boards: fvp_baser_aemv8r: enable cache_state_modelled

Let FVP_BaseR_AEMv8R simulate the operation of cache-related.

This will slow the speed of FVP_BaseR_AEMv8R but it can obtain realistic
simulation results as much as possible. So the timeout_multiplier needs
to be increased.

Signed-off-by: Huifeng Zhang <Huifeng.Zhang@arm.com>

boards/arm64/fvp_baser_aemv8r/board.cmake

@@ -35,5 +35,5 @@ set(ARMFVP_FLAGS
 
   -C bp.vis.disable_visualisation=1
   -C bp.vis.rate_limit-enable=0
-  -C cache_state_modelled=0
+  -C cache_state_modelled=1
   )

boards/arm64/fvp_baser_aemv8r/fvp_baser_aemv8r.yaml

@@ -10,3 +10,5 @@ toolchain:
   - cross-compile
 ram: 2048
 flash: 64
+testing:
+  timeout_multiplier: 8

boards/arm64/fvp_baser_aemv8r/fvp_baser_aemv8r_smp.yaml

@@ -13,4 +13,4 @@ flash: 64
 supported:
   - smp
 testing:
-  timeout_multiplier: 8
+  timeout_multiplier: 20