mariadb/mysys/my_pthread.c
Michael Widenius 32f81bab7d WL#3262 add mutex lock order checking to safemutex (also called safe_mutex_deadlock_detector)
This writes a warning on stderr if one uses mutex in different order,
like if one in one case would lock mutex in the order A,B and in another case
would lock mutex in the order B,A
This is inspired by and loosely based on the LOCKDEP patch by Jonas
Wrong mutex order is either fixed or mutex are marked with MYF_NO_DEADLOCK_DETECTION
if used inconsistently (need to be fixed by server team)

KNOWN_BUGS.txt:
  Added information that one need to dump and restore Maria tables
include/hash.h:
  Added prototype function for walking over all elements in a hash
include/my_pthread.h:
  Added my_pthread_mutex_init() and my_pthread_mutex_lock(); These should be used if one wants to disable mutex order checking.
  Changed names of the nonposix mutex_init functions to not conflict with my_phread_mutex_init()
  Added and extended structures for mutex deadlock detection.
  New arguments to sage_mutex_init() and safe_mutex_lock() to allow one to disable mutex order checking.
  Added variable 'safe_mutex_deadlock_detector' to enable/disable deadlock detection for all pthread_mutex_init()
mysys/Makefile.am:
  Added cleaning of test files
  Added test_thr_mutex
mysys/hash.c:
  Added hash_iterate() to iterate over all elements in a hash
  More comments
mysys/my_init.c:
  Added calls to destory all mutex uses by mysys()
  Added waiting for threads to end before calling TERMINATE() to list not freed memory
mysys/my_pthread.c:
  Changed names to free my_pthread_mutex_init() for mutex-lock-order-checking
mysys/my_sleep.c:
  Fixed too long wait if using 1000000L as argument
mysys/my_thr_init.c:
  Mark THR_LOCK_threads and THR_LOCK_malloc to not have mutex deadlock detection.
  (We can't have it enabled for this as these are internal mutex used by the detector  
  Call my_thread_init() early as we need thread specific variables enabled for the following pthread_mutex_init()
  Move code to wait for threads to end to my_wait_for_other_threads_to_die()
  Don't destroy mutex and conditions unless all threads have died
  Added my_thread_destroy_mutex() to destroy all mutex used by the mysys thread system
  Name the thread specific mutex as "mysys_var->mutex"
  Added my_thread_var_mutex_in_use() to return pointer to mutex in use or 0 if thread variables are not initialized
mysys/mysys_priv.h:
  Added prototypes for functions used internally with mutex-wrong-usage detection
mysys/thr_mutex.c:
  Added runtime detection of mutex used in conflicting order
  See WL#3262 or test_thr_mutex.c for examples
  The base idea is for each mutex have two hashes:
  - mutex->locked_mutex points to all mutex used after this one
  - mutex->used_mutex points to all mutex which has this mutex in it's mutex->locked_mutex
  There is a wrong mutex order if any mutex currently locked before this mutex is in the mutex->locked_mutex hash
sql/event_queue.cc:
  Mark mutex used inconsistently (need to be fixed by server team)
sql/event_scheduler.cc:
  Declare the right order to take the mutex
sql/events.cc:
  Mark mutex used inconsistently (need to be fixed by server team)
sql/ha_ndbcluster_binlog.cc:
  Mark mutex used inconsistently (need to be fixed by server team)
sql/log.cc:
  Mark mutex used inconsistently (need to be fixed by server team)
sql/mysqld.cc:
  Use pthread_mutex_trylock instead of pthread_mutex_unlock() when sending kill signal to thread
  This is needed to avoid wrong mutex order as normally one takes 'current_mutex' before mysys_var->mutex.
  Added call to free sp cache.
  Add destruction of LOCK_server_started and COND_server_started.
  Added register_mutex_order() function to register in which order mutex should be taken
  (to initiailize mutex_deadlock_detector).
  Added option to turn off safe_mutex_deadlock_detector
sql/protocol.cc:
  Fixed wrong argument to DBUG_PRINT (found by valgrind)
sql/rpl_mi.cc:
  Mark mutex used inconsistently (need to be fixed by server team)
sql/set_var.cc:
  Remove wrong locking of LOCK_global_system_variables when reading and setting log variables
  (would cause inconsistent mutex order).
