| T42_ | <Mister_Magister> @b100dian what fix | 00:09 |
|---|---|---|
| T42_ | <b100dian> Sorry I was literally in airplane mode. Did you search atel? | 01:22 |
| T42_ | <Mister_Magister> no | 01:23 |
| T42_ | <Mister_Magister> there isn't much info in logs either | 01:25 |
| T42_ | <b100dian> https://github.com/sailfishos-sony-nagara/fake-qcrilmsgtunnel | 01:28 |
| T42_ | <b100dian> pihgz: landing changed amazfit hour on my old zeppos now \o/ | 01:37 |
| T42_ | <b100dian> Piggz: didnt help me writing your handle correctly though | 01:38 |
| T42_ | <Mister_Magister> @b100dian on op6 receiving sms works, just not for everybody | 01:55 |
| T42_ | <b100dian> Yup, depends on operator | 01:55 |
| T42_ | <Mister_Magister> ohhh | 01:56 |
| T42_ | <Mister_Magister> so like, you just build it and install it and thats it? | 01:56 |
| T42_ | <adampigg> on here its @adampigg to get a notify :) (re @b100dian: Piggz: didnt help me...) | 08:31 |
| Mister_Magister | back in my day there was piggz on irc >:( | 09:24 |
| T42_ | <adampigg> yeah, matrix bridge / tg allows me to always be here though! | 11:06 |
| T42_ | <branja6> Hey, good folks. I recently purchased a used Xperia 10 II and wanted to flash SFOS on it. After flashing, no matter what I try, there's no sound on the device (the speaker and microphone are not working, although, I have not tried Bluetooth headphones). | 11:49 |
| T42_ | <branja6> Does anyone have any ideas?? | 11:49 |
| T42_ | <branja6> | 11:49 |
| T42_ | <branja6> Btw, I flashed the latest SFOS, on the Android 11 base (I also tried flashing while having 10 and 12 as a base), and have used the recommendes binaries. | 11:49 |
| mal | did you test audio in android before flashing sailfish? | 12:01 |
| T42_ | <branja6> Yeah. Everything is working as expected. | 12:07 |
| mal | get journal log (journalctl -b --no-pager) from sailfish, at least messages about pulseaudio | 12:25 |
| T42_ | <branja6> Will do. As soon as I come home. | 12:46 |
| T42_ | <branja6> I used utilities to reset audio system, this is the journal log after I did that: | 14:02 |
| T42_ | <branja6> | 14:02 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II systemd[2413]: Starting PulseAudio... | 14:02 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II estart[21342]: [D] :37 - Done: Restart | 14:02 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II estart[21342]: [D] actionIsDone:59 - Notify Restart: OK | 14:02 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II pulseaudio[23090]: W: [pulseaudio] pid.c: Stale PID file, overwriting. | 14:02 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II pulseaudio[23090]: library "/vendor/lib64/libsndmonitor.so" not found | 14:02 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II pulseaudio[23090]: library "/vendor/lib64/libssrec.so" not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II pulseaudio[23090]: library "/vendor/lib64/libhfp.so" not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II ngfd[2527]: [58501.423] ERROR: tonegen-ausrv: ausrv: server connection failure: Connection refused | 14:03 |
| T42_ | <branja6> Oct 31 15:00:49 Xperia10II kernel: msm_qti_pp_get_rms_value_control, back not active to query rms be_idx:3 | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: core_get_license_status: cmdrsp_license_result.result = 0x15 for module 0x131ff | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm-ext-disp-audio-codec-rx soc:qcom,msm-ext-disp:qcom,msm-ext-disp-audio-codec-rx: msm_ext_disp_audio_type_get: codec_data, get_audio_edid_blk() or get_intf_id is NULL | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm-ext-disp-audio-codec-rx soc:qcom,msm-ext-disp:qcom,msm-ext-disp-audio-codec-rx: