The Linux Kernel Mailing List
 help / color / mirror / Atom feed
* [patch] ext3 inode accounting fix
@ 2002-11-07  3:23 Andrew Morton
  0 siblings, 0 replies; only message in thread
From: Andrew Morton @ 2002-11-07  3:23 UTC (permalink / raw)
  To: Linus Torvalds, lkml


ext3 in the 2.5.46 kernel is double-counting inodes in the per-blockgroup
accounting structures.  When a blockgroup fills up this triggers a
consistency check and the filesystem is remounted read-only.

There is no data loss, but all ext3 filesystems which have been mounted
read/write under the 2.5.46 kernel will have incorrect accounting.   They
must be fixed with a fsck.  If this is not done, the filesystem will
be remounted read-only at some time in the future - when a blockgroup
runs out of inodes and the consistency check detects the error.

Step 1: patch the 2.5.46 kernel with the below fix

Step 2: reboot into the new kernel, forcing a fsck against all ext3
        filesystems.

Alternatively, wait for 2.5.47 and then fsck all filesystems.  The problem
only manifests if all of a blockgroup's inodes are consumed, and that is
rare.


--- 25/fs/ext3/ialloc.c~ext3-inodes-count-fix	Wed Nov  6 16:16:55 2002
+++ 25-akpm/fs/ext3/ialloc.c	Wed Nov  6 16:24:20 2002
@@ -227,11 +227,6 @@ static int find_group_dir(struct super_b
 	}
 	if (!best_desc)
 		return -1;
-	best_desc->bg_free_inodes_count =
-		cpu_to_le16(le16_to_cpu(best_desc->bg_free_inodes_count) - 1);
-	best_desc->bg_used_dirs_count =
-		cpu_to_le16(le16_to_cpu(best_desc->bg_used_dirs_count) + 1);
-	mark_buffer_dirty(best_bh);
 	return best_group;
 }
 
@@ -355,14 +350,7 @@ fallback:
 	}
 
 	return -1;
-
 found:
-	desc->bg_free_inodes_count =
-		cpu_to_le16(le16_to_cpu(desc->bg_free_inodes_count) - 1);
-	desc->bg_used_dirs_count =
-		cpu_to_le16(le16_to_cpu(desc->bg_used_dirs_count) + 1);
-	sbi->s_dir_count++;
-	mark_buffer_dirty(bh);
 	return group;
 }
 
@@ -410,9 +398,6 @@ static int find_group_other(struct super
 	return -1;
 
 found:
-	desc->bg_free_inodes_count =
-		cpu_to_le16(le16_to_cpu(desc->bg_free_inodes_count) - 1);
-	mark_buffer_dirty(bh);
 	return group;
 }
 
@@ -521,9 +506,11 @@ repeat:
 	if (err) goto fail;
 	gdp->bg_free_inodes_count =
 		cpu_to_le16(le16_to_cpu(gdp->bg_free_inodes_count) - 1);
-	if (S_ISDIR(mode))
+	if (S_ISDIR(mode)) {
 		gdp->bg_used_dirs_count =
 			cpu_to_le16(le16_to_cpu(gdp->bg_used_dirs_count) + 1);
+		EXT3_SB(sb)->s_dir_count++;
+	}
 	BUFFER_TRACE(bh2, "call ext3_journal_dirty_metadata");
 	err = ext3_journal_dirty_metadata(handle, bh2);
 	if (err) goto fail;

_

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2002-11-07  3:16 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-11-07  3:23 [patch] ext3 inode accounting fix Andrew Morton

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox