Hello.
This is a try to keep other thread clean. Please post everything regarding kerneldevelopment in this thread.
Github: Original Elephone P9000 Kernel
Github: Forked Kernel by vsrookie
Github: Forked Kernel by @BlueFlame4
In BlueFlames kernelsources are already commits. These commits, correct wrong includes. Without you can't compile the kernel.
Thank you!
greeting
vsrookie
@BlueFlame4
I compiled the kernel with another cmdline:
Code:
console=tty0 console=ttyMT0,921600n1 root=/dev/ram vmalloc=496M slub_max_order=0 slub_debug=O androidboot.hardware=mt6755 androidboot.verifiedbootstate=green bootopt=64S3,32N2,64N2 lcm=1-R63315_GWMD13752A_drv fps=5540 vram=29229056 printk.disable_uart=1 bootprof.pl_t=658 bootprof.lk_t=2992 boot_reason=4 androidboot.serialno=K7CUOZJRJBAERG5D androidboot.bootreason=wdt_by_pass_pwk gpt=1 usb2jtag_mode=0
This is taken from /proc/cmdline.
The resulting kmsg is something weired. http://grork.lohnadmin.de/kernel/lstkms.txt
greetings
vsrookie
Ps: @BlueFlame4
I've testes your kernel also. This one:
BlueFlame4 said:
Here's a testkernel. I am pretty sure it will not boot. If it crashes, please reboot into recovery and post /proc/last_kmsg here. I just repacked stock boot.img with p9000_defconfig kernel.
http://d-h.st/clHI
Click to expand...
Click to collapse
As expected it didnt boot. Here is the last_kmsg from this.
http://grork.lohnadmin.de/kernel/test.txt
It looks like there is any encoding problem, or?
greetings
vsrookie
It looks like the video driver is crashing:
Code:
BUG: failure at /homeproject3/25%/2000706_c238_dx-andrOid+alps/kdrnel-3.18/drivers.misc/mediat%k/video/mt6755/dirp3ys/ddp[manafer.c:1616/dpmgr_ah%ck_stAtus
BlueFlame4 said:
It looks like the video driver is crashing:
Code:
BUG: failure at /homeproject3/25%/2000706_c238_dx-andrOid+alps/kdrnel-3.18/drivers.misc/mediat%k/video/mt6755/dirp3ys/ddp[manafer.c:1616/dpmgr_ah%ck_stAtus
Click to expand...
Click to collapse
I compile it with another cmdline to see what happens. I will leave the "lcm"-Parameter.
Two lines above in the kmsg there is:
Code:
[0.617501] (7)Y12swapper/0][DISP][disp^lcm_prbe #219]EBROR:[email protected] EROR: aan't found lcm drIver*R63315_GWMD3752A_drv�in linux kerNel driver [ 0.607528] (7)[0:swapp%r/ ][DISP][disp_lcm_probe #029]ERROP:[email protected]@L ERROR!!!Fo LCM Driver defined
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
Oddly enough the mt65xx_lcm_list.c is prebuilt as well as many other .c files which normally are not. Even every single LCM driver is prebuilt... I would say that Elephone wants to screw us but from my past experience with them it is more likely that they just have no clue what they are doing
Luckily we have other 6.0 sources (for example from Vernee) so we could reverse engineer them but I want to avoid that for now. I attached a new version of the kernel with a possible fix for the crash implemented. Let me know if it boots please
http://d-h.st/6hwF
BlueFlame4 said:
Oddly enough the mt65xx_lcm_list.c is prebuilt as well as many other .c files which normally are not. Even every single LCM driver is prebuilt... I would say that Elephone wants to screw us but from my past experience with them it is more likely that they just have no clue what they are doing
Luckily we have other 6.0 sources (for example from Vernee) so we could reverse engineer them but I want to avoid that for now. I attached a new version of the kernel with a possible fix for the crash implemented. Let me know if it boots please
http://d-h.st/6hwF
Click to expand...
Click to collapse
Yepp, just grep for project3. You will see all shipped drivers.
Your kernel dont boot. There is also a one line last_kmsg. Strange. Only ram console header, hw_status: 0, fiq step 0.
Did you compile with the original cmdline?
greetings
vsrookie
EDIT:
Compiled without lcm Parameter, also same problem. last_kmsg
I used the default cmdline, yeah. Essentially I just repacked the stock boot.img without changing anything else than the kernel.
BlueFlame4 said:
I used the default cmdline, yeah. Essentially I just repacked the stock boot.img without changing anything else than the kernel.
Click to expand...
Click to collapse
Hi.
I've a last_kmsg that is readable! It took even longer till crash.
What I've done:
I've setted up a user project3 and put the sources into a directory you have to create:
Code:
/home/project3/6755/20160706_c239_dx/android/alps/kernel-3.18/
Also you have to put the correct config_cmdline in your config. (look in second post).
Hope that will help other so that we have a running kernel.
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
Good job! Still weird that you have to use a static path but it is MediaTek code so everything is possible I tried to build lcm_list myself like so but then I get linker errors complaining for some "agold_lcm_power_down" symbol. When I grep for it, it is only found in the .shipped files so I think we have a dead end here and we somehow have to use the .shipped files.
What I am seeing from you log is that
Code:
[ 158.303970] (4)[1:init]init: Starting service 'md_monitor'...
[ 158.334455] (4)[264:logd.auditd]type=1400 audit(1469711965.478:374): avc: denied { open } for pid=5690 comm="md_monitor" path="/data/md_mon/mdlog_mon1_config" dev="mmcblk0p28" ino=1589251 scontext=u:r:md_monitor:s0 tcontext=u:object_r:system_data_file:s0 tclass=file permissive=0
[ 158.336867] (4)[1:init]init: Service 'md_monitor' (pid 5690) exited with status 1
[ 158.336890] (4)[1:init]init: Service 'md_monitor' (pid 5690) killing any children in process group
md_monitor crashes which is probably leading to the reboot. Could you boot it permissively (androidboot.selinux=permissive in cmdline) and see if that helps?
