Mikahlo commented 6 years ago

A continuation of #10, #18, #19, #20, #21, #22, #23 and #24

Spookcity commented 5 years ago

@MikeTeh Yeah that is so strange. I really have no idea what it could be or why it works without issue for me and not you. Also not sure why just this particular ROM. Just one of those bizarre things I guess. Happy weekend to you as well!

rodman01 commented 5 years ago

hi @Spookcity do you know whether and how I can force or verify the reboots? I have added graphics.composer to smdk3470-common.mk and do not have reboots so far. Also the ui feels a bit smoother and faster. I tried videos from youtube and google play film, they are working and no reboots so far. So I am not sure whether the include changed something. Also I am wondering whether we maybe need also additional the hwcomposer.xxx? How would be the right file name?

Spookcity commented 5 years ago

Hey @rodman01 First,thanks a lot for taking the time to work on this. I really appreciate it. I thought about perhaps needing to add a hwc.cpp (or multiple) cpp files to our repo as I've seen this in other devices. I didn't get any further than thinking about it though haha. As far as forcing or verifying reboots I'm really not sure. It seems to be rather sporadic. I've had about 5 over the last week. All seem to be when I'm using the browser,but nothing more specific really and I haven't been able to grab a log. If you don't apply This patch it will be very easy though. you will get reboots anytime you unlock your sim,try to sign into WiFi,or if you go to settings _> about phone and then android version and click it a few times to bring up the P easter egg you should trigger a reboot with this as well. I just did a build with These patches applies by hand. I have tried before and they didnt fix it,but I was hoping maybe along with the libhwc2on1adapter patch maybe that would do the trick. I have not tested yet though.

rodman01 commented 5 years ago

Many thanks for the info and yes such reboots (most of them were soft reboots as far as I remember) I also noticed some times....Yesterday I mixed up something during my changes and what broke my wlan. So I couldn't get a connection anymore gggr :-( so I reverted everything step by step and sure how it is in most of such cases, only the very last revert brought my wifi back. Anyways I ll add graphics composer service and a few other things and give it a new try :-) I will also add the commits from leos you linked. Will see whether we could narrow down the problem a bit. To all: I will upload my RR Testversion and if it makes sense a few test roms too, ist someone here who could test it (them) to verify whether something has changed and maybe fixed? Or should I better post it on xda in the RR thread?

rodman01 commented 5 years ago

OK I have made an RR Test build with some changes and its uploaded. I haven't tested it much and not grapped a logcat yet, so if somebody has the some time and motivation, I would be glad if we can get some info. As for now, I have no major issues. What I did and changed (will update my github sources, once I have sorted out the commits and when there are no bugs and changes needs to be reverted):

synched hw samsung with latest leos commits
in hidl manifest: included impl level to generic at graphics allocator
in hidl manifest: changed graphic composer to hwbinder
in hidl manifest: included impl level to generic at hardware wifi
in common.mk: included graphics composer but not in hidl manifest (I think then it makes no sense at all :-)
in overlay config.xml included some gps and wifi softapp, as well as hwc overlays, also dozealwaysonenabled it set to false
in overlay powerprofile included some settings found on other device, cannot remember which one sorry
included this 2 commits spook mentioned
included the wifi patch form here

Hope I haven't forgotten to mention something :-)

So please test WIFI stability, for example disconnecting after going in deep sleep), GPS (I am getting gps fix after less than a minute with gps status, but only after reboot after a clean installation) and of course possible reboots caused by HWC

I did a complete new flash and installed also opengapps nano, please note that mobile data is disabled by default in system.prop. Rom is based, beside the changes above, on last weeks RR and LEOS sources.

And here is the link, so download and have fun :-) TEST_RR-P-v7.0.2-20190526-kminilte-Unofficial.zip: https://www.androidfilehost.com/?fid=6006931924117881841

Spookcity commented 5 years ago

@rodman01 I will gladly test out your RR build. I cant really test WiFi as I have never had any issues with it,but I will test other things. As far as the SkipValidate patches. Sadly they do not change anything. I had a soft reboot when I was out and about watching YouTube videos via the browser app. I then got home and connected to adb to try to reproduce this and of course couldn't. It just wouldn't reboot. Its a very random issue it seems. Thanks again for your work. I will let you know my findings.

Please do share your updated github sources whenver you have the time cause I will add some of your changes to my main repos if all seems OK. Thanks again. It is really appreciated!

rodman01 commented 5 years ago

Yes seems that you are right. Youtube browser app ist starting and videos also begin to play As soon as I tab on full screen mode, the app freezes and doesn't go back. App stops and there is a kind of soft reboot. I will revert the 2 commits and make a new build and upload this is a few. Google play movie instead is changing to full screen and film is running but reverting back to the film list is not smooth, its more or similar to freezing.

Spookcity commented 5 years ago

Yeah,I'm not sure what the cause is. I will try to grab a log later.... First things Ive noticed about your RR build in just a minute of using... LiveDisplay works! Did you do something specific for this? Is it perhaps just the updated hardware/samsung sources. I couldn't get this to work for anything. Also noticed that the camera works (as well as before) I thought I remember the main reason i switched our graphic composer to passthrough was because hwbinder broke the camera. Either way,it seems to work now! I will test more and report back later.

rodman01 commented 5 years ago

No I added nothing about livedisplay additional to the latest leos commits. What I also added to boardconfigcommon.mk are this lines which I took from here, but I guess that they are useless anyways because not relevant for our device. `USE_OPENGL_RENDERER := true NUM_FRAMEBUFFER_SURFACE_BUFFERS := 3

Gralloc

TARGET_USES_GRALLOC1_ADAPTER := true

(G)SCALER

BOARD_USES_SCALER := true BOARD_USES_DT := true

HWCServices

BOARD_USES_HWC_SERVICES := true SKIP_DISPLAY_BLANK_CTRL := true`

rodman01 commented 5 years ago

also with the reverted commits SkipValidate I have the same problems with youtube, its almost freezing when changing to fullscreen and so to horizontal view and back to vertical. But I only did a dirty flash with the updated rom version, will try a clean install later.

rodman01 commented 5 years ago

New Test rom is uploaded: https://www.androidfilehost.com/?fid=6006931924117881894

Spookcity commented 5 years ago

@rodman01 Thanks for the new build. Will flash and test some. Good to know about LiveDisplay. Hopefully that means it will work on LOS as well with updated sources.

USE_OPENGL_RENDERER := true NUM_FRAMEBUFFER_SURFACE_BUFFERS := 3 We had these two flags originally,but I removed them as I thought they were not needed and/or deprecated. I havent noticed any difference without them. I am not real sure about the others honestly. I think I remember trying the "TARGET_USES_GRALLOC1_ADAPTER := true" one,but didn't notice any changes. I'm not positive though.

I've not installed GApps on this ROM,so i am not sure about YouTube videos. Videos work in the browser though. Can't say if it will cause soft reboots though. I will continue to test.

rodman01 commented 5 years ago

Yes can confirm youtube in browser is working much better as the youtube app.

rodman01 commented 5 years ago

my repos are updated, but I reverted all the entries made in boradconfigcommon. https://github.com/rodman01/android_device_samsung_smdk3470-common https://github.com/rodman01/android_device_samsung_kminilte

Spookcity commented 5 years ago

@rodman01 Great,thank you very much! I will build soon with some (all) of your commits and see how things work out. Then if all is good I will cherry-pick them. Especially the power profile update. Thanks for that. That is something that I have been meaning to do,but kept forgetting as I knew that this changed in Pie. Your RR build is running good for me. I will keep testing. Strange that videos with the YouTube app are kinda messed up. Not sure what is causing this.

MikeTeh commented 5 years ago

@rodman01 @Spookcity Testbuid2 seems to be a good build, smooth, no reboots so far. WLAN Hotspot Error, disables itself, error message under settings - hotspot and tethering :( I don't use localizations, so what's the shortest way to test gps? Localization and mobile data are disabled with installation. youtube vanced brakes the device to standstill, "system ui does not react", only (soft)reboot helps (works in crdroid 5.4 ie)

