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/
eschnett has joined #ste||ar
jaafar has quit [Ping timeout: 264 seconds]
ste||ar-github has joined #ste||ar
<ste||ar-github>
[hpx] msimberg opened pull request #3502: Fix bogus `throws` bindings in scheduled_thread_pool_impl (master...fix-std-move-ref-thread-pool) https://github.com/STEllAR-GROUP/hpx/pull/3502
<ste||ar-github>
hpx/msimberg-patch-1 152aba2 Mikael Simberg: Fix comparison for expect_connecting_localities config entry...
<ste||ar-github>
hpx/msimberg-patch-1 ff0595d Mikael Simberg: Fix launch_process test...
ste||ar-github has left #ste||ar [#ste||ar]
anushi has joined #ste||ar
anushi has quit [Ping timeout: 264 seconds]
anushi has joined #ste||ar
anushi has quit [Ping timeout: 252 seconds]
jaafar has joined #ste||ar
jaafar has quit [Ping timeout: 250 seconds]
mcopik has quit [Read error: Connection reset by peer]
anushi has joined #ste||ar
david_pfander has joined #ste||ar
anushi has quit [Ping timeout: 250 seconds]
anushi has joined #ste||ar
anushi has quit [Ping timeout: 244 seconds]
anushi has joined #ste||ar
anushi has quit [Ping timeout: 244 seconds]
Anushi1998 has joined #ste||ar
Anushi1998 has quit [Ping timeout: 240 seconds]
Anushi1998 has joined #ste||ar
Anushi1998 has quit [Ping timeout: 252 seconds]
Anushi1998 has joined #ste||ar
<zao>
jbjnr_: Having tried to get the gosh-darn BSD builds working, I would love a topology that also handles the case where you don't have the PCI info and have a flat topology.
<zao>
(regarding the SoC mail)
<zao>
That whole branch has been blocked on that I can literally not spoof enough topology junk to get HPX to not die whining about not being able to something OS threads.
<jbjnr_>
oh dear
<jbjnr_>
Well ideally, we would have a smart toplogy parsing thingy that would handle all cases
<jbjnr_>
so if your toplogy is unknown or flat, it jsut would say, N cores and that's all
<jbjnr_>
the locality/affinity is for optimization, but the basic funcrionality needs to work
<zao>
In mine, I had a test in the ctor that checks if the topology is simplified/flattened, and branch on that in most functions that care, but there's a lot of icky assumptions.
<zao>
I got it to the point that the topology looked sane, but something else threw a hissy-fit.
<zao>
Something about binding OS threads, but I don't have the gist around anymor.
<zao>
(nor do I seem to have pushed my branch to my github, should probably do it before all the BSD VMs disappear)
Anushi1998 has quit [Ping timeout: 252 seconds]
Anushi1998 has joined #ste||ar
Anushi1998 has quit [Remote host closed the connection]
<jbjnr_>
zao: does hwloc run ok on your machine? cos if so, that's all we should need, parse what it generates for our use
<zao>
jbjnr_: hwloc runs, the topology it exposes has IIRC Machine level, with the number of hyperthreads as direct children.
Anushi1998 has joined #ste||ar
nikunj has joined #ste||ar
<zao>
My source code is currently on a machine without memory, as I looted all the RAM to build HPX in another machine :)