Firebird 2 Bug Fixes

(CORE-4302)     Lookup (or scan) in a descending index could be very inefficient for some keys.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-4214)     Global temporary tables (GTTs) were able to make references to permanent relations, which they must not.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-4144)     The error “context already in use (BLR error)” could occur when preparing a query with UNION.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-4117)     A COMPUTED BY field was being evaluated as NULL if used as an exception parameter directly.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-4051)     The engine was exhibiting a memory leak when sorting records larger than 32 KB.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-4036)     A bugcheck, or even database corruption, could occur when attempting to store long, uncompressible data into a table.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-4002)     The error message “index unexpectedly deleted” could occur in a database trigger ON COMMIT TRANSACTION.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3940)     An I/O error could occur during a GetFileSize operation.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3895)     High memory usage was exhibited when PSQL code was selecting from a stored procedure that modified some data.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3855)     Blobs, inserted into GLOBAL TEMPORARY TABLE ON COMMIT DELETE ROWS, could be placed into newly allocated pages even if there was enough free space on some existing data page.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3860)     UPDATE ... RETURNING ... would raise error -551 (no permission to update) for a column present only in the RETURNING clause.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2966)     Wrong results or unexpected errors could occur while sorting a large data set.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2165)     An unnecessary index read could occur when using a strict inequality condition.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-4127)     The server would crash instead of reporting the error “key size exceeds implementation restriction.”.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-4123)     The server could crash when executing a stored procedure called by a trigger.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-4093)     The server could crash while converting an overscaled numeric to a string.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3859)     Too many different dynamic events could crash the server.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-4058)     A remote stack buffer overflow could occur.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2349)     A false “Invalid SQLDA” error could occur.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2912)     An exception would occur when uppercasing a string with lowercase y trema (code 0xFF in ISO8859_1).

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-4011)     Start/stop status was not being detected for SuperClassic and SuperServer running under Red Hat and Mageia.

fixed by P. Makowski

    ~ ~ ~     

(CORE-4323)     The V.2.1.5 64-bit Windows zip kit was missing fbsvcmgr.exe.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3844)     Validation was failing to detect one specific case of index corruption.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3841)     A database could become corrupted after inserting rows.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3834)     Using a NATURAL JOIN with a derived table could crash the server.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3807)     The error “Invalid expression in the select list” could be raised unexpectedly if a string literal was used inside a GROUP BY clause in a multi-byte connection.

fixed by D. Yemanov, A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3791)     Performance would degrade when actively working with databases bigger than the available amount of RAM.

fixed by N. Samofatov, D. Yemanov

    ~ ~ ~     

(CORE-3761)     A conversion error would occur when using a blob as an argument for the EXCEPTION statement.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3736)     The WITH LOCK clause was being allowed for users with read-only rights on some table, causing others to be blocked from updating that table.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3730)     isc_dsql_exec_immed2() would lose input parameter values when they had been specified for a RETURNING clause.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3722)     IS NOT DISTINCT FROM NULL was not using an index.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3697)     A string truncation error could occur when selecting from a VIEW containing a UNION.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3691)     The foreign key error message was missing the constraint name in Firebird 2.1.4.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3690)     Ambiguous queries were throwing the wrong warning message.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3677)     Firebird's executables were exporting entry points, causing bugs when used in combination with some UDFs in embedded mode. The only exception is the Superserver binary, which must export the ISC API functions.

stopped by A. Peshkov

    ~ ~ ~     

(CORE-3658)     The fbsvcmgr utility was connecting to the server under the operating system user name, rather than value of the ISC_USER environment variable.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3631)     Duplicate records with NULLs were being checked incorrectly.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3579)     Dropping a table was blocked when a computed field depended on another column that was created after it.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3569)     CHAR(32767) was presenting in the XSQLVAR structure with a length of 32765.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3553)     A nested loop plan would be chosen instead of the sort merge for joining independent streams using keys of different types.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3549)     A database corruption could occur after the end of a session, reporting the error “page xxx is of wrong type expected 4 found 7”.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3515)     When validation detected index corruption, it was putting a “missing entries” message in the firebird log.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3510)     The gds_drop utility would print an error when trying to shut down a missing lock manager process.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3479)     The internal function ASCII_VAL() would raise an exception when the argument was an empty string, instead of returning 0 as it should.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3464)     A dirty page could be written without releasing (or downgrading) its page lock. The result would be a bugcheck indicating either “lock denied” or “lock conversion denied”.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3418)     A database trigger created as INACTIVE would actually be created active.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3409)     A segmentation fault could occur after upgrading from v.2.1.3 to v.2.1.4.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3397)     The intl and trace libraries had unresolved symbols.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3394)     A failed attempt to violate a unique constraint could leave a superfluous “lock conflict” error in the status vector.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3355)     DATE and TIMESTAMP would be compared wrongly if an index was used.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3353)     A predicate (blob_field LIKE ?) would describe the parameter as VARCHAR(30) rather than as BLOB.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3335)     Internal wrapping was occurring for the multi-byte blob SUBSTRING function and its boundary arguments, producing wrong results.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3296)     A “context already in use” error would occur when the simple CASE function had an operand that was a subquery.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3245)     A SUBSTRING operation on a long BLOB would truncate the result to 32767 if the optional third argument was not present.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3233)     LIKE, STARTING and CONTAINING predicates would fail if the second operand was 32KB or larger.

fixed by A. dos Santos Fernandes, D. Yemanov

    ~ ~ ~     

(CORE-3231)     The OVERLAY() function would fail when used on text BLOBs containing multi-byte characters.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3092)     ROW_COUNT was not being cleared before a singleton INSERT statement.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2422)     The server was failing to switch between multiple locations configured in TempDirectories when the first one ran out of free space.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1992)     A valid but complex UNION query could fail with the error “Bad BLR -- invalid stream for union select”.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3557)     The server would crash while preparing a query against a table that was in the process of being dropped.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3554)     The server would crash during prepare, or throw an incorrect parsing error, if the remotely passed SQL query was empty.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3524)     The server would crash while compiling a stored procedure that was in use.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3517)     The built-in function LPAD() would crash the server with a string passed as the second parameter.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3440)     The server would crash if isc_que_events() had no events queued.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3374)     The server could crash or corrupt data if SELECT WITH LOCK was issued against records not in the latest format.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3320)     Some forms of the MERGE syntax could cause the server to crash.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3625)     MON$IO_STATS was not reporting page writes performed asynchronously (at the AST level).

