aix_ppc64.s

cryptlib安全工具包

	mulhdu		r8,r5,r6
	addc		r11,r7,r11
	adde		r9,r8,r9
	addze		r10,r0
	addc		r11,r7,r11
	adde		r9,r8,r9
	addze		r10,r10
						#sqr_add_c2(a,6,5,c3,c1,c2);
	ld		r5,40(r4)
	ld		r6,48(r4)
	mulld		r7,r5,r6
	mulhdu		r8,r5,r6
	addc		r11,r7,r11
	adde		r9,r8,r9
	addze		r10,r10
	addc		r11,r7,r11
	adde		r9,r8,r9
	addze		r10,r10
	std		r11,88(r3)	#r[11]=c3;
						#sqr_add_c(a,6,c1,c2,c3);
	mulld		r7,r6,r6
	mulhdu		r8,r6,r6
	addc		r9,r7,r9
	adde		r10,r8,r10
	addze		r11,r0
						#sqr_add_c2(a,7,5,c1,c2,c3)
	ld		r6,56(r4)
	mulld		r7,r5,r6
	mulhdu		r8,r5,r6
	addc		r9,r7,r9
	adde		r10,r8,r10
	addze		r11,r11
	addc		r9,r7,r9
	adde		r10,r8,r10
	addze		r11,r11
	std		r9,96(r3)	#r[12]=c1;
	
						#sqr_add_c2(a,7,6,c2,c3,c1)
	ld		r5,48(r4)
	mulld		r7,r5,r6
	mulhdu		r8,r5,r6
	addc		r10,r7,r10
	adde		r11,r8,r11
	addze		r9,r0
	addc		r10,r7,r10
	adde		r11,r8,r11
	addze		r9,r9
	std		r10,104(r3)	#r[13]=c2;
						#sqr_add_c(a,7,c3,c1,c2);
	mulld		r7,r6,r6
	mulhdu		r8,r6,r6
	addc		r11,r7,r11
	adde		r9,r8,r9
	std		r11,112(r3)	#r[14]=c3;
	std		r9, 120(r3)	#r[15]=c1;


	bclr	BO_ALWAYS,CR0_LT

	.long	0x00000000

#
#	NOTE:	The following label name should be changed to
#		"bn_mul_comba4" i.e. remove the first dot
#		for the gcc compiler. This should be automatically
#		done in the build
#

.align	4
.bn_mul_comba4:
#
# This is an optimized version of the bn_mul_comba4 routine.
#
# void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
# r3 contains r
# r4 contains a
# r5 contains b
# r6, r7 are the 2 BN_ULONGs being multiplied.
# r8, r9 are the results of the 32x32 giving 64 multiply.
# r10, r11, r12 are the equivalents of c1, c2, and c3.
#
	xor	r0,r0,r0		#r0=0. Used in addze below.
					#mul_add_c(a[0],b[0],c1,c2,c3);
	ld	r6,0(r4)		
	ld	r7,0(r5)		
	mulld	r10,r6,r7		
	mulhdu	r11,r6,r7		
	std	r10,0(r3)	#r[0]=c1
					#mul_add_c(a[0],b[1],c2,c3,c1);
	ld	r7,8(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r8,r11
	adde	r12,r9,r0
	addze	r10,r0
					#mul_add_c(a[1],b[0],c2,c3,c1);
	ld	r6, 8(r4)		
	ld	r7, 0(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r8,r11
	adde	r12,r9,r12
	addze	r10,r10
	std	r11,8(r3)	#r[1]=c2
					#mul_add_c(a[2],b[0],c3,c1,c2);
	ld	r6,16(r4)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r8,r12
	adde	r10,r9,r10
	addze	r11,r0
					#mul_add_c(a[1],b[1],c3,c1,c2);
	ld	r6,8(r4)		
	ld	r7,8(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r8,r12
	adde	r10,r9,r10
	addze	r11,r11
					#mul_add_c(a[0],b[2],c3,c1,c2);
	ld	r6,0(r4)		
	ld	r7,16(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r8,r12
	adde	r10,r9,r10
	addze	r11,r11
	std	r12,16(r3)	#r[2]=c3
					#mul_add_c(a[0],b[3],c1,c2,c3);
	ld	r7,24(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r8,r10
	adde	r11,r9,r11
	addze	r12,r0
					#mul_add_c(a[1],b[2],c1,c2,c3);
	ld	r6,8(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r8,r10
	adde	r11,r9,r11
	addze	r12,r12
					#mul_add_c(a[2],b[1],c1,c2,c3);
	ld	r6,16(r4)
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r8,r10
	adde	r11,r9,r11
	addze	r12,r12
					#mul_add_c(a[3],b[0],c1,c2,c3);
	ld	r6,24(r4)
	ld	r7,0(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r8,r10
	adde	r11,r9,r11
	addze	r12,r12
	std	r10,24(r3)	#r[3]=c1
					#mul_add_c(a[3],b[1],c2,c3,c1);
	ld	r7,8(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r8,r11
	adde	r12,r9,r12
	addze	r10,r0
					#mul_add_c(a[2],b[2],c2,c3,c1);
	ld	r6,16(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r8,r11
	adde	r12,r9,r12
	addze	r10,r10
					#mul_add_c(a[1],b[3],c2,c3,c1);
	ld	r6,8(r4)
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r8,r11
	adde	r12,r9,r12
	addze	r10,r10
	std	r11,32(r3)	#r[4]=c2
					#mul_add_c(a[2],b[3],c3,c1,c2);
	ld	r6,16(r4)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r8,r12
	adde	r10,r9,r10
	addze	r11,r0
					#mul_add_c(a[3],b[2],c3,c1,c2);
	ld	r6,24(r4)
	ld	r7,16(r4)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r8,r12
	adde	r10,r9,r10
	addze	r11,r11
	std	r12,40(r3)	#r[5]=c3
					#mul_add_c(a[3],b[3],c1,c2,c3);
	ld	r7,24(r5)		
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r8,r10
	adde	r11,r9,r11

