CM-10.1 - Samsung Galaxy S Blaze 4G

I'm posting this here because I need some help from other devs. I'm in the process of porting ubuntu touch for you guys but I
m having a hard time getting the cm-10.1 kernel built (booting). I'm wondering if any of you other guys have been working on cm-10.1 and have a gotten a successfully booting kernel/ROM. If not, let's team up and see what we can come up with to get the blaze bumped up to the latest version.
So with that said, let's share our notes, findings, and issues...
I was able to get a booting kernel (just not the ROM) by doing a backwards merge with my jellybean LZ kernel (downloaded my kernel to my repo and merged with cm 10.1 sources). I'm thinking that this may be the way to go to get us rolling on cm 10.1

thomas.raines said:
I'm posting this here because I need some help from other devs. I'm in the process of porting ubuntu touch for you guys but I
m having a hard time getting the cm-10.1 kernel built (booting). I'm wondering if any of you other guys have been working on cm-10.1 and have a gotten a successfully booting kernel/ROM. If not, let's team up and see what we can come up with to get the blaze bumped up to the latest version.
So with that said, let's share our notes, findings, and issues...
I was able to get a booting kernel (just not the ROM) by doing a backwards merge with my jellybean LZ kernel (downloaded my kernel to my repo and merged with cm 10.1 sources). I'm thinking that this may be the way to go to get us rolling on cm 10.1
Click to expand...
Click to collapse
I have been working on it for a couple weeks... we can not seem to get a booting kernel... once we get a kernel, porting ubuntu phone looks to be a breeze... any help will be greatly appreciated.. Ericmm was working on it too but I have not been in touch with him lately..

Racing19969 said:
I have been working on it for a couple weeks... we can not seem to get a booting kernel... once we get a kernel, porting ubuntu phone looks to be a breeze... any help will be greatly appreciated.. Ericmm was working on it too but I have not been in touch with him lately..
Click to expand...
Click to collapse
I already have the Ubuntu ROM portion made. It's the kernel that is getting me stopped...lol
I know eric has been really busy lately and has about 500 projects going on as well. I'm taking a off the wall approach to it tonight and I will let you know how it turns out, and if all goes well, I'll update my repo...

thomas.raines said:
I already have the Ubuntu ROM portion made. It's the kernel that is getting me stopped...lol
I know eric has been really busy lately and has about 500 projects going on as well. I'm taking a off the wall approach to it tonight and I will let you know how it turns out, and if all goes well, I'll update my repo...
Click to expand...
Click to collapse
Your not alone, ericmm, dr4stic, uberpinguin and racing have all been working at it. If you need help, hit uberpinguin up on IrC.freenode.net at #Cyanogenmod or #cyanogenmod-dev. He's been trying to get help from somebody that has more experience with kernels. Pm dr4stic or beast mode, that's the best way to reach them. Dr4stic only pops in bout once a week. Good luck and if you need a tester. I gotcha:good:

Yep, the ubuntu part is easy. have had that since the day it was released... The kernel is where we are all stumped for some reason...
I am guessing that sooner or later one of us will figure it out... by then source will get released.. ha ha.. would be great if you could figure it out..
thanks for any info you can pass on..

possible help
Check into the galaxy exhibit 4g forums and get ahold of smartguy044! Great kernel creator!!! He created the first operating 3.0 overclock/ under/over volt kernel to work on the exhibit 4g and to my knowledge the only one! Possibly a great dev by the name of camcory could help also! Great devs and great guys

I think i got it!!!!!!!!!!!
Ok, I think I found the underlying issue!
Although the ROM didn't boot, the kernel sure did cuz I got a logcat...Lots of egl errors (/system/lib/egl/***) Which tells me that the kernel at least booted. We can work the hardware errors off as we go. However, with what I did, was a bit "dirty", so if you other devs (rancing, dr4stic, eric) could check this against yours, we might be able to come up with a definitive fix and have official cm-10.1. Now I wold like to say in concern with the ROM not booting, I do not have cm-10.1 repo on my computer, and I really don't want to because my android partition only has about 20gb left on it which is what I use for ccache. Instead, I used my ubuntu repo.
Here's what I did:
First, I add my jellybean branch of the lz kernel and synced it
I then added Cyanogenmod's cm-10.1 branch to my git config file:
Code:
git remote add Cyanogenmod https://github.com/CyanogenMod/android_kernel_samsung_msm8660-common.git
git pull Cyanogenmod cm-10.1
Then I built the ubuntu ROM using brunch t769 command
After it was finished, I pushed it to the phone, flashed and rebooted.
I got nothing...lol So I restored from my nandroid backup...
I diffed the defconfig for the skyrocket between jb and cm10.1, and found a few differences. So I made note of them and then check the defconfig of the t769 for those diffs. I added the missing ones, changed the differences, and removed the ones that was suppose to be there. I ran make clean then brunch t769, pushed it to the phone and rebooted and saw this:
Code:
I/cm ( 345): ____ _ _ ____ _ _ ____ ____ ____ _ _ _ _ ____ ___
I/cm ( 348): | \_/ |__| |\ | | | | __ |___ |\ | |\/| | | | \
I/cm ( 350): |___ | | | | \| |__| |__] |___ | \| | | |__| |__/
I/cm ( 356): Welcome to Android 4.2.1 / CyanogenMod-10.1-20130228-UNOFFICIAL-t769
Along with a sleuth of errors.
Now, I'm sure your asking what are the diffs thomas?
Entries to add or change:
Code:
# Linux/arm 3.0.60 Kernel Configuration
# CONFIG_MSM_SLEEP_STATS_DEVICE is not set
# CONFIG_MSM_CPU_PWRCTL is not set
# CONFIG_ARM_ERRATA_764369 is not set
# CONFIG_ARM_ERRATA_775420 is not set
CONFIG_ARM_FLUSH_CONSOLE_ON_RESTART=y
CONFIG_SYNC=y
CONFIG_SW_SYNC=y
# CONFIG_SW_SYNC_USER is not set
# CONFIG_LIBRA_SDIOIF is not set
# CONFIG_FRAMEBUFFER_CONSOLE is not set
CONFIG_ANDROID_LOW_MEMORY_KILLER_AUTODETECT_OOM_ADJ_VALUES=y
# CONFIG_HARDLOCKUP_DETECTOR_NMI is not set
CONFIG_HARDLOCKUP_DETECTOR_OTHER_CPU=y
CONFIG_HARDLOCKUP_DETECTOR=y
Entries to be removed:
Code:
# CONFIG_SHAPE_TOUCH is not set
CONFIG_FB_MSM_VSYNC_SYSFS=y
CONFIG_FB_MSM_HDMI_MSM_PANEL_HDCP_SUPPORT=y
# CONFIG_USB_UAS is not set
Now, I could be way off base with this, but that's why I need some of you other devs to test it in a cleaner way.
sync cm-10.1 source without merging
diff defconfig and make necessary changes
Build with cm-10.1 ROM resources
Boot and keep fingers crossed
Report back
Later today/this evening, I will post my logcat from boot to bootloop so we can diffs those as well...
If necessary, I will remove my ubuntu repo and pull cm stuff, but I would really not like to since syncing ubuntu took 4 days...lol
**EDIT** I meant to mention, I think the first line, Linux/arm 3.0.60 Kernel Configuration, could be the biggest issue since I have ran into a similar issue sometime back when I was porting ICS to the SGS4G...

