Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kern…

…el/git/tytso/ext4

Pull ext4 updates from Ted Ts'o:
 "Scalability improvements when allocating inodes, and some
  miscellaneous bug fixes and cleanups"

* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: avoid Y2038 overflow in recently_deleted()
  ext4: fix fault handling when mounted with -o dax,ro
  ext4: fix quota inconsistency during orphan cleanup for read-only mounts
  ext4: fix incorrect quotaoff if the quota feature is enabled
  ext4: remove useless test and assignment in strtohash functions
  ext4: backward compatibility support for Lustre ea_inode implementation
  ext4: remove timebomb in ext4_decode_extra_time()
  ext4: use sizeof(*ptr)
  ext4: in ext4_seek_{hole,data}, return -ENXIO for negative offsets
  ext4: reduce lock contention in __ext4_new_inode
  ext4: cleanup goto next group
  ext4: do not unnecessarily allocate buffer in recently_deleted()

fs/ext4/dir.c

@@ -411,7 +411,7 @@ static struct dir_private_info *ext4_htree_create_dir_info(struct file *filp,
 {
 	struct dir_private_info *p;
 
-	p = kzalloc(sizeof(struct dir_private_info), GFP_KERNEL);
+	p = kzalloc(sizeof(*p), GFP_KERNEL);
 	if (!p)
 		return NULL;
 	p->curr_hash = pos2maj_hash(filp, pos);

fs/ext4/ext4.h

@@ -838,13 +838,11 @@ static inline void ext4_decode_extra_time(struct timespec *time, __le32 extra)
 {
 	if (unlikely(sizeof(time->tv_sec) > 4 &&
 			(extra & cpu_to_le32(EXT4_EPOCH_MASK)))) {
-#if LINUX_VERSION_CODE < KERNEL_VERSION(4,20,0)
+
+#if 1
 		/* Handle legacy encoding of pre-1970 dates with epoch
-		 * bits 1,1.  We assume that by kernel version 4.20,
-		 * everyone will have run fsck over the affected
-		 * filesystems to correct the problem.  (This
-		 * backwards compatibility may be removed before this
-		 * time, at the discretion of the ext4 developers.)
+		 * bits 1,1. (This backwards compatibility may be removed
+		 * at the discretion of the ext4 developers.)
 		 */
 		u64 extra_bits = le32_to_cpu(extra) & EXT4_EPOCH_MASK;
 		if (extra_bits == 3 && ((time->tv_sec) & 0x80000000) != 0)
@@ -1567,6 +1565,7 @@ enum {
 					   nolocking */
 	EXT4_STATE_MAY_INLINE_DATA,	/* may have in-inode data */
 	EXT4_STATE_EXT_PRECACHED,	/* extents have been precached */
+	EXT4_STATE_LUSTRE_EA_INODE,	/* Lustre-style ea_inode */
 };
 
 #define EXT4_INODE_BIT_FNS(name, field, offset)				\

fs/ext4/file.c

@@ -279,7 +279,20 @@ static int ext4_dax_huge_fault(struct vm_fault *vmf,
 	handle_t *handle = NULL;
 	struct inode *inode = file_inode(vmf->vma->vm_file);
 	struct super_block *sb = inode->i_sb;
-	bool write = vmf->flags & FAULT_FLAG_WRITE;
+
+	/*
+	 * We have to distinguish real writes from writes which will result in a
+	 * COW page; COW writes should *not* poke the journal (the file will not
+	 * be changed). Doing so would cause unintended failures when mounted
+	 * read-only.
+	 *
+	 * We check for VM_SHARED rather than vmf->cow_page since the latter is
+	 * unset for pe_size != PE_SIZE_PTE (i.e. only in do_cow_fault); for
+	 * other sizes, dax_iomap_fault will handle splitting / fallback so that
+	 * we eventually come back with a COW page.
+	 */
+	bool write = (vmf->flags & FAULT_FLAG_WRITE) &&
+		(vmf->vma->vm_flags & VM_SHARED);
 
