mariadb/storage/innobase/include/fsp0fsp.h

364 lines
14 KiB
C
Raw Normal View History

/*****************************************************************************
2011-08-29 11:22:43 +03:00
Copyright (c) 1995, 2011, Oracle and/or its affiliates. All Rights Reserved.
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 59 Temple
Place, Suite 330, Boston, MA 02111-1307 USA
*****************************************************************************/
/**************************************************//**
@file include/fsp0fsp.h
File space management
Created 12/18/1995 Heikki Tuuri
*******************************************************/
#ifndef fsp0fsp_h
#define fsp0fsp_h
#include "univ.i"
#include "mtr0mtr.h"
#include "fut0lst.h"
#include "ut0byte.h"
#include "page0types.h"
2009-07-30 17:42:56 +05:00
#include "fsp0types.h"
/**********************************************************************//**
Initializes the file space system. */
UNIV_INTERN
void
fsp_init(void);
/*==========*/
/**********************************************************************//**
Gets the current free limit of the system tablespace. The free limit
means the place of the first page which has never been put to the
free list for allocation. The space above that address is initialized
to zero. Sets also the global variable log_fsp_current_free_limit.
@return free limit in megabytes */
UNIV_INTERN
ulint
fsp_header_get_free_limit(void);
/*===========================*/
/**********************************************************************//**
Gets the size of the system tablespace from the tablespace header. If
we do not have an auto-extending data file, this should be equal to
the size of the data files. If there is an auto-extending data file,
this can be smaller.
@return size in pages */
UNIV_INTERN
ulint
fsp_header_get_tablespace_size(void);
/*================================*/
/**********************************************************************//**
Reads the file space size stored in the header page.
@return tablespace size stored in the space header */
UNIV_INTERN
ulint
fsp_get_size_low(
/*=============*/
page_t* page); /*!< in: header page (page 0 in the tablespace) */
/**********************************************************************//**
Reads the space id from the first page of a tablespace.
@return space id, ULINT UNDEFINED if error */
UNIV_INTERN
ulint
fsp_header_get_space_id(
/*====================*/
const page_t* page); /*!< in: first page of a tablespace */
/**********************************************************************//**
Reads the space flags from the first page of a tablespace.
@return flags */
UNIV_INTERN
ulint
fsp_header_get_flags(
/*=================*/
const page_t* page); /*!< in: first page of a tablespace */
/**********************************************************************//**
Reads the compressed page size from the first page of a tablespace.
@return compressed page size in bytes, or 0 if uncompressed */
UNIV_INTERN
ulint
fsp_header_get_zip_size(
/*====================*/
const page_t* page); /*!< in: first page of a tablespace */
/**********************************************************************//**
Writes the space id and compressed page size to a tablespace header.
This function is used past the buffer pool when we in fil0fil.c create
a new single-table tablespace. */
UNIV_INTERN
void
fsp_header_init_fields(
/*===================*/
page_t* page, /*!< in/out: first page in the space */
ulint space_id, /*!< in: space id */
ulint flags); /*!< in: tablespace flags (FSP_SPACE_FLAGS):
0, or table->flags if newer than COMPACT */
/**********************************************************************//**
Initializes the space header of a new created space and creates also the
insert buffer tree root if space == 0. */
UNIV_INTERN
void
fsp_header_init(
/*============*/
ulint space, /*!< in: space id */
ulint size, /*!< in: current size in blocks */
mtr_t* mtr); /*!< in: mini-transaction handle */
/**********************************************************************//**
Increases the space size field of a space. */
UNIV_INTERN
void
fsp_header_inc_size(
/*================*/
ulint space, /*!< in: space id */
ulint size_inc,/*!< in: size increment in pages */
mtr_t* mtr); /*!< in: mini-transaction handle */
/**********************************************************************//**
Creates a new segment.
@return the block where the segment header is placed, x-latched, NULL
if could not create segment because of lack of space */
UNIV_INTERN
buf_block_t*
fseg_create(
/*========*/
ulint space, /*!< in: space id */
ulint page, /*!< in: page where the segment header is placed: if
this is != 0, the page must belong to another segment,
if this is 0, a new page will be allocated and it
will belong to the created segment */
ulint byte_offset, /*!< in: byte offset of the created segment header
on the page */
mtr_t* mtr); /*!< in: mtr */
/**********************************************************************//**
Creates a new segment.
@return the block where the segment header is placed, x-latched, NULL
if could not create segment because of lack of space */
UNIV_INTERN
buf_block_t*
fseg_create_general(
/*================*/
ulint space, /*!< in: space id */
ulint page, /*!< in: page where the segment header is placed: if
this is != 0, the page must belong to another segment,
if this is 0, a new page will be allocated and it
will belong to the created segment */
ulint byte_offset, /*!< in: byte offset of the created segment header
on the page */
ibool has_done_reservation, /*!< in: TRUE if the caller has already
done the reservation for the pages with
fsp_reserve_free_extents (at least 2 extents: one for
the inode and the other for the segment) then there is
no need to do the check for this individual
operation */
mtr_t* mtr); /*!< in: mtr */
/**********************************************************************//**
Calculates the number of pages reserved by a segment, and how many pages are
currently used.
@return number of reserved pages */
UNIV_INTERN
ulint
fseg_n_reserved_pages(
/*==================*/
fseg_header_t* header, /*!< in: segment header */
ulint* used, /*!< out: number of pages used (<= reserved) */
mtr_t* mtr); /*!< in: mtr handle */
/**********************************************************************//**
Allocates a single free page from a segment. This function implements
the intelligent allocation strategy which tries to minimize
file space fragmentation.
2011-08-29 11:22:43 +03:00
@param[in/out] seg_header segment header
@param[in] hint hint of which page would be desirable
@param[in] direction if the new page is needed because
of an index page split, and records are
inserted there in order, into which
direction they go alphabetically: FSP_DOWN,
2011-08-29 11:22:43 +03:00
FSP_UP, FSP_NO_DIR
@param[in/out] mtr mini-transaction
@return the allocated page offset FIL_NULL if no page could be allocated */
#define fseg_alloc_free_page(seg_header, hint, direction, mtr) \
fseg_alloc_free_page_general(seg_header, hint, direction, \
FALSE, mtr, mtr)
/**********************************************************************//**
Allocates a single free page from a segment. This function implements
the intelligent allocation strategy which tries to minimize file space
fragmentation.
@return allocated page offset, FIL_NULL if no page could be allocated */
UNIV_INTERN
ulint
fseg_alloc_free_page_general(
/*=========================*/
2011-08-29 11:22:43 +03:00
fseg_header_t* seg_header,/*!< in/out: segment header */
ulint hint, /*!< in: hint of which page would be desirable */
byte direction,/*!< in: if the new page is needed because
of an index page split, and records are
inserted there in order, into which
direction they go alphabetically: FSP_DOWN,
FSP_UP, FSP_NO_DIR */
ibool has_done_reservation, /*!< in: TRUE if the caller has
already done the reservation for the page
with fsp_reserve_free_extents, then there
is no need to do the check for this individual
page */
2011-08-29 11:22:43 +03:00
mtr_t* mtr, /*!< in/out: mini-transaction */
mtr_t* init_mtr)/*!< in/out: mtr or another mini-transaction
Bug#12704861 Corruption after a crash during BLOB update The fix of Bug#12612184 broke crash recovery. When a record that contains off-page columns (BLOBs) is updated, we must first write redo log about the BLOB page writes, and only after that write the redo log about the B-tree changes. The buggy fix would log the B-tree changes first, meaning that after recovery, we could end up having a record that contains a null BLOB pointer. Because we will be redo logging the writes off the off-page columns before the B-tree changes, we must make sure that the pages chosen for the off-page columns are free both before and after the B-tree changes. In this way, the worst thing that can happen in crash recovery is that the BLOBs are written to free pages, but the B-tree changes are not applied. The BLOB pages would correctly remain free in this case. To achieve this, we must allocate the BLOB pages in the mini-transaction of the B-tree operation. A further quirk is that BLOB pages are allocated from the same file segment as leaf pages. Because of this, we must temporarily "hide" any leaf pages that were freed during the B-tree operation by "fake allocating" them prior to writing the BLOBs, and freeing them again before the mtr_commit() of the B-tree operation, in btr_mark_freed_leaves(). btr_cur_mtr_commit_and_start(): Remove this faulty function that was introduced in the Bug#12612184 fix. The problem that this function was trying to address was that when we did mtr_commit() the BLOB writes before the mtr_commit() of the update, the new BLOB pages could have overwritten clustered index B-tree leaf pages that were freed during the update. If recovery applied the redo log of the BLOB writes but did not see the log of the record update, the index tree would be corrupted. The correct solution is to make the freed clustered index pages unavailable to the BLOB allocation. This function is also a likely culprit of InnoDB hangs that were observed when testing the Bug#12612184 