[m-rev.] for review: add --resume dd option

Ian MacLarty maclarty at cs.mu.OZ.AU
Sat Feb 26 20:26:27 AEDT 2005


On Thu, Feb 24, 2005 at 05:47:31PM +1100, Julien Fischer wrote:
> >
> Don't forget to update the log mesage as well.
> 

Done.

> >  search mode to use.  Either @samp{top-down} or @samp{divide-and-query}
> >  may be specified.  See @ref{Search Modes} for a more detailed description of
> >  the available search modes.  @samp{top-down} is the default when this option is
> >  not given.
> > + at sp 1
> > +Give the @samp{-r} or @samp{--resume} option to continue your previous
> 
> I suggest: Use the @samp{-r} or @samp{--resume} option ...B
> 

Okay.

> > +declarative debugging session.
> 
> > This option should only be given if a
> > + at samp{dd} command was previously issued without the @samp{--resume} option.
> 
> This sentence seems redundant.
> 

Agreed.

> > If
> > +the @samp{--resume} option is given and there were no previous declarative
> > +debugging sessions then the option will be ignored.
> > +A @samp{dd --resume} command can be issued at any event.
> 
> An alternative (that Ian and I discussed in person), is to have dd resume
> a previous declarative debugging session if one exists and have a --new
> option for dd to start a new session if there is pre-existing one.
> 

I'll make this change later since it'll involve updating a lot of test cases
(all tests that call dd more than once will need to pass the -n option).

After bootchecking in the decldebug grade I've also made the following changes:

To the CVS log:

trace/mercury_trace_declarative.c
	Don't update the saved event details from the global event
	number/seqno/depth variables at the start of MR_decl_diagnosis.  The
	globals could have been updated by a previous call to Mercury code and
	could have incorrect values.
	
And to mercury_trace_declarative.c:

diff -u trace/mercury_trace_declarative.c trace/mercury_trace_declarative.c
--- trace/mercury_trace_declarative.c	24 Feb 2005 01:59:14 -0000
+++ trace/mercury_trace_declarative.c	26 Feb 2005 09:02:59 -0000
@@ -1642,10 +1643,6 @@
 	MR_Unsigned		io_start;
 	MR_Unsigned		io_end;
 
-	event_details->MR_call_seqno = MR_trace_call_seqno;
-	event_details->MR_call_depth = MR_trace_call_depth;
-	event_details->MR_event_number = MR_trace_event_number;
-
 	if (MR_edt_compiler_flag_warning) {
 		fflush(MR_mdb_out);
 		fprintf(MR_mdb_err, "Warning: some modules were compiled with"
@@ -1701,8 +1698,7 @@
 	MR_TRACE_CALL_MERCURY(
 		if (root == (MR_Trace_Node) NULL) {
 			MR_DD_decl_diagnosis_resume_previous(
-				MR_trace_node_store, 
-				use_old_io_map,
+				MR_trace_node_store, MR_FALSE,
 				MR_io_action_map_cache_start,
 				MR_io_action_map_cache_end,
 				&response, MR_trace_front_end_state,

The second change here is to tell the front not to replace its IO action
map if a previous session is being resumed.

Ian.
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------



More information about the reviews mailing list