xref: /netbsd-src/sys/kern/sys_pipe.c (revision 627f7eb200a4419d89b531d55fccd2ee3ffdcde0)
1 /*	$NetBSD: sys_pipe.c,v 1.152 2021/01/25 19:21:11 dholland Exp $	*/
2 
3 /*-
4  * Copyright (c) 2003, 2007, 2008, 2009 The NetBSD Foundation, Inc.
5  * All rights reserved.
6  *
7  * This code is derived from software contributed to The NetBSD Foundation
8  * by Paul Kranenburg, and by Andrew Doran.
9  *
10  * Redistribution and use in source and binary forms, with or without
11  * modification, are permitted provided that the following conditions
12  * are met:
13  * 1. Redistributions of source code must retain the above copyright
14  *    notice, this list of conditions and the following disclaimer.
15  * 2. Redistributions in binary form must reproduce the above copyright
16  *    notice, this list of conditions and the following disclaimer in the
17  *    documentation and/or other materials provided with the distribution.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
20  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
21  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
22  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
23  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29  * POSSIBILITY OF SUCH DAMAGE.
30  */
31 
32 /*
33  * Copyright (c) 1996 John S. Dyson
34  * All rights reserved.
35  *
36  * Redistribution and use in source and binary forms, with or without
37  * modification, are permitted provided that the following conditions
38  * are met:
39  * 1. Redistributions of source code must retain the above copyright
40  *    notice immediately at the beginning of the file, without modification,
41  *    this list of conditions, and the following disclaimer.
42  * 2. Redistributions in binary form must reproduce the above copyright
43  *    notice, this list of conditions and the following disclaimer in the
44  *    documentation and/or other materials provided with the distribution.
45  * 3. Absolutely no warranty of function or purpose is made by the author
46  *    John S. Dyson.
47  * 4. Modifications may be freely made to this file if the above conditions
48  *    are met.
49  */
50 
51 /*
52  * This file contains a high-performance replacement for the socket-based
53  * pipes scheme originally used.  It does not support all features of
54  * sockets, but does do everything that pipes normally do.
55  *
56  * This code has two modes of operation, a small write mode and a large
57  * write mode.  The small write mode acts like conventional pipes with
58  * a kernel buffer.  If the buffer is less than PIPE_MINDIRECT, then the
59  * "normal" pipe buffering is done.  If the buffer is between PIPE_MINDIRECT
60  * and PIPE_SIZE in size it is mapped read-only into the kernel address space
61  * using the UVM page loan facility from where the receiving process can copy
62  * the data directly from the pages in the sending process.
63  *
64  * The constant PIPE_MINDIRECT is chosen to make sure that buffering will
65  * happen for small transfers so that the system will not spend all of
66  * its time context switching.  PIPE_SIZE is constrained by the
67  * amount of kernel virtual memory.
68  */
69 
70 #include <sys/cdefs.h>
71 __KERNEL_RCSID(0, "$NetBSD: sys_pipe.c,v 1.152 2021/01/25 19:21:11 dholland Exp $");
72 
73 #include <sys/param.h>
74 #include <sys/systm.h>
75 #include <sys/proc.h>
76 #include <sys/fcntl.h>
77 #include <sys/file.h>
78 #include <sys/filedesc.h>
79 #include <sys/filio.h>
80 #include <sys/kernel.h>
81 #include <sys/ttycom.h>
82 #include <sys/stat.h>
83 #include <sys/poll.h>
84 #include <sys/signalvar.h>
85 #include <sys/vnode.h>
86 #include <sys/uio.h>
87 #include <sys/select.h>
88 #include <sys/mount.h>
89 #include <sys/syscallargs.h>
90 #include <sys/sysctl.h>
91 #include <sys/kauth.h>
92 #include <sys/atomic.h>
93 #include <sys/pipe.h>
94 
95 static int	pipe_read(file_t *, off_t *, struct uio *, kauth_cred_t, int);
96 static int	pipe_write(file_t *, off_t *, struct uio *, kauth_cred_t, int);
97 static int	pipe_close(file_t *);
98 static int	pipe_poll(file_t *, int);
99 static int	pipe_kqfilter(file_t *, struct knote *);
100 static int	pipe_stat(file_t *, struct stat *);
101 static int	pipe_ioctl(file_t *, u_long, void *);
102 static void	pipe_restart(file_t *);
103 
104 static const struct fileops pipeops = {
105 	.fo_name = "pipe",
106 	.fo_read = pipe_read,
107 	.fo_write = pipe_write,
108 	.fo_ioctl = pipe_ioctl,
109 	.fo_fcntl = fnullop_fcntl,
110 	.fo_poll = pipe_poll,
111 	.fo_stat = pipe_stat,
112 	.fo_close = pipe_close,
113 	.fo_kqfilter = pipe_kqfilter,
114 	.fo_restart = pipe_restart,
115 };
116 
117 /*
118  * Default pipe buffer size(s), this can be kind-of large now because pipe
119  * space is pageable.  The pipe code will try to maintain locality of
120  * reference for performance reasons, so small amounts of outstanding I/O
121  * will not wipe the cache.
122  */
123 #define	MINPIPESIZE	(PIPE_SIZE / 3)
124 #define	MAXPIPESIZE	(2 * PIPE_SIZE / 3)
125 
126 /*
127  * Limit the number of "big" pipes
128  */
129 #define	LIMITBIGPIPES	32
130 static u_int	maxbigpipes = LIMITBIGPIPES;
131 static u_int	nbigpipe = 0;
132 
133 /*
134  * Amount of KVA consumed by pipe buffers.
