btrfs: implement swap file support

index 1bf9f89..21b7eca 100644
@@ -502,7 +502,7 @@ static int btree_read_extent_buffer_pages(struct btrfs_root *root,
 
 static int csum_dirty_buffer(struct btrfs_root *root, struct page *page)
 {
-	u64 start = page_offset(page);
+	u64 start = page_file_offset(page);
 	u64 found_start;
 	struct extent_buffer *eb;
 

@@ -607,7 +607,7 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 	u64 found_start;
 	int found_level;
 	struct extent_buffer *eb;
-	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
+	struct btrfs_root *root = BTRFS_I(page_file_mapping(page)->host)->root;
 	int ret = 0;
 	int reads_done;
 

@@ -696,7 +696,7 @@ out:
 static int btree_io_failed_hook(struct page *page, int failed_mirror)
 {
 	struct extent_buffer *eb;
-	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
+	struct btrfs_root *root = BTRFS_I(page_file_mapping(page)->host)->root;
 
 	eb = (struct extent_buffer *)page->private;
 	set_bit(EXTENT_BUFFER_READ_ERR, &eb->bflags);

@@ -880,7 +880,7 @@ static int btree_csum_one_bio(struct bio *bio)
 	int i, ret = 0;
 
 	bio_for_each_segment_all(bvec, bio, i) {
-		root = BTRFS_I(bvec->bv_page->mapping->host)->root;
+		root = BTRFS_I(page_file_mapping(bvec->bv_page)->host)->root;
 		ret = csum_dirty_buffer(root, bvec->bv_page);
 		if (ret)
 			break;

@@ -1018,7 +1018,7 @@ static int btree_writepages(struct address_space *mapping,
 static int btree_readpage(struct file *file, struct page *page)
 {
 	struct extent_io_tree *tree;
-	tree = &BTRFS_I(page->mapping->host)->io_tree;
+	tree = &BTRFS_I(page_file_mapping(page)->host)->io_tree;
 	return extent_read_full_page(tree, page, btree_get_extent, 0);
 }
 

@@ -1034,13 +1034,13 @@ static void btree_invalidatepage(struct page *page, unsigned int offset,
 				 unsigned int length)
 {
 	struct extent_io_tree *tree;
-	tree = &BTRFS_I(page->mapping->host)->io_tree;
+	tree = &BTRFS_I(page_file_mapping(page)->host)->io_tree;
 	extent_invalidatepage(tree, page, offset);
 	btree_releasepage(page, GFP_NOFS);
 	if (PagePrivate(page)) {
-		btrfs_warn(BTRFS_I(page->mapping->host)->root->fs_info,
+		btrfs_warn(BTRFS_I(page_file_mapping(page)->host)->root->fs_info,
 			   "page private not zero on page %llu",
-			   (unsigned long long)page_offset(page));
+			   (unsigned long long)page_file_offset(page));
 		ClearPagePrivate(page);
 		set_page_private(page, 0);
 		page_cache_release(page);

376 lines not on this page · raw message

index 84a2d18..ba11623 100644
@@ -222,7 +222,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root,
 		offset = logical_offset;
 	while (bio_index < bio->bi_vcnt) {
 		if (!dio)
-			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
+			offset = page_file_offset(bvec->bv_page) + bvec->bv_offset;
 		count = btrfs_find_ordered_sum(inode, offset, disk_bytenr,
 					       (u32 *)csum, nblocks);
 		if (count)

@@ -448,7 +448,7 @@ int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
 	if (contig)
 		offset = file_start;
 	else
-		offset = page_offset(bvec->bv_page) + bvec->bv_offset;
+		offset = page_file_offset(bvec->bv_page) + bvec->bv_offset;
 
 	ordered = btrfs_lookup_ordered_extent(inode, offset);
 	BUG_ON(!ordered); /* Logic error */

