kai/samba.git
14 years agor21563: Fix a memleak: We only need dispinfo structs for "our" and for the builtin
Volker Lendecke [Tue, 27 Feb 2007 17:21:21 +0000 (17:21 +0000)]
r21563: Fix a memleak: We only need dispinfo structs for "our" and for the builtin
domain. Without this patch we leaked a DISPINFO for the (NULL) domain per
samr_connect*() call.

Volker

14 years agor21562: Regenerate gen_ndr after pidl changes.
Jelmer Vernooij [Tue, 27 Feb 2007 17:17:16 +0000 (17:17 +0000)]
r21562: Regenerate gen_ndr after pidl changes.

14 years agor21561: It makes absolutely no sense to call krb5_kt_resolve() two times
Günther Deschner [Tue, 27 Feb 2007 14:24:49 +0000 (14:24 +0000)]
r21561: It makes absolutely no sense to call krb5_kt_resolve() two times
directly after another.

Guenther

14 years agor21560: Convert name_to_fqdn to BOOL.
Günther Deschner [Tue, 27 Feb 2007 13:42:15 +0000 (13:42 +0000)]
r21560: Convert name_to_fqdn to BOOL.

Guenther

14 years agor21558: Safe more indent, again no code changes.
Günther Deschner [Tue, 27 Feb 2007 13:31:42 +0000 (13:31 +0000)]
r21558: Safe more indent, again no code changes.

Guenther

14 years agor21557: indent only fix. No code change.
Günther Deschner [Tue, 27 Feb 2007 13:27:31 +0000 (13:27 +0000)]
r21557: indent only fix. No code change.

Guenther

14 years agor21556: Remove superfluos return check in ads_keytab_verify_ticket().
Günther Deschner [Tue, 27 Feb 2007 13:25:42 +0000 (13:25 +0000)]
r21556: Remove superfluos return check in ads_keytab_verify_ticket().

Guenther

14 years agor21552: Ensure to check for proto_exists before linking any binary. No make
Lars Müller [Mon, 26 Feb 2007 23:07:54 +0000 (23:07 +0000)]
r21552: Ensure to check for proto_exists before linking any binary.  No make
proto should be required before creating any binary from now on.

Remove proto_exists from the all, pam_smbpass, and pam_bindind rule.

14 years agor21551: Ok, this is more subtle. More tomorrow :-)
Volker Lendecke [Mon, 26 Feb 2007 23:06:17 +0000 (23:06 +0000)]
r21551: Ok, this is more subtle. More tomorrow :-)

14 years agor21550: make disp_info_list static to get_samr_dispinfo_by_sid(), add a comment :-)
Volker Lendecke [Mon, 26 Feb 2007 22:48:42 +0000 (22:48 +0000)]
r21550: make disp_info_list static to get_samr_dispinfo_by_sid(), add a comment :-)

14 years agor21549: Only create DISP_INFO structs for domain handles, the others don't need
Volker Lendecke [Mon, 26 Feb 2007 22:44:24 +0000 (22:44 +0000)]
r21549: Only create DISP_INFO structs for domain handles, the others don't need
them. It just does not make sense to do a querydispinfo on an alias handle...

This fixes a memleak: Every samr_connect*() call leaked a DISP_INFO for the
(NULL) sid.

More cleanup pending: Essentially, we only need the DISP_INFO cache for the
get_global_sam_sid() domain. BUILTIN is fixed and small enough, and there are
no other domains around where enumerations could happen.

This also removes the explicit builtin_domain flags. I don't think this is
worth it. If this makes a significant difference, then we have a *VERY* tuned
RPC layer...

Jeremy, please check this. If it's ok, we might want to merge it across.

Volker

14 years agor21548: prevent segv (reference to -1 element of array)
Herb Lewis [Mon, 26 Feb 2007 22:19:23 +0000 (22:19 +0000)]
r21548: prevent segv (reference to -1 element of array)

14 years agor21547: Fix from Michael Adam <ma@sernet.de>: Refuse registry shares without path.
Volker Lendecke [Mon, 26 Feb 2007 20:14:35 +0000 (20:14 +0000)]
r21547: Fix from Michael Adam <ma@sernet.de>: Refuse registry shares without path.

Thanks,

Volker

14 years agor21546: remove duplicate lines
Herb Lewis [Mon, 26 Feb 2007 17:43:45 +0000 (17:43 +0000)]
r21546: remove duplicate lines

14 years agor21543: Fix 64bit build warning.
Günther Deschner [Mon, 26 Feb 2007 16:37:19 +0000 (16:37 +0000)]
r21543: Fix 64bit build warning.

Guenther

14 years agor21540: Fix Bug #3713 and readd reporting what the profiles tool does (when
Günther Deschner [Mon, 26 Feb 2007 13:07:34 +0000 (13:07 +0000)]
r21540: Fix Bug #3713 and readd reporting what the profiles tool does (when
called with the -v option).

Patch from William Jojo <jojowil@hvcc.edu>.

Guenther

14 years agor21537: Avoid to trigger the confusing "cached entry differs." warning when
Günther Deschner [Mon, 26 Feb 2007 09:53:35 +0000 (09:53 +0000)]
r21537: Avoid to trigger the confusing "cached entry differs." warning when
there is just no cache around for a user.

Guenther

14 years agor21536: Fix copy/paste typo.
Günther Deschner [Mon, 26 Feb 2007 09:46:05 +0000 (09:46 +0000)]
r21536: Fix copy/paste typo.

Guenther

14 years agor21530: Don't code with jet-lag and Volker looking over your
Jeremy Allison [Sun, 25 Feb 2007 09:04:06 +0000 (09:04 +0000)]
r21530: Don't code with jet-lag and Volker looking over your
shoulder.... Correct fix for warning :-)
Jeremy.

14 years agor21529: Fix warning from bad cast.
Jeremy Allison [Sun, 25 Feb 2007 09:00:37 +0000 (09:00 +0000)]
r21529: Fix warning from bad cast.
Jeremy.

14 years agor21526: Fix stray character in sys_memalign() that is only
Gerald Carter [Sat, 24 Feb 2007 13:03:59 +0000 (13:03 +0000)]
r21526: Fix stray character in sys_memalign() that is only
is the case where we don't have memalign() or posix_memalign().

14 years agor21525: Go ahead and checkin the mlock() & memalign() fixes so
Gerald Carter [Sat, 24 Feb 2007 12:40:43 +0000 (12:40 +0000)]
r21525: Go ahead and checkin the mlock() & memalign() fixes so
others don't get stuck with the winbindd hang.
Still waiting on additional confirmation from Guenther
that this fixes thes issues he was observing as well.
But it's been running in my local tree for a day without
problems.

14 years agor21517: Fix build warnings.
Günther Deschner [Fri, 23 Feb 2007 11:53:43 +0000 (11:53 +0000)]
r21517: Fix build warnings.

Guenther

14 years agor21509: Merge lp_passdb_backend() from rev 21506 to 3_0_25.
Lars Müller [Thu, 22 Feb 2007 23:12:36 +0000 (23:12 +0000)]
r21509: Merge lp_passdb_backend() from rev 21506 to 3_0_25.

Slightly change the DEBUG 0 message as suggested by Volker on
samba-technical.

14 years agor21508: Fix memleak in new idmap_tdb, thanks Herb.
Simo Sorce [Thu, 22 Feb 2007 21:59:54 +0000 (21:59 +0000)]
r21508: Fix memleak in new idmap_tdb, thanks Herb.

Jerry please check.

Simo.

14 years agor21507: Fix some "cannot access LDAP when no root" bugs.
Gerald Carter [Thu, 22 Feb 2007 20:52:27 +0000 (20:52 +0000)]
r21507: Fix some "cannot access LDAP when no root" bugs.
The two culprits were

* pdb_get_account_policy()
* pdb_get_group_sid()

14 years agor21506: Allow old pre 3.0.22 multi passdb backend configurations to work with
Lars Müller [Thu, 22 Feb 2007 17:52:23 +0000 (17:52 +0000)]
r21506: Allow old pre 3.0.22 multi passdb backend configurations to work with
post 3.0.23.

This implementation considers spaces in ldapsam configs.  Such configs
are trunkated after the closing quote.

14 years agor21505: make sure mlock()'d memory is aligned on a page boundary
Gerald Carter [Thu, 22 Feb 2007 17:21:27 +0000 (17:21 +0000)]
r21505: make sure mlock()'d memory is aligned on a page boundary

14 years agor21500: Fix inappropriate creation of a krb5 ticket refreshing event when a user
Günther Deschner [Thu, 22 Feb 2007 13:35:01 +0000 (13:35 +0000)]
r21500: Fix inappropriate creation of a krb5 ticket refreshing event when a user
changed a password via pam_chauthtok. Only do this if

a) a user logs on using an expired password (or a password that needs to
be changed immediately) or

b) the user itself changes his password.

Also make sure to delete the in-memory krb5 credential cache (when a
user did not request a FILE based cred cache).

Finally honor the krb5 settings in the first pam authentication in the
chauthtok block (PAM_PRELIM_CHECK). This circumvents confusion when
NTLM samlogon authentication is still possible with the old password after
the password has been already changed (on w2k3 sp1 dcs).

Guenther

14 years agor21483: Fix use of uninitialized variable.
Jeremy Allison [Wed, 21 Feb 2007 02:15:23 +0000 (02:15 +0000)]
r21483: Fix use of uninitialized variable.
Jeremy.

14 years agor21482: Use IPC$ not ipc$ for consistency.
Jeremy Allison [Wed, 21 Feb 2007 02:11:06 +0000 (02:11 +0000)]
r21482: Use IPC$ not ipc$ for consistency.
Jeremy.

14 years agor21481: No one said anything, so I'm disallowing anything
Jeremy Allison [Wed, 21 Feb 2007 02:04:28 +0000 (02:04 +0000)]
r21481: No one said anything, so I'm disallowing anything
but explicit shares in "default service" :-).
Jeremy.

14 years agor21480: Make fd_open match fd_close be translating
Jeremy Allison [Wed, 21 Feb 2007 01:55:08 +0000 (01:55 +0000)]
r21480: Make fd_open match fd_close be translating
errno into an NTSTATUS immediately.
Jeremy.

14 years agor21478: Add 65k length limit for split SPNEGO blobs.
Jeremy Allison [Tue, 20 Feb 2007 23:56:46 +0000 (23:56 +0000)]
r21478: Add 65k length limit for split SPNEGO blobs.
Jeremy

14 years agor21474: Ensure trustdom_cache_shutdown() gets called
Jeremy Allison [Tue, 20 Feb 2007 19:57:14 +0000 (19:57 +0000)]
r21474: Ensure trustdom_cache_shutdown() gets called
on terminate. Pointed out by Herb.
Jeremy.

14 years agor21467: Add GPFS-provided DMAPI support based on their GPL library
Alexander Bokovoy [Tue, 20 Feb 2007 15:55:12 +0000 (15:55 +0000)]
r21467: Add GPFS-provided DMAPI support based on their GPL library

14 years agor21462: Fix EnumValue (?)
Jelmer Vernooij [Tue, 20 Feb 2007 13:43:41 +0000 (13:43 +0000)]
r21462: Fix EnumValue (?)

14 years agor21461: Not strictly necessary, as data_blob() panics if it can't allocate. But I'd
Volker Lendecke [Tue, 20 Feb 2007 12:16:20 +0000 (12:16 +0000)]
r21461: Not strictly necessary, as data_blob() panics if it can't allocate. But I'd
see this as a design flaw in data_blob() and it made me look in that routine.

