| *** Mister_M- is now known as Mister_Magister | 01:38 | |
| *** Mister_M- is now known as Mister_Magister | 01:47 | |
| *** Mister_Magister_ is now known as Mister_Magister | 02:07 | |
| *** Mister_Magister_ is now known as Mister_Magister | 02:15 | |
| *** Mister_M- is now known as Mister_Magister | 02:35 | |
| T42 | <edp_17> Thaodan : Good morning. This is how ngfd.service looks like: https://paste.ubuntu.com/p/fJKrS5gB6T/ | 05:28 |
|---|---|---|
| T42 | <edp_17> And here is the journal : https://paste.ubuntu.com/p/FzWSX5MCqG/ | 05:28 |
| T42 | <edp_17> Please use this journal as in here I've tried to use vibra with csd : https://paste.ubuntu.com/p/V7HGHgykv6/ | 05:34 |
| T42 | <TheVancedGamer> /blocklist KuCoins | 07:01 |
| T42 | <elros34> @edp_17 did you notice this line: "native-vibrator: could not open native vibra control" | 09:01 |
| T42 | <edp_17> @elros34 : Yes, I did but don't know why. I'll search in the archive for this. | 09:08 |
| T42 | <elros34> probably open (duration_path, O_WRONLY) fails, there is no error message. Either path doesnt exist yet or most likely it's permission issue | 09:10 |
| T42 | <elros34> grep for vibra in whole system to check whether something change it permissions: udev rules, init*rc. | 09:11 |
| T42 | <edp_17> @elros34 : I only found /init.rc : https://paste.ubuntu.com/p/mfRPjS8ztZ/ | 09:31 |
| T42 | <edp_17> I don't have "/sys/class/leds/vibrator" | 09:34 |
| T42 | <elros34> doesn't matter, init.rc is not device specific. Maybe you can try to add 'ExecStartPre=/bin/ls -l /sys/class/timed_output/vibrator/' to ngfd.service | 09:46 |
| T42 | <edp_17> @elros34 : I've added and rebooted. Journal : https://paste.ubuntu.com/p/m39js3kC7n/ | 10:15 |
| T42 | <elros34> so it's clear now. In your grep command you only search in /init*rc but there are many more rc file in /system /vendor and even /usr/libexec | 10:19 |
| T42 | <edp_17> There is only ueventd.rc in /system/vendor and that doesn't have anything vibra related. And there is only another one in /usr/libexec. : https://paste.ubuntu.com/p/xrFGydxD6y/ | 10:42 |
| T42 | <elros34> not sure, ngfd is started after droid-hal-init which set permission to timed_output. My bet is there must be other init*rc file which overrides probable in /system | 10:48 |
| T42 | <edp_17> Okay, I do a find on /system. | 10:49 |
| T42 | <elros34> "grep for vibra in whole system" means whole / not only /vendor or /init*rc | 10:49 |
| T42 | <edp_17> Okay, should I look for vibra or timed_output ? | 10:51 |
| T42 | <elros34> doesn' t matter if you look at the path | 10:51 |
| T42 | <markvesime> @calebccff if I flashed your enchilada port to a poco f1 would it work? | 11:09 |
| T42 | <calebccff> no | 11:10 |
| T42 | <markvesime> pity! Thanks | 11:11 |
| T42 | <Verevka86> Hello, the microphone in the recorder works but does not work during a phone call, tell me how can I fix this? | 13:09 |
| T42 | <Verevka86> Hybris18.1 | 13:09 |
| T42 | <Verevka86> https://github.com/SailfishOS-vayu | 13:09 |
| T42 | <b100dian> Disable android.hardware.audio.service? | 13:31 |
| T42 | <edp_17> @elros34 : The grep is still running. :( So far it hasn't found anything. | 13:32 |
| T42 | <edp_17> I am thinking wouldn't be quicker/easier to re-starting of ngfd at a later point. (I know this is hacky.) | 13:33 |
| T42 | <elros34> you mean since few hours? Then it hang at some /dev or /proc. even grep -rni vibrator /system/etc may be enough | 13:36 |
| T42 | <elros34> @Verevka86 have you checked that you have no started mentioned previously audio service: 'pgrep -af audio'? | 13:38 |
| T42 | <Verevka86> Yes (re @b100dian: Disable android.hard...) | 13:38 |
| T42 | <edp_17> Yes, since you suggested to do it. (over two and half an hours ago) | 13:38 |
| T42 | <elros34> you should stop it after 5 minutes max | 13:39 |
| T42 | <edp_17> Ok, I stop it and only do /system/etc | 13:39 |
| T42 | <edp_17> In /system/etc it didn't find anything that uses "/sys/class/timed_output/vibrator/" | 13:40 |
| T42 | <edp_17> However, these are using vibrator : https://paste.ubuntu.com/p/Nq8jt5nS6H/ | 13:41 |
| mal | also check /usr/libexec/droid-hybris/system/etc/ and /*.rc | 13:42 |
| Thaodan | for easy grepping in those run find /usr/libexec/droid-hybris/system /system /vendor -name \*.rc -exec grep --color=auto -nH {} \; | 14:01 |
| T42 | <edp_17> @elros34 : Here they are (I think they all fine): https://paste.ubuntu.com/p/xpvN53KDSF/ | 14:05 |
| T42 | <edp_17> This last one meant to be for mal | 14:06 |
| T42 | <elros34> according to your logs ngfd starts about 4s after droid-hal-init finished so I guess it should already set. Do you have "processing action (boot) from " in dmesg and at which point of boot? | 14:08 |
