mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andrew Schwartzmeyer <and...@schwartzmeyer.com>
Subject Re: Review Request 63859: Windows: Fixed mock signal values in stout.
Date Sat, 02 Dec 2017 00:21:10 GMT


> On Nov. 30, 2017, 6:14 p.m., Andrew Schwartzmeyer wrote:
> > 3rdparty/stout/include/stout/windows.hpp
> > Lines 343-348 (original), 343-348 (patched)
> > <https://reviews.apache.org/r/63859/diff/3/?file=1901747#file1901747line343>
> >
> >     :/ I'd really like to fix whatever code is using these signals for logic. I
feel like the defining the signals for Windows was originally a band-aid, and understand this
patch didn't add them.
> >     
> >     The funny thing is that, since these values aren't defined on Windows, they
could be any number so long as only the symbol is used in the rest of the code base. I think
this is why this worked anyway.
> >     
> >     Akash, what bug did you run into that required correcting these?
> >     
> >     (And indeed, they now appear correct. SIGKILL is 9, SIGCONT is 18 in decimal,
and SIGSTOP is 19 in decimal).
> 
> Akash Gupta wrote:
>     I think they worked before because they were used only within stout. One of the docker
executor method signature is `docker->kill(ID, SIGNAL)`, which eventually calls `docker
kill -s SIGNAL ID`. Docker (and go standard library) defines the Linux signal values on Windows,
so it's expecting `docker kill -s 9`. If you want, I can fix the docker executor to ignore
the signal field on Windows and just send `docker kill` without the `-s`.

Gotcha! That makes sense. I could go either way on your proposition. It obviously works and
is supported by Docker on Windows, so it makes sense (now) to leave it. OTOH if removing it
would also let us remove these definitions, I like that too.


- Andrew


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/63859/#review192435
-----------------------------------------------------------


On Nov. 27, 2017, 9:36 a.m., Akash Gupta wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/63859/
> -----------------------------------------------------------
> 
> (Updated Nov. 27, 2017, 9:36 a.m.)
> 
> 
> Review request for mesos, Andrew Schwartzmeyer and John Kordich.
> 
> 
> Bugs: MESOS-7342
>     https://issues.apache.org/jira/browse/MESOS-7342
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Also fixed the WEXITSTATUS macro to cast the exit code instead of
> bit-masking it, since Windows exit codes are 32 bit unsigned ints.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/stout/windows.hpp 7aa0ba72c4cc3b688ce6374b3308945ea8cb7572 
> 
> 
> Diff: https://reviews.apache.org/r/63859/diff/3/
> 
> 
> Testing
> -------
> 
> See https://reviews.apache.org/r/63862/ for test results.
> 
> 
> Thanks,
> 
> Akash Gupta
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message