SHA512ROUND(V3, V0, V4, V2, V1, V25, V29, V17, V18, V16, V13, V14)
	SHA512ROUND(V2, V3, V1, V4, V0, V26, V30, V18, V19, V17, V14, V15)
	SHA512ROUND(V4, V2, V0, V1, V3, V27, V31, V19, V12, V18, V15, V16)
	SHA512ROUND(V1, V4, V3, V0, V2, V28, V24, V12, V13, V19, V16, V17)

	SHA512ROUND(V0, V1, V2, V3, V4, V29, V25, V13, V14, V12, V17, V18)
	SHA512ROUND(V3, V0, V4, V2, V1, V30, V26, V14, V15, V13, V18, V19)

	SHA1SU1	V4.S4, V5.S4
	VADD	V17.S4, V6.S4, V16.S4
	SHA1SU0	V4.S4, V7.S4, V6.S4
	HASHUPDATECHOOSE
	SHA1SU1	V5.S4, V6.S4

	VADD	V17.S4, V7.S4, V16.S4
	SHA1SU0	V5.S4, V4.S4, V7.S4
	HASHUPDATECHOOSE
	SHA1SU1	V6.S4, V7.S4

	VADD	V17.S4, V4.S4, V16.S4
	SHA1SU0	V6.S4, V5.S4, V4.S4
	HASHUPDATECHOOSE
	SHA1SU1	V7.S4, V4.S4

	VDUP	V19.S[2], V17.S4
	VADD	V18.S4, V5.S4, V16.S4
	SHA1SU0	V7.S4, V6.S4, V5.S4
	HASHUPDATEPARITY

	VADD	V10.S4, V14.S4, V10.S4
	VADD	V11.S4, V15.S4, V11.S4
	VEOR	V8.B16, V4.B16, V16.B16
	VEOR	V9.B16, V5.B16, V17.B16
	VEOR	V10.B16, V6.B16, V18.B16
	VEOR	V11.B16, V7.B16, V19.B16
	VSHL	$12, V16.S4, V4.S4
	VSHL	$12, V17.S4, V5.S4
	VSHL	$12, V18.S4, V6.S4
	VSHL	$12, V19.S4, V7.S4
	VSRI	$20, V16.S4, V4.S4
	VSRI	$20, V17.S4, V5.S4
	VSRI	$20, V18.S4, V6.S4
	VSRI	$20, V19.S4, V7.S4

	// V0..V3 += V4..V7

	LXVW4X (R0)(CONSTBASE), VS32
	LXVW4X (R8)(CONSTBASE), VS33
	LXVW4X (R9)(CONSTBASE), VS34
	LXVW4X (R10)(CONSTBASE), VS35

	// splat values from V17, V18 into V4-V11
	VSPLTW $0, V17, V4
	VSPLTW $1, V17, V5
	VSPLTW $2, V17, V6
	VSPLTW $3, V17, V7
	VSPLTW $0, V18, V8
	VSPLTW $1, V18, V9
	VSPLTW $2, V18, V10
	VSPLTW $3, V18, V11

	// VOR
	VOR V26, V26, V12

	CMPBGT	R8, $0, notaligned

aligned:
	ADD	R6, R4, R8
	MOVD	R8, R7
	AND	$-16, R7
	// replicate c across V17
	VLVGB	$0, R5, V19
	VREPB	$0, V19, V17

vectorloop:
	CMPBGE	R3, R7, residual
	VL	0(R3), V16    // load string to be searched into V16
	ADD	$16, R3
	VFEEBS	V16, V17, V18 // search V17 in V16 and set conditional code accordingly
	BVS	vectorloop

	// when vector search found c in the string

	MOVD	$1(R7), R7
	CMPBLE  R7, R2, index49to64loop
notfound:
	MOVD	$-1, (R5)
	RET

index65plus:
	// not implemented
	MOVD	$0, (R0)
	RET

foundV17: // index is in doubleword V17[0]
	VLGVG	$0, V17, R8
	ADD	R8, R7
found:
	SUB	R1, R7
	MOVD	R7, (R5)

// corresponding successor block directly. For example:
//
//	p:
//	  v11 = Less64 <bool> v10 v8
//	  If v11 goto b else u
//	b: <- p ...
//	  v17 = Leq64 <bool> v10 v8
//	  If v17 goto s else o
//
// We can redirect p to s directly.
//
// The implementation here borrows the framework of the prove pass.
//
//	1, Traverse all blocks of function f to find If blocks.

	MOVD	p_len+32(FP), R3                           // message length
	VLD1	(R0), [V0.S4, V1.S4]                       // load h(a,b,c,d,e,f,g,h)
	VLD1.P	64(R2), [V16.S4, V17.S4, V18.S4, V19.S4]
	VLD1.P	64(R2), [V20.S4, V21.S4, V22.S4, V23.S4]
	VLD1.P	64(R2), [V24.S4, V25.S4, V26.S4, V27.S4]
	VLD1	(R2), [V28.S4, V29.S4, V30.S4, V31.S4]     //load 64*4bytes K constant(K0-K63)

blockloop:

	AESIMC	V0.B16, V15.B16
	AESIMC	V1.B16, V14.B16
	AESIMC	V2.B16, V13.B16
	AESIMC	V3.B16, V12.B16
	VLD1	(R10), [V0.B16, V1.B16, V2.B16]
	AESIMC	V0.B16, V18.B16
	AESIMC	V1.B16, V17.B16
	VMOV	V2.B16, V16.B16
	VST1.P	[V16.B16, V17.B16, V18.B16], 48(R11)
	VST1.P	[V12.B16, V13.B16, V14.B16, V15.B16], 64(R11)
	VST1.P	[V8.B16, V9.B16, V10.B16, V11.B16], 64(R11)
	VST1	[V4.B16, V5.B16, V6.B16, V7.B16], (R11)
ksDone:

	F27
	F28
	F29
	F30
	F31
	V0 // VSX extension, F0 is V0[0:63].
	V1
	V2
	V3
	V4
	V5
	V6
	V7
	V8
	V9
	V10
	V11
	V12
	V13
	V14
	V15
	V16
	V17
	V18
	V19
	V20
	V21
	V22
	V23
	V24
	V25
	V26
	V27
	V28
	V29
	V30
	V31
	VS0
	VS1
	VS2
	VS3
	VS4
	VS5
	VS6
	VS7
	VS8
	VS9