	std	r10,48(r3)	#r[6]=c1
	std	r11,56(r3)	#r[7]=c2
	bclr	BO_ALWAYS,CR0_LT
	.long	0x00000000

#
#	NOTE:	The following label name should be changed to
#		"bn_mul_comba8" i.e. remove the first dot
#		for the gcc compiler. This should be automatically
#		done in the build
#
	
.align	4
.bn_mul_comba8:
#
# Optimized version of the bn_mul_comba8 routine.
#
# void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
# r3 contains r
# r4 contains a
# r5 contains b
# r6, r7 are the 2 BN_ULONGs being multiplied.
# r8, r9 are the results of the 32x32 giving 64 multiply.
# r10, r11, r12 are the equivalents of c1, c2, and c3.
#
	xor	r0,r0,r0		#r0=0. Used in addze below.
	
					#mul_add_c(a[0],b[0],c1,c2,c3);
	ld	r6,0(r4)	#a[0]
	ld	r7,0(r5)	#b[0]
	mulld	r10,r6,r7
	mulhdu	r11,r6,r7
	std	r10,0(r3)	#r[0]=c1;
					#mul_add_c(a[0],b[1],c2,c3,c1);
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	addze	r12,r9			# since we didnt set r12 to zero before.
	addze	r10,r0
					#mul_add_c(a[1],b[0],c2,c3,c1);
	ld	r6,8(r4)
	ld	r7,0(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
	std	r11,8(r3)	#r[1]=c2;
					#mul_add_c(a[2],b[0],c3,c1,c2);
	ld	r6,16(r4)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r0
					#mul_add_c(a[1],b[1],c3,c1,c2);
	ld	r6,8(r4)
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[0],b[2],c3,c1,c2);
	ld	r6,0(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
	std	r12,16(r3)	#r[2]=c3;
					#mul_add_c(a[0],b[3],c1,c2,c3);
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r0
					#mul_add_c(a[1],b[2],c1,c2,c3);
	ld	r6,8(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
		
