Path: news.gmane.org!not-for-mail From: Anton Blanchard Newsgroups: gmane.linux.network Subject: [PATCH 2/4] ibmveth: Fix issue with DMA mapping failure Date: Thu, 08 Sep 2011 10:41:04 +1000 Lines: 47 Approved: news@gmane.org Message-ID: <20110908004121.669225190@samba.org> References: <20110908004102.355674129@samba.org> NNTP-Posting-Host: lo.gmane.org X-Trace: dough.gmane.org 1315442760 18037 80.91.229.12 (8 Sep 2011 00:46:00 GMT) X-Complaints-To: usenet@dough.gmane.org NNTP-Posting-Date: Thu, 8 Sep 2011 00:46:00 +0000 (UTC) Cc: netdev@vger.kernel.org To: Santiago Leon , brking@linux.vnet.ibm.com, rcj@linux.vnet.ibm.com Original-X-From: netdev-owner@vger.kernel.org Thu Sep 08 02:45:55 2011 Return-path: Envelope-to: linux-netdev-2@lo.gmane.org Original-Received: from vger.kernel.org ([209.132.180.67]) by lo.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1R1Ska-0001cA-M3 for linux-netdev-2@lo.gmane.org; Thu, 08 Sep 2011 02:45:53 +0200 Original-Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757665Ab1IHApa (ORCPT ); Wed, 7 Sep 2011 20:45:30 -0400 Original-Received: from ozlabs.org ([203.10.76.45]:60840 "EHLO ozlabs.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757659Ab1IHAp1 (ORCPT ); Wed, 7 Sep 2011 20:45:27 -0400 Original-Received: from localhost (ppp121-44-79-234.lns20.syd6.internode.on.net [121.44.79.234]) (using TLSv1 with cipher DHE-RSA-AES128-SHA (128/128 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTPSA id 56E01B6F97; Thu, 8 Sep 2011 10:45:25 +1000 (EST) X-Mailbox-Line: From anton@samba.org Thu Sep 8 10:41:21 2011 User-Agent: quilt/0.48-1 Content-Disposition: inline; filename=ibmveth_dma_mapping_error.patch Original-Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Xref: news.gmane.org gmane.linux.network:205931 Archived-At: descs[].fields.address is 32bit which truncates any dma mapping errors so dma_mapping_error() fails to catch it. Use a dma_addr_t to do the comparison. With this patch I was able to transfer many gigabytes of data with IOMMU fault injection set at 10% probability. Signed-off-by: Anton Blanchard Cc: # v2.6.37+ --- Index: linux-build/drivers/net/ibmveth.c =================================================================== --- linux-build.orig/drivers/net/ibmveth.c 2011-09-01 15:01:18.066844039 +1000 +++ linux-build/drivers/net/ibmveth.c 2011-09-01 15:03:34.299079796 +1000 @@ -930,6 +930,7 @@ static netdev_tx_t ibmveth_start_xmit(st union ibmveth_buf_desc descs[6]; int last, i; int force_bounce = 0; + dma_addr_t dma_addr; /* * veth handles a maximum of 6 segments including the header, so @@ -994,17 +995,16 @@ retry_bounce: } /* Map the header */ - descs[0].fields.address = dma_map_single(&adapter->vdev->dev, skb->data, - skb_headlen(skb), - DMA_TO_DEVICE); - if (dma_mapping_error(&adapter->vdev->dev, descs[0].fields.address)) + dma_addr = dma_map_single(&adapter->vdev->dev, skb->data, + skb_headlen(skb), DMA_TO_DEVICE); + if (dma_mapping_error(&adapter->vdev->dev, dma_addr)) goto map_failed; descs[0].fields.flags_len = desc_flags | skb_headlen(skb); + descs[0].fields.address = dma_addr; /* Map the frags */ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) { - unsigned long dma_addr; skb_frag_t *frag = &skb_shinfo(skb)->frags[i]; dma_addr = dma_map_page(&adapter->vdev->dev, frag->page,