Skip to content

Add asserts to br_cdc_bit_pulse and scrub some obsolete TODOs #653

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.

Sign up for GitHub

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

Open
mgottscho wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Add asserts to br_cdc_bit_pulse and scrub some obsolete TODOs #653

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
fa30f87
Scrub some obsolete TODOs
mgottscho Apr 25, 2025
bd52c59
Merge branch 'main' into mgottscho/todo-scrub
mgottscho May 17, 2025
e9bb8e6
Merge branch 'main' into mgottscho/todo-scrub
mgottscho Nov 12, 2025
4fcab54
Merge branch 'main' into mgottscho/todo-scrub
mgottscho Nov 12, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 14 additions & 3 deletions cdc/rtl/br_cdc_bit_pulse.sv
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,10 +56,18 @@ module br_cdc_bit_pulse #(
input logic dst_rst,
output logic dst_pulse
);
// Integration Assertions

//------------------------------------------
// Integration checks
//------------------------------------------
// Relying on checks in br_cdc_bit_toggle

`BR_ASSERT_CR_INTG(src_pulse_high_then_low_a, src_pulse |=> !src_pulse, src_clk, src_rst)
Copy link
Copy Markdown
Contributor

@zhemao-openai zhemao-openai Apr 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I didn't add these in my PR because I think this is too stringent. If the source clock is much slower than the destination clock, it would be perfectly fine to have two back to back pulses.

Copy link
Copy Markdown
Contributor Author

@mgottscho mgottscho Apr 28, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I disagree: the purpose of the module is to transfer pulses. If you're saying you don't have to pulse on the sender side then it seems ill-defined?

Also "two back to back pulses" means there must be a falling edge between, no?

`BR_COVER_CR_INTG(back_to_back_src_pulse_c, src_pulse ##2 src_pulse, src_clk, src_rst)
Copy link
Copy Markdown
Contributor

@zhemao-openai zhemao-openai Apr 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Conversely, if the source clock is much faster than the destination clock, it may not be safe to have pulses two cycles apart on the source clock.

Copy link
Copy Markdown
Contributor Author

@mgottscho mgottscho Apr 28, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I suppose if this is true then you should probably use a 1-bit CDC FIFO?

Copy link
Copy Markdown
Contributor

@zhemao-openai zhemao-openai Apr 28, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If you can guarantee that the pulses are sufficiently spaced on the source side, you don't need to use a FIFO. I think if you want to add this cover, you should have a parameter like MinPulseSeparation.

Copy link
Copy Markdown
Contributor Author

@mgottscho mgottscho May 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Right so I only see two possibilities. Unless I'm missing something.

  1. If you want to guarantee no data loss - use a CDC FIFO.
  2. If you can guarantee the source pulses are far enough apart that they can be synchronized to destination - use this.

If 2, then shouldn't we have the source side assert MinPulseSeparation unconditionally? If it fails, the user should switch to CDC FIFO for their use case?

Copy link
Copy Markdown
Contributor

@zhemao-openai zhemao-openai May 7, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, I think have an assert that the pulses are minimum number of cycles apart and have the minimum be configurable.


//------------------------------------------
// Implementation
//------------------------------------------
logic src_level;
logic dst_level;
logic dst_level_d;
Expand Down Expand Up @@ -98,7 +106,10 @@ module br_cdc_bit_pulse #(
assign dst_pulse = dst_pulse_internal;
end

// Implementation assertions
// TODO(zhemao): Add some here
//------------------------------------------
// Implementation checks
//------------------------------------------
`BR_ASSERT_CR_IMPL(dst_pulse_high_then_low_a, dst_pulse |=> !dst_pulse, dst_clk, dst_rst)
`BR_COVER_CR_IMPL(back_to_back_dst_pulse_c, dst_pulse ##2 dst_pulse, dst_clk, dst_rst)

endmodule
3 changes: 0 additions & 3 deletions enc/sim/br_enc_bin2onehot_tb.sv
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -81,9 +81,6 @@ module br_enc_bin2onehot_tb;
// Finish simulation
#10;

// TODO(mgottscho): not enough information to know if test passed.
// If DUT asserts fire, we can't see that here.
// Need to determine pass/fail outside of this TB?
if (errors) begin
$display("Number of errors: %0d", errors);
$display("TEST FAILED");
Expand Down