[CM7][GP3.6][LOGCAT] CM7 porting - Need some help - Samsung Galaxy Player 4.0, 5.0

Hello XDA community for the GPs,
some of you may heard about my project to port CyanogenMod 7.2 to the Galaxy Player 3.6.
The thing is.. I got it successfully compiled and installed on my player but it won't boot (it is stuck at zygote).
So I thin here are some really good CM porters and experts which could may help ... so...
Please be a community and help me a bit
The poor thing is that I'm the only dev for this device.
Here are some files...
If you need more information or files.. LET ME KNOW!
Here's the init.rc which I'm using to boot (I used the stock one as a base to write a new init.rc)
http://pastebin.com/nbkubMyx
That's the init.<devicename>.rc (devicename: aalto) (I fixed some things on the kernel so it gets executed)
http://pastebin.com/g0FxQUja
That's the logcat which gets made via init.rc (service: logcat)
http://pastebin.com/bN4yfPiq
Here's the logcat which gets made via ADB
https://www.dropbox.com/s/j1zbganc8whowly/log
Device tree
https://github.com/androthan/android_device_samsung_aalto
Thanks for any help in advance

If the 3.6 is anything like 5.0, the stock gingerbread had much more in it's initramfs than CM puts in there. Alot of the paths in the stock init.rc and init.(device).rc may point to locations expected to be there in the initramfs which cm has moved to the /system dir. You might want to try looking at the init.rc and init.(device).rc for a device that has a similar partition layout( and CM7 working). Maybe use that as a base instead of stock init.rc???
I've never built CM7. I'm actually pretty new to building CM but it looks like you've got mount point problems....
Hope this helps

Meticulus said:
If the 3.6 is anything like 5.0, the stock gingerbread had much more in it's initramfs than CM puts in there. Alot of the paths in the stock init.rc and init.(device).rc may point to locations expected to be there in the initramfs which cm has moved to the /system dir. You might want to try looking at the init.rc and init.(device).rc for a device that has a similar partition layout( and CM7 working). Maybe use that as a base instead of stock init.rc???
I've never built CM7. I'm actually pretty new to building CM but it looks like you've got mount point problems....
Hope this helps
Click to expand...
Click to collapse
Yeah that's what I've seen. Though here's a beginning for what you can do @Androthan
1. Make a cleaner vold.fstab (yours look like stock samsung crap, even vold can't read it as per your logcat says ), you can base yourself here: https://github.com/dhiru1602/android_device_samsung_galaxysl/blob/gingerbread/etc/vold.fstab
Also you can remove that sdcard mounting in your init.rc as it is already supposed to be mounted with vold.
2. Looks like from the log that your device can't access dalvik cache needed for the apps. Here's a commit I've found that will replace the path location from /cache to /dbdata. Though perhaps it would be better if it could be /cache. For this I would need to check dmesg as well to see if something haven't broke during the process. Here's the commit for the path change in case you need it:
https://github.com/dhiru1602/android_device_samsung_galaxysl/commit/5c2a2a54541448565a8d2418cfd75a1a4882342d
My guess would be that maybe cache doesn't get mounted right but it's not only logcat that can tell me this.
Hope it helps!
EDIT: Looks like it's better to use /data for dalvik-cache instead of /cache then make a symlink to it as per the commit says...

This commit shows how mounting was done in for the 5.0 in CM7 : https://github.com/ambrice/android_...mmit/891c1f59058d4d7e082d44ca40f9fcf3addd6188 . Not exactly sure how far @ambrice got with CM7 though...
I know that the partition table is different for 3.6 but I thought it might give you an idea of how it's should be.
EDIT: I should mention that in the 5.0, we try to mount ext4 first but we fail back to rfs if we can't mount ext4....
Another thing I noticed is that you are trying to mount mmcblk0p1. I don't know how it is on the 3.6 but on the 5.0 that partition represents the entire extended partition and I don't think it can actually be mounted....

Meticulus said:
This commit shows how mounting was done in for the 5.0 in CM7 : https://github.com/ambrice/android_...mmit/891c1f59058d4d7e082d44ca40f9fcf3addd6188 . Not exactly sure how far @ambrice got with CM7 though...
I know that the partition table is different for 3.6 but I thought it might give you an idea of how it's should be.
EDIT: I should mention that in the 5.0, we try to mount ext4 first but we fail back to rfs if we can't mount ext4....
Another thing I noticed is that you are trying to mount mmcblk0p1. I don't know how it is on the 3.6 but on the 5.0 that partition represents the entire extended partition and I don't think it can actually be mounted....
Click to expand...
Click to collapse
The mmcblk0p1 is the efs partition which is why it is needed to be mounted as reperesented here:
https://github.com/zaclimon/initramfs-ypgs1/blob/master/init.aalto.rc#L68
However he looks like he doesn't put the mounting flags to his partition so that to be done as well. Also those devices mounts as ext4 by default as well.

Meticulus said:
If the 3.6 is anything like 5.0, the stock gingerbread had much more in it's initramfs than CM puts in there. Alot of the paths in the stock init.rc and init.(device).rc may point to locations expected to be there in the initramfs which cm has moved to the /system dir. You might want to try looking at the init.rc and init.(device).rc for a device that has a similar partition layout( and CM7 working). Maybe use that as a base instead of stock init.rc???
I've never built CM7. I'm actually pretty new to building CM but it looks like you've got mount point problems....
Hope this helps
Click to expand...
Click to collapse
zaclimon said:
Yeah that's what I've seen. Though here's a beginning for what you can do @Androthan
1. Make a cleaner vold.fstab (yours look like stock samsung crap, even vold can't read it as per your logcat says ), you can base yourself here: https://github.com/dhiru1602/android_device_samsung_galaxysl/blob/gingerbread/etc/vold.fstab
Also you can remove that sdcard mounting in your init.rc as it is already supposed to be mounted with vold.
2. Looks like from the log that your device can't access dalvik cache needed for the apps. Here's a commit I've found that will replace the path location from /cache to /dbdata. Though perhaps it would be better if it could be /cache. For this I would need to check dmesg as well to see if something haven't broke during the process. Here's the commit for the path change in case you need it:
https://github.com/dhiru1602/androi...mmit/5c2a2a54541448565a8d2418cfd75a1a4882342d
My guess would be that maybe cache doesn't get mounted right but it's not only logcat that can tell me this.
Hope it helps!
EDIT: Looks like it's better to use /data for dalvik-cache instead of /cache then make a symlink to it as per the commit says...
Click to expand...
Click to collapse
Thank you both for answering my thread
I used the init.rc from <cmsource>/system/core/rootdir and added the things from the stock init.rc. I also cleaned up init.aalto.rc and modified it to work with CM.
I already tried to symlink /cache/dalvik-cache and /data/dalvik-cache
but this doesn't fixed my issue that it can't create the DEX caches.
I think I'll have to fix...
... the vold.fstab (like you said)
... Add some folders to ramdisk
... libaudio.so (libaudio doesn't get located but its needed to start the system server
... Figure why it doesn't mount cache
... Figurebwhy dbus can't find etc/dbus.conf (even when /system/etc is symlinked to /etc)
BTW, zaclimon you said you need a 'dmesg'. What is it and how can I get it?
Thanks, Androthan
Sent from my Nexus 7 using xda app-developers app

