> On Fri, 17 Jul 1998, Kendall Bennett wrote:
>
> > Now this means that the kernel *could* be modified such that the I/O
> > bitmap is not allocated at all by default for a task, and the
> > allocation of the I/O bitmap is staved off until a task calls the
> > ioperm() system call. [...]
>
> the problem is that the ->iomap pointer within the TSS is only 16 bits :(
> This means we'd have to allocate the bitmap within 64k of the address of
> the (already allocated) task structure. It's highly unlikely that this can
> be done cleanly :( Maybe we could 'relocate' the task, but this is way too
> much work and complexity, as all possible data structures that somehow
> might reference this task have to be fixed up.
Yeah, I have looked into this and the TSS structure is firmly
embedded in the entire task structure, which means that moving it out
of there so that it can be resized is going to be rather complicated
to say the least. I am positive that it can be done, but it is going
to have wide reaching implications and could have undesired effects
on other parts of the kernel.
Hence I have decided that the simplest way to solve this problem is
for us to distribute a patch for the latest kernels with our
software, and if people have problems running the BIOS code on their
system they can apply the kernel patch and re-build a kernel that
will work. I believe that most cards should work just fine without
the patch, so this will affect only a few cards (but then again ATI
has shipped probably well in excess of 10 million graphics cards ;-(
).
All our future driver technology (6-12 month time frame and PCI/AGP
only) is all NoBIOS so that we wont need the BIOS access technology
anymore (it is also cross CPU portable so it can run on Alpha, MIPS,
PowerPC and Intel Merced). Hence this is only a necessary short term
solution (or a solution for people with ISA/VLB cards).
But of course if someone has a brainwave and figures out how to do
this cleanly I am all ears ;-)
Regards,
+--------------------------------------------------------------------------+
| SciTech Software - Building Truly Plug'n'Play Software! |
+--------------------------------------------------------------------------+
| Kendall Bennett | Email: KendallB@scitechsoft.com |
| Director of Engineering | Phone: (530) 894 8400 |
| SciTech Software, Inc. | Fax : (530) 894 9069 |
| 505 Wall Street | ftp : ftp.scitechsoft.com |
| Chico, CA 95928, USA | www : http://www.scitechsoft.com |
+--------------------------------------------------------------------------+
-
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.altern.org/andrebalsa/doc/lkml-faq.html