ipc/msg.c

   1 /*
   2  * linux/ipc/msg.c
   3  * Copyright (C) 1992 Krishna Balasubramanian
   4  *
   5  * Removed all the remaining kerneld mess
   6  * Catch the -EFAULT stuff properly
   7  * Use GFP_KERNEL for messages as in 1.2
   8  * Fixed up the unchecked user space derefs
   9  * Copyright (C) 1998 Alan Cox & Andi Kleen
  10  *
  11  * /proc/sysvipc/msg support (c) 1999 Dragos Acostachioaie <dragos@iname.com>
  12  *
  13  * mostly rewritten, threaded and wake-one semantics added
  14  * MSGMAX limit removed, sysctl's added
  15  * (c) 1999 Manfred Spraul <manfred@colorfullife.com>
  16  *
  17  * support for audit of ipc object properties and permission changes
  18  * Dustin Kirkland <dustin.kirkland@us.ibm.com>
  19  *
  20  * namespaces support
  21  * OpenVZ, SWsoft Inc.
  22  * Pavel Emelianov <xemul@openvz.org>
  23  */
  24
  25 #include <linux/capability.h>
  26 #include <linux/slab.h>
  27 #include <linux/msg.h>
  28 #include <linux/spinlock.h>
  29 #include <linux/init.h>
  30 #include <linux/proc_fs.h>
  31 #include <linux/list.h>
  32 #include <linux/security.h>
  33 #include <linux/sched.h>
  34 #include <linux/syscalls.h>
  35 #include <linux/audit.h>
  36 #include <linux/seq_file.h>
  37 #include <linux/mutex.h>
  38 #include <linux/nsproxy.h>
  39
  40 #include <asm/current.h>
  41 #include <asm/uaccess.h>
  42 #include "util.h"
  43
  44 /*
  45  * one msg_receiver structure for each sleeping receiver:
  46  */
  47 struct msg_receiver {
  48         struct list_head        r_list;
  49         struct task_struct      *r_tsk;
  50
  51         int                     r_mode;
  52         long                    r_msgtype;
  53         long                    r_maxsize;
  54
  55         struct msg_msg          *volatile r_msg;
  56 };
  57
  58 /* one msg_sender for each sleeping sender */
  59 struct msg_sender {
  60         struct list_head        list;
  61         struct task_struct      *tsk;
  62 };
  63
  64 #define SEARCH_ANY              1
  65 #define SEARCH_EQUAL            2
  66 #define SEARCH_NOTEQUAL         3
  67 #define SEARCH_LESSEQUAL        4
  68
  69 static atomic_t msg_bytes =     ATOMIC_INIT(0);
  70 static atomic_t msg_hdrs =      ATOMIC_INIT(0);
  71
  72 static struct ipc_ids init_msg_ids;
  73
  74 #define msg_ids(ns)     (*((ns)->ids[IPC_MSG_IDS]))
  75
  76 #define msg_lock(ns, id)        ((struct msg_queue*)ipc_lock(&msg_ids(ns), id))
  77 #define msg_unlock(msq)         ipc_unlock(&(msq)->q_perm)
  78 #define msg_rmid(ns, id)        ((struct msg_queue*)ipc_rmid(&msg_ids(ns), id))
  79 #define msg_checkid(ns, msq, msgid)     \
  80         ipc_checkid(&msg_ids(ns), &msq->q_perm, msgid)
  81 #define msg_buildid(ns, id, seq) \
  82         ipc_buildid(&msg_ids(ns), id, seq)
  83
  84 static void freeque (struct ipc_namespace *ns, struct msg_queue *msq, int id);
  85 static int newque (struct ipc_namespace *ns, key_t key, int msgflg);
  86 #ifdef CONFIG_PROC_FS
  87 static int sysvipc_msg_proc_show(struct seq_file *s, void *it);
  88 #endif
  89
  90 static void __msg_init_ns(struct ipc_namespace *ns, struct ipc_ids *ids)
  91 {
  92         ns->ids[IPC_MSG_IDS] = ids;
  93         ns->msg_ctlmax = MSGMAX;
  94         ns->msg_ctlmnb = MSGMNB;
  95         ns->msg_ctlmni = MSGMNI;
  96         ipc_init_ids(ids, ns->msg_ctlmni);
  97 }
  98
  99 int msg_init_ns(struct ipc_namespace *ns)
 100 {
 101         struct ipc_ids *ids;
 102
 103         ids = kmalloc(sizeof(struct ipc_ids), GFP_KERNEL);
 104         if (ids == NULL)
 105                 return -ENOMEM;
 106
 107         __msg_init_ns(ns, ids);
 108         return 0;
 109 }
 110
 111 void msg_exit_ns(struct ipc_namespace *ns)
 112 {
 113         int i;
 114         struct msg_queue *msq;
 115
 116         mutex_lock(&msg_ids(ns).mutex);