I will make the changes you listed and run a build on cm10.1 source. will let you know how I make out...
Update: Using the changes listed above the kernel will not build...
arch/arm/mach-msm/board-msm8x60_celox.c:16531: error: 'hdmi_panel_power' undeclared here (not in a function)
make[1]: *** [arch/arm/mach-msm/board-msm8x60_celox.o] Error 1
make: *** [arch/arm/mach-msm] Error 2
make: *** Waiting for unfinished jobs....
will keep trying

Racing19969 said:
I will make the changes you listed and run a build on cm10.1 source. will let you know how I make out...
Update: Using the changes listed above the kernel will not build...
arch/arm/mach-msm/board-msm8x60_celox.c:16531: error: 'hdmi_panel_power' undeclared here (not in a function)
make[1]: *** [arch/arm/mach-msm/board-msm8x60_celox.o] Error 1
make: *** [arch/arm/mach-msm] Error 2
make: *** Waiting for unfinished jobs....
will keep trying
Click to expand...
Click to collapse
Oops...I removed the entire struct entry (from }; to the next }) for now... you might have some libcopybit errors, something about mpd4_overlay, and pow too...since I'm away from my computer I can't tell you exactly what I did, but when I get to it, I'll post it... you can find the libcopybit and pow fix on github via Google...but the overlay one I just remove the line(s)I think...I'm sure we'll need to add them back in, but I'm just trying to get this to boot first...Lol
Sent from my SAMSUNG-SGH-I727 using xda app-developers app
Here's the fix for libcopybit:
https://github.com/UnicornButter/ha...mmit/d40e33e88aae9c0c9c9839d69b44ad828dc18d84
And the pow error:
https://github.com/aokpx/platform_bionic/commit/edd30e9e0b66b2604f3a94346e67422d619b0558
and the mdp4_overlay.c
I'm just removed the entries that were throwing the error. Around lines 3641 and 3657 (after you delete the first entry). I want to say for the one at/near 3641 it was only 1 line and the second entry near 3657 it was 3 lines
I know the mdp4_overlay.c error is definitely NOT a fix, but I was honestly a bit stumped so I just went around it for now. It could also be the cause of egl (since this is a video driver) not working/loading properly which would cause the ROM to not boot...
**EDIT**
Ok, here is the full booting logcat... I started it while the samsung screen was on and it stopped when the phone rebooted. Before this, there was nothing. The phone would be on for about 2 or 3 minutes after the samsung was displayed and the it would turn off... So, I would say this is progress. I'm taking a little break for tonight. Feel free to rummage thru it and see what you guys can figure out. Please post anything helpful for the good of the community.
cm-10.1 partial boot

@thomas , I posted a link to this page to uberpinguin, maybe he can help as well.

I regressed...sorta
I compared the defconfig with that of a known booting device and saw many differences. So I thought I would merge the 2 defconfigs together and now it doesn't boot. However, I did not get any of the build errors, like the board-msm8X60_celox.c or the mdp4_overlay whatever. So, I am going to think of this as progression with minor regression...lol
So, I think if we can figure out what really needs to be in the defconfig and fix thes disply/egl errors, we will have cm-10.1 running.

thomas.raines said:
I regressed...sorta
I compared the defconfig with that of a known booting device and saw many differences. So I thought I would merge the 2 defconfigs together and now it doesn't boot. However, I did not get any of the build errors, like the board-msm8X60_celox.c or the mdp4_overlay whatever. So, I am going to think of this as progression with minor regression...lol
So, I think if we can figure out what really needs to be in the defconfig and fix thes disply/egl errors, we will have cm-10.1 running.
Click to expand...
Click to collapse
You can't just go combining defconfigs, they're different for a reason
The T989 is the general model i used as the first port and the differences from it to ours is a matter of driver selection. Their builds work (obviously) and ours don't. I've got a build environment going, have been pouring over the source differences, and have attempted a few things but progress is definitely stalled.
UberPinguin has had more time than I to delve deeper and I've been following his progress closely. Just to give you a peek behind the curtain, he's got a pretty sophisticated debugging set up going trying to figure out exactly why this thing crashes He's added debugging just about everywhere and still can't find the hold up.
Anyone looking to get this working, I wish the whole community luck on this. This one is an all hands-on thing.
(and yes, i'm alive, and sorry I've been so missing...)

To what dr4stic said he's right defconfig merges won't work for us because the blaze is different even if similar to the other msm8660.
Also, I've been working on this for over a month now with racing and we're at the same spot we were back then. I'm sure with dr4stic and uber and even CM helping we might get something booting.
As for the kernel well that's been the only thing holding back any rom from being released I had a 4.2.2 ROM running on the old kernel but no point on releasing that if we have to fix so much so I'm working on this as well on my free time. Hopefully I can come up with something to help out the big guys upstream (dr4stic and Uber)
Sent from my Galaxy Nexus using xda premium

Beastmode said:
To what dr4stic said he's right defconfig merges won't work for us because the blaze is different even if similar to the other msm8660.
Also, I've been working on this for over a month now with racing and we're at the same spot we were back then. I'm sure with dr4stic and uber and even CM helping we might get something booting.
As for the kernel well that's been the only thing holding back any rom from being released I had a 4.2.2 ROM running on the old kernel but no point on releasing that if we have to fix so much so I'm working on this as well on my free time. Hopefully I can come up with something to help out the big guys upstream (dr4stic and Uber)
Sent from my Galaxy Nexus using xda premium
Click to expand...
Click to collapse
Yeah,I know I over did it with the defconfig...
Sent from my SAMSUNG-SGH-I727 using xda app-developers app

Outside the Box?
Have you guy's thought about talking to Morfic (Trinity Kernels) or Jellette? They don't have a Blaze but if someone had an extra dev Blaze for Them to work with, They may be game.
Morfic:http://forum.xda-developers.com/member.php?u=1122306 >
Jellette: http://forum.xda-developers.com/member.php?u=2967115 > Both of these guy's have a lot of experience with Samsung Devices and Kernels.......just a thought

thomas.raines said:
Yeah,I know I over did it with the defconfig...
Click to expand...
Click to collapse
I forgot to mention why I mentioned it... there are, absolutely, a number of drivers on the other devices that would keep a kernel from booting up on ours. The MHL adapters immediately come to mind (anything having to do with HDMI really), and I believe a few others might have the same effect as well. The kernel simply won't boot.
Just wanted to make sure you didn't end up on a wild goose chase.
I'm spending the weekend downloading new kernel sources and diffing them to all hell, see if something's changed for the other devices that should've changed for us. Too many custom files... this is annoying...

One question.. What is the problem as of now. Is it still not booting??
Sent from my Nexus 5 with Key Lime Orange Mango Beef Tuna Pie

a case of beer to the person who can get ubuntu on our phone

Well the T-Mobile Galaxy S2 received Jelly Bean today. Official from sammy mobile. Is that like our device? If so lol. Head over to their forums.
Sent from my Nexus 5 with Key Lime Orange Mango Beef Tuna Pie

im_awesome_right? said:
Well the T-Mobile Galaxy S2 received Jelly Bean today. Official from sammy mobile. Is that like our device? If so lol. Head over to their forums.
Sent from my Nexus 5 with Key Lime Orange Mango Beef Tuna Pie
Click to expand...
Click to collapse
That's awesome, were next. They got their security update before us, we just got ours. My best guestimate is in late may early June.

Related

[DEV_THREAD]CyanogenMod 7 for Droid 2

***DEVS ONLY***ALPHA***DEVS ONLY***ALPHA***
*** THIS ROM WILL NOT BOOT *** THIS IS A DEV THREAD ***
You will have to sbf back to usable state this rom will not boot so be sure you have a fresh backup and your logcat ready
Don't even bother flashing if you are not comfortable with sbf flashing and logcat as this won't boot yet
I just need help bughunting
Let's get this working
you can repo what I have so far via
Code:
repo init -u http://github.com/JBirdVegas/cm4d2.git
repo sync
Sources:
github.com/cvpcs #cvpcs did all the work
github.com/JBirdVegas #all I did was port his work
github.com/JBirdVegas/android_vendor_cyanogen -b cm4dx_gingerbread #the branch that contains cvpcs's shadow cm port and my droid2 port
github.com/JBirdVegas/proprietary_vendor_motorola_droid2 #prop files
github.com/JBirdVegas/android_device_motorola_droid2 #device setup tree
github.com/CyanogenMod #The goal is a stable release for droid 2
use pastebin to post your logcats and link the logcats to this post so the group can discuss
# I'm not responsible if you brick your device #
# I'm not responsible for your alien sex dreams #
# I'm not responsible for anything you do ... EVER! #
MOST IMPROTANTLY THANKS
Thanks to The CyanogenMod Project for their open source ROM
Thanks to cvpcs for doing all the legwork getting CM for DX to boot
Thanks to greenromproject for all being a great place for beginners to learn to make ROMs
Thanks to XDA because I'm sure with the help of the community we can get CM7 for our Droid2's
Links:
To recover your droid2
SBF SystemOnly: http://bit.ly/b5Zf4V *This is the only one that works for me
SBF FullSystem: http://bit.ly/dir7fq
sbf_flash: http://dl.opticaldelusion.org/sbf_flash
dorootv3: http://kimete.com/droid/beta/dorootv3.sh
instructions for dorootv3: http://forum.xda-developers.com/showthread.php?t=782556
koush's droid2 bootstrap
koush dl: http://koush.tandtgaming.com//motorola/droid2/Droid2Bootstrap.apk
mirror: http://download766.mediafire.com/88...ejtwk/Droid+2+Recovery+Bootstrap+v1.0.0.3.apk
koush's application notes: http://www.koushikdutta.com/2010/08/droid-x-recovery.html
*************************DANGER*************************************
**********THIS ROM IS BROKEN AND WILL NOT BOOT**********************
CyanogenMod 7 Droid2 ROM: http://www.mediafire.com/?a78kh6udv534s28
********************************************************************
updates ...
Changelog will be here
first problem to solve...
http://pastebin.com/b7Sj1rNN
The the process repeats till the error on line 38 making the zygote throw the error on line 49
I'm still looking for how to fix I'm thinking my problem is in https://github.com/JBirdVegas/android_device_motorola_droid2/tree/gingerbread/hijack
... but I'm exhausted and calling it a night for now
Thanks for taking up this project! Very excited to see what becomes of it.
I'm not the best at programming for android on the source level, but it seems like cvpcs is still trying to fix quite a few things for CM4DX at this moment. Just a thought, but it may be better to wait a day or two until the cm4dx beta is released before trying to bang your head on getting a droid 2 port to work, since the DX build isn't even fully finished. I've been reading up on how all this works and if I can get a bit clearer understanding I'd love to help, but I think it might be beneficial to wait a few days before really trying to do anything about porting this over.
Just trying to help your exhaustion and keep you from getting a headache to quickly I'd guess by this time next week it would be worthwhile to hammer full force at getting a D2 port going!
Additionally based on how the phones work and the similarities between them, its actually highly possible that the DX version may actually flash on the droid 2. It obviously wouldn't be perfect as we know from the handful of files that are different, but the majority of finalized code for the DX beta build can probably be used once cvpcs updates it all. If what I understood is correct, the actual DX CM7 source that he is using for the alpha doesn't seem to actually be up on a public server just yet, and that he was planning on adding it with the release of the beta in the next few days.
Guys his CM4DX beta boots on the D2
Sent from my ADR6300 using XDA App
This was only possible because of cvpcs
THANK YOU CVPCS!!!
DX version and D2 version are the same except in nomenclature in some prop files this version has been compiled with the D2 prop files ... check cvpcs's twitter for DX CyanogenMod Love.
If you want to build repo CyanogenMod's source then drop these packages on top overwriting everything you are asked about.
https://github.com/JBirdVegas/propri...otorola_droid2 branch master
https://github.com/JBirdVegas/androi...otorola_droid2 branch gingerbread
https://github.com/JBirdVegas/android_vendor_cyanogen branch cm4dx_gingerbread
Here you go ...
********THIS IS STILL BETA***********
http://www.mediafire.com/?19oq9aam9cprw#
I opened a thread for this topic here: http://forum.xda-developers.com/showthread.php?p=14739355#post14739355
Work in Progress
Great to hear CM7 on the Droid 2, I just tried it out and everything works pretty well except camera, vibrating buttons, and sometimes my calls drop but looking forward to the stable version.
I took a video of it real quick:
I just posted in the other forum but just incase. I think it might be a good idea to see if we can merge the D2G and D2 development together rather than being forked out. Not sure if it is possible or not but if it is it would make development much faster.
http://www.droidforums.net/forum/droid-2-global/153481-cyanogen-mod-dx-will-hit-d2g-future-8.html

