From owner-cvs-src@FreeBSD.ORG Sat Jun 28 13:52:19 2008 Return-Path: Delivered-To: cvs-src@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id CB6421065679 for ; Sat, 28 Jun 2008 13:52:19 +0000 (UTC) (envelope-from christoph.mallon@gmx.de) Received: from mail.gmx.net (mail.gmx.net [213.165.64.20]) by mx1.freebsd.org (Postfix) with SMTP id 1E37C8FC14 for ; Sat, 28 Jun 2008 13:52:18 +0000 (UTC) (envelope-from christoph.mallon@gmx.de) Received: (qmail invoked by alias); 28 Jun 2008 13:52:17 -0000 Received: from p54A3DE64.dip.t-dialin.net (EHLO tron.homeunix.org) [84.163.222.100] by mail.gmx.net (mp015) with SMTP; 28 Jun 2008 15:52:17 +0200 X-Authenticated: #1673122 X-Provags-ID: V01U2FsdGVkX18iNhld5FkismmHN+Gyx3CvHQqBomOZbt1PgS99TR GO2bNSe3JBdPyQ Message-ID: <48664210.80204@gmx.de> Date: Sat, 28 Jun 2008 15:52:16 +0200 From: Christoph Mallon User-Agent: Thunderbird 2.0.0.9 (X11/20071230) MIME-Version: 1.0 To: Marius Strobl References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628114417.GL1215@alchemy.franken.de> <486629AA.1050409@gmx.de> <20080628132241.GO1215@alchemy.franken.de> In-Reply-To: <20080628132241.GO1215@alchemy.franken.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 Cc: cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/sparc64/include in_cksum.h X-BeenThere: cvs-src@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: CVS commit messages for the src tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 28 Jun 2008 13:52:19 -0000 Marius Strobl wrote: > On Sat, Jun 28, 2008 at 02:08:10PM +0200, Christoph Mallon wrote: >> Marius Strobl wrote: >>>> On a related note: Is inline assembler really necessary here? For >>>> example couldn't in_addword() be written as >>>> static __inline u_short >>>> in_addword(u_short const sum, u_short const b) >>>> { >>>> u_int const t = sum + b; >>>> return t + (t >> 16); >>>> } ? >>>> This should at least produce equally good code and because the compiler >>>> has more knowledge about it than an assembler block, it potentially >>>> leads to better code. I have no SPARC compiler at hand, though. >>> With GCC 4.2.1 at -O2 the code generated for the above C version >>> takes on more instruction than the inline assembler so if one >> On SPARC? What code does it produce? I have not SPARC compiler at hand. >> Even if it is one more instruction, I think the reduced register >> pressure makes more than up for it. > > Correct, it only uses two registers: > > 0000000000000000 : > 0: 92 02 00 09 add %o0, %o1, %o1 > 4: 91 32 60 10 srl %o1, 0x10, %o0 > 8: 90 02 00 09 add %o0, %o1, %o0 > c: 91 2a 20 10 sll %o0, 0x10, %o0 > 10: 91 32 20 10 srl %o0, 0x10, %o0 > 14: 81 c3 e0 08 retl > 18: 91 3a 20 00 sra %o0, 0, %o0 > 1c: 01 00 00 00 nop One more instruction? That's five instructions for the actual calculation afaict, just like the inline assembler version. The sra in the delay slot should be present in the inline assembler version, too. >> This should work fine and only use two registers (though the compiler >> can choose to use three, if it deems it beneficial): >> >> static __inline u_short >> in_addword(u_short const sum, u_short const b) >> { >> u_long const sum16 = sum << 16; >> u_long const b16 = b << 16; >> u_long ret; >> >> __asm( >> "addcc %1, %2, %0\n\t" >> "srl %0, 16, %0\n\t" >> "addc %0, 0, %0\n" >> : "=r" (ret) : "r" (sum16), "r" (b16) : "cc"); >> >> return (ret); >> } > > This is ten instructions with two registers. Where is the > break even regarding instructions vs. registers for sparc64? :) I still have no SPARC compiler. Ten instructions? All I did was write the two shifts in C and adjust the register constraints. It should produce identical code. >> But I still prefer the C version. >> > > And I prefer to not re-write otherwise working code for > micro-optimizations, there are enough unfixed real bugs Obviously the inline assembler magic did not work and is/was a real bug. > to deal with. Similarly we should not waste time discussing > how to possibly optimize MD versions even more but rather > spend the time improving the MI version so it's good enough > that using MD versions isn't worth the effort. The C alternative is MI and in length on par with the inline assembler version, isn't it? Regards Christoph