nan11 has quit [Remote host closed the connection]
weilewei has quit [Remote host closed the connection]
hkaiser has quit [Quit: bye]
bita has joined #ste||ar
bita has quit [Ping timeout: 260 seconds]
nikunj has quit [Ping timeout: 260 seconds]
nikunj has joined #ste||ar
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
mdiers_ has quit [Remote host closed the connection]
mdiers_ has joined #ste||ar
nikunj has quit [Ping timeout: 260 seconds]
nikunj has joined #ste||ar
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
nikunj has quit [Ping timeout: 260 seconds]
nikunj has joined #ste||ar
kale_ has joined #ste||ar
karame_ has quit [Remote host closed the connection]
diehlpk_work_ has joined #ste||ar
diehlpk_work has quit [Ping timeout: 260 seconds]
jaafar_ has joined #ste||ar
jaafar has quit [Ping timeout: 252 seconds]
nikunj97 has joined #ste||ar
Nikunj__ has joined #ste||ar
nikunj97 has quit [Ping timeout: 260 seconds]
hkaiser has joined #ste||ar
<hkaiser>
simbergm: hey, do you feel better now?
Nikunj__ has quit [Ping timeout: 260 seconds]
<simbergm>
hkaiser: better, thanks
<simbergm>
don't know what it was...
<hkaiser>
good!
<hkaiser>
common flu?
<simbergm>
I guess
<hkaiser>
I'm glad you're over it, it's lonely here without you ;-)
<hkaiser>
simbergm: I have finalized #4277 and rebased #4311
<hkaiser>
if you have the time, could you please have a look?
<hkaiser>
simbergm: also PASC got postponed
<simbergm>
hkaiser: really? I'm sure you could do with less of me every now and then... ;)
<simbergm>
thanks a lot! I'll have a look
<simbergm>
I saw you had some build problems still yesterday, want me to have a look at that or is it resolved?
<simbergm>
yeah, just saw the email about pasc
<hkaiser>
simbergm: yah #4311 after rebasing has build issues, but I was hoping for John to have a look
<hkaiser>
this is nothing I introduced
<simbergm>
at least we now have the minisymposium finalized, and now you have no excuses for not being there next year ;)
<hkaiser>
indeed, let's see what new 'threat' the world will invent by then
Nikunj__ has joined #ste||ar
<hkaiser>
simbergm: John said that he stopped reading github messages - do you have a better way of talking to him?
kale_ has quit [Ping timeout: 258 seconds]
kale_ has joined #ste||ar
weilewei has joined #ste||ar
<simbergm>
hkaiser: yeah, I can try on slack, he's fairly responsive there
<hkaiser>
ok, thanks
<simbergm>
I looked through the prs, I don't have much to add on the mpi one (much less on the dataflow one), but all the open comments still stand
<simbergm>
I can have a go at some of them if you'd like
<simbergm>
I'll try getting john first though
<simbergm>
jbjnr: ping?
<hkaiser>
I think I fixed all the comments on the MPI futures one
<hkaiser>
that should be good to go now
<weilewei>
hkaiser nice!
<hkaiser>
weilewei: would you mind trying that out, pls?
<simbergm>
hmm, I hope I didn't review outdated commits...
<simbergm>
I'll check again
<weilewei>
hkaiser Yea, sure, I can test it today on my ring algorithm
<hkaiser>
simbergm: let me know if I missed something, pls
nan11 has joined #ste||ar
<simbergm>
hkaiser: I'm fine with leaving most of the comments for future prs, but the namespace and option names would be nice to have "right" now (the rest is mostly invisible to users)
<hkaiser>
I fixed the options, didn't I?
<hkaiser>
simbergm: the namespaces should be ok now as well, did I miss some ?
<simbergm>
john responded as well, he'll have a look at the prs
<hkaiser>
ok, thanks
<simbergm>
and thanks hkaiser for updating those prs!
<hkaiser>
the mpi one was a mess, frankly... - still has some wrinkles, but I could live with it
<zao>
Been porting some code to CUDA the last few days. I spy a Bryce in the github contributors of the thrust library I found :D
<hkaiser>
zao: Bryce owns thrust nowadays
<zao>
So nice, compared to the C I've been writing the last few weeks.
<weilewei>
I am on DCA++ project and its CI is hosted by CSCS Jerkins, and jbjnr did not response my request... sorry for cross posting
<simbergm>
weilewei: I just run the builders ;) I think most likely you won't get access to that because of CSCS policies...
<simbergm>
but I'll ask john
<weilewei>
simbergm thanks, that's also good to know, but I can't even see the dashboard after I created PR
<hkaiser>
simbergm: cool!
<simbergm>
weilewei: yep, I get it, I don't like those policies either (I may not be up to date on our policies either!)
<weilewei>
simbergm sure, thanks for checking
<simbergm>
it practically makes our ci useless for projects that have collaborators outside of cscs
<weilewei>
True though...
<simbergm>
hkaiser: I think 4487 can be merged unless you had something else you wanted to change on it?
<hkaiser>
simbergm: yep, good to go
<simbergm>
merged, thanks!
<hkaiser>
thank you!
<weilewei>
hkaiser the updated mpi_futures works on my ring G test! Thanks
nikunj97 has joined #ste||ar
<simbergm>
weilewei: btw, you do have john's email, right? just want to make sure irc isn't the *only* way you can reach him (rori or I are not always here...)
<hkaiser>
simbergm: is he still reading email?
<simbergm>
hkaiser: well, I don't know... I hope so
<simbergm>
he did respond to the pragma once email, so every once in a while at least
<hkaiser>
ok
<hkaiser>
weilewei: thanks for checking
<hkaiser>
weilewei: would you mind commenting on the ticket as well?
<weilewei>
hkaiser ok
kale_ has quit [Ping timeout: 260 seconds]
<nikunj97>
hkaiser, does HPX 1.4.1 have updates to the block executor as well?
<nikunj97>
I will be running my final sets of benchmarks
<hkaiser>
nikunj97: I don't think so
<nikunj97>
ohh I see
<nikunj97>
is there a tag I can use which contains the updates?
<nikunj97>
I'm not sure if it's right to use commit hash as HPX versions
<nikunj97>
for the report i.e.
<heller1>
Why not?
<nikunj97>
heller1, I've never seen that before in any reports
<nikunj97>
they usually mention version numbers or tag numbers
<nikunj97>
so just wanted to make sure
<heller1>
Well, a tag is just a name for a commit
<heller1>
If you want, you can fork hpx, create a tag and reference that
<nikunj97>
aah yea!
<nikunj97>
btw I got access to new nodes with thunderX2 so I'll be running the benchmark in there as well
<nikunj97>
and I may get access to A64FX processor from Fujitsu as well if things go the way they're going
<diehlpk_work_>
nikunj97, I would not count on A64FX, since they never responded to us
<nikunj97>
diehlpk_work_, the professor I'm working with (Prof. Dirk Pleiter) already has access to them
<nikunj97>
so I should get access to them if he finds the progress nice
<diehlpk_work_>
nikunj97, Cool
<nikunj97>
after last call he gave me access to thunderX2 and asked to emulate the code for SVE. He said if SVE codes work like the way it's intended to on the emulator, he should be able to provide me with proxy access
<nikunj97>
if not actual access
<diehlpk_work_>
Ok, I do not know why we coudl not get access
<nikunj97>
in any case if we get HPX on A64FX, we'll be known in community as well!
wate123_Jun has joined #ste||ar
Hashmi has joined #ste||ar
wate123__ has joined #ste||ar
wate123_Jun has quit [Ping timeout: 260 seconds]
wate123__ has quit [Remote host closed the connection]
<nikunj97>
hkaiser, could you tell me the cmake commands while building HPX that enables APEX and OTF2 support?
