Re: [RFC PATCH 2/3] RISC-V: Initial DTS for Microchip ICICLE board

From: Ben Dooks
Date: Fri Oct 30 2020 - 17:45:28 EST


On 30/10/2020 07:11, Atish Patra wrote:
On Thu, Oct 29, 2020 at 3:24 AM Ben Dooks <ben.dooks@xxxxxxxxxxxxxxx> wrote:

On 28/10/2020 23:27, Atish Patra wrote:
Add initial DTS for Microchip ICICLE board having only
essential devcies (clocks, sdhci, ethernet, serial, etc).

Signed-off-by: Atish Patra <atish.patra@xxxxxxx>
---
arch/riscv/boot/dts/Makefile | 1 +
arch/riscv/boot/dts/microchip/Makefile | 2 +
.../microchip/microchip-icicle-kit-a000.dts | 313 ++++++++++++++++++
3 files changed, 316 insertions(+)
create mode 100644 arch/riscv/boot/dts/microchip/Makefile
create mode 100644 arch/riscv/boot/dts/microchip/microchip-icicle-kit-a000.dts

diff --git a/arch/riscv/boot/dts/Makefile b/arch/riscv/boot/dts/Makefile
index ca1f8cbd78c0..3ea94ea0a18a 100644
--- a/arch/riscv/boot/dts/Makefile
+++ b/arch/riscv/boot/dts/Makefile
@@ -1,5 +1,6 @@
# SPDX-License-Identifier: GPL-2.0
subdir-y += sifive
subdir-y += kendryte
+subdir-y += microchip

