1
0
mirror of https://github.com/samba-team/samba.git synced 2025-01-25 06:04:04 +03:00

16624 Commits

Author SHA1 Message Date
Volker Lendecke
4fda68a62f r1338: A netlogon schannel failure is a normal event with XP clients. They cache the
netlogon session key and try to reconnect using that key. This fails with a
restarted smbd, we expect another serverauth2. XP falls back immediately.

Make the corresponding messages a debug level 3, not 0 to not flood log.smbd.

Volker
2007-10-10 10:52:07 -05:00
Richard Sharpe
609828444d r1333: NFS Quota support from Dan Peterson. 2007-10-10 10:52:07 -05:00
Rafal Szczesniak
659ddcc443 r1332: Typo fix. 2007-10-10 10:52:07 -05:00
Volker Lendecke
b7267121af r1330: Fix the build for systems without ldap headers 2007-10-10 10:52:07 -05:00
Volker Lendecke
49f3f0412f r1328: Fix the build 2007-10-10 10:52:06 -05:00
Jeremy Allison
4cc38b8aea r1326: Modification to get_dc_list to check negative cache. From "Joe Meadows" <jameadows@webopolis.com>.
Jeremy.
2007-10-10 10:52:06 -05:00
Jeremy Allison
fbc06831d3 r1325: Always use GetTimeOfDay() (wrapper). Ensure ldap replication
sleep time is not more than 5 seconds. Should fix issue
reported by Chris Garrigues <cwg@deepeddy.com>.
Jeremy.
2007-10-10 10:52:06 -05:00
Jelmer Vernooij
374f00b56b r1320: Return an error when the last command read from stdin fails in smbclient +
prepare for better error checking in tar..
2007-10-10 10:52:06 -05:00
Volker Lendecke
cc1881c143 r1318: Install libsmbclient into $(LIBDIR), not into hardcoded ${prefix}/lib. This
helps amd64 systems with /lib and /lib64 and an explicit configure --libdir
setting.

Thanks to Bjoern Jacke <bj@sernet.de>

Volker
2007-10-10 10:52:06 -05:00
Jeremy Allison
e5b3094c4c r1317: Patch from Joe Meadows "Joe Meadows" <jameadows@webopolis.com> to
add a timeout to the ldap open calls. New parameter, ldap timeout
added.
Jeremy.
2007-10-10 10:52:06 -05:00
Volker Lendecke
80e57d2790 r1314: Restore the 2.2 'force unknown acl user' parameter. When getting a security
descriptor for a file, if the owner sid is not known, the owner uid is set to
the current uid. Same for group sid.

This makes xcopy /o possible for files that are owned by local users/groups
(local administrators for example).

Thanks to Guenther for his persistence :-)

Volker
2007-10-10 10:52:05 -05:00
Jeremy Allison
ec614a8f7d r1311: Turn on sendfile for non-oplocked files. tridge proved
this is the correct thing to do on a train in Germany :-).
Turn on sendfile by default ! It's about time.....
Jeremy.
2007-10-10 10:52:05 -05:00
Jeremy Allison
38002bfc1c r1304: Turns out non-io opens that cause oplock breaks are a *different* set
of desired access flags from those that cause sharing violations.
Oplock breaks are caused by : ~(SYNCHRONIZE_ACCESS|FILE_READ_ATTRIBUTES|FILE_WRITE_ATTRIBUTES)
Sharing violations are caused by : ~(SYNCHRONIZE_ACCESS|READ_CONTROL_ACCESS|FILE_READ_ATTRIBUTES|FILE_WRITE_ATTRIBUTES)
We now pass the torture rename test again.
I still need to work out if subsequent opens will cause sharing violations with an existing open
mode of SYNCHRONIZE_ACCESS|READ_CONTROL_ACCESS|FILE_READ_ATTRIBUTES|FILE_WRITE_ATTRIBUTES;
Jeremy.
2007-10-10 10:52:05 -05:00
Jeremy Allison
a5a3df7853 r1301: Fix bogus error message when using "mangling method = hash" rather
than hash2. We are already calculating lengths so just use memcpy
not safe_strcpy().
Jeremy.
2007-10-10 10:52:05 -05:00
Jeremy Allison
451d289f69 r1299: Don't "break" - "continue" ! Otherwise we only expire the first name !
Doh !
Jeremy.
2007-10-10 10:52:05 -05:00
Volker Lendecke
29235f0c69 r1297: Yes, it does survive valgrind for my tests :-)
Check in the 'winbind proxy only' mode -- no new parameter required :-)

