try to fix MatMulCompressConvertCPUTest for Linux ARM64

Author: antonvor

src/plugins/intel_cpu/tests/functional/subgraph_tests/src/arm/matmul_compress_convert.cpp

@@ -175,7 +175,9 @@ const std::vector<std::vector<InputShape>> inputShapes2D = {
 
 const std::vector<element::Type> inferPrecisions = {
     element::f32,
+#if defined(OV_CPU_ARM_ENABLE_FP16)
     element::f16,
+#endif
 };
 
 const auto testParams2D_ARM_smoke = ::testing::Combine(