[dcmf] 32-bit ROMIO: What is the best solution?
Rob Ross
rross at mcs.anl.gov
Thu Feb 28 12:56:10 CST 2008
That's fine with me too. Michael, please provide us a list of
interested IBM parties, and we'll get this set up.
Pavan, can you put RobL, Kevin, Sam, and I on the mpich2-bg list
(along with Rajeev and others on the MPICH team as appropriate)?
Thanks,
Rob
On Feb 28, 2008, at 12:42 PM, Pavan Balaji wrote:
>
> There's an mpich2-bg list that I just created if you want to move
> there.
>
> -- Pavan
>
> On 02/28/2008 12:39 PM, Rob Ross wrote:
>> Rajeev had suggested this too. Let's go with mpich2-dev. Can you
>> give us a list of people that should be added to the list?
>> Thanks,
>> Rob
>> On Feb 28, 2008, at 12:29 PM, Michael Blocksome wrote:
>>>
>>> I agree .. we need to get over the discussion about the discussion
>>> as fast as possible. :)
>>>
>>> The dcmf list has worked well, but I think we started using it
>>> because there were no alternatives to the issue at hand. Given
>>> that the mpich2 and global arrays communities have different
>>> licenses, our long term direction for the dcmf community should be
>>> to support the dcmf library and move discussion of other
>>> "external" libraries (like mpich2 and global arrays) into thier
>>> communities.
>>>
>>> I just noticed that there is a mpich2-dev at mcs.anl.gov mailing list
>>> already set up. Could we use this list to collaborate on the
>>> mpich2 implementation?
>>>
>>> Michael Blocksome
>>> Blue Gene Messaging Team Lead
>>> Advanced Systems SW Development
>>> blocksom at us.ibm.com
>>>
>>>
>>> Rob Ross <rross at mcs.anl.gov> wrote on 02/28/2008 12:17:06 PM:
>>>
>>> > Hi,
>>> >
>>> > We've been using the DCMF list for all sorts of things that aren't
>>> > directly related to DCMF without any issues (as far as I know);
>>> it has
>>> > become the successful technical communication path between IBM
>>> and ANL
>>> > so far.
>>> >
>>> > However, if you'd like to create a separate list for discussing
>>> this,
>>> > we can do that. Is that what you would like to do? If so, can
>>> you give
>>> > us a list of people on the IBM side that should be on the new
>>> list?
>>> > I'd like to be done discussing where we're going to talk about
>>> this
>>> > issue ASAP and concentrate on fixing it.
>>> >
>>> > Thanks,
>>> >
>>> > Rob
>>> >
>>> > On Feb 28, 2008, at 11:18 AM, Michael Blocksome wrote:
>>> >
>>> > >
>>> > > Unfortunately, this MPI_Aint discussion really has nothing to do
>>> > > with DCMF. The entrire problem and solution is with MPICH2.
>>> > >
>>> > > Is there another, less broad, mpich mailing list that we could
>>> use?
>>> > > Maybe a "mpich-development" or something like that?
>>> > >
>>> > > Michael Blocksome
>>> > > Blue Gene Messaging Team Lead
>>> > > Advanced Systems SW Development
>>> > > blocksom at us.ibm.com
>>> > >
>>> > >
>>> > > dcmf-bounces at lists.anl-external.org wrote on 02/28/2008
>>> 11:01:09 AM:
>>> > >
>>> > > > I'd say this mailing list is better for it right now.
>>> Posting on the
>>> > > > broader mpich-discuss might just distract us.
>>> > > >
>>> > > > Rajeev
>>> > > >
>>> > > > From: dcmf-bounces at lists.anl-external.org [mailto:dcmf-
>>> > > > bounces at lists.anl-external.org] On Behalf Of Bob Cernohous
>>> > > > Sent: Wednesday, February 27, 2008 4:19 PM
>>> > > > To: dcmf at lists.anl-external.org
>>> > > > Subject: Re: [dcmf] 32-bit ROMIO: What is the best solution?
>>> > >
>>> > > >
>>> > > > bobc at us.ibm.com wrote on 02/08/2008 02:41:40 PM:
>>> > > >
>>> > > > >We all realize there are issues with MPI_Aint's in 32 bit
>>> > > implementations.
>>> > > > >Using signed 32 bit addresses along with 64 bit offsets can
>>> > > result in
>>> > > > >some pretty broken code. I've reproduced several problems on
>>> > > > >BGL/BGP/linux. The problems are most obvious with romio
>>> files >
>>> > > 2G or
>>> > > > >virtual addresses > 2G.
>>> > > >
>>> > > > Earlier this month I posted some comments to this mailing
>>> list about
>>> > > > 32-bit MPICH/ROMIO/MPI_Aint issues (see above and the
>>> archive).
>>> > > >
>>> > > > We've started to work on the problem. We're (optionally)
>>> making
>>> > > > MPI_Aint a 64 bit signed value even when integers/pointers
>>> are 32
>>> > > bit
>>> > > > values.
>>> > > >
>>> > > > We'd like to start sending fix patches to the community for
>>> comment.
>>> > > >
>>> > > > However, since we are (in parallel) contributing our BGP
>>> code back
>>> > > > to MPICH and trying to converge on a common code base of
>>> 1.0.7,
>>> > > > and because this problem applies to any 32-bit platform,
>>> > > > we thought we should move the discussion to mpich-discuss to
>>> get
>>> > > > the larger community's comments.
>>> > > >
>>> > > > It's unlikely, at this point, that our patches would apply to
>>> > > MPICH2,
>>> > > > but that is our goal after the 1.0.7/BGP patch settles.
>>> For now,
>>> > > we'd
>>> > > > be sending these as RFC (request for comment) patches.
>>> > > >
>>> > > > Please let us know if you agree with discussing this on mpich-
>>> > > discuss,
>>> > > > or if there is a better place.
>>> > > >
>>> > > > Bob Cernohous: (T/L 553) 507-253-6093
>>> > > >
>>> > > > BobC at us.ibm.com
>>> > > > IBM Rochester, Building 030-2(C335), Department 61L
>>> > > > 3605 Hwy 52 North, Rochester, MN 55901-7829
>>> > > >
>>> > > > > Chaos reigns within.
>>> > > > > Reflect, repent, and reboot.
>>> > > > > Order shall
>>> return._______________________________________________
>>> > > > dcmf mailing list
>>> > > > dcmf at lists.anl-external.org
>>> > > > http://lists.anl-external.org/cgi-bin/mailman/listinfo/dcmf
>>> > > > http://dcmf.anl-external.
>>> > org/wiki_______________________________________________
>>> > > dcmf mailing list
>>> > > dcmf at lists.anl-external.org
>>> > > http://lists.anl-external.org/cgi-bin/mailman/listinfo/dcmf
>>> > > http://dcmf.anl-external.org/wiki
>>> >
>> _______________________________________________
>> dcmf mailing list
>> dcmf at lists.anl-external.org
>> http://lists.anl-external.org/cgi-bin/mailman/listinfo/dcmf
>> http://dcmf.anl-external.org/wiki
>
> --
> Pavan Balaji
> http://www.mcs.anl.gov/~balaji
>
More information about the dcmf
mailing list