dstep | A tool for converting C and Objective-C headers to D modules | Compiler library

The error is caused by the version mismatching between the RTX GPU cards and the CUDA driver.

First of all, i'm not sure, if you want to use discrete time or continuous time, because you're using s operator for cont. time, the functions dstep and dlsim are used for discrete time representation. However, i used the continuous one in my example.

You can create a dlti object in python with scipy.signal's lti function. The created filter object has a method step where the first parameter is used for the initial time vector. lti.step So you can plot your step response with this snippet.

As suggested, you could get actual motivation from the OpenCV forums. Note however that historically OpenCV has been developed with real-time or near-real-time applications in mind (for example, the Darpa Grand Challenge), hence you'll find easily code that optimizes for speed over accuracy.

First, let me try give short answers to each of your questions:

1. I realise that proc 0 can't terminate until 1 and 2 have terminated, but why are the terminations of 2 and 1 being interleaved non-deterministically?

The processes always terminate in a deterministic fashion: 2 terminates before 1, 1 before 0 and 0 is always the last one. However, there is nothing special about process termination: it is simply the final transition that is taken by a process. As a result, process interleaving is possible at any point in time in which there are more than one process with an (immediately) executable instruction.

2. Any why is SPIN randomly choosing between executing proc 1 and proc 2 when the init function is atomic and therefore should be executed in order?

Although it is true that init executes both of his instructions atomically, the important fact to keep in mind is that step1 and step2 are independent processes and are executed after init exits its atomic block: run is not a function call, it just spawns a process inside the environment with absolutely no guarantee that such process is executed immediately. That might or might not happen depending on whether the spawned process has any executable instruction, whether the process that is currently executing is in an atomic sequence and on the outcome of the non-deterministic scheduler process selection.

3. And why can I get proc 2 to start and terminate before proc 1 (in Run 3) but not the other way around?

In Promela, processes can only die in reverse order of their creation, as indicated in the docs: