This is the mail archive of the
gdb@sources.redhat.com
mailing list for the GDB project.
Re: rdi debugger dies and ctrl-c stops working
- To: Grant Edwards <grante at visi dot com>
- Subject: Re: rdi debugger dies and ctrl-c stops working
- From: Andrew Cagney <ac131313 at cygnus dot com>
- Date: Tue, 11 Sep 2001 11:04:50 -0400
- Cc: Shaun Jackman <sjackman at pathwayconnect dot com>,gdb at sources dot redhat dot com
- References: <E15gUc5-0006J8-00@dazed> <20010910122522.A18172@visi.com>
> On Mon, Sep 10, 2001 at 11:12:45AM -0600, Shaun Jackman wrote:
>
>
>> I have a Jeeni debugger which occasionally hangs. When this
>> happens while the target is running ctrl-c can't be used to
>> grab a hold of gdb. I know gdb hasn't crashed though because
>> ctrl-c causes a packet to be sent out on ethernet to the Jeeni.
>> Is there another signal I can send to gdb to cause it to abort
>> the rdi connection and put me back at the prompt?
>
>
> In my experience that happens when the Jeeni and the ARM debug
> macro cell have gotten out of sync (e.g. the CPU has been
> reset). Even if you could get gdb back to the command prompt,
> I think you'd still have to reset the Jeeni and start over.
>
> It would be nice to be able to get gdb's attention without
> having to kill it from another window or put it into background
> and do a "kill %1".
remote.c counts cntrl-c. more than one and it asks if you want to
abandon the connection. perhaps you could modify your gdb backend so
that it has the same behaviour?
andrew