Androthan said:
Thank you both for answering my thread
I used the init.rc from <cmsource>/system/core/rootdir and added the things from the stock init.rc. I also cleaned up init.aalto.rc and modified it to work with CM.
I already tried to symlink /cache/dalvik-cache and /data/dalvik-cache
but this doesn't fixed my issue that it can't create the DEX caches.
I think I'll have to fix...
... the vold.fstab (like you said)
... Add some folders to ramdisk
... libaudio.so (libaudio doesn't get located but its needed to start the system server
... Figure why it doesn't mount cache
... Figurebwhy dbus can't find etc/dbus.conf (even when /system/etc is symlinked to /etc)
BTW, zaclimon you said you need a 'dmesg'. What is it and how can I get it?
Thanks, Androthan
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
Thing is the approach for the dalvik cache is different from the one you maybe made. This one implies that you make a dalvik-cache folder inside /dbdata which is in /data, then symlink it to make it /cache/dalvik-cache as well. For the moment these are the most important. Also as for cache, I didn't say that it won't mount it, I said that maybe it won't mount it because of some issues unless the dalvik-cache isn't made here.
Also the dmesg is the kernel log (it's basically a logcat but from a kernel side instead, here you can see how the kernel make it's initialization. You can have it either with this:
adb shell dmesg > dmesg.log
A proc/last_ksmg is a dmesg too but from the previous boot.
It's better to do one at a time then fix the other things after (this way you could get other problems than the one you have already)

Androthan said:
Thank you both for answering my thread
I used the init.rc from <cmsource>/system/core/rootdir and added the things from the stock init.rc. I also cleaned up init.aalto.rc and modified it to work with CM.
I already tried to symlink /cache/dalvik-cache and /data/dalvik-cache
but this doesn't fixed my issue that it can't create the DEX caches.
I think I'll have to fix...
... the vold.fstab (like you said)
... Add some folders to ramdisk
... libaudio.so (libaudio doesn't get located but its needed to start the system server
... Figure why it doesn't mount cache
... Figurebwhy dbus can't find etc/dbus.conf (even when /system/etc is symlinked to /etc)
BTW, zaclimon you said you need a 'dmesg'. What is it and how can I get it?
Thanks, Androthan
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
Here is the dmesg man:
http://unixhelp.ed.ac.uk/CGI/man-cgi?dmesg+8
Normally you could get a dmesg by being root and typing dmesg > nameoffile.log but you'd need to be in a directory that you can write to for that. You can type dmesg as root in the shell and it will output to the screen...
I normally just
adb root
adb shell
dmesg
EDIT: Sorry @zaclimon, I didn't think you were going to answer

Meticulus said:
Here is the dmesg man:
http://unixhelp.ed.ac.uk/CGI/man-cgi?dmesg+8
Normally you could get a dmesg by being root and typing dmesg > nameoffile.log but you'd need to be in a directory that you can write to for that. You can type dmesg as root in the shell and it will output to the screen...
I normally just
adb root
adb shell
dmesg
EDIT: Sorry @zaclimon, I didn't think you were going to answer
Click to expand...
Click to collapse
Haha ya well perhaps it wouldn't have worked if he didn't made adb root so I guess this is a team work after all.
Envoyé depuis mon Nexus 7 avec Tapatalk 4

zaclimon said:
Haha ya well perhaps it wouldn't have worked if he didn't made adb root so I guess this is a team work after all.
Envoyé depuis mon Nexus 7 avec Tapatalk 4
Click to expand...
Click to collapse
Meticulus said:
Here is the dmesg man:
http://unixhelp.ed.ac.uk/CGI/man-cgi?dmesg+8
Normally you could get a dmesg by being root and typing dmesg > nameoffile.log but you'd need to be in a directory that you can write to for that. You can type dmesg as root in the shell and it will output to the screen...
I normally just
adb root
adb shell
dmesg
EDIT: Sorry @zaclimon, I didn't think you were going to answer
Click to expand...
Click to collapse
Thanks for the tip and the explanation of dmsg.
I'm now testing a new build with the following changes :
- added lines to init.rc "export ANDROID_CACHE /cache and export DOWNLOD_CACHE /cache/download"
- Added mounting flags (for example: nodev,noexec,wait...)
- symlinked /dbdata/dalvik-cache /cache/dalvik-cache
Another question... Do I have to use the libaudio.so library (and libasound.so) from the device (proprietary way) or can I use the libaudio which gets compiled with CM7 (ALSA sound) ?
Greetings, Androthan
Sent from my Nexus 7 using xda app-developers app

Androthan said:
...
Another question... Do I have to use the libaudio.so library (and libasound.so) from the device (proprietary way) or can I use the libaudio which gets compiled with CM7 (ALSA sound) ?
...
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
I honestly don't know but If I were you, I'd do which ever one worked and if they both worked, then I'd do which ever one worked better...:cyclops:

Androthan said:
Thanks for the tip and the explanation of dmsg.
I'm now testing a new build with the following changes :
- added lines to init.rc "export ANDROID_CACHE /cache and export DOWNLOD_CACHE /cache/download"
- Added mounting flags (for example: nodev,noexec,wait...)
- symlinked /dbdata/dalvik-cache /cache/dalvik-cache
Another question... Do I have to use the libaudio.so library (and libasound.so) from the device (proprietary way) or can I use the libaudio which gets compiled with CM7 (ALSA sound) ?
Greetings, Androthan
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
As meticulus said, I'd probably test which one sound better but it would be better to use the open source variant. Though it would be better as I said earlier to make the thing booting first.
As for the dalvik cache, be sure to make a directory in dbdata before symlink it to /cache.
Envoyé depuis mon Nexus 7 avec Tapatalk 4

