**** BEGIN LOGGING AT Wed Jan 13 02:59:57 2021
Jan 13 07:43:38 <Tofe> Morning!
Jan 13 07:44:15 <Tofe> Herrie: and can even be further simplified, as they also provide a native implementation which will work for recent androids (or maybe even the older ones)
Jan 13 07:52:48 <Herrie> Tofe: yeah wasn't sure you saw it so just pasted it here for reference
Jan 13 07:54:54 <Tofe> Herrie: iirc, nyx isn't supposed to use Qt, right?
Jan 13 07:55:00 <Tofe> as it's lower level stuff
Jan 13 07:56:19 <Tofe> and the vibrator is mainly used through LS2, i.e. through nyx, skipping Qt
Jan 13 07:56:44 <Tofe> however, we could very well take their native implementation and put it into ours
Jan 13 07:58:08 <Herrie> Yeah I don't think nyx works in Qt, nizovn might know
Jan 13 07:58:13 <Tofe> speaking of which, isn't there already a non-hybris nyx vibrator module ?
Jan 13 07:58:37 <Herrie> nizovn: ^
Jan 13 07:59:47 <Tofe> https://github.com/webOS-ports/nyx-modules/blob/webOS-ports/webOS-OSE/src/haptics_timedoutput/haptics.c here
Jan 13 08:01:26 <Tofe> so we need to change things like https://github.com/webOS-ports/meta-webos-ports/blob/gatesgarth/meta-luneos/recipes-webos/nyx-modules/nyx-modules/rosy.cmake#L34 and disable the hybris haptic module
Jan 13 08:03:18 <Herrie> Tofe: Ah yes
Jan 13 08:03:32 <Herrie> So it could be something even more simple to fix?
Jan 13 08:03:50 <Tofe> yes, we already have the code !
Jan 13 08:03:57 <Tofe> we'll need to make https://github.com/webOS-ports/nyx-modules-hybris/blob/master/CMakeLists.txt#L33 more flexible
Jan 13 08:11:31 <Herrie> Tofe: Any thoughts on why my oFono doesn't like the RIL ?
Jan 13 08:11:34 <Herrie> 'Via binder
Jan 13 08:12:26 <Tofe> Herrie: can you paste your .conf ?
Jan 13 08:12:40 <Tofe> maybe there's a typo or something
Jan 13 08:13:20 <Tofe> though "Unknown RIL transport: binder " would mean that it doesn't find the ril-binder plugin for some reason
Jan 13 08:14:19 <JaMa> morning
Jan 13 08:21:30 <Tofe> Hey JaMa ! long time no see
Jan 13 08:21:57 <Tofe> We've been playing with Halium 9.0 these last few weeks
Jan 13 09:21:30 <Herrie> JaMa: Morning!
Jan 13 09:21:41 <Herrie> I remember you had issues with Tissot as well?
Jan 13 09:22:03 <Herrie> I have issues with touchscreen, just cannot figure our how to fix it...
Jan 13 09:22:10 <Herrie> Seems touch coordinates are off
Jan 13 09:24:16 <Herrie> Tofe: Don't we need to add it in depends somewhere or in packageconfig?
Jan 13 09:24:19 <Herrie> For the ril-binder?
Jan 13 09:27:01 <Tofe> on it
Jan 13 09:31:23 <Tofe> Herrie: done
Jan 13 09:31:49 <Herrie> See :P
Jan 13 09:31:55 <Tofe> "off", what do you mean ? offsetted ?
Jan 13 09:32:06 <Herrie> Tofe: Yeah like wrong coordinates
Jan 13 09:32:42 <Herrie> In such a way it's completely not possible to do anything...
Jan 13 09:40:12 <Herrie> Even in TWRP
Jan 13 09:40:28 <Herrie> Tried flashing factory 7.1.2 image, latest one from Xiaomi to no avail
Jan 13 09:44:23 <Tofe> mmh weird... firmware issue ?
