Re: [PATCH 2/3] zsmalloc: add generic path and remove x86 dependency

[Minchan Kim]

On 06/28/2012 04:09 AM, Seth Jennings wrote:

> On 06/27/2012 12:28 AM, Minchan Kim wrote:
>>> +{
>>> +	if (area->vm)
>>> +		return 0;
>>
>>
>> Just out of curiosity.
>> When do we need above check?
> 
> I did this in the case that there was a race between the for
> loop in zs_init(), calling zs_cpu_notifier(), and a CPU
> coming online.  I've never seen the condition hit, but if it
> did, it would leak memory without this check.
> 


Could you add this as a comment?



> I would move the cpu notifier registration after the loop in
> zs_init(), but then I could miss a cpu up event and we
> wouldn't have the needed per-cpu resources for mapping.
> 
> All other suggestions are accepted.  Thanks for the feedback!
> 



Thanks!

-- 
Kind regards,
Minchan Kim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/