BlueFlame4 said:
Good job! Still weird that you have to use a static path but it is MediaTek code so everything is possible I tried to build lcm_list myself like so but then I get linker errors complaining for some "agold_lcm_power_down" symbol. When I grep for it, it is only found in the .shipped files so I think we have a dead end here and we somehow have to use the .shipped files.
What I am seeing from you log is that
Code:
[ 158.303970] (4)[1:init]init: Starting service 'md_monitor'...
[ 158.334455] (4)[264:logd.auditd]type=1400 audit(1469711965.478:374): avc: denied { open } for pid=5690 comm="md_monitor" path="/data/md_mon/mdlog_mon1_config" dev="mmcblk0p28" ino=1589251 scontext=u:r:md_monitor:s0 tcontext=u:object_r:system_data_file:s0 tclass=file permissive=0
[ 158.336867] (4)[1:init]init: Service 'md_monitor' (pid 5690) exited with status 1
[ 158.336890] (4)[1:init]init: Service 'md_monitor' (pid 5690) killing any children in process group
md_monitor crashes which is probably leading to the reboot. Could you boot it permissively (androidboot.selinux=permissive in cmdline) and see if that helps?
Click to expand...
Click to collapse
I am tested my kernel on cm with permissive selinux, same
Deepflex said:
I am tested my kernel on cm with permissive selinux, same
Click to expand...
Click to collapse
Hi.
Have you compiled your own, or is it stock? How did you compiled it. (Toolchain, cmdline, config?)
Thank you for reply.
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
vsrookie said:
Hi.
Have you compiled your own, or is it stock? How did you compiled it. (Toolchain, cmdline, config?)
Thank you for reply.
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
Click to expand...
Click to collapse
Hi.
In sources of CM
Hi!
Hi!
The last last_kmsg wasnt correct i think, but this is a really nice one.
Its readable and from boot.
last_kmsg
There are at least three really big problems with the kernel sources:
1. incorrect sources; you have to do too much to compile a kernel.
2. incomplete; at least three drivers are missing:
Code:
/home/project3/6755/20160706_c239_dx/android/alps/kernel-3.18/drivers/misc/mediatek/base/power/ppm_v1/src/mt_ppm_main.c
/home/project3/6755/20160706_c239_dx/android/alps/kernel-3.18/drivers/misc/mediatek/agold_pm_info/agold_pm_info.c
/home/project3/6755/20160706_c239_dx/android/alps/kernel-3.18/drivers/misc/mediatek/agold_hw_compatible/agold_hw_compatible.c
3. most drivers are compiled binaries
Cause this is fun to me, i ill go on!
greetings
vsrookie
vsrookie said:
Hi!
Hi!
The last last_kmsg wasnt correct i think, but this is a really nice one.
Its readable and from boot.
last_kmsg
There are at least three really big problems with the kernel sources:
1. incorrect sources; you have to do too much to compile a kernel.
2. incomplete; at least one driver is missing:
[CODEt/home/project3/6755/20160706_c239_dx/android/alps/kernel-3.18/drivers/misc/mediatek/base/power/ppm_v1/src/mt_ppm_main.c
[/CODE]
3. most drivers are compiled binaries
Cause this is fun to me, i ill go on!
greetings
vsrookie
Click to expand...
Click to collapse
Welcome to the world of MediaTek! I had the same problems with a Kazam device I got at XDA DevCon a couple of years ago - the company sent me the source code to play with and it was horrendous, gave up on it in the end :laugh:
Sent from my Elephone P9000 using XDA Labs
Jonny said:
Welcome to the world of MediaTek! I had the same problems with a Kazam device I got at XDA DevCon a couple of years ago - the company sent me the source code to play with and it was horrendous, gave up on it in the end :laugh:
Sent from my Elephone P9000 using XDA Labs
Click to expand...
Click to collapse
Thats not very encouraging. Maybe they will release some more infos or sources.... I dont give up... at least the hope!
greetings
vsrookie
PS: Why isnt this a Sony Phone!?
Unlock bootloader and try your kernel
Deepflex said:
Unlock bootloader and try your kernel
Click to expand...
Click to collapse
Hi.
At present i'm struggeling with lcmdriver. I have to find the correct initialization for it. I think if I find it, then its a little milestone.
Kernel want to boot, but screen fades out slowly with no kernelpanic.
So far, I've replaced many shipped drivers with sourcecode from other devices. Looks good till now.
kmsg will print proper messages without any other weird codecs.
Do you have sources for uboot?
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
vsrookie said:
Hi.
At present i'm struggeling with lcmdriver. I have to find the correct initialization for it. I think if I find it, then its a little milestone.
Kernel want to boot, but screen fades out slowly with no kernelpanic.
So far, I've replaced many shipped drivers with sourcecode from other devices. Looks good till now.
kmsg will print proper messages without any other weird codecs.
Do you have sources for uboot?
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
Click to expand...
Click to collapse
Hi!
Go to settings/about phone/and enable developer settings, after, go to developers settings and choose "Factory Unlock (unlocking the bootloader)"
Deepflex said:
Hi!
Go to settings/about phone/and enable developer settings, after, go to developers settings and choose "Factory Unlock (unlocking the bootloader)"
Click to expand...
Click to collapse
Hi. There is no option for that. Only "OEM UNLOCKING" (allow the bootloader to be unlocked).
Also if I use fastboot OEM unlock, I can't use volume up? Anyone tried to unlock bootloader?
greetings
vsrookie
Sent from my P9000 using XDA-Developers mobile app
Related
i have pushed the board files to my github to compile a fully working kernel for the GSM Hero. i dont own the device, but Behnaam has been working with me testing and such. anyhow everything works. the link to my source can be found below:
GoDmOdE-KeRnEl-SoUrCe
anyhow enjoy i do this for the community and ask for nothing but in return. Devs have at it!!!
Behnaam already beat you to the punch, link.
gu1dry said:
Behnaam already beat you to the punch, link.
Click to expand...
Click to collapse
i seen but i can update this one plus i guess i forgot to add two files . btw they are there now and it will compile for sure. (i just did it) enjoy
will it say no more huge battery drain ?
Quick and dirty compile by me in boot.img format (too tired to hack an update.zip together)
Credit to: Benocharm, and well everyone else who worked on this project... i just compiled it lol
EDIT: Had a lapse and forgot to compile a new wlan module... will reupload after school today
Cant post links yet sooo h**p://w w w .mega upload. c o m/ ? d = 724NZ7Q1
** = tt
remove spaces
Method One Flash: (For those with unlocked SPL)
fastboot flash boot newBoot.img
fastboot reboot
Method Two Flash:
adb push boot.img /sdcard
adb shell
cat /dev/zero > /dev/mtd/mtd2
- write: No space left on device [this is ok, you can ignore]
flash_image boot /sdcard/boot.img
reboot your phone
zjr8 said:
Quick and dirty compile by me in boot.img format (too tired to hack an update.zip together)
Credit to: Benocharm, and well everyone else who worked on this project... i just compiled it lol
Cant post links yet sooo h**p://w w w .mega upload. c o m/ ? d = 724NZ7Q1
** = tt
remove spaces
Method One Flash: (For those with unlocked SPL)
fastboot flash boot newBoot.img
fastboot reboot
Method Two Flash:
adb push boot.img /sdcard
adb shell
cat /dev/zero > /dev/mtd/mtd2
- write: No space left on device [this is ok, you can ignore]
flash_image boot /sdcard/boot.img
reboot your phone
Click to expand...
Click to collapse
Here ya go mate
http://www.megaupload.com/?d=724NZ7Q1
Are the video drivers included?
MerXury13 said:
Are the video drivers included?
Click to expand...
Click to collapse
I think that's very basic that they check it .
I'm more curious about other things that are not so obvious while testing:
volume keys while sleeping, battery usage, g-sensor, correct behavior of the audio jack (audio switches to jack instead of speaker), etc...
'gsm long text > 160 chars' issues can't have anything to do with the kernel used as I see it. That's something of the messaging app and / or the radio RIL used.
Hmm, new kernel to play with... might be the push I needed to try building my own android stuff to learn
Am I missing something with regards to the wlan module sources? Didn't see them so far...
Or am I being stupid (highly likely)
Are the video drivers included?
Click to expand...
Click to collapse
I think that's very basic that they check it .
I'm more curious about other things that are not so obvious while testing:
volume keys while sleeping, battery usage, g-sensor, correct behavior of the audio jack (audio switches to jack instead of speaker), etc...
'gsm long text > 160 chars' issues can't have anything to do with the kernel used as I see it. That's something of the messaging app and / or the radio RIL used.
Hmm, new kernel to play with... might be the push I needed to try building my own android stuff to learn
Click to expand...
Click to collapse
I asked 'cause in the cronos Rom, the developer said that he excluded the Video drivers to f**k up the other devs ...what a jerk
MerXury13 said:
I asked 'cause in the cronos Rom, the developer said that he excluded the Video drivers to f**k up the other devs ...what a jerk
Click to expand...
Click to collapse
dude, he only did that on his 'OC' version. The normal 'stable' version has them... reading is an artform
I asked 'cause in the cronos Rom, the developer said that he excluded the Video drivers to f**k up the other devs ...what a jerk
Click to expand...
Click to collapse
dude, he only did that on his 'OC' version. The normal 'stable' version has them... reading is an artform
Click to expand...
Click to collapse
It's too early for me^^ sorry
dipje said:
dude, he only did that on his 'OC' version. The normal 'stable' version has them... reading is an artform
Click to expand...
Click to collapse
The normal 'stable' version of his ROM is built on the same test kernal that every other 2.1 ROM is built on.
MerXury13 said:
I asked 'cause in the cronos Rom, the developer said that he excluded the Video drivers to f**k up the other devs ...what a jerk
Click to expand...
Click to collapse
Again... getting boring
Can someone please compile this one with overclock patch?I have no idea how to compile from source else I would do it
l0st.prophet said:
The normal 'stable' version of his ROM is built on the same test kernal that every other 2.1 ROM is built on.
Click to expand...
Click to collapse
Hence it got video drivers
Can anyone point me in the direction of a good write up on how to build a boot.img from this git repo please?im using ubuntu 10.4
so can we use this kernel on the vanilla roms?
anon2122 said:
Am I missing something with regards to the wlan module sources? Didn't see them so far...
Or am I being stupid (highly likely)
Click to expand...
Click to collapse
Do you have the latest commit? I believe initially there were some missing files. I know when I grabbed the source the first time i had wlan drive compile issues
I applied this to both cronos rom and to superhero rom and both got really sluggish after installing it (I used the boot.img posted among the first posts).
Up-to-Date Bare Repositories for TW/AOSP 4.2.2 Kernelsamsung-kernel-jfltexx
Short Info:
I've made this repository to have a central place for everyone to fork their kernel base from. The advantage of this repo is that it is always up-to-date (as in the latest linux kernel version). Imagine the difference from starting at a 3.4.0 kernel base and 3.4.52 kernel base, there are thousands of new patches in that patch range. :good:
Repository Url:
Code:
https://github.com/broodplank/samsung-kernel-jfltexx
Repository Info:
Name: samsung-kernel-jfltexx
Description: Bare repositories for jfltexx (i9505)
Branches: tw-4.2.2 and aosp-4.2.2 (aosp 4.3 coming when 4.3 kernel sources are released)
tw-4.2.2 = Stock i9505 TouchWiz Kernel from Samsung sources patched until latest linux kernel version
aosp-4.2.2 = i9505g Google Edition Kernel (for AOSP 4.2.2 Google Edition and TouchWiz 4.2.2) (for i9505)
Getting the sources:
If you want to compile the kernel directly without much changes it's advised to clone it:
Code:
git clone [url]https://github.com/broodplank/samsung-kernel-jfltexx[/url] -b branchname
If you want to make your own kernel it's advised to fork the kernel:
- Login at github
- Go to https://github.com/broodplank/samsung-kernel-jfltexx
- Click Fork at the right top of the page
Click to expand...