Jeremy, revert or merge please :-)

Volker

14 years agor21460: Fix for server-side processing of SPNEGO auth
Jeremy Allison [Tue, 20 Feb 2007 06:22:20 +0000 (06:22 +0000)]
r21460: Fix for server-side processing of SPNEGO auth
fragmented into "max xmit" size security blob
chunks. Bug #4400. Needs limits adding, and also
a client-side version.
Jeremy.

14 years agor21454: Fix debug typo.
Günther Deschner [Mon, 19 Feb 2007 17:47:50 +0000 (17:47 +0000)]
r21454: Fix debug typo.

Guenther

14 years agor21453: use the tdb mapping file for groups by default for compatibility with the...
Gerald Carter [Mon, 19 Feb 2007 17:21:21 +0000 (17:21 +0000)]
r21453: use the tdb mapping file for groups by default for compatibility with the 3.0.25 tree (for now)

14 years agor21450: No need to TALLOC_FREE twice here.
Günther Deschner [Mon, 19 Feb 2007 14:43:31 +0000 (14:43 +0000)]
r21450: No need to TALLOC_FREE twice here.

Guenther

14 years agor21449: Talked to Jerry on irc, he does not see what might break with this: Change
Volker Lendecke [Mon, 19 Feb 2007 13:52:29 +0000 (13:52 +0000)]
r21449: Talked to Jerry on irc, he does not see what might break with this: Change
back the 'msdfs root = yes' default to 'no'.

Volker

14 years agor21446: Karolins "printjob username"
Volker Lendecke [Mon, 19 Feb 2007 13:30:07 +0000 (13:30 +0000)]
r21446: Karolins "printjob username"

14 years agor21445: Apply tdb_parse_record Tridges error return, merge to 3_0_25 and 4_0
Volker Lendecke [Mon, 19 Feb 2007 11:45:33 +0000 (11:45 +0000)]
r21445: Apply tdb_parse_record Tridges error return, merge to 3_0_25 and 4_0

14 years agor21444: Check in tdb_parse_record. Not merging to the other branches yet, we need to
Volker Lendecke [Mon, 19 Feb 2007 11:19:53 +0000 (11:19 +0000)]
r21444: Check in tdb_parse_record. Not merging to the other branches yet, we need to
agree on the behaviour of non-existing records.

Tridge, can you comment? Should we change tdb_fetch, or should we have
different concepts in tdb_fetch() and tdb_parse_record() ?

Volker

14 years agor21417: Janitor for Herb. Make sure sharesec.c is functionally
Jeremy Allison [Sun, 18 Feb 2007 01:31:50 +0000 (01:31 +0000)]
r21417: Janitor for Herb. Make sure sharesec.c is functionally
identical in 3.0.25 and 3.0.
Jeremy.

14 years agor21412: The last patch also incremented the seqnum when tdb_store failed. Not as bad
Volker Lendecke [Sat, 17 Feb 2007 23:41:45 +0000 (23:41 +0000)]
r21412: The last patch also incremented the seqnum when tdb_store failed. Not as bad
as not doing it at all, but needs fixing. Also simplify the logic, I had
missed the "goto out" at the end of the function.

Volker

14 years agor21410: We have to increment the sequence number also when tdb_update_hash()
Volker Lendecke [Sat, 17 Feb 2007 21:46:13 +0000 (21:46 +0000)]
r21410: We have to increment the sequence number also when tdb_update_hash()
succeeded. Found while testing the brlock seqnum patch.

Tridge, please check!

