PERF-2 P3-ii: the device V-cycle as the CG's preconditioner — poisson/device.rs (one CUDA runtime per process, persistent per-operator buffers, the mg_vcycle.cu kernels at K = 1; upload r, run the V-cycle, download z; the f64 CG unchanged), MultigridParameters::device, Prepared holds the device hierarchy, the CG driver destructures the prepared operator instead of cloning it; EmbeddedPisoSolver::set_poisson_device, overset pass-through, harness knob RTX_FSI2O_MG_DEVICE=1; export_levels factored out; the quarantine's dangling cfg attribute fixed

Co-Authored-By: Claude Fable 5.1 <[email protected]>
Claude-Session: https://claude.ai/code/session_01YJPeT6WA2e7YvAnS875AHL

crates/specialized/rtx-cfd/src/solvers/incompressible/overset/mod.rs

@@ -342,6 +342,11 @@ impl OversetPisoSolver {
         self.background.set_poisson_threads(threads);
     }
 
+    /// PERF-2 P3-ii: the background multigrid's V-cycle on the CUDA device.
+    pub fn set_poisson_device(&mut self, on: bool) {
+        self.background.set_poisson_device(on);
+    }
+
     /// The step timers, when profiling (`RTX_PROFILE`).
     pub fn timers(&self) -> Option<&StepTimers> {
         self.timers.as_deref()