February 2016 - mpiwg-tools - mailman3.cels.anl.gov

Re: [mpiwg-tools] Tools WG for Forum
by Mohror, Kathryn 29 Feb '16

29 Feb '16

Having trouble joining the webex. Is anyone out there trying to join? Kathryn _________________________________________________________________ Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA > -----Original Message----- > From: mpiwg-tools [mailto:[email protected]] On > Behalf Of Jeff Squyres (jsquyres) > Sent: Saturday, February 27, 2016 4:32 AM > To: MPI Tool WG <mpiwg-tools(a)lists.mpi-forum.org> > Subject: Re: [mpiwg-tools] Tools WG for Forum > > The webex sessions are now listed on the meeting agenda page: > > http://secure- > web.cisco.com/1LqwRRT2sNBLsvh9to7mFqVINDG1l0ikN3RTFvv1qICbQW35V > 89ZhrBYrLB0kTh4G427AHCm-Scfl-Ksv3qsaJ1DpSox2WETghP5omqf6jS-xW- > rwwHr9KhgslfjxQO89mSamwAjCTNQ4ztXD87AjTybrDyj5cyXMXxenXQ7lD_m8 > 0HDvgJU2lay6XAS4DoZoZZEpgujCrAkz7uD3aIJ3299hJDoGQiPxvi3R- > kOtWfDTuOWMW684g5rhzaGxVjSg3-OAuH7RVcD- > g9yKzR17OGSLmdmjP3NHh4WGYSzH0- > Qb81knmnrxUeDf1GauMsUPI9iy7qii_5uaYQHPR57fFBrNmNqdqfrACfQ_pgizC > tU/http%3A%2F%2Fmeetings.mpi- > forum.org%2Fsecretary%2F2016%2F03%2Fagenda.php > > > > On Feb 21, 2016, at 9:25 PM, Mohror, Kathryn <mohror1(a)llnl.gov> wrote: > > > > No trouble! We'll get it set up for all the tools WG sessions, since others > indicated they'd like to join one or more, too. > > > > Thanks, > > Kathryn > > > > From: mpiwg-tools <mpiwg-tools-bounces(a)lists.mpi-forum.org> on behalf > > of "s.c.rasmussen(a)gmail.com" <s.c.rasmussen(a)gmail.com> > > Reply-To: MPI Tools <mpiwg-tools(a)lists.mpi-forum.org> > > Date: Thursday, February 18, 2016 at 2:05 PM > > To: Kathryn Mohror <kathryn(a)llnl.gov>, MPI Tools > > <mpiwg-tools(a)lists.mpi-forum.org> > > Subject: Re: [mpiwg-tools] Tools WG for Forum > > > > H Kathryn, > > > > I'd be interested in joining a webex for the tools topics on Tuesday if it isn't > too much trouble to set it up. > > > > Thanks, Soren > > > > On Wed, Feb 17, 2016 at 1:56 PM, Kathryn Mohror <kathryn(a)llnl.gov> > wrote: > >> Hi all, > >> > >> We're trying to work out the schedule for our Tools WG meetings at > >> the Forum in case people who can't come to the physical meeting want > >> to join in by webex. The current plan is: > >> > >> (Times in CST - Chicago Time): > >> Monday 1:30 - 3:30 PM, 4:00 - 5:30 PM - Debugger Topics Tuesday 9:00 > >> - 10:30 AM, 11:00 AM - 12:30 PM - Tools Topics (MPI_T, PMPI > >> Replacement) > >> Tuesday 1:30 -3:00 PM - Overflow time for topics that still need > >> discussion, likely tools