arm64: use .subsections_via_symbols on Darwin

Apple's linker never removes anything from an object without this
directive, so a static link with -dead_strip kept every assembly kernel it
had loaded, used or not.  With it, a kernel that nothing refers to is left
out, as the C kernels already are.

This is safe now that every label inside a kernel is L(name), local to the
assembler: each kernel object defines one symbol, its entry point, and so is
one block for the linker.  EPILOGUE, until now empty, is added to the five
kernels that did not end in it.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

Author: topolarity

common_arm64.h

@@ -168,7 +168,11 @@ static inline int blas_quickdivide(blasint x, blasint y){
 .endm
 
 
+#if defined(__APPLE__)
+#define EPILOGUE .subsections_via_symbols
+#else
 #define EPILOGUE
+#endif
 
 #define PROFCODE
 

kernel/arm64/axpy.S

@@ -207,3 +207,5 @@ L(axpy_kernel_L999):
 
 	mov	w0, wzr
 	ret
+
+	EPILOGUE

kernel/arm64/daxpy_thunderx2t99.S

@@ -253,3 +253,5 @@ L(daxpy_kernel_L999):
 
 	mov	w0, wzr
 	ret
+
+	EPILOGUE

kernel/arm64/rot.S

@@ -243,3 +243,5 @@ L(rot_kernel_L999):
 
 	mov	w0, wzr
 	ret
+
+	EPILOGUE

kernel/arm64/zaxpy.S

@@ -322,3 +322,5 @@ L(zaxpy_kernel_L999):
 
 	mov	w0, wzr
 	ret
+
+	EPILOGUE

kernel/arm64/zrot.S

@@ -254,3 +254,5 @@ L(zrot_kernel_L999):
 
 	mov	w0, wzr
 	ret
+
+	EPILOGUE