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]
Yorlik has joined #ste||ar
bita__ has quit [Ping timeout: 260 seconds]
Yorlik has quit [Ping timeout: 260 seconds]
nikunj97 has joined #ste||ar
kale[m] has quit [Ping timeout: 264 seconds]
Nikunj__ has joined #ste||ar
nikunj97 has quit [Ping timeout: 260 seconds]
hkaiser has joined #ste||ar
kale[m] has joined #ste||ar
<tiagofg[m]> Is there any incompatibility with HPX/AGAS and the package libevent?
<tiagofg[m]> By the way, that errors always occur when I include #include<hpx/hpx.hpp>
<hkaiser> tiagofg[m]: could be
<hkaiser> not sure, actually
<Nikunj__> hkaiser, whoa you're up early ;-)
<hkaiser> can't sleep...
<tiagofg[m]> humm, do you know some other library like libevent that work fine with HPX?
<hkaiser> tiagofg[m]: frankly, it's difficult to see how libevent could interact with hpx
<tiagofg[m]> here in portugal it is already 10:34AM :P
<hkaiser> what problems do you see?
<hkaiser> for me it's 4:34am
<tiagofg[m]> :O
<Nikunj__> since we're mentioning times, here it's 3:05pm :D
* tiagofg[m] posted a file: erros_placor.txt (28KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/TXEqOTwwXkDWSNWiWmFYtfiF >
<tiagofg[m]> that errors...
<hkaiser> hmmm, I've never seen this
<hkaiser> looks like some system header incompatibility
<tiagofg[m]> exactly
<zao> Does it manifest even if you don't include <hpx/hpx.hpp> but have the target configured for HPX?
<hkaiser> this kind of problem could be the result of some macro name clash causing symbols to be expanded to something nonsensical
<zao> Are you building with some odd _GNU_SOURCE or other defines?
<zao> Ah yes, macros, our eternal friends :D
<hkaiser> tiagofg[m]: may I suggest you create a preprocessed file?
<hkaiser> looking at the preprocessed output helps figuring out what's wrong
<tiagofg[m]> <zao "Does it manifest even if you don"> No, only when I include
<tiagofg[m]> The program I am working on it wasn't made by me, but I'm going to look for macros
<tiagofg[m]> <zao "Are you building with some odd _"> I don't think so
<zao> I recommend going with hkaiser's approach, grab the command line from your build output, change -c into -E to preprocess the translation unit to a file.
<zao> Inspecting that might give some insight into what goes wrong.
<tiagofg[m]> yeh, I'm going to do that, this is a tricky problem :D Thank you guys for helping
kale[m] has quit [Ping timeout: 264 seconds]
kale[m] has joined #ste||ar
<Nikunj__> hkaiser, yt?
<hkaiser> Nikunj__: here
<Nikunj__> hkaiser, see pm please
joe[m] has quit [Quit: killed]
rori has quit [Quit: killed]
oleg[m] has quit [Quit: killed]
jbjnr has quit [Quit: killed]
heller1 has quit [Quit: killed]
diehlpk_mobile[m has quit [Quit: killed]
tiagofg[m] has quit [Quit: killed]
kordejong has quit [Quit: killed]
parsa[m] has quit [Quit: killed]
otto[m] has quit [Quit: killed]
gdaiss[m] has quit [Quit: killed]
wladimir[m] has quit [Quit: killed]
oscar[m]1 has quit [Quit: killed]
norbert[m] has quit [Quit: killed]
klaus[m]1 has quit [Quit: killed]
mariella[m] has quit [Quit: killed]
rajpalsanyam1997 has quit [Quit: killed]
neill[m] has quit [Quit: killed]
richard[m]1 has quit [Quit: killed]
gretax[m] has quit [Quit: killed]
mdiers[m] has quit [Quit: killed]
carola[m] has quit [Quit: killed]
ms[m] has quit [Quit: killed]
noise[m] has quit [Quit: killed]
gonidelis[m] has quit [Quit: killed]
smith[m] has quit [Quit: killed]
ralph[m] has quit [Quit: killed]
karame_ has quit [Remote host closed the connection]
<Nikunj__> hkaiser, yt?
<hkaiser> Nikunj__: here now
<Nikunj__> hkaiser, see pm please (again)
nanmiao11 has joined #ste||ar
norbert[m] has joined #ste||ar
kordejong has joined #ste||ar
kale[m] has quit [Ping timeout: 240 seconds]
kale[m] has joined #ste||ar
oscar[m] has joined #ste||ar
ralph[m] has joined #ste||ar
neill[m] has joined #ste||ar
rori has joined #ste||ar
joe[m] has joined #ste||ar
oleg[m] has joined #ste||ar
otto[m] has joined #ste||ar
smith[m] has joined #ste||ar
carola[m] has joined #ste||ar
wladimir[m] has joined #ste||ar
richard[m]1 has joined #ste||ar
noise[m] has joined #ste||ar
gdaiss[m] has joined #ste||ar
diehlpk_mobile[m has joined #ste||ar
gretax[m] has joined #ste||ar
klaus[m]1 has joined #ste||ar
heller1 has joined #ste||ar
mariella[m] has joined #ste||ar
jbjnr has joined #ste||ar
rajpalsanyam1997 has joined #ste||ar
mdiers[m] has joined #ste||ar
ms[m] has joined #ste||ar
tiagofg[m] has joined #ste||ar
weilewei has quit [Remote host closed the connection]
Yorlik has joined #ste||ar
parsa[m] has joined #ste||ar
gonidelis[m] has joined #ste||ar
weilewei has joined #ste||ar
<diehlpk_work_> SC main track paper notification and SC panel notification is today!
<gonidelis[m]> hkaiser: I think the fix on reduce_adapt branch is fine now ;)
<hkaiser> gonidelis[m]: great!
<hkaiser> let's wait for the tests to cycle, I'll merge it once it went through
<gonidelis[m]> great !
karame_ has joined #ste||ar
<diehlpk_work_> hkaiser, jbjnr parsa ms[m] June 29, 2020 - July 3, 2020 - Mentors and students submit their evaluations of one another. These evaluations are a required step of the program.
<hkaiser> yes, will do
<diehlpk_work_> hkaiser, jbjnr parsa ms[m] Please do that by Wednesday or I will spam you with messages and emails :)
<hkaiser> ok
<gonidelis[m]> The evaluations open in like 3 hours
<diehlpk_work_> Nikunj__, you are included as well
<Nikunj__> diehlpk_work_, will do
<diehlpk_work_> gonidelis[m], Yes, please do it as soon as poosible
<gonidelis[m]> diehlpk_work_: I am refreshing the gsoc webpage from this morning like every 30 minutes ;p
<diehlpk_work_> gonidelis[m], No hurry. you just have to do it before the deadline to be able to work with us
<gonidelis[m]> diehlpk_work_: Yeah sure. Will do it today
<gonidelis[m]> Can I git pull this branch (https://github.com/STEllAR-GROUP/hpx/pull/4790) immediately to a local branch of mine? Or should I expect conflicts?
<hkaiser> gonidelis[m]: just pulling will not create conflicts
<hkaiser> you can get conflicts on merge or rebase only
rtohid has joined #ste||ar
<gonidelis[m]> right... So what if I merge?
<hkaiser> gonidelis[m]: : I don't think you'll see any conflicts
<hkaiser> conflicts happen only if the two involved branches have changes to the same code lines
<gonidelis[m]> ok great thanks!
<gonidelis[m]> Do we have any suspicion on where the warnings keep comming from? I would like to give it a try...
<gonidelis[m]> I am talking about #4790
<K-ballo> sometimes pulling does cause a merge, so try not to get in that situation
Nikunj__ has quit [Read error: Connection reset by peer]
<gonidelis[m]> hkaiser: Important question: While I know that I should change that https://github.com/gonidelis/hpx/blob/2104ee1d166cb4cc18257ccd074b46232b54c938/libs/algorithms/include/hpx/parallel/algorithms/for_each.hpp#L511 to `FwdIterB`, `FwdIterE` (and I know that's according to the TS), can't `FwdIterE` be a Sentinel? I mean the naming `FwdIterE` is a specific case of a Sentinel while there could be another cases ...
