Probably fix TLB shootdown "diagonal" livelock.

Diagonal as in

Thread A        Thread B
Send to B       Send to A
Wait for B      Wait for A.

I haven't reproduced this bug, but this should fix it, since the wait
steps now checks the local percpu block for pending TLB shootdowns onto
itself.

src/context/memory.rs

@@ -2898,6 +2898,7 @@ impl<'guard, 'addrsp> Flusher<'guard, 'addrsp> {
         }
 
         while self.state.ackword.load(Ordering::SeqCst) < affected_cpu_count {
+            PercpuBlock::current().maybe_handle_tlb_shootdown();
             core::hint::spin_loop();
         }