source3/modules/vfs_aio_fork.c

   1 /*
   2  * Simulate the Posix AIO using mmap/fork
   3  *
   4  * Copyright (C) Volker Lendecke 2008
   5  * Copyright (C) Jeremy Allison 2010
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License
  18  * along with this program; if not, write to the Free Software
  19  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  20  */
  21
  22 #include "includes.h"
  23 #include "system/filesys.h"
  24 #include "system/shmem.h"
  25 #include "smbd/smbd.h"
  26 #include "smbd/globals.h"
  27 #include "lib/async_req/async_sock.h"
  28 #include "lib/util/tevent_unix.h"
  29
  30 #undef recvmsg
  31
  32 #ifndef MAP_FILE
  33 #define MAP_FILE 0
  34 #endif
  35
  36 struct aio_fork_config {
  37         bool erratic_testing_mode;
  38 };
  39
  40 struct mmap_area {
  41         size_t size;
  42         volatile void *ptr;
  43 };
  44
  45 static int mmap_area_destructor(struct mmap_area *area)
  46 {
  47         munmap((void *)area->ptr, area->size);
  48         return 0;
  49 }
  50
  51 static struct mmap_area *mmap_area_init(TALLOC_CTX *mem_ctx, size_t size)
  52 {
  53         struct mmap_area *result;
  54         int fd;
  55
  56         result = talloc(mem_ctx, struct mmap_area);
  57         if (result == NULL) {
  58                 DEBUG(0, ("talloc failed\n"));
  59                 goto fail;
  60         }
  61
  62         fd = open("/dev/zero", O_RDWR);
  63         if (fd == -1) {
  64                 DEBUG(3, ("open(\"/dev/zero\") failed: %s\n",
  65                           strerror(errno)));
  66                 goto fail;
  67         }
  68
  69         result->ptr = mmap(NULL, size, PROT_READ|PROT_WRITE,
  70                            MAP_SHARED|MAP_FILE, fd, 0);
  71         close(fd);
  72         if (result->ptr == MAP_FAILED) {
  73                 DEBUG(1, ("mmap failed: %s\n", strerror(errno)));
  74                 goto fail;
  75         }
  76
  77         result->size = size;
  78         talloc_set_destructor(result, mmap_area_destructor);
  79
  80         return result;
  81
  82 fail:
  83         TALLOC_FREE(result);
  84         return NULL;
  85 }
  86
  87 enum cmd_type {
  88         READ_CMD,
  89         WRITE_CMD,
  90         FSYNC_CMD
  91 };
  92
  93 static const char *cmd_type_str(enum cmd_type cmd)
  94 {
  95         const char *result;
  96
  97         switch (cmd) {
  98         case READ_CMD:
  99                 result = "READ";
 100                 break;
 101         case WRITE_CMD:
 102                 result = "WRITE";
 103                 break;
 104         case FSYNC_CMD:
 105                 result = "FSYNC";
 106                 break;
 107         default:
 108                 result = "<UNKNOWN>";
 109                 break;
 110         }
 111         return result;
 112 }
 113
 114 struct rw_cmd {
 115         size_t n;
 116         off_t offset;
 117         enum cmd_type cmd;
 118         bool erratic_testing_mode;
 119 };
 120
 121 struct rw_ret {
 122         ssize_t size;
 123         int ret_errno;
 124 };
 125
 126 struct aio_child_list;
 127
 128 struct aio_child {
 129         struct aio_child *prev, *next;
 130         struct aio_child_list *list;
 131         pid_t pid;
 132         int sockfd;
 133         struct mmap_area *map;
 134         bool dont_delete;       /* Marked as in use since last cleanup */
 135         bool busy;
 136 };
 137
 138 struct aio_child_list {
 139         struct aio_child *children;
 140         struct tevent_timer *cleanup_event;
 141 };
 142
 143 static void free_aio_children(void **p)
 144 {
 145         TALLOC_FREE(*p);
 146 }
 147
 148 static ssize_t read_fd(int fd, void *ptr, size_t nbytes, int *recvfd)
 149 {
 150         struct msghdr msg;
 151         struct iovec iov[1];
 152         ssize_t n;
 153 #ifndef HAVE_MSGHDR_MSG_CONTROL
 154         int newfd;
 155 #endif
 156
 157 #ifdef  HAVE_MSGHDR_MSG_CONTROL
 158         union {
 159           struct cmsghdr        cm;
 160           char                          control[CMSG_SPACE(sizeof(int))];
 161         } control_un;
 162         struct cmsghdr  *cmptr;
 163
 164         msg.msg_control = control_un.control;
 165         msg.msg_controllen = sizeof(control_un.control);