Volker

14 years agor21409: saved_status1 is not used anymore after aio write behind was removed
Volker Lendecke [Sat, 17 Feb 2007 18:47:40 +0000 (18:47 +0000)]
r21409: saved_status1 is not used anymore after aio write behind was removed

14 years agor21399: need to zero the request and response structures
Herb Lewis [Fri, 16 Feb 2007 20:27:30 +0000 (20:27 +0000)]
r21399: need to zero the request and response structures

14 years agor21397: revert accidential commit
Herb Lewis [Fri, 16 Feb 2007 20:02:13 +0000 (20:02 +0000)]
r21397: revert accidential commit

14 years agor21396: fix wbinfo --lookup-rids command
Herb Lewis [Fri, 16 Feb 2007 19:49:12 +0000 (19:49 +0000)]
r21396: fix wbinfo --lookup-rids command
allow detection of libbiconv if all others fail - need for FreeBSD

14 years agor21394: Prevent nscd crash due to potential NULL pointer dereference in
Günther Deschner [Fri, 16 Feb 2007 17:25:54 +0000 (17:25 +0000)]
r21394: Prevent nscd crash due to potential NULL pointer dereference in
_nss_winbind_initgroups_dyn() on an empty group list.

Guenther

14 years agor21393: Fix typo.
Günther Deschner [Fri, 16 Feb 2007 16:59:28 +0000 (16:59 +0000)]
r21393: Fix typo.

Guenther

14 years agor21388: Merge support for providing replacement system headers.
Jelmer Vernooij [Fri, 16 Feb 2007 15:35:26 +0000 (15:35 +0000)]
r21388: Merge support for providing replacement system headers.

14 years agor21387: Another important fix for non-AD domains:
Günther Deschner [Fri, 16 Feb 2007 15:13:57 +0000 (15:13 +0000)]
r21387: Another important fix for non-AD domains:

Avoid assigning 0 as primary group id for users in NSS calls.
Jerry, please check.

Guenther

14 years agor21385: Regenerate IDL after pidl changes, sync winreg.idl from samba4.
Jelmer Vernooij [Fri, 16 Feb 2007 14:50:57 +0000 (14:50 +0000)]
r21385: Regenerate IDL after pidl changes, sync winreg.idl from samba4.

14 years agor21383: More possible "security=share" fixes. If a client
Jeremy Allison [Fri, 16 Feb 2007 13:40:11 +0000 (13:40 +0000)]
r21383: More possible "security=share" fixes. If a client
is sending LMv2 make sure we test with the password
blob in the LM field as well as the NT field.
Jeremy.

14 years agor21382: Important fix for winbind when using non-AD domains.
Günther Deschner [Fri, 16 Feb 2007 13:30:19 +0000 (13:30 +0000)]
r21382: Important fix for winbind when using non-AD domains.

Jeremy, I'm afraid you removed the "domain->initialized" from the
set_dc_types_and_flags() call when the connect to PI_LSARPC_DS failed
(with rev. 19148).

This causes now that init_dc_connection_network is called again and
again which in turn rescans the DC each time (which of course fails each
time with NT_STATUS_BUFFER_TOO_SMALL). Just continue with the
non-PI_LSARPC_DS scan so that the domain is initialized properly.

Guenther

14 years agor21379: Attempt to fix the build on "gwen"
Volker Lendecke [Fri, 16 Feb 2007 12:13:52 +0000 (12:13 +0000)]
r21379: Attempt to fix the build on "gwen"

14 years agor21369: sys_disk_free return type is SMB_BIG_UINT.
Simo Sorce [Thu, 15 Feb 2007 16:50:14 +0000 (16:50 +0000)]
r21369: sys_disk_free return type is SMB_BIG_UINT.
Fix dfree_retval to be SMB_BIG_UINT as well,
otherwise we may wrap up on > 2T file systems.

Simo.

14 years agor21365: Lets see which build farm machines have makecontext & friends
Volker Lendecke [Thu, 15 Feb 2007 14:09:39 +0000 (14:09 +0000)]
r21365: Lets see which build farm machines have makecontext & friends

14 years agor21358: Some more debugging for _nss_winbind_initgroups_dyn() on Linux.
Günther Deschner [Thu, 15 Feb 2007 12:10:57 +0000 (12:10 +0000)]
r21358: Some more debugging for _nss_winbind_initgroups_dyn() on Linux.

