**** BEGIN LOGGING AT Fri Jan 27 02:59:57 2012 Jan 27 09:55:08 rbasak: did you try the new kernel? Jan 27 10:10:37 ppisati: I've just been using whatever the archive gives me, but I haven't done a reinstall in a little while Jan 27 10:11:46 rbasak: it should be the 1405 Jan 27 10:14:58 ppisati: trying now Jan 27 10:22:43 ppisati: yes, it's installing 1405.7 Jan 27 10:36:41 ppisati: 1405 booted with no special parameters needed - thanks! The installer still needs them though. Jan 27 10:38:22 rbasak: why the installer needs them? Jan 27 10:38:56 ppisati: for netinst. I assume the installer isn't using the latest kernel. Jan 27 10:39:25 rbasak: ah, i don't know. You better probe orga/Grue/ndec aabout it Jan 27 11:26:29 ogra_: so you have a beagle xm active by now, right? Jan 27 11:26:44 as my printserver, yes Jan 27 11:26:53 running lucid Jan 27 11:26:53 ogra_: which kernel are you running? and do you change the mac addres? (/etc/network/interfaces) Jan 27 11:27:26 no, its a stock install, i didnt touch the network settings beyond what the installer does Jan 27 11:27:33 ok Jan 27 11:28:14 oh, wait, thats a C4, i lied Jan 27 11:28:26 my XM actually sits in a bag in my office Jan 27 11:28:55 sorry Jan 27 11:30:22 ok Jan 27 11:30:32 because *i think* i met a regression Jan 27 11:30:59 my XM is a rev A though Jan 27 11:31:06 very first model Jan 27 11:31:10 k Jan 27 11:48:28 ndec, there was an Xorg ABI bump (bumped to 11) would it be possible to get a driver rebuild in the PPA since the current driver wont work with the 11 ABI Jan 27 11:55:43 hello - any change someone with ARM/Assembler experience could review https://code.launchpad.net/~james-page/ubuntu/precise/zookeeper/arm-ftbfs-fixes/+merge/89940 for me Jan 27 11:56:01 ? Jan 27 13:22:27 ogra_: you mean for O or P? Jan 27 13:22:30 jamespage, you don't need ARM experience to *remove* ARM code ;-P Jan 27 13:22:37 looks ine Jan 27 13:22:39 fine Jan 27 13:22:51 ndec, for P indeed, we didnt update xorg to a new ABI in O ;) Jan 27 13:22:57 ok Jan 27 13:23:13 we will rebuild, but likely will only support armhf, is that okay? Jan 27 13:23:14 doko: good - I know it works on armel/armhf and for x86 - but wanted someone who might actually understand why! Jan 27 13:23:16 ta Jan 27 13:23:17 ndec, but i think rsalveti is already mailing xavier about it Jan 27 13:23:29 ndec, fine with me Jan 27 14:30:48 ogra_, janimo: please could you install 'deb http://ppa.launchpad.net/ubuntu-toolchain-r/glibc/ubuntu precise main' and run this for a while on your arm machine(s)? Jan 27 14:33:16 doko, what packages are that ? just libc ? Jan 27 14:44:43 doko, installed, anything i should look for after i reboot ? Jan 27 14:58:50 * ogra_ reboots and will blame doko if the system doesnt come up again Jan 27 15:04:02 doko, booted fine, so what am i looking for now ? Jan 27 15:08:21 doko: during installation i see a warning locale: /lib/arm-linux-gnueabi/libc.so.6: version 'GLIBC_2.15' not found (required by locale) Jan 27 15:08:44 ah.. it have not been installed yet Jan 27 15:08:47 likely because the new libc isnt running while it generates the locales Jan 27 15:09:15 thats correct dpkg are Setting up libc6 (2.1.5... now Jan 27 15:09:20 yep Jan 27 15:09:21 2.15 Jan 27 15:10:12 bah, it doesnt fix my graphics issues ! Jan 27 15:10:16 heh Jan 27 15:10:19 :) Jan 27 15:10:24 you should have benchmarked before the upgrade Jan 27 15:10:30 things might be faster Jan 27 15:10:34 pfft Jan 27 15:10:40 whats the actual change ? Jan 27 15:11:06 if i would be after speed i wouldnt use ac100 as my main work machine :) Jan 27 15:12:03 * ogra_ only sees a reversion in the changelog that could be relevant Jan 27 15:22:32 is anyone else's panda netinst falling over today? with the same preseed I was using yesterday, I'm getting "!! ERROR: No root file system" and then "No root file system is defined." at the partition stage. I've tried the externally powered disk I was using weeks ago, and hopefully reverted everything else I've changed. Jan 27 15:22:38 jamespage, GrueMaster: ^^? Jan 27 15:26:34 ogra_: basically looks like we are testing the 2.15 release http://sourceware.org/ml/libc-alpha/2011-12/msg00085.html Jan 27 15:26:55 doh.. i though that statement want out yesterday Jan 27 15:27:11 its a 1 old month news entry Jan 27 15:28:10 ogra: please just run it for a while Jan 27 15:28:16 a week or so Jan 27 15:28:56 xranby, yeah, I need to fix this, dependency issue, but the locales get regenerated later anyway Jan 27 15:29:48 well, there are some neon optimized string operations like {str.mem}{cpy,mov,cmp} Jan 27 15:36:44 doko, well, i'm running on my ac100, NEON wont actually gain me much i think, is it still worth the test for you ? Jan 27 15:38:41 rbasak, not tried since I had to manually hack mine yesterday Jan 27 15:38:50 have been watching hadoop crash :-) Jan 27 15:43:22 ogra_, yes, just to know any issues with 2.15 vs. 2.13 Jan 27 15:47:18 k, no prob then, i'll keep it running ans scream and shout if it breaks :) Jan 27 16:02:15 rbasak: I'll try it here and see if it fails. Jan 27 16:02:21 thanks Jan 27 16:15:25 rbasak: I think it isyour settings. I'm