**** BEGIN LOGGING AT Sat Nov 12 02:59:56 2011 Nov 12 11:09:23 SHR: 03Martin.Jansa 07shr-chroot * rf61db9a94dda 10/ (268 files in 26 dirs): system upgrade Nov 12 11:13:52 lindi-_: so, about your last patch, do you want me to push it exactly the way you've sent it? Nov 12 11:17:47 PaulFertser: which ubifs change did you push? Nov 12 11:18:37 JaMa: http://git.openmoko.org/?p=qi.git;a=commitdiff;h=6c02038d6bab6977fddda7736887b95807b064ce Nov 12 11:19:19 ah, that's from qi-ubi OE recipe? Nov 12 11:19:49 then I guess I have to redo this patch to be reverted in classic qi(-jffs2) recipe when I update to latest SRCREV Nov 12 11:20:09 PaulFertser: (phonet under maemo) right you are... it does not seem to use kernel side things Nov 12 11:20:14 and I'm not sure that everybody will agree on ubi vol name ubi0:om-gta01-rootfs Nov 12 11:20:21 I have phonet0 and upnlink0 but pnroute output is empty Nov 12 11:20:38 probably doing it in proprietary user space... Nov 12 11:21:40 PaulFertser: at least that has been tested yes Nov 12 11:22:02 PaulFertser: unless you have some other objections than style, that is. you can fix style afterwards too :) Nov 12 11:22:44 and phonet shows an address for phonet0 but none for upnlink0 Nov 12 11:23:04 PaulFertser: now away for some hours Nov 12 11:29:16 mrmoku: so, what about the "correct" way that you got working some time ago? Nov 12 11:29:42 lindi-_: ok, pushing shortly. Nov 12 11:31:30 PaulFertser: my memory sucks :/ Nov 12 11:33:35 JaMa: why do you specify the erase size explicitely Nov 12 11:35:55 where? Nov 12 11:36:13 ubi.mtd param? Nov 12 11:37:45 isn't that "smallest flash I/O unit" ? Nov 12 11:39:25 or even offset "Why does ubiattach on a freshly formatted device fail with "Invalid argument"?" http://www.linux-mtd.infradead.org/faq/ubi.html Nov 12 11:45:39 JaMa: is this subpage problem still makes sense with newer kernels? Nov 12 11:46:13 haven't tested without, since .32 Nov 12 12:31:34 good morning Nov 12 12:40:32 mrmoku: did you rebuild already after TCLIBCAPPEND drop? Nov 12 12:41:23 mrmoku: if yes, I'll push that to master so everybody can rebuild or put it to local.conf Nov 12 12:56:14 JaMa: hmm Nov 12 12:57:11 OE om-gta04@shr ~/shr-core $ ls -ld tmp* Nov 12 12:57:11 drwxr-xr-x 12 bitbake bitbake 4096 Nov 1 15:38 tmp Nov 12 12:57:11 drwxr-xr-x 12 bitbake bitbake 4096 Nov 6 09:54 tmp-eglibc Nov 12 12:57:29 it is building in tmp-eglibc... so I would think yes Nov 12 12:58:14 actually... I never corrected the TCAPPEND --> TCLIBCAPPEND :P Nov 12 12:58:53 JaMa: btw. I had twice the libstdc++ problem with my current build... and both times it was fixable by touch + ldconfig Nov 12 13:00:56 JaMa: the dbus problem finally fixed? Nov 12 13:01:29 mrmoku: seems so Nov 12 13:01:41 great Nov 12 13:01:51 mrmoku: did you build gcc or binutils betwee those 2 libstdc++ problems? Nov 12 13:01:55 JaMa: my current build is for gta04 with systemd... there I have a serial console Nov 12 13:02:09 JaMa: not manually Nov 12 13:02:21 I'll drop dbus work around (keep only PR) after opkg change is applied upstream Nov 12 13:02:27 good Nov 12 13:02:46 mrmoku: it happens also when you build ie another MACHINE Nov 12 13:02:55 which needs another gcc/binutils Nov 12 13:03:05 yeah, I switched from qemuarmx86 to gta04 Nov 12 13:03:17 and then it happened twice Nov 12 13:03:20 that could explain it.. I'll experiment with it after libreoffice build Nov 12 13:03:28 ok :) Nov 12 13:03:48 * mrmoku hopes to be able to do some systemd work with gta04 and its serial console Nov 12 13:04:51 * JaMa will probably broke up