 166 #else
 167 #if HAVE_MSGHDR_MSG_ACCTRIGHTS
 168         msg.msg_accrights = (caddr_t) &newfd;
 169         msg.msg_accrightslen = sizeof(int);
 170 #else
 171 #error Can not pass file descriptors
 172 #endif
 173 #endif
 174
 175         msg.msg_name = NULL;
 176         msg.msg_namelen = 0;
 177         msg.msg_flags = 0;
 178
 179         iov[0].iov_base = (void *)ptr;
 180         iov[0].iov_len = nbytes;
 181         msg.msg_iov = iov;
 182         msg.msg_iovlen = 1;
 183
 184         if ( (n = recvmsg(fd, &msg, 0)) <= 0) {
 185                 return(n);
 186         }
 187
 188 #ifdef  HAVE_MSGHDR_MSG_CONTROL
 189         if ((cmptr = CMSG_FIRSTHDR(&msg)) != NULL
 190             && cmptr->cmsg_len == CMSG_LEN(sizeof(int))) {
 191                 if (cmptr->cmsg_level != SOL_SOCKET) {
 192                         DEBUG(10, ("control level != SOL_SOCKET"));
 193                         errno = EINVAL;
 194                         return -1;
 195                 }
 196                 if (cmptr->cmsg_type != SCM_RIGHTS) {
 197                         DEBUG(10, ("control type != SCM_RIGHTS"));
 198                         errno = EINVAL;
 199                         return -1;
 200                 }
 201                 memcpy(recvfd, CMSG_DATA(cmptr), sizeof(*recvfd));
 202         } else {
 203                 *recvfd = -1;           /* descriptor was not passed */
 204         }
 205 #else
 206         if (msg.msg_accrightslen == sizeof(int)) {
 207                 *recvfd = newfd;
 208         }
 209         else {
 210                 *recvfd = -1;           /* descriptor was not passed */
 211         }
 212 #endif
 213
 214         return(n);
 215 }
 216
 217 static ssize_t write_fd(int fd, void *ptr, size_t nbytes, int sendfd)
 218 {
 219         struct msghdr   msg;
 220         struct iovec    iov[1];
 221
 222 #ifdef  HAVE_MSGHDR_MSG_CONTROL
 223         union {
 224                 struct cmsghdr  cm;
 225                 char control[CMSG_SPACE(sizeof(int))];
 226         } control_un;
 227         struct cmsghdr  *cmptr;
 228
 229         ZERO_STRUCT(msg);
 230         ZERO_STRUCT(control_un);
 231
 232         msg.msg_control = control_un.control;
 233         msg.msg_controllen = sizeof(control_un.control);
 234
 235         cmptr = CMSG_FIRSTHDR(&msg);
 236         cmptr->cmsg_len = CMSG_LEN(sizeof(int));
 237         cmptr->cmsg_level = SOL_SOCKET;
 238         cmptr->cmsg_type = SCM_RIGHTS;
 239         memcpy(CMSG_DATA(cmptr), &sendfd, sizeof(sendfd));
 240 #else
 241         ZERO_STRUCT(msg);
 242         msg.msg_accrights = (caddr_t) &sendfd;
 243         msg.msg_accrightslen = sizeof(int);
 244 #endif
 245
 246         msg.msg_name = NULL;
 247         msg.msg_namelen = 0;
 248
 249         ZERO_STRUCT(iov);
 250         iov[0].iov_base = (void *)ptr;
 251         iov[0].iov_len = nbytes;
 252         msg.msg_iov = iov;
 253         msg.msg_iovlen = 1;
 254
 255         return (sendmsg(fd, &msg, 0));
 256 }
 257
 258 static void aio_child_cleanup(struct tevent_context *event_ctx,
 259                               struct tevent_timer *te,
 260                               struct timeval now,
 261                               void *private_data)
 262 {
 263         struct aio_child_list *list = talloc_get_type_abort(
 264                 private_data, struct aio_child_list);
 265         struct aio_child *child, *next;
 266
 267         TALLOC_FREE(list->cleanup_event);
 268
 269         for (child = list->children; child != NULL; child = next) {
 270                 next = child->next;
 271
 272                 if (child->busy) {
 273                         DEBUG(10, ("child %d currently active\n",
 274                                    (int)child->pid));
 275                         continue;
 276                 }
 277
 278                 if (child->dont_delete) {
 279                         DEBUG(10, ("Child %d was active since last cleanup\n",
 280                                    (int)child->pid));
 281                         child->dont_delete = false;
 282                         continue;
 283                 }
 284
 285                 DEBUG(10, ("Child %d idle for more than 30 seconds, "
 286                            "deleting\n", (int)child->pid));
