Jump to content

Android 2.2 Development


Guest eckengucker1
 Share

Recommended Posts

Did you read the thread about that? They are trying to get the U8150 IDEOS ROM working by changing a few files around. I can honestly say that they will fail unless they start looking at building from AOSP. I'm going to be quite blunt but they don't appear to know what they are doing. A ROM for one phone cannot be made to work on another device without some serious work. At the very least you'll need a compatible kernel (which the Pulse Mini has which enables it to run a custom mod based upon FLB-Mod).

The U8150 2.2 ROM is built for an MSM7201A device running a .32 kernel. They can't just slot in a .29 kernel and expect things to work, which they've already found out.

one of the custom roms for the 845 includes OC. seems some of them can use GCC.

was just an idea. no offense. sorry.

Link to comment
Share on other sites

That's true. A real FAIL about U8150 porting.

Good notices from huawei:

http://www.huaweidevice.com/community/thre...&orderStr=9

Anyway i'm thinking to sell my U8120, I'm too tired trying to work with the ROM, adjusting, building from AOSP and all the stuff about U8120, so I preffer get another device that could runs perfectly with 2.2 (Desire or Nexus One).

Anyway, i'm not an android developer / cooker , but when you releases some information seems that you are the "coder", not, i'm only a tester :-)

Good luck about 2.2 developing on Pulse/Pulse Mini , i'm waiting it while having the device.

Greetings, DarkPlace.

Link to comment
Share on other sites

  • 2 weeks later...

Just a quick update.

Some of you already know I have been working on a port of CM6 for the Pulse. At the moment not much is working, but most of it can be fixed very quickly.

I have been concentrating on getting audio going. CM6 has the same audio problem that I had with the AOSP code. The code I have been using for my roms uses a lot of codeaurora code to fix the audio and I always believed that was the cause of several of our problems (particularly the video issues). I tried to do it a bit different this time, and have got it down to patching just 4 files (2 in libaudio, 2 in bionic headers). Some of the changes still probably aren't needed, but the good news is in CM6 I now have working audio and working video.

I should be able to push this change back to the AOSP based build and fix the problems there as well.

I will try to get it all tested and into git over the next couple of days. I will also start a new topic to track the progress of the CM6 port.

PS. I will also test 2.2.1, but it doesn't look like much has changed so I don't expect any problems.

Edited by Tom G
Link to comment
Share on other sites

Just a quick update.

Some of you already know I have been working on a port of CM6 for the Pulse. At the moment not much is working, but most of it can be fixed very quickly.

I have been concentrating on getting audio going. CM6 has the same audio problem that I had with the AOSP code. The code I have been using for my roms uses a lot of codeaurora code to fix the audio and I always believed that was the cause of several of our problems (particularly the video issues). I tried to do it a bit different this time, and have got it down to patching just 4 files (2 in libaudio, 2 in bionic headers). Some of the changes still probably aren't needed, but the good news is in CM6 I now have working audio and working video.

I should be able to push this change back to the AOSP based build and fix the problems there as well.

I will try to get it all tested and into git over the next couple of days. I will also start a new topic to track the progress of the CM6 port.

PS. I will also test 2.2.1, but it doesn't look like much has changed so I don't expect any problems.

About the Cam:

Fetch AOSP from android.git. checkout eclair. fetch vendor/htc/dream-open, vendor/htc/sapphire-open, and hardware/msm7k from ncommander's gitorious (http://gitorious.org/~ncommander). checkout origin/eclair. copy recursive libaudio* from AOSP hardware/msm7k to the new hardware/msm7k directory, replacing what is there. your fetching ncommander's source so you have a reference point for future use.

(assuming your using an x86 host and bash/ksh)

export CCOMPILER=PATH_TO_AOSP_SOURCE/prebuilt/linux-x86/toolchain/arm-eabi-4.4.0/bin/arm-eabi-

fetch 2.6.29.6 'donut' kernel source.

copy your boot.img to the root of the kernel source dir.

scripts/extract-ikconfig boot.img >.config

make ARCH=arm CROSS_COMPILE=$CCOMPILER oldconfig

make clean; make ARCH=arm CROSS_COMPILE=$CCOMPILER

make ARCH=arm CROSS_COMPILE=$CCOMPILER INSTALL_MOD_PATH=./out/modules/ modules_install

(get squashfs-tools package)

cd out/modules

mksquashfs . ../modules.sqf -noappend

cd ../../

fetch this and unpack it:

http://android-dls.com/files/linux/split_bootimg.zip

./split_bootimg.pl boot.img

cp arch/arm/boot/zImage ./

(this util and more are found after you build AOSP, within out/host/linux-x86/bin)

mkbootimg --cmdline 'no_console_suspend=1 console=null' --kernel zImage --ramdisk boot.img-ramdisk.gz -o boot-new.img

you can flash boot-new.img to boot and push modules.sqf to system/modules

Cam working!? :lol:

Edited by hungary
Link to comment
Share on other sites

I have a bug to report , am using 2.2 with Jit , OC , Swap , it is stable and fast no reboots so far. . . But there is something wrong with the rotator , sometimes it go to landscape position and it stays like that until i reboot the phone , so i disabled the autorotation future and it is good , but this is a bug and it needs to be reviewed

Link to comment
Share on other sites

I have a bug to report , am using 2.2 with Jit , OC , Swap , it is stable and fast no reboots so far. . . But there is something wrong with the rotator , sometimes it go to landscape position and it stays like that until i reboot the phone , so i disabled the autorotation future and it is good , but this is a bug and it needs to be reviewed

Yeah, Tom G knows about this. Not entirely sure why it's happening :lol:

Link to comment
Share on other sites

Yeah, Tom G knows about this. Not entirely sure why it's happening :lol:

And it could be difficult to fix considering we don't have source for the sensors lib. I haven't been able to recreate the problem on my pulse, so unless someone can find a reliable way to recreate the problem I don't think I can do much.

