Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Instead of requiring applications use our configuration strings, translate | Keith Bostic | 2015-04-15 | 3 | -12/+90 |
| | | | | | | a short list of possible function or method names, where method names are the public handle (for example, "WT_SESSION"), a dot, and the method name, for example ("WT_SESSION.create"). | ||||
* | Merge branch 'develop' into validate-configuration-string | Keith Bostic | 2015-04-03 | 160 | -1639/+2175 |
|\ | |||||
| * | Add shebang, make db dir if does not exist | Pat Gunn | 2015-04-03 | 1 | -0/+5 |
| | | |||||
| * | Windows doesn't have <x86intrin.h>, it does have <intrin.h> which is | Keith Bostic | 2015-04-02 | 1 | -0/+2 |
| | | | | | | | | included by <msvc.h>. | ||||
| * | Make --with-spinlock=pthread_logging option compile again. | Keith Bostic | 2015-04-02 | 12 | -60/+49 |
| | | | | | | | | | | | | | | Remove the WT_SESSION_IMPL argument to the fprintf, vfprintf, fflush and fclose functions, there are places we want to use it that don't have session handles, and it's not currently needed. Clean up error handling in the vfprintf function. | ||||
| * | Gcc 4.1 can't figure out that copy.v can't be used uninitialized. | Keith Bostic | 2015-04-02 | 1 | -0/+1 |
| | | |||||
| * | Revert "Gcc 4.1 can't figure out that copy.v can't be used uninitialized." | Keith Bostic | 2015-04-02 | 1 | -8/+6 |
| | | | | | | | | This reverts commit fccb479aa9b97fb22d9ec2827c94ba49faa5ab79. | ||||
| * | Gcc 4.1 can't figure out that copy.v can't be used uninitialized. | Keith Bostic | 2015-04-02 | 1 | -6/+8 |
| | | | | | | | | Inline a length check in the short-key/value case to avoid it. | ||||
| * | Unused variables. | Keith Bostic | 2015-04-02 | 1 | -2/+1 |
| | | |||||
| * | Coverity 44262 (#1 of 1): Parse warning (PW.MISSING_INITIALIZER_ON_CONST) | Keith Bostic | 2015-04-02 | 2 | -26/+24 |
| | | | | | | | | | | | | | | | | | | 1. missing_initializer_on_const: const variable "__clear" requires an initializer. This isn't a bug, but we only use WT_CLEAR_INLINE() in a single place in WiredTiger, and we can probably do better by not clearing the unpack structure at all. | ||||
| * | Merge pull request #1841 from wiredtiger/log-cursor-fix | Michael Cahill | 2015-04-02 | 2 | -1/+10 |
| |\ | | | | | | | Reset eol if we continue. Return NOTFOUND if checksum mismatch. #1840 | ||||
| | * | Reset eol if we continue. Return NOTFOUND if checksum mismatch. #1840 | Susan LoVerso | 2015-04-01 | 2 | -1/+10 |
| | | | |||||
| * | | Merge pull request #1843 from wiredtiger/config-speedup | Michael Cahill | 2015-04-02 | 5 | -406/+489 |
| |\ \ | | | | | | | | | Change linear walk of configuration options at API layer to a binary search. | ||||
| | * | | Revert a change that wasn't necessary. | Keith Bostic | 2015-04-01 | 2 | -38/+38 |
| | | | | |||||
| | * | | Merge branch 'develop' into config-speedup | Keith Bostic | 2015-04-01 | 5 | -23/+13 |
| | |\ \ | | |/ / | |/| | | |||||
| * | | | Merge pull request #1842 from wiredtiger/build-fixes | sueloverso | 2015-04-01 | 5 | -23/+13 |
| |\ \ \ | | | | | | | | | | | Build fixes | ||||
| | * | | | Move local var into ifdef. | Susan LoVerso | 2015-04-01 | 1 | -6/+2 |
| | | | | | |||||
| | * | | | Fix warnings. | Susan LoVerso | 2015-04-01 | 2 | -2/+9 |
| | | | | | |||||
| | * | | | Use the existence of the <x86intrin.h> #include file to configure the | Keith Bostic | 2015-04-01 | 3 | -14/+8 |
| | | | | | | | | | | | | | | | | | | | | | | | | | x86 vector instructions. Some old versions of gcc don't have it, but I believe they're old enough that we don't care. | ||||
| | * | | | I can't find any pthreads_XXX_np functions (which is good, because the | Keith Bostic | 2015-04-01 | 3 | -7/+0 |
| |/ / / | | | | | | | | | | | | | "np" stands for "non-portable"); remove optional include of pthread_np.h. | ||||
| | * | | Do a binary search of the base configuration options when checking | Keith Bostic | 2015-04-01 | 4 | -441/+524 |
| | | | | | | | | | | | | | | | | configuration at the API layer instead of a linear walk. | ||||
| | * | | Trivial simplification of API_CALL, wrap the function call in WT_ERR, | Keith Bostic | 2015-04-01 | 1 | -3/+3 |
| |/ / | | | | | | | | | | not the argument test. | ||||
| * | | KNF | Keith Bostic | 2015-04-01 | 1 | -2/+2 |
| | | | |||||
| * | | Mark's fix to handle the style checks functions that have pointer return | Keith Bostic | 2015-04-01 | 4 | -13/+4 |
| | | | | | | | | | | | | types (for example, char *, void *). | ||||
