ARM binary code translator

[Dels]

Hi there,

As you may know that Buildroid team already port Intel Medfield ARM
binary code translator into their ICS build

http://www.buildroid.org/blog/?p=175

Is it possible to port this library from this build into android-x86
ISO?

[Marek Vasut]

Dear Dels,

> Hi there,
>
> As you may know that Buildroid team already port Intel Medfield ARM
> binary code translator into their ICS build
>
> http://www.buildroid.org/blog/?p=175

Very good catch. Looks like they run patched dalvik, so we'd have to wait for
patches. And those might never see the light of day, since it's all BSD
licensed. Shitty android, here it all comes.

>
> Is it possible to port this library from this build into android-x86
> ISO?

Best regards,
Marek Vasut

[Chih-Wei Huang]

2012/6/13 Dels <crushf...@gmail.com>:

Intel never release the source or patches of
its binary translator to public.
(correct me if I'm wrong).
I don't understand how Buildroid could do the "porting".

Besides, I think none of our developers have got
a medfield-powered smartphone.
(I've asked some friends in Intel Taiwan.
They said the phone is not available to Taiwan market)
If someone is going to donate the Intel phones to us,
I'm happy to try.


--
Chih-Wei
Android-x86 project
http://www.android-x86.org

[Dels]

I extracted libvm, libhoudini and arm library from Buildroid:
http://jumbofiles.com/6n9i5zznj1xw/houdini.zip.html

Next, i will tried injecting into eeepc ISO and test (i know it maybe
gives error or else non bootable system)

On 14 Jun, 11:15, Chih-Wei Huang <cwhu...@android-x86.org> wrote:
> 2012/6/13 Dels <crushfire2...@gmail.com>:

[Marek Vasut]

Dear Dels,

> I extracted libvm, libhoudini and arm library from Buildroid:
> http://jumbofiles.com/6n9i5zznj1xw/houdini.zip.html

So basically if I compiled libdvm with other optimizations than what's enabled
on the medfield phone, I'm pooped. But then, it indicates that intel added some
changes only to dalvik (libdvm) and added some other library that has the BT in
it, right?

> Next, i will tried injecting into eeepc ISO and test (i know it maybe
> gives error or else non bootable system)
>
> On 14 Jun, 11:15, Chih-Wei Huang <cwhu...@android-x86.org> wrote:
> > 2012/6/13 Dels <crushfire2...@gmail.com>:
> > > Hi there,
> > > As you may know that Buildroid team already port Intel Medfield ARM
> > > binary code translator into their ICS build
> > >
> > >http://www.buildroid.org/blog/?p=175
> > >
> > > Is it possible to port this library from this build into android-x86
> > > ISO?
> >
> > Intel never release the source or patches of
> > its binary translator to public.
> > (correct me if I'm wrong).
> > I don't understand how Buildroid could do the "porting".
> >
> > Besides, I think none of our developers have got
> > a medfield-powered smartphone.
> > (I've asked some friends in Intel Taiwan.
> > They said the phone is not available to Taiwan market)
> > If someone is going to donate the Intel phones to us,
> > I'm happy to try.
> >
> > --
> > Chih-Wei
> > Android-x86 projecthttp://www.android-x86.org

Best regards,
Marek Vasut

[Morten Sjøgren]

Dan commented on June 14, 2012 at 9:20 pm: 

"Hi,
I’ll try to write a post on how patching the android source code to support the libhoudini. This will enable android-x86 users to have ARM emulation.
Expect it in 1 or 2 days."

Looking forward to his post on it.

[Marek Vasut]

Dear Morten Sjøgren,

> Dan commented on June 14, 2012 at 9:20

> pm<http://www.buildroid.org/blog/?p=175#comment-1763>
>
> *"*Hi,

> I’ll try to write a post on how patching the android source code to support
> the libhoudini. This will enable android-x86 users to have ARM emulation.
> Expect it in 1 or 2 days."

Good to hear. Thanks for tracking the comments and updating us ;-)

> Looking forward to his post on it.
>
> On Wednesday, June 13, 2012 4:48:01 AM UTC+2, Dels wrote:
> > Hi there,
> >
> > As you may know that Buildroid team already port Intel Medfield ARM
> > binary code translator into their ICS build
> >
> > http://www.buildroid.org/blog/?p=175
> >
> > Is it possible to port this library from this build into android-x86
> > ISO?

Best regards,
Marek Vasut

[Dels]

Finally, good news for everyone

[TheARC77]

Oh yeah. Currently using Buildroid and wait powerup my Sammy Netbook with ICS

[Ollie]

Sounds very exciting news. Has anyone compiled this on a tega/viewpad build?

[Tim Kowalsky]

This is very exciting.  On his blog, Dan also states he recovered an x86 Flash library from the Medfield phone.

 

[Daniel Fages]

Hi all,

instructions posted here : http://www.buildroid.org/blog/?p=198

Enjoy

Dan.

2012/6/15 Tim Kowalsky <akro...@designby.com>

This is very exciting.  On his blog, Dan also states he recovered an x86 Flash library from the Medfield phone.

 

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/nDEYZ9ZhpLIJ.

To post to this group, send email to andro...@googlegroups.com.
To unsubscribe from this group, send email to android-x86...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/android-x86?hl=en.

[TheARC77]

OMG! Well time to a x86 build with this patchs/libs, pls someone make a nice try!! :D

El lunes, 18 de junio de 2012 13:30:53 UTC-3, dfages escribió:

Hi all,

instructions posted here : http://www.buildroid.org/blog/?p=198

Enjoy

Dan.

2012/6/15 Tim Kowalsky <akro...@designby.com>

This is very exciting.  On his blog, Dan also states he recovered an x86 Flash library from the Medfield phone.

 

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/nDEYZ9ZhpLIJ.

To post to this group, send email to andro...@googlegroups.com.

To unsubscribe from this group, send email to android-x86+unsubscribe@googlegroups.com.

[Rui Carmo]

This is very nice, but can we distribute the Intel binary? Can it be rebuilt from source somehow?

Otherwise it's just a nice (temporary) hack that may break soon.

R.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/vFV0iyTlz94J.

To post to this group, send email to andro...@googlegroups.com.

To unsubscribe from this group, send email to android-x86...@googlegroups.com.

[Daniel Fages]

No, no source code available, at least I didn't find it...

Dan.

2012/6/18 Rui Carmo <rui....@gmail.com>

[Marek Vasut]

Dear Daniel Fages,

> Hi all,
> instructions posted here : http://www.buildroid.org/blog/?p=198

Patch doesn't apply ...

>
> Enjoy
> Dan.
>
> 2012/6/15 Tim Kowalsky <akro...@designby.com>
>
> > This is very exciting. On his blog, Dan also states he recovered an x86
> > Flash library from the Medfield phone.
> >
> >
> >
> >
> > --
> > You received this message because you are subscribed to the Google Groups
> > "Android-x86" group.
> > To view this discussion on the web visit
> > https://groups.google.com/d/msg/android-x86/-/nDEYZ9ZhpLIJ.
> >
> > To post to this group, send email to andro...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > android-x86...@googlegroups.com.
> > For more options, visit this group at
> > http://groups.google.com/group/android-x86?hl=en.

Best regards,
Marek Vasut

[Daniel Fages]

Oops, here's the good one : http://www.buildroid.org/Download/patch_houdini_ics_20120618_v2

Dan.

2012/6/18 Marek Vasut <marek...@gmail.com>

[Marek Vasut]

Dear Daniel Fages,

> No, no source code available, at least I didn't find it...

No source, but what's the license? That's more of a concern ...

> Dan.
>
> 2012/6/18 Rui Carmo <rui....@gmail.com>
>
> > This is very nice, but can we distribute the Intel binary? Can it be
> > rebuilt from source somehow?
> >
> > Otherwise it's just a nice (temporary) hack that may break soon.
> >
> > R.
> >
> >
> > On 18/06/2012, at 19:03, TheARC77 <thea...@f4krom.com.ar> wrote:
> >
> > OMG! Well time to a x86 build with this patchs/libs, pls someone make a
> > nice try!! :D
> >
> > El lunes, 18 de junio de 2012 13:30:53 UTC-3, dfages escribió:
> >> Hi all,
> >> instructions posted here :

> >> http://www.buildroid.org/**blog/?p=198<http://www.buildroid.org/blog/?p
> >> =198>

> >>
> >> Enjoy
> >> Dan.
> >>
> >> 2012/6/15 Tim Kowalsky <akro...@designby.com>
> >>
> >>> This is very exciting. On his blog, Dan also states he recovered an
> >>> x86 Flash library from the Medfield phone.
> >>>
> >>>
> >>>
> >>>
> >>> --
> >>> You received this message because you are subscribed to the Google
> >>> Groups "Android-x86" group.

> >>> To view this discussion on the web visit https://groups.google.com/d/**
> >>> msg/android-x86/-/nDEYZ9ZhpLIJ<https://groups.google.com/d/msg/android-
> >>> x86/-/nDEYZ9ZhpLIJ> **.

> >>>
> >>> To post to this group, send email to andro...@googlegroups.com.
> >>> To unsubscribe from this group, send email to

> >>> android-x86+unsubscribe@** googlegroups.com
> >>> <android-x86%2Bunsu...@googlegroups.com>. For more options, visit
> >>> this group at http://groups.google.com/**
> >>> group/android-x86?hl=en<http://groups.google.com/group/android-x86?hl=
> >>> en> .

> >>>
> >> --
> >
> > You received this message because you are subscribed to the Google Groups
> > "Android-x86" group.
> > To view this discussion on the web visit
> > https://groups.google.com/d/msg/android-x86/-/vFV0iyTlz94J.
> > To post to this group, send email to andro...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > android-x86...@googlegroups.com.
> > For more options, visit this group at
> > http://groups.google.com/group/android-x86?hl=en.
> >
> > --
> >
> > You received this message because you are subscribed to the Google Groups
> > "Android-x86" group.
> > To post to this group, send email to andro...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > android-x86...@googlegroups.com.
> > For more options, visit this group at
> > http://groups.google.com/group/android-x86?hl=en.

Best regards,
Marek Vasut

[Daniel Fages]

Hi,

I agree that's a main concern.

Unfortunately I don't really know the license.
When I use the Intel phone, there's no special "license popup" which asks me to accept some special license (contrary to what we see when using Google Apps).

I'll have to check the physical documents that came with the device to see if there's any mention of this.

If there's nothing, I don't know what we can conclude - would it mean we can freely use the binaries ?

Is there a lawyer in the room ? :-)

Dan.

2012/6/19 Marek Vasut <marek...@gmail.com>

[Daniel Fages]

Hi,

I agree that's a main concern.

Unfortunately I don't really know the license.
When I use the Intel phone, there's no special "license popup" which asks me to accept some special license (contrary to what we see when using Google Apps).

I'll have to check the physical documents that came with the device to see if there's any mention of this.

If there's nothing, I don't know what we can conclude - would it mean we can freely use the binaries ?

Is there a lawyer in the room ? :-)

Dan.

2012/6/19 Marek Vasut <marek...@gmail.com>

Dear Daniel Fages,

[Marek Vasut]

Dear Daniel Fages,

> Hi,
> I agree that's a main concern.
> Unfortunately I don't really know the license.
> When I use the Intel phone, there's no special "license popup" which asks
> me to accept some special license (contrary to what we see when using
> Google Apps).
> I'll have to check the physical documents that came with the device to see
> if there's any mention of this.
> If there's nothing, I don't know what we can conclude - would it mean we
> can freely use the binaries ?

We can not. I'll ask at intel eventually (if I won't forget), but I don't expect
much.

Best regards,
Marek Vasut

[Ollie]

Has anyone tried this using the android-x86 ics-x86 branch? Would love to see it in a Tega v2 build but have no idea how to follow those instructions. Don't really have the processing power either.

On Monday, 18 June 2012 17:30:53 UTC+1, dfages wrote:

Hi all,

instructions posted here : http://www.buildroid.org/blog/?p=198

Enjoy

Dan.

2012/6/15 Tim Kowalsky <akro...@designby.com>

This is very exciting.  On his blog, Dan also states he recovered an x86 Flash library from the Medfield phone.

 

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/nDEYZ9ZhpLIJ.

To post to this group, send email to andro...@googlegroups.com.

To unsubscribe from this group, send email to android-x86+unsubscribe@googlegroups.com.

[Chih-Wei Huang]

2012/6/20 Ollie <ojre...@gmail.com>:

> Has anyone tried this using the android-x86 ics-x86 branch? Would love to
> see it in a Tega v2 build but have no idea how to follow those instructions.
> Don't really have the processing power either.

Yes. I tried on tegav2. It works well.
I've succeeded to run angry birds (all variants)
and osmos demo from the Google Play.

[Chih-Wei Huang]

2012/6/19 Daniel Fages <dfa...@gmail.com>:

> Hi all,
> instructions posted here : http://www.buildroid.org/blog/?p=198

The LD_PRELOAD seems to cause some problem
in debug mode. Some commands just got
segmentation fault.

Is there a way to avoid that?
For example, only export LD_PRELOAD for
dalvik processes?

[Rinaldo Jonathan]

what? angry birds on android-x86?

tell me if i dont read it wrong .... or someone can punch me in the face to make sure if i not dreaming? :D

FINALLY!!! 
i just need to buy x86 machine with touch screen and run android-x86 inside it, and im done!!

2012/6/20 Chih-Wei Huang <cwh...@android-x86.org>

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.

To post to this group, send email to andro...@googlegroups.com.

To unsubscribe from this group, send email to android-x86...@googlegroups.com.

[Marek Vasut]

Dear Chih-Wei Huang,

> 2012/6/19 Daniel Fages <dfa...@gmail.com>:
> > Hi all,
> > instructions posted here : http://www.buildroid.org/blog/?p=198
>
> The LD_PRELOAD seems to cause some problem
> in debug mode. Some commands just got
> segmentation fault.
>
> Is there a way to avoid that?
> For example, only export LD_PRELOAD for
> dalvik processes?

Well ... if you checked what the preloaded library is for, I believe it should
be possible to implement that fake function directly in dalvik. It actually was
attempted even, check the source.

Best regards,
Marek Vasut

[Ollie]

Very exciting news! I hope you don't mind me asking, but are you able to share the build? Or will it be incorporated in the Android-x86 source in the future?

Thanks for the test report anyway.

[Rinaldo Jonathan]

yeah, and for virtualbox / vmware too :D
i will love it :D

2012/6/20, Ollie <ojre...@gmail.com>:

> --
> You received this message because you are subscribed to the Google Groups
> "Android-x86" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/android-x86/-/9s3p9c1HJ0AJ.

> To post to this group, send email to andro...@googlegroups.com.
> To unsubscribe from this group, send email to

> android-x86...@googlegroups.com.

[Marek Vasut]

Dear Rinaldo Jonathan,

> yeah, and for virtualbox / vmware too :D
> i will love it :D

When it's ready ... every time you ask such stupid question and don't help, the
release gets delayed. Since we have to type these stupid reply emails if nothing
else.

Best regards,
Marek Vasut

[Gernot Hillier]

Hi there!

