| Age | Commit message (Expand) | Author |
| 2012-01-23 | Some cleanup of the hazard code. In particular: | Michael Cahill |
| 2012-01-22 | Update copyright notices to 2012. | Keith Bostic |
| 2011-12-05 | Parallelize eviction by having application threads evict pages while reads ar... | Michael Cahill |
| 2011-12-07 | Remove the read thread now that application threads handle reads directly. | Michael Cahill |
| 2011-12-05 | Make eviction parameters tunable, to force more or less aggressive eviction. | Michael Cahill |
| 2011-11-26 | Split the reconciliation into reconciliation and eviction pieces, and | Keith Bostic |
| 2011-11-28 | Make the size of the read and eviction request arrays dependent on the maximu... | Michael Cahill |
| 2011-11-11 | Add a force flag to eviction to block waiting for busy pages. | Michael Cahill |
| 2011-11-10 | Avoid session buffers, just call malloc/free directly. | Michael Cahill |
| 2011-11-04 | Rename WT_MTX to WT_CONDVAR | Michael Cahill |
| 2011-11-04 | Remove the workQ code entirely. | Michael Cahill |
| 2011-11-03 | Remove the workQ -- have application threads notify the read and eviction ser... | Michael Cahill |
| 2011-10-01 | Another run at memory-barriers | Keith Bostic |
| 2011-09-30 | Incorporate Keith's feedback on read/write barriers. | Michael Cahill |
| 2011-09-19 | Fix some test failures associated with signalling the I/O threads. During | Michael Cahill |
| 2011-09-11 | Replace column-store page-extension code with a new skiplist, maintained | Keith Bostic |
| 2011-08-09 | Turned on GCC "-Wpadded". | Michael Cahill |
| 2011-08-18 | Use size_t for size calculations, only store uint32_t in structs. | Michael Cahill |
| 2011-08-05 | Make reconciliation structures per-file instead of per-database. | Keith Bostic |
| 2011-08-04 | Shuffle some declarations around in include files (e.g., so WT_BTREE is in bt... | Michael Cahill |
| 2011-07-28 | Connect the old Btree statistics code to the statistics cursors: if you | Keith Bostic |
| 2011-07-08 | Add memory-footprint field to the in-memory page structure, it's used | Keith Bostic |
| 2011-07-07 | Re-organize the cache structure by functionality. | Keith Bostic |
| 2011-05-30 | Change all struct names to have the prefix "__wt_", and all corresponding typ... | Michael Cahill |
| 2011-05-16 | Add a BTREE flag, WT_BTREE_NO_EVICTION, to lock out eviction: none of | Keith Bostic |
| 2011-05-11 | Change the "caller" argument to page-reconciliation to be flags: the page | Keith Bostic |
| 2011-05-04 | Now that evict doesn't doing anything with the page's referencing WT_REF, | Keith Bostic |
| 2011-05-04 | Move the copying of the hazard references to the reconciliation code: eviction | Keith Bostic |
| 2011-05-04 | Fix the problem where a sync or close call could deadlock with a thread of | Keith Bostic |
| 2011-05-03 | A problem with previous versions of the code was we did not get exclusive | Keith Bostic |
| 2011-04-25 | Move the reconciliation structure into the bt_reconcile.c file, it's only used | Keith Bostic |
| 2011-04-18 | Small random changes: | Keith Bostic |
| 2011-04-18 | Re-write buffer management to fix a problem where we could overflow the | Keith Bostic |
| 2011-04-17 | The inactive array has to be reallocated, not allocated anew each time -- | Keith Bostic |
| 2011-04-16 | Fix a bug with the "remaining space" calculation at the end of the page: If | Keith Bostic |
| 2011-04-16 | Handle split-created internal pages having an incorrect first key. The | Keith Bostic |
| 2011-04-14 | Don't discard inactive pages immediately, because that corrupts the tree if | Keith Bostic |
| 2011-04-13 | Update a comment and remove an unused structure field. | Keith Bostic |
| 2011-04-13 | The reconciliation code tracks current information about the starting | Keith Bostic |
| 2011-04-13 | Bug fix for sync/close methods. The problem is pages that were logically | Keith Bostic |
| 2011-04-11 | Make cache bytes-in/bytes-out code inline functions so I can put diagnostic | Keith Bostic |
| 2011-04-09 | Branch merge. | Keith Bostic |
| 2011-04-09 | The recent core dump problem was caused by the sync call reconciling pages the | Keith Bostic |
| 2011-04-08 | Treating the statistics fields as an array makes it too easy to walk off the | Keith Bostic |
| 2011-04-03 | In diagnostic mode, track the location from which we acquire hazard references. | Keith Bostic |
| 2011-03-28 | Build the internal typedef list automatically. | Keith Bostic |
| 2011-03-21 | A simpler approach to handling eviction during file verification -- don't | Keith Bostic |
| 2011-03-20 | whitespace cleanup | Keith Bostic |
| 2011-03-11 | Rename a bunch of items to be cells. | Keith Bostic |
| 2011-03-09 | Use scratch buffers for the RLE expansion array so we don't allocate new memory | Keith Bostic |