Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

Subject: Re: [OMPI devel] [OMPI svn] svn:open-mpi r23767
From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2010-09-20 11:54:35


On Sep 20, 2010, at 11:48 AM, Barrett, Brian W wrote:

> Sorry, was out of town last week. The patch undoes the discussion we had -- we always run component macros, even if the component couldn't build, to solve the issue of AC_CONFIG_FILES needing to always be run.

Ah RIGHT -- *that* was the reason: AC_CONFIG_FILES. This is why I asked before I committed. :-)

> So the first part of the patch is incorrect and should not be committed. We might need to make the valgrind configure.m4 macro smarter, but the other option is really painful.

Ok. I'll do something here; I think the right solution is to:

- have valgrind AC_MSG_WARN instead of AC_MSG_ERROR (which I committed already)
- have the memchecker/configure.m4 error if you --enable-memchecker but then no components are available to be compiled

The latter is important because this is a static-only framework (i.e., a new component can't be introduced at run-tmie via DSO).

> I think the second issue is a typo and can be changed. Originally, I think we needed to know whether the component configure failed or the component was told not to build, but that no longer seems necessary.

Cool. I'll commit just that part tonight.

Thanks!

-- 
Jeff Squyres
jsquyres_at_[hidden]
For corporate legal information go to:
http://www.cisco.com/web/about/doing_business/legal/cri/