mkbootimg
Jan 16 05:47:45 <twb> The tool I have seen that builds those is called "abootimg", apt-get install it
Jan 16 05:47:57 <AdamOutler> I need some sort of guide though.
Jan 16 05:48:31 <AdamOutler> ah ha!  abootimg is a good tool
Jan 16 05:48:38 <twb> To build the kernel, get an environment that can compile for your target architecture, then do "make zImage" or "make deb-pkg" to taste
Jan 16 05:49:08 <twb> It would be a good idea to start with the stock ubuntu kernel and its omap4 .config
Jan 16 05:49:10 <AdamOutler> twb,  and abootimg handles zImage files?
Jan 16 05:49:19 <twb> AdamOutler: I don't remember offhand
Jan 16 05:49:26 <AdamOutler> ok, twb, what is a uImage?
Jan 16 05:49:50 <twb> Different flavour of zImage, it's for u-boot
Jan 16 05:50:03 <twb> If you are using the android bootloader (which takes abootimg boot.img files), then you don't want uImage.
Jan 16 05:50:49 <AdamOutler> I have uImage and uInitrd..  they need to get into a boot img...  I found a program that converts uImage to zImage, and now I need to convert uInitrd to initrd I guess.
Jan 16 05:50:49 <twb> You probably want to prepare an Ubuntu rootfs, generate a kernel .deb, install the latter into the former, then run abootimg on the resulting kernel and ramdisk from /boot
Jan 16 05:51:12 <twb> AdamOutler: if you already have a uImage why are you asking about kernel compilation?
Jan 16 05:51:24 <AdamOutler> twb.  you're opening my eyes.
Jan 16 05:51:32 <AdamOutler> :)
Jan 16 05:51:58 <AdamOutler> twb.. I've been compiling non-stop for a few days,  I have several test images to work with in various formats.
Jan 16 05:52:18 <twb> What hardware are you compiling on?
Jan 16 05:52:23 <AdamOutler> stupid boot.img is tearing me up.
Jan 16 05:52:51 <AdamOutler> twb, I'm cross-compiling from a i5 quad core for ARM arch
Jan 16 05:53:00 <twb> That should compile in deciminutes, not days
Jan 16 05:53:23 <AdamOutler> twb, it's operator error... this is why I asked for a guide when I came in ;)
Jan 16 05:53:25 <twb> Try turning off the debugging symbols
Jan 16 05:53:51 <twb> AdamOutler: it would be a good idea for you to get comfortable building normal x86 kernels first
Jan 16 05:54:07 <AdamOutler> twb I am comfortable building 86_64
Jan 16 05:54:08 <twb> AdamOutler: that way you will have a solid foundation when you try to get a handle on the arm randomness
Jan 16 05:54:50 <AdamOutler> twb, I've been operating several linux boxes in my home for years.  I've compiled several for my media center alone.
Jan 16 05:55:33 <AdamOutler> I'd just like some knoledge about all these formats and everything..  On my Samsung, I could just flash a zImage to my device and I was done..  This boot.img format is crazy.
Jan 16 05:56:04 <twb> IIUC samsungs use u-boot and are generally on the "less stupid" side of ARM'd spectrum of idiocy
Jan 16 05:56:19 <AdamOutler> zImage has Ramdisk+kernel.   For some reason Boot.img has zImage + ramdisk, + android headers?
Jan 16 05:56:39 <twb> zImage does not normally include the ramdisk
Jan 16 05:56:53 <twb> The kernel has support for tacking a ramdisk onto itself -- this should work in both cases
Jan 16 05:57:04 <twb> Normally the ramdisk is a separate file.
Jan 16 05:57:34 <AdamOutler> ok, and what is the point of a ramdisk when the files are already on the root of the storage medium?
Jan 16 05:57:36 <twb> To generate a boot.img using abootimg, you need a kernel, a ramdisk (optional), and a config file that basically consists of the options passed at the boot: prompt
Jan 16 05:58:09 <twb> AdamOutler: the primary reason to use a ramdisk, is so you can build a small portable kernel and then include device-specific drivers to *get to* the root fliesystem, in the ramdisk
Jan 16 05:58:28 <twb> That is how e.g. Ubuntu x86 servers all share the same kernel despite having different SATA controllers and video cards
Jan 16 05:59:01 <twb> You can also put various other clever code in the ramdisk, e.g. this is how live CDs are made to look writable
Jan 16 05:59:47 <twb> There is probably a paper somewhere that describes the linux boot process in detail; I don't have one handy to recommend.
Jan 16 06:00:01 <AdamOutler> I think this is what I need.
Jan 16 06:00:24 <AdamOutler> the paper describing the Uboot/Linux boot process.  I really need to sort out these formats.
Jan 16 06:01:01 <AdamOutler> you're making me much smarter though twb.  I appreciate it.
Jan 16 06:01:07 <AdamOutler> may I ask what you do for a living?
Jan 16 06:01:43 <twb> http://prisonpc.com/
Jan 16 06:07:03 <AdamOutler> thanks for your help.
Jan 16 06:09:38 <AdamOutler> twb..  one more thing before I go..  On the pre-compiled version, the touchscreen is 90 degrees off, and the wifi is totally inop.   Any suggestions before I go trying to mix and match and recompile a few more things?
