[pcsx_rearmed.git] / deps / lightning / check / range.tst

#define M64	67108864

#define aB1	(1<<1)
#define aB2	(1<<2)
#define aB3	(1<<3)
#define aB4	(1<<4)
#define aB5	(1<<5)
#define aB6	(1<<6)
#define aB7	(1<<7)
#define aB8	(1<<8)
#define aB9	(1<<9)
#define aB10	(1<<10)
#define aB11	(1<<11)
#define aB12	(1<<12)
#define aB13	(1<<13)
#define aB14	(1<<14)
#define aB15	(1<<15)
#define aB16	(1<<16)
#define aB17	(1<<17)
#define aB18	(1<<18)
#define aB19	(1<<19)
#define aB20	(1<<20)
#define aB21	(1<<21)
#define aB22	(1<<22)
#define aB23	(1<<23)
#define aB24	(1<<24)
#define aB25	(1<<25)
#define aB26	(1<<26)
#define bB1	(-aB1)
#define bB2	(-aB2)
#define bB3	(-aB3)
#define bB4	(-aB4)
#define bB5	(-aB5)
#define bB6	(-aB6)
#define bB7	(-aB7)
#define bB8	(-aB8)
#define bB9	(-aB9)
#define bB10	(-aB10)
#define bB11	(-aB11)
#define bB12	(-aB12)
#define bB13	(-aB13)
#define bB14	(-aB14)
#define bB15	(-aB15)
#define bB16	(-aB16)
#define bB17	(-aB17)
#define bB18	(-aB18)
#define bB19	(-aB19)
#define bB20	(-aB20)
#define bB21	(-aB21)
#define bB22	(-aB22)
#define bB23	(-aB23)
#define bB24	(-aB24)
#define bB25	(-aB25)
#define bB26	(-aB26)
#define cB1	(aB1-1)
#define cB2	(aB2-1)
#define cB3	(aB3-1)
#define cB4	(aB4-1)
#define cB5	(aB5-1)
#define cB6	(aB6-1)
#define cB7	(aB7-1)
#define cB8	(aB8-1)
#define cB9	(aB9-1)
#define cB10	(aB10-1)
#define cB11	(aB11-1)
#define cB12	(aB12-1)
#define cB13	(aB13-1)
#define cB14	(aB14-1)
#define cB15	(aB15-1)
#define cB16	(aB16-1)
#define cB17	(aB17-1)
#define cB18	(aB18-1)
#define cB19	(aB19-1)
#define cB20	(aB20-1)
#define cB21	(aB21-1)
#define cB22	(aB22-1)
#define cB23	(aB23-1)
#define cB24	(aB24-1)
#define cB25	(aB25-1)
#define cB26	(aB26-1)
#define dB1	(-aB1+1)
#define dB2	(-aB2+1)
#define dB3	(-aB3+1)
#define dB4	(-aB4+1)
#define dB5	(-aB5+1)
#define dB6	(-aB6+1)
#define dB7	(-aB7+1)
#define dB8	(-aB8+1)
#define dB9	(-aB9+1)
#define dB10	(-aB10+1)
#define dB11	(-aB11+1)
#define dB12	(-aB12+1)
#define dB13	(-aB13+1)
#define dB14	(-aB14+1)
#define dB15	(-aB15+1)
#define dB16	(-aB16+1)
#define dB17	(-aB17+1)
#define dB18	(-aB18+1)
#define dB19	(-aB19+1)
#define dB20	(-aB20+1)
#define dB21	(-aB21+1)
#define dB22	(-aB22+1)
#define dB23	(-aB23+1)
#define dB24	(-aB24+1)
#define dB25	(-aB25+1)
#define dB26	(-aB26+1)

#define add(a, b)		$(a + b)
#define sub(a, b)		$(a - b)
#define rsb(a, b)		$(b - a)
#define mul(a, b)		$(a * b)
#define div(a, b)		$(a / b)
#define rem(a, b)		$(a % b)
#define and(a, b)		$(a & b)
#define or(a, b)		$(a | b)
#define xor(a, b)		$(a ^ b)

