Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |  

This web mail archive is frozen.

This page is part of a frozen web archive of this mailing list.

You can still navigate around this archive, but know that no new mails have been added to it since July of 2016.

Click here to be taken to the new web archives of this list; it includes all the mails that are in this frozen archive plus all new mails that have been sent to the list since it was migrated to the new archives.

Subject: Re: [OMPI devel] MPIR attach from padb broken (1.5.5rc1)
From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2011-12-16 11:24:15

On Dec 16, 2011, at 10:56 AM, Nathan T. Hjelm wrote:

>> Why do the symbols need to be weak? Remember that not all platforms
>> support weak symbols.
> I don't know. The way we currently pull MPIR symbols into orterun breaks
> with some configuration. If we don't pull the symbols in then launchmon
> can't attach to orterun. Moving the symbols to weak fixes the problem
> Ashley is seeing without breaking launchmon.

We could pull them in the same way MPIR_Breakpoint is pulled in (i.e., via function call, not a symbol dereference -- which is not always enough for some linkers).

> What platforms configurations do not support weak?


>> The symbols don't need to be in the executable itself, right? It should
>> be fine for them to be a library (e.g.,
> In theory, yes, but in practice, no. To support launchmon orterun needs to
> "define" MPIR symbols. At the moment they are in the wrong section (B) and
> moving them to section V (when it is supported) should be better.

What is launchmon doing to find the symbols? I.e., why can't it find them in the BSS? If gdb and others can find them there, why can't launchmon? Does a dlsym not find these symbols?

Jeff Squyres
For corporate legal information go to: