code.grnet.gr Git - archipelago/blob - xseg/peers/kernel/xsegbd.c

   1 /* xsegbd.c
   2  *
   3  */
   4
   5 #include <linux/module.h>
   6 #include <linux/moduleparam.h>
   7 #include <linux/init.h>
   8 #include <linux/sched.h>
   9 #include <linux/kernel.h>
  10 #include <linux/slab.h>
  11 #include <linux/fs.h>
  12 #include <linux/errno.h>
  13 #include <linux/timer.h>
  14 #include <linux/types.h>
  15 #include <linux/vmalloc.h>
  16 #include <linux/genhd.h>
  17 #include <linux/blkdev.h>
  18 #include <linux/bio.h>
  19 #include <linux/device.h>
  20 #include <linux/completion.h>
  21 #include <linux/wait.h>
  22 #include <sys/kernel/segdev.h>
  23 #include "xsegbd.h"
  24 #include <xseg/protocol.h>
  25
  26 #define XSEGBD_MINORS 1
  27 /* define max request size to be used in xsegbd */
  28 //FIXME should we make this 4MB instead of 256KB ?
  29 #define XSEGBD_MAX_REQUEST_SIZE 262144U
  30
  31 MODULE_DESCRIPTION("xsegbd");
  32 MODULE_AUTHOR("XSEG");
  33 MODULE_LICENSE("GPL");
  34
  35 static long sector_size = 0;
  36 static long blksize = 512;
  37 static int major = 0;
  38 static int max_dev = 1024;
  39 static char name[XSEGBD_SEGMENT_NAMELEN] = "xsegbd";
  40 static char spec[256] = "segdev:xsegbd:4:1024:12";
  41
  42 module_param(sector_size, long, 0644);
  43 module_param(blksize, long, 0644);
  44 module_param(max_dev, int, 0644);
  45 module_param(major, int, 0644);
  46 module_param_string(name, name, sizeof(name), 0644);
  47 module_param_string(spec, spec, sizeof(spec), 0644);
  48
  49 static struct xsegbd xsegbd;
  50 static struct xsegbd_device **xsegbd_devices; /* indexed by portno */
  51 static DEFINE_MUTEX(xsegbd_mutex);
  52 static DEFINE_SPINLOCK(xsegbd_devices_lock);
  53
  54
  55 static void __xsegbd_get(struct xsegbd_device *xsegbd_dev)
  56 {
  57         atomic_inc(&xsegbd_dev->usercount);
  58 }
  59
  60 static void __xsegbd_put(struct xsegbd_device *xsegbd_dev)
  61 {
  62         atomic_dec(&xsegbd_dev->usercount);
  63         wake_up(&xsegbd_dev->wq);
  64 }
  65
  66 static struct xsegbd_device *__xsegbd_get_dev(unsigned long id)
  67 {
  68         struct xsegbd_device *xsegbd_dev = NULL;
  69
  70         spin_lock(&xsegbd_devices_lock);
  71         xsegbd_dev = xsegbd_devices[id];
  72         if (xsegbd_dev)
  73                 __xsegbd_get(xsegbd_dev);
  74         spin_unlock(&xsegbd_devices_lock);
  75
  76         return xsegbd_dev;
  77 }
  78
  79 /* ************************* */
  80 /* ***** sysfs helpers ***** */
  81 /* ************************* */
  82
  83 static struct xsegbd_device *dev_to_xsegbd(struct device *dev)
  84 {
  85         return container_of(dev, struct xsegbd_device, dev);
  86 }
  87
  88 static struct device *xsegbd_get_dev(struct xsegbd_device *xsegbd_dev)
  89 {
  90         /* FIXME */
  91         return get_device(&xsegbd_dev->dev);
  92 }
  93
  94 static void xsegbd_put_dev(struct xsegbd_device *xsegbd_dev)
  95 {
  96         put_device(&xsegbd_dev->dev);
  97 }
  98
  99 /* ************************* */
 100 /* ** XSEG Initialization ** */
 101 /* ************************* */
 102
 103 static void xseg_callback(uint32_t portno);
 104
 105 int xsegbd_xseg_init(void)
 106 {
 107         int r;
 108
 109         if (!xsegbd.name[0])
 110                 strncpy(xsegbd.name, name, XSEGBD_SEGMENT_NAMELEN);
 111
 112         r = xseg_initialize();
 113         if (r) {
 114                 XSEGLOG("cannot initialize 'segdev' peer");
 115                 goto err;
 116         }
 117
 118         r = xseg_parse_spec(spec, &xsegbd.config);
 119         if (r)
 120                 goto err;
 121
 122         if (strncmp(xsegbd.config.type, "segdev", 16))
 123                 XSEGLOG("WARNING: unexpected segment type '%s' vs 'segdev'",
 124                          xsegbd.config.type);
 125
 126         /* leave it here for now */
 127         XSEGLOG("joining segment");
 128         xsegbd.xseg = xseg_join(        xsegbd.config.type,
 129                                         xsegbd.config.name,
 130                                         "segdev",
 131                                         xseg_callback           );
 132         if (!xsegbd.xseg) {
 133                 XSEGLOG("cannot find segment");
 134                 r = -ENODEV;
 135                 goto err;
 136         }
 137
 138         return 0;
 139 err:
 140         return r;
 141
 142 }
 143
 144 int xsegbd_xseg_quit(void)
 145 {
 146         struct segdev *segdev;
 147
 148         /* make sure to unmap the segment first */
 149         segdev = segdev_get(0);
 150         clear_bit(SEGDEV_RESERVED, &segdev->flags);
 151         xsegbd.xseg->priv->segment_type.ops.unmap(xsegbd.xseg, xsegbd.xseg->segment_size);
 152         segdev_put(segdev);
 153
 154         return 0;
 155 }
 156
 157
 158 /* ***************************** */
 159 /* ** Block Device Operations ** */
 160 /* ***************************** */
 161
 162 static int xsegbd_open(struct block_device *bdev, fmode_t mode)
 163 {
 164         struct gendisk *disk = bdev->bd_disk;
 165         struct xsegbd_device *xsegbd_dev = disk->private_data;
 166
 167         xsegbd_get_dev(xsegbd_dev);
 168
 169         return 0;
 170 }
 171
 172 static int xsegbd_release(struct gendisk *gd, fmode_t mode)
 173 {
 174         struct xsegbd_device *xsegbd_dev = gd->private_data;
 175
 176         xsegbd_put_dev(xsegbd_dev);
 177
 178         return 0;
 179 }
 180
 181 static int xsegbd_ioctl(struct block_device *bdev, fmode_t mode,
 182                         unsigned int cmd, unsigned long arg)
 183 {
 184         return -ENOTTY;
 185 }
 186
 187 static const struct block_device_operations xsegbd_ops = {
 188         .owner          = THIS_MODULE,
 189         .open           = xsegbd_open,
 190         .release        = xsegbd_release,
 191         .ioctl          = xsegbd_ioctl
 192 };
 193
 194
 195 /* *************************** */
 196 /* ** Device Initialization ** */
 197 /* *************************** */
 198
 199 static void xseg_request_fn(struct request_queue *rq);
 200 static int xsegbd_get_size(struct xsegbd_device *xsegbd_dev);
 201 static int xsegbd_mapclose(struct xsegbd_device *xsegbd_dev);
 202
 203 static int xsegbd_dev_init(struct xsegbd_device *xsegbd_dev)
 204 {
 205         int ret = -ENOMEM;
 206         struct gendisk *disk;
 207         unsigned int max_request_size_bytes;
 208
 209         spin_lock_init(&xsegbd_dev->rqlock);
 210
 211         xsegbd_dev->xsegbd = &xsegbd;
 212
 213         xsegbd_dev->blk_queue = blk_alloc_queue(GFP_KERNEL);
 214         if (!xsegbd_dev->blk_queue)
 215                 goto out;
 216
 217         if (!blk_init_allocated_queue(xsegbd_dev->blk_queue,
 218                         xseg_request_fn, &xsegbd_dev->rqlock))
 219                 goto outqueue;
 220
 221         xsegbd_dev->blk_queue->queuedata = xsegbd_dev;
 222
 223         blk_queue_flush(xsegbd_dev->blk_queue, REQ_FLUSH | REQ_FUA);
 224         blk_queue_logical_block_size(xsegbd_dev->blk_queue, 512);
 225         blk_queue_physical_block_size(xsegbd_dev->blk_queue, blksize);
 226         blk_queue_bounce_limit(xsegbd_dev->blk_queue, BLK_BOUNCE_ANY);
 227
 228         //blk_queue_max_segments(dev->blk_queue, 512);
 229
 230         max_request_size_bytes = XSEGBD_MAX_REQUEST_SIZE;
 231         blk_queue_max_hw_sectors(xsegbd_dev->blk_queue, max_request_size_bytes >> 9);
 232         blk_queue_max_segment_size(xsegbd_dev->blk_queue, max_request_size_bytes);
 233         blk_queue_io_min(xsegbd_dev->blk_queue, max_request_size_bytes);
 234         blk_queue_io_opt(xsegbd_dev->blk_queue, max_request_size_bytes);
 235
 236         queue_flag_set_unlocked(QUEUE_FLAG_NONROT, xsegbd_dev->blk_queue);
 237
 238         /* vkoukis says we don't need partitions */
 239         xsegbd_dev->gd = disk = alloc_disk(1);
 240         if (!disk)
 241                 goto outqueue;
 242
 243         disk->major = xsegbd_dev->major;
 244         disk->first_minor = 0; // id * XSEGBD_MINORS;
 245         disk->fops = &xsegbd_ops;
 246         disk->queue = xsegbd_dev->blk_queue;
 247         disk->private_data = xsegbd_dev;
 248         disk->flags |= GENHD_FL_SUPPRESS_PARTITION_INFO;
 249         snprintf(disk->disk_name, 32, "xsegbd%u", xsegbd_dev->id);
 250
 251         ret = 0;
 252
 253         /* allow a non-zero sector_size parameter to override the disk size */
 254         if (sector_size)
 255                 xsegbd_dev->sectors = sector_size;
 256         else {
 257                 ret = xsegbd_get_size(xsegbd_dev);
 258                 if (ret)
 259                         goto outdisk;
 260         }
 261
 262         set_capacity(disk, xsegbd_dev->sectors);
 263         XSEGLOG("xsegbd active...");
 264         add_disk(disk); /* immediately activates the device */
 265
 266         return 0;
 267
 268
 269 outdisk:
 270         put_disk(xsegbd_dev->gd);
 271 outqueue:
 272         blk_cleanup_queue(xsegbd_dev->blk_queue);
 273         xsegbd_dev->blk_queue = NULL;
 274 out:
 275         xsegbd_dev->gd = NULL;
 276         return ret;
 277 }
 278
 279 static void xsegbd_dev_release(struct device *dev)
 280 {
 281         int ret;
 282         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 283
 284         xseg_cancel_wait(xsegbd_dev->xseg, xsegbd_dev->src_portno);
 285
 286         /* cleanup gendisk and blk_queue the right way */
 287         if (xsegbd_dev->gd) {
 288                 if (xsegbd_dev->gd->flags & GENHD_FL_UP)
 289                         del_gendisk(xsegbd_dev->gd);
 290
 291                 put_disk(xsegbd_dev->gd);
 292                 xsegbd_mapclose(xsegbd_dev);
 293         }
 294
 295         spin_lock(&xsegbd_devices_lock);
 296         BUG_ON(xsegbd_devices[xsegbd_dev->src_portno] != xsegbd_dev);
 297         xsegbd_devices[xsegbd_dev->src_portno] = NULL;
 298         spin_unlock(&xsegbd_devices_lock);
 299
 300         /* wait for all pending operations on device to end */
 301         wait_event(xsegbd_dev->wq, atomic_read(&xsegbd_dev->usercount) <= 1);
 302         if (xsegbd_dev->blk_queue)
 303                 blk_cleanup_queue(xsegbd_dev->blk_queue);
 304
 305
 306 //      if (xseg_free_requests(xsegbd_dev->xseg,
 307 //                      xsegbd_dev->src_portno, xsegbd_dev->nr_requests) < 0)
 308 //              XSEGLOG("Error trying to free requests!\n");
 309
 310
 311         //FIXME xseg_leave to free_up resources ?
 312         unregister_blkdev(xsegbd_dev->major, XSEGBD_NAME);
 313
 314         if (xsegbd_dev->blk_req_pending)
 315                 kfree(xsegbd_dev->blk_req_pending);
 316         xq_free(&xsegbd_dev->blk_queue_pending);
 317
 318         kfree(xsegbd_dev);
 319
 320         module_put(THIS_MODULE);
 321 }
 322
 323 /* ******************* */
 324 /* ** Critical Path ** */
 325 /* ******************* */
 326
 327 static void blk_to_xseg(struct xseg *xseg, struct xseg_request *xreq,
 328                         struct request *blkreq)
 329 {
 330         struct bio_vec *bvec;
 331         struct req_iterator iter;
 332         uint64_t off = 0;
 333         char *data = xseg_get_data(xseg, xreq);
 334         rq_for_each_segment(bvec, blkreq, iter) {
 335                 char *bdata = kmap_atomic(bvec->bv_page) + bvec->bv_offset;
 336                 memcpy(data + off, bdata, bvec->bv_len);
 337                 off += bvec->bv_len;
 338                 kunmap_atomic(bdata);
 339         }
 340 }
 341
 342 static void xseg_to_blk(struct xseg *xseg, struct xseg_request *xreq,
 343                         struct request *blkreq)
 344 {
 345         struct bio_vec *bvec;
 346         struct req_iterator iter;
 347         uint64_t off = 0;
 348         char *data = xseg_get_data(xseg, xreq);
 349         rq_for_each_segment(bvec, blkreq, iter) {
 350                 char *bdata = kmap_atomic(bvec->bv_page) + bvec->bv_offset;
 351                 memcpy(bdata, data + off, bvec->bv_len);
 352                 off += bvec->bv_len;
 353                 kunmap_atomic(bdata);
 354         }
 355 }
 356
 357 static void xseg_request_fn(struct request_queue *rq)
 358 {
 359         struct xseg_request *xreq;
 360         struct xsegbd_device *xsegbd_dev = rq->queuedata;
 361         struct request *blkreq;
 362         struct xsegbd_pending *pending;
 363         xqindex blkreq_idx;
 364         char *target;
 365         uint64_t datalen;
 366         xport p;
 367         int r;
 368         unsigned long flags;
 369
 370         __xsegbd_get(xsegbd_dev);
 371
 372         spin_unlock_irq(&xsegbd_dev->rqlock);
 373         for (;;) {
 374                 if (current_thread_info()->preempt_count || irqs_disabled()){
 375                         XSEGLOG("Current thread preempt_count: %d, irqs_disabled(): %lu ",
 376                                         current_thread_info()->preempt_count, irqs_disabled());
 377                 }
 378                 //XSEGLOG("Priority: %d", current_thread_info()->task->prio);
 379                 //XSEGLOG("Static priority: %d", current_thread_info()->task->static_prio);
 380                 //XSEGLOG("Normal priority: %d", current_thread_info()->task->normal_prio);
 381                 //XSEGLOG("Rt_priority: %u", current_thread_info()->task->rt_priority);
 382                 blkreq_idx = Noneidx;
 383                 xreq = xseg_get_request(xsegbd_dev->xseg, xsegbd_dev->src_portno,
 384                                 xsegbd_dev->dst_portno, X_ALLOC);
 385                 if (!xreq)
 386                         break;
 387
 388                 blkreq_idx = xq_pop_head(&xsegbd_dev->blk_queue_pending,
 389                                                 xsegbd_dev->src_portno);
 390                 if (blkreq_idx == Noneidx)
 391                         break;
 392
 393                 if (blkreq_idx >= xsegbd_dev->nr_requests) {
 394                         XSEGLOG("blkreq_idx >= xsegbd_dev->nr_requests");
 395                         BUG_ON(1);
 396                         break;
 397                 }
 398
 399
 400                 spin_lock_irqsave(&xsegbd_dev->rqlock, flags);
 401                 blkreq = blk_fetch_request(rq);
 402                 if (!blkreq){
 403                         spin_unlock_irqrestore(&xsegbd_dev->rqlock, flags);
 404                         break;
 405                 }
 406
 407                 if (blkreq->cmd_type != REQ_TYPE_FS) {
 408                         //we lose xreq here
 409                         XSEGLOG("non-fs cmd_type: %u. *shrug*", blkreq->cmd_type);
 410                         __blk_end_request_all(blkreq, 0);
 411                         spin_unlock_irqrestore(&xsegbd_dev->rqlock, flags);
 412                         continue;
 413                 }
 414                 spin_unlock_irqrestore(&xsegbd_dev->rqlock, flags);
 415                 if (current_thread_info()->preempt_count || irqs_disabled()){
 416                         XSEGLOG("Current thread preempt_count: %d, irqs_disabled(): %lu ",
 417                                         current_thread_info()->preempt_count, irqs_disabled());
 418                 }
 419
 420                 datalen = blk_rq_bytes(blkreq);
 421                 r = xseg_prep_request(xsegbd_dev->xseg, xreq,
 422                                         xsegbd_dev->targetlen, datalen);
 423                 if (r < 0) {
 424                         XSEGLOG("couldn't prep request");
 425                         blk_end_request_err(blkreq, r);
 426                         BUG_ON(1);
 427                         break;
 428                 }
 429                 r = -ENOMEM;
 430                 if (xreq->bufferlen - xsegbd_dev->targetlen < datalen){
 431                         XSEGLOG("malformed req buffers");
 432                         blk_end_request_err(blkreq, r);
 433                         BUG_ON(1);
 434                         break;
 435                 }
 436
 437                 target = xseg_get_target(xsegbd_dev->xseg, xreq);
 438                 strncpy(target, xsegbd_dev->target, xsegbd_dev->targetlen);
 439
 440                 pending = &xsegbd_dev->blk_req_pending[blkreq_idx];
 441                 pending->dev = xsegbd_dev;
 442                 pending->request = blkreq;
 443                 pending->comp = NULL;
 444
 445                 xreq->size = datalen;
 446                 xreq->offset = blk_rq_pos(blkreq) << 9;
 447                 xreq->priv = (uint64_t) blkreq_idx;
 448
 449                 /*
 450                 if (xreq->offset >= (sector_size << 9))
 451                         XSEGLOG("sector offset: %lu > %lu, flush:%u, fua:%u",
 452                                  blk_rq_pos(blkreq), sector_size,
 453                                  blkreq->cmd_flags & REQ_FLUSH,
 454                                  blkreq->cmd_flags & REQ_FUA);
 455                 */
 456
 457                 if (blkreq->cmd_flags & REQ_FLUSH)
 458                         xreq->flags |= XF_FLUSH;
 459
 460                 if (blkreq->cmd_flags & REQ_FUA)
 461                         xreq->flags |= XF_FUA;
 462
 463                 if (rq_data_dir(blkreq)) {
 464                         /* unlock for data transfers? */
 465                         blk_to_xseg(xsegbd_dev->xseg, xreq, blkreq);
 466                         xreq->op = X_WRITE;
 467                 } else {
 468                         xreq->op = X_READ;
 469                 }
 470
 471
 472                 r = -EIO;
 473                 p = xseg_submit(xsegbd_dev->xseg, xreq,
 474                                         xsegbd_dev->src_portno, X_ALLOC);
 475                 if (p == NoPort) {
 476                         XSEGLOG("coundn't submit req");
 477                         BUG_ON(1);
 478                         blk_end_request_err(blkreq, r);
 479                         break;
 480                 }
 481                 WARN_ON(xseg_signal(xsegbd_dev->xsegbd->xseg, p) < 0);
 482         }
 483         if (xreq)
 484                 BUG_ON(xseg_put_request(xsegbd_dev->xsegbd->xseg, xreq,
 485                                         xsegbd_dev->src_portno) == -1);
 486         if (blkreq_idx != Noneidx)
 487                 BUG_ON(xq_append_head(&xsegbd_dev->blk_queue_pending,
 488                                 blkreq_idx, xsegbd_dev->src_portno) == Noneidx);
 489         spin_lock_irq(&xsegbd_dev->rqlock);
 490         __xsegbd_put(xsegbd_dev);
 491 }
 492
 493 int update_dev_sectors_from_request(    struct xsegbd_device *xsegbd_dev,
 494                                         struct xseg_request *xreq       )
 495 {
 496         void *data;
 497         if (!xreq) {
 498                 XSEGLOG("Invalid xreq");
 499                 return -EIO;
 500         }
 501
 502         if (xreq->state & XS_FAILED)
 503                 return -ENOENT;
 504
 505         if (!(xreq->state & XS_SERVED))
 506                 return -EIO;
 507
 508         data = xseg_get_data(xsegbd_dev->xseg, xreq);
 509         if (!data) {
 510                 XSEGLOG("Invalid req data");
 511                 return -EIO;
 512         }
 513         if (!xsegbd_dev) {
 514                 XSEGLOG("Invalid xsegbd_dev");
 515                 return -ENOENT;
 516         }
 517         xsegbd_dev->sectors = *((uint64_t *) data) / 512ULL;
 518         return 0;
 519 }
 520
 521 static int xsegbd_get_size(struct xsegbd_device *xsegbd_dev)
 522 {
 523         struct xseg_request *xreq;
 524         char *target;
 525         uint64_t datalen;
 526         xqindex blkreq_idx;
 527         struct xsegbd_pending *pending;
 528         struct completion comp;
 529         xport p;
 530         void *data;
 531         int ret = -EBUSY, r;
 532
 533         __xsegbd_get(xsegbd_dev);
 534
 535         xreq = xseg_get_request(xsegbd_dev->xseg, xsegbd_dev->src_portno,
 536                         xsegbd_dev->dst_portno, X_ALLOC);
 537         if (!xreq)
 538                 goto out;
 539
 540         BUG_ON(xseg_prep_request(xsegbd_dev->xseg, xreq, xsegbd_dev->targetlen,
 541                                 sizeof(struct xseg_reply_info)));
 542
 543         init_completion(&comp);
 544         blkreq_idx = xq_pop_head(&xsegbd_dev->blk_queue_pending, 1);
 545         if (blkreq_idx == Noneidx)
 546                 goto out_put;
 547
 548         pending = &xsegbd_dev->blk_req_pending[blkreq_idx];
 549         pending->dev = xsegbd_dev;
 550         pending->request = NULL;
 551         pending->comp = &comp;
 552
 553
 554         xreq->priv = (uint64_t) blkreq_idx;
 555
 556         target = xseg_get_target(xsegbd_dev->xseg, xreq);
 557         strncpy(target, xsegbd_dev->target, xsegbd_dev->targetlen);
 558         xreq->size = xreq->datalen;
 559         xreq->offset = 0;
 560         xreq->op = X_INFO;
 561
 562         xseg_prepare_wait(xsegbd_dev->xseg, xsegbd_dev->src_portno);
 563         p = xseg_submit(xsegbd_dev->xseg, xreq,
 564                                 xsegbd_dev->src_portno, X_ALLOC);
 565         if ( p == NoPort) {
 566                 XSEGLOG("couldn't submit request");
 567                 BUG_ON(1);
 568                 goto out_queue;
 569         }
 570         WARN_ON(xseg_signal(xsegbd_dev->xseg, p) < 0);
 571         XSEGLOG("Before wait for completion, comp %lx [%llu]", (unsigned long) pending->comp, (unsigned long long) blkreq_idx);