#define alu2(N, X, L, R, V)					\
	movi %r1 L						\
	N##i %r0 %r1 R						\
	beqi X %r0 V						\
	calli @abort						\
X:
#define alu1(N, M)						\
	alu2(N, N##M##1, 3, $(M##1), N(3, M##1))		\
	alu2(N, N##M##2, 3, $(M##2), N(3, M##2))		\
	alu2(N, N##M##3, 3, $(M##3), N(3, M##3))		\
	alu2(N, N##M##4, 3, $(M##4), N(3, M##4))		\
	alu2(N, N##M##5, 3, $(M##5), N(3, M##5))		\
	alu2(N, N##M##6, 3, $(M##6), N(3, M##6))		\
	alu2(N, N##M##7, 3, $(M##7), N(3, M##7))		\
	alu2(N, N##M##8, 3, $(M##8), N(3, M##8))		\
	alu2(N, N##M##9, 3, $(M##9), N(3, M##9))		\
	alu2(N, N##M##10, 3, $(M##10), N(3, M##10))		\
	alu2(N, N##M##11, 3, $(M##11), N(3, M##11))		\
	alu2(N, N##M##12, 3, $(M##12), N(3, M##12))		\
	alu2(N, N##M##13, 3, $(M##13), N(3, M##13))		\
	alu2(N, N##M##14, 3, $(M##14), N(3, M##14))		\
	alu2(N, N##M##15, 3, $(M##15), N(3, M##15))		\
	alu2(N, N##M##16, 3, $(M##16), N(3, M##16))		\
	alu2(N, N##M##17, 3, $(M##17), N(3, M##17))		\
	alu2(N, N##M##18, 3, $(M##18), N(3, M##18))		\
	alu2(N, N##M##19, 3, $(M##19), N(3, M##19))		\
	alu2(N, N##M##20, 3, $(M##20), N(3, M##20))		\
	alu2(N, N##M##21, 3, $(M##21), N(3, M##21))		\
	alu2(N, N##M##22, 3, $(M##22), N(3, M##22))		\
	alu2(N, N##M##23, 3, $(M##23), N(3, M##23))		\
	alu2(N, N##M##24, 3, $(M##24), N(3, M##24))		\
	alu2(N, N##M##25, 3, $(M##25), N(3, M##25))		\
	alu2(N, N##M##26, 3, $(M##26), N(3, M##26))

#define alu(N)							\
	alu1(N, aB)						\
	alu1(N, bB)						\
	alu1(N, cB)						\
	alu1(N, dB)

#define _lsh(N)							\
	alu2(lsh, L##N, 3, N, $(3<<N))
#define _rsh(N)							\
	alu2(rsh, R##N, $(1<<63), N, $((1<<63)>>N))

#if __WORDSIZE == 32
#  define xsh64(X)			/**/
#else
#  define xsh64(X)						\
	_##X##sh(32)						\
	_##X##sh(33)						\
	_##X##sh(34)						\
	_##X##sh(35)						\
	_##X##sh(36)						\
	_##X##sh(37)						\
	_##X##sh(38)						\
	_##X##sh(39)						\
	_##X##sh(40)						\
	_##X##sh(41)						\
	_##X##sh(42)						\
	_##X##sh(43)						\
	_##X##sh(44)						\
	_##X##sh(45)						\
	_##X##sh(46)						\
	_##X##sh(47)						\
	_##X##sh(48)						\
	_##X##sh(49)						\
	_##X##sh(50)						\
	_##X##sh(51)						\
	_##X##sh(52)						\
	_##X##sh(53)						\
	_##X##sh(54)						\
	_##X##sh(55)						\
	_##X##sh(56)						\
	_##X##sh(57)						\
	_##X##sh(58)						\
	_##X##sh(59)						\
	_##X##sh(60)						\
	_##X##sh(61)						\
	_##X##sh(62)						\
	_##X##sh(63)
#endif

#define xsh(X)							\
	_##X##sh(0)						\
	_##X##sh(1)						\
	_##X##sh(2)						\
	_##X##sh(3)						\
	_##X##sh(4)						\
	_##X##sh(5)						\
	_##X##sh(6)						\
	_##X##sh(7)						\
	_##X##sh(8)						\
	_##X##sh(9)						\
	_##X##sh(10)						\
	_##X##sh(11)						\
	_##X##sh(12)						\
	_##X##sh(13)						\
	_##X##sh(14)						\
	_##X##sh(15)						\
	_##X##sh(16)						\
	_##X##sh(17)						\
	_##X##sh(18)						\
	_##X##sh(19)						\
	_##X##sh(20)						\
	_##X##sh(21)						\
	_##X##sh(22)						\
	_##X##sh(23)						\
	_##X##sh(24)						\
	_##X##sh(25)						\
	_##X##sh(26)						\
	_##X##sh(27)						\
	_##X##sh(28)						\
	_##X##sh(29)						\
	_##X##sh(30)						\
	_##X##sh(31)						\
	xsh64(X)

#define lsh()							\
	xsh(l)
#define rsh()							\
	xsh(r)

#define reset(V)						\
	prepare							\
		pushargi buf					\
		pushargi V					\
		pushargi $(M64 + 8)				\
	finishi @memset

