Various bug fixes for ext4. Perhaps the most serious bug fixed is one

which could cause file system corruptions when performing file punch
 operations.
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1.4.12 (GNU/Linux)
 
 iQIcBAABCAAGBQJQ374OAAoJENNvdpvBGATwEGAP/jKUwjQhBZiF0k9dg1kQ5eTz
 bdli4fy1vxrEMIOym8IZa4nBQJVCkArwRgjc28gCBD6k9u6X3GPa26vUydsoPfP6
 odPdc9c9HtsbYQGuaq1SohID5HfjxHewTcUmCs4X4SpGcSurUcT7eQYWqSuIxFHR
 0nKk8NO4EcWh2uqIoGPrc8QpSdor0DXXYYjZmHCeVLH1n6PyoMsnrFMfO9KqMLUL
 vNR54CX9n1GRTfAfJNkNzcwfs8IfNkDUyv5hFpDh15tLltogU0TqnlAl3vSeZGSx
 vVfhwHmQTK/bJyC3YaoRZqq9CQJVk2f/OTBpJDFY/USaapuitJd6vqbmh7NiRNAN
 LaKmFt99MPfwyjEhIA7+J0LCTraAxc536q43oWWK5dAJhWI7DW0lbHARVeQTixNy
 KJ1Lp0pmmz1mX8/lugOnK1SPBF525kTaoiz2bWqg4oQgn7mBzUlgj+EV22/6Rq83
 TpKOKstl4BiZi8t5AhmFiwqtknCDiT5vUKQNy2kuM/oXtPJID/lM/TJbR5viYD3l
 AH3Ef7xj61CynFZ0oBeraGwtXc2BHJpJdWz+8uj0/VhFfC+uNUYapSLFwyiAVZKO
 xxaItT3ylfKpa0AWK6HBc2SLuL72SCHAPks06YKFtSyHtr5C8SCcafxU2DSOSi7K
 VrhkcH6STa77Br7a1ORt
 =9R/D
 -----END PGP SIGNATURE-----

Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4

Pull ext4 bug fixes from Ted Ts'o:
 "Various bug fixes for ext4.  Perhaps the most serious bug fixed is one
  which could cause file system corruptions when performing file punch
  operations."

* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: avoid hang when mounting non-journal filesystems with orphan list
  ext4: lock i_mutex when truncating orphan inodes
  ext4: do not try to write superblock on ro remount w/o journal
  ext4: include journal blocks in df overhead calcs
  ext4: remove unaligned AIO warning printk
  ext4: fix an incorrect comment about i_mutex
  ext4: fix deadlock in journal_unmap_buffer()
  ext4: split off ext4_journalled_invalidatepage()
  jbd2: fix assertion failure in jbd2_journal_flush()
  ext4: check dioread_nolock on remount
  ext4: fix extent tree corruption caused by hole punch

fs/ext4/extents.c

@@ -2226,13 +2226,14 @@ errout:
  * removes index from the index block.
  */
 static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
-			struct ext4_ext_path *path)
+			struct ext4_ext_path *path, int depth)
 {
 	int err;
 	ext4_fsblk_t leaf;
 
 	/* free index block */
-	path--;
+	depth--;
+	path = path + depth;
 	leaf = ext4_idx_pblock(path->p_idx);
 	if (unlikely(path->p_hdr->eh_entries == 0)) {
 		EXT4_ERROR_INODE(inode, "path->p_hdr->eh_entries == 0");
@@ -2257,6 +2258,19 @@ static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
 
 	ext4_free_blocks(handle, inode, NULL, leaf, 1,
 			 EXT4_FREE_BLOCKS_METADATA | EXT4_FREE_BLOCKS_FORGET);
+
+	while (--depth >= 0) {
+		if (path->p_idx != EXT_FIRST_INDEX(path->p_hdr))
+			break;
+		path--;
+		err = ext4_ext_get_access(handle, inode, path);
+		if (err)
+			break;
+		path->p_idx->ei_block = (path+1)->p_idx->ei_block;
+		err = ext4_ext_dirty(handle, inode, path);
+		if (err)
+			break;
+	}
 	return err;
 }
 
