dimitr
477af6bb96
Checking the datatype at compile time is unreliable: some nodes return one type at compile time and another type at runtime, thus causing unexpectedly missing casts. I'm moving the check to runtime, this should fix the remaining regressions (fingers crossed).
2016-02-24 20:11:34 +00:00
dimitr
04f5f59ff8
Yet another postfix for CORE-5097, it resolves regressions for core_1173.fbt and core_1525.fbt.
2016-02-23 08:20:02 +00:00
dimitr
fb49b002c6
Fixed the comment.
2016-02-21 09:02:11 +00:00
dimitr
67151a6b43
Corrected my old commit re. optimization of the system requests.
2016-02-21 08:54:31 +00:00
dimitr
ff1222c994
Fixed CORE-5220: Regression: ORDER BY clause on compound index may disable usage of other indices.
2015-11-17 10:47:59 +00:00
dimitr
7134690bf3
Fixed CORE-4984: Ordering by compound index together with a range condition gives wrong results.
2015-10-30 06:08:19 +00:00
alexpeshkoff
139331c7e3
Enhanced memory leaks search - changed FB_NEW to FB_NEW_POOL,
...
plain new to FB_NEW, making allocated with it blocks tagged with file/line.
Make related code slightly better readable.
2015-10-12 14:26:00 +00:00
dimitr
a14e993fbe
Fixed inconsistency between unique vs non-unique retrievals (see CORE-4702).
2015-03-06 14:04:07 +00:00
dimitr
12da25ce6d
Minor adjustments to the join order selection.
2015-03-06 10:24:29 +00:00
dimitr
8508fea434
Fixed CORE-4665: Wrong result when use "where <field_C> STARTING WITH <:value> ORDER BY <field_N>' and field_C is leading part of compound index key: { field_C, field_N }.
2015-01-12 11:28:10 +00:00
dimitr
b9178d4df1
Small refactoring.
2014-12-21 19:16:51 +00:00
alexpeshkoff
c14a1210b0
Work in progress on CORE-3632: make all file descriptors, opened by firebird, close on exec()
2014-12-17 14:31:02 +00:00
dimitr
263b396cd2
Postfix for CORE-1550: Unnecessary index scan happens when the same index is mapped to both WHERE and ORDER BY clauses.
2014-12-16 11:40:42 +00:00
dimitr
182390b49e
Fixed CORE-4530: DB_KEY based join of two tables may be ineffective.
2014-08-25 13:28:19 +00:00
dimitr
7d802db4f5
Avoid metadata lookups while retrieving the plan (especially important in ASTs).
2014-07-30 09:50:33 +00:00
dimitr
7f90a395c4
Fixed CORE-4498: FB 3.0 crashes when getting an explained plan for a DBKEY based retrieval.
2014-07-24 19:37:58 +00:00
skidder
7c743fc291
Fix build warnings
2014-07-17 18:48:46 +00:00
robocop
66af445e2e
Misc.
2014-03-22 20:51:24 +00:00
dimitr
ef040723f7
This should fix the regression reported in fb-devel.
2014-03-21 18:57:41 +00:00
dimitr
f9c9257fc9
Totally misc.
2014-03-08 07:26:08 +00:00
dimitr
bd410d08aa
Fixed refactoring errors that caused some optimizer regressions. Also, simplified code a little.
2014-02-09 15:58:33 +00:00
dimitr
994b04ea98
Fixed the assertion.
2014-02-06 09:52:16 +00:00
dimitr
39574ba899
Front-ported my recent commit re. possibly slow restore.
2014-01-26 15:27:15 +00:00
dimitr
3ee2025161
Fixed CORE-4118: Expression index may be not used for derived fields or view fields.
2014-01-06 09:24:46 +00:00
asfernandes
0ec5daac05
Misc.
2014-01-06 00:51:37 +00:00
dimitr
15113ae2dd
Resolved CORE-1482: Make optimizer to consider ORDER BY optimization when making decision about join order.
...
Also, improved (hopefully) cost calculation for navigational retrievals.
