**** BEGIN LOGGING AT Tue Nov 13 02:59:57 2012 Nov 13 18:04:16 Hm, i wonder how fsogsmd is supposed to be loaded. When i simply request GSM from fsousaged, it says it was never registered. Nov 13 18:04:29 What and when is usually dbus-activating or explicitely loading it? Nov 13 18:05:05 on debian? Nov 13 18:05:14 does debian uses systemd or sysvinit? Nov 13 18:06:53 GNUtoo-x60: on SHR Nov 13 18:07:26 ok Nov 13 18:07:34 let me ssh and look Nov 13 18:09:25 PaulFertser, sorry it's not started by systemd Nov 13 18:09:29 probably by dbus then Nov 13 18:11:07 GNUtoo-x60: yes, it might be. But something should be trying to use org.freesmartphone.ogsmd then. Nov 13 18:11:13 I wonder what does it. Nov 13 18:11:59 hmmm Nov 13 18:12:06 maybe SHR deamons Nov 13 18:15:31 GNUtoo-x60: yes, the SHR daemons are requesting the GSM resource Nov 13 18:15:44 ok Nov 13 18:16:33 morphis: what do you mean? When i try to request GSM fsodeviced just tells me it was never registered. Nov 13 18:17:10 PaulFertser: fsousaged should list the GSM resource Nov 13 18:17:32 it depends wether the shadow resource feature is enabled or not Nov 13 18:17:45 if it is then fsousaged will activate fsogsmd through dbus Nov 13 18:17:57 if not somebody else has to activate fsogsmd Nov 13 18:18:02 morphis: ah, shadow resource feature, right. Nov 13 18:18:08 morphis: it seems to be disabled by default Nov 13 18:18:34 morphis: it's too long i was working with FSO, have absolutely forgotten about the shadow resource feature. Nov 13 18:18:35 yes Nov 13 18:18:39 :) Nov 13 18:19:54 morphis: thanks! Nov 13 18:20:31 np Nov 13 18:30:16 PaulFertser, what's the shadow resource? Nov 13 18:32:38 GNUtoo-x60: it's a special trick to have fso daemons auto-activated when something requests them from fsousaged. Nov 13 18:33:27 ok Nov 13 18:59:21 morphis: on my system it got compiled the way that ${datarootdir}/dbus-1/system-services is used to search for the shadow resources, and also i notice none of the supplied configs uses that option. Nov 13 18:59:49 I mean literally Nov 13 19:04:57 In the generated Makefile DBUS_SYSTEM_SERVICE_DIR goes before datarootdir = ${prefix}/share so no wonder. Nov 13 19:09:32 GNUtoo-x60: now pushed another improvements to fso-el, you can even easily run it on your host and access the phone via dbus-proxy-daemon Nov 13 19:09:52 ok Nov 13 19:09:53 nice Nov 13 19:11:12 I'm not sure what Emacs version to recommend though. Emacs24 in Debian unstable misses signals but self-built today version works fine. Nov 13 19:11:24 ok Nov 13 19:12:39 And Emacs23 doesn't allow to specify custom dbus address, so with dbus-proxy-daemon you'll need to run the whole Emacs prosses with a special env variable and specify :session bus for fso.el Nov 13 19:19:02 On gta04 emacs is fast anyway so it's easier to run it on device itself. Nov 13 19:25:18 GNUtoo-x60: it looks like the only way fsogsmd can properly establish "data service" is when fsonetworkd running... Nov 13 19:25:23 I'm not sure i like that. Nov 13 19:26:13 hmmm Nov 13 19:46:15 I mean is that an official FSO policy? Nov 13 19:50:32 I mean people often like to run other network managers and that would probably not play along well etc. Nov 13 20:09:53 hey! Nov 13 20:10:15 GNUtoo-x60, were you able to try the headset stuff yesterday? Nov 13 20:23:59 slyon, not yet Nov 13 20:24:27 ok Nov 13 20:31:32 slyon, opkg upgrading now Nov 13 20:40:54 cool :) Nov 13 20:50:48 slyon, strange....I've still the init wich asks me to choose between nand