Androthan said:
Thank you both for answering my thread
I used the init.rc from <cmsource>/system/core/rootdir and added the things from the stock init.rc. I also cleaned up init.aalto.rc and modified it to work with CM.
I already tried to symlink /cache/dalvik-cache and /data/dalvik-cache
but this doesn't fixed my issue that it can't create the DEX caches.
I think I'll have to fix...
... the vold.fstab (like you said)
... Add some folders to ramdisk
... libaudio.so (libaudio doesn't get located but its needed to start the system server
... Figure why it doesn't mount cache
... Figurebwhy dbus can't find etc/dbus.conf (even when /system/etc is symlinked to /etc)
BTW, zaclimon you said you need a 'dmesg'. What is it and how can I get it?
Thanks, Androthan
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
So, I reworked some things and fixed...
... vold.fstab (rewrote it)
... dbus.conf (deleted etc in the ramdisk)
... dalvik-cache (the rootfilesystem '/' system was mounted as 'read only' so we hadn't the right permissions // symlinked /cache/dalvik-cache and /dbdata/dalvik-cache )
My/Our problem is now the libaudio.so:
It is needed to load the SystemServer.
I'm actually using the stock libaudio files (libaudiopolicy.so, libaudioflinger.so, libasound.so, libsoundpool.so,...) but theres an error in libaudio.so , something doesn't get located.
Some websites are saying that libaudio.so contains the proprietary code (HAL implementation) for the audio device (like libcamera.so for the camera).
How to fix that? Did I missed a proprietary file?
Greetings, Androthan

Androthan said:
So, I reworked some things and fixed...
... vold.fstab (rewrote it)
... dbus.conf (deleted etc in the ramdisk)
... dalvik-cache (the rootfilesystem '/' system was mounted as 'read only' so we hadn't the right permissions // symlinked /cache/dalvik-cache and /dbdata/dalvik-cache )
My/Our problem is now the libaudio.so:
It is needed to load the SystemServer.
I'm actually using the stock libaudio files (libaudiopolicy.so, libaudioflinger.so, libasound.so, libsoundpool.so,...) but theres an error in libaudio.so , something doesn't get located.
Some websites are saying that libaudio.so contains the proprietary code (HAL implementation) for the audio device (like libcamera.so for the camera).
How to fix that? Did I missed a proprietary file?
Greetings, Androthan
Click to expand...
Click to collapse
Logs?

Androthan said:
So, I reworked some things and fixed...
... vold.fstab (rewrote it)
... dbus.conf (deleted etc in the ramdisk)
... dalvik-cache (the rootfilesystem '/' system was mounted as 'read only' so we hadn't the right permissions // symlinked /cache/dalvik-cache and /dbdata/dalvik-cache )
My/Our problem is now the libaudio.so:
It is needed to load the SystemServer.
I'm actually using the stock libaudio files (libaudiopolicy.so, libaudioflinger.so, libasound.so, libsoundpool.so,...) but theres an error in libaudio.so , something doesn't get located.
Some websites are saying that libaudio.so contains the proprietary code (HAL implementation) for the audio device (like libcamera.so for the camera).
How to fix that? Did I missed a proprietary file?
Greetings, Androthan
Click to expand...
Click to collapse
Well as a temporary solution you could use the one found in the stock rom and then copy it to your vendor files so it can be copied into your lib folder. This way it'll work no problem. Also as @Meticulus said, a bunch of logs is always appreciated.
Envoyé depuis mon Nexus 7 avec Tapatalk 4

Related

CWM for Galaxy Player 5.0 US (YP-G70XAA)

