Jump to content

Advent Vega kernel source code now available!


Guest PaulOBrien

Recommended Posts

I think that if i get it right, it will be better than backporting 3.0 filters to 2.6.36 filters ... ;)

..

Still have to figure out the hwcomposer thing. I assume they changed some structs from HC to ICS. It should be possible to check the change history of the associated header to be sure ;)

Good to know, ill stop trying to port the filter :)

Which header do you refer to for hwcomposer? 3.2 -> 4.0 changes in Android side or the Nvidia stuff in kernel?

Cheers

Cass

Link to comment
Share on other sites

Guest ejtagle

Good to know, ill stop trying to port the filter :)

Which header do you refer to for hwcomposer? 3.2 -> 4.0 changes in Android side or the Nvidia stuff in kernel?

Cheers

Cass

Changes in the Android side. the libhardware and libhardware_legacy projects. One of them has the header that defines the structures required to implement the hwcomposer module. If the memory layout of those structs didn't change, i think the hwcomposer should be working...

I am still porting to the 3.0 kernel.. This time, the porting process is completely different to the 2.6.36, as we are adding support for the tegra arch to a kernel that does not fully support it... the omap panda kernel ;)

Link to comment
Share on other sites

Changes in the Android side. the libhardware and libhardware_legacy projects. One of them has the header that defines the structures required to implement the hwcomposer module. If the memory layout of those structs didn't change, i think the hwcomposer should be working...

Ill have a look to those later tonight and see if anything stands out .. hopefully it wont be something major ..

Cheers

Cass

Link to comment
Share on other sites

Hi guys,

I keep getting questions about sound problems. I never had any trouble with it but is there a way to fix it? Problems reported are: Sound disturtion etc.

Keep up the good work!

Kind Regards,

Areo

Link to comment
Share on other sites

Guest ejtagle

Hi guys,

I keep getting questions about sound problems. I never had any trouble with it but is there a way to fix it? Problems reported are: Sound disturtion etc.

Keep up the good work!

Kind Regards,

Areo

Please, disregard sound problems. Sound took a huge amount of time to get it right, and actual measurements on real devices using professional measuring equipment. I am absolutely sure there are no distorsions there. What we have here is people that want a louder volume, or that have blown speakers ... We should not/can't fix those problems ;)

Link to comment
Share on other sites

Hi,

I want to say big thank you to Cass, Eduardo and Wooshy for incredible work. Today I have compiled the kernel with patches by Eduardo...and it's amazing, working flawless! My Vega is resurrected. :) Thank you guys for sharing your knowlegde and work with community.

Link to comment
Share on other sites

Ill have a look to those later tonight and see if anything stands out .. hopefully it wont be something major ..

Re:- libharware .. i took a look tonight at the headers and they look to be the same as the references i could find ..

Checked

EGL/egl.h << included in the hwcomposer stub

include/hardware/hwcomposer.h << included in the hwcomposer stub

modules/hwcomposer/hwcomposer.cpp << the stub

include/hardware/gralloc.h << related

modules/gralloc/gralloc.cpp << related

modules/gralloc/gr. h << related to gralloc

These files are al the same as the reference ones i could find from > 1 year ago ... surfaceflinger is next i guess ..

This issue has to be something stupid ...

Cass

Link to comment
Share on other sites

Guest ejtagle

Re:- libharware .. i took a look tonight at the headers and they look to be the same as the references i could find ..

Checked

EGL/egl.h << included in the hwcomposer stub

include/hardware/hwcomposer.h << included in the hwcomposer stub

modules/hwcomposer/hwcomposer.cpp << the stub

include/hardware/gralloc.h << related

modules/gralloc/gralloc.cpp << related

modules/gralloc/gr. h << related to gralloc

These files are al the same as the reference ones i could find from > 1 year ago ... surfaceflinger is next i guess ..

This issue has to be something stupid ...

