From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stanislaw Gruszka Subject: Re: [PATCH 3/3 v3] AT91: initialize Compact Flash on AT91SAM9263 cpu Date: Fri, 13 Feb 2009 11:48:08 +0100 Message-ID: <200902131148.08728.stf_xl@wp.pl> References: <200902091145.11266.stf_xl@wp.pl> <200902121005.46772.stf_xl@wp.pl> <49944F44.1080807@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]:61306 "EHLO mx1.wp.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751013AbZBMKs2 convert rfc822-to-8bit (ORCPT ); Fri, 13 Feb 2009 05:48:28 -0500 In-Reply-To: <49944F44.1080807@ru.mvista.com> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Sergei Shtylyov Cc: linux-ide@vger.kernel.org, Andrew Victor , linux-arm-kernel@lists.arm.linux.org.uk Hi. Thursday 12 February 2009 17:33:08 Sergei Shtylyov napisa=C5=82(a): > >> I'm still not sure why this is needed. Do you think thta power-o= n=20 > >>reset is not enough? >=20 > > HW reset is not needed when board is powered-on. But situation is > > diffrent when we have reboot due to board reset booton press. > > There are configuration where signal from reset button is connected > > to CF card (or disk), in some cases device reset signal is connecte= d to > > GPIO with assumption to be asserted by software. >=20 > OK, thanks for the explanation. >=20 > > I don't know if we can remove this. >=20 > Well, seems worth keeping. >=20 > > In hw reset case system/driver will run with undefined device state= , >=20 > What do you mean by undefined? It could have not come out of rese= t which=20 > shoiuld mean that the IDE probing code will have to wait some time be= fore BSY=20 > clears -- see ide_port_wait_ready(). >=20 > > in example device can generate interrupts. >=20 > Interrupts after any kind of reset? That just shouldn't happen --= it's=20 > against the ATA spec at least. I mean undefined state and interrupts where is _no_ IDE device reset and system boots because of board button h/w reset (or watchdog reset). Ok, IDE layer will cope with any registers state during initialization,= but when not handled interrupt is generated we have spurious interrupt which is some trouble for kernel. Cheers Stanislaw Gruszka