Now available: The Design and Implementation of the FreeBSD Operating System (Second Edition)
[ source navigation ] [ diff markup ] [ identifier search ] [ freetext search ] [ file search ] [ list types ] [ track identifier ]
FreeBSD/Linux Kernel Cross Reference
sys/kern/sys_pipe.c
Version:
- FREEBSD - FREEBSD-13-STABLE - FREEBSD-13-0 - FREEBSD-12-STABLE - FREEBSD-12-0 - FREEBSD-11-STABLE - FREEBSD-11-0 - FREEBSD-10-STABLE - FREEBSD-10-0 - FREEBSD-9-STABLE - FREEBSD-9-0 - FREEBSD-8-STABLE - FREEBSD-8-0 - FREEBSD-7-STABLE - FREEBSD-7-0 - FREEBSD-6-STABLE - FREEBSD-6-0 - FREEBSD-5-STABLE - FREEBSD-5-0 - FREEBSD-4-STABLE - FREEBSD-3-STABLE - FREEBSD22 - l41 - OPENBSD - linux-2.6 - MK84 - PLAN9 - xnu-8792
SearchContext: - none - 3 - 10
SearchContext: - none - 3 - 10
1 /* $NetBSD: sys_pipe.c,v 1.55.2.2 2004/07/23 15:49:46 tron Exp $ */ 2 3 /*- 4 * Copyright (c) 2003 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. 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 * 3. All advertising materials mentioning features or use of this software 19 * must display the following acknowledgement: 20 * This product includes software developed by the NetBSD 21 * Foundation, Inc. and its contributors. 22 * 4. Neither the name of The NetBSD Foundation nor the names of its 23 * contributors may be used to endorse or promote products derived 24 * from this software without specific prior written permission. 25 * 26 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 27 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 28 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 29 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 30 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 31 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 32 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 33 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 34 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 36 * POSSIBILITY OF SUCH DAMAGE. 37 */ 38 39 /* 40 * Copyright (c) 1996 John S. Dyson 41 * All rights reserved. 42 * 43 * Redistribution and use in source and binary forms, with or without 44 * modification, are permitted provided that the following conditions 45 * are met: 46 * 1. Redistributions of source code must retain the above copyright 47 * notice immediately at the beginning of the file, without modification, 48 * this list of conditions, and the following disclaimer. 49 * 2. Redistributions in binary form must reproduce the above copyright 50 * notice, this list of conditions and the following disclaimer in the 51 * documentation and/or other materials provided with the distribution. 52 * 3. Absolutely no warranty of function or purpose is made by the author 53 * John S. Dyson. 54 * 4. Modifications may be freely made to this file if the above conditions 55 * are met. 56 * 57 * $FreeBSD: src/sys/kern/sys_pipe.c,v 1.95 2002/03/09 22:06:31 alfred Exp $ 58 */ 59 60 /* 61 * This file contains a high-performance replacement for the socket-based 62 * pipes scheme originally used in FreeBSD/4.4Lite. It does not support 63 * all features of sockets, but does do everything that pipes normally 64 * do. 65 * 66 * Adaption for NetBSD UVM, including uvm_loan() based direct write, was 67 * written by Jaromir Dolecek. 68 */ 69 70 /* 71 * This code has two modes of operation, a small write mode and a large 72 * write mode. The small write mode acts like conventional pipes with 73 * a kernel buffer. If the buffer is less than PIPE_MINDIRECT, then the 74 * "normal" pipe buffering is done. If the buffer is between PIPE_MINDIRECT 75 * and PIPE_SIZE in size it is mapped read-only into the kernel address space 76 * using the UVM page loan facility from where the receiving process can copy 77 * the data directly from the pages in the sending process. 78 * 79 * The constant PIPE_MINDIRECT is chosen to make sure that buffering will 80 * happen for small transfers so that the system will not spend all of 81 * its time context switching. PIPE_SIZE is constrained by the 82 * amount of kernel virtual memory. 83 */ 84 85 #include <sys/cdefs.h> 86 __KERNEL_RCSID(0, "$NetBSD: sys_pipe.c,v 1.55.2.2 2004/07/23 15:49:46 tron Exp $"); 87 88 #include <sys/param.h> 89 #include <sys/systm.h> 90 #include <sys/proc.h> 91 #include <sys/fcntl.h> 92 #include <sys/file.h> 93 #include <sys/filedesc.h> 94 #include <sys/filio.h> 95 #include <sys/kernel.h> 96 #include <sys/lock.h> 97 #include <sys/ttycom.h> 98 #include <sys/stat.h> 99 #include <sys/malloc.h> 100 #include <sys/poll.h> 101 #include <sys/signalvar.h> 102 #include <sys/vnode.h> 103 #include <sys/uio.h> 104 #include <sys/lock.h> 105 #include <sys/select.h> 106 #include <sys/mount.h> 107 #include <sys/sa.h> 108 #include <sys/syscallargs.h> 109 #include <uvm/uvm.h> 110 #include <sys/sysctl.h> 111 #include <sys/kernel.h> 112 113 #include <sys/pipe.h> 114 115 /* 116 * Avoid microtime(9), it's slow. We don't guard the read from time(9) 117 * with splclock(9) since we don't actually need to be THAT sure the access 118 * is atomic. 