Commit Graph

826 Commits

Author SHA1 Message Date
Howard Chu
7e58cc0fd3 updates 2013-09-05 16:23:49 -07:00
Salvador Ortiz
cfc13e2b48 ITS#7682 In mdb_env_copy retry open if O_DIRECT fails 2013-09-05 16:20:18 -07:00
Howard Chu
90c55178db More for 5f6fa8c042
Reset delete flag in cursor_put().
2013-09-05 06:02:07 -07:00
Claude Brisson
18a07eb7c2 ITS#7681 fix mdb_cursor_set behaviour for MDB_SET_RANGE 2013-09-05 05:57:03 -07:00
Howard Chu
5f6fa8c042 Remember if last cursor op was a delete
If so, don't advance cursor on next NEXT. Also, init xcursor
on the next GET_CURRENT.
2013-09-03 16:10:20 -07:00
Howard Chu
3d59ca3558 Add a changelog 2013-08-28 20:22:41 -07:00
Howard Chu
5c5300c604 Bump version to 0.9.8 2013-08-28 19:17:25 -07:00
Howard Chu
2e7130cab0 Allow mdb_env_set_mapsize() on an open environment
The caller is responsible for making sure no transactions are
active in this process before resizing. This is slightly lighter
weight than doing a full env_close/env_open cycle.
2013-08-28 19:12:59 -07:00
Howard Chu
2a28686e4f ITS#7672 fix mdb_dbi_flags
Use a txn instead of env argument. Only return persistent flags.
2013-08-27 06:28:30 -07:00
Hallvard Furuseth
4dd10d70a5 mdb_page_unspill(): Fix dirty_room in nested txn 2013-08-25 23:59:58 +02:00
Howard Chu
56c2a8d04a Another take on spill_pgs and deletes 2013-08-24 20:37:45 +01:00
Howard Chu
b917266ca1 Cleanup prev commit 2013-08-24 11:54:05 -07:00
Howard Chu
41bf5c906a Speedup deletion from spill_pgs 2013-08-24 19:51:47 +01:00
Howard Chu
6423c87749 ITS#7671 always return the key for a dup val 2013-08-23 07:36:07 -07:00
Howard Chu
e15a90c762 ITS#7670 Tweak cursor_del0
Always leave cursor pointing at "next" node, if any.
Find next sibling if we're already at end of current page.
2013-08-23 04:55:41 -07:00
Hallvard Furuseth
6616034eb9 mdb_page_spill(): Fix nested txn mt_dirty_room 2013-08-23 13:52:35 +02:00
Hallvard Furuseth
392be3a7c6 Simplify last commits 2013-08-23 13:52:35 +02:00
Howard Chu
45c4ed1805 Tweak prev commit
If somehow "need" is larger than the list size, don't
try to fake out the dirty list.
2013-08-22 21:29:53 -07:00
Howard Chu
7b1db13050 Less aggressive page_spill
spilling all possible pages is a waste because the majority will be
needed again. Just do 1/8th of the dirty list instead.
2013-08-22 18:51:48 -07:00
Howard Chu
c22a7d98ec Update caveats now that mdb_reader_check exists 2013-08-20 17:37:53 -07:00
Howard Chu
214e9ed07d Silence warnings 2013-08-14 09:57:25 -07:00
Hallvard Furuseth
82b27c09c9 Tweak ITS#7615 fix: Obey strict-aliasing rules.
Do not access a long long as an array of longs.
2013-08-14 15:21:59 +02:00
Howard Chu
76f0a02485 ITS#7664 better fix
For RDONLY, don't get lockfile until we know datafile exists.
Also, don't try to create a new datafile for me_mfd if someone
deleted it after we got me_fd.
2013-08-13 13:12:47 -07:00
Howard Chu
a77ffaaa5d Revert "ITS#7664 delete lockfile if there is no environment"
This reverts commit 8725838409.
2013-08-13 13:05:32 -07:00
Howard Chu
8725838409 ITS#7664 delete lockfile if there is no environment 2013-08-13 11:34:32 -07:00
Howard Chu
eaeeecc3b2 Fix uninit'd mem ref in mdb_page_split
Don't compare cursors beyond their depth. (detected by valgrind)
2013-08-12 12:43:42 -07:00
Howard Chu
6ecee1cbf9 Tweak NODUPDATA flag check
Should allow for the possibility of future flag combos.
2013-08-11 17:25:00 -07:00
Howard Chu
5a5d33ff33 Fix obscure MDB_MULTIPLE bug
If a key has a single existing value, and then a put (MDB_MULTIPLE)
is done where the first of the multiple values matches the existing
value, the put would return SUCCESS without writing any of the
values. Fixed to loop to the next value as intended.
2013-08-11 17:15:03 -07:00
Howard Chu
76f8b77e36 Point to mdb.c for docs on macros in Makefile 2013-08-10 16:19:44 -07:00
Hallvard Furuseth
9c49ef1d71 Avoid compiling debug args with C90 compilers.
Use double DPRINTF() parenthesis, so varargs macros are not needed.
Parenthesize MDB_DEBUG, in case the user defines it as foo || bar.
2013-08-10 13:50:10 +02:00
Hallvard Furuseth
76d31573a4 Refuse ops on bad txns 2013-08-10 13:23:30 +02:00
Hallvard Furuseth
9beea541e1 Replace EINVALs with generalized MDB_INCOMPATIBLE.
Also check xcursor instead of DUPSORT, it's slightly briefer.
2013-08-10 13:23:30 +02:00
John Hewson
ef899fe84e ITS#7656 fix install target 2013-08-10 04:13:24 -07:00
Salvador Ortiz
86551828ab ITS#7660 Add mdb_txn_env to API 2013-08-10 04:06:16 -07:00
Hallvard Furuseth
5645812f41 mdb_txn_commit(): Always commit if MDB_TXN_SPILLS.
Checking dirty_list was insufficient after a spill
with no named databases and no positioned cursors.
2013-08-10 12:29:45 +02:00
Salvador Ortiz
92eb220184 ITS#7661 mdb_dbi_flags(): Allow main DBI 2013-08-10 09:42:19 +02:00
Salvador Ortiz
335f0e4cee mdb_del must ignore data if db not opened with MDB_DUPSORT 2013-08-09 09:05:44 -07:00
Howard Chu
1ac3c8308f Add mdb_env_get_maxkeysize() 2013-08-09 04:51:33 -07:00
Hallvard Furuseth
1ecd86b14c Replace unpredictable EINVAL error returns.
Return EINVAL only for simple programmer errors.
2013-08-09 13:05:14 +02:00
Hallvard Furuseth
22c104c5d7 Re-fix reader-pid code 2013-08-09 12:54:42 +02:00
Hallvard Furuseth
38b11b1de1 mdb_cursorpages_mark: Mark current txn and no more.
Ignore parent txn cursors since it is the current txn's dirty_list
which will be flushed. But check the current txn also when clearing,
since cursors can have pages which are dirty in a parent.