Hello all,
I'm working on porting CWM to our beloved Galaxy Player 5.0 US edition. I'm loosely following the guide here. I cannot promise that I'll be able to port it successfully, but I can promise that I'll brick my device at least once.
tl;dr : This post says some guy is working on CWM. Woot!
-apapousek
Your sending it to the people are you? And good luck, I only wish I knew how to do this
Apapouseck-
Are you a dev? PM me and give me your email address please. I have a question.
Sent from my Galaxy Nexus using xda premium
azoller1 said:
Your sending it to the people are you? And good luck, I only wish I knew how to do this
Click to expand...
Click to collapse
Of course I'm distributing it. Once I get a working copy, it's being thrown on github, or my own server, and given freely, open source, blah blah blah. No proprietary crap, except maybe enough drivers to get it up and running. But sources for those, too.
cramjammer said:
Apapouseck-
Are you a dev? PM me and give me your email address please. I have a question.
Sent from my Galaxy Nexus using xda premium
Click to expand...
Click to collapse
Well, technically not a dev. I just like taking software, messing around with it, seeing what it can do. My email address is [email protected]
Can't wait
Sent from my YP-G70 using Xparent Blue Tapatalk
That post, as it says, doesn't cover Samsung devices. It turns out that a while back, Koush made a commit to mkvendor.sh that allows it to operate without an example boot.img.
However, the problem is that there's a remaining bug in mkvendor.sh that TMPDIR is not set in the code path that is triggered when you don't specify boot.img
Right now I'm at a state where CWM starts and runs, but somehow it is treating the tilt sensor as volup/voldn input - so the thing scrolls eternally.
Entropy512 said:
That post, as it says, doesn't cover Samsung devices. It turns out that a while back, Koush made a commit to mkvendor.sh that allows it to operate without an example boot.img.
However, the problem is that there's a remaining bug in mkvendor.sh that TMPDIR is not set in the code path that is triggered when you don't specify boot.img
Right now I'm at a state where CWM starts and runs, but somehow it is treating the tilt sensor as volup/voldn input - so the thing scrolls eternally.
Click to expand...
Click to collapse
Maybe you need the s3c-keypad file to correct this or changing somewhere in the source /dev/input/eventX to another one.
Looks like I need to figure out how to filter out /dev/input/event1, and only pay attention to event4 and event5 (or just event4)
I've got something that seems to be working... Needs a lot more testing. Recovery is important, and if it's broken, can really **** up someone's device.
Also, the way I got it working seems to be a massive hackjob, I need to chat with Koush if there's a better way to do things. Either that or a LOT of CWM porters have done source code modifications without publishing them (Which, while not ideal is allowed for CWM since it is Apache-licensed not GPL.)
Entropy512 said:
I've got something that seems to be working... Needs a lot more testing. Recovery is important, and if it's broken, can really **** up someone's device.
Also, the way I got it working seems to be a massive hackjob, I need to chat with Koush if there's a better way to do things. Either that or a LOT of CWM porters have done source code modifications without publishing them (Which, while not ideal is allowed for CWM since it is Apache-licensed not GPL.)
Click to expand...
Click to collapse
I'm glad you got a clockworkmod which seems to work. Let us know what koush says and if it'll be official support to CWM on our devices.
Yup.
Main things you need to consider when building:
1) The recovery.fstab for Samsung devices isn't appropriate for CWM. One appropriate for CWM has just three entries:
mount point, filesystem, device
The filesystem format options (fourth and fifth columns) seen in stock recovery.fstab will screw up CWM
2) /dev/input/event1 MUST be filtered on SGP5 US devices (and probably international too) - the accelerometer spams keystrokes. This is in ev_init() in minui/events.c
3) Recovery.fstab can't be in /etc - but this path is hardcoded in CWM so you must change the path in roots.c - SGSII seems to use /misc instead of /etc - this is in load_volume_table() in roots.c
Entropy512 said:
Yup.
Main things you need to consider when building:
1) The recovery.fstab for Samsung devices isn't appropriate for CWM. One appropriate for CWM has just three entries:
mount point, filesystem, device
The filesystem format options (fourth and fifth columns) seen in stock recovery.fstab will screw up CWM
2) /dev/input/event1 MUST be filtered on SGP5 US devices (and probably international too) - the accelerometer spams keystrokes. This is in ev_init() in minui/events.c
3) Recovery.fstab can't be in /etc - but this path is hardcoded in CWM so you must change the path in roots.c - SGSII seems to use /misc instead of /etc - this is in load_volume_table() in roots.c
Click to expand...
Click to collapse
Koush put this up at rootzwiki.
apapousek said:
Koush put this up at rootzwiki.
Click to expand...
Click to collapse
I've seen it - if you read it, it's just some "best practices" document that says what you should do after you've managed to compile a working CWM recovery binary - however, I have not seen a single kernel that integrated CWM the way he says, even ones coming from very talented developers (like codeworkx)
Entropy512 said:
I've got something that seems to be working... Needs a lot more testing. Recovery is important, and if it's broken, can really **** up someone's device.
Also, the way I got it working seems to be a massive hackjob, I need to chat with Koush if there's a better way to do things. Either that or a LOT of CWM porters have done source code modifications without publishing them (Which, while not ideal is allowed for CWM since it is Apache-licensed not GPL.)
Click to expand...
Click to collapse
Could you post how you achieved this partial state? I'm busy downloading cm, which is just fun
I'll try to when I have some time... next few days are going to be busy.
Very quick summary:
Need to fix mkvendor.sh so TMPDIR is set when you don't provide a bootimage argument, otherwise the device tree doesn't get created properly
Need to pull recovery.fstab from the device and make it CWM-compatible (Only three columns: mount point, fstype, mount device) by removing all the Samsung-specific cruft - look at the SGSII recovery.fstab for an example of what it should look like
Put in the partition sizes and such in BoardConfig.mk
Need to do the source edits to CWM I mention a few posts above - both are oneliners. I'll try to post a patch when I have the time
Once you build by "make recoveryzip" you need to manually go and copy the recovery binary and all its symlinks into the initramfs - it's in out/target/product/samsung/yourdevicenamehere/ somewhere
dammit...
Something is not working right. Nandroid restores aren't restoring properly... The restores always seem to get corrupted.
****ing RFS... I hate it.
Somehow restored directories are FUBAR:
Code:
ls: ./tts: No such file or directory
ls: ./bin: No such file or directory
ls: ./app: No such file or directory
ls: ./xbin: No such file or directory
ls: ./vendor: No such file or directory
ls: ./firmware: No such file or directory
ls: ./usr: No such file or directory
ls: ./fonts: No such file or directory
ls: ./lib: No such file or directory
ls: ./media: No such file or directory
ls: ./modules: No such file or directory
ls: ./vsc: No such file or directory
-rw-r--r-- 1 root root 2437 Sep 16 21:35 build.prop
drwxr-xr-x 1 root root 0 Jan 9 23:24 cameradata
-rw-r--r-- 1 root root 44 Sep 16 21:35 default.prop
drwxr-xr-x 1 root root 0 Jan 9 23:24 etc
drwxr-xr-x 1 root root 0 Jan 9 23:24 framework
Entropy512 said:
I'll try to when I have some time... next few days are going to be busy.
Very quick summary:
Need to fix mkvendor.sh so TMPDIR is set when you don't provide a bootimage argument, otherwise the device tree doesn't get created properly
Need to pull recovery.fstab from the device and make it CWM-compatible (Only three columns: mount point, fstype, mount device) by removing all the Samsung-specific cruft - look at the SGSII recovery.fstab for an example of what it should look like
Put in the partition sizes and such in BoardConfig.mk
Need to do the source edits to CWM I mention a few posts above - both are oneliners. I'll try to post a patch when I have the time
Once you build by "make recoveryzip" you need to manually go and copy the recovery binary and all its symlinks into the initramfs - it's in out/target/product/samsung/yourdevicenamehere/ somewhere
Click to expand...
Click to collapse
Can you post the patch with the modifications in the source? Thanks!
Attached patch and my current recovery.fstab - remove the .txt extensions
Still not exactly working... RFS formatting seems busted.
Figured out the formatting problems - a few services running from /system were preventing it from unmounting properly.
I did a successful Nandroid restore - things are close now. I may just need to tweak recovery.fstab a bit more to have RFS format/restore/backup fully working.
Voodoo Lagfix - whole other story...
Entropy512 said:
Figured out the formatting problems - a few services running from /system were preventing it from unmounting properly.
I did a successful Nandroid restore - things are close now. I may just need to tweak recovery.fstab a bit more to have RFS format/restore/backup fully working.
Voodoo Lagfix - whole other story...
Click to expand...
Click to collapse
It's a step closer and a BIG advance if we got CWM working on our devices. Really thank you.

di11i kernel beta .01 (unsecure boot.img- root shell, adb remount) ext4 tweaks/init.d

