**** BEGIN LOGGING AT Thu Jul 15 02:59:56 2010 Jul 15 09:01:05 JaMa: what's all the java talk in #oe about? are you working on jalimo for SHR? Jul 15 09:01:32 zub: yup, building openjdk in feeds Jul 15 09:01:58 zub: we had cacao,jamvm before.. but now with all jalimo stuff in oe.dev I want to build openjdk too Jul 15 09:02:19 interesting Jul 15 09:02:23 good luck :) Jul 15 09:59:47 Ola. Jul 15 10:21:06 omg dos1 Jul 15 10:47:16 mrmoku: http://enivax.net/jk/kswapd.png swap-fragmentation Jul 15 10:48:05 X = writecount, Y = blocknumber written to Jul 15 10:48:51 causes N900 to slow down after some days uptime Jul 15 10:50:00 swapon tempswap; swapoff mainswap; swappon mainswap; swapoff tempswap Jul 15 10:50:08 will 'fix' it Jul 15 10:51:12 <[Rui]> DocScrutinizer: ouch! Jul 15 10:51:34 better you have two persistent swaps, so you can do "swapon swap2; swapoff swap1" Jul 15 10:52:43 <[Rui]> DocScrutinizer: yeah, but even better would be that something would deal with that silently under the hood Jul 15 10:52:46 [Rui]: the main problem is 256k blocksize penalty on MMC, when writing random Jul 15 10:52:57 <[Rui]> DocScrutinizer: ok... too bad :( Jul 15 10:53:41 so you *need* a way to 'defragment' the swap storage area Jul 15 10:54:30 no matter if you do 'on the fly' by moving around single blocks while swapperd is idle, or you trigger the huge defrag by swapoff Jul 15 10:54:52 DocScrutinizer: how did you get that log? Jul 15 10:55:20 shadowjk did it, a nice clever guy on #maemo Jul 15 10:55:43 kernel patching? Jul 15 10:56:29 http://mg.pov.lt/maemo-irclog/%23maemo.2010-07-15.log.html#t2010-07-15T13:07:55 Jul 15 11:00:51 DocScrutinizer: hmm, he doesn't mention how he got the data Jul 15 11:01:05 lindi-: I dunno Jul 15 11:01:54 he's after that for quite some weeks, so either he found a sysnode or debugfs node, or he patched swapperd Jul 15 11:07:50 DocScrutinizer: heh, fun Jul 15 11:08:02 yeah, actually Jul 15 11:10:44 DocScrutinizer: are you using swap file or partition? Jul 15 11:11:05 my use of swap file makes a simple block_dump impossible Jul 15 11:11:05 N800 has a 768M partition for swap Jul 15 11:11:24 N900* Jul 15 11:12:04 in future I'm quite sure it will have TWO :-P Jul 15 11:13:22 but hey, N900 has 32GB builtin eMMC, so what's wrong with two 0.7G swap partitions Jul 15 11:15:29 mickeyl: the official kernel now works with htcdream Jul 15 11:15:46 we can change the recipe for htcdream to standartkernel soon Jul 15 11:17:18 mickeyl: hi btw Jul 15 11:17:22 :-) Jul 15 11:26:39 mrmoku: JaMa: i have issue with reiserfs home partition Jul 15 11:26:52 after suspend it becames mounted read only Jul 15 13:25:44 freesmartphone.org: 03mlauer 07cornucopia * rf697473df5da 10/fsogsmd/src/plugins/modem_qualcomm_palm/channel.vala: fsogsmd: modem_qualcomm_palm: yielding the super-class is supposed to be fixed now Jul 15 13:31:09 hi. i'm trying to install navit. I added the feed to the repo and when doing opkg info it shows the two versions, but when installing it, it uses the "old" version from SHR (maybe because the "wrong" version number is higher). how can I force the other version ? opkg --help did not help, and the main unsolved thread I found is http://www.mail-archive.com/community@lists.openmoko.org/msg25606.html Jul 15 13:31:42 tsaitgaist: SHR version is not so "old" :) Jul 15 13:31:43 hi Jul 15 13:32:03 mickeyl, I just finished & installed fso2-demo-image for my htcleo Jul 15 13:32:06 tsaitgaist: please don't mix shr/navit feeds without _good_ reason Jul 15 13:32:20 using the force option can be used to install it when giving the URL, but the next update will revert this Jul 15 13:32:41 dcordes: wow, does anything work? Jul 15 13:33:04 JaMa: I wanted to test a newer version (more the 100 svn commit diff) Jul 15 13:33:18 mickeyl, not yet. I am a bit puzzled it doesn't have Xorg installed Jul 15 13:33:24 tsaitgaist: read "Navit packages" thread for newer SHR navit builds.. Jul 15 13:34:59 dcordes: I don't care too much about xorg yet Jul 15 13:35:10 mickeyl, ok I thought zhone2 is gui Jul 15 13:35:14 yes, it is Jul 15 13:35:15 based on X Jul 15 13:35:19 no Jul 15 13:35:28 it can be work on multiple display backends Jul 15 13:35:35 everywhere where EFL runs Jul 15 13:35:40 s/be// Jul 15 13:35:54 then it might be an EFL configuration issue Jul 15 13:36:38 btw I tried first zhone version in ubuntu maverick ARM Jul 15 13:36:57 with both, ubuntu fso and latest fso from git Jul 15 13:37:03 couldn't get to work either Jul 15 13:37:37 although frameworkd will report it can communicate with smd0, displays some processed network information and the unprocessed htc prop. signal strength Jul 15 13:37:47 zhone will say it can't access gsm Jul 15 13:37:59 don't use frameworkd Jul 15 13:38:09 it doesn't contain full htc support Jul 15 13:38:30 but since frameworkd can operate with qualcomm_msm I assumed zhone will understand it Jul 15 13:38:41 because it just uses the unified dbus stuff Jul 15 13:38:47 which I didn't understand Jul 15 13:39:24 JaMa : in the SHR ml ? I didn't found relevant info Jul 15 13:39:33 just wanted to let you know the qualcomm_msm is broken in ubuntu as I was quite excited that it exists Jul 15 13:39:55 and you also have competition. some ofono is available via apt Jul 15 13:39:57 well, i didn't work on frameworkd for years Jul 15 13:40:12 fsogsmd's support for qualcomm_htc is lightyears ahead Jul 15 13:40:26 ofono? heh, oh well Jul 15 13:40:50 that's the FOSS way Jul 15 13:40:58 thousands of solutions for one problem Jul 15 13:41:00 :D Jul 15 13:41:06 tsaitgaist: http://www.mail-archive.com/shr-user@lists.shr-project.org/msg04762.html Jul 15 13:41:44 mickeyl, I have seen some people using that ofono on htc rhodium in ubuntu and it seems tow work Jul 15 13:41:58 mickeyl, anyway. does that efl interface allow any configuration ? Jul 15 13:41:58 thx. I only looked in http://openmoko-public-mailinglists.1958.n2.nabble.com/ Jul 15 13:41:58 tsaitgaist: navit_0.1.99+svnr3497-r8.8.5_armv4t.ipk is there Jul 15 13:41:59 dcordes: sure, i don't doubt that Jul 15 13:42:44 dcordes: it it was using a more sane language and a better architecture, i would have stopped working on fsogsmd Jul 15 13:42:49 but as things stand, i can't :) Jul 15 13:43:13 ok keep it up. I'm looking forward to try it especially with the rmnet interface Jul 15 13:43:19 so I can setup 3g data easily Jul 15 13:43:34 yep. it runs great on the dream Jul 15 13:43:47 should give the same experience on yours Jul 15 13:43:57 mickey|: I cloned ofono today to check state of n900... strange code ;) Jul 15 13:45:19 mickeyl, hm yeah htcdream and htcleo have similar fb driver. htcleo has wvga resolution though Jul 15 13:45:20 mrmoku: :) lets see whether electranox delivers on adding a plugin for fsogsmd Jul 15 13:46:06 mickey|: if there won't hit something nice and new the hardware scene soon I'm seriously considering to get one :/ Jul 15 13:47:46 mickey|: and of course I want it to run FSO/SHR :-) Jul 15 13:47:49 mickeyl, it seems like it freezes during bootup. I will try the fso2-console-image first Jul 15 13:48:01 mrmoku: hmmm, why no palm pre or HTC Dream/Nexus one? Jul 15 13:48:05 mickeyl, might be related to 4.5 as I didn't try it before in any image Jul 15 13:48:19 gcc-4.5 that is Jul 15 13:48:34 mickey|: I need something more terminal like... to use as mobile computer Jul 15 13:48:38 the bigger the screen the better Jul 15 13:48:50 mrmoku: i see, that leaves only the pre Jul 15 13:48:51 err Jul 15 13:48:51 dcordes: do you have patched kernel for 4.5? Jul 15 13:48:52 dream Jul 15 13:48:53 (resolution wise of course... having a good pair of glasses :P) Jul 15 13:49:03 JaMa, I don't know. I run 2.6.32 Jul 15 13:49:07 mrmoku: hmm, ok, if you get one, i might be interested in obtaining one as well Jul 15 13:49:18 needs to have a good price though Jul 15 13:49:23 dcordes: it was merged in 2.6.35-rc3... Jul 15 13:49:47 