From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [parisc-linux] [PATCH] timer_interrupt and gettimeoffset. Date: Sun, 03 Sep 2006 11:14:53 -0500 Message-ID: <1157300093.3486.15.camel@mulgrave.il.steeleye.com> References: <20060830165211.GA3999@colo.lackof.org> <119aab440608301323w309debf2g6635ce4757ac024b@mail.gmail.com> <1157212326.4041.26.camel@mulgrave.il.steeleye.com> <119aab440609030800h264c16bcn2958a78558a2ef48@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain Cc: Joel Soete , parisc-linux To: "Carlos O'Donell" Return-Path: In-Reply-To: <119aab440609030800h264c16bcn2958a78558a2ef48@mail.gmail.com> List-Id: parisc-linux developers list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: parisc-linux-bounces@lists.parisc-linux.org On Sun, 2006-09-03 at 11:00 -0400, Carlos O'Donell wrote: > Your guess that this is the solution to problem 2) is not quite right. > I will argue that the "halftick" counting is at fault here. When a > "halftick" is counted as a full tick the system time moves into the > future. All calls to gettimeoffset must therefore return negative > adjustments. I think removing the erroneous halftick counting is the > solution to issue 2. Actually, I think we just accept it as a feature rather than a bug. Since the timer interrupts miss so many ticks, halftick processing looks like a good idea to me. Thus, we have to assume gettimeoffset() can be negative by up to half a tick. The current processing in do_gettimeofday() already copes nicely with this. James _______________________________________________ parisc-linux mailing list parisc-linux@lists.parisc-linux.org http://lists.parisc-linux.org/mailman/listinfo/parisc-linux