Hotspot: 05-27 03:05:07.687 V/WifiManager(2673): registerSoftApCallback: callback=com.android.systemui.statusbar.policy.HotspotControllerImpl@ca7c317, handler=Handler (android.os.Handler) {ba94304} 05-27 03:05:07.713 V/WifiManager(2673): SoftApCallbackProxy: onStateChanged: state=14, failureReason=0 05-27 03:05:07.714 V/WifiManager(2673): SoftApCallbackProxy: onNumClientsChanged: numClients=0 05-27 03:05:10.294 I/WifiService(2271): startSoftAp uid=1000 05-27 03:05:10.296 D/WifiStateMachinePrime(2271): Starting SoftApModeManager 05-27 03:05:10.303 I/android.hardware.wifi@1.0-service(2072): Wifi HAL started 05-27 03:05:10.308 I/WifiP2pNative(2271): Registering for interface available listener 05-27 03:05:10.307 I/WifiVendorHal(2271): Vendor Hal started successfully 05-27 03:05:10.323 D/WifiP2pNative(2271): P2P InterfaceAvailableListener true 05-27 03:05:10.323 D/WifiP2pService(2271): Wifi enabled=false, P2P Interface availability=true, Number of clients=0 05-27 03:05:10.325 D/WifiP2pService(2271): Wifi enabled=false, P2P Interface availability=true 05-27 03:05:10.340 I/e (2072): type=1400 audit(0.0:1671): avc: denied { write } for name="firmware_path" dev="sysfs" ino=1738 scontext=u:r:hal_wifi_default:s0 tcontext=u:object_r:sysfs:s0 tclass=file permissive=1 05-27 03:05:10.340 W/e (2072): type=1300 audit(0.0:1671): arch=40000028 syscall=322 per=800008 success=yes exit=6 a0=ffffff9c a1=b068f5cc a2=20001 a3=0 items=1 ppid=1 auid=4294967295 uid=1010 gid=1010 euid=1010 suid=1010 fsuid=1010 egid=1010 sgid=1010 fsgid=1010 tty=(none) ses=4294967295 exe="/system/vendor/bin/hw/android.hardware.wifi@1.0-service" subj=u:r:hal_wifi_default:s0 key=(null) 05-27 03:05:10.340 W/auditd (2038): type=1327 audit(0.0:1671): proctitle="/vendor/bin/hw/android.hardware.wifi@1.0-service" 05-27 03:05:10.345 I/e (2072): type=1400 audit(0.0:1672): avc: denied { search } for name="wifi" dev="mmcblk0p3" ino=21 scontext=u:r:hal_wifi_default:s0 tcontext=u:object_r:wifi_efs_file:s0 tclass=dir permissive=1 05-27 03:05:10.345 I/e (2072): type=1400 audit(0.0:1672): avc: denied { read } for name=".mac.info" dev="mmcblk0p3" ino=22 scontext=u:r:hal_wifi_default:s0 tcontext=u:object_r:wifi_data_file:s0 tclass=file permissive=1 05-27 03:05:10.345 I/e (2072): type=1400 audit(0.0:1672): avc: denied { open } for name=".mac.info" dev="mmcblk0p3" ino=22 scontext=u:r:hal_wifi_default:s0 tcontext=u:object_r:wifi_data_file:s0 tclass=file permissive=1 05-27 03:05:11.532 I/WifiHAL (2072): Initializing wifi 05-27 03:05:11.532 I/WifiHAL (2072): Creating socket 05-27 03:05:10.345 W/e (2072): type=1300 audit(0.0:1672): arch=40000028 syscall=54 per=800008 success=yes exit=0 a0=6 a1=8914 a2=be91d204 a3=be91d1ec items=5 ppid=1 auid=4294967295 uid=1010 gid=1010 euid=1010 suid=1010 fsuid=1010 egid=1010 sgid=1010 fsgid=1010 tty=(none) ses=4294967295 exe="/system/vendor/bin/hw/android.hardware.wifi@1.0-service" subj=u:r:hal_wifi_default:s0 key=(null) 05-27 03:05:11.547 I/WifiHAL (2072): Initialized Wifi HAL Successfully; vendor cmd = 103 05-27 03:05:11.548 I/android.hardware.wifi@1.0-service(2072): Adding interface handle for wlan0 05-27 03:05:11.548 I/android.hardware.wifi@1.0-service(2072): Adding interface handle for p2p0 05-27 03:05:11.548 E/android.hardware.wifi@1.0-service(2072): Failed to register radio mode change callback 05-27 03:05:11.548 I/android.hardware.wifi@1.0-service(2072): Configured chip in mode 1 05-27 03:05:11.570 D/WifiP2pNative(2271): P2P InterfaceAvailableListener false 05-27 03:05:11.570 