d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
824a5e96dc
freebsd-dev/lib/libthr/thread/thr_rwlock.c

474 lines
12 KiB
C
Raw Normal View History

- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
/*-
* Copyright (c) 1998 Alex Nash
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
* Copyright (c) 2004 Michael Telahun Makonnen
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*/
#include <errno.h>
#include <limits.h>
#include <stdlib.h>
#include <pthread.h>
#include "thr_private.h"
/* maximum number of times a read lock may be obtained */
#define MAX_READ_LOCKS (INT_MAX - 1)
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
/*
* For distinguishing operations on read and write locks.
*/
enum rwlock_type {RWT_READ, RWT_WRITE};
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
/* Support for staticaly initialized mutexes. */
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
static struct umtx init_lock = UMTX_INITIALIZER;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
__weak_reference(_pthread_rwlock_destroy, pthread_rwlock_destroy);
__weak_reference(_pthread_rwlock_init, pthread_rwlock_init);
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
__weak_reference(_pthread_rwlock_rdlock, pthread_rwlock_rdlock);
__weak_reference(_pthread_rwlock_timedrdlock, pthread_rwlock_timedrdlock);
__weak_reference(_pthread_rwlock_timedwrlock, pthread_rwlock_timedwrlock);
__weak_reference(_pthread_rwlock_tryrdlock, pthread_rwlock_tryrdlock);
__weak_reference(_pthread_rwlock_trywrlock, pthread_rwlock_trywrlock);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
__weak_reference(_pthread_rwlock_unlock, pthread_rwlock_unlock);
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
__weak_reference(_pthread_rwlock_wrlock, pthread_rwlock_wrlock);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
static int insert_rwlock(struct pthread_rwlock *, enum rwlock_type);
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
static int rwlock_init_static(struct pthread_rwlock **rwlock);
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
static int rwlock_rdlock_common(pthread_rwlock_t *, int,
const struct timespec *);
static int rwlock_wrlock_common(pthread_rwlock_t *, int,
const struct timespec *);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
int
_pthread_rwlock_destroy (pthread_rwlock_t *rwlock)
{
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
pthread_rwlock_t prwlock;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (rwlock == NULL || *rwlock == NULL)
return (EINVAL);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
prwlock = *rwlock;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (prwlock->state != 0)
return (EBUSY);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
pthread_mutex_destroy(&prwlock->lock);
pthread_cond_destroy(&prwlock->read_signal);
pthread_cond_destroy(&prwlock->write_signal);
free(prwlock);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
*rwlock = NULL;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
return (0);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
}
int
_pthread_rwlock_init (pthread_rwlock_t *rwlock, const pthread_rwlockattr_t *attr)
{
pthread_rwlock_t prwlock;
int ret;
/* allocate rwlock object */
prwlock = (pthread_rwlock_t)malloc(sizeof(struct pthread_rwlock));
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
if (prwlock == NULL) {
ret = ENOMEM;
goto out;
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
/* initialize the lock */
if ((ret = pthread_mutex_init(&prwlock->lock, NULL)) != 0)
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
goto out;
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
/* initialize the read condition signal */
if ((ret = pthread_cond_init(&prwlock->read_signal, NULL)) != 0)
goto out_readcond;
/* initialize the write condition signal */
if ((ret = pthread_cond_init(&prwlock->write_signal, NULL)) != 0)
goto out_writecond;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
/* success */
prwlock->state = 0;
prwlock->blocked_writers = 0;
*rwlock = prwlock;
return (0);
out_writecond:
pthread_cond_destroy(&prwlock->read_signal);
out_readcond:
pthread_mutex_destroy(&prwlock->lock);
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
out:
if (prwlock != NULL)
free(prwlock);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
return(ret);
}
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
/*
* If nonblocking is 0 this function will wait on the lock. If
* it is greater than 0 it will return immediately with EBUSY.
*/
static int
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
rwlock_rdlock_common(pthread_rwlock_t *rwlock, int nonblocking,
const struct timespec *timeout)
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
{
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
struct rwlock_held *rh;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
pthread_rwlock_t prwlock;
int ret;
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
rh = NULL;
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
if (rwlock == NULL)
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
return(EINVAL);
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
/*
* Check for validity of the timeout parameter.
*/
if (timeout != NULL &&
(timeout->tv_nsec < 0 || timeout->tv_nsec >= 1000000000))
return (EINVAL);
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
if ((ret = rwlock_init_static(rwlock)) !=0 )
return (ret);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
prwlock = *rwlock;
/* grab the monitor lock */
if ((ret = pthread_mutex_lock(&prwlock->lock)) != 0)
return(ret);
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
/* check lock count */
if (prwlock->state == MAX_READ_LOCKS) {
pthread_mutex_unlock(&prwlock->lock);
return (EAGAIN);
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
/* give writers priority over readers */
while (prwlock->blocked_writers || prwlock->state < 0) {
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (nonblocking) {
pthread_mutex_unlock(&prwlock->lock);
return (EBUSY);
}
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
/*
* If this lock is already held for writing we have
* a deadlock situation.
*/
if (curthread->rwlockList != NULL && prwlock->state < 0) {
LIST_FOREACH(rh, curthread->rwlockList, rh_link) {
if (rh->rh_rwlock == prwlock &&
rh->rh_wrcount > 0) {
pthread_mutex_unlock(&prwlock->lock);
return (EDEADLK);
}
}
}
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
if (timeout == NULL)
ret = pthread_cond_wait(&prwlock->read_signal,
&prwlock->lock);
else
ret = pthread_cond_timedwait(&prwlock->read_signal,
&prwlock->lock, timeout);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (ret != 0 && ret != EINTR) {
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
/* can't do a whole lot if this fails */
pthread_mutex_unlock(&prwlock->lock);
return(ret);
}
}
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
++prwlock->state; /* indicate we are locked for reading */
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
ret = insert_rwlock(prwlock, RWT_READ);
if (ret != 0) {
pthread_mutex_unlock(&prwlock->lock);
return (ret);
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
/*
* Something is really wrong if this call fails. Returning
* error won't do because we've already obtained the read
* lock. Decrementing 'state' is no good because we probably
* don't have the monitor lock.
*/
pthread_mutex_unlock(&prwlock->lock);
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
return(0);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
}
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
int
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
_pthread_rwlock_rdlock (pthread_rwlock_t *rwlock)
{
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
return (rwlock_rdlock_common(rwlock, 0, NULL));
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
}
int
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
_pthread_rwlock_timedrdlock(pthread_rwlock_t *rwlock,
const struct timespec *timeout)
{
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
return (rwlock_rdlock_common(rwlock, 0, timeout));
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
}
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
int
_pthread_rwlock_tryrdlock (pthread_rwlock_t *rwlock)
{
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
return (rwlock_rdlock_common(rwlock, 1, NULL));
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
int
_pthread_rwlock_unlock (pthread_rwlock_t *rwlock)
{
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
struct rwlock_held *rh;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
pthread_rwlock_t prwlock;
int ret;
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
rh = NULL;
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (rwlock == NULL || *rwlock == NULL)
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
return(EINVAL);
prwlock = *rwlock;
/* grab the monitor lock */
if ((ret = pthread_mutex_lock(&prwlock->lock)) != 0)
return(ret);
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
if (curthread->rwlockList != NULL) {
LIST_FOREACH(rh, curthread->rwlockList, rh_link) {
if (rh->rh_rwlock == prwlock)
break;
}
}
if (rh == NULL) {
ret = EPERM;
goto out;
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
if (prwlock->state > 0) {
o Assertions to catch that stuff that shouldn't happen is not happening. o In the rwlock code: move a duplicated check inside an if..else to after the if...else clause. o When initializing a static rwlock move the initialization check inside the lock. o In thr_setschedparam.c: When breaking out of the trylock...retry if busy loop make sure to reset the mtx pointer to null if the mutex is nolonger in a queue.
2004-07-30 17:13:00 +00:00
PTHREAD_ASSERT(rh->rh_wrcount == 0,
"write count on a readlock should be zero!");
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
rh->rh_rdcount--;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
if (--prwlock->state == 0 && prwlock->blocked_writers)
ret = pthread_cond_signal(&prwlock->write_signal);
} else if (prwlock->state < 0) {
o Assertions to catch that stuff that shouldn't happen is not happening. o In the rwlock code: move a duplicated check inside an if..else to after the if...else clause. o When initializing a static rwlock move the initialization check inside the lock. o In thr_setschedparam.c: When breaking out of the trylock...retry if busy loop make sure to reset the mtx pointer to null if the mutex is nolonger in a queue.
2004-07-30 17:13:00 +00:00
PTHREAD_ASSERT(rh->rh_rdcount == 0,
"read count on a writelock should be zero!");
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
rh->rh_wrcount--;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
prwlock->state = 0;
if (prwlock->blocked_writers)
ret = pthread_cond_signal(&prwlock->write_signal);
else
ret = pthread_cond_broadcast(&prwlock->read_signal);
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
} else {
/*
* No thread holds this lock. We should never get here.
