**** BEGIN LOGGING AT Tue Apr 13 02:59:56 2010 Apr 13 06:45:03 GrueMaster: You can't loop mount cramfs; you have to gunzip + cpio -i it Apr 13 07:50:38 hrw|gone: yet another ARM netbook, http://www.linuxfordevices.com/c/a/News/Science-and-Technology-Co-Ltd-Wabook/ Apr 13 07:54:49 VIA arm, wtf. Apr 13 08:05:05 They build any sort of processor :) Apr 13 08:13:57 ndec, http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-live/current/ has a live image for the C4 beagle that should boot and run without hitting OOM (still not installable and indeed very slow on that HW, but workable and our future base for later releases) Apr 13 08:14:18 ndec, from now on we should have "usable" dailies Apr 13 08:14:25 ogra: thx. Apr 13 08:14:54 ogra: does the installer support installation on the SD card, or does it require USB drive? Apr 13 08:15:06 ogra: btw, I will be in vacation next 3 days... ;-) Apr 13 08:15:09 ndec, for lucid it requires a USB drive Apr 13 08:15:46 ndec, for later we'll do with other setups, given that thu. is the last day to upload anything to lucid i had to make many compromises ... Apr 13 08:16:10 ndec, the lucid image will also re-use the SD as /boot Apr 13 08:16:14 ogra: ok. but if there is a workaround I can do on the image, please let me know. Apr 13 08:16:16 zumbi: this netbook is parody of all netbooks... Apr 13 08:16:38 ndec, but we're not there yet, the partitioner seems to hit a kernel bug atm Apr 13 08:17:06 suihkulokki: Hey, I couldn't get qemu-maemo/-meego to work with SD card images; it hangs agter Uncompressing linux... done., can't get any kernel output; does it work for you with some combination of SD image + qemu-system-arm invocation? NB: I'm building without the gles forward libs Apr 13 08:17:22 (This is with -M beagle) Apr 13 08:17:23 ndec, Bug 561426 ... probably your kernel folks have an idea and could help amitk Apr 13 08:17:25 Launchpad bug 561426 in linux-ti-omap (Ubuntu Lucid) (and 1 other project) "partman dies when trying to detect disks due to kernel error (affects: 1) (heat: 8)" [High,New] https://launchpad.net/bugs/561426 Apr 13 08:17:29 10" 1024x600 are fine. but 300MHz arm926 cpu and 2h battery life??? my 8 years zaurus is faster and have better battery life Apr 13 08:18:14 lool: http://wiki.meego.com/ARM/Meego_on_Qemu Apr 13 08:24:50 ndec, oh, and i forgot, http://ports.ubuntu.com/ubuntu-ports/dists/lucid/main/installer-armel/current/images/omap/netboot/omap/ has a working netinstall image too now (as well as http://cdimage.ubuntu.com/ubuntu-server/ports/daily/current/) ... they all suffer the same partitioner issue though, but these two are more lightweight Apr 13 08:26:03 ogra: its interesting though that I cannot trigger this on a USB stick (running fdisk/formatting, etc.) from rootfs Apr 13 08:26:41 amitk, well, i would happily give more debug info if i could switch consoles in d-i Apr 13 08:27:07 ogra: serial console? Apr 13 08:27:20 doesnt give me tty4 :) Apr 13 08:27:37 ogra: the exact set of commands that trigger this would be useful Apr 13 08:27:37 i tried already, i cant get any output from the partman script at all Apr 13 08:28:06 live images are not very useful for development, I want to recreate this on an 'installed' system Apr 13 08:28:11 i'll try to attach the script that runs last to the bug ... but i cant identify which command causes the actual failure Apr 13 08:30:14 ogra: tty4? Apr 13 08:30:36 amitk, thats where d-i runs and shows output ... Apr 13 08:31:16 amitk, its not that important ... i tried to redirect script output to a file but didnt get any info from that either, it dies very early Apr 13 08:31:24 suihkulokki: thanks Apr 13 08:31:27 like ... before it execs commands it seems Apr 13 08:33:04 suihkulokki: Do you run your tree with beagle images from time to time? Apr 13 08:35:34 randomly Apr 13 08:43:13 hello Apr 13 08:43:17 what is the current status? Apr 13 08:43:26 great! Apr 13 08:43:49 where can i get image i can try? Apr 13 08:44:38 you haven't specified what status you're inquiring about nor which image you're interested in :) Apr 13 08:45:06 i want to try ubuntu on my beagleboard Apr 13 08:45:16 does keyboard, mouse, network work? Apr 13 08:45:32 neure: yes, on the live image. Apr 13 08:45:40 where do i get one? Apr 13 08:45:41 neure, what beagle (revision) Apr 13 08:45:47 rev c i think Apr 13 08:45:50 this is 256MB Apr 13 08:46:03 revC should work (but very slow) with the live image Apr 13 08:46:12 very slow? Apr 13 08:46:14 why? Apr 13 08:46:14 http://cdimage.ubuntu.com/cdimage/ubuntu-netbook/ports/daily-live/ Apr 13 08:46:19 http://cdimage.ubuntu.com/ubuntu-netbook/ports/daily-live/current/ Apr 13 08:46:33 whicih one? Apr 13 08:46:37 i mean which image Apr 13 08:46:43 omap? Apr 13 08:46:48 yes Apr 13 08:47:11 neure, because our live images are rather built for 512M ... the 256M HW makes everything slow and the image has to use several hacks to even make it work at all Apr 13 08:47:52 can i tell it not to boot to X? Apr 13 08:48:14 something like server install?) Apr 13 08:48:39 http://cdimage.ubuntu.com/ubuntu-server/ports/daily/current/ Apr 13 08:48:44 try that image then Apr 13 08:49:01 note though that we are working on a bug with the partitioner atm, you wont get very far Apr 13 08:55:17 suihkulokki: The meego n900 drop boots fine for me with your cmdline, but e.g. an Angstrom beagleboard SD card image doesn't; I'm using: Apr 13 08:55:20 qemu-maemo-system-arm -M beagle -m 256 -sd Angstrom-Beagleboard-demo-image-glibc-ipk-2010.3-beagleboard.sd-image-2GiB.img -serial stdio -clock unix Apr 13 08:55:47 suihkulokki: Do you use any extra args which could help? Apr 13 08:56:53 ogra i need something i can use to build my app Apr 13 08:58:44 lool: try building the kernel with early printk on to see where it ends Apr 13 08:58:47 amitk, so it does seem this time testing i dont even get as far as partman ... seems it rather fails in disk-detect Apr 13 08:58:56 which gets me even less logs :/ Apr 13 08:59:31 suihkulokki: Ok thanks Apr 13 09:00:54 ogra: seems like disk-detect is d-i-only? Apr 13 09:01:08 no, ubiquity uses it as well Apr 13 09:01:29 (I meant it isn't part of the standard install) Apr 13 09:02:01 it is Apr 13 09:02:12 oh, you mean on the installed system ? no Apr 13 09:02:33 its a script that checks the disk availability Apr 13 09:02:41 * ogra treis to log to a file Apr 13 09:02:57 script with no debug output, I presume? Apr 13 09:04:35 hi Apr 13 09:06:24 amitk, debug output added to the bug now (running disk-detect with set -x) Apr 13 09:06:35 amitk, smells like it dies even before Apr 13 09:07:15 ogra: question is, will all debugging have to be done in a live environment only? Or does the installed system have similar commands... Apr 13 09:07:41 amitk, i doubt it Apr 13 09:07:56 lets iedntify what triggers it first Apr 13 09:10:59 ogra: can the disk-detect script be used outside d-i? Apr 13 09:11:20 i dont think so, it sources several d-i componentws Apr 13 09:11:42 http://launchpadlibrarian.net/44009972/syslog somewhat indicates that the issue lies even before disk-detect Apr 13 09:13:25 Apr 13 08:47:27 kernel: [ 318.895812] Unhandled fault: external abort on non-linefetch (0x1018) at 0x40200000 Apr 13 09:13:30 right Apr 13 09:13:39 thats right *before* disk-detect runs Apr 13 09:13:53 i wonder if it has something to do with the check for missing firmware Apr 13 09:14:05 i'll set -x that one (if i can find it) Apr 13 09:14:05 Apr 13 08:47:23 main-menu[238]: INFO: Menu item 'disk-detect' selected Apr 13 09:14:06 Apr 13 08:47:24 kernel: [ 