mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Benjamin Bannier" <benjamin.bann...@mesosphere.io>
Subject Re: Review Request 40326: libprocess: Marked Boost as a "system" header.
Date Mon, 16 Nov 2015 08:02:08 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/40326/#review106604
-----------------------------------------------------------

Ship it!



3rdparty/libprocess/3rdparty/Makefile.am (line 212)
<https://reviews.apache.org/r/40326/#comment165331>

    I think the emphasis is slightly off here; we aren't working around noisy warnings (which
are all valid), but the fact that these warnings are useless to us as we have no intention
of fixing them. I'd also drop the ref to the jira issue since it can be reached via the commit
message.
    
    How about:
    
    > We use "-isystem" instead of "-I" to add Boost to the include search path. This disables
compiler warnings inside Boost headers since we have no intention of fixing those.



3rdparty/libprocess/3rdparty/Makefile.am (line 216)
<https://reviews.apache.org/r/40326/#comment165332>

    I think it would make sense to also update the mesos `CMakeLists.txt`.



3rdparty/libprocess/Makefile.am (line 89)
<https://reviews.apache.org/r/40326/#comment165329>

    I think the emphasis is slightly off here; we aren't working around noisy warnings (which
are all valid), but the fact that these warnings are useless to us as we have no intention
of fixing them. I'd also drop the ref to the jira issue since it can be reached via the commit
message.
    
    How about:
    
    > We use "-isystem" instead of "-I" to add Boost to the include search path. This disables
compiler warnings inside Boost headers since we have no intention of fixing those.



3rdparty/libprocess/Makefile.am (line 92)
<https://reviews.apache.org/r/40326/#comment165335>

    I think it would make sense to also update the mesos `CMakeLists.txt`.



3rdparty/libprocess/configure.ac (line 207)
<https://reviews.apache.org/r/40326/#comment165330>

    I think the emphasis is slightly off here; we aren't working around noisy warnings (which
are all valid), but the fact that these warnings are useless to us as we have no intention
of fixing them. I'd also drop the ref to the jira issue since it can be reached via the commit
message.
    
    How about:
    
    > We use "-isystem" instead of "-I" to add Boost to the include search path. This disables
compiler warnings inside Boost headers since we have no intention of fixing those.


- Benjamin Bannier


On Nov. 15, 2015, 10:30 a.m., Neil Conway wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/40326/
> -----------------------------------------------------------
> 
> (Updated Nov. 15, 2015, 10:30 a.m.)
> 
> 
> Review request for mesos, Benjamin Bannier and Joris Van Remoortere.
> 
> 
> Bugs: MESOS-3799
>     https://issues.apache.org/jira/browse/MESOS-3799
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This instructs the compiler to ignore warnings in the Boost headers. This is
> useful, because those headers often contain hard-to-silence warnings that are
> not otherwise useful to display. See MESOS-3799.
> 
> 
> Diffs
> -----
> 
>   3rdparty/libprocess/3rdparty/Makefile.am 0adbe539afaf683e4a85582463a2930049a63998 
>   3rdparty/libprocess/Makefile.am cdefa37528ea69422978a9772f955042e882dde4 
>   3rdparty/libprocess/configure.ac 40801653a7fb9a943dfe33913161d28ef24040c3 
> 
> Diff: https://reviews.apache.org/r/40326/diff/
> 
> 
> Testing
> -------
> 
> Re-bootstrapped and tested compilation with Ubuntu Wily and OSX 10.10.
> 
> Note that the cmake build still emits boost-related warnings: I didn't try to fix this,
since cmake emits other warnings that autotools suppresses (e.g., no-unused-local-typedefs).
> 
> 
> Thanks,
> 
> Neil Conway
> 
>


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