[iterate-4A] jit: cache MemEnv fast_mem + expose mmio_count (prep for block chaining)

Behavior-neutral prep on top of the inline-load parity milestone:

  * JitCache caches mem.fast_mem() (resolved once — the mapping is invariant for
    a run) so compiled blocks build their MemEnv without a virtual call per
    execution. MemEnv::from_fast(mem, Option<FastMem>) is the non-virtual
    constructor. Measured impact is within noise (96.5 -> 96.4 MIPS) but it
    removes redundant per-block work.
  * FastMem/MemEnv gain mmio_count: a pointer to GuestMemory's monotonic MMIO
    access counter. Unused for now; the upcoming superblock-chaining JIT will
    sample it across a block boundary to stop chaining on an MMIO touch
    (preserving the interpreter's fine-grained MMIO ordering).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

crates/xenia-memory/src/access.rs

@@ -22,6 +22,11 @@ pub struct FastMem {
     /// condition for `addr` to be MMIO, so `!=` proves non-MMIO.
     pub mmio_mask: u32,
     pub mmio_value: u32,
+    /// Pointer to the monotonic MMIO-access counter (`AtomicU64` as `*const
+    /// u64`). A superblock JIT samples this across a block boundary to detect an
+    /// MMIO touch and stop chaining there (preserving the interpreter's
+    /// fine-grained MMIO ordering).
+    pub mmio_count: *const u64,
 }
 
 /// Trait for all guest memory access. Every load/store goes through this,

crates/xenia-memory/src/heap.rs

@@ -519,6 +519,7 @@ impl MemoryAccess for GuestMemory {
             page_table: self.page_table.as_ptr() as *const u64,
             mmio_mask: self.mmio_aperture_mask,
             mmio_value: self.mmio_aperture_value,
+            mmio_count: &self.mmio_access_count as *const AtomicU64 as *const u64,
         })
     }