@@ -457,7 +457,7 @@ int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
 
 	while (bio_index < bio->bi_vcnt) {
 		if (!contig)
-			offset = page_offset(bvec->bv_page) + bvec->bv_offset;
+			offset = page_file_offset(bvec->bv_page) + bvec->bv_offset;
 
 		if (offset >= ordered->file_offset + ordered->len ||
 		    offset < ordered->file_offset) {

152 lines not on this page · raw message

index 4399f0c..e3b458a 100644
@@ -1151,7 +1151,7 @@ again:
 		if (!page)
 			break;
 
-		page_start = page_offset(page);
+		page_start = page_file_offset(page);
 		page_end = page_start + PAGE_CACHE_SIZE - 1;
 		while (1) {
 			lock_extent_bits(tree, page_start, page_end,

@@ -1171,7 +1171,7 @@ again:
 			 * we unlocked the page above, so we need check if
 			 * it was released or not.
 			 */
-			if (page->mapping != inode->i_mapping) {
+			if (page_file_mapping(page) != inode->i_mapping) {
 				unlock_page(page);
 				page_cache_release(page);
 				goto again;

@@ -1189,7 +1189,7 @@ again:
 			}
 		}
 
-		if (page->mapping != inode->i_mapping) {
+		if (page_file_mapping(page) != inode->i_mapping) {
 			unlock_page(page);
 			page_cache_release(page);
 			goto again;

@@ -1211,8 +1211,8 @@ again:
 	for (i = 0; i < i_done; i++)
 		wait_on_page_writeback(pages[i]);
 
-	page_start = page_offset(pages[0]);
-	page_end = page_offset(pages[i_done - 1]) + PAGE_CACHE_SIZE;
+	page_start = page_file_offset(pages[0]);
+	page_end = page_file_offset(pages[i_done - 1]) + PAGE_CACHE_SIZE;
 
 	lock_extent_bits(&BTRFS_I(inode)->io_tree,
 			 page_start, page_end - 1, 0, &cached_state);

index 74257d6..ad1abd3 100644
@@ -3163,7 +3163,7 @@ static int relocate_file_extent_cluster(struct inode *inode,
 			}
 		}
 
-		page_start = page_offset(page);
+		page_start = page_file_offset(page);
 		page_end = page_start + PAGE_CACHE_SIZE - 1;
 
 		lock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end);

index efa0831..ed7e275 100644
@@ -668,7 +668,7 @@ static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx)
 		}
 		ret = repair_io_failure(inode, offset, PAGE_SIZE,
 					fixup->logical, page,
-					offset - page_offset(page),
+					offset - page_file_offset(page),
 					fixup->mirror_num);
 		unlock_page(page);
 		corrected = !ret;

@@ -3411,7 +3411,7 @@ again:
 			 * old one, the new data may be written into the new
 			 * page in the page cache.
 			 */
-			if (page->mapping != inode->i_mapping) {
+			if (page_file_mapping(page) != inode->i_mapping) {
 				unlock_page(page);
 				page_cache_release(page);
 				goto again;

index e3b458a..7aee8cf 100644
@@ -293,14 +293,21 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 		}
 	} else {
 		/*
-		 * Revert back under same assuptions as above
+		 * swap_activate checks that we don't swapon a copy-on-write
+		 * file, but we must also make sure that it doesn't become
+		 * copy-on-write.
 		 */
-		if (S_ISREG(mode)) {
-			if (inode->i_size == 0)
-				ip->flags &= ~(BTRFS_INODE_NODATACOW
-				             | BTRFS_INODE_NODATASUM);
-		} else {
-			ip->flags &= ~BTRFS_INODE_NODATACOW;
+		if (!IS_SWAPFILE(inode)) {
+			/*
+			 * Revert back under same assumptions as above
+			 */
+			if (S_ISREG(mode)) {
+				if (inode->i_size == 0)
+					ip->flags &= ~(BTRFS_INODE_NODATACOW |
+						       BTRFS_INODE_NODATASUM);
+			} else {
+				ip->flags &= ~BTRFS_INODE_NODATACOW;
+			}
 		}
 	}
 

