Hello Ralph and Gilles
Thanks for the clarification. My understanding was that if a BTL was
specified to mpirun, then only BTL (and, therefore, the ob1 PML) will be
used. However, I always saw that is not the case and now I know why.
I do have PSM capable cards (Qlogic IB) in my nodes, and this t
With patch
https://github.com/torvalds/linux/commit/e15f431fe2d53cd4673510736da7d4fa1090e096,
the use of ENOSYS has been clarified.
/*
* This error code is special: arch syscall entry code will return
* -ENOSYS if users try to call a syscall that doesn't exist. To keep
* failures of syscalls
Thanks! I've filed https://github.com/open-mpi/ompi/issues/1607 to follow up.
> On Apr 29, 2016, at 6:43 AM, Håkon Bugge wrote:
>
> With patch
> https://github.com/torvalds/linux/commit/e15f431fe2d53cd4673510736da7d4fa1090e096,
> the use of ENOSYS has been clarified.
>
> /*
> * This error co
Certainly we need to communicate / advertise / evangelize the improvements
in job launch - the largest and most substantial change between the two
branches - and provide some best practice guidelines for usage (use direct
modex for applications with sparse communication patterns and full modex
for
All “green” from me:
Passed: 1347
Failed: 0
> On Apr 28, 2016, at 4:01 PM, Jeff Squyres (jsquyres)
> wrote:
>
> At long last, here's the next v2.0.0 release candidate: 2.0.0rc2:
>
>https://www.open-mpi.org/software/ompi/v2.x/
>
> We didn't keep a good list of all the things that have ch
How about for “developers that have not been following the transition from 1.x
to 2.0? Particularly myself ☺. I started contributing to some specific parts
(psm2 mtl) and following changes. However, I don’t have details of what is
changing in 2.0. I see there could be different level of details
Hi Jeff,
checkpoint/restart is not supported in this release.
Does this release work with totalview? I recall we had some problems,
and do not remember if they were resolved.
We may also want to clarify if any PML/MTLs are experimental in this
release.
MPI_THREAD_MULTIPLE support.
Howard
2
Matias --
You're probably already in pretty good shape. If you're following master and
the v2.x branch (and you are), then your code is already in good shape.
I was thinking mostly for users: transitioning from v1.8/v1.10 series to the
v2.x series -- what kinds of user-noticeable things will t
I'm thinking something like a simple "User's migration guide: 1.8.x/1.10.x -->
2.0.0"
Here's big topics I see so far:
User-Noticeable changes
(i.e., things that may prevent users from simply re-compiling / re-mpirun'ing
their existing MPI app)
---
- mpirun -np behavior
- OMP
FWIW: I think Matias has a good point, though perhaps it belongs on a wiki
page. When we moved the BTLs down to OPAL, for example, it doesn’t impact
users, but would be worth ensuring developer’s and ISVs had a convenient place
to see what changed.
> On Apr 29, 2016, at 10:47 AM, Jeff Squyres
Didn’t OSHMEM up-level its API?
I believe we also have some early support in there for DVM and Singularity, but
not the full-blown capability that is in master. Unsure if we want to advertise
that for 2.0, maybe wait for the updates in 2.1?
> On Apr 29, 2016, at 10:55 AM, Jeff Squyres (jsquyre
Exactly on the nail! Thanks Ralph. I've seen several comments about patches
needing to be ported from 2.x to 1.x. That will definitely be transparent to
users (apologize if I diverged the email intention), but maybe few others will
be grateful to read that wiki entry.
Thanks!
_MAC
-Orig
On Apr 29, 2016, at 2:00 PM, Ralph Castain wrote:
>
> Didn’t OSHMEM up-level its API?
Yeah, actually, we don't have much in NEWS about OSHMEM -- Mellanox: what
should be in there?
> I believe we also have some early support in there for DVM and Singularity,
> but not the full-blown capability
Let's iterate over this in a wiki page -- seems easier than iterating over
email:
https://github.com/open-mpi/ompi/wiki/User-Migration-Guide%3A-1.8.x-and-v1.10.x-to-v2.0.0
(I put a link to this on the main OMPI wiki page too, but currently have it
labeled as DRAFT, just to make it clear that th
Ok, I started a wiki page for this one, too:
https://github.com/open-mpi/ompi/wiki/Developer-Migration-Guide:-v1.8.x-and-v1.10.x-to-v2.x
> On Apr 29, 2016, at 2:02 PM, Cabral, Matias A
> wrote:
>
> Exactly on the nail! Thanks Ralph. I've seen several comments about patches
> needing to be po
OpenSHMEM API is V1.3 at this point. I can send some info if it would help.
Best wishes
Steve...
Sent from my iPhone
> On Apr 29, 2016, at 11:00, Ralph Castain wrote:
>
> Didn’t OSHMEM up-level its API?
>
> I believe we also have some early support in there for DVM and Singularity,
> but not
I don't think all the OSHMEM v1.3 updates made it into v2.0.0...?
> On Apr 29, 2016, at 2:19 PM, Swpoole-Gmail wrote:
>
> OpenSHMEM API is V1.3 at this point. I can send some info if it would help.
>
> Best wishes
> Steve...
> Sent from my iPhone
>
>> On Apr 29, 2016, at 11:00, Ralph Castain
Correct, Jeff. OMPI 2.0.0 will not be OSHMEM 1.3 compliant, but OMPI 2.1.0
will be.
Josh
On Fri, Apr 29, 2016 at 2:21 PM, Jeff Squyres (jsquyres) > wrote:
> I don't think all the OSHMEM v1.3 updates made it into v2.0.0...?
>
>
> > On Apr 29, 2016, at 2:19 PM, Swpoole-Gmail > wrote:
> >
> > Open
Do you guys want to add anything into NEWS about OSHMEM improvements in 2.0.0
(even though it won't be 1.3)?
Or were such improvements hidden down in UCX / MXM?
> On Apr 29, 2016, at 5:40 PM, Joshua Ladd wrote:
>
> Correct, Jeff. OMPI 2.0.0 will not be OSHMEM 1.3 compliant, but OMPI 2.1.0
>
We didn't introduce any improvements to OSHMEM at the OMPI level in 2.0.0
save for the improved job launch times. In OMPI 2.1.0 we will add the
non-blocking iput/iget and alltoall operations - whose inclusion will make
OMPI's OSHMEM a bona fide OSHMEM 1.3.
Josh
On Fri, Apr 29, 2016 at 5:46 PM, Je
Got it; thanks.
> On Apr 29, 2016, at 5:52 PM, Joshua Ladd wrote:
>
> We didn't introduce any improvements to OSHMEM at the OMPI level in 2.0.0
> save for the improved job launch times. In OMPI 2.1.0 we will add the
> non-blocking iput/iget and alltoall operations - whose inclusion will make
21 matches
Mail list logo
