freebsd-nq/lib/libc_r/uthread/uthread_file.c
Daniel Eischen fc8f3f5bfe Fix problems with cancellation while in critical regions.
o Cancellation flags were not getting properly set/cleared.
  o Loops waiting for internal locks were not being exited
    correctly by a cancelled thread.
  o Minor spelling (cancelation -> cancellation) and formatting
    corrections (missing tab).

Found by:	tg
Reviewed by:	jasone
1999-12-17 00:57:54 +00:00

476 lines
12 KiB
C

/*
* Copyright (c) 1995 John Birrell <jb@cimlogic.com.au>.
* 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.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by John Birrell.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY JOHN BIRRELL 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.
*
* $FreeBSD$
*
* POSIX stdio FILE locking functions. These assume that the locking
* is only required at FILE structure level, not at file descriptor
* level too.
*
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/queue.h>
#ifdef _THREAD_SAFE
#include <pthread.h>
#include "pthread_private.h"
/*
* Weak symbols for externally visible functions in this file:
*/
#pragma weak flockfile=_flockfile
#pragma weak ftrylockfile=_ftrylockfile
#pragma weak funlockfile=_funlockfile
/*
* The FILE lock structure. The FILE *fp is locked if the owner is
* not NULL. If not locked, the file lock structure can be
* reassigned to a different file by setting fp.
*/
struct file_lock {
LIST_ENTRY(file_lock) entry; /* Entry if file list. */
TAILQ_HEAD(lock_head, pthread)
l_head; /* Head of queue for threads */
/* waiting on this lock. */
FILE *fp; /* The target file. */
pthread_t owner; /* Thread that owns lock. */
int count; /* Lock count for owner. */
};
/*
* The number of file lock lists into which the file pointer is
* hashed. Ideally, the FILE structure size would have been increased,
* but this causes incompatibility, so separate data structures are
* required.
*/
#define NUM_HEADS 128
/*
* This macro casts a file pointer to a long integer and right
* shifts this by the number of bytes in a pointer. The shifted
* value is then remaindered using the maximum number of hash
* entries to produce and index into the array of static lock
* structures. If there is a collision, a linear search of the
* dynamic list of locks linked to each static lock is perfomed.
*/
#define file_idx(_p) ((((u_long) _p) >> sizeof(void *)) % NUM_HEADS)
/*
* Global array of file locks. The first lock for each hash bucket is
* allocated statically in the hope that there won't be too many
* collisions that require a malloc and an element added to the list.
*/
struct static_file_lock {
LIST_HEAD(file_list_head, file_lock) head;
struct file_lock fl;
} flh[NUM_HEADS];
/* Set to non-zero when initialisation is complete: */
static int init_done = 0;
/* Lock for accesses to the hash table: */
static spinlock_t hash_lock = _SPINLOCK_INITIALIZER;
/*
* Find a lock structure for a FILE, return NULL if the file is
* not locked:
*/
static
struct file_lock *
find_lock(int idx, FILE *fp)
{
struct file_lock *p;
/* Check if the file is locked using the static structure: */
if (flh[idx].fl.fp == fp && flh[idx].fl.owner != NULL)
/* Return a pointer to the static lock: */
p = &flh[idx].fl;
else {
/* Point to the first dynamic lock: */
p = flh[idx].head.lh_first;
/*
* Loop through the dynamic locks looking for the
* target file:
*/
while (p != NULL && (p->fp != fp || p->owner == NULL))
/* Not this file, try the next: */
p = p->entry.le_next;
}
return(p);
}
/*
* Lock a file, assuming that there is no lock structure currently
* assigned to it.
*/
static
struct file_lock *
do_lock(int idx, FILE *fp)
{
struct file_lock *p;
/* Check if the static structure is not being used: */
if (flh[idx].fl.owner == NULL) {
/* Return a pointer to the static lock: */
p = &flh[idx].fl;
}
else {
/* Point to the first dynamic lock: */
p = flh[idx].head.lh_first;
/*
* Loop through the dynamic locks looking for a
* lock structure that is not being used:
*/
while (p != NULL && p->owner != NULL)
/* This one is used, try the next: */
p = p->entry.le_next;
}
/*
* If an existing lock structure has not been found,
* allocate memory for a new one:
*/
if (p == NULL && (p = (struct file_lock *)
malloc(sizeof(struct file_lock))) != NULL) {
/* Add the new element to the list: */
LIST_INSERT_HEAD(&flh[idx].head, p, entry);
}
/* Check if there is a lock structure to acquire: */
if (p != NULL) {
/* Acquire the lock for the running thread: */
p->fp = fp;
p->owner = _thread_run;
p->count = 1;
TAILQ_INIT(&p->l_head);
}
return(p);
}
void
_flockfile_debug(FILE * fp, char *fname, int lineno)
{
int idx = file_idx(fp);
struct file_lock *p;
/* Check if this is a real file: */
if (fp->_file >= 0) {
/* Lock the hash table: */
_SPINLOCK(&hash_lock);
/* Check if the static array has not been initialised: */
if (!init_done) {
/* Initialise the global array: */
memset(flh,0,sizeof(flh));
/* Flag the initialisation as complete: */
init_done = 1;
}
/* Get a pointer to any existing lock for the file: */
if ((p = find_lock(idx, fp)) == NULL) {
/*
* The file is not locked, so this thread can
* grab the lock:
*/
p = do_lock(idx, fp);
/* Unlock the hash table: */
_SPINUNLOCK(&hash_lock);
/*
* The file is already locked, so check if the
* running thread is the owner:
*/
} else if (p->owner == _thread_run) {
/*
* The running thread is already the
* owner, so increment the count of
* the number of times it has locked
* the file:
*/
p->count++;
/* Unlock the hash table: */
_SPINUNLOCK(&hash_lock);
} else {
/* Clear the interrupted flag: */
_thread_run->interrupted = 0;
/*
* Prevent being context switched out while
* adding this thread to the file lock queue.
*/
_thread_kern_sig_defer();
/*
* The file is locked for another thread.
* Append this thread to the queue of
* threads waiting on the lock.
*/
TAILQ_INSERT_TAIL(&p->l_head,_thread_run,qe);
_thread_run->flags |= PTHREAD_FLAGS_IN_FILEQ;
/* Unlock the hash table: */
_SPINUNLOCK(&hash_lock);
/* Wait on the FILE lock: */
_thread_kern_sched_state(PS_FILE_WAIT, fname, lineno);
if ((_thread_run->flags & PTHREAD_FLAGS_IN_FILEQ) != 0) {
TAILQ_REMOVE(&p->l_head,_thread_run,qe);
_thread_run->flags &= ~PTHREAD_FLAGS_IN_FILEQ;
}
_thread_kern_sig_undefer();
if (((_thread_run->cancelflags & PTHREAD_CANCEL_NEEDED) != 0) &&
(_thread_run->cancelflags & PTHREAD_CANCEL_DISABLE) != 0) {
_thread_run->cancelflags &= ~PTHREAD_CANCEL_NEEDED;
_thread_exit_cleanup();
pthread_exit(PTHREAD_CANCELED);
}
}
}
return;
}
void
_flockfile(FILE * fp)
{
_flockfile_debug(fp, __FILE__, __LINE__);
return;
}
int
_ftrylockfile(FILE * fp)
{
int ret = -1;
int idx = file_idx(fp);
struct file_lock *p;
/* Check if this is a real file: */
if (fp->_file >= 0) {
/* Lock the hash table: */
_SPINLOCK(&hash_lock);
/* Get a pointer to any existing lock for the file: */
if ((p = find_lock(idx, fp)) == NULL) {
/*
* The file is not locked, so this thread can
* grab the lock:
*/
p = do_lock(idx, fp);
/*
* The file is already locked, so check if the
* running thread is the owner:
*/
} else if (p->owner == _thread_run) {
/*
* The running thread is already the
* owner, so increment the count of
* the number of times it has locked
* the file:
*/
p->count++;
} else {
/*
* The file is locked for another thread,
* so this try fails.
