[PATCH v2 0/2] Reconcile Encryption and Casefolding in Ext4

[PATCH v2 0/2] Reconcile Encryption and Casefolding in Ext4

[Daniel Rosenberg]

These patches add support for ext4 encryption and casefolding at the same time.
Since the hash for encrypted casefolded directory names cannot be computed
without the key, we need to store the hash on disk. We only do so for encrypted
and casefolded directories to avoid on disk format changes.

e2fsprogs has already been updated with support for casefolding and encryption.

v2 changes:
	When checking for 'fake' entries (which do not include the extra hash bytes)
	-Check for . and .. using names instead of position
	-Check for csum entries via file_type instead of position
	-Assume last entry in directory will be csum for __ext4_check_entry if csum enabled
	
	This means we don't need to pass along lblk all over the place
	
	-Don't use siphash value for find_group_orlov, just use regular hash

Daniel Rosenberg (2):
  ext4: Handle casefolding with encryption
  ext4: Optimize match for casefolded encrypted dirs

 Documentation/filesystems/ext4/directory.rst |  27 +++
 fs/ext4/dir.c                                |  37 +++-
 fs/ext4/ext4.h                               |  73 +++++--
 fs/ext4/hash.c                               |  25 ++-
 fs/ext4/inline.c                             |  25 ++-
 fs/ext4/namei.c                              | 213 ++++++++++++++-----
 fs/ext4/super.c                              |   6 -
 7 files changed, 303 insertions(+), 103 deletions(-)


base-commit: f296bfd5cd04cbb49b8fc9585adc280ab2b58624
-- 
2.31.0.rc2.261.g7f71774620-goog

[PATCH v2 1/2] ext4: Handle casefolding with encryption

[Daniel Rosenberg]

This adds support for encryption with casefolding.

Since the name on disk is case preserving, and also encrypted, we can no
longer just recompute the hash on the fly. Additionally, to avoid
leaking extra information from the hash of the unencrypted name, we use
siphash via an fscrypt v2 policy.

The hash is stored at the end of the directory entry for all entries
inside of an encrypted and casefolded directory apart from those that
deal with '.' and '..'. This way, the change is backwards compatible
with existing ext4 filesystems.

Signed-off-by: Daniel Rosenberg <drosen@google.com>
---
 Documentation/filesystems/ext4/directory.rst |  27 +++
 fs/ext4/dir.c                                |  37 +++-
 fs/ext4/ext4.h                               |  56 +++++-
 fs/ext4/hash.c                               |  25 ++-
 fs/ext4/inline.c                             |  25 ++-
 fs/ext4/namei.c                              | 198 ++++++++++++++-----
 fs/ext4/super.c                              |   6 -
 7 files changed, 285 insertions(+), 89 deletions(-)

diff --git a/Documentation/filesystems/ext4/directory.rst b/Documentation/filesystems/ext4/directory.rst
index 073940cc64ed..55f618b37144 100644
--- a/Documentation/filesystems/ext4/directory.rst
+++ b/Documentation/filesystems/ext4/directory.rst
@@ -121,6 +121,31 @@ The directory file type is one of the following values:
    * - 0x7
      - Symbolic link.
 
+To support directories that are both encrypted and casefolded directories, we
+must also include hash information in the directory entry. We append
+``ext4_extended_dir_entry_2`` to ``ext4_dir_entry_2`` except for the entries
+for dot and dotdot, which are kept the same. The structure follows immediately
+after ``name`` and is included in the size listed by ``rec_len`` If a directory
+entry uses this extension, it may be up to 271 bytes.
+
+.. list-table::
+   :widths: 8 8 24 40
+   :header-rows: 1
+
+   * - Offset
+     - Size
+     - Name
+     - Description
+   * - 0x0
+     - \_\_le32
+     - hash
+     - The hash of the directory name
+   * - 0x4
+     - \_\_le32
+     - minor\_hash
+     - The minor hash of the directory name
+
+
 In order to add checksums to these classic directory blocks, a phony
 ``struct ext4_dir_entry`` is placed at the end of each leaf block to
 hold the checksum. The directory entry is 12 bytes long. The inode
@@ -322,6 +347,8 @@ The directory hash is one of the following values:
      - Half MD4, unsigned.
    * - 0x5
      - Tea, unsigned.
+   * - 0x6
+     - Siphash.
 
 Interior nodes of an htree are recorded as ``struct dx_node``, which is
 also the full length of a data block:
diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
index 5ed870614c8d..21a98288de49 100644
--- a/fs/ext4/dir.c
+++ b/fs/ext4/dir.c
@@ -55,6 +55,18 @@ static int is_dx_dir(struct inode *inode)
 	return 0;
 }
 
+static bool is_fake_dir_entry(struct ext4_dir_entry_2 *de)
+{
+	/* Check if . or .. , or skip if namelen is 0 */
+	if ((de->name_len > 0) && (de->name_len <= 2) && (de->name[0] == '.') &&
+	    (de->name[1] == '.' || de->name[1] == '\0'))
+		return true;
+	/* Check if this is a csum entry */
+	if (de->file_type == EXT4_FT_DIR_CSUM)
+		return true;
+	return false;
+}
+
 /*
  * Return 0 if the directory entry is OK, and 1 if there is a problem
  *
@@ -73,16 +85,20 @@ int __ext4_check_dir_entry(const char *function, unsigned int line,
 	const int rlen = ext4_rec_len_from_disk(de->rec_len,
 						dir->i_sb->s_blocksize);
 	const int next_offset = ((char *) de - buf) + rlen;
+	bool fake = is_fake_dir_entry(de);
+	bool has_csum = ext4_has_metadata_csum(dir->i_sb);
 
-	if (unlikely(rlen < EXT4_DIR_REC_LEN(1)))
+	if (unlikely(rlen < ext4_dir_rec_len(1, fake ? NULL : dir)))
 		error_msg = "rec_len is smaller than minimal";
 	else if (unlikely(rlen % 4 != 0))
 		error_msg = "rec_len % 4 != 0";
-	else if (unlikely(rlen < EXT4_DIR_REC_LEN(de->name_len)))
+	else if (unlikely(rlen < ext4_dir_rec_len(de->name_len,
+							fake ? NULL : dir)))
 		error_msg = "rec_len is too small for name_len";
 	else if (unlikely(next_offset > size))
 		error_msg = "directory entry overrun";
-	else if (unlikely(next_offset > size - EXT4_DIR_REC_LEN(1) &&
+	else if (unlikely(next_offset > size - ext4_dir_rec_len(1,
+						  has_csum ? NULL : dir) &&
 			  next_offset != size))
 		error_msg = "directory entry too close to block end";
 	else if (unlikely(le32_to_cpu(de->inode) >
@@ -94,15 +110,15 @@ int __ext4_check_dir_entry(const char *function, unsigned int line,
 	if (filp)
 		ext4_error_file(filp, function, line, bh->b_blocknr,
 				"bad entry in directory: %s - offset=%u, "
-				"inode=%u, rec_len=%d, name_len=%d, size=%d",
+				"inode=%u, rec_len=%d, size=%d fake=%d",
 				error_msg, offset, le32_to_cpu(de->inode),
-				rlen, de->name_len, size);
+				rlen, size, fake);
 	else
 		ext4_error_inode(dir, function, line, bh->b_blocknr,
 				"bad entry in directory: %s - offset=%u, "
-				"inode=%u, rec_len=%d, name_len=%d, size=%d",
+				"inode=%u, rec_len=%d, size=%d fake=%d",
 				 error_msg, offset, le32_to_cpu(de->inode),
-				 rlen, de->name_len, size);
+				 rlen, size, fake);
 