 287
 288                 TALLOC_FREE(child);
 289                 child = next;
 290         }
 291
 292         if (list->children != NULL) {
 293                 /*
 294                  * Re-schedule the next cleanup round
 295                  */
 296                 list->cleanup_event = tevent_add_timer(server_event_context(), list,
 297                                                       timeval_add(&now, 30, 0),
 298                                                       aio_child_cleanup, list);
 299
 300         }
 301 }
 302
 303 static struct aio_child_list *init_aio_children(struct vfs_handle_struct *handle)
 304 {
 305         struct aio_child_list *data = NULL;
 306
 307         if (SMB_VFS_HANDLE_TEST_DATA(handle)) {
 308                 SMB_VFS_HANDLE_GET_DATA(handle, data, struct aio_child_list,
 309                                         return NULL);
 310         }
 311
 312         if (data == NULL) {
 313                 data = talloc_zero(NULL, struct aio_child_list);
 314                 if (data == NULL) {
 315                         return NULL;
 316                 }
 317         }
 318
 319         /*
 320          * Regardless of whether the child_list had been around or not, make
 321          * sure that we have a cleanup timed event. This timed event will
 322          * delete itself when it finds that no children are around anymore.
 323          */
 324
 325         if (data->cleanup_event == NULL) {
 326                 data->cleanup_event = tevent_add_timer(server_event_context(), data,
 327                                                       timeval_current_ofs(30, 0),
 328                                                       aio_child_cleanup, data);
 329                 if (data->cleanup_event == NULL) {
 330                         TALLOC_FREE(data);
 331                         return NULL;
 332                 }
 333         }
 334
 335         if (!SMB_VFS_HANDLE_TEST_DATA(handle)) {
 336                 SMB_VFS_HANDLE_SET_DATA(handle, data, free_aio_children,
 337                                         struct aio_child_list, return False);
 338         }
 339
 340         return data;
 341 }
 342
 343 static void aio_child_loop(int sockfd, struct mmap_area *map)
 344 {
 345         while (true) {
 346                 int fd = -1;
 347                 ssize_t ret;
 348                 struct rw_cmd cmd_struct;
 349                 struct rw_ret ret_struct;
 350
 351                 ret = read_fd(sockfd, &cmd_struct, sizeof(cmd_struct), &fd);
 352                 if (ret != sizeof(cmd_struct)) {
 353                         DEBUG(10, ("read_fd returned %d: %s\n", (int)ret,
 354                                    strerror(errno)));
 355                         exit(1);
 356                 }
 357
 358                 DEBUG(10, ("aio_child_loop: %s %d bytes at %d from fd %d\n",
 359                            cmd_type_str(cmd_struct.cmd),
 360                            (int)cmd_struct.n, (int)cmd_struct.offset, fd));
 361
 362                 if (cmd_struct.erratic_testing_mode) {
 363                         /*
 364                          * For developer testing, we want erratic behaviour for
 365                          * async I/O times
 366                          */
 367                         uint8_t randval;
 368                         unsigned msecs;
 369                         /*
 370                          * use generate_random_buffer, we just forked from a
 371                          * common parent state
 372                          */
 373                         generate_random_buffer(&randval, sizeof(randval));
 374                         msecs = randval + 20;
 375                         DEBUG(10, ("delaying for %u msecs\n", msecs));
 376                         smb_msleep(msecs);
 377                 }
 378
 379                 ZERO_STRUCT(ret_struct);
 380
 381                 switch (cmd_struct.cmd) {
 382                 case READ_CMD:
 383                         ret_struct.size = sys_pread(
 384                                 fd, (void *)map->ptr, cmd_struct.n,
 385                                 cmd_struct.offset);
 386 #if 0
 387 /* This breaks "make test" when run with aio_fork module. */
 388 #ifdef DEVELOPER
 389                         ret_struct.size = MAX(1, ret_struct.size * 0.9);
 390 #endif
 391 #endif
 392                         break;
 393                 case WRITE_CMD:
 394                         ret_struct.size = sys_pwrite(
 395                                 fd, (void *)map->ptr, cmd_struct.n,
 396                                 cmd_struct.offset);
 397                         break;
 398                 case FSYNC_CMD:
 399                         ret_struct.size = fsync(fd);
 400                         break;
 401                 default:
 402                         ret_struct.size = -1;
 403                         errno = EINVAL;
 404                 }
 405
 406                 DEBUG(10, ("aio_child_loop: syscall returned %d\n",
 407                            (int)ret_struct.size));
 408
 409                 if (ret_struct.size == -1) {
 410                         ret_struct.ret_errno = errno;
 411                 }
 412
 413                 /*
 414                  * Close the fd before telling our parent we're done. The
 415                  * parent might close and re-open the file very quickly, and
 416                  * with system-level share modes (GPFS) we would get an
 417                  * unjustified SHARING_VIOLATION.
 418                  */
 419                 close(fd);
 420
 421                 ret = write_data(sockfd, (char *)&ret_struct,
 422                                  sizeof(ret_struct));
 423                 if (ret != sizeof(ret_struct)) {
 424                         DEBUG(10, ("could not write ret_struct: %s\n",
 425                                    strerror(errno)));
 426                         exit(2);
 427                 }
 428         }
 429 }
 430
 431 static int aio_child_destructor(struct aio_child *child)
 432 {
 433         char c=0;
 434
 435         SMB_ASSERT(!child->busy);
 436
 437         DEBUG(10, ("aio_child_destructor: removing child %d on fd %d\n",
 438                         child->pid, child->sockfd));
 439
 440         /*
 441          * closing the sockfd makes the child not return from recvmsg() on RHEL
 442          * 5.5 so instead force the child to exit by writing bad data to it
 443          */
 444         write(child->sockfd, &c, sizeof(c));
 445         close(child->sockfd);
 446         DLIST_REMOVE(child->list->children, child);
 447         return 0;
 448 }
 449
 450 /*
 451  * We have to close all fd's in open files, we might incorrectly hold a system
 452  * level share mode on a file.
 453  */
 454
 455 static struct files_struct *close_fsp_fd(struct files_struct *fsp,
 456                                          void *private_data)
 457 {
 458         if ((fsp->fh != NULL) && (fsp->fh->fd != -1)) {
 459                 close(fsp->fh->fd);
 460                 fsp->fh->fd = -1;
 461         }
 462         return NULL;
 463 }
 464
 465 static int create_aio_child(struct smbd_server_connection *sconn,
 466                             struct aio_child_list *children,
 467                             size_t map_size,
 468                             struct aio_child **presult)
 469 {
 470         struct aio_child *result;
 471         int fdpair[2];
 472         int ret;
 473
 474         fdpair[0] = fdpair[1] = -1;
 475
 476         result = talloc_zero(children, struct aio_child);
 477         if (result == NULL) {
 478                 return ENOMEM;
 479         }
 480
 481         if (socketpair(AF_UNIX, SOCK_STREAM, 0, fdpair) == -1) {
 482                 ret = errno;
 483                 DEBUG(10, ("socketpair() failed: %s\n", strerror(errno)));
 484                 goto fail;
 485         }
 486
 487         DEBUG(10, ("fdpair = %d/%d\n", fdpair[0], fdpair[1]));
 488
 489         result->map = mmap_area_init(result, map_size);
 490         if (result->map == NULL) {
 491                 ret = errno;
 492                 DEBUG(0, ("Could not create mmap area\n"));
 493                 goto fail;
 494         }
 495
 496         result->pid = fork();
 497         if (result->pid == -1) {
 498                 ret = errno;
 499                 DEBUG(0, ("fork failed: %s\n", strerror(errno)));
 500                 goto fail;
 501         }
 502
