Let's fix up CM10.1 (bugs + discussion)
- Thread starter tonyp
- Start date
Okay, I committed a ****load of bluetooth patches and compiled a new (newbl) testbuild.
Works perfectly fine on my Nokia-BH504 headset while listening to mp3s - no speedsups whatsoever.
But this just means that my patches didn't break anything, as I never had any issues.
http://tonyp.hopto.org/temp/cm-10.1-20130323-UNOFFICIAL-p990.zip
Please give some feedback
If it doesn't work attach logcats again - and try to find out which A2DP profile your headset is using.
Mine is using 1.0 - pengus one (which suffers from the accellerations) has version 1.2, so it's possibly related to that.
Edit: I just got my first speedup - after listening for half an hour and pausing/resuming a couple times.
After a bluetooth disable+enable it works again for 10 minutes now. If this would be the case for everyone it would be pretty awesome!
Edit2: I think I can trigger it by pausing apollo for some time, then pressing play again.
Works perfectly fine on my Nokia-BH504 headset while listening to mp3s - no speedsups whatsoever.
But this just means that my patches didn't break anything, as I never had any issues.
http://tonyp.hopto.org/temp/cm-10.1-20130323-UNOFFICIAL-p990.zip
Please give some feedback
If it doesn't work attach logcats again - and try to find out which A2DP profile your headset is using.
Mine is using 1.0 - pengus one (which suffers from the accellerations) has version 1.2, so it's possibly related to that.
Edit: I just got my first speedup - after listening for half an hour and pausing/resuming a couple times.
After a bluetooth disable+enable it works again for 10 minutes now. If this would be the case for everyone it would be pretty awesome!
Edit2: I think I can trigger it by pausing apollo for some time, then pressing play again.
Last edited:
Hi tonyp, i dont use apollo music player and instead use google player. I also freeze apollo with titanium backup. This problem of accelerated streaming stays the same no matter which player is used. The problem is bluetooth code not the player.
Hi Tony.
You got it. T'hat's one of the many problems we got.
As I explained some time ago, for me, the best way to reproduce the problem is : The player (any player and any music) works fine just after a phone restart and some time after, the music begin to be accelerated.
I'll try to test your build and to colect some logcat i'll post tonight for analyse.
Could you add the MDM6600 spi driver in the "to-fix" list ?
Devs from the P999 forums would surely join you guys to try to finally get a fully working CM10.1 build, but without radio we are somewhat stuck without a working phone while debugging.
I know this is not easy to work on if you can't test it by yourself, but many users could try beta builds and provide logcats/kmsgs since this bug is definitely killing our device right now.
We are still trying to find out why PTS1-32 nodes are not answering to RIL calls. GPIO settings looks all right (in our sources, not CM's) and this radio driver has been used, again via a SPI bus, in other devices such as the startablet with a >2.6.39 kernel and a working JB ROM...so it should work somehow.
Devs from the P999 forums would surely join you guys to try to finally get a fully working CM10.1 build, but without radio we are somewhat stuck without a working phone while debugging.
I know this is not easy to work on if you can't test it by yourself, but many users could try beta builds and provide logcats/kmsgs since this bug is definitely killing our device right now.
We are still trying to find out why PTS1-32 nodes are not answering to RIL calls. GPIO settings looks all right (in our sources, not CM's) and this radio driver has been used, again via a SPI bus, in other devices such as the startablet with a >2.6.39 kernel and a working JB ROM...so it should work somehow.
Can this be of any help for HDMI orientation broken bug:
http://xdaforums.com/showthread.php?t=1963785
?
http://xdaforums.com/showthread.php?t=1963785
?
bluetooth fixed!
Sorry I didn't update this thread for a while, I have been terribly busy with university stuff.
I tried lots of different configs, fixes, reverts and so on.
The best result has been reverting this commit:
https://github.com/CyanogenMod/andr...mmit/ff8e103bae6deafbce4bfc2b4bedcada6f0b7752
Without this BT worked - until the phone went into deep sleep., then it sped up.
But I didn't make any further progress looking into the ICS_AUDIO_BLOB implementation and it's connection to the phone sleeping.
Soo.. screw it, I chose the last resort straw and went back to replace bluedroid with bluez.
Thanks to Andrei Emeltchenko for his work and kind help I managed to do it.
And.. *drumroll* I succeeded Bluetooth fully works with the bluez stack, confirmed by chp2 who used to get audio speedups with all my former testbuilds.