and microsd Nov 13 20:51:08 Linux om-gta04 3.4.9 #1 PREEMPT Tue Nov 13 00:07:03 CET 2012 armv7l GNU/Linux Nov 13 20:56:22 GNUtoo-x60, hmm. that's strange... I've defenitely removed gta04-init: http://git.shr-project.org/git/?p=meta-smartphone.git;a=commit;h=8b81a1be326117acc3f2c7e15cc79acdd70baf17 Nov 13 20:56:35 yes Nov 13 20:56:39 and I have that commit Nov 13 20:57:40 GNUtoo-x60, do you have the "Voice route" alsa control in alsamixer? Nov 13 20:58:07 how to grep that? Nov 13 20:59:24 amixer list or something? Nov 13 20:59:33 it's just amixer Nov 13 20:59:34 ah yes Nov 13 20:59:38 without parameters Nov 13 20:59:45 there is also alsactl -f foo store Nov 13 20:59:47 vim foo Nov 13 20:59:49 yes Nov 13 21:00:08 Simple mixer control 'Voice route',0 Nov 13 21:00:52 hmm do you have another system installed? Nov 13 21:01:04 afaik gta04-init does some symlinking magic Nov 13 21:01:17 maybe it started with an older/other kernel and then booted the new one? Nov 13 21:02:12 maybe Nov 13 21:02:20 anyway I want to try the patches Nov 13 21:03:04 yes. but as "Voice route" is available you should be able to test my fsodeviced Nov 13 21:03:12 yes Nov 13 21:10:41 now what should I test? Nov 13 21:13:58 GNUtoo-x60, trying to play sound, using the 'headset' scenario Nov 13 21:14:13 and trying to make a call, using the 'gsmheadset' scenario Nov 13 21:14:45 i assume you'll have noise using 'gsmheadset' then you'd need to flip the AVADC Clock priority control Nov 13 21:14:55 I've no noise Nov 13 21:15:04 or very minor maybe Nov 13 21:15:27 OK. even better. using gsmhandset I had very loud noise in the background Nov 13 21:15:30 how does the noise looks like? Nov 13 21:15:47 should I push audio on my router? Nov 13 21:16:03 paulk had noise long time ago too Nov 13 21:16:31 because there is noise Nov 13 21:16:36 but the voice is way louder Nov 13 21:16:41 which router? Nov 13 21:17:19 ah ouch Nov 13 21:17:24 wait one second Nov 13 21:17:28 I've noise Nov 13 21:17:40 arecord -r 8000 -c 1 > rec.wav => noise Nov 13 21:17:59 arecord -f cd > cd.wav => very minor noise Nov 13 21:18:31 hmm i'm not sure about arecord... Nov 13 21:18:40 ok Nov 13 21:18:41 afaik the AVADC has something to do with the modem Nov 13 21:18:43 what should I do? Nov 13 21:18:46 real calls? Nov 13 21:18:52 so i think it should be tested in a call Nov 13 21:18:53 yes Nov 13 21:19:00 like calling the mailbox or something Nov 13 21:19:34 note that it stiill jumps a bit back and forth between headset and normal during calls Nov 13 21:19:52 I ear nothing during that call Nov 13 21:19:57 before I did ear something Nov 13 21:20:04 and forwarder is fixed Nov 13 21:20:06 ah wait a bit Nov 13 21:20:11 maybe not in your branch Nov 13 21:20:22 i've rebased. should include you latest stuff Nov 13 21:20:51 ok Nov 13 21:20:56 fsoaudiod is still there Nov 13 21:20:58 but I do: Nov 13 21:21:01 4242 => works Nov 13 21:21:06 4200 => fails Nov 13 21:22:07 voice route should be set to something? Nov 13 21:22:09 to twl? Nov 13 21:22:11 to SOC? Nov 13 21:22:50 it should be set to SoC for swrouting Nov 13 21:22:55 and to twl for hwrouting Nov 13 21:23:33 but it should be set to SOC in the gsmheadset scenario Nov 13 21:23:58 ok Nov 13 21:25:21 maybe this has something to do with this strange bug, the people on gta04-owner complain about for QtMoko where the sound card is kept open, so no other sound can be played? Nov 13 21:25:39 no idea but we fixed that long ago in SHR Nov 13 21:25:45 ok Nov 13 21:25:54 it was ringing