*/
p = NULL;
}
/* Check if the lock was obtained: */
if (p != NULL)
/* Return success: */
ret = 0;
/* Unlock the hash table: */
_SPINUNLOCK(&hash_lock);
}
return (ret);
}
void
_funlockfile(FILE * fp)
{
int idx = file_idx(fp);
struct file_lock *p;
/* Check if this is a real file: */
if (fp->_file >= 0) {
/*
* Defer signals to protect the scheduling queues from
* access by the signal handler:
*/
_thread_kern_sig_defer();
/* Lock the hash table: */
_SPINLOCK(&hash_lock);
/*
* Get a pointer to the lock for the file and check that
* the running thread is the one with the lock:
*/
if ((p = find_lock(idx, fp)) != NULL &&
p->owner == _thread_run) {
/*
* Check if this thread has locked the FILE
* more than once:
*/
if (p->count > 1)
/*
* Decrement the count of the number of
* times the running thread has locked this
* file:
*/
p->count--;
else {
/*
* The running thread will release the
* lock now:
*/
p->count = 0;
/* Get the new owner of the lock: */
while ((p->owner = TAILQ_FIRST(&p->l_head)) != NULL) {
/* Pop the thread off the queue: */
TAILQ_REMOVE(&p->l_head,p->owner,qe);
p->owner->flags &= ~PTHREAD_FLAGS_IN_FILEQ;
if (p->owner->interrupted == 0) {
/*
* This is the first lock for
* the new owner:
*/
p->count = 1;
/* Allow the new owner to run: */
PTHREAD_NEW_STATE(p->owner,PS_RUNNING);
/* End the loop when we find a
* thread that hasn't been
* cancelled or interrupted;
*/
break;
}
}
}
}
/* Unlock the hash table: */
_SPINUNLOCK(&hash_lock);
/*
* Undefer and handle pending signals, yielding if
* necessary:
*/
_thread_kern_sig_undefer();
}
return;
}
void
_funlock_owned(pthread_t pthread)
{
int idx;
struct file_lock *p, *next_p;
/*
* Defer signals to protect the scheduling queues from
* access by the signal handler:
*/
_thread_kern_sig_defer();
/* Lock the hash table: */
_SPINLOCK(&hash_lock);
for (idx = 0; idx < NUM_HEADS; idx++) {
/* Check the static file lock first: */
p = &flh[idx].fl;
next_p = LIST_FIRST(&flh[idx].head);
while (p != NULL) {
if (p->owner == pthread) {
/*
* The running thread will release the
* lock now:
*/
p->count = 0;
/* Get the new owner of the lock: */
while ((p->owner = TAILQ_FIRST(&p->l_head)) != NULL) {
/* Pop the thread off the queue: */
TAILQ_REMOVE(&p->l_head,p->owner,qe);
p->owner->flags &= ~PTHREAD_FLAGS_IN_FILEQ;
if (p->owner->interrupted == 0) {
/*
* This is the first lock for
* the new owner:
*/
p->count = 1;
/* Allow the new owner to run: */
PTHREAD_NEW_STATE(p->owner,PS_RUNNING);
/* End the loop when we find a
* thread that hasn't been
* cancelled or interrupted;
*/
break;
}
}
}
p = next_p;
if (next_p != NULL)
next_p = LIST_NEXT(next_p, entry);
}
}
/* Unlock the hash table: */
_SPINUNLOCK(&hash_lock);
/*
* Undefer and handle pending signals, yielding if
* necessary:
*/
_thread_kern_sig_undefer();
}
#endif