Alldocube X Neo

According to my last today finding (but not confirmed yet) the tablet is not much sensitive to 5GHz signal.
I did some more testing and have to say that the tablet have either a very bad 5GHz antenna or maybe it misses it at all. The signal reception sensitivity is really very bad. I believe it cannot be problem of the SD660 chipset, but the signal reception stuff.
 
Last edited:
In addition to that it may be a software problem too. There are two different states of working 5 GHz on my device - if I boot up with airplane mode and have 5 GHz, the tablet sees only 1 of 3 access points with 2.4 GHz in my house. In the rare case of having 5 GHz with normal boot up it sees all 3 access points with 2.4 GHz. And this is reproduceable.
 
So I flashed the LOS 17.1 to my slot B (I keep QiDroid in slot A for now) and briefly checked the 5GHz WiFi. It behaves the same as on QiDroid, so no good news here.

Then I tried to flash the LeOS, but was successful only with the LeOS-Q variant, which is basically the LOS 17.1 on Android 10. I was not successful with all the LeOS-R-arm64_bv* variants, including the Mini. It just doesn't boot. Or maybe does, but... I would say the tablet has powered off during the boot (shortly after the boot logo disappears) in these cases. But I noticed one detail I didn't noticed earlier in many same cases so far. When the tablet "powers off" during the boot, the USB plugged in sound plays on the connected computer. And the running QFIL shows the opened port. Then, when I want to power on the tablet again (because I believe it is turned off), I have to press the power button really long. So I'm not just powering on, I'm forcing it to reboot in fact. It means the tablet does not shuts down itself, as it looks like. On the contrary it is running and maybe it has already booted up and just shows the USB debugging confirmation screen. But nothing is being visible, the screen is dark as it would be when powered off. Short pressing the power button does nothing. So maybe just little is missing to working state. But don't have any clue what...
 
Wow, cannot believe that! The LOS 16.0 (lineage-16.0-20200516-UNOFFICIAL-t1009) works really great in terms of WiFi 5GHz performance! I get approx. 300Mbit/s download as well as upload speeds in Ubiquti WiFiman speed test and some 220Mbit/s download speed in real life download test from FTP server. The signal performance is also great and is comparable to the older Samsung phone. It keeps the connection in places where it dropped it in all my other tests so far. Yeah taro4, you were right.;) The only drawback is that it still doesn't support the channel 149 I have been using on my home router till now. But it seems to be a device limitation I have to live with. Despite this one I'm really impressed.

Could someone (probably CrimsonKnight13) tell where the lineage-16.0-20200516-UNOFFICIAL-t1009 comes from and if there is any way how to upgrade it to Android 10 (LOS 17.1 or any other usable distribution)? I think we should start right from here if we want to achieve a good WiFi performance (maybe HW compatibility in general).

Edit:
The above do apply for running LOS 16.0 with its default recovery. After flashing TWRP, the WiFi cannot be turned on again. Installing Magisk doesn't help.:(
 
Last edited:
You can't upgrade from the unofficial LOS16 to LOS17.1 or any other GSI ROM. It's a custom ROM I found from the person that created TWRP for the X Neo. They have not updated anything regarding LOS16 since the spring of this year.
 
You can't upgrade from the unofficial LOS16 to LOS17.1 or any other GSI ROM. It's a custom ROM I found from the person that created TWRP for the X Neo. They have not updated anything regarding LOS16 since the spring of this year.
It's a pity. But thank you for the info anyway. I thought to keep LOS 16.0. But had issues with bluetooth. Was not able to connect speaker nor headphone. Showstopper for me. So decided to flash LOS 17.1 GSI system image over LOS 16.0. It boots, it works, but 5GHz WiFi is again back where it was with any other ROM tested besides LOS 16.0. Still wondering why it works so great only on the custom LOS 16.0. And thinking how to fool it to make it work so fine on Android 10 GSI too. But already out of ideas...
 
Maybe found the cause of the 5GHz issue. See the logcat outputs the from booted custom LOS 16.0 and GSI LOS 17.1 below. Just to remind you, in custom LOS 16.0 the WiFi 5GHz works nice, but in LOS 17.1 GIS (and other GSI ROMs) has issues (slow transfer rates, weak signal, has to be enabled by booting in airplane mode).

LOS 16.0
Code:
qucii_qa6210_64:/ $ logcat | grep wifi
01-03 11:03:59.482   813   813 I [email protected]: Wifi Hal is booting up...
12-02 22:48:39.175  1168  1168 I wificond: wificond is starting up...
12-02 22:48:41.257   984   984 I SystemServiceManager: Starting com.android.server.wifi.WifiService
12-02 22:48:41.384   984   984 I SystemServiceManager: Starting com.android.server.wifi.scanner.WifiScanningService
12-02 22:48:41.387   984   984 I SystemServiceManager: Starting com.android.server.wifi.p2p.WifiP2pService
12-02 22:48:41.410   984   984 D ConnectivityService: wifiOnly=false
12-02 22:48:42.322   984  1397 E BatteryExternalStatsWorker: no controller energy info supplied for wifi
12-02 22:48:43.685   813   813 I [email protected]: Wifi driver is ready: Success
12-02 22:48:43.687   813   813 I [email protected]: Wifi HAL started
12-02 22:48:43.766   984  1617 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.829   813   813 I WifiHAL : Initializing wifi
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: initialized exit socket pair
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: host_logs mcast_id: 10
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: fw_logs mcast_id: 11
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: per_pkt_stats mcast_id: 12
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: diag_events mcast_id: 13
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: fatal_events mcast_id: 14
12-02 22:48:43.841   813   813 I [email protected]: Adding interface handle for wlan0
12-02 22:48:43.841   813   813 I [email protected]: Adding interface handle for p2p0
12-02 22:48:43.841   813   813 I [email protected]: Configured chip in mode 2
12-02 22:48:43.894  1168  1168 I wificond: create scanner for interface with index: 26
12-02 22:48:43.894  1168  1168 I wificond: subscribe scan result for interface with index: 26
12-02 22:48:43.895   593   593 W hwservicemanager: getTransport: Cannot find entry [email protected]::IOffload/default in either framework or device manifest.
12-02 22:48:43.895  1168  1168 E wificond: No Offload Service available
12-02 22:48:43.895  1168  1168 I wificond: Offload HAL not supported
12-02 22:48:43.901   984  1648 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.906   984  1648 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Override interface parameter: ctrl_interface ('(null)' -> '/data/vendor/wifi/wpa/sockets')
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Initializing interface 'wlan0' conf '/data/vendor/wifi/wpa/wpa_supplicant.conf' driver 'nl80211' ctrl_interface '/data/vendor/wifi/wpa/sockets' bridge 'N/A'
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Configuration file '/data/vendor/wifi/wpa/wpa_supplicant.conf' -> '/data/vendor/wifi/wpa/wpa_supplicant.conf'
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Reading configuration file '/data/vendor/wifi/wpa/wpa_supplicant.conf'
12-02 22:48:43.911  2278  2278 D wpa_supplicant: Reading configuration file '/vendor/etc/wifi/wpa_supplicant_overlay.conf'
12-02 22:48:43.914  2278  2278 I chatty  : uid=1010(wifi) /vendor/bin/hw/wpa_supplicant identical 2 lines
12-02 22:48:43.932   984  1648 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.945  2278  2278 I chatty  : uid=1010(wifi) /vendor/bin/hw/wpa_supplicant identical 2 lines
12-02 22:48:43.949   984  1688 I WifiScanningService: wifi driver unloaded
12-02 22:48:43.964   984  1688 I WifiScanningService: wifi driver loaded with scan capabilities: max buckets=6
12-02 22:48:48.557  1168  1168 I wificond: Regulatory domain changed to country:

LOS 17.1 GSI
Code:
qucii_qa6210_64:/ # logcat | grep wifi
12-02 21:19:48.691  1238  1738 I PackageParser: Skipping target and overlay pair android and null: overlay ignored due to required system property: persist.sys.overlay.wifi5g with value: true
12-02 21:19:48.691  1238  1238 W PackageManager: Failed to parse /product/overlay/treble-overlay-wifi5g.apk: /product/overlay/treble-overlay-wifi5g.apk (at Binary XML file line #5): null
12-02 21:19:52.495  1238  1238 I SystemServiceManager: Starting com.android.server.wifi.WifiService
12-02 21:19:52.599  1238  1238 I SystemServiceManager: Starting com.android.server.wifi.scanner.WifiScanningService
12-02 21:19:52.601  1238  1238 I SystemServiceManager: Starting com.android.server.wifi.p2p.WifiP2pService
12-02 21:19:52.612  1238  1238 D ConnectivityService: wifiOnly=false
12-02 21:19:53.513  1093  1093 I [email protected]: Wifi driver is ready: Success
12-02 21:19:53.516  1093  1093 I [email protected]: Wifi HAL started
12-02 21:19:53.544  1093  1093 I [email protected]: Adding interface handle for wlan0
12-02 21:19:53.544  1093  1093 I [email protected]: Adding interface handle for p2p0
12-02 21:19:53.544  1093  1093 I [email protected]: Configured chip in mode 2
12-02 21:19:53.556  1438  1438 I wificond: create scanner for interface with index: 26
12-02 21:19:53.557  1438  1438 I wificond: subscribe scan result for interface with index: 26
12-02 21:19:53.561  1438  1438 E wificond: No Offload Service available
12-02 21:19:53.561  1438  1438 I wificond: Offload HAL not supported
12-02 21:19:54.006  1238  1238 D SystemServiceManager: onStartUser-0 com.android.server.wifi.WifiService took to complete: 0ms
12-02 21:19:54.006  1238  1238 D SystemServiceManager: onStartUser-0 com.android.server.wifi.scanner.WifiScanningService took to complete: 0ms
12-02 21:19:54.006  1238  1238 D SystemServiceManager: onStartUser-0 com.android.server.wifi.p2p.WifiP2pService took to complete: 0ms
12-02 21:19:56.316  1438  1438 W wificond: Scan is not started. Ignore abort request
12-02 21:19:56.358  1438  1438 I wificond: Regulatory domain changed to country:

Those first two lines in LOS 17.1 GSI output look suspicious to me. Any idea what it might be and if really could it be the cause? Only found some little info on issues with overlays in treble ROMs, but no solution applicable to this one.
 
  • Like
Reactions: taro4
like this
Maybe found the cause of the 5GHz issue. See the logcat outputs the from booted custom LOS 16.0 and GSI LOS 17.1 below. Just to remind you, in custom LOS 16.0 the WiFi 5GHz works nice, but in LOS 17.1 GIS (and other GSI ROMs) has issues (slow transfer rates, weak signal, has to be enabled by booting in airplane mode).