@@ -2599,7 +2613,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
 	/* if this leaf is free, then we should
 	 * remove it from index block above */
 	if (err == 0 && eh->eh_entries == 0 && path[depth].p_bh != NULL)
-		err = ext4_ext_rm_idx(handle, inode, path + depth);
+		err = ext4_ext_rm_idx(handle, inode, path, depth);
 
 out:
 	return err;
@@ -2802,7 +2816,7 @@ again:
 				/* index is empty, remove it;
 				 * handle must be already prepared by the
 				 * truncatei_leaf() */
-				err = ext4_ext_rm_idx(handle, inode, path + i);
+				err = ext4_ext_rm_idx(handle, inode, path, i);
 			}
 			/* root level has p_bh == NULL, brelse() eats this */
 			brelse(path[i].p_bh);

fs/ext4/file.c

@@ -108,14 +108,6 @@ ext4_file_dio_write(struct kiocb *iocb, const struct iovec *iov,
 
 	/* Unaligned direct AIO must be serialized; see comment above */
 	if (unaligned_aio) {
-		static unsigned long unaligned_warn_time;
-
-		/* Warn about this once per day */
-		if (printk_timed_ratelimit(&unaligned_warn_time, 60*60*24*HZ))
-			ext4_msg(inode->i_sb, KERN_WARNING,
-				 "Unaligned AIO/DIO on inode %ld by %s; "
-				 "performance will be poor.",
-				 inode->i_ino, current->comm);
 		mutex_lock(ext4_aio_mutex(inode));
 		ext4_unwritten_wait(inode);
 	}

fs/ext4/fsync.c

@@ -109,8 +109,6 @@ static int __sync_inode(struct inode *inode, int datasync)
  *
  * What we do is just kick off a commit and wait on it.  This will snapshot the
  * inode to disk.
- *
- * i_mutex lock is held when entering and exiting this function
  */
 
 int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)

fs/ext4/inode.c

@@ -2880,8 +2880,6 @@ static void ext4_invalidatepage_free_endio(struct page *page, unsigned long offs
 
 static void ext4_invalidatepage(struct page *page, unsigned long offset)
 {
-	journal_t *journal = EXT4_JOURNAL(page->mapping->host);
-
 	trace_ext4_invalidatepage(page, offset);
 
 	/*
@@ -2889,16 +2887,34 @@ static void ext4_invalidatepage(struct page *page, unsigned long offset)
 	 */
 	if (ext4_should_dioread_nolock(page->mapping->host))
 		ext4_invalidatepage_free_endio(page, offset);
+
+	/* No journalling happens on data buffers when this function is used */
+	WARN_ON(page_has_buffers(page) && buffer_jbd(page_buffers(page)));
+
+	block_invalidatepage(page, offset);
+}
+
+static int __ext4_journalled_invalidatepage(struct page *page,
+					    unsigned long offset)
+{
+	journal_t *journal = EXT4_JOURNAL(page->mapping->host);
+
+	trace_ext4_journalled_invalidatepage(page, offset);
+
 	/*
 	 * If it's a full truncate we just forget about the pending dirtying
 	 */
 	if (offset == 0)
 		ClearPageChecked(page);
 
-	if (journal)
+	return jbd2_journal_invalidatepage(journal, page, offset);
-		jbd2_journal_invalidatepage(journal, page, offset);
+}
-	else
+
-		block_invalidatepage(page, offset);
+/* Wrapper for aops... */
+static void ext4_journalled_invalidatepage(struct page *page,
+					   unsigned long offset)
+{
+	WARN_ON(__ext4_journalled_invalidatepage(page, offset) < 0);
 }
 
 static int ext4_releasepage(struct page *page, gfp_t wait)
@@ -3264,7 +3280,7 @@ static const struct address_space_operations ext4_journalled_aops = {
 	.write_end		= ext4_journalled_write_end,
 	.set_page_dirty		= ext4_journalled_set_page_dirty,
 	.bmap			= ext4_bmap,
-	.invalidatepage		= ext4_invalidatepage,
+	.invalidatepage		= ext4_journalled_invalidatepage,
 	.releasepage		= ext4_releasepage,
 	.direct_IO		= ext4_direct_IO,
 	.is_partially_uptodate  = block_is_partially_uptodate,
@@ -4304,6 +4320,47 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc)
 	return err;
 }
 
