i compiled htcs official kernel and made some mods to it and it worked well, some wifi bug made me retry making it from 0, but now everytime i compile the kernel with the same toolchain but no modifications to it, lcd image disapears when booting, any help?, its a rare bug that shouldnt be happening since i already compiled and nothing like this happened before
Related
Just compiled AOSP master. It boots OK - however it looks like the screen driver and the radio are not working. I get excessive screen flicker and radio doesn't work at all.
Any ideas? Do I need to pull the proprietary drivers from the phone itself like we used to have to if using cyanogen's passion vendor overlay?
riddelln said:
Just compiled AOSP master. It boots OK - however it looks like the screen driver and the radio are not working. I get excessive screen flicker and radio doesn't work at all.
Any ideas? Do I need to pull the proprietary drivers from the phone itself like we used to have to if using cyanogen's passion vendor overlay?
Click to expand...
Click to collapse
I would suggest checking out Cyanogen's IRC. There are a number of people in there who could help you with this.
You need to run the extract-files script or the unzip-files scripts in devices/htc/passion before you can build. the product should be "full_passion"
After that step there are some things still not working such as camera and playing videos. Logcat says that the permissions in /dev are not correct and they cannot access the /dev/vdec and audio as well. I havnt got past that part yet using either 2.6.32 or 2.6.34 kernels.
I had the problem with replacing the broken glass on my TF201 that caused an inverted X-axis. Thanks to the solution provided by flumpster, I was able to get the fix the problem in the stock JB ROM (10.4.2.18) so that the inversion is corrected there, however the problem still exists in TWRP. I understand why. I asked in the TWRP IRC chat support if there were any fixes or patches that could be applied that would correct the issue. I was told that I needed to compile my own TWRP to load with the fix and that there were flags that could be set to flip either axis. I've looked into it but at the moment I am a little overwhelmed with the process of setting up the build environment and making sure I have everything in order to do this as I have never compiled anything from source before. Is there anyone out there that could possibly do a build of a compatible version of TWRP that would fix this issue? Many thanks to anyone who can.
Hello, I'm trying to port the franco kernel, but, it (given that the hardware is the same) should at least boot to the splash (you can flash OG kernels on the N4). Anyways, the kernel messages keep erroring with "[ 18.169082] msmfb_overlay_set: ioctl failed, rc=-34"
What I can't figure out is what msmfb error is. If I understand correctly, it is the adreno framebuffer having a write error.
It's initializing and the kernel boots correctly, but, I can't figure out what the issue with the framebuffer is...(The screen just shows a bunch of lines)
Here's the full log: http://pastebin.com/Gfs5ks8B
alastair53 said:
Hello, I'm trying to port the franco kernel, but, it (given that the hardware is the same) should at least boot to the splash (you can flash OG kernels on the N4). Anyways, the kernel messages keep erroring with "[ 18.169082] msmfb_overlay_set: ioctl failed, rc=-34"
What I can't figure out is what msmfb error is. If I understand correctly, it is the adreno framebuffer having a write error.
It's initializing and the kernel boots correctly, but, I can't figure out what the issue with the framebuffer is...(The screen just shows a bunch of lines)
Here's the full log: http://pastebin.com/Gfs5ks8B
Click to expand...
Click to collapse
I am having the same issue (different kernel, CM11). Did you found a fix?
ssrij said:
I am having the same issue (different kernel, CM11). Did you found a fix?
Click to expand...
Click to collapse
alastair53 said:
Hello, I'm trying to port the franco kernel, but, it (given that the hardware is the same) should at least boot to the splash (you can flash OG kernels on the N4). Anyways, the kernel messages keep erroring with "[ 18.169082] msmfb_overlay_set: ioctl failed, rc=-34"
What I can't figure out is what msmfb error is. If I understand correctly, it is the adreno framebuffer having a write error.
It's initializing and the kernel boots correctly, but, I can't figure out what the issue with the framebuffer is...(The screen just shows a bunch of lines)
Here's the full log: http://pastebin.com/Gfs5ks8B
Click to expand...
Click to collapse
hope you find a solution. What features will it have? It would be nice to have a kernel like Neighborhood but updated and for cm11.
Hello friends. So I am officially stumped. I have an issue with my kernel bootlooping. The interesting thing is, eventually it boots. After about 5-10 minutes of bootloop, it just boots. Weird right?
See, I merged a lot of code with my kernel and I'm not sure what is causing the bootloop. Here is my source: https://github.com/Snepsts/kernel_e980
It started on the add intelli-plug commit (which also includes intelli-thermal) it is probably important to mention disabling intelli-plug and thermal (and only one of each) all have the same outcome. I have also tried wiping the board back to its original state. I'm assuming its one of the core changes or nr_running changes. But I cannot find any way to get a log of the issue as it is not in a stable state during bootloop to give me logs from adb. I also cannot compile intelli-plug without the changes I made. And I really want to have it. I have looked for code I may be missing but cannot find any. I have made many different test kernels and taken the ramdisk and config from the latest official cm build for my device. Does anybody have any ideas? If you need anymore info please let me know. Thanks for any help in advance.
Sent from my LG-E980 running Carbon KitKat & E980 Kernel+
Hi all,
I have been trying to port TWRP for my device (a somewhat dated tablet, the Toshiba AT100 Thrive) but am running into problems with the kernel. It seems that no matter what I try, I am unable to get a kernel that actually boots.
I have the full OmniROM 4.4 source (with the end intention of building OmniROM for the tablet) as well as the kernel source from Toshiba (it is their patched version of 2.6.39.4). I also have a known working recovery (built around CWM with a 2.6.39 kernel) and stock boot.img and ICS ROM for the tablet.
I have extracted the kernel config from the ICS ROM as well as the kernel command line parameters.
Additionally, I also have access to two other ports (CyanogenMod) with kernel sources for this tablet and have found the AOSP source for the 2.6.39 and 3.10 kernels.
I am using CentOS 6.7 64-bit as my build OS, as that is what I have running on my Linux machine. I have OpenJDK 1.6 installed as opposed to the Oracle version (this might affect building OmniROM, but certainly shouldn't affect the kernel build).
I can successfully compile the kernel from the Toshiba sources using the tablet's config, as well as the CM-based port kernel sources and the 3.10 kernel (2.6.39 has some bugs straight out of the Google repository, which I can only get around if I disable yaffs2 in that kernel config or replace the yaffs2 source for it with the equivalent from the Toshiba sources).
I have the Android emulator set up on my Windows machine with both a ICS and KitKat AVD. After being unable to get the Toshiba-based kernel to start on my tablet, I thought I'd try building a kernel from the sources to run in the emulator.
At this stage, not a single one of my kernels has successfully booted either the tablet or the emulator. I have tried using the kernel config from the ICS AVD to build both the Toshiba 2.6.39 and AOSP 3.10 sources. The compile completes with no errors and a zImage is produced. However, it fails to boot the emulator.
Out of desperation, I have tried downloading the toolchains for Gingerbread and ICS and using those to build the kernels instead of the toolchain provided with OmniROM 4.4. That hasn't made a difference, either.
On the tablet, I have tried extracting the known working recovery kernel from the CWM recovery and using it as a prebuilt kernel. TWRP fails to load and the tablet always reboots after a few seconds, regardless of whether I set the kernel commandline parameters or not.
I am puzzled by this. Clearly there is something wrong but I am at a loss as to how to debug it, especially as do not appear to be able to produce a working kernel either for the tablet or emulator, despite having the sources and the kernel configs.
If anyone has any suggestions, they would be much appreciated.
Still have the same problem.
I've now tried building an Ubuntu 12.04 VM and compiling the kernels within that, but still get the same results.
Does anyone have any ideas or has anyone else ever run into this problem?
Did it ever work? I am experiencing this with SM-G950F. Very keen.
Unfortunately I was not able to resolve the issue. After spending a fair bit of time on it, I gave up and moved onto other projects. I've still got everything I need to try again, so at some point I may do so (especially as a use case for my tablet is possible again in the near future).