Related
Hi,
I had installed xrecovery and tried to add JIT, but wrong xrecovery (with sdcard bug) and JIT causes this error (boot loop):
~ # [[email protected] ~]$ adb logcat
- waiting for device -
link_image[1965]: 1120 could not load needed library 'libm.so' for '/system/bin/seserve_mem_region[834]: OOPS: 1120 prelinked library 'libm.so' mapped at 0x400, not at 0xafc00000)CANNOT LINK EXECUTABLE
Before i did a backup with xrecovery, byt i cant recover it, because it gives me sdcard mount error. So I had stuck on it!
I cant use flashtool to reflash my phone, because I am on linux...And need backed up data of course...
I hope, because adb shell still works, that recovery of libraries could work, could anyone good man post for me these libraries (and binaries)?
/system/bin/dalvikvm
/system/bin/dexopt
/system/bin/logcat
/system/lib/libcutils.so
/system/lib/libdl.so
/system/lib/libdvm.so
/system/lib/liblog.so
/system/lib/libm.so
/system/lib/libnativehelper.so
/system/lib/libz.so
Thank you in advance. Hope anyone help me!
HeliumX10 said:
Hi,
I had installed xrecovery and tried to add JIT, but wrong xrecovery (with sdcard bug) and JIT causes this error (boot loop):
~ # [[email protected] ~]$ adb logcat
- waiting for device -
link_image[1965]: 1120 could not load needed library 'libm.so' for '/system/bin/seserve_mem_region[834]: OOPS: 1120 prelinked library 'libm.so' mapped at 0x400, not at 0xafc00000)CANNOT LINK EXECUTABLE
Before i did a backup with xrecovery, byt i cant recover it, because it gives me sdcard mount error. So I had stuck on it!
I cant use flashtool to reflash my phone, because I am on linux...And need backed up data of course...
I hope, because adb shell still works, that recovery of libraries could work, could anyone good man post for me these libraries (and binaries)?
/system/bin/dalvikvm
/system/bin/dexopt
/system/bin/logcat
/system/lib/libcutils.so
/system/lib/libdl.so
/system/lib/libdvm.so
/system/lib/liblog.so
/system/lib/libm.so
/system/lib/libnativehelper.so
/system/lib/libz.so
Thank you in advance. Hope anyone help me!
Click to expand...
Click to collapse
Why were you trying instal JIT?
Anyway, would it be easier to repair phone with pccompanion rather thatn do this?
Its only a suggestion, I am not savvy with linux
I was trying to install JIT because of speed up my phone. Lags more than few seconds were getting me really angry. even with few apps on background and good task manager...
You`re true, it COULD be easier. But in linux it is not possible to run PC companion or SEUS...
That`s why I am looking for files. I am not a noob
HeliumX10 said:
I was trying to install JIT because of speed up my phone. Lags more than few seconds were getting me really angry. even with few apps on background and good task manager...
You`re true, it COULD be easier. But in linux it is not possible to run PC companion or SEUS...
That`s why I am looking for files. I am not a noob
Click to expand...
Click to collapse
I never said you were, I just thought JIT was not needed on GB roms...and I have no experience with linux.
Cant you just download thos files from a custom rom, or extract them a stock fw?
Again, apologies if this is not possible on linux, but I dont have access to those files at the moment
gregbradley said:
I never said you were, I just thought JIT was not needed on GB roms...and I have no experience with linux.
Cant you just download thos files from a custom rom, or extract them a stock fw?
Again, apologies if this is not possible on linux, but I dont have access to those files at the moment
Click to expand...
Click to collapse
No need to apologize... I am discussing a lot of time with totally non experienced users (because I am linux forum administrator)
So I havent read that JIT is not needed on GB... Damn, if it is true I made a really stupid and unnessesary mistake with horrible effect...
I`ll google how it is about it, but want to recover my phone again first
Yes, I had downloaded few stock roms, but I am not able to extract files from them! Looking for some ROM dissasembler or something like that... I notice, that many windows programs (but not all - especially flash tools) works under linux properly, so looking for ANY extractor...
Maybe unyaffs will work. Doomlord had those files. Send him a pm, he is very knowledgeable about this
Also, can you not use win zip (or similar) on Linux....
Sent from my X10i using xda premium
HeliumX10 said:
Hi,
I had installed xrecovery and tried to add JIT, but wrong xrecovery (with sdcard bug) and JIT causes this error (boot loop):
~ # [[email protected] ~]$ adb logcat
- waiting for device -
link_image[1965]: 1120 could not load needed library 'libm.so' for '/system/bin/seserve_mem_region[834]: OOPS: 1120 prelinked library 'libm.so' mapped at 0x400, not at 0xafc00000)CANNOT LINK EXECUTABLE
Before i did a backup with xrecovery, byt i cant recover it, because it gives me sdcard mount error. So I had stuck on it!
I cant use flashtool to reflash my phone, because I am on linux...And need backed up data of course...
I hope, because adb shell still works, that recovery of libraries could work, could anyone good man post for me these libraries (and binaries)?
/system/bin/dalvikvm
/system/bin/dexopt
/system/bin/logcat
/system/lib/libcutils.so
/system/lib/libdl.so
/system/lib/libdvm.so
/system/lib/liblog.so
/system/lib/libm.so
/system/lib/libnativehelper.so
/system/lib/libz.so
Thank you in advance. Hope anyone help me!
Click to expand...
Click to collapse
dalvikvm
dexopt
logcat
libcutils.so
libdl.so
libdvm.so
liblog.so
libm.so
libnativehelper.so
libz.so
Thanks a lot matthew5025 ! (thanks button pressed of course )
I have rewrited all the files, which has been changed... But without effect... This is really strange, I cant understand it...
So, I have to flash, its bad...
But my first flashing ROM I`d like to flash some of custom rom (not official) - any suggestions?
I like fast and stable system, dont need any special features and apps, animations and laggy eye candy launchers.
HeliumX10 said:
Thanks a lot matthew5025 ! (thanks button pressed of course )
I have rewrited all the files, which has been changed... But without effect... This is really strange, I cant understand it...
So, I have to flash, its bad...
But my first flashing ROM I`d like to flash some of custom rom (not official) - any suggestions?
I like fast and stable system, dont need any special features and apps, animations and laggy eye candy launchers.
Click to expand...
Click to collapse
only 2 ROMS.
FXP045 (FXP will be release anytime soon this week)
Here
zdzihu ZMOD
Here
cheers
HeliumX10 said:
Thanks a lot matthew5025 ! (thanks button pressed of course )
I have rewrited all the files, which has been changed... But without effect... This is really strange, I cant understand it...
So, I have to flash, its bad...
But my first flashing ROM I`d like to flash some of custom rom (not official) - any suggestions?
I like fast and stable system, dont need any special features and apps, animations and laggy eye candy launchers.
Click to expand...
Click to collapse
I like wolfbreak's gamer's rom as it's FAST! link but you should go on a flashing spree, than choose one you like.
VMODS 5 rom +5.1 Update with doomkernel 0.5a!!
Thanks guys for your tips !
Tried both ROMs and I chose Wolfbreak`s gaming ROM, really fast
Sweet
As of you guys have found out there has been a few threads that have been started that throw around theories about getting CWMR to work and such and I'm here to finally give you guys the progress from the dev that's working on it.
I have been working on getting CWMR working off and on for a few months now to some avail. I have been slowly testing it and sending out alphas to people on the IRC and now I'm here to tell you where I'm at and what has been accomplished.
Part 1: Getting into CWMR
I started working with our current bootstrap recovery as a base for CWMR. It was of course in charge mode, but it was a step that had to be done first. I got a hold on a CWMR binary and after a bunch of shell commands managed to get the mangled version of CWMR that we've seen before. From there I had the idea to inject the ATRIX binary into the recovery and try it from there. It was a big success and I could use some (very few) of the functions that if offered. But I was finally in CWMR on the X2. (I posted the picture on my twitter).
