mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2024-05-18 03:22:40 +00:00
Code Issues Projects Releases Wiki Activity
cosmopolitan/libc/calls/fcntl.c
Justine Tunney 791f79fcb3
Make improvements 
- We now serialize the file descriptor table when spawning / executing
  processes on Windows. This means you can now inherit more stuff than
  just standard i/o. It's needed by bash, which duplicates the console
  to file descriptor #255. We also now do a better job serializing the
  environment variables, so you're less likely to encounter E2BIG when
  using your bash shell. We also no longer coerce environ to uppercase

- execve() on Windows now remotely controls its parent process to make
  them spawn a replacement for itself. Then it'll be able to terminate
  immediately once the spawn succeeds, without having to linger around
  for the lifetime as a shell process for proxying the exit code. When
  process worker thread running in the parent sees the child die, it's
  given a handle to the new child, to replace it in the process table.

- execve() and posix_spawn() on Windows will now provide CreateProcess
  an explicit handle list. This allows us to remove handle locks which
  enables better fork/spawn concurrency, with seriously correct thread
  safety. Other codebases like Go use the same technique. On the other
  hand fork() still favors the conventional WIN32 inheritence approach
  which can be a little bit messy, but is *controlled* by guaranteeing
  perfectly clean slates at both the spawning and execution boundaries

- sigset_t is now 64 bits. Having it be 128 bits was a mistake because
  there's no reason to use that and it's only supported by FreeBSD. By
  using the system word size, signal mask manipulation on Windows goes
  very fast. Furthermore @asyncsignalsafe funcs have been rewritten on
  Windows to take advantage of signal masking, now that it's much more
  pleasant to use.

- All the overlapped i/o code on Windows has been rewritten for pretty
  good signal and cancelation safety. We're now able to ensure overlap
  data structures are cleaned up so long as you don't longjmp() out of
  out of a signal handler that interrupted an i/o operation. Latencies
  are also improved thanks to the removal of lots of "busy wait" code.
  Waits should be optimal for everything except poll(), which shall be
  the last and final demon we slay in the win32 i/o horror show.

- getrusage() on Windows is now able to report RUSAGE_CHILDREN as well
  as RUSAGE_SELF, thanks to aggregation in the process manager thread.
2023-10-08 08:59:53 -07:00

