From d7ce91a2b7d322ffce68e9eea917fa481e578d29 Mon Sep 17 00:00:00 2001 From: cvs2svn Date: Sat, 9 Sep 2006 00:11:04 +0000 Subject: This commit was manufactured by cvs2git to create tag 'OPENBSD_4_0_BASE'. --- src/lib/libcrypto/bn/asm/bn-586.pl | 593 --------- src/lib/libcrypto/bn/asm/co-586.pl | 286 ----- src/lib/libcrypto/bn/asm/ia64.S | 1560 ----------------------- src/lib/libcrypto/bn/asm/pa-risc2.s | 1618 ------------------------ src/lib/libcrypto/bn/asm/pa-risc2W.s | 1605 ------------------------ src/lib/libcrypto/bn/asm/ppc.pl | 2078 ------------------------------- src/lib/libcrypto/bn/asm/sparcv8.S | 1458 ---------------------- src/lib/libcrypto/bn/asm/sparcv8plus.S | 1547 ----------------------- src/lib/libcrypto/bn/asm/x86.pl | 28 - src/lib/libcrypto/bn/asm/x86/add.pl | 76 -- src/lib/libcrypto/bn/asm/x86/comba.pl | 277 ---- src/lib/libcrypto/bn/asm/x86/div.pl | 15 - src/lib/libcrypto/bn/asm/x86/mul.pl | 77 -- src/lib/libcrypto/bn/asm/x86/mul_add.pl | 87 -- src/lib/libcrypto/bn/asm/x86/sqr.pl | 60 - src/lib/libcrypto/bn/asm/x86/sub.pl | 76 -- src/lib/libcrypto/bn/asm/x86_64-gcc.c | 593 --------- 17 files changed, 12034 deletions(-) delete mode 100644 src/lib/libcrypto/bn/asm/bn-586.pl delete mode 100644 src/lib/libcrypto/bn/asm/co-586.pl delete mode 100644 src/lib/libcrypto/bn/asm/ia64.S delete mode 100644 src/lib/libcrypto/bn/asm/pa-risc2.s delete mode 100644 src/lib/libcrypto/bn/asm/pa-risc2W.s delete mode 100644 src/lib/libcrypto/bn/asm/ppc.pl delete mode 100644 src/lib/libcrypto/bn/asm/sparcv8.S delete mode 100644 src/lib/libcrypto/bn/asm/sparcv8plus.S delete mode 100644 src/lib/libcrypto/bn/asm/x86.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/add.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/comba.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/div.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/mul.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/mul_add.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/sqr.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86/sub.pl delete mode 100644 src/lib/libcrypto/bn/asm/x86_64-gcc.c (limited to 'src/lib/libcrypto/bn/asm') diff --git a/src/lib/libcrypto/bn/asm/bn-586.pl b/src/lib/libcrypto/bn/asm/bn-586.pl deleted file mode 100644 index c4de4a2bee..0000000000 --- a/src/lib/libcrypto/bn/asm/bn-586.pl +++ /dev/null @@ -1,593 +0,0 @@ -#!/usr/local/bin/perl - -push(@INC,"perlasm","../../perlasm"); -require "x86asm.pl"; - -&asm_init($ARGV[0],$0); - -&bn_mul_add_words("bn_mul_add_words"); -&bn_mul_words("bn_mul_words"); -&bn_sqr_words("bn_sqr_words"); -&bn_div_words("bn_div_words"); -&bn_add_words("bn_add_words"); -&bn_sub_words("bn_sub_words"); -#&bn_sub_part_words("bn_sub_part_words"); - -&asm_finish(); - -sub bn_mul_add_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $Low="eax"; - $High="edx"; - $a="ebx"; - $w="ebp"; - $r="edi"; - $c="esi"; - - &xor($c,$c); # clear carry - &mov($r,&wparam(0)); # - - &mov("ecx",&wparam(2)); # - &mov($a,&wparam(1)); # - - &and("ecx",0xfffffff8); # num / 8 - &mov($w,&wparam(3)); # - - &push("ecx"); # Up the stack for a tmp variable - - &jz(&label("maw_finish")); - - &set_label("maw_loop",0); - - &mov(&swtmp(0),"ecx"); # - - for ($i=0; $i<32; $i+=4) - { - &comment("Round $i"); - - &mov("eax",&DWP($i,$a,"",0)); # *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+= *r - &mov($c,&DWP($i,$r,"",0)); # L(t)+= *r - &adc("edx",0); # H(t)+=carry - &add("eax",$c); # L(t)+=c - &adc("edx",0); # H(t)+=carry - &mov(&DWP($i,$r,"",0),"eax"); # *r= L(t); - &mov($c,"edx"); # c= H(t); - } - - &comment(""); - &mov("ecx",&swtmp(0)); # - &add($a,32); - &add($r,32); - &sub("ecx",8); - &jnz(&label("maw_loop")); - - &set_label("maw_finish",0); - &mov("ecx",&wparam(2)); # get num - &and("ecx",7); - &jnz(&label("maw_finish2")); # helps branch prediction - &jmp(&label("maw_end")); - - &set_label("maw_finish2",1); - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov("eax",&DWP($i*4,$a,"",0));# *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+=c - &mov($c,&DWP($i*4,$r,"",0)); # L(t)+= *r - &adc("edx",0); # H(t)+=carry - &add("eax",$c); - &adc("edx",0); # H(t)+=carry - &dec("ecx") if ($i != 7-1); - &mov(&DWP($i*4,$r,"",0),"eax"); # *r= L(t); - &mov($c,"edx"); # c= H(t); - &jz(&label("maw_end")) if ($i != 7-1); - } - &set_label("maw_end",0); - &mov("eax",$c); - - &pop("ecx"); # clear variable from - - &function_end($name); - } - -sub bn_mul_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $Low="eax"; - $High="edx"; - $a="ebx"; - $w="ecx"; - $r="edi"; - $c="esi"; - $num="ebp"; - - &xor($c,$c); # clear carry - &mov($r,&wparam(0)); # - &mov($a,&wparam(1)); # - &mov($num,&wparam(2)); # - &mov($w,&wparam(3)); # - - &and($num,0xfffffff8); # num / 8 - &jz(&label("mw_finish")); - - &set_label("mw_loop",0); - for ($i=0; $i<32; $i+=4) - { - &comment("Round $i"); - - &mov("eax",&DWP($i,$a,"",0)); # *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+=c - # XXX - - &adc("edx",0); # H(t)+=carry - &mov(&DWP($i,$r,"",0),"eax"); # *r= L(t); - - &mov($c,"edx"); # c= H(t); - } - - &comment(""); - &add($a,32); - &add($r,32); - &sub($num,8); - &jz(&label("mw_finish")); - &jmp(&label("mw_loop")); - - &set_label("mw_finish",0); - &mov($num,&wparam(2)); # get num - &and($num,7); - &jnz(&label("mw_finish2")); - &jmp(&label("mw_end")); - - &set_label("mw_finish2",1); - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov("eax",&DWP($i*4,$a,"",0));# *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+=c - # XXX - &adc("edx",0); # H(t)+=carry - &mov(&DWP($i*4,$r,"",0),"eax");# *r= L(t); - &mov($c,"edx"); # c= H(t); - &dec($num) if ($i != 7-1); - &jz(&label("mw_end")) if ($i != 7-1); - } - &set_label("mw_end",0); - &mov("eax",$c); - - &function_end($name); - } - -sub bn_sqr_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $r="esi"; - $a="edi"; - $num="ebx"; - - &mov($r,&wparam(0)); # - &mov($a,&wparam(1)); # - &mov($num,&wparam(2)); # - - &and($num,0xfffffff8); # num / 8 - &jz(&label("sw_finish")); - - &set_label("sw_loop",0); - for ($i=0; $i<32; $i+=4) - { - &comment("Round $i"); - &mov("eax",&DWP($i,$a,"",0)); # *a - # XXX - &mul("eax"); # *a * *a - &mov(&DWP($i*2,$r,"",0),"eax"); # - &mov(&DWP($i*2+4,$r,"",0),"edx");# - } - - &comment(""); - &add($a,32); - &add($r,64); - &sub($num,8); - &jnz(&label("sw_loop")); - - &set_label("sw_finish",0); - &mov($num,&wparam(2)); # get num - &and($num,7); - &jz(&label("sw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov("eax",&DWP($i*4,$a,"",0)); # *a - # XXX - &mul("eax"); # *a * *a - &mov(&DWP($i*8,$r,"",0),"eax"); # - &dec($num) if ($i != 7-1); - &mov(&DWP($i*8+4,$r,"",0),"edx"); - &jz(&label("sw_end")) if ($i != 7-1); - } - &set_label("sw_end",0); - - &function_end($name); - } - -sub bn_div_words - { - local($name)=@_; - - &function_begin($name,""); - &mov("edx",&wparam(0)); # - &mov("eax",&wparam(1)); # - &mov("ebx",&wparam(2)); # - &div("ebx"); - &function_end($name); - } - -sub bn_add_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $a="esi"; - $b="edi"; - $c="eax"; - $r="ebx"; - $tmp1="ecx"; - $tmp2="edx"; - $num="ebp"; - - &mov($r,&wparam(0)); # get r - &mov($a,&wparam(1)); # get a - &mov($b,&wparam(2)); # get b - &mov($num,&wparam(3)); # get num - &xor($c,$c); # clear carry - &and($num,0xfffffff8); # num / 8 - - &jz(&label("aw_finish")); - - &set_label("aw_loop",0); - for ($i=0; $i<8; $i++) - { - &comment("Round $i"); - - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0)); # *b - &add($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &add($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - } - - &comment(""); - &add($a,32); - &add($b,32); - &add($r,32); - &sub($num,8); - &jnz(&label("aw_loop")); - - &set_label("aw_finish",0); - &mov($num,&wparam(3)); # get num - &and($num,7); - &jz(&label("aw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0));# *b - &add($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &add($tmp1,$tmp2); - &adc($c,0); - &dec($num) if ($i != 6); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &jz(&label("aw_end")) if ($i != 6); - } - &set_label("aw_end",0); - -# &mov("eax",$c); # $c is "eax" - - &function_end($name); - } - -sub bn_sub_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $a="esi"; - $b="edi"; - $c="eax"; - $r="ebx"; - $tmp1="ecx"; - $tmp2="edx"; - $num="ebp"; - - &mov($r,&wparam(0)); # get r - &mov($a,&wparam(1)); # get a - &mov($b,&wparam(2)); # get b - &mov($num,&wparam(3)); # get num - &xor($c,$c); # clear carry - &and($num,0xfffffff8); # num / 8 - - &jz(&label("aw_finish")); - - &set_label("aw_loop",0); - for ($i=0; $i<8; $i++) - { - &comment("Round $i"); - - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0)); # *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - } - - &comment(""); - &add($a,32); - &add($b,32); - &add($r,32); - &sub($num,8); - &jnz(&label("aw_loop")); - - &set_label("aw_finish",0); - &mov($num,&wparam(3)); # get num - &and($num,7); - &jz(&label("aw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0));# *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &dec($num) if ($i != 6); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &jz(&label("aw_end")) if ($i != 6); - } - &set_label("aw_end",0); - -# &mov("eax",$c); # $c is "eax" - - &function_end($name); - } - -sub bn_sub_part_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $a="esi"; - $b="edi"; - $c="eax"; - $r="ebx"; - $tmp1="ecx"; - $tmp2="edx"; - $num="ebp"; - - &mov($r,&wparam(0)); # get r - &mov($a,&wparam(1)); # get a - &mov($b,&wparam(2)); # get b - &mov($num,&wparam(3)); # get num - &xor($c,$c); # clear carry - &and($num,0xfffffff8); # num / 8 - - &jz(&label("aw_finish")); - - &set_label("aw_loop",0); - for ($i=0; $i<8; $i++) - { - &comment("Round $i"); - - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0)); # *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - } - - &comment(""); - &add($a,32); - &add($b,32); - &add($r,32); - &sub($num,8); - &jnz(&label("aw_loop")); - - &set_label("aw_finish",0); - &mov($num,&wparam(3)); # get num - &and($num,7); - &jz(&label("aw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov($tmp1,&DWP(0,$a,"",0)); # *a - &mov($tmp2,&DWP(0,$b,"",0));# *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP(0,$r,"",0),$tmp1); # *r - &add($a, 4); - &add($b, 4); - &add($r, 4); - &dec($num) if ($i != 6); - &jz(&label("aw_end")) if ($i != 6); - } - &set_label("aw_end",0); - - &cmp(&wparam(4),0); - &je(&label("pw_end")); - - &mov($num,&wparam(4)); # get dl - &cmp($num,0); - &je(&label("pw_end")); - &jge(&label("pw_pos")); - - &comment("pw_neg"); - &mov($tmp2,0); - &sub($tmp2,$num); - &mov($num,$tmp2); - &and($num,0xfffffff8); # num / 8 - &jz(&label("pw_neg_finish")); - - &set_label("pw_neg_loop",0); - for ($i=0; $i<8; $i++) - { - &comment("dl<0 Round $i"); - - &mov($tmp1,0); - &mov($tmp2,&DWP($i*4,$b,"",0)); # *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - } - - &comment(""); - &add($b,32); - &add($r,32); - &sub($num,8); - &jnz(&label("pw_neg_loop")); - - &set_label("pw_neg_finish",0); - &mov($tmp2,&wparam(4)); # get dl - &mov($num,0); - &sub($num,$tmp2); - &and($num,7); - &jz(&label("pw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("dl<0 Tail Round $i"); - &mov($tmp1,0); - &mov($tmp2,&DWP($i*4,$b,"",0));# *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &dec($num) if ($i != 6); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &jz(&label("pw_end")) if ($i != 6); - } - - &jmp(&label("pw_end")); - - &set_label("pw_pos",0); - - &and($num,0xfffffff8); # num / 8 - &jz(&label("pw_pos_finish")); - - &set_label("pw_pos_loop",0); - - for ($i=0; $i<8; $i++) - { - &comment("dl>0 Round $i"); - - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &sub($tmp1,$c); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &jnc(&label("pw_nc".$i)); - } - - &comment(""); - &add($a,32); - &add($r,32); - &sub($num,8); - &jnz(&label("pw_pos_loop")); - - &set_label("pw_pos_finish",0); - &mov($num,&wparam(4)); # get dl - &and($num,7); - &jz(&label("pw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("dl>0 Tail Round $i"); - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &sub($tmp1,$c); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &jnc(&label("pw_tail_nc".$i)); - &dec($num) if ($i != 6); - &jz(&label("pw_end")) if ($i != 6); - } - &mov($c,1); - &jmp(&label("pw_end")); - - &set_label("pw_nc_loop",0); - for ($i=0; $i<8; $i++) - { - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &set_label("pw_nc".$i,0); - } - - &comment(""); - &add($a,32); - &add($r,32); - &sub($num,8); - &jnz(&label("pw_nc_loop")); - - &mov($num,&wparam(4)); # get dl - &and($num,7); - &jz(&label("pw_nc_end")); - - for ($i=0; $i<7; $i++) - { - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - &set_label("pw_tail_nc".$i,0); - &dec($num) if ($i != 6); - &jz(&label("pw_nc_end")) if ($i != 6); - } - - &set_label("pw_nc_end",0); - &mov($c,0); - - &set_label("pw_end",0); - -# &mov("eax",$c); # $c is "eax" - - &function_end($name); - } - diff --git a/src/lib/libcrypto/bn/asm/co-586.pl b/src/lib/libcrypto/bn/asm/co-586.pl deleted file mode 100644 index 5d962cb957..0000000000 --- a/src/lib/libcrypto/bn/asm/co-586.pl +++ /dev/null @@ -1,286 +0,0 @@ -#!