dcordes: see OE patches for needed patch and add it to dream kernel Jul 15 13:49:47 JaMa, you have a link to the patch by any chance ? Jul 15 13:49:53 ok nice Jul 15 13:50:10 dcordes: http://gitorious.org/~jama/htc-msm-2-6-32/openmoko-kernel/commit/2938af9681c96b5b1332982d534e719a960dbff3 Jul 15 13:50:13 mickey|: http://www.heise.de/preisvergleich/a457621.html still 400€ :/ Jul 15 13:50:35 darn, that's alot Jul 15 13:50:39 JaMa, ok doesn't seem too heavy Jul 15 13:50:42 yup Jul 15 13:55:17 mrmoku, palm pr´e Jul 15 13:55:22 ´e Jul 15 13:55:26 damn Jul 15 13:56:24 playya_: ? Jul 15 13:56:36 buy a palm pre Jul 15 13:56:53 resolution too low :/ Jul 15 13:57:05 better buy a desire Jul 15 13:57:24 3rd qsd8250 device in oe Jul 15 13:58:04 JaMa, do you know what kind of problems are expected without the patch ? Jul 15 13:58:14 dcordes: missing kbd :/ Jul 15 13:58:21 * mrmoku wants a hw kbd this time Jul 15 13:58:35 that's why i loved the raphael Jul 15 13:58:35 dcordes: yes.. hanging boot Jul 15 13:58:43 dream's is much worse Jul 15 13:59:08 but the raphael's kernel went the wrong direction Jul 15 13:59:14 dcordes: I can find my old conversation with khem on #oe for more details.. but this patch is definetly needed so try it first Jul 15 13:59:18 (warning, subjective, as always...) Jul 15 13:59:36 mickeyl, yes what we talked about last time I was on here Jul 15 13:59:50 mickeyl, with the androids taking over git. no good. Jul 15 14:00:12 mickeyl, but I'm still looking forward to the mainline msm/qsd . they recently added initial fb support Jul 15 14:01:09 JaMa, khem is the fastest oe low level expert Jul 15 14:01:24 nice. if it wasn't to take a year or more i'd be enlightened ;) Jul 15 14:01:24 he helped me with many thing Jul 15 14:02:17 JaMa, kernel almost done Jul 15 14:05:39 the touch pro2 is 383€ too... that's almost a N900 then :/ Jul 15 14:07:27 can somebody point me to the touchscreen driver gta02 ueses ? (maybe some git link?) Jul 15 14:07:37 s3c24xx-ts or so Jul 15 14:07:37 mrmoku, rhodium ? Jul 15 14:07:47 don't recommend WinCE devices Jul 15 14:08:04 dcordes: be aware that driver changed from om specific in 2.6.32 to upstream version in 2.6.34 Jul 15 14:08:05 they still have a high probability that some peripherals will never be supported Jul 15 14:08:43 JaMa, can you point me to the 32 one ? Jul 15 14:08:52 hi guys ! hey what about Samsung Wave phone ? could it be a target for FSO ? :) Jul 15 14:09:06 I just need ts driver that works with tslib/Xorg (out of the box) in recent kernel Jul 15 14:09:10 I mean is it "opened enought "? Jul 15 14:09:13 so I have example how to do it Jul 15 14:09:33 dcordes: http://git.openmoko.org/?p=kernel.git;a=commit;h=7f90a049d8ddcb6c485e2a6889cb3b5d4a5be1c0 Jul 15 14:10:23 GarthPS: unfortunately this remains unclear as long as someone runs linux on it w/ some peripherals supported Jul 15 14:10:34 JaMa: thats the old driver Jul 15 14:11:00 mickey|: as usual ... :) Jul 15 14:11:07 yep Jul 15 14:11:13 larsc: he wanted old driver.. Jul 15 14:11:17 ok Jul 15 14:12:44 freesmartphone.org: 03mlauer 07cornucopia * r4cd9ca1d3f54 10/libfsotransport/fsotransport/commandqueue.vala: libfsotransport: retry -> int Jul 15 14:14:37 JaMa, thanks Jul 15 14:15:10 dcordes: (rhodium) is that the o2 branded thing? Jul 15 14:15:44 JaMa, it works. I can now login Jul 15 14:16:14 great Jul 15 14:16:41 freesmartphone.org: 03mlauer 07cornucopia * red93e714f0d4 10/libfsotransport/fsotransport/commandqueue.vala: libfsotransport: remove default retry constant Jul 15 14:16:42 freesmartphone.org: 03mlauer 07cornucopia * re7e64326f070 10/fsogsmd/src/lib/atcommandqueue.vala: fsogsmd: prepare for reintroducing command timeouts Jul 15 14:17:20 mickeyl, I still get a problem when starting the zhone2 init script Jul 15 14:17:42 mickeyl, screen looks garbled Jul 15 14:18:05 yea, the fb config is probably different to the dream Jul 15 14:18:07 mickeyl, it only gets to draw a garbeld bar