Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
50 descendants and 50 ancestors of 521d72bdf67b4b19
|
2011-12-22
| ||
| 01:46 | Remove a redundant assignment operator. (check-in: 19a6852a1e user: drh tags: statvfs) | |
|
2011-12-21
| ||
| 18:04 | Fix other problems where 32-bit integer overflow may cause a problem. Two that require an improbably large sub-journal and two in test harness code. Ticket [ac0ff496b7e2] (check-in: c723e3e18a user: dan tags: trunk) | |
| 17:00 | Avoid 32-bit overflow when calculating the byte offset of an overflow page in SQLITE_DIRECT_OVERFLOW_READ code. Fix for [ac0ff496b7]. (check-in: c5256b59ad user: dan tags: trunk) | |
| 14:42 | Add the ability to enable or disable SQLITE_IOCAP_ZERO_DAMAGE using a URI parameter for both unix and windows. Add a file-control to query or disable the ZERO_DAMAGE setting. Add the -uri option to the "sqlite3" TCL command in tclsqlite3.c. Allow the sqlite3_uri_parameter() interface to accept a NULL pointer for its first parameter. (check-in: c83627b732 user: drh tags: statvfs) | |
|
2011-12-20
| ||
| 22:18 | Remove the code that tries to detect OOO header writes on a WAL recovery. The code is made obsolete by syncing the WAL header. (check-in: 7ac713a14e user: drh tags: statvfs) | |
| 20:13 | Refactor the sqlite3WalFrames() routine for clarity of presentation. Do the padded transaction sync as the write pointer crosses the final sector boundary instead of at the end, for efficiency. Always sync the WAL header immediately after it is written. (check-in: 92c73b421b user: drh tags: statvfs) | |
| 13:13 | Assert that the isCommit parameter to sqlite3WalFrames() is zero if and only if the nTruncate parameter is zero. (check-in: 979daf92e0 user: drh tags: trunk) | |
|
2011-12-19
| ||
| 15:46 | Add test for WAL mode to unixexcl.test. (check-in: 4d518bd480 user: dan tags: trunk) | |
| 11:57 | Merge [21b76af6ed] into statvfs branch. (check-in: e694f7b166 user: dan tags: statvfs) | |
| 11:16 | Fix a couple of test cases to account for the master-journal name related change in [cf3bccc2]. (check-in: 21b76af6ed user: dan tags: trunk) | |
| 10:07 | Modify test cases to account for the ZERO_DAMAGE change. (check-in: 68684495f1 user: dan tags: statvfs) | |
| 00:31 | Some fixes to the test suite so that it works with ZERO_DAMAGE set to true. Still lots more problems remain. (check-in: 41891b231e user: drh tags: statvfs) | |
|
2011-12-17
| ||
| 20:02 | For improved clarity of presentation, refactor some of the code associated with ZERO_DAMAGE and sector-size. (check-in: 1dde96c9ee user: drh tags: statvfs) | |
| 19:49 | Add SQLITE_IOCAP_ZERO_DAMAGE and enable it for both unix and windows. Use this device characteristic to reduce the required work in journaling. A side effect is that this changes the default page exists back to 1024 even with the use of statvfs(). (check-in: a0be6ea464 user: drh tags: statvfs) | |
| 16:25 | Fix a bad #endif with the previous check-in on this branch. (check-in: 915713ffe4 user: drh tags: statvfs) | |
| 16:09 | Add support for statvfs() in os_unix.c, for determining the sector size. This causes many TCL test failures under Linux. (check-in: e0d44450b9 user: drh tags: statvfs) | |
| 13:45 | Merge in changes that cause the first sector of the WAL file to be synced when the WAL restarts. This is a fix for the power-loss corruption problem described in ticket [ff5be73dee086] (check-in: 44ca4d1233 user: drh tags: trunk) | |
| 08:10 | Add tests to walcrash3.test. (check-in: d768804280 user: dan tags: trunk) | |
| 01:31 | Make sure the Salt-1 value in the WAL file is sequentially numbered at each checkpoint, as it says it should be in the documentation. (check-in: 7b63b11b93 user: drh tags: trunk) | |
|
2011-12-16
| ||
| 21:26 | Enhance the WAL header sync so that it honors the various synchronous pragmas, settings, and device characteristics. (Closed-Leaf check-in: 9799241f7d user: drh tags: wal-header-sync) | |
| 19:34 | Proposed changes that ensure that the WAL header is written prior to the first commit mark. (check-in: 91d0437c07 user: drh tags: wal-header-sync) | |
| 17:01 | Add code for a test that was failing before the persistent-wal related changes of [09ccc4a1be]. (check-in: 49d21ce50f user: dan tags: trunk) | |
| 15:38 | Merge the fix for [a1fa75cbdd02] from the experimental branch. Also fix the persistent-wal mode feature of truncating the WAL on close so that it always truncates the WAL to zero bytes. (check-in: 09ccc4a1be user: drh tags: trunk) | |
| 15:11 | Improved logging of master-journal name conflicts. (check-in: b1005ef46c user: drh tags: trunk) | |
| 13:42 | Merge the nx-devkit changes into trunk. (check-in: cf3bccc2e9 user: drh tags: trunk) | |
| 13:24 | Experimental fix for [a1fa75cbdd]. (Closed-Leaf check-in: 6492af76ea user: dan tags: experimental) | |
|
2011-12-15
| ||
| 17:44 | Add stdio support to the quota VFS. (check-in: 322bd15f97 user: drh tags: trunk) | |
|
2011-12-14
| ||
| 18:33 | Merge the nx-devkit changes into trunk. This includes the new SQLITE_FCNTL_VFSNAME file-control. (check-in: da118e02c0 user: drh tags: trunk) | |
| 00:37 | Treat a zero return value from the Win32 APIs MultiByteToWideChar and WideCharToMultiByte as an error condition. (check-in: c65e5a36f1 user: mistachkin tags: trunk) | |
|
2011-12-13
| ||
| 15:37 | Merge the nx-devkit changes into trunk. (check-in: 03a70c3dae user: drh tags: trunk) | |
| 12:10 | This branch merged with nx-devkit and from there into trunk. Was: In the multiplexor, instead of generating a unique file-name when SQLite opens a temp file, allow the underlying VFS to generate a different temp file for each chunk. Given the changes to the xFileSize method, it is no longer necessary for the different chunks of a temp file to use the same base name. (Closed-Leaf check-in: 199f52bced user: dan tags: experimental) | |
| 11:15 | Change the way IO errors are handled in the xFileSize method of the multiplexor VFS. Add test file multiplex3.test. (check-in: 255d21499b user: dan tags: experimental) | |
|
2011-12-12
| ||
| 19:48 | Fix os_unix.c so that, unless 8.3 filenames are actually in use, journal and wal file permissions are assigned correctly even if SQLITE_ENABLE_8_3_NAMES is defined. (check-in: 169e12295c user: dan tags: trunk) | |
| 18:55 | Change the multiplexor VFS so that xTruncate is a no-op on database files. (check-in: ee1e012256 user: dan tags: trunk) | |
| 18:21 | Fix for the xFileSize method of test_multiplex.c when used on a file opened with the DELETEONCLOSE flag set. (check-in: 713071c493 user: dan tags: trunk) | |
| 16:51 | Add magic comments to shell.c to make it easier for scripts to insert additional version information displays for custom builds. (check-in: 9b2691aa91 user: drh tags: trunk) | |
|
2011-12-11
| ||
| 22:44 | In the multiplexor shim, when using 8+3 filenames, begin numbering journal overflow files with 101 instead of 001 to avoid name collisions with the main database file. (check-in: 05bf8a0b17 user: drh tags: trunk) | |
| 21:51 | Update comments describing the flattening optimization in select.c. (check-in: dab4c137a8 user: drh tags: trunk) | |
