From: "Toke Høiland-Jørgensen" <toke@toke.dk> To: Rajkumar Manoharan <rmanohar@codeaurora.org> Cc: linux-wireless@vger.kernel.org, ath10k@lists.infradead.org Subject: Re: [PATCH 3/6] mac80211: Add airtime accounting and scheduling to TXQs Date: Sun, 28 Oct 2018 16:48:09 +0100 [thread overview] Message-ID: <87woq2843q.fsf@toke.dk> (raw) In-Reply-To: <9c2b790132a9a89fecd7dd79dc67d891@codeaurora.org> Rajkumar Manoharan <rmanohar@codeaurora.org> writes: > On 2018-10-26 07:16, Toke Høiland-Jørgensen wrote: >> Rajkumar Manoharan <rmanohar@codeaurora.org> writes: >> >>> From: Toke Høiland-Jørgensen <toke@toke.dk> > [...] >>> u8 max_nan_de_entries; >>> u8 tx_sk_pacing_shift; >>> + u32 airtime_weight; >>> }; >> >> This doesn't make sense. Airtime weights can be set by userspace, so >> even if a driver sets another default it is not guaranteed to be >> honoured. So what's the point? >> > The reason for driver specific default is to avoid performance impact > in ath10k when the user is using vanilla ath10k with default airtime. > As I mentioned earlier, mac80211 default (256us) is too low for 11ac > devices especially with driver is bursting aggregation. > > Yes. I do understand the user can change airtime at anytime but It > must be noted that different airtime weight will result in different > throughput. IMHO the defaults should not impact current benchmark. > Otherwise it will be alarmed as regression later. isn't it? My point is that if the user has to know the implementation-specific limitations of each driver before setting a weight, then it's not a particularly friendly API. I think we should be able to do better than that... >> So since we're rotating the queue on every call to the function, I'm >> wondering if this actually succeeds in throttling the slow station's >> airtime usage enough to achieve fairness? So I'll ask again: Did you >> test the fairness performance, and how? >> > We are collecting data of mixed clients (11ac, 11n and legacy) keeping > them at same distance and different distance. So that lower rate > clients will interfere higher MCS rate station. Also configuring > different airtime weight for each stations so that throttling low rate > clients more should help higher performing(better MCS) clients. > > By allowing different airtime for each stations, the user can control > guest network over primary network. Also It helped to improve > performance of preferred station and algo. to control station is given > to cloud or user application. > > As of now, We are testing with 4 11ac clients of same distance. And > collecting the performance data in multiple iteration. Below are > current data of station's performance (Mbps) against airtime weight. > > airtime station1(%airtime) station2 station3 station4 > (Mbps) > > No ATF 182 168 166 169 > > 4ms 170 (100%) 164 (100%) 185 (100%) 175 (100%) > > 4ms 277 (70%) 115 (10%) 103 (10%) 105 (10%) > > 4ms 223 (40%) 214 (40%) 109 (10%) 94 (10%) > > 4ms 337 (90%) 182 (8%) 23 (1%) 30 (1%) So this looks like it's doing *something*, but not like it's succeeding in achieving the set percentages. Did you check if the actual airtime values (in debugfs) corresponds to the configured weights? > > STA1(11ac) STA2 (11n) STA3(11a) > ========== ========== ========= > > No ATF 225 166 3.5 > > ATF (4ms) 234 151 3.5 This also shows like ATF has no effect? >> Also, taking a step back: >> >> With this, we're doing lots of work to make sure that the hardware's >> round-robin scheduling queue lines up with mac80211; so if we do this >> anyway, why can't we just control the order directly from mac80211 >> (which is what we do with the other next_txq() API)? >> > The only way to enforce mac80211 ordering in ath10k is to disable pull > mode in firmware and always operates in push mode similar to ath9k. And I assume that is not too likely to happen, or? What is the benefit of pull mode at the firmware level? -Toke
WARNING: multiple messages have this Message-ID (diff)
From: "Toke Høiland-Jørgensen" <toke@toke.dk> To: Rajkumar Manoharan <rmanohar@codeaurora.org> Cc: linux-wireless@vger.kernel.org, ath10k@lists.infradead.org Subject: Re: [PATCH 3/6] mac80211: Add airtime accounting and scheduling to TXQs Date: Sun, 28 Oct 2018 16:48:09 +0100 [thread overview] Message-ID: <87woq2843q.fsf@toke.dk> (raw) In-Reply-To: <9c2b790132a9a89fecd7dd79dc67d891@codeaurora.org> Rajkumar Manoharan <rmanohar@codeaurora.org> writes: > On 2018-10-26 07:16, Toke Høiland-Jørgensen wrote: >> Rajkumar Manoharan <rmanohar@codeaurora.org> writes: >> >>> From: Toke Høiland-Jørgensen <toke@toke.dk> > [...] >>> u8 max_nan_de_entries; >>> u8 tx_sk_pacing_shift; >>> + u32 airtime_weight; >>> }; >> >> This doesn't make sense. Airtime weights can be set by userspace, so >> even if a driver sets another default it is not guaranteed to be >> honoured. So what's the point? >> > The reason for driver specific default is to avoid performance impact > in ath10k when the user is using vanilla ath10k with default airtime. > As I mentioned earlier, mac80211 default (256us) is too low for 11ac > devices especially with driver is bursting aggregation. > > Yes. I do understand the user can change airtime at anytime but It > must be noted that different airtime weight will result in different > throughput. IMHO the defaults should not impact current benchmark. > Otherwise