Page MenuHomeFreeBSD

Use the presence of a susp_clean method to test for suspendability.
ClosedPublic

Authored by markj on Jun 29 2015, 2:29 AM.
Tags
None
Referenced Files
Unknown Object (File)
Fri, Nov 22, 4:54 PM
Unknown Object (File)
Tue, Nov 19, 7:10 PM
Unknown Object (File)
Tue, Nov 19, 7:07 PM
Unknown Object (File)
Fri, Nov 15, 10:10 AM
Unknown Object (File)
Wed, Nov 13, 3:47 AM
Unknown Object (File)
Tue, Nov 5, 12:17 PM
Unknown Object (File)
Thu, Oct 31, 6:24 AM
Unknown Object (File)
Mon, Oct 28, 4:07 AM
Subscribers
None

Details

Summary

This change removes the MNTK_SUSPENDABLE flag and instead tests the mount
point returned by VOP_GETWRITEMOUNT. This removes the need for
passthrough filesystems like nullfs and unionfs to inherit mount flags
from lower layers.

Test Plan

Verified that vn_start_write_locked() was or was not being called under the same conditions as before, using ufs, tmpfs, nullfs on tmpfs, and devfs.

Diff Detail

Lint
Lint Passed
Unit
No Test Coverage

Event Timeline

markj retitled this revision from to Use the presence of a susp_clean method to test for suspendability..
markj edited the test plan for this revision. (Show Details)
markj updated this object.
markj added reviewers: kib, mjg.
mjg edited edge metadata.

I would possibly rename vn_suspendable_mp to vn_suspendable, but looks good enough to me as it is.

This revision is now accepted and ready to land.Jun 29 2015, 3:49 AM
kib edited edge metadata.
kib added inline comments.
sys/fs/nullfs/null_vfsops.c
202

This is weird reformatting.

In D2937#57266, @mjg wrote:

I would possibly rename vn_suspendable_mp to vn_suspendable, but looks good enough to me as it is.

I'll fix that in the commit, thanks.

sys/fs/nullfs/null_vfsops.c
202

It's a bit easier to read when the RHS is on a single line, but I agree that it looks weird. I'll fix it before committing.

This revision was automatically updated to reflect the committed changes.