02:44
hkaiser has quit [Quit: bye]
03:33
nanashi55 has quit [Ping timeout: 252 seconds]
03:35
nanashi55 has joined #ste||ar
04:02
Anushi1998 has joined #ste||ar
04:36
eschnett_ has quit [Quit: eschnett_]
06:37
jaafar has quit [Ping timeout: 272 seconds]
08:45
Anushi1998 has quit [Ping timeout: 252 seconds]
09:38
ste||ar-github has joined #ste||ar
09:38
<
ste||ar-github >
hpx/remove_thread_specific_ptr 4a33893 Thomas Heller: Use HPX_NATIVE_TLS instead of hpx::util::thread_specific_ptr...
09:38
ste||ar-github has left #ste||ar [#ste||ar]
11:04
jbjnr__ has joined #ste||ar
11:08
jbjnr_ has quit [Ping timeout: 276 seconds]
12:14
jbjnr_ has joined #ste||ar
12:17
jbjnr__ has quit [Ping timeout: 252 seconds]
12:37
hkaiser has joined #ste||ar
12:41
Anushi1998 has joined #ste||ar
14:30
mcopik has quit [Ping timeout: 252 seconds]
14:33
mcopik has joined #ste||ar
14:44
eschnett_ has joined #ste||ar
16:39
hkaiser has quit [Quit: bye]
16:39
heller__ has joined #ste||ar
16:47
Anushi1998 has quit [Ping timeout: 252 seconds]
16:52
Anushi1998 has joined #ste||ar
18:20
eschnett_ has quit [Quit: eschnett_]
19:17
ste||ar-github has joined #ste||ar
19:17
<
ste||ar-github >
hpx/fibhash e24602a Thomas Heller: Fixing dataflow...
19:17
<
ste||ar-github >
hpx/fibhash 03131a1 Thomas Heller: Improving spinlook pools by using a multiplicative fibonacci based hash...
19:17
<
ste||ar-github >
hpx/fibhash ba3127d Thomas Heller: Getting rid of modulo operation
19:17
ste||ar-github has left #ste||ar [#ste||ar]
19:34
Anushi1998 has quit [Ping timeout: 252 seconds]
20:42
<
K-ballo >
anyone tried building HPX on WSL yet?
20:44
hkaiser has joined #ste||ar
20:47
<
zao >
Not recently, how so?
20:48
<
K-ballo >
just curious
20:49
<
zao >
The only weirdness I've observed building a fair bit of software with EasyBuild on Debian under WSL, is that Tcl's networking tests hang.
20:50
<
heller__ >
hkaiser: I fixed the compilation errors
20:51
<
zao >
Started a build on my home machine of an early 1.2.0-rc1, but it'll take a while.
20:51
<
zao >
Not overly WSL-dependent, as I've got my toolchain and Boost from modules, and I'm not running Ubuntu - but the kernel's the same :)
21:11
<
K-ballo >
don't thank me, it's a team effort
21:13
mcopik has quit [Ping timeout: 252 seconds]
21:27
<
hkaiser >
heller__: thanks!
21:30
mcopik has joined #ste||ar
21:36
<
heller__ >
hkaiser: please let me know how it goes!
21:42
<
zao >
K-ballo: Build succeded at least, tests are running.
21:43
<
zao >
HPX-1.2.0-rc1-g38ecfb0ec6c-foss-2018b-cxx14.eb
21:45
<
zao >
EasyBuild releases toolchains twice a year, this is GCC + OpenMPI + OpenBLAS.
21:45
<
zao >
GCC 7.3.0 (binutils 2.30), OpenMPI 3.1.1, OpenBLAS 0.3.1
21:46
<
zao >
Tests succeeded \i/
22:09
<
K-ballo >
hkaiser: remind me why are we (optionally) using return type deduction in execution traits?
22:56
<
hkaiser >
K-ballo: what are you referring to?
23:01
<
hkaiser >
K-ballo: we've had some problems making things compile on modern compilers without that
23:01
<
hkaiser >
don't remember the details, hodl on
23:02
<
K-ballo >
would that be issues with modern compilers when compiling with C++11?
23:02
<
K-ballo >
let's see
23:03
<
hkaiser >
it was related to NVCC
23:03
<
K-ballo >
looks broken
23:04
<
hkaiser >
seems to compile, though
23:04
<
K-ballo >
yeah, seems it was massaged enough so that it would compile
23:04
<
K-ballo >
but it doesn't make much sense
23:05
<
hkaiser >
NVCC never makes any sense
23:05
<
K-ballo >
do we have an NVCC builder?
23:05
<
hkaiser >
not right now I believe, simbergm?
23:06
<
K-ballo >
I'd need to replicate those incomplete type errors, to actually fix them
23:07
<
hkaiser >
think gentryx was compiling on linux
23:48
ste||ar-github has joined #ste||ar
23:48
<
ste||ar-github >
hpx/no-return-type-deduction 75a82f6 Agustin K-ballo Berge: (trigger incomplete type failures)
23:48
ste||ar-github has left #ste||ar [#ste||ar]
23:59
<
zao >
Aaw, build failed, but not in the way I wanted - "virtual memory exhausted: Cannot allocate memory"