> On March 20, 2016, 2:33 a.m., Neil Conway wrote: > > 3rdparty/libprocess/3rdparty/stout/include/stout/os/posix/rmdir.hpp, line 71 > > <https://reviews.apache.org/r/45003/diff/1/?file=1304569#file1304569line71> > > > > This comment seems incorrect: > > > > (1) The relevant parameters are `FTS_PHYSICAL` vs. `FTS_LOGICAL`, as > > well as `FTS_COMFOLLOW`. > > > > (2) Since we _do_ set `FTS_PHYSICAL`, isn't it possible for > > `FTS_SLNONE` to be returned? > > Jojy Varghese wrote: > I believe FTS_LOGICAL or FTS_COMFOLLOW needs to be set in order for > FTS_SLNONE to be returned. > > Jie Yu wrote: > Can you do a simple test to validate that (see if FTS_SLNONE is hit or > not)? My understanding is that FTS_SLNONE will still be hit if we're using > FTS_PHYSICAL. > > Cong Wang wrote: > Looking at > https://sourceware.org/git/?p=glibc.git;a=blob;f=io/fts.c;h=0c5abfcdd660871d876751fba351ab25b921e88a;hb=HEAD#l901 > I think Jojy is correct here, I don't see FTS_SLNONE could be set without > FTS_LOGICAL or FTS_COMFOLLOW. > > Jojy Varghese wrote: > @jie. I tested it already and verified that FTS_SLNONE is not hit (using > logs). > > Neil Conway wrote: > Can we add a unit test anyway, please? This is corner-case behavior that > we should be validating we handle correctly, regardless of whether > `FTS_SLNONE` is actually set. > > Jojy Varghese wrote: > I have added unit tests in patch https://reviews.apache.org/r/45002/. > Would love to hear more test ideas. > > Neil Conway wrote: > I'd like a test case that covers `rmdir` of a symbolic link that points > to a non-existent file/directory; as far as I can see, the existing tests do > not cover this case.
What do you think about the test case `RemoveDirectoryWithNoTargetSymbolicLink` in the test file? - Jojy ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/45003/#review124433 ----------------------------------------------------------- On March 24, 2016, 1:10 a.m., Jojy Varghese wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/45003/ > ----------------------------------------------------------- > > (Updated March 24, 2016, 1:10 a.m.) > > > Review request for mesos, Jie Yu and Neil Conway. > > > Repository: mesos > > > Description > ------- > > Fixed rmdir comment for FTS_SLNONE as per coding guidelines. > > > Diffs > ----- > > 3rdparty/libprocess/3rdparty/stout/include/stout/os/posix/rmdir.hpp > cbc97596cd8ed1e6d4261568fd0086c86e063232 > > Diff: https://reviews.apache.org/r/45003/diff/ > > > Testing > ------- > > make check. > > > Thanks, > > Jojy Varghese > >
