[DEV] AOSP Dev for Zte Blade

[Guest davidnintendo]

One quick question: Is it possible to turn off surface dithering and to use 16 bit transparency on ICS? And would it increase responsiveness of the UI on the Blade? Perhaps it's worth the try to see if it works better. In Gingerbread there is a huge difference between using 16bit transparency and the default one, specially when scrolling.

[Guest fonix232]

One quick question: Is it possible to turn off surface dithering and to use 16 bit transparency on ICS? And would it increase responsiveness of the UI on the Blade? Perhaps it's worth the try to see if it works better. In Gingerbread there is a huge difference between using 16bit transparency and the default one, specially when scrolling.

First we need to get HW accel working :D

[Guest ThermalDroid]

One quick question: Is it possible to turn off surface dithering and to use 16 bit transparency on ICS? And would it increase responsiveness of the UI on the Blade? Perhaps it's worth the try to see if it works better. In Gingerbread there is a huge difference between using 16bit transparency and the default one, specially when scrolling.

The fact that it allows the UI to be drawn by the GPU will probably be the main focus of the DEVs.

[Guest jonas.lindberg]

Of course, per request:

http://www.mediafire.com/?8u84b76ngwta2nn

Changes to Skywave's Build #10:

- Removed OpenWnn

- Added Google Apps (Gingerbread version but works)

- Added RIL V3 disabling

Thanks for your effort, though it does nothing but stays at the boot screen for me. Any thoughts on what to do?

Sorry for disturbing the thread with these kind of questions, but would love to take my first bite of that tasty ice cream sandwich.

[Guest fonix232]

Thanks for your effort, though it does nothing but stays at the boot screen for me. Any thoughts on what to do?

Sorry for disturbing the thread with these kind of questions, but would love to take my first bite of that tasty ice cream sandwich.

As I asked not long before, provide more info about your device. Then we can figure out the cause of the problem.

[Guest evilbrain]

Thanks for your effort, though it does nothing but stays at the boot screen for me. Any thoughts on what to do?

Sorry for disturbing the thread with these kind of questions, but would love to take my first bite of that tasty ice cream sandwich.

same for me here. Fortunately, I've made a backup of my existing rom.

It just kept on booting for somewhere about 10 minutes when I got bored and did a restore

Edit: I think from fonix232's description my blade comes along the fourth wave, if that is any help to you

Edited December 4, 2011 by evilbrain

[Guest fonix232]

same for me here. Fortunately, I've made a backup of my existing rom.

It just kept on booting for somewhere about 10 minutes when I got bored and did a restore

Edit: I think from fonix232's description my blade comes along the fourth wave, if that is any help to you

Weird thing is that mine came with wave4 too, and it boots, works perfectly, etc.

[Guest evilbrain]

Weird thing is that mine came with wave4 too, and it boots, works perfectly, etc.

what did you use for flashing the thing?

[Guest fonix232]

what did you use for flashing the thing?

The gen1-2 mix v9 (160MB system), plus it's 5.0.2.0 ClockWorkMod.

[Guest evilbrain]

The gen1-2 mix v9 (160MB system), plus it's 5.0.2.0 ClockWorkMod.

strange, about the same as I have used. but while flashing the progress bar was only grey and in the end i got some message I dont remember now

[Guest skywave]

I tinkered a little with the permissions ril got, now the same as CM7. Also deleted per fonix532 suggestion the wnn apks

http://www.mediafire.com/?thtxki9hjg8ftob

[Guest sej7278]

The answer is in the hardware differences - the Blade has more generations, and not the NAND partitioning one (it is possible that I missed a generation, please fix me up if I did):

nah i don't buy that - there's something not right in the kernel config i guess as the cm7 n255 its based on works fine.

is the kernel blob on github not actually compiled with the synaptics fix - do we have to rebuild it? i saw something about splitting it from one of the attached boot.img files, so is that not the same file that's on github?

i see skywave just uploaded the debuggerd fix from cm9 thanks, ics4blade compiles again now.

edit: ah i see the kernel_ics4blade.config from github is not the same as the one used to create the kernel blob on github (lol, its like zte here!) so do we need to compile our own kernel (from tom_g's?) and not use the blog in github?

diff kernel_ics4blade.config ~/extracted.config |grep CONFIG

< # CONFIG_ZTE_FTM_FLAG_SUPPORT is not set

> CONFIG_ZTE_FTM_FLAG_SUPPORT=y

< CONFIG_SENSORS_AKM8973=y

> # CONFIG_SENSORS_AKM8973 is not set

< # CONFIG_TUN is not set

> CONFIG_TUN=y

< # CONFIG_PCMCIA_RAYCS is not set

> CONFIG_PCMCIA_RAYCS=y

< CONFIG_LIBRA_SDIOIF=y

> # CONFIG_LIBRA_SDIOIF is not set

< # CONFIG_TOUCHSCREEN_FOCALTECH is not set

> CONFIG_TOUCHSCREEN_FOCALTECH=y

< # CONFIG_IR_CORE is not set

> CONFIG_IR_CORE=y

> CONFIG_VIDEO_IR=y

> CONFIG_RC_MAP=y

> CONFIG_IR_NEC_DECODER=y

> CONFIG_IR_RC5_DECODER=y

> CONFIG_IR_RC6_DECODER=y

> CONFIG_IR_JVC_DECODER=y

> CONFIG_IR_SONY_DECODER=y

> # CONFIG_IR_IMON is not set

> CONFIG_VIDEO_IR_I2C=y

< CONFIG_MT9T013=y

> # CONFIG_MT9T013 is not set

< # CONFIG_QUP_EXCLUSIVE_TO_CAMERA is not set

> CONFIG_QUP_EXCLUSIVE_TO_CAMERA=y

< CONFIG_RADIO_ADAPTERS=y

< # CONFIG_I2C_SI4713 is not set

< # CONFIG_RADIO_SI4713 is not set

< CONFIG_RADIO_SI470X=y

< # CONFIG_I2C_SI470X is not set

< # CONFIG_RADIO_TEA5764 is not set

< # CONFIG_RADIO_SAA7706H is not set

< # CONFIG_RADIO_TEF6862 is not set

> # CONFIG_RADIO_ADAPTERS is not set

< CONFIG_BCM_WIFI=y

> # CONFIG_BCM_WIFI is not set

< # CONFIG_LED_ON_AFTER_SUSPEND is not set

> CONFIG_LED_ON_AFTER_SUSPEND=y

< CONFIG_ZTE_FIX_ALARM_SYNC=y

> # CONFIG_ZTE_FIX_ALARM_SYNC is not set

< CONFIG_IIO=y

< # CONFIG_IIO_RING_BUFFER is not set

< # CONFIG_IIO_TRIGGER is not set

< # CONFIG_MAX1363 is not set

< # CONFIG_SENSORS_TSL2563 is not set

> # CONFIG_IIO is not set

< # CONFIG_ZRAM_DEBUG is not set

Edited December 4, 2011 by sej7278

[Guest skywave]

Correct, kernel builds without touchscreen.

[Guest OneB1t]

i have only bootloop with update 10 :-/ on san francisco gen1 512mb ram tft @ gen2 tpt

[Guest fonix232]

I tinkered a little with the permissions ril got, now the same as CM7. Also deleted per fonix532 suggestion the wnn apks

http://www.mediafire...thtxki9hjg8ftob

So this one should fix RIL PIN loops?

OpenWnn should be disabled in the build bases - in blade.mk, you call some base.mk and device.mk makefiles, edit those (you should edit base.mk), and remove the OpenWnn, etc lines.

Edited December 4, 2011 by fonix232

[Guest birdibird]

So this one should fix RIL PIN loops?

Nop, not on mine. Still bootloops

(OSF UK OLED)

[Guest fonix232]

Nop, not on mine. Still bootloops

(OSF UK OLED)

/* facepalm */

I'm not talking about bootloops, but a little bug, where the system would always pop the PIN request window up.

[Guest skywave]

So this one should fix RIL PIN loops?

OpenWnn should be disabled in the build bases - in blade.mk, you call some base.mk and device.mk makefiles, edit those (you should edit base.mk), and remove the OpenWnn, etc lines.

Could fix ril.

Turned out the prebuild kernel was (probably) never updated, just commited to git a version found in t0mm13b's boot.img

[Guest sej7278]

Correct, kernel builds without touchscreen.

ok thanks. i tried to build the kernel using tom_g's github repo that cm7 is based on, and the kernel config from ics4blade's github but it doesn't work, seems to be having problems with the new stuff tom_g put in since the kernel that was in n255 that ics4blade is based on:

  LD  	.tmp_vmlinux1

drivers/built-in.o: In function `libra_sdio_suspend':

/data3/useful_software/android/sdk/kernel_zte_blade/drivers/net/wireless/libra/libra_sdioif.c:302: undefined reference to `sdio_set_host_pm_flags'

/data3/useful_software/android/sdk/kernel_zte_blade/drivers/net/wireless/libra/libra_sdioif.c:313: undefined reference to `sdio_set_host_pm_flags'

make: *** [.tmp_vmlinux1] Error 1

the zImage builds but none of the modules.

[Guest t0mm13b]

@fonix232: The poster birdibird is talking about where the services die, bootloops, which I am experiencing too so no need for facepalms! FelixL,is another user experiencing same regardless if java's RIL layer is patched or not.

Edited December 4, 2011 by t0mm13b

[Guest fp30]

nah i don't buy that - there's something not right in the kernel config i guess as the cm7 n255 its based on works fine.

is the kernel blob on github not actually compiled with the synaptics fix - do we have to rebuild it? i saw something about splitting it from one of the attached boot.img files, so is that not the same file that's on github?

i see skywave just uploaded the debuggerd fix from cm9 thanks, ics4blade compiles again now.

edit: ah i see the kernel_ics4blade.config from github is not the same as the one used to create the kernel blob on github (lol, its like zte here!) so do we need to compile our own kernel (from tom_g's?) and not use the blog in github?

I thought people were just iusing t0mm13b's boot.img from an earlier post. that had working touchscreen and RIL. I guess if you want to compile your own you'll have to make changes to the source.

[Guest sej7278]

I thought people were just iusing t0mm13b's boot.img from an earlier post. that had working touchscreen and RIL. I guess if you want to compile your own you'll have to make changes to the source.

but that should be in github otherwise its duplication of effort.

[Guest JegZ]

gen1 TFT here with update10 works fine. no bootloops. only the prob with calls without sound, so i suppose the gen is not the prob

[Guest JimJam707]

If the Nexus One has the same GPU as the Blade, and the Desire == Nexus One, then surely this would help with Hardware Acceleration?

[Guest leopesto]

update10b --- osf tft 512mb gen1 with tpt gen2

if I put the pin code the phone will restart after some seconds... if I don't the phone doesn't restart....

I attache the logcat (start → pin → restart → nothing)

logcat.txt