Open Hub
Open Hub

 Settings |  Report Duplicate

0
I Use This!
Inactive

Commits : Listings

Analyzed about 3 hours ago. based on code collected about 4 hours ago.
Aug 19, 2024 — Aug 19, 2025
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
Re-work ENV/IENV and DB/IDB creation, initialization and destruction -- we can't re-allocate all of them, because we may not have all the handles that reference them (DB ref's IENV, for example, and on ENV->close we don't have the DB handles).
Keith Bostic
More... over 16 years ago
Re-work ENV/IENV and DB/IDB creation, initialization and destruction -- we can't re-allocate all of them, because we may not have all the handles that reference them (DB ref's IENV, for example, and on ENV->close we don't have the DB handles). Add WT_FREE_AND_CLEAR macro to clean memory.
Keith Bostic
More... over 16 years ago
Re-use pages from the cache without freeing and re-allocating the memory. Re-order arguments to __wt_cache_db_in, put side-effect return at the end. Set file handle reference to NULL after closing it.
Keith Bostic
More... over 16 years ago
Re-use pages from the cache without freeing and re-allocating the memory. Quit using a macro for __wt_bt_page_in(), it's not worth it. Add WT_FREE_AND_CLEAR macro to clean memory.
Keith Bostic
More... over 16 years ago
Re-use pages from the cache without freeing and re-allocating the memory. Don't allocate more WT_INDX structures than we need for the in-memory info, most page-types can only have entries / 2 index items. Quit using a macro for __wt_bt_page_in(), it's not worth it. Add WT_FREE_AND_CLEAR macro to clean memory. Re-order arguments to __wt_cache_db_in, put side-effect return at the end. Rename __wt_bt_ovfl_page_in -> __wt_bt_ovfl_in, it's not a page.
Keith Bostic
More... over 16 years ago
Quit using a macro for __wt_bt_page_in(), it's not worth it. Rename __wt_bt_ovfl_page_in -> __wt_bt_ovfl_in, it's not a page.
Keith Bostic
More... over 16 years ago
Quit using a macro for __wt_bt_page_in(), it's not worth it. Add WT_FREE_AND_CLEAR macro to clean memory. Rename __wt_bt_ovfl_page_in -> __wt_bt_ovfl_in, it's not a page.
Keith Bostic
More... over 16 years ago
Re-order arguments to __wt_cache_db_in, put side-effect return at the end.
Keith Bostic
More... over 16 years ago
Re-order arguments to __wt_cache_db_in, put side-effect return at the end. Rename __wt_bt_ovfl_page_in -> __wt_bt_ovfl_in, it's not a page. Add __wt_bt_dump_ipage function to dump in-memory page information.
Keith Bostic
More... over 16 years ago
Quit using a macro for __wt_bt_page_in(), it's not worth it. Add WT_FREE_AND_CLEAR macro to clean memory.
Keith Bostic
More... over 16 years ago
Update.
Keith Bostic
More... over 16 years ago
Add DBT's to the IDB structure so we can free allocated memory when we close the DB handle.
Keith Bostic
More... over 16 years ago
Add debugging support to the memory allocation routines.
Keith Bostic
More... over 16 years ago
Fix memory leaks.
Keith Bostic
More... over 16 years ago
Add internal, leaf, frag and extent sizes to the statistics.
Keith Bostic
More... over 16 years ago
Add debugging support to the allocation routines.
Keith Bostic
More... over 16 years ago
Add routine to configure the IDB structure on creation. Add DBT's to the IDB structure so we can free allocated memory when we close the DB handle.
Keith Bostic
More... over 16 years ago
Close private environments when the DB handle is closed. Move free of IDB structure field into IDB destroy routine.
Keith Bostic
More... over 16 years ago
Add --enable-diagnostic_memory
Keith Bostic
More... over 16 years ago
Add statistics on database leaf, node, fragment & extent sizes.
Keith Bostic
More... over 16 years ago
Store returned data DBT in IDB structure, otherwise we can't free it when we close the DBT.
Keith Bostic
More... over 16 years ago
Checkpoint.
Keith Bostic
More... over 16 years ago
Add $Date$ instead of a version #.
Keith Bostic
More... over 16 years ago
Initial version, 1.0.
Keith Bostic
More... over 16 years ago
We have to allocate page 0 first, otherwise we can end up with an overflow key or data item trying to allocate the first page (page 0), which loses big.
Keith Bostic
More... over 16 years ago
Now the {leaf,intl} node sizes are stored in the DESC structure, we have to set those values in the DESC structure, and read it much more carefully at startup.
Keith Bostic
More... over 16 years ago
Re-work the "isleaf" code -- we have to figure out if the first retrival is a leaf or an internal page, and then use the real level from the pages from then on to keep the value correct. This is a fix to the original work to support node and leaf pages with different sizes.
Keith Bostic
More... over 16 years ago
WT_INDX->addr wasn't being set to invalid for overflow items on leaf pages. Fixing that triggered a but in __wt_bt_dbt_return.
Keith Bostic
More... over 16 years ago
Fix bug in overflow items on leaf pages -- the code checking WT_INDX->addr was wrong.
Keith Bostic
More... over 16 years ago
Describe with WT_INDX->addr is set.
Keith Bostic
More... over 16 years ago