:victory: :fingers-crossed: :good: HAPPY :good: :fingers-crossed: :victory:
Finally! Lots of work has been put into this, and it's been especially annoying that my headset didn't suffer which made it hard to debug.
There is a drawback though - this will never make it into official CM, it required (partially deep) changes in ten different android packages.
As bluedroid is the future it might be worth looking further into it, but for now I'll do a break and am happy about using bluez and a fully functional bluetooth.
I need to clean everything up now. Once all testers gave a green light I will post my testbuild here to get feedback from a broader test cycle, then get a release ready.
Stay tuned - and patient.
Sorry I didn't update this thread for a while, I have been terribly busy with university stuff.
I tried lots of different configs, fixes, reverts and so on.
The best result has been reverting this commit:
https://github.com/CyanogenMod/andr...mmit/ff8e103bae6deafbce4bfc2b4bedcada6f0b7752
Without this BT worked - until the phone went into deep sleep., then it sped up.
But I didn't make any further progress looking into the ICS_AUDIO_BLOB implementation and it's connection to the phone sleeping.
Soo.. screw it, I chose the last resort straw and went back to replace bluedroid with bluez.
Thanks to Andrei Emeltchenko for his work and kind help I managed to do it.
And.. *drumroll* I succeeded
Bluetooth fully works with the bluez stack, confirmed by chp2 who used to get audio speedups with all my former testbuilds.:victory: :fingers-crossed: :good: HAPPY :good: :fingers-crossed: :victory:
Finally! Lots of work has been put into this, and it's been especially annoying that my headset didn't suffer which made it hard to debug.
There is a drawback though - this will never make it into official CM, it required (partially deep) changes in ten different android packages.
As bluedroid is the future it might be worth looking further into it, but for now I'll do a break and am happy about using bluez and a fully functional bluetooth.
I need to clean everything up now. Once all testers gave a green light I will post my testbuild here to get feedback from a broader test cycle, then get a release ready.
Stay tuned - and patient.
Last edited:
Great work then tony... and don't worry because it won't be the first time that this LG won't be upstream.
I strongly believe this CM10.1 will be the last one really possible for this phone, and your modifications won't change Cyanogen flavour at all.
Congratulations meanwhile
I strongly believe this CM10.1 will be the last one really possible for this phone, and your modifications won't change Cyanogen flavour at all.
Congratulations meanwhile
Okay, here's the build (newbl): http://tonyp.hopto.org/cm-10.1-20130410-UNOFFICIAL-p990.zip
The speed-up bug should be totally gone and everything work exactly as in CM10 (it's the same bluetooth stack).
If something isn't please report it with an attached logcat.
Development related posts are totally okay in here, for everything else you can use the CM 10.1 Bug Thread in the General section.
The speed-up bug should be totally gone and everything work exactly as in CM10 (it's the same bluetooth stack).
If something isn't please report it with an attached logcat.
Development related posts are totally okay in here, for everything else you can use the CM 10.1 Bug Thread in the General section.
For me it work fine now for all my BT devices.
Tony, Don't forget to update the 1st page if BT is OK
hell no. the only way to have this in a flashable zip is a the zip file I posted above
Sent from my LG-P990 using xda app-developers app
Good morning!
took yesterday with the rom and now goes very smoothly.
The only problem I had was with the camera this morning. When taking a picture has frozen image and saved me but to teach the rom after a classmate to teach him the problem I could do 7 shots in a row without problems.
I'm the only one with this problem?
Not if I worked for ever or will fail again the next time you need to use ...
took yesterday with the rom and now goes very smoothly.
The only problem I had was with the camera this morning. When taking a picture has frozen image and saved me but to teach the rom after a classmate to teach him the problem I could do 7 shots in a row without problems.
I'm the only one with this problem?
Not if I worked for ever or will fail again the next time you need to use ...
SoftAP fixed
Okay guys - another great news
I had some time today again and managed to fix the Wifi hotspot!
I finished the work I started 2 weeks ago (with forwardporting the CM10 netd).
Last time I missed a spot in the framework, I tracked it down, fixed it as well and booom
My testbuild (newbl):
http://tonyp.hopto.org/cm-10.1-20130415-UNOFFICIAL-p990.zip
(note that it's an engineering build, you will see flashing red borders from time to time)
pengus asked me to hold back on releasing the sources - while I've been working on forwardporting the old CM10 code into CM10.1 he's working on another approach:
Updating the kernel to the newer nvidia sources, including the bcmdhd wifi driver.
