From mboxrd@z Thu Jan 1 00:00:00 1970 From: "George Spelvin" Subject: Re: [RFC PATCH 1/2] arch/m68k/lib/mulsi3.S: Optimize] Date: 12 May 2016 16:52:17 -0400 Message-ID: <20160512205217.19233.qmail@ns.horizon.com> References: <57347B1E.9030108@westnet.com.au> Return-path: Received: from ns.horizon.com ([71.41.210.147]:51233 "HELO ns.horizon.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1751902AbcELUwT (ORCPT ); Thu, 12 May 2016 16:52:19 -0400 In-Reply-To: <57347B1E.9030108@westnet.com.au> Sender: linux-m68k-owner@vger.kernel.org List-Id: linux-m68k@vger.kernel.org To: geert@linux-m68k.org, gregungerer@westnet.com.au, linux-m68k@lists.linux-m68k.org, linux@horizon.com Thank you very much! Greg Ungerer wrote: > This syntax fails for me (using a binutils-2.25.1 based toolchain). > Registers must be prefixed with a "%", so here %sp and %a0. > > arch/m68k/lib/mulsi3.S: Assembler messages: > arch/m68k/lib/mulsi3.S:12: Error: syntax error -- statement `lea 4(sp),a0' ignored > > That was compiling with just this one patch for a ColdFire target. Well, *that* is an embarrassing oversight. If you fix that obvious typo (I used ":%s/[ad][01]/%&/g"), do you have a way of testing it? Setting up a suitable Qemu environment is many times the effort needed to write the code, so I was hoping someone with the facilities already in place would be willing to test. I've stared at the code and am convinced it's right, but I remember Knuth's words: "Beware of bugs in the above code; I have only proved it correct, not tried it."