*/
PTHREAD_ASSERT(0, "state=0 on read-write lock held by thread");
ret = EPERM;
goto out;
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
}
o Assertions to catch that stuff that shouldn't happen is not happening. o In the rwlock code: move a duplicated check inside an if..else to after the if...else clause. o When initializing a static rwlock move the initialization check inside the lock. o In thr_setschedparam.c: When breaking out of the trylock...retry if busy loop make sure to reset the mtx pointer to null if the mutex is nolonger in a queue.
2004-07-30 17:13:00 +00:00
if (rh->rh_wrcount == 0 && rh->rh_rdcount == 0) {
LIST_REMOVE(rh, rh_link);
free(rh);
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
out:
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
/* see the comment on this in rwlock_rdlock_common */
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
pthread_mutex_unlock(&prwlock->lock);
return(ret);
}
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
int
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
_pthread_rwlock_wrlock (pthread_rwlock_t *rwlock)
{
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
return (rwlock_wrlock_common(rwlock, 0, NULL));
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
}
int
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
_pthread_rwlock_timedwrlock (pthread_rwlock_t *rwlock,
const struct timespec *timeout)
{
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
return (rwlock_wrlock_common(rwlock, 0, timeout));
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
}
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
int
_pthread_rwlock_trywrlock (pthread_rwlock_t *rwlock)
{
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
return (rwlock_wrlock_common(rwlock, 1, NULL));
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
}
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
/*
* If nonblocking is 0 this function will wait on the lock. If
* it is greater than 0 it will return immediately with EBUSY.
*/
static int
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
rwlock_wrlock_common(pthread_rwlock_t *rwlock, int nonblocking,
const struct timespec *timeout)
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
{
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
struct rwlock_held *rh;
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
pthread_rwlock_t prwlock;
int ret;
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
rh = NULL;
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
if (rwlock == NULL)
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
return(EINVAL);
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
/*
* Check the timeout value for validity.
*/
if (timeout != NULL &&
(timeout->tv_nsec < 0 || timeout->tv_nsec >= 1000000000))
return (EINVAL);
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
if ((ret = rwlock_init_static(rwlock)) !=0 )
return (ret);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
prwlock = *rwlock;
/* grab the monitor lock */
if ((ret = pthread_mutex_lock(&prwlock->lock)) != 0)
return(ret);
while (prwlock->state != 0) {
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (nonblocking) {
pthread_mutex_unlock(&prwlock->lock);
return (EBUSY);
}
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
/*
* If this thread already holds the lock for reading
* or writing we have a deadlock situation.
*/
if (curthread->rwlockList != NULL) {
LIST_FOREACH(rh, curthread->rwlockList, rh_link) {
o Assertions to catch that stuff that shouldn't happen is not happening. o In the rwlock code: move a duplicated check inside an if..else to after the if...else clause. o When initializing a static rwlock move the initialization check inside the lock. o In thr_setschedparam.c: When breaking out of the trylock...retry if busy loop make sure to reset the mtx pointer to null if the mutex is nolonger in a queue.
2004-07-30 17:13:00 +00:00
if (rh->rh_rwlock == prwlock) {
PTHREAD_ASSERT((rh->rh_rdcount > 0 ||
rh->rh_wrcount > 0),
"Invalid 0 R/RW count!");
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
pthread_mutex_unlock(&prwlock->lock);
return (EDEADLK);
break;
}
}
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
++prwlock->blocked_writers;
Add an implementation of pthread_rwlock_timed{rd,wr}lock() to libthr with attendant documentation.
2004-01-16 10:52:10 +00:00
if (timeout == NULL)
ret = pthread_cond_wait(&prwlock->write_signal,
&prwlock->lock);
else
ret = pthread_cond_timedwait(&prwlock->write_signal,
&prwlock->lock, timeout);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
if (ret != 0 && ret != EINTR) {
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
--prwlock->blocked_writers;
pthread_mutex_unlock(&prwlock->lock);
return(ret);
}
--prwlock->blocked_writers;
}
/* indicate we are locked for writing */
prwlock->state = -1;
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
ret = insert_rwlock(prwlock, RWT_WRITE);
if (ret != 0) {
pthread_mutex_unlock(&prwlock->lock);
return (ret);
}
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
/* see the comment on this in pthread_rwlock_rdlock */
pthread_mutex_unlock(&prwlock->lock);
o We are not required to initialize an invalid rwlock. So axe all that code and simply return EINVAL (which is allowed by the standard) in all those pthread functions that previously initialized it. o Refactor the pthread_rwlock_[try]rdlock() and pthread_rwlock_[try]wrlock() functions. They are now completeley condensed into rwlock_rdlock_common() and rwlock_wrlock_common(), respectively. o If the application tries to destroy an rwlock that is currently held by a thread return EBUSY where it previously went ahead and freed all resources associated with the lock. o Refactor _pthread_rwlock_init() to make it look (relatively) sane. o When obtaining a read lock on an rwlock the check for whether it would exceed the maximum allowed read locks should happen *before* we obtain the lock. o The pthread_rwlock_* functions shall *never* return EINTR, so make sure to requeue/resuspend the thread if it encounters such an error. o Make a note that pthread_rwlock_unlock() needs to ensure it holds a lock on an rwlock it tries to unlock. It will be implemented in a separate commit because it requires some additional rwlock infrastructure.
