Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 1 Nov 1997 09:32:15 +0100
From:      Christoph Kukulies <kuku@gilberto.physik.RWTH-Aachen.DE>
To:        dg@root.com
Cc:        "Guillermo A. Loyola" <gmo@epigram.com>, Christoph Kukulies <kuku@gilberto.physik.RWTH-Aachen.DE>, freebsd-hackers@freefall.FreeBSD.org
Subject:   Re: mmap/mlock problem
Message-ID:  <19971101093215.27334@gil.physik.rwth-aachen.de>
In-Reply-To: <199710281036.CAA12019@implode.root.com>; from David Greenman on Tue, Oct 28, 1997 at 02:36:27AM -0800
References:  <3455B267.FBC7DD67@epigram.com> <199710281036.CAA12019@implode.root.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Oct 28, 1997 at 02:36:27AM -0800, David Greenman wrote:
> >Worse yet, the interface for drivers to do it pmap_mapdev()
> >claims to make the mapping non-cached in the comment, but the code
> >does not seem to be doing it. The non-cached bits are defined in
> >pmap.h:
> >
> >i386/include/pmap.h:75:#define PG_N             (PG_NC_PWT|PG_NC_PCD)   /* Non-cacheable */
> >
> >but are only used in machdep.c when probing memory.
> >
> >Questions for FreeBSD hackers:
> >
> >1) Shouldn't the line that sets the ptes in pmap_mapdev read:          
> >	*pte = pa | PG_RW | PG_N | PG_V;
> >
> >   instead of
> >	*pte = pa | PG_RW | PG_V;
> >   ?
> >
> >2) WOuld it be reasonalbe to add a new flag to mmap to allow
> >   specification of uncached mappings?
> 
>    The code used to set the no-cache flags, but I changed it to do a normal
> mapping after seeing relibility problems with certain memory mapped devices (I
> suspect this was caused by PWT rather than PCD). The cachability is normally
> determined by chipset settings and the page table flags are not needed to
> disable it for non-RAM areas of the address space.

When my BIOS doesn't have settings to make ISA memory regions
unchacheable (I assume it's not the 'shadow ram' section of the
chipset setup), can I be sure that the chipset does that for me?

Or would adding the PG_N flag to the device mapping code
assure this. I would not mind adding this in my special
case if only I could get the memory region behave like it
does under DOS.

> 
> -DG
> 
> David Greenman
> Core-team/Principal Architect, The FreeBSD Project

-- 
Chris Christoph P. U. Kukulies kuku@gil.physik.rwth-aachen.de



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19971101093215.27334>