From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 28 Sep 2001 11:30:58 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 28 Sep 2001 11:30:48 -0400 Received: from ns.caldera.de ([212.34.180.1]:27050 "EHLO ns.caldera.de") by vger.kernel.org with ESMTP id ; Fri, 28 Sep 2001 11:30:27 -0400 Date: Fri, 28 Sep 2001 17:30:37 +0200 Message-Id: <200109281530.f8SFUbN18754@ns.caldera.de> From: Christoph Hellwig To: n.roos@berlin.de (Norbert Roos) Cc: linux-kernel@vger.kernel.org, kernelnewbies@nl.linux.org Subject: Re: inverse mmap() available? X-Newsgroups: caldera.lists.linux.kernel In-Reply-To: <3BB49438.1F2C59B3@berlin.de> User-Agent: tin/1.4.4-20000803 ("Vet for the Insane") (UNIX) (Linux/2.4.2 (i686)) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org In article <3BB49438.1F2C59B3@berlin.de> you wrote: > Hello! > > Is there a way to map user space memory into kernel address space, e.g. > that i don't have to call get_user(var,...) but simply use var = > *user_space_ptr? > > What i intend to do (as the next step) is to DMA-transfer data directly > between a PCI device and user space memory. The buffer in user space > should be allocated with malloc(), so allocating a buffer in kernel and > mmap()-ping it to user space is not the solution.. > > I guess this has been asked before, any links to further information > would be great. map_user_kiobuf is what you want :) Some more explanation at: ftp://ftp.openlinux.org/pub/people/hch/kio/kiobuf.txt