 	if (write) {
 		sb_start_pagefault(sb);
@@ -595,7 +608,7 @@ static loff_t ext4_seek_data(struct file *file, loff_t offset, loff_t maxsize)
 	inode_lock(inode);
 
 	isize = i_size_read(inode);
-	if (offset >= isize) {
+	if (offset < 0 || offset >= isize) {
 		inode_unlock(inode);
 		return -ENXIO;
 	}
@@ -658,7 +671,7 @@ static loff_t ext4_seek_hole(struct file *file, loff_t offset, loff_t maxsize)
 	inode_lock(inode);
 
 	isize = i_size_read(inode);
-	if (offset >= isize) {
+	if (offset < 0 || offset >= isize) {
 		inode_unlock(inode);
 		return -ENXIO;
 	}

fs/ext4/hash.c

@@ -148,8 +148,6 @@ static void str2hashbuf_signed(const char *msg, int len, __u32 *buf, int num)
 	if (len > num*4)
 		len = num * 4;
 	for (i = 0; i < len; i++) {
-		if ((i % 4) == 0)
-			val = pad;
 		val = ((int) scp[i]) + (val << 8);
 		if ((i % 4) == 3) {
 			*buf++ = val;
@@ -176,8 +174,6 @@ static void str2hashbuf_unsigned(const char *msg, int len, __u32 *buf, int num)
 	if (len > num*4)
 		len = num * 4;
 	for (i = 0; i < len; i++) {
-		if ((i % 4) == 0)
-			val = pad;
 		val = ((int) ucp[i]) + (val << 8);
 		if ((i % 4) == 3) {
 			*buf++ = val;

fs/ext4/ialloc.c

@@ -692,24 +692,25 @@ static int find_group_other(struct super_block *sb, struct inode *parent,
  * somewhat arbitrary...)
  */
 #define RECENTCY_MIN	5
-#define RECENTCY_DIRTY	30
+#define RECENTCY_DIRTY	300
 
 static int recently_deleted(struct super_block *sb, ext4_group_t group, int ino)
 {
 	struct ext4_group_desc	*gdp;
 	struct ext4_inode	*raw_inode;
 	struct buffer_head	*bh;
-	unsigned long		dtime, now;
-	int	inodes_per_block = EXT4_SB(sb)->s_inodes_per_block;
-	int	offset, ret = 0, recentcy = RECENTCY_MIN;
+	int inodes_per_block = EXT4_SB(sb)->s_inodes_per_block;
+	int offset, ret = 0;
+	int recentcy = RECENTCY_MIN;
+	u32 dtime, now;
 
 	gdp = ext4_get_group_desc(sb, group, NULL);
 	if (unlikely(!gdp))
 		return 0;
 
-	bh = sb_getblk(sb, ext4_inode_table(sb, gdp) +
+	bh = sb_find_get_block(sb, ext4_inode_table(sb, gdp) +
 		       (ino / inodes_per_block));
-	if (unlikely(!bh) || !buffer_uptodate(bh))
+	if (!bh || !buffer_uptodate(bh))
 		/*
 		 * If the block is not in the buffer cache, then it
 		 * must have been written out.
@@ -718,18 +719,45 @@ static int recently_deleted(struct super_block *sb, ext4_group_t group, int ino)
 
 	offset = (ino % inodes_per_block) * EXT4_INODE_SIZE(sb);
 	raw_inode = (struct ext4_inode *) (bh->b_data + offset);
+
+	/* i_dtime is only 32 bits on disk, but we only care about relative
+	 * times in the range of a few minutes (i.e. long enough to sync a
+	 * recently-deleted inode to disk), so using the low 32 bits of the
+	 * clock (a 68 year range) is enough, see time_before32() */
 	dtime = le32_to_cpu(raw_inode->i_dtime);
-	now = get_seconds();
+	now = ktime_get_real_seconds();
 	if (buffer_dirty(bh))
 		recentcy += RECENTCY_DIRTY;
 
-	if (dtime && (dtime < now) && (now < dtime + recentcy))
+	if (dtime && time_before32(dtime, now) &&
+	    time_before32(now, dtime + recentcy))
 		ret = 1;
 out:
 	brelse(bh);
 	return ret;
 }
 
