[RESEND][PATCH] dir_index: error out instead of BUG on corrupt hash dir limit

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
From: Eric Sandeen
Date: Friday, August 31, 2007 - 7:48 pm

(resend, this one got lost?  Got an acked-by from Andreas
last go-round)

(Andrew, Ted, should I be splitting out ext3 and ext4 patches and
sending separately...?)

Thanks,
-Eric

----------

A corrupt ondisk hash dir limit will trip an assert in dx_probe,
which calls BUG().  Instead, we can just issue the warning and
fail dx_probe like the other 3 tests just before it.  Thanks
to aviro for suggesting this...  Tested with a hand-crafted
corrupt ext3 image, issues:

EXT3-fs warning (device loop0): dx_probe: Corrupt limit in dir inode 14337

vs. previous:

Assertion failure in dx_probe() at fs/ext3/namei.c:383: "dx_get_limit(entries) == dx_root_limit(dir, root->info.info_length)"
------------[ cut here ]------------
kernel BUG at fs/ext3/namei.c:383!
...


Signed-off-by: Eric Sandeen <sandeen@sandeen.net>

Index: linux-2.6.22-rc4/fs/ext3/namei.c
===================================================================
--- linux-2.6.22-rc4.orig/fs/ext3/namei.c
+++ linux-2.6.22-rc4/fs/ext3/namei.c
@@ -379,8 +379,16 @@ dx_probe(struct dentry *dentry, struct i
 
 	entries = (struct dx_entry *) (((char *)&root->info) +
 				       root->info.info_length);
-	assert(dx_get_limit(entries) == dx_root_limit(dir,
-						      root->info.info_length));
+
+	if (dx_get_limit(entries) != dx_root_limit(dir,
+						   root->info.info_length)) {
+		ext3_warning(dir->i_sb, __FUNCTION__,
+			     "Corrupt limit in dir inode %ld\n", dir->i_ino);
+		brelse(bh);
+		*err = ERR_BAD_DX_DIR;
+		goto fail;
+	}
+
 	dxtrace (printk("Look up %x", hash));
 	while (1)
 	{
Index: linux-2.6.22-rc4/fs/ext4/namei.c
===================================================================
--- linux-2.6.22-rc4.orig/fs/ext4/namei.c
+++ linux-2.6.22-rc4/fs/ext4/namei.c
@@ -379,8 +379,16 @@ dx_probe(struct dentry *dentry, struct i
 
 	entries = (struct dx_entry *) (((char *)&root->info) +
 				       root->info.info_length);
-	assert(dx_get_limit(entries) == dx_root_limit(dir,
-						      root->info.info_length));
+
+	if (dx_get_limit(entries) != dx_root_limit(dir,
+						   root->info.info_length)) {
+		ext4_warning(dir->i_sb, __FUNCTION__,
+			     "Corrupt limit in dir inode %ld\n", dir->i_ino);
+		brelse(bh);
+		*err = ERR_BAD_DX_DIR;
+		goto fail;
+	}
+
 	dxtrace (printk("Look up %x", hash));
 	while (1)
 	{

-
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[RESEND][PATCH] dir_index: error out instead of BUG on cor ..., Eric Sandeen, (Fri Aug 31, 7:48 pm)