msm_ext_disp_audio_type_get: codec_data, get_audio_edid_blk() or get_intf_id is NULL | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm_pcm_path_latency_ctl_get substream runtime not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm_pcm_path_latency_ctl_get substream runtime not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm_pcm_path_latency_ctl_get substream runtime not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm_pcm_path_latency_ctl_get substream runtime not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm_pcm_path_latency_ctl_get substream runtime not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:50 Xperia10II kernel: msm_pcm_path_latency_ctl_get substream runtime not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II kernel: i2c_geni 4a88000.i2c: i2c error :-107 | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II kernel: wcdcal_hwdep_ioctl_shared: codec didn't set this 0!! | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II kernel: wcdcal_hwdep_ioctl_shared: codec didn't set this 3!! | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II pulseaudio[23090]: "/vendor/lib/soundfx/libqcomvisualizer.so" is 32-bit instead of 64-bit | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II pulseaudio[23090]: library "/vendor/lib/libqtigef.so" not found | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II pulseaudio[23090]: "/vendor/lib/soundfx/libqcompostprocbundle.so" is 32-bit instead of 64-bit | 14:03 |
| T42_ | <branja6> Oct 31 15:00:51 Xperia10II pulseaudio[23090]: "/vendor/lib/hw/sound_trigger.primary.sm6125.so" is 32-bit instead of 64-bit | 14:03 |
| T42_ | <branja6> Oct 31 15:00:58 Xperia10II kernel: perf: interrupt took too long (5599 > 4978), lowering kernel.perf_event_max_sample_rate to 35000 | 14:03 |
| T42_ | <branja6> Oct 31 15:00:59 Xperia10II kernel: BUG: Bad rss-counter state mm:ffffffc0a508b100 idx:4 val:64 | 14:03 |
| T42_ | <branja6> [defaultuser@Xperia10II ~]$ | 14:03 |
| mal | please use pastebin or similar service when you paste more than 3 lines next time | 14:32 |
| T42_ | <branja6> Sure thing. I'm not too familiar with all of this. I apologize | 14:39 |
| T42_ | <branja6> https://pastebin.com/DgjvSj66 | 14:44 |
| T42_ | <adampigg> quite a few missing libraries from the android-vendor side | 14:49 |
| T42_ | <branja6> How could this happen? I followed the instructions correctly. Is there anything I can do to obtain those libraries? | 14:54 |
| mal | those are normal | 14:58 |
| mal | so pulseaudio is running? check output of "ps aux | grep pulseaudio" | 14:58 |
| T42_ | <branja6> 24440 defaultu grep pulseaudio - this is what I'm getting | 15:08 |
| mal | only that? | 15:12 |
| T42_ | <branja6> yup | 15:29 |
| T42_ | <branja6> is there anything else I can check? | 15:32 |
| T42_ | <branja6> when I tried to pulseaudio —kill, this is what I got - "Failed to kill daemon: No such file or directory" | 15:33 |
| T42_ | <branja6> When I tried to start/kill pulseaudio - "WARNING: stream-restore: failed to open connection to pulseaudio: Failed to connect to socket /run/user/100000/pulse/dbus-socket: No such file or directory" | 15:38 |
| mal | how did you try to start it? | 15:48 |
| mal | systemctl --user restart pulseaudio | 15:48 |
| T42_ | <branja6> pulseaudio —start | 15:48 |
| T42_ | <branja6> let me try (re @SailfishFreenodeIRCBridgeBot: <mal>systemctl --use...) | 15:48 |
| mal | and check journal log after that | 15:49 |
| T42_ | <branja6> https://pastebin.com/hpVipQRB | 15:52 |
| mal | that says pulseaudio is running | 15:53 |
| T42_ | <branja6> Is there any other way of turning the volume up from the terminal? I mean, if it's running and everything seems okay, maybe it just needs to turn it up...like with pavucontrol or something. | 15:58 |
