mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-01-16 12:02:42 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions
mariadb/sql/records.cc
monty@mysql.com/narttu.mysql.fi 088e2395f1 WL#3817: Simplify string / memory area types and make things more consistent (first part) 
The following type conversions was done:

- Changed byte to uchar
- Changed gptr to uchar*
- Change my_string to char *
- Change my_size_t to size_t
- Change size_s to size_t

Removed declaration of byte, gptr, my_string, my_size_t and size_s. 

Following function parameter changes was done:
- All string functions in mysys/strings was changed to use size_t
  instead of uint for string lengths.
- All read()/write() functions changed to use size_t (including vio).
- All protocoll functions changed to use size_t instead of uint
- Functions that used a pointer to a string length was changed to use size_t*
- Changed malloc(), free() and related functions from using gptr to use void *
  as this requires fewer casts in the code and is more in line with how the
  standard functions work.
- Added extra length argument to dirname_part() to return the length of the
  created string.
- Changed (at least) following functions to take uchar* as argument:
  - db_dump()
  - my_net_write()
  - net_write_command()
  - net_store_data()
  - DBUG_DUMP()
  - decimal2bin() & bin2decimal()
- Changed my_compress() and my_uncompress() to use size_t. Changed one
  argument to my_uncompress() from a pointer to a value as we only return
  one value (makes function easier to use).
- Changed type of 'pack_data' argument to packfrm() to avoid casts.
- Changed in readfrm() and writefrom(), ha_discover and handler::discover()
  the type for argument 'frmdata' to uchar** to avoid casts.
- Changed most Field functions to use uchar* instead of char* (reduced a lot of
  casts).
- Changed field->val_xxx(xxx, new_ptr) to take const pointers.

Other changes:
- Removed a lot of not needed casts
- Added a few new cast required by other changes
- Added some cast to my_multi_malloc() arguments for safety (as string lengths
  needs to be uint, not size_t).
- Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done
  explicitely as this conflict was often hided by casting the function to
  hash_get_key).
- Changed some buffers to memory regions to uchar* to avoid casts.
- Changed some string lengths from uint to size_t.
- Changed field->ptr to be uchar* instead of char*. This allowed us to
  get rid of a lot of casts.
- Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar
- Include zlib.h in some files as we needed declaration of crc32()
- Changed MY_FILE_ERROR to be (size_t) -1.
- Changed many variables to hold the result of my_read() / my_write() to be
  size_t. This was needed to properly detect errors (which are
  returned as (size_t) -1).
- Removed some very old VMS code
- Changed packfrm()/unpackfrm() to not be depending on uint size
  (portability fix)
- Removed windows specific code to restore cursor position as this
  causes slowdown on windows and we should not mix read() and pread()
  calls anyway as this is not thread safe. Updated function comment to
  reflect this. Changed function that depended on original behavior of
  my_pwrite() to itself restore the cursor position (one such case).
- Added some missing checking of return value of malloc().
- Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow.
- Changed type of table_def::m_size from my_size_t to ulong to reflect that
  m_size is the number of elements in the array, not a string/memory
  length.
- Moved THD::max_row_length() to table.cc (as it's not depending on THD).
  Inlined max_row_length_blob() into this function.
- More function comments
- Fixed some compiler warnings when compiled without partitions.
- Removed setting of LEX_STRING() arguments in declaration (portability fix).
- Some trivial indentation/variable name changes.
- Some trivial code simplifications:
  - Replaced some calls to alloc_root + memcpy to use
    strmake_root()/strdup_root().
  - Changed some calls from memdup() to strmake() (Safety fix)
  - Simpler loops in client-simple.c
2007-05-10 12:59:39 +03:00

614 lines
18 KiB
C++
Raw Blame History

/* Copyright (C) 2000-2006 MySQL AB
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 */
/* Functions for easy reading of records, possible through a cache */
#include "mysql_priv.h"
static int rr_quick(READ_RECORD *info);
int rr_sequential(READ_RECORD *info);
static int rr_from_tempfile(READ_RECORD *info);
static int rr_unpack_from_tempfile(READ_RECORD *info);
static int rr_unpack_from_buffer(READ_RECORD *info);
static int rr_from_pointers(READ_RECORD *info);
static int rr_from_cache(READ_RECORD *info);
static int init_rr_cache(THD *thd, READ_RECORD *info);
static int rr_cmp(uchar *a,uchar *b);
static int rr_index_first(READ_RECORD *info);
static int rr_index(READ_RECORD *info);
/*
Initialize READ_RECORD structure to perform full index scan
SYNOPSIS
init_read_record_idx()
info READ_RECORD structure to initialize.
thd Thread handle
table Table to be accessed
print_error If true, call table->file->print_error() if an error
occurs (except for end-of-records error)
idx index to scan
DESCRIPTION
Initialize READ_RECORD structure to perform full index scan (in forward
direction) using read_record.read_record() interface.