LOS 16.0
Code:
qucii_qa6210_64:/ $ logcat | grep wifi
01-03 11:03:59.482   813   813 I [email protected]: Wifi Hal is booting up...
12-02 22:48:39.175  1168  1168 I wificond: wificond is starting up...
12-02 22:48:41.257   984   984 I SystemServiceManager: Starting com.android.server.wifi.WifiService
12-02 22:48:41.384   984   984 I SystemServiceManager: Starting com.android.server.wifi.scanner.WifiScanningService
12-02 22:48:41.387   984   984 I SystemServiceManager: Starting com.android.server.wifi.p2p.WifiP2pService
12-02 22:48:41.410   984   984 D ConnectivityService: wifiOnly=false
12-02 22:48:42.322   984  1397 E BatteryExternalStatsWorker: no controller energy info supplied for wifi
12-02 22:48:43.685   813   813 I [email protected]: Wifi driver is ready: Success
12-02 22:48:43.687   813   813 I [email protected]: Wifi HAL started
12-02 22:48:43.766   984  1617 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.829   813   813 I WifiHAL : Initializing wifi
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: initialized exit socket pair
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: host_logs mcast_id: 10
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: fw_logs mcast_id: 11
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: per_pkt_stats mcast_id: 12
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: diag_events mcast_id: 13
12-02 22:48:43.835   813   813 I CLD80211: /vendor/bin/hw/[email protected]: nlctrl family id: 16 group: fatal_events mcast_id: 14
12-02 22:48:43.841   813   813 I [email protected]: Adding interface handle for wlan0
12-02 22:48:43.841   813   813 I [email protected]: Adding interface handle for p2p0
12-02 22:48:43.841   813   813 I [email protected]: Configured chip in mode 2
12-02 22:48:43.894  1168  1168 I wificond: create scanner for interface with index: 26
12-02 22:48:43.894  1168  1168 I wificond: subscribe scan result for interface with index: 26
12-02 22:48:43.895   593   593 W hwservicemanager: getTransport: Cannot find entry [email protected]::IOffload/default in either framework or device manifest.
12-02 22:48:43.895  1168  1168 E wificond: No Offload Service available
12-02 22:48:43.895  1168  1168 I wificond: Offload HAL not supported
12-02 22:48:43.901   984  1648 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.906   984  1648 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Override interface parameter: ctrl_interface ('(null)' -> '/data/vendor/wifi/wpa/sockets')
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Initializing interface 'wlan0' conf '/data/vendor/wifi/wpa/wpa_supplicant.conf' driver 'nl80211' ctrl_interface '/data/vendor/wifi/wpa/sockets' bridge 'N/A'
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Configuration file '/data/vendor/wifi/wpa/wpa_supplicant.conf' -> '/data/vendor/wifi/wpa/wpa_supplicant.conf'
12-02 22:48:43.907  2278  2278 D wpa_supplicant: Reading configuration file '/data/vendor/wifi/wpa/wpa_supplicant.conf'
12-02 22:48:43.911  2278  2278 D wpa_supplicant: Reading configuration file '/vendor/etc/wifi/wpa_supplicant_overlay.conf'
12-02 22:48:43.914  2278  2278 I chatty  : uid=1010(wifi) /vendor/bin/hw/wpa_supplicant identical 2 lines
12-02 22:48:43.932   984  1648 I android_os_HwBinder: HwBinder: Starting thread pool for default::[email protected]::ISupplicant
12-02 22:48:43.945  2278  2278 I chatty  : uid=1010(wifi) /vendor/bin/hw/wpa_supplicant identical 2 lines
12-02 22:48:43.949   984  1688 I WifiScanningService: wifi driver unloaded
12-02 22:48:43.964   984  1688 I WifiScanningService: wifi driver loaded with scan capabilities: max buckets=6
12-02 22:48:48.557  1168  1168 I wificond: Regulatory domain changed to country:

LOS 17.1 GSI
Code:
qucii_qa6210_64:/ # logcat | grep wifi
12-02 21:19:48.691  1238  1738 I PackageParser: Skipping target and overlay pair android and null: overlay ignored due to required system property: persist.sys.overlay.wifi5g with value: true
12-02 21:19:48.691  1238  1238 W PackageManager: Failed to parse /product/overlay/treble-overlay-wifi5g.apk: /product/overlay/treble-overlay-wifi5g.apk (at Binary XML file line #5): null
12-02 21:19:52.495  1238  1238 I SystemServiceManager: Starting com.android.server.wifi.WifiService
12-02 21:19:52.599  1238  1238 I SystemServiceManager: Starting com.android.server.wifi.scanner.WifiScanningService
12-02 21:19:52.601  1238  1238 I SystemServiceManager: Starting com.android.server.wifi.p2p.WifiP2pService
12-02 21:19:52.612  1238  1238 D ConnectivityService: wifiOnly=false
12-02 21:19:53.513  1093  1093 I [email protected]: Wifi driver is ready: Success
12-02 21:19:53.516  1093  1093 I [email protected]: Wifi HAL started
12-02 21:19:53.544  1093  1093 I [email protected]: Adding interface handle for wlan0
12-02 21:19:53.544  1093  1093 I [email protected]: Adding interface handle for p2p0
12-02 21:19:53.544  1093  1093 I [email protected]: Configured chip in mode 2
12-02 21:19:53.556  1438  1438 I wificond: create scanner for interface with index: 26
12-02 21:19:53.557  1438  1438 I wificond: subscribe scan result for interface with index: 26
12-02 21:19:53.561  1438  1438 E wificond: No Offload Service available
12-02 21:19:53.561  1438  1438 I wificond: Offload HAL not supported
12-02 21:19:54.006  1238  1238 D SystemServiceManager: onStartUser-0 com.android.server.wifi.WifiService took to complete: 0ms
12-02 21:19:54.006  1238  1238 D SystemServiceManager: onStartUser-0 com.android.server.wifi.scanner.WifiScanningService took to complete: 0ms
12-02 21:19:54.006  1238  1238 D SystemServiceManager: onStartUser-0 com.android.server.wifi.p2p.WifiP2pService took to complete: 0ms
12-02 21:19:56.316  1438  1438 W wificond: Scan is not started. Ignore abort request
12-02 21:19:56.358  1438  1438 I wificond: Regulatory domain changed to country:

Those first two lines in LOS 17.1 GSI output look suspicious to me. Any idea what it might be and if really could it be the cause? Only found some little info on issues with overlays in treble ROMs, but no solution applicable to this one.
Might need to test with Magisk modules that deal with overlays or any root apps that do so.
 