Nov 13 21:25:58 and no sound during call Nov 13 21:26:03 but here it's not ringing Nov 13 21:26:06 and it's fixed already Nov 13 21:26:09 since ages Nov 13 21:26:35 ok. so it shouldn't be that problem Nov 13 21:27:49 hmm. but was the sound OK when you called 4242? Nov 13 21:28:07 wow I know why it jumps Nov 13 21:28:09 [ 1210.239990] twl4030_madc twl4030_madc: conversion timeout! Nov 13 21:28:25 that's a very usefull info Nov 13 21:28:35 yes Nov 13 21:28:37 wow. nice find! Nov 13 21:28:39 but I don't record Nov 13 21:28:49 oops Nov 13 21:28:51 I meant: Nov 13 21:29:05 on 4242 the remote is only a dumb interactive menu Nov 13 21:29:15 it doesn't feed you back your voice Nov 13 21:29:58 that should be OK... I had the noise in the background of my dumb mailbox menu, not on my own speech Nov 13 21:32:41 no I mean that I want my speech back Nov 13 21:32:49 like the echo in asterisk Nov 13 21:32:58 like call 500 and get your voice back in asterisk Nov 13 21:33:26 yes, i know what you mean. Nov 13 21:34:27 just my noise-problem wasn't on my own voice, but on the sound i heard from the other side Nov 13 21:34:45 ok Nov 13 21:34:54 hmmm Nov 13 21:34:59 now what should I do? Nov 13 21:35:01 call you? Nov 13 21:35:01 so if this works for you (you hear no voice on in your earpiece/headset) it should be allright Nov 13 21:35:14 and you record me? Nov 13 21:35:24 or just ear me Nov 13 21:35:25 ? Nov 13 21:36:06 hmm if you like to Nov 13 21:36:14 ok Nov 13 21:36:18 what's your number? Nov 13 21:36:22 maybe PM it Nov 13 21:43:14 so summup of private conversation: Nov 13 21:43:15 no echo Nov 13 21:43:21 and no noise Nov 13 21:43:26 yes, on both sides Nov 13 21:44:02 but: slight delay (silence) at the beginning (few seconds) Nov 13 21:45:00 yes Nov 13 21:45:03 we should fix it Nov 13 21:45:13 also I've made a fix for the forwarder crash Nov 13 21:45:19 some days ago Nov 13 21:45:36 the glibc malloc/calloc/free etc... failures Nov 13 21:46:25 yeah. that fix you did is really important Nov 13 21:47:45 ok Nov 13 21:47:50 for A3.... Nov 13 21:48:11 yes Nov 13 21:48:44 what you did for A4 is even more important Nov 13 21:48:54 because way more people have > A3 than A3 Nov 13 21:50:33 yes. now that my hwrouting work seems to work, I'll switch hwrouting on by default. and we A3 users will have to disable the extra_path option in fsodeviced.conf. OK? Nov 13 21:51:50 can't we make out of the box work for both? Nov 13 21:53:32 what does /proc/cpuinfo says for > A3 ? Nov 13 21:53:40 that would be possible too, but i'd need to hack device specific code into the generic alsa_router plugin then... Nov 13 21:53:56 because we have that on fsodeviced: Nov 13 21:53:59 mdbus2 -s org.freesmartphone.odeviced /org/freesmartphone/Device/Info org.freesmartphone.Device.Info.GetCpuInfo Nov 13 21:54:05 hmmm Nov 13 21:54:31 since it's vala can't we extends the standard class? Nov 13 21:54:39 but if you want to push, push Nov 13 21:55:40 it looked like the cleanest method to me... there are just ~20 A3s out there... so those people would need to disable "extra_path" after installation Nov 13 21:55:51 i'll document this on the installtion page Nov 13 21:56:02 (in our trac wiki) Nov 13 21:56:24 ok Nov 13 21:56:51 slyon: GNUtoo-x60: hi Nov 13 21:56:52 buying a new GTA04 is not cheap for me tough Nov 13 21:57:02 so I'm improving the A3 to make it become usable for me Nov 13 21:57:05 hi jake42 Nov 13 21:57:07 morphis, the slyon/gta04-hwrouting branch seems to be working. if you're OK with