You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@mesos.apache.org by Benjamin Bannier <be...@mesosphere.io> on 2019/02/23 20:41:00 UTC

Review Request 70046: Prevented closing invalid file descriptors.

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/
-----------------------------------------------------------

Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.


Repository: mesos


Description
-------

During the refactoring of `d838f2958e` we reorganized the code flow to
propagate failures from `os::close` and introduce code which migh
close an invalid file descriptor. While this would produce an `Error`
result from `mktemp` in any case, it would have lead to confusing
error messages in that scenario. This patch prevents closing invalid
file descriptors altogether which is consistent with the
pre-refactoring behavior.


Diffs
-----

  3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 


Diff: https://reviews.apache.org/r/70046/diff/1/


Testing
-------

`make check`


Thanks,

Benjamin Bannier


Re: Review Request 70046: Prevented closing invalid file descriptors.

Posted by Benjamin Mahler <bm...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/#review213186
-----------------------------------------------------------


Ship it!




Ship It!

- Benjamin Mahler


On Feb. 25, 2019, 9:19 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70046/
> -----------------------------------------------------------
> 
> (Updated Feb. 25, 2019, 9:19 p.m.)
> 
> 
> Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> During the refactoring of `d838f2958e` we reorganized the code flow to
> propagate failures from `os::close` and introduce code which migh
> close an invalid file descriptor. While this would produce an `Error`
> result from `mktemp` in any case, it would have lead to confusing
> error messages in that scenario. This patch prevents closing invalid
> file descriptors altogether which is consistent with the
> pre-refactoring behavior.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 
> 
> 
> Diff: https://reviews.apache.org/r/70046/diff/2/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>


Re: Review Request 70046: Prevented closing invalid file descriptors.

Posted by Benjamin Bannier <be...@mesosphere.io>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/
-----------------------------------------------------------

(Updated Feb. 25, 2019, 10:19 p.m.)


Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.


Changes
-------

Simplify code more as suggested by Ben


Repository: mesos


Description
-------

During the refactoring of `d838f2958e` we reorganized the code flow to
propagate failures from `os::close` and introduce code which migh
close an invalid file descriptor. While this would produce an `Error`
result from `mktemp` in any case, it would have lead to confusing
error messages in that scenario. This patch prevents closing invalid
file descriptors altogether which is consistent with the
pre-refactoring behavior.


Diffs (updated)
-----

  3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 


Diff: https://reviews.apache.org/r/70046/diff/2/

Changes: https://reviews.apache.org/r/70046/diff/1-2/


Testing
-------

`make check`


Thanks,

Benjamin Bannier


Re: Review Request 70046: Prevented closing invalid file descriptors.

Posted by Mesos Reviewbot <re...@mesos.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/#review213124
-----------------------------------------------------------



Patch looks great!

Reviews applied: [70046]

Passed command: export OS='ubuntu:14.04' BUILDTOOL='autotools' COMPILER='gcc' CONFIGURATION='--verbose --disable-libtool-wrappers --disable-parallel-test-execution' ENVIRONMENT='GLOG_v=1 MESOS_VERBOSE=1'; ./support/docker-build.sh

- Mesos Reviewbot


On Feb. 23, 2019, 12:41 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70046/
> -----------------------------------------------------------
> 
> (Updated Feb. 23, 2019, 12:41 p.m.)
> 
> 
> Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> During the refactoring of `d838f2958e` we reorganized the code flow to
> propagate failures from `os::close` and introduce code which migh
> close an invalid file descriptor. While this would produce an `Error`
> result from `mktemp` in any case, it would have lead to confusing
> error messages in that scenario. This patch prevents closing invalid
> file descriptors altogether which is consistent with the
> pre-refactoring behavior.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 
> 
> 
> Diff: https://reviews.apache.org/r/70046/diff/1/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>


Re: Review Request 70046: Prevented closing invalid file descriptors.

Posted by Benjamin Bannier <be...@mesosphere.io>.