obj-$(CONFIG_BUILTIN_DTB) := $(addsuffix /, $(subdir-y))
diff --git a/arch/riscv/boot/dts/microchip/Makefile b/arch/riscv/boot/dts/microchip/Makefile
new file mode 100644
index 000000000000..55ad77521304
--- /dev/null
+++ b/arch/riscv/boot/dts/microchip/Makefile
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0
+dtb-$(CONFIG_SOC_MICROCHIP_POLARFIRE) += microchip-icicle-kit-a000.dtb
diff --git a/arch/riscv/boot/dts/microchip/microchip-icicle-kit-a000.dts b/arch/riscv/boot/dts/microchip/microchip-icicle-kit-a000.dts
new file mode 100644
index 000000000000..5848920af55c
--- /dev/null
+++ b/arch/riscv/boot/dts/microchip/microchip-icicle-kit-a000.dts
@@ -0,0 +1,313 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (c) 2020 Microchip Technology Inc */
+
+/dts-v1/;
+
+/* Clock frequency (in Hz) of the rtcclk */
+#define RTCCLK_FREQ 1000000
+
+/ {
+ #address-cells = <2>;
+ #size-cells = <2>;
+ model = "Microchip PolarFire-SoC";
+ compatible = "microchip,polarfire-soc";
+
+ chosen {
+ stdout-path = &serial0;
+ };
+
+ cpus {
+ #address-cells = <1>;
+ #size-cells = <0>;
+ timebase-frequency = <RTCCLK_FREQ>;
+
+ cpu@0 {
+ clock-frequency = <0>;
+ compatible = "sifive,rocket0", "riscv";
+ device_type = "cpu";
+ i-cache-block-size = <64>;
+ i-cache-sets = <128>;
+ i-cache-size = <16384>;
+ reg = <0>;
+ riscv,isa = "rv64imac";
+ status = "disabled";
+
+ cpu0_intc: interrupt-controller {
+ #interrupt-cells = <1>;
+ compatible = "riscv,cpu-intc";
+ interrupt-controller;
+ };
+ };
+
+ cpu@1 {
+ clock-frequency = <0>;
+ compatible = "sifive,rocket0", "riscv";
+ d-cache-block-size = <64>;
+ d-cache-sets = <64>;
+ d-cache-size = <32768>;
+ d-tlb-sets = <1>;
+ d-tlb-size = <32>;
+ device_type = "cpu";
+ i-cache-block-size = <64>;
+ i-cache-sets = <64>;
+ i-cache-size = <32768>;
+ i-tlb-sets = <1>;
+ i-tlb-size = <32>;
+ mmu-type = "riscv,sv39";
+ reg = <1>;
+ riscv,isa = "rv64imafdc";
+ tlb-split;
+ status = "okay";
+
+ cpu1_intc: interrupt-controller {
+ #interrupt-cells = <1>;
+ compatible = "riscv,cpu-intc";
+ interrupt-controller;
+ };
+ };
+
+ cpu@2 {
+ clock-frequency = <0>;
+ compatible = "sifive,rocket0", "riscv";
+ d-cache-block-size = <64>;
+ d-cache-sets = <64>;
+ d-cache-size = <32768>;
+ d-tlb-sets = <1>;
+ d-tlb-size = <32>;
+ device_type = "cpu";
+ i-cache-block-size = <64>;
+ i-cache-sets = <64>;
+ i-cache-size = <32768>;
+ i-tlb-sets = <1>;
+ i-tlb-size = <32>;
+ mmu-type = "riscv,sv39";
+ reg = <2>;
+ riscv,isa = "rv64imafdc";
+ tlb-split;
+ status = "okay";
+
+ cpu2_intc: interrupt-controller {
+ #interrupt-cells = <1>;
+ compatible = "riscv,cpu-intc";
+ interrupt-controller;
+ };
+ };
+
+ cpu@3 {
+ clock-frequency = <0>;
+ compatible = "sifive,rocket0", "riscv";
+ d-cache-block-size = <64>;
+ d-cache-sets = <64>;
+ d-cache-size = <32768>;
+ d-tlb-sets = <1>;
+ d-tlb-size = <32>;
+ device_type = "cpu";
+ i-cache-block-size = <64>;
+ i-cache-sets = <64>;
+ i-cache-size = <32768>;
+ i-tlb-sets = <1>;
+ i-tlb-size = <32>;
+ mmu-type = "riscv,sv39";
+ reg = <3>;
+ riscv,isa = "rv64imafdc";
+ tlb-split;
+ status = "okay";
+
+ cpu3_intc: interrupt-controller {
+ #interrupt-cells = <1>;
+ compatible = "riscv,cpu-intc";
+ interrupt-controller;
+ };
+ };
+
+ cpu@4 {
+ clock-frequency = <0>;
+ compatible = "sifive,rocket0", "riscv";
+ d-cache-block-size = <64>;
+ d-cache-sets = <64>;
+ d-cache-size = <32768>;
+ d-tlb-sets = <1>;
+ d-tlb-size = <32>;
+ device_type = "cpu";
+ i-cache-block-size = <64>;
+ i-cache-sets = <64>;
+ i-cache-size = <32768>;
+ i-tlb-sets = <1>;
+ i-tlb-size = <32>;
+ mmu-type = "riscv,sv39";
+ reg = <4>;
+ riscv,isa = "rv64imafdc";
+ tlb-split;
+ status = "okay";
+ cpu4_intc: interrupt-controller {
+ #interrupt-cells = <1>;
+ compatible = "riscv,cpu-intc";
+ interrupt-controller;
+ };
+ };
+ };
+
+ memory@80000000 {
+ device_type = "memory";
+ reg = <0x0 0x80000000 0x0 0x40000000>;
+ clocks = <&clkcfg 26>;
+ };

U-boot doesn't seem to be updating this properly.

The board should have 2GiB, confirmed by looking at the device's
chip markings. We only see 1GiB memory. The 0x80000000 bus window
is only capable of dealing with 1GiB memory. The higher 64-bit one
can have 16GiB mapped.

Do we need a second node for the second GiB of memory?

We could just modify the reg size but to allow more memory. I tried
that for Linux but it didn't boot.
Probably, DDR init code in HSS only initialized 1GB of memory.

Yes, it is only looking at the low window which is 1GiB max.
If it used the upper window it would get the 16GiB.

I don't know how no-one noticed this issue before shipping a board
out with this. I have updated the firmware on my second board but
this only seems to currently fix a reboot issue with the eMMC.