with gf and have more time for systemd work .. Nov 12 13:05:03 heh Nov 12 13:06:02 JaMa: I hope systemd won't be the only reason :P Nov 12 13:06:21 that would be a pity then... Nov 12 13:06:36 yeah :) not the only reason Nov 12 13:06:53 ok :) Nov 12 13:06:58 brb Nov 12 13:07:11 * JaMa is not _that_ crazy :) Nov 12 13:09:58 SHR: 03Martin.Jansa 07meta-smartphone * r11f5fab0c4e3 10/meta-shr/conf/distro/shr.conf: meta-shr: drop TCLIBCAPPEND, add it to local.conf if you don't want to rebuild now Nov 12 13:10:08 SHR: 03Martin.Jansa 07meta-smartphone * re51b68e076d6 10/meta-openmoko/recipes-multimedia/mplayer/ (mplayer2/tremor.build.fix.patch mplayer2_git.bbappend): mplayer2: force tremor instead of libvorbis Nov 12 13:23:26 ahh... image finished :-) Nov 12 13:37:46 JaMa: reading that mtd more closely, it appears you're doing it wrong then. You're forcing a kernel driver to not use subpages and that makes ubi less efficient. Nov 12 13:38:17 JaMa: of course the ubi image should be created with the right subpage value in the first place. But since you know the target flash, it shouldn't be a problem. Nov 12 13:38:59 yes it's less efficient but it didn't work without (with .32 or .31) Nov 12 13:39:01 What am i missing? Nov 12 13:40:27 JaMa: i guess that's because you're creating the images without specifying the right subpage size. Nov 12 13:41:42 too long for me to remember the exact error http://git.openembedded.org/openembedded/commit/?id=e7a640aa3d9288df15a3a5a8750fab4a8be276e9 Nov 12 13:43:08 according to git log I was using .32 around that time Nov 12 13:45:19 that commit does not unfortunately explain why it is done.. Nov 12 13:48:35 JaMa: hmm... udevd got pulled in Nov 12 13:48:40 and started Nov 12 13:49:29 yes.. it's from systemd Nov 12 13:50:07 JaMa: does it essentially depend on it? or will it work without? Nov 12 13:50:09 we can get rid of it later Nov 12 13:50:14 ok Nov 12 13:50:25 it should work without imho Nov 12 13:50:32 but udev is not so slow anymore Nov 12 13:50:33 * mrmoku needs to build a kernel with CGROUPS Nov 12 13:50:41 maybe we could enable it again later Nov 12 13:54:43 what about old kernels Nov 12 13:54:51 like palm pre Nov 12 13:55:10 udev is slow on theses devices Nov 12 13:55:50 and what about n900 that has a 2.6.37 kernel Nov 12 13:56:03 it's too old for last udev Nov 12 13:56:19 i'm not taking a side... Nov 12 13:56:35 i'm just asking... Nov 12 13:56:45 GNUtoo-N900: you can just backport the select4 patch Nov 12 13:56:54 ah? Nov 12 13:57:04 GNUtoo-N900: I used latest udev on 2.6.34 for a long time Nov 12 13:57:12 sys-accept4 is easy to backport Nov 12 13:57:22 ah, accept4, not select4 Nov 12 13:57:28 GNUtoo-N900: there's some new issue? Nov 12 13:57:34 does it have a bug report? Nov 12 13:57:39 but thats not the only thing to backport Nov 12 13:58:07 GNUtoo-N900: interesting, afaik that was enough with 2.6.34 Nov 12 13:58:35 i backported sys accept 4 on 2.6.27....and it didnt made udev work Nov 12 13:59:01 and if i remember well the palm pre kernel is even older than that Nov 12 13:59:14 25 or somethinh like that Nov 12 13:59:42 GNUtoo-N900: 2.6.27 is too old, you mean 2.6.37? Nov 12 13:59:58 no I meant 27 Nov 12 14:00:05 GNUtoo-N900: ok, that is too old definitely Nov 12 14:00:33 the bug device 1.x still has 27 Nov 12 14:00:43 and the palm pre is even older Nov 12 14:00:52 let me look Nov 12 14:00:56 GNUtoo-N900: yeah they need newer kernels Nov 12 14:00:57 .24 iirc Nov 12 14:01:15 but I didn't mean that we have to enable it for all devices.. Nov 12 14:01:39 just reevaluate it later for devices with new enough kernels if it's fast Nov 12 14:02:32 ok Nov 12 14:02:41 what device systemd images work on right now? Nov 12 14:03:01 all wip Nov 12 14:03:36 GNUtoo: I'm trying it on gta04 as there I have a nice serial console Nov 12 14:04:09 ok Nov 12 14:04:18 should I try on gta02? Nov 12 14:04:21 I've serial console Nov 12 14:04:31 I've serial console on crespo too Nov 12 14:04:49 but it crashes during boot if I enable it Nov 12 14:04:52 some bad kernel option Nov 12 14:07:59 it's not really working yet.. Nov 12 14:08:34 so it depends if you have something more important to work on and need working image then stay with sysinit a bit longer Nov 12 14:10:39 GNUtoo: yeah, JaMa is right. We should not all three work on systemd in that state Nov 12 14:11:40 ok Nov 12 14:12:11 btw is the midori segfault at startup fixed? Nov 12 14:25:39 SHR: 03Martin.Jansa 07shr-makefile * radd788602ec6 10/conf/shr-core/ (local.conf site.conf): site.conf, local.conf: add few usefull examples and drop unneeded stuff from site.conf Nov 12 14:53:45 hi Nov 12 15:06:51 leviathan: ping Nov 12 15:07:05 leviathan: possible current state of gta04 master is not good? Nov 12 15:11:28 : which shr image should i install on an GTA02 ? is shr allready migrated to oe-core ? Nov 12 15:13:12 pwgen: we work on shr-core only Nov 12 15:13:35 though things are not yet settled down enough to be equivalent to shr-unstable I think Nov 12 15:13:47 you can give it a try though Nov 12 15:14:00 is the shr-unstable allready build with oe-core ? Nov 12 15:14:04 no Nov 12 15:14:14 only shr-core is built with oe-core Nov 12 15:17:54 * DocScrutinizer yawns Nov 12 15:18:51 ~nuke systemd Nov 12 15:18:52 * apt prepares 100 missle silos, and targets them at systemd ... B☢☢M! Nov 12 15:19:20 * DocScrutinizer hopes Poettering been no farther than 10m from it Nov 12 15:20:05 why is systemd bad(I don't know enough of it) Nov 12 15:20:10 *about Nov 12 15:20:42 rnadom quote of last 48h: >>systemd is a PITA to debug, no logs, no way to intercept anywhere...<< Nov 12 15:20:57 ouch Nov 12 15:21:05 >>there's really little that can be done to find out *why* things don't work...<< Nov 12 15:21:39 and then it's Nov 12 15:21:42 ~poettering Nov 12 15:21:43 'sth is poettering' means it acts invasive, possessive, destructive, and generally in an egocentric exacerbating negative way. ``this cancer is extremely poettering'' Nov 12 15:21:50 lol Nov 12 15:22:05 like pulse Nov 12 15:22:19 yeah, like polypaudio Nov 12 15:23:16 this guy is such a douche bag Nov 12 15:23:46 lol hmmm Nov 12 15:26:42 may the ghost of Dennis MacAlistair Ritchie possess him and curse him forever Nov 12 15:29:50 I sometimes wonder "if unix community is a meritocracy then WTF is grabting this Poettering guy credit? He never went to pay back" Nov 12 15:30:06 granting* Nov 12 15:31:53 what unix community? Nov 12 15:31:54 sometimes it seems to me linux is a pretentiocracy rather than a meritocracy Nov 12 15:34:52 one thing's for sure: with guys like Shuttleworth shoving BS like buntkuh down the throat of millions of igmorant unix noobs linux hasn't become better over-all quality, neither in design nor in spirit of user base, and the very least in quality of distributions published Nov 12 15:36:30 hell, meanwhile I hear a "but it doesn't look/do-things like winYZ" every now and then, and it's meant as criticism :-(( Nov 12 15:39:04 my answer to that frequently would be a "IDIOTS, if you want sth to look like windoze, go for M$ redmond BS and leave linux alone" - but I prefer to shut up and shake head in silent