Re: [RFC PATCH] mm: net: disable kswapd for high-order network buffer allocation

[Shakeel Butt]

On Wed, Oct 15, 2025 at 04:28:17AM +0800, Barry Song wrote:
> On Tue, Oct 14, 2025 at 10:38 PM Shakeel Butt <shakeel.butt@xxxxxxxxx> wrote:
> 
> > >
> > > It might be worth exploring these settings further, but I can’t quite see
> > > their connection to high-order allocations,
> >
> > I don't think there is a connection between them. Is there a reason you
> > are expecting a connection/relation between them?
> 
> Eric replied to my email about frequent high-order allocation requests,
> suggesting that I might be missing some proper configurations for these
> settings[1]. So I’m trying to understand whether these configurations affect
> the frequency of high-order allocations.

If I understand Eric correctly, those configurations do indirectly
affect the number of memory allocations and their lifetime (irrespective
of order). In one scenario, setting tcp_wmem[0] higher, allow the kernel
to allocate more memory even when the system is under memory pressure.
See tcp_wmem_schedule(). In your case it would be up to 0.5MiB per
socket.

Have you tested the configuration values suggested by Eric?