mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-01-23 23:34:34 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions
mariadb/storage/innobase/include/buf0buf.ic
aivanov@mysql.com d33b523271 Applied innodb-5.1-ss269 snapshot. 
Fixed BUGS:
  #3300: "UPDATE statement with no index column in where condition locks
    all rows"
    Implement semi-consistent read to reduce lock conflicts at the cost
    of breaking serializability.
    ha_innobase::unlock_row(): reset the "did semi consistent read" flag
    ha_innobase::was_semi_consistent_read(),
    ha_innobase::try_semi_consistent_read(): new methods
    row_prebuilt_t, row_create_prebuilt(): add field row_read_type for
    keeping track of semi-consistent reads
    row_vers_build_for_semi_consistent_read(),
    row_sel_build_committed_vers_for_mysql(): new functions
    row_search_for_mysql(): implement semi-consistent reads

  #9802: "Foreign key checks disallow alter table".
    Added test cases.

  #12456: "Cursor shows incorrect data - DML does not affect,
    probably caching"
    This patch implements a high-granularity read view to be used with
    cursors. In this high-granularity consistent read view modifications 
    done by the creating transaction after the cursor is created or 
   future transactions are not visible. But those modifications that 
   transaction did before the cursor was created are visible.

  #12701: "Support >4GB buffer pool and log files on 64-bit Windows"
    Do not call os_file_create_tmpfile() at runtime. Instead, create all
    tempfiles at startup and guard access to them with mutexes.

  #13778: "If FOREIGN_KEY_CHECKS=0, one can create inconsistent FOREIGN KEYs".
    When FOREIGN_KEY_CHECKS=0 we still need to check that datatypes between
    foreign key references are compatible.

  #14189: "VARBINARY and BINARY variables: trailing space ignored with InnoDB"
    innobase_init(): Assert that
    DATA_MYSQL_BINARY_CHARSET_COLL == my_charset_bin.number.
    dtype_get_pad_char(): Do not pad VARBINARY or BINARY columns.
    row_ins_cascade_calc_update_vec(): Refuse ON UPDATE CASCADE when trying
    to change the length of a VARBINARY column that refers to or is referenced
    by a BINARY column. BINARY columns are no longer padded on comparison,
    and thus they cannot be padded on storage either.

  #14747: "Race condition can cause btr_search_drop_page_hash_index() to crash"
    Note that buf_block_t::index should be protected by btr_search_latch
    or an s-latch or x-latch on the index page.
    btr_search_drop_page_hash_index(): Read block->index while holding
    btr_search_latch and use the cached value in the loop.  Remove some
    redundant assertions.

  #15108: "mysqld crashes when innodb_log_file_size is set > 4G"

  #15308: "Problem of Order with Enum Column in Primary Key"

  #15550: "mysqld crashes in printing a FOREIGN KEY error in InnoDB"
    row_ins_foreign_report_add_err(): When printing the parent record,
    use the index in the parent table rather than the index in the child table.

  #15653: "Slow inserts to InnoDB if many thousands of .ibd files"
    Keep track on unflushed modifications to file spaces.  When there are tens
    of thousands of file spaces, flushing all files in fil_flush_file_spaces()
    would be very slow.
    fil_flush_file_spaces(): Only flush unflushed file spaces.
    fil_space_t, fil_system_t: Add a list of unflushed spaces.

  #15991: "innodb-file-per-table + symlink database + rename = cr"
   os_file_handle_error(): Map the error codes EXDEV, ENOTDIR, and EISDIR
   to the new code OS_FILE_PATH_ERROR. Treat this code as OS_FILE_PATH_ERROR.
   This fixes the crash on RENAME TABLE when the .ibd file is a symbolic link
   to a different file system.

  #16157: "InnoDB crashes when main location settings are empty"
    This patch is from Heikki.

  #16298: "InnoDB segfaults in INSERTs in upgrade of 4.0 -> 5.0 tables
    with VARCHAR BINARY"
    dict_load_columns(): Set the charset-collation code
    DATA_MYSQL_BINARY_CHARSET_COLL for those binary string columns
    that lack a charset-collation code, i.e., the tables were created
    with an older version of MySQL/InnoDB than 4.1.2.