+/*
+ * In data=journal mode ext4_journalled_invalidatepage() may fail to invalidate
+ * buffers that are attached to a page stradding i_size and are undergoing
+ * commit. In that case we have to wait for commit to finish and try again.
+ */
+static void ext4_wait_for_tail_page_commit(struct inode *inode)
+{
+	struct page *page;
+	unsigned offset;
+	journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
+	tid_t commit_tid = 0;
+	int ret;
+
+	offset = inode->i_size & (PAGE_CACHE_SIZE - 1);
+	/*
+	 * All buffers in the last page remain valid? Then there's nothing to
+	 * do. We do the check mainly to optimize the common PAGE_CACHE_SIZE ==
+	 * blocksize case
+	 */
+	if (offset > PAGE_CACHE_SIZE - (1 << inode->i_blkbits))
+		return;
+	while (1) {
+		page = find_lock_page(inode->i_mapping,
+				      inode->i_size >> PAGE_CACHE_SHIFT);
+		if (!page)
+			return;
+		ret = __ext4_journalled_invalidatepage(page, offset);
+		unlock_page(page);
+		page_cache_release(page);
+		if (ret != -EBUSY)
+			return;
+		commit_tid = 0;
+		read_lock(&journal->j_state_lock);
+		if (journal->j_committing_transaction)
+			commit_tid = journal->j_committing_transaction->t_tid;
+		read_unlock(&journal->j_state_lock);
+		if (commit_tid)
+			jbd2_log_wait_commit(journal, commit_tid);
+	}
+}
+
 /*
  * ext4_setattr()
  *
@@ -4417,16 +4474,28 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
-		if (attr->ia_size != i_size_read(inode)) {
+		if (attr->ia_size != inode->i_size) {
-			truncate_setsize(inode, attr->ia_size);
+			loff_t oldsize = inode->i_size;
-			/* Inode size will be reduced, wait for dio in flight.
+
-			 * Temporarily disable dioread_nolock to prevent
+			i_size_write(inode, attr->ia_size);
-			 * livelock. */
+			/*
+			 * Blocks are going to be removed from the inode. Wait
+			 * for dio in flight.  Temporarily disable
+			 * dioread_nolock to prevent livelock.
+			 */
 			if (orphan) {
-				ext4_inode_block_unlocked_dio(inode);
+				if (!ext4_should_journal_data(inode)) {
-				inode_dio_wait(inode);
+					ext4_inode_block_unlocked_dio(inode);
-				ext4_inode_resume_unlocked_dio(inode);
+					inode_dio_wait(inode);
+					ext4_inode_resume_unlocked_dio(inode);
+				} else
+					ext4_wait_for_tail_page_commit(inode);
 			}
+			/*
+			 * Truncate pagecache after we've waited for commit
+			 * in data=journal mode to make pages freeable.
+			 */
+			truncate_pagecache(inode, oldsize, inode->i_size);
 		}
 		ext4_truncate(inode);
 	}

fs/ext4/namei.c

@@ -2648,7 +2648,8 @@ int ext4_orphan_del(handle_t *handle, struct inode *inode)
 	struct ext4_iloc iloc;
 	int err = 0;
 
