Received-SPF: pass (google.com: domain of linux-kernel+bounces-139357-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Precedence: bulk
MIME-Version: 1.0
References: <a022f547c43e40e2b9f0aebb2bd0bfa8@kioxia.com>
In-Reply-To: <a022f547c43e40e2b9f0aebb2bd0bfa8@kioxia.com>
From: Song Liu <song@kernel.org>
Date: Wed, 10 Apr 2024 14:13:50 -0700
Message-ID: <CAPhsuW6iDjPp2B-NEqMEjb9v40US9eN88CuOc+882h6+BRCAQw@mail.gmail.com>
Subject: Re: [PATCH 00/11] md: avoid counter operation conflicts
To: tada keisuke <keisuke1.tada@kioxia.com>
Cc: "yukuai3@huawei.com" <yukuai3@huawei.com>, 
	"linux-raid@vger.kernel.org" <linux-raid@vger.kernel.org>, 
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Thanks for the patchset!

On Tue, Mar 26, 2024 at 3:29=E2=80=AFAM tada keisuke <keisuke1.tada@kioxia.=
com> wrote:
>
> Currently, active_aligned_reads and nr_pending used as counters are atomi=
c types.
> Therefore, when inc/dec in a multi-core results in conflicts.
>
> To solve this problem, use "percpu_ref" counters that can avoid conflicts=
 and maintain consistency.
>
> Switch modes of percpu_ref to achieve both consistency and conflict avoid=
ance.
> During normal operations such as inc/dec, it operates as percpu mode.
> When consistency is required, it operates as atomic mode.
> The operations that require consistency are as follows:
> =E3=83=BBZero check for the counter
> =E3=83=BBAll operations in RAID 1/10

IIUC, the patchset is to resolve race conditions and thus improve performan=
ce.
If this is the case, could you please include some performance numbers in t=
he
next version?

Thanks,
Song