This is the mail archive of the
gdb@sources.redhat.com
mailing list for the GDB project.
Re: extended remote protocol
- To: jtc at redback dot com
- Subject: Re: extended remote protocol
- From: Andrew Cagney <ac131313 at cygnus dot com>
- Date: Wed, 01 Nov 2000 23:44:09 +1100
- Cc: gdb at sourceware dot cygnus dot com
- References: <5mu2bmjz1h.fsf@jtc.redback.com>
"J.T. Conklin" wrote:
>
> The remote protocol description in gdb.texinfo states:
> Stubs that support the extended remote protocol return @samp{}
> which, unfortunately, is identical to the response returned by
> stubs that do not support protocol extensions.
>
> which I've believed without question until earlier today. I was
> poking around gdbserver/server.c trying to tighten up command parsing,
> and I found:
>
> case '!':
> extended_protocol = 1;
> prepare_resume_reply (own_buf, status, signal);
> break;
> case '?':
> prepare_resume_reply (own_buf, status, signal);
> break;
1.7 (law 26-Jul-95): case '!':
1.7 (law 26-Jul-95): extended_protocol = 1;
1.7 (law 26-Jul-95): prepare_resume_reply
(own_buf, status, signal);
1.7 (law 26-Jul-95): break;
>
> gdbserver is the only stub I know of that supports the extended remote
> protocol. Has it always output the resume reply like the '?' command?
> Are there other stubs that support the extended remote protocol?
I'll do a little bit more digging but it looks like I was wrong.
(Checking two other stub implementations neither recognize the ``!''
packet.)
Andrew