Check !mc_xcursor instead of !MDB_DUPSORT.  Equivalent for valid
data, but a bit safer if the sub-DB flags are corrupt.
2013-08-08 19:57:52 +02:00
Hallvard Furuseth
2bd5d8102e Fix mdb_ovpage_free() vs. spill.
Ensure me_pghead has room before removing from spill/dirty list.
Don't return pages to me_pghead in nested txns, use mt_free_pgs.
2013-08-08 19:57:52 +02:00
Hallvard Furuseth
c3547e81f3 Fix page spilling when MDB_WRITEMAP.
mdb_page_spill(): Don't binary-search the unsorted dirty_list.
mdb_page_flush(): Don't overwrite unprocessed dirty_list items.
2013-08-08 19:57:51 +02:00
Hallvard Furuseth
ee5ba85565 Set MDB_TXN_ERROR when inconsistent txn state 2013-08-08 19:57:51 +02:00
Hallvard Furuseth
ffd2287b55 Factor out MDB_env. 2013-08-08 19:54:54 +02:00
Hallvard Furuseth
d66739f947 MDB_LOCK_VERSION -> MDB_LOCK_FORMAT.
Pid locking needs a different lockfile-version: MDB_env's with and
without pid locking must not coexist, they can sabotage each other.
Store MDB_LOCK_FORMAT = (version | "use locking" flag) instead.
2013-08-08 19:43:04 +02:00
Hallvard Furuseth
1878213092 Fix mdb_reader_pid().
Treat unexpected errors as "don't know".  Invert Pidcheck return
value, so nonzero including error codes = "the process may exist".

On Windows: Catch exited but still existing processes.  Handle
undefined PROCESS_QUERY_LIMITED_INFORMATION.
On Unix: don't trust F_GETLK error to leave the input alone,
the fcntl() doc seems unclear.
2013-08-08 19:43:04 +02:00
Howard Chu
636c2d2a29 Use proper printf format on Windows 2013-08-07 12:42:46 -07:00
Hallvard Furuseth
74ed1757a4 Silence warnings 2013-08-05 10:01:39 +02:00
Hallvard Furuseth
27435aa5ac Tweak comments 2013-08-05 09:55:57 +02:00