From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stanislaw Gruszka Subject: Re: [RFC][PATCH] at91_ide driver Date: Thu, 22 Jan 2009 11:15:40 +0100 Message-ID: <200901221115.41110.stf_xl@wp.pl> References: <200901141345.42583.stf_xl@wp.pl> <200901211133.05095.stf_xl@wp.pl> <49783FEA.3010302@ru.mvista.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mx1.wp.pl ([212.77.101.5]:38611 "EHLO mx1.wp.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754154AbZAVKPy convert rfc822-to-8bit (ORCPT ); Thu, 22 Jan 2009 05:15:54 -0500 In-Reply-To: <49783FEA.3010302@ru.mvista.com> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Sergei Shtylyov Cc: Andrew Victor , Nicolas Ferre , Haavard Skinnemoen , linux-ide@vger.kernel.org Thursday 22 January 2009 10:44:10 Sergei Shtylyov napisa=C5=82(a): > Hello. >=20 > Stanislaw Gruszka wrote: >=20 > >> Besides, we have ide_timing_compute() doing the same thing. > >> =20 > > I'm trying use it, but have too bigger results - 1 or 2 cycles is a= dded. > > =20 >=20 > That's most probably because ide_timing_compute() assumes non-zero= =20 > minimum recovery time for PIO modes 0 to 2 (libata does the same) --=20 > that actually smells of over-caution. It then tries to stretch the=20 > active time if the sum of active and recovery times is less than cycl= e=20 > time (all quantized already). Hmm, why active time is stretched. IMHO it is better to stretch inactiv= e time (t2i), such it can be divided into setup (t1) and recovery time (t= 9 with t6z). Otherwise sum of t1 + t2 + max(t9, t2i - t1) will be longer than cycl= e time t0. Stanislaw Gruszka