desperation Nov 12 15:47:04 good, recommending windows would be recommending non-free software :) Nov 12 15:49:37 DocScrutinizer: http://fedoraproject.org/wiki/How_to_debug_Systemd_problems Nov 12 15:49:46 looks like there's lots of possibility to debug stuff Nov 12 15:50:00 I'm just quoting Nov 12 15:50:33 * mrmoku too Nov 12 15:50:38 did not try any of the stuff there Nov 12 15:50:45 and hoping all my systems will live on with sysV init rather than systemd or upstart for a loooong time Nov 12 15:50:54 that will hopefully change when I get the 3.1 running on my gta04 Nov 12 15:51:03 DocScrutinizer: even your phones? ;) Nov 12 15:51:15 as trying to deal with poettering stuff inevitably causes nausea on me Nov 12 15:57:29 mrmoku: >>Add systemd.log_level=debug to the kernel command line to log systemd with debug output enabled.<< UHUH, the most easy and obvious thing to do, esp on embedded like N900/NOLO-bootloader Nov 12 15:58:12 DocScrutinizer: you can't add boot options on n900? Nov 12 15:58:26 how ould you do that? Nov 12 15:58:59 no idea, I didn't buy it Nov 12 15:59:09 but sounds quite crippled if you can't do that :/ Nov 12 15:59:18 ...when the bootloader is closed and has hardcoded kernel cmdline Nov 12 15:59:38 and generally no way to edit boot parameters Nov 12 16:00:10 I'd blame non-free software and not systemd here :) Nov 12 16:00:17 HAHAHA Nov 12 16:01:03 yeah, poettering inventing new shit, and if it doesn't work then it's the fault of everybody else not complying with what great MrPoettering invented Nov 12 16:01:37 seen same argument for polypaudio so many times Nov 12 16:01:54 I think it is called pulseaudio. Nov 12 16:02:04 ooh, he renamed it? Nov 12 16:02:50 I'll stay with the term polypaudio, sounds more descriptive for this PoS Nov 12 16:03:06 your argumentation does not impress me here :) Nov 12 16:04:00 which argumentation? Nov 12 16:05:42 this guy is such a douche bag Nov 12 16:07:06 you are entitled to an opinion but with that kind of statements you can't expect any useful results :/ Nov 12 16:08:05 I don't expect any results on those statements Nov 12 16:10:02 and Lennart has proven to not adhere to Nov 12 16:10:05 ~djs Nov 12 16:10:06 methinks djs is http://fun.sdinet.de/pics/english/conversation-flowchart.jpg Nov 12 16:10:15 on several occasions Nov 12 16:11:35 you can find mail threads, blog threads, even videos of CCC talks where he popped up showing cause for me to call hime douche bag Nov 12 16:19:36 PaulFertser: if you had committed the qi patch yesterday I could have used version number 20111111-1 Nov 12 17:09:02 I get libc error even on my gta02 -,- Downloading http://build.shr-project.org/shr-core/ipk//all/Packages.gz. Nov 12 17:09:03 wget: res_query.c:251: __libc_res_nquery: Assertion `hp != hp2' failed. Nov 12 17:10:40 lindi-: i did offer that to you yesterday ;) Nov 12 17:11:03 DocScrutinizer: hey Doc :) Nov 12 17:11:20 hey paul! :-D Nov 12 17:12:01 DocScrutinizer, CONFIG_CMDLINE Nov 12 17:15:14 freesmartphone.org: 03angelo 07aurora * r593b1db5c3d8 10/aurora-daemon/src/applications/app-settings/ (Makefile.am app-settings.info): aurora-daemon: add settings file for app-settings Nov 12 17:47:22 pespin: that error likes you :P Nov 12 17:47:37 pespin: I still think it has to do with your providers nameserver Nov 12 17:47:45 mrmoku, yeah, who the hell needs dns servers? let's use ips directly Nov 12 17:47:55 xD Nov 12 17:47:58 indeed ;) Nov 12 17:49:00 hmm I tried mplayer2 with -ac mad... it's better than mp3float, but still not usable enough.... Nov 12 17:51:29 pespin, do you care enough