Added ability top change empty space filler of the loghandler.
Fixed end of log reaction.
Fixed memory corruprion bug caused by reading non-filled hage header.
Added debug output.
storage/maria/ma_bitmap.c:
Compiler warnings removed.
storage/maria/ma_blockrec.c:
Compiler warnings removed.
storage/maria/ma_loghandler.c:
Added ability top change empty space filler of the loghandler.
Fixed end of log reaction.
Fixed memory corruprion bug caused by reading non-filled hage header.
Added debug output.
storage/maria/ma_loghandler_lsn.h:
Compiler warnings removed.
moved debugging defines.
Fixed buffer flushing (we copied last page
content before it was actually written, now
we abvance pointer in new buffer and unlock
it while waiting for filling old buffer)
Misc changes:
- fix for benign Valgrind error, compiler warnings
- fix for a segfault in execution of maria_delete_all_rows() and one
when taking multiple checkpoints
- fix for too paranoid assertion
- adding ability to take checkpoints at the end of the REDO phase
and at the end of recovery.
- other minor changes
storage/maria/ha_maria.cc:
The checkpoint done after Recovery is finished, is moved to
maria_recover().
storage/maria/ma_bitmap.c:
fix for Valgrind error: the "shadow debug copy" of the bitmap page
started unitialized and so ma_print_bitmap() would use it uninitialized
storage/maria/ma_checkpoint.c:
* reset pointers to NULL after freeing them, or we segfault at
next checkpoint in my_realloc().
* fix for compiler warnings.
storage/maria/ma_delete_all.c:
info->trn is NULL for non-transactional tables
storage/maria/ma_locking.c:
correct assertion (it fired wrongly in execution of REDO_DROP_TABLE
due to the maria_extra(HA_PREPARE_FOR_DROP)->_ma_decrement_open_count()
->maria_lock_database(F_UNLCK); another solution would have been to
not call _ma_decrement_open_count() (it's ok to have a wrong open
count in a table which we are dropping), but the same problem
would still exist for REDO_RENAME_TABLE.
storage/maria/ma_loghandler.c:
fail early if UNRECOVERABLE_ERROR
storage/maria/ma_recovery.c:
* new argument to maria_apply_log(): should it take checkpoints
(at end of REDO phase and at the very end) or no.
* moving the call to translog_next_LSN() into
parse_checkpoint_record() ("hide the details").
* Refining an error detection for something which could happen
if there is a checkpoint record in the log.
* Using close_one_table() instead of maria_extra(HA_EXTRA_PREPARE_FOR_DROP|RENAME),
as it looks safer, and also changing how close_one_table() works:
it now limits itself to scanning all_tables[], thus having one loopp
instead of two, which should be faster (as a result, it does not
close tables not registered in this array, which is ok as there
should not be any).
storage/maria/ma_recovery.h:
new parameter
storage/maria/maria_read_log.c:
update to new prototype
Debug output information fixed.
Fixed direct page referencing in write mode.
storage/maria/ma_loghandler.c:
TODO added.
storage/maria/ma_pagecache.c:
Debug output information fixed.
Fixed direct page referencing in write mode.
* Recovery of the table's live checksum (CREATE TABLE ... CHECKSUM=1)
is achieved in this patch. The table's live checksum
(info->s->state.state.checksum) is updated in inwrite_rec_hook's
under the log mutex when writing UNDO_ROW_INSERT|UPDATE|DELETE
and REDO_DELETE_ALL. The checksum variation caused by the operation
is stored in these UNDOs, so that the REDO phase, when it sees such
UNDOs, can update the live checksum if it is older (state.is_of_lsn is
lower) than the record. It is also used, as a nice add-on with no
cost, to do less row checksum computation during the UNDO phase
(as we have it in the record already).
Doing this work, it became pressing to move in-write hooks
(write_hook_for_redo() et al) to ma_blockrec.c.
The 'parts' argument of inwrite_rec_hook is unpredictable (it comes
mangled at this stage, for example by LSN compression) so it is
replaced by a 'void* hook_arg', which is used to pass down information,
currently only to write_hook_for_clr_end() (previous undo_lsn and
type of undone record).
* If from ha_maria, we print to stderr how many seconds (with one
fractional digit) the REDO phase took, same for UNDO phase and for
final table close. Just to give an indication for debugging and maybe
also for Support.
storage/maria/ha_maria.cc:
question for Monty
storage/maria/ma_blockrec.c:
* log in-write hooks (write_hook_for_redo() etc) move from
ma_loghandler.c to here; this is natural: the hooks are coupled
to their callers (functions in ma_blockrec.c).
* translog_write_record() now has a new argument "hook_arg";
using it to pass down to write_hook_for_clr_end() the transaction's
previous_undo_lsn and the type of the being undone record, and also
to pass down to all UNDOs the live checksum variation caused by the
operation.
* If table has live checksum, store in UNDO_ROW_INSERT|UPDATE|DELETE
and in CLR_END the checksum variation ("delta") caused by the
operation. For example if a DELETE caused the table's live checksum
to change from 123 to 456, we store in the UNDO_ROW_DELETE, in 4 bytes,
the value 333 (456-123).
* Instead of hard-coded "1" as length of the place where we store
the undone record's type in CLR_END, use a symbol CLR_TYPE_STORE_SIZE;
use macros clr_type_store and clr_type_korr.
* write_block_record() has a new parameter 'old_record_checksum'
which is the pre-computed checksum of old_record; that value is used
to update the table's live checksum when writing UNDO_ROW_UPDATE|CLR_END.
* In allocate_write_block_record(), if we are executing UNDO_ROW_DELETE
the row's checksum is already computed.
* _ma_update_block_record2() now expect the new row's checksum into
cur_row.checksum (was already true) and the old row's checksum into
new_row.checksum (that's new). Its two callers, maria_update() and
_ma_apply_undo_row_update(), honour this.
* When executing an UNDO_ROW_INSERT|UPDATE|DELETE in UNDO phase, pick
up the checksum delta from the log record. It is then used to update
the table's live checksum when writing CLR_END, and saves us a
computation of record.
storage/maria/ma_blockrec.h:
in-write hooks move from ma_loghandler.c
storage/maria/ma_check.c:
more straightforward size of buffer
storage/maria/ma_checkpoint.c:
<= is enough
storage/maria/ma_commit.c:
new prototype of translog_write_record()
storage/maria/ma_create.c:
new prototype of translog_write_record()
storage/maria/ma_delete.c:
The row's checksum must be computed before calling(*delete_record)(),
not after, because it must be known inside _ma_delete_block_record()
(to update the table's live checksum when writing UNDO_ROW_DELETE).
If deleting from a transactional table, live checksum was already updated
when writing UNDO_ROW_DELETE.
storage/maria/ma_delete_all.c:
@todo is now done (in ma_loghandler.c)
storage/maria/ma_delete_table.c:
new prototype of translog_write_record()
storage/maria/ma_loghandler.c:
* in-write hooks move to ma_blockrec.c.
* translog_write_record() gets a new argument 'hook_arg' which is
passed down to pre|inwrite_rec_hook. It is more useful that 'parts'
for those hooks, because when those hooks are called, 'parts' has
possibly been mangled (like with LSN compression) and is so
unpredictable.
* fix for compiler warning (unused buffer_start when compiling without
debug support)
* Because checksum delta is stored into UNDO_ROW_INSERT|UPDATE|DELETE
and CLR_END, but only if the table has live checksum, these records
are not PSEUDOFIXEDLENGTH anymore, they are now VARIABLE_LENGTH (their
length is X if no live checksum and X+4 otherwise).
* add an inwrite_rec_hook for UNDO_ROW_UPDATE, which updates the
table's live checksum. Update it also in hooks of UNDO_ROW_INSERT|
DELETE and REDO_DELETE_ALL and CLR_END.
* Bugfix: when reading a record in translog_read_record(), it happened
that "length" became negative, because the function assumed that
the record extended beyond the page's end, whereas it may be shorter.
storage/maria/ma_loghandler.h:
* Instead of hard-coded "1" and "4", use symbols and macros
to store/retrieve the type of record which the CLR_END corresponds
to, and the checksum variation caused by the operation which logs the
record
* translog_write_record() gets a new argument 'hook_arg' which is
passed down to pre|inwrite_rec_hook. It is more useful that 'parts'
for those hooks, because when those hooks are called, 'parts' has
possibly been mangled (like with LSN compression) and is so
unpredictable.
storage/maria/ma_open.c:
fix for "empty body in if() statement" (when compiling without safemutex)
storage/maria/ma_pagecache.c:
<= is enough
storage/maria/ma_recovery.c:
* print the time that each recovery phase (REDO/UNDO/flush) took;
this is enabled only when recovering from ha_maria. Is it printed
n seconds with a fractional part of one digit (like 123.4 seconds).