 	return 1;
 }
@@ -224,7 +240,8 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx)
 				 * failure will be detected in the
 				 * dirent test below. */
 				if (ext4_rec_len_from_disk(de->rec_len,
-					sb->s_blocksize) < EXT4_DIR_REC_LEN(1))
+					sb->s_blocksize) < ext4_dir_rec_len(1,
+									inode))
 					break;
 				i += ext4_rec_len_from_disk(de->rec_len,
 							    sb->s_blocksize);
@@ -265,7 +282,9 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx)
 
 					/* Directory is encrypted */
 					err = fscrypt_fname_disk_to_usr(inode,
-						0, 0, &de_name, &fstr);
+						EXT4_DIRENT_HASH(de),
+						EXT4_DIRENT_MINOR_HASH(de),
+						&de_name, &fstr);
 					de_name = fstr;
 					fstr.len = save_len;
 					if (err)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 644fd69185d3..dafa528c4d9f 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2195,6 +2195,17 @@ struct ext4_dir_entry {
 	char	name[EXT4_NAME_LEN];	/* File name */
 };
 
+
+/*
+ * Encrypted Casefolded entries require saving the hash on disk. This structure
+ * followed ext4_dir_entry_2's name[name_len] at the next 4 byte aligned
+ * boundary.
+ */
+struct ext4_dir_entry_hash {
+	__le32 hash;
+	__le32 minor_hash;
+};
+
 /*
  * The new version of the directory entry.  Since EXT4 structures are
  * stored in intel byte order, and the name_len field could never be
@@ -2209,6 +2220,22 @@ struct ext4_dir_entry_2 {
 	char	name[EXT4_NAME_LEN];	/* File name */
 };
 
+/*
+ * Access the hashes at the end of ext4_dir_entry_2
+ */
+#define EXT4_DIRENT_HASHES(entry) \
+	((struct ext4_dir_entry_hash *) \
+		(((void *)(entry)) + \
+		((8 + (entry)->name_len + EXT4_DIR_ROUND) & ~EXT4_DIR_ROUND)))
+#define EXT4_DIRENT_HASH(entry) le32_to_cpu(EXT4_DIRENT_HASHES(de)->hash)
+#define EXT4_DIRENT_MINOR_HASH(entry) \
+		le32_to_cpu(EXT4_DIRENT_HASHES(de)->minor_hash)
+
+static inline bool ext4_hash_in_dirent(const struct inode *inode)
+{
+	return IS_CASEFOLDED(inode) && IS_ENCRYPTED(inode);
+}
+
 /*
  * This is a bogus directory entry at the end of each leaf block that
  * records checksums.
@@ -2250,10 +2277,24 @@ struct ext4_dir_entry_tail {
  */
 #define EXT4_DIR_PAD			4
 #define EXT4_DIR_ROUND			(EXT4_DIR_PAD - 1)
-#define EXT4_DIR_REC_LEN(name_len)	(((name_len) + 8 + EXT4_DIR_ROUND) & \
-					 ~EXT4_DIR_ROUND)
 #define EXT4_MAX_REC_LEN		((1<<16)-1)
 
+/*
+ * The rec_len is dependent on the type of directory. Directories that are
+ * casefolded and encrypted need to store the hash as well, so we add room for
+ * ext4_extended_dir_entry_2. For all entries related to '.' or '..' you should
+ * pass NULL for dir, as those entries do not use the extra fields.
+ */
+static inline unsigned int ext4_dir_rec_len(__u8 name_len,
+						const struct inode *dir)
+{
+	int rec_len = (name_len + 8 + EXT4_DIR_ROUND);
+
+	if (dir && ext4_hash_in_dirent(dir))
+		rec_len += sizeof(struct ext4_dir_entry_hash);
+	return (rec_len & ~EXT4_DIR_ROUND);
+}
+
 /*
  * If we ever get support for fs block sizes > page_size, we'll need
  * to remove the #if statements in the next two functions...
@@ -2310,6 +2351,7 @@ static inline __le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize)
 #define DX_HASH_LEGACY_UNSIGNED		3
 #define DX_HASH_HALF_MD4_UNSIGNED	4
 #define DX_HASH_TEA_UNSIGNED		5
+#define DX_HASH_SIPHASH			6
 
 static inline u32 ext4_chksum(struct ext4_sb_info *sbi, u32 crc,
 			      const void *address, unsigned int length)
@@ -2364,6 +2406,7 @@ struct ext4_filename {
 };
 
 #define fname_name(p) ((p)->disk_name.name)
+#define fname_usr_name(p) ((p)->usr_fname->name)
 #define fname_len(p)  ((p)->disk_name.len)
 
