Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Jon Pike]

[-- Attachment #1: Type: text/plain, Size: 2559 bytes --]

HNY to all...

Did a Sysupgrade of the 12-27-16 LEDE, from Toke's 11-24 build,  running
nicely.

Not sure if I should have done the full FW update,  or not tried saving
settings, but after luci install everything seems to be working fine.
Better than before, actually.

I did see the same errors in the syslog that Dave reports on the ath10k.
Wondering if I should, on a TP-Link C7?

Tnx...

Jon

---------- Forwarded message ----------
From: "Dave Täht" <dave@taht.net>
To: make-wifi-fast@lists.bufferbloat.net
Cc:
Date: Sun, 25 Dec 2016 12:16:59 -0500
Subject: Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged
upstream, and the Turris Omnia debloated


I just installed the dec 25th build for the ubnt uaplite.

airtime fairness/fq_codel appears in there for the ath9k.

regrettably the ath10k in that box is not picking up the fq_codel code.
Too old firmware? Too old chipset?

root@fqcodel:/etc/config# dmesg | grep ath10k
[   27.720253] ath10k_pci 0000:00:00.0: pci irq legacy oper_irq_mode 1
irq_mode 0 reset_mode 0
[   27.938406] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/pre-cal-pci-0000:00:00.0.bin failed with error -2
[   27.949483] ath10k_pci 0000:00:00.0: Falling back to user helper
[   28.140816] firmware ath10k!pre-cal-pci-0000:00:00.0.bin:
firmware_loading_store: map pages failed
[   28.150362] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/cal-pci-0000:00:00.0.bin failed with error -2
[   28.161085] ath10k_pci 0000:00:00.0: Falling back to user helper
[   35.436780] ath10k_pci 0000:00:00.0: qca988x hw2.0 target 0x4100016c
chip_id 0x043222ff sub 0000:0000
[   35.446345] ath10k_pci 0000:00:00.0: kconfig debug 0 debugfs 1
tracing 0 dfs 1 testmode 1
[   35.459369] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.70.59-2 api
5 features no-p2p,raw-mode,mfp crc32 4159f498
[   35.470177] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
[   35.480968] ath10k_pci 0000:00:00.0: Falling back to user helper
[   35.558782] firmware ath10k!QCA988X!hw2.0!board-2.bin:
firmware_loading_store: map pages failed
[   35.575232] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
crc32 bebc7c08
[   36.684593] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
cal file max-sta 128 raw 0 hwcrypto 1


>
> -Toke
> _______________________________________________
> Make-wifi-fast mailing list
> Make-wifi-fast@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/make-wifi-fast
>

[-- Attachment #2: Type: text/html, Size: 3581 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Jaap Buurman]

[-- Attachment #1: Type: text/plain, Size: 3148 bytes --]

Is this planned to be added in the current LEDE trunk? And if so, is there
an ETA? Sounds like an amazing feature to have by default in all LEDE
builds, similar to the current Ath9k airtime fairness patch :)

On Dec 25, 2016 19:41, "Noah Causin" <n0manletter@gmail.com> wrote:

> It probably does not have the ath10k softqueues patch from Toke's
> repository.
>
>
> On 12/25/2016 12:16 PM, Dave Täht wrote:
>
>>
>> On 12/25/16 12:06 PM, Toke Høiland-Jørgensen wrote:
>>
>>> Valent Turkovic <valent@otvorenamreza.org> writes:
>>>
>>> Please send just a short notification as soon as Lede with airfairness
>>>> has been released, I have really good real world location that can
>>>> test how effective it works...
>>>>
>>> It should be in the nightly LEDE builds by now :)
>>>
>> I just installed the dec 25th build for the ubnt uaplite.
>>
>> airtime fairness/fq_codel appears in there for the ath9k.
>>
>> regrettably the ath10k in that box is not picking up the fq_codel code.
>> Too old firmware? Too old chipset?
>>
>> root@fqcodel:/etc/config# dmesg | grep ath10k
>> [   27.720253] ath10k_pci 0000:00:00.0: pci irq legacy oper_irq_mode 1
>> irq_mode 0 reset_mode 0
>> [   27.938406] ath10k_pci 0000:00:00.0: Direct firmware load for
>> ath10k/pre-cal-pci-0000:00:00.0.bin failed with error -2
>> [   27.949483] ath10k_pci 0000:00:00.0: Falling back to user helper
>> [   28.140816] firmware ath10k!pre-cal-pci-0000:00:00.0.bin:
>> firmware_loading_store: map pages failed
>> [   28.150362] ath10k_pci 0000:00:00.0: Direct firmware load for
>> ath10k/cal-pci-0000:00:00.0.bin failed with error -2
>> [   28.161085] ath10k_pci 0000:00:00.0: Falling back to user helper
>> [   35.436780] ath10k_pci 0000:00:00.0: qca988x hw2.0 target 0x4100016c
>> chip_id 0x043222ff sub 0000:0000
>> [   35.446345] ath10k_pci 0000:00:00.0: kconfig debug 0 debugfs 1
>> tracing 0 dfs 1 testmode 1
>> [   35.459369] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.70.59-2 api
>> 5 features no-p2p,raw-mode,mfp crc32 4159f498
>> [   35.470177] ath10k_pci 0000:00:00.0: Direct firmware load for
>> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
>> [   35.480968] ath10k_pci 0000:00:00.0: Falling back to user helper
>> [   35.558782] firmware ath10k!QCA988X!hw2.0!board-2.bin:
>> firmware_loading_store: map pages failed
>> [   35.575232] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
>> crc32 bebc7c08
>> [   36.684593] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
>> cal file max-sta 128 raw 0 hwcrypto 1
>>
>>
>> -Toke
>>> _______________________________________________
>>> Make-wifi-fast mailing list
>>> Make-wifi-fast@lists.bufferbloat.net
>>> https://lists.bufferbloat.net/listinfo/make-wifi-fast
>>>
>>> _______________________________________________
>> Make-wifi-fast mailing list
>> Make-wifi-fast@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/make-wifi-fast
>>
>
> _______________________________________________
> Make-wifi-fast mailing list
> Make-wifi-fast@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/make-wifi-fast

[-- Attachment #2: Type: text/html, Size: 4678 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Noah Causin]

It probably does not have the ath10k softqueues patch from Toke's 
repository.


On 12/25/2016 12:16 PM, Dave Täht wrote:
>
> On 12/25/16 12:06 PM, Toke Høiland-Jørgensen wrote:
>> Valent Turkovic <valent@otvorenamreza.org> writes:
>>
>>> Please send just a short notification as soon as Lede with airfairness
>>> has been released, I have really good real world location that can
>>> test how effective it works...
>> It should be in the nightly LEDE builds by now :)
> I just installed the dec 25th build for the ubnt uaplite.
>
> airtime fairness/fq_codel appears in there for the ath9k.
>
> regrettably the ath10k in that box is not picking up the fq_codel code.
> Too old firmware? Too old chipset?
>
> root@fqcodel:/etc/config# dmesg | grep ath10k
> [   27.720253] ath10k_pci 0000:00:00.0: pci irq legacy oper_irq_mode 1
> irq_mode 0 reset_mode 0
> [   27.938406] ath10k_pci 0000:00:00.0: Direct firmware load for
> ath10k/pre-cal-pci-0000:00:00.0.bin failed with error -2
> [   27.949483] ath10k_pci 0000:00:00.0: Falling back to user helper
> [   28.140816] firmware ath10k!pre-cal-pci-0000:00:00.0.bin:
> firmware_loading_store: map pages failed
> [   28.150362] ath10k_pci 0000:00:00.0: Direct firmware load for
> ath10k/cal-pci-0000:00:00.0.bin failed with error -2
> [   28.161085] ath10k_pci 0000:00:00.0: Falling back to user helper
> [   35.436780] ath10k_pci 0000:00:00.0: qca988x hw2.0 target 0x4100016c
> chip_id 0x043222ff sub 0000:0000
> [   35.446345] ath10k_pci 0000:00:00.0: kconfig debug 0 debugfs 1
> tracing 0 dfs 1 testmode 1
> [   35.459369] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.70.59-2 api
> 5 features no-p2p,raw-mode,mfp crc32 4159f498
> [   35.470177] ath10k_pci 0000:00:00.0: Direct firmware load for
> ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
> [   35.480968] ath10k_pci 0000:00:00.0: Falling back to user helper
> [   35.558782] firmware ath10k!QCA988X!hw2.0!board-2.bin:
> firmware_loading_store: map pages failed
> [   35.575232] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
> crc32 bebc7c08
> [   36.684593] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
> cal file max-sta 128 raw 0 hwcrypto 1
>
>
>> -Toke
>> _______________________________________________
>> Make-wifi-fast mailing list
>> Make-wifi-fast@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/make-wifi-fast
>>
> _______________________________________________
> Make-wifi-fast mailing list
> Make-wifi-fast@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/make-wifi-fast

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Täht]

On 12/25/16 12:06 PM, Toke Høiland-Jørgensen wrote:
> Valent Turkovic <valent@otvorenamreza.org> writes:
> 
>> Please send just a short notification as soon as Lede with airfairness
>> has been released, I have really good real world location that can
>> test how effective it works...
> 
> It should be in the nightly LEDE builds by now :)