#define stx(T, N, O, V)						\
	movi %r0 V						\
	stxi##T O %v0 %r0
#define stx8(T, M, V)						\
	stx(T, 3, $(M##B3), V)					\
	stx(T, 4, $(M##B4), V)					\
	stx(T, 5, $(M##B5), V)					\
	stx(T, 6, $(M##B6), V)					\
	stx(T, 7, $(M##B7), V)					\
	stx(T, 8, $(M##B8), V)					\
	stx(T, 9, $(M##B9), V)					\
	stx(T, 10, $(M##B10), V)				\
	stx(T, 11, $(M##B11), V)				\
	stx(T, 12, $(M##B12), V)				\
	stx(T, 13, $(M##B13), V)				\
	stx(T, 14, $(M##B14), V)				\
	stx(T, 15, $(M##B15), V)				\
	stx(T, 16, $(M##B16), V)				\
	stx(T, 17, $(M##B17), V)				\
	stx(T, 18, $(M##B18), V)				\
	stx(T, 19, $(M##B19), V)				\
	stx(T, 20, $(M##B20), V)				\
	stx(T, 21, $(M##B21), V)				\
	stx(T, 22, $(M##B22), V)				\
	stx(T, 23, $(M##B23), V)				\
	stx(T, 24, $(M##B24), V)				\
	stx(T, 25, $(M##B25), V)				\
	stx(T, 26, $(M##B26), V)
#define stx4(T, M, V)						\
	stx(T, 2, $(M##B2), V)					\
	stx8(T, M, V)
#define stx2(T, M, V)						\
	stx(T, 1, $(M##B1), V)					\
	stx4(T, M, V)
#define ldx(T, N, M, O, V)					\
	movi %r0 0						\
	ldxi##T %r0 %v0 O					\
	beqi ldx##T##N##M %r0 V					\
	calli @abort						\
ldx##T##N##M:
#define ldx8(T, M, V)						\
	ldx(T, 3, M, $(M##B3), V)				\
	ldx(T, 4, M, $(M##B4), V)				\
	ldx(T, 5, M, $(M##B5), V)				\
	ldx(T, 6, M, $(M##B6), V)				\
	ldx(T, 7, M, $(M##B7), V)				\
	ldx(T, 8, M, $(M##B8), V)				\
	ldx(T, 9, M, $(M##B9), V)				\
	ldx(T, 10, M, $(M##B10), V)				\
	ldx(T, 11, M, $(M##B11), V)				\
	ldx(T, 12, M, $(M##B12), V)				\
	ldx(T, 13, M, $(M##B13), V)				\
	ldx(T, 14, M, $(M##B14), V)				\
	ldx(T, 15, M, $(M##B15), V)				\
	ldx(T, 16, M, $(M##B16), V)				\
	ldx(T, 17, M, $(M##B17), V)				\
	ldx(T, 18, M, $(M##B18), V)				\
	ldx(T, 19, M, $(M##B19), V)				\
	ldx(T, 20, M, $(M##B20), V)				\
	ldx(T, 21, M, $(M##B21), V)				\
	ldx(T, 22, M, $(M##B22), V)				\
	ldx(T, 23, M, $(M##B23), V)				\
	ldx(T, 24, M, $(M##B24), V)				\
	ldx(T, 25, M, $(M##B25), V)				\
	ldx(T, 26, M, $(M##B26), V)
#define ldx4(T, M, V)						\
	ldx(T, 2, M, $(M##B2), V)				\
	ldx8(T, M, V)
#define ldx2(T, M, V)						\
	ldx(T, 1, M, $(M##B1), V)				\
	ldx4(T, M, V)

#define stf(T, N, O, V)						\
	movi##T %f0 V						\
	stxi##T O %v0 %f0
#define stf8(T, M, V)						\
	stf(T, 3, $(M##B3), V)					\
	stf(T, 4, $(M##B4), V)					\
	stf(T, 5, $(M##B5), V)					\
	stf(T, 6, $(M##B6), V)					\
	stf(T, 7, $(M##B7), V)					\
	stf(T, 8, $(M##B8), V)					\
	stf(T, 9, $(M##B9), V)					\
	stf(T, 10, $(M##B10), V)				\
	stf(T, 11, $(M##B11), V)				\
	stf(T, 12, $(M##B12), V)				\
	stf(T, 13, $(M##B13), V)				\
	stf(T, 14, $(M##B14), V)				\
	stf(T, 15, $(M##B15), V)				\
	stf(T, 16, $(M##B16), V)				\
	stf(T, 17, $(M##B17), V)				\
	stf(T, 18, $(M##B18), V)				\
	stf(T, 19, $(M##B19), V)				\
	stf(T, 20, $(M##B20), V)				\
	stf(T, 21, $(M##B21), V)				\
	stf(T, 22, $(M##B22), V)				\
	stf(T, 23, $(M##B23), V)				\
	stf(T, 24, $(M##B24), V)				\
	stf(T, 25, $(M##B25), V)				\
	stf(T, 26, $(M##B26), V)
#define stf4(T, M, V)						\
	stf(T, 2, $(M##B2), V)					\
	stf8(T, M, V)
#define ldf(T, N, M, O, V)					\
	movi##T %f0 0						\
	ldxi##T %f0 %v0 O					\
	beqi##T ldf##T##N##M %f0 V				\
	calli @abort						\
ldf##T##N##M:
#define ldf8(T, M, V)						\
	ldf(T, 3, M, $(M##B3), V)				\
	ldf(T, 4, M, $(M##B4), V)				\
	ldf(T, 5, M, $(M##B5), V)				\
	ldf(T, 6, M, $(M##B6), V)				\
	ldf(T, 7, M, $(M##B7), V)				\
	ldf(T, 8, M, $(M##B8), V)				\
	ldf(T, 9, M, $(M##B9), V)				\
	ldf(T, 10, M, $(M##B10), V)				\
	ldf(T, 11, M, $(M##B11), V)				\
	ldf(T, 12, M, $(M##B12), V)				\
	ldf(T, 13, M, $(M##B13), V)				\
	ldf(T, 14, M, $(M##B14), V)				\
	ldf(T, 15, M, $(M##B15), V)				\
	ldf(T, 16, M, $(M##B16), V)				\
	ldf(T, 17, M, $(M##B17), V)				\
	ldf(T, 18, M, $(M##B18), V)				\
	ldf(T, 19, M, $(M##B19), V)				\
	ldf(T, 20, M, $(M##B20), V)				\
	ldf(T, 21, M, $(M##B21), V)				\
	ldf(T, 22, M, $(M##B22), V)				\
	ldf(T, 23, M, $(M##B23), V)				\
	ldf(T, 24, M, $(M##B24), V)				\
	ldf(T, 25, M, $(M##B25), V)				\
	ldf(T, 26, M, $(M##B26), V)
#define ldf4(T, M, V)						\
	ldf(T, 2, M, $(M##B2), V)				\
	ldf8(T, M, V)