 503         if (result->pid == 0) {
 504                 close(fdpair[0]);
 505                 result->sockfd = fdpair[1];
 506                 files_forall(sconn, close_fsp_fd, NULL);
 507                 aio_child_loop(result->sockfd, result->map);
 508         }
 509
 510         DEBUG(10, ("Child %d created with sockfd %d\n",
 511                         result->pid, fdpair[0]));
 512
 513         result->sockfd = fdpair[0];
 514         close(fdpair[1]);
 515
 516         result->list = children;
 517         DLIST_ADD(children->children, result);
 518
 519         talloc_set_destructor(result, aio_child_destructor);
 520
 521         *presult = result;
 522
 523         return 0;
 524
 525  fail:
 526         if (fdpair[0] != -1) close(fdpair[0]);
 527         if (fdpair[1] != -1) close(fdpair[1]);
 528         TALLOC_FREE(result);
 529
 530         return ret;
 531 }
 532
 533 static int get_idle_child(struct vfs_handle_struct *handle,
 534                           struct aio_child **pchild)
 535 {
 536         struct aio_child_list *children;
 537         struct aio_child *child;
 538
 539         children = init_aio_children(handle);
 540         if (children == NULL) {
 541                 return ENOMEM;
 542         }
 543
 544         for (child = children->children; child != NULL; child = child->next) {
 545                 if (!child->busy) {
 546                         break;
 547                 }
 548         }
 549
 550         if (child == NULL) {
 551                 int ret;
 552
 553                 DEBUG(10, ("no idle child found, creating new one\n"));
 554
 555                 ret = create_aio_child(handle->conn->sconn, children,
 556                                           128*1024, &child);
 557                 if (ret != 0) {
 558                         DEBUG(10, ("create_aio_child failed: %s\n",
 559                                    strerror(errno)));
 560                         return ret;
 561                 }
 562         }
 563
 564         child->dont_delete = true;
 565         child->busy = true;
 566
 567         *pchild = child;
 568         return 0;
 569 }
 570
 571 struct aio_fork_pread_state {
 572         struct aio_child *child;
 573         ssize_t ret;
 574         int err;
 575 };
 576
 577 static void aio_fork_pread_done(struct tevent_req *subreq);
 578
 579 static struct tevent_req *aio_fork_pread_send(struct vfs_handle_struct *handle,
 580                                               TALLOC_CTX *mem_ctx,
 581                                               struct tevent_context *ev,
 582                                               struct files_struct *fsp,
 583                                               void *data,
 584                                               size_t n, off_t offset)
 585 {
 586         struct tevent_req *req, *subreq;
 587         struct aio_fork_pread_state *state;
 588         struct rw_cmd cmd;
 589         ssize_t written;
 590         int err;
 591         struct aio_fork_config *config;
 592
 593         SMB_VFS_HANDLE_GET_DATA(handle, config,
 594                                 struct aio_fork_config,
 595                                 return NULL);
 596
 597         req = tevent_req_create(mem_ctx, &state, struct aio_fork_pread_state);
 598         if (req == NULL) {
 599                 return NULL;
 600         }
 601
 602         if (n > 128*1024) {
 603                 /* TODO: support variable buffers */
 604                 tevent_req_error(req, EINVAL);
 605                 return tevent_req_post(req, ev);
 606         }
 607
 608         err = get_idle_child(handle, &state->child);
 609         if (err != 0) {
 610                 tevent_req_error(req, err);
 611                 return tevent_req_post(req, ev);
 612         }
 613
 614         ZERO_STRUCT(cmd);
 615         cmd.n = n;
 616         cmd.offset = offset;
 617         cmd.cmd = READ_CMD;
 618         cmd.erratic_testing_mode = config->erratic_testing_mode;
 619
 620         DEBUG(10, ("sending fd %d to child %d\n", fsp->fh->fd,
 621                    (int)state->child->pid));
 622
 623         /*
 624          * Not making this async. We're writing into an empty unix
 625          * domain socket. This should never block.