<gonidelis[m]> What am I missing :| ?
<gonidelis[m]> K-ballo: too
<K-ballo> I don't understand the question
<gonidelis[m]> In order to expose the range based interface we should change `typename FwIter` to `typename FwIterB , typename FwIterE`
<K-ballo> I don't approve of those names btw, this is like the third time they come up
<gonidelis[m]> But `FwIterE` doesn't have to be a forward iterator specificaly but it could be any kind of sentinel
<gonidelis[m]> K-ballo: So I reckon we agree on that
<K-ballo> indeed
Nikunj__ has joined #ste||ar
<gonidelis[m]> Do you know why `ForwardIterator first, ForwardIterator last` are not differentiated here http://eel.is/c++draft/alg.foreach?
nanmiao11 has quit [Remote host closed the connection]
<K-ballo> old algorithms only support iterator pairs, new ranges support sentinels
<gonidelis[m]> I thought this site was referred to the latest TS
nanmiao11 has joined #ste||ar
akheir has joined #ste||ar
<K-ballo> you probably don't mean TS, but even in the latest draft and standard the old algorithms still only support iterator pairs
<hkaiser> K-ballo: I think we should support Iter/Sentinel pairs for the C++17 algorithms as well
<hkaiser> otherwise we end up with more code duplication that actually needed
<gonidelis[m]> Aren't ranges proposed through a TS?
<hkaiser> ranges are part of c++10
<hkaiser> c++20
<K-ballo> ranges were proposed through a TS, the TS was eventually adopted in the draft for C++20
<K-ballo> hkaiser: we definitely should
<gonidelis[m]> Where could I find this draft?
<hkaiser> gonidelis[m]: http://eel.is/c++draft/
<K-ballo> heh
<gonidelis[m]> Thank you both
<gonidelis[m]> That's where I was lookking for
<gonidelis[m]> But I can't find our implementation there :/
<hkaiser> what implementations?
<gonidelis[m]> the `for_each(typename FwdIterB, typename FwdIterE)` one
<gonidelis[m]> I can only find `for_each(typename FwdIter)`
<hkaiser> gonidelis[m]: you will not find that, as K-ballo said those have not been changed to support Iter/Sentinel pairs
<K-ballo> ranges::for_each(I first, S last, Fun f, Proj proj = {}); ?
<gonidelis[m]> ok I get it.
<gonidelis[m]> So these guys don't provide a uniform implementation like ours which dispatches according to what FwIterB, FwIterE actually are during specialization... (?)
<K-ballo> uhm
<K-ballo> more like we don't provide the plain old algorithms, kinda?
<K-ballo> the reason we end up with both old algorithms and new range algorithms is mostly backwards compatibility, due to concept requirements mainly
nikunj97 has joined #ste||ar
<K-ballo> but at the same time, only the old algorithms are parallelized... so, we provide more of a hybrid
<gonidelis[m]> oh ok... But aren't plain old algorithms supported by our implementation ?
<gonidelis[m]> K-ballo:
<K-ballo> I have no idea what your mental picture for these things is
<K-ballo> the underlying "algorithms", the textbook meaning of algorithms, is the same for "old algorithms" and new "range algorithms"
Nikunj__ has quit [Ping timeout: 240 seconds]
<gonidelis[m]> K-ballo: ok great! thank you :)
<gonidelis[m]> When I `salloc -p marvin` on ROSTAM for example and then I run `srun make -j40 tests.unit` what's the difference if I ran `make -j40 tests.unit` without the `srun` ?
<hkaiser> gonidelis[m]: without the srun things would run on the headnode (except if you loged into your node explicitly
<hkaiser> gonidelis[m]: the best is to request an interactive allocation (using srun), and then to directly log in to the allocated node
noise[m] has quit [Quit: Idle for 30+ days]
<gonidelis[m]> "except if you loged into your node explicitly" isn't that what I do when I `salloc -p marvin` ?