This basically (for right now) the stock kernel image with some tweaks to init.
-unsecure boot.img-(perma- rooted adb shell)
-adb remount(mount /system r/w)
-init.d support added in for init scripts(survive after reboot)
-ext4 filesystem mount tweaks-/system/data/cache(faster I/O)
(Actually have a full ext2 version for,****s and giggles, but it boots into the encryption error screen, so cant release that. May have a fix though)
-mem and cache tweaks coming soon
still working on the actual kernel, will release that next, still have lots of work to do
boot.img MD5- fca41dba8f4699b67fd461a1632b65cf
MAKE BACKUP FIRST
You will not have issues if you wipe data, chances are youll be fine if you dont, but if anything starts acting up just wipe data, then install boot.img
This is the actual boot.img and for now has to be flashed with adb with these EXACT commands--
Code:
adb push boot.img /sdcard/
adb shell
su
dd if=/sdcard/boot.img of=/dev/block/mmcblk0 seek=3968 bs=4096 count=2048
reboot
#Thanks Diamondback and friends for offsets
This is not the ideal way to flash this, but i can not get the blob file to staging partition to actually write the image. Im pretty sure its an issue with the signature of the blob file, so if someone wants to pack this into a blob id be glad test it and then update the OP.
My primes been using this boot.img for a few days, along with one other. Flash at your own risk.
NOT COMPATIBLE WITH CM-9
di11igaf said:
This basically (for right now) the stock kernel image with some tweaks to init.
-unsecure boot.img-(perma- rooted adb shell)
-adb remount(mount /system r/w)
-init.d support added in for init scripts(survive after reboot)
-ext4 filesystem mount tweaks-/system/data/cache(faster I/O)
(Actually have a full ext2 version for,****s and giggles, but it boots into the encryption error screen, so cant release that. May have a fix though)
-mem and cache tweaks coming soon
still working on the actual kernel, will release that next, still have lots of work to do
boot.img MD5- fca41dba8f4699b67fd461a1632b65cf
MAKE BACKUP FIRST
You will not have issues if you wipe data, chances are youll be fine if you dont, but if anything starts acting up just wipe data, then install boot.img
This is the actual boot.img and for now has to be flashed with adb with these EXACT commands--
Code:
adb push boot.img /sdcard/
adb shell
su
dd if=/sdcard/boot.img of=/dev/block/mmcblk0 seek=3968 bs=4096 count=2048
reboot
#Thanks Diamondback and friends for offsets
This is not the ideal way to flash this, but i can not get the blob file to staging partition to actually write the image. Im pretty sure its an issue with the signature of the blob file, so if someone wants to pack this into a blob id be glad test it and then update the OP.
My primes been using this boot.img for a few days, along with one other. Flash at your own risk.
Click to expand...
Click to collapse
Great! I'm not the only one who as this problem. For what ever reason I too cannot get the blob to write.
I've tried running the blobpack -s to sign the blob but after trying to write it to the staging partition its a no go.....
EDIT: how are you working on the kernel if it has not been released yet?? JW, not trying to say your not but it would be cool if someone found a way to edit a compiled kernel.
MasterZen88 said:
Great! I'm not the only one who as this problem. For what ever reason I too cannot get the blob to write.
I've tried running the blobpack -s to sign the blob but after trying to write it to the staging partition its a no go.....
EDIT: how are you working on the kernel if it has not been released yet?? JW, not trying to say your not but it would be cool if someone found a way to edit a compiled kernel.
Click to expand...
Click to collapse
ICS kernel source is on the global download site.
Doktaphex said:
ICS kernel source is on the global download site.
Click to expand...
Click to collapse
Correct but thats an older source code V9.4.2.7 right?
MasterZen88 said:
Correct but thats an older source code V9.4.2.7 right?
Click to expand...
Click to collapse
True, I don't know how much they have changed since then. Surely they need to be releasing new source pretty quickly to be in compliance with the license. I don't know enough about that side of things though.
prime has a BLOBSIGN signature in the first 28 bytes of the blob files for prime.
you just need to copy those 28bytes from an existing blob file that works.
this is of course only working on a unlocked prime. then the BLOBSIGN signature is not checked, so anything can be pasted in front of it.
Rayman will update the packblob to include the signature later on.
MasterZen88 said:
Great! I'm not the only one who as this problem. For what ever reason I too cannot get the blob to write.
I've tried running the blobpack -s to sign the blob but after trying to write it to the staging partition its a no go.....
EDIT: how are you working on the kernel if it has not been released yet?? JW, not trying to say your not but it would be cool if someone found a way to edit a compiled kernel.
Click to expand...
Click to collapse
I'm confused. I thought after our PMs you had it flashing just fine?
Sent from my Galaxy Nexus using Tapatalk
jermaine151 said:
I'm confused. I thought after our PMs you had it flashing just fine?
Sent from my Galaxy Nexus using Tapatalk
Click to expand...
Click to collapse
Yes it flashed but later I realize non of my settings/changes took effect.
When I stated it work, I had took the stock kernel/initramfs from asus blob, extracted, made no changes at all, recompile like we talked about in our PM's then flashed it. It work but then after I made changes, recompiled. Non of my changes took effect.
Again Jermaine you have been a BIG help. and can only hope one day I can return the favor!!!
jermaine151 said:
I'm confused. I thought after our PMs you had it flashing just fine?
Sent from my Galaxy Nexus using Tapatalk
Click to expand...
Click to collapse
MasterZen88 said:
Yes it flashed but later I realize non of my settings/changes took effect.
When I stated it work, I had took the stock kernel/initramfs from asus blob, extracted, made no changes at all, recompile like we talked about in our PM's then flashed it. It work but then after I made changes, recompiled. Non of my changes took effect.
Again Jermaine you have been a BIG help. and can only hope one day I can return the favor!!!
Click to expand...
Click to collapse
Flashing blobs does NOT always work. No idea why yet. Directly dd'ing the boot.img to the right place works BETTER than flashing a blob in some cases...
Asus really screwed up with this Unlocker thing....
MasterZen88 said:
Yes it flashed but later I realize non of my settings/changes took effect.
When I stated it work, I had took the stock kernel/initramfs from asus blob, extracted, made no changes at all, recompile like we talked about in our PM's then flashed it. It work but then after I made changes, recompiled. Non of my changes took effect.
Again Jermaine you have been a BIG help. and can only hope one day I can return the favor!!!
Click to expand...
Click to collapse
No problem. I'm surprised it didn't work for you after you made changes.
Diamondback said:
Flashing blobs does NOT always work. No idea why yet. Directly dd'ing the boot.img to the right place works BETTER than flashing a blob in some cases...
Asus really screwed up with this Unlocker thing....
Click to expand...
Click to collapse
Flashing them to staging seems to work perfectly for me. That's how I made my insecure boot.blob and flashed it. It doesn't matter whether you're on unofficial CWM or Official since if you start your updater-script with unmounting staging, it doesn't hurt anything if staging wasn't mounted. Then you can either dd the blob to mmcblk0p4 or you can directly flash it via:
