From mboxrd@z Thu Jan 1 00:00:00 1970 From: david mosberger Date: Wed, 07 Sep 2005 17:59:45 +0000 Subject: Re: [patch] delete sync.i in ia64_switch_to() Message-Id: List-Id: References: <200509070723.j877N9g25144@unix-os.sc.intel.com> In-Reply-To: <200509070723.j877N9g25144@unix-os.sc.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org On 9/7/05, Jim Hull wrote: > That sync.i is not old code - it is required. Yes, the sync.i was added by Asit at some point precisely for the reasons you explained. I believe we concluded at the time that the "mf" wasn't needed since we'd be guaranteed to have a fence in the execution-path anyhow (IIRC, it was due to locks being acquired/released on the paths in question). I'm not sure, however, whether this point has been considered when kernel-preemption support was added. The folks caring about CONFIG_PREEMPT may want to reconsider that issue if not. --david