 /*
@@ -2706,9 +2749,9 @@ extern int __ext4_check_dir_entry(const char *, unsigned int, struct inode *,
 				  struct ext4_dir_entry_2 *,
 				  struct buffer_head *, char *, int,
 				  unsigned int);
-#define ext4_check_dir_entry(dir, filp, de, bh, buf, size, offset)	\
+#define ext4_check_dir_entry(dir, filp, de, bh, buf, size, offset) \
 	unlikely(__ext4_check_dir_entry(__func__, __LINE__, (dir), (filp), \
-					(de), (bh), (buf), (size), (offset)))
+				(de), (bh), (buf), (size), (offset)))
 extern int ext4_htree_store_dirent(struct file *dir_file, __u32 hash,
 				__u32 minor_hash,
 				struct ext4_dir_entry_2 *dirent,
@@ -2719,7 +2762,7 @@ extern int ext4_find_dest_de(struct inode *dir, struct inode *inode,
 			     void *buf, int buf_size,
 			     struct ext4_filename *fname,
 			     struct ext4_dir_entry_2 **dest_de);
-void ext4_insert_dentry(struct inode *inode,
+void ext4_insert_dentry(struct inode *dir, struct inode *inode,
 			struct ext4_dir_entry_2 *de,
 			int buf_size,
 			struct ext4_filename *fname);
@@ -3516,9 +3559,6 @@ extern void ext4_initialize_dirent_tail(struct buffer_head *bh,
 					unsigned int blocksize);
 extern int ext4_handle_dirty_dirblock(handle_t *handle, struct inode *inode,
 				      struct buffer_head *bh);
-extern int ext4_ci_compare(const struct inode *parent,
-			   const struct qstr *fname,
-			   const struct qstr *entry, bool quick);
 extern int __ext4_unlink(handle_t *handle, struct inode *dir, const struct qstr *d_name,
 			 struct inode *inode);
 extern int __ext4_link(struct inode *dir, struct inode *inode,
diff --git a/fs/ext4/hash.c b/fs/ext4/hash.c
index a92eb79de0cc..f34f4176c1e7 100644
--- a/fs/ext4/hash.c
+++ b/fs/ext4/hash.c
@@ -197,7 +197,7 @@ static void str2hashbuf_unsigned(const char *msg, int len, __u32 *buf, int num)
  * represented, and whether or not the returned hash is 32 bits or 64
  * bits.  32 bit hashes will return 0 for the minor hash.
  */
-static int __ext4fs_dirhash(const char *name, int len,
+static int __ext4fs_dirhash(const struct inode *dir, const char *name, int len,
 			    struct dx_hash_info *hinfo)
 {
 	__u32	hash;
@@ -259,6 +259,22 @@ static int __ext4fs_dirhash(const char *name, int len,
 		hash = buf[0];
 		minor_hash = buf[1];
 		break;
+	case DX_HASH_SIPHASH:
+	{
+		struct qstr qname = QSTR_INIT(name, len);
+		__u64	combined_hash;
+
+		if (fscrypt_has_encryption_key(dir)) {
+			combined_hash = fscrypt_fname_siphash(dir, &qname);
+		} else {
+			ext4_warning_inode(dir, "Siphash requires key");
+			return -1;
+		}
+
+		hash = (__u32)(combined_hash >> 32);
+		minor_hash = (__u32)combined_hash;
+		break;
+	}
 	default:
 		hinfo->hash = 0;
 		return -1;
@@ -280,7 +296,8 @@ int ext4fs_dirhash(const struct inode *dir, const char *name, int len,
 	unsigned char *buff;
 	struct qstr qstr = {.name = name, .len = len };
 
-	if (len && IS_CASEFOLDED(dir) && um) {
+	if (len && IS_CASEFOLDED(dir) && um &&
+	   (!IS_ENCRYPTED(dir) || fscrypt_has_encryption_key(dir))) {
 		buff = kzalloc(sizeof(char) * PATH_MAX, GFP_KERNEL);
 		if (!buff)
 			return -ENOMEM;
@@ -291,12 +308,12 @@ int ext4fs_dirhash(const struct inode *dir, const char *name, int len,
 			goto opaque_seq;
 		}
 
-		r = __ext4fs_dirhash(buff, dlen, hinfo);
+		r = __ext4fs_dirhash(dir, buff, dlen, hinfo);
 
 		kfree(buff);
 		return r;
 	}
 opaque_seq:
 #endif
-	return __ext4fs_dirhash(name, len, hinfo);
+	return __ext4fs_dirhash(dir, name, len, hinfo);
 }
diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
index b41512d1badc..ba960e6e7e3a 100644
--- a/fs/ext4/inline.c
+++ b/fs/ext4/inline.c
@@ -1031,7 +1031,7 @@ static int ext4_add_dirent_to_inline(handle_t *handle,
 	err = ext4_journal_get_write_access(handle, iloc->bh);
 	if (err)
 		return err;
-	ext4_insert_dentry(inode, de, inline_size, fname);
+	ext4_insert_dentry(dir, inode, de, inline_size, fname);
 
 	ext4_show_inline_dir(dir, iloc->bh, inline_start, inline_size);
 
@@ -1100,7 +1100,7 @@ static int ext4_update_inline_dir(handle_t *handle, struct inode *dir,
 	int old_size = EXT4_I(dir)->i_inline_size - EXT4_MIN_INLINE_DATA_SIZE;
 	int new_size = get_max_inline_xattr_value_size(dir, iloc);
 
-	if (new_size - old_size <= EXT4_DIR_REC_LEN(1))
+	if (new_size - old_size <= ext4_dir_rec_len(1, NULL))
 		return -ENOSPC;
 
 	ret = ext4_update_inline_data(handle, dir,
@@ -1380,8 +1380,8 @@ int ext4_inlinedir_to_tree(struct file *dir_file,
 			fake.name_len = 1;
 			strcpy(fake.name, ".");
 			fake.rec_len = ext4_rec_len_to_disk(
-						EXT4_DIR_REC_LEN(fake.name_len),
-						inline_size);
+					  ext4_dir_rec_len(fake.name_len, NULL),
+					  inline_size);
 			ext4_set_de_type(inode->i_sb, &fake, S_IFDIR);
 			de = &fake;
 			pos = EXT4_INLINE_DOTDOT_OFFSET;
@@ -1390,8 +1390,8 @@ int ext4_inlinedir_to_tree(struct file *dir_file,
 			fake.name_len = 2;
 			strcpy(fake.name, "..");
 			fake.rec_len = ext4_rec_len_to_disk(
-						EXT4_DIR_REC_LEN(fake.name_len),
-						inline_size);
+					  ext4_dir_rec_len(fake.name_len, NULL),
+					  inline_size);
 			ext4_set_de_type(inode->i_sb, &fake, S_IFDIR);
 			de = &fake;
 			pos = EXT4_INLINE_DOTDOT_SIZE;
@@ -1406,7 +1406,12 @@ int ext4_inlinedir_to_tree(struct file *dir_file,
 			}
 		}
 
-		ext4fs_dirhash(dir, de->name, de->name_len, hinfo);
+		if (ext4_hash_in_dirent(dir)) {
+			hinfo->hash = EXT4_DIRENT_HASH(de);
+			hinfo->minor_hash = EXT4_DIRENT_MINOR_HASH(de);
+		} else {
+			ext4fs_dirhash(dir, de->name, de->name_len, hinfo);
+		}
 		if ((hinfo->hash < start_hash) ||
 		    ((hinfo->hash == start_hash) &&
 		     (hinfo->minor_hash < start_minor_hash)))
@@ -1488,8 +1493,8 @@ int ext4_read_inline_dir(struct file *file,
 	 * So we will use extra_offset and extra_size to indicate them
 	 * during the inline dir iteration.
 	 */
-	dotdot_offset = EXT4_DIR_REC_LEN(1);
-	dotdot_size = dotdot_offset + EXT4_DIR_REC_LEN(2);
+	dotdot_offset = ext4_dir_rec_len(1, NULL);
+	dotdot_size = dotdot_offset + ext4_dir_rec_len(2, NULL);
 	extra_offset = dotdot_size - EXT4_INLINE_DOTDOT_SIZE;
 	extra_size = extra_offset + inline_size;
 
@@ -1524,7 +1529,7 @@ int ext4_read_inline_dir(struct file *file,
 			 * failure will be detected in the
 			 * dirent test below. */
 			if (ext4_rec_len_from_disk(de->rec_len, extra_size)
-				< EXT4_DIR_REC_LEN(1))
+				< ext4_dir_rec_len(1, NULL))
 				break;
 			i += ext4_rec_len_from_disk(de->rec_len,
 						    extra_size);
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 686bf982c84e..97d2755b9775 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -280,9 +280,11 @@ static int dx_make_map(struct inode *dir, struct ext4_dir_entry_2 *de,
 		       unsigned blocksize, struct dx_hash_info *hinfo,
 		       struct dx_map_entry map[]);
 static void dx_sort_map(struct dx_map_entry *map, unsigned count);
-static struct ext4_dir_entry_2 *dx_move_dirents(char *from, char *to,
-		struct dx_map_entry *offsets, int count, unsigned blocksize);
-static struct ext4_dir_entry_2* dx_pack_dirents(char *base, unsigned blocksize);
+static struct ext4_dir_entry_2 *dx_move_dirents(struct inode *dir, char *from,
+					char *to, struct dx_map_entry *offsets,
+					int count, unsigned int blocksize);
+static struct ext4_dir_entry_2 *dx_pack_dirents(struct inode *dir, char *base,
+						unsigned int blocksize);
 static void dx_insert_block(struct dx_frame *frame,
 					u32 hash, ext4_lblk_t block);
 static int ext4_htree_next_block(struct inode *dir, __u32 hash,
@@ -574,8 +576,9 @@ static inline void dx_set_limit(struct dx_entry *entries, unsigned value)
 
 static inline unsigned dx_root_limit(struct inode *dir, unsigned infosize)
 {
-	unsigned entry_space = dir->i_sb->s_blocksize - EXT4_DIR_REC_LEN(1) -
-		EXT4_DIR_REC_LEN(2) - infosize;
+	unsigned int entry_space = dir->i_sb->s_blocksize -
+			ext4_dir_rec_len(1, NULL) -
+			ext4_dir_rec_len(2, NULL) - infosize;
 
 	if (ext4_has_metadata_csum(dir->i_sb))
 		entry_space -= sizeof(struct dx_tail);
@@ -584,7 +587,8 @@ static inline unsigned dx_root_limit(struct inode *dir, unsigned infosize)
 
 static inline unsigned dx_node_limit(struct inode *dir)
 {
-	unsigned entry_space = dir->i_sb->s_blocksize - EXT4_DIR_REC_LEN(0);
+	unsigned int entry_space = dir->i_sb->s_blocksize -
+			ext4_dir_rec_len(0, dir);
 
 	if (ext4_has_metadata_csum(dir->i_sb))
 		entry_space -= sizeof(struct dx_tail);
@@ -673,7 +677,10 @@ static struct stats dx_show_leaf(struct inode *dir,
 						name = fname_crypto_str.name;
 						len = fname_crypto_str.len;
 					}
-					ext4fs_dirhash(dir, de->name,
+					if (IS_CASEFOLDED(dir))
+						h.hash = EXT4_DIRENT_HASH(de);
+					else
+						ext4fs_dirhash(dir, de->name,
 						       de->name_len, &h);
 					printk("%*.s:(E)%x.%u ", len, name,
 					       h.hash, (unsigned) ((char *) de
@@ -689,7 +696,7 @@ static struct stats dx_show_leaf(struct inode *dir,
 				       (unsigned) ((char *) de - base));
 #endif
 			}
-			space += EXT4_DIR_REC_LEN(de->name_len);
+			space += ext4_dir_rec_len(de->name_len, dir);
 			names++;
 		}
 		de = ext4_next_entry(de, size);
@@ -784,11 +791,25 @@ dx_probe(struct ext4_filename *fname, struct inode *dir,
 	root = (struct dx_root *) frame->bh->b_data;
 	if (root->info.hash_version != DX_HASH_TEA &&
 	    root->info.hash_version != DX_HASH_HALF_MD4 &&
-	    root->info.hash_version != DX_HASH_LEGACY) {
+	    root->info.hash_version != DX_HASH_LEGACY &&
+	    root->info.hash_version != DX_HASH_SIPHASH) {
 		ext4_warning_inode(dir, "Unrecognised inode hash code %u",
 				   root->info.hash_version);
 		goto fail;
 	}
+	if (ext4_hash_in_dirent(dir)) {
+		if (root->info.hash_version != DX_HASH_SIPHASH) {
+			ext4_warning_inode(dir,
+				"Hash in dirent, but hash is not SIPHASH");
+			goto fail;
+		}
+	} else {
+		if (root->info.hash_version == DX_HASH_SIPHASH) {
+			ext4_warning_inode(dir,
+				"Hash code is SIPHASH, but hash not in dirent");
+			goto fail;
+		}
+	}
 	if (fname)
 		hinfo = &fname->hinfo;
 	hinfo->hash_version = root->info.hash_version;
@@ -997,6 +1018,7 @@ static int htree_dirblock_to_tree(struct file *dir_file,
 	struct ext4_dir_entry_2 *de, *top;
 	int err = 0, count = 0;
 	struct fscrypt_str fname_crypto_str = FSTR_INIT(NULL, 0), tmp_str;
+	int csum = ext4_has_metadata_csum(dir->i_sb);
 
 	dxtrace(printk(KERN_INFO "In htree dirblock_to_tree: block %lu\n",
 							(unsigned long)block));
@@ -1005,9 +1027,11 @@ static int htree_dirblock_to_tree(struct file *dir_file,
 		return PTR_ERR(bh);
 
 	de = (struct ext4_dir_entry_2 *) bh->b_data;
+	/* csum entries are not larger in the casefolded encrypted case */
 	top = (struct ext4_dir_entry_2 *) ((char *) de +
 					   dir->i_sb->s_blocksize -
-					   EXT4_DIR_REC_LEN(0));
+					   ext4_dir_rec_len(0,
+							   csum ? NULL : dir));
 	/* Check if the directory is encrypted */
 	if (IS_ENCRYPTED(dir)) {
 		err = fscrypt_prepare_readdir(dir);
@@ -1031,7 +1055,17 @@ static int htree_dirblock_to_tree(struct file *dir_file,
 			/* silently ignore the rest of the block */
 			break;
 		}
-		ext4fs_dirhash(dir, de->name, de->name_len, hinfo);
+		if (ext4_hash_in_dirent(dir)) {
+			if (de->name_len && de->inode) {
+				hinfo->hash = EXT4_DIRENT_HASH(de);
+				hinfo->minor_hash = EXT4_DIRENT_MINOR_HASH(de);
+			} else {
+				hinfo->hash = 0;
+				hinfo->minor_hash = 0;
+			}
+		} else {
+			ext4fs_dirhash(dir, de->name, de->name_len, hinfo);
+		}
 		if ((hinfo->hash < start_hash) ||
 		    ((hinfo->hash == start_hash) &&
 		     (hinfo->minor_hash < start_minor_hash)))
@@ -1100,7 +1134,11 @@ int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash,
 		       start_hash, start_minor_hash));
 	dir = file_inode(dir_file);
 	if (!(ext4_test_inode_flag(dir, EXT4_INODE_INDEX))) {
-		hinfo.hash_version = EXT4_SB(dir->i_sb)->s_def_hash_version;
+		if (ext4_hash_in_dirent(dir))
+			hinfo.hash_version = DX_HASH_SIPHASH;
+		else
+			hinfo.hash_version =
+					EXT4_SB(dir->i_sb)->s_def_hash_version;
 		if (hinfo.hash_version <= DX_HASH_TEA)
 			hinfo.hash_version +=
 				EXT4_SB(dir->i_sb)->s_hash_unsigned;
@@ -1218,7 +1256,10 @@ static int dx_make_map(struct inode *dir, struct ext4_dir_entry_2 *de,
 
