**** BEGIN LOGGING AT Thu Sep 01 02:59:57 2011 Sep 01 09:26:35 SHR: 03Martin.Jansa 07shr-chroot * rfc78d6e7bf55 10/ (2441 files in 227 dirs): system upgrade Sep 01 11:11:32 dos1: hi :) Sep 01 11:11:42 dos1: good that I did let you send those patches to oe-core ML :) Sep 01 11:36:10 SHR: 03Martin.Jansa 07meta-smartphone * r3aec4e8c106e 10/meta-openmoko/recipes-kernel/linux/ (linux-2.6.39/shr.patch linux_2.6.39.bb): linux-openmoko: upgrade to 2.6.39.4 and apply touchscreen patches from 2.6.37 Sep 01 11:36:20 SHR: 03Martin.Jansa 07meta-smartphone * r36ad029194d2 10/meta-zaurus/recipes-kernel/linux/ (linux-kexecboot_2.6.39.bb linux_2.6.39.bb): meta-zaurus: linux(-kexecboot): upgrade to 2.6.39.4 Sep 01 11:46:23 SHR: 03Martin.Jansa 07meta-smartphone * red82e43171d3 10/meta-shr/recipes-shr/tasks/task-shr-feed.bb: task-shr-feed: remove empty line from multiline variable Sep 01 12:20:16 JaMa: hi! what is the status of shr-core? because it is more and more difficult to build with shr-unstable.. last one is bitbake updated and not liking comments in recipes but recipes were not updated so ... can't parse Sep 01 12:24:34 GarthPS: use older bitbake if you care only about oe.dev.. Sep 01 12:27:27 JaMa: yep but as i sue the shr-chroot how do I do that ? Sep 01 12:31:08 make setup-bitbake Sep 01 12:31:45 JaMa: even with shr-chroot ? shrchroot with use it then ? Sep 01 12:32:08 setup-env should put it to PATH before system bitbake IIRC Sep 01 12:32:17 at least I had it this way for shr-core Sep 01 12:32:48 JaMa: ok I will try . thx Sep 01 12:34:06 GNUtoo|htcdream: pong Sep 01 12:35:34 JaMa: guess we have to find some solution for disk space? :/ Sep 01 12:40:17 mrmoku: well as long as we wont provide sstate -> can delete it every 2nd week, we're fine for few more years :) Sep 01 12:42:56 JaMa: for what do we need sstate exactly? Rebuild with correct revisions? Sep 01 12:49:11 JaMa: I did setup-bitbake but stil, setup-env use shr-chroot's bitbake as it is 1.13.3 and not 1.12 as fetched by setup-bitbake Sep 01 12:54:57 GNUtoo|laptop: os the pin problem fixed by reverting dos1's patch? Sep 01 12:55:13 mrmoku: which patch? ;o Sep 01 12:55:14 s/o/i Sep 01 12:55:31 dos1: phonefsod... backlight... which you reverted again :P Sep 01 12:56:11 dos1: don't think I'm not watching you... just because I'm sitting in front of a tent ;) Sep 01 12:56:26 mrmoku: yep, i reverted cause it was hack, and when i fixed fsodeviced hack was not necessary anymore :P Sep 01 12:56:47 and pin works again? Sep 01 12:57:12 mrmoku: no, pin wasn't working even before i looked at phonefsod source :P Sep 01 12:57:40 it is working for me though... problem probably is somewhere else then Sep 01 12:57:45 (on gta02 Sep 01 12:57:45 ) Sep 01 12:58:03 mrmoku: i haven't tested it on gta02 Sep 01 12:58:05 only n900 Sep 01 12:58:05 did not try on n900 for some time Sep 01 12:58:20 mrmoku: and on n900 there's some issue with fsogsmd i think Sep 01 12:58:26 if it works after restarting phonefsod Sep 01 12:58:26 yeah Sep 01 12:58:27 mrmoku: as it does not register to network automatically Sep 01 12:58:32 wanted to say just that :) Sep 01 12:59:24 i always have to call org.freesmartphone.GSM.Network.Register() manually, even after restarting phonefsod Sep 01 13:00:51 IIRC I implemented SetFunctionality and it even worked some weeks ago Sep 01 13:00:56 maybe months even Sep 01 13:02:24 ♬ regression here, regression there, bugs and regressions everywhere! ♬ Sep 01 13:02:26 ;) Sep 01 13:02:35 your new song? Sep 01 13:02:41 the regression blues? Sep 01 13:03:17 SHR tune :D Sep 01 13:03:21 :P Sep 01 13:04:50 anyway, we are here to proudly fight with them :] Sep 01 13:05:09 bugs busters :] Sep 01 13:05:23 unfortunatelly before 10th I won't be able to fight anything :/ Sep 01 13:06:23 mrmoku: it doesn't matter "when", it only matters "if" ;) Sep 01 13:06:56 ok :) Sep 01 13:08:34 (that could be some kind of my motto :D) Sep 01 13:13:18 * mrmoku back to vacation Sep 01 13:13:20 bbl maybe Sep 01 13:16:54 mrmoku: sstate is to download already prebuilt ie gcc/eglibc without need to build it locally Sep 01 13:18:04 GarthPS: what's in PATH? Sep 01 14:14:26 JaMa: /OE/bin:/usr/local/bin:/usr/bin:/bin:/opt/bin:/usr/x86_64-pc-linux-gnu/gcc-bin/4.6.1 Sep 01 15:11:21 GarthPS: do you have symlink from shr-unstable/bitbake to bitbake checkout? Sep 01 15:11:26 GarthPS: if no, add it Sep 01 15:11:35 GarthPS: http://git.shr-project.org/git/?p=shr-makefile.git;a=commit;h=6bead857240d13b0d81e46419f824d3d29fdc709 Sep 01 15:12:15 GarthPS: or better create different checkout of bitbake in shr-unstable/bitbake Sep 01 15:12:26 because makefile defaults to bitbake/master not bitbake/1.12 Sep 01 15:12:44 at least after http://git.shr-project.org/git/?p=shr-makefile.git;a=commit;h=e2991b2db8f806b087b74ed7931915a196f12cfa Sep 01 15:49:03 JaMa: meta-handheld? Sep 01 16:02:15 dos1: for my spitz :) Sep 01 16:02:38 dos1: I'll remove meta-zaurus from meta-smartphone as soon as meta-handheld is published properly Sep 01 16:02:51 JaMa: ok :) Sep 01 17:20:31 PaulFertser: ping? Sep 01 17:21:35 lindi-: pong Sep 01 17:21:38 lindi-: hi :) Sep 01 17:22:09 PaulFertser: qi fails to build due to new GCC being pedantic Sep 01 17:22:19 PaulFertser: I sent you a patch to fercerpav@gmail.com Sep 01 17:26:36 PaulFertser: I was just about to upload this to debian when I noticed the problem :) Sep 01 17:28:36 lindi-: ok, pushing this in 5 mins. Sep 01 17:28:43 PaulFertser: thanks Sep 01 17:30:21 lindi-: i see we didn't finish with that \r thing... :( Sep 01 17:30:31 * PaulFertser is feeling shame Sep 01 17:30:54 PaulFertser: it can wait Sep 01 17:31:16 I do have all those things written down in TODO Sep 01 17:33:21 lindi-: do you prefer to not use S-o-b? Or can i add it for you? Sep 01 17:34:45 PaulFertser: sure you can add Sep 01 17:35:12 I did carefully review that only the unused "resp" instances were removed Sep 01 17:35:28 heyho Sep 01 17:43:20 lindi-: can't push for some unknown reasons, http://paste.debian.net/128106/ :( Sep 01 17:43:43 lindi-: google suggests that might be wrong permissions in the git repo (no write access for the git user). Sep 01 17:44:04 PaulFertser: possibly Sep 01 17:44:22 PaulFertser: I would really have wanted to use "Applied-Upstream" DEP-3 tag for the patch when I send the package for review Sep 01 17:44:56 lindi-: i'm still online, let's try to get ahold of roh and/or gizmo? DocScrutinizer, can you help? Sep 01 17:45:14 wazzup? Sep 01 17:45:24 DocScrutinizer: http://paste.debian.net/128106/ Sep 01 17:45:51 errr Sep 01 17:45:58 lindi-: this is what i was going to do for the \r, not sure why i changed your version... Sep 01 17:46:23 DocScrutinizer: i'm trying to push to the Qi repo at git.openmoko.org but something goes wrong and i suspect shell access is required to sort it out. Sep 01 17:46:58 I don't think roh has an idea about what to do to fix it Sep 01 17:47:06 DocScrutinizer: but i do :) Sep 01 17:47:38 so you're asking for ssh access to ... which box? which user? Sep 01 17:47:45 DocScrutinizer: i'd check permissions for the repository, the objects/ and all the subdirs should be writable by the "git" user afaik. Sep 01 17:48:08 gimme domain name Sep 01 17:48:13 DocScrutinizer: git.openmoko.org Sep 01 17:48:19 hmmm Sep 01 17:48:48 I'm in Sep 01 17:48:49 DocScrutinizer: i do not really need/want ssh access there. Sep 01 17:49:14 waddyawannado? Sep 01 17:49:20 DocScrutinizer: ok, so please check the permissions for the qi repository, does git user has access there (usual ls -l) should suffice. Sep 01 17:49:32 mickeyl: ping Sep 01 17:49:42 gimme a cmdline Sep 01 17:50:55 ls -alR? Sep 01 18:03:57 lindi-: will push in 5 mins Sep 01 18:04:04 lindi-: what about the \r patch? Sep 01 18:05:06 PaulFertser: if you have it ready, sure go ahead Sep 01 18:05:22 lindi-: i'd like you to review it http://paste.debian.net/128107/ Sep 01 18:05:34 lindi-: because i do not remember why but it's not the same as yours. Sep 01 18:06:04 I probably won't add it to the package upload now since I have not tested it for sufficiently long time Sep 01 18:06:12 but in the next upload it will be there :) Sep 01 18:06:29 PaulFertser: it does look fine to me Sep 01 18:06:38 lindi-: so, pushed your patch already. Sep 01 18:06:56 Big thanks to DocScrutinizer! Sep 01 18:06:58 mplayer pauses but doesnt resume. related to alsa. there was even a bug report a long time ago. Sep 01 18:07:19 lindi-: so i'm pushing my \r patch and you're testing it later, ok? Sep 01 18:07:27 strace shows loop with nanosleep and ioctl with audio device (pcm...) Sep 01 18:08:06 mpd works. Sep 01 18:08:32 (but its not packaged on shr...) Sep 01 18:08:52 PaulFertser: yes Sep 01 18:09:14 PaulFertser: the todo list is http://lindi.iki.fi/lindi/openmoko/qi-todo.txt in case you feel bored.. Sep 01 18:15:01 lindi-: i see, thanks. \r pushed out. Sep 01 18:15:37 I already booted up my crypto-laptop :) Sep 01 18:15:53 lindi-: vt-d supported? Sep 01 18:16:14 no Sep 01 18:16:37 just a separate machine for doing crypto stuff Sep 01 18:20:35 PaulFertser: please ask Roh to add your pubkey to authorized_keys Sep 01 18:21:11 DocScrutinizer: ok, i'll add a ticket on the admin trac if i'll be able to :) Sep 01 18:21:26 you *should* be able to Sep 01 18:21:46 chandra works atm :-D Sep 01 18:24:12 PaulFertser: qi uploaded to debian :-) now let's wait what the ftpmaster has to say about it Sep 01 18:26:22 lindi-: yay. I appreciate your efforts and i hope i'm not the only one :) Sep 01 18:29:49 the copyright information in qi is not exactly simple so it might take a while for the ftpmaster to review it Sep 01 18:57:02 mrmoku: ping Sep 01 19:11:09 morphis: pong Sep 01 19:11:36 JaMa: ic ... so we can survive without Sep 01 19:22:30 hi gnutoo Sep 01 19:29:29 hi Sep 01 19:34:51 hi mrmoku Sep 01 19:35:42 gnutoo: still having pin problems? Sep 01 19:36:36 yes, and some more huge problems too: Sep 01 19:36:50 I got nervous because of u-boot Sep 01 19:37:03 why that? Sep 01 19:37:05 and the hdd + the white htcdream broke because of that Sep 01 19:37:26 pin does not work there too? Sep 01 19:37:31 u-boot for n900 doesn't work Sep 01 19:37:39 pin is broken everywhere Sep 01 19:37:43 gnutoo: hello... i am giving up... Sep 