From mboxrd@z Thu Jan 1 00:00:00 1970 From: jbrunet@baylibre.com (Jerome Brunet) Date: Tue, 15 Nov 2016 12:27:56 +0100 Subject: stmmac/RTL8211F/Meson GXBB: TX throughput problems In-Reply-To: <24b4d9b3-0a9d-5f6c-7dc2-5fcbb1012481@suse.de> References: <20161031112524.4c4ee8f8@gmail.com> <1478192276.6632.34.camel@baylibre.com> <3c21b0a4-43c9-0257-f8f2-c8e02cf94fbf@st.com> <24b4d9b3-0a9d-5f6c-7dc2-5fcbb1012481@suse.de> Message-ID: <1479209276.2619.10.camel@baylibre.com> To: linus-amlogic@lists.infradead.org List-Id: linus-amlogic.lists.infradead.org On Mon, 2016-11-14 at 16:00 +0100, Andreas F?rber wrote: > Hi, > > Am 07.11.2016 um 18:37 schrieb Martin Blumenstingl: > > > > The same problem still appears on my Tronsmart Vega S95 Meta even > > with > > the patched PHY driver. > > > > Unfortunately I don't have a second device to rule out that my > > Tronsmart Vega S95 Meta could be broken (not unlikely, I get DDR > > errors from time to time in u-boot). Maybe Andreas Faerber can test > > ethernet with and without Jerome's patch on one of his Tronsmart > > devices. > > So far I only ran into the stall issue on my Odroid-C2, not on my > Vega > S95 Telos, both on v4.10/integ branch with a rebased MMC patch of > yours > on top. The status of vega/mxq boards seems a bit confused. This is the status I have so far: These designs are not affected by the EEE issue. The patch mentioned has no effect on these boards. Some boards run fine at GBit speed (like yours) , some have terrible Tx throughput (like mine and Martin's one). However, unlike on the odroidc2, the link is not broken after the test, it is not necessary to ifdown/ifup the interface to do another test. I'm wondering if the vega/mxq designs all use the rtl8211f phy ? > > I'm assuming you mean realtek8211f-disable-eee-1000.patch from Nov 3. > Indeed. I won't submit anything for the vega since it has no effect. > Regards, > Andreas > From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jerome Brunet Subject: Re: stmmac/RTL8211F/Meson GXBB: TX throughput problems Date: Tue, 15 Nov 2016 12:27:56 +0100 Message-ID: <1479209276.2619.10.camel@baylibre.com> References: <20161031112524.4c4ee8f8@gmail.com> <1478192276.6632.34.camel@baylibre.com> <3c21b0a4-43c9-0257-f8f2-c8e02cf94fbf@st.com> <24b4d9b3-0a9d-5f6c-7dc2-5fcbb1012481@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8bit Cc: Giuseppe CAVALLARO , Alexandre Torgue , netdev@vger.kernel.org, =?ISO-8859-1?Q?Andr=E9?= Roth , Johnson Leung , linux-amlogic@lists.infradead.org To: Andreas =?ISO-8859-1?Q?F=E4rber?= , Martin Blumenstingl Return-path: Received: from mail-wm0-f44.google.com ([74.125.82.44]:36998 "EHLO mail-wm0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751540AbcKOL17 (ORCPT ); Tue, 15 Nov 2016 06:27:59 -0500 Received: by mail-wm0-f44.google.com with SMTP id t79so161895879wmt.0 for ; Tue, 15 Nov 2016 03:27:58 -0800 (PST) In-Reply-To: <24b4d9b3-0a9d-5f6c-7dc2-5fcbb1012481@suse.de> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 2016-11-14 at 16:00 +0100, Andreas Färber wrote: > Hi, > > Am 07.11.2016 um 18:37 schrieb Martin Blumenstingl: > > > > The same problem still appears on my Tronsmart Vega S95 Meta even > > with > > the patched PHY driver. > > > > Unfortunately I don't have a second device to rule out that my > > Tronsmart Vega S95 Meta could be broken (not unlikely, I get DDR > > errors from time to time in u-boot). Maybe Andreas Faerber can test > > ethernet with and without Jerome's patch on one of his Tronsmart > > devices. > > So far I only ran into the stall issue on my Odroid-C2, not on my > Vega > S95 Telos, both on v4.10/integ branch with a rebased MMC patch of > yours > on top. The status of vega/mxq boards seems a bit confused. This is the status I have so far: These designs are not affected by the EEE issue. The patch mentioned has no effect on these boards. Some boards run fine at GBit speed (like yours) , some have terrible Tx throughput (like mine and Martin's one). However, unlike on the odroidc2, the link is not broken after the test, it is not necessary to ifdown/ifup the interface to do another test. I'm wondering if the vega/mxq designs all use the rtl8211f phy ? > > I'm assuming you mean realtek8211f-disable-eee-1000.patch from Nov 3. > Indeed. I won't submit anything for the vega since it has no effect. > Regards, > Andreas >