D/WifiP2pService(2271): Wifi enabled=false, P2P Interface availability=false 05-27 03:05:11.582 D/WificondControl(2271): Setting up interface for soft ap mode 05-27 03:05:11.592 I/WifiNative(2271): Interface state changed on Iface:{Name=wlan0,Id=17,Type=AP}, isUp=true 05-27 03:05:11.592 I/WifiNative(2271): Successfully setup Iface:{Name=wlan0,Id=17,Type=AP} 05-27 03:05:11.593 V/WifiManager(2673): SoftApCallbackProxy: onStateChanged: state=12, failureReason=0 05-27 03:05:11.662 W/BroadcastQueue(2271): Background execution not allowed: receiving Intent { act=android.net.wifi.WIFI_AP_STATE_CHANGED flg=0x4000010 (has extras) } to com.google.android.gms/.chimera.GmsIntentOperationService$PersistentTrustedReceiver 05-27 03:05:11.667 D/WifiService(2271): handleWifiApStateChange: currentState=12 previousState=11 errorCode= -1 ifaceName=wlan0 mode=1 05-27 03:05:11.676 E/WifiVendorHal(2271): setCountryCodeHal(l.1831) failed {.code = ERROR_NOT_AVAILABLE, .description = } 05-27 03:05:11.681 W/wificond(2128): Copy constructor is only used for unit tests 05-27 03:05:11.682 W/wificond(2128): Failed to get NL80211_ATTR_EXT_FEATURES 05-27 03:05:11.820 I/android_os_HwBinder(2271): HwBinder: Starting thread pool for default::android.hardware.wifi.hostapd@1.0::IHostapd 05-27 03:05:11.991 E/WifiNative(2271): Failed to add acccess point 05-27 03:05:11.993 V/WifiManager(2673): SoftApCallbackProxy: onStateChanged: state=14, failureReason=0 05-27 03:05:11.994 E/WifiController(2271): SoftAP start failed 05-27 03:05:11.995 D/WifiService(2271): handleWifiApStateChange: currentState=14 previousState=12 errorCode= 0 ifaceName=wlan0 mode=1 05-27 03:05:11.995 D/WifiService(2271): updateInterfaceIpState: ifaceName=null mode=-1 previous mode= -1 05-27 03:05:11.995 I/WifiService(2271): stopSoftAp uid=1000 05-27 03:05:11.996 W/BroadcastQueue(2271): Background execution not allowed: receiving Intent { act=android.net.wifi.WIFI_AP_STATE_CHANGED flg=0x4000010 (has extras) } to com.google.android.gms/.chimera.GmsIntentOperationService$PersistentTrustedReceiver 05-27 03:05:12.210 D/WificondControl(2271): tearing down interfaces in wificond 05-27 03:05:12.216 E/WifiHAL (2072): wifi_event_loop: Read after POLL returned 4, error no = 0 (Success) 05-27 03:05:12.216 D/WifiHAL (2072): Got a signal to exit!!! 05-27 03:05:12.216 I/WifiHAL (2072): Exit wifi_event_loop 05-27 03:05:12.216 E/WifiHAL (2072): wifi_cleanup: Read after POLL returned 4, error no = 0 (Success) 05-27 03:05:12.216 E/WifiHAL (2072): Event processing terminated 05-27 03:05:12.216 I/android.hardware.wifi@1.0-service(2072): Legacy HAL stop complete callback received 05-27 03:05:12.217 I/WifiHAL (2072): Internal cleanup completed 05-27 03:05:12.419 I/android.hardware.wifi@1.0-service(2072): Wifi HAL stopped 05-27 03:05:12.420 I/WifiVendorHal(2271): Vendor Hal stopped 05-27 03:05:12.421 I/WifiNative(2271): Successfully torn down Iface:{Name=wlan0,Id=17,Type=AP} 05-27 03:05:12.423 I/WifiNative(2271): Successfully initiated teardown for iface=wlan0 05-27 03:05:12.424 V/BatteryExternalStatsWorker(2271): no controller energy info supplied for wifi 05-27 03:05:13.171 V/WifiManager(2673): registerSoftApCallback: callback=com.android.systemui.statusbar.policy.HotspotControllerImpl@ca7c317, handler=Handler (android.os.Handler) {ba94304} 05-27 03:05:13.173 V/WifiManager(2673): SoftApCallbackProxy: onStateChanged: state=14, failureReason=0 05-27 03:05:13.174 V/WifiManager(2673): SoftApCallbackProxy: onNumClientsChanged: numClients=0

