btrfs-progs: check: add duplicate filename check

If when doing btrfs check we encounter a DIR_ITEM slot with multiple
entries due to a hash collision, loop through them to make sure that
they are all distinct names.

Pull-request: #1021
Signed-off-by: Mark Harmstone <[email protected]>
Signed-off-by: David Sterba <[email protected]>

Author: maharmstone

check/main.c

@@ -653,6 +653,8 @@ static void print_inode_error(struct btrfs_root *root, struct inode_record *rec)
 			rec->nlink);
 	if (errors & I_ERR_INVALID_XATTR)
 		fprintf(stderr, ", invalid xattr");
+	if (errors & I_ERR_DUP_FILENAME)
+		fprintf(stderr, ", dup filename");
 	fprintf(stderr, "\n");
 
 	/* Print the holes if needed */
@@ -1434,6 +1436,40 @@ static int add_mismatch_dir_hash(struct inode_record *dir_rec,
 	return 0;
 }
 
+static void check_for_dupe_filenames(struct extent_buffer *eb, int slot,
+				     int nritems, struct inode_record *rec)
+{
+	struct btrfs_dir_item *di, *di2;
+	char namebuf[BTRFS_NAME_LEN], namebuf2[BTRFS_NAME_LEN];
+	u32 len, len2;
+
+	di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
+
+	for (int i = 0; i < nritems - 1; i++) {
+		len = btrfs_dir_name_len(eb, di) + btrfs_dir_data_len(eb, di);
+
+		read_extent_buffer(eb, namebuf, (unsigned long)(di + 1), len);
+
+		di2 = di;
+		len2 = len;
+
+		for (int j = i + 1; j < nritems; j++) {
+			di2 = (struct btrfs_dir_item *)((char *)di2 + sizeof(*di2) + len2);
+			len2 = btrfs_dir_name_len(eb, di2) + btrfs_dir_data_len(eb, di2);
+
+			if (len != len2)
+				continue;
+
+			read_extent_buffer(eb, namebuf2, (unsigned long)(di2 + 1), len2);
+
+			if (memcmp(namebuf, namebuf2, len) == 0)
+				rec->errors |= I_ERR_DUP_FILENAME;
+		}
+
+		di = (struct btrfs_dir_item *)((char *)di + sizeof(*di) + len);
+	}
+}
+
 static int process_dir_item(struct extent_buffer *eb,
 			    int slot, struct btrfs_key *key,
 			    struct shared_node *active_node)
@@ -1525,6 +1561,9 @@ static int process_dir_item(struct extent_buffer *eb,
 	if (key->type == BTRFS_DIR_INDEX_KEY && nritems > 1)
 		rec->errors |= I_ERR_DUP_DIR_INDEX;
 
+	if (key->type == BTRFS_DIR_ITEM_KEY && nritems > 1)
+		check_for_dupe_filenames(eb, slot, nritems, rec);
+
 	return 0;
 }
 

check/mode-lowmem.c

@@ -1661,6 +1661,47 @@ static void print_dir_item_err(struct btrfs_root *root, struct btrfs_key *key,
 
 }
 
+static bool check_for_dupe_filenames_lowmem(struct extent_buffer *eb, int slot,
+					    int nritems, struct btrfs_root *root,
+					    struct btrfs_key *di_key)
+{
+	struct btrfs_dir_item *di, *di2;
+	char namebuf[BTRFS_NAME_LEN], namebuf2[BTRFS_NAME_LEN];
+	u32 len, len2;
+
+	di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
+
+	for (int i = 0; i < nritems - 1; i++) {
+		len = btrfs_dir_name_len(eb, di) + btrfs_dir_data_len(eb, di);
+
+		read_extent_buffer(eb, namebuf, (unsigned long)(di + 1), len);
+
+		di2 = di;
+		len2 = len;
+
+		for (int j = i + 1; j < nritems; j++) {
+			di2 = (struct btrfs_dir_item *)((char *)di2 + sizeof(*di2) + len2);
+			len2 = btrfs_dir_name_len(eb, di2) + btrfs_dir_data_len(eb, di2);
+
+			if (len != len2)
+				continue;
+
+			read_extent_buffer(eb, namebuf2, (unsigned long)(di2 + 1), len2);
+
+			if (!memcmp(namebuf, namebuf2, len)) {
+				error("root %llu inode %llu dup filename %.*s",
+				      root->objectid, di_key->objectid,
+				      len, namebuf);
+				return true;
+			}
+		}
+
+		di = (struct btrfs_dir_item *)((char *)di + sizeof(*di) + len);
+	}
+
+	return false;
+}
+
 /*
  * Traverse the given DIR_ITEM/DIR_INDEX and check related INODE_ITEM and
  * call find_inode_ref() to check related INODE_REF/INODE_EXTREF.
@@ -1695,6 +1736,7 @@ static int check_dir_item(struct btrfs_root *root, struct btrfs_key *di_key,
 	int err;
 	int tmp_err;
 	int need_research = 0;
+	int nritems = 1;
 
 begin:
 	err = 0;
@@ -1835,7 +1877,16 @@ static int check_dir_item(struct btrfs_root *root, struct btrfs_key *di_key,
 			      di_key->offset);
 			break;
 		}
+
+		if (cur < total)
+			nritems++;
 	}
+
+	if (nritems > 1) {
+		if (check_for_dupe_filenames_lowmem(node, slot, nritems, root, di_key))
+			err |= DUP_FILENAME_ERROR;
+	}
+
 out:
 	/* research path */
 	btrfs_release_path(path);

check/mode-lowmem.h

@@ -47,6 +47,7 @@
 #define INODE_MODE_ERROR	(1U << 25)	/* Bad inode mode */
 #define INVALID_GENERATION	(1U << 26)	/* Generation is too new */
 #define SUPER_BYTES_USED_ERROR	(1U << 27)	/* Super bytes_used is invalid */
+#define DUP_FILENAME_ERROR	(1U << 28)	/* DIR_ITEM contains duplicate names */
 
 /*
  * Error bit for low memory mode check.

check/mode-original.h

@@ -195,6 +195,7 @@ struct unaligned_extent_rec_t {
 #define I_ERR_INVALID_NLINK		(1U << 21)
 #define I_ERR_INVALID_XATTR		(1U << 22)
 #define I_ERR_DEPRECATED_FREE_INO	(1U << 23)
+#define I_ERR_DUP_FILENAME		(1U << 24)
 
 struct inode_record {
 	struct list_head backrefs;

tests/fsck-tests/067-dupe-filename/default.img.xz

@@ -0,0 +1,14 @@
+#!/bin/bash
+#
+# Verify that check can report duplicate filename as an error
+
+source "$TEST_TOP/common" || exit
+
+check_prereq btrfs
+
+check_image() {
+	run_mustfail "duplicate filename not reported as error" \
+		"$TOP/btrfs" check "$1"
+}
+
+check_all_images