2004-01-16 07:10:30 +00:00
return(0);
- Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code.
2003-04-01 03:46:29 +00:00
}
Implement reference counting of read-write locks. This uses a list in the thread structure to keep track of the locks and how many times they have been locked. This list is checked on every lock and unlock. The traversal through the list is O(n). Most applications don't hold so many locks at once that this will become a problem. However, if it does become a problem it might be a good idea to review this once libthr is off probation and in the optimization cycle. This fixes: o deadlock when a thread tries to recursively acquire a read lock when a writer is waiting on the lock. o a thread could previously successfully unlock a lock it did not own o deadlock when a thread tries to acquire a write lock on a lock it already owns for reading or writing [ this is admittedly not required by POSIX, but is nice to have ]
2004-01-19 14:51:45 +00:00
static int
insert_rwlock(struct pthread_rwlock *prwlock, enum rwlock_type rwt)
{
struct rwlock_held *rh;
/*
* Initialize the rwlock list in the thread. Although this function
* may be called for many read-write locks, the initialization
* of the the head happens only once during the lifetime of
* the thread.
*/
if (curthread->rwlockList == NULL) {
curthread->rwlockList =
(struct rwlock_listhead *)malloc(sizeof(struct rwlock_listhead));
if (curthread->rwlockList == NULL) {
return (ENOMEM);
}
LIST_INIT(curthread->rwlockList);
}
LIST_FOREACH(rh, curthread->rwlockList, rh_link) {
if (rh->rh_rwlock == prwlock) {
if (rwt == RWT_READ)
rh->rh_rdcount++;
else if (rwt == RWT_WRITE)
rh->rh_wrcount++;
return (0);
}
}
/*
* This is the first time we're holding this lock,
* create a new entry.
*/
rh = (struct rwlock_held *)malloc(sizeof(struct rwlock_held));
if (rh == NULL)
return (ENOMEM);
rh->rh_rwlock = prwlock;
rh->rh_rdcount = 0;
rh->rh_wrcount = 0;
if (rwt == RWT_READ)
rh->rh_rdcount = 1;
else if (rwt == RWT_WRITE)
rh->rh_wrcount = 1;
LIST_INSERT_HEAD(curthread->rwlockList, rh, rh_link);
return (0);
}
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
/*
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
* There are consumers of rwlocks, inluding our own libc, that depend on
* a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
* a similarly named symbol does for statically initialized mutexes.
There are consumers of rwlocks, inluding our own libc, that depend on a PTHREAD_RWLOCK_INITIALIZER to do for rwlocks what a similarly named symbol does for statically initialized mutexes. This symbol was dropped in The Open Group Base Specifications Issue 6 and does not exist in IEEE Std 1003.1, 2003, but it should still be supported for backwards compatibility. Pointy hat: mtm
2004-02-18 15:30:10 +00:00
* This symbol was dropped in The Open Group Base Specifications Issue 6
* and does not exist in IEEE Std 1003.1, 2003, but it should still be
* supported for backwards compatibility.
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
*/
static int
rwlock_init_static(struct pthread_rwlock **rwlock)
{
int error;
error = 0;
o Assertions to catch that stuff that shouldn't happen is not happening. o In the rwlock code: move a duplicated check inside an if..else to after the if...else clause. o When initializing a static rwlock move the initialization check inside the lock. o In thr_setschedparam.c: When breaking out of the trylock...retry if busy loop make sure to reset the mtx pointer to null if the mutex is nolonger in a queue.
2004-07-30 17:13:00 +00:00
UMTX_LOCK(&init_lock);
if (*rwlock == PTHREAD_RWLOCK_INITIALIZER)
error = _pthread_rwlock_init(rwlock, NULL);
UMTX_UNLOCK(&init_lock);
I update the rwlock code in libthr to be more standards compliant and what do I get for my troubles? libc breaks offcourse! Reimplement a hack (in libthr) that allows libc to use rwlocks without initializing them first. The hack was reimplemented so that only a private libc version of the rwlock locking functions initializes an uninitialized rwlock. The application version will correctly fail.
2004-01-29 12:03:17 +00:00
return (error);
}
Reference in New Issue Copy Permalink