Jan 16 06:09:48 <AdamOutler> https://plus.google.com/u/0/104711040110222472212/posts
Jan 16 06:10:09 <twb> xrandr may help the latter if you are using X
Jan 16 06:10:15 <twb> Er, the former
Jan 16 06:10:25 <twb> wifi will probably be a missing firmware blob or soemthing
Jan 16 06:10:54 <twb> http://openbsd.org/lyrics.html#39
Jan 16 06:13:32 <twb> You'll need to find someone who knows about your hw
Jan 16 06:14:37 <AdamOutler> ok..  I found out about the xrandr thing, but I'm having a problem getting getty to work.  I have a UART console hooked up, but I can't get shell access no matter what I do.
Jan 16 06:15:04 <AdamOutler> I think it may have to do with the kernel being compiled with ttyS2 instead of ttyO2 for the default?
Jan 16 06:15:12 <AdamOutler> I'm monitoring on ttyO2
Jan 16 06:15:26 <twb> Never heard of ttyO2
Jan 16 06:15:40 <twb> I've only seen ttyACM0 or so, plus the usual tty1 and ttyS0
Jan 16 06:15:45 <AdamOutler> /dev/ttyO2 is Texas Instrument's serial administration console.
Jan 16 06:15:50 <twb> ok
Jan 16 06:15:56 <AdamOutler> On samsung devices it's /dev/ttySAC2
Jan 16 06:16:28 <twb> Well, you can just pass console=/dev/ttyO2
Jan 16 06:16:33 <twb> On the boot: prompt
Jan 16 06:16:39 <AdamOutler> I can make this device and get output from it.. I just can't seem to send any input.  this would solve about 99% of my problems.
Jan 16 06:16:45 <twb> Also in /etc/init/ there is a ttyN.conf you need to make one for O2
Jan 16 06:16:58 <AdamOutler> I can't pass arguments.
Jan 16 06:17:24 <twb> What output are you seeing?  The linux kernel boot messages?  A tty login prompt?  Both?
Jan 16 06:19:14 <AdamOutler> I'm only getting kernel boot messages.   I have tried several things including creating a S80StartConsole in /etc/RCS.d, creating my device on the /dev/ folder with the proper chmod params, and creating a /etc/init/ttyO2.conf
Jan 16 06:19:57 <twb> you probably got ttyO2.conf wrong, pastebin it
Jan 16 06:20:35 <AdamOutler> http://pastebin.ubuntu.com/805911/
Jan 16 06:21:07 <twb> Assuming the baud rate is right, looks OK to me.
Jan 16 06:21:16 <AdamOutler> I've tried with those params and a -L instead of 8n
Jan 16 06:21:17 <twb> Oh, and "rc RUNLEVEL=..." looks wrong.
Jan 16 06:21:45 <twb> Never mind, it looks like that's the style used for other getty .conf
Jan 16 06:21:57 <twb> I'm used to "start on runlevel [2345]" and "stop on runlevel [^2345]"
Jan 16 06:22:19 <twb> I'm used to "start on runlevel [2345]" and "stop on runlevel [!2345]" (braino)
Jan 16 06:23:16 <AdamOutler> I'll try that.
Jan 16 06:23:19 <AdamOutler> next time.
Jan 16 06:24:14 <AdamOutler> my s80StartConsole contains "start ttyO2" and a bunch of other things.  I have determined it is run as root so I have a way to make things happen, I just don't know what needs to happen.
Jan 16 12:31:18 <S0NiC> hi
Jan 16 12:32:20 <S0NiC> hmm is it possoible to play on a pandaboard any mp4 video? is this supported?
Jan 16 12:32:47 <taruti> S0NiC: just use mplayer?
Jan 16 12:33:50 <ogra_> only supported in totem
Jan 16 12:34:29 <LetoThe2nd> ogra_: depends on the desired playback speed ;)
Jan 16 12:34:37 <ogra_> mplayer would only try to do it in sw ... all gstreamer based players can use the hw acceleration of the panda (if you have the codecs installed=
Jan 16 12:35:00 <S0NiC> ogra_: wich codecs would that be?
Jan 16 12:35:12 <ogra_> LetoThe2nd, indeed, i should have said that mplayer is great for turning mp4 movies into slideshows instead :)
Jan 16 12:35:49 <ogra_> S0NiC, just install the ubuntu-omap4-extras or the ubuntu-omap4-multimedia packages after you enabled the PPA
Jan 16 12:35:50 <LetoThe2nd> S0NiC: maybe have a good look at http://rsalveti.wordpress.com/2012/01/06/hw-video-decode-and-xbmc-ubuntu-linaro/
Jan 16 12:36:07 <LetoThe2nd> ogra_: cool use case that is ;)
Jan 16 12:36:14 <ogra_> (they get installed by default after you clicked the TI desktop item on the images)
Jan 16 12:36:39 <ogra_> LetoThe2nd, well, ubuntu-tv extensively makes use of xbmc ;)
Jan 16 12:36:45 <S0NiC> thx
Jan 16 12:36:55 <LetoThe2nd> ogra_: havent gotten round to try it yet :/
Jan 16 12:36:57 <S0NiC> hmm wich kind of video works out of the box?