  #16229: "MySQL/InnoDB uses full explicit table locks in trigger processing"
    Take a InnoDB table lock only if user has explicitly requested a table
    lock. Added some additional comments to store_lock() and external_lock().

  #16387: "InnoDB crash when dropping a foreign key <table>_ibfk_0"
    Do not mistake TABLENAME_ibfk_0 for auto-generated id.
    dict_table_get_highest_foreign_id(): Ignore foreign constraint
    identifiers starting with the pattern TABLENAME_ibfk_0.

  #16582: "InnoDB: Error in an adaptive hash index pointer to page"
    Account for a race condition when dropping the adaptive hash index
    for a B-tree page.
    btr_search_drop_page_hash_index(): Retry the operation if a hash index
    with different parameters was built meanwhile.  Add diagnostics for the
    case that hash node pointers to the page remain.
    btr_search_info_update_hash(), btr_search_info_update_slow():
    Document the parameter "info" as in/out.

  #16814: "SHOW INNODB STATUS format error in LATEST FOREIGN KEY ERROR
    section"
    Add a missing newline to the LAST FOREIGN KEY ERROR section in SHOW
    INNODB STATUS output.
    dict_foreign_error_report(): Always print a newline after invoking
    dict_print_info_on_foreign_key_in_create_format().

  #16827: "Better InnoDB error message if ibdata files omitted from my.cnf"

  #17126: "CHECK TABLE on InnoDB causes a short hang during check of adaptive
    hash"
    CHECK TABLE blocking other queries, by releasing the btr_search_latch
    periodically during the adaptive hash table validation.

  #17405: "Valgrind: conditional jump or move depends on unititialised values"
    buf_block_init(): Reset magic_n, buf_fix_count and io_fix to avoid
    testing uninitialized variables.
2006-03-10 19:22:21 +03:00