316.195953] sd 0:0:0:0: Attached scsi generic sg0 type 0 Apr 13 09:14:09 Apr 13 08:47:24 net/hw-detect.hotplug: Detected hotpluggable network interface eth0 Apr 13 09:14:12 Apr 13 08:47:24 net/hw-detect.hotplug: Detected hotpluggable network interface lo Apr 13 09:14:15 Apr 13 08:47:26 check-missing-firmware: no missing firmware in /tmp/missing-firmware Apr 13 09:14:19 Apr 13 08:47:27 kernel: [ 318.895812] Unhandled fault: external abort on non-linefetch (0x1018) at 0x40200000 Apr 13 09:14:22 yeah Apr 13 09:14:22 ogra: it says you selected disk-detect there... Apr 13 09:14:35 d-i selects it Apr 13 09:14:57 or rather the main-menu process (which is the core of d-i) Apr 13 09:16:27 Apr 13 09:15:46 main-menu[240]: (process:5940): + log no missing firmware in /tmp/missing-firmware Apr 13 09:16:27 Apr 13 09:15:46 main-menu[240]: (process:5940): + logger -t check-missing-firmware no missing firmware in /tmp/missing-firmware Apr 13 09:16:27 Apr 13 09:15:46 main-menu[240]: (process:5940): + return 1 Apr 13 09:16:27 Apr 13 09:15:46 main-menu[240]: (process:5940): Bus error Apr 13 09:16:27 Apr 13 09:15:46 main-menu[240]: (process:5940): Bus error Apr 13 09:16:29 Apr 13 09:15:46 main-menu[240]: (process:5940): Bus error Apr 13 09:16:45 so thats with set -x in check-missing-firmware Apr 13 09:17:03 the non-linefetch error still shows above that Apr 13 09:17:45 * ogra puts set -x into net/hw-detect.hotplug Apr 13 09:20:20 amitk, https://bugs.edge.launchpad.net/ubuntu/+source/linux-ti-omap/+bug/561426/comments/3 Apr 13 09:20:25 Launchpad bug 561426 in linux-ti-omap (Ubuntu Lucid) (and 1 other project) "partman dies when trying to detect disks due to kernel error (affects: 1) (heat: 8)" [High,In progress] Apr 13 09:25:15 hmm, adding set -x to net/hw-detect.hotplug doesnt get me any more output Apr 13 09:28:07 ogra: no strace possible? Apr 13 09:28:26 ~ # Apr 13 09:28:26 ~ # parted_devices Apr 13 09:28:28 [ 1034.439666] Unhandled fault: external abort on non-linefetch (0x1018) at 0x40200000 Apr 13 09:28:30 Bus error Apr 13 09:28:32 ~ # Apr 13 09:28:34 got it ! Apr 13 09:31:10 ogra, I have the same on my ti-omap variant Apr 13 09:31:55 hence I doubt it's a ubuntu regression Apr 13 09:39:18 nosse1, no matter if its a regression or not, it breaks the ubuntu (and likely also debian) installer Apr 13 10:00:00 lool, you wanted to change the omap naming scheme, do you plan to do that for lucid ? (i'm just patching libdebian-installer to add proper subarch detection) Apr 13 10:02:21 anyone know if efence / electric fence is thread safe ? Apr 13 10:02:37 or rather - i think its not - does anyone know of any thread safe memory bounds tools to run on the arm ? Apr 13 10:05:13 ogra: You mean omap3 versus omap? After yours and amitks explanation, I'm happy to keep it if it means any OMAP Apr 13 10:05:20 At least 3 and 4 that is Apr 13 10:05:25 ok Apr 13 10:05:43 amitk: Hmm open(/dev/mem) :-) Apr 13 10:05:46 just wanted to know if i need to take that into account for d-i Apr 13 10:06:05 lool, yeah, it tries to run dmidecode :P Apr 13 10:06:24 or parts of it at least which are copied into parted_devices Apr 13 10:08:21 * ogra takes a break Apr 13 10:11:17 lool: yeah, /dev/mem is so irresistible :) Apr 13 12:14:54 ogra: do you know already if the parted fixes fix the problem? Apr 13 12:15:10 amitk, still building with the fix Apr 13 12:15:14 i'll test asap Apr 13 12:15:24 but the parted build takes a while Apr 13 12:16:32 ok Apr 13 12:27:10 ogra: is the mobile IRC meeting cancelled today? Apr 13 12:27:28 dmart, its in 30min Apr 13 12:27:45 dmart, do you share calendars with asac ? he asked the same some mins ago :) Apr 13 12:28:08 are there any popcon figures available for arm? Apr 13 12:28:23 Oh, ... ah my calendar doesn't understand that meeting's in