Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id <S284264AbSAPRZG>; Wed, 16 Jan 2002 12:25:06 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org
	id <S284285AbSAPRZA>; Wed, 16 Jan 2002 12:25:00 -0500
Received: from ns.ithnet.com ([217.64.64.10]:26635 "HELO heather.ithnet.com")
	by vger.kernel.org with SMTP id <S284987AbSAPRYT>;
	Wed, 16 Jan 2002 12:24:19 -0500
Date: Wed, 16 Jan 2002 18:24:09 +0100
From: Stephan von Krawczynski <skraw@ithnet.com>
To: Gerd Knorr <kraxel@bytesex.org>
Cc: linux-kernel@vger.kernel.org
Subject: Re: Memory problem with bttv driver
Message-Id: <20020116182409.326473de.skraw@ithnet.com>
In-Reply-To: <20020116105747.B1190@bytesex.org>
In-Reply-To: <20020114210039.180c0438.skraw@ithnet.com>
	<E16QETz-0002yD-00@the-village.bc.nu>
	<20020115004205.A12407@werewolf.able.es>
	<slrna480cv.68d.kraxel@bytesex.org>
	<20020115121424.10bb89b2.skraw@ithnet.com>
	<20020115142017.D8191@bytesex.org>
	<20020116105747.B1190@bytesex.org>
Organization: ith Kommunikationstechnik GmbH
X-Mailer: Sylpheed version 0.7.0 (GTK+ 1.2.10; i686-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org

On Wed, 16 Jan 2002 10:57:47 +0100
Gerd Knorr <kraxel@bytesex.org> wrote:

> > > > > Yes.  Instead of remapping vmalloced kernel memory it gives you
shared> > > > > anonymous pages, then does zerocopy DMA using kiobufs.  You may
run in> > > > > trouble with >4GB machines.
> > > > 
> > > > Interesting.
> > > > What's the problem on > 4GB ?
> > > 
> > > The bt878/848 is a 32bit PCI device, it simply can't go DMA to main
> > > memory above 4GB.  At least on ia32, on architecures with a iommu
> > > (sparc, ...) it should work without trouble.
> > 
> > Sorry, maybe I should have clarified: what is the problem with allocating
those> > pages below 4GB in a >4GB box?
> 
> do_anonymous_page() may give you pages above 4GB.

Guess it is a bit ... flaky ... to just do a

                page = alloc_page(GFP_HIGHUSER);
                if (!page)
                        goto no_mem;

in

do_anonymous_page()

I guess it should be somehow configurable (parameter). Your idea of using it
may grow and mem too, for sure.

Regards,
Stephan


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/