broken smbtorture lease cases

Steve French smfrench at gmail.com
Sat Mar 14 10:22:10 MDT 2015


On Wed, Mar 11, 2015 at 7:33 PM, Michael Adam <obnox at samba.org> wrote:
> On 2015-03-11 at 18:09 -0500, Steve French wrote:
>> On Wed, Mar 11, 2015 at 6:16 PM, Michael Adam <obnox at samba.org> wrote:
<snip>
>> > Looking at the source code of the actual test
>> > only reveals half the truth.
>>
>>
>> In any case - loading the four vfs modules mentioned in Samba3.pm does
>> cause the mystery test (smb2.lease.request) to work ie
>>
>> vfs objects = acl_xattr fake_acls xattr_tdb streams_depot
>
> On a FS with proper xattr support you should be able to
> omit xattr_tdb.
>
> With proper posix acl support you should be able to omit
> fake_acls.
>
>> There isn't much else useful about leases that I could track down in
>> the test setup directories you mention though. e.g.
>> I grep -i for lease and nothing in the selftest directories and the
>> whole category (smb2.lease) is in the skip category
>
> You need to quote the whole line:
> ^samba4.smb2.lease
>
> This means that the leases subtests of the smb2 tets
> are skipped against the samba4 envs, which basically
> means the ntvfs server basically.
>
> Looking for smb2 in source{3,4}/selftest/tests.py
> gives a clou against which envs the smb2 tests are
> run and with which name. The smbd-envs are planned
> source3/selftest/tests.py under the namespace of
> source3.smb2 while source4/selftest/tests.py plans
> the smb2 torture tests against the 'dc' env (ntvfs)
> under the namespace of source4.smb2.
>
>> and nothing in the knownfail.
>
> Right, which means that the complete testsuite
> is run against the appropriated s3/smbd envs
> and expected to fully succeed. Which leads us
> to the requirements for the samba config.
>
> Hope this helps - Michael


Thanks - that is helpful.  It is very strange that '4' refers
to the obsolete target target, ie the old source4 file server, not
the location of the test, ie most torture cases are in source4.
Testing latest version of file server with latest version of
the test tools, ie source4 (smbtorture) testing source3 (bin/smbd)
is not quite as easy as "make test bin/smbd" (maybe we can
get there someday) but the normal case - where you run
smbtorture against a known server to find a bug (eg against Fedora or
Ubuntu or NAS) now makes more sense just needs
some documentation about what the configuration
prereqs would be (is there a start on documentation
somewhere for that topic).    A 'normal' distro config doesn't
need many changes, as noted above, to pass many (at
least of the lease category).


-- 
Thanks,

Steve


More information about the samba-technical mailing list