Click to collapse
Usage Info:
Initializing first config:
Code:
make VARIANT_DEFCONFIG=jf_eur_defconfig jf_defconfig
Configuration:
Easiest variant (QT based GUI, needs libqt4-dev libqt4-core and libqt4-gui)
Code:
make xconfig
Default variant:
Code:
make menuconfig
Other less popular variants:
(GTK based, doesn't need any special packages)
Code:
make gconfig
(Ncurses based, needs libncurses5, libncurses5-dev and libncursesw5)\
Code:
make nconfig
Building Kernel:
Code:
export ARCH=arm
export CROSS_COMPILE=/path/to/toolchain/bin/arm-eabi-
make -jx (where x is the amount of cpu cores + 1)
Toolchains:
Toolchains that are suited for building the sources are:
arm-eabi-4.6
Click to expand...
Click to collapse
Building boot.img:
Code:
./build_bootimg
(Choose the rom type you will be using and it will be created in build/output)
Flashing boot.img and modules:
Code:
./flash_bootimg
(This is a process that pushes boot.img to your phone and then to the kernel partition, after that you can choose whether to push modules or not, and finally you can choose to wipe dalvik-cache or not and it will reboot.
Updating Info:
If you have forked the repository do this to fetch the updates from this repository:
First add it to the remote:
Code:
git remote add bare [url]https://github.com/broodplank/samsung-kernel-jfltexx[/url]
git fetch bare
Then search for the commit(s) that you want to merge
For 1 commit:
Code:
git cherry-pick [commit hash here, this is the last part of a commits url]
(for example: https://github.com/broodplank/samsung-kernel-jfltexx/commit/612bbea08b67f16dd5628cb8cd8406973f8fe821 would be: 612bbea08b67f16dd5628cb8cd8406973f8fe821)
For a range of commits:
Code:
git cherry-pick [commit a hash]^..[commit b hash]
Where commit a is the oldest commit and commit b the newest.
for example:
Code:
git cherry-pick 0bff3bdba64c58ea6100d72763f2be3acc9659db^..612bbea08b67f16dd5628cb8cd8406973f8fe821
This would patch the sources from 3.4.40 to 3.4.52
Latest Repository Changes:
3-08-2013
- Switched to real samsung 4.3 sources
29-07-2013
- Updated aosp-4.2.2 and tw-4.2.2 to 3.4.55
25-07-2013
- NEW BRANCH: AOSP-4.3
- NEW BRANCH: TW-4.3
- Added MSM8960 Headers from Google!
23-07-2013
- Update to 3.4.54
14-07-2013
- Update to 3.4.53
- Merge sensor fix from faux123
13-07-2013
- Updated defconfigs for proper building
09-07-2013
- Update to 3.4.52
- Sensor fix by ausdim
- Large update to readme
06-07-2013
- Add samsung sources
- Patch to latest version 3.4.51
- add boot.img creator
05-07-2013
- Initial commit, make repository
Thanks, I just forked it.
I'm a noob at this but am learning fast. I've modified and compiled the 3.0.31 kernel for the 9210T but I'd like to now try 3.4. Appears none of the celox changes for board_msm8x60.c source came across from 3.0 to 3.4 so I think I've got my work cut out for me.
m0tter
This guide is not working for me.
If I follow your guide the resulting kernel boots, but I have no sound and no 3G (PIN pad doesn't show up).
I'm on MF8 stock ROM. Yours, Ausdims or my own (based on Samsung Update 2 sources) work fine.
This is what I did:
$ git clone https://github.com/broodplank/samsung-kernel-jfltexx -b tw-4.2.2
$ make VARIANT_DEFCONFIG=jf_eur_defconfig jf_defconfig
$ export ARCH=arm
$ export CROSS_COMPILE=/path/to/toolchain/bin/arm-eabi-
$ make xconfig (no changes!)
$ make -j3
$ ./build_bootimg && ./flash_bootimg (selected 1 for TW)
What am I doing wrong?
harise100 said:
This guide is not working for me.
If I follow your guide the resulting kernel boots, but I have no sound and no 3G (PIN pad doesn't show up).
I'm on MF8 stock ROM. Yours, Ausdims or my own (based on Samsung Update 2 sources) work fine.
This is what I did:
$ git clone https://github.com/broodplank/samsung-kernel-jfltexx -b tw-4.2.2
$ make VARIANT_DEFCONFIG=jf_eur_defconfig jf_defconfig
$ export ARCH=arm
$ export CROSS_COMPILE=/path/to/toolchain/bin/arm-eabi-
$ make xconfig (no changes!)
$ make -j3
$ ./build_bootimg && ./flash_bootimg (selected 1 for TW)
What am I doing wrong?
Click to expand...
Click to collapse
InitRamfs??
Permisions of initramfs?
ausdim said:
InitRamfs??
Permisions of initramfs?
Click to expand...
Click to collapse
Don't know.
I used the script to build the boot.img. If I do the same with my sources, it works.
In both cases, I haven't touched those files.
Sent from my GT-I9505 using xda app-developers app
harise100 said:
Don't know.
I used the script to build the boot.img. If I do the same with my sources, it works.
In both cases, I haven't touched those files.
Sent from my GT-I9505 using xda app-developers app
Click to expand...
Click to collapse
Im not on pc to check that and help you but your problem about radio (pin) sounds initramfs problem.
Unpack an stock boot.img and use that ramfs.
Thanks
ausdim said:
Im not on pc to check that and help you but your problem about radio (pin) sounds initramfs problem.
Unpack an stock boot.img and use that ramfs.
Thanks
Click to expand...
Click to collapse
Thank you for trying to help me... but it still doesn't work. Same result.
harise100 said:
Thank you for trying to help me... but it still doesn't work. Same result.
Click to expand...
Click to collapse
Tonight Ill do a little research on this and tell you the solution.
I remember when I started to make my first kernel, I had similar problems with the Samsung sources.
Wanam gave me his jf_defconfig file and it worked. He said that the config file by Samsung is crap. But I forgot to ask for details.
If I use my config file on your sources, it results in a bootloop, whereas they work fine with Samsung update 2 sources.
Does that make sense?
Sent from my GT-I9505 using xda app-developers app
Still figuring out, why i cant even successfully compile the sources. With ktoonse sources it works.
But i'm still a beginner. Will see about that xP
Gesendet von meinem GT-I9505 mit Tapatalk 2
.c0dy said:
Still figuring out, why i cant even successfully compile the sources. With ktoonse sources it works.
