bug#17162: 24.3; GDB interface stops working in some circumstances

[Zach Kost-Smith <zachkostsmith@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1647 bytes --]

Hi Barry,

Perhaps there is some kind of misunderstanding?  M-x gud-gdb seems to
produce a prompt for gdb (have to remember not to use -i=mi for that one),
and it provides some simple support for setting breakpoints and printing
variables via the C-x C-a C-b / M-x gud-break sort of commands, but this
seems to be barely better than simpling running in a terminal.  There seems
to be no graphical representation of the point of execution, the location
of breakpoints either within the buffer or in a special breakpoint buffer,
nor a backtrace buffer, or the list of local variables (which is the only
one that I never really found very useful).

But, yes, using gud-gdb does not show the same behavior where a C-c C-c
will make the M-x gud-<something> commands stop working.  In some sense it
does indeed work better, but I have fond memories of a time where I think
that the GDB interface in Emacs was working.  Is there at least of
confirmation that others are seeing the same issue with things like M-x
gud-print stop working with M-x gdb?  I was troubleshooting with someone on
Reddit and they never saw this issue.

Zach



On Tue, Apr 1, 2014 at 1:39 PM, Barry OReilly <gundaetiapo@gmail.com> wrote:

> > we just stopped using GDB in Emacs long ago but I haven't given up
> > hope
>
> Does
>
>   M-x gud-gdb
>
> work better for you? The interface to GDB accessed via gdb command is
> dreadful.
>
> > As an aside, when troubleshooting this there seems to be more severe
> > but harder to reproduce issues where C-c C-c doesn't actually
> > interrupt the execution.
>
> Possibly related: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=16745
>
>

[-- Attachment #2: Type: text/html, Size: 2203 bytes --]