 572         wait_for_completion_interruptible(&comp);
 573         XSEGLOG("Woken up after wait_for_completion_interruptible(), comp: %lx [%llu]", (unsigned long) pending->comp, (unsigned long long) blkreq_idx);
 574         ret = update_dev_sectors_from_request(xsegbd_dev, xreq);
 575         //XSEGLOG("get_size: sectors = %ld\n", (long)xsegbd_dev->sectors);
 576 out_put:
 577         BUG_ON(xseg_put_request(xsegbd_dev->xseg, xreq, xsegbd_dev->src_portno) == -1);
 578 out:
 579         __xsegbd_put(xsegbd_dev);
 580         return ret;
 581
 582 out_queue:
 583         pending->dev = NULL;
 584         pending->comp = NULL;
 585         xq_append_head(&xsegbd_dev->blk_queue_pending, blkreq_idx, 1);
 586
 587         goto out;
 588 }
 589
 590 static int xsegbd_mapclose(struct xsegbd_device *xsegbd_dev)
 591 {
 592         struct xseg_request *xreq;
 593         char *target;
 594         uint64_t datalen;
 595         xqindex blkreq_idx;
 596         struct xsegbd_pending *pending;
 597         struct completion comp;
 598         xport p;
 599         void *data;
 600         int ret = -EBUSY, r;
 601
 602         __xsegbd_get(xsegbd_dev);
 603         xreq = xseg_get_request(xsegbd_dev->xseg, xsegbd_dev->src_portno,
 604                         xsegbd_dev->dst_portno, X_ALLOC);
 605         if (!xreq)
 606                 goto out;
 607
 608         BUG_ON(xseg_prep_request(xsegbd_dev->xseg, xreq, xsegbd_dev->targetlen, 0));
 609
 610         init_completion(&comp);
 611         blkreq_idx = xq_pop_head(&xsegbd_dev->blk_queue_pending, 1);
 612         if (blkreq_idx == Noneidx)
 613                 goto out_put;
 614
 615         pending = &xsegbd_dev->blk_req_pending[blkreq_idx];
 616         pending->dev = xsegbd_dev;
 617         pending->request = NULL;
 618         pending->comp = &comp;
 619
 620
 621         xreq->priv = (uint64_t) blkreq_idx;
 622
 623         target = xseg_get_target(xsegbd_dev->xseg, xreq);
 624         strncpy(target, xsegbd_dev->target, xsegbd_dev->targetlen);
 625         xreq->size = xreq->datalen;
 626         xreq->offset = 0;
 627         xreq->op = X_CLOSE;
 628
 629         xseg_prepare_wait(xsegbd_dev->xseg, xsegbd_dev->src_portno);
 630         p = xseg_submit(xsegbd_dev->xseg, xreq,
 631                                 xsegbd_dev->src_portno, X_ALLOC);
 632         if ( p == NoPort) {
 633                 XSEGLOG("couldn't submit request");
 634                 BUG_ON(1);
 635                 goto out_queue;
 636         }
 637         WARN_ON(xseg_signal(xsegbd_dev->xseg, p) < 0);
 638         wait_for_completion_interruptible(&comp);
 639         ret = 0;
 640         if (xreq->state & XS_FAILED)
 641                 XSEGLOG("Couldn't close disk on mapper");
 642 out_put:
 643         BUG_ON(xseg_put_request(xsegbd_dev->xseg, xreq, xsegbd_dev->src_portno) == -1);
 644 out:
 645         __xsegbd_put(xsegbd_dev);
 646         return ret;
 647
 648 out_queue:
 649         pending->dev = NULL;
 650         pending->comp = NULL;
 651         xq_append_head(&xsegbd_dev->blk_queue_pending, blkreq_idx, 1);
 652
 653         goto out;
 654 }
 655
 656 static void xseg_callback(xport portno)
 657 {
 658         struct xsegbd_device *xsegbd_dev;
 659         struct xseg_request *xreq;
 660         struct request *blkreq;
 661         struct xsegbd_pending *pending;
 662         unsigned long flags;
 663         xqindex blkreq_idx, ridx;
 664         int err;
 665         void *data;
 666
 667         xsegbd_dev  = __xsegbd_get_dev(portno);
 668         if (!xsegbd_dev) {
 669                 XSEGLOG("portno: %u has no xsegbd device assigned", portno);
 670                 WARN_ON(1);
 671                 return;
 672         }
 673
 674         for (;;) {
 675                 xseg_prepare_wait(xsegbd_dev->xseg, xsegbd_dev->src_portno);
 676                 xreq = xseg_receive(xsegbd_dev->xseg, portno);
 677                 if (!xreq)
 678                         break;
 679
 680                 xseg_cancel_wait(xsegbd_dev->xseg, xsegbd_dev->src_portno);
 681
 682                 blkreq_idx = (xqindex) xreq->priv;
 683                 if (blkreq_idx >= xsegbd_dev->nr_requests) {
 684                         WARN_ON(1);
 685                         //FIXME maybe put request?
 686                         continue;
 687                 }
 688
 689                 pending = &xsegbd_dev->blk_req_pending[blkreq_idx];
 690                 if (pending->comp) {
 691                         /* someone is blocking on this request
 692                            and will handle it when we wake them up. */
 693                         complete(pending->comp);
 694                         /* the request is blocker's responsibility so
 695                            we will not put_request(); */
 696
 697                         continue;
 698                 }
 699
 700                 /* this is now treated as a block I/O request to end */
 701                 blkreq = pending->request;
 702                 pending->request = NULL;
 703                 if (xsegbd_dev != pending->dev) {
 704                         //FIXME maybe put request?
 705                         XSEGLOG("xsegbd_dev != pending->dev");
 706                         BUG_ON(1);
 707                         continue;
 708                 }
 709                 pending->dev = NULL;
 710                 if (!blkreq){
 711                         //FIXME maybe put request?
