[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Zl5eat0sh7rrspUG@google.com>
Date: Mon, 3 Jun 2024 17:23:06 -0700
From: Sean Christopherson <seanjc@...gle.com>
To: James Houghton <jthoughton@...gle.com>
Cc: Yu Zhao <yuzhao@...gle.com>, Andrew Morton <akpm@...ux-foundation.org>,
Paolo Bonzini <pbonzini@...hat.com>, Albert Ou <aou@...s.berkeley.edu>,
Ankit Agrawal <ankita@...dia.com>, Anup Patel <anup@...infault.org>,
Atish Patra <atishp@...shpatra.org>, Axel Rasmussen <axelrasmussen@...gle.com>,
Bibo Mao <maobibo@...ngson.cn>, Catalin Marinas <catalin.marinas@....com>,
David Matlack <dmatlack@...gle.com>, David Rientjes <rientjes@...gle.com>,
Huacai Chen <chenhuacai@...nel.org>, James Morse <james.morse@....com>,
Jonathan Corbet <corbet@....net>, Marc Zyngier <maz@...nel.org>, Michael Ellerman <mpe@...erman.id.au>,
Nicholas Piggin <npiggin@...il.com>, Oliver Upton <oliver.upton@...ux.dev>,
Palmer Dabbelt <palmer@...belt.com>, Paul Walmsley <paul.walmsley@...ive.com>,
Raghavendra Rao Ananta <rananta@...gle.com>, Ryan Roberts <ryan.roberts@....com>,
Shaoqin Huang <shahuang@...hat.com>, Shuah Khan <shuah@...nel.org>,
Suzuki K Poulose <suzuki.poulose@....com>, Tianrui Zhao <zhaotianrui@...ngson.cn>,
Will Deacon <will@...nel.org>, Zenghui Yu <yuzenghui@...wei.com>, kvm-riscv@...ts.infradead.org,
kvm@...r.kernel.org, kvmarm@...ts.linux.dev,
linux-arm-kernel@...ts.infradead.org, linux-doc@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-kselftest@...r.kernel.org,
linux-mips@...r.kernel.org, linux-mm@...ck.org,
linux-riscv@...ts.infradead.org, linuxppc-dev@...ts.ozlabs.org,
loongarch@...ts.linux.dev
Subject: Re: [PATCH v4 2/7] mm: multi-gen LRU: Have secondary MMUs participate
in aging
On Mon, Jun 03, 2024, James Houghton wrote:
> On Mon, Jun 3, 2024 at 4:03 PM Sean Christopherson <seanjc@...gle.com> wrote:
> > But before we do that, I think we need to perform due dilegence (or provide data)
> > showing that having KVM take mmu_lock for write in the "fast only" API provides
> > better total behavior. I.e. that the additional accuracy is indeed worth the cost.
>
> That sounds good to me. I'll drop the Kconfig. I'm not really sure
> what to do about the self-test, but that's not really all that
> important.
Enable it only on architectures+setups that are guaranteed to implement the
fast-only API? E.g. on x86, it darn well better be active if the TDP MMU is
enabled. If the test fails because that doesn't hold true, then we _want_ the
failure.
Powered by blists - more mailing lists