TC-enhanced_Cross-correlation_Function/Makefile at main · ymagchi/TC-enhanced_Cross-correlation_Function · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
FC = nvfortran
NVCC = nvcc

GPU_VER1 = -Mcuda=cuda11.2 -ta=tesla:cc80,loadcache:L1
GPU_VER2 = arch=compute_80,code=sm_80

BASEFLAGS = -DNBLOCK=6 -DNTENSOR=16
BASEFLAGS += -DVERIFICATION

OBJS = \
	compdef_gpu.o \
	precision.o \
	cublas_wrapper.o \
	main.o

FFLAGS = -fastsse -O3 $(GPU_VER1) $(BASEFLAGS)
NVCCFLAGS = --generate-code $(GPU_VER2) -O3 --use_fast_math -lineinfo -Xptxas="-v" $(BASEFLAGS)

PROGRAM = a.out

.SUFFIXES: $(SUFFIXES) .F90 .cu

all: $(PROGRAM)

.cu.o:
	$(NVCC) $(NVCCFLAGS) -c $<
.F.o:
	$(FC) $(FFLAGS) -c $<

$(PROGRAM):$(OBJS)
	$(FC) $(FFLAGS) $(OBJS)  -o $@

clean:
	rm -f *.o *.ptx *~ *.mod $(PROGRAM)