Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 10 Aug 2001 22:50:03 -0600
From:      "Kenneth D. Merry" <ken@kdm.org>
To:        "Sergey A. Osokin" <osa@freebsd.org.ru>
Cc:        audit@FreeBSD.ORG
Subject:   Re: [PATCH] libdevstat have two compute statistics functions
Message-ID:  <20010810225003.B75606@panzer.kdm.org>
In-Reply-To: <20010810161632.A7827@freebsd.org.ru>; from osa@freebsd.org.ru on Fri, Aug 10, 2001 at 04:16:32PM %2B0400
References:  <20010808184710.A87195@freebsd.org.ru> <20010809102140.A62588@panzer.kdm.org> <20010810161632.A7827@freebsd.org.ru>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Aug 10, 2001 at 16:16:32 +0400, Sergey A. Osokin wrote:
> On Thu, Aug 09, 2001 at 10:21:40AM -0600, Kenneth D. Merry wrote:
> > On Wed, Aug 08, 2001 at 18:47:11 +0400, Sergey A. Osokin wrote:
> > > Hello.
> > > After last commit into libdevstat, devstat.c containes
> > > 2 compute statistics functions: old compute_stats() and
> > > new devstat_compute_statistics().
> > > 
> > > I believe that compute_stats() is not necessary any more
> > > and we should remove it. devstat_compute_statistics() is
> > > a superset for compute_stats().
> > 
> > This patch won't work.  The reason is that the arguments to compute_stats()
> > can be NULL, which will cause it not to calculate that particular
> > statistic.
> 
> OK. I mistake about goto's in switch(devstat_arg_list[metric].argtype).
> Look at attached patch. I make a little hack for
> devstat_compute_statistics(). After this patch function try to
> resolve all pairs of parameters.

You cut out the part of my message where I discussed, indirectly, this
approach:

> > Your patch passes in all of the arguments to devstat_compute_statistics().
> > If one of the initial arguments is NULL, it will bail out and not compute
> > the rest of the statistics.  IMO this is the correct behavior -- the user
> > has specified a statistic to calculate, and if he has not also given a
> > valid variable to put the result in, that is an error.

The problem with this patch is that you can now supply bogus arguments.  I
don't think that's the right approach.

My preference is to leave things as-is.

Ken
-- 
Kenneth Merry
ken@kdm.org

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-audit" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010810225003.B75606>