Code:
package_extract_file("boot.blob", "/dev/block/mmcblk0p4");
I have not found an occasion where this didn't work. You need to make sure that you don't add any --cmdline parameters when you mkbootimg or you will get a bootloop. Staging is definitely the safest way to flash to any partitions on the Prime.
MasterZen88 said:
Correct but thats an older source code V9.4.2.7 right?
Click to expand...
Click to collapse
The kernel versions seem to be the same with the latest .15 build as the previous one. I wonder if they just recompiled the same source and that's why the host name changed to Mercury.
EDIT: I'm going to pack this kernel into a flashable blob and attach it here.
jermaine151 said:
No problem. I'm surprised it didn't work for you after you made changes.
Flashing them to staging seems to work perfectly for me. That's how I made my insecure boot.blob and flashed it. It doesn't matter whether you're on unofficial CWM or Official since if you start your updater-script with unmounting staging, it doesn't hurt anything if staging wasn't mounted. Then you can either dd the blob to mmcblk0p4 or you can directly flash it via:
Code:
package_extract_file("boot.blob", "/dev/block/mmcblk0p4");
I have not found an occasion where this didn't work. You need to make sure that you don't add any --cmdline parameters when you mkbootimg or you will get a bootloop. Staging is definitely the safest way to flash to any partitions on the Prime.
Click to expand...
Click to collapse
Hmm.. not sure whats going on but I did use this command
Code:
mkbootimg --kernel boot.img-kernel.gz --ramdisk newramdisk.cpio.gz -o newboot.img
To recompile the boot.img then used the new blobtoolsv2 to repack the blob
Code:
blobpack -s kernelblob blob.LNX newboot.img
I know staging is unmounted because I'm using your updater-script
jermaine151 said:
Flashing them to staging seems to work perfectly for me. That's how I made my insecure boot.blob and flashed it. It doesn't matter whether you're on unofficial CWM or Official since if you start your updater-script with unmounting staging, it doesn't hurt anything if staging wasn't mounted. Then you can either dd the blob to mmcblk0p4 or you can directly flash it via:
Code:
package_extract_file("boot.blob", "/dev/block/mmcblk0p4");
I have not found an occasion where this didn't work. You need to make sure that you don't add any --cmdline parameters when you mkbootimg or you will get a bootloop. Staging is definitely the safest way to flash to any partitions on the Prime.
Click to expand...
Click to collapse
if you copy the blob, while p4 is mounted to staging in cwm, you will have a problem when you reboot, as the filesystem umount, and that cause a small change to the mounted filesystem, to indicate it is umounted normally. and the p4 partition becomes invalid.
so it can easily become a problem if p4 is mounted.
at least thats how i experienced the problem when seen initially.
Dexter_nlb said:
if you copy the blob, while p4 is mounted to staging in cwm, you will have a problem when you reboot, as the filesystem umount, and that cause a small change to the mounted filesystem, to indicate it is umounted normally. and the p4 partition becomes invalid.
so it can easily become a problem if p4 is mounted.
at least thats how i experienced the problem when seen initially.
Click to expand...
Click to collapse
That's why I said to always add an unmount("/staging"); to the top of your script. The attached flashable zip works perfectly on the official CWM; staging is not mounted by default, but I told the script to unmount it anyway in case someone with the unofficial CWM tries to flash it. If it's already unmounted, the script just continues.
Anyone may use this as a template for consistently flashing a kernel to staging.
MasterZen88 said:
Hmm.. not sure whats going on but I did use this command
Code:
mkbootimg --kernel boot.img-kernel.gz --ramdisk newramdisk.cpio.gz -o newboot.img
To recompile the boot.img then used the new blobtoolsv2 to repack the blob
Code:
blobpack -s kernelblob blob.LNX newboot.img
I know staging is unmounted because I'm using your updater-script
Click to expand...
Click to collapse
Hmm... That all looks good. What happens after you attempt to flash this and what changes did you make to the ramdisk?
EDIT: Hey, I just re-read your PM. It sounds like you're editing the wrong file in the ramdisk. To make an insecure boot image, you should be editing init.rc, NOT init.cardhu.rc.
Here's a great guide to follow. It's the same procedure for the Prime as the original Transformer:
http://forum.xda-developers.com/showthread.php?t=1193737
jermaine151 said:
Hmm... That all looks good. What happens after you attempt to flash this and what changes did you make to the ramdisk?
EDIT: Hey, I just re-read your PM. It sounds like you're editing the wrong file in the ramdisk. To make an insecure boot image, you should be editing init.rc, NOT init.cardhu.rc.
Here's a great guide to follow. It's the same procedure for the Prime as the original Transformer:
http://forum.xda-developers.com/showthread.php?t=1193737
Click to expand...
Click to collapse
Okay that makes sense. One f my changes was to default.prop for insecure boot.img
Code:
ro.secure=0
This did not show after I flashed. But I will try again in a couple hour's once my workload calms down... SysAdmin life is not easy...lol
MasterZen88 said:
Okay that makes sense. One f my changes was to default.prop for insecure boot.img
Code:
ro.secure=0
This did not show after I flashed. But I will try again in a couple hour's once my workload calms down... SysAdmin life is not easy...lol
Click to expand...
Click to collapse
LOL! I meant to say default.prop and not init.rc. I totally understand. I used to be a SysSdmin too. Now I moved more towards the networking side of things. Much more calm.
jermaine151 said:
LOL! I meant to say default.prop and not init.rc. I totally understand. I used to be a SysSdmin too. Now I moved more towards the networking side of things. Much more calm.
Click to expand...
Click to collapse
Would love to be on that side of the tracks. I'm banging my head against the wall right now trying to figure out how to implement Microsoft System Center Service Manager 2010 and on top of that one of our Citrix Broker boxes went down last night... Not a funny day at all. But I'll keep this thread about Android...for now....
No, unfortunately flashing a blob does not always do the same as dd'ing boot.img directly. We recently had a case were someone lost root access. the only way to recover was to directly dd a boot.img.
Flashing the SAME boot.img via blob did not work (nor did any other blob we tested)
So blob flashing obviously has some quirks....
Diamondback said:
No, unfortunately flashing a blob does not always do the same as dd'ing boot.img directly. We recently had a case were someone lost root access. the only way to recover was to directly dd a boot.img.
Flashing the SAME boot.img via blob did not work (nor did any other blob we tested)
So blob flashing obviously has some quirks....
Click to expand...
Click to collapse
I just haven't see that issue flashing any of my blobs and CM9 is using the same method and it's working. Who knows.
jermaine151 said:
I just haven't see that issue flashing any of my blobs and CM9 is using the same method and it's working. Who knows.
Click to expand...
Click to collapse
We only had it one time too, but one time is enough to scare me further
This whole blob stuff just is crap