Am 19.06.2012 07:36, schrieb Daniel Fages:
> I agree that's a main concern.
> Unfortunately I don't really know the license.
> When I use the Intel phone, there's no special "license popup" which
> asks me to accept some special license (contrary to what we see when
> using Google Apps).
> I'll have to check the physical documents that came with the device to
> see if there's any mention of this.
> If there's nothing, I don't know what we can conclude - would it mean we
> can freely use the binaries ?
> Is there a lawyer in the room ? :-)

I'm no layer, but as far as I know, in most contries, you're not allowed
to use not speaking of distribute any copyrighted work if you have no
license contract. So if you don't find anything, you're not allowed to
do anything with it...

--
Gernot

[Dels]

For VirtualBox you can already use Buildroid for that, "ignorant will
ignore"

On 20 Jun, 20:49, Rinaldo Jonathan <rinaldo.jonat...@gmail.com> wrote:
> yeah, and for virtualbox / vmware too :D
> i will love it :D
>

> 2012/6/20, Ollie <ojrear...@gmail.com>:

>
>
>
>
>
>
>
> > Very exciting news! I hope you don't mind me asking, but are you able to
> > share the build? Or will it be incorporated in the Android-x86 source in
> > the future?
>
> > Thanks for the test report anyway.
>
> > On Wednesday, 20 June 2012 05:00:52 UTC+1, Chih-Wei Huang wrote:
>

> >> 2012/6/20 Ollie <ojrear...@gmail.com>:

> >> > Has anyone tried this using the android-x86 ics-x86 branch? Would love
> >> to
> >> > see it in a Tega v2 build but have no idea how to follow those
> >> instructions.
> >> > Don't really have the processing power either.
>
> >> Yes. I tried on tegav2. It works well.
> >> I've succeeded to run angry birds (all variants)
> >> and osmos demo from the Google Play.
>
> >> --
> >> Chih-Wei
> >> Android-x86 project
> >>http://www.android-x86.org
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "Android-x86" group.
> > To view this discussion on the web visit
> >https://groups.google.com/d/msg/android-x86/-/9s3p9c1HJ0AJ.

> > To post to this group, send email toandr...@googlegroups.com.

[Chih-Wei Huang]

However, it seems we cannot support both
armeabi and armeabi-v7a abis.
Settings TARGET_CPU_ABI2 to armeabi,
we are unable to support armeabi-v7a only app
like Temple Run.

https://play.google.com/store/apps/details?id=com.imangi.templerun

Is there a solution?

(fortunately most apps containing native code
support both)

2012/6/20 prober <holger...@googlemail.com>:
> I'd just like to confirm that libhoudini is also working on a WeTab/EXOpc
> Atom N450 tablet with Android 4.0.4. Runs fast and fine ;-)

[Suman Saraf]

Chih-Wei,

As far as I know, the Intel translator can currently only translate ARM v5. The bluestacks binary translator can do both v5 and v7.

Suman

> --
> You received this message because you are subscribed to the Google Groups "Android-x86" group.

[Chih-Wei Huang]

2012/6/21 Suman Saraf <su...@android-x86.org>:

> Chih-Wei,
> As far as I know, the Intel translator can currently only translate ARM v5.

Does it?

I've tried to set abi2 to armeabi-v7a and
install angry birds. I've checked armeabi-v7a .so
are installed and it plays normal.
So I think it supports armeabi-v7a.

> The bluestacks binary translator can do both v5 and v7.

Cool! Could we integrate it into our isos?

And how did it set abi2?

[Rui Carmo]

On 21/06/2012, at 09:08, Chih-Wei Huang wrote:

Cool! Could we integrate it into our isos?

Doing a few searches, it seems that they want to turn their binary translator into a business.

You can get a feel for it here: https://getsatisfaction.com/bstk

I'm wondering if QEMU might help here somehow (not sure about whether you'd need to do all the ABI bits and just hand ARM code over to it...)

R.

[Marek Vasut]

Dear Rui Carmo,

well the application is either evaluating it's own functions or functions from
android ... so bridging it via qemu might be possible, but non-trivial

>
> R.

Best regards,
Marek Vasut

[Dels]

Some detail regarding binary translator:
http://www.anandtech.com/show/5770/lava-xolo-x900-review-the-first-intel-medfield-phone/3
http://software.intel.com/en-us/articles/creating-and-porting-ndk-based-android-apps-for-ia/

And since Lava Xolo X900 use gingerbread 2.3.7, is it possible that
libhoudini actually included on Intel Android* 2.3.7 (Gingerbread) x86
Emulator Image?
http://software.intel.com/en-us/articles/android-237-gingerbread-x86-emulator-image-add-on/

[Chih-Wei Huang]

2012/6/22 Dels <crushf...@gmail.com>:

> And since Lava Xolo X900 use gingerbread 2.3.7, is it possible that
> libhoudini actually included on Intel Android* 2.3.7 (Gingerbread) x86
> Emulator Image?
> http://software.intel.com/en-us/articles/android-237-gingerbread-x86-emulator-image-add-on/

Why don't you check yourself and tell us the result?

[Stefan]

Salam Rinaldo,

if your only target is to get Angry Birds running on X86, that is an easy job. Just take any Windows box and download the Windows version of Angry Birds and install... Download adresses via Google.

:D *scnr*

Stefan

[Marek Vasut]

Dear Thofen86,

> Suman,
>
> "Bluestacks binary translator" uses a modified version of QEMU which
> uses GPL license. Where can I download its source code?

Don't ask here, ask them. And let us know the result

> Thomas

>
> On Wednesday, June 20, 2012 11:59:13 PM UTC-7, Suman Saraf wrote:
> > Chih-Wei,
> >
> > As far as I know, the Intel translator can currently only translate ARM
> > v5. The bluestacks binary translator can do both v5 and v7.
> >
> > Suman
> >
> > On Jun 21, 2012, at 9:38 AM, Chih-Wei Huang wrote:
> > > However, it seems we cannot support both
> > > armeabi and armeabi-v7a abis.
> > > Settings TARGET_CPU_ABI2 to armeabi,
> > > we are unable to support armeabi-v7a only app
> > > like Temple Run.
> > >
> > > https://play.google.com/store/apps/details?id=com.imangi.templerun
> > >
> > > Is there a solution?
> > >
> > > (fortunately most apps containing native code
> > > support both)
> > >

> > > 2012/6/20 prober:

> > >> I'd just like to confirm that libhoudini is also working on a
> >
> > WeTab/EXOpc
> >
> > >> Atom N450 tablet with Android 4.0.4. Runs fast and fine ;-)

Best regards,
Marek Vasut

[Anunakin]

Nokia used, some time ago, a apps called ScratchBox, that was made for developers, and run native x86 apps under CPU and ARM under QEmu ... I think it cold be better then used a hacked version of Intel Phone..

Em terça-feira, 12 de junho de 2012 23h48min01s UTC-3, Dels escreveu:

Hi there,

As you may know that Buildroid team already port Intel Medfield ARM
binary code translator into their ICS build

http://www.buildroid.org/blog/?p=175

Is it possible to port this library from this build into android-x86
ISO?

[Rui Carmo]

Yes, sb2 was used as part of their (Maemo?) tool chain, and it basically ran qemu-user inside a shell prompt to enable you to execute ELF ARM binaries. Last I saw of it was as part of a Raspberry Pi cross-compiling setup (sb2 is a set of packages that also includes a cross-compiler that runs on Intel and compiles ARM binaries), and that was the reason I mentioned QEMU a few days back.

