#ste||ar on 2017-10-24 — 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/

00:08 EverYoung has joined #ste||ar

00:10 EverYou__ has quit [Ping timeout: 252 seconds]

00:11 <K-ballo> what the, what from?

00:17 EverYoung has quit [Ping timeout: 252 seconds]

00:31 EverYoung has joined #ste||ar

00:34 gedaj_ has quit [Remote host closed the connection]

00:34 gedaj_ has joined #ste||ar

00:55 EverYoung has quit [Ping timeout: 258 seconds]

01:10 EverYoung has joined #ste||ar

01:14 EverYoung has quit [Ping timeout: 248 seconds]

01:18 bobakk3r has quit [Remote host closed the connection]

02:07 Guest17093 has joined #ste||ar

02:19 K-ballo has quit [Quit: K-ballo]

02:27 parsa has quit [Quit: Zzzzzzzzzzzz]

03:03 Guest17093 has quit [Quit: See you later]

03:05 jaafar_ has joined #ste||ar

03:09 jaafar has quit [Ping timeout: 252 seconds]

03:20 hkaiser has quit [Quit: bye]

04:11 EverYoung has joined #ste||ar

04:16 EverYoung has quit [Ping timeout: 240 seconds]

04:19 zbyerly_ has joined #ste||ar

04:44 gedaj_ has quit [Remote host closed the connection]

04:44 gedaj_ has joined #ste||ar

05:11 jaafar_ has quit [Ping timeout: 255 seconds]

06:31 wash has joined #ste||ar

07:25 parsa has joined #ste||ar

07:27 parsa has quit [Client Quit]

07:38 david_pfander has joined #ste||ar

09:07 <github> [hpx] StellarBot pushed 1 new commit to gh-pages: https://git.io/vdhsX

09:07 <github> hpx/gh-pages f5be8cc StellarBot: Updating docs

10:30 K-ballo has joined #ste||ar

10:54 zbyerly_ has quit [Ping timeout: 255 seconds]

10:57 <zao> K-ballo: The output of each test run for this commit is 800K of CTest Test.xml

10:58 <zao> In total 294 test runs, each with the usual 554 tests or something, for a total of 116k individual tests.

11:02 <zao> Some legitimate test outputs are up towards 34 KiB, with a steady spread down to a kilobyte or lower.

11:35 hkaiser has joined #ste||ar

13:01 <zao> Ah, it's the gosh-darned spam from tests.unit.parallel.reduce_by_key

13:05 <zao> https://zao.se/~zao/hpx/reduce_by_key.txt

13:05 <zao> I question the usefulness of such test output.

13:05 <github> [hpx] K-ballo force-pushed dataflow-api from 306ec04 to 1739ace: https://git.io/vdbHQ

13:05 <github> hpx/dataflow-api 1de303b Agustin K-ballo Berge: Cleaning up dataflow overload set

13:05 <github> hpx/dataflow-api 170efa9 Agustin K-ballo Berge: Consolidate local and remote dataflow implementations

13:05 <github> hpx/dataflow-api e75d775 Agustin K-ballo Berge: Simplify dataflow implementation

13:06 <K-ballo> zao: cut it

13:07 eschnett has quit [Quit: eschnett]

13:09 <zao> Not overly keen on truncating output in the case where there may be actually useful output. Maybe drop the output for Passed tests.

13:09 <zao> Ah, this one is indeed a passed test.

13:10 <zao> Your typical crash output from something like tests.regressions.lcos_dir.wait_for_1751 is 30 KiB too, but those have a bit more value.

13:11 <zao> https://zao.se/~zao/hpx/wait_for_1751.txt

13:12 <K-ballo> any chance you could capture stderr but drop stdout?

13:12 <K-ballo> and, do we respect that? :P

13:13 parsa has joined #ste||ar

13:14 <zao> I have no idea what's being done, I just consume what CTest spits out.

13:14 <zao> As far as I can tell, the "value" of a test is a single stream.