119 */ 120 #define PIPE_TIMESTAMP(tvp) (*(tvp) = time) 121 122 123 /* 124 * Use this define if you want to disable *fancy* VM things. Expect an 125 * approx 30% decrease in transfer rate. 126 */ 127 /* #define PIPE_NODIRECT */ 128 129 /* 130 * interfaces to the outside world 131 */ 132 static int pipe_read(struct file *fp, off_t *offset, struct uio *uio, 133 struct ucred *cred, int flags); 134 static int pipe_write(struct file *fp, off_t *offset, struct uio *uio, 135 struct ucred *cred, int flags); 136 static int pipe_close(struct file *fp, struct proc *p); 137 static int pipe_poll(struct file *fp, int events, struct proc *p); 138 static int pipe_fcntl(struct file *fp, u_int com, void *data, 139 struct proc *p); 140 static int pipe_kqfilter(struct file *fp, struct knote *kn); 141 static int pipe_stat(struct file *fp, struct stat *sb, struct proc *p); 142 static int pipe_ioctl(struct file *fp, u_long cmd, void *data, 143 struct proc *p); 144 145 static struct fileops pipeops = { 146 pipe_read, pipe_write, pipe_ioctl, pipe_fcntl, pipe_poll, 147 pipe_stat, pipe_close, pipe_kqfilter 148 }; 149 150 /* 151 * Default pipe buffer size(s), this can be kind-of large now because pipe 152 * space is pageable. The pipe code will try to maintain locality of 153 * reference for performance reasons, so small amounts of outstanding I/O 154 * will not wipe the cache. 155 */ 156 #define MINPIPESIZE (PIPE_SIZE/3) 157 #define MAXPIPESIZE (2*PIPE_SIZE/3) 158 159 /* 160 * Maximum amount of kva for pipes -- this is kind-of a soft limit, but 161 * is there so that on large systems, we don't exhaust it. 162 */ 163 #define MAXPIPEKVA (8*1024*1024) 164 static int maxpipekva = MAXPIPEKVA; 165 166 /* 167 * Limit for direct transfers, we cannot, of course limit 168 * the amount of kva for pipes in general though. 169 */ 170 #define LIMITPIPEKVA (16*1024*1024) 171 static int limitpipekva = LIMITPIPEKVA; 172 173 /* 174 * Limit the number of "big" pipes 175 */ 176 #define LIMITBIGPIPES 32 177 static int maxbigpipes = LIMITBIGPIPES; 178 static int nbigpipe = 0; 179 180 /* 181 * Amount of KVA consumed by pipe buffers. 182 */ 183 static int amountpipekva = 0; 184 185 MALLOC_DEFINE(M_PIPE, "pipe", "Pipe structures"); 186 187 static void pipeclose(struct file *fp, struct pipe *pipe); 188 static void pipe_free_kmem(struct pipe *pipe); 189 static int pipe_create(struct pipe **pipep, int allockva); 190 static int pipelock(struct pipe *pipe, int catch); 191 static __inline void pipeunlock(struct pipe *pipe); 192 static void pipeselwakeup(struct pipe *pipe, struct pipe *sigp, void *data, 193 int code); 194 #ifndef PIPE_NODIRECT 195 static int pipe_direct_write(struct file *fp, struct pipe *wpipe, 196 struct uio *uio); 197 #endif 198 static int pipespace(struct pipe *pipe, int size); 199 200 #ifndef PIPE_NODIRECT 201 static int pipe_loan_alloc(struct pipe *, int); 202 static void pipe_loan_free(struct pipe *); 203 #endif /* PIPE_NODIRECT */ 204 205 static struct pool pipe_pool; 206 207 /* 208 * The pipe system call for the DTYPE_PIPE type of pipes 209 */ 210 211 /* ARGSUSED */ 212 int 213 sys_pipe(l, v, retval) 214 struct lwp *l; 215 void *v; 216 register_t *retval; 217 { 218 struct file *rf, *wf; 219 struct pipe *rpipe, *wpipe; 220 int fd, error; 221 struct proc *p; 222 223 p = l->l_proc; 224 rpipe = wpipe = NULL; 225 if (pipe_create(&rpipe, 1) || pipe_create(&wpipe, 0)) { 226 pipeclose(NULL, rpipe); 227 pipeclose(NULL, wpipe); 228 return (ENFILE); 229 } 230 231 /* 232 * Note: the file structure returned from falloc() is marked 233 * as 'larval' initially. Unless we mark it as 'mature' by 234 * FILE_SET_MATURE(), any attempt to do anything with it would 235 * return EBADF, including e.g. dup(2) or close(2). This avoids 236 * file descriptor races if we block in the second falloc(). 237 */ 238 239 error = falloc(p, &rf, &fd); 240 if (error) 241 goto free2; 242 retval[0] = fd; 243 rf->f_flag = FREAD; 244 rf->f_type = DTYPE_PIPE; 245 rf->f_data = (caddr_t)rpipe; 246 rf->f_ops = &pipeops; 247 248 error = falloc(p, &wf, &fd); 249 if (error) 250 goto free3; 251 retval[1] = fd; 252 wf->f_flag = FWRITE; 253 wf->f_type = DTYPE_PIPE; 254 wf->f_data = (caddr_t)wpipe; 255 wf->f_ops = &pipeops; 256 257 rpipe->pipe_peer = wpipe; 258 wpipe->pipe_peer = rpipe; 259 260 FILE_SET_MATURE(rf); 261 FILE_SET_MATURE(wf); 262 FILE_UNUSE(rf, p); 263 FILE_UNUSE(wf, p); 264 return (0); 265 free3: 266 FILE_UNUSE(rf, p); 267 ffree(rf); 268 fdremove(p->p_fd, retval[0]); 269 free2: 270 pipeclose(NULL, wpipe); 271 pipeclose(NULL, rpipe); 272 273 return (error); 274 } 275 276 /* 277 * Allocate kva for pipe circular buffer, the space is pageable 278 * This routine will 'realloc' the size of a pipe safely, if it fails 279 * it will retain the old buffer. 280 * If it fails it will return ENOMEM. 281 */ 282 static int 283 pipespace(pipe, size) 284 struct pipe *pipe; 285 int size; 286 { 287 caddr_t buffer; 288 /* 289 * Allocate pageable virtual address space. Physical memory is 290 * allocated on demand. 