664 lines
17 KiB
Text
Raw Blame History

/******************************************************
The database buffer buf_pool
(c) 1995 Innobase Oy
Created 11/5/1995 Heikki Tuuri
*******************************************************/
#include "buf0flu.h"
#include "buf0lru.h"
#include "buf0rea.h"
#include "mtr0mtr.h"
#ifdef UNIV_DEBUG
extern ulint buf_dbg_counter; /* This is used to insert validation
operations in execution in the
debug version */
#endif /* UNIV_DEBUG */
/************************************************************************
Recommends a move of a block to the start of the LRU list if there is danger
of dropping from the buffer pool. NOTE: does not reserve the buffer pool
mutex. */
UNIV_INLINE
ibool
buf_block_peek_if_too_old(
/*======================*/
/* out: TRUE if should be made younger */
buf_block_t* block) /* in: block to make younger */
{
return(buf_pool->freed_page_clock >= block->freed_page_clock
+ 1 + (buf_pool->curr_size / 1024));
}
/*************************************************************************
Gets the current size of buffer buf_pool in bytes. In the case of AWE, the
size of AWE window (= the frames). */
UNIV_INLINE
ulint
buf_pool_get_curr_size(void)
/*========================*/
/* out: size in bytes */
{
return((buf_pool->n_frames) * UNIV_PAGE_SIZE);
}
/*************************************************************************
Gets the maximum size of buffer buf_pool in bytes. In the case of AWE, the
size of AWE window (= the frames). */
UNIV_INLINE
ulint
buf_pool_get_max_size(void)
/*=======================*/
/* out: size in bytes */
{
return((buf_pool->n_frames) * UNIV_PAGE_SIZE);
}
/***********************************************************************
Accessor function for block array. */
UNIV_INLINE
buf_block_t*
buf_pool_get_nth_block(
/*===================*/
/* out: pointer to block */
buf_pool_t* buf_pool,/* in: buf_pool */
ulint i) /* in: index of the block */
{
ut_ad(buf_pool);
ut_ad(i < buf_pool->max_size);
return(i + buf_pool->blocks);
}
/***********************************************************************
Checks if a pointer points to the block array of the buffer pool (blocks, not
the frames). */
UNIV_INLINE
ibool
buf_pool_is_block(
/*==============*/
/* out: TRUE if pointer to block */
void* ptr) /* in: pointer to memory */
{
if ((buf_pool->blocks <= (buf_block_t*)ptr)
&& ((buf_block_t*)ptr < buf_pool->blocks
+ buf_pool->max_size)) {
return(TRUE);
}
return(FALSE);
}
/************************************************************************
Gets the smallest oldest_modification lsn for any page in the pool. Returns
ut_dulint_zero if all modified pages have been flushed to disk. */
UNIV_INLINE
dulint
buf_pool_get_oldest_modification(void)
/*==================================*/
/* out: oldest modification in pool,
ut_dulint_zero if none */
{
buf_block_t* block;
dulint lsn;
mutex_enter(&(buf_pool->mutex));
block = UT_LIST_GET_LAST(buf_pool->flush_list);
if (block == NULL) {
lsn = ut_dulint_zero;
} else {
lsn = block->oldest_modification;
}
mutex_exit(&(buf_pool->mutex));
return(lsn);
}
/***********************************************************************
Increments the buf_pool clock by one and returns its new value. Remember
that in the 32 bit version the clock wraps around at 4 billion! */
UNIV_INLINE
ulint
buf_pool_clock_tic(void)
/*====================*/
/* out: new clock value */
{
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&(buf_pool->mutex)));
#endif /* UNIV_SYNC_DEBUG */
buf_pool->ulint_clock++;
return(buf_pool->ulint_clock);
}
/*************************************************************************
Gets a pointer to the memory frame of a block. */
UNIV_INLINE
buf_frame_t*
buf_block_get_frame(
/*================*/
/* out: pointer to the frame */
buf_block_t* block) /* in: pointer to the control block */
{
ut_ad(block);
ut_ad(block >= buf_pool->blocks);
ut_ad(block < buf_pool->blocks + buf_pool->max_size);
ut_ad(block->state != BUF_BLOCK_NOT_USED);
ut_ad((block->state != BUF_BLOCK_FILE_PAGE)
|| (block->buf_fix_count > 0));
return(block->frame);
}
/*************************************************************************
Gets the space id of a block. */
UNIV_INLINE
ulint
buf_block_get_space(
/*================*/
/* out: space id */
buf_block_t* block) /* in: pointer to the control block */