Part 2: Making the Functions Work
From here the first thing I decided to do was to get the "Fix Permissions" script working. By getting this to work I know I could place my own code in there that will log to the file that I could open up and see whatever I wanted as ADB didn't work YET (more on that later). I fiddled with it a bit and got it working.
This also allowed me to access the log and find out why the unmounting and formatting wasn't working either. This was due to an issue with the mke2fs and other complementary binaries not being appropriately compiled for our phone. As I was working on CM7 at the time, i had access to the sources of these and to the recovery binary as well. I decided to build a recovery binary that was tailored to our phone and not the ATRIX and the binaries that handle formatting partitions, etc. These worked on some partitions I could wipe data and cache properly by reformatting, I could easily erase a partition at will except for one: /system.
Part 3: Getting /system Formatted (WIP)
At this point I decided to find out why /system could never be formatted. When a partition needs formatting it must be unmounted first. But I could never unmount /system. By using the logs I found out that the partition was still in use. By using the "Fix Permissions" script, I managed to find that a lot of daemons (aka services) were being ran from the /system sub directories. This meant that in order to format, I had to have them not running. Easy enough as there's a script in CWMR that allows us to kill certain processes before the new recovery starts. I decided to kill all of the services with this script but found that the display would go blank. From there I found that two services are REQUIRED to be running on the phone for drawing (making things show on the screen) to take place: servicemanager and nvrm_daemon.
The servicemanager does exactly what it's title says: it manages the currently running services. This has to be running to keep nvrm_daemon running or else things get kinda freaky.
nvrm_daemon is what is used to manage drawing commands to the screen (the nv stands for nVidia which is who made the Tegra 2). With this service running I was able to (while in charge mode still) draw to the screen and show all of the menus for CWMR.
This created a problem. As these HAD to be running /system could never unmount because they were in there. Unless we used 2nd-init. From here I created my own hijack script that allowed me to copy over the necessary daemons into /sbin (which is just a sub directory of the rootfs). Then by using a new init.rc I could start these daemons from /sbin and see if they worked for CWMR.
After a lot of coding I managed to get the daemons running from /sbin and I thought I was free from there. And while I was at it, I might as well tried to get it out of charge mode.
Part 4: Untethering the Recovery (WIP)
I figured that since I was using 2nd-init anyway that I might as well have been working on trying to get it untethered also. After a bunch of looking into why charge mode was the only way to get it working, I found that before charge mode was started the init.rc started the "pre-zygote" services. These include usbd (the usb daemon), battd (the battery daemon), etc. What was also started was nvrm_daemon (another proof that we needed this running) to draw on the screen (charge mode draws the battery indicator to show charging progress).
I decided to move to mot_boot_mode to do the hijacking. I eventually managed to get things running but it still wouldn't draw to the screen. A look at the logs showed that for whatever reason we still didn't have access to the framebuffer (the memory space aka buffer that is drawn to the screen). I was tired of rebooting over and over again to get results and by now I had SBF'd I don't know how many times, only to not be able to access the logs due to a wipe. So next thing I decided was to get ADB working at boot.
Part 5: Starting ADB at Boot (The Short Version)
I took a look through the init.rc and found that Motorola had disabled the default way to get ADB working at boot. Through a bunch of digging I found that I could re-enable the method but it started usbd also (yet another daemon I'd have to move and thus more time, etc.). But eventually I found another property (aka system variable) that i could set that would start ADB without having usbd started: persist.adb.enable_NV_DISABLED. Setting that to 1 started ADB whenever I wanted it to. This allowed me access to the system (as root actually) without having to reboot and try to get into logs and whatnot. I could see everything that was running, mount/unmount and do everything I normally could and see better why they were failing.
