mirror of https://github.com/YosysHQ/yosys.git
Updated TODOs in README file
This commit is contained in:
parent
e340532ce5
commit
d248419fe0
32
README
32
README
|
@ -287,17 +287,37 @@ and after each occurrence of PRIi64 in the header file:
|
|||
Roadmap / Large-scale TODOs
|
||||
===========================
|
||||
|
||||
- Technology mapping for real-world applications (specific FPGAs and ASIC processes)
|
||||
- Implement SAT-based formal equivialence checker based on existing SAT framework
|
||||
- Re-implement Verilog frontend (cleaner AST format, pipeline of well structured AST transformations)
|
||||
- Verification and Regression Tests
|
||||
- VlogHammer: http://www.clifford.at/yosys/vloghammer.html
|
||||
- yosys-bigsim: https://github.com/cliffordwolf/yosys-bigsim
|
||||
|
||||
- Missing Verilog-2005 features to be implemented soon:
|
||||
- Fix corner cases with contextual name lookup
|
||||
- Part select on memory read
|
||||
- Indexed part selects
|
||||
|
||||
- Technology mapping for real-world applications
|
||||
- Add "mini synth script" feature to techmap pass
|
||||
- Add const-folding via cell parameters to techmap pass
|
||||
- Rewrite current stdcells.v techmap rules (modular and clean)
|
||||
- Improve Xilinx FGPA synthesis (RAMB, CARRY4, SLR, etc.)
|
||||
|
||||
- Implement SAT-based formal equivialence checker
|
||||
- Add x-state support to SAT model generator
|
||||
- Rewrite freduce pass with input-cone analysis
|
||||
- Write equiv pass, base hypothesis on input cones
|
||||
|
||||
- Re-implement Verilog frontend (far future)
|
||||
- cleaner (easier to use, harder to use wrong) AST format
|
||||
- pipeline of well structured AST transformations
|
||||
- true contextual name lookup
|
||||
|
||||
|
||||
TODOs / Open Bugs
|
||||
=================
|
||||
Other Unsorted TODOs
|
||||
====================
|
||||
|
||||
- Implement missing Verilog 2005 features:
|
||||
|
||||
- Indexed part selects
|
||||
- Multi-dimensional arrays
|
||||
- ROM modeling using "initial" blocks
|
||||
- Ignore what needs to be ignored (e.g. drive and charge strengths)
|
||||
|
|
Loading…
Reference in New Issue