K-ballo 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/
sivoais has quit [*.net *.split]
zao has quit [*.net *.split]
zao_ is now known as zao
sivoais_ has quit [Quit: leaving]
sivoais has joined #ste||ar
parsa| has joined #ste||ar
parsa has quit [Ping timeout: 245 seconds]
parsa| is now known as parsa
bering[m] has quit [Ping timeout: 245 seconds]
mdiers[m] has quit [Ping timeout: 245 seconds]
gnikunj[m] has quit [Ping timeout: 245 seconds]
mdiers[m] has joined #ste||ar
bering[m] has joined #ste||ar
gnikunj[m] has joined #ste||ar
chuanqiu has joined #ste||ar
chuanqiu has quit [Quit: Connection closed]
chuanqiu has joined #ste||ar
chuanqiu has quit [Client Quit]
hkaiser has quit [Quit: bye]
diehlpk_work has quit [Remote host closed the connection]
<Deepak1411[m]> hkaiser: zao I am trying to update VS19 and trying builld procedures once again, I'll let you know by evening (Timezone: IST) .
<Deepak1411[m]> hkaiser: I am getting a new problem while cloning the repo
<srinivasyadav227> Deepak1411: did you try "Retry clone"?
<Deepak1411[m]> srinivasyadav227: Obviously I tried 5-6 times, everytime same error message
<Deepak1411[m]> thanks I'll look into this :)
<zao> It’d still be interesting to see the actual errors to be able to figure out if we have broken older compilers somehow, but I guess it may be a bit too late now.
<zao> Your internet connection sounds like a fun ride :)
<zao> The HPX repo is a bit hefty :D
gdaiss[m]1 has joined #ste||ar
gdaiss[m] has quit [Ping timeout: 245 seconds]
<tarzeau_> is there a reason why hpx builds dynamically linked AND statically linked libs? is there a preferation which ones people should use?
<zao> Note that those are not the same library, the shared one is the main HPX library, the others are support infrastructure that probably need to be static in order to be properly linked to whatever they're supposed to be used in.
<zao> The ones mentioned seem to be categorized under HPXInternal:: in CMake, but it feels like lintian is way too picky :P
<zao> (none of this actual feedback from someone in the HPX group, just musings)
<tarzeau_> that is already helpful information, thank you
<zao> I had kind of expected things like hpx_init.a to appear there, but maybe they're hiding in a corner somewhere.
<zao> Or I guess they're sufficiently stripped to please lintian.
<ms[m]> tarzeau_: zao is on point, we're moving the static libs into object libraries which eventually should never end up in the install, they're currently there only to satisfy cmake
<ms[m]> the static libs are also not meant to be used, they just get put into the shared libs
<zao> I noticed that they appear in the CMake targets in the installation but not in the pkg-config, which hints that they're probably a bit bonus :)
<tarzeau_> ms[m]: i see. so i'll wait for the announced next released version
<ms[m]> tarzeau_: they'll still be there in the next version, but the one after that might already be free of the static libs
<Deepak1411[m]> zao: Yes still facing the same errors, tried many approaches shown in that bitbucket link but, I'm going to keep try. feel free to give any suggestionssrinivasyadav227 hkaiser
<Deepak1411[m]> * zao: Yes still facing the same errors, tried many approaches shown in that bitbucket link but, I'm going to keep trying. feel free to give any suggestionssrinivasyadav227 hkaiser
<zao> Least horrible I can think of is to either try to use a proxy, or download a source tarball from GitHub.
<Deepak1411[m]> zao: fine I'll try that
<Deepak1411[m]> zao: it takes time for me to reply sorry about that, as my college exams are going
<zao> You could also try cloning over SSH instead of HTTPS.
<zao> That is, "git@github.com:stellar-group/hpx.git" instead of "https://github.com/stellar-group/hpx.git" as the remote.
<zao> I kind of assumed that you had something like that going on, given the reply rate :D
<zao> Good luck with those ^_^
hkaiser has joined #ste||ar
<Deepak1411[m]> <zao "That is, "git@github.com:stellar"> thanks
chuanqiu has joined #ste||ar
<hkaiser> mdiers[m]: yes, the branch had another unrelated problem
<hkaiser> could you please try with #5117?
