Hi Jeff,
> inside of GA, I have to manually do everything that ARMCI_Malloc did> previously?
yes the user has to. Do you think that is a bad idea?
> Will it be possible for RMA to distinguish between segments which are> locally allocated as shared versus private? Presumably, it is much
Certainly, the proposal mentions a memory attribute to inform the MPI implementation of any special memory like say preregistered memory.
Thanks,
Vinod.> Date: Mon, 8 Dec 2008 13:55:16 -0600> From: jeff.science(a)gmail.com> To: mpi3-rma(a)lists.mpi-forum.org> Subject: Re: [Mpi3-rma] FW: draft of a proposal for RMA interfaces> > I apologize in advance for potential noobiness...> > The GA user manual has an extensive discussion of the coupling between> GA, ARMCI and MA> (http://www.emsl.pnl.gov/docs/global/um/init.html#ma) If MPI3> one-sided communication replaces ARMCI as the messaging layer of GA,> what happens at the interface between these libraries? In particular,> in what way would the sentence "shared memory is used to store global> array data and is allocated by the Global Arrays run-time system> called ARMCI" change if "ARMCI" changes to "MPI3"?> > I imagine that it is not within the scope of MPI to start explicitly> managing local memory but does that mean that if I happen to get> inside of GA, I have to manually do everything that ARMCI_Malloc did> previously?> > Will it be possible for RMA to distinguish between segments which are> locally allocated as shared versus private? Presumably, it is much> harder to make RMA on shared segments thread safe and where one is> RMA-ing only between private segments then performance is increased in> the context of multithreading if the more rigorous locking procedure> used for the former is turned off.> > Thanks,> > Jeff> > On Mon, Dec 8, 2008 at 1:12 PM, Vinod tipparaju <tipparajuv(a)hotmail.com> wrote:> >> >> > As I am preparing the faq Rajeev suggested, please send me any more> > questions you want included.> > Thanks,> > Vinod.> >> > ________________________________> > From: thakur(a)mcs.anl.gov> > To: mpi3-rma(a)lists.mpi-forum.org> > Date: Mon, 8 Dec 2008 12:04:14 -0600> > Subject: [Mpi3-rma] FW: draft of a proposal for RMA interfaces> >> > Just resending this as a reminder. It would be good to have an FAQ that> > answers commonly asked questions.> >> > Rajeev> > ________________________________> > From: Rajeev Thakur [mailto:[email protected]]> > Sent: Sunday, October 19, 2008 6:43 PM> > To: 'MPI 3.0 Remote Memory Access working group'> > Subject: RE: [Mpi3-rma] draft of a proposal for RMA interfaces> >> > I think it would be good to add an FAQ section at the end, containing> > answers to questions that will be asked of any RMA proposal, such as> > non-cache-coherent, does it meet the needs of PGAS/Global Arrays, support> > for heterogeneous, how does the target know of completion, how does it> > interplay with existing RMA spec, etc. It will make sure that the proposal> > addresses those issues, that we ourselves are clear of the answers, and that> > they are not repeatedly raised at each meeting.> > Rajeev> >> >> Richard Graham wrote:> >>> >>> Just to get discussion going again. Talking with several folks I have> >>> heard several concerns expressed about the proposal. I think it would> >>> be good if these (and others) could be raised on the list, so we can> >>> start discussion. We can continue this next week in Chicago, but> >>> Vinod will not be able to make this meeting, so an e-mail discussion> >>> will help.> >>>> >>> Here are the issues I have hear of so far:> >>> - May not work well on current h/w that is not cache coherent, as it> >>> requires a remote thread in this case. I believe this is for the SX> >>> series of machines, but Jesper please correct me if I am wrong here.> >>> What would be an alternative approach that could provide expected> >>> performance on platforms that may require work on the remote end for> >>> RMA for correctness, and work well on platforms that do require very> >>> specific remote cache management (or other actions) for correctness ?> >>> - Concern about future high-end platforms, under that assumption that> >>> these will not be cache coherent (and will actually have caches – if> >>> they don't this is not a concern), and therefore this proposal is> >>> aimed at a short-lived technical capability.> >>> - What is missing ?> >> > _______________________________________________> > mpi3-rma mailing list> > mpi3-rma(a)lists.mpi-forum.org> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma> >> >> > > > -- > Jeff Hammond> The University of Chicago> http://home.uchicago.edu/~jhammond/> > _______________________________________________> mpi3-rma mailing list> mpi3-rma(a)lists.mpi-forum.org> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
I apologize in advance for potential noobiness...
The GA user manual has an extensive discussion of the coupling between
GA, ARMCI and MA
(http://www.emsl.pnl.gov/docs/global/um/init.html#ma) If MPI3
one-sided communication replaces ARMCI as the messaging layer of GA,
what happens at the interface between these libraries? In particular,
in what way would the sentence "shared memory is used to store global
array data and is allocated by the Global Arrays run-time system
called ARMCI" change if "ARMCI" changes to "MPI3"?
I imagine that it is not within the scope of MPI to start explicitly
managing local memory but does that mean that if I happen to get
inside of GA, I have to manually do everything that ARMCI_Malloc did
previously?
Will it be possible for RMA to distinguish between segments which are
locally allocated as shared versus private? Presumably, it is much
harder to make RMA on shared segments thread safe and where one is
RMA-ing only between private segments then performance is increased in
the context of multithreading if the more rigorous locking procedure
used for the former is turned off.
Thanks,
Jeff
On Mon, Dec 8, 2008 at 1:12 PM, Vinod tipparaju <tipparajuv(a)hotmail.com> wrote:
>
>
> As I am preparing the faq Rajeev suggested, please send me any more
> questions you want included.
> Thanks,
> Vinod.
>
> ________________________________
> From: thakur(a)mcs.anl.gov
> To: mpi3-rma(a)lists.mpi-forum.org
> Date: Mon, 8 Dec 2008 12:04:14 -0600
> Subject: [Mpi3-rma] FW: draft of a proposal for RMA interfaces
>
> Just resending this as a reminder. It would be good to have an FAQ that
> answers commonly asked questions.
>
> Rajeev
> ________________________________
> From: Rajeev Thakur [mailto:[email protected]]
> Sent: Sunday, October 19, 2008 6:43 PM
> To: 'MPI 3.0 Remote Memory Access working group'
> Subject: RE: [Mpi3-rma] draft of a proposal for RMA interfaces
>
> I think it would be good to add an FAQ section at the end, containing
> answers to questions that will be asked of any RMA proposal, such as
> non-cache-coherent, does it meet the needs of PGAS/Global Arrays, support
> for heterogeneous, how does the target know of completion, how does it
> interplay with existing RMA spec, etc. It will make sure that the proposal
> addresses those issues, that we ourselves are clear of the answers, and that
> they are not repeatedly raised at each meeting.
> Rajeev
>
>> Richard Graham wrote:
>>
>>> Just to get discussion going again. Talking with several folks I have
>>> heard several concerns expressed about the proposal. I think it would
>>> be good if these (and others) could be raised on the list, so we can
>>> start discussion. We can continue this next week in Chicago, but
>>> Vinod will not be able to make this meeting, so an e-mail discussion
>>> will help.
>>>
>>> Here are the issues I have hear of so far:
>>> - May not work well on current h/w that is not cache coherent, as it
>>> requires a remote thread in this case. I believe this is for the SX
>>> series of machines, but Jesper please correct me if I am wrong here.
>>> What would be an alternative approach that could provide expected
>>> performance on platforms that may require work on the remote end for
>>> RMA for correctness, and work well on platforms that do require very
>>> specific remote cache management (or other actions) for correctness ?
>>> - Concern about future high-end platforms, under that assumption that
>>> these will not be cache coherent (and will actually have caches � if
>>> they don't this is not a concern), and therefore this proposal is
>>> aimed at a short-lived technical capability.
>>> - What is missing ?
>
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
>
--
Jeff Hammond
The University of Chicago
http://home.uchicago.edu/~jhammond/
As I am preparing the faq Rajeev suggested, please send me any more questions you want included.Thanks,Vinod.From: thakur(a)mcs.anl.govTo: mpi3-rma(a)lists.mpi-forum.orgDate: Mon, 8 Dec 2008 12:04:14 -0600Subject: [Mpi3-rma] FW: draft of a proposal for RMA interfaces
RE: [Mpi3-rma] draft of a proposal for RMA interfaces
Just resending this as a reminder. It would be good to have
an FAQ that answers commonly asked questions.
Rajeev
From: Rajeev Thakur [mailto:[email protected]]
Sent: Sunday, October 19, 2008 6:43 PMTo: 'MPI 3.0 Remote
Memory Access working group'Subject: RE: [Mpi3-rma] draft of a
proposal for RMA interfaces
I think it would be good to add an FAQ section at the end,
containing answers to questions that will be asked of any RMA proposal, such as
non-cache-coherent, does it meet the needs of PGAS/Global Arrays, support for
heterogeneous, how does the target know of completion, how does it interplay
with existing RMA spec, etc. It will make sure that the proposal addresses those
issues, that we ourselves are clear of the answers, and that they are not
repeatedly raised at each meeting.
Rajeev
> Richard Graham wrote: >
>> Just to get discussion going again. Talking with
several folks I have >> heard several concerns
expressed about the proposal. I think it would >>
be good if these (and others) could be raised on the list, so we can
>> start discussion. We can continue this next
week in Chicago, but >> Vinod will not be able to
make this meeting, so an e-mail discussion >> will
help. >> >> Here are
the issues I have hear of so far: >> - May not
work well on current h/w that is not cache coherent, as it >> requires a remote thread in this case. I believe this is for the
SX >> series of machines, but Jesper please
correct me if I am wrong here. >> What would be an
alternative approach that could provide expected >> performance on platforms that may require work on the remote end
for >> RMA for correctness, and work well on
platforms that do require very >> specific remote
cache management (or other actions) for correctness ? >> - Concern about future high-end platforms, under that assumption
that >> these will not be cache coherent (and will
actually have caches – if >> they don’t this is
not a concern), and therefore this proposal is >>
aimed at a short-lived technical capability. >> -
What is missing ?
Just resending this as a reminder. It would be good to have an FAQ that
answers commonly asked questions.
Rajeev
_____
From: Rajeev Thakur [mailto:[email protected]]
Sent: Sunday, October 19, 2008 6:43 PM
To: 'MPI 3.0 Remote Memory Access working group'
Subject: RE: [Mpi3-rma] draft of a proposal for RMA interfaces
I think it would be good to add an FAQ section at the end, containing
answers to questions that will be asked of any RMA proposal, such as
non-cache-coherent, does it meet the needs of PGAS/Global Arrays, support
for heterogeneous, how does the target know of completion, how does it
interplay with existing RMA spec, etc. It will make sure that the proposal
addresses those issues, that we ourselves are clear of the answers, and that
they are not repeatedly raised at each meeting.
Rajeev
> Richard Graham wrote:
>
>> Just to get discussion going again. Talking with several folks I have
>> heard several concerns expressed about the proposal. I think it would
>> be good if these (and others) could be raised on the list, so we can
>> start discussion. We can continue this next week in Chicago, but
>> Vinod will not be able to make this meeting, so an e-mail discussion
>> will help.
>>
>> Here are the issues I have hear of so far:
>> - May not work well on current h/w that is not cache coherent, as it
>> requires a remote thread in this case. I believe this is for the SX
>> series of machines, but Jesper please correct me if I am wrong here.
>> What would be an alternative approach that could provide expected
>> performance on platforms that may require work on the remote end for
>> RMA for correctness, and work well on platforms that do require very
>> specific remote cache management (or other actions) for correctness ?
>> - Concern about future high-end platforms, under that assumption that
>> these will not be cache coherent (and will actually have caches - if
>> they don't this is not a concern), and therefore this proposal is
>> aimed at a short-lived technical capability.
>> - What is missing ?
The phone call is set up for Tue, Dec 9, 9.00am Central (GMT-6) = 10.00am
Eastern, 4.00pm Germany.
Dial-in Number: 1-218-486-8700
Participant Access Code: 35704
Agenda:
* Next steps for RMA proposal from Vinod Tipparaju et al (original version
from Dublin meeting attached)
* coherency issues in specific architectures (need input from participants
representing those architectures)
If time permits:
* Ticket #37 for MPI 2.2 (examples for the one-sided synchronization
semantics https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/37 )
* Ticket #35 for MPI 2.2 (registration of datatypes on MPI_Ops for RMA
https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/35 )
If you have other agenda items, let me know.
Rajeev
same here
Jesper
On Thu, Dec 04, 2008 at 05:14:29PM +0100, Hubert Ritzdorf wrote:
> An hour earlier is also fine for me.
>
> Hubert
>
> Richard Graham wrote:
> >Can we do this an hour earlier ?
> >
> >Rich
> >
> >
> >On 12/3/08 12:28 PM, "Hubert Ritzdorf" <ritzdorf(a)it.neclab.eu> wrote:
> >
> > It works also for me.
> >
> > Hubert
> >
> > Rajeev Thakur wrote:
> > > We are planning to have a conference call next Tuesday (Dec 9)
> > for the RMA
> > > working group. Since there are many time zones, would 10.00am
> > Central Time
> > > (GMT-6) work for everyone who would like to participate? I will
> > send the
> > > calling details later.
> > >
> > > Rajeev
> > >
> > > _______________________________________________
> > > mpi3-rma mailing list
> > > mpi3-rma(a)lists.mpi-forum.org
> > > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> > >
> > >
> >
> > ------------------------------------------------------------------------
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma(a)lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> >------------------------------------------------------------------------
> >
> >_______________________________________________
> >mpi3-rma mailing list
> >mpi3-rma(a)lists.mpi-forum.org
> >http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
An hour earlier is also fine for me.
Hubert
Richard Graham wrote:
> Can we do this an hour earlier ?
>
> Rich
>
>
> On 12/3/08 12:28 PM, "Hubert Ritzdorf" <ritzdorf(a)it.neclab.eu> wrote:
>
> It works also for me.
>
> Hubert
>
> Rajeev Thakur wrote:
> > We are planning to have a conference call next Tuesday (Dec 9)
> for the RMA
> > working group. Since there are many time zones, would 10.00am
> Central Time
> > (GMT-6) work for everyone who would like to participate? I will
> send the
> > calling details later.
> >
> > Rajeev
> >
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma(a)lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> >
>
> ------------------------------------------------------------------------
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
That might make it difficult for folks in Europe to join in.
Rajeev
_____
From: mpi3-rma-bounces(a)lists.mpi-forum.org
[mailto:[email protected]] On Behalf Of Su Huang
Sent: Wednesday, December 03, 2008 1:55 PM
To: MPI 3.0 Remote Memory Access working group
Subject: Re: [Mpi3-rma] RMA phone call next Tuesday, Dec 9
Rajeev,
I work with Dick and Robert at IBM and would like to participate the
conference call. Can we schedule the call in the afternoon?
