WHATSNEW: Update changes.

Karolin
(cherry picked from commit efc59974e7b893627ad3c4a07e35f1429fc1044d)

s3: fix the fix for bug #6195 - dont let smbd child processes panic

This patch makes sure the original and temporary TDBs are closed
_before_ the rename. Originally, the open TDB was renamed, and so
the name passdb.tdb.tmp stayed around in the db context. Hence
upon client connect, the smbd children died because reinit_after_fork()
calling tdb_reopen_all() would try to reopen passdb.tdb.tmp which
existed no longer...

Michael
(cherry picked from commit 735c27c7176a9fb5b20ad02dc7cdd1ef23ff3e81)

VERSION: Raise version number up to 3.3.3.

Karolin
(cherry picked from commit 9d647667a5e4814a6e80b9d3b6507f9ccccfe9ee)

WHATSNEW: Update changes since 3.3.2.

Karolin
(cherry picked from commit 9356163d00a18c2d34e6df259e189b95fd5ad162)

Tidy up some convert_string_internal error cases, found by Andrew Bartlett.
Jeremy.
(cherry picked from commit e0f3ed481137d83087c4c6da6526eae3b9265838)

s3:build: fix build of pam_winbind.so with static linking

Static talloc libs were missing since conversion to talloc.

Michael
(cherry picked from commit 51b10a39a764e91303ee40fde1b0c4b3a6f4574c)

Ensure files starting with multiple dots are hidden
if "hide dot files" is set. Thanks to Barry Kelly <bkelly.ie@gmail.com>
for pointing this one out.
Jeremy.
(cherry picked from commit beeb86618e3af1478708d996b118856a4f9a0c0b)

Fix the problem of 3.0.x passdb databases being version
3 but using a different hash calculation than 3.2.x passwd
databases (also version 3). Introduces a minor version
number.
Jeremy.
(cherry picked from commit d30f1fc69dd1e56d46f90f7e60f13c1d383f6376)

[Bug 6228] SMBC_open_ctx failure due to path resolve failure doesn't set errno

Fixed.

It turns out there were a number of places where cli_resolve_path() was called
and the error path upon that function failing did not set errno. There were a
couple of places the failure handling code did set errno to ENOENT, so I made
them all consistent, although I think better errno choices for this condition
exist, e.g.  EHOSTUNREACH.

Derrell
(cherry picked from commit d72271908e0d67eb31fbc1d818d6f2c720bd7fbb)

Fix bug #6195 - Migrating from 3.0.x to 3.3.x can fail to update passdb.tdb correctly. For the clustering case.
Clustered setups should have only ever used
the unsigned version of TDB_DATA in the
first place so they can't be in this mess :-).
Just do the normal upgrade in the clustered case.
Jeremy.
(cherry picked from commit 58d3ec1cb81d6086d65cd12acd16cd591cf0c71f)

WHATSNEW: Update changes since 3.3.2.

Karolin
(cherry picked from commit 81d7d62c339c836ddd7053b5b3624ba36a2c0174)

Try and fix the build farm RAW-STREAMS errors. Ordering of
modules shouldn't matter, so as vfs_streams_depot doesn't
implement get/setxattrs then call into the full VFS stack
at the top.
Jeremy
(cherry picked from commit 05c038fb86555e2ee26d051ca002eeac1c1669ce)

Fix bug #6224 - nmbd waits 5 minutes at startup before checking if it needs to run elections
Fix logic bug that causes nmbd to wait 5 minutes before
looking for a master browser. This one is *old* :-). Thanks
for Simo for bugging me on this.
Jeremy.
(cherry picked from commit 15718a5d3edc3a65256533dd898ef582ded7f833)

s3: parse_packet can return NULL which is then dereferenced in match_mailslot_name
(cherry picked from commit e452955c077a761cbadd27bede8d6844edbb2e5e)

Fix a scary "fill_share_mode_lock failed" message

To me "fill_share_mode_lock failed" is a "can't happen" alert. There is
however a perfectly valid case in get_file_infos() when the file is not open.

Change the corresponding debug message to level 10 and explain more.
(cherry picked from commit 74efc72bca5707ff35df9282c843d62c4cf331c3)

s3-net: Fix Bug #6102. NetQueryDisplayInformation could return wrong information.

Guenther
(cherry picked from commit fe102f53d1ac6f888bd559eaf52b6a166d5f78d3)

Avahi disables a timer by tv=NULL in avahi_timeout_update(), do not crash
(cherry picked from commit 57301725e1c92676d3dbb62c81a41507ed1adb7b)

s3-libnet: Fix Bug #6193: avoid messing with sync_context in libnet_samsync_delta().

We absolutely need to avoid messing with the sync_context as that breaks the
stream of replication data coming from the DC (only replicates ~350 instead of
~4000 groups).

Guenther
(cherry picked from commit e3f7057b0942793543c215ab45176c4280bd7d51)
(cherry picked from commit efa7f6bff9bec89214dab368ff5efc61fa1cc212)

Fix two memleaks in the encryption code

ntlmssp_seal_packet creates its own signature data blob, which we then
have to free.

Jeremy, please check and merge appropriately (Yes, I'm asking you to do the
janitor work, I want you to *look* at this :-))

Volker
(cherry picked from commit 4d0cfb46e449e85646e05df2c4efe7dffa670edd)

s3:smbd: if we allow trans2 on the IPC$ share, then we have to allow transs2 too.

Otherwise we'll confuse the client signing engine, when we reply an error to each transs2.

metze
(cherry picked from commit 97cdf68a42bd9d5ec312151bcca9830228caeee1)

Fix bug 6097

A client sent a SID with authority 0 and 0 sub-authorities. W2k3 replies with
NT_STATUS_INVALID_SID, even if other SIDs in the list are valid.

Thanks to Pavel <wylda@volny.cz> for the bug report!
(cherry picked from commit c38c80fcd3f63e60450fd95c27d842082e8e00fd)

s3:dsgetdcname: use parentheses in if condition to make negation clear

Signed-off-by: Günther Deschner <gd@samba.org>
(cherry picked from commit 87b428e424e2e3cca975ecd0efed327e72950a1d)
(cherry picked from commit deba6c883965f99bf70744a5a4fb8124a73910fd)

Use avahi to register _smb._tcp in smbd
(cherry picked from commit b0ad52693d4ee548a2d3870e28f6499f827bed31)

Add event avahi binding
(cherry picked from commit cb18c1686b3cd20f0535458f8cee24a772560350)

Add avahi detection to configure
(cherry picked from commit dcf1df139f02d54f040e0e649f6b9bf388fdfb00)

Disable dns_sd by default
(cherry picked from commit 28abdd1ee9559bf53cbbc3372d947cbb77ea1cd5)

=?utf-8?q?=D1=953/winbind=5Fpam:=20fix=20gcc=204.4=20compile=20warning

=20(cherry=20picked=20from=20commit=20f4aec7b5907065a2f48d1b19065c6673aff6a190)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
(cherry picked from commit 86abc613f119e7685f0664033317b5d191cc88e8)

Only build the dirsort module if we have the dirfd function

Fix the build on HP/UX
(cherry picked from commit a2cb872e58e6db5fe9c5fbd575e0a4776e02b9e5)

Document the dirsort module
(cherry picked from commit a4badda87bc10a820f3591a1232d840fa4262bf3)

Use StrCaseCmp in the dirsort module
(cherry picked from commit 9ada8991505b43a81687076b35828f21598b5ca0)

Add dirsort module
(cherry picked from commit 3bf35e844cc60d911b3b2eff19db0df6348a57b5)

s3-netapi: Fix Coverity #774 (REVERSE_INULL).