 626          */
 627         written = write_fd(state->child->sockfd, &cmd, sizeof(cmd),
 628                            fsp->fh->fd);
 629         if (written == -1) {
 630                 err = errno;
 631
 632                 TALLOC_FREE(state->child);
 633
 634                 DEBUG(10, ("write_fd failed: %s\n", strerror(err)));
 635                 tevent_req_error(req, err);
 636                 return tevent_req_post(req, ev);
 637         }
 638
 639         subreq = read_packet_send(state, ev, state->child->sockfd,
 640                                   sizeof(struct rw_ret), NULL, NULL);
 641         if (tevent_req_nomem(subreq, req)) {
 642                 TALLOC_FREE(state->child); /* we sent sth down */
 643                 return tevent_req_post(req, ev);
 644         }
 645         tevent_req_set_callback(subreq, aio_fork_pread_done, req);
 646         return req;
 647 }
 648
 649 static void aio_fork_pread_done(struct tevent_req *subreq)
 650 {
 651         struct tevent_req *req = tevent_req_callback_data(
 652                 subreq, struct tevent_req);
 653         struct aio_fork_pread_state *state = tevent_req_data(
 654                 req, struct aio_fork_pread_state);
 655         ssize_t nread;
 656         uint8_t *buf;
 657         int err;
 658         struct rw_ret *retbuf;
 659
 660         nread = read_packet_recv(subreq, talloc_tos(), &buf, &err);
 661         TALLOC_FREE(subreq);
 662         if (nread == -1) {
 663                 TALLOC_FREE(state->child);
 664                 tevent_req_error(req, err);
 665                 return;
 666         }
 667
 668         state->child->busy = false;
 669
 670         retbuf = (struct rw_ret *)buf;
 671         state->ret = retbuf->size;
 672         state->err = retbuf->ret_errno;
 673         tevent_req_done(req);
 674 }
 675
 676 static ssize_t aio_fork_pread_recv(struct tevent_req *req, int *err)
 677 {
 678         struct aio_fork_pread_state *state = tevent_req_data(
 679                 req, struct aio_fork_pread_state);
 680
 681         if (tevent_req_is_unix_error(req, err)) {
 682                 return -1;
 683         }
 684         if (state->ret == -1) {
 685                 *err = state->err;
 686         }
 687         return state->ret;
 688 }
 689
 690 struct aio_fork_pwrite_state {
 691         struct aio_child *child;
 692         ssize_t ret;
 693         int err;
 694 };
 695
 696 static void aio_fork_pwrite_done(struct tevent_req *subreq);
 697
 698 static struct tevent_req *aio_fork_pwrite_send(
 699         struct vfs_handle_struct *handle, TALLOC_CTX *mem_ctx,
 700         struct tevent_context *ev, struct files_struct *fsp,
 701         const void *data, size_t n, off_t offset)
 702 {
 703         struct tevent_req *req, *subreq;
 704         struct aio_fork_pwrite_state *state;
 705         struct rw_cmd cmd;
 706         ssize_t written;
 707         int err;
 708         struct aio_fork_config *config;
 709         SMB_VFS_HANDLE_GET_DATA(handle, config,
 710                                 struct aio_fork_config,
 711                                 return NULL);
 712
 713         req = tevent_req_create(mem_ctx, &state, struct aio_fork_pwrite_state);
 714         if (req == NULL) {
 715                 return NULL;
 716         }
 717
 718         if (n > 128*1024) {
 719                 /* TODO: support variable buffers */
 720                 tevent_req_error(req, EINVAL);
 721                 return tevent_req_post(req, ev);
 722         }
 723
 724         err = get_idle_child(handle, &state->child);
 725         if (err != 0) {
 726                 tevent_req_error(req, err);
 727                 return tevent_req_post(req, ev);
 728         }
 729
 730         ZERO_STRUCT(cmd);
 731         cmd.n = n;
 732         cmd.offset = offset;
 733         cmd.cmd = WRITE_CMD;
 734         cmd.erratic_testing_mode = config->erratic_testing_mode;
 735
 736         DEBUG(10, ("sending fd %d to child %d\n", fsp->fh->fd,
 737                    (int)state->child->pid));
 738
 739         /*
 740          * Not making this async. We're writing into an empty unix
 741          * domain socket. This should never block.