rodman01 commented 5 years ago

sorry I wasn't online yesterday anymore... @Spookcity yes the youtube issue is strange, because after I reverted everything, I haven't had this problem. @MikeTeh great many thanks for the log. Yes I also noticed some problems with wifi disconnections. But I am not using the hotspot. So the problem might be related to the changes in overlay: `

true

<!-- Enable 802.11ac for Wifi hotspot (SAP) -->
<bool translatable="false" name="config_wifi_softap_ieee80211ac_supported">true</bool>`

I will reverse them in next build and see whether it is fixed.

Spookcity commented 5 years ago

Yeah @rodman01 strange issue with YouTube indeed. Perhaps it was related to switching graphic composer to hwbinder. I can't really think of what else it would be. I've had 2 reboots since using your RR build sadly. So it seems as though this issue is still there. I wasn't able to grab a log. Hopefully at some point I will be able to. It seems,at least for me,that HWC works about 95% of the time without problems. I am just not sure what the remaining issue might be.

A bit off topic,but LiveDisplay just refuses to work in LOS. I have no idea why. It also now works in AEX. I assume they have just added this feature to their frameworks/base..Which is prob what RR did as well,so I guess thats why it works there. I just get a black screen on boot anytime I build any version of LiveDisplay for Lineage. kinda confusing and frustrating.

rodman01 commented 5 years ago

ggr I thought too, that we are on a good way. I haven't had any reboots so far anymore, but do not use the mobile as daily driver, only when I flash new things and for testing. Anyways. You are right, after reverting the composer back to passthrough the youtube app returns to normal :-) Switching from portrait to landscape is working again and is quite fast again. After I changed other things by try and error previously, I got the change too, but very slow, but the app was not crashing anymore. Not even freezing, only slow movements. Also systemui didn't stopped anymore. What I also did is set treble omx in system.prop to false which I found for example here. What I also did is this kernel commit, but not sure whether this helped anything, but it works :-): Kbuild: Add missing videodev2 exynos headers. And at least I copied the open max video board configurations from gts2-common(https://github.com/Exynos5433/android_device_samsung_gts2-common/blob/lineage-16.0/BoardConfigCommon.mk). Think that this alone is also not sufficient or maybe even wrong. I will upload a new RR Testbuild in a few, if somebody is interested in a test

About livedisplay: usually RR doesn't add not RR specific comits to their synced or forked leo repos. Or better thats they used to do it in the past :-). And I haven't found any related commits at their fwb or somewhere, which is not also included in leos repos. Haggerty just merged liveisplay related commits in the msm8976 common device repo. Maybe this helps and points to the right direction? Although I haven't included something of that in my device sources, as far as I remember :-)

rodman01 commented 5 years ago

ok new RR test build is online. I also reverted the overlays relatet to wifi and softap. So if you can test also the hotspot again, it would be great. Test3-RR-P-v7.0.2-20190529-kminilte-Unofficial.zip: https://www.androidfilehost.com/?fid=6006931924117884250

MikeTeh commented 5 years ago

Thanks, I'll try RR3, I had today the first two reboots in RR2 (no logs, connected with car over Bluetooth this time). No other problems so far..

Spookcity commented 5 years ago

@rodman01 Thanks a lot for the info and again also for working on this. Yeah,the HWC issue is weird as it seems to work without issues most of the time,but then a few times causes soft reboots,so I'm not sure. I believe that the treble omx prop is deprecated as we had this at one point and I removed it. Not sure about the open max video ones,but I can't imagine they are going to hurt. Same with the videodev2 exynos headers kernel commit. Not real sure about this,but hey if it builds it probably doeasnt hurt anything. I tried to update our CameraWrapper.cpp with This commit but sadly it didnt change anything,so I'm really at a loss at to what exactly is wrong with our camera.

Thanks for the info on LiveDisplay. Haggerty's commit is pretty similar to one of the things that I have tried (except using the exynos version) I also tried building the LiveDisplay sysfs. It's weird because if I don't add LiveDisplay to out boardconfig and the HAL to our manifest it doesn't work and when I do I get a black screen. Thats why i assumed that RR added this to their fwb. Either way,its a minor thing. I would just like to get it working. I will contimue to mess with it when I have time.

I'll flash your new RR build later and report back my findings. Thanks!

rodman01 commented 5 years ago

Maybe later today or tomorrow I can try to build a leos version with your sources too, just to verify and check how it is working on my device. Perhaps so we could get a hint about the not working livedisplay.

In the meantime I tried several other thing and at the moment with some commits included in openmax/osal RR is running good and youtube is almost smooth and fast :-) I took this: https://github.com/LineageOS/android_hardware_samsung_slsi-cm_openmax/commit/faf9a76b1e26a78d65d5b61ff5137aae63ad8a20 and some of this: https://github.com/Exynos5433/android_hardware_samsung_slsi-cm_exynos5433/commits/lineage-16.0. Will add the repo to my github and push the commits I cherry picked.

Currently and still no reboots :-). But I have 2 Warnings in logcat for which I cannot find something with google W MapperHal: buffer descriptor with invalid usage bits 0x2000 (in youtube "section" W MapperHal: buffer descriptor with invalid usage bits 0x400 (paired with) E ion : ioctl c0044901 failed with code -1: Invalid argument

Camera I haven't tested yet.

Spookcity commented 5 years ago

@rodman01 Great,thanks! I think I remember seeing these errors in our logs for awhile now,so they may be (mostly) harmless hopefully. Interesting commits. Hopefully they will indeed help. Thanks a lot!

rodman01 commented 5 years ago

github updated: https://github.com/rodman01/android_hardware_samsung_slsi_exynos3470 mmh I notice no significant change :-) so this is a good new, but also a bad :-)

But I haven't done a clean flash for a while, maybe its time for it now :-)

Edit: camera is crashing gggr, will do a reset and new installation now.

rodman01 commented 5 years ago

Gave up the hwc thingis and camera problems for now :-), I am lost at the moment... But synced the leos sources already as said and its building now. I noticed when I tried to merge the latest hardware/samsung leo commits to your repo, many conflicts. I set everything to theirs, and now its still building, but maybe that there is the reason why livedisplay is not working for you? Have you also merged the leo commits from 24. and 25.5?

Spookcity commented 5 years ago

Yeah,the HWC issue is confusing. I think I will just go ahead and disable HWC again in my next LOS build and revert the related commits. I honestly haven't really noticed any downside to this. Battery seems to be OK and the system seems pretty fluid with it disabled.

As far as my hardware/samsung repo: Yeah, I have merged the latest LOS hardware/samsung into my repo. I just keep forgetting to push. I did have merge conflicts I think because i had repo picked some LiveDisplay stuff that eventually became part of the repo,but in a different form. I solved these conflicts I think properly,but it is def possible that I messed something up along the way. I will wait to see what your results of are with LiveDisplay before I push because if it works for you then I def messed something up. You are going to have some RIL problems since you are using LOS's hardware/samsung repo as there are patches applied to mine,but this will obviously not effect LiveDisplay. Thanks again for helping out!

rodman01 commented 5 years ago

okidoki build finished fine and in my opinion live display is working. It is different to RR, if I remember right, there is one colour setting missing, but not sure. What is not working or greyed out is the colour settings the others are there but the effects are lower then on RR. But that is just a personal feeling. Anyways here is the link, if you and all others what to test my leos build. I also used your hardware/samsung, but merged the leos 16 branche and solved the conflicts choosing: set the conflicted files to theirs (leos 16). If you want I can also put it on my github

lineage-16.0-20190530-UNOFFICIAL-kminilte.zip: https://www.androidfilehost.com/?fid=6006931924117884987

Spookcity commented 5 years ago

@rodman01 Ya know what? I decided to start a clean build when I got home as its been a bit since Ive done so....and LiveDisplay is indeed working for me now too! I have no idea why. I did add the commit you made to our overlay "config_setColorTransformAccelerated" I am not sure what that does honestly,but could this be it? Did you use this in your build as well or did you use my repo? Perhaps I will try to add livedisplay stuff to our manifest and BoardConfig and see what happens,but either way,its working now,so great! Thanks again for all of your work! I will push my hardware/samsung repo and I cherry-picked about 20 kernel commits relating mostly to bluetooth and wireless. I will push this as well. Thanks again!!

Edit: All of my repos have been updated and pushed

MikeTeh commented 5 years ago

RR test3: Hotspot is working again, phone app is loosing favourites sometime. I'd changed live display over the qs-tile from auto to night, off and on again and I had a reboot

05-31 18:23:03.648 I/LineageSystemServer(9712): Attempting to start service org.lineageos.platform.internal.display.LiveDisplayService 05-31 18:23:03.662 I/LineageSystemServer(9712): Starting service org.lineageos.platform.internal.display.LiveDisplayService 05-31 18:23:03.662 I/SystemServiceManager(9712): Starting org.lineageos.platform.internal.display.LiveDisplayService 05-31 18:23:21.544 W/hwservicemanager(1967): getTransport: Cannot find entry vendor.lineage.livedisplay@2.0::IDisplayModes/default in either framework or device manifest. 05-31 18:23:21.552 W/hwservicemanager(1967): getTransport: Cannot find entry vendor.lineage.livedisplay@2.0::IReadingEnhancement/default in either framework or device manifest. 05-31 18:23:21.572 W/hwservicemanager(1967): getTransport: Cannot find entry vendor.lineage.livedisplay@2.0::IColorBalance/default in either framework or device manifest. 05-31 18:23:21.578 W/hwservicemanager(1967): getTransport: Cannot find entry vendor.lineage.livedisplay@2.0::ISunlightEnhancement/default in either framework or device manifest. 05-31 18:23:21.591 W/hwservicemanager(1967): getTransport: Cannot find entry vendor.lineage.livedisplay@2.0::IPictureAdjustment/default in either framework or device manifest. 05-31 18:23:21.610 E/LiveDisplay(9712): Color temperature out of range: 0 05-31 18:23:21.620 W/SystemServiceManager(9712): Service org.lineageos.platform.internal.display.LiveDisplayService took 112 ms in onBootPhase

rodman01 commented 5 years ago

hello and good morning! @MikeTeh many thanks. I have added livedisplay-sdm and sysfs as well as advanced display to the device and have no reboot after doing the tests you mentioned above. I cannot get rid of the warnings in logcat, but I hope the reboots are gone for you too anymays. New RR testbuild is up: Test4-RR-P-v7.0.2-20190601-kminilte-Unofficial.zip @Spookcity I have added the missing entries of livedisplay to hidl manifest, but depending on which one I all included, I get difficulties from black screen to not booting over endless phone is updating message. Do you know why? and/or what is to do to solve the warnings? Do we need to change to passthrough? I also notices some of this warnings for example related to audio.

My logcat seems to look not to bad now, but I have crash regarding to fingerprint validity service. I changed biometrics fingerprint to 2.1 but no change. Is fingerprint working in general with P, and with your leos P? Or is it currently broken in general. If you want to check my RR log, please find it attached. Clean install of RR without wifi and without gapps. logcat.txt

Spookcity commented 5 years ago

@rodman01 I have no idea about the LiveDisplay issue. This is the problem I was experiencing. I didn't grab any recent logs from the blackscreen caused by livedisplay,but older ones showed missing libs. No idea what that was about,but it was a bunch of them. For whatever reason it just doesn't seem to work if we include the entries in the manifest. Not sure why at this point. I personally don't use the FP scanner,but I think that in general it works. The fatal errors are always there upon boot,but it seems to connect after a bit. I think (if i remember correctly) that the FP works decently at least,but there are issues with initially registering the fingerprint. This is something I really havent looked into at all. I do know that This was something I wanted to look into,but just havent had the time recently.

rodman01 commented 5 years ago

okidoki and thanks....i will have a look at it and try to cherry pick / "port" the commit. Will see how it works :-)

rodman01 commented 5 years ago

stupid question, because I cannot find the settings: where to enable fingerprint unlock? I recently saw it in setup wizard and skipped it all the time, but in RR and without gapps I do not have a setup wizard. maybe I cannot find it because its not working and not properly included? But with the klte changes the fatal error in logcat disappeared, but others related to that appeared :-). logcat.txt

edit: never mind I found it after flashing leos 16 spookies original and also in my leos build it is in settings available.

Spookcity commented 5 years ago

It should be under security and location if it built properly. Otherwise I'd guess something didn't work. I'll have to have a look at your logcat when I'm at my laptop as in the fasthub app I can't for whatever reason.

Edit: You beat me to it :smile:

aelfinn commented 5 years ago

@Spookcity The overclocked kernel works well, btw. Would it be hard to enable the higher and lower frequencies in the default kernel of one of your ROMs, or does one have to recompile the kernel? Also, would it be possible to include a version of IntelliPlug?

Spookcity commented 5 years ago

@aelfinn Glad to hear its working out! I guess I could try to use anykernel2 so it would just work with any Pie ROM. I'm just not sure of any changes to anykernel and how easy/difficult it would be. I will look into it when I have some time. As far as adding IntelliPlug... It really should just be a matter of cherry-picking the commit since I already added it to my N_custom kernel. Again,I will look into this when I have some free time.

MikeTeh commented 5 years ago

I have tested RR Test4 for a few days now in everyday use and it works fine for me, no reboots, hotspot works again. Is it planned to put it in xda? Thanks for your work on it @rodman01

rodman01 commented 5 years ago

Thanks for your feedback. I am currently in holiday, but after my return middle of next week, I will update my sources and try another and new build. Maybe I will put a release note in the current RR thread from @friedensfurz if he don‘t mind, but at the moment I am not planning to open a new thread of and for my own.

Spookcity commented 5 years ago

AospExtended update

This build seems to be running smoothly unlike the previous one. I've been using it for a day plus. Also fixed the messaging app not being built (stupid mistake on my part last time) Updated AEX sources. June security patch and same changes from the LineageOS 16 build.

