----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/7001/#review11625 -----------------------------------------------------------
src/tests/cgroups_tests.cpp <https://reviews.apache.org/r/7001/#comment25091> use the same style (one arg per line) as you did in the rest of the file and in configurator_tests src/tests/cgroups_tests.cpp <https://reviews.apache.org/r/7001/#comment25092> ditto third_party/libprocess/include/stout/os.hpp <https://reviews.apache.org/r/7001/#comment25126> from the man page, i'm not sure when NULL happens but errno is not equal to 0. i think you should always return error if tree == NULL. This makes the return value of rmdir to be Try<Nothing> instead of Try<bool>. I like Try<Nothin> because its symmetrical with mkdir. third_party/libprocess/include/stout/os.hpp <https://reviews.apache.org/r/7001/#comment25127> again, similar to mkdir you should do if (::rmdir() < 0 && errno != ENOENT) {..} this way removing non-existent files/directories is not an error. third_party/libprocess/include/stout/os.hpp <https://reviews.apache.org/r/7001/#comment25128> ditto - Vinod Kone On Sept. 15, 2012, 3:16 a.m., Ben Mahler wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/7001/ > ----------------------------------------------------------- > > (Updated Sept. 15, 2012, 3:16 a.m.) > > > Review request for mesos, Benjamin Hindman, Vinod Kone, and Jie Yu. > > > Description > ------- > > We unnecessarily have Try<bool>s all over the place, these are tri-state: > {error, some:true, some:false}. It seems most cases, we never use > {some:false} in the function or the caller. > > So, this restores some sanity to use two-state Try<void>s: {error, some} > > > Diffs > ----- > > src/linux/cgroups.hpp 1a3cdc2 > src/linux/cgroups.cpp 53d611f > src/linux/fs.hpp 31a6100 > src/linux/fs.cpp 744aea6 > src/log/replica.cpp 2bba6fc > src/logging/logging.cpp d6d31ec > src/slave/cgroups_isolation_module.hpp 00255b5 > src/slave/cgroups_isolation_module.cpp 34db9af > src/slave/gc.hpp 3760d09 > src/slave/gc.cpp 5212a41 > src/slave/process_based_isolation_module.cpp c0576bd > src/slave/slave.cpp 4ea1db1 > src/tests/cgroups_tests.cpp fbaa046 > src/tests/configurator_tests.cpp 8baed76 > src/tests/files_tests.cpp 6ef2004 > src/tests/stout_tests.cpp f690fac > src/tests/zookeeper_server.hpp 4f34910 > src/webui/webui.cpp d4f2ab9 > third_party/libprocess/include/stout/os.hpp d08b0cb > third_party/libprocess/include/stout/try.hpp ec0a7b6 > third_party/libprocess/src/process.cpp 2d2b56c > > Diff: https://reviews.apache.org/r/7001/diff/ > > > Testing > ------- > > osx 10.7 gcc 4.2.1 > redhat Red Hat 4.1.2-48 gcc 4.1.2 > > make > make check > > note that SampleFrameworks.PythonFramework is consistently failing on red > hat, unrelated to this change > > > Thanks, > > Ben Mahler > >