Jan 13 09:47:03 <Herrie> Probably
Jan 13 09:48:27 <Herrie> Seems so, but factory reset should solve that
Jan 13 09:48:31 <Herrie> Which it didn
Jan 13 09:48:32 <Herrie> 't
Jan 13 09:48:39 <Herrie> I recall JaMa also had some issues
Jan 13 09:59:02 <Herrie> Tofe: I noticed that the boot of Android part seems a bit slow, but you noticed that also right?
Jan 13 09:59:08 <Herrie> The wait-for-android stuff probably
Jan 13 10:11:29 <Tofe> Herrie: wait-for-android probably doesn't work anymore
Jan 13 10:11:58 <Tofe> actually it's pretty sure :)
Jan 13 10:12:25 <Tofe> as we used to patch on the fly the anroid init.rc, and now it's read-only because it's system-as-root
Jan 13 10:13:10 <Tofe> but we can probably change the shell to take benefit from the other way implemented in halium 9 + patches
Jan 13 10:13:18 <Tofe> (through properties)
Jan 13 10:13:58 <Tofe> so you're just experiencing a systemd timeout for the .sh
Jan 13 10:14:18 <Tofe> just remove it from the android-system.service file
Jan 13 10:14:32 <Herrie> Yeah I noticed they did some things in init now
Jan 13 10:14:58 <Herrie> Also reboot takes very long, but that issue was there before too
Jan 13 10:20:18 <Tofe> reboot is probably some process that's not responding to TERM signal
Jan 13 10:23:19 <JaMa> Herrie: yes, I had the issues with android on tissot as well, in the logs it was complaining about the firmware version of the TS itself, don't remember the details, but IIRC I got it working only after reflashing older version and then going back to newer which finally updated the fw correctly, before it was stuck in some inconsistent state
Jan 13 10:24:18 <Herrie> JaMa: Hmmz OK.. I tried oldest & newest stock but no luck for some reason
Jan 13 10:24:23 <Herrie> Still have off offsets in TWRP
Jan 13 10:29:32 <JaMa> in my case it looked like this:
Jan 13 10:29:35 <JaMa> 2019-03-21.log:16:32 <+JaMa> dmesg.tissot_images_7.12.19_7.1:[    1.740509] ft5435_ts 3-0038: Firmware version = 10.0.0
Jan 13 10:29:38 <JaMa> 2019-03-21.log:16:32 <+JaMa> dmesg.tissot_images_7.12.19_7.1:[    1.740676] ft5435_ts 3-0038: Current firmware: 0x0a.0.0
Jan 13 10:29:41 <JaMa> 2019-03-21.log:16:32 <+JaMa> dmesg.tissot_images_7.12.19_7.1:[    1.740681] ft5435_ts 3-0038: New firmware: 0x10.0.0
Jan 13 10:29:44 <JaMa> 2019-03-21.log:16:32 <+JaMa> dmesg.tissot_images_7.12.19_7.1:[   12.450463] ft5435_ts 3-0038: Firmware version = 16.0.0
Jan 13 10:30:38 <JaMa> 16:33 <+JaMa> future JaMa: bad TS fw is 0x0a, good one is 0x10 and 7.12.19_7.1 is what you need to flash
Jan 13 10:38:51 <Herrie> JaMa: Thnx that's helpful
Jan 13 10:40:26 <Herrie> [    1.091069] ft5435_ts 3-0038: Firmware version = 10.0.0
Jan 13 10:40:26 <Herrie> [    1.091211] [Fu]fw_vendor_id=0x3b
Jan 13 10:40:27 <Herrie> [    1.091215] upgrade,fts_fw_vendor_id=0x3b
Jan 13 10:40:27 <Herrie> [    1.091219] ft5435_fw_upgrade_by_array_data, suspended=0
Jan 13 10:40:27 <Herrie> [    1.091225] ft5435_ts 3-0038: Current firmware: 0x0a.0.0
Jan 13 10:40:27 <Herrie> [    1.091229] ft5435_ts 3-0038: New firmware: 0x0a.0.0
Jan 13 10:40:27 <Herrie> [    1.091234] ft5435_ts 3-0038: Exiting fw upgrade...
