2021-07-01 15:44
hkaiser changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar-group.org | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | This channel is logged: irclog.cct.lsu.edu | Everybody: please respond to the documentation survey:
https://forms.gle/aCpNkhjSfW55isXGA
01:00
qiu has joined #ste||ar
01:15
qiu has quit [Quit: Ping timeout (120 seconds)]
01:49
Yorlik_ has joined #ste||ar
01:51
K-ballo has quit [Quit: K-ballo]
01:52
Yorlik has quit [Ping timeout: 255 seconds]
01:57
hkaiser has quit [Quit: Bye!]
03:16
diehlpk_work has quit [Remote host closed the connection]
06:08
Yorlik_ is now known as Yorlik
08:22
<
heller[m] >
ms yes. Do you have credentials for it?
08:22
<
ms[m] >
heller: nope, I was mostly responding to hkaiser's comment from yesterday that he has the password but no username
08:22
<
ms[m] >
so I think we have the pieces, we just need to put them together...
08:23
<
heller[m] >
I might have the credentials somewhere... But shit is hitting the fan here. Sorry for not being able to help
08:32
<
ms[m] >
heller: no worries, there's no rush with it
08:32
<
ms[m] >
it's mostly nice to have so that github statuses aren't reported through gdaiss's or my github account ;)
11:50
K-ballo has joined #ste||ar
12:47
qiu has joined #ste||ar
12:56
hkaiser has joined #ste||ar
13:06
ms[m] has quit [Quit: Reconnecting]
13:20
qiu has quit [Ping timeout: 246 seconds]
13:21
qiu has joined #ste||ar
13:43
qiu has quit [Quit: Client closed]
13:47
ms[m] has joined #ste||ar
13:48
<
hkaiser >
any idea what they have in common that could cause this?
13:49
<
gnikunj[m] >
Let me have a look
13:53
<
hkaiser >
gnikunj[m]: it's most likely caused by somethiing I did, but I'm a bit clueless atm
13:54
<
gnikunj[m] >
It seems to be arriving at the time of HPX initialization itself
13:55
<
hkaiser >
gnikunj[m]: you think so?
13:55
<
gnikunj[m] >
hkaiser: the stack trace hasn't gone into any of the resilience related functions
13:55
<
hkaiser >
gnikunj[m]: could you try reproducing it?
13:55
<
hkaiser >
I know :/
13:56
<
hkaiser >
but it's happening for the resilience tests only
13:56
<
gnikunj[m] >
Yes, let me try reproducing it
13:56
qiu has joined #ste||ar
13:57
<
hkaiser >
gnikunj[m]: also: meeting
13:57
<
gnikunj[m] >
Yes, installing zoom on my other laptop atm. Will join asap.
14:07
<
ms[m] >
hkaiser, gnikunj, I don't have any particularly useful hints for you, but it's not necessarily during initialization
14:08
<
ms[m] >
it's happening while trying to run a thread function, asan is just reporting that os thread was spawned from the init
14:08
<
gnikunj[m] >
ms: I'm currently trying to reproduce it. Thanks for the heads up!
14:08
qiu has quit [Quit: Client closed]
14:28
mdiers[m] has quit [*.net *.split]
14:28
Deepak1411[m] has quit [*.net *.split]
14:28
gdaiss[m] has quit [*.net *.split]
14:49
<
hkaiser >
ms[m]: thanks
14:50
<
hkaiser >
I can't reproduce it here, unfortunately, even with ASAN enabled
15:22
<
gnikunj[m] >
hkaiser: why does HPX throws a cmake error searching for Asio when the backend parcelport is set to MPI?
15:23
<
gnikunj[m] >
ms: ^^
15:27
diehlpk_work has joined #ste||ar
15:28
<
ms[m] >
gnikunj: good question
15:28
<
ms[m] >
can you post the cmake config and output?
15:28
<
gnikunj[m] >
sure, let me raise a ticket
15:31
<
hkaiser >
gnikunj[m]: we need asio for the timers, not only the tcp parcelport
15:41
<
ms[m] >
gnikunj: well, do you have asio somewhere in a system location?
15:41
<
gnikunj[m] >
No, but why do we need asio when we're using MPI backend and not TCP?
15:41
<
ms[m] >
and does it really have anything to do with mpi? do you mean it works fine if you turn off the mpi parcelport?
15:41
<
ms[m] >
for timed execution
15:42
<
gnikunj[m] >
aah, I thought it was only used in the parcelport backends
15:42
<
hkaiser >
gnikunj[m]: I don't think we can get rid of the asio dependency atm
15:43
<
ms[m] >
well, timed execution and a few utilities (in addition to the parcelport)
15:43
<
gnikunj[m] >
got it
16:25
<
hkaiser >
gnikunj[m]: your do{ if(cond()) break; } while(true); is now a more complicated version of while(cond()) {} :/
16:26
<
hkaiser >
what's wrong with do { yield(); } while(!try_lock()); ?
16:59
<
diehlpk_work >
hkaiser, gnikunj[m] gonidelis[m] Have you finished the GSoC evaluation?
16:59
<
diehlpk_work >
Deadline is this Friday
17:47
<
gnikunj[m] >
hkaiser: see pm pls
17:52
<
hkaiser >
gnikunj[m]: what's up?
18:37
hkaiser has quit [Read error: Connection reset by peer]
19:29
hkaiser has joined #ste||ar
19:30
ms[m] has quit [Read error: Connection reset by peer]
19:30
rori[m] has quit [Read error: Connection reset by peer]
19:30
KordeJong[m] has quit [Read error: Connection reset by peer]
19:30
gonidelis[m] has quit [Read error: Connection reset by peer]
19:30
jedi18[m] has quit [Read error: Connection reset by peer]
19:30
sestro[m] has quit [Read error: Connection reset by peer]
19:30
rachitt_shah[m] has quit [Read error: Connection reset by peer]
19:30
PatrickDiehl[m] has quit [Read error: Connection reset by peer]
19:30
pedro_barbosa[m] has quit [Read error: Connection reset by peer]
19:30
heller[m] has quit [Read error: Connection reset by peer]
19:30
gnikunj[m] has quit [Read error: Connection reset by peer]
19:30
srinivasyadav227 has quit [Read error: Connection reset by peer]
19:30
jpinto[m] has quit [Read error: Connection reset by peer]
20:52
PatrickDiehl[m] has joined #ste||ar
21:02
jedi18[m] has joined #ste||ar
21:02
ms[m] has joined #ste||ar
21:02
rori[m] has joined #ste||ar
21:02
gdaiss[m] has joined #ste||ar
21:02
KordeJong[m] has joined #ste||ar
21:02
rachitt_shah[m] has joined #ste||ar
21:02
pedro_barbosa[m] has joined #ste||ar
21:02
sestro[m] has joined #ste||ar
21:02
heller[m] has joined #ste||ar
21:02
srinivasyadav227 has joined #ste||ar
21:02
jpinto[m] has joined #ste||ar
21:02
Deepak1411[m] has joined #ste||ar
21:02
gonidelis[m] has joined #ste||ar
21:02
gnikunj[m] has joined #ste||ar
21:02
mdiers[m] has joined #ste||ar