mirror of
https://github.com/jart/cosmopolitan.git
synced 2025-02-12 01:08:00 +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.
146 lines
4.9 KiB
C
146 lines
4.9 KiB
C
/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:2;tab-width:8;coding:utf-8 -*-│
|
|
│ vi: set et ft=c ts=8 sts=2 sw=2 fenc=utf-8 :vi │
|
|
╞══════════════════════════════════════════════════════════════════════════════╡
|
|
│ Copyright 2020 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/struct/iovec.internal.h"
|
|
#include "libc/errno.h"
|
|
#include "libc/fmt/conv.h"
|
|
#include "libc/macros.h"
|
|
#include "libc/stdckdint.h"
|
|
#include "libc/stdio/internal.h"
|
|
#include "libc/stdio/stdio.h"
|
|
#include "libc/str/str.h"
|
|
#include "libc/sysv/consts/o.h"
|
|
|
|
static ssize_t writevall(int fd, struct iovec *iov, int iovlen) {
|
|
int olde;
|
|
ssize_t rc;
|
|
size_t got, toto;
|
|
toto = 0;
|
|
olde = errno;
|
|
do {
|
|
if ((rc = writev(fd, iov, iovlen)) == -1) {
|
|
if (toto && errno == EINTR) {
|
|
errno = olde;
|
|
continue;
|
|
}
|
|
return -1;
|
|
}
|
|
got = rc;
|
|
toto += got;
|
|
for (;;) {
|
|
if (!iov->iov_len) {
|
|
--iovlen;
|
|
++iov;
|
|
} else if (got >= iov->iov_len) {
|
|
got -= iov->iov_len;
|
|
--iovlen;
|
|
++iov;
|
|
} else {
|
|
iov->iov_base += got;
|
|
iov->iov_len -= got;
|
|
break;
|
|
}
|
|
}
|
|
} while (got && iovlen);
|
|
return toto;
|
|
}
|
|
|
|
/**
|
|
* Writes data to stream.
|
|
*
|
|
* @param stride specifies the size of individual items
|
|
* @param count is the number of strides to write
|
|
* @return count on success, [0,count) on EOF, 0 on error or count==0
|
|
*/
|
|
size_t fwrite_unlocked(const void *data, size_t stride, size_t count, FILE *f) {
|
|
ldiv_t d;
|
|
ssize_t rc;
|
|
size_t n, m;
|
|
const char *p;
|
|
struct iovec iov[2];
|
|
if (!stride || !count)
|
|
return 0;
|
|
if ((f->oflags & O_ACCMODE) == O_RDONLY) {
|
|
f->state = errno = EBADF;
|
|
return 0;
|
|
}
|
|
if (ckd_mul(&n, stride, count)) {
|
|
f->state = errno = EOVERFLOW;
|
|
return 0;
|
|
}
|
|
m = f->size - f->beg;
|
|
if (n <= m && f->bufmode != _IONBF) {
|
|
// this isn't an unbuffered stream, and
|
|
// there's enough room in the buffer for the request
|
|
memcpy(f->buf + f->beg, data, n);
|
|
f->beg += n;
|
|
if (f->fd != -1 && f->bufmode == _IOLBF &&
|
|
(p = memrchr(f->buf, '\n', f->beg))) {
|
|
// write out as many lines as possible
|
|
n = p + 1 - f->buf;
|
|
if ((rc = write(f->fd, f->buf, n)) == -1) {
|
|
// the write() system call failed
|
|
if (errno == EINTR || errno == EAGAIN) {
|
|
return count;
|
|
} else {
|
|
f->state = errno;
|
|
return 0;
|
|
}
|
|
}
|
|
// copy backwards last line fragment or uncompleted data
|
|
n = rc;
|
|
if (f->beg - n) {
|
|
memmove(f->buf, f->buf + n, f->beg - n);
|
|
}
|
|
f->beg -= n;
|
|
}
|
|
return count;
|
|
}
|
|
// what's happening is either
|
|
// (1) a fully buffered stream, or
|
|
// (2) no room in buffer to hold full request
|
|
if (f->fd == -1) {
|
|
// this is an in-memory stream
|
|
// store as much of request as we can hold
|
|
n = MIN(n, m);
|
|
d = ldiv(n, stride);
|
|
n -= d.rem;
|
|
if (n) {
|
|
memcpy(f->buf + f->beg, data, n);
|
|
f->beg += n;
|
|
}
|
|
// trigger eof condition
|
|
f->state = EOF;
|
|
return d.quot;
|
|
}
|
|
// perform a fragmented write
|
|
// (1) we avoid needless copies in fully buffered mode
|
|
// (2) we avoid need for malloc() when it's out of room
|
|
iov[0].iov_base = f->buf;
|
|
iov[0].iov_len = f->beg;
|
|
iov[1].iov_base = (void *)data;
|
|
iov[1].iov_len = n;
|
|
n += f->beg;
|
|
if ((rc = writevall(f->fd, iov, 2)) == -1) {
|
|
f->state = errno;
|
|
return 0;
|
|
}
|
|
f->beg = 0;
|
|
return count;
|
|
}
|