 712                         XSEGLOG("blkreq does not exist");
 713                         BUG_ON(1);
 714                         continue;
 715                 }
 716
 717                 err = -EIO;
 718                 if (!(xreq->state & XS_SERVED))
 719                         goto blk_end;
 720
 721                 if (xreq->serviced != blk_rq_bytes(blkreq))
 722                         goto blk_end;
 723
 724                 err = 0;
 725                 /* unlock for data transfer? */
 726                 if (!rq_data_dir(blkreq)){
 727                         xseg_to_blk(xsegbd_dev->xseg, xreq, blkreq);
 728                 }
 729 blk_end:
 730                 blk_end_request_all(blkreq, err);
 731
 732                 ridx = xq_append_head(&xsegbd_dev->blk_queue_pending,
 733                                         blkreq_idx, xsegbd_dev->src_portno);
 734                 if (ridx == Noneidx) {
 735                         XSEGLOG("couldnt append blkreq_idx");
 736                         WARN_ON(1);
 737                 }
 738
 739                 if (xseg_put_request(xsegbd_dev->xseg, xreq,
 740                                                 xsegbd_dev->src_portno) < 0){
 741                         XSEGLOG("couldn't put req");
 742                         BUG_ON(1);
 743                 }
 744         }
 745         if (xsegbd_dev) {
 746                 spin_lock_irqsave(&xsegbd_dev->rqlock, flags);
 747                 xseg_request_fn(xsegbd_dev->blk_queue);
 748                 spin_unlock_irqrestore(&xsegbd_dev->rqlock, flags);
 749                 __xsegbd_put(xsegbd_dev);
 750         }
 751 }
 752
 753
 754 /* sysfs interface */
 755
 756 static struct bus_type xsegbd_bus_type = {
 757         .name   = "xsegbd",
 758 };
 759
 760 static ssize_t xsegbd_size_show(struct device *dev,
 761                                         struct device_attribute *attr, char *buf)
 762 {
 763         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 764
 765         return sprintf(buf, "%llu\n", (unsigned long long) xsegbd_dev->sectors * 512ULL);
 766 }
 767
 768 static ssize_t xsegbd_major_show(struct device *dev,
 769                                         struct device_attribute *attr, char *buf)
 770 {
 771         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 772
 773         return sprintf(buf, "%d\n", xsegbd_dev->major);
 774 }
 775
 776 static ssize_t xsegbd_srcport_show(struct device *dev,
 777                                         struct device_attribute *attr, char *buf)
 778 {
 779         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 780
 781         return sprintf(buf, "%u\n", (unsigned) xsegbd_dev->src_portno);
 782 }
 783
 784 static ssize_t xsegbd_dstport_show(struct device *dev,
 785                                         struct device_attribute *attr, char *buf)
 786 {
 787         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 788
 789         return sprintf(buf, "%u\n", (unsigned) xsegbd_dev->dst_portno);
 790 }
 791
 792 static ssize_t xsegbd_id_show(struct device *dev,
 793                                         struct device_attribute *attr, char *buf)
 794 {
 795         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 796
 797         return sprintf(buf, "%u\n", (unsigned) xsegbd_dev->id);
 798 }
 799
 800 static ssize_t xsegbd_reqs_show(struct device *dev,
 801                                         struct device_attribute *attr, char *buf)
 802 {
 803         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 804
 805         return sprintf(buf, "%u\n", (unsigned) xsegbd_dev->nr_requests);
 806 }
 807
 808 static ssize_t xsegbd_target_show(struct device *dev,
 809                                         struct device_attribute *attr, char *buf)
 810 {
 811         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 812
 813         return sprintf(buf, "%s\n", xsegbd_dev->target);
 814 }
 815
 816 static ssize_t xsegbd_image_refresh(struct device *dev,
 817                                         struct device_attribute *attr,
 818                                         const char *buf,
 819                                         size_t size)
 820 {
 821         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 822         int rc, ret = size;
 823
 824         mutex_lock_nested(&xsegbd_mutex, SINGLE_DEPTH_NESTING);
 825
 826         rc = xsegbd_get_size(xsegbd_dev);
 827         if (rc < 0) {
 828                 ret = rc;
 829                 goto out;
 830         }
 831
 832         set_capacity(xsegbd_dev->gd, xsegbd_dev->sectors);
 833
 834 out:
 835         mutex_unlock(&xsegbd_mutex);
 836         return ret;
 837 }
 838
 839 static ssize_t xsegbd_cleanup(struct device *dev,
 840                                         struct device_attribute *attr,
 841                                         const char *buf,
 842                                         size_t size)
 843 {
 844         struct xsegbd_device *xsegbd_dev = dev_to_xsegbd(dev);
 845         int ret = size, i;
 846         struct request *blkreq = NULL;
 847         struct xsegbd_pending *pending = NULL;
 848         struct completion *comp = NULL;
 849
 850         mutex_lock_nested(&xsegbd_mutex, SINGLE_DEPTH_NESTING);
 851         xlock_acquire(&xsegbd_dev->blk_queue_pending.lock,
 852                                 xsegbd_dev->src_portno);
 853         for (i = 0; i < xsegbd_dev->nr_requests; i++) {
 854                 if (!__xq_check(&xsegbd_dev->blk_queue_pending, i)) {
