----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/65469/#review197045 -----------------------------------------------------------
3rdparty/libprocess/src/subprocess.cpp Lines 259 (patched) <https://reviews.apache.org/r/65469/#comment277146> Actually, should we enable inheritance in `internal::windows::create_process` and then immediately disable it after the proccess finshes initializing? That way, we keep the handle inheritable as little as possible to avoid leaks. I think `create_process` is the only case where you actually need inheritable handles. - Akash Gupta On Feb. 2, 2018, 8:12 p.m., Andrew Schwartzmeyer wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/65469/ > ----------------------------------------------------------- > > (Updated Feb. 2, 2018, 8:12 p.m.) > > > Review request for mesos, Akash Gupta, Jie Yu, and Joseph Wu. > > > Bugs: MESOS-8512 > https://issues.apache.org/jira/browse/MESOS-8512 > > > Repository: mesos > > > Description > ------- > > On Windows, we make all file handles non-inheritable by default. > Therefore when creating a `Subprocess::FD` for a subprocess to inherit a > file descriptor, we have to explicitly make it inheritable. This was > already happening as a side-effect of `os::dup` for `IO::DUPLICATED` > file descriptors, but not for `IO::OWNED`. Both, however, are meant to > be inherited. > > > Diffs > ----- > > 3rdparty/libprocess/src/subprocess.cpp > 785e2e1083d115d25fffde2df74ed8bc1726511c > > > Diff: https://reviews.apache.org/r/65469/diff/1/ > > > Testing > ------- > > > Thanks, > > Andrew Schwartzmeyer > >