On 8/1/06, Denis Oliver Kropp <[EMAIL PROTECTED]> wrote:
Albert Torras wrote:
>
>
> On 7/29/06, *Denis Oliver Kropp* < [EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
>
>     Most of the CPU access to the frame buffer happens in the software
>     rendering code in src/gfx/generic/generic.c, but there's also some
>     code in src/core/surfacemanager.c for "upload/download" of buffers.
>
>     Another place is src/misc/gfx_util.c for scaling and converting
>     (loaded) image data to the various formats a surface can have.
>
>     Furthermore, applications can also access the framebuffer using the
>     public API IDirectFBSurface::Lock().
>
>     Endianness problems? Or no direct mapping of the framebuffer?
>
>     --
>     Best regards,
>        Denis Oliver Kropp
>
>
>
>
> Hi,
>
> as I am working with a Motorola CPU, I defined WORDS_BIGENDIAN in
> config.h in order to solve the possible endianness problem.
>
> I have seen  the DirectFB library detects my framebuffer works with
> ARGB1555, when it actually does with RGB16.
>
> How can this be possible? Is there any way to solve this?

You can hack the detection in systems/fbdev/fbdev.c or fix the
frame buffer driver.

> I have tried to adjust colors manually, but without success...

If you have a big endian CPU with a little endian bus it's not easy...

--
 
 
 
Is the detection done in "static DFBSurfacePixelFormat dfb_fbdev_get_pixelformat( struct fb_var_screeninfo *var )"  ??
 
I will try in a couple of days... I hope it works.
 
Regards,
Albert Torras
 

_______________________________________________
directfb-dev mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev

Reply via email to