I thought it does boot?
BTW I have the WiFi-only version, so this 3G version will probably brick my Tab...
Sent from my GT-P5210 using Tapatalk
[Development][Source]CyanogenMod 11 for SGT-3 10.1
- Thread starter Angel_666
- Start date
I thought it does boot?
BTW I have the WiFi-only version, so this 3G version will probably brick my Tab...
Sent from my GT-P5210 using Tapatalk
Ok, boot and work properly
Sent from my SPH-L720 using XDA Premium 4 mobile app
---------- Post added at 10:24 AM ---------- Previous post was at 10:24 AM ----------
No this tab hasn't been updated yet
Sent from my SPH-L720 using XDA Premium 4 mobile app
Sent from my SPH-L720 using XDA Premium 4 mobile app
---------- Post added at 10:24 AM ---------- Previous post was at 10:24 AM ----------
No this tab hasn't been updated yet
Sent from my SPH-L720 using XDA Premium 4 mobile app
Let's wait. If no updates until july - i start CM11 anyway. Just to clear the things - if Sammy realise update, it will be very easy to port CM. If not - it will be harder, but we'll do it anyway, right?
M
moonbutt74
Guest
patches
on debian/kde if the diff/patch is correct it should open in kompare automatically /else start kompare manually
open diff/patch as source and file to be patched as target. if the diff/patch is malformed you will/may have issues
with saving. in console i believe you can diff and when done patch but i haven't gone that route yet. OR if probs with
kompare saving because of malformed diff look at the path to files and work back and forth between two instances
of text editor using kompare as your guide.
m
on debian/kde if the diff/patch is correct it should open in kompare automatically /else start kompare manually
open diff/patch as source and file to be patched as target. if the diff/patch is malformed you will/may have issues
with saving. in console i believe you can diff and when done patch but i haven't gone that route yet. OR if probs with
kompare saving because of malformed diff look at the path to files and work back and forth between two instances
of text editor using kompare as your guide.
m
For those with a 5210, if you go back a few pages I have a modded boot image posted to allow this to boot on the wifi version . just saying
It's on this thread, just look back a bit.
It's on this thread, just look back a bit.
M
moonbutt74
Guest
bear with me
@Angel_666 @Restl3ss
so this will sound really amateur but hear goes
at least on gt-p5210 rfkill list all only returns bluetooth entry,
0: bcm Bluetooth: bluetooth
Softblocked:yes [no after rfkill unblock bluetooth] wrong or missing library?
Hard blocked:no [no change from rfkill unblock]
whereas rfkill list all should return
0: bcm Bluetooth: bluetooth
Softblocked:yes [no after rfkill unblock bluetooth]
Hard blocked:no
1: phy0: wlan
soft blocked:no [if you keep wifi on like i do]
hard blocked:no
2: brcmfmac-wifi: wlan
soft blocked:no
hard blocked:no
now, i had asked you [angel] if you had all apprpriate libraries in dev-tree. maybe recheck
there is also an open source brcmfmac/bcm4334 driver that can be included in kernel build if i read it right.
i have to find my way back to the page though
ok onto storage for gt-p5210 anyway [no extsdcard!]
i think this might connect to some issues in your {angel} twrp build as well .
in /dev/block/vold there is nothing.......[crickets...]
there should be something like 179:50 @ 50bytes and 179:51 @ about the same.
typing vold should unmount card maybe see if and what error message you get.
( i believe this is a kernel level issue)
also root is not quite right , after a fight ultimately manually pushing and setting perms [6775-for my purposes , may need recheck daemon setting in .rc ]on su binary through adb
some things started working more correctly.
lastly since i couldn't do a dirty flash, i did a dirty restore instead, which unlocked the device from airplane mode
which suggests to me anyway the possibility for a lot of profitable hackery in /data to maybe force/trick/correct/better diagnose
fixes etc. i hope this was helpful, in spite of version diffs i'm going to try some cross-rom surgery frankenstein style
thanks for reading
m
@Angel_666 @Restl3ss
so this will sound really amateur but hear goes
at least on gt-p5210 rfkill list all only returns bluetooth entry,
0: bcm Bluetooth: bluetooth
Softblocked:yes [no after rfkill unblock bluetooth] wrong or missing library?
Hard blocked:no [no change from rfkill unblock]
whereas rfkill list all should return
0: bcm Bluetooth: bluetooth
Softblocked:yes [no after rfkill unblock bluetooth]
Hard blocked:no
1: phy0: wlan
soft blocked:no [if you keep wifi on like i do]
hard blocked:no
2: brcmfmac-wifi: wlan
soft blocked:no
hard blocked:no
now, i had asked you [angel] if you had all apprpriate libraries in dev-tree. maybe recheck
there is also an open source brcmfmac/bcm4334 driver that can be included in kernel build if i read it right.
i have to find my way back to the page though
ok onto storage for gt-p5210 anyway [no extsdcard!]
i think this might connect to some issues in your {angel} twrp build as well .
in /dev/block/vold there is nothing.......[crickets...]
there should be something like 179:50 @ 50bytes and 179:51 @ about the same.
typing vold should unmount card maybe see if and what error message you get.
( i believe this is a kernel level issue)
also root is not quite right , after a fight ultimately manually pushing and setting perms [6775-for my purposes , may need recheck daemon setting in .rc ]on su binary through adb
some things started working more correctly.
lastly since i couldn't do a dirty flash, i did a dirty restore instead, which unlocked the device from airplane mode
which suggests to me anyway the possibility for a lot of profitable hackery in /data to maybe force/trick/correct/better diagnose
fixes etc. i hope this was helpful, in spite of version diffs i'm going to try some cross-rom surgery frankenstein style
thanks for reading
m
Last edited:
M
moonbutt74
Guest
more - from logcat
concerning Not working for now:
Wi-Fi (can not obtain mac)
logcat - I/macloader(4533):WIFI MAC is valid [?]
rkill switches maybe?
only other thing in logcat was E/DataRouter(1911)pening of the USB File failed fd is -1 & errno is 13
note- what i din't see in logcat were any bluetooth/wifi Errors
from dmesg
init] init: starting 'macloader'
init] init: waitpid returned pid 12517, status = 00000000
init] init: process 'macloader', pid 12517 exited
init] init: process 'macloader' killing any children in process group - why return 00000000 ?
does macloader have uid gid settings?
dhd.ko does not insmod , returns no such device
concerning Not working for now:
Wi-Fi (can not obtain mac)
logcat - I/macloader(4533):WIFI MAC is valid [?]
rkill switches maybe?
only other thing in logcat was E/DataRouter(1911)
pening of the USB File failed fd is -1 & errno is 13note- what i din't see in logcat were any bluetooth/wifi Errors
from dmesg
init] init: starting 'macloader'
init] init: waitpid returned pid 12517, status = 00000000
init] init: process 'macloader', pid 12517 exited
init] init: process 'macloader' killing any children in process group - why return 00000000 ?
does macloader have uid gid settings?
dhd.ko does not insmod , returns no such device
M
moonbutt74
Guest
what we need
@Angel_666
if there's a chance of sorting this out, we need the kernel source you used and hopefully still have.
if the kernel was plucked from another device and dropped in for prebuilt then we need to know the device.
this thread could also use a cleaning. i don't think we should wait for samsung to spit out something
that would probably be suspect at best.
seriously.
m
@Angel_666
if there's a chance of sorting this out, we need the kernel source you used and hopefully still have.
if the kernel was plucked from another device and dropped in for prebuilt then we need to know the device.
this thread could also use a cleaning. i don't think we should wait for samsung to spit out something
that would probably be suspect at best.
seriously.
m
kernel source is in free access on Samsung Opensorce site http://opensource.samsung.com/reception/receptionSub.do?method=sub&sub=F&searchValue=p5200
M
moonbutt74
Guest
angel,
i have the source. what modifications did you make to bring it up to 4.4 from 4.2?
when i build bootimage it remains 4.2.
to be clear, i want your source after modifications that you used to build the kernel that you included into
your device tree , bzImage,vmlinux.
m
Last edited:
I just add new KitKat security method support (present in kernel, just check the boxes in kernel config).
bootimage - it's a kernel with RamFS. What do you mean by " bootimage remains 4.2"?
M
moonbutt74
Guest
ahh ok
angel,
sorry, i should have been clearer. when i use my prebuilt kernel from my source i output a kernel that boots android 4.2.
when i use the prebuilt kernel you included in your device tree it boot android 4.4.
what modifications in your kernel source did you make?
m
angel,
sorry, i should have been clearer. when i use my prebuilt kernel from my source i output a kernel that boots android 4.2.
when i use the prebuilt kernel you included in your device tree it boot android 4.4.
what modifications in your kernel source did you make?
m
In other words - i turn on SElinux support in kernel. All other modifications we need to do in init***.rc (at ramfs). Check my init.common.rc in git sources. You need to change "BOOTCLASSPATH" parameter.
M
moonbutt74
Guest
ok wait
so the only difference between the kernel for 4.2 and 4.4 is selinux and init***.rc modifications?
if so when i swapped my zimage replacing yours your rom should have booted but, it didn't.
meaning i build my kernel just to get zimage and then replace yours with mine and
no boot. i'm using your ramfs unchanged.
m
so the only difference between the kernel for 4.2 and 4.4 is selinux and init***.rc modifications?
if so when i swapped my zimage replacing yours your rom should have booted but, it didn't.
meaning i build my kernel just to get zimage and then replace yours with mine and
no boot. i'm using your ramfs unchanged.
m
Last edited:
You device is 5200? If so, just turn on SeLinux support in kernel AND replace my kernel modules by your's.
M
moonbutt74
Guest
okay
angel,
okay, found selinux enable, question is, yes to evrything across the board or just leave it enabled with default selections?
m
and i have a p5210, now the only change i'm going to make is enabling selinux but, will that require me to enable force loading of modules?
angel,
okay, found selinux enable, question is, yes to evrything across the board or just leave it enabled with default selections?
m
and i have a p5210, now the only change i'm going to make is enabling selinux but, will that require me to enable force loading of modules?
You need you own builded modules to let system work correctly. As for selinux - i don't remember for now (sources on another PC), but try to leave default variant... Or search on xda about kernel config for kitkat.
M
moonbutt74
Guest
ok
angel,
so first let me say, I [bleeping] despise selinux. that's ridiculous...anyway, thank you, kernel built/booted.... Joy? no......
bluetooth is in part built into the kernel as opposed to module?
and wifi/bluetooth both require broadcom drivers/libraries?
i say/ask that because with my kernel the result is still the same with macloader/rfkill list all bringing up only bluetooth at 0/ and dhd.ko insmod error with no such device. so if it not in the kernel it's in the rom and meaning something missing.?
most of the above are questions. a while back i found an open source bcm4334 driver that can be built in as opposed to what i guess
is the module/library setup if i understood that right. gonna try that. that still doesn't help me to understand why macloader registers wifi address as valid and then dies. if i find anything i will post. selinux sucks.
also external storage still doesn't register, though cat /proc/partitions brings it up, it lives in /dev/block
but won't mount, not through terminal or file browser. edit-after alot of reading, kitkat has been intentionally sleazed up by google to break external storage. the nexus should have been a clue.
also there are major vold/fstab issues that need to be sorted. cyanogenmod should be a little more concerned with getting their **** right than throwing up their colors.
adb runs fine and root shell is go. selinux sucks.
m
(selinux sucks) shouldn't bmgr.jar and com.broadcom.bt.jar be in classpath? edit-did that no dice.
angel,
so first let me say, I [bleeping] despise selinux. that's ridiculous...anyway, thank you, kernel built/booted.... Joy? no......
bluetooth is in part built into the kernel as opposed to module?
and wifi/bluetooth both require broadcom drivers/libraries?
i say/ask that because with my kernel the result is still the same with macloader/rfkill list all bringing up only bluetooth at 0/ and dhd.ko insmod error with no such device. so if it not in the kernel it's in the rom and meaning something missing.?
most of the above are questions. a while back i found an open source bcm4334 driver that can be built in as opposed to what i guess
is the module/library setup if i understood that right. gonna try that. that still doesn't help me to understand why macloader registers wifi address as valid and then dies. if i find anything i will post. selinux sucks.
also external storage still doesn't register, though cat /proc/partitions brings it up, it lives in /dev/block
but won't mount, not through terminal or file browser. edit-after alot of reading, kitkat has been intentionally sleazed up by google to break external storage. the nexus should have been a clue.
also there are major vold/fstab issues that need to be sorted. cyanogenmod should be a little more concerned with getting their **** right than throwing up their colors.
adb runs fine and root shell is go. selinux sucks.
m
(selinux sucks) shouldn't bmgr.jar and com.broadcom.bt.jar be in classpath? edit-did that no dice.
Last edited:
Similar threads
Top Liked Posts
-
There are no posts matching your filters.
-
52This is development thread for CM11 building on Samsung GalaxyTab3 10.1 devices, based on Intel Clover Trail platform. (x86)
Device tree: https://github.com/Angel666/android_device_samsung_santos103g
CM11 source repo:Patches for CM source needed to build rom. Made by HazouPH:Code:[FONT=Arial]repo init -u git://github.com/CyanogenMod/android.git -b cm-11.0[/FONT]
https://github.com/HazouPH/android_device_motorola_smi/tree/jellybean/patch
Working for now:
hardware accel (with some bugs)
storages and mtp mounting
Ril (partially)
Lights and buttons
Houdini service
anything else I can forgot..
Not working for now:
Wi-Fi (can not obtain mac)
BT (can not obtain mac)
Camera (service works, app can't connect)
SeLinux security (disabled)
Sound (Samsung lib won't work)
Basic CM Launcher3 is glitchy! I don't know why. Third-party launcers works well (testing by GOlauncher)
anything else I can forgot..
Images for testing: >DOWNLOAD< 3g version only for now!
WARNING! This images are for testing purposes only! Don't flash it if you don't know what you doing! Rom is almost UNUSABLE for enduser! Work in progress....
0.Unpack zip's.
1. Flash boot by Odin
2. Flash recovery by Odin
3. Enter to recovery and Wipe cache, dalvik, system and data.
4 Flash system by Odin
I appreciate any possible help (testing, donations, devs and experienced users advices). Let's speedup the process!
18
Yes, after fixes and testing, i will create device tree for wi-fi version too.12
Let's wait. If no updates until july - i start CM11 anyway. Just to clear the things - if Sammy realise update, it will be very easy to port CM. If not - it will be harder, but we'll do it anyway, right?
