From mboxrd@z Thu Jan 1 00:00:00 1970 From: Todd E Brandt Subject: Re: [PATCH/RESEND v2 1/2] Hard disk S3 resume time optimization Date: Mon, 13 Jan 2014 12:06:48 -0800 Message-ID: <20140113200648.GB13900@linux.intel.com> References: <20140108005607.GB29556@linux.intel.com> <52CED67E.7080706@ubuntu.com> <11E08D716F0541429B7042699DD5C1A17073C362@FMSMSX103.amr.corp.intel.com> <52D0A8B4.90905@ubuntu.com> Reply-To: todd.e.brandt@linux.intel.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mga09.intel.com ([134.134.136.24]:26802 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751353AbaAMUH2 (ORCPT ); Mon, 13 Jan 2014 15:07:28 -0500 Content-Disposition: inline In-Reply-To: Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Dan Williams Cc: tj@kernel.org, James.Bottomley@hansenpartnership.com, linux-ide@vger.kernel.org, linux-scsi@vger.kernel.org, psusi@ubuntu.com On Fri, Jan 10, 2014 at 07:13:11PM -0800, Dan Williams wrote: > On Fri, Jan 10, 2014 at 6:13 PM, Phillip Susi wrote: > > -----BEGIN PGP SIGNED MESSAGE----- > > Hash: SHA512 > > > > On 01/10/2014 06:11 PM, Brandt, Todd E wrote: > >> Yes yours is simpler, but it also opens a potential memory issue > >> by passing a static int as the return location for the error value. > >> I think it's just safer to tell the callback to attempt no return > >> value at all, and for that you need to expand it into two > >> arguments, one for selection, the other for the output address. > > > > What sort of memory issue? Also isn't there a system NULL page > > somewhere that could be used? > > > > I think the static variable is ok. We can be sure that all eh threads > are torn down before libata.ko is unloaded. Actually there's one other reason. In the ata_port_request_pm function it checks to see if there's a previous resume operation pending, and if so it calls ata_port_wait_eh in order to wait for it to complete before issuing the new suspend. If you just use the (int*)async parameter it will return immediately and defer to the caller to try again, like is does with SAS. But in our case we *don't* try again, so it would result in the resume being skipped. There needs to be a new case where the caller wants the call to be asynchronous, and it wants ata_port_request_pm to do its own waiting, but doesn't care about the return value. Thus the additional parameter.