01 19:37:48 hmm Sep 01 19:38:02 see everyone gives up for uboot, JaMa nschle85 and me Sep 01 19:38:14 nothing changed on the shr side regarding pin Sep 01 19:38:42 sure it's not some fsogsmd problem? Sep 01 19:39:05 no Sep 01 19:39:12 I think it's some phonefsod issue Sep 01 19:39:18 restarting it makes the pin appear Sep 01 19:39:22 nothing changed there Sep 01 19:39:24 but I'm unable to bisect it Sep 01 19:39:24 yeah Sep 01 19:39:31 gnutoo: it can be race condition in fsogsmd as well Sep 01 19:39:38 ok Sep 01 19:39:39 but that is probably more related to a fsogsmd segfault Sep 01 19:39:43 gnutoo: i wrote an mail to JaMa and asked how to compile a working u-boot he answered me very fast but the result for my builds is alwys the same Sep 01 19:40:20 ok Sep 01 19:40:35 I cannot help on that ....sorry Sep 01 19:40:55 I already tried and broke an hdd and a phone..... Sep 01 19:41:04 and lost all mydata.... Sep 01 19:41:45 gnutoo: i tried openmoko toolchain, sourcery 2008.1 2008.3 2009.x .. n900 sdk and differnet versions of sources ... old oe dev, shr-core, ali1234 original patches... nothing will boot maemo kernel, u-boot starts but maemo kernel does not start Sep 01 19:42:21 maybe try assembly with the binutils.... Sep 01 19:42:35 or with gdb Sep 01 19:43:05 basically try to find out why it doesn't work looking at its assembly output Sep 01 19:43:18 what u-boot version are you trying? Sep 01 19:45:14 anyway... can't much help before 10th anyway Sep 01 19:46:05 ok Sep 01 19:46:13 nschle85: can't you use precompiled u-boot until it's resolved? Sep 01 19:46:19 gnutoo: hi Sep 01 19:46:39 hi Sep 01 19:47:04 gnutoo: i have the following schemes on implementing sound routing. i want to implement variant 2. any objections? Sep 01 19:47:06 http://pastebin.com/wgnLR7i7 Sep 01 19:47:13 mrmoku: dos1 i wanted to resolve or support a little bit but to start with that i nee a working u-boot Sep 01 19:47:48 nschle85: and u-boots from JaMa does not work for you too? Sep 01 19:49:29 let me look.... Sep 01 19:49:40 dos1: no, i tried one Sep 01 19:50:20 dos1: that did not work, do you have a link of an working u-boot from Jama ? Sep 01 19:50:42 nschle85: what maemo kernel do you need? Sep 01 19:51:13 dos1: is it relevant ? i combine it myself Sep 01 19:51:34 nschle85: it may be, as i don't know where i have non-combined parts on my hdd Sep 01 19:51:49 nschle85: and i have working u-boot here Sep 01 19:51:57 on my phone Sep 01 19:52:07 dos1: where is it from ? Sep 01 19:52:27 nschle85: from JaMa - downloaded from build.shr-project.org Sep 01 19:52:33 nschle85: but i did that almost year ago ;] Sep 01 19:52:49 nschle85: and didn't touch it since then Sep 01 19:53:05 and is that version still there ? Sep 01 19:53:18 dunno Sep 01 19:53:29 let me check if i have it somewhere Sep 01 19:53:31 hmmm Sep 01 19:53:40 dos1: thank you Sep 01 19:53:46 if not, i could try to read it from the device Sep 01 19:53:53 no objection Sep 01 19:54:05 but i have it combined with stock PR1.3 kernel Sep 01 19:54:23 dos1: that is the kernel i also use Sep 01 19:54:27 ok Sep 01 19:54:28 http://build.shr-project.org/shr-unstable/images/nokia900/u-boot-nokia900-2010.06+gitr0+bd2313078114c4b44c4a5ce149af43bcb7fc8854-r68.bin.ok Sep 01 19:54:36 looks like a good candidate to try :) Sep 01 19:54:43 gnutoo: ok, will try to implement it then. i just need to find some time since university has begun Sep 01 19:54:53 <[Rui]> hi all Sep 01 19:55:03 u-boot-nokia900-2010.06+gitr0+bd2313078114c4b44c4a5ce149af43bcb7fc8854-r68.bin.dos-like Sep 01 19:55:05 mrmoku: i think this one i tried and did not work but ill check immedialtely ! Sep 01 19:55:17 i have this one ;) Sep 01 19:55:21 nschle85: you still have to combine it IIRC Sep 01 19:55:38 mrmoku: i ll do Sep 01 19:56:48 I'll go bye Sep 01 19:56:52 nschle85: anyway, i've found http://dosowisko.net/combined.bin on my disk Sep 01 19:56:54 gnutoo: ^^^ Sep 01 19:57:21 ? Sep 01 19:57:31 gnutoo: you could try it if you want Sep 01 19:57:43 I've a working-for-me binary Sep 01 19:57:50 ok Sep 01 19:58:02 but I want that when you do an oe build that it works by default..... Sep 01 19:58:12 gnutoo: yup... Sep 01 19:58:34 ok ill try the version from shr build host first Sep 01 19:58:36 because the non-working binaries are ending up in the feeds! Sep 01 19:58:51 I'll go Sep 01 19:58:52 bye Sep 01 19:59:43 dos1: the ill try your version Sep 01 20:00:10 then Sep 01 20:06:39 mrmoku: the shr version does not work Sep 01 20:07:20 mrmoku: after typing run maemo last message is: starting kernel ... Sep 01 20:07:44 mrmoku: this is waht i get in all versions Sep 01 20:08:32 hmm Sep 01 20:09:11 nschle85: does it boot other kernels? Sep 01 20:13:47 dos11: i dont know i always tried to boot the standard maemo kernel Sep 01 20:14:07 nschle85: try to boot some from sd card Sep 01 20:15:31 dos1: do you know the excact layout of sd card for u-boot-pr13 to boot shr kernel ? the shr wiki is confusing me Sep 01 20:16:04 and the putput of N900 cat /proc/partitions and mount is confusing me too ! Sep 01 20:22:18 dos1: ping Sep 01 20:24:51 mrmoku: hihi, your link seems