I just installed the dec 25th build for the ubnt uaplite.

airtime fairness/fq_codel appears in there for the ath9k.

regrettably the ath10k in that box is not picking up the fq_codel code.
Too old firmware? Too old chipset?

root@fqcodel:/etc/config# dmesg | grep ath10k
[   27.720253] ath10k_pci 0000:00:00.0: pci irq legacy oper_irq_mode 1
irq_mode 0 reset_mode 0
[   27.938406] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/pre-cal-pci-0000:00:00.0.bin failed with error -2
[   27.949483] ath10k_pci 0000:00:00.0: Falling back to user helper
[   28.140816] firmware ath10k!pre-cal-pci-0000:00:00.0.bin:
firmware_loading_store: map pages failed
[   28.150362] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/cal-pci-0000:00:00.0.bin failed with error -2
[   28.161085] ath10k_pci 0000:00:00.0: Falling back to user helper
[   35.436780] ath10k_pci 0000:00:00.0: qca988x hw2.0 target 0x4100016c
chip_id 0x043222ff sub 0000:0000
[   35.446345] ath10k_pci 0000:00:00.0: kconfig debug 0 debugfs 1
tracing 0 dfs 1 testmode 1
[   35.459369] ath10k_pci 0000:00:00.0: firmware ver 10.2.4.70.59-2 api
5 features no-p2p,raw-mode,mfp crc32 4159f498
[   35.470177] ath10k_pci 0000:00:00.0: Direct firmware load for
ath10k/QCA988X/hw2.0/board-2.bin failed with error -2
[   35.480968] ath10k_pci 0000:00:00.0: Falling back to user helper
[   35.558782] firmware ath10k!QCA988X!hw2.0!board-2.bin:
firmware_loading_store: map pages failed
[   35.575232] ath10k_pci 0000:00:00.0: board_file api 1 bmi_id N/A
crc32 bebc7c08
[   36.684593] ath10k_pci 0000:00:00.0: htt-ver 2.1 wmi-op 5 htt-op 2
cal file max-sta 128 raw 0 hwcrypto 1


> 
> -Toke
> _______________________________________________
> Make-wifi-fast mailing list
> Make-wifi-fast@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/make-wifi-fast
> 

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Toke Høiland-Jørgensen]

Valent Turkovic <valent@otvorenamreza.org> writes:

> Please send just a short notification as soon as Lede with airfairness
> has been released, I have really good real world location that can
> test how effective it works...

It should be in the nightly LEDE builds by now :)

-Toke

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Valent Turkovic]

[-- Attachment #1: Type: text/plain, Size: 169 bytes --]

Please send just a short notification as soon as Lede with airfairness has
been released, I have really good real world location that can test how
effective it works...

[-- Attachment #2: Type: text/html, Size: 194 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

Heh. Of all the hackerboards I tried, the odroid c2 was the fastest
and is generally my default target these days.

The "chip" is coming along nicely but lacks network.

The odroid c1 and rpi2 have been the most reliable.

(I'm still looking for a board with a decent toslink port....)

...

I just foolishly tried to update my linksys 1200ac, but... it turned
out the cable modem behind it was acting up and I lost touch with it
(lacking a babel pre-installed)... aha! (/me worms his way through
multiple hops through another modem)

It looks like my default command line, post install is basically:

opkg install luci-ssl luci-app-sqm babeld kmod-sched-cake netperf

The 1200ac still has bug

https://bugs.lede-project.org/index.php?do=details&task_id=294&order=status&sort=asc&order2=severity&sort2=desc

I triggered it with:

root@apu2:~/t# flent -H 172.26.64.1 -t 'isitfixed'
--te=download_streams=12 tcp_ndown

Warning: Program exited non-zero (1).
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4  -H 172.26.64.1 -t
TCP_MAERTS -l 60 -f m   --   -H 172.26.64.1
Program output:
  netperf: send_omni: connect_data_socket failed: No route to host

Warning: Command produced no valid data.
Data series: TCP download::5
Runner: NetperfDemoRunner
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4  -H 172.26.64.1 -t
TCP_MAERTS -l 60 -f m   --   -H 172.26.64.1
Standard error output:
  netperf: send_omni: connect_data_socket failed: No route to host

The way I was dealing with it was with running cake at 900mbit on the
internal ethernet. I guess it's one way around bugs like this.




On Fri, Dec 23, 2016 at 6:51 PM, Aaron Wood <woody77@gmail.com> wrote:
> Excellent. I'll run more tests, and see if I can get some more boxes with
> flent installed. I'm short of test clients, myself.
>
> -Aaron
>
> On Fri, Dec 23, 2016 at 18:29 Dave Taht <dave.taht@gmail.com> wrote:
>>
>> On Fri, Dec 23, 2016 at 6:18 PM, Aaron Wood <woody77@gmail.com> wrote:
>> > On Fri, Dec 23, 2016 at 6:03 PM, Dave Taht <dave.taht@gmail.com> wrote:
>> >>
>> >> The simplest way to verify if you have the airtime fairness stuff is
>> >> to look for a sysfs file named "airtime". The best way to test it is
>> >> to drive more than one wifi device over it simultaneously from a wired
>> >> link, and not through the ISP link.
>> >
>> >
>> > /sys/kernel/debug/ieee80211/phy[N]/ath9k/airtime_flags?
>>
>> yep. there's also a per station one that only a parser could love.
>> Example:
>>
>>
>> ./kernel/debug/ieee80211/phy0/netdev:wlp2s0/stations/04:f0:21:1f:36:e2/airtime
>>
>> I'm getting on a plane to florida in a few hours, but I'm packing a
>> ubntlite with me (as a gift for mom). Sounds like no more debugging
>> required....
>>
>> It sounds like we are on the verge of breaking out the eggnog, and
>> raising a glass... (but I'd like a few more testers first - got a c7
>> anyone?)
>>
>> A merry bloat-free christmas to all, and to all, a good firmware!
>>
>> >>
>> >>
>> >> flent -H device1 -H device2 -H device1 -H device2 rtt_fair_up # example
>> >>
>> >> What is your wifi client presently?
>> >
>> >
>> > Mid 2015 MacBook Pro, very short range (<10 feet).  My test bed is down
>> > for
>> > the evening, sadly.
>>
>>
>>
>> --
>> Dave Täht
>> Let's go make home routers and wifi faster! With better software!
>> http://blog.cerowrt.org



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 1657 bytes --]

Excellent. I'll run more tests, and see if I can get some more boxes with
flent installed. I'm short of test clients, myself.

-Aaron

On Fri, Dec 23, 2016 at 18:29 Dave Taht <dave.taht@gmail.com> wrote:

> On Fri, Dec 23, 2016 at 6:18 PM, Aaron Wood <woody77@gmail.com> wrote:
> > On Fri, Dec 23, 2016 at 6:03 PM, Dave Taht <dave.taht@gmail.com> wrote:
> >>
> >> The simplest way to verify if you have the airtime fairness stuff is
> >> to look for a sysfs file named "airtime". The best way to test it is
> >> to drive more than one wifi device over it simultaneously from a wired
> >> link, and not through the ISP link.
> >
> >
> > /sys/kernel/debug/ieee80211/phy[N]/ath9k/airtime_flags?
>
> yep. there's also a per station one that only a parser could love. Example:
>
>
> ./kernel/debug/ieee80211/phy0/netdev:wlp2s0/stations/04:f0:21:1f:36:e2/airtime
>
> I'm getting on a plane to florida in a few hours, but I'm packing a
> ubntlite with me (as a gift for mom). Sounds like no more debugging
> required....
>
> It sounds like we are on the verge of breaking out the eggnog, and
> raising a glass... (but I'd like a few more testers first - got a c7
> anyone?)
>
> A merry bloat-free christmas to all, and to all, a good firmware!
>
> >>
> >>
> >> flent -H device1 -H device2 -H device1 -H device2 rtt_fair_up # example
> >>
> >> What is your wifi client presently?
> >
> >
> > Mid 2015 MacBook Pro, very short range (<10 feet).  My test bed is down
> for
> > the evening, sadly.
>
>
>
> --
> Dave Täht
> Let's go make home routers and wifi faster! With better software!
> http://blog.cerowrt.org
>

[-- Attachment #2: Type: text/html, Size: 3113 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 1407 bytes --]

I tried higher power early on with the wrt1900ac, but it seemed to cause
problems, although I don't think I was on that high of a channel.

-Aaron

On Fri, Dec 23, 2016 at 18:33 Dave Taht <dave.taht@gmail.com> wrote:

> On Fri, Dec 23, 2016 at 6:25 PM, Aaron Wood <woody77@gmail.com> wrote:
> > On Fri, Dec 23, 2016 at 6:24 PM, Dave Taht <dave.taht@gmail.com> wrote:
> >>
> >> On Fri, Dec 23, 2016 at 6:22 PM, Aaron Wood <woody77@gmail.com> wrote:
> >> > On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com>
> wrote:
> >> >>
> >> >> your 3800 is outperforming the ac1900 across the board at this
> >> >> distance????
> >> >
> >> >
> >> > At short range, the wrt1900ac running OpenWRT is still out-performing
> >> > the
> >> > 3800 running LEDE.  The wrt1900ac has some _very_ nice radios (or
> maybe
> >> > it's
> >> > just the much better antennas).  I'll do longer-range tests tomorrow,
> if
> >> > I
> >> > get a chance.
> >>
> >> Are you in HT40 mode?
> >
> >
> > Yes, on both.  Also on ch 44 for both.
>
> At distance I have generally found the higher tx power available helps
> on the higher channels, btw. I can't remember right now, but 153 and
> higher could do like txpower 30, and are less commonly used.
>
> > -Aaron
>
>
>
> --
> Dave Täht
> Let's go make home routers and wifi faster! With better software!
> http://blog.cerowrt.org
>