[REF] Github repo for SGH-I997R Source Release

Inspired by supercurio creating a master git repo for SGH-I997R development, I'm in the process of uploading today's source release from Samsung of the Gingerbread kernel for the Rogers Infuse 4G. (I'm surprised none of the other kernel hackers here have already done it... )
See supercurio's post http://forum.xda-developers.com/showthread.php?t=1054738 for more info on why git is good, and superior to kernel devs using tarballs directly from Samsung. This method is already unofficially how the kernel devs have been working - nearly every kernel here is forked from gtg465x's github repo.
I don't plan on directly doing development on this repo - It's just here as a starting point for everyone else to fork from. Some of us have slow upstream, so it's easier to pull a fork than it is to push a tarball import. However I might consider pulling in basic fixes to this repo (.gitignore, Makefile fixes - basically the same things supercurio has fixed in his unified git repos). (Edit: I'm starting to consider doing things a bit differently in order to allow this to stay as a relevant reference - After all there are certain "core" features like Voodoo Lagfix/Sound that have no known detrimental effects and are basically expected from any kernel used by the community.)
One thing is that most of the other ref repos are on a github "team" account - mine is on a personal account. I'm going to look into whether I can change this without breaking the fork relationships.
The repo is located at https://github.com/Entropy512/linux_kernel_sgh-i997r
Initramfs from Wednesday's dump is at https://github.com/Entropy512/initramfs_sgh-i997r
Note: If someone creates a better "master" repo for everyone to track, I'll happily edit this post to point there.
Update: Merged a pull request from LinuxBozo that has a bunch of .gitignore fixes and Makefile fixes.
Good work! Looking now.
Entropy512 said:
Inspired by supercurio creating a master git repo for SGH-I997R development, I'm in the process of uploading today's source release from Samsung of the Gingerbread kernel for the Rogers Infuse 4G. (I'm surprised none of the other kernel hackers here have already done it... )
See supercurio's post http://forum.xda-developers.com/showthread.php?t=1054738 for more info on why git is good, and superior to kernel devs using tarballs directly from Samsung. This method is already unofficially how the kernel devs have been working - nearly every kernel here is forked from gtg465x's github repo.
I don't plan on directly doing development on this repo - It's just here as a starting point for everyone else to fork from. Some of us have slow upstream, so it's easier to pull a fork than it is to push a tarball import. However I might consider pulling in basic fixes to this repo (.gitignore, Makefile fixes - basically the same things supercurio has fixed in his unified git repos)
The repo is located at https://github.com/Entropy512/linux_kernel_sgh-i997r - There is nothing visible there yet as the push is still in progress. It should be done within 30 minutes or so.
I don't have an initramfs repo yet - we need an initramfs dump from a device first.
Note: If someone creates a better "master" repo for everyone to track, I'll happily edit this post to point there.
Click to expand...
Click to collapse
Absolutely awesome news :-D
Sent from my SAMSUNG-SGH-I997 using XDA Premium App
Oh blast, hadn't even seen that kernel code was available... thanks Entropy!
A personal favor I'd like to ask of people forking - try to keep things to one feature per commit. An example being LinuxBozo's ext4 backport to the Froyo kernels - it was simple and easy to apply to other kernels. As opposed to someone had a massive supercommit that added a whole bunch of stuff at once, making it hard to either pick-and-choose or hard to apply if they had already applied a single-feature commit for one of the features. (One example being a commit that combined something like TinyRCU, jhash3, and BFQ all in one. I don't remember exactly what it was.)
Entropy512 said:
A personal favor I'd like to ask of people forking - try to keep things to one feature per commit. An example being LinuxBozo's ext4 backport to the Froyo kernels - it was simple and easy to apply to other kernels. As opposed to someone had a massive supercommit that added a whole bunch of stuff at once, making it hard to either pick-and-choose or hard to apply if they had already applied a single-feature commit for one of the features. (One example being a commit that combined something like TinyRCU, jhash3, and BFQ all in one. I don't remember exactly what it was.)
Click to expand...
Click to collapse
Defuse kernel
Sent from my SAMSUNG-SGH-I997 using XDA Premium App
Is that all we needed for gb roms or is there still more needed?
Sent from my SAMSUNG-SGH-I997 using XDA Premium App
nelomen said:
Is that all we needed for gb roms or is there still more needed?
Sent from my SAMSUNG-SGH-I997 using XDA Premium App
Click to expand...
Click to collapse
Initramfs, radio firmware dump, proper /system dump. Ours is incomplete.
Updates:
Merged LinuxBozo's pull request with build fixes
Merged LinuxBozo's pull request with the rainbow fix
Added initramfs repo
Entropy512 said:
Updates:
Merged LinuxBozo's pull request with build fixes
Merged LinuxBozo's pull request with the rainbow fix
Added initramfs repo
Click to expand...
Click to collapse
Nice!! Anything buildable yet?
ookba said:
Nice!! Anything buildable yet?
Click to expand...
Click to collapse
should be buildable since initramfs is available, tomorrow gtg's voodoo commit should be available too. this is all very exciting.
ookba said:
Nice!! Anything buildable yet?
Click to expand...
Click to collapse
A few more things got pulled in last night.
Kernel - not much
Initramfs - Working CWM - or at least it comes up, it hasn't been fully tested functionally yet
As of last night a few people have custom kernels running including myself, although some people still can't boot them. CWM wasn't working until past midnight...
Next up is voodoo - Not sure if I'll include that in the main initramfs repo or not. Maybe as a branch. It is a fairly standard mod, so it should be in a readily accessible repo
Bad news - as I said before CWM hadn't been fully tested.
Apparently, after further testing, it still does indeed have some pretty severe issues.
Sadface
10c
P1 Wookie said:
Sadface
10c
Click to expand...
Click to collapse
We have what seems to be working CWM and Voodoo thanks to gtg - in testing now.
The trolls in IRC aren't helping. Not a good time for the ops to be AFK.
trolls suck.
GL guys.
Entropy512 said:
We have what seems to be working CWM and Voodoo thanks to gtg - in testing now.
The trolls in IRC aren't helping. Not a good time for the ops to be AFK.
Click to expand...
Click to collapse
Gotta love the trolls.
Good luck guys. I know you all will figure it out
Well, thanks to the blasted trolls and the unfortunate circumstance of my only internet being my phone, I'm borked from being able to participate in the main infuse4g room as webchat is banned (for a good reason) and I can't get past the SASL issue otherwise.... oh well.
bedwa said:
Well, thanks to the blasted trolls and the unfortunate circumstance of my only internet being my phone, I'm borked from being able to participate in the main infuse4g room as webchat is banned (for a good reason) and I can't get past the SASL issue otherwise.... oh well.
Click to expand...
Click to collapse
why no use of a client?
drowningchild said:
why no use of a client?
Click to expand...
Click to collapse
Finally got it to work, had to hack a command line client.... urg.