Guenther

14 years agor21357: Fix typo.
Günther Deschner [Thu, 15 Feb 2007 12:05:19 +0000 (12:05 +0000)]
r21357: Fix typo.

Guenther

14 years agor21353: In the turn of tracking down nss_winbind related bugs on Linux:
Günther Deschner [Thu, 15 Feb 2007 11:44:06 +0000 (11:44 +0000)]
r21353: In the turn of tracking down nss_winbind related bugs on Linux:
print NSS_STATUS code with DEBUG_NSS when leaving a function.

Guenther

14 years agor21352: Let ads_upn_suffixes() return a pointer to an array of suffixes.
Günther Deschner [Thu, 15 Feb 2007 00:03:38 +0000 (00:03 +0000)]
r21352: Let ads_upn_suffixes() return a pointer to an array of suffixes.

Guenther

14 years agor21349: Fix memleak in ads_upn_suffixes().
Günther Deschner [Wed, 14 Feb 2007 20:52:23 +0000 (20:52 +0000)]
r21349: Fix memleak in ads_upn_suffixes().

Guenther

14 years agor21346: FIXME ! Our parsing here is wrong I think,
Jeremy Allison [Wed, 14 Feb 2007 17:17:32 +0000 (17:17 +0000)]
r21346: FIXME ! Our parsing here is wrong I think,
but for a level3 it makes no sense for
ptr_sec_desc to be NULL. JRA. Based on
a Vista sniff from Martin Zielinski <mz@seh.de>.
Jerry - part of the Vista patchset.
Jeremy.

14 years agor21345: Smaller fixes for adssearch:
Günther Deschner [Wed, 14 Feb 2007 16:20:38 +0000 (16:20 +0000)]
r21345: Smaller fixes for adssearch:

* get rid of horrible ads.h parsing
* add LDAP_SERVER_SHUTDOWN_NOTIFY_OID
* display hex bitmasks

Guenther

14 years agor21339: Fix the non-linux build. This is more evidence that this needs to be
Jim McDonough [Wed, 14 Feb 2007 14:25:56 +0000 (14:25 +0000)]
r21339: Fix the non-linux build.  This is more evidence that this needs to be
moved up one layer.

14 years agor21336: Fix indent (as pointed out by Volker).
Günther Deschner [Wed, 14 Feb 2007 13:51:42 +0000 (13:51 +0000)]
r21336: Fix indent (as pointed out by Volker).

Guenther

14 years agor21324: Add linux setlease to the vfs layer. Next round, as Volker points out,
Jim McDonough [Wed, 14 Feb 2007 02:37:14 +0000 (02:37 +0000)]
r21324: Add linux setlease to the vfs layer.  Next round, as Volker points out,
it should be abstracted a little higher up so other os'es can have an
entry, but it will take a bit more work.  Thanks to Chetan Shringarpure
and Mathias Dietz.

I didn't increment the vfs number again because the kernel change notify
stuff hasn't been released yet anyway.

14 years agor21320: Display query_user info level 16 in rpcclient.
Günther Deschner [Tue, 13 Feb 2007 16:04:36 +0000 (16:04 +0000)]
r21320: Display query_user info level 16 in rpcclient.

Guenther

14 years agor21319: Remove functions not needed anymore
Volker Lendecke [Tue, 13 Feb 2007 15:57:54 +0000 (15:57 +0000)]
r21319: Remove functions not needed anymore

14 years agor21318: Fix Bug #4225.
Günther Deschner [Tue, 13 Feb 2007 15:56:09 +0000 (15:56 +0000)]
r21318: Fix Bug #4225.

Cached logon with pam_winbind should work now also for NT4 and samba3
domains.

Guenther

14 years agor21312: merge from SAMBA_4_0:
Stefan Metzmacher [Tue, 13 Feb 2007 12:42:28 +0000 (12:42 +0000)]
r21312: merge from SAMBA_4_0:

fix memory hierachy, and access to already freed memory

metze

14 years agor21310: Fix invalid printfs in pam_winbind.
Günther Deschner [Tue, 13 Feb 2007 11:04:10 +0000 (11:04 +0000)]
r21310: Fix invalid printfs in pam_winbind.

