mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alexander Rukletsov <ruklet...@gmail.com>
Subject Re: Review Request 64686: Fixed the type-punned pointer and strict aliasing issue.
Date Fri, 22 Dec 2017 13:41:35 GMT


> On Dec. 21, 2017, 8 p.m., Benjamin Bannier wrote:
> > src/linux/ns.cpp
> > Line 397 (original), 397-399 (patched)
> > <https://reviews.apache.org/r/64686/diff/2/?file=1926225#file1926225line397>
> >
> >     Why do we put `struct` here? This is C++ (and the `memcpy` fits on a single
line w/o it).
> >     
> >     Not yours, but we should probably also include `sys/socket.h` for `ucred` in
this file.

See https://reviews.apache.org/r/64807/


> On Dec. 21, 2017, 8 p.m., Benjamin Bannier wrote:
> > src/linux/ns.cpp
> > Line 455 (original), 459 (patched)
> > <https://reviews.apache.org/r/64686/diff/2/?file=1926225#file1926225line459>
> >
> >     `struct`?

See https://reviews.apache.org/r/64807/


> On Dec. 21, 2017, 8 p.m., Benjamin Bannier wrote:
> > src/linux/ns.cpp
> > Lines 472 (patched)
> > <https://reviews.apache.org/r/64686/diff/2/?file=1926225#file1926225line472>
> >
> >     `struct`?

See https://reviews.apache.org/r/64807/


- Alexander


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64686/#review194361
-----------------------------------------------------------


On Dec. 22, 2017, 1:39 p.m., Alexander Rukletsov wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/64686/
> -----------------------------------------------------------
> 
> (Updated Dec. 22, 2017, 1:39 p.m.)
> 
> 
> Review request for mesos, Armand Grillet, Benjamin Bannier, and Michael Park.
> 
> 
> Bugs: MESOS-6616
>     https://issues.apache.org/jira/browse/MESOS-6616
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Dereferencing a pointer cast from a different type of pointer
> violates the so-called “strict aliasing” rule, which is undefined
> behaviour and might lead to bugs when compiler optimizations are
> enabled.
> 
> For more information on this topic, see
> https://blog.regehr.org/archives/959
> http://alas.matf.bg.ac.rs/manuals/lspe/snode=153.html
> 
> 
> Diffs
> -----
> 
>   src/linux/ns.cpp 5e2df1ed56432c6c1bfa04a31dba7f9f547d6139 
> 
> 
> Diff: https://reviews.apache.org/r/64686/diff/2/
> 
> 
> Testing
> -------
> 
> make check on Mac OS 10.11.6 and various Linux distros
> 
> 
> Thanks,
> 
> Alexander Rukletsov
> 
>


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