summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2010-08-14s3:idmap_tdb2: remove unused idmap_tdb2_alloc_close().Michael Adam1-9/+0
2010-08-14s3:idmap_tdb2: remove unused idmap_alloc_methods.Michael Adam1-6/+0
2010-08-14s3:idmap_tdb2: don't call smb_register_idmap_alloc() in idmap_tdb2_initMichael Adam1-9/+0
The registering of alloc backends is being removed. The idmap backends are responsible for initializing their alloc code on their own if necessary.
2010-08-14s3:idmap_tdb: make idmap_alloc_tdb_init() static.Michael Adam2-2/+1
2010-08-14s3:idmap: remove unused idmap_alloc_init().Michael Adam1-79/+0
2010-08-14s3:idmap: use allocate_id() from the idmap_methods in idmap_allocate_unixid()Michael Adam1-4/+12
The idmap alloc methods are being removed.
2010-08-14s3:idmap: add an allocate_id method to the idmap_methods struct.Michael Adam1-0/+3
The idmap_alloc_methods are being removed, but this single "alloc" method is still needed for the samba-wide Unix-ID allocator, which is used in group-mapping and ldapsam:editposix. This method should ultimately also disappear.
2010-08-14s3:idmap: factor out common code of idmap_allocate_uid|gid()Michael Adam1-10/+9
into new idmap_allocate_unixid().
2010-08-14s3:idmap: remove the set_mapping method from the idmap APIMichael Adam5-18/+0
2010-08-14s3:idmap: remove idmap_new_mapping() - now implemented in the backendsMichael Adam1-68/+0
2010-08-14s3:idmap: add a debug message to idmap_sid_to_gidMichael Adam1-0/+1
2010-08-14s3:idmap: add a debug message to idmap_sid_to_uidMichael Adam1-0/+1
2010-08-14s3:idmap: don't call idmap_new_mapping idmap_sid_to_gidMichael Adam1-25/+16
The setting of a new mapping is moved into the backend code to achieve atomicity and greater flexibility. Michael
2010-08-14s3:idmap: don't call idmap_new_mapping idmap_sid_to_unixid.Michael Adam1-25/+15
The setting of a new mapping is moved into the backend code to achieve atomicity and greater flexibility. Michael
2010-08-14s3:winbind: increase interface version after removing callsMichael Adam1-1/+4
Michael
2010-08-14s3:idmap: remove unused method set_id_hwm from idmap APIMichael Adam4-177/+0
Michael
2010-08-14s3:idmap: remove unused alloc method get_id_hwm from idmap APIMichael Adam4-194/+0
Michael
2010-08-14s3:idmap: remove unused method dump_data() from the idmap APIMichael Adam5-120/+0
Michael
2010-08-14s3:idmap: remove the remove_mapping method from API and backendsMichael Adam4-147/+0
Michael
2010-08-14s3:idmap: remove unused idmap_remove_mapping().Michael Adam2-18/+0
Michael
2010-08-14s4: remove REMOVE_MAPPING from wb_samba3_protocolMichael Adam1-2/+0
Michael
2010-08-14s3:winbind: remove the method REMOVE_MAPPING from winbind's APIMichael Adam7-153/+0
Michael
2010-08-14s3:idmap: remove unused idmap_set_mapping().Michael Adam2-18/+0
Michael
2010-08-14s4: remove SET_MAPPING from wb_samba3_protocolMichael Adam1-2/+0
Michael
2010-08-14s3:winbind: remove the method SET_MAPPING from winbind's APIMichael Adam7-146/+0
Michael
2010-08-14libwbclient: unimplement wbcRemoveGidMapping()Michael Adam1-33/+2
Michael
2010-08-14libwbclient: unimplement wbcRemoveUidMapping()Michael Adam1-33/+2
Michael
2010-08-14libwbclient: unimplement wbcSetGidMapping()Michael Adam1-33/+2
Michael
2010-08-14libwbclient: unimplement wbcSetUidMapping()Michael Adam1-33/+2
Michael
2010-08-14s3:idmap: remove unused idmap_set_gid_hwm()Michael Adam2-14/+0
Michael
2010-08-14s3:idmap: remove unused idmap_set_uid_hwm()Michael Adam2-14/+0
Michael
2010-08-14s4: remove SET_HWM and SET_DUAL_HWM from wb_samba3_protocolMichael Adam1-2/+0
Michael
2010-08-14s3:winbind: remove SET_HWM from winbind's API.Michael Adam7-135/+0
2010-08-14libwbclient: unimplement wbcSetGidHwm()Michael Adam1-21/+2
Michael
2010-08-14libwbclient: unimplement wbcSetUidHwm()Michael Adam1-21/+2
Michael
2010-08-14s3:net: rewrite "net idmap restore" using dbwrapMichael Adam1-52/+130
This removes the use of libwbclient from "net idmap". This detects the idmap backend from the current smb.conf and stores the id mapping tables in the corresponding data bases. Currently, only tdb and tdb2 backends are supported. Support for the ldap backend may be implemented later.