(I've been using qemu-user for a good while now to build ARM binaries for Debian. It works, and I've even got it to build under Lion in a sensible way. But I have no idea what it would take to load Android ARM binaries and execute them inside an Intel runtime)

I don't know about Intel's stuff, but BlueStack uses a modified QEMU library to do the job. Given the complexity of doing any sort of decent emulation, it's pretty likely that Intel does so as well, and I think that's the way to go eventually.

R.

--

You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/NUmGLn0fJXkJ.

[Thofen86]

I sent an email asking Bluestacks for the ARM binary translator source code three days ago. Still waiting for a response.

If you install Bluestacks (beta or the production version in Asus Vibe), using adb, you can easily extract a file /system/bin/arm-runtime. This is the key file which Bluestacks uses to run ARM apps. Perform a binary analysis, and you will see it is in fact Qemu. In newer version of its beta, the process name of arm-runtime is changed to zygote_arm. So discovering this fact becomes a little harder.

From a news link, http://download.cnet.com/8301-2007_4-57404865-12/latest-bluestacks-arms-your-pc/, the company called this LayerCake, without mentioning Qemu. Go figure.

[Dels]

Likely Daniel Fages already discovered the qemu stuff:
http://www.buildroid.org/blog/?p=12

"One of my interests in Android is to run it on x86 platform,
especially in virtual environments. From the beginning (?), Android
comes with qemu support to run Android ARM-code on x86 platform but as
you can expect, the performance is very poor… A good surprise came
with Ice Cream Sandwich (Android 4.0) : there is now a target in AOSP
to build Android for VirtualBox"

Surely you can ask him :D

On 25 Jun, 15:12, Thofen86 <xism...@gmail.com> wrote:
> I sent an email asking Bluestacks for the ARM binary translator source code
> three days ago. Still waiting for a response.
>
> If you install Bluestacks (beta or the production version in Asus Vibe),
> using adb, you can easily extract a file /system/bin/arm-runtime. This is
> the key file which Bluestacks uses to run ARM apps. Perform a binary
> analysis, and you will see it is in fact Qemu. In newer version of its
> beta, the process name of arm-runtime is changed to zygote_arm. So
> discovering this fact becomes a little harder.
>

> From a news link,http://download.cnet.com/8301-2007_4-57404865-12/latest-bluestacks-ar...,

[Daniel Fages]

Hi,

my post might be not very clear on that : I was only speaking of the qemu emulator coming with the SDK...

Dan.

2012/6/25 Dels <crushf...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.

[Rui Carmo]

The Android SDK ships with QEMU as part of the Android emulator. That's what he was talking about.

[Thofen86]

The QEMU implementation in Android SDK is QEMU full system emulation. QEMU runs whole Android ARM-based OS. Hence it is slow.
The QEMU implementation in Bluestacks (and maybe Intel?) is QEMU user mode emulation. QEMU user mode is ported as separate processes running INSIDE Android x86-based OS. It should have better performance.

These two implementations are different. Both are from QEMU open source with GPL license.

On Monday, June 25, 2012 1:38:36 AM UTC-7, Rui Carmo wrote:

The Android SDK ships with QEMU as part of the Android emulator. That's what he was talking about.

[dscm]

License FWIW:

from Intel Evaluation License Agreement:

THE EVALUATION MATERIALS ARE PROVIDED FOR EVALUATION PURPOSES ONLY AND
MAY NOT BE DISTRIBUTED BY RECIPIENT OR INCORPORATED INTO RECIPIENT’S
PRODUCTS OR SOFTWARE. PLEASE CONTACT AN INTEL SALES REPRESENTATIVE TO
LEARN ABOUT THE AVAILABILITY AND COST OF A COMMERICAL VERSION OF THE
EVALUATION SOFTWARE.

from here:

http://software.intel.com/en-us/articles/intel-eula-x86-android-2/

though section 10.8 may cover all but the libhoudini...???

On Jun 18, 7:37 pm, Marek Vasut <marek.va...@gmail.com> wrote:
> Dear Daniel Fages,
>
> > No, no source code available, at least I didn't find it...
>
> No source, but what's the license? That's more of a concern ...
>
>
>
>
>
>
>
>
>
> > Dan.
>
> > 2012/6/18 Rui Carmo <rui.ca...@gmail.com>
>
> > > This is very nice, but can we distribute the Intel binary? Can it be
> > > rebuilt from source somehow?
>
> > > Otherwise it's just a nice (temporary) hack that may break soon.
>
> > > R.
>
> > > On 18/06/2012, at 19:03, TheARC77 <thear...@f4krom.com.ar> wrote:
>
> > > OMG! Well time to a x86 build with this patchs/libs, pls someone make a
> > > nice try!! :D

>
> > > El lunes, 18 de junio de 2012 13:30:53 UTC-3, dfages escribió:
> > >> Hi all,
> > >> instructions posted here :

> > >>http://www.buildroid.org/**blog/?p=198<http://www.buildroid.org/blog/?p
> > >> =198>
>
> > >> Enjoy
> > >> Dan.
>
> > >> 2012/6/15 Tim Kowalsky <akropo...@designby.com>
>
> > >>> This is very exciting.  On his blog, Dan also states he recovered an
> > >>> x86 Flash library from the Medfield phone.

>
> > >>> --
> > >>> You received this message because you are subscribed to the Google
> > >>> Groups "Android-x86" group.

> > >>> To view this discussion on the web visithttps://groups.google.com/d/**
> > >>> msg/android-x86/-/nDEYZ9ZhpLIJ<https://groups.google.com/d/msg/android-
> > >>> x86/-/nDEYZ9ZhpLIJ> **.

>
> > >>> To post to this group, send email to andro...@googlegroups.com.
> > >>> To unsubscribe from this group, send email to

> > >>> android-x86+unsubscribe@** googlegroups.com
> > >>> <android-x86%2Bunsu...@googlegroups.com>. For more options, visit
> > >>> this group athttp://groups.google.com/**
> > >>> group/android-x86?hl=en<http://groups.google.com/group/android-x86?hl=
> > >>> en> .

>
> > >>  --
>
> > > You received this message because you are subscribed to the Google Groups
> > > "Android-x86" group.

> > > To view this discussion on the web visit

> > >https://groups.google.com/d/msg/android-x86/-/vFV0iyTlz94J.

> > > To post to this group, send email to andro...@googlegroups.com.
> > > To unsubscribe from this group, send email to
> > > android-x86...@googlegroups.com.
> > > For more options, visit this group at
> > >http://groups.google.com/group/android-x86?hl=en.
>
> > >  --
>
> > > You received this message because you are subscribed to the Google Groups
> > > "Android-x86" group.
> > > To post to this group, send email to andro...@googlegroups.com.
> > > To unsubscribe from this group, send email to
> > > android-x86...@googlegroups.com.
> > > For more options, visit this group at
> > >http://groups.google.com/group/android-x86?hl=en.
>

> Best regards,
> Marek Vasut

[Holger Schurig]

from Intel Evaluation License Agreement:

THE EVALUATION MATERIALS ARE PROVIDED FOR EVALUATION PURPOSES ONLY AND
MAY NOT BE DISTRIBUTED BY RECIPIENT OR INCORPORATED INTO RECIPIENT’S
PRODUCTS OR SOFTWARE. PLEASE CONTACT AN INTEL SALES REPRESENTATIVE TO
LEARN ABOUT THE AVAILABILITY AND COST OF A COMMERICAL VERSION OF THE
EVALUATION SOFTWARE.

The GPL doesn't know a "evaluation version". If you distribute binaries of GPLed programs, you have to submit the source as well. Period.

QEMU itself is mostly GPLv2. Just libqemu.a is LGPL. So, if they used libqemu.a unchanged, then they can re-license the result under something different. But if they made changes to libqemu.a (the the source codes leading to libqemu.a), then they have to re-distribute that source code ... and non one can hinder you on using that.