* In the REDO phase, update the table's live checksum by using
the checksum delta stored in UNDO_ROW_INSERT|DELETE|UPDATE and CLR_END.
Update it too when seeing REDO_DELETE_ALL.
* In the UNDO phase, when executing UNDO_ROW_INSERT, if the table does
not have live checksum then reading the record's header (as done by
the master loop of run_undo_phase()) is enough; otherwise we
do a translog_read_record() to have the checksum delta ready
for _ma_apply_undo_row_insert().
* When at the end of the REDO phase we notice that there is an unfinished
group of REDOs, don't assert in debug binaries, as I verified that it
can happen in real life (with kill -9)
* removing ' in #error as it confuses gcc3
storage/maria/ma_rename.c:
new prototype of translog_write_record()
storage/maria/ma_test_recovery.expected:
Change in output of ma_test_recovery: now all live checksums of
original tables equal those of tables recreated by the REDO phase
and those of tables fixed by the UNDO phase. I.e. recovery of
the live checksum looks like working (which was after all the only
goal of this changeset).
I checked by hand that it's not just all live checksums which are
now 0 and that's why they match. They are the old values like
3757530372. maria.test has hard-coded checksum values in its result
file so checks this too.
storage/maria/ma_update.c:
* It's useless to put up HA_STATE_CHANGED in 'key_changed',
as we put up HA_STATE_CHANGED in info->update anyway.
* We need to compute the old and new rows' checksum before calling
(*update_record)(), as checksum delta must be known when logging
UNDO_ROW_UPDATE which is done by _ma_update_block_record(). Note that
some functions change the 'newrec' record (at least _ma_check_unique()
does) so we cannot move the checksum computation too early in the
function.
storage/maria/ma_write.c:
If inserting into a transactional table, live's checksum was
already updated when writing UNDO_ROW_INSERT. The multiplication
is a trick to save an if().
storage/maria/unittest/ma_test_loghandler-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_first_lsn-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_max_lsn-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_multigroup-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_multithread-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_noflush-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_pagecache-t.c:
new prototype of translog_write_record()
storage/maria/unittest/ma_test_loghandler_purge-t.c:
new prototype of translog_write_record()
storage/myisam/sort.c:
fix for compiler warnings in pushbuild (write_merge_key* functions
didn't have their declaration match MARIA_HA::write_key).
gprof build for amd64.
storage/maria/ma_loghandler.c:
fater way to detect last page address for the last log file.
BUILD/compile-amd64-gprof-no-ndb:
New BitKeeper file ``BUILD/compile-amd64-gprof-no-ndb''
REDO optimization (Bascily avoid moving blocks from/to pagecache)
More command line arguments to maria_read_log
Fixed recovery bug when recreating table
sql/opt_range.cc:
Remove SAFE_MODE for opt_range as it disables UPDATE to use keys
storage/maria/ma_blockrec.c:
REDO optimization
Use new interface for pagecache_reads to avoid copying page buffers
storage/maria/ma_loghandler.c:
Patch from Sanja:
- Added new parameter to translog_get_page to use direct links to pagecache
- Changed scanner to be able to use direct links
This avoids a lot of calls to bmove512() in page cache.
storage/maria/ma_loghandler.h:
Added direct link to pagecache objects
storage/maria/ma_open.c:
Added const to parameter
Added missing braces
storage/maria/ma_pagecache.c:
From Sanja:
- Added direct links to pagecache (from pagecache_read())
Dirrect link means that on pagecache_read we get back a pointer to the pagecache buffer
From Monty:
- Fixed arguments to init_page_cache to handle big page caches
- Fixed compiler warnings
- Replaced PAGECACHE_PAGE_LINK with PAGECACHE_BLOCK_LINK * to catch errors
storage/maria/ma_pagecache.h:
Changed block numbers from int to long to be able to handle big page caches
Changed some PAGECACHE_PAGE_LINK to PAGECACHE_BLOCK_LINK
storage/maria/ma_recovery.c:
Fixed recovery bug when recreating table (table was kept open)
Moved some variables to function start (portability)
Added space to some print messages
storage/maria/maria_chk.c:
key_buffer_size -> page_buffer_size
storage/maria/maria_def.h:
Changed default page_buffer_size to 10M
storage/maria/maria_read_log.c:
Added more startup options:
--version
--undo (apply undo)
--page_cache_size (to run with big cache sizes)
--silent (to not get any output from --apply)
storage/maria/unittest/ma_control_file-t.c:
Fixed compiler warning
storage/maria/unittest/ma_test_loghandler-t.c:
Added new argument to translog_init_scanner()
storage/maria/unittest/ma_test_loghandler_multigroup-t.c:
Added new argument to translog_init_scanner()
storage/maria/unittest/ma_test_loghandler_multithread-t.c:
Added new argument to translog_init_scanner()
record (for example in ma_test_loghandler-t). Reason was
wrongly matched () in DBUG.
storage/maria/ma_loghandler.c:
Wrongly matched parenthesis:
DBUG_PRINT(keyword, argslist) expands to roughly
_db_doprnt arglist;
So DBUG_PRINT("enter",(a);
b;
c);
expands to roughly
_db_doprnt(a);b;c;
which is valid code. Except that in non-debug builds, DBUG_PRINT(
expands to nothing so the wrongly "included" code is thrown away,
leading to some members of "scanner" to not be initialized.
horizon addres.
storage/maria/ma_loghandler.c:
New function to get correct LSN from chunk address.
storage/maria/ma_loghandler.h:
New function to get correct LSN from chunk address.
storage/maria/ma_checkpoint.c:
The new macro for easier printing LSN added.
storage/maria/ma_loghandler.c:
The assertion returned.
The new macro for easier printing LSN added.
storage/maria/ma_loghandler_lsn.h:
The new macro for easier printing LSN added.
storage/maria/ma_pagecache.c:
The new macro for easier printing LSN added.
storage/maria/ma_recovery.c:
Recovery checks empty log state.
RECHEADER_READ_ERROR means some real error.
storage/maria/maria_read_log.c:
Read log starts from real beggining of the log and precess
error and empty log states.
The new macro for easier printing LSN added.
storage/maria/unittest/ma_test_loghandler-t.c:
The new macro for easier printing LSN added.
storage/maria/unittest/ma_test_loghandler_first_lsn-t.c:
The new macro for easier printing LSN added.
storage/maria/unittest/ma_test_loghandler_max_lsn-t.c:
The new macro for easier printing LSN added.
storage/maria/unittest/ma_test_loghandler_multigroup-t.c:
The new macro for easier printing LSN added.
storage/maria/unittest/ma_test_loghandler_multithread-t.c:
The new macro for easier printing LSN added.
storage/maria/unittest/ma_test_loghandler_noflush-t.c:
The new macro for easier printing LSN added.
Finally this is the real checkpoint code.
It however exhibits unstabilities when a checkpoint runs concurrently
with data-modifying clients (table corruption, transaction log's
assertions) so for now a checkpoint is taken only at startup after
recovery and at shutdown, i.e. not in concurrent situations. Later
we will let it run periodically, as well as flush dirty pages
periodically (almost all needed code is there already, only pagecache
code is written but not committed).
WL#3072 Maria recovery
* replacing UNDO_ROW_PURGE with CLR_END; testing of those CLR_END via
ma_test2 which has INSERTs failing with duplicate keys.
* replaying of REDO_RENAME_TABLE
Now, off to test Recovery in ha_maria :)
BitKeeper/deleted/.del-ma_least_recently_dirtied.c:
Delete: storage/maria/ma_least_recently_dirtied.c
BitKeeper/deleted/.del-ma_least_recently_dirtied.h:
Delete: storage/maria/ma_least_recently_dirtied.h
storage/maria/Makefile.am:
compile Checkpoint module
storage/maria/ha_maria.cc:
When ha_maria starts, do a recovery from last checkpoint.
Take a checkpoint when that recovery has ended and when ha_maria
shuts down cleanly.
storage/maria/ma_blockrec.c:
* even if my_sync() fails we have to my_close() (otherwise we leak
a descriptor)
* UNDO_ROW_PURGE is replaced by a simple CLR_END for UNDO_ROW_INSERT,
as promised in the old comment; it gives us skipping during the
UNDO phase.
storage/maria/ma_check.c:
All REDOs before create_rename_lsn are ignored by Recovery. So
create_rename_lsn must be set only after all data/index has been
flushed and forced to disk. We thus move write_log_record_for_repair()
to after _ma_flush_tables_files_after_repair().
storage/maria/ma_checkpoint.c:
Checkpoint module.
storage/maria/ma_checkpoint.h:
optional argument if caller wants a thread to periodically take
checkpoints and flush dirty pages.
storage/maria/ma_create.c:
* no need to init some vars as the initial bzero(share) takes care of this.
* update to new function's name
* even if we fail in my_sync() we have to my_close()
storage/maria/ma_extra.c:
Checkpoint reads share->last_version under intern_lock, so we make
maria_extra() update it under intern_lock. THR_LOCK_maria still needed
because of _ma_test_if_reopen().
storage/maria/ma_init.c:
destroy checkpoint module when Maria shuts down.
storage/maria/ma_loghandler.c:
* UNDO_ROW_PURGE gone (see ma_blockrec.c)
* we need to remember the LSN of the LOGREC_FILE_ID for a share,
because this LSN is needed into the checkpoint record (Recovery wants
to know the validity domain of an id->name mapping)
* translog_get_horizon_no_lock() needed for Checkpoint
* comment about failing assertion (Sanja knows)
* translog_init_reader_data() thought that translog_read_record_header_scan()
returns 0 in case of error, but 0 just means "0-length header".
* translog_assign_id_to_share() now needs the MARIA_HA because
LOGREC_FILE_ID uses a log-write hook.
* Verify that (de)assignment of share->id happens only under intern_lock,
as Checkpoint reads this id with intern_lock.
* translog_purge() can accept TRANSLOG_ADDRESS, not necessarily
a real LSN.
storage/maria/ma_loghandler.h:
prototype updates
storage/maria/ma_open.c:
no need to initialize "res"
storage/maria/ma_pagecache.c:
When taking a checkpoint, we don't need to know the maximum rec_lsn
of dirty pages; this LSN was intended to be used in the two-checkpoint
rule, but last_checkpoint_lsn is as good.
4 bytes for stored_list_size is enough as PAGECACHE::blocks (number
of blocks which the pagecache can contain) is int.
storage/maria/ma_pagecache.h:
new prototype
storage/maria/ma_recovery.c:
* added replaying of REDO_RENAME_TABLE
* UNDO_ROW_PURGE gone (see ma_blockrec.c), replaced by CLR_END
* Recovery from the last checkpoint record now possible
* In new_table() we skip the table if the id->name mapping is older than
create_rename_lsn (mapping dates from lsn_of_file_id).
* in get_MARIA_HA_from_REDO_record() we skip the record
if the id->name mapping is newer than the record (can happen if processing
a record which is before the checkpoint record).
* parse_checkpoint_record() has to return a LSN, that's what caller expects
storage/maria/ma_rename.c:
new function's name; log end zeroes of tables' names (ease recovery)
storage/maria/ma_test2.c:
* equivalent of ma_test1's --test-undo added (named -u here).
* -t=1 now stops right after creating the table, so that
we can test undoing of INSERTs with duplicate keys (which tests the
CLR_END logged by _ma_write_abort_block_record()).
storage/maria/ma_test_recovery.expected:
Result of testing undoing of INSERTs with duplicate keys; there are
some differences in maria_chk -dvv but they are normal (removing
records does not shrink data/index file, does not put back the
"analyzed, optimized keys"(etc) index state.
storage/maria/ma_test_recovery:
Test undoing of INSERTs with duplicate keys, using ma_test2;
when such INSERT happens, it logs REDO_INSERT, UNDO_INSERT, REDO_DELETE,
CLR_END; we abort after that, and test that CLR_END causes recovery
to jump over UNDO_INSERT.
storage/maria/ma_write.c:
comment
storage/maria/maria_chk.c:
comment
storage/maria/maria_def.h:
* a new bit in MARIA_SHARE::in_checkpoint, used to build a list
of unique shares during Checkpoint.
* MARIA_SHARE::lsn_of_file_id added: the LSN of the last LOGREC_FILE_ID
for this share; needed to know to which LSN domain the mappings
found in the Checkpoint record apply (new mappings should not apply
to old REDOs).
storage/maria/trnman.c:
* small changes to how trnman_collect_transactions() fills its buffer;
it also uses a non-dummy lsn_read_non_atomic() found in ma_checkpoint.h
* testing of execution of UNDO_ROW_UPDATE
* when executing an UNDO_ROW_UPDATE, store "UNDO_ROW_UPDATE" as
"type of undone record" into the CLR_END record.
storage/maria/ma_blockrec.c:
When logging a CLR_END in write_block_record(), it can be for
a DELETE or for an UPDATE (now that Monty has coded execution of
UNDO_UPDATE)
storage/maria/ma_loghandler.c:
UNDO_ROW_UPDATE's execution coded, so no crash
storage/maria/ma_recovery.c:
UNDO_ROW_UPDATE's execution now coded, so no crash
storage/maria/ma_test1.c:
upper case letter
storage/maria/ma_test_recovery.expected:
output of testing execution of UNDO_ROW_UPDATE. Table's checksum
not recovered (known issue not specific to UPDATE).
storage/maria/ma_test_recovery:
Test execution of UNDO_ROW_UPDATE: first we stop ma_test1 after
deletes and commit, then we stop ma_test1 after updates and abort;
we verify that updates are rolled back by comparing tables
into mysql.com:/home/my/mysql-maria
storage/maria/ma_check.c:
Auto merged
storage/maria/ma_locking.c:
Auto merged
storage/maria/ma_loghandler.c:
Auto merged
storage/maria/ma_open.c:
Auto merged
storage/maria/ma_recovery.c:
Auto merged
storage/maria/maria_def.h:
Auto merged
storage/maria/maria_read_log.c:
Auto merged
storage/maria/ma_blockrec.c:
Manual merge
storage/maria/ma_test1.c:
Manual merge (using Guilhems code)
Fixed bug in duplicate key handling for block records during repair
All read-row methods now return error number in case of error
Don't calculate checksum for null fields
Fixed bug when running maria_read_log with -o
BUILD/SETUP.sh:
Added STACK_DIRECTION
BUILD/compile-pentium-debug-max:
Moved STACK_DIRECTION to SETUP
include/myisam.h:
Added extra parameter to write_key
storage/maria/ma_blockrec.c:
Added applying of undo for updates
Fixed indentation
Removed some not needed casts
Fixed wrong logging of CLR record
Split ma_update_block_record to two functions to be able to reuse it from undo-applying
Simplify filling of packed fields
ma_record_block_record) now returns error number on failure
Sligtly changed log record information for undo-update
storage/maria/ma_check.c:
Fixed bug in duplicate key handling for block records during repair
storage/maria/ma_checksum.c:
Don't calculate checksum for null fields
storage/maria/ma_dynrec.c:
_ma_read_dynamic_reocrd() now returns error number on error
Rest of the changes are code simplification and indentation fixes
storage/maria/ma_locking.c:
Added comment
storage/maria/ma_loghandler.c:
More debugging
Removed printing of total_record_length as this was always same as record_length
storage/maria/ma_open.c:
Allocate bitmap for changed fields
storage/maria/ma_packrec.c:
read_record now returns error number on error
storage/maria/ma_recovery.c:
Fixed wrong arguments to undo_row_update
storage/maria/ma_statrec.c:
read_record now returns error number on error (not 1)
Code simplification
storage/maria/ma_test1.c:
Added exit possibility after update phase (to test undo of updates)
storage/maria/maria_def.h:
Include bitmap header file
storage/maria/maria_read_log.c:
Fixed bug when running with -o
end of translog_flush() when datadir was in /dev/shm.
storage/maria/ma_loghandler.c:
directory syncing can fail on shared memory devices (/dev/shm on Linux
in this case); see my_sync_dir().
Recovery of state.records (the count of records which is stored into
the header of the index file). For that, state.is_of_lsn is introduced;
logic is explained in ma_recovery.c (look for "Recovery of the state").
The net gain is that in case of crash, we now recover state.records,
and it is idempotent (ma_test_recovery tests it).
state.checksum is not recovered yet, mail sent for discussion.
- WL#3071 Maria Checkpoint: preparation for it, by protecting
all modifications of the state in memory or on disk with intern_lock
(with the exception of the really-often-modified state.records,
which is now protected with the log's lock, see ma_recovery.c
(look for "Recovery of the state"). Also, if maria_close() sees that
Checkpoint is looking at this table it will not my_free() the share.
- don't compute row's checksum twice in case of UPDATE (correction
to a bugfix I made yesterday).
storage/maria/ha_maria.cc:
protect state write with intern_lock (against Checkpoint)
storage/maria/ma_blockrec.c:
* don't reset trn->rec_lsn in _ma_unpin_all_pages(), because it
should wait until we have corrected the allocation in the bitmap
(as the REDO can serve to correct the allocation during Recovery);
introducing _ma_finalize_row() for that.
