mirror of
https://github.com/jart/cosmopolitan.git
synced 2025-02-22 22:19:02 +00:00
This change introduces a new deadlock detector for Cosmo's POSIX threads implementation. Error check mutexes will now track a DAG of nested locks and report EDEADLK when a deadlock is theoretically possible. These will occur rarely, but it's important for production hardening your code. You don't even need to change your mutexes to use the POSIX error check mode because `cosmocc -mdbg` will enable error checking on mutexes by default globally. When cycles are found, an error message showing your demangled symbols describing the strongly connected component are printed and then the SIGTRAP is raised, which means you'll also get a backtrace if you're using ShowCrashReports() too. This new error checker is so low-level and so pure that it's able to verify the relationships of every libc runtime lock, including those locks upon which the mutex implementation depends.
167 lines
6.9 KiB
C
167 lines
6.9 KiB
C
/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:2;tab-width:8;coding:utf-8 -*-│
|
|
│ vi: set et ft=c ts=2 sts=2 sw=2 fenc=utf-8 :vi │
|
|
╞══════════════════════════════════════════════════════════════════════════════╡
|
|
│ Copyright 2022 Justine Alexandra Roberts Tunney │
|
|
│ │
|
|
│ Permission to use, copy, modify, and/or distribute this software for │
|
|
│ any purpose with or without fee is hereby granted, provided that the │
|
|
│ above copyright notice and this permission notice appear in all copies. │
|
|
│ │
|
|
│ THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL │
|
|
│ WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED │
|
|
│ WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE │
|
|
│ AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL │
|
|
│ DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR │
|
|
│ PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER │
|
|
│ TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR │
|
|
│ PERFORMANCE OF THIS SOFTWARE. │
|
|
╚─────────────────────────────────────────────────────────────────────────────*/
|
|
#include "libc/calls/calls.h"
|
|
#include "libc/calls/cp.internal.h"
|
|
#include "libc/calls/struct/timespec.h"
|
|
#include "libc/cosmo.h"
|
|
#include "libc/dce.h"
|
|
#include "libc/errno.h"
|
|
#include "libc/intrin/atomic.h"
|
|
#include "libc/sysv/consts/clock.h"
|
|
#include "libc/thread/lock.h"
|
|
#include "libc/thread/posixthread.internal.h"
|
|
#include "libc/thread/thread.h"
|
|
#include "libc/thread/thread2.h"
|
|
#include "third_party/nsync/common.internal.h"
|
|
#include "third_party/nsync/cv.h"
|
|
#include "third_party/nsync/time.h"
|
|
|
|
__static_yoink("nsync_mu_lock");
|
|
__static_yoink("nsync_mu_unlock");
|
|
__static_yoink("nsync_mu_trylock");
|
|
|
|
struct PthreadWait {
|
|
pthread_cond_t *cond;
|
|
pthread_mutex_t *mutex;
|
|
};
|
|
|
|
static bool can_use_nsync(uint64_t muword) {
|
|
return !IsXnuSilicon() && //
|
|
MUTEX_TYPE(muword) != PTHREAD_MUTEX_RECURSIVE &&
|
|
MUTEX_PSHARED(muword) == PTHREAD_PROCESS_PRIVATE;
|
|
}
|
|
|
|
static void pthread_cond_leave(void *arg) {
|
|
struct PthreadWait *wait = (struct PthreadWait *)arg;
|
|
if (pthread_mutex_lock(wait->mutex))
|
|
__builtin_trap();
|
|
atomic_fetch_sub_explicit(&wait->cond->_waiters, 1, memory_order_acq_rel);
|
|
}
|
|
|
|
static errno_t pthread_cond_timedwait_impl(pthread_cond_t *cond,
|
|
pthread_mutex_t *mutex,
|
|
const struct timespec *abstime) {
|
|
|
|
// this is a cancelation point
|
|
// check the cancelation status before we begin waiting
|
|
if (pthread_testcancel_np() == ECANCELED)
|
|
return ECANCELED;
|
|
|
|
// get original monotonic sequence while lock is held
|
|
uint32_t seq1 = atomic_load_explicit(&cond->_sequence, memory_order_relaxed);
|
|
|
|
// start waiting on condition variable
|
|
atomic_fetch_add_explicit(&cond->_waiters, 1, memory_order_acq_rel);
|
|
if (pthread_mutex_unlock(mutex))
|
|
__builtin_trap();
|
|
|
|
// wait for sequence change, timeout, or cancelation
|
|
int rc;
|
|
struct PthreadWait waiter = {cond, mutex};
|
|
pthread_cleanup_push(pthread_cond_leave, &waiter);
|
|
rc = cosmo_futex_wait((atomic_int *)&cond->_sequence, seq1, cond->_pshared,
|
|
cond->_clock, abstime);
|
|
pthread_cleanup_pop(true);
|
|
if (rc == -EAGAIN)
|
|
rc = 0;
|
|
|
|
// turn linux syscall status into posix errno
|
|
return -rc;
|
|
}
|
|
|
|
/**
|
|
* Waits for condition with optional time limit, e.g.
