One more question and just a double check --- have you copied bzImage to
vendor/asus/eee_701 after you build the kernel and before you build
installer_img?

On Mon, Apr 27, 2009 at 2:49 AM, FrankieCZ <[email protected]> wrote:

>
> I did check of kernel config file, VESA fb. is included in the kernel
> for sure. When I do /proc/fb I cannot see VESA, actually, where should
> I see it?
> It seems VESA fb is included but probably not initialized. I have also
> downloaded one Android image (compiled by someone else) and this image
> runs perfectly in VirtualBox (but don't know details about this
> image).
>
> Now, I'm trying to compile everything from scratch (with the lastest
> version of Android from git) and I will see if something change...it
> takes some time to compile it etc. :)
>
> Frank
>
> On Apr 27, 12:10 am, Yi Sun <[email protected]> wrote:
> > Please also check the content of /proc/fb If u vessfb driver is
> > installed coreectly, you should see VESA in that file. The
> > surfaceflinger crash is because it can not find /dev/graphics/fb
> > during the init process and it is caused by wrong fb driver
> > installation and configuration.
> > Please do check .config file to make sure u indeed enabled VESA fb.
> > Even u have already applied patches  there are still many other things
> > can went wrong
> >
> > On Apr 26, 2009, at 3:04 AM, Chen Yang <[email protected]> wrote:
> >
> >
> >
> > > oops, have you compiled the kernel with builtin vesafb?
> > > it is not there by default.
> > > --
> > > Chen
> >
>  > > On 4/26/09, FrankieCZ <[email protected]> wrote:
> >
> > >> Unfortunately, still the same.
> > >> I removed vesafb, add vga=788, I can boot and see black screen only
> > >> (I'm waiting enough time to be sure that Android GUI will not
> > >> appear).
> > >> I've tried to change vesafb into intelfb (in cmdline), just if
> > >> something will change, but nothing.
> >
> > >> Obviously, there is some problem with SurfaceFlinger / GUI
> > >> initialization, maybe because I'm running it under VirtualBox? Is
> > >> there any special FB or video driver to be compiled into the Kernel?
> >
> > >> Frank
> >
> > >> On Apr 26, 10:24 am, Chen Yang <[email protected]> wrote:
> > >>> please remove the vesafb in cmdline and append vga=788
> > >>> and have a try. :)
> > >>> --
> > >>> Chen
> >
> > >>> On Sun, Apr 26, 2009 at 3:57 PM, FrankieCZ <[email protected]>
> > >>> wrote:
> >
> > >>>> Hi,
> >
> > >>>> for booting I'm using these cmdline parameters:
> >
> > >>>> console=tty0 console=ttyS1,115200n8 console=tty0
> > >>>> androidboot.hardware=eee_701 vesafb
> >
> > >>>> (during Kernel compilation, VESAFB drivers in menuconfig were
> > >>>> selected. Generally, I've applied all patches provided on the wiki
> > >>>> (except those 2 for atl2 driver)
> >
> > >>>> (if I don't specify vga parameter I can boot to console mode and
> > >>>> access console via alt+f1, if I specify vga parameter (does not
> > >>>> matter, which resolution / bit depth, I cannot access console
> > >>>> after)
> >
> > >>>> the logcat oputput:
> >
> > >>>> I/SurfaceFlinger(2536): SurfaceFlinger is starting
> > >>>> I/SurfaceFlinger(2536): SurfaceFlinger's main thread ready to run.
> > >>>> Initializing graphics H/W...
> > >>>> E/MemoryHeapBase(2536): error opening /dev/pmem: No such file or
> > >>>> directory
> > >>>> E/GLLogger(2536): couldn't load <libhgl.so> library (Cannot find
> > >>>> library)
> > >>>> I/SurfaceFlinger(2536): EGL informations:
> > >>>> I/SurfaceFlinger(2536): # of configs : 6
> > >>>> I/SurfaceFlinger(2536): vendor  : Android
> > >>>> I/SurfaceFlinger(2536): version : 1.31 Android META-EGL
> > >>>> I/SurfaceFlinger(2536): extensions:
> > >>>> I/SurfaceFlinger(2536): Client API: OpenGL ES
> > >>>> E/GLLogger (2536) validate_display_surface:488 error 300d
> > >>>> (EGL_BAD_SURFACE)
> > >>>> W/SurfaceFlinger(2536): ro.sf.lcd_density not defined, using 160
> > >>>> dpi
> > >>>> by default.
> > >>>> E/GLLogger (2536): call to OpenGL ES API with no current context
> > >>>> I/SurfaceFlinger(2536): OpenGL informations:
> > >>>> E/GLLogger (2536): call to OpenGL ES API with no current context
> > >>>> D/Zygote (2535): Process 2536 terminated by signal (11)
> > >>>> I/Zygote (2535): Exit Zygote because system server (2536) has
> > >>>> terminated
> >
> > >>>> As you said, libhgl.so problem is probably not important but for
> > >>>> some
> > >>>> reason SurfaceFlinger is terminated (segfault) and therefore,
> > >>>> Zygote
> > >>>> fails to start (better say is terminated) as well.
> >
> > >>>> BTW: Is there any possibility how to access console if I specify
> > >>>> vga
> > >>>> parameter? And, is there any possibility how to get logcat output
> > >>>> from
> > >>>> VirtualBox window - some copy&paste or accessing logcat output file
> > >>>> (if I redirected logcat output to the file)?
> >
> > >>>> Frank
> >
> > >>>> On Apr 26, 3:56 am, Yi Sun <[email protected]> wrote:
> > >>>>> The libhgl error is harmless, please send us the logcat output and
> > >>>>> make sure you are using right frame buffer driver as mentioned
> > >>>>> in the
> > >>>>> wiki
> >
> > >>>>> On Apr 25, 2009, at 10:12 AM, FrankieCZ <[email protected]>
> > >>>>> wrote:
> >
> > >>>>>> Problem is solved. I did everything again and problem with
> > >>>>>> dalvik-
> > >>>>>> cache is not present (seems there was some problem during
> > >>>>>> creating img
> > >>>>>> file, evenly i've tried it twice...). Now, I have another
> > >>>>>> problem,
> > >>>>>> which is related to SurfaceFlinger and problem with:
> >
> > >>>>>> 608 E/GLLogger(  441): couldn't load <libhgl.so> library
> > >>>>>> (Cannot find
> > >>>>>> library)
> >
> > >>>>>> and segfault problem...so GUI is not starting still :(
> >
> > >>>>>> This seems to be more common problem (as far as I could see more
> > >>>>>> people facing with it), so I'm trying to figure out how to solve
> > >>>>>> it....
> >
> > >>>>>> Anyway, thanks,
> >
> > >>>>>> Frank
> >
> > >>>>>> On Apr 25, 6:06 pm, Chen Yang <[email protected]> wrote:
> > >>>>>>> You may be able to reference the init.rc to see which
> > >>>>>>> partition is
> > >>>>>>> mounted as /data( by default it seems to be /dev/block/sda8),
> > >>>>>>> do you
> > >>>>>>> have the right partition and file system type(ext3)?
> > >>>>>>> --
> > >>>>>>> Chen
> >
> > >>>>>>> On Sat, Apr 25, 2009 at 11:57 PM, FrankieCZ
> > >>>>>>> <[email protected]>
> > >>>>>>> wrote:
> >
> > >>>>>>>> Hi all,
> >
> > >>>>>>>> I've tried to port Android on my x86 laptop and run it inside
> > >>>>>>>> VirtualBox. I was following this tutorial:
> >
> > >>>>>>>>
> http://code.google.com/p/patch-hosting-for-android-x86-support/wiki/B
> > >>>>>>>> ...
> >
> > >>>>>>>> Everything went smoothly but when I try to boot, I'm staying
> > >>>>>>>> stuck
> > >>>>>>>> with black screen (alt+f1 console does not work). If  I don't
> > >>>>>>>> specify
> > >>>>>>>> vga=XXX parameter (cmdline during booting) I can boot to
> > >>>>>>>> console
> > >>>>>>>> mode,
> > >>>>>>>> after I can use logcat, which tells me:
> >
> > >>>>>>>> ...
> > >>>>>>>> Can't open dex cache '/data/dalvik-cache/
> > >>>>>>>> sys...@framew...@[email protected]': No such file or
> > >>>>>>>> directory
> > >>>>>>>> Unable to open or create cache for /system/framework/core.jar
> > >>>>>>>> ...
> >
> > >>>>>>>> and therefore, Dalvik does not start.
> >
> > >>>>>>>> If I try to delete all content of /data/ directory, I'm getting
> > >>>>>>>> error
> > >>>>>>>> that file-system is a read-only.
> >
> > >>>>>>>> Any suggestions or similar problems?
> >
> > >>>>>>>> Thanks for help,
> >
> > >>>>>>>> Frank
> >
> >
> >
>

--~--~---------~--~----~------------~-------~--~----~
unsubscribe: [email protected]
website: http://groups.google.com/group/android-porting
-~----------~----~----~----~------~----~------~--~---

Reply via email to