Linux MIPS Architecture development
 help / color / mirror / Atom feed
* More on modules...
@ 1998-01-27  0:45 Alex deVries
  1998-01-27  1:51 ` William J. Earl
  0 siblings, 1 reply; 4+ messages in thread
From: Alex deVries @ 1998-01-27  0:45 UTC (permalink / raw)
  To: SGI Linux


The modutils in the CVS appears to work; I'll get an RPM out tonight for
others.

But, I use the term 'work' with a grain of salt; the modules appear to
load, but I get errors like:
Illegal instruction at c000c0d0 ra=88034ed4
Unable to handle kernel paging request at virtual address c0008113, epc ==
c000c0d8, ra == 88034ed4
Killed

- A

-- 
      Alex deVries          Run Linux on everything,
  System Administrator      run everything on Linux.
   The EngSoc Project       Send spam to spam@engsoc.carleton.ca.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: More on modules...
  1998-01-27  0:45 More on modules Alex deVries
@ 1998-01-27  1:51 ` William J. Earl
  1998-01-27  6:41   ` ralf
  0 siblings, 1 reply; 4+ messages in thread
From: William J. Earl @ 1998-01-27  1:51 UTC (permalink / raw)
  To: Alex deVries; +Cc: SGI Linux

Alex deVries writes:
 > 
 > The modutils in the CVS appears to work; I'll get an RPM out tonight for
 > others.
 > 
 > But, I use the term 'work' with a grain of salt; the modules appear to
 > load, but I get errors like:
 > Illegal instruction at c000c0d0 ra=88034ed4
 > Unable to handle kernel paging request at virtual address c0008113, epc ==
 > c000c0d8, ra == 88034ed4
 > Killed
...

      I wonder if the kernel is properly synchronzing the I-cache with
the D-cache after loading the module.  In general, you need to
writeback the primary D-cache and invalidate the primary I-cache for
the range of addresses occupied by the driver (or simply
index-writeback-invalidate all of the D-cache and index-invalidate all
of the I-cache, if the driver is larger than the cache size).  

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: More on modules...
  1998-01-27  1:51 ` William J. Earl
@ 1998-01-27  6:41   ` ralf
  1998-01-27  8:05     ` Alex deVries
  0 siblings, 1 reply; 4+ messages in thread
From: ralf @ 1998-01-27  6:41 UTC (permalink / raw)
  To: William J. Earl; +Cc: Alex deVries, SGI Linux

On Mon, Jan 26, 1998 at 05:51:40PM -0800, William J. Earl wrote:

>  > But, I use the term 'work' with a grain of salt; the modules appear to
>  > load, but I get errors like:
>  > Illegal instruction at c000c0d0 ra=88034ed4
>  > Unable to handle kernel paging request at virtual address c0008113, epc ==
>  > c000c0d8, ra == 88034ed4
>  > Killed
> ...
> 
>       I wonder if the kernel is properly synchronzing the I-cache with
> the D-cache after loading the module.  In general, you need to
> writeback the primary D-cache and invalidate the primary I-cache for
> the range of addresses occupied by the driver (or simply
> index-writeback-invalidate all of the D-cache and index-invalidate all
> of the I-cache, if the driver is larger than the cache size).  

Yes, we do that properly.

Alex, when compiling modules, does the compiler pass the flag -mlong-jumps?

  Ralf

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: More on modules...
  1998-01-27  6:41   ` ralf
@ 1998-01-27  8:05     ` Alex deVries
  0 siblings, 0 replies; 4+ messages in thread
From: Alex deVries @ 1998-01-27  8:05 UTC (permalink / raw)
  To: ralf; +Cc: William J. Earl, SGI Linux


On Tue, 27 Jan 1998 ralf@uni-koblenz.de wrote:
> On Mon, Jan 26, 1998 at 05:51:40PM -0800, William J. Earl wrote:
> >       I wonder if the kernel is properly synchronzing the I-cache with
> > the D-cache after loading the module.  In general, you need to
> > writeback the primary D-cache and invalidate the primary I-cache for
> > the range of addresses occupied by the driver (or simply
> > index-writeback-invalidate all of the D-cache and index-invalidate all
> > of the I-cache, if the driver is larger than the cache size).  
> 
> Yes, we do that properly.

> Alex, when compiling modules, does the compiler pass the flag -mlong-jumps?

No, but it does compile them with -mlong-calls.

- Alex "damned insomnia" deVries

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~1998-01-27  8:09 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
1998-01-27  0:45 More on modules Alex deVries
1998-01-27  1:51 ` William J. Earl
1998-01-27  6:41   ` ralf
1998-01-27  8:05     ` Alex deVries

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox