| error404notfound | Hi! | 06:55 |
|---|---|---|
| error404notfound | how easy it is to install ubuntu on an arm machine with no usb port, no cd rom, no ethernet, just wifi and 3g? | 07:11 |
| persia | Not very. You'd need some way to download enough to boot, and then be able to boot off that. | 07:14 |
| error404notfound | persia: hmmm, actually I am really confused that if it doesn't have a usb port, a cdrom, or ethernet, how can even copy anything from network to it. I don't know from where should I even start | 07:15 |
| persia | I'd recommend starting by looking for someone else who has gotten any linux running on your device. | 07:16 |
| persia | You can probably adjust those instructions as required to have the install be Ubuntu. | 07:17 |
| error404notfound | I am following your first recommendation by joining this channel... | 07:18 |
| persia | In that case, you might want to mention the specific device name :) | 07:20 |
| error404notfound | hmmm, wanna know the truth? I still haven't got one, its stuck for clearance on airport and I am waiting for it... may take couple of days.. :P | 07:21 |
| Stskeeps | sometimes i want to shoot whoever chose right click as paste. | 07:46 |
| persia | Oughtn't be right-click. Ought be middle-click. | 07:47 |
| Stskeeps | yeah | 07:47 |
| tomsh | is it normal when using build-arm-rootfs | 08:02 |
| tomsh | * Setting kernel variables (/etc/sysctl.d/10-network-security.conf)... error: "net.ipv4.tcp_syncookies" is an unknown key | 08:02 |
| tomsh | * Setting kernel variables (/etc/sysctl.d/10-process-security.conf)... error: "vm.mmap_min_addr" is an unknown key | 08:02 |
| tomsh | ? | 08:02 |
| persia | tomsh: It depends on your kernel configuration, really. It's been reported before. | 08:33 |
| lool | persia: Do you know whether we have bug for these? It shouldn't happen with the versatile kernel as it's a distro kernel | 08:56 |
| lool | And the imx51 one is as well now | 08:57 |
| persia | We don't have a bug for it: it's never been reported against a distro kernel (that I've heard). | 08:59 |
| lool | NCommander: Do you think you could report these issues as well as the babbage ones? | 14:00 |
| lool | NCommander: Tag them arm | 14:01 |
| ogra | lool, they are on my list of bugs for amitk_ | 14:04 |
| ogra | net.ipv4.tcp_syncookies definately shows on babbage as well | 14:04 |
| lool | ogra: Which list? | 14:04 |
| ogra | my personal one i'nm assembling while testing the babbage kernel here | 14:04 |
| ogra | klogd not starting is another one ... | 14:05 |
| ogra | though currently i'm after finding out whats wrong with aufs/apparmor that makes all ttys die and X not start up | 14:06 |
| ogra | thats more important than cosmetical boormessages | 14:06 |
| ogra | *boot | 14:06 |
| ogra | lool, i wont manage to file these bugs right now, can you mention in the release meeting that there are imx51 bugs to come we will need to fix before beta ? | 14:10 |
| lool | ogra: No | 14:11 |
| lool | I'll give a general status update | 14:11 |
| ogra | ok | 14:11 |
| lool | I remember quite well the day were I was shot in place for mentionning bugs which I hadn't filed yet | 14:11 |
| ogra | well, they are all under investigation only atm ... | 14:12 |
| ogra | the imx51 kernel isnt 100% merged and i assume a lot will be solved once that happened | 14:13 |
| lool | NCommander: Please Cc: slangasek on the RedBoot build issues with native toolchain | 14:20 |
| obra | Hi. I'm seeing some weird behaviour with ubuntu (and debian) arm ports on qemu. It _appears_ that some of the time the SCSI device just stops responding. It appears to be fairly random. When it happens, spurious "a"s get printed to my terminal. | 14:22 |
| obra | I'm seeing the behaviour across several versions of qemu on OSX and Linux. | 14:23 |
| obra | Is there an obvious FAQ I'm missing? | 14:23 |
