**** BEGIN LOGGING AT Sun Jan 16 02:59:58 2011 Jan 16 05:32:42 hi, Jan 16 05:32:46 I'm trying and trying Jan 16 05:32:53 and I can't enable wireless manually Jan 16 05:33:20 uname -a says: Jan 16 05:33:21 Linux om-gta02 2.6.34.8 #1 Wed Jan 12 08:00:21 CET 2011 armv4tl GNU/Linux Jan 16 05:33:32 what's the correct command? Jan 16 05:33:54 echo gta02-pm-wlan > /sys/bus/platform/drivers/gta02-pm-wlan/bind Jan 16 05:33:57 doesn't work Jan 16 05:35:57 ahh ok Jan 16 05:36:00 bad update Jan 16 05:36:09 which caused wrong kernel modules Jan 16 05:36:25 serial even if painfull(only one hand to type) is handy Jan 16 05:45:16 hi Lopi Jan 16 10:22:32 <[Rui]> moanin' Jan 16 11:16:52 mrmoku, hi Jan 16 11:17:08 I think it's the lack of bq27xxx that makes it segfault Jan 16 11:17:26 ok Jan 16 11:17:34 I'm recompiling an image Jan 16 11:21:36 GNUtoo|laptop: didn't we already have alsa scenarios for n900? Jan 16 11:21:46 we had Jan 16 11:21:54 we lost them somewhere Jan 16 11:22:01 heh, ok Jan 16 11:22:13 we also had keyboard illumination on htcdream, was lost too Jan 16 11:22:21 they have not been committed to fsodeviced then? Jan 16 11:22:26 or have they been and got lost then Jan 16 11:22:36 * mrmoku checks git log Jan 16 11:22:40 I think it was commited Jan 16 11:23:12 hmm... git log does say no Jan 16 11:23:26 ah ok Jan 16 11:23:33 how do we handle multiple kernels? Jan 16 11:23:54 I mean 2.6.28 and 2.6.3x have different scenario needs Jan 16 11:25:29 GNUtoo|laptop: we have the same thing for gta02 Jan 16 11:25:40 ok Jan 16 11:27:52 btw. are they actually different? Jan 16 11:29:12 freesmartphone.org: 03mok 07cornucopia * rf04c2d07cb88 10/fsodeviced/conf/nokia_n900/ (alsa-default/headset alsa-default/stereoout alsa.conf): Jan 16 11:29:12 freesmartphone.org: fsodeviced: add headset and stereout scenarii for n900 Jan 16 11:29:12 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 11:29:17 alsamixer output is Jan 16 11:31:24 mickeyl: hmm... fsogsmd is segfaulting on n900 Jan 16 11:31:25 1970-01-01T01:11:56.809631Z [CRITICAL] fsogsmd : GLib <>: string_to_string: assertion `self != NULL' failed Jan 16 11:31:40 first I thought it would be my fault in the lowlevel plugin Jan 16 11:31:45 but same with it disabled Jan 16 11:34:12 mickeyl: it does not segfault when disabling modem_isi Jan 16 11:38:54 gdb? Jan 16 11:40:25 will try Jan 16 11:40:57 might as well be my lowlevel plugin which stops segfault when nobody is using it .P Jan 16 11:41:20 hello there! About my vibrator problems, the guy who owned the neo before I do said that it stopped working after a #1024 hw fix. Is it possible that there is a relation ? (The C1009 capacitor has been changed.) Jan 16 11:43:47 well, if the fix is wrong ... Jan 16 11:45:58 is there a way to get the capacitor's capacity (in uF) using a multimeter ? Jan 16 11:52:32 paulk: in case you multimeter supports that, yes. Jan 16 11:52:41 paulk: but that cap must be desoldered. Jan 16 11:52:52 not a good idea ^^ Jan 16 11:53:08 paulk: 1024 fix should be unrelated. Jan 16 11:53:25 paulk: but some of the vibrator-related components might have been damaged. Jan 16 11:53:32 paulk: check the transistor etc. Jan 16 11:54:09 I mad some tests : when I echo 0 > /sys/class/*/brightness, I have 1.1V, when I echo 65 > /sys/class/*/brightness, I have 0.6V, and when I echo 120 > /sys/class/*/brightness I have 0.3V Jan 16 11:55:03 paulk: 4-0.3 is 3.7, so that should be enough for the DC motor to start spinning. Jan 16 11:56:12 why 4-0.3 ? I just have 0.3 on the vibrator connectors Jan 16 11:57:43 paulk: ah, i thought you've measuring voltage to the GND. Jan 16 11:58:18 nope Jan 16 12:02:38 + to GND is always 2.26V Jan 16 12:12:54 GNUtoo|laptop: hmm... actually it does not segfault, but abort Jan 16 12:13:00 ah ok Jan 16 12:13:15 I'll look for fsodeviced Jan 16 12:13:21 how to enable debug output? Jan 16 12:13:30 GNUtoo|laptop: talking about fsogsmd Jan 16 12:13:33 isn't there a -d 1 Jan 16 12:13:35 I know Jan 16 12:13:36 ahh Jan 16 12:13:37 ok Jan 16 12:13:38 :) Jan 16 12:13:43 but maybe both have similar issues Jan 16 12:13:49 both can be debuged the same way Jan 16 12:13:51 etc... Jan 16 12:13:54 just set loglevel = DEBUG in conf? Jan 16 12:14:45 well... it is DEBUG by default for n900 Jan 16 12:15:03 and I don't know of an option to make it not daemonize Jan 16 12:15:22 ah ok Jan 16 12:16:17 # fsodeviced Jan 16 12:16:17 Aborting due to critical error:'GLib <>: kernel_idle_status_onState: assertion `self != NULL' failed' Jan 16 12:16:28 I bet you've the same error right? Jan 16 12:19:47 maybe I've an idea Jan 16 12:19:55 let's use gdb in a more advanced way Jan 16 12:20:00 put a breakpoint on this assert Jan 16 12:20:20 or on the print of the assert Jan 16 12:20:23 or something like that Jan 16 12:25:59 yeah, that looks like being the same Jan 16 12:27:03 mickeyl: any idea? Jan 16 12:28:32 I'll eat Jan 16 12:28:42 and I have to do some garden work :P Jan 16 12:28:53 kernel_idle_status_onState seem a function in C Jan 16 12:29:05 in kernel_idle/plugin.c Jan 16 12:29:18 maybe we disable that plugin and retry Jan 16 12:48:45 mrmoku, disabling the plugin makes it work Jan 16 12:51:40 GNUtoo|laptop: ok Jan 16 12:51:43 hi folks. which plugin is broken? Jan 16 12:51:57 we have a similiary segfault in fsodeviced and fsogsmd Jan 16 12:52:00 ok Jan 16 12:52:05 can i get a backtrace? Jan 16 12:52:06 so it might be something broken in the infra Jan 16 12:52:10 well... no :P Jan 16 12:52:13 darn Jan 16 12:52:15 it's an abort actually Jan 16 12:52:15 :) Jan 16 12:52:20 13:11 < GNUtoo|laptop> Aborting due to critical error:'GLib <>: kernel_idle_status_onState: assertion `self != NULL' failed' Jan 16 12:52:21 [fsodevice.kernel_idle] Jan 16 12:52:23 that plugin Jan 16 12:52:41 I have the same self != NULL assertion with fsogsmd on n900 Jan 16 12:52:49 please set a breakpoint in g_critical Jan 16 12:53:00 indeed my idea of breakpoint was a good idea Jan 16 12:53:06 but I didn't know where to set it Jan 16 12:53:11 I'll try Jan 16 12:53:22 mrmoku: where's it happening in fsogsmd? Jan 16 12:53:39 'GLib <>: string_to_string: assertion `self != NULL' failed' Jan 16 12:53:48 when you do what? Jan 16 12:53:54 start fsogsmd Jan 16 12:54:04 with isimodem and lowlevel_n900 enabled Jan 16 12:54:04 d'oh, on start Jan 16 12:54:12 fsogsmd: symbol lookup error: /usr/lib/cornucopia/modules/fsogsm/lowlevel_nokia900.so: undefined symbol: netlink_start Jan 16 12:54:48 GNUtoo|laptop: that should be old... what rev is it? Jan 16 12:55:01 ok Jan 16 12:55:05 first thing we should do is bump vala Jan 16 12:55:11 there has been an official release yesterday Jan 16 12:55:12 I've issues with revs Jan 16 12:55:14 just in case Jan 16 12:55:31 ok Jan 16 12:55:41 fsodeviced - 2:0.9.4+gitr6+f04c2d07cb880078eeb8546f10345a3322fbec63-r4.20.6 Jan 16 12:55:45 oops Jan 16 12:55:47 2nd, i will hunt for the fsodeviced thing Jan 16 12:55:56 fsogsmd - 1:0.5.0+gitr6+f04c2d07cb880078eeb8546f10345a3322fbec63-r4.5.6 Jan 16 12:56:00 checking whether i can reproduce it on the pc Jan 16 12:56:11 mickeyl, I'll try to give you a bt Jan 16 12:56:16 thanks Jan 16 12:56:31 GNUtoo|laptop: hmm... that would be current head Jan 16 12:56:39 but I don't have that Jan 16 12:57:32 maybe libnl issue Jan 16 12:57:35 brb Jan 16 12:57:36 anyway first the bt Jan 16 12:59:22 maybe I should rebuild n900 from scratch Jan 16 12:59:32 it would prevent some libs errors Jan 16 13:04:01 mickeyl, g_critical is in glib? Jan 16 13:05:12 because it didn't took the breakpoint Jan 16 13:06:43 yes, g_critical is in glib Jan 16 13:08:20 else can I break on glib loading? Jan 16 13:10:35 http://pastebin.com/xTMLWmPK Jan 16 13:10:50 here's what I have: b g_critical doesn't work Jan 16 13:11:29 for me neither Jan 16 13:13:57 mickeyl, what's the glib loop function? Jan 16 13:14:02 or init function Jan 16 13:14:25 maybe I could try to breakpoint on that Jan 16 13:15:47 _vala_main maybe? Jan 16 13:16:11 ok, _vala_main hits Jan 16 13:16:25 g_main_loop_run Jan 16 13:18:08 hits too Jan 16 13:18:38 hehe Jan 16 13:18:38 'Could not connect to DBus System bus: The connection is closed. dbus-daemon started?' Jan 16 13:18:42 too slow :P Jan 16 13:19:29 heh Jan 16 13:19:35 * mickeyl bumps vala in OE Jan 16 13:22:32 mickeyl: btw. for me on new kernel fsodeviced works fine Jan 16 13:23:03 ok Jan 16 13:23:06 it's a race condition Jan 16 13:23:18 I'll commit something to streamline the init process Jan 16 13:23:23 ok Jan 16 13:23:23 please build it and tell me whether it still works for you Jan 16 13:24:19 mickeyl: and if you have time you could take a look at lowlevel_n900... to check if it looks ok Jan 16 13:25:16 will do Jan 16 13:25:35 freesmartphone.org: 03mickey 07cornucopia * r69e11f74f348 10/libfsoframework/fsoframework/ (dbus.vala subsystem.vala): libfsoframework: send a signal once we have acquired a servicename Jan 16 13:25:36 freesmartphone.org: 03mickey 07cornucopia * r87197d555d10 10/fsodatad/README: fsodatad: fix README Jan 16 13:25:39 freesmartphone.org: 03mickey 07cornucopia * rdf22b1f6302c 10/libfsoresource/ (configure.ac fsoresource/dbusresource.vala): libfsoresource: rewrite resource init, bump Vala dependency Jan 16 13:25:41 freesmartphone.org: 03mickey 07cornucopia * r440556754b9d 10/libfsobasics/ (configure.ac fsobasics/utilities.vala): libfsobasics: fix yield syntax for Vala 0.11.4; bump Vala dependency Jan 16 13:25:43 freesmartphone.org: 03mickey 07cornucopia * re6a761361dc6 10/fsodeviced/src/plugins/kernel_idle/plugin.vala: fsodeviced: kernel_idle: rewrite init path to be more secure against race conditions Jan 16 13:25:45 freesmartphone.org: 03mickey 07cornucopia * r534f25d2ffa2 10/fsotdld/README: fsotdld: fix typo