@@ -317,20 +324,25 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 		if (ret && ret != -ENODATA)
 			goto out_drop;
 	} else if (flags & FS_COMPR_FL) {
-		const char *comp;
-
-		ip->flags |= BTRFS_INODE_COMPRESS;
-		ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
+		/*
+		 * Like nodatacow, swap_activate checks that we don't swapon a
+		 * compressed file, so we shouldn't let it become compressed.
+		 */
+		if (!IS_SWAPFILE(inode)) {
+			const char *comp;
 
-		if (root->fs_info->compress_type == BTRFS_COMPRESS_LZO)
-			comp = "lzo";
-		else
-			comp = "zlib";
-		ret = btrfs_set_prop(inode, "btrfs.compression",
-				     comp, strlen(comp), 0);
-		if (ret)
-			goto out_drop;
+			ip->flags |= BTRFS_INODE_COMPRESS;
+			ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
 
+			if (root->fs_info->compress_type == BTRFS_COMPRESS_LZO)
+				comp = "lzo";
+			else
+				comp = "zlib";
+			ret = btrfs_set_prop(inode, "btrfs.compression",
+					     comp, strlen(comp), 0);
+			if (ret)
+				goto out_drop;
+		}
 	} else {
 		ret = btrfs_set_prop(inode, "btrfs.compression", NULL, 0, 0);
 		if (ret && ret != -ENODATA)

index 9b67b37..54b2d00 100644
@@ -2824,6 +2824,7 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree,
 static void attach_extent_buffer_page(struct extent_buffer *eb,
 				      struct page *page)
 {
+	BUG_ON(PageSwapCache(page));
 	if (!PagePrivate(page)) {
 		SetPagePrivate(page);
 		page_cache_get(page);

@@ -2835,6 +2836,7 @@ static void attach_extent_buffer_page(struct extent_buffer *eb,
 
 void set_page_extent_mapped(struct page *page)
 {
+	BUG_ON(PageSwapCache(page));
 	if (!PagePrivate(page)) {
 		SetPagePrivate(page);
 		page_cache_get(page);

@@ -2903,7 +2905,8 @@ static int __do_readpage(struct extent_io_tree *tree,
 	size_t blocksize = inode->i_sb->s_blocksize;
 	unsigned long this_bio_flag = *bio_flags & EXTENT_BIO_PARENT_LOCKED;
 
-	set_page_extent_mapped(page);
+	if (likely(!PageSwapCache(page)))
+		set_page_extent_mapped(page);
 
 	end = page_end;
 	if (!PageUptodate(page)) {

index 54b2d00..b8dc256 100644
@@ -2904,13 +2904,14 @@ static int __do_readpage(struct extent_io_tree *tree,
 	size_t disk_io_size;
 	size_t blocksize = inode->i_sb->s_blocksize;
 	unsigned long this_bio_flag = *bio_flags & EXTENT_BIO_PARENT_LOCKED;
+	int swapcache = PageSwapCache(page);
 
-	if (likely(!PageSwapCache(page)))
+	if (likely(!swapcache))
 		set_page_extent_mapped(page);
 
 	end = page_end;
 	if (!PageUptodate(page)) {
-		if (cleancache_get_page(page) == 0) {
+		if (likely(!swapcache) && cleancache_get_page(page) == 0) {
 			BUG_ON(blocksize != PAGE_SIZE);
 			unlock_extent(tree, start, end);
 			goto out;

index b8dc256..ca696d5 100644
@@ -2496,12 +2496,12 @@ static void end_bio_extent_writepage(struct bio *bio, int err)
 
 static void
 endio_readpage_release_extent(struct extent_io_tree *tree, u64 start, u64 len,
-			      int uptodate)
+			      int uptodate, int swapcache)
 {
 	struct extent_state *cached = NULL;
 	u64 end = start + len - 1;
 
-	if (uptodate && tree->track_uptodate)
+	if (likely(!swapcache) && uptodate && tree->track_uptodate)
 		set_extent_uptodate(tree, start, end, &cached, GFP_ATOMIC);
 	unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC);
 }

@@ -2532,6 +2532,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 	int mirror;
 	int ret;
 	int i;
+	int swapcache = 0;
 
 	if (err)
 		uptodate = 0;

@@ -2539,6 +2540,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
 	bio_for_each_segment_all(bvec, bio, i) {
 		struct page *page = bvec->bv_page;
 		struct inode *inode = page_file_mapping(page)->host;
+		swapcache |= PageSwapCache(page);
 
 		pr_debug("end_bio_extent_readpage: bi_sector=%llu, err=%d, "
 			 "mirror=%u\n", (u64)bio->bi_iter.bi_sector, err,

@@ -2631,12 +2633,14 @@ readpage_ok:
 			if (extent_len) {
 				endio_readpage_release_extent(tree,
 							      extent_start,
-							      extent_len, 1);
+							      extent_len, 1,
+							      swapcache);
 				extent_start = 0;
 				extent_len = 0;
 			}
 			endio_readpage_release_extent(tree, start,
-						      end - start + 1, 0);
+						      end - start + 1, 0,
+						      swapcache);
 		} else if (!extent_len) {
 			extent_start = start;
 			extent_len = end + 1 - start;

@@ -2644,7 +2648,8 @@ readpage_ok:
 			extent_len += end + 1 - start;
 		} else {
 			endio_readpage_release_extent(tree, extent_start,
-						      extent_len, uptodate);
+						      extent_len, uptodate,
+						      swapcache);
 			extent_start = start;
 			extent_len = end + 1 - start;
 		}

@@ -2652,7 +2657,7 @@ readpage_ok:
 
 	if (extent_len)
 		endio_readpage_release_extent(tree, extent_start, extent_len,
-					      uptodate);
+					      uptodate, swapcache);
 	if (io_bio->end_io)
 		io_bio->end_io(io_bio, err);
 	bio_put(bio);

@@ -2942,8 +2947,10 @@ static int __do_readpage(struct extent_io_tree *tree,
 			memset(userpage + pg_offset, 0, iosize);
 			flush_dcache_page(page);
 			kunmap_atomic(userpage);
-			set_extent_uptodate(tree, cur, cur + iosize - 1,
-					    &cached, GFP_NOFS);
+			if (likely(!swapcache))
+				set_extent_uptodate(tree, cur,
+						    cur + iosize - 1,
+						    &cached, GFP_NOFS);
 			if (!parent_locked)
 				unlock_extent_cached(tree, cur,
 						     cur + iosize - 1,

@@ -2995,8 +3002,9 @@ static int __do_readpage(struct extent_io_tree *tree,
 			flush_dcache_page(page);
 			kunmap_atomic(userpage);
 
-			set_extent_uptodate(tree, cur, cur + iosize - 1,
-					    &cached, GFP_NOFS);
+			if (likely(!swapcache))
+				set_extent_uptodate(tree, cur, cur + iosize - 1,
+						    &cached, GFP_NOFS);
 			unlock_extent_cached(tree, cur, cur + iosize - 1,
 			                     &cached, GFP_NOFS);
 			cur = cur + iosize;

@@ -3006,6 +3014,7 @@ static int __do_readpage(struct extent_io_tree *tree,
 		/* the get_extent function already copied into the page */
 		if (test_range_bit(tree, cur, cur_end,
 				   EXTENT_UPTODATE, 1, NULL)) {
+			WARN_ON(swapcache);
 			check_page_uptodate(tree, page);
 			if (!parent_locked)
 				unlock_extent(tree, cur, cur + iosize - 1);

index 0e84316..c7cce4e 100644
@@ -9442,6 +9442,75 @@ out_inode:
 
