Remove unused assembler modules.

This commit is contained in:
Andy Polyakov 2005-02-06 13:43:02 +00:00
parent 67ea999d4a
commit da30c74a27
34 changed files with 0 additions and 10814 deletions

File diff suppressed because it is too large Load Diff

View File

@ -1,533 +0,0 @@
# DEC Alpha assember
# The bn_div64 is actually gcc output but the other parts are hand done.
# Thanks to tzeruch@ceddec.com for sending me the gcc output for
# bn_div64.
# I've gone back and re-done most of routines.
# The key thing to remeber for the 164 CPU is that while a
# multiply operation takes 8 cycles, another one can only be issued
# after 4 cycles have elapsed. I've done modification to help
# improve this. Also, normally, a ld instruction will not be available
# for about 3 cycles.
.file 1 "bn_asm.c"
.set noat
gcc2_compiled.:
__gnu_compiled_c:
.text
.align 3
.globl bn_mul_add_words
.ent bn_mul_add_words
bn_mul_add_words:
bn_mul_add_words..ng:
.frame $30,0,$26,0
.prologue 0
.align 5
subq $18,4,$18
bis $31,$31,$0
blt $18,$43 # if we are -1, -2, -3 or -4 goto tail code
ldq $20,0($17) # 1 1
ldq $1,0($16) # 1 1
.align 3
$42:
mulq $20,$19,$5 # 1 2 1 ######
ldq $21,8($17) # 2 1
ldq $2,8($16) # 2 1
umulh $20,$19,$20 # 1 2 ######
ldq $27,16($17) # 3 1
ldq $3,16($16) # 3 1
mulq $21,$19,$6 # 2 2 1 ######
ldq $28,24($17) # 4 1
addq $1,$5,$1 # 1 2 2
ldq $4,24($16) # 4 1
umulh $21,$19,$21 # 2 2 ######
cmpult $1,$5,$22 # 1 2 3 1
addq $20,$22,$20 # 1 3 1
addq $1,$0,$1 # 1 2 3 1
mulq $27,$19,$7 # 3 2 1 ######
cmpult $1,$0,$0 # 1 2 3 2
addq $2,$6,$2 # 2 2 2
addq $20,$0,$0 # 1 3 2
cmpult $2,$6,$23 # 2 2 3 1
addq $21,$23,$21 # 2 3 1
umulh $27,$19,$27 # 3 2 ######
addq $2,$0,$2 # 2 2 3 1
cmpult $2,$0,$0 # 2 2 3 2
subq $18,4,$18
mulq $28,$19,$8 # 4 2 1 ######
addq $21,$0,$0 # 2 3 2
addq $3,$7,$3 # 3 2 2
addq $16,32,$16
cmpult $3,$7,$24 # 3 2 3 1
stq $1,-32($16) # 1 2 4
umulh $28,$19,$28 # 4 2 ######
addq $27,$24,$27 # 3 3 1
addq $3,$0,$3 # 3 2 3 1
stq $2,-24($16) # 2 2 4
cmpult $3,$0,$0 # 3 2 3 2
stq $3,-16($16) # 3 2 4
addq $4,$8,$4 # 4 2 2
addq $27,$0,$0 # 3 3 2
cmpult $4,$8,$25 # 4 2 3 1
addq $17,32,$17
addq $28,$25,$28 # 4 3 1
addq $4,$0,$4 # 4 2 3 1
cmpult $4,$0,$0 # 4 2 3 2
stq $4,-8($16) # 4 2 4
addq $28,$0,$0 # 4 3 2
blt $18,$43
ldq $20,0($17) # 1 1
ldq $1,0($16) # 1 1
br $42
.align 4
$45:
ldq $20,0($17) # 4 1
ldq $1,0($16) # 4 1
mulq $20,$19,$5 # 4 2 1
subq $18,1,$18
addq $16,8,$16
addq $17,8,$17
umulh $20,$19,$20 # 4 2
addq $1,$5,$1 # 4 2 2
cmpult $1,$5,$22 # 4 2 3 1
addq $20,$22,$20 # 4 3 1
addq $1,$0,$1 # 4 2 3 1
cmpult $1,$0,$0 # 4 2 3 2
addq $20,$0,$0 # 4 3 2
stq $1,-8($16) # 4 2 4
bgt $18,$45
ret $31,($26),1 # else exit
.align 4
$43:
addq $18,4,$18
bgt $18,$45 # goto tail code
ret $31,($26),1 # else exit
.end bn_mul_add_words
.align 3
.globl bn_mul_words
.ent bn_mul_words
bn_mul_words:
bn_mul_words..ng:
.frame $30,0,$26,0
.prologue 0
.align 5
subq $18,4,$18
bis $31,$31,$0
blt $18,$143 # if we are -1, -2, -3 or -4 goto tail code
ldq $20,0($17) # 1 1
.align 3
$142:
mulq $20,$19,$5 # 1 2 1 #####
ldq $21,8($17) # 2 1
ldq $27,16($17) # 3 1
umulh $20,$19,$20 # 1 2 #####
ldq $28,24($17) # 4 1
mulq $21,$19,$6 # 2 2 1 #####
addq $5,$0,$5 # 1 2 3 1
subq $18,4,$18
cmpult $5,$0,$0 # 1 2 3 2
umulh $21,$19,$21 # 2 2 #####
addq $20,$0,$0 # 1 3 2
addq $17,32,$17
addq $6,$0,$6 # 2 2 3 1
mulq $27,$19,$7 # 3 2 1 #####
cmpult $6,$0,$0 # 2 2 3 2
addq $21,$0,$0 # 2 3 2
addq $16,32,$16
umulh $27,$19,$27 # 3 2 #####
stq $5,-32($16) # 1 2 4
mulq $28,$19,$8 # 4 2 1 #####
addq $7,$0,$7 # 3 2 3 1
stq $6,-24($16) # 2 2 4
cmpult $7,$0,$0 # 3 2 3 2
umulh $28,$19,$28 # 4 2 #####
addq $27,$0,$0 # 3 3 2
stq $7,-16($16) # 3 2 4
addq $8,$0,$8 # 4 2 3 1
cmpult $8,$0,$0 # 4 2 3 2
addq $28,$0,$0 # 4 3 2
stq $8,-8($16) # 4 2 4
blt $18,$143
ldq $20,0($17) # 1 1
br $142
.align 4
$145:
ldq $20,0($17) # 4 1
mulq $20,$19,$5 # 4 2 1
subq $18,1,$18
umulh $20,$19,$20 # 4 2
addq $5,$0,$5 # 4 2 3 1
addq $16,8,$16
cmpult $5,$0,$0 # 4 2 3 2
addq $17,8,$17
addq $20,$0,$0 # 4 3 2
stq $5,-8($16) # 4 2 4
bgt $18,$145
ret $31,($26),1 # else exit
.align 4
$143:
addq $18,4,$18
bgt $18,$145 # goto tail code
ret $31,($26),1 # else exit
.end bn_mul_words
.align 3
.globl bn_sqr_words
.ent bn_sqr_words
bn_sqr_words:
bn_sqr_words..ng:
.frame $30,0,$26,0
.prologue 0
subq $18,4,$18
blt $18,$543 # if we are -1, -2, -3 or -4 goto tail code
ldq $20,0($17) # 1 1
.align 3
$542:
mulq $20,$20,$5 ######
ldq $21,8($17) # 1 1
subq $18,4
umulh $20,$20,$1 ######
ldq $27,16($17) # 1 1
mulq $21,$21,$6 ######
ldq $28,24($17) # 1 1
stq $5,0($16) # r[0]
umulh $21,$21,$2 ######
stq $1,8($16) # r[1]
mulq $27,$27,$7 ######
stq $6,16($16) # r[0]
umulh $27,$27,$3 ######
stq $2,24($16) # r[1]
mulq $28,$28,$8 ######
stq $7,32($16) # r[0]
umulh $28,$28,$4 ######
stq $3,40($16) # r[1]
addq $16,64,$16
addq $17,32,$17
stq $8,-16($16) # r[0]
stq $4,-8($16) # r[1]
blt $18,$543
ldq $20,0($17) # 1 1
br $542
$442:
ldq $20,0($17) # a[0]
mulq $20,$20,$5 # a[0]*w low part r2
addq $16,16,$16
addq $17,8,$17
subq $18,1,$18
umulh $20,$20,$1 # a[0]*w high part r3
stq $5,-16($16) # r[0]
stq $1,-8($16) # r[1]
bgt $18,$442
ret $31,($26),1 # else exit
.align 4
$543:
addq $18,4,$18
bgt $18,$442 # goto tail code
ret $31,($26),1 # else exit
.end bn_sqr_words
.align 3
.globl bn_add_words
.ent bn_add_words
bn_add_words:
bn_add_words..ng:
.frame $30,0,$26,0
.prologue 0
subq $19,4,$19
bis $31,$31,$0 # carry = 0
blt $19,$900
ldq $5,0($17) # a[0]
ldq $1,0($18) # b[1]
.align 3
$901:
addq $1,$5,$1 # r=a+b;
ldq $6,8($17) # a[1]
cmpult $1,$5,$22 # did we overflow?
ldq $2,8($18) # b[1]
addq $1,$0,$1 # c+= overflow
ldq $7,16($17) # a[2]
cmpult $1,$0,$0 # overflow?
ldq $3,16($18) # b[2]
addq $0,$22,$0
ldq $8,24($17) # a[3]
addq $2,$6,$2 # r=a+b;
ldq $4,24($18) # b[3]
cmpult $2,$6,$23 # did we overflow?
addq $3,$7,$3 # r=a+b;
addq $2,$0,$2 # c+= overflow
cmpult $3,$7,$24 # did we overflow?
cmpult $2,$0,$0 # overflow?
addq $4,$8,$4 # r=a+b;
addq $0,$23,$0
cmpult $4,$8,$25 # did we overflow?
addq $3,$0,$3 # c+= overflow
stq $1,0($16) # r[0]=c
cmpult $3,$0,$0 # overflow?
stq $2,8($16) # r[1]=c
addq $0,$24,$0
stq $3,16($16) # r[2]=c
addq $4,$0,$4 # c+= overflow
subq $19,4,$19 # loop--
cmpult $4,$0,$0 # overflow?
addq $17,32,$17 # a++
addq $0,$25,$0
stq $4,24($16) # r[3]=c
addq $18,32,$18 # b++
addq $16,32,$16 # r++
blt $19,$900
ldq $5,0($17) # a[0]
ldq $1,0($18) # b[1]
br $901
.align 4
$945:
ldq $5,0($17) # a[0]
ldq $1,0($18) # b[1]
addq $1,$5,$1 # r=a+b;
subq $19,1,$19 # loop--
addq $1,$0,$1 # c+= overflow
addq $17,8,$17 # a++
cmpult $1,$5,$22 # did we overflow?
cmpult $1,$0,$0 # overflow?
addq $18,8,$18 # b++
stq $1,0($16) # r[0]=c
addq $0,$22,$0
addq $16,8,$16 # r++
bgt $19,$945
ret $31,($26),1 # else exit
$900:
addq $19,4,$19
bgt $19,$945 # goto tail code
ret $31,($26),1 # else exit
.end bn_add_words
#
# What follows was taken directly from the C compiler with a few
# hacks to redo the lables.
#
.text
.align 3
.globl bn_div64
.ent bn_div64
bn_div64:
ldgp $29,0($27)
bn_div64..ng:
lda $30,-48($30)
.frame $30,48,$26,0
stq $26,0($30)
stq $9,8($30)
stq $10,16($30)
stq $11,24($30)
stq $12,32($30)
stq $13,40($30)
.mask 0x4003e00,-48
.prologue 1
bis $16,$16,$9
bis $17,$17,$10
bis $18,$18,$11
bis $31,$31,$13
bis $31,2,$12
bne $11,$119
lda $0,-1
br $31,$136
.align 4
$119:
bis $11,$11,$16
jsr $26,BN_num_bits_word
ldgp $29,0($26)
subq $0,64,$1
beq $1,$120
bis $31,1,$1
sll $1,$0,$1
cmpule $9,$1,$1
bne $1,$120
# lda $16,_IO_stderr_
# lda $17,$C32
# bis $0,$0,$18
# jsr $26,fprintf
# ldgp $29,0($26)
jsr $26,abort
ldgp $29,0($26)
.align 4
$120:
bis $31,64,$3
cmpult $9,$11,$2
subq $3,$0,$1
addl $1,$31,$0
subq $9,$11,$1
cmoveq $2,$1,$9
beq $0,$122
zapnot $0,15,$2
subq $3,$0,$1
sll $11,$2,$11
sll $9,$2,$3
srl $10,$1,$1
sll $10,$2,$10
bis $3,$1,$9
$122:
srl $11,32,$5
zapnot $11,15,$6
lda $7,-1
.align 5
$123:
srl $9,32,$1
subq $1,$5,$1
bne $1,$126
zapnot $7,15,$27
br $31,$127
.align 4
$126:
bis $9,$9,$24
bis $5,$5,$25
divqu $24,$25,$27
$127:
srl $10,32,$4
.align 5
$128:
mulq $27,$5,$1
subq $9,$1,$3
zapnot $3,240,$1
bne $1,$129
mulq $6,$27,$2
sll $3,32,$1
addq $1,$4,$1
cmpule $2,$1,$2
bne $2,$129
subq $27,1,$27
br $31,$128
.align 4
$129:
mulq $27,$6,$1
mulq $27,$5,$4
srl $1,32,$3
sll $1,32,$1
addq $4,$3,$4
cmpult $10,$1,$2
subq $10,$1,$10
addq $2,$4,$2
cmpult $9,$2,$1
bis $2,$2,$4
beq $1,$134
addq $9,$11,$9
subq $27,1,$27
$134:
subl $12,1,$12
subq $9,$4,$9
beq $12,$124
sll $27,32,$13
sll $9,32,$2
srl $10,32,$1
sll $10,32,$10
bis $2,$1,$9
br $31,$123
.align 4
$124:
bis $13,$27,$0
$136:
ldq $26,0($30)
ldq $9,8($30)
ldq $10,16($30)
ldq $11,24($30)
ldq $12,32($30)
ldq $13,40($30)
addq $30,48,$30
ret $31,($26),1
.end bn_div64
.set noat
.text
.align 3
.globl bn_sub_words
.ent bn_sub_words
bn_sub_words:
bn_sub_words..ng:
.frame $30,0,$26,0
.prologue 0
subq $19, 4, $19
bis $31, $31, $0
blt $19, $100
ldq $1, 0($17)
ldq $2, 0($18)
$101:
ldq $3, 8($17)
cmpult $1, $2, $4
ldq $5, 8($18)
subq $1, $2, $1
ldq $6, 16($17)
cmpult $1, $0, $2
ldq $7, 16($18)
subq $1, $0, $23
ldq $8, 24($17)
addq $2, $4, $0
cmpult $3, $5, $24
subq $3, $5, $3
ldq $22, 24($18)
cmpult $3, $0, $5
subq $3, $0, $25
addq $5, $24, $0
cmpult $6, $7, $27
subq $6, $7, $6
stq $23, 0($16)
cmpult $6, $0, $7
subq $6, $0, $28
addq $7, $27, $0
cmpult $8, $22, $21
subq $8, $22, $8
stq $25, 8($16)
cmpult $8, $0, $22
subq $8, $0, $20
addq $22, $21, $0
stq $28, 16($16)
subq $19, 4, $19
stq $20, 24($16)
addq $17, 32, $17
addq $18, 32, $18
addq $16, 32, $16
blt $19, $100
ldq $1, 0($17)
ldq $2, 0($18)
br $101
$102:
ldq $1, 0($17)
ldq $2, 0($18)
cmpult $1, $2, $27
subq $1, $2, $1
cmpult $1, $0, $2
subq $1, $0, $1
stq $1, 0($16)
addq $2, $27, $0
addq $17, 8, $17
addq $18, 8, $18
addq $16, 8, $16
subq $19, 1, $19
bgt $19, $102
ret $31,($26),1
$100:
addq $19, 4, $19
bgt $19, $102
$103:
ret $31,($26),1
.end bn_sub_words

View File

@ -1,119 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_add_words
{
local($name)=@_;
local($cc,$a,$b,$r);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
$count=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&br(&label("finish"));
&blt($count,&label("finish"));
($a0,$b0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
##########################################################
&set_label("loop");
($a1)=&NR(1); &ld($a1,&QWPw(1,$ap));
($b1)=&NR(1); &ld($b1,&QWPw(1,$bp));
($a2)=&NR(1); &ld($a2,&QWPw(2,$ap));
($b2)=&NR(1); &ld($b2,&QWPw(2,$bp));
($a3)=&NR(1); &ld($a3,&QWPw(3,$ap));
($b3)=&NR(1); &ld($b3,&QWPw(3,$bp));
($o0,$t0)=&NR(2);
&add($a0,$b0,$o0);
&cmpult($o0,$b0,$t0);
&add($o0,$cc,$o0);
&cmpult($o0,$cc,$cc);
&add($cc,$t0,$cc); &FR($t0);
($t1,$o1)=&NR(2);
&add($a1,$b1,$o1); &FR($a1);
&cmpult($o1,$b1,$t1); &FR($b1);
&add($o1,$cc,$o1);
&cmpult($o1,$cc,$cc);
&add($cc,$t1,$cc); &FR($t1);
($t2,$o2)=&NR(2);
&add($a2,$b2,$o2); &FR($a2);
&cmpult($o2,$b2,$t2); &FR($b2);
&add($o2,$cc,$o2);
&cmpult($o2,$cc,$cc);
&add($cc,$t2,$cc); &FR($t2);
($t3,$o3)=&NR(2);
&add($a3,$b3,$o3); &FR($a3);
&cmpult($o3,$b3,$t3); &FR($b3);
&add($o3,$cc,$o3);
&cmpult($o3,$cc,$cc);
&add($cc,$t3,$cc); &FR($t3);
&st($o0,&QWPw(0,$rp)); &FR($o0);
&st($o1,&QWPw(0,$rp)); &FR($o1);
&st($o2,&QWPw(0,$rp)); &FR($o2);
&st($o3,&QWPw(0,$rp)); &FR($o3);
&sub($count,4,$count); # count-=4
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
##################################################
# Do the last 0..3 words
($t0,$o0)=&NR(2);
&set_label("last_loop");
&ld($a0,&QWPw(0,$ap)); # get a
&ld($b0,&QWPw(0,$bp)); # get b
&add($a0,$b0,$o0);
&cmpult($o0,$b0,$t0); # will we borrow?
&add($o0,$cc,$o0); # will we borrow?
&cmpult($o0,$cc,$cc); # will we borrow?
&add($cc,$t0,$cc); # add the borrows
&st($o0,&QWPw(0,$rp)); # save
&add($ap,$QWS,$ap);
&add($bp,$QWS,$bp);
&add($rp,$QWS,$rp);
&sub($count,1,$count);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&FR($o0,$t0,$a0,$b0);
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,144 +0,0 @@
#!/usr/local/bin/perl
sub bn_div64
{
local($data)=<<'EOF';
#
# What follows was taken directly from the C compiler with a few
# hacks to redo the lables.
#
.text
.set noreorder
.set volatile
.align 3
.globl bn_div64
.ent bn_div64
bn_div64:
ldgp $29,0($27)
bn_div64..ng:
lda $30,-48($30)
.frame $30,48,$26,0
stq $26,0($30)
stq $9,8($30)
stq $10,16($30)
stq $11,24($30)
stq $12,32($30)
stq $13,40($30)
.mask 0x4003e00,-48
.prologue 1
bis $16,$16,$9
bis $17,$17,$10
bis $18,$18,$11
bis $31,$31,$13
bis $31,2,$12
bne $11,$9119
lda $0,-1
br $31,$9136
.align 4
$9119:
bis $11,$11,$16
jsr $26,BN_num_bits_word
ldgp $29,0($26)
subq $0,64,$1
beq $1,$9120
bis $31,1,$1
sll $1,$0,$1
cmpule $9,$1,$1
bne $1,$9120
# lda $16,_IO_stderr_
# lda $17,$C32
# bis $0,$0,$18
# jsr $26,fprintf
# ldgp $29,0($26)
jsr $26,abort
ldgp $29,0($26)
.align 4
$9120:
bis $31,64,$3
cmpult $9,$11,$2
subq $3,$0,$1
addl $1,$31,$0
subq $9,$11,$1
cmoveq $2,$1,$9
beq $0,$9122
zapnot $0,15,$2
subq $3,$0,$1
sll $11,$2,$11
sll $9,$2,$3
srl $10,$1,$1
sll $10,$2,$10
bis $3,$1,$9
$9122:
srl $11,32,$5
zapnot $11,15,$6
lda $7,-1
.align 5
$9123:
srl $9,32,$1
subq $1,$5,$1
bne $1,$9126
zapnot $7,15,$27
br $31,$9127
.align 4
$9126:
bis $9,$9,$24
bis $5,$5,$25
divqu $24,$25,$27
$9127:
srl $10,32,$4
.align 5
$9128:
mulq $27,$5,$1
subq $9,$1,$3
zapnot $3,240,$1
bne $1,$9129
mulq $6,$27,$2
sll $3,32,$1
addq $1,$4,$1
cmpule $2,$1,$2
bne $2,$9129
subq $27,1,$27
br $31,$9128
.align 4
$9129:
mulq $27,$6,$1
mulq $27,$5,$4
srl $1,32,$3
sll $1,32,$1
addq $4,$3,$4
cmpult $10,$1,$2
subq $10,$1,$10
addq $2,$4,$2
cmpult $9,$2,$1
bis $2,$2,$4
beq $1,$9134
addq $9,$11,$9
subq $27,1,$27
$9134:
subl $12,1,$12
subq $9,$4,$9
beq $12,$9124
sll $27,32,$13
sll $9,32,$2
srl $10,32,$1
sll $10,32,$10
bis $2,$1,$9
br $31,$9123
.align 4
$9124:
bis $13,$27,$0
$9136:
ldq $26,0($30)
ldq $9,8($30)
ldq $10,16($30)
ldq $11,24($30)
ldq $12,32($30)
ldq $13,40($30)
addq $30,48,$30
ret $31,($26),1
.end bn_div64
EOF
&asm_add($data);
}
1;