Jan 13 10:40:39 <Herrie> That's what I have
Jan 13 10:40:52 <Herrie> In my TWRP dmesg
Jan 13 10:42:05 <Herrie> JaMa: YOu don't happen to have that version laying around?
Jan 13 10:42:59 <Herrie> 7.12.19_7.1 I mean
Jan 13 10:50:34 <JaMa> yes tissot_images_7.12.19_20171219.0000.00_7.1_d39fa89aaf.tgz
Jan 13 10:50:53 <JaMa> https://androidfilehost.com/?fid=673791459329072901
Jan 13 10:52:33 <Herrie> Thnx
Jan 13 10:52:44 <JaMa> md5 of my .tgz d39fa89aafcaaaa2203324313c964706 , if it cannot be downloaded from here or has different checksum let me know will upload it somewhere
Jan 13 10:52:58 <Herrie> Well I guess it should be OK
Jan 13 10:53:44 <Herrie> I tried 7.8.23_7.1 and latest 9.0 from Xiaomi without results
Jan 13 10:53:47 <Herrie> So good to have this one
Jan 13 10:54:04 <Herrie> Can get the Halium 9.0 to build for it as well, so can work with Tofe on the same target :P
Jan 13 10:54:14 <Herrie> Easier v.s. Mido and Rosy at the same time
Jan 13 10:54:22 <Herrie> Though they are pretty similar
Jan 13 10:54:29 <Herrie> And we're almost at the same state
Jan 13 10:56:26 <Tofe> Herrie: as long as we're not blocked, it's a bit more practical to have a halium build ready at home
Jan 13 10:57:22 <Tofe> I still have two daemon (camera and time_daemon) spamming a bit my logs
Jan 13 10:57:31 <Herrie> Yeah time_daemon here too
Jan 13 10:57:34 <Tofe> and the firmware loading is sometimes crashing the phone
Jan 13 10:58:02 <Herrie> I'm missing adsp firmware it seems
Jan 13 10:58:07 <Herrie> Which gives me a lot of spam
Jan 13 10:59:40 <Herrie> Sorry msadp that should be
Jan 13 10:59:47 <Herrie> Jan 12 11:15:42 mido ueventd: firmware: could not find firmware for msadp
Jan 13 11:00:21 <Tofe> I have /android/vendor/firmware_mnt/image/adsp.b00
Jan 13 11:01:05 <Tofe> Ah, I have wifi too -- I forgot to unmask the wifi module service :p
Jan 13 11:01:14 <Herrie> Hhee
Jan 13 11:01:17 <Herrie> Wifi is OK for me
Jan 13 11:01:23 <Herrie> Status a bit flaky at times, but it connects
Jan 13 11:01:26 <Herrie> We had similar on 7.1
Jan 13 11:01:35 <Herrie> In FirstUse, but it's merely cosmetics, it connects
Jan 13 11:02:06 <Herrie> OK Tissot is back working
Jan 13 11:02:14 <Herrie> JaMa: ^ Thanks :)
Jan 13 11:02:37 <Tofe> mmh I don't have bluetooth though
Jan 13 11:02:49 <Tofe> but that's fine for now. Sound is more important.
Jan 13 11:03:11 <Herrie> Well mine scans & connects
Jan 13 11:05:48 <Tofe> I think I found the sound change I talked about: "Thaodan: either use pulseaudio-modules-droid-glue+audioflingerglue OR use pulseaudio-modules-droid-hidl+audiosystem-passthrough"
Jan 13 11:06:38 <JaMa> cool, let me crawl back to bed :)
Jan 13 11:07:28 <Tofe> Herrie: for sound, I think we can rely on that https://github.com/mer-hybris/audiosystem-passthrough/blob/master/README.md
Jan 13 11:11:20 <Tofe> JaMa: tired at noon ? or in different TZ ?