Part 6: Recent Discoveries (as of 1/3/12) On the /system Unmount
I decided that I was going to go back to charge mode for now. After causing too much pain with framebuffer access I decided to look into why /system STILL wouldn't unmount after moving everything out of there. With the help of mastafunk and DrPenguin I was able to get a hold of a program named strace. This allows you to debug what binaries are doing down to pretty much every comparison, file opening, memory mapping, etc. I had found that nvrm_daemon (although started from /sbin) was looking for and loading drivers found in /system/lib. This was a problem and I thought that we may never get /system unmounted. Until I found out one other thing.
The nvrm_daemon doesn't look for the drivers inside of /system/lib first. It looks for them in /vendor/lib first and then if it can't find them, goes to /system/lib. So simple enough. I thought I could just make a /vendor/lib directory, copy the libs over and have them load from there. Well it's not that simple as there is already a file in the rootfs named vendor. But the file is a symlink to /system/vendor so if it looks at /vendor/lib it really looks at /system/vendor/lib. I thought I was stuck again until I took a closer look at the /system directory. There is no ./vendor sub directory in /system. That being said, I was able to delete the symlink in the rootfs and then create a /vendor/lib directory copy over the libs to it and nvrm_daemon successfully loaded the libs from there and not from system.
Conclusion
You are now caught up on when we've basically done to try to get CWMR working on the X2. Being that we're the only Tegra 2 phone from Motorola that is locked, we're kinda stuck trying to get things working. For now I consider this progress. Albeit slow progress, but progress none the less. If you took the time to read this, I hope you now understand the frustrations I've had to deal with this. All the while trying to help masta get 2nd-init fully working for our phones.
Thank you.
Thank you very much for the detailed explanation - well written so that even a novice like myself could understand your progress and frustrations. Thank you to you and your peers also for your hard work and determination to see this through. I think I can speak for all X2 users and say this is most appreciated!
Great writeup. Very informative. Thank you for your dedication.
I thought 2nd init was done and masta was working on aosp? I mean we can log from our new init.rc so it is using it. In the meantime i've been working on my barebones so we can get new blood using 2nd init and because I want a decent daily driver in the event we cant get stuff going.
To echo what others have said, thank you very much for the information. This kind of transparency is always very much appreciated when there's some spare time to put that out there. It not only is really helpful to make it understandable, but it's nice to know that these projects are still active.
Perhaps most importantly though in the context of potential progress: having specific information is a great way for knowledgeable people reading to recognize if they might be able to help out in certain areas. You might get some fresh ideas from people who hadn't considered themselves familiar enough with the situation to be useful. Who knows.
Anyways, thank you very much again. It's very encouraging to hear progress reports in general, not to mention ones like this that involve so many successes.
Wow thanks so much for your hard work and dedication can't wait until you conquer this sounds like you are closer then ever
Sent from my DROID X2 using XDA App
I am so glad to hear the "behind the scene" drama. Thank You DK sharing. I AM sorry for what bad drama I may have started. Will you forgive me?
skwoodwiva said:
I am so glad to hear the "behind the scene" drama. Thank You DK sharing. I AM sorry for what bad drama I may have started. Will you forgive me?
Click to expand...
Click to collapse
No need to appologize man... Stuff like this should have been posted or known. To be honest, I thought Masta was finishing 2nd INIT and starting to work on getting AOSP running and CWM was out the window, thats why I jumped in to try and get progress on CWM. Thanks DK and now I know the IRC is more active, I will be on there more to try and help out.
Thank you!
Thanks a ton for all the effort you're putting in. When I got my DX2 I thought that despite the crazily over-bloated Motoblur, I would be OK due to the Tegra 2 power and wonderfully innovative developer community. I didn't count on their being such a small number of us out there for this device or the complexity that Moto would unnecessarily bake in. So, after many months of gut wrenching disappointment, I finally have hope again.
You are a true nerd hero, sir!