to improve mplayer1 enough to submit it to oe? Nov 12 17:51:40 because I've ported mplayer1 which is *FAST* Nov 12 17:51:46 I would if I had time Nov 12 17:52:18 but I have too much work from univ u.u Nov 12 17:56:30 PaulFertser: I started writing a brief qi man page Nov 12 18:00:30 PaulFertser: anything obviously broken in http://lindi.iki.fi/lindi/openmoko/qi.html ? (Yes I know it does not mention LEDs or command line arguments) Nov 12 18:06:49 lindi-: s/initrd/initramfs/ Nov 12 18:08:26 lindi-: "ext2 compatible" might benefit from more details. Also worth mentioning all the uSD partitions should be "primary"; and that it's allowed to use symlinks for the files. Nov 12 18:08:43 PaulFertser: fixed initramfs, refresh Nov 12 18:08:54 lindi-: other than that, looks cool Nov 12 18:08:58 PaulFertser: I don't know what "ext2 compatible" actually means here Nov 12 18:09:47 PaulFertser: "something that can be parsed as an ext2 filesystem"? Nov 12 18:11:16 lindi-: i think it should say it's possible to use ext3 (with or without a journal) and ext4 (without extents). I can't tell for sure if those are the only requirements. Nov 12 18:12:28 PaulFertser: rephrased, refresh Nov 12 18:13:19 man2html seems to add extra empty lines there, they are not visible with real man : Nov 12 18:14:07 lindi-: i think s/OpenMoko/Freerunner/ in the beginning of the Storage organisation section. Nov 12 18:15:04 PaulFertser: ah sure, updated Nov 12 18:16:06 lindi-: btw, Qi also works on gta01. uSD support should probably mention SDHC too. Nov 12 18:16:25 PaulFertser: I have never seen anybody use it on gta01 :) Nov 12 18:16:41 lindi-: i used it, and josch too :) Nov 12 18:16:46 so I'd rather not advertise it :) Nov 12 18:16:54 since I have not tested it Nov 12 18:17:12 the debian package also only is built for gta02 Nov 12 18:17:26 I'm trying to avoid getting hatemail from people who brick their gta01 :P Nov 12 18:17:52 lindi-: understandable ;) Nov 12 18:18:11 I'm sure gta01 users can read the source/wiki/README Nov 12 18:19:19 PaulFertser: is it important to mention uSDHC there? Nov 12 18:20:22 maybe I'll call it "SD" and then explain that both uSD and uSDHC work Nov 12 18:20:23 lindi-: well, some users might have questions if it is possible to use SDHC or not, i'd mention just in case i think. Nov 12 18:21:57 PaulFertser: added Nov 12 18:24:57 PaulFertser: "The kernel can be either in uImage or zImage format. Qi understand Nov 12 18:24:57 symlinks. Nov 12 18:24:59 "? Nov 12 18:25:21 lindi-: understands. Nov 12 18:25:56 lindi-: and probably SEE ALSO section mentioning README, not sure if that's needed or not. Nov 12 18:26:34 PaulFertser: sure Nov 12 18:26:45 PaulFertser: my problem is that the README is not really up to date Nov 12 18:28:30 leviathan: nvm. was my stupidity Nov 12 18:28:49 leviathan: but I don't get a serial console on ttyO2 - did that change to something else? Nov 12 18:31:13 PaulFertser: "Additional documentation is available in the README file and OpenMoko wiki. " Nov 12 18:32:28 lindi-: archlinux did it with the devtools devtools-20111111-1 :-P Nov 12 18:32:49 PaulFertser: my gta01 has been converted to a gta04 ;) Nov 12 18:33:01 (especially nice because of stereo audio) Nov 12 18:33:04 PaulFertser: ha! Nov 12 18:33:28 josch: can you summarize the status of gta04 kernel? Nov 12 18:33:53 josch: what are you using for gsm telephony atm? Nov 12 18:33:59 josch: n900 with maemo? ;) Nov 12 18:34:08 lindi-: there is 2.6.32 which is ancient but works quite well - and there is 3.0 which has its rough edges... Nov 12 18:34:23 