Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
100 check-ins occurring around 0f0959c6f95046e8.
|
2021-06-02
| ||
| 18:50 | Fix an ALWAYS which can be false. (check-in: a87f8ecd user: drh tags: seekscan-improvements) | |
| 17:32 | Tighten assert() restrictions on the OP_SeekScan opcode. (check-in: dbc33d3a user: drh tags: seekscan-improvements) | |
| 17:14 | Never use WHERE_IN_SEEKSCAN for the 2nd or subsequent IN operator. (check-in: d0b98619 user: drh tags: seekscan-improvements) | |
| 12:44 | This is an attempt to address the OP_SeekScan performance issue identified by forum post b4fcb8a598. (check-in: d58efb3f user: drh tags: seekscan-improvements) | |
|
2021-06-01
| ||
| 21:07 | Fix issue in 'mkopcodeh.tcl' tool that prevented its termination in rare circumstances. (check-in: a8d92113 user: mistachkin tags: trunk) | |
| 15:37 | Fix a problem with running ALTER TABLE against schemas that contain compound SELECT statements with ORDER BY clauses containing one or more references to the second or subsequent component SELECT statements. This fixes the problem reported by forum post 31ceaa8461. (check-in: 587a3044 user: dan tags: trunk) | |
|
2021-05-29
| ||
| 23:07 | Do not confuse the ">" operator with IS NOT NULL when checking to see if a partial index is usable. Forum post d813704d7c. (check-in: 8cc23931 user: drh tags: trunk) | |
| 21:50 | Fix a subtle error in an assert(). dbsqlfuzz eefbd0215c0c1b4bcc32b8141b48b35f7b431300 (check-in: a5ec81eb user: drh tags: trunk) | |
|
2021-05-28
| ||
| 14:28 | If a subquery is used to drive an index, this change avoids making an extra copy of the result of that subquery. But, such situations are probably sufficiently unusual that the added complexity of this enhancement is not worth the performance gain. So I'm going to park this check-in on a branch. If we later find a use case to justify it, we can merge it to trunk then. This is the "further change" that was promised by the prior check-in comment. (Leaf check-in: 4488cb88 user: drh tags: copy-optimization) | |
| 12:48 | Make a deep copy of the result of a subquery in case the subquery is reused. Fix for the problem reported by forum post 28216b36ac and introduced by check-in [f30fb19ff763a7cb]. Further changes to try to optimize the new OP_Copy opcode back into either OP_SCopy or OP_Move will be attempted separately. A test case will be in TH3. (check-in: c9f0b9cb user: drh tags: trunk) | |
| 12:15 | Fix a potential memory leak in json_group_object() following an error. dbsqlfuzz cd32630de3ff039d97089592b63cb3616f8ec9dd (check-in: 21676731 user: drh tags: trunk) | |
|
2021-05-27
| ||
| 16:31 | Add the merge-test.tcl script to ease verification of LTS branches. (check-in: f958ffbc user: drh tags: trunk) | |
| 15:24 | The SQLITE_ENABLE_SHARED_SCHEMA compile-time option is on by default in the configure-generated makefile, permitting a complete test of this branch using "./configure && make test". (check-in: e867d226 user: drh tags: reuse-schema) | |
| 14:09 | Improved comment on the prpagateConstants() routine. No changes to executable code. (check-in: d61a875d user: drh tags: trunk) | |
| 10:23 | Remove a NEVER that is reachable following OOM in writable_schema=ON mode with a corrupt schema. (check-in: 4eb80b0c user: drh tags: trunk) | |
|
2021-05-26
| ||
| 23:10 | Improvements to the constant-propagation optimization in order to better deal with unusual affinities. See forum post 6a06202608 for more detail. (check-in: 9be208a6 user: drh tags: trunk) | |
| 19:52 | Minor optimization to opcode comparison logic in the fixes to the constant-propagation optimization on this branch. (Closed-Leaf check-in: f17dec40 user: drh tags: constant-propagation-fix) | |
| 19:37 | Add new test file whereM.test. Containing tests for the change on this branch. (check-in: c4c76b82 user: dan tags: constant-propagation-fix) | |
| 18:51 | Avoid assuming that an expression like "x=10" in a WHERE clause implies that CASE(x AS TEXT)=='10'. (check-in: 389ec669 user: dan tags: constant-propagation-fix) | |
| 18:46 | Take care that the code is not generated for the same Select object more than once, as transformations that apply during the first pass might cause problems for the second pass. dbsqlfuzz 836b625cd8a41809ef80fc7ebaa6554357bcb463. (check-in: f30fb19f user: drh tags: trunk) | |
| 14:32 | Update an assert() in wherecode.c that might fail following an unrelated SQL error. (check-in: 3e2c36a8 user: dan tags: trunk) | |
|
2021-05-25
| ||
| 16:10 | Do not run sessionbig.test as part of the session_strm permutation. It is too slow. (check-in: 708ce7ad user: dan tags: trunk) | |
| 15:21 | Update an allocation routine in the sessions module to allow it to allocate the maximum size permitted by sqlite3_realloc64(). (check-in: 0b45e821 user: dan tags: trunk) | |
| 11:39 | Enhance the shell tool ".dump PATTERN" command so that it dumps the contents of shadow tables when a virtual table is identified by the PATTERN. (check-in: b0bc5ab9 user: dan tags: trunk) | |
|
2021-05-24
| ||
| 14:35 | Fix a problem in the in-memory journal code that could occasionally lead to a segfault when a sub-transaction that modified zero pages was committed. (check-in: 17960165 user: dan tags: trunk) | |
| 12:28 | Minor comment improvements in fuzzcheck.c. (check-in: 7aca8d52 user: drh tags: trunk) | |
| 12:13 | New dbsqlfuzz cases added to test/fuzzdata8.db. (check-in: 5bb5c9ec user: drh tags: trunk) | |
| 11:35 | Fix the TreeView module so that it works with the TK_ERROR exprssion. Also fix an assert() that was failing due to the recent TK_ERROR addition. (check-in: dd56fbe0 user: drh tags: trunk) | |
| 00:17 | Additional defenses (above and beyond [b986600520696b0c]) to prevent an invalid subquery from causing problems downstream. If an error is found while analyzing a subquery expression, change the expression to TK_ERROR so inhibit further processing on that expression. dbsqlfuzz cf624b8c0484c66e0f552bf6475e3e3f2c22b24e. (check-in: 0be6b6c9 user: drh tags: trunk) | |
|
2021-05-23
| ||
| 17:47 | Do not push a WITH clause onto the processing stack if prior errors have occurred. dbsqlfuzz 6b7a144674e215f06ddfeb9042c873d9ee956ac0. (check-in: c2066dde user: drh tags: trunk) | |
|
2021-05-22
| ||
| 15:05 | Merge latest trunk changes into this branch. (check-in: dbc65a69 user: dan tags: reuse-schema) | |
| 11:23 | Do not invoke sqlite3ExprAffinity() after a syntax error that might have left the tree in an inconsistent state. See also [e8a1515b44380cc5] and forum post 7e484e225c. (check-in: b9866005 user: drh tags: trunk) | |
| 11:00 | The fix in the previous check-in was only correct if the OOM occurs on the initial allocation. This changes should make it correct for a resize as well. (check-in: 57087ab2 user: drh tags: trunk) | |
| 01:30 | If an FTS5 Cursor fails to enlarge the space for the aInst array, set the size of the aInst array to zero. dbsqlfuzz 294254b8105cca409f27a711f1eb2e9e63cbcac5. (check-in: 4ae5e5b5 user: drh tags: trunk) | |
|
2021-05-21
| ||
| 21:49 | If there are errors in a nested CTE, be sure to abandon processing. Do not continue since the parse tree may have been left in a goofy state which could cause use-after-free and segfaults. See forum post aa4a7a3980 for an example. (check-in: 94225d69 user: drh tags: trunk) | |
| 17:31 | Merge latest changes from wal2 into this branch. (check-in: 4efefb59 user: dan tags: begin-concurrent-pnu-wal2) | |
| 16:42 | Merge the latest trunk enhancements into the wal2 branch. (check-in: 95cc7783 user: drh tags: wal2) | |
| 16:41 | Fix a problem with SQLITE_MAX_MEMORY in malloc.c. (check-in: c18dbe2f user: dan tags: trunk) | |
| 16:32 | Merge the latest trunk changes into begin-concurrent-report branch via the begin-concurrent-pnu branch. (check-in: 1dc6cf51 user: drh tags: begin-concurrent-report) | |
| 16:17 | Merge recent trunk changes into the begin-concurrent-pnu branch. (check-in: a8d05f8e user: drh tags: begin-concurrent-pnu) | |
| 15:33 | Merge all recent trunk changes into the begin-concurrent branch. (check-in: 0dba9010 user: drh tags: begin-concurrent) | |
| 13:32 | Add a new sqlite3_config() option for setting the maximum precision of a printf() substitition. The default value is 100,000. It was formerly more than 2 billion. The default can be changed using the SQLITE_PRINTF_PRECISION_LIMIT compile-time option. (Leaf check-in: fd8b68a4 user: drh tags: compile-time-precision-limit) | |
|
2021-05-20
| ||
| 23:25 | Fix a faulty assert() statement in sqlite3ExprListDup(). This is a continuation of the fix at [59812e7ef705226c]. (check-in: 240f7494 user: drh tags: trunk) | |
| 18:11 | Enhance one test case to use various alternative definitions of DUAL. (check-in: 179dcb6b user: drh tags: trunk) | |
| 17:15 | Ensure that objects within view definitions are not incorrectly resolved to CTEs that are part of the statement using the view. (check-in: f7dcc4b5 user: dan tags: trunk) | |
| 11:42 | Ensure the required b-tree mutexes are held for "CREATE TABLE IF NOT EXISTS" and "DROP TABLE IF EXISTS" statements on attached databases. (check-in: 67bde016 user: dan tags: trunk) | |
| 00:44 | Replace [0f0959c6f95046e8] with a new and better solution that also fixes the CTE name resolution problem described in forum post 8590e3f6dc. Test cases for both problems added. (check-in: 5614279d user: drh tags: trunk) | |
| 00:10 | Fixes and improvements to PRAGMA compile_options. (check-in: 34579549 user: drh tags: trunk) | |
|
2021-05-19
| ||
| 23:29 | final merge from trunk (Leaf check-in: 06629569 user: larrybr tags: compile_options) | |
| 21:55 | When constructing the synthensized SELECT statement that is used to choose the rows in an UPDATE FROM, make sure the first table is really the table being updated, and not some common-table expression that happens to have the same name. forum post a274248080. More changes associated with CTE name resolution are pending. (check-in: 0f0959c6 user: drh tags: trunk) | |
| 20:01 | Incorporate ..._DESERIALIZE revision. (check-in: b70587cc user: larrybr tags: compile_options) | |
| 19:55 | merge from trunk (check-in: 19ffe3cf user: larrybr tags: compile_options) | |
| 19:28 | Add a comment with useful information about SHM locking. No changes to deliverable code. (check-in: 4e81ea3b user: drh tags: trunk) | |
| 19:27 | Fix a formatting error on the comment. (Closed-Leaf check-in: 6f2fb60e user: drh tags: better-comment) | |
| 16:55 | Add a table of shared-memory lock offsets in a comment in the os_unix.c code. Include some assert()s to help verify the numbers. (check-in: b480aacb user: drh tags: better-comment) | |
| 14:49 | Allow aggregate sub-selects within ORDER BY and PARTITION BY clauses of window frame definitions. (check-in: 3daab949 user: dan tags: trunk) | |
| 12:17 | Improved column name and column type determination for the RETURNING clause. (check-in: 699c3399 user: drh tags: trunk) | |
| 02:33 | Cure some TCL test failures and narrow an object scope. (check-in: 1155696c user: larrybr tags: trunk) | |
|
2021-05-18
| ||
| 19:10 | In the MULTI-INDEX OR query plan, code for sub-expressions can sometimes be generated twice. But for some subqueries, generating code off of the same tree twice causes problems. So now MULTI-INDEX OR makes a copy of the sub-expressions it uses to avoid code-generating them more than once. dbsqlfuzz 9ebd2140e7206ff724e665f172faea28af801635. (check-in: 4a55f725 user: drh tags: trunk) | |
| 12:36 | Improved comments on the generated opcodes.h file. (check-in: f2a17f1f user: drh tags: trunk) | |
| 00:52 | Add a CORRUPT_DB term to an assert() added earlier today. (check-in: 304739d2 user: drh tags: trunk) | |
|
2021-05-17
| ||
| 17:14 | Fix harmless compiler warnings. (check-in: ace12a39 user: drh tags: trunk) | |
| 16:54 | Enhance the integer-comparison optimization on the OP_Eq and similar opcodes so that it avoids a lot of useless work. (check-in: 4221f41a user: drh tags: trunk) | |
| 16:20 | Fix problems with refering to CTEs from within sub-selects in PARTITION BY or ORDER BY clauses of window frame definitions. Also a problem with renaming a column when the schema contains a trigger containing a correlated sub-select within a window frames PARTITION BY or ORDER BY clause. (check-in: 4c6cd54a user: dan tags: trunk) | |
| 13:11 | When deleting an SQL function that does not exist, return without doing anything at all rather than creating a tombstone function. In this way, function deletes that happen inside virtual-table destructors that are run when a database connection is closing do not create new tombstones in the function table after the function table has already been purged. forum post 726219164b. (check-in: 391c7313 user: drh tags: trunk) | |
| 11:19 | Performance improvement in sqlite3TableLock(). (check-in: 0d77e371 user: drh tags: trunk) | |
|
2021-05-15
| ||
| 19:36 | Performance optimization in the memory allocation of allocateCursor. (check-in: 9d16323d user: drh tags: trunk) | |
| 13:08 | Performance optimization to the clearCell() routine inside of btree. (check-in: c4e02619 user: drh tags: trunk) | |
|
2021-05-14
| ||
| 20:01 | Small performance improvement for sqlite3BtreeInsert(). (check-in: 4ae64484 user: drh tags: trunk) | |
| 15:37 | Avoid adding superfluous virtual WHERE clause terms that might arise due to the constant propagation optimization. (check-in: cf63abbe user: drh tags: trunk) | |
| 14:26 | Attempt the constant propagation optimization on any WHERE clause that has a top-level AND operator, even if the query is not a join. This is an attempt to partially address the concern raised in forum post 830d37b928. (check-in: e994c9f2 user: drh tags: trunk) | |
| 13:32 | Fix an over-length source code comment in whereexpr.c. No logic changes. (check-in: af5eb902 user: drh tags: trunk) | |
|
2021-05-13
| ||
| 18:24 | Modify the sqlite3_stmt_readonly() interface so that it returns false for CREATE TABLE IF NOT EXISTS statements even if the table already exists and the statement is really a read-only no-op. Likewise for DROP TABLE, CREATE INDEX, and DROP INDEX. Update the documentation for sqlite3_stmt_readonly() to reflect this new behavior. (check-in: cf8eb465 user: drh tags: trunk) | |
| 13:43 | The content columns of the index-btree that implements a WITHOUT ROWID table are not ordered and so the query planner should not assume they are ordered. Fix for the issue identified by forum post 6c8960f545. (check-in: c21bc5a2 user: drh tags: trunk) | |
|
2021-05-12
| ||
| 22:15 | Further simplification of the reverse-order scan logic of the previous check-in. (check-in: b2b0e23b user: drh tags: trunk) | |
| 22:02 | Fix the query plan for an indexed lookup on a WITHOUT ROWID table with a DESC primary key when the primary key is constrained by an inequality. See forum post 8988341615. Test cases in TH3. (check-in: f65c929b user: drh tags: trunk) | |
| 15:39 | Fix a race condition that can lead to deadlock in the memdb VFS if one thread is trying to open an existing database at the same moment that another thread that is the only prior user of that database is trying to close it. (check-in: b635375d user: drh tags: trunk) | |
| 14:17 | Add the new threadtest5 test program for stressing multiple database connections in the same process hammering on a single database. Primarily designed to test memdb, but works on any database. (check-in: 8db1c069 user: drh tags: trunk) | |
