> >>>>> "Jon" == Jon M Taylor <taylorj@ecs.csus.edu> writes:
>
> Jon> On 18 Nov 1998, Jes Sorensen wrote:
> >> The current virgefb is just a hack to allow people to use the card
> >> as a framebuffer after the video mode has been set by the boot
> >> `firmware' if you like. The driver really needs someone who cares
> >> to sit down and write the part of the code that initialised the gfx
> >> modes.
>
> Jon> The KGI ViRGE chipset driver is currently in active
> Jon> development and does not have these problems. It is a
> Jon> full-featured driver, with support for MMIO relocation,
> Jon> accelerations (not many yet but the framework is present and
> Jon> works), and true VGA and S3 'fasttext' textmode support. It
> Jon> should also be fairly easy to port to other architectures due to
> Jon> the KGI abstract device I/O scheme used.
>
> The natural thing he is of course to look at the code and port it to
> become a real fbdev driver.
Why? We have a *real* fbdev driver now - the KGIcon drivers are
API conformant. How else do you want to define "real"?
> However one should make an effort not to
> break it for non-Intel architectures.
That is the point of the KGI abstract I/O scheme. This is
somewhat dumbed-down in KGIcon, but when the new Degas-KGI is released
(i.e. we are back to "full" KGI as a total replacement for fbcon/fbdev) we
will have a *very* nice system with separation of autodetection,
configuration (multihead, etc) and I/O methods. Currently the KGI chipset
driver handles all this stuff, which is leading to bloat. It needs to be
broken out into another driver subsection. One single chipset (like an S3
Trio64) can be found on ISA, PCI-x86, PCI-Alpha, PCI-PPC, Amiga Zorro, and
even (IIRC) sbus cards. But the actual chipset is the same on all of
them. The chipset driver shoould not have to deal with all this crap
itself. It should be handed a properly initialized card/bus setup and
just deal with programming the video chipset.
Jon
--- 'Cloning and the reprogramming of DNA is the first serious step in becoming one with God.' - Scientist G. Richard Seed
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/