Jan 13 11:14:54 <Herrie> Tofe: Yeah seems decent
Jan 13 11:25:26 <Herrie> Seeing Mer is being used on "Production" devices, I'm happy to piggyback on their work :P
Jan 13 11:41:24 <JaMa> Tofe: still being sick a bit and sleepy a lot
Jan 13 13:04:43 <Herrie> Tofe: Since my Tissot is up and running, will try a quick build for it
Jan 13 13:05:16 <Tofe> :) ok
Jan 13 13:12:07 <Herrie> Should be easy now hopefully since we're using the same common ;)
Jan 13 13:14:10 <Herrie> Interesting would be the 4.9 kernel
Jan 13 13:14:17 <Herrie> That should make Anbox easier as well ;)
Jan 13 13:32:54 <Herrie> Halium build OK, no vendor.img but a large system.img it seems
Jan 13 13:38:32 <Tofe> system.img contains /vendor then ?
Jan 13 13:49:19 <Herrie> Didn't check yet
Jan 13 13:49:24 <Herrie> Building image now, will flash and see
Jan 13 13:55:28 <Tofe> looks like lineageos for tissot isn't treble-compatible, they don't separate vendor.img
Jan 13 13:55:44 <Tofe> maybe because it's a A/B device, and there's only one "cust" partition
Jan 13 14:04:55 <Herrie> Yes LOS isn't Treble it seems
Jan 13 14:05:05 <Herrie> Hmmzm black screen, no ADB
Jan 13 14:05:24 <Herrie> Though also no reboot
Jan 13 14:05:27 <Herrie> I'll wait a bit
Jan 13 14:07:30 <Herrie> Could be I need to tweak my defconfig a bit still
Jan 13 14:11:47 <Herrie> Ah I was using the LuciferKerne;l
Jan 13 14:12:01 <Herrie> Not helping ;)
Jan 13 14:15:24 <Herrie> OK found an issue... Seems my fork was still using 3.18 kernel not 4.9 which device tree specifies 4.9 for Tissot
Jan 13 14:15:29 <Herrie> So they got out of sync somewhere
Jan 13 14:15:32 <Herrie> Retry :P
Jan 13 14:24:55 <Tofe> hoepfully we're getting good at it
Jan 13 14:31:25 <Herrie> Tofe: Well if this kernel check script has the right flags it's very quick :P
Jan 13 14:31:44 <Herrie> 100's of defconfig settings updated in a second :p
Jan 13 14:32:37 <Herrie> 315 changes in LOS defconfig :P
Jan 13 14:41:59 <Herrie> Kernel doesn't like to compile, so cherry picking a few fixes from Lucifer
Jan 13 15:06:37 <Herrie> Too many changes it seems
Jan 13 15:06:50 <Herrie> Let's try luciferkenel with standard tissot_defconfig instead of the tweaked one
Jan 13 15:52:10 <Herrie> Hmmz failing on do_configure :S
Jan 13 15:52:17 <Herrie> Taking 15 mins and then failing
Jan 13 15:52:20 <Herrie> Maybe just need to reboot
Jan 13 16:05:58 <Herrie> Nope :S
Jan 13 16:06:24 <Tofe> mmh
Jan 13 16:06:59 <Herrie> Log from do_configure: 101GB :S
Jan 13 16:07:01 <Herrie> WTF
Jan 13 16:07:44 <Herrie> Per run
Jan 13 16:07:48 <Herrie> Let me do cleanall
Jan 13 16:07:51 <Herrie> That's weird
Jan 13 16:08:23 <Tofe> ahah, wow
Jan 13 16:08:27 <Herrie> Nope
Jan 13 16:08:31 <Herrie> 1GB within seconds :S
Jan 13 16:08:49 <Tofe> what does it contain ? must be the same thing over and over ?