/usr/local/bin/perl - -push(@INC,"perlasm","../../perlasm"); -require "x86asm.pl"; - -&asm_init($ARGV[0],$0); - -&bn_mul_comba("bn_mul_comba8",8); -&bn_mul_comba("bn_mul_comba4",4); -&bn_sqr_comba("bn_sqr_comba8",8); -&bn_sqr_comba("bn_sqr_comba4",4); - -&asm_finish(); - -sub mul_add_c - { - local($a,$ai,$b,$bi,$c0,$c1,$c2,$pos,$i,$na,$nb)=@_; - - # pos == -1 if eax and edx are pre-loaded, 0 to load from next - # words, and 1 if load return value - - &comment("mul a[$ai]*b[$bi]"); - - # "eax" and "edx" will always be pre-loaded. - # &mov("eax",&DWP($ai*4,$a,"",0)) ; - # &mov("edx",&DWP($bi*4,$b,"",0)); - - &mul("edx"); - &add($c0,"eax"); - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 0; # laod next a - &mov("eax",&wparam(0)) if $pos > 0; # load r[] - ### - &adc($c1,"edx"); - &mov("edx",&DWP(($nb)*4,$b,"",0)) if $pos == 0; # laod next b - &mov("edx",&DWP(($nb)*4,$b,"",0)) if $pos == 1; # laod next b - ### - &adc($c2,0); - # is pos > 1, it means it is the last loop - &mov(&DWP($i*4,"eax","",0),$c0) if $pos > 0; # save r[]; - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 1; # laod next a - } - -sub sqr_add_c - { - local($r,$a,$ai,$bi,$c0,$c1,$c2,$pos,$i,$na,$nb)=@_; - - # pos == -1 if eax and edx are pre-loaded, 0 to load from next - # words, and 1 if load return value - - &comment("sqr a[$ai]*a[$bi]"); - - # "eax" and "edx" will always be pre-loaded. - # &mov("eax",&DWP($ai*4,$a,"",0)) ; - # &mov("edx",&DWP($bi*4,$b,"",0)); - - if ($ai == $bi) - { &mul("eax");} - else - { &mul("edx");} - &add($c0,"eax"); - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 0; # load next a - ### - &adc($c1,"edx"); - &mov("edx",&DWP(($nb)*4,$a,"",0)) if ($pos == 1) && ($na != $nb); - ### - &adc($c2,0); - # is pos > 1, it means it is the last loop - &mov(&DWP($i*4,$r,"",0),$c0) if $pos > 0; # save r[]; - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 1; # load next b - } - -sub sqr_add_c2 - { - local($r,$a,$ai,$bi,$c0,$c1,$c2,$pos,$i,$na,$nb)=@_; - - # pos == -1 if eax and edx are pre-loaded, 0 to load from next - # words, and 1 if load return value - - &comment("sqr a[$ai]*a[$bi]"); - - # "eax" and "edx" will always be pre-loaded. - # &mov("eax",&DWP($ai*4,$a,"",0)) ; - # &mov("edx",&DWP($bi*4,$a,"",0)); - - if ($ai == $bi) - { &mul("eax");} - else - { &mul("edx");} - &add("eax","eax"); - ### - &adc("edx","edx"); - ### - &adc($c2,0); - &add($c0,"eax"); - &adc($c1,"edx"); - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 0; # load next a - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 1; # load next b - &adc($c2,0); - &mov(&DWP($i*4,$r,"",0),$c0) if $pos > 0; # save r[]; - &mov("edx",&DWP(($nb)*4,$a,"",0)) if ($pos <= 1) && ($na != $nb); - ### - } - -sub bn_mul_comba - { - local($name,$num)=@_; - local($a,$b,$c0,$c1,$c2); - local($i,$as,$ae,$bs,$be,$ai,$bi); - local($tot,$end); - - &function_begin_B($name,""); - - $c0="ebx"; - $c1="ecx"; - $c2="ebp"; - $a="esi"; - $b="edi"; - - $as=0; - $ae=0; - $bs=0; - $be=0; - $tot=$num+$num-1; - - &push("esi"); - &mov($a,&wparam(1)); - &push("edi"); - &mov($b,&wparam(2)); - &push("ebp"); - &push("ebx"); - - &xor($c0,$c0); - &mov("eax",&DWP(0,$a,"",0)); # load the first word - &xor($c1,$c1); - &mov("edx",&DWP(0,$b,"",0)); # load the first second - - for ($i=0; $i<$tot; $i++) - { - $ai=$as; - $bi=$bs; - $end=$be+1; - - &comment("################## Calculate word $i"); - - for ($j=$bs; $j<$end; $j++) - { - &xor($c2,$c2) if ($j == $bs); - if (($j+1) == $end) - { - $v=1; - $v=2 if (($i+1) == $tot); - } - else - { $v=0; } - if (($j+1) != $end) - { - $na=($ai-1); - $nb=($bi+1); - } - else - { - $na=$as+($i < ($num-1)); - $nb=$bs+($i >= ($num-1)); - } -#printf STDERR "[$ai,$bi] -> [$na,$nb]\n"; - &mul_add_c($a,$ai,$b,$bi,$c0,$c1,$c2,$v,$i,$na,$nb); - if ($v) - { - &comment("saved r[$i]"); - # &mov("eax",&wparam(0)); - # &mov(&DWP($i*4,"eax","",0),$c0); - ($c0,$c1,$c2)=($c1,$c2,$c0); - } - $ai--; - $bi++; - } - $as++ if ($i < ($num-1)); - $ae++ if ($i >= ($num-1)); - - $bs++ if ($i >= ($num-1)); - $be++ if ($i < ($num-1)); - } - &comment("save r[$i]"); - # &mov("eax",&wparam(0)); - &mov(&DWP($i*4,"eax","",0),$c0); - - &pop("ebx"); - &pop("ebp"); - &pop("edi"); - &pop("esi"); - &ret(); - &function_end_B($name); - } - -sub bn_sqr_comba - { - local($name,$num)=@_; - local($r,$a,$c0,$c1,$c2)=@_; - local($i,$as,$ae,$bs,$be,$ai,$bi); - local($b,$tot,$end,$half); - - &function_begin_B($name,""); - - $c0="ebx"; - $c1="ecx"; - $c2="ebp"; - $a="esi"; - $r="edi"; - - &push("esi"); - &push("edi"); - &push("ebp"); - &push("ebx"); - &mov($r,&wparam(0)); - &mov($a,&wparam(1)); - &xor($c0,$c0); - &xor($c1,$c1); - &mov("eax",&DWP(0,$a,"",0)); # load the first word - - $as=0; - $ae=0; - $bs=0; - $be=0; - $tot=$num+$num-1; - - for ($i=0; $i<$tot; $i++) - { - $ai=$as; - $bi=$bs; - $end=$be+1; - - &comment("############### Calculate word $i"); - for ($j=$bs; $j<$end; $j++) - { - &xor($c2,$c2) if ($j == $bs); - if (($ai-1) < ($bi+1)) - { - $v=1; - $v=2 if ($i+1) == $tot; - } - else - { $v=0; } - if (!$v) - { - $na=$ai-1; - $nb=$bi+1; - } - else - { - $na=$as+($i < ($num-1)); - $nb=$bs+($i >= ($num-1)); - } - if ($ai == $bi) - { - &sqr_add_c($r,$a,$ai,$bi, - $c0,$c1,$c2,$v,$i,$na,$nb); - } - else - { - &sqr_add_c2($r,$a,$ai,$bi, - $c0,$c1,$c2,$v,$i,$na,$nb); - } - if ($v) - { - &comment("saved r[$i]"); - #&mov(&DWP($i*4,$r,"",0),$c0); - ($c0,$c1,$c2)=($c1,$c2,$c0); - last; - } - $ai--; - $bi++; - } - $as++ if ($i < ($num-1)); - $ae++ if ($i >= ($num-1)); - - $bs++ if ($i >= ($num-1)); - $be++ if ($i < ($num-1)); - } - &mov(&DWP($i*4,$r,"",0),$c0); - &pop("ebx"); - &pop("ebp"); - &pop("edi"); - &pop("esi"); - &ret(); - &function_end_B($name); - } diff --git a/src/lib/libcrypto/bn/asm/ia64.S b/src/lib/libcrypto/bn/asm/ia64.S deleted file mode 100644 index 7b82b820e6..0000000000 --- a/src/lib/libcrypto/bn/asm/ia64.S +++ /dev/null @@ -1,1560 +0,0 @@ -.explicit -.text -.ident "ia64.S, Version 2.1" -.ident "IA-64 ISA artwork by Andy Polyakov " - -// -// ==================================================================== -// Written by Andy Polyakov for the OpenSSL -// project. -// -// Rights for redistribution and usage in source and binary forms are -// granted according to the OpenSSL license. Warranty of any kind is -// disclaimed. -// ==================================================================== -// -// Version 2.x is Itanium2 re-tune. Few words about how Itanum2 is -// different from Itanium to this module viewpoint. Most notably, is it -// "wider" than Itanium? Can you experience loop scalability as -// discussed in commentary sections? Not really:-( Itanium2 has 6 -// integer ALU ports, i.e. it's 2 ports wider, but it's not enough to -// spin twice as fast, as I need 8 IALU ports. Amount of floating point -// ports is the same, i.e. 2, while I need 4. In other words, to this -// module Itanium2 remains effectively as "wide" as Itanium. Yet it's -// essentially different in respect to this module, and a re-tune was -// required. Well, because some intruction latencies has changed. Most -// noticeably those intensively used: -// -// Itanium Itanium2 -// ldf8 9 6 L2 hit -// ld8 2 1 L1 hit -// getf 2 5 -// xma[->getf] 7[+1] 4[+0] -// add[->st8] 1[+1] 1[+0] -// -// What does it mean? You might ratiocinate that the original code -// should run just faster... Because sum of latencies is smaller... -// Wrong! Note that getf latency increased. This means that if a loop is -// scheduled for lower latency (as they were), then it will suffer from -// stall condition and the code will therefore turn anti-scalable, e.g. -// original bn_mul_words spun at 5*n or 2.5 times slower than expected -// on Itanium2! What to do? Reschedule loops for Itanium2? But then -// Itanium would exhibit anti-scalability. So I've chosen to reschedule -// for worst latency for every instruction aiming for best *all-round* -// performance. - -// Q. How much faster does it get? -// A. Here is the output from 'openssl speed rsa dsa' for vanilla -// 0.9.6a compiled with gcc version 2.96 20000731 (Red Hat -// Linux 7.1 2.96-81): -// -// sign verify sign/s verify/s -// rsa 512 bits 0.0036s 0.0003s 275.3 2999.2 -// rsa 1024 bits 0.0203s 0.0011s 49.3 894.1 -// rsa 2048 bits 0.1331s 0.0040s 7.5 250.9 -// rsa 4096 bits 0.9270s 0.0147s 1.1 68.1 -// sign verify sign/s verify/s -// dsa 512 bits 0.0035s 0.0043s 288.3 234.8 -// dsa 1024 bits 0.0111s 0.0135s 90.0 74.2 -// -// And here is similar output but for this assembler -// implementation:-) -// -// sign verify sign/s verify/s -// rsa 512 bits 0.0021s 0.0001s 549.4 9638.5 -// rsa 1024 bits 0.0055s 0.0002s 183.8 4481.1 -// rsa 2048 bits 0.0244s 0.0006s 41.4 1726.3 -// rsa 4096 bits 0.1295s 0.0018s 7.7 561.5 -// sign verify sign/s verify/s -// dsa 512 bits 0.0012s 0.0013s 891.9 756.6 -// dsa 1024 bits 0.0023s 0.0028s 440.4 376.2 -// -// Yes, you may argue that it's not fair comparison as it's -// possible to craft the C implementation with BN_UMULT_HIGH -// inline assembler macro. But of course! Here is the output -// with the macro: -// -// sign verify sign/s verify/s -// rsa 512 bits 0.0020s 0.0002s 495.0 6561.0 -// rsa 1024 bits 0.0086s 0.0004s 116.2 2235.7 -// rsa 2048 bits 0.0519s 0.0015s 19.3 667.3 -// rsa 4096 bits 0.3464s 0.0053s 2.9 187.7 -// sign verify sign/s verify/s -// dsa 512 bits 0.0016s 0.0020s 613.1 510.5 -// dsa 1024 bits 0.0045s 0.0054s 221.0 183.9 -// -// My code is still way faster, huh:-) And I believe that even -// higher performance can be achieved. Note that as keys get -// longer, performance gain is larger. Why? According to the -// profiler there is another player in the field, namely -// BN_from_montgomery consuming larger and larger portion of CPU -// time as keysize decreases. I therefore consider putting effort -// to assembler implementation of the following routine: -// -// void bn_mul_add_mont (BN_ULONG *rp,BN_ULONG *np,int nl,BN_ULONG n0) -// { -// int i,j; -// BN_ULONG v; -// -// for (i=0; i= 0,num,bn_mul_add_words_exit ; if (num <= 0) then exit - LDO 128(%sp),%sp ; bump stack - - ; - ; The loop is unrolled twice, so if there is only 1 number - ; then go straight to the cleanup code. - ; - CMPIB,= 1,num,bn_mul_add_words_single_top - FLDD -184(%sp),fw ; (-56-128) load up w into fw (fw_h/fw_l) - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; - ; PA-RISC 2.0 chips have two fully pipelined multipliers, thus - ; two 32-bit mutiplies can be issued per cycle. - ; -bn_mul_add_words_unroll2 - - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - FLDD 8(a_ptr),t_float_1 ; load up 64-bit value (fr8L) ht(L)/lt(R) - LDD 0(r_ptr),rp_val ; rp[0] - LDD 8(r_ptr),rp_val_1 ; rp[1] - - XMPYU fht_0,fw_l,fm1 ; m1[0] = fht_0*fw_l - XMPYU fht_1,fw_l,fm1_1 ; m1[1] = fht_1*fw_l - FSTD fm1,-16(%sp) ; -16(sp) = m1[0] - FSTD fm1_1,-48(%sp) ; -48(sp) = m1[1] - - XMPYU flt_0,fw_h,fm ; m[0] = flt_0*fw_h - XMPYU flt_1,fw_h,fm_1 ; m[1] = flt_1*fw_h - FSTD fm,-8(%sp) ; -8(sp) = m[0] - FSTD fm_1,-40(%sp) ; -40(sp) = m[1] - - XMPYU fht_0,fw_h,ht_temp ; ht_temp = fht_0*fw_h - XMPYU fht_1,fw_h,ht_temp_1 ; ht_temp_1 = fht_1*fw_h - FSTD ht_temp,-24(%sp) ; -24(sp) = ht_temp - FSTD ht_temp_1,-56(%sp) ; -56(sp) = ht_temp_1 - - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - XMPYU flt_1,fw_l,lt_temp_1 ; lt_temp = lt*fw_l - FSTD lt_temp,-32(%sp) ; -32(sp) = lt_temp - FSTD lt_temp_1,-64(%sp) ; -64(sp) = lt_temp_1 - - LDD -8(%sp),m_0 ; m[0] - LDD -40(%sp),m_1 ; m[1] - LDD -16(%sp),m1_0 ; m1[0] - LDD -48(%sp),m1_1 ; m1[1] - - LDD -24(%sp),ht_0 ; ht[0] - LDD -56(%sp),ht_1 ; ht[1] - ADD,L m1_0,m_0,tmp_0 ; tmp_0 = m[0] + m1[0]; - ADD,L m1_1,m_1,tmp_1 ; tmp_1 = m[1] + m1[1]; - - LDD -32(%sp),lt_0 - LDD -64(%sp),lt_1 - CMPCLR,*>>= tmp_0,m1_0, %r0 ; if (m[0] < m1[0]) - ADD,L ht_0,top_overflow,ht_0 ; ht[0] += (1<<32) - - CMPCLR,*>>= tmp_1,m1_1,%r0 ; if (m[1] < m1[1]) - ADD,L ht_1,top_overflow,ht_1 ; ht[1] += (1<<32) - EXTRD,U tmp_0,31,32,m_0 ; m[0]>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1[0] = m[0]<<32 - - EXTRD,U tmp_1,31,32,m_1 ; m[1]>>32 - DEPD,Z tmp_1,31,32,m1_1 ; m1[1] = m[1]<<32 - ADD,L ht_0,m_0,ht_0 ; ht[0]+= (m[0]>>32) - ADD,L ht_1,m_1,ht_1 ; ht[1]+= (m[1]>>32) - - ADD lt_0,m1_0,lt_0 ; lt[0] = lt[0]+m1[0]; - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - ADD lt_1,m1_1,lt_1 ; lt[1] = lt[1]+m1[1]; - ADD,DC ht_1,%r0,ht_1 ; ht[1]++ - - ADD %ret1,lt_0,lt_0 ; lt[0] = lt[0] + c; - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - ADD lt_0,rp_val,lt_0 ; lt[0] = lt[0]+rp[0] - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - - LDO -2(num),num ; num = num - 2; - ADD ht_0,lt_1,lt_1 ; lt[1] = lt[1] + ht_0 (c); - ADD,DC ht_1,%r0,ht_1 ; ht[1]++ - STD lt_0,0(r_ptr) ; rp[0] = lt[0] - - ADD lt_1,rp_val_1,lt_1 ; lt[1] = lt[1]+rp[1] - ADD,DC ht_1,%r0,%ret1 ; ht[1]++ - LDO 16(a_ptr),a_ptr ; a_ptr += 2 - - STD lt_1,8(r_ptr) ; rp[1] = lt[1] - CMPIB,<= 2,num,bn_mul_add_words_unroll2 ; go again if more to do - LDO 16(r_ptr),r_ptr ; r_ptr += 2 - - CMPIB,=,N 0,num,bn_mul_add_words_exit ; are we done, or cleanup last one - - ; - ; Top of loop aligned on 64-byte boundary - ; -bn_mul_add_words_single_top - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - LDD 0(r_ptr),rp_val ; rp[0] - LDO 8(a_ptr),a_ptr ; a_ptr++ - XMPYU fht_0,fw_l,fm1 ; m1 = ht*fw_l - FSTD fm1,-16(%sp) ; -16(sp) = m1 - XMPYU flt_0,fw_h,fm ; m = lt*fw_h - FSTD fm,-8(%sp) ; -8(sp) = m - XMPYU fht_0,fw_h,ht_temp ; ht_temp = ht*fw_h - FSTD ht_temp,-24(%sp) ; -24(sp) = ht - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - FSTD lt_temp,-32(%sp) ; -32(sp) = lt - - LDD -8(%sp),m_0 - LDD -16(%sp),m1_0 ; m1 = temp1 - ADD,L m_0,m1_0,tmp_0 ; tmp_0 = m + m1; - LDD -24(%sp),ht_0 - LDD -32(%sp),lt_0 - - CMPCLR,*>>= tmp_0,m1_0,%r0 ; if (m < m1) - ADD,L ht_0,top_overflow,ht_0 ; ht += (1<<32) - - EXTRD,U tmp_0,31,32,m_0 ; m>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1 = m<<32 - - ADD,L ht_0,m_0,ht_0 ; ht+= (m>>32) - ADD lt_0,m1_0,tmp_0 ; tmp_0 = lt+m1; - ADD,DC ht_0,%r0,ht_0 ; ht++ - ADD %ret1,tmp_0,lt_0 ; lt = lt + c; - ADD,DC ht_0,%r0,ht_0 ; ht++ - ADD lt_0,rp_val,lt_0 ; lt = lt+rp[0] - ADD,DC ht_0,%r0,%ret1 ; ht++ - STD lt_0,0(r_ptr) ; rp[0] = lt - -bn_mul_add_words_exit - .EXIT - - EXTRD,U %ret1,31,32,%ret0 ; for 32-bit, return in ret0/ret1 - LDD -80(%sp),%r9 ; restore r9 - LDD -88(%sp),%r8 ; restore r8 - LDD -96(%sp),%r7 ; restore r7 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 ; restore r3 - .PROCEND ;in=23,24,25,26,29;out=28; - -;---------------------------------------------------------------------------- -; -;BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w) -; -; arg0 = rp -; arg1 = ap -; arg3 = num -; w on stack at -56(sp) - -bn_mul_words - .proc - .callinfo frame=128 - .entry - .EXPORT bn_mul_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - NOP - STD %r5,16(%sp) ; save r5 - - STD %r6,24(%sp) ; save r6 - STD %r7,32(%sp) ; save r7 - COPY %r0,%ret1 ; return 0 by default - DEPDI,Z 1,31,1,top_overflow ; top_overflow = 1 << 32 - - CMPIB,>= 0,num,bn_mul_words_exit - LDO 128(%sp),%sp ; bump stack - - ; - ; See if only 