* In a changeset yesterday I moved computation of the checksum
into write_block_record(), to fix a bug in UPDATE. Now I notice
that maria_update() already computes the checksum, it's just that
it puts it into info->cur_row while _ma_update_block_record()
uses info->new_row; so, removing the checksum computation from
write_block_record(), putting it back into allocate_and_write_block_record()
(which is called only by INSERT and UNDO_DELETE), and copying
cur_row->checksum into new_row->checksum in _ma_update_block_record().
storage/maria/ma_check.c:
new prototypes, they will take intern_lock when writing the state;
also take intern_lock when changing share->kfile. In both cases
this is to protect against Checkpoint reading/writing the state or reading
kfile at the same time.
Not updating create_rename_lsn directly at end of write_log_record_for_repair()
as it wouldn't have intern_lock.
storage/maria/ma_close.c:
Checkpoint builds a list of shares (under THR_LOCK_maria), then it
handles each such share (under intern_lock) (doing flushing etc);
if maria_close() freed this share between the two, Checkpoint
would see a bad pointer. To avoid this, when building the list Checkpoint
marks each share, so that maria_close() knows it should not free it
and Checkpoint will free it itself.
Extending the zone covered by intern_lock to protect against
Checkpoint reading kfile, writing state.
storage/maria/ma_create.c:
When we update create_rename_lsn, we also update is_of_lsn to
the same value: it is logical, and allows us to test in maria_open()
that the former is not bigger than the latter (the contrary is a sign
of index header corruption, or severe logging bug which hinders
Recovery, table needs a repair).
_ma_update_create_rename_lsn_on_disk() also writes is_of_lsn;
it now operates under intern_lock (protect against Checkpoint),
a shortcut function is available for cases where acquiring
intern_lock is not needed (table's creation or first open).
storage/maria/ma_delete.c:
if table is transactional, "records" is already decremented
when logging UNDO_ROW_DELETE.
storage/maria/ma_delete_all.c:
comments
storage/maria/ma_extra.c:
Protect modifications of the state, in memory and/or on disk,
with intern_lock, against a concurrent Checkpoint.
When state goes to disk, update it's is_of_lsn (by calling
the new _ma_state_info_write()).
In HA_EXTRA_FORCE_REOPEN, don't set share->changed to 0 (undoing
a change I made a few days ago) and ASK_MONTY
storage/maria/ma_locking.c:
no real code change here.
storage/maria/ma_loghandler.c:
Log-write-hooks for updating "state.records" under log's mutex
when writing/updating/deleting a row or deleting all rows.
storage/maria/ma_loghandler_lsn.h:
merge (make LSN_ERROR and LSN_REPAIRED_BY_MARIA_CHK different)
storage/maria/ma_open.c:
When opening a table verify that is_of_lsn >= create_rename_lsn; if
false the header must be corrupted.
_ma_state_info_write() is split in two: _ma_state_info_write_sub()
which is the old _ma_state_info_write(), and _ma_state_info_write()
which additionally takes intern_lock if requested (to protect
against Checkpoint) and updates is_of_lsn.
_ma_open_keyfile() should change kfile.file under intern_lock
to protect Checkpoint from reading a wrong kfile.file.
storage/maria/ma_recovery.c:
Recovery of state.records: when the REDO phase sees UNDO_ROW_INSERT
which has a LSN > state.is_of_lsn it increments state.records.
Same for UNDO_ROW_DELETE and UNDO_ROW_PURGE.
When closing a table during Recovery, we know its state is at least
as new as the current log record we are looking at, so increase
is_of_lsn to the LSN of the current log record.
storage/maria/ma_rename.c:
update for new behaviour of _ma_update_create_rename_lsn_on_disk().
storage/maria/ma_test1.c:
update to new prototype
storage/maria/ma_test2.c:
update to new prototype (actually prototype was changed days ago,
but compiler does not complain about the extra argument??)
storage/maria/ma_test_recovery.expected:
new result file of ma_test_recovery. Improvements: record
count read from index's header is now always correct.
storage/maria/ma_test_recovery:
"rm" fails if file does not exist. Redirect stderr of script.
storage/maria/ma_write.c:
if table is transactional, "records" is already incremented when
logging UNDO_ROW_INSERT. Comments.
storage/maria/maria_chk.c:
update is_of_lsn too
storage/maria/maria_def.h:
- MARIA_STATE_INFO::is_of_lsn which is used by Recovery. It is stored
into the index file's header.
- Checkpoint can now mark a table as "don't free this", and maria_close()
can reply "ok then you will free it".
- new functions
storage/maria/maria_pack.c:
update for new name
misc fixes of execution of UNDOs in the UNDO phase:
- into the CLR_END, store the LSN of the _previous_ UNDO (we debated
what was best, so far we're going with "previous"; later we can change
to "current" if needed), and store the type of record which is being
undone (needed to know how to update state.records when we see the
CLR_END during the REDO phase).
- declaring all UNDOs and CLR_END as "compressed"
- when executing an UNDO in the UNDO phase, state.records is updated
as a hook when writing CLR_END (needed for "recovery of the state"),
and so is trn->undo_lsn (needed for when we have checkpoints).
- bugfix (execution of UNDO_ROW_DELETE didn't store the correct checksum
into the re-inserted row, maria_chk -r thus threw the row away).
- modifications of ma_test1: where to stop is now driven by --testflag;
--test-undo just tells how to stop (flush data, flush log, nothing).
- ma_test_recovery: testing of the UNDO phase, more testing of the
REDO phase, identification of a bug.
storage/maria/ma_blockrec.c:
- bugfix: execution of UNDO_ROW_DELETE didn't store the correct
checksum into the row (leading to "maria_chk -r" eliminating the
re-inserted row, net effect was that rollback appeared to have
rolled back no deletion). Reason was that write_block_record() used
info->cur_row.checksum, while "row" can be != &info->cur_row
(case of UNDO_ROW_DELETE). After fixing this, problems with
_ma_update_block_record() appeared; indeed checksum was computed
by allocate_and_write_block_record() while _ma_update_block_record()
directly calls write_block_record(). Solution is to compute checksum
in write_block_record() instead.
- when executing an UNDO, we now pass the LSN of the _previous_ UNDO
to block_format functions. This LSN can be 0 (if the being-executed UNDO
was the transaction's first UNDO), so "undo_lsn==0" cannot work
anymore to indicate "this is not UNDO work". Using undo_lsn==LSN_ERROR
instead (this is an impossible LSN).
- store into CLR_END the type of log record which was undone
(INSERT/UPDATE/DELETE); needed for Recovery to know if/how it has
to update state.records if it sees this CLR_END in the REDO phase.
- when writing the CLR_END in _ma_apply_undo_row_insert(),
the place to store file's id is log_data+LSN_STORE_SIZE.
- in _ma_apply_undo_row_insert(), the records-- is moved
to a hook when writing the CLR_END (this way it is under log's mutex
which is needed for "recovery of the state")
storage/maria/ma_loghandler.c:
- all UNDOs, and CLR_END, start with the LSN of another UNDO; so
we can declare them "compressed".
- write_hook_for_clr_end() to set trn->undo_lsn (to the previous
UNDO's LSN) under log's lock (like UNDOs set trn->undo_lsn under log's
lock), and also update, if appropriate, state.records.
- reset share->id to 0 when deassigning; not useful for now but
sounds logical.
storage/maria/ma_recovery.c:
- if no table is found for a REDO, it's not an error; for an UNDO, it is
- in the REDO phase, when we see a CLR_END we must update trn->undo_lsn
and sometimes state.records.
- in the UNDO phase, when we execute an UNDO_ROW_INSERT:
* update trn->undo_lsn only after executing the record
* store the _previous_ undo_lsn into the CLR_END
- at the end of the REDO phase, when we recreate TRN objects, they
have already their long id in the log (either via a
LOGREC_LONG_TRANSACTION_ID, or in a checkpoint record), don't write
a new, useless LOGREC_LONG_TRANSACTION_ID for them.
storage/maria/ma_test1.c:
* where to stop execution is now driven by --testflag and not --test-undo
(ma_test2 already has --testflag for the same purpose). This allows
us to do a clean stop (with commit) at any point.
* --test-undo=# tells how to abort (flush all pages (which implies
flushing log) or only log or nothing); all such "ways of crashing"
are tested in ma_test_recovery
storage/maria/ma_test_recovery:
* Testing execution of UNDOs, with and without BLOBs.
* Testing idempotency of REDOs.
* See @todo for a probable bug with BLOBs.
* maria_chk -rq instead of -r, as with -q it nicely stops on any
problem in the data file (like the checksum bug see comment of
ma_blockrec.c).
* Testing if log was written by UNDO phase (often expected),
not written by REDO phase (always expected).
* Less output on the screen, compares with expected output in the end.