12-02 21:19:48.691 1238 1738 I PackageParser: Skipping target and overlay pair android and null: overlay ignored due to required system property: persist.sys.overlay.wifi5g with value: true
I think I have another 5 GHz solution. I searched for the text persist.sys.overlay.wifi5g in the root-folder and below. The text is included in the file /bin/rw-system.sh:

if getprop ro.wlan.mtk.wifi.5g | grep -q 1; then
setprop persist.sys.overlay.wifi5g true
fi

As we don't have a MTK-device, ro.wlan.mtk.wifi.5g is not set to 1 and therefore persist.sys.overlay.wifi5g is not set to true.

So I added the line "ro.wlan.mtk.wifi.5g=1" to the build.prop, saved it, rebootet and now have a normal working 5 GHz wifi. :) The connection at the same place like written some posts before (3 m away from router) is around 800 Mbit/sec. with -55 dBm, channel 100, which are better values compared to the former "magic" 5 GHz mode. All my access points are visible at the moment.

Update: F*ck, after some more reboots 5 GHz is absent again. But the error message like above does not appear in the log. For testing I enabled airplane mode, reenabled wifi and have the (bad) 5 GHz back again. So weird!
 
Last edited:
Update: F*ck, after some more reboots 5 GHz is absent again. But the error message like above does not appear in the log. For testing I enabled airplane mode, reenabled wifi and have the (bad) 5 GHz back again. So weird!
Hmm, strange. Earlier today I tried to add directly the
Code:
 persist.sys.overlay.wifi5g=true
to /system/build.prop, but had no effect. After your promising message I added the
Code:
ro.wlan.mtk.wifi.5g=1
too there. But again without any effect. So unlike you I didn't recognized any effect of these changes. But at least you initial success give us some hope we are following the right track.

BTW: Do we try to set these properties in the right way? Do we use the right build.prop file? There are several other instances of this file in root, but with different content (in /system/product and in /vendor too). Because if I check by
Code:
getprop | grep wifi
I do not see our properties there. So it seems these are not set. Although added in /system/build.prop and restarted.
 
After restart only persist.sys.overlay.wifi5g remains to be set (according to getprop). ro.wlan.mtk.wifi.5g doesn't. Although both still in /system/build.prop. Even manual setting ro.wlan.mtk.wifi.5g in shell and cycling WiFi off and on doesn't help. Still the "bad" WiFi.
 
I'm still wondering why I had a good working 5 GHz wifi after the first reboot... Without any airplane mode. It seems to be an initialization problem.

Update: Because I had to flash my QiDroid again I did the same change in the /system/build.prop again like yesterday, rebootet and... no 5 GHz at all. Getprop shows:

[init.svc.vendor.wifi_hal_legacy]: [running]
[init.svc.wificond]: [running]
[init.svc.wifidisplayhalservice]: [running]
[net.tcp.buffersize.wifi]: [524288,2097152,4194304,262144,524288,1048576]
[net.tcp.delack.wifi]: [20]
[net.tcp.usercfg.wifi]: [1]
[persist.sys.overlay.wifi5g]: [true]
[ro.wifi.channels]: []
[ro.wlan.mtk.wifi.5g]: [1]
[sys.wifitracing.started]: [1]
[wifi.interface]: [wlan0]
 
Last edited:
CrimsonKnight13 Perhaps you can add to the guide that a backup/restore in TWRP does not use the opposite slot, but the active slot.

I just tried the TWRP-restore to both slots, but boot was always stuck showing the boot logo of QiDroid. After flashing QiDroid, GApps and Magisk to both slots the system started as normal. The TWRP-backup was the normal one, not the image backup, what maybe was the problem.

And like written in the guide, is it really neccessary to flash the same GSI-rom to both slots? I tried to flash LeOS and CAOS to slot_b with TWRP (with and without Magisk), but booting slot_b always ended in the bootlader, what seems to confirm the guide. Booting slot_a boots up to my regular QiDroid, what seems to not confirm the guide. :)
 
CrimsonKnight13 Perhaps you can add to the guide that a backup/restore in TWRP does not use the opposite slot, but the active slot.
I do confirm that. It is according to my experience too. I tried to find any confirmation of that on Internet, but didn't find any though. The same do apply for installing Magisk or GApps in TWRP, or resizing the partition. I would say that besides flashing the ROM everything else applies to the current slot, not the other one.
I just tried the TWRP-restore to both slots, but boot was always stuck showing the boot logo of QiDroid. After flashing QiDroid, GApps and Magisk to both slots the system started as normal. The TWRP-backup was the normal one, not the image backup, what maybe was the problem.
I do have similar experience. I've never be able to boot from slot_b after restoring the TWRP backup. It stuck exactly as for you. Even restoring all the partitions (not just boot, system and data) didn't help.
And like written in the guide, is it really neccessary to flash the same GSI-rom to both slots? I tried to flash LeOS and CAOS to slot_b with TWRP (with and without Magisk), but booting slot_b always ended in the bootlader, what seems to confirm the guide. Booting slot_a boots up to my regular QiDroid, what seems to not confirm the guide. :)
I do not think that is necessary. I do almost all my testing with different ROMs in each slot. But I'm not able to boot the slot_b always. I do restore the correct data backup before booting to the other slot or do wipe data of course. My life saver is the custom LOS16.0 in zip file flashed from the opposite slot in TWRP. Every time I can't boot I do flash custom LOS16.0 to this slot and it boots then. Then I can flash GSI ROM system image over it. But sometimes I have the "WiFi can't be turned on" issue I can't recover from otherwise than by flashing custom LOS16.0 again.

I already thought of not using the A/B ROMs' versions, but just the A versions. That make sense to me considering the way I do the testing (with different ROMs in each slot). But I haven't verified this theory yet.
 
I'm still wondering why I had a good working 5 GHz wifi after the first reboot... Without any airplane mode. It seems to be an initialization problem.

Update: Because I had to flash my QiDroid again I did the same change in the /system/build.prop again like yesterday, rebootet and... no 5 GHz at all. Getprop shows:
I wonder how did you managed to get your hack with ro.wlan.mtk.wifi.5g working at least. Even I added ro.wlan.mtk.wifi.5g=1 and (later) persist.sys.overlay.wifi5g=true as well to /system/build.prop and restarted I cannot get neither one in getprop output listed. And still do have those error messages in log. Currently I'm on QiDroid, bud had a similar experience on GSI LOS17.1. yesterday. I checked the build.prop file content and both lines are there.

BTW: Those two error messages (from my post #207) do appear in log after installing the Wifi Bonding Magisk module. The are not there before that.
 
Last edited:
Both slots need to be flashed since userdata is not a split partition. You can't have different ROMs jacking with the other's data. You'll wind up having boot failures.

As far as TWRP, that's known for Treble with restoration of only 1 slot. I honestly hate that TWRP doesn't have better freatures for Treble ROMs. You would need to backup both slots to make it work right, with probably some caveats involved.
 
Last edited:
Both slots need to be flashed since userdata is not a split partition. You can't have different ROMs jacking with the other's data. You'll wind up having boot failures.
Sure, I know that. That's why I always restore the correct/other data partition from backup before booting to the other slot. Or at least wipe data when want to start from the beginning for testing purposes. I believe that it should work this way and can have different ROMs in each slot. Just unsure whether I should use the A only ROMs rather then the AB ones in this scenario. To be honest I don't need the seamless update feature on tablet. To have the possibility to test some other ROM while keeping the working one make more sense to me.
As far as TWRP, that's known for Treble with restoration of only 1 slot. I honestly hate that TWRP doesn't have better freatures for Treble ROMs. You would need to backup both slots to make it work right, with probably some caveats involved.
I read somewhere that TWRP is not going to fully support the A/B (or maybe Treble) devices/ROMs. But not sure how valid information was that.
 
Sure, I know that. That's why I always restore the correct/other data partition from backup before booting to the other slot. Or at least wipe data when want to start from the beginning for testing purposes. I believe that it should work this way and can have different ROMs in each slot. Just unsure whether I should use the A only ROMs rather then the AB ones in this scenario. To be honest I don't need the seamless update feature on tablet. To have the possibility to test some other ROM while keeping the working one make more sense to me.

I read somewhere that TWRP is not going to fully support the A/B (or maybe Treble) devices/ROMs. But not sure how valid information was that.

Not to beatdown the TWRP devs, but that seems like they're falling behind the times (if this is true) since it's the main required feature for all new certified Android devices. I wonder if it will ever get outclassed by a new custom recovery that fully supports Treble going forward. One can only hope for one direction or another.
 
If you read this and consider that our TWRP-version 3.3.1-0 is from May 2019 (some month before Android 10 started), we can be glad that it works "somehow". :) From what source is this TWRP?
 
If you read this and consider that our TWRP-version 3.3.1-0 is from May 2019 (some month before Android 10 started), we can be glad that it works "somehow". :) From what source is this TWRP?
The TWRP backup/restore doesn't work as it should unfortunately. At least for me. Having the October Chinese ROM installed I flashed TWRP into both slots and did the backup of all the partitions in each slot that TWRP provides for backing up. Then flashed both slots with QiDroid. After some time I wanted to get back to the Chinese ROM, so I restored all the previously backed up partitions into both slots. But it doesn't boot to neither one now. Interestingly when I restored all the Chinese ROM backups right after I made them (just for a trial) it booted well into both slots. It is mystery for me why it doesn't work after flashing the GSI QiDroid ROM any more. The only explanation that comes to mind to me is that I resized system partitions in both slots in order to install GApps into QIDroid. So now the system size is larger than it was when I made the Chinese ROM backup. But I would expect the restore process to handle with that. There are no errors during restoring.
 
I still do struggle with the 5GHz WiFi. Cannot make it work well on GSI ROM. Therefore I moved back to October Chinese ROM where 5GHz WiFi works well. I searched through all the files in root filesystem for both the persist.sys.overlay.wifi5g and ro.wlan.mtk.wifi.5g strings and didn't find any file that contains at least one of these. No Magisk modules installed yet.
 
In TWRP backup, did you try the "System-Image" and "Vendor-Image" instead of "System" and "Vendor"? This will make an exact image of the whole partition (with dd) and restoring it should restore the former file system size too (just to exclude this as a possible source of error).
 
In TWRP backup, did you try the "System-Image" and "Vendor-Image" instead of "System" and "Vendor"? This will make an exact image of the whole partition (with dd) and restoring it should restore the former file system size too (just to exclude this as a possible source of error).
Hmm, are you telling me that System and System image are not two backups of two different partitions but two different ways of backing up the single partition? One as simple copy of its content and the other as the dd image? The same for Vendor. Anyway if so, it still doesn't explain the problem I'm facing to. Because in the restore process the System image is restored after System (i.e. it overwrites the restored System), so the result should be the same as I would restore System image only. I guess.
 