Its nice to have an explanation of things and why they work or dont work. A lot of us are just trying to learn so we can be useful.
Sent from my Eclipsed and ICS themed X2.
Dragonz...after reading the info about 1/3/12 I am curious...
Does this mean that you might be able to get it working and unmount system as u need it to?
Or were there still other processes using system that you will need to mess with and change around as you did with the vendor ?
Do you still need to move the servicemanager also?
O btw imma start Hopn on the irc. I have a cm7 nightly booting thanks to second init but am having some crashes, not to mention other horrific bugs.. I need some help
Sent from my DROID X2 using xda premium
PM me with a copy of your nightly so I can help as well please
kaliblazin707 said:
O btw imma start Hopn on the irc. I have a cm7 nightly booting thanks to second init but am having some crashes, not to mention other horrific bugs.. I need some help
Sent from my DROID X2 using xda premium
Click to expand...
Click to collapse
Kali send me a pm, I have the week off and work as a mobile developer... would like to help get rid of some of those crashes. Send me a link to that irc and ill get cracking on it 2moro
Sent from my DROID X2 using xda premium
drapp said:
Kali send me a pm, I have the week off and work as a mobile developer... would like to help get rid of some of those crashes. Send me a link to that irc and ill get cracking on it 2moro
Sent from my DROID X2 using xda premium
Click to expand...
Click to collapse
Webchat.freenode.net
Channel #X2-AOSP
AND #DX2 for "off topic"
kaliblazin707 said:
O btw imma start Hopn on the irc. I have a cm7 nightly booting thanks to second init but am having some crashes, not to mention other horrific bugs.. I need some help
Sent from my DROID X2 using xda premium
Click to expand...
Click to collapse
I know why you have a cm7 nightly "booting" and I'm sorry to say that it's not for good cause. What is happening is that the system isn't being formatted properly (hence a point in my post) and you're just flashing a bunch of new APKs from CyanogenMod on top of BLUR. I'm sorry to say, but it's not CyanogenMod.
dragonzkiller said:
I know why you have a cm7 nightly "booting" and I'm sorry to say that it's not for good cause. What is happening is that the system isn't being formatted properly (hence a point in my post) and you're just flashing a bunch of new APKs from CyanogenMod on top of BLUR. I'm sorry to say, but it's not CyanogenMod.
Click to expand...
Click to collapse
I honestly think the way to go is to get back to that step with proper 2nd init and then replace everything blur with as much CM as possible
aceoyame said:
I honestly think the way to go is to get back to that step with proper 2nd init and then replace everything blur with as much CM as possible
Click to expand...
Click to collapse
But that's not the right way to go. All that is is basically taking the BLUR base ROM and replacing the things with CM7's stuff. That's basically taking the stock rom and turning it into another blur-based ROM with CM7 apks. That's basically like Eclipse.
dragonzkiller said:
But that's not the right way to go. All that is is basically taking the BLUR base ROM and replacing the things with CM7's stuff. That's basically taking the stock rom and turning it into another blur-based ROM with CM7 apks. That's basically like Eclipse.
Click to expand...
Click to collapse
Let me rephrase that, add CM stuff and then remove Moto stuff. So get back to that, delete moto apk's and then edit init.rc for the framework. All those FC's we are getting if when were in that state are vital clues as to whats isn't playing nice with CM's apk's and framework. The fact it can even somewhat boot in a state like that means it's fairly close to how it wants. If it didn't wipe /system/ then it sounds like you just need to properly 2nd init that ***** and then in moto's init.rc remove moto jar's and apks's. Remember if it didn't wipe then you replaced the framework with CM and it's blur running ontop of it.
ashclepdia said:
Webchat.freenode.net
Channel #X2-AOSP
AND #DX2 for "off topic"
Click to expand...
Click to collapse
Thanks ashclepdia.... helpful as always...
Sent from my DROID X2 using xda premium
Hi,
I'm a noob, so i'm not allowed to post in developpement forum even if believe it's the correct place for this post.
I'm following the B2G porting guide for a Galaxy s3 device. (wiki.mozilla.org/B2G/Porting)
I'm starting from the samsung rom (XXBLH3) rooted.
Here are my files :
device profile : github.com/nka11/android-device-galaxys3
B2G fork for config/build files : github.com/nka11/B2G
and you will find here some notes taken during build : gist.github.com/3826507
Right now, the system.img (file available, see gist ) builds successfully and i modified the rooted boot.img to include init.b2g.rc.
I'm flashing with heimdall successfully but it fails booting (device hangs showing Samsung Galaxy S3 GT-I9300, first screen).
As i'm a noob (can't make link nor post in good places), i'm stalled here, i don't know what i can do to figure out why it don't work.
So i'm looking for an experienced rom builder on gs3 (or a firefox os expert) to review the profile, and would like to learn some tricks which may help me to debug on my own.
Hope someone can help
Regards
still stalled
Hello,
Thanks for the silence.
I'm stalled, so changing my strategy and using the CM9.1 ROM as codebase.
I'll keey forum in touch anyway.
nka11 said:
Hello,
Thanks for the silence.
I'm stalled, so changing my strategy and using the CM9.1 ROM as codebase.
I'll keey forum in touch anyway.
Click to expand...
Click to collapse
Thanks for the silence.
Thats not the way a forum works and with that attitude you may find many developers not bothering to reply .
jje
JJEgan said:
Thanks for the silence.
Thats not the way a forum works and with that attitude you may find many developers not bothering to reply .
jje
Click to expand...
Click to collapse
Sorry, i was deep in dev and quite impatient. Just like many people here, I spend my spare time on this, so i can understand you felt insulted.
Anyway, I have a bugfix for heimdall 1.4, userdata image partition flash is now supported for GT-I9300 : github.com/Benjamin-Dobell/Heimdall/pull/56 . It may be useful for some people here.
News about porting
With CM9.1 rom :
Build successful but boot stalled at Samsung splash screen.
adb daemon active, so i can shell into the device.
After parsing dmesg, no error message found.
Going for further investigation.
I can give latest rom images on request.
Best regards
Interesting! I'll subscribe to this thread!
a new attempt
Here are the extracts from dmesg i can suspect.
around line 263 :
<3>[ 2.222980] c1 init: could not import file /system/etc/init.local.rc
<4>[ 2.225921] c1 init (1): /proc/1/oom_adj is deprecated, please use /proc/1/oom_score_adj instead.
<3>[ 2.409017] c1 init: cannot open '/initlogo.rle'
but as the kernel startup goes further, i don't know if i can suspect one of those lines. I've just commented
reference to init.local.rc in init.rc as it contains only CM specific service loading. Let's see later if we'll need
it with Firefox OS.
Later i can find these lines
<3>[ 2.643364] c2 init: cannot find '/system/bin/dmbserver', disabling 'dmb'
<3>[ 2.643446] c2 init: cannot find '/system/bin/bintvoutservice', disabling 'TvoutService_C'
but after checking in the CM9.1 backup, it appears that these files are not present in the rom.
it seems to reference proprietary binaries. I've checked in the samsung rom and found bintvoutservice but no dmbserver.
I've found the reference to these files in the init.smdk4x12.rc, so I commented the lines out for 'dmb'.
As above, it doesn't block the boot process.
I haven't noticed anything else, but something bothered me...
init.smdk4x12.rc isn't referenced in init.rc, however, init.rc content doesn't seems to be interpreted. So I moved
the import init.b2g.rc and chmod instructions to init.smdk4x12.rc.
Rebuilded the updated rom files for a new attempt.
Comportement changed : when i boot, the samsung splash screen disappears but the screen remains black and I have no access to adb...
Hard buttons seems to react as they light on when i touch them, but nothing else seems to work.