<mdiers[m]> hkaiser: Yes, but I don't think I can do it today.
<hkaiser> sure, np, take your time
hkaiser has quit [Quit: bye]
hkaiser has joined #ste||ar
diehlpk_work has joined #ste||ar
<mdiers[m]> hkaiser: Thank you very much. My problem #5117 is now solved with the changes.
diehlpk_work has quit [Changing host]
diehlpk_work has joined #ste||ar
<hkaiser> mdiers[m]: \o/
<hkaiser> sorry it took so long to fix it
<mdiers[m]> hkaiser: No problem. Is there already a plan when there is the 1.7 release?
<hkaiser> mdiers[m]: somewhere during the next 2-3 weeks
<mdiers[m]> hkaiser: Thank you. I will then probably take over the 1.7 release with us.
<hkaiser> nice
<gonidelis[m]> hkaiser: you have a mail ;)
<ms[m]> hkaiser: neat, good call on moving that "users and applications" issue to a discussion; didn't know you could do that!
nanmiao has quit [Quit: Connection closed]
<Deepak1411[m]> zao: cloning the repo is done with ssh key and works fine. but, a build problem appeared during cmake build
<Deepak1411[m]> * zao: cloning the repo is done with ssh key and works fine.
nanmiao has joined #ste||ar
hkaiser has quit [Quit: bye]
chuanqiu has quit [Quit: Connection closed]
hkaiser has joined #ste||ar
<zao> Deepak1411[m]: It's again probably using HTTP(S) to clone Asio. Easiest thing there is probably to separately download ASIO and set -DASIO_ROOT on the CMake command line, like how you may specify -DBOOST_ROOT and -DHWLOC_ROOT already.
<zao> *Asio, sorry, ASIO is something else ;)
<hkaiser> zao: but why should that fail?
<zao> I assume either broken/paranoid security software on the computer, rickety ISP, or broken router.
<hkaiser> great :/
<zao> I'm curious if the build failure cited earlier was with a HPX release or master.
<zao> Not sure if I can spin up a VM with older VS versions, seems to be some available in https://docs.microsoft.com/en-us/visualstudio/releases/2019/history#release-dates-and-build-numbers
<hkaiser> nod, I'm using 16.11 preview
<zao> Install of VS Build Tools worked, got 16.9.5 in a VM now.
<zao> hkaiser: Current master builds for me with 64-bit VS 16.9.5, Boost 1.76, hwloc 2.4.1.
<zao> Including the transpose_smp example that I'm using as a canary.
<zao> Gonna try HPX/1.6.0 for good measure.
<zao> hkaiser: Good news, 1.6.0 fails in a manner similar to Deepak1411[m]'s build output.
<hkaiser> grrr
<hkaiser> thanks for diagnosing this, I'll have a look
<zao> Managed to exfiltrate one of the build outputs: https://gist.github.com/zao/460be4355a485d848e4459265468cf1e
<hkaiser> zao: but it fails for older VS versions only, does it?
<zao> Haven't tested yet.
<hkaiser> I meant, does it compile with recent VS?
<hkaiser> I can try myself, no worries - just trying to understand if this is a comiler issue
<hkaiser> or if it's a problem in our code
<zao> I have only tried 1.6.0 with 16.9.5, so no info yet.
<hkaiser> ok
<hkaiser> thanks
<hkaiser> zaobut you tried master, no?
<hkaiser> zao^^
<zao> Correct, master with 16.9.5 built properly.
<hkaiser> ok
<zao> I have not tried 1.6.0 with 19.10.0
<hkaiser> I can do that
<zao> hkaiser: Had one queued up, it fails with that combination as well so the compiler version was not to blame.
<zao> In summary, 1.6.0 fails on both 16.9.5 and 16.10.0, master succeeds on both.
nanmiao has quit [Quit: Connection closed]
jpinto[m] has quit [Ping timeout: 245 seconds]
jpinto[m] has joined #ste||ar