#ste||ar on 2017-11-06 — irc logs at irclog.cct.lsu.edu

2017-05-17 13:54 aserio 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/

00:00 <wash[m]> hkaiser: yep

00:00 <hkaiser> wash[m]: just arrived, any plans for dinner yet?

00:01 <wash[m]> I am at another thing until 7:00/7:30, I won't be back until then, but then I'll be going for dinner

00:01 <hkaiser> ok - until then I've starved to death ;)

00:01 <hkaiser> will get something myself

00:02 <wash[m]> K

00:02 <wash[m]> Some people are going out at 6 I think, ask in the ISO irc

00:05 <hkaiser> wash[m]: could you remind me the channel name, pls

00:06 <wash[m]> ##C++-meeting

00:06 <wash[m]> Also #wg21 on slack

00:06 <wash[m]> And #future_standard on slack

00:07 <hkaiser> ks

00:08 parsa has quit [Quit: Zzzzzzzzzzzz]

00:28 parsa has joined #ste||ar

00:36 hkaiser has quit [Ping timeout: 240 seconds]

01:11 gedaj has joined #ste||ar

01:21 EverYoun_ has quit [Remote host closed the connection]

01:31 hkaiser has joined #ste||ar

01:34 ct_clmsn_ has quit [Ping timeout: 246 seconds]

01:58 ct_clmsn_ has joined #ste||ar

02:25 jaafar_ has quit [Ping timeout: 252 seconds]

02:49 parsa has quit [Quit: Zzzzzzzzzzzz]

02:50 ct-clmsn_ has joined #ste||ar

03:09 ct-clmsn_ has quit [Quit: Leaving]

03:13 hkaiser has quit [Quit: bye]

03:22 quaz0r has quit [Ping timeout: 240 seconds]

03:22 EverYoung has joined #ste||ar

03:35 quaz0r has joined #ste||ar

03:45 <K-ballo> hkaiser: ping

04:03 gedaj_ has joined #ste||ar

04:04 gedaj has quit [Ping timeout: 248 seconds]

04:05 eschnett has quit [Quit: eschnett]

05:12 parsa has joined #ste||ar

05:35 jaafar_ has joined #ste||ar

05:44 parsa has quit [Quit: Zzzzzzzzzzzz]

05:44 parsa has joined #ste||ar

05:45 parsa has quit [Client Quit]

05:45 parsa has joined #ste||ar

05:45 parsa has quit [Client Quit]

06:31 K-ballo has quit [*.net *.split]

06:40 jaafar_ has quit [Ping timeout: 260 seconds]

07:12 K-ballo has joined #ste||ar

07:22 taeguk[m] has quit [Ping timeout: 248 seconds]

07:23 thundergroudon[m has quit [Ping timeout: 276 seconds]

08:46 david_pfander has joined #ste||ar

09:13 thundergroudon[m has joined #ste||ar

09:22 taeguk[m] has joined #ste||ar

09:37 thundergroudon[m has quit [Remote host closed the connection]

09:37 taeguk[m] has quit [Read error: Connection reset by peer]

09:41 EverYoung has quit [Remote host closed the connection]

09:41 EverYoung has joined #ste||ar

09:41 thundergroudon[m has joined #ste||ar

09:42 EverYoung has quit [Remote host closed the connection]

09:43 EverYoung has joined #ste||ar

09:51 taeguk[m] has joined #ste||ar

10:27 EverYoun_ has joined #ste||ar

10:27 EverYoun_ has quit [Remote host closed the connection]

10:27 EverYoun_ has joined #ste||ar

10:28 EverYoun_ has quit [Remote host closed the connection]

10:28 EverYoun_ has joined #ste||ar

10:29 EverYoun_ has quit [Remote host closed the connection]

10:29 EverYoun_ has joined #ste||ar

10:29 EverYoung has quit [Ping timeout: 258 seconds]

10:30 EverYoun_ has quit [Remote host closed the connection]

10:30 EverYoung has joined #ste||ar

10:30 EverYoung has quit [Remote host closed the connection]

10:31 EverYoung has joined #ste||ar

10:31 EverYoung has quit [Remote host closed the connection]

10:31 EverYoung has joined #ste||ar

10:32 EverYoung has quit [Remote host closed the connection]

10:32 EverYoung has joined #ste||ar

10:33 EverYoung has quit [Remote host closed the connection]

10:33 EverYoung has joined #ste||ar

10:34 EverYoung has quit [Remote host closed the connection]

10:55 jbjnr_ has joined #ste||ar

10:57 jbjnr has quit [Ping timeout: 246 seconds]

10:57 jbjnr_ is now known as jbjnr

11:03 <jbjnr> heller: do we have a tuple iteration mechanism for compile time extraction of hpx::util::tuple contents?

