• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

Root [DEV] Continuing Triumph ICS Development

Whyzor said:
Actually on the service wpa_supplicant line, I just noticed the convention is to NOT have a space in between. So either consistently have no spaces or have all spaces.
Click to expand...

So will what you put with the spaces is fine or will that cause a problem
 
mantera said:
Yumbrad's changes have been pushed up. Do a repo sync and you should be good to go.

IMO, the following needs to get fixed before an alpha release:

The web browser doesn't recognize touches.
The keyboard is still messed up.

Even with data, without the above, it's not very usable. Although, the market works fine. :)

Overall, data seems slower than with gb but it could just be a symptom of the other issues. working is better than not though.

Also, I ran a quadrant just to see what it would be like. I got a score of about 2100 at stock. So it's not the greatest but I think once we finish getting things working, and optimizing stuff, it will probably get better.
Click to expand...

Have you tried an alternate browser?
And I have been using the slideIt keyboard and it has been working fine.
Sometimes it comes up missing all the letters but if I close it and re open it its fine.
 
mantera said:
I tried dolphin hd and same crap.
Click to expand...

Ok good that means its not a problem with the browser code its a problem with the touchscreen driver......

Does the touchscreen driver have different modes that it enters?
Like an "app mode" and when the screen is off "mode" because it works okay on the homescreen but it seems when apps try to utlize it, it freaks out.
 
I think i pulled all of your changes but same problem

D/Beautiful Widgets( 1162): WIFI_STATE_CHANGED_ACTION
E/WifiStateMachine( 235): Failed to load driver!
E/WifiStateMachine( 235): DriverFailedState
D/Beautiful Widgets( 1162): WIFI_STATE_CHANGED_ACTION
 
konspiracy said:
Okay so pinch to zoom works on the browser so it seems we have a swiping problem/scrolling problem...
Click to expand...

If you install the "Multitouch Vis Test" app it'll show whether the touch input data is read correctly or not. I suspect there's broken code in CM9 still (it's not released yet).

So will what you put with the spaces is fine or will that cause a problem
Click to expand...
I would try no spaces first since that's what it was before, but that seems unconventional. Then with spaces in another build if it still doesn't work.
 
Whyzor said:
Code: 
- setprop wifi.interface wlan0
+ setprop wifi.interface "wlan0"
+ setprop wlan.driver.status "ok"
Click to expand...

wlan.driver.status is a runtime, dynamically set property rather than a configuration option - may cause problems to hardcode it as it should cycle through "loading," "ok" .. and there is code that checks those states in order to do the right thing at the right time.

I tried cherry picking Isaac's wifi changes (even though we should probably make use of the vendor specific makefile config "TwistedUmbrella" added in early January to avoid changing the base wifi.c). These changes are necessary, or at least the spirit of them are - they didn't just get things working. Someone who wants to investigate might grab Isaac's changes and then start tracing why the driver failed - I started with a grep through the sources to find where the error is issued:

Code: 
frameworks/base/wifi/java/android/net/wifi/WifiStateMachine.java:   loge("Failed to load driver!");

But I won't be able to look at anything for a while, so hopefully when I get back I can tackle some other problem because wifi's working :)

Oh, and one more thing - graphics seem a bit snappier with the change I just sent mantera on github to enable gpu overlay, but it could just be my imagination.
 
yumbrad said:
wlan.driver.status is a runtime, dynamically set property rather than a configuration option - may cause problems to hardcode it as it should cycle through "loading," "ok" .. and there is code that checks those states in order to do the right thing at the right time.

I tried cherry picking Isaac's wifi changes (even though we should probably make use of the vendor specific makefile config "TwistedUmbrella" added in early January to avoid changing the base wifi.c). These changes are necessary, or at least the spirit of them are - they didn't just get things working. Someone who wants to investigate might grab Isaac's changes and then start tracing why the driver failed - I started with a grep through the sources to find where the error is issued:

Code: 
frameworks/base/wifi/java/android/net/wifi/WifiStateMachine.java:   loge("Failed to load driver!");

But I won't be able to look at anything for a while, so hopefully when I get back I can tackle some other problem because wifi's working :)

Oh, and one more thing - graphics seem a bit snappier with the change I just sent mantera on github to enable gpu overlay, but it could just be my imagination.
Click to expand...

you sound like you made android! :D (thats a compliment lol)

so you said wifi is working? :confused: cause what you said was kind of confusing lol

But I won't be able to look at anything for a while, so hopefully when I get back I can tackle some other problem because wifi's working :)
Click to expand...
 