X10 [/data/media] / [/sdcard] wiped & not available?!

Alright, starting here in this thread you can see what I've tried to fix this:
http://forum.xda-developers.com/showthread.php?p=43218631#post43218631
My internal sdcard has been wiped and is no longer available in CWM or the OS. Have tried multiple ftf flashes from stock 2.1, 2.3 and CM10 but no luck. If I have an external sdcard in the phone while flashing it uses that as the internal space, but it's not seeing the flash memory that should be available in the actual device.
Everything works 'fine' without that space available, just that file explorers can't see it, camera can't take pictures, about what you may expect.
Anybody experience anything similar?
Been around Android a long time now and quite familiar with the ins-and-outs but this has me stumped.
Thanks!
Try Another Rom My Friend
Ronald G V A said:
Try Another Rom My Friend
Click to expand...
Click to collapse
It's definitely not just a ROM install issue as I've tried 3 or 4 different ones now and they don't fix the underlying issue.
Need to properly repartition my internal sdcard so it's visible to recovery/Android.
seanpr123 said:
It's definitely not just a ROM install issue as I've tried 3 or 4 different ones now and they don't fix the underlying issue.
Need to properly repartition my internal sdcard so it's visible to recovery/Android.
Click to expand...
Click to collapse
jaja ok Senior Member All as you said (Y)
I tried another ROM and kernel, no luck.
Anybody else have any thoughts/suggestions?
Please please
Please, anybody else have any input?
Stuck in the mud here... :-\
Hi there, if been wondering the same things and i almost sure how it works and that particular service is set from the services in one of the .rc files but we doesnt seem to have it.
Its about a binary and needs to be set from the kernels ramdisk.
This is what i found so far on info about it.
the /system/bin/sdcard binary, which mounts the /data/media folder at /mnt/sdcard and does some magic with the user/group ids. from one of the
/init.????.rc files.
This i saw in another devices .rc file.
# create virtual SD card at /mnt/sdcard, based on the /data/media directory
# deamon will drop to user/group system/media_rw after initializing
# underlying files in /data/media wil be created with user and group media_rw (1023)
service sdcard /system/bin/sdcard /data/media 1023 1023
class late_start
If i have some time i will try to set that services in x10i ramdisk with kernel to see if that will work.
Sent from my x10i
Danny23 said:
Hi there, if been wondering the same things and i almost sure how it works and that particular service is set from the services in one of the .rc files but we doesnt seem to have it.
Its about a binary and needs to be set from the kernels ramdisk.
If i have some time i will try to set that services in x10i ramdisk with kernel to see if that will work.
Click to expand...
Click to collapse
Not really following that, what are you suggesting I try?
Simply a new kernel or something else?
QUOTE=seanpr123;43754685]Not really following that, what are you suggesting I try?
Simply a new kernel or something else?[/QUOTE]
I am sorry it still not 100% clear to me.
But you need a new ramdisk inside the kernel.
And the framework-res.apk storage-list.xml needs also to be edited and recovery.fstab and vold.fstab.
And a file which we dont have some devices have it and i got it from Google search.
Its a file named "vold.conf".
I did these mods and cwm recovery is acting weird crying about log but some new options apeared and roms stil boots.
Here are few pictures of it with my gtn-n7000.
Sent from my x10i

Getting INT2EXT4 working on M4?

Good afternoon. As the 8 GB aren't enough for me, I've been looking for a solution. While doing so, I found the INT2EXT4 script, which mounts /data on the second partition of the SD card. The first time I flashed it (after enabling init.d support with an app) it didn't work at all. Then, I realised this script is written assuming that the second partition of the card is mmcblk0p2. But on the M4, it's mmcblk1p2 (found it out with another app).
I changed those lines on the script, but it didn't work either. I analyzed all the lines, and I can't see where the problem could be. Has anyone tried this, or could somebody help me with this issue? I think it could be really useful for a lot of users.
Another thing: flashing the file through TWRP gave me an error, so I copied the file directly to /system/etc/init.d and set the permissions manually. I hope this is not a problem, but let me know if it could be...
sounds interesting... ill take a look at it soon!
Have you checked the "adress" of internal sd? In the script there is a connection between internal and external storage so i would assume that both "adresses" need to be correct in order for the script to work. Just an opinion....
Sent from my SM-G901F using XDA Free mobile app
vader20 said:
Have you checked the "adress" of internal sd? In the script there is a connection between internal and external storage so i would assume that both "adresses" need to be correct in order for the script to work. Just an opinion....
Sent from my SM-G901F using XDA Free mobile app
Click to expand...
Click to collapse
several1989 said:
sounds interesting... ill take a look at it soon!
Click to expand...
Click to collapse
I think there's a problem with the commands that the script uses. If I try to do umount from the terminal, nothing happens. So maybe the script doesn't work in this phone, but if any of you are able to make it do so, let me know. So I think it's better to build the kernel from source, but before that make changes in the fstab file. With that, /data will be mounted at the second partition of the sdcard directly at boot time. Trying to do this now, but I'm facing some difficulties building the kernel. If I make it work, I'll post it
Testé folder Mouth playstore