1 word to do, thus just do cleanup - ; - CMPIB,= 1,num,bn_mul_words_single_top - FLDD -184(%sp),fw ; (-56-128) load up w into fw (fw_h/fw_l) - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; - ; PA-RISC 2.0 chips have two fully pipelined multipliers, thus - ; two 32-bit mutiplies can be issued per cycle. - ; -bn_mul_words_unroll2 - - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - FLDD 8(a_ptr),t_float_1 ; load up 64-bit value (fr8L) ht(L)/lt(R) - XMPYU fht_0,fw_l,fm1 ; m1[0] = fht_0*fw_l - XMPYU fht_1,fw_l,fm1_1 ; m1[1] = ht*fw_l - - FSTD fm1,-16(%sp) ; -16(sp) = m1 - FSTD fm1_1,-48(%sp) ; -48(sp) = m1 - XMPYU flt_0,fw_h,fm ; m = lt*fw_h - XMPYU flt_1,fw_h,fm_1 ; m = lt*fw_h - - FSTD fm,-8(%sp) ; -8(sp) = m - FSTD fm_1,-40(%sp) ; -40(sp) = m - XMPYU fht_0,fw_h,ht_temp ; ht_temp = fht_0*fw_h - XMPYU fht_1,fw_h,ht_temp_1 ; ht_temp = ht*fw_h - - FSTD ht_temp,-24(%sp) ; -24(sp) = ht - FSTD ht_temp_1,-56(%sp) ; -56(sp) = ht - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - XMPYU flt_1,fw_l,lt_temp_1 ; lt_temp = lt*fw_l - - FSTD lt_temp,-32(%sp) ; -32(sp) = lt - FSTD lt_temp_1,-64(%sp) ; -64(sp) = lt - LDD -8(%sp),m_0 - LDD -40(%sp),m_1 - - LDD -16(%sp),m1_0 - LDD -48(%sp),m1_1 - LDD -24(%sp),ht_0 - LDD -56(%sp),ht_1 - - ADD,L m1_0,m_0,tmp_0 ; tmp_0 = m + m1; - ADD,L m1_1,m_1,tmp_1 ; tmp_1 = m + m1; - LDD -32(%sp),lt_0 - LDD -64(%sp),lt_1 - - CMPCLR,*>>= tmp_0,m1_0, %r0 ; if (m < m1) - ADD,L ht_0,top_overflow,ht_0 ; ht += (1<<32) - CMPCLR,*>>= tmp_1,m1_1,%r0 ; if (m < m1) - ADD,L ht_1,top_overflow,ht_1 ; ht += (1<<32) - - EXTRD,U tmp_0,31,32,m_0 ; m>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1 = m<<32 - EXTRD,U tmp_1,31,32,m_1 ; m>>32 - DEPD,Z tmp_1,31,32,m1_1 ; m1 = m<<32 - - ADD,L ht_0,m_0,ht_0 ; ht+= (m>>32) - ADD,L ht_1,m_1,ht_1 ; ht+= (m>>32) - ADD lt_0,m1_0,lt_0 ; lt = lt+m1; - ADD,DC ht_0,%r0,ht_0 ; ht++ - - ADD lt_1,m1_1,lt_1 ; lt = lt+m1; - ADD,DC ht_1,%r0,ht_1 ; ht++ - ADD %ret1,lt_0,lt_0 ; lt = lt + c (ret1); - ADD,DC ht_0,%r0,ht_0 ; ht++ - - ADD ht_0,lt_1,lt_1 ; lt = lt + c (ht_0) - ADD,DC ht_1,%r0,ht_1 ; ht++ - STD lt_0,0(r_ptr) ; rp[0] = lt - STD lt_1,8(r_ptr) ; rp[1] = lt - - COPY ht_1,%ret1 ; carry = ht - LDO -2(num),num ; num = num - 2; - LDO 16(a_ptr),a_ptr ; ap += 2 - CMPIB,<= 2,num,bn_mul_words_unroll2 - LDO 16(r_ptr),r_ptr ; rp++ - - CMPIB,=,N 0,num,bn_mul_words_exit ; are we done? - - ; - ; Top of loop aligned on 64-byte boundary - ; -bn_mul_words_single_top - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - - XMPYU fht_0,fw_l,fm1 ; m1 = ht*fw_l - FSTD fm1,-16(%sp) ; -16(sp) = m1 - XMPYU flt_0,fw_h,fm ; m = lt*fw_h - FSTD fm,-8(%sp) ; -8(sp) = m - XMPYU fht_0,fw_h,ht_temp ; ht_temp = ht*fw_h - FSTD ht_temp,-24(%sp) ; -24(sp) = ht - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - FSTD lt_temp,-32(%sp) ; -32(sp) = lt - - LDD -8(%sp),m_0 - LDD -16(%sp),m1_0 - ADD,L m_0,m1_0,tmp_0 ; tmp_0 = m + m1; - LDD -24(%sp),ht_0 - LDD -32(%sp),lt_0 - - CMPCLR,*>>= tmp_0,m1_0,%r0 ; if (m < m1) - ADD,L ht_0,top_overflow,ht_0 ; ht += (1<<32) - - EXTRD,U tmp_0,31,32,m_0 ; m>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1 = m<<32 - - ADD,L ht_0,m_0,ht_0 ; ht+= (m>>32) - ADD lt_0,m1_0,lt_0 ; lt= lt+m1; - ADD,DC ht_0,%r0,ht_0 ; ht++ - - ADD %ret1,lt_0,lt_0 ; lt = lt + c; - ADD,DC ht_0,%r0,ht_0 ; ht++ - - COPY ht_0,%ret1 ; copy carry - STD lt_0,0(r_ptr) ; rp[0] = lt - -bn_mul_words_exit - .EXIT - EXTRD,U %ret1,31,32,%ret0 ; for 32-bit, return in ret0/ret1 - LDD -96(%sp),%r7 ; restore r7 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 ; restore r3 - .PROCEND - -;---------------------------------------------------------------------------- -; -;void bn_sqr_words(BN_ULONG *rp, BN_ULONG *ap, int num) -; -; arg0 = rp -; arg1 = ap -; arg2 = num -; - -bn_sqr_words - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_sqr_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - NOP - STD %r5,16(%sp) ; save r5 - - CMPIB,>= 0,num,bn_sqr_words_exit - LDO 128(%sp),%sp ; bump stack - - ; - ; If only 1, the goto straight to cleanup - ; - CMPIB,= 1,num,bn_sqr_words_single_top - DEPDI,Z -1,32,33,high_mask ; Create Mask 0xffffffff80000000L - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; - -bn_sqr_words_unroll2 - FLDD 0(a_ptr),t_float_0 ; a[0] - FLDD 8(a_ptr),t_float_1 ; a[1] - XMPYU fht_0,flt_0,fm ; m[0] - XMPYU fht_1,flt_1,fm_1 ; m[1] - - FSTD fm,-24(%sp) ; store m[0] - FSTD fm_1,-56(%sp) ; store m[1] - XMPYU flt_0,flt_0,lt_temp ; lt[0] - XMPYU flt_1,flt_1,lt_temp_1 ; lt[1] - - FSTD lt_temp,-16(%sp) ; store lt[0] - FSTD lt_temp_1,-48(%sp) ; store lt[1] - XMPYU fht_0,fht_0,ht_temp ; ht[0] - XMPYU fht_1,fht_1,ht_temp_1 ; ht[1] - - FSTD ht_temp,-8(%sp) ; store ht[0] - FSTD ht_temp_1,-40(%sp) ; store ht[1] - LDD -24(%sp),m_0 - LDD -56(%sp),m_1 - - AND m_0,high_mask,tmp_0 ; m[0] & Mask - AND m_1,high_mask,tmp_1 ; m[1] & Mask - DEPD,Z m_0,30,31,m_0 ; m[0] << 32+1 - DEPD,Z m_1,30,31,m_1 ; m[1] << 32+1 - - LDD -16(%sp),lt_0 - LDD -48(%sp),lt_1 - EXTRD,U tmp_0,32,33,tmp_0 ; tmp_0 = m[0]&Mask >> 32-1 - EXTRD,U tmp_1,32,33,tmp_1 ; tmp_1 = m[1]&Mask >> 32-1 - - LDD -8(%sp),ht_0 - LDD -40(%sp),ht_1 - ADD,L ht_0,tmp_0,ht_0 ; ht[0] += tmp_0 - ADD,L ht_1,tmp_1,ht_1 ; ht[1] += tmp_1 - - ADD lt_0,m_0,lt_0 ; lt = lt+m - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - STD lt_0,0(r_ptr) ; rp[0] = lt[0] - STD ht_0,8(r_ptr) ; rp[1] = ht[1] - - ADD lt_1,m_1,lt_1 ; lt = lt+m - ADD,DC ht_1,%r0,ht_1 ; ht[1]++ - STD lt_1,16(r_ptr) ; rp[2] = lt[1] - STD ht_1,24(r_ptr) ; rp[3] = ht[1] - - LDO -2(num),num ; num = num - 2; - LDO 16(a_ptr),a_ptr ; ap += 2 - CMPIB,<= 2,num,bn_sqr_words_unroll2 - LDO 32(r_ptr),r_ptr ; rp += 4 - - CMPIB,=,N 0,num,bn_sqr_words_exit ; are we done? - - ; - ; Top of loop aligned on 64-byte boundary - ; -bn_sqr_words_single_top - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - - XMPYU fht_0,flt_0,fm ; m - FSTD fm,-24(%sp) ; store m - - XMPYU flt_0,flt_0,lt_temp ; lt - FSTD lt_temp,-16(%sp) ; store lt - - XMPYU fht_0,fht_0,ht_temp ; ht - FSTD ht_temp,-8(%sp) ; store ht - - LDD -24(%sp),m_0 ; load m - AND m_0,high_mask,tmp_0 ; m & Mask - DEPD,Z m_0,30,31,m_0 ; m << 32+1 - LDD -16(%sp),lt_0 ; lt - - LDD -8(%sp),ht_0 ; ht - EXTRD,U tmp_0,32,33,tmp_0 ; tmp_0 = m&Mask >> 32-1 - ADD m_0,lt_0,lt_0 ; lt = lt+m - ADD,L ht_0,tmp_0,ht_0 ; ht += tmp_0 - ADD,DC ht_0,%r0,ht_0 ; ht++ - - STD lt_0,0(r_ptr) ; rp[0] = lt - STD ht_0,8(r_ptr) ; rp[1] = ht - -bn_sqr_words_exit - .EXIT - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - .PROCEND ;in=23,24,25,26,29;out=28; - - -;---------------------------------------------------------------------------- -; -;BN_ULONG bn_add_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n) -; -; arg0 = rp -; arg1 = ap -; arg2 = bp -; arg3 = n - -t .reg %r22 -b .reg %r21 -l .reg %r20 - -bn_add_words - .proc - .entry - .callinfo - .EXPORT bn_add_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .align 64 - - CMPIB,>= 0,n,bn_add_words_exit - COPY %r0,%ret1 ; return 0 by default - - ; - ; If 2 or more numbers do the loop - ; - CMPIB,= 1,n,bn_add_words_single_top - NOP - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; -bn_add_words_unroll2 - LDD 0(a_ptr),t - LDD 0(b_ptr),b - ADD t,%ret1,t ; t = t+c; - ADD,DC %r0,%r0,%ret1 ; set c to carry - ADD t,b,l ; l = t + b[0] - ADD,DC %ret1,%r0,%ret1 ; c+= carry - STD l,0(r_ptr) - - LDD 8(a_ptr),t - LDD 8(b_ptr),b - ADD t,%ret1,t ; t = t+c; - ADD,DC %r0,%r0,%ret1 ; set c to carry - ADD t,b,l ; l = t + b[0] - ADD,DC %ret1,%r0,%ret1 ; c+= carry - STD l,8(r_ptr) - - LDO -2(n),n - LDO 16(a_ptr),a_ptr - LDO 16(b_ptr),b_ptr - - CMPIB,<= 2,n,bn_add_words_unroll2 - LDO 16(r_ptr),r_ptr - - CMPIB,=,N 0,n,bn_add_words_exit ; are we done? - -bn_add_words_single_top - LDD 0(a_ptr),t - LDD 0(b_ptr),b - - ADD t,%ret1,t ; t = t+c; - ADD,DC %r0,%r0,%ret1 ; set c to carry (could use CMPCLR??) - ADD t,b,l ; l = t + b[0] - ADD,DC %ret1,%r0,%ret1 ; c+= carry - STD l,0(r_ptr) - -bn_add_words_exit - .EXIT - BVE (%rp) - EXTRD,U %ret1,31,32,%ret0 ; for 32-bit, return in ret0/ret1 - .PROCEND ;in=23,24,25,26,29;out=28; - -;---------------------------------------------------------------------------- -; -;BN_ULONG bn_sub_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n) -; -; arg0 = rp -; arg1 = ap -; arg2 = bp -; arg3 = n - -t1 .reg %r22 -t2 .reg %r21 -sub_tmp1 .reg %r20 -sub_tmp2 .reg %r19 - - -bn_sub_words - .proc - .callinfo - .EXPORT bn_sub_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - CMPIB,>= 0,n,bn_sub_words_exit - COPY %r0,%ret1 ; return 0 by default - - ; - ; If 2 or more numbers do the loop - ; - CMPIB,= 1,n,bn_sub_words_single_top - NOP - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; -bn_sub_words_unroll2 - LDD 0(a_ptr),t1 - LDD 0(b_ptr),t2 - SUB t1,t2,sub_tmp1 ; t3 = t1-t2; - SUB sub_tmp1,%ret1,sub_tmp1 ; t3 = t3- c; - - CMPCLR,*>> t1,t2,sub_tmp2 ; clear if t1 > t2 - LDO 1(%r0),sub_tmp2 - - CMPCLR,*= t1,t2,%r0 - COPY sub_tmp2,%ret1 - STD sub_tmp1,0(r_ptr) - - LDD 8(a_ptr),t1 - LDD 8(b_ptr),t2 - SUB t1,t2,sub_tmp1 ; t3 = t1-t2; - SUB sub_tmp1,%ret1,sub_tmp1 ; t3 = t3- c; - CMPCLR,*>> t1,t2,sub_tmp2 ; clear if t1 > t2 - LDO 1(%r0),sub_tmp2 - - CMPCLR,*= t1,t2,%r0 - COPY sub_tmp2,%ret1 - STD sub_tmp1,8(r_ptr) - - LDO -2(n),n - LDO 16(a_ptr),a_ptr - LDO 16(b_ptr),b_ptr - - CMPIB,<= 2,n,bn_sub_words_unroll2 - LDO 16(r_ptr),r_ptr - - CMPIB,=,N 0,n,bn_sub_words_exit ; are we done? - -bn_sub_words_single_top - LDD 0(a_ptr),t1 - LDD 0(b_ptr),t2 - SUB t1,t2,sub_tmp1 ; t3 = t1-t2; - SUB sub_tmp1,%ret1,sub_tmp1 ; t3 = t3- c; - CMPCLR,*>> t1,t2,sub_tmp2 ; clear if t1 > t2 - LDO 1(%r0),sub_tmp2 - - CMPCLR,*= t1,t2,%r0 - COPY sub_tmp2,%ret1 - - STD sub_tmp1,0(r_ptr) - -bn_sub_words_exit - .EXIT - BVE (%rp) - EXTRD,U %ret1,31,32,%ret0 ; for 32-bit, return in ret0/ret1 - .PROCEND ;in=23,24,25,26,29;out=28; - -;------------------------------------------------------------------------------ -; -; unsigned long bn_div_words(unsigned long h, unsigned long l, unsigned long d) -; -; arg0 = h -; arg1 = l -; arg2 = d -; -; This is mainly just output from the HP C compiler. -; -;------------------------------------------------------------------------------ -bn_div_words - .PROC - .EXPORT bn_div_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR,LONG_RETURN - .IMPORT BN_num_bits_word,CODE - ;--- not PIC .IMPORT __iob,DATA - ;--- not PIC .IMPORT fprintf,CODE - .IMPORT abort,CODE - .IMPORT $$div2U,MILLICODE - .CALLINFO CALLER,FRAME=144,ENTRY_GR=%r9,SAVE_RP,ARGS_SAVED,ORDERING_AWARE - .ENTRY - STW %r2,-20(%r30) ;offset 0x8ec - STW,MA %r3,192(%r30) ;offset 0x8f0 - STW %r4,-188(%r30) ;offset 0x8f4 - DEPD %r5,31,32,%r6 ;offset 0x8f8 - STD %r6,-184(%r30) ;offset 0x8fc - DEPD %r7,31,32,%r8 ;offset 0x900 - STD %r8,-176(%r30) ;offset 0x904 - STW %r9,-168(%r30) ;offset 0x908 - LDD -248(%r30),%r3 ;offset 0x90c - COPY %r26,%r4 ;offset 0x910 - COPY %r24,%r5 ;offset 0x914 - DEPD %r25,31,32,%r4 ;offset 0x918 - CMPB,*<> %r3,%r0,$0006000C ;offset 0x91c - DEPD %r23,31,32,%r5 ;offset 0x920 - MOVIB,TR -1,%r29,$00060002 ;offset 0x924 - EXTRD,U %r29,31,32,%r28 ;offset 0x928 -$0006002A - LDO -1(%r29),%r29 ;offset 0x92c - SUB %r23,%r7,%r23 ;offset 0x930 -$00060024 - SUB %r4,%r31,%r25 ;offset 0x934 - AND %r25,%r19,%r26 ;offset 0x938 - CMPB,*<>,N %r0,%r26,$00060046 ;offset 0x93c - DEPD,Z %r25,31,32,%r20 ;offset 0x940 - OR %r20,%r24,%r21 ;offset 0x944 - CMPB,*<<,N %r21,%r23,$0006002A ;offset 0x948 - SUB %r31,%r2,%r31 ;offset 0x94c -$00060046 -$0006002E - DEPD,Z %r23,31,32,%r25 ;offset 0x950 - EXTRD,U %r23,31,32,%r26 ;offset 0x954 - AND %r25,%r19,%r24 ;offset 0x958 - ADD,L %r31,%r26,%r31 ;offset 0x95c - CMPCLR,*>>= %r5,%r24,%r0 ;offset 0x960 - LDO 1(%r31),%r31 ;offset 0x964 -$00060032 - CMPB,*<<=,N %r31,%r4,$00060036 ;offset 0x968 - LDO -1(%r29),%r29 ;offset 0x96c - ADD,L %r4,%r3,%r4 ;offset 0x970 -$00060036 - ADDIB,=,N -1,%r8,$D0 ;offset 0x974 - SUB %r5,%r24,%r28 ;offset 0x978 -$0006003A - SUB %r4,%r31,%r24 ;offset 0x97c - SHRPD %r24,%r28,32,%r4 ;offset 0x980 - DEPD,Z %r29,31,32,%r9 ;offset 0x984 - DEPD,Z %r28,31,32,%r5 ;offset 0x988 -$0006001C - EXTRD,U %r4,31,32,%r31 ;offset 0x98c - CMPB,*<>,N %r31,%r2,$00060020 ;offset 0x990 - MOVB,TR %r6,%r29,$D1 ;offset 0x994 - STD %r29,-152(%r30) ;offset 0x998 -$0006000C - EXTRD,U %r3,31,32,%r25 ;offset 0x99c - COPY %r3,%r26 ;offset 0x9a0 - EXTRD,U %r3,31,32,%r9 ;offset 0x9a4 - EXTRD,U %r4,31,32,%r8 ;offset 0x9a8 - .CALL ARGW0=GR,ARGW1=GR,RTNVAL=GR ;in=25,26;out=28; - B,L BN_num_bits_word,%r2 ;offset 0x9ac - EXTRD,U %r5,31,32,%r7 ;offset 0x9b0 - LDI 64,%r20 ;offset 0x9b4 - DEPD %r7,31,32,%r5 ;offset 0x9b8 - DEPD %r8,31,32,%r4 ;offset 0x9bc - DEPD %r9,31,32,%r3 ;offset 0x9c0 - CMPB,= %r28,%r20,$00060012 ;offset 0x9c4 - COPY %r28,%r24 ;offset 0x9c8 - MTSARCM %r24 ;offset 0x9cc - DEPDI,Z -1,%sar,1,%r19 ;offset 0x9d0 - CMPB,*>>,N %r4,%r19,$D2 ;offset 0x9d4 -$00060012 - SUBI 64,%r24,%r31 ;offset 0x9d8 - CMPCLR,*<< %r4,%r3,%r0 ;offset 0x9dc - SUB %r4,%r3,%r4 ;offset 0x9e0 -$00060016 - CMPB,= %r31,%r0,$0006001A ;offset 0x9e4 - COPY %r0,%r9 ;offset 0x9e8 - MTSARCM %r31 ;offset 0x9ec - DEPD,Z %r3,%sar,64,%r3 ;offset 0x9f0 - SUBI 64,%r31,%r26 ;offset 0x9f4 - MTSAR %r26 ;offset 0x9f8 - SHRPD %r4,%r5,%sar,%r4 ;offset 0x9fc - MTSARCM %r31 ;offset 0xa00 - DEPD,Z %r5,%sar,64,%r5 ;offset 0xa04 -$0006001A - DEPDI,Z -1,31,32,%r19 ;offset 0xa08 - AND %r3,%r19,%r29 ;offset 0xa0c - EXTRD,U %r29,31,32,%r2 ;offset 0xa10 - DEPDI,Z -1,63,32,%r6 ;offset 0xa14 - MOVIB,TR 2,%r8,$0006001C ;offset 0xa18 - EXTRD,U %r3,63,32,%r7 ;offset 0xa1c -$D2 - ;--- not PIC ADDIL LR'__iob-$global$,%r27,%r1 ;offset 0xa20 - ;--- not PIC LDIL LR'C$7,%r21 ;offset 0xa24 - ;--- not PIC LDO RR'__iob-$global$+32(%r1),%r26 ;offset 0xa28 - ;--- not PIC .CALL ARGW0=GR,ARGW1=GR,ARGW2=GR,RTNVAL=GR ;in=24,25,26;out=28; - ;--- not PIC B,L fprintf,%r2 ;offset 0xa2c - ;--- not PIC LDO RR'C$7(%r21),%r25 ;offset 0xa30 - .CALL ; - B,L abort,%r2 ;offset 0xa34 - NOP ;offset 0xa38 - B $D3 ;offset 0xa3c - LDW -212(%r30),%r2 ;offset 0xa40 -$00060020 - COPY %r4,%r26 ;offset 0xa44 - EXTRD,U %r4,31,32,%r25 ;offset 0xa48 - COPY %r2,%r24 ;offset 0xa4c - .CALL ;in=23,24,25,26;out=20,21,22,28,29; (MILLICALL) - B,L $$div2U,%r31 ;offset 