11:03 <heller> more or less

11:03 <jbjnr> PS. Good morning :)

11:03 <heller> what do you need it for?

11:04 <jbjnr> I've got a future of tuple<future<int>, future<double, ...> I need to extract the internal items

11:04 <jbjnr> (and do magic with them)

11:04 <heller> what's the usecase?

11:04 <jbjnr> intercepting when_all or dataflow and calling my numa_hint_function with the args

11:05 <heller> so you want to call numa_hint on all arguments?

11:06 <jbjnr> essentially yes, but the args are wrapped in hpx::lcos::future<hpx::util::tuple<hpx::lcos::future<int>, hpx::lcos::future<double> > >

11:06 <jbjnr> &etc etc

11:06 <heller> "fun"

11:06 <jbjnr> indeed

11:06 <heller> so it's always a future?

11:06 <jbjnr> after dataflow or when_all yes. I already have async and .then working

11:07 <heller> my initial feeling would be that the new unpack is the tool to go there

11:07 <heller> sorry, unwrap

11:07 <heller> give me a second

11:08 <heller> at least, the usecase sounds simila

11:08 <heller> r

11:11 <jbjnr> pack_traversal look interesting

11:13 <heller> something like this: https://gist.github.com/sithhell/a7d516eedbcb9f6b3bc288bfd91c76d2

11:14 <jbjnr> no, can't use that one. sorry. I'll play with pack_traversal after lunch. looks very interesting. mappers and stuff ...

11:14 <heller> yes, that should do the trick

11:14 <heller> traverse_pack should be the thing to use then

11:15 <heller> or well, map_pack

11:15 <heller> yeah

12:12 pree has joined #ste||ar

12:35 EverYoung has joined #ste||ar

12:37 ct_clmsn_ has quit [Ping timeout: 240 seconds]

13:26 hkaiser has joined #ste||ar

13:30 rod_t has joined #ste||ar

13:46 <pree> hkaiser, heller : Are you interested in having the Distributed GRAPH Data Structure in HPX. that is similar to " Graphx - distributed graph " on the top spark. Their way of implementing distributed graph is very much appealing. Currently HPX don't have Graph. So i think it will good to add it.

13:46 <pree> Thanks

13:46 <pree> pree

13:53 <heller> sure, why not

13:56 eschnett has joined #ste||ar

13:56 <pree> heller : Glad to know that, I will soon write some basic proposal for that,

13:57 <pree> Graph + graph algorithms

13:58 <zao> Clearly need Tensorflow integration, for all the ML/DL shenanigans!

13:58 <zao> (please don't)

14:00 <pree> zao : integrating tensorflow + hpx ??

14:01 <zao> pree: Mostly joking.

14:02 <hkaiser> pree: that has been done already (prrof of concept), no progress on this for a while however

14:02 <pree> zao : I think it's possible but requires some terrible work

14:03 <zao> Actual regular graphs would be more fun.

14:03 <pree> hkaiser : Could please elaborate ?

14:03 <pree> zao : lol

14:04 <pree> hkaiser : could you point me to the proof of concept please

14:04 <hkaiser> pree: https://github.com/ltroska/tensorflow, http://stellar.cct.lsu.edu/pubs/lukas_troska_hpx_tensorflow_04.05.17.pdf

14:05 <pree> hkaiser : Soory I misunderstand between GRAPHx and tensorflows : )

14:05 <pree> I actually mean graphx and you mean tensorflow

14:05 <pree> : )

14:05 <hkaiser> I do

14:06 <pree> shh

14:07 eschnett has quit [Quit: eschnett]

14:07 hkaiser has quit [Quit: bye]

14:41 hkaiser has joined #ste||ar

14:47 aserio has joined #ste||ar

14:58 eschnett has joined #ste||ar

15:04 aserio has quit [Read error: Connection reset by peer]

15:05 aserio has joined #ste||ar

15:06 <aserio> K-ballo: Morning

15:06 <aserio> K-ballo: Whom did you leave your key with?

15:07 <aserio> K-ballo: nvm found it!

15:10 daissgr has joined #ste||ar

15:23 <K-ballo> yep, right there

15:29 <jbjnr> heller: hpx::util::map_pack - totally awesome!