Link to comment
Share on other sites

About the Cam:

...

Cam working!? :lol:

Is that a tested & working solution, or just a suggestion based on other devices?

Cam will be the next challenge, the video problems would have been contributing to the cam problems (our main problem is the driver which is unrelated to the video problem, but with a working driver we wouldn't be able to test it until video worked).

Link to comment
Share on other sites

Is that a tested & working solution, or just a suggestion based on other devices?

Cam will be the next challenge, the video problems would have been contributing to the cam problems (our main problem is the driver which is unrelated to the video problem, but with a working driver we wouldn't be able to test it until video worked).

So now on your CM6 port the only thing not working is the cam? If so that's amazing, great work :lol: Does the 3D Gallery work like it should, is the rotation proper? Cheers!

Link to comment
Share on other sites

So now on your CM6 port the only thing not working is the cam? If so that's amazing, great work :lol: Does the 3D Gallery work like it should, is the rotation proper? Cheers!

There are a lot of things not working (because I haven't set them up yet), but they will be easy to fix. I am just focusing on the things that caused the most problems previously.

The main reason for trying CM6 was that we were looking at porting code from CM6 to fix the cam, so why not try porting all of CM6.

The GPS reboot problem is still happening.

Link to comment
Share on other sites

There are a lot of things not working (because I haven't set them up yet), but they will be easy to fix. I am just focusing on the things that caused the most problems previously.

The main reason for trying CM6 was that we were looking at porting code from CM6 to fix the cam, so why not try porting all of CM6.

The GPS reboot problem is still happening.

Cool, so you're saying that the thing that used to work will after you set them up, plus video. So still no cam and still GPS issues. To be honest for me the video was the deal breaker for going to Froyo, I'll patiently wait till you release something new to the masses :lol:

Also, by porting CM will we get all the extra goodies they have, like the possibility to change text colors in notification bar and such through Spare Parts?

Link to comment
Share on other sites

Cool, so you're saying that the thing that used to work will after you set them up, plus video. So still no cam and still GPS issues. To be honest for me the video was the deal breaker for going to Froyo, I'll patiently wait till you release something new to the masses :lol:

Also, by porting CM will we get all the extra goodies they have, like the possibility to change text colors in notification bar and such through Spare Parts?

I don't know much about CM, but I think most of the extra bits should work as it looks like a lot of it is changes within android and should not be device specific. As long as the device has the hardware features needed and appropriate drivers, then the software features should work. As an example the turn over to silence ring feature should work since the pulse has the necessary sensors.

Froyo tethering still won't work. I think we will need a .32 kernel and USB gadget support for USB tethering, and for wifi tethering we will need a wifi driver that can run as an AP (I don't think ours supports that). I was never able get the ar6k source from the tegra or openmoko to work (I didn't try very hard), but I think I've found a better place to get source, so we might be able to build the wifi driver from source soon (which is important because at the moment we can't use wifi on a .32 kernel).

For those interested, ChromeOS has support for AR6k (and has had for a while) and it is tested and working on MSM hardware. I haven't looked at it yet, I found it while looking for some other stuff.

http://codereview.chromium.org/646055

http://git.chromium.org/cgi-bin/gitweb.cgi.../ath6kl;hb=HEAD

Link to comment
Share on other sites

Tom, you may be right that the video playback problem may come from mix aosp and codeaurora. when tracking down the situation it seems that it is at a higher level, msm copybit, msm gralloc and the kernel hw accelerated blitter are fine, and they seem to be compatible with each other. so the problem is at higher level, but at higher level it is so complicated, that I couldn't track down the whole situation.

(btw replacing all the video-related libraries with Ideos ones also didn't help, so the incompatibility might be between androdid and its video libraries, but its not worth to debug if cm6 plays video well.)

Link to comment
Share on other sites

I can't test with a sim at the moment (the phone crashes if I boot it with a sim in and the FT232R board connected), but it looks like the audio/video changes may have also fixed the netcfg problem.

I noticed what looks like a bug. Most (but not all) of the ringtones appear multiple times in the list. I assume that also happened in my older froyo roms (I had never looked before).

Here is a rom with the changes I mentioned yesterday. Video now works.

Froyo-Pulse-0.60

I will get the changes into git tomorrow. Let me know if you find any new bugs (or if you find that old ones are fixed).

PS. This is 2.2.1/FRG83.

Edited by Tom G
Link to comment
Share on other sites

I can't test with a sim at the moment (the phone crashes if I boot it with a sim in and the FT232R board connected), but it looks like the audio/video changes may have also fixed the netcfg problem.

PS. This is 2.2.1/FRG83.

Boots fine with network :lol:

post-656233-1285947117_thumb.png post-656233-1285947209_thumb.png

EDIT: Wifi doesn't work, can't connect. Network disconnects briefly.

Edited by ogiogi
Link to comment
Share on other sites

I noticed what looks like a bug. Most (but not all) of the ringtones appear multiple times in the list. I assume that also happened in my older froyo roms (I had never looked before).

This isn't the case for me, all ringtones appear only once :lol:

However, ringtones that are stored on the sdcard under /media/audio/ringtones now appear in the list, this never worked for me before in 2.1

Edited by Schwertfish
Link to comment
Share on other sites

Boots fine with network :lol:

post-656233-1285947117_thumb.png post-656233-1285947209_thumb.png

EDIT: Wifi doesn't work, can't connect. Network disconnects briefly.

What do the logs say when you try to enable wifi?

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share


×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.