MySQL 派生表查詢導(dǎo)致 Crash 的根源分析與解決方案
作者：wuyy
對于復(fù)雜的 SQL 會執(zhí)行復(fù)雜的 Item 變換和刪除不需要的 Item，但是正是由于這樣才更容易導(dǎo)致 Crash 的出現(xiàn)。分析類似這樣的 Crash 問題的時候，因為涉及代碼量大，代碼邏輯復(fù)雜往往很難找到相關(guān)修復(fù)代碼，因此需要對代碼運行流程比較熟悉，同時要有相關(guān)復(fù)雜問題解決的經(jīng)驗才能更好的應(yīng)對這類問題。?
一、問題發(fā)現(xiàn)
在之前的 MySQL 8.0.32 使用中，發(fā)現(xiàn)使用以下帶有派生表的 SQL 會導(dǎo)致 MySQL Crash，以下的sequence_table(2)替換為任何非常量表都行：
僅 MySQL 8.0.32 版本有影響。
EXPLAIN FORMAT=TREE
select
trim(ref_15.c_ogj),
0<>0as c_lrcm63eani
from
(select
0<>0as c_ogj
from
sequence_table(2) t1
where0<>0
orderby c_ogj asc) as ref_15;Crash 的堆棧如下：
Thread 55 "mysqld" received signal SIGSEGV, Segmentation fault.
Item_view_ref::used_tables (this=0x7fff2418f410)
at sql/item.h:6670
6670 table_map inner_map = ref_item()->used_tables(); ==> ref_item()為空指針，因此crash了
(gdb) bt
#0 Item_view_ref::used_tables (this=0x7fff2418f410)
at sql/item.h:6670
#1 0x0000555558e978d1 in Item::const_item (this=0x7fff2418f410)
at sql/item.h:2342
#2 0x0000555558ecc765 in Item_ref::print (this=0x7fff2418f410, thd=0x7fff24001050,
str=0x7fffc83ee7e0, query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/item.cc:9993
#3 0x000055555903b839 in Item_func_trim::print (this=0x7fff24120d20, thd=0x7fff24001050,
str=0x7fffc83ee7e0, query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/item_strfunc.cc:3244
#4 0x0000555558ea7fc5 in Item::print_item_w_name (this=0x7fff24120d20, thd=0x7fff24001050,
str=0x7fffc83ee7e0, query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/item.cc:727
#5 0x00005555593f18c0 in Query_block::print_item_list (this=0x7fff24120768, thd=0x7fff24001050,
str=0x7fffc83ee7e0, query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/sql_lex.cc:4041
#6 0x00005555593efb50 in Query_block::print_query_block (this=0x7fff24120768,
thd=0x7fff24001050, str=0x7fffc83ee7e0,
query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/sql_lex.cc:3614
#7 0x00005555593efa3d in Query_block::print (this=0x7fff24120768, thd=0x7fff24001050,
str=0x7fffc83ee7e0, query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/sql_lex.cc:3598
#8 0x00005555593ee556 in Query_expression::print (this=0x7fff24120670, thd=0x7fff24001050,
str=0x7fffc83ee7e0, query_type=(QT_TO_SYSTEM_CHARSET | QT_SHOW_SELECT_NUMBER))
at sql/sql_lex.cc:3232
#9 0x0000555559a89c2c in print_query_for_explain (query_thd=0x7fff24001050,
unit=0x7fff24120670, str=0x7fffc83ee7e0)
at sql/opt_explain.cc:2288
#10 0x0000555559a10b11 in PrintQueryPlan[abi:cxx11](THD*, THD const*, Query_expression*) (
ethd=0x7fff24001050, query_thd=0x7fff24001050, unit=0x7fff24120670)
at sql/join_optimizer/explain_access_path.cc:1894
#11 0x0000555559a8985a in ExplainIterator (ethd=0x7fff24001050, query_thd=0x7fff24001050,
unit=0x7fff24120670) at sql/opt_explain.cc:2205
#12 0x0000555559a89e91 in explain_query (explain_thd=0x7fff24001050, query_thd=0x7fff24001050,
unit=0x7fff24120670) at sql/opt_explain.cc:2359
#13 0x000055555955cd46 in Sql_cmd_dml::execute_inner (this=0x7fff24165630, thd=0x7fff24001050)二、問題調(diào)查過程
調(diào)查執(zhí)行 SQL 的 optimize 的過程，分析發(fā)現(xiàn)該 SQL 的 SQL 變換情況如下：