If they don't comply, they risk of revoking their own license grant and might not be able to use QEMU any longer.

Please note that this doesn't hold for BSDish licensed software.

[Chih-Wei Huang]

2012/6/20 Chih-Wei Huang <cwh...@android-x86.org>:
> 2012/6/19 Daniel Fages <dfa...@gmail.com>:
>> Hi all,
>> instructions posted here : http://www.buildroid.org/blog/?p=198

>
> The LD_PRELOAD seems to cause some problem
> in debug mode. Some commands just got
> segmentation fault.
>
> Is there a way to avoid that?
> For example, only export LD_PRELOAD for
> dalvik processes?

I've found a way to avoid LD_PRELOAD.
(and libnativehelper_GBfake.so)
Now the solution looks promising. :)

[Holger Pandel]

Sounds very promising! I'm absolutely curious about how you solved that....

[Thofen86]

So far, I do not see traces of QEMU in Intel's libhoudini, but arm-runtime in Bluestacks is from QEMU.

[Dels]

Likely Intel use another approach to modify .so file(s) on runtime
directly without recompiling, similar attempt already done for MIPS
based processor on XDA Developer forum
http://forum.xda-developers.com/showthread.php?t=1411879

If that is true, then thats why some application will run smoothly and
full speed while some would lag as hell :D

[Chih-Wei Huang]

2012/6/26 Holger Pandel <holger...@googlemail.com>:

> Sounds very promising! I'm absolutely curious about how you solved that....

It's already in our git repository.

[Daniel Fages]

Congrats Chih-Wei,

much more elegant way than my "fake library".
Here's the patch for the curious : 

diff --git a/libnativehelper/Register.cpp b/libnativehelper/Register.cpp

index b6b1b1f..18b53f1 100644

--- a/libnativehelper/Register.cpp

+++ b/libnativehelper/Register.cpp

@@ -22,7 +22,7 @@ extern int registerJniHelp(JNIEnv* env);

 /*

  * Register all methods for system classes.

  */

-int jniRegisterSystemMethods(JNIEnv* env) {

+extern "C" int jniRegisterSystemMethods(JNIEnv* env) {

     // JniHelp depends on core library classes such as java.io.FileDescriptor.

     return registerCoreLibrariesJni(env) != -1 && registerJniHelp(env) != -1;

 }

diff --git a/vm/Init.cpp b/vm/Init.cpp

index 96a7c2a..e8b2f9b 100644

--- a/vm/Init.cpp

+++ b/vm/Init.cpp

@@ -38,7 +38,7 @@

 /*

  * Register VM-agnostic native methods for system classes.

  */

-extern int jniRegisterSystemMethods(JNIEnv* env);

+extern "C" int jniRegisterSystemMethods(JNIEnv* env);

 /* fwd */

 static bool registerSystemNatives(JNIEnv* pEnv);

Dan.

2012/6/27 Chih-Wei Huang <cwh...@android-x86.org>

--

[Holger Pandel]

Many thanks! I'll pull the changes later!

[dscm]

APP_ABI := armeabi x86 armeabi-v7a mips in Application.mk maybe???

On Jun 21, 12:08 am, Chih-Wei Huang <cwhu...@android-x86.org> wrote:
> However, it seems we cannot support both
> armeabi and armeabi-v7a abis.
> Settings TARGET_CPU_ABI2 to armeabi,
> we are unable to support armeabi-v7a only app
> like Temple Run.
>
> https://play.google.com/store/apps/details?id=com.imangi.templerun
>
> Is there a solution?
>
> (fortunately most apps containing native code
> support both)
>

> 2012/6/20 prober <holger.pan...@googlemail.com>:

>
> > I'd just like to confirm that libhoudini is also working on a WeTab/EXOpc
> > Atom N450 tablet with Android 4.0.4. Runs fast and fine ;-)
>

[dscm]

so the:

modified dalvik (libdvm) to dynamically load a library named
libdvm_houdini.so (which is just a rename of the libdvm.so from the
Intel phone) and call the needed functions to load and run ARM code –
it was quite tricky ’cause I had to disassemble the lib to know the
exact function parameters to be used. An other problem was that the
libdvm_houdini.so has some external dependency which is not resolved
in ICS (it looks for a jniRegisterSystemMethods() C function which has
been replaced by a same-name C++ function in ICS) ; to solve that,
I’ve compiled a “libnativehelper_GBfake” library which only implements
this function ; final thing : this “libnativehelper_fake” library has
to be preloaded before executing dalvik, to do this I’ve the “export
LD_PRELOAD /system/lib/libnativehelper_GBfake.so” line to the init.rc
(just after export BOOTCLASSPATH)

from your instructions is not needed if one uses the latest repo????

That also does not affect the patch to dalvik correct???

> 2012/6/27 Chih-Wei Huang <cwhu...@android-x86.org>
>
>
>
>
>
>
>
> > 2012/6/26 Holger Pandel <holger.pan...@googlemail.com>:

> > > Sounds very promising! I'm absolutely curious about how you solved
> > that....
>
> > It's already in our git repository.
>
> > > Am Dienstag, 26. Juni 2012 12:36:59 UTC+2 schrieb Chih-Wei Huang:
>

> > >> 2012/6/20 Chih-Wei Huang <cwhu...@android-x86.org>:

[Holger Pandel]

Dan, would you be so nice to describe, what changed exactly in handling your code and the new diff? I looked at the dalvik source, last commit, and i don't get it. Your commit includes the fake lib but Chih-Wei's last commit doesn't remove it, it's still in the source. Does that mean, the library is still needed for building, but not used anymore? I thought, every reference to the fake lib should have been removed now... And also, there is some slight difference between the patch from buildroid and in the source here:

diff --git a/vm/oo/Object.h b/vm/oo/Object.h
index b361926..41b8c99 100644
--- a/vm/oo/Object.h
+++ b/vm/oo/Object.h
@@ -580,6 +580,9 @@ struct Method {
 
     /* set if method was called during method profiling */
     bool            inProfile;
+
+    // is ARM ?
+    int        is_arm;  <- your original patch, but in the source now it's    + bool   is_arm; does that make any difference?
 };

I'm asking because I get a libdvm.so segfault with a fresh repo build, but my older build which incorporates your original patch does still work.

2012/6/27 Chih-Wei Huang <cwh...@android-x86.org>

To unsubscribe from this group, send email to android-x86+unsubscribe@googlegroups.com.

[Holger Pandel]

Ok. Everythings working now. Thanks!

[Ollie]

Thanks for these. Unfortunately the asus_laptop iso didn't seem to boot on any of my machines. Which is funny since I swear that an older asus_laptop image worked on my netbook previously. It just reboots when selecting a choice from the unetbootin menu. Same thing happens when I use the dd command. Is anyone erlse getting this issue?

Also, tabletsx86 are you planning on uploading more builds? Specifically the tegav2 version? Great work anyway, thank you so much for spending time, processing power and bandwidth to offer your excellent testing build service!

On Monday, 2 July 2012 14:56:00 UTC+1, tabletsx86 wrote:

testing builds available...

   http://www.tabletsx86.org/


On Jun 20, 9:26 am, Ollie <ojrear...@gmail.com> wrote:
> Very exciting news! I hope you don't mind me asking, but are you able to
> share the build? Or will it be incorporated in the Android-x86 source in
> the future?
>
> Thanks for the test report anyway.
>
>
>
>
>
>
>
> On Wednesday, 20 June 2012 05:00:52 UTC+1, Chih-Wei Huang wrote:
>
> > 2012/6/20 Ollie <ojrear...@gmail.com>:

> > > Has anyone tried this using the android-x86 ics-x86 branch? Would love
> > to
> > > see it in a Tega v2 build but have no idea how to follow those
> > instructions.
> > > Don't really have the processing power either.
>
> > Yes. I tried on tegav2. It works well.
> > I've succeeded to run angry birds (all variants)
> > and osmos demo from the Google Play.
>

[Ollie]

I have a Viewpad 10 which is just a TegaV2 so it is Atom based. :) I'll try another USB later.

On Tuesday, 3 July 2012 14:55:57 UTC+1, tabletsx86 wrote:

do you have an Intel Atom based asus laptop???

both builds were tested for booting on an Intel Atom based tablet...
and both booted fine....READ THE ARM-EMULATION note on the site...

check and redo the USB your building if other asus_laptops worked
on your laptop before...

> uploading more builds? Specifically
> the tegav2 version?

we will defer to Chih-Wei on the tegav2 builds but there maybe
other Intel Atom based builds we can make available....

[Ollie]

Tested out your new arm translator asus_laptop image on my Netbook and it turns out the first USB was funny. Redid it made it work. Unfortunately the hid-multitouch drivers aren't available like in the Tega v2 targets so I couldn't use the touchscreen on my Viewpad 10. Will there be more arm translator enabled images for different targets some time like you have done previously? That would be awesome. :) 

On Tuesday, 3 July 2012 16:05:58 UTC+1, tabletsx86 wrote:

targets from android-x86.org are optimized for different x86 platforms
this means more than you have the same cpu...!!!

[TheARC77]

Except for ICS OpenGL, the apps appears and can be download from store, but with the first several don't work. Or i'm missing something?

Samsung N150Plus (latin american version)

Ty in advance

[TheARC77]

Is a pitty you dont share your EE custom build. Run awesome in yours videos.

El lunes, 25 de junio de 2012 00:50:13 UTC-3, ALoGeNo escribió:

Hi there, I have that stuff working, and works pretty fine (Thank you Dan for this contribution, Marex for your help in the IRC, and of course cwhuang and all of you guys).

I did a How-to detailed and step by step, you can find that here in my blog : http://android-x86.sceners.org/en/?p=536

I did some benchmarks with quake3droid ARM and i was get fantastics results, so BRAVO.! this is a very nice advance.

I hope to be helpful.

Regards.

[notandyet]

Great job !!!

you are getting closer to a universal build, very useable

first iso to run on all 3 diverse platforms and speed is good on all,
even atom was useable, but got better after install, it got stable.

feedback on android-x86-4.0-asus_laptop-amei-20120702.iso

Machine 1=HP dv6753cl  intel  centrino
   internal wifi
      after updates
      books work
             sound very low even with sound boost app
      camera preview best ever in low light situations,
                 but  snap shots are overexposed and slow
                 completely overcast outside and side facing window
nearly white
      voice (mic) not working
      Chrome not compatible
      downloaded word search works
      downloaded angry birds screen goes black then app dies

Machine 2=MSI Brazos e-350is motherboard
      runs fine and fast (boots best with video=HDMI-A-1:d)
 usb wifi
 ( have no attached mic or camera )
 HDMI over scans on TV and no sound via HDMI (has refresh rate of 30)
 VGA 1920x1080 works fine and as well as HDMI on ASUS 24" monitor
 external powered speakers has good volume  yeah!!!
        usb Bluetooth pairs but no volume at all

Machine 3=intel dg945glf2 atom (has svideo out)
      runs fine and fast
 ( have no attached mic or camera )
 Ethernet connection
 VGA 1280x768 works fine 17" Samsung monitor
 ear buds and sound boost almost useable
        stability issues much tweaking needed to boot without
        browser, settings, you-tube etc crashing.

installed to USB 8gb split into 2 partitions 2nd partition as ext3
        apps on atom stabilized after install and reboot.

will test on these old motherboards in a few days

intel dg945glfD atom (no  svideo out)
ASUS M4A785-M with amd dual core rana cpu

thank you

[ALoGeNo]

Hello again, I did a find, I was looking for some dump of the intel phones this weeks without results I was look for Xolo x900 (only India). Lenovo K800 (only China) and now the new Orange San Diego aka Santa Clara (Only in UK, but maybe for the rest of Europe too) all have the same android version 2.3.7.

Yesterday nigth, I found "two" not one but TWO haha, one dump of Orange San diego 2.3.7, and another more of 4.0.4 (from april of this year), the first one works, the last one didnt, I was see many changes in the /lib/arm/ files, and i was see an increase in the houdini stuff, but I did some tests and dont works, maybe Mr Fages can do something more than me, but looks like isnt a full dump.

In any case, I was find more houdini files inside the roms, this is inside of /bin and are:

 "enable_houdini and disable_houdini"

This two files are two scripts wich logically start and stop houdini

"houdini"

Wich is 2.1MB big is a binary executable and is called by the scripts.


And letme show you some thing..

#!/system/bin/sh

# if you don't see the files 'register' and 'status' in /proc/sys/fs/binfmt_misc
# then run the following command:
# mount -t binfmt_misc none /proc/sys/fs/binfmt_misc

# this is to add the supported binary formats via binfmt_misc

houdini_bin=0

if [ ! -e /proc/sys/fs/binfmt_misc ]; then
    modprobe binfmt_misc
fi

if [ ! -e /proc/sys/fs/binfmt_misc/register ]; then
    mount -t binfmt_misc none /proc/sys/fs/binfmt_misc
fi

if [ -e /proc/sys/fs/binfmt_misc/register ]; then
    # register Houdini for arm binaries
    echo ':arm:M::\\x7f\\x45\\x4c\\x46\\x01\\x01\\x01\\x00\\x00\\x00\\x00\\x00\\x00\\x00\\x00\\x00\\x02\\x00\\x28::/system/bin/houdini:' > /proc/sys/fs/binfmt_misc/register
    if [ -e /proc/sys/fs/binfmt_misc/arm ]; then
        houdini_bin=1
    fi
else
    echo "No binfmt_misc support"
fi

if [ $houdini_bin -eq 0 ]; then
    echo "houdini enabling failed !!!"
fi
if [ $houdini_bin -eq 1 ]; then
    echo "houdini enabled"
fi

exit 0

Like you see we need to do modprobe binfmt_misc, if you look in our kernel configs, you'll see that flag.

#
# Executable file formats / Emulations
#
CONFIG_BINFMT_ELF=y
# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
CONFIG_HAVE_AOUT=y
# CONFIG_BINFMT_AOUT is not set
CONFIG_BINFMT_MISC=y
CONFIG_HAVE_ATOMIC_IOMAP=y
CONFIG_HAVE_TEXT_POKE_SMP=y
CONFIG_NET=y

 
So we cant load this module because this is loaded in kernel, and not like module

We will must to set in kernel CONFIG_BINFMT_MISC=m

In any case the system seems work with that module in kernel, and like external module ( I did the try)
But this module looks important, wikipedia say:

binfmt_misc is a capability of the Linux kernel which allows arbitrary executable file formats to be recognized and passed to certain user space applications, such as emulators and virtual machines.

Then that has more sense, but the question is.. WTF is the houdini binary?, and another question is.. why we are not using this files?, and another one more question is.. WTF launch the enable_houdini here?, is the houdini binary an essential part of the Binary translator?