with blue zhone background pattern on top of the screen Jul 15 14:18:15 mickeyl, where is it located ? Jul 15 14:18:28 just remove the echo from the init script Jul 15 14:18:49 mickeyl, ok Jul 15 14:19:06 iirc it pipes a picture into the fb first Jul 15 14:19:51 ok not exciting it looks wrong then if it is just being dumped in the fb Jul 15 14:19:55 with wrong config Jul 15 14:20:16 yeah, it's a raw image Jul 15 14:22:40 it's so nice to debug with full qwerty directly attached to device Jul 15 14:24:47 mickeyl, ok still freezes Jul 15 14:25:02 mickeyl, is there any extra configuration ? Jul 15 14:25:37 mickeyl, it might be gcc 4.5 problem affecting fb in some way Jul 15 14:26:16 mickeyl, let's see if I can do stuff in fso cli Jul 15 14:26:52 mickeyl, any fso2 Xorg based gui available ? Jul 15 14:28:23 I see some fsogsmd configurations for htcdream. how can I make fsogsmd use them ? Jul 15 14:30:28 two ways Jul 15 14:30:32 1.) manual Jul 15 14:31:03 copy /etc/freesmartphone/conf/htc_qualcomm_dream/fsogsmd.conf to /etc/ Jul 15 14:31:05 2.) automatic Jul 15 14:31:57 add proper symlinks so that board name (/proc/cpuinfo Hardware field minus spaces) points to the right conf directory Jul 15 14:34:30 ok Jul 15 14:36:01 dcordes: I've got the official kernel running here Jul 15 14:36:04 on htcdream Jul 15 14:36:53 leviathan, what's wrong with the patched (leviathan etc) kernels ? Jul 15 14:37:03 hmm Jul 15 14:37:22 I mean I've got the linus linux running on dream ;-) Jul 15 14:37:35 vanilliest vanilla Jul 15 14:37:37 with all hardware supported ? Jul 15 14:37:38 if you like Jul 15 14:37:48 misc: nope Jul 15 14:37:55 only framebuffer Jul 15 14:37:57 ^^ Jul 15 14:40:26 but its a beginning Jul 15 14:41:29 ahh, and sd card also works Jul 15 14:45:48 i wonder how far you can get with that Jul 15 14:46:18 i'll be impressed once you have reliable modem working Jul 15 14:46:23 leviathan, hmmm nice ! Jul 15 14:46:28 since that involves nasty wakelocks and BP communication Jul 15 14:46:36 leviathan, I have seen the fb patch Jul 15 14:47:57 mickey|: due to your last changes we are calling SetFunctionality too early... while modem is still initting Jul 15 14:48:19 mickey|: check for that error and retry later? or listen to some signal? Jul 15 14:49:37 mickey|: like device-no-sim ? Jul 15 14:49:44 err alive-no sim Jul 15 14:50:12 mrmoku: either check and retry or wait for modem_state to become one of the alive states Jul 15 14:50:23 mickeyl: modem is no big problem, while connected to internal uart Jul 15 14:50:25 :-) Jul 15 14:50:34 mickeyl, it seems like I have a problem running the fso2 binaries Jul 15 14:50:40 mickey|: ok Jul 15 14:50:59 mickeyl: I'm compiling fso2-demo-image Jul 15 14:51:09 mickeyl: I'll try it with the kernel Jul 15 14:51:11 mrmoku: i'm still pondering whether we can delay the RequestResource even longer to wait until the modem is actually ready to accept all kinds of commands; but due to the default dbus timeouts it probably is not a good idea Jul 15 14:51:56 dcordes: ok... ? Jul 15 14:52:53 mrmoku: i will probably rather document the fact that there are subsystem specific states of "readyness" and that clients should be aware of dealing with command errors Jul 15 14:53:03 mickeyl, have you tried to compile them with gcc 4.5 ? Jul 15 14:53:17 dcordes: no, still using 4.3 and 4.4 here Jul 15 14:54:03 mickeyl, is fsogsmd --help supposed to print stuff ? Jul 15 14:56:53 no, the fso2 daemons have no CLI Jul 15 14:57:06 they're configured through configuration files Jul 15 14:57:23 ah ok Jul 15 14:57:40 you will probabyl want to edit logging as well Jul 15 14:57:50 e.g. start with setting LOG_LEVEL = DEBUG Jul 15 14:57:54 just looked at zhone2 output Jul 15 14:58:01 and log_to = stderr:colors Jul 15 14:58:03 everything seems good Jul 15 14:58:06 ok Jul 15 14:58:06 s/LOG_LEVEL/log_level/ Jul 15 14:58:08 no errors are visible Jul 15 14:58:19 it freezes at the following line: Jul 15 14:58:20 