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/
EverYoung has joined #ste||ar
EverYoun_ has quit [Ping timeout: 255 seconds]
EverYoung has quit [Ping timeout: 276 seconds]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
vamatya has quit [Ping timeout: 260 seconds]
EverYoung has quit [Read error: Connection reset by peer]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Ping timeout: 248 seconds]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Ping timeout: 260 seconds]
hkaiser has quit [Quit: bye]
mcopik has joined #ste||ar
anushi has quit [Remote host closed the connection]
anushi has joined #ste||ar
hkaiser has joined #ste||ar
daissgr has quit [Ping timeout: 248 seconds]
vamatya has joined #ste||ar
mcopik has quit [Ping timeout: 260 seconds]
EverYoung has joined #ste||ar
Nikunj has joined #ste||ar
EverYoung has quit [Ping timeout: 276 seconds]
nanashi55 has quit [Ping timeout: 268 seconds]
nanashi55 has joined #ste||ar
hkaiser has quit [Quit: bye]
Nikunj has quit [Ping timeout: 260 seconds]
CaptainRubik_ has joined #ste||ar
CaptainRubik_ has quit [Ping timeout: 260 seconds]
<jbjnr_>
nikunj: I had a quick look at your cmakelists last night - what is it you are trying to do?
<jbjnr_>
I didn't understand why you are calling scripts - and not just doing a normal build - so I must not understand the purpose
<Nikunj>
@jbjnr: @hkaiser asked me to actually build the file by executing same commands that hello.sh uses
<Nikunj>
So I decided to simply execute the hello.sh file from examples/hello_world_component
<Nikunj>
This way any linux user can run cmakelists to build the hello_world_component and the CI can also build it the same way
<Nikunj>
@jbjnr_: If you feel that this is not the right way, then how should I build the hello_world_component?
david_pfander has joined #ste||ar
<jbjnr_>
why do you need a script to do it? is my question. look at all the other examples - they just use cmake commands to build libs/componsnts. I just wonder why you are using a script
<simbergm>
jbjnr_: I think the purpose of the example was to show how to build a component without cmake
<simbergm>
so the goal is
<simbergm>
1. make sure that the component builds at all
<simbergm>
2. make sure that the script is able to build the component
<simbergm>
(I think)
<simbergm>
but I've been confused before, so...
<Nikunj>
@jbnjr_: @hkaiser asked me to build it directly using cmake but to mimic commands given in hello.sh
<Nikunj>
*not to build it directly using cmake ^^
<Nikunj>
@jbjnr_: I initially used cmake to build the file like other examples to which @hkaiser asked me to not build it that way. He told me that the goal is to build like hello.sh is building
<Nikunj>
That is why I resorted to using simple scripts to simply run hello.sh instead of letting cmake build the file.
<Nikunj>
@jbjnr_: what simbergm is explaining is what I've been told to do.
<jbjnr_>
ok if you want to use a script - then check the dirs you are pointint to - you use '../build/' but this seems to imply some kind of hardcoded location - my buid and src trees are not near eah other, so check this is correct
<jbjnr_>
you should use a variable from cmake like ${CMAKE_ binary/build/etc } var
<Nikunj>
I want cmake to point to the actual image directory. The one which we use to build. Btw, I ../build/ that I used yesterday was by-mistake, I didn't realise that ../build/ contains the build elements and not the image itself.
<Nikunj>
So I changed the scripts to something like this: ../examples/hello_world_component/hello.sh. If I'm correct the circleci creates a directory of the repo of hpx then builds hpx into it.
<Nikunj>
So what I implemented was to create another directory into repo directory by the name of test and invoked cmake from there itself by using ../examples/hello_world_component/
<Nikunj>
It seemed to work fine in circleci since it was able to find the directory
<Nikunj>
However the make -j2 could not find hello.sh which either means that I'm pointing to wrong directory(which is confusing since cmake did work) or hello.sh actually does not exists.
simbergm has quit [Ping timeout: 240 seconds]
simbergm has joined #ste||ar
<jbjnr_>
if hello doesn't exist then you need to point to the src/ dir - unless you copy it to the build dir
<Nikunj>
@jbjnr: I didn't get you. I just want to run the script hello.sh which I'm first making it executable then simply running it.
<Nikunj>
How will pointing to src/dir will correct the situation?
<jbjnr_>
yes bt that script is in the source tree of hpx, not the build one - unless you copied it using cmake copy(...)
<Nikunj>
The above script does run the cmake command but fails to execute make saying that hello.sh does not exists in the desired folder
<jbjnr_>
so you need to check what the 'desired folder' it thinks it is using is the one that you think it is using and make sure you get it right! (captain obvious speaking)
<github>
hpx/release 7f2ed88 Mikael Simberg: Small change to release_procedure.rst
<Nikunj>
simbergm: I have a slight confusion. If we want to try building hello_world_component through hello.sh why don't we directly run the hello.sh file from circleci?
<Nikunj>
Why do we need to create another cmakelists to mimic the same?
<simbergm>
nikunj: I guess there are many ways it could be done
<simbergm>
the reason to add a CMakeLists.txt would be that one can integrate it so that when you run make examples the hello_world example would also be built
<simbergm>
and no special case is needed for circleci
<simbergm>
I think everyone will have their own idea of the best way to do it
<Nikunj>
simbergm: Hold on, wasn't the hello_world_component supposed to be command line only example?
