I found the problem - someone had a typo in btl_openib_mca.c. The threshold need to be set to the module eager limit as that is the only thing defined at that point.
Thanks for bringing it to our attention! I’ll set it up to go into 1.8.6 > On May 25, 2015, at 3:04 AM, Xavier Besseron <xavier.besse...@uni.lu> wrote: > > Hi, > > Thanks for your reply Ralph. > > The option only I'm using when configuring OpenMPI is '--prefix'. > When checking the config.log file, I see > > configure:208504: checking whether the openib BTL will use malloc hooks > configure:208510: result: yes > > so I guess it is properly enabled (full config.log in attachment of this > email). > > > > However, I think I have the reason of the bug (lines refer to source code of > OpenMPI 1.8.5): > > The default value of memalign_threshold is taken from eager_limit in function > btl_openib_register_mca_params() in btl_openib_mca.c line 717. > But the default value is eager_limit is set in btl_openib_component.c at line > 193 right after the call to btl_openib_register_mca_params(). > > To summarize, memalign_threshold gets its value from eager_limit before this > one gets its value assigned. > > > > Best regards, > > Xavier > > > > > > > > > On Mon, May 25, 2015 at 2:27 AM, Ralph Castain <r...@open-mpi.org > <mailto:r...@open-mpi.org>> wrote: > Looking at the code, we do in fact set the memalign_threshold = eager_limit > by default, but only if you configured with > —enable-btl-openib-malloc-alignment AND/OR we found the malloc hook functions > were available. > > You might check config.log to see if the openib malloc hooks were enabled. My > guess is that they weren’t, for some reason. > > >> On May 24, 2015, at 9:07 AM, Xavier Besseron <xavier.besse...@uni.lu >> <mailto:xavier.besse...@uni.lu>> wrote: >> >> Dear OpenMPI developers / users, >> >> This is much more a comment than a question since I believe I have already >> solved my issue. But I would like to report it. >> >> I have noticed my code performed very badly with OpenMPI when Infinand is >> enabled, sometime +50% or even +100% overhead. >> I also have this slowdown when running with one thread and one process. In >> such case, there is no other MPI call than MPI_Init() and MPI_Finalize(). >> This overhead disappears if I disable at runtime the openib btl, ie with >> '--mca btl ^openib'. >> After further investigation, I figured out it comes from the memory >> allocator which is aligning every memory allocation when Infiniband is used. >> This makes sense because my code is a large irregular C++ code creating and >> deleting many objects. >> >> Just below is the documentation of the relevant MCA parameters coming >> ompi_info: >> >> MCA btl: parameter "btl_openib_memalign" (current value: "32", data source: >> default, level: 9 dev/all, type: int) >> [64 | 32 | 0] - Enable (64bit or 32bit)/Disable(0) memoryalignment >> for all malloc calls if btl openib is used. >> >> MCA btl: parameter "btl_openib_memalign_threshold" (current value: "0", data >> source: default, level: 9 dev/all, type: size_t) >> Allocating memory more than btl_openib_memalign_threshholdbytes >> will automatically be algined to the value of btl_openib_memalign >> bytes.memalign_threshhold defaults to the same value as >> mca_btl_openib_eager_limit. >> >> MCA btl: parameter "btl_openib_eager_limit" (current value: "12288", data >> source: default, level: 4 tuner/basic, type: size_t) >> Maximum size (in bytes, including header) of "short" messages (must >> be >= 1). >> >> >> In the end, the problem is that the default value for >> btl_openib_memalign_threshold is 0, which means that all memory allocations >> are aligned to 32 bits. >> The documentation says that the default value of >> btl_openib_memalign_threshold should be the the same as >> btl_openib_eager_limit, ie 12288 instead of 0. >> >> On my side, changing btl_openib_memalign_threshold to 12288 fixes my >> performance issue. >> However, I believe that the default value of btl_openib_memalign_threshold >> should be fixed in the OpenMPI code (or at least the documentation should be >> fixed). >> >> I tried OpenMPI 1.8.5, 1.7.3 and 1.6.4 and it's all the same. >> >> >> Bonus question: >> As this issue might impact other users, I'm considering applying a global >> fix on our clusters by setting this default value >> etc/openmpi-mca-params.conf. >> Do you see any good reason not doing it? >> >> Thank you for your comments. >> >> Best regards, >> >> Xavier >> >> >> -- >> Dr Xavier BESSERON >> Research associate >> FSTC, University of Luxembourg >> Campus Kirchberg, Office E-007 >> Phone: +352 46 66 44 5418 <tel:%2B352%2046%2066%2044%205418> >> http://luxdem.uni.lu/ <http://luxdem.uni.lu/> >> >> _______________________________________________ >> users mailing list >> us...@open-mpi.org <mailto:us...@open-mpi.org> >> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users >> <http://www.open-mpi.org/mailman/listinfo.cgi/users> >> Link to this post: >> http://www.open-mpi.org/community/lists/users/2015/05/26913.php >> <http://www.open-mpi.org/community/lists/users/2015/05/26913.php> > > > > -- > Dr Xavier BESSERON > Research associate > FSTC, University of Luxembourg > Campus Kirchberg, Office E-007 > Phone: +352 46 66 44 5418 > http://luxdem.uni.lu/ <http://luxdem.uni.lu/> > > > _______________________________________________ > users mailing list > us...@open-mpi.org > Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/users > Link to this post: > http://www.open-mpi.org/community/lists/users/2015/05/26915.php
