empty result when using DESC
Problem: fetching MyISAM keys we copy a key block pointer to the end of the key buffer.
However, we don't take into account the pointer length calculatig the buffer size,
that may leads to memory overwriting and in turn to unpredictable results.
Fix: increase key buffer size by length of the key block pointer.
Note: no simple test case.
at page 1024 with ucs2_bin
Inserting strings with a common prefix into a table with
characterset UCS2 corrupted the table.
An efficient search method was used, which compares end space
with ASCII blank. This doesn't work for character sets like UCS2,
which do not encode blank like ASCII does.
Use the less efficient search method _mi_seq_search()
for charsets with mbminlen > 1.
SPATIAL key is fine actually, but the chk_key() function
mistakenly returns error. It tries to compare checksums
of btree and SPATIAL keys while the checksum for the SPATIAL isn't
calculated (always 0). Same thing with FULLTEXT keys is handled
using full_text_keys counter, so fixed by counting both
SPATIAL and FULLTEXT keys in that counter.
When we insert a record into MYISAM table which is almost 'full',
we first write record data in the free space inside a file, and then
check if we have enough space after the end of the file.
So if we don't have the space, table will left corrupted.
Similar error also happens when we updata MYISAM tables.
Fixed by modifying write_dynamic_record and update_dynamic_record functions
to check for free space before writing parts of a record
With certain data sets (when compressed record length gets bigger than
uncompressed) myisamchk --unpack may corrupt data file.
Fixed that record length was wrongly restored from compressed table.
Disabling and enabling indexes on a non-empty table grows the
index file.
Disabling indexes just sets a flag per non-unique index and does not
free the index blocks of the affected indexes. Re-enabling indexes
creates new indexes with new blocks. The old blocks remain unused
in the index file.
Fixed by dropping and re-creating all indexes if non-empty disabled
indexes exist when enabling indexes. Dropping all indexes resets
the internal end-of-file marker to the end of the index file header.
It also clears the root block pointers of every index and clears the
deleted blocks chains. This way all blocks are declared as free.
fulltext index
Having a table with broken multibyte characters may cause fulltext
parser dead-loop.
Since normally it is not possible to insert broken multibyte sequence
into a table, this problem may arise only if table is damaged.
Affected statements are:
- CHECK/REPAIR against damaged table with fulltext index;
- boolean mode phrase search against damaged table with or
without fulltext inex;
- boolean mode searches without index;
- nlq searches.
No test case for this fix. Affects 5.0 only.
Fulltext boolean mode phrase search may crash server on platforms
where size of pointer is not equal to size of unsigned integer
(in other words some 64-bit platforms).
The problem was integer overflow.
Affects 4.1 only.
CPUs / Intel's ICC compile
The bug is a combination of two problems:
1. IA64/ICC MySQL binaries use glibc's qsort(), not the one in mysys.
2. The order relation implemented by join_tab_cmp() is not transitive,
i.e. it is possible to choose such a, b and c that (a < b) && (b < c)
but (c < a). This implies that result of a sort using the relation
implemented by join_tab_cmp() depends on the order in which
elements are compared, i.e. the result is implementation-specific. Since
choose_plan() uses qsort() to pre-sort the
join tables using join_tab_cmp() as a compare function, the results of
the sorting may vary depending on qsort() implementation.
It is neither possible nor important to implement a better ordering
algorithm in join_tab_cmp(). Therefore the only way to fix it is to
force our own qsort() to be used by renaming it to my_qsort(), so we don't depend
on linker to decide that.
This patch also "fixes" bug #20530: qsort redefinition violates the
standard.
myisam_sort_buffer_size.
An incorrect length of the sort buffer was used when calculating the
maximum number of keys. When myisam_sort_buffer_size is small enough,
this could result in the number of keys < number of
BUFFPEK structures which in turn led to use of uninitialized BUFFPEKs.
Fixed by correcting the buffer length calculation.
As the result of DOUBLE claculations can be bigger
than DBL_MAX constant we use in code, we shouldn't use this constatn
as a biggest possible value.
Particularly the rtree_pick_key function set 'min_area= DBL_MAX' relying
that any rtree_area_increase result will be less so we return valid
key. Though in rtree_area_increase function we calculate the area
of the rectangle, so the result can be 'inf' if the rectangle is
huge enough, which is bigger than DBL_MAX.
Code of the rtree_pick_key modified so we always return a valid key.
When using concurrent insert with parallel index reads, it could
happen that reading sessions found keys that pointed to records
yet to be written to the data file. The result was a report of
a corrupted table. But it was false alert.
When inserting a record in a table with indexes, the keys are
inserted into the indexes before the record is written to the data
file. When the insert happens concurrently to selects, an
index read can find a key that references the record that is not
yet written to the data file. To avoid any access to such record,
the select saves the current end of file position when it starts.
Since concurrent inserts are always appended at end of the data
file, the select can easily ignore any concurrently inserted record.
The problem was that the ignore was only done for non-exact key
searches (partial key or using >, >=, < or <=).
The fix is to ignore concurrently inserted records also for
exact key searches.
No test case. Concurrent inserts cannot be tested with the test
suite. Test cases are attached to the bug report.
