**** BEGIN LOGGING AT Thu Jul 08 02:59:57 2010 Jul 08 05:15:13 [2010-07-08 00:30:23] https://elektranox.org/website/debian_on_n900.html Jul 08 05:15:15 [2010-07-08 00:30:30] look the script at the end Jul 08 05:15:16 [2010-07-08 00:31:03] Battery Charging is done by closed source BME daemon in Maemo. Pancake wrote a script to charge the battery without this daemon based on information from the public available specs and the not working script from forum. Jul 08 05:15:32 Quite offensive. Jul 08 05:16:02 pnacake 'wrote a script' ? Bwahahaha Jul 08 05:19:33 seems to me like he just switched the status byte condition from 0x90 to 0x10, btw sth I never bothered to do in the "non working script" as I felt everybody savvy enough to feel he's 'allowed to use' my original script is also capable to check out the correct statusbit to tst for, for the paricular charging situation Jul 08 05:38:10 http://talk.maemo.org/showthread.php?p=649339#post649339 and esp http://talk.maemo.org/showthread.php?p=658278#post658278 Jul 08 05:39:24 proper attributing instead of claiming "wrote 'own script' form not working arbitrary unknown one" would have been much better style :-/ Jul 08 05:40:47 TAsn: thanks... but we were playing lousy this time :( Jul 08 05:43:19 moo mrmoku Jul 08 05:44:12 well... at least we had one evening without excessive noise Jul 08 05:44:31 mrmoku, aye. Can't say you were good. Jul 08 05:46:12 btw I told mickey|sad|zzZZz frequently about there's no more real showstopper to use N900. He never seemed interested, so it's quite bewildering to me he now thinks N900 *suddenly* became a candidate again. I mean the original charging script is of 05-14-10, 02:56 PM, and I'm sure I told mickey Jul 08 05:46:26 DocScrutinizer: hey :) Jul 08 05:47:09 DocScrutinizer: I heard about charging script only yesterday Jul 08 05:48:00 lindi-: maybe that's because you are not following #maemo closely, neither do you check my activities on talk.maemo.org Jul 08 05:48:34 possibly Jul 08 05:48:38 or occasionally look to sites like jrbme.garage.maemo.org Jul 08 05:49:12 admittedly rather empty there Jul 08 05:49:17 yet Jul 08 05:53:22 such a breakthrough should have been announced on openmoko list too :) Jul 08 05:53:52 lol, what has N900 charging to do with Openmoko? Jul 08 05:56:09 DocScrutinizer: well there's regular talk on alternative hardware Jul 08 06:05:34 lindi-: http://talk.maemo.org/showthread.php?t=57734 Jul 08 06:07:09 lindi-: I kept N900 operating without any battery for ~30min, and I'd be able to extend that if only I had used another "script" like http://talk.maemo.org/showthread.php?p=658278#post658278 Jul 08 06:07:32 lindi-: JFYI Jul 08 08:49:29 mrmoku`: if you're intersted to test the latest stuff, i've updated the ticket. Jul 08 08:50:14 mrmoku`: btw, if you want to help debugging fsogsmd i propose to always run it under strace, i suspect there might be some other parser problem. Jul 08 08:50:27 Q-Master: ^^^ same for you ;) Jul 08 08:50:55 PaulFertser: url? Jul 08 08:52:07 PaulFertser: you need some sort of log rotation for that Jul 08 08:52:39 lindi-: indeed. But i do not care, i've some Gb left there, and i'm not going to run it for ages. Jul 08 08:52:50 Q-Master: http://trac.freesmartphone.org/attachment/ticket/567/0001-fsogsmd-distinguish-between-pending-unsolicited-and-.patch Jul 08 08:54:56 PaulFertser: is this patch already applied and building? Jul 08 08:55:04 lindi-: while i was walking to work today, i was thinking about you a lot :) Jul 08 08:55:13 lindi-: ^_^ Jul 08 08:55:20 Q-Master: only locally on my device. Jul 08 08:55:33 bad Jul 08 08:55:52 Q-Master: it would be nice if you run fsogsmd straced on your device, as you say you can often reproduce some strange issues. Jul 08 08:56:33 lindi-: i mean it would be cool if you started testing fsogsmd, but i understand it's a considerable hassle for you. Jul 08 08:57:46 PaulFertser: current fsogsmd + that patch, right? Jul 08 08:57:47 PaulFertser: theoretically I can, but I have not much space left afaik. let me check later a bit Jul 08 08:58:58 mrmoku: right. The initial patch that solves the problem uncovered by "flowcontrolled" introduction is already in git, and this additional patch i cooked yesterday after we understood what's wrong with the message recetion (read the ticket for details). Jul 08 08:59:22 PaulFertser: yeah, read it... and your discussion last night Jul 08 08:59:36 Q-Master: stracing is better than DEBUG log because as we can see raw input, we can "replay" it through the parser whenever needed to check it's working ok. Jul 08 08:59:44 BTW, seems that after fsogsmd goes to it's dbus-stuck state the GSM starts to drain power from accum. I've woken up today by ffalarms and saw that it is almost 0% left in accum... Jul 08 08:59:45 PaulFertser: I still have GPRS troubles... so I will give it a try Jul 08 09:00:11 PaulFertser: I am unable to build new fsogsmd myself. Jul 08 09:00:12 Q-Master: I have logs of power drain by something too :/ Jul 08 09:00:14 Q-Master: would be very nice to have an strace of events leading to the "dbus-stuck" state. Jul 08 09:00:44 mrmoku: It started just after 2.6.32 update. Jul 08 09:00:46 PaulFertser: ok, then I will first try to strace it... as I can reproduce it quite reliably Jul 08 09:00:51 Q-Master: regardless, probably it's good because you'll probably be able to reproduce dbus-stuck state without it faster. Jul 08 09:00:56 Q-Master: yup Jul 08 09:01:03 PaulFertser: ok. let me try Jul 08 09:03:31 JaMa: fix plz /etc/profile.d/mc.sh and mc.csh. add -c after a wrapper script otherwise mc will be colourless. Jul 08 09:06:27 * PaulFertser . o O ( Why are russians so attached to NC-derivatives? ) Jul 08 09:07:39 * Q-Master attached to Amiga Directory Opus 4, but console's best is mc 8) Jul 08 09:15:05 PaulFertser: started it in strace mode. 8) will check today later. Jul 08 09:15:11 Q-Master: thanks :) Jul 08 09:16:14 PaulFertser: but if the bug will not happen today - experiments will continue only on monday, coz I'will move to DiHalt 2010 tomorrow morning and will return only on sunday night. Jul 08 09:16:54 Q-Master: wow, cool! Jul 08 09:40:19 PaulFertser: hmm... does not hit me under strace :/ Jul 08 09:41:35 PaulFertser: do you manually start fsogsmd or do you put strace into the activation file? Jul 08 09:44:44 mrmoku: I've started it manually Jul 08 09:45:07 Q-Master: me too Jul 08 09:57:51 PaulFertser: ok, got it hanging now :-) Jul 08 09:58:56 265 select(12, [0 11], [], NULL, NULL) = 1 (in [11]) Jul 08 09:58:56 265 read(11, "~!E\0\0T\0\0@\0@\1\220\325\n>M\335\301c\220U\10\0B\31B"..., 1504) = 89 Jul 08 09:58:59 265 write(1, "~!E\0\0T\0\0@\0@\1\220\325\n>M\335\301c\220U\10\0B\31B"..., 89) = 89 Jul 08 09:59:03 is what it's doing endlessly... Jul 08 10:00:11 before it was writing to 2 Jul 08 10:00:20 mickey|office: ^^ Jul 08 10:01:24 220 write(30, "~\23\357}^!E\0\0T\0\0@\0@\1\220\325\n>M\335\301c\220U\10"..., 96 Jul 08 10:01:26 265 <... select resumed> ) = 1 (in [11]) Jul 08 10:01:35 was the last thing before endless select-read-write loop started Jul 08 10:03:17 hmm Jul 08 10:03:19 l-wx------ 1 root root 64 Jan 1 19:53 /proc/220/fd/11 -> /var/volatile/log/fsogsmd.log Jul 08 10:03:27 why is it reading from 11? Jul 08 10:07:23 ohh... wrong process :P Jul 08 10:07:36 lrwx------ 1 root root 64 Jan 1 19:58 /proc/265/fd/11 -> /dev/ptmx