file.c 33.4 KB
Newer Older
1
#include <linux/ceph/ceph_debug.h>
S
Sage Weil 已提交
2

3
#include <linux/module.h>
S
Sage Weil 已提交
4
#include <linux/sched.h>
5
#include <linux/slab.h>
S
Sage Weil 已提交
6
#include <linux/file.h>
S
Sage Weil 已提交
7
#include <linux/mount.h>
S
Sage Weil 已提交
8 9
#include <linux/namei.h>
#include <linux/writeback.h>
10
#include <linux/aio.h>
L
Li Wang 已提交
11
#include <linux/falloc.h>
S
Sage Weil 已提交
12 13 14

#include "super.h"
#include "mds_client.h"
15
#include "cache.h"
S
Sage Weil 已提交
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

/*
 * Ceph file operations
 *
 * Implement basic open/close functionality, and implement
 * read/write.
 *
 * We implement three modes of file I/O:
 *  - buffered uses the generic_file_aio_{read,write} helpers
 *
 *  - synchronous is used when there is multi-client read/write
 *    sharing, avoids the page cache, and synchronously waits for an
 *    ack from the OSD.
 *
 *  - direct io takes the variant of the sync path that references
 *    user pages directly.
 *
 * fsync() flushes and waits on dirty pages, but just queues metadata
 * for writeback: since the MDS can recover size and mtime there is no
 * need to wait for MDS acknowledgement.
 */


/*
 * Prepare an open request.  Preallocate ceph_cap to avoid an
 * inopportune ENOMEM later.
 */
static struct ceph_mds_request *
prepare_open_request(struct super_block *sb, int flags, int create_mode)
{
46 47
	struct ceph_fs_client *fsc = ceph_sb_to_client(sb);
	struct ceph_mds_client *mdsc = fsc->mdsc;
S
Sage Weil 已提交
48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
	struct ceph_mds_request *req;
	int want_auth = USE_ANY_MDS;
	int op = (flags & O_CREAT) ? CEPH_MDS_OP_CREATE : CEPH_MDS_OP_OPEN;

	if (flags & (O_WRONLY|O_RDWR|O_CREAT|O_TRUNC))
		want_auth = USE_AUTH_MDS;

	req = ceph_mdsc_create_request(mdsc, op, want_auth);
	if (IS_ERR(req))
		goto out;
	req->r_fmode = ceph_flags_to_mode(flags);
	req->r_args.open.flags = cpu_to_le32(flags);
	req->r_args.open.mode = cpu_to_le32(create_mode);
out:
	return req;
}

/*
 * initialize private struct file data.
 * if we fail, clean up by dropping fmode reference on the ceph_inode
 */
static int ceph_init_file(struct inode *inode, struct file *file, int fmode)
{
	struct ceph_file_info *cf;
	int ret = 0;
73 74 75
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_fs_client *fsc = ceph_sb_to_client(inode->i_sb);
	struct ceph_mds_client *mdsc = fsc->mdsc;
S
Sage Weil 已提交
76 77 78

	switch (inode->i_mode & S_IFMT) {
	case S_IFREG:
79 80 81 82 83 84 85 86 87 88 89
		/* First file open request creates the cookie, we want to keep
		 * this cookie around for the filetime of the inode as not to
		 * have to worry about fscache register / revoke / operation
		 * races.
		 *
		 * Also, if we know the operation is going to invalidate data
		 * (non readonly) just nuke the cache right away.
		 */
		ceph_fscache_register_inode_cookie(mdsc->fsc, ci);
		if ((fmode & CEPH_FILE_MODE_WR))
			ceph_fscache_invalidate(inode);
S
Sage Weil 已提交
90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
	case S_IFDIR:
		dout("init_file %p %p 0%o (regular)\n", inode, file,
		     inode->i_mode);
		cf = kmem_cache_alloc(ceph_file_cachep, GFP_NOFS | __GFP_ZERO);
		if (cf == NULL) {
			ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */
			return -ENOMEM;
		}
		cf->fmode = fmode;
		cf->next_offset = 2;
		file->private_data = cf;
		BUG_ON(inode->i_fop->release != ceph_release);
		break;

	case S_IFLNK:
		dout("init_file %p %p 0%o (symlink)\n", inode, file,
		     inode->i_mode);
		ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */
		break;

	default:
		dout("init_file %p %p 0%o (special)\n", inode, file,
		     inode->i_mode);
		/*
		 * we need to drop the open ref now, since we don't
		 * have .release set to ceph_release.
		 */
		ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */
		BUG_ON(inode->i_fop->release == ceph_release);

		/* call the proper open fop */
		ret = inode->i_fop->open(inode, file);
	}
	return ret;
}

/*
 * If we already have the requisite capabilities, we can satisfy
 * the open request locally (no need to request new caps from the
 * MDS).  We do, however, need to inform the MDS (asynchronously)
 * if our wanted caps set expands.
 */
int ceph_open(struct inode *inode, struct file *file)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
135 136
	struct ceph_fs_client *fsc = ceph_sb_to_client(inode->i_sb);
	struct ceph_mds_client *mdsc = fsc->mdsc;
S
Sage Weil 已提交
137 138
	struct ceph_mds_request *req;
	struct ceph_file_info *cf = file->private_data;
S
Sage Weil 已提交
139
	struct inode *parent_inode = NULL;
S
Sage Weil 已提交
140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163
	int err;
	int flags, fmode, wanted;

	if (cf) {
		dout("open file %p is already opened\n", file);
		return 0;
	}

	/* filter out O_CREAT|O_EXCL; vfs did that already.  yuck. */
	flags = file->f_flags & ~(O_CREAT|O_EXCL);
	if (S_ISDIR(inode->i_mode))
		flags = O_DIRECTORY;  /* mds likes to know */

	dout("open inode %p ino %llx.%llx file %p flags %d (%d)\n", inode,
	     ceph_vinop(inode), file, flags, file->f_flags);
	fmode = ceph_flags_to_mode(flags);
	wanted = ceph_caps_for_mode(fmode);

	/* snapped files are read-only */
	if (ceph_snap(inode) != CEPH_NOSNAP && (file->f_mode & FMODE_WRITE))
		return -EROFS;

	/* trivially open snapdir */
	if (ceph_snap(inode) == CEPH_SNAPDIR) {
164
		spin_lock(&ci->i_ceph_lock);
S
Sage Weil 已提交
165
		__ceph_get_fmode(ci, fmode);
166
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
167 168 169 170
		return ceph_init_file(inode, file, fmode);
	}

	/*
171 172
	 * No need to block if we have caps on the auth MDS (for
	 * write) or any MDS (for read).  Update wanted set
S
Sage Weil 已提交
173 174
	 * asynchronously.
	 */
175
	spin_lock(&ci->i_ceph_lock);
176 177
	if (__ceph_is_any_real_caps(ci) &&
	    (((fmode & CEPH_FILE_MODE_WR) == 0) || ci->i_auth_cap)) {
S
Sage Weil 已提交
178 179 180 181 182 183 184
		int mds_wanted = __ceph_caps_mds_wanted(ci);
		int issued = __ceph_caps_issued(ci, NULL);

		dout("open %p fmode %d want %s issued %s using existing\n",
		     inode, fmode, ceph_cap_string(wanted),
		     ceph_cap_string(issued));
		__ceph_get_fmode(ci, fmode);
185
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
186 187 188 189 190 191 192 193 194 195 196

		/* adjust wanted? */
		if ((issued & wanted) != wanted &&
		    (mds_wanted & wanted) != wanted &&
		    ceph_snap(inode) != CEPH_SNAPDIR)
			ceph_check_caps(ci, 0, NULL);

		return ceph_init_file(inode, file, fmode);
	} else if (ceph_snap(inode) != CEPH_NOSNAP &&
		   (ci->i_snap_caps & wanted) == wanted) {
		__ceph_get_fmode(ci, fmode);
197
		spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
198 199
		return ceph_init_file(inode, file, fmode);
	}
200

201
	spin_unlock(&ci->i_ceph_lock);
S
Sage Weil 已提交
202 203 204 205 206 207 208

	dout("open fmode %d wants %s\n", fmode, ceph_cap_string(wanted));
	req = prepare_open_request(inode->i_sb, flags, 0);
	if (IS_ERR(req)) {
		err = PTR_ERR(req);
		goto out;
	}
209 210
	req->r_inode = inode;
	ihold(inode);
211

S
Sage Weil 已提交
212
	req->r_num_caps = 1;
213
	if (flags & O_CREAT)
S
Sage Weil 已提交
214
		parent_inode = ceph_get_dentry_parent_inode(file->f_dentry);
S
Sage Weil 已提交
215
	err = ceph_mdsc_do_request(mdsc, parent_inode, req);
S
Sage Weil 已提交
216
	iput(parent_inode);
S
Sage Weil 已提交
217 218 219 220 221 222 223 224 225 226
	if (!err)
		err = ceph_init_file(inode, file, req->r_fmode);
	ceph_mdsc_put_request(req);
	dout("open result=%d on %llx.%llx\n", err, ceph_vinop(inode));
out:
	return err;
}


/*
S
Sage Weil 已提交
227 228
 * Do a lookup + open with a single request.  If we get a non-existent
 * file or symlink, return 1 so the VFS can retry.
S
Sage Weil 已提交
229
 */
S
Sage Weil 已提交
230
int ceph_atomic_open(struct inode *dir, struct dentry *dentry,
A
Al Viro 已提交
231
		     struct file *file, unsigned flags, umode_t mode,
A
Al Viro 已提交
232
		     int *opened)
S
Sage Weil 已提交
233
{
234 235
	struct ceph_fs_client *fsc = ceph_sb_to_client(dir->i_sb);
	struct ceph_mds_client *mdsc = fsc->mdsc;
S
Sage Weil 已提交
236
	struct ceph_mds_request *req;
S
Sage Weil 已提交
237
	struct dentry *dn;
S
Sage Weil 已提交
238 239
	int err;

S
Sage Weil 已提交
240 241 242 243 244 245 246 247 248 249
	dout("atomic_open %p dentry %p '%.*s' %s flags %d mode 0%o\n",
	     dir, dentry, dentry->d_name.len, dentry->d_name.name,
	     d_unhashed(dentry) ? "unhashed" : "hashed", flags, mode);

	if (dentry->d_name.len > NAME_MAX)
		return -ENAMETOOLONG;

	err = ceph_init_dentry(dentry);
	if (err < 0)
		return err;
S
Sage Weil 已提交
250 251 252 253

	/* do the open */
	req = prepare_open_request(dir->i_sb, flags, mode);
	if (IS_ERR(req))
A
Al Viro 已提交
254
		return PTR_ERR(req);
S
Sage Weil 已提交
255 256 257 258 259 260 261
	req->r_dentry = dget(dentry);
	req->r_num_caps = 2;
	if (flags & O_CREAT) {
		req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
		req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
	}
	req->r_locked_dir = dir;           /* caller holds dir->i_mutex */
262 263 264
	err = ceph_mdsc_do_request(mdsc,
				   (flags & (O_CREAT|O_TRUNC)) ? dir : NULL,
				   req);
265 266 267
	if (err)
		goto out_err;

268
	err = ceph_handle_snapdir(req, dentry, err);
S
Sage Weil 已提交
269
	if (err == 0 && (flags & O_CREAT) && !req->r_reply_info.head->is_dentry)
S
Sage Weil 已提交
270
		err = ceph_handle_notrace_create(dir, dentry);
271

S
Sage Weil 已提交
272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287
	if (d_unhashed(dentry)) {
		dn = ceph_finish_lookup(req, dentry, err);
		if (IS_ERR(dn))
			err = PTR_ERR(dn);
	} else {
		/* we were given a hashed negative dentry */
		dn = NULL;
	}
	if (err)
		goto out_err;
	if (dn || dentry->d_inode == NULL || S_ISLNK(dentry->d_inode->i_mode)) {
		/* make vfs retry on splice, ENOENT, or symlink */
		dout("atomic_open finish_no_open on dn %p\n", dn);
		err = finish_no_open(file, dn);
	} else {
		dout("atomic_open finish_open on dn %p\n", dn);
288
		if (req->r_op == CEPH_MDS_OP_CREATE && req->r_reply_info.has_create_ino) {
289
			ceph_init_acl(dentry, dentry->d_inode, dir);
290 291
			*opened |= FILE_CREATED;
		}
S
Sage Weil 已提交
292 293
		err = finish_open(file, dentry, ceph_open, opened);
	}
294

