dimitr
bfc9a9c929
Frontported CORE-3305: "BLOB not found" error after creation/altering of the invalid trigger.
2014-06-13 11:28:33 +00:00
hvlad
759bf6a592
Front ported fix for bug CORE-4444 : Engine could hung and block all attachments in out of disk space condition during physical backup
2014-06-10 09:31:35 +00:00
alexpeshkoff
6a2016235a
Fixed CORE-4435: After calling release() for attachment to database (instead detach) in embedded mode attachment remains active forever (and some cleanup re. reference counters)
2014-06-10 07:13:34 +00:00
hvlad
6a806677fd
Front ported:
...
1. Improvement CORE-4431 : Reduce contention for allocation table lock while database is in stalled physical backup state
2. Improvement CORE-4432 : Let attachments to not block others when allocation table is read first time
3. On Windows, file can not be deleted while system writes cached data into it, even if file is not open by anyone.
Therefore flush delta file implicitly before closing it.
2014-05-16 12:07:08 +00:00
asfernandes
40b38014e4
Misc.
2014-05-08 23:26:40 +00:00
alexpeshkoff
d83d4d084a
Fixed a bug, privately reported by Pavel Zotov - segfault when closing trace
2014-05-06 12:25:35 +00:00
alexpeshkoff
9bd0939f09
Implemented CORE-1377: Add an ability to change role without reconnecting to database
2014-04-30 15:12:12 +00:00
alexpeshkoff
a98e565571
Fixed CORE-4303: Possible races while Service destruction, related code cleanup
2014-04-25 10:59:34 +00:00
dimitr
62405ad22f
Fixed ON DISCONNECT triggers I accidentally broke more than a year ago.
2014-04-20 10:09:05 +00:00
asfernandes
913c435667
Misc.
2014-04-19 21:36:55 +00:00
alexpeshkoff
58ec2954c1
Fixed CORE-4394: Cursor not found error when using legacy API
2014-04-18 12:08:17 +00:00
alexpeshkoff
5bfe0dfa97
Fixed some errors in mapping code, reported by Vlad privately, when using serverwide auth plugin
2014-04-15 12:18:28 +00:00
dimitr
7c540ae0db
1) Prevent shutdown/cancellation for system attachments.
...
2) Fixed incorrect generation of attachment IDs causing a hangup during gfix -mode read_only.
2014-04-14 15:27:27 +00:00
asfernandes
8608223f53
Misc.
2014-04-13 01:42:10 +00:00
alexpeshkoff
82a3a54896
Fixed mapping errors reported by Claudio & Treeve
2014-04-10 16:43:10 +00:00
alexpeshkoff
8b28dd1da9
Implemented improvement CORE-4387: IStatement::execute() and IAttachment::execute() functions should return on error pointer to old transaction interface
2014-04-08 13:45:21 +00:00
hvlad
a0ddc6771a
Patch by Alex (re.CORE-3370 and deadlock in user mapping code)
2014-04-08 07:15:46 +00:00
alexpeshkoff
a66ac62755
Resolved first problem, reported by Vlad privately - it was bad idea to try to load cache from database which is not created yet
2014-04-07 09:28:08 +00:00
robocop
c90d9d4547
Misc.
2014-04-05 01:44:32 +00:00
alexpeshkoff
20e846369e
Implemented CORE-3370: Resolve additional aspects of multiple security databases from services and cross-database requests POV (mapping names)
2014-04-04 15:57:18 +00:00
robocop
fa2f26b645
A bit less repetitive code.
2014-04-04 07:51:24 +00:00
robocop
29e36dc4f3
The 0th generator becomes unnamed. Only internal requests can access it (zero length names are not allowed from outside). RDB$GENERATORS is no longer a reserved name.System generators cannot be changed by user requests, but can be queried with gen_id(g, 0) if the permissions (ACL) allow that. We don't track dependencies on sys generators, it's not necessary.
2014-04-04 06:39:09 +00:00
robocop
58654b05cf
Try to drop as many shadow files as possible.
2014-03-26 19:10:54 +00:00
robocop
66af445e2e
Misc.
2014-03-22 20:51:24 +00:00
hvlad
d125f5617a
Fixed warning by PVS-Studio
...
Review is appreciated
2014-02-25 14:18:32 +00:00
alexpeshkoff
26b090c93b
Added access to configuration interfaces from IMaster and support for getting configured directories
2014-02-12 11:20:55 +00:00
asfernandes
48ea42f1c0
Misc.
