This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patch, rfc, rft] Multi-threaded single-step vs. breakpoint problems (prepare_to_proceed)
- 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, 10 Sep 2007 20:53:22 +0200 (CEST)
- Subject: Re: [patch, rfc, rft] Multi-threaded single-step vs. breakpoint problems (prepare_to_proceed)
Daniel Jacobowitz wrote:
> On Fri, Aug 03, 2007 at 06:54:34PM +0200, Ulrich Weigand wrote:
> > ChangeLog:
> >
> > * infrun.c (stepping_past_breakpoint): New global variable.
> > (stepping_past_breakpoint_ptid): Likewise.
> > (prepare_to_proceed): Add STEP parameter. Do not check for Ctrl-C.
> > Only switch threads if we need to single-step over a breakpoint hit
> > in the previously selected thread. If stepping, remember previous
> > thread to switch back to in STEPPING_PAST_BREAKPOINT[_PTID]. Call
> > switch_to_thread instead of copying its contents.
> > (proceed): Pass STEP to prepare_to_proceed. Always set ONEPROC if
> > prepare_to_proceed returns true.
> > (init_wait_for_inferior): Reset STEPPING_PAST_BREAKPOINT.
> > (context_switch): Call switch_to_thread.
> > (handle_inferior_event): Switch back to previous thread if requested
> > in STEPPING_PAST_BREAKPOINT[_PTID] by prepare_to_proceed.
> > * gdbthread.h (switch_to_thread): Add prototype.
> > * thread.c (switch_to_thread): Make global.
>
> No one else has commented on this patch in the last month; I think you
> should probably commit it (though you may wish to wait until after
> the release branch - I've no preference).
Thanks for reminding me of this. I've checked this in now. That means
it'll not be part of the release branch -- I think that's OK.
Bye,
Ulrich
--
Dr. Ulrich Weigand
GNU Toolchain for Linux on System z and Cell BE
Ulrich.Weigand@de.ibm.com