* some shell thingies like "set --" and $# are courtesy of
Danny and Pekka.
storage/maria/maria_read_log.c:
when only displaying the records, don't do an UNDO phase
storage/maria/ma_test_recovery.expected:
This is the expected output of a great part of ma_test_recovery.
ma_test_recovery compares its output to the expected output
and tells if different.
If we look at this file it mentions differences in checksum
(normal, it's not recovered yet) and in records count
(getting a correct records' count when recovery starts on an
already existing table, like when testing rollback,
is coded but not yet pushed).
Small fixes made.
storage/maria/ma_loghandler.c:
Check of transaction log descriptor table consistance added.\
Incorrect record description fixed.
Compiler warning fixed.
storage/maria/ma_loghandler.h:
fixed ident.
storage/maria/unittest/ma_test_loghandler-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_first_lsn-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_max_lsn-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_multigroup-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_multithread-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_noflush-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_pagecache-t.c:
Suppressing of automatic record writing
storage/maria/unittest/ma_test_loghandler_purge-t.c:
Suppressing of automatic record writing
storage/maria/ma_blockrec.c:
Added UNDO handling of insert during recovery
To do this, I also had to add write locking of tail pages during undo phase (As we need to access the same page twice if extents are split over two pages)
Another way to handle the undo of insert would be to store the extent information as part of the UNDO_INSERT block.
storage/maria/ma_blockrec.h:
Added new prototype
storage/maria/ma_loghandler.c:
Changed type of CLR_END (to avoid crash in log handler)
Removed not used variable
storage/maria/ma_loghandler.h:
Added TRN argument to record_execute_in_undo_phase()
storage/maria/ma_pagecache.c:
Hack for undo phase of recovery. During REDO we work with PLAIN pages, but UNDO works with LSN pages, which caused an abort when trying to access a cached page.
storage/maria/ma_recovery.c:
Added execution of UNDO_ROW_INSERT
storage/maria/ma_test1.c:
Added option --test-undo for testing recovery with undo
storage/maria/maria_read_log.c:
Added processing of undos
into desktop.sanja.is.com.ua:/home/bell/mysql/bk/work-maria-purge
storage/maria/ma_loghandler.c:
Auto merged
storage/maria/ma_loghandler.h:
Auto merged
into gbichot4.local:/home/mysql_src/mysql-maria-for-undo-phase
storage/maria/ha_maria.cc:
Auto merged
storage/maria/ma_blockrec.c:
Auto merged
storage/maria/ma_loghandler.c:
Auto merged
storage/maria/ma_loghandler.h:
Auto merged
storage/maria/ma_loghandler_lsn.h:
Auto merged
storage/maria/maria_chk.c:
Auto merged
storage/maria/maria_read_log.c:
Auto merged
* create page cache before initializing engine and not after, because
Maria's recovery needs a page cache
* make the creation of a bitmap page more crash-resistent
* bugfix (see ma_blockrec.c)
* back to old way: create an 8k bitmap page when creating table
* preparations for the UNDO phase: recreate TRNs
* preparations for Checkpoint: list of dirty pages, testing
of rec_lsn to know if page should be skipped during Recovery
(unused in this patch as no Checkpoint module pushed yet)
* maria_chk tags repaired table with a special LSN
* reworking all around in ma_recovery.c (less duplication)
mysys/my_realloc.c:
noted an issue in my_realloc()
sql/mysqld.cc:
page cache needs to be created before engines are initialized,
because Maria's initialization may do a recovery which needs
the page cache.
storage/maria/ha_maria.cc:
update to new prototype
storage/maria/ma_bitmap.c:
when creating the first bitmap page we used chsize to 8192 bytes then
pwrite (overwrite) the last 2 bytes (8191-8192). If crash between
the two operations, this leaves a bitmap page full without its end
marker. A later recovery may try to read this page and find it
exists and misses a marker and conclude it's corrupted and fail.
Changing the chsize to only 8190 bytes: recovery will then find
the page is too short and recreate it entirely.
storage/maria/ma_blockrec.c:
Fix for a bug: when executing a REDO, if the data page is created,
data_file_length was increased before _ma_bitmap_set():
_ma_bitmap_set() called _ma_read_bitmap_page() which, due to the
increased data_file_length, expected to find a bitmap page on disk
with a correct end marker; if the bitmap page didn't exist already
in fact, this failed. Fixed by increasing data_file_length only after
_ma_read_bitmap_page() has created the new bitmap page correctly.
This bug could happen every time a REDO is about creating a new
bitmap page.
storage/maria/ma_check.c:
empty data file has a bitmap page
storage/maria/ma_control_file.c:
useless parameter to ma_control_file_create_or_open(), just
test if this is recovery.
storage/maria/ma_control_file.h:
new prototype
storage/maria/ma_create.c:
Back to how it was before: maria_create() creates an 8k bitmap page.
Thus (bugfix) data_file_length needs to reflect this instead of being 0.
storage/maria/ma_loghandler.c:
as ma_test1 and ma_test2 now use real transactions and not
dummy_transaction_object, REDO for INSERT/UPDATE/DELETE are always
about real transactions, can assert this.
A function for Recovery to assign a short id to a table.
storage/maria/ma_loghandler.h:
new function
storage/maria/ma_loghandler_lsn.h:
maria_chk tags repaired tables with this LSN
storage/maria/ma_open.c:
* enforce that DMLs on transactional tables use real transactions
and not dummy_transaction_object.
* test if table was repaired with maria_chk (which has to been
seen as an import of an external table into the server), test
validity of create_rename_lsn (header corruption detection)
* comments.
storage/maria/ma_recovery.c:
* preparations for the UNDO phase: recreate TRNs
* preparations for Checkpoint: list of dirty pages, testing
of rec_lsn to know if page should be skipped during Recovery
(unused in this patch as no Checkpoint module pushed yet)
* reworking all around (less duplication)
storage/maria/ma_recovery.h:
a parameter to say if the UNDO phase should be skipped
storage/maria/maria_chk.c:
tag repaired tables with a special LSN
storage/maria/maria_read_log.c:
* update to new prototype
* no UNDO phase in maria_read_log for now
storage/maria/trnman.c:
* a function for Recovery to create a transaction (TRN), needed
in the UNDO phase
* a function for Recovery to grab an existing transaction, needed
in the UNDO phase (rollback all existing transactions)
storage/maria/trnman_public.h:
new functions
Now ma_test1 -M -T and ma_test2 -M -T produces readable, applyable logs
Note: The .MAD file is not binary identical after applying redo compare to a an original file.
(This is becasue we don't have full information which function called PURGE_REDO_BLOCKS).
To verify if a file was correctly applied, we now instead compare row checksums
BitKeeper/etc/ignore:
added storage/maria/tmp/*
include/maria.h:
Added maria_commit() and maria_begin() to be used with external tests
storage/maria/ha_maria.cc:
Ensure maria_def. is read in C mode
storage/maria/ma_blockrec.c:
Fixed redo handling.
_ma_apply_redo_purge_blocks() updated to handle any number of purged blocks
Removed code to make data file idenitcal after redo (can't easily be done). See changeset comments
Now ma_test1 -M -T and ma_test2 -M -T produces readable, applyable logs
storage/maria/ma_commit.c:
More DBUG statements
Moved variable declaration to start of function (portability fix)
Added helper functions 'maria_commit()' and 'maria_begin()'
storage/maria/ma_loghandler.c:
Fixed wrong REDO_PURGE_BLOCKS initialization
storage/maria/ma_recovery.c:
Added UNDO_ROW_UPDATE
Removed wrong setting of lsn (there was no lsn at the used position)
Fixed REDO_PURGE_BLOCKS to handle any number of blocks
storage/maria/ma_test1.c:
Added transaction support (via maria_begin() & maria_commit()) to get a log that can be applied with maria_read_log
storage/maria/ma_test2.c:
Added transaction support (via maria_begin() & maria_commit()) to get a log that can be applied with maria_read_log
storage/maria/ma_test_recovery:
Create temporary files in maria/tmp
Verify files with checksums instead of byte comparisons
storage/maria/maria_chk.c:
When using with -dss we only get filename, records and checksum.
This is useful to do a quick comparision if a files is identical to another one.
storage/maria/maria_def.h:
Added ma_commit()
storage/maria/maria_read_log.c:
Added --help
File number types fixed.
storage/maria/unittest/Makefile.am:
Test of purger added.
storage/maria/unittest/ma_test_loghandler_purge-t.c:
New BitKeeper file ``storage/maria/unittest/ma_test_loghandler_purge-t.c''
in the file to the file header.
storage/maria/ma_loghandler.h:
Getting maximum LSN of the record which parts written
in the file to the file header.
storage/maria/unittest/Makefile.am:
Test suite for getting max LSN added.
storage/maria/unittest/ma_test_loghandler_first_lsn-t.c:
Spelling fixed.
Cleanup fixed.
storage/maria/unittest/ma_test_loghandler_noflush-t.c:
Cleanup fixed.
storage/maria/unittest/ma_test_loghandler_max_lsn-t.c:
New BitKeeper file ``storage/maria/unittest/ma_test_loghandler_max_lsn-t.c''
storage/maria/ma_checkpoint.c:
Definitions of LSN should be collected in
the one file (ma_loghandler_lsn.h)
storage/maria/ma_loghandler.c:
New calls to get first theoretical and first real LSN.
storage/maria/ma_loghandler.h:
New calls to get first theoretical and first real LSN.
storage/maria/ma_loghandler_lsn.h:
Defined yet another impossible LSN to indicate error.
storage/maria/ma_recovery.c:
The first LSN call changed.
storage/maria/maria_read_log.c:
The first LSN call changed.
storage/maria/unittest/Makefile.am:
New unittest added.
storage/maria/unittest/ma_test_loghandler_first_lsn-t.c:
New BitKeeper file ``storage/maria/unittest/ma_test_loghandler_first_lsn-t.c''
into desktop.sanja.is.com.ua:/home/bell/mysql/bk/work-maria
storage/maria/ma_pagecache.c:
Auto merged
storage/maria/unittest/ma_test_loghandler-t.c:
Auto merged
storage/maria/ma_loghandler.c:
merge
storage/maria/ma_pagecache.h:
merge
Already written pages injection to the cache fixed.
storage/maria/ma_loghandler.c:
Ability to read unflushed data added.
storage/maria/ma_page.c:
Parameters added
storage/maria/ma_pagecache.c:
Already written pages injection to the cache fixed.
Validator for case of page content injection added.
storage/maria/ma_pagecache.h:
Validator for case of page content injection added.
storage/maria/unittest/Makefile.am:
Test of reading unflushed data
storage/maria/unittest/ma_test_loghandler-t.c:
Define fixed.
Restart of the log removed.
storage/maria/unittest/ma_test_loghandler_noflush-t.c:
New BitKeeper file ``storage/maria/unittest/ma_test_loghandler_noflush-t.c''
Preparation: for the two-checkpoint rule we will flush all pages
which have block->rec_lsn <= last_checkpoint_lsn. Pages with
rec_lsn not yet set (i.e. ==0) were dirtied very recently and so should
not be flushed (otherwise it's not efficient: we should try to let
a page be modified several times in memory before we flush it to disk).
To make this easy, "block->rec_lsn not yet set" is now expressed
with block->rec_lsn==LSN_MAX, not block->rec_lsn==0 anymore.
It is easier this way because LSN_MAX>last_checkpoint_lsn whereas
0<=last_checkpoint_lsn.
storage/maria/ma_blockrec.c:
typo
storage/maria/ma_loghandler.c:
typo
storage/maria/ma_loghandler_lsn.h:
LSN_MAX
storage/maria/ma_pagecache.c:
"block->rec_lsn not yet set" is now expressed by block->rec_lsn==LSN_MAX,
not block_rec_lsn==0.
preparation to enable recoverability of the table's state, more exactly
info->state->checksum. This will require that info->state->checksum
be updated in a inwrite_hook when writing an UNDO record, thus
info->cur_row.checksum needs to be accessible to inwrite_hook, so we
make translog_write_record() accept a MARIA_HA* (info) instead of
MARIA_SHARE* (info->s); with this, we will be able to access
info->cur_row.checksum. Old code which needed the MARIA_SHARE
can derive it from MARIA_HA. Fix for typos and compiler warnings.
storage/maria/ma_blockrec.c:
fix for new loghandler API. Removing strange lines (how could gcc
accept that?)
storage/maria/ma_check.c:
fix for new loghandler API
storage/maria/ma_delete_all.c:
fix for new loghandler API
storage/maria/ma_loghandler.c:
functions now take a MARIA_HA in argument, this is more powerful
than a MARIA_SHARE (MARIA_SHARE can be derived from MARIA_HA, not
the other way around). MARIA_HA will be needed to allow recoverability
of the table's state.
Fixing wrong DBUG_PRINT ('i' is not the id).
When writing the LOGREC_FILE_ID, we don't have a MARIA_HA around,
so we cannot ask translog_write_record() to store the id for us;
we thus store the file's id by ourselves. Alternative would have been
to pass MARIA_HA to translog_assign_id_to_share() but I didn't like it.
storage/maria/ma_loghandler.h:
new loghandler API
storage/maria/tablockman.c:
fix for compiler warning (intptr is int on my machine)
errors when reading a log record which has a 0-length header
(like LOGREC_REDO_DROP_TABLE).
storage/maria/ma_loghandler.c:
Functions reading record's header now don't use 0 to indicate error,
as some valid records have a 0-length header (like REDO_DROP_TABLE).
Instead, negative values are used for EOF and error.
storage/maria/ma_loghandler.h:
functions to read record's header now return an int
(either the length of this header (>=0) or some negative values
for EOF or error).
storage/maria/ma_recovery.c:
update to the new log handler's behaviour. Note the @todo.
storage/maria/maria_read_log.c:
inform when program failed
storage/maria/unittest/ma_test_loghandler-t.c:
update to new log handler's API
storage/maria/unittest/ma_test_loghandler_multigroup-t.c:
update to new log handler's API
storage/maria/unittest/ma_test_loghandler_multithread-t.c:
update to new log handler's API
Jul 7th). "maria.test" and "ps_maria.test" still fail;
"ma_test_all" starts failing (MyISAM has the same issue see BUG#30094).
include/maria.h:
merging MyISAM changes into Maria
mysys/mf_keycache.c:
mi_test_all showed "floating point exception", this was already
fixed in the latest 5.1, importing fix.
sql/item_xmlfunc.cc:
compiler warning (already fixed in latest 5.1)
storage/maria/ha_maria.cc:
merging MyISAM changes into Maria. See #ifdef ASK_MONTY.
storage/maria/ha_maria.h:
merging MyISAM changes into Maria
storage/maria/ma_cache.c:
merging MyISAM changes into Maria
storage/maria/ma_check.c:
merging MyISAM changes into Maria
storage/maria/ma_create.c:
merging MyISAM changes into Maria
storage/maria/ma_dynrec.c:
merging MyISAM changes into Maria
storage/maria/ma_extra.c:
merging MyISAM changes into Maria
storage/maria/ma_ft_boolean_search.c:
merging MyISAM changes into Maria
storage/maria/ma_ft_nlq_search.c:
merging MyISAM changes into Maria
storage/maria/ma_info.c:
merging MyISAM changes into Maria
storage/maria/ma_key.c:
merging MyISAM changes into Maria
storage/maria/ma_loghandler.c:
compiler warning (part->length is size_t)
storage/maria/ma_open.c:
merging MyISAM changes into Maria
storage/maria/ma_preload.c:
merging MyISAM changes into Maria
storage/maria/ma_range.c:
merging MyISAM changes into Maria
storage/maria/ma_rkey.c:
merging MyISAM changes into Maria
storage/maria/ma_rt_index.c:
merging MyISAM changes into Maria
storage/maria/ma_rt_key.c:
merging MyISAM changes into Maria
storage/maria/ma_rt_split.c:
merging MyISAM changes into Maria
storage/maria/ma_search.c:
merging MyISAM changes into Maria
storage/maria/ma_sort.c:
merging MyISAM changes into Maria
storage/maria/maria_def.h:
merging MyISAM changes into Maria
found in pushbuild on sapsrv1. Some not fixed as not repeatable
on my machine (32/64 bit issue?).
Fixes for some test failures:
- "maria-connect" now passes;
- "maria": after fixing the obvious reasons for failures, the test
went further and hit a more complex issues: difference in the output
of EXPLAIN output; not fixed;
- "ps_maria" still crashes in assertion
mysqld: ha_maria.cc:1627: virtual int ha_maria::index_read(uchar*, const uchar*, uint, ha_rkey_function): Ass
ertion `inited == INDEX' failed, as already observable in pushbuild.
All this might just be due to an incomplete merge of MyISAM changes
into Maria when 5.1 was last merged to mysql-maria.
include/my_global.h:
temporary fix until next merge of 5.1; without this it does not build
mysql-test/r/maria-connect.result:
position changed
mysql-test/t/maria-connect.test:
If one wants to use the binlog it has to ask for it.
1582 is not used for dup entry error anymore (it was in older 5.1).
