Trond Myklebust wrote:
> On Wed, 2007-08-29 at 16:30 -0400, Chuck Lever wrote:
>> Yep, this is the same fix as one provided in the clean-up/bug-fix 
>> patches I just resent today.
> 
> Sure, but we shouldn't be sending pure cleanups to Linus when we're
> already in an -rc4 release. OTOH, I'd like to fix genuine bugs like
> this, so I'll probably send in James' patch for now, and then let the
> cleanups lie until the 2.6.24 merge window opens.

Bruce already (sternly) lectured me on this.

These are not just clean-ups.  They really are bug-fixes for the API. 
I'm sorry I didn't recognize the semantic distinction in the patch 
description.

As Tom said, these really are needed to make the interface useful.

>> James Lentini wrote:
>>> While experimenting with the new string mount infrastructure, I 
>>> noticed an off by one error in the computation of the hostname's 
>>> length. This patch fixes the length computation to match the same 
>>> computation in nfs4_validate_mount_data().
>>>
>>> Signed-off-by: James Lentini <jlentini@netapp.com>
>>> ---
>>>
>>>  fs/nfs/super.c |    2 +-
>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> --- fs/nfs/super.c	2007-08-29 15:06:59.320932000 -0400
>>> +++ fs/nfs/super.c	2007-08-29 15:07:51.439960000 -0400
>>> @@ -1153,7 +1153,7 @@ static int nfs_validate_mount_data(struc
>>>  		c = strchr(dev_name, ':');
>>>  		if (c == NULL)
>>>  			return -EINVAL;
>>> -		len = c - dev_name - 1;
>>> +		len = c - dev_name;
>>>  		if (len > sizeof(data->hostname))
>>>  			return -EINVAL;
>>>  		strncpy(data->hostname, dev_name, len);
>