topics > >> > >> If you are interested in joining the discussion for a particular > >> topic group, let me know so we can get webex set up. Also let me know > >> if the proposed schedule won't work for you in terms of which topics > >> are discussed when and we can possibly switch things up. > >> > >> Thanks, > >> Kathryn > >> > ________________________________________________________________ > _ > >> Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ > >> Scalability Team @ Lawrence Livermore National Laboratory, Livermore, > >> CA, USA > >> > >> > >> _______________________________________________ > >> mpiwg-tools mailing list > >> mpiwg-tools(a)lists.mpi-forum.org > >> http://secure- > web.cisco.com/18mBVqFffhiXsJ1_iLBbikHle8ZcNsoD6EOFqTcZd > >> > 8GD0bmwFwiRsKL95cZ_EpVQTNfIrvl2IGEe_qbKZfgndzhYyxnVZs05CQIMY8bu > WQTLWz > >> > Lf9x_afJByY3656wQFvSOrOCeNjWXPzu7bnjHFFmsAsfQAUZBDFi9xmhrcmLUd > 6sTFf9c > >> ep9j_ojRnD- > OpcU31Xkl1XVV22qsonkVdobX21AoEWgdipBxyf0UKojWjgAynO8LHhEcP > >> Rq92n7GiSR34KVTUaBn85TuRHSOL7AuKN2tutqxHw1Mp- > MhhlebEl6DSxTAYvgNfoWenW > >> > mTXt5R6YIrmMYdquZbo8ag7maiwHoMQWvQ5NXqIy6lIYCK8/http%3A%2F%2 > Flists.mp > >> i-forum.org%2Fmailman%2Flistinfo.cgi%2Fmpiwg-tools > > > > _______________________________________________ > > mpiwg-tools mailing list > > mpiwg-tools(a)lists.mpi-forum.org > > http://secure- > web.cisco.com/1E1mRmadpWQczv_6LyFPXohGYJ9tJiMWjZktfB6u7H > > > Otz_XcBFJiA_gW9t4ODWyHJ8s7XfgajodRpKtzAkOsszFQuboSPuXGyeLqp0fjOv > QyQrNX > > > Sxpk2hG0ATIn0RkYno49YJwovGGBfWj1qaoiztFnnYtHaoeC5sxbuln3fhih3iKfKir > Zo_ > > Nk4n- > b4mo3_d3_Z7hjrr9WWtY2_kvnpjaPzGsKfIANpZW5JOalu1t_HsCVzIfc2xMvSah > 1 > > > APfaln3zm3cu5f0T1eyxfjS35bkgi1QYVNdaZvN5qlDK9iu_wQp0xmUbZzEetTEN > 13mLbt > > > o6qsvl8gbAYyxNwsHaMYcvQLGSF9sz2srpg8sPtaGU/http%3A%2F%2Flists.mpi > -foru > > m.org%2Fmailman%2Flistinfo.cgi%2Fmpiwg-tools > > > -- > Jeff Squyres > jsquyres(a)cisco.com > For corporate legal information go to: > http://www.cisco.com/web/about/doing_business/legal/cri/ > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > http://secure- > web.cisco.com/1E1mRmadpWQczv_6LyFPXohGYJ9tJiMWjZktfB6u7HOtz_XcBF > JiA_gW9t4ODWyHJ8s7XfgajodRpKtzAkOsszFQuboSPuXGyeLqp0fjOvQyQrNXSx > pk2hG0ATIn0RkYno49YJwovGGBfWj1qaoiztFnnYtHaoeC5sxbuln3fhih3iKfKirZ > o_Nk4n- > b4mo3_d3_Z7hjrr9WWtY2_kvnpjaPzGsKfIANpZW5JOalu1t_HsCVzIfc2xMvSah > 1APfaln3zm3cu5f0T1eyxfjS35bkgi1QYVNdaZvN5qlDK9iu_wQp0xmUbZzEetTE > N13mLbto6qsvl8gbAYyxNwsHaMYcvQLGSF9sz2srpg8sPtaGU/http%3A%2F%2 > Flists.mpi-forum.org%2Fmailman%2Flistinfo.cgi%2Fmpiwg-tools

