[PATCH][malloc] Avoid atomics in have_fastchunks

[Arjan van de Ven]

On 9/19/2017 6:57 PM, Carlos O'Donell wrote:
> 
>> An alternative would be to move all the fastbin handling into the t-cache - but
>> then I bet it's much easier just to write a fast modern allocator...
> 
> Writing a fast modern allocator is not easy :-)
> 

and since glibc 2.26, glibc's malloc() is actually no longer losing against some of those
alternative allocators as well, for example see

https://www.phoronix.com/scan.php?page=news_item&px=Glibc-2.26-Redis-Test