summaryrefslogtreecommitdiff
path: root/src/include/bitstring.i
AgeCommit message (Collapse)Author
2014-02-24Switch the WT_PAGE union #defines from pu_xxx to pg_xxx.Keith Bostic
2014-02-05Add pu_xxx_field #defines for the rest of the page-union fields.Keith Bostic
2014-01-07Update copyright notices for 2014.Keith Bostic
Move lang/java and lang/python into the public domain.
2013-11-22Pedantic lint: don't return a void expression.Michael Cahill
2013-01-01Update copyright notice to 2013.Keith Bostic
2012-08-27nitAlex Gorrod
2012-08-27Update btree usage of 64 bitstring implementation, so it's cleaner.Alex Gorrod
2012-08-22Update the bitstring implementation to use 64 bit length strings.Alex Gorrod
2012-08-15We're sending a mixed message -- remove the copyright notices from filesKeith Bostic
where we've placed our content in the public domain.
2012-08-07Whitespace cleanups, no semantic changes.Keith Bostic
2012-04-27Fix uninitialized variables as reported by GCC 4.6.Michael Cahill
2012-04-19Snapshot verification code. The verification statements:Keith Bostic
Every btree page found in a snapshot should be located in the file before the recorded maximum file size, so, each snapshot's pages must be less than the snapshot's recorded file size. The sum of the allocation lists in every snapshot minus the sum of the discard lists in every snapshot should map one-to-one to the tree pages in that snapshot. The sum of all the pages allocated by every snapshot plus the pages on the avail list (plus blocks used to hold extent lists), should map one-to-one to the entire file. Salvage has to create valid alloc/discard/avail lists, so it starts off with the entire file on the alloc list, and then "frees" any pages it doesn't want. Don't read the avail list when we're deleting snapshots, it doesn't ever change and there's no reason to read or re-write it. Fix a bug where the live system's alloc/discard lists weren't merged during a snapshot, if no snapshot was being deleted. We no longer reset the snapshot's file size when writing a snapshot, we want to check the snapshot's blocks against a reasonable maximum file size during verification. This won't work if there's ever an API to roll-forward from an intermediate snapshot. Pass the snapshot-array into the verify-start routine: we have to read the avail list from the last snapshot, those are blocks that must not be allocated in any snapshot. Add a check: the earliest snapshot in the system should never have blocks on its discard list. Replace the __wt_block_free_ext() function with the more general-purpose __wt_block_insert_ext() function, which inserts a range onto an extent list.
2012-01-31Fix uninitialized warnings with GCC 4.6.Michael Cahill
2012-01-22Update copyright notices to 2012.Keith Bostic
Change the copyright check script to always use the current year. Change the example code copyright notices to document them as freeware, anybody using WiredTiger example code is free to use it in any way they choose, including commercially, WiredTiger claims no rights in such a copy. Minor re-ordering of the standard copyright notice so the preamble is always the copyright notice itself, not the mention of the LICENSE file.
2011-12-18Split u.col_leaf into two parts (u.col_fix and u.col_var), and move theKeith Bostic
column-store append/update lists from WT_PAGE into WT_PAGE_MODIFY in order to shrink WT_PAGE. --HG-- extra : rebase_source : 6d7df0f5b50351cabfec464eccfa0b49b8b0f35a
2011-08-19Fix a comment.Keith Bostic
2011-08-05Automated change whacked some copyright statements.Keith Bostic
2011-08-04Remove declarations for static inline functions, when mixed with non-inlined ↵Michael Cahill
code, move the inlined functions to the top. This is required for some compilers to inline at all, and avoids the need for a "STATIN" to shorten "static inline" declarations.
2011-08-03Fast-path 8-bit fixed-length column-store fields: I think they'll be commonKeith Bostic
enough we should notice them. I looked at slice-and-dice for the rest of the field sizes, but it's going to require tests (if not a loop), and I bet the compiler is going to do a better job than I will.
2011-08-03Redo Paul Vixie's bitstring.h #defines as inline functions, usingKeith Bostic
uint32_t max-bits instead of "int", and rename include/bitstring.h to include/bitstring.i. Change the names of the fixed-length column-store get/set value functions to match the other bitstring functions, and move them from include/btree.i to include/bitstring.i. --HG-- rename : src/include/bitstring.h => src/include/bitstring.i