Jul 08 10:32:53 hi Jul 08 10:33:09 mrmoku: did you get to testing the speaker during phone call? Jul 08 10:34:03 I'm looking into webkit-eff. Anybody knows what's the relation between upstream webkit and git://gitorious.org/webkit-efl/webkit-efl.git? Jul 08 10:35:37 looking into log of the git tree, it seems not much has been going on there Jul 08 10:35:55 while upstream seems to be getting some kind of activity on the efl front Jul 08 10:46:32 mrmoku: please put it somewhere Jul 08 10:47:13 zub: yep, does not work for me too :/ Jul 08 10:47:22 zub: did not come around investigating yet though Jul 08 10:47:48 PaulFertser: yeah, did another run with timestamp Jul 08 10:48:04 mrmoku: ok, at least I know it's global issue and I'm not seeing ghosts :) Jul 08 10:48:40 PaulFertser: have to run for lunch first though :/ Jul 08 10:48:43 bbiab Jul 08 10:51:45 hi mickey|office Jul 08 10:57:31 zub: webkit-efl was merged to upstream webkit Jul 08 10:58:05 zub: but nobody merged webkit-gtk and webkit-efl recipes to build shared parts just once and provide webkit-gtk and -efl only as subpackages Jul 08 11:00:38 JaMa: so, if OE recipes were updated, we could switch to upstream? Jul 08 11:00:46 and ventura would have to be updated... Jul 08 11:02:06 yes Jul 08 11:02:18 but if you look at those recipes it doesn't look so easy Jul 08 11:02:57 well, so far I only looked at webkit upstream code, and it's complex enough Jul 08 11:03:17 in the git tree, autotools is used, in upstream they use cmake Jul 08 11:05:52 zub: upstream webkit-efl also use cmake Jul 08 11:06:10 mickey|office: ping Jul 08 11:06:37 zub: https://lists.webkit.org/pipermail/webkit-dev/2010-May/012863.html Jul 08 11:08:13 zub: http://trac.webkit.org/changeset/59537 Jul 08 11:09:12 PaulFertser: http://build.shr-project.org/tests/mrmoku/fsogsmd/fsogsmd.strace.2 Jul 08 11:11:08 mrmoku: next time please use -s 9999 Jul 08 11:12:58 PaulFertser: I can redo it if it helps :) Jul 08 11:13:56 mrmoku: well, i need time to analise it. What exactly were you doing btw? using gprs? Jul 08 11:14:08 PaulFertser: just activated gprs and had a ping running Jul 08 11:14:10 nothing else Jul 08 11:14:12 mickey|office: i want to patch fsodeviced in order to support UsbHost at 2.6.32, should i support both .29 and .32 kernels or can i ignore .29? Jul 08 11:14:25 somehow when plugged into usb it does not trigger... or at least not that easily Jul 08 11:15:23 mrmoku: do you have some additional debug enabled or something? it looks like it mirrors all input to the stdout. Jul 08 11:16:03 PaulFertser: not that I would know of... I saw fsogsmd putting out PPP stuff to stdout or stderr Jul 08 11:16:19 ~nf Jul 08 11:16:20 The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-03-28 21:09 UTC]:buildhost got rebooted... new RAM will arrive beginning of next week Jul 08 11:16:37 fresh news ;] Jul 08 11:17:03 ~nf-help Jul 08 11:17:04 >>list old content (~literal NF); ~forget #openmoko-cdevel NF; c&p and edit old content, then set new (~#openmoko-cdevel NF is whatever-new-content). The leading #openmoko-cdevel makes the factoid channel specific. If you edit in a query, you invoke by '#openmoko-cdevel NF' Jul 08 11:18:17 NF Jul 08 11:18:24 NF Jul 08 11:18:34 ~NF Jul 08 11:18:35 The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-03-28 21:09 UTC]:buildhost got rebooted... new RAM will arrive beginning of next week Jul 08 11:19:27 ~NF2 Jul 08 11:19:28 (here you'll find additional info, if newsflash on ~NF is too short to hold all the info) Jul 08 11:19:42 ~RL Jul 08 11:19:43 methinks rl is Real Life (the bad world out there), or known as the Big Blue Box (tm). Conways game using only rational numbers Jul 08 11:19:54 ;o Jul 08 11:19:56 hello Jul 08 11:20:20 ~forget #openmoko-cdevel NF Jul 08 11:20:20 dos1: i forgot #openmoko-cdevel nf Jul 08 11:20:22 did someone has already try to make a map for mcnavi ? Jul 08 11:20:39 tried Jul 08 11:22:08 mrmoku: and what happens when you see that loop? Does fsogsmd stop responding on dbus? Jul 08 11:22:54 ~#openmoko-cdevel NF is SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster. Jul 08 11:22:55 okay, dos1 Jul 08 11:22:58 ~nf Jul 08 11:22:59 it has been said that nf is SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster. Jul 08 11:23:27 PaulFertser: yep, fsogsmd is playing dead Jul 08 11:23:41 PaulFertser: does not signal incoming calls either Jul 08 11:24:35 PaulFertser: it is ringing though, which means it is still registered Jul 08 11:25:03 ~#openmoko-cdevel NF is The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-07-08 11:25 UTC]: SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster. Jul 08 11:25:04 ...but #openmoko-cdevel nf is already something else... Jul 08 11:25:19 ~forget #openmoko-cdevel NF Jul 08 11:25:19 dos1: i forgot #openmoko-cdevel nf Jul 08 11:25:24 ~#openmoko-cdevel NF is The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-07-08 11:25 UTC]: SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster. Jul 08 11:25:25 dos1: okay Jul 08 11:25:26 ~NF Jul 08 11:25:27 somebody said nf was The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-07-08 11:25 UTC]: SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster. Jul 08 11:25:31 :( Jul 08 11:26:37 PaulFertser: I still have it in that state... in case you want some more info Jul 08 11:27:12 mrmoku: i'm not sure i can do anything right atm. You should have mickey|office take a look at that. Jul 08 11:27:41 PaulFertser: yep, will wait for mickey|office then Jul 08 11:27:52 * mrmoku reboots to get a working phone again :-) Jul 08 11:41:39 JaMa: thx for links Jul 08 11:42:10 * JaMa looking for patch from zub :) Jul 08 11:42:29 JaMa sounds like too much for me :-/ Jul 08 11:42:38 but, I'm poking into webkit as part of my job Jul 08 11:43:44 looking around so far Jul 08 11:51:47 gena2x: what settings did you use for the Glamo bankcon? Was it the same as http://www.mail-archive.com/openmoko-kernel@lists.openmoko.org/msg03657.html > Jul 08 11:53:01 or, if I remember correctly, you're altering the page settings and not the wait states? Jul 08 12:10:00 hi mickey|office Jul 08 12:10:08 could you ack/look at my oe patch? Jul 08 12:26:00 done Jul 08 12:34:09 mickey|office: i've thought a little about my misterious lost message issue and the only explanation that comes to my mind is that the at parser gets confused somehow. Alas i do not have strace log to see what lead to that state, i think it's time now to always run fsogsmd straced. Reading the source didn't help either, i couldn't construct a scenario in which the parser unrecoverrably fails. Jul 08 12:36:20 * DocScrutinizer sighs Jul 08 12:37:24 mickey|office, thanks a lot!!! Jul 08 12:37:45 which part of "">>list old content (~literal NF); ~forget #openmoko-cdevel NF; c&p and edit old content, then set new (~#openmoko-cdevel NF is whatever-new-content)"" was so hard to get? Jul 08 12:38:42 hmm, is "" part of instruction or part of command? ;x Jul 08 12:39:42 yes, of course it's a part Jul 08 12:40:19 as is ~literal, which had shown you the part Jul 08 12:40:26 reply Jul 08 12:40:56 ~literal #openmoko NF Jul 08 12:41:09 mhm Jul 08 12:41:34 ~nf Jul 08 12:41:35 The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-07-08 11:25 UTC]: SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster. Jul 08 12:41:44 #openmoko defaults to #openmoko-cdevel