Iox #1054 unix platform support

[elfenpiff]

Pre-Review Checklist for the PR Author

1. Code follows the coding style of CONTRIBUTING.md
2. Tests follow the best practice for testing
3. Changelog updated in the unreleased section including API breaking changes
4. Branch follows the naming format (iox-#123-this-is-a-branch)
5. Commits messages are according to this guideline
   • Commit messages have the issue ID (iox-#123 commit text)
   • Commit messages are signed (git commit -s)
   • Commit author matches Eclipse Contributor Agreement (and ECA is signed)
6. Update the PR title
   • Follow the same conventions as for commit messages
   • Link to the relevant issue
7. Relevant issues are linked
8. Add sensible notes for the reviewer
9. All checks have passed (except task-list-completed)
10. Assign PR to reviewer

Notes for Reviewer

• The platform files are all copies of the linux platform with minor adjustments
  • acl was replaced with the empty implementation version from the windows platform since they do not work with ZFS out of the box
  • IOX_UDS_SOCKET_MAX_MESSAGE_SIZE was reduced to 1024 in platform_settings.hpp
  • Add #include <unistd.h> in grp.cpp.
  • Add #define PTHREAD_MUTEX_RECURSIVE_NP PTHREAD_MUTEX_RECURSIVE and #define PTHREAD_MUTEX_FAST_NP PTHREAD_MUTEX_DEFAULT in pthread.hpp since our implementation is not POSIX conform
• Fix wrong handling of the mqd_t message queue descriptor. A message queue descriptor is NOT a file descriptor. Stated even as a warning in the posix manual! But we use it nevertheless as a filedescriptor for instance in combination with fstat or when defining an invalid version. This was corrected mainly in the message queue posix wrapper.
• Fix mixup between mode_t and oflags in the semaphore wrapper. The values are somehow compatible in linux, in FreeBSD we get luckily an invalid argument error.
• When no platform is successfully detected we always assume its Unix but with an I am feeling lucky warning.

Checklist for the PR Reviewer

• Commits are properly organized and messages are according to the guideline
• Code according to our coding style and naming conventions
• Unit tests have been written for new behavior
  • Each unit test case has a unique UUID
• Public API changes are documented via doxygen
• Copyright owner are updated in the changed files
• PR title describes the changes

Post-review Checklist for the PR Author

1. All open points are addressed and tracked via issues

References

• Closes TBD

[codecov]

Codecov Report

Merging #1055 (9571a90) into master (1db8749) will decrease coverage by 1.53%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##           master    #1055      +/-   ##
==========================================
- Coverage   77.78%   76.24%   -1.54%     
==========================================
  Files         343      343              
  Lines       12957    12953       -4     
  Branches     1864     1863       -1     
==========================================
- Hits        10078     9876     -202     
- Misses       2262     2463     +201     
+ Partials      617      614       -3     

Flag	Coverage Δ
unittests	75.47% <0.00%> (-1.53%)	⬇️
unittests_timing	12.58% <0.00%> (-0.02%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...eoryx_hoofs/source/posix_wrapper/message_queue.cpp	0.00% <0.00%> (-79.04%)	⬇️
iceoryx_hoofs/platform/linux/source/mqueue.cpp	0.00% <0.00%> (-50.00%)	⬇️
...oryx_hoofs/source/posix_wrapper/access_control.cpp	37.09% <0.00%> (-29.04%)	⬇️
...hoofs/source/posix_wrapper/posix_access_rights.cpp	67.02% <0.00%> (-6.39%)	⬇️
iceoryx_hoofs/source/posix_wrapper/named_pipe.cpp	58.54% <0.00%> (-4.15%)	⬇️
...e/iceoryx_hoofs/internal/concurrent/smart_lock.inl	92.45% <0.00%> (-1.89%)	⬇️
...fs/include/iceoryx_hoofs/internal/cxx/expected.inl	96.34% <0.00%> (-1.83%)	⬇️
iceoryx_hoofs/source/posix_wrapper/timer.cpp	63.07% <0.00%> (-0.83%)	⬇️
iceoryx_posh/source/roudi/port_manager.cpp	78.70% <0.00%> (+0.32%)	⬆️
... and 1 more

[elfenpiff]

Hint: mqd_t is not a file descriptor and therefore not convertable to an int32_t this worked just by accident.

[elfenpiff]

Hint: fstat requires as first input parameter a file descriptor. According to the posix manual mqd_t is a message queue descriptor and not a file descriptor. This was just a happy little accident that in linux there was actually a file descriptor used in the implementation.
But this could change from one linux version to another and then the whole code breaks.

In the end we exploited undefined behavior.

[elBoberido]

In this case this function could be removed or we need to find another way how to check whether the MessageQueue is outdated. On the other hand, this was a workaround to fix a startup race with RouDi and the applications and since we do not use the MessageQueue anymore, it is probably no harm to remove this.

[elfenpiff]

We already have a ticket and I added this as a task into the ticket #832

[elfenpiff]

Hint: Another bug uncovered by a unix posix system. The second parameter are the oflags and not mode_t.
By accident the mode_t values were compatible to the oflag values in this test. But who knows what they meant here.

[elBoberido]

Looks good. Just a minor issue. I wish we could use pragma once

[elBoberido]

In this case this function could be removed or we need to find another way how to check whether the MessageQueue is outdated. On the other hand, this was a workaround to fix a startup race with RouDi and the applications and since we do not use the MessageQueue anymore, it is probably no harm to remove this.