All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Toke Høiland-Jørgensen" <toke@redhat.com>
To: Kan Yan <kyan@google.com>, Johannes Berg <johannes@sipsolutions.net>
Cc: linux-wireless@vger.kernel.org,
	Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>,
	ath10k@lists.infradead.org, John Crispin <john@phrozen.org>,
	Lorenzo Bianconi <lorenzo@kernel.org>,
	Felix Fietkau <nbd@nbd.name>,
	Rajkumar Manoharan <rmanohar@codeaurora.org>,
	Kevin Hayes <kevinhayes@google.com>
Subject: Re: [PATCH v6 3/4] mac80211: Implement Airtime-based Queue Limit (AQL)
Date: Sat, 09 Nov 2019 12:22:49 +0100	[thread overview]
Message-ID: <87k189qoom.fsf@toke.dk> (raw)
In-Reply-To: <CA+iem5saT0ntxnSV1ZatFuRRcaBzivkQ_MU66ZdV1kfb1dC6Qg@mail.gmail.com>

Kan Yan <kyan@google.com> writes:

> It is most likely just insufficient locking. active_txq_lock is per
> AC, can't protect local->aql_total_pending_airtime against racing
> conditions:
> void ieee80211_sta_update_pending_airtime(...)
> {
>         spin_lock_bh(&local->active_txq_lock[ac]);
>         ...
>         local->aql_total_pending_airtime -= tx_airtime;
>         ...
>         spin_unlock_bh(&local->active_txq_lock[ac]);
> }

Ohh, right; didn't even realise those were not per-AC as well...

> After changing it to atomic_t, no more aql_total_pending_airtime
> underflow so far :). Using atomic operation should also help reduce
> CPU overhead due to lock contention, as
> ieee80211_sta_update_pending_airtime() is often called from the tx
> completion routine triggered by interrupts, often in a different core
> than where __ieee80211_schedule_txq() is running.
>
> I will post a new version a bit later if the test goes well.

Awesome! :)

-Toke


WARNING: multiple messages have this Message-ID (diff)
From: "Toke Høiland-Jørgensen" <toke@redhat.com>
To: Kan Yan <kyan@google.com>, Johannes Berg <johannes@sipsolutions.net>
Cc: Rajkumar Manoharan <rmanohar@codeaurora.org>,
	Kevin Hayes <kevinhayes@google.com>,
	Make-Wifi-fast <make-wifi-fast@lists.bufferbloat.net>,
	linux-wireless@vger.kernel.org, ath10k@lists.infradead.org,
	John Crispin <john@phrozen.org>,
	Lorenzo Bianconi <lorenzo@kernel.org>,
	Felix Fietkau <nbd@nbd.name>
Subject: Re: [PATCH v6 3/4] mac80211: Implement Airtime-based Queue Limit (AQL)
Date: Sat, 09 Nov 2019 12:22:49 +0100	[thread overview]
Message-ID: <87k189qoom.fsf@toke.dk> (raw)
In-Reply-To: <CA+iem5saT0ntxnSV1ZatFuRRcaBzivkQ_MU66ZdV1kfb1dC6Qg@mail.gmail.com>

Kan Yan <kyan@google.com> writes:

> It is most likely just insufficient locking. active_txq_lock is per
> AC, can't protect local->aql_total_pending_airtime against racing
> conditions:
> void ieee80211_sta_update_pending_airtime(...)
> {
>         spin_lock_bh(&local->active_txq_lock[ac]);
>         ...
>         local->aql_total_pending_airtime -= tx_airtime;
>         ...
>         spin_unlock_bh(&local->active_txq_lock[ac]);
> }

Ohh, right; didn't even realise those were not per-AC as well...

> After changing it to atomic_t, no more aql_total_pending_airtime
> underflow so far :). Using atomic operation should also help reduce
> CPU overhead due to lock contention, as
> ieee80211_sta_update_pending_airtime() is often called from the tx
> completion routine triggered by interrupts, often in a different core
> than where __ieee80211_schedule_txq() is running.
>
> I will post a new version a bit later if the test goes well.