2014-01-05 19:40:07 +00:00
dimitr
a5cde2bbb4
Correction for CORE-1550: Unnecessary index scan happens when the same index is mapped to both WHERE and ORDER BY clauses.
...
Also resolved CORE-1846: Allow index walk (ORDER plan) when there's a composite index {A, B} and the query looks like WHERE A = ? ORDER BY B.
2013-11-28 17:26:54 +00:00
dimitr
74fbaf5ec1
Another solution for CORE-1550: Unnecessary index scan happens when the same index is mapped to both WHERE and ORDER BY clauses. It's intended to fix the reported issues.
...
Also resolved CORE-4285: Choose the best matching index for navigation.
2013-11-28 09:12:02 +00:00
alexpeshkoff
d6297bebd4
Fixed some gcc warnings
2013-11-19 13:19:11 +00:00
dimitr
8ee2b22263
Fixed the optimization improvement accidentally broken by a recent commit.
2013-07-29 06:56:44 +00:00
dimitr
d11d3be507
Sync'ed with v2.5 and fixed wrong stream comparison checks.
2013-07-07 16:11:28 +00:00
dimitr
6961e4066d
Fixed the assertion, thanks to Adriano.
2013-07-07 07:50:44 +00:00
dimitr
fec38d76b2
Slightly refactored the optimizer regarding navigational walks.
2013-06-30 17:33:02 +00:00
dimitr
a1d9ca6726
A few additional adjustments for CORE-1295: Bad optimization of queries with DB_KEY.
2012-11-27 17:36:13 +00:00
dimitr
b07ef45ae6
A few additional adjustments for CORE-1295: Bad optimization of queries with DB_KEY.
2012-11-27 17:22:42 +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
robocop
70d5ebdfdf
- Got rid of the fixed array that kept the stream count in the first element and 255 streams after it.
...
- Changed streams from UCHAR to ULONG as Dmitry wanted
- Got rid of some redundant flags to indicate BLR level
- Misc changes (for example, converted to test for DBB_read_only in Database to a function)
- BLR v6 is left commented (waiting for a better solution to appear, if any exists)
- This change is not going to raise the engine limits until a solution is found to express more than 255 streams in BLR.
2012-04-12 09:02:13 +00:00
asfernandes
a0194aeaf3
Simplificate CastNode. There is no need for a format there.
2012-03-17 16:06:05 +00:00
alexpeshkoff
4e795f4047
Include platform definitions file common.h in firebird.h. Use OS-specific ThreadId instead artificial FB_THREAD_ID, avoiding unneeded casts in the code. This should fix BSD and MAC ports.
2012-03-01 08:55:43 +00:00
asfernandes
a62c4b17d3
Fixed CORE-2457 - UNICODE_CI internal gds software consistency check.
2012-01-24 00:16:50 +00:00
asfernandes
42850d55b8
Misc.
2011-07-16 18:49:28 +00:00
dimitr
55b37f2648
Methinks this is a more correct logic than the one committed yesterday. Also, fixed the stream comparison for views.
2011-07-14 11:54:10 +00:00
asfernandes
03ce334e2b
Misc
2011-04-04 13:37:47 +00:00
robocop
82403a5621
Misc, style, etc.
2011-04-02 04:45:26 +00:00
dimitr
82b595ae7e
Misc adjustments.
2011-03-10 08:13:02 +00:00
dimitr
d97e9492a9
Fixed the regression I introduced recently. Temporary solution, to be reworked soon.
2011-02-26 09:02:19 +00:00
asfernandes
92fc79f8a9
Correction.
2011-02-22 15:29:21 +00:00
dimitr
af76dc8de8
Better solution for CORE-3355, as suggested by Vlad.
2011-02-22 12:43:14 +00:00
dimitr
b0e3c9f94d
Fixed CORE-3355: Wrong comparison of DATE and TIMESTAMP if index is used.
2011-02-21 11:18:29 +00:00
dimitr
993788f9bc
Misc optimizer tuning.
2011-02-21 10:09:30 +00:00
