Honeycomb LX2160 - First bootup RAM training problems

NXP LX2160
1

Hi,

I just assembled everything in the rack and tried to boot up. But I am facing some problems.

I guess thats because of RAM training.

NOTICE:  BL2: v2.6(release):lf-5.15.71-2.2.0-14-g38bb9e491
NOTICE:  BL2: Built : 12:19:39, Nov  7 2024
NOTICE:  UDIMM CT32G4SFD832A.C16F
ERROR:   Execution FW failed (error code -5)
ERROR:   Calculating DDR PHY registers failed.

NOTICE:  32 GB DDR4, 64-bit, CL=17, ECC off, CS0+CS1
NOTICE:  BL2: v2.6(release):lf-5.15.71-2.2.0-14-g38bb9e491
NOTICE:  BL2: Built : 12:19:39, Nov  7 2024
NOTICE:  UDIMM CT32G4SFD832A.C16F
ERROR:   Execution FW failed (error code -5)
ERROR:   Calculating DDR PHY registers failed.

Here is what I got in Console. And that’s it. I saw similar problem 2 years ago here:

Does it mean it’s not solved?

I just tried smalled RAM I had arround but similar problem:

NOTICE:  BL2: v2.6(release):lf-5.15.71-2.2.0-14-g38bb9e491
NOTICE:  BL2: Built : 12:19:39, Nov  7 2024
NOTICE:  UDIMM AD4B320038G22-BSSC
ERROR:   Execution FW failed (error code -5)
ERROR:   Calculating DDR PHY registers failed.
ERROR:   Found training error(s): 0x100
ERROR:   Error: Waiting for D_INIT timeout.
ERROR:   Writing DDR register(s) failed
ERROR:   Programing DDRC error
ERROR:   DDR init failed.
ERROR:   Asserting as the DDR is not initialized yet.

One is 32G Crucial the other one is 8GB Synology from NAS

Edit:

Here is also similar topic and some newer image worked.

2

I guess the timing data could be looked up easily in datasheet and this could be supported as well. Or should I return it (10 days left)? @jnettlet sorry for pinging if its agains forum rules

Edit: I also tried this older image same error:

3

So I’ve tried third set of RAM sticks. This time Kingston Hynix 32G (non-ECC):

NOTICE:  BL2: v2.6(release):lf-5.15.71-2.2.0-14-g38bb9e491
NOTICE:  BL2: Built : 12:19:39, Nov  7 2024
NOTICE:  UDIMM KF3200C20S4/32GX  
ERROR:   Execution FW failed (error code -5)
ERROR:   Calculating DDR PHY registers failed.

NOTICE:  32 GB DDR4, 64-bit, CL=16, ECC off, CS0+CS1

Same problem. Its a biy anoyying that my old Thinkpad from more than 10 years ago doesn’t care about what RAM I put there and simply runs even when overclocked. But brand new server is failing at third attempt.

Is there anything that may be done? Possibly about this particular model as I have 2 and will prefer to use them.

Edit:
Ok I take back anoyment cuz it doesnt lead to anything productive. But could someone from the SolidRun team take a look? I am willing to build and test images.

4

Please test booting an image from my older UEFI builds. SolidRun Images They include patches to the SPD parsing and calibration process. I am pretty sure that the memory you have is not CL=16, which means the settings are trying to overclock your memory which is why it is failing to initialize.

But I am more concerned with the “Execution FW failed” However test with an EFI firmware image first to see if the memory initializes and then we will work on the NXP BSP based images to see why they are failing.

1 Like
5

Ok this image booted into UEFI:
https://solid-run-images.sos-de-fra-1.exo.io/LX2k/lx2160a_uefi/lx2160acex7_2000_700_3200_8_5_2_sd_ee5c233.img.xz

I need to work now so cannot further test. But after work I will try to write UEFI to SPI and boot from there. But I guess for that I will need a U-Boot image that works. According to your instructions here: