bodhi Wrote: ------------------------------------------------------- > wacke, > > You are running OpenWrt, so I don't know what > caused that error in the dmesg. > That error was the resault of devmem 0xd0008100, and now I can get the resault: root@OpenWrt:/# devmem 0xf1008100 0x00000001 root@OpenWrt:/# devmem 0xf1008104 0x12086300 root@OpenWrt:/# devmem 0xf1by wacke - Debian
Hi bodhi, Use 0xf1 address got the resualt: root@OpenWrt:/# devmem 0xf1008100 0x00000001 root@OpenWrt:/# devmem 0xf1008104 0x12086300 root@OpenWrt:/# devmem 0xf1020200 0x01000002 But seems still not work properly.by wacke - Debian
Hi, It did not work: root@OpenWrt:/# dmesg [ 0.000000] Booting Linux on physical CPU 0x0 [ 0.000000] Linux version 5.9.8 (Wacke@HOME-Server) (arm-openwrt-linux-muslgnueabi-gcc (OpenWrt GCC 10.2.0 r14945-47e089e30e) 10.2.0, GNU ld (GNU Binutils) 2.35.1) #0 SMP Sun Nov 10 [ 0.000000] CPU: ARMv7 Processor [561f5811] revision 1 (ARMv7), cr=10c5387d [ 0.000000] CPU: PIPT / VIPT noby wacke - Debian
Hi bodhi, My box got this: root@OpenWrt:/# devmem 0xd0008100 [ 46.565761] 8<--- cut here --- [ 46.568848] Unhandled fault: external abort on non-linefetch (0x1018) at 0xb6f31100 [ 46.576541] pgd = cdebe8d4 [ 46.579259] *pgd=2d01d831, *pte=d0008383, *ppte=d0008a33 Bus error Device tree: // SPDX-License-Identifier: (GPL-2.0+ OR MIT) /* * Device Tree file for Marveby wacke - Debian
Hi bodhi, Acctully, I've contacted to his team member, and was told that there won't be any progress anymore, so I'm on my own now. The last info I got from him, was the openwrt cc system, I decompiled the dtb from that system (kernel image): /dts-v1/; / { #address-cells = <0x1>; #size-cells = <0x1>; model = "RTNAS V3"; compatible = "marby wacke - uBoot
Hi bodhi, Unfortunately, there was no GPL source for my box. The oem system was a customized openwrt(AA) that build from marvell sdk. The bootlog see: https://forum.archive.openwrt.org/viewtopic.php?id=50173&p=17#p237049by wacke - uBoot
And the boot log: [ 0.000000] Booting Linux on physical CPU 0x0 [ 0.000000] Linux version 4.20.6-mvebu-tld-1 (root@lam-mint19) (gcc version 7.3.0 (Ubuntu/Linaro 7.3.0-27ubuntu1~18.04)) #2 SMP PREEMPT Tue Feb 5 16:14:01 PST 2019 [ 0.0by wacke - uBoot
Hi bodhi, I've booted up my box with this dts: // SPDX-License-Identifier: (GPL-2.0+ OR MIT) /* * Device Tree file for Marvell Armada 370 Reference Design board * (RD-88F6710-A1) * * Copied from arch/arm/boot/dts/armada-370-db.dts * * Copyright (C) 2013 Florian Fainelli <florian@openwrt.org> * * Note: this Device Tree assumes that the bootloader has remapped theby wacke - uBoot
Hi bodhi, I was just copy the dts from armada-370-rd.dts, I'll try your idea.by wacke - uBoot
Hi bodhi, This is my box current DTS file (modified for kernel 4.19 above): // SPDX-License-Identifier: (GPL-2.0+ OR MIT) /* * Device Tree file for Marvell Armada 370 Reference Design board * (RD-88F6710-A1) * * Copied from arch/arm/boot/dts/armada-370-db.dts * * Copyright (C) 2013 Florian Fainelli <florian@openwrt.org> * * Note: this Device Tree assumes that the bootby wacke - uBoot
Hi bodhi, I don't know why box get the low perfomance and I've tried every way to modify the device tree, but got no help....So maybe I should give up....by wacke - uBoot
Hi bodhi, Did the EA4500 using the kernel DSA switch driver or using the mvsw61xx driver?by wacke - uBoot
Hi bodhi, After a lot of tests, I think it's still the HW IO problem, because my box don't have phy chip at all, the 8e6171r switch is link to the cpu directly.by wacke - uBoot
