[m-dev.] for review: interface tracing, part 1

[Fergus Henderson]

On 15-May-1998, Zoltan Somogyi <zs at cs.mu.OZ.AU> wrote:
> 
> > By the way, have you got some benchmarks for us -- how much bigger
> > and slower is it with interface tracing and full tracing?
> 
> I haven't done timings yet. As far as sizes are concerned:
> 
> text	data	bss
> 9338880 6766592 98944   full tracing
> 7692288 3973120 103792	interface tracing
> 4276224 827392  104400  no tracing

Or to summarize,

		full		interface	none
text+data	15.4M		11.1M		4.9M.

That's certainly within tolerable bounds.  

Of course it would be nice if we could get it smaller ;-)
Particularly for interface tracing.

> 	--trace all
> 	--trace interfaces
> 	--trace minimum

I'd be happy with that.

> > The documentation should explain the effect of each of these
> > different tracing levels.
> > 
> > (In addition, the user guide should probably have a new section
> > on debugging.  But that can be committed as a seperate change.)
> 
> The documentation of the option cannot contain a *full* discussion
> of the trace levels, since that would require duplicating a large part
> of that future debugging section.

OK, but you should at least include

	@c XXX should include a pointer here to the as-yet-unwritten
	@c     section on debugging

in the documentation source.

> What follows is a relative diff for both parts 1 & 2.

That diff looks great, thanks.

-- 
Fergus Henderson <fjh at cs.mu.oz.au>  |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>  |  of excellence is a lethal habit"
PGP: finger fjh at 128.250.37.3        |     -- the last words of T. S. Garp.