1 0

Issue 33: MPI_T Errata: Scope of guarantees on variable information
by Marc-Andre Hermanns 29 Feb '16

29 Feb '16

Dear all, sorry for the vast delay on my part. Today I finally got to work on the text regarding the clarification of variable scope. The changes are currently local to a branch of mine but I attached an annotated PDF to the issue at: https://github.com/mpi-forum/mpi-issues/issues/33 The document will be the basis for our discussion tomorrow during the performance-centric discussion. Personally, I am not 100% happy with "MPI tool epoch", but as the MPI_T interface is always referenced as "MPI tool information interface" and never directly as MPI_T interface, I chose it for the lack of a better idea. I am open to any suggestions. ;-) Cheers, Marc-Andre -- Marc-Andre Hermanns Jülich Aachen Research Alliance, High Performance Computing (JARA-HPC) Jülich Supercomputing Centre (JSC) Schinkelstrasse 2 52062 Aachen Germany Phone: +49 2461 61 2509 | +49 241 80 24381 Fax: +49 2461 80 6 99753 www.jara.org/jara-hpc email: hermanns(a)jara.rwth-aachen.de

1 0

Re: [mpiwg-tools] Tools WG for Forum
by Jeff Squyres (jsquyres) 27 Feb '16

27 Feb '16

The webex sessions are now listed on the meeting agenda page: http://meetings.mpi-forum.org/secretary/2016/03/agenda.php > On Feb 21, 2016, at 9:25 PM, Mohror, Kathryn <mohror1(a)llnl.gov> wrote: > > No trouble! We’ll get it set up for all the tools WG sessions, since others indicated they’d like to join one or more, too. > > Thanks, > Kathryn > > From: mpiwg-tools <mpiwg-tools-bounces(a)lists.mpi-forum.org> on behalf of "s.c.rasmussen(a)gmail.com" <s.c.rasmussen(a)gmail.com> > Reply-To: MPI Tools <mpiwg-tools(a)lists.mpi-forum.org> > Date: Thursday, February 18, 2016 at 2:05 PM > To: Kathryn Mohror <kathryn(a)llnl.gov>, MPI Tools <mpiwg-tools(a)lists.mpi-forum.org> > Subject: Re: [mpiwg-tools] Tools WG for Forum > > H Kathryn, > > I'd be interested in joining a webex for the tools topics on Tuesday if it isn't too much trouble to set it up. > > Thanks, Soren > > On Wed, Feb 17, 2016 at 1:56 PM, Kathryn Mohror <kathryn(a)llnl.gov> wrote: >> Hi all, >> >> We're trying to work out the schedule for our Tools WG meetings at the Forum >> in case people who can't come to the physical meeting want to join in by >> webex. The current plan is: >> >> (Times in CST - Chicago Time): >> Monday 1:30 - 3:30 PM, 4:00 - 5:30 PM - Debugger Topics >> Tuesday 9:00 - 10:30 AM, 11:00 AM - 12:30 PM - Tools Topics (MPI_T, PMPI >> Replacement) >> Tuesday 1:30 -3:00 PM - Overflow time for topics that still need discussion, >> likely tools topics >> >> If you are interested in joining the discussion for a particular topic >> group, let me know so we can get webex set up. Also let me know if the >> proposed schedule won't work for you in terms of which topics are discussed >> when and we can possibly switch things up. >> >> Thanks, >> Kathryn >> _________________________________________________________________ >> Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ >> Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, >> USA >> >> >> _______________________________________________ >> mpiwg-tools mailing list >> mpiwg-tools(a)lists.mpi-forum.org >> http://secure-web.cisco.com/18mBVqFffhiXsJ1_iLBbikHle8ZcNsoD6EOFqTcZd8GD0bm… > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Cancel meeting today Feb 25, 2016 ?
by Marc-Andre Hermanns 24 Feb '16

24 Feb '16

Dear all, Martin and Kathryn are on travel, and I did not get to write the text on either the MPI_T clarification or an initial draft on MPI_T callbacks, yet. As we have the face-to-face meeting next week, I would cancel today's call, unless someone else would like to discuss anything specifically. Once the webex information for next week is available we will post it here. We will have two 1.5h slots, both on Monday and Tuesday. You can find details here: http://meetings.mpi-forum.org/secretary/2016/03/agenda.php Cheers, Marc-Andre -- Marc-Andre Hermanns Jülich Aachen Research Alliance, High Performance Computing (JARA-HPC) Jülich Supercomputing Centre (JSC) Schinkelstrasse 2 52062 Aachen Germany Phone: +49 2461 61 2509 | +49 241 80 24381 Fax: +49 2461 80 6 99753 www.jara.org/jara-hpc email: hermanns(a)jara.rwth-aachen.de

