From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH] libata: implement per-dev EH action mask eh_info->dev_action[] Date: Tue, 20 Jun 2006 17:08:32 +0900 Message-ID: <4497AD00.6000703@gmail.com> References: <20060619092723.GF11755@htj.dyndns.org> <1150786964.7132.117.camel@forrest26.sh.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from nz-out-0102.google.com ([64.233.162.196]:32077 "EHLO nz-out-0102.google.com") by vger.kernel.org with ESMTP id S964991AbWFTIIb (ORCPT ); Tue, 20 Jun 2006 04:08:31 -0400 Received: by nz-out-0102.google.com with SMTP id s18so1941335nze for ; Tue, 20 Jun 2006 01:08:30 -0700 (PDT) In-Reply-To: <1150786964.7132.117.camel@forrest26.sh.intel.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: "zhao, forrest" Cc: Jeff Garzik , linux-ide@vger.kernel.org zhao, forrest wrote: > On Mon, 2006-06-19 at 18:27 +0900, Tejun Heo wrote: >> Note that per-dev actions >> can still be specified at port-level and it has the same effect of >> specifying the action for all devices on the port. >> > > Why not use eh_info->action for port-level only, and eh_info->dev_action > [] for device-level only? >>>From my understanding, this clear separation can at least make code easy > to read. The distinction between port-wide (later it becomes link-wide) and device-wide EH operations is meaningful only for PATA devices. For a SATA LLD, device-wide operations is port-wide operation and using different fields doesn't make much sense. So, I chose to put a bit more complexity where those per-dev actions are implemented while allowing LLDs and other parts of EH not to distinguish between those two unless it specifically wants to. -- tejun