+static int find_inode_bit(struct super_block *sb, ext4_group_t group,
+			  struct buffer_head *bitmap, unsigned long *ino)
+{
+next:
+	*ino = ext4_find_next_zero_bit((unsigned long *)
+				       bitmap->b_data,
+				       EXT4_INODES_PER_GROUP(sb), *ino);
+	if (*ino >= EXT4_INODES_PER_GROUP(sb))
+		return 0;
+
+	if ((EXT4_SB(sb)->s_journal == NULL) &&
+	    recently_deleted(sb, group, *ino)) {
+		*ino = *ino + 1;
+		if (*ino < EXT4_INODES_PER_GROUP(sb))
+			goto next;
+		return 0;
+	}
+
+	return 1;
+}
+
 /*
  * There are two policies for allocating an inode.  If the new inode is
  * a directory, then a forward search is made for a block group with both
@@ -892,47 +920,34 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
 		/*
 		 * Check free inodes count before loading bitmap.
 		 */
-		if (ext4_free_inodes_count(sb, gdp) == 0) {
-			if (++group == ngroups)
-				group = 0;
-			continue;
-		}
+		if (ext4_free_inodes_count(sb, gdp) == 0)
+			goto next_group;
 
 		grp = ext4_get_group_info(sb, group);
 		/* Skip groups with already-known suspicious inode tables */
-		if (EXT4_MB_GRP_IBITMAP_CORRUPT(grp)) {
-			if (++group == ngroups)
-				group = 0;
-			continue;
-		}
+		if (EXT4_MB_GRP_IBITMAP_CORRUPT(grp))
+			goto next_group;
 
 		brelse(inode_bitmap_bh);
 		inode_bitmap_bh = ext4_read_inode_bitmap(sb, group);
 		/* Skip groups with suspicious inode tables */
 		if (EXT4_MB_GRP_IBITMAP_CORRUPT(grp) ||
 		    IS_ERR(inode_bitmap_bh)) {
 			inode_bitmap_bh = NULL;
-			if (++group == ngroups)
-				group = 0;
-			continue;
+			goto next_group;
 		}
 
 repeat_in_this_group:
-		ino = ext4_find_next_zero_bit((unsigned long *)
-					      inode_bitmap_bh->b_data,
-					      EXT4_INODES_PER_GROUP(sb), ino);
-		if (ino >= EXT4_INODES_PER_GROUP(sb))
+		ret2 = find_inode_bit(sb, group, inode_bitmap_bh, &ino);
+		if (!ret2)
 			goto next_group;
-		if (group == 0 && (ino+1) < EXT4_FIRST_INO(sb)) {
+
+		if (group == 0 && (ino + 1) < EXT4_FIRST_INO(sb)) {
 			ext4_error(sb, "reserved inode found cleared - "
 				   "inode=%lu", ino + 1);
-			continue;
-		}
-		if ((EXT4_SB(sb)->s_journal == NULL) &&
-		    recently_deleted(sb, group, ino)) {
-			ino++;
-			goto next_inode;
+			goto next_group;
 		}
+
 		if (!handle) {
 			BUG_ON(nblocks <= 0);
 			handle = __ext4_journal_start_sb(dir->i_sb, line_no,
@@ -952,11 +967,23 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir,
 		}
 		ext4_lock_group(sb, group);
 		ret2 = ext4_test_and_set_bit(ino, inode_bitmap_bh->b_data);
+		if (ret2) {
+			/* Someone already took the bit. Repeat the search
+			 * with lock held.
+			 */
+			ret2 = find_inode_bit(sb, group, inode_bitmap_bh, &ino);
+			if (ret2) {
+				ext4_set_bit(ino, inode_bitmap_bh->b_data);
+				ret2 = 0;
+			} else {
+				ret2 = 1; /* we didn't grab the inode */
+			}
+		}
 		ext4_unlock_group(sb, group);
 		ino++;		/* the inode bitmap is zero-based */
 		if (!ret2)
 			goto got; /* we grabbed the inode! */
-next_inode:
+
 		if (ino < EXT4_INODES_PER_GROUP(sb))
 			goto repeat_in_this_group;
 next_group:

fs/ext4/inode.c

@@ -4897,14 +4897,6 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino)
 	brelse(iloc.bh);
 	ext4_set_inode_flags(inode);
 