| T42_ | <branja6> But, again, why would it be muted from the get-go? | 15:58 |
| mal | check again "ps aux | grep pulseaudio" | 16:00 |
| T42_ | <branja6> [defaultuser@Xperia10II ~]$ ps aux | grep pulseaudio | 16:01 |
| T42_ | <branja6> 26096 defaultu pulseaudio --start | 16:01 |
| T42_ | <branja6> 27004 defaultu grep pulseaudio | 16:01 |
| T42_ | <branja6> | 16:01 |
| T42_ | <branja6> This is all I get. :/ | 16:01 |
| mal | "killall -KILL pulseaudio" | 16:02 |
| mal | then try running the systemctl command again | 16:02 |
| T42_ | <branja6> Let me get this straight... | 16:11 |
| T42_ | <branja6> | 16:11 |
| T42_ | <branja6> 1. killall -KILL pulseaudio | 16:11 |
| T42_ | <branja6> 2. systemctl --user restart pulseaudio | 16:11 |
| T42_ | <branja6> 3. journalctl -b --no-pager | 16:11 |
| T42_ | <branja6> | 16:11 |
| T42_ | <branja6> Right? | 16:11 |
| mal | yes | 16:11 |
| T42_ | <branja6> https://pastebin.com/AckpZBui | 16:15 |
| T42_ | <elros34> could be related https://forum.sailfishos.org/t/8268. But don't give up yet, mal can do miracles | 16:28 |
| mal | maybe try this, first "systemctl --user stop pulseaudio" then remove the folder "/data/vendor/audio" and reboot, after booting check journal again | 16:29 |
| mal | if that doesn't help we'll debug it with different tools | 16:29 |
| T42_ | <branja6> https://pastebin.com/j9sHqtze | 16:36 |
| T42_ | <branja6> darn (re @elros34: could be related htt...) | 16:36 |
| T42_ | <branja6> Thanks for your patience, everyone | 16:37 |
| mal | can you check also output of /system/bin/logcat for anything audio related | 16:38 |
| T42_ | <branja6> https://pastebin.com/pWzbx7Z6 | 16:41 |
| mal | can you crab another journal log, make sure it as similar timestamps as logcat (ignore the hours and focus on minutes and seconds of the timestamp) | 16:46 |
| mal | *grab | 16:46 |
| mal | I need to see both logs from same start attemp of pulseaudio so I can see if there is some timing issue | 16:47 |
| T42_ | <branja6> I'll do it. Give me just a minute or two. | 16:50 |
| mal | also don't filter the logcat, just copy the parts after you try to start pulseaudio, might be something is hidden | 16:52 |
| T42_ | <elros34> not sure what do you prefer but I remember logcat as user instead root might be different | 16:54 |
| T42_ | <branja6> when I type in "/system/bin/logcat", it's showing data from an hour ago. | 17:01 |
| T42_ | <branja6> this is the latest journalctl - https://pastebin.com/2GuWSRhy | 17:02 |
| mal | like I said, ignore the hours | 17:02 |
| mal | there is timezone mismatch in journal vs logcat | 17:02 |
| T42_ | <branja6> I see. Sorry | 17:02 |
| T42_ | <branja6> I'll do it asap | 17:02 |
| T42_ | <branja6> https://pastebin.com/8E7NjBM2 | 17:03 |
| T42_ | <branja6> this should be it | 17:03 |
| T42_ | <branja6> https://pastebin.com/8E7NjBM2 | 17:04 |
| T42_ | <branja6> this is the updated version to include everything | 17:04 |
| mal | install some packages, first "pkcon refresh", then "pkcon install gdb pulseaudio-modules-droid-jb2q-debuginfo" | 17:11 |
| mal | as root | 17:11 |
| mal | (using devel-su) | 17:12 |
| mal | then once those are installed run this as defaultuser: | 17:13 |
| mal | HYBRIS_ENABLE_LINKER_DEBUG_MAP=1 HYBRIS_USE_VENDOR_NAMESPACE=yes gdb --args /usr/bin/pulseaudio --daemonize=no -n --file=/etc/pulse/default_sailfish.pa | 17:14 |
| mal | and once it starts you probably need to type "run" into the terminal that appears | 17:14 |
| mal | and once it crashes type "bt" | 17:14 |
| T42_ | <branja6> just a sec | 17:17 |
| T42_ | <branja6> should I paste the output of everything? | 17:19 |
