On Tue, Dec 11, 2018 at 11:18:41AM +0100, Florian Westphal wrote:
> Steffen Klassert <steffen.klass...@secunet.com> wrote:
> > On Mon, Dec 10, 2018 at 03:50:06PM +0100, Florian Westphal wrote:
> > > }
> > > @@ -552,11 +517,6 @@ void __init xfrm_input_init(void)
> > > if (err)
> > > gro_cells.cells = NULL;
> > >
> > > - secpath_cachep = kmem_cache_create("secpath_cache",
> > > - sizeof(struct sec_path),
> > > - 0, SLAB_HWCACHE_ALIGN|SLAB_PANIC,
> > > - NULL);
> >
> > This is not so nice. Usually we need a secpath per packet for IPsec.
> > With removing the cache, we have to kmalloc a secpath for each packet.
> > This might have some performance impact.
>
> I would expect that the extension allocations come from
> kmalloc-96 cache in 'ipsec only' case.
>
> I can run a few IPSEC benchmark tests to see if there is measureable
> impact.
That would be good, thanks!