to be target dos1 :-) Sep 01 20:25:18 nschle85: i have just one ext3 partition Sep 01 20:26:23 dos1: then you do not use the maemo u-boot-pr13 Sep 01 20:26:35 nschle85: ah, that one Sep 01 20:26:49 nschle85: no, i'm using old SHR u-boot, as i said Sep 01 20:27:29 nschle85: use those u-boot links from wiki :) Sep 01 20:27:29 dos1: does it boot maemo Sep 01 20:27:35 nschle85: yup Sep 01 20:27:54 nschle85: and pali from kernel-power wrote me an e-mail about upstreaming u-boot patches Sep 01 20:28:24 JaMa|Off: for gta02? Sep 01 20:28:27 nschle85: so he has now R/W to gitorious u-boot repo and is trying to push it upstream (see u-boot ML for details) so it looks better Sep 01 20:28:31 lindi-: no, n900 Sep 01 20:28:37 JaMa|Off: ok :( Sep 01 20:29:06 lindi-: but he has also added simplified bootmenu Sep 01 20:29:57 JaMa|Off: ok i am too stupid: please send me a link to an u-boot which boots default maemo kernel, if it does not work, then something with me or my device is wrong Sep 01 20:30:12 and I've got just email about ali's new patchset for u-boto master :) https://github.com/ali1234/u-boot-n900 Sep 01 20:30:42 nschle85: which version do you want (dos-like or normal?) Sep 01 20:31:22 JaMa|Off: i want a version which boots a combined standard maemo kernel Sep 01 20:32:25 nschle85: did you really read http://wiki.shr-project.org/trac/wiki/Devices/NokiaN900/InstallGuide ? Sep 01 20:32:38 JaMa|Off: yes Sep 01 20:32:40 nschle85: because the difference is what should be booted by default with keyboard closed Sep 01 20:32:58 I don't know what you prefers and both versions boot Sep 01 20:34:15 JaMa|Off: i would to be happy to see any non u-boot-pr13 is booting a maemo kernel Sep 01 20:34:45 nschle85: wget http://build.shr-project.org/shr-unstable/images/nokia900/u-boot-nokia900-2010.06+gitr0+bd2313078114c4b44c4a5ce149af43bcb7fc8854-r68.bin.mmcblk0.dos-like Sep 01 20:35:22 wget http://build.shr-project.org/shr-kernels/nokia900/uImage-maemo-2.6.28-20103103+0m5 -O uImage Sep 01 20:35:53 JaMa|Off: ill combine it with maemo kernel and try to boot the combined kernel Sep 01 20:36:11 JaMa|Off: please wait.... Sep 01 20:37:45 JaMa|Off: ohh, patches against u-boot master... very nice :-D Sep 01 20:38:24 nschle85: gtg in few minutes.. Sep 01 20:40:46 JaMa|Off: did not work but i used my extracted maemo kernel which works with u-boot-pr13 Sep 01 20:41:04 JaMa|Off: starting kernel... is the last message Sep 01 20:42:09 should i also try the kernel provided by build host ? Sep 01 20:44:15 JaMa|Off: the md5sums are different Sep 01 20:45:16 nschle85: uploading combined images now Sep 01 20:45:37 u-boot-pr13 has different offset iirc Sep 01 20:45:46 but I'm using power-kernel only Sep 01 20:46:13 nschle85: http://build.shr-project.org/shr-kernels/nokia900/ Sep 01 20:46:47 nschle85: and gtg now, sorry Sep 01 20:47:17 JaMa|Off: which offset ? Sep 01 20:47:27 the 0x40000 ? Sep 01 20:48:34 JaMa|Off: ill flash the combined doslike now Sep 01 20:48:58 nschle85: https://gitorious.org/u-boot-shr/u-boot/commit/c76c4a060f7a92bc0974a231d40121844f29882e Sep 01 20:49:28 yes 0x38000 instead of 0x40000, which is what builder.py is using Sep 01 20:50:03 JaMa|Off: tried this too.. Sep 01 20:50:14 but different offset did not help Sep 01 20:50:22 ok let me flash ... Sep 01 20:55:19 JaMa|Off: same ffect the combined image does not boot its combined kernel , run maemo, starting kernel ... thats it Sep 01 20:56:45 JaMa|Off: did doslike combined image boot maemo ? Sep 01 20:58:35 JaMa|Off: so i should use power kernel ? Sep 01 20:58:51 JaMa|Off: as starting point ? Sep 01 20:59:19 * juri_ charges her 1973s. Sep 01 21:08:25 mrmoku: ping Sep 01 21:09:49 nschle85: I'm off :P Sep 01 21:10:09 mrmoku: ok :-( Sep 01 21:10:39 sorry and gnight Sep 01 21:13:17 nschle85, have you tested my combined.bin? Sep 01 21:15:52 no Sep 01 21:16:08 dos1|N900: sorry ill test now ! Sep 01 21:19:52 dos1|N900: hmm same effect, does not boot combined kernel Sep 01 21:20:09 dos1|N900: does it work on your phone ? Sep 01 21:28:44 dos1|N900: ping .... Sep 01 21:32:48 nschle85, yes, it works here :x Sep 01 21:39:41 nschle85: our phones want not to be hacked :D Sep 01 21:45:26 dos1|N900: strange Sep 01 21:45:57 angelox|laptop: i think too Sep 01 21:47:02 dos1|N900: puh... Sep 01 21:47:45 nschle85, wild guess: do you have R&D mode enabled or not? Sep 01 21:48:56 dos1|N900: i heard about it but i dont know, i bought the phone used by some other developer Sep 01 21:49:20 dos1|N900: how to check that ? Sep 01 21:49:41 dos1|N900: and what should it be on or off ? Sep 01 21:51:10 nschle85, dunno, I just want to find difference between our phones Sep 01 21:51:24 as it's working for me, and for you it's not Sep 01 21:51:30 i have R&D mode turned on Sep 01 21:51:38 so watchdog is disabled Sep 01 21:52:01 dos1|N900: ahh and how do i disable watchdog ? Sep 01 21:52:17 nschle85, google for enabling R&D mode on N900 ;) Sep 01 21:52:41 nschle85, using flasher Sep 01 21:52:59 (or 0xFFFF, but dunno if it's implemented there) Sep 01 21:53:20 is that persistent ? Sep 01 21:54:19 nschle85, yup Sep 01 21:54:47 i mean, you