[MITgcm-support] testreport and genmake2

Ed Hill ed at eh3.com
Sat Sep 27 10:07:20 EDT 2003


On Sat, 2003-09-27 at 09:27, Martin Losch wrote:
> Hi,
> 
> I have figured out how to use gm_local, thanks for your patience while 
> I was whining.

No problem!  I'm glad its (somewhat) usable.


> But, .... I'd like to whine some more about ./testreport and genmake2:
> Is it really necessary that, everytime you rerun testreport, genmake2 
> overwrites the CPP_OPTIONS.h file, thereby forcing a complete rebuild? 
> Even on fast machines this really slows down progess to such a degree, 
> that I really don't feel like using this script at all. Something like 
> the -quick option of the old testscript would be nice ...

Yes, Alistair pointed out this problem earlier this week.  Theres a few
issues that this touches upon -- the most serious being a bug that we've
encountered where re-running genmake (1 or 2) results in malformed
(circular) links every *other* time.  I'll try to check in some fixes
soon.


> Equally, is it essential, that genmake2 by itself overwrites 
> CPP_OPTIONS.h every time you run it? 

Alistair also complained about that yesterday.  I'm working on a fix for
that right now.


> Also, is it intended that the 
> ALLOW_PACKAGE flags in CPP_OPTIONS.h are obsolete? They get redefined 
> at the end of the file by genmake2, so that I have to decide which pkg 
> I want to use at the genmake2 level, and having them in the 
> CPP_OPTIONS.h is even misleading ...

Again, you're right.

CNH, AA, JMC, and I agree that the way packages are currently built is
messy and far from optimal in terms of compilation effort.  The only
thing good about the way that its now handled is that its "safe".

So if the goal is:

 "adding a package to a previous build should not mean a complete 
  recompile of everything"

then how do we get there?  Chris explained a fix to me yesterday and,
again, I'll try to get it checked in soon.

Ed

-- 
Edward H. Hill III, PhD
office:  MIT Dept. of EAPS;  Room 54-1424;  77 Massachusetts Ave.
            Cambridge, MA 02139-4307
email:   eh3 at mit.edu,  ed at eh3.com
URL:     http://web.mit.edu/eh3/
phone:   617-253-0098
fax:     617-253-4464
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20030927/1b2856ef/attachment.sig>


More information about the MITgcm-support mailing list