 }
 
+static int btrfs_swap_activate(struct swap_info_struct *sis, struct file *file,
+			       sector_t *span)
+{
+	struct inode *inode = file_inode(file);
+	struct btrfs_inode *ip = BTRFS_I(inode);
+	int ret = 0;
+	u64 isize = inode->i_size;
+	struct extent_state *cached_state = NULL;
+	struct extent_map *em;
+	u64 start, len;
+
+	if (ip->flags & BTRFS_INODE_COMPRESS) {
+		/* Can't do direct I/O on a compressed file. */
+		pr_err("BTRFS: swapfile is compressed");
+		return -EINVAL;
+	}
+	if (!(ip->flags & BTRFS_INODE_NODATACOW)) {
+		/* The swap file can't be copy-on-write. */
+		pr_err("BTRFS: swapfile is copy-on-write");
+		return -EINVAL;
+	}
+
+	lock_extent_bits(&ip->io_tree, 0, isize - 1, 0, &cached_state);
+
+	/*
+	 * All of the extents must be allocated and support direct I/O. Inline
+	 * extents and compressed extents fall back to buffered I/O, so those
+	 * are no good.
+	 */
+	start = 0;
+	while (start < isize) {
+		len = isize - start;
+		em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
+		if (IS_ERR(em)) {
+			ret = PTR_ERR(em);
+			goto out;
+		}
+
+		if (test_bit(EXTENT_FLAG_VACANCY, &em->flags) ||
+		    em->block_start == EXTENT_MAP_HOLE) {
+			pr_err("BTRFS: swapfile has holes");
+			ret = -EINVAL;
+			goto out;
+		}
+		if (em->block_start == EXTENT_MAP_INLINE) {
+			pr_err("BTRFS: swapfile is inline");
+			ret = -EINVAL;
+			goto out;
+		}
+		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) {
+			pr_err("BTRFS: swapfile is compresed");
+			ret = -EINVAL;
+			goto out;
+		}
+
+		start = extent_map_end(em);
+		free_extent_map(em);
+	}
+
+out:
+	unlock_extent_cached(&ip->io_tree, 0, isize - 1, &cached_state,
+			     GFP_NOFS);
+	return ret;
+}
+
+static void btrfs_swap_deactivate(struct file *file)
+{
+}
+
 static const struct inode_operations btrfs_dir_inode_operations = {
 	.getattr	= btrfs_getattr,
 	.lookup		= btrfs_lookup,

@@ -9519,6 +9588,8 @@ static const struct address_space_operations btrfs_aops = {
 	.releasepage	= btrfs_releasepage,
 	.set_page_dirty	= btrfs_set_page_dirty,
 	.error_remove_page = generic_error_remove_page,
+	.swap_activate	= btrfs_swap_activate,
+	.swap_deactivate = btrfs_swap_deactivate,
 };
 
 static const struct address_space_operations btrfs_symlink_aops = {

index e181b6b..e542ce4 100644
@@ -120,6 +120,7 @@ struct dio {
 	spinlock_t bio_lock;		/* protects BIO fields below */
 	int page_errors;		/* errno from get_user_pages() */
 	int is_async;			/* is IO async ? */
+	int should_dirty;		/* should we mark read pages dirty? */
 	bool defer_completion;		/* defer AIO completion to workqueue? */
 	int io_error;			/* IO error in completion path */
 	unsigned long refcount;		/* direct_io_worker() and bios */

@@ -392,7 +393,7 @@ static inline void dio_bio_submit(struct dio *dio, struct dio_submit *sdio)
 	dio->refcount++;
 	spin_unlock_irqrestore(&dio->bio_lock, flags);
 
-	if (dio->is_async && dio->rw == READ)
+	if (dio->is_async && dio->rw == READ && dio->should_dirty)
 		bio_set_pages_dirty(bio);
 
 	if (sdio->submit_io)

