From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265243AbUD3T5R (ORCPT ); Fri, 30 Apr 2004 15:57:17 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265244AbUD3T5R (ORCPT ); Fri, 30 Apr 2004 15:57:17 -0400 Received: from vena.lwn.net ([206.168.112.25]:55465 "HELO lwn.net") by vger.kernel.org with SMTP id S265243AbUD3T5N (ORCPT ); Fri, 30 Apr 2004 15:57:13 -0400 Message-ID: <20040430195711.16275.qmail@lwn.net> To: linux-kernel@vger.kernel.org Subject: [MICROPATCH] Make x86_64 build work without GART_IOMMU cc: ak@suse.de From: Jonathan Corbet Date: Fri, 30 Apr 2004 13:57:11 -0600 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org If you try to build a 2.6.6-rc3 kernel with CONFIG_GART_IOMMU turned off, the link fails because bad_dma_address is undefined. This little hack just defines the variable in pci-nommu.c as well. It may not be an optimal fix, but it does make the kernel build. Why do I care? My Radeon 9200SE goes nuts if I build a GART-enabled kernel. Haven't figured out why... jon --- 2.6.6-rc3-slab/arch/x86_64/kernel/pci-nommu.c.orig 2004-02-06 00:51:20.000000000 -0700 +++ 2.6.6-rc3-slab/arch/x86_64/kernel/pci-nommu.c 2004-04-30 13:40:35.000000000 -0600 @@ -5,6 +5,7 @@ #include int iommu_merge = 0; +dma_addr_t bad_dma_address; /* * Dummy IO MMU functions