Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <8d6d9329-f5f1-2f15-f578-e4f8010b9b02@gmail.com>
 <5E28252D-6EA1-4DD9-A5B3-957E13589982@redhat.com> <CAB6yy359Gvdu=v1ZvTLXPoY2EMtER3_cBVDKc4MQYhaMOjcUSw@mail.gmail.com>
 <70B9A332-FDA0-418E-81CF-962229E93AC5@redhat.com>
In-Reply-To: <70B9A332-FDA0-418E-81CF-962229E93AC5@redhat.com>
From:   Kinglong Mee <kinglongmee@gmail.com>
Date:   Sat, 15 Jul 2023 20:06:28 +0800
Message-ID: <CAB6yy36uWjnay8OTbZCvHGNih8mjCASc7yDnNeDp2KqxfA8Dzw@mail.gmail.com>
Subject: Re: [PATCH] nfs: fix redundant readdir request after get eof
To:     Benjamin Coddington <bcodding@redhat.com>
Cc:     Anna Schumaker <anna@kernel.org>,
        Linux NFS Mailing List <linux-nfs@vger.kernel.org>,
        Trond Myklebust <trondmy@hammerspace.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Fri, Jul 14, 2023 at 11:19=E2=80=AFPM Benjamin Coddington
<bcodding@redhat.com> wrote:
>
> On 13 Jul 2023, at 23:07, Kinglong Mee wrote:
>
> > Hi Ben,
> >
> ...
> > Comparing with the above one, this seems work.
>
> This fixes it for me and keeps the optimization.  Its quite a subtle bit =
of
> logic - maybe a comment is appropriate?

Thanks for your testing.
I will send a new patch with a comment.

>
> One non-intuitive thing here is
> that array->size =3D=3D 19 for a directory with 18 entries, since we coun=
t the
> "eof" entry as a blank entry instead of the last real entry.

No.
This is not a blank entry, every entry is a real one.
For the first emit, only returns 18 entries to the caller,
the next emit will return the 19th one.

thanks,
Kinglong Mee