1 #include <linux/ceph/ceph_debug.h> 2 3 #include <linux/module.h> 4 #include <linux/sched.h> 5 #include <linux/slab.h> 6 #include <linux/file.h> 7 #include <linux/mount.h> 8 #include <linux/namei.h> 9 #include <linux/writeback.h> 10 #include <linux/falloc.h> 11 12 #include "super.h" 13 #include "mds_client.h" 14 #include "cache.h" 15 16 static __le32 ceph_flags_sys2wire(u32 flags) 17 { 18 u32 wire_flags = 0; 19 20 switch (flags & O_ACCMODE) { 21 case O_RDONLY: 22 wire_flags |= CEPH_O_RDONLY; 23 break; 24 case O_WRONLY: 25 wire_flags |= CEPH_O_WRONLY; 26 break; 27 case O_RDWR: 28 wire_flags |= CEPH_O_RDWR; 29 break; 30 } 31 32 #define ceph_sys2wire(a) if (flags & a) { wire_flags |= CEPH_##a; flags &= ~a; } 33 34 ceph_sys2wire(O_CREAT); 35 ceph_sys2wire(O_EXCL); 36 ceph_sys2wire(O_TRUNC); 37 ceph_sys2wire(O_DIRECTORY); 38 ceph_sys2wire(O_NOFOLLOW); 39 40 #undef ceph_sys2wire 41 42 if (flags) 43 dout("unused open flags: %x", flags); 44 45 return cpu_to_le32(wire_flags); 46 } 47 48 /* 49 * Ceph file operations 50 * 51 * Implement basic open/close functionality, and implement 52 * read/write. 53 * 54 * We implement three modes of file I/O: 55 * - buffered uses the generic_file_aio_{read,write} helpers 56 * 57 * - synchronous is used when there is multi-client read/write 58 * sharing, avoids the page cache, and synchronously waits for an 59 * ack from the OSD. 60 * 61 * - direct io takes the variant of the sync path that references 62 * user pages directly. 63 * 64 * fsync() flushes and waits on dirty pages, but just queues metadata 65 * for writeback: since the MDS can recover size and mtime there is no 66 * need to wait for MDS acknowledgement. 67 */ 68 69 /* 70 * Calculate the length sum of direct io vectors that can 71 * be combined into one page vector. 72 */ 73 static size_t dio_get_pagev_size(const struct iov_iter *it) 74 { 75 const struct iovec *iov = it->iov; 76 const struct iovec *iovend = iov + it->nr_segs; 77 size_t size; 78 79 size = iov->iov_len - it->iov_offset; 80 /* 81 * An iov can be page vectored when both the current tail 82 * and the next base are page aligned. 83 */ 84 while (PAGE_ALIGNED((iov->iov_base + iov->iov_len)) && 85 (++iov < iovend && PAGE_ALIGNED((iov->iov_base)))) { 86 size += iov->iov_len; 87 } 88 dout("dio_get_pagevlen len = %zu\n", size); 89 return size; 90 } 91 92 /* 93 * Allocate a page vector based on (@it, @nbytes). 94 * The return value is the tuple describing a page vector, 95 * that is (@pages, @page_align, @num_pages). 96 */ 97 static struct page ** 98 dio_get_pages_alloc(const struct iov_iter *it, size_t nbytes, 99 size_t *page_align, int *num_pages) 100 { 101 struct iov_iter tmp_it = *it; 102 size_t align; 103 struct page **pages; 104 int ret = 0, idx, npages; 105 106 align = (unsigned long)(it->iov->iov_base + it->iov_offset) & 107 (PAGE_SIZE - 1); 108 npages = calc_pages_for(align, nbytes); 109 pages = kvmalloc(sizeof(*pages) * npages, GFP_KERNEL); 110 if (!pages) 111 return ERR_PTR(-ENOMEM); 112 113 for (idx = 0; idx < npages; ) { 114 size_t start; 115 ret = iov_iter_get_pages(&tmp_it, pages + idx, nbytes, 116 npages - idx, &start); 117 if (ret < 0) 118 goto fail; 119 120 iov_iter_advance(&tmp_it, ret); 121 nbytes -= ret; 122 idx += (ret + start + PAGE_SIZE - 1) / PAGE_SIZE; 123 } 124 125 BUG_ON(nbytes != 0); 126 *num_pages = npages; 127 *page_align = align; 128 dout("dio_get_pages_alloc: got %d pages align %zu\n", npages, align); 129 return pages; 130 fail: 131 ceph_put_page_vector(pages, idx, false); 132 return ERR_PTR(ret); 133 } 134 135 /* 136 * Prepare an open request. Preallocate ceph_cap to avoid an 137 * inopportune ENOMEM later. 138 */ 139 static struct ceph_mds_request * 140 prepare_open_request(struct super_block *sb, int flags, int create_mode) 141 { 142 struct ceph_fs_client *fsc = ceph_sb_to_client(sb); 143 struct ceph_mds_client *mdsc = fsc->mdsc; 144 struct ceph_mds_request *req; 145 int want_auth = USE_ANY_MDS; 146 int op = (flags & O_CREAT) ? CEPH_MDS_OP_CREATE : CEPH_MDS_OP_OPEN; 147 148 if (flags & (O_WRONLY|O_RDWR|O_CREAT|O_TRUNC)) 149 want_auth = USE_AUTH_MDS; 150 151 req = ceph_mdsc_create_request(mdsc, op, want_auth); 152 if (IS_ERR(req)) 153 goto out; 154 req->r_fmode = ceph_flags_to_mode(flags); 155 req->r_args.open.flags = ceph_flags_sys2wire(flags); 156 req->r_args.open.mode = cpu_to_le32(create_mode); 157 out: 158 return req; 159 } 160 161 /* 162 * initialize private struct file data. 163 * if we fail, clean up by dropping fmode reference on the ceph_inode 164 */ 165 static int ceph_init_file(struct inode *inode, struct file *file, int fmode) 166 { 167 struct ceph_file_info *cf; 168 int ret = 0; 169 170 switch (inode->i_mode & S_IFMT) { 171 case S_IFREG: 172 ceph_fscache_register_inode_cookie(inode); 173 ceph_fscache_file_set_cookie(inode, file); 174 case S_IFDIR: 175 dout("init_file %p %p 0%o (regular)\n", inode, file, 176 inode->i_mode); 177 cf = kmem_cache_zalloc(ceph_file_cachep, GFP_KERNEL); 178 if (cf == NULL) { 179 ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */ 180 return -ENOMEM; 181 } 182 cf->fmode = fmode; 183 cf->next_offset = 2; 184 cf->readdir_cache_idx = -1; 185 file->private_data = cf; 186 BUG_ON(inode->i_fop->release != ceph_release); 187 break; 188 189 case S_IFLNK: 190 dout("init_file %p %p 0%o (symlink)\n", inode, file, 191 inode->i_mode); 192 ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */ 193 break; 194 195 default: 196 dout("init_file %p %p 0%o (special)\n", inode, file, 197 inode->i_mode); 198 /* 199 * we need to drop the open ref now, since we don't 200 * have .release set to ceph_release. 201 */ 202 ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */ 203 BUG_ON(inode->i_fop->release == ceph_release); 204 205 /* call the proper open fop */ 206 ret = inode->i_fop->open(inode, file); 207 } 208 return ret; 209 } 210 211 /* 212 * try renew caps after session gets killed. 