[Oberon] Unlimited Oberon System for any board
joerg.straube at iaeth.ch
Sun May 10 00:22:18 CEST 2020
The „right one“: With the 32bit address space there are two flavours of the RISC-V ISA:
RV32I and its subset RV32E for small embedded controllers (only 16 registers and no floating point)
RV32I has several extensions
M: multiplication/division of integers
A: atomic instruction for interprocess communication
F: single precision floating point instructions
D: double precision floating point instructions
If a processor implements all these 4 defined extensions (IMAFD) it is called G.
So, a RISC-V processor called „RV32G“, implements all of the above instructions.
> Am 08.05.2020 um 18:24 schrieb Guy T. <turgu666 at gmail.com>:
> Thanks Wojtek,
> I only hope that if RISC-V is targeted, it will be the right one, whatever that means. I’ve looked to the makefile of the project you mention. They seems to be using the rv32i architecture as defined for gcc ("-march=rv32i” parameter to the compiler). Could be related to “the RISC-V” then.
>> On May 8, 2020, at 11:19 AM, Skulski, Wojciech <skulski at pas.rochester.edu> wrote:
>>> What about using RISC-V ISA on FPGA, allowing for “some kind of” easy path between RISC-V chips and FPGA made systems?
>> Have a look at
>> http://opencores.org/project,hf-risc (older version)
>> https://github.com/sjohann81/hf-risc (newer version)
>> The older version from Open Cores runs on the same Spartan-3 starter kit which originally hosted RISC5. The newer version removed that support and added some other stuff. The highlights which are worth exploring:
>> 1. The project is mostly in VHDL which I like much better than NW Verilog.
>> 2. The structure of the HF-RISC-V seems more clear to me than RISC5, which is IMHO too terse with too few comments (ETH specialty, I guess).
>> 3. The HF CPU is explicitly pipelined and there is some discussion. The discussion is a little bit above my head, but at least there is hope of understanding.
>> 4. There is also a MIPS variant in the same archive. It may be interesting to compare the three approaches, HF-RISC-V, MIPS, and RISC5.
>> Please note that this is "a RISC-V" rather than "the RISC-V". I am not sure what is the connection between *this* RISC-V and the official one. I suspect that just like there is no "the ARM", there is no such thing as "the RISC-V" either.
>> Hope it helps,
>> Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
> Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
More information about the Oberon