Oliver Kowalke
89dbbbdf69
fix example ping_pong
2017-02-26 11:11:37 +01:00
Oliver Kowalke
a0f27c8b04
buffered_channel's capacity must be a power of 2
...
- fixes issue #111
2017-02-25 10:27:30 +01:00
oliver Kowalke
c0b5beb971
remove unused header
2017-02-23 08:14:17 +01:00
Oliver Kowalke
bd59759fcb
add reference to skynet test
2017-02-19 19:46:15 +01:00
oliver Kowalke
d38d709205
additional comments
2017-02-16 08:15:40 +01:00
oliver Kowalke
5618d4549f
resume() -> continuation::operator()
2017-02-02 09:30:42 +01:00
oliver Kowalke
b8fa122594
rename transfer_data() -> get_data()
2017-02-01 09:34:06 +01:00
oliver Kowalke
2f3ca28af7
modify example ping-pong
2017-01-24 13:54:20 +01:00
oliver Kowalke
e87f700ded
remove tabs and non-ASCII characters
2017-01-24 08:03:43 +01:00
oliver Kowalke
b027ba3f77
fix Jamfile for doc generation
2017-01-18 10:24:00 +01:00
Oliver Kowalke
26b0a5e871
Merge pull request #107 from brandon-kohn/test/skynet_async
...
Test/skynet async
2017-01-13 07:48:31 +01:00
Brandon Kohn
114e815260
Added test entry for skynet_async.
2017-01-12 09:32:11 -05:00
Brandon Kohn
63c6427a2d
More formatting changes.
2017-01-12 09:25:09 -05:00
Brandon Kohn
dec1701e30
Updated formatting to match existing style.
2017-01-12 09:23:32 -05:00
Brandon Kohn
155ad42842
Added skynet test using async and futures
...
to compare with HPX.
2017-01-12 09:18:06 -05:00
oliver Kowalke
a8d5772ac6
supress clang warnings caused to padding
2017-01-12 09:27:27 +01:00
oliver Kowalke
20ca3705e7
pass stack-allocator to fiber in async()
...
- stack-allocator was mistakenly passed as argument to packaged_task<>
- additional overload for async() taking stack-allocator for fiber and
allocator for packaged_task<>
2017-01-12 09:11:25 +01:00
Oliver Kowalke
b5e62ab70d
callcc() -> resume()
2017-01-05 19:10:07 +01:00
Oliver Kowalke
bfb1c0a864
remove html files from doc
2016-12-27 19:08:42 +01:00
Oliver Kowalke
c8cdfefd25
implicitly declared default constructor would not be constexpr
2016-12-16 19:10:00 +01:00
oliver Kowalke
fffb7e7f32
check support for std::mutex and std::thread
2016-12-13 08:19:15 +01:00
Oliver Kowalke
f39e8e6dbc
comment added
2016-12-06 21:13:57 +01:00
Oliver Kowalke
ba5d7a4702
add copyright notice for Dmitry Vyukov
2016-11-27 11:25:15 +01:00
Oliver Kowalke
85ae668f53
documentation: support of valgrind
2016-11-24 19:26:53 +01:00
Oliver Kowalke
7843617641
fix asio integration, thx to Arnaud Kapp
2016-11-24 19:25:21 +01:00
Oliver Kowalke
5d0b7f3d84
use multiset for sleep-queue
2016-11-22 20:02:41 +01:00
Oliver Kowalke
1d5834d996
fix meta infos
2016-11-22 19:59:10 +01:00
oliver Kowalke
de58b076eb
fix asio example: scheduler-algorithm should not spin
2016-11-18 08:29:20 +01:00
oliver Kowalke
a989468064
documentation: fix perfromance numbers for std::thread
2016-11-18 07:50:09 +01:00
oliver Kowalke
648de6a665
documentation: fix performance numbers for pthreads
2016-11-17 15:13:09 +01:00
oliver Kowalke
22e9c60960
Fix typos in README.md
2016-11-17 14:41:31 +01:00
Oliver Kowalke
5d2c4b701c
make pointer to scheduler an atomic type
2016-11-17 08:17:28 +01:00
xaqq
e9092f5d0d
Fix allocator-aware fiber::async() overload allocator forwarding.
...
The fiber::async() overload which is aware of allocator doesn't use the
allocator to contruct its internal fiber object.
Signed-off-by: Oliver Kowalke <oliver.kowalke@gmx.de >
2016-11-16 08:47:50 +01:00
oliver Kowalke
df8f8ac41f
documentation: perforamnce numbers of threads
2016-11-14 08:02:09 +01:00
Oliver Kowalke
1402998cf2
add performance test with pthreads
2016-11-12 13:23:20 +01:00
oliver Kowalke
1ea326d956
documentation: reorder columns in performance table
2016-11-11 07:46:34 +01:00
Oliver Kowalke
2f03181ea9
Merge pull request #95 from nat-goodspeed/develop
...
Retitle tables; remove mu from "systems".
2016-11-10 19:43:29 +01:00
Nat Goodspeed
5bf854eff6
Retitle tables; remove mu from "systems".
2016-11-10 10:34:55 -08:00
Oliver Kowalke
6a10579e95
documentation: use micro seconds for performance measurements
2016-11-10 18:59:59 +01:00
Oliver Kowalke
9ac7608e33
Merge pull request #93 from nat-goodspeed/develop
...
Editorial pass through updated performance.qbk.
2016-11-10 18:24:41 +01:00
Nat Goodspeed
f026de1458
Editorial pass through updated performance.qbk.
...
Also a couple of tweaks to referenced sections in scheduling.qbk.
2016-11-10 07:14:39 -08:00
oliver Kowalke
e7019e98a1
add deprecated (un)bounded_channel
...
- (un)bounded_channel are deprected
- added in order to get boost.fiber released with 1.63 (breaking change)
. will be removed or renamed in next release (1.64?)
2016-11-09 08:17:14 +01:00
Oliver Kowalke
290f1ff5ed
update documentation
...
- section performance updated
- subsection tweaking
2016-11-09 08:16:28 +01:00
Oliver Kowalke
3a5dd3b657
re-factor performance tests
2016-11-05 16:30:23 +01:00
oliver Kowalke
418f6c60b2
some adjustment of ttas spinlocks
2016-11-05 15:34:59 +01:00
Oliver Kowalke
fc054f3f97
intro of BOOST_FIBERS_SPIN_SINGLE_CORE
...
- if a multi-threaded app runs on single-core systems, cpu_relax()
makes no sense
2016-11-03 20:05:06 +01:00
Oliver Kowalke
fa071c27cf
use SPMC (chase-lev algoritm) for ready-queue (work stealing)
2016-11-03 11:05:51 +01:00
Oliver Kowalke
03e11d6538
spinlocks refactored
...
- use cpu_relax() to delay (using pause and equivalent mnemonic)
- exponential backoff if lock could no be acquired
- ttas, adaptive ttas
2016-11-03 11:05:05 +01:00
Oliver Kowalke
38f1bc5945
use MPSC queue for remote-ready-queue
2016-10-30 19:41:42 +01:00
Oliver Kowalke
b53e167a68
support for channels refactored
...
- buffered_channel: MPMC with lock-free guarantees
- unbuffered_channel: rendezvous point
2016-10-30 19:33:49 +01:00
