| sharks_ | Thanks Elros, It was that simple: I did not have the latest build of droid-hal installed. I now have working wi-fi! If I just get audio and bluetooth now, I'll have a port I'm happy with! | 00:32 |
|---|---|---|
| sharks_ | Well except that most apps still don't open, have to work out what's wrong there... | 00:33 |
| sharks_ | Weird, if I run sailfish-browser from an ssh session it opens on the screen, but if I hit the app icon it tries to load on the homescreen and disappears. | 00:53 |
| sharks_ | Trying to debug pulseaudio. I understand hybris-18.1 at least used to have some troubles with audio, running `pulseaudio -v` as user spits out the following: | 02:40 |
| sharks_ | E: [pulseaudio] module-alsa-card.c: Failed to find a working profile. | 02:40 |
| sharks_ | E: [pulseaudio] module.c: Failed to load module "module-alsa-card" (argument: "device_id="0" name="platform-sound.29" card_name="alsa_card.platform-sound.29" namereg_fail=false tsched=yes fixed_latency_range=no ignore_dB=no deferred_volume=yes use_ucm=yes avoid_resampling=no card_properties="module-udev-detect.discovered=1""): initialization failed. | 02:40 |
| sharks_ | Any idea where I should be looking? | 02:41 |
| T42 | <elros34> how do you start it? with all arguments | 03:31 |
| sharks_ | At the minute after booting to UI I have `Process: 2236 ExecStart=/usr/bin/pulseaudio --daemonize=no $CONFIG (code=dumped, signal=ABRT)`. If I run `pulseaudio -v` I get the error pasted above. If I run `pulseaudio --start` I get no output, but instead of `(core=dumped, signal=ABRT)` I have `(core=exited, status=1/FAILURE)` on the same line. | 04:04 |
| sharks_ | If I run `pulseaudio -v --start` I just get a single line of output stating `I: [pulseaudio] main.c: Daemon startup successful.`, but that doesn't change the fact that I have no volume control or sound. | 04:07 |
| T42 | <elros34> how about you start it correctly with all arguments, see hadk-hot pulseaudio section | 04:09 |
| T42 | <elros34> make sure you stop pulseaudio service first | 04:10 |
| sharks_ | Okay, so you're saying I should be running it as `pulseaudio --daemonize=no -n --file=/etc/pulse/default_sailfish.pa` | 04:13 |
| sharks_ | My apologies, I've got no idea what I'm doing | 04:14 |
| sharks_ | The output of that is below: | 04:14 |
| sharks_ | E: [pulseaudio] droid-util.c: Internal data structures are confused. | 04:14 |
| sharks_ | E: [pulseaudio] droid-util.c: Code should not be reached at ../src/common/droid-util.c:278, function update_mapping(). Aborting. | 04:14 |
| sharks_ | Verbose output -> https://paste.opensuse.org/pastes/541b0bec593d | 04:15 |
| sharks_ | Looks like perhaps the issue is with audio_policy_configuration.xml? | 04:16 |
| T42 | <elros34> doesn't look correct, where did you get this args from? Did you read /etc/sysconfig/pulseaudio? | 04:18 |
| T42 | <elros34> do you really have something like default_sailfish.pa? | 04:19 |
| sharks_ | $ cat /etc/sysconfig/pulseaudio | 04:19 |
| sharks_ | HYBRIS_USE_VENDOR_NAMESPACE=yes | 04:19 |
| sharks_ | CONFIG="-n --file=/etc/pulse/default_sailfish.pa" | 04:19 |
| sharks_ | Yes, /etc/pulse/default_sailfish.pa is a real file | 04:20 |
| sharks_ | There is also default.pa and system.pa | 04:20 |
| T42 | <elros34> I don't get is this your change or something I miss in droid-config recent changes? | 04:21 |
| sharks_ | To my knowledge I have not changed anything. I have just checked my Xperia 10 IV and found it uses arm_droid_default.pa, is this what is missing? | 04:22 |
| sharks_ | Hang on, I think I found the problem. I should have copied droid-configs-device/sparse/* as well as droid-configs-device/sparse-11/* into my sparse, shouldn't I? I only did the latter, hence arm_droid_default.pa is missing | 04:26 |
| T42 | <elros34> no, you shouldn't copy andy sparse. .pa name changed recently, just make sure you have matching name in /etc/pulseaudio, use env variable for sysconfig and try again | 04:30 |
| T42 | <elros34> andy* -> any | 04:30 |
| T42 | <elros34> you never copy sparse* from submodule | 04:31 |
| T42 | <elros34> check also logcat | 04:33 |
| sharks_ | So are you saying there should not be hadk/hybris/droid-configs/sparse/etc/pulse/ directory? | 04:33 |
| T42 | <elros34> directory itself is harmless | 04:35 |