0xa50 - EXTRD,U %r2,31,32,%r23 ;offset 0xa54 - DEPD %r28,31,32,%r29 ;offset 0xa58 -$00060022 - STD %r29,-152(%r30) ;offset 0xa5c -$D1 - AND %r5,%r19,%r24 ;offset 0xa60 - EXTRD,U %r24,31,32,%r24 ;offset 0xa64 - STW %r2,-160(%r30) ;offset 0xa68 - STW %r7,-128(%r30) ;offset 0xa6c - FLDD -152(%r30),%fr4 ;offset 0xa70 - FLDD -152(%r30),%fr7 ;offset 0xa74 - FLDW -160(%r30),%fr8L ;offset 0xa78 - FLDW -128(%r30),%fr5L ;offset 0xa7c - XMPYU %fr8L,%fr7L,%fr10 ;offset 0xa80 - FSTD %fr10,-136(%r30) ;offset 0xa84 - XMPYU %fr8L,%fr7R,%fr22 ;offset 0xa88 - FSTD %fr22,-144(%r30) ;offset 0xa8c - XMPYU %fr5L,%fr4L,%fr11 ;offset 0xa90 - XMPYU %fr5L,%fr4R,%fr23 ;offset 0xa94 - FSTD %fr11,-112(%r30) ;offset 0xa98 - FSTD %fr23,-120(%r30) ;offset 0xa9c - LDD -136(%r30),%r28 ;offset 0xaa0 - DEPD,Z %r28,31,32,%r31 ;offset 0xaa4 - LDD -144(%r30),%r20 ;offset 0xaa8 - ADD,L %r20,%r31,%r31 ;offset 0xaac - LDD -112(%r30),%r22 ;offset 0xab0 - DEPD,Z %r22,31,32,%r22 ;offset 0xab4 - LDD -120(%r30),%r21 ;offset 0xab8 - B $00060024 ;offset 0xabc - ADD,L %r21,%r22,%r23 ;offset 0xac0 -$D0 - OR %r9,%r29,%r29 ;offset 0xac4 -$00060040 - EXTRD,U %r29,31,32,%r28 ;offset 0xac8 -$00060002 -$L2 - LDW -212(%r30),%r2 ;offset 0xacc -$D3 - LDW -168(%r30),%r9 ;offset 0xad0 - LDD -176(%r30),%r8 ;offset 0xad4 - EXTRD,U %r8,31,32,%r7 ;offset 0xad8 - LDD -184(%r30),%r6 ;offset 0xadc - EXTRD,U %r6,31,32,%r5 ;offset 0xae0 - LDW -188(%r30),%r4 ;offset 0xae4 - BVE (%r2) ;offset 0xae8 - .EXIT - LDW,MB -192(%r30),%r3 ;offset 0xaec - .PROCEND ;in=23,25;out=28,29;fpin=105,107; - - - - -;---------------------------------------------------------------------------- -; -; Registers to hold 64-bit values to manipulate. The "L" part -; of the register corresponds to the upper 32-bits, while the "R" -; part corresponds to the lower 32-bits -; -; Note, that when using b6 and b7, the code must save these before -; using them because they are callee save registers -; -; -; Floating point registers to use to save values that -; are manipulated. These don't collide with ftemp1-6 and -; are all caller save registers -; -a0 .reg %fr22 -a0L .reg %fr22L -a0R .reg %fr22R - -a1 .reg %fr23 -a1L .reg %fr23L -a1R .reg %fr23R - -a2 .reg %fr24 -a2L .reg %fr24L -a2R .reg %fr24R - -a3 .reg %fr25 -a3L .reg %fr25L -a3R .reg %fr25R - -a4 .reg %fr26 -a4L .reg %fr26L -a4R .reg %fr26R - -a5 .reg %fr27 -a5L .reg %fr27L -a5R .reg %fr27R - -a6 .reg %fr28 -a6L .reg %fr28L -a6R .reg %fr28R - -a7 .reg %fr29 -a7L .reg %fr29L -a7R .reg %fr29R - -b0 .reg %fr30 -b0L .reg %fr30L -b0R .reg %fr30R - -b1 .reg %fr31 -b1L .reg %fr31L -b1R .reg %fr31R - -; -; Temporary floating point variables, these are all caller save -; registers -; -ftemp1 .reg %fr4 -ftemp2 .reg %fr5 -ftemp3 .reg %fr6 -ftemp4 .reg %fr7 - -; -; The B set of registers when used. -; - -b2 .reg %fr8 -b2L .reg %fr8L -b2R .reg %fr8R - -b3 .reg %fr9 -b3L .reg %fr9L -b3R .reg %fr9R - -b4 .reg %fr10 -b4L .reg %fr10L -b4R .reg %fr10R - -b5 .reg %fr11 -b5L .reg %fr11L -b5R .reg %fr11R - -b6 .reg %fr12 -b6L .reg %fr12L -b6R .reg %fr12R - -b7 .reg %fr13 -b7L .reg %fr13L -b7R .reg %fr13R - -c1 .reg %r21 ; only reg -temp1 .reg %r20 ; only reg -temp2 .reg %r19 ; only reg -temp3 .reg %r31 ; only reg - -m1 .reg %r28 -c2 .reg %r23 -high_one .reg %r1 -ht .reg %r6 -lt .reg %r5 -m .reg %r4 -c3 .reg %r3 - -SQR_ADD_C .macro A0L,A0R,C1,C2,C3 - XMPYU A0L,A0R,ftemp1 ; m - FSTD ftemp1,-24(%sp) ; store m - - XMPYU A0R,A0R,ftemp2 ; lt - FSTD ftemp2,-16(%sp) ; store lt - - XMPYU A0L,A0L,ftemp3 ; ht - FSTD ftemp3,-8(%sp) ; store ht - - LDD -24(%sp),m ; load m - AND m,high_mask,temp2 ; m & Mask - DEPD,Z m,30,31,temp3 ; m << 32+1 - LDD -16(%sp),lt ; lt - - LDD -8(%sp),ht ; ht - EXTRD,U temp2,32,33,temp1 ; temp1 = m&Mask >> 32-1 - ADD temp3,lt,lt ; lt = lt+m - ADD,L ht,temp1,ht ; ht += temp1 - ADD,DC ht,%r0,ht ; ht++ - - ADD C1,lt,C1 ; c1=c1+lt - ADD,DC ht,%r0,ht ; ht++ - - ADD C2,ht,C2 ; c2=c2+ht - ADD,DC C3,%r0,C3 ; c3++ -.endm - -SQR_ADD_C2 .macro A0L,A0R,A1L,A1R,C1,C2,C3 - XMPYU A0L,A1R,ftemp1 ; m1 = bl*ht - FSTD ftemp1,-16(%sp) ; - XMPYU A0R,A1L,ftemp2 ; m = bh*lt - FSTD ftemp2,-8(%sp) ; - XMPYU A0R,A1R,ftemp3 ; lt = bl*lt - FSTD ftemp3,-32(%sp) - XMPYU A0L,A1L,ftemp4 ; ht = bh*ht - FSTD ftemp4,-24(%sp) ; - - LDD -8(%sp),m ; r21 = m - LDD -16(%sp),m1 ; r19 = m1 - ADD,L m,m1,m ; m+m1 - - DEPD,Z m,31,32,temp3 ; (m+m1<<32) - LDD -24(%sp),ht ; r24 = ht - - CMPCLR,*>>= m,m1,%r0 ; if (m < m1) - ADD,L ht,high_one,ht ; ht+=high_one - - EXTRD,U m,31,32,temp1 ; m >> 32 - LDD -32(%sp),lt ; lt - ADD,L ht,temp1,ht ; ht+= m>>32 - ADD lt,temp3,lt ; lt = lt+m1 - ADD,DC ht,%r0,ht ; ht++ - - ADD ht,ht,ht ; ht=ht+ht; - ADD,DC C3,%r0,C3 ; add in carry (c3++) - - ADD lt,lt,lt ; lt=lt+lt; - ADD,DC ht,%r0,ht ; add in carry (ht++) - - ADD C1,lt,C1 ; c1=c1+lt - ADD,DC,*NUV ht,%r0,ht ; add in carry (ht++) - LDO 1(C3),C3 ; bump c3 if overflow,nullify otherwise - - ADD C2,ht,C2 ; c2 = c2 + ht - ADD,DC C3,%r0,C3 ; add in carry (c3++) -.endm - -; -;void bn_sqr_comba8(BN_ULONG *r, BN_ULONG *a) -; arg0 = r_ptr -; arg1 = a_ptr -; - -bn_sqr_comba8 - .PROC - .CALLINFO FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_sqr_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .ENTRY - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z -1,32,33,high_mask ; Create Mask 0xffffffff80000000L - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - FLDD 32(a_ptr),a4 - FLDD 40(a_ptr),a5 - FLDD 48(a_ptr),a6 - FLDD 56(a_ptr),a7 - - SQR_ADD_C a0L,a0R,c1,c2,c3 - STD c1,0(r_ptr) ; r[0] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1 - STD c2,8(r_ptr) ; r[1] = c2; - COPY %r0,c2 - - SQR_ADD_C a1L,a1R,c3,c1,c2 - SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2 - STD c3,16(r_ptr) ; r[2] = c3; - COPY %r0,c3 - - SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3 - SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3 - STD c1,24(r_ptr) ; r[3] = c1; - COPY %r0,c1 - - SQR_ADD_C a2L,a2R,c2,c3,c1 - SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1 - SQR_ADD_C2 a4L,a4R,a0L,a0R,c2,c3,c1 - STD c2,32(r_ptr) ; r[4] = c2; - COPY %r0,c2 - - SQR_ADD_C2 a5L,a5R,a0L,a0R,c3,c1,c2 - SQR_ADD_C2 a4L,a4R,a1L,a1R,c3,c1,c2 - SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2 - STD c3,40(r_ptr) ; r[5] = c3; - COPY %r0,c3 - - SQR_ADD_C a3L,a3R,c1,c2,c3 - SQR_ADD_C2 a4L,a4R,a2L,a2R,c1,c2,c3 - SQR_ADD_C2 a5L,a5R,a1L,a1R,c1,c2,c3 - SQR_ADD_C2 a6L,a6R,a0L,a0R,c1,c2,c3 - STD c1,48(r_ptr) ; r[6] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a7L,a7R,a0L,a0R,c2,c3,c1 - SQR_ADD_C2 a6L,a6R,a1L,a1R,c2,c3,c1 - SQR_ADD_C2 a5L,a5R,a2L,a2R,c2,c3,c1 - SQR_ADD_C2 a4L,a4R,a3L,a3R,c2,c3,c1 - STD c2,56(r_ptr) ; r[7] = c2; - COPY %r0,c2 - - SQR_ADD_C a4L,a4R,c3,c1,c2 - SQR_ADD_C2 a5L,a5R,a3L,a3R,c3,c1,c2 - SQR_ADD_C2 a6L,a6R,a2L,a2R,c3,c1,c2 - SQR_ADD_C2 a7L,a7R,a1L,a1R,c3,c1,c2 - STD c3,64(r_ptr) ; r[8] = c3; - COPY %r0,c3 - - SQR_ADD_C2 a7L,a7R,a2L,a2R,c1,c2,c3 - SQR_ADD_C2 a6L,a6R,a3L,a3R,c1,c2,c3 - SQR_ADD_C2 a5L,a5R,a4L,a4R,c1,c2,c3 - STD c1,72(r_ptr) ; r[9] = c1; - COPY %r0,c1 - - SQR_ADD_C a5L,a5R,c2,c3,c1 - SQR_ADD_C2 a6L,a6R,a4L,a4R,c2,c3,c1 - SQR_ADD_C2 a7L,a7R,a3L,a3R,c2,c3,c1 - STD c2,80(r_ptr) ; r[10] = c2; - COPY %r0,c2 - - SQR_ADD_C2 a7L,a7R,a4L,a4R,c3,c1,c2 - SQR_ADD_C2 a6L,a6R,a5L,a5R,c3,c1,c2 - STD c3,88(r_ptr) ; r[11] = c3; - COPY %r0,c3 - - SQR_ADD_C a6L,a6R,c1,c2,c3 - SQR_ADD_C2 a7L,a7R,a5L,a5R,c1,c2,c3 - STD c1,96(r_ptr) ; r[12] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a7L,a7R,a6L,a6R,c2,c3,c1 - STD c2,104(r_ptr) ; r[13] = c2; - COPY %r0,c2 - - SQR_ADD_C a7L,a7R,c3,c1,c2 - STD c3, 112(r_ptr) ; r[14] = c3 - STD c1, 120(r_ptr) ; r[15] = c1 - - .EXIT - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - -;----------------------------------------------------------------------------- -; -;void bn_sqr_comba4(BN_ULONG *r, BN_ULONG *a) -; arg0 = r_ptr -; arg1 = a_ptr -; - -bn_sqr_comba4 - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_sqr_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z -1,32,33,high_mask ; Create Mask 0xffffffff80000000L - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - FLDD 32(a_ptr),a4 - FLDD 40(a_ptr),a5 - FLDD 48(a_ptr),a6 - FLDD 56(a_ptr),a7 - - SQR_ADD_C a0L,a0R,c1,c2,c3 - - STD c1,0(r_ptr) ; r[0] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1 - - STD c2,8(r_ptr) ; r[1] = c2; - COPY %r0,c2 - - SQR_ADD_C a1L,a1R,c3,c1,c2 - SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2 - - STD c3,16(r_ptr) ; r[2] = c3; - COPY %r0,c3 - - SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3 - SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3 - - STD c1,24(r_ptr) ; r[3] = c1; - COPY %r0,c1 - - SQR_ADD_C a2L,a2R,c2,c3,c1 - SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1 - - STD c2,32(r_ptr) ; r[4] = c2; - COPY %r0,c2 - - SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2 - STD c3,40(r_ptr) ; r[5] = c3; - COPY %r0,c3 - - SQR_ADD_C a3L,a3R,c1,c2,c3 - STD c1,48(r_ptr) ; r[6] = c1; - STD c2,56(r_ptr) ; r[7] = c2; - - .EXIT - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - - -;--------------------------------------------------------------------------- - -MUL_ADD_C .macro A0L,A0R,B0L,B0R,C1,C2,C3 - XMPYU A0L,B0R,ftemp1 ; m1 = bl*ht - FSTD ftemp1,-16(%sp) ; - XMPYU A0R,B0L,ftemp2 ; m = bh*lt - FSTD ftemp2,-8(%sp) ; - XMPYU A0R,B0R,ftemp3 ; lt = bl*lt - FSTD ftemp3,-32(%sp) - XMPYU A0L,B0L,ftemp4 ; ht = bh*ht - FSTD ftemp4,-24(%sp) ; - - LDD -8(%sp),m ; r21 = m - LDD -16(%sp),m1 ; r19 = m1 - ADD,L m,m1,m ; m+m1 - - DEPD,Z m,31,32,temp3 ; (m+m1<<32) - LDD -24(%sp),ht ; r24 = ht - - CMPCLR,*>>= m,m1,%r0 ; if (m < m1) - ADD,L ht,high_one,ht ; ht+=high_one - - EXTRD,U m,31,32,temp1 ; m >> 32 - LDD -32(%sp),lt ; lt - ADD,L ht,temp1,ht ; ht+= m>>32 - ADD lt,temp3,lt ; lt = lt+m1 - ADD,DC ht,%r0,ht ; ht++ - - ADD C1,lt,C1 ; c1=c1+lt - ADD,DC ht,%r0,ht ; bump c3 if overflow,nullify otherwise - - ADD C2,ht,C2 ; c2 = c2 + ht - ADD,DC C3,%r0,C3 ; add in carry (c3++) -.endm - - -; -;void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b) -; arg0 = r_ptr -; arg1 = a_ptr -; arg2 = b_ptr -; - -bn_mul_comba8 - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_mul_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - FSTD %fr12,32(%sp) ; save r6 - FSTD %fr13,40(%sp) ; save r7 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - FLDD 32(a_ptr),a4 - FLDD 40(a_ptr),a5 - FLDD 48(a_ptr),a6 - FLDD 56(a_ptr),a7 - - FLDD 0(b_ptr),b0 - FLDD 8(b_ptr),b1 - FLDD 16(b_ptr),b2 - FLDD 24(b_ptr),b3 - FLDD 32(b_ptr),b4 - FLDD 40(b_ptr),b5 - FLDD 48(b_ptr),b6 - FLDD 56(b_ptr),b7 - - MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3 - STD c1,0(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1 - STD c2,8(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2 - STD c3,16(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3 - MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3 - STD c1,24(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a4L,a4R,b0L,b0R,c2,c3,c1 - MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1 - MUL_ADD_C a0L,a0R,b4L,b4R,c2,c3,c1 - STD c2,32(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a0L,a0R,b5L,b5R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b4L,b4R,c3,c1,c2 - MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2 - MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2 - MUL_ADD_C a4L,a4R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a5L,a5R,b0L,b0R,c3,c1,c2 - STD c3,40(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a6L,a6R,b0L,b0R,c1,c2,c3 - MUL_ADD_C a5L,a5R,b1L,b1R,c1,c2,c3 - MUL_ADD_C a4L,a4R,b2L,b2R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a2L,a2R,b4L,b4R,c1,c2,c3 - MUL_ADD_C a1L,a1R,b5L,b5R,c1,c2,c3 - MUL_ADD_C a0L,a0R,b6L,b6R,c1,c2,c3 - STD c1,48(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a0L,a0R,b7L,b7R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b6L,b6R,c2,c3,c1 - MUL_ADD_C a2L,a2R,b5L,b5R,c2,c3,c1 - MUL_ADD_C a3L,a3R,b4L,b4R,c2,c3,c1 - MUL_ADD_C a4L,a4R,b3L,b3R,c2,c3,c1 - MUL_ADD_C a5L,a5R,b2L,b2R,c2,c3,c1 - MUL_ADD_C a6L,a6R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a7L,a7R,b0L,b0R,c2,c3,c1 - STD c2,56(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a7L,a7R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a6L,a6R,b2L,b2R,c3,c1,c2 - MUL_ADD_C a5L,a5R,b3L,b3R,c3,c1,c2 - MUL_ADD_C a4L,a4R,b4L,b4R,c3,c1,c2 - MUL_ADD_C a3L,a3R,b5L,b5R,c3,c1,c2 - MUL_ADD_C a2L,a2R,b6L,b6R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b7L,b7R,c3,c1,c2 - STD c3,64(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a2L,a2R,b7L,b7R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b6L,b6R,c1,c2,c3 - MUL_ADD_C a4L,a4R,b5L,b5R,c1,c2,c3 - MUL_ADD_C a5L,a5R,b4L,b4R,c1,c2,c3 - MUL_ADD_C a6L,a6R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a7L,a7R,b2L,b2R,c1,c2,c3 - STD c1,72(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a7L,a7R,b3L,b3R,c2,c3,c1 - MUL_ADD_C a6L,a6R,b4L,b4R,c2,c3,c1 - MUL_ADD_C a5L,a5R,b5L,b5R,c2,c3,c1 - MUL_ADD_C a4L,a4R,b6L,b6R,c2,c3,c1 - MUL_ADD_C a3L,a3R,b7L,b7R,c2,c3,c1 - STD c2,80(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a4L,a4R,b7L,b7R,c3,c1,c2 - MUL_ADD_C a5L,a5R,b6L,b6R,c3,c1,c2 - MUL_ADD_C a6L,a6R,b5L,b5R,c3,c1,c2 - MUL_ADD_C a7L,a7R,b4L,b4R,c3,c1,c2 - STD c3,88(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a7L,a7R,b5L,b5R,c1,c2,c3 - MUL_ADD_C a6L,a6R,b6L,b6R,c1,c2,c3 - MUL_ADD_C a5L,a5R,b7L,b7R,c1,c2,c3 - STD c1,96(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a6L,a6R,b7L,b7R,c2,c3,c1 - MUL_ADD_C a7L,a7R,b6L,b6R,c2,c3,c1 - STD c2,104(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a7L,a7R,b7L,b7R,c3,c1,c2 - STD c3,112(r_ptr) - STD c1,120(r_ptr) - - .EXIT - FLDD -88(%sp),%fr13 - FLDD -96(%sp),%fr12 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - -;----------------------------------------------------------------------------- -; -;void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b) -; arg0 = r_ptr -; arg1 = a_ptr -; arg2 = b_ptr -; - -bn_mul_comba4 - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_mul_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - FSTD %fr12,32(%sp) ; save r6 - FSTD %fr13,40(%sp) ; save r7 