DPDK patches and discussions
 help / color / mirror / Atom feed
From: Thomas Monjalon <thomas@monjalon.net>
To: Eli Britstein <elibr@nvidia.com>
Cc: dev@dpdk.org, Gaetan Rivet <gaetanr@nvidia.com>,
	Asaf Penso <asafp@nvidia.com>, Ori Kam <orika@nvidia.com>,
	stable@dpdk.org, matan@nvidia.com, viacheslavo@nvidia.com
Subject: Re: [dpdk-dev] [PATCH] doc: add Linux rawio capability requirement in mlx5 guide
Date: Thu, 01 Apr 2021 13:39:21 +0200	[thread overview]
Message-ID: <7504044.RxHU5scrEf@thomas> (raw)
In-Reply-To: <20210401074913.16563-1-elibr@nvidia.com>

+Cc mlx5 maintainers

About the title: "rawio" is not a word.
"raw IO" would be better.

01/04/2021 09:49, Eli Britstein:
> For an application to be able to create "transfer" rte_flows for mlx5

What is "tranfer" rule?

rte_flows does not exist. You probably mean rte_flow rule.

> devices, it should have cap_sys_rawio capability.
> Document it.
> Fixes: f772cc424c45 ("doc: add required Linux capabilities in mlx5 guide")
> Cc: stable@dpdk.org
> Signed-off-by: Eli Britstein <elibr@nvidia.com>
> Reviewed-by: Gaetan Rivet <gaetanr@nvidia.com>
>  In order to run as a non-root user,
> -some capabilities must be granted to the application::
> +some capabilities must be granted to the application, while others are

If you wish to show that all are not a must,
better to replace "must be granted" by "may be required".

> +optional for some use cases. Capabilities can be set for example using

Please start new sentence on next line.

> +the setcap utilitiy::

typo: utility

> -   setcap cap_sys_admin,cap_net_admin,cap_net_raw,cap_ipc_lock+ep <dpdk-app>
> +   setcap cap_net_admin,cap_net_raw,cap_ipc_lock,cap_sys_admin,cap_sys_rawio+ep <dpdk-app>
>  Below are the reasons of the need for each capability:
> -``cap_sys_admin``
> -   When using physical addresses (PA mode), with Linux >= 4.0,
> -   for access to ``/proc/self/pagemap``.
> -
>  ``cap_net_admin``
>     For device configuration.
> @@ -477,6 +475,14 @@ Below are the reasons of the need for each capability:
>  ``cap_ipc_lock``
>     For DMA memory pinning.
> +``cap_sys_admin`` optional
> +   When using physical addresses (PA mode), with Linux >= 4.0,
> +   for access to ``/proc/self/pagemap``.
> +
> +``cap_sys_rawio`` optional
> +   For the application to be able to apply rte_flow flows in transfer
> +   mode.

"For the application" looks useless.
We don't apply "rte_flow flows" but "flow rules".
What is "transfer mode"? It is not mentioned elsewhere in this doc.
How this capability is related to dv_esw_en?
Do we need it in case of trusted VF?

  reply	other threads:[~2021-04-01 11:39 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01  7:49 Eli Britstein
2021-04-01 11:39 ` Thomas Monjalon [this message]
2021-04-01 12:06   ` Eli Britstein
2021-04-01 13:15     ` Thomas Monjalon

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:

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

  git send-email \
    --in-reply-to=7504044.RxHU5scrEf@thomas \
    --to=thomas@monjalon.net \
    --cc=asafp@nvidia.com \
    --cc=dev@dpdk.org \
    --cc=elibr@nvidia.com \
    --cc=gaetanr@nvidia.com \
    --cc=matan@nvidia.com \
    --cc=orika@nvidia.com \
    --cc=stable@dpdk.org \
    --cc=viacheslavo@nvidia.com \


* 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).