291 */ 292 buffer = (caddr_t) uvm_km_valloc(kernel_map, round_page(size)); 293 if (buffer == NULL) 294 return (ENOMEM); 295 296 /* free old resources if we're resizing */ 297 pipe_free_kmem(pipe); 298 pipe->pipe_buffer.buffer = buffer; 299 pipe->pipe_buffer.size = size; 300 pipe->pipe_buffer.in = 0; 301 pipe->pipe_buffer.out = 0; 302 pipe->pipe_buffer.cnt = 0; 303 amountpipekva += pipe->pipe_buffer.size; 304 return (0); 305 } 306 307 /* 308 * Initialize and allocate VM and memory for pipe. 309 */ 310 static int 311 pipe_create(pipep, allockva) 312 struct pipe **pipep; 313 int allockva; 314 { 315 struct pipe *pipe; 316 int error; 317 318 pipe = *pipep = pool_get(&pipe_pool, PR_WAITOK); 319 320 /* Initialize */ 321 memset(pipe, 0, sizeof(struct pipe)); 322 pipe->pipe_state = PIPE_SIGNALR; 323 324 PIPE_TIMESTAMP(&pipe->pipe_ctime); 325 pipe->pipe_atime = pipe->pipe_ctime; 326 pipe->pipe_mtime = pipe->pipe_ctime; 327 simple_lock_init(&pipe->pipe_slock); 328 lockinit(&pipe->pipe_lock, PSOCK | PCATCH, "pipelk", 0, 0); 329 330 if (allockva && (error = pipespace(pipe, PIPE_SIZE))) 331 return (error); 332 333 return (0); 334 } 335 336 337 /* 338 * Lock a pipe for I/O, blocking other access 339 * Called with pipe spin lock held. 340 * Return with pipe spin lock released on success. 341 */ 342 static int 343 pipelock(pipe, catch) 344 struct pipe *pipe; 345 int catch; 346 { 347 int error; 348 349 LOCK_ASSERT(simple_lock_held(&pipe->pipe_slock)); 350 351 while (1) { 352 error = lockmgr(&pipe->pipe_lock, LK_EXCLUSIVE | LK_INTERLOCK, 353 &pipe->pipe_slock); 354 if (error == 0) 355 break; 356 357 simple_lock(&pipe->pipe_slock); 358 if (catch || (error != EINTR && error != ERESTART)) 359 break; 360 /* 361 * XXX XXX XXX 362 * The pipe lock is initialised with PCATCH on and we cannot 363 * override this in a lockmgr() call. Thus a pending signal 364 * will cause lockmgr() to return with EINTR or ERESTART. 365 * We cannot simply re-enter lockmgr() at this point since 366 * the pending signals have not yet been posted and would 367 * cause an immediate EINTR/ERESTART return again. 368 * As a workaround we pause for a while here, giving the lock 369 * a chance to drain, before trying again. 370 * XXX XXX XXX 371 * 372 * NOTE: Consider dropping PCATCH from this lock; in practice 373 * it is never held for long enough periods for having it 374 * interruptable at the start of pipe_read/pipe_write to be 375 * beneficial. 376 */ 377 (void) ltsleep(&lbolt, PSOCK, "rstrtpipelock", hz, 378 &pipe->pipe_slock); 379 } 380 return (error); 381 } 382 383 /* 384 * unlock a pipe I/O lock 385 */ 386 static __inline void 387 pipeunlock(pipe) 388 struct pipe *pipe; 389 { 390 391 lockmgr(&pipe->pipe_lock, LK_RELEASE, NULL); 392 } 393 394 /* 395 * Select/poll wakup. This also sends SIGIO to peer connected to 396 * 'sigpipe' side of pipe. 397 */ 398 static void 399 pipeselwakeup(selp, sigp, data, code) 400 struct pipe *selp, *sigp; 401 void *data; 402 int code; 403 { 404 int band; 405 406 selnotify(&selp->pipe_sel, NOTE_SUBMIT); 407 408 if (sigp == NULL || (sigp->pipe_state & PIPE_ASYNC) == 0) 409 return; 410 411 switch (code) { 412 case POLL_IN: 413 band = POLLIN|POLLRDNORM; 414 break; 415 case POLL_OUT: 416 band = POLLOUT|POLLWRNORM; 417 break; 418 case POLL_HUP: 419 band = POLLHUP; 420 break; 421 #if POLL_HUP != POLL_ERR 422 case POLL_ERR: 423 band = POLLERR; 424 break; 425 #endif 426 default: 427 band = 0; 428 #ifdef DIAGNOSTIC 429 printf("bad siginfo code %d in pipe notification.\n", code); 430 #endif 431 break; 432 } 433 434 fownsignal(sigp->pipe_pgid, SIGIO, code, band, selp); 435 } 436 437 /* ARGSUSED */ 438 static int 439 pipe_read(fp, offset, uio, cred, flags) 440 struct file *fp; 441 off_t *offset; 442 struct uio *uio; 443 struct ucred *cred; 444 int flags; 445 { 446 struct pipe *rpipe = (struct pipe *) fp->f_data; 447 struct pipebuf *bp = &rpipe->pipe_buffer; 448 int error; 449 size_t nread = 0; 450 size_t size; 451 size_t ocnt; 452 453 PIPE_LOCK(rpipe); 454 ++rpipe->pipe_busy; 455 ocnt = bp->cnt; 456 457 again: 458 error = pipelock(rpipe, 1); 459 if (error) 460 goto unlocked_error; 461 462 while (uio->uio_resid) { 463 /* 464 * normal pipe buffer receive 465 */ 466 if (bp->cnt > 0) { 467 size = bp->size - bp->out; 468 if (size > bp->cnt) 469 size = bp->cnt; 470 if (size > uio->uio_resid) 471 size = uio->uio_resid; 472 473 error = uiomove(&bp->buffer[bp->out], size, uio); 474 if (error) 475 break; 476 477 bp->out += size; 478 if (bp->out >= bp->size) 479 bp->out = 0; 480 481 bp->cnt -= size; 482 483 /* 484 * If there is no more to read in the pipe, reset 485 * its pointers to the beginning. This improves 486 * cache hit stats. 487 */ 488 if (bp->cnt == 0) { 489 bp->in = 0; 490 bp->out = 0; 491 } 492 nread += size; 493 #ifndef PIPE_NODIRECT 494 } else if ((rpipe->pipe_state & PIPE_DIRECTR) != 0) { 495 /* 496 * Direct copy, bypassing a kernel buffer. 497 */ 498 caddr_t va; 499 500 KASSERT(rpipe->pipe_state & PIPE_DIRECTW); 501 502 size = rpipe->pipe_map.cnt; 503 if (size > uio->uio_resid) 504 size = uio->uio_resid; 505 506 va = (caddr_t) rpipe->pipe_map.kva + 507 rpipe->pipe_map.pos; 508 error = uiomove(va, size, uio); 509 if (error) 510 break; 511 nread += size; 512 rpipe->pipe_map.pos += size; 513 rpipe->pipe_map.cnt -= size; 514 if (rpipe->pipe_map.cnt == 0) { 515 PIPE_LOCK(rpipe); 516 rpipe->pipe_state &= ~PIPE_DIRECTR; 517 wakeup(rpipe); 518 PIPE_UNLOCK(rpipe); 519 } 520 #endif 521 } else { 522 /* 523 * Break if some data was read. 524 */ 525 if (nread > 0) 526 break; 527 528 PIPE_LOCK(rpipe); 529 530 /* 531 * detect EOF condition 532 * read returns 0 on EOF, no need to set error 533 */ 534 if (rpipe->pipe_state & PIPE_EOF) { 535 PIPE_UNLOCK(rpipe); 536 break; 537 } 538 539 /* 540 * don't block on non-blocking I/O 541 */ 542 if (fp->f_flag & FNONBLOCK) { 543 PIPE_UNLOCK(rpipe); 544 error = EAGAIN; 545 break; 546 } 547 548 /* 549 * Unlock the pipe buffer for our remaining processing. 550 * We will either break out with an error or we will 551 * sleep and relock to loop. 552 */ 553 pipeunlock(rpipe); 554 555 /* 556 * The PIPE_DIRECTR flag is not under the control 557 * of the long-term lock (see pipe_direct_write()), 558 * so re-check now while holding the spin lock. 559 */ 560 if ((rpipe->pipe_state & PIPE_DIRECTR) != 0) 561 goto again; 562 563 /* 564 * We want to read more, wake up select/poll. 565 */ 566 pipeselwakeup(rpipe, rpipe->pipe_peer, fp->f_data, 567 POLL_IN); 568 569 /* 570 * If the "write-side" is blocked, wake it up now. 571 */ 572 if (rpipe->pipe_state & PIPE_WANTW) { 573 rpipe->pipe_state &= ~PIPE_WANTW; 574 wakeup(rpipe); 575 } 576 577 /* Now wait until the pipe is filled */ 578 rpipe->pipe_state |= PIPE_WANTR; 579 error = ltsleep(rpipe, PSOCK | PCATCH, 580 "piperd", 0, &rpipe->pipe_slock); 581 if (error != 0) 582 goto unlocked_error; 583 goto again; 584 } 585 } 586 587 if (error == 0) 588 PIPE_TIMESTAMP(&rpipe->pipe_atime); 589 590 PIPE_LOCK(rpipe); 591 pipeunlock(rpipe); 592 593 unlocked_error: 594 --rpipe->pipe_busy; 595 596 /* 597 * PIPE_WANTCLOSE processing only makes sense if pipe_busy is 0. 598 */ 599 if ((rpipe->pipe_busy == 0) && (rpipe->pipe_state & PIPE_WANTCLOSE)) { 600 rpipe->pipe_state &= ~(PIPE_WANTCLOSE|PIPE_WANTW); 601 wakeup(rpipe); 602 } else if (bp->cnt < MINPIPESIZE) { 603 /* 604 * Handle write blocking hysteresis. 605 */ 606 if (rpipe->pipe_state & PIPE_WANTW) { 607 rpipe->pipe_state &= ~PIPE_WANTW; 608 wakeup(rpipe); 609 } 610 } 611 612 /* 613 * If anything was read off the buffer, signal to the writer it's 614 * possible to write more data. Also send signal if we are here for the 615 * first time after last write. 616 */ 617 if ((bp->size - bp->cnt) >= PIPE_BUF 618 && (ocnt != bp->cnt || (rpipe->pipe_state & PIPE_SIGNALR))) { 619 pipeselwakeup(rpipe, rpipe->pipe_peer, fp->f_data, POLL_OUT); 620 rpipe->pipe_state &= ~PIPE_SIGNALR; 621 } 622 623 PIPE_UNLOCK(rpipe); 624 return (error); 625 } 626 627 #ifndef PIPE_NODIRECT 628 /* 629 * Allocate structure for loan transfer. 630 */ 631 static int 632 pipe_loan_alloc(wpipe, npages) 633 struct pipe *wpipe; 634 int npages; 635 { 636 vsize_t len; 637 638 len = (vsize_t)npages << PAGE_SHIFT; 639 wpipe->pipe_map.kva = uvm_km_valloc_wait(kernel_map, len); 640 if (wpipe->pipe_map.kva == 0) 641 return (ENOMEM); 642 643 amountpipekva += len; 644 wpipe->pipe_map.npages = npages; 645 wpipe->pipe_map.pgs = malloc(npages * sizeof(struct vm_page *), M_PIPE, 646 M_WAITOK); 647 return (0); 648 } 649 650 /* 651 * Free resources allocated for loan transfer. 652 */ 653 static void 654 pipe_loan_free(wpipe) 655 struct pipe *wpipe; 656 { 657 vsize_t len; 658 659 len = (vsize_t)wpipe->pipe_map.npages << PAGE_SHIFT; 660 uvm_km_free(kernel_map, wpipe->pipe_map.kva, len); 661 wpipe->pipe_map.kva = 0; 662 amountpipekva -= len; 663 free(wpipe->pipe_map.pgs, M_PIPE); 664 wpipe->pipe_map.pgs = NULL; 665 } 666 667 /* 668 * NetBSD direct write, using uvm_loan() mechanism. 669 * This implements the pipe buffer write mechanism. Note that only 670 * a direct write OR a normal pipe write can be pending at any given time. 671 * If there are any characters in the pipe buffer, the direct write will 672 * be deferred until the receiving process grabs all of the bytes from 673 * the pipe buffer. Then the direct mapping write is set-up. 674 * 675 * Called with the long-term pipe lock held. 676 */ 677 static int 678 pipe_direct_write(fp, wpipe, uio) 679 struct file *fp; 680 struct pipe *wpipe; 681 struct uio *uio; 682 { 683 int error, npages, j; 684 struct vm_page **pgs; 685 vaddr_t bbase, kva, base, bend; 686 vsize_t blen, bcnt; 687 voff_t bpos; 688 689 KASSERT(wpipe->pipe_map.cnt == 0); 690 691 /* 692 * Handle first PIPE_CHUNK_SIZE bytes of buffer. Deal with buffers 693 * not aligned to PAGE_SIZE. 694 */ 695 bbase = (vaddr_t)uio->uio_iov->iov_base; 696 base = trunc_page(bbase); 697 bend = round_page(bbase + uio->uio_iov->iov_len); 698 blen = bend - base; 699 bpos = bbase - base; 700 701 if (blen > PIPE_DIRECT_CHUNK) { 702 blen = PIPE_DIRECT_CHUNK; 703 bend = base + blen; 704 bcnt = PIPE_DIRECT_CHUNK - bpos; 705 } else { 706 bcnt = uio->uio_iov->iov_len; 707 } 708 npages = blen >> PAGE_SHIFT; 709 710 /* 711 * Free the old kva if we need more pages than we have 712 * allocated. 