 855                         pending = &xsegbd_dev->blk_req_pending[i];
 856                         blkreq = pending->request;
 857                         pending->request = NULL;
 858                         comp = pending->comp;
 859                         pending->comp = NULL;
 860                         if (blkreq){
 861                                 XSEGLOG("Cleaning up blkreq %lx [%d]", (unsigned long) blkreq, i);
 862                                 blk_end_request_all(blkreq, -EIO);
 863                         }
 864                         if (comp){
 865                                 XSEGLOG("Cleaning up comp %lx [%d]", (unsigned long) comp, i);
 866                                 complete(comp);
 867                         }
 868                         __xq_append_tail(&xsegbd_dev->blk_queue_pending, i);
 869                 }
 870         }
 871         xlock_release(&xsegbd_dev->blk_queue_pending.lock);
 872
 873         mutex_unlock(&xsegbd_mutex);
 874         return ret;
 875 }
 876
 877 static DEVICE_ATTR(size, S_IRUGO, xsegbd_size_show, NULL);
 878 static DEVICE_ATTR(major, S_IRUGO, xsegbd_major_show, NULL);
 879 static DEVICE_ATTR(srcport, S_IRUGO, xsegbd_srcport_show, NULL);
 880 static DEVICE_ATTR(dstport, S_IRUGO, xsegbd_dstport_show, NULL);
 881 static DEVICE_ATTR(id , S_IRUGO, xsegbd_id_show, NULL);
 882 static DEVICE_ATTR(reqs , S_IRUGO, xsegbd_reqs_show, NULL);
 883 static DEVICE_ATTR(target, S_IRUGO, xsegbd_target_show, NULL);
 884 static DEVICE_ATTR(refresh , S_IWUSR, NULL, xsegbd_image_refresh);
 885 static DEVICE_ATTR(cleanup , S_IWUSR, NULL, xsegbd_cleanup);
 886
 887 static struct attribute *xsegbd_attrs[] = {
 888         &dev_attr_size.attr,
 889         &dev_attr_major.attr,
 890         &dev_attr_srcport.attr,
 891         &dev_attr_dstport.attr,
 892         &dev_attr_id.attr,
 893         &dev_attr_reqs.attr,
 894         &dev_attr_target.attr,
 895         &dev_attr_refresh.attr,
 896         &dev_attr_cleanup.attr,
 897         NULL
 898 };
 899
 900 static struct attribute_group xsegbd_attr_group = {
 901         .attrs = xsegbd_attrs,
 902 };
 903
 904 static const struct attribute_group *xsegbd_attr_groups[] = {
 905         &xsegbd_attr_group,
 906         NULL
 907 };
 908
 909 static void xsegbd_sysfs_dev_release(struct device *dev)
 910 {
 911 }
 912
 913 static struct device_type xsegbd_device_type = {
 914         .name           = "xsegbd",
 915         .groups         = xsegbd_attr_groups,
 916         .release        = xsegbd_sysfs_dev_release,
 917 };
 918
 919 static void xsegbd_root_dev_release(struct device *dev)
 920 {
 921 }
 922
 923 static struct device xsegbd_root_dev = {
 924         .init_name      = "xsegbd",
 925         .release        = xsegbd_root_dev_release,
 926 };
 927
 928 static int xsegbd_bus_add_dev(struct xsegbd_device *xsegbd_dev)
 929 {
 930         int ret = -ENOMEM;
 931         struct device *dev;
 932
 933         mutex_lock_nested(&xsegbd_mutex, SINGLE_DEPTH_NESTING);
 934         dev = &xsegbd_dev->dev;
 935
 936         dev->bus = &xsegbd_bus_type;
 937         dev->type = &xsegbd_device_type;
 938         dev->parent = &xsegbd_root_dev;
 939         dev->release = xsegbd_dev_release;
 940         dev_set_name(dev, "%d", xsegbd_dev->id);
 941
 942         ret = device_register(dev);
 943
 944         mutex_unlock(&xsegbd_mutex);
 945         return ret;
 946 }
 947
 948 static void xsegbd_bus_del_dev(struct xsegbd_device *xsegbd_dev)
 949 {
 950         device_unregister(&xsegbd_dev->dev);
 951 }
 952
 953 static ssize_t xsegbd_add(struct bus_type *bus, const char *buf, size_t count)
 954 {
 955         struct xsegbd_device *xsegbd_dev;
 956         struct xseg_port *port;
 957         ssize_t ret = -ENOMEM;
 958
 959         if (!try_module_get(THIS_MODULE))
 960                 return -ENODEV;
 961
 962         xsegbd_dev = kzalloc(sizeof(*xsegbd_dev), GFP_KERNEL);
 963         if (!xsegbd_dev)
 964                 goto out;
 965
 966         spin_lock_init(&xsegbd_dev->rqlock);
 967         INIT_LIST_HEAD(&xsegbd_dev->node);
 968         init_waitqueue_head(&xsegbd_dev->wq);
 969         atomic_set(&xsegbd_dev->usercount, 0);
 970
 971         /* parse cmd */
 972         if (sscanf(buf, "%" __stringify(XSEGBD_TARGET_NAMELEN) "s "
 973                         "%d:%d:%d", xsegbd_dev->target, &xsegbd_dev->src_portno,
 974                         &xsegbd_dev->dst_portno, &xsegbd_dev->nr_requests) < 3) {
 975                 ret = -EINVAL;
 976                 goto out_dev;
 977         }
 978         xsegbd_dev->targetlen = strlen(xsegbd_dev->target);
 979
 980         spin_lock(&xsegbd_devices_lock);
 981         if (xsegbd_devices[xsegbd_dev->src_portno] != NULL) {
 982                 ret = -EINVAL;
 983                 goto out_unlock;
 984         }
 985         xsegbd_devices[xsegbd_dev->src_portno] = xsegbd_dev;
 986         xsegbd_dev->id = xsegbd_dev->src_portno;
 987         spin_unlock(&xsegbd_devices_lock);
 988
 989         XSEGLOG("registering block device major %d", major);
 990         ret = register_blkdev(major, XSEGBD_NAME);
 991         if (ret < 0) {
 992                 XSEGLOG("cannot register block device!");
 993                 ret = -EBUSY;
 994                 goto out_delentry;
 995         }
 996         xsegbd_dev->major = ret;
 997         XSEGLOG("registered block device major %d", xsegbd_dev->major);
 998
 999         ret = xsegbd_bus_add_dev(xsegbd_dev);
1000         if (ret)