in README Jan 16 13:26:36 mickeyl: heh, better not :P Jan 16 13:26:43 fsogsmd: symbol lookup error: /usr/lib/cornucopia/modules/fsogsm/lowlevel_nokia900.so: undefined symbol: netlink_start Jan 16 13:26:48 now I have that too Jan 16 13:26:54 and I know why it was segfaulting Jan 16 13:26:58 /dev/cmt was not populated Jan 16 13:27:15 it's normal that it doesn't work Jan 16 13:27:20 g_critical doesn't exist Jan 16 13:27:23 nm confirmed it Jan 16 13:27:31 but g_assert_warning does Jan 16 13:28:44 mickeyl: looks like it's not linking in netlink.o Jan 16 13:29:38 in what lib g_critical should be? Jan 16 13:29:43 because it's valid source Jan 16 13:30:48 maybe I should just rebuild from scratch....again? Jan 16 13:30:52 mrmoku: do you have it in Makefile.am ? Jan 16 13:31:09 I've no such issues on htcdream that was rebuild from scratch more recently than nokia n900 Jan 16 13:31:12 check that the functions are not static Jan 16 13:31:58 heh, guess I'm just very stupid... moment Jan 16 13:32:09 maybe it's defined in a header? Jan 16 13:32:17 (the function) Jan 16 13:32:26 like boost does Jan 16 13:34:23 freesmartphone.org: 03mok 07cornucopia * rd2e211b55bc2 10/fsogsmd/src/plugins/lowlevel_nokia900/plugin.vala: Jan 16 13:34:23 freesmartphone.org: fsogsmd: use the correct names for extern netlink.c functions :P Jan 16 13:34:23 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 13:34:25 mickeyl: nvm. indeed being stupid :) Jan 16 13:34:54 mrmoku, does that fix my fsogsmd issue? Jan 16 13:35:02 GNUtoo|laptop: I hope so Jan 16 13:35:03 no need to rebuild from scratch? Jan 16 13:35:05 ok Jan 16 13:35:08 I'll try Jan 16 13:35:28 GNUtoo|laptop: thing is... I didn't have that, because it segfaulted before that point Jan 16 13:35:35 because /dev/cmt was not populated Jan 16 13:35:57 which is a 'new kernel' thing to have links to the gpio sys nodes for the modem there Jan 16 13:36:14 after populating it I had the same undefined symbol Jan 16 13:36:20 rebuilding to confirm Jan 16 13:36:34 ok Jan 16 13:37:24 I'll git pull in oe because of valac Jan 16 13:38:00 * mrmoku too Jan 16 13:40:00 freesmartphone.org: 03mickey 07cornucopia * raa22da1b6c32 10/fsophoned/ (28 files in 10 dirs): fsophoned: build skeleton for new subsystem daemon Jan 16 13:41:20 freesmartphone.org: 03mickey 07cornucopia * r671b9b054ca9 10/Makefile: update top-level Makefile Jan 16 13:42:14 freesmartphone.org: 03mickey 07cornucopia * r546fae01c222 10/README: update toplevel README Jan 16 13:43:21 freesmartphone.org: 03mickey 07cornucopia * rfe0c0eef993e 10/THANKS: add GNUtoo to contributors Jan 16 13:43:30 thanks Jan 16 13:43:50 :) Jan 16 13:44:51 maybe we should make some plans Jan 16 13:44:54 like: Jan 16 13:45:09 *make modem work and fix bugs that stands between that and us Jan 16 13:45:24 *do some kernel work Jan 16 13:45:34 and some uboot work Jan 16 13:46:01 in the order of importance Jan 16 13:46:14 mickeyl: | plugin.c:3894:2: error: too many arguments to function 'free_smartphone_usage_system_action_from_string' Jan 16 13:46:35 mickeyl: while building fsousaged Jan 16 13:47:28 I was wondering,all the people that uses the nokia900 with SHR are devs right Jan 16 13:47:52 so maybe that kernel stuff downgrade is not so important Jan 16 13:48:23 I mean if it's unusable but that we could develop on it and advance.... Jan 16 13:48:28 *nod* Jan 16 13:48:29 who cares.... Jan 16 13:48:32 mrmoku: d'oh... let me check Jan 16 13:49:03 only us crazy folks are using n900 with != ( maemo || meego ) Jan 16 13:49:05 in the meantime I reverted some stuff locally Jan 16 13:49:12 lol Jan 16 13:49:23 mrmoku: try to rebuild libfso-glib Jan 16 13:49:34 there are also some crazier folks with nitdroid Jan 16 13:49:34 mrmoku: vala needs a rebuild of libfso-glib Jan 16 13:49:40 since they included an important bugfix Jan 16 13:49:46 that changes the number of arguments in one function Jan 16 13:49:55 heh, right Jan 16 13:50:05 i can not understand the nitdroid at all Jan 16 13:50:15 there is so many cool android hardware Jan 16 13:50:15 nitdroid has telephony but no voice calls Jan 16 13:50:32 then again, i can't understand the htclinux folks either Jan 16 13:50:32 so if they have voice calls a day Jan 16 13:50:43 we could copy their voice-call alsa stuff Jan 16 13:50:55 sounds good Jan 16 13:51:16 I can understand nitdroid people, they want to dual-boot many OS/distros on the same hardware Jan 16 13:51:32 no need to buy 1 android phone, one maemo phone, one foo phone etc... Jan 16 13:52:09 also the developper could have bought the n900 long time ago Jan 16 13:52:17 but htc-linux...hmmmm Jan 16 13:52:42 I mean between a half-working hd2 and a nexusone or nexus S or something like that Jan 16 13:53:50 freesmartphone.org: 03mickey 07cornucopia * reee3c3381a0c 10/fsophoned/ (3 files in 3 dirs): fsophoned: fix default configuration file Jan 16 13:53:51 yep Jan 16 13:53:52 freesmartphone.org: 03mickey 07cornucopia * r4e841bf22c5f 10/ (2 files in 2 dirs): libfsoframework: interfaces: add Phone and reorder alphabetically Jan 16 13:54:55 mickeyl: thx, built successfully now Jan 16 13:55:25 good Jan 16 13:55:30 let me have a look at the netlink stuff now Jan 16 13:57:09 that looks pretty clean Jan 16 13:57:10 i like it Jan 16 13:57:18 please wrap all calls to logger.debug with an assert thoguh Jan 16 13:57:31 so we can automatically remove them at some point of time Jan 16 13:59:00 ahh, ok Jan 16 13:59:21 i plan that we eventually compile all the stuff with G_DISABLE_ASSERT Jan 16 13:59:23 (or something) Jan 16 13:59:31 so only info, warning, and error continue to be in the binaries Jan 16 13:59:51 that should give some nice speedup Jan 16 14:00:00 :) Jan 16 14:00:02 especially for stuff that logs a lot to debug Jan 16 14:00:04 (fsogsmd...) Jan 16 14:02:30 btw. Jan 16 14:03:03 we should use -flto at some point of time Jan 16 14:03:15 (whole program optimization) Jan 16 14:03:43 (should be an OE setting in CFLAGS and LDFLAGS) Jan 16 14:04:40 especially for vala programs this should help a bit Jan 16 14:04:48 as they often contain repetetive patterns Jan 16 14:05:00 due to all the bunch of generated helper functions Jan 16 14:05:42 bbl, gone for a walk, it's nice weather here Jan 16 14:05:44 speedup are nice on lower end platform(like freerunner) Jan 16 14:06:09 freesmartphone.org: 03mok 07cornucopia * rd9f5a265593f 10/fsogsmd/src/plugins/lowlevel_nokia900/plugin.vala: Jan 16 14:06:09 freesmartphone.org: fsogsmd: wrap all logger.debug in assert within lowlevel_nokia900 Jan 16 14:06:09 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 14:08:54 * mrmoku lunch Jan 16 14:09:05 a bit late... but as part of the family is italian... that should be ok ;) Jan 16 14:26:32 * GNUtoo|laptop will wait for | plugin.c:3894:2: error: too many arguments to function 'free_smartphone_usage_system_action_from_string' Jan 16 14:40:09 GNUtoo|laptop: did you rebuild libfso-glib? Jan 16 14:40:19 that fixed it for me Jan 16 14:40:21 no, I should Jan 16 14:40:23 ok Jan 16 14:50:08 mrmoku, thanks it did an shr-image now Jan 16 14:52:55 good Jan 16 14:53:08 and I'm back to 1970-01-01T04:28:54.442780Z [CRITICAL] fsogsmd : GLib <>: string_to_string: assertion `self != NULL' failed Jan 16 14:53:14 at a different point :P Jan 16 15:00:34 hmm Jan 16 15:01:30 during operation or right ons tartup? Jan 16 15:02:06 hmm Jan 16 15:02:07 fso_gsm_abstract_modem_constructor Jan 16 15:02:11 mrmoku, what are the messages around that entry? Jan 16 15:02:34 aah Jan 16 15:02:45 mrmoku: i think you're not allowed to use the logger in construct Jan 16 15:02:49 it's not created yet Jan 16 15:02:52 looks like an output went wrong Jan 16 15:02:52 ohh Jan 16 15:03:00 or something like tht Jan 16 15:03:06 try removing that statement Jan 16 15:03:27 let me check here Jan 16 15:03:34 it should also fail here on x86 Jan 16 15:04:00 hmm... the logger in construct works though Jan 16 15:04:04 and it aborts afterwards Jan 16 15:04:59 are there any string templates? Jan 16 15:05:28 hmm in fso_gsm_abstract_modem_constructor ? Jan 16 15:06:24 * mrmoku fires gdbserver to see actual code Jan 16 15:07:50 on x86 that looks good so far Jan 16 15:07:53 (of course it can't open the modem) Jan 16 15:08:11 it can trough g_nokia no? Jan 16 15:08:22 and maemo forwarding etc... Jan 16 15:08:26 no Jan 16 15:08:34 lowlevel will never operate on x86 Jan 16 15:08:38 i don't have any gpio nodes Jan 16 15:08:39 etc. Jan 16 15:08:40 ah lowlevel Jan 16 15:08:44 yes indeed Jan 16 15:08:51 I didn't know it was lowlevel Jan 16 15:09:22 strace shows that: Jan 16 15:09:52 http://pastebin.com/uRsuKLSd Jan 16 15:10:07 so right after gpio switch stuff Jan 16 15:10:34 ok, lets validate all logger calls in the lowlevel plugin Jan 16 15:10:51 and the log says: Jan 16 15:11:05 http://pastebin.com/6tDamkaV Jan 16 15:11:09 altough Jan 16 15:11:12 beware Jan 16 15:11:16 I'm on 2.6.28 Jan 16 15:11:31 GNUtoo|laptop: do I have to use substitute-path too when running gdb on my buildhost? Jan 16 15:11:47 mrmoku, if I remember well that's for the sources Jan 16 15:11:59 yup Jan 16 15:12:01 mrmoku, if you build on your laptop, no need Jan 16 15:12:08 /media/port4 beeing where the root filesystem of the build system is mounted(could be sshfs,NSF etc...) Jan 16 15:12:11 ok Jan 16 15:12:13 is $rapu_type garanteed to be != nul? Jan 16 15:12:26 just set sysroot Jan 16 15:12:30 yup and file Jan 16 15:12:37 mickeyl: I think so Jan 16 15:12:39 hmm Jan 16 15:12:44 a backtrace would really ease our search Jan 16 15:12:58 it's an if else Jan 16 15:12:59 so yes Jan 16 15:13:03 we must find a good breakpoint then Jan 16 15:13:06 maybe printf Jan 16 15:13:09 working on the backtrace :) Jan 16 15:13:21 for me