 117         for (i = 0; i <= msg_ids(ns).max_id; i++) {
 118                 msq = msg_lock(ns, i);
 119                 if (msq == NULL)
 120                         continue;
 121
 122                 freeque(ns, msq, i);
 123         }
 124         mutex_unlock(&msg_ids(ns).mutex);
 125
 126         ipc_fini_ids(ns->ids[IPC_MSG_IDS]);
 127         kfree(ns->ids[IPC_MSG_IDS]);
 128         ns->ids[IPC_MSG_IDS] = NULL;
 129 }
 130
 131 void __init msg_init(void)
 132 {
 133         __msg_init_ns(&init_ipc_ns, &init_msg_ids);
 134         ipc_init_proc_interface("sysvipc/msg",
 135                                 "       key      msqid perms      cbytes       qnum lspid lrpid   uid   gid  cuid  cgid      stime      rtime      ctime\n",
 136                                 IPC_MSG_IDS, sysvipc_msg_proc_show);
 137 }
 138
 139 static int newque (struct ipc_namespace *ns, key_t key, int msgflg)
 140 {
 141         struct msg_queue *msq;
 142         int id, retval;
 143
 144         msq = ipc_rcu_alloc(sizeof(*msq));
 145         if (!msq)
 146                 return -ENOMEM;
 147
 148         msq->q_perm.mode = msgflg & S_IRWXUGO;
 149         msq->q_perm.key = key;
 150
 151         msq->q_perm.security = NULL;
 152         retval = security_msg_queue_alloc(msq);
 153         if (retval) {
 154                 ipc_rcu_putref(msq);
 155                 return retval;
 156         }
 157
 158         id = ipc_addid(&msg_ids(ns), &msq->q_perm, ns->msg_ctlmni);
 159         if (id == -1) {
 160                 security_msg_queue_free(msq);
 161                 ipc_rcu_putref(msq);
 162                 return -ENOSPC;
 163         }
 164
 165         msq->q_id = msg_buildid(ns, id, msq->q_perm.seq);
 166         msq->q_stime = msq->q_rtime = 0;
 167         msq->q_ctime = get_seconds();
 168         msq->q_cbytes = msq->q_qnum = 0;
 169         msq->q_qbytes = ns->msg_ctlmnb;
 170         msq->q_lspid = msq->q_lrpid = 0;
 171         INIT_LIST_HEAD(&msq->q_messages);
 172         INIT_LIST_HEAD(&msq->q_receivers);
 173         INIT_LIST_HEAD(&msq->q_senders);
 174         msg_unlock(msq);
 175
 176         return msq->q_id;
 177 }
 178
 179 static inline void ss_add(struct msg_queue *msq, struct msg_sender *mss)
 180 {
 181         mss->tsk = current;
 182         current->state = TASK_INTERRUPTIBLE;
 183         list_add_tail(&mss->list, &msq->q_senders);
 184 }
 185
 186 static inline void ss_del(struct msg_sender *mss)
 187 {
 188         if (mss->list.next != NULL)
 189                 list_del(&mss->list);
 190 }
 191
 192 static void ss_wakeup(struct list_head *h, int kill)
 193 {
 194         struct list_head *tmp;
 195
 196         tmp = h->next;
 197         while (tmp != h) {
 198                 struct msg_sender *mss;
 199
 200                 mss = list_entry(tmp, struct msg_sender, list);
 201                 tmp = tmp->next;
 202                 if (kill)
 203                         mss->list.next = NULL;
 204                 wake_up_process(mss->tsk);
 205         }
 206 }
 207
 208 static void expunge_all(struct msg_queue *msq, int res)
 209 {
 210         struct list_head *tmp;
 211
 212         tmp = msq->q_receivers.next;
 213         while (tmp != &msq->q_receivers) {
 214                 struct msg_receiver *msr;
 215
 216                 msr = list_entry(tmp, struct msg_receiver, r_list);
 217                 tmp = tmp->next;
 218                 msr->r_msg = NULL;
 219                 wake_up_process(msr->r_tsk);
 220                 smp_mb();
 221                 msr->r_msg = ERR_PTR(res);
 222         }
 223 }
 224
 225 /*
 226  * freeque() wakes up waiters on the sender and receiver waiting queue,
 227  * removes the message queue from message queue ID
 228  * array, and cleans up all the messages associated with this queue.