for undefined factoids Jul 08 12:42:06 ~literal #openmoko-cdevel nf Jul 08 12:42:07 "#openmoko-cdevel nf" is "The #openmoko-cdevel Newsflash Bulletin Board. (continued at ~NF2. For help see ~NF-help) - - - Recommended URLs and channels(chanlogs etc):see ~RL - - - NEWSFLASH [2010-07-08 11:25 UTC]: SHR-unstable moved to 2.6.32 kernel; gena2x works on overclocking Neo and making Glamo faster." Jul 08 12:42:38 somebody ruined the RL factoid. Dunno what happened to it Jul 08 12:42:57 weird thing is I can't even unforget it Jul 08 12:43:54 ~literal nf-help Jul 08 12:43:55 "#openmoko-cdevel nf-help" is ">>list old content (~literal NF); ~forget #openmoko-cdevel NF; c&p and edit old content, then set new (~#openmoko-cdevel NF is whatever-new-content). The leading #openmoko-cdevel makes the factoid channel specific. If you edit in a query, you invoke by '#openmoko-cdevel NF'" Jul 08 12:44:22 ~literal #openmoko-cdevel nf-help Jul 08 12:44:24 "#openmoko-cdevel nf-help" is ">>list old content (~literal NF); ~forget #openmoko-cdevel NF; c&p and edit old content, then set new (~#openmoko-cdevel NF is whatever-new-content). The leading #openmoko-cdevel makes the factoid channel specific. If you edit in a query, you invoke by '#openmoko-cdevel NF'" Jul 08 12:44:45 ~literal #openmoko nf-help Jul 08 12:44:57 ~rl Jul 08 12:44:58 Test Jul 08 12:45:20 dos!!!! Jul 08 12:45:34 now you even fucked up the real life! Jul 08 12:45:42 ~factinfo rl Jul 08 12:45:42 rl -- last modified at Tue Aug 26 04:38:55 2003 by sgeigerbot!~user@139.55.144.113; it has been requested 5 times, last by dos1, 1m 19s ago. Jul 08 12:45:53 not really :-P Jul 08 12:46:00 ~rl Jul 08 12:46:01 i heard rl is Real Life (the bad world out there), or known as the Big Blue Box (tm). Conways game using only rational numbers Jul 08 12:46:04 DocScrutinizer: :PPP Jul 08 12:46:04 ~factinfo #openmoko-cdevel rl Jul 08 12:46:04 there's no such factoid as #openmoko-cdevel rl, DocScrutinizer Jul 08 12:47:32 ~factinfo logs Jul 08 12:47:32 logs -- last modified at Mon Mar 1 19:53:28 2010 by TimRiker!~timr@bzflag/projectlead/TimRiker; it has been requested 298 times, last by hape, 3h 15m 40s ago. Jul 08 12:47:42 ~factinfo #openmoko-cdevel logs Jul 08 12:47:42 #openmoko-cdevel logs -- created by DocScrutinizer <~jr@openmoko/engineers/joerg> at Tue Mar 30 02:13:19 2010 (100 days); last modified 2d 38m 36s ago by DocScrutinizer!~halley@openmoko/engineers/joerg; it has been requested 9 times, last by DocScrutinizer, 2d 38m 17s ago. Jul 08 12:47:55 ~literal logs Jul 08 12:47:56 "#openmoko-cdevel logs" is " logs are on http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged. logs are updated daily --- livelogs are http://logs.nslu2-linux.org/livelogs/openmoko-cdevel.txt and http://hentges.net/tmp/logs/irc/livelogs/#openmoko-cdevel.livelog" Jul 08 12:48:21 get the picture? Jul 08 12:49:40 so please do NOT redefine ~RL, rather define ~#openmoko-cdevel RL Jul 08 12:49:58 i didn't redefine ~RL ;] Jul 08 12:50:12 somebody did Jul 08 12:50:27 as I created it some day in the past Jul 08 12:50:31 for THIS channel Jul 08 12:50:37 sgeigerbot!~user@139.55.144.113 Jul 08 12:50:50 otherwise I wouldn't have included it to the /topic Jul 08 12:51:13 mickey|office, JaMa thanks pushed Jul 08 12:51:25 last modified at Tue Aug 26 04:38:55 2003 by sgeigerbot Jul 08 12:51:27 !!! Jul 08 12:52:00 oh :D Jul 08 12:52:10 dos1: *I* created a #openmoko-cdevel RL factoid *this* year, not 2002 Jul 08 12:52:43 and it vanished Jul 08 12:52:48 DocScrutinizer: but now i don't understand you ;x Jul 08 12:52:54 [14:49] so please do NOT redefine ~RL, rather define ~#openmoko-cdevel RL Jul 08 12:53:20 which is a pitty as it contained all the recommended links I cleaned out of the overlength /topic Jul 08 12:53:51 dos1: this was a generic advice, no criticism addressed at you Jul 08 12:56:16 hi JaMa Jul 08 12:56:31 you use initscripts-openmoko right? Jul 08 12:56:37 for gta-0* Jul 08 12:57:00 yes Jul 08 12:57:10 PaulFertser: ok, thanks, keep me posted Jul 08 12:57:41 GNUtoo|laptop: IMAGE_INITSCRIPTS = "initscripts-shr" in shr.conf (should be probably in shr-image.inc) Jul 08 12:58:05 JaMa, should I use that too for dream Jul 08 12:58:05 ? Jul 08 12:58:07 I bet so Jul 08 12:58:10 because of: Jul 08 12:58:19 .etc/init.d/devices Jul 08 12:58:21 oops Jul 08 12:58:30 /etc/init.d/devices Jul 08 12:58:45 basically it get pulled with normal initscrpits Jul 08 12:58:48 yes you should use it.. it's not machine specific Jul 08 12:58:51 and nothing work after that Jul 08 12:58:53 ok thanks a lot Jul 08 12:59:16 GNUtoo|laptop: that's probably because normal initscripts doesn't mount /dev/pts for you Jul 08 12:59:48 I have devtmpfs Jul 08 12:59:57 so I don't want /etc/init.d/devices Jul 08 13:00:19 GNUtoo|laptop: you need this http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=517816ac3c3d43cb53569955ad6705f468da6cf4 Jul 08 13:00:20 removing it fix issues like gsm not available because no device node etc... Jul 08 13:01:10 initscripts - 1.0-r121.5 Jul 08 13:01:15 that's what I have on device Jul 08 13:01:21 I really want shr initscripts Jul 08 13:01:24 I'll look Jul 08 13:01:34 which image are you building? Jul 08 13:01:58 shr-image Jul 08 13:02:03 that's why it's strange Jul 08 13:02:05 JaMa: btw, i had old initscripts-shr, due to changed versioning scheme ;x Jul 08 13:02:11 maybe it's because I've an old image Jul 08 13:02:14 JaMa: so that could be a reason of non-working gsm Jul 08 13:02:36 dos1: non-working gsm was probably this http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=38ee732803984de88a449b8ba8296a5deda62771 Jul 08 13:03:17 hmmm no Jul 08 13:03:19 ah ok Jul 08 13:03:20 JaMa: yup, if i had old initscripts-shr, i couldn't have this fix :P Jul 08 13:03:21 sorry Jul 08 13:03:51 I'll grep for IMAGE_INITSCRIPTS Jul 08 13:03:54 it's set in: Jul 08 13:04:02 distro/shr.conf Jul 08 13:08:59 JaMa, that's all I could find: Jul 08 13:09:04 image.bbclass:IMAGE_INITSCRIPTS ?= "initscripts" Jul 08 13:09:05 image.bbclass:IMAGE_BOOT ?= "${IMAGE_INITSCRIPTS} Jul 08 13:09:12 not in shr-image Jul 08 13:10:01 strange then Jul 08 13:10:04 GNUtoo|laptop: in shr.conf is enough, I meant only that it would be better in shr-image.inc because there is IMAGE_DEV_MANAGER and IMAGE_LOGIN_MANAGER defined Jul 08 13:10:17 I'll bitbake a new image Jul 08 13:10:24 GNUtoo|laptop: but it should work in both.. just run bitbake -e -b shr-image.bb a check the value Jul 08 13:10:42 ok Jul 08 13:10:46 GNUtoo|laptop: don't you have it in local.conf? Jul 08 13:11:03 no I've no initscript in local.conf Jul 08 13:11:24 but my images are old Jul 08 13:11:32 and updated with opkg update;opkg upgrade Jul 08 13:11:40 s/images/image Jul 08 13:11:45 I mean the one I've on device Jul 08 13:12:23 btw for release I've also another issue: Jul 08 13:12:35 images creations fails without this patch which was rejected in the past: Jul 08 13:13:38 http://pastebin.com/v6b8b4BV Jul 08 13:14:26 GNUtoo|laptop: do you have binary locales generation enabled? Jul 08 13:14:51 ENABLE_BINARY_LOCALE_GENERATION ?