Commit f847e588 authored by mithun's avatar mithun

Bug #17708621 : EXCEEDING SORT_BUFFER_SIZE (FILE SORT)

                WITH SORT ABORTED LEAKS FILE DESCRIPTORS

ISSUE : IO_CACHE used for index_merge quick select
is freed only on successful retrieval of all rows
from index merge.
Suppose if there is a interrupt( or failure) to
this operation of row retrieval (let it be a
KILL_QUERY signal) then we are not freeing the IO_CACHE
resources allocated by index_merge quick select.
And hence temp file associated with it is also not closed.
This lead to a file descriptor leak.

SOLUTION : As part of file sort operation now we always 
free the IO_CACHE allocated by index_merge quick select.
parent 4f5eab39
......@@ -34,6 +34,7 @@
#include "sql_test.h" // TEST_filesort
#include "opt_range.h" // SQL_SELECT
#include "debug_sync.h"
#include "sql_base.h"
/// How to write record_ref.
#define WRITE_REF(file,from) \
......@@ -337,6 +338,14 @@ ha_rows filesort(THD *thd, TABLE *table, SORT_FIELD *sortorder, uint s_length,
{
int kill_errno= thd->killed_errno();
DBUG_ASSERT(thd->is_error() || kill_errno);
/*
We replace the table->sort at the end.
Hence calling free_io_cache to make sure table->sort.io_cache
used for QUICK_INDEX_MERGE_SELECT is free.
*/
free_io_cache(table);
my_printf_error(ER_FILSORT_ABORT,
"%s: %s",
MYF(ME_ERROR + ME_WAITTANG),
......@@ -363,6 +372,10 @@ ha_rows filesort(THD *thd, TABLE *table, SORT_FIELD *sortorder, uint s_length,
#ifdef SKIP_DBUG_IN_FILESORT
DBUG_POP(); /* Ok to DBUG */
#endif
/* table->sort.io_cache should be free by this time */
DBUG_ASSERT(NULL == table->sort.io_cache);
memcpy(&table->sort, &table_sort, sizeof(FILESORT_INFO));
DBUG_PRINT("exit",("num_rows: %ld", (long) num_rows));
MYSQL_FILESORT_DONE(error, num_rows);
......@@ -602,6 +615,7 @@ static ha_rows find_all_keys(SORTPARAM *param, SQL_SELECT *select,
(uchar*) sort_form);
sort_form->column_bitmaps_set(&sort_form->tmp_set, &sort_form->tmp_set);
DEBUG_SYNC(thd, "after_index_merge_phase1");
for (;;)
{
if (quick_select)
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment