From mboxrd@z Thu Jan 1 00:00:00 1970 Received: with ECARTIS (v1.0.0; list linux-mips); Mon, 07 Oct 2002 16:48:01 +0200 (CEST) Received: from buserror-extern.convergence.de ([212.84.236.66]:14084 "EHLO hell") by linux-mips.org with ESMTP id ; Mon, 7 Oct 2002 16:48:01 +0200 Received: from js by hell with local (Exim 3.35 #1 (Debian)) id 17yZAj-0004Mp-00 for ; Mon, 07 Oct 2002 16:47:49 +0200 Date: Mon, 7 Oct 2002 16:47:49 +0200 From: Johannes Stezenbach To: linux-mips@linux-mips.org Subject: Re: Once again: test_and_set for CPUs w/o LL/SC Message-ID: <20021007144749.GB16641@convergence.de> Mail-Followup-To: Johannes Stezenbach , linux-mips@linux-mips.org References: <20020916164034.GA12489@convergence.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20020916164034.GA12489@convergence.de> User-Agent: Mutt/1.4i Return-Path: X-Envelope-To: <"|/home/ecartis/ecartis -s linux-mips"> (uid 0) X-Orcpt: rfc822;linux-mips@linux-mips.org Original-Recipient: rfc822;linux-mips@linux-mips.org X-archive-position: 390 X-ecartis-version: Ecartis v1.0.0 Sender: linux-mips-bounce@linux-mips.org Errors-to: linux-mips-bounce@linux-mips.org X-original-sender: js@convergence.de Precedence: bulk X-list: linux-mips Hi all, On Mon, Sep 16, 2002 at 06:40:34PM +0200, I wrote: > > The NEC VR41xx CPU has no LL/SC instructions, so they must > be emulated by the kernel, which slows down the test-and-set > and compare-and-swap operations (used by linux-threads) > considerably. For the VR41xx (and other CPUs which have > branch-likely instructions), there exisits a workaround > which enables userspace-only atomic operations, with minor > help from the kernel: The kernel must guarantee that register > k1 is not equal to some magic value after every transition > to userspace. > > Two things were left open in July: > - find out the minimal amount of changes to the kernel > to guarantee k1 != MAGIC after eret > - determine how to tell glibc to use the branch-likely > workaround instead of emulated LL/SC Since there have been no follow-ups I must assume that this topic is no longer of interest. Is this so? Or is the way I approach it deemed inappropriate? Regards, Johannes