View File

@ -1,116 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_mul_words
{
local($name)=@_;
local($cc,$a,$b,$r,$couny);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$count=&wparam(2);
$word=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&br(&label("finish"));
&blt($count,&label("finish"));
($a0,$r0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($r0,&QWPw(0,$rp));
$a=<<'EOF';
##########################################################
&set_label("loop");
($a1)=&NR(1); &ld($a1,&QWPw(1,$ap));
($b1)=&NR(1); &ld($b1,&QWPw(1,$bp));
($a2)=&NR(1); &ld($a2,&QWPw(2,$ap));
($b2)=&NR(1); &ld($b2,&QWPw(2,$bp));
($a3)=&NR(1); &ld($a3,&QWPw(3,$ap));
($b3)=&NR(1); &ld($b3,&QWPw(3,$bp));
($o0,$t0)=&NR(2);
&add($a0,$b0,$o0);
&cmpult($o0,$b0,$t0);
&add($o0,$cc,$o0);
&cmpult($o0,$cc,$cc);
&add($cc,$t0,$cc); &FR($t0);
($t1,$o1)=&NR(2);
&add($a1,$b1,$o1); &FR($a1);
&cmpult($o1,$b1,$t1); &FR($b1);
&add($o1,$cc,$o1);
&cmpult($o1,$cc,$cc);
&add($cc,$t1,$cc); &FR($t1);
($t2,$o2)=&NR(2);
&add($a2,$b2,$o2); &FR($a2);
&cmpult($o2,$b2,$t2); &FR($b2);
&add($o2,$cc,$o2);
&cmpult($o2,$cc,$cc);
&add($cc,$t2,$cc); &FR($t2);
($t3,$o3)=&NR(2);
&add($a3,$b3,$o3); &FR($a3);
&cmpult($o3,$b3,$t3); &FR($b3);
&add($o3,$cc,$o3);
&cmpult($o3,$cc,$cc);
&add($cc,$t3,$cc); &FR($t3);
&st($o0,&QWPw(0,$rp)); &FR($o0);
&st($o1,&QWPw(0,$rp)); &FR($o1);
&st($o2,&QWPw(0,$rp)); &FR($o2);
&st($o3,&QWPw(0,$rp)); &FR($o3);
&sub($count,4,$count); # count-=4
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
EOF
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap)); # get a
&mul($a0,$word,($l0)=&NR(1));
&add($ap,$QWS,$ap);
&muh($a0,$word,($h0)=&NR(1)); &FR($a0);
&add($l0,$cc,$l0);
&add($rp,$QWS,$rp);
&sub($count,1,$count);
&cmpult($l0,$cc,$cc);
&st($l0,&QWPw(-1,$rp)); &FR($l0);
&add($h0,$cc,$cc); &FR($h0);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,120 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_mul_add_words
{
local($name)=@_;
local($cc,$a,$b,$r,$couny);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$count=&wparam(2);
$word=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&br(&label("finish"));
&blt($count,&label("finish"));
($a0,$r0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($r0,&QWPw(0,$rp));
$a=<<'EOF';
##########################################################
&set_label("loop");
($a1)=&NR(1); &ld($a1,&QWPw(1,$ap));
($b1)=&NR(1); &ld($b1,&QWPw(1,$bp));
($a2)=&NR(1); &ld($a2,&QWPw(2,$ap));
($b2)=&NR(1); &ld($b2,&QWPw(2,$bp));
($a3)=&NR(1); &ld($a3,&QWPw(3,$ap));
($b3)=&NR(1); &ld($b3,&QWPw(3,$bp));
($o0,$t0)=&NR(2);
&add($a0,$b0,$o0);
&cmpult($o0,$b0,$t0);
&add($o0,$cc,$o0);
&cmpult($o0,$cc,$cc);
&add($cc,$t0,$cc); &FR($t0);
($t1,$o1)=&NR(2);
&add($a1,$b1,$o1); &FR($a1);
&cmpult($o1,$b1,$t1); &FR($b1);
&add($o1,$cc,$o1);
&cmpult($o1,$cc,$cc);
&add($cc,$t1,$cc); &FR($t1);
($t2,$o2)=&NR(2);
&add($a2,$b2,$o2); &FR($a2);
&cmpult($o2,$b2,$t2); &FR($b2);
&add($o2,$cc,$o2);
&cmpult($o2,$cc,$cc);
&add($cc,$t2,$cc); &FR($t2);
($t3,$o3)=&NR(2);
&add($a3,$b3,$o3); &FR($a3);
&cmpult($o3,$b3,$t3); &FR($b3);
&add($o3,$cc,$o3);
&cmpult($o3,$cc,$cc);
&add($cc,$t3,$cc); &FR($t3);
&st($o0,&QWPw(0,$rp)); &FR($o0);
&st($o1,&QWPw(0,$rp)); &FR($o1);
&st($o2,&QWPw(0,$rp)); &FR($o2);
&st($o3,&QWPw(0,$rp)); &FR($o3);
&sub($count,4,$count); # count-=4
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
EOF
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap)); # get a
&ld(($r0)=&NR(1),&QWPw(0,$rp)); # get b
&mul($a0,$word,($l0)=&NR(1));
&sub($count,1,$count);
&add($ap,$QWS,$ap);
&muh($a0,$word,($h0)=&NR(1)); &FR($a0);
&add($r0,$l0,$r0);
&add($rp,$QWS,$rp);
&cmpult($r0,$l0,($t0)=&NR(1)); &FR($l0);
&add($r0,$cc,$r0);
&add($h0,$t0,$h0); &FR($t0);
&cmpult($r0,$cc,$cc);
&st($r0,&QWPw(-1,$rp)); &FR($r0);
&add($h0,$cc,$cc); &FR($h0);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,213 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub mul_add_c
{
local($a,$b,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
&mul($a,$b,($l1)=&NR(1));
&muh($a,$b,($h1)=&NR(1));
&add($c0,$l1,$c0);
&cmpult($c0,$l1,($t1)=&NR(1)); &FR($l1);
&add($t1,$h1,$h1); &FR($t1);
&add($c1,$h1,$c1);
&cmpult($c1,$h1,($t2)=&NR(1)); &FR($h1);
&add($c2,$t2,$c2); &FR($t2);
}
sub bn_mul_comba4
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(3);
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($b[0])=&NR(1),&QWPw(0,$bp));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($b[1])=&NR(1),&QWPw(1,$bp));
&mul($a[0],$b[0],($r00)=&NR(1));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($b[2])=&NR(1),&QWPw(2,$bp));
&muh($a[0],$b[0],($r01)=&NR(1));
&FR($ap); &ld(($a[3])=&NR(1),&QWPw(3,$ap));
&FR($bp); &ld(($b[3])=&NR(1),&QWPw(3,$bp));
&mul($a[0],$b[1],($r02)=&NR(1));
($R,$H1,$H2)=&NR(3);
&st($r00,&QWPw(0,$rp)); &FR($r00);
&mov("zero",$R);
&mul($a[1],$b[0],($r03)=&NR(1));
&mov("zero",$H1);
&mov("zero",$H0);
&add($R,$r01,$R);
&muh($a[0],$b[1],($r04)=&NR(1));
&cmpult($R,$r01,($t01)=&NR(1)); &FR($r01);
&add($R,$r02,$R);
&add($H1,$t01,$H1) &FR($t01);
&muh($a[1],$b[0],($r05)=&NR(1));
&cmpult($R,$r02,($t02)=&NR(1)); &FR($r02);
&add($R,$r03,$R);
&add($H2,$t02,$H2) &FR($t02);
&mul($a[0],$b[2],($r06)=&NR(1));
&cmpult($R,$r03,($t03)=&NR(1)); &FR($r03);
&add($H1,$t03,$H1) &FR($t03);
&st($R,&QWPw(1,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r04,$R);
&mov("zero",$H2);
&mul($a[1],$b[1],($r07)=&NR(1));
&cmpult($R,$r04,($t04)=&NR(1)); &FR($r04);
&add($R,$r05,$R);
&add($H1,$t04,$H1) &FR($t04);
&mul($a[2],$b[0],($r08)=&NR(1));
&cmpult($R,$r05,($t05)=&NR(1)); &FR($r05);
&add($R,$r01,$R);
&add($H2,$t05,$H2) &FR($t05);
&muh($a[0],$b[2],($r09)=&NR(1));
&cmpult($R,$r06,($t06)=&NR(1)); &FR($r06);
&add($R,$r07,$R);
&add($H1,$t06,$H1) &FR($t06);
&muh($a[1],$b[1],($r10)=&NR(1));
&cmpult($R,$r07,($t07)=&NR(1)); &FR($r07);
&add($R,$r08,$R);
&add($H2,$t07,$H2) &FR($t07);
&muh($a[2],$b[0],($r11)=&NR(1));
&cmpult($R,$r08,($t08)=&NR(1)); &FR($r08);
&add($H1,$t08,$H1) &FR($t08);
&st($R,&QWPw(2,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r09,$R);
&mov("zero",$H2);
&mul($a[0],$b[3],($r12)=&NR(1));
&cmpult($R,$r09,($t09)=&NR(1)); &FR($r09);
&add($R,$r10,$R);
&add($H1,$t09,$H1) &FR($t09);
&mul($a[1],$b[2],($r13)=&NR(1));
&cmpult($R,$r10,($t10)=&NR(1)); &FR($r10);
&add($R,$r11,$R);
&add($H1,$t10,$H1) &FR($t10);
&mul($a[2],$b[1],($r14)=&NR(1));
&cmpult($R,$r11,($t11)=&NR(1)); &FR($r11);
&add($R,$r12,$R);
&add($H1,$t11,$H1) &FR($t11);
&mul($a[3],$b[0],($r15)=&NR(1));
&cmpult($R,$r12,($t12)=&NR(1)); &FR($r12);
&add($R,$r13,$R);
&add($H1,$t12,$H1) &FR($t12);
&muh($a[0],$b[3],($r16)=&NR(1));
&cmpult($R,$r13,($t13)=&NR(1)); &FR($r13);
&add($R,$r14,$R);
&add($H1,$t13,$H1) &FR($t13);
&muh($a[1],$b[2],($r17)=&NR(1));
&cmpult($R,$r14,($t14)=&NR(1)); &FR($r14);
&add($R,$r15,$R);
&add($H1,$t14,$H1) &FR($t14);
&muh($a[2],$b[1],($r18)=&NR(1));
&cmpult($R,$r15,($t15)=&NR(1)); &FR($r15);
&add($H1,$t15,$H1) &FR($t15);
&st($R,&QWPw(3,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r16,$R);
&mov("zero",$H2);
&muh($a[3],$b[0],($r19)=&NR(1));
&cmpult($R,$r16,($t16)=&NR(1)); &FR($r16);
&add($R,$r17,$R);
&add($H1,$t16,$H1) &FR($t16);
&mul($a[1],$b[3],($r20)=&NR(1));
&cmpult($R,$r17,($t17)=&NR(1)); &FR($r17);
&add($R,$r18,$R);
&add($H1,$t17,$H1) &FR($t17);
&mul($a[2],$b[2],($r21)=&NR(1));
&cmpult($R,$r18,($t18)=&NR(1)); &FR($r18);
&add($R,$r19,$R);
&add($H1,$t18,$H1) &FR($t18);
&mul($a[3],$b[1],($r22)=&NR(1));
&cmpult($R,$r19,($t19)=&NR(1)); &FR($r19);
&add($R,$r20,$R);
&add($H1,$t19,$H1) &FR($t19);
&muh($a[1],$b[3],($r23)=&NR(1));
&cmpult($R,$r20,($t20)=&NR(1)); &FR($r20);
&add($R,$r21,$R);
&add($H1,$t20,$H1) &FR($t20);
&muh($a[2],$b[2],($r24)=&NR(1));
&cmpult($R,$r21,($t21)=&NR(1)); &FR($r21);
&add($R,$r22,$R);
&add($H1,$t21,$H1) &FR($t21);
&muh($a[3],$b[1],($r25)=&NR(1));
&cmpult($R,$r22,($t22)=&NR(1)); &FR($r22);
&add($H1,$t22,$H1) &FR($t22);
&st($R,&QWPw(4,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r23,$R);
&mov("zero",$H2);
&mul($a[2],$b[3],($r26)=&NR(1));
&cmpult($R,$r23,($t23)=&NR(1)); &FR($r23);
&add($R,$r24,$R);
&add($H1,$t23,$H1) &FR($t23);
&mul($a[3],$b[2],($r27)=&NR(1));
&cmpult($R,$r24,($t24)=&NR(1)); &FR($r24);
&add($R,$r25,$R);
&add($H1,$t24,$H1) &FR($t24);
&muh($a[2],$b[3],($r28)=&NR(1));
&cmpult($R,$r25,($t25)=&NR(1)); &FR($r25);
&add($R,$r26,$R);
&add($H1,$t25,$H1) &FR($t25);
&muh($a[3],$b[2],($r29)=&NR(1));
&cmpult($R,$r26,($t26)=&NR(1)); &FR($r26);
&add($R,$r27,$R);
&add($H1,$t26,$H1) &FR($t26);
&mul($a[3],$b[3],($r30)=&NR(1));
&cmpult($R,$r27,($t27)=&NR(1)); &FR($r27);
&add($H1,$t27,$H1) &FR($t27);
&st($R,&QWPw(5,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r28,$R);
&mov("zero",$H2);
&muh($a[3],$b[3],($r31)=&NR(1));
&cmpult($R,$r28,($t28)=&NR(1)); &FR($r28);
&add($R,$r29,$R);
&add($H1,$t28,$H1) &FR($t28);
############
&cmpult($R,$r29,($t29)=&NR(1)); &FR($r29);
&add($R,$r30,$R);
&add($H1,$t29,$H1) &FR($t29);
############
&cmpult($R,$r30,($t30)=&NR(1)); &FR($r30);
&add($H1,$t30,$H1) &FR($t30);
&st($R,&QWPw(6,$rp));
&add($H1,$H2,$R);
&add($R,$r31,$R); &FR($r31);
&st($R,&QWPw(7,$rp));
&FR($R,$H1,$H2);
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,98 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub mul_add_c
{
local($a,$b,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
print STDERR "count=$cnt\n"; $cnt++;
&mul($a,$b,($l1)=&NR(1));
&muh($a,$b,($h1)=&NR(1));
&add($c0,$l1,$c0);
&cmpult($c0,$l1,($t1)=&NR(1)); &FR($l1);
&add($t1,$h1,$h1); &FR($t1);
&add($c1,$h1,$c1);
&cmpult($c1,$h1,($t2)=&NR(1)); &FR($h1);
&add($c2,$t2,$c2); &FR($t2);
}
sub bn_mul_comba4
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(3);
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($b[0])=&NR(1),&QWPw(0,$bp));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($b[1])=&NR(1),&QWPw(1,$bp));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($b[2])=&NR(1),&QWPw(2,$bp));
&ld(($a[3])=&NR(1),&QWPw(3,$ap)); &FR($ap);
&ld(($b[3])=&NR(1),&QWPw(3,$bp)); &FR($bp);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$b[0],$c0);
&muh($a[0],$b[0],$c1);
&st($c0,&QWPw(0,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[1],$c0,$c1,$c2);
&mul_add_c($a[1],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[1],$b[1],$c0,$c1,$c2);
&mul_add_c($a[0],$b[2],$c0,$c1,$c2);
&mul_add_c($a[2],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[3],$c0,$c1,$c2); &FR($a[0]);
&mul_add_c($a[1],$b[2],$c0,$c1,$c2);
&mul_add_c($a[2],$b[1],$c0,$c1,$c2);
&mul_add_c($a[3],$b[0],$c0,$c1,$c2); &FR($b[0]);
&st($c0,&QWPw(3,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[1],$b[3],$c0,$c1,$c2); &FR($a[1]);
&mul_add_c($a[2],$b[2],$c0,$c1,$c2);
&mul_add_c($a[3],$b[1],$c0,$c1,$c2); &FR($b[1]);
&st($c0,&QWPw(4,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[2],$b[3],$c0,$c1,$c2); &FR($a[2]);
&mul_add_c($a[3],$b[2],$c0,$c1,$c2); &FR($b[2]);
&st($c0,&QWPw(5,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[3],$b[3],$c0,$c1,$c2); &FR($a[3],$b[3]);
&st($c0,&QWPw(6,$rp));
&st($c1,&QWPw(7,$rp));
&FR($c0,$c1,$c2);
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,177 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_mul_comba8
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(3);
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
&function_begin($name,"");
&comment("");
&stack_push(2);
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($b[0])=&NR(1),&QWPw(0,$bp));
&st($reg_s0,&swtmp(0)); &FR($reg_s0);
&st($reg_s1,&swtmp(1)); &FR($reg_s1);
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($b[1])=&NR(1),&QWPw(1,$bp));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($b[2])=&NR(1),&QWPw(2,$bp));
&ld(($a[3])=&NR(1),&QWPw(3,$ap));
&ld(($b[3])=&NR(1),&QWPw(3,$bp));
&ld(($a[4])=&NR(1),&QWPw(1,$ap));
&ld(($b[4])=&NR(1),&QWPw(1,$bp));
&ld(($a[5])=&NR(1),&QWPw(1,$ap));
&ld(($b[5])=&NR(1),&QWPw(1,$bp));
&ld(($a[6])=&NR(1),&QWPw(1,$ap));
&ld(($b[6])=&NR(1),&QWPw(1,$bp));
&ld(($a[7])=&NR(1),&QWPw(1,$ap)); &FR($ap);
&ld(($b[7])=&NR(1),&QWPw(1,$bp)); &FR($bp);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$b[0],$c0);
&muh($a[0],$b[0],$c1);
&st($c0,&QWPw(0,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[1],$c0,$c1,$c2);
&mul_add_c($a[1],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[2],$c0,$c1,$c2);
&mul_add_c($a[1],$b[1],$c0,$c1,$c2);
&mul_add_c($a[2],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[3],$c0,$c1,$c2);
&mul_add_c($a[1],$b[2],$c0,$c1,$c2);
&mul_add_c($a[2],$b[1],$c0,$c1,$c2);
&mul_add_c($a[3],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(3,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[4],$c0,$c1,$c2);
&mul_add_c($a[1],$b[3],$c0,$c1,$c2);
&mul_add_c($a[2],$b[2],$c0,$c1,$c2);
&mul_add_c($a[3],$b[1],$c0,$c1,$c2);
&mul_add_c($a[4],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(4,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[5],$c0,$c1,$c2);
&mul_add_c($a[1],$b[4],$c0,$c1,$c2);
&mul_add_c($a[2],$b[3],$c0,$c1,$c2);
&mul_add_c($a[3],$b[2],$c0,$c1,$c2);
&mul_add_c($a[4],$b[1],$c0,$c1,$c2);
&mul_add_c($a[5],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(5,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[6],$c0,$c1,$c2);
&mul_add_c($a[1],$b[5],$c0,$c1,$c2);
&mul_add_c($a[2],$b[4],$c0,$c1,$c2);
&mul_add_c($a[3],$b[3],$c0,$c1,$c2);
&mul_add_c($a[4],$b[2],$c0,$c1,$c2);
&mul_add_c($a[5],$b[1],$c0,$c1,$c2);
&mul_add_c($a[6],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(6,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[7],$c0,$c1,$c2); &FR($a[0]);
&mul_add_c($a[1],$b[6],$c0,$c1,$c2);
&mul_add_c($a[2],$b[5],$c0,$c1,$c2);
&mul_add_c($a[3],$b[4],$c0,$c1,$c2);
&mul_add_c($a[4],$b[3],$c0,$c1,$c2);
&mul_add_c($a[5],$b[2],$c0,$c1,$c2);
&mul_add_c($a[6],$b[1],$c0,$c1,$c2);
&mul_add_c($a[7],$b[0],$c0,$c1,$c2); &FR($b[0]);
&st($c0,&QWPw(7,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[1],$b[7],$c0,$c1,$c2); &FR($a[1]);
&mul_add_c($a[2],$b[6],$c0,$c1,$c2);
&mul_add_c($a[3],$b[5],$c0,$c1,$c2);
&mul_add_c($a[4],$b[4],$c0,$c1,$c2);
&mul_add_c($a[5],$b[3],$c0,$c1,$c2);
&mul_add_c($a[6],$b[2],$c0,$c1,$c2);
&mul_add_c($a[7],$b[1],$c0,$c1,$c2); &FR($b[1]);
&st($c0,&QWPw(8,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[2],$b[7],$c0,$c1,$c2); &FR($a[2]);
&mul_add_c($a[3],$b[6],$c0,$c1,$c2);
&mul_add_c($a[4],$b[5],$c0,$c1,$c2);
&mul_add_c($a[5],$b[4],$c0,$c1,$c2);
&mul_add_c($a[6],$b[3],$c0,$c1,$c2);
&mul_add_c($a[7],$b[2],$c0,$c1,$c2); &FR($b[2]);
&st($c0,&QWPw(9,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[3],$b[7],$c0,$c1,$c2); &FR($a[3]);
&mul_add_c($a[4],$b[6],$c0,$c1,$c2);
&mul_add_c($a[5],$b[5],$c0,$c1,$c2);
&mul_add_c($a[6],$b[4],$c0,$c1,$c2);
&mul_add_c($a[7],$b[3],$c0,$c1,$c2); &FR($b[3]);
&st($c0,&QWPw(10,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[4],$b[7],$c0,$c1,$c2); &FR($a[4]);
&mul_add_c($a[5],$b[6],$c0,$c1,$c2);
&mul_add_c($a[6],$b[5],$c0,$c1,$c2);
&mul_add_c($a[7],$b[4],$c0,$c1,$c2); &FR($b[4]);
&st($c0,&QWPw(11,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[5],$b[7],$c0,$c1,$c2); &FR($a[5]);
&mul_add_c($a[6],$b[6],$c0,$c1,$c2);
&mul_add_c($a[7],$b[5],$c0,$c1,$c2); &FR($b[5]);
&st($c0,&QWPw(12,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[6],$b[7],$c0,$c1,$c2); &FR($a[6]);
&mul_add_c($a[7],$b[6],$c0,$c1,$c2); &FR($b[6]);
&st($c0,&QWPw(13,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[7],$b[7],$c0,$c1,$c2); &FR($a[7],$b[7]);
&st($c0,&QWPw(14,$rp));
&st($c1,&QWPw(15,$rp));
&FR($c0,$c1,$c2);
&ld($reg_s0,&swtmp(0));
&ld($reg_s1,&swtmp(1));
&stack_pop(2);
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,113 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_sqr_words
{
local($name)=@_;
local($cc,$a,$b,$r,$couny);
&init_pool(3);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$count=&wparam(2);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&br(&label("finish"));
&blt($count,&label("finish"));
($a0,$r0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($r0,&QWPw(0,$rp));
$a=<<'EOF';
##########################################################
&set_label("loop");
($a1)=&NR(1); &ld($a1,&QWPw(1,$ap));
($b1)=&NR(1); &ld($b1,&QWPw(1,$bp));
($a2)=&NR(1); &ld($a2,&QWPw(2,$ap));
($b2)=&NR(1); &ld($b2,&QWPw(2,$bp));
($a3)=&NR(1); &ld($a3,&QWPw(3,$ap));
($b3)=&NR(1); &ld($b3,&QWPw(3,$bp));
($o0,$t0)=&NR(2);
&add($a0,$b0,$o0);
&cmpult($o0,$b0,$t0);
&add($o0,$cc,$o0);
&cmpult($o0,$cc,$cc);
&add($cc,$t0,$cc); &FR($t0);
($t1,$o1)=&NR(2);
&add($a1,$b1,$o1); &FR($a1);
&cmpult($o1,$b1,$t1); &FR($b1);
&add($o1,$cc,$o1);
&cmpult($o1,$cc,$cc);
&add($cc,$t1,$cc); &FR($t1);
($t2,$o2)=&NR(2);
&add($a2,$b2,$o2); &FR($a2);
&cmpult($o2,$b2,$t2); &FR($b2);
&add($o2,$cc,$o2);
&cmpult($o2,$cc,$cc);
&add($cc,$t2,$cc); &FR($t2);
($t3,$o3)=&NR(2);
&add($a3,$b3,$o3); &FR($a3);
&cmpult($o3,$b3,$t3); &FR($b3);
&add($o3,$cc,$o3);
&cmpult($o3,$cc,$cc);
&add($cc,$t3,$cc); &FR($t3);
&st($o0,&QWPw(0,$rp)); &FR($o0);
&st($o1,&QWPw(0,$rp)); &FR($o1);
&st($o2,&QWPw(0,$rp)); &FR($o2);
&st($o3,&QWPw(0,$rp)); &FR($o3);
&sub($count,4,$count); # count-=4
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
EOF
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap)); # get a
&mul($a0,$a0,($l0)=&NR(1));
&add($ap,$QWS,$ap);
&add($rp,2*$QWS,$rp);
&sub($count,1,$count);
&muh($a0,$a0,($h0)=&NR(1)); &FR($a0);
&st($l0,&QWPw(-2,$rp)); &FR($l0);
&st($h0,&QWPw(-1,$rp)); &FR($h0);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,109 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub sqr_add_c
{
local($a,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
&mul($a,$a,($l1)=&NR(1));
&muh($a,$a,($h1)=&NR(1));
&add($c0,$l1,$c0);
&add($c1,$h1,$c1);
&cmpult($c0,$l1,($t1)=&NR(1)); &FR($l1);
&cmpult($c1,$h1,($t2)=&NR(1)); &FR($h1);
&add($c1,$t1,$c1); &FR($t1);
&add($c2,$t2,$c2); &FR($t2);
}
sub sqr_add_c2
{
local($a,$b,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
&mul($a,$b,($l1)=&NR(1));
&muh($a,$b,($h1)=&NR(1));
&cmplt($l1,"zero",($lc1)=&NR(1));
&cmplt($h1,"zero",($hc1)=&NR(1));
&add($l1,$l1,$l1);
&add($h1,$h1,$h1);
&add($h1,$lc1,$h1); &FR($lc1);
&add($c2,$hc1,$c2); &FR($hc1);
&add($c0,$l1,$c0);
&add($c1,$h1,$c1);
&cmpult($c0,$l1,($lc1)=&NR(1)); &FR($l1);
&cmpult($c1,$h1,($hc1)=&NR(1)); &FR($h1);
&add($c1,$lc1,$c1); &FR($lc1);
&add($c2,$hc1,$c2); &FR($hc1);
}
sub bn_sqr_comba4
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(2);
$rp=&wparam(0);
$ap=&wparam(1);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($a[3])=&NR(1),&QWPw(3,$ap)); &FR($ap);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$a[0],$c0);
&muh($a[0],$a[0],$c1);
&st($c0,&QWPw(0,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[0],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[1],$c0,$c1,$c2);
&sqr_add_c2($a[2],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[3],$a[0],$c0,$c1,$c2);
&sqr_add_c2($a[2],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(3,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[2],$c0,$c1,$c2);
&sqr_add_c2($a[3],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(4,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[3],$a[2],$c0,$c1,$c2);
&st($c0,&QWPw(5,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[3],$c0,$c1,$c2);
&st($c0,&QWPw(6,$rp));
&st($c1,&QWPw(7,$rp));
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,132 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_sqr_comba8
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(2);
$rp=&wparam(0);
$ap=&wparam(1);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($a[3])=&NR(1),&QWPw(3,$ap));
&ld(($a[4])=&NR(1),&QWPw(4,$ap));
&ld(($a[5])=&NR(1),&QWPw(5,$ap));
&ld(($a[6])=&NR(1),&QWPw(6,$ap));
&ld(($a[7])=&NR(1),&QWPw(7,$ap)); &FR($ap);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$a[0],$c0);
&muh($a[0],$a[0],$c1);
&st($c0,&QWPw(0,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[1],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[1],$c0,$c1,$c2);
&sqr_add_c2($a[2],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[2],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[3],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(3,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[2],$c0,$c1,$c2);
&sqr_add_c2($a[3],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[4],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(4,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[3],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[4],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(5,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[3],$c0,$c1,$c2);
&sqr_add_c2($a[4],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(6,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[4],$a[3],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(7,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[4],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[3],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(8,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[5],$a[4],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[3],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[2],$c0,$c1,$c2);
&st($c0,&QWPw(9,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[5],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[4],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[3],$c0,$c1,$c2);
&st($c0,&QWPw(10,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[6],$a[5],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[4],$c0,$c1,$c2);
&st($c0,&QWPw(11,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[6],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[5],$c0,$c1,$c2);
&st($c0,&QWPw(12,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[7],$a[6],$c0,$c1,$c2);
&st($c0,&QWPw(13,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[7],$c0,$c1,$c2);
&st($c0,&QWPw(14,$rp));
&st($c1,&QWPw(15,$rp));
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,108 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_sub_words
{
local($name)=@_;
local($cc,$a,$b,$r);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
$count=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&blt($count,&label("finish"));
($a0,$b0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
##########################################################
&set_label("loop");
($a1,$tmp,$b1,$a2,$b2,$a3,$b3,$o0)=&NR(8);
&ld($a1,&QWPw(1,$ap));
&cmpult($a0,$b0,$tmp); # will we borrow?
&ld($b1,&QWPw(1,$bp));
&sub($a0,$b0,$a0); # do the subtract
&ld($a2,&QWPw(2,$ap));
&cmpult($a0,$cc,$b0); # will we borrow?
&ld($b2,&QWPw(2,$bp));
&sub($a0,$cc,$o0); # will we borrow?
&ld($a3,&QWPw(3,$ap));
&add($b0,$tmp,$cc); ($t1,$o1)=&NR(2); &FR($tmp);
&cmpult($a1,$b1,$t1); # will we borrow?
&sub($a1,$b1,$a1); # do the subtract
&ld($b3,&QWPw(3,$bp));
&cmpult($a1,$cc,$b1); # will we borrow?
&sub($a1,$cc,$o1); # will we borrow?
&add($b1,$t1,$cc); ($tmp,$o2)=&NR(2); &FR($t1,$a1,$b1);
&cmpult($a2,$b2,$tmp); # will we borrow?
&sub($a2,$b2,$a2); # do the subtract
&st($o0,&QWPw(0,$rp)); &FR($o0); # save
&cmpult($a2,$cc,$b2); # will we borrow?
&sub($a2,$cc,$o2); # will we borrow?
&add($b2,$tmp,$cc); ($t3,$o3)=&NR(2); &FR($tmp,$a2,$b2);
&cmpult($a3,$b3,$t3); # will we borrow?
&sub($a3,$b3,$a3); # do the subtract
&st($o1,&QWPw(1,$rp)); &FR($o1);
&cmpult($a3,$cc,$b3); # will we borrow?
&sub($a3,$cc,$o3); # will we borrow?
&add($b3,$t3,$cc); &FR($t3,$a3,$b3);
&st($o2,&QWPw(2,$rp)); &FR($o2);
&sub($count,4,$count); # count-=4
&st($o3,&QWPw(3,$rp)); &FR($o3);
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld($a0,&QWPw(0,$ap)); # get a
&ld($b0,&QWPw(0,$bp)); # get b
&cmpult($a0,$b0,$tmp); # will we borrow?
&sub($a0,$b0,$a0); # do the subtract
&cmpult($a0,$cc,$b0); # will we borrow?
&sub($a0,$cc,$a0); # will we borrow?
&st($a0,&QWPw(0,$rp)); # save
&add($b0,$tmp,$cc); # add the borrows
&add($ap,$QWS,$ap);
&add($bp,$QWS,$bp);
&add($rp,$QWS,$rp);
&sub($count,1,$count);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&FR($a0,$b0);
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,118 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_add_words
{
local($name)=@_;
local($cc,$a,$b,$r);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
$count=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&blt($count,&label("finish"));
($a0,$b0)=&NR(2);
##########################################################
&set_label("loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap));
&ld(($b0)=&NR(1),&QWPw(0,$bp));
&ld(($a1)=&NR(1),&QWPw(1,$ap));
&ld(($b1)=&NR(1),&QWPw(1,$bp));
($o0,$t0)=&NR(2);
&add($a0,$b0,$o0);
&ld(($a2)=&NR(1),&QWPw(2,$ap));
&cmpult($o0,$b0,$t0);
&add($o0,$cc,$o0);
&cmpult($o0,$cc,$cc);
&ld(($b2)=&NR(1),&QWPw(2,$bp));
&add($cc,$t0,$cc); &FR($t0);
($t1,$o1)=&NR(2);
&add($a1,$b1,$o1); &FR($a1);
&cmpult($o1,$b1,$t1); &FR($b1);
&add($o1,$cc,$o1);
&cmpult($o1,$cc,$cc);
&ld(($a3)=&NR(1),&QWPw(3,$ap));
&add($cc,$t1,$cc); &FR($t1);
($t2,$o2)=&NR(2);
&add($a2,$b2,$o2); &FR($a2);
&cmpult($o2,$b2,$t2); &FR($b2);
&add($o2,$cc,$o2);
&cmpult($o2,$cc,$cc);
&ld(($b3)=&NR(1),&QWPw(3,$bp));
&st($o0,&QWPw(0,$rp)); &FR($o0);
&add($cc,$t2,$cc); &FR($t2);
($t3,$o3)=&NR(2);
&st($o1,&QWPw(0,$rp)); &FR($o1);
&add($a3,$b3,$o3); &FR($a3);
&cmpult($o3,$b3,$t3); &FR($b3);
&add($o3,$cc,$o3);
&st($o2,&QWPw(0,$rp)); &FR($o2);
&cmpult($o3,$cc,$cc);
&st($o3,&QWPw(0,$rp)); &FR($o3);
&add($cc,$t3,$cc); &FR($t3);
&sub($count,4,$count); # count-=4
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
###
&bge($count,&label("loop"));
###
&br(&label("finish"));
##################################################
# Do the last 0..3 words
($t0,$o0)=&NR(2);
&set_label("last_loop");
&ld($a0,&QWPw(0,$ap)); # get a
&ld($b0,&QWPw(0,$bp)); # get b
&add($ap,$QWS,$ap);
&add($bp,$QWS,$bp);
&add($a0,$b0,$o0);
&sub($count,1,$count);
&cmpult($o0,$b0,$t0); # will we borrow?
&add($o0,$cc,$o0); # will we borrow?
&cmpult($o0,$cc,$cc); # will we borrow?
&add($rp,$QWS,$rp);
&st($o0,&QWPw(-1,$rp)); # save
&add($cc,$t0,$cc); # add the borrows
###
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&FR($o0,$t0,$a0,$b0);
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,144 +0,0 @@
#!/usr/local/bin/perl
sub bn_div_words
{
local($data)=<<'EOF';
#
# What follows was taken directly from the C compiler with a few
# hacks to redo the lables.
#
.text
.set noreorder
.set volatile
.align 3
.globl bn_div_words
.ent bn_div_words
bn_div_words
ldgp $29,0($27)
bn_div_words.ng:
lda $30,-48($30)
.frame $30,48,$26,0
stq $26,0($30)
stq $9,8($30)
stq $10,16($30)
stq $11,24($30)
stq $12,32($30)
stq $13,40($30)
.mask 0x4003e00,-48
.prologue 1
bis $16,$16,$9
bis $17,$17,$10
bis $18,$18,$11
bis $31,$31,$13
bis $31,2,$12
bne $11,$9119
lda $0,-1
br $31,$9136
.align 4
$9119:
bis $11,$11,$16
jsr $26,BN_num_bits_word
ldgp $29,0($26)
subq $0,64,$1
beq $1,$9120
bis $31,1,$1
sll $1,$0,$1
cmpule $9,$1,$1
bne $1,$9120
# lda $16,_IO_stderr_
# lda $17,$C32
# bis $0,$0,$18
# jsr $26,fprintf
# ldgp $29,0($26)
jsr $26,abort
ldgp $29,0($26)
.align 4
$9120:
bis $31,64,$3
cmpult $9,$11,$2
subq $3,$0,$1
addl $1,$31,$0
subq $9,$11,$1
cmoveq $2,$1,$9
beq $0,$9122
zapnot $0,15,$2
subq $3,$0,$1
sll $11,$2,$11
sll $9,$2,$3
srl $10,$1,$1
sll $10,$2,$10
bis $3,$1,$9
$9122:
srl $11,32,$5
zapnot $11,15,$6
lda $7,-1
.align 5
$9123:
srl $9,32,$1
subq $1,$5,$1
bne $1,$9126
zapnot $7,15,$27
br $31,$9127
.align 4
$9126:
bis $9,$9,$24
bis $5,$5,$25
divqu $24,$25,$27
$9127:
srl $10,32,$4
.align 5
$9128:
mulq $27,$5,$1
subq $9,$1,$3
zapnot $3,240,$1
bne $1,$9129
mulq $6,$27,$2
sll $3,32,$1
addq $1,$4,$1
cmpule $2,$1,$2
bne $2,$9129
subq $27,1,$27
br $31,$9128
.align 4
$9129:
mulq $27,$6,$1
mulq $27,$5,$4
srl $1,32,$3
sll $1,32,$1
addq $4,$3,$4
cmpult $10,$1,$2
subq $10,$1,$10
addq $2,$4,$2
cmpult $9,$2,$1
bis $2,$2,$4
beq $1,$9134
addq $9,$11,$9
subq $27,1,$27
$9134:
subl $12,1,$12
subq $9,$4,$9
beq $12,$9124
sll $27,32,$13
sll $9,32,$2
srl $10,32,$1
sll $10,32,$10
bis $2,$1,$9
br $31,$9123
.align 4
$9124:
bis $13,$27,$0
$9136:
ldq $26,0($30)
ldq $9,8($30)
ldq $10,16($30)
ldq $11,24($30)
ldq $12,32($30)
ldq $13,40($30)
addq $30,48,$30
ret $31,($26),1
.end bn_div_words
EOF
&asm_add($data);
}
1;

View File

@ -1,104 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_mul_words
{
local($name)=@_;
local($cc,$a,$b,$r,$couny);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$count=&wparam(2);
$word=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
###
&blt($count,&label("finish"));
($a0)=&NR(1); &ld($a0,&QWPw(0,$ap));
&set_label("loop");
($a1)=&NR(1); &ld($a1,&QWPw(1,$ap));
($a2)=&NR(1); &ld($a2,&QWPw(2,$ap));
&muh($a0,$word,($h0)=&NR(1)); &FR($a0);
($a3)=&NR(1); &ld($a3,&QWPw(3,$ap));
### wait 8
&mul($a0,$word,($l0)=&NR(1)); &FR($a0);
### wait 8
&muh($a1,$word,($h1)=&NR(1)); &FR($a1);
&add($l0,$cc,$l0); ### wait 8
&mul($a1,$word,($l1)=&NR(1)); &FR($a1);
&cmpult($l0,$cc,$cc); ### wait 8
&muh($a2,$word,($h2)=&NR(1)); &FR($a2);
&add($h0,$cc,$cc); &FR($h0); ### wait 8
&mul($a2,$word,($l2)=&NR(1)); &FR($a2);
&add($l1,$cc,$l1); ### wait 8
&st($l0,&QWPw(0,$rp)); &FR($l0);
&cmpult($l1,$cc,$cc); ### wait 8
&muh($a3,$word,($h3)=&NR(1)); &FR($a3);
&add($h1,$cc,$cc); &FR($h1);
&mul($a3,$word,($l3)=&NR(1)); &FR($a3);
&add($l2,$cc,$l2);
&st($l1,&QWPw(1,$rp)); &FR($l1);
&cmpult($l2,$cc,$cc);
&add($h2,$cc,$cc); &FR($h2);
&sub($count,4,$count); # count-=4
&st($l2,&QWPw(2,$rp)); &FR($l2);
&add($l3,$cc,$l3);
&cmpult($l3,$cc,$cc);
&add($bp,4*$QWS,$bp); # count+=4
&add($h3,$cc,$cc); &FR($h3);
&add($ap,4*$QWS,$ap); # count+=4
&st($l3,&QWPw(3,$rp)); &FR($l3);
&add($rp,4*$QWS,$rp); # count+=4
###
&blt($count,&label("finish"));
($a0)=&NR(1); &ld($a0,&QWPw(0,$ap));
&br(&label("finish"));
##################################################
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap)); # get a
###
###
###
&muh($a0,$word,($h0)=&NR(1));
### Wait 8 for next mul issue
&mul($a0,$word,($l0)=&NR(1)); &FR($a0)
&add($ap,$QWS,$ap);
### Loose 12 until result is available
&add($rp,$QWS,$rp);
&sub($count,1,$count);
&add($l0,$cc,$l0);
###
&st($l0,&QWPw(-1,$rp)); &FR($l0);
&cmpult($l0,$cc,$cc);
&add($h0,$cc,$cc); &FR($h0);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,123 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_mul_add_words
{
local($name)=@_;
local($cc,$a,$b,$r,$couny);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$count=&wparam(2);
$word=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
###
&blt($count,&label("finish"));
&ld(($a0)=&NR(1),&QWPw(0,$ap));
$a=<<'EOF';
##########################################################
&set_label("loop");
&ld(($r0)=&NR(1),&QWPw(0,$rp));
&ld(($a1)=&NR(1),&QWPw(1,$ap));
&muh($a0,$word,($h0)=&NR(1));
&ld(($r1)=&NR(1),&QWPw(1,$rp));
&ld(($a2)=&NR(1),&QWPw(2,$ap));
###
&mul($a0,$word,($l0)=&NR(1)); &FR($a0);
&ld(($r2)=&NR(1),&QWPw(2,$rp));
&muh($a1,$word,($h1)=&NR(1));
&ld(($a3)=&NR(1),&QWPw(3,$ap));
&mul($a1,$word,($l1)=&NR(1)); &FR($a1);
&ld(($r3)=&NR(1),&QWPw(3,$rp));
&add($r0,$l0,$r0);
&add($r1,$l1,$r1);
&cmpult($r0,$l0,($t0)=&NR(1)); &FR($l0);
&cmpult($r1,$l1,($t1)=&NR(1)); &FR($l1);
&muh($a2,$word,($h2)=&NR(1));
&add($r0,$cc,$r0);
&add($h0,$t0,$h0); &FR($t0);
&cmpult($r0,$cc,$cc);
&add($h1,$t1,$h1); &FR($t1);
&add($h0,$cc,$cc); &FR($h0);
&mul($a2,$word,($l2)=&NR(1)); &FR($a2);
&add($r1,$cc,$r1);
&cmpult($r1,$cc,$cc);
&add($r2,$l2,$r2);
&add($h1,$cc,$cc); &FR($h1);
&cmpult($r2,$l2,($t2)=&NR(1)); &FR($l2);
&muh($a3,$word,($h3)=&NR(1));
&add($r2,$cc,$r2);
&st($r0,&QWPw(0,$rp)); &FR($r0);
&add($h2,$t2,$h2); &FR($t2);
&st($r1,&QWPw(1,$rp)); &FR($r1);
&cmpult($r2,$cc,$cc);
&mul($a3,$word,($l3)=&NR(1)); &FR($a3);
&add($h2,$cc,$cc); &FR($h2);
&st($r2,&QWPw(2,$rp)); &FR($r2);
&sub($count,4,$count); # count-=4
&add($rp,4*$QWS,$rp); # count+=4
&add($r3,$l3,$r3);
&add($ap,4*$QWS,$ap); # count+=4
&cmpult($r3,$l3,($t3)=&NR(1)); &FR($l3);
&add($r3,$cc,$r3);
&add($h3,$t3,$h3); &FR($t3);
&cmpult($r3,$cc,$cc);
&st($r3,&QWPw(-1,$rp)); &FR($r3);
&add($h3,$cc,$cc); &FR($h3);
###
&blt($count,&label("finish"));
&ld(($a0)=&NR(1),&QWPw(0,$ap));
&br(&label("loop"));
EOF
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap)); # get a
&ld(($r0)=&NR(1),&QWPw(0,$rp)); # get b
###
###
&muh($a0,$word,($h0)=&NR(1)); &FR($a0);
### wait 8
&mul($a0,$word,($l0)=&NR(1)); &FR($a0);
&add($rp,$QWS,$rp);
&add($ap,$QWS,$ap);
&sub($count,1,$count);
### wait 3 until l0 is available
&add($r0,$l0,$r0);
###
&cmpult($r0,$l0,($t0)=&NR(1)); &FR($l0);
&add($r0,$cc,$r0);
&add($h0,$t0,$h0); &FR($t0);
&cmpult($r0,$cc,$cc);
&add($h0,$cc,$cc); &FR($h0);
&st($r0,&QWPw(-1,$rp)); &FR($r0);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,215 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
# upto
sub mul_add_c
{
local($a,$b,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
&mul($a,$b,($l1)=&NR(1));
&muh($a,$b,($h1)=&NR(1));
&add($c0,$l1,$c0);
&cmpult($c0,$l1,($t1)=&NR(1)); &FR($l1);
&add($t1,$h1,$h1); &FR($t1);
&add($c1,$h1,$c1);
&cmpult($c1,$h1,($t2)=&NR(1)); &FR($h1);
&add($c2,$t2,$c2); &FR($t2);
}
sub bn_mul_comba4
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(3);
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($b[0])=&NR(1),&QWPw(0,$bp));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($b[1])=&NR(1),&QWPw(1,$bp));
&mul($a[0],$b[0],($r00)=&NR(1));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($b[2])=&NR(1),&QWPw(2,$bp));
&muh($a[0],$b[0],($r01)=&NR(1));
&FR($ap); &ld(($a[3])=&NR(1),&QWPw(3,$ap));
&FR($bp); &ld(($b[3])=&NR(1),&QWPw(3,$bp));
&mul($a[0],$b[1],($r02)=&NR(1));
($R,$H1,$H2)=&NR(3);
&st($r00,&QWPw(0,$rp)); &FR($r00);
&mov("zero",$R);
&mul($a[1],$b[0],($r03)=&NR(1));
&mov("zero",$H1);
&mov("zero",$H0);
&add($R,$r01,$R);
&muh($a[0],$b[1],($r04)=&NR(1));
&cmpult($R,$r01,($t01)=&NR(1)); &FR($r01);
&add($R,$r02,$R);
&add($H1,$t01,$H1) &FR($t01);
&muh($a[1],$b[0],($r05)=&NR(1));
&cmpult($R,$r02,($t02)=&NR(1)); &FR($r02);
&add($R,$r03,$R);
&add($H2,$t02,$H2) &FR($t02);
&mul($a[0],$b[2],($r06)=&NR(1));
&cmpult($R,$r03,($t03)=&NR(1)); &FR($r03);
&add($H1,$t03,$H1) &FR($t03);
&st($R,&QWPw(1,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r04,$R);
&mov("zero",$H2);
&mul($a[1],$b[1],($r07)=&NR(1));
&cmpult($R,$r04,($t04)=&NR(1)); &FR($r04);
&add($R,$r05,$R);
&add($H1,$t04,$H1) &FR($t04);
&mul($a[2],$b[0],($r08)=&NR(1));
&cmpult($R,$r05,($t05)=&NR(1)); &FR($r05);
&add($R,$r01,$R);
&add($H2,$t05,$H2) &FR($t05);
&muh($a[0],$b[2],($r09)=&NR(1));
&cmpult($R,$r06,($t06)=&NR(1)); &FR($r06);
&add($R,$r07,$R);
&add($H1,$t06,$H1) &FR($t06);
&muh($a[1],$b[1],($r10)=&NR(1));
&cmpult($R,$r07,($t07)=&NR(1)); &FR($r07);
&add($R,$r08,$R);
&add($H2,$t07,$H2) &FR($t07);
&muh($a[2],$b[0],($r11)=&NR(1));
&cmpult($R,$r08,($t08)=&NR(1)); &FR($r08);
&add($H1,$t08,$H1) &FR($t08);
&st($R,&QWPw(2,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r09,$R);
&mov("zero",$H2);
&mul($a[0],$b[3],($r12)=&NR(1));
&cmpult($R,$r09,($t09)=&NR(1)); &FR($r09);
&add($R,$r10,$R);
&add($H1,$t09,$H1) &FR($t09);
&mul($a[1],$b[2],($r13)=&NR(1));
&cmpult($R,$r10,($t10)=&NR(1)); &FR($r10);
&add($R,$r11,$R);
&add($H1,$t10,$H1) &FR($t10);
&mul($a[2],$b[1],($r14)=&NR(1));
&cmpult($R,$r11,($t11)=&NR(1)); &FR($r11);
&add($R,$r12,$R);
&add($H1,$t11,$H1) &FR($t11);
&mul($a[3],$b[0],($r15)=&NR(1));
&cmpult($R,$r12,($t12)=&NR(1)); &FR($r12);
&add($R,$r13,$R);
&add($H1,$t12,$H1) &FR($t12);
&muh($a[0],$b[3],($r16)=&NR(1));
&cmpult($R,$r13,($t13)=&NR(1)); &FR($r13);
&add($R,$r14,$R);
&add($H1,$t13,$H1) &FR($t13);
&muh($a[1],$b[2],($r17)=&NR(1));
&cmpult($R,$r14,($t14)=&NR(1)); &FR($r14);
&add($R,$r15,$R);
&add($H1,$t14,$H1) &FR($t14);
&muh($a[2],$b[1],($r18)=&NR(1));
&cmpult($R,$r15,($t15)=&NR(1)); &FR($r15);
&add($H1,$t15,$H1) &FR($t15);
&st($R,&QWPw(3,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r16,$R);
&mov("zero",$H2);
&muh($a[3],$b[0],($r19)=&NR(1));
&cmpult($R,$r16,($t16)=&NR(1)); &FR($r16);
&add($R,$r17,$R);
&add($H1,$t16,$H1) &FR($t16);
&mul($a[1],$b[3],($r20)=&NR(1));
&cmpult($R,$r17,($t17)=&NR(1)); &FR($r17);
&add($R,$r18,$R);
&add($H1,$t17,$H1) &FR($t17);
&mul($a[2],$b[2],($r21)=&NR(1));
&cmpult($R,$r18,($t18)=&NR(1)); &FR($r18);
&add($R,$r19,$R);
&add($H1,$t18,$H1) &FR($t18);
&mul($a[3],$b[1],($r22)=&NR(1));
&cmpult($R,$r19,($t19)=&NR(1)); &FR($r19);
&add($R,$r20,$R);
&add($H1,$t19,$H1) &FR($t19);
&muh($a[1],$b[3],($r23)=&NR(1));
&cmpult($R,$r20,($t20)=&NR(1)); &FR($r20);
&add($R,$r21,$R);
&add($H1,$t20,$H1) &FR($t20);
&muh($a[2],$b[2],($r24)=&NR(1));
&cmpult($R,$r21,($t21)=&NR(1)); &FR($r21);
&add($R,$r22,$R);
&add($H1,$t21,$H1) &FR($t21);
&muh($a[3],$b[1],($r25)=&NR(1));
&cmpult($R,$r22,($t22)=&NR(1)); &FR($r22);
&add($H1,$t22,$H1) &FR($t22);
&st($R,&QWPw(4,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r23,$R);
&mov("zero",$H2);
&mul($a[2],$b[3],($r26)=&NR(1));
&cmpult($R,$r23,($t23)=&NR(1)); &FR($r23);
&add($R,$r24,$R);
&add($H1,$t23,$H1) &FR($t23);
&mul($a[3],$b[2],($r27)=&NR(1));
&cmpult($R,$r24,($t24)=&NR(1)); &FR($r24);
&add($R,$r25,$R);
&add($H1,$t24,$H1) &FR($t24);
&muh($a[2],$b[3],($r28)=&NR(1));
&cmpult($R,$r25,($t25)=&NR(1)); &FR($r25);
&add($R,$r26,$R);
&add($H1,$t25,$H1) &FR($t25);
&muh($a[3],$b[2],($r29)=&NR(1));
&cmpult($R,$r26,($t26)=&NR(1)); &FR($r26);
&add($R,$r27,$R);
&add($H1,$t26,$H1) &FR($t26);
&mul($a[3],$b[3],($r30)=&NR(1));
&cmpult($R,$r27,($t27)=&NR(1)); &FR($r27);
&add($H1,$t27,$H1) &FR($t27);
&st($R,&QWPw(5,$rp));
&add($H1,$H2,$R);
&mov("zero",$H1);
&add($R,$r28,$R);
&mov("zero",$H2);
&muh($a[3],$b[3],($r31)=&NR(1));
&cmpult($R,$r28,($t28)=&NR(1)); &FR($r28);
&add($R,$r29,$R);
&add($H1,$t28,$H1) &FR($t28);
############
&cmpult($R,$r29,($t29)=&NR(1)); &FR($r29);
&add($R,$r30,$R);
&add($H1,$t29,$H1) &FR($t29);
############
&cmpult($R,$r30,($t30)=&NR(1)); &FR($r30);
&add($H1,$t30,$H1) &FR($t30);
&st($R,&QWPw(6,$rp));
&add($H1,$H2,$R);
&add($R,$r31,$R); &FR($r31);
&st($R,&QWPw(7,$rp));
&FR($R,$H1,$H2);
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,98 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub mul_add_c
{
local($a,$b,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
print STDERR "count=$cnt\n"; $cnt++;
&mul($a,$b,($l1)=&NR(1));
&muh($a,$b,($h1)=&NR(1));
&add($c0,$l1,$c0);
&cmpult($c0,$l1,($t1)=&NR(1)); &FR($l1);
&add($t1,$h1,$h1); &FR($t1);
&add($c1,$h1,$c1);
&cmpult($c1,$h1,($t2)=&NR(1)); &FR($h1);
&add($c2,$t2,$c2); &FR($t2);
}
sub bn_mul_comba4
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(3);
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($b[0])=&NR(1),&QWPw(0,$bp));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($b[1])=&NR(1),&QWPw(1,$bp));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($b[2])=&NR(1),&QWPw(2,$bp));
&ld(($a[3])=&NR(1),&QWPw(3,$ap)); &FR($ap);
&ld(($b[3])=&NR(1),&QWPw(3,$bp)); &FR($bp);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$b[0],$c0);
&muh($a[0],$b[0],$c1);
&st($c0,&QWPw(0,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[1],$c0,$c1,$c2);
&mul_add_c($a[1],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[1],$b[1],$c0,$c1,$c2);
&mul_add_c($a[0],$b[2],$c0,$c1,$c2);
&mul_add_c($a[2],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[3],$c0,$c1,$c2); &FR($a[0]);
&mul_add_c($a[1],$b[2],$c0,$c1,$c2);
&mul_add_c($a[2],$b[1],$c0,$c1,$c2);
&mul_add_c($a[3],$b[0],$c0,$c1,$c2); &FR($b[0]);
&st($c0,&QWPw(3,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[1],$b[3],$c0,$c1,$c2); &FR($a[1]);
&mul_add_c($a[2],$b[2],$c0,$c1,$c2);
&mul_add_c($a[3],$b[1],$c0,$c1,$c2); &FR($b[1]);
&st($c0,&QWPw(4,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[2],$b[3],$c0,$c1,$c2); &FR($a[2]);
&mul_add_c($a[3],$b[2],$c0,$c1,$c2); &FR($b[2]);
&st($c0,&QWPw(5,$rp)); &FR($c0); ($c0)=&NR($c0);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[3],$b[3],$c0,$c1,$c2); &FR($a[3],$b[3]);
&st($c0,&QWPw(6,$rp));
&st($c1,&QWPw(7,$rp));
&FR($c0,$c1,$c2);
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,177 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_mul_comba8
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(3);
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
&function_begin($name,"");
&comment("");
&stack_push(2);
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($b[0])=&NR(1),&QWPw(0,$bp));
&st($reg_s0,&swtmp(0)); &FR($reg_s0);
&st($reg_s1,&swtmp(1)); &FR($reg_s1);
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($b[1])=&NR(1),&QWPw(1,$bp));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($b[2])=&NR(1),&QWPw(2,$bp));
&ld(($a[3])=&NR(1),&QWPw(3,$ap));
&ld(($b[3])=&NR(1),&QWPw(3,$bp));
&ld(($a[4])=&NR(1),&QWPw(1,$ap));
&ld(($b[4])=&NR(1),&QWPw(1,$bp));
&ld(($a[5])=&NR(1),&QWPw(1,$ap));
&ld(($b[5])=&NR(1),&QWPw(1,$bp));
&ld(($a[6])=&NR(1),&QWPw(1,$ap));
&ld(($b[6])=&NR(1),&QWPw(1,$bp));
&ld(($a[7])=&NR(1),&QWPw(1,$ap)); &FR($ap);
&ld(($b[7])=&NR(1),&QWPw(1,$bp)); &FR($bp);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$b[0],$c0);
&muh($a[0],$b[0],$c1);
&st($c0,&QWPw(0,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[1],$c0,$c1,$c2);
&mul_add_c($a[1],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[2],$c0,$c1,$c2);
&mul_add_c($a[1],$b[1],$c0,$c1,$c2);
&mul_add_c($a[2],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[3],$c0,$c1,$c2);
&mul_add_c($a[1],$b[2],$c0,$c1,$c2);
&mul_add_c($a[2],$b[1],$c0,$c1,$c2);
&mul_add_c($a[3],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(3,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[4],$c0,$c1,$c2);
&mul_add_c($a[1],$b[3],$c0,$c1,$c2);
&mul_add_c($a[2],$b[2],$c0,$c1,$c2);
&mul_add_c($a[3],$b[1],$c0,$c1,$c2);
&mul_add_c($a[4],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(4,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[5],$c0,$c1,$c2);
&mul_add_c($a[1],$b[4],$c0,$c1,$c2);
&mul_add_c($a[2],$b[3],$c0,$c1,$c2);
&mul_add_c($a[3],$b[2],$c0,$c1,$c2);
&mul_add_c($a[4],$b[1],$c0,$c1,$c2);
&mul_add_c($a[5],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(5,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[6],$c0,$c1,$c2);
&mul_add_c($a[1],$b[5],$c0,$c1,$c2);
&mul_add_c($a[2],$b[4],$c0,$c1,$c2);
&mul_add_c($a[3],$b[3],$c0,$c1,$c2);
&mul_add_c($a[4],$b[2],$c0,$c1,$c2);
&mul_add_c($a[5],$b[1],$c0,$c1,$c2);
&mul_add_c($a[6],$b[0],$c0,$c1,$c2);
&st($c0,&QWPw(6,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[0],$b[7],$c0,$c1,$c2); &FR($a[0]);
&mul_add_c($a[1],$b[6],$c0,$c1,$c2);
&mul_add_c($a[2],$b[5],$c0,$c1,$c2);
&mul_add_c($a[3],$b[4],$c0,$c1,$c2);
&mul_add_c($a[4],$b[3],$c0,$c1,$c2);
&mul_add_c($a[5],$b[2],$c0,$c1,$c2);
&mul_add_c($a[6],$b[1],$c0,$c1,$c2);
&mul_add_c($a[7],$b[0],$c0,$c1,$c2); &FR($b[0]);
&st($c0,&QWPw(7,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[1],$b[7],$c0,$c1,$c2); &FR($a[1]);
&mul_add_c($a[2],$b[6],$c0,$c1,$c2);
&mul_add_c($a[3],$b[5],$c0,$c1,$c2);
&mul_add_c($a[4],$b[4],$c0,$c1,$c2);
&mul_add_c($a[5],$b[3],$c0,$c1,$c2);
&mul_add_c($a[6],$b[2],$c0,$c1,$c2);
&mul_add_c($a[7],$b[1],$c0,$c1,$c2); &FR($b[1]);
&st($c0,&QWPw(8,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[2],$b[7],$c0,$c1,$c2); &FR($a[2]);
&mul_add_c($a[3],$b[6],$c0,$c1,$c2);
&mul_add_c($a[4],$b[5],$c0,$c1,$c2);
&mul_add_c($a[5],$b[4],$c0,$c1,$c2);
&mul_add_c($a[6],$b[3],$c0,$c1,$c2);
&mul_add_c($a[7],$b[2],$c0,$c1,$c2); &FR($b[2]);
&st($c0,&QWPw(9,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[3],$b[7],$c0,$c1,$c2); &FR($a[3]);
&mul_add_c($a[4],$b[6],$c0,$c1,$c2);
&mul_add_c($a[5],$b[5],$c0,$c1,$c2);
&mul_add_c($a[6],$b[4],$c0,$c1,$c2);
&mul_add_c($a[7],$b[3],$c0,$c1,$c2); &FR($b[3]);
&st($c0,&QWPw(10,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[4],$b[7],$c0,$c1,$c2); &FR($a[4]);
&mul_add_c($a[5],$b[6],$c0,$c1,$c2);
&mul_add_c($a[6],$b[5],$c0,$c1,$c2);
&mul_add_c($a[7],$b[4],$c0,$c1,$c2); &FR($b[4]);
&st($c0,&QWPw(11,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[5],$b[7],$c0,$c1,$c2); &FR($a[5]);
&mul_add_c($a[6],$b[6],$c0,$c1,$c2);
&mul_add_c($a[7],$b[5],$c0,$c1,$c2); &FR($b[5]);
&st($c0,&QWPw(12,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[6],$b[7],$c0,$c1,$c2); &FR($a[6]);
&mul_add_c($a[7],$b[6],$c0,$c1,$c2); &FR($b[6]);
&st($c0,&QWPw(13,$rp)); &FR($c0); ($c0)=&NR(1);
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&mul_add_c($a[7],$b[7],$c0,$c1,$c2); &FR($a[7],$b[7]);
&st($c0,&QWPw(14,$rp));
&st($c1,&QWPw(15,$rp));
&FR($c0,$c1,$c2);
&ld($reg_s0,&swtmp(0));
&ld($reg_s1,&swtmp(1));
&stack_pop(2);
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,113 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_sqr_words
{
local($name)=@_;
local($cc,$a,$b,$r,$couny);
&init_pool(3);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$count=&wparam(2);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&br(&label("finish"));
&blt($count,&label("finish"));
($a0,$r0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($r0,&QWPw(0,$rp));
$a=<<'EOF';
##########################################################
&set_label("loop");
($a1)=&NR(1); &ld($a1,&QWPw(1,$ap));
($b1)=&NR(1); &ld($b1,&QWPw(1,$bp));
($a2)=&NR(1); &ld($a2,&QWPw(2,$ap));
($b2)=&NR(1); &ld($b2,&QWPw(2,$bp));
($a3)=&NR(1); &ld($a3,&QWPw(3,$ap));
($b3)=&NR(1); &ld($b3,&QWPw(3,$bp));
($o0,$t0)=&NR(2);
&add($a0,$b0,$o0);
&cmpult($o0,$b0,$t0);
&add($o0,$cc,$o0);
&cmpult($o0,$cc,$cc);
&add($cc,$t0,$cc); &FR($t0);
($t1,$o1)=&NR(2);
&add($a1,$b1,$o1); &FR($a1);
&cmpult($o1,$b1,$t1); &FR($b1);
&add($o1,$cc,$o1);
&cmpult($o1,$cc,$cc);
&add($cc,$t1,$cc); &FR($t1);
($t2,$o2)=&NR(2);
&add($a2,$b2,$o2); &FR($a2);
&cmpult($o2,$b2,$t2); &FR($b2);
&add($o2,$cc,$o2);
&cmpult($o2,$cc,$cc);
&add($cc,$t2,$cc); &FR($t2);
($t3,$o3)=&NR(2);
&add($a3,$b3,$o3); &FR($a3);
&cmpult($o3,$b3,$t3); &FR($b3);
&add($o3,$cc,$o3);
&cmpult($o3,$cc,$cc);
&add($cc,$t3,$cc); &FR($t3);
&st($o0,&QWPw(0,$rp)); &FR($o0);
&st($o1,&QWPw(0,$rp)); &FR($o1);
&st($o2,&QWPw(0,$rp)); &FR($o2);
&st($o3,&QWPw(0,$rp)); &FR($o3);
&sub($count,4,$count); # count-=4
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
EOF
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld(($a0)=&NR(1),&QWPw(0,$ap)); # get a
&mul($a0,$a0,($l0)=&NR(1));
&add($ap,$QWS,$ap);
&add($rp,2*$QWS,$rp);
&sub($count,1,$count);
&muh($a0,$a0,($h0)=&NR(1)); &FR($a0);
&st($l0,&QWPw(-2,$rp)); &FR($l0);
&st($h0,&QWPw(-1,$rp)); &FR($h0);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,109 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub sqr_add_c
{
local($a,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
&mul($a,$a,($l1)=&NR(1));
&muh($a,$a,($h1)=&NR(1));
&add($c0,$l1,$c0);
&add($c1,$h1,$c1);
&cmpult($c0,$l1,($t1)=&NR(1)); &FR($l1);
&cmpult($c1,$h1,($t2)=&NR(1)); &FR($h1);
&add($c1,$t1,$c1); &FR($t1);
&add($c2,$t2,$c2); &FR($t2);
}
sub sqr_add_c2
{
local($a,$b,$c0,$c1,$c2)=@_;
local($l1,$h1,$t1,$t2);
&mul($a,$b,($l1)=&NR(1));
&muh($a,$b,($h1)=&NR(1));
&cmplt($l1,"zero",($lc1)=&NR(1));
&cmplt($h1,"zero",($hc1)=&NR(1));
&add($l1,$l1,$l1);
&add($h1,$h1,$h1);
&add($h1,$lc1,$h1); &FR($lc1);
&add($c2,$hc1,$c2); &FR($hc1);
&add($c0,$l1,$c0);
&add($c1,$h1,$c1);
&cmpult($c0,$l1,($lc1)=&NR(1)); &FR($l1);
&cmpult($c1,$h1,($hc1)=&NR(1)); &FR($h1);
&add($c1,$lc1,$c1); &FR($lc1);
&add($c2,$hc1,$c2); &FR($hc1);
}
sub bn_sqr_comba4
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(2);
$rp=&wparam(0);
$ap=&wparam(1);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($a[3])=&NR(1),&QWPw(3,$ap)); &FR($ap);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$a[0],$c0);
&muh($a[0],$a[0],$c1);
&st($c0,&QWPw(0,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[0],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[1],$c0,$c1,$c2);
&sqr_add_c2($a[2],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[3],$a[0],$c0,$c1,$c2);
&sqr_add_c2($a[2],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(3,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[2],$c0,$c1,$c2);
&sqr_add_c2($a[3],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(4,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[3],$a[2],$c0,$c1,$c2);
&st($c0,&QWPw(5,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[3],$c0,$c1,$c2);
&st($c0,&QWPw(6,$rp));
&st($c1,&QWPw(7,$rp));
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,132 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_sqr_comba8
{
local($name)=@_;
local(@a,@b,$r,$c0,$c1,$c2);
$cnt=1;
&init_pool(2);
$rp=&wparam(0);
$ap=&wparam(1);
&function_begin($name,"");
&comment("");
&ld(($a[0])=&NR(1),&QWPw(0,$ap));
&ld(($a[1])=&NR(1),&QWPw(1,$ap));
&ld(($a[2])=&NR(1),&QWPw(2,$ap));
&ld(($a[3])=&NR(1),&QWPw(3,$ap));
&ld(($a[4])=&NR(1),&QWPw(4,$ap));
&ld(($a[5])=&NR(1),&QWPw(5,$ap));
&ld(($a[6])=&NR(1),&QWPw(6,$ap));
&ld(($a[7])=&NR(1),&QWPw(7,$ap)); &FR($ap);
($c0,$c1,$c2)=&NR(3);
&mov("zero",$c2);
&mul($a[0],$a[0],$c0);
&muh($a[0],$a[0],$c1);
&st($c0,&QWPw(0,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[1],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(1,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[1],$c0,$c1,$c2);
&sqr_add_c2($a[2],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(2,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[2],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[3],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(3,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[2],$c0,$c1,$c2);
&sqr_add_c2($a[3],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[4],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(4,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[3],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[4],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(5,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[3],$c0,$c1,$c2);
&sqr_add_c2($a[4],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(6,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[4],$a[3],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[1],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[0],$c0,$c1,$c2);
&st($c0,&QWPw(7,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[4],$c0,$c1,$c2);
&sqr_add_c2($a[5],$a[3],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[2],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[1],$c0,$c1,$c2);
&st($c0,&QWPw(8,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[5],$a[4],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[3],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[2],$c0,$c1,$c2);
&st($c0,&QWPw(9,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[5],$c0,$c1,$c2);
&sqr_add_c2($a[6],$a[4],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[3],$c0,$c1,$c2);
&st($c0,&QWPw(10,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[6],$a[5],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[4],$c0,$c1,$c2);
&st($c0,&QWPw(11,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[6],$c0,$c1,$c2);
&sqr_add_c2($a[7],$a[5],$c0,$c1,$c2);
&st($c0,&QWPw(12,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c2($a[7],$a[6],$c0,$c1,$c2);
&st($c0,&QWPw(13,$rp));
($c0,$c1,$c2)=($c1,$c2,$c0);
&mov("zero",$c2);
&sqr_add_c($a[7],$c0,$c1,$c2);
&st($c0,&QWPw(14,$rp));
&st($c1,&QWPw(15,$rp));
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,108 +0,0 @@
#!/usr/local/bin/perl
# alpha assember
sub bn_sub_words
{
local($name)=@_;
local($cc,$a,$b,$r);
&init_pool(4);
($cc)=GR("r0");
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
$count=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&blt($count,&label("finish"));
($a0,$b0)=&NR(2);
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
##########################################################
&set_label("loop");
($a1,$tmp,$b1,$a2,$b2,$a3,$b3,$o0)=&NR(8);
&ld($a1,&QWPw(1,$ap));
&cmpult($a0,$b0,$tmp); # will we borrow?
&ld($b1,&QWPw(1,$bp));
&sub($a0,$b0,$a0); # do the subtract
&ld($a2,&QWPw(2,$ap));
&cmpult($a0,$cc,$b0); # will we borrow?
&ld($b2,&QWPw(2,$bp));
&sub($a0,$cc,$o0); # will we borrow?
&ld($a3,&QWPw(3,$ap));
&add($b0,$tmp,$cc); ($t1,$o1)=&NR(2); &FR($tmp);
&cmpult($a1,$b1,$t1); # will we borrow?
&sub($a1,$b1,$a1); # do the subtract
&ld($b3,&QWPw(3,$bp));
&cmpult($a1,$cc,$b1); # will we borrow?
&sub($a1,$cc,$o1); # will we borrow?
&add($b1,$t1,$cc); ($tmp,$o2)=&NR(2); &FR($t1,$a1,$b1);
&cmpult($a2,$b2,$tmp); # will we borrow?
&sub($a2,$b2,$a2); # do the subtract
&st($o0,&QWPw(0,$rp)); &FR($o0); # save
&cmpult($a2,$cc,$b2); # will we borrow?
&sub($a2,$cc,$o2); # will we borrow?
&add($b2,$tmp,$cc); ($t3,$o3)=&NR(2); &FR($tmp,$a2,$b2);
&cmpult($a3,$b3,$t3); # will we borrow?
&sub($a3,$b3,$a3); # do the subtract
&st($o1,&QWPw(1,$rp)); &FR($o1);
&cmpult($a3,$cc,$b3); # will we borrow?
&sub($a3,$cc,$o3); # will we borrow?
&add($b3,$t3,$cc); &FR($t3,$a3,$b3);
&st($o2,&QWPw(2,$rp)); &FR($o2);
&sub($count,4,$count); # count-=4
&st($o3,&QWPw(3,$rp)); &FR($o3);
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld($a0,&QWPw(0,$ap)); # get a
&ld($b0,&QWPw(0,$bp)); # get b
&cmpult($a0,$b0,$tmp); # will we borrow?
&sub($a0,$b0,$a0); # do the subtract
&cmpult($a0,$cc,$b0); # will we borrow?
&sub($a0,$cc,$a0); # will we borrow?
&st($a0,&QWPw(0,$rp)); # save
&add($b0,$tmp,$cc); # add the borrows
&add($ap,$QWS,$ap);
&add($bp,$QWS,$bp);
&add($rp,$QWS,$rp);
&sub($count,1,$count);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&FR($a0,$b0);
&set_label("end");
&function_end($name);
&fin_pool;
}
1;

