2.4.19-pre2: ufs problems

PATCH [email protected] 2 participants 2 messages Mar 4, 2002 – Mar 5, 2002

Share & tools

[email protected]

author

2002-03-04 00:35 UTC #0 lore raw

Both Gerd Knorr and Jurgen Philippaerts complain about problems mounting an ufs filesystem, one for BSD, the other for Solaris.
The reason is the patch fragment in patch-2.4.19-pre2:

fs/ufs/super.c Wed Feb 27 20:34:30 2002 renamed from fs/ufs/super.c Thu Feb 28 18:24:57 2002 +5 -1

@@ -597,7 +597,11 @@
        }
 
 again:
-       set_blocksize (sb->s_dev, block_size);
+       if (!set_blocksize (sb->s_dev, block_size)) {
+               printk(KERN_ERR "UFS: failed to set blocksize\n");
+               goto failed;
+       }
+
        sb->s_blocksize = block_size;
 
        /*

Indeed, set_blocksize returns 0 when all is well.
Thus, this change will always cause a failure.

Andries

[so, this patch fragment must be reverted, or the '!' must be removed]

Marcelo Tosatti

2002-03-05 19:16 UTC #1 lore raw

Quoted 21 lines · On Mon, 4 Mar 2002 [email protected] wrote:

Both Gerd Knorr and Jurgen Philippaerts complain about problems mounting an ufs filesystem, one for BSD, the other for Solaris.
The reason is the patch fragment in patch-2.4.19-pre2:
fs/ufs/super.c Wed Feb 27 20:34:30 2002 renamed from fs/ufs/super.c Thu Feb 28 18:24:57 2002 +5 -1
@@ -597,7 +597,11 @@
        }
 
 again:
-       set_blocksize (sb->s_dev, block_size);
+       if (!set_blocksize (sb->s_dev, block_size)) {
+               printk(KERN_ERR "UFS: failed to set blocksize\n");
+               goto failed;
+       }
+
        sb->s_blocksize = block_size;
 
        /*

Quoted 2 lines

Indeed, set_blocksize returns 0 when all is well.
Thus, this change will always cause a failure.

Fixed in my tree.

Thanks