Re: [hpx-users] HPX and mvapich2

est, > > Karol > On 04/09/2019 21:19, Thomas Heller wrote: > > Hi Karol, > > As Hartmut said, CMake should be able to detect your mvapich2 > installation. Please check the following: > Is the MPI compiler wrapper in your path? > > Another reason could be, that your

Re: [hpx-users] HPX and mvapich2

Hi Karol, As Hartmut said, CMake should be able to detect your mvapich2 installation. Please check the following: Is the MPI compiler wrapper in your path? Another reason could be, that your cmake version is too old. Which cmake version, and which version of mvapich2 are you using? Regards,

Re: [hpx-users] Debugging bad locality number on MPI?

ble networking in console mode, or if I enable networking > > if the node number equals -1, then it works. > > > > Looks like the this was introduced with https://github.com/STEllAR- > > GROUP/hpx/commit/ffb8470e6a1143e9e1c95c39eff58eec322148d3#diff- > > 86850321552971332a5de97

Re: [hpx-users] Debugging bad locality number on MPI?

/mpi_environment.cpp#L30-L52 On Fri, Jul 5, 2019 at 10:41 AM Thomas Heller wrote: > The initialization of the MPI parcelport is done by checking environment > variables set by the most common MPI implementations. Which MPI > implementation do you use? Can you attach the output of `mpirun e

Re: [hpx-users] Debugging bad locality number on MPI?

The initialization of the MPI parcelport is done by checking environment variables set by the most common MPI implementations. Which MPI implementation do you use? Can you attach the output of `mpirun env` maybe? Andreas Schäfer schrieb am Fr., 5. Juli 2019 08:58: > > On 01:11 Fri 05 Jul ,

Re: [hpx-users] Benchmarks

Hi Karol, The below looks like you are on a Cray machine, forcing static linking to off (it is off by default anyway) and using the Cray compiler wrapper, which adds some special static linking flags by default. This is the reason for the error you are seeing. CMake thinks it needs to link

Re: [hpx-users] performance study of HPX runtime

s of syntax and semantics) to shared memory and distributed memory parallelization which is hard to find in any other solution. So even if there are initial hiccups, I'd like to invite you to work with us to make your experience and HPX in general better. > > Thanks, > Yongkee

Re: [hpx-users] performance study of HPX runtime

Hi Yongkee, In addition to the performance tests, we published a wide range of papers looking at the performance. Pleas have a look here: http://stellar-group.org/publications/ On Sun, Jan 27, 2019 at 6:16 PM Hartmut Kaiser wrote: > > Hey Yongkee, > > Thanks for your interest in HPX! > > >

Re: [hpx-users] Contributing to HPX

On Wed, Oct 24, 2018 at 10:14 PM Biddiscombe, John A. wrote: > Ahmed > > I'm a little bit frightened by the idea of you working on hwloc+hpx - the > reason is the same as the one that caused problems in gsoc. > Something I have in mind, which would be a good fit with clear direction and

Re: [hpx-users] Segmentation fault with mpi4py

Hi, >From looking at the stacktrace, it is a segfault coming directly out of MPI, which is then caught by our signal handlers. In theory, there shouldn't be any problem with having multiple MPI libraries running within HPX. The HPX parcelport tries to be a good citizen and creates its own

Re: [hpx-users] list of OS names

Patrick Welche schrieb am Mo., 13. Aug. 2018 16:16: > On Mon, Aug 13, 2018 at 08:06:25AM -0500, Hartmut Kaiser wrote: > > FWIW, I have merged #3402 just now. > > Thanks! > > I have got as far as linking libhpx now - somehow there is a -ldl in > link.txt (and in the generated pkg-config .pc

Re: [hpx-users] CMake Targets for HPX

On Sun, Aug 12, 2018 at 7:13 PM Jayesh Badwaik wrote: > Is it possible to use HPX with CMake with Targets? For example, can I > do `target_link_libraries(mylib INTERFACE/PUBLIC/PRIVATE HPX::HPX)`? > Please see my reply to your ticket here: https://github.com/STEllAR-GROUP/hpx/issues/3408 > >

Re: [hpx-users] list of OS names

Hi Patrick, Yes, it's hard to maintain. The problem is, that we have to rely on some OS specific parts in very few places. From the top of my that includes: - User level context switching (that's the most crucial part) - Dynamic plugin loading (the one you stumbled over in your second mail) -

Re: [hpx-users] Regarding GSoc 2018 Project

Hi, Did you already approach me on IRC? I answered some questions there already. If you haven't read them, I'll be happy to forward them again. Please also use the public mailing list, so others can chime in and your fellow students are aware of what's being discussed. Am 17.02.2018 9:46 vorm.

Re: [hpx-users] Performance Counter Data Interpretation

Hi Kilian, The cumulative overhead counter indeed looks very odd. What it means is the accumulated time *not* spend doing useful work. Let's break it appart (For the single thread case): - The overall runtime of your application appears to be 2 seconds. - The number of total tasks run is 127704

Re: [hpx-users] Strong scalability of hpx dataflow and async

Hi Steve, On Mittwoch, 11. Oktober 2017 18:24:36 CEST Steve Petruzza wrote: > Going back to the channels discussion, how do you know if a channel has been > already registered or not? > > Look at the following code below (running with 2 localities): > - they create or connect to the same channel

Re: [hpx-users] thread_stacksize_large for HPX_PLAIN_ACTION not working?

On 10/08/2017 08:14 PM, Hartmut Kaiser wrote: > Steve, > >> Thank you Harmut, I will report the bug. > > > Thanks! > >> But so this means that the only way to use my function remotely is to get >> rid of all possible stack allocations? >> >> Or is there any setting/parameter of hpx to ask for

Re: [hpx-users] HPX concurrent queue ?

Hi Jean-Loup, Am 24.08.2017 1:07 nachm. schrieb "Jean-Loup Tastet" < jean-loup.tas...@epfl.ch>: Hi all, I am currently looking for a data structure that basically behaves like TBB's `concurrent_queue`, except for the `pop()` operation which would return a `hpx::future`, and would never fail.

Re: [hpx-users] Runtime error with simple class as component

Hi Tobias, On Donnerstag, 1. Juni 2017 14:49:58 CEST Tobias Gauweiler wrote: > Hello, > > After compiling and running the code below i get the error message "*** > stack smashing detected ***". > > I would highly appreciate it if you could help me. > > Best regards > Tobias Gauweiler > > >

Re: [hpx-users] partitioned_vectors questions

ks in advance, > > ct > > On Mon, Apr 24, 2017 at 4:52 PM, Thomas Heller <thom.hel...@gmail.com> > > wrote: > > Hi Chris, > > > > Am 24.04.2017 7:30 nachm. schrieb "ct clmsn" <ct.cl...@gmail.com>: > > > > two questions for

Re: [hpx-users] [GSoC17] Am I write only backgrounds related in my proposal?

Am 24.03.2017 2:58 vorm. schrieb "권태국" : Hi, I have a question. In educational and programming background, am I write only things related in my proposal topic or not? Whatever you think might be helpful for us to consider you a good candidate... Taeguk Kwon,

Re: [hpx-users] GSoC 2017 Proposal Work on Parallel Algorithms for HPX

Hi again I've been through the full proposal now, and I think what I wrote below still holds. On Dienstag, 21. März 2017 19:53:10 CET you wrote: > On Dienstag, 21. März 2017 19:46:39 CET you wrote: > > Hi Aja, > > > > I think you should focus your proposal. As it stands right now, you will

Re: [hpx-users] linking/cmake problem

Hi Alex, On Mittwoch, 15. März 2017 22:40:16 CET Alexander Neundorf wrote: > Hi, > > I'm building hpx without cuda, and in examples/qt/, tools/ and > tools/inspect/ cmake errors out with the following error: > This is embarrassing, please see inline for more discussion > CMake Error at

Re: [hpx-users] SLURM: requesting a different number of localities than have been assigned

On Freitag, 24. Februar 2017 01:32:24 CET Tim Biedert wrote: > Below you find the gdb thread infos for the single idling process (recall: > all processes have 100% but one, which idles). As you can see, there a > quite a few threads stuck in some kind of TCMalloc spin lock. > > Most backtraces

Re: [hpx-users] SLURM: requesting a different number of localities than have been assigned

t fixes your deadlocks? > > Best, > > Tim > > > > On 02/22/2017 10:30 PM, Thomas Heller wrote: >> If you are running with at least 256 localities (or are using >> hpx::lcos::barrier with more than 256 participating threads), please >> be aware >

Re: [hpx-users] SLURM: requesting a different number of localities than have been assigned

On Mittwoch, 22. Februar 2017 22:02:57 CET Tim Biedert wrote: > > We were having troubles when running with multiple localities up until > > recently. If you update to today's top of master, at least the problems > > coming from HPX directly should go away. > > Hi, thanks for the feedback!

Re: [hpx-users] SLURM: requesting a different number of localities than have been assigned

On Mittwoch, 22. Februar 2017 22:02:57 CET Tim Biedert wrote: > > We were having troubles when running with multiple localities up until > > recently. If you update to today's top of master, at least the problems > > coming from HPX directly should go away. > > Hi, thanks for the feedback!