Cass

I also think so... Perhaps, we could try to spot the crash we are having... the logcat says exactly the place, now we must find why :) --

Link to comment
Share on other sites

I also think so... Perhaps, we could try to spot the crash we are having... the logcat says exactly the place, now we must find why :) --

yeah ill try to get an strace up on boot to catch it ...

Yeah strace told me to look elsewhere .. so i did ... i looked to the adam build i have here .. and tadaa ... something is weird with my build it seems ... grrrrrrrrrrrrrrrr and did i also say .. grrrrrrrrrrrrrrrrrrrr ahem ... grr :)

just hear this adam build is based from an aosp meant for streak :)

guess i need to check the diffs between our nv file and this one ... or forget it and move on .. hmm what to do :)

green screen only on layer, turns green when we hit power button to shut down .. progress still :)

IMAG0167.jpg

Edited by Cass67
Link to comment
Share on other sites

Guest ejtagle

yeah ill try to get an strace up on boot to catch it ...

Yeah strace told me to look elsewhere .. so i did ... i looked to the adam build i have here .. and tadaa ... something is weird with my build it seems ... grrrrrrrrrrrrrrrr and did i also say .. grrrrrrrrrrrrrrrrrrrr ahem ... grr :)

just hear this adam build is based from an aosp meant for streak :)

guess i need to check the diffs between our nv file and this one ... or forget it and move on .. hmm what to do :)

green screen only on layer, turns green when we hit power button to shut down .. progress still :)

IMAG0167.jpg

Yes, there must be something very silly going on ;) -- I am still at the kernel porting ... ;)

Link to comment
Share on other sites

Yes, there must be something very silly going on ;) -- I am still at the kernel porting ... ;)

take your time .. no rush .. sleep is good also :)

ill keep this as a good base and over the next few days work out what is dumb that is missing ..

Link to comment
Share on other sites

Guest mg.rionline

hi there,

do not wanna distirb all your ics efforts, but could someone inform me, if there is already an installable kernel with microphone fixed? i have seen some note on rebel1s github changlog. i would be great, if there were a fully functional kernel for honeycomb. without micro there is almost no need for camera. two major things in my opinion. would be great if someone could give me a heads up on this.

thanks for all your hard work.

bb and greets from germany

Link to comment
Share on other sites

hi there,

do not wanna distirb all your ics efforts, but could someone inform me, if there is already an installable kernel with microphone fixed? i have seen some note on rebel1s github changlog. i would be great, if there were a fully functional kernel for honeycomb. without micro there is almost no need for camera. two major things in my opinion. would be great if someone could give me a heads up on this.

thanks for all your hard work.

bb and greets from germany

Hi,

Post a link to the notes, I don't see them (could be me being blind).

If it is really fixed I will compile a new kernel and post it soon.

Areo

Link to comment
Share on other sites

Guest ejtagle

Hi,

Post a link to the notes, I don't see them (could be me being blind).

If it is really fixed I will compile a new kernel and post it soon.

Areo

mic works since a long time ago. VC3.2 update3 already contains the fixes. The fact that the mic input is noisy and sometimes is not sensitive enough is just a result of a low quality hw design... People, why don't you try to test with a mic test program (available for free from the market, not skype!) , talk closer to the mic,, and see for yourselves... ?

The only chance to improve things here is to make ICS work, as ICS has a software implementation of noise cancelling for the mic... HC does not have it.

Edited by ejtagle
Link to comment
Share on other sites

Guest ejtagle

Hi masterminds,

Quick question again: Which files should I look to regarding uncapping the fps?

Kind Regards,

Areo

No idea... fps perhaps could be capped to the display refresh rate inside the nvidia propietary libraries. It makes no sense to go faster than that ;)

Link to comment
Share on other sites

No idea... fps perhaps could be capped to the display refresh rate inside the nvidia propietary libraries. It makes no sense to go faster than that ;)

Ok, so where are you guys now with ICS progress? I know still a long list has to be done, but could you make a list of what's already done?

I will try to help but I am not the best with kernel, but doesn't hurt to try again.