#define ldst_c()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx2(_c, a, 0x5a)					\
	ldx2(_c, a, 0x5a)					\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stx2(_c, b, 0x5a)					\
	ldx2(_c, b, 0x5a)
#define ldst_uc()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx2(_c, a, 0x5a)					\
	ldx2(_uc, a, 0x5a)					\
	movi %v0 $(buf + M64)					\
	stx2(_c, b, 0x5a)					\
	ldx2(_uc, b, 0x5a)
#define ldst_s()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx2(_s, a, 0x5a5a)					\
	ldx2(_s, a, 0x5a5a)					\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stx2(_s, b, 0x5a5a)					\
	ldx2(_s, b, 0x5a5a)
#define ldst_us()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx2(_s, a, 0x5a5a)					\
	ldx2(_us, a, 0x5a5a)					\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stx2(_s, b, 0x5a5a)					\
	ldx2(_us, b, 0x5a5a)
#define ldst_i()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx4(_i, a, 0x5a5a5a5a)					\
	ldx4(_i, a, 0x5a5a5a5a)					\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stx4(_i, b, 0x5a5a5a5a)					\
	ldx4(_i, b, 0x5a5a5a5a)
#define ldst_ui()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx4(_i, a, 0x5a5a5a5a)					\
	ldx4(_ui, a, 0x5a5a5a5a)				\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stx4(_i, b, 0x5a5a5a5a)					\
	ldx4(_ui, b, 0x5a5a5a5a)
#define ldst_l()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stx8(_l, a, 0x5a5a5a5a5a5a5a5a)				\
	ldx8(_l, a, 0x5a5a5a5a5a5a5a5a)				\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stx8(_l, b, 0x5a5a5a5a5a5a5a5a)				\
	ldx8(_l, b, 0x5a5a5a5a5a5a5a5a)
#define ldst_f()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stf4(_f, a, 0.5)					\
	ldf4(_f, a, 0.5)					\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stf4(_f, b, 0.5)					\
	ldf4(_f, b, 0.5)
#define ldst_d()						\
	reset(0xa5)						\
	movi %v0 buf						\
	stf8(_d, a, 0.5)					\
	ldf8(_d, a, 0.5)					\
	reset(0xa5)						\
	movi %v0 $(buf + M64)					\
	stf8(_d, b, 0.5)					\
	ldf8(_d, b, 0.5)

.data		67112960
buf:
.size		M64
.size		8
ok:
.c		"ok"

.code
	prolog

	alu(add)
	alu(sub)
	alu(rsb)
	alu(mul)
	alu(div)
	alu(rem)
	lsh()
	rsh()
	alu(and)
	alu(or)
	alu(xor)
	ldst_c()
	ldst_uc()
	ldst_s()
	ldst_us()
	ldst_i()
#if __WORDSIZE == 64
	ldst_ui()
	ldst_l()
#endif
	ldst_f()
	ldst_d()

