MOVL R13, (5*4)(BP)
	ADDL (6*4)(BP), R14 // H6 = g + H6
	MOVL R14, (6*4)(BP)
	ADDL (7*4)(BP), R15 // H7 = h + H7
	MOVL R15, (7*4)(BP)

	ADDQ $64, SI
	CMPQ SI, 256(SP)
	JB   loop

end:
	RET

// shuffle byte order from LE to BE
DATA flip_mask<>+0x00(SB)/8, $0x0405060700010203
DATA flip_mask<>+0x08(SB)/8, $0x0c0d0e0f08090a0b
DATA flip_mask<>+0x10(SB)/8, $0x0405060700010203

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...

			instructions[s] = obj.As(i) + obj.ABaseAMD64
		}
	}
	// Annoying aliases.
	instructions["JA"] = x86.AJHI   /* alternate */
	instructions["JAE"] = x86.AJCC  /* alternate */
	instructions["JB"] = x86.AJCS   /* alternate */
	instructions["JBE"] = x86.AJLS  /* alternate */
	instructions["JC"] = x86.AJCS   /* alternate */
	instructions["JCC"] = x86.AJCC  /* carry clear (CF = 0) */

	/*5201*/ uint16(xSetOp), uint16(JB),
	/*5203*/ uint16(xReadCw),
	/*5204*/ uint16(xArgRel16),
	/*5205*/ uint16(xMatch),
	/*5206*/ uint16(xSetOp), uint16(JB),
	/*5208*/ uint16(xReadCd),
	/*5209*/ uint16(xArgRel32),
	/*5210*/ uint16(xMatch),
	/*5211*/ uint16(xCondDataSize), 5215, 5206, 5220,
	/*5215*/ uint16(xSetOp), uint16(JB),
	/*5217*/ uint16(xReadCd),
	/*5218*/ uint16(xArgRel32),

	ROUND4(CX, DX, BP, AX, BX, 78)
	ROUND4(BX, CX, DX, BP, AX, 79)

	ADDL	R11, AX
	ADDL	R12, BX
	ADDL	R13, CX
	ADDL	R14, DX
	ADDL	R15, BP

	ADDQ	$64, SI
	CMPQ	SI, DI
	JB	loop

end:
	MOVQ	dig+0(FP), DI
	MOVL	AX, (0*4)(DI)
	MOVL	BX, (1*4)(DI)
	MOVL	CX, (2*4)(DI)
	MOVL	DX, (3*4)(DI)
	MOVL	BP, (4*4)(DI)
	RET

y z I"},C:{"1":"2 3 4 5 6 7 8 9 AB BB CB DB EB FB GB HB IB JB KB LB MB NB OB PB QB RB SB TB UB VB WB XB YB ZB aB bB cB 0B dB 1B eB fB gB hB iB jB kB lB mB nB oB m pB qB rB sB tB P Q R 2B S T U V W X Y Z a b c d e f g h i j n o p q r s t u v w x y z I uB 3B 4B","2":"0 1 KC zB J K E F G A B C L M H N D O k l LC MC"},D:{"1":"8 9 AB BB CB DB EB FB GB HB IB JB KB LB MB NB OB PB QB RB SB TB UB VB WB XB YB ZB aB bB cB 0B dB 1B eB fB gB hB iB jB kB lB mB nB oB m pB qB rB sB tB P Q R S T U V W X Y Z a b c...

	MOVQ	R13, (5*8)(BP)
	ADDQ	(6*8)(BP), R14	// H6 = g + H6
	MOVQ	R14, (6*8)(BP)
	ADDQ	(7*8)(BP), R15	// H7 = h + H7
	MOVQ	R15, (7*8)(BP)

	ADDQ	$128, SI
	CMPQ	SI, 640(SP)
	JB	loop

end:
	RET

// Version below is based on "Fast SHA512 Implementations on Intel
// Architecture Processors" White-paper

��_�tQe�̲�\��u��u��K>+( ��Oǹ��p���2rQ8n\�URw�E]��YAa�E�ݢ.��熶uR-6O�2�����Խ�� m���0��0����")6�̲�\ԕ6\#u�W�'qՕ%}EM��%��iHH\�/ �v�����!Q��T{�j�w�c�ʆ���}Gm���S��� ��|��|�x�̧`��Jg���'wzbf�V���.��nmC8����bz��rfw�mv�T^��;��^*��ٝ��*�����X9T^鳫ϯ�Py�O�>�jB�>����K*�����W]Rw�Mnש��W�� ����#颙(yK��@gaO��Ң҇ �D.��own֡�>��%Zl���b�1�r�3r���������X��KW�I�Z��,�(M��^�����G����N�ٙ��v�X����>���{���py���G�m+ul����j���݇����/��כ���r��?��Ͽ�n����GD� >�J������6Bi��#�*�UG>k���KJTk�[\q����y\+��W���[��֢a�nek)...

,IAE5BnG,KAGTuhB,cAAA,WACE,OAAO9jB,KAAKojB,iBAAmBzY,OAC3B3K,KAAKojB,eAAemB,YAAcvkB,KAAKojB,eAAe9L,aAG5DyM,iBAAA,WACE,OAAO/jB,KAAKojB,eAAe1K,cAAgBjY,KAAK+jB,IAC9C7jB,SAASoT,KAAK2E,aACd/X,SAAS8C,gBAAgBiV,iBAI7B+L,iBAAA,WACE,OAAOzkB,KAAKojB,iBAAmBzY,OAC3BA,OAAO+Z,YAAc1kB,KAAKojB,eAAe1R,wBAAwByS,UAGvEV,SAAA,WACE,IAAMnM,EAAetX,KAAK8jB,gBAAkB9jB,KAAKoK,QAAQwI,OACnD8F,EAAe1Y,KAAK+jB,mBACpBY,EAAe3kB,KAAKoK,QAAQwI,OAChC8F,EACA1Y,KAAKykB,mBAMP,GAJIzkB,KAAKwjB,gBAAkB9K,GACzB1Y,KAAK0jB,UAGUiB,GAAbrN,EAAJ,CA...

xU���eD�H���'�Abd�D`����bu;-i}�jb{T��Hl�!��M�s�N�_7d�Iv�n�:7,�ڟf$���ᩇ�Y��}���AL�&���Z�V|�hy�1a���
#$��i�9	E�%��KW+	�·纋:�N�>�e:�x�#��)�7��0��{Q��IO>Y�$h�w�3՞�<�l�!򳽐^͵w6	��g�#�F	���[o'8�W���&W_st9�����r���p?{��Qa������������2V��z2
������+O�T���.U���K�y��U�<...