Areo

Link to comment
Share on other sites

Guest mg.rionline

mic works since a long time ago. VC3.2 update3 already contains the fixes. The fact that the mic input is noisy and sometimes is not sensitive enough is just a result of a low quality hw design... People, why don't you try to test with a mic test program (available for free from the market, not skype!) , talk closer to the mic,, and see for yourselves... ?

The only chance to improve things here is to make ICS work, as ICS has a software implementation of noise cancelling for the mic... HC does not have it.

hmm. thanks for the heads up. i was just wondering, because on stock it works far better...

Edited by mg.rionline
Link to comment
Share on other sites

Guest richardmlea

mic works since a long time ago. VC3.2 update3 already contains the fixes. The fact that the mic input is noisy and sometimes is not sensitive enough is just a result of a low quality hw design... People, why don't you try to test with a mic test program (available for free from the market, not skype!) , talk closer to the mic,, and see for yourselves... ?

The only chance to improve things here is to make ICS work, as ICS has a software implementation of noise cancelling for the mic... HC does not have it.

hmm. thanks for the heads up. i was just wondering, because on stock it works far better...

The main problem is the location and orientation of the microphone. The hole for sound to get in is far too small and it, and the microphone face straight up (if using the Vaga for video chat) at a 90 degree angle to the screen. It will pick up the noise coming from the ceiling perfectly but the sound coming from the front where your talking from will be to low. So you increase the mic volume to compensate and pick up loads of background noise as well.

As eduardo has said only noise cancelation will help.

The mic may have seemed better on the stock rom because the input level may have been lower. Lower input level would sound better but quieter (you would need to speak louder).

Edited by richardmlea
Link to comment
Share on other sites

Guest ejtagle

Ok, so where are you guys now with ICS progress? I know still a long list has to be done, but could you make a list of what's already done?

I will try to help but I am not the best with kernel, but doesn't hurt to try again.

Areo

kernel porting... ;)

Link to comment
Share on other sites

Ok, so where are you guys now with ICS progress? I know still a long list has to be done, but could you make a list of what's already done?

I will try to help but I am not the best with kernel, but doesn't hurt to try again.

Areo

Yeah as Eduardo points out kernel porting is where its at right now ... Nothing really to be done at the moment until we get up on the required kernel... Stupid thing to work out with libs but there is no rush on the right now until kernel is done...

Current kernel works to a point Eduardo is porting 3.0 kernel so we don't have missing stuff we need to look at later.

Edited by Cass67
Link to comment
Share on other sites

Guest ejtagle

Yeah as Eduardo points out kernel porting is where its at right now ... Nothing really to be done at the moment until we get up on the required kernel... Stupid thing to work out with libs but there is no rush on the right now until kernel is done...

Current kernel works to a point Eduardo is porting 3.0 kernel so we don't have missing stuff we need to look at later.

The only (important) thing not related to kernel is the hwcomposer green layer issue. I dissasembled the hw composer, but see no real problems. Seems to be quite standard, by the way. Even if it is referencing nvidia libs, seems not to be calling them.... I think the problem is when it actually composes layers (by default, it does not, except with emerging popups) . And, something funny i found out: This hwcomposer is the responsible for hdmi output, and it has specific properties to tell how to display into the screen:

nvidia.hwc.mirror_mode

It can be set to the following strings

crop , scale or center

Link to comment
Share on other sites

The only (important) thing not related to kernel is the hwcomposer green layer issue. I dissasembled the hw composer, but see no real problems. Seems to be quite standard, by the way. Even if it is referencing nvidia libs, seems not to be calling them.... I think the problem is when it actually composes layers (by default, it does not, except with emerging popups) . And, something funny i found out: This hwcomposer is the responsible for hdmi output, and it has specific properties to tell how to display into the screen:

nvidia.hwc.mirror_mode

It can be set to the following strings

crop , scale or center

Thank you very much Eduardo! Works perfectly.

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
×
×
  • Create New...

Important Information

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