Re: NanoPC-T6 LTS (rk3588) with edk2(UEFI)

[mlelstv%serpens.de@localhost (Michael van Elst)]

mochid%netside.co.jp@localhost (MOCHIDA Shuji) writes:

Hi,

> dmesg:
>  https://dmesgd.nycbug.org/dmesgd?do=view&id=8575

> friendlyelec NanoPC-T6 LTS
>  https://www.friendlyelec.com/index.php?route=product/product&product_id=292

Yes. This is similar to my NanoPi R6S.

I built UEFI myself from a checkout in May and changed a few things:
- PLL configuration in Emmc.asl
- sdhci capabilities in Sdhc.asl
- added a RNG driver to generate an entropy seed value.


>  = EDK2 is on eMM
>  = bootaa64.efi in FAT fs and NetBSD FFSs are on M.2 NVMe
>  * EDK2 can be loaded from eMM or microSD
>  * EDK2 can load bootaa64.efi from NVMe (or USB mass strage)
>  * bootaa64.efi can load NetBSD kernel from NVMe (or USB mass strage)
>  * NetBSD kernel can use NVMe (or USB mass strage) as root file system
>  * and, NetBSD 10.1/aarch64 cannot use eMM nor microSD as root (yet).

No NVMe here, so the choice is between eMM, SDcard and USB storage.
I added a small delay to sdhc(4) which made eMM work for NetBSD
(but I don't understand why this made a difference).

[     1.000004] sdhc0 at acpi0 (SDC3, RKCP0D40-3): mem 0xfe2e0000-0xfe2effff irq 237
[     1.000004] sdhc0: SDHC 4.2, rev 0, caps <226dc881/08000007>, 32-bit ADMA2, 200000 kHz, HS SDR50 DDR50 SDR104 HS200 3.0V, re-tuning mode 1, 1024 byte blocks
[     1.000004] sdmmc0 at sdhc0 slot 0
...
[     1.439014] ld0 at sdmmc0: <0xd6:0x0103:A3A551:0x00:0xc7de08d1:0x000>
[     1.439014] ld0: 29600 MB, 7517 cyl, 128 head, 63 sec, 512 bytes/sect x 60620800 sectors
[     1.439014] ld0: GPT GUID: 25a6920b-1ce1-4a88-9fa7-72ecbf641dc7



The microSD card slot also works, but EDK2 configures it to minimal speed:

[     1.000004] dwcmmc0 at acpi0 (SDHC, RKCPFE2C-0): mem 0xfe2c0000-0xfe2c3fff irq 235
[     1.239009] sdmmc1 at dwcmmc0
...
[ 827098.588202] sdmmc1: SD card status: 4-bit, C10, U3, V30, A2, Host/Card self-maintenance
[ 827098.588202] sdmmc1: cache enabled
[ 827098.588202] ld1 at sdmmc1: <0x03:0x5344:SN128:0x86:0x9af3bda2:0x188>
[ 827098.588202] ld1: 119 GB, 15545 cyl, 255 head, 63 sec, 512 bytes/sect x 249737216 sectors
[ 827098.598202] ld1: 4-bit width, High-Speed/SDR25, 50.000 MHz


> I use urtwn(4) for Wi-Fi. stable.
> on board rge0, rge1 at pci shows (six-zeros)
>    rge0: Ethernet address 00:00:00:00:00:00

Yes. The MAC addresses are neither set by EDK2 nor the NetBSD kernel.

[     1.000004] acpipchb0 at acpi0 (PCI3, PNP0A08-3): PCI Express Host Bridge
[     1.000004] pci0 at acpipchb0 bus 1
[     1.000004] pci0: i/o space, memory space enabled, rd/line, rd/mult, wr/inv ok
[     1.000004] rge0 at pci0 dev 0 function 0: vendor 10ec product 8125 (rev. 0x05)
[     1.000004] rge0: interrupting at irq 277
[     1.000004] rge0: HW rev. B
[     1.000004] rge0: Ethernet address 00:00:00:00:00:00
[     1.000004] acpipchb1 at acpi0 (PCI4, PNP0A08-4): PCI Express Host Bridge
[     1.000004] pci1 at acpipchb1 bus 1
[     1.000004] pci1: i/o space, memory space enabled, rd/line, rd/mult, wr/inv ok
[     1.000004] rge1 at pci1 dev 0 function 0: vendor 10ec product 8125 (rev. 0x05)
[     1.000004] rge1: interrupting at irq 282
[     1.000004] rge1: HW rev. B
[     1.000004] rge1: Ethernet address 00:00:00:00:00:00

This is different from the eqos0 interface which is part of the SoC
and also supported by EDK2.

[     1.000004] eqos0 at acpi0 (MAC1, RKCP6543-61): mem 0xfe1c0000-0xfe1cffff irq 266,265
[     1.000004] eqos0: DesignWare EQOS ver 0x51 (0x30)
[     1.000004] eqos0: hw features 1a1173f7 111e01e8 11041041 0c370031
[     1.000004] eqos0: using 32-bit DMA
[     1.000004] eqos0: Ethernet address d6:de:d3:52:de:dc


So far there is no way for NetBSD to access the EEPROM where u-boot
is located. While EDK2 does expose GPIO and I2C interfaces, there
is no SPI driver yet.