135  */
136 static u_int	amountpipekva = 0;
137 
138 static void	pipeclose(struct pipe *);
139 static void	pipe_free_kmem(struct pipe *);
140 static int	pipe_create(struct pipe **, pool_cache_t);
141 static int	pipelock(struct pipe *, bool);
142 static inline void pipeunlock(struct pipe *);
143 static void	pipeselwakeup(struct pipe *, struct pipe *, int);
144 static int	pipespace(struct pipe *, int);
145 static int	pipe_ctor(void *, void *, int);
146 static void	pipe_dtor(void *, void *);
147 
148 static pool_cache_t	pipe_wr_cache;
149 static pool_cache_t	pipe_rd_cache;
150 
151 void
152 pipe_init(void)
153 {
154 
155 	/* Writer side is not automatically allocated KVA. */
156 	pipe_wr_cache = pool_cache_init(sizeof(struct pipe), 0, 0, 0, "pipewr",
157 	    NULL, IPL_NONE, pipe_ctor, pipe_dtor, NULL);
158 	KASSERT(pipe_wr_cache != NULL);
159 
160 	/* Reader side gets preallocated KVA. */
161 	pipe_rd_cache = pool_cache_init(sizeof(struct pipe), 0, 0, 0, "piperd",
162 	    NULL, IPL_NONE, pipe_ctor, pipe_dtor, (void *)1);
163 	KASSERT(pipe_rd_cache != NULL);
164 }
165 
166 static int
167 pipe_ctor(void *arg, void *obj, int flags)
168 {
169 	struct pipe *pipe;
170 	vaddr_t va;
171 
172 	pipe = obj;
173 
174 	memset(pipe, 0, sizeof(struct pipe));
175 	if (arg != NULL) {
176 		/* Preallocate space. */
177 		va = uvm_km_alloc(kernel_map, PIPE_SIZE, 0,
178 		    UVM_KMF_PAGEABLE | UVM_KMF_WAITVA);
179 		KASSERT(va != 0);
180 		pipe->pipe_kmem = va;
181 		atomic_add_int(&amountpipekva, PIPE_SIZE);
182 	}
183 	cv_init(&pipe->pipe_rcv, "pipe_rd");
184 	cv_init(&pipe->pipe_wcv, "pipe_wr");
185 	cv_init(&pipe->pipe_draincv, "pipe_drn");
186 	cv_init(&pipe->pipe_lkcv, "pipe_lk");
187 	selinit(&pipe->pipe_sel);
188 	pipe->pipe_state = PIPE_SIGNALR;
189 
190 	return 0;
191 }
192 
193 static void
194 pipe_dtor(void *arg, void *obj)
195 {
196 	struct pipe *pipe;
197 
198 	pipe = obj;
199 
200 	cv_destroy(&pipe->pipe_rcv);
201 	cv_destroy(&pipe->pipe_wcv);
202 	cv_destroy(&pipe->pipe_draincv);
203 	cv_destroy(&pipe->pipe_lkcv);
204 	seldestroy(&pipe->pipe_sel);
205 	if (pipe->pipe_kmem != 0) {
206 		uvm_km_free(kernel_map, pipe->pipe_kmem, PIPE_SIZE,
207 		    UVM_KMF_PAGEABLE);
208 		atomic_add_int(&amountpipekva, -PIPE_SIZE);
209 	}
210 }
211 
212 /*
213  * The pipe system call for the DTYPE_PIPE type of pipes
214  */
215 int
216 pipe1(struct lwp *l, int *fildes, int flags)
217 {
218 	struct pipe *rpipe, *wpipe;
219 	file_t *rf, *wf;
220 	int fd, error;
221 	proc_t *p;
222 
223 	if (flags & ~(O_CLOEXEC|O_NONBLOCK|O_NOSIGPIPE))
224 		return EINVAL;
225 	p = curproc;
226 	rpipe = wpipe = NULL;
227 	if ((error = pipe_create(&rpipe, pipe_rd_cache)) ||
228 	    (error = pipe_create(&wpipe, pipe_wr_cache))) {
229 		goto free2;
230 	}
231 	rpipe->pipe_lock = mutex_obj_alloc(MUTEX_DEFAULT, IPL_NONE);
232 	wpipe->pipe_lock = rpipe->pipe_lock;
233 	mutex_obj_hold(wpipe->pipe_lock);
234 
235 	error = fd_allocfile(&rf, &fd);
236 	if (error)
237 		goto free2;
238 	fildes[0] = fd;
239 
240 	error = fd_allocfile(&wf, &fd);
241 	if (error)
242 		goto free3;
243 	fildes[1] = fd;
244 
245 	rf->f_flag = FREAD | flags;
246 	rf->f_type = DTYPE_PIPE;
247 	rf->f_pipe = rpipe;
248 	rf->f_ops = &pipeops;
249 	fd_set_exclose(l, fildes[0], (flags & O_CLOEXEC) != 0);
250 
251 	wf->f_flag = FWRITE | flags;
252 	wf->f_type = DTYPE_PIPE;
253 	wf->f_pipe = wpipe;
254 	wf->f_ops = &pipeops;
255 	fd_set_exclose(l, fildes[1], (flags & O_CLOEXEC) != 0);
256 
257 	rpipe->pipe_peer = wpipe;
258 	wpipe->pipe_peer = rpipe;
259 
260 	fd_affix(p, rf, fildes[0]);
261 	fd_affix(p, wf, fildes[1]);
262 	return (0);
263 free3:
264 	fd_abort(p, rf, fildes[0]);
265 free2:
266 	pipeclose(wpipe);
267 	pipeclose(rpipe);
268 
269 	return (error);
270 }
271 
272 /*
273  * Allocate kva for pipe circular buffer, the space is pageable
274  * This routine will 'realloc' the size of a pipe safely, if it fails
275  * it will retain the old buffer.