And now another one more thing, I did another find:

If we check the build.prop in the ICS build, we will see that:

ro.product.cpu.abi=x86
ro.product.locale.language=en
ro.wifi.channels=
ro.board.platform=mrst
# ro.build.product is obsolete; use ro.product.device
ro.build.product=mfld_pr2
# Do not try to parse ro.build.description or .fingerprint
ro.build.characteristics=default
# end build properties

#
# ADDITIONAL_BUILD_PROPERTIES
#
keyguard.no_require_sim=true
ro.com.android.dataroaming=true
ro.com.android.dateformat=MM-dd-yyyy
ro.config.ringtone=Ring_Synth_04.ogg
ro.config.notification_sound=pixiedust.ogg
ro.config.alarm_alert=Alarm_Classic.ogg
ro.ril.status.polling.enable=0
dalvik.vm.heapsize=64m
ro.product.cpu.abi2=armeabi
ro.product.cpu.upgradeabi=armeabi-v7a

OMG,..what is that hehe... looks like they are working to can emulate armv7, NICE!, but I did some tests (lot of them) and the armv6 or armv7 apps dont work.

Well and this is all.

You can download the dumps from here:

http://loadbalancing.modaco.com/download.php?file=system.sandiego.zip < 2.3.7 working in real phones.
http://loadbalancing.modaco.com/download.php?file=supk_eng.release.mfld_pr2.BKB4OUK.devr3.i2847.zip < 4.0.4 DONT WORK in real phones.

Cheers. 

[Edward Peek]

Is it possible binfmt is only to allow direct execution of pure ARM executables via the kernel whereas libhoudini does the actual binary translation? If this were the case binfmt would not be necessary as it seems most of the trouble with ARM is the shared objects embedded in APKs and referenced via the NDK which directly invokes houdini withoout the need for binfmt.

[Ollie]

Will there be an updated Tega v2 build with arm library enabled by any chance?

On Thursday, 5 July 2012 02:04:04 UTC+1, tabletsx86 wrote:

testing builds HAVE BEEN UPDATED...

   http://www.tabletsx86.org/

On Jul 3, 1:41 pm, TheARC77 <thear...@f4krom.com.ar> wrote:
> Except for ICS OpenGL, the apps appears and can be download from store, but
> with the first several don't work. Or i'm missing something?
>

> Samsung N150Plus<http://www.samsung.com/latin/consumer/monitor-peripherals-printer/not...>(latin american version)

[TheARC77]

Downloading and testing!!!

El miércoles, 4 de julio de 2012 22:04:04 UTC-3, tabletsx86 escribió:

testing builds HAVE BEEN UPDATED...

   http://www.tabletsx86.org/


On Jul 3, 1:41 pm, TheARC77 <thear...@f4krom.com.ar> wrote:

> Except for ICS OpenGL, the apps appears and can be download from store, but
> with the first several don't work. Or i'm missing something?
>

> Samsung N150Plus<http://www.samsung.com/latin/consumer/monitor-peripherals-printer/not...>(latin american version)

[TheARC77]

Angry Birds Space Premiun works !!!!!

Google Plus, Opera Mobile (not x86 test build), Mx Video Player, Friendcaster Pro, SocialFrame HD, The Amazing SpiderMan Live Wallpapper and so on. Games Sentinel 3, Shoot the apple, Magic piano (:P), Iced Age 4 and moooooooooooaaaaaaaaaare games works perfectly!!!

And a lot of paid stuff is working or test for 15 minutes :D

Not working usuals brightness, mic, video recording, lives wallpapers included, Flash Player from Store (I suggest include the San Diego plugin like Alogeno post), Poweramp (FC), Chrome and others (perhaps plataform/graphic detection)

Another suggestion: when i setup dpi at 120 for get more space in launcher, after reboot i cant use "Settings" (FC), can you fix it? Another, can you integrate newer version of SuperUser (3.1.3 x86 here or SuperSU). There is newer version of Google Apps (Store/Gmail/Youtube) and incorporate others like G+ and Currents at least. And if it possible, anothers phisical layouts. Sorry for this :P

Really thank you, at least some love for X86 and is getting so interesting!!!

Please view pics from my G+ Album in special about Antutu

El miércoles, 4 de julio de 2012 22:04:04 UTC-3, tabletsx86 escribió:

testing builds HAVE BEEN UPDATED...

   http://www.tabletsx86.org/


On Jul 3, 1:41 pm, TheARC77 <thear...@f4krom.com.ar> wrote:

> Except for ICS OpenGL, the apps appears and can be download from store, but
> with the first several don't work. Or i'm missing something?
>

> Samsung N150Plus<http://www.samsung.com/latin/consumer/monitor-peripherals-printer/not...>(latin american version)

[TheARC77]

Hi there, if you see my Antutu information screen cap this other pic from Alogeno build (nop i don't have it, i take it from his site, he don't share his build)
he has OpenGL2 running. I think he take some tips from www.buildroid.org

El miércoles, 4 de julio de 2012 22:04:04 UTC-3, tabletsx86 escribió:

testing builds HAVE BEEN UPDATED...

   http://www.tabletsx86.org/


On Jul 3, 1:41 pm, TheARC77 <thear...@f4krom.com.ar> wrote:

> Except for ICS OpenGL, the apps appears and can be download from store, but
> with the first several don't work. Or i'm missing something?
>

> Samsung N150Plus<http://www.samsung.com/latin/consumer/monitor-peripherals-printer/not...>(latin american version)

[Stefan]

Hello,

I already played Angry Birds on my WeTab (exoPC). This is based on the tegav2 iso image, see wetab-community.com, there is an android section with the Arm enabled image, but this won't work anymore on the Tegav2 as there are some differences in the hal. But maybe one can backport the Arm support to Tegav2...?

Stefan

--

You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/r8XsuRh3BsMJ.

To post to this group, send email to andro...@googlegroups.com.

To unsubscribe from this group, send email to android-x86...@googlegroups.com.

[Stefan]

Hello,

yes I can confirm, a lot more software is running, and the performance is good. Angry Birds really makes fun! But some apps are not offered, like the Firefox beta. I think this is because the arm emulator only emulates an old subset of the Arm assembler language. Is there any way to test which subset is emulated, and is there a good benchmark tool, which differentiates between ARM, X86 and Dalvik apps and compares to other Android devices? I have downloaded one free benchmark, and there my WeTab only has about 3800 points, while a Samsung Galaxy S3 has abou 18.000 points, I can't believe that...

Stefan

--

You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/osD3UiF2l2MJ.

[TheARC77]

Dear "ppl" behind Tabletsx86 i will suggest a new discussion about yours builds and issues.

El miércoles, 4 de julio de 2012 22:04:04 UTC-3, tabletsx86 escribió:

testing builds HAVE BEEN UPDATED...

   http://www.tabletsx86.org/


On Jul 3, 1:41 pm, TheARC77 <thear...@f4krom.com.ar> wrote:

> Except for ICS OpenGL, the apps appears and can be download from store, but
> with the first several don't work. Or i'm missing something?
>

> Samsung N150Plus<http://www.samsung.com/latin/consumer/monitor-peripherals-printer/not...>(latin american version)

[ALoGeNo]

On Thursday, July 5, 2012 7:40:50 PM UTC+2, Stefan wrote:

Hello,

yes I can confirm, a lot more software is running, and the performance is good. Angry Birds really makes fun! But some apps are not offered, like the Firefox beta. I think this is because the arm emulator only emulates an old subset of the Arm assembler language. Is there any way to test which subset is emulated, and is there a good benchmark tool, which differentiates between ARM, X86 and Dalvik apps and compares to other Android devices? I have downloaded one free benchmark, and there my WeTab only has about 3800 points, while a Samsung Galaxy S3 has abou 18.000 points, I can't believe that...

Stefan

Hey Stefan, you saw my benchmarks? take a look to get a more close comparision :)  http://android-x86.sceners.org/en/?p=580

Regards, and thanks for your work.

[blackhawk]

Using tabletsx86 asus_laptop ISO and ALoGeNo San Diego's houdini library, now I have a playable Angry Bird on my Lenovo Ideapad S100, including San Diego's Intel Inside boot animation. Thanks to Android-x86 developers, tabletsx86, and ALoGeNo.

Salam,
Ketut P. Kumajaya

[notandyet]

after install to EXT3 partition build
android-x86-4.0-asus_laptop-v7a-20120706.iso
running on MSI E-350is amd brazos motherboard 4gb ram
wifi very slow on RT28xx
when boot from CD boot it is very fast but from hdd with 55gb free
space very slow
any thoughts?

additional boot options used DPi=160 video=HDMI-A-1:d


On Jul 5, 10:35 pm, tabletsx86 <rbg.tablets...@gmail.com> wrote:
>  new testing builds available...
>
>    http://www.tabletsx86.org/

[John McHugh]

@tabletsx86 - I have an Acer ZG5, I cant seem to use the webcam with external applications like google plus for hangouts but it seems to detect it in the default camera app. Do you have any idea what might be causing this issue. Is it that my specific netbook is not compatible or is it a macro issue?

On Tuesday, 10 July 2012 14:52:17 UTC+1, tabletsx86 wrote:

updated testing builds available...

   http://www.tabletsx86.org/

PLEASE read arm-emulation note...

On Jul 4, 9:26 pm, Ollie <ojrear...@gmail.com> wrote:

[Anunakin]

Hi all I using Android-x86 test build (android-x86-4.0-asus_laptop-20120710.iso) from tabletsx86.org

First all, thanks for that work from Android-x86 team and Tabletsx86!

I have a Dell XPS15 L502x, isos for asus_laptop works nice here, but not detects all my 6GB of RAM (only 1GB) dont know how 

I found some problems on using your test build:

build.prop using armeabi-v7a and armeabi

With that titanium backup all near all apps who needs root access dont works, I think is because that apps detects x86 or arm boards...

If using x86, armeabi and then armeabi-v7a on build.prop (ro....cpu cpu2 and cpu3) .. that way all apps works, and Google Store show arm apps too...

Em terça-feira, 10 de julho de 2012 10h52min17s UTC-3, tabletsx86 escreveu:

updated testing builds available... 

   http://www.tabletsx86.org/ 

PLEASE read arm-emulation note... 


On Jul 4, 9:26 pm, Ollie <ojrear...@gmail.com> wrote: 

[Ketut Putu Kumajaya]

@tabletsx86
1. Since 2012-07-06 asus_laptop build, internal sdcard support removed from /system/etc/init.sh :(
2. APK installer extract wrong Titanium Backup's busybox and sqlite3 binary, extract wrong AnTuTu Benchmark library. Both applicatian have native x86 support, no x86 cpu.abi since 2012-07-06 build :)
3. 2012-07-10 build extracted to /android-2012-07-09 directory. Android-x86 refuse to boot since my manually edited Grub map the kernel and initrd to /android-2012-07-10 :)
But of course, big thanks for your build.

@Anunakin
I will try your cpu.abi tips, thank you.

Best regards,
Ketut P. Kumajaya

[ALoGeNo]

On Thursday, July 12, 2012 2:09:23 PM UTC+2, Anunakin wrote:

Hi all I using Android-x86 test build (android-x86-4.0-asus_laptop-20120710.iso) from tabletsx86.org

First all, thanks for that work from Android-x86 team and Tabletsx86!

I have a Dell XPS15 L502x, isos for asus_laptop works nice here, but not detects all my 6GB of RAM (only 1GB) dont know how 

I found some problems on using your test build:

build.prop using armeabi-v7a and armeabi

With that titanium backup all near all apps who needs root access dont works, I think is because that apps detects x86 or arm boards...

If using x86, armeabi and then armeabi-v7a on build.prop (ro....cpu cpu2 and cpu3) .. that way all apps works, and Google Store show arm apps too...

 
Hey Anunakin can you explain a bit more your method please?, so you are using then:

ro.product.cpu.abi1=x86
ro.product.cpu2.abi1=armeabi
ro.product.cpu3.abi1=armeabi-v7a

In that way works????? (Amazing) :O

Did you try? (this method is used in the leaked orange san diego ICS build.prop) :

ro.product.cpu.abi1=x86
ro.product.cpu.abi2=armeabi
ro.product.cpu.upgradeabi=armeabi-v7a

Or did you try?:

ro.product.cpu.abi1=armeabi
ro.product.cpu.abi2=armeabi-v7a

This methods also works for me, in any case I did lot of tests in armv7 mode and all the armv7 apps dont works at the moment, I did some tests using armv6 apps and nothing not results.
We will must to wait some San diego ICS dump but funtional one.

Cheers

[Ron M]

I think armeabi-v7a is not supported, or at least not fully supported, regardless of what you put in  ro.product.cpu.abi[2].

Did anyone manage to run Fruit Ninja or Temple Run on their build? Both crash.
(all) Angry birds do work fine however when GPU is working. 

@Daniel Fages

A bit unrelated, but did you manage to run Angry Birds in one of your VirtualBox/VmWare build? I just want to make sure it is a GPU accelration topic and not related to the emulation in any way.

[Daniel Fages]

Hi,

yes I was able to run Angry Birds on VirtualBox, with and without GPU acceleration.

DAn.

2012/7/16 Ron M <ron...@gmail.com>

--

You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/WtFP16w-meMJ.

[Thofen86]

Here is the result.

I emailed Bluestacks, stated the fact and requested the source code of their qemu based ARM binary translator. The company asked my mailing address, which then I provided.  A few days later I received a CD containing the ARM binary translator source code from Bluestacks. It's qemu-user based.

Thomas

On Friday, June 22, 2012 4:24:38 AM UTC-7, Marek Vašut wrote:

Dear Thofen86,

> Suman,
>
>     "Bluestacks binary translator" uses a modified version of QEMU which
> uses GPL license. Where can I download its source code?

Don't ask here, ask them. And let us know the result

> Thomas
>
> On Wednesday, June 20, 2012 11:59:13 PM UTC-7, Suman Saraf wrote:
> > Chih-Wei,
> >
> > As far as I know, the Intel translator can currently only translate ARM
> > v5. The bluestacks binary translator can do both v5 and v7.
> >
> > Suman
> >
> > On Jun 21, 2012, at 9:38 AM, Chih-Wei Huang wrote:
> > > However, it seems we cannot support both
> > > armeabi and armeabi-v7a abis.
> > > Settings TARGET_CPU_ABI2 to armeabi,
> > > we are unable to support armeabi-v7a only app
> > > like Temple Run.
> > >
> > > https://play.google.com/store/apps/details?id=com.imangi.templerun
> > >
> > > Is there a solution?
> > >
> > > (fortunately most apps containing native code
> > > support both)
> > >
> > > 2012/6/20 prober:
> > >> I'd just like to confirm that libhoudini is also working on a
> >
> > WeTab/EXOpc
> >
> > >> Atom N450 tablet with Android 4.0.4. Runs fast and fine ;-)

[dscm]

and the licensing says.............????????????????

[Thofen86]

In the CD-ROM, there are no specific licensing statements,  except those standard ones found in qemu code base.  For example, in the main.c, it says,

"
/*
 *  qemu user main
 *
 *  Copyright (c) 2003-2008 Fabrice Bellard
 *
 *  This program is free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 2 of the License, or
 *  (at your option) any later version.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program; if not, see <http://www.gnu.org/licenses/>.
 */

[Rui Carmo]

Okay, so it's usable. Does it build?

--
You received this message because you are subscribed to the Google Groups "Android-x86" group.

To view this discussion on the web visit https://groups.google.com/d/msg/android-x86/-/3PLKlOMarK4J.