	prepare
		pushargi ok
	finishi @puts
	ret
	epilog
Commit	Line	Data
4a71579b PC	1	#define M64 67108864
	2
	3	#define aB1 (1<<1)
	4	#define aB2 (1<<2)
	5	#define aB3 (1<<3)
	6	#define aB4 (1<<4)
	7	#define aB5 (1<<5)
	8	#define aB6 (1<<6)
	9	#define aB7 (1<<7)
	10	#define aB8 (1<<8)
	11	#define aB9 (1<<9)
	12	#define aB10 (1<<10)
	13	#define aB11 (1<<11)
	14	#define aB12 (1<<12)
	15	#define aB13 (1<<13)
	16	#define aB14 (1<<14)
	17	#define aB15 (1<<15)
	18	#define aB16 (1<<16)
	19	#define aB17 (1<<17)
	20	#define aB18 (1<<18)
	21	#define aB19 (1<<19)
	22	#define aB20 (1<<20)
	23	#define aB21 (1<<21)
	24	#define aB22 (1<<22)
	25	#define aB23 (1<<23)
	26	#define aB24 (1<<24)
	27	#define aB25 (1<<25)
	28	#define aB26 (1<<26)
	29	#define bB1 (-aB1)
	30	#define bB2 (-aB2)
	31	#define bB3 (-aB3)
	32	#define bB4 (-aB4)
	33	#define bB5 (-aB5)
	34	#define bB6 (-aB6)
	35	#define bB7 (-aB7)
	36	#define bB8 (-aB8)
	37	#define bB9 (-aB9)
	38	#define bB10 (-aB10)
	39	#define bB11 (-aB11)
	40	#define bB12 (-aB12)
	41	#define bB13 (-aB13)
	42	#define bB14 (-aB14)
	43	#define bB15 (-aB15)
	44	#define bB16 (-aB16)
	45	#define bB17 (-aB17)
	46	#define bB18 (-aB18)
	47	#define bB19 (-aB19)
	48	#define bB20 (-aB20)
	49	#define bB21 (-aB21)
	50	#define bB22 (-aB22)
	51	#define bB23 (-aB23)
	52	#define bB24 (-aB24)
	53	#define bB25 (-aB25)
	54	#define bB26 (-aB26)
	55	#define cB1 (aB1-1)
	56	#define cB2 (aB2-1)
	57	#define cB3 (aB3-1)
	58	#define cB4 (aB4-1)
	59	#define cB5 (aB5-1)
	60	#define cB6 (aB6-1)
	61	#define cB7 (aB7-1)
	62	#define cB8 (aB8-1)
	63	#define cB9 (aB9-1)
	64	#define cB10 (aB10-1)
65	#define cB11 (aB11-1)
66	#define cB12 (aB12-1)
67	#define cB13 (aB13-1)
68	#define cB14 (aB14-1)
69	#define cB15 (aB15-1)
70	#define cB16 (aB16-1)
71	#define cB17 (aB17-1)
72	#define cB18 (aB18-1)
73	#define cB19 (aB19-1)
74	#define cB20 (aB20-1)
75	#define cB21 (aB21-1)
76	#define cB22 (aB22-1)
77	#define cB23 (aB23-1)
78	#define cB24 (aB24-1)
79	#define cB25 (aB25-1)
80	#define cB26 (aB26-1)
81	#define dB1 (-aB1+1)
82	#define dB2 (-aB2+1)
83	#define dB3 (-aB3+1)
84	#define dB4 (-aB4+1)
85	#define dB5 (-aB5+1)
86	#define dB6 (-aB6+1)
87	#define dB7 (-aB7+1)
88	#define dB8 (-aB8+1)
89	#define dB9 (-aB9+1)
90	#define dB10 (-aB10+1)
91	#define dB11 (-aB11+1)
92	#define dB12 (-aB12+1)
93	#define dB13 (-aB13+1)
94	#define dB14 (-aB14+1)
95	#define dB15 (-aB15+1)
96	#define dB16 (-aB16+1)
97	#define dB17 (-aB17+1)
98	#define dB18 (-aB18+1)
99	#define dB19 (-aB19+1)
100	#define dB20 (-aB20+1)
101	#define dB21 (-aB21+1)
102	#define dB22 (-aB22+1)
103	#define dB23 (-aB23+1)
104	#define dB24 (-aB24+1)
105	#define dB25 (-aB25+1)
106	#define dB26 (-aB26+1)
107
108	#define add(a, b) $(a + b)
109	#define sub(a, b) $(a - b)
110	#define rsb(a, b) $(b - a)
111	#define mul(a, b) $(a * b)
112	#define div(a, b) $(a / b)
113	#define rem(a, b) $(a % b)
114	#define and(a, b) $(a & b)
115	#define or(a, b) $(a \| b)
116	#define xor(a, b) $(a ^ b)
117
118	#define alu2(N, X, L, R, V) \
119	movi %r1 L \
120	N##i %r0 %r1 R \
121	beqi X %r0 V \
122	calli @abort \
123	X:
124	#define alu1(N, M) \
125	alu2(N, N##M##1, 3, $(M##1), N(3, M##1)) \
126	alu2(N, N##M##2, 3, $(M##2), N(3, M##2)) \
127	alu2(N, N##M##3, 3, $(M##3), N(3, M##3)) \
128	alu2(N, N##M##4, 3, $(M##4), N(3, M##4)) \
129	alu2(N, N##M##5, 3, $(M##5), N(3, M##5)) \
130	alu2(N, N##M##6, 3, $(M##6), N(3, M##6)) \
131	alu2(N, N##M##7, 3, $(M##7), N(3, M##7)) \
132	alu2(N, N##M##8, 3, $(M##8), N(3, M##8)) \
133	alu2(N, N##M##9, 3, $(M##9), N(3, M##9)) \
134	alu2(N, N##M##10, 3, $(M##10), N(3, M##10)) \
135	alu2(N, N##M##11, 3, $(M##11), N(3, M##11)) \
136	alu2(N, N##M##12, 3, $(M##12), N(3, M##12)) \
137	alu2(N, N##M##13, 3, $(M##13), N(3, M##13)) \
138	alu2(N, N##M##14, 3, $(M##14), N(3, M##14)) \
139	alu2(N, N##M##15, 3, $(M##15), N(3, M##15)) \
140	alu2(N, N##M##16, 3, $(M##16), N(3, M##16)) \
141	alu2(N, N##M##17, 3, $(M##17), N(3, M##17)) \
142	alu2(N, N##M##18, 3, $(M##18), N(3, M##18)) \
143	alu2(N, N##M##19, 3, $(M##19), N(3, M##19)) \
144	alu2(N, N##M##20, 3, $(M##20), N(3, M##20)) \
145	alu2(N, N##M##21, 3, $(M##21), N(3, M##21)) \
146	alu2(N, N##M##22, 3, $(M##22), N(3, M##22)) \
147	alu2(N, N##M##23, 3, $(M##23), N(3, M##23)) \
148	alu2(N, N##M##24, 3, $(M##24), N(3, M##24)) \
149	alu2(N, N##M##25, 3, $(M##25), N(3, M##25)) \
150	alu2(N, N##M##26, 3, $(M##26), N(3, M##26))
151
152	#define alu(N) \
153	alu1(N, aB) \
154	alu1(N, bB) \
155	alu1(N, cB) \
156	alu1(N, dB)
157
158	#define _lsh(N) \
159	alu2(lsh, L##N, 3, N, $(3<<N))
160	#define _rsh(N) \
161	alu2(rsh, R##N, $(1<<63), N, $((1<<63)>>N))
162
163	#if __WORDSIZE == 32
164	# define xsh64(X) /**/
165	#else
166	# define xsh64(X) \
167	_##X##sh(32) \
168	_##X##sh(33) \
169	_##X##sh(34) \
170	_##X##sh(35) \
171	_##X##sh(36) \
172	_##X##sh(37) \
173	_##X##sh(38) \
174	_##X##sh(39) \
175	_##X##sh(40) \
176	_##X##sh(41) \
177	_##X##sh(42) \
178	_##X##sh(43) \
179	_##X##sh(44) \
180	_##X##sh(45) \
181	_##X##sh(46) \
182	_##X##sh(47) \
183	_##X##sh(48) \
184	_##X##sh(49) \
185	_##X##sh(50) \
186	_##X##sh(51) \
187	_##X##sh(52) \
188	_##X##sh(53) \
189	_##X##sh(54) \
190	_##X##sh(55) \
191	_##X##sh(56) \
192	_##X##sh(57) \
193	_##X##sh(58) \
194	_##X##sh(59) \
195	_##X##sh(60) \
196	_##X##sh(61) \
197	_##X##sh(62) \
198	_##X##sh(63)
199	#endif
200
201	#define xsh(X) \
202	_##X##sh(0) \
203	_##X##sh(1) \
204	_##X##sh(2) \
205	_##X##sh(3) \
206	_##X##sh(4) \
207	_##X##sh(5) \
208	_##X##sh(6) \
209	_##X##sh(7) \
210	_##X##sh(8) \
211	_##X##sh(9) \
212	_##X##sh(10) \
213	_##X##sh(11) \
214	_##X##sh(12) \
215	_##X##sh(13) \
216	_##X##sh(14) \
217	_##X##sh(15) \
218	_##X##sh(16) \
219	_##X##sh(17) \
220	_##X##sh(18) \
221	_##X##sh(19) \
222	_##X##sh(20) \
223	_##X##sh(21) \
224	_##X##sh(22) \
225	_##X##sh(23) \
226	_##X##sh(24) \
227	_##X##sh(25) \
228	_##X##sh(26) \
229	_##X##sh(27) \
230	_##X##sh(28) \
231	_##X##sh(29) \
232	_##X##sh(30) \
233	_##X##sh(31) \
234	xsh64(X)
235
236	#define lsh() \
237	xsh(l)
238	#define rsh() \
239	xsh(r)
240
241	#define reset(V) \
242	prepare \
243	pushargi buf \
244	pushargi V \
245	pushargi $(M64 + 8) \
246	finishi @memset
247
248	#define stx(T, N, O, V) \
249	movi %r0 V \
250	stxi##T O %v0 %r0
251	#define stx8(T, M, V) \
252	stx(T, 3, $(M##B3), V) \
253	stx(T, 4, $(M##B4), V) \