 	while ((char *) de < base + blocksize) {
 		if (de->name_len && de->inode) {
-			ext4fs_dirhash(dir, de->name, de->name_len, &h);
+			if (ext4_hash_in_dirent(dir))
+				h.hash = EXT4_DIRENT_HASH(de);
+			else
+				ext4fs_dirhash(dir, de->name, de->name_len, &h);
 			map_tail--;
 			map_tail->hash = h.hash;
 			map_tail->offs = ((char *) de - base)>>2;
@@ -1282,31 +1323,47 @@ static void dx_insert_block(struct dx_frame *frame, u32 hash, ext4_lblk_t block)
  * Returns: 0 if the directory entry matches, more than 0 if it
  * doesn't match or less than zero on error.
  */
-int ext4_ci_compare(const struct inode *parent, const struct qstr *name,
-		    const struct qstr *entry, bool quick)
+static int ext4_ci_compare(const struct inode *parent, const struct qstr *name,
+			   u8 *de_name, size_t de_name_len, bool quick)
 {
 	const struct super_block *sb = parent->i_sb;
 	const struct unicode_map *um = sb->s_encoding;
+	struct fscrypt_str decrypted_name = FSTR_INIT(NULL, de_name_len);
+	struct qstr entry = QSTR_INIT(de_name, de_name_len);
 	int ret;
 
+	if (IS_ENCRYPTED(parent)) {
+		const struct fscrypt_str encrypted_name =
+				FSTR_INIT(de_name, de_name_len);
+
+		decrypted_name.name = kmalloc(de_name_len, GFP_KERNEL);
+		if (!decrypted_name.name)
+			return -ENOMEM;
+		ret = fscrypt_fname_disk_to_usr(parent, 0, 0, &encrypted_name,
+						&decrypted_name);
+		if (ret < 0)
+			goto out;
+		entry.name = decrypted_name.name;
+		entry.len = decrypted_name.len;
+	}
+
 	if (quick)
-		ret = utf8_strncasecmp_folded(um, name, entry);
+		ret = utf8_strncasecmp_folded(um, name, &entry);
 	else
-		ret = utf8_strncasecmp(um, name, entry);
-
+		ret = utf8_strncasecmp(um, name, &entry);
 	if (ret < 0) {
 		/* Handle invalid character sequence as either an error
 		 * or as an opaque byte sequence.
 		 */
 		if (sb_has_strict_encoding(sb))
-			return -EINVAL;
-
-		if (name->len != entry->len)
-			return 1;
-
-		return !!memcmp(name->name, entry->name, name->len);
+			ret = -EINVAL;
+		else if (name->len != entry.len)
+			ret = 1;
+		else
+			ret = !!memcmp(name->name, entry.name, entry.len);
 	}
-
+out:
+	kfree(decrypted_name.name);
 	return ret;
 }
 
@@ -1342,14 +1399,11 @@ void ext4_fname_setup_ci_filename(struct inode *dir, const struct qstr *iname,
  *
  * Return: %true if the directory entry matches, otherwise %false.
  */
-static inline bool ext4_match(const struct inode *parent,
+static bool ext4_match(struct inode *parent,
 			      const struct ext4_filename *fname,
-			      const struct ext4_dir_entry_2 *de)
+			      struct ext4_dir_entry_2 *de)
 {
 	struct fscrypt_name f;
-#ifdef CONFIG_UNICODE
-	const struct qstr entry = {.name = de->name, .len = de->name_len};
-#endif
 
 	if (!de->inode)
 		return false;
@@ -1365,10 +1419,23 @@ static inline bool ext4_match(const struct inode *parent,
 		if (fname->cf_name.name) {
 			struct qstr cf = {.name = fname->cf_name.name,
 					  .len = fname->cf_name.len};
-			return !ext4_ci_compare(parent, &cf, &entry, true);
+			if (IS_ENCRYPTED(parent)) {
+				struct dx_hash_info hinfo;
+
+				hinfo.hash_version = DX_HASH_SIPHASH;
+				hinfo.seed = NULL;
+				ext4fs_dirhash(parent, fname->cf_name.name,
+						fname_len(fname), &hinfo);
+				if (hinfo.hash != EXT4_DIRENT_HASH(de) ||
+						hinfo.minor_hash !=
+						    EXT4_DIRENT_MINOR_HASH(de))
+					return 0;
+			}
+			return !ext4_ci_compare(parent, &cf, de->name,
+							de->name_len, true);
 		}
-		return !ext4_ci_compare(parent, fname->usr_fname, &entry,
-					false);
+		return !ext4_ci_compare(parent, fname->usr_fname, de->name,
+						de->name_len, false);
 	}
 #endif
 
@@ -1765,7 +1832,8 @@ struct dentry *ext4_get_parent(struct dentry *child)
  * Returns pointer to last entry moved.
  */
 static struct ext4_dir_entry_2 *
-dx_move_dirents(char *from, char *to, struct dx_map_entry *map, int count,
+dx_move_dirents(struct inode *dir, char *from, char *to,
+		struct dx_map_entry *map, int count,
 		unsigned blocksize)
 {
 	unsigned rec_len = 0;
@@ -1773,7 +1841,8 @@ dx_move_dirents(char *from, char *to, struct dx_map_entry *map, int count,
 	while (count--) {
 		struct ext4_dir_entry_2 *de = (struct ext4_dir_entry_2 *)
 						(from + (map->offs<<2));
-		rec_len = EXT4_DIR_REC_LEN(de->name_len);
+		rec_len = ext4_dir_rec_len(de->name_len, dir);
+
 		memcpy (to, de, rec_len);
 		((struct ext4_dir_entry_2 *) to)->rec_len =
 				ext4_rec_len_to_disk(rec_len, blocksize);
@@ -1788,7 +1857,8 @@ dx_move_dirents(char *from, char *to, struct dx_map_entry *map, int count,
  * Compact each dir entry in the range to the minimal rec_len.
  * Returns pointer to last entry in range.
  */
-static struct ext4_dir_entry_2* dx_pack_dirents(char *base, unsigned blocksize)
+static struct ext4_dir_entry_2 *dx_pack_dirents(struct inode *dir, char *base,
+							unsigned int blocksize)
 {
 	struct ext4_dir_entry_2 *next, *to, *prev, *de = (struct ext4_dir_entry_2 *) base;
 	unsigned rec_len = 0;
@@ -1797,7 +1867,7 @@ static struct ext4_dir_entry_2* dx_pack_dirents(char *base, unsigned blocksize)
 	while ((char*)de < base + blocksize) {
 		next = ext4_next_entry(de, blocksize);
 		if (de->inode && de->name_len) {
-			rec_len = EXT4_DIR_REC_LEN(de->name_len);
+			rec_len = ext4_dir_rec_len(de->name_len, dir);
 			if (de > to)
 				memmove(to, de, rec_len);
 			to->rec_len = ext4_rec_len_to_disk(rec_len, blocksize);
@@ -1887,9 +1957,9 @@ static struct ext4_dir_entry_2 *do_split(handle_t *handle, struct inode *dir,
 					hash2, split, count-split));
 