Jan 13 16:09:33 <Herrie> Too big to check :P
Jan 13 16:09:43 <Herrie> Probably some loop somehow
Jan 13 16:09:46 <Tofe> with "less", "more" or even vim
Jan 13 16:14:19 <Herrie> https://imgur.com/a/viK88au
Jan 13 16:14:23 <Herrie> Seems the bitmask stuff
Jan 13 16:22:55 <Tofe> must be a bug in the yocto script
Jan 13 16:26:21 <Herrie> Probably somehow
Jan 13 16:26:23 <Herrie> Weird
Jan 13 16:26:27 <Herrie> Just with this defconfig...
Jan 13 16:26:33 <Herrie> Let me try our 7.1 defconfig....
Jan 13 16:27:09 <Herrie> Though it's 3.18 kernel
Jan 13 16:27:12 <Herrie> Probably not a good idea
Jan 13 16:29:39 <Tofe> Herrie: I suggest you remove any occurrence of CONFIG_LITTLE_CPU_MASK and retry
Jan 13 16:33:08 <Herrie> Well I took LOS defconfig with changes now
Jan 13 16:33:12 <Herrie> For Halium
Jan 13 16:33:15 <Herrie> And see
Jan 13 16:33:21 <Herrie> If thta doesn't work will try that
Jan 13 16:42:49 <Tofe> ok, heading back home...
Jan 13 16:43:29 <Tofe> ... ok, done
Jan 13 16:55:50 <Herrie> Hmmz same, weird stuff
Jan 13 17:04:18 <Tofe> Herrie: I... I just heard a sound ?! and I didn't touch anything regarding the pulseaudio studd
Jan 13 17:04:20 <Tofe> stuff
Jan 13 17:04:40 <Tofe> I've just modified wait-for-android to make it work with the new init
Jan 13 17:07:51 <Herrie> Tofe: I had sound from the start on Mido
Jan 13 17:08:35 <Herrie> Tofe: After I added these 5 config values (taken from Lucifer defconfig) it at least starts to compile again https://github.com/Herrie82/LuciferKernel/commit/b4a08a4dbbb8b7386049d212bfbca267563105be
Jan 13 17:10:27 <Tofe> ok, quite weird behavior
Jan 13 17:10:30 <Herrie> Yup
Jan 13 17:10:37 <Herrie> Fails with some redefinitions now :S
Jan 13 17:12:26 <Tofe> I've pushed the latest stuff to meta-smartphone
Jan 13 17:25:32 <Herrie> Tofe: nice
Jan 13 17:26:10 <Tofe> yes, the waiting script looks good now
Jan 13 17:26:38 <Tofe> could the persist mount explain your firmware issues ?...
Jan 13 17:26:48 <Tofe> I think you already had it
Jan 13 17:32:20 <Tofe> mmh I have the same error as this one: https://review.lineageos.org/c/LineageOS/android_device_motorola_griffin/+/241011
Jan 13 17:33:31 <Tofe> looking at the pointed change in Android https://android.googlesource.com/platform/frameworks/native/+/47f876b74be0a2d6d417f5ff385f85d835a97fe9%5E%21/#F4 I wonder if this error is a consequence of having added BOARD_VNDK_VERSION and/or BOARD_VNDK_RUNTIME_DISABLE
Jan 13 18:08:06 <Tofe> Herrie: https://github.com/webOS-ports/nyx-modules-hybris/pull/6/commits/8fc4296392cf2b5a34345ab70718e919a3717d44 and https://github.com/Tofee/meta-webos-ports/commit/2a02246a2e850cacbacbaaed4dc72ab72a155504 should work; I've migrated only rosy.cmake
Jan 13 19:13:00 <Herrie> Tofe: Nice for the fixes, I'll check
Jan 13 19:13:08 <Herrie> My Mido didn't build with the VNDK flags
Jan 13 19:13:36 <Herrie> Since Mido/Tissot/Rosy share the same common, it's probably in my device tree somewhere
Jan 13 19:14:56 <Herrie> Tofe: PR https://github.com/webOS-ports/nyx-modules-hybris/pull/6/files still includes the commit from yesterday. Is that still needed?