 229  *
 230  * msg_ids.mutex and the spinlock for this message queue is hold
 231  * before freeque() is called. msg_ids.mutex remains locked on exit.
 232  */
 233 static void freeque(struct ipc_namespace *ns, struct msg_queue *msq, int id)
 234 {
 235         struct list_head *tmp;
 236
 237         expunge_all(msq, -EIDRM);
 238         ss_wakeup(&msq->q_senders, 1);
 239         msq = msg_rmid(ns, id);
 240         msg_unlock(msq);
 241
 242         tmp = msq->q_messages.next;
 243         while (tmp != &msq->q_messages) {
 244                 struct msg_msg *msg = list_entry(tmp, struct msg_msg, m_list);
 245
 246                 tmp = tmp->next;
 247                 atomic_dec(&msg_hdrs);
 248                 free_msg(msg);
 249         }
 250         atomic_sub(msq->q_cbytes, &msg_bytes);
 251         security_msg_queue_free(msq);
 252         ipc_rcu_putref(msq);
 253 }
 254
 255 asmlinkage long sys_msgget(key_t key, int msgflg)
 256 {
 257         struct msg_queue *msq;
 258         int id, ret = -EPERM;
 259         struct ipc_namespace *ns;
 260
 261         ns = current->nsproxy->ipc_ns;
 262
 263         mutex_lock(&msg_ids(ns).mutex);
 264         if (key == IPC_PRIVATE)
 265                 ret = newque(ns, key, msgflg);
 266         else if ((id = ipc_findkey(&msg_ids(ns), key)) == -1) { /* key not used */
 267                 if (!(msgflg & IPC_CREAT))
 268                         ret = -ENOENT;
 269                 else
 270                         ret = newque(ns, key, msgflg);
 271         } else if (msgflg & IPC_CREAT && msgflg & IPC_EXCL) {
 272                 ret = -EEXIST;
 273         } else {
 274                 msq = msg_lock(ns, id);
 275                 BUG_ON(msq == NULL);
 276                 if (ipcperms(&msq->q_perm, msgflg))
 277                         ret = -EACCES;
 278                 else {
 279                         int qid = msg_buildid(ns, id, msq->q_perm.seq);
 280
 281                         ret = security_msg_queue_associate(msq, msgflg);
 282                         if (!ret)
 283                                 ret = qid;
 284                 }
 285                 msg_unlock(msq);
 286         }
 287         mutex_unlock(&msg_ids(ns).mutex);
 288
 289         return ret;
 290 }
 291
 292 static inline unsigned long
 293 copy_msqid_to_user(void __user *buf, struct msqid64_ds *in, int version)
 294 {
 295         switch(version) {
 296         case IPC_64:
 297                 return copy_to_user(buf, in, sizeof(*in));
 298         case IPC_OLD:
 299         {
 300                 struct msqid_ds out;
 301
 302                 memset(&out, 0, sizeof(out));
 303
 304                 ipc64_perm_to_ipc_perm(&in->msg_perm, &out.msg_perm);
 305
 306                 out.msg_stime           = in->msg_stime;
 307                 out.msg_rtime           = in->msg_rtime;
 308                 out.msg_ctime           = in->msg_ctime;
 309
 310                 if (in->msg_cbytes > USHRT_MAX)
 311                         out.msg_cbytes  = USHRT_MAX;
 312                 else
 313                         out.msg_cbytes  = in->msg_cbytes;
 314                 out.msg_lcbytes         = in->msg_cbytes;
 315
 316                 if (in->msg_qnum > USHRT_MAX)
 317                         out.msg_qnum    = USHRT_MAX;
 318                 else
 319                         out.msg_qnum    = in->msg_qnum;
 320
 321                 if (in->msg_qbytes > USHRT_MAX)
 322                         out.msg_qbytes  = USHRT_MAX;
 323                 else
 324                         out.msg_qbytes  = in->msg_qbytes;
 325                 out.msg_lqbytes         = in->msg_qbytes;
 326
 327                 out.msg_lspid           = in->msg_lspid;
 328                 out.msg_lrpid           = in->msg_lrpid;
 329
 330                 return copy_to_user(buf, &out, sizeof(out));
 331         }
 332         default:
 333                 return -EINVAL;
 334         }
 335 }
 336
 337 struct msq_setbuf {
 338         unsigned long   qbytes;
 339         uid_t           uid;