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - - FLDD 0(b_ptr),b0 - FLDD 8(b_ptr),b1 - FLDD 16(b_ptr),b2 - FLDD 24(b_ptr),b3 - - MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3 - STD c1,0(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1 - STD c2,8(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2 - STD c3,16(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3 - MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3 - STD c1,24(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1 - STD c2,32(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2 - MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2 - STD c3,40(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3 - STD c1,48(r_ptr) - STD c2,56(r_ptr) - - .EXIT - FLDD -88(%sp),%fr13 - FLDD -96(%sp),%fr12 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - - -;--- not PIC .SPACE $TEXT$ -;--- not PIC .SUBSPA $CODE$ -;--- not PIC .SPACE $PRIVATE$,SORT=16 -;--- not PIC .IMPORT $global$,DATA -;--- not PIC .SPACE $TEXT$ -;--- not PIC .SUBSPA $CODE$ -;--- not PIC .SUBSPA $LIT$,ACCESS=0x2c -;--- not PIC C$7 -;--- not PIC .ALIGN 8 -;--- not PIC .STRINGZ "Division would overflow (%d)\n" - .END diff --git a/src/lib/libcrypto/bn/asm/pa-risc2W.s b/src/lib/libcrypto/bn/asm/pa-risc2W.s deleted file mode 100644 index a99545754d..0000000000 --- a/src/lib/libcrypto/bn/asm/pa-risc2W.s +++ /dev/null @@ -1,1605 +0,0 @@ -; -; PA-RISC 64-bit implementation of bn_asm code -; -; This code is approximately 2x faster than the C version -; for RSA/DSA. -; -; See http://devresource.hp.com/ for more details on the PA-RISC -; architecture. Also see the book "PA-RISC 2.0 Architecture" -; by Gerry Kane for information on the instruction set architecture. -; -; Code written by Chris Ruemmler (with some help from the HP C -; compiler). -; -; The code compiles with HP's assembler -; - - .level 2.0W - .space $TEXT$ - .subspa $CODE$,QUAD=0,ALIGN=8,ACCESS=0x2c,CODE_ONLY - -; -; Global Register definitions used for the routines. -; -; Some information about HP's runtime architecture for 64-bits. -; -; "Caller save" means the calling function must save the register -; if it wants the register to be preserved. -; "Callee save" means if a function uses the register, it must save -; the value before using it. -; -; For the floating point registers -; -; "caller save" registers: fr4-fr11, fr22-fr31 -; "callee save" registers: fr12-fr21 -; "special" registers: fr0-fr3 (status and exception registers) -; -; For the integer registers -; value zero : r0 -; "caller save" registers: r1,r19-r26 -; "callee save" registers: r3-r18 -; return register : r2 (rp) -; return values ; r28 (ret0,ret1) -; Stack pointer ; r30 (sp) -; global data pointer ; r27 (dp) -; argument pointer ; r29 (ap) -; millicode return ptr ; r31 (also a caller save register) - - -; -; Arguments to the routines -; -r_ptr .reg %r26 -a_ptr .reg %r25 -b_ptr .reg %r24 -num .reg %r24 -w .reg %r23 -n .reg %r23 - - -; -; Globals used in some routines -; - -top_overflow .reg %r29 -high_mask .reg %r22 ; value 0xffffffff80000000L - - -;------------------------------------------------------------------------------ -; -; bn_mul_add_words -; -;BN_ULONG bn_mul_add_words(BN_ULONG *r_ptr, BN_ULONG *a_ptr, -; int num, BN_ULONG w) -; -; arg0 = r_ptr -; arg1 = a_ptr -; arg2 = num -; arg3 = w -; -; Local register definitions -; - -fm1 .reg %fr22 -fm .reg %fr23 -ht_temp .reg %fr24 -ht_temp_1 .reg %fr25 -lt_temp .reg %fr26 -lt_temp_1 .reg %fr27 -fm1_1 .reg %fr28 -fm_1 .reg %fr29 - -fw_h .reg %fr7L -fw_l .reg %fr7R -fw .reg %fr7 - -fht_0 .reg %fr8L -flt_0 .reg %fr8R -t_float_0 .reg %fr8 - -fht_1 .reg %fr9L -flt_1 .reg %fr9R -t_float_1 .reg %fr9 - -tmp_0 .reg %r31 -tmp_1 .reg %r21 -m_0 .reg %r20 -m_1 .reg %r19 -ht_0 .reg %r1 -ht_1 .reg %r3 -lt_0 .reg %r4 -lt_1 .reg %r5 -m1_0 .reg %r6 -m1_1 .reg %r7 -rp_val .reg %r8 -rp_val_1 .reg %r9 - -bn_mul_add_words - .export bn_mul_add_words,entry,NO_RELOCATION,LONG_RETURN - .proc - .callinfo frame=128 - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - NOP ; Needed to make the loop 16-byte aligned - NOP ; Needed to make the loop 16-byte aligned - - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - STD %r7,32(%sp) ; save r7 - STD %r8,40(%sp) ; save r8 - - STD %r9,48(%sp) ; save r9 - COPY %r0,%ret0 ; return 0 by default - DEPDI,Z 1,31,1,top_overflow ; top_overflow = 1 << 32 - STD w,56(%sp) ; store w on stack - - CMPIB,>= 0,num,bn_mul_add_words_exit ; if (num <= 0) then exit - LDO 128(%sp),%sp ; bump stack - - ; - ; The loop is unrolled twice, so if there is only 1 number - ; then go straight to the cleanup code. - ; - CMPIB,= 1,num,bn_mul_add_words_single_top - FLDD -72(%sp),fw ; load up w into fp register fw (fw_h/fw_l) - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; - ; PA-RISC 2.0 chips have two fully pipelined multipliers, thus - ; two 32-bit mutiplies can be issued per cycle. - ; -bn_mul_add_words_unroll2 - - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - FLDD 8(a_ptr),t_float_1 ; load up 64-bit value (fr8L) ht(L)/lt(R) - LDD 0(r_ptr),rp_val ; rp[0] - LDD 8(r_ptr),rp_val_1 ; rp[1] - - XMPYU fht_0,fw_l,fm1 ; m1[0] = fht_0*fw_l - XMPYU fht_1,fw_l,fm1_1 ; m1[1] = fht_1*fw_l - FSTD fm1,-16(%sp) ; -16(sp) = m1[0] - FSTD fm1_1,-48(%sp) ; -48(sp) = m1[1] - - XMPYU flt_0,fw_h,fm ; m[0] = flt_0*fw_h - XMPYU flt_1,fw_h,fm_1 ; m[1] = flt_1*fw_h - FSTD fm,-8(%sp) ; -8(sp) = m[0] - FSTD fm_1,-40(%sp) ; -40(sp) = m[1] - - XMPYU fht_0,fw_h,ht_temp ; ht_temp = fht_0*fw_h - XMPYU fht_1,fw_h,ht_temp_1 ; ht_temp_1 = fht_1*fw_h - FSTD ht_temp,-24(%sp) ; -24(sp) = ht_temp - FSTD ht_temp_1,-56(%sp) ; -56(sp) = ht_temp_1 - - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - XMPYU flt_1,fw_l,lt_temp_1 ; lt_temp = lt*fw_l - FSTD lt_temp,-32(%sp) ; -32(sp) = lt_temp - FSTD lt_temp_1,-64(%sp) ; -64(sp) = lt_temp_1 - - LDD -8(%sp),m_0 ; m[0] - LDD -40(%sp),m_1 ; m[1] - LDD -16(%sp),m1_0 ; m1[0] - LDD -48(%sp),m1_1 ; m1[1] - - LDD -24(%sp),ht_0 ; ht[0] - LDD -56(%sp),ht_1 ; ht[1] - ADD,L m1_0,m_0,tmp_0 ; tmp_0 = m[0] + m1[0]; - ADD,L m1_1,m_1,tmp_1 ; tmp_1 = m[1] + m1[1]; - - LDD -32(%sp),lt_0 - LDD -64(%sp),lt_1 - CMPCLR,*>>= tmp_0,m1_0, %r0 ; if (m[0] < m1[0]) - ADD,L ht_0,top_overflow,ht_0 ; ht[0] += (1<<32) - - CMPCLR,*>>= tmp_1,m1_1,%r0 ; if (m[1] < m1[1]) - ADD,L ht_1,top_overflow,ht_1 ; ht[1] += (1<<32) - EXTRD,U tmp_0,31,32,m_0 ; m[0]>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1[0] = m[0]<<32 - - EXTRD,U tmp_1,31,32,m_1 ; m[1]>>32 - DEPD,Z tmp_1,31,32,m1_1 ; m1[1] = m[1]<<32 - ADD,L ht_0,m_0,ht_0 ; ht[0]+= (m[0]>>32) - ADD,L ht_1,m_1,ht_1 ; ht[1]+= (m[1]>>32) - - ADD lt_0,m1_0,lt_0 ; lt[0] = lt[0]+m1[0]; - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - ADD lt_1,m1_1,lt_1 ; lt[1] = lt[1]+m1[1]; - ADD,DC ht_1,%r0,ht_1 ; ht[1]++ - - ADD %ret0,lt_0,lt_0 ; lt[0] = lt[0] + c; - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - ADD lt_0,rp_val,lt_0 ; lt[0] = lt[0]+rp[0] - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - - LDO -2(num),num ; num = num - 2; - ADD ht_0,lt_1,lt_1 ; lt[1] = lt[1] + ht_0 (c); - ADD,DC ht_1,%r0,ht_1 ; ht[1]++ - STD lt_0,0(r_ptr) ; rp[0] = lt[0] - - ADD lt_1,rp_val_1,lt_1 ; lt[1] = lt[1]+rp[1] - ADD,DC ht_1,%r0,%ret0 ; ht[1]++ - LDO 16(a_ptr),a_ptr ; a_ptr += 2 - - STD lt_1,8(r_ptr) ; rp[1] = lt[1] - CMPIB,<= 2,num,bn_mul_add_words_unroll2 ; go again if more to do - LDO 16(r_ptr),r_ptr ; r_ptr += 2 - - CMPIB,=,N 0,num,bn_mul_add_words_exit ; are we done, or cleanup last one - - ; - ; Top of loop aligned on 64-byte boundary - ; -bn_mul_add_words_single_top - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - LDD 0(r_ptr),rp_val ; rp[0] - LDO 8(a_ptr),a_ptr ; a_ptr++ - XMPYU fht_0,fw_l,fm1 ; m1 = ht*fw_l - FSTD fm1,-16(%sp) ; -16(sp) = m1 - XMPYU flt_0,fw_h,fm ; m = lt*fw_h - FSTD fm,-8(%sp) ; -8(sp) = m - XMPYU fht_0,fw_h,ht_temp ; ht_temp = ht*fw_h - FSTD ht_temp,-24(%sp) ; -24(sp) = ht - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - FSTD lt_temp,-32(%sp) ; -32(sp) = lt - - LDD -8(%sp),m_0 - LDD -16(%sp),m1_0 ; m1 = temp1 - ADD,L m_0,m1_0,tmp_0 ; tmp_0 = m + m1; - LDD -24(%sp),ht_0 - LDD -32(%sp),lt_0 - - CMPCLR,*>>= tmp_0,m1_0,%r0 ; if (m < m1) - ADD,L ht_0,top_overflow,ht_0 ; ht += (1<<32) - - EXTRD,U tmp_0,31,32,m_0 ; m>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1 = m<<32 - - ADD,L ht_0,m_0,ht_0 ; ht+= (m>>32) - ADD lt_0,m1_0,tmp_0 ; tmp_0 = lt+m1; - ADD,DC ht_0,%r0,ht_0 ; ht++ - ADD %ret0,tmp_0,lt_0 ; lt = lt + c; - ADD,DC ht_0,%r0,ht_0 ; ht++ - ADD lt_0,rp_val,lt_0 ; lt = lt+rp[0] - ADD,DC ht_0,%r0,%ret0 ; ht++ - STD lt_0,0(r_ptr) ; rp[0] = lt - -bn_mul_add_words_exit - .EXIT - LDD -80(%sp),%r9 ; restore r9 - LDD -88(%sp),%r8 ; restore r8 - LDD -96(%sp),%r7 ; restore r7 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 ; restore r3 - .PROCEND ;in=23,24,25,26,29;out=28; - -;---------------------------------------------------------------------------- -; -;BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w) -; -; arg0 = rp -; arg1 = ap -; arg2 = num -; arg3 = w - -bn_mul_words - .proc - .callinfo frame=128 - .entry - .EXPORT bn_mul_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - - STD %r7,32(%sp) ; save r7 - COPY %r0,%ret0 ; return 0 by default - DEPDI,Z 1,31,1,top_overflow ; top_overflow = 1 << 32 - STD w,56(%sp) ; w on stack - - CMPIB,>= 0,num,bn_mul_words_exit - LDO 128(%sp),%sp ; bump stack - - ; - ; See if only 1 word to do, thus just do cleanup - ; - CMPIB,= 1,num,bn_mul_words_single_top - FLDD -72(%sp),fw ; load up w into fp register fw (fw_h/fw_l) - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; - ; PA-RISC 2.0 chips have two fully pipelined multipliers, thus - ; two 32-bit mutiplies can be issued per cycle. - ; -bn_mul_words_unroll2 - - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - FLDD 8(a_ptr),t_float_1 ; load up 64-bit value (fr8L) ht(L)/lt(R) - XMPYU fht_0,fw_l,fm1 ; m1[0] = fht_0*fw_l - XMPYU fht_1,fw_l,fm1_1 ; m1[1] = ht*fw_l - - FSTD fm1,-16(%sp) ; -16(sp) = m1 - FSTD fm1_1,-48(%sp) ; -48(sp) = m1 - XMPYU flt_0,fw_h,fm ; m = lt*fw_h - XMPYU flt_1,fw_h,fm_1 ; m = lt*fw_h - - FSTD fm,-8(%sp) ; -8(sp) = m - FSTD fm_1,-40(%sp) ; -40(sp) = m - XMPYU fht_0,fw_h,ht_temp ; ht_temp = fht_0*fw_h - XMPYU fht_1,fw_h,ht_temp_1 ; ht_temp = ht*fw_h - - FSTD ht_temp,-24(%sp) ; -24(sp) = ht - FSTD ht_temp_1,-56(%sp) ; -56(sp) = ht - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - XMPYU flt_1,fw_l,lt_temp_1 ; lt_temp = lt*fw_l - - FSTD lt_temp,-32(%sp) ; -32(sp) = lt - FSTD lt_temp_1,-64(%sp) ; -64(sp) = lt - LDD -8(%sp),m_0 - LDD -40(%sp),m_1 - - LDD -16(%sp),m1_0 - LDD -48(%sp),m1_1 - LDD -24(%sp),ht_0 - LDD -56(%sp),ht_1 - - ADD,L m1_0,m_0,tmp_0 ; tmp_0 = m + m1; - ADD,L m1_1,m_1,tmp_1 ; tmp_1 = m + m1; - LDD -32(%sp),lt_0 - LDD -64(%sp),lt_1 - - CMPCLR,*>>= tmp_0,m1_0, %r0 ; if (m < m1) - ADD,L ht_0,top_overflow,ht_0 ; ht += (1<<32) - CMPCLR,*>>= tmp_1,m1_1,%r0 ; if (m < m1) - ADD,L ht_1,top_overflow,ht_1 ; ht += (1<<32) - - EXTRD,U tmp_0,31,32,m_0 ; m>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1 = m<<32 - EXTRD,U tmp_1,31,32,m_1 ; m>>32 - DEPD,Z tmp_1,31,32,m1_1 ; m1 = m<<32 - - ADD,L ht_0,m_0,ht_0 ; ht+= (m>>32) - ADD,L ht_1,m_1,ht_1 ; ht+= (m>>32) - ADD lt_0,m1_0,lt_0 ; lt = lt+m1; - ADD,DC ht_0,%r0,ht_0 ; ht++ - - ADD lt_1,m1_1,lt_1 ; lt = lt+m1; - ADD,DC ht_1,%r0,ht_1 ; ht++ - ADD %ret0,lt_0,lt_0 ; lt = lt + c (ret0); - ADD,DC ht_0,%r0,ht_0 ; ht++ - - ADD ht_0,lt_1,lt_1 ; lt = lt + c (ht_0) - ADD,DC ht_1,%r0,ht_1 ; ht++ - STD lt_0,0(r_ptr) ; rp[0] = lt - STD lt_1,8(r_ptr) ; rp[1] = lt - - COPY ht_1,%ret0 ; carry = ht - LDO -2(num),num ; num = num - 2; - LDO 16(a_ptr),a_ptr ; ap += 2 - CMPIB,<= 2,num,bn_mul_words_unroll2 - LDO 16(r_ptr),r_ptr ; rp++ - - CMPIB,=,N 0,num,bn_mul_words_exit ; are we done? - - ; - ; Top of loop aligned on 64-byte boundary - ; -bn_mul_words_single_top - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - - XMPYU fht_0,fw_l,fm1 ; m1 = ht*fw_l - FSTD fm1,-16(%sp) ; -16(sp) = m1 - XMPYU flt_0,fw_h,fm ; m = lt*fw_h - FSTD fm,-8(%sp) ; -8(sp) = m - XMPYU fht_0,fw_h,ht_temp ; ht_temp = ht*fw_h - FSTD ht_temp,-24(%sp) ; -24(sp) = ht - XMPYU flt_0,fw_l,lt_temp ; lt_temp = lt*fw_l - FSTD lt_temp,-32(%sp) ; -32(sp) = lt - - LDD -8(%sp),m_0 - LDD -16(%sp),m1_0 - ADD,L m_0,m1_0,tmp_0 ; tmp_0 = m + m1; - LDD -24(%sp),ht_0 - LDD -32(%sp),lt_0 - - CMPCLR,*>>= tmp_0,m1_0,%r0 ; if (m < m1) - ADD,L ht_0,top_overflow,ht_0 ; ht += (1<<32) - - EXTRD,U tmp_0,31,32,m_0 ; m>>32 - DEPD,Z tmp_0,31,32,m1_0 ; m1 = m<<32 - - ADD,L ht_0,m_0,ht_0 ; ht+= (m>>32) - ADD lt_0,m1_0,lt_0 ; lt= lt+m1; - ADD,DC ht_0,%r0,ht_0 ; ht++ - - ADD %ret0,lt_0,lt_0 ; lt = lt + c; - ADD,DC ht_0,%r0,ht_0 ; ht++ - - COPY ht_0,%ret0 ; copy carry - STD lt_0,0(r_ptr) ; rp[0] = lt - -bn_mul_words_exit - .EXIT - LDD -96(%sp),%r7 ; restore r7 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 ; restore r3 - .PROCEND ;in=23,24,25,26,29;out=28; - -;---------------------------------------------------------------------------- -; -;void bn_sqr_words(BN_ULONG *rp, BN_ULONG *ap, int num) -; -; arg0 = rp -; arg1 = ap -; arg2 = num -; - -bn_sqr_words - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_sqr_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - NOP - STD %r5,16(%sp) ; save r5 - - CMPIB,>= 0,num,bn_sqr_words_exit - LDO 128(%sp),%sp ; bump stack - - ; - ; If only 1, the goto straight to cleanup - ; - CMPIB,= 1,num,bn_sqr_words_single_top - DEPDI,Z -1,32,33,high_mask ; Create Mask 0xffffffff80000000L - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; - -bn_sqr_words_unroll2 - FLDD 0(a_ptr),t_float_0 ; a[0] - FLDD 8(a_ptr),t_float_1 ; a[1] - XMPYU fht_0,flt_0,fm ; m[0] - XMPYU fht_1,flt_1,fm_1 ; m[1] - - FSTD fm,-24(%sp) ; store m[0] - FSTD fm_1,-56(%sp) ; store m[1] - XMPYU flt_0,flt_0,lt_temp ; lt[0] - XMPYU flt_1,flt_1,lt_temp_1 ; lt[1] - - FSTD lt_temp,-16(%sp) ; store lt[0] - FSTD lt_temp_1,-48(%sp) ; store lt[1] - XMPYU fht_0,fht_0,ht_temp ; ht[0] - XMPYU fht_1,fht_1,ht_temp_1 ; ht[1] - - FSTD ht_temp,-8(%sp) ; store ht[0] - FSTD ht_temp_1,-40(%sp) ; store ht[1] - LDD -24(%sp),m_0 - LDD -56(%sp),m_1 - - AND m_0,high_mask,tmp_0 ; m[0] & Mask - AND m_1,high_mask,tmp_1 ; m[1] & Mask - DEPD,Z m_0,30,31,m_0 ; m[0] << 32+1 - DEPD,Z m_1,30,31,m_1 ; m[1] << 32+1 - - LDD -16(%sp),lt_0 - LDD -48(%sp),lt_1 - EXTRD,U tmp_0,32,33,tmp_0 ; tmp_0 = m[0]&Mask >> 32-1 - EXTRD,U tmp_1,32,33,tmp_1 ; tmp_1 = m[1]&Mask >> 32-1 - - LDD -8(%sp),ht_0 - LDD -40(%sp),ht_1 - ADD,L ht_0,tmp_0,ht_0 ; ht[0] += tmp_0 - ADD,L ht_1,tmp_1,ht_1 ; ht[1] += tmp_1 - - ADD lt_0,m_0,lt_0 ; lt = lt+m - ADD,DC ht_0,%r0,ht_0 ; ht[0]++ - STD lt_0,0(r_ptr) ; rp[0] = lt[0] - STD ht_0,8(r_ptr) ; rp[1] = ht[1] - - ADD lt_1,m_1,lt_1 ; lt = lt+m - ADD,DC ht_1,%r0,ht_1 ; ht[1]++ - STD lt_1,16(r_ptr) ; rp[2] = lt[1] - STD ht_1,24(r_ptr) ; rp[3] = ht[1] - - LDO -2(num),num ; num = num - 2; - LDO 16(a_ptr),a_ptr ; ap += 2 - CMPIB,<= 2,num,bn_sqr_words_unroll2 - LDO 32(r_ptr),r_ptr ; rp += 4 - - CMPIB,=,N 0,num,bn_sqr_words_exit ; are we done? - - ; - ; Top of loop aligned on 64-byte boundary - ; -bn_sqr_words_single_top - FLDD 0(a_ptr),t_float_0 ; load up 64-bit value (fr8L) ht(L)/lt(R) - - XMPYU fht_0,flt_0,fm ; m - FSTD fm,-24(%sp) ; store m - - XMPYU flt_0,flt_0,lt_temp ; lt - FSTD lt_temp,-16(%sp) ; store lt - - XMPYU fht_0,fht_0,ht_temp ; ht - FSTD ht_temp,-8(%sp) ; store ht - - LDD -24(%sp),m_0 ; load m - AND m_0,high_mask,tmp_0 ; m & Mask - DEPD,Z m_0,30,31,m_0 ; m << 32+1 - LDD -16(%sp),lt_0 ; lt - - LDD -8(%sp),ht_0 ; ht - EXTRD,U tmp_0,32,33,tmp_0 ; tmp_0 = m&Mask >> 32-1 - ADD m_0,lt_0,lt_0 ; lt = lt+m - ADD,L ht_0,tmp_0,ht_0 ; ht += tmp_0 - ADD,DC ht_0,%r0,ht_0 ; ht++ - - STD lt_0,0(r_ptr) ; rp[0] = lt - STD ht_0,8(r_ptr) ; rp[1] = ht - -bn_sqr_words_exit - .EXIT - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - .PROCEND ;in=23,24,25,26,29;out=28; - - -;---------------------------------------------------------------------------- -; -;BN_ULONG bn_add_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n) -; -; arg0 = rp -; arg1 = ap -; arg2 = bp -; arg3 = n - -t .reg %r22 -b .reg %r21 -l .reg %r20 - -bn_add_words - .proc - .entry - .callinfo - .EXPORT bn_add_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .align 64 - - CMPIB,>= 0,n,bn_add_words_exit - COPY %r0,%ret0 ; return 0 by default - - ; - ; If 2 or more numbers do the loop - ; - CMPIB,= 1,n,bn_add_words_single_top - NOP - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; -bn_add_words_unroll2 - LDD 0(a_ptr),t - LDD 0(b_ptr),b - ADD t,%ret0,t ; t = t+c; - ADD,DC %r0,%r0,%ret0 ; set c to carry - ADD t,b,l ; l = t + b[0] - ADD,DC %ret0,%r0,%ret0 ; c+= carry - STD l,0(r_ptr) - - LDD 8(a_ptr),t - LDD 8(b_ptr),b - ADD t,%ret0,t ; t = t+c; - ADD,DC %r0,%r0,%ret0 ; set c to carry - ADD t,b,l ; l = t + b[0] - ADD,DC %ret0,%r0,%ret0 ; c+= carry - STD l,8(r_ptr) - - LDO -2(n),n - LDO 16(a_ptr),a_ptr - LDO 16(b_ptr),b_ptr - - CMPIB,<= 2,n,bn_add_words_unroll2 - LDO 16(r_ptr),r_ptr - - CMPIB,=,N 0,n,bn_add_words_exit ; are we done? - -bn_add_words_single_top - LDD 0(a_ptr),t - LDD 0(b_ptr),b - - ADD t,%ret0,t ; t = t+c; - ADD,DC %r0,%r0,%ret0 ; set c to carry (could use CMPCLR??) - ADD t,b,l ; l = t + b[0] - ADD,DC %ret0,%r0,%ret0 ; c+= carry - STD l,0(r_ptr) - -bn_add_words_exit - .EXIT - BVE (%rp) - NOP - .PROCEND ;in=23,24,25,26,29;out=28; - -;---------------------------------------------------------------------------- -; -;BN_ULONG bn_sub_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n) -; -; arg0 = rp -; arg1 = ap -; arg2 = bp -; arg3 = n - -t1 .reg %r22 -t2 .reg %r21 -sub_tmp1 .reg %r20 -sub_tmp2 .reg %r19 - - -bn_sub_words - .proc - .callinfo - .EXPORT bn_sub_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - CMPIB,>= 0,n,bn_sub_words_exit - COPY %r0,%ret0 ; return 0 by default - - ; - ; If 2 or more numbers do the loop - ; - CMPIB,= 1,n,bn_sub_words_single_top - NOP - - ; - ; This loop is unrolled 2 times (64-byte aligned as well) - ; -bn_sub_words_unroll2 - LDD 0(a_ptr),t1 - LDD 0(b_ptr),t2 - SUB t1,t2,sub_tmp1 ; t3 = t1-t2; - SUB sub_tmp1,%ret0,sub_tmp1 ; t3 = t3- c; - - CMPCLR,*>> t1,t2,sub_tmp2 ; clear if t1 > t2 - LDO 1(%r0),sub_tmp2 - - CMPCLR,*= t1,t2,%r0 - COPY sub_tmp2,%ret0 - STD sub_tmp1,0(r_ptr) - - LDD 8(a_ptr),t1 - LDD 8(b_ptr),t2 - SUB t1,t2,sub_tmp1 ; t3 = t1-t2; - SUB sub_tmp1,%ret0,sub_tmp1 ; t3 = t3- c; - CMPCLR,*>> t1,t2,sub_tmp2 ; clear if t1 > t2 - LDO 1(%r0),sub_tmp2 - - CMPCLR,*= t1,t2,%r0 - COPY sub_tmp2,%ret0 - STD sub_tmp1,8(r_ptr) - - LDO -2(n),n - LDO 16(a_ptr),a_ptr - LDO 16(b_ptr),b_ptr - - CMPIB,<= 2,n,bn_sub_words_unroll2 - LDO 16(r_ptr),r_ptr - - CMPIB,=,N 0,n,bn_sub_words_exit ; are we done? - -bn_sub_words_single_top - LDD 0(a_ptr),t1 - LDD 0(b_ptr),t2 - SUB t1,t2,sub_tmp1 ; t3 = t1-t2; - SUB sub_tmp1,%ret0,sub_tmp1 ; t3 = t3- c; - CMPCLR,*>> t1,t2,sub_tmp2 ; clear if t1 > t2 - LDO 1(%r0),sub_tmp2 - - CMPCLR,*= t1,t2,%r0 - COPY sub_tmp2,%ret0 - - STD sub_tmp1,0(r_ptr) - -bn_sub_words_exit - .EXIT - BVE (%rp) - NOP - .PROCEND ;in=23,24,25,26,29;out=28; - -;------------------------------------------------------------------------------ -; -; unsigned long bn_div_words(unsigned long h, unsigned long l, unsigned long d) -; -; arg0 = h -; arg1 = l -; arg2 = d -; -; This is mainly just modified assembly from the compiler, thus the -; lack of variable names. -; -;------------------------------------------------------------------------------ -bn_div_words - .proc - .callinfo CALLER,FRAME=272,ENTRY_GR=%r10,SAVE_RP,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_div_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .IMPORT BN_num_bits_word,CODE,NO_RELOCATION - .IMPORT __iob,DATA - .IMPORT fprintf,CODE,NO_RELOCATION - .IMPORT abort,CODE,NO_RELOCATION - .IMPORT $$div2U,MILLICODE - .entry - STD %r2,-16(%r30) - STD,MA %r3,352(%r30) - STD %r4,-344(%r30) - STD %r5,-336(%r30) - STD %r6,-328(%r30) - STD %r7,-320(%r30) - STD %r8,-312(%r30) - STD %r9,-304(%r30) - STD %r10,-296(%r30) - - STD %r27,-288(%r30) ; save gp - - COPY %r24,%r3 ; save d - COPY %r26,%r4 ; save h (high 64-bits) - LDO -1(%r0),%ret0 ; return -1 by default - - CMPB,*= %r0,%arg2,$D3 ; if (d == 0) - COPY %r25,%r5 ; save l (low 64-bits) - - LDO -48(%r30),%r29 ; create ap - .CALL ;in=26,29;out=28; - B,L BN_num_bits_word,%r2 - COPY %r3,%r26 - LDD -288(%r30),%r27 ; restore gp - LDI 64,%r21 - - CMPB,= %r21,%ret0,$00000012 ;if (i == 64) (forward) - COPY %ret0,%r24 ; i - MTSARCM %r24 - DEPDI,Z -1,%sar,1,%r29 - CMPB,*<<,N %r29,%r4,bn_div_err_case ; if (h > 1<= d) - SUB %r4,%r3,%r4 ; h -= d - CMPB,= %r31,%r0,$0000001A ; if (i) - COPY %r0,%r10 ; ret = 0 - MTSARCM %r31 ; i to shift - DEPD,Z %r3,%sar,64,%r3 ; d <<= i; - SUBI 64,%r31,%r19 ; 64 - i; redundent - MTSAR %r19 ; (64 -i) to shift - SHRPD %r4,%r5,%sar,%r4 ; l>> (64-i) - MTSARCM %r31 ; i to shift - DEPD,Z %r5,%sar,64,%r5 ; l <<= i; - -$0000001A - DEPDI,Z -1,31,32,%r19 - EXTRD,U %r3,31,32,%r6 ; dh=(d&0xfff)>>32 - EXTRD,U %r3,63,32,%r8 ; dl = d&0xffffff - LDO 2(%r0),%r9 - STD %r3,-280(%r30) ; "d" to stack - -$0000001C - DEPDI,Z -1,63,32,%r29 ; - EXTRD,U %r4,31,32,%r31 ; h >> 32 - CMPB,*=,N %r31,%r6,$D2 ; if ((h>>32) != dh)(forward) div - COPY %r4,%r26 - EXTRD,U %r4,31,32,%r25 - COPY %r6,%r24 - .CALL ;in=23,24,25,26;out=20,21,22,28,29; (MILLICALL) - B,L $$div2U,%r2 - EXTRD,U %r6,31,32,%r23 - DEPD %r28,31,32,%r29 -$D2 - STD %r29,-272(%r30) ; q - AND %r5,%r19,%r24 ; t & 0xffffffff00000000; - EXTRD,U %r24,31,32,%r24 ; ??? - FLDD -272(%r30),%fr7 ; q - FLDD -280(%r30),%fr8 ; d - XMPYU %fr8L,%fr7L,%fr10 - FSTD %fr10,-256(%r30) - XMPYU %fr8L,%fr7R,%fr22 - FSTD %fr22,-264(%r30) - XMPYU %fr8R,%fr7L,%fr11 - XMPYU %fr8R,%fr7R,%fr23 - FSTD %fr11,-232(%r30) - FSTD %fr23,-240(%r30) - LDD -256(%r30),%r28 - DEPD,Z %r28,31,32,%r2 - LDD -264(%r30),%r20 - ADD,L %r20,%r2,%r31 - LDD -232(%r30),%r22 - DEPD,Z %r22,31,32,%r22 - LDD -240(%r30),%r21 - B $00000024 ; enter loop - ADD,L %r21,%r22,%r23 - -$0000002A - LDO -1(%r29),%r29 - SUB %r23,%r8,%r23 -$00000024 - SUB %r4,%r31,%r25 - AND %r25,%r19,%r26 - CMPB,*<>,N %r0,%r26,$00000046 ; (forward) - DEPD,Z %r25,31,32,%r20 - OR %r20,%r24,%r21 - CMPB,*<<,N %r21,%r23,$0000002A ;(backward) - SUB %r31,%r6,%r31 -;-------------Break path--------------------- - -$00000046 - DEPD,Z %r23,31,32,%r25 ;tl - EXTRD,U %r23,31,32,%r26 ;t - AND %r25,%r19,%r24 ;tl = (tl<<32)&0xfffffff0000000L - ADD,L %r31,%r26,%r31 ;th += t; - CMPCLR,*>>= %r5,%r24,%r0 ;if (l>32)); - DEPD,Z %r29,31,32,%r10 ; ret = q<<32 - b $0000001C - DEPD,Z %r28,31,32,%r5 ; l = l << 32 - -$D1 - OR %r10,%r29,%r28 ; ret |= q -$D3 - LDD -368(%r30),%r2 -$D0 - LDD -296(%r30),%r10 - LDD -304(%r30),%r9 - LDD -312(%r30),%r8 - LDD -320(%r30),%r7 - LDD -328(%r30),%r6 - LDD -336(%r30),%r5 - LDD -344(%r30),%r4 - BVE (%r2) - .EXIT - LDD,MB -352(%r30),%r3 - -bn_div_err_case - MFIA %r6 - ADDIL L'bn_div_words-bn_div_err_case,%r6,%r1 - LDO R'bn_div_words-bn_div_err_case(%r1),%r6 - ADDIL LT'__iob,%r27,%r1 - LDD RT'__iob(%r1),%r26 - ADDIL L'C$4-bn_div_words,%r6,%r1 - LDO R'C$4-bn_div_words(%r1),%r25 - LDO 64(%r26),%r26 - .CALL ;in=24,25,26,29;out=28; - B,L fprintf,%r2 - LDO -48(%r30),%r29 - LDD -288(%r30),%r27 - .CALL ;in=29; - B,L abort,%r2 - LDO -48(%r30),%r29 - LDD -288(%r30),%r27 - B $D0 - LDD -368(%r30),%r2 - .PROCEND ;in=24,25,26,29;out=28; - -;---------------------------------------------------------------------------- -; -; Registers to hold 64-bit values to manipulate. The "L" part -; of the register corresponds to the upper 32-bits, while the "R" -; part corresponds to the lower 32-bits -; -; Note, that when using b6 and b7, the code must save these before -; using them because they are callee save registers -; -; -; Floating point registers to use to save values that -; are manipulated. These don't collide with ftemp1-6 and -; are all caller save registers -; -a0 .reg %fr22 -a0L .reg %fr22L -a0R .reg %fr22R - -a1 .reg %fr23 -a1L .reg %fr23L -a1R .reg %fr23R - -a2 .reg %fr24 -a2L .reg %fr24L -a2R .reg %fr24R - -a3 .reg %fr25 -a3L .reg %fr25L -a3R .reg %fr25R - -a4 .reg %fr26 -a4L .reg %fr26L -a4R .reg %fr26R - -a5 .reg %fr27 -a5L .reg %fr27L -a5R .reg %fr27R - -a6 .reg %fr28 -a6L .reg %fr28L -a6R .reg %fr28R - -a7 .reg %fr29 -a7L .reg %fr29L -a7R .reg %fr29R - -b0 .reg %fr30 -b0L .reg %fr30L -b0R .reg %fr30R - -b1 .reg %fr31 -b1L .reg %fr31L -b1R .reg %fr31R - -; -; Temporary floating point variables, these are all caller save -; registers -; -ftemp1 .reg %fr4 -ftemp2 .reg %fr5 -ftemp3 .reg %fr6 -ftemp4 .reg %fr7 - -; -; The B set of registers when used. -; - -b2 .reg %fr8 -b2L .reg %fr8L -b2R .reg %fr8R - -b3 .reg %fr9 -b3L .reg %fr9L -b3R .reg %fr9R - -b4 .reg %fr10 -b4L .reg %fr10L -b4R .reg %fr10R - -b5 .reg %fr11 -b5L .reg %fr11L -b5R .reg %fr11R - -b6 .reg %fr12 -b6L .reg %fr12L -b6R .reg %fr12R - -b7 .reg %fr13 -b7L .reg %fr13L -b7R .reg %fr13R - -c1 .reg %r21 ; only reg -temp1 .reg %r20 ; only reg -temp2 .reg %r19 ; only reg -temp3 .reg %r31 ; only reg - -m1 .reg %r28 -c2 .reg %r23 -high_one .reg %r1 -ht .reg %r6 -lt .reg %r5 -m .reg %r4 -c3 .reg %r3 - -SQR_ADD_C .macro A0L,A0R,C1,C2,C3 - XMPYU A0L,A0R,ftemp1 ; m - FSTD ftemp1,-24(%sp) ; store m - - XMPYU A0R,A0R,ftemp2 ; lt - FSTD ftemp2,-16(%sp) ; store lt - - XMPYU A0L,A0L,ftemp3 ; ht - FSTD ftemp3,-8(%sp) ; store ht - - LDD -24(%sp),m ; load m - AND m,high_mask,temp2 ; m & Mask - DEPD,Z m,30,31,temp3 ; m << 32+1 - LDD -16(%sp),lt ; lt - - LDD -8(%sp),ht ; ht - EXTRD,U temp2,32,33,temp1 ; temp1 = m&Mask >> 32-1 - ADD temp3,lt,lt ; lt = lt+m - ADD,L ht,temp1,ht ; ht += temp1 - ADD,DC ht,%r0,ht ; ht++ - - ADD C1,lt,C1 ; c1=c1+lt - ADD,DC ht,%r0,ht ; ht++ - - ADD C2,ht,C2 ; c2=c2+ht - ADD,DC C3,%r0,C3 ; c3++ -.endm - -SQR_ADD_C2 .macro A0L,A0R,A1L,A1R,C1,C2,C3 - XMPYU A0L,A1R,ftemp1 ; m1 = bl*ht - FSTD ftemp1,-16(%sp) ; - XMPYU A0R,A1L,ftemp2 ; m = bh*lt - FSTD ftemp2,-8(%sp) ; - XMPYU A0R,A1R,ftemp3 ; lt = bl*lt - FSTD ftemp3,-32(%sp) - XMPYU A0L,A1L,ftemp4 ; ht = bh*ht - FSTD ftemp4,-24(%sp) ; - - LDD -8(%sp),m ; r21 = m - LDD -16(%sp),m1 ; r19 = m1 - ADD,L m,m1,m ; m+m1 - - DEPD,Z m,31,32,temp3 ; (m+m1<<32) - LDD -24(%sp),ht ; r24 = ht - - CMPCLR,*>>= m,m1,%r0 ; if (m < m1) - ADD,L ht,high_one,ht ; ht+=high_one - - EXTRD,U m,31,32,temp1 ; m >> 32 - LDD -32(%sp),lt ; lt - ADD,L ht,temp1,ht ; ht+= m>>32 - ADD lt,temp3,lt ; lt = lt+m1 - ADD,DC ht,%r0,ht ; ht++ - - ADD ht,ht,ht ; ht=ht+ht; - ADD,DC C3,%r0,C3 ; add in carry (c3++) - - ADD lt,lt,lt ; lt=lt+lt; - ADD,DC ht,%r0,ht ; add in carry (ht++) - - ADD C1,lt,C1 ; c1=c1+lt - ADD,DC,*NUV ht,%r0,ht ; add in carry (ht++) - LDO 1(C3),C3 ; bump c3 if overflow,nullify otherwise - - ADD C2,ht,C2 ; c2 = c2 + ht - ADD,DC C3,%r0,C3 ; add in carry (c3++) -.endm - -; -;void bn_sqr_comba8(BN_ULONG *r, BN_ULONG *a) -; arg0 = r_ptr -; arg1 = a_ptr -; - -bn_sqr_comba8 - .PROC - .CALLINFO FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_sqr_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .ENTRY - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z -1,32,33,high_mask ; Create Mask 0xffffffff80000000L - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - FLDD 32(a_ptr),a4 - FLDD 40(a_ptr),a5 - FLDD 48(a_ptr),a6 - FLDD 56(a_ptr),a7 - - SQR_ADD_C a0L,a0R,c1,c2,c3 - STD c1,0(r_ptr) ; r[0] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1 - STD c2,8(r_ptr) ; r[1] = c2; - COPY %r0,c2 - - SQR_ADD_C a1L,a1R,c3,c1,c2 - SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2 - STD c3,16(r_ptr) ; r[2] = c3; - COPY %r0,c3 - - SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3 - SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3 - STD c1,24(r_ptr) ; r[3] = c1; - COPY %r0,c1 - - SQR_ADD_C a2L,a2R,c2,c3,c1 - SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1 - SQR_ADD_C2 a4L,a4R,a0L,a0R,c2,c3,c1 - STD c2,32(r_ptr) ; r[4] = c2; - COPY %r0,c2 - - SQR_ADD_C2 a5L,a5R,a0L,a0R,c3,c1,c2 - SQR_ADD_C2 a4L,a4R,a1L,a1R,c3,c1,c2 - SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2 - STD c3,40(r_ptr) ; r[5] = c3; - COPY %r0,c3 - - SQR_ADD_C a3L,a3R,c1,c2,c3 - SQR_ADD_C2 a4L,a4R,a2L,a2R,c1,c2,c3 - SQR_ADD_C2 a5L,a5R,a1L,a1R,c1,c2,c3 - SQR_ADD_C2 a6L,a6R,a0L,a0R,c1,c2,c3 - STD c1,48(r_ptr) ; r[6] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a7L,a7R,a0L,a0R,c2,c3,c1 - SQR_ADD_C2 a6L,a6R,a1L,a1R,c2,c3,c1 - SQR_ADD_C2 a5L,a5R,a2L,a2R,c2,c3,c1 - SQR_ADD_C2 a4L,a4R,a3L,a3R,c2,c3,c1 - STD c2,56(r_ptr) ; r[7] = c2; - COPY %r0,c2 - - SQR_ADD_C a4L,a4R,c3,c1,c2 - SQR_ADD_C2 a5L,a5R,a3L,a3R,c3,c1,c2 - SQR_ADD_C2 a6L,a6R,a2L,a2R,c3,c1,c2 - SQR_ADD_C2 a7L,a7R,a1L,a1R,c3,c1,c2 - STD c3,64(r_ptr) ; r[8] = c3; - COPY %r0,c3 - - SQR_ADD_C2 a7L,a7R,a2L,a2R,c1,c2,c3 - SQR_ADD_C2 a6L,a6R,a3L,a3R,c1,c2,c3 - SQR_ADD_C2 a5L,a5R,a4L,a4R,c1,c2,c3 - STD c1,72(r_ptr) ; r[9] = c1; - COPY %r0,c1 - - SQR_ADD_C a5L,a5R,c2,c3,c1 - SQR_ADD_C2 a6L,a6R,a4L,a4R,c2,c3,c1 - SQR_ADD_C2 a7L,a7R,a3L,a3R,c2,c3,c1 - STD c2,80(r_ptr) ; r[10] = c2; - COPY %r0,c2 - - SQR_ADD_C2 a7L,a7R,a4L,a4R,c3,c1,c2 - SQR_ADD_C2 a6L,a6R,a5L,a5R,c3,c1,c2 - STD c3,88(r_ptr) ; r[11] = c3; - COPY %r0,c3 - - SQR_ADD_C a6L,a6R,c1,c2,c3 - SQR_ADD_C2 a7L,a7R,a5L,a5R,c1,c2,c3 - STD c1,96(r_ptr) ; r[12] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a7L,a7R,a6L,a6R,c2,c3,c1 - STD c2,104(r_ptr) ; r[13] = c2; - COPY %r0,c2 - - SQR_ADD_C a7L,a7R,c3,c1,c2 - STD c3, 112(r_ptr) ; r[14] = c3 - STD c1, 120(r_ptr) ; r[15] = c1 - - .EXIT - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - -;----------------------------------------------------------------------------- -; -;void bn_sqr_comba4(BN_ULONG *r, BN_ULONG *a) -; arg0 = r_ptr -; arg1 = a_ptr -; - -bn_sqr_comba4 - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_sqr_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z -1,32,33,high_mask ; Create Mask 0xffffffff80000000L - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - FLDD 32(a_ptr),a4 - FLDD 40(a_ptr),a5 - FLDD 48(a_ptr),a6 - FLDD 56(a_ptr),a7 - - SQR_ADD_C a0L,a0R,c1,c2,c3 - - STD c1,0(r_ptr) ; r[0] = c1; - COPY %r0,c1 - - SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1 - - STD c2,8(r_ptr) ; r[1] = c2; - COPY %r0,c2 - - SQR_ADD_C a1L,a1R,c3,c1,c2 - SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2 - - STD c3,16(r_ptr) ; r[2] = c3; - COPY %r0,c3 - - SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3 - SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3 - - STD c1,24(r_ptr) ; r[3] = c1; - COPY %r0,c1 - - SQR_ADD_C a2L,a2R,c2,c3,c1 - SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1 - - STD c2,32(r_ptr) ; r[4] = c2; - COPY %r0,c2 - - SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2 - STD c3,40(r_ptr) ; r[5] = c3; - COPY %r0,c3 - - SQR_ADD_C a3L,a3R,c1,c2,c3 - STD c1,48(r_ptr) ; r[6] = c1; - STD c2,56(r_ptr) ; r[7] = c2; - - .EXIT - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - - -;--------------------------------------------------------------------------- - -MUL_ADD_C .macro A0L,A0R,B0L,B0R,C1,C2,C3 - XMPYU A0L,B0R,ftemp1 ; m1 = bl*ht - FSTD ftemp1,-16(%sp) ; - XMPYU A0R,B0L,ftemp2 ; m = bh*lt - FSTD ftemp2,-8(%sp) ; - XMPYU A0R,B0R,ftemp3 ; lt = bl*lt - FSTD ftemp3,-32(%sp) - XMPYU A0L,B0L,ftemp4 ; ht = bh*ht - FSTD ftemp4,-24(%sp) ; - - LDD -8(%sp),m ; r21 = m - LDD -16(%sp),m1 ; r19 = m1 - ADD,L m,m1,m ; m+m1 - - DEPD,Z m,31,32,temp3 ; (m+m1<<32) - LDD -24(%sp),ht ; r24 = ht - - CMPCLR,*>>= m,m1,%r0 ; if (m < m1) - ADD,L ht,high_one,ht ; ht+=high_one - - EXTRD,U m,31,32,temp1 ; m >> 32 - LDD -32(%sp),lt ; lt - ADD,L ht,temp1,ht ; ht+= m>>32 - ADD lt,temp3,lt ; lt = lt+m1 - ADD,DC ht,%r0,ht ; ht++ - - ADD C1,lt,C1 ; c1=c1+lt - ADD,DC ht,%r0,ht ; bump c3 if overflow,nullify otherwise - - ADD C2,ht,C2 ; c2 = c2 + ht - ADD,DC C3,%r0,C3 ; add in carry (c3++) -.endm - - -; -;void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b) -; arg0 = r_ptr -; arg1 = a_ptr -; arg2 = b_ptr -; - -bn_mul_comba8 - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_mul_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - FSTD %fr12,32(%sp) ; save r6 - FSTD %fr13,40(%sp) ; save r7 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - FLDD 32(a_ptr),a4 - FLDD 40(a_ptr),a5 - FLDD 48(a_ptr),a6 - FLDD 56(a_ptr),a7 - - FLDD 0(b_ptr),b0 - FLDD 8(b_ptr),b1 - FLDD 16(b_ptr),b2 - FLDD 24(b_ptr),b3 - FLDD 32(b_ptr),b4 - FLDD 40(b_ptr),b5 - FLDD 48(b_ptr),b6 - FLDD 56(b_ptr),b7 - - MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3 - STD c1,0(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1 - STD c2,8(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2 - STD c3,16(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3 - MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3 - STD c1,24(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a4L,a4R,b0L,b0R,c2,c3,c1 - MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1 - MUL_ADD_C a0L,a0R,b4L,b4R,c2,c3,c1 - STD c2,32(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a0L,a0R,b5L,b5R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b4L,b4R,c3,c1,c2 - MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2 - MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2 - MUL_ADD_C a4L,a4R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a5L,a5R,b0L,b0R,c3,c1,c2 - STD c3,40(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a6L,a6R,b0L,b0R,c1,c2,c3 - MUL_ADD_C a5L,a5R,b1L,b1R,c1,c2,c3 - MUL_ADD_C a4L,a4R,b2L,b2R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a2L,a2R,b4L,b4R,c1,c2,c3 - MUL_ADD_C a1L,a1R,b5L,b5R,c1,c2,c3 - MUL_ADD_C a0L,a0R,b6L,b6R,c1,c2,c3 - STD c1,48(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a0L,a0R,b7L,b7R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b6L,b6R,c2,c3,c1 - MUL_ADD_C a2L,a2R,b5L,b5R,c2,c3,c1 - MUL_ADD_C a3L,a3R,b4L,b4R,c2,c3,c1 - MUL_ADD_C a4L,a4R,b3L,b3R,c2,c3,c1 - MUL_ADD_C a5L,a5R,b2L,b2R,c2,c3,c1 - MUL_ADD_C a6L,a6R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a7L,a7R,b0L,b0R,c2,c3,c1 - STD c2,56(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a7L,a7R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a6L,a6R,b2L,b2R,c3,c1,c2 - MUL_ADD_C a5L,a5R,b3L,b3R,c3,c1,c2 - MUL_ADD_C a4L,a4R,b4L,b4R,c3,c1,c2 - MUL_ADD_C a3L,a3R,b5L,b5R,c3,c1,c2 - MUL_ADD_C a2L,a2R,b6L,b6R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b7L,b7R,c3,c1,c2 - STD c3,64(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a2L,a2R,b7L,b7R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b6L,b6R,c1,c2,c3 - MUL_ADD_C a4L,a4R,b5L,b5R,c1,c2,c3 - MUL_ADD_C a5L,a5R,b4L,b4R,c1,c2,c3 - MUL_ADD_C a6L,a6R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a7L,a7R,b2L,b2R,c1,c2,c3 - STD c1,72(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a7L,a7R,b3L,b3R,c2,c3,c1 - MUL_ADD_C a6L,a6R,b4L,b4R,c2,c3,c1 - MUL_ADD_C a5L,a5R,b5L,b5R,c2,c3,c1 - MUL_ADD_C a4L,a4R,b6L,b6R,c2,c3,c1 - MUL_ADD_C a3L,a3R,b7L,b7R,c2,c3,c1 - STD c2,80(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a4L,a4R,b7L,b7R,c3,c1,c2 - MUL_ADD_C a5L,a5R,b6L,b6R,c3,c1,c2 - MUL_ADD_C a6L,a6R,b5L,b5R,c3,c1,c2 - MUL_ADD_C a7L,a7R,b4L,b4R,c3,c1,c2 - STD c3,88(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a7L,a7R,b5L,b5R,c1,c2,c3 - MUL_ADD_C a6L,a6R,b6L,b6R,c1,c2,c3 - MUL_ADD_C a5L,a5R,b7L,b7R,c1,c2,c3 - STD c1,96(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a6L,a6R,b7L,b7R,c2,c3,c1 - MUL_ADD_C a7L,a7R,b6L,b6R,c2,c3,c1 - STD c2,104(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a7L,a7R,b7L,b7R,c3,c1,c2 - STD c3,112(r_ptr) - STD c1,120(r_ptr) - - .EXIT - FLDD -88(%sp),%fr13 - FLDD -96(%sp),%fr12 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - -;----------------------------------------------------------------------------- -; -;void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b) -; arg0 = r_ptr -; arg1 = a_ptr -; arg2 = b_ptr -; - -bn_mul_comba4 - .proc - .callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE - .EXPORT bn_mul_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN - .entry - .align 64 - - STD %r3,0(%sp) ; save r3 - STD %r4,8(%sp) ; save r4 - STD %r5,16(%sp) ; save r5 - STD %r6,24(%sp) ; save r6 - FSTD %fr12,32(%sp) ; save r6 - FSTD %fr13,40(%sp) ; save r7 - - ; - ; Zero out carries - ; - COPY %r0,c1 - COPY %r0,c2 - COPY %r0,c3 - - LDO 128(%sp),%sp ; bump stack - DEPDI,Z 1,31,1,high_one ; Create Value 1 << 32 - - ; - ; Load up all of the values we are going to use - ; - FLDD 0(a_ptr),a0 - FLDD 8(a_ptr),a1 - FLDD 16(a_ptr),a2 - FLDD 24(a_ptr),a3 - - FLDD 0(b_ptr),b0 - FLDD 8(b_ptr),b1 - FLDD 16(b_ptr),b2 - FLDD 24(b_ptr),b3 - - MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3 - STD c1,0(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1 - STD c2,8(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2 - MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2 - MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2 - STD c3,16(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3 - MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3 - MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3 - MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3 - STD c1,24(r_ptr) - COPY %r0,c1 - - MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1 - MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1 - MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1 - STD c2,32(r_ptr) - COPY %r0,c2 - - MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2 - MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2 - STD c3,40(r_ptr) - COPY %r0,c3 - - MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3 - STD c1,48(r_ptr) - STD c2,56(r_ptr) - - .EXIT - FLDD -88(%sp),%fr13 - FLDD -96(%sp),%fr12 - LDD -104(%sp),%r6 ; restore r6 - LDD -112(%sp),%r5 ; restore r5 - LDD -120(%sp),%r4 ; restore r4 - BVE (%rp) - LDD,MB -128(%sp),%r3 - - .PROCEND - - - .SPACE $TEXT$ - .SUBSPA $CODE$ - .SPACE $PRIVATE$,SORT=16 - .IMPORT $global$,DATA - .SPACE $TEXT$ - .SUBSPA $CODE$ - .SUBSPA $LIT$,ACCESS=0x2c -C$4 - .ALIGN 8 - .STRINGZ "Division would overflow (%d)\n" - .END diff --git a/src/lib/libcrypto/bn/asm/ppc.pl b/src/lib/libcrypto/bn/asm/ppc.pl deleted file mode 100644 index 08e0053473..0000000000 --- a/src/lib/libcrypto/bn/asm/ppc.pl +++ /dev/null @@ -1,2078 +0,0 @@ -#!/usr/bin/env perl -# -# Implemented as a Perl wrapper as we want to support several different -# architectures with single file. We pick up the target based on the -# file name we are asked to generate. -# -# It should be noted though that this perl code is nothing like -# /crypto/perlasm/x86*. In this case perl is used pretty much -# as pre-processor to cover for platform differences in name decoration, -# linker tables, 32-/64-bit instruction sets... -# -# As you might know there're several PowerPC ABI in use. Most notably -# Linux and AIX use different 32-bit ABIs. Good news are that these ABIs -# are similar enough to implement leaf(!) functions, which would be ABI -# neutral. And that's what you find here: ABI neutral leaf functions. -# In case you wonder what that is... -# -# AIX performance -# -# MEASUREMENTS WITH cc ON a 200 MhZ PowerPC 604e. -# -# The following is the performance of 32-bit compiler -# generated code: -# -# OpenSSL 0.9.6c 21 dec 2001 -# built on: Tue Jun 11 11:06:51 EDT 2002 -# options:bn(64,32) ... -#compiler: cc -DTHREADS -DAIX -DB_ENDIAN -DBN_LLONG -O3 -# sign verify sign/s verify/s -#rsa 512 bits 0.0098s 0.0009s 102.0 1170.6 -#rsa 1024 bits 0.0507s 0.0026s 19.7 387.5 -#rsa 2048 bits 0.3036s 0.0085s 3.3 117.1 -#rsa 4096 bits 2.0040s 0.0299s 0.5 33.4 -#dsa 512 bits 0.0087s 0.0106s 114.3 94.5 -#dsa 1024 bits 0.0256s 0.0313s 39.0 32.0 -# -# Same bechmark with this assembler code: -# -#rsa 512 bits 0.0056s 0.0005s 178.6 2049.2 -#rsa 1024 bits 0.0283s 0.0015s 35.3 674.1 -#rsa 2048 bits 0.1744s 0.0050s 5.7 201.2 -#rsa 4096 bits 1.1644s 0.0179s 0.9 55.7 -#dsa 512 bits 0.0052s 0.0062s 191.6 162.0 -#dsa 1024 bits 0.0149s 0.0180s 67.0 55.5 -# -# Number of operations increases by at almost 75% -# -# Here are performance numbers for 64-bit compiler -# generated code: -# -# OpenSSL 0.9.6g [engine] 9 Aug 2002 -# built on: Fri Apr 18 16:59:20 EDT 2003 -# options:bn(64,64) ... -# compiler: cc -DTHREADS -D_REENTRANT -q64 -DB_ENDIAN -O3 -# sign verify sign/s verify/s -#rsa 512 bits 0.0028s 0.0003s 357.1 3844.4 -#rsa 1024 bits 0.0148s 0.0008s 67.5 1239.7 -#rsa 2048 bits 0.0963s 0.0028s 10.4 353.0 -#rsa 4096 bits 0.6538s 0.0102s 1.5 98.1 -#dsa 512 bits 0.0026s 0.0032s 382.5 313.7 -#dsa 1024 bits 0.0081s 0.0099s 122.8 100.6 -# -# Same benchmark with this assembler code: -# -#rsa 512 bits 0.0020s 0.0002s 510.4 6273.7 -#rsa 1024 bits 0.0088s 0.0005s 114.1 2128.3 -#rsa 2048 bits 0.0540s 0.0016s 18.5 622.5 -#rsa 4096 bits 0.3700s 0.0058s 2.7 171.0 -#dsa 512 bits 0.0016s 0.0020s 610.7 507.1 -#dsa 1024 bits 0.0047s 0.0058s 212.5 173.2 -# -# Again, performance increases by at about 75% -# -# Mac OS X, Apple G5 1.8GHz (Note this is 32 bit code) -# OpenSSL 0.9.7c 30 Sep 2003 -# -# Original code. -# -#rsa 512 bits 0.0011s 0.0001s 906.1 11012.5 -#rsa 1024 bits 0.0060s 0.0003s 166.6 3363.1 -#rsa 2048 bits 0.0370s 0.0010s 27.1 982.4 -#rsa 4096 bits 0.2426s 0.0036s 4.1 280.4 -#dsa 512 bits 0.0010s 0.0012s 1038.1 841.5 -#dsa 1024 bits 0.0030s 0.0037s 329.6 269.7 -#dsa 2048 bits 0.0101s 0.0127s 98.9 78.6 -# -# Same benchmark with this assembler code: -# -#rsa 512 bits 0.0007s 0.0001s 1416.2 16645.9 -#rsa 1024 bits 0.0036s 0.0002s 274.4 5380.6 -#rsa 2048 bits 0.0222s 0.0006s 45.1 1589.5 -#rsa 4096 bits 0.1469s 0.0022s 6.8 449.6 -#dsa 512 bits 0.0006s 0.0007s 1664.2 1376.2 -#dsa 1024 bits 0.0018s 0.0023s 545.0 442.2 -#dsa 2048 bits 0.0061s 0.0075s 163.5 132.8 -# -# Performance increase of ~60% -# -# If you have comments or suggestions to improve code send -# me a note at schari@us.ibm.com -# - -$opf = shift; - -if ($opf =~ /32\.s/) { - $BITS= 32; - $BNSZ= $BITS/8; - $ISA= "\"ppc\""; - - $LD= "lwz"; # load - $LDU= "lwzu"; # load and update - $ST= "stw"; # store - $STU= "stwu"; # store and update - $UMULL= "mullw"; # unsigned multiply low - $UMULH= "mulhwu"; # unsigned multiply high - $UDIV= "divwu"; # unsigned divide - $UCMPI= "cmplwi"; # unsigned compare with immediate - $UCMP= "cmplw"; # unsigned compare - $CNTLZ= "cntlzw"; # count leading zeros - $SHL= "slw"; # shift left - $SHR= "srw"; # unsigned shift right - $SHRI= "srwi"; # unsigned shift right by immediate - $SHLI= "slwi"; # shift left by immediate - $CLRU= "clrlwi"; # clear upper bits - $INSR= "insrwi"; # insert right - $ROTL= "rotlwi"; # rotate left by immediate - $TR= "tw"; # conditional trap -} elsif ($opf =~ /64\.s/) { - $BITS= 64; - $BNSZ= $BITS/8; - $ISA= "\"ppc64\""; - - # same as above, but 64-bit mnemonics... - $LD= "ld"; # load - $LDU= "ldu"; # load and update - $ST= "std"; # store - $STU= "stdu"; # store and update - $UMULL= "mulld"; # unsigned multiply low - $UMULH= "mulhdu"; # unsigned multiply high - $UDIV= "divdu"; # unsigned divide - $UCMPI= "cmpldi"; # unsigned compare with immediate - $UCMP= "cmpld"; # unsigned compare - $CNTLZ= "cntlzd"; # count leading zeros - $SHL= "sld"; # shift left - $SHR= "srd"; # unsigned shift right - $SHRI= "srdi"; # unsigned shift right by immediate - $SHLI= "sldi"; # shift left by immediate - $CLRU= "clrldi"; # clear upper bits - $INSR= "insrdi"; # insert right - $ROTL= "rotldi"; # rotate left by immediate - $TR= "td"; # conditional trap -} else { die "nonsense $opf"; } - -( defined shift || open STDOUT,">$opf" ) || die "can't open $opf: $!"; - -# function entry points from the AIX code -# -# There are other, more elegant, ways to handle this. We (IBM) chose -# this approach as it plays well with scripts we run to 'namespace' -# OpenSSL .i.e. we add a prefix to all the public symbols so we can -# co-exist in the same process with other implementations of OpenSSL. -# 'cleverer' ways of doing these substitutions tend to hide data we -# need to be obvious. -# -my @items = ("bn_sqr_comba4", - "bn_sqr_comba8", - "bn_mul_comba4", - "bn_mul_comba8", - "bn_sub_words", - "bn_add_words", - "bn_div_words", - "bn_sqr_words", - "bn_mul_words", - "bn_mul_add_words"); - -if ($opf =~ /linux/) { do_linux(); } -elsif ($opf =~ /aix/) { do_aix(); } -elsif ($opf =~ /osx/) { do_osx(); } -else { do_bsd(); } - -sub do_linux { - $d=&data(); - - if ($BITS==64) { - foreach $t (@items) { - $d =~ s/\.