the good bp is fso_gsm_abstract_modem_constructor Jan 16 15:15:27 hmm. maybe i forgot a flag that critical shouldn't abort? Jan 16 15:17:09 G_DEBUG_FATAL_CRITICALS Jan 16 15:17:09 GNUtoo|laptop: heh, I _have_ to use substitute-path... I'm building in a chroot :P Jan 16 15:17:16 ok Jan 16 15:18:58 hmm... still no code in bt Jan 16 15:20:51 when does it exactly happens? when requesting the resource? Jan 16 15:20:58 i.e. during lowlevel_poweron? Jan 16 15:20:58 mrmoku, I've a hint Jan 16 15:21:01 no Jan 16 15:21:15 I comment lowlevel and there is still that or maybe another Jan 16 15:21:36 *that critical error Jan 16 15:22:29 aah Jan 16 15:22:34 i can reproduce it now Jan 16 15:22:51 even with modem on it does it Jan 16 15:23:00 hmm Jan 16 15:23:02 fsogsmd : GLib <> Jan 16 15:23:03 mickeyl: great, as I'm unable to give a goot bt Jan 16 15:23:06 yup Jan 16 15:23:12 that's playya's new code Jan 16 15:23:32 i can't get a backtrace either Jan 16 15:23:39 something messes with the stack Jan 16 15:25:03 hmm Jan 16 15:25:16 or rather that the logger now picks up the glib critical Jan 16 15:25:34 that prevents debugging Jan 16 15:25:44 as we probably get an infinite recursion Jan 16 15:27:35 ok Jan 16 15:29:36 ah Jan 16 15:33:41 yes. i think glib aborts on a not so critical critical ;) Jan 16 15:41:56 mrmoku: show me your modem_access line Jan 16 15:43:28 mickeyl: in fsogsmd.conf? Jan 16 15:43:32 no such thing Jan 16 15:43:39 I've none too Jan 16 15:43:44 ah Jan 16 15:43:47 I wondered about that and forgott Jan 16 15:43:50 that would be a problem Jan 16 15:44:50 modem_access = phonet:phonet0: is the right thing according to my irc logs Jan 16 15:44:55 freesmartphone.org: 03mickey 07cornucopia * rebcbd9d9f58f 10/fsogsmd/src/lib/modem.vala: fsogsmd: don't bail out on invalid modem_access line Jan 16 15:44:55 correct Jan 16 15:45:08 now it seem to work Jan 16 15:45:48 root@nokia900 ~ # fsogsmd Jan 16 15:45:48 Aborting due to critical error: Jan 16 15:45:48 'Interface phonet0 not available' Jan 16 15:45:56 that's better Jan 16 15:46:00 yup Jan 16 15:47:55 for me it works Jan 16 15:48:08 after modprobe omap_ssi I have a phonet0 Jan 16 15:48:24 and fsogsmd segfaults now Jan 16 15:48:28 this time a real one :) Jan 16 15:48:59 http://shr.pastebin.com/ia6X8bqQ Jan 16 15:49:29 in libisi Jan 16 15:50:01 ya, seen that before Jan 16 15:50:06 something is broken in destroy Jan 16 15:50:15 ah now it segfault Jan 16 15:50:16 however if init works fine, it shouldn't destroy Jan 16 15:50:26 but I had to start phonefsod to have that Jan 16 15:50:30 so i guess the real problem might be somewhere else Jan 16 15:51:53 1970-01-01T05:28:29.472687Z [DEBUG] NokiaIsiModem : Created; configured for phonet:phonet0@0 Jan 16 15:51:56 1970-01-01T05:28:29.478760Z [DEBUG] fsogsmd : GLib <>: g_pn_netlink_add_route: Invalid argument Jan 16 15:52:17 and the callback from netlink.c works :-) Jan 16 15:52:17 1970-01-01T05:28:29.480835Z [DEBUG] fsogsmd : GLib <>: Phonet: phonet0 is down, idx=0x5 Jan 16 15:52:21 good so far Jan 16 15:52:48 ahh... no Jan 16 15:52:51 that's not my callback :P Jan 16 15:53:40 is the lowlevel getting called at all? Jan 16 15:54:23 #0 0x406cb31c in g_isi_client_destroy () from /home/gnutoo/embedded/sshfs/usr/lib/libisi.so.0 Jan 16 15:55:48 hmm... Jan 16 15:55:50 * mrmoku has an idea Jan 16 15:57:27 freesmartphone.org: 03mok 07cornucopia * re489d6fe9fc4 10/fsogsmd/src/plugins/lowlevel_nokia900/netlink.c: Jan 16 15:57:27 freesmartphone.org: fsogsmd: fix indentation in lowlevel_nokia900/netlink.c Jan 16 15:57:27 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 15:57:30 freesmartphone.org: 03mok 07cornucopia * recfe78cdf462 10/fsogsmd/src/plugins/lowlevel_nokia900/netlink.c: Jan 16 15:57:30 freesmartphone.org: fsogsmd: use the correct inteface index for phonet in lowlevel_nokia900 Jan 16 15:57:30 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 15:57:42 freesmartphone.org: 03morphis 07aurora * rca251c5e0473 10/zhone2/src/locationpage.vala: zhone2: add missing parameters to printf statement Jan 16 15:58:46 mickeyl: yeah, lots of loggers from lowlevel :-) Jan 16 15:59:20 ok Jan 16 15:59:39 I know where Jan 16 15:59:46 indeed Jan 16 15:59:48 lowlevel Jan 16 15:59:53 I just desactivated it Jan 16 16:00:02 and enabled the modem manually Jan 16 16:00:06 mickeyl: http://shr.pastebin.com/t0RQiq71 is the complete log Jan 16 16:00:09 and it didn't segfault so far Jan 16 16:00:54 it asked for my pin Jan 16 16:01:12 wow :) Jan 16 16:03:26 1970-01-01T00:03:48.882568Z [DEBUG] fsogsmd : GLib <>: mediators.vala:74: code = 9, ISI_SIM_AUTH_STATUS_AUTHORIZED Jan 16 16:03:33 1970-01-01T00:03:48.882720Z [WARN] NokiaIsiModem : Unhandled ISI SIMAuth.Status ISI_SIM_AUTH_STATUS_AUTHORIZED Jan 16 16:03:42 maybe we must fix this clock Jan 16 16:03:55 else people would think that our phone is so 70's Jan 16 16:04:49 maybe dos1 could fix the clock? Jan 16 16:04:49 :P Jan 16 16:04:59 it's also an issue with savegames Jan 16 16:05:08 and a lot of stuff that uses dates for saving stuff Jan 16 16:07:33 ok, that looks better Jan 16 16:07:38 still segfaulting though Jan 16 16:07:57 mickeyl: it's disabling the modem on startup... isn't that why client_destroy is called? Jan 16 16:08:16 i don't think so Jan 16 16:08:21 freesmartphone.org: 03morphis 07aurora * rfb6cd87c7325 10/zhone2/src/locationpage.vala: zhone2: it's a double and not an integer ... Jan 16 16:08:23 disabling the modem does not remove the modem object Jan 16 16:08:45 http://shr.pastebin.com/eFCmBRk1 Jan 16 16:08:46 once the modem object is being garbage collected, the isi client is being destroyed Jan 16 16:09:01 1970-01-01T05:46:45.201813Z [DEBUG] fsogsmd : GLib <>: Phonet: phonet0 is down, idx=0x5 Jan 16 16:09:08 this one is from where? Jan 16 16:09:13 libisi Jan 16 16:09:21 this is the last thing it logs Jan 16 16:09:29 this glib logging is a good idea, but confuses us Jan 16 16:09:35 i'd vote for disabling it temporarily Jan 16 16:10:50 mrmoku: i'd ignore the SIGSEGV in destroy Jan 16 16:10:56 the problem is that it doesn't startup Jan 16 16:10:57 mickeyl: libisi is calling g_isi_client_destroy(modem->client); Jan 16 16:11:00 when the link goes down Jan 16 16:11:07 hmm Jan 16 16:11:41 which file is this? Jan 16 16:11:47 isi/modem.c Jan 16 16:12:00 line 150 Jan 16 16:12:12 ok Jan 16 16:12:14 argh Jan 16 16:12:18 so there's a bug in g_isi Jan 16 16:12:23 fun Jan 16 16:12:26 that's the ofono code Jan 16 16:13:16 let me unpack the n900 Jan 16 16:18:33 ok, via forwarding it looks like that: Jan 16 16:18:51 http://pastebin.com/47yMhaTT Jan 16 16:20:22 so, on the real device lowlevel is turning down the modem, which destroys the client? Jan 16 16:20:52 looks like Jan 16 16:21:00 hmm Jan 16 16:21:28 isn't destroying the client something we don't want in any case? Jan 16 16:21:31 apart from shutdown? Jan 16 16:21:34 yes Jan 16 16:21:48 though Jan 16 16:21:51 then we should remove that in the netlink callback in libisi Jan 16 16:22:00 we can do that Jan 16 16:22:07 note that the init sequence is broken though Jan 16 16:22:11 it has been only written for a quick test Jan 16 16:22:21 ok Jan 16 16:22:22 take a look at the modem plugin Jan 16 16:22:29 already opened in vim Jan 16 16:22:36 it tries to instanciate the modem in construct Jan 16 16:22:40 so _before_ lowlevel is being called Jan 16 16:22:49 if we were to properly init it Jan 16 16:23:00 the client would be initialized only _after_ the modem has been turned on Jan 16 16:23:05 (isiclient) Jan 16 16:23:22 ok, that we should fix first then Jan 16 16:23:25 right Jan 16 16:23:42 moving modem init from construct into an actual dummy transport Jan 16 16:24:01 for me it works Jan 16 16:24:05 I start it manually Jan 16 16:24:16 trough the C activator Jan 16 16:24:19 for 2.6.28 Jan 16 16:24:53 sure, because you do that before starting fsogsmd, right? Jan 16 16:25:03 yes Jan 16 16:25:17 and I disabled the low-level plugin too Jan 16 16:25:31 altough I don't remember exactly when I did it Jan 16 16:25:48 mickeyl: how does the isi modem know when it's up? should the lowlevel plugin do something? Jan 16 16:26:11 no, we're going to rely on libisi here Jan 16 16:26:31 we only are interested in the highlevel reachability Jan 16 16:26:36 let me quickly change the init order Jan 16 16:26:38 ok Jan 16 16:44:31 so much for quickly... Jan 16 16:44:34 will take a bit Jan 16 16:48:15 all of libisi is in cornucopia? Jan 16 16:48:16 hehe, ok Jan 16 16:48:25 I mean all is implemented in fsogsmd Jan 16 16:49:11 what's fsophoned for? Jan 16 16:49:16 I've read the README Jan 16 16:51:17 GNUtoo|laptop: abstract interface for different type of transports like GSM or VOIP for telephony Jan 16 16:51:24 wow Jan 16 16:51:33 :) Jan 16 16:51:50 will it work in zhone2? Jan 16 16:52:15 when it's ready and somebody switch zhone2 to use it (which mickeyl will probably do) Jan 16 16:52:40 ohh nice Jan 16 16:53:47 maybe we need to map what's implemented and what isn't Jan 16 16:53:53 in libisi+fsogsmd Jan 16 16:53:57 I mean in detail Jan 16 16:54:04 not sim pin Jan 16 16:54:09 but each callback Jan 16 16:54:16 or maybe we just read the code Jan 16 16:54:20 it was a bad idea Jan 16 16:57:52 or maybe Jan 16 16:58:01 when we implement something we write it somewhere? Jan 16 16:58:49 because I'd like to add somehting small right now Jan 16 16:58:52 and I'm lost Jan 16 16:59:15 * GNUtoo|laptop thinks of mdbus2 -s org.freesmartphone.ogsmd /org/freesmartphone/GSM/Device Jan 16 16:59:27 there is a lot of methods Jan 16 16:59:33 and I don't know which ones work Jan 16 17:00:12 GNUtoo|laptop: on n900 ? Jan 16 17:00:24 yes Jan 16 17:00:26 isi modem Jan 16 17:00:41 ok Jan 16 17:00:48 no idea either Jan 16 