Guenther
(cherry picked from commit 9bf3e129605d51483fd5ce47c90642f7e82ac102)

s3-netapi: Fix Coverity #775 (REVERSE_INULL).

Guenther
(cherry picked from commit 3b0f612c52e992cd130f5176432722bada3c64e4)

s3-netapi: Fix Coverity #776 (REVERSE_INULL).

Guenther
(cherry picked from commit d2e348b191ada5492538b7bdae1bb7cd3f639aba)
(cherry picked from commit 76c858c6fcb8111e3fd87a2a8aa2606ab06bb90a)

s3-krb5: Fix Coverity #762 (REVERSE_INULL).

Guenther
(cherry picked from commit 97190ae184dff6450b1390c854f7426e2ee3f980)
(cherry picked from commit 0f7757e94bfa35d22a7349d78578685c0dba22d7)

s3-krb5: Fix Coverity #722 (RESOURCE_LEAK).

Guenther
(cherry picked from commit 1524abd8bf12d82e1fb0063585fc9a465fc7bf9c)
(cherry picked from commit 3517388b5d5439ffe3f9629aaf826fa1dfbb4ba7)

version: fix handling of SAMBA_VERSION_VENDOR_PATCH.

We need a string version of this, or else version.c does not compile.

Michael
(cherry picked from commit f846e0fcdf62333b44a16e5a392930189b512f9b)

s3:libsmb: fix smb signing for fragmented trans/trans2/nttrans requests

Before we send the secondary requests we need to remove the
old mid=>seqnum mapping and reset cli->mid and make the new
mid=>seqnum mapping "persistent".

The bug we had in cli_send_trans was this:
The first cli_send_smb() incremented cli->mid
and the secondary requests used the incremented mid,
but as cli->outbuf still had the correct mid,
we send the correct mid to the server. The real problem
was that the cli_send_smb() function stored the seqnum
under the wrong mid.

cli_send_nttrans() was totally broken and now follows the
same logic as cli_send_trans().

The good thing is that in practice the problem is unlikely to happen,
because max_xmit is large enough to avoid secondary requests.

metze
(cherry picked from commit 880fbc4e8cd67de73c4bcda94489eb1e1422a04b)
(cherry picked from commit 70466990b4b7c68ae95dbbcf741cd3f41f2dd0b3)
(cherry picked from commit d01cca5e3ddb925696d49a1ea728013ec1032372)

fix build on old Heimdal based systems

Signed-off-by: Günther Deschner <gd@samba.org>
(cherry picked from commit 011ad7245d53a716c4c766f5ef8d317bb3a53d0f)
(cherry picked from commit fecb980bc90ba4be1fa95b4a6e2aa6967d036656)

s3-spoolss: fix notify_printer_status_byname.

This took me almost a week to find, so here a little longer explanation:

When a windows client registers printer *status* change notifies using
spoolss_RemoteFindFirstChangeNotify, it registers them to a print server handle,
not a printer handle. We were then correctly monitoring the printer status
changes but were sending out the spoolss_RouterReplyPrinterEx via the back-channel
connection with job_id set to 0 (which we only may do for monitored printer
change status notifies on printer handlers, not print server handles). Windows
was then showing a new empty dummy printer icon in the explorer as it cannot
route the notify event to the approriate local handle. It also discarded the
content of the notify event message of course. With this, printer change notify for
pausing, resuming and purging printers nicely works again here.

Jerry, Tim and all other printing gurus, please check.

Guenther
(cherry picked from commit 69a8eeb16af7cccdaa524dd85c7e2b648c0547a3)

build: fix detection of netinet/ip.h on FreeBSD

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit a3d04991ef409591a7c8b9fbb01e024bd5aaa041)
(cherry picked from commit ddcf94787f59cf682c51010b87d3624bbdd6b4a3)

build: format the header check for netinet/ip.h more nicely

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit e8aa7169a6672bd32af6c5580f6e8ec7e99511bc)
(cherry picked from commit 9baa0c3b119f41c7d65bc13f4f34c6e242ea1abc)

s3: Fix bugs in the detection of the GNU ld version (Bug #6147)

This bug results in a failure to use linker scripts to limit the set of symbols
exported by our shared libraries.

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit c6b570ce30b5cc3631c0ed780826d5450f681800)
(cherry picked from commit e1a79d063de536969732b63ed9be24c4fd2aa505)

s3:registry: tighten the subkey loop in reg_deletekey_recursive()

and loop from the end to the beginning so that we don't need
to rehash the subkeys...

This gets "net conf drop" with 2000 shares down to 14 seconds
on my box.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit db367c32756fcdf669a2ea7dad51d4daf40770e0)

s3:registry: use delete_reg_subkey() in reg_deletekey()

This further speeds up net conf drop.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 0045eb93bfb8285bcf8165b9cb2c39ade59a0782)

s3:registry: implement delete_subkey in the smbconf backend

delegating the call to the db backend

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit e574048770d8e40490f949b6e0673f1f44b6e644)

s3:registry: implement delete_subkey in the db backend

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 5ad50bd6c0f5e030105b225f15a4e8e1d5c18999)

s3:registry: add a delete_subkey method to the backend ops.

This is to provide a more atomic means of deleting a subkey of a key.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit eced96cf26b08b1fd606381eeced5abec9fac3c4)

s3:registry: refactor deletion of various subkey lists out of regdb_store_keys()

Micheal

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 4c0a996876115e2877c3903b7b9953b527175a16)

s3:registry: streamline and cleanup regdb_set_secdesc() somewhat.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 5446e47968f1c0ff2f6390e00b5483b6163be78f)

s3:registry: refactor deletion of value/secdesc/subkey list tdb records out

of regdb_store_values().

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 22c71fc6f054ffbe73346a8c20cc48b141312d95)

s3:registry: streamline reg_deletekey() somewhat.

use W_ERROR_... macros and separate assignments from checks

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit ce1051dd5498fb864e959c66b15a16d6d0cadd31)

s3:registry: use create_reg_subkey() in reg_createkey().

Instead of calling store_reg_keys()

On my box, importing 2000 shares into an empty smbconf key
now takes some 32 seconds.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit da272a52894314bd12f43fbb52a24c8b9cc024e8)

s3:registry: implement create_subkey for the smbconf backend

by delegating to the db backend

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 0b99c03824886432a86dcdc161a219a5dbad7c58)

s3:registry: implement create_subkey for the db backend.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 049d5745cbf2bba954e9a06cc0f255506805ae0f)

s3:registry: add a create_subkey method to the backend ops.

This is to provide a more atomic means of adding a subkey of a key.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 453c7e1503f814034765ebc6aac06dd11f89b381)

source/include/nt_status.h: add macro W_ERROR_NOT_OK_GOTO(x, y)

a goto destination can be specified as a second parameter.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit d316cab3cb5037909121494d57cc2173e4c66a23)

s3: move definition of W_ERROR_NOT_OK_GOTO_DONE down to nt_status.h

where all the other W_ERROR_xyz macros are found as well.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 76a460aecab6d5a03af9b8b5d97cba15d364cfde)

s3:net rpc registry: use regsubkey_ctr_init() instead of using talloc directly.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 7679a4b9179dd32599d3799fa02f42e1acfce3ef)

s3:registry: hash the list of subkeys in the regsubkey_ctr

This removes many loops over all the arrays
(from regsubkey_ctr_key_exists) and thus
reduces "net conf drop" from 1m55 to 48seconds
and "net conf import" from 1m55 to 58 seconds
for 2000 shares on my box.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 1eaff4f96768912b31795c1c8b6a4741e2e746e1)

s3:registry: remove definition of regsubkey_ctr from the surface.

