From mboxrd@z Thu Jan 1 00:00:00 1970 Received: with ECARTIS (v1.0.0; list linux-mips); Tue, 04 Nov 2008 07:16:49 +0000 (GMT) Received: from qmta07.westchester.pa.mail.comcast.net ([76.96.62.64]:11999 "EHLO QMTA07.westchester.pa.mail.comcast.net") by ftp.linux-mips.org with ESMTP id S23098494AbYKDHQq (ORCPT ); Tue, 4 Nov 2008 07:16:46 +0000 Received: from OMTA08.westchester.pa.mail.comcast.net ([76.96.62.12]) by QMTA07.westchester.pa.mail.comcast.net with comcast id av791a0030Fqzac57vGB3i; Tue, 04 Nov 2008 07:16:11 +0000 Received: from [192.168.1.13] ([69.140.18.238]) by OMTA08.westchester.pa.mail.comcast.net with comcast id avGf1a00G58Be2l3UvGfLV; Tue, 04 Nov 2008 07:16:40 +0000 X-Authority-Analysis: v=1.0 c=1 a=se7utVhDntQA:10 a=sj6Exy8oVBoA:10 a=MYXyyX_-kbTf0ocPCMQA:9 a=y_SrRMj2HFMETmdtJKYA:7 a=P6CIcvTiMdE7gXlkRojVOyVIo3UA:4 a=WeOa-AV5lc8A:10 a=QJAqVYndk0IA:10 Message-ID: <490FF6B3.9030906@gentoo.org> Date: Tue, 04 Nov 2008 02:16:03 -0500 From: Kumba User-Agent: Thunderbird 2.0.0.17 (Windows/20080914) MIME-Version: 1.0 To: libc-ports@sources.redhat.com CC: Ralf Baechle , Daniel Jacobowitz , Linux MIPS List Subject: Re: [PATCH]: R10000 Needs LL/SC Workaround in Glibc References: <490A912A.8030901@gentoo.org> <20081101112643.GA2249@linux-mips.org> In-Reply-To: <20081101112643.GA2249@linux-mips.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit 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: 21180 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: kumba@gentoo.org Precedence: bulk X-list: linux-mips Ralf Baechle wrote: > > In the kernel we have very good knowledge about what types of processors > are being used for what configuration; much less in userland and the code > as suggested by you would result in a silent failure on affected R10000 > machines if version built not for the R10000 was being used - iow no > improvment over what we have right now. So for userland I'd prefer to > > o MIPS I builds: use the some 28 nops. > o Builds for MIPS II or better: always use the branch likely > o A runtime test would have to be implemented pessimisticall because it > would have to rely on /proc being mounted which isn't available early in > the boot process. It's probably going to add more overhead than it > saves anyway. > > There is a price for using branch likely - but not that high. In the grand > picture it'll almost certainly vanish in the benchmarking noise. Good idea. I'll tinker with this once I wrap my head around the gcc-side of things. -- Joshua Kinard Gentoo/MIPS kumba@gentoo.org "The past tempts us, the present confuses us, the future frightens us. And our lives slip away, moment by moment, lost in that vast, terrible in-between." --Emperor Turhan, Centauri Republic