**** BEGIN LOGGING AT Wed Jul 11 02:59:58 2012 Jul 11 04:15:46 Hello, could I please have an account for the Trac Wiki @ shr-project.org. Thanks. Jul 11 09:17:25 mrmoku: ^^ Jul 11 13:59:15 paulk-desktop: hi Jul 11 13:59:31 hi Alex[sp3dev] Jul 11 14:00:31 paulk-desktop: I've spent today fscking with GPS. I don't have a clue why the proprietary library fails to run with our libsamsung-ril-client. I've made dumps of IPC and ril is unused for gps Jul 11 14:01:00 paulk-desktop: also, I've dumped some uart traffic and once firmware is loaded it seems to use good old binary sirf protocol Jul 11 14:02:16 Alex[sp3dev], ok. I'd be interested to see how the SIRF logs look like Jul 11 14:06:59 paulk-desktop: hope you can download from here http://rghost.ru/39155751 . the 'uart' file is my bad attempt at hacking kernel uart driver to print buffer char by char. hope you can correlate that with the daemon log Jul 11 14:07:24 paulk-desktop: if anything, make sure s2 uses the same chip/driver. I think it does, but it was long since I looked Jul 11 14:08:35 ok, thanks Jul 11 14:08:42 I'll look at galaxys2 eventually Jul 11 14:08:57 (I'm working on another phone ATM) Jul 11 14:09:11 and I'm badly sick so I don't do much… Jul 11 14:09:14 I wonder why RIL has gps ipc stuff if it is unused Jul 11 14:09:49 I think it's because some modems do GPS too Jul 11 14:09:58 like in Samsung H1 IIRC Jul 11 14:10:05 so to get GPS data, you ask the modem directly Jul 11 14:10:06 ok, try to recover then. sucks to be sick in summer Jul 11 14:10:13 yeah… Jul 11 14:11:24 on the other hand, I could binary-patch the gps library and disable error checks. that should make it work and that'll be ok until we decide what to do about gps Jul 11 16:43:05 morphis, hi Jul 11 16:43:06 hmmm Jul 11 16:43:12 GNUtoo-desktop: heyho Jul 11 16:43:24 1) I don't remember well what happened but I got audio working again thanks to the wiki with http://pastie.org/private/bz6zunzteudcjaltrc5uq Jul 11 16:43:28 but that's the problem Jul 11 16:43:40 in one hand we have mark brown's incomplete driver Jul 11 16:43:50 and in the other hand we have an unusable driver Jul 11 16:44:08 I mean messing with alsamixer is not working as expected either Jul 11 16:44:16 when you change route it reset the controls Jul 11 16:44:22 and doesn't change the control status Jul 11 16:44:37 hm Jul 11 16:44:39 maybe it doesn't show that behaviour in other insterfaces like direct alsa or amixer Jul 11 16:45:08 I tried it this morning too Jul 11 16:45:18 with a plain shr-lite-image Jul 11 16:45:22 as it comes out of OE Jul 11 16:45:28 one test.wav and aplay Jul 11 16:45:37 disabled alsa-stream-keeper in fsodeviced Jul 11 16:45:42 # aplay test.wav Jul 11 16:45:57 sounds come through the speakers after enabling in alsamixer Jul 11 16:46:06 second # aplay test.wav Jul 11 16:46:11 mixing works Jul 11 16:46:20 third # aplay test.wav Jul 11 16:46:26 mixing still works Jul 11 16:46:43 ah? Jul 11 16:46:49 I didn't tried with mplayer Jul 11 16:47:01 ok Jul 11 16:47:10 the format of test.wav is important Jul 11 16:47:27 /usr/share/sounds/female_ringtone.wav works? Jul 11 16:47:51 pcm_s16le Jul 11 16:47:53 was it Jul 11 16:47:58 stereo? Jul 11 16:47:59 I