Re: [PATCH 1/1] tty: vt: Remove redundant null check before kfree.

[Jiri Slaby <jslaby@suse.cz>]

On 11/22/2012 12:00 AM, Greg KH wrote:
> On Wed, Nov 21, 2012 at 11:43:34PM +0100, Jiri Slaby wrote:
>> On 11/21/2012 03:46 PM, Greg KH wrote:
>>> On Wed, Nov 21, 2012 at 10:49:07AM +0530, Sachin Kamat wrote:
>>>> --- a/drivers/tty/vt/consolemap.c
>>>> +++ b/drivers/tty/vt/consolemap.c
>>>> @@ -410,10 +410,8 @@ static void con_release_unimap(struct uni_pagedir *p)
>>>>  		kfree(p->inverse_translations[i]);
>>>>  		p->inverse_translations[i] = NULL;
>>>>  	}
>>>> -	if (p->inverse_trans_unicode) {
>>>> -		kfree(p->inverse_trans_unicode);
>>>> -		p->inverse_trans_unicode = NULL;
>>>> -	}
>>>> +	kfree(p->inverse_trans_unicode);
>>>> +	p->inverse_trans_unicode = NULL;
>>>
>>> kfree with NULL is a no-op, but the line after that just caused a kernel
>>> crash if it was NULL, so I can't accept this type of thing.
>>
>> Greg, I'm not sure -- what do you mean here? The change actually looks
>> fine to me... We do not dereference p->inverse_trans_unicode there.
> 
> If we never dereference it, why is it being set to NULL?

I'm not disputing the assignment. Let it there. But I do not see any
issues with the patch. What problem do you see exactly?


-- 
js
suse labs