Update readme.md

pull/18/head
Bill Flynn 2 years ago committed by GitHub
parent 659fd8b7b4
commit 3b494a09a4
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

@ -6,16 +6,17 @@


* updated source to remove a bunch of Verilator warnings * updated source to remove a bunch of Verilator warnings
* updated source for compatibility with Icaraus -g2012 * updated source for compatibility with Icaraus -g2012
* changed the GPR/FPR macro to remove need for clk4x on FPGA; was ugly and not working with either simulator * changed arrays to use clk1x

* refactored nclk[] to separate clk, rst signals and removed lcb's from clock path
* Verilator * Verilator


* can IFETCH some ops in ST and SMT2 mode * too big(?) to build with verilator --public with cocotb
* can't get cocotb to build with Verilator (very long compile times) * executing boot code and tst with a2o_litex wrapper


* Icarus (w/cocotb) * Icarus (w/cocotb)


* executing boot code until test call with python A2L2 interface * executing boot code and tst with core/node/a2o_litex


* Yosys * Yosys


@ -24,18 +25,11 @@


## Next To Do ## Next To Do


* create A2L2 cpp model that can be used by Verilator and cocotb (Cython wrapper) * create Cython code for sharing between cocotb and verilator tb
* eventually; don't need right now if can skip to RTL for A2L2

* create simple A2L2-WB RTL for easily connecting to Litex, etc.
* version 1 is pre-WB; uses RAM interface which could be sim driver or BRAM

* create Litex core wrapper
* OpenLane experiments with blackbox arrays and yosys/abc/sta * OpenLane experiments with blackbox arrays and yosys/abc/sta
* create FPGA version of GPR/FPR (4R4W) using (4)4R1W banks and *valid* table * create FPGA version of GPR/FPR (4R4W) using (4)4R1W banks and *valid* table
* parse vcd/fst and serve browser code for custom trace screens (handle spec/arch mapped facilities, arrays, etc.) * parse vcd/fst and serve browser code for custom trace screens (handle spec/arch mapped facilities, arrays, etc.)



#### node #### node


* test BE/LE versions; kernel can stay BE until jump to BIOS; any problem with BIOS to initial ROM copy/zero or is it always words? * test BE/LE versions; kernel can stay BE until jump to BIOS; any problem with BIOS to initial ROM copy/zero or is it always words?

Loading…
Cancel
Save