Where else could i watch ? Can someone explain (or reference a thread) init.rc/init.smdk4x12.rc on GT-I9300 ?
Any news ?
RuisuMercha said:
Any news ?
Click to expand...
Click to collapse
I remember seeing someone upload some Galaxy S III stuff for Firefox OS to Github a couple of weeks ago, but I can't find the link anywhere.
Found those links, I hope that they'll help anyone trying to port Firefox OS to the Galaxy S III.
Godspeed
https://github.com/Droid2Fox/android_device_samsung_i9300
And another attempt
https://gist.github.com/nka11/3826507
nka11 said:
Here are the extracts from dmesg i can suspect.
around line 263 :
<3>[ 2.222980] c1 init: could not import file /system/etc/init.local.rc
<4>[ 2.225921] c1 init (1): /proc/1/oom_adj is deprecated, please use /proc/1/oom_score_adj instead.
<3>[ 2.409017] c1 init: cannot open '/initlogo.rle'
but as the kernel startup goes further, i don't know if i can suspect one of those lines. I've just commented
reference to init.local.rc in init.rc as it contains only CM specific service loading. Let's see later if we'll need
it with Firefox OS.
Later i can find these lines
<3>[ 2.643364] c2 init: cannot find '/system/bin/dmbserver', disabling 'dmb'
<3>[ 2.643446] c2 init: cannot find '/system/bin/bintvoutservice', disabling 'TvoutService_C'
but after checking in the CM9.1 backup, it appears that these files are not present in the rom.
it seems to reference proprietary binaries. I've checked in the samsung rom and found bintvoutservice but no dmbserver.
I've found the reference to these files in the init.smdk4x12.rc, so I commented the lines out for 'dmb'.
As above, it doesn't block the boot process.
I haven't noticed anything else, but something bothered me...
init.smdk4x12.rc isn't referenced in init.rc, however, init.rc content doesn't seems to be interpreted. So I moved
the import init.b2g.rc and chmod instructions to init.smdk4x12.rc.
Rebuilded the updated rom files for a new attempt.
Comportement changed : when i boot, the samsung splash screen disappears but the screen remains black and I have no access to adb...
Hard buttons seems to react as they light on when i touch them, but nothing else seems to work.
Where else could i watch ? Can someone explain (or reference a thread) init.rc/init.smdk4x12.rc on GT-I9300 ?
Click to expand...
Click to collapse
I don't pretend to be a developer for one minute, but my understanding is that as far as those last two errors go, they are unsupported in CM because the code is propriety.
Sorry I can't help further - this does sound like an interesting thread - I only hope someone can give you some assistance.
Still recruiting? I know someone who can help
Sent from my Samsung Galaxy SIII GT-I9305 LTE using XDA Premium HD
I think I'm missing cyttsp_fwloader on my phone. If anyone with an sk17i, and possibly 4.0 / cm9/10 could do a;
Code:
find / -name "cyttsp_fwloader"
for me (in terminal emulator, adb or other means), post the result here, and if the file is found; upload it here!
Background
After flashing the wrong rom for my phone my touch screen stopped working. I have tried a bunch of different things to get it back up; numerous different mods/kernels etc; to no avail. For instance, also wiping the touchscreen drivers etc, even wiping the whole system directory .
Anyway, opening the file hw_config.sh I can see a switch case being run, determining my touch screen driver based on my phone model as specified by the file
/sys/devices/platform/spi_qsd.0/spi0.0/appid
Checking it manually with cat; it seems to be correct!
The two last lines calles two different binaries
cyttsp_fwloader (some arguments, one being the path to a .hex file which I believe is the touch screen driver?)
bq275xx_fwloader
Code:
find / -name "bq275xx_fwloader"
results in hit(s) (have some symlinks due to botbrew-basil)
Code:
find / -name "cyttsp_fwloader"
- nothing!
Hence, I'm led to believe that cyttsp_fwloader is missing from my system.
Who wants a cookie!?
