1/1 
|
Starting with Or1Ksim: delayr & delayw
by mbelaoucha on Sep 12, 2011 |
mbelaoucha
Posts: 3 Joined: Sep 2, 2011 Last seen: Mar 23, 2012 |
||
|
Hi all,
I did start playing with or1ksim, and i wish to understand its API! Is there any dev docs (even a draft)? what's the easiest way (entry point) to understand the API in the shortest time possible. Otherwise, i've a question for you ... Are delayr and delayw information (in config memory sections) accurate for the current release? Were they for earlier releases? Do they, realy, influence the simulation process?! Thank you, |
|||
|
RE: Starting with Or1Ksim: delayr & delayw
by jeremybennett on Sep 12, 2011 |
jeremybennett
Posts: 815 Joined: May 29, 2008 Last seen: Jun 13, 2019 |
||
|
Hi all, I did start playing with or1ksim, and i wish to understand its API! Is there any dev docs (even a draft)? what's the easiest way (entry point) to understand the API in the shortest time possible. Hi mbelaoucha, The public API for the library is in the user guide. There is a small section at the end of the User Guide on internal structure, but that is quite small. Where I have updated the code, it is comprehensively documented, in a form suitable for Doxygen. That will give you some low level documentation. The timing stuff in Or1ksim is a mess. It does not have a timing model that is of any great meaning. It is based on 1 cycle per instruction + memory access cycles if appropriate as specified by delayr/delayw. This is not useful. Or1ksim really ought just to report an instruction count. It you want accurate timings, use the Verilator model. HTH Jeremy
-- |
|||
|
RE: Starting with Or1Ksim: delayr & delayw
by mbelaoucha on Sep 12, 2011 |
mbelaoucha
Posts: 3 Joined: Sep 2, 2011 Last seen: Mar 23, 2012 |
||
|
Hi all, I did start playing with or1ksim, and i wish to understand its API! Is there any dev docs (even a draft)? what's the easiest way (entry point) to understand the API in the shortest time possible. Hi mbelaoucha, The public API for the library is in the user guide. There is a small section at the end of the User Guide on internal structure, but that is quite small. Where I have updated the code, it is comprehensively documented, in a form suitable for Doxygen. That will give you some low level documentation. The timing stuff in Or1ksim is a mess. It does not have a timing model that is of any great meaning. It is based on 1 cycle per instruction + memory access cycles if appropriate as specified by delayr/delayw. This is not useful. Or1ksim really ought just to report an instruction count. It you want accurate timings, use the Verilator model. HTH Jeremy
-- Hi Jeremy, Thanks a lot for the prompt reply, Great, that will make my life easier. I was wondering if my binaries would be accurate for the simulator (with empirical delay values). I'll take a look on the source. Thank you, |
|||
1/1