= Failed tests =
== samba3.smbtorture_s3.plain(dc).POSIX ==
command: /home/pelletm/src/samba-3.6.0/source3/script/tests/test_smbtorture_s3.sh POSIX //$SERVER_IP/tmp $USERNAME $PASSWORD  -l $LOCAL_PATH 2>&1 | /home/pelletm/src/samba-3.6.0/selftest/filter-subunit --fail-on-empty --prefix="samba3.smbtorture_s3.plain(dc).POSIX."
expanded command: /home/pelletm/src/samba-3.6.0/source3/script/tests/test_smbtorture_s3.sh POSIX //127.0.0.2/tmp root locals3dc2pass  -l /var/tmp/samba-3.6.0/source3/st/s3dc/share 2>&1 | /home/pelletm/src/samba-3.6.0/selftest/filter-subunit --fail-on-empty --prefix="samba3.smbtorture_s3.plain(dc).POSIX."
ERROR: Testsuite[samba3.smbtorture_s3.plain(dc).POSIX]
REASON: Exit code was 1


== samba3.smbtorture_s3.crypt(dc).POSIX ==
command: /home/pelletm/src/samba-3.6.0/source3/script/tests/test_smbtorture_s3.sh POSIX //$SERVER_IP/tmp $USERNAME $PASSWORD -e -l $LOCAL_PATH 2>&1 | /home/pelletm/src/samba-3.6.0/selftest/filter-subunit --fail-on-empty --prefix="samba3.smbtorture_s3.crypt(dc).POSIX."
expanded command: /home/pelletm/src/samba-3.6.0/source3/script/tests/test_smbtorture_s3.sh POSIX //127.0.0.2/tmp root locals3dc2pass -e -l /var/tmp/samba-3.6.0/source3/st/s3dc/share 2>&1 | /home/pelletm/src/samba-3.6.0/selftest/filter-subunit --fail-on-empty --prefix="samba3.smbtorture_s3.crypt(dc).POSIX."
ERROR: Testsuite[samba3.smbtorture_s3.crypt(dc).POSIX]
REASON: Exit code was 1


= Skipped tests =
UNKNOWN
        samba3.smbtorture_s3.plain(dc).CHAIN1
        samba3.smbtorture_s3.crypt(dc).CHAIN1

====================

Is this related to http://samba.2283325.n4.nabble.com/return-codes-standardized-td2521714.html by any chance?
-----
I've seen that the POSIX test in the samba build farm fails on some non-Linux 
platforms. One of the reasons is that different OSes return different errnos on 
posix open calls on symlinks. The torture test expects to receive 
NT_STATUS_OBJECT_PATH_NOT_FOUND from the server but the return codes of other 
Unixes are mapped to different NT status codes. I wonder where is defined that 
NT_STATUS_OBJECT_PATH_NOT_FOUND is the right bit that we should return on a 
posix open on a symlink. Do we have an RFC draft or anything like that which 
covers this? 
=====

Or is there something else going on with symlinks? These were the only two errors from the selftests on Solaris 8.