UTC don't move with daylight savings time Apr 13 12:28:41 looked @ http://popcon.ubuntu.com/ and the latest seem to be dapper Apr 13 12:28:43 ian_brasil, http://popcon.ubuntu.com/ Apr 13 12:29:19 ian_brasil, i simply doubt anyone enabled popcon on the few armel installs we have Apr 13 12:30:25 ogra, this could be quite useful..could this be encouraged in the installer? Apr 13 12:37:46 amitk, fixed :) Apr 13 12:38:03 Creating ext4 file system for / in partition #1 of SCSI1 (0,0,0) (sda) ... Apr 13 12:38:06 :D Apr 13 12:39:31 wohoo ... i'm in base-install Apr 13 12:42:29 nice Apr 13 12:43:54 ogra: nice! Apr 13 12:44:13 and cjwatson already uploaded ... Apr 13 13:42:35 I have an ever-eating netbook-launcher running on my target. Now it has consumed 525M of ram, on a 256M system.... Anyone else seen this behaviour? Apr 13 13:44:29 netbook-launcher ? or netbook-launcher-efl ? Apr 13 13:44:39 (dont run the former on armel) Apr 13 13:46:27 ogra, It's the efl version. It is the one started automatically Apr 13 13:46:55 nosse1: I've not seen it take up that much, no. What version, and what are you running it on? Apr 13 13:47:09 Now its 534M, so I suspect a leak. Apr 13 13:47:35 ii netbook-launcher-efl 0.2.6-0ubuntu1 A lite version of netbook launcher written in EFL Apr 13 13:47:41 * ogra has only seen netbook-launcher misbehave Apr 13 13:47:48 but never efl Apr 13 13:48:22 * nosse1 cant tell if it should behave this way or if its a leak Apr 13 13:48:38 it shouldnt Apr 13 13:49:04 Hmm. Its consuming 50% CPU constantly as well Apr 13 13:49:35 nosse1: how long has it been running? Apr 13 13:49:51 demo 894 58.8 67.8 538024 168144 ? RLl 13:36 77:41 /usr/bin/netbook-launcher-efl Apr 13 13:51:41 I want to kill it, unless theres something you would like to know Apr 13 14:06:53 nosse1: What version is it? I have had it running overnight on my systems and it is only using 56104k memory. Apr 13 14:11:39 It's the 0.2.6-0ubuntu1 version Apr 13 14:11:49 I updated this morning Apr 13 14:17:26 ogra: uploading a 'final' kernel to people.canonical.com/~amitk/ti . Please test and let me know if you see anything untoward. I'll test it overnight here and upload first thing tomorrow morning. Apr 13 14:17:48 amitk, ok, will do, thanks a lot for the awesome work the last days Apr 13 14:17:53 It has OTG support plus some config fixes, everything else will have to go as SRUs Apr 13 14:18:11 right Apr 13 14:18:12 ogra: thank you :) Apr 13 14:22:35 amitk, oh, we might need to have a uboot-mkimage dependency in linux-image-2.6.33-500-omap Apr 13 14:22:38 apw, ^^^ Apr 13 14:22:42 can that be added Apr 13 14:23:05 ogra, i'll let amitk handlle that Apr 13 14:23:10 ok Apr 13 14:23:23 as i don't have an uptodate omap tree at the mo Apr 13 14:23:27 morning Apr 13 14:24:07 * Martyn is dissapointed .. TI has said there are no plans to bring an omap44xx beagleboard into play. *sigh* Apr 13 14:24:19 and nVidia are completely out of tegra250's Apr 13 14:29:56 lool: Thanks for reminding me about initrd being cramfs. It has been a long time since I debugged an initrd issue. Apr 13 14:38:30 GrueMaster: it does seem to grow very gradually, I've been up for just about 30 min on this system and n-b-l-e is at 24164 RSS, if you are at 54k from overnight, that would indicate it seems to have grown over time, but it's hard to imagine it getting to 525M as nosse1 is seeing very soon at that rate Apr 13 14:38:37 something has to be setting it off Apr 13 14:38:59 nosse1: are you doing a lot from the console? anything you can think of that may have cause it to suddenly grow rapidly? Apr 13 14:39:18 No, I just left the target while going for lunch Apr 13 14:43:22 amitk, is your OTG fix supposed to fix that: http://paste.ubuntu.com/413681/ (i get it on an installed system) Apr 13 15:02:59 Martyn, there will be no OMAP4 'BeagleBoard' but there will certainly be a low cost development board from us... just with a different name Apr 13 15:03:03 cheer up :) Apr 13 15:03:24 cpearson: Yeah .. but I will miss the compact cuteness of the beagleboard Apr 13 15:03:48 um... let's say that it'll be about the same form factor and backwards compatible Apr 13 15:03:56 * Martyn perks Apr 13 15:11:42 Is is possible to install hooks on update-initramfs in order to generate uInitrd images? Apr 13 15:12:03 How do you do it (i.e. workflow) when u-boot is involved Apr 13 15:13:12 I'm not sure, but I think the kernel post install script does that for the dove kernels. Apr 13 15:13:28 It may be easier to create a wrapper script. Apr 13 15:14:30 nosse1, thast done by flash-kernel Apr 13 15:14:46 nosse1, i'm working on getting that pieces sorted before thu. Apr 13 15:15:42 cpearson: first day at the conference go well? Apr 13 15:20:23 yes, just VERY long. The Ubuntu on ARM BoF didn't wrap up until 10pm Apr 13 15:20:54 wow Apr 13 15:21:00 sounds like lots of interest Apr 13 15:21:21 Dave's talk was good, and our the Ubuntu on OMAP4 demo sparked a lot of conversation Apr 13 15:21:51 sweet Apr 13 15:27:08 * cpearson packs up and heads over to the conference Apr 13 15:42:00 ogra: I fixed the dove netboot issue (well figured out what was wrong). Need a fix for debian-installer. See my post on bug 541399 Apr 13 15:42:06 Launchpad bug 541399 in debian-installer (Ubuntu) "netboot image fails to boot. (affects: 1) (heat: 8)" [Medium,In progress] https://launchpad.net/bugs/541399 Apr 13 15:42:13 * ogra reloads Apr 13 15:43:12 GrueMaster, that proves my point ;) Apr 13 15:43:15 ogra@osiris:~/Devel/branches/debian-installer/ubuntu$ grep INITRD_FS ./build/config/armel/dove/cdrom.cfg Apr 13 15:43:15 INITRD_FS = initramfs Apr 13 15:43:33 ogra@osiris:~/Devel/branches/debian-installer/ubuntu$ grep INITRD_FS ./build/config/armel/dove/netboot.cfg Apr 13 15:43:33 ogra@osiris:~/Devel/branches/debian-installer/ubuntu$ Apr 13 15:43:48 which conference? ELC? Apr 13 15:44:46 So, it should be a very simple fix, right? Apr 13 15:46:15 GrueMaster, committed Apr 13 15:46:23 cool, thanks. Apr 13 19:25:25 ogra: this oops occurs _after_ the install? Apr 14 01:37:34 If I use u-boot to configure my beagleboard to use SPI, will the stock Ubuntu kernel pick up on the device? where will it show up in /sys/? Apr 14 02:06:20 andruk, those u-boot spi setting in 2010-03 are still in their infancy.. I'm not even sure if Angstrom's 2.6.32-psp kernel does that either... Apr 14 02:17:18 rcn-ee: oh, okay. so, to adjust the SPI settings in the kernel, I just need to get the source from https://code.launchpad.net/~beagleboard-kernel/+junk/2.6-stable ? will it be blindingly obvious how to compile a new kernel? Apr 14 02:17:53 andruk, you have a zippy1 right? Apr 14 02:18:08 rcn-ee: nope, just a vanilla bb Apr 14 02:19:05 ah yes, i remember now, you were just wanting to get "spidev" to work... this patch works: http://elinux.org/BeagleBoard/SPIPatch-2.6.32 Apr 14 02:20:28 all you need to do is "bzr branch" the 2.6-stable... add that patch anywhere in the patch dir, then add a "patch -p1 -s < ${DIR..etc" to patch.sh then run build_kernel.sh... ALl you need is CC defined in system.sh Apr 14 02:23:39 something like this: http://pastebin.com/hZd2XLu6 (had it saved in my temp since our last conversation) Apr 14 02:25:14 andruk, if it actually works for you and your spi devices work, I'll rework the patch and actually commit it... Apr 14 02:55:03 rcn-ee: so, is that a patch for patch.sh? Apr 14 02:55:24 rcn-ee: the pastebin link, I mean Apr 14 02:55:47 yeah, it's a patch for patch.sh and it adds the SPIPatch-2.6.32.diff file... **** ENDING LOGGING AT Wed Apr 14 02:59:56 2010