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
Yorlik_ has joined #ste||ar
Yorlik has quit [Ping timeout: 268 seconds]
hkaiser has quit [Quit: Bye!]
K-ballo1 has joined #ste||ar
K-ballo has quit [Ping timeout: 268 seconds]
K-ballo1 is now known as K-ballo
K-ballo has quit [Ping timeout: 240 seconds]
K-ballo1 has joined #ste||ar
K-ballo1 is now known as K-ballo
hkaiser has joined #ste||ar
diehlpk_work has joined #ste||ar
<pansysk75[m]> should it be possible to use hpx::annotated_function or hpx::scoped_annotation inside HPX's scan_partitioner.hpp?
<pansysk75[m]> I can't get them to work and mr Hartmut said something about circular dependencies last time i tried to use somthing like hpx::cout inside an HPX header
<pansysk75[m]> pansysk75[m]: if that doesnt't work, I can probably directly use __itt functions provided by the intel API
<hkaiser> pansysk75[m]: well, those should actually work
<hkaiser> the circular dependency I mentioned was related to performance counters
<pansysk75[m]> oh ok, i will troubleshoot some more i guess :)
<pansysk75[m]> so hpx::cout would work inside an hpx header?
<hkaiser> don't use hpx::cout, use std::cout if you have to
<pansysk75[m]> aha ok, thank you
<hkaiser> all: Giannis just created the HPX V1.8.1-rc2 release candidate
<hkaiser> please try it out
<hkaiser> Thanks to everybody helping with this!
<gonidelis[m]> hkaiser: I love this guy
<gonidelis[m]> So basically we guessed correctly that `__awaitable` encapsulates the Promise type defaulted to void within its definition
<gonidelis[m]> Doesn
<gonidelis[m]> Doesn't that answer our question? satacker
<satacker[m]> This is different from my question
<satacker[m]> The standard implementation should have extracted it's `promise_type`
<satacker[m]> There are 2 coroutine contexts
<satacker[m]> 1. Natural
<satacker[m]> 2. By promise_type using await transform
<satacker[m]> It is written more clearly in P12880
<satacker[m]> P1288*
<satacker[m]> s/P12880/P1288/
<satacker[m]> An even more primitive question is - is http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2018/p1288r0.pdf an accepted paper?
<gonidelis[m]> seeing it's from the same author as p2300 fells like they embeded 1288 to 2300
<gonidelis[m]> but in any case, the question is why does it not extract the promise that is inherent to the the `__awaitable` type?
<gonidelis[m]> satacker: ^^
<satacker[m]> <gonidelis[m]> "but in any case, the question is..." <- yes, +1
<john98zakaria[m]> Is there an example of distributed load balancing?
<john98zakaria[m]> I was thinking of creating an hpx component that has a function getLoad() which I will then query and use to move the component to a another locality whose load is lower.
<john98zakaria[m]> Is there an Hpx way of writing distributed executors?
<hkaiser> john98zakaria[m]: no full example
<hkaiser> but the API is hpx::components::migrate(comp_id, whereto_loc_id)
<hkaiser> john98zakaria[m]: some tests demonstrate its use, though
<john98zakaria[m]> hkaiser: With load balancing?
<hkaiser> no, using this api
<john98zakaria[m]> Ah ok I'll cook my solution using the migration call then
<hkaiser> if you come up with a small example while you do, I'd appreciate you contributed it ;-)
K-ballo1 has joined #ste||ar
<john98zakaria[m]> What happens when I move a component that is actively computing something?
<john98zakaria[m]> Does it continue on the other locality elsewhere?
K-ballo has quit [Ping timeout: 240 seconds]
K-ballo1 is now known as K-ballo
<hkaiser> migrate will wait automatically untill all pending work for the component ceases
<john98zakaria[m]> hkaiser: I'll publish that after the thesis so my code example wouldn't get caught in a plagurism detector
<hkaiser> john98zakaria[m]: thanks!
diehlpk_work has quit [Remote host closed the connection]