[CLOSED]||Evervolv 2.2]

This thread is no longer active. Jomeister plans to continue work on Evervolv. He will open a new thread for that purpose.
Introduction
The purpose of this thread is to coordinate work on the AOSP-based Evervolv 2.2 ROM for Skyrocket (and its sister, T989).
Skyrocket is not an officially supported ROM by Evervolv, but we have fully integrated device and vendor directories in their tree. Therefore we get fixes from the main tree and also have the ability to push fixes to Evervolv's tree. This is something we could never do when we worked directly off the Code Aurora source.
Our desire is to give the dev community a single codebase to coalesce around. Hopefully, with more devs working on the same project, we can progress faster towards AOSP.
To build, follow the instructions here (choose ev_skyrocket-eng at the lunch menu):
https://github.com/Evervolv/android
Or, the short, short version:
Run "repo status" and make sure you have no local mods.
Run "repo sync" and make sure it completes without error.
Run "rm -rf out" to remove any old compiled stuff
Run ". build/envsetup.sh" to set up your environment
Run "lunch ev_skyrocket-eng" to select the target
Run "make -j4 otapackage" (adjust -j option to match your # of processors)
What works:
2D/3D graphics, hw accel of gui, bluetooth, gps, nfc, wifi, sdcard (int+ext), MTP/PTP over USB
What doesn't work:
audio, calls, camera, cell data, sms, anything else you can think of
Issues are being managed on Github:
https://github.com/Evervolv/android_device_samsung_skyrocket/issues
Want a ROM link?
Then follow the instructions to build a ROM! The point, at this time, is to get devs involved in the process. The ROM is not ready for anyone to try out or use at this point. There are many problems left to solve.
Want to help?
Are you a dev with a Skyrocket and a burning desire to get AOSP for your phone? Please, check out Evervolv and start debugging! The Github Issues link above is a great starting point.
Jomeister and RaverX3X graciously agreed to join forces here. We hope other devs join too!
Current Dev Contributors:
jomeister15
RaverX3X
topprospect
Credit and Thanks:
Xboarder56 for the original Evervolv Skyrocket device tree
romracer for the unsecure mods to the ICS kernel
Changes Log:
04/11/12: Posted succint build instructions
04/11/12: Fixed unstable-ness of GUI (link)
04/06/12: Build now boots on its own boot.img (link)
this is on 4.0.4 now?
Sent from my GT-P7510 using XDA Premium HD app
BaconStep said:
this is on 4.0.4 now?
Sent from my GT-P7510 using XDA Premium HD app
Click to expand...
Click to collapse
yes because google has released 4.0.4 sources and ota updates to nexus.
wali01x said:
yes because google has released 4.0.4 sources and ota updates to nexus.
Click to expand...
Click to collapse
Sweet!
Sent from my GT-P7510 using XDA Premium HD app
Looks great, will try building/debugging when I get my Ubuntu box up again!
Sent from my SAMSUNG-SGH-I727 using XDA
I'll be helping out as well, just got me a T989. I was working on cm9, but I'll switch to this. Hope I can help.
This is obviously a stupid question, and probably not the place to be asking it, but it is going to be asked anyway so.. I understand what the device tree is, but why is it not possible to pull one from the existing ICS leak?
giant420 said:
This is obviously a stupid question, and probably not the place to be asking it, but it is going to be asked anyway so.. I understand what the device tree is, but why is it not possible to pull one from the existing ICS leak?
Click to expand...
Click to collapse
No stupid questions here. A fully compiled ROM like the ICS leak obscures a lot of the details about how to build the ROM from scratch. You have to reverse engineer a lot of what Samsung did to compile AOSP, plus how to avoid pulling in too much Samsung bloat code. Hope that helps.
more or less were going to keep as much samsung out of it as we possibly can . Yes u can kang in stuff from the leak but then that actually causes more issues in the long run then doing it right the first time ;P.
bsbachert said:
I'll be helping out as well, just got me a T989. I was working on cm9, but I'll switch to this. Hope I can help.
Click to expand...
Click to collapse
I think we as a community should still continue working on things such as CM9 and AOKP in addition to Evervolv, just to give people as many options as possible. The good thing that Evervolv device trees are very similar to those required by CM9 and AOKP, so with a few hacks here and there I'm sure we can get those building as well. I don't know about the other devs, but I definitely have enough time to collaborate on multiple projects, as long as I'm not the sole carrier.
ur killing me. I finally managed to get media server to stay up for msm8660 audio pulled from the leak with jos isotope a4 with some other libs and bins. acdb pulling audio policy was still a problem though.
This is definitely for the better though. I was uncomfortable with all those Samsung libs bins and blobs, especially after staring at them under IDA. like raver said it woulda been a nightmare later on.
This is great though, everyone on the same pile should make this go a lot faster.
im very excited for this.
What's IDA?
Sent from my SGH-T989 using XDA
Wow. This build is going to b epic! Can't wait guys. Nice work so far
LawStud3nt said:
What's IDA?
Click to expand...
Click to collapse
http://en.m.wikipedia.org/wiki/Interactive_Disassembler
crazy what u can find in those blobs. helps a lot for debugging though
all I know is I love evervolv on my nexus one its super baller... Thanks to all 3 of you guys
Sent from my SAMSUNG-SGH-I727 using xda premium
Any updates? Got ants in my pants waiting for this one!
softballjunkee13 said:
Any updates? Got ants in my pants waiting for this one!
Click to expand...
Click to collapse
Thanks for the interest.
We will post updates as they happen. Right now we are just trying to organize and get everyone on the same page.
Sent from my SAMSUNG-SGH-I727 using Tapatalk
topprospect said:
Thanks for the interest.
We will post updates as they happen. Right now we are just trying to organize and get everyone on the same page.
Sent from my SAMSUNG-SGH-I727 using Tapatalk
Click to expand...
Click to collapse
OK...enjoy the holiday as well!

[DEV] CM7.2 Motorola Triumph

More information is listed at the site below
http://androidforums.com/triumph-al...-files-7-09-2012-5-30-pm-mst.html#post4641932
Just love the fact that we're basically redoing this whole phone. Lol.
Yeah I am slowly getting there with the device files. I will be updating the post each time I push back up to github and what works and doesn't.
I am stopping work on this with the gingerbread branch and now going to use the new gb-branch-7.2 branch as it was finally released on June 16th I forgot to check back.
http://www.cyanogenmod.com/category/blog
I will be updating the WX435 github during the day.
Edit: Update is done!
I believe this is the best way to go about doing this! If we can get by on base cm7 code then staying updated with the latest from cm7 shouldn't be a problem. Also it in therory should make cm9/ aokp etc easier to iron out all the bugs!
Whenever I get my laptop set back up to build, I'm going to work on picking out the best suited branch to start with from Code Aurora and try to get a vanilla froyo kernel based off the code Aurora sources working. Then from there try and apply those changes to the newer kernels off code Aurora.
It maybe a good idea to try to contact Motorola and see if they wont give us the branch and possibly even commit tag where they forked there kernel from cause I'm pretty sure they started making cchanges to a code Aurora branch on there kernel source. Just which one and where they forked it from is the question. That way we can apply the Motorola source over top that and get the exact changes moto made to the kernel!
If anyone else has anything to add to this or any insights pleas let us know!!
Sent from my Triumph using Tapatalk
Hey g60 if you cab get some logcats on some of the stuff that ain't working right ill look at em and see what I can find!
Sent from my Triumph using Tapatalk
b_randon14 said:
Hey g60 if you cab get some logcats on some of the stuff that ain't working right ill look at em and see what I can find!
Sent from my Triumph using Tapatalk
Click to expand...
Click to collapse
The stuff that is not working has not been added into the device files yet. There is a lot of missing code and lib files. Once I implement more of the changes and stuff isn't working I will holler at you, I have been really focused on fixing the video drivers. I want them playing all formats as perfect as possible before I start adding more stuff. :good:
Hey on the default.xml under the github remote you just have .. so it causes problems when trying to repo sync. Didn't know if you knew it was like that or not!
It should work fine, I just did a repo sync. Make sure you are running the right branch as CM7 changed it from "gingerbread"
repo init -u git://github.com/WX435/android.git -b gb-release-7.2
I just went back in and deleted my folder, then ran repo init again, copying directly from here and it still did the same thing when I tried to run repo sync.
Hop on IRC
b_randon14 said:
I just went back in and deleted my folder, then ran repo init again, copying directly from here and it still did the same thing when I tried to run repo sync.
Click to expand...
Click to collapse
I just looked at it myself. Under remote it shows "..", shouldn't it be git://github.com/
dsmryder said:
I just looked at it myself. Under remote it shows "..", shouldn't it be git://github.com/
Click to expand...
Click to collapse
Nope this is correct it works fine. There are changes CM Team did.
https://github.com/CyanogenMod/android/blob/gb-release-7.2/default.xml
g60madman said:
Nope this is correct it works fine. There are changes CM Team did.
https://github.com/CyanogenMod/android/blob/gb-release-7.2/default.xml
Click to expand...
Click to collapse
So with that, would you have to setup all of your own remotes?
I thought that was to have a Pre-entry for "git remote"?
If you're having issues with the .. in the manifest, see the comments on this commit from cm/android/ where this was originally introduced. The solution is in there:
https://github.com/CyanogenMod/android/commit/89acec784fd50305cc55d05ecb3416fcd7c3eb0e
Yeah i fixed mine by deleting the .repo folder, curling the latest repo into myh ~/bin folder, then pulling the repo in again.
Man I was surprised. I finally got my cm7-bROM repo setup on my new Linux mint install and finally got everything updated from upstream and added some more open commits from gerrit and built it and to my surprise, it built fine with no errors a d I flashed it and it worked great! I figured it would blow up where it had been over a month or two since I had synced and built my ROM and with all the 7.2 release stuff I thought for sure it wouldn't build and add I was building on Linux mind that I had never built on!
Hey g60 you need any help with the device files? I noticed you ain't even got the triumph device repo in the manifest anymore. If you need me to help with anything my box is up and running great! I built my rom in like 47 minutes first time and I was using ccache so hopefully the next build will be faster!
I am digging into my stock code Aurora kernel nor booting. I thought it built fine but I noticed there were some warning spit out by the linker during the build and I tried with linaro and the included 4.4.3 with aosp and it did the sane thing so I'm gonna try to get the bottom of those warnings because I think that may have something to do with it!
Ill post the pastebin of the warning up in the kernel thread later!
Sent from my Triumph using Tapatalk
No updates for a while, still nothing works?
I have been really busy and have had maybe had 1-2 hours to work on the device files in the last 6 days. I should have some more time this week so will see what I can get out.
headset does not register on boot if plugged in
Click to expand...
Click to collapse
Pretty sure this bug was in all CM7 builds before. I also experiences the same thing with MIUI rom.

[ROM][4 May 2013] CM7 for HTC Merge/Lexikon [Kernel: msm7x30 CM]

I originally released this rom on 2fastforums.com (http://2fastroms.com/showthread.php?5435-WIP-18-Dec-2012-CM7-Kernel-CM-msm7x30) and now feel it is ready for release on XDA. (Site is currently down for unknown reasons)
CM7 for HTC Merge/Lexikon
Click to expand...
Click to collapse
Code:
Features:
-100% stock CM7 from source
What works:
Almost everything
Click to expand...
Click to collapse
What doesn't work:
MMS on USC?
Jittery Compass
Click to expand...
Click to collapse
Instructions
1) Flash recovery through bootloader or fastboot
2) Boot into recovery and factory reset, also go into mounts and format system
3) Flash Rom
4) Flash GApps
Click to expand...
Click to collapse
Downloads
Recovery: http://d-h.st/users/e334/?fld_id=6766 (Either one will work, I personally prefer v5.0.2.8 Capacitive Touch)
ROM: http://d-h.st/users/e334/?fld_id=6727
GApps: https://www.box.com/s/79bpq0ywdxcf4093t926 (thanks liamf1!)
Click to expand...
Click to collapse
Source
Kernel: https://github.com/jianC/htc-kernel-msm7x30
Lexikon Device Tree: https://github.com/jianC/android_device_htc_lexikon
Lexikon Proprietary Vendor Tree: https://github.com/jianC/android_vendor_htc
Click to expand...
Click to collapse
All my changes are documented on github.
Credits: Cyberblade (CDMA fix), liamf1 (beta tester/helpful feedback) 2fast4u88, Cyanogenmod & Team, Kali-, Neopeek & Team, UnitedDev & Team, and everyone else who contributed.
Please leave any feedback about bugs/problems/glitches/etc and I will try to fix them.
Thank you ;D
Sent from my HTC Merge using Tapatalk 2
faille13 said:
Thank you ;D
Sent from my HTC Merge using Tapatalk 2
Click to expand...
Click to collapse
Nice to see you on here too
Lol yeah, I'm everywhere.
Sent from my HTC Merge using Tapatalk 2
---------- Post added at 01:34 PM ---------- Previous post was at 01:12 PM ----------
I don't get around here too often. I haven't seen a dedicated forum for my merge yet.
Sent from my HTC Merge using Tapatalk 2
GPS problem
How to fix the GPS ?
It won't turn on , I mean GPS icon is not shown in the notifications bar and it searches for the satellites forever .
I use 20130121 version of this mod and gsm mode .
Gytixas said:
How to fix the GPS ?
It won't turn on , I mean GPS icon is not shown in the notifications bar and it searches for the satellites forever .
I use 20130121 version of this mod and gsm mode .
Click to expand...
Click to collapse
Hmm.. i'll look into it.
I am so happy that new version established but i GSM doesn't work. Please fix the bug
Thanks
Mistake
sorry sorry, GSM works perfect, Just needed to reboot. Thanks a lot. but Bluetooth and GPS doesn't work either. Plz fix these bugs ! Thanks
nikoloz995 said:
I am so happy that new version established but i GSM doesn't work. Please fix the bug
Thanks
Click to expand...
Click to collapse
+1
development of the firmware is?
Thank you so much, this is the best rom for this phone (the only working cyanogen). I am happy to report that after a few weeks use it has across the board improved my phones responsiveness, battery life, and usability. However... I have one problem. I live in Alaska and have GCI as a carrier... when I alter my cdma apn info, everything works up to the point I want to change my MCC and MNC... it's defaulting to 310 and 012 respectively. When I leave those values alone it saves the apn... when I alter them, the apn entry disappears! How can I fix this? I I need it to be 311 and 370 but can't get it to save my apn.
If this never works, I don't care, I wil continue using this rom without mobile internet or mms downloads... it's just that good.
lexikonlover said:
Thank you so much, this is the best rom for this phone (the only working cyanogen). I am happy to report that after a few weeks use it has across the board improved my phones responsiveness, battery life, and usability. However... I have one problem. I live in Alaska and have GCI as a carrier... when I alter my cdma apn info, everything works up to the point I want to change my MCC and MNC... it's defaulting to 310 and 012 respectively. When I leave those values alone it saves the apn... when I alter them, the apn entry disappears! How can I fix this? I I need it to be 311 and 370 but can't get it to save my apn.
If this never works, I don't care, I wil continue using this rom without mobile internet or mms downloads... it's just that good.
Click to expand...
Click to collapse
APN backup and restore app worked for me on US Cellular (backup apn from stock install, then restore to CM7), though I had to switch to GoSMS as a messaging app for MMS to send properly (stock app would receive but not send MMS - SMS worked normally in both). I had no success manually entering the apn info - didn't seem to have any effect when I tried.
@JianC
I found an issue with the 20130121 version from Jan 21 you posted on 2fastroms: If the phone is left on the charger it will charge all the way up, then discharge all the way back to zero while still plugged in. It still displays that it is at 100%, but then abruptly drops to whatever the actual charge level is after a few minutes of use. I've tried wiping battery stats to no effect and it works properly in stock rom. This is the only bug I can't seem to work around... Any fix or suggestions for things to try would be most appreciated! Thanks again for your work on this ROM!
Fixed GPS (and other issues), please see OP!
EDIT: I'll look into the apns issue... I don't use my Merge on CDMA so it's a stab in the dark.
The final version of CM7 has been uploaded (charging bug *should* be fixed).. I will not be supporting CM7 anymore but will be working on Evervolv/CM10.1.
Thanks for your feedback and support!
jianC said:
The final version of CM7 has been uploaded (charging bug *should* be fixed).. I will not be supporting CM7 anymore but will be working on Evervolv/CM10.1.
Thanks for your feedback and support!
Click to expand...
Click to collapse
CM10.1 availible to test?
DJ_MiX said:
CM10.1 availible to test?
Click to expand...
Click to collapse
Nope, he working on kernel .
http://review.evervolv.com
Sent from my HTC Merge using xda premium
If any help with the kernel is needed I will try my best to free up some time and pitch in some help. I have made several kernels for several other devices and the most popular being for the Acer Iconia tablets here on XDA see my signature. Great to see a CM rom for the merge on here though!! GB works fine for me, not that much cool factor from cm 10 or cm 10.1 as I have compiled quiet a few nightlies for the A100.. Other than the Project Butter and the overall UI I could care less about JB/ICS.
linuxsociety said:
If any help with the kernel is needed I will try my best to free up some time and pitch in some help. I have made several kernels for several other devices and the most popular being for the Acer Iconia tablets here on XDA see my signature. Great to see a CM rom for the merge on here though!! GB works fine for me, not that much cool factor from cm 10 or cm 10.1 as I have compiled quiet a few nightlies for the A100.. Other than the Project Butter and the overall UI I could care less about JB/ICS.
Click to expand...
Click to collapse