  Update global variables outside of logger.unlock() as LOCK_global_system_variables has to be taken before logger locks
  Reviewed by gluh
sql/sp_cache.cc:
  Added function to destroy mutex used by sp cache
sql/sp_cache.h:
  Added function to destroy mutex used by sp cache
sql/sql_class.cc:
  Use pthread_mutex_trylock instead of pthread_mutex_unlock() when sending kill signal to thread
  This is needed to avoid wrong mutex order as normally one takes 'current_mutex' before mysys_var->mutex.
  Register order in which LOCK_delete and mysys_var->mutex is taken
sql/sql_insert.cc:
  Give a name for Delayed_insert::mutex
  Mark mutex used inconsistently (need to be fixed by server team)
  Move closing of tables outside of di->mutex (to avoid wrong mutex order)
sql/sql_show.cc:
  Don't keep LOCK_global_system_variables locked over value->show_type() as this leads to wrong mutex order
storage/innobase/handler/ha_innodb.cc:
  Disable safe_muted_deadlock_detector for innobase intern mutex (to speed up page cache initialization)
storage/maria/ha_maria.cc:
  Added flag to ha_maria::info() to signal if we need to lock table share or not.
  This is needed to avoid locking mutex in wrong order
storage/maria/ha_maria.h:
  Added flag to ha_maria::info() to signal if we need to lock table share or not.
storage/maria/ma_close.c:
  Destroy key_del_lock
  Simplify freeing ftparser_param
storage/maria/ma_key.c:
  Better comment
storage/maria/ma_loghandler.c:
  Mark mutex used inconsistently (need to be fixed by sanja)
storage/maria/ma_state.c:
  More comments
storage/maria/ma_test1.c:
  Ensure that safe_mutex_deadlock_detector is always on (should be, this is just for safety)
storage/maria/ma_test2.c:
  Ensure that safe_mutex_deadlock_detector is always on (should be, this is just for safety)
2008-12-03 00:02:52 +02:00

547 lines
14 KiB
C

/* Copyright (C) 2000-2003 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 to get threads more portable */
#define DONT_REMAP_PTHREAD_FUNCTIONS
#include "mysys_priv.h"
#ifdef THREAD
#include <signal.h>
#include <m_string.h>
#include <thr_alarm.h>
#if (defined(__BSD__) || defined(_BSDI_VERSION))
#define SCHED_POLICY SCHED_RR
#else
#define SCHED_POLICY SCHED_OTHER
#endif
uint thd_lib_detected= 0;
#ifndef my_pthread_setprio
void my_pthread_setprio(pthread_t thread_id,int prior)
{
#ifdef HAVE_PTHREAD_SETSCHEDPARAM
struct sched_param tmp_sched_param;
bzero((char*) &tmp_sched_param,sizeof(tmp_sched_param));
tmp_sched_param.sched_priority=prior;
VOID(pthread_setschedparam(thread_id,SCHED_POLICY,&tmp_sched_param));
#endif
}
#endif
#ifndef my_pthread_getprio
int my_pthread_getprio(pthread_t thread_id)
{
#ifdef HAVE_PTHREAD_SETSCHEDPARAM
struct sched_param tmp_sched_param;
int policy;
if (!pthread_getschedparam(thread_id,&policy,&tmp_sched_param))
{
return tmp_sched_param.sched_priority;
}
#endif
return -1;
}
#endif
#ifndef my_pthread_attr_setprio
void my_pthread_attr_setprio(pthread_attr_t *attr, int priority)
{
#ifdef HAVE_PTHREAD_SETSCHEDPARAM
struct sched_param tmp_sched_param;
bzero((char*) &tmp_sched_param,sizeof(tmp_sched_param));
tmp_sched_param.sched_priority=priority;
VOID(pthread_attr_setschedparam(attr,&tmp_sched_param));
#endif
}
#endif
/* To allow use of pthread_getspecific with two arguments */
#ifdef HAVE_NONPOSIX_PTHREAD_GETSPECIFIC
#undef pthread_getspecific
void *my_pthread_getspecific_imp(pthread_key_t key)
{
void *value;
if (pthread_getspecific(key,(void *) &value))
return 0;
return value;
}
#endif
#ifdef __NETWARE__
/*
Don't kill the LibC Reaper thread or the main thread
*/
#include <nks/thread.h>
#undef pthread_exit
void my_pthread_exit(void *status)
{
NXThreadId_t tid;
