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/ | GSoD: https://developers.google.com/season-of-docs/
Guest65867 has quit [Ping timeout: 276 seconds]
Guest65867 has joined #ste||ar
hkaiser has quit [Quit: bye]
K-ballo has quit [Quit: K-ballo]
Guest65867 has quit [Ping timeout: 240 seconds]
Guest65867 has joined #ste||ar
nikunj has joined #ste||ar
nikunj has quit [Remote host closed the connection]
Guest65867 has quit [Ping timeout: 264 seconds]
Guest65867 has joined #ste||ar
Guest65867 has quit [Ping timeout: 265 seconds]
Guest65867 has joined #ste||ar
Guest65867 has quit [Ping timeout: 245 seconds]
Guest65867 has joined #ste||ar
Guest65867 has quit [Ping timeout: 240 seconds]
Guest65867 has joined #ste||ar
Guest65867 has quit [Ping timeout: 250 seconds]
Guest65867 has joined #ste||ar
K-ballo has joined #ste||ar
hkaiser has joined #ste||ar
<hkaiser> simbergm: hey, g'morning
<hkaiser> simbergm: not sure if you can help, ubt I'd like to ask you for comments: https://circleci.com/gh/STEllAR-GROUP/phylanx/32049
<hkaiser> could it be that hpx_wrap is not exported as a target (hpx_init is)?
<simbergm> hkaiser: I can't really, but hpx_wrap is getting linked so I don't think it's a problem of exporting
<simbergm> do you use any special options in phylanx related to that hpx_init/wrap business?
<hkaiser> simbergm: the -W wrap flag is not propagated
<simbergm> it could be a problem of linking order
<simbergm> ah
<hkaiser> do we have a hpx_wrap target?
<hkaiser> or is it part of hpx_init?
<simbergm> no, there's a separate target
<simbergm> let me ping rori
<hkaiser> ok, then the target is Linux specific - good point
<hkaiser> yes
<simbergm> I'm not in the office right now so can't look right away but I've written to rori
<hkaiser> thanks
<simbergm> our build unit tests work :/ probably because they don't use the hpx_main wrapper
<hkaiser> simbergm: the hpx_main wrapper is on by default, I think
<simbergm> yeah, that's why I was wondering about the build unit tests
<simbergm> but I think they don't need the wrapper, that's why they're ok
<simbergm> i.e. they use hpx::init
<hkaiser> right
<hkaiser> heller: thanks for fixing the 'noexcept'! Excellent catch!
<heller> hkaiser: np, took me a while ;)
<K-ballo> now I'm curious, which noexcept was broken?
<heller> K-ballo: nothing on master, hkaiser added a noexcept on the execution_context branch to a function which was supposed to return an exception in the first place
<hkaiser> K-ballo: that caused the exception to be ignored, obviously
<heller> btw, we have a small QoI issue with our function implementation: https://pasteboard.co/IE3ZGJi.png
<K-ballo> what is it?
<heller> that it doesn't seem to get optimized as nicely as std::function
<K-ballo> needs clang's __builtin_operator_new to be able to do heap elision
<heller> no
<heller> just needs to inline the constructors and destructor again
<K-ballo> there's no allocation in that case
<heller> sure
<heller> but the compiler can't see into function_base::destroy, for example
<K-ballo> I may be able to look at it during the weekend
<K-ballo> maybe we can also stop emitting trivial functions, save us some bloat
<heller> yeah
<heller> I have a potential fix that leads at least to the same assembly for std::function and hpx::function
<heller> let's see if it has some real impact
<heller> no noticable difference for the future_overhead test
hkaiser has quit [Quit: bye]
aserio has joined #ste||ar
rori has joined #ste||ar
<rori> sorry for not being around, I don't know what phylanx is but will try to solve the problem, maybe changing some cmake stuff inside phylanx
<rori> sorry for the disturbments
hkaiser has joined #ste||ar
rtohid has joined #ste||ar
<simbergm> hkaiser: how did you end up linking to hpx in phylanx now
<simbergm> ?
<simbergm> is this the commit that is failing for you?
<simbergm> would hpx_setup_target work for you? we should still make target_link_libraries(... hpx) work but that wasn't expected to work yet
weilewei has joined #ste||ar
<weilewei> heller my email sender said the email sending to you: Delivery has failed to these recipients or groups: your_gmail_account
<heller> weilewei: ugh, ok, will have a look
<weilewei> Any other email is good?
<heller> maybe that's the reason why I didn't get any mails recently ;)
<weilewei> Or send me an email with your another email account, I can forward the email to you
aserio1 has joined #ste||ar
<heller> weilewei: I'll be home soon, I'll tell you once it is working again
weilewei has quit [Remote host closed the connection]
aserio has quit [Ping timeout: 250 seconds]
aserio1 has quit [Quit: aserio1]
rtohid has left #ste||ar ["Konversation terminated!"]
weilewei has joined #ste||ar
<hkaiser> simbergm: yah, that's what I'm using now target_link_libraries(... hpx)
aserio has joined #ste||ar
weilewei has quit [Remote host closed the connection]
<simbergm> hkaiser: if you need it fixed quickly use hpx_setup_target for now, if you can wait we'll get that working eventually
<simbergm> you shouldn't need to set that manually, it should come from the hpx target (eventually...)
