[Open SoC Debug] Queries Regarding GSoC and Open SoC

Philipp Wagner philipp.wagner at tum.de
Fri Mar 23 15:37:36 CET 2018


Hi,

On 03/21/2018 02:11 AM, Anant Sharma wrote:
> Its been a while. I will tell you the things I have tried from my side 
> to try and implement a breakpoint in the mor1kx module. I will try to be 
> as specific as I can so that you might be able to help me out. I will 
> also talk about the problems I faced.
> 
> 1. I understand the registers on which I have to load bits to get things 
> done. But, I had to do a lot of research on the repository of mor1kx to 
> find out which register does what. I am still uncertain but I know the 
> path that is being followed by each bit put in the 'du_*' registers. 
> There are things that I am still uncertain about like the difference 
> between the data and address register.

The address bus is for the address of the register to read/write, the 
data bus is (the name says it all) for the data stored in or written to 
the register specified on the address bus/port.

The semantics behind the addresses and values of this du* bus are given 
in the ISA specification of or1k 
(https://raw.githubusercontent.com/openrisc/doc/master/openrisc-arch-1.2-rev0.pdf).

> 2. Any values I load on the registers on the mor1kx_module file is not 
> effecting the output at the gtkwave at all. One of the possible reasons 
> might be that I am not in Supervisor mode but I am having trouble 
> figuring that out.

What code did you use to interact with these the du_* bus? Where did you 
insert it? Please push your modified code somewhere to a repository if 
you have questions on that.

> 3. In my experience with Verilog. You simply cannot load stuff into 
> input ports of the module. I obviously know how to write to a bus but 
> that is not exactly an editable bus here. 

What do you mean by "editable bus"?

 > I would have to deal with the
> 'from where' are the bits in the register coming from so that this can 
> be edited and changes can be made at the source. What I am doing as of 
> now is creating registers of similar names and using those registers to 
> push bits into the next channel.
> 
> If I am able to solve this out. I will start writing my proposal.

Looking forward to that!

Philipp


More information about the OpenSoCDebug mailing list