Guenther

14 years agor21309: Add PRINTF_ATTRIBUTE checks for log statements.
Günther Deschner [Tue, 13 Feb 2007 10:56:04 +0000 (10:56 +0000)]
r21309: Add PRINTF_ATTRIBUTE checks for log statements.

Guenther

14 years agor21308: Fix some typos and ensure to null terminate the correct strings.
Günther Deschner [Tue, 13 Feb 2007 10:42:53 +0000 (10:42 +0000)]
r21308: Fix some typos and ensure to null terminate the correct strings.

Guenther

14 years agor21303: As discussed on samba-technical: Change the static array for the in-memory
Volker Lendecke [Mon, 12 Feb 2007 23:16:02 +0000 (23:16 +0000)]
r21303: As discussed on samba-technical: Change the static array for the in-memory
mirrors of the hash chain locks to a dynamically allocated one.

Jeremy, I count on you to revert it if the build farm freaks out, it's after
midnight here :-)

Volker

14 years agor21284: Fix some unitilized variable warnings pointed out by Volker.
Gerald Carter [Sun, 11 Feb 2007 19:47:38 +0000 (19:47 +0000)]
r21284: Fix some unitilized variable warnings pointed out by Volker.

14 years agor21283: Do not display away debug output of cifs dfs resolver by default
Steve French [Sun, 11 Feb 2007 19:28:52 +0000 (19:28 +0000)]
r21283: Do not display away debug output of cifs dfs resolver by default
per Dave Howell's suggestion

14 years agor21280: Fix an uninitialized variable warning. Jeremy, please check.
Volker Lendecke [Sun, 11 Feb 2007 14:55:21 +0000 (14:55 +0000)]
r21280: Fix an uninitialized variable warning. Jeremy, please check.

Volker

14 years agor21279: Get rid of 'aio write behind', this is broken.
Volker Lendecke [Sun, 11 Feb 2007 14:39:21 +0000 (14:39 +0000)]
r21279: Get rid of 'aio write behind', this is broken.

It should probably better be integrated with our write cache.

Volker

14 years agor21278: The main goal of this was to get rid of the NetInBuffer / set_InBuffer. But it
Volker Lendecke [Sun, 11 Feb 2007 14:07:50 +0000 (14:07 +0000)]
r21278: The main goal of this was to get rid of the NetInBuffer / set_InBuffer. But it
turns out that this patch actually speeds up the async writes considerably.

I tested writing 100.000 times 65535 bytes with the allowed 10 ops in
parallel. Without this patch it took about 32 seconds on my dual-core 1.6GHz
laptop. With this patch it dropped to about 26 seconds. I can only explain it
by better cache locality, NewInBuffer allocates more than 128k, so we jump
around in memory more.

Jeremy, please check!

Volker

14 years agor21277: Fix an off by one error in the signal handler for aio: We can actually receive
Volker Lendecke [Sun, 11 Feb 2007 13:42:48 +0000 (13:42 +0000)]
r21277: Fix an off by one error in the signal handler for aio: We can actually receive
AIO_PENDING_SIZE signals, not one less.

Jeremy I'm not merging this to 3_0_25, I want you to look at it first!

Volker

14 years agor21274: As we now have POSIX opens which can unlink
Jeremy Allison [Sun, 11 Feb 2007 04:37:56 +0000 (04:37 +0000)]
r21274: As we now have POSIX opens which can unlink
with other open files we may have taken
the delete code path with more than one share mode
entry - ensure we only delete once by resetting
the delete on close flag.
Jeremy.

