netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jakub Kicinski <kuba@kernel.org>
To: Tariq Toukan <ttoukan.linux@gmail.com>
Cc: Saeed Mahameed <saeed@kernel.org>,
	"David S. Miller" <davem@davemloft.net>,
	Paolo Abeni <pabeni@redhat.com>,
	Eric Dumazet <edumazet@google.com>,
	Saeed Mahameed <saeedm@nvidia.com>,
	netdev@vger.kernel.org, Tariq Toukan <tariqt@nvidia.com>,
	Gal Pressman <gal@nvidia.com>,
	Leon Romanovsky <leonro@nvidia.com>,
	Carolina Jubran <cjubran@nvidia.com>
Subject: Re: [net 06/10] net/mlx5: RSS, Block changing channels number when RXFH is configured
Date: Mon, 1 Apr 2024 07:34:16 -0700	[thread overview]
Message-ID: <20240401073416.39fb2234@kernel.org> (raw)
In-Reply-To: <87ca050f-5643-4b90-8768-1d624e367cac@gmail.com>

On Mon, 1 Apr 2024 09:54:26 +0300 Tariq Toukan wrote:
> The rationale of having a "single flow" or "single "logic" is to make it 
> simple, and achieve a fine user experience.
> 
> Otherwise, users would, for example, question why increasing the number 
> of channels (after setting the indir table) from 24 channels to 120 
> works, but doesn't work when trying with 130 channels, although max num 
> channels is much higher.

Any way to preserve the indir table in case it has to grow?
If it increases by pow2 maybe we can "duplicate" the old table?
90% of the time when user changes the settings it's to exclude
a queue from RSS, the remaining 10% to change the balance. 
In both cases "mirroring" the settings would be fine.

> The required order looks pretty natural: first set the desired num of 
> channels, and only then set your indirection table.

Say user allocated 16 queues for RSS and 4 for flow rules and/or other
RSS context. Now they want to bump the 4 to 8. Resetting RSS and to be
able to allocate new queues may not be an option, as traffic from the
two "domains" would start mixing. Admittedly a bit contrived but not
impossible, so my vote would be to only nak the cases we really can't
reliably support :(

> At the end, there are pros and cons for each solution.
> If you still strongly prefer narrowing it down only to the truly 
> problematic transitions, then we'll have no big issue in changing this.

  reply	other threads:[~2024-04-01 14:34 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-26 14:46 [pull request][net 00/10] mlx5 fixes 2024-03-26 Saeed Mahameed
2024-03-26 14:46 ` [net 01/10] net/mlx5: E-switch, store eswitch pointer before registering devlink_param Saeed Mahameed
2024-03-26 14:46 ` [net 02/10] net/mlx5: Register devlink first under devlink lock Saeed Mahameed
2024-03-26 14:46 ` [net 03/10] net/mlx5: offset comp irq index in name by one Saeed Mahameed
2024-03-26 14:46 ` [net 04/10] net/mlx5: Properly link new fs rules into the tree Saeed Mahameed
2024-03-26 14:46 ` [net 05/10] net/mlx5: Correctly compare pkt reformat ids Saeed Mahameed
2024-03-26 14:46 ` [net 06/10] net/mlx5: RSS, Block changing channels number when RXFH is configured Saeed Mahameed
2024-03-29  5:31   ` Jakub Kicinski
2024-04-01  6:54     ` Tariq Toukan
2024-04-01 14:34       ` Jakub Kicinski [this message]
2024-03-26 14:46 ` [net 07/10] net/mlx5e: Fix mlx5e_priv_init() cleanup flow Saeed Mahameed
2024-03-26 14:46 ` [net 08/10] net/mlx5e: HTB, Fix inconsistencies with QoS SQs number Saeed Mahameed
2024-03-26 14:46 ` [net 09/10] net/mlx5e: Do not produce metadata freelist entries in Tx port ts WQE xmit Saeed Mahameed
2024-03-26 14:46 ` [net 10/10] net/mlx5e: RSS, Block XOR hash with over 128 channels Saeed Mahameed

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=20240401073416.39fb2234@kernel.org \
    --to=kuba@kernel.org \
    --cc=cjubran@nvidia.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=gal@nvidia.com \
    --cc=leonro@nvidia.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=saeed@kernel.org \
    --cc=saeedm@nvidia.com \
    --cc=tariqt@nvidia.com \
    --cc=ttoukan.linux@gmail.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;
as well as URLs for NNTP newsgroup(s).