public inbox for virtio-dev@lists.linux.dev
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Heng Qi <hengqi@linux.alibaba.com>
Cc: virtio-dev@lists.oasis-open.org,
	virtio-comment@lists.oasis-open.org, jasowang@redhat.com,
	parav@nvidia.com, xuanzhuo@linux.alibaba.com
Subject: [virtio-dev] Re: [PATCH] virtio-net: support the outer source port hash for UDP-encapsulated packets
Date: Mon, 7 Aug 2023 10:56:41 -0400	[thread overview]
Message-ID: <20230807104317-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <20230728084635.19916-1-hengqi@linux.alibaba.com>

On Fri, Jul 28, 2023 at 04:46:35PM +0800, Heng Qi wrote:
> UDP tunneling protocols usually identify a flow using the outer source UDP
> port, which is useful for processing UDP-encapsulated packets, such as
> increasing entropy for RSS queue selection or hashing the same flow to the
> same receive queue. Therefore, we extend it as a new RSS configuration.
>

that's not a real motivation is it? it's already included for
increasing entropy.

the real motivation is ignoring source and destination ip
because they can change if tunneling gateway changes, right?

 
> Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
> Reviewed-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
> ---
>  device-types/net/description.tex        | 49 ++++++++++++++++++++++++-
>  device-types/net/device-conformance.tex |  1 +
>  2 files changed, 48 insertions(+), 2 deletions(-)
> 
> diff --git a/device-types/net/description.tex b/device-types/net/description.tex
> index 76585b0..366775a 100644
> --- a/device-types/net/description.tex
> +++ b/device-types/net/description.tex
> @@ -88,6 +88,9 @@ \subsection{Feature bits}\label{sec:Device Types / Network Device / Feature bits
>  \item[VIRTIO_NET_F_CTRL_MAC_ADDR(23)] Set MAC address through control
>      channel.
>  
> +\item[VIRTIO_NET_F_HASH_UDP_TUNNEL_SRCPORT(50)] Device supports outer source port hash
> +    for UDP-encapsulated packets.
> +
>  \item[VIRTIO_NET_F_HASH_TUNNEL(51)] Device supports inner header hash for encapsulated packets.
>  
>  \item[VIRTIO_NET_F_VQ_NOTF_COAL(52)] Device supports virtqueue notification coalescing.
> @@ -150,6 +153,7 @@ \subsubsection{Feature bit requirements}\label{sec:Device Types / Network Device
>  \item[VIRTIO_NET_F_RSS] Requires VIRTIO_NET_F_CTRL_VQ.
>  \item[VIRTIO_NET_F_VQ_NOTF_COAL] Requires VIRTIO_NET_F_CTRL_VQ.
>  \item[VIRTIO_NET_F_HASH_TUNNEL] Requires VIRTIO_NET_F_CTRL_VQ along with VIRTIO_NET_F_RSS or VIRTIO_NET_F_HASH_REPORT.
> +\item[VIRTIO_NET_F_HASH_UDP_TUNNEL_SRCPORT] Requires VIRTIO_NET_F_CTRL_VQ along with VIRTIO_NET_F_RSS or VIRTIO_NET_F_HASH_REPORT.
>  \end{description}
>  
>  \subsubsection{Legacy Interface: Feature bits}\label{sec:Device Types / Network Device / Feature bits / Legacy Interface: Feature bits}
> @@ -927,11 +931,29 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
>  #define VIRTIO_NET_HASH_TYPE_UDP_EX            (1 << 8)
>  \end{lstlisting}
>  
> +A UDP-encapsulated packet uses UDP as the transport protocol in the outer header
> +\ref{sec:Device Types / Network Device / Device Operation / Processing of Incoming Packets / Hash calculation for incoming packets / Encapsulated packet}.
> +UDP tunneling protocols usually use the outer source UDP port to identify a specific flow.
> +The hash calculated over the outer source UDP port
> can thus be used by the device to queue these different flows

what does this "can be used" mean here is unclear.

> +on different receive queues or queue the same specific flow on the same receive queue.

different from what? same as what?


> +
> +If additionally the feature VIRTIO_NET_F_HASH_UDP_TUNNEL_SRCPORT was negotiated,
> +the following hash types are applicable for UDP-encapsulated packets:
> +\begin{lstlisting}
> +#define VIRTIO_NET_HASH_TYPE_UDPv4_TUNNEL_SRCPORT (1 << 9)
> +#define VIRTIO_NET_HASH_TYPE_UDPv6_TUNNEL_SRCPORT (1 << 10)
> +\end{lstlisting}
> +
>  \subparagraph{IPv4 packets}
>  \label{sec:Device Types / Network Device / Device Operation / Processing of Incoming Packets / Hash calculation for incoming packets / IPv4 packets}
>  The device calculates the hash on IPv4 packets according to 'Enabled hash types' bitmask as follows:
>  \begin{itemize}
> -\item If VIRTIO_NET_HASH_TYPE_TCPv4 is set and the packet has
> +\item If additionally VIRTIO_NET_HASH_TYPE_UDPv4_TUNNEL_SRCPORT is set and a received encapsulated

what makes the packet "encapsulated" here is unclear.