PaulFertser: nope - still my 10 year old nokia :( Nov 12 18:34:41 josch: so "quite well" means that suspend is reliable, audio/wifi works? Nov 12 18:34:52 josch: at least that's reliable Nov 12 18:35:05 lindi-: wifi works, audio works - suspend i didnt check Nov 12 18:35:12 PaulFertser: indeed :) Nov 12 18:35:19 josch: jikes, suspend is kind of critical :) Nov 12 18:35:31 or some other lowpower mode Nov 12 18:35:36 lindi-: i never suspend - my gta0X are always at usb0 ;) Nov 12 18:37:42 lindi-: is there a checklist of what should work? there is a list of what goldendelicious checks on each phone before shipping it Nov 12 18:38:08 lindi-: http://projects.goldelico.com/p/gta04-tester/ Nov 12 18:38:12 josch: but those checks are done using a non-free OS, I won't trust it :P Nov 12 18:38:41 lindi-: haha :) Nov 12 18:43:47 a personal blocker for me right now is, that there are some troubles with the framebuffer driver Nov 12 18:43:51 in the 3.0 kernel Nov 12 18:44:10 the image will be output a few pixel shifted down Nov 12 18:44:27 so that the upper border is pixel gibberish and the lower border of the image is off the screen Nov 12 18:44:49 josch: ok Nov 12 18:45:06 josch: I need a mobile device so energy consumption needs to be low Nov 12 18:45:28 lindi-: are you subscribed to the ml? Nov 12 18:45:41 josch: gta04? nope Nov 12 18:46:02 there was some recent mail about some power consumption improvements... wait... Nov 12 18:47:09 lindi-: http://lists.goldelico.com/pipermail/gta04-owner/2011-November/000638.html Nov 12 18:48:31 Sigh, why it's so damn hard to get power saving working right with those omaps? Nov 12 18:50:58 whoa, I keep forgetting that git-buildpackage nowadays has dbus freedesktop.org notification support Nov 12 18:51:33 and that xpra nowadays forwards such things to my local display even when I am building things on a remote machine Nov 12 18:51:54 wonders of modern technology for sure :) Nov 12 18:55:15 lol Nov 12 18:58:33 PaulFertser: anyways, I'm adding the man page to the debian package at least, you can merge it upstream later Nov 12 18:58:43 I'll need to test the package for some time and then upload to debian Nov 12 19:00:10 lindi-: that's really cool you're doing that packaging. Nov 12 20:19:14 JaMa: http://paste.pocoo.org/show/506870/ Nov 12 20:24:18 Starting sshd.service ESC[1;31mfailedESC[0m, see 'systemctl status sshd.service' for details. Nov 12 20:24:22 Starting Serial Getty on ttyO2 ESC[1;31mfailedESC[0m, see 'systemctl status serial-getty@ttyO2.service' for details. Nov 12 20:24:31 together mean I'm still doomed :/ Nov 12 20:34:42 JaMa: possible that volatile is problematic with systemd? Nov 12 20:41:58 JaMa: removing the run -> volatile/run link makes sshd start Nov 12 20:42:13 still no usb network Nov 12 20:42:31 I get an usb0 on my laptop but cant ping... looks like it is not configured Nov 12 20:44:58 I have a getty running on tty1... but the usb keyboard of my desktop does not work either Nov 12 21:26:24 mrmoku, do you have serial console? Nov 12 21:26:42 ah right gta04 Nov 12 21:37:17 GNUtoo: I could need some advise Nov 12 21:37:26 I do *not* get kernel output Nov 12 21:37:35 neither the getty on ttyO2 Nov 12 21:37:49 but I get syslog output Nov 12 21:38:06 mrmoku, ok Nov 12 21:38:11 then kernel cmdline is wrong Nov 12 21:38:14 back when I tried 3.1 the first time both worked Nov 12 21:38:16 ok Nov 12 21:38:30 CMDLINE_om-gta04 = "root=/dev/mmcblk0p2 rootwait ro console=ttyO2,115200n8 mpurate=800 vram=12M omapfb.mode=dvi:1024x768MR-16@60 omapfb.debug=y omapdss.def_disp=lcd" Nov 12 