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

Also fuzz OCLK and OCLKB inputs to the ISERDESE2. Fix typo. #1830

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

bl0x
Copy link
Contributor

@bl0x bl0x commented Jan 24, 2022

I have not run the fuzzer, but am mainly asking if this is a sane change, since this is the first time I'm looking into this code.

I have hopes that this may fix a problem I've encountered here:

f4pga/f4pga-arch-defs#2367 (comment)

while adding support for the OCLK and OCLKB inputs reqiured for the OVERSAMPLE mode of ISERDESE2 in symbiflow-arch-defs.

@acomodi
Copy link
Contributor

acomodi commented Jan 24, 2022

Hi @bl0x, thanks for this PR. When dealing with the ISERDES fuzzer we actually noticed that setting the OCLK and OCLKB inputs, the PIPs responsible for those were not actually producing any unknown bit.

I think that the best way of verifying that we already have all the information is using the minitest approach:

  1. Isolate the ISERDES from your design with the exact features you want to use
  2. Generate a bitstream with Vivado
  3. Use bit2fasm with the --verbose option enabled
  4. Check the number of unknown bits (if any) in the resulting FASM file

@bl0x
Copy link
Contributor Author

bl0x commented Jan 24, 2022

How do I notice unknown bits in the resulting FASM file?
bit2fasm prints everything to stdout by default.

In order to properly generate a bitstream with Vivado I had to make a few changes to the code, namely adding proper clock sources and making connections to the iserdes.
I also needed to set IOSTANDARD attributes to the I/O pins used (though I doubt that that is relevant here).

With these changes the bit2fasm output did not show any obvious unknown bits.
Also, I could generate a bitstream from the modified (fixed) code using the Symbiflow toolchain.

It looks to me, as if this PR is then not needed.

@acomodi
Copy link
Contributor

acomodi commented Jan 24, 2022

In that case all the bits should actually be there

With these changes the bit2fasm output did not show any obvious unknown bits.

Just to double-check, did you run with the --verbose flag? If yes and no unknown bits are reported, than I think that this is another confirmation that all bits are documented.

@bl0x
Copy link
Contributor Author

bl0x commented Jan 24, 2022

Yes, I did run with the --verbose flag. How would unknown bits be reported?
Like so?

Found unknown bits: XX, YY

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants