• Introducing XDA Computing: Discussion zones for Hardware, Software, and more!    Check it out!

[EOL][ROM][SDCLANG-6][microg-ready][OMS Service Mode]NLOS-15.1 - 20190208

Search This thread

64ashg

Senior Member
May 18, 2014
548
351
Redmi K20 Pro
Thank you so much man for integrating battery current and cycles in userspace. You claim its inaccurate, and i believe that, but at least we can know bigger degradations better now, at least in a relative sense now that this rom has it. Will be comparing values today and a month later :)
 
Last edited:

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
Thank you so much man for integrating battery current and cycles in userspace. You calim its inaccurate, and i believe that, but at least we can know bigger degradations better now, at least in a relative sense now that this rom has it. Will be comparing values today and a month later :)

The relative point of view was my first thought, too. But when checking with a bat capacity of 22% I had about 2840 mAh (with a nominal capacity of 3400 mAh), when testing the sysfs node a second time (after some charging with the car dash charger, to ~90%) max capacity has droped to 2710 mAh... this makes a degradation of ~4.5 % in about 30 minutes or so...

Tempreture, presure and humidity hasn't changed in that half hour.

So even if you read the value relativly, I doubt it's accurate. Take the values with a grain of salt... ;)

The cycle count seems plausible to me.

EDIT: Just came to my mind: Perhaps the increaswd bat temp (due to charging) may be a factor on the max capacity value.
 
  • Like
Reactions: thomasnsr

knpk13

Senior Member
Dec 11, 2015
2,115
1,231
OnePlus 3
OnePlus 3T
@nvertigo67 I know you like microg and when I came across this I thought of you lol. Its available for the OP3/T as well. I hadn't realized they had already started rolling out builds for Eelo. Looks like they renamed the project though.

https://e.foundation/mobile-phone-os/

Tried it on my old Moto G (osprey), was pretty sweet and polished. The launcher is exactly like iOS lol, although I don't have any complaints about that.

https://hackernoon.com/leaving-apple-google-e-first-beta-is-here-89e39f492c6f

Thought you might wanna check it out. It's still based on Nougat though. They're looking for developers so maybe you could join them since they seem to share your philosophy. Right now it's not much more than https://lineage.microg.org/ , but seems like they have a solid infrastructure and resources and will reach the ambitious goals they mentioned in their documentation. I'd keep an eye on them if I were you.
 
Last edited:
  • Like
Reactions: nvertigo67

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
@nvertigo67 Maybe it's time to re-synchronize the repo with the LineageOS sources instead of the Omni.

https://review.lineageos.org/#/c/LineageOS/android_external_chromium-webview/+/230039/

Best Regards

I'm thinking of this since a couple of days.

My problem is, that I've never could reproduce the issues some people had. Those issues were the reason to change to omnis webview. If I switch back to los webview and doesn't expirience any issues, this doesn't mean the origin issues (which made me switch webview) is solved... I'm really unsure how to proceed. :(

Disclaimer: I really hate issues, I can't reproduce...
 
  • Like
Reactions: thomasnsr

64ashg

Senior Member
May 18, 2014
548
351
Redmi K20 Pro
The relative point of view was my first thought, too. But when checking with a bat capacity of 22% I had about 2840 mAh (with a nominal capacity of 3400 mAh), when testing the sysfs node a second time (after some charging with the car dash charger, to ~90%) max capacity has droped to 2710 mAh... this makes a degradation of ~4.5 % in about 30 minutes or so...

Tempreture, presure and humidity hasn't changed in that half hour.

So even if you read the value relativly, I doubt it's accurate. Take the values with a grain of salt... ;)

The cycle count seems plausible to me.

EDIT: Just came to my mind: Perhaps the increaswd bat temp (due to charging) may be a factor on the max capacity value.
Yes its the temperature mate. I've been monitoring and using my phone 80-20% cycles and the degradation is minimal. You can see posts where i posted 27xxish mah since march 2018, its 2691 yesterday. I usually measure this at every discharge with around 50ish percent battery. My battery performance is remaining constant-ish with the mods i use. And a superb full 100-0 battery that falls in line with sultans rom and the degradation that came after switching to oreo and oneplus' unoptimised (relatively to m and n) blobs.
 
  • Like
Reactions: nvertigo67

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
UPDATE (maintainance; WireGuard 20181021; minor cleanup)

Build 20181021

Releasenotes:

NLOS-15.1 relays on Oreo firmware. OxygenOS 5.0.6 firmware or later is needed:

Changelog (Last repo sync: 21 Okt 2018, 13:09:00 CEST / 21 Okt 2018, 11:09:00 UTC):

NLOS-Bootlogo
nlos_bootlogo-v0.1.zip
back_in_black_bootlogo-v1.0.zip(pre ob16 OxygenOS Bootlogo)


[SIZE="+2"]DOWNLOAD[/SIZE]​

Happy flashing!
 

