hkaiser changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar-group.org | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | This channel is logged: irclog.cct.lsu.edu
hkaiser has quit [Quit: Bye!]
K-ballo1 has joined #ste||ar
K-ballo has quit [Ping timeout: 246 seconds]
K-ballo1 is now known as K-ballo
hkaiser has joined #ste||ar
K-ballo1 has joined #ste||ar
K-ballo has quit [Ping timeout: 256 seconds]
K-ballo1 is now known as K-ballo
hkaiser has quit [Quit: Bye!]
hkaiser has joined #ste||ar
diehlpk_work has joined #ste||ar
hkaiser has quit [Ping timeout: 246 seconds]
K-ballo1 has joined #ste||ar
K-ballo has quit [Ping timeout: 252 seconds]
K-ballo1 is now known as K-ballo
hkaiser has joined #ste||ar
<gonidelis[m]> dkaratza: hkaiser https://hpx-docs.stellar-group.org/latest/html/libs/core/iterator_support/docs/index.html?highlight=is_iterator should id adding links here an overdemand?
hkaiser has quit [Quit: Bye!]
beojan has joined #ste||ar
<beojan> Hi. I'm working on porting (/ evaluating porting) Gaudi (the ATLAS and LHCb core framework) to use HPX, in order to build a multi-node version
<beojan> I have a few questions. First, is it possible to use HPX without a persistent runtime that has to be initialized at program start. We're envisioning something along the lines of a TBB task_arena.
<beojan> Gaudi's architecture is such that it consists of a number of modules that are essentially Python modules. The main process is python, and these modules are all loaded and glued together with Python.
<beojan> The second, following on from this, is whether it's possible to essentially have nested HPX "instances"? We'd like to have one instance coordinating our nodes, and a separate instance on each node coordinating the threads.
apop has quit [Ping timeout: 260 seconds]
hkaiser has joined #ste||ar
<beojan> Hi. Reposting in case someone missed it earlier.
<beojan> I'm working on porting (/ evaluating porting) Gaudi (the ATLAS and LHCb core framework) to use HPX, in order to build a multi-node version
<beojan> have a few questions. First, is it possible to use HPX without a persistent runtime that has to be initialized at program start. We're envisioning something along the lines of a TBB task_arena.
<beojan> Gaudi's architecture is such that it consists of a number of modules that are essentially Python modules. The main process is python, and these modules are all loaded and glued together with Python.
<beojan> The second, following on from this, is whether it's possible to essentially have nested HPX "instances"? We'd like to have one instance coordinating our nodes, and a separate instance on each node coordinating the threads.
<hkaiser> beojan: you can initialize hpx at any point and exit it later as well
<hkaiser> you can also restart it, if you want
<hkaiser> hpx can not be nested inside the same executable
<hkaiser> (I meant process)
<beojan> Thanks
<beojan> Is it OK to fake the argc and argv passed to HPX_Start, and do you have any guidance on pitfalls to be aware of if we do this?
<hkaiser> beojan: we've been initializing hpx from a python module before
<hkaiser> that also demonstrates how to get at argc/argv
<hkaiser> essentially you expose init/exit functions to Python and invoke it as needed (e.g. https://github.com/STEllAR-GROUP/phylanx/blob/master/python/phylanx/core/config.py)
<hkaiser> in this case we called init from Python and exit on module unload (https://github.com/STEllAR-GROUP/phylanx/blob/master/python/phylanx/core/config.py)
<beojan> I see.
<beojan> That's similar to how I started prototyping (but my prototype is much simpler because I don't want the user fiddling with HPX configuration arguments anyway), but my prototype seemed a little too hacky
beojan has quit [Quit: Konversation terminated!]
tufei__ has joined #ste||ar
K-ballo has quit [Ping timeout: 252 seconds]
K-ballo has joined #ste||ar