 	/* Fancy dance to stay within two buffers */
-	de2 = dx_move_dirents(data1, data2, map + split, count - split,
+	de2 = dx_move_dirents(dir, data1, data2, map + split, count - split,
 			      blocksize);
-	de = dx_pack_dirents(data1, blocksize);
+	de = dx_pack_dirents(dir, data1, blocksize);
 	de->rec_len = ext4_rec_len_to_disk(data1 + (blocksize - csum_size) -
 					   (char *) de,
 					   blocksize);
@@ -1937,7 +2007,7 @@ int ext4_find_dest_de(struct inode *dir, struct inode *inode,
 		      struct ext4_dir_entry_2 **dest_de)
 {
 	struct ext4_dir_entry_2 *de;
-	unsigned short reclen = EXT4_DIR_REC_LEN(fname_len(fname));
+	unsigned short reclen = ext4_dir_rec_len(fname_len(fname), dir);
 	int nlen, rlen;
 	unsigned int offset = 0;
 	char *top;
@@ -1950,7 +2020,7 @@ int ext4_find_dest_de(struct inode *dir, struct inode *inode,
 			return -EFSCORRUPTED;
 		if (ext4_match(dir, fname, de))
 			return -EEXIST;
-		nlen = EXT4_DIR_REC_LEN(de->name_len);
+		nlen = ext4_dir_rec_len(de->name_len, dir);
 		rlen = ext4_rec_len_from_disk(de->rec_len, buf_size);
 		if ((de->inode ? rlen - nlen : rlen) >= reclen)
 			break;
@@ -1964,7 +2034,8 @@ int ext4_find_dest_de(struct inode *dir, struct inode *inode,
 	return 0;
 }
 
-void ext4_insert_dentry(struct inode *inode,
+void ext4_insert_dentry(struct inode *dir,
+			struct inode *inode,
 			struct ext4_dir_entry_2 *de,
 			int buf_size,
 			struct ext4_filename *fname)
@@ -1972,7 +2043,7 @@ void ext4_insert_dentry(struct inode *inode,
 
 	int nlen, rlen;
 
-	nlen = EXT4_DIR_REC_LEN(de->name_len);
+	nlen = ext4_dir_rec_len(de->name_len, dir);
 	rlen = ext4_rec_len_from_disk(de->rec_len, buf_size);
 	if (de->inode) {
 		struct ext4_dir_entry_2 *de1 =
@@ -1986,6 +2057,17 @@ void ext4_insert_dentry(struct inode *inode,
 	ext4_set_de_type(inode->i_sb, de, inode->i_mode);
 	de->name_len = fname_len(fname);
 	memcpy(de->name, fname_name(fname), fname_len(fname));
+	if (ext4_hash_in_dirent(dir)) {
+		struct dx_hash_info hinfo;
+
+		hinfo.hash_version = DX_HASH_SIPHASH;
+		hinfo.seed = NULL;
+		ext4fs_dirhash(dir, fname_usr_name(fname),
+				fname_len(fname), &hinfo);
+		EXT4_DIRENT_HASHES(de)->hash = cpu_to_le32(hinfo.hash);
+		EXT4_DIRENT_HASHES(de)->minor_hash =
+				cpu_to_le32(hinfo.minor_hash);
+	}
 }
 
 /*
@@ -2022,7 +2104,7 @@ static int add_dirent_to_buf(handle_t *handle, struct ext4_filename *fname,
 	}
 
 	/* By now the buffer is marked for journaling */
-	ext4_insert_dentry(inode, de, blocksize, fname);
+	ext4_insert_dentry(dir, inode, de, blocksize, fname);
 
 	/*
 	 * XXX shouldn't update any times until successful
@@ -2114,11 +2196,16 @@ static int make_indexed_dir(handle_t *handle, struct ext4_filename *fname,
 
 	/* Initialize the root; the dot dirents already exist */
 	de = (struct ext4_dir_entry_2 *) (&root->dotdot);
-	de->rec_len = ext4_rec_len_to_disk(blocksize - EXT4_DIR_REC_LEN(2),
-					   blocksize);
+	de->rec_len = ext4_rec_len_to_disk(
+			blocksize - ext4_dir_rec_len(2, NULL), blocksize);
 	memset (&root->info, 0, sizeof(root->info));
 	root->info.info_length = sizeof(root->info);
-	root->info.hash_version = EXT4_SB(dir->i_sb)->s_def_hash_version;
+	if (ext4_hash_in_dirent(dir))
+		root->info.hash_version = DX_HASH_SIPHASH;
+	else
+		root->info.hash_version =
+				EXT4_SB(dir->i_sb)->s_def_hash_version;
+
 	entries = root->entries;
 	dx_set_block(entries, 1);
 	dx_set_count(entries, 1);
@@ -2129,7 +2216,12 @@ static int make_indexed_dir(handle_t *handle, struct ext4_filename *fname,
 	if (fname->hinfo.hash_version <= DX_HASH_TEA)
 		fname->hinfo.hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned;
 	fname->hinfo.seed = EXT4_SB(dir->i_sb)->s_hash_seed;
-	ext4fs_dirhash(dir, fname_name(fname), fname_len(fname), &fname->hinfo);
+	if (ext4_hash_in_dirent(dir))
+		ext4fs_dirhash(dir, fname_usr_name(fname),
+				fname_len(fname), &fname->hinfo);
+	else
+		ext4fs_dirhash(dir, fname_name(fname),
+				fname_len(fname), &fname->hinfo);
 
 	memset(frames, 0, sizeof(frames));
 	frame = frames;
@@ -2722,7 +2814,7 @@ struct ext4_dir_entry_2 *ext4_init_dot_dotdot(struct inode *inode,
 {
 	de->inode = cpu_to_le32(inode->i_ino);
 	de->name_len = 1;
-	de->rec_len = ext4_rec_len_to_disk(EXT4_DIR_REC_LEN(de->name_len),
+	de->rec_len = ext4_rec_len_to_disk(ext4_dir_rec_len(de->name_len, NULL),
 					   blocksize);
 	strcpy(de->name, ".");
 	ext4_set_de_type(inode->i_sb, de, S_IFDIR);
@@ -2732,11 +2824,12 @@ struct ext4_dir_entry_2 *ext4_init_dot_dotdot(struct inode *inode,
 	de->name_len = 2;
 	if (!dotdot_real_len)
 		de->rec_len = ext4_rec_len_to_disk(blocksize -
-					(csum_size + EXT4_DIR_REC_LEN(1)),
+					(csum_size + ext4_dir_rec_len(1, NULL)),
 					blocksize);
 	else
 		de->rec_len = ext4_rec_len_to_disk(
-				EXT4_DIR_REC_LEN(de->name_len), blocksize);
+					ext4_dir_rec_len(de->name_len, NULL),
+					blocksize);
 	strcpy(de->name, "..");
 	ext4_set_de_type(inode->i_sb, de, S_IFDIR);
 