If you don't set idmap uid or idmap gid, winbind will not do idmap stuff, it
will only proxy the netlogon request and thus speed up the authentication of
domain users.

Volker
2007-10-10 10:52:05 -05:00
Volker Lendecke
c66be874d8 r1295: To be able to send a message to the background queue updated, we need to be
root. Otherwise the USR1 signal will not be delivered.

Volker
2007-10-10 10:52:04 -05:00
Jeremy Allison
e6ff6d95d2 r1290: Ensure we remove DNS and DNSFAIL records immediately on timeout.
Fix for #607.
Jeremy.
2007-10-10 10:52:04 -05:00
Volker Lendecke
43020cf459 r1287: Attempt to fix the build for systems without kerberos headers.
Volker
2007-10-10 10:52:04 -05:00
Volker Lendecke
e8786506b8 r1285: Put variable definitions into a block before the statements...
Volker
2007-10-10 10:52:04 -05:00
Volker Lendecke
3af4348faf r1283: Fix typo 2007-10-10 10:52:04 -05:00
Volker Lendecke
b491e76625 r1282: gcc 3.2 on SuSE 8.2 does not like #ifdefs inside a macro argument (DEBUG).
Volker
2007-10-10 10:52:04 -05:00
Jeremy Allison
42cdeccc36 r1263: Make "defer sharing violations" a global parameter.
Jeremy.
2007-10-10 10:52:04 -05:00
Jeremy Allison
bee0619cc4 r1260: Added new parameter in the protocol section. "defer sharing violations",
by default set to "yes" (to correctly emulate Windows). I've added this to
ensure if we find a critical problem with this new code when 3.0.5 ships
it can be turned off to test for bugs.
Jeremy.
2007-10-10 10:52:03 -05:00
Jeremy Allison
756a004311 r1259: Ensure we pass Samba4 RAW-RENAME test.
Jeremy.
2007-10-10 10:52:03 -05:00
Jeremy Allison
b52a04a5cd r1257: Ensure we deferr a sharing violation on rename correctly.
Jeremy.
2007-10-10 10:52:03 -05:00
Jeremy Allison
f17382ad8a r1255: Ensure we check attributes correctly on rename request. Gets us further with
Samba4 RAW-RENAME test.
Jeremy.
2007-10-10 10:52:03 -05:00
Jeremy Allison
c892545960 r1248: Fix from Nick Wellnhofer <wellnhofer@aevum.de> to prevent
lp_interfaces() list from being corrupted.
Jeremy.
2007-10-10 10:52:03 -05:00
Jeremy Allison
b462b8fa2f r1247: Final fix to make this compile on Heimdal.
Jeremy.
2007-10-10 10:52:02 -05:00
Jeremy Allison
d4abeefe3e r1245: I think the parameter for "password" and "data" was reversed.
CHECK THIS !
Jeremy.
2007-10-10 10:52:02 -05:00
Jeremy Allison
92a5dc1880 r1244: More Heimdal compile fixes.
Jeremy.
2007-10-10 10:52:02 -05:00
Jeremy Allison
940f893d48 r1243: Fix so this compiles with Heimdal (in Heimdal krb5_kt_cursor is a struct not a pointer).
Jeremy.
2007-10-10 10:52:02 -05:00
Jeremy Allison
a535a05975 r1241: Fix incorrect type in printer publishing (struct uuid, not UUID_FLAT).
Jeremy.
2007-10-10 10:52:02 -05:00
Jeremy Allison
464d2e9048 r1240: Ensure we don't shadow Heimdal globals.
Jeremy.
2007-10-10 10:52:02 -05:00
Volker Lendecke
fece76cfc2 r1237: Correctly honor infolevel argument. Thanks to Guenther Deschner <gd@sernet.de>.
Volker
2007-10-10 10:52:01 -05:00
Jeremy Allison
dd07278b89 r1236: Heimdal fixes from Guenther Deschner <gd@sernet.de>, more to come before