View File

@ -1,571 +0,0 @@
#!/usr/local/bin/perl
# I have this in perl so I can use more usefull register names and then convert
# them into alpha registers.
#
$d=&data();
$d =~ s/CC/0/g;
$d =~ s/R1/1/g;
$d =~ s/R2/2/g;
$d =~ s/R3/3/g;
$d =~ s/R4/4/g;
$d =~ s/L1/5/g;
$d =~ s/L2/6/g;
$d =~ s/L3/7/g;
$d =~ s/L4/8/g;
$d =~ s/O1/22/g;
$d =~ s/O2/23/g;
$d =~ s/O3/24/g;
$d =~ s/O4/25/g;
$d =~ s/A1/20/g;
$d =~ s/A2/21/g;
$d =~ s/A3/27/g;
$d =~ s/A4/28/g;
if (0){
}
print $d;
sub data
{
local($data)=<<'EOF';
# DEC Alpha assember
# The bn_div_words is actually gcc output but the other parts are hand done.
# Thanks to tzeruch@ceddec.com for sending me the gcc output for
# bn_div_words.
# I've gone back and re-done most of routines.
# The key thing to remeber for the 164 CPU is that while a
# multiply operation takes 8 cycles, another one can only be issued
# after 4 cycles have elapsed. I've done modification to help
# improve this. Also, normally, a ld instruction will not be available
# for about 3 cycles.
.file 1 "bn_asm.c"
.set noat
gcc2_compiled.:
__gnu_compiled_c:
.text
.align 3
.globl bn_mul_add_words
.ent bn_mul_add_words
bn_mul_add_words:
bn_mul_add_words..ng:
.frame $30,0,$26,0
.prologue 0
.align 5
subq $18,4,$18
bis $31,$31,$CC
blt $18,$43 # if we are -1, -2, -3 or -4 goto tail code
ldq $A1,0($17) # 1 1
ldq $R1,0($16) # 1 1
.align 3
$42:
mulq $A1,$19,$L1 # 1 2 1 ######
ldq $A2,8($17) # 2 1
ldq $R2,8($16) # 2 1
umulh $A1,$19,$A1 # 1 2 ######
ldq $A3,16($17) # 3 1
ldq $R3,16($16) # 3 1
mulq $A2,$19,$L2 # 2 2 1 ######
ldq $A4,24($17) # 4 1
addq $R1,$L1,$R1 # 1 2 2
ldq $R4,24($16) # 4 1
umulh $A2,$19,$A2 # 2 2 ######
cmpult $R1,$L1,$O1 # 1 2 3 1
addq $A1,$O1,$A1 # 1 3 1
addq $R1,$CC,$R1 # 1 2 3 1
mulq $A3,$19,$L3 # 3 2 1 ######
cmpult $R1,$CC,$CC # 1 2 3 2
addq $R2,$L2,$R2 # 2 2 2
addq $A1,$CC,$CC # 1 3 2
cmpult $R2,$L2,$O2 # 2 2 3 1
addq $A2,$O2,$A2 # 2 3 1
umulh $A3,$19,$A3 # 3 2 ######
addq $R2,$CC,$R2 # 2 2 3 1
cmpult $R2,$CC,$CC # 2 2 3 2
subq $18,4,$18
mulq $A4,$19,$L4 # 4 2 1 ######
addq $A2,$CC,$CC # 2 3 2
addq $R3,$L3,$R3 # 3 2 2
addq $16,32,$16
cmpult $R3,$L3,$O3 # 3 2 3 1
stq $R1,-32($16) # 1 2 4
umulh $A4,$19,$A4 # 4 2 ######
addq $A3,$O3,$A3 # 3 3 1
addq $R3,$CC,$R3 # 3 2 3 1
stq $R2,-24($16) # 2 2 4
cmpult $R3,$CC,$CC # 3 2 3 2
stq $R3,-16($16) # 3 2 4
addq $R4,$L4,$R4 # 4 2 2
addq $A3,$CC,$CC # 3 3 2
cmpult $R4,$L4,$O4 # 4 2 3 1
addq $17,32,$17
addq $A4,$O4,$A4 # 4 3 1
addq $R4,$CC,$R4 # 4 2 3 1
cmpult $R4,$CC,$CC # 4 2 3 2
stq $R4,-8($16) # 4 2 4
addq $A4,$CC,$CC # 4 3 2
blt $18,$43
ldq $A1,0($17) # 1 1
ldq $R1,0($16) # 1 1
br $42
.align 4
$45:
ldq $A1,0($17) # 4 1
ldq $R1,0($16) # 4 1
mulq $A1,$19,$L1 # 4 2 1
subq $18,1,$18
addq $16,8,$16
addq $17,8,$17
umulh $A1,$19,$A1 # 4 2
addq $R1,$L1,$R1 # 4 2 2
cmpult $R1,$L1,$O1 # 4 2 3 1
addq $A1,$O1,$A1 # 4 3 1
addq $R1,$CC,$R1 # 4 2 3 1
cmpult $R1,$CC,$CC # 4 2 3 2
addq $A1,$CC,$CC # 4 3 2
stq $R1,-8($16) # 4 2 4
bgt $18,$45
ret $31,($26),1 # else exit
.align 4
$43:
addq $18,4,$18
bgt $18,$45 # goto tail code
ret $31,($26),1 # else exit
.end bn_mul_add_words
.align 3
.globl bn_mul_words
.ent bn_mul_words
bn_mul_words:
bn_mul_words..ng:
.frame $30,0,$26,0
.prologue 0
.align 5
subq $18,4,$18
bis $31,$31,$CC
blt $18,$143 # if we are -1, -2, -3 or -4 goto tail code
ldq $A1,0($17) # 1 1
.align 3
$142:
mulq $A1,$19,$L1 # 1 2 1 #####
ldq $A2,8($17) # 2 1
ldq $A3,16($17) # 3 1
umulh $A1,$19,$A1 # 1 2 #####
ldq $A4,24($17) # 4 1
mulq $A2,$19,$L2 # 2 2 1 #####
addq $L1,$CC,$L1 # 1 2 3 1
subq $18,4,$18
cmpult $L1,$CC,$CC # 1 2 3 2
umulh $A2,$19,$A2 # 2 2 #####
addq $A1,$CC,$CC # 1 3 2
addq $17,32,$17
addq $L2,$CC,$L2 # 2 2 3 1
mulq $A3,$19,$L3 # 3 2 1 #####
cmpult $L2,$CC,$CC # 2 2 3 2
addq $A2,$CC,$CC # 2 3 2
addq $16,32,$16
umulh $A3,$19,$A3 # 3 2 #####
stq $L1,-32($16) # 1 2 4
mulq $A4,$19,$L4 # 4 2 1 #####
addq $L3,$CC,$L3 # 3 2 3 1
stq $L2,-24($16) # 2 2 4
cmpult $L3,$CC,$CC # 3 2 3 2
umulh $A4,$19,$A4 # 4 2 #####
addq $A3,$CC,$CC # 3 3 2
stq $L3,-16($16) # 3 2 4
addq $L4,$CC,$L4 # 4 2 3 1
cmpult $L4,$CC,$CC # 4 2 3 2
addq $A4,$CC,$CC # 4 3 2
stq $L4,-8($16) # 4 2 4
blt $18,$143
ldq $A1,0($17) # 1 1
br $142
.align 4
$145:
ldq $A1,0($17) # 4 1
mulq $A1,$19,$L1 # 4 2 1
subq $18,1,$18
umulh $A1,$19,$A1 # 4 2
addq $L1,$CC,$L1 # 4 2 3 1
addq $16,8,$16
cmpult $L1,$CC,$CC # 4 2 3 2
addq $17,8,$17
addq $A1,$CC,$CC # 4 3 2
stq $L1,-8($16) # 4 2 4
bgt $18,$145
ret $31,($26),1 # else exit
.align 4
$143:
addq $18,4,$18
bgt $18,$145 # goto tail code
ret $31,($26),1 # else exit
.end bn_mul_words
.align 3
.globl bn_sqr_words
.ent bn_sqr_words
bn_sqr_words:
bn_sqr_words..ng:
.frame $30,0,$26,0
.prologue 0
subq $18,4,$18
blt $18,$543 # if we are -1, -2, -3 or -4 goto tail code
ldq $A1,0($17) # 1 1
.align 3
$542:
mulq $A1,$A1,$L1 ######
ldq $A2,8($17) # 1 1
subq $18,4
umulh $A1,$A1,$R1 ######
ldq $A3,16($17) # 1 1
mulq $A2,$A2,$L2 ######
ldq $A4,24($17) # 1 1
stq $L1,0($16) # r[0]
umulh $A2,$A2,$R2 ######
stq $R1,8($16) # r[1]
mulq $A3,$A3,$L3 ######
stq $L2,16($16) # r[0]
umulh $A3,$A3,$R3 ######
stq $R2,24($16) # r[1]
mulq $A4,$A4,$L4 ######
stq $L3,32($16) # r[0]
umulh $A4,$A4,$R4 ######
stq $R3,40($16) # r[1]
addq $16,64,$16
addq $17,32,$17
stq $L4,-16($16) # r[0]
stq $R4,-8($16) # r[1]
blt $18,$543
ldq $A1,0($17) # 1 1
br $542
$442:
ldq $A1,0($17) # a[0]
mulq $A1,$A1,$L1 # a[0]*w low part r2
addq $16,16,$16
addq $17,8,$17
subq $18,1,$18
umulh $A1,$A1,$R1 # a[0]*w high part r3
stq $L1,-16($16) # r[0]
stq $R1,-8($16) # r[1]
bgt $18,$442
ret $31,($26),1 # else exit
.align 4
$543:
addq $18,4,$18
bgt $18,$442 # goto tail code
ret $31,($26),1 # else exit
.end bn_sqr_words
.align 3
.globl bn_add_words
.ent bn_add_words
bn_add_words:
bn_add_words..ng:
.frame $30,0,$26,0
.prologue 0
subq $19,4,$19
bis $31,$31,$CC # carry = 0
blt $19,$900
ldq $L1,0($17) # a[0]
ldq $R1,0($18) # b[1]
.align 3
$901:
addq $R1,$L1,$R1 # r=a+b;
ldq $L2,8($17) # a[1]
cmpult $R1,$L1,$O1 # did we overflow?
ldq $R2,8($18) # b[1]
addq $R1,$CC,$R1 # c+= overflow
ldq $L3,16($17) # a[2]
cmpult $R1,$CC,$CC # overflow?
ldq $R3,16($18) # b[2]
addq $CC,$O1,$CC
ldq $L4,24($17) # a[3]
addq $R2,$L2,$R2 # r=a+b;
ldq $R4,24($18) # b[3]
cmpult $R2,$L2,$O2 # did we overflow?
addq $R3,$L3,$R3 # r=a+b;
addq $R2,$CC,$R2 # c+= overflow
cmpult $R3,$L3,$O3 # did we overflow?
cmpult $R2,$CC,$CC # overflow?
addq $R4,$L4,$R4 # r=a+b;
addq $CC,$O2,$CC
cmpult $R4,$L4,$O4 # did we overflow?
addq $R3,$CC,$R3 # c+= overflow
stq $R1,0($16) # r[0]=c
cmpult $R3,$CC,$CC # overflow?
stq $R2,8($16) # r[1]=c
addq $CC,$O3,$CC
stq $R3,16($16) # r[2]=c
addq $R4,$CC,$R4 # c+= overflow
subq $19,4,$19 # loop--
cmpult $R4,$CC,$CC # overflow?
addq $17,32,$17 # a++
addq $CC,$O4,$CC
stq $R4,24($16) # r[3]=c
addq $18,32,$18 # b++
addq $16,32,$16 # r++
blt $19,$900
ldq $L1,0($17) # a[0]
ldq $R1,0($18) # b[1]
br $901
.align 4
$945:
ldq $L1,0($17) # a[0]
ldq $R1,0($18) # b[1]
addq $R1,$L1,$R1 # r=a+b;
subq $19,1,$19 # loop--
addq $R1,$CC,$R1 # c+= overflow
addq $17,8,$17 # a++
cmpult $R1,$L1,$O1 # did we overflow?
cmpult $R1,$CC,$CC # overflow?
addq $18,8,$18 # b++
stq $R1,0($16) # r[0]=c
addq $CC,$O1,$CC
addq $16,8,$16 # r++
bgt $19,$945
ret $31,($26),1 # else exit
$900:
addq $19,4,$19
bgt $19,$945 # goto tail code
ret $31,($26),1 # else exit
.end bn_add_words
.align 3
.globl bn_sub_words
.ent bn_sub_words
bn_sub_words:
bn_sub_words..ng:
.frame $30,0,$26,0
.prologue 0
subq $19,4,$19
bis $31,$31,$CC # carry = 0
br $800
blt $19,$800
ldq $L1,0($17) # a[0]
ldq $R1,0($18) # b[1]
.align 3
$801:
addq $R1,$L1,$R1 # r=a+b;
ldq $L2,8($17) # a[1]
cmpult $R1,$L1,$O1 # did we overflow?
ldq $R2,8($18) # b[1]
addq $R1,$CC,$R1 # c+= overflow
ldq $L3,16($17) # a[2]
cmpult $R1,$CC,$CC # overflow?
ldq $R3,16($18) # b[2]
addq $CC,$O1,$CC
ldq $L4,24($17) # a[3]
addq $R2,$L2,$R2 # r=a+b;
ldq $R4,24($18) # b[3]
cmpult $R2,$L2,$O2 # did we overflow?
addq $R3,$L3,$R3 # r=a+b;
addq $R2,$CC,$R2 # c+= overflow
cmpult $R3,$L3,$O3 # did we overflow?
cmpult $R2,$CC,$CC # overflow?
addq $R4,$L4,$R4 # r=a+b;
addq $CC,$O2,$CC
cmpult $R4,$L4,$O4 # did we overflow?
addq $R3,$CC,$R3 # c+= overflow
stq $R1,0($16) # r[0]=c
cmpult $R3,$CC,$CC # overflow?
stq $R2,8($16) # r[1]=c
addq $CC,$O3,$CC
stq $R3,16($16) # r[2]=c
addq $R4,$CC,$R4 # c+= overflow
subq $19,4,$19 # loop--
cmpult $R4,$CC,$CC # overflow?
addq $17,32,$17 # a++
addq $CC,$O4,$CC
stq $R4,24($16) # r[3]=c
addq $18,32,$18 # b++
addq $16,32,$16 # r++
blt $19,$800
ldq $L1,0($17) # a[0]
ldq $R1,0($18) # b[1]
br $801
.align 4
$845:
ldq $L1,0($17) # a[0]
ldq $R1,0($18) # b[1]
cmpult $L1,$R1,$O1 # will we borrow?
subq $L1,$R1,$R1 # r=a-b;
subq $19,1,$19 # loop--
cmpult $R1,$CC,$O2 # will we borrow?
subq $R1,$CC,$R1 # c+= overflow
addq $17,8,$17 # a++
addq $18,8,$18 # b++
stq $R1,0($16) # r[0]=c
addq $O2,$O1,$CC
addq $16,8,$16 # r++
bgt $19,$845
ret $31,($26),1 # else exit
$800:
addq $19,4,$19
bgt $19,$845 # goto tail code
ret $31,($26),1 # else exit
.end bn_sub_words
#
# What follows was taken directly from the C compiler with a few
# hacks to redo the lables.
#
.text
.align 3
.globl bn_div_words
.ent bn_div_words
bn_div_words:
ldgp $29,0($27)
bn_div_words..ng:
lda $30,-48($30)
.frame $30,48,$26,0
stq $26,0($30)
stq $9,8($30)
stq $10,16($30)
stq $11,24($30)
stq $12,32($30)
stq $13,40($30)
.mask 0x4003e00,-48
.prologue 1
bis $16,$16,$9
bis $17,$17,$10
bis $18,$18,$11
bis $31,$31,$13
bis $31,2,$12
bne $11,$119
lda $0,-1
br $31,$136
.align 4
$119:
bis $11,$11,$16
jsr $26,BN_num_bits_word
ldgp $29,0($26)
subq $0,64,$1
beq $1,$120
bis $31,1,$1
sll $1,$0,$1
cmpule $9,$1,$1
bne $1,$120
# lda $16,_IO_stderr_
# lda $17,$C32
# bis $0,$0,$18
# jsr $26,fprintf
# ldgp $29,0($26)
jsr $26,abort
ldgp $29,0($26)
.align 4
$120:
bis $31,64,$3
cmpult $9,$11,$2
subq $3,$0,$1
addl $1,$31,$0
subq $9,$11,$1
cmoveq $2,$1,$9
beq $0,$122
zapnot $0,15,$2
subq $3,$0,$1
sll $11,$2,$11
sll $9,$2,$3
srl $10,$1,$1
sll $10,$2,$10
bis $3,$1,$9
$122:
srl $11,32,$5
zapnot $11,15,$6
lda $7,-1
.align 5
$123:
srl $9,32,$1
subq $1,$5,$1
bne $1,$126
zapnot $7,15,$27
br $31,$127
.align 4
$126:
bis $9,$9,$24
bis $5,$5,$25
divqu $24,$25,$27
$127:
srl $10,32,$4
.align 5
$128:
mulq $27,$5,$1
subq $9,$1,$3
zapnot $3,240,$1
bne $1,$129
mulq $6,$27,$2
sll $3,32,$1
addq $1,$4,$1
cmpule $2,$1,$2
bne $2,$129
subq $27,1,$27
br $31,$128
.align 4
$129:
mulq $27,$6,$1
mulq $27,$5,$4
srl $1,32,$3
sll $1,32,$1
addq $4,$3,$4
cmpult $10,$1,$2
subq $10,$1,$10
addq $2,$4,$2
cmpult $9,$2,$1
bis $2,$2,$4
beq $1,$134
addq $9,$11,$9
subq $27,1,$27
$134:
subl $12,1,$12
subq $9,$4,$9
beq $12,$124
sll $27,32,$13
sll $9,32,$2
srl $10,32,$1
sll $10,32,$10
bis $2,$1,$9
br $31,$123
.align 4
$124:
bis $13,$27,$0
$136:
ldq $26,0($30)
ldq $9,8($30)
ldq $10,16($30)
ldq $11,24($30)
ldq $12,32($30)
ldq $13,40($30)
addq $30,48,$30
ret $31,($26),1
.end bn_div_words
EOF
return($data);
}

View File

@ -1,33 +0,0 @@
#!/usr/local/bin/perl
# I have this in perl so I can use more usefull register names and then convert
# them into alpha registers.
#
push(@INC,"perlasm","../../perlasm");
require "alpha.pl";
require "alpha/mul_add.pl";
require "alpha/mul.pl";
require "alpha/sqr.pl";
require "alpha/add.pl";
require "alpha/sub.pl";
require "alpha/mul_c8.pl";
require "alpha/mul_c4.pl";
require "alpha/sqr_c4.pl";
require "alpha/sqr_c8.pl";
require "alpha/div.pl";
&asm_init($ARGV[0],$0);
&bn_mul_words("bn_mul_words");
&bn_sqr_words("bn_sqr_words");
&bn_mul_add_words("bn_mul_add_words");
&bn_add_words("bn_add_words");
&bn_sub_words("bn_sub_words");
&bn_div_words("bn_div_words");
&bn_mul_comba8("bn_mul_comba8");
&bn_mul_comba4("bn_mul_comba4");
&bn_sqr_comba4("bn_sqr_comba4");
&bn_sqr_comba8("bn_sqr_comba8");
&asm_finish();

View File

@ -1,116 +0,0 @@
#!/usr/local/bin/perl
# I have this in perl so I can use more usefull register names and then convert
# them into alpha registers.
#
push(@INC,"perlasm","../../perlasm");
require "alpha.pl";
&asm_init($ARGV[0],$0);
print &bn_sub_words("bn_sub_words");
&asm_finish();
sub bn_sub_words
{
local($name)=@_;
local($cc,$a,$b,$r);
$cc="r0";
$a0="r1"; $b0="r5"; $r0="r9"; $tmp="r13";
$a1="r2"; $b1="r6"; $r1="r10"; $t1="r14";
$a2="r3"; $b2="r7"; $r2="r11";
$a3="r4"; $b3="r8"; $r3="r12"; $t3="r15";
$rp=&wparam(0);
$ap=&wparam(1);
$bp=&wparam(2);
$count=&wparam(3);
&function_begin($name,"");
&comment("");
&sub($count,4,$count);
&mov("zero",$cc);
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
##########################################################
&set_label("loop");
&ld($a1,&QWPw(1,$ap));
&cmpult($a0,$b0,$tmp); # will we borrow?
&ld($b1,&QWPw(1,$bp));
&sub($a0,$b0,$a0); # do the subtract
&ld($a2,&QWPw(2,$ap));
&cmpult($a0,$cc,$b0); # will we borrow?
&ld($b2,&QWPw(2,$bp));
&sub($a0,$cc,$a0); # will we borrow?
&ld($a3,&QWPw(3,$ap));
&add($b0,$tmp,$cc); # add the borrows
&cmpult($a1,$b1,$t1); # will we borrow?
&sub($a1,$b1,$a1); # do the subtract
&ld($b3,&QWPw(3,$bp));
&cmpult($a1,$cc,$b1); # will we borrow?
&sub($a1,$cc,$a1); # will we borrow?
&add($b1,$t1,$cc); # add the borrows
&cmpult($a2,$b2,$tmp); # will we borrow?
&sub($a2,$b2,$a2); # do the subtract
&st($a0,&QWPw(0,$rp)); # save
&cmpult($a2,$cc,$b2); # will we borrow?
&sub($a2,$cc,$a2); # will we borrow?
&add($b2,$tmp,$cc); # add the borrows
&cmpult($a3,$b3,$t3); # will we borrow?
&sub($a3,$b3,$a3); # do the subtract
&st($a1,&QWPw(1,$rp)); # save
&cmpult($a3,$cc,$b3); # will we borrow?
&sub($a3,$cc,$a3); # will we borrow?
&add($b3,$t3,$cc); # add the borrows
&st($a2,&QWPw(2,$rp)); # save
&sub($count,4,$count); # count-=4
&st($a3,&QWPw(3,$rp)); # save
&add($ap,4*$QWS,$ap); # count+=4
&add($bp,4*$QWS,$bp); # count+=4
&add($rp,4*$QWS,$rp); # count+=4
&blt($count,&label("finish"));
&ld($a0,&QWPw(0,$ap));
&ld($b0,&QWPw(0,$bp));
&br(&label("loop"));
##################################################
# Do the last 0..3 words
&set_label("last_loop");
&ld($a0,&QWPw(0,$ap)); # get a
&ld($b0,&QWPw(0,$bp)); # get b
&cmpult($a0,$b0,$tmp); # will we borrow?
&sub($a0,$b0,$a0); # do the subtract
&cmpult($a0,$cc,$b0); # will we borrow?
&sub($a0,$cc,$a0); # will we borrow?
&st($a0,&QWPw(0,$rp)); # save
&add($b0,$tmp,$cc); # add the borrows
&add($ap,$QWS,$ap);
&add($bp,$QWS,$bp);
&add($rp,$QWS,$rp);
&sub($count,1,$count);
&bgt($count,&label("last_loop"));
&function_end_A($name);
######################################################
&set_label("finish");
&add($count,4,$count);
&bgt($count,&label("last_loop"));
&set_label("end");
&function_end($name);
}

