From mint-bounce@lists.fishpool.fi Tue Dec 22 15:31:38 2009 Date: Tue, 22 Dec 2009 21:26:28 +0100 (CET) From: arnaud.bercegeay@free.fr To: =?utf-8?Q?Vincent_Rivi=C3=A8re?= Cc: mint@lists.fishpool.fi Message-ID: <4257750.4409381261513588817.JavaMail.root@spooler5-g27.priv.proxad.net> In-Reply-To: <27658765.4409291261513523912.JavaMail.root@spooler5-g27.priv.proxad.net> Subject: Re: [MiNT] Possible bug in cflib's font selector MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-Originating-IP: [172.20.243.76] X-Mailer: Zimbra 5.0.11_GA_2627.UBUNTU8 (ZimbraWebClient - FF3.0 (Linux)/5.0.11_GA_2627.UBUNTU8) X-ecartis-version: Ecartis v1.0.0 Sender: mint-bounce@lists.fishpool.fi Errors-to: mint-bounce@lists.fishpool.fi X-original-sender: arnaud.bercegeay@free.fr Precedence: bulk List-help: List-unsubscribe: List-Id: X-List-ID: List-subscribe: List-owner: List-post: Hello, > I found some documentation here: > http://toshyp.atari.org/00d00e.htm#Pure-C_20and_20_27cdecl_27 > It says that the API is "Pure C cdecl". > Does this mean that the clobbered registers are d0/d1/a0/a1, like GCC > ? > Do you agree, Henk ? > > It this is right, d2 should not be present in the clobber lists, and > a1 > is missing. Even worse for xfsl_exit(). > > Am I right ? for cdecl functions, list of clobbered registers depends on the compiler. IIRC: - for gcc, it is D0/D1/A0/A1. - for purec, D2 should be added on this list (D0/D1/D2/A0/A1) - for sozobonc, A2 is in the list too. A "stdcall" layer has been added in LDG to allow branchement between code compiled by various compilers. see the comment in the source code : http://ldg.svn.sourceforge.net/viewvc/ldg/trunk/ldg/src/devel/stdcall.s?view=markup Arnaud.