<simbergm>
yes
<Nikunj>
I don't think that the idea was to integrate it in make examples since @hkaiser specifically asked it to remain command line
<simbergm>
but at least in my mind having a CMakeLists.txt doesn't mean that it has to do it the "normal" cmake way, it's also possible to run scripts such as hello.sh using cmake
<Nikunj>
or was it @k-ballo, I can't recall correctly :-(
<simbergm>
yeah, might be that they have some other idea
<Nikunj>
simbergm: Yeah
<simbergm>
it's probably best if you discuss it with them, I'm just confusing everyone even more
<Nikunj>
Creating a cmakelists has been a really easy task. Integrating it with circle is quite tricky though
<Nikunj>
Every time I send a pull request perfectly working fine on my PC tends to fail in circle..
Zwei_ has joined #ste||ar
mcopik has joined #ste||ar
Zwei has quit [Ping timeout: 264 seconds]
Nikunj has quit [Ping timeout: 260 seconds]
hkaiser has joined #ste||ar
<hkaiser>
heller_: yt?
Nikunj has joined #ste||ar
auviga has quit [Ping timeout: 240 seconds]
auviga has joined #ste||ar
anushi has quit [Read error: Connection reset by peer]
anushi has joined #ste||ar
CaptainRubik has quit [Ping timeout: 260 seconds]
mcopik has quit [Ping timeout: 265 seconds]
hkaiser has quit [Quit: bye]
eschnett has quit [Quit: eschnett]
eschnett has joined #ste||ar
Nikunj has quit [Quit: Page closed]
hkaiser has joined #ste||ar
<zao>
I wonder how much HPX tests break if you run them from a batch job.
<zao>
Time will tell, some day when I migrate my SLURM out of a VM to iron.
V|r has quit [Ping timeout: 265 seconds]
V|r has joined #ste||ar
V|r has quit [Changing host]
V|r has joined #ste||ar
<jbjnr_>
zao: pycicle is running all the tests on the dashboard from a slurm job on compute nodes of daint - so they should not break.
jakub_golinowski has joined #ste||ar
aserio has joined #ste||ar
<zao>
Sweet.
<zao>
jbjnr_: I went and set up a single-node SLURM instance in a VM yesterday, pestered my local batch wranglers with questions today on how to set it up properly :)
<zao>
I stopped myself trying to build a job runner myself, when I realized that SLURM actually does what I want it to :D
akheir has joined #ste||ar
<github>
[hpx] hkaiser created cmake_policy (+1 new commit): https://git.io/vAgLh
<victor_ludorum>
Here only addition,subtraction,multiplication and division is there .
EverYoung has quit [Remote host closed the connection]
<victor_ludorum>
Is the statistics/mean and arithmetic/mean same ??
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
<hkaiser>
victor_ludorum: we have two different types of statistics/arithmetic counters
<hkaiser>
victor_ludorum: one type is initialized from a set of counters and a time interval
<hkaiser>
those call the underlying counters themselves once every interval and calculate the statistic value from that - exposing the value whenever queried
<victor_ludorum>
Okay sir!
<hkaiser>
sorry , not initialized from a set of counters, but using one other counter - they calculate the statistic value from several invocations of the same counter over time
<hkaiser>
the other type of counters is initialized from a set of counters and they calculate the statistic value based on the values returned by those counters at the same point in time
<hkaiser>
while the firts type invokes the underlying counter once evry interval, the second type incokes all underlying counters only if somebody queries their value
<hkaiser>
victor_ludorum: I hope this makes sense
<victor_ludorum>
Yeah sir!!
<victor_ludorum>
Sir, then is arithmetic counters first type of counter ?
<hkaiser>
victor_ludorum: there are both types
<victor_ludorum>
yeah sorry I have misread something
<hkaiser>
feel free to create a pr removin the first of those two
<victor_ludorum>
Thanks Sir :)
<hkaiser>
;)
shahrzad has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
EverYoung has quit [Ping timeout: 260 seconds]
aserio has joined #ste||ar
EverYoung has joined #ste||ar
EverYoung has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
<github>
[hpx] victor-ludorum opened pull request #3189: Deleting one include of median from BOOST (master...include/median-twice) https://git.io/vAgox
anushi has quit [Ping timeout: 252 seconds]
nanashi64 has joined #ste||ar
nanashi55 has quit [Ping timeout: 240 seconds]
nanashi64 is now known as nanashi55
anushi has joined #ste||ar
aserio1 has joined #ste||ar
aserio has quit [Ping timeout: 276 seconds]
aserio1 is now known as aserio
victor_ludorum has quit [Quit: Page closed]
CaptainRubik has joined #ste||ar
CaptainRubik has quit [Client Quit]
hkaiser has quit [Quit: bye]
EverYoun_ has joined #ste||ar
aserio has quit [Ping timeout: 240 seconds]
EverYoung has quit [Ping timeout: 276 seconds]
aserio has joined #ste||ar
EverYoun_ has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
hkaiser has joined #ste||ar
EverYoun_ has joined #ste||ar
EverYoung has quit [Ping timeout: 260 seconds]
EverYoun_ has quit [Remote host closed the connection]
EverYoung has joined #ste||ar
david_pfander has quit [Ping timeout: 276 seconds]
<github>
[hpx] biddisco created topology_improvements (+3 new commits): https://git.io/vA2kD
<github>
hpx/topology_improvements e88fe81 John Biddiscombe: Add TLS bitmap for get_area_membind_nodeset and numa domain queries...
<github>
hpx/topology_improvements cd7711c John Biddiscombe: Add a user membind enum for custom allocator usage
<github>
hpx/topology_improvements 9eeb7df John Biddiscombe: Fix build error/compatibility with hwloc-2.0.0