Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add DDR4 SPD EEPROM data parser #204

Merged
merged 3 commits into from
Jun 4, 2020

Conversation

jedrzejboczar
Copy link
Collaborator

@jedrzejboczar jedrzejboczar commented Jun 2, 2020

This adds a parser for data from DDR4 SPD EEPROM. It uses tXX_L timings as we do not support tXX_S yet. It also checks page size to determine minimal tFAW in clocks.

I've tested this on ZCU104 with MTA4ATF51264HZ and the timings taken from SPD are correct (I also verified SPD data with reference on Micron's website), but when I was using the timings taken from SPD the module failed during leveling.

I checked that this is not due to the generation process, because changing the timings in the module definition gave the same results. I found that the only timings that differed enough to result in different timing_settings were tFAW (5 -> 7) and tCCD (1 -> 2). I wonder why this would result in leveling failing, when the timings increased, not decreased. Could this somehow influence leveling process?

I added SPD parsing tests in test_modules.py but one is being skipped for now. Maybe we can merge this and I can create an issue with information about this problem?

@enjoy-digital
Copy link
Owner

Thanks @jedrzejboczar, i will have a closer look, but just for info, it seems there is a regression on the ZCU104: #200 I will look at this, but this could explain your issue. Have you tested with the default SDRAMModule?.

@jedrzejboczar
Copy link
Collaborator Author

@enjoy-digital I was already testing with the change described in #200 that seems to fix the problem. Leveling stopped working only after changing the timings.

@enjoy-digital
Copy link
Owner

Thanks @jedrzejboczar, the parser support itself seems fine, so we can indeed merge it. For the issue with on hardware, can you create an issue for it as you were suggesting? I could have a closer look and could also do some tests on other boards. The leveling is slow process so tFAW and tCCD should not be impacting it but we will need to investigate.

@enjoy-digital enjoy-digital merged commit c4c8803 into enjoy-digital:master Jun 4, 2020
@jedrzejboczar jedrzejboczar deleted the jboc/spd-read branch June 5, 2020 14:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants