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/ | GSoC2018: https://wp.me/p4pxJf-k1
jaafar has quit [Remote host closed the connection]
jaafar has joined #ste||ar
Viraj has joined #ste||ar
Viraj has quit [Client Quit]
mcopik has joined #ste||ar
mcopik has quit [Ping timeout: 264 seconds]
K-ballo has quit [Quit: K-ballo]
eschnett has joined #ste||ar
__gedaj__ has joined #ste||ar
__gedaj__ has quit [Client Quit]
__gedaj__ has joined #ste||ar
__gedaj__ has quit [Client Quit]
Anushi1998 has quit [Ping timeout: 260 seconds]
Anushi1998 has joined #ste||ar
nanashi55 has quit [Ping timeout: 260 seconds]
nanashi55 has joined #ste||ar
nanashi55 has quit [Remote host closed the connection]
nanashi55 has joined #ste||ar
Anushi1998 has quit [Read error: Connection reset by peer]
Anushi1998 has joined #ste||ar
diehlpk_mobile has joined #ste||ar
diehlpk_mobile has quit [Read error: Connection reset by peer]
Anushi1998 has quit [Ping timeout: 256 seconds]
Anushi1998 has joined #ste||ar
stmatengss has joined #ste||ar
stmatengss has quit [Read error: No route to host]
stmatengss1 has joined #ste||ar
Anushi1998 has quit [Ping timeout: 248 seconds]
Anushi1998 has joined #ste||ar
parsa has quit [Read error: Connection reset by peer]
Anushi1998 has quit [Ping timeout: 240 seconds]
Anushi1998 has joined #ste||ar
jaafar has quit [Ping timeout: 276 seconds]
stmatengss has joined #ste||ar
stmatengss1 has quit [Ping timeout: 240 seconds]
Anushi1998 has quit [Ping timeout: 276 seconds]
Anushi1998 has joined #ste||ar
stmatengss has quit [Ping timeout: 264 seconds]
stmatengss has joined #ste||ar
stmatengss1 has joined #ste||ar
stmatengss has quit [Ping timeout: 260 seconds]
Anushi1998 has quit [Remote host closed the connection]
<marco>
Hi, Is it possible to compile hpx with the intel compiler 2018?
<hkaiser>
marco: no idea
<hkaiser>
have you tried?
<marco>
No, not really. I have it compiled and installed with g++7. Now I want to compile my program with intel, and got some errors.
<hkaiser>
marco: I don't think there are any specific reasons for it not to compile, but intel has proven to be finicky about things, so it might not work
<marco>
Yes, I have already seen that the boost config for intel is specially adapted
mcopik has joined #ste||ar
<heller>
marco: what problems do you see?
K-ballo has joined #ste||ar
<marco>
heller: Lots of problems with the noexcept specifier, ...
<heller>
can you show an example error?
<hkaiser>
did we prematurely removed the NOEXCEPT macro?
<heller>
not sure...
<marco>
I added the BOOST_NO_CXX11_NOEXCEPT to the compiler/intel.hpp, and some errors gone away
<zao>
Build seems to mostly work for me.
<zao>
icc/2018.1.163-GCC-6.4.0-2.28
<zao>
(Boost 1.66.0)
<marco>
But there are still many variadic template errors.
<marco>
icpc version 18.0.3 (gcc version 7.3.1 compatibility)
<K-ballo>
more likely to be problems with 17's meaning of noexcept than lack of 11's noexcept
<marco>
zao: We have a lot of problems with the 2018 version, and I want to check if some have been fixed.
<marco>
Currently my problems look like this: /opt/rh/devtoolset-7/root/usr/include/c++/7/variant(1258): error: parameter pack "_Types" was referenced but not expanded
<heller>
that's a problem coming out of the standard librar
<heller>
y
<marco>
heller: Yes, but over hpx_init.hpp -> boost/program_options/options_description.hpp -> ...
<zao>
We haven't dared go for 7.x yet, for good reason :P
<zao>
heller: Any clue to my linker error?
<heller>
sure, should be reproducible with a minimal example just including variant
<zao>
Built with C++17, Boost built with heaven knows what standard, via modules.
* zao
jiggles off for meetings
<K-ballo>
what's the gcc personality for our icc 18?
<heller>
zao: yeah, the usual boost not being compiled in C++11 mode
<heller>
K-ballo: we don't test icc at the moment
<marco>
No problems with aocc 1.2 (clang 6) and gcc 7, only icpc do not want
<heller>
marco: could you a try an empty main just including the variant header?
<marco>
Yes, just a moment
<marco>
heller: Ok, I'm defeated ;-), It is the c++17 standard with the intel compiler 2018.
<zao>
marco: not the first time they are optimistic about the GCC they pretend to support.
elfring has joined #ste||ar
<marco>
zao: Ok, then I'll put our intel compilation to sleep for a few months for our new development.
diehlpk_work has joined #ste||ar
stmatengss has quit [Quit: Leaving.]
mbremer has joined #ste||ar
eschnett has quit [Quit: eschnett]
mbremer has quit [Ping timeout: 260 seconds]
david_pfander has quit [Ping timeout: 264 seconds]
bibek has joined #ste||ar
mbremer has joined #ste||ar
<heller>
marco: FWIW, we were seeing significant performance regressions with the intel compilers a year ago
eschnett has joined #ste||ar
RostamLog has joined #ste||ar
Anushi1998 has joined #ste||ar
jakub_golinowski has joined #ste||ar
stmatengss has joined #ste||ar
mbremer_ has joined #ste||ar
mbremer_ has quit [Client Quit]
Anushi1998 has quit [Remote host closed the connection]
galabc has joined #ste||ar
Anushi1998 has joined #ste||ar
jaafar has joined #ste||ar
Anushi1998 has quit [Ping timeout: 276 seconds]
galabc has quit [Quit: Leaving]
Anushi1998 has joined #ste||ar
Anushi1998 has quit [Read error: Connection reset by peer]
Anushi1998 has joined #ste||ar
galabc has joined #ste||ar
diehlpk has joined #ste||ar
<mcopik>
galabc: yt?
mcopik_mobile has joined #ste||ar
Anushi1998 has quit [Ping timeout: 240 seconds]
Anushi1998 has joined #ste||ar
AndroUser has joined #ste||ar
stmatengss has quit [Quit: Leaving.]
mcopik has quit [Ping timeout: 260 seconds]
mcopik_mobile has quit [Ping timeout: 256 seconds]
AndroUser has quit [Client Quit]
nikunj has joined #ste||ar
rtohid has joined #ste||ar
AndroUser2 has joined #ste||ar
quaz0r has quit [Ping timeout: 268 seconds]
<M-ms>
jakub_golinowski: yt?
<jakub_golinowski>
yes
<M-ms>
just wanted to add a small thing: remember that you can test if different chunking makes the hpx backend faster just by setting the nstripes parameter explicitly in the call to parallel_for (perhaps obvious... but saves you recompiling opencv to see if it makes a difference) and
<jakub_golinowski>
ah yes! Very nice and can be done quickly
<M-ms>
as a mini-benchmark you could do a sweep over different nstripes and you should end with a "bathtub" plot (a U-shaped plot)
<M-ms>
it's just a nice way to demonstrate that task granularity matters a lot for performance
<M-ms>
that's all, I don't need to see a graph or anything, just try it out for yourself to see the difference
<jakub_golinowski>
ok
quaz0r has joined #ste||ar
diehlpk has quit [Ping timeout: 265 seconds]
diehlpk has joined #ste||ar
Anushi1998 has quit [Read error: Connection reset by peer]
Anushi1998 has joined #ste||ar
anushi has joined #ste||ar
diehlpk has quit [Ping timeout: 276 seconds]
verganz has joined #ste||ar
AndroUser2 has quit [Read error: Connection reset by peer]
akheir has quit [Quit: Leaving]
diehlpk has joined #ste||ar
Anushi1998 has quit [Ping timeout: 256 seconds]
Anushi1998 has joined #ste||ar
quaz0r has quit [Ping timeout: 248 seconds]
mcopik has joined #ste||ar
rod_ has joined #ste||ar
diehlpk has quit [Ping timeout: 260 seconds]
rod_ has left #ste||ar [#ste||ar]
diehlpk has joined #ste||ar
quaz0r has joined #ste||ar
bibek has quit [Quit: Leaving]
diehlpk has quit [Remote host closed the connection]
diehlpk has joined #ste||ar
diehlpk has quit [Ping timeout: 256 seconds]
jakub_golinowski has quit [Quit: Ex-Chat]
nikunj has quit [Quit: Leaving]
bobakk3r has joined #ste||ar
bobakk3r has quit [Quit: Leaving.]
nikunj has joined #ste||ar
elfring has quit [Quit: Konversation terminated!]
bobakk3r has joined #ste||ar
diehlpk has joined #ste||ar
rtohid has quit [Ping timeout: 260 seconds]
bobakk3r has quit [Quit: Leaving.]
bobakk3r has joined #ste||ar
bobakk3r has quit [Client Quit]
diehlpk has quit [Ping timeout: 276 seconds]
diehlpk_work has quit [Quit: Leaving]
eschnett has quit [Quit: eschnett]
diehlpk_mobile has joined #ste||ar
Viraj has joined #ste||ar
Viraj has quit [Client Quit]
<nikunj>
I have another good update to give. I can now run main on HPX threads even for statically linked executable
<nikunj>
I'll update my github repo with an experimental code for normal code today. I might be able to make it work for HPX by the end of the day too. Else, it will be available by tomorrow.
<K-ballo>
\o/
rtohid has joined #ste||ar
rtohid has quit [Read error: Connection reset by peer]
mcopik_ has joined #ste||ar
diehlpk_mobile has quit [Read error: Connection reset by peer]