S
Sage Weil 已提交
295 296 297
out_err:
	ceph_mdsc_put_request(req);
	dout("atomic_open result=%d\n", err);
A
Al Viro 已提交
298
	return err;
S
Sage Weil 已提交
299 300 301 302 303 304 305 306 307 308 309 310 311 312 313
}

int ceph_release(struct inode *inode, struct file *file)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_file_info *cf = file->private_data;

	dout("release inode %p file %p\n", inode, file);
	ceph_put_fmode(ci, cf->fmode);
	if (cf->last_readdir)
		ceph_mdsc_put_request(cf->last_readdir);
	kfree(cf->last_name);
	kfree(cf->dir_info);
	dput(cf->dentry);
	kmem_cache_free(ceph_file_cachep, cf);
314 315

	/* wake up anyone waiting for caps on this inode */
316
	wake_up_all(&ci->i_cap_wq);
S
Sage Weil 已提交
317 318 319 320 321 322 323 324 325 326 327 328
	return 0;
}

/*
 * Read a range of bytes striped over one or more objects.  Iterate over
 * objects we stripe over.  (That's not atomic, but good enough for now.)
 *
 * If we get a short result from the OSD, check against i_size; we need to
 * only return a short read to the caller if we hit EOF.
 */
static int striped_read(struct inode *inode,
			u64 off, u64 len,
329
			struct page **pages, int num_pages,
330
			int *checkeof, bool o_direct,
331
			unsigned long buf_align)
S
Sage Weil 已提交
332
{
333
	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
S
Sage Weil 已提交
334
	struct ceph_inode_info *ci = ceph_inode(inode);
335
	u64 pos, this_len, left;
336
	int io_align, page_align;
337
	int pages_left;
S
Sage Weil 已提交
338 339 340 341 342 343 344 345 346 347 348 349 350
	int read;
	struct page **page_pos;
	int ret;
	bool hit_stripe, was_short;

	/*
	 * we may need to do multiple reads.  not atomic, unfortunately.
	 */
	pos = off;
	left = len;
	page_pos = pages;
	pages_left = num_pages;
	read = 0;
351
	io_align = off & ~PAGE_MASK;
S
Sage Weil 已提交
352 353

more:
354
	if (o_direct)
355
		page_align = (pos - io_align + buf_align) & ~PAGE_MASK;
356 357
	else
		page_align = pos & ~PAGE_MASK;
S
Sage Weil 已提交
358
	this_len = left;
359
	ret = ceph_osdc_readpages(&fsc->client->osdc, ceph_vino(inode),
S
Sage Weil 已提交
360 361 362
				  &ci->i_layout, pos, &this_len,
				  ci->i_truncate_seq,
				  ci->i_truncate_size,
363
				  page_pos, pages_left, page_align);
S
Sage Weil 已提交
364 365
	if (ret == -ENOENT)
		ret = 0;
366 367
	hit_stripe = this_len < left;
	was_short = ret >= 0 && ret < this_len;
368
	dout("striped_read %llu~%llu (read %u) got %d%s%s\n", pos, left, read,
S
Sage Weil 已提交
369 370
	     ret, hit_stripe ? " HITSTRIPE" : "", was_short ? " SHORT" : "");

371 372 373 374 375 376 377 378 379 380
	if (ret >= 0) {
		int didpages;
		if (was_short && (pos + ret < inode->i_size)) {
			u64 tmp = min(this_len - ret,
					inode->i_size - pos - ret);
			dout(" zero gap %llu to %llu\n",
				pos + ret, pos + ret + tmp);
			ceph_zero_page_vector_range(page_align + read + ret,
							tmp, pages);
			ret += tmp;
S
Sage Weil 已提交
381
		}
382 383

		didpages = (page_align + ret) >> PAGE_CACHE_SHIFT;
S
Sage Weil 已提交
384 385 386 387 388 389
		pos += ret;
		read = pos - off;
		left -= ret;
		page_pos += didpages;
		pages_left -= didpages;

390 391
		/* hit stripe and need continue*/
		if (left && hit_stripe && pos < inode->i_size)
S
Sage Weil 已提交
392 393 394
			goto more;
	}

395
	if (read > 0) {
396
		ret = read;
397 398 399
		/* did we bounce off eof? */
		if (pos + left > inode->i_size)
			*checkeof = 1;
S
Sage Weil 已提交
400 401 402 403 404 405 406 407 408 409 410 411
	}

	dout("striped_read returns %d\n", ret);
	return ret;
}

/*
 * Completely synchronous read and write methods.  Direct from __user
 * buffer to osd, or directly to user pages (if O_DIRECT).
 *
 * If the read spans object boundary, just do multiple reads.
 */
412 413
static ssize_t ceph_sync_read(struct kiocb *iocb, struct iov_iter *i,
				int *checkeof)
S
Sage Weil 已提交
414
{
415
	struct file *file = iocb->ki_filp;
A
Al Viro 已提交
416
	struct inode *inode = file_inode(file);
S
Sage Weil 已提交
417
	struct page **pages;
418
	u64 off = iocb->ki_pos;
419
	int num_pages, ret;
420
	size_t len = i->count;
S
Sage Weil 已提交
421

422 423
	dout("sync_read on file %p %llu~%u %s\n", file, off,
	     (unsigned)len,
S
Sage Weil 已提交
424
	     (file->f_flags & O_DIRECT) ? "O_DIRECT" : "");
425 426 427 428 429 430
	/*
	 * flush any page cache pages in this range.  this
	 * will make concurrent normal and sync io slow,
	 * but it will at least behave sensibly when they are
	 * in sequence.
	 */
431 432
	ret = filemap_write_and_wait_range(inode->i_mapping, off,
						off + len);
433
	if (ret < 0)
434
		return ret;
435

436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
	if (file->f_flags & O_DIRECT) {
		while (iov_iter_count(i)) {
			void __user *data = i->iov[0].iov_base + i->iov_offset;
			size_t len = i->iov[0].iov_len - i->iov_offset;

			num_pages = calc_pages_for((unsigned long)data, len);
			pages = ceph_get_direct_page_vector(data,
							    num_pages, true);
			if (IS_ERR(pages))
				return PTR_ERR(pages);

			ret = striped_read(inode, off, len,
					   pages, num_pages, checkeof,
					   1, (unsigned long)data & ~PAGE_MASK);
			ceph_put_page_vector(pages, num_pages, true);

			if (ret <= 0)
				break;
			off += ret;
			iov_iter_advance(i, ret);
			if (ret < len)
				break;
		}
	} else {
		num_pages = calc_pages_for(off, len);
		pages = ceph_alloc_page_vector(num_pages, GFP_NOFS);
		if (IS_ERR(pages))
			return PTR_ERR(pages);
		ret = striped_read(inode, off, len, pages,
					num_pages, checkeof, 0, 0);
		if (ret > 0) {
			int l, k = 0;
			size_t left = len = ret;

			while (left) {
				void __user *data = i->iov[0].iov_base
							+ i->iov_offset;
				l = min(i->iov[0].iov_len - i->iov_offset,
					left);

				ret = ceph_copy_page_vector_to_user(&pages[k],
								    data, off,
								    l);
				if (ret > 0) {
					iov_iter_advance(i, ret);
					left -= ret;
					off += ret;
					k = calc_pages_for(iocb->ki_pos,
							   len - left + 1) - 1;
					BUG_ON(k >= num_pages && left);
				} else
					break;
			}
		}
		ceph_release_page_vector(pages, num_pages);
	}
S
Sage Weil 已提交
492

493 494 495 496
	if (off > iocb->ki_pos) {
		ret = off - iocb->ki_pos;
		iocb->ki_pos = off;
	}
S
Sage Weil 已提交
497 498 499 500 501 502

	dout("sync_read result %d\n", ret);
	return ret;
}

/*
503 504 505 506 507 508 509 510 511 512
 * Write commit request unsafe callback, called to tell us when a
 * request is unsafe (that is, in flight--has been handed to the
 * messenger to send to its target osd).  It is called again when
 * we've received a response message indicating the request is
 * "safe" (its CEPH_OSD_FLAG_ONDISK flag is set), or when a request
 * is completed early (and unsuccessfully) due to a timeout or
 * interrupt.
 *
 * This is used if we requested both an ACK and ONDISK commit reply
 * from the OSD.
S
Sage Weil 已提交
513
 */
514
static void ceph_sync_write_unsafe(struct ceph_osd_request *req, bool unsafe)
S
Sage Weil 已提交
515 516 517
{
	struct ceph_inode_info *ci = ceph_inode(req->r_inode);

518 519 520 521 522 523 524 525 526 527 528 529 530 531
	dout("%s %p tid %llu %ssafe\n", __func__, req, req->r_tid,
		unsafe ? "un" : "");
	if (unsafe) {
		ceph_get_cap_refs(ci, CEPH_CAP_FILE_WR);
		spin_lock(&ci->i_unsafe_lock);
		list_add_tail(&req->r_unsafe_item,
			      &ci->i_unsafe_writes);
		spin_unlock(&ci->i_unsafe_lock);
	} else {
		spin_lock(&ci->i_unsafe_lock);
		list_del_init(&req->r_unsafe_item);
		spin_unlock(&ci->i_unsafe_lock);
		ceph_put_cap_refs(ci, CEPH_CAP_FILE_WR);
	}
S
Sage Weil 已提交
532 533
}

534

S
Sage Weil 已提交
535
/*
536
 * Synchronous write, straight from __user pointer or user pages.
S
Sage Weil 已提交
537 538 539 540 541
 *
 * If write spans object boundary, just do multiple writes.  (For a
 * correct atomic write, we should e.g. take write locks on all
 * objects, rollback on failure, etc.)
 */
542 543 544
static ssize_t
ceph_sync_direct_write(struct kiocb *iocb, const struct iovec *iov,
		       unsigned long nr_segs, size_t count)
S
Sage Weil 已提交
545
{
546
	struct file *file = iocb->ki_filp;
A
Al Viro 已提交
547
	struct inode *inode = file_inode(file);
S
Sage Weil 已提交
548
	struct ceph_inode_info *ci = ceph_inode(inode);
549
	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
550 551
	struct ceph_snap_context *snapc;
	struct ceph_vino vino;
S
Sage Weil 已提交
552 553 554 555 556 557
	struct ceph_osd_request *req;
	struct page **pages;
	int num_pages;
	int written = 0;
	int flags;
	int check_caps = 0;
558
	int page_align;
S
Sage Weil 已提交
559 560
	int ret;
	struct timespec mtime = CURRENT_TIME;
561 562
	loff_t pos = iocb->ki_pos;
	struct iov_iter i;
S
Sage Weil 已提交
563

A
Al Viro 已提交
564
	if (ceph_snap(file_inode(file)) != CEPH_NOSNAP)
S
Sage Weil 已提交
565 566
		return -EROFS;

567 568
	dout("sync_direct_write on file %p %lld~%u\n", file, pos,
	     (unsigned)count);
S
Sage Weil 已提交
569

570
	ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + count);
571 572 573 574 575
	if (ret < 0)
		return ret;

	ret = invalidate_inode_pages2_range(inode->i_mapping,
					    pos >> PAGE_CACHE_SHIFT,
576
					    (pos + count) >> PAGE_CACHE_SHIFT);
577 578 579
	if (ret < 0)
		dout("invalidate_inode_pages2_range returned %d\n", ret);

S
Sage Weil 已提交
580 581 582 583
	flags = CEPH_OSD_FLAG_ORDERSNAP |
		CEPH_OSD_FLAG_ONDISK |
		CEPH_OSD_FLAG_WRITE;

584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604
	iov_iter_init(&i, iov, nr_segs, count, 0);

	while (iov_iter_count(&i) > 0) {
		void __user *data = i.iov->iov_base + i.iov_offset;
		u64 len = i.iov->iov_len - i.iov_offset;

		page_align = (unsigned long)data & ~PAGE_MASK;

		snapc = ci->i_snap_realm->cached_context;
		vino = ceph_vino(inode);
		req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout,
					    vino, pos, &len,
					    2,/*include a 'startsync' command*/
					    CEPH_OSD_OP_WRITE, flags, snapc,
					    ci->i_truncate_seq,
					    ci->i_truncate_size,
					    false);
		if (IS_ERR(req)) {
			ret = PTR_ERR(req);
			goto out;
		}
S
Sage Weil 已提交
605

