Hi!
Calls for isa_memset_io() and isa_writeb() in hga_clear_screen() and
hga_show_logo() make garbage and bad things since 2.4.13. (Unable to
handle kernel paging request at virtual address...)
Anybody experiencing problems may want to revert changes made in
2.4.13. It solves the problem.
I have two questions about asm-i386/io.h:
1. Why #define __ISA_IO_base ((char *)(PAGE_OFFSET)) ?
2. Why not #define __ISA_IO_base ((char *)0) ?
Regards:
Ferenc Bakonyi
> I have two questions about asm-i386/io.h:
> 1. Why #define __ISA_IO_base ((char *)(PAGE_OFFSET)) ?
> 2. Why not #define __ISA_IO_base ((char *)0) ?
For ix86 the ISA IO space start is at the very begining of memory. On
other platforms like the PPC the ISA IO space is seperate from the
regular memory space. It starts after the regular memory space thus the
__ISA_IO_base will not be 0x0.
On Thu, 8 Nov 2001, James Simmons wrote:
> > I have two questions about asm-i386/io.h:
> > 1. Why #define __ISA_IO_base ((char *)(PAGE_OFFSET)) ?
> > 2. Why not #define __ISA_IO_base ((char *)0) ?
>
> For ix86 the ISA IO space start is at the very begining of memory. On
> other platforms like the PPC the ISA IO space is seperate from the
> regular memory space. It starts after the regular memory space thus the
> __ISA_IO_base will not be 0x0.
He's talking about asm-i386/io.h.
__ISA_IO_base (what's in a name, why not call it __ISA_MEM_base?!?!?) is the
kernel virtual address of the start of ISA _memory_ space. On ia32 PCs, ISA
memory space overlaps with the first 16 MB of real RAM, at CPU physical address
0.
(DISCLAIMER: my ia32 knowledge is limited)
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
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