| Stskeeps | versatile kernel or? | 14:40 |
| obra | versatile kernel. | 14:41 |
| obra | /usr/local/qemu/bin/qemu-system-arm -M versatilepb -m 256 -kernel vmlinuz-2.6.28-8-versatile -hda armel.img -append "mem=256M root=/dev/sda1" | 14:41 |
| obra | (Starting from the initrd images and doing an install) | 14:42 |
| obra | I get the same behaviour using http://people.debian.org/~aurel32/qemu/armel/ | 14:42 |
| Martyn | Good morning. | 16:25 |
| Martyn | Ogra, look, amit .. any of you online? | 16:25 |
| Martyn | lool rather | 16:26 |
| lool | Martyn: I'm around, but I'm out for an hour or so | 16:27 |
| lool | I'll be back here later | 16:27 |
| Martyn | Allright. I'm trying to do a first bringup of ubuntu on the board, and wanted to verify some settings | 16:32 |
| amitk_ | Martyn: is this something quick? | 16:33 |
| * amitk_ is about to wrap up for the day | 16:33 | |
| === amitk_ is now known as amitk | ||
| Martyn | amitk : Relatively. The image that ogra put up (the SD card image) seems to have reboot on it, and the first partition is a few sectors from the beginning of the card... | 16:39 |
| ogra | right | 16:39 |
| ogra | i think i left 20M ... though 15M shoudl suffice | 16:40 |
| Martyn | When I put it in the board and power up, I do see the keyboard flash, but no console output or display output.. is that normal for this image? | 16:40 |
| Martyn | yep, about 20M | 16:40 |
| ogra | you should get console output on VGA | 16:40 |
| Martyn | dreck, I don't. | 16:40 |
| ogra | after redboot decompressed the kernel | 16:40 |
| ogra | probably an issue with your DIP switch settings | 16:40 |
| Martyn | would you mind going through a quick board comparison? I want to make sure the DIP switches on mine match yours. | 16:40 |
| Martyn | the two that are together at the top, read from SATA->pushbutton are: | 16:41 |
| Martyn | 0-0-0-0-0-0-0-1 1-0-0-0-0-0-0-0 | 16:41 |
| ogra | the middle ones are on | 16:41 |
| ogra | right | 16:42 |
| Martyn | the one next to the SD card is (read from VGA to serial breakout) | 16:42 |
| Martyn | sorry, read from serial->vga direction | 16:42 |
| ogra | the second should be on | 16:42 |
| Martyn | 0-1-0-0-0-0-0-0 | 16:42 |
| Martyn | Okay, those match | 16:42 |
| ogra | looks ok | 16:42 |
| Martyn | So, in theory it should have just worked | 16:42 |
| ogra | yes | 16:43 |
| Martyn | I'll try a different SD card then. Darnit! | 16:43 |
| ogra | note the kernel doesnt support DVI | 16:43 |
| Martyn | Wait, I thought the kernel didn't support VGA... | 16:43 |
| ogra | you need to use VGA | 16:43 |
| ogra | no | 16:43 |
| Martyn | okay, switching connectors. | 16:43 |
| Martyn | and switching to a real SanDisk card | 16:43 |
| ogra | yup, sandisk ++ | 16:44 |
| Martyn | okay, I'm going to re dd it to a sandisk 2Gb card. | 16:44 |
| Martyn | allright, once the DD is finished, I'll try again. | 16:46 |
| * ogra finally got the initramfs booting for amitk's latest kernel | 16:46 | |
| Martyn | congrats! | 16:47 |
| ogra | well, i dont see the bugs fixed yet ... | 16:47 |
| * ogra waits for casper to find the squashfs | 16:47 | |
| ogra | amitk, tty's still die :( | 16:48 |
| ogra | amitk, oh, but you added fusectl .... | 16:48 |
| ogra | at least it doesnt complain anymore | 16:48 |
| ogra | oh, and X now starts up but i get a "your session lasted less than 10 secs."+ message now | 16:49 |
| amitk | hmm | 16:49 |
| Martyn | ogra : At least X starts up. | 16:49 |
| ogra | hmm, that looks a bit like the former error though, i see attempts fopr autologin on tty1 | 16:50 |
| amitk | ogra: this is live image? | 16:50 |
| ogra | yep | 16:50 |
| amitk | with union=aufs? | 16:50 |
| ogra | without union= | 16:50 |
| ogra | which defaults to aufs | 16:50 |
| ogra | i see it loading aufs | 16:50 |
| amitk | ok | 16:50 |
| ogra | switching to console i see multiple motd messages | 16:50 |
| ogra | something is wrong with the login, but its hard to debug since i konw if i boot into single mode i will see everything with the user is right | 16:51 |
| * ogra will boot into single and start executing the initscripts piece by piece ... time consuming, but probably i see where i hit the issue | 16:52 | |
| ogra | oh | 16:53 |
| * Martyn chuckles, the dd is still going :) | 16:53 | |
| Martyn | this isn't my first dance with new hardware, but I'm frankly amazed at how fast you guys at canonical are doing the bring-up of this one | 16:53 |
| ogra | well, we know our tools :) | 16:54 |
| ogra | so its just a matter of adapting them | 16:54 |
| Martyn | HUZZAH | 16:54 |
| Martyn | we're up. | 16:54 |
| ogra | the main issue was to get initramfs working with redboot | 16:54 |
| ogra | everything beyond that is just small adjustments | 16:55 |
| ogra | congrats ! | 16:55 |
| Martyn | gets as far as sd | 16:55 |
| Martyn | then stops | 16:55 |
| ogra | give it time :) | 16:55 |
| Martyn | timeout issue? | 16:55 |
| ogra | it looks for the squashfs that takes some seconds ... | 16:55 |
| ogra | amitk, klogd still dies | 16:56 |
| Martyn | there it goes.. | 16:56 |
| Martyn | I should probably disable SATA | 16:56 |
| Martyn | okay, I'm seing the tty messages | 16:56 |
| Martyn | nice X | 16:56 |
| Martyn | and there we go. Thanks ogra :) | 16:56 |
| ogra | enjoy | 16:57 |
| ogra | note that its not installable | 16:57 |
| Martyn | This is a much faster bring-up than I usually have to do (kernel, then busybox, then ...) | 16:57 |
| Martyn | That's fine. This is enough for us to do some very, very important tests for the work we're doing. | 16:57 |
| ogra | but you could for example use debootstrap or the rootfs script from the channel topic and install a system to a usb disk | 16:58 |
| Martyn | *nod* | 16:58 |
| Martyn | Or sata? | 16:58 |
| ogra | then just change the cmdline in redboot to not use casper | 16:58 |
| Martyn | Or is boot-to-sata not quite yet ready for prime time? | 16:58 |
| ogra | and add root=UUID=<uuid of your disk partition> | 16:58 |
| ogra | i havent tried sata at all yet | 16:58 |
| amitk | usb disk should work well | 16:58 |
| ogra | and i think sata is just a fake ... | 16:59 |
| ogra | in the backend its attached to usb | 16:59 |
| Martyn | Oh, you think there's a bridge chip? | 16:59 |
| ogra | yes | 16:59 |
| Martyn | The mx51's spec shows a SATA controller though. | 16:59 |
| Martyn | well, doesn't matter :) | 16:59 |
| ogra | well, feel free to try it :) | 16:59 |
| ogra | (if amitk has the sata driver enabled) | 17:00 |
| Martyn | I'm pleased as punch to have a working install. I'll track your builds from this point. | 17:00 |
| Martyn | It looked like it was at least trying to bring sata up during kernel initialization | 17:00 |
| amitk | ogra: no sata drivers yet IIRC | 17:00 |
| Martyn | amitk : Danke. | 17:00 |
| ogra | right, thats what i thought | 17:00 |
| amitk | Martyn: I'll flesh out support for _every_ storage driver as a module in the next upload | 17:01 |
| ogra | if we find one that doesnt make the consoles die :P | 17:02 |
| Martyn | amitk : Ooooo :) | 17:02 |
| Martyn | ogra : Just out of curiosity, where are the authentication token error messages coming from? | 17:03 |
| ogra | which one ? "your password is expired" ? | 17:03 |
| Martyn | yep | 17:04 |
| ogra | run "sudo hwclock --systohc" and reboot ... | 17:04 |
| ogra | (if your system clock is right at least) | 17:04 |
| ogra | that fixes it | 17:04 |
| ogra | useradd somehow sets the expiry time to zero if the system date is at 01 01 1970 | 17:05 |
| ogra | setting the hwclock fixes it | 17:05 |
| Martyn | fixed, and thanks. | 17:07 |
| Martyn | worth adding to your README | 17:07 |
| Martyn | now I want the serial debug dongle ... -sigh- | 17:08 |
| ogra | worth fixing the bug rather :) | 17:08 |
| Martyn | being a unionfs based image, I'm assuming any changes I make to the FS will be lost at reboot? | 17:09 |
| Martyn | very much worth fixing the bug! It's just more of an appnote for that very first release of the image | 17:09 |
| Martyn | thanks :) You guys just made my day. | 17:14 |
| ogra | amitk, hmm, klogd tries to access /dev/console | 17:16 |
| ogra | and then dies | 17:16 |
| Martyn | interesting! | 17:16 |
| Martyn | when the hwclock is set correctly, the tty's also come to life | 17:16 |
| ogra | yep | 17:16 |
| ogra | the autologin works then | 17:17 |
| Martyn | it does indeed | 17:17 |
| ogra | upstart makes them die if the user isnt set up correctly | 17:17 |
| Martyn | no modules.dep file? | 17:17 |
| ogra | nope, next image will have the kernel package inside | 17:17 |
| Martyn | heh, no modules at all :) | 17:17 |
| Martyn | *chuckle* | 17:17 |
| ogra | and have the modules.dep | 17:17 |
| Martyn | should I apt-get it? | 17:17 |
| ogra | apt-get install linux-imx51 | 17:18 |
| Martyn | done | 17:18 |
| ogra | that will get you everything you need | 17:18 |
| ogra | note that the NIC driver is shaky | 17:18 |
| Martyn | Yeah, I know. Works okay though. | 17:18 |
| Martyn | you guys should be very, very proud of this. Most hardware first bring-ups are a hell of a lot more shaky than this. | 17:19 |
| Martyn | This is practically almost useable in comparison to some projects I've worked on. | 17:19 |
| ogra | :) | 17:20 |
| Martyn | Wow, everything is arm V7 clean? | 17:23 |
| Martyn | no v4 compiles? | 17:23 |
| Martyn | what's the process I have to kill to prevent the 100% cpu spinning? | 17:36 |
| Martyn | gnome-keyring? | 17:36 |
| Martyn | ogra, how do I get the live image to boot single user, if I don't have access to the serial console? | 17:46 |
| Stskeeps | Martyn: think it's armv5te targeted, not armv7 | 17:55 |
| Martyn | Stskeeps : Thanks. | 18:05 |
| Martyn | I was wondering how far the system was going to be optimised... | 18:06 |
| Martyn | v5 is good enough | 18:06 |
| Stskeeps | think they're looking at HWCAP of libraries too, which makes sense | 18:06 |
| ogra | Martyn, we might switch to v7 completely with 9.10 | 18:14 |
| ogra | its still being benchmarked what we gain from it and will be discussed at next UDS | 18:14 |
| Stskeeps | or loose in terms of supporting less devices, heh | 18:15 |
| Martyn | ogra : *nod* | 18:15 |
| Martyn | ogra : If I don't have access to the redboot monitor (thanks to having no serial debug board), how can I get the liveimage you've posted to come up single user? | 18:16 |
| ogra | not at all ... | 18:16 |
| * Martyn tried hexediting the redboot, to change the kernel bootparams, but that didn't work :) | 18:17 | |
| ogra | lool is working on something we should see next week that will make it possible to dd changes into the fis partitions | 18:17 |
| Martyn | Is there a way to bring down X safely once the livecd has booted, without triggering a reboot? | 18:17 |
| ogra | you can switch to a console and sudo /etc/init.d/gdm stop i guess | 18:18 |
| Martyn | Tried that. No joy. | 18:18 |
| ogra | well, both options you can set are cmdline options to get into single | 18:19 |
| ogra | and the cmdline is only accessible via serial atm | 18:19 |
| ogra | one would be to add "text" to it, the other would be "single" | 18:19 |
| Martyn | yep | 18:19 |
| ogra | text actually gets you a full system, while single gets you only a recovery console | 18:19 |
| Martyn | Oh well :) I guess I have to be patient and wait until Monday when a serial debug dongle will arrive as if by magic. | 18:20 |
| * ogra was pondering to add a menu to the boot .... | 18:20 | |
| ogra | but i have to much to do to track down the kernel issues on the weekend ... | 18:20 |
| ogra | if i magically find the solution i might look into adding some bootmenu | 18:21 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!