netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stanislav Fomichev <stfomichev@gmail.com>
To: Jakub Kicinski <kuba@kernel.org>
Cc: davem@davemloft.net, netdev@vger.kernel.org, edumazet@google.com,
	pabeni@redhat.com, andrew+netdev@lunn.ch, horms@kernel.org,
	donald.hunter@gmail.com, sdf@fomichev.me, almasrymina@google.com,
	dw@davidwei.uk, asml.silence@gmail.com, ap420073@gmail.com,
	jdamato@fastly.com, dtatulea@nvidia.com,
	michael.chan@broadcom.com
Subject: Re: [RFC net-next 17/22] netdev: add support for setting rx-buf-len per queue
Date: Tue, 22 Apr 2025 09:15:59 -0700	[thread overview]
Message-ID: <aAfAvxX707BGyLOZ@mini-arch> (raw)
In-Reply-To: <20250421222827.283737-18-kuba@kernel.org>

On 04/21, Jakub Kicinski wrote:
> Zero-copy APIs increase the cost of buffer management. They also extend
> this cost to user space applications which may be used to dealing with
> much larger buffers. Allow setting rx-buf-len per queue, devices with
> HW-GRO support can commonly fill buffers up to 32k (or rather 64k - 1
> but that's not a power of 2..)
> 
> The implementation adds a new option to the netdev netlink, rather
> than ethtool. The NIC-wide setting lives in ethtool ringparams so
> one could argue that we should be extending the ethtool API.
> OTOH netdev API is where we already have queue-get, and it's how
> zero-copy applications bind memory providers.
> 
> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
> ---
>  Documentation/netlink/specs/netdev.yaml | 15 ++++
>  include/net/netdev_queues.h             |  5 ++
>  include/net/netlink.h                   | 19 +++++
>  include/uapi/linux/netdev.h             |  2 +
>  net/core/netdev-genl-gen.h              |  1 +
>  tools/include/uapi/linux/netdev.h       |  2 +
>  net/core/netdev-genl-gen.c              | 15 ++++
>  net/core/netdev-genl.c                  | 92 +++++++++++++++++++++++++
>  net/core/netdev_config.c                | 16 +++++
>  9 files changed, 167 insertions(+)
> 
> diff --git a/Documentation/netlink/specs/netdev.yaml b/Documentation/netlink/specs/netdev.yaml
> index f5e0750ab71d..b0dfa970ee83 100644
> --- a/Documentation/netlink/specs/netdev.yaml
> +++ b/Documentation/netlink/specs/netdev.yaml
> @@ -324,6 +324,10 @@ name: netdev
>          doc: XSK information for this queue, if any.
>          type: nest
>          nested-attributes: xsk-info
> +      -
> +        name: rx-buf-len
> +        doc: Per-queue configuration of ETHTOOL_A_RINGS_RX_BUF_LEN.
> +        type: u32
>    -
>      name: qstats
>      doc: |
> @@ -743,6 +747,17 @@ name: netdev
>              - defer-hard-irqs
>              - gro-flush-timeout
>              - irq-suspend-timeout
> +    -
> +      name: queue-set
> +      doc: Set per-queue configurable options.
> +      attribute-set: queue
> +      do:
> +        request:
> +          attributes:
> +            - ifindex
> +            - type
> +            - id
> +            - rx-buf-len

Do we want some guidance going forward on what belongs to queue-set
vs napi-set? (mostly) HW settings for the queue-set and (mostly) SW
settings for (serving the queues) in napi-set?

  reply	other threads:[~2025-04-22 16:16 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-04-21 22:28 [RFC net-next 00/22] net: per-queue rx-buf-len configuration Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 01/22] docs: ethtool: document that rx_buf_len must control payload lengths Jakub Kicinski
2025-04-22 16:19   ` David Wei
2025-04-22 19:48   ` Joe Damato
2025-04-23 20:08   ` Mina Almasry
2025-04-25 22:50     ` Jakub Kicinski
2025-04-25 23:20       ` Joe Damato
2025-04-21 22:28 ` [RFC net-next 02/22] net: ethtool: report max value for rx-buf-len Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 03/22] net: use zero value to restore rx_buf_len to default Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 04/22] net: clarify the meaning of netdev_config members Jakub Kicinski
2025-04-22 19:57   ` Joe Damato
2025-04-21 22:28 ` [RFC net-next 05/22] net: add rx_buf_len to netdev config Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 06/22] eth: bnxt: read the page size from the adapter struct Jakub Kicinski
2025-04-23 20:35   ` Mina Almasry
2025-04-25 22:51     ` Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 07/22] eth: bnxt: set page pool page order based on rx_page_size Jakub Kicinski
2025-04-22 15:32   ` Stanislav Fomichev
2025-04-22 15:52     ` Jakub Kicinski
2025-04-22 17:27       ` Stanislav Fomichev
2025-04-21 22:28 ` [RFC net-next 08/22] eth: bnxt: support setting size of agg buffers via ethtool Jakub Kicinski
2025-04-23 21:00   ` Mina Almasry
2025-04-25 22:58     ` Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 09/22] net: move netdev_config manipulation to dedicated helpers Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 10/22] net: reduce indent of struct netdev_queue_mgmt_ops members Jakub Kicinski
2025-04-23 21:04   ` Mina Almasry
2025-04-21 22:28 ` [RFC net-next 11/22] net: allocate per-queue config structs and pass them thru the queue API Jakub Kicinski
2025-04-23 21:17   ` Mina Almasry
2025-04-25 23:24     ` Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 12/22] net: pass extack to netdev_rx_queue_restart() Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 13/22] net: add queue config validation callback Jakub Kicinski
2025-04-22 15:49   ` Stanislav Fomichev
2025-04-22 20:16   ` Joe Damato
2025-04-21 22:28 ` [RFC net-next 14/22] eth: bnxt: always set the queue mgmt ops Jakub Kicinski
2025-04-22 15:50   ` Stanislav Fomichev
2025-04-22 20:18   ` Joe Damato
2025-04-21 22:28 ` [RFC net-next 15/22] eth: bnxt: store the rx buf size per queue Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 16/22] eth: bnxt: adjust the fill level of agg queues with larger buffers Jakub Kicinski
2025-04-22 16:13   ` Stanislav Fomichev
2025-04-21 22:28 ` [RFC net-next 17/22] netdev: add support for setting rx-buf-len per queue Jakub Kicinski
2025-04-22 16:15   ` Stanislav Fomichev [this message]
2025-04-25 23:41     ` Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 18/22] net: wipe the setting of deactived queues Jakub Kicinski
2025-04-22 16:21   ` Stanislav Fomichev
2025-04-25 23:42     ` Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 19/22] eth: bnxt: use queue op config validate Jakub Kicinski
2025-04-23 10:00   ` Dragos Tatulea
2025-04-23 13:46     ` Jakub Kicinski
2025-04-23 14:24       ` Dragos Tatulea
2025-04-23 15:33         ` Jakub Kicinski
2025-06-12 11:56   ` Dragos Tatulea
2025-06-12 14:10     ` Jakub Kicinski
2025-06-12 15:52       ` Dragos Tatulea
2025-06-12 22:30         ` Jakub Kicinski
2025-06-13 19:02           ` Dragos Tatulea
2025-06-13 23:16             ` Jakub Kicinski
2025-06-17 12:36               ` Dragos Tatulea
2025-04-21 22:28 ` [RFC net-next 20/22] eth: bnxt: support per queue configuration of rx-buf-len Jakub Kicinski
2025-04-21 22:28 ` [RFC net-next 21/22] selftests: drv-net: add helper/wrapper for bpftrace Jakub Kicinski
2025-04-22 16:36   ` Stanislav Fomichev
2025-04-22 16:39     ` Stanislav Fomichev
2025-06-25 12:23   ` Breno Leitao
2025-04-21 22:28 ` [RFC net-next 22/22] selftests: drv-net: add test for rx-buf-len Jakub Kicinski
2025-04-22 17:06   ` Stanislav Fomichev
2025-04-25 23:52     ` Jakub Kicinski
2025-04-23 20:02 ` [RFC net-next 00/22] net: per-queue rx-buf-len configuration Mina Almasry
2025-04-25 23:55   ` Jakub Kicinski

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=aAfAvxX707BGyLOZ@mini-arch \
    --to=stfomichev@gmail.com \
    --cc=almasrymina@google.com \
    --cc=andrew+netdev@lunn.ch \
    --cc=ap420073@gmail.com \
    --cc=asml.silence@gmail.com \
    --cc=davem@davemloft.net \
    --cc=donald.hunter@gmail.com \
    --cc=dtatulea@nvidia.com \
    --cc=dw@davidwei.uk \
    --cc=edumazet@google.com \
    --cc=horms@kernel.org \
    --cc=jdamato@fastly.com \
    --cc=kuba@kernel.org \
    --cc=michael.chan@broadcom.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=sdf@fomichev.me \
    /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;
as well as URLs for NNTP newsgroup(s).