mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Qian Zhang <zhang...@cn.ibm.com>
Subject Re: Review Request 53041: Added `setxattr()` and `getxattr()` in stout.
Date Fri, 21 Oct 2016 05:57:13 GMT


> On Oct. 20, 2016, 12:03 p.m., James Peach wrote:
> > 3rdparty/stout/include/stout/os/posix/xattr.hpp, line 24
> > <https://reviews.apache.org/r/53041/diff/1/?file=1541992#file1541992line24>
> >
> >     Since the flags differ between platform, it might be a good idea to define stout-specific
versions.

Can you elaborate on stout-specific versions? Did you mean we define two versions of setxattr()
here, one for Linux and another for APPLE?


> On Oct. 20, 2016, 12:03 p.m., James Peach wrote:
> > 3rdparty/stout/include/stout/os/posix/xattr.hpp, line 51
> > <https://reviews.apache.org/r/53041/diff/1/?file=1541992#file1541992line51>
> >
> >     Add a flags argument? Darwin's ``XATTR_NOFOLLOW`` can be mapped to ``lgetxattr``
on Linux.

Did you mean if caller passes `XATTR_NOFOLLOW` as flag to this method on Linux, internally
we call `lgetxattr()`? But I think it may confuse the caller since on Linux `getxattr()` does
not support flags at all, so I would suggest not to mix `getxattr()` and `lgetxattr()` in
one method.


> On Oct. 20, 2016, 12:03 p.m., James Peach wrote:
> > 3rdparty/stout/include/stout/os/posix/xattr.hpp, line 64
> > <https://reviews.apache.org/r/53041/diff/1/?file=1541992#file1541992line64>
> >
> >     You could use ``std::vector<char>`` here to avoid manual memory management.

Can you please let me know the advantage of std::vector<char>? To me, new + memset can
satisfy my requirement here.


> On Oct. 20, 2016, 12:03 p.m., James Peach wrote:
> > 3rdparty/stout/include/stout/os/posix/xattr.hpp, line 80
> > <https://reviews.apache.org/r/53041/diff/1/?file=1541992#file1541992line80>
> >
> >     Should you add ``removexattr`` for completeness?

Agree!


> On Oct. 20, 2016, 12:03 p.m., James Peach wrote:
> > 3rdparty/stout/include/stout/os/xattr.hpp, line 21
> > <https://reviews.apache.org/r/53041/diff/1/?file=1541994#file1541994line21>
> >
> >     I don't think this API makes sense for Windows. NTFS has streams, which have
different semantics that extended attributes. I doubt that you would want this API on Windows.

Yes, I agree, let me mark those functions as delete on Windows.


- Qian


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/53041/#review153347
-----------------------------------------------------------


On Oct. 20, 2016, 11:20 a.m., Qian Zhang wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/53041/
> -----------------------------------------------------------
> 
> (Updated Oct. 20, 2016, 11:20 a.m.)
> 
> 
> Review request for mesos and Jie Yu.
> 
> 
> Bugs: MESOS-6360
>     https://issues.apache.org/jira/browse/MESOS-6360
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Added `setxattr()` and `getxattr()` in stout.
> 
> 
> Diffs
> -----
> 
>   3rdparty/stout/include/Makefile.am 1eb9c146d3eebacaea32e0f724d145f0c1dd676f 
>   3rdparty/stout/include/stout/os.hpp 96e8621b198a3ec4cce78e0a6ff5f271eda05ff1 
>   3rdparty/stout/include/stout/os/posix/xattr.hpp PRE-CREATION 
>   3rdparty/stout/include/stout/os/windows/xattr.hpp PRE-CREATION 
>   3rdparty/stout/include/stout/os/xattr.hpp PRE-CREATION 
> 
> Diff: https://reviews.apache.org/r/53041/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Qian Zhang
> 
>


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