1 0

Re: [mpiwg-tools] Tools WG for Forum
by Mohror, Kathryn 21 Feb '16

21 Feb '16

No trouble! We’ll get it set up for all the tools WG sessions, since others indicated they’d like to join one or more, too. Thanks, Kathryn From: mpiwg-tools <mpiwg-tools-bounces(a)lists.mpi-forum.org<mailto:[email protected]>> on behalf of "s.c.rasmussen(a)gmail.com<mailto:[email protected]>" <s.c.rasmussen(a)gmail.com<mailto:[email protected]>> Reply-To: MPI Tools <mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]>> Date: Thursday, February 18, 2016 at 2:05 PM To: Kathryn Mohror <kathryn(a)llnl.gov<mailto:[email protected]>>, MPI Tools <mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]>> Subject: Re: [mpiwg-tools] Tools WG for Forum H Kathryn, I'd be interested in joining a webex for the tools topics on Tuesday if it isn't too much trouble to set it up. Thanks, Soren On Wed, Feb 17, 2016 at 1:56 PM, Kathryn Mohror <kathryn(a)llnl.gov<mailto:[email protected]>> wrote: Hi all, We're trying to work out the schedule for our Tools WG meetings at the Forum in case people who can't come to the physical meeting want to join in by webex. The current plan is: (Times in CST - Chicago Time): Monday 1:30 - 3:30 PM, 4:00 - 5:30 PM - Debugger Topics Tuesday 9:00 - 10:30 AM, 11:00 AM - 12:30 PM - Tools Topics (MPI_T, PMPI Replacement) Tuesday 1:30 -3:00 PM - Overflow time for topics that still need discussion, likely tools topics If you are interested in joining the discussion for a particular topic group, let me know so we can get webex set up. Also let me know if the proposed schedule won't work for you in terms of which topics are discussed when and we can possibly switch things up. Thanks, Kathryn _________________________________________________________________ Kathryn Mohror, kathryn(a)llnl.gov<mailto:[email protected]>, http://scalability.llnl.gov/ Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA _______________________________________________ mpiwg-tools mailing list mpiwg-tools(a)lists.mpi-forum.org<mailto:[email protected]> http://secure-web.cisco.com/18mBVqFffhiXsJ1_iLBbikHle8ZcNsoD6EOFqTcZd8GD0bm…

1 0

Re: [mpiwg-tools] Tools WG for Forum
by Soren Rasmussen 18 Feb '16

18 Feb '16

H Kathryn, I'd be interested in joining a webex for the tools topics on Tuesday if it isn't too much trouble to set it up. Thanks, Soren On Wed, Feb 17, 2016 at 1:56 PM, Kathryn Mohror <kathryn(a)llnl.gov> wrote: > Hi all, > > We're trying to work out the schedule for our Tools WG meetings at the > Forum > in case people who can't come to the physical meeting want to join in by > webex. The current plan is: > > (Times in CST - Chicago Time): > Monday 1:30 - 3:30 PM, 4:00 - 5:30 PM - Debugger Topics > Tuesday 9:00 - 10:30 AM, 11:00 AM - 12:30 PM - Tools Topics (MPI_T, PMPI > Replacement) > Tuesday 1:30 -3:00 PM - Overflow time for topics that still need > discussion, > likely tools topics > > If you are interested in joining the discussion for a particular topic > group, let me know so we can get webex set up. Also let me know if the > proposed schedule won't work for you in terms of which topics are discussed > when and we can possibly switch things up. > > Thanks, > Kathryn > _________________________________________________________________ > Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ > Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, > USA > > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-tools >

1 0

Tools WG for Forum
by Kathryn Mohror 17 Feb '16

17 Feb '16

Hi all, We're trying to work out the schedule for our Tools WG meetings at the Forum in case people who can't come to the physical meeting want to join in by webex. The current plan is: (Times in CST - Chicago Time): Monday 1:30 - 3:30 PM, 4:00 - 5:30 PM - Debugger Topics Tuesday 9:00 - 10:30 AM, 11:00 AM - 12:30 PM - Tools Topics (MPI_T, PMPI Replacement) Tuesday 1:30 -3:00 PM - Overflow time for topics that still need discussion, likely tools topics If you are interested in joining the discussion for a particular topic group, let me know so we can get webex set up. Also let me know if the proposed schedule won't work for you in terms of which topics are discussed when and we can possibly switch things up. Thanks, Kathryn _________________________________________________________________ Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA

1 0

Meeting reminder - Thursday Feb 11
by Kathryn Mohror 09 Feb '16

09 Feb '16

Hey all, This is a reminder of our regular meeting on Thursday Feb 11 at 8:00 AM PST, 11:00 AM EST, 5:00 PM CET. To join the webex, go here: https://cisco.webex.com/ciscosales/j.php?MTID=m2c3f15fb156cb498d61af2e40e4da 26b For meeting information and past meeting notes: https://github.com/mpiwg-tools/tools-issues/wiki/Meetings Agenda: - Quick topic: New errata item to clarify cvar value guarantees (see email thread) - Quick topic: Double check we are ready for forum readings for existing items (no new news is good news) - Debugger issues - In the last debugger-focused call, we had a good discussion of what might be done and how we might approach it: https://github.com/mpiwg-tools/tools-issues/wiki/Notes-2016-01-14 - Let's discuss concrete items to move forward on this and what the priorities are Talk to you soon! Kathryn _________________________________________________________________ Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA

1 0

Re: [mpiwg-tools] MPI_T question about cvar value guarantees
by Mohror, Kathryn 09 Feb '16

09 Feb '16

Excellent catch. I'll try to have it ready to discuss Thursday so we can get it in for reading by the Monday 2/15 deadline. Thanks! Kathryn _________________________________________________________________ Kathryn Mohror, kathryn(a)llnl.gov, http://scalability.llnl.gov/ Scalability Team @ Lawrence Livermore National Laboratory, Livermore, CA, USA > -----Original Message----- > From: mpiwg-tools [mailto:[email protected]] On > Behalf Of Schulz, Martin > Sent: Friday, February 5, 2016 5:24 PM > To: mpiwg-tools(a)lists.mpi-forum.org > Subject: Re: [mpiwg-tools] MPI_T question about cvar value guarantees > > Yes, I agree, we should probably submit an errata for this text. > > Martin > > > On 2/5/16, 4:50 PM, "mpiwg-tools on behalf of Jeff Squyres (jsquyres)" > <mpiwg-tools-bounces(a)lists.mpi-forum.org on behalf of jsquyres(a)cisco.com> > wrote: > > >> On Feb 5, 2016, at 4:26 PM, Schulz, Martin <schulz6(a)llnl.gov> wrote: > >> > >> That¹s a good question and I don¹t think we say - my original intent > >>when writing the document (based on my personal model that I had in > >>my head for > >> this) was to provide no guarantees once MPI_T gets uninitialized, so > >>in this case both versions would be OK. In fact, it could be even > >>worse - what happens in case 1 if the component is not reloaded? Do > >>you still have access to the variable at all? This is probably more > >>likely if the MPI_T only block is after MPI_Finalize (since this > >>would unload components), but the principle is the same. > >> > >> For the discovery/query routines we state that the numbering is not > >>allowed to change between ³runs², which may have been the wrong thing > >>to say - it probably should say ³while MPI_T is initialized². In this > >>case, there would really not be any guarantees at all. If you wanted > >>guarantees, you would have to keep MPI_T initialized. > > > >That sounds fine to me. I think that this is worth an errata. > > > >-- > >Jeff Squyres > >jsquyres(a)cisco.com > >For corporate legal information go to: > >http://www.cisco.com/web/about/doing_business/legal/cri/ > > > >_______________________________________________ > >mpiwg-tools mailing list > >mpiwg-tools(a)lists.mpi-forum.org > >http://secure-web.cisco.com/1YQzVZ9p1gLekoVYv9VGQ_- > CNEYsPlQQtkva9wBeUEu > >V5I > >NONI4Nm-vTVCCxm4xjArNq97pldPUVCOWShL7- > hhjDh_0uYjZ9MYdA28tvg1pIvOl8shxbR > >xZh > >3FbapA2wu5Q99HsmgPLjCxBBGktJijRvb9BIcaGKHBzMAFlDIX_mz5BZNMo2JR > mvTX1neOf > >byF > >PTV_rVef2O7blagdgDD7YOeu54XlV9ImkS8qcvLeGLcxfmzeu1pPNMN4IoiiAao7I > z2Xh9y > >7MM > >dtqOZi6XcylO9kxyregmJifR44SqEhZukURI8THsglPLQvSmkYM0iVJuc9ZlEXl9nru > au-A > >FKa > >VH8WgAzBQkDzVeG2StEuxo/http%3A%2F%2Flists.mpi- > forum.org%2Fmailman%2Flis > >tin > >fo.cgi%2Fmpiwg-tools > > > > _______________________________________________ > mpiwg-tools mailing list > mpiwg-tools(a)lists.mpi-forum.org > http://secure- > web.cisco.com/1RsFhgnyhJqbl3e3u1W6tmBFwlbZcXXlimgtJ1E7tDwUt3VZ96biQz > 4_-IVFynNm- > cO3cyxVRGHEZ2_OySkm1e63TxATLQaJLPPZByzyLF0sq1VN11SXNfPvyvXsOkMPN > VB9z6XPPa25CTesfw0HpSrJsg-ThVxy3IIvUbR_rQHOLuMdmbS- > ffFmqsw9A7FD1uJMxPWMB7HMyflXJpSCgKCNJ5myWylldNAnIALjyjtsJnn7KTWv > puf3XPhClSCzvRwur9iQEZJmC7goyLuHzB01moGDZ_0AWo1GNaMuYSwkfjZlrsRY > nyqTiQs3CDyL2H4oup0366qOpXPHY_oZGNJ2AVj4UMvKx65oMXzEP9Vk/http%3 > A%2F%2Flists.mpi-forum.org%2Fmailman%2Flistinfo.cgi%2Fmpiwg-tools

