Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

From: Andreas Knüpfer (andreas.knuepfer_at_[hidden])
Date: 2007-10-08 16:00:50


Hi Jeff, hi everybody,

quite a few things to comment on. hope i catch them all:

- yes, we might move vanmpirtrace to ./ompi/contrib/vampirtrace/, why not. but
since we agreed on the current location ./tracing/vampirtrace/ we should not
rush it just because another software is coming to openmpi, should we?

- the "call home" feature. I understand you concerns perfectly well. we'll
remove this, If anyone is asking us to. but please let me explain first:

1) it's not in vampirtrace itself but in the parts we added to ompi on behalf
of vampirtrace
2) it is never active by default, instead you need a
special '--update-vampirtrace' switch on your configure command. otherwise
nothing is calling nowhere. is this equivalent to "totally deactivated" or
still not good enough?
3) it was our idea to make it easier to replace the included vampirtrace
version if it was necessary. manually download and untar is not that much
harder, though.
4) we check if 'wget' is present and try _not_ to make ./configure fail

once again, if anybody want's this removed, please say so.

- about the two vampirtrace-specific spots in the .m4 files: they correspont
to two tasks: firstly, decide if you want vampirtrace at all or (if you might
want to update) and secondly, passing configure options to vampirtrace.
we need to do the first before the second, of course. maybe we can move
everything to "our" .m4 file, let me check ...

- btw: so far the vampirtrace distribution tarball is brought to openmpi
under ./tracing/vampirtrace with no modifications

- the mpicc-vt (and friends) compiler wrappers: this is not part of
vampirtrace but a new thing that only makes sense together with openmpi.
therefore, they stay next to 'mpicc' and all others. in fact we're following
a earlier suggestion from you, Jeff: 'mpicc-vt' is just like 'mpicc' but
calls the 'vtcc' compier wrapper instead of 'cc'.

this makes everything much simpler, because we can handle all special cases in
vtcc. the wrapper config for 'mpicc-vt' is almost a mere copy of mpicc's one.
therefore, I'd like to keep them where they are right now. is this o.k. with
everyone?

So, I agree to some of your proposed changes. what about the rest? I'll be
back in the office next week. Please, get back to Matthias and Matthias in
the meantime.

knue

-- 
Dipl. Math. Andreas Knuepfer, 
ZIH, TU Dresden, 
Willersbau A114, Zellescher Weg 12, 01062 Dresden
phone +49-351-463-38323, fax +49-351-463-37773