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3138)     An internal error or a crash would occur when accessing any MON$ table after altering its structure.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2286)     Selecting from MON$CALL_STACK from within a PSQL module could return no rows in some cases.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3724)     The server could hang when working with events.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3351)     A Windows client could put 10054 network error messages into the firebird log at connection time.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3721)     The multi-user server startup script (/etc/init.d) was picking up the ISC_ variables if they were set.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3447)     Code changes in the Linux ICU libraries beyond ICU v.4.2 were preventing the installation of collations.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3388)     Firebird would not build on Fedora Core.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3285)     100% CPU usage (endless loop) inside the lock manager.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3262)     The LIST function would truncate the last part of the result sometimes.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3244)     The POSITION function would give a wrong result with an empty string, if the optional third argument was present.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3228)     The RIGHT() function would fail with multibyte text blobs longer than 1024 characters.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3227)     The ASCII_VAL() function would fail if the argument contained a multi-byte character anywhere.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3208)     Recursive queries were prone to significant memory leaks.

fixed by D. Yemanov, V. Khorsun

    ~ ~ ~     

(CORE-3188)     The exception “Page 0 is of wrong type (expected 6, found 1)” was occurring when the exception condition did not really exist.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3148)     Dangerous code was discovered in SQZ_apply_differences.

fixed by D. Kovalenko, A. Peshkov

    ~ ~ ~     

(CORE-3137)     Partial rollback was possible for a selectable procedure that was also modifying data.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3115)     Bugs were found in the internal record compression routines.

fixed by D. Kovalenko, A. Peshkov

    ~ ~ ~     

(CORE-3091)     The built-in function POWER(X, Y) would not work when the X argument was negative and the Y value was a scaled numeric with scale 0.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3079)     A transaction that involved events would slow down inserts in the same transaction markedly.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3067)     Objects were not being unmapped when shared memory was closed.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3064)     Using both the procedure name and an alias for it inside an explicit plan would crash the server.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3041)     The engine could hang when working with a read-only database.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3034)     An unhandled exception condition could throw “Bugcheck 300 (can't find shared latch)”.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3016)     On disconnect, the engine could erroneously report “Fatal lock manager error: invalid lock id (0), errno: 0” in firebird.log.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3015)     Various occurrences of “Cannot initialize the shared memory region” errors were being reported.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2999)     The message “SCH_validate -- not entered” was being logged in firebird.log when a Classic server connection that had used events was finished.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2995)     The same error was being reported twice in the status vector.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2977)     The server worked incorrectly with indexed fields of type DATE in ODS 9.x and older databases.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2969)     RDB$SET_CONTEXT would not allow any variables to be overwritten once the number of variables had reached the 1000 limit.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2956)     Some problems were occurring with requests involving procedure parameters.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2943)     Errors could occur while parsing a recursive query with more than one recursive part.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2936)     The error “Wrong page type (expected 7 found N)” was being reported when it should not have been.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2916)     Error handling was found to be broken for the condition where a conversion error happened during index creation.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2879)     A sweep could raise the error “Page 0 is of wrong type (expected 6, found 1)”.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2871)     An outer ORDER BY clause would have no effect.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2858)     Memory could be trashed when raising some exceptions to signal failed security checks.

fixed by C. Valderrama C.

    ~ ~ ~     

(CORE-2856)     A non-NULL key in a unique index could not be found when the key was removed

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2833)     Changing data that affected an expression index would fail if that data contained references to null date fields.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2827)     The Prepare for complex interrelated metadata, where many triggers were indirectly involved in the operation being prepared, was very slow.

fixed by D. Yemanov

(CORE-2826)     Join condition would fail for UTF-8 databases.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2822)     The error “no current row for fetch operation” was being thrown when a subquery included a non-trivial derived table.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2820)     Queries with PLAN ORDER were exhibiting small memory leaks as a side effect of an earlier, major fix.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2815)     Tidy-up of a logic condition that could cause the page inventory page to be marked for change after it was already changed.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2785)     An improperly handled BLOB transliteration problem showed up under specific conditions, viz., where an input argument string in single-byte Cyrillic text, for storing as COMMENT with an object definition, overstepped the maximum 64 KB size of a non-initial BLOB segment when transliterated. Its surplus bytes were causing an immediate transliteration error instead of being deferred for inclusion in the subsequent chunk, as they should have been.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2783)     An access violation would occur when a recursive query was used as a subquery in the SELECT list and its result was specified as an ORDER BY criterion.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2720)     Divide and unary plus\minus operations were being performed with the wrong evaluation result.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2698)     Infinite wait was occurring in LocksCache::get.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2684)     Error “Page NNN is of wrong type (expected 7, found N)” was being reported erroneously.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2679)     Applications using the embedded server could hang while exiting.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2675)     Every blob transliteration would cause a small memory leak.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2640)     Under some conditions, the lock manager could fail to detect a regular deadlock and cause the server to hang.

fixed by V. Khorsun, D. Yemanov

    ~ ~ ~     

(CORE-2635)     A unique index could be corrupted at level 1 if it contained a lot of NULL keys.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2616)     Error “page <N> is of wrong type (expected 7, found 5)” could occur under load, giving the impression that something had corrupted the database. On restart, there would be no evidence of corruption.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2591)     High mutex wait ratio and degraded performance would start to show up after a period of normal performance.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2578)     Selecting RDB$DBKEY from a view with more than one table joined would return a conversion error.