+
+ soc {
+ #address-cells = <2>;
+ #size-cells = <2>;
+ compatible = "simple-bus";
+ ranges;
+
+ cache-controller@2010000 {
+ compatible = "sifive,fu540-c000-ccache", "cache";
+ cache-block-size = <64>;
+ cache-level = <2>;
+ cache-sets = <1024>;
+ cache-size = <2097152>;
+ cache-unified;
+ interrupt-parent = <&plic>;
+ interrupts = <1 2 3>;
+ reg = <0x0 0x2010000 0x0 0x1000>;
+ };
+
+ clint@2000000 {
+ compatible = "riscv,clint0";
+ reg = <0x0 0x2000000 0x0 0xC000>;
+ interrupts-extended = <&cpu0_intc 3 &cpu0_intc 7
+ &cpu1_intc 3 &cpu1_intc 7
+ &cpu2_intc 3 &cpu2_intc 7
+ &cpu3_intc 3 &cpu3_intc 7
+ &cpu4_intc 3 &cpu4_intc 7>;
+ };
+
+ plic: interrupt-controller@c000000 {
+ #interrupt-cells = <1>;
+ compatible = "sifive,plic-1.0.0";
+ reg = <0x0 0xc000000 0x0 0x4000000>;
+ riscv,ndev = <53>;
+ interrupt-controller;
+ interrupts-extended = <&cpu0_intc 11
+ &cpu1_intc 11 &cpu1_intc 9
+ &cpu2_intc 11 &cpu2_intc 9
+ &cpu3_intc 11 &cpu3_intc 9
+ &cpu4_intc 11 &cpu4_intc 9>;
+ };
+
+ dma@3000000 {
+ compatible = "sifive,fu540-c000-pdma";
+ reg = <0x0 0x3000000 0x0 0x8000>;
+ interrupt-parent = <&plic>;
+ interrupts = <23 24 25 26 27 28 29 30>;
+ #dma-cells = <1>;
+ };
+
+ refclk: refclk {
+ compatible = "fixed-clock";
+ #clock-cells = <0>;
+ clock-frequency = <600000000>;
+ clock-output-names = "msspllclk";
+ };
+
+ clkcfg: clkcfg@20002000 {
+ compatible = "microchip,pfsoc-clkcfg";
+ reg = <0x0 0x20002000 0x0 0x1000>;
+ reg-names = "mss_sysreg";
+ clocks = <&refclk>;
+ #clock-cells = <1>;
+ clock-output-names = "cpuclk", "axiclk", "ahbclk", "ENVMclk", "MAC0clk", "MAC1clk", "MMCclk", "TIMERclk", "MMUART0clk", "MMUART1clk", "MMUART2clk", "MMUART3clk", "MMUART4clk", "SPI0clk", "SPI1clk", "I2C0clk", "I2C1clk", "CAN0clk", "CAN1clk", "USBclk", "RESERVED", "RTCclk", "QSPIclk", "GPIO0clk", "GPIO1clk", "GPIO2clk", "DDRCclk", "FIC0clk", "FIC1clk", "FIC2clk", "FIC3clk", "ATHENAclk", "CFMclk";

Any chance of making this list multi-line, it is difficult to read as-is.


Yes. We can also get rid of a few names that are not used. I will fix it in v2.

