Commit Graph

456 Commits

Author SHA1 Message Date
Eugene Kabanov 57ebe84d17
Add Defect check for arguments `pbytes` and `nbytes`. (#128)
* Add Defect check for arguments `pbytes` and `nbytes`.

* Allow some edge cases to be not Defect.
2020-09-10 03:50:06 +03:00
Eugene Kabanov 126ea4bc56
Fix newFuture[T] location source for generated async procedures. (#125)
Add tests to for locations.
2020-09-03 14:37:53 +03:00
Eugene Kabanov a5442edfc0
Add asyncSpawn() procedure and tests. (#123)
Deprecated asyncDiscard() procedure.
Bump version to 2.5.2.
2020-09-01 21:41:18 +03:00
Zahary Karadjov e6d50b7736 Revert "some metrics for monitoring futures (#85)"
This reverts commit e45ef32b5b.

Metrics implemented this way, with a lock inside the otherwise tight
event loop are not consistent with the chronos architecture that for
good or bad uses thread local variables to avoid them - the solution
does not have rough consensus behind it, and other avenues should be
explored for this generally useful functionality.
2020-08-16 01:48:25 +03:00
Zahary Karadjov 826d48c4aa Revert "more docs (#111)"
This reverts commit d0a17d551f.

Moving CancelledError outside of the established Nim hierarchy is
not a solution that has rough consensus and has an unknown impact
on compatibility with otherwise correctly implemented cancellation
code (for example when `CatchableError` is caught, cleanup is done,
then the exception is reraised). Further, this breaks the established
convention in the Nim community that Exception should not be inherited
from, complicating compatibility with future Nim versions that may
enforce this more strongly.
2020-08-16 01:48:25 +03:00
Zahary Karadjov 03f4a26829 Revert "yieldAsync() (#120)"
This reverts commit 284d677815.

There is no rough consensus for how this should be implemented and
whether it's needed indeed - a use case that cannot be solved with
a queue or a lock/event should be identified before pursuing this
functionality.
2020-08-16 01:48:25 +03:00
Zahary Karadjov 34b81961d3 Revert "asyncCheck(): don't raise exception on cancellation (#121)"
This reverts commit baf33f04ea.

This change hides exceptions that should be handled explicitly -
asyncCheck is a debugging tool that should only be used in
exception-free situations (indeed, to fail when the operation
unexpectedly raises). There is no rough consensus behind this
change.
2020-08-16 01:48:25 +03:00
Ștefan Talpalaru baf33f04ea
asyncCheck(): don't raise exception on cancellation (#121) 2020-08-12 16:24:55 +02:00
Ștefan Talpalaru 284d677815
yieldAsync() (#120) 2020-08-10 15:31:21 +02:00
Ștefan Talpalaru d0a17d551f
more docs (#111)
* more docs

* make CancelledError and object of Exception
2020-08-06 20:52:50 +02:00
Ștefan Talpalaru e45ef32b5b
some metrics for monitoring futures (#85) 2020-08-06 19:30:53 +02:00
zah 0d4d0002b0
Produce an error when the result variable is used in void async procs (#117) 2020-08-05 16:28:11 +03:00
Eugene Kabanov 73de13ca7a
Fix SIGSEGV while consuming errors more then once. (#114) 2020-08-05 00:23:43 +03:00
Ștefan Talpalaru f5634f499d
cancelAndWait(): add comments (#112)
The cancellation process has a very complicated control flow, split over
different files and hidden by macros. These comments should help shed a
light on this madness.
2020-07-30 02:47:11 +02:00
cheatfate 3968f09ae1
Skip flaky test on MacOS. 2020-07-17 09:38:58 +03:00
Eugene Kabanov 31fec25063
Fix accept races. (#110)
* Fix accept races.
2020-07-15 11:09:34 +03:00
cheatfate 783f84aa4b
Fix EAGAIN issue in acceptLoop(). 2020-07-15 01:07:03 +03:00
Ștefan Talpalaru 63041b2d8f
start documenting the library (#109)
* start documenting the library

* introduce futures

* running the event loop

* async procs and "await"

