ad7933ba0c
Optimization for mdb_rpage_get()
...
The caller already knows if it's using an overflow page, so
pass the number of expected pages in instead of having to
map the page first and check the count there.
4 years ago
53799e51da
RPAGE_CACHE is now dynamically selectable
...
Behavior is controlled by MDB_REMAP_CHUNKS envflag
Remapping is always enabled in MDB_VL32
4 years ago
ac047d1eff
Add new MDB_RPAGE_CACHE def
...
Separate 32/64 dependency from rpage buffer mechanism
4 years ago
775be5e50f
Fix Android recognition
...
The official macro is __ANDROID__; ANDROID may or may not be defined.
8 years ago
348885413e
MDB_VL32
...
Use the same size dirty list for both 64 and 32 bit.
8 years ago
4d2154397a
ITS#8622 fix xcursor after cursor_del
...
Re-fix 6b1df0e4c7
8 years ago
b5e5fcc31d
ITS#8582 Fill in MDB_LOCK_FORMAT
...
Attempt to stop liblmdb variants compiled with conflicting options
from using the lockfile at the same time and thus breaking it.
8 years ago
52c0df1d50
Clear any struct padding in idbuf
8 years ago
172d825155
Simplify mdb_hash_val() -> mdb_hash()
...
Simpler usage since it's only called once, rename to match new usage,
and drop 3 loop pessimizations which were optimizations 20 years ago.
8 years ago
68eda68f0b
Store lock ID instead of pathname in lockfile
...
This limits the namespace which the user can meddle with for
POSIX semaphores and Windows mutexes. Their names change a
bit, they no longer have fixed lengths.
8 years ago
58ba039b8f
ITS#8582 MDB_LOCK_VERSION = 2 due to format change
8 years ago
e36517dbbe
ITS#8582 keep mutex at end of struct
...
since it's variable size on Linux/glibc
8 years ago
1db9f32ae2
Workaround VL32 cursor refcounting miscount
...
Don't try to deref cursor page if txn's pagelist is empty
8 years ago
882e27c1b4
Further fix f8ce8a8271
...
Fully revert the change to GET_MULTIPLE
8 years ago
5eae7aad7e
Fix f8ce8a8271
...
GET_MULTIPLE was broken
8 years ago
6ac9aa666e
Happy New Year
8 years ago
f8ce8a8271
More C_EOF tweaks
8 years ago
c44b29eaa8
ITS#8557 fix prev commit
8 years ago
511f5880e2
Tweak cursor_next C_EOF check
...
Allow C_EOF flag to be stale
8 years ago
d84dee516f
ITS#8557 fix mdb_cursor_last
...
Optimize mdb_page_search_root(PS_LAST) when cursor is already near
last position, ignoring C_EOF flag for now.
8 years ago
59ac317d2a
ITS#8558 fix mdb_load with escaped plaintext
8 years ago
2e3eaf2ce2
ITS#8554 kFreeBSD is like BSD
...
Doesn't have POSIX robust mutexes - GNU userland on BSD kernel
8 years ago
4bc270a2cb
More MDB_node doc
8 years ago
be94a7565b
Doxygen fixes. Use DISTRIBUTE_GROUP_DOC.
...
- DISTRIBUTE_GROUP_DOC makes doxygen give several fields the
same doc: mn_hi + mn_lo in MDB_node.
- With mdb_mutex_t + mdb_mutexref_t, instead split them up.
- Don't hide a doxygen #name inside double quotes.
8 years ago
72c893fc82
Mention MDB_PREV_MULTIPLE along with MDB_NEXT_MULTIPLE
8 years ago
e539654051
ITS#8504 Fix prev commit: mc_error, #ifdef SIGPIPE
...
Never clear mc_error, we could lose a failure in the other thread.
8 years ago
3e7a8e26e6
ITS#8504 mdb_env_copyfd2(): Don't abort on SIGPIPE
...
Return EPIPE instead.
8 years ago
65e95ffccf
ITS#7377 Catch mdb_cursor_sibling() error
8 years ago
d78c80d902
Clean up and comment C_UNTRACK
...
Don't use it as a "cursor is tracked" hint in mdb_pages_xkeep().
It's been harmless so far, but would break after mdb_cursor_copy().
Checking m0 directly short-circuits better anyway.
8 years ago
1fb0822b40
ITS#8355 cleanup
...
Drop ~C_EOF, pointless after 8c215aa970
8 years ago
c0ff9a267a
MDB_VL32 cleanup: MDB_env.mm_mapsize type
8 years ago
e1be73c771
ITS#8542 mdb_dbi_open(): Protect mainDB cursors
8 years ago
ffd13db3f8
Cleanup: Add flag DB_DUPDATA, drop DB_DIRTY hack
8 years ago
ef066598b5
MDB_CP_COMPACT comments
8 years ago
3dda2bfa4a
doxygen cleanup
8 years ago
72f875b48a
Note functions which must set MDB_TXN_ERROR on failure
...
Other functions depend on them to do so.
For mdb_node_read(), instead remove such a dependence.
8 years ago
08e4684d37
MDB_VL32 mdb_page_get(): Set MDB_TXN_ERROR on failure.
8 years ago
a70200f15f
Fix mdb_env_open() with MDB_VL32
...
Init me_rpmutex independently of MDB_NOLOCK.
Plug leaks on mdb_env_open() failure.
Tweak mdb_env_close0() to handle the rearranged mdb_env_open().
8 years ago
f3ab0d23a4
Factor out refreshing sub-page pointers
8 years ago
fa83b25ec5
More for ITS#8406
...
xcursor fixup depends on init state
8 years ago
e58db7d556
More for ITS#8406
...
Revert excess cursor fixup
8 years ago
f7e85d7804
Only set me_mfd if needed. Drop unused read access.
8 years ago
77845345ca
ITS#8505 Clarify fork() caveat, mdb_env_get_fd(), flock->fcntl.
8 years ago
6355dade31
ITS#8505 Protect parent from fork()-pthread_exit()
8 years ago
04acac634a
ITS#8505 Set FD_CLOEXEC for me_mfd,env_copy as well
8 years ago
15666878af
Move opening files to mdb_fopen()
...
No change in functionality.
8 years ago
cdc3f9cc2d
Factor filename handling out to mdb_fname_*()
...
No change in functionality, except needs less mallocing.
8 years ago
67fb3c746a
ITS#7992 Tighter utf8_to_utf16(), fix errcodes
...
The 0xFFFD check seems due to misleading MultiByteToWideChar() doc.
Bad UTF-8 gives 0xFFFD in the output string, not the return value.
8 years ago
d87ee20e0b
Clean up strange fcntl result check
...
...and check !MDB_CLOEXEC in an 'if' rather than '#if'
to match its non-zero usage.
8 years ago
3dd2d207d7
Drop spurious Errcode() call
8 years ago
Howard Chu