Xilinx / open-nic

AMD OpenNIC Project Overview
Apache License 2.0
222 stars 37 forks source link

RX not aligned on U50 #14

Open yrpang opened 1 year ago

yrpang commented 1 year ago

Environments

What did i do?

I generate bitstream and loaded it onto an Alveo U50 and reboot the machine. After that I try to insmod onic.ko.

What happend?

After reboot I can see the 01:00.0 Memory controller: Xilinx Corporation Device 903f on lspci result. But when I try to insmod I got CMAC 0 RX not aligned after waiting in the dmesg log and I cannot ping another machine it connected to.

The full log is following:

[  213.858893] OpenNIC Linux Kernel Driver 0.21
[  213.859161] onic 0000:01:00.0 onic1s0f0 (uninitialized): Set MAC address to 00:0a:35:9d:56:f1
[  213.859169] onic 0000:01:00.0: device is a master PF
[  213.859380] onic 0000:01:00.0: Allocated 8 queue vectors
[  213.859440] onic 0000:01:00.0: CMAC 0 RX_STATUS_REG value: 0x000000C0
[  218.832614] onic 0000:01:00.0: CMAC 0 RX_STATUS_REG value: 0x000000C0
[  218.832620] onic 0000:01:00.0: CMAC 0 RX not aligned after waiting
[  218.832627] onic 0000:01:00.0: Number of CMAC instances = 1
[  218.832689] onic 0000:01:00.0: Setup IRQ vector 146 with name onic1s0f0-0
[  218.832701] onic 0000:01:00.0: Setup IRQ vector 147 with name onic1s0f0-1
[  218.832713] onic 0000:01:00.0: Setup IRQ vector 148 with name onic1s0f0-2
[  218.832725] onic 0000:01:00.0: Setup IRQ vector 149 with name onic1s0f0-3
[  218.832736] onic 0000:01:00.0: Setup IRQ vector 150 with name onic1s0f0-4
[  218.832746] onic 0000:01:00.0: Setup IRQ vector 151 with name onic1s0f0-5
[  218.832758] onic 0000:01:00.0: Setup IRQ vector 152 with name onic1s0f0-6
[  218.832768] onic 0000:01:00.0: Setup IRQ vector 153 with name onic1s0f0-7
[  218.838351] onic 0000:01:00.0 enp1s0: renamed from onic1s0f0
[  218.906384] onic 0000:01:00.0 enp1s0: ethtool: onic_get_link port: 0   carrier ok: 1 -- rx status ok: 0
[  218.910071] onic 0000:01:00.0 enp1s0: ethtool: onic_get_link port: 0   carrier ok: 1 -- rx status ok: 0
[  219.840626] IPv6: ADDRCONF(NETDEV_CHANGE): enp1s0: link becomes ready
[  220.041218] userif-3: sent link down event.
[  220.041225] userif-3: sent link up event.

I checked the CMAC document and learned that the RX_STATUS_REG 0xC0 means "rx_internal_local_fault". Could you please give me some suggestions for troubleshooting or some ideas of potential problems?