NXContext_t ctx;
char name[NX_MAX_OBJECT_NAME_LEN+1] = "";
tid= NXThreadGetId();
if (tid == NX_INVALID_THREAD_ID || !tid)
return;
if (NXThreadGetContext(tid, &ctx) ||
NXContextGetName(ctx, name, sizeof(name)-1))
return;
/*
"MYSQLD.NLM's LibC Reaper" or "MYSQLD.NLM's main thread"
with a debug build of LibC the reaper can have different names
*/
if (!strindex(name, "\'s"))
pthread_exit(status);
}
#endif
/*
Some functions for RTS threads, AIX, Siemens Unix and UnixWare 7
(and DEC OSF/1 3.2 too)
*/
int my_pthread_create_detached=1;
#if defined(HAVE_NONPOSIX_SIGWAIT) || defined(HAVE_DEC_3_2_THREADS)
int my_sigwait(const sigset_t *set,int *sig)
{
int signal=sigwait((sigset_t*) set);
if (signal < 0)
return errno;
*sig=signal;
return 0;
}
#endif
/* localtime_r for SCO 3.2V4.2 */
#if !defined(HAVE_LOCALTIME_R) || !defined(HAVE_GMTIME_R)
extern pthread_mutex_t LOCK_localtime_r;
#endif
#if !defined(HAVE_LOCALTIME_R)
struct tm *localtime_r(const time_t *clock, struct tm *res)
{
struct tm *tmp;
pthread_mutex_lock(&LOCK_localtime_r);
tmp=localtime(clock);
*res= *tmp;
pthread_mutex_unlock(&LOCK_localtime_r);
return res;
}
#endif
#if !defined(HAVE_GMTIME_R)
/*
Reentrant version of standard gmtime() function.
Needed on some systems which don't implement it.
*/
struct tm *gmtime_r(const time_t *clock, struct tm *res)
{
struct tm *tmp;
pthread_mutex_lock(&LOCK_localtime_r);
tmp= gmtime(clock);
*res= *tmp;
pthread_mutex_unlock(&LOCK_localtime_r);
return res;
}
#endif
/****************************************************************************
** Replacement of sigwait if the system doesn't have one (like BSDI 3.0)
**
** Note:
** This version of sigwait() is assumed to called in a loop so the signalmask
** is permanently modified to reflect the signal set. This is done to get
** a much faster implementation.
**
** This implementation isn't thread safe: It assumes that only one
** thread is using sigwait.
**
** If one later supplies a different signal mask, all old signals that
** was used before are unblocked and set to SIGDFL.
**
** Author: Gary Wisniewski <garyw@spidereye.com.au>, much modified by Monty
****************************************************************************/
#if !defined(HAVE_SIGWAIT) && !defined(sigwait) && !defined(__WIN__) && !defined(HAVE_rts_threads) && !defined(HAVE_NONPOSIX_SIGWAIT) && !defined(HAVE_DEC_3_2_THREADS)
#if !defined(DONT_USE_SIGSUSPEND)
static sigset_t sigwait_set,rev_sigwait_set,px_recd;
void px_handle_sig(int sig)
{
sigaddset(&px_recd, sig);
}
void sigwait_setup(sigset_t *set)
{
int i;
struct sigaction sact,sact1;
sigset_t unblock_mask;
sact.sa_flags = 0;
sact.sa_handler = px_handle_sig;
memcpy_fixed(&sact.sa_mask,set,sizeof(*set)); /* handler isn't thread_safe */
sigemptyset(&unblock_mask);
pthread_sigmask(SIG_UNBLOCK,(sigset_t*) 0,&rev_sigwait_set);
for (i = 1; i <= sizeof(sigwait_set)*8; i++)
{
if (sigismember(set,i))
{
sigdelset(&rev_sigwait_set,i);
if (!sigismember(&sigwait_set,i))
sigaction(i, &sact, (struct sigaction*) 0);
}
else
{
sigdelset(&px_recd,i); /* Don't handle this */
if (sigismember(&sigwait_set,i))
{ /* Remove the old handler */
sigaddset(&unblock_mask,i);
sigdelset(&rev_sigwait_set,i);
sact1.sa_flags = 0;
sact1.sa_handler = SIG_DFL;
sigemptyset(&sact1.sa_mask);
sigaction(i, &sact1, 0);
}
}
}
memcpy_fixed(&sigwait_set,set,sizeof(*set));
pthread_sigmask(SIG_BLOCK,(sigset_t*) set,(sigset_t*) 0);
pthread_sigmask(SIG_UNBLOCK,&unblock_mask,(sigset_t*) 0);
}
int sigwait(sigset_t *setp, int *sigp)
{
if (memcmp(setp,&sigwait_set,sizeof(sigwait_set)))
sigwait_setup(setp); /* Init or change of set */
for (;;)
{
/*
This is a fast, not 100% portable implementation to find the signal.
