[20210914]探究mutex的值 5.txt
[20210914]探究mutex的值 5.txt
--//前幾天做了做library_cache轉儲時,顯示的mutex結構體裡面相關資訊的探究.
Bucket: #=102650 Mutex=0x80528f40(0, 19, 0, 6)
--//注：11g 下每個library cache bucket佔用16位元組，後面跟著mutex，mutex結構佔用24位元組，這樣整個佔用40位元組。可以參考我前面
--//的測試 [20210524]分析library cache轉儲 3.txt
--//大概猜測出第2,3數字表示gets,sleeps的數量.後面的dump顯示總是6,不知道為什麼我感覺應該對應mutex結構體的20~23位元組.
--//前面第1個數字,轉儲總是顯示0,如果使用oradebug poke前面0~3,4-7位元組非0,dump總是掛起.導致我無法猜測第1個數字表示什麼.
oradebug poke 0x0000000080528f40 4 0x00000001
oradebug poke 0x0000000080528f44 4 0x00000002
--//我仔細看了以前我的測試,發現以前在理解上出現一些偏差,這個測試主要目的看看muetx的0-3,4-7到底那個是表示sid,那個是ref count.
--//透過例子說明問題:
1.環境：
SCOTT@book> @ ver1
PORT_STRING VERSION BANNER
------------------------------ -------------- --------------------------------------------------------------------------------
x86_64/Linux 2.4.xx 11.2.0.4.0 Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
SYS@book> alter system set session_cached_cursors=0 scope=spfile;
System altered.
--//重啟略,設定目的主要保證每次都是軟解析,這樣都會訪問library cache mutex.
--//session 1:
SCOTT@book> @ spid
SID SERIAL# PROCESS SERVER SPID PID P_SERIAL# C50
---------- ---------- ------------------------ --------- ------ ------- ---------- --------------------------------------------------
295 5 9539 DEDICATED 9540 21 3 alter system kill session '295,5' immediate;
--//sid=295 = 0x127.
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//執行5次以上。
SCOTT@book> @ hash
HASH_VALUE SQL_ID CHILD_NUMBER HASH_HEX
---------- ------------- ------------ ---------
95129850 80baj2c2ur47u 0 5ab90fa
--//95129850%131072 = 102650,確定library cache mutex 地址如下,透過轉儲library cache,過程略.
oradebug setmypid
oradebug dump library_cache 10;
--//檢查轉儲,搜尋Bucket: #=102650,發現如下:
Bucket: #=102650 Mutex=0x80528f40(0, 6, 0, 6)
2.測試:
--//session 2:
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 00000000 00000051 0000AE84 000190FA 00000000
SYS@book> oradebug poke 0x0000000080528f40 4 0x00000127
BEFORE: [080528F40, 080528F44) = 00000000
AFTER: [080528F40, 080528F44) = 00000127
SYS@book> oradebug poke 0x0000000080528f44 4 0x00000127
BEFORE: [080528F44, 080528F48) = 00000000
AFTER: [080528F44, 080528F48) = 00000127
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000127 00000127 00000051 0000AE84 000190FA 00000000
--//0x51 = 81
--//session 1:
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//你可以發現sid=127並沒有阻塞.
--//session 2:
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000127 00000127 00000051 0000AE84 000190FA 00000000
--//0x51 = 81,可以發現並沒有增加.
SYS@book> oradebug poke 0x0000000080528f44 4 0x00000000
BEFORE: [080528F44, 080528F48) = 00000127
AFTER: [080528F44, 080528F48) = 00000000
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000127 00000000 00000051 0000AE84 000190FA 00000000
--//session 1:
SCOTT@book> select * from dept where deptno=20;
..
--//這樣反而掛起.
--//session 2, 修改回來.
SYS@book> oradebug poke 0x0000000080528f44 4 0x00000127
BEFORE: [080528F44, 080528F48) = 00000000
AFTER: [080528F44, 080528F48) = 00000127
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000127 00000127 00000051 0000AE84 000190FA 00000000
--//還是掛起.
