Its MAP_ANON on OS X... It works on Linux as well but it is marked deprecated.
-Nathan
On Wed, Jul 24, 2013 at 01:10:38PM +, Joshua Ladd wrote:
> Great feedback. Thanks, Jeff. We always compile on IB machines (in fact, no
> one in MLNX has a Mac to test on OSX.) We appreciate the help!! We w
FWIW: my report was from a non-IB Linux box.
On Jul 24, 2013, at 6:10 AM, Joshua Ladd wrote:
> Great feedback. Thanks, Jeff. We always compile on IB machines (in fact, no
> one in MLNX has a Mac to test on OSX.) We appreciate the help!! We will
> iterate on this.
>
> Josh
>
> -Original M
Thanks, Ralph!!! I know the guys have been working hard trying to get this to
work with an arbitrary BTL, seems we aren't there yet...
Josh
From: devel-boun...@open-mpi.org [mailto:devel-boun...@open-mpi.org] On Behalf
Of Ralph Castain
Sent: Wednesday, July 24, 2013 8:20 AM
To: Open MPI Devel
Nope, not intended behavior. It's a bug.
Josh
-Original Message-
From: devel-boun...@open-mpi.org [mailto:devel-boun...@open-mpi.org] On Behalf
Of Jeff Squyres (jsquyres)
Sent: Wednesday, July 24, 2013 8:37 AM
To: Open MPI Developers
Cc: Mike Dubman
Subject: Re: [OMPI devel] OpenSHMEM u
Great feedback. Thanks, Jeff. We always compile on IB machines (in fact, no one
in MLNX has a Mac to test on OSX.) We appreciate the help!! We will iterate on
this.
Josh
-Original Message-
From: devel-boun...@open-mpi.org [mailto:devel-boun...@open-mpi.org] On Behalf
Of Jeff Squyres (j
On Jul 23, 2013, at 9:43 AM, Joshua Ladd wrote:
> To build with OSHMEM support, build as you would OMPI but simply include
> ‘--with-oshmem’ on your configure line. This will get you started.
FWIW, Open shmem builds even if you don't specify --with-oshmem.
Is this intended?
--
Jeff Squyres
j
In addition to the I-don't-have-verbs compile errors on my Mac, I also get
these:
base/memheap_base_alloc.c: In function '__shm_attach':
base/memheap_base_alloc.c:216: error: 'SHM_HUGETLB' undeclared (first use in
this function)
base/memheap_base_alloc.c:216: error: (Each undeclared identifier i
I think this is going to need to be refactored, Josh. Looks to me like whoever
wrote this "locked" it to IB. That's a problem for me, at least - I believe
that OpenShmem should use the MPI transports, but respect the OMPI abstraction
layer.
In other words, you should use the transports to move
HI'm afraid this doesn't build for us folks that don't have IB on their
system. From what I can see, it looks like you broke a fundamental abstraction
layer rule by including a header from a component into a completely different
component. That is definitely *not* allowed for precisely t
I'm afraid Josh got bit by grabbing the wrong URL - here's the correct clone
command for anyone not Josh:
git clone https://bitbucket.org/jladd_math/mlnx-oshmem.git
On Jul 23, 2013, at 6:43 AM, Joshua Ladd wrote:
> Dear OMPI Developers,
>
> I have put Mellanox OpenSHMEM up for review on my
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 24/07/13 09:42, Ralph Castain wrote:
> Not to 1.6 series, but it is in the about-to-be-released 1.7.3,
> and will be there from that point onwards.
Oh dear, I cannot delay this machine any more to change to 1.7.x. :-(
> Still waiting to see if it
11 matches
Mail list logo
