Optimize bssn_rhs calculation with cache blocking and vectorization

- Implemented cache blocking (BLK=8) in bssn_rhs_opt.f90 to improve L1/L2 cache hit rate.
- Introduced bssn_rhs_opt.f90 module with vectorized derivative and physics kernels.
- Renamed original implementation to bssn_rhs_legacy.f90 for fallback.
- Updated bssn_rhs.f90 to act as a dispatcher, using the optimized path for ghost_width=3.
- Updated makefile to include new source files.
- Added DEBUG_NAN_CHECK macro to optionally disable NaN checks in production.

AMSS_NCKU_source/makefile

@@ -34,7 +34,7 @@ C++FILES_GPU = ABE.o Ansorg.o Block.o misc.o monitor.o Parallel.o MPatch.o var.o
 
 F90FILES = enforce_algebra.o fmisc.o initial_puncture.o prolongrestrict.o\
 	   prolongrestrict_cell.o prolongrestrict_vertex.o\
-	   rungekutta4_rout.o bssn_rhs.o diff_new.o kodiss.o kodiss_sh.o\
+	   rungekutta4_rout.o bssn_rhs_opt.o bssn_rhs.o bssn_rhs_legacy.o diff_new.o kodiss.o kodiss_sh.o\
 	   lopsidediff.o sommerfeld_rout.o getnp4.o diff_new_sh.o\
 	   shellfunctions.o bssn_rhs_ss.o Set_Rho_ADM.o\
            getnp4EScalar.o bssnEScalar_rhs.o bssn_constraint.o ricci_gamma.o\