[-- Attachment #2: Type: text/html, Size: 2956 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Fri, Dec 23, 2016 at 6:25 PM, Aaron Wood <woody77@gmail.com> wrote:
> On Fri, Dec 23, 2016 at 6:24 PM, Dave Taht <dave.taht@gmail.com> wrote:
>>
>> On Fri, Dec 23, 2016 at 6:22 PM, Aaron Wood <woody77@gmail.com> wrote:
>> > On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:
>> >>
>> >> your 3800 is outperforming the ac1900 across the board at this
>> >> distance????
>> >
>> >
>> > At short range, the wrt1900ac running OpenWRT is still out-performing
>> > the
>> > 3800 running LEDE.  The wrt1900ac has some _very_ nice radios (or maybe
>> > it's
>> > just the much better antennas).  I'll do longer-range tests tomorrow, if
>> > I
>> > get a chance.
>>
>> Are you in HT40 mode?
>
>
> Yes, on both.  Also on ch 44 for both.

At distance I have generally found the higher tx power available helps
on the higher channels, btw. I can't remember right now, but 153 and
higher could do like txpower 30, and are less commonly used.

> -Aaron



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Fri, Dec 23, 2016 at 6:22 PM, Aaron Wood <woody77@gmail.com> wrote:
> On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:
>>
>> your 3800 is outperforming the ac1900 across the board at this
>> distance????
>
>
> At short range, the wrt1900ac running OpenWRT is still out-performing the
> 3800 running LEDE.  The wrt1900ac has some _very_ nice radios (or maybe it's
> just the much better antennas).  I'll do longer-range tests tomorrow, if I
> get a chance.

I'd like you to use tcp_ndown and tcp_nup for a fairer comparison than
rrul, with 1,2,3,8,12 streams... but I am quite happy with all this so
far. Looks great.




> -Aaron
>



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Fri, Dec 23, 2016 at 6:18 PM, Aaron Wood <woody77@gmail.com> wrote:
> On Fri, Dec 23, 2016 at 6:03 PM, Dave Taht <dave.taht@gmail.com> wrote:
>>
>> The simplest way to verify if you have the airtime fairness stuff is
>> to look for a sysfs file named "airtime". The best way to test it is
>> to drive more than one wifi device over it simultaneously from a wired
>> link, and not through the ISP link.
>
>
> /sys/kernel/debug/ieee80211/phy[N]/ath9k/airtime_flags?

yep. there's also a per station one that only a parser could love. Example:

./kernel/debug/ieee80211/phy0/netdev:wlp2s0/stations/04:f0:21:1f:36:e2/airtime

I'm getting on a plane to florida in a few hours, but I'm packing a
ubntlite with me (as a gift for mom). Sounds like no more debugging
required....

It sounds like we are on the verge of breaking out the eggnog, and
raising a glass... (but I'd like a few more testers first - got a c7
anyone?)

A merry bloat-free christmas to all, and to all, a good firmware!

>>
>>
>> flent -H device1 -H device2 -H device1 -H device2 rtt_fair_up # example
>>
>> What is your wifi client presently?
>
>
> Mid 2015 MacBook Pro, very short range (<10 feet).  My test bed is down for
> the evening, sadly.



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 657 bytes --]

On Fri, Dec 23, 2016 at 6:24 PM, Dave Taht <dave.taht@gmail.com> wrote:

> On Fri, Dec 23, 2016 at 6:22 PM, Aaron Wood <woody77@gmail.com> wrote:
> > On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:
> >>
> >> your 3800 is outperforming the ac1900 across the board at this
> >> distance????
> >
> >
> > At short range, the wrt1900ac running OpenWRT is still out-performing the
> > 3800 running LEDE.  The wrt1900ac has some _very_ nice radios (or maybe
> it's
> > just the much better antennas).  I'll do longer-range tests tomorrow, if
> I
> > get a chance.
>
> Are you in HT40 mode?
>

Yes, on both.  Also on ch 44 for both.

