Accelerated framebuffer driver

Jan 13, 2004 3 Replies

Hello, because the generic VESA framebuffer on Geode-boards runs slowly, I compiled the AMD (Nat.Semi.) framebuffer driver (nsc-kfb-driver-2.7.7.tar.gz) into the kernel. Now I would like to use the accelerated drawing functions (for lines, bitBlt's, filled rectangles...).



May I use the functions which (might be available inside the kernel) or do I have to compile the accelerated functions a second time and to link it to my program file?



What's the correct way to use the advantages of the framebuffer driver?



Thanks an advance, Juergen


While searching the net for something completely different a few days ago, I stumbled over this article:

formatting link

At first glance, it seems to talk about similar issues, so maybe it is helpful to you..

Cheers

Rob

Robert Kaiser email: rkaiser AT sysgo DOT com SYSGO AG http://www.elinos.com Klein-Winternheim / Germany http://www.sysgo.com

To take advantage of the the NSC accelerated framebuffer, you've got to call the accelerated ioctls() that the NSC driver supports. They do provide some headers (or they do in the version that I've got) in the galproto.h and some interface source in galfns.c.

This would have to be done in Xfree86 if you're using that, and you'd have to modify the Xserver to take advantage of that. Somewhere on my list of things to do if I ever get time is to modify the stock Xfbdev server to support some of the basic acceleration functions.

If you're using the framebuffer directly, then you can do whatever you want, of course.

Alex Pavloff - remove BLAH to email Software Engineer, ESA Technology

IMHO, the correct way is to use a standard api like DirectFB

formatting link
maybe you can improve the GX1 directfb driver insead of rolling your own one.

clemens

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required