feat: full runtime redesign (v0.6)

Complete rewrite with improved architecture & correctness:
- src/runtime.rs: Simplified task scheduling with proper state transitions
- src/scheduler.rs: Decoupled from runtime, pure task queue logic
- src/io.rs, src/mutex.rs: Refactored for clarity & performance
- New actor model framework (src/actor.rs, src/context.rs)
- Channel primitives (src/channel.rs) & process IDs (src/pid.rs)
- Preemption framework (src/preempt.rs) for fair timeslicing
- Expanded benchmarks & tests (multi_scheduler, primes, runtime)

src/pid.rs

@@ -0,0 +1,38 @@
+//! Process identifiers.
+//!
+//! A `Pid` is `(index, generation)`. The index is a slot in the scheduler's
+//! actor table; the generation increments every time that slot is reused.
+//! A stale `Pid` (correct index, wrong generation) is a detectable error,
+//! not a silent misdirection — solves the ABA problem without exhausting
+//! the PID space.
+
+#[derive(Copy, Clone, PartialEq, Eq, Hash)]
+pub struct Pid {
+    index: u32,
+    generation: u32,
+}
+
+impl Pid {
+    #[inline]
+    pub const fn new(index: u32, generation: u32) -> Self {
+        Self { index, generation }
+    }
+
+    #[inline]
+    pub const fn index(self) -> u32 { self.index }
+
+    #[inline]
+    pub const fn generation(self) -> u32 { self.generation }
+}
+
+impl std::fmt::Debug for Pid {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "Pid({}.{})", self.index, self.generation)
+    }
+}
+
+impl std::fmt::Display for Pid {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "<{}.{}>", self.index, self.generation)
+    }
+}