-	if (!EXT4_SB(inode->i_sb)->s_journal)
+	if ((!EXT4_SB(inode->i_sb)->s_journal) &&
+	    !(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_ORPHAN_FS))
 		return 0;
 
 	mutex_lock(&EXT4_SB(inode->i_sb)->s_orphan_lock);

fs/ext4/super.c

@@ -1645,9 +1645,7 @@ static int parse_options(char *options, struct super_block *sb,
 			 unsigned int *journal_ioprio,
 			 int is_remount)
 {
-#ifdef CONFIG_QUOTA
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
-#endif
 	char *p;
 	substring_t args[MAX_OPT_ARGS];
 	int token;
@@ -1696,6 +1694,16 @@ static int parse_options(char *options, struct super_block *sb,
 		}
 	}
 #endif
+	if (test_opt(sb, DIOREAD_NOLOCK)) {
+		int blocksize =
+			BLOCK_SIZE << le32_to_cpu(sbi->s_es->s_log_block_size);
+
+		if (blocksize < PAGE_CACHE_SIZE) {
+			ext4_msg(sb, KERN_ERR, "can't mount with "
+				 "dioread_nolock if block size != PAGE_SIZE");
+			return 0;
+		}
+	}
 	return 1;
 }
 
@@ -2212,7 +2220,9 @@ static void ext4_orphan_cleanup(struct super_block *sb,
 				__func__, inode->i_ino, inode->i_size);
 			jbd_debug(2, "truncating inode %lu to %lld bytes\n",
 				  inode->i_ino, inode->i_size);
+			mutex_lock(&inode->i_mutex);
 			ext4_truncate(inode);
+			mutex_unlock(&inode->i_mutex);
 			nr_truncates++;
 		} else {
 			ext4_msg(sb, KERN_DEBUG,
@@ -3223,6 +3233,10 @@ int ext4_calculate_overhead(struct super_block *sb)
 			memset(buf, 0, PAGE_SIZE);
 		cond_resched();
 	}
+	/* Add the journal blocks as well */
+	if (sbi->s_journal)
+		overhead += EXT4_B2C(sbi, sbi->s_journal->j_maxlen);
+
 	sbi->s_overhead = overhead;
 	smp_wmb();
 	free_page((unsigned long) buf);
@@ -3436,15 +3450,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 			clear_opt(sb, DELALLOC);
 	}
 
-	blocksize = BLOCK_SIZE << le32_to_cpu(es->s_log_block_size);
-	if (test_opt(sb, DIOREAD_NOLOCK)) {
-		if (blocksize < PAGE_SIZE) {
-			ext4_msg(sb, KERN_ERR, "can't mount with "
-				 "dioread_nolock if block size != PAGE_SIZE");
-			goto failed_mount;
-		}
-	}
-
 	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
 		(test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
 
@@ -3486,6 +3491,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	if (!ext4_feature_set_ok(sb, (sb->s_flags & MS_RDONLY)))
 		goto failed_mount;
 
+	blocksize = BLOCK_SIZE << le32_to_cpu(es->s_log_block_size);
 	if (blocksize < EXT4_MIN_BLOCK_SIZE ||
 	    blocksize > EXT4_MAX_BLOCK_SIZE) {
 		ext4_msg(sb, KERN_ERR,
@@ -4725,7 +4731,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 	}
 
 	ext4_setup_system_zone(sb);
-	if (sbi->s_journal == NULL)
+	if (sbi->s_journal == NULL && !(old_sb_flags & MS_RDONLY))
 		ext4_commit_super(sb, 1);
 
 #ifdef CONFIG_QUOTA

fs/jbd2/transaction.c

@@ -209,7 +209,8 @@ repeat:
 		if (!new_transaction)
 			goto alloc_transaction;
 		write_lock(&journal->j_state_lock);
-		if (!journal->j_running_transaction) {
+		if (!journal->j_running_transaction &&
+		    !journal->j_barrier_count) {
 			jbd2_get_transaction(journal, new_transaction);
 			new_transaction = NULL;
 		}
@@ -1839,7 +1840,6 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh,
 
 	BUFFER_TRACE(bh, "entry");
 