* more "await" examples

* clarify internal callback creation

* error handling

* address review comments

* remove TODO item

* more about future completion
2020-07-13 17:59:11 +02:00
Jacek Sieka f856c885fa
fix endian conversion issues (#82)
* fixes call to `bigEndian32` on a uint64 which breaks on big endian
platforms
* prefer endians2 for less and safer code
2020-07-12 18:22:47 +02:00
Jacek Sieka 8b8a1e793d
bump supported nim version 2020-07-12 18:20:21 +02:00
Ștefan Talpalaru 0ad7321aa0
ServerFlags.ReusePort support on POSIX (#95) 2020-07-12 18:13:22 +02:00
Eugene Kabanov ce6e7d17b1
Make Future tracking and stack traces optional (#108)
* Make Future tracking optional via -d:chronosDutureTracking compilation flag.
* Stack traces is now optional, use -d:chronosStackTraces.
* Fix mistypes and add test for chronosStackTrace option.
2020-07-08 19:48:01 +03:00
Eugene Kabanov 5629b3c41f
[WIP] Zero-cost unattended Future[T] tracking mechanism. (#106)
* Zero-cost unattended Future[T] tracking mechanism with tests and tracking of test suite.
2020-07-06 09:33:13 +03:00
Eugene Kabanov 16ed169f25
Fix cancellation race when low level futures are already completed, while cancellation process is pending. (#107)
Added test.
2020-07-03 15:03:59 +03:00
Eugene Kabanov 528688d01e
Add sanity check. (#105) 2020-06-25 09:32:41 +03:00
Eugene Kabanov 319e2bfc09
Fix Nim's issue #13899 using #14723 and add tests. (#104) 2020-06-24 13:03:36 +03:00
Eugene Kabanov 02b8da986b
Add accept() call (#103)
* Add accept() call and tests.
* Fix rare fd leaks on Windows.
* Fix compilation warnings.
* Add fd leak test.
* Bump version to 2.4.0.
2020-06-24 11:21:52 +03:00
Eugene Kabanov bedd1ded5e
Properly fix cancellation race and not introduce FD leaks. (#102) 2020-06-05 19:11:51 +03:00
Eugene Kabanov d6d0084333
Fix cancellation race when connect() call is getting cancelled. (#101) 2020-06-04 13:07:52 +03:00
Kim De Mey bce0f878d1
Remove unreachable raise from initTAddress (#97) 2020-05-15 14:09:21 +03:00
Jacek Sieka aff69aaf29
fix closed error type (#99) 2020-05-13 22:45:40 +03:00
cheatfate 357baa52a0
Fix documentation for `or`. 2020-04-21 07:33:31 +03:00
Eugene Kabanov 2ecc5500c2
Undeprecate `or` operation. (#93)
* Undeprecate `or` operation.
Fix `or` for already finished futures.
Add tests.

* Bump version to 2.3.9.
2020-04-21 07:07:49 +03:00
Giovanni Petrantoni 9ea1017a06
Fix possible race condition in asyncstream close (#91) 2020-04-09 17:56:56 +03:00
Eugene Kabanov 3d745a4b0c
Fix Nim's issue #13889 https://github.com/nim-lang/Nim/issues/13889. (#90) 2020-04-06 15:49:09 +03:00
Eugene Kabanov 4e2810cfe0
Fix issue with allFinished(), allFutures(), one() behavior when Futures passed are already finished. (#89)
Added test.
2020-04-06 13:56:24 +03:00
Eugene Kabanov cbd8e03823
Add allFinished() primitive. (#87)
* Add allCompleted() primitive.

* Rename it to allFinished().

* Fix allFinished Future's static name.
Use `mitems()` instead of `for`.
2020-04-01 12:10:56 +03:00
andri lim 72b4f14427
fix nim 1.2 compilation issue on linux 2020-03-26 16:46:31 +07:00
Araq f3827a13d1 prepare nim-chronos for Nim version 1.2 2020-03-23 20:17:29 +02:00
cheatfate 7ed9f1431a
Bump version to 2.3.8. 2020-03-05 20:54:57 +02:00
Jacek Sieka d8f8e3d9fc
readMsg branch by @arnetheduck with some changes. (#83)
Co-authored-by: Eugene Kabanov <ka@hardcore.kiev.ua>
2020-03-05 10:59:10 +01:00
cheatfate 56fbdff096
Async transformed procedures will not catch Exception anymore.
Bump version to 2.3.7.
2020-03-03 13:42:43 +02:00
Eugene Kabanov fe1b1a6983
Merge pull request #79 from status-im/dgramanyaddr
Fix Windows datagram's AnyAddress issue.
2020-02-26 00:09:51 +02:00
cheatfate 64583b4269
Fix Windows datagram's AnyAddress issue.
Add test for datagram's AnyAddress.
Bump version to 2.3.6.
2020-02-25 23:50:39 +02:00
Eugene Kabanov 3367b6ca90
Merge pull request #75 from status-im/fix-future-defect
FutureDefect is a type of Defect
2020-02-18 05:54:55 +02:00
Zahary Karadjov a9ed7e4f3f
FutureDefect is a type of Defect 2020-02-18 00:35:03 +02:00
cheatfate 8e70d1be7a
Fix macos warnings. 2020-02-12 23:01:41 +02:00
cheatfate 80351cb928
Fix #73. 2020-02-12 22:54:05 +02:00
Eugene Kabanov e3ced62d4b
Merge pull request #69 from status-im/fix-64
Fix #64
2020-01-28 16:08:56 +02:00
cheatfate 4a2416fe5e
Allow to setup datagram servers with 0 port. 2020-01-28 15:46:57 +02:00