But i'm still a beginner. Will see about that xP
Gesendet von meinem GT-I9505 mit Tapatalk 2
Click to expand...
Click to collapse
What is your problem?
Sent from my GT-I9505 using xda app-developers app
harise100 said:
What is your problem?
Sent from my GT-I9505 using xda app-developers app
Click to expand...
Click to collapse
I hate it -.- xD
Now it worked. And I do not know why... As I am pretty new to this, as already said, - could this be the problem? :
I think I made this:
$ make VARIANT_DEFCONFIG=jf_eur_defconfig jf_defconfig
$ make xconfig
$ export ARCH=arm
$ export CROSS_COMPILE=/path/to/toolchain/bin/arm-eabi-
Click to expand...
Click to collapse
instead of:
$ make VARIANT_DEFCONFIG=jf_eur_defconfig jf_defconfig
$ export ARCH=arm
$ export CROSS_COMPILE=/path/to/toolchain/bin/arm-eabi-
$ make xconfig
Click to expand...
Click to collapse
So I used xconfig before setting ARCH and CROSS_COMPILE.
Could that be the problem?
Atm i can not paste the error. The first one had something to do with a buffersize. Could "fix" that by setting "CONFIG_FRAME_WARN=1024" higher. Because it said it was 1032.
Second error was something like "‘ret’ may be used uninitialized in this function [-Wuninitialized]".
---------- Post added at 08:08 PM ---------- Previous post was at 07:32 PM ----------
Will add a new post for this, if it's okay.
And i hope it's okay to ask in here
So, first, as wrote in my previous post the compiling worked.
Now, i did the following:
1. Changed in the config: "kernel compression mode -> LZO
2. Removed "Optimize for size"
3. And changed the flag in the makefile to O3.
kernel/workqueue.c: In function 'freeze_workqueues_begin':
kernel/workqueue.c:483:11: warning: array subscript is above array bounds [-Warray-bounds]
error, forbidden warning: workqueue.c:483
make[1]: *** [kernel/workqueue.o] Error 1
make[1]: *** Waiting for unfinished jobs....
While writing this I reset "Optimize for size" and restart compiling. Now it does seem to work.
So i guess there is a problem with the flag. Where is the right place of the flag?
If you want to use O2 or O3, it requires some modification of config and make files.
The build boot.img script worked for me right away.
Sent from my GT-I9505 using xda app-developers app
harise100 said:
If you want to use O2 or O3, it requires some modification of config and make files.
The build boot.img script worked for me right away.
Sent from my GT-I9505 using xda app-developers app
Click to expand...
Click to collapse
Ah okay, thanks. Will see if I can find a few more infos about that
broodplank1337 said:
Tonight Ill do a little research on this and tell you the solution.
Click to expand...
Click to collapse
Did you find anything?
I do have a working build environment and I can compile and install a kernel that is better than stock kernel.
Only these sources don't work. How is that possible?
Sent from my GT-I9505 using xda app-developers app
.c0dy said:
Ah okay, thanks. Will see if I can find a few more infos about that
Click to expand...
Click to collapse
check out this commit: https://github.com/broodplank/andro...mmit/69b6e56fb4151f547368fc77d5ad682fbf99ea7b
broodplank1337 said:
check out this commit: https://github.com/broodplank/andro...mmit/69b6e56fb4151f547368fc77d5ad682fbf99ea7b
Click to expand...
Click to collapse
I'll give it a try when I get home.
Sent from my GT-I9505 using xda app-developers app
harise100 said:
I'll give it a try when I get home.
Sent from my GT-I9505 using xda app-developers app
Click to expand...
Click to collapse
It will work if you have the right toolchain, I use this one:
http://www.mediafire.com/download/i...linaro_4.7.4-2013.07-build_2013_07_12.tar.bz2
It's an optimized Linaro toolchain (http://forum.xda-developers.com/showthread.php?t=2098133)
broodplank1337 said:
check out this commit: https://github.com/broodplank/andro...mmit/69b6e56fb4151f547368fc77d5ad682fbf99ea7b
Click to expand...
Click to collapse
Thanks Will try that.
I already compiled a kernel, but it just booted until the samsung logo.. after that instant reboot... :/
Since Spyke555 is busy, I updated cm10.1 with his last device tree.
You can download from http://dualhoki.vim.hu/bubor/dev/10.1/cm-10.1-20131024-UNOFFICIAL-i927.zip
Please read http://forum.xda-developers.com/showthread.php?t=2217288
It has same issues. Cm has security bugfix, and I want to keep cm up to date.
- Gps was fixed.
20131025: update to cm 10.1.3, increaes dalvik.vm.heapsize, android_packages_app_settings at b1755aad3d3ad2ac0e72297a6b4b60031d15ff69
20150826: cm security update
issue
same as other cm 10.1
settings -> about crash
ghost call: please try litekernel
Updated kernel here: http://dualhoki.vim.hu/bubor/dev/11.0/tegra-reserved-192mb-kernel-only-oldone.zip with lowered tegra reserved memory ( more usable memory)
Litekernel found here
Androide3 said:
Home long press not work
Sent from my SGH-I927 using xda premium
Click to expand...
Click to collapse
settings -> system -> hardware keys -> enable custom actions (check) and home key (long press) set to Recent app switcher
And you may set app switch key to voice search, so mic key will work as voice search.
Didn't have time to test everything but gave it a quick spin but noticed something right away... Couldn't download from playstore. Gave some error about couldn't find information.
Had to go back to spykes last version which is dd, will try to play more later. for reference I don't have any playstore issue after restore.
Neuntoter said:
Didn't have time to test everything but gave it a quick spin but noticed something right away... Couldn't download from playstore. Gave some error about couldn't find information.
Had to go back to spykes last version which is dd, will try to play more later. for reference I don't have any playstore issue after restore.
Click to expand...
Click to collapse
Was that clean install?
I can install app from play store.
bubor said:
Was that clean install?
I can install app from play store.
Click to expand...
Click to collapse
The only thing I didn't wipe was the sdcard.
I will give it a better shake later today. Were you using tonyg-nonneon gaps?
Neuntoter said:
The only thing I didn't wipe was the sdcard.
I will give it a better shake later today. Were you using tonyg-nonneon gaps?