21:38:32 for instance on crepspo I've something strange Nov 12 21:38:43 /dev/s3c2410_serial2 is the serial Nov 12 21:38:52 but you have to do: console=/dev/ttySAC2 Nov 12 21:38:58 ,115200 Nov 12 21:38:58 ok, that would have changed then Nov 12 21:39:00 oh Nov 12 21:39:06 really? Nov 12 21:39:09 yes Nov 12 21:39:19 that's with the default kernel Nov 12 21:39:21 android.... Nov 12 21:39:31 »·······console=ttyS1,9600 console=tty0 Nov 12 21:39:41 is from kernel docs Nov 12 21:39:49 GNUtoo: does the kernel actually name these on android? Nov 12 21:39:52 do you still have your first kernel Nov 12 21:40:02 GNUtoo: or is it userland? Nov 12 21:40:08 no, I'm trying with 3.1. built in OE Nov 12 21:40:10 lindi-, booting the android kernel under GNU/Linux with devtmpfs produce that Nov 12 21:40:19 GNUtoo: ok Nov 12 21:42:11 for crespo I've an idea Nov 12 21:42:14 mtdoops Nov 12 21:50:22 maybe it's not a good idea Nov 12 21:51:44 GNUtoo: ok, I hacked an ifup -a into the phonefsod start script :P Nov 12 21:51:52 now usbnet + ssh is working :-) Nov 12 21:51:55 ok Nov 12 21:55:05 I think getty not working is due to udev Nov 12 21:55:22 ok Nov 12 21:55:27 I have no /dev/ttyO2 anymore Nov 12 21:55:28 not on 3.1 I guess Nov 12 21:55:40 https://bugzilla.redhat.com/show_bug.cgi?id=678875 Nov 12 21:56:35 indeed I'm aware of ttyS->ttyO change Nov 12 21:57:17 I have no console_check though... hmm Nov 12 21:58:09 GNUtoo: heh, indeed Nov 12 21:58:13 WOW!!!!!!!!!! Nov 12 21:58:15 root@om-gta04:~# mount -t devtmpfs devtmpfs /foo Nov 12 21:58:19 early printk is nice Nov 12 21:58:21 root@om-gta04:~# ls /foo/ttyO* Nov 12 21:58:21 /foo/ttyO0 /foo/ttyO1 /foo/ttyO2 /foo/ttyO3 Nov 12 21:58:25 it needed a kernel parameter..... Nov 12 21:59:06 http://www.pastie.org/private/og42qxiinpw3iphghnfxyg Nov 12 22:01:30 bootconsole Nov 12 22:01:31 hmm Nov 12 22:15:37 ok its not only udev Nov 12 22:15:46 ah? Nov 12 22:15:51 even with it disabled and an existing /dev/ttyO2 it does not work Nov 12 22:16:13 what about devtmpfs? Nov 12 22:16:16 is it activated? Nov 12 22:16:52 ahh Nov 12 22:16:54 wrong baudrate Nov 12 22:17:01 hmm... that might explain why it does not work Nov 12 22:17:13 calling agetty ttyO2 115200 manually gives me a login Nov 12 22:17:14 ok Nov 12 22:17:27 38400 is the default for getty Nov 12 22:17:53 yup and systemd has no freaking idea about the 115200 Nov 12 22:18:35 ExecStart=-/sbin/agetty %I 38400 Nov 12 23:00:31 http://shr-project.org/trac/wiki/Devices/GTA/InstallGuide still talks about jffs2 Nov 12 23:02:42 oh, gnutoo is not here anymore Nov 12 23:07:10 what should I flash to test that qi can boot SHR that uses 2.6.39? Nov 12 23:07:57 http://shr.bearstech.com/shr-core/images/om-gta02/shr-full-20111109-om-gta02.rootfs.ubifs Nov 12 23:08:00 this one? Nov 12 23:36:06 I get "UBI error: ubi_read_volume_table: the layout volume was not found" Nov 12 23:36:19 "UBI error: ubi_init: cannot attach mtd6" Nov 12 23:36:36 "Kernel panic - noy syncing: VFS: Unable to mount root fs on unknown-block(0,0)" Nov 12 23:36:41 ... Nov 12 23:38:42 does anybody else know? ;) Nov 12 23:54:16 "VFS: Cannot open root device "ubi0:om-gta02rootfs" or unknown-block(0,0)" Nov 12 23:54:50 and then it lists mtdblock0 .. mtdblock6 with "(driver?)" after each line Nov 12 23:55:20 and for completeness the kernel I tried is http://shr.bearstech.com/shr-core/images/om-gta02/uImage-2.6.39-r6-oe13-om-gta02-20111031080338.bin Nov 13 00:39:11 seems ".ubi" is the right one and boots. ".ubifs" doesn't :) Nov 13 00:43:58 good night people! **** ENDING LOGGING AT Sun Nov 13 02:59:56 2011