[Mpi3-rma] nested locks

Rajeev Thakur thakur at mcs.anl.gov
Mon Aug 19 16:31:12 CDT 2013


Access epoch refers to an epoch at the process that is initiating the RMA calls (origin process). The epoch can have multiple targets, as in the start-complete case.  The sentence (437:28) applies to all synchronization methods.

To me it sounds like the sentence is saying you cannot start multiple access epochs on the same (origin) process.

Rajeev


On Aug 19, 2013, at 3:58 PM, Jim Dinan wrote:

> Hi Rajeev,
> 
> Re-pasting:
> 
> ---8<---
> pg 437, ln 28: "Distinct access epochs for win at the same process must be disjoint."
> ---8<---
> 
> I think it's still ok (sorry Jeff, I think this is the same thing you were just saying).  That sentence states that one cannot have more than a single access epoch per target concurrently.  I don't think it means that one cannot have concurrent epochs.
> 
>  ~Jim.
> 
> 
> On Mon, Aug 19, 2013 at 4:49 PM, Rajeev Thakur <thakur at mcs.anl.gov> wrote:
> Those are two access epochs on the same rank (say rank 0), as defined by the paragraph above that sentence (pg 437, ln 28).
> 
> On Aug 19, 2013, at 2:52 PM, Jeff Hammond wrote:
> 
> > you mean lock at those ranks, not by those ranks?  those aren't conflicting so i don't see the issue.  maybe i am a dolt though.
> >
> > jeff
> >
> >
> > On Mon, Aug 19, 2013 at 2:44 PM, Rajeev Thakur <thakur at mcs.anl.gov> wrote:
> > Did we make any change in MPI 3.0 to allow nesting of MPI_Win_lock calls? For example,
> >
> > Win_lock(rank 1)
> >     Win_lock(rank 2)
> >
> >     Win_unlock(rank 2)
> > Win_unlock(rank 1)
> >
> > I can't find the text we added or deleted to allow this.
> >
> > But I do see text that disallows this:
> > pg 437, ln 28: "Distinct access epochs for win at the same process must be disjoint."
> >
> > Rajeev
> >
> >
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma at lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> >
> >
> > --
> > Jeff Hammond
> > jeff.science at gmail.com
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma at lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> 
> 
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> 
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma at lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma





More information about the mpiwg-rma mailing list