used my own test.wav Jul 11 16:48:03 indeed Jul 11 16:48:19 I wondered the format of your own test.wav Jul 11 16:48:25 AUDIO: 44100 Hz, 2 ch, s16le, 1411.2 kbit Jul 11 16:48:29 ok Jul 11 16:48:32 that works Jul 11 16:48:33 so stereo Jul 11 16:48:49 whats the format of /usr/share/sounds/female_ringtone.wav ? Jul 11 16:48:57 how to disable stream keeper already? Jul 11 16:49:01 mono s16_le Jul 11 16:49:24 and 44100 Jul 11 16:49:44 you tried with what .asoundrc? Jul 11 16:49:58 GNUtoo-desktop: if you have latest FSO and no [fsodeviced.herring_qurisk/alsa_stream_keeper] in your fsodeviced.conf it's disabled Jul 11 16:50:13 ok Jul 11 16:50:16 I used a plain shr-lite-image without any modification Jul 11 16:50:22 ah ok Jul 11 16:50:29 now it works with my test.wav too Jul 11 16:50:30 so no .asoundrc Jul 11 16:50:34 hm Jul 11 16:51:27 indeed Jul 11 16:51:33 before messing with it it didn't work Jul 11 16:51:41 ok Jul 11 16:51:58 like fresh image Jul 11 16:52:04 maybe because fsoaudiod is now disabled Jul 11 16:52:08 I'll reboot Jul 11 16:52:24 I used our kernel we have in OE Jul 11 16:52:28 not broonie's Jul 11 16:52:42 I disabled it because it ate the ram space on /var/log/fsoaudiod.log Jul 11 16:52:47 me too Jul 11 16:52:51 stock oe stuff Jul 11 16:53:09 I'll reboot without fsoaudiod and see Jul 11 16:53:31 ok Jul 11 16:53:32 (space means Out of space because of the 200M of logs) Jul 11 16:53:36 :) Jul 11 16:53:43 FSO_LOG_LEVEL=DEBUG :) Jul 11 16:53:50 default stuff Jul 11 16:54:08 full of: Jul 11 16:54:16 2012-07-09T21:03:30.895262Z [ERROR] RouterLibAlsa <>: Could not switch to new mode FREE_SMARTPHONE_AUDIO_MODE_CALL; switching back to old mode FREE_SMARTPHONE_AUDIO_MODE_NORMAL ... Jul 11 16:54:21 hm Jul 11 16:54:22 ok Jul 11 16:54:30 we should disable the router stuff anyway Jul 11 16:54:37 as we continue to use fsodeviced for this Jul 11 16:54:54 now after reboot it blocks again Jul 11 16:54:58 at least until we found out what is the right way for us to go for audio Jul 11 16:55:10 with mplayer? Jul 11 16:55:21 yes Jul 11 16:55:24 I tested aplay too Jul 11 16:55:35 same test.wav? Jul 11 16:55:44 yes Jul 11 16:55:46 ok Jul 11 16:55:58 I will try here again in half an hour Jul 11 16:56:11 but.... Jul 11 16:56:15 arecord -f cd Jul 11 16:56:19 ctrl+c Jul 11 16:56:23 then playback Jul 11 16:56:25 makes it work Jul 11 16:56:28 hm Jul 11 16:56:36 so the driver is buggy somewhere Jul 11 16:56:49 maybe easy to fix when we know where? Jul 11 16:56:53 yes in a lot of places I guess Jul 11 16:56:53 ok Jul 11 16:57:03 or we could record shortly Jul 11 16:57:12 in the fsodeviced quirk? Jul 11 16:57:37 or do what the free software android libaudio does Jul 11 16:57:43 like open for recording or something like that Jul 11 16:57:54 GNUtoo-desktop: it does? Jul 11 16:58:22 no idea Jul 11 16:58:24 GNUtoo-desktop: we should test if it stays usable after the short arecord on startup Jul 11 16:58:30 let me verify the arecord theory Jul 11 16:59:34 arecord + enabling the mic for arecord it seems Jul 11 16:59:35 ok Jul 11 17:01:57 and together with the stream-keeper? Jul 11 17:02:13 I'll try more stuff Jul 11 17:02:16 it's too early right now Jul 11 17:02:19 ok Jul 11 17:10:24 