{
ut_ad(block);
ut_ad(block >= buf_pool->blocks);
ut_ad(block < buf_pool->blocks + buf_pool->max_size);
ut_a(block->state == BUF_BLOCK_FILE_PAGE);
ut_ad(block->buf_fix_count > 0);
return(block->space);
}
/*************************************************************************
Gets the page number of a block. */
UNIV_INLINE
ulint
buf_block_get_page_no(
/*==================*/
/* out: page number */
buf_block_t* block) /* in: pointer to the control block */
{
ut_ad(block);
ut_ad(block >= buf_pool->blocks);
ut_ad(block < buf_pool->blocks + buf_pool->max_size);
ut_a(block->state == BUF_BLOCK_FILE_PAGE);
ut_ad(block->buf_fix_count > 0);
return(block->offset);
}
/***********************************************************************
Gets the block to whose frame the pointer is pointing to. */
UNIV_INLINE
buf_block_t*
buf_block_align(
/*============*/
/* out: pointer to block */
byte* ptr) /* in: pointer to a frame */
{
buf_block_t* block;
buf_frame_t* frame_zero;
ut_ad(ptr);
frame_zero = buf_pool->frame_zero;
if (UNIV_UNLIKELY((ulint)ptr < (ulint)frame_zero)
|| UNIV_UNLIKELY((ulint)ptr > (ulint)(buf_pool->high_end))) {
ut_print_timestamp(stderr);
fprintf(stderr,
"InnoDB: Error: trying to access a stray pointer %p\n"
"InnoDB: buf pool start is at %p, end at %p\n"
"InnoDB: Probable reason is database corruption or memory\n"
"InnoDB: corruption. If this happens in an InnoDB database recovery,\n"
"InnoDB: you can look from section 6.1 at http://www.innodb.com/ibman.html\n"
"InnoDB: how to force recovery.\n",
ptr, frame_zero,
buf_pool->high_end);
ut_error;
}
block = *(buf_pool->blocks_of_frames + (((ulint)(ptr - frame_zero))
>> UNIV_PAGE_SIZE_SHIFT));
return(block);
}
/***********************************************************************
Gets the frame the pointer is pointing to. */
UNIV_INLINE
buf_frame_t*
buf_frame_align(
/*============*/
/* out: pointer to frame */
byte* ptr) /* in: pointer to a frame */
{
buf_frame_t* frame;
ut_ad(ptr);
frame = ut_align_down(ptr, UNIV_PAGE_SIZE);
if (UNIV_UNLIKELY((ulint)frame < (ulint)(buf_pool->frame_zero))
|| UNIV_UNLIKELY((ulint)frame >= (ulint)(buf_pool->high_end))) {
ut_print_timestamp(stderr);
fprintf(stderr,
"InnoDB: Error: trying to access a stray pointer %p\n"
"InnoDB: buf pool start is at %p, end at %p\n"
"InnoDB: Probable reason is database corruption or memory\n"
"InnoDB: corruption. If this happens in an InnoDB database recovery,\n"
"InnoDB: you can look from section 6.1 at http://www.innodb.com/ibman.html\n"
"InnoDB: how to force recovery.\n",
ptr, buf_pool->frame_zero,
buf_pool->high_end);
ut_error;
}
return(frame);
}
/**************************************************************************
Gets the page number of a pointer pointing within a buffer frame containing
a file page. */
UNIV_INLINE
ulint
buf_frame_get_page_no(
/*==================*/
/* out: page number */
byte* ptr) /* in: pointer to within a buffer frame */
{
return(buf_block_get_page_no(buf_block_align(ptr)));
}
/**************************************************************************
Gets the space id of a pointer pointing within a buffer frame containing a
file page. */
UNIV_INLINE
ulint
buf_frame_get_space_id(
/*===================*/
/* out: space id */
byte* ptr) /* in: pointer to within a buffer frame */
{
return(buf_block_get_space(buf_block_align(ptr)));
}
/**************************************************************************
Gets the space id, page offset, and byte offset within page of a
pointer pointing to a buffer frame containing a file page. */
UNIV_INLINE
void
buf_ptr_get_fsp_addr(
/*=================*/
byte* ptr, /* in: pointer to a buffer frame */
ulint* space, /* out: space id */
fil_addr_t* addr) /* out: page offset and byte offset */
{
buf_block_t* block;
block = buf_block_align(ptr);
*space = buf_block_get_space(block);
addr->page = buf_block_get_page_no(block);
addr->boffset = ptr - buf_frame_align(ptr);
}
/**************************************************************************
Gets the hash value of the page the pointer is pointing to. This can be used
in searches in the lock hash table. */
UNIV_INLINE
ulint
buf_frame_get_lock_hash_val(
/*========================*/
/* out: lock hash value */
byte* ptr) /* in: pointer to within a buffer frame */
{
buf_block_t* block;
block = buf_block_align(ptr);