Hmm, are you telling me that System and System image are not two backups of two different partitions but two different ways of backing up the single partition? One as simple copy of its content and the other as the dd image? The same for Vendor.
OK, so I checked the TWRP's recovery.log file and can confirm that the thing is exactly as you describe. But still do not have an idea why it doesn't boot after full TWRP restore. It still means that the original and the restored states are not identical. And that's something I wouldn't expect from restoring. I cannot tell what might be wrong. Maybe it is not a good idea to restore Vendor, Modem, EFS partitions and these should stay untouched?

Edit:
There are other partitions being backed up by full TWRP backup. These are bluetooth, dsp, efs1, efs2, efsg partitions. But these cannot be selected for backing up nor for restoring from TWRP UI. So maybe these make some mess. Who knows...?
 
Last edited:
Back to the 5GHz WiFi mystery. I went through the bootloader opening procedure again. Then played with October Chinese ROM (where 5GHz WiFi works alright) for a while. Afterwards I flashed GSI QiDroid. During the very first boot to QiDroid I tried to connect to 5GHz WiFi. But after entering in the password the setting app has stuck. Killing it didn't help. Rebooting the tablet from UI has stuck too. Only hard restart using the long press of the power button worked. I managed to grab the logcat before tablet has rebooted:
Code:
qucii_qa6210_64:/ $ logcat | grep wifi
10-21 23:50:54.565  1187  1187 I [email protected]: Wifi HAL started
10-21 23:50:54.589  1187  1187 I [email protected]: Adding interface handle for p2p0
10-21 23:50:54.589  1187  1187 I [email protected]: Adding interface handle for wlan0
10-21 23:50:54.589  1187  1187 I [email protected]: Configured chip in mode 0
10-21 23:50:54.597  1378  1378 I wificond: create scanner for interface with index: 26
10-21 23:50:54.598  1378  1378 I wificond: subscribe scan result for interface with index: 26
10-21 23:50:54.601  1378  1378 E wificond: No Offload Service available
10-21 23:50:54.601  1378  1378 I wificond: Offload HAL not supported
10-21 23:51:00.848  1378  1378 I wificond: Regulatory domain changed to country:

So no overlay errors there. Then just after tablet has rebooted I grabbed the logcat again:
Code:
13|qucii_qa6210_64:/ $ logcat | grep wifi
01-02 04:00:02.234  1169  1169 I [email protected]: Wifi Hal is booting up...
01-02 04:00:02.146  1320  1320 I wificond: wificond is starting up...
01-02 04:00:05.534  1559  2009 I PackageParser: Skipping target and overlay pair android and null: overlay ignored due to required system property: persist.sys.overlay.wifi5g with value: true
01-02 04:00:05.534  1559  1559 W PackageManager: Failed to parse /product/overlay/treble-overlay-wifi5g.apk: /product/overlay/treble-overlay-wifi5g.apk (at Binary XML file line #5): null
10-21 23:50:50.210  1559  1559 I SystemServiceManager: Starting com.android.server.wifi.WifiService
10-21 23:50:50.318  1559  1559 I SystemServiceManager: Starting com.android.server.wifi.scanner.WifiScanningService
10-21 23:50:50.320  1559  1559 I SystemServiceManager: Starting com.android.server.wifi.p2p.WifiP2pService
10-21 23:50:50.336  1559  1559 D ConnectivityService: wifiOnly=false
10-21 23:50:51.379  1169  1169 I [email protected]: Wifi HAL started
10-21 23:50:51.450  1169  1169 I [email protected]: Adding interface handle for wlan0
10-21 23:50:51.450  1169  1169 I [email protected]: Adding interface handle for p2p0
10-21 23:50:51.450  1169  1169 I [email protected]: Configured chip in mode 0
10-21 23:50:51.479  1320  1320 I wificond: create scanner for interface with index: 26
10-21 23:50:51.479  1320  1320 I wificond: subscribe scan result for interface with index: 26
10-21 23:50:51.483  1320  1320 E wificond: No Offload Service available
10-21 23:50:51.484  1320  1320 I wificond: Offload HAL not supported
10-21 23:50:52.744  1320  1320 I wificond: Regulatory domain changed to country:

This time I was already able to connect to WiFi, but overlay errors appeared there. Just after the reboot! Nothing else has changed. As you can see, the first 8 lines from the second listing are missing in the firs one. The thing is that this is not a random behavior. It does the same every time during the very first boot to QiDroid (i.e. after factory reset too). Maybe not related to our 5GHz WiFi problem, but...

Edit:
Seems it doesn't have nothing to do with the WiFi connecting problem during first boot to QiDroid and the following hard restart. I rebooted in airplane mode now. Just after reboot these two overlay related error lines are in logcat. After some 4 minutes of doing nothing (not switching off the airplane mode, nor doing anything else), just waiting, these lines just disappear from logcat. They are no longer there. To be honest the more I want to understand the thing, the more I'm getting lost... :(
 
Last edited:
The current Android GSI has the following configuration:
  • Treble. The GSI includes full support for the HIDL-based architectural changes (also known as Treble) introduced in Android 8.0, including support for the HIDL interfaces. You can use the GSI on any Android device that uses HIDL vendor interfaces. (For more details, see Architecture resources.)
  • Verify boot. The GSI doesn't include a verify boot solution (such as vboot 1.0 or AVB). To flash the GSI to an device launching on Android 9 or earlier, the device must have a method for disabling verify boot.
  • File system. The GSI uses the ext4 file system.
  • Partition layout. The GSI uses system-as-root partition layout.
(From https://source.android.com/setup/build/gsi)


What's new in TWRP 3.4.0-0:

System As Root (SAR)​

  • Fix backup and restore using SAR - dianlujitao
  • System mount point - Chaosmaster
  • ORS - Chaosmaster
  • Zip install - Chaosmaster
  • system_root bind mount to /system - Chaosmaster
  • Autodetection of SAR - Chaosmaster
(From https://twrp.me/site/update/2020/06/24/twrp-3.4.0-0-released.html)


That may explain why 3.3.1-0 has some difficulties with backup/restore. Is someone fit in chinese to ask the dev of the X Neo TWRP to build an update? :)
 
That may explain why 3.3.1-0 has some difficulties with backup/restore. Is someone fit in chinese to ask the dev of the X Neo TWRP to build an update? :)
Not me unfortunately.:) But thank you for providing info on the current TWRP situation anyway.
 