fix. btr_mark_freed_leaves(): Mark all freed clustered index leaf pages of a mini-transaction allocated (nonfree=TRUE) before storing the BLOBs, or freed (nonfree=FALSE) before committing the mini-transaction. btr_freed_leaves_validate(): A debug function for checking that all clustered index leaf pages that have been marked free in the mini-transaction are consistent (have not been zeroed out). btr_page_alloc_low(): Refactored from btr_page_alloc(). Return the number of the allocated page, or FIL_NULL if out of space. Add the parameter "mtr_t* init_mtr" for specifying the mini-transaction where the page should be initialized, or if this is a "fake allocation" (init_mtr=NULL) by btr_mark_freed_leaves(nonfree=TRUE). btr_page_alloc(): Add the parameter init_mtr, allowing the page to be initialized and X-latched in a different mini-transaction than the one that is used for the allocation. Invoke btr_page_alloc_low(). If a clustered index leaf page was previously freed in mtr, remove it from the memo of previously freed pages. btr_page_free(): Assert that the page is a B-tree page and it has been X-latched by the mini-transaction. If the freed page was a leaf page of a clustered index, link it by a MTR_MEMO_FREE_CLUST_LEAF marker to the mini-transaction. btr_store_big_rec_extern_fields_func(): Add the parameter alloc_mtr, which is NULL (old behaviour in inserts) and the same as local_mtr in updates. If alloc_mtr!=NULL, the BLOB pages will be allocated from it instead of the mini-transaction that is used for writing the BLOBs. fsp_alloc_from_free_frag(): Refactored from fsp_alloc_free_page(). Allocate the specified page from a partially free extent. fseg_alloc_free_page_low(), fseg_alloc_free_page_general(): Add the parameter "mtr_t* init_mtr" for specifying the mini-transaction where the page should be initialized, or NULL if this is a "fake allocation" that prevents the reuse of a previously freed B-tree page for BLOB storage. If init_mtr==NULL, try harder to reallocate the specified page and assert that it succeeded. fsp_alloc_free_page(): Add the parameter "mtr_t* init_mtr" for specifying the mini-transaction where the page should be initialized. Do not allow init_mtr == NULL, because this function is never to be used for "fake allocations". mtr_t: Add the operation MTR_MEMO_FREE_CLUST_LEAF and the flag mtr->freed_clust_leaf for quickly determining if any MTR_MEMO_FREE_CLUST_LEAF operations have been posted. row_ins_index_entry_low(): When columns are being made off-page in insert-by-update, invoke btr_mark_freed_leaves(nonfree=TRUE) and pass the mini-transaction as the alloc_mtr to btr_store_big_rec_extern_fields(). Finally, invoke btr_mark_freed_leaves(nonfree=FALSE) to avoid leaking pages. row_build(): Correct a comment, and add a debug assertion that a record that contains NULL BLOB pointers must be a fresh insert. row_upd_clust_rec(): When columns are being moved off-page, invoke btr_mark_freed_leaves(nonfree=TRUE) and pass the mini-transaction as the alloc_mtr to btr_store_big_rec_extern_fields(). Finally, invoke btr_mark_freed_leaves(nonfree=FALSE) to avoid leaking pages. buf_reset_check_index_page_at_flush(): Remove. The function fsp_init_file_page_low() already sets bpage->check_index_page_at_flush=FALSE. There is a known issue in tablespace extension. If the request to allocate a BLOB page leads to the tablespace being extended, crash recovery could see BLOB writes to pages that are off the tablespace file bounds. This should trigger an assertion failure in fil_io() at crash recovery. The safe thing would be to write redo log about the tablespace extension to the mini-transaction of the BLOB write, not to the mini-transaction of the record update. However, there is no redo log record for file extension in the current redo log format. rb:693 approved by Sunny Bains
2011-08-29 11:16:42 +03:00
in which the page should be initialized,
or NULL if this is a "fake allocation" of
a page that was previously freed in mtr */
2011-08-29 11:22:43 +03:00
__attribute__((warn_unused_result, nonnull(1,5)));
/**********************************************************************//**
Reserves free pages from a tablespace. All mini-transactions which may
use several pages from the tablespace should call this function beforehand
and reserve enough free extents so that they certainly will be able
to do their operation, like a B-tree page split, fully. Reservations
must be released with function fil_space_release_free_extents!
The alloc_type below has the following meaning: FSP_NORMAL means an
operation which will probably result in more space usage, like an
insert in a B-tree; FSP_UNDO means allocation to undo logs: if we are
deleting rows, then this allocation will in the long run result in
less space usage (after a purge); FSP_CLEANING means allocation done
in a physical record delete (like in a purge) or other cleaning operation
which will result in less space usage in the long run. We prefer the latter
two types of allocation: when space is scarce, FSP_NORMAL allocations
will not succeed, but the latter two allocations will succeed, if possible.