Jan 16 12:37:12 <S0NiC> *codec
Jan 16 12:39:41 <S0NiC> does anybody know?
Jan 16 12:40:34 <LetoThe2nd> S0NiC: any video works out of the box, but none with hw accel. thats why the ppa is there.
Jan 16 12:41:12 <S0NiC> LetoThe2nd: ok that means i have to enable hw acceleration. what means ppa in this case?
Jan 16 12:42:30 <LetoThe2nd> S0NiC: basically http://omappedia.org/wiki/PandaBoard_Ubuntu_PPA, no idea how uptodate the instructions are.
Jan 16 12:43:48 <ndec> the codecs which are h/w accelerated on OMAP4 are MPEG2, MPEG4, H263, H264, VC1 (WMV)
Jan 16 12:43:50 <S0NiC> LetoThe2nd: but i think your link http://rsalveti.wordpress.com/2012/01/06/hw-video-decode-and-xbmc-ubuntu-linaro/ is good. but i have to create a new image...
Jan 16 12:44:24 <ndec> just start with the default 11.10, upgrade it, reboot, and install the TI adds on.
Jan 16 12:44:31 <ndec> that should be a good start.
Jan 16 12:45:00 <S0NiC> ndec: mom
Jan 16 12:46:02 <S0NiC> ndec: ok, iam using 10.10 at the moment...
Jan 16 12:46:09 <ndec> ouch...
Jan 16 12:46:09 <LetoThe2nd> urgh
Jan 16 12:46:19 <ndec> any good reason to stick to 10.10?
Jan 16 12:46:29 <LetoThe2nd> "ancient technologies and wisdom of the olde"?
Jan 16 12:46:38 <S0NiC> ndec: no
Jan 16 12:46:40 <ndec> 11.10 has better OMAP support, and more codecs. and is maintained in case you face an issue
Jan 16 12:46:41 <S0NiC> ok, dann i update,
Jan 16 12:46:55 <S0NiC> what is ment by the "ti adds"
Jan 16 12:46:58 <S0NiC> ?
Jan 16 12:47:11 <ndec> hum, my advice is to restart from scratch, don't dist-upgrade a 10.10 system into 11.10.
Jan 16 12:47:22 <LetoThe2nd> S0NiC: the ti addons for hw accel
Jan 16 12:47:39 <ndec> once you install 11.10 image, there will be an icon to install the TI 'addons' on the desktop. just click
Jan 16 12:47:40 <S0NiC> i get them with apt-get....?
Jan 16 12:47:45 <ndec> yes you can too.
Jan 16 12:47:47 <S0NiC> ndec: ah cool
Jan 16 12:47:51 <LetoThe2nd> S0NiC: have you actuelly read the links?
Jan 16 12:47:51 <S0NiC> thanks guys
Jan 16 12:48:07 <S0NiC> LetoThe2nd: only your link
Jan 16 12:48:16 <S0NiC> now i have a lot to do ;D
Jan 16 12:48:17 <ndec> but make sure that you run apt-get update & apt-get dist-upgrade before you install the TI stuff. some people have reported that there are issues otherwise
Jan 16 12:48:34 <LetoThe2nd> S0NiC: i doubt it, because then you would know how to install them ;)
Jan 16 12:51:08 <LetoThe2nd> one could also aim directly for the tv release that ricardo pointed to...
Jan 16 12:52:24 <S0NiC> ok
Jan 16 12:52:27 <S0NiC> thx
Jan 16 12:52:56 <S0NiC> ok i have to leave cya later...
Jan 16 12:52:58 <S0NiC> thx bye
Jan 16 12:54:43 <LetoThe2nd> side note: his original question was in #ubuntu-de why 'apt-cache search vlc' did not yield any results :)
Jan 16 12:55:25 <ogra_> oh, it should
Jan 16 12:56:01 <LetoThe2nd> ogra_: on ancient 10.10, and a user that doesn't know how to handle a ppa? ;)
Jan 16 12:56:02 * ogra_ gets more than a terminal page 
Jan 16 12:56:27 <ogra_> doesnt matter, vlc is in universe, he should even have seen it on 10.10
Jan 16 12:56:35 <ndec> i guess he didn't enable universe
Jan 16 12:56:43 <ogra_> yeah, smells like
Jan 16 12:56:54 <ogra_> and 10.10 had a bug with the default sources.list iirc
Jan 16 12:57:01 <ogra_> which we fixed in an update
Jan 16 12:57:15 <LetoThe2nd> smells like nirvana... erm, teen spirit ;)
Jan 16 12:57:19 <ogra_> hehe
Jan 16 12:58:01 * ogra_ had totally forgotten ubuntu-de ... i used to be resident there ... but i find IRC in german weird and have probs to not write half my sentences in english
Jan 16 12:59:16 <LetoThe2nd> ogra_: the traffic there receded quite a bit over the last 3 or 4 years, after we enforced more and more to stay on topic, and in the meantime rejected to support weird offshoots/derivates/whatevers ;)
Jan 16 12:59:52 <ogra_> yeah, i remember that kind of discuaaions
Jan 16 12:59:55 <LetoThe2nd> ogra_: you wouldn't believe how many people come there bit***ing around when you reject to support mint or debian.
