**** BEGIN LOGGING AT Tue Jun 29 02:59:57 2010
Jun 29 03:09:43 <padovan> denkenz: where in the Makefile.am you set gatchat as a lib? I can't find that to reproduce on bluetooth.c
Jun 29 07:42:52 <kvalo> oh, this is _so_ nice. holtmann's disconnect hack works just great with huawei. no more unplugging the modem during test \o/
Jun 29 12:05:25 <holtmann> ofonod[13480]: App:< \r\n+STUD: 13,""\r\n
Jun 29 12:05:26 <holtmann> ofonod[13480]: App:< \r\n_OSIGQ: 18,0\r\n
Jun 29 12:05:27 <holtmann> ofonod[13480]: App:< \r\n+STC: 2\r\n
Jun 29 12:05:27 <holtmann> ofonod[13480]: App:< \r\n_OSIGQ: 20,0\r\n
Jun 29 12:06:00 <holtmann> denkenz: Saw that on my old Option card now. Any idea waht STUD or STC is?
Jun 29 12:54:14 <kvalo> denkenz: ofono crashed in g_at_ppp_unref(). will send backtrace soon
Jun 29 14:04:51 <pessi> holtmann: sim toolkit stuff, stud 13 indicates satk sent sms ?
Jun 29 14:08:03 <holtmann> pessi: That is on an HSO modem. We never enabled SIM Toolkit there. And why does it send an SMS.
Jun 29 14:08:14 <holtmann> Do you happen to have the proper documentation for this.
Jun 29 14:15:53 <pessi> nope, sorry. I have docs for some chinese modem which supports +stud
Jun 29 14:29:56 <holtmann> Hah.
Jun 29 14:30:15 <holtmann> Seriously, we never enabled this with the HSO one, but in the end it is a Qualcomm modem.
Jun 29 14:30:31 <holtmann> So maybe these things are on by default.
Jun 29 14:52:11 <holtmann> kvalo: So I really want a fix for that crash. Otherwise I move the target date for the release to tomorrow. I am not getting to it today.
Jun 29 15:03:27 <kvalo> holtmann: sorry, I don't have time to work on ofono this week at all. I need to finish another project before I leave for holidays on friday
Jun 29 15:04:40 <denkenz> kvalo: I don't get how yu trigger your crash
Jun 29 15:05:17 <kvalo> denkenz: I did connect "saunalahti", disconnect, connect, disconnect...
Jun 29 15:05:34 <denkenz> kvalo: were you removing the dongle during activate / deactivate?
Jun 29 15:06:07 <kvalo> denkenz: ofono crashed twice for me today, but I don't know if the second crash was similar as the one I reported
Jun 29 15:06:30 <kvalo> denkenz: no dongle removals. I used the huawei hack patch I posted this morning.
Jun 29 15:08:10 <denkenz> holtmann: We really need timing info
Jun 29 15:10:40 <denkenz> kvalo: try this
Jun 29 15:10:41 <denkenz> -       if (gcd->state != STATE_IDLE) {
Jun 29 15:10:43 <denkenz> +       if (gcd->state != STATE_IDLE && gcd->ppp != NULL) {
Jun 29 15:11:42 <denkenz> looks like you're connecting very close to a disconnect, so the modem sends a HUP
Jun 29 15:12:21 <denkenz> Your GPRS context will get removed and the core will never return, but at least it won't crash
Jun 29 15:15:53 <denkenz> So unfortunately holtmann's hack won't work too well in that case
Jun 29 15:24:07 <kvalo> denkenz: I can send you my syslog, it has seconds
Jun 29 15:27:32 <denkenz> ofonod[2365]: drivers/atmodem/gprs-context.c:ppp_disconnect() 
Jun 29 15:27:33 <denkenz> ofonod[2365]: Pcui:< \r\n^BOOT:38645652,0,0,0,87\r\n
Jun 29 15:27:35 <denkenz> ofonod[2365]: Modem:> AT+CGDCONT=1,"IP","internet"\r
Jun 29 15:27:36 <denkenz> ofonod[2365]: plugins/huawei.c:huawei_disconnect() 
Jun 29 15:27:38 <denkenz> ofonod[2365]: src/gprs.c:gprs_context_remove() atom: 0x6e4460