1001                 goto out_blkdev;
1002
1003         if (!xq_alloc_seq(&xsegbd_dev->blk_queue_pending,
1004                                 xsegbd_dev->nr_requests,
1005                                 xsegbd_dev->nr_requests))
1006                 goto out_bus;
1007
1008         xsegbd_dev->blk_req_pending = kzalloc(
1009                         xsegbd_dev->nr_requests *sizeof(struct xsegbd_pending),
1010                                    GFP_KERNEL);
1011         if (!xsegbd_dev->blk_req_pending)
1012                 goto out_freeq;
1013
1014
1015         XSEGLOG("joining segment");
1016         //FIXME use xsebd module config for now
1017         xsegbd_dev->xseg = xseg_join(   xsegbd.config.type,
1018                                         xsegbd.config.name,
1019                                         "segdev",
1020                                         xseg_callback           );
1021         if (!xsegbd_dev->xseg)
1022                 goto out_freepending;
1023
1024
1025         XSEGLOG("binding to source port %u (destination %u)",
1026                         xsegbd_dev->src_portno, xsegbd_dev->dst_portno);
1027         port = xseg_bind_port(xsegbd_dev->xseg, xsegbd_dev->src_portno, NULL);
1028         if (!port) {
1029                 XSEGLOG("cannot bind to port");
1030                 ret = -EFAULT;
1031
1032                 goto out_xseg;
1033         }
1034
1035         if (xsegbd_dev->src_portno != xseg_portno(xsegbd_dev->xseg, port)) {
1036                 XSEGLOG("portno != xsegbd_dev->src_portno");
1037                 BUG_ON(1);
1038                 ret = -EFAULT;
1039                 goto out_xseg;
1040         }
1041
1042
1043         /* make sure we don't get any requests until we're ready to handle them */
1044         xseg_cancel_wait(xsegbd_dev->xseg, xseg_portno(xsegbd_dev->xseg, port));
1045
1046         ret = xsegbd_dev_init(xsegbd_dev);
1047         if (ret)
1048                 goto out_xseg;
1049
1050         xseg_prepare_wait(xsegbd_dev->xseg, xseg_portno(xsegbd_dev->xseg, port));
1051         return count;
1052
1053 out_xseg:
1054         xseg_leave(xsegbd_dev->xseg);
1055
1056 out_freepending:
1057         kfree(xsegbd_dev->blk_req_pending);
1058
1059 out_freeq:
1060         xq_free(&xsegbd_dev->blk_queue_pending);
1061
1062 out_bus:
1063         xsegbd_bus_del_dev(xsegbd_dev);
1064         return ret;
1065
1066 out_blkdev:
1067         unregister_blkdev(xsegbd_dev->major, XSEGBD_NAME);
1068
1069 out_delentry:
1070         spin_lock(&xsegbd_devices_lock);
1071         xsegbd_devices[xsegbd_dev->src_portno] = NULL;
1072
1073 out_unlock:
1074         spin_unlock(&xsegbd_devices_lock);
1075
1076 out_dev:
1077         kfree(xsegbd_dev);
1078
1079 out:
1080         return ret;
1081 }
1082
1083 static ssize_t xsegbd_remove(struct bus_type *bus, const char *buf, size_t count)
1084 {
1085         struct xsegbd_device *xsegbd_dev = NULL;
1086         int id, ret;
1087         unsigned long ul_id;
1088
1089         ret = strict_strtoul(buf, 10, &ul_id);
1090         if (ret)
1091                 return ret;
1092
1093         id = (int) ul_id;
1094         if (id != ul_id)
1095                 return -EINVAL;
1096
1097         mutex_lock_nested(&xsegbd_mutex, SINGLE_DEPTH_NESTING);
1098
1099         ret = count;
1100         //FIXME when to put dev?
1101         xsegbd_dev = __xsegbd_get_dev(id);
1102         if (!xsegbd_dev) {
1103                 ret = -ENOENT;
1104                 goto out_unlock;
1105         }
1106         xsegbd_bus_del_dev(xsegbd_dev);
1107
1108 out_unlock:
1109         mutex_unlock(&xsegbd_mutex);
1110         return ret;
1111 }
1112
1113 static struct bus_attribute xsegbd_bus_attrs[] = {
1114         __ATTR(add, S_IWUSR, NULL, xsegbd_add),
1115         __ATTR(remove, S_IWUSR, NULL, xsegbd_remove),
1116         __ATTR_NULL
1117 };
1118
1119 static int xsegbd_sysfs_init(void)
1120 {
1121         int ret;
1122
1123         ret = device_register(&xsegbd_root_dev);
1124         if (ret < 0)
1125                 return ret;
1126
1127         xsegbd_bus_type.bus_attrs = xsegbd_bus_attrs;
1128         ret = bus_register(&xsegbd_bus_type);
1129         if (ret < 0)
1130                 device_unregister(&xsegbd_root_dev);
1131
1132         return ret;
1133 }
1134
1135 static void xsegbd_sysfs_cleanup(void)
1136 {
1137         bus_unregister(&xsegbd_bus_type);
1138         device_unregister(&xsegbd_root_dev);
1139 }
1140
1141 /* *************************** */
1142 /* ** Module Initialization ** */
1143 /* *************************** */
1144
1145 static int __init xsegbd_init(void)
1146 {
1147         int ret = -ENOMEM;
1148         xsegbd_devices = kzalloc(max_dev * sizeof(struct xsegbd_devices *), GFP_KERNEL);
1149         if (!xsegbd_devices)
1150                 goto out;
1151
1152         spin_lock_init(&xsegbd_devices_lock);
1153
1154         ret = -ENOSYS;
1155         ret = xsegbd_xseg_init();
1156         if (ret)
1157                 goto out_free;
1158
1159         ret = xsegbd_sysfs_init();
1160         if (ret)
1161                 goto out_xseg;
1162
1163         XSEGLOG("initialization complete");
1164
1165 out:
1166         return ret;
1167
1168 out_xseg:
1169         xsegbd_xseg_quit();
1170
1171 out_free:
1172         kfree(xsegbd_devices);
1173
1174         goto out;
1175 }
1176
1177 static void __exit xsegbd_exit(void)
1178 {
1179         xsegbd_sysfs_cleanup();
1180         xsegbd_xseg_quit();
1181 }
1182
1183 module_init(xsegbd_init);
1184 module_exit(xsegbd_exit);
1185