From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752785Ab1KMChm (ORCPT ); Sat, 12 Nov 2011 21:37:42 -0500 Received: from casper.infradead.org ([85.118.1.10]:50086 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752139Ab1KMChl (ORCPT ); Sat, 12 Nov 2011 21:37:41 -0500 Date: Sat, 12 Nov 2011 18:37:35 -0800 From: Arjan van de Ven To: Xin Tong Cc: linux-kernel Subject: Re: INVLPG Instruction Message-ID: <20111112183735.333b747f@infradead.org> In-Reply-To: References: Organization: Intel X-Mailer: Claws Mail 3.7.9 (GTK+ 2.24.4; i386-redhat-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-SRS-Rewrite: SMTP reverse-path rewritten from by casper.infradead.org See http://www.infradead.org/rpr.html Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, 12 Nov 2011 19:59:44 -0500 Xin Tong wrote: > Hello > > When a process has a page fault and Linux swaps out a page ( no more > free page ) from another process and give it to the page-faulting > process. Typically a INVLPG is used to invalidate entries in the TLB. > I would like to know whether INVLPG is global ( i.e. across all cores > )? How does the process which Linux stole the page from get informed ( > how is its tlb invalidated) ? invlpg is not "global" in the way you mean it (nor in the sense of "global pte's"). The kernel will do an IPI to the processors that have a possibility to have this page in the TLB and do a local flush there. -- Arjan van de Ven Intel Open Source Technology Centre For development, discussion and tips for power savings, visit http://www.lesswatts.org