17:01:13 org.freesmartphone.GSM.SIM.GetAuthStatus says org.freesmartphone.Unavailable: This function is not available while modem is in state FSO_GSM_MODEM_STATUS_INITIALIZING Jan 16 17:01:23 so I bet the modem has to advance to make that available Jan 16 17:01:55 1970-01-01T00:03:48.882720Z [WARN] NokiaIsiModem : Unhandled ISI SIMAuth.Status ISI_SIM_AUTH_STATUS_AUTHORIZED Jan 16 17:02:01 maybe I'll try to fix that Jan 16 17:02:12 * mrmoku dinner meanwhile Jan 16 17:03:36 freesmartphone.org: 03mickey 07cornucopia * rbd97d8a255e3 10/fsogsmd/src/plugins/modem_nokia_isi/ (Makefile.am channel.vala isitransport.vala plugin.vala): fsogsmd: modem_nokia_isi: use dedicated IsiTransport; refactor libisi client initialization Jan 16 17:03:39 hmmm Jan 16 17:03:44 maybe it's because of my sim Jan 16 17:03:51 my sim is not authorized Jan 16 17:03:58 because I didn't bother to go activate it Jan 16 17:04:04 I'll put my other sim inside Jan 16 17:07:00 freesmartphone.org: 03mickey 07cornucopia * r38eadb370c35 10/fsogsmd/src/plugins/modem_nokia_isi/isitransport.vala: fsogsmd: modem_nokia_isi: warn only, if several subsystems are not reachable (yet?) Jan 16 17:08:05 mrmoku: ok, worth another try now Jan 16 17:08:27 mrmoku, what was the fix to start phonefsod automatically? Jan 16 17:09:37 mickeyl, he's eating Jan 16 17:09:42 mickeyl, what should I implement Jan 16 17:09:49 I'm looking for something quick and easy Jan 16 17:09:55 as I'm tired now Jan 16 17:10:03 headache Jan 16 17:10:34 FsoGsmNetworkGetSignalStrength? Jan 16 17:11:04 do you want to do something in libisi or fsogsmd? Jan 16 17:11:10 yes Jan 16 17:11:17 I did all the setup for that Jan 16 17:11:29 that was an 'or' question :) Jan 16 17:11:32 or both? Jan 16 17:11:47 fsogsmd first Jan 16 17:12:17 ok, warm up yourself by implementing org.freesmartphone.GSM.Network.GetSignalStrength Jan 16 17:12:23 ok Jan 16 17:12:26 libisi already exports that Jan 16 17:12:34 in what source file is it? Jan 16 17:12:38 in fsogsmd Jan 16 17:12:42 mediator? Jan 16 17:13:00 yes, nokia_isi_modem/mediators.c Jan 16 17:13:01 err, vala Jan 16 17:13:05 ok Jan 16 17:13:07 you need Jan 16 17:13:11 a) implement the new mediator Jan 16 17:13:13 b) register it Jan 16 17:13:19 (same file) Jan 16 17:13:32 ok so I don't piggy-back on an existing mediator Jan 16 17:13:33 ok Jan 16 17:13:36 I understand Jan 16 17:13:48 you override the generic mediators (which are empty and only define the interface) Jan 16 17:13:50 every AT-command equivalent has his own mediator Jan 16 17:14:08 lets say every high level command Jan 16 17:14:13 ok Jan 16 17:14:22 usually one mediator combines multiple modem commands Jan 16 17:14:27 ok Jan 16 17:14:29 not in the case of signal strength though, so you're right Jan 16 17:14:37 ok lol Jan 16 17:14:43 I picked the right onew Jan 16 17:14:44 *one Jan 16 17:15:54 lol Jan 16 17:16:03 it seems that you can set the signal strength Jan 16 17:16:07 maybe for dummy modem Jan 16 17:16:20 hmm? Jan 16 17:16:44 http://pastebin.com/kNGjRqME Jan 16 17:16:46 get set Jan 16 17:17:05 gena2x: hi, I remember you told me that it's possible to use touchscreen via xf86-input-evdev, can you explain how to calibrate it? Jan 16 17:17:17 that's a property Jan 16 17:17:23 someone needs to set it Jan 16 17:17:32 (i.e. the mediator) Jan 16 17:17:33 ok Jan 16 17:17:37 to the value it gathers from the modem Jan 16 17:17:39 understood Jan 16 17:17:54 and the dbus service gets it Jan 16 17:17:56 anarsoul: hi, i personally prefer tslib, evdev is shr guys favorite (last time i checked) Jan 16 17:17:56 the fact that the property was public instead of protected or something like that mislead me Jan 16 17:17:59 and writes it back to the responder Jan 16 17:18:13 it's public because the dbus service does not derive from the mediator Jan 16 17:18:17 anarsoul: probably JaMa|Off may tell more Jan 16 17:18:30 have a look at the dbus service to get the whole picture Jan 16 17:18:33 it's actually pretty simple Jan 16 17:18:40 it seems he's not avail right now :( Jan 16 17:19:00 I'll fist implement the org.freesmartphone.GSM.Network.GetSignalStrength first Jan 16 17:19:31 anarsoul: probably he'll answer than he will be back and available Jan 16 17:19:47 gena2x: but what about right button emulation with xf86-input-tslib? Jan 16 17:20:09 anarsoul: rmb emulation is strong reason to use tslib :) Jan 16 17:20:19 anarsoul: it works Jan 16 17:20:30 ok Jan 16 17:25:08 <[Rui]> gena2x: shr has tslib right now Jan 16 17:25:38 <[Rui]> gena2x: but rmb emulation should be configurable Jan 16 17:26:00 <[Rui]> gena2x: if an app (say eve) uses rmb a lot, when you drag pages you're frequently calling up the menu Jan 16 17:26:08 [Rui]: but what had happen with evdev? i thought it were used? Jan 16 17:26:36 <[Rui]> gena2x: it was by the time of the filtering issue, and tslib also had better results (I was also insisting on it because of rmb) Jan 16 17:29:38 mickeyl, what is run.callback(); Jan 16 17:30:02 how does the whole callback stuff works in the mediator Jan 16 17:30:15 in the network.vala in libisi it seem much more simplier Jan 16 17:31:07 yes, demo programs do not need the complexity of full programs Jan 16 17:31:15 I know Jan 16 17:31:29 one thing you have to understand is how coroutines work in vala Jan 16 17:31:34 basically if you have an async method Jan 16 17:31:36 ahh ok Jan 16 17:31:38 you can yield at any time Jan 16 17:31:40 however Jan 16 17:31:41 ok Jan 16 17:31:52 someone needs to call you, so execution continues after the yield Jan 16 17:32:00 ok Jan 16 17:32:01 and this thing is .callback Jan 16 17:32:29 is there some coroutines howto? Jan 16 17:32:40 in case I need to know something more Jan 16 17:32:44 btw thanks a lot Jan 16 17:32:55 http://blogs.gnome.org/juergbi/2009/09/18/closures-and-asynchronous-methods-in-vala/ Jan 16 17:33:03 thanks Jan 16 17:33:05 for the general concept, read the wikipedia on closures Jan 16 17:33:37 we use closures in our vala stuff so not have millions of callback functions Jan 16 17:33:44 s/to/ Jan 16 17:33:50 s/so/to/ Jan 16 17:34:08 ok Jan 16 17:34:29 that way one high level function (i.e. get network status) maps nicely in one mediator function (run), although we call multiple async functions in libisi with callbacks returning the results to us Jan 16 17:35:15 bbl, dinner Jan 16 17:35:23 I should go to eat too Jan 16 17:35:29 thanks Jan 16 17:37:19 [Rui]: yes. probably increasing rmb timeout may help a lot. also it is possible to rmb with some special programs. but i think problem is more complicated. from first glance, X should announce what kind of pointer is user, so programs which have support of single-button operation, should turn off emulation. Jan 16 17:37:37 s/user/used/ Jan 16 17:37:38 gena2x meant: [Rui]: yes. probably increasing rmb timeout may help a lot. also it is possible to rmb with some special programs. but i think problem is more complicated. from first glance, X should announce what kind of pointer is used, so programs which have support o... Jan 16 17:37:54 [Rui]: ... of single-button operation, should turn off emulation. Jan 16 17:38:15 [Rui]: of something like this Jan 16 17:38:19 *or Jan 16 17:39:19 hi guys Jan 16 17:39:36 i was wondering... development on embbeded devices is very hard Jan 16 17:39:59 reaching the level of flexibility and ease of development of applications of android is hard and loooong task Jan 16 17:40:16 i really like android... but i don't want java! :( Jan 16 17:44:49 daniele_athome, hi Jan 16 17:45:01 daniele_athome, developping for android is harder Jan 16 17:45:35 daniele_athome, is it possible to use your keyboard outside your mokosuite environment? Jan 16 17:45:59 GNUtoo|laptop, mmm... not outside the mokosuite window manager Jan 16 17:46:12 but you can bring out the source module managing it, it's pretty indipendent Jan 16 17:46:17 s/pretty/quite/ Jan 16 17:46:17 daniele_athome meant: but you can bring out the source module managing it, it's quite indipendent Jan 16 17:46:32 I'ld like to have it as a normal illume keyboard for now Jan 16 17:46:42 because it doesn't work well in mokosuite Jan 16 17:47:04 under the nexus one Jan 16 17:47:22 (the phone that would benefit the most from this kind of keyboards) Jan 16 17:47:59 GNUtoo|laptop, if you tell me what's wrong i could make some adjustments Jan 16 17:48:10 otherwise you'll have to develop an illume keyboard module Jan 16 17:49:03 GNUtoo|laptop, why do you say it's harder on android Jan 16 17:49:13 i mean integration level and api support is huge Jan 16 17:50:23 daniele_athome, you can develop like for a desktop on GNU/Linux phones Jan 16 17:50:27 you can re-use stuff Jan 16 17:50:28 etc... Jan 16 17:50:30 no java Jan 16 17:50:40 or rather java is not mandatory Jan 16 17:50:49 you have the standard libc Jan 16 17:50:50 yes but except for that Jan 16 17:50:50 etc... Jan 16 17:51:06 exept for that I didn't try developping applications under android Jan 16 17:51:10 I only did low level Jan 16 17:52:35 anarsoul: check xinput-calibrator Jan 16 17:54:24 daniele_athome, are screenshots enough? Jan 16 17:54:36 GNUtoo|laptop, it's a start :) Jan 16 17:54:42 ok Jan 16 17:54:48 I'll try to do some screenshots Jan 16 17:54:49 mickeyl: still same segfault in destroy Jan 16 17:54:54 altough it's hard Jan 16 17:55:08 because the keyboard mask the send to scap.linuxtogo Jan 16 17:55:14 in gpe-scap Jan 16 17:55:20 I'll have to assign some buttons Jan 16 18:09:19 <[Rui]> daniele_athome: does mokosuite's keyboard handle dicts like raster's? does it support special characters unlike raster's? Jan 16 18:09:54 <[Rui]> if yes to both, it would be useful to make a new keyboard based on it Jan 16 18:12:00 