This function has been added at late stage and is used only by
UPDATE/DELETE. Other statements perform index scans using
join_read_first/next functions.
*/
void init_read_record_idx(READ_RECORD *info, THD *thd, TABLE *table,
bool print_error, uint idx)
{
bzero((char*) info,sizeof(*info));
info->table= table;
info->file= table->file;
info->record= table->record[0];
info->print_error= print_error;
table->status=0; /* And it's always found */
if (!table->file->inited)
table->file->ha_index_init(idx, 1);
/* read_record will be changed to rr_index in rr_index_first */
info->read_record= rr_index_first;
}
/*
init_read_record is used to scan by using a number of different methods.
Which method to use is set-up in this call so that later calls to
the info->read_record will call the appropriate method using a function
pointer.
There are five methods that relate completely to the sort function
filesort. The result of a filesort is retrieved using read_record
calls. The other two methods are used for normal table access.
The filesort will produce references to the records sorted, these
references can be stored in memory or in a temporary file.
The temporary file is normally used when the references doesn't fit into
a properly sized memory buffer. For most small queries the references
are stored in the memory buffer.
The temporary file is also used when performing an update where a key is
modified.
Methods used when ref's are in memory (using rr_from_pointers):
rr_unpack_from_buffer:
----------------------
This method is used when table->sort.addon_field is allocated.
This is allocated for most SELECT queries not involving any BLOB's.
In this case the records are fetched from a memory buffer.
rr_from_pointers:
-----------------
Used when the above is not true, UPDATE, DELETE and so forth and
SELECT's involving BLOB's. It is also used when the addon_field
buffer is not allocated due to that its size was bigger than the
session variable max_length_for_sort_data.
In this case the record data is fetched from the handler using the
saved reference using the rnd_pos handler call.
Methods used when ref's are in a temporary file (using rr_from_tempfile)
rr_unpack_from_tempfile:
------------------------
Same as rr_unpack_from_buffer except that references are fetched from
temporary file. Should obviously not really happen other than in
strange configurations.
rr_from_tempfile:
-----------------
Same as rr_from_pointers except that references are fetched from
temporary file instead of from
rr_from_cache:
--------------
This is a special variant of rr_from_tempfile that can be used for
handlers that is not using the HA_FAST_KEY_READ table flag. Instead
of reading the references one by one from the temporary file it reads
a set of them, sorts them and reads all of them into a buffer which
is then used for a number of subsequent calls to rr_from_cache.
It is only used for SELECT queries and a number of other conditions
on table size.
All other accesses use either index access methods (rr_quick) or a full
table scan (rr_sequential).
rr_quick:
---------
rr_quick uses one of the QUICK_SELECT classes in opt_range.cc to
perform an index scan. There are loads of functionality hidden
in these quick classes. It handles all index scans of various kinds.
rr_sequential:
--------------
This is the most basic access method of a table using rnd_init,
rnd_next and rnd_end. No indexes are used.
*/
void init_read_record(READ_RECORD *info,THD *thd, TABLE *table,
SQL_SELECT *select,
int use_record_cache, bool print_error)
{
IO_CACHE *tempfile;
DBUG_ENTER("init_read_record");
bzero((char*) info,sizeof(*info));
info->thd=thd;
info->table=table;
info->file= table->file;
info->forms= &info->table; /* Only one table */
if (table->s->tmp_table == NON_TRANSACTIONAL_TMP_TABLE &&
!table->sort.addon_field)
VOID(table->file->extra(HA_EXTRA_MMAP));
if (table->sort.addon_field)
{
info->rec_buf= table->sort.addon_buf;
info->ref_length= table->sort.addon_length;
}
else
{
info->record= table->record[0];
info->ref_length= table->file->ref_length;
}
info->select=select;
info->print_error=print_error;
info->ignore_not_found_rows= 0;
table->status=0; /* And it's always found */
if (select && my_b_inited(&select->file))
tempfile= &select->file;
else
tempfile= table->sort.io_cache;
if (tempfile && my_b_inited(tempfile)) // Test if ref-records was used
{
DBUG_PRINT("info",("using rr_from_tempfile"));
info->read_record= (table->sort.addon_field ?
