**** BEGIN LOGGING AT Sat Nov 05 02:59:57 2011 Nov 05 05:41:54 Oneric on Pandaboard. Distorted display at install and afterwards. Are there known issues with HDMI to DVI? HDMI to DVDI-D Cable is known good. Nov 05 07:16:37 Hello. I've compiled a kernel for my BeagleBoard-xM rev C using Linus' git tree and when I boot it, it hangs on mmcblk0: p1 p2 Nov 05 07:16:43 how would I debug this problem? Nov 05 07:30:17 Well, that line is it finding two partitions on the first MMC (SD card) block device. Nov 05 07:30:31 Not sure why that's hanging, tho Nov 05 07:36:49 it is finding those partitions Nov 05 07:38:30 here's full boot log: https://gist.github.com/1341243 Nov 05 07:41:13 at least I think it's finding 2 partitions Nov 05 13:55:00 hello i am trying to run the ubuntu oneiric preinstalled image on a beagleboard xm and cannot get the board to boot Nov 05 13:55:32 boot has errors like - Wrong Image Format for bootm command, ERROR: can't get kernel image Nov 05 13:57:07 are there any folks here successfully using the beagleboard xm? Nov 05 14:08:25 Hello. May I ask a question regarding an armv5 chip in this channel? Nov 05 16:43:39 lilstevie: whats this I hear about new flashkit and stuff :D :D :D Nov 05 22:25:20 Any activity here now? Anyone running i386 Natty? Would be nice if anyone could confirm #886690. Nov 05 22:25:28 bug 886690 Nov 05 22:25:30 Launchpad bug 886690 in qemu-linaro "qemu segfaults in armel chroot on i386 host" [Undecided,New] https://launchpad.net/bugs/886690 Nov 06 00:25:29 sveinse: did you try with latest qemu-linaro from https://launchpad.net/~linaro-maintainers/+archive/tools? Nov 06 00:26:16 rsalveti: Yes. I just commented it on the bug. It works using the newer backported qemu Nov 06 00:26:43 sveinse: great, so at least you have a working version Nov 06 00:27:31 Yes. This was a serious blocker for us, since the build server is running i386 natty... Nov 06 00:29:48 rsalveti: Did you see my question the other day? For some reason rootstock adds the fuse group, but without any explanation to why. Do you recall if it was a particular reason? Nov 06 00:31:12 sveinse: hm, don't remember the reason why it was added Nov 06 00:31:17 sveinse: any issues with it? Nov 06 00:34:35 rsalveti: Nope. Just ooi, so don't bother. In fact, I'm not using rootstock any more. I'm using qemu-debootstrap instead. But when changing over from rootstock to deboostrap, I had to ensure my script had similar setup to rootstock, and that's when I discovered the fuse setting. Nov 06 00:35:04 sveinse: I'd recommend you to use live-build if you can Nov 06 00:35:16 I know there's support for qemu at the latest live-build version Nov 06 00:35:25 and you can hook scripts and etc Nov 06 00:40:27 rsalveti: I'll check it out. However we are satisfied with way debootstrap works. I.e. run debootstrap and then apt-get install everything under chroot qemu emulation. The only risk is the qemu armel emulation during dpkg install when building the rootfs on an intel host. But it has proved to work successfully on a larger set of prototypes. Nov 06 00:40:47 sveinse: yeah, great Nov 06 00:41:04 and it seems for precise mono, tomboy and banshee was removed from desktop Nov 06 00:41:17 then no mono issue when cross debootstraping the desktop anymore :-) Nov 06 00:42:30 rsalveti: I'm doing it in two steps: I'm debootstrapping to minimal, and then include the rest under chroot. And our app is not using desktop, its Qt QWS, so no X11. Nov 06 00:42:52 cool, then you can build it quickly Nov 06 00:43:10 But one of the developers are talking about adding mono. Should I expect issues with installing it under chroot? Nov 06 00:46:43 sveinse: yup, there are quite a few issues when installing mono packages Nov 06 00:47:31 sveinse: bug 530000 Nov 06 00:47:32 Launchpad bug 530000 in qemu-kvm "mono assembly installation under qemu-arm-static hangs" [Medium,Won't fix] https://launchpad.net/bugs/530000 Nov 06 00:47:42 rsalveti: good to know. Does this also include native installation of mono on armel? Nov 06 00:47:58 nops, just under qemu iirc Nov 06 00:50:26 rsalveti: One thing I've been considering, is to only run dpkg unpack in image creation, and wait to do configure until the customer boots the product. I suppose this is the same as running the second stage natively Nov 06 00:50:43 sveinse: ye Nov 06 00:50:44 yes Nov 06 00:50:53 depending on the amount of packages you have, it's quite a sane solution Nov 06 00:51:11 and you don't rely on qemu anymore Nov 06 00:51:29 Except that "See I bought this product. Now I just have to wait 20 minutes for it to start for the first time" Nov 06 00:54:05 well, it all depends on the amount of packages you have Nov 06 00:54:19 and for sure some packages could be changed to avoid doing stuff at the post inst Nov 06 00:55:10 rsalveti: Would it be possible to debootstrap the first stage on intel, and then move the root into a chroot on a armel native and run the second stage using a chroot there? Nov 06 00:55:35 E.g. does that resolve the mono issues, right? Nov 06 00:55:45 sveinse: yes Nov 06 00:59:54 rsalveti: BTW I don't know if there is much activity on rootstock, but during an installation of a package I discovered that it can be smart to dpkg-divert /sbin/start since some packages tries to start the services from their postinst's. Nov 06 01:00:58 yeah, this would help avoiding starting services while running with qemu Nov 06 01:01:10 I need to spend some time on it, to at least fix some of the bugs Nov 06 01:01:20 but we're kind of moving to live-build now Nov 06 01:01:55 yup, just a suggestion Nov 06 01:02:19 but yeah, useful, thanks anyway :-) **** ENDING LOGGING AT Sun Nov 06 02:59:57 2011