mickeyl: complete log here http://paste.pocoo.org/show/321661/ Jan 16 18:13:07 mickeyl: 1970-01-01T07:48:09.901795Z [DEBUG] fsogsmd : GLib <>: g_pn_netlink_add_route: Invalid argument Jan 16 18:13:48 yes Jan 16 18:14:04 lets outcomment this for now Jan 16 18:14:57 (g_isi_client_destroy) Jan 16 18:15:13 mickeyl, I didn't undersood wikipedia, but fortunately I undersood your code Jan 16 18:15:27 [Rui], both no :) Jan 16 18:15:28 public FreeSmartphone.GSM.NetworkProvider[] providers { get; set; } Jan 16 18:15:31 was the missing part Jan 16 18:15:36 it's just a simple qwerty keyboard Jan 16 18:15:38 and far from complete Jan 16 18:15:42 <[Rui]> daniele_athome: damn Jan 16 18:15:44 some buttons don't work yet Jan 16 18:15:45 :) Jan 16 18:16:15 mickeyl: ok Jan 16 18:16:43 i'm not sure whether libisi supports multiple up/downs atm. Jan 16 18:16:52 but one cycle should be enough for us to continue working Jan 16 18:16:57 (i.e. off -> on Jan 16 18:17:30 not worse than calypso :P Jan 16 18:17:35 hehe, right Jan 16 18:18:00 are you on .37 atm.? Jan 16 18:18:12 yup Jan 16 18:18:15 could you have a quick look at whether forwarding works? Jan 16 18:18:32 do they know in #ofono or #meego-arm ? Jan 16 18:18:44 with 'fixed' libisi ? Jan 16 18:19:11 GNUtoo|laptop: feel free to ask. i did ask a couple of times without results, but perhaps the right persons were not in Jan 16 18:19:20 ok Jan 16 18:19:36 strkeeps didn't know? Jan 16 18:19:46 i don't remember any names Jan 16 18:19:47 *Stskeeps Jan 16 18:19:49 ok Jan 16 18:19:49 mickeyl: will do when finished cleaning the three dents of the young man :) Jan 16 18:19:56 hehe Jan 16 18:19:57 righto Jan 16 18:27:56 mickeyl: hmm... looking at the libisi code I would have to change more than just disabling the destroy Jan 16 18:28:06 like just creating the modem once Jan 16 18:28:12 which it does on every link up now Jan 16 18:28:56 hmm Jan 16 18:29:25 i don't know why he did it that way Jan 16 18:29:34 perhaps we can just ignore the link status completely Jan 16 18:29:43 (on that layer) Jan 16 18:32:11 mickeyl, I should suscribe strength right Jan 16 18:32:12 ? Jan 16 18:32:22 because there is already a request strength that I found Jan 16 18:33:09 there is request or suscribe strength Jan 16 18:33:13 request is ok Jan 16 18:33:27 * GNUtoo|laptop started to suscribe strength Jan 16 18:33:40 as it's callback based.... Jan 16 18:33:46 hmm Jan 16 18:33:48 but not in that request Jan 16 18:33:54 ah ok Jan 16 18:33:57 ah right Jan 16 18:34:08 if you want to subscribe for signal strength updates, it needs to be done in the equivalent of channel init Jan 16 18:34:10 it's the client that request the strength Jan 16 18:34:13 not a signal Jan 16 18:34:16 yes Jan 16 18:34:22 ok got it Jan 16 18:41:44 GNUtoo|laptop: so you see... noone has a clue :( Jan 16 18:42:13 ? Jan 16 18:42:16 #meego-arm Jan 16 18:42:20 since you didn't ask, i did again Jan 16 18:42:23 ah ok Jan 16 18:48:51 make[4]: *** [commands.c] Bus error Jan 16 18:48:59 * GNUtoo|laptop is frightened and go checks dmesg Jan 16 18:49:36 nothing in dmesg .... Jan 16 18:49:39 apart overheat Jan 16 18:49:47 and recompiling works Jan 16 18:50:49 mickeyl: I'm wondering if libisi isn't actually correct... Jan 16 18:50:55 and we should just find where it segfaults Jan 16 18:51:15 you mean gisi might be wrong? Jan 16 18:51:18 as it's not destroying the modem... but a client that connected to it Jan 16 18:51:31 which might be you from your laptop Jan 16 18:51:51 shit nslu-log, useless as it only has recent day Jan 16 18:51:54 hmm? Jan 16 18:51:55 ~log Jan 16 18:51:55 [log] http://logs.nslu2-linux.org/livelogs/openmoko-cdevel/ Jan 16 18:52:02 ~logs Jan 16 18:52:03 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 Jan 16 18:53:12 mickeyl: nvm. just forget what I said :P Jan 16 18:53:20 ok :) Jan 16 18:53:51 will do a local build of libisi with just the destroy disabled and see what will happen :) Jan 16 18:54:00 meh!!! Jan 16 18:54:01 ok Jan 16 18:54:15 mickeyl: what's been the AT cmd for cellbroadcast? Jan 16 18:54:23 +CSCB iirc Jan 16 18:54:28 thanks Jan 16 18:54:35 =1,"","" Jan 16 18:56:13 private int s = 0; Jan 16 18:56:23 the compiler errors on that Jan 16 18:56:31 vala compiler Jan 16 18:57:05 in which scope? Jan 16 18:57:14 private is only available in class scope Jan 16 18:57:42 http://pastebin.com/ddNFrNMe Jan 16 18:58:12 private is not available in method scope Jan 16 18:58:14 just remove it Jan 16 18:58:21 makes no sense in method scope Jan 16 18:58:31 as it's a class-wide visibility attribute Jan 16 18:58:35 ok Jan 16 18:58:44 I'll try something Jan 16 18:59:15 ok seem to work now Jan 16 18:59:30 that confused me: Jan 16 18:59:39 var p = new FreeSmartphone.GSM.NetworkProvider[] {}; Jan 16 18:59:45 which was in method scope Jan 16 19:00:25 softupd! Jan 16 19:00:37 ? Jan 16 19:00:51 it's softupd Jan 16 19:01:13 initctl stop softupd kills forwarding Jan 16 19:01:22 initctl start softupd makes it working again Jan 16 19:01:29 ok Jan 16 19:02:28 GNUtoo|laptop, have to go now Jan 16 19:02:54 ok Jan 16 19:03:00 see you later Jan 16 19:04:10 could anyone here tell me where I can find the alsa scenarios (states?) for the current unstable SHR? Jan 16 19:05:07 polz: /etc/freesmartphone/conf/openmoko_gta/alsa-default Jan 16 19:05:22 polz: but take care... they're *not* alsamixer compatible Jan 16 19:06:10 softupd contains some interesting strings: Jan 16 19:06:23 CMT: Unknown RPC message Jan 16 19:06:24 Unhandled message received while offline (%02x %02x) Jan 16 19:06:24 CMT: sending packet with length %d Jan 16 19:06:34 mrmoku: are they supposed to be without endings? Jan 16 19:06:36 that somewhat supports my theory of userland forwarding Jan 16 19:06:56 and are they supposed to contain carriage returns before newlines? Jan 16 19:07:30 mickey@saphir:/tmp$ strings softupd |grep upnlink0 Jan 16 19:07:30 upnlink0 Jan 16 19:07:33 hah Jan 16 19:07:58 mickeyl: iirc (judging by softupd disassmebly i tried to do) it looked like softupd indeed knows how to reflash the modem. Jan 16 19:08:05 mickeyl: yay... found the segfault :-) Jan 16 19:08:14 mrmoku: congrats! Jan 16 19:08:25 mickeyl: not sure if it actively participates in phonet forwarding. Jan 16 19:08:29 well Jan 16 19:08:34 it looks like it Jan 16 19:08:37 oh how nice :) Jan 16 19:08:37 1970-01-01T08:49:09.034729Z [INFO] NokiaIsiModem : Modem Status changed to FSO_GSM_MODEM_STATUS_ALIVE_SIM_LOCKED Jan 16 19:08:41 mrmoku: :D Jan 16 19:08:42 oh, and has anyone here managed to get a kernel newer than 2.6.29 working on a gta01? Jan 16 19:08:49 * mrmoku tries forwarding now Jan 16 19:09:46 mickeyl: should be easy to check: someone who has forwarding with maemo working can kill softupd and try again. Jan 16 19:10:03 PaulFertser: that mr. someone is me Jan 16 19:10:06 confirmed. Jan 16 19:10:11 hmm... no Jan 16 19:10:20 mrmoku: doesn't surprise me Jan 16 19:11:00 can someone with an strace binary strace it? Jan 16 19:11:19 or does anyone have an strace for maemo handy? Jan 16 19:13:26 and actually softupd is the _only_ binary in /usr/sbin that has the string upnlink0 embedded Jan 16 19:13:30 i guess that pretty much nails it Jan 16 19:16:15 lsof would be interesting too Jan 16 19:17:01 http://pastebin.com/HRqDeHau Jan 16 19:17:43 unfortunately it's not a static binary Jan 16 19:17:57 so we really need to find out what it does Jan 16 19:18:05 or rather... how it does what it does Jan 16 19:19:36 sre: oh and... for the records... getting the revision is indeed broken with the firmware that ships with PR <= 1.3 Jan 16 19:19:46 sre: it only started to work after i upgraded Jan 16 19:19:59 s/<=/ sre, hi Jan 16 19:21:07 1970-01-01T00:09:34.027221Z [WARN] NokiaIsiModem : Unhandled ISI SIMAuth.Status ISI_SIM_AUTH_STATUS_AUTHORIZED Jan 16 19:21:10 hmmm Jan 16 19:21:17 Is there any description about how you currently do that forwarding? Jan 16 19:21:28 PaulFertser: yes, here it comes: Jan 16 19:21:36 1.) connect your n900 to the PC Jan 16 19:21:42 2.) on the PC: ifconfig usbpn0 up Jan 16 19:21:52 3.) launch any client program Jan 16 19:22:00 1.5) select pc-suite mode Jan 16 19:22:22 mickeyl: ok Jan 16 19:22:30 freesmartphone.org: 03mok 07libisi * r4a5cf7420e74 10/isi/modem.c: Jan 16 19:22:30 freesmartphone.org: initialize client to NULL in isi_modem_create Jan 16 19:22:30 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 19:22:31 1.6)it will do some magic Jan 16 19:22:40 1.7)udev will be called Jan 16 19:22:44 sre: hope that is correct so :-) Jan 16 19:22:46 which will setup the interfaces Jan 16 19:22:47 etc... Jan 16 19:22:54 with the phonet binary Jan 16 19:23:16 mickeyl: hmm... I did not do the ifconfig usbpn0 up on the pc side... Jan 16 19:23:33 mrmoku: ok,i guess gisi does that for you then Jan 16 19:23:37 or libisi Jan 16 19:23:40 yeah, as it's up Jan 16 19:24:28 I'll fix the sim auth error too Jan 16 19:25:04 ya Jan 16 19:25:16 send it to READY as well Jan 16 19:25:29 i don't know why there are so many different states that mean the same Jan 16 19:25:34 need to investigate the differences Jan 16 19:25:36 there must be some Jan 16 19:25:50 1970-01-01T09:01:01.001648Z [ERROR] NokiaIsiModem : Can't open main channel; open returned false Jan 16 19:26:46 1970-01-01T00:16:46.748352Z [WARN] fsogsmd : GLib <>: Request failed: NET_CAUSE_COMMUNICATION_ERROR Jan 16 19:26:50 hmmm Jan 16 19:27:04 mickeyl, should I push my changes? Jan 16 19:27:13 let me see a diff first Jan 16 19:27:26 ok Jan 16 19:27:30 thanks for the review Jan 16 19:28:00 np, it's great to see people helping with fsogsmd. i feel we're making really good progress here Jan 16 19:28:19 unfortunately the weekend is almost over... Jan 16 19:28:43 1970-01-01T09:08:51.471650Z [DEBUG] fsogsmd : GLib <>: Phonet: phonet0 is up, idx=0x5 Jan 16 19:28:46 1970-01-01T09:08:51.472687Z [DEBUG] fsogsmd : GLib <>: Phonet: phonet0 is up, idx=0x5 Jan 16 19:28:49 1970-01-01T09:08:51.473297Z [INFO] fsogsmd : received signal -11, exiting. Jan 16 19:28:57 bummer, bt? Jan 16 19:29:11 not yet Jan 16 19:29:12 run it a couple of times Jan 16 19:29:16 there might be more race conditions Jan 16 19:29:19 i don't know enough about libisi Jan 16 19:29:24 http://pastebin.com/js3qQB4U Jan 16 19:29:26 btw. that reminds me I should have done some daywork this weekend too :P Jan 16 19:29:30 (which I did not) Jan 16 19:29:31 2 changes will go in 2 different commits Jan 16 19:29:36 same here... Jan 16 19:29:47 brb to get a bt Jan 16 19:29:52 te first is the Status.AUTHORIZED Jan 16 19:30:00 GNUtoo|laptop: that looks good. clear to go. Jan 16 19:30:04 ok Jan 16 19:30:13 I'll ident correctly and push and build+test Jan 16 19:30:18 ok Jan 16 19:30:28 test were minimal Jan 16 19:30:31 as it doesn't go far Jan 16 19:30:48 1970-01-01T00:16:46.748352Z [WARN] fsogsmd : GLib <>: Request failed: NET_CAUSE_COMMUNICATION_ERROR Jan 16 19:31:41 never seen that one Jan 16 19:31:47 but i don't have any valid SIM card atm. Jan 16 19:31:51 so i can only test without :/ Jan 16 19:45:36 how are alsa states supposed to be set? Jan 16 19:45:56 by odeviced or by some other means? Jan 16 19:49:46 polz: odeviced, yup Jan 16 19:50:28 freesmartphone.org: 03GNUtoo 07cornucopia * rf62b40a489e8 10/fsogsmd/src/plugins/modem_nokia_isi/mediators.vala: (log message trimmed) Jan 16 19:50:28 freesmartphone.org: fsogsmd: modem_nokia_isi: fix Unhandled ISI SIMAuth.Status ISI_SIM_AUTH_STATUS_AUTHORIZED Jan 16 19:50:28 freesmartphone.org: Without this commit we have the following warning: Jan 16 19:50:28 freesmartphone.org: 1970-01-01T00:09:34.027221Z [WARN] NokiaIsiModem : Unhandled ISI SIMAuth.Status ISI_SIM_AUTH_STATUS_AUTHORIZED Jan 16 19:50:29 freesmartphone.org: Which resulted in the modem staying in FSO_GSM_MODEM_STATUS_INITIALIZING state: Jan 16 19:50:29 freesmartphone.org: root@nokia900 ~ # mdbus2 -s org.freesmartphone.ogsmd /org/freesmartphone/GSM/Device org.freesmartphone.GSM.Network.GetStatus Jan 16 19:50:29 freesmartphone.org: org.freesmartphone.Unavailable: This function is not available while modem is in state FSO_GSM_MODEM_STATUS_INITIALIZING Jan 16 19:50:29 freesmartphone.org: 03GNUtoo 07cornucopia * r13244acdd760 10/fsogsmd/src/plugins/modem_nokia_isi/mediators.vala: Jan 16 19:50:30 freesmartphone.org: fsogsmd: modem_nokia_isi: implement NetworkGetSignalStrength Jan 16 19:50:30 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jan 16 19:50:40 mrmoku: and if I'm reading the source correctly, odeviced reads alsa states Jan 16 19:50:49 mrmoku: using alsactl -f Jan 16 19:50:55 or am I missing something here? Jan 16 19:51:49 polz: are you reading old python code? Jan 16 19:52:04 mrmoku: yep. Jan 16 19:52:19 polz: that's obsolete for the odeviced part Jan 16 19:52:22 mrmoku: that's what I've found on my phone and in the FSO GIT Jan 16 19:52:25 oh... Jan 16 19:52:37 FSO GIT has a cornucopia repo Jan 16 19:52:42 so I guess I should disable odeviced in frameworkd.conf? Jan 16 19:52:43 polz: cornucopia.git, look for fsodeviced Jan 16 19:52:55 like in the config for gta02 ? Jan 16 19:53:05 dos1: thanks. Jan 16 19:59:11 polz, you've got a second hand freerunner and the previous owner didn't update it because he didn't use it for a while? Jan 16 19:59:40 GNUtoo|laptop: no, I have an original neo1973 and I've taken it off the shell after about a year Jan 16 19:59:46 wow Jan 16 20:00:05 GNUtoo|laptop: I've applied the 1024 fix and I'd like to get it to work as a phone Jan 16 20:00:11 s/shell/shelF/ Jan 16 20:00:11 neo1973 lacks wifi but got clear sound, right? Jan 16 20:00:21 ok nice Jan 16 20:00:22 well, the sound isn't all that clear Jan 16 20:00:28 ah Jan 16 20:00:34 om-gta02 has a capacitor issue Jan 16 20:00:40 and it was useless until someone came up with the echo cancelation init string Jan 16 20:00:41 for audio headphones out Jan 16 20:00:49 ah ok Jan 16 20:00:57 oh, and the headphones were broken when I got it Jan 16 20:01:17 when I bought om-gta02 I knew it was not very usable as a phone, but I had a lot of fun wiht it Jan 16 20:02:07 I had a fix idea: porting wesnoth on it, I did it(with some help) and got commit access to openembedded Jan 16 20:02:22 at the end(I also ported some other games) Jan 16 20:02:28 when I bought the neo1973, I thought I would be able to throw away all the other phones. Too bad it's broken in so many ways :) Jan 16 20:02:51 I didn't know neo1973 was also broken Jan 16 20:03:09 I tought there were not the same flaws than om-gta02 Jan 16 20:03:14 well, the 1024 took a while to fix Jan 16 20:03:19 it has no wifi Jan 16 20:03:24 it's horribly slow Jan 16 20:03:27 1024 is recamping Jan 16 20:03:30 ah right Jan 16 20:03:33 200Mhz Jan 16 20:03:36 yep. Jan 16 20:04:00 but it might still be useful if the software worked Jan 16 20:04:08 indeed Jan 16 20:04:11 it never has :) Jan 16 20:04:15 you can do many things with it Jan 16 20:04:18 not only phone Jan 16 20:04:36 polz, it's not that stable nowadays but it's a lot nicer Jan 16 20:04:46 yeah, I might use it for something someday Jan 16 20:04:52 that is to say we just got a gdbus migration, things should work tough Jan 16 20:05:00 but a lot of stuff changes Jan 16 20:05:00 but first of all, I'll have to change the battery Jan 16 20:05:03 *changed, Jan 16 20:05:08 a lot of stuff got fixed Jan 16 20:05:11 it became usable Jan 16 20:05:22 if you don't opkg update;opkg upgrade all the time Jan 16 20:05:23 yeah, the GTA02 became usable :) Jan 16 20:05:27 yes Jan 16 20:05:32 or if you use shr-testing Jan 16 20:05:40 well... the latest kernel that works for me is 2.6.29 Jan 16 20:05:50 oh, and I don't have a debug board Jan 16 20:05:53 ah ok Jan 16 20:06:08 I wanted to ask exactly that: to check the serial with the debug board Jan 16 20:06:15 mickeyl: http://repository.maemo.org/pool/fremantle/free/s/strace/strace_4.5.18-1.1+0m5_armel.deb Jan 16 20:06:17 *serial console Jan 16 20:06:46 GNUtoo|laptop: I might do that in the next year :) Jan 16 20:06:54 ok Jan 16 20:07:20 GNUtoo|laptop: are you still using your gta02? Jan 16 20:07:28 I keep it updated Jan 16 20:07:31 hi, is om-gta02 usable as phone with last content in feeds? Jan 16 20:07:46 basically it's my reference platform Jan 16 20:07:55 sometimes I use it Jan 16 20:07:58 but not all the time Jan 16 20:08:00 GNUtoo|laptop: for what? Jan 16 20:08:06 just to use it Jan 16 20:08:08 phone Jan 16 20:08:12 surfing Jan 16 20:08:13 etc... Jan 16 20:08:19 irc Jan 16 20:08:26 I check from time to time the status Jan 16 20:08:33 how fast it is etc... Jan 16 20:09:06 are you developing some other phone? Jan 16 20:09:08 maybe I could convert it in a VOIP phone once VOPI would work Jan 16 20:09:10 yes Jan 16 20:09:21 I did some porting to other phones Jan 16 20:09:23 porting SHR Jan 16 20:09:34 htcdream and now nokia n900 Jan 16 20:09:39 but I've less time nowadays Jan 16 20:09:43 so I do less work Jan 16 20:09:44 but still Jan 16 20:09:52 I do some Jan 16 20:10:27 so neo1973 has the bass issue for music? Jan 16 20:10:35 I don't know Jan 16 20:10:39 ok Jan 16 20:10:44 I guess not... Jan 16 20:10:49 but I'm not sure :) Jan 16 20:10:51 om-gta02 has horrible sound on headphones Jan 16 20:10:56 and it's not the headphones Jan 16 20:11:03 it's a capacitor Jan 16 20:11:12 oh... the sound here is OK Jan 16 20:11:13 that has a value that is too low Jan 16 20:11:22 but it does it only for headphones Jan 16 20:11:33 as that capacitor is connected to the headphones output Jan 16 20:11:42 while the headphones still haf-worked, the sound was OK Jan 16 20:11:48 ok Jan 16 20:12:21 how do I check which odeviced implementation is being used? Jan 16 20:12:45 pidof fsodeviced Jan 16 20:13:06 OK, fsodeviced is running Jan 16 20:13:14 basically the new daemons are written in vala, it's a java-like language but it's compiled into C Jan 16 20:13:19 which is compiled into machine language Jan 16 20:13:23 so it's faster Jan 16 20:13:38 and much more difficult do debug Jan 16 20:13:47 not for me Jan 16 20:13:58 how so? Jan 16 20:14:09 because I'm used to gdb+gdbserver Jan 16 20:14:44 yeah, but if you want to try out a fix, you still have to first set up the whole build system Jan 16 20:14:48 and as I build the images for some of my devices Jan 16 20:14:50 rebuild the executable Jan 16 20:14:53 upload it Jan 16 20:14:54 and that I know badly vala Jan 16 20:14:57 it's a PITA Jan 16 20:14:57 etc... Jan 16 20:15:04 yes Jan 16 20:15:19 but makes sense if the stuff is supposed to be working :) Jan 16 20:15:19 but don't you prefer speed on neo1973 Jan 16 20:15:20 ? Jan 16 20:15:30 it is Jan 16 20:15:36 I'd rather have working sound at the moment :) Jan 16 20:17:49 mrmoku, hi Jan 16 20:17:59 mrmoku, what was the fix for phonefsod not starting? Jan 16 20:18:13 I've that on n900 but not on htcdream(built later) Jan 16 20:18:13 GNUtoo|laptop: it starts ... Jan 16 20:18:27 oh, right, it's one of _your_ problems :) Jan 16 20:18:30 s/built/rebuilt from scratch/ Jan 16 20:18:30 GNUtoo|laptop meant: I've that on n900 but not on htcdream(rebuilt from scratch later) Jan 16 20:18:50 yes in order to develop I've to build my own images Jan 16 20:18:53 and I like that Jan 16 20:19:30 GNUtoo|laptop: hmmm... then something is still wrong Jan 16 20:19:41 mrmoku, it's on nokia n900 Jan 16 20:19:47 mrmoku, where are the logs for that? Jan 16 20:19:53 /var/log/phonefsod Jan 16 20:20:01 yup Jan 16 20:20:26 http://pastebin.com/Sa28Hsbp Jan 16 20:21:09 we should push new config for