606
		num_pages = calc_pages_for(page_align, len);
607
		pages = ceph_get_direct_page_vector(data, num_pages, false);
S
Sage Weil 已提交
608 609 610 611 612 613 614 615 616
		if (IS_ERR(pages)) {
			ret = PTR_ERR(pages);
			goto out;
		}

		/*
		 * throw out any page cache pages in this range. this
		 * may block.
		 */
S
Sage Weil 已提交
617
		truncate_inode_pages_range(inode->i_mapping, pos,
618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729
				   (pos+len) | (PAGE_CACHE_SIZE-1));
		osd_req_op_extent_osd_data_pages(req, 0, pages, len, page_align,
						false, false);

		/* BUG_ON(vino.snap != CEPH_NOSNAP); */
		ceph_osdc_build_request(req, pos, snapc, vino.snap, &mtime);

		ret = ceph_osdc_start_request(&fsc->client->osdc, req, false);
		if (!ret)
			ret = ceph_osdc_wait_request(&fsc->client->osdc, req);

		ceph_put_page_vector(pages, num_pages, false);

out:
		ceph_osdc_put_request(req);
		if (ret == 0) {
			pos += len;
			written += len;
			iov_iter_advance(&i, (size_t)len);

			if (pos > i_size_read(inode)) {
				check_caps = ceph_inode_set_size(inode, pos);
				if (check_caps)
					ceph_check_caps(ceph_inode(inode),
							CHECK_CAPS_AUTHONLY,
							NULL);
			}
		} else
			break;
	}

	if (ret != -EOLDSNAPC && written > 0) {
		iocb->ki_pos = pos;
		ret = written;
	}
	return ret;
}


/*
 * Synchronous write, straight from __user pointer or user pages.
 *
 * If write spans object boundary, just do multiple writes.  (For a
 * correct atomic write, we should e.g. take write locks on all
 * objects, rollback on failure, etc.)
 */
static ssize_t ceph_sync_write(struct kiocb *iocb, const struct iovec *iov,
			       unsigned long nr_segs, size_t count)
{
	struct file *file = iocb->ki_filp;
	struct inode *inode = file_inode(file);
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
	struct ceph_snap_context *snapc;
	struct ceph_vino vino;
	struct ceph_osd_request *req;
	struct page **pages;
	u64 len;
	int num_pages;
	int written = 0;
	int flags;
	int check_caps = 0;
	int ret;
	struct timespec mtime = CURRENT_TIME;
	loff_t pos = iocb->ki_pos;
	struct iov_iter i;

	if (ceph_snap(file_inode(file)) != CEPH_NOSNAP)
		return -EROFS;

	dout("sync_write on file %p %lld~%u\n", file, pos, (unsigned)count);

	ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + count);
	if (ret < 0)
		return ret;

	ret = invalidate_inode_pages2_range(inode->i_mapping,
					    pos >> PAGE_CACHE_SHIFT,
					    (pos + count) >> PAGE_CACHE_SHIFT);
	if (ret < 0)
		dout("invalidate_inode_pages2_range returned %d\n", ret);

	flags = CEPH_OSD_FLAG_ORDERSNAP |
		CEPH_OSD_FLAG_ONDISK |
		CEPH_OSD_FLAG_WRITE |
		CEPH_OSD_FLAG_ACK;

	iov_iter_init(&i, iov, nr_segs, count, 0);

	while ((len = iov_iter_count(&i)) > 0) {
		size_t left;
		int n;

		snapc = ci->i_snap_realm->cached_context;
		vino = ceph_vino(inode);
		req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout,
					    vino, pos, &len, 1,
					    CEPH_OSD_OP_WRITE, flags, snapc,
					    ci->i_truncate_seq,
					    ci->i_truncate_size,
					    false);
		if (IS_ERR(req)) {
			ret = PTR_ERR(req);
			goto out;
		}

		/*
		 * write from beginning of first page,
		 * regardless of io alignment
		 */
		num_pages = (len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;

730
		pages = ceph_alloc_page_vector(num_pages, GFP_NOFS);
S
Sage Weil 已提交
731 732 733 734
		if (IS_ERR(pages)) {
			ret = PTR_ERR(pages);
			goto out;
		}
735 736 737

		left = len;
		for (n = 0; n < num_pages; n++) {
738
			size_t plen = min_t(size_t, left, PAGE_SIZE);
739 740 741 742 743 744 745 746 747
			ret = iov_iter_copy_from_user(pages[n], &i, 0, plen);
			if (ret != plen) {
				ret = -EFAULT;
				break;
			}
			left -= ret;
			iov_iter_advance(&i, ret);
		}

S
Sage Weil 已提交
748 749 750 751 752
		if (ret < 0) {
			ceph_release_page_vector(pages, num_pages);
			goto out;
		}

753 754 755
		/* get a second commit callback */
		req->r_unsafe_callback = ceph_sync_write_unsafe;
		req->r_inode = inode;
S
Sage Weil 已提交
756

757 758
		osd_req_op_extent_osd_data_pages(req, 0, pages, len, 0,
						false, true);
759

760 761
		/* BUG_ON(vino.snap != CEPH_NOSNAP); */
		ceph_osdc_build_request(req, pos, snapc, vino.snap, &mtime);
S
Sage Weil 已提交
762

763 764 765
		ret = ceph_osdc_start_request(&fsc->client->osdc, req, false);
		if (!ret)
			ret = ceph_osdc_wait_request(&fsc->client->osdc, req);
S
Sage Weil 已提交
766 767

out:
768 769 770 771 772 773 774 775 776 777 778 779 780 781 782
		ceph_osdc_put_request(req);
		if (ret == 0) {
			pos += len;
			written += len;

			if (pos > i_size_read(inode)) {
				check_caps = ceph_inode_set_size(inode, pos);
				if (check_caps)
					ceph_check_caps(ceph_inode(inode),
							CHECK_CAPS_AUTHONLY,
							NULL);
			}
		} else
			break;
	}
S
Sage Weil 已提交
783

784
	if (ret != -EOLDSNAPC && written > 0) {
S
Sage Weil 已提交
785
		ret = written;
786
		iocb->ki_pos = pos;
S
Sage Weil 已提交
787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
	}
	return ret;
}

