On Mon, Oct 10, 2005 at 06:25:47PM -0500, Brian Barrett wrote:
> On Oct 10, 2005, at 6:12 PM, Brooks Davis wrote:
> 
> > I'm looking at porting Open MPI to FreeBSD and am trying to get a  
> > handle
> > on the requirements.  The issue I'm hitting so far is the opal timer
> > code which won't finish configuration at the moment.  Where is the API
> > to implement documented?  Is there an intent to implement a sub- 
> > optimal
> > version with POSIX calls at some point?  If I understand the current
> > code correctly, something based on clock_gettime() and clock_getres()
> > should do it.
> >
> > Are there other things I can expect to have to implement?
> 
> That should actually be about it.  In theory, Open MPI should have  
> fallen back to using gettimeofday().  Can you send the output of  
> configure and the config.log file?  As for implementing a FreeBSD- 
> specific timer component, if clock_gettime() and clock_getres()  
> provide reasonably low perturbationtime reference, they would be be  
> ideal.

The configure output ends with:

...
config.status: creating test/util/Makefile
config.status: creating include/ompi_config.h
config.status: creating include/mpi.h
config.status: include/mpi.h is unchanged
config.status: linking ./opal/mca/timer/base/timer_base_null.h to 
opal/mca/timer/base/base_impl.h

I've attached gziped copies of the configure output and config.log.

-- Brooks

-- 
Any statement of the form "X is the one, true Y" is FALSE.
PGP fingerprint 655D 519C 26A7 82E7 2529  9BF0 5D8E 8BE9 F238 1AD4

Attachment: config.log.gz
Description: GNU Zip compressed data

Attachment: configure.out.gz
Description: GNU Zip compressed data

Attachment: pgprKzVHuhPIG.pgp
Description: PGP signature

Reply via email to