00:17
parsa has quit [Quit: Zzzzzzzzzzzz]
00:33
<
github >
hpx/fixing_2847 82835bd Hartmut Kaiser: Allowing for non-default-constructible action argument and action return types
00:54
ajaivgeorge has joined #ste||ar
00:54
jaafar_ has joined #ste||ar
00:55
bikineev has quit [Remote host closed the connection]
00:56
bikineev has joined #ste||ar
01:00
bikineev has quit [Ping timeout: 240 seconds]
01:02
ajaivgeorge has quit [Quit: ajaivgeorge]
01:13
mcopik has quit [Ping timeout: 255 seconds]
01:35
jaafar_ has quit [Ping timeout: 240 seconds]
02:09
<
github >
hpx/fixing_2847 bc7eb90 Hartmut Kaiser: Allowing for non-default-constructible action argument and action return types
02:21
hkaiser has quit [Quit: bye]
03:18
parsa has joined #ste||ar
03:29
K-ballo has quit [Quit: K-ballo]
05:32
parsa has quit [Quit: Zzzzzzzzzzzz]
05:36
parsa has joined #ste||ar
05:40
parsa has quit [Client Quit]
05:44
parsa has joined #ste||ar
05:51
parsa has quit [Ping timeout: 246 seconds]
06:01
bikineev has joined #ste||ar
06:05
bikineev has quit [Ping timeout: 246 seconds]
06:08
<
jbjnr >
(I'm just about to go away for a few days, so won't be here much this week)
06:08
<
jbjnr >
hope you had a nice vacation
06:26
Matombo has joined #ste||ar
06:33
<
heller >
jbjnr: i did
06:34
<
heller >
how much did I miss?
07:21
bikineev has joined #ste||ar
08:02
david_pfander has joined #ste||ar
08:39
Matombo has quit [Remote host closed the connection]
09:07
<
github >
hpx/gh-pages 4c0262b StellarBot: Updating docs
09:26
bikineev has quit [Remote host closed the connection]
09:50
bikineev has joined #ste||ar
10:06
mcopik has joined #ste||ar
10:20
Matombo has joined #ste||ar
10:24
denis_blank has joined #ste||ar
10:24
Matombo has quit [Ping timeout: 248 seconds]
10:37
Matombo has joined #ste||ar
10:47
bikineev has quit [Ping timeout: 240 seconds]
10:52
Matombo has quit [Ping timeout: 240 seconds]
11:03
bikineev has joined #ste||ar
11:09
hkaiser has joined #ste||ar
11:19
hkaiser has quit [Read error: Connection reset by peer]
11:27
hkaiser has joined #ste||ar
11:46
mcopik has quit [Ping timeout: 240 seconds]
11:49
<
Vir >
hkaiser: I just pushed Vc 1.71
11:49
<
hkaiser >
Vir: ohh, cool!
11:49
<
Vir >
heller: on the master branch...
11:50
<
Vir >
so mkretz/datapar is still 1.70
11:50
<
hkaiser >
ohh heller! you're back!
11:50
<
heller >
for a little while
11:50
<
heller >
need to clean up all the mess that has been created while I was MIA :/
11:51
<
heller >
having a guest from allscale here to make progress ...
11:51
<
heller >
they messed up a bit
11:52
<
hkaiser >
the mice dance on the table if the cat's away
11:52
<
heller >
also trying to clean up your mess :P
11:52
<
heller >
the channel failure
11:53
<
hkaiser >
heller: also working on it
11:53
<
hkaiser >
but feel free
11:53
<
heller >
hkaiser: do you wanna know what I found out so far?
11:54
<
heller >
it looks like there is mismatch in the dynamic type of the created components
11:54
<
heller >
that's what ubsan is telling me
11:55
<
heller >
that is, in simple_component_base.hpp line 304, we cast to the wrong type
11:56
K-ballo has joined #ste||ar
12:02
<
hkaiser >
heller: uuh
12:02
<
hkaiser >
what test do you look at?
12:03
<
heller >
hkaiser: tests/unit/lcos/channel.cpp
12:03
<
hkaiser >
k, will look closer
12:12
eschnett has quit [Quit: eschnett]
12:48
<
heller >
hkaiser: nice. coverage increases
12:49
<
hkaiser >
nod, we're listed as well ;)
12:50
<
heller >
now, we only need a green buildbot :P
12:50
<
hkaiser >
things have improved massively over the last days
12:51
<
heller >
did it ;)?
12:51
<
heller >
I obviously only looked at it today
12:51
<
hkaiser >
believe me, I'm working hard on this front
12:52
<
heller >
what I am really hinting at that our PRs should be tested more thoroughly
12:52
<
hkaiser >
heller: absolutely
12:53
<
hkaiser >
compiling is obviously not sufficient
12:53
<
heller >
and it is frustrating to have it breaking almost every other merge
12:54
<
heller >
this current failure is very strange though :/
12:55
hkaiser has quit [Read error: Connection reset by peer]
12:55
eschnett has joined #ste||ar
12:55
hkaiser has joined #ste||ar
12:56
<
hkaiser >
heller: not having anybody to review PRs is not helping the issue either
12:57
<
heller >
hkaiser: yes :/
12:57
<
heller >
hkaiser: I am back now ...
12:57
<
heller >
but even manual code review doesn't catch everything
13:04
<
heller >
hkaiser: but I am still thinking that we should only start merging stuff once we have a (almost) green master branch ,)
13:15
<
hkaiser >
sure, that's what I'm doing
13:16
eschnett has quit [Quit: eschnett]
13:16
hkaiser has quit [Quit: bye]
13:18
eschnett has joined #ste||ar
13:23
eschnett has quit [Ping timeout: 240 seconds]
13:24
eschnett has joined #ste||ar
13:51
mcopik has joined #ste||ar
14:13
hkaiser has joined #ste||ar
14:18
<
heller >
hkaiser: got it fixed!
14:18
<
heller >
Will commit the fix later today
14:22
hkaiser has quit [Ping timeout: 276 seconds]
14:26
hkaiser has joined #ste||ar
14:29
hkaiser has quit [Read error: Connection reset by peer]
14:30
hkaiser has joined #ste||ar
14:31
<
hkaiser >
heller: I got the channel fixed - stupid typo after all
14:32
parsa has joined #ste||ar
14:32
hkaiser has quit [Read error: Connection reset by peer]
14:33
hkaiser has joined #ste||ar
14:37
david_pfander has quit [Ping timeout: 246 seconds]
14:38
pree has joined #ste||ar
14:38
parsa has quit [Read error: Connection reset by peer]
14:38
pree has quit [Read error: Connection reset by peer]
14:39
parsa| has joined #ste||ar
14:41
pree has joined #ste||ar
14:44
<
github >
hpx/master 4cb1934 Hartmut Kaiser: Fixing channel test
14:45
hkaiser has quit [Read error: Connection reset by peer]
14:46
hkaiser has joined #ste||ar
14:48
bikineev has quit [Ping timeout: 240 seconds]
14:50
parsa| has quit [Read error: Connection reset by peer]
14:51
parsa has joined #ste||ar
14:58
hkaiser has quit [Read error: Connection reset by peer]
14:59
hkaiser has joined #ste||ar
15:01
hkaiser has quit [Read error: Connection reset by peer]
15:01
parsa has quit [Read error: Connection reset by peer]
15:02
parsa has joined #ste||ar
15:05
hkaiser has joined #ste||ar
15:08
bikineev has joined #ste||ar
15:08
hkaiser has quit [Read error: Connection reset by peer]
15:08
parsa has quit [Read error: Connection reset by peer]
15:08
hkaiser_ has joined #ste||ar
15:08
parsa| has joined #ste||ar
15:10
hkaiser_ has quit [Read error: Connection reset by peer]
15:14
hendrx has joined #ste||ar
15:14
parsa| has quit [Read error: Connection reset by peer]
15:14
parsa has joined #ste||ar
15:14
jaafar_ has joined #ste||ar
15:17
hkaiser has joined #ste||ar
15:17
parsa has quit [Read error: Connection reset by peer]
15:18
parsa has joined #ste||ar
15:18
parsa has quit [Client Quit]
15:21
parsa has joined #ste||ar
15:22
hkaiser has quit [Read error: Connection reset by peer]
15:22
hkaiser has joined #ste||ar
15:22
<
K-ballo >
interesting... what about an indecisive codebase with different styles per file?