SYS@book> oradebug poke 0x0000000080528f40 4 0x00000000
BEFORE: [080528F40, 080528F44) = 00000127
AFTER: [080528F40, 080528F44) = 00000000
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 00000127 00000051 0000AE84 000190FA 00000000
--//還是掛起.
SYS@book> oradebug poke 0x0000000080528f44 4 0x00000000
BEFORE: [080528F44, 080528F48) = 00000127
AFTER: [080528F44, 080528F48) = 00000000
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 00000000 00000052 0001819C 000190FA 00000000
--//session 1:
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//到底那個表示sid,ref count我還是猜測不出來.
3.繼續:
--//開啟新會話,執行:
--//session 3:
SCOTT@book> @ spid
SID SERIAL# PROCESS SERVER SPID PID P_SERIAL# C50
---------- ---------- ------------------------ --------- ------ ------- ---------- --------------------------------------------------
58 85 10235 DEDICATED 10236 28 36 alter system kill session '58,85' immediate;
--//58 = 0x3a
--//session 2:
oradebug poke 0x0000000080528f40 4 0x0000003a
oradebug poke 0x0000000080528f44 4 0x0000003a
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 0000003A 0000003A 0000005B 0001819C 000190FA 00000000
--//session 3:
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//session 2:
SYS@book> oradebug poke 0x0000000080528f40 4 0x00000000
BEFORE: [080528F40, 080528F44) = 0000003A
AFTER: [080528F40, 080528F44) = 00000000
SYS@book> oradebug poke 0x0000000080528f44 4 0x0000003a
BEFORE: [080528F44, 080528F48) = 0000003A
AFTER: [080528F44, 080528F48) = 0000003A
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 0000003A 0000005B 0001819C 000190FA 00000000
--//session 3:
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//可以正常執行,也就是第4-7位元組=58(sid)的情況下可以正常執行.
SYS@book> oradebug poke 0x0000000080528f44 4 0x00000127
BEFORE: [080528F44, 080528F48) = 0000003A
AFTER: [080528F44, 080528F48) = 00000127
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 00000127 0000005B 0001819C 000190FA 00000000
--//session 3:
SCOTT@book> select * from dept where deptno=20;
--//掛起!!
--//session 1:
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//你可以發現在這樣的情況下sid=297(0x127)的會話可以正常執行.
--//session 2:
SYS@book> oradebug poke 0x0000000080528f44 4 0x00000000
BEFORE: [080528F44, 080528F48) = 00000127
AFTER: [080528F44, 080528F48) = 00000000
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 00000000 0000005B 0001B0F9 000190FA 00000000
--//session 3:
SCOTT@book> select * from dept where deptno=20;
DEPTNO DNAME LOC
---------- -------------- -------------
20 RESEARCH DALLAS
--//從這個測試可以看出第4-7位元組應該表示sid. 這樣0-3 表示ref count.
4.再繼續:
--//session 2:
SYS@book> oradebug poke 0x0000000080528f40 4 0x00000001
BEFORE: [080528F40, 080528F44) = 00000000
AFTER: [080528F40, 080528F44) = 00000001
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000001 00000000 00000060 0002CA11 000190FA 00000000
--//session 1:
SCOTT@book> select * from dept where deptno=20;
--//session 3:
SCOTT@book> select * from dept where deptno=20;
--//掛起!!
--//session 2:
SYS@book> @ wait
P1RAW P2RAW P3RAW P1 P2 P3 SID SERIAL# SEQ# EVENT STATUS STATE WAIT_TIME_MICRO SECONDS_IN_WAIT WAIT_CLASS CLIENT_INFO
---------------- ---------------- ---------------- ---------- ---------- ---------- ---------- ---------- ---------- ---------------------------------------- -------- ------------------- --------------- --------------- -------------------- --------------------
00000000000190FA 0000000000000001 000000000000003E 102650 1 62 58 85 31967 library cache: mutex X ACTIVE WAITED SHORT TIME 2 0 Concurrency
00000000000190FA 0000000000000001 000000000000003E 102650 1 62 295 5 16799 library cache: mutex X ACTIVE WAITED SHORT TIME 4 0 Concurrency
SYS@book> @ ev_name "library cache: mutex X"
EVENT# EVENT_ID NAME PARAMETER1 PARAMETER2 PARAMETER3 WAIT_CLASS_ID WAIT_CLASS# WAIT_CLASS
---------- ---------- ---------------------------------------- -------------------- -------------------- -------------------- ------------- ----------- --------------------
289 1646780882 library cache: mutex X idn value where 3875070507 4 Concurrency
--//P2=1 表示muext的第4-7,0-3位元組資訊,注意intel系列CPU大小頭問題.另外如果修改8個位元組,特別注意這個問題,比如:
SYS@book> oradebug poke 0x0000000080528f40 8 0x0000000100000002
BEFORE: [080528F40, 080528F48) = 00000000 00000000
AFTER: [080528F40, 080528F48) = 00000002 00000001
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000002 00000001 00000062 000486F7 000190FA 00000000
--//0x00000002在前,0x00000001在後
--//不知道為什麼我一直以為這樣的情況如果前面是00000002 00000000時不會阻塞,看來我搞錯了.
SYS@book> @ tpt/ash/ash_wait_chains BLOCKING_SESSION||','||BLOCKING_SESSION_SERIAL#||'@'||BLOCKING_INST_ID||'=>'||session_id||','||SESSION_SERIAL#||'@'||inst_id||'=>'||event 1=1 sysdate-1/1440 sysdate
-- Display ASH Wait Chain Signatures script v0.2 BETA by Tanel Poder ( http://blog.tanelpoder.com )
%This SECONDS AAS WAIT_CHAIN
------ ---------- ---------- ----------------------------------------------------------------------
32% 37 .6 -> ,@=>58,85@1=>
28% 32 .5 -> ,@=>295,5@1=>
23% 26 .4 -> ,@=>295,5@1=>library cache: mutex X
17% 19 .3 -> ,@=>58,85@1=>library cache: mutex X
1% 1 0 -> ,@=>44,23@1=>
5.知道這些第4-7位元組等於dump會話的sid,應該不會阻塞:
--//從測試可以看出mutex結構體,第0-3位元組表示ref count,第4-7位元組表示阻塞的持有的sid???
--//如果第3-7位元組等於執行會話sid,可以正常執行,有點奇怪為什麼是不出現ora-04024錯誤.也許這裡不是cursor的原因,或者mutex型別不一樣.
$ oerr ora 04024
04024, 00000, "self-deadlock detected while trying to mutex pin cursor %s"
// *Cause: While trying to mutex pin a cursor, a self-deadlock is detected.
// *Action: Retry the operation later.
SYS@book> oradebug poke 0x0000000080528f40 8 0x0000000000000000
BEFORE: [080528F40, 080528F48) = 00000002 00000001
AFTER: [080528F40, 080528F48) = 00000000 00000000
--//session 2:
SYS@book> @ spid
SID SERIAL# PROCESS SERVER SPID PID P_SERIAL# C50
---------- ---------- ------------------------ --------- ------ ------- ---------- --------------------------------------------------
44 23 9542 DEDICATED 9543 27 9 alter system kill session '44,23' immediate;
--//44 = 0x2c
SYS@book> oradebug poke 0x0000000080528f40 8 0x0000002c00000000
BEFORE: [080528F40, 080528F48) = 00000000 00000000
AFTER: [080528F40, 080528F48) = 00000000 0000002C
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 0000002C 00000062 000486F7 000190FA 00000000
SYS@book> oradebug dump library_cache 8;
Statement processed.
--//ok,,在sid=第3-7位元組內容的情況下,可以dump.