 340         gid_t           gid;
 341         mode_t          mode;
 342 };
 343
 344 static inline unsigned long
 345 copy_msqid_from_user(struct msq_setbuf *out, void __user *buf, int version)
 346 {
 347         switch(version) {
 348         case IPC_64:
 349         {
 350                 struct msqid64_ds tbuf;
 351
 352                 if (copy_from_user(&tbuf, buf, sizeof(tbuf)))
 353                         return -EFAULT;
 354
 355                 out->qbytes             = tbuf.msg_qbytes;
 356                 out->uid                = tbuf.msg_perm.uid;
 357                 out->gid                = tbuf.msg_perm.gid;
 358                 out->mode               = tbuf.msg_perm.mode;
 359
 360                 return 0;
 361         }
 362         case IPC_OLD:
 363         {
 364                 struct msqid_ds tbuf_old;
 365
 366                 if (copy_from_user(&tbuf_old, buf, sizeof(tbuf_old)))
 367                         return -EFAULT;
 368
 369                 out->uid                = tbuf_old.msg_perm.uid;
 370                 out->gid                = tbuf_old.msg_perm.gid;
 371                 out->mode               = tbuf_old.msg_perm.mode;
 372
 373                 if (tbuf_old.msg_qbytes == 0)
 374                         out->qbytes     = tbuf_old.msg_lqbytes;
 375                 else
 376                         out->qbytes     = tbuf_old.msg_qbytes;
 377
 378                 return 0;
 379         }
 380         default:
 381                 return -EINVAL;
 382         }
 383 }
 384
 385 asmlinkage long sys_msgctl(int msqid, int cmd, struct msqid_ds __user *buf)
 386 {
 387         struct kern_ipc_perm *ipcp;
 388         struct msq_setbuf uninitialized_var(setbuf);
 389         struct msg_queue *msq;
 390         int err, version;
 391         struct ipc_namespace *ns;
 392
 393         if (msqid < 0 || cmd < 0)
 394                 return -EINVAL;
 395
 396         version = ipc_parse_version(&cmd);
 397         ns = current->nsproxy->ipc_ns;
 398
 399         switch (cmd) {
 400         case IPC_INFO:
 401         case MSG_INFO:
 402         {
 403                 struct msginfo msginfo;
 404                 int max_id;
 405
 406                 if (!buf)
 407                         return -EFAULT;
 408                 /*
 409                  * We must not return kernel stack data.
 410                  * due to padding, it's not enough
 411                  * to set all member fields.
 412                  */
 413                 err = security_msg_queue_msgctl(NULL, cmd);
 414                 if (err)
 415                         return err;
 416
 417                 memset(&msginfo, 0, sizeof(msginfo));
 418                 msginfo.msgmni = ns->msg_ctlmni;
 419                 msginfo.msgmax = ns->msg_ctlmax;
 420                 msginfo.msgmnb = ns->msg_ctlmnb;
 421                 msginfo.msgssz = MSGSSZ;
 422                 msginfo.msgseg = MSGSEG;
 423                 mutex_lock(&msg_ids(ns).mutex);
 424                 if (cmd == MSG_INFO) {
 425                         msginfo.msgpool = msg_ids(ns).in_use;
 426                         msginfo.msgmap = atomic_read(&msg_hdrs);
 427                         msginfo.msgtql = atomic_read(&msg_bytes);
 428                 } else {
 429                         msginfo.msgmap = MSGMAP;
 430                         msginfo.msgpool = MSGPOOL;
 431                         msginfo.msgtql = MSGTQL;
 432                 }
 433                 max_id = msg_ids(ns).max_id;
 434                 mutex_unlock(&msg_ids(ns).mutex);
 435                 if (copy_to_user(buf, &msginfo, sizeof(struct msginfo)))
 436                         return -EFAULT;
 437                 return (max_id < 0) ? 0 : max_id;
 438         }
 439         case MSG_STAT:
 440         case IPC_STAT:
 441         {
 442                 struct msqid64_ds tbuf;
 443                 int success_return;
 444
 445                 if (!buf)
 446                         return -EFAULT;
