| *** amccarthy is now known as Guest28979 | 09:48 | |
| *** amccarthy_ is now known as amccarthy | 09:48 | |
| T42_ | <adampigg> fun bug ... on quintus sometimes after a call, general audio stops working. call audio still works though. restarting pq doenst help, need to compare pactl list | 13:56 |
|---|---|---|
| T42_ | <adampigg> seems like the profile gets stuck in voicecall, not defualt | 14:32 |
| rinigus | @b100dian: interesting fix to our BT issues is to apply a quirk to VHCI driver: set_bit(HCI_QUIRK_NO_SUSPEND_NOTIFIER, &hdev->quirks); | 18:06 |
| T42_ | <TheVancedGamer> oh hey seems like you finally figured it out :D | 18:06 |
| rinigus | testing it right now. so far suspend worked without devices connected and with headphones connected | 18:07 |
| mal | rinigus: which issue was that? | 18:08 |
| T42_ | <TheVancedGamer> BT would cause device to not suspend | 18:08 |
| T42_ | <TheVancedGamer> same fix as mine on FP5 | 18:08 |
| rinigus | mal: when enabling BT, nagara wouldn't suspend. | 18:08 |
| rinigus | damn! @TheVancedGamer - somehow missed your fix | 18:09 |
| T42_ | <TheVancedGamer> I've had it for quite a while now actually | 18:09 |
| T42_ | <TheVancedGamer> i also sent the kernel here a while ago :P | 18:09 |
| rinigus | I should have followed it | 18:09 |
| rinigus | but it makes sense if you look into all logs. | 18:09 |
| T42_ | <TheVancedGamer> I also tried to optimize it further by tweaking suspend notifier itself | 18:10 |
| T42_ | <TheVancedGamer> but I haven't followed up on it yet | 18:10 |
| rinigus | @TheVancedGamer: I wonder whether BT stack switches to lower power mode on suspend. or whether we should notify HAL somehow | 18:11 |
| rinigus | probably android sources have notification sorted out there | 18:11 |
| T42_ | <TheVancedGamer> technically there is no notification | 18:12 |
| T42_ | <TheVancedGamer> the HAL only serves as a layer to send HCI/ISO data | 18:12 |
| rinigus | what we had was notification from kernel to HCI -> bluebinder -> HAL. and that killed suspend. I guess same for you | 18:13 |
| rinigus | or am I mixing terms? | 18:13 |
| T42_ | <TheVancedGamer> the problem is that Linux will initiate a disconnect for all BT devices on suspend | 18:14 |
| T42_ | <TheVancedGamer> not the HAL itself | 18:14 |
| T42_ | <TheVancedGamer> since all BT is handled in userspace in android, they don't have to worry about this, because the OS will wake up whenever it needs to process BT data | 18:14 |
| rinigus | but in this respect that quirk is "problem solved" | 18:15 |
| T42_ | <TheVancedGamer> yup | 18:15 |
| rinigus | how did you patch your FP5 builds? commit ID? would be happy to refer to that from the issue :) | 18:15 |
| mal | was that suspend issue only when bt was used or always? | 18:16 |
| rinigus | mal: always | 18:16 |
| rinigus | mal: https://github.com/sailfishos-sony-nagara/main/issues/38 | 18:17 |
| T42_ | <TheVancedGamer> https://gitlab.com/ubports/porting/reference-device-ports/android11/fairphone-5/kernel-fairphone-qcm6490/-/blob/halium-11.0-rebase/backports/drivers/bluetooth/hci_vhci.c?ref_type=heads#L128 | 18:18 |
| T42_ | <TheVancedGamer> rinigus: I did the change back when I backported the newer BT stack | 18:18 |
| T42_ | <TheVancedGamer> almost... 5 months ago? damn | 18:18 |
| mal | what is that HCI_QUIRK_BROKEN_ERR_DATA_REPORTING ? | 18:18 |
| T42_ | <TheVancedGamer> a quirk needed for newer BT stacks, because some quirks got inverted | 18:19 |
| T42_ | <TheVancedGamer> it was defaulted to broken before, now it needs to be exclusively selected | 18:19 |
| rinigus | yes, saw that invert in mainline kernel. was looking through commits over there | 18:20 |
| mal | any idea if the need for that HCI_QUIRK_NO_SUSPEND_NOTIFIER is specific to device or kernel version or something else? | 18:24 |
| T42_ | <TheVancedGamer> that is specific to all android | 18:24 |
| rinigus | I am rather surprised that it has not been used throughout on ports | 18:26 |
| T42_ | <TheVancedGamer> nobody really works on bluetooth | 18:26 |
| mal | I don't see that being used in any other kernel and those don't have suspend issues | 18:26 |
| T42_ | <TheVancedGamer> the only reason I put in so much effort was for Pebbles to work on UT | 18:26 |
| T42_ | <TheVancedGamer> mal: it's the flag invert issue | 18:26 |
| mal | even that HCI_QUIRK_NO_SUSPEND_NOTIFIER ? | 18:27 |
| T42_ | <TheVancedGamer> all quirks were made opt-out iirc | 18:27 |
| T42_ | <TheVancedGamer> in a later kernel | 18:27 |
| mal | so it's kernel version specific then | 18:27 |
| mal | i.e. after some version it's needed | 18:27 |
| rinigus | nagara is on 5.10 | 18:28 |
| mal | maybe that is the key reason then, for example x10iv and v have 5.4 kernel | 18:30 |
| mal | I'll check one device to see if it has that issue | 18:31 |
| mal | I remember it having suspend issues | 18:31 |
| T42_ | <abranson> Did you get anywhere with the BLE in Rockwork in the end? I'd love to get Rockpool working with the new Pebble 2 Duo. I guess the Time 2 will be easier as I guess it will still support classic BT (re @TheVancedGamer: the only reason I pu...) | 18:42 |
| T42_ | <Hty909> Forwarded from Пермь 360°: *В Пермском крае женщина из ревности облила соседку кипятком во время совместного застолья. * | 18:42 |
| T42_ | <Hty909> | 18:42 |
| T42_ | <Hty909> 58-летняя пермячка увидела, как её соперница положила голову на плечо знакомого, и в приступе ревности совершила нападение. Пострадавшая скончалась в больнице от ожогов. | 18:42 |
| T42_ | <Hty909> | 18:42 |
| T42_ | <Hty909> Возбуждено уголовное дело по статье «Умышленное причинение тяжкого вреда здоровью, повлекшее смерть потерпевшего». | 18:42 |
| T42_ | <Hty909> | 18:42 |
| T42_ | <Hty909> Пермь 360° (https://t.me/+g4rQFE49g8Q4MTM6) | 18:42 |
| T42_ | <TheVancedGamer> yup, I've shipped the BLE app already | 18:42 |
| T42_ | <TheVancedGamer> and the Time 2 won't support RFCOMM | 18:42 |
| T42_ | <TheVancedGamer> it's BLE only | 18:42 |
| T42_ | <abranson> boo | 18:42 |
| T42_ | <abranson> smashing news about ble though! i haven't checked your repo lately. | 18:43 |
| T42_ | <TheVancedGamer> too bad I never got my beta P2D | 18:43 |
| T42_ | <TheVancedGamer> would've been nice | 18:43 |
| T42_ | <abranson> it's really funny seeing all these ios users celebrating being allowed to use their pebbles again after seven years or so | 18:44 |
| T42_ | <abranson> i've got a time 2 on pre-order. it really is amazing it's all coming back. | 18:44 |
| T42_ | <TheVancedGamer> yup, same | 18:44 |
| T42_ | <TheVancedGamer> I was promised a devkit by Eric but that went pretty much nowhere | 18:45 |
| T42_ | <abranson> i saw that libpebble3 is kotlin :( | 18:45 |
| T42_ | <TheVancedGamer> who needs that anyway :) | 18:45 |
| T42_ | <TheVancedGamer> my implementation is pretty much completed anyway | 18:45 |
| T42_ | <TheVancedGamer> had all missing features (health, voice, etc) before the firmware was released | 18:46 |
| T42_ | <abranson> yeah I saw you'd merged our stuff in and added a lot. great progress! would be much better to have a common libpebble though, written in a language that's not ultra platform specific. | 18:49 |
| T42_ | <TheVancedGamer> I'd say Qt/C++ is very much platform-agnostic | 18:50 |
| T42_ | <abranson> I mean the kotlin... | 18:50 |
| T42_ | <TheVancedGamer> and I mean we can make *ours* the common libpebble :P | 18:51 |
| T42_ | <TheVancedGamer> I wanted to make a rockwork for android | 18:51 |
| T42_ | <abranson> now that's the spirit :D | 18:51 |
| T42_ | <TheVancedGamer> you could join the rockwork group if you're interested btw | 18:51 |
| T42_ | <abranson> I think I did. on the discord? | 18:52 |
| T42_ | <TheVancedGamer> nope, we have ours on telegram | 18:52 |
| T42_ | <TheVancedGamer> t.me/rockwork_ut | 18:52 |
| T42_ | <b100dian> rinigus: wow great! | 19:28 |
| T42_ | <b100dian> @TheVancedGamer I do remember I read through your backport diff, but it was huge :)) I might have missed the direction | 19:28 |
| rinigus | @b100dian: so, kernel rebuild time and testing it :) | 19:29 |
| rinigus | or are you on LOS and trying to fix esim? | 19:29 |
| T42_ | <b100dian> no, not at all. I am not even close :) I think I should re-try to add secondary sim in qcrild directly under sailfish and then re-read https://lucaweiss.eu/post/2024-06-24-esim-manager-for-mobile-linux/ | 19:30 |
| T42_ | <b100dian> but truly, in this stanza I don't even have a bt device to test, left the airpod trackers behind | 19:31 |
| T42_ | <b100dian> So it is just a one-line kernel change rinigus? | 19:32 |
| rinigus | @b100dian: yes, one-liner | 19:33 |
| T42_ | <b100dian> https://github.com/sony-nagara/android_kernel_sony_sm8450/commit/e0cbff362295bfb886c8e2526ea20d8db0458379 GG | 19:33 |
| rinigus | that's the one. git pull away | 19:33 |
| T42_ | <b100dian> rinigus: so are you going to ship mce with default = external, so I must build the android suspend patches, that true? | 20:13 |
| T42_ | <b100dian> Curious since you mentioned there might not be a difference after all - did you have more measurements or it is just for the possibility of it | 20:14 |
| T42_ | <ZweiD> I am looking for a second hand 10 V atm but seems like the 5 IV is the better device. Should I look for the 5 IV can it be a daily? (re @SailfishFreenodeIRCBridgeBot: <rinigus>that's the ...) | 22:45 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!