Skip to content

Fix sky130 LitePCI clock period parsing - #241

Open
appleweiping wants to merge 1 commit into
VLSIDA:mainfrom
appleweiping:fix/litepci-sky130-clock-period-238
Open

appleweiping wants to merge 1 commit into
VLSIDA:mainfrom
appleweiping:fix/litepci-sky130-clock-period-238

Conversation

@appleweiping

Copy link
Copy Markdown

Fixes #238.

The sky130 LitePCI SDC had a historical reference-clock example before the real clock-period assignment. The sed expression in the pinned bazel-orfs revision takes the first matching period token, so it captured the historical value and its closing backtick and generated an invalid shell command.

This moves the real 20 ns assignment above the comments and rewrites the historical note. It also updates the LitePCI decision record and workaround table. This restores synthesis startup only; the later sky130 routing problem in #204 remains.

Validation:

  • Ran the exact clock-period sed expression from the pinned bazel-orfs revision under GNU sed
  • Before the fix, sky130hd returned 10000 followed by a backtick; after the fix, it returns 20
  • The same check still returns 10 for nangate45 and 3600 for asap7
  • git diff --check passes
  • Full Bazel synthesis was not run because Bazel is unavailable in the local WSL environment

This branch has not been deployed

No deployments
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.

litepci-sky130hd: synth fails at clock_period.txt — stray backtick in SDC clock-period (sed-poisoning)

1 participant