hkaiser changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar.cct.lsu.edu | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | Buildbot: http://rostam.cct.lsu.edu/ | Log: http://irclog.cct.lsu.edu/ | GSoC: https://github.com/STEllAR-GROUP/hpx/wiki/Google-Summer-of-Code-%28GSoC%29-2020
hkaiser has quit [Quit: bye]
nan11 has quit [Remote host closed the connection]
karame_ has quit [Quit: Ping timeout (120 seconds)]
Amy2 has quit [Ping timeout: 240 seconds]
Amy2 has joined #ste||ar
weilewei has quit [Remote host closed the connection]
akheir has quit [Quit: Leaving]
bita_ has quit [Read error: Connection reset by peer]
Amy2 has quit [Ping timeout: 256 seconds]
Amy2 has joined #ste||ar
<jbjnr> From what I can see, HPX_SetupCUDA creates a Cuda::cuda interface library, but doesn't actually link cuda to it
<heller1> jbjnr: phew, finally making progress on the sender/receiver front....
<jbjnr> cool
<jbjnr> I must take a look very soon
Yorlik has quit [Read error: Connection reset by peer]
<heller1> let's see. My goal for today is to get the sender concept in...
hkaiser has joined #ste||ar
hkaiser has quit [Ping timeout: 252 seconds]
hkaiser has joined #ste||ar
diehlpk_work_ has quit [Remote host closed the connection]
diehlpk_work_ has joined #ste||ar
nan11 has joined #ste||ar
<heller1> forgive me for spamming you with pull requests ... I hope those are easier to digest then
<hkaiser> heller1: a couple of smaller PRs are better than one giant one
<heller1> That's the idea...
<hkaiser> heller1: those customization points introduced... should they use the CP mechanisms we use for executors?
<heller1> What do we use there?
<heller1> The tag_invoke mechanism has the promise to avoid name collisions
<heller1> hkaiser: I jumped on the hype train of this: http://wg21.link/P1895
karame_ has joined #ste||ar
<heller1> hkaiser: This might make the transition easier, as it is easier to disambiguate expressions with the same name
<hkaiser> nod
<hkaiser> should we go back and use the same for the executors?
<heller1> I don't think it is worth the effort right now
<hkaiser> ok
<heller1> I suggest to make that change once we are closer to have an actual P0443 compliant implementation
<hkaiser> ok
<hkaiser> will your tag_invoke implementation be able coexist with what the standard might/will have?
<heller1> my roadmap: 1) implement the sender concept 2) adapt hpx::future to accept senders and act as a sender 3) port wait_all 4) port when_all 5) port then
<hkaiser> sounds good
<heller1> yes, I discussed that with Eric, the different tag types that will exist until we have a full standardization are a potential problem. However, it should be rather trivial to hook into other implementations with different tag types. That is, just create a simple mapping for, for example hpx::basic_execution::set_value_t <-> decltype(libunifex::set_value)
<heller1> or just namespace hpx { namespace basic_execution { using set_value_t = std::set_value_t; }}
<heller1> the nice thing here however is, if there happens to be a convergence in semantics between the different tag types, it should be possible to close that gap with providing the appropriate algorithm
<hkaiser> ok, makes sense
<heller1> not convergence, divergence is what I meant
<hkaiser> nod
weilewei has joined #ste||ar
karame_ has quit [Ping timeout: 245 seconds]
<weilewei> \o/ wow
<weilewei> This paper is published on 05/29/2016 but it was mentioning history that dated 05/30/2016, a day ahead of itself: http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2016/p0233r1.pdf
<weilewei> 1st time to read a C++ proposal...
weilewei has quit [Remote host closed the connection]
weilewei has joined #ste||ar
akheir has joined #ste||ar
<hkaiser> weilewei: 5/30 most probably was the mailing deadline, and 5/29 was the day they put the document together ;-)
<weilewei> hkaiser Ah, IC
<hkaiser> heller1: pls send this to Katie
<hkaiser> heller1: is Thomas (Fahringer) still working with HPX?
<heller1> hkaiser: not as far as I know
<heller1> hkaiser: you can always ask him ;)
<hkaiser> nah, I'm just curious
<hkaiser> so AllScale is dead?
<heller1> there is no more funding for it
<heller1> but still paper output
<heller1> braindead, I suppose is the right term
<heller1> it was supposed to be my ticket into acadmia :P
bita_ has joined #ste||ar
<gonidelis[m]> hkaiser: rori been studying membes-initializer lists and they are like WOW! Truly magnificent concept... Anyways as for the test that I am trying to implement in order to check the iterator type change, is it ok if I imitate this one? I mean, if this one works with different iterator types then my changes are fine (???)
<hkaiser> gonidelis[m]: sure, steal as much code as you like ;-)
<gonidelis[m]> What is the difference between `foreach.cpp` under `tests/unit/algorithms' and `foreach_range.cpp` under 'tests/unit/container_algorithms`?
<hkaiser> gonidelis[m]: well, the code is different, no?
<K-ballo> I bet one of those is range-aware
<gonidelis[m]> hkaiser well, as far as I see it's the same ;p . And the reason I am asking is because I thought foreach_range had not been implemented yet (isn't that what I am doing?)
<hkaiser> good catch
<hkaiser> I wasn't aware of that
<rori> yes just did a diff and they pretty much identical
<hkaiser> probably somebody died over implementing things - most likely me ;-)
<hkaiser> gonidelis[m]: if that's the case, then feel free to change the foreach_range.cpp as needed
<gonidelis[m]> hkaiser: I just wanted to make sure (mostly that I am aware of my job here ;p). Thank you both. I will proceed accordingly
<hkaiser> heller1: any ideas ^^??
<heller1> Maybe due to this: https://status.circleci.com/
<heller1> ?
<heller1> Partial outages
<hkaiser> could be
<hkaiser> it's just a docker pull, you'd think this couldn't cause any problems as long as the dockerhub is up
<ms[m]> hkaiser: uhh... I'm sure that's temporary
<ms[m]> the image is definitely still on dockerhub
<ms[m]> most other builds seem to be running ok now
nan11 has quit [Remote host closed the connection]
nan111 has joined #ste||ar
weilewei has quit [Remote host closed the connection]
Yorlik has joined #ste||ar
karame_ has joined #ste||ar
weilewei has joined #ste||ar
nikunj has quit [Ping timeout: 260 seconds]
nikunj has joined #ste||ar
nan111 has quit [Remote host closed the connection]