GR0S

Senior Member
Jun 10, 2012
889
192
nvertigo67 said:
GR0S said:
Could you please include ttl iptable module in NLOS future builds? Would be useful for tethering
In which way? Would that need some additional iptables rule?
BTW: why not discussing this in thee thread?
Some carriers drop packets with TTL other than 128 or 64 to prevent tethering.
CONFIG_IP_NF_TARGET_TTL is useful for fixing TTL values to certain number.
 

pedritopl

Senior Member
Apr 3, 2015
316
267
Nexus 7 (2013)
OnePlus 3T
Hi,

I noticed that I can't watch YouTube videos linked in articles in apps like Feedly (RSS reader). If I try to play them I only get the sound. No video. As I suspected it's related to WebView implementation. If I switch the WebView to Chrome in Developer Options the videos in Feedly work.
In Play Store the current versions are:
- Chrome 69.0
- Android System WebView 70.0
The WebView included in latest NLOS build is 68.0 and it's not attached to Play Store implementation so it does not update.
I remember there were some changes specific to NLOS related to WebView to address some problems (which ones?).
Questions which come to my mind:
- shouldn't the WebView in NLOS be updated?
- shouldn't it attach to Play Store version at least when gapps are installed?
- are there any disadvantages of selecting Chrome as the WebView implementation in Developer Options?
Thanks in advance for clarifying it for me :)
 

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
Some carriers drop packets with TTL other than 128 or 64 to prevent tethering.
CONFIG_IP_NF_TARGET_TTL is useful for fixing TTL values to certain number.

Activating the kernel option doesn't change anything. To change the ttl in iptables something like this is needed: https://rlworkman.net/howtos/iptables/chunkyhtml/x4820.html

To what new value should it be configuered? Since the new value would kick in for all carriers and tethered connection, is the new value proofed to NOT break the currently working tethers?

Is it possible to make iptables rules that only apply to one carrier?

Just activating the the kernel config doesn't make much sense to me.

- shouldn't the WebView in NLOS be updated?

Since you expirience issues with the included one: yes, of course.

- shouldn't it attach to Play Store version at least when gapps are installed?

No. They have different package names (nlos shipped: com.android.webview; playstore: com.google.android.webview). This is necessary to have more then one version installed.

- are there any disadvantages of selecting Chrome as the WebView implementation in Developer Options?
Thanks in advance for clarifying it for me :)

@Dirk Has answered allready.
 
  • Like
Reactions: pedritopl and Dirk

GR0S

Senior Member
Jun 10, 2012
889
192
Activating the kernel option doesn't change anything. To change the ttl in iptables something like this is needed: https://rlworkman.net/howtos/iptables/chunkyhtml/x4820.html

To what new value should it be configuered? Since the new value would kick in for all carriers and tethered connection, is the new value proofed to NOT break the currently working tethers?

Is it possible to make iptables rules that only apply to one carrier?

Just activating the the kernel config doesn't make much sense to me.
128 probably will be good choice, since it is default value for majority of devices.
However, my real intention for having ttl target module is to enable apps like this one https://play.google.com/store/apps/details?id=ru.glebchajah.ttlmaster&hl=ru
It simply runs iptable commands in root shell. I know that roms probably should not include features requiring root, but that kernel option is useful and has no drawbacks.

Also lineageos kernel doesn't build on ubuntu 18.10 (both in VM and WSL). It complains about missing files. Will debug that later. BTW: which ubuntu version build servers have?
 
  • Like
Reactions: nvertigo67

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
128 probably will be good choice, since it is default value for majority of devices.
However, my real intention for having ttl target module is to enable apps like this one https://play.google.com/store/apps/details?id=ru.glebchajah.ttlmaster&hl=ru
It simply runs iptable commands in root shell.

:good:

Also lineageos kernel doesn't build on ubuntu 18.10 (both in VM and WSL). It complains about missing files. Will debug that later. BTW: which ubuntu version build servers have?

"Ubuntu" is a slang acronym for "Dude, gentoo is to hard for me!"... ;)

Code:
adler ~ # cat /etc/gentoo-release
Gentoo Base System release 2.4.1
adler ~ #

Show me the first error or the complete make log - the final error message is pointless.

EDIT:
What exactly are you doing? Do you try to build the kernel stand alone (to use the kernelmwith i.e. anykernel)? Do you try to build a bootimage? Do you try to build the complete rom?