<hkaiser>
nikunj97: no, I can't - this is described somewhere, possibly on the APEX website
<nikunj97>
aah alright, will take a look
<hkaiser>
nikunj97: Patrick's Octo-Tiger scripts could give you some inside as well
<tiagofg[m]>
I've been doing some tests with hpx channels and when I use get(), all cores of my cpu stay in busy-waiting.
<tiagofg[m]>
is it supposed?
<tiagofg[m]>
what I want is to wait for messages idly, but I don't know how
<hkaiser>
tiagofg[m]: hmmm
<hkaiser>
the schedulers will try finding work
<hkaiser>
what do you mean by 'wait idly'?
<tiagofg[m]>
wait without wasting cpu time
<hkaiser>
well, your CPUs run anyways, don't they?
<tiagofg[m]>
yes of course
<hkaiser>
tiagofg[m]: you can instruct the HPX scheduler to step back quickly by setting fast_idle mode
<tiagofg[m]>
my idea of all this is to have a server that waits for messages from clients.
<tiagofg[m]>
ok, i will figure that out, thank you
<hkaiser>
nod, let me know if you need help, I think the API is hpx::set_scheduler_mode()
<hkaiser>
or add_scheduler_mode, I don't remember
<tiagofg[m]>
I'll find
Hashmi has quit [Quit: Connection closed for inactivity]
K-ballo has quit [Remote host closed the connection]
K-ballo has joined #ste||ar
<diehlpk_work_>
Can anyone please check if he can access his old home dir on rostam?
<diehlpk_work_>
Mine is not available
<hkaiser>
diehlpk_work_: ping Alireza
<diehlpk_work_>
Ok
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
nikunj has quit [Read error: Connection reset by peer]
nikunj has joined #ste||ar
nikunj has quit [Ping timeout: 256 seconds]
nikunj has joined #ste||ar
ibalampanis has quit [Remote host closed the connection]
nan11 has quit [Remote host closed the connection]
nan11 has joined #ste||ar
ibalampanis has joined #ste||ar
<diehlpk_mobile[m>
hkaiser see pm
ibalampanis has quit [Quit: Ping timeout (120 seconds)]
ibalampanis has joined #ste||ar
<ibalampanis>
Hello hkaiser , bita . Provided that my thesis work is going better than the predicted concerning the timeline, I'd like to work on hpx/phylanx/blaze to make myself more familiar. Have you in your mind something to advise me about that?
<bita>
I think hkaiser would answer better. In my opinion, you can run als algorithm, see what it needs and get familiar with its python, physl and c++ version. When you did it, we can discuss further about the LDA
<hkaiser>
+1
<ibalampanis>
Thank you for your response. Always feel like home here with you!
<bita>
np :)
ibalampanis has quit [Remote host closed the connection]
nikunj97 has quit [Ping timeout: 260 seconds]
K-ballo has quit [Remote host closed the connection]
<hkaiser>
all of Phylanx works with int64_t internally (because of the node_data<int64_t>), but I know thi sis a weak argument in this case
<hkaiser>
bita: so my answer is: I don't know why it is the way it is ;-)
<bita>
Okay thank you
<hkaiser>
is that a problem for you?
<hkaiser>
bita: ^^
<bita>
Igot this: 1>c:\repos\phylanx\src\plugins\dist_matrixops\retile_annotations.cpp(163): error C2397: conversion from 'int64_t' to 'size_t' requires a narrowing conversion
<hkaiser>
ahh, right
<bita>
and I was curios, but it's not a problem, I am gonna use int64_t in my struct
<hkaiser>
use () instead of {} ;-)
<hkaiser>
that will shut up the compiler
<bita>
well it now does not find the constructor :D
<hkaiser>
lol
<hkaiser>
ok
<bita>
no worries, I will change the type in my struct
<hkaiser>
or cast it explicitly
<bita>
got it thanks
nan11 has quit [Remote host closed the connection]
weilewei has quit [Remote host closed the connection]