fixed by A. dos Santos Fernandes, A Peshkov

    ~ ~ ~     

(CORE-2532)     Volume sizes in a multi-volume database were being assigned incorrectly.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2507)     CreateFile could fail intermittently.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2301)     A query such as select list(xxx, ';') from table into :xxx could cause a memory leak.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2289)     The wrong name was being reported for the referenced primary key when a foreign key violation occurred during foreign key creation.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2215)     A GROUP BY expression involving concatenation with an empty string could cause problems.

fixed by A. dos Santos Fernandes

(CORE-1752)     Results of a join with different collations varied depending on the execution plan.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1346)     LPAD() and RPAD() functions hit implementation restrictions when applied to more than one column in a statement.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1274)     When PLAN MERGE was chosen and the data types of the equality predicate arguments were different, the query produced wrong results.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1089)     Selecting from a view that used DISTINCT and LEFT JOIN returned records in the wrong order if the ORDER BY clause did not include columns from the right-side (non-mandatory) table.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-195)     Regression of an old bug, previously fixed in v.1.5.1, whereby a bugcheck 291 (cannnot find back record version) would occur when updating the same record that had already fired an action in a BEFORE UPDATE trigger. The regression that was reintroduced in v.2.0 was less destructive, insofar as it affected only the record that was physically first in the table.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3050)     A race condition in LocksCache::get() could lead to an AV in the engine.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2914)     The server would crash while restoring a database having an expression index that referred to a non-existent UDF.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2908)     The engine could crash or raise unexpected errors working with ODS 8.x database.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2900)     The server could crash while evaluating the DISTINCT aggregate.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2888)     A memory corruption could occur that caused incorrect query evaluation and could crash the server.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-3218)     A statement cancellation request could be silently ignored by the currently running SQL code.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3011)     The server could hang or crash while monitoring connections that were repeatedly attaching and detaching.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2993)     The fatal lock manager error “Invalid lock id (NNN)” could occur while working with monitoring tables on a highly loaded system.

fixed by V. Khorsun, D. Yemanov

    ~ ~ ~     

(CORE-2632)     Unexpected “Invalid BLOB ID” errors could occur when working with the monitoring tables.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3220)     The service function call isc_info_svc_get_users was returning the error message in the result cluster.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3205)     The DSQL function call isc_dsql_exec_immed2() was not returning the error codes for isc_stream_eof nor isc_sing_select_err.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3202)     The execute_immediate family of API calls family could crash the remote server.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-3199)     The remote protocol code related to events processing could go into an endless loop, causing CPU usage to go to 100 per cent and stay there.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2084)     A security problem was discovered in the Services API.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3131)     The WIN1257_LV (Latvian) collation for four letters (A, E, I and U) was incorrect.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2122)     Translation of large text blobs between UNICODE_FSS / UTF8 and other charsets had bugs.

fixed by A. dos Santos Fernandes, D. Kovalenko

    ~ ~ ~     

(CORE-2356)     On Windows, the “listener” process of Classic server could not create the necessary resources after restart if any worker process was present.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2107)     Establishing a TCP\IP connection to the Windows Classic Server could fail under high load.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-3019)     Superserver startup was failing on Linux Gentoo.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-3001)     The POSIX installation routine was failing to create the “firebird” user and group.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2919)     The POSIX installation script was ignoring non-standard ports.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2865)     A segmentation fault would occur when the Superserver binary on POSIX was restarted.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2845)     The build process on Solaris would stop with the message: “need to use SFIO”. On the whole, Solaris 10 no longer requires SFIO - the 64-bit builds do not suffer the issue with the maximum 255 file descriptors and there are simple work-arounds for 32-bit builds. See this article for details.

However, the issue still affects Solaris 10 3/05 through Solaris 10 11/06. Some patches are needed to correct the problem.

For the time being, the easiest solution is to leave the define in the code, but comment it with the relevant information for Solaris 10. The define can be removed if the patches have been applied or the user is running an unaffected version of Solaris 10.

Updated information concerning this issue has been inserted into common.h. Anyone building 32-bit Firebird on Solaris 10 will have to make an informed decision according to the version and build/patch level of Solaris before uncommenting the SFIO define.

fixed by P. Beach

    ~ ~ ~     

(CORE-2814)     On SPARC, the server was crashing in the routine map_sort_data with a bus error.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2572)     Locks of type LCK_page_space were being processed incorrectly on big-endian machines.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2564)     Bus errors would be thrown when working with the monitoring tables on RISC machines.

fixed by A. Peshkov

(CORE-2550)     Bus errors would be thrown when working with DB_KEY on big-Endian machines.

fixed by A. Peshkov

(CORE-2475)     In Classic, external table data was visible only to the session that had operated on it.

fixed by V. Khorsun

(CORE-2449)     An unexpected “lock conflict” error could be thrown instead of the appropriate exception.

fixed by D. Yemanov

(CORE-2444)     The engine could hang when too many attachments were registering their interest in events simultaneously and exhausting free space in the events table.

fixed by V. Khorsun

(CORE-2441)     The server could crash on executing an 'UPDATE OR INSERT' statement.

fixed by A. Peshkov

(CORE-2416)     Preparing a query with aggregation over a derived table could cause an access violation.

fixed by V. Khorsun

(CORE-2415)     Firebird would crash when temporary space was exhausted.

fixed by A. Peshkov

(CORE-2397)     Database corruption could occur if two different indexes on a table were dropped within the same transaction.

fixed by V. Khorsun

(CORE-2355)     LOWER() and UPPER() functions would misbehave when the byte length of the result was smaller than the input string.

fixed by A. dos Santos Fernandes

(CORE-2348)     Problems with transaction numbers overflowing 32-bit signed integer and corrupting databases were appearing again.

fixed by V. Khorsun

(CORE-2340)     Bugcheck 258 (page slot not empty) could occur under high concurrent load.

fixed by V. Khorsun

(CORE-2320)     Complex recursive queries might not produce all rows.

fixed by V. Khorsun

(CORE-2313)     A boundary condition could cause INF_* functions to invalidate the whole output buffer with isc_info_truncated at the beginning.

fixed by C. Valderrama

(CORE-2311)     A memory leak was possible in a WITH RECURSIVE query.