hmmm I cannot reproduce anymore Jul 11 17:10:28 stuck again Jul 11 17:10:34 (me and mplayer) Jul 11 17:11:24 hm Jul 11 17:14:07 hmmm I've another theory Jul 11 17:14:12 mv asounrc .asounrc Jul 11 17:14:18 mplayer 02.wav Jul 11 17:14:28 that plays music Jul 11 17:14:37 then mv .aosoundrc asoundrc Jul 11 17:14:40 and mplayer 02.wav Jul 11 17:14:44 and it seems to work Jul 11 17:15:19 status_t AudioHardware::AudioStreamInALSA::open_l() Jul 11 17:15:42 http://pastie.org/private/3mz69c62uidat4egpneg Jul 11 17:18:35 ? Jul 11 17:19:26 the pastie is from libaudio for crespo Jul 11 17:20:03 they use the following config: Jul 11 17:20:06 44100 Jul 11 17:20:08 stereo Jul 11 17:20:21 ok Jul 11 17:20:26 so they never play something else Jul 11 17:20:28 #define AUDIO_HW_OUT_PERIOD_SZ 1024 Jul 11 17:20:28 #define AUDIO_HW_OUT_PERIOD_CNT 4 Jul 11 17:20:34 I don't know yet Jul 11 17:20:37 I must read more Jul 11 17:21:26 ok Jul 11 17:22:07 http://pastie.org/private/wyjqsv7eoc6tejccwci4g sounds googd Jul 11 17:22:23 mmap works Jul 11 17:22:30 so it's jut the initialization I think Jul 11 17:23:00 I'll try more the mv .asoundrc asoundrc Jul 11 17:23:01 etc... Jul 11 17:23:04 to be sure Jul 11 17:30:07 what's in your .asoundrc? Jul 11 17:30:23 I pasted it before, I'll repaste Jul 11 17:30:43 http://pastie.org/private/ozff9ntquil3udmtuulsww Jul 11 17:30:58 ah ok Jul 11 17:31:02 missed that one Jul 11 17:32:07 fsodeviced.herring_qurisk/alsa_stream_keeper has to be disabled also Jul 11 17:32:23 ah strange Jul 11 17:32:26 qurisk Jul 11 17:33:24 maybe come back when I'll have found the magic thing Jul 11 17:33:54 because I guess I will need your help for sound in-calls Jul 11 17:35:17 I'm lost again Jul 11 17:35:19 see Jul 11 17:35:44 one time it's arecord Jul 11 17:35:51 one time it's mv asoundrc.... Jul 11 17:36:29 + messing with the controls Jul 11 17:37:45 basically this time: Jul 11 17:37:54 fsodeviced reset the controls Jul 11 17:37:58 so it make it blocks Jul 11 17:38:12 then I do arecord + playing with controls Jul 11 17:38:16 and it works again Jul 11 17:38:38 it's still strange Jul 11 17:38:56 aha there is a timeout in the reset of the paths Jul 11 17:39:10 hmmm Jul 11 17:39:22 for instance: Jul 11 17:39:27 # fsodeviced Jul 11 17:39:32 /wait Jul 11 17:39:42 the # mplayer 02.wav -> silence Jul 11 17:40:00 if I remove the //wait Jul 11 17:40:04 I've music Jul 11 17:40:24 some new bug? Jul 11 17:40:31 crespo device Jul 11 17:40:44 *device: crespo/herring/nexus S Jul 11 17:41:09 GNUtoo-desktop: I am trying here with fso2-console-image only Jul 11 17:41:11 right now Jul 11 17:41:15 ok Jul 11 17:41:20 maybe I should do the same Jul 11 17:41:25 with a console-only image Jul 11 17:41:31 right after boot: http://pastie.org/4239045 Jul 11 17:41:48 so I don't mess up with fsoaudiod/fsodeviced while playing Jul 11 17:41:59 ah ok Jul 11 17:42:06 I do mess up with stuff while playing Jul 11 17:42:21 like alsamixer Jul 11 17:42:39 hm killing aplay with CTRL+C make it block Jul 11 17:42:47 "Aborted by signal Interrupt..." Jul 11 17:42:53 ok Jul 11 17:43:25 and on the console I get this: http://pastie.org/4239055 Jul 11 17:43:54 that's dmesg Jul 11 17:44:46 yes Jul 11 17:44:53 