<hkaiser> no
<hkaiser> salloc just allocated a node (or several)
<gonidelis[m]> oh ok. So I have just `srun -p medusa --pty /bin/bash -l`. and now my bash looks like this
<gonidelis[m]> `[giannis@medusa01 ~]$ `
<gonidelis[m]> So whatever I type here, runs on medusa, right?
bita_ has joined #ste||ar
kale[m] has quit [Ping timeout: 260 seconds]
kale[m] has joined #ste||ar
parsa[m] has quit [Quit: Idle for 30+ days]
<gonidelis[m]> What was that ? :p
<gonidelis[m]> K-ballo: Does this test check the case of predicate based sentinels? https://github.com/STEllAR-GROUP/hpx/blob/master/libs/iterator_support/tests/unit/is_sentinel_for.cpp
<gonidelis[m]> (I think not)
<K-ballo> gonidelis[m]: there's no real difference to Sentinel<int64_t>, it essentially predicates on equal_to(stop_value())
<gonidelis[m]> thanks! btw is there any goos source to read about sentinels?
<gonidelis[m]> good^^
<K-ballo> the original post from ranges would be a good start
<gonidelis[m]> you mean the TS?
<K-ballo> no, I mean eric's blog
<K-ballo> http://ericniebler.com/ anything you can find on ranges there would be valuable
<gonidelis[m]> ah ok got it
<K-ballo> ranges talk should start around 2013-2014
<gonidelis[m]> thank you!
<K-ballo> gonidelis[m]: DON'T read the TS
<gonidelis[m]> Great... I rember reading the blog during my preperation on April. I will give at a more detailed look
<gonidelis[m]> Could someone help me with this https://gist.github.com/gonidelis/b90424fa9f23ca9ffc3aa3a06c206cc4 ?
<gonidelis[m]> I am trying to compile this test
<K-ballo> looks like broken code.. that's not with master, is it?
<K-ballo> it's as if the algorithm internals confused the sentinel for the return value
<gonidelis[m]> yes of course that's not from master
<gonidelis[m]> It's the current for_each adaptation I am working in
<gonidelis[m]> Trying to make it accept for the end and begin iterators to be different
<gonidelis[m]> Before posting my for_each.hpp though, should this test compile fine?
<K-ballo> does it work for two iterators on master?
<nikunj97> hkaiser, yt?
<gonidelis[m]> yes of course. all these tests pass just fine: https://github.com/gonidelis/hpx/blob/master/libs/execution/tests/unit/foreach_tests.hpp
<hkaiser> nikunj97: here
<gonidelis[m]> The error seems to come from here
<nikunj97> hkaiser, a bit late. I have sent the email to Britney. You should've seen it as well.
<gonidelis[m]> `/home/giannis/foreach_adapt/libs/algorithms/include/hpx/parallel/util/detail/algorithm_result.hpp:36:38: error: could not convert ‘std::forward<Sentinel<long int> >((* & t))’ from ‘Sentinel<long int>’ to ‘hpx::parallel::util::detail::algorithm_result_impl<hpx::parallel::execution::sequenced_policy, Iterator<long int> >::type’ {aka ‘Iterator<long int>’}
<gonidelis[m]> `
<hkaiser> nikunj97: ok
<gonidelis[m]> Is `std::forward` the problem?
<hkaiser> no
<hkaiser> the algorithm_result_impl<..., Iterator> should be algorithm_result_impl<..., Sentinel<>>
<gonidelis[m]> ahhh
<gonidelis[m]> ?
<gonidelis[m]> and not `FwdIterB`
<gonidelis[m]> hkaiser:
sayefsakin has joined #ste||ar
bita_ has quit [Read error: Connection reset by peer]
bita_ has joined #ste||ar
nikunj97 has quit [Read error: Connection reset by peer]
nikunj97 has joined #ste||ar
kale[m] has quit [Ping timeout: 260 seconds]
kale[m] has joined #ste||ar
nikunj97 has quit [Ping timeout: 260 seconds]
<hkaiser> gonidelis[m]: yes
nikunj has quit [Remote host closed the connection]
nikunj has joined #ste||ar