15:23
<
K-ballo >
mispoke, not per file
15:24
hkaiser has quit [Read error: Connection reset by peer]
15:25
hkaiser has joined #ste||ar
15:25
EverYoung has joined #ste||ar
15:25
<
zao >
(tool brute-forces possible settings by trying to format all the files indicated and picking the least mutating results)
15:25
pree has quit [Ping timeout: 276 seconds]
15:26
pree has joined #ste||ar
15:31
parsa has quit [Read error: Connection reset by peer]
15:31
parsa| has joined #ste||ar
15:32
hkaiser has quit [Ping timeout: 240 seconds]
15:34
parsa| has quit [Read error: Connection reset by peer]
15:34
parsa has joined #ste||ar
15:37
hkaiser has joined #ste||ar
15:42
hkaiser has quit [Read error: Connection reset by peer]
15:44
hkaiser has joined #ste||ar
15:45
<
github >
hpx/master 7a01561 Hartmut Kaiser: Attempting to fix shared_future test
15:48
<
heller >
hkaiser: actually, it's a bit more than that
15:49
<
heller >
hkaiser: base_lco_with_value is not derived from any of the components.
15:51
hkaiser has quit [Read error: Connection reset by peer]
15:51
hkaiser has joined #ste||ar
15:51
<
hkaiser >
heller: it shouldn't be
15:54
<
heller >
hkaiser: sure, look at how it is created though new_ though
15:55
<
heller >
we static_cast to component_type and then back to Component
15:55
hkaiser_ has joined #ste||ar
15:55
hkaiser has quit [Read error: Connection reset by peer]
15:56
parsa has quit [Read error: Connection reset by peer]
15:56
parsa has joined #ste||ar
15:56
<
heller >
hkaiser_: I'll commit what I have in an hour or two
15:57
<
hkaiser_ >
heller: I don't think base_lco_with_value should be derived from a component base class
16:03
<
github >
hpx/partitioned_vector a6749f7 Hartmut Kaiser: Merge branch 'master' into partitioned_vector
16:04
hkaiser_ has quit [Read error: Connection reset by peer]
16:04
hkaiser has joined #ste||ar
16:05
<
heller >
hkaiser: I agree, it shouldn't
16:06
hkaiser_ has joined #ste||ar
16:06
hkaiser has quit [Read error: Connection reset by peer]
16:06
parsa has quit [Read error: Connection reset by peer]
16:06
parsa| has joined #ste||ar
16:08
hkaiser_ has quit [Read error: Connection reset by peer]
16:08
hkaiser has joined #ste||ar
16:11
hkaiser has quit [Read error: Connection reset by peer]
16:14
hkaiser has joined #ste||ar
16:14
parsa has joined #ste||ar
16:16
parsa| has quit [Ping timeout: 255 seconds]
16:16
parsa has quit [Read error: Connection reset by peer]
16:16
hkaiser has quit [Read error: Connection reset by peer]
16:16
parsa has joined #ste||ar
16:18
hkaiser has joined #ste||ar
16:18
parsa has quit [Read error: Connection reset by peer]
16:20
<
heller >
denis_blank: the reset was to break cycles
16:20
parsa has joined #ste||ar
16:21
<
denis_blank >
heller: when are such cycles occuring?