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000000 0000002C 00000062 000486F7 000190FA 00000000
--//檢查轉儲檔案內容.
Bucket: #=102650 Mutex=0x80528f40(0, 98, 296695, 0)
--//98 = 0x62,這樣的情況gets計數不增加,是否可以理解為獨佔持有該mutex.
LibraryHandle: Address=0x7d3c9600 Hash=5ab90fa LockMode=N PinMode=0 LoadLockMode=0 Status=VALD
ObjectName: Name=select * from dept where deptno=20
FullHashValue=e8ec445edab00042802d511305ab90fa Namespace=SQL AREA(00) Type=CURSOR(00) Identifier=95129850 OwnerIdn=83
Statistics: InvalidationCount=0 ExecutionCount=97 LoadCount=2 ActiveLocks=1 TotalLockCount=97 TotalPinCount=1
Counters: BrokenCount=1 RevocablePointer=1 KeepDependency=1 Version=0 BucketInUse=96 HandleInUse=96 HandleReferenceCount=0
Concurrency: DependencyMutex=0x7d3c96b0(0, 4, 0, 0) Mutex=0x7d3c9740(44, 444, 0, 6)
Flags=RON/PIN/TIM/PN0/DBN/[10012841]
WaitersLists:
Lock=0x7d3c9690[0x7d3c9690,0x7d3c9690]
Pin=0x7d3c9670[0x7d3c9670,0x7d3c9670]
LoadLock=0x7d3c96e8[0x7d3c96e8,0x7d3c96e8]
Timestamp: Current=09-14-2021 09:20:40
HandleReference: Address=0x7d3c97d0 Handle=(nil) Flags=[00]
ReferenceList:
Reference: Address=0x7d06d128 Handle=0x7d6c0cc8 Flags=ROD[21]
LibraryObject: Address=0x7d06ee18 HeapMask=0000-0001-0001-0000 Flags=EXS[0000] Flags2=[0000] PublicFlags=[0000]
DataBlocks:
Block: #='0' name=KGLH0^5ab90fa pins=0 Change=NONE
Heap=0x7c185b68 Pointer=0x7d06eeb8 Extent=0x7d06ed98 Flags=I/-/P/A/-/-
FreedLocation=0 Alloc=2.437500 Size=3.976562 LoadTime=22924574690
ChildTable: size='16'
Child: id='0' Table=0x7d06fcc8 Reference=0x7d06f708 Handle=0x7d5e6c68
NamespaceDump:
Parent Cursor: sql_id=80baj2c2ur47u parent=0x7d06eeb8 maxchild=1 plk=y ppn=n
SYS@book> oradebug poke 0x0000000080528f40 8 0x0000002c00000011
BEFORE: [080528F40, 080528F48) = 00000000 0000002C
AFTER: [080528F40, 080528F48) = 00000011 0000002C
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 00000011 0000002C 00000062 000486F7 000190FA 00000000
SYS@book> @ tix
New tracefile_identifier = /u01/app/oracle/diag/rdbms/book/book/trace/book_ora_9543_0001.trc
SYS@book> oradebug dump library_cache 8;
Statement processed.
--//檢查轉儲檔案內容.
Bucket: #=102650 Mutex=0x80528f40(11, 98, 296695, 0)
--//98 = 0x62,
--//11? 嗯這裡是16進位制嗎?正常應該顯示17才對啊,oracle怎麼這樣dump呢?
LibraryHandle: Address=0x7d3c9600 Hash=5ab90fa LockMode=N PinMode=0 LoadLockMode=0 Status=VALD
ObjectName: Name=select * from dept where deptno=20
FullHashValue=e8ec445edab00042802d511305ab90fa Namespace=SQL AREA(00) Type=CURSOR(00) Identifier=95129850 OwnerIdn=83
Statistics: InvalidationCount=0 ExecutionCount=97 LoadCount=2 ActiveLocks=1 TotalLockCount=97 TotalPinCount=1
Counters: BrokenCount=1 RevocablePointer=1 KeepDependency=1 Version=0 BucketInUse=96 HandleInUse=96 HandleReferenceCount=0
Concurrency: DependencyMutex=0x7d3c96b0(0, 5, 0, 0) Mutex=0x7d3c9740(44, 445, 0, 6)
Flags=RON/PIN/TIM/PN0/DBN/[10012841]
WaitersLists:
Lock=0x7d3c9690[0x7d3c9690,0x7d3c9690]
Pin=0x7d3c9670[0x7d3c9670,0x7d3c9670]
LoadLock=0x7d3c96e8[0x7d3c96e8,0x7d3c96e8]
Timestamp: Current=09-14-2021 09:20:40
HandleReference: Address=0x7d3c97d0 Handle=(nil) Flags=[00]
ReferenceList:
Reference: Address=0x7d06d128 Handle=0x7d6c0cc8 Flags=ROD[21]
LibraryObject: Address=0x7d06ee18 HeapMask=0000-0001-0001-0000 Flags=EXS[0000] Flags2=[0000] PublicFlags=[0000]
DataBlocks:
Block: #='0' name=KGLH0^5ab90fa pins=0 Change=NONE
Heap=0x7c185b68 Pointer=0x7d06eeb8 Extent=0x7d06ed98 Flags=I/-/P/A/-/-
FreedLocation=0 Alloc=2.437500 Size=3.976562 LoadTime=22924574690
ChildTable: size='16'
Child: id='0' Table=0x7d06fcc8 Reference=0x7d06f708 Handle=0x7d5e6c68
NamespaceDump:
Parent Cursor: sql_id=80baj2c2ur47u parent=0x7d06eeb8 maxchild=1 plk=y ppn=n
SYS@book> @ tix
New tracefile_identifier = /u01/app/oracle/diag/rdbms/book/book/trace/book_ora_9543_0002.trc
SYS@book> oradebug poke 0x0000000080528f40 4 0x000001ab
BEFORE: [080528F40, 080528F44) = 00000100
AFTER: [080528F40, 080528F44) = 000001AB
SYS@book> oradebug peek 0x80528f40 24
[080528F40, 080528F58) = 000001AB 0000002C 00000062 000486F7 000190FA 00000000
SYS@book> oradebug dump library_cache 8;
Statement processed.
--//檢查轉儲檔案內容.
Bucket: #=102650 Mutex=0x80528f40(1ab, 98, 296695, 0)
--//oracle這裡很奇葩,mutex的值第1個數字竟然是16進位制的.
LibraryHandle: Address=0x7d3c9600 Hash=5ab90fa LockMode=N PinMode=0 LoadLockMode=0 Status=VALD
ObjectName: Name=select * from dept where deptno=20
FullHashValue=e8ec445edab00042802d511305ab90fa Namespace=SQL AREA(00) Type=CURSOR(00) Identifier=95129850 OwnerIdn=83
Statistics: InvalidationCount=0 ExecutionCount=97 LoadCount=2 ActiveLocks=1 TotalLockCount=97 TotalPinCount=1
Counters: BrokenCount=1 RevocablePointer=1 KeepDependency=1 Version=0 BucketInUse=96 HandleInUse=96 HandleReferenceCount=0
Concurrency: DependencyMutex=0x7d3c96b0(0, 7, 0, 0) Mutex=0x7d3c9740(44, 447, 0, 6)
Flags=RON/PIN/TIM/PN0/DBN/[10012841]
WaitersLists:
Lock=0x7d3c9690[0x7d3c9690,0x7d3c9690]
Pin=0x7d3c9670[0x7d3c9670,0x7d3c9670]
LoadLock=0x7d3c96e8[0x7d3c96e8,0x7d3c96e8]
Timestamp: Current=09-14-2021 09:20:40
HandleReference: Address=0x7d3c97d0 Handle=(nil) Flags=[00]
ReferenceList:
Reference: Address=0x7d06d128 Handle=0x7d6c0cc8 Flags=ROD[21]
LibraryObject: Address=0x7d06ee18 HeapMask=0000-0001-0001-0000 Flags=EXS[0000] Flags2=[0000] PublicFlags=[0000]
DataBlocks:
Block: #='0' name=KGLH0^5ab90fa pins=0 Change=NONE
Heap=0x7c185b68 Pointer=0x7d06eeb8 Extent=0x7d06ed98 Flags=I/-/P/A/-/-
FreedLocation=0 Alloc=2.437500 Size=3.976562 LoadTime=22924574690
ChildTable: size='16'
Child: id='0' Table=0x7d06fcc8 Reference=0x7d06f708 Handle=0x7d5e6c68
NamespaceDump:
Parent Cursor: sql_id=80baj2c2ur47u parent=0x7d06eeb8 maxchild=1 plk=y ppn=n
6.總結:
--//至此可以知道dump library cache,mutex的值應該表示mutex結構體的第0-3,第8-11,第12-15,第20-23位元組.分別對應ref count,gets,sleep.???.
--//同樣道理mutex結構體24位元組裡面的資訊(各佔4個位元組)分別表示ref_count,sid,gets,sleeps,mutex標識(這裡對應bucket值,sql語句可能對應hash_value),???.
--//最後第20-23,我還是不知道表示什麼?
來自 “ ITPUB部落格 ” ，連結：http://blog.itpub.net/267265/viewspace-2791949/，如需轉載，請註明出處，否則將追究法律責任。
相關文章
- [20210914]探究mutex的值 4.txtMutex
- [20210903]探究mutex的值.txtMutex
- [20210916]探究mutex的值 8.txtMutex
- [20210915]探究mutex的值 6.txtMutex
- [20201203]探究library cache mutex X 3.txtMutex
- Swift列舉關聯值的記憶體探究Swift記憶體
- oracle mutexOracleMutex
- 'cursor:mutex ..'/ 'cursor:pin ..'/ 'library cache:mutex ..'型別的等待事件Mutex型別事件
- [20210520]11g shared pool latch與library cache mutex的簡單探究.txtMutex
- [20210521]11g shared pool latch與library cache mutex的簡單探究4.txtMutex
- 體驗mutexMutex
- Go語言的互斥鎖MutexGoMutex
- 深入理解Oracle中的MutexOracleMutex
- 如何查MUTEX的的持有者Mutex
- [20210520]11g shared pool latch與library cache mutex的簡單探究3.txtMutex
- Oracle：cursor:mutex XOracleMutex
- Oracle Mutex 等待事件OracleMutex事件
- mutex compare latchMutex
- 探究CRM系統能給企業帶來哪些價值？
- 互斥鎖mutex的簡單實現Mutex
- golang 中 sync.Mutex 的實現GolangMutex
- _gcry_ath_mutex_lock: Assertion `*lock == ((ath_mutex_t) 0)' failed.GCMutexAI
- 當 Go struct 遇上 MutexGoStructMutex
- cursor: mutex S等待事件Mutex事件
- C#互斥體——MutexC#Mutex
- 【轉】spin lock 和mutexMutex
- [20200401]優化的困惑5.txt優化
- 深度探究MMO社交對話系統（一）：聊天系統的進化與價值
- 啟用資料價值，探究DataOps下的資料架構及其實踐架構
- Go 中的鎖原始碼實現：MutexGo原始碼Mutex
- library cache: mutex X引發的故障Mutex
- 互斥鎖pthread_mutex_t的使用threadMutex
- IHSE KVM坐席管理系統應用案例價值探究與評價
- Go 系列教程 —— 25. MutexGoMutex
- spinlock和mutex選用方法Mutex
- [20160213]閉包傳遞5.txt
- GO: sync.Mutex 的實現與演進GoMutex
- golang中的Mutex設計原理詳解（一）GolangMutex