| mal | just what you get after bt, and maybe couple of lines before that | 17:19 |
| T42_ | <branja6> okay | 17:19 |
| T42_ | <branja6> https://pastebin.com/YLGndRBC | 17:21 |
| mal | you are sure you installed pulseaudio-modules-droid-jb2q-debuginfo ? | 17:21 |
| mal | try running zypper install -C "debuginfo(build-id)=cd8d5ea89bd39f44673ec516e2dd562b2e7c93df" as root | 17:22 |
| mal | wait, in case zypper is not installed first run "pkcon install zypper" | 17:23 |
| T42_ | <branja6> just a sec | 17:24 |
| T42_ | <branja6> wait wait...so, I should type "devel-su zypper install -C "debuginfo(build-id)=cd8d5ea89bd39f44673ec516e2dd562b2e7c93df", right? | 17:25 |
| mal | yes | 17:26 |
| mal | or just devel-su and then in the root terminal the zypper part | 17:27 |
| T42_ | <branja6> sure | 17:27 |
| T42_ | <branja6> should I remove the quotation marks? | 17:28 |
| mal | from what? | 17:28 |
| mal | zypper install -C "debuginfo(build-id)=cd8d5ea89bd39f44673ec516e2dd562b2e7c93df" | 17:29 |
| mal | no | 17:29 |
| T42_ | <branja6> this is what I get - No provider of 'debuginfo(build-id)=cd8d5ea89bd39f44673ec516e2dd562b2e7c93df' found. ['--plus-content debug'?] | 17:29 |
| mal | what does "zypper in pulseaudio-modules-droid-jb2q-debuginfo" say? | 17:30 |
| mal | or maybe those debug symbols are not included, which would be annoying | 17:30 |
| T42_ | <branja6> Package 'pulseaudio-modules-droid-jb2q-debuginfo' not found. | 17:30 |
| mal | you have 5.0 sailfish release? | 17:33 |
| T42_ | <branja6> The latest one provided on the download page. 5.0.0.68 | 17:33 |
| mal | copy the rpms from here and install on your device http://sailfish-fairphone.dy.fi/tmp/ | 17:47 |
| T42_ | <branja6> ok. Now what? | 17:56 |
| mal | try the gdb command again | 17:56 |
| mal | this one: HYBRIS_ENABLE_LINKER_DEBUG_MAP=1 HYBRIS_USE_VENDOR_NAMESPACE=yes gdb --args /usr/bin/pulseaudio --daemonize=no -n --file=/etc/pulse/default_sailfish.pa | 18:04 |
| T42_ | <branja6> followed by run and bt? | 18:06 |
| T42_ | <branja6> https://pastebin.com/qiYiZwf7 | 18:07 |
| mal | where is the crash and bt? | 18:09 |
| mal | I don't see the crash in that | 18:10 |
| T42_ | <branja6> it stopped and then I typed in bt | 18:11 |
| mal | stopped how? | 18:11 |
| T42_ | <branja6> I'll try it one more time, and wait a bit | 18:11 |
| mal | is there a file /data/vendor/audio/cirrus_sony.cal now? | 18:12 |
| T42_ | <branja6> let me try running the command again, and I'll report back shortly | 18:13 |
| T42_ | <branja6> it stopped here, there was no crash : https://irc.thaodan.de/.imgstore/697240b4/file_7849.jpg | 18:17 |
| mal | check if the file I mentioned exists | 18:17 |
| T42_ | <branja6> acdbdata - this is the only thing that lists in that folder | 18:19 |
| mal | hmm | 18:19 |
| mal | if you have pulseaudio running try what happens if you try to play some audio | 18:20 |
| T42_ | <branja6> lets see | 18:20 |
| T42_ | <branja6> I'm playing a YT video. No sound | 18:22 |
| mal | but pulseaudio is running? | 18:22 |
| T42_ | <branja6> I didn't stop | 18:23 |
| T42_ | <branja6> it | 18:24 |
| T42_ | <branja6> I guess it's running | 18:25 |
| mal | stop it, then make sure there are no pulseaudio processes running by checking ps aux | grep pulseaudio | 18:25 |
| T42_ | <branja6> 12637 defaultu grep pulseaudio - this is the only thing it's shown when I type "ps aux | grep pulseaudio" after "killall -KILL pulseaudio" | 18:27 |
| mal | ok then run this: HYBRIS_USE_VENDOR_NAMESPACE=yes /usr/bin/pulseaudio --daemonize=no -n --file=/etc/pulse/default_sailfish.pa | 18:30 |
| mal | does it crash | 18:30 |
| T42_ | <branja6> as a user or root? | 18:30 |
| mal | normal user | 18:32 |
| T42_ | <branja6> it says the file dosn't exist | 18:35 |