As this kernel might make my fixes obsolete it could create a mess if lots of people try to include these fixes into their rom now, as they would break the newer kernel.
I will polish everything and release a CM10.1 for both bootloaders with my bluetooth, wifi hotspot and whatever other fixes I stumble across.
Gimme a couple days to wrap everything up.
We're LG users - we know how to be patient, right?
Okay guys - another great news
I had some time today again and managed to fix the Wifi hotspot!
I finished the work I started 2 weeks ago (with forwardporting the CM10 netd).
Last time I missed a spot in the framework, I tracked it down, fixed it as well and booom
My testbuild (newbl):
http://tonyp.hopto.org/cm-10.1-20130415-UNOFFICIAL-p990.zip
(note that it's an engineering build, you will see flashing red borders from time to time)
pengus asked me to hold back on releasing the sources - while I've been working on forwardporting the old CM10 code into CM10.1 he's working on another approach:
Updating the kernel to the newer nvidia sources, including the bcmdhd wifi driver.
As this kernel might make my fixes obsolete it could create a mess if lots of people try to include these fixes into their rom now, as they would break the newer kernel.
I will polish everything and release a CM10.1 for both bootloaders with my bluetooth, wifi hotspot and whatever other fixes I stumble across.
Gimme a couple days to wrap everything up.
We're LG users - we know how to be patient, right?
Last edited:
I've been asked to create a proper CM10.1 discussion thread withour the posting restrictions like in here.
So here you go: http://xdaforums.com/showthread.php?t=2236544
At first I need to know on which basebands call rejecting works, read the other thread for more information.
So here you go: http://xdaforums.com/showthread.php?t=2236544
At first I need to know on which basebands call rejecting works, read the other thread for more information.
Hey Tony, I read on pengus kernel thread that you'll post a test built for CM10.1 with brightness bug fixed. The thread you'll post it is this one isn't it?
Thank you
Sent from my LG-P990 using xda app-developers app
Thank you
Sent from my LG-P990 using xda app-developers app
As this is the development thread I will only post development related stuff here - e.g. instructions for other developers.
You'll find my testbuilds in the discussion thread I posted above - I think that's the better approach and will keep this thread cleaner.
But as I'm posting here anyways I can link to the build as well:
http://tonyp.hopto.org/cm-10.1-20130421-UNOFFICIAL-p990.zip
Note that this build isn't meant to be used as a daily driver. This build is a development preview of the upcoming CM10.1 ROM.
For one it's an engineering build again - you will get some red flashing borders when the cpu ramps up.
Secondly it requires kk 119 to work as intended - which isn't publicly released, yet.
If you flash an earlier kernel you won't be able to change the brightness at all.
You need to keep the included kernel where everything works, but that isn't optimized performance-wise unlike the official release kernels.
If you don't touch the kernel it should fix the low brightness after unlocking bug.
Additionally I included the v30 automatic time fix properly this time, there was a fishy bug in the last version which I accidently introduced when fixing call rejecting.
Please check if everything (brightness after unlocking, automatic time, call rejecting) works and tell me if you notice any issues (in the thread in the general section!).
I'm not yet in release mode, so don't complain about anything else.
But as the most annoying bugs are fixed now you can expect a release "soon" - no ETA of course.
Last edited:
Hey ho,
I found a problem with bluetooth, my bluetooth keyboard isn't working. I'm able to pair with it but if i want to type something nothing happens.
I flashed your last build from this thread: cm-10.1-20130410 with a full wipe
only app I installed was aLogCat for the logs
and now the strange thing, my bluetooth mouse (Logitech M555b) works fine, and so does file transfer with my pc
this problem is reproducible also in your PA roms.
lg sudden
I found a problem with bluetooth, my bluetooth keyboard isn't working. I'm able to pair with it but if i want to type something nothing happens.
I flashed your last build from this thread: cm-10.1-20130410 with a full wipe
only app I installed was aLogCat for the logs
and now the strange thing, my bluetooth mouse (Logitech M555b) works fine, and so does file transfer with my pc
this problem is reproducible also in your PA roms.
lg sudden
Similar threads
- Locked
Top Liked Posts
-
There are no posts matching your filters.
-
66SoftAP fixed
Okay guys - another great news
I had some time today again and managed to fix the Wifi hotspot!
