jffs2: Fix corruption when flash erase/write failure
Joakim Tjernlund [Thu, 28 May 2009 15:43:59 +0000 (17:43 +0200)]
Erase errors such as:
"Newly-erased block contained word 0xa4ef223e at offset 0x0296a014"
and failure to write the clean marker,
moves the offending erase block to erasing list before calling
jffs2_erase_failed(). This is bad as jffs2_erase_failed() will
also move the block to the bad_list, but is now moving the
wrong block, causing FS corruption.

Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>

fs/jffs2/erase.c

index c32b4a1..a024474 100644 (file)
@@ -480,13 +480,6 @@ static void jffs2_mark_erased_block(struct jffs2_sb_info *c, struct jffs2_eraseb
        return;
 
 filebad:
-       mutex_lock(&c->erase_free_sem);
-       spin_lock(&c->erase_completion_lock);
-       /* Stick it on a list (any list) so erase_failed can take it
-          right off again.  Silly, but shouldn't happen often. */
-       list_move(&jeb->list, &c->erasing_list);
-       spin_unlock(&c->erase_completion_lock);
-       mutex_unlock(&c->erase_free_sem);
        jffs2_erase_failed(c, jeb, bad_offset);
        return;