 447                 if (cmd == MSG_STAT && msqid >= msg_ids(ns).entries->size)
 448                         return -EINVAL;
 449
 450                 memset(&tbuf, 0, sizeof(tbuf));
 451
 452                 msq = msg_lock(ns, msqid);
 453                 if (msq == NULL)
 454                         return -EINVAL;
 455
 456                 if (cmd == MSG_STAT) {
 457                         success_return = msg_buildid(ns, msqid, msq->q_perm.seq);
 458                 } else {
 459                         err = -EIDRM;
 460                         if (msg_checkid(ns, msq, msqid))
 461                                 goto out_unlock;
 462                         success_return = 0;
 463                 }
 464                 err = -EACCES;
 465                 if (ipcperms(&msq->q_perm, S_IRUGO))
 466                         goto out_unlock;
 467
 468                 err = security_msg_queue_msgctl(msq, cmd);
 469                 if (err)
 470                         goto out_unlock;
 471
 472                 kernel_to_ipc64_perm(&msq->q_perm, &tbuf.msg_perm);
 473                 tbuf.msg_stime  = msq->q_stime;
 474                 tbuf.msg_rtime  = msq->q_rtime;
 475                 tbuf.msg_ctime  = msq->q_ctime;
 476                 tbuf.msg_cbytes = msq->q_cbytes;
 477                 tbuf.msg_qnum   = msq->q_qnum;
 478                 tbuf.msg_qbytes = msq->q_qbytes;
 479                 tbuf.msg_lspid  = msq->q_lspid;
 480                 tbuf.msg_lrpid  = msq->q_lrpid;
 481                 msg_unlock(msq);
 482                 if (copy_msqid_to_user(buf, &tbuf, version))
 483                         return -EFAULT;
 484                 return success_return;
 485         }
 486         case IPC_SET:
 487                 if (!buf)
 488                         return -EFAULT;
 489                 if (copy_msqid_from_user(&setbuf, buf, version))
 490                         return -EFAULT;
 491                 break;
 492         case IPC_RMID:
 493                 break;
 494         default:
 495                 return  -EINVAL;
 496         }
 497
 498         mutex_lock(&msg_ids(ns).mutex);
 499         msq = msg_lock(ns, msqid);
 500         err = -EINVAL;
 501         if (msq == NULL)
 502                 goto out_up;
 503
 504         err = -EIDRM;
 505         if (msg_checkid(ns, msq, msqid))
 506                 goto out_unlock_up;
 507         ipcp = &msq->q_perm;
 508
 509         err = audit_ipc_obj(ipcp);
 510         if (err)
 511                 goto out_unlock_up;
 512         if (cmd == IPC_SET) {
 513                 err = audit_ipc_set_perm(setbuf.qbytes, setbuf.uid, setbuf.gid,
 514                                          setbuf.mode);
 515                 if (err)
 516                         goto out_unlock_up;
 517         }
 518
 519         err = -EPERM;
 520         if (current->euid != ipcp->cuid &&
 521             current->euid != ipcp->uid && !capable(CAP_SYS_ADMIN))
 522                 /* We _could_ check for CAP_CHOWN above, but we don't */
 523                 goto out_unlock_up;
 524
 525         err = security_msg_queue_msgctl(msq, cmd);
 526         if (err)
 527                 goto out_unlock_up;
 528
 529         switch (cmd) {
 530         case IPC_SET:
 531         {
 532                 err = -EPERM;
 533                 if (setbuf.qbytes > ns->msg_ctlmnb && !capable(CAP_SYS_RESOURCE))
 534                         goto out_unlock_up;
 535
 536                 msq->q_qbytes = setbuf.qbytes;
 537
 538                 ipcp->uid = setbuf.uid;
 539                 ipcp->gid = setbuf.gid;
 540                 ipcp->mode = (ipcp->mode & ~S_IRWXUGO) |
 541                              (S_IRWXUGO & setbuf.mode);
 542                 msq->q_ctime = get_seconds();
 543                 /* sleeping receivers might be excluded by
 544                  * stricter permissions.
 545                  */
 546                 expunge_all(msq, -EAGAIN);
 547                 /* sleeping senders might be able to send
 548                  * due to a larger queue size.