 742          */
 743         written = write_fd(state->child->sockfd, &cmd, sizeof(cmd),
 744                            fsp->fh->fd);
 745         if (written == -1) {
 746                 err = errno;
 747
 748                 TALLOC_FREE(state->child);
 749
 750                 DEBUG(10, ("write_fd failed: %s\n", strerror(err)));
 751                 tevent_req_error(req, err);
 752                 return tevent_req_post(req, ev);
 753         }
 754
 755         subreq = read_packet_send(state, ev, state->child->sockfd,
 756                                   sizeof(struct rw_ret), NULL, NULL);
 757         if (tevent_req_nomem(subreq, req)) {
 758                 TALLOC_FREE(state->child); /* we sent sth down */
 759                 return tevent_req_post(req, ev);
 760         }
 761         tevent_req_set_callback(subreq, aio_fork_pwrite_done, req);
 762         return req;
 763 }
 764
 765 static void aio_fork_pwrite_done(struct tevent_req *subreq)
 766 {
 767         struct tevent_req *req = tevent_req_callback_data(
 768                 subreq, struct tevent_req);
 769         struct aio_fork_pwrite_state *state = tevent_req_data(
 770                 req, struct aio_fork_pwrite_state);
 771         ssize_t nread;
 772         uint8_t *buf;
 773         int err;
 774         struct rw_ret *retbuf;
 775
 776         nread = read_packet_recv(subreq, talloc_tos(), &buf, &err);
 777         TALLOC_FREE(subreq);
 778         if (nread == -1) {
 779                 TALLOC_FREE(state->child);
 780                 tevent_req_error(req, err);
 781                 return;
 782         }
 783
 784         state->child->busy = false;
 785
 786         retbuf = (struct rw_ret *)buf;
 787         state->ret = retbuf->size;
 788         state->err = retbuf->ret_errno;
 789         tevent_req_done(req);
 790 }
 791
 792 static ssize_t aio_fork_pwrite_recv(struct tevent_req *req, int *err)
 793 {
 794         struct aio_fork_pwrite_state *state = tevent_req_data(
 795                 req, struct aio_fork_pwrite_state);
 796
 797         if (tevent_req_is_unix_error(req, err)) {
 798                 return -1;
 799         }
 800         if (state->ret == -1) {
 801                 *err = state->err;
 802         }
 803         return state->ret;
 804 }
 805
 806 struct aio_fork_fsync_state {
 807         struct aio_child *child;
 808         ssize_t ret;
 809         int err;
 810 };
 811
 812 static void aio_fork_fsync_done(struct tevent_req *subreq);
 813
 814 static struct tevent_req *aio_fork_fsync_send(
 815         struct vfs_handle_struct *handle, TALLOC_CTX *mem_ctx,
 816         struct tevent_context *ev, struct files_struct *fsp)
 817 {
 818         struct tevent_req *req, *subreq;
 819         struct aio_fork_fsync_state *state;
 820         struct rw_cmd cmd;
 821         ssize_t written;
 822         int err;
 823         struct aio_fork_config *config;
 824
 825         SMB_VFS_HANDLE_GET_DATA(handle, config,
 826                                 struct aio_fork_config,
 827                                 return NULL);
 828
 829         req = tevent_req_create(mem_ctx, &state, struct aio_fork_fsync_state);
 830         if (req == NULL) {
 831                 return NULL;
 832         }
 833
 834         err = get_idle_child(handle, &state->child);
 835         if (err != 0) {
 836                 tevent_req_error(req, err);
 837                 return tevent_req_post(req, ev);
 838         }
 839
 840         ZERO_STRUCT(cmd);
 841         cmd.cmd = FSYNC_CMD;
 842         cmd.erratic_testing_mode = config->erratic_testing_mode;
 843
 844         DEBUG(10, ("sending fd %d to child %d\n", fsp->fh->fd,
 845                    (int)state->child->pid));
 846
 847         /*
 848          * Not making this async. We're writing into an empty unix
 849          * domain socket. This should never block.