16:23
hkaiser_ has joined #ste||ar
16:23
hkaiser has quit [Read error: Connection reset by peer]
16:25
parsa has quit [Read error: Connection reset by peer]
16:25
parsa| has joined #ste||ar
16:28
parsa| has quit [Read error: Connection reset by peer]
16:28
hkaiser_ has quit [Read error: Connection reset by peer]
16:28
hkaiser has joined #ste||ar
16:28
parsa has joined #ste||ar
16:32
parsa has quit [Read error: Connection reset by peer]
16:33
parsa| has joined #ste||ar
16:33
hkaiser_ has joined #ste||ar
16:33
EverYoung has quit [Ping timeout: 246 seconds]
16:33
<
hkaiser_ >
denis_blank: because futures are move-only and they are moved into the callback
16:35
hkaiser has quit [Read error: Connection reset by peer]
16:37
hkaiser_ has quit [Read error: Connection reset by peer]
16:39
hkaiser has joined #ste||ar
16:39
parsa| has quit [Read error: Connection reset by peer]
16:39
<
hkaiser >
denis_blank: yah, good catch - this needs fixing
16:40
parsa has joined #ste||ar
16:40
<
hkaiser >
K-ballo: use of a moved away tuple
16:41
<
K-ballo >
I don't see an issue with that
16:43
<
hkaiser >
see line 218
16:43
<
K-ballo >
yes, I see the full picture, I don't see an issue with it
16:43
<
hkaiser >
hmm, I do see a problem
16:44
<
denis_blank >
hkaiser: I removed it in my local copy already. I'll mention it in the upcoming PR again, so we can discuss about this in detail.
16:45
<
K-ballo >
in which situation would it be problematic?
16:45
hkaiser has quit [Read error: Connection reset by peer]
16:45
hkaiser_ has joined #ste||ar
16:45
* K-ballo
has a vague recollection of having removed this and later putting it back in
16:47
<
K-ballo >
say you call dataflow([](auto const& sf) {}, shared_future(...));, without the reset the decayed-copy of the shared future argument is artificially extended to the end of the dataflow frame lifetime
16:47
hkaiser_ has quit [Read error: Connection reset by peer]
16:47
hkaiser has joined #ste||ar
16:47
<
K-ballo >
er, I meant to say it artificially extends the lifetime of the shared state
16:48
EverYoung has joined #ste||ar
16:49
<
hkaiser >
K-ballo: if the dataflow callback takes the futures by value the reset accesses a zombie future
16:49
<
hkaiser >
isn't that UB?
16:49
<
K-ballo >
no, it's a valid but unspecified state
16:49
<
K-ballo >
in particular re-assignment is well defined
16:50
<
hkaiser >
well yah, it's re-assigned only
16:54
<
heller >
denis_blank: for shared_futures, you might get cycles
16:54
hkaiser has quit [Read error: Connection reset by peer]
16:55
<
K-ballo >
at the very least I remember strongly wanting to remove it
17:00
parsa has quit [Read error: Connection reset by peer]
17:00
parsa| has joined #ste||ar
17:02
parsa| has quit [Remote host closed the connection]
17:03
parsa has joined #ste||ar
17:03
<
denis_blank >
Mh ok, I'll leave it as it is, however I have to reset the futures inside nested packs then...
17:04
hkaiser has joined #ste||ar
17:11
parsa has quit [Read error: Connection reset by peer]
17:11
parsa| has joined #ste||ar
17:14
hkaiser has quit [Read error: Connection reset by peer]
17:14
<
heller >
K-ballo: yup. Me too
17:16
hkaiser has joined #ste||ar
17:21
bikineev has quit [Read error: No route to host]
17:22
bikineev has joined #ste||ar
17:22
hkaiser has quit [Read error: Connection reset by peer]
17:23
hkaiser has joined #ste||ar
17:26
<
K-ballo >
heller: do you remember if it breaks one of the examples or something like that?
17:26
hkaiser has quit [Read error: Connection reset by peer]
17:29
hkaiser has joined #ste||ar
17:30
parsa| has quit [Read error: Connection reset by peer]
17:30
parsa has joined #ste||ar
17:33
hkaiser_ has joined #ste||ar
17:33
hkaiser has quit [Read error: Connection reset by peer]
17:35
parsa has quit [Read error: Connection reset by peer]
17:35
parsa| has joined #ste||ar
17:36
<
github >
hpx/fixing_2847 aa5bf3e Hartmut Kaiser: Fixing C++11 compatibility issue
17:36
hkaiser_ has quit [Read error: Connection reset by peer]
17:36
hkaiser has joined #ste||ar
17:37
<
K-ballo >
we don't have a constexpr friendly HPX_ASSERT?
17:38
<
hkaiser >
K-ballo: no
17:39
parsa| has quit [Read error: Connection reset by peer]
17:39
parsa has joined #ste||ar
17:41
<
pree >
HPX_ASSERT == assert ===> 1 right ?
17:41
<
heller >
K-ballo: iirc it breaks the stencil applications
17:41
<
pree >
not a static_assert
17:43
EverYoun_ has joined #ste||ar
17:46
parsa has quit [Read error: Connection reset by peer]
17:46
parsa has joined #ste||ar
17:47
EverYoung has quit [Ping timeout: 246 seconds]
17:47
hkaiser has quit [Read error: Connection reset by peer]
17:47
hkaiser has joined #ste||ar
17:47
<
hkaiser >
heller: ok, the channel test passes now
17:52
bikineev has quit [Remote host closed the connection]
17:52
hkaiser_ has joined #ste||ar
17:53
hkaiser has quit [Ping timeout: 240 seconds]
17:54
parsa has quit [Read error: Connection reset by peer]
17:55
parsa has joined #ste||ar
17:55
EverYoun_ has quit [Remote host closed the connection]
17:56
EverYoung has joined #ste||ar
18:00
parsa has quit [Ping timeout: 255 seconds]
18:02
hkaiser_ has quit [Ping timeout: 276 seconds]
18:35
hkaiser has joined #ste||ar
18:38
hkaiser has quit [Read error: Connection reset by peer]
18:39
bikineev has joined #ste||ar
18:40
pree has quit [Ping timeout: 276 seconds]
18:41
hkaiser has joined #ste||ar
18:44
<
github >
hpx/fixing_2847 7d428a2 Hartmut Kaiser: Fixing C++11 compatibility issue
18:52
pree has joined #ste||ar
18:53
bikineev has quit [Remote host closed the connection]
18:54
<
pree >
I got it " hpx::get_os_thread_count() ".
18:55
<
pree >
whether this taken "hyper threading into account ??
18:56
bikineev has joined #ste||ar
18:59
parsa has joined #ste||ar
19:04
pree has quit [Ping timeout: 276 seconds]
19:10
<
hkaiser >
heller: yay! first builders are green!
19:13
<
github >
hpx/master 5368318 Taeguk Kwon: Fix a strange thing in parallel::util::detail::handle_local_exceptions. (Fix #2818)
19:13
<
github >
hpx/master 19453bd Hartmut Kaiser: Merge pull request #2832 from taeguk/tg_fix_2818...
19:15
parsa has quit [Ping timeout: 246 seconds]
19:17
pree has joined #ste||ar
19:24
Matombo has joined #ste||ar
19:25
denis_blank has quit [Quit: denis_blank]
20:04
pree has quit [Ping timeout: 276 seconds]
20:17
pree has joined #ste||ar
20:34
<
github >
[hpx] ShmuelLevine opened pull request #2853: Update Table 17. Non-modifying Parallel Algorithms in Documentation (master...patch-1)
https://git.io/v5UZL
20:36
eschnett has quit [Quit: eschnett]
20:46
<
github >
hpx/fixing_2847 13525b2 Hartmut Kaiser: Fixing C++11 compatibility issue
21:08
pree has quit [Ping timeout: 276 seconds]
21:21
pree has joined #ste||ar
21:26
pree has quit [Ping timeout: 276 seconds]
22:11
hendrx has quit [Ping timeout: 260 seconds]
22:11
<
heller >
hkaiser: this still leads to diagnostics when running with ubsan
22:14
bikineev_ has joined #ste||ar
22:16
bikineev has quit [Ping timeout: 248 seconds]
22:27
mcopik has quit [Ping timeout: 240 seconds]
22:28
<
heller >
hkaiser: you'll see the complete PR tomorrow, I'm still not entirely sure if it's too pedantic though
22:29
<
heller >
But failing with ubsan is not a good feat
22:34
mcopik has joined #ste||ar
22:45
jaafar_ has quit [Quit: Konversation terminated!]
22:46
jaafar has joined #ste||ar
23:08
mcopik has quit [Ping timeout: 240 seconds]
23:29
EverYoun_ has joined #ste||ar
23:32
EverYoung has quit [Ping timeout: 246 seconds]
23:33
EverYoun_ has quit [Ping timeout: 246 seconds]
23:36
EverYoung has joined #ste||ar
23:38
EverYoung has quit [Remote host closed the connection]
23:38
EverYoung has joined #ste||ar