-Aaron

[-- Attachment #2: Type: text/html, Size: 1189 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Fri, Dec 23, 2016 at 6:22 PM, Aaron Wood <woody77@gmail.com> wrote:
> On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:
>>
>> your 3800 is outperforming the ac1900 across the board at this
>> distance????
>
>
> At short range, the wrt1900ac running OpenWRT is still out-performing the
> 3800 running LEDE.  The wrt1900ac has some _very_ nice radios (or maybe it's
> just the much better antennas).  I'll do longer-range tests tomorrow, if I
> get a chance.

Are you in HT40 mode?

> -Aaron
>



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1.1: Type: text/plain, Size: 402 bytes --]

On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:

> your 3800 is outperforming the ac1900 across the board at this distance????
>

At short range, the wrt1900ac running OpenWRT is still out-performing the
3800 running LEDE.  The wrt1900ac has some _very_ nice radios (or maybe
it's just the much better antennas).  I'll do longer-range tests tomorrow,
if I get a chance.

-Aaron

[-- Attachment #1.2: Type: text/html, Size: 753 bytes --]

[-- Attachment #2: boxes_wifi.png --]
[-- Type: image/png, Size: 45428 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 602 bytes --]

On Fri, Dec 23, 2016 at 6:03 PM, Dave Taht <dave.taht@gmail.com> wrote:

> The simplest way to verify if you have the airtime fairness stuff is
> to look for a sysfs file named "airtime". The best way to test it is
> to drive more than one wifi device over it simultaneously from a wired
> link, and not through the ISP link.
>

/sys/kernel/debug/ieee80211/phy[N]/ath9k/airtime_flags?


>
> flent -H device1 -H device2 -H device1 -H device2 rtt_fair_up # example
>
> What is your wifi client presently?
>

Mid 2015 MacBook Pro, very short range (<10 feet).  My test bed is down for
the evening, sadly.

[-- Attachment #2: Type: text/html, Size: 1236 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 993 bytes --]

On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:

> your 3800 is outperforming the ac1900 across the board at this distance????
>

Wired, 3800 running the latest LEDE snapshot, and the ac1900 running the
OpenWRT CC release.  They have _slightly_ different SQM settings, though:

3800:
LEDE 2016/12/22 snapshot
"piece of cake"
135Mbps down
12Mbps up
DSCP no squash, don't ignore on ingress
ECN

wrt1900ac:
OpenWRT Chaos Calmer rc2
simple.qos
160Mbps down
12Mbps up
DSCP no squash, allow on ingress
ECN

So I'm running a single-queue on the 3800, and multi-queue on the wrt.  The
WRT has always counted bytes in an "off" manner, limiting to _far_ below
the set level.  Unlimited on this modem is up at 180Mbps downstream today.

Today's tests with LEDE on the 3800 were a warm-up for moving the wrt1900ac
over to LEDE.  But I wanted a backup in case something goes awry (nearly
bricked the 3800 as I forgot to uncheck "keep settings" on the
OpenWRT->LEDE upgrade.

-Aaron

[-- Attachment #2: Type: text/html, Size: 1597 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

and your rrul result I regard to be impossible unless wifi is not your
bottleneck or all your markings are getting washed away or you have
802.11e disabled (which I long have recommended disabling anyway).

On Fri, Dec 23, 2016 at 6:07 PM, Dave Taht <dave.taht@gmail.com> wrote:
> your 3800 is outperforming the ac1900 across the board at this distance????



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

your 3800 is outperforming the ac1900 across the board at this distance????

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

The simplest way to verify if you have the airtime fairness stuff is
to look for a sysfs file named "airtime". The best way to test it is
to drive more than one wifi device over it simultaneously from a wired
link, and not through the ISP link.

flent -H device1 -H device2 -H device1 -H device2 rtt_fair_up # example

What is your wifi client presently?

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Jonathan Morton]

> On 24 Dec, 2016, at 03:59, Dave Taht <dave.taht@gmail.com> wrote:
> 
>> I, however, am _very_ happy with the results at those levels (using piece of
>> cake, not layer-cake, layer-cake is flapping bandwidth between BK and the
>> rest in my tests).
> 
> Impressive.

I did pay some attention to optimising the core algorithms recently.  I’m glad that seems to be paying off.