Emui 9.1. The system partition does not go in rw

Hello. I installed emui 9.1 on my mate20pro with the dload method. I unlocked the bootloader again and rooted the system with patched recovery. The phone is rooted, but the system partition remains in ro. I have tried various apps to mount the system in rw, but without success. I have also tried some commands found on the net both through adb and emulator, but nothing to do. I have also tried using rootexplorer which has been granted root permissions, but fails to modify anything, nor to insert or delete files within the system folders. What can be done?
ernia66 said:
Hello. I installed emui 9.1 on my mate20pro with the dload method. I unlocked the bootloader again and rooted the system with patched recovery. The phone is rooted, but the system partition remains in ro. I have tried various apps to mount the system in rw, but without success. I have also tried some commands found on the net both through adb and emulator, but nothing to do. I have also tried using rootexplorer which has been granted root permissions, but fails to modify anything, nor to insert or delete files within the system folders. What can be done?
Click to expand...
Click to collapse
it's the new filesystem called erofs (Extendable Read-Only File System)
So there's nothing to do?
Sorry OP if this is off topic , but its about magisk and root in 9.1
I have the "system partition remains in ro" issue too , its annoying yes , but I have a bigger problem
When "rooted" , the phone app ( dialer ), and the contacts app both get broken and crash .. So I have to fastbootflash twrp and lose root to make those apps to work normally again
Anyone with the same problem ?
I need root so I can change and use the font ( I like ) with a magisk module
ernia66 said:
So there's nothing to do?
Click to expand...
Click to collapse
Magisk can already 'replace on the fly' file from /system systemless-y.
It dies already support erofs.
Check Magisk documentations.
There is also a way to mount custom erofs partition....
which is?
ernia66 said:
which is?
Click to expand...
Click to collapse
Build your own erofs partitions (cf. erofs.mkfs)
Unmount stock ones and mount yours.
I just did some test 48h ago on erofs /cust and it seems to work...
-Mounting other file system type on same folder could even work !
oslo83 said:
Build your own erofs partitions (cf. erofs.mkfs)
Unmount stock ones and mount yours.
I just did some test 48h ago on erofs /cust and it seems to work...
-Mounting other file system type on same folder could even work !
Click to expand...
Click to collapse
Unfortunately I am not able to do it
:crying:
pvillasuso said:
Sorry OP if this is off topic , but its about magisk and root in 9.1
I have the "system partition remains in ro" issue too , its annoying yes , but I have a bigger problem
When "rooted" , the phone app ( dialer ), and the contacts app both get broken and crash .. So I have to fastbootflash twrp and lose root to make those apps to work normally again
Anyone with the same problem ?
I need root so I can change and use the font ( I like ) with a magisk module
Click to expand...
Click to collapse
use search button pls, you will find the solution already posted
ilos said:
use search button pls, you will find the solution already posted
Click to expand...
Click to collapse
Shhh
Here Rootexplorer paying app does not help writing on erofs partitions...
- Or did you meant something else ?
Magisk Doc. :
https://topjohnwu.github.io/Magisk/
So far, out-of-the-box, Magisk is already ready to replace systemless-ly files and paths on /system easily when used with modules.
-Module just need a a small tweak to add Magisk systemless-ly compatibility to other erofs partitions like /cust
Another way would be to bind or mount...
ilos said:
use search button pls, you will find the solution already posted
Click to expand...
Click to collapse
Wow , what , where ???
I did search but maybe I didnt use the right keywords for such an specific issue
Any extra help , would be very appreciated !
oslo83 said:
Build your own erofs partitions (cf. erofs.mkfs)
Unmount stock ones and mount yours.
I just did some test 48h ago on erofs /cust and it seems to work...
-Mounting other file system type on same folder could even work !
Click to expand...
Click to collapse
I'd like to create my own partition but I don't know how to do it. Is there any online guide for beginners that I can try?
oslo83 said:
Shhh
Here Rootexplorer paying app does not help writing on erofs partitions...
- Or did you meant something else ?
Click to expand...
Click to collapse
my reply was not for you but for the off topic before my reply: contact, ph app and message force close after root.
as I said, there is already a solution about force close.
Anyway I can only write in /system/etc and some over folders
oslo83 said:
Build your own erofs partitions (cf. erofs.mkfs)
Unmount stock ones and mount yours.
I just did some test 48h ago on erofs /cust and it seems to work...
-Mounting other file system type on same folder could even work !
Click to expand...
Click to collapse
Hi, I saw that you managed to get a custom erofs partition mounted using magisk. I was wondering if the same could be done for the/product folder because that would allow me to fix the DPI scaling issues found on EMUI 9.0-9.1. Would it be possible to see the code for your magisk module?
Ok, So.
After having a look at what I did with this WorkInProgress...
So far there is 2 main way I thought about and played with:
Method A
A1/
Using the mkfs.erofs tool to make your custom erofs partition and mount it over the original one. (mayne umount it before)
Thanks @ph03n!x for his work on mkfs.erofs.
Thanks @ante0 for his android arm build.
A2/
I did not yet tryed to flash via fastboot or adb these modded erofs images.
Method B
What I'm actually doing is just mounting systemless-y with a magisk module whole folders over erofs one ;
In post-fs-data.sh, I'm doing:
mount -o bind $MODDIR/cust /cust
mount -o bind $MODDIR/version /version
With a ~150MB custom /cust it is working ; It should be tryed/perfected for bigger folders.
Another way, to not replace/spoof the whole folders could be to mount the whole list of custom files at their respective path (-to be tested)
PS: I'm not sure if update-binary module's file should be modded with some:
mount /cust 2>/dev/null
@ line 42
Sources:
https://forum.xda-developers.com/huawei-p30-pro/how-to/erofs-mounting-accessing-img-t3944193
oslo83 said:
Ok, So.
After having a look at what I did with this WorkInProgress...
So far there is 2 main way I thought about and played with:
Method A
A1/
Using the mkfs.erofs tool to make your custom erofs partition and mount it over the original one. (mayne umount it before)
Thanks @ph03n!x for his work on mkfs.erofs.
Thanks @ante0 for his android arm build.
A2/
I did not yet tryed to flash via fastboot or adb these modded erofs images.
Method B
What I'm actually doing is just mounting systemless-y with a magisk module whole folders over erofs one ;
In post-fs-data.sh, I'm doing:
mount -o bind $MODDIR/cust /cust
mount -o bind $MODDIR/version /version
With a ~150MB custom /cust it is working ; It should be tryed/perfected for bigger folders.
Another way, to not replace/spoof the whole folders could be to mount the whole list of custom files at their respective path (-to be tested)
PS: I'm not sure if update-binary module's file should be modded with some:
mount /cust 2>/dev/null
@ line 42
Sources:
https://forum.xda-developers.com/huawei-p30-pro/how-to/erofs-mounting-accessing-img-t3944193
Click to expand...
Click to collapse
I'm rather new to this trying to do something similar on my EMUI9.1 using Magisk, I'm having a problem that i can replace anything in most other folders without much problem like /etc/ for example, but when i try to substitute /vendor/, whatever i put it the folder does not only fail to get substituted, it even disappears from the $MODDIR/system/vendor folder upon restart!
I.e. let's say I create a new file in $MODDIR/system/etc. Upon restart i'm successfully seeing that same file appear /etc.
However when I create a new file under $MODDIR/system/vendor, Upon restart i'm not seeing that under /vendor. ALSO the file is gone from $MODDIR/system/vendor like it was never there.
Any idea what is happening? I've also tried manually mounting in the post-fs-data.sh, where i put the following:
mount -o bind $MODDIR/system/vendor /vendor
That actually caused my system to freeze on the loading screen. I had to reset magisk to non-modded settings in order to get it to run again.
Any help's appreciated.
pvillasuso said:
Sorry OP if this is off topic , but its about magisk and root in 9.1
I have the "system partition remains in ro" issue too , its annoying yes , but I have a bigger problem
When "rooted" , the phone app ( dialer ), and the contacts app both get broken and crash .. So I have to fastbootflash twrp and lose root to make those apps to work normally again
Anyone with the same problem ?
I need root so I can change and use the font ( I like ) with a magisk module
Click to expand...
Click to collapse
i got the solution of above problem from this https://forum.xda-developers.com/huawei-p30-pro/how-to/guide-magisk-root-p30-pro-bl-code-t3938285
Hi I have almost the same issue after updating to EMUI 9.1 the phone hardbricked. It only shows as fastboot on my PC I tried flashing but not sending the img files please help

Categories

Resources