From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 28889 invoked by alias); 17 May 2002 13:51:42 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 28845 invoked from network); 17 May 2002 13:51:38 -0000 Received: from unknown (HELO dell-pe2450-3.cambridge.redhat.com) (195.224.55.225) by sources.redhat.com with SMTP; 17 May 2002 13:51:38 -0000 Received: from north-pole.nickc.cambridge.redhat.com (host217-39-2-165.in-addr.btopenworld.com [217.39.2.165]) by dell-pe2450-3.cambridge.redhat.com (Postfix) with ESMTP id E7681857E6 for ; Fri, 17 May 2002 14:51:37 +0100 (BST) Received: from north-pole.nickc.cambridge.redhat.com.nickc.cambridge.redhat.com (localhost [127.0.0.1]) by north-pole.nickc.cambridge.redhat.com (Postfix) with ESMTP id 33C971C39D; Fri, 17 May 2002 14:51:28 +0100 (BST) To: Richard.Earnshaw@arm.com Cc: gdb-patches@sources.redhat.com Subject: Re: ARM SIM - add support to select SWI emulations to support. References: <200205171238.NAA22578@cam-mail2.cambridge.arm.com> From: Nick Clifton Date: Fri, 17 May 2002 06:51:00 -0000 In-Reply-To: <200205171238.NAA22578@cam-mail2.cambridge.arm.com> Message-ID: User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/21.1 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-SW-Source: 2002-05/txt/msg00711.txt.bz2 Hi Richard, > + extern unsigned int swi_mask; > + > + #define SWI_MASK_DEMON (1 << 0) > + #define SWI_MASK_ANGEL (1 << 1) > + #define SWI_MASK_REDBOOT (1 << 2) > > Why make this a mask? A major part of the problem here is that > these SWI sets may be mutually incompatible. Err really ? I was under the impression that all three SWI protocols used completely different number sets and so could be supported simultaneously. It certainly looks that way in the ARMul_OSHandleSWI function. > Why not just require the user to select the right one? The user may need to support more than one protocol and the user/script setting this option might not know which protocol is going to be used by a given binary. For example the current newlib implementation defaults to using ANGEL, but the libgloss implementation uses REDBOOT. Besides there are enough free bits that there is no reason not to be flexible and provide the ability to support more than one protocol at the same time. Cheers Nick