Download Standard Build - AEX-6.6 MicroG Build - AEX-6.6

Enjoy!

aelfinn commented 5 years ago

@Spookcity I’d be happy to test a kernel! :)

Oleggluk commented 5 years ago

@Spookcity Hi! Glad to see you are fine. I bought an RN7. I'm happy, 10 hours of screen work in 4G+. Thank you for everything that you did for Samsung, I will never come back))))). Join us). (Google translator)

rodman01 commented 5 years ago

so finally is done :-) New RR build is up dan posted on xda. Here it is for your interest too. Download RR-P-v7.0.2-20190624-kminilte-Unofficial.zip: https://www.androidfilehost.com/?fid=6006931924117902206 https://forum.xda-developers.com/showpost.php?p=79788255&postcount=85 @Spookcity I have added 2 wifi commits to the kernel, which I found in leos review and are related to error lines in logcat of my previous build. Also in device repos I did a few changes wifi and media codecs related. All are or is uploaded to my github. Till now it seems to be stable and after my simple tests (no whats up, no instagram etc etc) only youtube (browser and app) and google play film rom seems to be stable. What I still haven't got solved is the crash in logcat related to validity service, so I reverted all my tries and it is working with your code anyways :-). And the slow or very long lasting of gps fix in gps status. Maybe some other can test and verify whether gps is working, not working or working good.

edit: download link fixed

Spookcity commented 5 years ago

@Oleggluk Great to see you around here!! I hope all is well with you. Sounds like it definitely is when it comes to your phone. Ya know,I have been due a new device for well over a year now. It would actually save me a little money on my contract to get a new one,but I am so "out of the loop" on what phones are good and what-not at this point. I am ashamed to admit i needed to look up what the RN&7 was :) I will certainly keep that in mind whenever I finally do decide to move on from my mini.

@rodman01 Thanks for the RR build and also the info on your commits. I will look into adding some of these to my main repo as well. Especially the WiFi ones. I don't see how they would be anything but helpful,so thanks!

@aelfinn OK,I have built a P_custom kernel and used anykernel2. It should flash on any P ROM. I have basically just added things (optimization flags,CPU governors,IO schedulers etc) that were not a huge pain in the a$$ to bring over from my N_custom branch. When I have time I will try to get more added. I am not going to give a ton of detail about what the kernel includes,so just have a look Here and you'll get a good idea. I used a linaro 7.2 toolchain to build this. I have tested for a few days and it seems stable fro me. Although Ive only really used the zzmoove governor and messed with voltage some. Anyway...Hope it works out and there is some noticed benefits.

Download SpookcityKernel-P

aelfinn commented 5 years ago

@Spookcity Will try a.s.a.p.!

Incidentally, the one device in my circle of friends that is running your LOS16 build (2019-06-09) has sporadic reboots, which seem to be associated to incoming/outgoing calls in every instance. Any idea what might cause that or what to try?

rodman01 commented 5 years ago

@Spookcity thanks for your new kernel and already downloaded :-) Working well also on RR build. According to my kernel commits: I noticed a more stable connection with the changes, but I still have disconnections especially during the restore of apps and settings from google play. So from my point of view it is still not perfect.

brAvEiitoOr commented 5 years ago

Thanks @Spookcity for the kernel. Unfortunately, the phone can´t boot. Could you try to make an 1.5Ghz version, like in Nougat? Or how to solve it?. Thanks in advance

Spookcity commented 5 years ago

@aelfinn Thats a weird issue your friend is experiencing with the ROM. I honestly have no idea what the cause may be. I haven't had any reboots while the HWC has been disabled. Perhaps they can get a log to see exactly what is happening.

@rodman01 Yeah,it probably will be a series of commits that hopefully will eventually solve the WiFi issue,so these commits are def a step in the right direction. I haven't had any time lately to really look into anything,but hopefully i will have a bit more in the coming weeks.

@brAvEiitoOr So weird how the 1.6 overclock works for some but not all. I guess this is just something I will never know the reason for. I mean we are all using the same device with the same CPU. Very weird. Either way, i will do a 1.5 version whenever I have some time. It is not very difficult or time consuming,but right now I just really don't have any time. Hopefully within a week or 2 though.

aelfinn commented 5 years ago

@Spookcity She depends on the phone as her main device, so I can’t take it away from her for extended testing. Is there maybe an app that can help in such circumstances? (By constantly taking a log, of fixed length, until some trigger event such as a reboot?)

Spookcity commented 5 years ago

@aelfinn Ahh,yes I can understand this. I always used Matlog in the past. Although I haven't in some time as I usually just connect to adb shell now, but this isn't super user friendly. There is probably an option in Matlog for something that would be helpful though I would think. Root is required. Unless someone else has a better idea that's all I can really think of off hand.

Spookcity / ROMS-G800F

Test Builds #9 #25

Gralloc

(G)SCALER

HWCServices