Jan 16 13:00:25 <ogra_> oh, i would :) i see the same bitching in bugs ...
Jan 16 13:00:38 <LetoThe2nd> ogra_: hehe
Jan 16 13:01:01 <ogra_> rarely debian, but often mint users freak out that we dont support their ubiquity bugs for the hacked up mint installer
Jan 16 13:01:25 <LetoThe2nd> ogra_: it has emerged that when the slightest doubts arise we ask for lsb_release -a and uname -a.
Jan 16 13:01:29 <ogra_> and mint does serious bad things when customizing the distro
Jan 16 13:01:43 <LetoThe2nd> ogra_: yeah, thats why we reject to support it.
Jan 16 13:01:57 <ogra_> like disabling security updates for things like X11, kernel, dbus etc
Jan 16 13:02:20 <ogra_> (or bootloaders)
Jan 16 13:02:38 <ogra_> s/security updates/all updates/
Jan 16 13:03:06 <LetoThe2nd> ogra_: hehe
Jan 16 13:04:15 <LetoThe2nd> lets be glad they don't have an arm port yet.
Jan 16 16:04:26 <hrw> heartbeat led trigger is not in kernel in 3.2.0/omap4
Jan 16 16:14:17 <rsalveti> hrw: config issue I believE?
Jan 16 16:18:50 <hrw> yes - it is a module now
Jan 16 16:19:09 <hrw> I was surprised when one of my pandas stopped blinking
Jan 16 16:20:38 <GrueMaster> It's a module?  figures.  Hard to track when I have 4 pandas in a tower.
Jan 16 16:22:28 <hrw> ledtrig-heartbeat
Jan 16 17:37:37 <S0NiC> re guys
Jan 16 17:38:05 <S0NiC> so i installed ubuntu 11.10 and doing an apt-get update /dist upgrade
Jan 16 17:38:27 <S0NiC> but there is no icon with Ti stuff... i only have installOAMP4addons...
Jan 16 17:38:29 <S0NiC> is taht the same=?
Jan 16 17:39:18 <ndec> yes
Jan 16 17:42:18 <S0NiC> ndec: thx
Jan 16 18:27:57 <vani> hello what's the arm ports repo for precise
Jan 16 18:31:49 <vani> how can I point to the repos, I want to use live-build
Jan 16 18:33:28 <rbasak> deb http://ports.ubuntu.com/ubuntu-ports precise main
Jan 16 18:34:50 <vani> thank you.
Jan 16 18:58:40 <NotJimCarrey> how do you add eabi support?
Jan 16 20:20:10 <AdamOutler> HI!
Jan 16 20:20:36 <AdamOutler> I've sucessfully created a boot.img that apparently works, however I am running into an error.   "Unrecognized/unsupported machine ID"..
Jan 16 20:20:43 <AdamOutler> Is there a way to change this in my kernel?
Jan 16 20:20:57 <AdamOutler> I'm using precompiled binaries
Jan 16 20:21:22 <AdamOutler> Here is the pastebin: http://pastebin.com/7cBB05KF
Jan 16 21:20:37 <vani> <AdamOutler> sorry I cannot help you on this as I'm new to ARM ports. But, I'm trying to piggy back on you.
Jan 16 21:20:54 <vani> How did you create the boot.img...
Jan 16 21:21:17 <vani> any notes or references...I'm having a board with ARMv7
Jan 16 21:21:40 <vani> trying to build a custom image for it...
Jan 16 21:26:04 <LetoThe2nd> the machine ID refers to the ID that the first/second stage bootloader passes to the kernel. the kernel uses this ID to decide which board file to use.
Jan 16 21:26:35 <LetoThe2nd> if the kernel doesn't know about the ID, it will not boot. so make sure kernel and bootloader (u-boot in most cases) match.
Jan 16 22:16:55 <LetoThe2nd> the machine ID refers to the ID that the first/second stage bootloader passes to the kernel. the kernel uses this ID to decide which board file to use.
Jan 16 22:17:08 <LetoThe2nd> if the kernel doesn't know about the ID, it will not boot. so make sure kernel and bootloader (u-boot in most cases) match.
Jan 16 22:17:13 <LetoThe2nd> AdamOutler: ^^^^^^^^^^
Jan 16 22:17:39 <AdamOutler> oh
Jan 16 22:17:40 <AdamOutler> hi
Jan 16 22:18:01 <LetoThe2nd> concerning your question from some time ago.
Jan 16 22:18:20 <AdamOutler> ok... so can I spoof this from U-Boot prompt somehow?
Jan 16 22:18:29 <LetoThe2nd> hopefully not.
Jan 16 22:18:53 <AdamOutler> LetoThe2nd, this is a Blaze Tablet based device.
Jan 16 22:19:10 <AdamOutler> LetoThe2nd, I'm pretty sure I should be able to get something from the device by spoofing the id
Jan 16 22:19:46 <LetoThe2nd> AdamOutler: its three parts altogether 1) get/register the machine ID at arm.linux.co.uk 2) make sure the bootloader passes this ID 3) make sure the kernel contains a board support file bound to this id.