can disable it later of course Sep 01 22:00:06 dos1|N900: ok, so i try to enable it now Sep 01 22:00:19 dos1|N900: i have to google .... Sep 01 22:04:44 dos1|N900: i enabled it but it does not help Sep 01 22:14:51 dos1|N900: i think during booting any watchdog is restarting my phone Sep 01 22:15:13 dos1|N900: but how to disable it ? Sep 01 22:24:40 dos1|N900: tried to disable some watchdogs, but nothing helps Sep 01 22:42:13 nschle85, i have no idea :( Sep 01 22:43:47 dos1|N900: i am very frustrated because i may have a problem who nobody else has Sep 01 22:44:14 else me.. Sep 01 22:44:59 angelox|laptop: that gives me hope that iam not soooo stupid Sep 01 22:45:27 lol Sep 01 22:45:52 angelox|laptop: i think we should compare our hardware Sep 01 22:46:17 nschle85: how that? Sep 01 22:46:28 does N900 sput its hardware revision ? Sep 01 22:46:55 sput? sorry bad english Sep 01 22:47:15 german: spucken Sep 01 22:47:45 ahh spit Sep 01 22:49:01 i read in one of u-boot patches about problems with a specific hardware revision Sep 01 22:59:00 angelox|laptop: please google how to find the hardware revision of N900 Sep 01 23:00:58 flasher will tell you Sep 01 23:08:32 angelox|laptop: please cat /proc/cpuinfo Sep 01 23:09:14 my is: Hardware : Nokia RX-51 board Sep 01 23:09:14 Revision : 2104 Sep 01 23:09:30 my N900 won't boot Sep 01 23:09:52 angelox|laptop: battery problem ? Sep 01 23:09:56 I've sent it to the Support and they gave me a 'test' battery Sep 01 23:10:02 nschle85: you a) made sure your size of uboot+kernel fits into the mtd-partition? b) your offset where kernel starts in that partition is correctly set in uBoot? c) your MODULES in rootfs actually match the kernel in your combined uBoot+kernel image you flash to N900 NAND? d) you hurry a little to not spend most of the time til watchdog triggered reboot in uBoot menu? Sep 01 23:10:04 which doesn't neither boot Sep 01 23:11:16 DocScrutinizer: a) yes, builder.py checks that Sep 01 23:11:53 DocScrutinizer: b) using 0x40000 Sep 01 23:13:13 c) sure, because i am always combine the same kernel and unsing u-boot-pr13 with that kernel works Sep 01 23:13:20 0x40000 doesn't mean anything to me - I just know uBoot was already too large for that scenario without kicking a lot out of it, and there were problems with some builds of powerkernel as they were too large to leave space for uBoot Sep 01 23:14:54 DocScrutinizer: d) phone is rebooting but some seconds after u-boot tries to start the kernel Sep 01 23:15:22 also powerkernels with enabled framebiffer once been notorious to only boot 1 out of 3 times Sep 01 23:16:35 hmm, that also doesn't mean anything to me. I just know the time until watchdog reboot is sth like 30s from power-up. maybe less, but always constant Sep 01 23:19:03 so let's assume it's 30s, then when NOLO takes like 5s after press of power button, then uBoot takes a few seconds to load and start, then it lingers on in bootmenu and times out after 20s, then odds are kernel won't get up to a state where it starts to tickle watchdog Sep 01 23:19:24 DocScrutinizer: does u-boot send any heratbeats to the watchdog ? Sep 01 23:19:29 nope Sep 01 23:19:59 unless you got Pali's most recent (not yet submitted) uBoot patches Sep 01 23:20:01 DocScrutinizer: and which event is the the watchdog waiting for ? Sep 01 23:20:22 err, some read of a register afaik Sep 01 23:20:50 or write to a register? Sep 01 23:21:02 DocScrutinizer: i dont know Sep 01 23:21:30 me neither, just check watchdog driver of kernel, it will tell you Sep 01 23:21:49 DocScrutinizer: but why does maemo-pr13 uboot work and any others not ? Sep 01 23:22:03 original uBoot knew how to tickle watchdogs, in N900 uBoot it's configured out afaik Sep 01 23:22:38 DocScrutinizer: so which watchdoc has to be disbaled ? Sep 01 23:22:51 nschle85: I just can assume that's because they are somehow different -as obvious as it may sound Sep 01 23:23:04 probably all Sep 01 23:23:48 iirc there are at least 2 hw-watchdogs Sep 01 23:23:48 may this help to answer my question ? http://wiki.maemo.org/R%26D_mode Sep 01 23:23:58 sure Sep 01 23:24:47 use my addition for N900 flasher though, the original page was about N8x0 flasher Sep 01 23:24:49 DocScrutinizer: i am not an expert here, i need explicite instructions and ill test them Sep 01 23:26:49 no-omap-wd no-ext-wd (the latter in GAIA afaik) Sep 01 23:27:47 DocScrutinizer: ill try .... Sep 01 23:28:17 i.e. flasher-3.5 --enable-rd-mode --set-rd-flags= no-omap-wd,no-ext-wd [,?] Sep 01 23:28:59 DocScrutinizer51: what is [,?] for ? Sep 01 23:29:09 maybe flasher-3.5 --enable-rd-mode --set-rd-flags=no-omap-wd --set-rd-flags=no-ext-wd Sep 01 23:29:33 [,?] means I'm not sure about the comma Sep 01 23:29:34 DocScrutinizer: ok, that was also my question Sep 01 23:30:06 DocScrutinizer: ill try first with commy Sep 01 23:30:10 comma Sep 01 23:31:03 there's areadme and a manpage cming with flasher-3.5.tgz Sep 01 23:33:47 DocScrutinizer: nschle85@dual-wohnzimmer:~/u-boot-n900$ sudo ./flasher-3.5 --enable-rd-mode --set-rd-flag=no-omap-wd,noext-wd -f -R -k combined.bin Sep 01 23:33:48 [sudo] password