linux-mips
[Top] [All Lists]

Re: [PATCH 2.5] SGI O2 framebuffer driver

To: Alan Cox <alan@lxorguk.ukuu.org.uk>
Subject: Re: [PATCH 2.5] SGI O2 framebuffer driver
From: Ralf Baechle <ralf@linux-mips.org>
Date: Thu, 12 Dec 2002 03:33:07 +0100
Cc: Vivien Chappelier <vivienc@nerim.net>, linux-mips@linux-mips.org, Ilya Volynets <ilya@theIlya.com>
In-reply-to: <1039656676.18587.63.camel@irongate.swansea.linux.org.uk>; from alan@lxorguk.ukuu.org.uk on Thu, Dec 12, 2002 at 01:31:16AM +0000
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <Pine.LNX.4.21.0212120004330.2300-100000@melkor> <1039656676.18587.63.camel@irongate.swansea.linux.org.uk>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mutt/1.2.5.1i
On Thu, Dec 12, 2002 at 01:31:16AM +0000, Alan Cox wrote:

> On Wed, 2002-12-11 at 23:41, Vivien Chappelier wrote:
> > linear framebuffer (up to 8MB with 64kB granularity). I'm then remapping
> > all those pages to one virtual region obtained from get_vm_area so that
> > 1. caching attributes can be set to cacheable write-through no WA
> 
> Ick. The framebuffer can't handle cached and write barriers ?

The O2 is non-cache coherent.  So with the fairly large write-back second
level caches enabled frame buffer write could potencially be delayed
indefinately but in any case quite long.  Frame buffers are usually only
written to, so the cache mode "uncached accelerated" seems preferable
but only the R10000 provides this mode, so for the R5000 write-though no
write allocation is the next best solution.

  Ralf

<Prev in Thread] Current Thread [Next in Thread>