return(block->lock_hash_val);
}
/**************************************************************************
Gets the mutex number protecting the page record lock hash chain in the lock
table. */
UNIV_INLINE
mutex_t*
buf_frame_get_lock_mutex(
/*=====================*/
/* out: mutex */
byte* ptr) /* in: pointer to within a buffer frame */
{
buf_block_t* block;
block = buf_block_align(ptr);
return(block->lock_mutex);
}
/*************************************************************************
Copies contents of a buffer frame to a given buffer. */
UNIV_INLINE
byte*
buf_frame_copy(
/*===========*/
/* out: buf */
byte* buf, /* in: buffer to copy to */
buf_frame_t* frame) /* in: buffer frame */
{
ut_ad(buf && frame);
ut_memcpy(buf, frame, UNIV_PAGE_SIZE);
return(buf);
}
/************************************************************************
Calculates a folded value of a file page address to use in the page hash
table. */
UNIV_INLINE
ulint
buf_page_address_fold(
/*==================*/
/* out: the folded value */
ulint space, /* in: space id */
ulint offset) /* in: offset of the page within space */
{
return((space << 20) + space + offset);
}
/************************************************************************
This function is used to get info if there is an io operation
going on on a buffer page. */
UNIV_INLINE
ibool
buf_page_io_query(
/*==============*/
/* out: TRUE if io going on */
buf_block_t* block) /* in: buf_pool block, must be bufferfixed */
{
mutex_enter(&(buf_pool->mutex));
ut_ad(block->state == BUF_BLOCK_FILE_PAGE);
ut_ad(block->buf_fix_count > 0);
if (block->io_fix != 0) {
mutex_exit(&(buf_pool->mutex));
return(TRUE);
}
mutex_exit(&(buf_pool->mutex));
return(FALSE);
}
/************************************************************************
Gets the youngest modification log sequence number for a frame. Returns zero
if not a file page or no modification occurred yet. */
UNIV_INLINE
dulint
buf_frame_get_newest_modification(
/*==============================*/
/* out: newest modification to the page */
buf_frame_t* frame) /* in: pointer to a frame */
{
buf_block_t* block;
dulint lsn;
ut_ad(frame);
block = buf_block_align(frame);
mutex_enter(&(buf_pool->mutex));
if (block->state == BUF_BLOCK_FILE_PAGE) {
lsn = block->newest_modification;
} else {
lsn = ut_dulint_zero;
}
mutex_exit(&(buf_pool->mutex));
return(lsn);
}
/************************************************************************
Increments the modify clock of a frame by 1. The caller must (1) own the
buf_pool mutex and block bufferfix count has to be zero, (2) or own an x-lock
on the block. */
UNIV_INLINE
dulint
buf_frame_modify_clock_inc(
/*=======================*/
/* out: new value */
buf_frame_t* frame) /* in: pointer to a frame */
{
buf_block_t* block;
ut_ad(frame);
block = buf_block_align(frame);
#ifdef UNIV_SYNC_DEBUG
ut_ad((mutex_own(&(buf_pool->mutex)) && (block->buf_fix_count == 0))
|| rw_lock_own(&(block->lock), RW_LOCK_EXCLUSIVE));
#endif /*UNIV_SYNC_DEBUG */
UT_DULINT_INC(block->modify_clock);
return(block->modify_clock);
}
/************************************************************************
Increments the modify clock of a frame by 1. The caller must (1) own the
buf_pool mutex and block bufferfix count has to be zero, (2) or own an x-lock
on the block. */
UNIV_INLINE
dulint
buf_block_modify_clock_inc(
/*=======================*/
/* out: new value */
buf_block_t* block) /* in: block */
{
#ifdef UNIV_SYNC_DEBUG
ut_ad((mutex_own(&(buf_pool->mutex)) && (block->buf_fix_count == 0))
|| rw_lock_own(&(block->lock), RW_LOCK_EXCLUSIVE));
#endif /* UNIV_SYNC_DEBUG */
UT_DULINT_INC(block->modify_clock);
return(block->modify_clock);
}
/************************************************************************
Returns the value of the modify clock. The caller must have an s-lock
or x-lock on the block. */
UNIV_INLINE
dulint
buf_block_get_modify_clock(
/*=======================*/
/* out: value */
buf_block_t* block) /* in: block */
{
#ifdef UNIV_SYNC_DEBUG
ut_ad(rw_lock_own(&(block->lock), RW_LOCK_SHARED)
|| rw_lock_own(&(block->lock), RW_LOCK_EXCLUSIVE));
#endif /* UNIV_SYNC_DEBUG */
return(block->modify_clock);
}
#ifdef UNIV_SYNC_DEBUG
/***********************************************************************
Increments the bufferfix count. */
UNIV_INLINE
void
buf_block_buf_fix_inc_debug(
