summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2009-02-26s3:registry: use regsubkey_ctr_init() in reg_eventlog.cMichael Adam1-4/+10
Michael
2009-02-26s3:services_db: use regsubkey_ctr_init() instead of using talloc directlyMichael Adam1-2/+4
Michael
2009-02-26s3:registry: use regsubkey_ctr_get_seqnum() in reg_backend_db.cMichael Adam1-1/+1
Michael
2009-02-26s3:registry: use regsubkey_ctr_set_seqnum() in reg_backend_db.cMichael Adam1-1/+4
Michael
2009-02-26s3:registry: use regsubkey_ctr_init() in reg_backend_db.cMichael Adam1-8/+14
instead of using talloc directly. Michael
2009-02-26s3:registry: use regsubkey_ctr_init() in reg_api.cMichael Adam1-13/+10
instead of using talloc on struct regsubkey_ctr. Michael
2009-02-26s3:registry: add regsubkey_ctr_get_seqnum() to hide implementationMichael Adam2-0/+10
Michael
2009-02-26s3:registry: add regsubkey_ctr_set_seqnum to hide implementation from caller.Michael Adam2-0/+12
Michael
2009-02-26s3:registry: add a regsubkey_ctr_init function for allocating a regsubkey_ctrMichael Adam2-1/+16
Michael
2009-02-26s3:registry: fix a commentMichael Adam1-3/+3
Michael
2009-02-26s3:registry: don't directly access key->subkeys->subkeys[] in reg_backend_db.cMichael Adam1-3/+5
Use the reg_objects accessor regsubkey_ctr_specific_key() instead. Michael
2009-02-26s3:registry: don't directly access key->subkeys->subkeys[] in reg_api.cMichael Adam1-2/+5
Use the reg_objects accessor regsubkey_ctr_specific_key() instead. Michael
2009-02-26s3:registry: don't directly access key->subkeys->num_subkeys in reg_backend_db.Michael Adam1-12/+16
Use the reg_objects api instead. Michael
2009-02-26s3:registry: don't directly access key->subkeys->num_subkeys in reg_api.cMichael Adam1-5/+5
Use the reg_objects api instead. Michael
2009-02-26s3:registry: replace typedef "REGSUBKEY_CTR" by "struct regsubkey_ctr"Michael Adam21-85/+85
This paves the way for hiding the typedef and the implementation from the surface. Michael
2009-02-26Revert "Fix a O(n^2) algorithm in regdb_fetch_keys()"Michael Adam1-27/+5
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
2009-02-26Remove a temporary debug message -- sorryVolker Lendecke1-1/+0
2009-02-26s3-spoolss: fix search_notify(). sorry...Günther Deschner1-1/+1
Guenther
2009-02-26s3: re-run make samba3-idl.Günther Deschner3-3/+54
Guenther
2009-02-26s3-spoolss: job status defines moved to IDL.Günther Deschner2-16/+1
Guenther
2009-02-26spoolss: add spoolss_JobStatus.Günther Deschner1-1/+22
Guenther
2009-02-26s3-spoolss: remove some unused defines and structs.Günther Deschner2-27/+1
Guenther
2009-02-26s4:heimdal_build: correrctly switch closefrom() and get/setprogname() ↵Stefan Metzmacher1-4/+6
replacements on and off metze
2009-02-26Convert open_socket_out to tevent_reqVolker Lendecke2-47/+47
2009-02-26Add tevent_req helpers to includes.hVolker Lendecke1-0/+2
2009-02-26Add tevent_ntstatusVolker Lendecke3-1/+84
2009-02-26s3:net: wrap net conf import into one big transactionMichael Adam1-6/+27
This speeds up "net conf import" of a file with 2000 shares from 11 minutest to 1m50s on my box. Michael
2009-02-26s3:libsmbconf: add transactions to the libsmbconf apiMichael Adam5-0/+57
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
2009-02-26s3:registry: wrap deletekey_recursive in one big transaction.Michael Adam1-2/+38
This speeds up "net conf drop" with 2000 shares on my box from 4m40s to 1m50s, leaving virtually only cpu load. Michael
2009-02-26s3:registry: provide transaction_start|commit|cancel fns for the registry tdbMichael Adam2-0/+21
Michael
2009-02-26Wrap creating the sorted subkey cache in a transactionVolker Lendecke1-3/+31
Signed-off-by: Michael Adam <obnox@samba.org>
2009-02-26Add a comment describing the sorted subkeysVolker Lendecke1-0/+23
Signed-off-by: Michael Adam <obnox@samba.org>
2009-02-26Add sorted subkey cacheVolker Lendecke2-19/+175
On my box this gets net conf list for 1000 records down to .1 seconds Signed-off-by: Michael Adam <obnox@samba.org>
2009-02-26Add db_tdb_parseVolker Lendecke1-0/+12
Signed-off-by: Michael Adam <obnox@samba.org>
2009-02-26Add dbwrap->parse_recordVolker Lendecke2-0/+30
Signed-off-by: Michael Adam <obnox@samba.org>
2009-02-26Speed up "net conf list"Volker Lendecke1-115/+41
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>
2009-02-26Speed up "net conf list"Volker Lendecke2-4/+3
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>
2009-02-26s4:Makefile: try to fix 'make installheaders' on *BSD systemsStefan Metzmacher1-1/+1
metze
2009-02-26s4:heimdal_build: don't always build setprogname.oStefan Metzmacher2-17/+34
This hopefully fixes the build on Mac OS 10, where setprogname() is available via libc. asn1_compile and compile_et use only .ho files again and don't use socket_wrapper anymore. metze
2009-02-26libreplace: make it possible to disable socket_wrapper via ↵Stefan Metzmacher1-2/+4
-DSOCKET_WRAPPER_DISABLE=1 metze
2009-02-26s4:ldb_map: include ldb_includes.h first to get replace.h as first headerStefan Metzmacher3-0/+3
This should bring the build on AIX one step further. metze
2009-02-26s4:VERSION: mark as 4.0.0-alpha8 gitsnapshotStefan Metzmacher1-2/+2
metze
2009-02-26Mark as the Samba 4.0 alpha7 releaseAndrew Bartlett1-1/+1
2009-02-26A simple hack to avoid the segfault in #6138Andrew Bartlett1-1/+1
Thanks to Andrew Kroeger <andrew@id10ts.net> for reporting this. This fix just for the release. A better fix will make it into the master branch soon. Andrew Bartlett
2009-02-25Fix more POSIX path lstat calls. Fix bug where close can returnJeremy Allison1-5/+17
failure if we have a pending modtime and the containing directory of the file has been renamed (there is no POSIX "update time by fd" call). This can't happen on Windows as the rename will fail if there are open files beneath it. Will add a torture test for this. Jeremy.
2009-02-25Use fsp->posix_open in preference if we have it.Jeremy Allison2-7/+7
Jeremy.
2009-02-25s3-spoolss: fix _spoolss_AddJob behaviour.Günther Deschner1-0/+4
Guenther
2009-02-25s4-spoolss: implement dcesrv_spoolss_AddJob which always is a dummy.Günther Deschner1-1/+5
Guenther
2009-02-25s4-smbtorture: add test_AddJob test to RPC-SPOOLSS test.Günther Deschner1-0/+29
Guenther
2009-02-25Ensure ACL modules work with POSIX paths.Jeremy Allison2-22/+79
Jeremy.