-retry:
 	/*
 	 * It is safe to proceed here without the j_list_lock because the
 	 * buffers cannot be stolen by try_to_free_buffers as long as we are
@@ -1934,14 +1934,11 @@ retry:
 		 * for commit and try again.
 		 */
 		if (partial_page) {
-			tid_t tid = journal->j_committing_transaction->t_tid;
-
 			jbd2_journal_put_journal_head(jh);
 			spin_unlock(&journal->j_list_lock);
 			jbd_unlock_bh_state(bh);
 			write_unlock(&journal->j_state_lock);
-			jbd2_log_wait_commit(journal, tid);
+			return -EBUSY;
-			goto retry;
 		}
 		/*
 		 * OK, buffer won't be reachable after truncate. We just set
@@ -2002,21 +1999,23 @@ zap_buffer_unlocked:
  * @page:    page to flush
  * @offset:  length of page to invalidate.
  *
- * Reap page buffers containing data after offset in page.
+ * Reap page buffers containing data after offset in page. Can return -EBUSY
- *
+ * if buffers are part of the committing transaction and the page is straddling
+ * i_size. Caller then has to wait for current commit and try again.
  */
-void jbd2_journal_invalidatepage(journal_t *journal,
+int jbd2_journal_invalidatepage(journal_t *journal,
-		      struct page *page,
+				struct page *page,
-		      unsigned long offset)
+				unsigned long offset)
 {
 	struct buffer_head *head, *bh, *next;
 	unsigned int curr_off = 0;
 	int may_free = 1;
+	int ret = 0;
 
 	if (!PageLocked(page))
 		BUG();
 	if (!page_has_buffers(page))
-		return;
+		return 0;
 
 	/* We will potentially be playing with lists other than just the
 	 * data lists (especially for journaled data mode), so be
@@ -2030,9 +2029,11 @@ void jbd2_journal_invalidatepage(journal_t *journal,
 		if (offset <= curr_off) {
 			/* This block is wholly outside the truncation point */
 			lock_buffer(bh);
-			may_free &= journal_unmap_buffer(journal, bh,
+			ret = journal_unmap_buffer(journal, bh, offset > 0);
-							 offset > 0);
 			unlock_buffer(bh);
+			if (ret < 0)
+				return ret;
+			may_free &= ret;
 		}
 		curr_off = next_off;
 		bh = next;
@@ -2043,6 +2044,7 @@ void jbd2_journal_invalidatepage(journal_t *journal,
 		if (may_free && try_to_free_buffers(page))
 			J_ASSERT(!page_has_buffers(page));
 	}
+	return 0;
 }
 
 /*

include/linux/jbd2.h

@@ -1098,7 +1098,7 @@ void		 jbd2_journal_set_triggers(struct buffer_head *,
 extern int	 jbd2_journal_dirty_metadata (handle_t *, struct buffer_head *);
 extern int	 jbd2_journal_forget (handle_t *, struct buffer_head *);
 extern void	 journal_sync_buffer (struct buffer_head *);
-extern void	 jbd2_journal_invalidatepage(journal_t *,
+extern int	 jbd2_journal_invalidatepage(journal_t *,
 				struct page *, unsigned long);
 extern int	 jbd2_journal_try_to_free_buffers(journal_t *, struct page *, gfp_t);
 extern int	 jbd2_journal_stop(handle_t *);

include/trace/events/ext4.h

@@ -451,7 +451,7 @@ DEFINE_EVENT(ext4__page_op, ext4_releasepage,
 	TP_ARGS(page)
 );
 
-TRACE_EVENT(ext4_invalidatepage,
+DECLARE_EVENT_CLASS(ext4_invalidatepage_op,
 	TP_PROTO(struct page *page, unsigned long offset),
 
 	TP_ARGS(page, offset),
@@ -477,6 +477,18 @@ TRACE_EVENT(ext4_invalidatepage,
 		  (unsigned long) __entry->index, __entry->offset)
 );
 
+DEFINE_EVENT(ext4_invalidatepage_op, ext4_invalidatepage,
+	TP_PROTO(struct page *page, unsigned long offset),
+
+	TP_ARGS(page, offset)
+);
+
+DEFINE_EVENT(ext4_invalidatepage_op, ext4_journalled_invalidatepage,
+	TP_PROTO(struct page *page, unsigned long offset),
+
+	TP_ARGS(page, offset)
+);
+
 TRACE_EVENT(ext4_discard_blocks,
 	TP_PROTO(struct super_block *sb, unsigned long long blk,
 			unsigned long long count),