Jan 13 19:16:03 <Tofe> Herrie: well, it's pretty sure it'll crash anyway on Halium 9+, so...
Jan 13 19:16:21 <Herrie> Ah OK
Jan 13 19:16:24 <Herrie> I thought you fixed it?
Jan 13 19:16:38 <Tofe> no, I deactivated it :)
Jan 13 19:16:42 <Tofe> with this commit
Jan 13 19:17:20 <Herrie> Ah OK
Jan 13 19:17:21 <Tofe> but in the end, maybe it's better to let it crash ?
Jan 13 19:17:32 <Tofe> so that we don't search why nothing vibrates
Jan 13 19:17:36 <Herrie> I don't know...
Jan 13 19:17:56 <Herrie> As long as it does it gracefully and not SEGFAULTS ;)
Jan 13 19:18:43 <Tofe> I'll leave the commit as-is for now
Jan 13 19:19:00 <Herrie> OK
Jan 13 19:19:10 <Herrie> So vibration works or doesn't now?
Jan 13 19:19:13 <Herrie> Not clear to me
Jan 13 19:23:04 <Herrie> Just started a build ;)
Jan 13 19:23:17 <Tofe> I didn't rebuild yet, but with my commit, it should work *on rosy*
Jan 13 19:25:25 <Herrie> Yeah I kniw
Jan 13 19:25:32 <Herrie> I did the same change but for Mido now ;)
Jan 13 19:25:56 <Herrie> Head was tortured by 3 running and shouting kids today, so brain is a bit boiled LOL
Jan 13 19:29:05 <Tofe> at least, you can relax and look at those beautiful stairs
Jan 13 19:29:21 <Herrie> Yup ;)
Jan 13 19:29:28 <Herrie> Which reminds me I should do some more painting LOL
Jan 13 19:29:31 <Herrie> But it's final bits now
Jan 13 19:29:34 <Herrie> And a lot less
Jan 13 19:29:46 <Herrie> Touch up of damages on wall etc
Jan 13 19:29:51 <Herrie> So a quick roller does the trick
Jan 13 19:29:55 <Herrie> Spots here and there
Jan 13 19:30:02 <Herrie> And some door frames so all small stuff :)
Jan 13 19:30:18 <Herrie> And a few wooden panels, but it's altogether < 1hr work I guess
Jan 13 19:30:32 <Herrie> So will first try a new build and have some tea
Jan 13 19:31:34 <Tofe> I'm rebuilding halium from scratch, but without the VNDK stuff, to see if it still work as well as before -- or better
Jan 13 19:32:02 <Herrie> Might be good to do a file compare of the 2 images
Jan 13 19:32:08 <Herrie> See what's different inside
Jan 13 19:33:45 <Herrie> BTW I love how quickly images install on the Xiaomi devices :D
Jan 13 19:35:56 <Tofe> yes, that's quite nice !
