Move more definitions from row0mysql.h to row0mysql.c. Remove the
unused definition of merge_thread.
merge_index_def_t: Replace merge_index_field_t** with merge_index_field_t*.
Use mem_heap_strdup() when copying strings.
ha_innobase::add_index(): Avoid excessive use of current_thd.
assignments. A memory leak remains at "goto error_handling": the memory
allocated with row_merge_block_create() in the beginning of this function
will not be freed.
Define the merge_rec_ structures in row0merge.c instead of
row0merge.h. They are not used in the interface defined in row0merge.h.
row_merge_create_file_structure(): Rename to row_merge_file_create().
Input a pointer to a pre-allocated block of memory.
row_merge_read_clustered_index(): Replace merge_file_t** with merge_file_t*.
row_merge_sort_and_insert_thread(): Remove the declaration.
This function was never defined or called.
row_merge_is_index_usable(): Correct the formatting.
row_build_index_for_mysql(): Allocate an array of merge_file_t instead
of an array of pointers to merge_file_t.
ha_innodb.cc: Add the columns COMPRESSED, COMPRESSED_OK, DECOMPRESSED
to INFORMATION_SCHEMA.INNODB_BUDDY.
page_zip_compress_count[], page_zip_compress_ok[]: New statistic counters,
incremented in page_zip_compress().
page_zip_decompress_count[]: New statistic counter,
incremented in page_zip_decompress().
innodb.test: The error code 1540 has been renumbered to 1542.
innodb.result: Adjust the expected value of Innodb_buffer_pool_pages_total.
Something may have changed in the BLOB handling in branches/zip.
ha_innodb.cc: Add ha_innobase:: prefix to some DBUG_ENTER macros.
row_merge_create_temporary_table(): Free the heap only after invoking
row_create_table_for_mysql(). The table->col_names points to the heap
until the system columns are added by row_create_table_for_mysql().
TODO: use the filename-safe encoding in temporary table names. Test
extensively under Valgrind and fix all errors found. Eliminate
merge_rec_t and rewrite row0merge.c to write the records directly to
the merge blocks.
dict_truncate_index_tree(): Add the parameter space for specifying the
new tablespace identifier of a single-table tablespace that has been
recreated. When SYS_INDEXES.PAGE_NO == FIL_NULL, do not abort but
create the index tree.
fil_discard_tablespace(): Pass the return code from fil_delete_tablespace().
row_truncate_table_for_mysql(): Attempt to discard and recreate
single-table tablespaces. Reassign the tablespace identifier both in
the data dictionary (SYS_TABLES and SYS_INDEXES) and in the data
dictionary cache.
trx_undo_left(): Add const qualifiers.
trx_undo_page_report_insert(): Use exact trx_undo_left() limit.
Remove a duplicated trx_undo_left() check.
trx_undo_page_report_modify(): Eliminate the local variable len.
Document that no prefix for BLOBs needs to be stored in the undo log.
Lump two trx_undo_left() checks together.
Previously, when big_rec was returned, the fields would point to
freed memory. The memory heap was allocated locally, and the data tuple
was allocated from the heap, and the big_rec would point to some fields
in the data tuple.
row_ins_clust_index_entry_by_modify(): Add parameter heap,
for the same reason.
allocated before overwriting some fields and flagging the frame freed.
Some of the buffer frame may have been flagged freed by the user
(such as mem_heap).
functions.
ut_malloc_low(): Flag the block with UNIV_MEM_ALLOC(). Do not flag the
block with UNIV_MEM_FREE() in ut_free(), because it would cause bogus
Valgrind warnings in the underlying memory allocator.
mem_pool_create(): Flag the data area with UNIV_MEM_FREE().
mem_pool_fill_free_list(): Flag the area header with UNIV_MEM_ALLOC().
mem_area_alloc(): Flag the data area with UNIV_MEM_ALLOC().
mem_area_free(): Flag the data area with UNIV_MEM_FREE().
mem_heap_alloc(): Flag the buffer with UNIV_MEM_ALLOC().
mem_heap_block_free(): Flag the block with UNIV_MEM_FREE().
mem_heap_free_top(): Flag the block with UNIV_MEM_FREE().
btr_rec_free_updated_extern_fields(): Remove the parameter
do_not_free_inherited, which is always passed as TRUE.
btr_cur_unmark_extern_fields(): Invoke rec_offs_any_extern().
btr_cur_unmark_dtuple_extern_fields(): Remove a local variable.
for displaying buddy allocator statistics.
buf_buddy_used[], buf_buddy_relocated[]: Declare as global symbols.
mysql_declare_plugin(innobase): Add MYSQL_INFORMATION_SCHEMA_PLUGIN
for innobase_is_buddy ("INNODB_BUDDY").
pass max_ins_size=UNIV_PAGE_SIZE and increase=ULINT_UNDEFINED to
ibuf_update_free_bits_if_full(). Otherwise, a bogus assertion failure
may occur when the page is recompressed and there is more space available
in the modification log after inserting the record.
