Follow up ideas from #924

[ludfjig]

nit: Could we have LinuxInterruptHandle::new(vm, config) and WindowsInterruptHandle::new(vm, config)? Can we also have struct GenericInterruptHandle(Arc<dyn InterruptHandleImpl>) and GenericInterruptHandle::new(vm, config)? see here

========

suggestion: it would be nice if interupt_handle had a method

enum InterruptibleResult<T> {
    Completed(T),
    Interrupted,
    Cancelled,
}
fn run_interruptible<T>(&mut self, func: impl FnOnce() -> T) -> InterruptibleResult<T>

and all the complexity of the interrupt handler can be contained there

then you can do

let exit_reason = self.interrupt_handle.run_interruptible(|| {
    #[cfg(feature = "trace_guest")]
    tc.setup_guest_trace(Span::current().context());
    let result = self.vm.run_vcpu();

    // End current host trace by closing the current span that captures traces
    // happening when a guest exits and re-enters.
    #[cfg(feature = "trace_guest")]
    {
        // end tracing stuff
    }
    result
})

But I would say not for this PR, it would just add noise
see here

=====

Can we reach this point? is is possible to call this method, match a region, not be a guard page, and neither an access violation?
I assume this matches the previous behavior, if it does, and we are still not sure if this is reachable, consider adding a TODO comment to check if this is reachable
see here

=====

Restructure feature-gated fields in HyperlightVm struct

        #[cfg(target_os = "windows")] handle: HandleWrapper,
        #[cfg(target_os = "windows")] raw_size: usize,
        #[cfg(gdb)] gdb_conn: Option<DebugCommChannel<DebugResponse, DebugMsg>>,
        #[cfg(crashdump)] rt_cfg: SandboxRuntimeConfig,
        #[cfg(feature = "mem_profile")] trace_info: MemTraceInfo,

=====

OK, so we no longer capture registers in the log? Again I think that was there because it was used in the past to try and narrow down where issues were (registers such as RIP are/were useful to try and determine what\where errors were happening) I suspect that this isnt as important now that we have crashdump capability

see here