Jan 13 19:36:58 <Herrie> Hmmz mine reboots :S
Jan 13 19:38:43 <Herrie> And now boots :S
Jan 13 19:39:11 <Herrie> And now crashes in firstuse
Jan 13 19:39:13 <Herrie> Weird
Jan 13 19:39:47 <Herrie> Maybe it didn't flash properly somehow
Jan 13 19:48:02 <Herrie> 2nd flash it's behaving better
Jan 13 19:48:08 <Herrie> No vibration though in Testr
Jan 13 20:23:55 <Tofe> ah
Jan 13 20:24:15 <Herrie> OK testing some changes in the common for time_daemon and adsprpcd
Jan 13 20:24:25 <Herrie> I.e.: https://github.com/Herrie82/android_device_xiaomi_msm8953-common/commits/halium-9.0
Jan 13 20:28:45 <Herrie> Tofe: I get "com.palm.vibrate" is not available it seems
Jan 13 20:29:23 <Herrie> At least in UI
Jan 13 20:30:35 <Herrie> In /var/log/messages: 2021-01-13T19:50:57.960607Z [514.801420113] user.warning ls-hubd [] ls-hubd LSHUB_NO_SERVICE {} _LSHubSendQueryNameReplyMessage: Failed Connecting to Service err_code: -3, service_name: "com.palm.vibrate", unique_name: "(null)", static, fd -1
Jan 13 20:31:27 <Herrie> Tofe: Ah
Jan 13 20:31:38 <Herrie> I need to revert my "fix" in luna-sysmgr :P
Jan 13 20:33:12 <Herrie> Where I disabled the whole castle haptics bits :P
Jan 13 20:33:42 <Herrie> That's quick enough to test
Jan 13 20:36:57 <Herrie> OK installed new IPK, rebooting
Jan 13 20:39:19 <Herrie> Ok now it works for 1/3 of the tests
Jan 13 20:39:40 <Herrie> "500" 0.2S doesn't work, "1000" 0.4S also not, but "300" 0.6S works
Jan 13 20:40:45 <Tofe> ok, well, it's better than nothing I guess
Jan 13 20:41:47 <Herrie> And journal for all 3 calls: https://paste.ubuntu.com/p/G4c8MqG5Df/
Jan 13 20:41:53 <Herrie> Tofe: Yeah and it works somewhat
Jan 13 20:41:56 <Tofe> the native nyx code probably needs some attention
Jan 13 20:44:41 <Tofe> I've rebuilt my halium, now let's see if it's better or worse
Jan 13 20:48:01 <Herrie> My Halium also ready
Jan 13 20:48:05 <Herrie> Redoing image and will test too
Jan 13 20:48:10 <Herrie> See if I got rid of the logspam :P
Jan 13 20:48:12 <Tofe> it boots as well as before :)
Jan 13 20:48:28 <Herrie> In both journalctl and logcat
Jan 13 20:48:36 <Herrie> That would help to debug stuff a bit
Jan 13 20:53:55 <Tofe> looks like I'm missing some things: libts_face_beautify_hal.so , libtime_genoff.so , libsdm-disp-vndapis.so
Jan 13 20:54:38 <Herrie> face_beautify is OK to miss
Jan 13 20:54:44 <Herrie> It's one of the things we don't build
Jan 13 20:54:54 <Herrie> Because it requires Java bits
Jan 13 20:55:02 <Herrie> At least that's my initial thought with it
Jan 13 20:55:42 <Tofe> well, it's present in lineage's vendor bits
Jan 13 20:55:55 <Tofe> and require by some camera service
Jan 13 20:55:59 <Tofe> required*
Jan 13 20:56:18 <Herrie> I think Snap uses it which we removed: https://github.com/Herrie82/android_device_xiaomi_msm8953-common/commit/d074102d607bef455a8f3a8f0cc28b37938097a0
Jan 13 20:56:22 <Herrie> If memory serves me well
Jan 13 20:56:52 <Tofe> I'll just add the library itself
Jan 13 20:57:12 <Herrie> Anyway it shouldn't be to hard to add if needed
Jan 13 20:57:33 <Tofe> sure
Jan 13 20:58:12 <Herrie> Flashing new image here
Jan 13 20:59:56 <Tofe> ah, it's just a proprietary file
Jan 13 21:06:39 <Tofe> Herrie: when a proprietary file is missing, what's the best fix ? add it to the device-vendor.mk file ?
Jan 13 21:07:38 <Tofe> there's this proprietary-files.txt, and I never know which one is the right source
Jan 13 21:08:49 <Herrie> Well it's a bit tricky
Jan 13 21:09:04 <Herrie> Since the Halium setup script does "magic" ;)
Jan 13 21:09:14 <Herrie> I.e. is will regenerate based upon the .txt
Jan 13 21:09:45 <Tofe> ok, so it's the .txt that is important
Jan 13 21:09:59 <Herrie> https://github.com/Halium/halium-devices/blob/halium-7.1/setup#L77
Jan 13 21:10:22 <Herrie> If my mind serves me well, it will take the .txt and then regenerate the .mk ;)
Jan 13 21:11:35 <Herrie> But sometimes you need to change things in both...
Jan 13 21:11:49 <Herrie> It's always hit & miss with me too
Jan 13 21:12:01 <Herrie> With poor device trees it's usually problematic
Jan 13 21:12:05 <Tofe> well I'll never be able to push the .mk
Jan 13 21:12:10 <Herrie> The LOS ones are OK usually
Jan 13 21:14:21 <Herrie> My adsprpc spam is gone ;)
Jan 13 21:14:33 <Tofe> so maybe it was the persist mount ?
Jan 13 21:15:16 <Herrie> Tofe: No I disabled it in the init.rc ;)
Jan 13 21:17:38 <Tofe> all mounts are disabled with the Halium patches
Jan 13 21:18:02 <Tofe> oh you mean adsprpc
Jan 13 21:18:53 <Herrie> Yes
Jan 13 21:19:05 <Herrie> That one was spamming my journalctl & logcat
Jan 13 21:19:10 <Herrie> Logcat is nice and clean now: https://paste.ubuntu.com/p/5dGVSJ5wVR/
Jan 13 21:19:18 <Herrie> Killed time_daemon as well ;)
Jan 13 21:19:50 <Tofe> good idea :)
Jan 13 21:22:42 <Herrie> 09-20 10:16:17.409   145   145 E QSEECOMD: : Init dlopen(libspl.so, RLTD_NOW) is failed.... dlopen failed: library "libspl.so" not found
Jan 13 21:22:51 <Herrie> Still not sure what Qseecomd does exactly
Jan 13 21:22:56 <Herrie> I think I killed it in 7.1
Jan 13 21:24:06 <Tofe> it's not just libspl.so that's missing in vendor?
Jan 13 21:25:02 <Herrie> Tofe: Yeah it probably is
Jan 13 21:25:23 <Herrie> Also I have this: 09-20 10:16:17.640   156   156 I LocSvc_xtra2: dlopen for libqdma_file_agent.so failed, error = dlopen failed: library "libqdma_file_agent.so" not found
Jan 13 21:26:31 <Herrie> Seems it's not there for 8953, I do see it for sdm845 and msm8998 :S
Jan 13 21:26:40 <Herrie> libspl.so that is
Jan 13 21:28:18 <Herrie> Ah here's some description of what qseecom does: https://github.com/Halium/android_kernel_fairphone_msm8974/blob/8772974506c3b9909d0fb28f85f98a5fc06b67b5/Documentation/devicetree/bindings/qseecom/qseecom.txt
Jan 13 21:35:23 <Tofe> I don't really see why we don't copy all the vendor/lib and vendor/lib64 content
Jan 13 22:10:04 <Herrie> Tofe: Yeah might be a good thing to do
Jan 13 22:10:20 <Herrie> Probably just being lazy :P
Jan 13 22:10:34 <Herrie> Because it's generated from the proprietary-files.txt and you'd need to update it completely
Jan 13 22:16:46 <Tofe> ok now the only spam left is time_daemon
Jan 13 23:34:45 <Herrie> Tofe: Well I killed that one at my end
Jan 13 23:34:51 <Herrie> When you would re-build it's gone ;)
Jan 13 23:35:03 <Herrie> Since it's in 8953-common
**** ENDING LOGGING AT Thu Jan 14 02:59:57 2021