Jan 16 22:51:45 <AdamOutler> LetoThe2nd, the link you provided is bad.
Jan 16 22:52:19 <LetoThe2nd> AdamOutler: the link was just from memory please google the correct site.
Jan 16 23:12:46 <rsalveti> hrw: it should be built-in to work during boot I
Jan 16 23:13:06 <rsalveti> something to ping ppisati to change
Jan 16 23:24:36 <mythos> register what?
Jan 16 23:39:55 <mythos> is it possible to start a dovel-kernel with qemu?
Jan 16 23:44:17 <GrueMaster> mythos: A which kernel?  I think qemu wants a versatile kernel only.
Jan 16 23:45:39 <mythos> yes, versatile runs perfectly. are you sure, that those dove-kernels don't run with qemu - i speak from those in maverick
Jan 16 23:45:43 <mythos> 10.10
Jan 16 23:46:06 <mythos> and btw thanks for your response, GrueMaster =)
Jan 16 23:46:11 <GrueMaster> The dove kernels are specifically for Marvell hardware.
Jan 16 23:46:44 <GrueMaster> Same as the omap4 kernels are for TI omap4 processors.
Jan 16 23:47:08 <mythos> hmm... ok, i see
Jan 16 23:47:32 <mythos> good to know why qemu pretends to stay black =)
Jan 16 23:47:34 <GrueMaster> And the Marvell kernel for Maverick was a bit of a mess anyways.  It was the same as the Lucid kernel, just built with the Maverick tool chain.
Jan 16 23:48:34 <mythos> i have a hp device here, which uses a maverick-debootstrab
Jan 16 23:48:51 <mythos> it has marvell-hardware in it, so that fits
Jan 16 23:48:54 <GrueMaster> Is it using a marvel soc?
Jan 16 23:48:58 <GrueMaster> Cool.
Jan 16 23:49:22 * NCommander screams in horror
Jan 16 23:49:36 <GrueMaster> heh
Jan 16 23:49:45 <NCommander> mythos: is that a thinclient by any chance?
Jan 16 23:49:54 <mythos> NCommander, yes
Jan 16 23:50:00 <mythos> a hp 5335z
Jan 16 23:50:03 <NCommander> mythos: ....
Jan 16 23:50:23 <mythos> i broke yesterday one
Jan 16 23:50:46 <mythos> was not able to boot up anymore, after i zeroed /dev/sda
Jan 16 23:50:51 <mythos> ^^"
Jan 16 23:52:06 <GrueMaster> Yea, that would be a bad thing.
Jan 16 23:52:45 <NCommander> mythos: I loathe to say this, but that thing is based around a Marvell Dove board. If they took my patches wholesale, it *might* boot with an Ubuntu maverick dove image off USB
Jan 16 23:52:56 <mythos> no, it should boot up from a usb-device. but it simply did not
Jan 16 23:53:10 <NCommander> mythos: open it up and find the serial header to get uboot access :-)
Jan 16 23:53:14 <NCommander> mythos: what image did you try?
Jan 16 23:53:57 <mythos> NCommander, i only messed around with the the image, which is downloadable from hp
Jan 16 23:54:13 * GrueMaster watches NCommander's version of the crying game, laughing hysterically.
Jan 16 23:54:34 <NCommander> mythos: http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/maverick/release/
Jan 16 23:54:40 <NCommander> try one of the dove images, and see if it boots
Jan 16 23:54:50 <NCommander> instructions are ... somewhere
Jan 16 23:54:59 <NCommander> ^- GrueMaster
Jan 16 23:56:04 <mythos> NCommander, i'm going to try that =)
Jan 16 23:56:16 <NCommander> mythos: if it works, please don't call me ;.;
Jan 16 23:56:24 * NCommander horrorifically flashs back to China
Jan 16 23:56:31 <mythos> don't?
Jan 16 23:56:46 <NCommander> mythos: I was assigned to Antartica^W China for a month to do that image
Jan 16 23:56:47 <mythos> ok, i won't
Jan 16 23:56:52 <NCommander> I'm still scared from the experience
Jan 16 23:57:01 <mythos> omg ^^"
Jan 16 23:57:13 <NCommander> mythos: (actually, feel free to ping me, but what support I can give is limited)
Jan 16 23:58:03 <mythos> my plan was to use the hp-image as skeleton
Jan 16 23:58:06 <NCommander> mythos: alternatively, you might need to open it up, pop the HDD, and then do some creative repartitioning
Jan 16 23:58:32 <mythos> hdd? there is no hdd inside them ^^"
Jan 16 23:58:32 <NCommander> what compelled you to zero sda though?
Jan 16 23:58:42 <NCommander> oh good, they fixed that 'bug'
Jan 16 23:59:19 <mythos> omg xD
Jan 17 00:00:10 <NCommander> mythos: so depending on how they smacked uboot, the thing will either try and load an image off USB, or try and TFTP a blob off somewhere (you'll probably need to use ethereal to see where its trying to get a blob)
Jan 17 00:00:10 <twb> NCommander: why did you need to go to .cn to make an ISO9660?