713 */ 714 if (wpipe->pipe_map.kva != 0 && npages > wpipe->pipe_map.npages) 715 pipe_loan_free(wpipe); 716 717 /* Allocate new kva. */ 718 if (wpipe->pipe_map.kva == 0) { 719 error = pipe_loan_alloc(wpipe, npages); 720 if (error) 721 return (error); 722 } 723 724 /* Loan the write buffer memory from writer process */ 725 pgs = wpipe->pipe_map.pgs; 726 error = uvm_loan(&uio->uio_procp->p_vmspace->vm_map, base, blen, 727 pgs, UVM_LOAN_TOPAGE); 728 if (error) { 729 pipe_loan_free(wpipe); 730 return (error); 731 } 732 733 /* Enter the loaned pages to kva */ 734 kva = wpipe->pipe_map.kva; 735 for (j = 0; j < npages; j++, kva += PAGE_SIZE) { 736 pmap_kenter_pa(kva, VM_PAGE_TO_PHYS(pgs[j]), VM_PROT_READ); 737 } 738 pmap_update(pmap_kernel()); 739 740 /* Now we can put the pipe in direct write mode */ 741 wpipe->pipe_map.pos = bpos; 742 wpipe->pipe_map.cnt = bcnt; 743 wpipe->pipe_state |= PIPE_DIRECTW; 744 745 /* 746 * But before we can let someone do a direct read, 747 * we have to wait until the pipe is drained. 748 */ 749 750 /* Relase the pipe lock while we wait */ 751 PIPE_LOCK(wpipe); 752 pipeunlock(wpipe); 753 754 while (error == 0 && wpipe->pipe_buffer.cnt > 0) { 755 if (wpipe->pipe_state & PIPE_WANTR) { 756 wpipe->pipe_state &= ~PIPE_WANTR; 757 wakeup(wpipe); 758 } 759 760 wpipe->pipe_state |= PIPE_WANTW; 761 error = ltsleep(wpipe, PSOCK | PCATCH, "pipdwc", 0, 762 &wpipe->pipe_slock); 763 if (error == 0 && wpipe->pipe_state & PIPE_EOF) 764 error = EPIPE; 765 } 766 767 /* Pipe is drained; next read will off the direct buffer */ 768 wpipe->pipe_state |= PIPE_DIRECTR; 769 770 /* Wait until the reader is done */ 771 while (error == 0 && (wpipe->pipe_state & PIPE_DIRECTR)) { 772 if (wpipe->pipe_state & PIPE_WANTR) { 773 wpipe->pipe_state &= ~PIPE_WANTR; 774 wakeup(wpipe); 775 } 776 pipeselwakeup(wpipe, wpipe, fp->f_data, POLL_IN); 777 error = ltsleep(wpipe, PSOCK | PCATCH, "pipdwt", 0, 778 &wpipe->pipe_slock); 779 if (error == 0 && wpipe->pipe_state & PIPE_EOF) 780 error = EPIPE; 781 } 782 783 /* Take pipe out of direct write mode */ 784 wpipe->pipe_state &= ~(PIPE_DIRECTW | PIPE_DIRECTR); 785 786 /* Acquire the pipe lock and cleanup */ 787 (void)pipelock(wpipe, 0); 788 if (pgs != NULL) { 789 pmap_kremove(wpipe->pipe_map.kva, blen); 790 uvm_unloan(pgs, npages, UVM_LOAN_TOPAGE); 791 } 792 if (error || amountpipekva > maxpipekva) 793 pipe_loan_free(wpipe); 794 795 if (error) { 796 pipeselwakeup(wpipe, wpipe, fp->f_data, POLL_ERR); 797 798 /* 799 * If nothing was read from what we offered, return error 800 * straight on. Otherwise update uio resid first. Caller 801 * will deal with the error condition, returning short 802 * write, error, or restarting the write(2) as appropriate. 803 */ 804 if (wpipe->pipe_map.cnt == bcnt) { 805 wpipe->pipe_map.cnt = 0; 806 wakeup(wpipe); 807 return (error); 808 } 809 810 bcnt -= wpipe->pipe_map.cnt; 811 } 812 813 uio->uio_resid -= bcnt; 814 /* uio_offset not updated, not set/used for write(2) */ 815 uio->uio_iov->iov_base = (char *)uio->uio_iov->iov_base + bcnt; 816 uio->uio_iov->iov_len -= bcnt; 817 if (uio->uio_iov->iov_len == 0) { 818 uio->uio_iov++; 819 uio->uio_iovcnt--; 820 } 821 822 wpipe->pipe_map.cnt = 0; 823 return (error); 824 } 825 #endif /* !PIPE_NODIRECT */ 826 827 static int 828 pipe_write(fp, offset, uio, cred, flags) 829 struct file *fp; 830 off_t *offset; 831 struct uio *uio; 832 struct ucred *cred; 833 int flags; 834 { 835 struct pipe *wpipe, *rpipe; 836 struct pipebuf *bp; 837 int error; 838 839 /* We want to write to our peer */ 840 rpipe = (struct pipe *) fp->f_data; 841 842 retry: 843 error = 0; 844 PIPE_LOCK(rpipe); 845 wpipe = rpipe->pipe_peer; 846 847 /* 848 * Detect loss of pipe read side, issue SIGPIPE if lost. 849 */ 850 if (wpipe == NULL) 851 error = EPIPE; 852 else if (simple_lock_try(&wpipe->pipe_slock) == 0) { 853 /* Deal with race for peer */ 854 PIPE_UNLOCK(rpipe); 855 goto retry; 856 } else if ((wpipe->pipe_state & PIPE_EOF) != 0) { 857 PIPE_UNLOCK(wpipe); 858 error = EPIPE; 859 } 860 861 PIPE_UNLOCK(rpipe); 862 if (error != 0) 863 return (error); 864 865 ++wpipe->pipe_busy; 866 867 /* Aquire the long-term pipe lock */ 868 if ((error = pipelock(wpipe,1)) != 0) { 869 --wpipe->pipe_busy; 870 if (wpipe->pipe_busy == 0 871 && (wpipe->pipe_state & PIPE_WANTCLOSE)) { 872 wpipe->pipe_state &= ~(PIPE_WANTCLOSE | PIPE_WANTR); 873 wakeup(wpipe); 874 } 875 PIPE_UNLOCK(wpipe); 876 return (error); 877 } 878 879 bp = &wpipe->pipe_buffer; 880 881 /* 882 * If it is advantageous to resize the pipe buffer, do so. 883 */ 884 if ((uio->uio_resid > PIPE_SIZE) && 885 (nbigpipe < maxbigpipes) && 886 #ifndef PIPE_NODIRECT 887 (wpipe->pipe_state & PIPE_DIRECTW) == 0 && 888 #endif 889 (bp->size <= PIPE_SIZE) && (bp->cnt == 0)) { 890 891 if (pipespace(wpipe, BIG_PIPE_SIZE) == 0) 892 nbigpipe++; 893 } 894 895 while (uio->uio_resid) { 896 size_t space; 897 898 #ifndef PIPE_NODIRECT 899 /* 900 * Pipe buffered writes cannot be coincidental with 901 * direct writes. Also, only one direct write can be 902 * in progress at any one time. We wait until the currently 903 * executing direct write is completed before continuing. 904 * 905 * We break out if a signal occurs or the reader goes away. 906 */ 907 while (error == 0 && wpipe->pipe_state & PIPE_DIRECTW) { 908 PIPE_LOCK(wpipe); 909 if (wpipe->pipe_state & PIPE_WANTR) { 910 wpipe->pipe_state &= ~PIPE_WANTR; 911 wakeup(wpipe); 912 } 913 pipeunlock(wpipe); 914 error = ltsleep(wpipe, PSOCK | PCATCH, 915 "pipbww", 0, &wpipe->pipe_slock); 916 917 (void)pipelock(wpipe, 0); 918 if (wpipe->pipe_state & PIPE_EOF) 919 error = EPIPE; 920 } 921 if (error) 922 break; 923 924 /* 925 * If the transfer is large, we can gain performance if 926 * we do process-to-process copies directly. 