Ok I guess after boiling things down further, this is the important line that someone should try in:

files/etc/wifi/wpa_supplicant.conf

Code: 
- ctrl_interface=DIR=/data/misc/wifi/wpa_supplicant GROUP=wifi
+ ctrl_interface=wlan0 DIR=/data/misc/wifi/wpa_supplicant GROUP=wifi

On the xda ideos x6 thread, new kernel 3.0 source code for MSM8255 by Huawei was posted, maybe someone will port this over eventually.

https://www.codeaurora.org/gitweb/quic/la/?p=kernel/msm.git;a=shortlog;h=refs/heads/msm-3.0
 
Sorry Whyzor, you were right about wlan.driver.status ... It is dynamic, but *we* need to set it to inform wpa_supplicant about the kernel driver being loaded. That link konspiracy sent is a really helpful overview, I should have looked at it like you did rather than go on my memory of what I saw late last night :) ! Errr, edit... on second read, we only need to set it if the driver is built into the kernel - since we have a module, wifi.c should set it after all :)

yumbrad said:
wlan.driver.status is a runtime, dynamically set property rather than a configuration option - may cause problems to hardcode it as it should cycle through "loading," "ok" .. and there is code that checks those states in order to do the right thing at the right time.
Click to expand...
 
yumbrad said:
Sorry Whyzor, you were right about wlan.driver.status ... It is dynamic, but *we* need to set it to inform wpa_supplicant about the kernel driver being loaded. That link konspiracy sent is a really helpful overview, I should have looked at it like you did rather than go on my memory of what I saw late last night :) ! Errr, edit... on second read, we only need to set it if the driver is built into the kernel - since we have a module, wifi.c should set it after all :)
Click to expand...


I try to help y'all out as much as possible, because I'm still learning how to apply the info they are giving me into my build environment.

So any info. I can dig up on the net I will post here. Im pretty good at digging around on the net. The thing that bugs me the most is when its in a different language and the translation is really really crude.

The good thing is I can look at the fixes yall have used and learn from it.
 
Kudos to yumbrad for getting 3g working. Thank you Whyzor also for helping, and mantera for merging the file commits.

I plan to help some with the wifi and touchscreen drivers a little tomorrow.

Also, yumbrad, I found a function in IccUtils that converts bytes to a hex string: IccUtils.bytesToHexString(byte arr[])

It's doesn't matter much, but it could save a little time when debugging code.
 
Failed to build for me... some borked file I'd imagine. Posted in the ICS build thread with the fail from terminal. If anyone could take a quick look and see what it might be that would be a great help. I was able to add whyzor's additions, but no build. Anyhow, thanks guys, still looking forward to it.
 
I tried to pull from Mantera's source, is the repo down for anyone else? Keep on getting this:
Code: 
Receiving objects: 100% (719/719), 1.33 MiB | 1.48 MiB/s, done.
Resolving deltas: 100% (451/451), done.
From https://github.com/CyanogenMod/android_external_libpng
 * [new branch]      froyo      -> github/froyo
 * [new branch]      froyo-stable -> github/froyo-stable
 * [new branch]      gingerbread -> github/gingerbread
 * [new branch]      gingerbread-release -> github/gingerbread-release
 * [new branch]      ics        -> github/ics
 * [new tag]         cm-7.0.0   -> cm-7.0.0
 * [new tag]         cm-7.0.1   -> cm-7.0.1
 * [new tag]         cm-7.0.2.1 -> cm-7.0.2.1
 * [new tag]         cm-7.0.3   -> cm-7.0.3
 * [new tag]         cm-7.1.0   -> cm-7.1.0
Fetching projects:  32% (75/233)  
error: Exited sync due to fetch errors
root@tusing-ubuntu:~/android#
Used repo sync -j4 on a 4 core virtual machine.
 
I just wanna say thanks to everybody involved in this project. you guys are amazing at what you do. I just have one question, what all isn't working so far? I'm not gonna ask for an E.T.A. on when it'll be done, hell I don't Even have a triumph yet.lol. I'm gonna pick one up from best buy tomorrow though. the old optimus V is on her last leg.
 
You must log in or register to reply here.
Back
Top Bottom