> +packet has an outer UDPv4 header, the hash is calculated over the following fields:
> +\begin{itemize}
> +\item Outer source UDP port
> +\end{itemize}
> +\item Else if VIRTIO_NET_HASH_TYPE_TCPv4 is set and the packet has
>  a TCP header, the hash is calculated over the following fields:
>  \begin{itemize}
>  \item Source IP address
> @@ -961,7 +983,12 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
>  The device calculates the hash on IPv6 packets without extension
>  headers according to 'Enabled hash types' bitmask as follows:
>  \begin{itemize}
> -\item If VIRTIO_NET_HASH_TYPE_TCPv6 is set and the packet has
> +\item If additionally VIRTIO_NET_HASH_TYPE_UDPv6_TUNNEL_SRCPORT is set and a received encapsulated
> +packet has an outer UDPv6 header, the hash is calculated over the following fields:
> +\begin{itemize}
> +\item Outer source UDP port
> +\end{itemize}
> +\item Else if VIRTIO_NET_HASH_TYPE_TCPv6 is set and the packet has
>  a TCPv6 header, the hash is calculated over the following fields:
>  \begin{itemize}
>  \item Source IPv6 address
> @@ -991,6 +1018,11 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
>  The device calculates the hash on IPv6 packets with extension
>  headers according to 'Enabled hash types' bitmask as follows:
>  \begin{itemize}
> +\item If additionally VIRTIO_NET_HASH_TYPE_UDPv6_TUNNEL_SRCPORT is set and a received encapsulated
> +packet has an outer UDPv6 header, the hash is calculated over the following fields:
> +\begin{itemize}
> +\item Outer source UDP port
> +\end{itemize}
>  \item If VIRTIO_NET_HASH_TYPE_TCP_EX is set and the packet
>  has a TCPv6 header, the hash is calculated over the following fields:
>  \begin{itemize}
> @@ -1139,8 +1171,21 @@ \subsubsection{Processing of Incoming Packets}\label{sec:Device Types / Network
>  #define VIRTIO_NET_HASH_REPORT_IPv6_EX         7
>  #define VIRTIO_NET_HASH_REPORT_TCPv6_EX        8
>  #define VIRTIO_NET_HASH_REPORT_UDPv6_EX        9
> +#define VIRTIO_NET_HASH_REPORT_UDPv4_TUNNEL_SRCPORT 10 /* Only if additionally VIRTIO_NET_F_HASH_UDP_TUNNEL_SRCPORT was negotiated */
> +#define VIRTIO_NET_HASH_REPORT_UDPv6_TUNNEL_SRCPORT 11 /* Only if additionally VIRTIO_NET_F_HASH_UDP_TUNNEL_SRCPORT was negotiated */
>  \end{lstlisting}
>  
> +\devicenormative{\subparagraph}{Outer Source UDP Port}{Device Types / Network Device / Device Operation / Outer Source UDP Port}
> +
> +When all the following cases occur:
> +\begin{itemize}
> +\item Both VIRTIO_NET_F_HASH_UDP_TUNNEL_SRCPORT and VIRTIO_NET_F_HASH_TUNNEL are negotiated.
> +\item A received UDP-encapsulated packet's outer header matches one of the encapsulation types enabled in \field{enabled_tunnel_types}.
> +\item VIRTIO_NET_HASH_TYPE_UDPv4_TUNNEL_SRCPORT is set in \field{hash_types} and a received encapsulated packet has an outer UDPv4 header or
> +      VIRTIO_NET_HASH_TYPE_UDPv6_TUNNEL_SRCPORT is set in \field{hash_types} and a received encapsulated packet has an outer UDPv6 header.
> +\end{itemize}
> +the device MUST calculate the hash over the inner header rather than the outer source UDP port.
> +
>  \subsubsection{Control Virtqueue}\label{sec:Device Types / Network Device / Device Operation / Control Virtqueue}
>  
>  The driver uses the control virtqueue (if VIRTIO_NET_F_CTRL_VQ is
> diff --git a/device-types/net/device-conformance.tex b/device-types/net/device-conformance.tex
> index f88f48b..4fd036b 100644
> --- a/device-types/net/device-conformance.tex
> +++ b/device-types/net/device-conformance.tex
> @@ -7,6 +7,7 @@
>  \item \ref{devicenormative:Device Types / Network Device / Device Operation / Packet Transmission}
>  \item \ref{devicenormative:Device Types / Network Device / Device Operation / Setting Up Receive Buffers}
>  \item \ref{devicenormative:Device Types / Network Device / Device Operation / Processing of Incoming Packets}
> +\item \ref{devicenormative:Device Types / Network Device / Device Operation / Outer Source UDP Port}
>  \item \ref{devicenormative:Device Types / Network Device / Device Operation / Control Virtqueue / Packet Receive Filtering}
>  \item \ref{devicenormative:Device Types / Network Device / Device Operation / Control Virtqueue / Setting MAC Address Filtering}
>  \item \ref{devicenormative:Device Types / Network Device / Device Operation / Control Virtqueue / VLAN Filtering}
> -- 
> 2.19.1.6.gb485710b


---------------------------------------------------------------------
To unsubscribe, e-mail: virtio-dev-unsubscribe@lists.oasis-open.org
For additional commands, e-mail: virtio-dev-help@lists.oasis-open.org


  parent reply	other threads:[~2023-08-07 14:57 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-28  8:46 [virtio-dev] [PATCH] virtio-net: support the outer source port hash for UDP-encapsulated packets Heng Qi
2023-07-28  8:53 ` [virtio-dev] Re: [virtio-comment] " Heng Qi
2023-08-07  7:34 ` [virtio-dev] " Heng Qi
2023-08-07 14:56 ` Michael S. Tsirkin [this message]
2023-08-08  2:42   ` [virtio-dev] Re: [virtio-comment] " Heng Qi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230807104317-mutt-send-email-mst@kernel.org \
    --to=mst@redhat.com \
    --cc=hengqi@linux.alibaba.com \
    --cc=jasowang@redhat.com \
    --cc=parav@nvidia.com \
    --cc=virtio-comment@lists.oasis-open.org \
    --cc=virtio-dev@lists.oasis-open.org \
    --cc=xuanzhuo@linux.alibaba.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox