Re: deprecated use of octave_allocator

[John W. Eaton]

On 01/20/2015 08:38 AM, John W. Eaton wrote:
> On 01/20/2015 07:19 AM, Andreas Weber wrote:
> > Am 20.01.2015 um 11:24 schrieb John W. Eaton:
> > > On 01/17/2015 12:10 PM, Andreas Weber wrote:
> >
> > > > Then I run "time make check"
> > > >
> > > > on the built with default configure flags:
> > > > real    6m20.855s
> > > > user    5m29.124s
> > > > sys    0m48.276s
> > > >
> > > > and the built with "--enable-octave-allocator"
> > > > real    6m16.529s
> > > > user    5m29.368s
> > > > sys    0m47.540s
> > > > ....
> >
> > > I don't think you are going to get useful data from just one simple
> > > operation.  A better test might be to compare the time it takes to run
> > > the full test suite.  That will create many objects, not just a few.
> >
> > The above tests are from "time make check". Do we have another testsuite
> > I'm not aware of?
>
> Sorry, I was focused on the --eval test you did, which doesn't really
> show anything.
>
> I think the results you see show that the custom allocator doesn't
> really make much difference.  So it seems to me that we might as well
> remove it completely.

I checked in the following change to remove the allocator:

  http://hg.savannah.gnu.org/hgweb/octave/rev/76478d2da117

The uses of these macros in Octave Forge code can be removed at any 
time.  In a few releases, we can remove the oct-alloc.h header file.

jwe