fixed by V. Khorsun

(CORE-2306)     Superserver would terminate abnormally when a thread start failed.

fixed by A. Peshkov

(CORE-2291)     Bugcheck 284 (cannot restore singleton select data) would occur in PSQL code involving some subqueries.

fixed by V. Khorsun

(CORE-1961)     Bugcheck 210 (page in use during flush) was being thrown during database validation.

fixed by D. Yemanov, R. Simakov

(CORE-1690)     Arithmetic exception, numeric overflow, or string truncation in UTF8 tables.

fixed by A. dos Santos Fernandes

(CORE-1647)     The file associated with an external table needed to be closed after use, even if it was still apparently in use by some cached (inactive) requests.

fixed by V. Khorsun

(CORE-2411)     The optimizer was choosing slower plans in v.2.0.5 and v2.1.2 than it would for the same queries in v.2.0.4 and v.2.1.1.

fixed by D. Yemanov

(CORE-1607)     A correlated subquery that depended on the UNION stream would be optimized poorly.

fixed by D. Yemanov

(CORE-2531)     The famous “cannot transliterate” error could be thrown when selecting data from the monitoring tables.

fixed by D. Yemanov

(CORE-2483)     The database permanent pool could get corrupted when working with monitoring tables.

fixed by D. Yemanov

(CORE-2482)     Monitoring tables data collection could become unstable when attaching to or detaching from database.

fixed by A. Peshkov

(CORE-2563)     It was possible to shut down the Superserver's main port (3050 by default) by sending a malformed packet of some special format, that would lead to a Denial of Service condition for new incoming connections. This exploit could be used by an unauthenticated client.

Reported 15-Jul-2009 by Core Security Technologies.

fixed by D. Yemanov

(CORE-2368)     An isc_cancel_events() call would be succeeded by an access violation if the event was not found.

fixed by V. Khorsun

(CORE-2437)     A buffer overflow could occur on a client during delivery of events.

fixed by A. Peshkov

(CORE-2272)     The server would start returning garbage when killing an attempt by an event to connect.

fixed by A. Peshkov

(CORE-1923)     “instsvc remove” would return 1 (failure), even when execution was successful.

fixed by D. Yemanov

(CORE-2249)     Error while building Firebird 2.1 on FreeBSD 7.0.

fixed by A. Peshkov

(CORE-2247)     On RISC platforms, message and descriptor buffers were not aligned properly.

fixed by A. Peshkov

(CORE-2282)     *truncate UDFs were broken for numbers smaller than -1.

fixed by C. Valderrama

(CORE-2281)     *round UDFs were broken for negative numbers.

fixed by C. Valderrama

(CORE-2329)     Discovered a significant source of performance degradation in comparison with V.1.5.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2326)     Committing a new user object (a view, for example) caused an access violation if a user-defined trigger had been applied to the system table RDB$RELATIONS.

It should be noted that no Firebird server version either supports, or retains after a backup and restore, any user-defined trigger on a system table. The strong recommendation against defining such triggers remains. The fix recognises one way that user interference with system tables can compromise internal operations and disarms it.

The ability to define “DDL triggers” through the regular DDL mechanisms is on the drawing board for V.3.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2242)     The engine was incorrectly populating integer containers in the blob parameter buffer (BPB) with integers in machine-local format, causing errors on Big Endian platforms.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2241)     If an ALTER TABLE ALTER COLUMN.. operation was performed on a table in the course of a bulk insert operation, minor index corruption could occur causing subsequent queries to return the wrong number of records. The bug was traced to legacy code in BTR\compress_root().

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2227)     Problems were occurring in some environments when creating triggers that referred to column names with accented characters.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2222)     Storing a text blob with a transliteration blob filter could cause an access violation in the engine.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2184)     Superserver could hang when multiple clients were creating tables simultaneously.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2182)     It was not possible to drop an existing UDF whose name was duplicated by the name of a new built-in function.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2173)     The server would crash after an abnormal disconnection if there was an open ExecuteStatement call.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2151)     When a temporary directory path had spaces within it, it was (wrongly) being truncated at the rightmost space.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2137)     A database restore could crash the server when the configuration parameter DummyPacketInterval was set explicitly.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2098)     A bug in the implementation of Global Temporary Tables (GTT) allowed a GTT to be referred to in COMPUTED BY field expression. Since GTTs do not store global data, such an attempt should have been before the engine could attempt to store dependency records for the expression.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2045)     A regression occurred at v.2.1, whereby references to non-existent system fields with blr_field were not being resolved to NULL, whereas a parallel change involving blr_fld was exhibiting the proper corrective behaviour.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2039)     Domain-level CHECK constraints were (wrongly) processing NULLs as “zero values”, i.e. null dates as “date zero” (1858-11-17), null numerics as zero, null strings as empty strings, etc.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2031)     The first record output from a query with a search condition on RDB$DB_KEY was being returned with NULL in the first column.

fixed by A. A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2026)     Temporary BLOBs (including implicit ones created when transliterating text from the BLOB character set to the connection one) could not be created while accessing a read-only database.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-2008)     Procedure parameters that were declared NOT NULL were nevertheless being stored as nullable in RDB$PROCEDURE_PARAMETERS.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2000)     Under high load conditions, the lock manager could report false deadlocks.

fixed by V. Horsun

    ~ ~ ~     

(CORE-1984)     Lock manager would report false deadlocks if one of the deadlock participants was in WAIT with a permitted timeout.

fixed by V. Horsun

    ~ ~ ~     

(CORE-1970)     A “Lock conversion denied (bugcheck 215)” error could occur. This fix is related to CORE-1984 and CORE-2000 (above).

fixed by V. Horsun

    ~ ~ ~     

(CORE-1963)     A server could crash on commit when privileges were being granted or revoked simultaneously from multiple connections.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1958)     When attempting to update the same record multiple times, a “Bugcheck 179 (decompression overran buffer)” failure could occur.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1930)     In a situation where a stored procedure was altered to remove output parameters and dependent procedures are not recompiled, the engine should properly track the dependencies and return an exception when the altered procedure is called. Instead, it was crashing.