| * | | Merge pull request #1838 from wiredtiger/log-cursor-compression | Don Anderson | 2015-04-01 | 6 | -12/+144 |
| |\ \ | | |/ | |/| | Modify log_scan callback args to send in next LSN. #1837 | ||||
| | * | Increased test size for log cursors. Fixed how the 'none' compressor | Don Anderson | 2015-04-01 | 1 | -125/+125 |
| | | | | | | | | | | | | is configured. Removed DOS carriage returns. | ||||
| | * | Merge branch 'develop' into log-cursor-compression | Don Anderson | 2015-04-01 | 3 | -101/+187 |
| | |\ | |||||
| | * | | Modify log_scan callback args to send in next LSN. #1837 | Susan LoVerso | 2015-03-31 | 6 | -12/+144 |
| | | | | |||||
| * | | | Don't #include <x86intrin.h>, gcc 4.1.2 doesn't have it, use | Keith Bostic | 2015-04-01 | 2 | -3/+5 |
| | | | | | | | | | | | | | | | | <emmintrin.h> instead. | ||||
| * | | | gcc 4.1 compiled failed, check for __SSE2__ explicitly. | Keith Bostic | 2015-04-01 | 1 | -0/+3 |
| | | | | |||||
| * | | | fix a comment. | Keith Bostic | 2015-04-01 | 1 | -1/+1 |
| | | | | |||||
| * | | | Clean up intrinsic #includes: use <intrin.h> on Microsoft | Keith Bostic | 2015-04-01 | 1 | -4/+10 |
| | |/ | |/| | | | | | | | | | | C/C++-compatible compiler, use <x86intrin.h> for GCC/x86 builds. | ||||
| * | | warning: declaration of 'remainder' shadows a global declaration | Keith Bostic | 2015-04-01 | 1 | -8/+8 |
| | | | | | | | | | | | | /usr/include/math.h:288: warning: shadowed declaration is here | ||||
| * | | Merge pull request #1823 from wiredtiger/vectorize-search | Keith Bostic | 2015-04-01 | 3 | -101/+187 |
| |\ \ | | |/ | |/| | Vectorize search | ||||
| | * | Merge branch 'develop' into vectorize-search | Keith Bostic | 2015-04-01 | 87 | -824/+704 |
| | |\ | | |/ | |/| | |||||
| * | | Coverity 1199719 (#1 of 1): Dereference after null check (FORWARD_NULL) | Keith Bostic | 2015-03-31 | 2 | -5/+4 |
| | | | | | | | | | | | | var_deref_op: Dereferencing null pointer "ref". | ||||
| * | | Coverty 1129071 (#1 of 1): Dereference before null check (REVERSE_INULL) | Keith Bostic | 2015-03-31 | 1 | -1/+1 |
| | | | | | | | | | | | | | | | check_after_deref: Null-checking "conn" suggests that it may be null, but it has already been dereferenced on all paths leading to the check. | ||||
| * | | Coverity 1129018, 1129019, 1129020, 1129021: Side effect in assertion | Keith Bostic | 2015-03-31 | 1 | -42/+49 |
| | | | | | | | | | | | | (ASSERT_SIDE_EFFECT). | ||||
| * | | update | Keith Bostic | 2015-03-31 | 1 | -0/+1 |
| | | | |||||
| * | | Merge pull request #1835 from wiredtiger/var-col-store-bug | Keith Bostic | 2015-03-31 | 2 | -14/+53 |
| |\ \ | | | | | | | | | Add a test case for a new bug in variable length column stores. | ||||
| | * | | When looking for the next-larger item (__col_insert_search_gt), stay | Keith Bostic | 2015-03-31 | 1 | -4/+4 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | at the same level if the checked record is equal to the searched for record (can't happen in the current use of this code, but it's the correct thing to do in a skiplist). When looking for the next-smaller item (__col_insert_search_lt), the search still has to be greater-than, reference #1835. | ||||
| | * | | Add a __setitem__ to cursors in the Python API. | Keith Bostic | 2015-03-31 | 1 | -12/+3 |
| | | | | |||||
| | * | | Merge branch 'develop' into var-col-store-bug | Alex Gorrod | 2015-03-31 | 60 | -366/+150 |
| | |\ \ | | |/ / | |/| | | |||||
| * | | | Merge pull request #1836 from wiredtiger/python-cursor-setitem | Alex Gorrod | 2015-03-31 | 59 | -366/+147 |
| |\ \ \ | | | | | | | | | | | Add a __setitem__ to cursors in the Python API | ||||
| | * | | | Add a __setitem__ to cursors in the Python API, remove lots of boilerplate ↵ | Michael Cahill | 2015-03-31 | 59 | -366/+147 |
| |/ / / | | | | | | | | | | | | | c.set_key ... c.set_value ... c.insert code. | ||||
| * | | | Add more spelling words required on Fedora Linux. | Michael Cahill | 2015-03-31 | 1 | -0/+3 |
| | | | | |||||
| | * | | Add the rest of the col gap test case. | Alex Gorrod | 2015-03-31 | 1 | -11/+20 |
| | | | | |||||
| | * | | Add a test case for a new bug in variable length column stores. | Alex Gorrod | 2015-03-31 | 1 | -0/+39 |
| |/ / | | | | | | | | | | Refs #1834 | ||||
| * | | Merge pull request #1833 from wiredtiger/column-store-perf | Alex Gorrod | 2015-03-31 | 7 | -12/+276 |
| |\ \ | | | | | | | | | Column store name-space gap handling. | ||||
| | * \ | Merge branch 'develop' into column-store-perf | Michael Cahill | 2015-03-30 | 5 | -343/+59 |
| | |\ \ | | | | | | | | | | | | | | | | | | | | | Conflicts: dist/s_string.ok |