[Tarantool-patches] [PATCH] lua: handle fiber cancellation for fiber.cond
Oleg Babin
olegrok at tarantool.org
Fri Oct 30 09:47:53 MSK 2020
On 29/10/2020 20:14, Sergey Ostanevich wrote:
> Hi Oleg!
>
> Thanks for the comment, I believe I have to resend the patch along with
> C API changes.
>
> What I see at the moment - failures like
>
> [076] test_run:wait_upstream(1, {message_re = 'Missing %.xlog file',
> status = 'loading'})
> [076] ---
> [076] -- true
> [076] +- false
>
> which incurred by the intended behavior change, AFAIU. We no longer
> return 'true' in case the fiber is cancelled. The wait_upstream returns
> the cond wait result, hence the output differs.
>
> Did you observe any other failures?
Yes, seems something goes wrong with "applier" status. It looks like
behaviour change.
[018] xlog/panic_on_wal_error.test.lua [
fail ]
[018]
[018] Test failed! Result content mismatch:
[018] --- xlog/panic_on_wal_error.result Sun Sep 6 06:00:02 2020
[018] +++ xlog/panic_on_wal_error.reject Sun Sep 6 06:05:59 2020
[018] @@ -134,14 +134,18 @@
[018] ...
[018] while box.info.replication[1].upstream.status ~= "stopped" do
fiber.sleep(0.001) end
[018] ---
[018] +- error: '[string "while
box.info.replication[1].upstream.status..."]:1: attempt to
[018] + index field ''upstream'' (a nil value)'
[018] ...
[018] box.info.replication[1].upstream.status
[018] ---
[018] -- stopped
[018] +- error: '[string "return box.info.replication[1].upstream.status
"]:1: attempt to
[018] + index field ''upstream'' (a nil value)'
[018] ...
[018] box.info.replication[1].upstream.message
[018] ---
[018] -- 'Missing .xlog file between LSN 6 {1: 6} and 8 {1: 8}'
[018] +- error: '[string "return
box.info.replication[1].upstream.message "]:1: attempt to
[018] + index field ''upstream'' (a nil value)'
[018] ...
[018] box.space.test:select{}
Also see https://travis-ci.org/github/tarantool/tarantool/jobs/724596598
> Regards,
> Sergos
>
>
> On 28 окт 20:14, Oleg Babin wrote:
>> Hi! Thanks for the patch!
>>
>> The main reason I didn't send this patch is inconsistency between Lua and C
>> API.
>>
>> To eliminate such inconsistency [1] was implemented. Unfortunately some test
>> fails and I've failed to find root of the problem.
>>
>> However I still believe that the problem should be solved at "core" level.
>>
>> Anyway, it would be great if some solution will be pushed to upstream.
>>
>>
>> [1] https://github.com/tarantool/tarantool/commit/e7a5120603f362364953315d31aa529342263e0b
>>
>>
>> On 28/10/2020 19:22, sergos at tarantool.org wrote:
>>> From: Sergey Ostanevich <sergos at tarantool.org>
>>>
>>> Before this patch fiber.cond():wait() just returns for cancelled
>>> fiber. In contrast fiber.channel():get() threw "fiber is
>>> canceled" error.
>>> This patch unify behaviour of channels and condvars and also fixes
>>> related net.box module problem - it was impossible to interrupt
>>> net.box call with fiber.cancel because it used fiber.cond under
>>> the hood. Test cases for both bugs are added.
>>>
>>> Closes #4834
>>> Closes #5013
>>>
>>> Co-authored-by: Oleg Babin <olegrok at tarantool.org>
>>>
>>> @TarantoolBot document
>>> Title: fiber.cond():wait() throws if fiber is cancelled
>>>
>>> Currently fiber.cond():wait() throws an error if waiting fiber is
>>> cancelled like in case with fiber.channel():get().
>>> ---
>>>
More information about the Tarantool-patches
mailing list