 850          */
 851         written = write_fd(state->child->sockfd, &cmd, sizeof(cmd),
 852                            fsp->fh->fd);
 853         if (written == -1) {
 854                 err = errno;
 855
 856                 TALLOC_FREE(state->child);
 857
 858                 DEBUG(10, ("write_fd failed: %s\n", strerror(err)));
 859                 tevent_req_error(req, err);
 860                 return tevent_req_post(req, ev);
 861         }
 862
 863         subreq = read_packet_send(state, ev, state->child->sockfd,
 864                                   sizeof(struct rw_ret), NULL, NULL);
 865         if (tevent_req_nomem(subreq, req)) {
 866                 TALLOC_FREE(state->child); /* we sent sth down */
 867                 return tevent_req_post(req, ev);
 868         }
 869         tevent_req_set_callback(subreq, aio_fork_fsync_done, req);
 870         return req;
 871 }
 872
 873 static void aio_fork_fsync_done(struct tevent_req *subreq)
 874 {
 875         struct tevent_req *req = tevent_req_callback_data(
 876                 subreq, struct tevent_req);
 877         struct aio_fork_fsync_state *state = tevent_req_data(
 878                 req, struct aio_fork_fsync_state);
 879         ssize_t nread;
 880         uint8_t *buf;
 881         int err;
 882         struct rw_ret *retbuf;
 883
 884         nread = read_packet_recv(subreq, talloc_tos(), &buf, &err);
 885         TALLOC_FREE(subreq);
 886         if (nread == -1) {
 887                 TALLOC_FREE(state->child);
 888                 tevent_req_error(req, err);
 889                 return;
 890         }
 891
 892         state->child->busy = false;
 893
 894         retbuf = (struct rw_ret *)buf;
 895         state->ret = retbuf->size;
 896         state->err = retbuf->ret_errno;
 897         tevent_req_done(req);
 898 }
 899
 900 static int aio_fork_fsync_recv(struct tevent_req *req, int *err)
 901 {
 902         struct aio_fork_fsync_state *state = tevent_req_data(
 903                 req, struct aio_fork_fsync_state);
 904
 905         if (tevent_req_is_unix_error(req, err)) {
 906                 return -1;
 907         }
 908         if (state->ret == -1) {
 909                 *err = state->err;
 910         }
 911         return state->ret;
 912 }
 913
 914 static int aio_fork_connect(vfs_handle_struct *handle, const char *service,
 915                             const char *user)
 916 {
 917         int ret;
 918         struct aio_fork_config *config;
 919         ret = SMB_VFS_NEXT_CONNECT(handle, service, user);
 920
 921         if (ret < 0) {
 922                 return ret;
 923         }
 924
 925         config = talloc_zero(handle->conn, struct aio_fork_config);
 926         if (!config) {
 927                 SMB_VFS_NEXT_DISCONNECT(handle);
 928                 DEBUG(0, ("talloc_zero() failed\n"));
 929                 return -1;
 930         }
 931
 932         config->erratic_testing_mode = lp_parm_bool(SNUM(handle->conn), "vfs_aio_fork",
 933                                                     "erratic_testing_mode", false);
 934
 935         SMB_VFS_HANDLE_SET_DATA(handle, config,
 936                                 NULL, struct aio_fork_config,
 937                                 return -1);
 938
 939         /*********************************************************************
 940          * How many threads to initialize ?
 941          * 100 per process seems insane as a default until you realize that
 942          * (a) Threads terminate after 1 second when idle.
 943          * (b) Throttling is done in SMB2 via the crediting algorithm.
 944          * (c) SMB1 clients are limited to max_mux (50) outstanding
 945          *     requests and Windows clients don't use this anyway.
 946          * Essentially we want this to be unlimited unless smb.conf
 947          * says different.
 948          *********************************************************************/
 949         aio_pending_size = 100;
 950         return 0;
 951 }
 952
 953 static struct vfs_fn_pointers vfs_aio_fork_fns = {
 954         .connect_fn = aio_fork_connect,
 955         .pread_send_fn = aio_fork_pread_send,
 956         .pread_recv_fn = aio_fork_pread_recv,
 957         .pwrite_send_fn = aio_fork_pwrite_send,
 958         .pwrite_recv_fn = aio_fork_pwrite_recv,
 959         .fsync_send_fn = aio_fork_fsync_send,
 960         .fsync_recv_fn = aio_fork_fsync_recv,
 961 };
 962
 963 NTSTATUS vfs_aio_fork_init(void);
 964 NTSTATUS vfs_aio_fork_init(void)
 965 {
 966         return smb_register_vfs(SMB_VFS_INTERFACE_VERSION,
 967                                 "aio_fork", &vfs_aio_fork_fns);
 968 }