| T42 | <TheVancedGamer> mal: has fast dormancy been implemented for ofono in SFOS? | 15:28 |
|---|---|---|
| mal | hmm, not sure | 15:38 |
| mal | need to check what that is | 15:38 |
| thevancedgamer | basically low power mode for modems | 15:39 |
| mal | we have some low power mode stuff there | 15:39 |
| thevancedgamer | yeah but that's just indication filters | 15:40 |
| thevancedgamer | the actual device state methods aren't implemented | 15:40 |
| mal | we do use RADIO_IND_FILTER_DATA_CALL_DORMANCY | 15:40 |
| thevancedgamer | https://cs.android.com/android/platform/superproject/main/+/main:hardware/interfaces/radio/1.0/IRadio.hal;l=1553 | 15:40 |
| thevancedgamer | android uses this too | 15:40 |
| mal | https://github.com/mer-hybris/ofono-binder-plugin/blob/master/src/binder_devmon_ds.c#L177 | 15:41 |
| mal | we do send device state | 15:41 |
| thevancedgamer | aha | 15:42 |
| thevancedgamer | hmm | 15:42 |
| thevancedgamer | I've noticed really bad battery drain on FP5 with a SIM installed, maybe we don't properly handle MCE events | 15:42 |
| thevancedgamer | need to try | 15:42 |
| mal | did you find anything for the wlan issue? | 15:43 |
| mal | need to test that modem stuff with wlan off and sim in | 15:43 |
| thevancedgamer | I did some kernel hacks and wifi started working again, no idea if that fixed it or something else | 15:43 |
| thevancedgamer | I can send it | 15:43 |
| mal | yeah, since it reboots always for me it should be easy to see if that fixes it | 15:44 |
| thevancedgamer | https://github.com/mer-hybris/ofono-binder-plugin/blob/master/src/binder_plugin.c#L1741 | 15:44 |
| thevancedgamer | do I need to enable devmon_ds in ofono config? | 15:44 |
| mal | thevancedgamer: https://github.com/mer-hybris/ofono-binder-plugin/blob/master/src/binder_plugin.c#L176 seems to be all by default | 15:47 |
| mal | but if display state is not reaching ofono then it might not handle things correctly | 15:48 |
| mal | debug log should tell if it sends device state | 15:48 |
| thevancedgamer | what is actually the difference between devmon_if and devmon_ds? | 15:48 |
| mal | those control different things | 15:48 |
| mal | if = indication filter control, ds = device state control | 15:49 |
| T42 | <TheVancedGamer> ah, ok | 15:50 |
| T42 | <TheVancedGamer> https://paste.myself5.de/hecizoseti.diff | 15:56 |
| T42 | <TheVancedGamer> mal: this is what i have | 15:56 |
| mal | thevancedgamer: what is that CONFIG_LOG_BUF_SHIFT ? | 16:03 |
| mal | is that needed? | 16:03 |
| T42 | <TheVancedGamer> mal: that's for debugging, you can omit it if you want | 16:09 |
| T42 | <TheVancedGamer> to increase dmesg size | 16:09 |
| mal | started kernel build now, will test later today and probably overnight if it doesn't cause reboot | 16:11 |
| thevancedgamer | ok | 16:21 |
| mal | I also need to see how fp4 behaves | 16:25 |
| mal | need to find fp4, too many phones everywhere | 16:26 |
| thevancedgamer | hmmm i'm quite confused as to why i'm suddenly draining 2.3% per hour on FP5 | 16:26 |
| thevancedgamer | I guess modem pulls too much power | 16:26 |
| thevancedgamer | or mce events aren't working | 16:26 |
| mal | is the battery drain with or without mobile data enabled? | 16:29 |
| T42 | <TheVancedGamer> without | 16:29 |
| T42 | <TheVancedGamer> wifi connected | 16:29 |
| T42 | <TheVancedGamer> without a sim i'm getting 0.3-0.5% per hour | 16:30 |
| mal | ok, I will test without wifi with one sim card | 16:30 |
| T42 | <TheVancedGamer> maybe try that patch first and see if it fixes wlan? :p | 16:30 |
| mal | still building | 16:30 |
| mal | I started wifi battery testing with fp4 also | 16:30 |
| thevancedgamer | it takes 3 minutes to build for me | 16:31 |
| thevancedgamer | unless you're doing full LTO I guess | 16:31 |
| mal | ah, it had built, just the terminal was in in copy mode and didn't update | 16:31 |
| thevancedgamer | :D | 16:32 |
