Re: [net-next rfc 1/3] net: avoid high order memory allocation for queues by using flex array

["Michael S. Tsirkin" <mst@redhat.com>]

On Wed, Jun 19, 2013 at 02:56:03AM -0700, Eric Dumazet wrote:
> On Wed, 2013-06-19 at 12:11 +0300, Michael S. Tsirkin wrote:
> 
> > Well KVM supports up to 160 VCPUs on x86.
> > 
> > Creating a queue per CPU is very reasonable, and
> > assuming cache line size of 64 bytes, netdev_queue seems to be 320
> > bytes, that's 320*160 = 51200. So 12.5 pages, order-4 allocation.
> > I agree most people don't have such systems yet, but
> > they do exist.
> 
> Even so, it will just work, like a fork() is likely to work, even if a
> process needs order-1 allocation for kernel stack.
> 
> Some drivers still use order-10 allocations with kmalloc(), and nobody
> complained yet.
> 
> We had complains with mlx4 driver lately only bcause kmalloc() now gives
> a warning if allocations above MAX_ORDER are attempted.
> 
> Having a single pointer means that we can :
> 
> - Attempts a regular kmalloc() call, it will work most of the time.
> - fallback to vmalloc() _if_ kmalloc() failed.

That's a good trick too - vmalloc memory is a bit slower
on x86 since it's not using a huge page, but that's only
when we have lots of CPUs/queues...

Short term - how about switching to vmalloc if > 32 queues?

> Frankly, if you want one tx queue per cpu, I would rather use
> NETIF_F_LLTX, like some other virtual devices.
> 
> This way, you can have real per cpu memory, with proper NUMA affinity.
> 
>