254	stx(T, 5, $(M##B5), V) \
255	stx(T, 6, $(M##B6), V) \
256	stx(T, 7, $(M##B7), V) \
257	stx(T, 8, $(M##B8), V) \
258	stx(T, 9, $(M##B9), V) \
259	stx(T, 10, $(M##B10), V) \
260	stx(T, 11, $(M##B11), V) \
261	stx(T, 12, $(M##B12), V) \
262	stx(T, 13, $(M##B13), V) \
263	stx(T, 14, $(M##B14), V) \
264	stx(T, 15, $(M##B15), V) \
265	stx(T, 16, $(M##B16), V) \
266	stx(T, 17, $(M##B17), V) \
267	stx(T, 18, $(M##B18), V) \
268	stx(T, 19, $(M##B19), V) \
269	stx(T, 20, $(M##B20), V) \
270	stx(T, 21, $(M##B21), V) \
271	stx(T, 22, $(M##B22), V) \
272	stx(T, 23, $(M##B23), V) \
273	stx(T, 24, $(M##B24), V) \
274	stx(T, 25, $(M##B25), V) \
275	stx(T, 26, $(M##B26), V)
276	#define stx4(T, M, V) \
277	stx(T, 2, $(M##B2), V) \
278	stx8(T, M, V)
279	#define stx2(T, M, V) \
280	stx(T, 1, $(M##B1), V) \
281	stx4(T, M, V)
282	#define ldx(T, N, M, O, V) \
283	movi %r0 0 \
284	ldxi##T %r0 %v0 O \
285	beqi ldx##T##N##M %r0 V \
286	calli @abort \
287	ldx##T##N##M:
288	#define ldx8(T, M, V) \
289	ldx(T, 3, M, $(M##B3), V) \
290	ldx(T, 4, M, $(M##B4), V) \
291	ldx(T, 5, M, $(M##B5), V) \
292	ldx(T, 6, M, $(M##B6), V) \
293	ldx(T, 7, M, $(M##B7), V) \
294	ldx(T, 8, M, $(M##B8), V) \
295	ldx(T, 9, M, $(M##B9), V) \
296	ldx(T, 10, M, $(M##B10), V) \
297	ldx(T, 11, M, $(M##B11), V) \
298	ldx(T, 12, M, $(M##B12), V) \
299	ldx(T, 13, M, $(M##B13), V) \
300	ldx(T, 14, M, $(M##B14), V) \
301	ldx(T, 15, M, $(M##B15), V) \
302	ldx(T, 16, M, $(M##B16), V) \
303	ldx(T, 17, M, $(M##B17), V) \
304	ldx(T, 18, M, $(M##B18), V) \
305	ldx(T, 19, M, $(M##B19), V) \
306	ldx(T, 20, M, $(M##B20), V) \
307	ldx(T, 21, M, $(M##B21), V) \
308	ldx(T, 22, M, $(M##B22), V) \
309	ldx(T, 23, M, $(M##B23), V) \
310	ldx(T, 24, M, $(M##B24), V) \
311	ldx(T, 25, M, $(M##B25), V) \
312	ldx(T, 26, M, $(M##B26), V)
313	#define ldx4(T, M, V) \
314	ldx(T, 2, M, $(M##B2), V) \
315	ldx8(T, M, V)
316	#define ldx2(T, M, V) \
317	ldx(T, 1, M, $(M##B1), V) \
318	ldx4(T, M, V)
319
320	#define stf(T, N, O, V) \
321	movi##T %f0 V \
322	stxi##T O %v0 %f0
323	#define stf8(T, M, V) \
324	stf(T, 3, $(M##B3), V) \
325	stf(T, 4, $(M##B4), V) \
326	stf(T, 5, $(M##B5), V) \
327	stf(T, 6, $(M##B6), V) \
328	stf(T, 7, $(M##B7), V) \
329	stf(T, 8, $(M##B8), V) \
330	stf(T, 9, $(M##B9), V) \
331	stf(T, 10, $(M##B10), V) \
332	stf(T, 11, $(M##B11), V) \
333	stf(T, 12, $(M##B12), V) \
334	stf(T, 13, $(M##B13), V) \
335	stf(T, 14, $(M##B14), V) \
336	stf(T, 15, $(M##B15), V) \
337	stf(T, 16, $(M##B16), V) \
338	stf(T, 17, $(M##B17), V) \
339	stf(T, 18, $(M##B18), V) \
340	stf(T, 19, $(M##B19), V) \
341	stf(T, 20, $(M##B20), V) \
342	stf(T, 21, $(M##B21), V) \
343	stf(T, 22, $(M##B22), V) \
344	stf(T, 23, $(M##B23), V) \
345	stf(T, 24, $(M##B24), V) \
346	stf(T, 25, $(M##B25), V) \
347	stf(T, 26, $(M##B26), V)
348	#define stf4(T, M, V) \
349	stf(T, 2, $(M##B2), V) \
350	stf8(T, M, V)
351	#define ldf(T, N, M, O, V) \
352	movi##T %f0 0 \
353	ldxi##T %f0 %v0 O \
354	beqi##T ldf##T##N##M %f0 V \
355	calli @abort \
356	ldf##T##N##M:
357	#define ldf8(T, M, V) \
358	ldf(T, 3, M, $(M##B3), V) \
359	ldf(T, 4, M, $(M##B4), V) \
360	ldf(T, 5, M, $(M##B5), V) \
361	ldf(T, 6, M, $(M##B6), V) \
362	ldf(T, 7, M, $(M##B7), V) \
363	ldf(T, 8, M, $(M##B8), V) \
364	ldf(T, 9, M, $(M##B9), V) \
365	ldf(T, 10, M, $(M##B10), V) \
366	ldf(T, 11, M, $(M##B11), V) \
367	ldf(T, 12, M, $(M##B12), V) \
368	ldf(T, 13, M, $(M##B13), V) \
369	ldf(T, 14, M, $(M##B14), V) \
370	ldf(T, 15, M, $(M##B15), V) \
371	ldf(T, 16, M, $(M##B16), V) \
372	ldf(T, 17, M, $(M##B17), V) \
373	ldf(T, 18, M, $(M##B18), V) \
374	ldf(T, 19, M, $(M##B19), V) \
375	ldf(T, 20, M, $(M##B20), V) \
376	ldf(T, 21, M, $(M##B21), V) \
377	ldf(T, 22, M, $(M##B22), V) \
378	ldf(T, 23, M, $(M##B23), V) \
379	ldf(T, 24, M, $(M##B24), V) \
380	ldf(T, 25, M, $(M##B25), V) \
381	ldf(T, 26, M, $(M##B26), V)
382	#define ldf4(T, M, V) \
383	ldf(T, 2, M, $(M##B2), V) \
384	ldf8(T, M, V)
385
386	#define ldst_c() \
387	reset(0xa5) \
388	movi %v0 buf \
389	stx2(_c, a, 0x5a) \
390	ldx2(_c, a, 0x5a) \
391	reset(0xa5) \
392	movi %v0 $(buf + M64) \
393	stx2(_c, b, 0x5a) \
394	ldx2(_c, b, 0x5a)
395	#define ldst_uc() \
396	reset(0xa5) \
397	movi %v0 buf \
398	stx2(_c, a, 0x5a) \
399	ldx2(_uc, a, 0x5a) \
400	movi %v0 $(buf + M64) \
401	stx2(_c, b, 0x5a) \
402	ldx2(_uc, b, 0x5a)
403	#define ldst_s() \
404	reset(0xa5) \
405	movi %v0 buf \
406	stx2(_s, a, 0x5a5a) \
407	ldx2(_s, a, 0x5a5a) \
408	reset(0xa5) \
409	movi %v0 $(buf + M64) \
410	stx2(_s, b, 0x5a5a) \
411	ldx2(_s, b, 0x5a5a)
412	#define ldst_us() \
413	reset(0xa5) \
414	movi %v0 buf \
415	stx2(_s, a, 0x5a5a) \
416	ldx2(_us, a, 0x5a5a) \
417	reset(0xa5) \
418	movi %v0 $(buf + M64) \
419	stx2(_s, b, 0x5a5a) \
420	ldx2(_us, b, 0x5a5a)
421	#define ldst_i() \
422	reset(0xa5) \
423	movi %v0 buf \
424	stx4(_i, a, 0x5a5a5a5a) \
425	ldx4(_i, a, 0x5a5a5a5a) \
426	reset(0xa5) \
427	movi %v0 $(buf + M64) \
428	stx4(_i, b, 0x5a5a5a5a) \
429	ldx4(_i, b, 0x5a5a5a5a)
430	#define ldst_ui() \
431	reset(0xa5) \
432	movi %v0 buf \
433	stx4(_i, a, 0x5a5a5a5a) \
434	ldx4(_ui, a, 0x5a5a5a5a) \
435	reset(0xa5) \
436	movi %v0 $(buf + M64) \
437	stx4(_i, b, 0x5a5a5a5a) \
438	ldx4(_ui, b, 0x5a5a5a5a)
439	#define ldst_l() \
440	reset(0xa5) \
441	movi %v0 buf \
442	stx8(_l, a, 0x5a5a5a5a5a5a5a5a) \
443	ldx8(_l, a, 0x5a5a5a5a5a5a5a5a) \
444	reset(0xa5) \
445	movi %v0 $(buf + M64) \
446	stx8(_l, b, 0x5a5a5a5a5a5a5a5a) \
447	ldx8(_l, b, 0x5a5a5a5a5a5a5a5a)
448	#define ldst_f() \
449	reset(0xa5) \
450	movi %v0 buf \
451	stf4(_f, a, 0.5) \
452	ldf4(_f, a, 0.5) \
453	reset(0xa5) \
454	movi %v0 $(buf + M64) \
455	stf4(_f, b, 0.5) \
456	ldf4(_f, b, 0.5)
457	#define ldst_d() \
458	reset(0xa5) \
459	movi %v0 buf \
460	stf8(_d, a, 0.5) \
461	ldf8(_d, a, 0.5) \
462	reset(0xa5) \
463	movi %v0 $(buf + M64) \
464	stf8(_d, b, 0.5) \
465	ldf8(_d, b, 0.5)
466
467	.data 67112960
468	buf:
469	.size M64
470	.size 8
471	ok:
472	.c "ok"
473
474	.code
475	prolog
476
477	alu(add)
478	alu(sub)
479	alu(rsb)
480	alu(mul)
481	alu(div)
482	alu(rem)
483	lsh()
484	rsh()
485	alu(and)
486	alu(or)
487	alu(xor)
488	ldst_c()
489	ldst_uc()
490	ldst_s()
491	ldst_us()
492	ldst_i()
493	#if __WORDSIZE == 64
494	ldst_ui()
495	ldst_l()
496	#endif
497	ldst_f()
498	ldst_d()
499
500	prepare
501	pushargi ok
502	finishi @puts
503	ret
504	epilog