15:30 hkaiser has quit [Quit: bye]

15:41 daissgr has quit [Remote host closed the connection]

15:41 daissgr has joined #ste||ar

15:45 eschnett has quit [Read error: Connection reset by peer]

15:47 eschnett has joined #ste||ar

15:48 <heller> jbjnr: good

15:52 eschnett has quit [Ping timeout: 260 seconds]

16:03 eschnett has joined #ste||ar

16:05 <jbjnr> K-ballo: got a mo?

16:10 <jbjnr> nvm, fixed it

16:11 hkaiser has joined #ste||ar

16:16 patg[[w]] has joined #ste||ar

16:17 elfring has joined #ste||ar

16:41 <aserio> zbyerly: yt?

16:42 pree has quit [Ping timeout: 240 seconds]

16:44 EverYoung has quit [Remote host closed the connection]

16:44 EverYoung has joined #ste||ar

16:45 EverYoung has quit [Remote host closed the connection]

16:46 EverYoung has joined #ste||ar

16:49 hkaiser has quit [Quit: bye]

16:57 pree has joined #ste||ar

17:10 jakemp has joined #ste||ar

17:12 pree has quit [Quit: Bye dudes]

17:33 hkaiser has joined #ste||ar

17:35 eschnett has quit [Quit: eschnett]

17:38 eschnett has joined #ste||ar

17:44 eschnett has quit [Ping timeout: 240 seconds]

17:46 eschnett has joined #ste||ar

18:10 david_pfander has quit [Ping timeout: 255 seconds]

18:10 <wash> hkaiser, K-ballo: can std::mutex try_lock be called from an HPX-thread? It's non blocking per the standard, but does that mean it would not block an HPX thread?

18:11 jakemp has quit [Ping timeout: 268 seconds]

18:19 <jbjnr> wash if it's non-blocking, then it should be fine. (though I wouldn't use it on principle)

18:26 <hkaiser> wash: can be called from an hpx thread, shouldn't cause any problems

18:27 <hkaiser> wash: an hpx thread is just a std::thread with some additional strings attached

18:40 <github> [hpx] hkaiser created fixing_2991 (+1 new commit): https://git.io/vF8rR

18:40 <github> hpx/fixing_2991 0c40763 Hartmut Kaiser: Adding non-blocking (on destruction) service executors/future_then_async_executor.cpp...

18:47 <github> [hpx] hkaiser force-pushed fixing_2991 from 0c40763 to 9cdb22e: https://git.io/vF8oL

18:47 <github> hpx/fixing_2991 9cdb22e Hartmut Kaiser: Adding non-blocking (on destruction) service executors...

18:57 eschnett has quit [Quit: eschnett]

19:00 hkaiser has quit [Quit: bye]

19:11 jaafar_ has joined #ste||ar

19:44 diehlpk_work has joined #ste||ar

19:49 hkaiser has joined #ste||ar

19:49 patg[[w]] has quit [Quit: Leaving]

19:54 parsa has joined #ste||ar

19:55 parsa has quit [Client Quit]

20:14 thundergroudon[m has quit [Ping timeout: 240 seconds]

20:15 taeguk[m] has quit [Ping timeout: 276 seconds]

20:25 hkaiser has quit [Quit: bye]

20:28 elfring has quit [Quit: Konversation terminated!]

20:28 hkaiser has joined #ste||ar

20:42 eschnett has joined #ste||ar

21:37 <aserio> zbyerly: yt?

21:43 <hkaiser> heller: yt?

21:47 hkaiser has quit [Quit: bye]

21:57 hkaiser has joined #ste||ar

21:57 thundergroudon[m has joined #ste||ar

22:04 <heller> hkaiser: thanks, got it!

22:04 <hkaiser> heller: k

22:05 <hkaiser> doyou have an updated version, perhaps?

22:05 eschnett has quit [Quit: eschnett]

22:05 <heller> Unfortunately not :/

22:05 <hkaiser> k

22:05 <hkaiser> could you ate least do a spell check before I continue reading?

22:05 <heller> Haven't worked on it since August...

22:05 <heller> Sure, didn't mean to waste your time

22:05 taeguk[m] has joined #ste||ar

23:00 thundergroudon[m has quit [Read error: Connection reset by peer]

23:00 taeguk[m] has quit [Remote host closed the connection]

23:03 aserio has quit [Quit: aserio]

23:11 thundergroudon[m has joined #ste||ar

23:20 taeguk[m] has joined #ste||ar

23:35 taeguk[m] has quit [Remote host closed the connection]

23:35 thundergroudon[m has quit [Read error: Connection reset by peer]

23:36 hkaiser has quit [Quit: bye]

23:42 thundergroudon[m has joined #ste||ar

23:43 thundergroudon[m has quit [Remote host closed the connection]

23:53 thundergroudon[m has joined #ste||ar