1 0

Re: [mpiwg-tools] [mpiwg-p2p] Message matching for tools
by Jim Dinan 09 Feb '16

09 Feb '16

Hi All, I'm preparing the updated draft of info assertions for a reading in March. Where did we land on an advice regarding tools? Do we want advice (1) to users, that they info keys may impact tools and/or (2) to tools that they should check info? For instance, we could extend the current advice with the following sentence: Setting info hints on the predefined communicators \const{MPI\_COMM\_WORLD} and \const{MPI\_COMM\_SELF} may have unintended effects, as changes to these global objects may affect all components of the application, including libraries. The usage of info hints may also impact the effectiveness of tools. ~Jim. On Fri, Dec 18, 2015 at 5:24 AM, Marc-Andre Hermanns < hermanns(a)jara.rwth-aachen.de> wrote: > Hi Jeff, > > > at the moment we don't handle MPI_THREAD_MULTIPLE at all. But we want > > to get there ;-) > > > > > > You should vote for endpoints, as this may help you out here, > > particularly if users start mapping endpoints 1:1 w/ threads. > > That would certainly ease things for us in these situations. > Unfortunately endpoints force use to adapt other infrastructure in our > measurement system. > > > b) Creating a derived datatype on the fly to add tool-level data to > > the original payload may induce a large overhead in practically > > _every_ send & receive operation and perturb the measurement. > > > > > > You should evaluate this experimentally. I wrote a simple test > > (https://github.com/jeffhammond/BigMPI/blob/master/test/perf/typepiggy.c > ) > > and measured 1.5 us per call of overhead to create a datatype. That > > is not significant except for very small messages. > > Thanks for the pointer. You are right. I should evaluate this further. > 1.5us does indeed seem tolerable. I wonder how the influence of the > derived datatype is on overall messaging performance, though. > > This is also something I should evaluate in the process. > > Cheers, > Marc-Andre > > -- > Marc-Andre Hermanns > Jülich Aachen Research Alliance, > High Performance Computing (JARA-HPC) > Jülich Supercomputing Centre (JSC) > > Schinkelstrasse 2 > 52062 Aachen > Germany > > Phone: +49 2461 61 2509 | +49 241 80 24381 > Fax: +49 2461 80 6 99753 > www.jara.org/jara-hpc > email: hermanns(a)jara.rwth-aachen.de > > > _______________________________________________ > mpiwg-p2p mailing list > mpiwg-p2p(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-p2p >

1 0