cosmopolitan/libc/intrin/reservefd.c
Justine Tunney 2f48a02b44
Make recursive mutexes faster
Recursive mutexes now go as fast as normal mutexes. The tradeoff is they
are no longer safe to use in signal handlers. However you can still have
signal safe mutexes if you set your mutex to both recursive and pshared.
You can also make functions that use recursive mutexes signal safe using
sigprocmask to ensure recursion doesn't happen due to any signal handler

The impact of this change is that, on Windows, many functions which edit
the file descriptor table rely on recursive mutexes, e.g. open(). If you
develop your app so it uses pread() and pwrite() then your app should go
very fast when performing a heavily multithreaded and contended workload

For example, when scaling to 40+ cores, *NSYNC mutexes can go as much as
1000x faster (in CPU time) than the naive recursive lock implementation.
Now recursive will use *NSYNC under the hood when it's possible to do so
2024-09-10 00:08:59 -07:00

93 lines
3.6 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 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/internal.h"
#include "libc/calls/state.internal.h"
#include "libc/intrin/atomic.h"
#include "libc/intrin/cmpxchg.h"
#include "libc/intrin/extend.h"
#include "libc/intrin/fds.h"
#include "libc/macros.h"
#include "libc/runtime/memtrack.internal.h"
#include "libc/str/str.h"
#include "libc/sysv/consts/map.h"
/**
* Grows file descriptor array memory if needed.
*
* @see libc/runtime/memtrack64.txt
* @see libc/runtime/memtrack32.txt
* @asyncsignalsafe
*/
int __ensurefds_unlocked(int fd) {
size_t n;
if (fd < g_fds.n)
return fd;
n = fd + 1;
g_fds.e = _extend(g_fds.p, n * sizeof(*g_fds.p), g_fds.e, MAP_PRIVATE,
kMemtrackFdsStart + kMemtrackFdsSize);
g_fds.n = n;
return fd;
}
/**
* Grows file descriptor array memory if needed.
* @asyncsignalsafe if signals are blocked
*/
int __ensurefds(int fd) {
__fds_lock();
fd = __ensurefds_unlocked(fd);
__fds_unlock();
return fd;
}
/**
* Finds open file descriptor slot.
* @asyncsignalsafe
*/
int __reservefd_unlocked(int start) {
int fd, f1, f2;
for (;;) {
f1 = atomic_load_explicit(&g_fds.f, memory_order_acquire);
for (fd = MAX(start, f1); fd < g_fds.n; ++fd)
if (!g_fds.p[fd].kind)
break;
fd = __ensurefds_unlocked(fd);
bzero(g_fds.p + fd, sizeof(*g_fds.p));
if (_cmpxchg(&g_fds.p[fd].kind, kFdEmpty, kFdReserved)) {
// g_fds.f isn't guarded by our mutex
do {
f2 = MAX(fd + 1, f1);
} while (!atomic_compare_exchange_weak_explicit(
&g_fds.f, &f1, f2, memory_order_release, memory_order_relaxed));
return fd;
}
}
}
/**
* Finds open file descriptor slot.
* @asyncsignalsafe if signals are blocked
*/
int __reservefd(int start) {
int fd;
__fds_lock();
fd = __reservefd_unlocked(start);
__fds_unlock();
return fd;
}