Re: Connect 2 mips simulator..

["Frank Ch. Eigler" <fche@redhat.com>]

Hi -

On Thu, Sep 06, 2001 at 02:58:51PM -0400, Zampieri Marco wrote:
: I'm looking for the option --memory-mapfile FILE, but I can't find it in
: the simulator parameters or gdb option.
: Where I can find it? It is a configure option or a parameter of Make
: program?

The --memory-mapfile option for the sim/common-based simulators such
as mips is supported automatically if your host supports mmap.  From
gdb, use "(gdb) [target] sim memory-mapfile FILE"; from the command
line, use "TARGET-run --memory-mapfile FILE".

Have you built your own cross-compiler/debugger/simulator recently?

- FChE
-- 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE7l34WVZbdDOm/ZT0RAkqLAJ488NHiRorasTKjKRneh7XoHMKFGwCeKBYT
r9XRR2/9QW801SabikVI/0k=
=m3+N
-----END PGP SIGNATURE-----
From dcoutts@cray.com Thu Sep 06 07:24:00 2001
From: Duncan Coutts <dcoutts@cray.com>
To: Andrew Cagney <ac131313@cygnus.com>
Cc: <gdb@sources.redhat.com>
Subject: Re: right way to_post_attach ?
Date: Thu, 06 Sep 2001 07:24:00 -0000
Message-id: <Pine.SGI.4.33.0109060911120.1220-100000@underbird7.mw.cray.com>
References: <3B970771.5060508@cygnus.com>
X-SW-Source: 2001-09/msg00044.html
Content-length: 2807

Hi,

On Thu, 6 Sep 2001, Andrew Cagney wrote:

> > Hi,
> >
> > I can't quite figure out what the right way to setup a
> > target_ops.to_post_attach hook. I'm not defining a whole new target (am
> > I?), I'm using the procfs target. I just need to hook the to_post_attach
> > so I can do some random symbol fiddling just after the attach.
> >
> > I notice that several targets do a push_target during their attach, so
> > that their operations become avaliable,  but presumably I still want to
> > use the procfs target so I can't have the attach operation push my extra
> > target vector.
> >
> > In oop terms, I just want to 'override' the to_post_attach
> > method. I can't find anywhere that uses the to_post_attach in any
> > substantial way, so I'm rather short on clues.
> >
> > What's the way to do this, or is to_post_attach not the tool to be using
> > here?
> >
> > Thank's for any tips.
>
>
> i cant comment on post attach but i can comment on the underlying
> problem.  yes, the push / pop mechanism and gdb's target stack is
> primative at best - little wonder you're having problems.  the oo
> concept is simple, what gdb did to it is not, sigh,

Well as a simple hack, I've done to procfs what the child target does,
which is to #define some flag CHILD[/PROCFS]_POST_ATTACH which indicates
that the to_post_attach hook should be set to an externally defined
function child[/procfs]_post_attach.

So I've modified procfs but done it in a similar way to how the child
target provides it's own hook. If there's a better way that doesn't
invlove changing generic code I'd be happy to use it.

To be more specific here's what I did:
Comments welcome.

Duncan

Index: procfs.c
===================================================================
diff -c -r1.7 procfs.c
*** procfs.c	2001/08/27 17:32:52	1.7
--- procfs.c	2001/09/04 17:03:09
***************
*** 93,98 ****
--- 93,101 ----

  static void procfs_open              PARAMS((char *, int));
  static void procfs_attach            PARAMS ((char *, int));
+ #if defined(PROCFS_POST_ATTACH)
+ extern void procfs_post_attach       PARAMS ((int));
+ #endif
  static void procfs_detach            PARAMS ((char *, int));
  static void procfs_resume            PARAMS ((int, int, enum target_signal));
  static int  procfs_can_run           PARAMS ((void));
***************
*** 131,136 ****
--- 134,142 ----
    procfs_ops.to_kill               = procfs_kill_inferior;
    procfs_ops.to_mourn_inferior     = procfs_mourn_inferior;
    procfs_ops.to_attach             = procfs_attach;
+ #if defined(PROCFS_POST_ATTACH)
+   procfs_ops.to_post_attach        = procfs_post_attach;
+ #endif
    procfs_ops.to_detach             = procfs_detach;
    procfs_ops.to_wait               = procfs_wait;
    procfs_ops.to_resume             = procfs_resume;