fixed by V. Horsun

    ~ ~ ~     

(CORE-1506)     The server would crash with isc_dsql_execute_immediate and a zero-length string.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1596)     There was a length-related bug in CsConvert::convert that had to be revisited.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1982)     Simultaneous backups or restores using the Services API under Superserver could interfere with one another.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2209)     Monitoring requests in high load conditions could become very slow and even block other activity during that time.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2171)     The column MON$CALLER_ID of table MON$CALL_STACK was reporting invalid IDs.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2017)     Monitoring tables were not keeping account of I/O statistics for stored procedures.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1944)     Monitoring tables contained wrong data on big-endian machines.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1926)     MON$DATABASE was returning outdated transaction counters.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2078)     The optimizer always had some trivial heuristics to estimate the effective stream selectivity, even if no indices could be used for the retrieval. This code missed being migrated into the ODS11 optimizer logic. The effect was that join orders chosen for cases involving non-indexed predicates were likely to be ineffective.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2053)     Computed expressions may be optimized badly if used inside the RETURNING clause of the INSERT statement

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2073)     The implementation of expression indexes exhibited a bug whereby an incorrect result was returned when an inverted Boolean predicate was applied to test an indexed expression.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2118)     With an UPDATE OR INSERT statement, an insert would fail silently if the MATCHING operand was a subquery.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1962)     Incorrect extraction of MILLISECONDs:

select extract(millisecond from time '01:00:10.1234')
from rdb$database
        

would return 10123 instead of 123.4.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2117)     Incorrect ROW_COUNT values were being returned with indexed retrieval and subqueries.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1919)     Memory corruptions in EXECUTE STATEMENT could crash the server.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-2087)     When the configuration parameter RemoteBindAddress specified a hostname instead of an IP address, or specified a non-existent IP address, it would be silently ignored and the server would bind to all interfaces, without any notification in firebird.log or the system log. This was considered a potential security risk if the system had ports open to the Internet. Now, an invalid or unavailable IP address will be resolved to localhost (127.0.0.1).

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2084)     Services API security problem, specific to Firebird 2.1: services would ignore the setting of the Authentication parameter in firebird.conf. Once a user was logged, any server-level operations not requiring a further login, such as viewing the firebird.log file or getting information about the server, etc., were accessible by any valid domain user.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2055)     Backported a fix for a known buffer overflow in the Firebird client library.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1972)     A non-SYSDBA user was able to change the Forced Writes mode of any database, along with several other database characteristics that should be restricted to the SYSDBA. This long-standing, legacy loophole in the handling of DPB parameters could lead to database corruptions or give ordinary users access to SYSDBA-only operations. The changes could affect several existing applications, database tools and connectivity layers (drivers, components). Details are in Chapter 3 of the V.2.1.2 Release Notes, in Changes to the Firebird API and ODS.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1957)     Because of a change done in the conversion to C++ at v.1.5, ACLs (Access Control Lists) longer than about 20 characters were being truncated. This has caused particular problems for applications that construct access privileges in run-time and has also given rise to privileges “going missing” when there are more than about 2000 privileges (for a report of the latter, see Tracker issue CORE-216).

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1922)     Trusted authentication would not work with the Services API.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2221)     On POSIX platforms, any attachment to any database would fail after the access rights for security2.fdb were modified from 0660 to 0666.

fixed by P. Beach, A. Peshkov

    ~ ~ ~     

(CORE-2157)     Known issue: a bug in gcc 3.2.x, the compiler used to build the official x86 Linux packages, can cause problems when people try to build binaries that depend on the Firebird client without using the -pthread switch. Setting the -pthread switch removes the dependency of the output binary on libpthread.

Reported by A. Peshkov

    ~ ~ ~     

(CORE-2077)     On POSIX platforms, the Classic server in embedded mode, i.e., loaded into the user's application space, would handle the TERM signal but would fail to call any other signal handlers in the queue. The effect was that signal handlers set by the application were not executed and the application would keep working after the termination. It was a bad idea to invoke ISC_signal_cancel() from the signal handler and the mechanism has been reworked.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2050)     Fixed a performance regression resulting from a surfeit of semop() system calls.

fixed by V. Horsun

    ~ ~ ~     

(CORE-2049)     Fixed a performance regression resulting from a surfeit of sigprocmask() system calls.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2033)     The symbol _Unwind_GetIP in the client library was being left unresolved due to a missing static library linkage.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1983)     In any POSIX environment except Solaris, the engine was mishandling the “out of memory” condition, causing the server to crash.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1909)     Garbage text was being printed to firebird.log on AMD64 Linux.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2093)     Superserver startup was failing on Solaris 64-bit.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-2234)     Sometimes, terminated worker processes in Classic on Windows were still considered to be alive after termination, due to improper checks on the Firebird server's part. The same bug could cause the Firebird server to misbehave with prolonged deadlocks when under load.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-2108)     When using the new implementation of Windows local protocol (XNET), the next available map number was calculated incorrectly, thus allowing the server to try to reuse a map number that already existed. If the “new” map's timestamp was equal to the timestamp of the pre-existing map, it would cause the get_free_slot() function to fail.

fixed by V. Khorsun

    ~ ~ ~     

Unregistered bug     When Firebird is configured to run in some specific directory (/usr/local/firebird, /opt/firebird or any other) the @prefix@ macro should be substituted with that directory path. On MacOS it was not done and caused exceptions to be thrown when the engine tried to locate some of its components.

fixed by P. Beach

    ~ ~ ~     

(CORE-2065)     The MacOSX installation package was in violation of platform rules by not including the client library in the dynamic loader search paths.

fixed by P. Beach

    ~ ~ ~     

(CORE-1890)     The database monitoring process could hang under high load.

fixed by D. Yemanov, V. Khorsun

    ~ ~ ~     

(CORE-1881)     Database monitoring could crash the server or mess up its page-locking logic.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1889)     Builds were creating the security database with Forced Writes OFF in all Firebird versions, allowing the possibility of corruption in unstable environments. From 2.1. 1 onward, builds will create security2.fdb with FW ON.