927 * If the write is non-blocking, we don't use the 928 * direct write mechanism. 929 * 930 * The direct write mechanism will detect the reader going 931 * away on us. 932 */ 933 if ((uio->uio_iov->iov_len >= PIPE_MINDIRECT) && 934 (fp->f_flag & FNONBLOCK) == 0 && 935 (wpipe->pipe_map.kva || (amountpipekva < limitpipekva))) { 936 error = pipe_direct_write(fp, wpipe, uio); 937 938 /* 939 * Break out if error occurred, unless it's ENOMEM. 940 * ENOMEM means we failed to allocate some resources 941 * for direct write, so we just fallback to ordinary 942 * write. If the direct write was successful, 943 * process rest of data via ordinary write. 944 */ 945 if (error == 0) 946 continue; 947 948 if (error != ENOMEM) 949 break; 950 } 951 #endif /* PIPE_NODIRECT */ 952 953 space = bp->size - bp->cnt; 954 955 /* Writes of size <= PIPE_BUF must be atomic. */ 956 if ((space < uio->uio_resid) && (uio->uio_resid <= PIPE_BUF)) 957 space = 0; 958 959 if (space > 0) { 960 int size; /* Transfer size */ 961 int segsize; /* first segment to transfer */ 962 963 /* 964 * Transfer size is minimum of uio transfer 965 * and free space in pipe buffer. 966 */ 967 if (space > uio->uio_resid) 968 size = uio->uio_resid; 969 else 970 size = space; 971 /* 972 * First segment to transfer is minimum of 973 * transfer size and contiguous space in 974 * pipe buffer. If first segment to transfer 975 * is less than the transfer size, we've got 976 * a wraparound in the buffer. 977 */ 978 segsize = bp->size - bp->in; 979 if (segsize > size) 980 segsize = size; 981 982 /* Transfer first segment */ 983 error = uiomove(&bp->buffer[bp->in], segsize, uio); 984 985 if (error == 0 && segsize < size) { 986 /* 987 * Transfer remaining part now, to 988 * support atomic writes. Wraparound 989 * happened. 990 */ 991 #ifdef DEBUG 992 if (bp->in + segsize != bp->size) 993 panic("Expected pipe buffer wraparound disappeared"); 994 #endif 995 996 error = uiomove(&bp->buffer[0], 997 size - segsize, uio); 998 } 999 if (error) 1000 break; 1001 1002 bp->in += size; 1003 if (bp->in >= bp->size) { 1004 #ifdef DEBUG 1005 if (bp->in != size - segsize + bp->size) 1006 panic("Expected wraparound bad"); 1007 #endif 1008 bp->in = size - segsize; 1009 } 1010 1011 bp->cnt += size; 1012 #ifdef DEBUG 1013 if (bp->cnt > bp->size) 1014 panic("Pipe buffer overflow"); 1015 #endif 1016 } else { 1017 /* 1018 * If the "read-side" has been blocked, wake it up now. 1019 */ 1020 PIPE_LOCK(wpipe); 1021 if (wpipe->pipe_state & PIPE_WANTR) { 1022 wpipe->pipe_state &= ~PIPE_WANTR; 1023 wakeup(wpipe); 1024 } 1025 PIPE_UNLOCK(wpipe); 1026 1027 /* 1028 * don't block on non-blocking I/O 1029 */ 1030 if (fp->f_flag & FNONBLOCK) { 1031 error = EAGAIN; 1032 break; 1033 } 1034 1035 /* 1036 * We have no more space and have something to offer, 1037 * wake up select/poll. 1038 */ 1039 if (bp->cnt) 1040 pipeselwakeup(wpipe, wpipe, fp->f_data, 1041 POLL_OUT); 1042 1043 PIPE_LOCK(wpipe); 1044 pipeunlock(wpipe); 1045 wpipe->pipe_state |= PIPE_WANTW; 1046 error = ltsleep(wpipe, PSOCK | PCATCH, "pipewr", 0, 1047 &wpipe->pipe_slock); 1048 (void)pipelock(wpipe, 0); 1049 if (error != 0) 1050 break; 1051 /* 1052 * If read side wants to go away, we just issue a signal 1053 * to ourselves. 1054 */ 1055 if (wpipe->pipe_state & PIPE_EOF) { 1056 error = EPIPE; 1057 break; 1058 } 1059 } 1060 } 1061 1062 PIPE_LOCK(wpipe); 1063 --wpipe->pipe_busy; 1064 if ((wpipe->pipe_busy == 0) && (wpipe->pipe_state & PIPE_WANTCLOSE)) { 1065 wpipe->pipe_state &= ~(PIPE_WANTCLOSE | PIPE_WANTR); 1066 wakeup(wpipe); 1067 } else if (bp->cnt > 0) { 1068 /* 1069 * If we have put any characters in the buffer, we wake up 1070 * the reader. 1071 */ 1072 if (wpipe->pipe_state & PIPE_WANTR) { 1073 wpipe->pipe_state &= ~PIPE_WANTR; 1074 wakeup(wpipe); 1075 } 1076 } 1077 1078 /* 1079 * Don't return EPIPE if I/O was successful 1080 */ 1081 if (error == EPIPE && bp->cnt == 0 && uio->uio_resid == 0) 1082 error = 0; 1083 1084 if (error == 0) 1085 PIPE_TIMESTAMP(&wpipe->pipe_mtime); 1086 1087 /* 1088 * We have something to offer, wake up select/poll. 1089 * wpipe->pipe_map.cnt is always 0 in this point (direct write 1090 * is only done synchronously), so check only wpipe->pipe_buffer.cnt 1091 */ 1092 if (bp->cnt) 1093 pipeselwakeup(wpipe, wpipe, fp->f_data, POLL_OUT); 1094 1095 /* 1096 * Arrange for next read(2) to do a signal. 1097 */ 1098 wpipe->pipe_state |= PIPE_SIGNALR; 1099 1100 pipeunlock(wpipe); 1101 PIPE_UNLOCK(wpipe); 1102 return (error); 1103 } 1104 1105 /* 1106 * we implement a very minimal set of ioctls for compatibility with sockets. 