					#mul_add_c(a[2],b[1],c1,c2,c3);
	ld	r6,16(r4)
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[3],b[0],c1,c2,c3);
	ld	r6,24(r4)
	ld	r7,0(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
	std	r10,24(r3)	#r[3]=c1;
					#mul_add_c(a[4],b[0],c2,c3,c1);
	ld	r6,32(r4)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r0
					#mul_add_c(a[3],b[1],c2,c3,c1);
	ld	r6,24(r4)
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[2],b[2],c2,c3,c1);
	ld	r6,16(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[1],b[3],c2,c3,c1);
	ld	r6,8(r4)
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[0],b[4],c2,c3,c1);
	ld	r6,0(r4)
	ld	r7,32(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
	std	r11,32(r3)	#r[4]=c2;
					#mul_add_c(a[0],b[5],c3,c1,c2);
	ld	r7,40(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r0
					#mul_add_c(a[1],b[4],c3,c1,c2);
	ld	r6,8(r4)		
	ld	r7,32(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[2],b[3],c3,c1,c2);
	ld	r6,16(r4)		
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[3],b[2],c3,c1,c2);
	ld	r6,24(r4)		
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[4],b[1],c3,c1,c2);
	ld	r6,32(r4)		
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[5],b[0],c3,c1,c2);
	ld	r6,40(r4)		
	ld	r7,0(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
	std	r12,40(r3)	#r[5]=c3;
					#mul_add_c(a[6],b[0],c1,c2,c3);
	ld	r6,48(r4)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r0
					#mul_add_c(a[5],b[1],c1,c2,c3);
	ld	r6,40(r4)
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[4],b[2],c1,c2,c3);
	ld	r6,32(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[3],b[3],c1,c2,c3);
	ld	r6,24(r4)
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[2],b[4],c1,c2,c3);
	ld	r6,16(r4)
	ld	r7,32(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[1],b[5],c1,c2,c3);
	ld	r6,8(r4)
	ld	r7,40(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[0],b[6],c1,c2,c3);
	ld	r6,0(r4)
	ld	r7,48(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
	std	r10,48(r3)	#r[6]=c1;
					#mul_add_c(a[0],b[7],c2,c3,c1);
	ld	r7,56(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r0
					#mul_add_c(a[1],b[6],c2,c3,c1);
	ld	r6,8(r4)
	ld	r7,48(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[2],b[5],c2,c3,c1);
	ld	r6,16(r4)
	ld	r7,40(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[3],b[4],c2,c3,c1);
	ld	r6,24(r4)
	ld	r7,32(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[4],b[3],c2,c3,c1);
	ld	r6,32(r4)
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[5],b[2],c2,c3,c1);
	ld	r6,40(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[6],b[1],c2,c3,c1);
	ld	r6,48(r4)
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
					#mul_add_c(a[7],b[0],c2,c3,c1);
	ld	r6,56(r4)
	ld	r7,0(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r11,r11,r8
	adde	r12,r12,r9
	addze	r10,r10
	std	r11,56(r3)	#r[7]=c2;
					#mul_add_c(a[7],b[1],c3,c1,c2);
	ld	r7,8(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r0
					#mul_add_c(a[6],b[2],c3,c1,c2);
	ld	r6,48(r4)
	ld	r7,16(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[5],b[3],c3,c1,c2);
	ld	r6,40(r4)
	ld	r7,24(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[4],b[4],c3,c1,c2);
	ld	r6,32(r4)
	ld	r7,32(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[3],b[5],c3,c1,c2);
	ld	r6,24(r4)
	ld	r7,40(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[2],b[6],c3,c1,c2);
	ld	r6,16(r4)
	ld	r7,48(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
					#mul_add_c(a[1],b[7],c3,c1,c2);
	ld	r6,8(r4)
	ld	r7,56(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r12,r12,r8
	adde	r10,r10,r9
	addze	r11,r11
	std	r12,64(r3)	#r[8]=c3;
					#mul_add_c(a[2],b[7],c1,c2,c3);
	ld	r6,16(r4)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r0
					#mul_add_c(a[3],b[6],c1,c2,c3);
	ld	r6,24(r4)
	ld	r7,48(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[4],b[5],c1,c2,c3);
	ld	r6,32(r4)
	ld	r7,40(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[5],b[4],c1,c2,c3);
	ld	r6,40(r4)
	ld	r7,32(r5)
	mulld	r8,r6,r7
	mulhdu	r9,r6,r7
	addc	r10,r10,r8
	adde	r11,r11,r9
	addze	r12,r12
					#mul_add_c(a[6],b[3],c1,c2,c3);
	ld	r6,48(r4)