This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [commit] Support displaced stepping on S/390
- From: "Ulrich Weigand" <uweigand at de dot ibm dot com>
- To: drow at false dot org (Daniel Jacobowitz)
- Cc: gdb-patches at sourceware dot org
- Date: Mon, 28 Sep 2009 11:24:22 +0200 (CEST)
- Subject: Re: [commit] Support displaced stepping on S/390
Daniel Jacobowitz wrote:
> FYI, I have a local patch to issue an error if you try to connect to a
> remote target in non-stop mode and the target supports non-stop but
> gdb does not know how to do displaced stepping. Clearly the check
> shouldn't be remote specific - but I didn't post it because I couldn't
> figure out how where to put the error to work for native too.
>
> Attached for reference. I am perfectly happy to commit this if
> someone thinks it's an improvement, but I'd rather figure out where to
> handle native first...
FWIW I agree that we should handle all targets ...
Does the decision need to be made at "attach" time? Maybe we should
allow attaching in non-stop mode, but refuse to install breakpoints
if displaced stepping is not available? That would at least allow
attaching to a process in non-stop mode and stopping/restarting
selected threads and inspecting their state.
Bye,
Ulrich
--
Dr. Ulrich Weigand
GNU Toolchain for Linux on System z and Cell BE
Ulrich.Weigand@de.ibm.com