nokia n900 Jan 16 20:21:12 for fsogsmd Jan 16 20:21:20 for the modem_access Jan 16 20:21:27 GNUtoo|laptop: well that looks like fsogsmd is not starting Jan 16 20:21:33 and yes for the config Jan 16 20:21:39 yes because of modem_access maybe Jan 16 20:23:24 * pespin adding little UI for fso alsa scenario support in shr-settings Jan 16 20:23:31 [SIGNAL] org.freesmartphone.Device.Audio.Scenario( s:scenario, s:reason ) Jan 16 20:23:42 mrmoku, now it started Jan 16 20:23:49 good Jan 16 20:23:52 mickeyl, permission to push modem_access? Jan 16 20:23:55 is this signal produced when someone changes scenario? Jan 16 20:27:26 as it's a fix I'll get the error message and push Jan 16 20:27:44 anyway mrmoku and you have overriden config on hosts Jan 16 20:27:53 like in /etc/fsodeviced.conf Jan 16 20:28:51 funny the error is: Jan 16 20:28:56 1970-01-01T00:01:03.089233Z [CRITICAL] fsogsmd : GLib <>: string_to_string: assertion `self != NULL' failed Jan 16 20:28:58 lol Jan 16 20:30:19 mrmoku, I push? Jan 16 20:30:45 glib_timeout_add_seconds(10,push_callback); Jan 16 20:31:21 what's the format of these state files? Jan 16 20:31:23 I mean... Jan 16 20:31:46 is the first number the number of the alsa control? Jan 16 20:33:13 freesmartphone.org: 03GNUtoo 07cornucopia * rd771d8b4902f 10/fsogsmd/conf/nokia_n900/fsogsmd.conf: Jan 16 20:33:13 freesmartphone.org: fsogsmd: nokia_n900 configuration: add modem_transport Jan 16 20:33:13 freesmartphone.org: Without that fix fsogsmd can't start on the nokia n900: Jan 16 20:33:13 freesmartphone.org: 1970-01-01T00:01:03.088317Z [DEBUG] LowLevelNokia900 <>: gpio_probe: rapu is LOW_LEVEL_NOKIA900_RAPU_TYPE_TYPE_1 Jan 16 20:33:13 freesmartphone.org: 1970-01-01T00:01:03.089233Z [CRITICAL] fsogsmd : GLib <>: string_to_string: assertion `self != NULL' failed Jan 16 20:33:13 freesmartphone.org: 1970-01-01T00:01:03.495086Z [INFO] fsogsmd : Binary launched successful (FsoFrameworkFileLogger created as theLogger) Jan 16 20:33:14 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jan 16 20:38:36 btw what's the status of the powercontrl plugins? Jan 16 20:38:44 lile the wifi, bt etc... Jan 16 20:39:01 is there a new method to follow now? Jan 16 20:39:23 GNUtoo|laptop: that's another thing missing in fsodeviced.conf Jan 16 20:39:40 [fsodevice.n900_powercontrol] Jan 16 20:39:40 [fsodevice.powersupply_n900] Jan 16 20:39:42 I was thinking of fsodeviced not the config Jan 16 20:39:44 is what I added Jan 16 20:39:48 ok Jan 16 20:39:54 and then I can request WiFi and wlan0 comes up Jan 16 20:39:54 there is the battery charging too Jan 16 20:39:58 yup Jan 16 20:40:09 I'll re-explain Jan 16 20:40:12 there are 2 issues Jan 16 20:40:15 the config Jan 16 20:40:17 and the code Jan 16 20:40:25 I was talking aobut the code Jan 16 20:40:35 (and wanted to talk about config too but later Jan 16 20:41:03 the code is that fsodeviced refuse to work when more than one thing handle the resource Jan 16 20:41:05 example: Jan 16 20:41:08 *rfkill + Jan 16 20:41:14 *ifconfig + Jan 16 20:41:23 *modprobe Jan 16 20:41:29 all handling one resource Jan 16 20:42:00 that is still an open issu Jan 16 20:42:00 e Jan 16 20:42:14 ok Jan 16 20:42:18 then I'll wait Jan 16 20:43:40 maybe I should do more modem work? Jan 16 20:43:58 very welcome :) Jan 16 20:44:00 SHR: 03pespin.shar 07shr-settings * ra3cdb92ee824 10/ (shr-settings shr_settings_modules/shr_alsa_scenarios.py): Add support for manual alsa scenario switch in shr_alsa_scenarios.py Jan 16 20:44:13 could you push the configs and warn about that on mailing lists? Jan 16 20:44:40 which configs? Jan 16 20:44:52 [fsodevice.n900_powercontrol] Jan 16 20:44:52 [fsodevice.powersupply_n900] Jan 16 20:44:56 ahh, ok Jan 16 20:44:58 the one for battery charging Jan 16 20:45:01 and the one for wifi Jan 16 20:45:16 the warning I sent some day ago... and I thought it was already enabled Jan 16 20:45:18 for alsa-scenarios.... Jan 16 20:45:29 it's not else I would have seen it Jan 16 20:45:50 alsa-scenarios I pushed earlier... but probably forgot to add it to the Makefile Jan 16 20:46:12 no Jan 16 20:46:15 wait a sec Jan 16 20:46:23 I didn't finish what I wanted to say Jan 16 20:46:37 for alsa scenario....you had 2.6.28 scenario from maemo Jan 16 20:46:44 could you add them too while you're at it? Jan 16 20:47:39 ugh I'm looking now at shr-settings recipe, and I thought it was in AUTOREV but now I see that the rev in OE is quite old Jan 16 20:48:12 pespin, ah it's not autorev? Jan 16 20:48:18 why did it rebuilt it for me then? Jan 16 20:48:31 NOTE: package shr-settings-1_0.1.1+gitr1+a3cdb92ee8246648cc5d0de949afc7f1284f3f0e-r12: task do_package_write_ipk: Started Jan 16 20:48:47 GNUtoo|laptop, is this one right? http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/shr/shr-settings_git.bb Jan 16 20:48:50 pespin, autorev must not be set in the recipe itself Jan 16 20:48:59 pespin, it should be in shr-autorev.inc Jan 16 20:49:03 aah ok Jan 16 20:49:05 in the distro/include dir Jan 16 20:49:08 GNUtoo|laptop, but what's this then? SRCREV = "e94cb974d08a242c8def89aca43da00124080368" Jan 16 20:49:22 the revision for people not using autorev Jan 16 20:49:32 like angstrom Jan 16 20:49:47 or shr built without including the fso-autorev and shr-autorev in local.conf Jan 16 20:50:33 lol OE is really complex Jan 16 20:50:44 hmmm Jan 16 20:50:53 GNUtoo|laptop: I only have those... did not adjust them for the new kernel Jan 16 20:51:05 ah ok Jan 16 20:51:07 then I'll try Jan 16 20:51:15 freesmartphone.org: 03mok 07cornucopia * rfa5306cb821e 10/fsodeviced/conf/nokia_n900/Makefile.am: Jan 16 20:51:15 freesmartphone.org: fsodeviced: actually install alsa-scenarios for nokia900 Jan 16 20:51:16 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 20:51:31 pespin, you found a rare case where there is hidden stuff but usually it's pretty clear Jan 16 20:51:47 * GNUtoo|laptop rebuilds again Jan 16 20:51:57 I keep rebuilding without interuption.... Jan 16 20:52:03 :) Jan 16 20:52:28 I'm looking at distro/shr.conf, I've never read it :) Jan 16 20:52:34 ok Jan 16 20:52:37 GNUtoo|laptop: me too :P Jan 16 20:52:49 lol ok Jan 16 20:53:02 pespin, why not reading machines configs Jan 16 20:53:06 that is more interesting Jan 16 20:53:15 so you can add your other phone Jan 16 20:53:23 the wing-linux supported phone Jan 16 20:53:25 right? Jan 16 20:53:43 GNUtoo|laptop, I have some configs for it already :) Jan 16 20:53:52 but indeed I have to look more into it. Jan 16 20:54:40 ok Jan 16 20:54:45 as I read they moved some stuff upstream and then they are using another repo. It has been a while since I built that kernel. Jan 16 20:54:55 ok Jan 16 20:54:56 but not now, this week I have exams :P Jan 16 20:55:18 next week I'll rebuilt an image for it Jan 16 20:55:36 | make[4]: Entering directory `/home/gnutoo/embedded/oe/oetmps/nokia900/work/armv7a-oe-linux-gnueabi/fsodeviced-2_0.9.4+gitr15+fa5306cb821ea285f3279c5a8207edcbeb1eaf96-r4.20/git/fsodeviced/conf/nokia_n900/alsa-default' =>| make[4]: *** No rule to make target `all'. Stop. Jan 16 20:55:38 I'll retry Jan 16 20:55:48 ok Jan 16 20:57:33 btw what's the current limit in mA for the usb host on nokia900 and on om-gta02 Jan 16 20:58:42 hmm Jan 16 20:58:54 mrmoku, maybe you forgott to include the Makefile Jan 16 20:59:01 the generated one Jan 16 20:59:04 it must be called Jan 16 20:59:29 sorry Jan 16 20:59:33 it's rather the countrary Jan 16 20:59:44 you may have forgetten to generate the Makefile Jan 16 20:59:58 isn't that autotools job? Jan 16 21:00:05 yes it is Jan 16 21:00:12 I'll look Jan 16 21:00:31 I forgot to add a Makefile to the subdir Jan 16 21:00:32 :/ Jan 16 21:00:46 ok Jan 16 21:02:02 example: 709bcd36cfa33b144ca8bf4c190dde7f15f21bab Jan 16 21:02:10 it's a new fsogsmd plugin Jan 16 21:02:24 and it has the bits necessary to add a new plugin in the build system Jan 16 21:03:19 Makefile? Jan 16 21:03:35 autotools Jan 16 21:03:45 hi dos1 Jan 16 21:03:52 mrmoku: lemme check, i have already added new plugins to fsodeviced :P Jan 16 21:03:54 hi ;) Jan 16 21:04:05 freesmartphone.org: 03mok 07cornucopia * r9de30ab3782c 10/fsodeviced/conf/nokia_n900/ (Makefile.am alsa-default/Makefile.am): Jan 16 21:04:05 freesmartphone.org: fsodeviced: fix Makefiles for n900 scenario installation Jan 16 21:04:05 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 21:04:07 GNUtoo|laptop: hope it works now Jan 16 21:04:17 ok,I'll rebuild Jan 16 21:04:35 dos1: too late :P Jan 16 21:04:39 (hopefully) Jan 16 21:04:46 mrmoku, if it doesn't look at my commit hash Jan 16 21:05:43 GNUtoo|laptop: it's not a new plugin... and I looked it what is done for the gta02 scenarios Jan 16 21:05:46 must work ;) Jan 16 21:06:01 ah ok Jan 16 21:08:21 mrmoku, I'll clean to retry Jan 16 21:08:26 but it seem that it didn't work Jan 16 21:08:32 :( Jan 16 21:08:45 | make[4]: Entering directory `/home/gnutoo/embedded/oe/oetmps/nokia900/work/armv7a-oe-linux-gnueabi/fsodeviced-2_0.9.4+gitr16+9de30ab3782cc4ecd533dec5de34e3e7331bda30-r4.20/git/fsodeviced/conf/nokia_n900/alsa-default'=>| make[4]: *** No rule to make target `all'. Stop. Jan 16 21:09:00 it seem the same revision than your last commit Jan 16 21:11:17 freesmartphone.org: 03mok 07cornucopia * rbcd4005f4b9c 10/fsodeviced/configure.ac: Jan 16 21:11:17 freesmartphone.org: fsodeviced: and actually make autoconf output the Makefile Jan 16 21:11:17 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jan 16 21:11:18 GNUtoo|laptop: next try :-) Jan 16 21:16:47 NOTE: package fsodeviced-2_0.9.4+gitr17+bcd4005f4b9c3c633fe44e683fa80f0764f6c7b8-r4.20: task do_install: Started Jan 16 21:16:54 PaulFertser: great, thanks Jan 16 21:17:28 mickeyl: libisi from the freesmartphone repo doesn't