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
GiannisProkopiou has quit [Remote host closed the connection]
diehlpk has joined #ste||ar
diehlpk has quit [Changing host]
diehlpk has joined #ste||ar
diehlpk has quit [Ping timeout: 240 seconds]
nikunj has quit [Ping timeout: 256 seconds]
nikunj has joined #ste||ar
hkaiser has quit [Quit: bye]
nikunj has quit [Ping timeout: 265 seconds]
nikunj has joined #ste||ar
jaafar has quit [Ping timeout: 265 seconds]
nikunj has quit [Ping timeout: 265 seconds]
nikunj has joined #ste||ar
<mdiers_> i think that confirms again #4306 and #4410
nikunj has quit [Ping timeout: 240 seconds]
nikunj has joined #ste||ar
nikunj has quit [Ping timeout: 256 seconds]
nikunj has joined #ste||ar
K-ballo has quit [Quit: K-ballo]
nikunj has quit [Ping timeout: 255 seconds]
nikunj has joined #ste||ar
K-ballo has joined #ste||ar
Abhishek09 has joined #ste||ar
<simbergm> mdiers_: thanks! that's very helpful
<simbergm> I haven't had time to look at things myself yet, but it would help if you comment on 4306 as well
<simbergm> we can try to get that in first and see what does to your results
Abhishek09 has quit [Ping timeout: 240 seconds]
gonidelis has joined #ste||ar
<simbergm> mdiers_: btw, we did cherry pick changes from 4306 into 1.4.1 (and the fixes are not in master), would you mind trying that?
<simbergm> mdiers_: yep
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
<mdiers_> simbergm: updated the table.
<simbergm> mdiers_: that looks better, except for the block_executor
<mdiers_> yeap, can you still a take look at my testimplementation of the block_executor? is there something wrong?
<simbergm> can you work with that? block_executor uses the attached executor so should in principle work the same assuming the indices are correct
<simbergm> the indices look like they should be correct (although a bit complicated)
<simbergm> mdiers_: you could also do just `block_executor<>(get_local_targets()[num_pus.first])`
<mdiers_> yeap, but i use only one target, that's a little untypical.
<simbergm> that should even be better because it schedules the threads on the correct worker thread directly
<simbergm> it looks like what you have there puts tasks on the first worker thread of each numa domain
<simbergm> *puts all the tasks
gonidelis has quit [Ping timeout: 240 seconds]
<mdiers_> simbergm: thanks have found the error: s/get_local_targets/numa_domains/g
K-ballo1 has joined #ste||ar
K-ballo has quit [Ping timeout: 258 seconds]
K-ballo1 is now known as K-ballo
<simbergm> mdiers_: great! so that means all three executors behave the same now?
<simbergm> on 1.4.1 that is
<mdiers_> simbergm: yeap on 1.4.1
<simbergm> very good, thanks for your patience with this
<simbergm> we'll try to get that fix into master soon as well, and make sure we don't break things again when we clean up the executors
<mdiers_> simbergm: updated the table: the block_executor is also correct in 1.3.
<mdiers_> simbergm: thanks a lot for your help
<mdiers_> simbergm: unfortunately the scheduler dead loop still exists when i take out the sleep
<simbergm> mdiers_: nothing stands out immediately as something that would cause a deadlock, but in general it's good to avoid calling `get` explicitly (simply return a future instead if possible)
<simbergm> in these kind of situations the debugger is not always that helpful
<simbergm> printf debugging is much more effective since you can see where your program is making progress
<simbergm> and where it's not
<simbergm> if you can identify where it's stuck I might be able to help you better
<mdiers_> simbergm: I'll look into it when I have some free time, thanks
hkaiser has joined #ste||ar
nikunj has quit [Ping timeout: 240 seconds]
nikunj has joined #ste||ar
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
gonidelis has joined #ste||ar
hkaiser has quit [Quit: bye]
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
nikunj has quit [Remote host closed the connection]
nikunj has joined #ste||ar
hkaiser has joined #ste||ar
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
gonidelis has quit [Remote host closed the connection]
parsa has quit [Quit: Free ZNC ~ Powered by LunarBNC: https://LunarBNC.net]
nikunj has quit [Ping timeout: 256 seconds]
parsa has joined #ste||ar
nikunj has joined #ste||ar
nikunj has quit [Ping timeout: 260 seconds]
iti has joined #ste||ar
<iti> Hello, I was trying to build hpx with https://github.com/STEllAR-GROUP/phylanx/wiki/Build-Instructions build instructions, and when I try to run cmake command I am getting error "Could NOT find TCMalloc" I did use the instruction -DCMAKE_C_COMPILER=gcc-7 DCMAKE_CXX_COMPILER=g++-7 but still getting it
<iti> any suggestions ?
diehlpk has joined #ste||ar
diehlpk has quit [Changing host]
diehlpk has joined #ste||ar
<simbergm> iti: (note that those are phylanx instructions, not hpx instructions)
<simbergm> you have two options: either install tcmalloc, or use the system allocator
<simbergm> tcmalloc is part of the gperftools package, it might have slightly different names depending on which distro you're on
<simbergm> if you want to skip it for now (not recommended for performance, but ok for playing around) you can say `-DHPX_WITH_MALLOC=system`
<iti> okay, thanks for informing. Can you let me know if there is any documentation for building hpx?
<simbergm> but specifically you need boost, hwloc, and preferably a custom allocator
<simbergm> we require a few compiled boost libraries, if the cmake configuration says it can't find them try searching for that package
jaafar has joined #ste||ar
<iti> I see! Thanks simbergm :D Will update here after building it
iti has quit [Ping timeout: 258 seconds]
diehlpk__ has joined #ste||ar
GiannisProkopiou has joined #ste||ar
GiannisProkopiou has quit [Remote host closed the connection]
diehlpk has quit [Ping timeout: 240 seconds]
<diehlpk_work> hkaiser, https://joss.theoj.org/
<diehlpk_work> We should consider to submit hpx there. 1) we will get few reviews of the code
<diehlpk_work> 2) It is a full journal article and we can cite hpx having a DOI which will show up in Google Scholar
<diehlpk_work> We just need to write a half page to one page sumamry, like here https://joss.readthedocs.io/en/latest/submitting.html#example-paper-and-bibliography
<hkaiser> diehlpk_work: sounds like a good idea
<diehlpk_work> I will add the paper and bibtex file to our repo
<diehlpk_work> I think best wat to proceed would be to send an email to hpx-devel and let everyone know that they should add their authors details until a certain deadline in alphabetical order
hkaiser has quit [Quit: bye]
hkaiser has joined #ste||ar
<diehlpk_work> hkaiser, I will start today to upload the workshop recordings to the ste||ar channel
<diehlpk_work> Should we just generate a new playlist
<hkaiser> diehlpk_work: yah, that might be a good idea
<hkaiser> let's create a new playlist
<diehlpk_work> Ok, sounds good
diehlpk__ has quit [Ping timeout: 265 seconds]
<bita_> hkaiser, the changes in HighFive to solve our issue are merged
zao has quit [Remote host closed the connection]
zao has joined #ste||ar
<hkaiser> bita_: thanks a lot for making this happen!
bita_ has quit [Read error: Connection reset by peer]