Re: [hpx-users] SLURM: requesting a different number of localities than have been assigned

Hi Tim, On Mittwoch, 22. Februar 2017 17:03:45 CET Tim Biedert wrote: > Hi, > > when running my application on our (new) SLURM cluster, I receive the > following warning (if using more than one node): > > hpx::init: command line warning: --hpx:localities used when running with > SLURM,

Re: [hpx-users] HPX build fail

Hi Justs, I am not exactly sure what's causing your issues. When looking at your script, I noticed that you don't use the CMake Toolchain file I strongly recommend to use when working on a cray (See here for more info:

Re: [hpx-users] archive data bstream data chunk size mismatch: HPX(serialization_error)

On Montag, 12. Dezember 2016 10:12:11 CET Tim Biedert wrote: > Thank you for the feedback! Unfortunately, this hasn't solved the > deserialization error. However, I'm sure this has made the code more > robust! ;-) > > What do you suggest to localize the deserialization error? Is it > possible

Re: [hpx-users] Attempting to do sparse triangular solve

Hi Riccardo, The problem seems to be that you capture everything by reference. You might want to consider to capture some variables by value. HTH, Thomas Am 08.12.2016 4:57 nachm. schrieb "Riccardo Rossi" : > Dear List, > > this is a continuation of previous

Re: [hpx-users] the simple code of previous message...

Hi Riccardo, >From a first glance, the second parameter of do_work is a non-const reference. We don't support non-const references here. >From a first glance Am 06.12.2016 5:51 nachm. schrieb "Riccardo Rossi" : any help in understanding why this does not compile would be

Re: [hpx-users] On-the-fly (de)compression in serialization

Hi Tim, Just a quick answer upfront, we support compression of action arguments using snappy or bzip2 already. There is no need for you to do it yourself. See here for an example on how to enable it for given actions (you can use async/apply as usual, the parcel stuff is merely to unit test) :

Re: [hpx-users] Serializing type-erased classes

On Dienstag, 20. September 2016 13:44:21 CEST Michael Levine wrote: > Thanks for getting back to me so quickly. It's great that the boilerplate is > able to make polymorphic serialization so straightforward and effortless. > > Likewise, thanks for the information about those macros. Having gone

Re: [hpx-users] Serializing type-erased classes

Am 20.09.2016 4:02 nachm. schrieb "Hartmut Kaiser" : > > > Sorry, I forgot to add macros to the classes, please see below: Just for completeness, serialization over unique_ptr works just as fine. The important piece are the serialization boilerplate macros. > > > >

Re: [hpx-users] Get raw address of component instance

On Mittwoch, 31. August 2016 17:35:18 CEST Tim Biedert wrote: > Hi, > > on each locality I have several components, which directly interact > (e.g. one "big" component has data which is read by lots of other > components). > > These components will never be migrated, so the pinning is alright. >

Re: [hpx-users] Memory management in a distributed app

On Montag, 29. August 2016 14:29:32 CEST Michael Levine wrote: > Hi Thomas, > > > -Original Message- > > From: hpx-users-boun...@stellar.cct.lsu.edu [mailto:hpx-users- > > boun...@stellar.cct.lsu.edu] On Behalf Of Thomas Heller > > Sent: August-2

Re: [hpx-users] Memory management in a distributed app

ic allocation_list_type allocation_list_; > > }; // class Matrix_Allocator > > The allocation_list_ is used to track the allocated size of a given > pointer, to determine to which free_list should the pointer be added > upon destruct

Re: [hpx-users] HPX futures and random number generators

On Mittwoch, 22. Juni 2016 16:24:01 CEST Michael Levine wrote: > I’m looking for some advice regarding creating an > std::vector > where the elements of the > innermost vector use a random number generator. > > > > Typically, if I were to use to generate this vector of random

Re: [hpx-users] clang-format

of hpx, > > then > > >> my dream of having each project I work on, formatted according to the > > >> style of that project could be achieved. > > >> > > >> Does anyone have such a file for hpx, or can one convert easily one of > > >>

Re: [hpx-users] hpx 0.9.11 segmentation fault running on multiple localities

"info threads". This will give you a list of all operating system threads running. One (or more) is then sitting in nanosleep. Switch to that thread with "thread I". Once you are in the correct frame, you can navigate the frame up until it reaches your code. The variable i is in the han

Re: [hpx-users] Program failing when not compiling with -O3

Am 17.11.2015 11:39 nachm. schrieb "Andreas Schäfer" : > > On 15:08 Tue 17 Nov , Hartmut Kaiser wrote: > > Why do you need your own component registry etc.? > > I can't use macros because a user would need to instantiate them > manually. Users forget about that stuff, so we do