Click to expand...
Click to collapse
yes
bubor said:
Since Spyke555 is busy, I updated cm10.1 with his last device tree.
You can download from http://dualhoki.vim.hu/bubor/cm-10.1-20130725-UNOFFICIAL-i927.zip
Please read http://forum.xda-developers.com/showthread.php?t=2217288
Click to expand...
Click to collapse
kudos bubor !
Unfortunately it seems the same issues that Kick is having with calls is present here as well so it's definitely seems to be something from CM. I haven't had anyone try to call me but I made a test call and was unable to end the call.
m1batt1 said:
Unfortunately it seems the same issues that Kick is having with calls is present here as well so it's definitely seems to be something from CM. I haven't had anyone try to call me but I made a test call and was unable to end the call.
Click to expand...
Click to collapse
please try this:
http://dualhoki.vim.hu/bubor/cm-10.1-20130726-UNOFFICIAL-i927.zip
ITs use opensource libril. Please feedback ASAP
bubor said:
please try this:
http://dualhoki.vim.hu/bubor/cm-10.1-20130726-UNOFFICIAL-i927.zip
ITs use opensource libril. Please feedback ASAP
Click to expand...
Click to collapse
I just got to work and won't be able to test until later tonight but I'll check it out and let you know.
Sent from my SGH-I927 using xda app-developers app
bubor said:
please try this:
http://dualhoki.vim.hu/bubor/cm-10.1-20130726-UNOFFICIAL-i927.zip
ITs use opensource libril. Please feedback ASAP
Click to expand...
Click to collapse
i tried it and calls work normally, i can end a call and i can recieve calls with no problem!
bubor said:
please try this:
http://dualhoki.vim.hu/bubor/cm-10.1-20130726-UNOFFICIAL-i927.zip
ITs use opensource libril. Please feedback ASAP
Click to expand...
Click to collapse
Hey! I'm at work currently but got chance to test it so here it goes what I found.
Calls: ok. I did 2 calls, receive 2 calls, did 1 outgoing call and put speaker, and receive 1 call and also test with speaker. So far all calls had audio.
Skype, and this is interesting. I tested from the glide to my work skype account, and in the glide the audio is still broken:
- Calling from glide to pc: I could hear myself ok ON the PC (so lets say the other side could hear me), but not FROM the pc, I heard it was all all deep and slow (nothing new to me there). I could even turn on the video call and it worked.
- Calling from pc to glide, skype in the glide just dies and restarts.
I'm sorry I cannot test more, I'm at work. I was using CM 10.1 RC2 from spyke555 and dirty flashed this one (just cleared cache and dalvik).
Bubor, if you can think a protocol for testing specific things (for example 1-test X and note results, 2-test Y and note resulted, after all tests report back to forum with results) somehting like that that would be equal among all that we are testing and might give you insight of the problems or if you need to test just 1 particulare thing, well, for that thing.
Keep it up people, let's give the devs what they need, tests, reports, results, they give us their time doing this, we must at least do the same!
Bubor, its CM10.1.2 o 10.1
Sent from my SGH-I927 using xda app-developers app
ugoPS said:
Bubor, its CM10.1.2 o 10.1
Sent from my SGH-I927 using xda app-developers app
Click to expand...
Click to collapse
10.1.2, you can check in about phone -> "view changelog"
Skype doesnt work. This build should be same as before, same bugs.
Okay. Did full wipe of sdcard and all, now no issues with play store.
Now I have never had voice search work with hot word detection, and that is still the case. Am I just lame or is this the way it should be. I had heard people talk about it working before.
Calls work. Both directions 4 times each tested. can hang up etc.
Unless there is something else supposedly fixed I guess it works for me.
How did you include the kernel? Just put the zip in the kernel/samsung/i927 folder?
Also are you still using the proprietary vendor files from The Covenant?
I know libril is a library for radio access but can you help me understand which one we need and how to include it in the build?
Thanks for any help you can provide still working on my first build after hours of downloading cm repo.
Neuntoter said:
How did you include the kernel? Just put the zip in the kernel/samsung/i927 folder?
Also are you still using the proprietary vendor files from The Covenant?
I know libril is a library for radio access but can you help me understand which one we need and how to include it in the build?
Thanks for any help you can provide still working on my first build after hours of downloading cm repo.
Click to expand...
Click to collapse
this device tree including kernel binary and modules. It would fail to build, if you include kernel source with this device tree.
Both device and vendor files you can found http://dualhoki.vim.hu/bubor/dev/cm-10.1-20130726-UNOFFICIAL-i927.tgz
You need one more thing that is https://github.com/CyanogenMod/android_hardware_samsung (just change _ to / for correct path)
bubor said:
this device tree including kernel binary and modules. It would fail to build, if you include kernel source with this device tree.
Both device and vendor files you can found http://dualhoki.vim.hu/bubor/dev/cm-10.1-20130726-UNOFFICIAL-i927.tgz
You need one more thing that is https://github.com/CyanogenMod/android_hardware_samsung (just change _ to / for correct path)
Click to expand...
Click to collapse
When you say kernel binary you mean the zip I download from kickass right? Are you saying I just put that in the kernel/samsung/i927 folder and it will automatically be included?
Just to clarify.
Get rid of device tree I got from spyke555's GIT and use the ones from your tgz. Now I assume I still put them in device/samsung/i927 and vendor/samsung/i927 right?
for the https://github.com/CyanogenMod/android_hardware_samsung
Do I just git clone that into the root? Or where should I be putting this?
Sorry for all the questions but I want to make sure I learn this and am a help later on.
Neuntoter said:
When you say kernel binary you mean the zip I download from kickass right? Are you saying I just put that in the kernel/samsung/i927 folder and it will automatically be included?
Just to clarify.
Get rid of device tree I got from spyke555's GIT and use the ones from your tgz. Now I assume I still put them in device/samsung/i927 and vendor/samsung/i927 right?
for the https://github.com/CyanogenMod/android_hardware_samsung
Do I just git clone that into the root? Or where should I be putting this?
Sorry for all the questions but I want to make sure I learn this and am a help later on.
Click to expand...