| mal | thevancedgamer: patch didn't help | 16:45 |
| T42 | <TheVancedGamer> what the heck | 16:48 |
| T42 | <TheVancedGamer> can you send me the journal? | 16:48 |
| mal | thevancedgamer: https://pastebin.com/FPpf1P0b | 16:50 |
| T42 | <TheVancedGamer> oh, you have pstore functioning? | 16:51 |
| T42 | <TheVancedGamer> interesting | 16:51 |
| mal | ? | 16:52 |
| T42 | <TheVancedGamer> I thought FP5 had the same pstore issue as FP4 | 16:52 |
| mal | that is from ssh using dmesg -w | 16:53 |
| T42 | <TheVancedGamer> ``` | 16:55 |
| T42 | <TheVancedGamer> [ 42.737106] [kworke][0x41045a32][16:44:48.358880] wlan: [336:E:QDF] Cannot malloc 0 bytes @ policy_mgr_init_dbs_hw_mode:771``` | 16:55 |
| T42 | <TheVancedGamer> this is interesting | 16:55 |
| mal | you don't have that? | 16:56 |
| T42 | <TheVancedGamer> checking | 17:12 |
| T42 | <TheVancedGamer> interestingly there's no kernel bug or anything | 17:12 |
| T42 | <TheVancedGamer> mal: what if you drop my last commit (the one that loads WLAN driver without manual invokation) and try it again? | 17:13 |
| mal | ok | 17:50 |
| mal | @TheVancedGamer fp4 with wlan on has lost about 1% of battery in about 1.5 hours | 17:52 |
| mal | could be less since I didn't know how close it was to the threshold | 17:52 |
| T42 | <TheVancedGamer> yeah, that's fine | 17:54 |
| T42 | <TheVancedGamer> without SIM? | 17:54 |
| mal | yes | 17:54 |
| T42 | <TheVancedGamer> hmm | 17:54 |
| mal | just for comparing to fp5 wlan | 17:54 |
| T42 | <TheVancedGamer> I get about 1% every 2 hours | 17:54 |
| mal | with the fix? | 17:54 |
| T42 | <TheVancedGamer> but I have a smartwatch connected too | 17:54 |
| T42 | <TheVancedGamer> yeah | 17:54 |
| mal | ok so about similar maybe, will test overnight | 17:54 |
| T42 | <TheVancedGamer> I know it wakes up every 15-30 minutes in my case, because the watch dumps health and analytics | 17:55 |
| mal | I think that kind of wakeup is usual anyway | 17:55 |
| T42 | <TheVancedGamer> probably yeah | 17:55 |
| mal | there is some timer I think to wake up sometimes | 17:55 |
| T42 | <TheVancedGamer> need to fix this WLAN issue though, so weird | 17:55 |
| mal | did you manage to reproduce it how often? | 17:56 |
| T42 | <TheVancedGamer> I reproduced it thrice, got home, tried to reproduce again and it just wouldn't | 17:56 |
| mal | so odd why it's so different for me | 17:57 |
| T42 | <TheVancedGamer> I think it's some sort of race condition | 17:57 |
| T42 | <TheVancedGamer> if you can disable auto loading we can rule that out too | 17:57 |
| mal | @TheVancedGamer reverting that didn't help, I do have the other patch from today though, should I revert that also? | 18:27 |
| T42 | <TheVancedGamer> yeah | 18:28 |
| T42 | <TheVancedGamer> so now you have to do the echo ON stuff to load driver, right? | 18:28 |
| mal | hmm, now wlan worked after reboot | 18:29 |
| mal | so maybe it's less bad now but need to see | 18:29 |
| mal | I have systemd service to echo ON to /dev/wlan | 18:30 |
| T42 | <TheVancedGamer> yeah, it's a firmware load race condition | 18:30 |
| T42 | <TheVancedGamer> I think | 18:30 |
| T42 | <TheVancedGamer> if I let the phone sit for 3-4 minutes and then turned on wlan, it magically started working | 18:31 |
| T42 | <TheVancedGamer> and then kept working | 18:31 |
| mal | hmm | 18:31 |
| mal | we also use firmware loading script from userspace | 18:31 |
| mal | if kernel loading fails | 18:32 |
| T42 | <TheVancedGamer> it might not be firmware, but that's my guess | 18:32 |
| T42 | <TheVancedGamer> we don't have a stack trace either | 18:33 |
| mal | but probably some race condition since it was very reproducible until I reverted the autoloading commit | 18:33 |
| mal | I had always reboot before that | 18:34 |
| T42 | <TheVancedGamer> need to figure it out | 18:34 |
| mal | yes, it's annoying there is no clear error | 18:34 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!