[RFA] Quiet a tracepoint notification with gdbserver

[RFA] Quiet a tracepoint notification with gdbserver

[Daniel Jacobowitz]

Since gdbserver now defines trace_timestamp, we get a notice after
connection.  This is confusing if you're not trying to use
tracepoints.  I suggest the attached patch; OK?

-- 
Daniel Jacobowitz
CodeSourcery

2010-10-08  Daniel Jacobowitz  <dan@codesourcery.com>

	* tracepoint.c (merge_uploaded_trace_state_variables): Only print
	messages if info_verbose.

Index: tracepoint.c
===================================================================
--- tracepoint.c	(revision 301900)
+++ tracepoint.c	(working copy)
@@ -3139,13 +3139,17 @@ merge_uploaded_trace_state_variables (st
     {
       tsv = find_matching_tsv (utsv);
       if (tsv)
-	printf_filtered (_("Assuming trace state variable $%s is same as target's variable %d.\n"),
-			 tsv->name, utsv->number);
+	{
+	  if (info_verbose)
+	    printf_filtered (_("Assuming trace state variable $%s is same as target's variable %d.\n"),
+			     tsv->name, utsv->number);
+	}
       else
 	{
 	  tsv = create_tsv_from_upload (utsv);
-	  printf_filtered (_("Created trace state variable $%s for target's variable %d.\n"),
-			   tsv->name, utsv->number);
+	  if (info_verbose)
+	    printf_filtered (_("Created trace state variable $%s for target's variable %d.\n"),
+			     tsv->name, utsv->number);
 	}
       /* Give precedence to numberings that come from the target.  */
       if (tsv)

Re: [RFA] Quiet a tracepoint notification with gdbserver

[Michael Snyder]

Daniel Jacobowitz wrote:
> Since gdbserver now defines trace_timestamp, we get a notice after
> connection.  This is confusing if you're not trying to use
> tracepoints.  I suggest the attached patch; OK?
> 

Thank you!  I've found those messages bothersome, not because they
confuse me but because they will confuse my users.

Re: [RFA] Quiet a tracepoint notification with gdbserver

[Stan Shebs]

Daniel Jacobowitz wrote:
> Since gdbserver now defines trace_timestamp, we get a notice after
> connection.  This is confusing if you're not trying to use
> tracepoints.  I suggest the attached patch; OK?
>
>   

It's OK with me.  I think what we really want is some kind of "I'm a 
tracepoint user" hint or heuristic that activates this class of warning, 
but we don't yet have enough user feedback to know how much effort we 
need to make.

Stan

Re: [RFA] Quiet a tracepoint notification with gdbserver

[Daniel Jacobowitz]

On Fri, Oct 08, 2010 at 11:52:21AM -0700, Stan Shebs wrote:
> Daniel Jacobowitz wrote:
> >Since gdbserver now defines trace_timestamp, we get a notice after
> >connection.  This is confusing if you're not trying to use
> >tracepoints.  I suggest the attached patch; OK?
> >
> 
> It's OK with me.

Thanks, checked in.

-- 
Daniel Jacobowitz
CodeSourcery