Improvement implemented by A. Peshkoff

    ~ ~ ~     

(CORE-1826)     The changeRunUser.sh and restoreRootRunUser.sh scripts were not changing the run user in init.d scripts.

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1887)     New databases were being created with the wrong access privileges.

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1854)     The value of CURRENT_USER was not always being converted to upper case when using operating system authentication on POSIX.

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1845)     Some standard calls would show the server installation directory to regular users.

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1885)     Processing a CREATE COLLATION statement on POSIX 64-bit versions would cause the server to crash.

fixed by A. dos Santos Fernandes, A. Peshkov

    ~ ~ ~     

(CORE-1802)     Wrong maximum key size was being calculated when the collation used was PXW_CSY.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1884)     Random crashes could occur when using stored procedures that were defined with input parameters having expressions as the default value.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1876)     

fixed by N. SamofatovThe nbackup incremental backup utility was broken.

    ~ ~ ~     

(CORE-1875)     An isql script that accessed CURRENT_DATE could throw an error or crash the server.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1864)     In isql on big-endian hosts, the SELECT operator did not work.

fixed by A. Peshkoff, P. Beach

    ~ ~ ~     

(CORE-1868)     The server could crash in isc_dsql_free_statement().

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1859)     “Arithmetic overflow or division by zero” errors were occurring while processing calls to the MAX() function.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1839)     The server could crash when sorting on a field that was calculated using a recursive common table expression (CTE).

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1798)     RDB$DB_KEY was evaluated as NULL in INSERT ... RETURNING

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1793)     On preparing a parameterised CTE with unused parameters, the server would crash.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1781)     LIKE, STARTING WITH and CONTAINING could return false positives.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1851)     Windows applications using fbembed.dll could hang during the process termination if there were open connections.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1844)     The Valgrind tools were frequently reporting the error “Conditional jump or move depends on uninitialised value(s)” in the function check_status_vector().

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1841)     A view using derived table names or aliases that were too long could cause RDB$VIEW_RELATIONS.RDB$CONTEXT_NAME to overflow.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1840)     Each DDL execution was producing a small memory leak.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1838)     SET STATISTICS INDEX on an index of a global temporary table (GTT) could wrongly change the index ID by an amount equivalent to the maximum available number for the database page size.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1830)     Index corruption was possible when multiple updates of the same record were performed in the same transaction with savepoints in use.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1819)     A more efficient solution was implemented for a bug that had caused lower-level index pages to be missed in indirection from the parent page (replacing the original fix for CORE-1300).

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1812)     Indexes were being overlooked for some date/time expressions in dialect 1.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1807)     On POSIX, the fbserver process was getting assigned to a random, non-canonical port after an abnormal termination.

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1731)     The engine would sometimes engine hang for several minutes, using 100% CPU load, with no sign of I/O activity.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1421)     Superserver failed to shut down immediately after a shutdown request if the request was preceded by a failed login attempt.

fixed by A. Peshkoff

    ~ ~ ~     

(CORE-1817)     The RelaxedAliasChecking parameter was having no effect on the RDB$DB_KEY field

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1810)     

fixed by A. PeshkoffThere was an issue with user names containing the '.' character.

    ~ ~ ~     

(CORE-1775)     The performance of security class checking during prepares was poor compared with v.1.5.x.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1357)     The DummyPacketInterval mechanism was broken.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1713)     Legacy behaviour of SELECT DISTINCT * on a table structure containing BLOB columns has been reverted and, along with it, the legacy behaviours when UNION, GROUP BY and ORDER BY operations refer to BLOB columns.

Earlier changes CORE-859 and CORE-1530 are thus also implicitly rolled back.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1724)     Common table expressions could not be used in computed columns and quantified predicates (IN / ANY / ALL). Now the behaviour of CTEs in SELECTS is aligned with that of other virtual table types.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1716)     Variable initialization in recursive procedures was being performed wrongly.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1729)     The server process would terminate when selecting the MON$ data was attempted in a system with a heavy load of concurrent connections.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1818)     Temporary files used for temporary page spaces were not being deleted after use on POSIX platforms.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1728)     After a fresh Linux install, monitoring tables would be found to be not working.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1726)     Failure was exhibited in isc_service_start().

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1746)     It was (wrongly) possible to create an expression index while inserts into the table were under way.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1715)     A “key size exceeds implementation restriction for index” exception was being wrongly returned under some conditions.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1694)     A bug was exhibited in creating or altering database triggers, related to attempts to include comment statements using a Russian character set.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1657)     Memory access violations (segfaults) could occur where read-only read committed transactions were left idle for a long time in a system under heavy load.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1648)     RDB$TYPES was failing to list the types for database triggers.

fixed by A. Dos Santos Fernandes

    ~ ~ ~     

(CORE-1644)     Compiling on GCC 4.1.1 was throwing a compilation error.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1641)     CREATE TRIGGER was causing two copies of the trigger object to exist in the metadata cache simultaneously.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1624)     MERGE worked incorrectly when there were parameters in a MATCHING clause

fixed by A. Dos Santos Fernandes

    ~ ~ ~     

(CORE-1610)     A full shutdown could cause database corruption.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1597)     It was (erroneously) possible to create global temporary tables in a database with an ODS lower than 11.1.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1579)     On 64-bit builds, a UDF that carried a BLOB as an argument could corrupt the BLOB on the stack by partly overwriting it if another argument was a string.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1574)     A number of problems appeared concerning multi-file databases.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1562)     Shutdown would fail to kill high-load query connections.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1549)     A performance-killer regression dating back to v.1.5 meant subquery-based predicates (IN, EXISTS, etc.) were not evaluated early enough in the join order to enable efficient filtering.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1539)     A transliteration error could occur in a SELECT statement on a system table, where the wildcard ('%') element of the search argument for 'LIKE' substituted metadata elements containing non-ASCII characters, e.g., in a CHECK constraint.