213 */ 214 int ceph_renew_caps(struct inode *inode) 215 { 216 struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; 217 struct ceph_inode_info *ci = ceph_inode(inode); 218 struct ceph_mds_request *req; 219 int err, flags, wanted; 220 221 spin_lock(&ci->i_ceph_lock); 222 wanted = __ceph_caps_file_wanted(ci); 223 if (__ceph_is_any_real_caps(ci) && 224 (!(wanted & CEPH_CAP_ANY_WR) || ci->i_auth_cap)) { 225 int issued = __ceph_caps_issued(ci, NULL); 226 spin_unlock(&ci->i_ceph_lock); 227 dout("renew caps %p want %s issued %s updating mds_wanted\n", 228 inode, ceph_cap_string(wanted), ceph_cap_string(issued)); 229 ceph_check_caps(ci, 0, NULL); 230 return 0; 231 } 232 spin_unlock(&ci->i_ceph_lock); 233 234 flags = 0; 235 if ((wanted & CEPH_CAP_FILE_RD) && (wanted & CEPH_CAP_FILE_WR)) 236 flags = O_RDWR; 237 else if (wanted & CEPH_CAP_FILE_RD) 238 flags = O_RDONLY; 239 else if (wanted & CEPH_CAP_FILE_WR) 240 flags = O_WRONLY; 241 #ifdef O_LAZY 242 if (wanted & CEPH_CAP_FILE_LAZYIO) 243 flags |= O_LAZY; 244 #endif 245 246 req = prepare_open_request(inode->i_sb, flags, 0); 247 if (IS_ERR(req)) { 248 err = PTR_ERR(req); 249 goto out; 250 } 251 252 req->r_inode = inode; 253 ihold(inode); 254 req->r_num_caps = 1; 255 req->r_fmode = -1; 256 257 err = ceph_mdsc_do_request(mdsc, NULL, req); 258 ceph_mdsc_put_request(req); 259 out: 260 dout("renew caps %p open result=%d\n", inode, err); 261 return err < 0 ? err : 0; 262 } 263 264 /* 265 * If we already have the requisite capabilities, we can satisfy 266 * the open request locally (no need to request new caps from the 267 * MDS). We do, however, need to inform the MDS (asynchronously) 268 * if our wanted caps set expands. 269 */ 270 int ceph_open(struct inode *inode, struct file *file) 271 { 272 struct ceph_inode_info *ci = ceph_inode(inode); 273 struct ceph_fs_client *fsc = ceph_sb_to_client(inode->i_sb); 274 struct ceph_mds_client *mdsc = fsc->mdsc; 275 struct ceph_mds_request *req; 276 struct ceph_file_info *cf = file->private_data; 277 int err; 278 int flags, fmode, wanted; 279 280 if (cf) { 281 dout("open file %p is already opened\n", file); 282 return 0; 283 } 284 285 /* filter out O_CREAT|O_EXCL; vfs did that already. yuck. */ 286 flags = file->f_flags & ~(O_CREAT|O_EXCL); 287 if (S_ISDIR(inode->i_mode)) 288 flags = O_DIRECTORY; /* mds likes to know */ 289 290 dout("open inode %p ino %llx.%llx file %p flags %d (%d)\n", inode, 291 ceph_vinop(inode), file, flags, file->f_flags); 292 fmode = ceph_flags_to_mode(flags); 293 wanted = ceph_caps_for_mode(fmode); 294 295 /* snapped files are read-only */ 296 if (ceph_snap(inode) != CEPH_NOSNAP && (file->f_mode & FMODE_WRITE)) 297 return -EROFS; 298 299 /* trivially open snapdir */ 300 if (ceph_snap(inode) == CEPH_SNAPDIR) { 301 spin_lock(&ci->i_ceph_lock); 302 __ceph_get_fmode(ci, fmode); 303 spin_unlock(&ci->i_ceph_lock); 304 return ceph_init_file(inode, file, fmode); 305 } 306 307 /* 308 * No need to block if we have caps on the auth MDS (for 309 * write) or any MDS (for read). Update wanted set 310 * asynchronously. 311 */ 312 spin_lock(&ci->i_ceph_lock); 313 if (__ceph_is_any_real_caps(ci) && 314 (((fmode & CEPH_FILE_MODE_WR) == 0) || ci->i_auth_cap)) { 315 int mds_wanted = __ceph_caps_mds_wanted(ci, true); 316 int issued = __ceph_caps_issued(ci, NULL); 317 318 dout("open %p fmode %d want %s issued %s using existing\n", 319 inode, fmode, ceph_cap_string(wanted), 320 ceph_cap_string(issued)); 321 __ceph_get_fmode(ci, fmode); 322 spin_unlock(&ci->i_ceph_lock); 323 324 /* adjust wanted? */ 325 if ((issued & wanted) != wanted && 326 (mds_wanted & wanted) != wanted && 327 ceph_snap(inode) != CEPH_SNAPDIR) 328 ceph_check_caps(ci, 0, NULL); 329 330 return ceph_init_file(inode, file, fmode); 331 } else if (ceph_snap(inode) != CEPH_NOSNAP && 332 (ci->i_snap_caps & wanted) == wanted) { 333 __ceph_get_fmode(ci, fmode); 334 spin_unlock(&ci->i_ceph_lock); 335 return ceph_init_file(inode, file, fmode); 336 } 337 338 spin_unlock(&ci->i_ceph_lock); 339 340 dout("open fmode %d wants %s\n", fmode, ceph_cap_string(wanted)); 341 req = prepare_open_request(inode->i_sb, flags, 0); 342 if (IS_ERR(req)) { 343 err = PTR_ERR(req); 344 goto out; 345 } 346 req->r_inode = inode; 347 ihold(inode); 348 349 req->r_num_caps = 1; 350 err = ceph_mdsc_do_request(mdsc, NULL, req); 351 if (!err) 352 err = ceph_init_file(inode, file, req->r_fmode); 353 ceph_mdsc_put_request(req); 354 dout("open result=%d on %llx.%llx\n", err, ceph_vinop(inode)); 355 out: 356 return err; 357 } 358 359 360 /* 361 * Do a lookup + open with a single request. If we get a non-existent 362 * file or symlink, return 1 so the VFS can retry. 363 */ 364 int ceph_atomic_open(struct inode *dir, struct dentry *dentry, 365 struct file *file, unsigned flags, umode_t mode, 366 int *opened) 367 { 368 struct ceph_fs_client *fsc = ceph_sb_to_client(dir->i_sb); 369 struct ceph_mds_client *mdsc = fsc->mdsc; 370 struct ceph_mds_request *req; 371 struct dentry *dn; 372 struct ceph_acls_info acls = {}; 373 int mask; 374 int err; 375 376 dout("atomic_open %p dentry %p '%pd' %s flags %d mode 0%o\n", 377 dir, dentry, dentry, 378 d_unhashed(dentry) ? "unhashed" : "hashed", flags, mode); 379 380 if (dentry->d_name.len > NAME_MAX) 381 return -ENAMETOOLONG; 382 383 if (flags & O_CREAT) { 384 err = ceph_pre_init_acls(dir, &mode, &acls); 385 if (err < 0) 386 return err; 387 } 388 389 /* do the open */ 390 req = prepare_open_request(dir->i_sb, flags, mode); 391 if (IS_ERR(req)) { 392 err = PTR_ERR(req); 393 goto out_acl; 394 } 395 req->r_dentry = dget(dentry); 396 req->r_num_caps = 2; 397 if (flags & O_CREAT) { 398 req->r_dentry_drop = CEPH_CAP_FILE_SHARED; 399 req->r_dentry_unless = CEPH_CAP_FILE_EXCL; 400 if (acls.pagelist) { 401 req->r_pagelist = acls.pagelist; 402 acls.pagelist = NULL; 403 } 404 } 405 406 mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED; 407 if (ceph_security_xattr_wanted(dir)) 408 mask |= CEPH_CAP_XATTR_SHARED; 409 req->r_args.open.mask = cpu_to_le32(mask); 410 411 req->r_parent = dir; 412 set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags); 413 err = ceph_mdsc_do_request(mdsc, 414 (flags & (O_CREAT|O_TRUNC)) ? dir : NULL, 415 req); 416 err = ceph_handle_snapdir(req, dentry, err); 417 if (err) 418 goto out_req; 419 420 if ((flags & O_CREAT) && !req->r_reply_info.head->is_dentry) 421 err = ceph_handle_notrace_create(dir, dentry); 422 423 if (d_in_lookup(dentry)) { 424 dn = ceph_finish_lookup(req, dentry, err); 425 if (IS_ERR(dn)) 426 err = PTR_ERR(dn); 427 } else { 428 /* we were given a hashed negative dentry */ 429 dn = NULL; 430 } 431 if (err) 432 goto out_req; 433 if (dn || d_really_is_negative(dentry) || d_is_symlink(dentry)) { 434 /* make vfs retry on splice, ENOENT, or symlink */ 435 dout("atomic_open finish_no_open on dn %p\n", dn); 436 err = finish_no_open(file, dn); 437 } else { 438 dout("atomic_open finish_open on dn %p\n", dn); 439 if (req->r_op == CEPH_MDS_OP_CREATE && req->r_reply_info.has_create_ino) { 440 ceph_init_inode_acls(d_inode(dentry), &acls); 441 *opened |= FILE_CREATED; 442 } 443 err = finish_open(file, dentry, ceph_open, opened); 444 } 445 out_req: 446 if (!req->r_err && req->r_target_inode) 447 ceph_put_fmode(ceph_inode(req->r_target_inode), req->r_fmode); 448 ceph_mdsc_put_request(req); 449 out_acl: 450 ceph_release_acls_info(&acls); 451 dout("atomic_open result=%d\n", err); 452 return err; 453 } 454 455 int ceph_release(struct inode *inode, struct file *file) 456 { 457 struct ceph_inode_info *ci = ceph_inode(inode); 458 struct ceph_file_info *cf = file->private_data; 459 460 dout("release inode %p file %p\n", inode, file); 461 ceph_put_fmode(ci, cf->fmode); 462 if (cf->last_readdir) 463 ceph_mdsc_put_request(cf->last_readdir); 464 kfree(cf->last_name); 465 kfree(cf->dir_info); 466 kmem_cache_free(ceph_file_cachep, cf); 467 468 /* wake up anyone waiting for caps on this inode */ 469 wake_up_all(&ci->i_cap_wq); 470 return 0; 471 } 472 473 enum { 474 HAVE_RETRIED = 1, 475 CHECK_EOF = 2, 476 READ_INLINE = 3, 477 }; 478 479 /* 480 * Read a range of bytes striped over one or more objects. Iterate over 481 * objects we stripe over. (That's not atomic, but good enough for now.) 482 * 483 * If we get a short result from the OSD, check against i_size; we need to 484 * only return a short read to the caller if we hit EOF. 485 */ 486 static int striped_read(struct inode *inode, 487 u64 pos, u64 len, 488 struct page **pages, int num_pages, 489 int page_align, int *checkeof) 490 { 491 struct ceph_fs_client *fsc = ceph_inode_to_client(inode); 492 struct ceph_inode_info *ci = ceph_inode(inode); 493 u64 this_len; 494 loff_t i_size; 495 int page_idx; 496 int ret, read = 0; 497 bool hit_stripe, was_short; 498 499 /* 500 * we may need to do multiple reads. not atomic, unfortunately. 501 */ 502 more: 503 this_len = len; 504 page_idx = (page_align + read) >> PAGE_SHIFT; 505 ret = ceph_osdc_readpages(&fsc->client->osdc, ceph_vino(inode), 506 &ci->i_layout, pos, &this_len, 507 ci->i_truncate_seq, ci->i_truncate_size, 508 pages + page_idx, num_pages - page_idx, 509 ((page_align + read) & ~PAGE_MASK)); 510 if (ret == -ENOENT) 511 ret = 0; 512 hit_stripe = this_len < len; 513 was_short = ret >= 0 && ret < this_len; 514 dout("striped_read %llu~%llu (read %u) got %d%s%s\n", pos, len, read, 515 ret, hit_stripe ? " HITSTRIPE" : "", was_short ? " SHORT" : ""); 516 517 i_size = i_size_read(inode); 518 if (ret >= 0) { 519 if (was_short && (pos + ret < i_size)) { 520 int zlen = min(this_len - ret, i_size - pos - ret); 521 int zoff = page_align + read + ret; 522 dout(" zero gap %llu to %llu\n", 523 pos + ret, pos + ret + zlen); 524 ceph_zero_page_vector_range(zoff, zlen, pages); 525 ret += zlen; 526 } 527 528 read += ret; 529 pos += ret; 530 len -= ret; 531 532 /* hit stripe and need continue*/ 533 if (len && hit_stripe && pos < i_size) 534 goto more; 535 } 536 537 if (read > 0) { 538 ret = read; 539 /* did we bounce off eof? */ 540 if (pos + len > i_size) 541 *checkeof = CHECK_EOF; 542 } 543 544 dout("striped_read returns %d\n", ret); 545 return ret; 546 } 547 548 /* 549 * Completely synchronous read and write methods. Direct from __user 550 * buffer to osd, or directly to user pages (if O_DIRECT). 551 * 552 * If the read spans object boundary, just do multiple reads. 553 */ 554 static ssize_t ceph_sync_read(struct kiocb *iocb, struct iov_iter *to, 555 int *checkeof) 556 { 557 struct file *file = iocb->ki_filp; 558 struct inode *inode = file_inode(file); 559 struct page **pages; 560 u64 off = iocb->ki_pos; 561 int num_pages; 562 ssize_t ret; 563 size_t len = iov_iter_count(to); 564 565 dout("sync_read on file %p %llu~%u %s\n", file, off, 566 (unsigned)len, 567 (file->f_flags & O_DIRECT) ? "O_DIRECT" : ""); 568 569 if (!len) 570 return 0; 571 /* 572 * flush any page cache pages in this range. this 573 * will make concurrent normal and sync io slow, 574 * but it will at least behave sensibly when they are 575 * in sequence. 576 */ 577 ret = filemap_write_and_wait_range(inode->i_mapping, off, 578 off + len); 579 if (ret < 0) 580 return ret; 581 582 if (unlikely(to->type & ITER_PIPE)) { 583 size_t page_off; 584 ret = iov_iter_get_pages_alloc(to, &pages, len, 585 &page_off); 586 if (ret <= 0) 587 return -ENOMEM; 588 num_pages = DIV_ROUND_UP(ret + page_off, PAGE_SIZE); 589 590 ret = striped_read(inode, off, ret, pages, num_pages, 591 page_off, checkeof); 592 if (ret > 0) { 593 iov_iter_advance(to, ret); 594 off += ret; 595 } else { 596 iov_iter_advance(to, 0); 597 } 598 ceph_put_page_vector(pages, num_pages, false); 599 } else { 600 num_pages = calc_pages_for(off, len); 601 pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); 602 if (IS_ERR(pages)) 603 return PTR_ERR(pages); 604 605 ret = striped_read(inode, off, len, pages, num_pages, 606 (off & ~PAGE_MASK), checkeof); 607 if (ret > 0) { 608 int l, k = 0; 609 size_t left = ret; 610 611 while (left) { 612 size_t page_off = off & ~PAGE_MASK; 613 size_t copy = min_t(size_t, left, 614 PAGE_SIZE - page_off); 615 l = copy_page_to_iter(pages[k++], page_off, 616 copy, to); 617 off += l; 618 left -= l; 619 if (l < copy) 620 break; 621 } 622 } 623 ceph_release_page_vector(pages, num_pages); 624 } 625 626 if (off > iocb->ki_pos) { 627 ret = off - iocb->ki_pos; 628 iocb->ki_pos = off; 629 } 630 631 dout("sync_read result %zd\n", ret); 632 return ret; 633 } 634 635 struct ceph_aio_request { 636 struct kiocb *iocb; 637 size_t total_len; 638 int write; 639 int error; 640 struct list_head osd_reqs; 641 unsigned num_reqs; 642 atomic_t pending_reqs; 643 struct timespec mtime; 644 struct ceph_cap_flush *prealloc_cf; 645 }; 646 647 struct ceph_aio_work { 648 struct work_struct work; 649 struct ceph_osd_request *req; 650 }; 651 652 static void ceph_aio_retry_work(struct work_struct *work); 653 654 static void ceph_aio_complete(struct inode *inode, 655 struct ceph_aio_request *aio_req) 656 { 657 struct ceph_inode_info *ci = ceph_inode(inode); 658 int ret; 659 660 if (!atomic_dec_and_test(&aio_req->pending_reqs)) 661 return; 662 663 ret = aio_req->error; 664 if (!ret) 665 ret = aio_req->total_len; 666 667 dout("ceph_aio_complete %p rc %d\n", inode, ret); 668 669 if (ret >= 0 && aio_req->write) { 670 int dirty; 671 672 loff_t endoff = aio_req->iocb->ki_pos + aio_req->total_len; 673 if (endoff > i_size_read(inode)) { 674 if (ceph_inode_set_size(inode, endoff)) 675 ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL); 676 } 677 678 spin_lock(&ci->i_ceph_lock); 679 ci->i_inline_version = CEPH_INLINE_NONE; 680 dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR, 681 &aio_req->prealloc_cf); 682 spin_unlock(&ci->i_ceph_lock); 683 if (dirty) 684 __mark_inode_dirty(inode, dirty); 685 686 } 687 688 ceph_put_cap_refs(ci, (aio_req->write ? CEPH_CAP_FILE_WR : 689 CEPH_CAP_FILE_RD)); 690 691 aio_req->iocb->ki_complete(aio_req->iocb, ret, 0); 692 693 ceph_free_cap_flush(aio_req->prealloc_cf); 694 kfree(aio_req); 695 } 696 697 static void ceph_aio_complete_req(struct ceph_osd_request *req) 698 { 699 int rc = req->r_result; 700 struct inode *inode = req->r_inode; 701 struct ceph_aio_request *aio_req = req->r_priv; 702 struct ceph_osd_data *osd_data = osd_req_op_extent_osd_data(req, 0); 703 int num_pages = calc_pages_for((u64)osd_data->alignment, 704 osd_data->length); 705 706 dout("ceph_aio_complete_req %p rc %d bytes %llu\n", 707 inode, rc, osd_data->length); 708 709 if (rc == -EOLDSNAPC) { 710 struct ceph_aio_work *aio_work; 711 BUG_ON(!aio_req->write); 712 713 aio_work = kmalloc(sizeof(*aio_work), GFP_NOFS); 714 if (aio_work) { 715 INIT_WORK(&aio_work->work, ceph_aio_retry_work); 716 aio_work->req = req; 717 queue_work(ceph_inode_to_client(inode)->wb_wq, 718 &aio_work->work); 719 return; 720 } 721 rc = -ENOMEM; 722 } else if (!aio_req->write) { 723 if (rc == -ENOENT) 724 rc = 0; 725 if (rc >= 0 && osd_data->length > rc) { 726 int zoff = osd_data->alignment + rc; 727 int zlen = osd_data->length - rc; 728 /* 729 * If read is satisfied by single OSD request, 730 * it can pass EOF. Otherwise read is within 731 * i_size. 732 */ 733 if (aio_req->num_reqs == 1) { 734 loff_t i_size = i_size_read(inode); 735 loff_t endoff = aio_req->iocb->ki_pos + rc; 736 if (endoff < i_size) 737 zlen = min_t(size_t, zlen, 738 i_size - endoff); 739 aio_req->total_len = rc + zlen; 740 } 741 742 if (zlen > 0) 743 ceph_zero_page_vector_range(zoff, zlen, 744 osd_data->pages); 745 } 746 } 747 748 ceph_put_page_vector(osd_data->pages, num_pages, !aio_req->write); 749 ceph_osdc_put_request(req); 750 751 if (rc < 0) 752 cmpxchg(&aio_req->error, 0, rc); 753 754 ceph_aio_complete(inode, aio_req); 755 return; 756 } 757 758 static void ceph_aio_retry_work(struct work_struct *work) 759 { 760 struct ceph_aio_work *aio_work = 761 container_of(work, struct ceph_aio_work, work); 762 struct ceph_osd_request *orig_req = aio_work->req; 763 struct ceph_aio_request *aio_req = orig_req->r_priv; 764 struct inode *inode = orig_req->r_inode; 765 struct ceph_inode_info *ci = ceph_inode(inode); 766 struct ceph_snap_context *snapc; 767 struct ceph_osd_request *req; 768 int ret; 769 770 spin_lock(&ci->i_ceph_lock); 771 if (__ceph_have_pending_cap_snap(ci)) { 772 struct ceph_cap_snap *capsnap = 773 list_last_entry(&ci->i_cap_snaps, 774 struct ceph_cap_snap, 775 ci_item); 776 snapc = ceph_get_snap_context(capsnap->context); 777 } else { 778 BUG_ON(!ci->i_head_snapc); 779 snapc = ceph_get_snap_context(ci->i_head_snapc); 780 } 781 spin_unlock(&ci->i_ceph_lock); 782 783 req = ceph_osdc_alloc_request(orig_req->r_osdc, snapc, 2, 784 false, GFP_NOFS); 785 if (!req) { 786 ret = -ENOMEM; 787 req = orig_req; 788 goto out; 789 } 790 791 req->r_flags = CEPH_OSD_FLAG_ORDERSNAP | CEPH_OSD_FLAG_WRITE; 792 ceph_oloc_copy(&req->r_base_oloc, &orig_req->r_base_oloc); 793 ceph_oid_copy(&req->r_base_oid, &orig_req->r_base_oid); 794 795 ret = ceph_osdc_alloc_messages(req, GFP_NOFS); 796 if (ret) { 797 ceph_osdc_put_request(req); 798 req = orig_req; 799 goto out; 800 } 801 802 req->r_ops[0] = orig_req->r_ops[0]; 803 osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC, 0); 804 805 req->r_mtime = aio_req->mtime; 806 req->r_data_offset = req->r_ops[0].extent.offset; 807 808 ceph_osdc_put_request(orig_req); 809 810 req->r_callback = ceph_aio_complete_req; 811 req->r_inode = inode; 812 req->r_priv = aio_req; 813 req->r_abort_on_full = true; 814 815 ret = ceph_osdc_start_request(req->r_osdc, req, false); 816 out: 817 if (ret < 0) { 818 req->r_result = ret; 819 ceph_aio_complete_req(req); 820 } 821 822 ceph_put_snap_context(snapc); 823 kfree(aio_work); 824 } 825 826 static ssize_t 827 ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, 828 struct ceph_snap_context *snapc, 829 struct ceph_cap_flush **pcf) 830 { 831 struct file *file = iocb->ki_filp; 832 struct inode *inode = file_inode(file); 833 struct ceph_inode_info *ci = ceph_inode(inode); 834 struct ceph_fs_client *fsc = ceph_inode_to_client(inode); 835 struct ceph_vino vino; 836 struct ceph_osd_request *req; 837 struct page **pages; 838 struct ceph_aio_request *aio_req = NULL; 839 int num_pages = 0; 840 int flags; 841 int ret; 842 struct timespec mtime = current_time(inode); 843 size_t count = iov_iter_count(iter); 844 loff_t pos = iocb->ki_pos; 845 bool write = iov_iter_rw(iter) == WRITE; 846 847 if (write && ceph_snap(file_inode(file)) != CEPH_NOSNAP) 848 return -EROFS; 849 850 dout("sync_direct_read_write (%s) on file %p %lld~%u\n", 851 (write ? "write" : "read"), file, pos, (unsigned)count); 852 853 ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + count); 854 if (ret < 0) 855 return ret; 856 857 if (write) { 858 int ret2 = invalidate_inode_pages2_range(inode->i_mapping, 859 pos >> PAGE_SHIFT, 860 (pos + count) >> PAGE_SHIFT); 861 if (ret2 < 0) 862 dout("invalidate_inode_pages2_range returned %d\n", ret2); 863 864 flags = CEPH_OSD_FLAG_ORDERSNAP | CEPH_OSD_FLAG_WRITE; 865 } else { 866 flags = CEPH_OSD_FLAG_READ; 867 } 868 869 while (iov_iter_count(iter) > 0) { 870 u64 size = dio_get_pagev_size(iter); 871 size_t start = 0; 872 ssize_t len; 873 874 vino = ceph_vino(inode); 875 req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout, 876 vino, pos, &size, 0, 877 /*include a 'startsync' command*/ 878 write ? 2 : 1, 879 write ? CEPH_OSD_OP_WRITE : 880 CEPH_OSD_OP_READ, 881 flags, snapc, 882 ci->i_truncate_seq, 883 ci->i_truncate_size, 884 false); 885 if (IS_ERR(req)) { 886 ret = PTR_ERR(req); 887 break; 888 } 889 890 len = size; 891 pages = dio_get_pages_alloc(iter, len, &start, &num_pages); 892 if (IS_ERR(pages)) { 893 ceph_osdc_put_request(req); 894 ret = PTR_ERR(pages); 895 break; 896 } 897 898 /* 899 * To simplify error handling, allow AIO when IO within i_size 900 * or IO can be satisfied by single OSD request. 901 */ 902 if (pos == iocb->ki_pos && !is_sync_kiocb(iocb) && 903 (len == count || pos + count <= i_size_read(inode))) { 904 aio_req = kzalloc(sizeof(*aio_req), GFP_KERNEL); 905 if (aio_req) { 906 aio_req->iocb = iocb; 907 aio_req->write = write; 908 INIT_LIST_HEAD(&aio_req->osd_reqs); 909 if (write) { 910 aio_req->mtime = mtime; 911 swap(aio_req->prealloc_cf, *pcf); 912 } 913 } 914 /* ignore error */ 915 } 916 917 if (write) { 918 /* 919 * throw out any page cache pages in this range. this 920 * may block. 921 */ 922 truncate_inode_pages_range(inode->i_mapping, pos, 923 (pos+len) | (PAGE_SIZE - 1)); 924 925 osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC, 0); 926 req->r_mtime = mtime; 927 } 928 929 osd_req_op_extent_osd_data_pages(req, 0, pages, len, start, 930 false, false); 931 932 if (aio_req) { 933 aio_req->total_len += len; 934 aio_req->num_reqs++; 935 atomic_inc(&aio_req->pending_reqs); 936 937 req->r_callback = ceph_aio_complete_req; 938 req->r_inode = inode; 939 req->r_priv = aio_req; 940 list_add_tail(&req->r_unsafe_item, &aio_req->osd_reqs); 941 942 pos += len; 943 iov_iter_advance(iter, len); 944 continue; 945 } 946 947 ret = ceph_osdc_start_request(req->r_osdc, req, false); 948 if (!ret) 949 ret = ceph_osdc_wait_request(&fsc->client->osdc, req); 950 951 size = i_size_read(inode); 952 if (!write) { 953 if (ret == -ENOENT) 954 ret = 0; 955 if (ret >= 0 && ret < len && pos + ret < size) { 956 int zlen = min_t(size_t, len - ret, 957 size - pos - ret); 958 ceph_zero_page_vector_range(start + ret, zlen, 959 pages); 960 ret += zlen; 961 } 962 if (ret >= 0) 963 len = ret; 964 } 965 966 ceph_put_page_vector(pages, num_pages, !write); 967 968 ceph_osdc_put_request(req); 969 if (ret < 0) 970 break; 971 972 pos += len; 973 iov_iter_advance(iter, len); 974 975 if (!write && pos >= size) 976 break; 977 978 if (write && pos > size) { 979 if (ceph_inode_set_size(inode, pos)) 980 ceph_check_caps(ceph_inode(inode), 981 CHECK_CAPS_AUTHONLY, 982 NULL); 983 } 984 } 985 986 if (aio_req) { 987 LIST_HEAD(osd_reqs); 988 989 if (aio_req->num_reqs == 0) { 990 kfree(aio_req); 991 return ret; 992 } 993 994 ceph_get_cap_refs(ci, write ? CEPH_CAP_FILE_WR : 995 CEPH_CAP_FILE_RD); 996 997 list_splice(&aio_req->osd_reqs, &osd_reqs); 998 while (!list_empty(&osd_reqs)) { 999 req = list_first_entry(&osd_reqs, 1000 struct ceph_osd_request, 1001 r_unsafe_item); 1002 list_del_init(&req->r_unsafe_item); 1003 if (ret >= 0) 1004 ret = ceph_osdc_start_request(req->r_osdc, 1005 req, false); 1006 if (ret < 0) { 1007 req->r_result = ret; 1008 ceph_aio_complete_req(req); 1009 } 1010 } 1011 return -EIOCBQUEUED; 1012 } 1013 1014 if (ret != -EOLDSNAPC && pos > iocb->ki_pos) { 1015 ret = pos - iocb->ki_pos; 1016 iocb->ki_pos = pos; 1017 } 1018 return ret; 1019 } 1020 1021 /* 1022 * Synchronous write, straight from __user pointer or user pages. 1023 * 1024 * If write spans object boundary, just do multiple writes. (For a 1025 * correct atomic write, we should e.g. take write locks on all 1026 * objects, rollback on failure, etc.) 1027 */ 1028 static ssize_t 1029 ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, 1030 struct ceph_snap_context *snapc) 1031 { 1032 struct file *file = iocb->ki_filp; 1033 struct inode *inode = file_inode(file); 1034 struct ceph_inode_info *ci = ceph_inode(inode); 1035 struct ceph_fs_client *fsc = ceph_inode_to_client(inode); 1036 struct ceph_vino vino; 1037 struct ceph_osd_request *req; 1038 struct page **pages; 1039 u64 len; 1040 int num_pages; 1041 int written = 0; 1042 int flags; 1043 int check_caps = 0; 1044 int ret; 1045 struct timespec mtime = current_time(inode); 1046 size_t count = iov_iter_count(from); 1047 1048 if (ceph_snap(file_inode(file)) != CEPH_NOSNAP) 1049 return -EROFS; 1050 1051 dout("sync_write on file %p %lld~%u\n", file, pos, (unsigned)count); 1052 1053 ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + count); 1054 if (ret < 0) 1055 return ret; 1056 1057 ret = invalidate_inode_pages2_range(inode->i_mapping, 1058 pos >> PAGE_SHIFT, 1059 (pos + count) >> PAGE_SHIFT); 1060 if (ret < 0) 1061 dout("invalidate_inode_pages2_range returned %d\n", ret); 1062 1063 flags = CEPH_OSD_FLAG_ORDERSNAP | CEPH_OSD_FLAG_WRITE; 1064 1065 while ((len = iov_iter_count(from)) > 0) { 1066 size_t left; 1067 int n; 1068 1069 vino = ceph_vino(inode); 1070 req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout, 1071 vino, pos, &len, 0, 1, 1072 CEPH_OSD_OP_WRITE, flags, snapc, 1073 ci->i_truncate_seq, 1074 ci->i_truncate_size, 1075 false); 1076 if (IS_ERR(req)) { 1077 ret = PTR_ERR(req); 1078 break; 1079 } 1080 1081 /* 1082 * write from beginning of first page, 1083 * regardless of io alignment 1084 */ 1085 num_pages = (len + PAGE_SIZE - 1) >> PAGE_SHIFT; 1086 1087 pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); 1088 if (IS_ERR(pages)) { 1089 ret = PTR_ERR(pages); 1090 goto out; 1091 } 1092 1093 left = len; 1094 for (n = 0; n < num_pages; n++) { 1095 size_t plen = min_t(size_t, left, PAGE_SIZE); 1096 ret = copy_page_from_iter(pages[n], 0, plen, from); 1097 if (ret != plen) { 1098 ret = -EFAULT; 1099 break; 1100 } 1101 left -= ret; 1102 } 1103 1104 if (ret < 0) { 1105 ceph_release_page_vector(pages, num_pages); 1106 goto out; 1107 } 1108 1109 req->r_inode = inode; 1110 1111 osd_req_op_extent_osd_data_pages(req, 0, pages, len, 0, 1112 false, true); 1113 1114 req->r_mtime = mtime; 1115 ret = ceph_osdc_start_request(&fsc->client->osdc, req, false); 1116 if (!ret) 1117 ret = ceph_osdc_wait_request(&fsc->client->osdc, req); 1118 1119 out: 1120 ceph_osdc_put_request(req); 1121 if (ret != 0) { 1122 ceph_set_error_write(ci); 1123 break; 1124 } 1125 1126 ceph_clear_error_write(ci); 1127 pos += len; 1128 written += len; 1129 if (pos > i_size_read(inode)) { 1130 check_caps = ceph_inode_set_size(inode, pos); 1131 if (check_caps) 1132 ceph_check_caps(ceph_inode(inode), 1133 CHECK_CAPS_AUTHONLY, 1134 NULL); 1135 } 1136 1137 } 1138 1139 if (ret != -EOLDSNAPC && written > 0) { 1140 ret = written; 1141 iocb->ki_pos = pos; 1142 } 1143 return ret; 1144 } 1145 1146 /* 1147 * Wrap generic_file_aio_read with checks for cap bits on the inode. 1148 * Atomically grab references, so that those bits are not released 1149 * back to the MDS mid-read. 1150 * 1151 * Hmm, the sync read case isn't actually async... should it be? 1152 */ 1153 static ssize_t ceph_read_iter(struct kiocb *iocb, struct iov_iter *to) 1154 { 1155 struct file *filp = iocb->ki_filp; 1156 struct ceph_file_info *fi = filp->private_data; 1157 size_t len = iov_iter_count(to); 1158 struct inode *inode = file_inode(filp); 1159 struct ceph_inode_info *ci = ceph_inode(inode); 1160 struct page *pinned_page = NULL; 1161 ssize_t ret; 1162 int want, got = 0; 1163 int retry_op = 0, read = 0; 1164 1165 again: 1166 dout("aio_read %p %llx.%llx %llu~%u trying to get caps on %p\n", 1167 inode, ceph_vinop(inode), iocb->ki_pos, (unsigned)len, inode); 1168 1169 if (fi->fmode & CEPH_FILE_MODE_LAZY) 1170 want = CEPH_CAP_FILE_CACHE | CEPH_CAP_FILE_LAZYIO; 1171 else 1172 want = CEPH_CAP_FILE_CACHE; 1173 ret = ceph_get_caps(ci, CEPH_CAP_FILE_RD, want, -1, &got, &pinned_page); 1174 if (ret < 0) 1175 return ret; 1176 1177 if ((got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0 || 1178 (iocb->ki_flags & IOCB_DIRECT) || 1179 (fi->flags & CEPH_F_SYNC)) { 1180 1181 dout("aio_sync_read %p %llx.%llx %llu~%u got cap refs on %s\n", 1182 inode, ceph_vinop(inode), iocb->ki_pos, (unsigned)len, 1183 ceph_cap_string(got)); 1184 1185 if (ci->i_inline_version == CEPH_INLINE_NONE) { 1186 if (!retry_op && (iocb->ki_flags & IOCB_DIRECT)) { 1187 ret = ceph_direct_read_write(iocb, to, 1188 NULL, NULL); 1189 if (ret >= 0 && ret < len) 1190 retry_op = CHECK_EOF; 1191 } else { 1192 ret = ceph_sync_read(iocb, to, &retry_op); 1193 } 1194 } else { 1195 retry_op = READ_INLINE; 1196 } 1197 } else { 1198 dout("aio_read %p %llx.%llx %llu~%u got cap refs on %s\n", 1199 inode, ceph_vinop(inode), iocb->ki_pos, (unsigned)len, 1200 ceph_cap_string(got)); 1201 current->journal_info = filp; 1202 ret = generic_file_read_iter(iocb, to); 1203 current->journal_info = NULL; 1204 } 1205 dout("aio_read %p %llx.%llx dropping cap refs on %s = %d\n", 1206 inode, ceph_vinop(inode), ceph_cap_string(got), (int)ret); 1207 if (pinned_page) { 1208 put_page(pinned_page); 1209 pinned_page = NULL; 1210 } 1211 ceph_put_cap_refs(ci, got); 1212 if (retry_op > HAVE_RETRIED && ret >= 0) { 1213 int statret; 1214 struct page *page = NULL; 1215 loff_t i_size; 1216 if (retry_op == READ_INLINE) { 1217 page = __page_cache_alloc(GFP_KERNEL); 1218 if (!page) 1219 return -ENOMEM; 1220 } 1221 1222 statret = __ceph_do_getattr(inode, page, 1223 CEPH_STAT_CAP_INLINE_DATA, !!page); 1224 if (statret < 0) { 1225 if (page) 1226 __free_page(page); 1227 if (statret == -ENODATA) { 1228 BUG_ON(retry_op != READ_INLINE); 1229 goto again; 1230 } 1231 return statret; 1232 } 1233 1234 i_size = i_size_read(inode); 1235 if (retry_op == READ_INLINE) { 1236 BUG_ON(ret > 0 || read > 0); 1237 if (iocb->ki_pos < i_size && 1238 iocb->ki_pos < PAGE_SIZE) { 1239 loff_t end = min_t(loff_t, i_size, 1240 iocb->ki_pos + len); 1241 end = min_t(loff_t, end, PAGE_SIZE); 1242 if (statret < end) 1243 zero_user_segment(page, statret, end); 1244 ret = copy_page_to_iter(page, 1245 iocb->ki_pos & ~PAGE_MASK, 1246 end - iocb->ki_pos, to); 1247 iocb->ki_pos += ret; 1248 read += ret; 1249 } 1250 if (iocb->ki_pos < i_size && read < len) { 1251 size_t zlen = min_t(size_t, len - read, 1252 i_size - iocb->ki_pos); 1253 ret = iov_iter_zero(zlen, to); 1254 iocb->ki_pos += ret; 1255 read += ret; 1256 } 1257 __free_pages(page, 0); 1258 return read; 1259 } 1260 1261 /* hit EOF or hole? */ 1262 if (retry_op == CHECK_EOF && iocb->ki_pos < i_size && 1263 ret < len) { 1264 dout("sync_read hit hole, ppos %lld < size %lld" 1265 ", reading more\n", iocb->ki_pos, i_size); 1266 1267 read += ret; 1268 len -= ret; 1269 retry_op = HAVE_RETRIED; 1270 goto again; 1271 } 1272 } 1273 1274 if (ret >= 0) 1275 ret += read; 1276 1277 return ret; 1278 } 1279 1280 /* 1281 * Take cap references to avoid releasing caps to MDS mid-write. 1282 * 1283 * If we are synchronous, and write with an old snap context, the OSD 1284 * may return EOLDSNAPC. In that case, retry the write.. _after_ 1285 * dropping our cap refs and allowing the pending snap to logically 1286 * complete _before_ this write occurs. 1287 * 1288 * If we are near ENOSPC, write synchronously. 