All access is now through accessor functions in reg_objects.c
This allows for performance tuning under the hood in the next step.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit ed8e84dbda344b19f1c6037a630e60885f722f13)

s3:profiles: use regsubkey_ctr_init() instead of using talloc directly.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 5fc572e9e9ddc1c87ddd4622428e41d3503ccc4b)

s3:registry: use regsubkey_ctr_init() in reg_eventlog.c

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit a6c9cd899a52cf093f557ddac9a7fb42cbae35d0)

s3:services_db: use regsubkey_ctr_init() instead of using talloc directly

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 3e8393af5fc40116305af6f155de9bad203e0a8d)

s3:registry: use regsubkey_ctr_get_seqnum() in reg_backend_db.c

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit bcbe20b30017f22f1f01d9620f149fdeefe3a9e0)

s3:registry: use regsubkey_ctr_set_seqnum() in reg_backend_db.c

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 3a0f670d425a37e435fecf575730834017e30940)

s3:registry: use regsubkey_ctr_init() in reg_backend_db.c

instead of using talloc directly.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 3f2255ad1c67bcea2d812eedbb4def8c35ff110a)

s3:registry: use regsubkey_ctr_init() in reg_api.c

instead of using talloc on struct regsubkey_ctr.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit f1c0b65da5cf90740c2013e20991a87f5a16fc2c)

s3:registry: add regsubkey_ctr_get_seqnum() to hide implementation

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 811b3fefab29c109c6c134339a438c7e4ffb13fe)

s3:registry: add regsubkey_ctr_set_seqnum to hide implementation from caller.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit d119cd129c6acd2f0eb16fcf3281744d717cd291)

s3:registry: add a regsubkey_ctr_init function for allocating a regsubkey_ctr

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit b26f384649b4285748bf0fb78d826651882208e7)

s3:registry: fix a comment

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 391027d79eeb897c539d35877f5b27ea90efb40b)

s3:registry: don't directly access key->subkeys->subkeys[] in reg_backend_db.c

Use the reg_objects accessor regsubkey_ctr_specific_key() instead.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 078d121b4d4d74237ea11c39754fb2ba0dc496a9)

s3:registry: don't directly access key->subkeys->subkeys[] in reg_api.c

Use the reg_objects accessor regsubkey_ctr_specific_key() instead.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit c861f55a35e315f3561bc9b5a4d4d2e1b782fa31)

s3:registry: don't directly access key->subkeys->num_subkeys in reg_backend_db.

Use the reg_objects api instead.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 128f91d4c26dd4635a88c4bc8939061300cf81b7)

s3:registry: don't directly access key->subkeys->num_subkeys in reg_api.c

Use the reg_objects api instead.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 6a9c73e5c58837c1a6e0abc03e90e616520cdd91)

s3:registry: replace typedef "REGSUBKEY_CTR" by "struct regsubkey_ctr"

This paves the way for hiding the typedef and the implementation from the
surface.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 9ca48a86fb6cf5fa30b622162877859b0c335f7e)

Revert "Fix a O(n^2) algorithm in regdb_fetch_keys()"

This reverts commit a13f065bad0f4d21a67e68b743f17f45bf0a4691.

This fix is reverted, because the speedup is going to move
further down into reg_objects.c. The unsorted list of subkey names
is going to be indexed: This O(n^2) search bites us in more places.
This re-establishes the abstraction of reg_objects.c.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit abc457fe4c167bfcd7c8a4a654e13884b8e1ef90)

s3:net conf: reduce memory usage of "net conf import".

