From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: *
X-Spam-Status: No, score=1.3 required=5.0 tests=BAYES_00,INVALID_MSGID,
	MSGID_RANDY autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 1025b4,959627a08fbc77c5
X-Google-Attributes: gid1025b4,public
X-Google-Thread: 103376,30a9bb3017fa58dd
X-Google-Attributes: gid103376,public
From: Robert Dewar <robert_dewar@my-dejanews.com>
Subject: Re: GNAT versions ( was :Ada compiler for PC?)
Date: 1999/05/06
Message-ID: <7gssdo$gta$1@nnrp1.deja.com>#1/1
X-Deja-AN: 474855550
References: <7fndu7$im4$1@nnrp1.dejanews.com> <7gbjhg$s98$1@rtl.cygnus.com>
 <FB8DLq.BH6@kithrup.com> <7gpsrd$qc7$1@nnrp1.dejanews.com>
 <FBAEH2.9C0@kithrup.com> <7gs0d5$lvh$1@nnrp1.deja.com>
 <m3so9aci9g.fsf@fred.muc.de>
X-Http-Proxy: 1.0 x15.dejanews.com:80 (Squid/1.1.22) for client 205.232.38.31
Organization: Deja News - The Leader in Internet Discussion
X-Article-Creation-Date: Thu May 06 20:00:59 1999 GMT
Newsgroups: comp.lang.ada,gnu.misc.discuss
X-Http-User-Agent: Mozilla/4.51 [en] (WinNT; U)
Date: 1999-05-06T00:00:00+00:00
List-Id: <comp.lang.ada>

In article <m3so9aci9g.fsf@fred.muc.de>,
  Andi Kleen <ak@muc.de> wrote:

> As a first impression the GNAT 3.11 "zero-cost" EH implementation
> seems to usually generate smaller unwind tables than the egcs C++
> EH implementation
>
> (IMNSHO that is the main problem with the current egcs EH - too much
> executable bloat)
>
> It would be great if the best points of both could be merged.
>
> -Andi


I suspect this comparing apples and oranges, I am not sure what target
you are on, but I think you may be comparing the stack unwinding tables
of gcc with the static exception tables of GNAT, but maybe not.
Certainly the approach used for finalization in GNAT will use less
table space than the approach used in G++ but who knows if that is
significant.

As far as unwinding goes, on DEC Unix, we use the standard DEC Unix
library calls for unwinding, and on SGI, we use the standard SGI
routines. On the x86, we do prolog interpretation, which is very
messy but is highly space efficient. On Solaris, we plan to use the
standard gcc mechanism, but have not figured it out yet.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own