Thanks, Su
Communication Protocols,
Advance Clustering Technology Team (ACTT)
IBM System and Technology Group
Poughkeepsie, NY 12601
Phone : (845)-433-7927 (T/L: 293)
Inactive hide details for Darius Buntinas ---12/03/2008 02:23:46 PM---That's
fine with me. I'm available all day (at least forDarius Buntinas
---12/03/2008 02:23:46 PM---That's fine with me. I'm available all day (at
least for now :-) ). -d
From:
Darius Buntinas <buntinas(a)mcs.anl.gov>
To:
"MPI 3.0 Remote Memory Access working group" <mpi3-rma(a)lists.mpi-forum.org>
Date:
12/03/2008 02:23 PM
Subject:
Re: [Mpi3-rma] RMA phone call next Tuesday, Dec 9
_____
That's fine with me. I'm available all day (at least for now :-) ).
-d
On 12/03/2008 11:37 AM, Richard Graham wrote:
> Can we do this an hour earlier ?
>
> Rich
>
>
> On 12/3/08 12:28 PM, "Hubert Ritzdorf" <ritzdorf(a)it.neclab.eu> wrote:
>
> It works also for me.
>
> Hubert
>
> Rajeev Thakur wrote:
> > We are planning to have a conference call next Tuesday (Dec 9) for
> the RMA
> > working group. Since there are many time zones, would 10.00am
> Central Time
> > (GMT-6) work for everyone who would like to participate? I will
> send the
> > calling details later.
> >
> > Rajeev
> >
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma(a)lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> >
>
>
------------------------------------------------------------------------
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
_______________________________________________
mpi3-rma mailing list
mpi3-rma(a)lists.mpi-forum.org
http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
Rajeev,
I work with Dick and Robert at IBM and would like to participate the
conference call. Can we schedule the call in the afternoon?
Thanks, Su
Communication Protocols,
Advance Clustering Technology Team (ACTT)
IBM System and Technology Group
Poughkeepsie, NY 12601
Phone : (845)-433-7927 (T/L: 293)
From: Darius Buntinas <buntinas(a)mcs.anl.gov>
To: "MPI 3.0 Remote Memory Access working group"
<mpi3-rma(a)lists.mpi-forum.org>
Date: 12/03/2008 02:23 PM
Subject: Re: [Mpi3-rma] RMA phone call next Tuesday, Dec 9
That's fine with me. I'm available all day (at least for now :-) ).
-d
On 12/03/2008 11:37 AM, Richard Graham wrote:
> Can we do this an hour earlier ?
>
> Rich
>
>
> On 12/3/08 12:28 PM, "Hubert Ritzdorf" <ritzdorf(a)it.neclab.eu> wrote:
>
> It works also for me.
>
> Hubert
>
> Rajeev Thakur wrote:
> > We are planning to have a conference call next Tuesday (Dec 9) for
> the RMA
> > working group. Since there are many time zones, would 10.00am
> Central Time
> > (GMT-6) work for everyone who would like to participate? I will
> send the
> > calling details later.
> >
> > Rajeev
> >
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma(a)lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> >
>
>
------------------------------------------------------------------------
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
_______________________________________________
mpi3-rma mailing list
mpi3-rma(a)lists.mpi-forum.org
http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
That's fine with me. I'm available all day (at least for now :-) ).
-d
On 12/03/2008 11:37 AM, Richard Graham wrote:
> Can we do this an hour earlier ?
>
> Rich
>
>
> On 12/3/08 12:28 PM, "Hubert Ritzdorf" <ritzdorf(a)it.neclab.eu> wrote:
>
> It works also for me.
>
> Hubert
>
> Rajeev Thakur wrote:
> > We are planning to have a conference call next Tuesday (Dec 9) for
> the RMA
> > working group. Since there are many time zones, would 10.00am
> Central Time
> > (GMT-6) work for everyone who would like to participate? I will
> send the
> > calling details later.
> >
> > Rajeev
> >
> > _______________________________________________
> > mpi3-rma mailing list
> > mpi3-rma(a)lists.mpi-forum.org
> > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
> >
> >
>
> ------------------------------------------------------------------------
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> mpi3-rma mailing list
> mpi3-rma(a)lists.mpi-forum.org
> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-rma
