| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | Move the append array out of the last page into the WT_BTREE structure: it's | Keith Bostic | 2011-12-28 | 1 | -2/+8 |
| * | Create a per-block structure, WT_BLOCK, and move all of the block-specific | Keith Bostic | 2011-12-27 | 1 | -1/+1 |
| * | lint | Keith Bostic | 2011-12-25 | 1 | -2/+0 |
| * | We can't discard pages or overflow values immediately after reconciling | Keith Bostic | 2011-12-24 | 1 | -23/+9 |
| * | unpack->prefix is 0 for overflow keys, no need for both tests. | Keith Bostic | 2011-12-24 | 1 | -2/+1 |
| * | Rework verify to work in the new, block-manager world. | Keith Bostic | 2011-12-23 | 1 | -7/+5 |
| * | Work through the block manager vs. block functions naming: the former are | Keith Bostic | 2011-12-20 | 1 | -2/+2 |
| * | Michael's change to handle eviction of just-created trees: mark the leaf page | Keith Bostic | 2011-12-20 | 1 | -7/+4 |
| * | bug, weren't freeing a temporary buffer on normal exit. | Keith Bostic | 2011-12-19 | 1 | -1/+1 |
| * | Rename __wt_rec_modify_init __wt_page_modify_init, it's the page | Keith Bostic | 2011-12-18 | 1 | -11/+1 |
| * | Split u.col_leaf into two parts (u.col_fix and u.col_var), and move the | Keith Bostic | 2011-12-18 | 1 | -5/+5 |
| * | Use a single mechanism (the tracking list) to schedule blocks to be | Keith Bostic | 2011-12-18 | 1 | -2/+8 |
| * | Replace WT_OFF/WT_OFF_RECORD structures on row- and column-store internal | Keith Bostic | 2011-12-17 | 1 | -160/+240 |
| * | Check WT_PAGE_FORCE_EVICT in __wt_rec_evict. | Michael Cahill | 2011-12-12 | 1 | -3/+0 |
| * | Rename a variable from "tmp" to "tmpkey" to make the intent clearer. | Michael Cahill | 2011-12-09 | 1 | -19/+20 |
| * | Update some comments and change a WT_RET to a WT_ERR so a scratch buffer | Keith Bostic | 2011-12-08 | 1 | -16/+26 |
| * | Move the reconcile state from btree handles to session handles: it is not sha... | Michael Cahill | 2011-12-06 | 1 | -29/+29 |
| * | Cache and use prefixes from the previous key when reconciling row store pages. | Michael Cahill | 2011-12-05 | 1 | -5/+49 |
| * | typo | Keith Bostic | 2011-12-07 | 1 | -1/+1 |
| * | whitespace | Keith Bostic | 2011-12-07 | 1 | -1/+1 |
| * | We can't evict or reconcile split-merge pages, that is, the newly | Keith Bostic | 2011-12-06 | 1 | -42/+57 |
| * | Expand "illegal-format" to include "bad value" (we use it in places | Keith Bostic | 2011-12-05 | 1 | -7/+7 |
| * | When reconciling column-store variable-length pages, avoid the data copy | Keith Bostic | 2011-12-05 | 1 | -6/+15 |
| * | Discard memory allocated to hold overflow items during variable-length | Keith Bostic | 2011-12-04 | 1 | -10/+13 |
| * | Move a couple of routines to more logical places in the file, no real | Keith Bostic | 2011-12-03 | 1 | -120/+120 |
| * | Change the bulk-load code to call the reconciliation code directly. | Keith Bostic | 2011-12-03 | 1 | -99/+203 |
| * | Split WT_PAGE_REC_SPLIT into two cases: WT_PAGE_REC_SPLIT which means | Keith Bostic | 2011-12-02 | 1 | -9/+16 |
| * | simplify an expression, no real change. | Keith Bostic | 2011-12-01 | 1 | -2/+2 |
| * | We can't assume the blocks underlying row-store overflow keys will be | Keith Bostic | 2011-12-01 | 1 | -16/+32 |
| * | The WT_RECONCILE->salvage field is no longer used, remove it. | Keith Bostic | 2011-12-01 | 1 | -7/+3 |
| * | __wt_rec_track_discard can fail, check the error return. | Keith Bostic | 2011-12-01 | 1 | -1/+1 |
| * | Give up trying to use the originating memory address to track overflow items | Keith Bostic | 2011-12-01 | 1 | -31/+25 |
| * | Do the __wt_rec_track_restart_row and __wt_rec_track_restart_col | Keith Bostic | 2011-11-30 | 1 | -3/+21 |
| * | Minor shuffling of one of the tracking routines. | Keith Bostic | 2011-11-30 | 1 | -7/+17 |
| * | Minor shuffling of one of the tracking routines. | Keith Bostic | 2011-11-30 | 1 | -4/+4 |
| * | Minor performance improvement, there's no need to copy column-store values | Keith Bostic | 2011-11-30 | 1 | -4/+14 |
| * | Missed a place in column-store pages where we have to check the | Keith Bostic | 2011-11-30 | 1 | -5/+4 |
| * | fix a comment. | Keith Bostic | 2011-11-30 | 1 | -1/+1 |
| * | Don't track the original page's disk blocks more than once, clear the | Keith Bostic | 2011-11-30 | 1 | -6/+15 |
| * | Split the object tracking routines out from rec_write.c into rec_track.c. | Keith Bostic | 2011-11-30 | 1 | -260/+11 |
| * | There may be multiple requests to track a single block. For example, an | Keith Bostic | 2011-11-30 | 1 | -6/+22 |
| * | We have to track overflow keys that reference split pages in the same | Keith Bostic | 2011-11-30 | 1 | -1/+13 |
| * | Distinguish between tracking, and retaining an overflow object. | Keith Bostic | 2011-11-29 | 1 | -1/+3 |
| * | lint. | Keith Bostic | 2011-11-29 | 1 | -1/+3 |
| * | Removing the bytes_allocated field from the WT_PAGE_MODIFY structure | Keith Bostic | 2011-11-29 | 1 | -3/+14 |
| * | Clean up verbose messages when discarding the WT_PAGE_MODIFY tracked | Keith Bostic | 2011-11-29 | 1 | -0/+39 |
| * | We don't need to track the bytes allocated in the WT_PAGE_MODIFY structure, | Keith Bostic | 2011-11-29 | 1 | -1/+1 |
| * | Don't write the new root page until the root page is evicted, if we're just | Keith Bostic | 2011-11-29 | 1 | -49/+0 |
| * | Always put the page address first in verbose messages. | Keith Bostic | 2011-11-29 | 1 | -3/+3 |
| * | automatically prepend verbose messages with their particular tag line | Keith Bostic | 2011-11-29 | 1 | -22/+18 |