"net conf import" was wrapped in one big transaction.
This lead to MAX_TALLOC_SIZE being exceeded at roughly
1500 shares. This patch resolves that problem by
limiting the top level transactions in "net conf import"
to 100 shares.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 026b72903c852e46012ac60d9d42b14c6860a159)

s3:net: wrap net conf import into one big transaction

This speeds up "net conf import" of a file with 2000 shares
from 11 minutest to 1m50s on my box.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit e95bcaefc98f4ddc461d60d2f6f812ef24c2f234)

s3:libsmbconf: add transactions to the libsmbconf api

This is useful for wrapping higher level aggregate operations
in transactions. The text backend implementations just return
WERR_OK, the registry backend implementatoins use the
regdb_transaction_start|commit|cancel routines just added.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 8dd46903944c48e171837729cbe937f54d9554b7)

s3:registry: wrap deletekey_recursive in one big transaction.

This speeds up "net conf drop" with 2000 shares on my box from
4m40s to 1m50s, leaving virtually only cpu load.

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 8353d9b764ec90948fb8a5cb39689ea39acc7f9f)

s3:registry: provide transaction_start|commit|cancel fns for the registry tdb

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 0509eee47ac2da1042bd1bb0d98695abc8ee44a3)

Wrap creating the sorted subkey cache in a transaction

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 4ddeed513ce5e2bb839570212d046145da70768b)

Add a comment describing the sorted subkeys

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit caa654d31fc65e3a08327e430ba5332776a21f11)

Add sorted subkey cache

On my box this gets net conf list for 1000 records down to .1 seconds

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 32524d865158080421867b36f9b64bdc98ff07c9)

Add db_tdb_parse

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit e5936d910cfc3590a3b015e627a5fa308861eaf4)

Add dbwrap->parse_record

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 68d4e25aad23b670e4ff735377de59ba396940cd)

s3:dbwrap_ctdb_marshall_add: don't leak the ctdb_rec_data to the outside

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit a470af9a18b328548f241dca4f744d25c7cdc13e)

s3:smbconf: move smbconf_share_exists checks into backend

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 11ed4d41906084a604797e761ffd27caf5c67ac3)

Speed up "net conf list"

For 1000 shares this speeds up net conf list from .6 to .25 seconds on my box

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit f8fa246bbae0db96e16e8f3e327577920c6ecec5)

Speed up "net conf list"

With 1000 shares in the registry, this changed the time of "net conf list" from
1.1 seconds to .6 seconds.

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit c4a1fc734b70efa932df4d4f8d9c66ea4f83ad0b)

Make use of gpfs_get_real_filename optional

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 36f2aa636d1eb5394469c09f76fe2fb676dfcac7)

s3: Fix a memleak in dbwrap_rbt.

The SMB_MALLOC'ed rbt node data was not free'd on talloc free of
the db context. This is a quick fix using talloc instead of malloc
for allocation of the node data.

Since malloc was originally used for performance reasons, one
might want to reverse to malloc and create a talloc destructor
that walks the tree and frees all the node data if this talloc
approach proves to be too slow..

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 6facbbb4020369c18b4312115fab9b8d6eceaf80)

s3:smbconftort: don't use reserved words ('test:', 'failure:', 'success:')

Temporary results printfs should not contain reserved subunit words.

metze

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit d5750eca159e0bf1095b102545838d771f26a7f7)

s3:libsmbconf: include smbconf headers with lib/smbconf prefix

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 5887fa35156df68416b4da4a45ef0576d8b203c1)

s3:libsmbconf: use talloc_free instead of TALLOC_FREE in txt backend

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 8589fa2cfe8fb549b013d2f15a598e43e28b64d7)

s3:libsmbconf: use talloc_free instead of TALLOC_FREE in smbconf_init

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit dc0583613dfd8ce8f9209a8fb56e6c41a14efd0a)

s3:libsmbconf: use talloc_free instead of TALLOC_FREE in testsuite

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit 9dfc6c9de92e58173e73d67a6dd56b4b4336909b)

s3:libsmbconf: use talloc_free instead of TALLOC_FREE in smbconf_util.c

Michael

Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit a077a710b9d402ed7ac86ab98dc2a98776b8f9bc)