[PULL] bcache: fix cache_set_flush() NULL pointer dereference on OOM

[PULL] bcache: fix cache_set_flush() NULL pointer dereference on OOM

[Eric Wheeler]

Hi Jens,

Here is another OOM fix for bcache.  Please pull:

  git pull https://bitbucket.org/ewheelerinc/linux.git v4.5-rc7-bcache-fixes 

 # git diff --stat v4.5-rc7
 drivers/md/bcache/super.c | 3 +++
 1 file changed, 3 insertions(+)

-Eric

commit 87264e96886f785922d741750a8a1794e71af1e7
Author: Eric Wheeler <git@linux.ewheeler.net>
Date:   Mon Mar 7 15:17:50 2016 -0800

    bcache: fix cache_set_flush() NULL pointer dereference on OOM
    
    When bch_cache_set_alloc() fails to kzalloc the cache_set, the
    asyncronous closure handling tries to dereference a cache_set that
    hadn't yet been allocated inside of cache_set_flush() which is called
    by __cache_set_unregister() during cleanup.  This appears to happen only
    during an OOM condition on bcache_register.
    
    Signed-off-by: Eric Wheeler <bcache@linux.ewheeler.net>
    Cc: stable@vger.kernel.org

Re: [PULL] bcache: fix cache_set_flush() NULL pointer dereference on OOM

[Jens Axboe]

On 03/07/2016 04:36 PM, Eric Wheeler wrote:
> Hi Jens,
>
> Here is another OOM fix for bcache.  Please pull:
>
>    git pull https://bitbucket.org/ewheelerinc/linux.git v4.5-rc7-bcache-fixes
>
>   # git diff --stat v4.5-rc7
>   drivers/md/bcache/super.c | 3 +++
>   1 file changed, 3 insertions(+)

I pulled this and the previous, but for 4.6. They are all marked stable, 
so that should be fine.

Note that your 2nd patch in the -rc6 branch, this one:

bcache: cleaned up error handling around register_cache()

had a 4 cases of trailing white space. Please ensure that future patches 
are cleaned up, I fixed that one up.

-- 
Jens Axboe

Re: [PULL] bcache: fix cache_set_flush() NULL pointer dereference on OOM

[Eric Wheeler]

On Tue, 8 Mar 2016, Jens Axboe wrote:

> On 03/07/2016 04:36 PM, Eric Wheeler wrote:
> > Hi Jens,
> >
> > Here is another OOM fix for bcache.  Please pull:
> >
> >    git pull https://bitbucket.org/ewheelerinc/linux.git
> >    v4.5-rc7-bcache-fixes
> >
> >   # git diff --stat v4.5-rc7
> >   drivers/md/bcache/super.c | 3 +++
> >   1 file changed, 3 insertions(+)
> 
> I pulled this and the previous, but for 4.6. They are all marked stable, so
> that should be fine.
> 
> Note that your 2nd patch in the -rc6 branch, this one:
> 
> bcache: cleaned up error handling around register_cache()
> 
> had a 4 cases of trailing white space. Please ensure that future patches are
> cleaned up, I fixed that one up.

Interesting, do you mean whitespace at the end of a line with code or 
empty lines containing whitespace or both?  Something matched by [ ]+$  ?

Is it acceptable to have empty likes whith whitespace leading up to the 
indent level?  eg:
	if (1) {
		foo
		<<whitespace-to-here but line is blank
		bar
	}

-Eric

> 
> -- 
> Jens Axboe
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 

Re: [PULL] bcache: fix cache_set_flush() NULL pointer dereference on OOM

[Jens Axboe]

On 03/08/2016 08:36 PM, Eric Wheeler wrote:
> On Tue, 8 Mar 2016, Jens Axboe wrote:
>
>> On 03/07/2016 04:36 PM, Eric Wheeler wrote:
>>> Hi Jens,
>>>
>>> Here is another OOM fix for bcache.  Please pull:
>>>
>>>     git pull https://bitbucket.org/ewheelerinc/linux.git
>>>     v4.5-rc7-bcache-fixes
>>>
>>>    # git diff --stat v4.5-rc7
>>>    drivers/md/bcache/super.c | 3 +++
>>>    1 file changed, 3 insertions(+)
>>
>> I pulled this and the previous, but for 4.6. They are all marked stable, so
>> that should be fine.
>>
>> Note that your 2nd patch in the -rc6 branch, this one:
>>
>> bcache: cleaned up error handling around register_cache()
>>
>> had a 4 cases of trailing white space. Please ensure that future patches are
>> cleaned up, I fixed that one up.
>
> Interesting, do you mean whitespace at the end of a line with code or
> empty lines containing whitespace or both?  Something matched by [ ]+$  ?

Both. You had both trailing white space at the end of a line, and a 
space at the start of a new line.

> Is it acceptable to have empty likes whith whitespace leading up to the
> indent level?  eg:
> 	if (1) {
> 		foo
> 		<<whitespace-to-here but line is blank
> 		bar
> 	}

Nope

-- 
Jens Axboe