The purpose is to avoid dead end where the database is full but the
user cannot free any space because these freeing operations temporarily
reserve some space.
Single-table tablespaces whose size is < 32 pages are a special case. In this
function we would liberally reserve several 64 page extents for every page
split or merge in a B-tree. But we do not want to waste disk space if the table
only occupies < 32 pages. That is why we apply different rules in that special
case, just ensuring that there are 3 free pages available.
@return TRUE if we were able to make the reservation */
UNIV_INTERN
ibool
fsp_reserve_free_extents(
/*=====================*/
ulint* n_reserved,/*!< out: number of extents actually reserved; if we
return TRUE and the tablespace size is < 64 pages,
then this can be 0, otherwise it is n_ext */
ulint space, /*!< in: space id */
ulint n_ext, /*!< in: number of extents to reserve */
ulint alloc_type,/*!< in: FSP_NORMAL, FSP_UNDO, or FSP_CLEANING */
mtr_t* mtr); /*!< in: mtr */
/**********************************************************************//**
This function should be used to get information on how much we still
will be able to insert new data to the database without running out the
tablespace. Only free extents are taken into account and we also subtract
the safety margin required by the above function fsp_reserve_free_extents.
@return available space in kB */
UNIV_INTERN
ullint
fsp_get_available_space_in_free_extents(
/*====================================*/
ulint space); /*!< in: space id */
/**********************************************************************//**
Frees a single page of a segment. */
UNIV_INTERN
void
fseg_free_page(
/*===========*/
fseg_header_t* seg_header, /*!< in: segment header */
ulint space, /*!< in: space id */
ulint page, /*!< in: page offset */
mtr_t* mtr); /*!< in: mtr handle */
/**********************************************************************//**
Frees part of a segment. This function can be used to free a segment
by repeatedly calling this function in different mini-transactions.
Doing the freeing in a single mini-transaction might result in
too big a mini-transaction.
@return TRUE if freeing completed */
UNIV_INTERN
ibool
fseg_free_step(
/*===========*/
fseg_header_t* header, /*!< in, own: segment header; NOTE: if the header
resides on the first page of the frag list
of the segment, this pointer becomes obsolete
after the last freeing step */
mtr_t* mtr); /*!< in: mtr */
/**********************************************************************//**
Frees part of a segment. Differs from fseg_free_step because this function
leaves the header page unfreed.
@return TRUE if freeing completed, except the header page */
UNIV_INTERN
ibool
fseg_free_step_not_header(
/*======================*/
fseg_header_t* header, /*!< in: segment header which must reside on
the first fragment page of the segment */
mtr_t* mtr); /*!< in: mtr */
/***********************************************************************//**
Checks if a page address is an extent descriptor page address.
@return TRUE if a descriptor page */
UNIV_INLINE
ibool
fsp_descr_page(
/*===========*/
ulint zip_size,/*!< in: compressed page size in bytes;
0 for uncompressed pages */
ulint page_no);/*!< in: page number */
/***********************************************************//**
Parses a redo log record of a file page init.
@return end of log record or NULL */
UNIV_INTERN
byte*
fsp_parse_init_file_page(
/*=====================*/
byte* ptr, /*!< in: buffer */
byte* end_ptr, /*!< in: buffer end */
buf_block_t* block); /*!< in: block or NULL */
/*******************************************************************//**
Validates the file space system and its segments.
@return TRUE if ok */
UNIV_INTERN
ibool
fsp_validate(
/*=========*/
ulint space); /*!< in: space id */
/*******************************************************************//**
Prints info of a file space. */
UNIV_INTERN
void
fsp_print(
/*======*/
ulint space); /*!< in: space id */
2009-07-30 17:42:56 +05:00
#ifdef UNIV_DEBUG
/*******************************************************************//**
Validates a segment.
@return TRUE if ok */
UNIV_INTERN
ibool
fseg_validate(
/*==========*/
fseg_header_t* header, /*!< in: segment header */
2009-07-30 17:42:56 +05:00
mtr_t* mtr); /*!< in: mtr */
#endif /* UNIV_DEBUG */
#ifdef UNIV_BTR_PRINT
/*******************************************************************//**
Writes info of a segment. */
UNIV_INTERN
void
fseg_print(
/*=======*/
fseg_header_t* header, /*!< in: segment header */
mtr_t* mtr); /*!< in: mtr */
#endif /* UNIV_BTR_PRINT */
#ifndef UNIV_NONINL
#include "fsp0fsp.ic"
#endif
#endif