Jan 17 00:00:21 <mythos> even though it is limited, that offer is very nice, NCommander =)
Jan 17 00:00:30 <twb> ethereal is called wireshark these days btw
Jan 17 00:00:41 <NCommander> twb: old names die hard
Jan 17 00:00:54 <NCommander> twb: and thats a story that can only be told with loads of booze and a NDA :-/
Jan 17 00:01:06 <twb> Ugh
Jan 17 00:01:20 * NCommander actually loved China TBH
Jan 17 00:01:24 <twb> Embedded hw vendors need to get a damn clue
Jan 17 00:01:33 <NCommander> I didn't like having four days notice going to china for a month
Jan 17 00:01:46 <mythos> NCommander, i already read out those u-boot parameters and wondered about those ip-addresses...
Jan 17 00:01:50 <twb> You managed to get a visa in four days?  Wow.
Jan 17 00:02:00 <NCommander> twb: heh, I already had the visa
Jan 17 00:02:12 <NCommander> Which was obtained for the same reason in the lucid cycle
Jan 17 00:02:30 <twb> Are you a canonical flunky?
Jan 17 00:02:36 <NCommander> twb: yes :-P
Jan 17 00:02:48 * twb adds NCommander to list
Jan 17 00:02:55 * NCommander is listed
Jan 17 00:03:32 <NCommander> mythos: if you got the u-boot paramters, do you have the bootcmd the board uses?
Jan 17 00:03:43 <NCommander> that might give me a better idea on a scale of 1-5 on how screwed you are
Jan 17 00:03:56 <mythos> NCommander, i'm not sure
Jan 17 00:04:11 <mythos> but i'm going to nopaste it
Jan 17 00:04:26 <NCommander> how'd you fish out the uboot paramaters? I never got uboot-tools to work properly (granted, I didn't try very hard)
Jan 17 00:06:20 <mythos> NCommander, http://pastebin.de/22381
Jan 17 00:07:06 <mythos> i booted up up with a usb-device dded /dev/mtd1
Jan 17 00:07:18 <NCommander> mythos: oh, so it boots off USB?
Jan 17 00:07:23 <NCommander> Sounds like ytour less screwed
Jan 17 00:07:30 <mythos> not quite stable, but yes
Jan 17 00:07:38 <NCommander> But yeah, thats a completely custom boot command compared to what I wrote for Ubuntu
Jan 17 00:09:08 <twb> NCommander: what'd you write, the u-boot script?
Jan 17 00:09:30 <NCommander> twb: basically hacked the bootloader to do something sane
Jan 17 00:09:46 <NCommander> it probes all devices on USB/SATA, and looks for a boot.scr, then chainloads into that
Jan 17 00:10:09 <twb> I guess probing mtd as well was nontivial?
Jan 17 00:10:10 <mythos> the device has no initrd... the init is on /dev/sda1 and chroots into a sqashfs-filesystem
Jan 17 00:10:18 <NCommander> twb: we didn't support it
Jan 17 00:10:22 <mythos> i never saw anything like that
Jan 17 00:10:26 <NCommander> Still don't support it actually (installer limitation)
Jan 17 00:10:43 <twb> As in d-i can't install to mtd?
Jan 17 00:11:03 <NCommander> twb: yeah. Its getting support via ubifs, but that still requires bootloader support to some extent
Jan 17 00:11:10 <twb> I thought ubuntu arm installs were all still "dd this image to <magic place> and hope for the best"
Jan 17 00:11:23 <NCommander> No, we had live and alternate and netboot for Dove
Jan 17 00:11:29 <twb> Cool
Jan 17 00:11:30 <NCommander> Panda is a new type of image called a preinstall
Jan 17 00:11:36 * NCommander implemented that deeper voodo
Jan 17 00:11:40 <NCommander> *voodoo
Jan 17 00:11:49 <twb> "preinstall" sounds like dd-and-hope to me ;-P
Jan 17 00:12:28 <GrueMaster> Pretty much.  The main idea was that beagle/panda boots from removable SD.
Jan 17 00:13:03 <GrueMaster> And ubiquity has a hard time repartitioning and reimaging the device it boots from.
Jan 17 00:13:04 <NCommander> twb: pandas can only boot from SD*
Jan 17 00:13:21 <NCommander> * - technically USB too, but thats not useful for standalone systems
Jan 17 00:13:33 <GrueMaster> No, that just is the secondary boot device.  It can also boot from USB-Gadget.
Jan 17 00:13:57 <GrueMaster> And it can be modified to boot from other devices.
Jan 17 00:14:08 <NCommander> GrueMaster: that's what I meant when I said can boot from USB
Jan 17 00:14:16 <GrueMaster> But that requires a soldering iron and steady hand.
Jan 17 00:14:17 <NCommander> it can't load its bootloader from USB :-P
Jan 17 00:14:19 <twb> GrueMaster: that's because ubiquity doesn't run 100% out of ram like d-i does
Jan 17 00:14:23 <twb> boo ubiquity, stupid GUI junk
Jan 17 00:14:32 <NCommander> twb: you can run ubiquity in 364MiB
Jan 17 00:14:37 <NCommander> We did that in jaunty for imx51
Jan 17 00:14:40 <NCommander> (karmic too)
Jan 17 00:14:57 <NCommander> and d-i can be a serious RAM hog if it can
Jan 17 00:14:59 <GrueMaster> twb: It is more than just ubiquity.  The packages would also be on the boot device.
