> On March 19, 2016, 10:10 p.m., Jie Yu wrote: > > src/linux/xfs.cpp, line 53 > > <https://reviews.apache.org/r/44946/diff/3/?file=1304831#file1304831line53> > > > > Instead of relying on parameter, can we use os::stat::isdir here? > > James Peach wrote: > ``isdir`` always follows symlinks. From the caller of this we always make > sure what kind of inode we have so we don't need to ``stat(2)`` again and > potentially get it wrong. > > Jie Yu wrote: > From the API perspective, i just feel it strange that we need to pass > around 'isdir' which we can always derive. > > Regarding the symlink semantics, we should clearly document the semantics > for those public functions (i.e., do we follow symlink or not). > > I am wondering if we can do a symlink (`os::stat::islink`) check in those > public functions. For instance, return Nothing if 'path' passed to > clearProjectId is a symlink.
Yeh I can do this. We'll just end up ``stat(2)``ing the path multiple times at different layers. - James ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/44946/#review124427 ----------------------------------------------------------- On March 21, 2016, 6:30 p.m., James Peach wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/44946/ > ----------------------------------------------------------- > > (Updated March 21, 2016, 6:30 p.m.) > > > Review request for mesos, Jie Yu and Jiang Yan Xu. > > > Bugs: MESOS-4828 > https://issues.apache.org/jira/browse/MESOS-4828 > > > Repository: mesos > > > Description > ------- > > Add utility functions to manipulate XFS project quotas. > > > Diffs > ----- > > src/Makefile.am 9dd21b56af0500f7125b07bf535b45fe5c544aaf > src/linux/xfs.hpp PRE-CREATION > src/linux/xfs.cpp PRE-CREATION > > Diff: https://reviews.apache.org/r/44946/diff/ > > > Testing > ------- > > Make check. Manual verification. Tests in subsequent patches. > > > Thanks, > > James Peach > >
