M-Labs
/
riscv-formal-nmigen
6
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Remove "Known issues" section in README

master
Donald Sebastian Leung 2020-09-16 17:12:40 +08:00
parent a6b072efcf
commit e50d35ead5
1 changed files with 0 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
README.md
View File

@ -43,13 +43,6 @@ This should run in the order of a few hours.
The RV32I, RV32M, RV64I and RV64M ISAs are currently implemented but only RV32IM are being tested by integrating with the Minerva core.
## Known Issues
- ~~21/08/2020: Verification passes unconditionally, even in the presence of obvious bugs. This is due to the `rvfi.valid` signal from the CPU being held at constant 0. Suitable instruction and data buses as well as interrupt signals have not been wired to Minerva which is likely preventing the core and verification from functioning properly~~
- ~~25/08/2020: Interrupts have been hardwired to zero and instruction/data buses attached to the Minerva core; however, running the verification quickly exhausts available memory on a machine with 32G RAM. Possible culprit: huge number of `Assume` statements on memory values~~
- ~~07/09/2020: Instruction checks fail at line 201 asserting the value of `rd_wdata`. At least we now have a proper failure instead of an unconditional pass or an out of memory issue. Whether the translated specs themselves are in error is under investigation.~~
- 15/09/2020: Apart from the occasional failed assertion in specific verification tasks (likely due to bugs in translation), all of the memory-related instruction checks fail with a "parser error: invalid slice" preventing the assertions from ever running. The root cause of this issue is yet to be determined.
## License
See [LICENSE](./LICENSE)