s/console/dmesg/ Jul 11 17:44:56 ok Jul 11 17:45:26 how to prevent X from starting with systemd? Jul 11 17:45:27 I am installing mplayer right now Jul 11 17:45:31 ok Jul 11 17:45:32 puuh Jul 11 17:45:39 rename xserver-nodm.service Jul 11 17:45:42 if aplay works mplayer should works Jul 11 17:45:45 *work Jul 11 17:45:46 ok Jul 11 17:46:34 ok Jul 11 17:46:59 if one works the other works too Jul 11 17:47:02 same for blocking Jul 11 17:47:04 hm Jul 11 17:47:17 I don't get it blocking like you Jul 11 17:47:32 always without a .asoundrc Jul 11 17:47:50 just the global /etc/asound.conf Jul 11 17:48:00 [ 38.430132] s3c2412_snd_lrsync: timeout Jul 11 17:48:14 I'll try your image soon Jul 11 17:48:26 I'm compiling it right now Jul 11 17:48:38 ok, fsodeviced is up with stream-keeper enabled Jul 11 17:48:41 and it blocks now Jul 11 17:49:04 ah ok Jul 11 17:49:15 stopping fsodeviced and starting mplayer again plays music again Jul 11 17:49:30 so our stream-keeper is the bad thing Jul 11 17:49:44 or one of the bad things Jul 11 17:49:52 I'm not sure it was started for me Jul 11 17:49:58 ok Jul 11 17:50:13 grep stream /var/log/fsodeviced.log returns nothing for instance Jul 11 17:50:24 A: 0.0 (00.0) of 248.6 (04:08.5) ??,?% Jul 11 17:50:29 and mplayer blocks ^^^ Jul 11 17:51:16 ok but thats not reproduceable Jul 11 17:51:28 I have now fsodeviced running and mplayer works Jul 11 17:51:44 is log_level = DEBUG set? Jul 11 17:51:56 let me look Jul 11 17:52:10 with the new fsodeviced there is no more output for the stream keeper afaik Jul 11 17:52:45 fsogsmd fails in do_configure Jul 11 17:53:22 really? Jul 11 17:53:52 seems like you're my CI build :) Jul 11 17:54:53 http://pastebin.com/Y3jdLCSU Jul 11 17:55:13 | No package 'fsotest-2.0' found Jul 11 17:55:21 ah :) Jul 11 17:55:28 libfsotest is missing in OE Jul 11 17:55:33 ok Jul 11 17:55:36 what should I do then Jul 11 17:55:49 will add it Jul 11 17:55:53 ok Jul 11 17:55:54 thanks Jul 11 17:55:59 GNUtoo-desktop: I rebooted my fso-console-image Jul 11 17:56:03 started fsodeviced Jul 11 17:56:08 and started playing Jul 11 17:56:09 it works Jul 11 17:56:12 ok Jul 11 17:56:15 mixings works too Jul 11 17:56:17 yes it's strange Jul 11 17:56:26 I'll test your fso-console-image Jul 11 17:56:36 I suspect that the /etc/asound.conf is different Jul 11 17:56:41 always with stream-keeper in enabled Jul 11 17:57:01 http://pastie.org/4239109 Jul 11 17:57:10 it's the default one from SHR Jul 11 17:57:20 ok yes Jul 11 17:57:59 I'll test your image as soon as I can compile fsodeviced Jul 11 17:58:02 *fsogsmd Jul 11 18:01:19 freesmartphone.org: 03morphis 07cornucopia * rc55d9f763c91 10/fsogsmd/tests/integration-tests.vala: fsogsmd: tests: integration: wait until modem leaves INITIALIZING state or fail otherwise Jul 11 18:01:20 freesmartphone.org: 03morphis 07cornucopia * r8386a7d21f66 10/fsogsmd/tests/ (integration-tests.vala lt-integration_tests.conf): fsogsmd: tests: integration: rework configuration handling and add missing default config Jul 11 18:01:21 freesmartphone.org: 03morphis 07cornucopia * r954852fe7b50 10/fsogsmd/tests/integration-tests.vala: fsogsmd: tests: integration: add missing exception handling to avoid compiler warnings Jul 