it compiles with Heimdal.
Jeremy.
2007-10-10 10:52:01 -05:00
Gerald Carter
e3d4fea780 r1230: (merges from HP PSA) fixing a couple of caching bugs in the printing code. (a) make sure to clear jobs_changed list when deleting a job and, (b) invalidate the printer handle cache when we get a notification that something has changed on that printer 2007-10-10 10:52:01 -05:00
Jeremy Allison
c378f3c325 r1224: Fix longstanding memleak bug with logfile name being set before lp_load()
loads the initial config file and overwrites the written logfile name with
zeros in init_globals(). Ensure we do a string_free() on Globals.szLogFile
if it isn't NULL.
Jeremy.
2007-10-10 10:52:01 -05:00
Jeremy Allison
5a1d8c3c9b r1223: Fix valgrind error with realm manipulation.... Damn macros :-(.
Jeremy.
2007-10-10 10:52:01 -05:00
Jeremy Allison
ad440213aa r1222: Valgrind memory leak fixes. Still tracking down a strange one...
Can't fix the krb5 memory leaks inside that library :-(.
Jeremy.
2007-10-10 10:52:00 -05:00
Jeremy Allison
286f4c809c r1221: Added the last of the system keytab patch from "Dan Perry" <dperry@pppl.gov>,
fixed valgrind detected mem corruption in libads/kerberos_keytab.c.
Jeremy.
2007-10-10 10:52:00 -05:00
Jelmer Vernooij
0ca894d5cf r1220: Revert last change. Apparently, NOTPARALLEL blocks all parallel builds 2007-10-10 10:52:00 -05:00
Jelmer Vernooij
d08ce198b5 r1219: Don't make proto parallel 2007-10-10 10:52:00 -05:00
Jeremy Allison
f0f2e28958 r1215: Intermediate checkin of the new keytab code. I need to make sure I
haven't broken krb5 ticket verification in the mainline code path,
also need to check with valgrind. Everything now compiles (MIT, need
to also check Heimdal) and the "net keytab" utility code will follow.
Jeremy.
2007-10-10 10:52:00 -05:00
Jeremy Allison
be8a2dc00d r1214: Now compiles. Changed krb5_kt_free_entry to krb5_free_keytab_entry_contents
Jeremy.
2007-10-10 10:52:00 -05:00
Gerald Carter
951a885194 r1212: small change to prevent home directories added during the SMBsesssetup&X tfrom being removed as unused services 2007-10-10 10:51:59 -05:00
Volker Lendecke
af251f4ea6 r1202: This hopefully fixes our memory use when unmarshalling strings. The test case
was 'rpcclient -c "enumprinters 2"' with 4000 printers. At some point this
completely exploded in memory usage. For every string we talloc'ed memory up
to the end of the buffer. -> O(n^2).

This survives valgrind with this number of printers. It might also have
influence on winbind with a large number of users.

All those who dare to look at samba3 rpc code, could you please take a look? I
know this is a burden, but I would like comments ;-)))

Volker
2007-10-10 10:51:59 -05:00
Jeremy Allison
7067e274dc r1195: Ensure libsmb/clikrb5.c compiles.
Jeremy.
2007-10-10 10:51:59 -05:00
Jeremy Allison
82c219ea02 r1194: Definition of krb5_free_unparsed_name() if we do't have it.
Jeremy.
2007-10-10 10:51:59 -05:00
Jeremy Allison
af5a08f5ad r1193: Ensure we check for and use krb5_free_unparsed_name().
Jeremy.
2007-10-10 10:51:59 -05:00