+ };
+
+ serial0: serial@20000000 {
+ compatible = "ns16550a";
+ reg = <0x0 0x20000000 0x0 0x400>;
+ reg-io-width = <4>;
+ reg-shift = <2>;
+ interrupt-parent = <&plic>;
+ interrupts = <90>;
+ current-speed = <115200>;
+ clocks = <&clkcfg 8>;
+ status = "okay";
+ };
+
+ serial1: serial@20100000 {
+ compatible = "ns16550a";
+ reg = <0x0 0x20100000 0x0 0x400>;
+ reg-io-width = <4>;
+ reg-shift = <2>;
+ interrupt-parent = <&plic>;
+ interrupts = <91>;
+ current-speed = <115200>;
+ clocks = <&clkcfg 9>;
+ status = "okay";
+ };
+
+ serial2: serial@20102000 {
+ compatible = "ns16550a";
+ reg = <0x0 0x20102000 0x0 0x400>;
+ reg-io-width = <4>;
+ reg-shift = <2>;
+ interrupt-parent = <&plic>;
+ interrupts = <92>;
+ current-speed = <115200>;
+ clocks = <&clkcfg 10>;
+ status = "okay";
+ };
+
+ serial3: serial@20104000 {
+ compatible = "ns16550a";
+ reg = <0x0 0x20104000 0x0 0x400>;
+ reg-io-width = <4>;
+ reg-shift = <2>;
+ interrupt-parent = <&plic>;
+ interrupts = <93>;
+ current-speed = <115200>;
+ clocks = <&clkcfg 11>;
+ status = "okay";
+ };
+
+ sdcard: sdhc@20008000 {
+ compatible = "cdns,sd4hc";
+ reg = <0x0 0x20008000 0x0 0x1000>;
+ interrupt-parent = <&plic>;
+ interrupts = <88>;
+ pinctrl-names = "default";
+ clocks = <&clkcfg 6>;
+ bus-width = <4>;
+ disable-wp;
+ no-1-8-v;
+ cap-mmc-highspeed;
+ cap-sd-highspeed;
+ card-detect-delay = <200>;
+ sd-uhs-sdr12;
+ sd-uhs-sdr25;
+ sd-uhs-sdr50;
+ sd-uhs-sdr104;
+ max-frequency = <200000000>;
+ status = "okay";
+ };

Given eMMC is the default device, shouldn't that be default for the
device tree too? Even if not, having the emmc node here would be a
good thing as it is different to the SD node.


I tested this device tree with sdcard. That's why, I just picked the
sdcard node.
I am not sure if both eMMC & sdcard node can co-exist. The polar fire
soc github repo
seems to point that both of them have the same address and only 1 can be enabled
at that time. That may not be true now as the github repo has not been
updated in
couple of months.

@Cyril : Can we enable both eMMC & sdcard at the same time ?

I would put /both/ in but only enable the one in use for the moment.
Our boards are booting of eMMC as supplied, so this isn't going to work as well. The eMMC is 8bit wide, and thus is only delivering 11MB/sec
instead of 22MB/sec. This performance is still not great, but losing
half the data-rate is just not good.

+
+ emac1: ethernet@20112000 {
+ compatible = "cdns,macb";
+ reg = <0x0 0x20112000 0x0 0x2000>;
+ interrupt-parent = <&plic>;
+ interrupts = <70 71 72 73>;
+ mac-address = [56 34 12 00 FC 00];
+ phy-mode = "sgmii";
+ clocks = <&clkcfg 5>, <&clkcfg 2>;
+ clock-names = "pclk", "hclk";
+ #address-cells = <1>;
+ #size-cells = <0>;
+ phy1: ethernet-phy@9 {
+ reg = <9>;
+ ti,fifo-depth = <0x01>;
+ };
+ };

Aren't there two ethernet ports on the board?


Yes. I hadn't tested that out. I will test it and add the 2nd one as well.

Also, at the moment u-boot is not filling the MAC address parameter
in so we've got at two boards on the network with the same MAC until
we override it in the device tree for the second.


Looking at latest U-Boot patches, it seems it updates the mac address
from the serial number.

Ok, the one supplied on the boards does not. And the boards /both/ have
the same MAC address!

+
+ uio_axi_lsram@2030000000 {
+ compatible = "generic-uio";
+ reg = <0x20 0x30000000 0 0x80000000 >;
+ status = "okay";
+ };
+ };
+};



--
Ben Dooks http://www.codethink.co.uk/
Senior Engineer Codethink - Providing Genius

https://www.codethink.co.uk/privacy.html

_______________________________________________
linux-riscv mailing list
linux-riscv@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/linux-riscv





--
Ben Dooks http://www.codethink.co.uk/
Senior Engineer Codethink - Providing Genius

https://www.codethink.co.uk/privacy.html