[ICS] [CM9] [4.0.4] CyanogenMod 9 for the ZTE Crescent [ROM] [Last updated: 13/07]

[Guest Dazzozo]

Wow, that post was scary big and I'm worried people are completely lost right now.

Basically, we have a bunch of Wi-Fi modules.

• ZTE's (I'm assuming the one in FLB2 is the same as stock), which has the behaviour demonstrated above, memory leak when firmware_path is blank and for some reason doesn't start without insmod anymore.
  
• Fuzzra's old module, which has identical behaviour to ZTE's, but md5sum claims they're different files.
  
• Fuzzra's experimental source containing the bcm4330_b1_b2_zte driver. SPIO times out and I'm not sure if this is a dead end and honestly, I have no idea how to fix something like this.
  
• Regular DHD stuff in the kernel_zte_blade repository, which I believe builds bcm4319. That either timed out too or had some other failure behaviour that I probably pasted somewhere back in the thread that I forgot.
  

I'm starting to lose the plot with all this. There's really not much information to go on and I think I'm going around in circles. I have no idea why an update like this broke it in the first place.

[Guest tilal6991]

We really need fuzra's help on this. I'm all out of ideas.

[Guest Dazzozo]

We really need fuzra's help on this. I'm all out of ideas.

I'm going to try flashing a pre-Adreno build and see what I can get out of insmod there.

Also, I'm going to be spending time with the missus tomorrow so I probably won't be around until late. I'll make sure to catch up with anything when I get back.

Edited April 9, 2012 by Dazzozo

[Guest PsYcHoKiLLa]

I think it's time to send up the bat signal to TomG :)

[Guest Dazzozo]

Oh wow, that's interesting. I flashed a pre-Adreno build, looks like I was wrong.

lsmod reports nothing until something is loaded with insmod. I ran "insmod dhd.ko" like earlier, and...

<4>[04-09 22:28:29.290000] [1037: insmod]request success

<4>[04-09 22:28:29.290000] [1037: insmod]wlan_wake_up_gpio request success

<4>[04-09 22:28:29.290000] [1037: insmod]WLAN_REG_ON--->0

<4>[04-09 22:28:29.500000] [1037: insmod]bcm_detect_card: (ed8e0400), call

<4>[04-09 22:28:29.510000] [1037: insmod]

<4>[04-09 22:28:29.510000] [1037: insmod]Dongle Host Driver, version 5.90.125.16.1

<6>[04-09 22:28:29.580000] [12: kmmcd]mmc1: new high speed SDIO card at address0001

<4>[04-09 22:28:29.580000] [12: kmmcd]dhd_customer_oob_irq_map: customer specific Host GPIO number is (19)

<4>[04-09 22:28:29.580000] [12: kmmcd]BCM4330B2 found!

<4>[04-09 22:28:29.590000] [12: kmmcd]DHD: dongle ram size is set to 294912(orig 294912)

<4>[04-09 22:28:29.590000] [12: kmmcd]shaohua set pfn_dhd cd5770b0

<4>[04-09 22:28:29.590000] [12: kmmcd]dhdsdio_probe: failed

<4>[04-09 22:28:29.590000] [12: kmmcd]dhd_osl_detach: MEMORY LEAK 140 bytes[/code]




Memory leak. Wi-Fi was working fine before I used insmod on this build, and now it's doing the switch snapping we all now know and love.



Logcat reports:

[code]E/WifiStateMachine(  309): Failed to load driver!

E/WifiStateMachine(  309): DriverFailedState

As you would expect. Hang on...

Edit: I passed the arguments defined in BoardConfig to insmod which should load the module as it does during boot up, to no avail. Behaves identically to the current build.

Edit 2: Oh, hang on. I rebooted and now there IS an entry in lsmod. Hmm.

Edit 3: Right. When you turn the Wi-Fi switch on in Settings, that creates a lsmod entry which effectively does what can be achieved through insmod. But Settings does something else that I'm missing in the new build.

Edit 4: Would you guys hurt me if I said I think it might not actually be the Wi-Fi module that's broken?

Edited April 9, 2012 by Dazzozo

[Guest PsYcHoKiLLa]

Edit 4: Would you guys hurt me if I said I think it might not actually be the Wi-Fi module that's broken?

Quite the contrary, I might have an accident in my pants :)

[Guest raverrr]

Quite the contrary, I might have an accident in my pants :)

Androphelia....XD

watching this thread closely! Brilliant to see the progress :)

[Guest Dazzozo]

Right. I have a couple of ideas, and I'll approach this with a fresh head in around a day's time. I'm outta here, see you guys! :P

[Guest Reider59]

Take a break anyway lads. Try too hard and its impossible to spot stuff. You need a clear head and mind and maybe been on this too long without a break. I`m sure a few more days won`t hurt with the progress so far. Take your time, just hurry up

[Guest eventus]

Do I understand correctly? This ROM does not support WiFi yet. Right?

I see many people say Blade 2 and Crescent are the same model. I'm using Blade 2 ROM from GiffGaff and front camera and flash light are not working from applications. Hence, I assume Blade 2 and Crescent on hardware level two different devices and ROMs must be different.

Where, can I find original ZTE ROM for Crescent?

[Guest PsYcHoKiLLa]

The WiFi doesn't work in this build, it did in the last build. It's a work in progress. As for the codename, ZTE didn't help. At first we were told it was the Blade 2, now they've released another phone called the Blade 2. So ours is either ZTE Crescent, Orange San Francisco II, T-Mobile Vivacity or Blade S (in France).

[Guest Dazzozo]

The WiFi doesn't work in this build, it did in the last build. It's a work in progress. As for the codename, ZTE didn't help. At first we were told it was the Blade 2, now they've released another phone called the Blade 2. So ours is either ZTE Crescent, Orange San Francisco II, T-Mobile Vivacity or Blade S (in France).

This. It seems the umbrella term that describes these phones is Crescent, as Blade S and such refer to the specific models.

Edit: In other news, I return! However, I'm pretty shattered. I'll take a look at the Wi-Fi issue when I feel compelled to do anything.

Edit 2: Regarding the device name, I've been using "ZTE Crescent" in all ROMs. Seems to look the most polished when you view it in about phone and Google Play.

Edited April 10, 2012 by Dazzozo

[Guest Reider59]

It seems to vary. Some apps previously working are declared as not available/working on your phone when you visit the Market, depends on the name the Dev uses. But in reality they still do work with those apps if they were put on before the name change. To avoid the confusion it would be better if all used the name ZTE Crescent IMHO. I always thought I read their was a proper Blade2 on the way. I have seen mine termed as a Crescent in the market I`m sure. Maybe it plucks the name from the phone itself and uses that to describe it, hence apps still working.

Edited April 10, 2012 by Reider59

[Guest PsYcHoKiLLa]

I always thought I read their was a proper Blade2 on the way.

Indeed there is and it isn't ours...

http://www.ADVERT LI...Blade-II_id7013

http://www.gsmarena.com/zte_blade_ii-4608.php

Edited April 10, 2012 by PsYcHoKiLLa

[Guest Dazzozo]

Indeed there is and it isn't ours...

http://www.ADVERT LI...Blade-II_id7013

http://www.gsmarena.com/zte_blade_ii-4608.php

It pulls "Carrier Manufacturer Device" from the phone.

[Guest Dazzozo]

Right, I'm somewhat physically stable now, going to approach the Wi-Fi issue.

[Guest welshrage]

well i will certainly check forum when i arise from my grave,gotta be up for 4.30 for work,so will have quick check in before i slave away

Right, I'm somewhat physically stable now, going to approach the Wi-Fi issue.

[Guest misodoctakleidist]

It pulls "Carrier Manufacturer Device" from the phone.

That's what I always assumed.

It would be nice if we could spoof the device name for market purposes so we could download "incompatible" apps.

[Guest Dazzozo]

That's what I always assumed.

It would be nice if we could spoof the device name for market purposes so we could download "incompatible" apps.

The device name isn't relevant, it's just for show as far as I'm aware. I could call it "Carrier ZTE Waste of Time" in the ROM and the market would let you download the same.

Edited April 10, 2012 by Dazzozo

[Guest welshrage]

cant it be done through build.prop editing? or even using rom toolbox by jrummy i think (that how i done it anyways)

That's what I always assumed.

It would be nice if we could spoof the device name for market purposes so we could download "incompatible" apps.

[Guest Dazzozo]

cant it be done through build.prop editing? or even using rom toolbox by jrummy i think (that how i done it anyways)

build.prop values determine what you can download I believe, so yes.

[Guest misodoctakleidist]

cant it be done through build.prop editing? or even using rom toolbox by jrummy i think (that how i done it anyways)

I've been considering getting Rom Toolbox. Does it definitely fool the market?

Edited April 10, 2012 by misodoctakleidist

[Guest Dazzozo]

I've been considering getting Rom Toolbox. Does it definitely fool the market?

I couldn't vouch for it. A lot of blocks are due to ARMv6 and LCD density and they're mostly done for a reason.

[Guest welshrage]

dont know to be honest, i was messin around with the features and saw the build prop editor so i took a look at the build prop to try n learn something

I've been considering getting Rom Toolbox. Does it definitely fool the market?

[Guest misodoctakleidist]

I couldn't vouch for it. A lot of blocks are due to ARMv6 and LCD density and they're mostly done for a reason.

Ah. I don't suppose I'd want to get around those restrictions. I just meant for the apps that work on stock ROM, but not on some custom ROMs.

I have downloaded Rom Toolbox anyway so I'll try it at some point.