mickey|: will waiting for alive-sim-locked, SendAuthCode and calling SetFunctionality without PIN on alive-sim-ready work? Jul 15 14:58:49 mickeyl, [INFO] zhone2: Using EFL engine fb w/ screen size 480 * 800 Jul 15 14:59:30 mrmoku: yes, that should work (if it doesn't i need to fix it) Jul 15 14:59:34 dcordes: good Jul 15 14:59:40 mickeyl: I look forward to see the patches floating into 2.6.36 Jul 15 14:59:53 so that we can get our alsa and fb patches there too Jul 15 15:00:05 mickeyl, it doesn't draw anything to the screen, cursor freezes Jul 15 15:00:23 mickey|: ok, will do it that way then, thanks :) Jul 15 15:00:51 dcordes: have a look at the init script again and try adusting the touchscreen node Jul 15 15:01:13 mickeyl, I run it without the init script Jul 15 15:01:50 JaMa, did you try zhone2 with gcc 4.5 ? Jul 15 15:01:58 dcordes: no Jul 15 15:02:39 dcordes: try strace Jul 15 15:02:45 are you sure it can open the touchscreen? Jul 15 15:03:02 mickeyl, ok Jul 15 15:03:07 mickeyl, no, but why would ts kill it ? Jul 15 15:03:17 dcordes: is your Xorg exiting with zhone2? Jul 15 15:03:35 JaMa, no zhone2 doesn't use Xorg Jul 15 15:03:47 dcordes: iirc zhone1 called exit or something in Xsession script, which is not allowed since we moved from xserver-kdrive-common to xserver-common Jul 15 15:05:34 dcordes: dunno, i have seen that in EFL/fb before Jul 15 15:07:55 mickeyl, is there any other EFL program I could test ? Jul 15 15:08:53 mouse opening is right after fb open, so i don#t think you'll find anything else Jul 15 15:09:07 try ts_calibrate and ts_draw Jul 15 15:09:37 or ts_test Jul 15 15:11:34 mickeyl, ok I will set right ts device Jul 15 15:13:07 hm with the aim to get additional ttys I added this to /etc/inittab Jul 15 15:13:14 2:23:respawn:/sbin/getty 38400 tty2 Jul 15 15:13:15 3:23:respawn:/sbin/getty 38400 tty3 Jul 15 15:13:15 4:23:respawn:/sbin/getty 38400 tty4 Jul 15 15:13:26 but still only 1 is spawned Jul 15 15:13:28 what's wrong ? Jul 15 15:14:10 are you sure that your kernel supports virtual terminals? Jul 15 15:14:19 could be the problem for EFL lock as well btw. Jul 15 15:14:27 since it tries to switch to another VT on startup Jul 15 15:14:43 � Jul 15 15:14:45 whoops Jul 15 15:16:52 mickeyl, same freeze with correct TSLIB_TSDEVICE passed Jul 15 15:17:02 mickeyl, what is the CONFIG_ for vt ? Jul 15 15:17:14 mickeyl, oh well.. it works in ubuntu with same kernel Jul 15 15:22:41 CONFIG_VT or so Jul 15 15:22:52 what does strace say btw? Jul 15 15:22:58 dcordes: I'm try'n to fix the lsof issue Jul 15 15:23:34 mickeyl, I will try witjh strace Jul 15 15:25:13 freesmartphone.org: 03mickey 07cornucopia * rcf1c2d369f8d 10/libfsotransport/fsotransport/commandqueue.vala: libfsotransport: support command timeouts Jul 15 15:26:30 freesmartphone.org: 03mickey 07cornucopia * r11ce13e9cb65 10/fsogsmd/src/ (2 files in 2 dirs): fsogsmd: support command timeouts Jul 15 15:29:55 freesmartphone.org: 03mickey 07cornucopia * rbb71f79a446a 10/libfsotransport/fsotransport/commandqueue.vala: libfsotransport: allow clearing the timeout for the current command Jul 15 15:29:57 freesmartphone.org: 03mickey 07cornucopia * r56c149057b3b 10/fsogsmd/src/lib/atcommandqueue.vala: fsogsmd: clear timeout if solicited result code comes in in-time Jul 15 15:39:38 ~seen gnutoo Jul 15 15:39:41 gnutoo <~GNUtoo@host12-84-dynamic.51-79-r.retail.telecomitalia.it> was last seen on IRC in channel #openmoko-cdevel, 7d 17h 1m 8s ago, saying: 'there is a free tool to interact with their bootloader...'