fixed by Adriano Dos Santos Fernandes

    ~ ~ ~     

(CORE-1533)     A JOIN on the first record of an ordered derived table would return the wrong record.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1509)     Extraneous trailing spaces were being inserted when casting DOUBLE PRECISION numbers to VARCHAR().

fixed by A. Dos Santos Fernandes

    ~ ~ ~     

(CORE-1501)     SLONG data in dsql_nod was being accessed wrongly.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1500)     Data in the internal buffer of EXECUTE STATEMENT was not being aligned correctly.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1492)     Compatibility between a text BLOB and a CHAR or VARCHAR was not working for COALESCE().

fixed by A. Dos Santos Fernandes

    ~ ~ ~     

(CORE-1681)     Garbage data in the incoming remote packet could crash the server

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1649)     Access violations could occur when a recursive query used the MERGE JOIN method in its execution plan.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1601)     The server could crash inside the page validation routine.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1519)     Memory access violations were occuring in isc_dsql_fetch().

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1199)     Internal GDS software consistency check (CCH_precedence: block marked (212))

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1589)     The start-stop script for SuperServer on Linux sometimes failed to stop the service.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1602)     Trusted authentication was not mapping domain administrators to the SYSDBA login.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1593)     Windows rules for full domain user names allow names longer than the 31 characters allowed by Firebird. Firebird was accepting such names via trusted authentication and failure was messy. Now, the 31-character limit is enforced and logins passing longer names are disabled. (This will remain the situation until the mapping of OS objects to database objects is implemented in a later Firebird version.)

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1543)     A bug in the remote interface checking for trusted authentication was enabling any OS user to log in as SYSDBA.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1679)     The isc_service_query() function was returning garbage bytes in the output.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1651)     The client library could sometimes falsely report a request synchronization error (isc_req_sync) to a client application.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1510)     XSQLVAR [NULL flags] for (2*COALESCE(NULL,NULL)) were bad.

fixed by A. Dos Santos Fernandes, D. Kovalenko

    ~ ~ ~     

(CORE-1594)     An alignment issue in character set conversion could cause extraneous trailing spaces on some platforms. Code CsConvert::convert() was refactored slightly to address the issue.

fixed by A. Dos Santos Fernandes

    ~ ~ ~     

(CORE-1375)     Wrong mapping for 0x212C in cs_gb2312.h

fixed by A. Dos Santos Fernandes

    ~ ~ ~     

(CORE-1642)     Monitoring under non-privileged login credentials would report wrong attachment data.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1584)     Inserting into monitoring tables could cause a crash or a bugcheck.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1567)     Regression behaviour exhibited in Beta 2 Embedded, where selecting from MON$STATEMENTS would yield an error and an exit() call.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1561)     Selecting from MON$STATEMENTS could cause the server to lock up.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1551)     A memory access violation would occur if all statements and all attachments were cancelled via the monitoring tables.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1603)     

fixed by A. PeshkovA long user name was a potential source of buffer overflow.

    ~ ~ ~     

(CORE-1476)     Forced writes have never actually worked on Linux, leaving open the potential for system trauma to break databases even with FW=ON. It has actually been known to happen on Linux.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1468)     Database corruption was possible when database file expansion and read\write activity were being performed simultaneously.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1440)     Transaction options were dangerously lacking in validation.

fixed by C. Valderrama

    ~ ~ ~     

(CORE-1418)     Rapidly starting and shutting down could cause a race condition in the blocking AST thread due to poor synchronization.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1401)     Instances of a global temporary table were not always picking up all indices.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1361)     Index operations for global temporary tables were not visible to the active connection.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1380)     Changing the Forced Writes setting for a database would cause I/O errors if the database had existing attachments.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1408)     UDF names using reserved words were being extracted with the double quotes missing.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1379)     Invalid parameter type ("Data type unknown" error) when passing the argument to the CHAR_LENGTH function as a parameter.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1347)     Unexpected "cannot transliterate" error.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1332)     The SQLSCALE member of a text BLOB column can carry the BLOB's character set. In some documentation it wrongly says it should always be there. Text BLOBs needed to be brought into line with character types, i.e., if the connection character set is other than NONE and the BLOB's character set is not NONE or OCTETS, then it should be the character set of the connection.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1470)     The server would crash if a secondary file name was longer than 127 characters.

fixed by C. Valderrama

    ~ ~ ~     

(CORE-1457)     The server would crash when attempting to deliver events for a session that had just disconnected.

fixed by V. Khorsun, D. Yemanov

    ~ ~ ~     

(CORE-1451)     Using RDB$DB_KEY in the WHERE clause of a SELECT from a stored procedure would crash the server.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1338)     Connection lost (error 335544721, Unable to complete network request to host ...) when selecting from a view having a derived field defined with ROUND().

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1334)     Joins with a NULL RDB$DB_KEY would crash the server.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1456)     Wrong events delivery could occur where there were concurrent XNET connections.

fixed by V. Khorsun, D. Yemanov

    ~ ~ ~     

(CORE-1443)     On 64-bit Windows 2003 Server, the embedded engine could cause an application to hang on exit if no database access was performed.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1403)     The server under Windows would crash if multiple events were being registered simultaneously by a client connected via the XNET protocol.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1395)     CHECK constraints on domains were demonstrating a few problems.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1378)     A number of issues were reported regarding domain names and character sets.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1466)     The SUBSTRING() function could return a truncated substring for some multi-byte BLOBs.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1428)     Timestamp subtraction in dialect 3 was incorrect if the calculation would result in a negative number.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1417)     Error “Invalid BLOB ID” error could occur when performing an insert using InterBaseXpress.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1373)     A recursive CTE query would produce incorrect results when the recursive member's SELECT list contained an expression involving self-referencing fields.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1434)     EXECUTE STATEMENT was truncating the last two bytes of VARCHAR columns.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1419)     CURRENT_TIMESTAMP was being wrongly evaluated during the execution of selectable procedures.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1371)     An EXECUTE BLOCK sequence would fail if it was passed within an EXECUTE STATEMENT string.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1370)     Use of CTE within procedures was causing memory leaks.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1331)     Character set transliterations would not work with EXECUTE STATEMENT.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1455)     Crash in fbclient after an unsuccessful user management API call.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1452)     The client library would crash when attempting to process an event received just before disconnection. (Did not affect libfbembed.so.)