-	if (ei->i_flags & EXT4_EA_INODE_FL) {
-		ext4_xattr_inode_set_class(inode);
-
-		inode_lock(inode);
-		inode->i_flags |= S_NOQUOTA;
-		inode_unlock(inode);
-	}
-
 	unlock_new_inode(inode);
 	return inode;
 

fs/ext4/mmp.c

@@ -367,7 +367,7 @@ int ext4_multi_mount_protect(struct super_block *sb,
 		goto failed;
 	}
 
-	mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL);
+	mmpd_data = kmalloc(sizeof(*mmpd_data), GFP_KERNEL);
 	if (!mmpd_data) {
 		ext4_warning(sb, "not enough memory for mmpd_data");
 		goto failed;

fs/ext4/super.c

@@ -2404,6 +2404,7 @@ static void ext4_orphan_cleanup(struct super_block *sb,
 	unsigned int s_flags = sb->s_flags;
 	int ret, nr_orphans = 0, nr_truncates = 0;
 #ifdef CONFIG_QUOTA
+	int quota_update = 0;
 	int i;
 #endif
 	if (!es->s_last_orphan) {
@@ -2442,14 +2443,32 @@ static void ext4_orphan_cleanup(struct super_block *sb,
 #ifdef CONFIG_QUOTA
 	/* Needed for iput() to work correctly and not trash data */
 	sb->s_flags |= MS_ACTIVE;
-	/* Turn on quotas so that they are updated correctly */
+
+	/*
+	 * Turn on quotas which were not enabled for read-only mounts if
+	 * filesystem has quota feature, so that they are updated correctly.
+	 */
+	if (ext4_has_feature_quota(sb) && (s_flags & MS_RDONLY)) {
+		int ret = ext4_enable_quotas(sb);
+
+		if (!ret)
+			quota_update = 1;
+		else
+			ext4_msg(sb, KERN_ERR,
+				"Cannot turn on quotas: error %d", ret);
+	}
+
+	/* Turn on journaled quotas used for old sytle */
 	for (i = 0; i < EXT4_MAXQUOTAS; i++) {
 		if (EXT4_SB(sb)->s_qf_names[i]) {
 			int ret = ext4_quota_on_mount(sb, i);
-			if (ret < 0)
+
+			if (!ret)
+				quota_update = 1;
+			else
 				ext4_msg(sb, KERN_ERR,
 					"Cannot turn on journaled "
-					"quota: error %d", ret);
+					"quota: type %d: error %d", i, ret);
 		}
 	}
 #endif
@@ -2510,10 +2529,12 @@ static void ext4_orphan_cleanup(struct super_block *sb,
 		ext4_msg(sb, KERN_INFO, "%d truncate%s cleaned up",
 		       PLURAL(nr_truncates));
 #ifdef CONFIG_QUOTA
-	/* Turn quotas off */
-	for (i = 0; i < EXT4_MAXQUOTAS; i++) {
-		if (sb_dqopt(sb)->files[i])
-			dquot_quota_off(sb, i);
+	/* Turn off quotas if they were enabled for orphan cleanup */
+	if (quota_update) {
+		for (i = 0; i < EXT4_MAXQUOTAS; i++) {
+			if (sb_dqopt(sb)->files[i])
+				dquot_quota_off(sb, i);
+		}
 	}
 #endif
 	sb->s_flags = s_flags; /* Restore MS_RDONLY status */
@@ -5512,6 +5533,9 @@ static int ext4_enable_quotas(struct super_block *sb)
 				DQUOT_USAGE_ENABLED |
 				(quota_mopt[type] ? DQUOT_LIMITS_ENABLED : 0));
 			if (err) {
+				for (type--; type >= 0; type--)
+					dquot_quota_off(sb, type);
+
 				ext4_warning(sb,
 					"Failed to enable quota tracking "
 					"(type=%d, err=%d). Please run "