View File

@ -1,539 +0,0 @@
/* This assember is for R2000/R3000 machines, or higher ones that do
* no want to do any 64 bit arithmatic.
* Make sure that the SSLeay bignum library is compiled with
* THIRTY_TWO_BIT set.
* This must either be compiled with the system CC, or, if you use GNU gas,
* cc -E mips1.s|gas -o mips1.o
*/
.set reorder
.set noat
#define R1 $1
#define CC $2
#define R2 $3
#define R3 $8
#define R4 $9
#define L1 $10
#define L2 $11
#define L3 $12
#define L4 $13
#define H1 $14
#define H2 $15
#define H3 $24
#define H4 $25
#define P1 $4
#define P2 $5
#define P3 $6
#define P4 $7
.align 2
.ent bn_mul_add_words
.globl bn_mul_add_words
.text
bn_mul_add_words:
.frame $sp,0,$31
.mask 0x00000000,0
.fmask 0x00000000,0
#blt P3,4,$lab34
subu R1,P3,4
move CC,$0
bltz R1,$lab34
$lab2:
lw R1,0(P1)
lw L1,0(P2)
lw R2,4(P1)
lw L2,4(P2)
lw R3,8(P1)
lw L3,8(P2)
lw R4,12(P1)
lw L4,12(P2)
multu L1,P4
addu R1,R1,CC
mflo L1
sltu CC,R1,CC
addu R1,R1,L1
mfhi H1
sltu L1,R1,L1
sw R1,0(P1)
addu CC,CC,L1
multu L2,P4
addu CC,H1,CC
mflo L2
addu R2,R2,CC
sltu CC,R2,CC
mfhi H2
addu R2,R2,L2
addu P2,P2,16
sltu L2,R2,L2
sw R2,4(P1)
addu CC,CC,L2
multu L3,P4
addu CC,H2,CC
mflo L3
addu R3,R3,CC
sltu CC,R3,CC
mfhi H3
addu R3,R3,L3
addu P1,P1,16
sltu L3,R3,L3
sw R3,-8(P1)
addu CC,CC,L3
multu L4,P4
addu CC,H3,CC
mflo L4
addu R4,R4,CC
sltu CC,R4,CC
mfhi H4
addu R4,R4,L4
subu P3,P3,4
sltu L4,R4,L4
addu CC,CC,L4
addu CC,H4,CC
subu R1,P3,4
sw R4,-4(P1) # delay slot
bgez R1,$lab2
bleu P3,0,$lab3
.align 2
$lab33:
lw L1,0(P2)
lw R1,0(P1)
multu L1,P4
addu R1,R1,CC
sltu CC,R1,CC
addu P1,P1,4
mflo L1
mfhi H1
addu R1,R1,L1
addu P2,P2,4
sltu L1,R1,L1
subu P3,P3,1
addu CC,CC,L1
sw R1,-4(P1)
addu CC,H1,CC
bgtz P3,$lab33
j $31
.align 2
$lab3:
j $31
.align 2
$lab34:
bgt P3,0,$lab33
j $31
.end bn_mul_add_words
.align 2
# Program Unit: bn_mul_words
.ent bn_mul_words
.globl bn_mul_words
.text
bn_mul_words:
.frame $sp,0,$31
.mask 0x00000000,0
.fmask 0x00000000,0
subu P3,P3,4
move CC,$0
bltz P3,$lab45
$lab44:
lw L1,0(P2)
lw L2,4(P2)
lw L3,8(P2)
lw L4,12(P2)
multu L1,P4
subu P3,P3,4
mflo L1
mfhi H1
addu L1,L1,CC
multu L2,P4
sltu CC,L1,CC
sw L1,0(P1)
addu CC,H1,CC
mflo L2
mfhi H2
addu L2,L2,CC
multu L3,P4
sltu CC,L2,CC
sw L2,4(P1)
addu CC,H2,CC
mflo L3
mfhi H3
addu L3,L3,CC
multu L4,P4
sltu CC,L3,CC
sw L3,8(P1)
addu CC,H3,CC
mflo L4
mfhi H4
addu L4,L4,CC
addu P1,P1,16
sltu CC,L4,CC
addu P2,P2,16
addu CC,H4,CC
sw L4,-4(P1)
bgez P3,$lab44
b $lab45
$lab46:
lw L1,0(P2)
addu P1,P1,4
multu L1,P4
addu P2,P2,4
mflo L1
mfhi H1
addu L1,L1,CC
subu P3,P3,1
sltu CC,L1,CC
sw L1,-4(P1)
addu CC,H1,CC
bgtz P3,$lab46
j $31
$lab45:
addu P3,P3,4
bgtz P3,$lab46
j $31
.align 2
.end bn_mul_words
# Program Unit: bn_sqr_words
.ent bn_sqr_words
.globl bn_sqr_words
.text
bn_sqr_words:
.frame $sp,0,$31
.mask 0x00000000,0
.fmask 0x00000000,0
subu P3,P3,4
bltz P3,$lab55
$lab54:
lw L1,0(P2)
lw L2,4(P2)
lw L3,8(P2)
lw L4,12(P2)
multu L1,L1
subu P3,P3,4
mflo L1
mfhi H1
sw L1,0(P1)
sw H1,4(P1)
multu L2,L2
addu P1,P1,32
mflo L2
mfhi H2
sw L2,-24(P1)
sw H2,-20(P1)
multu L3,L3
addu P2,P2,16
mflo L3
mfhi H3
sw L3,-16(P1)
sw H3,-12(P1)
multu L4,L4
mflo L4
mfhi H4
sw L4,-8(P1)
sw H4,-4(P1)
bgtz P3,$lab54
b $lab55
$lab56:
lw L1,0(P2)
addu P1,P1,8
multu L1,L1
addu P2,P2,4
subu P3,P3,1
mflo L1
mfhi H1
sw L1,-8(P1)
sw H1,-4(P1)
bgtz P3,$lab56
j $31
$lab55:
addu P3,P3,4
bgtz P3,$lab56
j $31
.align 2
.end bn_sqr_words
# Program Unit: bn_add_words
.ent bn_add_words
.globl bn_add_words
.text
bn_add_words: # 0x590
.frame $sp,0,$31
.mask 0x00000000,0
.fmask 0x00000000,0
subu P4,P4,4
move CC,$0
bltz P4,$lab65
$lab64:
lw L1,0(P2)
lw R1,0(P3)
lw L2,4(P2)
lw R2,4(P3)
addu L1,L1,CC
lw L3,8(P2)
sltu CC,L1,CC
addu L1,L1,R1
sltu R1,L1,R1
lw R3,8(P3)
addu CC,CC,R1
lw L4,12(P2)
addu L2,L2,CC
lw R4,12(P3)
sltu CC,L2,CC
addu L2,L2,R2
sltu R2,L2,R2
sw L1,0(P1)
addu CC,CC,R2
addu P1,P1,16
addu L3,L3,CC
sw L2,-12(P1)
sltu CC,L3,CC
addu L3,L3,R3
sltu R3,L3,R3
addu P2,P2,16
addu CC,CC,R3
addu L4,L4,CC
addu P3,P3,16
sltu CC,L4,CC
addu L4,L4,R4
subu P4,P4,4
sltu R4,L4,R4
sw L3,-8(P1)
addu CC,CC,R4
sw L4,-4(P1)
bgtz P4,$lab64
b $lab65
$lab66:
lw L1,0(P2)
lw R1,0(P3)
addu L1,L1,CC
addu P1,P1,4
sltu CC,L1,CC
addu P2,P2,4
addu P3,P3,4
addu L1,L1,R1
subu P4,P4,1
sltu R1,L1,R1
sw L1,-4(P1)
addu CC,CC,R1
bgtz P4,$lab66
j $31
$lab65:
addu P4,P4,4
bgtz P4,$lab66
j $31
.end bn_add_words
# Program Unit: bn_div64
.set at
.set reorder
.text
.align 2
.globl bn_div64
# 321 {
.ent bn_div64 2
bn_div64:
subu $sp, 64
sw $31, 56($sp)
sw $16, 48($sp)
.mask 0x80010000, -56
.frame $sp, 64, $31
move $9, $4
move $12, $5
move $16, $6
# 322 BN_ULONG dh,dl,q,ret=0,th,tl,t;
move $31, $0
# 323 int i,count=2;
li $13, 2
# 324
# 325 if (d == 0) return(BN_MASK2);
bne $16, 0, $80
li $2, -1
b $93
$80:
# 326
# 327 i=BN_num_bits_word(d);
move $4, $16
sw $31, 16($sp)
sw $9, 24($sp)
sw $12, 32($sp)
sw $13, 40($sp)
.livereg 0x800ff0e,0xfff
jal BN_num_bits_word
li $4, 32
lw $31, 16($sp)
lw $9, 24($sp)
lw $12, 32($sp)
lw $13, 40($sp)
move $3, $2
# 328 if ((i != BN_BITS2) && (h > (BN_ULONG)1<<i))
beq $2, $4, $81
li $14, 1
sll $15, $14, $2
bleu $9, $15, $81
# 329 {
# 330 #if !defined(NO_STDIO) && !defined(WIN16)
# 331 fprintf(stderr,"Division would overflow (%d)\n",i);
# 332 #endif
# 333 abort();
sw $3, 8($sp)
sw $9, 24($sp)
sw $12, 32($sp)
sw $13, 40($sp)
sw $31, 26($sp)
.livereg 0xff0e,0xfff
jal abort
lw $3, 8($sp)
li $4, 32
lw $9, 24($sp)
lw $12, 32($sp)
lw $13, 40($sp)
lw $31, 26($sp)
# 334 }
$81:
# 335 i=BN_BITS2-i;
subu $3, $4, $3
# 336 if (h >= d) h-=d;
bltu $9, $16, $82
subu $9, $9, $16
$82:
# 337
# 338 if (i)
beq $3, 0, $83
# 339 {
# 340 d<<=i;
sll $16, $16, $3
# 341 h=(h<<i)|(l>>(BN_BITS2-i));
sll $24, $9, $3
subu $25, $4, $3
srl $14, $12, $25
or $9, $24, $14
# 342 l<<=i;
sll $12, $12, $3
# 343 }
$83:
# 344 dh=(d&BN_MASK2h)>>BN_BITS4;
# 345 dl=(d&BN_MASK2l);
and $8, $16, -65536
srl $8, $8, 16
and $10, $16, 65535
li $6, -65536
$84:
# 346 for (;;)
# 347 {
# 348 if ((h>>BN_BITS4) == dh)
srl $15, $9, 16
bne $8, $15, $85
# 349 q=BN_MASK2l;
li $5, 65535
b $86
$85:
# 350 else
# 351 q=h/dh;
divu $5, $9, $8
$86:
# 352
# 353 for (;;)
# 354 {
# 355 t=(h-q*dh);
mul $4, $5, $8
subu $2, $9, $4
move $3, $2
# 356 if ((t&BN_MASK2h) ||
# 357 ((dl*q) <= (
# 358 (t<<BN_BITS4)+
# 359 ((l&BN_MASK2h)>>BN_BITS4))))
and $25, $2, $6
bne $25, $0, $87
mul $24, $10, $5
sll $14, $3, 16
and $15, $12, $6
srl $25, $15, 16
addu $15, $14, $25
bgtu $24, $15, $88
$87:
# 360 break;
mul $3, $10, $5
b $89
$88:
# 361 q--;
addu $5, $5, -1
# 362 }
b $86
$89:
# 363 th=q*dh;
# 364 tl=q*dl;
# 365 t=(tl>>BN_BITS4);
# 366 tl=(tl<<BN_BITS4)&BN_MASK2h;
sll $14, $3, 16
and $2, $14, $6
move $11, $2
# 367 th+=t;
srl $25, $3, 16
addu $7, $4, $25
# 368
# 369 if (l < tl) th++;
bgeu $12, $2, $90
addu $7, $7, 1
$90:
# 370 l-=tl;
subu $12, $12, $11
# 371 if (h < th)
bgeu $9, $7, $91
# 372 {
# 373 h+=d;
addu $9, $9, $16
# 374 q--;
addu $5, $5, -1
# 375 }
$91:
# 376 h-=th;
subu $9, $9, $7
# 377
# 378 if (--count == 0) break;
addu $13, $13, -1
beq $13, 0, $92
# 379
# 380 ret=q<<BN_BITS4;
sll $31, $5, 16
# 381 h=((h<<BN_BITS4)|(l>>BN_BITS4))&BN_MASK2;
sll $24, $9, 16
srl $15, $12, 16
or $9, $24, $15
# 382 l=(l&BN_MASK2l)<<BN_BITS4;
and $12, $12, 65535
sll $12, $12, 16
# 383 }
b $84
$92:
# 384 ret|=q;
or $31, $31, $5
# 385 return(ret);
move $2, $31
$93:
lw $16, 48($sp)
lw $31, 56($sp)
addu $sp, 64
j $31
.end bn_div64

View File

@ -1,710 +0,0 @@
.SPACE $PRIVATE$
.SUBSPA $DATA$,QUAD=1,ALIGN=8,ACCESS=31
.SUBSPA $BSS$,QUAD=1,ALIGN=8,ACCESS=31,ZERO,SORT=82
.SPACE $TEXT$
.SUBSPA $LIT$,QUAD=0,ALIGN=8,ACCESS=44
.SUBSPA $CODE$,QUAD=0,ALIGN=8,ACCESS=44,CODE_ONLY
.IMPORT $global$,DATA
.IMPORT $$dyncall,MILLICODE
; gcc_compiled.:
.SPACE $TEXT$
.SUBSPA $CODE$
.align 4
.EXPORT bn_mul_add_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR
bn_mul_add_words
.PROC
.CALLINFO FRAME=0,CALLS,SAVE_RP
.ENTRY
stw %r2,-20(0,%r30)
ldi 0,%r28
extru %r23,31,16,%r2
stw %r2,-16(0,%r30)
extru %r23,15,16,%r23
ldil L'65536,%r31
fldws -16(0,%r30),%fr11R
stw %r23,-16(0,%r30)
ldo 12(%r25),%r29
ldo 12(%r26),%r23
fldws -16(0,%r30),%fr11L
L$0002
ldw 0(0,%r25),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0005
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi 1,%r19,%r19
ldw 0(0,%r26),%r28
addl %r20,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0003
stw %r20,0(0,%r26)
ldw -8(0,%r29),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0010
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi 1,%r19,%r19
ldw -8(0,%r23),%r28
addl %r20,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0003
stw %r20,-8(0,%r23)
ldw -4(0,%r29),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0015
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi 1,%r19,%r19
ldw -4(0,%r23),%r28
addl %r20,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0003
stw %r20,-4(0,%r23)
ldw 0(0,%r29),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0020
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi 1,%r19,%r19
ldw 0(0,%r23),%r28
addl %r20,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0003
stw %r20,0(0,%r23)
ldo 16(%r29),%r29
ldo 16(%r25),%r25
ldo 16(%r23),%r23
bl L$0002,0
ldo 16(%r26),%r26
L$0003
ldw -20(0,%r30),%r2
bv,n 0(%r2)
.EXIT
.PROCEND
.align 4
.EXPORT bn_mul_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR
bn_mul_words
.PROC
.CALLINFO FRAME=0,CALLS,SAVE_RP
.ENTRY
stw %r2,-20(0,%r30)
ldi 0,%r28
extru %r23,31,16,%r2
stw %r2,-16(0,%r30)
extru %r23,15,16,%r23
ldil L'65536,%r31
fldws -16(0,%r30),%fr11R
stw %r23,-16(0,%r30)
ldo 12(%r26),%r29
ldo 12(%r25),%r23
fldws -16(0,%r30),%fr11L
L$0026
ldw 0(0,%r25),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0029
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0027
stw %r20,0(0,%r26)
ldw -8(0,%r23),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0033
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0027
stw %r20,-8(0,%r29)
ldw -4(0,%r23),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0037
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0027
stw %r20,-4(0,%r29)
ldw 0(0,%r23),%r19
extru %r19,31,16,%r20
stw %r20,-16(0,%r30)
extru %r19,15,16,%r19
fldws -16(0,%r30),%fr22L
stw %r19,-16(0,%r30)
xmpyu %fr22L,%fr11R,%fr8
fldws -16(0,%r30),%fr22L
fstws %fr8R,-16(0,%r30)
xmpyu %fr11R,%fr22L,%fr10
ldw -16(0,%r30),%r2
stw %r20,-16(0,%r30)
xmpyu %fr22L,%fr11L,%fr9
fldws -16(0,%r30),%fr22L
fstws %fr10R,-16(0,%r30)
copy %r2,%r22
ldw -16(0,%r30),%r2
fstws %fr9R,-16(0,%r30)
xmpyu %fr11L,%fr22L,%fr8
copy %r2,%r19
ldw -16(0,%r30),%r2
fstws %fr8R,-16(0,%r30)
copy %r2,%r20
ldw -16(0,%r30),%r2
addl %r2,%r19,%r21
comclr,<<= %r19,%r21,0
addl %r20,%r31,%r20
L$0041
extru %r21,15,16,%r19
addl %r20,%r19,%r20
zdep %r21,15,16,%r19
addl %r22,%r19,%r22
comclr,<<= %r19,%r22,0
addi,tr 1,%r20,%r19
copy %r20,%r19
addl %r22,%r28,%r20
comclr,<<= %r28,%r20,0
addi,tr 1,%r19,%r28
copy %r19,%r28
addib,= -1,%r24,L$0027
stw %r20,0(0,%r29)
ldo 16(%r23),%r23
ldo 16(%r25),%r25
ldo 16(%r29),%r29
bl L$0026,0
ldo 16(%r26),%r26
L$0027
ldw -20(0,%r30),%r2
bv,n 0(%r2)
.EXIT
.PROCEND
.align 4
.EXPORT bn_sqr_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR
bn_sqr_words
.PROC
.CALLINFO FRAME=0,NO_CALLS
.ENTRY
ldo 28(%r26),%r23
ldo 12(%r25),%r28
L$0046
ldw 0(0,%r25),%r21
extru %r21,31,16,%r22
stw %r22,-16(0,%r30)
extru %r21,15,16,%r21
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
ldw -16(0,%r30),%r29
stw %r22,-16(0,%r30)
fldws -16(0,%r30),%fr10R
stw %r21,-16(0,%r30)
copy %r29,%r19
xmpyu %fr10L,%fr10R,%fr8
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
fstws %fr8R,-16(0,%r30)
extru %r19,16,17,%r20
zdep %r19,14,15,%r19
ldw -16(0,%r30),%r29
xmpyu %fr10L,%fr10R,%fr9
addl %r29,%r19,%r22
stw %r22,0(0,%r26)
fstws %fr9R,-16(0,%r30)
ldw -16(0,%r30),%r29
addl %r29,%r20,%r21
comclr,<<= %r19,%r22,0
addi 1,%r21,%r21
addib,= -1,%r24,L$0057
stw %r21,-24(0,%r23)
ldw -8(0,%r28),%r21
extru %r21,31,16,%r22
stw %r22,-16(0,%r30)
extru %r21,15,16,%r21
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
ldw -16(0,%r30),%r29
stw %r22,-16(0,%r30)
fldws -16(0,%r30),%fr10R
stw %r21,-16(0,%r30)
copy %r29,%r19
xmpyu %fr10L,%fr10R,%fr8
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
fstws %fr8R,-16(0,%r30)
extru %r19,16,17,%r20
zdep %r19,14,15,%r19
ldw -16(0,%r30),%r29
xmpyu %fr10L,%fr10R,%fr9
addl %r29,%r19,%r22
stw %r22,-20(0,%r23)
fstws %fr9R,-16(0,%r30)
ldw -16(0,%r30),%r29
addl %r29,%r20,%r21
comclr,<<= %r19,%r22,0
addi 1,%r21,%r21
addib,= -1,%r24,L$0057
stw %r21,-16(0,%r23)
ldw -4(0,%r28),%r21
extru %r21,31,16,%r22
stw %r22,-16(0,%r30)
extru %r21,15,16,%r21
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
ldw -16(0,%r30),%r29
stw %r22,-16(0,%r30)
fldws -16(0,%r30),%fr10R
stw %r21,-16(0,%r30)
copy %r29,%r19
xmpyu %fr10L,%fr10R,%fr8
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
fstws %fr8R,-16(0,%r30)
extru %r19,16,17,%r20
zdep %r19,14,15,%r19
ldw -16(0,%r30),%r29
xmpyu %fr10L,%fr10R,%fr9
addl %r29,%r19,%r22
stw %r22,-12(0,%r23)
fstws %fr9R,-16(0,%r30)
ldw -16(0,%r30),%r29
addl %r29,%r20,%r21
comclr,<<= %r19,%r22,0
addi 1,%r21,%r21
addib,= -1,%r24,L$0057
stw %r21,-8(0,%r23)
ldw 0(0,%r28),%r21
extru %r21,31,16,%r22
stw %r22,-16(0,%r30)
extru %r21,15,16,%r21
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
ldw -16(0,%r30),%r29
stw %r22,-16(0,%r30)
fldws -16(0,%r30),%fr10R
stw %r21,-16(0,%r30)
copy %r29,%r19
xmpyu %fr10L,%fr10R,%fr8
fldws -16(0,%r30),%fr10L
stw %r21,-16(0,%r30)
fldws -16(0,%r30),%fr10R
fstws %fr8R,-16(0,%r30)
extru %r19,16,17,%r20
zdep %r19,14,15,%r19
ldw -16(0,%r30),%r29
xmpyu %fr10L,%fr10R,%fr9
addl %r29,%r19,%r22
stw %r22,-4(0,%r23)
fstws %fr9R,-16(0,%r30)
ldw -16(0,%r30),%r29
addl %r29,%r20,%r21
comclr,<<= %r19,%r22,0
addi 1,%r21,%r21
addib,= -1,%r24,L$0057
stw %r21,0(0,%r23)
ldo 16(%r28),%r28
ldo 16(%r25),%r25
ldo 32(%r23),%r23
bl L$0046,0
ldo 32(%r26),%r26
L$0057
bv,n 0(%r2)
.EXIT
.PROCEND
.IMPORT BN_num_bits_word,CODE
.IMPORT fprintf,CODE
.IMPORT __iob,DATA
.SPACE $TEXT$
.SUBSPA $LIT$
.align 4
L$C0000
.STRING "Division would overflow\x0a\x00"
.IMPORT abort,CODE
.SPACE $TEXT$
.SUBSPA $CODE$
.align 4
.EXPORT bn_div64,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,RTNVAL=GR
bn_div64
.PROC
.CALLINFO FRAME=128,CALLS,SAVE_RP,ENTRY_GR=8
.ENTRY
stw %r2,-20(0,%r30)
stwm %r8,128(0,%r30)
stw %r7,-124(0,%r30)
stw %r4,-112(0,%r30)
stw %r3,-108(0,%r30)
copy %r26,%r3
copy %r25,%r4
stw %r6,-120(0,%r30)
ldi 0,%r7
stw %r5,-116(0,%r30)
movb,<> %r24,%r5,L$0059
ldi 2,%r6
bl L$0076,0
ldi -1,%r28
L$0059
.CALL ARGW0=GR
bl BN_num_bits_word,%r2
copy %r5,%r26
ldi 32,%r19
comb,= %r19,%r28,L$0060
subi 31,%r28,%r19
mtsar %r19
zvdepi 1,32,%r19
comb,>>= %r19,%r3,L$0060
addil LR'__iob-$global$+32,%r27
ldo RR'__iob-$global$+32(%r1),%r26
ldil LR'L$C0000,%r25
.CALL ARGW0=GR,ARGW1=GR
bl fprintf,%r2
ldo RR'L$C0000(%r25),%r25
.CALL
bl abort,%r2
nop
L$0060
comb,>> %r5,%r3,L$0061
subi 32,%r28,%r28
sub %r3,%r5,%r3
L$0061
comib,= 0,%r28,L$0062
subi 31,%r28,%r19
mtsar %r19
zvdep %r5,32,%r5
zvdep %r3,32,%r21
subi 32,%r28,%r20
mtsar %r20
vshd 0,%r4,%r20
or %r21,%r20,%r3
mtsar %r19
zvdep %r4,32,%r4
L$0062
extru %r5,15,16,%r23
extru %r5,31,16,%r28
L$0063
extru %r3,15,16,%r19
comb,<> %r23,%r19,L$0066
copy %r3,%r26
bl L$0067,0
zdepi -1,31,16,%r29
L$0066
.IMPORT $$divU,MILLICODE
bl $$divU,%r31
copy %r23,%r25
L$0067
stw %r29,-16(0,%r30)
fldws -16(0,%r30),%fr10L
stw %r28,-16(0,%r30)
fldws -16(0,%r30),%fr10R
stw %r23,-16(0,%r30)
xmpyu %fr10L,%fr10R,%fr8
fldws -16(0,%r30),%fr10R
fstws %fr8R,-16(0,%r30)
xmpyu %fr10L,%fr10R,%fr9
ldw -16(0,%r30),%r8
fstws %fr9R,-16(0,%r30)
copy %r8,%r22
ldw -16(0,%r30),%r8
extru %r4,15,16,%r24
copy %r8,%r21
L$0068
sub %r3,%r21,%r20
copy %r20,%r19
depi 0,31,16,%r19
comib,<> 0,%r19,L$0069
zdep %r20,15,16,%r19
addl %r19,%r24,%r19
comb,>>= %r19,%r22,L$0069
sub %r22,%r28,%r22
sub %r21,%r23,%r21
bl L$0068,0
ldo -1(%r29),%r29
L$0069
stw %r29,-16(0,%r30)
fldws -16(0,%r30),%fr10L
stw %r28,-16(0,%r30)
fldws -16(0,%r30),%fr10R
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
ldw -16(0,%r30),%r8
stw %r23,-16(0,%r30)
fldws -16(0,%r30),%fr10R
copy %r8,%r19
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
extru %r19,15,16,%r20
ldw -16(0,%r30),%r8
zdep %r19,15,16,%r19
addl %r8,%r20,%r20
comclr,<<= %r19,%r4,0
addi 1,%r20,%r20
comb,<<= %r20,%r3,L$0074
sub %r4,%r19,%r4
addl %r3,%r5,%r3
ldo -1(%r29),%r29
L$0074
addib,= -1,%r6,L$0064
sub %r3,%r20,%r3
zdep %r29,15,16,%r7
shd %r3,%r4,16,%r3
bl L$0063,0
zdep %r4,15,16,%r4
L$0064
or %r7,%r29,%r28
L$0076
ldw -148(0,%r30),%r2
ldw -124(0,%r30),%r7
ldw -120(0,%r30),%r6
ldw -116(0,%r30),%r5
ldw -112(0,%r30),%r4
ldw -108(0,%r30),%r3
bv 0(%r2)
ldwm -128(0,%r30),%r8
.EXIT
.PROCEND

View File

@ -1,416 +0,0 @@
.SPACE $PRIVATE$
.SUBSPA $DATA$,QUAD=1,ALIGN=8,ACCESS=31
.SUBSPA $BSS$,QUAD=1,ALIGN=8,ACCESS=31,ZERO,SORT=82
.SPACE $TEXT$
.SUBSPA $LIT$,QUAD=0,ALIGN=8,ACCESS=44
.SUBSPA $CODE$,QUAD=0,ALIGN=8,ACCESS=44,CODE_ONLY
.IMPORT $global$,DATA
.IMPORT $$dyncall,MILLICODE
; gcc_compiled.:
.SPACE $TEXT$
.SUBSPA $CODE$
.align 4
.EXPORT bn_mul_add_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR
bn_mul_add_words
.PROC
.CALLINFO FRAME=64,CALLS,SAVE_RP,ENTRY_GR=4
.ENTRY
stw %r2,-20(0,%r30)
stwm %r4,64(0,%r30)
copy %r24,%r31
stw %r3,-60(0,%r30)
ldi 0,%r20
ldo 12(%r26),%r2
stw %r23,-16(0,%r30)
copy %r25,%r3
ldo 12(%r3),%r1
fldws -16(0,%r30),%fr8L
L$0010
copy %r20,%r25
ldi 0,%r24
fldws 0(0,%r3),%fr9L
ldw 0(0,%r26),%r19
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r19,%r23
ldw -16(0,%r30),%r28
ldw -12(0,%r30),%r29
ldi 0,%r22
add %r23,%r29,%r29
addc %r22,%r28,%r28
add %r25,%r29,%r29
addc %r24,%r28,%r28
copy %r28,%r21
ldi 0,%r20
copy %r21,%r20
addib,= -1,%r31,L$0011
stw %r29,0(0,%r26)
copy %r20,%r25
ldi 0,%r24
fldws -8(0,%r1),%fr9L
ldw -8(0,%r2),%r19
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r19,%r23
ldw -16(0,%r30),%r28
ldw -12(0,%r30),%r29
ldi 0,%r22
add %r23,%r29,%r29
addc %r22,%r28,%r28
add %r25,%r29,%r29
addc %r24,%r28,%r28
copy %r28,%r21
ldi 0,%r20
copy %r21,%r20
addib,= -1,%r31,L$0011
stw %r29,-8(0,%r2)
copy %r20,%r25
ldi 0,%r24
fldws -4(0,%r1),%fr9L
ldw -4(0,%r2),%r19
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r19,%r23
ldw -16(0,%r30),%r28
ldw -12(0,%r30),%r29
ldi 0,%r22
add %r23,%r29,%r29
addc %r22,%r28,%r28
add %r25,%r29,%r29
addc %r24,%r28,%r28
copy %r28,%r21
ldi 0,%r20
copy %r21,%r20
addib,= -1,%r31,L$0011
stw %r29,-4(0,%r2)
copy %r20,%r25
ldi 0,%r24
fldws 0(0,%r1),%fr9L
ldw 0(0,%r2),%r19
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r19,%r23
ldw -16(0,%r30),%r28
ldw -12(0,%r30),%r29
ldi 0,%r22
add %r23,%r29,%r29
addc %r22,%r28,%r28
add %r25,%r29,%r29
addc %r24,%r28,%r28
copy %r28,%r21
ldi 0,%r20
copy %r21,%r20
addib,= -1,%r31,L$0011
stw %r29,0(0,%r2)
ldo 16(%r1),%r1
ldo 16(%r3),%r3
ldo 16(%r2),%r2
bl L$0010,0
ldo 16(%r26),%r26
L$0011
copy %r20,%r28
ldw -84(0,%r30),%r2
ldw -60(0,%r30),%r3
bv 0(%r2)
ldwm -64(0,%r30),%r4
.EXIT
.PROCEND
.align 4
.EXPORT bn_mul_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR
bn_mul_words
.PROC
.CALLINFO FRAME=64,CALLS,SAVE_RP,ENTRY_GR=3
.ENTRY
stw %r2,-20(0,%r30)
copy %r25,%r2
stwm %r4,64(0,%r30)
copy %r24,%r19
ldi 0,%r28
stw %r23,-16(0,%r30)
ldo 12(%r26),%r31
ldo 12(%r2),%r29
fldws -16(0,%r30),%fr8L
L$0026
fldws 0(0,%r2),%fr9L
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r28,%r21
ldi 0,%r20
ldw -16(0,%r30),%r24
ldw -12(0,%r30),%r25
add %r21,%r25,%r25
addc %r20,%r24,%r24
copy %r24,%r23
ldi 0,%r22
copy %r23,%r28
addib,= -1,%r19,L$0027
stw %r25,0(0,%r26)
fldws -8(0,%r29),%fr9L
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r28,%r21
ldi 0,%r20
ldw -16(0,%r30),%r24
ldw -12(0,%r30),%r25
add %r21,%r25,%r25
addc %r20,%r24,%r24
copy %r24,%r23
ldi 0,%r22
copy %r23,%r28
addib,= -1,%r19,L$0027
stw %r25,-8(0,%r31)
fldws -4(0,%r29),%fr9L
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r28,%r21
ldi 0,%r20
ldw -16(0,%r30),%r24
ldw -12(0,%r30),%r25
add %r21,%r25,%r25
addc %r20,%r24,%r24
copy %r24,%r23
ldi 0,%r22
copy %r23,%r28
addib,= -1,%r19,L$0027
stw %r25,-4(0,%r31)
fldws 0(0,%r29),%fr9L
xmpyu %fr8L,%fr9L,%fr9
fstds %fr9,-16(0,%r30)
copy %r28,%r21
ldi 0,%r20
ldw -16(0,%r30),%r24
ldw -12(0,%r30),%r25
add %r21,%r25,%r25
addc %r20,%r24,%r24
copy %r24,%r23
ldi 0,%r22
copy %r23,%r28
addib,= -1,%r19,L$0027
stw %r25,0(0,%r31)
ldo 16(%r29),%r29
ldo 16(%r2),%r2
ldo 16(%r31),%r31
bl L$0026,0
ldo 16(%r26),%r26
L$0027
ldw -84(0,%r30),%r2
bv 0(%r2)
ldwm -64(0,%r30),%r4
.EXIT
.PROCEND
.align 4
.EXPORT bn_sqr_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR
bn_sqr_words
.PROC
.CALLINFO FRAME=0,NO_CALLS
.ENTRY
ldo 28(%r26),%r19
ldo 12(%r25),%r28
L$0042
fldws 0(0,%r25),%fr8L
fldws 0(0,%r25),%fr8R
xmpyu %fr8L,%fr8R,%fr8
fstds %fr8,-16(0,%r30)
ldw -16(0,%r30),%r22
ldw -12(0,%r30),%r23
stw %r23,0(0,%r26)
copy %r22,%r21
ldi 0,%r20
addib,= -1,%r24,L$0049
stw %r21,-24(0,%r19)
fldws -8(0,%r28),%fr8L
fldws -8(0,%r28),%fr8R
xmpyu %fr8L,%fr8R,%fr8
fstds %fr8,-16(0,%r30)
ldw -16(0,%r30),%r22
ldw -12(0,%r30),%r23
stw %r23,-20(0,%r19)
copy %r22,%r21
ldi 0,%r20
addib,= -1,%r24,L$0049
stw %r21,-16(0,%r19)
fldws -4(0,%r28),%fr8L
fldws -4(0,%r28),%fr8R
xmpyu %fr8L,%fr8R,%fr8
fstds %fr8,-16(0,%r30)
ldw -16(0,%r30),%r22
ldw -12(0,%r30),%r23
stw %r23,-12(0,%r19)
copy %r22,%r21
ldi 0,%r20
addib,= -1,%r24,L$0049
stw %r21,-8(0,%r19)
fldws 0(0,%r28),%fr8L
fldws 0(0,%r28),%fr8R
xmpyu %fr8L,%fr8R,%fr8
fstds %fr8,-16(0,%r30)
ldw -16(0,%r30),%r22
ldw -12(0,%r30),%r23
stw %r23,-4(0,%r19)
copy %r22,%r21
ldi 0,%r20
addib,= -1,%r24,L$0049
stw %r21,0(0,%r19)
ldo 16(%r28),%r28
ldo 16(%r25),%r25
ldo 32(%r19),%r19
bl L$0042,0
ldo 32(%r26),%r26
L$0049
bv,n 0(%r2)
.EXIT
.PROCEND
.IMPORT BN_num_bits_word,CODE
.IMPORT fprintf,CODE
.IMPORT __iob,DATA
.SPACE $TEXT$
.SUBSPA $LIT$
.align 4
L$C0000
.STRING "Division would overflow (%d)\x0a\x00"
.IMPORT abort,CODE
.SPACE $TEXT$
.SUBSPA $CODE$
.align 4
.EXPORT bn_div64,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,RTNVAL=GR
bn_div64
.PROC
.CALLINFO FRAME=128,CALLS,SAVE_RP,ENTRY_GR=8
.ENTRY
stw %r2,-20(0,%r30)
stwm %r8,128(0,%r30)
stw %r7,-124(0,%r30)
stw %r4,-112(0,%r30)
stw %r3,-108(0,%r30)
copy %r26,%r3
copy %r25,%r4
stw %r6,-120(0,%r30)
ldi 0,%r7
stw %r5,-116(0,%r30)
movb,<> %r24,%r5,L$0051
ldi 2,%r6
bl L$0068,0
ldi -1,%r28
L$0051
.CALL ARGW0=GR
bl BN_num_bits_word,%r2
copy %r5,%r26
copy %r28,%r24
ldi 32,%r19
comb,= %r19,%r24,L$0052
subi 31,%r24,%r19
mtsar %r19
zvdepi 1,32,%r19
comb,>>= %r19,%r3,L$0052
addil LR'__iob-$global$+32,%r27
ldo RR'__iob-$global$+32(%r1),%r26
ldil LR'L$C0000,%r25
.CALL ARGW0=GR,ARGW1=GR,ARGW2=GR
bl fprintf,%r2
ldo RR'L$C0000(%r25),%r25
.CALL
bl abort,%r2
nop
L$0052
comb,>> %r5,%r3,L$0053
subi 32,%r24,%r24
sub %r3,%r5,%r3
L$0053
comib,= 0,%r24,L$0054
subi 31,%r24,%r19
mtsar %r19
zvdep %r5,32,%r5
zvdep %r3,32,%r21
subi 32,%r24,%r20
mtsar %r20
vshd 0,%r4,%r20
or %r21,%r20,%r3
mtsar %r19
zvdep %r4,32,%r4
L$0054
extru %r5,15,16,%r23
extru %r5,31,16,%r28
L$0055
extru %r3,15,16,%r19
comb,<> %r23,%r19,L$0058
copy %r3,%r26
bl L$0059,0
zdepi -1,31,16,%r29
L$0058
.IMPORT $$divU,MILLICODE
bl $$divU,%r31
copy %r23,%r25
L$0059
stw %r29,-16(0,%r30)
fldws -16(0,%r30),%fr10L
stw %r28,-16(0,%r30)
fldws -16(0,%r30),%fr10R
stw %r23,-16(0,%r30)
xmpyu %fr10L,%fr10R,%fr8
fldws -16(0,%r30),%fr10R
fstws %fr8R,-16(0,%r30)
xmpyu %fr10L,%fr10R,%fr9
ldw -16(0,%r30),%r8
fstws %fr9R,-16(0,%r30)
copy %r8,%r22
ldw -16(0,%r30),%r8
extru %r4,15,16,%r24
copy %r8,%r21
L$0060
sub %r3,%r21,%r20
copy %r20,%r19
depi 0,31,16,%r19
comib,<> 0,%r19,L$0061
zdep %r20,15,16,%r19
addl %r19,%r24,%r19
comb,>>= %r19,%r22,L$0061
sub %r22,%r28,%r22
sub %r21,%r23,%r21
bl L$0060,0
ldo -1(%r29),%r29
L$0061
stw %r29,-16(0,%r30)
fldws -16(0,%r30),%fr10L
stw %r28,-16(0,%r30)
fldws -16(0,%r30),%fr10R
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
ldw -16(0,%r30),%r8
stw %r23,-16(0,%r30)
fldws -16(0,%r30),%fr10R
copy %r8,%r19
xmpyu %fr10L,%fr10R,%fr8
fstws %fr8R,-16(0,%r30)
extru %r19,15,16,%r20
ldw -16(0,%r30),%r8
zdep %r19,15,16,%r19
addl %r8,%r20,%r20
comclr,<<= %r19,%r4,0
addi 1,%r20,%r20
comb,<<= %r20,%r3,L$0066
sub %r4,%r19,%r4
addl %r3,%r5,%r3
ldo -1(%r29),%r29
L$0066
addib,= -1,%r6,L$0056
sub %r3,%r20,%r3
zdep %r29,15,16,%r7
shd %r3,%r4,16,%r3
bl L$0055,0
zdep %r4,15,16,%r4
L$0056
or %r7,%r29,%r28
L$0068
ldw -148(0,%r30),%r2
ldw -124(0,%r30),%r7
ldw -120(0,%r30),%r6
ldw -116(0,%r30),%r5
ldw -112(0,%r30),%r4
ldw -108(0,%r30),%r3
bv 0(%r2)
ldwm -128(0,%r30),%r8
.EXIT
.PROCEND

View File

@ -1,646 +0,0 @@
.file 1 "../bn_mulw.c"
.set nobopt
.option pic2
# GNU C 2.6.3 [AL 1.1, MM 40] SGI running IRIX 5.0 compiled by GNU C
# Cc1 defaults:
# -mabicalls
# Cc1 arguments (-G value = 0, Cpu = 3000, ISA = 1):
# -quiet -dumpbase -O2 -o
gcc2_compiled.:
__gnu_compiled_c:
.rdata
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x34,0x39,0x20
.byte 0x24,0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x33,0x34,0x20
.byte 0x24,0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x35,0x20,0x24
.byte 0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x38,0x20,0x24
.byte 0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x32,0x33,0x20
.byte 0x24,0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x37,0x38,0x20
.byte 0x24,0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x33,0x2e,0x37,0x30,0x20
.byte 0x24,0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x32,0x20,0x24
.byte 0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x34,0x20,0x24
.byte 0x0
.byte 0x24,0x52,0x65,0x76,0x69,0x73,0x69,0x6f
.byte 0x6e,0x3a,0x20,0x31,0x2e,0x38,0x20,0x24
.byte 0x0
.text
.align 2
.globl bn_mul_add_words
.ent bn_mul_add_words
bn_mul_add_words:
.frame $sp,0,$31 # vars= 0, regs= 0/0, args= 0, extra= 0
.mask 0x00000000,0
.fmask 0x00000000,0
.set noreorder
.cpload $25
.set reorder
move $12,$4
move $14,$5
move $9,$6
move $13,$7
move $8,$0
addu $10,$12,12
addu $11,$14,12
$L2:
lw $6,0($14)
#nop
multu $13,$6
mfhi $6
mflo $7
#nop
move $5,$8
move $4,$0
lw $3,0($12)
addu $9,$9,-1
move $2,$0
addu $7,$7,$3
sltu $8,$7,$3
addu $6,$6,$2
addu $6,$6,$8
addu $7,$7,$5
sltu $2,$7,$5
addu $6,$6,$4
addu $6,$6,$2
srl $3,$6,0
move $2,$0
move $8,$3
.set noreorder
.set nomacro
beq $9,$0,$L3
sw $7,0($12)
.set macro
.set reorder
lw $6,-8($11)
#nop
multu $13,$6
mfhi $6
mflo $7
#nop
move $5,$8
move $4,$0
lw $3,-8($10)
addu $9,$9,-1
move $2,$0
addu $7,$7,$3
sltu $8,$7,$3
addu $6,$6,$2
addu $6,$6,$8
addu $7,$7,$5
sltu $2,$7,$5
addu $6,$6,$4
addu $6,$6,$2
srl $3,$6,0
move $2,$0
move $8,$3
.set noreorder
.set nomacro
beq $9,$0,$L3
sw $7,-8($10)
.set macro
.set reorder
lw $6,-4($11)
#nop
multu $13,$6
mfhi $6
mflo $7
#nop
move $5,$8
move $4,$0
lw $3,-4($10)
addu $9,$9,-1
move $2,$0
addu $7,$7,$3
sltu $8,$7,$3
addu $6,$6,$2
addu $6,$6,$8
addu $7,$7,$5
sltu $2,$7,$5
addu $6,$6,$4
addu $6,$6,$2
srl $3,$6,0
move $2,$0
move $8,$3
.set noreorder
.set nomacro
beq $9,$0,$L3
sw $7,-4($10)
.set macro
.set reorder
lw $6,0($11)
#nop
multu $13,$6
mfhi $6
mflo $7
#nop
move $5,$8
move $4,$0
lw $3,0($10)
addu $9,$9,-1
move $2,$0
addu $7,$7,$3
sltu $8,$7,$3
addu $6,$6,$2
addu $6,$6,$8
addu $7,$7,$5
sltu $2,$7,$5
addu $6,$6,$4
addu $6,$6,$2
srl $3,$6,0
move $2,$0
move $8,$3
.set noreorder
.set nomacro
beq $9,$0,$L3
sw $7,0($10)
.set macro
.set reorder
addu $11,$11,16
addu $14,$14,16
addu $10,$10,16
.set noreorder
.set nomacro
j $L2
addu $12,$12,16
.set macro
.set reorder
$L3:
.set noreorder
.set nomacro
j $31
move $2,$8
.set macro
.set reorder
.end bn_mul_add_words
.align 2
.globl bn_mul_words
.ent bn_mul_words
bn_mul_words:
.frame $sp,0,$31 # vars= 0, regs= 0/0, args= 0, extra= 0
.mask 0x00000000,0
.fmask 0x00000000,0
.set noreorder
.cpload $25
.set reorder
move $11,$4
move $12,$5
move $8,$6
move $6,$0
addu $10,$11,12
addu $9,$12,12
$L10:
lw $4,0($12)
#nop
multu $7,$4
mfhi $4
mflo $5
#nop
move $3,$6
move $2,$0
addu $8,$8,-1
addu $5,$5,$3
sltu $6,$5,$3
addu $4,$4,$2
addu $4,$4,$6
srl $3,$4,0
move $2,$0
move $6,$3
.set noreorder
.set nomacro
beq $8,$0,$L11
sw $5,0($11)
.set macro
.set reorder
lw $4,-8($9)
#nop
multu $7,$4
mfhi $4
mflo $5
#nop
move $3,$6
move $2,$0
addu $8,$8,-1
addu $5,$5,$3
sltu $6,$5,$3
addu $4,$4,$2
addu $4,$4,$6
srl $3,$4,0
move $2,$0
move $6,$3
.set noreorder
.set nomacro
beq $8,$0,$L11
sw $5,-8($10)
.set macro
.set reorder
lw $4,-4($9)
#nop
multu $7,$4
mfhi $4
mflo $5
#nop
move $3,$6
move $2,$0
addu $8,$8,-1
addu $5,$5,$3
sltu $6,$5,$3
addu $4,$4,$2
addu $4,$4,$6
srl $3,$4,0
move $2,$0
move $6,$3
.set noreorder
.set nomacro
beq $8,$0,$L11
sw $5,-4($10)
.set macro
.set reorder
lw $4,0($9)
#nop
multu $7,$4
mfhi $4
mflo $5
#nop
move $3,$6
move $2,$0
addu $8,$8,-1
addu $5,$5,$3
sltu $6,$5,$3
addu $4,$4,$2
addu $4,$4,$6
srl $3,$4,0
move $2,$0
move $6,$3
.set noreorder
.set nomacro
beq $8,$0,$L11
sw $5,0($10)
.set macro
.set reorder
addu $9,$9,16
addu $12,$12,16
addu $10,$10,16
.set noreorder
.set nomacro
j $L10
addu $11,$11,16
.set macro
.set reorder
$L11:
.set noreorder
.set nomacro
j $31
move $2,$6
.set macro
.set reorder
.end bn_mul_words
.align 2
.globl bn_sqr_words
.ent bn_sqr_words
bn_sqr_words:
.frame $sp,0,$31 # vars= 0, regs= 0/0, args= 0, extra= 0
.mask 0x00000000,0
.fmask 0x00000000,0
.set noreorder
.cpload $25
.set reorder
move $9,$4
addu $7,$9,28
addu $8,$5,12
$L18:
lw $2,0($5)
#nop
multu $2,$2
mfhi $2
mflo $3
#nop
addu $6,$6,-1
sw $3,0($9)
srl $3,$2,0
move $2,$0
.set noreorder
.set nomacro
beq $6,$0,$L19
sw $3,-24($7)
.set macro
.set reorder
lw $2,-8($8)
#nop
multu $2,$2
mfhi $2
mflo $3
#nop
addu $6,$6,-1
sw $3,-20($7)
srl $3,$2,0
move $2,$0
.set noreorder
.set nomacro
beq $6,$0,$L19
sw $3,-16($7)
.set macro
.set reorder
lw $2,-4($8)
#nop
multu $2,$2
mfhi $2
mflo $3
#nop
addu $6,$6,-1
sw $3,-12($7)
srl $3,$2,0
move $2,$0
.set noreorder
.set nomacro
beq $6,$0,$L19
sw $3,-8($7)
.set macro
.set reorder
lw $2,0($8)
#nop
multu $2,$2
mfhi $2
mflo $3
#nop
addu $6,$6,-1
sw $3,-4($7)
srl $3,$2,0
move $2,$0
.set noreorder
.set nomacro
beq $6,$0,$L19
sw $3,0($7)
.set macro
.set reorder
addu $8,$8,16
addu $5,$5,16
addu $7,$7,32
.set noreorder
.set nomacro
j $L18
addu $9,$9,32
.set macro
.set reorder
$L19:
j $31
.end bn_sqr_words
.rdata
.align 2
$LC0:
.byte 0x44,0x69,0x76,0x69,0x73,0x69,0x6f,0x6e
.byte 0x20,0x77,0x6f,0x75,0x6c,0x64,0x20,0x6f
.byte 0x76,0x65,0x72,0x66,0x6c,0x6f,0x77,0xa
.byte 0x0
.text
.align 2
.globl bn_div64
.ent bn_div64
bn_div64:
.frame $sp,56,$31 # vars= 0, regs= 7/0, args= 16, extra= 8
.mask 0x901f0000,-8
.fmask 0x00000000,0
.set noreorder
.cpload $25
.set reorder
subu $sp,$sp,56
.cprestore 16
sw $16,24($sp)
move $16,$4
sw $17,28($sp)
move $17,$5
sw $18,32($sp)
move $18,$6
sw $20,40($sp)
move $20,$0
sw $19,36($sp)
li $19,0x00000002 # 2
sw $31,48($sp)
.set noreorder
.set nomacro
bne $18,$0,$L26
sw $28,44($sp)
.set macro
.set reorder
.set noreorder
.set nomacro
j $L43
li $2,-1 # 0xffffffff
.set macro
.set reorder
$L26:
move $4,$18
jal BN_num_bits_word
move $4,$2
li $2,0x00000020 # 32
.set noreorder
.set nomacro
beq $4,$2,$L27
li $2,0x00000001 # 1
.set macro
.set reorder
sll $2,$2,$4
sltu $2,$2,$16
.set noreorder
.set nomacro
beq $2,$0,$L44
li $5,0x00000020 # 32
.set macro
.set reorder
la $4,__iob+32
la $5,$LC0
jal fprintf
jal abort
$L27:
li $5,0x00000020 # 32
$L44:
sltu $2,$16,$18
.set noreorder
.set nomacro
bne $2,$0,$L28
subu $4,$5,$4
.set macro
.set reorder
subu $16,$16,$18
$L28:
.set noreorder
.set nomacro
beq $4,$0,$L29
li $10,-65536 # 0xffff0000
.set macro
.set reorder
sll $18,$18,$4
sll $3,$16,$4
subu $2,$5,$4
srl $2,$17,$2
or $16,$3,$2
sll $17,$17,$4
$L29:
srl $7,$18,16
andi $9,$18,0xffff
$L30:
srl $2,$16,16
.set noreorder
.set nomacro
beq $2,$7,$L34
li $6,0x0000ffff # 65535
.set macro
.set reorder
divu $6,$16,$7
$L34:
mult $6,$9
mflo $5
#nop
#nop
mult $6,$7
and $2,$17,$10
srl $8,$2,16
mflo $4
$L35:
subu $3,$16,$4
and $2,$3,$10
.set noreorder
.set nomacro
bne $2,$0,$L36
sll $2,$3,16
.set macro
.set reorder
addu $2,$2,$8
sltu $2,$2,$5
.set noreorder
.set nomacro
beq $2,$0,$L36
subu $5,$5,$9
.set macro
.set reorder
subu $4,$4,$7
.set noreorder
.set nomacro
j $L35
addu $6,$6,-1
.set macro
.set reorder
$L36:
mult $6,$7
mflo $5
#nop
#nop
mult $6,$9
mflo $4
#nop
#nop
srl $3,$4,16
sll $2,$4,16
and $4,$2,$10
sltu $2,$17,$4
.set noreorder
.set nomacro
beq $2,$0,$L40
addu $5,$5,$3
.set macro
.set reorder
addu $5,$5,1
$L40:
sltu $2,$16,$5
.set noreorder
.set nomacro
beq $2,$0,$L41
subu $17,$17,$4
.set macro
.set reorder
addu $16,$16,$18
addu $6,$6,-1
$L41:
addu $19,$19,-1
.set noreorder
.set nomacro
beq $19,$0,$L31
subu $16,$16,$5
.set macro
.set reorder
sll $20,$6,16
sll $3,$16,16
srl $2,$17,16
or $16,$3,$2
.set noreorder
.set nomacro
j $L30
sll $17,$17,16
.set macro
.set reorder
$L31:
or $2,$20,$6
$L43:
lw $31,48($sp)
lw $20,40($sp)
lw $19,36($sp)
lw $18,32($sp)
lw $17,28($sp)
lw $16,24($sp)
addu $sp,$sp,56
j $31
.end bn_div64
.globl abort .text
.globl fprintf .text
.globl BN_num_bits_word .text