@@ -2869,7 +2962,8 @@ bool ext4_empty_dir(struct inode *inode)
 	}
 
 	sb = inode->i_sb;
-	if (inode->i_size < EXT4_DIR_REC_LEN(1) + EXT4_DIR_REC_LEN(2)) {
+	if (inode->i_size < ext4_dir_rec_len(1, NULL) +
+					ext4_dir_rec_len(2, NULL)) {
 		EXT4_ERROR_INODE(inode, "invalid size");
 		return true;
 	}
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index ad34a37278cd..312c7c03f401 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -4291,12 +4291,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 		struct unicode_map *encoding;
 		__u16 encoding_flags;
 
-		if (ext4_has_feature_encrypt(sb)) {
-			ext4_msg(sb, KERN_ERR,
-				 "Can't mount with encoding and encryption");
-			goto failed_mount;
-		}
-
 		if (ext4_sb_read_encoding(es, &encoding_info,
 					  &encoding_flags)) {
 			ext4_msg(sb, KERN_ERR,
-- 
2.31.0.rc2.261.g7f71774620-goog

Re: [PATCH v2 1/2] ext4: Handle casefolding with encryption

[Andreas Dilger]

[-- Attachment #1: Type: text/plain, Size: 764 bytes --]

On Mar 19, 2021, at 1:34 AM, Daniel Rosenberg <drosen@google.com> wrote:
> 
> This adds support for encryption with casefolding.
> 
> Since the name on disk is case preserving, and also encrypted, we can no
> longer just recompute the hash on the fly. Additionally, to avoid
> leaking extra information from the hash of the unencrypted name, we use
> siphash via an fscrypt v2 policy.
> 
> The hash is stored at the end of the directory entry for all entries
> inside of an encrypted and casefolded directory apart from those that
> deal with '.' and '..'. This way, the change is backwards compatible
> with existing ext4 filesystems.
> 
> Signed-off-by: Daniel Rosenberg <drosen@google.com>

Reviewed-by: Andreas Dilger <adilger@dilger.ca>

Cheers, Andreas






[-- Attachment #2: Message signed with OpenPGP --]
[-- Type: application/pgp-signature, Size: 873 bytes --]

Re: [PATCH v2 1/2] ext4: Handle casefolding with encryption

[Theodore Ts'o]

On Fri, Mar 19, 2021 at 07:34:13AM +0000, Daniel Rosenberg wrote:
> This adds support for encryption with casefolding.
> 
> Since the name on disk is case preserving, and also encrypted, we can no
> longer just recompute the hash on the fly. Additionally, to avoid
> leaking extra information from the hash of the unencrypted name, we use
> siphash via an fscrypt v2 policy.
> 
> The hash is stored at the end of the directory entry for all entries
> inside of an encrypted and casefolded directory apart from those that
> deal with '.' and '..'. This way, the change is backwards compatible
> with existing ext4 filesystems.
> 
> Signed-off-by: Daniel Rosenberg <drosen@google.com>

Applied, thanks with the following addition so that tests, e2fsprogs,
etc., can determine whether or not the currently running kernel has
this feature enabled:

diff --git a/fs/ext4/sysfs.c b/fs/ext4/sysfs.c
index a3d08276d441..7367ba406e01 100644
--- a/fs/ext4/sysfs.c
+++ b/fs/ext4/sysfs.c
@@ -313,6 +313,7 @@ EXT4_ATTR_FEATURE(verity);
 #endif
 EXT4_ATTR_FEATURE(metadata_csum_seed);
 EXT4_ATTR_FEATURE(fast_commit);
+EXT4_ATTR_FEATURE(encrypted_casefold);
 
 static struct attribute *ext4_feat_attrs[] = {
 	ATTR_LIST(lazy_itable_init),
@@ -330,6 +331,7 @@ static struct attribute *ext4_feat_attrs[] = {
 #endif
 	ATTR_LIST(metadata_csum_seed),
 	ATTR_LIST(fast_commit),
+	ATTR_LIST(encrypted_casefold),
 	NULL,
 };
 ATTRIBUTE_GROUPS(ext4_feat);


Future versions of e2fsprogs may issue a warning if tune2fs or mke2fs
tries to modify or create a file system such that both the encryption
and casefold feature is enabled if it appears that the kernel won't
support this combination.  Daniel, if you could try to get this change
into the Android kernels that are using encrypted casefold, that would
be a good thing.

					- Ted

[PATCH v2 2/2] ext4: Optimize match for casefolded encrypted dirs

[Daniel Rosenberg]

Matching names with casefolded encrypting directories requires
decrypting entries to confirm case since we are case preserving. We can
avoid needing to decrypt if our hash values don't match.

Signed-off-by: Daniel Rosenberg <drosen@google.com>
---
 fs/ext4/ext4.h  | 17 ++++++++-------
 fs/ext4/namei.c | 55 ++++++++++++++++++++++++++-----------------------
 2 files changed, 38 insertions(+), 34 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index dafa528c4d9f..181d07791efb 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2637,9 +2637,9 @@ extern unsigned ext4_free_clusters_after_init(struct super_block *sb,
 ext4_fsblk_t ext4_inode_to_goal_block(struct inode *);
 
 #ifdef CONFIG_UNICODE
-extern void ext4_fname_setup_ci_filename(struct inode *dir,
+extern int ext4_fname_setup_ci_filename(struct inode *dir,
 					 const struct qstr *iname,
-					 struct fscrypt_str *fname);
+					 struct ext4_filename *fname);
 #endif
 
 #ifdef CONFIG_FS_ENCRYPTION
@@ -2670,9 +2670,9 @@ static inline int ext4_fname_setup_filename(struct inode *dir,
 	ext4_fname_from_fscrypt_name(fname, &name);
 
 #ifdef CONFIG_UNICODE
-	ext4_fname_setup_ci_filename(dir, iname, &fname->cf_name);
+	err = ext4_fname_setup_ci_filename(dir, iname, fname);
 #endif
-	return 0;
+	return err;
 }
 
 static inline int ext4_fname_prepare_lookup(struct inode *dir,
@@ -2689,9 +2689,9 @@ static inline int ext4_fname_prepare_lookup(struct inode *dir,
 	ext4_fname_from_fscrypt_name(fname, &name);
 
