Per Jeff's suggestion, I tried recompiling. While I could not successfully recompile mpich, I could recompile lam and I was able to run the lam test successfully. Maybe we should offer two versions of some of these libraries... one using 2.9x and one using 3.x? OTOH, newer distros have compat- type packages which provide the older binaries and libraries (and headers, etc.). Any suggestions on what we should do?

Jason

At 06:37 PM 1/27/2003 -0500, Jeff Squyres wrote:
On Mon, 27 Jan 2003, Jason B. wrote:

> If anyone can assist in debugging this problem, I'd really like to add
> RH 8.0 support to OSCAR in the near future.  Even if people don't use
> x.0 releases, it will give us a "foot in the door" for future 8.x
> releases.
>
> I'm assuming this has something to do with gcc 3.2 being the only
> available gcc compiler... I'm looking into subbing the old version of
> them.

Yes, this is quite definitely C++ library linker problem.

Really, the best solution here is to have multiple RPMs for LAM and MPICH,
and the LAM/MPICH packages pick which RPMs to install.  Not only will it
fix this C++ issue, the MPI implementations will get better performance.

--
{+} Jeff Squyres
{+} [EMAIL PROTECTED]
{+} http://www.lam-mpi.org/


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com
_______________________________________________
Oscar-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/oscar-devel


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com
_______________________________________________
Oscar-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/oscar-devel

Reply via email to