summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
...
* - don't leak the poll fd array if we exit the thread on error (defect #10105)Nalin Dahyabhai2011-01-031-0/+1
|
* - don't leak an array of lengths (defect #10104)Nalin Dahyabhai2011-01-031-0/+1
|
* - don't leak memory on out-of-memory (defect #10103)Nalin Dahyabhai2011-01-031-0/+1
|
* - don't deref a null pointer on out-of-memory here (defect #10093)Nalin Dahyabhai2011-01-031-0/+6
|
* - allocate space for n integers, not n pointers to integers (defect #10110)Nalin Dahyabhai2011-01-031-1/+1
|
* - clear the whole array of structures before filling it, instead of theNalin Dahyabhai2011-01-031-2/+2
| | | | | first sizeof(size_t) bytes (which would have been fine for 1 entry) (defect #10089)
* - clear the whole reply structure before use instead of just the initial ↵Nalin Dahyabhai2011-01-031-1/+1
| | | | part (defect #10088)
* - compile and link also using LDAP_CFLAGS and LDAP_LIBSNalin Dahyabhai2010-12-221-5/+5
|
* cast ber_len_t lengths to (int) when we use them to limit string lengthsNalin Dahyabhai2010-12-221-15/+17
| | | | in diagnostic messages
* - for containers which are also the groups, don't return the containerNalin Dahyabhai2010-11-231-3/+5
| | | | | entry as a search result if it's also the group entry (in which case we already looked at the entry)
* - also, call the right functionNalin Dahyabhai2010-11-231-1/+1
|
* - whoops, enforce the right argument countNalin Dahyabhai2010-11-231-2/+2
|
* - break up %deref into %deref and %deref_fNalin Dahyabhai2010-11-231-44/+107
| | | | - add %deref_fr as an alias for %deref_rf
* - fix a brown-paper-bag crash in the schema-compat pluginslapi-nis-0.19Nalin Dahyabhai2010-11-221-1/+1
|
* - handle a group with no explicitly-named container by treating theNalin Dahyabhai2010-11-222-19/+55
| | | | entries as direct subordinates of the group entry
* - use the filter at the right point in walking a chainNalin Dahyabhai2010-11-221-3/+37
| | | | - add a test for deref_rf
* - teach wrap_search_internal_get_entry to take an optional filterNalin Dahyabhai2010-11-225-109/+21
| | | | | - where we were applying a filter test after retrieving an entry, just let the retrieval function do it
* - try to factor deref_r and its corresponding update code intoNalin Dahyabhai2010-11-223-50/+197
| | | | optionally handling filters at each step of the way
* Check that the search base is serviced by a real backend before we'll ↵Nalin Dahyabhai2010-11-221-0/+9
| | | | attempt to participate in a search request.
* - more loggingNalin Dahyabhai2010-11-191-0/+24
|
* - more debug loggingNalin Dahyabhai2010-11-191-10/+18
|
* - initialize map_supported so that when we debug log we output it rightNalin Dahyabhai2010-11-191-0/+1
|
* - do a full clear of some temporaries (valgrind)Nalin Dahyabhai2010-11-181-0/+2
|
* - don't walk off the end of the filter array, but expect it to be of theNalin Dahyabhai2010-11-181-1/+1
| | | | right length
* - add optional caching of filters to attribute listsNalin Dahyabhai2010-11-182-19/+51
|
* - use actual slapi_filter structures to avoid reparsingNalin Dahyabhai2010-11-182-4/+14
|
* - manage filters along with chains of attributes and search basesNalin Dahyabhai2010-11-182-4/+23
|
* - return unwilling-to-perform instead of insufficient-access in responseNalin Dahyabhai2010-08-021-2/+2
| | | | to write requests
* - note whether or not the map is "secure" regardless of whether or notNalin Dahyabhai2009-06-181-5/+5
| | | | we were able to find a matching key
* - actually send portmap registrations to the right serverslapi-nis-0.17Nalin Dahyabhai2009-05-141-1/+1
|
* - when not configured to use a specific port, bindresvport to a datagramNalin Dahyabhai2009-05-141-3/+4
| | | | | socket so that we avoid trying to use the same one that we might be using for a portmap client socket
* - try to set the reuse-address socket option at the right level (socket,Nalin Dahyabhai2009-05-141-2/+2
| | | | not IP) this time
* - changes to commentsNalin Dahyabhai2009-05-071-4/+5
|
* - remove that debug message, it's redundantNalin Dahyabhai2009-05-071-1/+0
|
* - log more detailsNalin Dahyabhai2009-05-071-9/+30
|
* - make sure that %referred() and %deref() don't return empty stringsNalin Dahyabhai2009-05-071-8/+10
|
* - remove unused variablesNalin Dahyabhai2009-05-063-7/+2
|
* - revert to registering plugins at init-time, which is what the docsNalin Dahyabhai2009-05-064-66/+90
| | | | | | say we should do - use whether or not the plugin_base is initialized as in indicator of whether the plugin's been started or not
* - register our internal pre-/post-/internalpre-op plugins atNalin Dahyabhai2009-05-062-78/+92
| | | | startup-time, so that the hooks will only be used if we're enabled
* - add the list of function parameters to debugged invocation failure logNalin Dahyabhai2009-04-301-2/+2
| | | | messages
* - fix format_link() to not corrupt the heap with some bad mathNalin Dahyabhai2009-04-301-29/+43
|
* - only clear the backend_data return pointer if we were given one -- it'sNalin Dahyabhai2009-04-301-2/+6
| | | | | an optional argument, and we were already checking for non-NULL later on when assigning it a real value
* - fix type mismatch warningsNalin Dahyabhai2009-04-274-25/+28
|
* - fix some crashes by dereferencing uninitialized pointers whenNalin Dahyabhai2009-04-271-0/+26
| | | | retrieving information from a map fails
* - make the defaults use recursive support for attribute valuesNalin Dahyabhai2009-03-041-4/+4
|
* - reject client requests > 64k in sizeNalin Dahyabhai2009-03-041-1/+13
|
* - assume that super-large fragments are an errorNalin Dahyabhai2009-02-121-0/+9
|
* - drop a debug messageNalin Dahyabhai2009-02-121-4/+0
| | | | - default to not using libtirpc
* - move client socket creation here, so that we don't have to know how toNalin Dahyabhai2009-02-122-183/+375
| | | | | | | | | | do it elsewhere - refactor the send/receive logic out so that it can be used for either rpcbind or portmap - try to connect to rpcbind over a local socket first, else assume we have to use portmap over IP - use our own rpcb encode/decode function to avoid mixing the one from libtirpc with the rest of libc's XDR functions -- THAT WOULD BE BAD
* - handle argument order and count changing for portmap registration functionsNalin Dahyabhai2009-02-121-115/+66
| | | | | - pass IPPROTO_XXX instead of SOCK_XXX to registration function - if we're registering/unregistering IPv6, also register/unregister IPv4
generated by cgit (git 2.34.1) at 2025-08-30 21:44:42 +0000