rr_unpack_from_tempfile : rr_from_tempfile);
info->io_cache=tempfile;
reinit_io_cache(info->io_cache,READ_CACHE,0L,0,0);
info->ref_pos=table->file->ref;
if (!table->file->inited)
table->file->ha_rnd_init(0);
/*
table->sort.addon_field is checked because if we use addon fields,
it doesn't make sense to use cache - we don't read from the table
and table->sort.io_cache is read sequentially
*/
if (!table->sort.addon_field &&
! (specialflag & SPECIAL_SAFE_MODE) &&
thd->variables.read_rnd_buff_size &&
!(table->file->ha_table_flags() & HA_FAST_KEY_READ) &&
(table->db_stat & HA_READ_ONLY ||
table->reginfo.lock_type <= TL_READ_NO_INSERT) &&
(ulonglong) table->s->reclength* (table->file->stats.records+
table->file->stats.deleted) >
(ulonglong) MIN_FILE_LENGTH_TO_USE_ROW_CACHE &&
info->io_cache->end_of_file/info->ref_length * table->s->reclength >
(my_off_t) MIN_ROWS_TO_USE_TABLE_CACHE &&
!table->s->blob_fields &&
info->ref_length <= MAX_REFLENGTH)
{
if (! init_rr_cache(thd, info))
{
DBUG_PRINT("info",("using rr_from_cache"));
info->read_record=rr_from_cache;
}
}
}
else if (select && select->quick)
{
DBUG_PRINT("info",("using rr_quick"));
info->read_record=rr_quick;
}
else if (table->sort.record_pointers)
{
DBUG_PRINT("info",("using record_pointers"));
table->file->ha_rnd_init(0);
info->cache_pos=table->sort.record_pointers;
info->cache_end=info->cache_pos+
table->sort.found_records*info->ref_length;
info->read_record= (table->sort.addon_field ?
rr_unpack_from_buffer : rr_from_pointers);
}
else
{
DBUG_PRINT("info",("using rr_sequential"));
info->read_record=rr_sequential;
table->file->ha_rnd_init(1);
/* We can use record cache if we don't update dynamic length tables */
if (!table->no_cache &&
(use_record_cache > 0 ||
(int) table->reginfo.lock_type <= (int) TL_READ_HIGH_PRIORITY ||
!(table->s->db_options_in_use & HA_OPTION_PACK_RECORD) ||
(use_record_cache < 0 &&
!(table->file->ha_table_flags() & HA_NOT_DELETE_WITH_CACHE))))
VOID(table->file->extra_opt(HA_EXTRA_CACHE,
thd->variables.read_buff_size));
}
/* Condition pushdown to storage engine */
if (thd->variables.engine_condition_pushdown &&
select && select->cond &&
(select->cond->used_tables() & table->map) &&
!table->file->pushed_cond)
table->file->cond_push(select->cond);
DBUG_VOID_RETURN;
} /* init_read_record */
void end_read_record(READ_RECORD *info)
{ /* free cache if used */
if (info->cache)
{
my_free_lock((char*) info->cache,MYF(0));
info->cache=0;
}
if (info->table)
{
filesort_free_buffers(info->table,0);
(void) info->file->extra(HA_EXTRA_NO_CACHE);
if (info->read_record != rr_quick) // otherwise quick_range does it
(void) info->file->ha_index_or_rnd_end();
info->table=0;
}
}
static int rr_handle_error(READ_RECORD *info, int error)
{
if (error == HA_ERR_END_OF_FILE)
error= -1;
else
{
if (info->print_error)
info->table->file->print_error(error, MYF(0));
if (error < 0) // Fix negative BDB errno
error= 1;
}
return error;
}
/* Read a record from head-database */
static int rr_quick(READ_RECORD *info)
{
int tmp;
while ((tmp= info->select->quick->get_next()))
{
if (info->thd->killed)
{
my_error(ER_SERVER_SHUTDOWN, MYF(0));
return 1;
}
if (tmp != HA_ERR_RECORD_DELETED)
{
tmp= rr_handle_error(info, tmp);
break;
}
}
return tmp;
}
/*
Reads first row in an index scan
SYNOPSIS
rr_index_first()
info Scan info
RETURN
0 Ok
-1 End of records
1 Error
*/
static int rr_index_first(READ_RECORD *info)
{
int tmp= info->file->index_first(info->record);
info->read_record= rr_index;
if (tmp)
tmp= rr_handle_error(info, tmp);
return tmp;
}
/*
Reads index sequentially after first row
SYNOPSIS
rr_index()
info Scan info
DESCRIPTION
Read the next index record (in forward direction) and translate return
value.