| 11:55 | Enhance the memdb VFS so that it is able to share databases among multiple database connections in the same process, as long as the database filename begins with "/". This provides a way for threads to share an in-memory database without the use of shared-cache mode. (check-in: 533fffc4 user: drh tags: trunk) | |
| 02:52 | Respond correctly to OOM during mutex allocation. (Closed-Leaf check-in: 98dae595 user: drh tags: memdb-enhancement) | |
| 02:09 | Mark an unreachable branch as NEVER(). (check-in: 6c20d9d4 user: drh tags: memdb-enhancement) | |
|
2021-05-11
| ||
| 11:40 | Fix a NEVER() that can sometimes be true. (check-in: 8f9f1cbc user: drh tags: memdb-enhancement) | |
| 11:33 | Merge changes from branch-3.35. Specifically, fixes for a running RBU with a cksumvfs VFS. (Leaf check-in: ca62c36b user: dan tags: reuse-schema-3.35) | |
| 11:25 | Fixes for cksmvfs and rbu so that they work together. (Leaf check-in: ce8b589e user: dan tags: branch-3.35) | |
| 10:47 | Fix a NEVER() that can sometimes be true. dbsqlfuzz 9a86fb3830977e216fde061bfbe1c1e5c2e9bfc4 (check-in: f39666e6 user: drh tags: trunk) | |
|
2021-05-10
| ||
| 23:48 | Enhance the memdb VFS to provide the ability to share a single database among multiple database connections. (check-in: 0617c66a user: drh tags: memdb-enhancement) | |
| 13:28 | Add new output columns to the "PRAGMA database_list" statement. (Leaf check-in: 905085a7 user: drh tags: database-list-enhancement) | |
|
2021-05-08
| ||
| 17:18 | Enable the sqlite3_serialize() and sqlite3_deserialize() interfaces by default. Omit the SQLITE_ENABLE_DESERIALIZE option and replace it with the SQLITE_OMIT_DESERIALIZE option. (check-in: 6df3b03e user: drh tags: trunk) | |
| 11:57 | Fix help-text typo in the CLI. (check-in: 56075314 user: drh tags: trunk) | |
|
2021-05-07
| ||
| 15:46 | Detect misuse of aggregate functions in the ORDER BY clause of a query even if the query also contains window functions. (check-in: 0d11d777 user: drh tags: trunk) | |
| 14:16 | This was suppose to be a merge from trunk. But something went wrong. (Closed-Leaf check-in: c5e2de1d user: dan tags: bad-merge) | |
|
2021-05-06
| ||
| 20:47 | Fixes for cksmvfs and rbu so that they work together. (check-in: f35f6972 user: dan tags: trunk) | |
| 13:45 | Remove an assert() in FTS5 that can (rarely) be false in the event of an OOM while processing a corrupt database file. (check-in: 5f12f891 user: drh tags: trunk) | |
| 11:02 | Improved detection of oversized cells in balance_nonroot(), especially in index b-trees when a cell is being moved from a child page into the parent page in order to become a new divider cell. (check-in: 9080d3d5 user: drh tags: trunk) | |
|
2021-05-05
| ||
| 19:46 | Provide the sqlite3PrintMem() routine on SQLITE_DEBUG builds, which can be invoked from a debugger to get a summary of the status of an sqlite3_value or Mem object. (check-in: 5ac24179 user: drh tags: trunk) | |
| 11:47 | Fix an undefined-integer-overflow problem in fts3.c. (check-in: a0bf931b user: dan tags: trunk) | |
|
2021-05-04
| ||
| 23:21 | When applying the optimization that disables WHERE clause terms that drive indexes, make sure not to do so if the term being disabled is a transitive constraint. Fix for the problem identified by forum post eb8613976a. (check-in: f1f9b5de user: drh tags: trunk) | |
| 18:24 | Also for the 0x20000 bit of ".wheretrace", show when WHERE clause terms are disabled. (check-in: 625fb253 user: drh tags: trunk) | |
| 16:51 | Additional debugging output for whereScanNext() showing equivalence classes when the ".wheretrace" setting contains the 0x20000 bit. (check-in: 9280e3d9 user: drh tags: trunk) | |
| 12:07 | Back out the EXISTS-to-IN optimization. It slows things down rather than speeds them up depending on the query. And (see forum post 8692d94725) it sometimes results in an incorrect answer. We may come back and revisit this optimization later, but for now it seems best just to disable it. (check-in: 16252d73 user: drh tags: trunk) | |