Fix error code on __zpl_ioctl_setflags()

[lumigch]

Other (all?) Linux filesystems seem to return -EPERM instead of -EACCESS when
trying to set FS_APPEND_FL or FS_IMMUTABLE_FL without the CAP_LINUX_IMMUTABLE
capability. This was detected by generic/545 test in the fstest suite.

Signed-off-by: Luis Henriques henrix@camandro.org

Motivation and Context

It normalizes error code across other filesystems.

Description

As mentioned in the commit changelog: simply return -EPERM instead of -EACCESS when trying to set FS_APPEND_FL or FS_IMMUTABLE_FL without the CAP_LINUX_IMMUTABLE capability.

How Has This Been Tested?

Executed a (modified) fstest suite, specifically generic/545

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Performance enhancement (non-breaking change which improves efficiency)
• Code cleanup (non-breaking change which makes code smaller or more readable)
• Breaking change (fix or feature that would cause existing functionality to change)
• Library ABI change (libzfs, libzfs_core, libnvpair, libuutil and libzfsbootenv)
• Documentation (a change to man pages or other documentation)

Checklist:

• My code follows the OpenZFS code style requirements.
• I have updated the documentation accordingly.
• I have read the contributing document.
• I have added tests to cover my changes.
• I have run the ZFS Test Suite with this change applied.
• All commit messages are properly formatted and contain Signed-off-by.

[behlendorf]

Thanks for catching this slight difference in behavior! Can you just address the style warning and then this should be good. You can run make checkstyle locally before updating the PR.

If you encounter any other unexpected discrepancies while testing with xfstests by all means let us know. Patches are even better! It's something we've wanted to add to our usual testing for a while now but haven't yet automated since it takes a little tweaking to run it against ZFS as I'm sure you know.

[lumigch]

Thanks for catching this slight difference in behavior! Can you just address the style warning and then this should be good. You can run make checkstyle locally before updating the PR.

Thanks for the review! I should have read more carefully the CONTRIBUTING.md before submitting my first fix. I've pushed a version with that style correction (TBH I assumed that change was making that if statement more readable; oh well...).

If you encounter any other unexpected discrepancies while testing with xfstests by all means let us know. Patches are even better! It's something we've wanted to add to our usual testing for a while now but haven't yet automated since it takes a little tweaking to run it against ZFS as I'm sure you know.

Yeah, it's tricky to run fstests against ZFS. I had to hack a few things and I'm sure there were a lot of tests skipped because some more hacks were missing. Is there a public tree with these tweaks available? (Obviously, I it would be nice to try to push these changes upstream.)

[behlendorf]

Is there a public tree with these tweaks available?

Nothing current that I'm aware of. It would be ideal if whatever changes are needed to support ZFS could be pushed upstream to xfstests. We could maintain our own repo with the changes, but being able to use an unmodified upstream version would be great. That would make it easier for us to integrate in to our CI workflow.