-rw-r--r-- 4400 lib1305-20250407/crypto_onetimeauth/poly1305/amd64-maa44-g4/poly1305_asm.h raw
#ifndef __POLY1305ASM__
#define __POLY1305ASM__

/* field arithmetic used for computing poly1305 */

#define fe1305_mul_taun(m,t)				\
							\
	movq	m+0(%rsi),%r14;			\
	movq	m+8(%rsi),%r15;			\
							\
	movq    %r14,%rbp;				\
	andq    pmask1(%rip),%r14;			\
	movq    %r15,%rbx;				\
	andq    pmask2(%rip),%rbp;			\
	shrq    $44,%rbp;				\
	andq    pmask3(%rip),%r15;			\
	shlq    $20,%r15;				\
	orq     %rbp,%r15;				\
	andq    pmask4(%rip),%rbx;			\
	shrq    $24,%rbx;				\
	orq     twoexp41(%rip),%rbx;			\
							\
	movq	%r14,120(%rsp);			\
	movq	%r15,128(%rsp);			\
	movq	%rbx,136(%rsp);			\
							\
	movq    t+0(%rdi),%rax;			\
	mulq	120(%rsp);				\
	movq    %rax,%r14;				\
	movq    %rdx,%r15;				\
							\
	movq    t+8(%rdi),%rax;			\
	imul    $20,%rax,%rax;				\
	mulq	136(%rsp);				\
	addq    %rax,%r14;				\
	adcq    %rdx,%r15;				\
							\
	movq    t+16(%rdi),%rax;			\
	imul    $20,%rax,%rax;				\
	movq    %rax,144(%rsp);			\
	mulq	128(%rsp);				\
	addq    %rax,%r14;				\
	adcq    %rdx,%r15;				\
							\
	movq    t+0(%rdi),%rax;			\
	mulq	128(%rsp);				\
	movq    %rax,%rbx;				\
	movq    %rdx,%rbp;				\
							\
	movq    t+8(%rdi),%rax;			\
	mulq	120(%rsp);				\
	addq    %rax,%rbx;				\
	adcq    %rdx,%rbp;				\
							\
	movq    144(%rsp),%rax;			\
	mulq	136(%rsp);				\
	addq    %rax,%rbx;				\
	adcq    %rdx,%rbp;				\
							\
	movq    t+0(%rdi),%rax;			\
	mulq	136(%rsp);				\
	movq    %rax,%rcx;				\
	movq    %rdx,144(%rsp);			\
							\
	movq    t+8(%rdi),%rax;			\
	mulq	128(%rsp);				\
	addq    %rax,%rcx;				\
	adcq    144(%rsp),%rdx;			\
	movq    %rdx,144(%rsp);			\
							\
	movq    t+16(%rdi),%rax;			\
	mulq	120(%rsp);				\
	addq    %rax,%rcx;				\
	adcq    144(%rsp),%rdx;			\
		
	
#define fe1305_mul_taunr(t)				\
							\
	movq	%r8,%r14;				\
	movq	%r10,%r15;				\
	movq	%r12,%rbx;				\
							\
	movq    t+0(%rdi),%rax;			\
	mulq	%r14;					\
	movq    %rax,%r8;				\
	movq    %rdx,%r9;				\
							\
	movq    t+8(%rdi),%rax;			\
	imul    $20,%rax,%rax;				\
	mulq	%rbx;					\
	addq    %rax,%r8;				\
	adcq    %rdx,%r9;				\
							\
	movq    t+16(%rdi),%rax;			\
	imul    $20,%rax,%rax;				\
	movq    %rax,%rbp;				\
	mulq	%r15;					\
	addq    %rax,%r8;				\
	adcq    %rdx,%r9;				\
							\
	movq    t+0(%rdi),%rax;			\
	mulq	%r15;					\
	movq    %rax,%r10;				\
	movq    %rdx,%r11;				\
							\
	movq    t+8(%rdi),%rax;			\
	mulq	%r14;					\
	addq    %rax,%r10;				\
	adcq    %rdx,%r11;				\
							\
	movq    %rbp,%rax;				\
	mulq	%rbx;					\
	addq    %rax,%r10;				\
	adcq    %rdx,%r11;				\
							\
	movq    t+0(%rdi),%rax;			\
	mulq	%rbx;					\
	movq    %rax,%r12;				\
	movq    %rdx,%r13;				\
							\
	movq    t+8(%rdi),%rax;			\
	mulq	%r15;					\
	addq    %rax,%r12;				\
	adcq    %rdx,%r13;				\
							\
	movq    t+16(%rdi),%rax;			\
	mulq	%r14;					\
	addq    %rax,%r12;				\
	adcq    %rdx,%r13;				\
	
	
#define fe1305_add_product()				\
							\
	addq	%r14,%r8;				\
	adcq	%r15,%r9;				\
							\
	addq	%rbx,%r10;				\
	adcq	%rbp,%r11;				\
							\
	addq	%rcx,%r12;				\
	adcq	%rdx,%r13;				\


#define fe1305_add_msg_block(m)			\
							\
	movq    m+0(%rsi),%r14;			\
	movq    m+8(%rsi),%r15;			\
							\
	movq    %r14,%rbp;				\
	andq    pmask1(%rip),%r14;			\
	movq    %r15,%rbx;				\
	andq    pmask2(%rip),%rbp;			\
	shrq    $44,%rbp;				\
	andq    pmask3(%rip),%r15;			\
	shlq    $20,%r15;				\
	orq     %rbp,%r15;				\
	andq    pmask4(%rip),%rbx;			\
	shrq    $24,%rbx;				\
	movq    $1,%rbp;				\
	shlq    $40,%rbp;				\
	orq     %rbp,%rbx;				\
							\
	addq	%r14,%r8;				\
	addq	%r15,%r10;				\
	addq	%rbx,%r12;				\

#define fe1305_reduce_3l_128bits()			\
							\
	shld 	$20,%r8,%r9;				\
	shld 	$20,%r10,%r11;				\
	shld 	$22,%r12,%r13;				\
							\
	movq 	mask44(%rip),%rbp;			\
							\
	andq  	%rbp,%r8;				\
							\
	andq  	%rbp,%r10;				\
	addq  	%r9,%r10;				\
							\
	andq  	mask42(%rip),%r12;			\
	addq  	%r11,%r12;				\
							\
	imul    $5,%r13,%r13;				\
	addq	%r13,%r8;				\
	
	
#define fe1305_reduce_3l_64bits()			\
							\
	movq 	mask44(%rip),%rbp;			\
							\
	movq  	%r8,%rdx;				\
	shrq 	$44,%rdx;				\
	addq  	%r10,%rdx;				\
	andq  	%rbp,%r8;				\
							\
	movq  	%rdx,%r10;				\
	shrq 	$44,%rdx;				\
	addq  	%r12,%rdx;				\
	andq 	%rbp,%r10;				\
							\
	movq  	%rdx,%r12;				\
	shrq 	$42,%rdx;				\
	imul    $5,%rdx,%rdx;				\
	addq    %rdx,%r8;				\
	andq 	mask42(%rip),%r12;			\
	
#endif