fixed by D. Yemanov, V. Khorsun

    ~ ~ ~     

(CORE-1430)     Access Violation in fbclient.dll if a statement was prepared and executed right after events were registered.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1388)     It was not possible to attach to the Service Manager remotely if the remote engine version was less than 2.0.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1349)     The remote interface was not validating the client-supplied message length against the message format length.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1485)     Fixed a very ancient bug whereby the sqllen field in the xsqlvar contained length of data in a varying structure, not its total size. In any OS environment it could cause an access violation when loading messages in msg.fdb

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1416)     Incorrect parameter order in a TPB was being accepted without returning an error.

fixed by C. Valderrama

    ~ ~ ~     

(CORE-1372)     If isc_dsql_fetch() is called after isc_commit_transaction() an exception should be raised. That was not happening.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1484)     INTL modules compiled in the Linux, gcc 4.1.2, amd64 environment would cause an access violation in Superserver, due to use of the standard operator new but the overloaded operator delete.

fixed by A. Peshkov

    ~ ~ ~     

(CORE-1446)     Problem with UNICODE collations from fbintl when using system ICU.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1431)     There were some inherent issues with uppercasing certain Greek characters in cp1251.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1384)     LIKE would not work correctly with collations using SPECIALS-FIRST=1.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1339)     The metadata character set upgrade script was generating garbage in descriptions.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1467)     A database attachment would go into some kind of invalid state after its long-running statement was canceled via MON$STATEMENTS, returning a 'database shutdown' error.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1441)     Query cancellation feature could not interrupt a long fetch.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1436)     Outer joins would not work properly with the MON$ tables.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1359)     The server would crash at the first operation with the monitoring tables if the filesystem lacked the necessary permissions for the shared-memory file.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1358)     Operations with MON$STATEMENTS were throwing "cannot transliterate" errors.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1330)     Semaphores were being double-locked when the monitoring tables were queried during long fetches.

fixed by D. Yemanov

    ~ ~ ~     

(CORE-1447)     Querying for database info on very long path through an isc_database_info() API call could cause a buffer overrun.

fixed by C. Valderrama

    ~ ~ ~     

(CORE-1397)     A possible vulnerability was discovered in the remote server attachment.

fixed by V. Khorsun

    ~ ~ ~     

(CORE-1312)     A remote attacker could check for the presence of a file on a system running the Firebird server.

fixed by A. Peshkov

    ~ ~ ~     

( CORE-1248)     Incorrect timestamp arithmetic was performed when one of the operands was a negative number.

fixed by V. Khorsun

    ~ ~ ~     

( CORE-1228)     Reports of database corruption after an out-of-disk-space condition.

fixed by V. Khorsun

    ~ ~ ~     

( CORE-1227)     LIST() function would seem to fail if used twice or more in a query.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1215)     Wrong SELECT query results using index to evaluate >= condition

fixed by V. Khorsun

    ~ ~ ~     

( CORE-1175)     Error “Data type unknown” when any UDF argument was a built-in function containing a DSQL parameter reference.

fixed by D. Yemanov

    ~ ~ ~     

( CORE-1244)     Server crash on “select * from <recursive CTE>”.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1207)     FB embedded would not load without extra OS privileges.

fixed by V. Khorsun

    ~ ~ ~     

( CORE-1240)     Any task on Darwin PPC that used libfbclient would hang on exit.

fixed by A. Peshkov

    ~ ~ ~     

( CORE-1223)     Wrong message in firebird.log on Open SuSe Linux 10.2 : Open file limit increased from 1024 to 0.

fixed by V. Khorsun

    ~ ~ ~     

( CORE-1292)     CREATE TABLE failed if using long username and UTF8 as attachment charset.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1271)     Engine was allowing creation of invalid procedures and triggers.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1183)     View could not be created if its WHERE clause contained IN <subquery> with a procedure reference.

fixed by D. Yemanov

    ~ ~ ~     

( CORE-1162)     Problem altering numeric field type.

fixed by C. Valderrama

    ~ ~ ~     

( CORE-1253)     LIST(DISTINCT) was concatenating VARCHAR values as CHAR

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1153)     Activating an index change would cause “STARTING” to work as “LIKE” in a join condition.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1267)     Small bug with default value for domains in PSQL

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1256)     Table columns were hiding the destination variables for RETURNING INTO.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1165)     WHEN <list of exceptions> was tracking dependencies only on the first exception in PSQL.

fixed by C. Valderrama

    ~ ~ ~     

( CORE-1218)     isc_dsql_info(isc_info_sql_stmt_type) did not set isc_info_end at the end of the passed user's buffer

fixed by V. Khorsun

    ~ ~ ~     

( CORE-1196)     Long SQL statements were breaking the TCP/IP connection.

fixed by V. Khorsun, A. Peshkov, D. Yemanov

    ~ ~ ~     

( CORE-885)     It was impossible to revoke rights on update of a column.

fixed by A. Peshkov

    ~ ~ ~     

( CORE-856)     Could not set FName, MName, LName fields in the Security database to blank.

fixed by A. Peshkov

    ~ ~ ~     

( CORE-981)     x86_64 RPM package missing “provides”.

fixed by A. Peshkov

    ~ ~ ~     

( CORE-107)     An instance of fb_lock_mgr would be left running after a build.

fixed by A. Peshkov

    ~ ~ ~     

( CORE-1286)     Bug with COMPUTED BY fields.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

( CORE-1167)     Character set GBK was not getting installed.

fixed by A. dos Santos Fernandes

    ~ ~ ~     

(CORE-1079)     Every attach of fbclient/fbembed library to the host process leaks 64KB of memory

No information available.

    ~ ~ ~