There are many complains on WiFi issues here https://github.com/phhusson/treble_experimentations/search?p=1&q=wifi&type=issues. And the Magisk is said to be the cause of the problem in some. But it seems we need Magisk to be able to boot up to GSI ROM. Right? Should we try a different Magisk versions? Version 19.0 is said to be the last working one with WiFi in one post there. But those posts are more than one year old mostly. Any idea what Magisk version could be the right one?
 
The current Android GSI has the following configuration:
  • Treble. The GSI includes full support for the HIDL-based architectural changes (also known as Treble) introduced in Android 8.0, including support for the HIDL interfaces. You can use the GSI on any Android device that uses HIDL vendor interfaces. (For more details, see Architecture resources.)
  • Verify boot. The GSI doesn't include a verify boot solution (such as vboot 1.0 or AVB). To flash the GSI to an device launching on Android 9 or earlier, the device must have a method for disabling verify boot.
  • File system. The GSI uses the ext4 file system.
  • Partition layout. The GSI uses system-as-root partition layout.
(From https://source.android.com/setup/build/gsi)


What's new in TWRP 3.4.0-0:

System As Root (SAR)​

  • Fix backup and restore using SAR - dianlujitao
  • System mount point - Chaosmaster
  • ORS - Chaosmaster
  • Zip install - Chaosmaster
  • system_root bind mount to /system - Chaosmaster
  • Autodetection of SAR - Chaosmaster
(From https://twrp.me/site/update/2020/06/24/twrp-3.4.0-0-released.html)


That may explain why 3.3.1-0 has some difficulties with backup/restore. Is someone fit in chinese to ask the dev of the X Neo TWRP to build an update? :)
 
Here for testing a twrp-3.4! https://www.daten-hoster.de/file/details/910818/TWRP-CUBE_X_Neo-3.4.rar
I can't do great tests because I can only install a GSI if CN-20200522 is installed. No matter if twrp-3.3.1-0 or twrp-3.4.
The problem with my Imei i can reconstructed but i have "service status radio off" and telephon or data does not work. My system is totally shot. Could someone make a complete backup of all the individual partitions (without userdata) and make it available?
via adb or terminal emulator like this example:

adb shell
su
ls -l /dev/block/by-name/

adb shell
su
dd if=/dev/block/sda? of=/sdcard/rom/???.img
 
  • Like
Reactions: taro4
like this
Thank you! Did you create this by your own and if yes, how?

qucii_qa6210_64:/ # ls -l /dev/block/by-name/
total 0
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 abl_a -> /dev/block/mmcblk0p32
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 abl_b -> /dev/block/mmcblk0p33
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 apdp -> /dev/block/mmcblk0p36
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 bluetooth_a -> /dev/block/mmcblk0p45
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 bluetooth_b -> /dev/block/mmcblk0p46
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 boot_a -> /dev/block/mmcblk0p12
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 boot_b -> /dev/block/mmcblk0p13
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 cmnlib64_a -> /dev/block/mmcblk0p21
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 cmnlib64_b -> /dev/block/mmcblk0p23
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 cmnlib_a -> /dev/block/mmcblk0p20
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 cmnlib_b -> /dev/block/mmcblk0p22
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 ddr -> /dev/block/mmcblk0p43
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 devcfg_a -> /dev/block/mmcblk0p54
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 devcfg_b -> /dev/block/mmcblk0p55
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 devinfo -> /dev/block/mmcblk0p35
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 dip -> /dev/block/mmcblk0p34
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 dpo -> /dev/block/mmcblk0p38
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 dsp_a -> /dev/block/mmcblk0p30
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 dsp_b -> /dev/block/mmcblk0p31
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 frp -> /dev/block/mmcblk0p56
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 fsc -> /dev/block/mmcblk0p47
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 fsg -> /dev/block/mmcblk0p11
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 hyp_a -> /dev/block/mmcblk0p7
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 hyp_b -> /dev/block/mmcblk0p8
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 keymaster_a -> /dev/block/mmcblk0p18
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 keymaster_b -> /dev/block/mmcblk0p19
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 keystore -> /dev/block/mmcblk0p53
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 limits -> /dev/block/mmcblk0p40
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 logdump -> /dev/block/mmcblk0p57
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 logfs -> /dev/block/mmcblk0p42
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 mdtp_a -> /dev/block/mmcblk0p26
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 mdtp_b -> /dev/block/mmcblk0p27
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 mdtpsecapp_a -> /dev/block/mmcblk0p24
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 mdtpsecapp_b -> /dev/block/mmcblk0p25
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 misc -> /dev/block/mmcblk0p52
lrwxrwxrwx 1 root root 18 1970-01-21 10:55 mmcblk0 -> /dev/block/mmcblk0
lrwxrwxrwx 1 root root 22 1970-01-21 10:55 mmcblk0rpmb -> /dev/block/mmcblk0rpmb
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 modem_a -> /dev/block/mmcblk0p28
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 modem_b -> /dev/block/mmcblk0p29
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 modemst1 -> /dev/block/mmcblk0p49
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 modemst2 -> /dev/block/mmcblk0p50
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 msadp -> /dev/block/mmcblk0p37
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 persist -> /dev/block/mmcblk0p51
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 pmic_a -> /dev/block/mmcblk0p9
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 pmic_b -> /dev/block/mmcblk0p10
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 rawdump -> /dev/block/mmcblk0p60
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 rpm_a -> /dev/block/mmcblk0p5
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 rpm_b -> /dev/block/mmcblk0p6
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 sec -> /dev/block/mmcblk0p44
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 splash -> /dev/block/mmcblk0p39
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 ssd -> /dev/block/mmcblk0p48
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 sti -> /dev/block/mmcblk0p58
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 storsec -> /dev/block/mmcblk0p59
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 system_a -> /dev/block/mmcblk0p14
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 system_b -> /dev/block/mmcblk0p15
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 toolsfv -> /dev/block/mmcblk0p41
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 tz_a -> /dev/block/mmcblk0p3
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 tz_b -> /dev/block/mmcblk0p4
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 userdata -> /dev/block/mmcblk0p63
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 vbmeta_a -> /dev/block/mmcblk0p61
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 vbmeta_b -> /dev/block/mmcblk0p62
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 vendor_a -> /dev/block/mmcblk0p16
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 vendor_b -> /dev/block/mmcblk0p17
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 xbl_a -> /dev/block/mmcblk0p1
lrwxrwxrwx 1 root root 20 1970-01-21 10:55 xbl_b -> /dev/block/mmcblk0p2

