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/
hkaiser has quit [Quit: bye]
nanashi55 has quit [Ping timeout: 244 seconds]
nanashi55 has joined #ste||ar
K-ballo has quit [Ping timeout: 244 seconds]
K-ballo has joined #ste||ar
david_pfander has joined #ste||ar
eschnett_ has quit [Read error: Connection reset by peer]
<ste||ar-github>
hpx/fixing_jemalloc_prefix aa1f44b Hartmut Kaiser: Fixing invalid cmake code if no jemalloc prefix was given
ste||ar-github has left #ste||ar [#ste||ar]
ste||ar-github has joined #ste||ar
<ste||ar-github>
[hpx] hkaiser opened pull request #3473: Fixing invalid cmake code if no jemalloc prefix was given (master...fixing_jemalloc_prefix) https://github.com/STEllAR-GROUP/hpx/pull/3473
ste||ar-github has left #ste||ar [#ste||ar]
<hkaiser>
heller: this ^^ should solve it, pls verify
aserio has joined #ste||ar
diehlpk has joined #ste||ar
K-ballo has quit [Quit: K-ballo]
K-ballo has joined #ste||ar
diehlpk has quit [Remote host closed the connection]
diehlpk has joined #ste||ar
david_pfander has quit [Ping timeout: 244 seconds]
aserio has quit [Ping timeout: 260 seconds]
diehlpk has quit [Ping timeout: 252 seconds]
aserio has joined #ste||ar
<heller>
hkaiser: hey
<hkaiser>
hey
<heller>
appear.in?
<hkaiser>
heller: I'm in
<hkaiser>
heller: can't hear you
diehlpk has joined #ste||ar
hkaiser has quit [Quit: bye]
<heller>
for the protocol: The MPI parcelport on Cray machines just sucks
<heller>
plain and simple...
<heller>
it needs to go ASAP
<khuck>
heller: is libfabric ready for prime-time?
<heller>
khuck: I think so yes.
<heller>
I was using it on a 9k node run recently
<aserio>
heller: slow is better than broken
<heller>
even the TCP parcelport is faster than my implementation
<heller>
than MPI with my applciation
<heller>
aserio: and really, I am using the libfabric parcelport exclusively on cori, no problems there
<heller>
consolidating the parcelports is something for post 1.2
<khuck>
HPX still calls a handful of posix routines from worker threads
<heller>
hkaiser: text pushed
<khuck>
but phylanx no longer does
<hkaiser>
heller: thanks! bravo! much appreciated!
<heller>
:D
<hkaiser>
khuck: thanks for this - I was not aware of those
<khuck>
np
<heller>
1.4 GB/s aggregated bandwidth isn't that much ;)
<hkaiser>
khuck: those calls look like to happen before the runtime is up, though
<khuck>
hkaiser: they were on the OS thread that was a "worker". Like I said, they are pretty minor - they hardly showed up on the profile and I had to hunt for them
<khuck>
sorry, across two different threads. but workers nonetheless
<hkaiser>
khuck: ok
<hkaiser>
I'll try to identify them
<heller>
maybe late module loading or ini parsing, maybe?
<hkaiser>
the parse_command_line should happen right at the start, no idea why it shows up to run on a worker thread
<khuck>
it's also possible it happens before that thread is labeled as a "worker"
quaz0r has joined #ste||ar
ste||ar-github has joined #ste||ar
<ste||ar-github>
[hpx] sithhell closed pull request #3473: Fixing invalid cmake code if no jemalloc prefix was given (master...fixing_jemalloc_prefix) https://github.com/STEllAR-GROUP/hpx/pull/3473