check for vala and so tests can't be built :) Jan 16 21:19:03 mickeyl: oh, my bad, me stupid Jan 16 21:19:03 sorry Jan 16 21:21:01 aha Jan 16 21:21:13 strace reveals that softupd is responsible for the forwarding Jan 16 21:21:27 so now we just need to find out how to replicate its behaviour Jan 16 21:21:48 i guess it's more or less just reading on one interface and writing to the other one Jan 16 21:21:49 bbl Jan 16 21:35:14 hi, is bluetooth headset working in SHR for phone calls? Jan 16 21:35:30 radekp: somebody reported it does not Jan 16 21:35:34 never tried myself though Jan 16 21:35:41 missing such a thing as a bt headset Jan 16 21:36:25 hmm i had it working in qtmoko with bluez3, but now i ported qtmoko bt stack to bluez4, but i got no sounds in handsfree profile Jan 16 21:36:52 so i wonder if anyone has it working with bluez4 Jan 16 21:37:54 GNUtoo|laptop: do you know if bt headset works with bluez4 on gta02? Jan 16 21:38:32 mrmoku, I've only bluetooth: Jan 16 21:38:34 *phones Jan 16 21:38:36 *keyboard Jan 16 21:38:37 *mouse Jan 16 21:38:43 not headsets Jan 16 21:38:46 ok :) Jan 16 21:38:59 I'm not sure if we can emulate a headset Jan 16 21:39:07 btw thanks a lot Jan 16 21:39:14 audio routing works fine Jan 16 21:39:21 great Jan 16 21:46:30 btw how's om-gta01 support in SHR? Jan 16 21:46:44 is framebuffer supported? Jan 16 21:46:49 good question Jan 16 21:46:53 is GPS supported? (I hope not) Jan 16 21:47:03 :P Jan 16 21:47:11 it was at least Jan 16 21:49:46 radekp: hm, it used to work Jan 16 21:49:57 radekp: but ophoned and some other fso-specific stuff made it unstable. Jan 16 21:50:08 radekp: other than that, yes, bluez4 worked with SCO headsets. Jan 16 21:51:04 radekp: does your bluetoothd config not lack SCORouting=true option? Jan 16 21:51:14 (or something like that) Jan 16 21:51:17 PaulFertser: do you know if there has something changed from bluez3->bluez4 Jan 16 21:51:28 radekp: first thing: check that option Jan 16 21:51:37 PaulFertser: oki Jan 16 21:51:51 radekp: i do not know bluez3 api at all, but i can help with bluez4 :) Jan 16 21:51:59 PaulFertser: great Jan 16 21:52:03 radekp: iirc there was simple Play() call you're to make. Jan 16 21:52:28 i thought bluez4 retain libc api Jan 16 21:52:32 PaulFertser: btw i have working buttons on the headset, i can hear ringing, but just no sound Jan 16 21:52:33 Having some trouble building SHR for the iPhone3G. Pidgin won't compile properly. I tried cleaning and rebuilding separate, but no luck. I'm not really sure what to do. Anyone mind taking a look at my pastebin? http://pastebin.com/zxDuWJtS Jan 16 21:52:35 radekp: btw, with bluez4 quite some calls activate something for as long as the process that requested it keeps dbus bus open. Jan 16 21:52:41 damn, s/libc/C/ Jan 16 21:52:56 gena2x: hm, no idea about bluez C api, i never cared :) Jan 16 21:53:11 gena2x: btw, wrt bootloader. I've just woken up and wasn't able to discuss it. Jan 16 21:53:50 PaulFertser: it will take some time to test - i dont have working SD card, so reflashing now... :( Jan 16 21:53:54 i can recall things like 'something working only wutg C api' Jan 16 21:54:03 s/wutg/with/ Jan 16 21:54:04 gena2x meant: i can recall things like 'something working only with C api' Jan 16 21:54:36 but may be it were related to digital interface Jan 16 21:54:52 radekp: SCORouting=PCM in [General] section. Jan 16 21:54:55 PaulFertser: the bootloader subject has no hurry :) Jan 16 21:55:16 it seems that qtopicomm module uses Dbus just for pairing and SDP queries, rest is done on rfcomm somehow, but i am really new to that code Jan 16 21:55:46 PaulFertser: yup, i wanted to try it, but right after i have started flashing... Jan 16 21:55:49 PaulFertser: (bootloader) as you can notice, i answered a week after our inital talk :) Jan 16 21:58:36 gena2x: when i was a kid i had a vynil disk with a short tale about trolls, the last phrase there was "i hate you chatters" Jan 16 22:00:54 PaulFertser: unsure if i got your idea, but i didn't start this topic Jan 16 22:01:28 gena2x: http://tombtroll.narod.ru/stih1.htm Jan 16 22:03:17 PaulFertser: hehe Jan 16 22:11:09 but i didn't want to interrupt bluez discussion :) Jan 16 22:18:17 Lopi,hmmm Jan 16 22:18:28 oops Jan 16 22:18:31 lol Jan 16 22:18:59 ?? Jan 16 22:19:55 mrmoku: (forwarding), hm, looks like it's not that easy, it seems sysupdate just communicates with another proprietary shit -- sscd Jan 16 22:20:10 :/ Jan 16 22:20:13 mrmoku: luckily there's ofono that can work natively on n900. Jan 16 22:20:37 so we should take a look how it's done in meego Jan 16 22:20:54 or is it done by ofono standalone? Jan 16 22:21:38 what's happening? ofono natively? Jan 16 22:21:56 i'll look at the logs Jan 16 22:22:02 23:15 < PaulFertser> mrmoku: (forwarding), hm, looks like it's not that easy, it seems sysupdate just communicates with another proprietary shit -- sscd Jan 16 22:23:27 mrmoku: anyway, reversing sysupdate and friends is not the way to go here because there's ofono working and it has nice sources. Jan 16 22:23:36 ah ok forwarding Jan 16 22:23:41 thanks Jan 16 22:24:37 but it could be easy? Jan 16 22:24:43 tcpdump Jan 16 22:24:51 on both interfaces Jan 16 22:25:00 and look what changed Jan 16 22:25:35 Lopi, you seem stuck Jan 16 22:25:41 PaulFertser: are we sure ofono is setting forwarding up correctly? we should try that Jan 16 22:25:42 Lopi, try oe ml Jan 16 22:25:42 GNUtoo: well... Apparently ofono folks needed some configuring via netlink for whatever reason. Jan 16 22:25:48 ok Jan 16 22:25:53 mrmoku: it doesn't use forwarding, it runs on device itself. Jan 16 22:26:02 sure Jan 16 22:26:21 GNUtoo: ml? Jan 16 22:26:36 mailing list Jan 16 22:26:56 PaulFertser: I mean we should try if for example when running meego the libisi tests work from the laptop Jan 16 22:27:43 mrmoku: why would they? phonet is p2p link, there must be some process that forwards the datagrams on n900, or else forwarding can't work. Jan 16 22:28:34 PaulFertser: they do work when I'm running maemo on the n900 Jan 16 22:30:39 PaulFertser: cool ot works now Jan 16 22:30:59 PaulFertser: SCORouting=PCM did the trick Jan 16 22:31:04 mrmoku: because it's sysupdate forwarding. I'm currently checking tcpdump Jan 16 22:31:07 radekp: :) Jan 16 22:31:21 PaulFertser: qtmoko is now ported to bluez4 :) Jan 16 22:31:34 PaulFertser: but then just looking at ofono code won't help Jan 16 22:31:46 brb Jan 16 22:35:26 radekp: cool :) Jan 16 22:36:03 i would never believe i can do it :) Jan 16 22:36:33 it took 10 days, not so bad... Jan 16 22:38:03 hmm Jan 16 22:38:12 i think it's relatively easy to get forwarding running Jan 16 22:38:22 at least it looks so: Jan 16 22:38:25 answer: Jan 16 22:38:36 read(5, "\25\20\0\33\0\26P\31\346\26\0\1\1\20\0\vNokia N900\0\0", 65541) = 28 Jan 16 22:38:37 write(3, "\20\0\33\0\26P\31\346\26\0\1\1\20\0\vNokia N900\0\0", 27) = 27 Jan 16 22:38:39 question: Jan 16 22:38:51 recv(3, "\0\20\33\0\5\0P\344\25\7", 49151, 0) = 10 Jan 16 22:38:51 write(5, "\25\0\20\33\0\5\0P\344\25\7", 11) = 11 Jan 16 22:39:06 could be exactly the same datagrams, just with source/destination address reversed Jan 16 22:40:27 and wrt. ofono, i don't think anyone used forwarding on a non-maemo system Jan 16 22:40:40 i'm pretty sure they have implemented it with softupd doing the forwarding Jan 16 22:40:54 since according to MH forwarding is just "automatic" Jan 16 22:42:48 mickeyl: yes Jan 16 22:43:09 mickeyl: i think i'll try to write the simplest forwarder now Jan 16 22:43:19 wow, awesome, thanks a lot for looking into this Jan 16 22:43:35 mickeyl: btw, it's strange nobody's noticed that phonet is strictly peer-to-peer and hence requires a forwarder. Jan 16 22:43:53 yeah Jan 16 22:49:13 PaulFertser: great Jan 16 22:52:29 * mrmoku heading to bed now Jan 16 22:52:35 happy forwarding and a good night :-) Jan 16 22:54:35 same here, g'night Jan 16 23:20:45 libfsobasic is failing to build: Jan 16 23:21:09 freesmartphone.org: 03mickey 07cornucopia * r70507074ac0d 10/fsogsmd/src/ (lib/modem.vala plugins/modem_nokia_isi/channel.vala): fsogsmd: really secure against missing modem_access configuration Jan 16 23:21:10 NOTE: package libfsobasics-1_0.9.10+gitr2+9e5ec26d3ceb6aa33242befd9af7322d7d66a8ab-r4.1: task do_compile: Failed Jan 16 23:21:15 mickey|zzZZzz, ^ Jan 16 23:22:12 yes, we need to bump it Jan 16 23:22:23 if we are going to use vala 0.11.4.1 Jan 16 23:23:19 feel free to include fso-autorev.inc in your local conf until then Jan 16 23:23:50 ok Jan 16 23:46:44 freesmartphone.org: 03mickey 07cornucopia * ra7ee7876453c 10/libfsobasics/fsobasics/ (libinitexit.vala logger.vala): libfsobasics: refactor glib logger and fix level switch Jan 16 23:46:56 mickey|zzZZzz, added fso-autorev and now fails this: Jan 16 23:46:57 NOTE: package fso-abyss-1_0.9.2+gitr3+70507074ac0df611d27f7a93472da7ec5e7b4388-r4.0: task do_compile: Failed Jan 16 23:47:22 hmm, who's building abyss these days? Jan 16 23:47:27 no idea Jan 16 23:47:42 just building shr-lite-image Jan 16 23:47:45 didn't look at that since long Jan 16 23:47:48 nothing should depend on it Jan 16 23:48:15 mickey|zzZZzz, maybe something in fso-autorev is adding it? Jan 16 23:48:45 no, fso-autorev is only specifying versions Jan 16 23:50:30 something in vala seems to have broken abyss Jan 16 23:50:53 I'll take that up to #vala Jan 16 23:50:57 but not tonight Jan 16 23:50:57 cu Jan 16 23:51:41 http://paste.pocoo.org/show/321842/ Jan 16 23:51:53 here's the log, save it if you wanna take a look tomorrow :) Jan 16 23:51:55 gningh! Jan 17 00:12:17 Can't quite figure how to use that fucking phonet yet, off to sleep now, continue tomorrow. **** ENDING LOGGING AT Mon Jan 17 02:59:57 2011