| T42_ | <branja6> https://pastebin.com/bBRRQSbL | 18:35 |
| mal | why did it not crash with gdb | 18:35 |
| mal | try again: HYBRIS_ENABLE_LINKER_DEBUG_MAP=1 HYBRIS_USE_VENDOR_NAMESPACE=yes gdb --args /usr/bin/pulseaudio --daemonize=no -n --file=/etc/pulse/default_sailfish.pa | 18:35 |
| T42_ | <branja6> will do | 18:36 |
| mal | hopefully we can get it to crash | 18:36 |
| T42_ | <branja6> it crashed this time...but only because I turned on the screen | 18:37 |
| mal | when you booted the device did you hear any sound? like a frequency sweep | 18:37 |
| T42_ | <branja6> nope...no sounds whatsoever (re @SailfishFreenodeIRCBridgeBot: <mal>when you booted...) | 18:37 |
| mal | that would be during early boot | 18:37 |
| T42_ | <branja6> https://pastebin.com/2W0MqcFV (re @SailfishFreenodeIRCBridgeBot: <mal>try again: HYBR...) | 18:38 |
| T42_ | <branja6> Shoud I reboot it? (re @SailfishFreenodeIRCBridgeBot: <mal>that would be d...) | 18:38 |
| mal | I'm thinking a bit, maybe I should give a patched package to test | 18:38 |
| mal | take the new rpms from here http://sailfish-fairphone.dy.fi/tmp/ install them and reboot | 18:44 |
| T42_ | <branja6> just a sec | 18:47 |
| T42_ | <branja6> Fatal error: nothing provides 'pulseaudio-modules-droid-jb2q-common = 14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla' needed by the to be installed pulseaudio-modules-droid-jb2q-14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla.aarch64 | 18:49 |
| mal | did but the common rpm is there also | 18:50 |
| mal | how did you try to install those? | 18:50 |
| T42_ | <branja6> via terminal pkcon install-local | 18:50 |
| mal | just use rpm -U and then list the files after that | 18:51 |
| T42_ | <branja6> I tried from the phone's file manager first | 18:51 |
| mal | from terminal | 18:51 |
| mal | you need to install at the same time | 18:51 |
| T42_ | <branja6> I'm sorry, what's the command if all the files are in the Download folder? | 18:55 |
| mal | in terminal go to the Downloads folder then check with "ls" the rpm filenames and then devel-su rpm -U some_files | 18:56 |
| mal | where some_files is all of the pulseaudio rpms you downloaded | 18:56 |
| mal | did you understand what to run? | 19:01 |
| T42_ | <branja6> not entirely...this is what I tried devel-su rpm -U pulseaudio-modules-droid-jb2q-14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla.aarch64.rpm | 19:02 |
| T42_ | <branja6> pulseaudio-modules-droid-jb2q-common-14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla.aarch64.rpm | 19:02 |
| T42_ | <branja6> pulseaudio-modules-droid-jb2q-debuginfo-14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla.aarch64.rpm | 19:02 |
| T42_ | <branja6> pulseaudio-modules-droid-jb2q-debugsource-14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla.aarch64.rpm | 19:02 |
| mal | yes, everything in one line | 19:03 |
| T42_ | <branja6> warning: pulseaudio-modules-droid-jb2q-14.2.102+test.20251031184134.1.g785490e-1.8.1.jolla.aarch64.rpm: Header V3 RSA/SHA256 Signature, key ID 47394f23: NOKEY | 19:04 |
| mal | that is normal | 19:06 |
| mal | did it install of the files now? | 19:06 |
| T42_ | <branja6> no | 19:07 |
| T42_ | <branja6> just this warning is all I got | 19:07 |
| mal | so you added the files in one line? | 19:07 |
| mal | you can check with "pkcon search pulseaudio-modules-droid-jb2q" which should show which version have been installed | 19:08 |
| T42_ | <branja6> I did (re @SailfishFreenodeIRCBridgeBot: <mal>so you added th...) | 19:08 |
| T42_ | <branja6> https://pastebin.com/dPCLM7Ut | 19:09 |
| mal | you can see the correct ones have been installed | 19:10 |
| mal | now reboot | 19:10 |
| T42_ | <branja6> will do | 19:10 |