compressed secondary index pages after attempting the insert.
btr_cur_update_alloc_zip(): Reset the insert buffer free bits if there is
insufficient space available on secondary index pages. Do not update
the bits after successful operation.
btr_cur_update_in_place(): Update the free bits in the index buffer.
btr_cur_optimistic_update(): First check rec_offs_any_extern(), because
checking it is cheaper than checking the "external storage" flags of the
update vector. Update the insert buffer free bits in any case after
successfully calling btr_cur_update_alloc_zip().
failing insert. Reorganization will have been attempted in
page_cur_tuple_insert() or page_cur_rec_insert().
page_zip_reorganize(): Recompute the insert buffer free bits for
leaf pages of secondary indexes.
ibuf_data_enough_free_for_insert(): Simplify.
were introduced when reducing the memory footprint of the data dictionary
cache (Bug #20877), around r834. This function is only compiled if
PAGE_CUR_LE_OR_EXTENDS is defined in page0cur.h.
compressed tablespaces.
btr_page_reorganize_low(): Recompute the insert buffer free bits
for compressed pages belonging to a non-clustered index.
btr_cur_optimistic_insert(): Determine if the page was reorganized by
page_cur_tuple_insert(). Do not attempt to reorganize compressed pages,
because that will be attempted by page_cur_tuple_insert().
page_zip_max_ins_size(): New function.
btr_cur_optimistic_insert(), btr_cur_optimistic_delete(),
btr_page_split_and_insert(), btr_compress(): Do not update the
ibuf free bits for non-leaf pages or pages belonging to a clustered index.
The insert buffer only covers operations on leaf pages of secondary indexes.
For pages covered by the insert buffer, limit the max_ins_size to
page_zip_max_ins_size().
buf_page_get_gen(): Merge the insert buffer after decompressing the page.
buf_page_io_complete(): Relax the assertion about ibuf_count. For
compressed-only pages, the insert buffer merge takes place
in buf_page_get_gen().
ibuf_index_page_calc_free_bits(), ibuf_index_page_calc_free_from_bits(),
ibuf_index_page_calc_free(), ibuf_update_free_bits_if_full(),
ibuf_update_free_bits_low(), ibuf_update_free_bits_for_two_pages_low(),
ibuf_set_free_bits_low(): Add the parameter zip_size. Limit the maximum
insert size to page_zip_max_ins_size().
compressed pages, because doing so would require a call to
buf_buddy_alloc(), which in turn can invoke buf_LRU_get_free_block()
and buf_LRU_search_and_free_block(), potentially infinite recursion.
buf_buddy_n_frames: Number of buffer frames allocated for the buddy system
buf_buddy_used[]: Counts of allocated blocks, by buddy size.
buf_buddy_relocated[]: Counts of blocks relocated, bu buddy size.
buf_buddy_alloc_clean(): Note the added configuration parameter
buf_buddy_min_n_frames: Preferred minimum buf_buddy_n_frames
buf_buddy_alloc() until after the block has been found not to exist
in the buffer pool.
buf_buddy_alloc(), buf_buddy_alloc_low(): Make lru an output parameter.
Improve documentation.
buf_page_init_for_read(): Simplify the code. Check buf_page_hash_get()
again if buf_buddy_alloc() released buf_pool->mutex.
frame when reading compressed pages, unless crash recovery is in progress.
buf_page_read_low(): Adapt for buf_page_init_for_read() returning
buf_page_t* instead of buf_block_t*.
but in buf_page_get_gen(). Adjust comments. This prevents
unnecessary decompression in read-ahead when the compressed block
exists in the buffer pool.
If the block does not exist in the buffer pool, read-ahead will still
allocate an uncompressed page and decompress the block.
Move buf_block_init_low() and buf_zip_decompress() earlier in the file,
because some compilers are unable to inline functions that are defined
after the invocation.
buf_page_get_gen(): Decompress the block if needed.
buf_page_init_for_read(): Do not decompress.
record does not contain externally stored columns.
page_zip_decompress_clust_ext(): New function for decompressing records
containing externally stored columns.
before testing create_info->avg_row_length, because otherwise the table
would be converted to uncompressed format by CREATE INDEX and possibly
other statements.
externally stored columns.
page_zip_compress_clust_ext(), page_zip_apply_log_ext():
New functions.
page_zip_compress_clust(), page_zip_apply_log(): Check rec_offs_any_extern()
and avoid invoking the costly loop in most cases.
Similar optimizations can be made in page_zip_decompress_clust() and
page_zip_write_rec().
by adding the REC_OFFS_EXTERNAL flag to rec_offs_base(offsets)[0].
This reduces the processor usage of page_zip_write_rec() by about 40%
in one test case. The code could be sped up further by testing
rec_offs_any_extern() outside of loops that check rec_offs_nth_extern().
The vast majority of records does not contain any externally stored columns.