1107 */ 1108 int 1109 pipe_ioctl(fp, cmd, data, p) 1110 struct file *fp; 1111 u_long cmd; 1112 void *data; 1113 struct proc *p; 1114 { 1115 struct pipe *pipe = (struct pipe *)fp->f_data; 1116 1117 switch (cmd) { 1118 1119 case FIONBIO: 1120 return (0); 1121 1122 case FIOASYNC: 1123 PIPE_LOCK(pipe); 1124 if (*(int *)data) { 1125 pipe->pipe_state |= PIPE_ASYNC; 1126 } else { 1127 pipe->pipe_state &= ~PIPE_ASYNC; 1128 } 1129 PIPE_UNLOCK(pipe); 1130 return (0); 1131 1132 case FIONREAD: 1133 PIPE_LOCK(pipe); 1134 #ifndef PIPE_NODIRECT 1135 if (pipe->pipe_state & PIPE_DIRECTW) 1136 *(int *)data = pipe->pipe_map.cnt; 1137 else 1138 #endif 1139 *(int *)data = pipe->pipe_buffer.cnt; 1140 PIPE_UNLOCK(pipe); 1141 return (0); 1142 1143 case TIOCSPGRP: 1144 case FIOSETOWN: 1145 return fsetown(p, &pipe->pipe_pgid, cmd, data); 1146 1147 case TIOCGPGRP: 1148 case FIOGETOWN: 1149 return fgetown(p, pipe->pipe_pgid, cmd, data); 1150 1151 } 1152 return (EPASSTHROUGH); 1153 } 1154 1155 int 1156 pipe_poll(fp, events, td) 1157 struct file *fp; 1158 int events; 1159 struct proc *td; 1160 { 1161 struct pipe *rpipe = (struct pipe *)fp->f_data; 1162 struct pipe *wpipe; 1163 int eof = 0; 1164 int revents = 0; 1165 1166 retry: 1167 PIPE_LOCK(rpipe); 1168 wpipe = rpipe->pipe_peer; 1169 if (wpipe != NULL && simple_lock_try(&wpipe->pipe_slock) == 0) { 1170 /* Deal with race for peer */ 1171 PIPE_UNLOCK(rpipe); 1172 goto retry; 1173 } 1174 1175 if (events & (POLLIN | POLLRDNORM)) 1176 if ((rpipe->pipe_buffer.cnt > 0) || 1177 #ifndef PIPE_NODIRECT 1178 (rpipe->pipe_state & PIPE_DIRECTR) || 1179 #endif 1180 (rpipe->pipe_state & PIPE_EOF)) 1181 revents |= events & (POLLIN | POLLRDNORM); 1182 1183 eof |= (rpipe->pipe_state & PIPE_EOF); 1184 PIPE_UNLOCK(rpipe); 1185 1186 if (wpipe == NULL) 1187 revents |= events & (POLLOUT | POLLWRNORM); 1188 else { 1189 if (events & (POLLOUT | POLLWRNORM)) 1190 if ((wpipe->pipe_state & PIPE_EOF) || ( 1191 #ifndef PIPE_NODIRECT 1192 (wpipe->pipe_state & PIPE_DIRECTW) == 0 && 1193 #endif 1194 (wpipe->pipe_buffer.size - wpipe->pipe_buffer.cnt) >= PIPE_BUF)) 1195 revents |= events & (POLLOUT | POLLWRNORM); 1196 1197 eof |= (wpipe->pipe_state & PIPE_EOF); 1198 PIPE_UNLOCK(wpipe); 1199 } 1200 1201 if (wpipe == NULL || eof) 1202 revents |= POLLHUP; 1203 1204 if (revents == 0) { 1205 if (events & (POLLIN | POLLRDNORM)) 1206 selrecord(td, &rpipe->pipe_sel); 1207 1208 if (events & (POLLOUT | POLLWRNORM)) 1209 selrecord(td, &wpipe->pipe_sel); 1210 } 1211 1212 return (revents); 1213 } 1214 1215 static int 1216 pipe_stat(fp, ub, td) 1217 struct file *fp; 1218 struct stat *ub; 1219 struct proc *td; 1220 { 1221 struct pipe *pipe = (struct pipe *)fp->f_data; 1222 1223 memset((caddr_t)ub, 0, sizeof(*ub)); 1224 ub->st_mode = S_IFIFO | S_IRUSR | S_IWUSR; 1225 ub->st_blksize = pipe->pipe_buffer.size; 1226 ub->st_size = pipe->pipe_buffer.cnt; 1227 ub->st_blocks = (ub->st_size) ? 1 : 0; 1228 TIMEVAL_TO_TIMESPEC(&pipe->pipe_atime, &ub->st_atimespec) 1229 TIMEVAL_TO_TIMESPEC(&pipe->pipe_mtime, &ub->st_mtimespec); 1230 TIMEVAL_TO_TIMESPEC(&pipe->pipe_ctime, &ub->st_ctimespec); 1231 ub->st_uid = fp->f_cred->cr_uid; 1232 ub->st_gid = fp->f_cred->cr_gid; 1233 /* 1234 * Left as 0: st_dev, st_ino, st_nlink, st_rdev, st_flags, st_gen. 1235 * XXX (st_dev, st_ino) should be unique. 1236 */ 1237 return (0); 1238 } 1239 1240 /* ARGSUSED */ 1241 static int 1242 pipe_close(fp, td) 1243 struct file *fp; 1244 struct proc *td; 1245 { 1246 struct pipe *pipe = (struct pipe *)fp->f_data; 1247 1248 fp->f_data = NULL; 1249 pipeclose(fp, pipe); 1250 return (0); 1251 } 1252 1253 static void 1254 pipe_free_kmem(pipe) 1255 struct pipe *pipe; 1256 { 1257 1258 if (pipe->pipe_buffer.buffer != NULL) { 1259 if (pipe->pipe_buffer.size > PIPE_SIZE) 1260 --nbigpipe; 1261 amountpipekva -= pipe->pipe_buffer.size; 1262 uvm_km_free(kernel_map, 1263 (vaddr_t)pipe->pipe_buffer.buffer, 1264 pipe->pipe_buffer.size); 1265 pipe->pipe_buffer.buffer = NULL; 1266 } 1267 #ifndef PIPE_NODIRECT 1268 if (pipe->pipe_map.kva != 0) { 1269 pipe_loan_free(pipe); 1270 pipe->pipe_map.cnt = 0; 1271 pipe->pipe_map.kva = 0; 1272 pipe->pipe_map.pos = 0; 1273 pipe->pipe_map.npages = 0; 1274 } 1275 #endif /* !PIPE_NODIRECT */ 1276 } 1277 1278 /* 1279 * shutdown the pipe 1280 */ 1281 static void 1282 pipeclose(fp, pipe) 1283 struct file *fp; 1284 struct pipe *pipe; 1285 { 1286 struct pipe *ppipe; 1287 1288 if (pipe == NULL) 1289 return; 1290 1291 retry: 1292 PIPE_LOCK(pipe); 1293 1294 if (fp) 1295 pipeselwakeup(pipe, pipe, fp->f_data, POLL_HUP); 1296 1297 /* 1298 * If the other side is blocked, wake it up saying that 1299 * we want to close it down. 1300 */ 1301 while (pipe->pipe_busy) { 1302 wakeup(pipe); 1303 pipe->pipe_state |= PIPE_WANTCLOSE | PIPE_EOF; 1304 ltsleep(pipe, PSOCK, "pipecl", 0, &pipe->pipe_slock); 1305 } 1306 1307 /* 1308 * Disconnect from peer 1309 */ 1310 if ((ppipe = pipe->pipe_peer) != NULL) { 1311 /* Deal with race for peer */ 1312 if (simple_lock_try(&ppipe->pipe_slock) == 0) { 1313 PIPE_UNLOCK(pipe); 1314 goto retry; 1315 } 1316 if (fp) 1317 pipeselwakeup(ppipe, ppipe, fp->f_data, POLL_HUP); 1318 1319 ppipe->pipe_state |= PIPE_EOF; 1320 wakeup(ppipe); 1321 ppipe->pipe_peer = NULL; 1322 PIPE_UNLOCK(ppipe); 1323 } 1324 1325 (void)lockmgr(&pipe->pipe_lock, LK_DRAIN | LK_INTERLOCK, 1326 &pipe->pipe_slock); 1327 1328 /* 1329 * free resources 1330 */ 1331 pipe_free_kmem(pipe); 1332 pool_put(&pipe_pool, pipe); 1333 } 1334 1335 static void 1336 filt_pipedetach(struct knote *kn) 1337 { 1338 struct pipe *pipe = (struct pipe *)kn->kn_fp->f_data; 1339 1340 switch(kn->kn_filter) { 1341 case EVFILT_WRITE: 1342 /* need the peer structure, not our own */ 1343 pipe = pipe->pipe_peer; 1344 /* XXXSMP: race for peer */ 1345 1346 /* if reader end already