From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ns.suse.de ([195.135.220.2]:47297 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S964850AbWBGSbb (ORCPT ); Tue, 7 Feb 2006 13:31:31 -0500 From: Andi Kleen Subject: Re: [patch 1/1] Fix spinlock debugging delays to not time out too early Date: Tue, 7 Feb 2006 19:29:24 +0100 References: In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200602071929.24762.ak@muc.de> Sender: linux-arch-owner@vger.kernel.org To: "Luck, Tony" Cc: akpm@osdl.org, linux-arch@vger.kernel.org, mingo@elte.hu List-ID: On Tuesday 07 February 2006 19:20, Luck, Tony wrote: > > The fix is to include a __delay(1) call in the loop, to correctly approximate > > the intended delay timeout of 1 second. The code assumes that every > > architecture implements __delay(1) to last around 1/(loops_per_jiffy*HZ) > > seconds. > > But we calculate loops_per_jiffy based on somewhat larger vales passed > to __delay() ... where the function call overhead is amortized away. > > I'd expect that __delay(1) would last quite a bit longer than > 1/(loops_per_jiffy * HZ) ... so we'll wait a lot more than a > second. For this case that would be ok, as long as it isn't a hour or so, but let's say < 1 minute. -Andi