From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 26364 invoked by alias); 5 Jan 2003 11:51:29 -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 26347 invoked from network); 5 Jan 2003 11:51:27 -0000 Received: from unknown (HELO executor.cambridge.redhat.com) (195.224.55.237) by 209.249.29.67 with SMTP; 5 Jan 2003 11:51:27 -0000 Received: from tantra.cambridge.redhat.com (tantra.cambridge.redhat.com [172.16.18.34]) by executor.cambridge.redhat.com (Postfix) with ESMTP id 3D537ABB20; Sun, 5 Jan 2003 11:51:15 +0000 (GMT) Received: (from rsandifo@localhost) by tantra.cambridge.redhat.com (8.11.6/8.11.0) id h05BpCp14442; Sun, 5 Jan 2003 11:51:12 GMT X-Authentication-Warning: tantra.cambridge.redhat.com: rsandifo set sender to rsandifo@redhat.com using -f To: cgd@broadcom.com Cc: gdb-patches@sources.redhat.com Subject: Re: sim/mips patch: add support for more NEC VR targets References: From: Richard Sandiford Date: Sun, 05 Jan 2003 11:51:00 -0000 In-Reply-To: Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.1 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-SW-Source: 2003-01/txt/msg00172.txt.bz2 Thanks for applying the patch ;) cgd@broadcom.com writes: > I believe there are two relatively minor issues with this patch > remaining: > > * Andrew wants to avoid explicit architecture checks in the mips.igen > file. That seems sensible enough, I guess. Without the code in > mips.igen:check_mf_cycles that he objects to, the vr5500 won't work > correctly but the rest of the new architectures being added should. > > I've mentioned previously to you that i'd look into this. I've not > yet had much time yet, but hope to in the next week or two. OK. One way (I think I mentioned it privately, maybe it's what you had in mind anyway) would be to add some sort of exclusion filter to igen. As well as "*FOO" to include something for architecture FOO there could be "!FOO" (or whatever) to exclude it. Then we could have something like: :function:::int:check_mt_hilo:hilo_history *history ... *mipsIV: ... !vr5500: ... version that calls check_mf_cycles ... :function:::int:check_mt_hilo:hilo_history *history *mips32: *mips64: *r3900: *vr5500: ... version that doesn't ... Would it be OK to change igen along these lines? I'll volunteer to try it if so... > * I believe that there are still a couple of files not cleaned > properly by 'make distclean'. > > Could you look into this? Heh, that was very sloppy of me, sorry. The files in question (m16m32*_run.c) should never have been created in the first place. The patch below gets rid of them and causes no change in the executable (except for a timestamp). Please install if OK. Richard * Makefile.in (tmp-run-multi): Fix mips16 filter. Index: sim/mips/Makefile.in =================================================================== RCS file: /cvs/src/src/sim/mips/Makefile.in,v retrieving revision 1.8 diff -u -d -p -F^\([(a-zA-Z0-9_]\|#define\) -r1.8 Makefile.in --- sim/mips/Makefile.in 5 Jan 2003 07:56:59 -0000 1.8 +++ sim/mips/Makefile.in 5 Jan 2003 11:28:14 -0000 @@ -371,7 +371,7 @@ tmp-itable-multi: $(IGEN_INSN) $(IGEN_DC tmp-run-multi: $(srcdir)/m16run.c for t in $(SIM_MULTI_IGEN_CONFIGS); do \ case $${t} in \ - *:*mips16*:*) \ + m16*) \ m=`echo $${t} | sed -e 's/^m16//' -e 's/:.*//'`; \ sed < $(srcdir)/m16run.c > tmp-run \ -e "s/^sim_/m16$${m}_/" \