| 02:30 | Fix STAT3 so that it works with the new uninitialized register logic of the VDBE. Ticket [7bbfb7d4422ff] (check-in: d11a57985c user: drh tags: trunk) | |
| 02:29 | Fix harmless compiler warnings. (check-in: 1e6a698aab user: drh tags: trunk) | |
|
2011-12-10
| ||
| 23:18 | Minor cleanups of the compound-subquery flattening logic. New test cases added for joins the compound subquery. (check-in: 5061d85ff9 user: drh tags: trunk) | |
| 17:17 | Allow UNION ALL compounds to be promoted up to replace a simple wrapper SELECT even if the compounds are joins. (check-in: 3d4b4f4fb7 user: drh tags: trunk) | |
| 15:55 | Import the experimental parse-tree explainer, with fixes, from the tree-explain branch. Disabled by default. Use SQLITE_ENABLE_TREE_EXPLAIN to turn it on. (check-in: bcbc7152d4 user: drh tags: trunk) | |
| 14:44 | Merge the latest trunk changes into tree-explain branch. (Closed-Leaf check-in: 1a360da0f8 user: drh tags: tree-explain) | |
| 13:49 | Always use _msize() to get memory allocation sizes on windows, without having to do anything special in the makefile. (check-in: 256e27bd11 user: drh tags: trunk) | |
|
2011-12-09
| ||
| 18:06 | Change the VDBE so that all registers are initialized to "Invalid" instead of NULL and report errors on any attempted read of an Invalid register. This will help prevent future bugs similar to [7bbfb7d442]. (check-in: 0064bab771 user: drh tags: trunk) | |
| 17:51 | Remove an unnecessary condition. (Closed-Leaf check-in: d9ba023c60 user: drh tags: uninit-vdbe-mem) | |
| 17:38 | Remove an unnecessary initialization of Vdbe.aOnceFlag. (check-in: 421714dad3 user: drh tags: uninit-vdbe-mem) | |
| 17:27 | Take out the OP_JumpOnce opcode. Revert compound SELECT to use OP_IfNot, which is the correct behavior. Mark trigger registers as initially invalid. (check-in: 6a9fb47d50 user: drh tags: uninit-vdbe-mem) | |
| 16:59 | Previous check-in broke auto-increment. This check-in appears to fix it. (check-in: 28ffd39c71 user: drh tags: uninit-vdbe-mem) | |
| 16:21 | Make no assumptions about the initial state of VDBE registers. (check-in: 521d72bdf6 user: drh tags: uninit-vdbe-mem) | |
| 13:24 | Modify the OP_Once opcode so that it works correctly in trigger sub-programs. This is a candidate fix for [7bbfb7d442]. (check-in: 557c69055a user: dan tags: trunk) | |
| 05:52 | The Windows OS flavor #ifdefs must be performed after the 'windows.h' file has been included. (check-in: 3702a31e56 user: mistachkin tags: trunk) | |
|
2011-12-08
| ||
| 20:41 | Hand merge the zone allocator for MacOS from the apple-osx branch. (check-in: 0d955c20c0 user: drh tags: trunk) | |
| 19:50 | In persistent WAL mode, truncate the WAL file to the size specified by the journal_size_limit pragma when disconnecting from the WAL. (check-in: 9687b305c2 user: drh tags: trunk) | |
| 03:51 | Follow the previously established pattern for detecting preprocessor defines for specific flavors of Windows (for NT in this case). (check-in: a0d92193dd user: mistachkin tags: trunk) | |
|
2011-12-07
| ||
| 01:55 | Cherrypick the [7e5b56b1c6] fix for the sqlite3SelectDup() routine into trunk. (check-in: 7fc535090c user: drh tags: trunk) | |
|
2011-12-06
| ||
| 13:46 | Add file tkt-3a77c9714e.test, containing tests to verify that the problem documented by ticket [3a77c9714e] has been fixed. (check-in: 162421dadf user: dan tags: trunk) | |
| 13:24 | Do not reuse temp registers originally allocated to a subroutine within the main body of the program logic, since if the subroutine is called while the reused temp registers are in use, their values will get clobbered. Candidate fix for ticket [3a77c9714e63330] (check-in: 092d53315e user: drh tags: trunk) | |
| 00:47 | Merge the winGetLastError fixes into trunk. (check-in: 5b03ba9db0 user: drh tags: trunk) | |
|
2011-11-29
| ||
| 15:40 | Remove unused fields from the Parse object. Documentation and formatting improvements on data structure definitions. (check-in: 431556cac0 user: drh tags: trunk) | |
| 15:04 | Remove unused boolean fields from the UnpackedRecord object. (check-in: b10d091ec0 user: drh tags: trunk) | |
| 14:46 | Removed the unused "type" field from the CollSeq (collating sequence) object. (check-in: ab89b047e8 user: drh tags: trunk) | |
|
2011-11-25
| ||
| 21:51 | Update the version number to 3.7.10. (check-in: ed0151ba83 user: drh tags: trunk) | |
| 17:51 | Fix a typo in the documentation for sqlite3_stmt_busy(). (check-in: 7812626d1d user: drh tags: trunk) | |
| 17:21 | Add the sqlite3_stmt_busy() interface. (check-in: 95cc4af686 user: drh tags: trunk) | |
|
2011-11-21
| ||
| 00:54 | Cleanup the semantics surrounding use of the GetLastError function on Windows. (Closed-Leaf check-in: 7e657bbb80 user: mistachkin tags: winGetLastError) | |
|
2011-11-18
| ||
| 13:10 | Change the multiplexor to use a 3-digit suffix. (check-in: 0b7edc4475 user: drh tags: trunk) | |
|
2011-11-17
| ||
| 11:49 | Improvements to the documentation of the sqlite3_db_filename() interface. (check-in: 1c45b2a0c0 user: drh tags: trunk) | |
| 11:43 | Restore the test for sqlite3OsFullPathname() failure that was mistakenly removed when [ceee03c79a] was backed out by [69ec53fc1c]. (check-in: 4d3cf9e1d8 user: drh tags: trunk) | |
| 00:56 | Add the sqlite3_db_filename() interface. (check-in: 93a947989b user: drh tags: trunk) | |
|
2011-11-16
| ||
| 23:29 | Back out the [ceee03c79a] change. (check-in: 69ec53fc1c user: drh tags: trunk) | |
| 19:29 | Add the sqlite3_db_release_memory() interface and the shrink_memory pragma. (check-in: 3f58e7c889 user: drh tags: trunk) | |
| 18:08 | Merge the PCACHE2 changes into trunk. (check-in: 457513f21f user: drh tags: trunk) | |
| 17:32 | Change the multiplexor suffix from 2 to 3 digits. (Closed-Leaf check-in: 06e0cdaf91 user: drh tags: 3-digit-multiplex-suffix) | |
| 16:23 | Remove code made obsolete by the changes to index processing that allow range search on the rowid. (check-in: a5418c7fc2 user: drh tags: trunk) | |
| 15:41 | Fix an invalid assert() statement added by [3b58f5f066]. (check-in: 888b09dd8f user: dan tags: trunk) | |
| 15:27 | Where possible, take advantage of the rowid at the end of index records to optimize range constraints (<, >, <=, >=) on the rowid column. (check-in: 3b58f5f066 user: dan tags: trunk) | |
| 08:18 | Update memsubsys1.test to account for the recently increased size of the MemPage structure in btreeInt.h. (check-in: 4fb3ca756a user: dan tags: trunk) | |
|
2011-11-14
| ||
| 03:00 | Forward port the 8-byte alignment fix from branch-3.7.9. (check-in: ebf6eb6ed7 user: drh tags: trunk) | |
| 02:53 | Fix a 8-byte alignment problem that causes a SIGBUS on Sparc. (check-in: 54cc119811 user: drh tags: branch-3.7.9) | |
|
2011-11-13
| ||
| 21:44 | Add a version number to the sqlite3_pcache_methods2 object. Other PCACHE2 documentation improvements. (Closed-Leaf check-in: 9f839ac05a user: drh tags: experimental-pcache) | |
|
2011-11-12
| ||
| 23:10 | Attempt to modify btree.c so that it assumes that calls to sqlite3PagerWrite() will reallocate the page buffer. As there is not good way to test this assumption yet, probably a few spots were missed. (check-in: ceee03c79a user: drh tags: experimental-pcache) | |
| 16:46 | Remove a couple of incorrect assert statements so that the test suite will run with -DSQLITE_DEFAULT_CACHE_SIZE=0. (check-in: 87614b62ac user: drh tags: trunk) | |
| 15:41 | Merge the windows xSyscall enhancements into trunk. (check-in: c1fab9aca1 user: drh tags: trunk) | |
| 03:17 | Make sure to flag benign malloc failures in the Windows VFS as such. Expand use of the DO_OS_MALLOC_TEST to cover the VFS functions that can now return an out of memory error. Support an experimental --match option to the test suite that will run only those tests matching the specified pattern. (Closed-Leaf check-in: 76dec8aa9d user: mistachkin tags: winSyscall) | |
|
2011-11-11
| ||
| 23:51 | Catch and report errors from sqlite3OsFullPathname(). (check-in: 77119785c8 user: drh tags: trunk) | |
| 23:31 | Remove use of malloc and free from the Windows VFS. Also, prevent sqlite3BtreeOpen from assuming that sqlite3OsFullPathname cannot fail. (check-in: 8966ec1797 user: mistachkin tags: winSyscall) | |
| 22:08 | Work in progress to implement the 'syscall' functionality for Windows. (check-in: ed88fb0024 user: mistachkin tags: winSyscall) | |
| 14:12 | Pull over all the latest changes from trunk. (check-in: 1bbbf8574a user: drh tags: experimental-pcache) | |
| 00:27 | Make sure a corrupt index does not cause a buffer overread in sqlite3VdbeRecordCompare(). (check-in: 471cf0d8e7 user: drh tags: trunk) | |
|
2011-11-10
| ||
| 21:45 | Expand passing of a last error argument to the getLastErrorMsg function. Also, remove unused SQLITE_W32_THREADS define. (check-in: 8f28797984 user: mistachkin tags: trunk) | |
| 20:21 | In winAccess, save the Win32 last error value prior to invoking user logging callback. Also, explicitly pass the Win32 last error value to winLogError in order to keep it accurate. Fixes a problem reported on the mailing list. (check-in: 32ab365715 user: mistachkin tags: trunk) | |
| 02:39 | Follow-on to the previous check-in to prevent a division by zero if the lookahead slot size is something goofy like 6 on a 32-bit machine. (check-in: 6bda711f93 user: drh tags: experimental-pcache) | |
| 02:24 | Use sqlite3MallocSize() to get the actual size of the memory allocation used for lookaside cache and increase the size of the cache to use the full allocation. (check-in: 0e53ecad94 user: drh tags: experimental-pcache) | |
|
2011-11-09
| ||
| 18:07 | Omit an unnecessary Sleep() call in windows pending-lock retry logic. Enhance the comment on that logic to discourage people from copying it into other VFSes. (check-in: 0c951a9704 user: drh tags: trunk) | |
| 17:01 | When compiling with MSVC, use the _msize function. (check-in: 797a147934 user: mistachkin tags: experimental-pcache) | |
| 16:12 | Minor changes needed to restore full branch test coverage. (check-in: bc10a75357 user: drh tags: experimental-pcache) | |
| 14:23 | A negative value N for the cache_size pragma adjusts the number of cache pages to use approximately N kibibytes of memory. (check-in: b3faa680ae user: drh tags: experimental-pcache) | |
| 01:53 | For the mem1.c system malloc implementation, use the malloc_usable_size() function if the HAVE_MALLOC_USABLE_SIZE macro is defined. Update autoconf to look for that function when configuring. (check-in: 2e8ab3cedf user: drh tags: experimental-pcache) | |
| 00:06 | Update the API documentation for the new pcache2 interface. Change the order of parameters on the xCreate method of pcache2. (check-in: 4da7095683 user: drh tags: experimental-pcache) | |