/*
 * Wrap generic_file_aio_read with checks for cap bits on the inode.
 * Atomically grab references, so that those bits are not released
 * back to the MDS mid-read.
 *
 * Hmm, the sync read case isn't actually async... should it be?
 */
static ssize_t ceph_aio_read(struct kiocb *iocb, const struct iovec *iov,
			     unsigned long nr_segs, loff_t pos)
{
	struct file *filp = iocb->ki_filp;
802
	struct ceph_file_info *fi = filp->private_data;
803
	size_t len = iocb->ki_nbytes;
A
Al Viro 已提交
804
	struct inode *inode = file_inode(filp);
S
Sage Weil 已提交
805 806
	struct ceph_inode_info *ci = ceph_inode(inode);
	ssize_t ret;
807
	int want, got = 0;
808
	int checkeof = 0, read = 0;
S
Sage Weil 已提交
809

810
again:
811 812 813
	dout("aio_read %p %llx.%llx %llu~%u trying to get caps on %p\n",
	     inode, ceph_vinop(inode), iocb->ki_pos, (unsigned)len, inode);

814 815 816 817 818
	if (fi->fmode & CEPH_FILE_MODE_LAZY)
		want = CEPH_CAP_FILE_CACHE | CEPH_CAP_FILE_LAZYIO;
	else
		want = CEPH_CAP_FILE_CACHE;
	ret = ceph_get_caps(ci, CEPH_CAP_FILE_RD, want, &got, -1);
S
Sage Weil 已提交
819
	if (ret < 0)
820
		return ret;
S
Sage Weil 已提交
821

822
	if ((got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0 ||
S
Sage Weil 已提交
823
	    (iocb->ki_filp->f_flags & O_DIRECT) ||
824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839
	    (fi->flags & CEPH_F_SYNC)) {
		struct iov_iter i;

		dout("aio_sync_read %p %llx.%llx %llu~%u got cap refs on %s\n",
		     inode, ceph_vinop(inode), iocb->ki_pos, (unsigned)len,
		     ceph_cap_string(got));

		if (!read) {
			ret = generic_segment_checks(iov, &nr_segs,
							&len, VERIFY_WRITE);
			if (ret)
				goto out;
		}

		iov_iter_init(&i, iov, nr_segs, len, read);

S
Sage Weil 已提交
840
		/* hmm, this isn't really async... */
841 842 843 844 845 846 847 848 849 850 851 852 853 854
		ret = ceph_sync_read(iocb, &i, &checkeof);
	} else {
		/*
		 * We can't modify the content of iov,
		 * so we only read from beginning.
		 */
		if (read) {
			iocb->ki_pos = pos;
			len = iocb->ki_nbytes;
			read = 0;
		}
		dout("aio_read %p %llx.%llx %llu~%u got cap refs on %s\n",
		     inode, ceph_vinop(inode), pos, (unsigned)len,
		     ceph_cap_string(got));
S
Sage Weil 已提交
855

856 857
		ret = generic_file_aio_read(iocb, iov, nr_segs, pos);
	}
S
Sage Weil 已提交
858 859 860 861
out:
	dout("aio_read %p %llx.%llx dropping cap refs on %s = %d\n",
	     inode, ceph_vinop(inode), ceph_cap_string(got), (int)ret);
	ceph_put_cap_refs(ci, got);
862 863

	if (checkeof && ret >= 0) {
864 865
		int statret = ceph_do_getattr(inode,
					      CEPH_STAT_CAP_SIZE);
866 867

		/* hit EOF or hole? */
868 869 870 871 872 873
		if (statret == 0 && iocb->ki_pos < inode->i_size &&
			ret < len) {
			dout("sync_read hit hole, ppos %lld < size %lld"
			     ", reading more\n", iocb->ki_pos,
			     inode->i_size);

874 875 876 877 878 879
			read += ret;
			len -= ret;
			checkeof = 0;
			goto again;
		}
	}
880

881 882 883
	if (ret >= 0)
		ret += read;

S
Sage Weil 已提交
884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900
	return ret;
}

/*
 * Take cap references to avoid releasing caps to MDS mid-write.
 *
 * If we are synchronous, and write with an old snap context, the OSD
 * may return EOLDSNAPC.  In that case, retry the write.. _after_
 * dropping our cap refs and allowing the pending snap to logically
 * complete _before_ this write occurs.
 *
 * If we are near ENOSPC, write synchronously.
 */
static ssize_t ceph_aio_write(struct kiocb *iocb, const struct iovec *iov,
		       unsigned long nr_segs, loff_t pos)
{
	struct file *file = iocb->ki_filp;
901
	struct ceph_file_info *fi = file->private_data;
A
Al Viro 已提交
902
	struct inode *inode = file_inode(file);
S
Sage Weil 已提交
903
	struct ceph_inode_info *ci = ceph_inode(inode);
904 905
	struct ceph_osd_client *osdc =
		&ceph_sb_to_client(inode->i_sb)->client->osdc;
906 907
	ssize_t count, written = 0;
	int err, want, got;
S
Sage Weil 已提交
908 909 910 911

	if (ceph_snap(inode) != CEPH_NOSNAP)
		return -EROFS;

912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935
	mutex_lock(&inode->i_mutex);

	err = generic_segment_checks(iov, &nr_segs, &count, VERIFY_READ);
	if (err)
		goto out;

	/* We can write back this queue in page reclaim */
	current->backing_dev_info = file->f_mapping->backing_dev_info;

	err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode));
	if (err)
		goto out;

	if (count == 0)
		goto out;

	err = file_remove_suid(file);
	if (err)
		goto out;

	err = file_update_time(file);
	if (err)
		goto out;

S
Sage Weil 已提交
936
retry_snap:
Y
Yan, Zheng 已提交
937
	if (ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL)) {
938
		err = -ENOSPC;
Y
Yan, Zheng 已提交
939 940
		goto out;
	}
941

942
	dout("aio_write %p %llx.%llx %llu~%zd getting caps. i_size %llu\n",
943
	     inode, ceph_vinop(inode), pos, count, inode->i_size);
S
Sage Weil 已提交
944 945 946 947
	if (fi->fmode & CEPH_FILE_MODE_LAZY)
		want = CEPH_CAP_FILE_BUFFER | CEPH_CAP_FILE_LAZYIO;
	else
		want = CEPH_CAP_FILE_BUFFER;
948 949 950
	got = 0;
	err = ceph_get_caps(ci, CEPH_CAP_FILE_WR, want, &got, pos + count);
	if (err < 0)
951
		goto out;
S
Sage Weil 已提交
952

953
	dout("aio_write %p %llx.%llx %llu~%zd got cap refs on %s\n",
954
	     inode, ceph_vinop(inode), pos, count, ceph_cap_string(got));
S
Sage Weil 已提交
955 956

	if ((got & (CEPH_CAP_FILE_BUFFER|CEPH_CAP_FILE_LAZYIO)) == 0 ||
957
	    (file->f_flags & O_DIRECT) || (fi->flags & CEPH_F_SYNC)) {
958
		mutex_unlock(&inode->i_mutex);
959 960 961 962 963
		if (file->f_flags & O_DIRECT)
			written = ceph_sync_direct_write(iocb, iov,
							 nr_segs, count);
		else
			written = ceph_sync_write(iocb, iov, nr_segs, count);
964 965 966 967 968 969 970 971
		if (written == -EOLDSNAPC) {
			dout("aio_write %p %llx.%llx %llu~%u"
				"got EOLDSNAPC, retrying\n",
				inode, ceph_vinop(inode),
				pos, (unsigned)iov->iov_len);
			mutex_lock(&inode->i_mutex);
			goto retry_snap;
		}
S
Sage Weil 已提交
972
	} else {
Y
Yan, Zheng 已提交
973 974 975 976 977 978 979
		/*
		 * No need to acquire the i_truncate_mutex. Because
		 * the MDS revokes Fwb caps before sending truncate
		 * message to us. We can't get Fwb cap while there
		 * are pending vmtruncate. So write and vmtruncate
		 * can not run at the same time
		 */
980 981 982
		written = generic_file_buffered_write(iocb, iov, nr_segs,
						      pos, &iocb->ki_pos,
						      count, 0);
Y
Yan, Zheng 已提交
983
		mutex_unlock(&inode->i_mutex);
S
Sage Weil 已提交
984
	}
985

986
	if (written >= 0) {
987
		int dirty;
988
		spin_lock(&ci->i_ceph_lock);
989
		dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR);
990
		spin_unlock(&ci->i_ceph_lock);
991 992
		if (dirty)
			__mark_inode_dirty(inode, dirty);
S
Sage Weil 已提交
993
	}
S
Sage Weil 已提交
994

S
Sage Weil 已提交
995
	dout("aio_write %p %llx.%llx %llu~%u  dropping cap refs on %s\n",
S
Sage Weil 已提交
996 997
	     inode, ceph_vinop(inode), pos, (unsigned)iov->iov_len,
	     ceph_cap_string(got));
S
Sage Weil 已提交
998
	ceph_put_cap_refs(ci, got);
S
Sage Weil 已提交
999

1000
	if (written >= 0 &&
Y
Yan, Zheng 已提交
1001 1002
	    ((file->f_flags & O_SYNC) || IS_SYNC(file->f_mapping->host) ||
	     ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_NEARFULL))) {
1003
		err = vfs_fsync_range(file, pos, pos + written - 1, 1);
Y
Yan, Zheng 已提交
1004
		if (err < 0)
1005
			written = err;
Y
Yan, Zheng 已提交
1006
	}
1007

1008 1009
	goto out_unlocked;

1010
out:
1011 1012
	mutex_unlock(&inode->i_mutex);
out_unlocked:
1013 1014
	current->backing_dev_info = NULL;
	return written ? written : err;
S
Sage Weil 已提交
1015 1016 1017 1018 1019
}

/*
 * llseek.  be sure to verify file size on SEEK_END.
 */
1020
static loff_t ceph_llseek(struct file *file, loff_t offset, int whence)
S
Sage Weil 已提交
1021 1022 1023 1024 1025
{
	struct inode *inode = file->f_mapping->host;
	int ret;

	mutex_lock(&inode->i_mutex);
1026

1027
	if (whence == SEEK_END || whence == SEEK_DATA || whence == SEEK_HOLE) {
S
Sage Weil 已提交
1028 1029 1030 1031 1032
		ret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE);
		if (ret < 0) {
			offset = ret;
			goto out;
		}
1033 1034
	}

1035
	switch (whence) {
1036
	case SEEK_END:
S
Sage Weil 已提交
1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051
		offset += inode->i_size;
		break;
	case SEEK_CUR:
		/*
		 * Here we special-case the lseek(fd, 0, SEEK_CUR)
		 * position-querying operation.  Avoid rewriting the "same"
		 * f_pos value back to the file because a concurrent read(),
		 * write() or lseek() might have altered it
		 */
		if (offset == 0) {
			offset = file->f_pos;
			goto out;
		}
		offset += file->f_pos;
		break;
1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
	case SEEK_DATA:
		if (offset >= inode->i_size) {
			ret = -ENXIO;
			goto out;
		}
		break;
	case SEEK_HOLE:
		if (offset >= inode->i_size) {
			ret = -ENXIO;
			goto out;
		}
		offset = inode->i_size;
		break;
S
Sage Weil 已提交
1065 1066
	}

J
Jie Liu 已提交
1067
	offset = vfs_setpos(file, offset, inode->i_sb->s_maxbytes);
S
Sage Weil 已提交
1068 1069 1070 1071 1072 1073

out:
	mutex_unlock(&inode->i_mutex);
	return offset;
}

L
Li Wang 已提交
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158
static inline void ceph_zero_partial_page(
	struct inode *inode, loff_t offset, unsigned size)
{
	struct page *page;
	pgoff_t index = offset >> PAGE_CACHE_SHIFT;

	page = find_lock_page(inode->i_mapping, index);
	if (page) {
		wait_on_page_writeback(page);
		zero_user(page, offset & (PAGE_CACHE_SIZE - 1), size);
		unlock_page(page);
		page_cache_release(page);
	}
}

static void ceph_zero_pagecache_range(struct inode *inode, loff_t offset,
				      loff_t length)
{
	loff_t nearly = round_up(offset, PAGE_CACHE_SIZE);
	if (offset < nearly) {
		loff_t size = nearly - offset;
		if (length < size)
			size = length;
		ceph_zero_partial_page(inode, offset, size);
		offset += size;
		length -= size;
	}
	if (length >= PAGE_CACHE_SIZE) {
		loff_t size = round_down(length, PAGE_CACHE_SIZE);
		truncate_pagecache_range(inode, offset, offset + size - 1);
		offset += size;
		length -= size;
	}
	if (length)
		ceph_zero_partial_page(inode, offset, length);
}

static int ceph_zero_partial_object(struct inode *inode,
				    loff_t offset, loff_t *length)
{
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
	struct ceph_osd_request *req;
	int ret = 0;
	loff_t zero = 0;
	int op;

	if (!length) {
		op = offset ? CEPH_OSD_OP_DELETE : CEPH_OSD_OP_TRUNCATE;
		length = &zero;
	} else {
		op = CEPH_OSD_OP_ZERO;
	}

	req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout,
					ceph_vino(inode),
					offset, length,
					1, op,
					CEPH_OSD_FLAG_WRITE |
					CEPH_OSD_FLAG_ONDISK,
					NULL, 0, 0, false);
	if (IS_ERR(req)) {
		ret = PTR_ERR(req);
		goto out;
	}

	ceph_osdc_build_request(req, offset, NULL, ceph_vino(inode).snap,
				&inode->i_mtime);

	ret = ceph_osdc_start_request(&fsc->client->osdc, req, false);
	if (!ret) {
		ret = ceph_osdc_wait_request(&fsc->client->osdc, req);
		if (ret == -ENOENT)
			ret = 0;
	}
	ceph_osdc_put_request(req);

out:
	return ret;
}

static int ceph_zero_objects(struct inode *inode, loff_t offset, loff_t length)
{
	int ret = 0;
	struct ceph_inode_info *ci = ceph_inode(inode);
S
Sage Weil 已提交
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
	s32 stripe_unit = ceph_file_layout_su(ci->i_layout);
	s32 stripe_count = ceph_file_layout_stripe_count(ci->i_layout);
	s32 object_size = ceph_file_layout_object_size(ci->i_layout);
	u64 object_set_size = object_size * stripe_count;
	u64 nearly, t;

	/* round offset up to next period boundary */
	nearly = offset + object_set_size - 1;
	t = nearly;
	nearly -= do_div(t, object_set_size);
L
Li Wang 已提交
1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204

	while (length && offset < nearly) {
		loff_t size = length;
		ret = ceph_zero_partial_object(inode, offset, &size);
		if (ret < 0)
			return ret;
		offset += size;
		length -= size;
	}
	while (length >= object_set_size) {
		int i;
		loff_t pos = offset;
		for (i = 0; i < stripe_count; ++i) {
			ret = ceph_zero_partial_object(inode, pos, NULL);
			if (ret < 0)
				return ret;
			pos += stripe_unit;
		}
		offset += object_set_size;
		length -= object_set_size;
	}
	while (length) {
		loff_t size = length;
		ret = ceph_zero_partial_object(inode, offset, &size);
		if (ret < 0)
			return ret;
		offset += size;
		length -= size;
	}
	return ret;
}

static long ceph_fallocate(struct file *file, int mode,
				loff_t offset, loff_t length)
{
	struct ceph_file_info *fi = file->private_data;
L
Libo Chen 已提交
1205
	struct inode *inode = file_inode(file);
L
Li Wang 已提交
1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271
	struct ceph_inode_info *ci = ceph_inode(inode);
	struct ceph_osd_client *osdc =
		&ceph_inode_to_client(inode)->client->osdc;
	int want, got = 0;
	int dirty;
	int ret = 0;
	loff_t endoff = 0;
	loff_t size;

	if (!S_ISREG(inode->i_mode))
		return -EOPNOTSUPP;

	if (IS_SWAPFILE(inode))
		return -ETXTBSY;

	mutex_lock(&inode->i_mutex);

	if (ceph_snap(inode) != CEPH_NOSNAP) {
		ret = -EROFS;
		goto unlock;
	}

	if (ceph_osdmap_flag(osdc->osdmap, CEPH_OSDMAP_FULL) &&
		!(mode & FALLOC_FL_PUNCH_HOLE)) {
		ret = -ENOSPC;
		goto unlock;
	}

	size = i_size_read(inode);
	if (!(mode & FALLOC_FL_KEEP_SIZE))
		endoff = offset + length;

	if (fi->fmode & CEPH_FILE_MODE_LAZY)
		want = CEPH_CAP_FILE_BUFFER | CEPH_CAP_FILE_LAZYIO;
	else
		want = CEPH_CAP_FILE_BUFFER;

	ret = ceph_get_caps(ci, CEPH_CAP_FILE_WR, want, &got, endoff);
	if (ret < 0)
		goto unlock;

	if (mode & FALLOC_FL_PUNCH_HOLE) {
		if (offset < size)
			ceph_zero_pagecache_range(inode, offset, length);
		ret = ceph_zero_objects(inode, offset, length);
	} else if (endoff > size) {
		truncate_pagecache_range(inode, size, -1);
		if (ceph_inode_set_size(inode, endoff))
			ceph_check_caps(ceph_inode(inode),
				CHECK_CAPS_AUTHONLY, NULL);
	}

	if (!ret) {
		spin_lock(&ci->i_ceph_lock);
		dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR);
		spin_unlock(&ci->i_ceph_lock);
		if (dirty)
			__mark_inode_dirty(inode, dirty);
	}

	ceph_put_cap_refs(ci, got);
unlock:
	mutex_unlock(&inode->i_mutex);
	return ret;
}

S
Sage Weil 已提交
1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
const struct file_operations ceph_file_fops = {
	.open = ceph_open,
	.release = ceph_release,
	.llseek = ceph_llseek,
	.read = do_sync_read,
	.write = do_sync_write,
	.aio_read = ceph_aio_read,
	.aio_write = ceph_aio_write,
	.mmap = ceph_mmap,
	.fsync = ceph_fsync,
G
Greg Farnum 已提交
1282 1283
	.lock = ceph_lock,
	.flock = ceph_flock,
S
Sage Weil 已提交
1284 1285 1286 1287
	.splice_read = generic_file_splice_read,
	.splice_write = generic_file_splice_write,
	.unlocked_ioctl = ceph_ioctl,
	.compat_ioctl	= ceph_ioctl,
L
Li Wang 已提交
1288
	.fallocate	= ceph_fallocate,
S
Sage Weil 已提交
1289 1290
};