Re: [Qemu-devel] [PATCH v5 13/13] cpu-exec: replace cpu->queued_work with GArray

[Alex Bennée]

Emilio G. Cota <address@hidden> writes:

> On Tue, Aug 02, 2016 at 18:27:44 +0100, Alex Bennée wrote:
>> Under times of high memory stress the additional small mallocs by a
>> linked list are source of potential memory fragmentation. As we have
>> worked hard to avoid mallocs elsewhere when queuing work we might as
>> well do the same for the list. We convert the lists to a auto-resizeing
>> GArray which will re-size in steps of powers of 2.
>
> Would be nice to see numbers on how this compares to simply using
> tcmalloc/jemalloc (or the glibc allocator, really).

glib just uses the glibc malloc/realloc underneath so it is all the same
allocator just a different usage pattern.

I was trying to find a decent way to measure the allocation usage and
fragmentation other than watching the differential in htop's memory
usage display with the two methods. Any ideas/suggestions?

>
> Thanks,
>
>               Emilio


--
Alex Bennée