2010-08-14s3:net: change "net idmap dump" to use dbwrap instead of direct tdb accessMichael Adam1-15/+17
This way, it will also work in a samba-ctdb cluster.
2010-08-14s3:net sam provision: compose sid before getting uid for GuestMichael Adam1-2/+2
Michael
2010-08-14s3:net sam provision: allocate the uid after composing the sid for AdministratorMichael Adam1-6/+7
Michael
2010-08-14s3:pdb_ldap: move some code in ldapsam_create_dom_group()Michael Adam1-22/+25
to make the flow more similar to ldapsam_create_user(). This prepares for calling winbind_sid_to_gid() instead of winbind_allocate_gid(): we need the group_sid for this... Michael
2010-08-14s3:groupdb: allocate a gid after allocating a rid in pdb_default_create_alias()Michael Adam1-6/+7
Michael
2010-08-14s3:idmap_tdb2: fix a debug messageMichael Adam1-1/+1
2010-08-14s3:test: add a new "valgrindtestenv" target, similar to gdbtestenvMichael Adam1-0/+7
this runs the daemins in xterm under valgrind instead of in gdb
2010-08-14s3:gdbtestenv: also start an xterm with nmbd. and start winbindd before smbd.Michael Adam1-2/+3
2010-08-14s3:gdbtestenv: also set the xterm's window title appropriatelyMichael Adam1-2/+2
2010-08-14s3:gdbtestenv: print the server's name in the icontitle instead of "server"Michael Adam1-2/+2
2010-08-13pidl: fix printing of server side ndr request debug messagesMichael Adam1-2/+2
The macros NDR_PRINT_IN_DEBUG and NDR_PRINT_OUT_DEBUG are made for the client side. For the server side, the NDR_SET_VALUES flag needs to be added for the OUT struct and not for the IN struct, otherwise, the OUT part can print uninitialized data and the IN part may recalculate string lengths illegally.
2010-08-13s3: Fix some error messagesVolker Lendecke3-7/+9
2010-08-13Eliminate any chance of a class of "uninitialized auto variable" errors.Jeremy Allison1-26/+24
Jeremy.
2010-08-14tdb: workaround starvation problem in locking entire database.Rusty Russell4-19/+131
We saw tdb_lockall() take 71 seconds under heavy load; this is because Linux (at least) doesn't prevent new small locks being obtained while we're waiting for a big log. The workaround is to do divide and conquer using non-blocking chainlocks: if we get down to a single chain we block. Using a simple test program where children did "hold lock for 100ms, sleep for 1 second" the time to do tdb_lockall() dropped signifiantly. There are ln(hashsize) locks taken in the contended case, but that's slow anyway. More analysis is given in my blog at http://rusty.ozlabs.org/?p=120 This may also help transactions, though in that case it's the initial read lock which uses this gradual locking routine; the update-to-write-lock code is separate and still tries to update in one go. Even though ABI doesn't change, minor version bumped so behavior change can be easily detected. Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
generated by cgit (git 2.34.1) at 2024-10-20 23:38:56 +0000