it will be alarmed as regression later. isn't it? My point is that if the user has to know the implementation-specific limitations of each driver before setting a weight, then it's not a particularly friendly API. I think we should be able to do better than that... >> So since we're rotating the queue on every call to the function, I'm >> wondering if this actually succeeds in throttling the slow station's >> airtime usage enough to achieve fairness? So I'll ask again: Did you >> test the fairness performance, and how? >> > We are collecting data of mixed clients (11ac, 11n and legacy) keeping > them at same distance and different distance. So that lower rate > clients will interfere higher MCS rate station. Also configuring > different airtime weight for each stations so that throttling low rate > clients more should help higher performing(better MCS) clients. > > By allowing different airtime for each stations, the user can control > guest network over primary network. Also It helped to improve > performance of preferred station and algo. to control station is given > to cloud or user application. > > As of now, We are testing with 4 11ac clients of same distance. And > collecting the performance data in multiple iteration. Below are > current data of station's performance (Mbps) against airtime weight. > > airtime station1(%airtime) station2 station3 station4 > (Mbps) > > No ATF 182 168 166 169 > > 4ms 170 (100%) 164 (100%) 185 (100%) 175 (100%) > > 4ms 277 (70%) 115 (10%) 103 (10%) 105 (10%) > > 4ms 223 (40%) 214 (40%) 109 (10%) 94 (10%) > > 4ms 337 (90%) 182 (8%) 23 (1%) 30 (1%) So this looks like it's doing *something*, but not like it's succeeding in achieving the set percentages. Did you check if the actual airtime values (in debugfs) corresponds to the configured weights? > > STA1(11ac) STA2 (11n) STA3(11a) > ========== ========== ========= > > No ATF 225 166 3.5 > > ATF (4ms) 234 151 3.5 This also shows like ATF has no effect? >> Also, taking a step back: >> >> With this, we're doing lots of work to make sure that the hardware's >> round-robin scheduling queue lines up with mac80211; so if we do this >> anyway, why can't we just control the order directly from mac80211 >> (which is what we do with the other next_txq() API)? >> > The only way to enforce mac80211 ordering in ath10k is to disable pull > mode in firmware and always operates in push mode similar to ath9k. And I assume that is not too likely to happen, or? What is the benefit of pull mode at the firmware level? -Toke _______________________________________________ ath10k mailing list ath10k@lists.infradead.org http://lists.infradead.org/mailman/listinfo/ath10k
next prev parent reply other threads:[~2018-10-28 15:48 UTC|newest] Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top 2018-10-20 11:05 [PATCH 0/6] Move TXQ scheduling and airtime fairness into mac80211 Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-10-20 11:05 ` [PATCH 1/6] mac80211: Add TXQ scheduling API Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-11-09 12:00 ` Johannes Berg 2018-11-09 12:00 ` Johannes Berg 2018-11-09 12:39 ` Toke Høiland-Jørgensen 2018-11-09 12:39 ` Toke Høiland-Jørgensen 2018-10-20 11:05 ` [PATCH 2/6] cfg80211: Add airtime statistics and settings Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-10-20 11:05 ` [PATCH 3/6] mac80211: Add airtime accounting and scheduling to TXQs Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-10-26 14:16 ` Toke Høiland-Jørgensen 2018-10-26 14:16 ` Toke Høiland-Jørgensen 2018-10-26 23:04 ` Rajkumar Manoharan 2018-10-26 23:04 ` Rajkumar Manoharan 2018-10-28 15:48 ` Toke Høiland-Jørgensen [this message] 2018-10-28 15:48 ` Toke Høiland-Jørgensen 2018-10-28 22:01 ` Rajkumar Manoharan 2018-10-28 22:01 ` Rajkumar Manoharan 2018-10-29 23:50 ` Rajkumar Manoharan 2018-10-29 23:50 ` Rajkumar Manoharan 2018-11-02 10:30 ` Toke Høiland-Jørgensen 2018-11-02 10:30 ` Toke Høiland-Jørgensen 2018-11-05 8:39 ` Rajkumar Manoharan 2018-11-05 8:39 ` Rajkumar Manoharan 2018-11-07 14:53 ` Toke Høiland-Jørgensen 2018-11-07 14:53 ` Toke Høiland-Jørgensen 2018-11-07 22:35 ` Rajkumar Manoharan 2018-11-07 22:35 ` Rajkumar Manoharan 2018-11-08 13:46 ` Toke Høiland-Jørgensen 2018-11-08 13:46 ` Toke Høiland-Jørgensen 2018-10-31 6:17 ` yiboz 2018-10-31 6:17 ` yiboz 2018-10-20 11:05 ` [PATCH 4/6] ath9k: Switch to mac80211 TXQ scheduling and airtime APIs Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-10-20 11:05 ` [PATCH 5/6] ath10k: migrate to mac80211 txq scheduling Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-10-24 8:33 ` Kalle Valo 2018-10-24 8:33 ` Kalle Valo 2018-10-24 18:55 ` Rajkumar Manoharan 2018-10-24 18:55 ` Rajkumar Manoharan 2018-10-20 11:05 ` [PATCH 6/6] ath10k: reporting estimated tx airtime for fairness Rajkumar Manoharan 2018-10-20 11:05 ` Rajkumar Manoharan 2018-10-24 8:35 ` Kalle Valo 2018-10-24 8:35 ` Kalle Valo 2018-10-21 11:27 ` [PATCH 0/6] Move TXQ scheduling and airtime fairness into mac80211 Toke Høiland-Jørgensen 2018-10-21 11:27 ` Toke Høiland-Jørgensen
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=87woq2843q.fsf@toke.dk \ --to=toke@toke.dk \ --cc=ath10k@lists.infradead.org \ --cc=linux-wireless@vger.kernel.org \ --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: linkBe 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.