it please merge it into master and into 0.12 branches Nov 13 21:57:10 hi jake42! Nov 13 21:57:30 GNUtoo-x60, it's the same here. I don't have the money to buy a A4/A5 Nov 13 21:57:37 slyon, can you do the work for merging it to master? Nov 13 21:58:30 slyon: I stumbled upon your rework of the shr wikipage, I like it Nov 13 21:58:37 GNUtoo-x60, it's just git merge slyon/gta04-hwrouting it will apply cleanly. but I'd like to have morphis to review it before. Nov 13 21:58:48 slyon: think we should make that the new one Nov 13 21:58:56 http://www.shr-project.org/trac/wiki/WikiStartNew Nov 13 21:58:57 ah ok Nov 13 21:59:04 jake42, yes i started it some time ago... it's not finished yet Nov 13 21:59:20 WOW Nov 13 21:59:26 jake42, if you want: go ahead and finish it and push it as our main page Nov 13 21:59:29 slyon: already looks alot better than the oldone Nov 13 21:59:31 I was just thinking that the front page needed to be updated Nov 13 21:59:40 like some hours ago Nov 13 21:59:47 GNUtoo-x60: :) Nov 13 22:00:06 but mainly it was about What is SHR Nov 13 22:00:17 because it lists some devices that we don't support anymore Nov 13 22:00:53 ah it's fixed in your version Nov 13 22:00:58 heh. but there are far too much links on that page. some of them should be removed or combined Nov 13 22:01:05 ok Nov 13 22:01:11 some links are for developers Nov 13 22:01:49 well. if you like it.... we could as well make it the default now and improve it afterwards Nov 13 22:02:03 yes I like it, but only if we don't loose content Nov 13 22:02:17 if nothing is lost merge it Nov 13 22:06:47 ok. i'll double check the links and merge it afterwards Nov 13 22:07:30 ok Nov 13 22:07:37 btw I've merged xv some time ago Nov 13 22:08:01 but I didn't find a way to make mplayer use it *and* fallback on X11 when it's not possible to use it Nov 13 22:09:08 the TODO here is nearly complete: http://wiki.freesmartphone.org/index.php/Hardware/GTA04 but still some stuff lacking Nov 13 22:09:38 I'll add gta04 A3 && > A3 detection in that list Nov 13 22:11:05 GNUtoo-x60, we already can detect gta04a3 Nov 13 22:11:23 we're doing this in the fsogsmd gta04 lowlevel plugin afaik Nov 13 22:11:24 in fso? Nov 13 22:11:29 ok nice Nov 13 22:11:29 yes Nov 13 22:11:31 let me look Nov 13 22:11:51 it's just checking if a specific gpio exists, which does only exist on gta04 >= a4 Nov 13 22:12:52 GNUtoo-x60, http://git.freesmartphone.org/?p=cornucopia.git;a=blob;f=fsogsmd/src/plugins/lowlevel_gta04/plugin.vala;h=736a66f4fcf28995dbea15ef0b0d97f034633726;hb=HEAD#l49 Nov 13 22:13:29 yes I forgott about that Nov 13 22:13:48 thanks Nov 13 22:14:53 also what would be nice would be to complete sound handling for telephony: Nov 13 22:15:01 Add gsmspeaker scenario Nov 13 22:15:01 Add more controls in /etc/libphoneui.conf Nov 13 22:15:18 these 2 things have to be done as soon as possible I guess Nov 13 22:15:43 yes... with /etc/libphoneui.conf we have the problem of hwrouting control vs swrouting controls Nov 13 22:16:17 my idea was to use hwrouting controls as default and comment the swrouting controls, so we A3 user would have to uncomment them Nov 13 22:18:04 ok Nov 13 22:20:05 slyon, why not doing a router_gta04? Nov 13 22:20:13 there was a router_palmpre Nov 13 22:20:19 I could look into it if you want Nov 13 22:22:13 ah sorry that was in fsoaudiod Nov 13 22:24:07 or maybe better Nov 13 22:24:10 here you go: http://www.shr-project.org/trac/wiki Nov 13 22:24:14 we do it in quirks Nov 13 22:24:23 nice thanks Nov 13 22:24:51 GNUtoo-x60, i thought about that as well.. but it would be a LOT of doublicate code... and the current solution is just a few lines of code Nov 13 22:25:19 in quirks it would work out of the box Nov 13 22:25:27 it would be clean(hopefully) Nov 13 22:25:28 if you think this would be better, we can still do it Nov 13 22:25:39 and it would use vala overrides etc... ont to ducplicate code Nov 13 22:25:47 *not to duplicate code Nov 13 22:25:56 sounds good Nov 13 22:26:13 the problem is the following: if we don't do it I'll end up with a non-working phone some day Nov 13 22:26:20 because I reinstall often Nov 13 22:26:32 and I will probably forget to delete the line Nov 13 22:26:41 but there would still be the problem with /etc/libphoneui.conf Nov 13 22:26:51 ah? Nov 13 22:27:07 hwrouting needs other controls there Nov 13 22:27:35 really? Nov 13 22:27:46 I didn't know that Nov 13 22:28:10 what kind of diff does it needs there? Nov 13 22:28:25 if it's that: Nov 13 22:28:26 -sw Nov 13 22:28:27 +hw Nov 13 22:28:30 http://paste.debian.net/209026/ that's jeepingben s libphoneui.conf Nov 13 22:28:30 should be ok Nov 13 22:28:43 if it's different at the volume control level it may be way harder Nov 13 22:28:59 DAC Voice Digital Downlink Volume Nov 13 22:29:00 ouch Nov 13 22:30:26 the problem is that we can't drop A3 Nov 13 22:30:34 because all our devs have A3 Nov 13 22:30:43 unless we both upgrade to A4? Nov 13 22:30:45 how? Nov 13 22:31:08 I've a way but it's quite long and complicated Nov 13 22:31:28 we ask morphis and JaMa to exchange our A3 against their A4 Nov 13 22:31:31 or something like that Nov 13 22:32:05 else we detect and make work out of the box Nov 13 22:33:01 basically all the devs have A3 and all the potential users will have A4 Nov 13 22:33:05 or more recent Nov 13 22:33:07 that's not good Nov 13 22:33:10 yes... Nov 13 22:33:24 a friend of mine will get an a4 soon Nov 13 22:33:28 ok Nov 13 22:33:36 he'll probably become a dev afterwards :) Nov 13 22:33:40 ok nice Nov 13 22:34:24 autodetection would work in FSO. but i think it's problematic in SHR (libphoneui), because libphoneui doen't have any infrastructure for such sort of things Nov 13 22:34:50 but we could still do an ugly device specific hack in libphoneui... Nov 13 22:35:15 yes Nov 13 22:35:22 but we would then somehow make sure, that there is no gpio186 on any other device... Nov 13 22:35:24 if you have an idea on how to do it in SHR? Nov 13 22:35:31 s/if/ Nov 13 22:35:50 slyon, we could use fso for that Nov 13 22:35:56 like detect me the hw revision Nov 13 22:36:04 yes... Nov 13 22:36:22 that would be a good solution i think Nov 13 22:38:20 so what would be needed is this: define a "device revision/version" API Nov 13 22:38:27 implement this API Nov 13 22:39:17 create a gta04 specific alsa_router inherited from the generic alsa router Nov 13 22:39:37 adopt libphoneui to use the new FSO api to load different libphoneui.conf on a3 or a4+ Nov 13 22:40:10 use that api in the gta04_alsa_router as well Nov 13 22:47:37 ok Nov 13 22:47:41 I'll go to sleep Nov 13 22:47:46 there is fsodeviced infos Nov 13 22:48:07 and devmem can work in userspace and get that revision Nov 13 22:48:12 *devmem2 Nov 13 22:48:36 bye Nov 13 23:00:20 GNUtoo, btw: here is /proc/cpuinfo for gta04A3 and gta04A4, they are the same: http://lists.goldelico.com/pipermail/gta04-owner/2012-July/002609.html Nov 13 23:00:42 i'll leave now as well. bye! **** ENDING LOGGING AT Wed Nov 14 02:59:59 2012