Evan,

The changes look good now.


thanks,
-ethan


Evan Layton wrote:
> Evan Layton wrote:
>> Ethan Quach wrote:
>>> Evan,
>>>
>>> be_mount.c
>>> ------------------
>>> 965 - I don't think you need to call to zfs_prop_set() here.
>>> We haven't changed the mountpoint yet.
>>
>> @$%$# left over cruft from other changes. Nice catch!
>>
>>>
>>> 1005 - In the error message, you want to use *orig_mntpnt
>>> here instead of *tmp_mnpnt because this is where we're trying
>>> to reset it back to the original mountpoint.
>>
>> Right it should have been *orig_mntpnt, changed.
>>
>> I also noticed while giving these error conditions another
>> look that I had missed setting *tmp_mntpnt and *orig_mntpnt
>> to NULL after freeing them so I fixed this as well.
>>
>>>
>>> 1029-1030 - this comment seems wrong.  In be_mount_pool(),
>>> we always populate orig_mntpnt, so this wouldn't ever be
>>> NULL.  For this to be true, you need to move line 970 up
>>> to be right after line 954, which I think would work.
>>
>> Yes it now reads:
>>
>>  *              orig_mntpnt - The original mountpoint for the pool 
>> this is
>>  *                            set the dataset mountpoint property 
>> back to
>>  *                            it's original value in the case where a
>>  *                            temporary mountpoint was used.
>
> that got a bit garbled...
>
>  *              orig_mntpnt - The original mountpoint for the pool. 
> This is
>  *                            used to set the dataset mountpoint property
>  *                            back to it's original value in the case 
> where a
>  *                            temporary mountpoint was used.
>
>
> I also moved lines 955 to 967 to after line 968 since we really don't 
> need to set orig_mntpnt unless we really do need to use a temporary 
> mount point.
>
> -evan

Reply via email to