RETURN
0 Ok
-1 End of records
1 Error
*/
static int rr_index(READ_RECORD *info)
{
int tmp= info->file->index_next(info->record);
if (tmp)
tmp= rr_handle_error(info, tmp);
return tmp;
}
int rr_sequential(READ_RECORD *info)
{
int tmp;
while ((tmp=info->file->rnd_next(info->record)))
{
if (info->thd->killed)
{
info->thd->send_kill_message();
return 1;
}
/*
rnd_next can return RECORD_DELETED for MyISAM when one thread is
reading and another deleting without locks.
*/
if (tmp != HA_ERR_RECORD_DELETED)
{
tmp= rr_handle_error(info, tmp);
break;
}
}
return tmp;
}
static int rr_from_tempfile(READ_RECORD *info)
{
int tmp;
for (;;)
{
if (my_b_read(info->io_cache,info->ref_pos,info->ref_length))
return -1; /* End of file */
if (!(tmp=info->file->rnd_pos(info->record,info->ref_pos)))
break;
/* The following is extremely unlikely to happen */
if (tmp == HA_ERR_RECORD_DELETED ||
(tmp == HA_ERR_KEY_NOT_FOUND && info->ignore_not_found_rows))
continue;
tmp= rr_handle_error(info, tmp);
break;
}
return tmp;
} /* rr_from_tempfile */
/*
Read a result set record from a temporary file after sorting
SYNOPSIS
rr_unpack_from_tempfile()
info Reference to the context including record descriptors
DESCRIPTION
The function first reads the next sorted record from the temporary file.
into a buffer. If a success it calls a callback function that unpacks
the fields values use in the result set from this buffer into their
positions in the regular record buffer.
RETURN
0 - Record successfully read.
-1 - There is no record to be read anymore.
*/
static int rr_unpack_from_tempfile(READ_RECORD *info)
{
if (my_b_read(info->io_cache, info->rec_buf, info->ref_length))
return -1;
TABLE *table= info->table;
(*table->sort.unpack)(table->sort.addon_field, info->rec_buf);
return 0;
}
static int rr_from_pointers(READ_RECORD *info)
{
int tmp;
uchar *cache_pos;
for (;;)
{
if (info->cache_pos == info->cache_end)
return -1; /* End of file */
cache_pos= info->cache_pos;
info->cache_pos+= info->ref_length;
if (!(tmp=info->file->rnd_pos(info->record,cache_pos)))
break;
/* The following is extremely unlikely to happen */
if (tmp == HA_ERR_RECORD_DELETED ||
(tmp == HA_ERR_KEY_NOT_FOUND && info->ignore_not_found_rows))
continue;
tmp= rr_handle_error(info, tmp);
break;
}
return tmp;
}
/*
Read a result set record from a buffer after sorting
SYNOPSIS
rr_unpack_from_buffer()
info Reference to the context including record descriptors
DESCRIPTION
The function first reads the next sorted record from the sort buffer.
If a success it calls a callback function that unpacks
the fields values use in the result set from this buffer into their
positions in the regular record buffer.
RETURN
0 - Record successfully read.
-1 - There is no record to be read anymore.