. Jul 15 15:47:29 mickeyl, I will further examine it later Jul 15 15:47:33 mickeyl, time to go to work Jul 15 15:47:35 ttyl Jul 15 15:49:09 k, cu Jul 15 15:49:50 freesmartphone.org: 03mickey 07cornucopia * r7107e46e0a05 10/fsogsmd/src/lib/atcall.vala: fsogsmd: atcall: don't reenter call synchronization method Jul 15 16:42:06 mickeyl: hey :) the commit messages look quite promising! ;) Jul 15 17:12:52 mickeyl: hi Jul 15 17:13:03 mickeyl: do you know where gnutoo is? Jul 15 18:12:47 GNUtoo|laptop: ping Jul 15 18:13:39 <[Rui]> hi Jul 15 18:13:51 PaulFertser: yeah, i realized it doesn't work completely without timeouts Jul 15 18:17:17 mickeyl: i think i saw an answer for one command was treated as an answer for an earlier command in one of my logs (due to them being on different channels). May i ask what do you think about that? Jul 15 18:18:42 hi mickeyl Jul 15 18:20:09 mickeyl: and i want to highlight it once again: you really move the project forward, and that's just great. Jul 15 18:21:24 <[Rui]> This doesn't look good, mickeyl ! 2010-07-15T18:20:59.929539Z [INFO] libfsotransport <0710:2>: SRC: "+COPS=3,1;+COPS?" -> [ "+CME ERROR: 3" ] Jul 15 18:24:04 PaulFertser: hmm, i'd need to see that log Jul 15 18:24:09 GNUtoo|laptop: did my patch work? Jul 15 18:24:27 mickeyl, I can't build because of this introspection issue Jul 15 18:24:36 should be fixed meanwhile Jul 15 18:24:51 ah I'll retry then Jul 15 18:25:49 [Rui]: 3 means you're not allowed to do that. SIM ok and booked to the network? Jul 15 18:26:12 <[Rui]> mickeyl: I inserted my pin code Jul 15 18:26:43 <[Rui]> and my libphoneui.conf somehow got "Switchitch" when it should have "Switch" at the alsa definitons Jul 15 18:27:21 <[Rui]> 2010-07-15T18:19:21.230598Z [INFO] libfsotransport <0710:2>: SRC: "+CPIN?" -> [ "+CPIN: READY", "OK" ] Jul 15 18:28:17 PaulFertser: i'd be very surprised though, since the parsers do not share state (we would get total chaos if they did) Jul 15 18:28:35 [Rui]: ok so far, but are you registered? Jul 15 18:29:10 <[Rui]> mickeyl: doesn't look like it (at least the gsm signal gadget is only a star and thos CME ERROR keep coming up Jul 15 18:29:17 [Rui]: oh, and... is the functionality set to "full" ? Jul 15 18:29:29 i have seen CME 003 in that case when +CFUN = 0 Jul 15 18:29:43 <[Rui]> waiting for another reboot to finish... Jul 15 18:29:57 ok, it's probably that phonefsod got confused Jul 15 18:30:15 * mickeyl puts the blame on other software ;) Jul 15 18:32:02 <[Rui]> mickeyl: what should I look at? Jul 15 18:32:30 <[Rui]> mickeyl: I've already found out that gps is being looked at at the wrong path Jul 15 18:33:02 hmm, i didn't follow all the kernel sysfs path shuffling around Jul 15 18:33:09 <[Rui]> I don't see no cfuns Jul 15 18:33:11 mickeyl: the same log as "clcc gone wild" Jul 15 18:33:49 [Rui]: if there's no CFUN, then for some reason SetFunctionality doesn't get called. I'm not at all familiar with phonefsod, so I'm afraid I have no idea where to look at Jul 15 18:34:47 PaulFertser: right. i'll have a look at that Jul 15 18:35:36 <[Rui]> mickeyl: damn :| Jul 15 18:36:12 <[Rui]> ok guys, who broke it? :) Jul 15 18:36:36 PaulFertser: i wonder if you could create such a log w/ "ultra debug" Jul 15 18:36:55 mickeyl: thnx :) sorry for (somewhat) obscure method of bugreporting. But at least i do supply (somewhat) meaningful logs ;) Jul 15 18:37:38 <[Rui]> 2010-07-15T18:37:13.991787Z [ERROR] UsageController <9 R>: Resource GSM can't be enabled: Can't open the modem.. Trying to disable instead Jul 15 18:39:20 wah, that clcc gone wild is crazy :) Jul 15 18:39:26 mickeyl: that clcc problem is unfortunately very hardly reproducable :/ Jul 15 18:39:54 i think 7107e46e0a05c99109ac41c782ed3674440ddf71 will keep this under control Jul 15 18:40:10 together with the timeout we will see what happens next time +CLCC will not answer in time Jul 15 18:40:22 mickeyl: basically that's what i was suggesting in the ticket :) Jul 15 18:40:44 eventually -- on the day i'm in a very good mood -- i will add a custom Calypso call handler based on %CPI support Jul 15 18:41:33 <[Rui]> the phone was working this afternoon Jul 15 18:41:39 <[Rui]> wtf happened? :| Jul 15 18:41:46 [Rui]: battery too empty? Jul 15 18:41:54 <[Rui]> mickeyl: nopes Jul 15 18:41:54 calypso will emit all kind of strange behaviour in that case Jul 15 18:42:07 <[Rui]> under an hour ago I even did a couple of calls Jul 15 18:42:08 mickeyl: so is it time i git pull? Jul 15 18:42:36 <[Rui]> I only powered the phone off, switched the sim card for another (which doesn't require sim code), noticed it didn't come up. Jul 15 18:42:37 PaulFertser: yeah. the timeout is set to a ridiculously high value atm. (2 minutes), but it's good for a start Jul 15 18:42:47 [Rui]: try reseating Jul 15 18:42:50 <[Rui]> swapped the sim back to mine, nothing... reboot, nothing Jul 15 18:42:53 hmm Jul 15 18:43:00 cleaning contacts a bit? Jul 15 18:43:02 checking card holder Jul 15 18:43:11 that thing is a bit fragile Jul 15 18:43:16 <[Rui]> meanwhile; i've been trying a mix of restarting fso services etc... Jul 15 18:43:46 <[Rui]> mickeyl: if it were that, would it negotiate pin as well as it appears to be doing so? Jul 15 18:43:51 mickeyl: there should be no need for a custom calypso call handler. CLCC should just work, and to the best of my understanding it doesn't confuse the modem. It's another question why and how an unanswered clcc is related to fsogsmd stopping answering on dbus calls. Jul 15 18:44:44 [Rui]: i'm not sure whether there can be "partial failure"... Jul 15 18:44:54 <[Rui]> damn Jul 15 18:45:08 PaulFertser: agreed Jul 15 18:45:25 <[Rui]> 2010-07-15T18:45:10.911675Z [ERROR] libfsotransport <0710:2>: Can't allocate channel #2 from MUX: Channel 2 is already taken. Jul 15 18:45:31 <[Rui]> does this ring any bells? Jul 15 18:45:47 sounds like you have multiple deamons running atm. Jul 15 18:46:10 try resetting the modem before restarting the services Jul 15 18:46:29 <[Rui]> atz ? Jul 15 18:46:43 no, with the sysfs dance Jul 15 18:47:13 although fsogsmd is supposed to do that automagically Jul 15 18:47:19 when it tries to request the resource Jul 15 18:47:30 you don't have any experimental kernels or so? Jul 15 18:50:14 <[Rui]> is it ./gta02-pm-bt.0/reset ? Jul 15 18:50:25 bt stands for bluetooth, so I'd say no Jul 15 18:50:29 gsm would be it Jul 15 18:50:30 <[Rui]> mickeyl: stock shr-u Jul 15 18:50:40 <[Rui]> ls | grep gsm == nothing! Jul 15 18:50:45 uh oh Jul 15 18:50:56 that's scary Jul 15 18:51:00 <[Rui]> ok it's hidden underneath Jul 15 18:51:07 <[Rui]> root@om-gta02 /sys/devices/platform # find . -name \*gsm\* Jul 15 18:51:07 <[Rui]> ./s3c2440-i2c/i2c-0/0-0073/pcf50633-gpio/reg-fixed-voltage.1/regulator/regulator.11/gta02-pm-gsm.0-GSM Jul 15 18:51:07 <[Rui]> ./s3c2440-i2c/i2c-0/0-0073/pcf50633-gpio/reg-fixed-voltage.1/gta02-pm-gsm.0 Jul 15 18:51:23 hmm Jul 15 18:51:27 that looks all new to me Jul 15 18:51:33 i never tested FSO on such a recent kernel Jul 15 18:52:11 power_node = /sys/bus/platform/devices/gta02-pm-gsm.0/power_on Jul 15 18:52:11 <[Rui]> yeah, it's weird Jul 15 18:52:20 <[Rui]> should be that Jul 15 18:53:12 <[Rui]> mrmoku: but it appears to be a link to the other Jul 15 18:53:22 <[Rui]> yes, it is a link Jul 15 18:53:39 yup Jul 15 18:55:00 [Rui]: poweroff, rip out the bat and retry after a moment? Jul 15 18:55:31 <[Rui]> 2010.07.15 19:55:13.912642 [phonefsod] MESSAGE: Error: Message did not receive a reply (timeout by message bus) Jul 15 18:56:03 [Rui]: does mdbus2 -s org.freesmartphone.ogsmd hang? Jul 15 18:58:10 <[Rui]> no it doesn't. I'm going to do yet another reboot. Jul 15 18:58:33 <[Rui]> taking out all power sources as I go get a glass of water Jul 15 19:01:46 hmm Jul 15 19:02:01 i guess i should grab http://cgit.freedesktop.org/dbus/dbus/commit/?id=6ff1d079316cb730a54b4e0e95bd3e6e31f439de on top of our 1.3.1 recipe Jul 15 19:02:52 <[Rui]> nice, the sim auth dialog just died as I unlocked the screen Jul 15 19:04:40 dbus folks finally started to try to solve the stability issues? Hm, kinda promising. Jul 15 19:05:43 "this isn't right either, but at least is correct" Jul 15 19:05:44 nice :P Jul 15 19:05:56 ironically rather stimulated by embedded folks Jul 15 19:09:30 pushed (via OE) Jul 15 19:09:30 bbl Jul 15 19:10:05 <[Rui]> *sigh* WTF