View File

@ -1,434 +0,0 @@
#!/usr/local/bin/perl
package alpha;
use Carp qw(croak cluck);
$label="100";
$n_debug=0;
$smear_regs=1;
$reg_alloc=1;
$align="3";
$com_start="#";
sub main'asm_init_output { @out=(); }
sub main'asm_get_output { return(@out); }
sub main'get_labels { return(@labels); }
sub main'external_label { push(@labels,@_); }
# General registers
%regs=( 'r0', '$0',
'r1', '$1',
'r2', '$2',
'r3', '$3',
'r4', '$4',
'r5', '$5',
'r6', '$6',
'r7', '$7',
'r8', '$8',
'r9', '$22',
'r10', '$23',
'r11', '$24',
'r12', '$25',
'r13', '$27',
'r14', '$28',
'r15', '$21', # argc == 5
'r16', '$20', # argc == 4
'r17', '$19', # argc == 3
'r18', '$18', # argc == 2
'r19', '$17', # argc == 1
'r20', '$16', # argc == 0
'r21', '$9', # save 0
'r22', '$10', # save 1
'r23', '$11', # save 2
'r24', '$12', # save 3
'r25', '$13', # save 4
'r26', '$14', # save 5
'a0', '$16',
'a1', '$17',
'a2', '$18',
'a3', '$19',
'a4', '$20',
'a5', '$21',
's0', '$9',
's1', '$10',
's2', '$11',
's3', '$12',
's4', '$13',
's5', '$14',
'zero', '$31',
'sp', '$30',
);
$main'reg_s0="r21";
$main'reg_s1="r22";
$main'reg_s2="r23";
$main'reg_s3="r24";
$main'reg_s4="r25";
$main'reg_s5="r26";
@reg=( '$0', '$1' ,'$2' ,'$3' ,'$4' ,'$5' ,'$6' ,'$7' ,'$8',
'$22','$23','$24','$25','$20','$21','$27','$28');
sub main'sub { &out3("subq",@_); }
sub main'add { &out3("addq",@_); }
sub main'mov { &out3("bis",$_[0],$_[0],$_[1]); }
sub main'or { &out3("bis",@_); }
sub main'bis { &out3("bis",@_); }
sub main'br { &out1("br",@_); }
sub main'ld { &out2("ldq",@_); }
sub main'st { &out2("stq",@_); }
sub main'cmpult { &out3("cmpult",@_); }
sub main'cmplt { &out3("cmplt",@_); }
sub main'bgt { &out2("bgt",@_); }
sub main'ble { &out2("ble",@_); }
sub main'blt { &out2("blt",@_); }
sub main'mul { &out3("mulq",@_); }
sub main'muh { &out3("umulh",@_); }
$main'QWS=8;
sub main'asm_add
{
push(@out,@_);
}
sub main'asm_finish
{
&main'file_end();
print &main'asm_get_output();
}
sub main'asm_init
{
($type,$fn)=@_;
$filename=$fn;
&main'asm_init_output();
&main'comment("Don't even think of reading this code");
&main'comment("It was automatically generated by $filename");
&main'comment("Which is a perl program used to generate the alpha assember.");
&main'comment("eric <eay\@cryptsoft.com>");
&main'comment("");
$filename =~ s/\.pl$//;
&main'file($filename);
}
sub conv
{
local($r)=@_;
local($v);
return($regs{$r}) if defined($regs{$r});
return($r);
}
sub main'QWPw
{
local($off,$reg)=@_;
return(&main'QWP($off*8,$reg));
}
sub main'QWP
{
local($off,$reg)=@_;
$ret="$off(".&conv($reg).")";
return($ret);
}
sub out3
{
local($name,$p1,$p2,$p3)=@_;
$p1=&conv($p1);
$p2=&conv($p2);
$p3=&conv($p3);
push(@out,"\t$name\t");
$l=length($p1)+1;
push(@out,$p1.",");
$ll=3-($l+9)/8;
$tmp1=sprintf("\t" x $ll);
push(@out,$tmp1);
$l=length($p2)+1;
push(@out,$p2.",");
$ll=3-($l+9)/8;
$tmp1=sprintf("\t" x $ll);
push(@out,$tmp1);
push(@out,&conv($p3)."\n");
}
sub out2
{
local($name,$p1,$p2,$p3)=@_;
$p1=&conv($p1);
$p2=&conv($p2);
push(@out,"\t$name\t");
$l=length($p1)+1;
push(@out,$p1.",");
$ll=3-($l+9)/8;
$tmp1=sprintf("\t" x $ll);
push(@out,$tmp1);
push(@out,&conv($p2)."\n");
}
sub out1
{
local($name,$p1)=@_;
$p1=&conv($p1);
push(@out,"\t$name\t".$p1."\n");
}
sub out0
{
push(@out,"\t$_[0]\n");
}
sub main'file
{
local($file)=@_;
local($tmp)=<<"EOF";
# DEC Alpha assember
# Generated from perl scripts contains in SSLeay
.file 1 "$file.s"
.set noat
EOF
push(@out,$tmp);
}
sub main'function_begin
{
local($func)=@_;
print STDERR "$func\n";
local($tmp)=<<"EOF";
.text
.align $align
.globl $func
.ent $func
${func}:
${func}..ng:
.frame \$30,0,\$26,0
.prologue 0
EOF
push(@out,$tmp);
$stack=0;
}
sub main'function_end
{
local($func)=@_;
local($tmp)=<<"EOF";
ret \$31,(\$26),1
.end $func
EOF
push(@out,$tmp);
$stack=0;
%label=();
}
sub main'function_end_A
{
local($func)=@_;
local($tmp)=<<"EOF";
ret \$31,(\$26),1
EOF
push(@out,$tmp);
}
sub main'function_end_B
{
local($func)=@_;
$func=$under.$func;
push(@out,"\t.end $func\n");
$stack=0;
%label=();
}
sub main'wparam
{
local($num)=@_;
if ($num < 6)
{
$num=20-$num;
return("r$num");
}
else
{ return(&main'QWP($stack+$num*8,"sp")); }
}
sub main'stack_push
{
local($num)=@_;
$stack+=$num*8;
&main'sub("sp",$num*8,"sp");
}
sub main'stack_pop
{
local($num)=@_;
$stack-=$num*8;
&main'add("sp",$num*8,"sp");
}
sub main'swtmp
{
return(&main'QWP(($_[0])*8,"sp"));
}
# Should use swtmp, which is above sp. Linix can trash the stack above esp
#sub main'wtmp
# {
# local($num)=@_;
#
# return(&main'QWP(-($num+1)*4,"esp","",0));
# }
sub main'comment
{
foreach (@_)
{
if (/^\s*$/)
{ push(@out,"\n"); }
else
{ push(@out,"\t$com_start $_ $com_end\n"); }
}
}
sub main'label
{
if (!defined($label{$_[0]}))
{
$label{$_[0]}=$label;
$label++;
}
return('$'.$label{$_[0]});
}
sub main'set_label
{
if (!defined($label{$_[0]}))
{
$label{$_[0]}=$label;
$label++;
}
# push(@out,".align $align\n") if ($_[1] != 0);
push(@out,'$'."$label{$_[0]}:\n");
}
sub main'file_end
{
}
sub main'data_word
{
push(@out,"\t.long $_[0]\n");
}
@pool_free=();
@pool_taken=();
$curr_num=0;
$max=0;
sub main'init_pool
{
local($args)=@_;
local($i);
@pool_free=();
for ($i=(14+(6-$args)); $i >= 0; $i--)
{
push(@pool_free,"r$i");
}
print STDERR "START :register pool:@pool_free\n";
$curr_num=$max=0;
}
sub main'fin_pool
{
printf STDERR "END %2d:register pool:@pool_free\n",$max;
}
sub main'GR
{
local($r)=@_;
local($i,@n,$_);
foreach (@pool_free)
{
if ($r ne $_)
{ push(@n,$_); }
else
{
$curr_num++;
$max=$curr_num if ($curr_num > $max);
}
}
@pool_free=@n;
print STDERR "GR:@pool_free\n" if $reg_alloc;
return(@_);
}
sub main'NR
{
local($num)=@_;
local(@ret);
$num=1 if $num == 0;
($#pool_free >= ($num-1)) || croak "out of registers: want $num, have @pool_free";
while ($num > 0)
{
push(@ret,pop @pool_free);
$curr_num++;
$max=$curr_num if ($curr_num > $max);
$num--
}
print STDERR "nr @ret\n" if $n_debug;
print STDERR "NR:@pool_free\n" if $reg_alloc;
return(@ret);
}
sub main'FR
{
local(@r)=@_;
local(@a,$v,$w);
print STDERR "fr @r\n" if $n_debug;
# cluck "fr @r";
for $w (@pool_free)
{
foreach $v (@r)
{
croak "double register free of $v (@pool_free)" if $w eq $v;
}
}
foreach $v (@r)
{
croak "bad argument to FR" if ($v !~ /^r\d+$/);
if ($smear_regs)
{ unshift(@pool_free,$v); }
else { push(@pool_free,$v); }
$curr_num--;
}
print STDERR "FR:@pool_free\n" if $reg_alloc;
}
1;

View File

@ -1,363 +0,0 @@
#!/usr/local/bin/perl
# x86 CodeWarrior assembler for NetWare
# This file is a slightly modified version of x86nasm.pl. The Metrowerks
# compiler for NetWare doesn't prefix symbols with an underscore.
#
$label="L000";
%lb=( 'eax', 'al',
'ebx', 'bl',
'ecx', 'cl',
'edx', 'dl',
'ax', 'al',
'bx', 'bl',
'cx', 'cl',
'dx', 'dl',
);
%hb=( 'eax', 'ah',
'ebx', 'bh',
'ecx', 'ch',
'edx', 'dh',
'ax', 'ah',
'bx', 'bh',
'cx', 'ch',
'dx', 'dh',
);
sub main'asm_init_output
{
@out=();
&comment("NetWare: assembly for CodeWarrior assembler (mwasmnlm)");
}
sub main'asm_get_output { return(@out); }
sub main'get_labels { return(@labels); }
sub main'external_label
{
push(@labels,@_);
foreach (@_) {
push(@out, ".extern\t$_\n");
}
}
sub main'LB
{
(defined($lb{$_[0]})) || die "$_[0] does not have a 'low byte'\n";
return($lb{$_[0]});
}
sub main'HB
{
(defined($hb{$_[0]})) || die "$_[0] does not have a 'high byte'\n";
return($hb{$_[0]});
}
sub main'BP
{
&get_mem("BYTE",@_);
}
sub main'DWP
{
&get_mem("DWORD",@_);
}
sub main'BC
{
return "@_";
}
sub main'DWC
{
return "@_";
}
sub main'stack_push
{
my($num)=@_;
$stack+=$num*4;
&main'sub("esp",$num*4);
}
sub main'stack_pop
{
my($num)=@_;
$stack-=$num*4;
&main'add("esp",$num*4);
}
sub get_mem
{
my($size,$addr,$reg1,$reg2,$idx)=@_;
my($t,$post);
my($ret)="$size PTR [";
$addr =~ s/^\s+//;
if ($addr =~ /^(.+)\+(.+)$/)
{
$reg2=&conv($1);
$addr="$2";
}
elsif ($addr =~ /^[_a-zA-Z]/)
{
$addr="$addr";
}
if ($addr =~ /^.+\-.+$/) { $addr="($addr)"; }
$reg1="$regs{$reg1}" if defined($regs{$reg1});
$reg2="$regs{$reg2}" if defined($regs{$reg2});
if (($addr ne "") && ($addr ne 0))
{
if ($addr !~ /^-/)
{ $ret.="${addr}+"; }
else { $post=$addr; }
}
if ($reg2 ne "")
{
$t="";
$t="*$idx" if ($idx != 0);
$reg1="+".$reg1 if ("$reg1$post" ne "");
$ret.="$reg2$t$reg1$post]";
}
else
{
$ret.="$reg1$post]"
}
$ret =~ s/\+\]/]/; # in case $addr was the only argument
return($ret);
}
sub main'mov { &out2("mov",@_); }
sub main'movb { &out2("mov",@_); }
sub main'and { &out2("and",@_); }
sub main'or { &out2("or",@_); }
sub main'shl { &out2("shl",@_); }
sub main'shr { &out2("shr",@_); }
sub main'xor { &out2("xor",@_); }
sub main'xorb { &out2("xor",@_); }
sub main'add { &out2("add",@_); }
sub main'adc { &out2("adc",@_); }
sub main'sub { &out2("sub",@_); }
sub main'rotl { &out2("rol",@_); }
sub main'rotr { &out2("ror",@_); }
sub main'exch { &out2("xchg",@_); }
sub main'cmp { &out2("cmp",@_); }
sub main'lea { &out2("lea",@_); }
sub main'mul { &out1("mul",@_); }
sub main'div { &out1("div",@_); }
sub main'dec { &out1("dec",@_); }
sub main'inc { &out1("inc",@_); }
sub main'jmp { &out1("jmp",@_); }
sub main'jmp_ptr { &out1p("jmp",@_); }
sub main'je { &out1("je ",@_); }
sub main'jle { &out1("jle ",@_); }
sub main'jz { &out1("jz ",@_); }
sub main'jge { &out1("jge ",@_); }
sub main'jl { &out1("jl ",@_); }
sub main'ja { &out1("ja ",@_); }
sub main'jae { &out1("jae ",@_); }
sub main'jb { &out1("jb ",@_); }
sub main'jbe { &out1("jbe ",@_); }
sub main'jc { &out1("jc ",@_); }
sub main'jnc { &out1("jnc ",@_); }
sub main'jnz { &out1("jnz ",@_); }
sub main'jne { &out1("jne ",@_); }
sub main'jno { &out1("jno ",@_); }
sub main'push { &out1("push",@_); $stack+=4; }
sub main'pop { &out1("pop",@_); $stack-=4; }
sub main'bswap { &out1("bswap",@_); &using486(); }
sub main'not { &out1("not",@_); }
sub main'call { &out1("call",$_[0]); }
sub main'ret { &out0("ret"); }
sub main'nop { &out0("nop"); }
sub out2
{
my($name,$p1,$p2)=@_;
my($l,$t);
push(@out,"\t$name\t");
$t=&conv($p1).",";
$l=length($t);
push(@out,$t);
$l=4-($l+9)/8;
push(@out,"\t" x $l);
push(@out,&conv($p2));
push(@out,"\n");
}
sub out0
{
my($name)=@_;
push(@out,"\t$name\n");
}
sub out1
{
my($name,$p1)=@_;
my($l,$t);
push(@out,"\t$name\t".&conv($p1)."\n");
}
sub conv
{
my($p)=@_;
$p =~ s/0x([0-9A-Fa-f]+)/0$1h/;
return $p;
}
sub using486
{
return if $using486;
$using486++;
grep(s/\.386/\.486/,@out);
}
sub main'file
{
push(@out, ".section .text\n");
}
sub main'function_begin
{
my($func,$extra)=@_;
push(@labels,$func);
my($tmp)=<<"EOF";
.global $func
$func:
push ebp
push ebx
push esi
push edi
EOF
push(@out,$tmp);
$stack=20;
}
sub main'function_begin_B
{
my($func,$extra)=@_;
my($tmp)=<<"EOF";
.global $func
$func:
EOF
push(@out,$tmp);
$stack=4;
}
sub main'function_end
{
my($func)=@_;
my($tmp)=<<"EOF";
pop edi
pop esi
pop ebx
pop ebp
ret
EOF
push(@out,$tmp);
$stack=0;
%label=();
}
sub main'function_end_B
{
$stack=0;
%label=();
}
sub main'function_end_A
{
my($func)=@_;
my($tmp)=<<"EOF";
pop edi
pop esi
pop ebx
pop ebp
ret
EOF
push(@out,$tmp);
}
sub main'file_end
{
}
sub main'wparam
{
my($num)=@_;
return(&main'DWP($stack+$num*4,"esp","",0));
}
sub main'swtmp
{
return(&main'DWP($_[0]*4,"esp","",0));
}
# Should use swtmp, which is above esp. Linix can trash the stack above esp
#sub main'wtmp
# {
# my($num)=@_;
#
# return(&main'DWP(-(($num+1)*4),"esp","",0));
# }
sub main'comment
{
foreach (@_)
{
push(@out,"\t; $_\n");
}
}
sub main'label
{
if (!defined($label{$_[0]}))
{
$label{$_[0]}="${label}${_[0]}";
$label++;
}
return($label{$_[0]});
}
sub main'set_label
{
if (!defined($label{$_[0]}))
{
$label{$_[0]}="${label}${_[0]}";
$label++;
}
push(@out,"$label{$_[0]}:\n");
}
sub main'data_word
{
push(@out,"\t.long\t$_[0]\n");
}
sub out1p
{
my($name,$p1)=@_;
my($l,$t);
push(@out,"\t$name\t ".&conv($p1)."\n");
}
sub main'picmeup
{
local($dst,$sym)=@_;
&main'lea($dst,&main'DWP($sym));
}
sub main'blindpop { &out1("pop",@_); }

View File

@ -1,364 +0,0 @@
#!/usr/local/bin/perl
# x86 nasm assembler for NetWare
# This file is a slightly modified version of x86nasm.pl. The Metrowerks
# compiler for NetWare doesn't prefix symbols with an underscore.
#
$label="L000";
%lb=( 'eax', 'al',
'ebx', 'bl',
'ecx', 'cl',
'edx', 'dl',
'ax', 'al',
'bx', 'bl',
'cx', 'cl',
'dx', 'dl',
);
%hb=( 'eax', 'ah',
'ebx', 'bh',
'ecx', 'ch',
'edx', 'dh',
'ax', 'ah',
'bx', 'bh',
'cx', 'ch',
'dx', 'dh',
);
sub main'asm_init_output
{
@out=();
&comment("NetWare: assembly for NASM assembler (nasmw)");
}
sub main'asm_get_output { return(@out); }
sub main'get_labels { return(@labels); }
sub main'external_label
{
push(@labels,@_);
foreach (@_) {
push(@out, "extern\t$_\n");
}
}
sub main'LB
{
(defined($lb{$_[0]})) || die "$_[0] does not have a 'low byte'\n";
return($lb{$_[0]});
}
sub main'HB
{
(defined($hb{$_[0]})) || die "$_[0] does not have a 'high byte'\n";
return($hb{$_[0]});
}
sub main'BP
{
&get_mem("BYTE",@_);
}
sub main'DWP
{
&get_mem("DWORD",@_);
}
sub main'BC
{
return "BYTE @_";
}
sub main'DWC
{
return "DWORD @_";
}
sub main'stack_push
{
my($num)=@_;
$stack+=$num*4;
&main'sub("esp",$num*4);
}
sub main'stack_pop
{
my($num)=@_;
$stack-=$num*4;
&main'add("esp",$num*4);
}
sub get_mem
{
my($size,$addr,$reg1,$reg2,$idx)=@_;
my($t,$post);
my($ret)="[";
$addr =~ s/^\s+//;
if ($addr =~ /^(.+)\+(.+)$/)
{
$reg2=&conv($1);
$addr="$2";
}
elsif ($addr =~ /^[_a-zA-Z]/)
{
$addr="$addr";
}
if ($addr =~ /^.+\-.+$/) { $addr="($addr)"; }
$reg1="$regs{$reg1}" if defined($regs{$reg1});
$reg2="$regs{$reg2}" if defined($regs{$reg2});
if (($addr ne "") && ($addr ne 0))
{
if ($addr !~ /^-/)
{ $ret.="${addr}+"; }
else { $post=$addr; }
}
if ($reg2 ne "")
{
$t="";
$t="*$idx" if ($idx != 0);
$reg1="+".$reg1 if ("$reg1$post" ne "");
$ret.="$reg2$t$reg1$post]";
}
else
{
$ret.="$reg1$post]"
}
$ret =~ s/\+\]/]/; # in case $addr was the only argument
return($ret);
}
sub main'mov { &out2("mov",@_); }
sub main'movb { &out2("mov",@_); }
sub main'and { &out2("and",@_); }
sub main'or { &out2("or",@_); }
sub main'shl { &out2("shl",@_); }
sub main'shr { &out2("shr",@_); }
sub main'xor { &out2("xor",@_); }
sub main'xorb { &out2("xor",@_); }
sub main'add { &out2("add",@_); }
sub main'adc { &out2("adc",@_); }
sub main'sub { &out2("sub",@_); }
sub main'rotl { &out2("rol",@_); }
sub main'rotr { &out2("ror",@_); }
sub main'exch { &out2("xchg",@_); }
sub main'cmp { &out2("cmp",@_); }
sub main'lea { &out2("lea",@_); }
sub main'mul { &out1("mul",@_); }
sub main'div { &out1("div",@_); }
sub main'dec { &out1("dec",@_); }
sub main'inc { &out1("inc",@_); }
sub main'jmp { &out1("jmp",@_); }
sub main'jmp_ptr { &out1p("jmp",@_); }
# This is a bit of a kludge: declare all branches as NEAR.
sub main'je { &out1("je NEAR",@_); }
sub main'jle { &out1("jle NEAR",@_); }
sub main'jz { &out1("jz NEAR",@_); }
sub main'jge { &out1("jge NEAR",@_); }
sub main'jl { &out1("jl NEAR",@_); }
sub main'ja { &out1("ja NEAR",@_); }
sub main'jae { &out1("jae NEAR",@_); }
sub main'jb { &out1("jb NEAR",@_); }
sub main'jbe { &out1("jbe NEAR",@_); }
sub main'jc { &out1("jc NEAR",@_); }
sub main'jnc { &out1("jnc NEAR",@_); }
sub main'jnz { &out1("jnz NEAR",@_); }
sub main'jne { &out1("jne NEAR",@_); }
sub main'jno { &out1("jno NEAR",@_); }
sub main'push { &out1("push",@_); $stack+=4; }
sub main'pop { &out1("pop",@_); $stack-=4; }
sub main'bswap { &out1("bswap",@_); &using486(); }
sub main'not { &out1("not",@_); }
sub main'call { &out1("call",$_[0]); }
sub main'ret { &out0("ret"); }
sub main'nop { &out0("nop"); }
sub out2
{
my($name,$p1,$p2)=@_;
my($l,$t);
push(@out,"\t$name\t");
$t=&conv($p1).",";
$l=length($t);
push(@out,$t);
$l=4-($l+9)/8;
push(@out,"\t" x $l);
push(@out,&conv($p2));
push(@out,"\n");
}
sub out0
{
my($name)=@_;
push(@out,"\t$name\n");
}
sub out1
{
my($name,$p1)=@_;
my($l,$t);
push(@out,"\t$name\t".&conv($p1)."\n");
}
sub conv
{
my($p)=@_;
$p =~ s/0x([0-9A-Fa-f]+)/0$1h/;
return $p;
}
sub using486
{
return if $using486;
$using486++;
grep(s/\.386/\.486/,@out);
}
sub main'file
{
push(@out, "segment .text\n");
}
sub main'function_begin
{
my($func,$extra)=@_;
push(@labels,$func);
my($tmp)=<<"EOF";
global $func
$func:
push ebp
push ebx
push esi
push edi
EOF
push(@out,$tmp);
$stack=20;
}
sub main'function_begin_B
{
my($func,$extra)=@_;
my($tmp)=<<"EOF";
global $func
$func:
EOF
push(@out,$tmp);
$stack=4;
}
sub main'function_end
{
my($func)=@_;
my($tmp)=<<"EOF";
pop edi
pop esi
pop ebx
pop ebp
ret
EOF
push(@out,$tmp);
$stack=0;
%label=();
}
sub main'function_end_B
{
$stack=0;
%label=();
}
sub main'function_end_A
{
my($func)=@_;
my($tmp)=<<"EOF";
pop edi
pop esi
pop ebx
pop ebp
ret
EOF
push(@out,$tmp);
}
sub main'file_end
{
}
sub main'wparam
{
my($num)=@_;
return(&main'DWP($stack+$num*4,"esp","",0));
}
sub main'swtmp
{
return(&main'DWP($_[0]*4,"esp","",0));
}
# Should use swtmp, which is above esp. Linix can trash the stack above esp
#sub main'wtmp
# {
# my($num)=@_;
#
# return(&main'DWP(-(($num+1)*4),"esp","",0));
# }
sub main'comment
{
foreach (@_)
{
push(@out,"\t; $_\n");
}
}
sub main'label
{
if (!defined($label{$_[0]}))
{
$label{$_[0]}="\$${label}${_[0]}";
$label++;
}
return($label{$_[0]});
}
sub main'set_label
{
if (!defined($label{$_[0]}))
{
$label{$_[0]}="${label}${_[0]}";
$label++;
}
push(@out,"$label{$_[0]}:\n");
}
sub main'data_word
{
push(@out,"\tDD\t$_[0]\n");
}
sub out1p
{
my($name,$p1)=@_;
my($l,$t);
push(@out,"\t$name\t ".&conv($p1)."\n");
}
sub main'picmeup
{
local($dst,$sym)=@_;
&main'lea($dst,&main'DWP($sym));
}
sub main'blindpop { &out1("pop",@_); }