Considering buying and question on the PCIe slot

[OwedDavin]

I think the time has come for ARM and chipmakers to use uefi on their chips since the u boot which is used is outdated now and as ARM is moving into laptop market, uefi might be a good use especially in driving pcie devices. With that being said, sbc makers should implement in their products the uefi for the good of its customers. Windows on ARM uses uefi, so as Linux must be. 

[ndb]

Hi guys,

I've just joined the forum because I am curious if anyone has made progress on getting a Nvidia card to work with their Rockpro64 (joe2gaan?)
I do not own a board yet but am strongly considering buying one if this works.

Might the "Generic display driver" and CUDA Toolkit from https://developer.nvidia.com/cuda-toolkit-65 (See Linux ARM tab) work with this board?

[tllim]

Hi guys,

I've just joined the forum because I am curious if anyone has made progress on getting a Nvidia card to work with their Rockpro64 (joe2gaan?)
I do not own a board yet but am strongly considering buying one if this works.

Might the "Generic display driver" and CUDA Toolkit from https://developer.nvidia.com/cuda-toolkit-65 (See Linux ARM tab) work with this board?

I don't think NVidia card can work on ROCKPro64, this due to graphic card needs large IO Memory range.

[jgalak]

I'm was hoping to use mine with my new PCIe based XTRX SDR (Software Defined Radio). But at first attempt the driver did not compile on arm64. So it seems like some debugging is required. But at least it's detected:

Code:

root@rockpro64:~# lspci
00:00.0 PCI bridge: Rockchip Inc. RK3399 PCI Express Root Port Device 0100
01:00.0 Memory controller: Xilinx Corporation Device 7012

Hey @zener, did you ever get this working?  I was looking at this board specifically to use with an Xtra, so hoping someone has blazed the trail

[zener]

I'm was hoping to use mine with my new PCIe based XTRX SDR (Software Defined Radio). But at first attempt the driver did not compile on arm64. So it seems like some debugging is required. But at least it's detected:

Code:

root@rockpro64:~# lspci
00:00.0 PCI bridge: Rockchip Inc. RK3399 PCI Express Root Port Device 0100
01:00.0 Memory controller: Xilinx Corporation Device 7012

Hey @zener, did you ever get this working?  I was looking at this board specifically to use with an Xtra, so hoping someone has blazed the trail

Sorry about the late reply.

No I could not get it working. The recent PCIe driver will load, the low level test application will not crash:

 

./test_xtrxll
16:08:23.219683 DEBUG:  xtrxllpciev0_discovery:264 [PCIE] pcie: Found `pcie:///dev/xtrx0`
16:08:23.220160 REGS:   internal_xtrxll_reg_in:156 [PCIE] PCI:/dev/xtrx0: Read  [001a] = 04000013
16:08:23.220212 INFO:   xtrxll_base_dev_init:1116 [CTRL] PCI:/dev/xtrx0: XTRX Rev4 (04000013)
16:08:23.220276 INFO:   xtrxllpciebase_dma_start:536 [BPCI] PCI:/dev/xtrx0: RX DMA STOP MIMO (BLK:0 TS:0); TX DMA STOP MIMO @0.0
16:08:23.220371 REGS:   internal_xtrxll_reg_out:147 [PCIE] PCI:/dev/xtrx0: Write [000d] = c0000000
16:08:23.220442 REGS:   internal_xtrxll_reg_out:147 [PCIE] PCI:/dev/xtrx0: Write [000d] = 42000000
16:08:23.220468 INFO:   xtrxllpciev0_open:360 [PCIE] PCI:/dev/xtrx0: Device `pcie:///dev/xtrx0` was opened
16:08:23.220494 REGS:   internal_xtrxll_reg_out:147 [PCIE] PCI:/dev/xtrx0: Write [0000] = 002fffff
16:08:23.220721 DEBUG:  xtrxllpciev0_lms7_spi_bulk:197 [PCIE] PCI:/dev/xtrx0: SPI[0/1] I:1
16:08:23.220796 REGS:   internal_xtrxll_reg_in:156 [PCIE] PCI:/dev/xtrx0: Read  [0000] = 00000000
16:08:23.220823 DEBUG:  xtrxllpciev0_lms7_spi_bulk:203 [PCIE] PCI:/dev/xtrx0: SPI[0/1] 002fffff => 00000000
Detected LMS #0: 00000000

But when I run the test application it will crash:

# ./test_xtrx -t -l2
Master: 32.000 MHz; RX rate: 4.000 MHz; TX rate: 0.000 MHz
RX tunned: 900000000.000000
RX bandwidth: 2000000.000000
RX LNA gain: 15.000000
RX PGA gain: 0.000000
RX TIA gain: 9.000000
RX CYCLES=1 SAMPLES=16384 SLICE=16384 (PARTS=1)
Bus error

In the kernel log I observe:

[  168.793490] xtrx: buf[30]=f1ee0000 [virt ffffff80098a1000] => f1ee0fff
[  168.793527] xtrx: buf[31]=f1ef0000 [virt ffffff80098b1000] => f1ef0fff
[  168.793744] Unhandled fault: implementation fault (unsupported exclusive) (0x92000035) at 0x0000007fae1e1008

[hmuller]

I think the time has come for ARM and chipmakers to use uefi on their chips since the u boot which is used is outdated now and as ARM is moving into laptop market, uefi might be a good use especially in driving pcie devices. With that being said, sbc makers should implement in their products the uefi for the good of its customers. Windows on ARM uses uefi, so as Linux must be. 

I was reading this post because of an interest in the pcie situation work the rk3399 SoCs.

I watched a presentation today from last October by some guy from ARM. The presentation was on just that topic. Do a search on "github ebbr arm" and you will find the specification they are leading work on. Good stuff.