-rw-r--r-- 4133 lib1305-20250415/crypto_onetimeauth/poly1305/amd64-mxaa-g4/poly1305_keypowers_asm.h raw
#ifndef __POLY1305KEYPOWERSASM__
#define __POLY1305KEYPOWERSASM__

/* field arithmetic used for computing the key powers */

#define fe1305_tau_square()				\
							\
	movq    %r14,%rdx;				\
    	mulx    %r15,%r9,%r10;				\
    	movq    $0,%r11;				\
    	shld    $1,%r10,%r11;				\
    	shld    $1,%r9,%r10;				\
    	shlq    $1,%r9;				\
  							\
      	xorq    %r12,%r12;				\
    	mulx    %rdx,%r8,%rbx;				\
    	addq    %rbx,%r9;				\
							\
    	movq    %r15,%rdx;				\
    	mulx    %rdx,%rax,%rbx;			\
    	adcq    %rax,%r10;				\
    	adcq    %rbx,%r11;				\
							\
	movq    %r10,%r13;				\
	andq    mask2(%rip),%r10;			\
	andq    mask2c(%rip),%r13;			\
							\
	addq    %r13,%r8;				\
	adcq    %r11,%r9;				\
	adcq    %r12,%r10;				\
							\
	shrd    $2,%r11,%r13;				\
	shrq    $2,%r11;				\
							\
	addq    %r13,%r8;				\
	adcq    %r11,%r9;				\
	adcq    %r12,%r10;				\
							\
	movq    %r10,%r13;				\
	andq    mask2(%rip),%r10;			\
	shrq    $2,%r13;				\
							\
	imul    $5,%r13,%r13;				\
	addq    %r13,%r8;				\
	adcq    $0,%r9;				\
	adcq    $0,%r10;				\
							\
	movq    %r8,24(%rdi);				\
	movq    %r9,32(%rdi);				\
	movq    %r10,40(%rdi);				\


#define fe1305_tau_squaren(x)				\
							\
	xorq    %r15,%r15;				\
    	movq    %r8,%rdx;				\
    							\
    	mulx    %r9,%r12,%r13;				\
							\
    	mulx    %r10,%rbx,%r14;			\
    	addq    %rbx,%r13;				\
    	adcq    %r15,%r14;				\
							\
    	xorq    %rax,%rax;				\
    	movq    %r9,%rdx;				\
    							\
    	mulx    %r10,%rbx,%rdx;			\
    	addq    %rbx,%r14;				\
    	adcq    %rdx,%r15;				\
    	adcq    %rax,%r15;				\
    							\
    	shld    $1,%r14,%r15;				\
    	shld    $1,%r13,%r14;				\
    	shld    $1,%r12,%r13;				\
    	shlq    $1,%r12;				\
     							\
    	xorq    %rdx,%rdx;				\
    	movq    %r8,%rdx;				\
    	mulx    %rdx,%r11,%rbx;			\
    	addq    %rbx,%r12;				\
							\
    	movq    %r9,%rdx;				\
    	mulx    %rdx,%r8,%r9;				\
    	adcq    %r13,%r8;				\
    	adcq    %r14,%r9;				\
							\
    	movq    %r10,%rdx;				\
    	mulx    %rdx,%rax,%rbx;			\
    	adcq    %rax,%r15;				\
							\
	movq    %r8,%r10;				\
	andq    mask2(%rip),%r10;			\
	andq    mask2c(%rip),%r8;			\
							\
	addq    %r8,%r11;				\
	adcq    %r9,%r12;				\
	adcq    %r15,%r10;				\
							\
	shrd    $2,%r9,%r8;				\
	shrd    $2,%r15,%r9;				\
	shrq    $2,%r15;				\
							\
	addq    %r11,%r8;				\
	adcq    %r12,%r9;				\
	adcq    %r15,%r10;				\
							\
	movq    %r10,%r15;				\
	andq    mask2(%rip),%r10;			\
	shrq    $2,%r15;				\
							\
	imul    $5,%r15,%r15;				\
	addq    %r15,%r8;				\
	adcq    $0,%r9;				\
	adcq    $0,%r10;				\
							\
	movq	%r8,24*x+0(%rdi);			\
	movq	%r9,24*x+8(%rdi);			\
	movq	%r10,24*x+16(%rdi);			\
	

#define fe1305_mul_tau_taun(x,y)			\
							\
	movq    24*x+0(%rdi),%rdx;			\
							\
	mulx    0(%rdi),%r8,%r9;			\
	mulx    8(%rdi),%rbx,%r10;			\
	addq    %rbx,%r9;				\
	adcq    $0,%r10;				\
							\
	movq    24*x+8(%rdi),%rdx;			\
	   						\
	mulx    0(%rdi),%rbx,%rbp;			\
	mulx    8(%rdi),%rdx,%r11;			\
	addq    %rdx,%rbp;				\
	adcq    $0,%r11;				\
	    						\
	addq    %rbx,%r9;				\
	adcq    %rbp,%r10;				\
	adcq    $0,%r11;				\
							\
	xorq    %rax,%rax;				\
	movq    24*x+16(%rdi),%rdx;			\
	    						\
	mulx    0(%rdi),%rbx,%rbp;			\
	mulx    8(%rdi),%rdx,%r12;			\
	addq    %rdx,%rbp;				\
	adcq    $0,%r12;				\
	    						\
	addq    %rbx,%r10;				\
	adcq    %rbp,%r11;				\
	adcq    %rax,%r12;				\
							\
	movq    %r10,%r13;				\
	andq    mask2(%rip),%r10;			\
	andq    mask2c(%rip),%r13;			\
							\
	addq    %r13,%r8;				\
	adcq    %r11,%r9;				\
	adcq    %r12,%r10;				\
							\
	shrd    $2,%r11,%r13;				\
	shrd    $2,%r12,%r11;				\
	shrq    $2,%r12;				\
							\
	addq    %r8,%r13;				\
	adcq    %r9,%r11;				\
	adcq    %r10,%r12;				\
							\
	movq    %r12,%rcx;				\
	andq    mask2(%rip),%rcx;			\
	shrq    $2,%r12;				\
							\
	imul    $5,%r12,%r12;				\
	addq    %r12,%r13;				\
	adcq    $0,%r11;				\
	adcq    $0,%rcx;				\
							\
	movq	%r13,24*y+0(%rdi);			\
	movq	%r11,24*y+8(%rdi);			\
	movq	%rcx,24*y+16(%rdi);			\
	
#endif