2014-01-19 20:38:16 +00:00
alexpeshkoff
411c3718ca
Fixed CORE-4286: "Statement already has a cursor assigned" error when trying to execute another SQL statement using different cursor name. Moved method setCursorName() from IStatement to IResultSet - it was my fault when splitting cursor from statement.
2014-01-17 12:33:23 +00:00
hvlad
5e3c596545
Fixed bug CORE-4309 : 'Cache Writer' record in MON$ATTACHMENTS dissapear when delete other connection that running some 'heavy update' on big table via delete from MON$ATTACHMENTS
2014-01-11 09:10:44 +00:00
robocop
2810026363
Misc and warnings.
2013-12-29 00:35:50 +00:00
asfernandes
2aa3e8dd4d
Working in progress on the fix for CORE-4271: unify as possible the handling of procedures and functions.
2013-12-12 14:42:31 +00:00
asfernandes
9c9dd72977
Misc.
2013-11-24 03:21:11 +00:00
alexpeshkoff
4a8fe5a0de
Implemented CORE-4273: Per-database control on access to database from remote stations
2013-11-22 15:21:11 +00:00
alexpeshkoff
c2efe1945a
Enable linger only in multiuser server mode. This should also fix windows build.
2013-11-21 07:31:30 +00:00
alexpeshkoff
83db69c849
Fixed new gcc 4.7 warning "deleting object of polymorphic class" in many places. Adjusted use of interface types (reference counted vs just vesrioned). Avoided some hacks caused by badly chosen interface type.
2013-11-18 12:20:23 +00:00
dimitr
8033131d18
Fixed possible server crash + minor refactoring.
2013-11-17 14:24:17 +00:00
alexpeshkoff
120b10a78a
Implemented CORE-4263: Database linger
2013-11-14 16:16:24 +00:00
dimitr
489480b774
Complementary patch for one of my prior commits (forced rollbacks should go through the TIP).
2013-11-13 15:39:32 +00:00
hvlad
37344e12bf
Fixed bug CORE-4260 : Bugcheck 165 (cannot find tip page) happens if engine crashed right after CREATE DATABASE
2013-11-11 14:44:52 +00:00
hvlad
34b99e6f3a
Don't force flush for database which creation is not complete.
2013-11-07 10:32:35 +00:00
hvlad
424fb1d75f
Avoid hack in create_database when we have FW=ON at header page disk image and FW=OFF in memory structures and in fact database file is created in FW=OFF mode.
...
Instead of setting flag on header page when it is formatted for the first usage, explicitly set FW=ON after database creation is complete, if there was no isc_dpb_force_write in DPB.
2013-11-07 10:06:36 +00:00
alexpeshkoff
ee748b192d
Make code (hopefully) better readable
2013-11-01 14:18:30 +00:00
alexpeshkoff
bd65abc1aa
Cleanup
2013-10-31 15:10:26 +00:00
alexpeshkoff
dd73db4494
Fixed CORE-4200: An uncommitted select of the pseudo table sec$users blocks new database connections
2013-10-30 12:11:40 +00:00
hvlad
88a470d22c
Fixed two more reasons for CORE-4235 : Deadlock is possible while accessing the monitoring tables under concurrent load.
2013-10-19 08:09:51 +00:00
alexpeshkoff
5a054cca16
Fix found by Jim Starkey man-in-themiddle attack on ATTACH packet.
...
That packet may contain database name, command line for service and/or some other important and critical information.
To avoid such attack wire encryption is started before attaching database or service manager, right after connect.
Also should fix a number of issues found in Alpha1 regarding services attach.
2013-10-16 12:53:04 +00:00
dimitr
c235774864
Attempted to fix CORE-4235: Deadlock is possible while accessing the monitoring tables under concurrent load.
...
Refactored the list of reference counted attachments to be useful in other cases.
Some minor adjustments in the monitoring code.
2013-09-22 16:10:06 +00:00
dimitr
3821ec9d41
Implemented CORE-4218 (database owner in MON$DATABASE) and CORE-4222 (auth method in MON$ATTACHMENTS).
2013-09-18 09:37:13 +00:00
alexpeshkoff
7c243418a9
Cleanup - removed unused (always NULL) field tra_sibling from transaction class in engine
2013-09-13 12:21:51 +00:00
dimitr
c9cc5294da
Avoid rescheduling if we're kindly asked to stop immediately.
2013-09-04 17:52:13 +00:00
dimitr
c367d1f958
Cleanup and use the explicit rollback via TIP for all kinds of forced disconnections.
2013-09-04 17:46:09 +00:00
dimitr
01b81b1edb
Code simplification and some corrections.