 - Jonathan Morton

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Fri, Dec 23, 2016 at 5:48 PM, Aaron Wood <woody77@gmail.com> wrote:
>
>
> On Fri, Dec 23, 2016 at 5:41 PM, Dave Taht <dave.taht@gmail.com> wrote:
>>
>> On Fri, Dec 23, 2016 at 5:32 PM, Aaron Wood <woody77@gmail.com> wrote:
>> > Toke,
>> >
>> > Do you know if it's made it into the LEDE nightlies, yet?  I loaded up
>> > the
>> > Dec 22 nightly onto my WNDR3800.  Running rrul I'm getting about 10-30ms
>> > of
>> > latency over wired on the same link (and wifi is definitely the limit,
>> > as
>>
>> I'm assuming you meant wireless above?
>
>
> Yes.
>
>>
>> > it's using piece_of_cake at 135M down, 12M up (and I'm seeing 0
>> > measurable
>>
>> I really don't recommend cake at much greater than 60Mbits on this
>> hardware inbound.
>
>
> I, however, am _very_ happy with the results at those levels (using piece of
> cake, not layer-cake, layer-cake is flapping bandwidth between BK and the
> rest in my tests).

Impressive.

I haven't gone and measured cake recently on low end routers. Certainly less
classes helps it. It would be good if this performance level is true
for the current
implementation with a single queue.


>> It's not clear to me if the quantum and codel targets were restored to
>> their defaults in this release (300->1514, 20ms -> 5ms)
>
>
> And... my WRT running OpenWRT CC is using quantum 300, target 5ms.

The actual target and quantum are in one of the sysfs aqm files, I believe.

Also I note most of the benefit of the new wifi code is at lower rates
from your AP. Try moving 20 or more feet away....

>
> -Aaron



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Aaron Wood]

[-- Attachment #1.1: Type: text/plain, Size: 1007 bytes --]

On Fri, Dec 23, 2016 at 5:41 PM, Dave Taht <dave.taht@gmail.com> wrote:

> On Fri, Dec 23, 2016 at 5:32 PM, Aaron Wood <woody77@gmail.com> wrote:
> > Toke,
> >
> > Do you know if it's made it into the LEDE nightlies, yet?  I loaded up
> the
> > Dec 22 nightly onto my WNDR3800.  Running rrul I'm getting about 10-30ms
> of
> > latency over wired on the same link (and wifi is definitely the limit, as
>
> I'm assuming you meant wireless above?
>

Yes.


> > it's using piece_of_cake at 135M down, 12M up (and I'm seeing 0
> measurable
>
> I really don't recommend cake at much greater than 60Mbits on this
> hardware inbound.
>

I, however, am _very_ happy with the results at those levels (using piece
of cake, not layer-cake, layer-cake is flapping bandwidth between BK and
the rest in my tests).

It's not clear to me if the quantum and codel targets were restored to
> their defaults in this release (300->1514, 20ms -> 5ms)
>

And... my WRT running OpenWRT CC is using quantum 300, target 5ms.

-Aaron

[-- Attachment #1.2: Type: text/html, Size: 1827 bytes --]

[-- Attachment #2: wndr3800_lede_piece_cake_135_12.all.png --]
[-- Type: image/png, Size: 576284 bytes --]

[-- Attachment #3: wndr3800_wired_before_after.png --]
[-- Type: image/png, Size: 46718 bytes --]

[-- Attachment #4: wndr3800_lede_vs_wrt1900ac_cc.png --]
[-- Type: image/png, Size: 43111 bytes --]

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Fri, Dec 23, 2016 at 5:32 PM, Aaron Wood <woody77@gmail.com> wrote:
> Toke,
>
> Do you know if it's made it into the LEDE nightlies, yet?  I loaded up the
> Dec 22 nightly onto my WNDR3800.  Running rrul I'm getting about 10-30ms of
> latency over wired on the same link (and wifi is definitely the limit, as

I'm assuming you meant wireless above?

> it's using piece_of_cake at 135M down, 12M up (and I'm seeing 0 measurable

I really don't recommend cake at much greater than 60Mbits on this
hardware inbound.

> additional latency from piece of cake).

rrul stresses out both sides of the wifi connection AND all the Qos queues.

rrul_be stresses out both sides but only one queue.

rrul remains generally awful on wifi unless your dscp markings have
been eliminated in transit.

for rrul_be - unless you've got the fq_codel code on both sides, you
aren't going to see as much benefit, and even if you are, the double
buffering on both sides leads to a typical latency of about 10ms each
way in this scenario. (so if you got it, you are inline with our
results).

I generally switched to using tcp_nup and tcp_ndown (with varying
numbers of flows), as well as the various rtt_fair_var tests against
multiple stations, to evaluate the wifi changes.