| T42_ | <branja6> What should I do after it boots up? | 19:11 |
| mal | get journal log, also during boot listen whether the device makes any sound | 19:12 |
| T42_ | <branja6> no sounds...I listened this time. I'll get the log (re @SailfishFreenodeIRCBridgeBot: <mal>get journal log...) | 19:13 |
| mal | why does the calibration fail on your device, very odd | 19:13 |
| T42_ | <branja6> https://pastebin.com/u0z6ENTG | 19:14 |
| T42_ | <branja6> I didn't copy everything. Is it needed? | 19:15 |
| mal | no | 19:16 |
| mal | that just shows it fails the same way | 19:16 |
| T42_ | <branja6> darn | 19:16 |
| mal | one possibility is to copy the calibration file from android or just try a file from another x10ii device | 19:19 |
| T42_ | <elros34> if you didn't read the 2 threads on forum there is user with same issue: coping files/ firmware change didn't help | 19:20 |
| mal | @elros34 which files? | 19:21 |
| mal | I was thinking of copying that cirrus_sony.cal file | 19:21 |
| T42_ | <elros34> yeah https://forum.sailfishos.org/t/4-1-0-24-4-2-0-21-xperia-10-ii-xq-au51-single-sim-pulseaudio-dies-with-coredump-no-sounds-in-device/8268/2 | 19:22 |
| mal | hmm | 19:22 |
| T42_ | <elros34> I am bad at gdb but shoudn't you folow this libthread_db related advice from gdb and then use "thread appply all bt"? | 19:24 |
| mal | maybe | 19:26 |
| mal | @branja6 run the gdb command again, before typing "run" type "set auto-load safe-path /" and once it crashes type "thread appply all bt" | 19:27 |
| mal | wait | 19:27 |
| mal | need to think how to do that | 19:28 |
| mal | well try those commands and see what happens | 19:28 |
| T42_ | <branja6> just a sec | 19:28 |
| T42_ | <branja6> https://pastebin.com/vYaDaaRg | 19:35 |
| mal | did you still see this message earlier: warning: File "/lib64/libthread_db-1.0.so" auto-loading has been declined by your ` | 19:36 |
| T42_ | <elros34> check in other terminal "cat /proc/$(pgrep -f pulse)/maps" while bt is still running. IIRC this way you can see what is hidden behind these 0x0000007fb64477a8 cryptic | 19:38 |
| T42_ | <elros34> while gdb* | 19:39 |
| mal | I suspect the crash is in android side library | 19:39 |
| mal | still can't understand what difference there could be in devices | 19:40 |
| T42_ | <branja6> nope...there wasn't mention of the https://libthread_db-1.0.so through the gdb (re @SailfishFreenodeIRCBridgeBot: <mal>did you still s...) | 19:47 |
| mal | ok | 19:49 |
| mal | just out of curiousity let's try something, download the older vendor blobs from sony website https://developer.sony.com/open-source/aosp-on-xperia-open-devices/downloads/software-binaries so the ones matching "Software binaries for AOSP Android 10.0 – Kernel 4.14 – Seine" at least v11 and maybe v10 and v9 also, then flash those manually to device and see what happens, start with v10 and if | 19:51 |
| mal | audio still fails, try one version older | 19:51 |
| mal | flashing can be done with fastboot using "fastboot flash oem_a some_image" where some_image if the .img file of the blobs you are flashing | 19:52 |
| T42_ | <branja6> I didn't know this can be achieved on 10 II | 19:53 |
| mal | I mean start with v11 and go to older if needed | 19:53 |
| mal | ? | 19:53 |
| T42_ | <branja6> I think I tried the v11 and the system didn't let me | 19:53 |
| mal | manually you can flash anything | 19:53 |
| mal | just need to be careful what you do | 19:53 |
| T42_ | <branja6> cool | 19:54 |
| mal | meaning download correct files from that website (android 10 seine ones) | 19:54 |
| T42_ | <branja6> should I flash for both oem_a and oem_b? | 19:55 |
| mal | so only flash the oem_a partition with the blobs, don't use flash.sh script | 19:55 |