well past the partitioning, and into the base installer section now. Jan 27 16:16:20 Although it looks like a kernel meta was updated before a new kernel was released into the pool. Jan 27 16:19:10 the kernel was released around tuesday or so Jan 27 16:19:32 not sure d-i has been uploaded to use the new kernel yet Jan 27 16:29:16 turns out it was a different (but odd) issue. netinstall was failing to resolve my mirror sewrver when installing the linux-headers packages. very odd. Jan 27 16:35:08 aww Jan 27 16:39:19 GrueMaster: ok, thanks for testing Jan 27 16:40:09 rbasak: I am still experiencing issues. Not what you are seeing, but very odd behavior. Jan 27 16:42:11 GrueMaster: I have been getting various panics, though that may be a power problem Jan 27 16:42:54 I'm seeing that too. Trying to capture a log now. Jan 27 16:45:55 Is David Duffey here? Jan 27 16:46:43 What I'm seeing is it fails at the kernel installation. Log shows issue resolving my mirror. Kicking it moves forward, but fails again with a failure to create /etc/resolv.conf. Then segfault. Jan 27 16:49:42 GrueMaster: You may want to bring it up with stgraber. He's been mucking with fixes from resolvconf fallout. Jan 27 16:51:19 http://paste.ubuntu.com/819007/ is the segfault. Jan 27 16:51:41 GrueMaster: Colin just fixed a similar issue in ubiquity Jan 27 16:52:34 interesting. Odd that this only just started happening yesterday. Jan 27 16:53:39 Yesterday was when the overrides got fixed to get resolvconf in important (and, by extension, in ubuntu-minimal) Jan 27 16:53:43 So, that timing seems right. Jan 27 16:54:02 Before said change your netinsts wouldn't have been installing resolvconf, I imagine. Jan 27 16:55:27 I just tried armel oneiric with exactly the same config, preseed, kernel opts etc, and it has gone past the problem partitioning stage with no issues. Jan 27 16:55:44 I think perhaps the panics I've been seeing over the last few days aren't a power issue as I thought, but a kernel issue Jan 27 16:56:08 There are a few different points where I seem to get different types of panic (from observation, I've not been logging) Jan 27 16:56:23 So something causing what appears to be pretty non-deterministic behaviour/failures Jan 27 16:56:52 Mine falls over exactly the same way on 2 different systems. Both well past the partitioning though. Jan 27 16:57:14 I quite often get a panic straight after the kernel loads, at around the point syslogd (?) starts Jan 27 16:57:15 janimo: Regarding bug 922558, are you actually suggesting LP collect current package info for every buildd and publish that, or did you mean to file that as an RT? Jan 27 16:57:15 Launchpad bug 922558 in launchpad-buildd "Page listing Ubuntu and kernel versions on ARM builders" [Undecided,New] https://launchpad.net/bugs/922558 Jan 27 16:57:25 GrueMaster: I'm running an amd64 netinstall now to confirm it's broken, then will look at a fix (after schroot, lxc, LTSP and ubiquity are fixed) Jan 27 16:57:35 janimo: (Note that indivudual build logs contain exactly the info you're looking for...) Jan 27 16:57:42 ok Jan 27 16:59:06 I don't know why it didn't occur to me to try an oneiric armel install to test my setup before. Now I can take my power hacking out of the equation Jan 27 16:59:17 stgraber: We wouldn't want you to get bored. :P Jan 27 17:00:46 GrueMaster: at least we only got install/build time problems, nobody reported breakage for existing systems :) Jan 27 17:01:30 Oooh, I haven't tried breaking my running installs. Thanks for the suggestion. :P Jan 27 17:03:14 :) Jan 27 17:04:24 I think I was stressing this kernel on a running install fairly hard the other day, and it didn't break Jan 27 17:04:31 not the latest kernel though Jan 27 17:04:35 or maybe it was Jan 27 17:04:48 that's not really any use is it? Jan 27 17:04:50 * rbasak shuts up Jan 27 17:16:48 Just found a USB power injection lead: http://linitx.com/product/12849 - similar to what I made up, but I don't know if it keeps the host connected on 5V or not. Jan 27 17:33:14 "Cannot allocate irq_descs @ IRQ16, assuming pre-allocated" Jan 27 17:33:29 http://paste.ubuntu.com/819067/ Jan 27 17:33:33 Is this important? Jan 27 17:33:35 Nice. I'll bet they have feedback diodes to prevent power pull to host. Jan 27 17:33:48 I've ordered one, will receive Tuesday Jan 27 17:34:48 That failure looks like what I am seeing. Jan 27 17:37:30 http://paste.ubuntu.com/819072/ is the most common failure I'm getting right now - no changes, just another run. No panic. Jan 27 17:42:40 A different type of panic: http://paste.ubuntu.com/819082/ Jan 27 17:43:27 Only changed was that I add DEBIAN_FRONTEND=text Jan 27 17:43:46 * rbasak brushes up on the english language Jan 27 17:43:46 Only change was that I added DEBIAN_FRONTEND=text Jan 27 17:51:29 This time I seem to have a random hang after "Starting system log daemon: syslogd, klogd." Jan 27 17:53:51 Another panic, looks like the same as one of the previous ones: http://paste.ubuntu.com/819098/ Jan 27 17:53:53 * rbasak gives up for now Jan 27 18:01:07 EOD. Enjoy your weekend when it starts! **** ENDING LOGGING AT Sat Jan 28 02:59:59 2012