diff options
Diffstat (limited to 'lib/kernel/test/file_SUITE.erl')
| -rw-r--r-- | lib/kernel/test/file_SUITE.erl | 56 |
1 files changed, 28 insertions, 28 deletions
diff --git a/lib/kernel/test/file_SUITE.erl b/lib/kernel/test/file_SUITE.erl index ede911c55b..f5f256a257 100644 --- a/lib/kernel/test/file_SUITE.erl +++ b/lib/kernel/test/file_SUITE.erl @@ -518,11 +518,11 @@ read_write_file(Config) when is_list(Config) -> ?line {error, enoent} = ?FILE_MODULE:read_file(""), ?line {error, enoent} = ?FILE_MODULE:read_file(''), - % Try writing to a bad filename + %% Try writing to a bad filename ?line {error, enoent} = ?FILE_MODULE:write_file("",term_to_binary(NullTerm)), - % Try writing something else than a binary + %% Try writing something else than a binary ?line {error, badarg} = ?FILE_MODULE:write_file(Name,{1,2,3}), ?line {error, badarg} = ?FILE_MODULE:write_file(Name,self()), @@ -550,9 +550,9 @@ make_del_dir(Config) when is_list(Config) -> ?line {error, eexist} = ?FILE_MODULE:make_dir(NewDir), ?line ok = ?FILE_MODULE:del_dir(NewDir), ?line {error, enoent} = ?FILE_MODULE:del_dir(NewDir), - % Make sure we are not in a directory directly under test_server - % as that would result in eacces errors when trying to delete '..', - % because there are processes having that directory as current. + %% Make sure we are not in a directory directly under test_server + %% as that would result in eacces errors when trying to delete '..', + %% because there are processes having that directory as current. ?line ok = ?FILE_MODULE:make_dir(NewDir), ?line {ok,CurrentDir} = file:get_cwd(), case {os:type(), length(NewDir) >= 260 } of @@ -1917,26 +1917,26 @@ allocate(Config) when is_list(Config) -> ok. allocate_and_assert(Fd, Offset, Length) -> - % Just verify that calls to ?PRIM_FILE:allocate/3 don't crash or have - % any other negative side effect. We can't really asssert against a - % specific return value, because support for file space pre-allocation - % depends on the OS, OS version and underlying filesystem. - % - % The Linux kernel added support for fallocate() in version 2.6.23, - % which currently works only for the ext4, ocfs2, xfs and btrfs file - % systems. posix_fallocate() is available in glibc as of version - % 2.1.94, but it was buggy until glibc version 2.7. - % - % Mac OS X, as of version 10.3, supports the fcntl operation F_PREALLOCATE. - % - % Solaris supports posix_fallocate() but only for the UFS file system - % apparently (not supported for ZFS). - % - % FreeBSD 9.0 is the first FreeBSD release supporting posix_fallocate(). - % - % For Windows there's apparently no way to pre-allocate file space, at - % least with same semantics as posix_fallocate(), fallocate() and - % fcntl F_PREALLOCATE. + %% Just verify that calls to ?PRIM_FILE:allocate/3 don't crash or have + %% any other negative side effect. We can't really asssert against a + %% specific return value, because support for file space pre-allocation + %% depends on the OS, OS version and underlying filesystem. + %% + %% The Linux kernel added support for fallocate() in version 2.6.23, + %% which currently works only for the ext4, ocfs2, xfs and btrfs file + %% systems. posix_fallocate() is available in glibc as of version + %% 2.1.94, but it was buggy until glibc version 2.7. + %% + %% Mac OS X, as of version 10.3, supports the fcntl operation F_PREALLOCATE. + %% + %% Solaris supports posix_fallocate() but only for the UFS file system + %% apparently (not supported for ZFS). + %% + %% FreeBSD 9.0 is the first FreeBSD release supporting posix_fallocate(). + %% + %% For Windows there's apparently no way to pre-allocate file space, at + %% least with same semantics as posix_fallocate(), fallocate() and + %% fcntl F_PREALLOCATE. Result = ?FILE_MODULE:allocate(Fd, Offset, Length), case os:type() of {win32, _} -> @@ -2543,9 +2543,9 @@ compress_async_crash(Config) when is_list(Config) -> ?line ok = ?FILE_MODULE:write(Fd, ExpectedData), ?line ok = ?FILE_MODULE:close(Fd), - % Test that when using async thread pool, the emulator doesn't crash - % when the efile port driver is stopped while a compressed file operation - % is in progress (being carried by an async thread). + %% Test that when using async thread pool, the emulator doesn't crash + %% when the efile port driver is stopped while a compressed file operation + %% is in progress (being carried by an async thread). ?line ok = compress_async_crash_loop(10000, Path, ExpectedData), ?line ok = ?FILE_MODULE:delete(Path), ok. |