From mboxrd@z Thu Jan 1 00:00:00 1970 From: ingrassia@epigenesys.com (Emiliano Ingrassia) Date: Fri, 6 Oct 2017 10:10:50 +0200 Subject: [PATCH 0/4] meson8b-odroidc1: ethernet support In-Reply-To: <20171002195432.GX6726@otheros> References: <20171002195432.GX6726@otheros> Message-ID: <20171006081050.GA17290@ingrassia.epigenesys.com> To: linus-amlogic@lists.infradead.org List-Id: linus-amlogic.lists.infradead.org Hi Linus, thanks for the review! On Mon, Oct 02, 2017 at 09:54:32PM +0200, Linus L?ssing wrote: > Hi Emiliano, > > On Wed, Sep 27, 2017 at 12:39:32PM +0200, Emiliano Ingrassia wrote: > > This patchset enables ethernet support on Odroid-C1/C1+ boards, > > improving and extending the ethernet description in the relative > > device tree. > > > > In particular: > > - the mpll2 clock is enabled to avoid timeout during ethernet DMA reset; > > - the Meson8b ethernet controller description is fixed and extended; > > - the Odroid-C1/C1+ ethernet PHY description is added; > > - the dwmac sleep and timeout values used during DMA reset are fixed. > > > > This patchset has been tested on Odroid-C1+ board. > > I tried this patchset on an Odroid C1+ with a v4.14-rc3 kernel > plus these currently pending patches from this mailing list [0]. > > I'm observing the following issue: > > 1) The C1+ successfully gets an IPv6 address from stateless > autoconfiguration > 2) IPv6 pinging from the C1+ to another host X works > C1+ -> X => OK > 3) IPv6 pinging from host X to the C1+ does not work: > X -> C1+ => not OK > 4) While ping'ing from X->C1+ after some seconds / a minute a > parallel C1+->X ping stops working too. After stopping the > X->C1+ ping the C1+->X ping quickly recovers within about > one or two seconds. > > I tried adding the "eee-broken-1000t;" parameter to > meson8b-odroidc1.dts similar to meson-gxbb-odroidc2.dts. Did not > seem to help though. > > Still, this behaviour very much feels like a powersaving issue. > And/or an incomplete setup of the random MAC address on the C1+, > for instance a missing notification of the MAC address for the > ethernet chip, leaving it unable to wake up / create interrupts > for incoming frames with the destination of the C1+. > > The C1+ and host X are both connected via cable to the same switch > of a router. > > Can you reproduce this issue on your side, Emiliano? > Ok. Actually I'm working on the dwmac-meson8b glue layer to obtain a functional ethernet MAC. Infact, as stated by Martin in [0], the correct PRG_ETHERNET_ADDR0 register address should be 0xc1108140. I'll try your tests as soon as I find the correct solution. > Regards, Linus > Regards, Emiliano [0] http://lists.infradead.org/pipermail/linux-amlogic/2017-September/004810.html > [0]: > > e008445 ARM: multi_v7_defconfig: Add Amlogic Meson MMC support > a315380 ARM: dts: meson8b: Add MMC nodes > 90a73b6 mmc: meson-mx-sdio: Add a driver for the Amlogic Meson8 and Meson8b SoCs > 2f0b808 dt-bindings: mmc: Document the Amlogic Meson8 and Meson8b SDIO bindings > 31247f2 net: stmmac: fixing DMA reset sleep and timeout values > 1f43db3 ARM: dts: meson8b-odroidc1: enabling ethernet support > 8ecc9d3 ARM: dts: meson8b: extending ethernet controller description > be4a137 clk: meson8b: keep mpll2 clock enabled > 54b818c ARM: dts: meson8b: add reserved memory zone to fix silent freezes > 113819d ARM: dts: meson: enabling the USB Host controller on Odroid-C1/C1+ board > 7240c6d ARM: dts: meson: fixing USB support on Meson8b > a9153ba ARM: dts: meson8b: add support for booting the secondary CPU cores > 8529b23 ARM: dts: meson8: add support for booting the secondary CPU cores > fd053c9 ARM: meson: Add SMP bringup code for Meson8 and Meson8b > d6b74cd ARM: smp_scu: allow the platform code to read the SCU CPU status > 822a0d8 ARM: smp_scu: add a helper for powering on a specific CPU > 7da8338 dt-bindings: Amlogic: Add Meson8 and Meson8b SMP related documentation