Jun 29 15:27:39 <denkenz> ofonod[2365]: drivers/atmodem/gprs-context.c:at_gprs_context_remove() 
Jun 29 15:27:51 <denkenz> If the difference between disconnect & remove is like < 1 second, then my hypothesis should be correct
Jun 29 15:28:43 <kvalo> denkenz: I'll find the relevant part from syslog
Jun 29 15:32:19 <kvalo> Jun 29 15:52:04 tukki ofonod[2365]: Pcui:< \r\n^RSSI:24\r\n
Jun 29 15:32:19 <kvalo> Jun 29 15:52:07 tukki ofonod[2365]: Pcui:< \r\n^RSSI:24\r\n
Jun 29 15:32:20 <kvalo> Jun 29 15:52:09 tukki ofonod[2365]: drivers/atmodem/gprs-context.c:at_gprs_deactivate_primary() 
Jun 29 15:32:23 <kvalo> Jun 29 15:52:09 tukki ofonod[2365]: drivers/atmodem/gprs-context.c:ppp_disconnect() 
Jun 29 15:32:26 <kvalo> Jun 29 15:52:09 tukki ofonod[2365]: Pcui:< \r\n^BOOT:38645652,0,0,0,87\r\n
Jun 29 15:32:28 <kvalo> Jun 29 15:52:11 tukki ofonod[2365]: Modem:> AT+CGDCONT=1,"IP","internet"\r
Jun 29 15:32:31 <kvalo> Jun 29 15:52:12 tukki ofonod[2365]: plugins/huawei.c:huawei_disconnect() 
Jun 29 15:32:33 <kvalo> Jun 29 15:52:12 tukki ofonod[2365]: src/gprs.c:gprs_context_remove() atom: 0x6e4460
Jun 29 15:32:36 <kvalo> Jun 29 15:52:12 tukki ofonod[2365]: drivers/atmodem/gprs-context.c:at_gprs_context_remove() 
Jun 29 15:32:39 <kvalo> Jun 29 15:56:42 tukki ofonod[2410]: oFono version 0.23
Jun 29 15:32:42 <kvalo> denkenz: ^
Jun 29 15:33:06 <denkenz> yeah, ok, so that's it
Jun 29 15:33:35 <denkenz> the ppp driver needs some more work, it needs to wait for at least NO CARRIER 
Jun 29 16:43:39 <denkenz> Our PPP stack is too fast :(
Jun 29 16:50:40 <kvalo> denkenz: how so?
Jun 29 16:51:21 <denkenz> oh I'm just adding server-side extensions, and we need to run a DHCP query for the peer address
Jun 29 16:51:39 <denkenz> I manage to exhaust the max-failure counter in about .2 seconds
Jun 29 17:36:50 <pessi> is supervisory tone generation out of ofono scope?
Jun 29 17:37:30 <denkenz> what exactly is it?
Jun 29 17:38:24 <pessi> like, ringing tone when the remote end is alerting
Jun 29 17:38:49 <pessi> or call waiting tone
Jun 29 17:39:29 <pessi> or different error tones when called number is busy, there is congestion or a special error occurs
Jun 29 17:39:40 <pessi> 22.001 annex F
Jun 29 17:39:53 <pessi> we have lot of code handling that in tp-ring
Jun 29 17:40:32 <pessi> I wonder if that should be ditched or moved to somewhere else
Jun 29 17:41:32 <pessi> and speaking of 22.001, is there something in 27.007 overriding annex E (aka Automatic calling repeat call attempt restrictions
Jun 29 17:42:16 <pessi> or Blacklisting)
Jun 29 17:45:45 <denkenz> You mean CCBS?
Jun 29 17:45:51 <denkenz> Nothing about CCBS in 27.007
Jun 29 17:46:03 <denkenz> And supervisory tones sound like a multimedia framework domain
Jun 29 17:46:11 <denkenz> Since they're not covered by AT commands
Jun 29 17:48:05 <pessi> ;)
Jun 29 17:54:25 <pessi> that is the problem... I'd push control of supervisory tones to ui, but how ui can know which tone to play and when? and how multimedia fw knowns when the play supervisory tone and when to play audio from network and when to mix it
Jun 29 17:55:07 <pessi> on n900, we have an elegant solution that uses ruleset written in Prolog ;)
Jun 29 18:07:30 <denkenz> how does oFono know when a supervisory tone needs to be played?
Jun 29 18:08:06 <denkenz> On most hw I've encountered, the modem takes care of audio generation
Jun 29 18:40:28 <Robot101> pessi: doesn't tonegend just listen to CSD and add the tones requested by the network from csd?
Jun 29 18:40:45 <Robot101> like, does it actually use the beautiful prolog policy so much?
Jun 29 19:15:50 <kvalo> pessi: moi. I tested n900 with a usb cable and I was able to get it work. but why did I have to manually set it to online mode?
Jun 29 19:31:56 <pessi> Robot101: the policy overrides the tones in some cases
Jun 29 19:32:11 <Robot101> just to keep you guessing :)
Jun 29 19:32:18 <pessi> yep
Jun 29 19:32:47 <pessi> or frustrated, whne you have managed to get the tones to work, they get silenced because a because
Jun 29 19:33:31 <pessi> kvalo: I needed a victim for flight mode ;)
Jun 29 19:33:39 <pessi> mbm is next to go ;)
Jun 29 19:38:04 <kvalo> pessi: sorry, I still don't understand
Jun 29 19:38:19 <kvalo> flight mode or not, why can't it be just in online mode by default?
Jun 29 19:42:08 <kvalo> pessi: or should connman always enable online mode?
Jun 29 19:42:39 <pessi> perhaps
Jun 29 19:43:19 <pessi> i'd rather default to offline, but there should be storage for online
Jun 29 19:43:48 <kvalo> I could use n900 directly from connman if we can solve this online problem somehow
Jun 29 19:51:24 <denkenz> kvalo: Please recompile your code with no CFLAGS before patch submissions ;)
Jun 29 19:51:26 <pessi> kvalo: you can always go to isimodem.c and comment out .set_online ?
Jun 29 19:52:45 <kvalo> denkenz: I think I did that. what was wrong?
Jun 29 19:52:57 <denkenz> 3 unused vars
Jun 29 19:53:20 <kvalo> denkenz: wow, sorry. I wonder why my compiler doesnt' warn anything
Jun 29 19:54:15 <denkenz> unsetting the env is not enough, have to bootstrap, but I'm sure you know that
Jun 29 19:54:34 <kvalo> V=1 shows "-Werror -Wextra"
Jun 29 19:54:54 <denkenz> Then get a better compiler ;)
Jun 29 19:55:28 <denkenz> Its the -O0 that turns off those kinds of warnings
Jun 29 19:55:42 <kvalo> pessi: of course. but I would like to keep distro modifications to the minimum.
Jun 29 19:56:18 <kvalo> pessi: and if this online is introduced to ofono, it should not prevent connman from using modems.
Jun 29 19:56:25 <denkenz> pessi: You need to get holtmann and sameo to figure out how they want to support Online mode in connman
Jun 29 19:56:48 <denkenz> The biggest problem is that Online mode is really what Powered used to be
Jun 29 19:57:08 <kvalo> denkenz: I wonder how I got that -O0
Jun 29 19:57:13 <denkenz> And right now we're in lala-land where some drivers do one, others do the other one
Jun 29 19:57:40 <pessi> biggest problem is tot get me to figure out how to use Online 
Jun 29 19:57:59 <denkenz> You're the one that wrote it ;)
Jun 29 19:58:11 <pessi> tomorrow some emergency calls with mbm ;)
Jun 29 19:58:19 <kvalo> pessi: if something is not clear yet, just default to online for now and change it later. don't make users suffer
Jun 29 19:59:48 <denkenz> pessi: Might want to keep logs and share them with us
Jun 29 19:59:58 <denkenz> pessi: particularly interesting sim/no sim emergency calls
Jun 29 20:00:25 <pessi> denkenz: sure
Jun 29 20:00:48 <denkenz> and also CFUN=4 / CFUN=1 emerency calls
**** ENDING LOGGING AT Wed Jun 30 02:59:57 2010