Hi bodhi, I've tried the turris omnia's switch part, and also tested with debian kernel 4.20.6, but nothing helped. And I found that the cpu usage was very high when the iperf3 was running....by wacke - uBoot
bodhi, Thanks. I'll try.by wacke - uBoot
Hi bodhi, So is there anyway to get the switch working properly?by wacke - uBoot
bodhi, nope. The dsa node was totally deprecated, it's in mdio node: &mdio { pinctrl-0 = <&mdio_pins>; pinctrl-names = "default"; switch0@0 { compatible = "marvell,mv88e6085"; #address-cells = <1>; #size-cells = <0>; reg = <0x0>;by wacke - uBoot
Hi bodhi, Currently my box runing openwrt with kernel 4.14.105, and if use base address 10, the kernel won't recognize the switch at all.by wacke - uBoot
Hi bodhi, My box recognize it with 6171: mv88e6085 f1072004.mdio-mii:00: switch 0x1710 detected: Marvell 88E6171, revision 2by wacke - uBoot
Hi bodhi, The dsa node was deprecated, and the current binding is mdio node: &mdio { pinctrl-0 = <&mdio_pins>; pinctrl-names = "default"; switch0@0 { compatible = "marvell,mv88e6085"; #address-cells = <1>; #size-cells = <0>; reg = <0x0>; ports { #address-cells = <1>; #size-cells = <0>; porby wacke - uBoot
Hi bodhi, So need rewrite the DTS? My box used the 88e6171r switch, and tested that base address <0x10> doesn't work. /* * Device Tree file for Marvell Armada 370 Reference Design board * (RD-88F6710-A1) * * Copied from arch/arm/boot/dts/armada-370-db.dts * * Copyright (C) 2013 Florian Fainelli <florian@openwrt.org> * * This file is dual-licensed: you canby wacke - uBoot
Hi bodhi, The iperf3 was not support bidirectional mode, so I just retested with bidirectional manually. My box to my router: root@debian:~# iperf3 -c 192.168.1.1 Connecting to host 192.168.1.1, port 5201 [ 4] local 192.168.1.238 port 40430 connected to 192.168.1.1 port 5201 [ ID] Interval Transfer Bandwidth Retr Cwnd [ 4] 0.00-1.01 sec 43.8 MBytes 363 Mbiby wacke - uBoot
Hi bodhi, Test of PC direct to router (armada 385, turris omnia): Wacke@HOME-Server:~> iperf3 -c 192.168.1.1 Connecting to host 192.168.1.1, port 5201 [ 5] local 192.168.1.2 port 45628 connected to 192.168.1.1 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 113 MBytes 946 Mbits/sec 0 474 KBytes [ 5] 1.00-2.00 seby wacke - uBoot
Hi bodhi, I used the iperf3 to test the IO performance.My box ip is 192.168.1.238. My box to router (armada 385,Turris Omnia) root@debian:~# iperf3 -c 192.168.1.1 Connecting to host 192.168.1.1, port 5201 [ 4] local 192.168.1.238 port 50948 connected to 192.168.1.1 port 5201 [ ID] Interval Transfer Bandwidth Retr Cwnd [ 4] 0.00-1.02 sec 45.0 MBytes 370 Mbiby wacke - uBoot
bodhi Wrote: ------------------------------------------------------- > wacke, > > So you should continue using the > 4.16.7-mvebu-tld-1 or which ever you are on, or > upgrade to the newer kernel 4.20.6-mvebu-tld-1 . OK. Does the kernel 4.20.6-mvebu-tld-1 include the HW I/O patch for armada 370?by wacke - uBoot
Hi bodhi, I'll try that kernel later,and my box uboot version seems is 2009.08....by wacke - uBoot
Hi bodhi, So my box also can use this kernel?by wacke - uBoot
Hi bodhi, Any update about your Armada 370 board?by wacke - uBoot
bodhi Wrote: ------------------------------------------------------- > wacke, > > Booted up my Armada 370 board. And having problem > right away :) it was previously own and root > credential seems to have been changed (not > root/nosoup4u anymore). > Hi bodhi, Great news! And the uboot for armada 370 will be the next step?by wacke - uBoot
Hi bodhi, Did you boot up your a370 board?by wacke - uBoot