Merge branch 'compression' into for-next-next-v4.14-20170721

7 files changed, 72 insertions, 26 deletions

diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index a68fbf6d4bc0..219e971f8577 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -179,9 +179,14 @@ struct btrfs_inode {
 	unsigned reserved_extents;
 
 	/*
-	 * always compress this one file
+	 * Cached values if inode properties
 	 */
-	unsigned force_compress;
+	unsigned prop_compress;		/* per-file compression algorithm */
+	/*
+	 * Force compression on the file using the defrag ioctl, could be
+	 * different from prop_compress and takes precedence if set
+	 */
+	unsigned defrag_compress;
 
 	struct btrfs_delayed_node *delayed_node;
 
diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index 5065d611a300..63f54bd2d5bb 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -1047,3 +1047,36 @@ int btrfs_decompress_buf2page(const char *buf, unsigned long buf_start,
 
 	return 1;
 }
+
+/*
+ * Compression heuristic.
+ *
+ * For now is's a naive and optimistic 'return true', we'll extend the logic to
+ * quickly (compared to direct compression) detect data characteristics
+ * (compressible/uncompressible) to avoid wasting CPU time on uncompressible
+ * data.
+ *
+ * The following types of analysis can be performed:
+ * - detect mostly zero data
+ * - detect data with low "byte set" size (text, etc)
+ * - detect data with low/high "core byte" set
+ *
+ * Return non-zero if the compression should be done, 0 otherwise.
+ */
+int btrfs_compress_heuristic(struct inode *inode, u64 start, u64 end)
+{
+	u64 index = start >> PAGE_SHIFT;
+	u64 end_index = end >> PAGE_SHIFT;
+	struct page *page;
+	int ret = 1;
+
+	while (index <= end_index) {
+		page = find_get_page(inode->i_mapping, index);
+		kmap(page);
+		kunmap(page);
+		put_page(page);
+		index++;
+	}
+
+	return ret;
+}
diff --git a/fs/btrfs/compression.h b/fs/btrfs/compression.h
index 89bcf975efb8..d1f4eee2d0af 100644
--- a/fs/btrfs/compression.h
+++ b/fs/btrfs/compression.h
@@ -129,4 +129,6 @@ struct btrfs_compress_op {
 extern const struct btrfs_compress_op btrfs_zlib_compress;
 extern const struct btrfs_compress_op btrfs_lzo_compress;
 
+int btrfs_compress_heuristic(struct inode *inode, u64 start, u64 end);
+
 #endif
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 04ee3720afc4..9a4413a176c8 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -392,20 +392,23 @@ static noinline int add_async_extent(struct async_cow *cow,
 	return 0;
 }
 
-static inline int inode_need_compress(struct inode *inode)
+static inline int inode_need_compress(struct inode *inode, u64 start, u64 end)
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 
 	/* force compress */
 	if (btrfs_test_opt(fs_info, FORCE_COMPRESS))
 		return 1;
+	/* defrag ioctl */
+	if (BTRFS_I(inode)->defrag_compress)
+		return 1;
 	/* bad compression ratios */
 	if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS)
 		return 0;
 	if (btrfs_test_opt(fs_info, COMPRESS) ||
 	    BTRFS_I(inode)->flags & BTRFS_INODE_COMPRESS ||
-	    BTRFS_I(inode)->force_compress)
-		return 1;
+	    BTRFS_I(inode)->prop_compress)
+		return btrfs_compress_heuristic(inode, start, end);
 	return 0;
 }
 
@@ -503,7 +506,7 @@ again:
 	 * inode has not been flagged as nocompress.  This flag can
 	 * change at any time if we discover bad compression ratios.
 	 */
-	if (inode_need_compress(inode)) {
+	if (inode_need_compress(inode, start, end)) {
 		WARN_ON(pages);
 		pages = kcalloc(nr_pages, sizeof(struct page *), GFP_NOFS);
 		if (!pages) {
@@ -511,8 +514,10 @@ again:
 			goto cont;
 		}
 
-		if (BTRFS_I(inode)->force_compress)
-			compress_type = BTRFS_I(inode)->force_compress;
+		if (BTRFS_I(inode)->defrag_compress)
+			compress_type = BTRFS_I(inode)->defrag_compress;
+		else if (BTRFS_I(inode)->prop_compress)
+			compress_type = BTRFS_I(inode)->prop_compress;
 
 		/*
 		 * we need to call clear_page_dirty_for_io on each
@@ -645,7 +650,7 @@ cont:
 
 		/* flag the file so we don't compress in the future */
 		if (!btrfs_test_opt(fs_info, FORCE_COMPRESS) &&
-		    !(BTRFS_I(inode)->force_compress)) {
+		    !(BTRFS_I(inode)->prop_compress)) {
 			BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
 		}
 	}
@@ -1577,7 +1582,7 @@ static int run_delalloc_range(void *private_data, struct page *locked_page,
 	} else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC && !force_cow) {
 		ret = run_delalloc_nocow(inode, locked_page, start, end,
 					 page_started, 0, nr_written);
-	} else if (!inode_need_compress(inode)) {
+	} else if (!inode_need_compress(inode, start, end)) {
 		ret = cow_file_range(inode, locked_page, start, end, end,
 				      page_started, nr_written, 1, NULL);
 	} else {
@@ -9431,7 +9436,8 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 	ei->reserved_extents = 0;
 
 	ei->runtime_flags = 0;
-	ei->force_compress = BTRFS_COMPRESS_NONE;
+	ei->prop_compress = BTRFS_COMPRESS_NONE;
+	ei->defrag_compress = BTRFS_COMPRESS_NONE;
 
 	ei->delayed_node = NULL;
 