@linuxsociety That would be awesome if you could. Elginsk8r and I have already ported the board files but I can't figure out what's keeping it from booting..
@ everyone: CM9 for the htc merge is here!: http://forum.xda-developers.com/showthread.php?p=44487058#post44487058
jianC said:
@linuxsociety That would be awesome if you could. Elginsk8r and I have already ported the board files but I can't figure out what's keeping it from booting..
Click to expand...
Click to collapse
Does it just boot loop or does it go black? does the CM logo appear? Guess there is no logcat output via adb either or you'd probably have nailed it by now. Maybe we will have to get jtag debugging output so we can see exactly what is causing the kernel panic and work from there.
I have just now had time to get on xda and check up on the new status. I will try to clone the git repository and check the kernel .config for the cm10 build. Just a thought but on most devices that have a working kernel with cm9 the exact same kernel + modules work fine with CM10. Have you all tried the kernel from CM9 that was just recently posted on any CM10 builds for the device? Not much has changed between ICS and JB that are dependent on in kernel features or that require bleeding edge mainline kernel and/or proprietary kernel patches. However there is a good possibility that the issue is more to do with the bootloader that is required to chainload the android OS for JB causing CM10 not to boot properly. So updating the bootloader may be necessary for JB/CM10 to boot properly. Nevertheless we need some debugging output in order to pinpoint what exactly is the problem going on. I am going to flash cm9 to my merge now and see how everything goes. I may experiment with the cm9 kernel and a cm10 build if I can find one for the merge. My hardware I have right now would take quiet some time to build JB as I haven't even used this old c2d laptop to build any android roms yet lol. Sorry for such a long posting but hopefully it is of some help in finding the resolution to the problems we have.
Thanks guys!
linuxsociety said:
Does it just boot loop or does it go black? does the CM logo appear? Guess there is no logcat output via adb either or you'd probably have nailed it by now. Maybe we will have to get jtag debugging output so we can see exactly what is causing the kernel panic and work from there.
I have just now had time to get on xda and check up on the new status. I will try to clone the git repository and check the kernel .config for the cm10 build. Just a thought but on most devices that have a working kernel with cm9 the exact same kernel + modules work fine with CM10. Have you all tried the kernel from CM9 that was just recently posted on any CM10 builds for the device? Not much has changed between ICS and JB that are dependent on in kernel features or that require bleeding edge mainline kernel and/or proprietary kernel patches. However there is a good possibility that the issue is more to do with the bootloader that is required to chainload the android OS for JB causing CM10 not to boot properly. So updating the bootloader may be necessary for JB/CM10 to boot properly. Nevertheless we need some debugging output in order to pinpoint what exactly is the problem going on. I am going to flash cm9 to my merge now and see how everything goes. I may experiment with the cm9 kernel and a cm10 build if I can find one for the merge. My hardware I have right now would take quiet some time to build JB as I haven't even used this old c2d laptop to build any android roms yet lol. Sorry for such a long posting but hopefully it is of some help in finding the resolution to the problems we have.
Thanks guys!
Click to expand...
Click to collapse
The .35 kernel works fine for CM10.0 but I'm having some issues booting cm10.1 with the .35 kernel because of some graphics memory allocation issues. The main issue is that there are a few issues that can't be fully fixed without the 3.0 kernel such as the camera/camcorder, ION support, etc (which is WIP here: http://review.evervolv.com/#/c/6609/ ). I have already tried probing for the UART TX/RX pins on the lexikon motherboard with a bus pirate but I have not found anything to date. I ordered a micro usb breakout board yesterday for USB UART debugging and am going to pick up some resistors from radioshack so we can hopefully get some progress on the 3.0 kernel because it won't boot far enough to adb or dmesg for our purposes. The problem isn't CM10 right now, but the kernel, which would require jtag or uart debugging, as you have mentioned. I hope that we can get the 3.0 kernel booting soon so we can have cm9/10 fully working.
jianC said:
The .35 kernel works fine for CM10.0 but I'm having some issues booting cm10.1 with the .35 kernel because of some graphics memory allocation issues. The main issue is that there are a few issues that can't be fully fixed without the 3.0 kernel such as the camera/camcorder, ION support, etc (which is WIP here: http://review.evervolv.com/#/c/6609/ ). I have already tried probing for the UART TX/RX pins on the lexikon motherboard with a bus pirate but I have not found anything to date. I ordered a micro usb breakout board yesterday for USB UART debugging and am going to pick up some resistors from radioshack so we can hopefully get some progress on the 3.0 kernel because it won't boot far enough to adb or dmesg for our purposes. The problem isn't CM10 right now, but the kernel, which would require jtag or uart debugging, as you have mentioned. I hope that we can get the 3.0 kernel booting soon so we can have cm9/10 fully working.
Click to expand...
Click to collapse
This sounds about like the exact same issues that was experienced when I was helping hashcode work on the original kindle fire device and getting it to work with ICS. I 'think' with the HTC it shouldn't be near as hard to resolve the issue to get full hw acceleration, memory mappings camera/gps and all the extra bells and whistles if we make an attempt at backporting the necessary support into the .35 kernel instead of trying to upgrade * to 3.x which will be adding to the work and break more things than it fixes. Since HTC apparently isn't going to be releasing any updates for the device it looks like it is going to be left all up to us to make one of two decisions:
1) backporting kernel - only the few things remaining needed for our current working GB kernel to have full support for ICS/JB roms
OR
2) updating to 3.x.x and merging HTC fixes and 3rd party patches -through several mainline kernel patches and keeping the proprietary htc addins compatible.
My opinion/suggestion:
I'd think of the two options that backporting would be the most logical answer to the equation since we are only missing support for few things and you already have everything working except the camera, ION, BINDER(?), and other >=3.0.0 android patches. Plus the device is pretty well coming to a good place to say cease development work and either enjoy the device or upgrade to something newer. So if we could get JB support or even just ICS working fully I think 99.99% of all owners/users would be more than 99.99% happy!! I know I would be very happy myself!

Categories

Resources