276  * If it fails it will return ENOMEM.
277  */
278 static int
279 pipespace(struct pipe *pipe, int size)
280 {
281 	void *buffer;
282 
283 	/*
284 	 * Allocate pageable virtual address space.  Physical memory is
285 	 * allocated on demand.
286 	 */
287 	if (size == PIPE_SIZE && pipe->pipe_kmem != 0) {
288 		buffer = (void *)pipe->pipe_kmem;
289 	} else {
290 		buffer = (void *)uvm_km_alloc(kernel_map, round_page(size),
291 		    0, UVM_KMF_PAGEABLE);
292 		if (buffer == NULL)
293 			return (ENOMEM);
294 		atomic_add_int(&amountpipekva, size);
295 	}
296 
297 	/* free old resources if we're resizing */
298 	pipe_free_kmem(pipe);
299 	pipe->pipe_buffer.buffer = buffer;
300 	pipe->pipe_buffer.size = size;
301 	pipe->pipe_buffer.in = 0;
302 	pipe->pipe_buffer.out = 0;
303 	pipe->pipe_buffer.cnt = 0;
304 	return (0);
305 }
306 
307 /*
308  * Initialize and allocate VM and memory for pipe.
309  */
310 static int
311 pipe_create(struct pipe **pipep, pool_cache_t cache)
312 {
313 	struct pipe *pipe;
314 	int error;
315 
316 	pipe = pool_cache_get(cache, PR_WAITOK);
317 	KASSERT(pipe != NULL);
318 	*pipep = pipe;
319 	error = 0;
320 	getnanotime(&pipe->pipe_btime);
321 	pipe->pipe_atime = pipe->pipe_mtime = pipe->pipe_btime;
322 	pipe->pipe_lock = NULL;
323 	if (cache == pipe_rd_cache) {
324 		error = pipespace(pipe, PIPE_SIZE);
325 	} else {
326 		pipe->pipe_buffer.buffer = NULL;
327 		pipe->pipe_buffer.size = 0;
328 		pipe->pipe_buffer.in = 0;
329 		pipe->pipe_buffer.out = 0;
330 		pipe->pipe_buffer.cnt = 0;
331 	}
332 	return error;
333 }
334 
335 /*
336  * Lock a pipe for I/O, blocking other access
337  * Called with pipe spin lock held.
338  */
339 static int
340 pipelock(struct pipe *pipe, bool catch_p)
341 {
342 	int error;
343 
344 	KASSERT(mutex_owned(pipe->pipe_lock));
345 
346 	while (pipe->pipe_state & PIPE_LOCKFL) {
347 		pipe->pipe_waiters++;
348 		KASSERT(pipe->pipe_waiters != 0); /* just in case */
349 		if (catch_p) {
350 			error = cv_wait_sig(&pipe->pipe_lkcv, pipe->pipe_lock);
351 			if (error != 0) {
352 				KASSERT(pipe->pipe_waiters > 0);
353 				pipe->pipe_waiters--;
354 				return error;
355 			}
356 		} else
357 			cv_wait(&pipe->pipe_lkcv, pipe->pipe_lock);
358 		KASSERT(pipe->pipe_waiters > 0);
359 		pipe->pipe_waiters--;
360 	}
361 
362 	pipe->pipe_state |= PIPE_LOCKFL;
363 
364 	return 0;
365 }
366 
367 /*
368  * unlock a pipe I/O lock
369  */
370 static inline void
371 pipeunlock(struct pipe *pipe)
372 {
373 
374 	KASSERT(pipe->pipe_state & PIPE_LOCKFL);
375 
376 	pipe->pipe_state &= ~PIPE_LOCKFL;
377 	if (pipe->pipe_waiters > 0) {
378 		cv_signal(&pipe->pipe_lkcv);
379 	}
380 }
381 
382 /*
383  * Select/poll wakup. This also sends SIGIO to peer connected to
384  * 'sigpipe' side of pipe.
385  */
386 static void
387 pipeselwakeup(struct pipe *selp, struct pipe *sigp, int code)
388 {
389 	int band;
390 
391 	switch (code) {
392 	case POLL_IN:
393 		band = POLLIN|POLLRDNORM;
394 		break;
395 	case POLL_OUT:
396 		band = POLLOUT|POLLWRNORM;
397 		break;
398 	case POLL_HUP:
399 		band = POLLHUP;
400 		break;
401 	case POLL_ERR:
402 		band = POLLERR;
403 		break;
404 	default:
405 		band = 0;
406 #ifdef DIAGNOSTIC
407 		printf("bad siginfo code %d in pipe notification.\n", code);
408 #endif
409 		break;
410 	}
411 
412 	selnotify(&selp->pipe_sel, band, NOTE_SUBMIT);
413 
414 	if (sigp == NULL || (sigp->pipe_state & PIPE_ASYNC) == 0)
415 		return;
416 
417 	fownsignal(sigp->pipe_pgid, SIGIO, code, band, selp);
418 }
419 
420 static int
421 pipe_read(file_t *fp, off_t *offset, struct uio *uio, kauth_cred_t cred,
422     int flags)
423 {
424 	struct pipe *rpipe = fp->f_pipe;
425 	struct pipebuf *bp = &rpipe->pipe_buffer;
426 	kmutex_t *lock = rpipe->pipe_lock;
427 	int error;
428 	size_t nread = 0;
429 	size_t size;
430 	size_t ocnt;
431 	unsigned int wakeup_state = 0;
432 
433 	mutex_enter(lock);
434 	++rpipe->pipe_busy;
435 	ocnt = bp->cnt;
436 
437 again:
438 	error = pipelock(rpipe, true);
439 	if (error)
440 		goto unlocked_error;
441 
442 	while (uio->uio_resid) {
443 		/*
444 		 * Normal pipe buffer receive.
445 		 */
446 		if (bp->cnt > 0) {
447 			size = bp->size - bp->out;
448 			if (size > bp->cnt)
449 				size = bp->cnt;
450 			if (size > uio->uio_resid)
451 				size = uio->uio_resid;
452 
453 			mutex_exit(lock);
454 			error = uiomove((char *)bp->buffer + bp->out, size, uio);
455 			mutex_enter(lock);
456 			if (error)
457 				break;
458 
459 			bp->out += size;
460 			if (bp->out >= bp->size)
461 				bp->out = 0;
462 
463 			bp->cnt -= size;
464 
465 			/*
466 			 * If there is no more to read in the pipe, reset
467 			 * its pointers to the beginning.  This improves
468 			 * cache hit stats.
469 			 */
470 			if (bp->cnt == 0) {
471 				bp->in = 0;
472 				bp->out = 0;
473 			}
474 			nread += size;
475 			continue;
476 		}
477 
478 		/*
479 		 * Break if some data was read.
480 		 */
481 		if (nread > 0)
482 			break;
483 
484 		/*
485 		 * Detect EOF condition.
486 		 * Read returns 0 on EOF, no need to set error.
487 		 */
488 		if (rpipe->pipe_state & PIPE_EOF)
489 			break;
490 
491 		/*
492 		 * Don't block on non-blocking I/O.
493 		 */
494 		if (fp->f_flag & FNONBLOCK) {
495 			error = EAGAIN;
496 			break;
497 		}
498 
499 		/*
500 		 * Unlock the pipe buffer for our remaining processing.
501 		 * We will either break out with an error or we will
502 		 * sleep and relock to loop.
503 		 */
504 		pipeunlock(rpipe);
505 
506 #if 1   /* XXX (dsl) I'm sure these aren't needed here ... */
507 		/*
508 		 * We want to read more, wake up select/poll.
509 		 */
510 		pipeselwakeup(rpipe, rpipe->pipe_peer, POLL_OUT);
511 
512 		/*
513 		 * If the "write-side" is blocked, wake it up now.
514 		 */
515 		cv_broadcast(&rpipe->pipe_wcv);
516 #endif
517 
518 		if (wakeup_state & PIPE_RESTART) {
519 			error = ERESTART;
520 			goto unlocked_error;
521 		}
522 
523 		/* Now wait until the pipe is filled */
524 		error = cv_wait_sig(&rpipe->pipe_rcv, lock);
525 		if (error != 0)
526 			goto unlocked_error;
527 		wakeup_state = rpipe->pipe_state;
528 		goto again;
529 	}
530 
531 	if (error == 0)
532 		getnanotime(&rpipe->pipe_atime);
533 	pipeunlock(rpipe);
534 
535 unlocked_error:
536 	--rpipe->pipe_busy;
537 	if (rpipe->pipe_busy == 0) {
538 		rpipe->pipe_state &= ~PIPE_RESTART;
539 		cv_broadcast(&rpipe->pipe_draincv);
540 	}
541 	if (bp->cnt < MINPIPESIZE) {
542 		cv_broadcast(&rpipe->pipe_wcv);
543 	}
544 
545 	/*
546 	 * If anything was read off the buffer, signal to the writer it's
547 	 * possible to write more data. Also send signal if we are here for the
548 	 * first time after last write.
549 	 */
550 	if ((bp->size - bp->cnt) >= PIPE_BUF
551 	    && (ocnt != bp->cnt || (rpipe->pipe_state & PIPE_SIGNALR))) {
552 		pipeselwakeup(rpipe, rpipe->pipe_peer, POLL_OUT);
553 		rpipe->pipe_state &= ~PIPE_SIGNALR;
554 	}
555 
556 	mutex_exit(lock);
557 	return (error);
558 }
559 
560 static int
561 pipe_write(file_t *fp, off_t *offset, struct uio *uio, kauth_cred_t cred,
562     int flags)
563 {
564 	struct pipe *wpipe, *rpipe;
565 	struct pipebuf *bp;
566 	kmutex_t *lock;
567 	int error;
568 	unsigned int wakeup_state = 0;
569 
570 	/* We want to write to our peer */
571 	rpipe = fp->f_pipe;
572 	lock = rpipe->pipe_lock;
573 	error = 0;
574 
575 	mutex_enter(lock);
576 	wpipe = rpipe->pipe_peer;
577 
578 	/*
579 	 * Detect loss of pipe read side, issue SIGPIPE if lost.
580 	 */
581 	if (wpipe == NULL || (wpipe->pipe_state & PIPE_EOF) != 0) {
582 		mutex_exit(lock);
583 		return EPIPE;
584 	}
585 	++wpipe->pipe_busy;
586 
587 	/* Aquire the long-term pipe lock */
588 	if ((error = pipelock(wpipe, true)) != 0) {
589 		--wpipe->pipe_busy;
590 		if (wpipe->pipe_busy == 0) {
591 			wpipe->pipe_state &= ~PIPE_RESTART;
592 			cv_broadcast(&wpipe->pipe_draincv);
593 		}
594 		mutex_exit(lock);
595 		return (error);
596 	}
597 
598 	bp = &wpipe->pipe_buffer;
599 
600 	/*
601 	 * If it is advantageous to resize the pipe buffer, do so.
602 	 */
603 	if ((uio->uio_resid > PIPE_SIZE) &&
604 	    (nbigpipe < maxbigpipes) &&
605 	    (bp->size <= PIPE_SIZE) && (bp->cnt == 0)) {
606 
607 		if (pipespace(wpipe, BIG_PIPE_SIZE) == 0)
608 			atomic_inc_uint(&nbigpipe);
609 	}
610 
611 	while (uio->uio_resid) {
612 		size_t space;
613 
614 		space = bp->size - bp->cnt;
615 
616 		/* Writes of size <= PIPE_BUF must be atomic. */
617 		if ((space < uio->uio_resid) && (uio->uio_resid <= PIPE_BUF))
618 			space = 0;
619 
620 		if (space > 0) {
621 			int size;	/* Transfer size */
622 			int segsize;	/* first segment to transfer */
623 
624 			/*
625 			 * Transfer size is minimum of uio transfer
626 			 * and free space in pipe buffer.
627 			 */
628 			if (space > uio->uio_resid)
629 				size = uio->uio_resid;
630 			else
631 				size = space;
632 			/*
633 			 * First segment to transfer is minimum of
634 			 * transfer size and contiguous space in
635 			 * pipe buffer.  If first segment to transfer
636 			 * is less than the transfer size, we've got
637 			 * a wraparound in the buffer.
638 			 */
639 			segsize = bp->size - bp->in;
640 			if (segsize > size)
641 				segsize = size;
642 
643 			/* Transfer first segment */
644 			mutex_exit(lock);
645 			error = uiomove((char *)bp->buffer + bp->in, segsize,
646 			    uio);
647 
648 			if (error == 0 && segsize < size) {
649 				/*
650 				 * Transfer remaining part now, to
651 				 * support atomic writes.  Wraparound
652 				 * happened.
653 				 */
654 				KASSERT(bp->in + segsize == bp->size);
655 				error = uiomove(bp->buffer,
656 				    size - segsize, uio);
657 			}
658 			mutex_enter(lock);
659 			if (error)
660 				break;
661 
662 			bp->in += size;
663 			if (bp->in >= bp->size) {
664 				KASSERT(bp->in == size - segsize + bp->size);
665 				bp->in = size - segsize;
666 			}
667 
668 			bp->cnt += size;
669 			KASSERT(bp->cnt <= bp->size);
670 			wakeup_state = 0;
671 		} else {
672 			/*
673 			 * If the "read-side" has been blocked, wake it up now.
674 			 */
675 			cv_broadcast(&wpipe->pipe_rcv);
676 
677 			/*
678 			 * Don't block on non-blocking I/O.
679 			 */
680 			if (fp->f_flag & FNONBLOCK) {
681 				error = EAGAIN;
682 				break;
683 			}
684 
685 			/*
686 			 * We have no more space and have something to offer,
687 			 * wake up select/poll.
688 			 */
689 			if (bp->cnt)
690 				pipeselwakeup(wpipe, wpipe, POLL_IN);
691 
692 			if (wakeup_state & PIPE_RESTART) {
693 				error = ERESTART;
694 				break;
695 			}
696 
697 			pipeunlock(wpipe);
698 			error = cv_wait_sig(&wpipe->pipe_wcv, lock);
699 			(void)pipelock(wpipe, false);
700 			if (error != 0)
701 				break;
702 			/*
703 			 * If read side wants to go away, we just issue a signal
704 			 * to ourselves.
705 			 */
706 			if (wpipe->pipe_state & PIPE_EOF) {
707 				error = EPIPE;
708 				break;
709 			}
710 			wakeup_state = wpipe->pipe_state;
711 		}
712 	}
713 
714 	--wpipe->pipe_busy;
715 	if (wpipe->pipe_busy == 0) {
716 		wpipe->pipe_state &= ~PIPE_RESTART;
717 		cv_broadcast(&wpipe->pipe_draincv);
718 	}
719 	if (bp->cnt > 0) {
720 		cv_broadcast(&wpipe->pipe_rcv);
721 	}
722 
723 	/*
724 	 * Don't return EPIPE if I/O was successful
725 	 */
726 	if (error == EPIPE && bp->cnt == 0 && uio->uio_resid == 0)
727 		error = 0;
728 
729 	if (error == 0)
730 		getnanotime(&wpipe->pipe_mtime);
731 
732 	/*
733 	 * We have something to offer, wake up select/poll.
734 	 * wmap->cnt is always 0 in this point (direct write
735 	 * is only done synchronously), so check only wpipe->pipe_buffer.cnt
736 	 */
737 	if (bp->cnt)
738 		pipeselwakeup(wpipe, wpipe, POLL_IN);
739 
740 	/*
741 	 * Arrange for next read(2) to do a signal.
742 	 */
743 	wpipe->pipe_state |= PIPE_SIGNALR;
744 
745 	pipeunlock(wpipe);
746 	mutex_exit(lock);
747 	return (error);
748 }
749 
750 /*
751  * We implement a very minimal set of ioctls for compatibility with sockets.
