jitterentropy vs. simulation

[Johannes Berg]

On Fri, 2023-12-01 at 14:25 -0500, Simo Sorce wrote:
> 
> Doesn't this imply the simulation is not complete

Kind of?

>  and you need to add
> clock jitter for the simulation to be more useful?
> 

No, it's more _intentionally_ incomplete. This works fine in normal
ARCH=um, but with time-travel variants we have more of a discrete event-
based simulation, to integrates well with other things (e.g. SystemC or
similar based device simulations). So this is quite intentional, and
yes, it breaks in a few spots such as this one.

johannes