Agilex 5 - HPS first - u-boot stuck if rebooting

[MM-ATH]

Hello,

I have problem after reboot of my system (reboot in linux console or press HPSrst button on AXE5-Eagle devboard), or if I fit fpgacore.rbf from u-boot twice, than boot from emmc stuck on:

Hit any key to stop autoboot:  0
2568192 bytes read in 124 ms (19.8 MiB/s)
...FPGA reconfiguration OK!
switch to partitions #0, OK
mmc0 is current device
Scanning mmc 0:1...
Found U-Boot script /boot.scr.uimg
2751 bytes read in 13 ms (206.1 KiB/s)
## Executing script at 81000000
crc32+ Trying to boot Linux from device mmc0
Found kernel in mmc0

First time (power-on or after press nconfig button on devboard AXE5-Eagle) it boots ok - succesfully continues to OS with:

crc32+ Trying to boot Linux from device mmc0
Found kernel in mmc0
5428330 bytes read in 245 ms (21.1 MiB/s)
## Loading kernel from FIT Image at 82000000 ...
   Using 'board-0' configuration
   Verifying Hash Integrity ... OK
   Trying 'kernel' kernel subimage
...

Can somebody help me to overcome this problem?

Thanks.

Martin

[MM-ATH]

Thank you, Jingyang.

Behaviour is same. After I removed using of bridges, I can boot linux even if FPGA is not programmed (or programmed more times). If I programm FPGA core more times, CPU very probably stucks when my application in linux (or devmem) tries to access HPS2FPGA bridge. I had about two situations when everything went OK after 5 tries of reprogramming in u-boot.

Hopefully I will be able to test the design on other AXE5-Eagle kit (we need two kits to work in parallel) and in the next two weeks it will be available.

We are now doing some experiments with Quartus 25.1, but it seems that more things must be changed. We can compile design OK, but if we use Q25.1 to generate (jic/rpd) files then we are not able to programm core.rbf even from u-boot and linux boot stucks with some errors. If we use Q24.3.1 for generating files compiled by Q25.1 it seems to work ok. So we will wait with migration, since Arrow provide GHRD, u-boot and linux examples compatible with Q25.1.

Regards,

Martin

[JingyangTeh]

Hi Martin

From what I have gathered. There is no workaround available as it is a silicone bug.

I assume if you were to use 25.1 it would be the same behavior too, but it is worth trying.

This issue will be fixed in the B0 silicone of the Agilex5.

Regards

Jingyang, Teh

[JingyangTeh]

Hi

Do you still have any follow up on this issue?

Regards

Jingyang, Teh

[JingyangTeh]

Hi

As we do not receive any response from you on the previous question/reply/answer that we have provided. Please login to ‘https://supporttickets.intel.com/s/?language=en_US’, view details of the desire request, and post a feed/response within the next 15 days to allow me to continue to support you. After 15 days, this thread will be transitioned to community support. The community users will be able to help you on your follow-up questions.

Regards

Jingyang,Teh