I'm not sure how to best get the attention of someone with an rooted sk17i running cm9 or equivalent who can help me.
I think nobody knows the answer Broo....
Sent from Hubble Telescope
I will try it later and let you know
I'm on sk17i, and I've found "bq275xx_fwloader" in /system/bin. This was on jelly bean.
On my MIUI rom, I've found "bq275xx_fwloader", "cyttsp_fwloader" and "cy8ctma300_fwloader". This is on ics.
I will zip these 3 files and you can flash it through cwm. But it's strange, why should the touchscreen drivers be in bin folder, if every rom has them in bin folder?!
You can download the zip here, but you do this on your own risk, maybe it will be completely destroyed. I don't think so, but only that you know it :good:
Many thanks man! The zip makes it easier to get the stuff over, cool. I'm not sure it matters, but I had to chmod +777 the two files to get them runnable on the device. Now, it seems I was also missing another binary, nvloader if memory serves me right.
I'm trying a few different mods, found a one that gave me back these binaries; I can run hw_config.sh fine now, but still no luck getting the touch to work. So, I have to dig somewhere else; judging from this forums and users similar problems this is not really solved.
Thanks for the help though, much appreciated!
ulph said:
Many thanks man! The zip makes it easier to get the stuff over, cool. I'm not sure it matters, but I had to chmod +777 the two files to get them runnable on the device. Now, it seems I was also missing another binary, nvloader if memory serves me right.
I'm trying a few different mods, found a one that gave me back these binaries; I can run hw_config.sh fine now, but still no luck getting the touch to work. So, I have to dig somewhere else; judging from this forums and users similar problems this is not really solved.
Thanks for the help though, much appreciated!
Click to expand...
Click to collapse
Another user reported the touch would work again after flashing an xperia active rom and the rom for your device again. But I can't promise, maybe your device would brick
any progress?
i have tried flashing the active rom and recieved the following error:
24/002/2012 02:02:05 - ERROR - ERR_SEVERITY="MAJOR";ERR_CLASS="SECURITY_CLASS";ERR_STATIC="IMAGE_ADDRESS_RANGE";ERR_DYNAMIC="SIN-block eraseblock mismatch";
I have attached a screenshot of the same as well. If you come up with a guide to replace the touchscreen files via emulator then it would be of immense help. I am pretty sure it is a software problem.
Another thing i noticed when i logged into the service menu>software info
touch firmware version:
"the phone doesnt support firmware version read." <<<<<< aint this fishy? coz the touchscreen aint taking any input i guess :cyclops:
fuel gauge firmware version
0x0507 0x0507 0x0046 MANG 1
Hello guys,
i decided to try building AOSP myself using the guide at: http://developer.sonymobile.com/kno...ld-aosp-lollipop-for-unlocked-xperia-devices/
however when i let it build with
source build/envsetup.sh && lunch
and then select sgp712 (mine is wifi) and then write
make –j 6 (8 core cpu with 2 cores turned off)
it works normally for a while spamming the including ./stuff/stuff and after a minute or two ends with
#### make failed to build some targets (01:53 (mm:ss)) ####(pic attached)
Anyone has any idea what could possibly be wrong or how i would go about fixing it?
nevermind it is supposed to be -j not the long line but the short one and its building now
Ah crap! After having to reinstall the whole system because a crash ruined it i now got this after a hour of building... System folder has some stuff in it but there are no .img files made yet if that helps with what could be to blame.. Any ideas on this one? If i need to tell some info that might help anyone figure out what is making it fail i'll be happy to do that.
I tried to cook the rom on my own too, but I haven't even come up to compilation. So, it's better to ask your question here. http://forum.xda-developers.com/z4-...-xperia-devices-project-t3178610/post65310749
Or on general xda topic devoted to cooking roms. It's interesting, roms by fxp and by Sony code are the same or not?
Also, I advice you to make twrp backup of your system. Thus, you lose nothing and can experiment boldly.