From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 25 May 2004 10:25:47 -0700 From: "David S. Miller" Subject: Re: [PATCH] ppc64: Fix possible race with set_pte on a present PTE Message-Id: <20040525102547.35207879.davem@redhat.com> In-Reply-To: References: <1085369393.15315.28.camel@gaston> <1085371988.15281.38.camel@gaston> <1085373839.14969.42.camel@gaston> <20040525034326.GT29378@dualathlon.random> <20040525114437.GC29154@parcelfarce.linux.theplanet.co.uk> <20040525153501.GA19465@foobazco.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Linus Torvalds Cc: wesolows@foobazco.org, willy@debian.org, andrea@suse.de, benh@kernel.crashing.org, akpm@osdl.org, linux-kernel@vger.kernel.org, mingo@elte.hu, bcrl@kvack.org, linux-mm@kvack.org, linux-arch@vger.kernel.org List-ID: On Tue, 25 May 2004 09:19:52 -0700 (PDT) Linus Torvalds wrote: > On Tue, 25 May 2004, Keith M Wesolowski wrote: > > > > Some sparc32 CPUs are also vulnerable to this race; in fact the > > supersparc manual describes it specifically and even outlines the > > compare-exchange loop using our rotten swap instruction. In our case, > > the race is with a hardware walker. > > Yes, but the sparc32 page tables are not the same as the linux kernel page > tables, so in your case it's a different path and a different page table. > Only the shared case really matters (ie things that do hw/microcode walk > of a page table _tree_ not a hash). Not true on 32-bit Sparc sun4m systems, it's exactly like i386 except the hardware is stupid and we only have an atomic swap instruction. :-) -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: aart@kvack.org