I finished the work I started 2 weeks ago (with forwardporting the CM10 netd).
Last time I missed a spot in the framework, I tracked it down, fixed it as well and booom
My testbuild (newbl):
http://tonyp.hopto.org/cm-10.1-20130415-UNOFFICIAL-p990.zip
(note that it's an engineering build, you will see flashing red borders from time to time)
pengus asked me to hold back on releasing the sources - while I've been working on forwardporting the old CM10 code into CM10.1 he's working on another approach:
Updating the kernel to the newer nvidia sources, including the bcmdhd wifi driver.
As this kernel might make my fixes obsolete it could create a mess if lots of people try to include these fixes into their rom now, as they would break the newer kernel.
I will polish everything and release a CM10.1 for both bootloaders with my bluetooth, wifi hotspot and whatever other fixes I stumble across.
Gimme a couple days to wrap everything up.
We're LG users - we know how to be patient, right?50Again: No chitchat, no "thank you" posts, no non-technical questions, no bug reports without logcats.
As the other CM10.1 threads usually are overrun by users let's try to make use of the new section and create a new bug thread in here.
Bug reports including a clear description and a logcat are welcome, every offtopic post will get deleted. Simple as that :fingers-crossed:
This thread can get used for discussion of these bugs as well - as long as it's closely development focused (e.g. "based on the logcat the problem is in package xyz, here's my idea how to fix it").
We have created a github organization where we'll push the fixes (and commit them to Gerrit if @aremcee is still interested at maintaining the P990).
https://github.com/p990-dev
If you want to help just send a pull request (or PM me about getting full access).
There are dedicated branches for the old bootloader ("cm10.1", following CM) and new bootloader ("cm10.1-newbl").
This will hopefully make it easier for devs to support both bootloaders.
Again, pull requests are always welcome.
Please don't discuss gapps related bugs (keyboard closing, gtalk video not working etc.) in here, there's a dedicated thread about non-neon gapps.47bluetooth fixed!
Sorry I didn't update this thread for a while, I have been terribly busy with university stuff.
I tried lots of different configs, fixes, reverts and so on.
The best result has been reverting this commit:
https://github.com/CyanogenMod/andr...mmit/ff8e103bae6deafbce4bfc2b4bedcada6f0b7752
Without this BT worked - until the phone went into deep sleep., then it sped up.
But I didn't make any further progress looking into the ICS_AUDIO_BLOB implementation and it's connection to the phone sleeping.
Soo.. screw it, I chose the last resort straw and went back to replace bluedroid with bluez.
Thanks to Andrei Emeltchenko for his work and kind help I managed to do it.
And.. *drumroll* I succeeded Bluetooth fully works with the bluez stack, confirmed by chp2 who used to get audio speedups with all my former testbuilds.
:victory: :fingers-crossed: :good: HAPPY :good: :fingers-crossed: :victory:
Finally! Lots of work has been put into this, and it's been especially annoying that my headset didn't suffer which made it hard to debug.
There is a drawback though - this will never make it into official CM, it required (partially deep) changes in ten different android packages.
As bluedroid is the future it might be worth looking further into it, but for now I'll do a break and am happy about using bluez and a fully functional bluetooth.
I need to clean everything up now. Once all testers gave a green light I will post my testbuild here to get feedback from a broader test cycle, then get a release ready.
Stay tuned - and patient.39Okay, here's the build (newbl): http://tonyp.hopto.org/cm-10.1-20130410-UNOFFICIAL-p990.zip
The speed-up bug should be totally gone and everything work exactly as in CM10 (it's the same bluetooth stack).
If something isn't please report it with an attached logcat.
Development related posts are totally okay in here, for everything else you can use the CM 10.1 Bug Thread in the General section.331st update:
I digged into the Wifi Hotspot issue.
One reason it doesn't work seems to be that cyanogen reverted a P990 specific patch in the netd package for merging Android 4.2 sources.
Link: https://github.com/CyanogenMod/android_system_netd/commit/d2fe72845348bb1c9e67d2cf3c648e68a93be05f
We need to bring that back. But the structure of the SoftapController.cpp changed substantially, I (or whoever beats me to it) need to look into it closer to determine who to apply the custom patch with the new format.
Additionally I just committed the bootanimation fix:
https://github.com/p990-dev/android...mmit/e298d5c00afb1eb057f1599dd8a71ba677c3c6ca