|
|
*
|
|
* struct timespec ts; // one second timeout
|
|
* ts = timespec_add(timespec_real(), timespec_frommillis(1000));
|
|
* if (pthread_cond_timedwait(cond, mutex, &ts) == ETIMEDOUT) {
|
|
* // handle timeout...
|
|
* }
|
|
*
|
|
* @param mutex needs to be held by thread when calling this function
|
|
* @param abstime is an absolute timestamp, which may be null to wait
|
|
* forever; it's relative to `clock_gettime(CLOCK_REALTIME)` by
|
|
* default; pthread_condattr_setclock() may be used to customize
|
|
* which system clock is used
|
|
* @return 0 on success, or errno on error
|
|
* @raise ETIMEDOUT if `abstime` was specified and the current time
|
|
* exceeded its value
|
|
* @raise EPERM if `mutex` is `PTHREAD_MUTEX_ERRORCHECK` and the lock
|
|
* isn't owned by the current thread
|
|
* @raise EINVAL if `0 ≤ abstime->tv_nsec < 1000000000` wasn't the case
|
|
* @raise ECANCELED if calling thread was cancelled in masked mode
|
|
* @see pthread_cond_broadcast()
|
|
* @see pthread_cond_signal()
|
|
* @cancelationpoint
|
|
*/
|
|
errno_t pthread_cond_timedwait(pthread_cond_t *cond, pthread_mutex_t *mutex,
|
|
const struct timespec *abstime) {
|
|
|
|
// validate arguments
|
|
struct PosixThread *pt;
|
|
if (!(pt = _pthread_self()))
|
|
return EINVAL;
|
|
if (abstime && !(0 <= abstime->tv_nsec && abstime->tv_nsec < 1000000000))
|
|
return EINVAL;
|
|
|
|
// look at the mutex argument
|
|
uint64_t muword = atomic_load_explicit(&mutex->_word, memory_order_relaxed);
|
|
|
|
// check that mutex is held by caller
|
|
if (IsModeDbg() || MUTEX_TYPE(muword) == PTHREAD_MUTEX_ERRORCHECK)
|
|
if (__deadlock_tracked(mutex) == 0)
|
|
return EPERM;
|
|
|
|
// if the cond is process shared then the mutex needs to be too
|
|
if ((cond->_pshared == PTHREAD_PROCESS_SHARED) ^
|
|
(MUTEX_PSHARED(muword) == PTHREAD_PROCESS_SHARED))
|
|
return EINVAL;
|
|
|
|
#if PTHREAD_USE_NSYNC
|
|
// the first time pthread_cond_timedwait() is called we learn if the
|
|
// associated mutex is normal and private. that means *NSYNC is safe
|
|
// this decision is permanent. you can't use a recursive mutex later
|
|
if (!atomic_load_explicit(&cond->_waited, memory_order_acquire)) {
|
|
cond->_footek = !can_use_nsync(muword);
|
|
atomic_store_explicit(&cond->_waited, true, memory_order_release);
|
|
} else if (!cond->_footek) {
|
|
if (!can_use_nsync(muword))
|
|
return EINVAL;
|
|
}
|
|
#endif
|
|
|
|
// now perform the actual wait
|
|
errno_t err;
|
|
BEGIN_CANCELATION_POINT;
|
|
#if PTHREAD_USE_NSYNC
|
|
// favor *NSYNC if this is a process private condition variable
|
|
// if using Mike Burrows' code isn't possible, use a naive impl
|
|
if (!cond->_footek) {
|
|
err = nsync_cv_wait_with_deadline(
|
|
(nsync_cv *)cond->_nsync, (nsync_mu *)mutex->_nsync, cond->_clock,
|
|
abstime ? *abstime : nsync_time_no_deadline, 0);
|
|
} else {
|
|
err = pthread_cond_timedwait_impl(cond, mutex, abstime);
|
|
}
|
|
#else
|
|
err = pthread_cond_timedwait_impl(cond, mutex, abstime);
|
|
#endif
|
|
END_CANCELATION_POINT;
|
|
return err;
|
|
}
|