EDIT2:
Unpack the zip, flash the contained img file as bootimage from twrp (don't fprget to reflash boot image altering packages like magisk).
 

Attachments

  • boot-CONFIG_IP_NF_TARGET_TTL.zip
    12.3 MB · Views: 4
Last edited:
  • Like
Reactions: GR0S

GR0S

Senior Member
Jun 10, 2012
889
192
:good:



"Ubuntu" is a slang acronym for "Dude, gentoo is to hard for me!"... ;)

Code:
adler ~ # cat /etc/gentoo-release
Gentoo Base System release 2.4.1
adler ~ #

Show me the first error or the complete make log - the final error message is pointless.

EDIT:
What exactly are you doing? Do you try to build the kernel stand alone (to use the kernelmwith i.e. anykernel)? Do you try to build a bootimage? Do you try to build the complete rom?

EDIT2:
Unpack the zip, flash the contained img file as bootimage from twrp (don't fprget to reflash boot image altering packages like magisk).

Kernel boots fine, tethering also works. I used TTLFixer.apk and latest magisk.
Will do magisk module soon to avoid running unknown apk.
 

Attachments

  • TTLFixer.apk
    1.4 MB · Views: 8
  • Like
Reactions: nvertigo67

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
EDIT:
What exactly are you doing? Do you try to build the kernel stand alone (to use the kernelmwith i.e. anykernel)? Do you try to build a bootimage? Do you try to build the complete rom?

GR0S said:
drivers/gpio/gpiolib.c:23:38: fatal error: ../power/oem_external_fg.h: No such file or directory
#include <../power/oem_external_fg.h>
^
compilation terminated.
make[2]: *** [scripts/Makefile.build:258: drivers/gpio/gpiolib.o] Error 1
make[1]: *** [scripts/Makefile.build:402: drivers/gpio] Error 2
make: *** [Makefile:958: drivers] Error 2
export ARCH=arm64
export CROSS_COMPILE=/home/alim/lin/android_prebuilts_gcc_linux-x86_aarch64_aarch64-linux-android-4.9/bin/aarch64-linux-android-
make lineageos_oneplus3_defconfig
make
I'm building just kernel, I don't even have android sources downloaded. Also Thanks for building for me.


Try:
Code:
make -j[number of cores + 1] ARCH=arm64 CROSS_COMPILE=/home/alim/lin/android_prebuilts_gcc_linux-x86_aarch64_aarch64-linux-android-4.9/bin/aarch64-linux-android- lineageos_oneplus3_defconfig && make -j[number of cores + 1] ARCH=arm64 CROSS_COMPILE=/home/alim/lin/android_prebuilts_gcc_linux-x86_aarch64_aarch64-linux-android-4.9/bin/aarch64-linux-android-

To my knowledge ARCH and CROSS_COMPILE are make variables and not environment variables.

The file is there:
Code:
[email protected] /usr/local/src/los15/kernel/oneplus/msm8996 $ find * -name oem_external_fg.h
drivers/power/oem_external_fg.h
[email protected] /usr/local/src/los15/kernel/oneplus/msm8996 $
 

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
UPDATE (maintainance; kernel: iptables: ttl; webview)

Build 20181029

Releasenotes:

NLOS-15.1 relays on Oreo firmware. OxygenOS 5.0.6 firmware or later is needed:

Changelog (Last repo sync: 29 Okt 2018, 10:48:27 CET / 29 Okt 2018, 09:48:27 UTC):

NLOS-Bootlogo
nlos_bootlogo-v0.1.zip
back_in_black_bootlogo-v1.0.zip(pre ob16 OxygenOS Bootlogo)


[SIZE="+2"]DOWNLOAD[/SIZE]​

Happy flashing!
 

nvertigo67

Senior Member
Dec 28, 2011
6,007
12,319
Snapdragon LLVM Compiler for Android v6.0.9

Is anybody interested in testing a build with Snapdragon LLVM Compiler for Android v6.0.9?

I've just finished a sdllvm-6.x the first time (havn't uploaded all necessary fixes to the code base to git). It's booting, basics (phone, mobile data, wifi, tethering, camera, gps, sensors) are working, so far. I've used it for 60 m, now. So, I can't say anything on stability and bat life. It feels snappy and fluid.

Nonetheless I can't promise this build will leave the cat alive - you need to check first.

It's the very same code base then the last stable build plus all needed fixes for llvm 6.x.

ROM: https://drive.google.com/open?id=1dhUMhOGpd87g9yjd1VfLxVedFW8BwKrP
MD5: https://drive.google.com/open?id=1eAmXnn3kxXn5J3kjJo_cRko7GMxjTFB2

Happy flashing and good luck to Schrödinger's cat!
 

Top Liked Posts