Re: Linux 2.6.27.26

[Milan Broz <mbroz@redhat.com>]

Arkadiusz Miskiewicz wrote:
> On Friday 03 of July 2009, Greg KH wrote:
> 
> Isn't this part broken?

yes, this part is broken, patch is misapplied to wrong function...
it applies only for for dm_get_from_kobject(), not dm_blk_open()

The patch

  4d89b7b4e4726893453d0fb4ddbb5b3e16353994 
  Do not process sysfs attributes when device is being destroyed.

is not needed for 2.6.27 (dm sysfs attributes are not there yet)

Milan


> 
> The code in 2.6.27 already has such thing and new code is trying to return 
> NULL even if function returns int
> 
> static int dm_blk_open(struct inode *inode, struct file *file)
> {
>         struct mapped_device *md;
> 
>         spin_lock(&_minor_lock);
> 
>         md = inode->i_bdev->bd_disk->private_data;
>         if (!md)
>                 goto out;
> 
>         if (test_bit(DMF_FREEING, &md->flags) ||
>             test_bit(DMF_DELETING, &md->flags)) {
>                 md = NULL;
>                 goto out;
>         }
> 
>         dm_get(md);
>         atomic_inc(&md->open_count);
> 
> out:
>         spin_unlock(&_minor_lock);
> 
>         return md ? 0 : -ENXIO;
> }
> 
> 
>> diff --git a/drivers/md/dm.c b/drivers/md/dm.c
>> index ace998c..925efaf 100644
>> --- a/drivers/md/dm.c
>> +++ b/drivers/md/dm.c
>> @@ -265,6 +265,10 @@ static int dm_blk_open(struct inode *inode, struct
>> file *file) goto out;
>>         }
>>  
>> +       if (test_bit(DMF_FREEING, &md->flags) ||
>> +           test_bit(DMF_DELETING, &md->flags))
>> +               return NULL;
>> +
>>         dm_get(md);
>>         atomic_inc(&md->open_count);
>>  
> 
>