From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: 2nd question on highmem in lldd Date: Sun, 13 Nov 2005 08:53:16 -0600 Message-ID: <1131893596.3536.1.camel@mulgrave> References: <200511131528.41121.oliver@neukum.org> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from stat9.steeleye.com ([209.192.50.41]:28338 "EHLO hancock.sc.steeleye.com") by vger.kernel.org with ESMTP id S932512AbVKMOxV (ORCPT ); Sun, 13 Nov 2005 09:53:21 -0500 In-Reply-To: <200511131528.41121.oliver@neukum.org> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Oliver Neukum Cc: linux-scsi@vger.kernel.org On Sun, 2005-11-13 at 15:28 +0100, Oliver Neukum wrote: > how are failures of dma_map_sg supposed to be reported to higher layers? return DID_REQUEUE in the host byte of the command result is probably closest to the behaviour you want: it will requeue until another command returns for the host which (presumably) frees up some IOMMU resources. James