00:02
EverYoung has joined #ste||ar
00:10
EverYoung has quit [Ping timeout: 255 seconds]
00:13
diehlpk has joined #ste||ar
00:15
<
diehlpk >
I try to fix hpxcl for the latest hpx master and now it compiles, but his here happens /usr/include/boost/thread/detail/thread.hpp:91: undefined reference to `boost::detail::thread_data_base::~thread_data_base()'
00:15
<
heller_ >
you forgot to link against boost thread
00:15
<
heller_ >
diehlpk: you should use hpx::compat
00:15
<
heller_ >
diehlpk: look at the PR I linked you earlier
00:16
<
diehlpk >
Ok, can you sent it again
00:16
<
diehlpk >
I am at home at a different machine
00:53
eschnett has joined #ste||ar
01:28
mcopik has quit [Ping timeout: 246 seconds]
01:43
diehlpk has quit [Remote host closed the connection]
01:53
K-ballo has quit [Quit: K-ballo]
02:30
hkaiser has quit [Quit: bye]
02:40
zbyerly_ has joined #ste||ar
03:08
<
zbyerly_ >
is rostam down?
05:37
wash has joined #ste||ar
07:17
dmarce1 has joined #ste||ar
07:18
<
dmarce1 >
heller: wash: Is parcel coalescing know to be buggy? If not, I have exposed one.
07:18
<
dmarce1 >
heller_:^
07:24
bikineev has joined #ste||ar
07:31
david_pfander has joined #ste||ar
09:06
<
github >
hpx/gh-pages 0ea588d StellarBot: Updating docs
09:49
david_pfander has quit [Ping timeout: 245 seconds]
11:10
david_pfander has joined #ste||ar
11:18
david_pfander has quit [Ping timeout: 240 seconds]
11:29
K-ballo has joined #ste||ar
11:33
bikineev has quit [Remote host closed the connection]
11:42
bikineev has joined #ste||ar
11:46
david_pf_ has joined #ste||ar
11:54
bikineev has quit [Remote host closed the connection]
12:04
denis_blank has joined #ste||ar
12:28
K-ballo1 has joined #ste||ar
12:28
jbjnr_ has joined #ste||ar
12:28
david_pf_ has quit [Remote host closed the connection]
12:30
jbjnr has quit [Ping timeout: 255 seconds]
12:30
jbjnr_ is now known as jbjnr
12:32
K-ballo has quit [Remote host closed the connection]
12:32
K-ballo1 is now known as K-ballo
12:47
zbyerly_ has quit [Ping timeout: 246 seconds]
13:12
hkaiser has joined #ste||ar
13:25
parsa[[[w]]] has joined #ste||ar
13:28
parsa[[w]] has quit [Ping timeout: 246 seconds]
13:32
mcopik has joined #ste||ar
14:52
<
K-ballo >
those MOVABLE_ONLY/NON_COPYABLE macros, do we still want them? for, I don't know, explicitness?
15:01
diehlpk has joined #ste||ar
16:03
denis_blank has quit [Quit: denis_blank]
16:12
mcopik has quit [Ping timeout: 240 seconds]
16:17
bikineev has joined #ste||ar
16:31
<
hkaiser >
K-ballo: they don't hurt, do they?
16:31
<
hkaiser >
they make things nicely explicit
16:31
<
K-ballo >
they kinda do, they lead to poor error messages
16:31
<
K-ballo >
if they stay they need to be made public:
16:33
<
hkaiser >
K-ballo: which would be fine, except this would need checking all uses to make sure adding a public does not change anything
16:34
<
K-ballo >
I didn't mean making the replacement public, but the usage
16:34
<
K-ballo >
the replacement should be access agnostic, it wasn't pre =delete which is what they were suggested to be used as private: HPX_MOVABLE_ONLY
16:35
<
K-ballo >
now they need to be public: HPX_MOVALBLE_ONLY
16:35
<
hkaiser >
sounds good to me
16:42
quaz0r has quit [Ping timeout: 260 seconds]
16:56
bikineev has quit [Ping timeout: 272 seconds]
16:56
quaz0r has joined #ste||ar
17:12
Matombo has joined #ste||ar
17:31
<
K-ballo >
I'm rethinking MOVABLE_ONLY now, it doesn't make things movable
17:31
<
K-ballo >
it just prevents them from becoming copyable
17:31
<
K-ballo >
copyable is not the right term either
17:32
<
K-ballo >
NON_COPYABLE on the other hand is safe
17:32
bikineev has joined #ste||ar
17:33
<
K-ballo >
I'll make a pass applying some of public:, some minor refaction, and some of idiomatic =default/=delete.. see where that takes us
17:49
<
K-ballo >
MOVABLE_ONLY is going, there's a considerable number of misuses, the rest are (should be) redundant
18:02
diehlpk has quit [Remote host closed the connection]
18:03
<
github >
hpx/emulate_deleted b92bfed Agustin K-ballo Berge: Drop misleading HPX_MOVABLE_ONLY...
18:13
bikineev has quit [Remote host closed the connection]
18:13
<
github >
hpx/emulate_deleted 51fa433 Agustin K-ballo Berge: Remove obsolete HPX_DELETE_* config macros
18:13
<
K-ballo >
hkaiser: you'll need top-of-trunk clang and libc++
18:13
<
hkaiser >
K-ballo: nod, ok
18:13
<
hkaiser >
need to create build system support etc.
18:14
<
hkaiser >
K-ballo: is the drop vc2013 PR ready to go now?
18:14
<
hkaiser >
(you force pushed something yesterday... is there more to come?)
18:14
<
K-ballo >
it is, let me check if I need to rebase
18:15
<
K-ballo >
that was a rebase on top of master for new commits, audited for conflicts
18:15
<
K-ballo >
seems it's up-to-date
18:16
<
github >
hpx/master bb0882c Khalid Hasanov: the first compiling version of a new throttling policy
18:16
<
github >
hpx/master 004f937 Khalid Hasanov: the first compiling version of a new throttling policy
18:16
<
github >
hpx/master bf8e7d5 Khalid Hasanov: first version that has been tested with the allscale runtime
18:16
<
K-ballo >
hkaiser: it's no longer ready
18:16
<
hkaiser >
K-ballo: no need to rebase
18:17
<
K-ballo >
no conflicts? no uses of things that have gone away, etc?
18:17
<
hkaiser >
as long as it does not create conflicts it's fine
18:17
<
K-ballo >
it can easily be using old things that don't exist anymore, that wouldn't cause conflicts
18:17
<
hkaiser >
it's a 100% non-msvc patch
18:17
<
K-ballo >
doesn't mean anything
18:17
<
K-ballo >
let me audit real quick
18:17
<
hkaiser >
ok, thanks
18:18
<
hkaiser >
K-ballo: I meant it shouldn't get in the way of dropping vs2013
18:18
<
K-ballo >
it may, it could for instance be using HPX_NOEXCEPT
18:18
<
K-ballo >
in which case merging vs2013 would break the build
18:19
<
hkaiser >
sorry, I should have merged yours first
18:19
<
K-ballo >
no, then his PR would have had to be audited
18:19
<
K-ballo >
and it's easier for me to do it anyways
18:21
<
K-ballo >
hkaiser: quick-audited, it
_should_ be safe
18:49
<
jbjnr >
hkaiser: heller_ K-ballo you all pinged me within the last couple of days.
18:49
<
jbjnr >
anything I need to do?
18:50
<
hkaiser >
jbjnr: nothing urgent (anymore)
18:50
<
jbjnr >
K-ballo: the std::move looks fine to me, but I'll have another look
18:50
<
jbjnr >
hkaiser: what did you find?
18:50
<
hkaiser >
jbjnr: heller_ tried to fix it already
18:50
<
jbjnr >
what ws it though
18:51
<
hkaiser >
see #2619
18:51
<
hkaiser >
jbjnr: heller_ was accessing a container after moving from it
18:52
<
hkaiser >
this probably has caused problems whenever more than one parcel was sent
18:53
<
jbjnr >
hmmm. That code is fairly recent, so I may not have had problems yet
18:54
<
hkaiser >
jbjnr: 8 month ago
18:54
<
jbjnr >
well ,we don't send multiple parcels in LF until a few weeks back when heller added stuff for that I mean
18:56
<
jbjnr >
I don't see what's wrong with it. .... [pause]
18:56
<
hkaiser >
jbjnr: every parcel except the first one will see a zombie
18:59
<
K-ballo >
jbjnr: clang said it's redundant and prevents elision.. maybe it's needed for intel or an older gcc? doesn't `f.get()` return by value?
18:59
mcopik has joined #ste||ar
19:00
<
jbjnr >
I'll try removing it and see if it work ok
19:00
<
hkaiser >
jbjnr: will not compile, use #2619 instead
19:00
zbyerly_ has joined #ste||ar
19:02
<
jbjnr >
hkaiser: sorry I meant the move that K-ballo referred to. I will remove it
19:03
<
K-ballo >
the number of lines over 90 cols with windows line endings went up to ~30 :/ not useful
20:44
bikineev has joined #ste||ar
21:02
diehlpk has joined #ste||ar
21:27
diehlpk has quit [Ping timeout: 268 seconds]
21:40
<
github >
hpx/emulate_deleted 0692fea Agustin K-ballo Berge: Improve uses of HPX_NON_COPYABLE...
22:12
Matombo has quit [Remote host closed the connection]
22:55
wash[m] has quit [Ping timeout: 245 seconds]
22:56
wash[m] has joined #ste||ar
23:00
bikineev has quit [Remote host closed the connection]
23:07
bikineev has joined #ste||ar
23:19
diehlpk has joined #ste||ar
23:20
diehlpk has quit [Client Quit]
23:32
bikineev has quit [Remote host closed the connection]