Implement the ffs and fls functions, and their longer counterparts, in

cpufunc, in terms of __builtin_ffs and the like, for arm64
architectures, and use those, rather than the simple libkern
implementations, in building arm64 kernels.

Tested by: greg_unrelenting.technology (earlier version)
Reviewed by: alc
Approved by: kib (mentor)
Differential Revision: https://reviews.freebsd.org/D20250

sys/arm64/include/cpufunc.h

@@ -38,6 +38,63 @@ breakpoint(void)
 
 #ifdef _KERNEL
 
+#define	HAVE_INLINE_FFS
+
+static __inline __pure2 int
+ffs(int mask)
+{
+
+	return (__builtin_ffs(mask));
+}
+
+#define	HAVE_INLINE_FFSL
+
+static __inline __pure2 int
+ffsl(long mask)
+{
+
+	return (__builtin_ffsl(mask));
+}
+
+#define	HAVE_INLINE_FFSLL
+
+static __inline __pure2 int
+ffsll(long long mask)
+{
+
+	return (__builtin_ffsll(mask));
+}
+
+#define	HAVE_INLINE_FLS
+
+static __inline __pure2 int
+fls(int mask)
+{
+
+	return (mask == 0 ? 0 :
+	    8 * sizeof(mask) - __builtin_clz((u_int)mask));
+}
+
+#define	HAVE_INLINE_FLSL
+
+static __inline __pure2 int
+flsl(long mask)
+{
+
+	return (mask == 0 ? 0 :
+	    8 * sizeof(mask) - __builtin_clzl((u_long)mask));
+}
+
+#define	HAVE_INLINE_FLSLL
+
+static __inline __pure2 int
+flsll(long long mask)
+{
+
+	return (mask == 0 ? 0 :
+	    8 * sizeof(mask) - __builtin_clzll((unsigned long long)mask));
+}
+
 #include <machine/armreg.h>
 
 void pan_enable(void);

sys/conf/files.arm64

@@ -258,12 +258,6 @@ kern/pic_if.m			optional	intrng
 kern/subr_devmap.c		standard
 kern/subr_intr.c		optional	intrng
 libkern/bcmp.c			standard
-libkern/ffs.c			standard
-libkern/ffsl.c			standard
-libkern/ffsll.c			standard
-libkern/fls.c			standard
-libkern/flsl.c			standard
-libkern/flsll.c			standard
 libkern/memcmp.c		standard
 libkern/memset.c		standard
 libkern/arm64/crc32c_armv8.S	standard