PCALIGN $16
loop64:
	LXVD2X	(R8+R0), V0
	LXVD2X	(R4+R0), V1
	VCMPEQUBCC V0, V1, V2	// compare, setting CR6
	BGELR_CR6
	LXVD2X	(R8+R14), V0
	LXVD2X	(R4+R14), V1
	VCMPEQUBCC	V0, V1, V2
	BGELR_CR6
	LXVD2X	(R8+R15), V0
	LXVD2X	(R4+R15), V1
	VCMPEQUBCC	V0, V1, V2
	BGELR_CR6
	LXVD2X	(R8+R16), V0
	LXVD2X	(R4+R16), V1
	VCMPEQUBCC	V0, V1, V2
	BGELR_CR6
	ADD	$64,R8		// bump up to next 64
	ADD	$64,R4

	LXVD2X	(R0)(R8),V2	// Scan 64 bytes at a time, starting at &s[0]
	VCMPEQUBCC	V2,V1,V6
	BNE	CR6,foundat0	// Match found at R8, jump out

	LXVD2X	(R11)(R8),V2
	VCMPEQUBCC	V2,V1,V6
	BNE	CR6,foundat1	// Match found at R8+16 bytes, jump out

	LXVD2X	(R12)(R8),V2
	VCMPEQUBCC	V2,V1,V6
	BNE	CR6,foundat2	// Match found at R8+32 bytes, jump out

	LXVD2X	(R6)(R8),V2
	VCMPEQUBCC	V2,V1,V6
	BNE	CR6,foundat3	// Match found at R8+48 bytes, jump out

	"VSRAD",
	"VSOI",
	"VSLDOI",
	"VCLZ",
	"VCLZB",
	"VCLZH",
	"VCLZW",
	"VCLZD",
	"VPOPCNT",
	"VPOPCNTB",
	"VPOPCNTH",
	"VPOPCNTW",
	"VPOPCNTD",
	"VCMPEQ",
	"VCMPEQUB",
	"VCMPEQUBCC",
	"VCMPEQUH",
	"VCMPEQUHCC",
	"VCMPEQUW",
	"VCMPEQUWCC",
	"VCMPEQUD",
	"VCMPEQUDCC",
	"VCMPGT",
	"VCMPGTUB",
	"VCMPGTUBCC",
	"VCMPGTUH",
	"VCMPGTUHCC",
	"VCMPGTUW",