This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch 5/5] Document


On Thursday 17 November 2011 14:30:37, Yao Qi wrote:
> On 11/17/2011 08:28 PM, Pedro Alves wrote:
> >> > +disconnects from the remote stub, pending tracepoints still exist but
> >> > +can not be resolved while @value{GDBN} is disconnected.
> > Sorry to be picky, but I'm trying to read this from a user's perspective,
> > and it still confuses me.  What does "pending tracepoints still exist"
> > mean?  Do you mean they still exist in GDB?  That's true for all kinds
> > of breakpoints, so it doesn't add anything.  If you mean that they exist
> > on the target, then what does it mean for a pending tracepoint to exist
> > on the target?  What we're really trying to say is that pending tracepoints
> > don't work with disconnected tracing.  How about:
> > 
> 
> I agree that "pending tracepoints still exist" is confusing, and we
> should remove this sentence.  However, I don't think we should express
> "pending tracepoints don't work with disconnected tracing.", because,
> "pending tracepoints" and "disconnected tracing" are orthogonal to each
> other.  A remote stub can support either/both/none of them.

But if the target doesn't support disconnected tracing, tracing always
stops on disconnection, and so mentioning that pending tracepoints don't
resolve isn't interesting for that case.  The ability to resolve or not
pending tracepoints while gdb is disconnected is only interesting from the
target's perpective _while_ a trace run is ongoing.  From gdb's
perspective, once the remote target goes away, there's nothing special about
pending tracepoints compared to other types of breakpoints.

> > The resolution of pending tracepoints requires @value{GDBN} support---
> > when debugging with the remote target, and @value{GDBN} disconnects from the
> > remote stub (@pxref{disconnected tracing}), pending tracepoints can not be
> 
> ...so I suggest remove "(@pxref{disconnected tracing})" here.  What do
> you think?

I do think we should have that reference there.

-- 
Pedro Alves


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]