/*========================*/
buf_block_t* block, /* in: block to bufferfix */
const char* file __attribute__ ((unused)), /* in: file name */
ulint line __attribute__ ((unused))) /* in: line */
{
#ifdef UNIV_SYNC_DEBUG
ibool ret;
ret = rw_lock_s_lock_func_nowait(&(block->debug_latch), file, line);
ut_ad(ret == TRUE);
#endif
block->buf_fix_count++;
}
#else /* UNIV_SYNC_DEBUG */
/***********************************************************************
Increments the bufferfix count. */
UNIV_INLINE
void
buf_block_buf_fix_inc(
/*==================*/
buf_block_t* block) /* in: block to bufferfix */
{
block->buf_fix_count++;
}
#endif /* UNIV_SYNC_DEBUG */
/**********************************************************************
Returns the control block of a file page, NULL if not found. */
UNIV_INLINE
buf_block_t*
buf_page_hash_get(
/*==============*/
/* out: block, NULL if not found */
ulint space, /* in: space id */
ulint offset) /* in: offset of the page within space */
{
buf_block_t* block;
ulint fold;
ut_ad(buf_pool);
#ifdef UNIV_SYNC_DEBUG
ut_ad(mutex_own(&(buf_pool->mutex)));
#endif /* UNIV_SYNC_DEBUG */
/* Look for the page in the hash table */
fold = buf_page_address_fold(space, offset);
HASH_SEARCH(hash, buf_pool->page_hash, fold, block,
(block->space == space) && (block->offset == offset));
ut_a(block == NULL || block->state == BUF_BLOCK_FILE_PAGE);
return(block);
}
/************************************************************************
Tries to get the page, but if file io is required, releases all latches
in mtr down to the given savepoint. If io is required, this function
retrieves the page to buffer buf_pool, but does not bufferfix it or latch
it. */
UNIV_INLINE
buf_frame_t*
buf_page_get_release_on_io(
/*=======================*/
/* out: pointer to the frame, or NULL
if not in buffer buf_pool */
ulint space, /* in: space id */
ulint offset, /* in: offset of the page within space
in units of a page */
buf_frame_t* guess, /* in: guessed frame or NULL */
ulint rw_latch, /* in: RW_X_LATCH, RW_S_LATCH,
or RW_NO_LATCH */
ulint savepoint, /* in: mtr savepoint */
mtr_t* mtr) /* in: mtr */
{
buf_frame_t* frame;
frame = buf_page_get_gen(space, offset, rw_latch, guess,
BUF_GET_IF_IN_POOL,
__FILE__, __LINE__,
mtr);
if (frame != NULL) {
return(frame);
}
/* The page was not in the buffer buf_pool: release the latches
down to the savepoint */
mtr_rollback_to_savepoint(mtr, savepoint);
buf_page_get(space, offset, RW_S_LATCH, mtr);
/* When we get here, the page is in buffer, but we release
the latches again down to the savepoint, before returning */
mtr_rollback_to_savepoint(mtr, savepoint);
return(NULL);
}
/************************************************************************
Decrements the bufferfix count of a buffer control block and releases
a latch, if specified. */
UNIV_INLINE
void
buf_page_release(
/*=============*/
buf_block_t* block, /* in: buffer block */
ulint rw_latch, /* in: RW_S_LATCH, RW_X_LATCH,
RW_NO_LATCH */
mtr_t* mtr) /* in: mtr */
{
ulint buf_fix_count;
ut_ad(block);
mutex_enter_fast(&(buf_pool->mutex));
ut_a(block->state == BUF_BLOCK_FILE_PAGE);
ut_a(block->buf_fix_count > 0);
if (rw_latch == RW_X_LATCH && mtr->modifications) {
buf_flush_note_modification(block, mtr);
}
#ifdef UNIV_SYNC_DEBUG
rw_lock_s_unlock(&(block->debug_latch));
#endif
buf_fix_count = block->buf_fix_count;
block->buf_fix_count = buf_fix_count - 1;
mutex_exit(&(buf_pool->mutex));
if (rw_latch == RW_S_LATCH) {
rw_lock_s_unlock(&(block->lock));
} else if (rw_latch == RW_X_LATCH) {
rw_lock_x_unlock(&(block->lock));
}
}
#ifdef UNIV_SYNC_DEBUG
/*************************************************************************
Adds latch level info for the rw-lock protecting the buffer frame. This
should be called in the debug version after a successful latching of a
page if we know the latching order level of the acquired latch. If
UNIV_SYNC_DEBUG is not defined, compiles to an empty function. */
UNIV_INLINE
void
buf_page_dbg_add_level(
/*===================*/
buf_frame_t* frame __attribute__((unused)), /* in: buffer page
where we have acquired latch */
ulint level __attribute__((unused))) /* in: latching order
level */
{
sync_thread_add_level(&(buf_block_align(frame)->lock), level);
}
#endif /* UNIV_SYNC_DEBUG */
Reference in a new issue View git blame Copy permalink