From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 29213 invoked by alias); 13 Jun 2007 15:05:09 -0000 Received: (qmail 29202 invoked by uid 22791); 13 Jun 2007 15:05:08 -0000 X-Spam-Check-By: sourceware.org Received: from cmburns.debconf.org (HELO cmburns.debconf.org) (80.68.86.241) by sourceware.org (qpsmtpd/0.31) with ESMTP; Wed, 13 Jun 2007 15:05:04 +0000 Received: from homer.dc7.debconf.org (unknown [129.215.255.137]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client CN "homer.dc7.debconf.org", Issuer "Certificate Authority - Email" (verified OK)) by cmburns.debconf.org (Postfix) with ESMTP id DA00D1C805F; Wed, 13 Jun 2007 15:05:00 +0000 (UTC) Received: from [10.8.249.248] (helo=lagash ident=Debian-exim) by homer.dc7.debconf.org with esmtpsa (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.63) (envelope-from ) id 1HyUP6-0006Kq-G0; Wed, 13 Jun 2007 16:05:00 +0100 Received: from ths by lagash with local (Exim 4.67) (envelope-from ) id 1HyUP8-0004d7-Bh; Wed, 13 Jun 2007 16:05:02 +0100 Date: Wed, 13 Jun 2007 15:05:00 -0000 To: "Maciej W. Rozycki" Cc: Daniel Jacobowitz , gdb-patches@sourceware.org Subject: Re: [rfc] Target-described register support for MIPS Message-ID: <20070613150502.GC8584@networkno.de> References: <20070521133939.GA8161@caradoc.them.org> <20070612182517.GA12526@caradoc.them.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.13 (2006-08-11) From: Thiemo Seufer X-IsSubscribed: yes Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2007-06/txt/msg00234.txt.bz2 Maciej W. Rozycki wrote: > On Tue, 12 Jun 2007, Daniel Jacobowitz wrote: > > > > Hmm, these look wrong -- no "ta" registers have been defined for old > > > ABIs. > > > > OK, I'll drop them. They came from O32_SYMBOLIC_REGISTER_NAMES in > > gas/config/tc-mips.c. > > Hmm, it looks like they may have been defined so that some weird > handcoded assembly works for both o32 and n32 by using $t0-$t3 and > $ta0-$ta3 -- note how the definitions are swapped for the two ABIs. > > But the names make no sense for o32 -- "ta" stands for > "temporary/argument" and these are obviously not argument registers. I do > not recall seeing any code using these names anyway and portable code will > probably refer to the registers by numbers as $4-$15 (once you have > recorded or discarded the arguments you are free to use all of these > registers as temporaries anyway), possibly hiding them under some > preprocessor macros. And GCC does not care. The "ta" aliases are supposed to help with writing multi-ABI assembler code, and are occasionally used for that purpose. I don't think they are useful for disassembling. When gdb knows the ABI it should IMHO display the "real" names instead of the "ta" aliases. Thiemo