Error in using sram macro(tiny) in openlane2 #247

Activity

1. mguthaus commented on Jul 19, 2024

   

   mguthaus

   on Jul 19, 2024 via email

   Collaborator

   More actions

   The layer errors are complaining about necessary layers in the macro. These are for OPC etc and cannot be removed.

   …

   On Fri, Jul 19, 2024, 02:52 K.Makise ***@***.***> wrote: *Describe the bug* Hello, I encountered some errors in trying to use sram macros in openlane2. The openlane2(magic) complains that some cells are placed at itself, and unknown layer/datatype in boundary. I also try just to use only the bank extracted from the sram macro, but failed with similar error. default.png (view on web) <https://github.com/user-attachments/assets/41988ad6-2584-4151-a34f-2a9bcd5344ae> I'm not sure if it's about the gds file, but it looks fine when opening with klayout There is also one similar error -> The-OpenROAD-Project/OpenLane#1040 <The-OpenROAD-Project/OpenLane#1040> *Version* Openram1.2.48, Openlane 2.0.0rc2(but 2.04 has same problem) *To Reproduce* macro_test.zip <https://github.com/user-attachments/files/16310272/macro_test.zip> *Expected behavior* No error in reading gds *Logs* Logs contained in the uploaded file — Reply to this email directly, view it on GitHub <#247>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC67SL7J7NOG2IFRIL6C5QTZNDOVFAVCNFSM6AAAAABLEIRVW2VHI2DSMVQWIX3LMV43ASLTON2WKOZSGQYTQNJRGQYTOMY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>
2. FriedrichWu commented on Jul 20, 2024

   

   FriedrichWu

   on Jul 20, 2024

   ContributorAuthor

   More actions

   Thanks for the quick response.

   I see. Then it's probably the problem of the openlane. Is there maybe any workaround to make the tool to recognize those necessary layers?
3. mguthaus commented on Jul 20, 2024

   

   mguthaus

   on Jul 20, 2024

   Collaborator

   More actions

   It's shown as an error, but it is really a warning.
4. mguthaus commented on Oct 13, 2026

   

   mguthaus

   on Oct 13, 2026

   Collaborator

   More actions

   Still open. This may be partly on the OpenLane side rather than OpenRAM's, since it concerns consuming the generated macro rather than producing it.

   One thing that has changed since: #322 just fixed the sky130 LEF header on dev, which declared DATABASE MICRONS 2000 when sky130 tech LEFs use 1000 (#278). That broke macro import in OpenROAD-based flows, so it is worth retesting against dev before digging further.

   If you would like to take it, a PR against dev would be very welcome and we will review and consider it — see CONTRIBUTING.md. Contributions go to the dev branch, and the regression suite is run with cd compiler/tests && make -k -j<N>.