Received-SPF: pass (google.com: domain of linux-kernel+bounces-91844-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Precedence: bulk
MIME-Version: 1.0
References: <20240207054701.616094-1-lulu@redhat.com> <20240207054701.616094-6-lulu@redhat.com>
 <CACGkMEuGwVwstzCGyYs90YjtE-QZGs1ztDKF4-gqFM_JZhSjmQ@mail.gmail.com>
In-Reply-To: <CACGkMEuGwVwstzCGyYs90YjtE-QZGs1ztDKF4-gqFM_JZhSjmQ@mail.gmail.com>
From: Cindy Lu <lulu@redhat.com>
Date: Tue, 5 Mar 2024 15:50:55 +0800
Message-ID: <CACLfguW1So+TtB9V04CB21jo_BTACZSX23S+_tASHsoM=_btVg@mail.gmail.com>
Subject: Re: [PATCH v4 5/5] Documentation: Add reconnect process for VDUSE
To: Jason Wang <jasowang@redhat.com>
Cc: mst@redhat.com, xieyongji@bytedance.com, linux-kernel@vger.kernel.org, 
	maxime.coquelin@redhat.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, Mar 4, 2024 at 1:11=E2=80=AFPM Jason Wang <jasowang@redhat.com> wro=
te:
>
> On Wed, Feb 7, 2024 at 1:47=E2=80=AFPM Cindy Lu <lulu@redhat.com> wrote:
> >
> > Add a document explaining the reconnect process, including what the
> > Userspace App needs to do and how it works with the kernel.
> >
> > Signed-off-by: Cindy Lu <lulu@redhat.com>
> > ---
> >  Documentation/userspace-api/vduse.rst | 32 +++++++++++++++++++++++++++
> >  1 file changed, 32 insertions(+)
> >
> > diff --git a/Documentation/userspace-api/vduse.rst b/Documentation/user=
space-api/vduse.rst
> > index bdb880e01132..a2be85e0e516 100644
> > --- a/Documentation/userspace-api/vduse.rst
> > +++ b/Documentation/userspace-api/vduse.rst
> > @@ -231,3 +231,35 @@ able to start the dataplane processing as follows:
> >     after the used ring is filled.
> >
> >  For more details on the uAPI, please see include/uapi/linux/vduse.h.
> > +
> > +HOW VDUSE devices reconnectoin works
>
> Typos, let's use a spell checker.
>
> > +----------------
> > +0. Userspace APP checks if the device /dev/vduse/vduse_name exists.
> > +   If it does not exist, need to create the instance.goto step 1
> > +   If it does exist, it means this is a reconnect and goto step 3.
> > +
> > +1. Create a new VDUSE instance with ioctl(VDUSE_CREATE_DEV) on
> > +   /dev/vduse/control.
> > +
> > +2. When the ioctl(VDUSE_CREATE_DEV) function is called, the kernel all=
ocates memory
> > +   to save the reconnect information.
> > +
> > +3. Userspace App need to mmap the pages to userspace
>
> Need to describe what kind of pages need to be mapped. And what's more:
>
> Does this require cooperation from the disconnected application? If
> yes, how to distinguish from the one that is not cooperative (doesn't
> support reconnection)?
>
> > +   Userspace App need to map Pages 0 to vq_number for vq status,
> > +   Users can define the structure for saving the reconnect information=
 themselves
> > +   in the userspace.
> > +
> > +4. Check if the infomatin sutiable for reconnect
>
> Typos again.
>
> > +   If this is reconnect:
> > +   Before attempting to reconnect, The userspace application need to t=
he
> > +   ioctl VDUSE_DEV_GET_CONFIG,VDUSE_DEV_GET_STATUS,VDUSE_DEV_GET_FEATU=
RES...
>
> "..." is not good, let's describe the steps in detail.
>
> > +   to get the and confirm if these information are suitable for reconn=
ecting.
> > +
> > +5. Start the userspace App.
>
> I had a hard time understanding this, does it mean the application
> needs to restart itself?
>
> > +   While running, the application should store the relevant informatio=
n about
> > +   reconnections in mapped pages.
> > +   When calling ioctl VDUSE_VQ_GET_INFO from the userspace APP
>
> Abbrev is not good here. Let's use "application" for consistency.
>
> > to get vq information, it is necessary
> > +   to check if this is a reconnection.
>
> How to check whether or not it is reconnected?
>
> > If a reconnection has occurred, the vq-related information
> > +   must be get from the mapped pages.
> > +
> > +6. When the Userspace App exits, it is necessary to unmap all the reco=
nnect pages.
> > --
> > 2.43.0
> >
>
> Thanks
Thanks Jason, I will rewrite this part and send the new version soon

Thanks
cindy
>