for nschle85: Sep 01 23:33:48 flasher v2.5.2 (Oct 21 2009) Sep 01 23:33:48 Sep 01 23:33:48 Invalid flag 'noext-wd' Sep 01 23:34:10 DocScrutinizer: so that was my fault Sep 01 23:34:31 dang you mustn't flash sth!!! Sep 01 23:34:53 DocScrutinizer: nschle85@dual-wohnzimmer:~/u-boot-n900$ sudo ./flasher-3.5 --enable-rd-mode --set-rd-flag=no-omap-wd,no-ext-wd -f -R -k combined.bin Sep 01 23:34:53 flasher v2.5.2 (Oct 21 2009) Sep 01 23:34:53 Sep 01 23:34:53 Suitable USB device not found, waiting. Sep 01 23:35:12 DO *NOT* use -f ! Sep 01 23:35:29 neither any file parameter Sep 01 23:35:40 nor -k Sep 01 23:36:10 DocScrutinizer: ok, but parameters may be comma separated Sep 01 23:36:41 DocScrutinizer: ok, what should i use ? Sep 01 23:36:59 flasher-3.5 --enable-rd-mode --set-rd-flags= no-omap-wd,no-ext-wd Sep 01 23:37:12 minus the space after = Sep 01 23:37:59 maybe even just flasher-3.5 --enable-rd-mode Sep 01 23:38:12 then flasher-3.5 --set-rd-flags= no-omap-wd,no-ext-wd Sep 01 23:38:25 minus the space after = Sep 01 23:38:30 GRRR Sep 01 23:40:35 http://paste.debian.net/128134/ Sep 01 23:41:00 notably: -h, --hw-revision=ARG Specify product code and HW revision Sep 01 23:41:22 err --query-rd-mode Query the device R&D mode status and flags Sep 01 23:41:48 -i, --read-device-id Print out the device type Sep 01 23:45:19 maybe there's even a bug in the comma separated list for set-rd-flags Sep 01 23:45:45 then you'd need to use a separate command for each single rd-flag to set Sep 01 23:46:14 plus I'd think you should have found out by now, so I could stop :-D Sep 01 23:46:39 nschle85: still here? Sep 01 23:46:45 DocScrutinizer: yes Sep 01 23:47:15 DocScrutinizer: giving a comma separated list is the right way Sep 01 23:47:53 so what's the result of --query-rd-mode now? Sep 01 23:48:08 DocScrutinizer: angelox|laptop: has the same problems like me so we should compare cat /pro/cpuinfo Sep 01 23:48:24 hmm Sep 01 23:48:38 DocScrutinizer: what does it give on your device ? Sep 01 23:48:50 othing special Sep 01 23:49:15 DocScrutinizer: the version numbers are interewrsting Sep 01 23:49:24 Revision : 2101 Sep 01 23:50:22 Processor : ARMv7 Processor rev 3 (v7l) Features : swp half fastmult vfp edsp neon vfpv3 Sep 01 23:50:40 CPU part : 0xc08 Sep 01 23:50:42 CPU revision : 3 Sep 01 23:50:52 Hardware : Nokia RX-51 board Revision : 2104 Sep 01 23:51:06 sounds like a newer one Sep 01 23:51:20 but,the problem is,uboot-pr13 works in any revision :\ Sep 01 23:51:50 angelox|laptop: thats no problem, thats a paradoxum :-) Sep 01 23:52:22 angelox|laptop: whats your board revision ? Sep 01 23:52:47 i can't check Sep 01 23:52:49 my phone won't boot Sep 01 23:53:01 i'm sending a email to Nokia :D Sep 01 23:53:14 hahahahaha Sep 01 23:53:43 what's paradox in sth doesn't work despite it's not related to the board revision? Sep 01 23:55:13 to me it simply looks like you're messing sth up when building your uboot+kernel img, and I honnestly doubt you got that right with the kernel modules Sep 01 23:56:25 DocScrutinizer: i am combining ubooot meego-pr13 with meego kernel and it works Sep 01 23:56:49 nschle85: to me too Sep 01 23:57:10 or did you tell you're using same procedure/script to build, with same kernel, just with a small modification of uBoot source, and one works and the other doesn't? Sep 01 23:57:21 nschle85: combining meego's uboot with my own kernel works,so meego's uboot has the magic too :D Sep 01 23:57:42 DocScrutinizer: i am combining a lots kernels with u-boot mmego-pr13 and they does not work Sep 01 23:58:15 wtf? how are the modules supposed to match "a lots kernels"??? Sep 01 23:59:10 DocScrutinizer: sorry i meand i combined a lot of u-boots with maemo kernels Sep 01 23:59:25 maemo-kernel Sep 01 23:59:35 the one maemo kernel Sep 01 23:59:40 you ever consiered trying to flash the kernel image without packing it together with uBoot? Sep 02 00:00:30 DocScrutinizer: no, because it works combined with u-boot-pr13 (maemo u-boot) Sep 02 00:00:53 *sigh* Sep 02 00:01:39 DocScrutinizer: i think ill start a wiki page on shr and document what i have done Sep 02 00:02:33 so you'd think it's obvious to change a minimal thing (like "printf("this is a slightly modified uBoot, hope it still works despite this printf()")") Sep 02 00:03:13 DocScrutinizer: so i think everybody (except angelox|laptop: ) and me is able to boot maemo kernel with u-boot provided by JaMa|Off: / SHR Sep 02 00:03:29 uhuh Sep 02 00:04:00 DocScrutinizer: in which country do you life ? Sep 02 00:04:16 well maybe you got two or 3 bad blocks in your kernel partition, so it doesn't fit in Sep 02 00:04:25 SOL3 Sep 02 00:05:06 SOL3 ? Sep 02 00:05:28 http://wiki.openmoko.org/wiki/User:JOERG Sep 02 00:06:30 SOL3 is the 3rd planet in system SOL Sep 02 00:07:01 DocScrutinizer: may i send my device to you and you try yourself to get one of JaMa|Off: u-boots work ? Sep 02 00:07:02 hahaha Sep 02 00:07:07 in a small area of this universe that particular planet is called "earth" Sep 02 00:07:18 DocScrutinizer: That's funny :D Sep 02 00:08:26 nschle85: I'd prefer not to do that, as it requires me to picjk up some packet at a station 3km away, and sending it back is equally cimbersome Sep 02 00:08:31 cumber* Sep 02 00:09:29 DocScrutinizer: 3km ??? in by best time i did the distance under 12 minutes :-) Sep 02 00:11:19 nschle85: there's nothing I could do better with the device on my desk rather than getting complete logs of what you're doing Sep 02 00:11:53 I frown on your 0x40000 actually Sep 02 00:12:16 isn't that supposed to be 0x30000 sometimes? Sep 02 00:13:05 I don't believe in scripts I haven't designed the bugs myself Sep 02 00:14:25 DocScrutinizer: frown ? google did not translate it Sep 02 00:14:35 and - as already mentioned - free space in mtd3: 00200000 00020000 "kernel" get you in a pinch really Sep 02 00:14:49 ~dict frown Sep 02 00:14:55 Dictionary 'frown' (3 of 6): \Frown\, v. t. To repress or repel by expressing displeasure or disapproval; to rebuke with a look; as, frown the impudent fellow into silence. [1913 Webster] ;; look angry or sullen, wrinkle one's forehead, as if to signal disapproval ;; a facial expression of dislike or displeasure . Sep 02 00:15:02 DocScrutinizer: lets meet in separate channel, and talk in german Sep 02 00:17:35 I already gave all the ideas I can share: check that magic 0x40000, check your script doing the calculations (hell, do the calculations on paper to cross-check), and last not least: try with a shorter image as your partition might contain bad blocks that reduce usable free space in that partition Sep 02 00:18:54 for the bad blocks remember that's NAND Sep 02 00:21:08 DocScrutinizer: weeks ago somebody told me I am shizophren when i told him that i would like to use external sd instead of internal flash because internal flash could be damaged Sep 02 00:22:17 DocScrutinizer: i am not an expert in hardware, what should i care about bad blocks and NAND ? Sep 02 00:22:56 so hw-revision and CPU version won't tell you that much, it's /sys/class/ubi/ubi0/bad-peb-count that counts Sep 02 00:23:26 iirc Sep 02 00:23:30 0 here Sep 02 00:23:59 though I'm afraid that's only the bad blocks of rootfs partition Sep 02 00:24:57 0 here too Sep 02 00:26:39 t900:~# mtdinfo -a Sep 02 00:27:33 kernel has 16 eraseblocks here. Dunno. If a bad block klls a whole eraseblock, then you'd lose 1/16 if you got one bad block in "kernel" partition Sep 02 00:28:33 I think i should meet with JaMa|Off: in munich and he can analyze my phone Sep 02 00:28:53 DocScrutinizer: mtdinfo is not installed Sep 02 00:29:41 duh Sep 02 00:29:51 DocScrutinizer: which package is it Sep 02 00:30:05 lemme check Sep 02 00:30:53 hmm you're sure yiu're root? Sep 02 00:31:22 Nokia-N900:~# whoami Sep 02 00:31:22 root Sep 02 00:31:22 Nokia-N900:~# Sep 02 00:31:33 mtd-utils Sep 02 00:33:45 DocScrutinizer: mtdinfo -a gives a lot of output Sep 02 00:33:53 yeah Sep 02 00:34:18 not exactly any info bout bad blocks though (though I don't have any probably) Sep 02 00:35:39 you could md5sum /dev/mtd3ro and compare this with results of those who are using same image but can boot with it Sep 02 00:36:13 there's also nandread and nandwrite Sep 02 00:36:24 for whatever it's worth Sep 02 00:36:30 DocScrutinizer: http://pastebin.com/kzLzj61p Sep 02 00:37:09 same here Sep 02 00:38:14 DocScrutinizer: so i must decide: send my phone to you or to szech republic :-) Sep 02 00:38:42 well, I'm friggin out of ideas and I think they have to take the seat with me now, after >9h Sep 02 00:39:23 DocScrutinizer: or we (you and JaMa|Off: ) meet here end of september to octoberfest :-) Sep 02 00:39:24 consider my suggestion about md5sums Sep 02 00:39:58 DocScrutinizer: i understand.. Sep 02 00:40:17 sorry, I'm really completely out of ideas I could do with my magic paws on your hw Sep 02 00:40:24 DocScrutinizer: tomorrow i have one day free Sep 02 00:40:47 DocScrutinizer: i am too.. Sep 02 00:41:36 do >> md5sum /dev/mtd3ro << Sep 02 00:41:42 t900:~# md5sum /dev/mtd3ro Sep 02 00:41:43 b8db739ea8f2aec4da18ddd550791590 /dev/mtd3ro Sep 02 00:41:46 DocScrutinizer: so i should ask here to flash some images and compare the results with my hardware Sep 02 00:42:16 ^^^ doesn't mean anything to you as that's some interim version of mohammad's uboot with a weird version of powerkernel Sep 02 00:42:57 DocScrutinizer: i also compiled mohamads sources... Sep 02 00:43:05 well, if you get same md5sum on the kernel partition as somebody who can boot with this, then it's evident your hw is "different" Sep 02 00:43:42 if your sum is different though you flashed same uBoot+kernel image as somebody else who can boot, then your NAND is borked Sep 02 00:44:19 DocScrutinizer: my is 90df1a347fbd6d2439b59fee559a91da Sep 02 00:44:33 id md5sum of your uboot+kernel.bin flahser img is different though you bult the same as somebody else, then your process is sick Sep 02 00:44:59 nschle85: that doesn't mean a thing to me Sep 02 00:45:16 I don't have your kernel partition image Sep 02 00:45:51 DocScrutinizer: i know, ... Sep 02 00:45:55 JaMa should flash your image, and see if it works for him, then do a md5sum Sep 02 00:46:19 DocScrutinizer: i think tooo Sep 02 00:46:37 DocScrutinizer: but JaMa|Off: is very busy Sep 02 00:47:30 so you'll need to find sb else who's using same image as you do Sep 02 00:47:30 DocScrutinizer: so we should go to bed Sep 02 00:47:44 DocScrutinizer: its very late Sep 02 00:47:55 fsck yeah Sep 02 00:48:25 DocScrutinizer: you know how to contact me ? Sep 02 00:48:30 I need my 3h "sleep" before leaving home in 5h Sep 02 00:48:45 i.e I will watch TV now Sep 02 00:49:12 at pro7 the alm is shown Sep 02 00:49:28 hmm Sep 02 00:50:07 DocScrutinizer: i know... its stupid Sep 02 00:51:11 >> It is our pleasure to invite you to an exclusive - and private - Qt Innovation Challenge that is seeking amazing Qt-based applications for use on the Nokia N9. You are among an elite group of developers worldwide that is receiving this invitation.<< BWAHAHAHA Sep 02 00:51:13 Sep 02 00:51:57 DocScrutinizer: wow... Sep 02 00:52:42 DocScrutinizer: but why invest in dead technology ? for mobile phones ? Sep 02 00:53:02 maybe I can get a t-shirt ;-P Sep 02 00:53:14 or another N950 - MUHAHAHA Sep 02 00:53:33 DocScrutinizer: and a beer opener :-) Sep 02 00:53:52 DocScrutinizer: do you have a N950 ? Sep 02 00:54:12 yoh, though for the special Finnish bottles that are not avalable here in Germany Sep 02 00:54:27 yup, aegis coffin here Sep 02 00:54:38 * DocScrutinizer glares at friggin N950 Sep 02 00:55:14 DocScrutinizer: please ask nokia to develop a linux phone with hardware keyboard Sep 02 00:55:45 DocScrutinizer: open all the sources would also be very nice Sep 02 00:55:54 that's what makes this nvitation particularly funny - I'd not have gotten it when I'd not already tried to develop for Qt on N9(50) Sep 02 00:56:43 I'm not quite sure they knew why they sent that mail to me Sep 02 00:57:13 DocScrutinizer: but you can influence nokia to be a little be more open Sep 02 00:57:19 nschle85: I can't tell Nokia any good advice. All I can do is bitch at them and annoy them Sep 02 00:58:00 honestly lost case Sep 02 00:58:02 DocScrutinizer: i think nokisa will die with its windows phone Sep 02 00:58:12 I bet on it Sep 02 00:58:26 dead man walking Sep 02 00:58:42 but nokia is a good customer of us Sep 02 01:00:00 they begged Trollop err Elop to bring the shotgun and clean up, and now he does and it will be pretty clean when he's done with Nokia Sep 02 01:01:18 you *have* to be depressed and desperate to pay somebody for killing you Sep 02 01:05:00 DocScrutinizer: my english is to bad to understand Sep 02 01:05:47 nschle85: we are two :s Sep 02 01:06:53 err what? Sep 02 01:07:48 DocScrutinizer: ok now i understand Sep 02 01:08:18 sie haben Trollop err Elop gebeten seine pumpgun mitzubringen und bei Nokia aufzuraeumen, und das tut er nun und wenn er mit Nokia fertig ist wird es verdammt sauber sein. Sep 02 01:09:28 man muss ganz schoen depressiv und verzweifelt sein um einen killer fuer sich selber anzuheuern Sep 02 01:09:53 DocScrutinizer: this i do understand Sep 02 01:11:04 DocScrutinizer: i think MS buys nokia and builds then MS7 phones Sep 02 01:11:18 sure Sep 02 01:11:35 actually they already did Sep 02 01:11:49 somehow Sep 02 01:12:19 DocScrutinizer: so qt, symbian, and mobile knowledge goes to microsoft Sep 02 01:12:32 DocScrutinizer: and some patents Sep 02 01:12:58 and nobody understands what's the benefit of switching to winP7. No faster time-to-market, no more mature system, not more devices per year Sep 02 01:14:07 DocScrutinizer: its not a question of understanding, its a question of explaoning :-) Sep 02 01:14:12 all the reasons Trollop put why Nokia is switching from maemo/meego to winP7 are actually moot Sep 02 01:14:15 explaining Sep 02 01:15:14 DocScrutinizer: yesterday i bought the huawei X3 android phone. Sep 02 01:15:42 trollop: "I noticed it takes us too long to come up with a system based on maemo that's ready to use, so we switch to winP7" Reality check: N9 rolls out *now*, where are the winPhones made by Nokia? Sep 02 01:16:43 DocScrutinizer: but the maemo n900 is much more usable than the android bullshit Sep 02 01:17:29 I bet it is, that's what all renegades told when they came back :-D Sep 02 01:17:31 DocScrutinizer: except navigation Sep 02 01:17:48 yeah, navi on N900 is BS Sep 02 01:18:15 DocScrutinizer: on google android its excellent Sep 02 01:18:30 seems Sygic was way better, and even mappero is better. On N950 the maps app and drive app are not that bad Sep 02 01:19:05 DocScrutinizer: but there in no real navigation Sep 02 01:19:14 where? Sep 02 01:19:20 N900 Sep 02 01:19:36 mappero has turn-by-turn voice navigation Sep 02 01:19:53 marble will follow this month Sep 02 01:20:07 marble even does on-board routing Sep 02 01:20:08 DocScrutinizer: i did not know about this software Sep 02 01:21:03 yeah there's usually more apps around than a single user is aware of Sep 02 01:23:22 DocScrutinizer: its very late and my son will ware up 6.00 in the mornin Sep 02 01:23:41 DocScrutinizer: so ill leave Sep 02 01:23:47 good night Sep 02 01:24:38 if you write an email to me mys n900 will inform me Sep 02 01:25:20 bye Sep 02 01:25:54 good night people! **** ENDING LOGGING AT Fri Sep 02 02:59:57 2011