@@ -463,13 +464,13 @@ static int dio_bio_complete(struct dio *dio, struct bio *bio)
 	if (!uptodate)
 		dio->io_error = -EIO;
 
-	if (dio->is_async && dio->rw == READ) {
+	if (dio->is_async && dio->rw == READ && dio->should_dirty) {
 		bio_check_pages_dirty(bio);	/* transfers ownership */
 	} else {
 		bio_for_each_segment_all(bvec, bio, i) {
 			struct page *page = bvec->bv_page;
 
-			if (dio->rw == READ && !PageCompound(page))
+			if (dio->rw == READ && !PageCompound(page) && dio->should_dirty)
 				set_page_dirty_lock(page);
 			page_cache_release(page);
 		}

@@ -1177,6 +1178,7 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 
 	dio->inode = inode;
 	dio->rw = rw;
+	dio->should_dirty = !(iter->type & ITER_BVEC);
 
 	/*
 	 * For AIO O_(D)SYNC writes we need to defer completions to a workqueue

index 955db8b..b9b84b2 100644
@@ -266,8 +266,8 @@ int __swap_writepage(struct page *page, struct writeback_control *wbc,
 		struct address_space *mapping = swap_file->f_mapping;
 		struct bio_vec bv = {
 			.bv_page = page,
-			.bv_len  = PAGE_SIZE,
-			.bv_offset = 0
+			.bv_len = PAGE_SIZE,
+			.bv_offset = 0,
 		};
 		struct iov_iter from = {
 			.type = ITER_BVEC | WRITE,

@@ -283,8 +283,7 @@ int __swap_writepage(struct page *page, struct writeback_control *wbc,
 
 		set_page_writeback(page);
 		unlock_page(page);
-		ret = mapping->a_ops->direct_IO(ITER_BVEC | WRITE,
-						&kiocb, &from,
+		ret = mapping->a_ops->direct_IO(WRITE, &kiocb, &from,
 						kiocb.ki_pos);
 		if (ret == PAGE_SIZE) {
 			count_vm_event(PSWPOUT);

@@ -303,7 +302,7 @@ int __swap_writepage(struct page *page, struct writeback_control *wbc,
 			set_page_dirty(page);
 			ClearPageReclaim(page);
 			pr_err_ratelimited("Write error on dio swapfile (%Lu)\n",
-				page_file_offset(page));
+					   page_file_offset(page));
 		}
 		end_page_writeback(page);
 		return ret;

@@ -348,12 +347,36 @@ int swap_readpage(struct page *page)
 	}
 
 	if (sis->flags & SWP_FILE) {
+		struct kiocb kiocb;
 		struct file *swap_file = sis->swap_file;
 		struct address_space *mapping = swap_file->f_mapping;
+		struct bio_vec bv = {
+			.bv_page = page,
+			.bv_len = PAGE_SIZE,
+			.bv_offset = 0,
+		};
+		struct iov_iter to = {
+			.type = ITER_BVEC | READ,
+			.count = PAGE_SIZE,
+			.iov_offset = 0,
+			.nr_segs = 1,
+		};
+		to.bvec = &bv;	/* older gcc versions are broken */
+
+		init_sync_kiocb(&kiocb, swap_file);
+		kiocb.ki_pos = page_file_offset(page);
+		kiocb.ki_nbytes = PAGE_SIZE;
 
-		ret = mapping->a_ops->readpage(swap_file, page);
-		if (!ret)
+		ret = mapping->a_ops->direct_IO(READ, &kiocb, &to,
+						kiocb.ki_pos);
+		if (ret == PAGE_SIZE) {
+			SetPageUptodate(page);
 			count_vm_event(PSWPIN);
+			ret = 0;
+		} else {
+			PageError(page); /* XXX: maybe? */
+		}
+		unlock_page(page);
 		return ret;
 	}