以下的 trim(ref_15.c_ogj) 執(zhí)行完 find_order_in_list 后，Item_func_trim的args[0]->m_ref_item[0] 等于0<>0 as c_lrcm63eani，而不是0<>0 as c_ogj，這是因為c_lrcm63eani和c_ogj的名字都一樣，都是0<>0，在find_order_in_list函數(shù)里面由于名字一樣因此內(nèi)層字段被外層替代了。而后在Item::clean_up_after_removal執(zhí)行的時候，Item_func_ne即c_lrcm63eani因為出現(xiàn)了2次，因此執(zhí)行了2次decrement_ref_count()，然而在Query_block::delete_unused_merged_columns函數(shù)卻把0<>0 as c_lrcm63eani的Item置為空了，因為這個時候c_lrcm63eani的item->decrement_ref_count()以后ref_count()為0因此繼續(xù)執(zhí)行Item::clean_up_after_removal了。
EXPLAIN FORMAT=TREE
select
trim(ref_15.c_ogj),
0<>0as c_lrcm63eani
from
(select
0<>0as c_ogj
from
sequence_table(2) t1
where0<>0
orderby c_ogj asc) as ref_15;查看函數(shù)調(diào)用過程發(fā)現(xiàn) Query_block 在 prepare 的時候執(zhí)行了 delete_unused_merged_columns，
-- 函數(shù)調(diào)用過程： Query_block::prepare -> Query_block::apply_local_transforms -> Query_block::delete_unused_merged_columns
bool find_order_in_list() {
if (select_item != not_found_item) {
if ((*order->item)->real_item() != (*select_item)->real_item()) {
Item::Cleanup_after_removal_context ctx(
thd->lex->current_query_block());
(*order->item)
->walk(&Item::clean_up_after_removal, walk_options, ==>Item_func_ne執(zhí)行了2次，也執(zhí)行了2次decrement_ref_count()
pointer_cast<uchar *>(&ctx));
}
}
}
bool Query_block::apply_local_transforms(THD *thd, bool prune) {
DBUG_TRACE;
assert(first_execution);
-- 這個函數(shù)把((Item_func *)&fields[0][0])->args[0]->m_ref_item[0]給刪了
if (derived_table_count) delete_unused_merged_columns(&m_table_nest);
}
void Query_block::delete_unused_merged_columns(
mem_root_deque<Table_ref *> *tables) {
DBUG_TRACE;
for (Table_ref *tl : *tables) {
if (tl->nested_join == nullptr) continue;
if (tl->is_merged()) {
for (Field_translator *transl = tl->field_translation;
transl < tl->field_translation_end; transl++) {
Item *const item = transl->item;
// Decrement the ref count as its no more used in
// select list.
if (item->decrement_ref_count()) continue; -- 因為執(zhí)行完decrement_ref_count()以后返回的m_ref_count=0因此不會跳出這個循環(huán)
// Cleanup the item since its not referenced from
// anywhere.
assert(item->fixed);
Item::Cleanup_after_removal_context ctx(this);
item->walk(&Item::clean_up_after_removal, walk_options,
pointer_cast<uchar *>(&ctx));
transl->item = nullptr; -- 這個地方把Item_view_ref引用的Item_func_ne對象置為空了，即把trim函數(shù)參數(shù)的c_lrcm63eani列刪除了
}
}
delete_unused_merged_columns(&tl->nested_join->m_tables);
}
}三、解決方案
通過上面的分析，我們可以發(fā)現(xiàn)問題在于多執(zhí)行了一次Item::clean_up_after_removal，隨后在 MySQL 最新代碼嘗試執(zhí)行以上 SQL 發(fā)現(xiàn)該 BUG 已經(jīng)被修復(fù)，找到相關(guān)修復(fù)代碼，可以發(fā)現(xiàn)以下修復(fù)代碼。
相關(guān)commit ID號為： 2171a1260e2cdbbd379646be8ff6413a92fd48f4
-- 相關(guān)修復(fù)代碼如下：
@@ -7575,7 +7865,6 @@ bool Item::clean_up_after_removal(uchar *arg) {
if (reference_count() > 1) {
(void)decrement_ref_count();
+ ctx->stop_at(this);
}
return false;
}修改完查看一下這個函數(shù)的堆棧信息：
#0 Item::clean_up_after_removal (this=0x2,
arg=0x41 <error: Cannot access memory at address 0x41>)
at sql/item.cc:9236
#1 0x0000555558fea5a8 in Item::walk (this=0x7fff2c338db8, processor=&virtual table offset 864,
walk=7, arg=0x7fffc83ee4b0 "") at sql/item.h:2543
#2 0x00005555596cc6f2 in find_order_in_list(thd=0x7fff2c001070, ref_item_array=...,
tables=0x7fff2c330b90, order=0x7fff2c32eae8, fields=0x7fff2c32fb20, is_group_field=false,
is_window_order=false) at sql/sql_resolver.cc:4625
#3 0x00005555596cd0ae in setup_order(thd=0x7fff2c001070, ref_item_array=...,
tables=0x7fff2c330b90, fields=0x7fff2c32fb20, order=0x7fff2c32eae8)
at sql/sql_resolver.cc:4811
#4 0x00005555596bf528 in Query_block::prepare(this=0x7fff2c32fae0, thd=0x7fff2c001070,
insert_field_list=0x0) at sql/sql_resolver.cc:400
#5 0x00005555597d035d in Query_expression::prepare(this=0x7fff2c32f9e8, thd=0x7fff2c001070,
sel_result=0x7fff2c33b2a8, insert_field_list=0x0, added_options=0, removed_options=0)
at sql/sql_union.cc:758
#6 0x0000555559590772 in Table_ref::resolve_derived(this=0x7fff2c339790, thd=0x7fff2c001070,
apply_semijoin=true) at sql/sql_derived.cc:451
#7 0x00005555596c2a80 in Query_block::resolve_placeholder_tables(this=0x7fff2c333f08,
thd=0x7fff2c001070, apply_semijoin=true)
at sql/sql_resolver.cc:1408
#8 0x00005555596bea62 in Query_block::prepare(this=0x7fff2c333f08, thd=0x7fff2c001070,
insert_field_list=0x0) at sql/sql_resolver.cc:265對于0<>0 as c_lrcm63eani這個Item_func_ne對象，執(zhí)行到Item::clean_up_after_removal的時候，因為reference_count() > 1因此會執(zhí)行新添加的ctx->stop_at(this)，等到下一次再執(zhí)行到這個Item_func_ne的clean_up_after_removal()函數(shù)的時候，就會因為ctx->is_stopped(this)而直接返回，不再執(zhí)行一次decrement_ref_count()，從而避免了執(zhí)行后面的transl->item = nullptr。
bool find_order_in_list() {
if (select_item != not_found_item) {
if ((*order->item)->real_item() != (*select_item)->real_item()) {
Item::Cleanup_after_removal_context ctx(
thd->lex->current_query_block());
(*order->item)
->walk(&Item::clean_up_after_removal, walk_options, -- Item_func_ne執(zhí)行了2次，而只執(zhí)行了一次decrement_ref_count()
pointer_cast<uchar *>(&ctx));
}
}
}
void Query_block::delete_unused_merged_columns(
mem_root_deque<Table_ref *> *tables) {
DBUG_TRACE;
for (Table_ref *tl : *tables) {
if (tl->nested_join == nullptr) continue;
if (tl->is_merged()) {
for (Field_translator *transl = tl->field_translation;
transl < tl->field_translation_end; transl++) {
Item *const item = transl->item;
// Decrement the ref count as its no more used in
// select list.
if (item->decrement_ref_count()) continue; 因為執(zhí)行完decrement_ref_count()以后返回的m_ref_count=1因此不會繼續(xù)執(zhí)行后面的置空設(shè)置
// Cleanup the item since its not referenced from
// anywhere.
assert(item->fixed);
Item::Cleanup_after_removal_context ctx(this);
item->walk(&Item::clean_up_after_removal, walk_options,
pointer_cast<uchar *>(&ctx));
transl->item = nullptr; ==>這個地方不會運行到
}
}
delete_unused_merged_columns(&tl->nested_join->m_tables);
}
}四、問題總結(jié)
通過以上分析我們可以發(fā)現(xiàn)，對于復(fù)雜的 SQL 會執(zhí)行復(fù)雜的 Item 變換和刪除不需要的 Item，但是正是由于這樣才更容易導(dǎo)致 Crash 的出現(xiàn)。分析類似這樣的 Crash 問題的時候，因為涉及代碼量大，代碼邏輯復(fù)雜往往很難找到相關(guān)修復(fù)代碼，因此需要對代碼運行流程比較熟悉，同時要有相關(guān)復(fù)雜問題解決的經(jīng)驗才能更好的應(yīng)對這類問題。
責(zé)任編輯：武曉燕
來源：
GreatSQL社區(qū)
