From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-oi0-x231.google.com (mail-oi0-x231.google.com [IPv6:2607:f8b0:4003:c06::231]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 759473B2A4; Thu, 30 Mar 2017 00:09:30 -0400 (EDT) Received: by mail-oi0-x231.google.com with SMTP id r203so18730238oib.3; Wed, 29 Mar 2017 21:09:30 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:sender:from:date:message-id:subject:to; bh=TbrG/WWDM2sEwhSL8TuKVVL8DGLQj/ChnhubdNgeAJw=; b=ehdnHUPp/z+hA5EiQ3w9YvvPUgFxF3HdiBQq/vhp5l5VduogbJXUqIslBPVQsXzfpY UcHECb7nfpddiflEkFASTaa8XOuyGqLZ+St7XFxaiPAe0RbfISgerK/WQic+TljMyJFP NRYqmDXUpIky/qgjdfkwlK+lkYA0yTF3PtVQnzUPlqaXHmJc/CnR0WAtmH2E8XGjBTjf r+1Zu6fY3rvtDpbeSZifG+cRy0PV2kdQ2vQ1NworE8lN9RrJJx5SJk53PGO2HhaZluxB 3omvoVmWMKvGNS5D4TcsVKDsehq2z2if3g7fmgbyICJ6cr6CUoD1hJ7bWDjQtswUiMIc LYTA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:sender:from:date:message-id:subject :to; bh=TbrG/WWDM2sEwhSL8TuKVVL8DGLQj/ChnhubdNgeAJw=; b=Ir17J9JfZK9tpaJmZw9U7nWX9rp3I9MBgV9+InfJnlL8mZC/aUCURIVIry2IiW/3yf yM8itUg1GuhMUszqB0jhNu8WHePTFP30/jHTWEnQp+/XqcdfpliiL5FBAGJP24st5GHs GgyABowizDh9oOuQiGp+wGSD8mvwcX2dbDX1HuQS37g1BeovVA8hkRC4pZKdw1MLWduL sb/e0XFcyC6YG3X97fJsYJuKyh5uHBQqSM+pK0wxVZBdWEHIAE2h3C4ESFMP8hv0erus umHuibNui+MCGw57ivAg1rmrDp1ZP4eVQ0xXXoCQ7PNiuJHj41+Ud7LlAU1se6H3wRxg H/LA== X-Gm-Message-State: AFeK/H00kVwrUClm7dH/SyTNmIqSwWsAphc3H/Y/DzipBquhFLYOSfTZGM+39CptFzd/DjbSSESjelRp7KIOWQ== X-Received: by 10.157.33.247 with SMTP id s110mr1976449otb.75.1490846969528; Wed, 29 Mar 2017 21:09:29 -0700 (PDT) MIME-Version: 1.0 Sender: white.phoenix@gmail.com Received: by 10.182.114.5 with HTTP; Wed, 29 Mar 2017 21:08:59 -0700 (PDT) From: leetminiwheat Date: Thu, 30 Mar 2017 00:08:59 -0400 X-Google-Sender-Auth: sAoibj2IYiNKqAYBzYK0ZGDfxBk Message-ID: To: cerowrt-devel , cake@lists.bufferbloat.net Content-Type: multipart/alternative; boundary=001a11493d9a0889e8054beadc50 Subject: [Cake] Current status and recommendations for APU2C*/ArcherC7 and ath10k platforms? X-BeenThere: cake@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: Cake - FQ_codel the next generation List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 30 Mar 2017 04:09:30 -0000 --001a11493d9a0889e8054beadc50 Content-Type: text/plain; charset=UTF-8 Apologies I haven't had time to follow recent developments but are there any recent bufferbloat/CAKE related news/recommendations for the APU2 platform? Specifically the APU2C (previous was marvell I believe) Hardware: APU2C4 (AMD x86_64 | 4GB | 3x i210-AT) WLE600VX (ath10k) 64GB mSATA Also have an Archer C7 V2 (ath10k) and a old WDR3800 (ath9k) still chugging along. Looking for an optimal Archer C7 build or recommendations if anyone has any experience on it. Use case: linespeed: 105/120 fiber -> APU2C4 primary gateway (and some WiFi) -> Archer C7 & WNDR3800 for internal & guest networks (undecided which of the last two for each task). My plan is to use separate networks for bulk traffic others for latency sensitive stuff. Anyways I searched through mailing lists and google but thought this might be the best place(s) to ask. I noticed dtaht is (or was) using an APU2 (though not sure which version of it since one of his posts refers to mvneta, the marvell platform). Also saw that Dave's tc-adv didn't make it upstream? >From what I gathered, I found these relevant but perhaps outdated: http://blog.cerowrt.org/post/birthday_problem/ (refers to APU2 and it's 4 hardware queues but was this marvell/mvneta?) https://github.com/riptidewave93/LEDE-APU2 (Older github repo for APU2 LEDE support but seems upstream now?) https://github.com/Gr4ffy/LEDE-APU2 (older fork from above riptidewave93 LEDE-APU2) https://github.com/greearb/lede-source-ct (ath10k related; LEDE fork from Candela in which they apparently obtained the ath10k firmware source code NDA) https://github.com/dtaht/sch_cake/issues/46 (Fairness issue and a reference to dthat's custom tc not going upstream) [multiple mailing list links] (References to airtime-fairness? is this primarily for ath9k?) [APU2 ubuntu repo link I lost] Saw a repo for dtaht's ubuntu build for APU2 somewhere?!? Or was it under another name? If anyone knows the URL I'd love to borrow some of the settings for testing. https://github.com/pcengines/coreboot https://github.com/pcengines/seabios (APU2 related, they seem to still update their bios but they don't release much. I also saw some commits referring to ECC support on APU3, I guess ECC on the APU2 is a lost cause) Any other more appropriate lists or IRC channels I might inquire in would also be helpful. P.S. My condolences for the loss of your lab @ Dave, that truly is a travesty. Thanks --001a11493d9a0889e8054beadc50 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Apologies I haven't had time to follow recent dev= elopments but are there
=C2=A0any recent bufferbloat/CAKE related= news/recommendations for the APU2
platform? Specifically the APU= 2C (previous was marvell I believe)