Click to collapse
Depend what is in your boardconfig.mk, if it has prebuild things, you would not put kernel source, but put kernel binary.
TARGET_PREBUILT_KERNEL := device/samsung/i927/prebuilt/kernel
TARGET_PREBUILT_RECOVERY := device/samsung/i927/prebuilt/recovery.img
If you use kernel source, it would be:
TARGET_KERNEL_SOURCE := kernel/samsung/i927
TARGET_KERNEL_CONFIG := tegra_bose_defconfig
Where keren source path, and the other is the defconfig what is in kernel/samsung/i927/arch/arm/configs
When you change _ to / in android_hardware_samsung you get android/hardware/samsung where android is the dir where you start, so you put to hardware/samsung. With command
git clone https://github.com/CyanogenMod/android_hardware_samsung hardware/samsung
Click to expand...
Click to collapse
Or you can put to the manifest, so repo sync will update to you.
If you wish, I could put whole directory to somewhere, but it would be something like 30giga.
Thank you Bubor! Couldn't of done it without you!
It appears to be building. I guess I won't know for sure until it is done.
bubor said:
Since Spyke555 is busy, I updated cm10.1 with his last device tree.
You can download from http://dualhoki.vim.hu/bubor/cm-10.1-20130726-UNOFFICIAL-i927.zip
Please read http://forum.xda-developers.com/showthread.php?t=2217288
It has same issues. Cm has security bugfix, and I want to keep cm up to date.
Last month I tried to fix tinyalsa, what fix most sound issue ( except bt calls). I'll keep trying.
Click to expand...
Click to collapse
Bubor, local forum users say that GPS is finally fixed in your build. I can confirm that, now GPS status and toolbox shows much more sattelites than previous CM versions. Thanx for that!
Today was released the source of the China version kernel
Can be downloaded here: http://consumer.huawei.com/cn/mobil...pport/downloads/detail/p6-u06-cn.htm?id=16381
Great new, but we need to support to unlock the bootloader
S34Qu4K3 said:
Today was released the source of the China version kernel
Can be downloaded here: http://consumer.huawei.com/cn/mobil...pport/downloads/detail/p6-u06-cn.htm?id=16381
Great new, but we need to support to unlock the bootloader
Click to expand...
Click to collapse
This is good, hopefully we can get some underclocking and some kernel governors on this beast of a phone
And most important, battery and wifi tweaks. There are the weakeness of this phone (i think)
Sent from my HUAWEI P6-U06 using xda app-developers app
S34Qu4K3 said:
Today was released the source of the China version kernel
Can be downloaded here: http://consumer.huawei.com/cn/mobil...pport/downloads/detail/p6-u06-cn.htm?id=16381
Great new, but we need to support to unlock the bootloader
Click to expand...
Click to collapse
Link is Dead
GuneetAtwal said:
Link is Dead
Click to expand...
Click to collapse
Here you got it, in my git repo: https://github.com/S34Qu4K3/P6-U06-JellyBean-Kernel-3.0.8---China-Version
S34Qu4K3 said:
Here you got it, in my git repo: https://github.com/S34Qu4K3/P6-U06-JellyBean-Kernel-3.0.8---China-Version
Click to expand...
Click to collapse
are you able to compile it ...?, how to get default device config file, adb pull /proc/config.gz gives file does not exists.
Latest Download Path : http://download-c.huawei.com/download/downloadCenter?downloadId=16381&siteCode=cn
Edit:
i am able to find default config for this kernel in this path :
/kernel/arch/arm/configs/hisi_k3v2oem1_defconfig
New modules CIFS/USB Ethernet
sreeramu said:
are you able to compile it ...?, how to get default device config file, adb pull /proc/config.gz gives file does not exists.
Latest Download Path : http://download-c.huawei.com/download/downloadCenter?downloadId=16381&siteCode=cn
Edit:
i am able to find default config for this kernel in this path :
/kernel/arch/arm/configs/hisi_k3v2oem1_defconfig
Click to expand...
Click to collapse
Thanks, got my Trendnet TU2-ET100 USB ethernet adapter working (modules attached)
And, CIFS support
oh, and ext3 filesystem also
Modules to be placed in /system/lib/modules are attached for kernel version 3.0.8-00788-g5d19ba2
build.prop
Nees to recompile for ext3 journaled
info:
this is based on the XZU recovery for locked bootloaders by @Krabappel2548
This thread will stay a dev and discussion thread for further development issues and fixes!
Download link can be found here: XZU recovery for locked bootloaders
THIS IS INTENDED FOR LOCKED BOOTLOADER DEVICES! But works on Unlocked bootloader too
THIS IS STILL IN DEVELOPMENT STAGE!
DO NOT EXPECT IT TO WORK PROBLEM FREE!
requirements:
rooted Xperia Z1
how to enter recovery:
when you see the Sony bootlogo and when device vibrates/leds light up u can use the entry keys to enter the corresponding recovery
VOLUME UP -> CWM
VOLUME DOWN -> TWRP
what works:
BOTH CWM 6.0.3.6 and TWRP 2.6.3.0 work
you can enter both recoveries and flash small files
what doesnt work:
in this particular version of CWM, adb doesnt work; it will be fixed later.. for debugging use TWRP
Shoudl be fixed:
TWRP and CWM both can destroy /cache and /system when trying to format it. Restoring your phone using an FTF is the only option left if that happens. You can however try to not format but use rm -rf instead in TWRP (TIC BOX!!), which will do for now.
regards,
Krabappel2548,
[NUT]
DooMLoRD
Last edit by Krabappel2548 on 19-10-'13
XDA:DevDB Information
Dual Recovery for Xperia Z1 (LOCKED BOOTLOADERS), a Tool/Utility for the Sony Xperia Z1
Contributors
DooMLoRD, [NUT], krabappel2548
Version Information
Status: Testing
Current Beta Version: 0.3
Beta Release Date: 2013-10-15
Created 2013-09-23
Last Updated 2013-10-19
Reserved
For anyone who does not own a Z1 (like me) and wants to search through the init actions.
Look also L2 Error detected! in 1089 Line. Usb don't work right and i think because we can't write in memory stack stack(0xe4859e70 to 0xe4859eb8) for lock bootloader. Problem it's in arch / arm / mach-msm but i don't find where!
As DooMLoRD's OP somehow does not fire the mention system... repeat for function
@Kali- @Bin4ry we need your help guys!
Oh now I got marked. I'll take a look at the weekend I don't have time Today sorry!
But just for me to know, ist that selfcompiled or a copy n paste version with a unpacked -> repacked ramdisk?
Have you checked against my posted recovery in Z Ultra forums? I can remember there was a similar problem on the locked bl recovery, you guys really seem to screw it somehow.
Also can i get a gdrive link to the files? Then I can grab it over my different systems more easy
Gesendet von meinem C6903 mit Tapatalk 4
Bin4ry said:
Oh now I got marked. I'll take a look at the weekend I don't have time Today sorry!
But just for me to know, ist that selfcompiled or a copy n paste version with a unpacked -> repacked ramdisk?
Have you checked against my posted recovery in Z Ultra forums? I can remember there was a similar problem on the locked bl recovery, you guys really seem to screw it somehow.
Also can i get a gdrive link to the files? Then I can grab it over my different systems more easy
Gesendet von meinem C6903 mit Tapatalk 4
Click to expand...
Click to collapse
It was a self-compiled version by @DooMLoRD, with a ramdisk slightly modified to get it to work for the Z1. What did you do to get it to work on the ZU? @krabappel2548 tried and failed there too...
EDIT: Just to be clear: it's not that the recovery does not work, it does, it's the kernel which triggers a reboot after 40 seconds because it seems to be unable to initialize the modem.
Bin4ry said:
Oh now I got marked. I'll take a look at the weekend I don't have time Today sorry!
But just for me to know, ist that selfcompiled or a copy n paste version with a unpacked -> repacked ramdisk?
Have you checked against my posted recovery in Z Ultra forums? I can remember there was a similar problem on the locked bl recovery, you guys really seem to screw it somehow.
Also can i get a gdrive link to the files? Then I can grab it over my different systems more easy
Gesendet von meinem C6903 mit Tapatalk 4
Click to expand...
Click to collapse
the boot hijack scripts, chargemon, are based on the ones for Xperia Z by @[NUT], with the necessary modifications, if any, for Xperia Z1
the ramdisks for the recoveries are using the same CPIO as the ones for unlocked bootloader...
recovery works well on unlocked bootloader with recovery integrated in kernel ramdisk...
will send u link over gtalk
did u post recovery for LOCKED bootloaders for Z Ultra?
Ok i did some first tests!
My own recovery works (see cm boot.img !). But this one does not work when ramdisk-recovery is integrated into your dualboot script!
So it is definetly a problem with the services from the "host" system. The ramdisks which get extracted also work fine when integrated into cm boot.img!
So wild guess here: You try to stop all services from host system but it seems not to work, since you dont really need anything running for recovery, but some of the host services dont want to stop here and waiting for the system to bootup. SOMETHING seems to reboot it then when nothing is happening
Need to see deeper whats happening, its quite a funny problem i think
Regards
I came to the same conclusion, but I had to give up as I had no idea what services could have been responsible
It might be the RIL service as it slingshots a lot of the radio/modem functionality... But I'm really not too sure...
Sent from my C6603 using xda app-developers app
Deleted
original ramdisk and the files from /system/etc are now in post 3 for anyone without a Z1 to check them.
Bin4ry said:
Ok i did some first tests!
My own recovery works (see cm boot.img !). But this one does not work when ramdisk-recovery is integrated into your dualboot script!
So it is definetly a problem with the services from the "host" system. The ramdisks which get extracted also work fine when integrated into cm boot.img!
So wild guess here: You try to stop all services from host system but it seems not to work, since you dont really need anything running for recovery, but some of the host services dont want to stop here and waiting for the system to bootup. SOMETHING seems to reboot it then when nothing is happening
Need to see deeper whats happening, its quite a funny problem i think
Regards
Click to expand...
Click to collapse
Did you find anything useful yet?
I've been wandering around in the files again myself, i've checked the qcom modem init already but i rechecked it, it seems that before chargemon is started, it already initialized a lot of the radio/modem hardware... it seems that at least one service is responsible for the communications layer between the kernel and modem, but I have been unable to determine which one... seeing the 40 second timeout before a reboot occurs it probably accounts for the service needing some time to start up.
Since kernel have initiated the modem, is it possible to write a procedure to set smem_get_entry flag. most likely to assign flag one to it.. New to android development..When i program a micro controller..we will be checking a flag in modem to see if it is initiated.. or sometimes.. i assigns a variable or memory address...so instruction will check that location for status rather than checking modem flag...as it takes more cycles...so i think this may be clearing some flags or memory or variable....
I did not had time yet to continue working but boyhood you guys should be correct. I am just thinking of the easiest way to fix it. Give me some more days please
Gesendet von meinem C6833 mit Tapatalk 4
Bin4ry said:
I did not had time yet to continue working but boyhood you guys should be correct. I am just thinking of the easiest way to fix it. Give me some more days please
Gesendet von meinem C6833 mit Tapatalk 4
Click to expand...
Click to collapse
I've had a chat about it with Dees Troy from TWRP about it, he suggested RIL as a possible culprit, or maybe watchdogd...
I don't have a z1 so I'm unable to test it on my own...
Sent from my C6603 using xda app-developers app
[NUT] said:
I've had a chat about it with Dees Troy from TWRP about it, he suggested RIL as a possible culprit, or maybe watchdogd...
I don't have a z1 so I'm unable to test it on my own...
Sent from my C6603 using xda app-developers app
Click to expand...
Click to collapse
For watchdog to reset an unexpected condition should occur.. for that something else have to go wrong...
and not sure- Radio Interface Layer is it not part of baseband?
jos_031 said:
For watchdog to reset an unexpected condition should occur.. for that something else have to go wrong...
and not sure- Radio Interface Layer is it not part of baseband?
Click to expand...
Click to collapse
Yes, it is: so it's tied in with the modem hardware...
Where do I download that recovery??
Jacob_Bentzen said:
Where do I download that recovery??
Click to expand...
Click to collapse
It's on the download tab in the project bar
Sent from my C6603 using xda app-developers app