13:15 <zao> Whatever would be done in the test runner would assumedly propagate here.

13:23 eschnett has joined #ste||ar

13:34 hkaiser has quit [Quit: bye]

13:45 aserio has joined #ste||ar

13:53 eschnett has quit [Quit: eschnett]

13:59 parsa has quit [Quit: Zzzzzzzzzzzz]

13:59 parsa has joined #ste||ar

14:05 aserio1 has joined #ste||ar

14:05 eschnett has joined #ste||ar

14:08 aserio has quit [Ping timeout: 255 seconds]

14:08 aserio1 is now known as aserio

14:45 gedaj_ has quit [Remote host closed the connection]

14:45 gedaj_ has joined #ste||ar

14:55 aserio1 has joined #ste||ar

14:56 hkaiser has joined #ste||ar

14:57 hkaiser_ has joined #ste||ar

14:58 aserio has quit [Ping timeout: 252 seconds]

14:58 aserio1 is now known as aserio

15:00 hkaiser has quit [Ping timeout: 240 seconds]

15:11 EverYoung has joined #ste||ar

15:15 EverYoung has quit [Ping timeout: 240 seconds]

15:22 eschnett has quit [Quit: eschnett]

15:29 aserio has quit [Ping timeout: 246 seconds]

15:31 rod_t has joined #ste||ar

15:36 <K-ballo> uh oh, I get runtime errors using the latest hwloc

15:39 <heller> K-ballo: 2.x?

15:39 <K-ballo> no, 1.11.8

15:40 <K-ballo> hwloc_topology_init returns -1, doing a clean full rebuild just in case

15:40 aserio has joined #ste||ar

15:55 EverYoung has joined #ste||ar

15:55 EverYoung has quit [Remote host closed the connection]

15:56 EverYoung has joined #ste||ar

15:57 <K-ballo> was a bad build issue on my side, I had switched back to binaries from before the hwloc switch

16:10 pree has joined #ste||ar

16:11 pree has quit [Read error: Connection reset by peer]

16:16 pree has joined #ste||ar

16:19 pree_ has joined #ste||ar

16:20 <pree_> Hi ! Is there any performance studies has been done between HPX && apache spark ?

16:21 pree has quit [Ping timeout: 240 seconds]

16:21 pree_ is now known as pree

16:30 pree has quit [Read error: Connection reset by peer]

16:35 david_pfander has quit [Ping timeout: 248 seconds]

16:37 aserio has quit [Ping timeout: 252 seconds]

16:39 parsa has quit [Quit: Zzzzzzzzzzzz]

16:43 pree has joined #ste||ar

16:55 jaafar_ has joined #ste||ar

16:58 parsa has joined #ste||ar

16:59 parsa has quit [Client Quit]

17:01 eschnett has joined #ste||ar

17:03 EverYoun_ has joined #ste||ar

17:05 EverYoung has quit [Ping timeout: 246 seconds]

17:45 pree has quit [Ping timeout: 240 seconds]

17:50 aserio has joined #ste||ar

17:53 eschnett has quit [Read error: Connection reset by peer]

17:54 eschnett has joined #ste||ar

17:59 pree has joined #ste||ar

18:06 pree has quit [Ping timeout: 240 seconds]

18:09 <github> [hpx] K-ballo pushed 1 new commit to dataflow-api: https://git.io/vdjWg

18:09 <github> hpx/dataflow-api 297a5a2 Agustin K-ballo Berge: Fix deprecated includes of dataflow

18:13 aserio has quit [Read error: Connection reset by peer]

18:14 aserio has joined #ste||ar

18:18 pree has joined #ste||ar

18:20 pree has quit [Client Quit]

18:21 pree has joined #ste||ar

18:22 rod_t has left #ste||ar [#ste||ar]

18:33 pree has quit [Ping timeout: 252 seconds]

18:46 pree has joined #ste||ar

18:55 pree has quit [Ping timeout: 246 seconds]

