Bug 9374 - Allow smb2.acls torture test to pass against smbd with a POSIX ACLs backend.
Summary: Allow smb2.acls torture test to pass against smbd with a POSIX ACLs backend.
Status: RESOLVED FIXED
Alias: None
Product: Samba 4.0
Classification: Unclassified
Component: File services (show other bugs)
Version: 4.0.0rc5
Hardware: All All
: P5 normal (vote)
Target Milestone: ---
Assignee: Karolin Seeger
QA Contact: Samba QA Contact
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-11-09 18:42 UTC by Jeremy Allison
Modified: 2020-12-11 08:01 UTC (History)
3 users (show)

See Also:

Attachments
git-am fix for 3.6.x. (8.80 KB, patch)
2012-11-13 00:33 UTC, Jeremy Allison 		jra: review? (rsharpe)
jra: review? (metze)
obnox: review+
jra: review? (ddiss)
 Details
git-am fix for 4.0.0rc.next (5.84 KB, patch)
2012-11-20 01:03 UTC, Jeremy Allison 		obnox: review+
Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Jeremy Allison 2012-11-09 18:42:18 UTC 
There are currently a small set of fixes needed for smbd to pass the smb2.acl test when back-ended onto a POSIX ACLs backend.

Patches to follow for 3.6.x and 4.0.x.

Jeremy.
Comment 1 Jeremy Allison 2012-11-13 00:33:11 UTC 
Created attachment 8186 [details]
git-am fix for 3.6.x.

With this patchset, 3.6.x passes master smbtorture raw.acls and smb2.acls (with one change to smb2.acls to make the default ACL match the Samba default ACL - i.e. the same change that is made in the smb1 raw.acls test - this didn't get ported to the smb2 ACL test).

Jeremy.
Comment 2 Richard Sharpe 2012-11-15 21:21:59 UTC 
I am starting to test this now.

My only comment so far is that for the OpenDir_fsp path to work requires that dirfd be found correctly on those platforms that it exists on.

Otherwise OpenDir_fsp falls back to SMB_VFS_OPENDIR which will, I believe, apply all the permission checks.
Comment 3 Jeremy Allison 2012-11-15 21:43:56 UTC 
Yes, this is true, but I think that's an acceptable limitation (this really is testing some edge cases). Plus your changes to autoconf (and hopefully waf) mean that we now have dirfd on both *BSD and Solaris platforms (which along with Linux are the important ones).

Jeremy.
Comment 4 Jeremy Allison 2012-11-20 01:03:37 UTC 
Created attachment 8213 [details]
git-am fix for 4.0.0rc.next

Cherry-picked from master commits:

c5ad5029fd87b36426927d57425d5debbb26394c
4985332b951d8cd46c9c0cd877875ab7839b4edb

Jeremy.
Comment 5 Jeremy Allison 2012-11-20 18:57:34 UTC 
Assigning to Karolin for inclusion in 4.0.0rc.next.

Jeremy.
Comment 6 Jeremy Allison 2012-11-21 07:09:36 UTC 
Ensuring this is logged against 4.0.0rc so it gets in the next release.

Jeremy.
Comment 7 Karolin Seeger 2012-11-21 07:42:43 UTC 
Pushed to autobuild-v4-0-test.
Waiting for review of the version for 3.6.
Comment 8 Karolin Seeger 2012-11-21 10:23:42 UTC 
Pushed to v4-0-test.
Comment 9 Karolin Seeger 2012-11-27 08:13:37 UTC 
Still waiting for review (3.6).
Comment 10 Jeremy Allison 2012-11-27 18:10:56 UTC 
Comment on attachment 8186 [details]
git-am fix for 3.6.x.

Help... someone ?

:-)

Jeremy.
Comment 11 Karolin Seeger 2012-11-28 07:57:08 UTC 
It would be good to include this into 3.6.10, which will be frozen on Monday December 3.
Comment 12 Karolin Seeger 2012-11-30 08:00:34 UTC 
Obnox, as you did the review for 4.0, could you review the 3.6 patches also, please? Thanks!
Comment 13 Michael Adam 2012-12-03 16:44:05 UTC 
Comment on attachment 8186 [details]
git-am fix for 3.6.x.

ACK
Comment 14 Michael Adam 2012-12-03 16:44:27 UTC 
Karolin, please pick for 3.6
Comment 15 Karolin Seeger 2012-12-03 19:21:39 UTC 
Pushed to v3-6-test.
Closing out bug report.

Thanks!