Jan 17 00:15:00 <mythos> NCommander, so, if i plug a hdd on the board, it would be able to boot from that?
Jan 17 00:15:20 <NCommander> (d-i will usually OoM if it doesn't have at least 128 MiB in total RAM+swap)
Jan 17 00:15:21 <twb> NCommander: what I mean is if I bootload the d-i netboot kernel and initrd, then I can do the install and blow away the boot medium entirely
Jan 17 00:15:48 <NCommander> twb: yes (with the cavet that the installer doesn't fully enforce panda's unique boot partition requirements)
Jan 17 00:15:49 <twb> NCommander: admittedly, the initrd needs a driver for the NIC and you can't use wpa yet
Jan 17 00:15:59 <NCommander> twb: no, we ship the NIC driver.
Jan 17 00:16:03 <GrueMaster> twb: That is different.  You aren't installing packages from the boot device to the repartitioned boot device with netboot.
Jan 17 00:16:06 <NCommander> twb: GrueMaster uses netboot all the time
Jan 17 00:16:52 <twb> GrueMaster: right, but what I *do* do sometimes, is put netinst d-i in /boot, point grub at it, then blow away the boot disk with a fresh install
Jan 17 00:17:08 <twb> ubiquity can't do that, which suck
Jan 17 00:17:10 <twb> ...s
Jan 17 00:18:03 <GrueMaster> twb: Do you know the kind of overhead to do ubiquity netboot?  Can't do that type of install on any Linux distro that I know of.
Jan 17 00:18:14 <twb> Exactly
Jan 17 00:18:49 <NCommander> twb: ubiquity is a serpate codebase from d-i that exists to copy a squashfs into target filesystem, even if you could load it over netboot, you'd still have to feed a squashfs to it
Jan 17 00:19:02 <GrueMaster> ubiquity is X based.  So is the Redhat installer (name escapes me atm) and the Suse installer.
Jan 17 00:19:05 <NCommander> twb: d-i is fully supported for those who want/need it (I always use d-i to reinstall my laptop)
Jan 17 00:19:17 <NCommander> and we have d-i gtk-installer if you really want flashy graphics
Jan 17 00:19:29 <NCommander> (though I don't think we build gtk-installer out of the box for ubuntu)
Jan 17 00:19:49 <twb> You don't as at lucid, anyway
Jan 17 00:19:57 <twb> (re gtk d-i)
Jan 17 00:20:04 <mythos> <NCommander> what compelled you to zero sda though? <-- i wanted to try, if it is able to boot after that. it did once, than it only beeps twice and does nothing
Jan 17 00:20:26 <mythos> sorry, i had to read it five-times to understand, what you meant
Jan 17 00:28:16 <GrueMaster> mythos: Wow.  I just downloaded the recovery image for your hp, and I am appalled at how they mucked it together.
Jan 17 00:30:02 <GrueMaster> Essentially, it looks like you format a USB drive, unzip the contents to it and reboot on your PC (not the thin client) to the USB stick.  It will boot a small linux image that on init will reflash the usb drive with the thin client image (which itsself is based on our Maverick image).
Jan 17 00:31:02 <mythos> yes
Jan 17 00:31:59 <mythos> indeed i reversed it to that state
Jan 17 00:32:03 <GrueMaster> If you want to speed up the process (and have less of a chance of clobbering your PC), use win32-image-writer.  It is a windows program designed for this purpose.
Jan 17 00:32:20 <twb> This is why we have FOSS
Jan 17 00:32:28 <twb> So that we don't need to deal with that kind of vendor bullshit
Jan 17 00:33:37 <mythos> GrueMaster, i don't know for what the tool is for
Jan 17 00:34:12 <mythos> GrueMaster, fdisk, mkfs.vfat and unzip is quite easy, so ^^"
Jan 17 00:35:02 <GrueMaster> If you are on a windows pc and you want to flash the recovery image for your HP POS^h Thin client, use win32-image-writer to flash the recovery image to the usb stick instead of their muck-around.
Jan 17 00:35:29 <GrueMaster> Yes, but then it requires you to reboot your PC to the usb stick.
Jan 17 00:35:43 <GrueMaster> Why reboot your PC?
Jan 17 00:36:19 <twb> It also requires that you still *ahve* an x86 pc
Jan 17 00:37:29 <GrueMaster> There is a compressed image in their zipfile at images/Z5D40019.dd.gz.  Uncompress that, and use win32-image-writer (windows) or dd (linux) to write that raw image to a usb stick, put it in the thin client and boot.
Jan 17 00:38:26 <GrueMaster> twb: Most companies deploying these thin clients will have PCs (usually laptops) for their IT support staff.
Jan 17 00:38:46 <mythos> GrueMaster, to whom are you responsing. for me, it is like i have someone on ignore <.<
Jan 17 00:38:51 <GrueMaster> These are mainly sold as kiosk or point of sale systems.
Jan 17 00:39:18 <GrueMaster> mythos: I was replying to twb.