18:59 aserio1 has joined #ste||ar

19:01 aserio has quit [Ping timeout: 258 seconds]

19:01 aserio1 is now known as aserio

19:06 jaafar_ has quit [Remote host closed the connection]

19:07 jaafar has joined #ste||ar

19:18 eschnett has quit [Quit: eschnett]

19:28 eschnett has joined #ste||ar

19:57 EverYoung has joined #ste||ar

20:00 EverYoun_ has quit [Ping timeout: 252 seconds]

20:02 rod_t has joined #ste||ar

20:09 patg[w] has joined #ste||ar

20:17 zbyerly_ has joined #ste||ar

20:31 hkaiser_ has quit [Ping timeout: 240 seconds]

20:32 patg[w] has quit [Quit: Leaving]

20:35 aserio has quit [Quit: aserio]

20:46 zbyerly_ has quit [Quit: Leaving]

20:54 eschnett has quit [Quit: eschnett]

20:57 hkaiser has joined #ste||ar

21:47 EverYoung has quit [Remote host closed the connection]

21:48 EverYoung has joined #ste||ar

21:54 <jbjnr> hkaiser: yt?

22:05 <jbjnr> gtg - here's my question if you see it : yesterday, you said that line 458 woukd be called, but I'm actually seeing this line being triggered https://github.com/STEllAR-GROUP/hpx/blob/master/hpx/parallel/executors/execution.hpp#L472 this then calls into the wrap_int version just before at line 430 - what am I missing to get the correct overload triggered - it prefers wrap_int to just int for...

22:05 <jbjnr> ...the zero.

22:09 <K-ballo> that means the int overload isn't viable

22:09 <K-ballo> (or should mean, modulo bugs in the implementation)

22:10 <K-ballo> jbjnr: you'd be missing a "then_execute"

22:10 <jbjnr> no, I've got that

22:11 <K-ballo> or the then_execute call is not otherwise viable

22:11 <K-ballo> show me your declaration

22:11 <jbjnr> hmm. ok. I'll check

22:14 <jbjnr> static_assert( has_then_execute_member<hpx::threads::executors::guided_pool_executor<hint_type2>>::value == std::true_type(),

22:14 <jbjnr> "has_then_execute_member<executor>::value");

22:14 <jbjnr> is ok

22:14 <jbjnr> is that good enough?

22:14 <K-ballo> where does has_then_execute_member come from?

22:15 <jbjnr> some traits macro in execution:: namesapce

22:15 <K-ballo> it's not enough for it to have a then_execute member, it has to be able to take the arguments too

22:15 <K-ballo> ok, let me see if I can find the definition

22:16 <jbjnr> I must go, so I'll play again tomorrow. thanks so far.

22:16 <K-ballo> no, it's not enough, it only checks for a member named then_execute

22:17 <jbjnr> ((my then_Execute is the same as all the others ...)

22:17 <K-ballo> where can I find the declaration?

22:29 gedaj_ has quit [Remote host closed the connection]

22:31 gedaj has joined #ste||ar

22:51 <github> [hpx] K-ballo deleted fix-spreading at 21d08b6: https://git.io/vdjNC

22:54 <K-ballo> appveyor running out of disk space :)

23:04 <hkaiser> \o/

23:12 EverYoun_ has joined #ste||ar

23:15 EverYoung has quit [Ping timeout: 258 seconds]

23:16 EverYoun_ has quit [Ping timeout: 246 seconds]

23:20 gedaj has quit [Remote host closed the connection]

23:20 gedaj has joined #ste||ar

23:25 <zao> Fun fact, the apparently static BuildStamp="20171018-0005-Experimental" in CTest XML, it changes daily on the same compile :)

23:25 <zao> (on multiple test runs of the same build, that is)

23:52 zao has quit [Ping timeout: 260 seconds]

23:53 zao has joined #ste||ar

23:53 EverYoung has joined #ste||ar

23:53 EverYoung has quit [Remote host closed the connection]

23:53 EverYoung has joined #ste||ar