752  */
753 int
754 pipe_ioctl(file_t *fp, u_long cmd, void *data)
755 {
756 	struct pipe *pipe = fp->f_pipe;
757 	kmutex_t *lock = pipe->pipe_lock;
758 
759 	switch (cmd) {
760 
761 	case FIONBIO:
762 		return (0);
763 
764 	case FIOASYNC:
765 		mutex_enter(lock);
766 		if (*(int *)data) {
767 			pipe->pipe_state |= PIPE_ASYNC;
768 		} else {
769 			pipe->pipe_state &= ~PIPE_ASYNC;
770 		}
771 		mutex_exit(lock);
772 		return (0);
773 
774 	case FIONREAD:
775 		mutex_enter(lock);
776 		*(int *)data = pipe->pipe_buffer.cnt;
777 		mutex_exit(lock);
778 		return (0);
779 
780 	case FIONWRITE:
781 		/* Look at other side */
782 		mutex_enter(lock);
783 		pipe = pipe->pipe_peer;
784 		if (pipe == NULL)
785 			*(int *)data = 0;
786 		else
787 			*(int *)data = pipe->pipe_buffer.cnt;
788 		mutex_exit(lock);
789 		return (0);
790 
791 	case FIONSPACE:
792 		/* Look at other side */
793 		mutex_enter(lock);
794 		pipe = pipe->pipe_peer;
795 		if (pipe == NULL)
796 			*(int *)data = 0;
797 		else
798 			*(int *)data = pipe->pipe_buffer.size -
799 			    pipe->pipe_buffer.cnt;
800 		mutex_exit(lock);
801 		return (0);
802 
803 	case TIOCSPGRP:
804 	case FIOSETOWN:
805 		return fsetown(&pipe->pipe_pgid, cmd, data);
806 
807 	case TIOCGPGRP:
808 	case FIOGETOWN:
809 		return fgetown(pipe->pipe_pgid, cmd, data);
810 
811 	}
812 	return (EPASSTHROUGH);
813 }
814 
815 int
816 pipe_poll(file_t *fp, int events)
817 {
818 	struct pipe *rpipe = fp->f_pipe;
819 	struct pipe *wpipe;
820 	int eof = 0;
821 	int revents = 0;
822 
823 	mutex_enter(rpipe->pipe_lock);
824 	wpipe = rpipe->pipe_peer;
825 
826 	if (events & (POLLIN | POLLRDNORM))
827 		if ((rpipe->pipe_buffer.cnt > 0) ||
828 		    (rpipe->pipe_state & PIPE_EOF))
829 			revents |= events & (POLLIN | POLLRDNORM);
830 
831 	eof |= (rpipe->pipe_state & PIPE_EOF);
832 
833 	if (wpipe == NULL)
834 		revents |= events & (POLLOUT | POLLWRNORM);
835 	else {
836 		if (events & (POLLOUT | POLLWRNORM))
837 			if ((wpipe->pipe_state & PIPE_EOF) || (
838 			     (wpipe->pipe_buffer.size - wpipe->pipe_buffer.cnt) >= PIPE_BUF))
839 				revents |= events & (POLLOUT | POLLWRNORM);
840 
841 		eof |= (wpipe->pipe_state & PIPE_EOF);
842 	}
843 
844 	if (wpipe == NULL || eof)
845 		revents |= POLLHUP;
846 
847 	if (revents == 0) {
848 		if (events & (POLLIN | POLLRDNORM))
849 			selrecord(curlwp, &rpipe->pipe_sel);
850 
851 		if (events & (POLLOUT | POLLWRNORM))
852 			selrecord(curlwp, &wpipe->pipe_sel);
853 	}
854 	mutex_exit(rpipe->pipe_lock);
855 
856 	return (revents);
857 }
858 
859 static int
860 pipe_stat(file_t *fp, struct stat *ub)
861 {
862 	struct pipe *pipe = fp->f_pipe;
863 
864 	mutex_enter(pipe->pipe_lock);
865 	memset(ub, 0, sizeof(*ub));
866 	ub->st_mode = S_IFIFO | S_IRUSR | S_IWUSR;
867 	ub->st_blksize = pipe->pipe_buffer.size;
868 	if (ub->st_blksize == 0 && pipe->pipe_peer)
869 		ub->st_blksize = pipe->pipe_peer->pipe_buffer.size;
870 	ub->st_size = pipe->pipe_buffer.cnt;
871 	ub->st_blocks = (ub->st_size) ? 1 : 0;
872 	ub->st_atimespec = pipe->pipe_atime;
873 	ub->st_mtimespec = pipe->pipe_mtime;
874 	ub->st_ctimespec = ub->st_birthtimespec = pipe->pipe_btime;
875 	ub->st_uid = kauth_cred_geteuid(fp->f_cred);
876 	ub->st_gid = kauth_cred_getegid(fp->f_cred);
877 
878 	/*
879 	 * Left as 0: st_dev, st_ino, st_nlink, st_rdev, st_flags, st_gen.
880 	 * XXX (st_dev, st_ino) should be unique.
881 	 */
882 	mutex_exit(pipe->pipe_lock);
883 	return 0;
884 }
885 
886 static int
887 pipe_close(file_t *fp)
888 {
889 	struct pipe *pipe = fp->f_pipe;
890 
891 	fp->f_pipe = NULL;
892 	pipeclose(pipe);
893 	return (0);
894 }
895 
896 static void
897 pipe_restart(file_t *fp)
898 {
899 	struct pipe *pipe = fp->f_pipe;
900 
901 	/*
902 	 * Unblock blocked reads/writes in order to allow close() to complete.
903 	 * System calls return ERESTART so that the fd is revalidated.
904 	 * (Partial writes return the transfer length.)