| mal | only oem_a | 19:55 |
| T42_ | <branja6> okay | 19:56 |
| T42_ | <branja6> any procedures after flashing and booting? | 19:59 |
| T42_ | <branja6> just flashing the earlier version, right? | 20:00 |
| T42_ | <branja6> i flashed a couple of earlier versions and it didn't help. Maybe I should flash a newer (A11 or 12) version? | 20:19 |
| mal | you can try but not sure if those work | 20:22 |
| T42_ | <branja6> Only one way to find out :D | 20:22 |
| T42_ | <branja6> the latest a11 and 12 versions don't working either | 20:33 |
| T42_ | <branja6> it looks like nothing can't be done about it | 20:33 |
| mal | did everything else seem to work with those other blobs? | 20:36 |
| T42_ | <branja6> from the short testing, it looks like it did (at least, wi-fi, I couldn't test the sim card). | 20:38 |
| mal | which was the older blob you tried? | 20:39 |
| mal | *oldest | 20:39 |
| T42_ | <branja6> v8 | 20:41 |
| mal | that seems to be the oldest available | 20:42 |
| mal | did you try what elros34 suggested about checking "cat /proc/$(pgrep -f pulse)/maps" when you when the gdb command is still running | 20:44 |
| mal | *when you have | 20:44 |
| T42_ | <branja6> nope | 20:46 |
| T42_ | <branja6> is this the exact command - cat /proc/$(pgrep -f pulse)/maps ? | 20:48 |
| mal | yes | 20:57 |
| mal | you then check from the bt command inside gdb the address where it crashes | 20:58 |
| T42_ | <branja6> so...I inter this (HYBRIS_ENABLE_LINKER_DEBUG_MAP=1 HYBRIS_USE_VENDOR_NAMESPACE=yes gdb --args /usr/bin/pulseaudio --daemonize=no -n --file=/etc/pulse/d | 21:01 |
| T42_ | <branja6> efault_sailfish.pa | 21:01 |
| T42_ | <branja6> ) and press run afterward | 21:01 |
| T42_ | <branja6> white that is working, I open a new a new tab in the terminal and enter cat /proc/$(pgrep -f pulse)/maps | 21:02 |
| T42_ | <branja6> is that correct? | 21:02 |
| mal | yes | 21:10 |
| mal | you can wait until it crashes also | 21:10 |
| mal | before running that last command | 21:10 |
| T42_ | <branja6> that command has no output for me...maybe I did it wrong | 21:14 |
| T42_ | <branja6> cat: can't open '4684/maps': No such file or directory | 21:15 |
| mal | hmm | 21:15 |
| mal | why is it missing the /proc part | 21:16 |
| mal | try manually cat /proc/4684/maps | 21:16 |
| T42_ | <branja6> https://pastebin.com/xbqwGMGK | 21:19 |
| T42_ | <branja6> This is after I typed run, and it finished | 21:19 |
| mal | now you just need to show the output of bt | 21:21 |
| T42_ | <branja6> sure | 21:21 |
| T42_ | <branja6> You mean showing the output of bt from the first tab, or output of the cat /proc/4684/maps from the second tab? | 21:23 |
| T42_ | <branja6> after running bt | 21:23 |
| mal | first tab | 21:23 |
| mal | I assume it crashed | 21:23 |
| T42_ | <branja6> (gdb) bt | 21:23 |
| T42_ | <branja6> #0 0x0000000000000000 in ?? () | 21:23 |
| T42_ | <branja6> #1 0x0000007fb646b7a8 in ?? () | 21:23 |
| T42_ | <branja6> #2 0x0000000000000300 in ?? () | 21:23 |
| T42_ | <branja6> Backtrace stopped: previous frame identical to this frame (corrupt stack?) | 21:23 |
| T42_ | <branja6> (gdb) | 21:23 |
| T42_ | <branja6> no changes there | 21:23 |
| mal | so that 0x0000007fb646b7a8 matches 7fb646b000-7fb646f000 r-xp 00003000 fc:00 6909 /vendor/lib64/libcirrusspkrprot.so | 21:24 |
| T42_ | <branja6> I'm not sure what I'm supposed to do here :D | 21:26 |
| mal | we just try to narrow down where it crashes | 21:26 |
| T42_ | <branja6> Imma be back in 10 minutes. Please let me know if there's anything else I can provide you with when I come back. | 21:30 |
| mal | @elros34 any ideas, that library is opensource but would need to build a debug version of it probably | 21:34 |