It's not clear to me if the quantum and codel targets were restored to
their defaults in this release (300->1514, 20ms -> 5ms)



> -Aaron
>
>
>
> On Mon, Dec 19, 2016 at 2:31 PM, Toke Høiland-Jørgensen <toke@toke.dk>
> wrote:
>>
>> The airtime fairness patch for ath9k has been accepted into the mainline
>> Linux kernel, and is queued to be merged for 4.11 (i.e. the next merge
>> window after the current 4.10 cycle completes).
>>
>> The patch has also been accepted into LEDE and is currently in Felix'
>> staging tree at
>> https://git.lede-project.org/?p=lede/nbd/staging.git;a=summary - so if
>> nothing surprising shows up, it should make it into the regular LEDE
>> nightlies before too long :)
>>
>>
>> In related news, the Turris Omnia had the WiFi queue restructure patches
>> added in the latest update; and the ath10k chip it ships with seems to
>> have the required hardware support, so both WiFi devices in the Omnia
>> now have nicely debloated queues. They also switched to sqm-scripts
>> (from wondershaper that they shipped before), and added Cake :)
>>
>> -Toke
>> _______________________________________________
>> Make-wifi-fast mailing list
>> Make-wifi-fast@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/make-wifi-fast
>
>
>
> _______________________________________________
> Cake mailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake
>



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

Also, I've been testing the new cake code and thus far it doesn't
crash, at least, and appears to be giving the desired results.

On Mon, Dec 19, 2016 at 4:10 PM, Dave Taht <dave.taht@gmail.com> wrote:
> I note that jon just merged cobalt and made the sqm mode (diffserv3)
> and triple-isolate the default.
>
> On Mon, Dec 19, 2016 at 2:31 PM, Toke Høiland-Jørgensen <toke@toke.dk> wrote:
>> The airtime fairness patch for ath9k has been accepted into the mainline
>> Linux kernel, and is queued to be merged for 4.11 (i.e. the next merge
>> window after the current 4.10 cycle completes).
>>
>> The patch has also been accepted into LEDE and is currently in Felix'
>> staging tree at
>> https://git.lede-project.org/?p=lede/nbd/staging.git;a=summary - so if
>> nothing surprising shows up, it should make it into the regular LEDE
>> nightlies before too long :)
>>
>>
>> In related news, the Turris Omnia had the WiFi queue restructure patches
>> added in the latest update; and the ath10k chip it ships with seems to
>> have the required hardware support, so both WiFi devices in the Omnia
>> now have nicely debloated queues. They also switched to sqm-scripts
>> (from wondershaper that they shipped before), and added Cake :)
>>
>> -Toke
>> _______________________________________________
>> Cake mailing list
>> Cake@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/cake
>
>
>
> --
> Dave Täht
> Let's go make home routers and wifi faster! With better software!
> http://blog.cerowrt.org



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

On Mon, Dec 19, 2016 at 2:31 PM, Toke Høiland-Jørgensen <toke@toke.dk> wrote:
> The airtime fairness patch for ath9k has been accepted into the mainline
> Linux kernel, and is queued to be merged for 4.11 (i.e. the next merge
> window after the current 4.10 cycle completes).
>
> The patch has also been accepted into LEDE and is currently in Felix'
> staging tree at
> https://git.lede-project.org/?p=lede/nbd/staging.git;a=summary - so if
> nothing surprising shows up, it should make it into the regular LEDE
> nightlies before too long :)

w00t.
>
> In related news, the Turris Omnia had the WiFi queue restructure patches
> added in the latest update; and the ath10k chip it ships with seems to
> have the required hardware support, so both WiFi devices in the Omnia
> now have nicely debloated queues. They also switched to sqm-scripts
> (from wondershaper that they shipped before), and added Cake :)

w00t w00t. Are you (or anyone here) in a position to bench the ath10k side?

> -Toke
> _______________________________________________
> Cake m.ailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Make-wifi-fast] [Cake] Airtime fairness patch merged upstream, and the Turris Omnia debloated

[Dave Taht]

I note that jon just merged cobalt and made the sqm mode (diffserv3)
and triple-isolate the default.

On Mon, Dec 19, 2016 at 2:31 PM, Toke Høiland-Jørgensen <toke@toke.dk> wrote:
> The airtime fairness patch for ath9k has been accepted into the mainline
> Linux kernel, and is queued to be merged for 4.11 (i.e. the next merge
> window after the current 4.10 cycle completes).
>
> The patch has also been accepted into LEDE and is currently in Felix'
> staging tree at
> https://git.lede-project.org/?p=lede/nbd/staging.git;a=summary - so if
> nothing surprising shows up, it should make it into the regular LEDE
> nightlies before too long :)
>
>
> In related news, the Turris Omnia had the WiFi queue restructure patches
> added in the latest update; and the ath10k chip it ships with seems to
> have the required hardware support, so both WiFi devices in the Omnia
> now have nicely debloated queues. They also switched to sqm-scripts
> (from wondershaper that they shipped before), and added Cake :)
>
> -Toke
> _______________________________________________
> Cake mailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org