Awesome! :)

-Toke


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k

  reply	other threads:[~2019-11-09 11:22 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-23  9:58 [PATCH v6 0/4] Add Airtime Queue Limits (AQL) to mac80211 Toke Høiland-Jørgensen
2019-10-23  9:58 ` Toke Høiland-Jørgensen
2019-10-23  9:59 ` [PATCH v6 1/4] mac80211: Shrink the size of ack_frame_id to make room for tx_time_est Toke Høiland-Jørgensen
2019-10-23  9:59   ` Toke Høiland-Jørgensen
2019-11-08 10:03   ` Johannes Berg
2019-11-08 10:03     ` Johannes Berg
2019-10-23  9:59 ` [PATCH v6 2/4] mac80211: Import airtime calculation code from mt76 Toke Høiland-Jørgensen
2019-10-23  9:59   ` Toke Høiland-Jørgensen
2019-11-08 10:07   ` Johannes Berg
2019-11-08 10:07     ` Johannes Berg
2019-11-08 10:55     ` Toke Høiland-Jørgensen
2019-11-08 10:55       ` Toke Høiland-Jørgensen
2019-11-08 10:57       ` Johannes Berg
2019-11-08 10:57         ` Johannes Berg
2019-10-23  9:59 ` [PATCH v6 3/4] mac80211: Implement Airtime-based Queue Limit (AQL) Toke Høiland-Jørgensen
2019-10-23  9:59   ` Toke Høiland-Jørgensen
2019-11-08 10:09   ` Johannes Berg
2019-11-08 10:09     ` Johannes Berg
2019-11-08 10:56     ` Toke Høiland-Jørgensen
2019-11-08 10:56       ` Toke Høiland-Jørgensen
2019-11-08 10:59       ` Johannes Berg
2019-11-08 10:59         ` Johannes Berg
2019-11-08 11:10         ` Toke Høiland-Jørgensen
2019-11-08 11:10           ` Toke Høiland-Jørgensen
2019-11-08 11:17           ` Johannes Berg
2019-11-08 11:17             ` Johannes Berg
2019-11-09  1:22             ` Kan Yan
2019-11-09  1:22               ` Kan Yan
2019-11-09 11:22               ` Toke Høiland-Jørgensen [this message]
2019-11-09 11:22                 ` Toke Høiland-Jørgensen
2019-10-23  9:59 ` [PATCH v6 4/4] mac80211: Use Airtime-based Queue Limits (AQL) on packet dequeue Toke Høiland-Jørgensen
2019-10-23  9:59   ` Toke Høiland-Jørgensen
2019-11-08 10:17   ` Johannes Berg
2019-11-08 10:17     ` Johannes Berg
2019-11-08 11:01     ` Toke Høiland-Jørgensen
2019-11-08 11:01       ` Toke Høiland-Jørgensen
2019-11-08 11:05       ` Johannes Berg
2019-11-08 11:05         ` Johannes Berg
2019-11-07  6:14 ` [PATCH v6 0/4] Add Airtime Queue Limits (AQL) to mac80211 Kan Yan
2019-11-07  6:14   ` Kan Yan
2019-11-07 10:55   ` Toke Høiland-Jørgensen
2019-11-07 10:55     ` Toke Høiland-Jørgensen
2019-11-07 21:24   ` [Make-wifi-fast] " Dave Taht

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=87k189qoom.fsf@toke.dk \
    --to=toke@redhat.com \
    --cc=ath10k@lists.infradead.org \
    --cc=johannes@sipsolutions.net \
    --cc=john@phrozen.org \
    --cc=kevinhayes@google.com \
    --cc=kyan@google.com \
    --cc=linux-wireless@vger.kernel.org \
    --cc=lorenzo@kernel.org \
    --cc=make-wifi-fast@lists.bufferbloat.net \
    --cc=nbd@nbd.name \
    --cc=rmanohar@codeaurora.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.