From mboxrd@z Thu Jan 1 00:00:00 1970 From: Brent Casavant Date: Fri, 20 Jan 2006 16:14:21 +0000 Subject: Re: [PATCH] SN2 user-MMIO CPU migration Message-Id: <20060120101221.R91550@chenjesu.americas.sgi.com> List-Id: References: <20060118163305.Y42462@chenjesu.americas.sgi.com> <20060120083651.GA3970@elte.hu> In-Reply-To: <20060120083651.GA3970@elte.hu> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Ingo Molnar Cc: linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org, jes@sgi.com, tony.luck@intel.com On Fri, 20 Jan 2006, Ingo Molnar wrote: > hm, why isnt the synchronization done in switch_to()? Your arch-level > switch_to() could have something like thread->last_cpu_sync, and if > thread->last_cpu_sync != this_cpu, do the flush. This would not only > keep this stuff out of the generic scheduler, but it would also optimize > things a bit more: the moment we do a set_task_cpu() it does not mean > that CPU _will_ run the task. Another CPU could grab that task later on. Very good points all around. I'll rework the changes in just the manner you mentioned. Brent -- Brent Casavant All music is folk music. I ain't bcasavan@sgi.com never heard a horse sing a song. Silicon Graphics, Inc. -- Louis Armstrong