 549                  */
 550                 ss_wakeup(&msq->q_senders, 0);
 551                 msg_unlock(msq);
 552                 break;
 553         }
 554         case IPC_RMID:
 555                 freeque(ns, msq, msqid);
 556                 break;
 557         }
 558         err = 0;
 559 out_up:
 560         mutex_unlock(&msg_ids(ns).mutex);
 561         return err;
 562 out_unlock_up:
 563         msg_unlock(msq);
 564         goto out_up;
 565 out_unlock:
 566         msg_unlock(msq);
 567         return err;
 568 }
 569
 570 static int testmsg(struct msg_msg *msg, long type, int mode)
 571 {
 572         switch(mode)
 573         {
 574                 case SEARCH_ANY:
 575                         return 1;
 576                 case SEARCH_LESSEQUAL:
 577                         if (msg->m_type <=type)
 578                                 return 1;
 579                         break;
 580                 case SEARCH_EQUAL:
 581                         if (msg->m_type == type)
 582                                 return 1;
 583                         break;
 584                 case SEARCH_NOTEQUAL:
 585                         if (msg->m_type != type)
 586                                 return 1;
 587                         break;
 588         }
 589         return 0;
 590 }
 591
 592 static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg)
 593 {
 594         struct list_head *tmp;
 595
 596         tmp = msq->q_receivers.next;
 597         while (tmp != &msq->q_receivers) {
 598                 struct msg_receiver *msr;
 599
 600                 msr = list_entry(tmp, struct msg_receiver, r_list);
 601                 tmp = tmp->next;
 602                 if (testmsg(msg, msr->r_msgtype, msr->r_mode) &&
 603                     !security_msg_queue_msgrcv(msq, msg, msr->r_tsk,
 604                                                msr->r_msgtype, msr->r_mode)) {
 605
 606                         list_del(&msr->r_list);
 607                         if (msr->r_maxsize < msg->m_ts) {
 608                                 msr->r_msg = NULL;
 609                                 wake_up_process(msr->r_tsk);
 610                                 smp_mb();
 611                                 msr->r_msg = ERR_PTR(-E2BIG);
 612                         } else {
 613                                 msr->r_msg = NULL;
 614                                 msq->q_lrpid = msr->r_tsk->pid;
 615                                 msq->q_rtime = get_seconds();
 616                                 wake_up_process(msr->r_tsk);
 617                                 smp_mb();
 618                                 msr->r_msg = msg;
 619
 620                                 return 1;
 621                         }
 622                 }
 623         }
 624         return 0;
 625 }
 626
 627 long do_msgsnd(int msqid, long mtype, void __user *mtext,
 628                 size_t msgsz, int msgflg)
 629 {
 630         struct msg_queue *msq;
 631         struct msg_msg *msg;
 632         int err;
 633         struct ipc_namespace *ns;
 634
 635         ns = current->nsproxy->ipc_ns;
 636
 637         if (msgsz > ns->msg_ctlmax || (long) msgsz < 0 || msqid < 0)
 638                 return -EINVAL;
 639         if (mtype < 1)
 640                 return -EINVAL;
 641
 642         msg = load_msg(mtext, msgsz);
 643         if (IS_ERR(msg))
 644                 return PTR_ERR(msg);
 645
 646         msg->m_type = mtype;
 647         msg->m_ts = msgsz;
 648
 649         msq = msg_lock(ns, msqid);
 650         err = -EINVAL;
 651         if (msq == NULL)
 652                 goto out_free;
 653
 654         err= -EIDRM;
 655         if (msg_checkid(ns, msq, msqid))
 656                 goto out_unlock_free;
 657
 658         for (;;) {
 659                 struct msg_sender s;
 660
 661                 err = -EACCES;
 662                 if (ipcperms(&msq->q_perm, S_IWUGO))
 663                         goto out_unlock_free;
 664
 665                 err = security_msg_queue_msgsnd(msq, msg, msgflg);
 666                 if (err)
 667                         goto out_unlock_free;
 668
 669                 if (msgsz + msq->q_cbytes <= msq->q_qbytes &&
 670                                 1 + msq->q_qnum <= msq->q_qbytes) {
 671                         break;
 672                 }
 673
 674                 /* queue full, wait: */
 675                 if (msgflg & IPC_NOWAIT) {
 676                         err = -EAGAIN;
 677                         goto out_unlock_free;
 678                 }
 679                 ss_add(msq, &s);
 680                 ipc_rcu_getref(msq);
 681                 msg_unlock(msq);
 682                 schedule();
 683
 684                 ipc_lock_by_ptr(&msq->q_perm);
 685                 ipc_rcu_putref(msq);
 686                 if (msq->q_perm.deleted) {
 687                         err = -EIDRM;
 688                         goto out_unlock_free;
 689                 }
 690                 ss_del(&s);
 691
 692                 if (signal_pending(current)) {
 693                         err = -ERESTARTNOHAND;
 694                         goto out_unlock_free;
 695                 }
 696         }
 697
 698         msq->q_lspid = current->tgid;
 699         msq->q_stime = get_seconds();
 700
 701         if (!pipelined_send(msq, msg)) {
 702                 /* noone is waiting for this message, enqueue it */
 703                 list_add_tail(&msg->m_list, &msq->q_messages);
 704                 msq->q_cbytes += msgsz;
 705                 msq->q_qnum++;
 706                 atomic_add(msgsz, &msg_bytes);
 707                 atomic_inc(&msg_hdrs);
 708         }
 709
 710         err = 0;
 711         msg = NULL;
 712
 713 out_unlock_free:
 714         msg_unlock(msq);
 715 out_free:
 716         if (msg != NULL)
 717                 free_msg(msg);
 718         return err;
 719 }
 720
 721 asmlinkage long
 722 sys_msgsnd(int msqid, struct msgbuf __user *msgp, size_t msgsz, int msgflg)
 723 {
 724         long mtype;
 725
 726         if (get_user(mtype, &msgp->mtype))
 727                 return -EFAULT;
 728         return do_msgsnd(msqid, mtype, msgp->mtext, msgsz, msgflg);
 729 }
 730
 731 static inline int convert_mode(long *msgtyp, int msgflg)
 732 {
 733         /*
 734          *  find message of correct type.
 735          *  msgtyp = 0 => get first.
 736          *  msgtyp > 0 => get first message of matching type.
 737          *  msgtyp < 0 => get message with least type must be < abs(msgtype).
 738          */
 739         if (*msgtyp == 0)
 740                 return SEARCH_ANY;
 741         if (*msgtyp < 0) {
 742                 *msgtyp = -*msgtyp;
 743                 return SEARCH_LESSEQUAL;
 744         }
 745         if (msgflg & MSG_EXCEPT)
 746                 return SEARCH_NOTEQUAL;
 747         return SEARCH_EQUAL;
 748 }
 749
 750 long do_msgrcv(int msqid, long *pmtype, void __user *mtext,
 751                 size_t msgsz, long msgtyp, int msgflg)
 752 {
 753         struct msg_queue *msq;
 754         struct msg_msg *msg;
 755         int mode;
 756         struct ipc_namespace *ns;
 757
 758         if (msqid < 0 || (long) msgsz < 0)
 759                 return -EINVAL;
 760         mode = convert_mode(&msgtyp, msgflg);
 761         ns = current->nsproxy->ipc_ns;
 762
 763         msq = msg_lock(ns, msqid);
 764         if (msq == NULL)
 765                 return -EINVAL;
 766
 767         msg = ERR_PTR(-EIDRM);
 768         if (msg_checkid(ns, msq, msqid))
 769                 goto out_unlock;
 770
 771         for (;;) {
 772                 struct msg_receiver msr_d;
 773                 struct list_head *tmp;
 774
 775                 msg = ERR_PTR(-EACCES);
 776                 if (ipcperms(&msq->q_perm, S_IRUGO))
 777                         goto out_unlock;
 778
 779                 msg = ERR_PTR(-EAGAIN);
 780                 tmp = msq->q_messages.next;
 781                 while (tmp != &msq->q_messages) {
 782                         struct msg_msg *walk_msg;
 783
 784                         walk_msg = list_entry(tmp, struct msg_msg, m_list);
 785                         if (testmsg(walk_msg, msgtyp, mode) &&
 786                             !security_msg_queue_msgrcv(msq, walk_msg, current,
 787                                                        msgtyp, mode)) {
 788
 789                                 msg = walk_msg;
 790                                 if (mode == SEARCH_LESSEQUAL &&
 791                                                 walk_msg->m_type != 1) {
 792                                         msg = walk_msg;
 793                                         msgtyp = walk_msg->m_type - 1;
 794                                 } else {
 795                                         msg = walk_msg;
 796                                         break;
 797                                 }
 798                         }
 799                         tmp = tmp->next;
 800                 }
 801                 if (!IS_ERR(msg)) {
 802                         /*
 803                          * Found a suitable message.
 804                          * Unlink it from the queue.
 805                          */
 806                         if ((msgsz < msg->m_ts) && !(msgflg & MSG_NOERROR)) {
 807                                 msg = ERR_PTR(-E2BIG);
 808                                 goto out_unlock;
 809                         }
 810                         list_del(&msg->m_list);
 811                         msq->q_qnum--;
 812                         msq->q_rtime = get_seconds();
 813                         msq->q_lrpid = current->tgid;
 814                         msq->q_cbytes -= msg->m_ts;
 815                         atomic_sub(msg->m_ts, &msg_bytes);
 816                         atomic_dec(&msg_hdrs);
 817                         ss_wakeup(&msq->q_senders, 0);
 818                         msg_unlock(msq);
 819                         break;
 820                 }
 821                 /* No message waiting. Wait for a message */
 822                 if (msgflg & IPC_NOWAIT) {
 823                         msg = ERR_PTR(-ENOMSG);
 824                         goto out_unlock;
 825                 }
 826                 list_add_tail(&msr_d.r_list, &msq->q_receivers);
 827                 msr_d.r_tsk = current;
 828                 msr_d.r_msgtype = msgtyp;
 829                 msr_d.r_mode = mode;
 830                 if (msgflg & MSG_NOERROR)
 831                         msr_d.r_maxsize = INT_MAX;
 832                 else
 833                         msr_d.r_maxsize = msgsz;
 834                 msr_d.r_msg = ERR_PTR(-EAGAIN);
 835                 current->state = TASK_INTERRUPTIBLE;
 836                 msg_unlock(msq);
 837
 838                 schedule();
 839
 840                 /* Lockless receive, part 1:
 841                  * Disable preemption.  We don't hold a reference to the queue
 842                  * and getting a reference would defeat the idea of a lockless
 843                  * operation, thus the code relies on rcu to guarantee the
 844                  * existance of msq:
 845                  * Prior to destruction, expunge_all(-EIRDM) changes r_msg.