| T42_ | <elros34> not really I am not that smart. I was following old instruction which helped me debuging some android crash long time ago: https://wiki.ubuntu.com/Touch/Core/UbuntuDebugAndroid | 21:37 |
| mal | yeah, so like I said need to build library manually to get debug symbols for it | 21:40 |
| mal | once the symbols exist that HYBRIS_ENABLE_LINKER_DEBUG_MAP=1 can help get the info to bt | 21:41 |
| mal | I'll try to build that | 21:42 |
| mal | it will take quite a while | 21:42 |
| T42_ | <elros34> this is the right file https://github.com/sonyxperiadev/vendor-qcom-opensource-audio-hal-primary-hal/blob/q-mr1/hal/audio_extn/cirrus_sony.c? | 21:43 |
| mal | yes | 21:44 |
| T42_ | <elros34> uncommenting #define LOG_NDEBUG 0 could be good idea but you definitelly know this. Shame there is no logcat from working device. | 21:51 |
| mal | well I have a working device | 21:51 |
| T42_ | <elros34> ah ok great so you know which errors are not really errors | 22:03 |
| mal | @elros34 the issue is get_ta_array | 22:07 |
| mal | probably | 22:08 |
| mal | yeah, that is the reason but why does tz not work properly | 22:14 |
| mal | at least that causes calibration to not be done | 22:14 |
| T42_ | <elros34> is this infameous sony TA partition? | 22:15 |
| mal | probably | 22:15 |
| mal | what happened to that on some devices | 22:15 |
| mal | that failing device has | 22:17 |
| mal | 10-31 16:57:28.905 4169 4169 E audio_hw_sony_cirrus_playback: get_ta_array: Cannot retrieve TA unit 4702 size: error 5 | 22:17 |
| mal | 10-31 16:57:28.905 4169 4169 I audio_hw_sony_cirrus_playback: get_ta_array: Read TA unit 4702 (size=0) values: 0x0 0x0 0x0 0x0 | 22:17 |
| mal | I have | 22:17 |
| mal | 10-31 22:04:19.939 4082 4082 I audio_hw_sony_cirrus_playback: get_ta_array: Read TA unit 4702 (size=4) values: 0x0 0x0 0x0 0x16 | 22:17 |
| T42_ | <elros34> so android or more likely aosp should also fail same? | 22:22 |
| mal | I would assume so but official android had working audio | 22:23 |
| mal | unless it does something very different, which is possibly | 22:23 |
| T42_ | <elros34> so this is TA_CIRRUS_CAL_GLOBAL_CAL_AMBIENT and go figure what's next:) | 22:24 |
| mal | it tries to read TA using some blob it loads | 22:25 |
| mal | miscta_get_unit_size = dlsym(ta_handle, "miscta_get_unit_size"); from libMiscTaWrapper.so | 22:25 |
| T42_ | <elros34> so assumin that fails same in aosp then maybe some developer from opendevices could help but it's quiet old device | 22:31 |
| mal | but still not sure what could cause TA partition to break only on some devices | 22:32 |
| T42_ | <elros34> according to this and my little understanding there is layer under layer under layer https://github.com/sonyxperiadev/vendor-qcom-opensource-audio-hal-primary-hal/commit/7e4c8601e13b6b26b534fef04e39797ed875940b libMiscTaWrapper -> binder -> misctacals | 22:38 |
| mal | wondering if the other end is tad_static or ta_qmi_service | 22:39 |
| mal | or vendor.somc.hardware.miscta@1.0-service | 22:39 |
| mal | would be interesting to strace those and see if anything happens in any of those | 22:40 |
| mal | I'll test | 22:40 |
| mal | I see activity in both vendor.somc.hardware.miscta@1.0-service and tad_static | 22:43 |
| T42_ | <elros34> @branja6 There is little chance that getting logcat as root user could potentially shed some lights on it. Maybe some failing services "ta" related | 22:43 |
| mal | the first one seems to read tad socket so probably tad_static is the one that does the real reading | 22:44 |
| mal | @branja6 check that you have the processes vendor.somc.hardware.miscta@1.0-service and tad_static running using "ps aux | grep ta" | 22:45 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