Size of first event in binlog was increased by 4 (when the new type
of event "gap" was added).
mysql-test/t/maria.test:
1582 not used anymore in this case
storage/maria/ha_maria.cc:
engine now has to say what binlogging it supports
storage/maria/ma_blockrec.c:
fix for compiler warnings ("comparison is always true" or "always
false")
storage/maria/ma_loghandler.c:
fix for compiler warnings (comparing char* to uchar*)
storage/maria/ma_packrec.c:
fix for compiler warning (fix simply merged from MyISAM)
storage/maria/ma_pagecache.c:
info_check_pin() was not used so gave a compiler warning.
storage/maria/ma_pagecache.h:
fixing typo from the last 5.1->maria merge.
storage/maria/ma_recovery.c:
my_free() has a void* argument, so why cast. byte->uchar.
storage/maria/ma_search.c:
fix for compiler warning (fix simply merged from MyISAM)
storage/maria/maria_read_log.c:
gptr->uchar*
storage/maria/trnman.c:
probable fix for warning found in pushbuild (but not on my machine):
storage/maria/trnman.c: 142
passing argument 6 of \u2018lf_hash_init\u2019 from incompatible pointer type
on sapsrv1.
sql/handler.cc:
Auto merged
storage/maria/ha_maria.cc:
Auto merged
storage/maria/ma_blockrec.c:
Auto merged
storage/maria/ma_close.c:
Auto merged
storage/maria/ma_create.c:
Auto merged
storage/maria/ma_loghandler.c:
Auto merged
storage/maria/ma_loghandler.h:
Auto merged
storage/maria/ma_test2.c:
Auto merged
storage/maria/maria_read_log.c:
will merge by hand
Unit test for recovery: runs ma_test1 and ma_test2 (both only with
INSERTs and DELETEs; UPDATEs disabled as not handled by recovery)
then moves the tables elswhere; recreates tables from the log, and
compares and fails if there is a difference. Passes now.
Most of maria_read_log.c moved to ma_recovery.c, as it will be re-used
for recovery-from-ha_maria.
Bugfixes of applying of REDO_INSERT, REDO_PURGE_ROW.
Applying of REDO_PURGE_BLOCKS, REDO_DELETE_ALL, REDO_DROP_TABLE,
UNDO_ROW_INSERT (in REDO phase only, i.e. just doing records++),
UNDO_ROW_DELETE, UNDO_ROW_PURGE.
Code cleanups.
Monty: please look for "QQ". Sanja: please look for "Sanja".
Future tasks: recovery of the bitmap (easy), recovery of the state
(make it idempotent), more REDOs (Monty to work on
REDO_UPDATE?), UNDO phase...
Pushing this cset as it looks safe, contains test and bugfixes which
will help Monty implement applying of REDO_UPDATE.
sql/handler.cc:
typo
storage/maria/Makefile.am:
Adding ma_test_recovery (which ma_test_all invokes, and which can
also be run alone). Most of maria_read_log.c moved to ma_recovery.c
storage/maria/ha_maria.cc:
comments
storage/maria/ma_bitmap.c:
fixing comments. 2 -> sizeof(maria_bitmap_marker).
Bitmap-related part of _ma_initialize_datafile() moves in bitmap module.
Now putting the "bm" signature when creating the first bitmap page
(it used to happen only at next open, but that
caused an annoying difference when testing Recovery if the original
run didn't open the table, and it looks more
logical like this: it goes to disk only with its signature correct);
see the "QQ" comment towards the _ma_initialize_data_file() call
in ma_create.c for more).
When reading a bitmap page, verify its signature (happens when normally
using the table or when CHECKing it; not when REPAIRing it).
storage/maria/ma_blockrec.c:
* no need to sync the data file if table is not transactional
* Comments, code cleanup (log-related data moved to log-related code
block, int5store->page_store).
* Store the table's short id into LOGREC_UNDO_ROW_PURGE, like we
do for other records (though this record will soon be replaced
with a CLR).
* If "page" is 1 it means the page which extends from byte
page*block_size+1 to (page+1)*block_size (byte number 1 being
the first byte of the file). The last byte of the file is
data_file_length (same convention).
A new page needs to be created if the last byte of the page is
beyond the last byte of the file, i.e.
(page+1)*block_size+1 > data_file_length, so we correct the test
(bug found when testing log applying for ma_test1 -M -T --skip-update).
* update the page's LSN when removing a row from it during
execution of a REDO_PURGE_ROW record (bug found when testing log
applying for ma_test1 -M -T --skip-update).
* applying of REDO_PURGE_BLOCKs (limited to a one-page range for now).
storage/maria/ma_blockrec.h:
new functions. maria_bitmap_marker does not need to be exported.
storage/maria/ma_close.c:
we can always flush the table's state when closing the last instance
of the table. And it is needed for maria_read_log (as it does
not use maria_lock_database()).
storage/maria/ma_control_file.c:
when in Recovery, some assertions should not be used.
storage/maria/ma_control_file.h:
double-inclusion safe
storage/maria/ma_create.c:
during recovery, don't log records. Comments.
Moving the creation of the first bitmap page to ma_bitmap.c
storage/maria/ma_delete_table.c:
during recovery, don't log records. Log the end-zero of the dropped
table's name, so that recovery can use the string in place without
extending it to fit an end zero.
storage/maria/ma_loghandler.c:
* inwrite_rec_hook also needs access to the MARIA_SHARE, like
prewrite_rec_hook. This will be needed to update
share->records_diff (in the upcoming patch "recovery of the state").
* LOG_DESC::record_ends_group changed to an enum.
* LOG_DESC for LOGREC_REDO_PURGE_BLOCKS and LOGREC_UNDO_ROW_PURGE
corrected
* Sanja please see the @todo LOG BUG
* avoiding DBUG_RETURN(func()) as it gives confusing debug traces.
storage/maria/ma_loghandler.h:
- log write hooks called while the log's lock is held (inwrite_rec_hook)
now need the MARIA_SHARE, like prewrite_rec_hook already had
- instead of a bool saying if this record's type ends groups or not,
we refine: it may not end a group, it may end a group, or it may
be a group in itself. Imagine that we had a physical write failure
to a table before we log the UNDO, we still end up in
external_lock(F_UNLCK) and then we log a COMMIT: we don't want
to consider this COMMIT as ending the group of REDOs (don't want
to execute those REDOs during Recovery), that's why we say "COMMIT
is a group in itself, it aborts any previous group". This also
gives one more sanity check in maria_read_log.
storage/maria/ma_recovery.c:
New Recovery code, replacing the old pseudocode.
Most of maria_read_log moved here.
Call-able from ha_maria, but not enabled yet.
Compared to the previous version of maria_read_log, some bugs have
been fixed, debugging output can go to stdout or a disk file (for now
it's useful for me, later it can be changed), execution of
REDO_DROP_TABLE, REDO_DELETE_ALL, REDO_PURGE_BLOCKS has been added. Duplicate code
has been factored into functions. We abort an unfinished group
of records if we see a record which is a group in itself (like COMMIT).
No need for maria_panic() after a bug (which caused tables to not
be closed) was fixed; if there is yet another bug I prefer to see it.
When opening a table for Recovery, set data_file_length
and key_file_length to their real physical value (these are the
easiest state members to restore :). Warn us if the last page
was truncated (but Recovery handles it).
MARIA_SHARE::state::state::records is now partly recovered (not
idempotent, but works if recreating tables from scracth).
When applying a REDO to a page, stamp it with the UNDO's LSN
(current_group_end_lsn), not with the REDO's LSN; it makes
the table more identical to the original table (easier to compare
the two tables in the end).
Big thing missing: some types of REDOs are not handled,
and the UNDO phase does not exist (missing functions to execute UNDOs
to actually rollback). So for now tests are only inserting/deleting
a few 100 rows, closing the table and seeing if the log is applied ok;
it works. UPDATE not handled.
storage/maria/ma_recovery.h:
new functions: ma_recover() for recovery from inside ha_maria;
_ma_apply_log() for maria_read_log (ma_recover() calls _ma_apply_log()).
Btw, we need to not use the word "recover" for REPAIR/maria_chk anymore.
storage/maria/ma_rename.c:
don't write log records during recovery
storage/maria/ma_test2.c:
- fail if maria_info() or other subtests find some wrong information
- new option -g to skip updates.
- init the translog before creating the table, so that log applying
can work.
- in "#if 0" you'll see some fixed bugs (will be removed).
storage/maria/ma_test_all.sh:
cleanup files. Test log applying.
storage/maria/maria_read_log.c:
most of the logic moves to ma_recovery.c to be shared between
maria_read_log and recovery-from-inside-mysqld.
See ma_recovery.c for additional changes made to the moved code.
storage/maria/ma_test_recovery:
unit test for Recovery. Tests insert and delete,
REDO_UPDATE not yet coded.
Script is called from ma_test_all. Can run standalone.
into labbari.dsl.inet.fi:/home/my/bk/mysql-maria.prod
mysys/thr_lock.c:
Auto merged
storage/csv/ha_tina.cc:
Auto merged
storage/csv/ha_tina.h:
Auto merged
storage/maria/ma_blockrec.c:
Auto merged
storage/maria/ma_check.c:
Auto merged
storage/maria/ma_loghandler.c:
Auto merged
storage/maria/trnman.c:
Auto merged
storage/myisam/mi_locking.c:
Auto merged
sql/lock.cc:
SCCS merged
mode) so ma_test_all works again; more error detection in ma_test_all;
maria_control renamed to maria_log_control (Monty's suggestion,
so that a "rm maria_log*" removes all log-related files). Disabling
failing wrong assertion.
storage/maria/ma_blockrec.c:
disabling assertion which fails because cur_block is a
local variable not initialized
storage/maria/ma_check.c:
comment
storage/maria/ma_control_file.h:
control file renamed
storage/maria/ma_loghandler.c:
assertions needed only in multi-threaded mode (ma_test1 and ma_test2
are single-threaded, it's ok for them to use dummy_transaction_object
with transactional tables: trn->rec_lsn can be set without interfering
with other threads).
storage/maria/ma_test_all.sh:
got caught by failures in some ma_test1 runs, which I didn't see
because ma_test_all returned 0 and I didn't scroll up in the window;
now using "set -e" to avoid that. Also testing that
we get the errors and warnings we expect.
storage/maria/unittest/Makefile.am:
maria_control renamed
into a88-113-38-195.elisa-laajakaista.fi:/home/my/bk/mysql-maria.prod
BitKeeper/etc/ignore:
auto-union
include/maria.h:
Auto merged
include/my_base.h:
Auto merged
mysql-test/include/ps_conv.inc:
Auto merged
mysql-test/r/ps_2myisam.result:
Auto merged
mysql-test/r/ps_3innodb.result:
Auto merged
mysql-test/r/ps_4heap.result:
Auto merged
mysql-test/r/ps_5merge.result:
Auto merged
mysql-test/r/ps_7ndb.result:
Auto merged
mysys/Makefile.am:
Auto merged
mysys/mf_iocache.c:
Auto merged
mysys/mf_tempfile.c:
Auto merged
mysys/my_error.c:
Auto merged
mysys/my_init.c:
Auto merged
mysys/my_open.c:
Auto merged
sql/handler.cc:
Auto merged
sql/handler.h:
Auto merged
sql/lex.h:
Auto merged
sql/sql_show.cc:
Auto merged
sql/sql_yacc.yy:
Auto merged
sql/table.cc:
Auto merged
sql/table.h:
Auto merged
storage/maria/ha_maria.cc:
Auto merged
storage/maria/ha_maria.h:
Auto merged
storage/maria/ma_bitmap.c:
Auto merged
storage/maria/ma_blockrec.h:
Auto merged
storage/maria/ma_create.c:
Auto merged
storage/maria/ma_dynrec.c:
Auto merged
storage/maria/ma_loghandler.h:
Auto merged
storage/maria/ma_loghandler_lsn.h:
Auto merged
storage/maria/ma_open.c:
Auto merged
storage/maria/ma_pagecache.c:
Auto merged
storage/maria/ma_pagecache.h:
Auto merged
storage/maria/ma_test1.c:
Auto merged
storage/maria/ma_test2.c:
Auto merged
storage/maria/ma_update.c:
Auto merged
storage/maria/maria_chk.c:
Auto merged
storage/maria/unittest/ma_test_loghandler-t.c:
Auto merged
storage/maria/unittest/ma_test_loghandler_multigroup-t.c:
Auto merged
storage/maria/unittest/ma_test_loghandler_multithread-t.c:
Auto merged
storage/maria/unittest/ma_test_loghandler_pagecache-t.c:
Auto merged
storage/myisam/ft_stopwords.c:
Auto merged
storage/myisam/mi_close.c:
Auto merged
support-files/compiler_warnings.supp:
Auto merged
mysys/mf_keycache.c:
Pulled from mysql-maria tree, manual merge.
mysys/my_handler.c:
Pulled from mysql-maria tree, manual merge.
mysys/my_symlink2.c:
Pulled from mysql-maria tree, manual merge.
sql/mysqld.cc:
Pulled from mysql-maria tree, manual merge.
sql/sql_table.cc:
Pulled from mysql-maria tree, manual merge.
storage/maria/ma_blockrec.c:
Pulled from mysql-maria tree, manual merge.
storage/maria/ma_check.c:
Pulled from mysql-maria tree, manual merge.
storage/maria/ma_close.c:
Pulled from mysql-maria tree, manual merge.
storage/maria/ma_loghandler.c:
Pulled from mysql-maria tree, manual merge.
storage/maria/maria_def.h:
Pulled from mysql-maria tree, manual merge.
* Don't modify share->base.born_transactional; now it is a value carved
in stone at creation time. share->now_transactional is what can be
modified: it starts at born_transactional, can become false during
ALTER TABLE (when we want no logging), and restored later.
* Not resetting create_rename_lsn to 0 during delete_all or repair.
* when we temporarily disable transactionality, we also change
the page type to PAGECACHE_PLAIN_PAGE: it bypasses some work in the
page cache (optimization), and avoids assertions related to LSNs.
* Disable INSERT DELAYED for transactional tables, because
durability could not be guaranteed (insertion may even not happen)
mysys/mf_keycache.c:
comment
storage/maria/ha_maria.cc:
* a transactional table cannot do INSERT DELAYED
* ha_maria::save_transactional not needed anymore, as now instead
we don't modify MARIA_SHARE::MARIA_BASE_INFO::born_transactional
(born_transactional plays the role of save_transactional), and modify
MARIA_SHARE::now_transactional.
* REPAIR_TABLE log record is now logged by maria_repair()
* comment why we rely on born_transactional to know if we should
skipping a transaction.
* putting together two if()s which test for F_UNLCK
storage/maria/ha_maria.h:
ha_maria::save_transactional not needed anymore (moved to the C layer)
storage/maria/ma_blockrec.c:
* For the block record's code (writing/updating/deleting records),
all that counts is now_transactional, not born_transactional.
* As we now set the page type to PAGECACHE_PLAIN_PAGE for tables
which have now_transactional==FALSE, pagecache will not expect
a meaningful LSN for them in pagecache_unlock_by_link(), so
we can pass it LSN_IMPOSSIBLE.
storage/maria/ma_check.c:
* writing LOGREC_REPAIR_TABLE moves from ha_maria::repair()
to maria_repair(), sounds cleaner (less functions to export).
* when opening a table during REPAIR, don't use the realpath-ed name,
as this may fail if the table has symlinked files (maria_open()
would try to find the data and index file in the directory
of unique_file_name, it would fail if data and index files are in
different dirs); use the unresolved name, open_file_name, which is
the argument which was passed to the maria_open() which created 'info'.
storage/maria/ma_close.c:
assert that when a statement is done with a table, it cleans up
storage/maria/ma_create.c:
new name
storage/maria/ma_delete_all.c:
* using now_transactional
* no reason to reset create_rename_lsn during delete_all (a bug);
also no reason to do it during repair: it was put there because
a positive create_rename_lsn caused a call to check_and_set_lsn()
which asserted in DBUG_ASSERT(block->type == PAGECACHE_LSN_PAGE);
first solution was to use LSN_IMPOSSIBLE in _ma_unpin_all_pages() if
not transactional; but then in the case of ALTER TABLE, with
transactionality temporarily disabled, it asserted in
DBUG_ASSERT(LSN_VALID(lsn)) in pagecache_fwrite() (PAGECACHE_LSN_PAGE
page with zero LSN - bad). The additional solution is to use
PAGECACHE_PLAIN_PAGE when we disable transactionality temporarily: this
avoids checks on the LSN, and also bypasses (optimization) the "flush
log up to LSN" call when the pagecache flushes our page (in other
words, no WAL needed).
storage/maria/ma_delete_table.c:
use now_transactional
storage/maria/ma_locking.c:
assert that when a statement is done with a table, it cleans up.
storage/maria/ma_loghandler.c:
* now_transactional should be used to test if we want a log record.
* Assertions to make sure dummy_transaction_object is not spoilt
by its many users.
storage/maria/ma_open.c:
base.transactional -> base.born_transactional
storage/maria/ma_pagecache.c:
missing name for page's type. Comment for future.
storage/maria/ma_rename.c:
use now_transactional
storage/maria/maria_chk.c:
use born_transactional
storage/maria/maria_def.h:
MARIA_BASE_INFO::transactional renamed to born_transactional.
MARIA_SHARE::now_transactional introduced.
_ma_repair_write_log_record() is made local to ma_check.c.
Macros to temporarily disable, and re-enable, transactionality for a
table.
storage/maria/maria_read_log.c:
assertions and using the new macros. Adding a forgotten resetting
when we finally close all tables.