 846                  * Thus if r_msg is -EAGAIN, then the queue not yet destroyed.
 847                  * rcu_read_lock() prevents preemption between reading r_msg
 848                  * and the spin_lock() inside ipc_lock_by_ptr().
 849                  */
 850                 rcu_read_lock();
 851
 852                 /* Lockless receive, part 2:
 853                  * Wait until pipelined_send or expunge_all are outside of
 854                  * wake_up_process(). There is a race with exit(), see
 855                  * ipc/mqueue.c for the details.
 856                  */
 857                 msg = (struct msg_msg*)msr_d.r_msg;
 858                 while (msg == NULL) {
 859                         cpu_relax();
 860                         msg = (struct msg_msg *)msr_d.r_msg;
 861                 }
 862
 863                 /* Lockless receive, part 3:
 864                  * If there is a message or an error then accept it without
 865                  * locking.
 866                  */
 867                 if (msg != ERR_PTR(-EAGAIN)) {
 868                         rcu_read_unlock();
 869                         break;
 870                 }
 871
 872                 /* Lockless receive, part 3:
 873                  * Acquire the queue spinlock.
 874                  */
 875                 ipc_lock_by_ptr(&msq->q_perm);
 876                 rcu_read_unlock();
 877
 878                 /* Lockless receive, part 4:
 879                  * Repeat test after acquiring the spinlock.
 880                  */
 881                 msg = (struct msg_msg*)msr_d.r_msg;
 882                 if (msg != ERR_PTR(-EAGAIN))
 883                         goto out_unlock;
 884
 885                 list_del(&msr_d.r_list);
 886                 if (signal_pending(current)) {
 887                         msg = ERR_PTR(-ERESTARTNOHAND);
 888 out_unlock:
 889                         msg_unlock(msq);
 890                         break;
 891                 }
 892         }
 893         if (IS_ERR(msg))
 894                 return PTR_ERR(msg);
 895
 896         msgsz = (msgsz > msg->m_ts) ? msg->m_ts : msgsz;
 897         *pmtype = msg->m_type;
 898         if (store_msg(mtext, msg, msgsz))
 899                 msgsz = -EFAULT;
 900
 901         free_msg(msg);
 902
 903         return msgsz;
 904 }
 905
 906 asmlinkage long sys_msgrcv(int msqid, struct msgbuf __user *msgp, size_t msgsz,
 907                            long msgtyp, int msgflg)
 908 {
 909         long err, mtype;
 910
 911         err =  do_msgrcv(msqid, &mtype, msgp->mtext, msgsz, msgtyp, msgflg);
 912         if (err < 0)
 913                 goto out;
 914
 915         if (put_user(mtype, &msgp->mtype))
 916                 err = -EFAULT;
 917 out:
 918         return err;
 919 }
 920
 921 #ifdef CONFIG_PROC_FS
 922 static int sysvipc_msg_proc_show(struct seq_file *s, void *it)
 923 {
 924         struct msg_queue *msq = it;
 925
 926         return seq_printf(s,
 927                         "%10d %10d  %4o  %10lu %10lu %5u %5u %5u %5u %5u %5u %10lu %10lu %10lu\n",
 928                         msq->q_perm.key,
 929                         msq->q_id,
 930                         msq->q_perm.mode,
 931                         msq->q_cbytes,
 932                         msq->q_qnum,
 933                         msq->q_lspid,
 934                         msq->q_lrpid,
 935                         msq->q_perm.uid,
 936                         msq->q_perm.gid,
 937                         msq->q_perm.cuid,
 938                         msq->q_perm.cgid,
 939                         msq->q_stime,
 940                         msq->q_rtime,
 941                         msq->q_ctime);
 942 }
 943 #endif