From: Simon Horman <simon.horman@corigine.com>
To: Vladimir Oltean <vladimir.oltean@nxp.com>
Cc: netdev@vger.kernel.org, "David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
Claudiu Manoil <claudiu.manoil@nxp.com>,
Vinicius Costa Gomes <vinicius.gomes@intel.com>,
Kurt Kanzenbach <kurt@linutronix.de>,
Jacob Keller <jacob.e.keller@intel.com>,
Jamal Hadi Salim <jhs@mojatatu.com>,
Cong Wang <xiyou.wangcong@gmail.com>,
Jiri Pirko <jiri@resnulli.us>
Subject: Re: [PATCH v5 net-next 08/17] net/sched: mqprio: allow reverse TC:TXQ mappings
Date: Fri, 3 Feb 2023 17:18:08 +0100 [thread overview]
Message-ID: <Y90zwMijGCw79Ulw@corigine.com> (raw)
In-Reply-To: <20230202003621.2679603-9-vladimir.oltean@nxp.com>
On Thu, Feb 02, 2023 at 02:36:12AM +0200, Vladimir Oltean wrote:
> By imposing that the last TXQ of TC i is smaller than the first TXQ of
> any TC j (j := i+1 .. n), mqprio imposes a strict ordering condition for
> the TXQ indices (they must increase as TCs increase).
>
> Claudiu points out that the complexity of the TXQ count validation is
> too high for this logic, i.e. instead of iterating over j, it is
> sufficient that the TXQ indices of TC i and i + 1 are ordered, and that
> will eventually ensure global ordering.
>
> This is true, however it doesn't appear to me that is what the code
> really intended to do. Instead, based on the comments, it just wanted to
> check for overlaps (and this isn't how one does that).
>
> So the following mqprio configuration, which I had recommended to
> Vinicius more than once for igb/igc (to account for the fact that on
> this hardware, lower numbered TXQs have higher dequeue priority than
> higher ones):
>
> num_tc 4 map 0 1 2 3 queues 1@3 1@2 1@1 1@0
>
> is in fact denied today by mqprio.
>
> The full story is that in fact, it's only denied with "hw 0"; if
> hardware offloading is requested, mqprio defers TXQ range overlap
> validation to the device driver (a strange decision in itself).
>
> This is most certainly a bug, but it's not one that has any merit for
> being fixed on "stable" as far as I can tell. This is because mqprio
> always rejected a configuration which was in fact valid, and this has
> shaped the way in which mqprio configuration scripts got built for
> various hardware (see igb/igc in the link below). Therefore, one could
> consider it to be merely an improvement for mqprio to allow reverse
> TC:TXQ mappings.
>
> Link: https://patchwork.kernel.org/project/netdevbpf/patch/20230130173145.475943-9-vladimir.oltean@nxp.com/#25188310
> Link: https://patchwork.kernel.org/project/netdevbpf/patch/20230128010719.2182346-6-vladimir.oltean@nxp.com/#25186442
> Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
next prev parent reply other threads:[~2023-02-03 16:18 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-02 0:36 [PATCH v5 net-next 00/17] ENETC mqprio/taprio cleanup Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 01/17] net: enetc: simplify enetc_num_stack_tx_queues() Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 02/17] net: enetc: allow the enetc_reconfigure() callback to fail Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 03/17] net: enetc: recalculate num_real_tx_queues when XDP program attaches Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 04/17] net: enetc: ensure we always have a minimum number of TXQs for stack Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 05/17] net/sched: mqprio: refactor nlattr parsing to a separate function Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 06/17] net/sched: mqprio: refactor offloading and unoffloading to dedicated functions Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 07/17] net/sched: move struct tc_mqprio_qopt_offload from pkt_cls.h to pkt_sched.h Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 08/17] net/sched: mqprio: allow reverse TC:TXQ mappings Vladimir Oltean
2023-02-03 16:18 ` Simon Horman [this message]
2023-02-05 11:55 ` Gerhard Engleder
2023-02-05 12:22 ` Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 09/17] net/sched: mqprio: allow offloading drivers to request queue count validation Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 10/17] net/sched: mqprio: add extack messages for " Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 11/17] net/sched: taprio: centralize mqprio qopt validation Vladimir Oltean
2023-02-03 16:32 ` Simon Horman
2023-02-05 12:10 ` Gerhard Engleder
2023-02-05 12:23 ` Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 12/17] net/sched: refactor mqprio qopt reconstruction to a library function Vladimir Oltean
2023-02-03 16:33 ` Simon Horman
2023-02-02 0:36 ` [PATCH v5 net-next 13/17] net/sched: taprio: pass mqprio queue configuration to ndo_setup_tc() Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 14/17] net/sched: taprio: only calculate gate mask per TXQ for igc, stmmac and tsnep Vladimir Oltean
2023-02-02 8:04 ` Roger Quadros
2023-02-02 12:31 ` Vladimir Oltean
2023-02-02 20:05 ` Roger Quadros
2023-02-02 0:36 ` [PATCH v5 net-next 15/17] net: enetc: request mqprio to validate the queue counts Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 16/17] net: enetc: act upon the requested mqprio queue configuration Vladimir Oltean
2023-02-02 0:36 ` [PATCH v5 net-next 17/17] net: enetc: act upon mqprio queue config in taprio offload Vladimir Oltean
2023-02-02 20:54 ` [PATCH v5 net-next 00/17] ENETC mqprio/taprio cleanup Vladimir Oltean
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=Y90zwMijGCw79Ulw@corigine.com \
--to=simon.horman@corigine.com \
--cc=claudiu.manoil@nxp.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=jacob.e.keller@intel.com \
--cc=jhs@mojatatu.com \
--cc=jiri@resnulli.us \
--cc=kuba@kernel.org \
--cc=kurt@linutronix.de \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=vinicius.gomes@intel.com \
--cc=vladimir.oltean@nxp.com \
--cc=xiyou.wangcong@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).