is going on... Jul 15 19:10:22 * [Rui] feels like cutting his wrists (not really, seriously guys!) Jul 15 19:16:22 [Rui]: Hey, you know it makes sense to activate DEBUG for fsogsmd and to pastebin the log after restarting fsogsmd. Why not simply do that? Jul 15 19:23:29 <[Rui]> 2010-07-15T19:22:55.281414Z [INFO] libfsotransport <0710:2>: SRC: "+COPS=3,0;+COPS?" -> [ "+CME ERROR: 3" ] Jul 15 19:25:15 [Rui]: sure it's not a bad sim connection? Jul 15 19:25:27 <[Rui]> PaulFertser: seems quite well Jul 15 19:26:10 mickeyl, http://pastebin.com/1K5W3uCx Jul 15 19:27:19 <[Rui]> 2010-07-15T19:27:05.256669Z [DEBUG] TiCalypsoModem <4C>: Did receive a valid response for TiCalypsoPercentCSTAT Jul 15 19:27:50 <[Rui]> but before that, 2010-07-15T19:26:54.890942Z [INFO] libfsotransport <0710:2>: SRC: "+CPIN?" -> [ "+CPIN: READY", "OK" ] Jul 15 19:30:12 mickeyl: 2010-07-15T19:29:17.333498Z [DEBUG] TiCalypsoModem <4C>: Parsing error: '%CSQ: 22, 99, 2' does not match '%CSQ: (?P\d+), (?:\d+), \d' Jul 15 19:30:27 is fixing that as easy as removing that space from the regex? Jul 15 19:30:55 <[Rui]> I'm going to try yesterday's image Jul 15 19:31:09 <[Rui]> will just untar it above current fs Jul 15 19:35:41 PaulFertser: already installed current fsogsmd? Jul 15 19:36:17 mrmoku: not yet Jul 15 19:36:22 mrmoku: compiling is slow Jul 15 19:36:27 PaulFertser: :) Jul 15 19:36:33 PaulFertser: we have problems with our rev :/ Jul 15 19:36:50 f248cb38db9f531393391984284de96e8ff4e264 Jul 15 19:36:55 <[Rui]> booting... Jul 15 19:37:25 mickeyl, could you fix it? Jul 15 19:40:16 mrmoku: what kind of problems? Jul 15 19:40:54 <[Rui]> if it's software, yesterday's image *also* has it Jul 15 19:41:15 mickeyl, I'll go, bb tomorrow Jul 15 19:41:16 <[Rui]> I'm going to try the one from 3 days ago Jul 15 19:42:14 PaulFertser: it looses registration Jul 15 19:42:26 PaulFertser: shows signal strength but lost the provider Jul 15 19:42:41 <[Rui]> with all this, it's 20:38, wife's not home yet and I can't call her to know whether all's fine Jul 15 19:44:08 mrmoku: how exactly? Jul 15 19:44:37 PaulFertser: just rebooted... will paste the log in a moment Jul 15 19:48:48 <[Rui]> ok, I've got GSM with 20100712 image, but not after it Jul 15 19:48:54 <[Rui]> I won't even do an upgrade Jul 15 19:48:58 <[Rui]> it's a software issue Jul 15 19:50:41 * von_fritz confirms :) Jul 15 19:50:48 * mrmoku too Jul 15 19:51:08 PaulFertser: did not even register now Jul 15 19:51:17 PaulFertser: and a manual SetFunctionality is hanging Jul 15 19:59:49 mrmoku: well, i was using a similar revision for a week. Jul 15 20:00:09 PaulFertser: http://pastie.org/1046184 Jul 15 20:00:44 brb Jul 15 20:02:14 mrmoku: strange. Anyway, i have to sleep now, sorry. Jul 15 20:06:09 here is mine : http://shr.pastebin.com/LeUHUC42 Jul 15 20:13:30 von_fritz: I fixed phonefsod to not call SetFunctionality too early Jul 15 20:15:18 hi Jul 15 20:22:54 shr-image and shr-image-lite are broken in org.openembedded.dev Jul 15 20:23:10 dcordes: in how far? Jul 15 20:23:28 matchbox-panel-2 do_fetch problem Jul 15 20:23:46 ibphone-ui some problem Jul 15 20:23:54 libphone-ui Jul 15 20:24:14 dcordes: we don't have matchbox-panel in our images... Jul 15 20:25:05 mrmoku, ok. when I build shr-image or shr-lite-image I have it in the dependency tree though. would you like to look at the dot files ? Jul 15 20:27:22 we have it in depends.. but do_fetch problem is only in your downloads dir.. Jul 15 20:28:00 * JaMa rebuild both images from scratch yesterday Jul 15 20:28:05 so shr+gta02 works ok Jul 15 20:28:39 JaMa, ok I will remove the matchbox-panel-2 from my sources directory, clean and rebuild the package Jul 15 20:32:05 JaMa, it is some net or other problem on my host side then. still doesn't work Jul 15 20:32:14 http://pastebin.ca/1901294 Jul 15 20:33:01 # Jul 15 20:33:01 svn: Failed to add directory 'applets': an unversioned directory of the same name already exists Jul 15 20:33:15