Hardware:=C2= =A0
APU2C4 (AMD x86_64 | 4GB | 3x i210-AT)
WLE600VX (at= h10k)
64GB mSATA

Also have an Archer C7 = V2 (ath10k) and a old WDR3800 (ath9k) still=C2=A0
chugging along.= =C2=A0

Looking for an optimal Archer C7 build or r= ecommendations if anyone=C2=A0
has any experience on it.

Use case:=C2=A0
linespeed: 105/120 fiber -> A= PU2C4 primary gateway (and some WiFi) ->=C2=A0
Archer C7 &= WNDR3800 for internal & guest networks=C2=A0
(undecided whic= h of the last two for each task). My plan is to use=C2=A0
separat= e networks for bulk traffic others for latency sensitive stuff.
<= br>
Anyways I searched through mailing lists and google but thoug= ht this=C2=A0
might be the best place(s) to ask.
I noti= ced dtaht is (or was) using an APU2 (though not sure which version
=C2=A0of it since one of his posts refers to mvneta, the marvell platform= ).=C2=A0
Also saw that Dave's tc-adv didn't make it upstr= eam?


From what I gathered, I found = these relevant but perhaps outdated:

(refers to APU2 and it's 4 hardwa= re queues but was this marvell/mvneta?)

(Older github repo for APU2 LEDE support but see= ms upstream now?)

(older fo= rk from above riptidewave93 LEDE-APU2)

(ath10k related; LEDE fork from Candela in which th= ey apparently=C2=A0
obtained the ath10k firmware source code NDA)=

(Fairness = issue and a reference to dthat's custom tc not going upstream)

[multiple mailing list links]
(References to air= time-fairness? is =C2=A0this primarily for ath9k?)

[APU2 ubuntu repo link I lost]
Saw a repo for dtaht's ubuntu= build for APU2 somewhere?!? Or was it
under another name? If any= one knows the URL I'd love to borrow some=C2=A0
of the settin= gs for testing.

(APU2 related, they seem to still update their bios but t= hey don't=C2=A0
release much.=C2=A0
I also saw some= commits referring to ECC support on APU3, I guess ECC on
=C2=A0t= he APU2 is a lost cause)


Any other = more appropriate lists or IRC channels I might inquire in would also
<= div>=C2=A0be helpful.=C2=A0


P.S. My= condolences for the loss of your lab @ Dave, that truly is a=C2=A0
travesty.

Thanks
--001a11493d9a0889e8054beadc50--