2013-08-19 18:50:28 +00:00
alexpeshkoff
1eace34614
CORE-3944 - fixed a bug in my previous commit, noticed by Adriano
2013-08-19 08:17:37 +00:00
alexpeshkoff
a8f8465366
Postfix for CORE-3935 & CORE-3944: helps avoid segfaults/deadlocks when shutting down firebird
2013-08-16 12:44:10 +00:00
asfernandes
b6a4d39596
Misc.
2013-07-06 23:11:13 +00:00
dimitr
9e8bbbdd13
Better protection against AST delivery during detaching.
2013-07-03 13:22:44 +00:00
hvlad
c12e731b6a
Fixed bug CORE-4134 : Race condition when auto-sweep is started
2013-06-29 11:32:55 +00:00
hvlad
ffa909f8d9
Fixed bug CORE-4135 : In SS sweep blocks establishment of concurrent attachments.
...
More details in tracker.
2013-06-29 11:25:07 +00:00
alexpeshkoff
7583535db5
Make dialect 3 default when creating database using OO API. Keep legacy behavior in ISC API.
2013-06-24 12:24:29 +00:00
alexpeshkoff
2a66ddf07f
Fixed assertion when restoring DB in non-existtent path with overwrite - reported privately by Dmitry
2013-06-18 09:10:09 +00:00
dimitr
0c6bc6844a
Protection against the not supported mode. I used a non-translatable error because it might be a temporary measure, time will show.
2013-06-15 14:12:25 +00:00
robocop
c0b8b68777
Misc.
2013-05-21 21:53:39 +00:00
alexpeshkoff
b9a96911ca
Avoid extra atomic operations when assigning pointer to the object with pre-incremented counter (like returned from the function) to reference-counted pointer holder
2013-04-17 11:16:06 +00:00
alexpeshkoff
6f5fedec9a
Fixes related with CORE-3993
2013-04-17 08:04:02 +00:00
alexpeshkoff
3746843930
Postfix for CORE-3935 and CORE-3993: avoid opening same database file twice
2013-04-15 16:51:20 +00:00
dimitr
551e2416c6
Add interface support for scrollable DSQL cursors. Implementation will follow later.
2013-04-13 07:11:58 +00:00
asfernandes
e0567463bf
Revert external routines to request-based (*) code, but now using a totally different internal approach.
...
* This is not ideal IMO, but much more easier to currently support stack traces, and domain's defaults and validations.
2013-04-11 15:44:17 +00:00
alexpeshkoff
8780552b94
Added support for changing configuration (client and embedded) from API
2013-04-04 13:17:57 +00:00
dimitr
dbe76bc342
The minimal patch for CORE-1604 (Error at database creation by user with non-ascii name) and CORE-3243 (CURRENT_USER and MON$USER don't work properly) that seems to work for the trusted authentication, but [at least] the legacy auth module has to be reviewed as it seems to deal with the user name "as is", without checking whether it arrived in UTF8 or not.
2013-03-24 16:16:29 +00:00
asfernandes
b9bc308a72
Work in progress fixing external triggers.
2013-03-23 21:12:27 +00:00
alexpeshkoff
afa0770781
1. Fixed broken by me a few time ago server shutdown.
...
2. Implemented prototype of TLS-based debugging helper holding information about 'What does firebird do now?"
Currently it's more or less ready for use with JAttachment::release().
2013-03-15 16:37:32 +00:00
asfernandes
0d299d455a
Remove FbMessage struct from the API.
...
This struct lost its main value (group various blr-related properties in a single parameter).
And since openCursor/fetch already split the message/buffer, it does not make sense anymore to have it.
2013-02-26 03:42:19 +00:00
asfernandes
b254fb11c7
Revert change of messages to void* in DSQL internals.
2013-02-24 16:26:00 +00:00
asfernandes
53d4d58113
Change message buffers type to void*. It adds some casts, but avoid them in users' code.
2013-02-23 03:09:17 +00:00
asfernandes
5c436f8d91
Misc.
2013-02-21 23:25:10 +00:00
alexpeshkoff
f1779a7398
Do not prefetch unneeded information in IAttachment::openCursor()
2013-02-19 12:03:07 +00:00
alexpeshkoff
ba9413b26a
Add IStatement::getFlags() to help user make a solution what to do with SQL statement
2013-02-19 11:20:49 +00:00
alexpeshkoff
37d353d392
Make openCursor() use default output format when none is given explicitly
2013-02-18 13:53:46 +00:00
asfernandes
ddc4d4eac3
Misc.
2013-02-17 23:06:40 +00:00
alexpeshkoff
fc12495d0e
Modified IStatement interface in our API:
...
1. Never allocate empty statement - always use att->prepare() to create statement interface
2. Separated IStatement into 2 parts - statement itself and resultset.
3. Added stmt->openCursor() (and att->openCursor() for unprepared statements) to create IResultSet.
4. Always use IMessageMetadata (former IParametersMetadata) to pass message fromat info from client,
therefore avoiding need in BLR generation in client applications.
2013-02-17 12:08:53 +00:00
asfernandes
dc8e42e9bb
Migrate external triggers to the new message-based style.
2013-01-29 01:21:32 +00:00
asfernandes
d9275d6567
Misc.
2013-01-07 01:15:24 +00:00
dimitr
e29b50f980
Front ported my yesterday's commits to trunk.
2012-12-23 07:10:56 +00:00
asfernandes
7e1ca71911
Misc.
2012-12-21 01:49:16 +00:00
alexpeshkoff
b3f9759960
Take locks in AST before building tdbb
...
Fixed assertion in CCH_exclusive_attachment
Rolled back my changes required to let cancelOperation() enter engine after shutdown
Better locks' history in SyncObject
2012-12-18 13:54:18 +00:00
alexpeshkoff
c7454a49a5
Work in progress (committed due to SF upgrade). Checked only on SS.
...
1. Frontported fix for CORE-3935 and CORE-3993.
2. Added debugging support for mutexes and rwlocks in Vulcan style.
Unfortunately after last 'svn up' build asserts in JrdStatement:71.
2012-12-14 17:59:02 +00:00
dimitr
4ff4290f11
Front-ported bugfixes related to CORE-4004: Sometimes long-running operations cannot be interrupted by asynchronous shutdown / cancellation requests.
2012-12-03 11:21:00 +00:00
dimitr
a85629a9de
Fixed CORE-3977: DELETE FROM MON$STATEMENTS does not interrupt a longish fetch.
2012-11-11 15:33:18 +00:00
dimitr
657511babe
Wiped out the legacy attachment level counters in favor of the already existing runtime statistics. Introduced a record lock counter (more to follow).
2012-11-01 00:16:55 +00:00
dimitr
9f2a992078
Implemented:
...
CORE-2780: Include client library version and protocol version in mon$attachments
CORE-2187: Return the Hostname in mon$attachments
CORE-3779: Report OS user name in MON$ATTACHMENTS
+ some cleanup.
2012-09-12 07:45:51 +00:00
dimitr
5f917a04ce
Minor cleanup and simplification.
2012-08-28 18:19:09 +00:00
alexpeshkoff
0fc893a4e1
Better names for API functions
2012-06-26 13:43:26 +00:00
dimitr
42a4c93b97
1) Removed the no longer actual "parent lock" concept from both the lock and event managers.
...
2) Reworked the lock manager synchronization: added the RAII mutex holders and avoided dirty reads for the shared memory.
3) Encapsulated lck_length into the Lock class + small related refactoring.
4) Front ported recent fixes from the v2.x branches.
5) Misc cleanup.
2012-06-21 15:37:38 +00:00
alexpeshkoff
2a01e4bcf9
Implemented CORE-3861: Make it possible to encrypt database
...
Also some cleanups, the most important are:
- meaningful ctor on Jrd::Lock, helping to avoid code dup
- avoid unneeded h-file dependencies, making boot build engine dependent
2012-05-31 16:53:42 +00:00
asfernandes
07557dc297
Misc.
2012-05-26 18:05:56 +00:00
robocop
642d506cd9
Change txn numbers to unsigned and commit after Vlad's review. If the compile-time checks (to ensure enough room for the new txn type) are too ugly, they may be deleted.
2012-05-19 12:04:37 +00:00
hvlad
d78703c9ae
Improvement CORE-3539 : TRACE: add ability for logging ERRORs that occur in runtime (lock conflicts, PK/UK/FK violations et al)
2012-05-15 07:56:24 +00:00
hvlad
02adff3b63
Improvement CORE-3598 : TRACE: add statistics of actions that were after transaction finished
2012-05-12 09:48:32 +00:00
asfernandes
580172419b
Misc.
2012-04-23 00:48:28 +00:00
hvlad
536c08cc01
Fixed bug CORE-3825 : If an autocommit transaction (i.e. transaction started with isc_tpb_autocommit option) run DDL using EXECUTE STATEMENT it will produce a bugcheck 287 (Too many savepoints)
2012-04-20 20:03:15 +00:00
