From mboxrd@z Thu Jan 1 00:00:00 1970 From: Robert Hancock Subject: Re: sata_sil & pmp ? Date: Sat, 11 Apr 2009 12:10:33 -0600 Message-ID: <49E0DD19.2080603@gmail.com> References: <49DF216D.6080208@gotshifted.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from yw-out-2324.google.com ([74.125.46.28]:26549 "EHLO yw-out-2324.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757731AbZDKSKi (ORCPT ); Sat, 11 Apr 2009 14:10:38 -0400 Received: by yw-out-2324.google.com with SMTP id 5so1593778ywb.1 for ; Sat, 11 Apr 2009 11:10:37 -0700 (PDT) In-Reply-To: <49DF216D.6080208@gotshifted.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: hasse@gotshifted.com Cc: Jeff Garzik , linux-ide@vger.kernel.org Hans Holmberg wrote: > I wonder if there is any plans to add pmp support to the sata_sil module ? > > I know there's an ugly patch that kind of works, but it seems nothing > has happened since it was posted. It looks like the controller does have some support for at least setting the PMP field in the transmitted FIS. I'm not sure if it includes everything that's needed for full PMP support though. I saw that there was some talk of a patch to add this, but I'm not sure how fully working it was. The Silicon Image 311x controllers are kind of funny beasts. Technically since they support transparent FIS reception and interlock for all FIS types, you could implement just about any SATA feature such as NCQ, ATAPI AN, etc. by overriding the handling of the required FISes. NCQ might be kind of expensive though, as you'd likely have to take a whole lot of extra interrupts for FISes to examine to get that to work, you'd basically be doing most of the SATA state machine in software..