Jan 17 00:39:19 <twb> I give all my support staff the same shitty thin clients as the prisoners
Jan 17 00:39:23 <twb> FWIW
Jan 17 00:39:55 <mythos> GrueMaster, hmm... ok... but i check my ignorelist anyway ^^"
Jan 17 00:40:25 <twb> mythos: probably my anti-ubiquity ranting :-)
Jan 17 00:41:34 <mythos> twb, maybe :o
Jan 17 00:43:47 <GrueMaster> Sadly I don't have an ignorefile.  I have to go on the basis that even trolls need help sometimes.  :P
Jan 17 00:45:38 <mythos> you are a very nice guy, with nervs out of steel, GrueMaster ^^"
Jan 17 00:54:50 <mythos> oh lucky, this channel is logged
Jan 17 00:55:01 <GrueMaster> yep
Jan 17 00:55:59 <mythos> then i don't need the verfification-client anymore
Jan 17 01:11:54 <mythos> hmm... the initrd in this dove-image http://cdimage.ubuntu.com/ubuntu-netbook/ports/releases/maverick/release/ (inside the squashfs) is a dead link
Jan 17 01:12:26 <GrueMaster> Yes, it is generated on install.
Jan 17 01:13:30 <mythos> ok
Jan 17 01:13:40 <mythos> but one question, if you don't mind
Jan 17 01:14:23 <mythos> how is this "/casper/uInitrd: u-boot/PPCBoot image" created and is it modifyable?
Jan 17 01:14:41 <GrueMaster> I'm listening (responding is still optional).  :P
Jan 17 01:15:02 <mythos> *g
Jan 17 01:15:12 <GrueMaster> The image was generated with our live-cd tools back when we were making them.
Jan 17 01:15:25 <GrueMaster> I think we have moved to a new tool base now.
Jan 17 01:15:58 <GrueMaster> As to modifying it, what needs to be modified?
Jan 17 01:16:10 <mythos> i only want to look inside =)
Jan 17 01:16:14 <GrueMaster> it's possible, but not easy.
Jan 17 01:16:17 <GrueMaster> Ah.
Jan 17 01:17:04 <GrueMaster> I have scripts for that.
Jan 17 01:17:46 <mythos> if you give me some hints, i'm going to study it for myself =)
Jan 17 01:17:53 <GrueMaster> mkdir initrd
Jan 17 01:17:53 <GrueMaster> dd if=$1 skip=64 bs=1|zcat | (cd initrd;sudo cpio -id)
Jan 17 01:18:33 <mythos> oh, that helps a lot, thx
Jan 17 01:18:58 <GrueMaster> That "may" work.  Depends on how much u-boot header cruft is in the uInitrd and also if the initrd is gzipped.
Jan 17 01:20:03 <GrueMaster> I've seen the header be as much as 72 bytes (skip=72) so play with it.  Make a backup first.
Jan 17 01:22:34 <mythos> ok
Jan 17 01:22:39 <mythos> nice, thanks =)
Jan 17 01:23:43 <twb> Is there a functional difference between uinitrd and a normal initrd?
Jan 17 01:24:35 <twb> Because if not the obvious way to build uinitrd (for boot, not install) would be update-initramfs
Jan 17 01:25:32 <GrueMaster> twb: uInitrd are checksum signed for u-boot.
Jan 17 01:25:50 <GrueMaster> same with uImage and boot.scr
Jan 17 01:26:08 <twb> Ah, OK.
Jan 17 01:26:47 <twb> How does u-boot know which key to check it against?  Wouldn't that be device/vendor specific?
Jan 17 01:27:21 <GrueMaster> twb it is a checksum, not a signed key.
Jan 17 01:27:49 <GrueMaster> Although it can be signed by the vendor (aka locked boot).
Jan 17 01:27:52 <twb> Ah, I misunderstood.  I thought you were saying it generated a checksum and then signed that, or something
Jan 17 01:28:30 <GrueMaster> No, it generates a checksum of the file and attaches it in a 64 byte header.
Jan 17 01:29:07 <twb> You know how you can tell the kernel (at compile time) "tack this initrd onto yourself" ?  Can you do that post-compile?
Jan 17 01:29:49 <GrueMaster> It can also generate a key signature that u-boot has built in so that vendors can lock users from rooting their devices (so far we're smart enough to get around that in most cases).
Jan 17 01:30:57 <GrueMaster> Not sure.  I know aboot can do that.  That is how I boot panda w/o SD.  I use abootimg to combine kernel, initrd, and boot args into a single blob that I push down with a usbboot utility on the host system.
Jan 17 01:31:18 <twb> Doesn't "get around it" usually involve exploiting bugs in their bootloader?
Jan 17 01:33:04 <GrueMaster> Not sure.  I don't work that end personally.
Jan 17 01:41:25 <AdamOutler> LetoThe2nd, I have performed step 1.
Jan 17 01:41:37 <AdamOutler> I'm not sure about 2 and 3...  2)make sure the bootloader passes this ID 3) make sure the kernel contains a board support file bound to this id.
Jan 17 01:44:54 <mythos> GrueMaster, it's lzma, thank you very much \o/
Jan 17 01:45:21 <GrueMaster> You're welcome.  I'm off for now.
Jan 17 01:45:34 <mythos> have a nice day
**** ENDING LOGGING AT Tue Jan 17 02:59:57 2012