= 1 in shr.conf Jul 08 13:15:00 I think so Jul 08 13:15:08 check your local.conf Jul 08 13:15:36 it's not disabled Jul 08 13:15:40 so I bet it's enabled Jul 08 13:16:09 and I've ipk locales Jul 08 13:16:10 then you should have %s-binary-localedata-%s from eglibc build.. Jul 08 13:16:48 JaMa: sent cookie with SHR_SPLASH_THEME ;) Jul 08 13:17:05 but this patch whould break other images where binary-localedata are available Jul 08 13:17:44 eglibc-localedata-zu-za_2.12-r11.2+svnr10809.5_armv6-novfp.ipk Jul 08 13:17:44 mickey|office: ping Jul 08 13:17:47 that's what I have Jul 08 13:17:50 dos1: DEPENDS is buildtime dep, I think you don't need it Jul 08 13:18:15 GNUtoo|laptop: which means you have ENABLE_BINARY_LOCALE_GENERATION disabled somewhere (od did have it disabled when building eglibc) Jul 08 13:18:16 JaMa, what's the proper patch then? Jul 08 13:18:21 JaMa: last time when i was playing with it, RRECOMMENDS wasn't automatically built on buildhost Jul 08 13:18:33 JaMa, ah ok, it should be disabled because of armv6-novfp Jul 08 13:18:44 I heard it was broken on this arch Jul 08 13:18:53 but before I didn't understand that it was really disabled Jul 08 13:19:02 I thought it was enabled because I got some ipks Jul 08 13:19:20 JaMa: so package had recommendation, but it wasn't available in feed Jul 08 13:20:01 dos1: strange last time I tried it was build here Jul 08 13:20:19 dos1: but ok Jul 08 13:20:21 JaMa: maybe something in bitbake was changed, it was loooong time ago :) Jul 08 13:26:37 dos1: what'sup? Jul 08 13:26:42 JaMa, so I ifdef-like that part for when locales generation are disabled? Jul 08 13:28:06 mickey|office: if i want to fix UsbHost in fsodeviced to work with 2.6.32, should i take care about compatibility with 2.6.29 kernel? Jul 08 13:31:37 GNUtoo|laptop: yes, that's better Jul 08 13:31:49 JaMa, ok thanks Jul 08 13:32:16 dos1: what are the various SHR versions shipping these days? Jul 08 13:33:25 strange Jul 08 13:33:32 mickey|office: shr-testing with 2.6.29, but it has very old fsodeviced Jul 08 13:33:32 if not enabled: return Jul 08 13:33:47 I'll do bitbake -e Jul 08 13:33:50 mickey|office: shr-unstable with 2.6.32, which has not-too-old fsodeviced Jul 08 13:34:37 mickey|office: i think there is also shr-unstable.29, but that's rather saved only for people who can't for some reason move to .32 and i don't think it will be upgraded Jul 08 13:35:35 dos1: ok, then i'd not bother until someone explicitly requests it Jul 08 13:35:47 I'm still on .29 Jul 08 13:35:49 * JaMa doesn't plan to continue updating shr-unstable.29 :) Jul 08 13:36:36 * mrmoku thinks next shr-testing update will include .32 too Jul 08 13:37:13 next shr-testing (not relased, but built) has .29 + fsogsmd, but maybe we can skip this one :) Jul 08 13:37:20 Missing or unbuildable dependency chain was: ['fso-apm'] Jul 08 13:37:29 which kernel has debian? Jul 08 13:37:52 JaMa: well, it won't have newest fsodeviced anyway ;) Jul 08 13:38:12 ahhh Jul 08 13:38:13 ok Jul 08 13:38:16 wrong target Jul 08 13:38:20 dos1: yup I agree with you (wanting to push fsogsmd config change) Jul 08 13:38:47 dos1: but there is also problem with gta01 Jul 08 13:39:29 dos1: we still don't have .32 for gta01 and sysfs nodes will be probably different between gta01 and gta02 (so fso couldn't use shared openmoko_gta anymore) :/ Jul 08 13:40:12 JaMa: weren't they already different? Jul 08 13:40:36 JaMa: and, IIRC gta01 doesn't have powered usb host mode Jul 08 13:40:51 SHR root@gojama ~ $ ll /etc/freesmartphone/conf/ Jul 08 13:40:51 total 24 Jul 08 13:40:51 lrwxrwxrwx 1 root root 14 Jul 1 11:13 GTA01 -> ./openmoko_gta Jul 08 13:40:51 lrwxrwxrwx 1 root root 14 Jul 1 11:13 GTA02 -> ./openmoko_gta Jul 08 13:41:07 hmm, ok, i didn't think of 01 Jul 08 13:41:14 in that case it should be in the config Jul 08 13:41:15 JaMa: i mean sysfs nodes ;) Jul 08 13:41:16 and we need to split configs Jul 08 13:41:54 dos1: ie for gsm they were both neo1973-pm-gsm.0 Jul 08 13:42:08 JaMa: yup, but i mean only usb host :D Jul 08 13:42:23 JaMa: ah, now they are different? Jul 08 13:42:57 dos1: now there is gta02 for gta02 and who knows if gta01 even works with new kernel Jul 08 13:43:31 om-gta01-2.6.31 is last om-gta01 branch... Jul 08 13:44:43 so no idea if there are now gta02 sysfs nodes also on gta01... Jul 08 13:46:31 ENABLE_BINARY_LOCALE_GENERATION="0" Jul 08 13:46:32 ok Jul 08 13:54:23 dos1: you were partially right, shr-splash-theme-logo is built only with BB_DEFAULT_TASK = "buildall" Jul 08 14:38:27 JaMa: where does one have to set that? Jul 08 15:05:58 PaulFertser: thanks for that patch, looks pretty good on first glance. I'll apply that and do some tests. It won't fix the underlying issue, but will make the parser more robust Jul 08 15:06:35 to fix the actual issue, i have to change the way we send the extraordinairy command blocks Jul 08 15:14:00 mrmoku: BB_DEFAULT_TASK? I had it in my local.conf only Jul 08 15:14:23 JaMa: and where we should put that? Or how to tell people where to put it? Jul 08 15:15:27 mrmoku: maybe it's not needed (dos's DEPENDS is workaround),I don't know if buildall has some unwelcome side-effects.. Jul 08 15:15:41 ahh ok :) Jul 08 15:15:42 mickeyl: :) Jul 08 15:15:44 gena2x: ping Jul 08 15:31:21 freesmartphone.org: 03ospite 07cornucopia * r27d36a824c87 10/fsogsmd/src/lib/atcommandqueue.vala: (log message trimmed) Jul 08 15:31:21 freesmartphone.org: fsogsmd: fix handling URCs of the form PREFIX:SUFFIX Jul 08 15:31:21 freesmartphone.org: The current code assumes that the first character after the ':' Jul 08 15:31:21 freesmartphone.org: separator can be discarded because it is usually a whitespace, but some Jul 08 15:31:22 freesmartphone.org: modems can send URCs without a whitespace after the separator and this Jul 08 15:31:22 freesmartphone.org: would lead to discard some info from the actual SUFFIX. Jul 08 15:31:23 freesmartphone.org: Use .strip() in place of .offset( 1 ) to cleanup the SUFFIX. Jul 08 15:44:56 mrmoku: have you created a ticket for your gprs issue? Jul 08 15:57:52 PaulFertser: not yet... wanted to get a DEBUG fsogsmd log too... will do later Jul 08 15:57:56 dos1: p o n g. i am almost not a developer today... >30C here gone crazy. it's very hard with our humidity Jul 08 15:57:57 * mrmoku still dayworking :/ Jul 08 15:58:19 Weiss: yes, page mode. Jul 08 15:58:45 gena2x: i've had >30C here whole last week :P Jul 08 15:59:12 gena2x: today it's "only" 27 Jul 08 15:59:28 Weiss: try memwrite from http://www.bsdmn.com/openmoko/glamo/timings/ Jul 08 15:59:37 gena2x: well, with 465/116 suspend/resume works :) Jul 08 15:59:46 gena2x: but, with loglevel lower than 7 i can't boot at all from SD Jul 08 15:59:58 Weiss: params 1207959560 7119 Jul 08 16:00:18 Weiss: you'll see hexvalues dump then you'll run tool. Jul 08 16:00:22 gena2x: rootwait just hangs forever, and rootdelay=x just results in kernel panic delayed x seconds :P Jul 08 16:00:41 gena2x: with loglevel=7 and 8 works like a charm Jul 08 16:00:48 dos1: ah resumes? with my kernel? Jul 08 16:00:55 gena2x: no, SHR Jul 08 16:01:05 gena2x: now testing overclocked device in real life Jul 08 16:01:06 dos1: hm... interesting. Jul 08 16:01:22 dos1: nice notice about loglevel. Jul 08 16:01:48 probably some race condition - it's too fast to boot now :D Jul 08 16:02:14 dos1: as now i know how to setup glamo, i wanted today try faster timings. Jul 08 16:02:24 gena2x: ok Jul 08 16:02:42 dos1: may be 500/125. i recall it were not working because of glamo. Jul 08 16:03:23 gena2x: could you also build u-boot with 465/116 and 1.8 voltage? Jul 08 16:03:24 dos1: but now unsure. i'm almost melted. Jul 08 16:03:43