Because the handler is blocked there should be at most 1 bit set, but
the specification on this is somewhat shady so we use a set instead a
single variable.
*/
ulong *ptr= (ulong*) &px_recd;
ulong *end=ptr+sizeof(px_recd)/sizeof(ulong);
for ( ; ptr != end ; ptr++)
{
if (*ptr)
{
ulong set= *ptr;
int found= (int) ((char*) ptr - (char*) &px_recd)*8+1;
while (!(set & 1))
{
found++;
set>>=1;
}
*sigp=found;
sigdelset(&px_recd,found);
return 0;
}
}
sigsuspend(&rev_sigwait_set);
}
return 0;
}
#else /* !DONT_USE_SIGSUSPEND */
/****************************************************************************
** Replacement of sigwait if the system doesn't have one (like BSDI 3.0)
**
** Note:
** This version of sigwait() is assumed to called in a loop so the signalmask
** is permanently modified to reflect the signal set. This is done to get
** a much faster implementation.
**
** This implementation uses a extra thread to handle the signals and one
** must always call sigwait() with the same signal mask!
**
** BSDI 3.0 NOTE:
**
** pthread_kill() doesn't work on a thread in a select() or sleep() loop?
** After adding the sleep to sigwait_thread, all signals are checked and
** delivered every second. This isn't that terrible performance vice, but
** someone should report this to BSDI and ask for a fix!
** Another problem is that when the sleep() ends, every select() in other
** threads are interrupted!
****************************************************************************/
static sigset_t pending_set;
static bool inited=0;
static pthread_cond_t COND_sigwait;
static pthread_mutex_t LOCK_sigwait;
void sigwait_handle_sig(int sig)
{
pthread_mutex_lock(&LOCK_sigwait);
sigaddset(&pending_set, sig);
VOID(pthread_cond_signal(&COND_sigwait)); /* inform sigwait() about signal */
pthread_mutex_unlock(&LOCK_sigwait);
}
void *sigwait_thread(void *set_arg)
{
sigset_t *set=(sigset_t*) set_arg;
int i;
struct sigaction sact;
sact.sa_flags = 0;
sact.sa_handler = sigwait_handle_sig;
memcpy_fixed(&sact.sa_mask,set,sizeof(*set)); /* handler isn't thread_safe */
sigemptyset(&pending_set);
for (i = 1; i <= sizeof(pending_set)*8; i++)
{
if (sigismember(set,i))
{
sigaction(i, &sact, (struct sigaction*) 0);
}
}
/* Ensure that init_thr_alarm() is called */
DBUG_ASSERT(thr_client_alarm);
sigaddset(set, thr_client_alarm);
pthread_sigmask(SIG_UNBLOCK,(sigset_t*) set,(sigset_t*) 0);
alarm_thread=pthread_self(); /* For thr_alarm */
for (;;)
{ /* Wait for signals */
#ifdef HAVE_NOT_BROKEN_SELECT
fd_set fd;
FD_ZERO(&fd);
select(0,&fd,0,0,0);
#else
sleep(1); /* Because of broken BSDI */
#endif
}
}
int sigwait(sigset_t *setp, int *sigp)
{
if (!inited)
{
pthread_attr_t thr_attr;
pthread_t sigwait_thread_id;
inited=1;
sigemptyset(&pending_set);
pthread_mutex_init(&LOCK_sigwait,MY_MUTEX_INIT_FAST);
pthread_cond_init(&COND_sigwait,NULL);
pthread_attr_init(&thr_attr);
pthread_attr_setscope(&thr_attr,PTHREAD_SCOPE_PROCESS);
pthread_attr_setdetachstate(&thr_attr,PTHREAD_CREATE_DETACHED);
pthread_attr_setstacksize(&thr_attr,8196);
my_pthread_attr_setprio(&thr_attr,100); /* Very high priority */
VOID(pthread_create(&sigwait_thread_id,&thr_attr,sigwait_thread,setp));
VOID(pthread_attr_destroy(&thr_attr));
}
pthread_mutex_lock(&LOCK_sigwait);
for (;;)
{
ulong *ptr= (ulong*) &pending_set;
ulong *end=ptr+sizeof(pending_set)/sizeof(ulong);
for ( ; ptr != end ; ptr++)
{
if (*ptr)
{
ulong set= *ptr;
int found= (int) ((char*) ptr - (char*) &pending_set)*8+1;
while (!(set & 1))
{
found++;
set>>=1;
}
*sigp=found;
sigdelset(&pending_set,found);
pthread_mutex_unlock(&LOCK_sigwait);
return 0;
}
}
VOID(pthread_cond_wait(&COND_sigwait,&LOCK_sigwait));
}
return 0;
}
#endif /* DONT_USE_SIGSUSPEND */
#endif /* HAVE_SIGWAIT */