905 	 */
906 	mutex_enter(pipe->pipe_lock);
907 	pipe->pipe_state |= PIPE_RESTART;
908 	/* Wakeup both cvs, maybe we only need one, but maybe there are some
909 	 * other paths where wakeup is needed, and it saves deciding which! */
910 	cv_broadcast(&pipe->pipe_rcv);
911 	cv_broadcast(&pipe->pipe_wcv);
912 	mutex_exit(pipe->pipe_lock);
913 }
914 
915 static void
916 pipe_free_kmem(struct pipe *pipe)
917 {
918 
919 	if (pipe->pipe_buffer.buffer != NULL) {
920 		if (pipe->pipe_buffer.size > PIPE_SIZE) {
921 			atomic_dec_uint(&nbigpipe);
922 		}
923 		if (pipe->pipe_buffer.buffer != (void *)pipe->pipe_kmem) {
924 			uvm_km_free(kernel_map,
925 			    (vaddr_t)pipe->pipe_buffer.buffer,
926 			    pipe->pipe_buffer.size, UVM_KMF_PAGEABLE);
927 			atomic_add_int(&amountpipekva,
928 			    -pipe->pipe_buffer.size);
929 		}
930 		pipe->pipe_buffer.buffer = NULL;
931 	}
932 }
933 
934 /*
935  * Shutdown the pipe.
936  */
937 static void
938 pipeclose(struct pipe *pipe)
939 {
940 	kmutex_t *lock;
941 	struct pipe *ppipe;
942 
943 	if (pipe == NULL)
944 		return;
945 
946 	KASSERT(cv_is_valid(&pipe->pipe_rcv));
947 	KASSERT(cv_is_valid(&pipe->pipe_wcv));
948 	KASSERT(cv_is_valid(&pipe->pipe_draincv));
949 	KASSERT(cv_is_valid(&pipe->pipe_lkcv));
950 
951 	lock = pipe->pipe_lock;
952 	if (lock == NULL)
953 		/* Must have failed during create */
954 		goto free_resources;
955 
956 	mutex_enter(lock);
957 	pipeselwakeup(pipe, pipe, POLL_HUP);
958 
959 	/*
960 	 * If the other side is blocked, wake it up saying that
961 	 * we want to close it down.
962 	 */
963 	pipe->pipe_state |= PIPE_EOF;
964 	if (pipe->pipe_busy) {
965 		while (pipe->pipe_busy) {
966 			cv_broadcast(&pipe->pipe_wcv);
967 			cv_wait_sig(&pipe->pipe_draincv, lock);
968 		}
969 	}
970 
971 	/*
972 	 * Disconnect from peer.
973 	 */
974 	if ((ppipe = pipe->pipe_peer) != NULL) {
975 		pipeselwakeup(ppipe, ppipe, POLL_HUP);
976 		ppipe->pipe_state |= PIPE_EOF;
977 		cv_broadcast(&ppipe->pipe_rcv);
978 		ppipe->pipe_peer = NULL;
979 	}
980 
981 	/*
982 	 * Any knote objects still left in the list are
983 	 * the one attached by peer.  Since no one will
984 	 * traverse this list, we just clear it.
985 	 *
986 	 * XXX Exposes select/kqueue internals.
987 	 */
988 	SLIST_INIT(&pipe->pipe_sel.sel_klist);
989 
990 	KASSERT((pipe->pipe_state & PIPE_LOCKFL) == 0);
991 	mutex_exit(lock);
992 	mutex_obj_free(lock);
993 
994 	/*
995 	 * Free resources.
996 	 */
997     free_resources:
998 	pipe->pipe_pgid = 0;
999 	pipe->pipe_state = PIPE_SIGNALR;
1000 	pipe->pipe_peer = NULL;
1001 	pipe->pipe_lock = NULL;
1002 	pipe_free_kmem(pipe);
1003 	if (pipe->pipe_kmem != 0) {
1004 		pool_cache_put(pipe_rd_cache, pipe);
1005 	} else {
1006 		pool_cache_put(pipe_wr_cache, pipe);
1007 	}
1008 }
1009 
1010 static void
1011 filt_pipedetach(struct knote *kn)
1012 {
1013 	struct pipe *pipe;
1014 	kmutex_t *lock;
1015 
1016 	pipe = ((file_t *)kn->kn_obj)->f_pipe;
1017 	lock = pipe->pipe_lock;
1018 
1019 	mutex_enter(lock);
1020 
1021 	switch(kn->kn_filter) {
1022 	case EVFILT_WRITE:
1023 		/* Need the peer structure, not our own. */
1024 		pipe = pipe->pipe_peer;
1025 
1026 		/* If reader end already closed, just return. */
1027 		if (pipe == NULL) {
1028 			mutex_exit(lock);
1029 			return;
1030 		}
1031 
1032 		break;
1033 	default:
1034 		/* Nothing to do. */
1035 		break;
1036 	}
1037 
1038 	KASSERT(kn->kn_hook == pipe);
1039 	selremove_knote(&pipe->pipe_sel, kn);
1040 	mutex_exit(lock);
1041 }
1042 
1043 static int
1044 filt_piperead(struct knote *kn, long hint)
1045 {
1046 	struct pipe *rpipe = ((file_t *)kn->kn_obj)->f_pipe;
1047 	struct pipe *wpipe;
1048 
1049 	if ((hint & NOTE_SUBMIT) == 0) {