<hkaiser> simbergm: sure, no worries
<heller> weilewei: my gmail stuff should work again
aserio has quit [Ping timeout: 276 seconds]
<rori> hkaiser: I'm trying to cmake phylanx, but I've still some troubles with pybind11, I cannot find pybind11Config.cmake nor pybind11-config.cmake in any of my source/build nor install pybind11 dirs, would you have an idea of how I can proceed ?
<hkaiser> rori: you need to at least configure pybind11
<hkaiser> rori: otoh, pls wait a bit, I think we know what's happening
<hkaiser> I'll get back to you if we get stuck again
<rori> configure how ? ^^ I just ran py3 setup.py build and py3 setup.py install with my respective build and install directories
<rori> yeah I can wait no pb
<hkaiser> rori: pybind11 is not part of python
weilewei has joined #ste||ar
<weilewei> heller I forward that email to you just now, let me know if you receive it or not, feel free to add comments if you have any. Thanks
<rori> yeah it's what I have but I just ran the setup.py
<rori> After cloning this repo
<rori> Ah ok understood ! sorry
<hkaiser> rori: I think it has its own cmake file you need to use
<rori> yep no worries thanks, I actually had another README that I followed..
<hkaiser> rori: btw, you should be able to use the hylanx docker base image
<hkaiser> that should give you what you need
<rori> aaaahh thanks !
<rori> And so what was exactly the pb ? I think I missed a part of the conv cause I was not online at that point
<zao> Some csv test doesn’t find real_main or something
<zao> Ah, there
<rori> Thanks !
<simbergm> btw, rori, don't remember if I showed you this: http://irclog.cct.lsu.edu/ste~b~~b~ar/2019-10-28
<simbergm> quite useful :)
<rori> Oh perfect !! Thanks !
<hkaiser> rori: the -W linker flags are missing when linking with hpx_wrap
<hkaiser> also, I'm not sure if the hpx_wrap target is pulled in if we do target_link_libraries(... hpx)
<hkaiser> well it should be pulled in whenever hpx_init is used as a dependency
<rori> yep I'm on it, just have to fix my compilation of phylanx cause I'm not making it to 6% ^^
<rori> no it's probably not but you can still do target_link_libraries to hpx_wrap
<rori> if I'm not mistaken
<rori> which is linking to hpx itself
<rori> I may have gotten the PUBLIC/PRIVATE flags wrong, I'll check that
<hkaiser> thanks
<rori> got*
<hkaiser> rori: will that add the -W linker flags?
<hkaiser> rori: also, I think hpx_init should depend on hpx_wrap
<rori> I have to check I'm not sure of where the linker flags are set I actually just transposed these directives, I come back to you as soon as I have a clue
<hkaiser> rori: thanks
<rori> yeah all link flags are INTERFACE so this should be transitive
<rori> hpx_init depend on hpx_wrap would make it depend on hpx, is it really what we want ?
aserio has joined #ste||ar
<hkaiser> rori: I think all is well now (it was a problem on our end), sorry for the noise (see: https://circleci.com/gh/STEllAR-GROUP/phylanx/32095)
<rori> no worried :)
<heller> weilewei: sent a reply
<rori> worries*
<weilewei> heller Thanks, just replied and forwarded it
<weilewei> heller the new 2019 tutorials look great, any video uploaded?
<heller> weilewei: we have been giving that for a while now. There is a video of the first time we gave it on youtube
<heller> we improved a bit since then, I think
<weilewei> I think I watched some part of old tutorials before, not all
<heller> there you go
<weilewei> yea, that's the one, I will check it out again
aserio1 has joined #ste||ar
aserio has quit [Ping timeout: 245 seconds]
aserio1 is now known as aserio
<weilewei> There is one physics student contacted me about stencil code, then I pointed him to our hpx code, he found out our tutorials are comparing hpx and openmp implementation. He found out that his visual studio has openmp, then he switched gear to openmp...
<zao> Helpful :)
aserio has quit [Ping timeout: 246 seconds]
rori has quit [Ping timeout: 276 seconds]
weilewei has quit [Remote host closed the connection]
bibek has quit [Quit: Konversation terminated!]
bibek has joined #ste||ar
hkaiser has quit [Read error: Connection reset by peer]
aserio has joined #ste||ar
weilewei has joined #ste||ar
weilewei has quit [Remote host closed the connection]
aserio has quit [Quit: aserio]
hkaiser has joined #ste||ar
coldblackice has joined #ste||ar
Coldblackice|2 has quit [Ping timeout: 240 seconds]
Coldblackice_ has joined #ste||ar
coldblackice has quit [Ping timeout: 276 seconds]
weilewei has joined #ste||ar
weilewei has quit [Remote host closed the connection]