| T42 | <Verevka86> https://paste.ubuntu.com/p/nTzpxp5GqK/ (re @elros34: @Verevka86 have you ...) | 14:11 |
| T42 | <edp_17> @elros34 : Yes, I do have that : https://paste.ubuntu.com/p/DvM2SWkVns/ | 14:12 |
| T42 | <Verevka86> @elros34 also in the bluetooth headset, the microphone works when making a phone call | 14:14 |
| T42 | <edp_17> @elros34 : In the current dmesg since last boot, there is no "processing action (boot) from" : https://paste.ubuntu.com/p/4Vks4KXVwc/ | 14:19 |
| T42 | <edp_17> However, in previous ones, that's there. | 14:20 |
| T42 | <edp_17> In here, for example: https://paste.ubuntu.com/p/nxfCr73JnP/ | 14:21 |
| T42 | <edp_17> Ah, I see why you asked. At those lines an *.rc file is initiated and I need to check them for whether any of them changes vibra's permission. | 14:23 |
| T42 | <elros34> because you did not wait long enough, use dmesg_logger without arguments and wait until it appears then compare with ngfd start up time to figure out whether chown happens after or before | 14:27 |
| T42 | <elros34> chown should be in ~28s after obot | 14:28 |
| T42 | <elros34> boot* | 14:28 |
| T42 | <elros34> or 38s | 14:28 |
| T42 | <edp_17> How do you mean didn't wait long enough? The device hasn't been rebooted for almost 5 hours. | 14:46 |
| T42 | <edp_17> The only difference between this and last boot, in this time I have not restarted ngfd.service. | 14:46 |
| T42 | <elros34> I only meant this log: https://irclogs.sailfishos.org/logs/%23sailfishos-porters/latest.log.html#t2022-05-20T14:19:57 | 14:47 |
| T42 | <edp_17> I've gained that log I posted from dmesg_logger that's been logging everything since yesterday. I though that's logging contsantly all dmesg message. Therefore, I though it should have recorded everything that happened since last reboot (over 4 hours ago). | 14:51 |
| T42 | <elros34> ok forget about it, it doesn't really matter, just figure out when chown happens. If after ngfd starts then you will add systemd service doing same thing before ngfd starts and that is all | 14:52 |
| T42 | <edp_17> Okay. | 14:53 |
| T42 | <edp_17> Okay, and how can I add systemd service doing same thing before ngfd starts? | 15:01 |
| T42 | <edp_17> I mean, what should I add and where to? | 15:01 |
| T42 | <elros34> there are many examples all over the place in official or community droid-configs, here: https://github.com/mer-hybris/droid-config-f5121/blob/master/sparse/usr/lib/systemd/system/zram.service | 15:05 |
| T42 | <edp_17> Oh, so you were talking about a new service that starts before ngfd. Ithough need to modify ngfd.service to run something before it starts ngfd. | 15:08 |
| T42 | <elros34> that will also work: systemctl-user edit ngfd.service, or you could probably use udev rules too, or use droid-hal-early script and so on. | 15:09 |
| T42 | <edp_17> Okay, and how I can add this into the build? (A new service is okay, but if I modify an existing one.) | 15:10 |
| T42 | <elros34> systemctl edit will create override which you can copy to /etc/systemd/user/ | 15:11 |
| T42 | <edp_17> So if sparse/etc/systemd/user I add a service with the same name that I have in /usr/lib/systemd/user, the one from /etc will be used? | 15:15 |
| T42 | <elros34> that will work too but systemctl edit is even better because you can change only part of the file | 15:16 |
| T42 | <edp_17> Yes, but that only works on device. I am talking about for new builds. I try that systsemctl edit to see what it's doing. | 15:17 |
| T42 | <edp_17> I've added to chown and chmod into droid-hal-early script but didn't help. | 15:29 |
| T42 | <elros34> have you confirmed it works like previously with ls command in ngfd.service? | 15:31 |
| T42 | <edp_17> No. | 15:31 |
| Thaodan | edp_17: Yes but first use overrides to not override the whole systemd unit. | 15:36 |
| T42 | <edp_17> How can I override if need to comment out something? | 15:36 |
| Thaodan | read systemd.unit and search for drop-in. | 15:37 |
| Thaodan | *man systemd.unit | 15:37 |
| T42 | <edp_17> Thanks. | 15:37 |
| T42 | <elros34> I suggest to focus at 1 issue then you can choose fix it "right way". | 15:38 |
| T42 | <edp_17> I've created a script that runs the chown and chmod in ExecStartPre of ngfd.service but didn't help. Perhaps, that runs as user and I need root permission to set those. I guess. I gave up for today. Thanks for your help so far! | 15:56 |
| T42 | <Spidey24Z> Hello | 17:01 |
| T42 | <Spidey24Z> Is hwc2-fallback available on sfos? | 17:03 |
| T42 | <Spidey24Z> Is hwc2-fallback package available on sfos? (edited) | 17:04 |
| T42 | <m49ux> Hi, what are the best low price phones to use with sailfish OS? | 17:07 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!