From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261975AbULKRae (ORCPT ); Sat, 11 Dec 2004 12:30:34 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261976AbULKRae (ORCPT ); Sat, 11 Dec 2004 12:30:34 -0500 Received: from mailout.stusta.mhn.de ([141.84.69.5]:34313 "HELO mailout.stusta.mhn.de") by vger.kernel.org with SMTP id S261975AbULKRa3 (ORCPT ); Sat, 11 Dec 2004 12:30:29 -0500 Date: Sat, 11 Dec 2004 18:30:16 +0100 From: Adrian Bunk To: Alan Cox Cc: Bartlomiej Zolnierkiewicz , Linux Kernel Mailing List Subject: [2.6 patch] IDE: remove WAIT_READY dependency on APM Message-ID: <20041211173016.GX22324@stusta.de> References: <20041209034438.GF22324@stusta.de> <1102692003.3201.5.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1102692003.3201.5.camel@localhost.localdomain> User-Agent: Mutt/1.5.6+20040907i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Dec 10, 2004 at 03:20:05PM +0000, Alan Cox wrote: > On Iau, 2004-12-09 at 03:44, Adrian Bunk wrote: > > The time for the !APM case was already increased from 30msec in 2.4 . > > Isn't there a timeout that is suitable for all cases? > > The five seconds should be just fine for all cases. The smaller value > with no > power manglement should help speed up recovery however. It probably > doesn't belong CONFIG_APM now ACPI and friends are involved either. Thanks for this information. A patch is below. Signed-off-by: Adrian Bunk --- linux-2.6.10-rc2-mm4-full/include/linux/ide.h.old 2004-12-11 18:11:20.000000000 +0100 +++ linux-2.6.10-rc2-mm4-full/include/linux/ide.h 2004-12-11 18:11:32.000000000 +0100 @@ -187,11 +187,7 @@ * Timeouts for various operations: */ #define WAIT_DRQ (HZ/10) /* 100msec - spec allows up to 20ms */ -#if defined(CONFIG_APM) || defined(CONFIG_APM_MODULE) #define WAIT_READY (5*HZ) /* 5sec - some laptops are very slow */ -#else -#define WAIT_READY (HZ/10) /* 100msec - should be instantaneous */ -#endif /* CONFIG_APM || CONFIG_APM_MODULE */ #define WAIT_PIDENTIFY (10*HZ) /* 10sec - should be less than 3ms (?), if all ATAPI CD is closed at boot */ #define WAIT_WORSTCASE (30*HZ) /* 30sec - worst case when spinning up */ #define WAIT_CMD (10*HZ) /* 10sec - maximum wait for an IRQ to happen */