177 lines
8.4 KiB
C
Raw Blame History

/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:2;tab-width:8;coding:utf-8 -*-│
│vi: set net 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/calls.h"
#include "libc/calls/cp.internal.h"
#include "libc/calls/internal.h"
#include "libc/calls/struct/flock.h"
#include "libc/calls/struct/flock.internal.h"
#include "libc/calls/syscall-nt.internal.h"
#include "libc/calls/syscall-sysv.internal.h"
#include "libc/dce.h"
#include "libc/intrin/describeflags.internal.h"
#include "libc/intrin/strace.internal.h"
#include "libc/intrin/weaken.h"
#include "libc/runtime/zipos.internal.h"
#include "libc/sysv/consts/f.h"
#include "libc/sysv/errfuns.h"
/**
* Does things with file descriptor, e.g.
*
* CHECK_NE(-1, fcntl(fd, F_SETFD, FD_CLOEXEC));
*
* This function lets you duplicate file descriptors without running
* into an edge case where they take over stdio handles:
*
* CHECK_GE((newfd = fcntl(oldfd, F_DUPFD, 3)), 3);
* CHECK_GE((newfd = fcntl(oldfd, F_DUPFD_CLOEXEC, 3)), 3);
*
* This function implements file record locking, which lets independent
* processes (and on Linux 3.15+, threads too!) lock arbitrary ranges
* associated with a file. See `test/libc/calls/lock_test.c` and other
* locking related tests in that folder.
*
* On Windows, the Cosmopolitan Libc polyfill for POSIX advisory locks
* only implements enough of its nuances to support SQLite's needs. Some
* possibilities, e.g. punching holes in lock, will raise `ENOTSUP`.
*
* @param fd is the file descriptor
* @param cmd can be one of:
* - `F_GETFD` gets `FD_CLOEXEC` status of `fd
* - `F_SETFD` sets `FD_CLOEXEC` status of `arg` file descriptor
* - `F_GETFL` returns file descriptor status flags
* - `F_SETFL` sets file descriptor status flags
* - `O_NONBLOCK` may be changed
* - `O_APPEND` may be changed
* - `O_DIRECT` may be changed
* - `O_SEQUENTIAL` may be changed (no-op on non-Windows)
* - `O_RANDOM` may be changed (no-op on non-Windows)
* - Other bits (`O_ACCMODE`, `O_CREAT`, etc.) are ignored
* - `F_DUPFD` is like dup() but `arg` is a minimum result, e.g. 3
* - `F_DUPFD_CLOEXEC` ditto but sets `O_CLOEXEC` on returned fd
* - `F_SETLK` for record locking where `arg` is `struct flock *`
* - `F_SETLKW` ditto but waits for lock (SQLite avoids this)
* - `F_GETLK` to retrieve information about a record lock
* - `F_OFD_SETLK` for better non-blocking lock (Linux 3.15+ only)
* - `F_OFD_SETLKW` for better blocking lock (Linux 3.15+ only)
* - `F_OFD_GETLK` for better lock querying (Linux 3.15+ only)
* - `F_FULLFSYNC` on MacOS for fsync() with release barrier
* - `F_BARRIERFSYNC` on MacOS for fsync() with even more barriers
* - `F_SETNOSIGPIPE` on MacOS and NetBSD to control `SIGPIPE`
* - `F_GETNOSIGPIPE` on MacOS and NetBSD to control `SIGPIPE`
* - `F_GETPATH` on MacOS and NetBSD where arg is `char[PATH_MAX]`
* - `F_MAXFD` on NetBSD to get max open file descriptor
* - `F_NOCACHE` on MacOS to toggle data caching
* - `F_GETPIPE_SZ` on Linux to get pipe size
* - `F_SETPIPE_SZ` on Linux to set pipe size
* - `F_NOTIFY` raise `SIGIO` upon `fd` events in `arg` (Linux only)
* - `DN_ACCESS` for file access
* - `DN_MODIFY` for file modifications
* - `DN_CREATE` for file creations
* - `DN_DELETE` for file deletions
* - `DN_RENAME` for file renames
* - `DN_ATTRIB` for file attribute changes
* - `DN_MULTISHOT` bitwise or for realtime signals (non-coalesced)
* @param arg can be FD_CLOEXEC, etc. depending
* @return 0 on success, or -1 w/ errno
* @raise EBADF if `fd` isn't a valid open file descriptor
* @raise EINVAL if `cmd` is unknown or unsupported by os
* @raise EINVAL if `cmd` is invalid or unsupported by os
* @raise EPERM if pledge() is in play w/o `stdio` or `flock` promise
* @raise ENOLCK if `F_SETLKW` would have exceeded `RLIMIT_LOCKS`
* @raise EPERM if `cmd` is `F_SETOWN` and we weren't authorized
* @raise ESRCH if `cmd` is `F_SETOWN` and process group not found
* @raise ENOTSUP on Windows if locking operation isn't supported yet
* @raise EDEADLK if `cmd` was `F_SETLKW` and waiting would deadlock
* @raise EMFILE if `cmd` is `F_DUPFD` or `F_DUPFD_CLOEXEC` and
* `RLIMIT_NOFILE` would be exceeded
* @cancelationpoint when `cmd` is `F_SETLKW` or `F_OFD_SETLKW`
* @asyncsignalsafe
* @restartable
*/
int fcntl(int fd, int cmd, ...) {
int rc;
va_list va;
uintptr_t arg;
va_start(va, cmd);
arg = va_arg(va, uintptr_t);
va_end(va);
if (fd >= 0) {
if (cmd >= 0) {
if (fd < g_fds.n && g_fds.p[fd].kind == kFdZip) {
rc = _weaken(__zipos_fcntl)(fd, cmd, arg);
} else if (!IsWindows()) {
if (cmd == F_SETLKW || cmd == F_OFD_SETLKW) {
BEGIN_CANCELATION_POINT;
rc = sys_fcntl(fd, cmd, arg, __sys_fcntl_cp);
END_CANCELATION_POINT;
} else {
rc = sys_fcntl(fd, cmd, arg, __sys_fcntl);
}
} else {
rc = sys_fcntl_nt(fd, cmd, arg);
}
} else {
rc = einval();
}
} else {
rc = ebadf();
}
#ifdef SYSDEBUG
if (rc != -1 && cmd == F_GETFL) {
STRACE("fcntl(%d, F_GETFL) → %s", fd, DescribeOpenFlags(rc));
} else if (cmd == F_SETFL) {
STRACE("fcntl(%d, F_SETFL, %s) → %d% m", fd, DescribeOpenFlags(arg), rc);
} else if (cmd == F_GETFD || //
cmd == F_GETOWN || //
cmd == F_FULLFSYNC || //
cmd == F_BARRIERFSYNC || //
cmd == F_MAXFD) {
STRACE("fcntl(%d, %s) → %d% m", fd, DescribeFcntlCmd(cmd), rc);
} else if (cmd == F_GETFL) {
STRACE("fcntl(%d, %s) → %s% m", fd, DescribeFcntlCmd(cmd),
DescribeOpenFlags(rc));
} else if (cmd == F_SETLK || //
cmd == F_SETLKW || //
cmd == F_GETLK || //
((SupportsLinux() || SupportsXnu()) && //
cmd != -1 && //
(cmd == F_OFD_SETLK || //
cmd == F_OFD_SETLKW || //
cmd == F_OFD_GETLK))) {
STRACE("fcntl(%d, %s, %s) → %d% m", fd, DescribeFcntlCmd(cmd),
DescribeFlock(cmd, (struct flock *)arg), rc);
} else if ((SupportsNetbsd() || SupportsXnu()) && cmd != -1 &&
cmd == F_GETPATH) {
STRACE("fcntl(%d, %s, [%s]) → %d% m", fd, DescribeFcntlCmd(cmd),
!rc ? (const char *)arg : "n/a", rc);
} else if (SupportsLinux() && cmd != -1 && cmd == F_NOTIFY) {
STRACE("fcntl(%d, %s, %s) → %d% m", fd, DescribeFcntlCmd(cmd),
DescribeDnotifyFlags(arg), rc);
} else {
STRACE("fcntl(%d, %s, %ld) → %d% m", fd, DescribeFcntlCmd(cmd), arg, rc);
}
#endif
return rc;
}
Reference in a new issue View git blame Copy permalink