From mboxrd@z Thu Jan  1 00:00:00 1970
From: Joel Brobecker <brobecker@gnat.com>
To: gdb-patches@sources.redhat.com
Subject: Re: [RFA/RFC] New command: ``start''
Date: Wed, 19 May 2004 16:10:00 -0000
Message-id: <20040519161006.GF10684@gnat.com>
References: <20040518024700.GV10684@gnat.com> <20040518214712.GD4036@nevyn.them.org> <20040518222740.GZ10684@gnat.com> <20040518224131.GA6026@nevyn.them.org> <20040519153615.GD10684@gnat.com> <20040519154155.GA8567@nevyn.them.org>
X-SW-Source: 2004-05/msg00575.html

> That sounds good to me.  Note that now you can avoid playing with
> xmalloc/xfree and just return the constant "main".

Before starting on this, there are a couple of things we might want
to discuss:

   . It think it would be better if we cached the main name, so as
     to avoid recomputing it each time we do a backtrace. But that
     causes a problem when debugging 2 different executables one
     after the other during the same GDB session. During the second
     debugging run, we'll end up using the wrong main procedure name.
     We need to insert an observer I think at the begining of
     exec_file_attach() to signal us that the cached main_name
     is obsolete.

   . Could you clarify a bit more what you mean about the xmalloc/xfree
     juggle? I see that set_main_name() takes care of this for me, so
     do you mean that the language method should be returning a string
     located in a temporary buffer? Something like: result is good until
     next call? (I am not too keen on that).

Thanks,
-- 
Joel