But I'm on GSI and unfortunately there are no sda-entries in /dev/block.
 
Thank you a lot! That's promising. But seems not to work well yet. It worked fine until I decided to install LOS 16.0, see below. After flashing the 3.3.1-0 back, the same operation passed well.
IMG_20201208_171507.jpg
 
I tried to install the enableAutoBrightness-twrp.zip from https://androidfilehost.com/?w=files&flid=312417 as the auto brightness is not working in GSI. It should just put one apk file into /vendor/overlay. Although the installation goes well in TWRP without any errors, the apk is not being installed into the directory. Is there any reason for that?
 
I tried to install the enableAutoBrightness-twrp.zip from https://androidfilehost.com/?w=files&flid=312417 as the auto brightness is not working in GSI. It should just put one apk file into /vendor/overlay. Although the installation goes well in TWRP without any errors, the apk is not being installed into the directory. Is there any reason for that?
Maybe because of partition layout (sar) or you used the wrong slot. I copied the included apk to /vendor/overlay, changed file permission to 777 and rebootet, but is does not work (I did not check the log). Did you find any documentation for this fix at xda or anywhere else?
 
Last edited:
Maybe because of partition layout (sar) or you used the wrong slot. I copied the included apk to /vendor/overlay, changed file permission to 777 and rebootet, but is does not work (I did not check the log). Did you find any documentation for this fix at xda or anywhere else?
I did the same as you and copied the apk into /vendor/overlay. I had to use a different file manager than I'm using to be able to change directory permissions. After restart I'm able to turn auto brightness on and off (was not possible till now), but it does nothing in fact. Here is the log output (those same lines do appear on each turning on or off the auto brightness):
Code:
qucii_qa6210_64:/ # logcat | grep brightness
12-10 18:06:01.403  1765  2090 I ActivityManager: Start proc 5681:com.google.android.apps.turbo:ssb/u0a299 for broadcast {com.google.android.apps.turbo/com.google.android.libraries.smartbattery.brightness.library.BootReceiver}
12-10 18:06:33.652  1765  1765 V SettingsProvider: Notifying for 0: content://settings/system/screen_brightness_mode
12-10 18:06:34.233  1765  2090 I ActivityManager: Start proc 9210:com.google.android.apps.turbo:ssb/u0a299 for broadcast {com.google.android.apps.turbo/com.google.android.libraries.smartbattery.brightness.library.PhenotypeUpdatedReceiver}
12-10 18:06:35.222  1765  1765 V SettingsProvider: Notifying for 0: content://settings/system/screen_brightness_mode
12-10 18:06:36.325  1765  1765 V SettingsProvider: Notifying for 0: content://settings/system/screen_brightness_mode
12-10 18:06:39.615  1765  1765 V SettingsProvider: Notifying for 0: content://settings/system/screen_brightness_mode
12-10 18:06:40.987  1765  1765 V SettingsProvider: Notifying for 0: content://settings/system/screen_brightness_mode
12-10 18:07:28.852  1765  1765 V SettingsProvider: Notifying for 0: content://settings/system/screen_brightness_mode

Although I searched for it, I didn't find any documentation for this one, nor for some other "GSI fixes" I'm interested in on that page. I'm really curious from where these comes and for what purpose they are intended for. Therefore the auto brightness fix was the only one I was courageous enough to try. :)
 
Thank you! Did you create this by your own and if yes, how?



But I'm on GSI and unfortunately there are no sda-entries in /dev/block.
All you need is the CarlivImageKitchen https://androidfilehost.com/?fid=1322778262904020080
Copy stock boot.img and a TWRP.img to recovery-resources. I was looking for a TWRP-3.4 for a device with a / b partitions and landscape orientation.
Extract both images. New folders are created. copy recovery.img kernel from stock recovery to TWRP. Copy ramdisk / etc / recovery.fstab from stock to TWRP. Copy ramdisk / default.prop + ueventd.rc from stock to TWRP. Copy or adapt prop.default. Have a recovery created. finished.

sda was just an example

new example:
lrwxrwxrwx 1 root root 21 1970-01-21 10:55 boot_a -> / dev / block / mmcblk0p12

adb shell
su
dd if=/dev/block/mmcblk0p12 of=/sdcard/rom/boot_a.img

From partitions with A / B, A is enough
 
Back
Top