/****************************************************************************
The following functions fixes that all pthread functions should work
according to latest posix standard
****************************************************************************/
/* Undefined wrappers set my_pthread.h so that we call os functions */
#undef pthread_mutex_init
#undef pthread_mutex_lock
#undef pthread_mutex_unlock
#undef pthread_mutex_destroy
#undef pthread_mutex_wait
#undef pthread_mutex_timedwait
#undef pthread_mutex_trylock
#undef pthread_mutex_t
#undef pthread_cond_init
#undef pthread_cond_wait
#undef pthread_cond_timedwait
#undef pthread_cond_t
#undef pthread_attr_getstacksize
/*****************************************************************************
** Patches for AIX and DEC OSF/1 3.2
*****************************************************************************/
#if defined(HAVE_NONPOSIX_PTHREAD_MUTEX_INIT)
#include <netdb.h>
int my_pthread_mutex_noposix_init(pthread_mutex_t *mp,
const pthread_mutexattr_t *attr)
{
int error;
if (!attr)
error=pthread_mutex_init(mp,pthread_mutexattr_default);
else
error=pthread_mutex_init(mp,*attr);
return error;
}
int my_pthread_cond_noposix_init(pthread_cond_t *mp,
const pthread_condattr_t *attr)
{
int error;
if (!attr)
error=pthread_cond_init(mp,pthread_condattr_default);
else
error=pthread_cond_init(mp,*attr);
return error;
}
#endif
/*****************************************************************************
Patches for HPUX
We need these because the pthread_mutex.. code returns -1 on error,
instead of the error code.
Note that currently we only remap pthread_ functions used by MySQL.
If we are depending on the value for some other pthread_xxx functions,
this has to be added here.
****************************************************************************/
#if defined(HPUX10) || defined(HAVE_BROKEN_PTHREAD_COND_TIMEDWAIT)
int my_pthread_cond_timedwait(pthread_cond_t *cond, pthread_mutex_t *mutex,
struct timespec *abstime)
{
int error=pthread_cond_timedwait(cond, mutex, abstime);
if (error == -1) /* Safety if the lib is fixed */
{
if (!(error=errno))
error= ETIMEDOUT; /* Can happen on HPUX */
}
if (error == EAGAIN) /* Correct errno to Posix */
error= ETIMEDOUT;
return error;
}
#endif
#if defined(HPUX10)
void my_pthread_attr_getstacksize(pthread_attr_t *connection_attrib,
size_t *stack_size)
{
*stack_size= pthread_attr_getstacksize(*connection_attrib);
}
#endif
#ifdef HAVE_POSIX1003_4a_MUTEX
/*
In HP-UX-10.20 and other old Posix 1003.4a Draft 4 implementations
pthread_mutex_trylock returns 1 on success, not 0 like
pthread_mutex_lock
From the HP-UX-10.20 man page:
RETURN VALUES
If the function fails, errno may be set to one of the following
values:
Return | Error | Description
_______|__________|_________________________________________
1 | | Successful completion.
0 | | The mutex is locked; therefore, it was
| | not acquired.
-1 | [EINVAL] | The value specified by mutex is invalid.
*/
/*
Convert pthread_mutex_trylock to return values according to latest POSIX
RETURN VALUES
0 If we are able successfully lock the mutex.
EBUSY Mutex was locked by another thread
# Other error number returned by pthread_mutex_trylock()
(Not likely)
*/
int my_pthread_mutex_trylock(pthread_mutex_t *mutex)
{
int error= pthread_mutex_trylock(mutex);
if (error == 1)
return 0; /* Got lock on mutex */
if (error == 0) /* Someon else is locking mutex */
return EBUSY;
if (error == -1) /* Safety if the lib is fixed */
error= errno; /* Probably invalid parameter */
return error;
}
#endif /* HAVE_POSIX1003_4a_MUTEX */
/* Some help functions */
int pthread_no_free(void *not_used __attribute__((unused)))
{
return 0;
}
int pthread_dummy(int ret)
{
return ret;
}
#endif /* THREAD */