1050 		mutex_enter(rpipe->pipe_lock);
1051 	}
1052 	wpipe = rpipe->pipe_peer;
1053 	kn->kn_data = rpipe->pipe_buffer.cnt;
1054 
1055 	if ((rpipe->pipe_state & PIPE_EOF) ||
1056 	    (wpipe == NULL) || (wpipe->pipe_state & PIPE_EOF)) {
1057 		kn->kn_flags |= EV_EOF;
1058 		if ((hint & NOTE_SUBMIT) == 0) {
1059 			mutex_exit(rpipe->pipe_lock);
1060 		}
1061 		return (1);
1062 	}
1063 
1064 	if ((hint & NOTE_SUBMIT) == 0) {
1065 		mutex_exit(rpipe->pipe_lock);
1066 	}
1067 	return (kn->kn_data > 0);
1068 }
1069 
1070 static int
1071 filt_pipewrite(struct knote *kn, long hint)
1072 {
1073 	struct pipe *rpipe = ((file_t *)kn->kn_obj)->f_pipe;
1074 	struct pipe *wpipe;
1075 
1076 	if ((hint & NOTE_SUBMIT) == 0) {
1077 		mutex_enter(rpipe->pipe_lock);
1078 	}
1079 	wpipe = rpipe->pipe_peer;
1080 
1081 	if ((wpipe == NULL) || (wpipe->pipe_state & PIPE_EOF)) {
1082 		kn->kn_data = 0;
1083 		kn->kn_flags |= EV_EOF;
1084 		if ((hint & NOTE_SUBMIT) == 0) {
1085 			mutex_exit(rpipe->pipe_lock);
1086 		}
1087 		return (1);
1088 	}
1089 	kn->kn_data = wpipe->pipe_buffer.size - wpipe->pipe_buffer.cnt;
1090 
1091 	if ((hint & NOTE_SUBMIT) == 0) {
1092 		mutex_exit(rpipe->pipe_lock);
1093 	}
1094 	return (kn->kn_data >= PIPE_BUF);
1095 }
1096 
1097 static const struct filterops pipe_rfiltops = {
1098 	.f_isfd = 1,
1099 	.f_attach = NULL,
1100 	.f_detach = filt_pipedetach,
1101 	.f_event = filt_piperead,
1102 };
1103 
1104 static const struct filterops pipe_wfiltops = {
1105 	.f_isfd = 1,
1106 	.f_attach = NULL,
1107 	.f_detach = filt_pipedetach,
1108 	.f_event = filt_pipewrite,
1109 };
1110 
1111 static int
1112 pipe_kqfilter(file_t *fp, struct knote *kn)
1113 {
1114 	struct pipe *pipe;
1115 	kmutex_t *lock;
1116 
1117 	pipe = ((file_t *)kn->kn_obj)->f_pipe;
1118 	lock = pipe->pipe_lock;
1119 
1120 	mutex_enter(lock);
1121 
1122 	switch (kn->kn_filter) {
1123 	case EVFILT_READ:
1124 		kn->kn_fop = &pipe_rfiltops;
1125 		break;
1126 	case EVFILT_WRITE:
1127 		kn->kn_fop = &pipe_wfiltops;
1128 		pipe = pipe->pipe_peer;
1129 		if (pipe == NULL) {
1130 			/* Other end of pipe has been closed. */
1131 			mutex_exit(lock);
1132 			return (EBADF);
1133 		}
1134 		break;
1135 	default:
1136 		mutex_exit(lock);
1137 		return (EINVAL);
1138 	}
1139 
1140 	kn->kn_hook = pipe;
1141 	selrecord_knote(&pipe->pipe_sel, kn);
1142 	mutex_exit(lock);
1143 
1144 	return (0);
1145 }
1146 
1147 /*
1148  * Handle pipe sysctls.
1149  */
1150 SYSCTL_SETUP(sysctl_kern_pipe_setup, "sysctl kern.pipe subtree setup")
1151 {
1152 
1153 	sysctl_createv(clog, 0, NULL, NULL,
1154 		       CTLFLAG_PERMANENT,
1155 		       CTLTYPE_NODE, "pipe",
1156 		       SYSCTL_DESCR("Pipe settings"),
1157 		       NULL, 0, NULL, 0,
1158 		       CTL_KERN, KERN_PIPE, CTL_EOL);
1159 
1160 	sysctl_createv(clog, 0, NULL, NULL,
1161 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1162 		       CTLTYPE_INT, "maxbigpipes",
1163 		       SYSCTL_DESCR("Maximum number of \"big\" pipes"),
1164 		       NULL, 0, &maxbigpipes, 0,
1165 		       CTL_KERN, KERN_PIPE, KERN_PIPE_MAXBIGPIPES, CTL_EOL);
1166 	sysctl_createv(clog, 0, NULL, NULL,
1167 		       CTLFLAG_PERMANENT,
1168 		       CTLTYPE_INT, "nbigpipes",
1169 		       SYSCTL_DESCR("Number of \"big\" pipes"),
1170 		       NULL, 0, &nbigpipe, 0,
1171 		       CTL_KERN, KERN_PIPE, KERN_PIPE_NBIGPIPES, CTL_EOL);
1172 	sysctl_createv(clog, 0, NULL, NULL,
1173 		       CTLFLAG_PERMANENT,
1174 		       CTLTYPE_INT, "kvasize",
1175 		       SYSCTL_DESCR("Amount of kernel memory consumed by pipe "
1176 				    "buffers"),
1177 		       NULL, 0, &amountpipekva, 0,
1178 		       CTL_KERN, KERN_PIPE, KERN_PIPE_KVASIZE, CTL_EOL);
1179 }
1180