| sharks_ | That directory doesn't currently exist. Should it? You're saying I shouldn't copy it in from droid-configs-device/sparse. If I copy it in then the contents on-device will include arm_droid_default.pa | 04:36 |
| T42 | <elros34> if it doesn't exist then what we are talking about? Just don't copy droid-configs-device/sparse* to your device sparse/ and it will be good. If you didn't don't forget about it | 04:38 |
| sharks_ | Fine, I'm sorry. Usually when I want to install an operating system I copy an *.iso to a thumb drive and boot to it. I'm trying something new here and am still learning. Let me try and decipher what you mean by the 'env variable for sysconfig' and I'll get back to you. | 04:39 |
| T42 | <elros34> <sharks_>HYBRIS_USE_VENDOR_NAMESPACE=yes | 04:40 |
| T42 | <elros34> this env variable | 04:40 |
| sharks_ | Okay, I made sure that env variable was set, re-ran the same command as above because we've confirmed default_sailfish.pa exists, and there is no change to the command output. It is still as per the paste.opensuse link above. | 04:42 |
| T42 | <elros34> try to add more -v, did you check logcat? | 04:43 |
| sharks_ | More -v: https://paste.opensuse.org/pastes/1a910a08c3ef | 04:45 |
| sharks_ | Relevant logcat lines: https://paste.opensuse.org/pastes/438cb7215967 | 04:45 |
| T42 | <elros34> do you build 32 or 64 bit sailfish? | 04:47 |
| sharks_ | Armv7hl, 32-bit. This device does not support 64 bits | 04:47 |
| T42 | <elros34> could you paste /vendor/etc/audio_policy_configuration.xml. I guess there is something wrong there but I am not an expert | 04:50 |
| sharks_ | I was looking at this file earlier but don't know enough about it either. Have a look: https://paste.opensuse.org/pastes/8599f60819c7 | 04:51 |
| T42 | <elros34> what about if you remove voip_rx from that file? I am not sure this one fails or next one (more -v maybe will reaveal it) | 04:59 |
| sharks_ | Even more -v: https://paste.opensuse.org/pastes/04c97e13f663 | 05:02 |
| sharks_ | Hang on, I will try removing voip_rx, one moment... | 05:02 |
| sharks_ | Sadly no change as a result of removing the voip_rx mixPort and all references in routes. | 05:10 |
| sharks_ | Are you saying to try the same thing now with the primary input? | 05:11 |
| T42 | <elros34> are you sure you remove it? | 05:12 |
| T42 | <elros34> I mean voip_rx section | 05:13 |
| sharks_ | Yes, I removed the <mixPort name="voip_rx"... all the way to the following </mixPort> | 05:14 |
| sharks_ | I had to remount /system/ as rw, then removed the lines and rebooted for good measure | 05:14 |
| T42 | <elros34> but it's still in pulseaudio output, what about grep voip_rx /vendor/etc/audio_policy_configuration.xml | 05:15 |
| sharks_ | The current pulseaudio output is here, apologies - it is still in the "module options" section but it does not get mapped: https://paste.opensuse.org/pastes/312dddfec506 | 05:17 |
| sharks_ | grep returns no output, its definitely not in the file anymore | 05:18 |
| T42 | <elros34> maybe something in other included xml files. trail and error method until somebody with better pulseaudio understanding will wake up | 05:38 |
| sharks_ | I removed the other xml files (except audio_policy_volumes.xml and default_volume_tables.xml) from audio_policy_configuration.xml so they wouldn't attempt to be loaded in, but that made no improvement either. That leads me to believe the issue is in audio_policy_configuration.xml | 05:47 |
| sharks_ | Thanks for your help so far elros. Hopefully someone else in some other part of the world has a better idea as you say. | 05:47 |
| T42 | <elros34> you can compare xml with some other similar device for anything obvious or gdb it to figure out that update_mapping port_type failure in sources | 05:51 |
| sharks_ | I'll have a quick look on my two other sailfish devices just in case, but I don't think either is similar enough. I'll let you know if I make any progress. | 06:00 |
| T42 | <fxclre> mal: refering to my past problem (with not full repo sync). I decided to do it on another machine, without anything installed, fresh. Here are the errors im getting: https://pastebin.com/yznDkbFN | 14:01 |
| T42 | <fxclre> As i said, okhttp is failing (for no obvious reason), but it says fetching 100%, so can i proceed further? I did repo init on hybris-16.0 | 14:01 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!