> On Feb. 25, 2019, 6:06 p.m., Benjamin Mahler wrote:
> > 3rdparty/stout/include/stout/os/posix/mktemp.hpp
> > Line 47 (original), 47 (patched)
> > <https://reviews.apache.org/r/70046/diff/1/?file=2126560#file2126560line47>
> >
> >     Can we just simplify this by returning early?
> >     
> >     ```
> >       if (fd < 0) {
> >         return ErrnoError();
> >       }
> >       
> >       Try<Nothing> close = os::close(fd);
> >     
> >       // We propagate `close` failures if creation of the temp file was successful.
> >       if (close.isError()) {
> >         return Error("Failed to close '" + stringify(fd) + "':" + close.error());
> >       }
> >     
> >       return temp;
> >     ```
> >     
> >     Seems easier to read as being correct too?
> >     
> >     Also, this lets `temp` become just `std::string` without the `Try`.

Much nicer solution, agreed!

I posted an additional cleanup as https://reviews.apache.org/r/70053/.


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/#review213170
-----------------------------------------------------------


On Feb. 25, 2019, 10:19 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70046/
> -----------------------------------------------------------
> 
> (Updated Feb. 25, 2019, 10:19 p.m.)
> 
> 
> Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> During the refactoring of `d838f2958e` we reorganized the code flow to
> propagate failures from `os::close` and introduce code which migh
> close an invalid file descriptor. While this would produce an `Error`
> result from `mktemp` in any case, it would have lead to confusing
> error messages in that scenario. This patch prevents closing invalid
> file descriptors altogether which is consistent with the
> pre-refactoring behavior.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 
> 
> 
> Diff: https://reviews.apache.org/r/70046/diff/2/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>


Re: Review Request 70046: Prevented closing invalid file descriptors.

Posted by Benjamin Mahler <bm...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/#review213170
-----------------------------------------------------------




3rdparty/stout/include/stout/os/posix/mktemp.hpp
Line 47 (original), 47 (patched)
<https://reviews.apache.org/r/70046/#comment299040>

    Can we just simplify this by returning early?
    
    ```
      if (fd < 0) {
        return ErrnoError();
      }
      
      Try<Nothing> close = os::close(fd);
    
      // We propagate `close` failures if creation of the temp file was successful.
      if (close.isError()) {
        return Error("Failed to close '" + stringify(fd) + "':" + close.error());
      }
    
      return temp;
    ```
    
    Seems easier to read as being correct too?
    
    Also, this lets `temp` become just `std::string` without the `Try`.


- Benjamin Mahler


On Feb. 23, 2019, 8:41 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70046/
> -----------------------------------------------------------
> 
> (Updated Feb. 23, 2019, 8:41 p.m.)
> 
> 
> Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> During the refactoring of `d838f2958e` we reorganized the code flow to
> propagate failures from `os::close` and introduce code which migh
> close an invalid file descriptor. While this would produce an `Error`
> result from `mktemp` in any case, it would have lead to confusing
> error messages in that scenario. This patch prevents closing invalid
> file descriptors altogether which is consistent with the
> pre-refactoring behavior.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 
> 
> 
> Diff: https://reviews.apache.org/r/70046/diff/1/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>


Re: Review Request 70046: Prevented closing invalid file descriptors.

Posted by Mesos Reviewbot Windows <re...@mesos.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70046/#review213123
-----------------------------------------------------------



PASS: Mesos patch 70046 was successfully built and tested.

Reviews applied: `['70046']`

All the build artifacts available at: http://dcos-win.westus2.cloudapp.azure.com/artifacts/mesos-reviewbot-testing/2926/mesos-review-70046

- Mesos Reviewbot Windows


On Feb. 23, 2019, 8:41 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70046/
> -----------------------------------------------------------
> 
> (Updated Feb. 23, 2019, 8:41 p.m.)
> 
> 
> Review request for mesos, Benjamin Mahler and Chun-Hung Hsiao.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> During the refactoring of `d838f2958e` we reorganized the code flow to
> propagate failures from `os::close` and introduce code which migh
> close an invalid file descriptor. While this would produce an `Error`
> result from `mktemp` in any case, it would have lead to confusing
> error messages in that scenario. This patch prevents closing invalid
> file descriptors altogether which is consistent with the
> pre-refactoring behavior.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/stout/os/posix/mktemp.hpp 8dab2599f13c3e1dab109423c8a938ec16540aaf 
> 
> 
> Diff: https://reviews.apache.org/r/70046/diff/1/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>