1289 */ 1290 static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) 1291 { 1292 struct file *file = iocb->ki_filp; 1293 struct ceph_file_info *fi = file->private_data; 1294 struct inode *inode = file_inode(file); 1295 struct ceph_inode_info *ci = ceph_inode(inode); 1296 struct ceph_osd_client *osdc = 1297 &ceph_sb_to_client(inode->i_sb)->client->osdc; 1298 struct ceph_cap_flush *prealloc_cf; 1299 ssize_t count, written = 0; 1300 int err, want, got; 1301 loff_t pos; 1302 1303 if (ceph_snap(inode) != CEPH_NOSNAP) 1304 return -EROFS; 1305 1306 prealloc_cf = ceph_alloc_cap_flush(); 1307 if (!prealloc_cf) 1308 return -ENOMEM; 1309 1310 inode_lock(inode); 1311 1312 /* We can write back this queue in page reclaim */ 1313 current->backing_dev_info = inode_to_bdi(inode); 1314 1315 if (iocb->ki_flags & IOCB_APPEND) { 1316 err = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE, false); 1317 if (err < 0) 1318 goto out; 1319 } 1320 1321 err = generic_write_checks(iocb, from); 1322 if (err <= 0) 1323 goto out; 1324 1325 pos = iocb->ki_pos; 1326 count = iov_iter_count(from); 1327 err = file_remove_privs(file); 1328 if (err) 1329 goto out; 1330 1331 err = file_update_time(file); 1332 if (err) 1333 goto out; 1334 1335 if (ci->i_inline_version != CEPH_INLINE_NONE) { 1336 err = ceph_uninline_data(file, NULL); 1337 if (err < 0) 1338 goto out; 1339 } 1340 1341 retry_snap: 1342 /* FIXME: not complete since it doesn't account for being at quota */ 1343 if (ceph_osdmap_flag(osdc, CEPH_OSDMAP_FULL)) { 1344 err = -ENOSPC; 1345 goto out; 1346 } 1347 1348 dout("aio_write %p %llx.%llx %llu~%zd getting caps. i_size %llu\n", 1349 inode, ceph_vinop(inode), pos, count, i_size_read(inode)); 1350 if (fi->fmode & CEPH_FILE_MODE_LAZY) 1351 want = CEPH_CAP_FILE_BUFFER | CEPH_CAP_FILE_LAZYIO; 1352 else 1353 want = CEPH_CAP_FILE_BUFFER; 1354 got = 0; 1355 err = ceph_get_caps(ci, CEPH_CAP_FILE_WR, want, pos + count, 1356 &got, NULL); 1357 if (err < 0) 1358 goto out; 1359 1360 dout("aio_write %p %llx.%llx %llu~%zd got cap refs on %s\n", 1361 inode, ceph_vinop(inode), pos, count, ceph_cap_string(got)); 1362 1363 if ((got & (CEPH_CAP_FILE_BUFFER|CEPH_CAP_FILE_LAZYIO)) == 0 || 1364 (iocb->ki_flags & IOCB_DIRECT) || (fi->flags & CEPH_F_SYNC) || 1365 (ci->i_ceph_flags & CEPH_I_ERROR_WRITE)) { 1366 struct ceph_snap_context *snapc; 1367 struct iov_iter data; 1368 inode_unlock(inode); 1369 1370 spin_lock(&ci->i_ceph_lock); 1371 if (__ceph_have_pending_cap_snap(ci)) { 1372 struct ceph_cap_snap *capsnap = 1373 list_last_entry(&ci->i_cap_snaps, 1374 struct ceph_cap_snap, 1375 ci_item); 1376 snapc = ceph_get_snap_context(capsnap->context); 1377 } else { 1378 BUG_ON(!ci->i_head_snapc); 1379 snapc = ceph_get_snap_context(ci->i_head_snapc); 1380 } 1381 spin_unlock(&ci->i_ceph_lock); 1382 1383 /* we might need to revert back to that point */ 1384 data = *from; 1385 if (iocb->ki_flags & IOCB_DIRECT) 1386 written = ceph_direct_read_write(iocb, &data, snapc, 1387 &prealloc_cf); 1388 else 1389 written = ceph_sync_write(iocb, &data, pos, snapc); 1390 if (written == -EOLDSNAPC) { 1391 dout("aio_write %p %llx.%llx %llu~%u" 1392 "got EOLDSNAPC, retrying\n", 1393 inode, ceph_vinop(inode), 1394 pos, (unsigned)count); 1395 inode_lock(inode); 1396 goto retry_snap; 1397 } 1398 if (written > 0) 1399 iov_iter_advance(from, written); 1400 ceph_put_snap_context(snapc); 1401 } else { 1402 /* 1403 * No need to acquire the i_truncate_mutex. Because 1404 * the MDS revokes Fwb caps before sending truncate 1405 * message to us. We can't get Fwb cap while there 1406 * are pending vmtruncate. So write and vmtruncate 1407 * can not run at the same time 1408 */ 1409 written = generic_perform_write(file, from, pos); 1410 if (likely(written >= 0)) 1411 iocb->ki_pos = pos + written; 1412 inode_unlock(inode); 1413 } 1414 1415 if (written >= 0) { 1416 int dirty; 1417 spin_lock(&ci->i_ceph_lock); 1418 ci->i_inline_version = CEPH_INLINE_NONE; 1419 dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR, 1420 &prealloc_cf); 1421 spin_unlock(&ci->i_ceph_lock); 1422 if (dirty) 1423 __mark_inode_dirty(inode, dirty); 1424 } 1425 1426 dout("aio_write %p %llx.%llx %llu~%u dropping cap refs on %s\n", 1427 inode, ceph_vinop(inode), pos, (unsigned)count, 1428 ceph_cap_string(got)); 1429 ceph_put_cap_refs(ci, got); 1430 1431 if (written >= 0) { 1432 if (ceph_osdmap_flag(osdc, CEPH_OSDMAP_NEARFULL)) 1433 iocb->ki_flags |= IOCB_DSYNC; 1434 1435 written = generic_write_sync(iocb, written); 1436 } 1437 1438 goto out_unlocked; 1439 1440 out: 1441 inode_unlock(inode); 1442 out_unlocked: 1443 ceph_free_cap_flush(prealloc_cf); 1444 current->backing_dev_info = NULL; 1445 return written ? written : err; 1446 } 1447 1448 /* 1449 * llseek. be sure to verify file size on SEEK_END. 1450 */ 1451 static loff_t ceph_llseek(struct file *file, loff_t offset, int whence) 1452 { 1453 struct inode *inode = file->f_mapping->host; 1454 loff_t i_size; 1455 loff_t ret; 1456 1457 inode_lock(inode); 1458 1459 if (whence == SEEK_END || whence == SEEK_DATA || whence == SEEK_HOLE) { 1460 ret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE, false); 1461 if (ret < 0) 1462 goto out; 1463 } 1464 1465 i_size = i_size_read(inode); 1466 switch (whence) { 1467 case SEEK_END: 1468 offset += i_size; 1469 break; 1470 case SEEK_CUR: 1471 /* 1472 * Here we special-case the lseek(fd, 0, SEEK_CUR) 1473 * position-querying operation. Avoid rewriting the "same" 1474 * f_pos value back to the file because a concurrent read(), 1475 * write() or lseek() might have altered it 1476 */ 1477 if (offset == 0) { 1478 ret = file->f_pos; 1479 goto out; 1480 } 1481 offset += file->f_pos; 1482 break; 1483 case SEEK_DATA: 1484 if (offset >= i_size) { 1485 ret = -ENXIO; 1486 goto out; 1487 } 1488 break; 1489 case SEEK_HOLE: 1490 if (offset >= i_size) { 1491 ret = -ENXIO; 1492 goto out; 1493 } 1494 offset = i_size; 1495 break; 1496 } 1497 1498 ret = vfs_setpos(file, offset, inode->i_sb->s_maxbytes); 1499 1500 out: 1501 inode_unlock(inode); 1502 return ret; 1503 } 1504 1505 static inline void ceph_zero_partial_page( 1506 struct inode *inode, loff_t offset, unsigned size) 1507 { 1508 struct page *page; 1509 pgoff_t index = offset >> PAGE_SHIFT; 1510 1511 page = find_lock_page(inode->i_mapping, index); 1512 if (page) { 1513 wait_on_page_writeback(page); 1514 zero_user(page, offset & (PAGE_SIZE - 1), size); 1515 unlock_page(page); 1516 put_page(page); 1517 } 1518 } 1519 1520 static void ceph_zero_pagecache_range(struct inode *inode, loff_t offset, 1521 loff_t length) 1522 { 1523 loff_t nearly = round_up(offset, PAGE_SIZE); 1524 if (offset < nearly) { 1525 loff_t size = nearly - offset; 1526 if (length < size) 1527 size = length; 1528 ceph_zero_partial_page(inode, offset, size); 1529 offset += size; 1530 length -= size; 1531 } 1532 if (length >= PAGE_SIZE) { 1533 loff_t size = round_down(length, PAGE_SIZE); 1534 truncate_pagecache_range(inode, offset, offset + size - 1); 1535 offset += size; 1536 length -= size; 1537 } 1538 if (length) 1539 ceph_zero_partial_page(inode, offset, length); 1540 } 1541 1542 static int ceph_zero_partial_object(struct inode *inode, 1543 loff_t offset, loff_t *length) 1544 { 1545 struct ceph_inode_info *ci = ceph_inode(inode); 1546 struct ceph_fs_client *fsc = ceph_inode_to_client(inode); 1547 struct ceph_osd_request *req; 1548 int ret = 0; 1549 loff_t zero = 0; 1550 int op; 1551 1552 if (!length) { 1553 op = offset ? CEPH_OSD_OP_DELETE : CEPH_OSD_OP_TRUNCATE; 1554 length = &zero; 1555 } else { 1556 op = CEPH_OSD_OP_ZERO; 1557 } 1558 1559 req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout, 1560 ceph_vino(inode), 1561 offset, length, 1562 0, 1, op, 1563 CEPH_OSD_FLAG_WRITE, 1564 NULL, 0, 0, false); 1565 if (IS_ERR(req)) { 1566 ret = PTR_ERR(req); 1567 goto out; 1568 } 1569 1570 req->r_mtime = inode->i_mtime; 1571 ret = ceph_osdc_start_request(&fsc->client->osdc, req, false); 1572 if (!ret) { 1573 ret = ceph_osdc_wait_request(&fsc->client->osdc, req); 1574 if (ret == -ENOENT) 1575 ret = 0; 1576 } 1577 ceph_osdc_put_request(req); 1578 1579 out: 1580 return ret; 1581 } 1582 1583 static int ceph_zero_objects(struct inode *inode, loff_t offset, loff_t length) 1584 { 1585 int ret = 0; 1586 struct ceph_inode_info *ci = ceph_inode(inode); 1587 s32 stripe_unit = ci->i_layout.stripe_unit; 1588 s32 stripe_count = ci->i_layout.stripe_count; 1589 s32 object_size = ci->i_layout.object_size; 1590 u64 object_set_size = object_size * stripe_count; 1591 u64 nearly, t; 1592 1593 /* round offset up to next period boundary */ 1594 nearly = offset + object_set_size - 1; 1595 t = nearly; 1596 nearly -= do_div(t, object_set_size); 1597 1598 while (length && offset < nearly) { 1599 loff_t size = length; 1600 ret = ceph_zero_partial_object(inode, offset, &size); 1601 if (ret < 0) 1602 return ret; 1603 offset += size; 1604 length -= size; 1605 } 1606 while (length >= object_set_size) { 1607 int i; 1608 loff_t pos = offset; 1609 for (i = 0; i < stripe_count; ++i) { 1610 ret = ceph_zero_partial_object(inode, pos, NULL); 1611 if (ret < 0) 1612 return ret; 1613 pos += stripe_unit; 1614 } 1615 offset += object_set_size; 1616 length -= object_set_size; 1617 } 1618 while (length) { 1619 loff_t size = length; 1620 ret = ceph_zero_partial_object(inode, offset, &size); 1621 if (ret < 0) 1622 return ret; 1623 offset += size; 1624 length -= size; 1625 } 1626 return ret; 1627 } 1628 1629 static long ceph_fallocate(struct file *file, int mode, 1630 loff_t offset, loff_t length) 1631 { 1632 struct ceph_file_info *fi = file->private_data; 1633 struct inode *inode = file_inode(file); 1634 struct ceph_inode_info *ci = ceph_inode(inode); 1635 struct ceph_osd_client *osdc = 1636 &ceph_inode_to_client(inode)->client->osdc; 1637 struct ceph_cap_flush *prealloc_cf; 1638 int want, got = 0; 1639 int dirty; 1640 int ret = 0; 1641 loff_t endoff = 0; 1642 loff_t size; 1643 1644 if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) 1645 return -EOPNOTSUPP; 1646 1647 if (!S_ISREG(inode->i_mode)) 1648 return -EOPNOTSUPP; 1649 1650 prealloc_cf = ceph_alloc_cap_flush(); 1651 if (!prealloc_cf) 1652 return -ENOMEM; 1653 1654 inode_lock(inode); 1655 1656 if (ceph_snap(inode) != CEPH_NOSNAP) { 1657 ret = -EROFS; 1658 goto unlock; 1659 } 1660 1661 if (ceph_osdmap_flag(osdc, CEPH_OSDMAP_FULL) && 1662 !(mode & FALLOC_FL_PUNCH_HOLE)) { 1663 ret = -ENOSPC; 1664 goto unlock; 1665 } 1666 1667 if (ci->i_inline_version != CEPH_INLINE_NONE) { 1668 ret = ceph_uninline_data(file, NULL); 1669 if (ret < 0) 1670 goto unlock; 1671 } 1672 1673 size = i_size_read(inode); 1674 if (!(mode & FALLOC_FL_KEEP_SIZE)) { 1675 endoff = offset + length; 1676 ret = inode_newsize_ok(inode, endoff); 1677 if (ret) 1678 goto unlock; 1679 } 1680 1681 if (fi->fmode & CEPH_FILE_MODE_LAZY) 1682 want = CEPH_CAP_FILE_BUFFER | CEPH_CAP_FILE_LAZYIO; 1683 else 1684 want = CEPH_CAP_FILE_BUFFER; 1685 1686 ret = ceph_get_caps(ci, CEPH_CAP_FILE_WR, want, endoff, &got, NULL); 1687 if (ret < 0) 1688 goto unlock; 1689 1690 if (mode & FALLOC_FL_PUNCH_HOLE) { 1691 if (offset < size) 1692 ceph_zero_pagecache_range(inode, offset, length); 1693 ret = ceph_zero_objects(inode, offset, length); 1694 } else if (endoff > size) { 1695 truncate_pagecache_range(inode, size, -1); 1696 if (ceph_inode_set_size(inode, endoff)) 1697 ceph_check_caps(ceph_inode(inode), 1698 CHECK_CAPS_AUTHONLY, NULL); 1699 } 1700 1701 if (!ret) { 1702 spin_lock(&ci->i_ceph_lock); 1703 ci->i_inline_version = CEPH_INLINE_NONE; 1704 dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR, 1705 &prealloc_cf); 1706 spin_unlock(&ci->i_ceph_lock); 1707 if (dirty) 1708 __mark_inode_dirty(inode, dirty); 1709 } 1710 1711 ceph_put_cap_refs(ci, got); 1712 unlock: 1713 inode_unlock(inode); 1714 ceph_free_cap_flush(prealloc_cf); 1715 return ret; 1716 } 1717 1718 const struct file_operations ceph_file_fops = { 1719 .open = ceph_open, 1720 .release = ceph_release, 1721 .llseek = ceph_llseek, 1722 .read_iter = ceph_read_iter, 1723 .write_iter = ceph_write_iter, 1724 .mmap = ceph_mmap, 1725 .fsync = ceph_fsync, 1726 .lock = ceph_lock, 1727 .flock = ceph_flock, 1728 .splice_read = generic_file_splice_read, 1729 .splice_write = iter_file_splice_write, 1730 .unlocked_ioctl = ceph_ioctl, 1731 .compat_ioctl = ceph_ioctl, 1732 .fallocate = ceph_fallocate, 1733 }; 1734 1735