large framebuffers and mmap

large framebuffers and mmap

[Jon Smirl]

 
 

=====
Jon Smirl
jonsmirl@yahoo.com

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

large framebuffers and mmap

[Jon Smirl]

Right now the framebuffer drivers are mapping the framebuffer into the kernel
address space. But shouldn't the fbconsole layer be doing the ioremap()? That
way there would be a single point of remap for all framebuffer drivers.  It
would also let all of the high mem access code be localized into fbconsole if
the fb was mapped high.

This would also get around my problem of using framebuffer drivers but not
using framebuffer console. In my case I don't have enough kernel address space
to do the remap. If fbconsole is doing the map, then the kernel map will never
happen on my system since I don't load fbconsole. 

I looked through a couple of drivers and none of them seem to touch the buffer
after it has been remapped. They only pass the pointer back to fbconsole.

=====
Jon Smirl
jonsmirl@yahoo.com

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: large framebuffers and mmap

[Geert Uytterhoeven]

On Sun, 21 Sep 2003, Jon Smirl wrote:
> Right now the framebuffer drivers are mapping the framebuffer into the kernel
> address space. But shouldn't the fbconsole layer be doing the ioremap()? That
> way there would be a single point of remap for all framebuffer drivers.  It
> would also let all of the high mem access code be localized into fbconsole if
> the fb was mapped high.
> 
> This would also get around my problem of using framebuffer drivers but not
> using framebuffer console. In my case I don't have enough kernel address space
> to do the remap. If fbconsole is doing the map, then the kernel map will never
> happen on my system since I don't load fbconsole. 
> 
> I looked through a couple of drivers and none of them seem to touch the buffer
> after it has been remapped. They only pass the pointer back to fbconsole.

What about
  - read from/write to /dev/fb*?
  - the logo code?

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

Re: large framebuffers and mmap

[James Simmons]

> > I looked through a couple of drivers and none of them seem to touch the buffer
> > after it has been remapped. They only pass the pointer back to fbconsole.
> 
> What about
>   - read from/write to /dev/fb*?
>   - the logo code?

Logo code should be using imageblit which in non accelerated cases uses 
cfb_imageblit. So we need the pointer. Also the software cursor uses the
ioremap memory space. Plus read/write to /dev/fb*. So it is not just the 
fbconsole that uses screen_base. The problem is broken drivers that 
allocate way to much memory.





-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf