Re: Optimization of add_dt_to_dt_list() in resolve.c

[Janus Weil <janus@gcc.gnu.org>]

Hi Andrew,

2018-07-20 18:29 GMT+02:00 Andrew Benson <abenson@carnegiescience.edu>:
> Pinging to see if anyone can take a look at this.

I think your patch has received a reasonable amount of reviewing
already. You seem to have responded to all the points that came up,
and I personally I don't see any further ones.

I had even tested your patch a few weeks ago. It did not bring much
speedup for the codes I've tried it on (probably because the number of
derived types is significantly lower than in your case), but it also
did not show any negative side effects either.

So, I'd say the patch is ok for trunk!

Do you want me to commit it for you, or do you already have a
sourceware account for svn access (see
https://gcc.gnu.org/svnwrite.html#authenticated) and prefer to do it
yourself?

Cheers,
Janus



> On Tuesday, July 10, 2018 7:32:58 PM PDT Andrew Benson wrote:
>> I now finally have heard back from the FSF, so my paperwork is all signed
>> and recorded.
>>
>> I've tested my patch on the latest trunk - slightly updated version which
>> applies cleanly to trunk is attached (along with the ChangeLog).
>>
>> -Andrew
>>
>> On Thursday, June 14, 2018 5:10:24 PM PDT Steve Kargl wrote:
>> > Ping me when you hear back from FSF.  I'll apply your
>> > current patch on the weekend to my tree and do some
>> > testing.
>> >
>> > > Steve,
>> > >
>> > > I have just sent my copyright assignment documents back to the FSF, so I
>> > > think it should be ok to submit my patch as soon as it has approval.
>> > >
>> > > -Andrew
>> > >
>> > > On Monday, June 11, 2018 11:17:10 AM PDT Steve Kargl wrote:
>> > > > On Fri, Jun 01, 2018 at 08:14:04AM +0200, Janus Weil wrote:
>> > > > > 2018-05-31 20:04 GMT+02:00 Andrew Benson
>>
>> <abenson@carnegiescience.edu>:
>> > > > > > One other question: for copyright assignment, who do I need to
>> > > > > > talk to to get the relevant form(s)?
>> > > > >
>> > > > > I think you need to send a request to assign@gnu.org and
>> > > > > gcc@gcc.gnu.org in order to get the copyright assignment form. After
>> > > > > the form is mailed to you, you sign it and send it back to the FSF.
>> > > >
>> > > > Andrew, I see that you've asked on gcc@gcc on June 1 about
>> > > > Copyright forms.  Has anyone responded?
>> > > >
>> > > > I ask because my patch for PR fortran/68544 walks the
>> > > > gfc_derived_types list.
>> > > >
>> > > > https://gcc.gnu.org/ml/fortran/2018-06/msg00054.html
>> > > >
>> > > > and Thomas has approved the patch.  In my patch, I
>> > > > have
>> > > >
>> > > > +static bool
>> > > > +is_dt_name (const char *name)
>> > > > +{
>> > > > +  gfc_dt_list *dt_list;
>> > > > +
>> > > > +  for (dt_list = gfc_derived_types; dt_list; dt_list = dt_list->next)
>> > > > +    if (strcmp(dt_list->derived->name, name) == 0)
>> > > > +      return true;
>> > > > +  return false;
>> > > > +}
>> > > >
>> > > > we'll need to update this to deal with your change for a
>> > > > circular linked list.
>
>
> --
>
> * Andrew Benson: http://users.obs.carnegiescience.edu/abenson/contact.html
>
> * Galacticus: https://bitbucket.org/abensonca/galacticus
>