From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael S. Tsirkin" Subject: Re: On guest free page hinting and OOM Date: Tue, 2 Apr 2019 13:53:12 -0400 Message-ID: <20190402134722-mutt-send-email-mst@kernel.org> References: <20190401104608-mutt-send-email-mst@kernel.org> <6a612adf-e9c3-6aff-3285-2e2d02c8b80d@redhat.com> <20190402112115-mutt-send-email-mst@kernel.org> <3dd76ce6-c138-b019-3a43-0bb0b793690a@redhat.com> <6b0a3610-0e7b-08dc-8b5f-707062f87bea@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Hildenbrand , Nitesh Narayan Lal , kvm list , LKML , linux-mm , Paolo Bonzini , lcapitulino@redhat.com, pagupta@redhat.com, Yang Zhang , Rik van Riel , dodgen@google.com, Konrad Rzeszutek Wilk , dhildenb@redhat.com, Andrea Arcangeli , Dave Hansen To: Alexander Duyck Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On Tue, Apr 02, 2019 at 10:45:43AM -0700, Alexander Duyck wrote: > We went through this back in the day with > networking. Adding more buffers is not the solution. The solution is > to have a way to gracefully recover and keep our hinting latency and > buffer bloat to a minimum. That's an interesting approach, I think that things that end up working well are NAPI (asychronous notifications), limited batching, XDP (big aligned buffers) and BQL (accounting). Is that your perspective too? -- MST