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 496e873b14de..bedeec66ca8b 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1255,6 +1255,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	unsigned long cluster = max_cluster;
 	u64 new_align = ~((u64)SZ_128K - 1);
 	struct page **pages = NULL;
+	bool do_compress = range->flags & BTRFS_DEFRAG_RANGE_COMPRESS;
 
 	if (isize == 0)
 		return 0;
@@ -1262,7 +1263,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	if (range->start >= isize)
 		return -EINVAL;
 
-	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
+	if (do_compress) {
 		if (range->compress_type > BTRFS_COMPRESS_TYPES)
 			return -EINVAL;
 		if (range->compress_type)
@@ -1341,8 +1342,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 
 		if (!should_defrag_range(inode, (u64)i << PAGE_SHIFT,
 					 extent_thresh, &last_len, &skip,
-					 &defrag_end, range->flags &
-					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
+					 &defrag_end, do_compress)){
 			unsigned long next;
 			/*
 			 * the should_defrag function tells us how much to skip
@@ -1370,8 +1370,8 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 		}
 
 		inode_lock(inode);
-		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
-			BTRFS_I(inode)->force_compress = compress_type;
+		if (do_compress)
+			BTRFS_I(inode)->defrag_compress = compress_type;
 		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
 		if (ret < 0) {
 			inode_unlock(inode);
@@ -1418,7 +1418,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 			filemap_flush(inode->i_mapping);
 	}
 
-	if ((range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)) {
+	if (do_compress) {
 		/* the filemap_flush will queue IO into the worker threads, but
 		 * we have to make sure the IO is actually started and that
 		 * ordered extents get created before we return
@@ -1440,9 +1440,9 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 	ret = defrag_count;
 
 out_ra:
-	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
+	if (do_compress) {
 		inode_lock(inode);
-		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
+		BTRFS_I(inode)->defrag_compress = BTRFS_COMPRESS_NONE;
 		inode_unlock(inode);
 	}
 	if (!file)
diff --git a/fs/btrfs/props.c b/fs/btrfs/props.c
index 4b23ae5d0e5c..09c0266f248d 100644
--- a/fs/btrfs/props.c
+++ b/fs/btrfs/props.c
@@ -403,28 +403,28 @@ static int prop_compression_apply(struct inode *inode,
 	if (len == 0) {
 		BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
 		BTRFS_I(inode)->flags &= ~BTRFS_INODE_COMPRESS;
-		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
+		BTRFS_I(inode)->prop_compress = BTRFS_COMPRESS_NONE;
 
 		return 0;
 	}
 
-	if (!strncmp("lzo", value, len))
+	if (!strncmp("lzo", value, 3))
 		type = BTRFS_COMPRESS_LZO;
-	else if (!strncmp("zlib", value, len))
+	else if (!strncmp("zlib", value, 4))
 		type = BTRFS_COMPRESS_ZLIB;
 	else
 		return -EINVAL;
 
 	BTRFS_I(inode)->flags &= ~BTRFS_INODE_NOCOMPRESS;
 	BTRFS_I(inode)->flags |= BTRFS_INODE_COMPRESS;
-	BTRFS_I(inode)->force_compress = type;
+	BTRFS_I(inode)->prop_compress = type;
 
 	return 0;
 }
 
 static const char *prop_compression_extract(struct inode *inode)
 {
-	switch (BTRFS_I(inode)->force_compress) {
+	switch (BTRFS_I(inode)->prop_compress) {
 	case BTRFS_COMPRESS_ZLIB:
 		return "zlib";
 	case BTRFS_COMPRESS_LZO:
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index eb0aeb81d576..ed8ce5edc9cc 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -499,14 +499,14 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
 				btrfs_test_opt(info, FORCE_COMPRESS);
 			if (token == Opt_compress ||
 			    token == Opt_compress_force ||
-			    strcmp(args[0].from, "zlib") == 0) {
+			    strncmp(args[0].from, "zlib", 4) == 0) {
 				compress_type = "zlib";
 				info->compress_type = BTRFS_COMPRESS_ZLIB;
 				btrfs_set_opt(info->mount_opt, COMPRESS);
 				btrfs_clear_opt(info->mount_opt, NODATACOW);
 				btrfs_clear_opt(info->mount_opt, NODATASUM);
 				no_compress = 0;
-			} else if (strcmp(args[0].from, "lzo") == 0) {
+			} else if (strncmp(args[0].from, "lzo", 3) == 0) {
 				compress_type = "lzo";
 				info->compress_type = BTRFS_COMPRESS_LZO;
 				btrfs_set_opt(info->mount_opt, COMPRESS);