-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/62050/#review184411
-----------------------------------------------------------


Fix it, then Ship it!





3rdparty/libprocess/src/tests/ssl_tests.cpp
Lines 939 (patched)
<https://reviews.apache.org/r/62050/#comment260597>

    This is not necessary to trigger the original oom issue, but i guess this 
more realistically simulate the real world scenario where the server recv 
returns 0 and triggers a shutdown of the server socket.
    
    I'd add some comments to explain here. Essentially, the bug might be 
triggered if send is called after shutdown. In that case, libevent will keep 
give us BEV_EVENT_ERROR with errno=0. Internally, SSL_write returns 
SSL_ERROR_SSL (`protocol is shutdown in SSL routines SSL_write`).
    
https://github.com/libevent/libevent/blob/release-2.0.22-stable/bufferevent_openssl.c#L504


- Jie Yu


On Sept. 2, 2017, 12:38 a.m., Benjamin Mahler wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/62050/
> -----------------------------------------------------------
> 
> (Updated Sept. 2, 2017, 12:38 a.m.)
> 
> 
> Review request for mesos, Chun-Hung Hsiao and Jie Yu.
> 
> 
> Bugs: MESOS-7934
>     https://issues.apache.org/jira/browse/MESOS-7934
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> See summary.
> 
> 
> Diffs
> -----
> 
>   3rdparty/libprocess/src/tests/ssl_tests.cpp 
> 6affda8e07d76d925a87f08b297bb6ebb4e3257f 
> 
> 
> Diff: https://reviews.apache.org/r/62050/diff/1/
> 
> 
> Testing
> -------
> 
> ran in repetition
> 
> 
> Thanks,
> 
> Benjamin Mahler
> 
>

Reply via email to