11 18:01:21 freesmartphone.org: 03morphis 07cornucopia * re65e1f14da1a 10/libfsotest/vapi/glib-2.0-ext.vapi: libfsotest: add missing glib-2.0-ext.vapi (blame .gitignore for this!) Jul 11 18:01:47 GNUtoo-desktop: recipes are finished, just running a test build Jul 11 18:01:52 ok Jul 11 18:02:03 GNUtoo-desktop: you have a different /etc/asound.conf? Jul 11 18:02:28 I don't think so Jul 11 18:03:21 ok Jul 11 18:04:33 looks good so far Jul 11 18:04:45 playya: you got any further with Galaxy Nexus support? Jul 11 18:05:22 GNUtoo-desktop: it would only be nice if we can avoid the stream-keeper Jul 11 18:05:35 ok Jul 11 18:05:41 maybe trigger fsodeviced with a alsa hook to reset the current scenario Jul 11 18:11:22 ping me when the recipe is ready Jul 11 18:11:30 SHR: 03morphis 07meta-smartphone * r1a3694ab4c9e 10/meta-fso/recipes-freesmartphone/cornucopia/fsogsmd_git.bb: meta-fso: cornucopia: fsogsmd: depend on libfsotest and bump PR Jul 11 18:11:39 GNUtoo-desktop: ^^ Jul 11 18:11:40 SHR: 03morphis 07meta-smartphone * r063009a1835d 10/meta-fso/recipes-freesmartphone/cornucopia/ (libfsotest.inc libfsotest_git.bb): meta-fso: cornucopia: add recipe for libfsotest - the FSO testing library Jul 11 18:12:33 ok thanks Jul 11 18:21:02 morphis, still not brave enough ti run it :( Jul 11 18:21:13 ok Jul 11 18:21:19 you need some help? Jul 11 18:22:27 I tried to create a backup over night but it took too long Jul 11 18:22:35 maybe tonight Jul 11 18:23:57 and the devshell/menuconfig sucks Jul 11 18:27:38 menuconfig? Jul 11 18:27:55 -c menuconfig Jul 11 18:28:05 yes Jul 11 18:28:50 should I use devtmpfs in my kernel? Jul 11 18:30:08 the touchscreen needs a firmware and I need to figure out where it is on my device Jul 11 18:30:22 ah ok Jul 11 18:30:34 currently there is udev installed in SHR so no need for devtmpfs Jul 11 18:38:33 no Jul 11 18:38:35 there is uedv Jul 11 18:38:37 but Jul 11 18:38:43 there is also systemd Jul 11 18:38:48 which require devtmpfs Jul 11 18:38:51 and devtmpfs mount Jul 11 18:42:58 hmm. yes Jul 11 18:44:45 ah yes, systemd needs it Jul 11 18:45:04 GNUtoo-desktop: hm, shr-lite-image powerd up gives no sound but no blocking aplay Jul 11 18:45:08 (with stream keeper enabled) Jul 11 18:45:35 ok Jul 11 18:45:47 without stream keeper? Jul 11 18:45:53 killing it, playing sound works Jul 11 18:47:08 ok Jul 11 18:47:10 let me retry Jul 11 18:48:04 somehow the mixer is messing up after some retries Jul 11 18:48:12 now I can't get sound anymore Jul 11 18:48:17 no fsodeviced is running Jul 11 18:48:24 ok Jul 11 18:48:24 just trying aplay + alsamixer Jul 11 18:48:27 ok Jul 11 18:48:46 [ 297.268326] s3c2412_snd_lrsync: timeout Jul 11 18:48:50 I have in dmesg Jul 11 18:48:51 ah ok Jul 11 18:48:53 right Jul 11 18:48:59 so you can reproduce my issue Jul 11 18:49:05 that was not there before Jul 11 18:49:17 I have no blocking aplay or mplayer Jul 11 18:49:19 just no sound Jul 11 18:49:21 maybe bad alsa state Jul 11 18:49:23 ok Jul 11 18:49:26 yes Jul 11 18:49:39 so we have to look in the kernel source Jul 11 18:49:41 by state I mean the mixer state Jul 11 18:49:44 like that Jul 11 18:49:45 yes Jul 11 18:49:49 alsactl -f foo store Jul 11 18:50:22 seems the mixer settings are not really safe to use when closing/opening the pcm often Jul 11 18:53:10 ok Jul 11 18:54:07 maybe it works right after flashing Jul 11 18:54:15 and something broke it Jul 11 18:54:20 s/flashing/adb push Jul 11 18:57:07 3642 of 4461 Jul 11 18:57:21 I am thinking more that setting a scenario with our router before the pcm was opend is causing the bad Jul 11 19:01:05 GNUtoo-desktop: can I disable debug mode of the wm8994 driver somewhere in sysfs or proc? Jul 11 19:08:20 thats wired something sets output path on boot to RCV Jul 11 19:08:52 I removed a lot components from the image, fsodeviced, fsoaudiod, phonefsod, phoneuid, e-wm, ... and it's still set? Jul 11 19:10:39 ok. I need to adjust the chroot-script first Jul 11 19:11:07 morphis, no idea Jul 11 19:12:45 can the codec hold it's state accross a reboot? Jul 11 19:13:32 possible Jul 11 19:13:45 in kernel hacking we have a trick Jul 11 19:13:56 on some board there is a reset button Jul 11 19:14:05 and when we have no serial output yet Jul 11 19:14:10 we hit the reset button Jul 11 19:14:14 and dump the ram Jul 11 19:14:22 specially the reagion where there is the dmesg Jul 11 19:14:24 and it does work Jul 11 19:14:34 so ram do stay across reboot Jul 11 19:14:39 so why not the CODEC state? Jul 11 19:14:46 hm ok Jul 11 19:14:51 but it's not the case Jul 11 19:14:53 I removed the battery Jul 11 19:14:57 and it's still set Jul 11 19:15:03 and I get the timeout in dmesg Jul 11 19:15:23 also you must take into account that the userspace save/restore the control states Jul 11 19:15:44 damn it Jul 11 19:16:04 for instnace the init scripts did that Jul 11 19:16:32 and also fsodeviced has support for save/restore as you know Jul 11 19:16:47 so some shr deamon uses that during calls Jul 11 19:17:19 yes Jul 11 19:17:27 but I think the init scripts are to blame Jul 11 19:17:29 disabled this now Jul 11 19:17:35 ok Jul 11 19:17:56 how is that called in systemd now? Jul 11 19:18:51 there is a service file for this in /lib/systemd/system Jul 11 19:19:03 yes I know but how is it called? Jul 11 19:19:41 that doesn't help Jul 11 19:19:53 it's a dependency on basic.target and shutdown.target Jul 11 19:20:13 ok Jul 11 19:20:26 is it possible to boot a kernel over USB like palm-pre? Jul 11 19:20:54 playya: with fastboot Jul 11 19:21:06 playya: go into bootloader mode Jul 11 19:21:14 and then: sudo fastboot boot zImage-crespo.fastboot Jul 11 19:21:31 thats what I am doing all the time Jul 11 19:21:40 ok Jul 11 19:22:28 GNUtoo-desktop: ok, removing /var/lib/alsa/asound.state helps Jul 11 19:22:30 I copying the image and rebuild the kernel with new init scrip Jul 11 19:22:48 no timeout any more Jul 11 19:23:21 GNUtoo-desktop: so adjusting the mixer when PCM is not used is very bad Jul 11 19:23:29 image finished Jul 11 19:23:38 do I need to change the build in command line? Jul 11 19:23:41 and puts the codec into a bad state Jul 11 19:23:48 playya: which build? Jul 11 19:24:00 I adb pushed the image Jul 11 19:24:03 and now it works Jul 11 19:24:11 morphis, for marugo Jul 11 19:24:12 I've to aslamixer to make it work tough Jul 11 19:24:39 playya: which build do you mean? Jul 11 19:24:56 playya, it depends Jul 11 19:25:04