 #ifdef CONFIG_UNICODE
-	ext4_fname_setup_ci_filename(dir, &dentry->d_name, &fname->cf_name);
+	err = ext4_fname_setup_ci_filename(dir, &dentry->d_name, fname);
 #endif
-	return 0;
+	return err;
 }
 
 static inline void ext4_fname_free_filename(struct ext4_filename *fname)
@@ -2716,15 +2716,16 @@ static inline int ext4_fname_setup_filename(struct inode *dir,
 					    int lookup,
 					    struct ext4_filename *fname)
 {
+	int err = 0;
 	fname->usr_fname = iname;
 	fname->disk_name.name = (unsigned char *) iname->name;
 	fname->disk_name.len = iname->len;
 
 #ifdef CONFIG_UNICODE
-	ext4_fname_setup_ci_filename(dir, iname, &fname->cf_name);
+	err = ext4_fname_setup_ci_filename(dir, iname, fname);
 #endif
 
-	return 0;
+	return err;
 }
 
 static inline int ext4_fname_prepare_lookup(struct inode *dir,
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 97d2755b9775..1fb7128220ce 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -816,7 +816,9 @@ dx_probe(struct ext4_filename *fname, struct inode *dir,
 	if (hinfo->hash_version <= DX_HASH_TEA)
 		hinfo->hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned;
 	hinfo->seed = EXT4_SB(dir->i_sb)->s_hash_seed;
-	if (fname && fname_name(fname))
+	/* hash is already computed for encrypted casefolded directory */
+	if (fname && fname_name(fname) &&
+				!(IS_ENCRYPTED(dir) && IS_CASEFOLDED(dir)))
 		ext4fs_dirhash(dir, fname_name(fname), fname_len(fname), hinfo);
 	hash = hinfo->hash;
 
@@ -1367,19 +1369,21 @@ static int ext4_ci_compare(const struct inode *parent, const struct qstr *name,
 	return ret;
 }
 
-void ext4_fname_setup_ci_filename(struct inode *dir, const struct qstr *iname,
-				  struct fscrypt_str *cf_name)
+int ext4_fname_setup_ci_filename(struct inode *dir, const struct qstr *iname,
+				  struct ext4_filename *name)
 {
+	struct fscrypt_str *cf_name = &name->cf_name;
+	struct dx_hash_info *hinfo = &name->hinfo;
 	int len;
 
 	if (!IS_CASEFOLDED(dir) || !dir->i_sb->s_encoding) {
 		cf_name->name = NULL;
-		return;
+		return 0;
 	}
 
 	cf_name->name = kmalloc(EXT4_NAME_LEN, GFP_NOFS);
 	if (!cf_name->name)
-		return;
+		return -ENOMEM;
 
 	len = utf8_casefold(dir->i_sb->s_encoding,
 			    iname, cf_name->name,
@@ -1387,10 +1391,18 @@ void ext4_fname_setup_ci_filename(struct inode *dir, const struct qstr *iname,
 	if (len <= 0) {
 		kfree(cf_name->name);
 		cf_name->name = NULL;
-		return;
 	}
 	cf_name->len = (unsigned) len;
+	if (!IS_ENCRYPTED(dir))
+		return 0;
 
+	hinfo->hash_version = DX_HASH_SIPHASH;
+	hinfo->seed = NULL;
+	if (cf_name->name)
+		ext4fs_dirhash(dir, cf_name->name, cf_name->len, hinfo);
+	else
+		ext4fs_dirhash(dir, iname->name, iname->len, hinfo);
+	return 0;
 }
 #endif
 
@@ -1420,16 +1432,12 @@ static bool ext4_match(struct inode *parent,
 			struct qstr cf = {.name = fname->cf_name.name,
 					  .len = fname->cf_name.len};
 			if (IS_ENCRYPTED(parent)) {
-				struct dx_hash_info hinfo;
-
-				hinfo.hash_version = DX_HASH_SIPHASH;
-				hinfo.seed = NULL;
-				ext4fs_dirhash(parent, fname->cf_name.name,
-						fname_len(fname), &hinfo);
-				if (hinfo.hash != EXT4_DIRENT_HASH(de) ||
-						hinfo.minor_hash !=
-						    EXT4_DIRENT_MINOR_HASH(de))
+				if (fname->hinfo.hash != EXT4_DIRENT_HASH(de) ||
+					fname->hinfo.minor_hash !=
+						EXT4_DIRENT_MINOR_HASH(de)) {
+
 					return 0;
+				}
 			}
 			return !ext4_ci_compare(parent, &cf, de->name,
 							de->name_len, true);
@@ -2058,15 +2066,11 @@ void ext4_insert_dentry(struct inode *dir,
 	de->name_len = fname_len(fname);
 	memcpy(de->name, fname_name(fname), fname_len(fname));
 	if (ext4_hash_in_dirent(dir)) {
-		struct dx_hash_info hinfo;
+		struct dx_hash_info *hinfo = &fname->hinfo;
 
-		hinfo.hash_version = DX_HASH_SIPHASH;
-		hinfo.seed = NULL;
-		ext4fs_dirhash(dir, fname_usr_name(fname),
-				fname_len(fname), &hinfo);
-		EXT4_DIRENT_HASHES(de)->hash = cpu_to_le32(hinfo.hash);
+		EXT4_DIRENT_HASHES(de)->hash = cpu_to_le32(hinfo->hash);
 		EXT4_DIRENT_HASHES(de)->minor_hash =
-				cpu_to_le32(hinfo.minor_hash);
+						cpu_to_le32(hinfo->minor_hash);
 	}
 }
 
@@ -2216,10 +2220,9 @@ static int make_indexed_dir(handle_t *handle, struct ext4_filename *fname,
 	if (fname->hinfo.hash_version <= DX_HASH_TEA)
 		fname->hinfo.hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned;
 	fname->hinfo.seed = EXT4_SB(dir->i_sb)->s_hash_seed;
-	if (ext4_hash_in_dirent(dir))
-		ext4fs_dirhash(dir, fname_usr_name(fname),
-				fname_len(fname), &fname->hinfo);
-	else
+
+	/* casefolded encrypted hashes are computed on fname setup */
+	if (!ext4_hash_in_dirent(dir))
 		ext4fs_dirhash(dir, fname_name(fname),
 				fname_len(fname), &fname->hinfo);
 
-- 
2.31.0.rc2.261.g7f71774620-goog

Re: [PATCH v2 2/2] ext4: Optimize match for casefolded encrypted dirs

[Theodore Ts'o]

On Fri, Mar 19, 2021 at 07:34:14AM +0000, Daniel Rosenberg wrote:
> Matching names with casefolded encrypting directories requires
> decrypting entries to confirm case since we are case preserving. We can
> avoid needing to decrypt if our hash values don't match.
> 
> Signed-off-by: Daniel Rosenberg <drosen@google.com>

Thanks, applied.

						- Ted