*/
static int rr_unpack_from_buffer(READ_RECORD *info)
{
if (info->cache_pos == info->cache_end)
return -1; /* End of buffer */
TABLE *table= info->table;
(*table->sort.unpack)(table->sort.addon_field, info->cache_pos);
info->cache_pos+= info->ref_length;
return 0;
}
/* cacheing of records from a database */
static int init_rr_cache(THD *thd, READ_RECORD *info)
{
uint rec_cache_size;
DBUG_ENTER("init_rr_cache");
info->struct_length= 3+MAX_REFLENGTH;
info->reclength= ALIGN_SIZE(info->table->s->reclength+1);
if (info->reclength < info->struct_length)
info->reclength= ALIGN_SIZE(info->struct_length);
info->error_offset= info->table->s->reclength;
info->cache_records= (thd->variables.read_rnd_buff_size /
(info->reclength+info->struct_length));
rec_cache_size= info->cache_records*info->reclength;
info->rec_cache_size= info->cache_records*info->ref_length;
// We have to allocate one more byte to use uint3korr (see comments for it)
if (info->cache_records <= 2 ||
!(info->cache=(uchar*) my_malloc_lock(rec_cache_size+info->cache_records*
info->struct_length+1,
MYF(0))))
DBUG_RETURN(1);
#ifdef HAVE_purify
// Avoid warnings in qsort
bzero(info->cache,rec_cache_size+info->cache_records* info->struct_length+1);
#endif
DBUG_PRINT("info",("Allocated buffert for %d records",info->cache_records));
info->read_positions=info->cache+rec_cache_size;
info->cache_pos=info->cache_end=info->cache;
DBUG_RETURN(0);
} /* init_rr_cache */
static int rr_from_cache(READ_RECORD *info)
{
reg1 uint i;
ulong length;
my_off_t rest_of_file;
int16 error;
uchar *position,*ref_position,*record_pos;
ulong record;
for (;;)
{
if (info->cache_pos != info->cache_end)
{
if (info->cache_pos[info->error_offset])
{
shortget(error,info->cache_pos);
if (info->print_error)
info->table->file->print_error(error,MYF(0));
}
else
{
error=0;
memcpy(info->record,info->cache_pos,
(size_t) info->table->s->reclength);
}
info->cache_pos+=info->reclength;
return ((int) error);
}
length=info->rec_cache_size;
rest_of_file=info->io_cache->end_of_file - my_b_tell(info->io_cache);
if ((my_off_t) length > rest_of_file)
length= (ulong) rest_of_file;
if (!length || my_b_read(info->io_cache,info->cache,length))
{
DBUG_PRINT("info",("Found end of file"));
return -1; /* End of file */
}
length/=info->ref_length;
position=info->cache;
ref_position=info->read_positions;
for (i=0 ; i < length ; i++,position+=info->ref_length)
{
memcpy(ref_position,position,(size_t) info->ref_length);
ref_position+=MAX_REFLENGTH;
int3store(ref_position,(long) i);
ref_position+=3;
}
qsort(info->read_positions,length,info->struct_length,(qsort_cmp) rr_cmp);
position=info->read_positions;
for (i=0 ; i < length ; i++)
{
memcpy(info->ref_pos,position,(size_t) info->ref_length);
position+=MAX_REFLENGTH;
record=uint3korr(position);
position+=3;
record_pos=info->cache+record*info->reclength;
if ((error=(int16) info->file->rnd_pos(record_pos,info->ref_pos)))
{
record_pos[info->error_offset]=1;
shortstore(record_pos,error);
DBUG_PRINT("error",("Got error: %d:%d when reading row",
my_errno, error));
}
else
record_pos[info->error_offset]=0;
}
info->cache_end=(info->cache_pos=info->cache)+length*info->reclength;
}
} /* rr_from_cache */
static int rr_cmp(uchar *a,uchar *b)
{
if (a[0] != b[0])
return (int) a[0] - (int) b[0];
if (a[1] != b[1])
return (int) a[1] - (int) b[1];
if (a[2] != b[2])
return (int) a[2] - (int) b[2];
#if MAX_REFLENGTH == 4
return (int) a[3] - (int) b[3];
#else
if (a[3] != b[3])
return (int) a[3] - (int) b[3];
if (a[4] != b[4])
return (int) a[4] - (int) b[4];
if (a[5] != b[5])
return (int) a[1] - (int) b[5];
if (a[6] != b[6])
return (int) a[6] - (int) b[6];
return (int) a[7] - (int) b[7];
#endif
}
Reference in a new issue View git blame Copy permalink