closed, just return */ 1347 if (pipe == NULL) 1348 return; 1349 1350 break; 1351 default: 1352 /* nothing to do */ 1353 break; 1354 } 1355 1356 #ifdef DIAGNOSTIC 1357 if (kn->kn_hook != pipe) 1358 panic("filt_pipedetach: inconsistent knote"); 1359 #endif 1360 1361 PIPE_LOCK(pipe); 1362 SLIST_REMOVE(&pipe->pipe_sel.sel_klist, kn, knote, kn_selnext); 1363 PIPE_UNLOCK(pipe); 1364 } 1365 1366 /*ARGSUSED*/ 1367 static int 1368 filt_piperead(struct knote *kn, long hint) 1369 { 1370 struct pipe *rpipe = (struct pipe *)kn->kn_fp->f_data; 1371 struct pipe *wpipe = rpipe->pipe_peer; 1372 1373 if ((hint & NOTE_SUBMIT) == 0) 1374 PIPE_LOCK(rpipe); 1375 kn->kn_data = rpipe->pipe_buffer.cnt; 1376 if ((kn->kn_data == 0) && (rpipe->pipe_state & PIPE_DIRECTW)) 1377 kn->kn_data = rpipe->pipe_map.cnt; 1378 1379 /* XXXSMP: race for peer */ 1380 if ((rpipe->pipe_state & PIPE_EOF) || 1381 (wpipe == NULL) || (wpipe->pipe_state & PIPE_EOF)) { 1382 kn->kn_flags |= EV_EOF; 1383 if ((hint & NOTE_SUBMIT) == 0) 1384 PIPE_UNLOCK(rpipe); 1385 return (1); 1386 } 1387 if ((hint & NOTE_SUBMIT) == 0) 1388 PIPE_UNLOCK(rpipe); 1389 return (kn->kn_data > 0); 1390 } 1391 1392 /*ARGSUSED*/ 1393 static int 1394 filt_pipewrite(struct knote *kn, long hint) 1395 { 1396 struct pipe *rpipe = (struct pipe *)kn->kn_fp->f_data; 1397 struct pipe *wpipe = rpipe->pipe_peer; 1398 1399 if ((hint & NOTE_SUBMIT) == 0) 1400 PIPE_LOCK(rpipe); 1401 /* XXXSMP: race for peer */ 1402 if ((wpipe == NULL) || (wpipe->pipe_state & PIPE_EOF)) { 1403 kn->kn_data = 0; 1404 kn->kn_flags |= EV_EOF; 1405 if ((hint & NOTE_SUBMIT) == 0) 1406 PIPE_UNLOCK(rpipe); 1407 return (1); 1408 } 1409 kn->kn_data = wpipe->pipe_buffer.size - wpipe->pipe_buffer.cnt; 1410 if (wpipe->pipe_state & PIPE_DIRECTW) 1411 kn->kn_data = 0; 1412 1413 if ((hint & NOTE_SUBMIT) == 0) 1414 PIPE_UNLOCK(rpipe); 1415 return (kn->kn_data >= PIPE_BUF); 1416 } 1417 1418 static const struct filterops pipe_rfiltops = 1419 { 1, NULL, filt_pipedetach, filt_piperead }; 1420 static const struct filterops pipe_wfiltops = 1421 { 1, NULL, filt_pipedetach, filt_pipewrite }; 1422 1423 /*ARGSUSED*/ 1424 static int 1425 pipe_kqfilter(struct file *fp, struct knote *kn) 1426 { 1427 struct pipe *pipe; 1428 1429 pipe = (struct pipe *)kn->kn_fp->f_data; 1430 switch (kn->kn_filter) { 1431 case EVFILT_READ: 1432 kn->kn_fop = &pipe_rfiltops; 1433 break; 1434 case EVFILT_WRITE: 1435 kn->kn_fop = &pipe_wfiltops; 1436 /* XXXSMP: race for peer */ 1437 pipe = pipe->pipe_peer; 1438 if (pipe == NULL) { 1439 /* other end of pipe has been closed */ 1440 return (EBADF); 1441 } 1442 break; 1443 default: 1444 return (1); 1445 } 1446 kn->kn_hook = pipe; 1447 1448 PIPE_LOCK(pipe); 1449 SLIST_INSERT_HEAD(&pipe->pipe_sel.sel_klist, kn, kn_selnext); 1450 PIPE_UNLOCK(pipe); 1451 return (0); 1452 } 1453 1454 static int 1455 pipe_fcntl(fp, cmd, data, p) 1456 struct file *fp; 1457 u_int cmd; 1458 void *data; 1459 struct proc *p; 1460 { 1461 if (cmd == F_SETFL) 1462 return (0); 1463 else 1464 return (EOPNOTSUPP); 1465 } 1466 1467 /* 1468 * Handle pipe sysctls. 1469 */ 1470 SYSCTL_SETUP(sysctl_kern_pipe_setup, "sysctl kern.pipe subtree setup") 1471 { 1472 1473 sysctl_createv(clog, 0, NULL, NULL, 1474 CTLFLAG_PERMANENT, 1475 CTLTYPE_NODE, "kern", NULL, 1476 NULL, 0, NULL, 0, 1477 CTL_KERN, CTL_EOL); 1478 sysctl_createv(clog, 0, NULL, NULL, 1479 CTLFLAG_PERMANENT, 1480 CTLTYPE_NODE, "pipe", 1481 SYSCTL_DESCR("Pipe settings"), 1482 NULL, 0, NULL, 0, 1483 CTL_KERN, KERN_PIPE, CTL_EOL); 1484 1485 sysctl_createv(clog, 0, NULL, NULL, 1486 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1487 CTLTYPE_INT, "maxkvasz", 1488 SYSCTL_DESCR("Maximum amount of kernel memory to be " 1489 "used for pipes"), 1490 NULL, 0, &maxpipekva, 0, 1491 CTL_KERN, KERN_PIPE, KERN_PIPE_MAXKVASZ, CTL_EOL); 1492 sysctl_createv(clog, 0, NULL, NULL, 1493 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1494 CTLTYPE_INT, "maxloankvasz", 1495 SYSCTL_DESCR("Limit for direct transfers via page loan"), 1496 NULL, 0, &limitpipekva, 0, 1497 CTL_KERN, KERN_PIPE, KERN_PIPE_LIMITKVA, CTL_EOL); 1498 sysctl_createv(clog, 0, NULL, NULL, 1499 CTLFLAG_PERMANENT|CTLFLAG_READWRITE, 1500 CTLTYPE_INT, "maxbigpipes", 1501 SYSCTL_DESCR("Maximum number of \"big\" pipes"), 1502 NULL, 0, &maxbigpipes, 0, 1503 CTL_KERN, KERN_PIPE, KERN_PIPE_MAXBIGPIPES, CTL_EOL); 1504 sysctl_createv(clog, 0, NULL, NULL, 1505 CTLFLAG_PERMANENT, 1506 CTLTYPE_INT, "nbigpipes", 1507 SYSCTL_DESCR("Number of \"big\" pipes"), 1508 NULL, 0, &nbigpipe, 0, 1509 CTL_KERN, KERN_PIPE, KERN_PIPE_NBIGPIPES, CTL_EOL); 1510 sysctl_createv(clog, 0, NULL, NULL, 1511 CTLFLAG_PERMANENT, 1512 CTLTYPE_INT, "kvasize", 1513 SYSCTL_DESCR("Amount of kernel memory consumed by pipe " 1514 "buffers"), 1515 NULL, 0, &amountpipekva, 0, 1516 CTL_KERN, KERN_PIPE, KERN_PIPE_KVASIZE, CTL_EOL); 1517 } 1518 1519 /* 1520 * Initialize pipe structs. 1521 */ 1522 void 1523 pipe_init(void) 1524 { 1525 1526 pool_init(&pipe_pool, sizeof(struct pipe), 0, 0, 0, "pipepl", 1527 &pool_allocator_nointr); 1528 }
Cache object: 8eb12d49ae7b672ace4d6102d7058ae1
[ source navigation ] [ diff markup ] [ identifier search ] [ freetext search ] [ file search ] [ list types ] [ track identifier ]