14 years agor21273: * Protect the sasl bind against a NULL principal string
Gerald Carter [Sat, 10 Feb 2007 20:29:09 +0000 (20:29 +0000)]
r21273: * Protect the sasl bind against a NULL principal string
  in the SPNEGO negTokenInit

14 years agor21269: Reformatting
Volker Lendecke [Sat, 10 Feb 2007 13:15:58 +0000 (13:15 +0000)]
r21269: Reformatting

14 years agor21265: Fix some valgrind errors -- not in 3_0
Volker Lendecke [Fri, 9 Feb 2007 23:14:23 +0000 (23:14 +0000)]
r21265: Fix some valgrind errors -- not in 3_0

14 years agor21264: LDAP_OPT_ERROR_NUMBER seems more portable than LDAP_OPT_RESULT_CODE
Volker Lendecke [Fri, 9 Feb 2007 21:25:22 +0000 (21:25 +0000)]
r21264: LDAP_OPT_ERROR_NUMBER seems more portable than LDAP_OPT_RESULT_CODE

14 years agor21263: getpeername() returning -1 is not a reliable indication if a TCP connection is
Volker Lendecke [Fri, 9 Feb 2007 20:58:17 +0000 (20:58 +0000)]
r21263: getpeername() returning -1 is not a reliable indication if a TCP connection is
dead. Might be my code, this rings a very distant bell...

Attempt to fix bug # 4372.

Volker

14 years agor21262: Final part of BUG 4093: fix %a with Windows XP 64bit
Gerald Carter [Fri, 9 Feb 2007 19:41:09 +0000 (19:41 +0000)]
r21262: Final part of BUG 4093: fix %a with Windows XP 64bit

14 years agor21259: Fix coverity bug id #340. No way to process
Jeremy Allison [Fri, 9 Feb 2007 05:16:25 +0000 (05:16 +0000)]
r21259: Fix coverity bug id #340. No way to process
logon parameter as the code was written.
Jeremy.

14 years agor21258: Fix for connecting printers from Vista by storing
Gerald Carter [Fri, 9 Feb 2007 02:12:12 +0000 (02:12 +0000)]
r21258: Fix for connecting printers from Vista by storing
the printer GUID as a REG_SZ as Vista seems to
whine about unknown REG_BINARY values.

Thanks to Martin Zielinski <mz@seh.de> for his excellent
analysis on this.

14 years agor21257: Better fix for bug #4188 :
Jeremy Allison [Fri, 9 Feb 2007 02:03:39 +0000 (02:03 +0000)]
r21257: Better fix for bug #4188 :
Windows Vista RC1 and RC2 can't delete directory on Samba share
based on work by Joe Meadows <jmeadows@webopolis.com>.
Jeremy.

14 years agor21251: Okay, after Jeremy has kindly tested this, check it in :-)
Volker Lendecke [Thu, 8 Feb 2007 21:57:41 +0000 (21:57 +0000)]
r21251: Okay, after Jeremy has kindly tested this, check it in :-)

Attached find a workaround that works for me. This is not the "correct" fix,
to me it seems our DFS referral marshalling is broken. Vista requests level 4,
we reply with level 2, and Vista seems not to like that. If we reply with
level 3 it seems more happy.

Needs more work!

Volker

14 years agor21250: Partial fix for BUG 4093: Make %a expand to "Vista"
Gerald Carter [Thu, 8 Feb 2007 21:48:09 +0000 (21:48 +0000)]
r21250: Partial fix for BUG 4093: Make %a expand to "Vista"
based on the flags2 values in the negprot request.

This also includes some code for testing the dialect
strings for "SMB 2.001" but this is unreliable as Vista
only sends that in the 1st negprot and caches the fact
that we don't support it.  Restartnig the WOrkstation
service on the client clears the cache.

14 years agor21247: In the refactoring I dropped a RESOLVE_DFSPATH call
Jeremy Allison [Thu, 8 Feb 2007 20:31:18 +0000 (20:31 +0000)]
r21247: In the refactoring I dropped a RESOLVE_DFSPATH call
from setpathinfo. Return it and make sure all paths
go through a dfs resolve.
Jeremy.