$t:/\ -\t.section\t".opd","aw"\ -\t.align\t3\ -\t.globl\t$t\ -$t:\ -\t.quad\t.$t,.TOC.\@tocbase,0\ -\t.size\t$t,24\ -\t.previous\n\ -\t.type\t.$t,\@function\ -\t.globl\t.$t\ -.$t:/g; - } - } - else { - foreach $t (@items) { - $d=~s/\.$t/$t/g; - } - } - # hide internal labels to avoid pollution of name table... - $d=~s/Lppcasm_/.Lppcasm_/gm; - print $d; -} - -sub do_aix { - # AIX assembler is smart enough to please the linker without - # making us do something special... - print &data(); -} - -# MacOSX 32 bit -sub do_osx { - $d=&data(); - # Change the bn symbol prefix from '.' to '_' - foreach $t (@items) { - $d=~s/\.$t/_$t/g; - } - # Change .machine to something OS X asm will accept - $d=~s/\.machine.*/.text/g; - $d=~s/\#/;/g; # change comment from '#' to ';' - print $d; -} - -# BSD (Untested) -sub do_bsd { - $d=&data(); - foreach $t (@items) { - $d=~s/\.$t/_$t/g; - } - print $d; -} - -sub data { - local($data)=< 0 then result !=0 - # In either case carry bit is set. - bc BO_IF,CR0_EQ,Lppcasm_sub_adios - addi r4,r4,-$BNSZ - addi r3,r3,-$BNSZ - addi r5,r5,-$BNSZ - mtctr r6 -Lppcasm_sub_mainloop: - $LDU r7,$BNSZ(r4) - $LDU r8,$BNSZ(r5) - subfe r6,r8,r7 # r6 = r7+carry bit + onescomplement(r8) - # if carry = 1 this is r7-r8. Else it - # is r7-r8 -1 as we need. - $STU r6,$BNSZ(r3) - bc BO_dCTR_NZERO,CR0_EQ,Lppcasm_sub_mainloop -Lppcasm_sub_adios: - subfze r3,r0 # if carry bit is set then r3 = 0 else -1 - andi. r3,r3,1 # keep only last bit. - bclr BO_ALWAYS,CR0_LT - .long 0x00000000 - - -# -# NOTE: The following label name should be changed to -# "bn_add_words" i.e. remove the first dot -# for the gcc compiler. This should be automatically -# done in the build -# - -.align 4 -.bn_add_words: -# -# Handcoded version of bn_add_words -# -#BN_ULONG bn_add_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n) -# -# r3 = r -# r4 = a -# r5 = b -# r6 = n -# -# Note: No loop unrolling done since this is not a performance -# critical loop. - - xor r0,r0,r0 -# -# check for r6 = 0. Is this needed? -# - addic. r6,r6,0 #test r6 and clear carry bit. - bc BO_IF,CR0_EQ,Lppcasm_add_adios - addi r4,r4,-$BNSZ - addi r3,r3,-$BNSZ - addi r5,r5,-$BNSZ - mtctr r6 -Lppcasm_add_mainloop: - $LDU r7,$BNSZ(r4) - $LDU r8,$BNSZ(r5) - adde r8,r7,r8 - $STU r8,$BNSZ(r3) - bc BO_dCTR_NZERO,CR0_EQ,Lppcasm_add_mainloop -Lppcasm_add_adios: - addze r3,r0 #return carry bit. - bclr BO_ALWAYS,CR0_LT - .long 0x00000000 - -# -# NOTE: The following label name should be changed to -# "bn_div_words" i.e. remove the first dot -# for the gcc compiler. This should be automatically -# done in the build -# - -.align 4 -.bn_div_words: -# -# This is a cleaned up version of code generated by -# the AIX compiler. The only optimization is to use -# the PPC instruction to count leading zeros instead -# of call to num_bits_word. Since this was compiled -# only at level -O2 we can possibly squeeze it more? -# -# r3 = h -# r4 = l -# r5 = d - - $UCMPI 0,r5,0 # compare r5 and 0 - bc BO_IF_NOT,CR0_EQ,Lppcasm_div1 # proceed if d!=0 - li r3,-1 # d=0 return -1 - bclr BO_ALWAYS,CR0_LT -Lppcasm_div1: - xor r0,r0,r0 #r0=0 - li r8,$BITS - $CNTLZ. r7,r5 #r7 = num leading 0s in d. - bc BO_IF,CR0_EQ,Lppcasm_div2 #proceed if no leading zeros - subf r8,r7,r8 #r8 = BN_num_bits_word(d) - $SHR. r9,r3,r8 #are there any bits above r8'th? - $TR 16,r9,r0 #if there're, signal to dump core... -Lppcasm_div2: - $UCMP 0,r3,r5 #h>=d? - bc BO_IF,CR0_LT,Lppcasm_div3 #goto Lppcasm_div3 if not - subf r3,r5,r3 #h-=d ; -Lppcasm_div3: #r7 = BN_BITS2-i. so r7=i - cmpi 0,0,r7,0 # is (i == 0)? - bc BO_IF,CR0_EQ,Lppcasm_div4 - $SHL r3,r3,r7 # h = (h<< i) - $SHR r8,r4,r8 # r8 = (l >> BN_BITS2 -i) - $SHL r5,r5,r7 # d<<=i - or r3,r3,r8 # h = (h<>(BN_BITS2-i)) - $SHL r4,r4,r7 # l <<=i -Lppcasm_div4: - $SHRI r9,r5,`$BITS/2` # r9 = dh - # dl will be computed when needed - # as it saves registers. - li r6,2 #r6=2 - mtctr r6 #counter will be in count. -Lppcasm_divouterloop: - $SHRI r8,r3,`$BITS/2` #r8 = (h>>BN_BITS4) - $SHRI r11,r4,`$BITS/2` #r11= (l&BN_MASK2h)>>BN_BITS4 - # compute here for innerloop. - $UCMP 0,r8,r9 # is (h>>BN_BITS4)==dh - bc BO_IF_NOT,CR0_EQ,Lppcasm_div5 # goto Lppcasm_div5 if not - - li r8,-1 - $CLRU r8,r8,`$BITS/2` #q = BN_MASK2l - b Lppcasm_div6 -Lppcasm_div5: - $UDIV r8,r3,r9 #q = h/dh -Lppcasm_div6: - $UMULL r12,r9,r8 #th = q*dh - $CLRU r10,r5,`$BITS/2` #r10=dl - $UMULL r6,r8,r10 #tl = q*dl - -Lppcasm_divinnerloop: - subf r10,r12,r3 #t = h -th - $SHRI r7,r10,`$BITS/2` #r7= (t &BN_MASK2H), sort of... - addic. r7,r7,0 #test if r7 == 0. used below. - # now want to compute - # r7 = (t<>BN_BITS4) - # the following 2 instructions do that - $SHLI r7,r10,`$BITS/2` # r7 = (t<>BN_BITS4) - $UCMP 1,r6,r7 # compare (tl <= r7) - bc BO_IF_NOT,CR0_EQ,Lppcasm_divinnerexit - bc BO_IF_NOT,CR1_FEX,Lppcasm_divinnerexit - addi r8,r8,-1 #q-- - subf r12,r9,r12 #th -=dh - $CLRU r10,r5,`$BITS/2` #r10=dl. t is no longer needed in loop. - subf r6,r10,r6 #tl -=dl - b Lppcasm_divinnerloop -Lppcasm_divinnerexit: - $SHRI r10,r6,`$BITS/2` #t=(tl>>BN_BITS4) - $SHLI r11,r6,`$BITS/2` #tl=(tl<=tl) goto Lppcasm_div7 - addi r12,r12,1 # th++ -Lppcasm_div7: - subf r11,r11,r4 #r11=l-tl - $UCMP 1,r3,r12 #compare h and th - bc BO_IF_NOT,CR1_FX,Lppcasm_div8 #if (h>=th) goto Lppcasm_div8 - addi r8,r8,-1 # q-- - add r3,r5,r3 # h+=d -Lppcasm_div8: - subf r12,r12,r3 #r12 = h-th - $SHLI r4,r11,`$BITS/2` #l=(l&BN_MASK2l)<>BN_BITS4))&BN_MASK2 - # the following 2 instructions will do this. - $INSR r11,r12,`$BITS/2`,`$BITS/2` # r11 is the value we want rotated $BITS/2. - $ROTL r3,r11,`$BITS/2` # rotate by $BITS/2 and store in r3 - bc BO_dCTR_ZERO,CR0_EQ,Lppcasm_div9#if (count==0) break ; - $SHLI r0,r8,`$BITS/2` #ret =q<> 2 - bc BO_IF,CR0_EQ,Lppcasm_mw_REM - mtctr r7 -Lppcasm_mw_LOOP: - #mul(rp[0],ap[0],w,c1); - $LD r8,`0*$BNSZ`(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - addc r9,r9,r12 - #addze r10,r10 #carry is NOT ignored. - #will be taken care of - #in second spin below - #using adde. - $ST r9,`0*$BNSZ`(r3) - #mul(rp[1],ap[1],w,c1); - $LD r8,`1*$BNSZ`(r4) - $UMULL r11,r6,r8 - $UMULH r12,r6,r8 - adde r11,r11,r10 - #addze r12,r12 - $ST r11,`1*$BNSZ`(r3) - #mul(rp[2],ap[2],w,c1); - $LD r8,`2*$BNSZ`(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - adde r9,r9,r12 - #addze r10,r10 - $ST r9,`2*$BNSZ`(r3) - #mul_add(rp[3],ap[3],w,c1); - $LD r8,`3*$BNSZ`(r4) - $UMULL r11,r6,r8 - $UMULH r12,r6,r8 - adde r11,r11,r10 - addze r12,r12 #this spin we collect carry into - #r12 - $ST r11,`3*$BNSZ`(r3) - - addi r3,r3,`4*$BNSZ` - addi r4,r4,`4*$BNSZ` - bc BO_dCTR_NZERO,CR0_EQ,Lppcasm_mw_LOOP - -Lppcasm_mw_REM: - andi. r5,r5,0x3 - bc BO_IF,CR0_EQ,Lppcasm_mw_OVER - #mul(rp[0],ap[0],w,c1); - $LD r8,`0*$BNSZ`(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - addc r9,r9,r12 - addze r10,r10 - $ST r9,`0*$BNSZ`(r3) - addi r12,r10,0 - - addi r5,r5,-1 - cmpli 0,0,r5,0 - bc BO_IF,CR0_EQ,Lppcasm_mw_OVER - - - #mul(rp[1],ap[1],w,c1); - $LD r8,`1*$BNSZ`(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - addc r9,r9,r12 - addze r10,r10 - $ST r9,`1*$BNSZ`(r3) - addi r12,r10,0 - - addi r5,r5,-1 - cmpli 0,0,r5,0 - bc BO_IF,CR0_EQ,Lppcasm_mw_OVER - - #mul_add(rp[2],ap[2],w,c1); - $LD r8,`2*$BNSZ`(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - addc r9,r9,r12 - addze r10,r10 - $ST r9,`2*$BNSZ`(r3) - addi r12,r10,0 - -Lppcasm_mw_OVER: - addi r3,r12,0 - bclr BO_ALWAYS,CR0_LT - .long 0x00000000 - -# -# NOTE: The following label name should be changed to -# "bn_mul_add_words" i.e. remove the first dot -# for the gcc compiler. This should be automatically -# done in the build -# - -.align 4 -.bn_mul_add_words: -# -# BN_ULONG bn_mul_add_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w) -# -# r3 = rp -# r4 = ap -# r5 = num -# r6 = w -# -# empirical evidence suggests that unrolled version performs best!! -# - xor r0,r0,r0 #r0 = 0 - xor r12,r12,r12 #r12 = 0 . used for carry - rlwinm. r7,r5,30,2,31 # num >> 2 - bc BO_IF,CR0_EQ,Lppcasm_maw_leftover # if (num < 4) go LPPCASM_maw_leftover - mtctr r7 -Lppcasm_maw_mainloop: - #mul_add(rp[0],ap[0],w,c1); - $LD r8,`0*$BNSZ`(r4) - $LD r11,`0*$BNSZ`(r3) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - addc r9,r9,r12 #r12 is carry. - addze r10,r10 - addc r9,r9,r11 - #addze r10,r10 - #the above instruction addze - #is NOT needed. Carry will NOT - #be ignored. It's not affected - #by multiply and will be collected - #in the next spin - $ST r9,`0*$BNSZ`(r3) - - #mul_add(rp[1],ap[1],w,c1); - $LD r8,`1*$BNSZ`(r4) - $LD r9,`1*$BNSZ`(r3) - $UMULL r11,r6,r8 - $UMULH r12,r6,r8 - adde r11,r11,r10 #r10 is carry. - addze r12,r12 - addc r11,r11,r9 - #addze r12,r12 - $ST r11,`1*$BNSZ`(r3) - - #mul_add(rp[2],ap[2],w,c1); - $LD r8,`2*$BNSZ`(r4) - $UMULL r9,r6,r8 - $LD r11,`2*$BNSZ`(r3) - $UMULH r10,r6,r8 - adde r9,r9,r12 - addze r10,r10 - addc r9,r9,r11 - #addze r10,r10 - $ST r9,`2*$BNSZ`(r3) - - #mul_add(rp[3],ap[3],w,c1); - $LD r8,`3*$BNSZ`(r4) - $UMULL r11,r6,r8 - $LD r9,`3*$BNSZ`(r3) - $UMULH r12,r6,r8 - adde r11,r11,r10 - addze r12,r12 - addc r11,r11,r9 - addze r12,r12 - $ST r11,`3*$BNSZ`(r3) - addi r3,r3,`4*$BNSZ` - addi r4,r4,`4*$BNSZ` - bc BO_dCTR_NZERO,CR0_EQ,Lppcasm_maw_mainloop - -Lppcasm_maw_leftover: - andi. r5,r5,0x3 - bc BO_IF,CR0_EQ,Lppcasm_maw_adios - addi r3,r3,-$BNSZ - addi r4,r4,-$BNSZ - #mul_add(rp[0],ap[0],w,c1); - mtctr r5 - $LDU r8,$BNSZ(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - $LDU r11,$BNSZ(r3) - addc r9,r9,r11 - addze r10,r10 - addc r9,r9,r12 - addze r12,r10 - $ST r9,0(r3) - - bc BO_dCTR_ZERO,CR0_EQ,Lppcasm_maw_adios - #mul_add(rp[1],ap[1],w,c1); - $LDU r8,$BNSZ(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - $LDU r11,$BNSZ(r3) - addc r9,r9,r11 - addze r10,r10 - addc r9,r9,r12 - addze r12,r10 - $ST r9,0(r3) - - bc BO_dCTR_ZERO,CR0_EQ,Lppcasm_maw_adios - #mul_add(rp[2],ap[2],w,c1); - $LDU r8,$BNSZ(r4) - $UMULL r9,r6,r8 - $UMULH r10,r6,r8 - $LDU r11,$BNSZ(r3) - addc r9,r9,r11 - addze r10,r10 - addc r9,r9,r12 - addze r12,r10 - $ST r9,0(r3) - -Lppcasm_maw_adios: - addi r3,r12,0 - bclr BO_ALWAYS,CR0_LT - .long 0x00000000 - .align 4 -EOF - $data =~ s/\`([^\`]*)\`/eval $1/gem; - - # if some assembler chokes on some simplified mnemonic, - # this is the spot to fix it up, e.g.: - # GNU as doesn't seem to accept cmplw, 32-bit unsigned compare - $data =~ s/^(\s*)cmplw(\s+)([^,]+),(.*)/$1cmpl$2$3,0,$4/gm; - # assembler X doesn't accept li, load immediate value - #$data =~ s/^(\s*)li(\s+)([^,]+),(.*)/$1addi$2$3,0,$4/gm; - return($data); -} diff --git a/src/lib/libcrypto/bn/asm/sparcv8.S b/src/lib/libcrypto/bn/asm/sparcv8.S deleted file mode 100644 index 88c5dc480a..0000000000 --- a/src/lib/libcrypto/bn/asm/sparcv8.S +++ /dev/null @@ -1,1458 +0,0 @@ -.ident "sparcv8.s, Version 1.4" -.ident "SPARC v8 ISA artwork by Andy Polyakov " - -/* - * ==================================================================== - * Written by Andy Polyakov for the OpenSSL - * project. - * - * Rights for redistribution and usage in source and binary forms are - * granted according to the OpenSSL license. Warranty of any kind is - * disclaimed. - * ==================================================================== - */ - -/* - * This is my modest contributon to OpenSSL project (see - * http://www.openssl.org/ for more information about it) and is - * a drop-in SuperSPARC ISA replacement for crypto/bn/bn_asm.c - * module. For updates see http://fy.chalmers.se/~appro/hpe/. - * - * See bn_asm.sparc.v8plus.S for more details. - */ - -/* - * Revision history. - * - * 1.1 - new loop unrolling model(*); - * 1.2 - made gas friendly; - * 1.3 - fixed problem with /usr/ccs/lib/cpp; - * 1.4 - some retunes; - * - * (*) see bn_asm.sparc.v8plus.S for details - */ - -.section ".text",#alloc,#execinstr -.file "bn_asm.sparc.v8.S" - -.align 32 - -.global bn_mul_add_words -/* - * BN_ULONG bn_mul_add_words(rp,ap,num,w) - * BN_ULONG *rp,*ap; - * int num; - * BN_ULONG w; - */ -bn_mul_add_words: - cmp %o2,0 - bg,a .L_bn_mul_add_words_proceed - ld [%o1],%g2 - retl - clr %o0 - -.L_bn_mul_add_words_proceed: - andcc %o2,-4,%g0 - bz .L_bn_mul_add_words_tail - clr %o5 - -.L_bn_mul_add_words_loop: - ld [%o0],%o4 - ld [%o1+4],%g3 - umul %o3,%g2,%g2 - rd %y,%g1 - addcc %o4,%o5,%o4 - addx %g1,0,%g1 - addcc %o4,%g2,%o4 - st %o4,[%o0] - addx %g1,0,%o5 - - ld [%o0+4],%o4 - ld [%o1+8],%g2 - umul %o3,%g3,%g3 - dec 4,%o2 - rd %y,%g1 - addcc %o4,%o5,%o4 - addx %g1,0,%g1 - addcc %o4,%g3,%o4 - st %o4,[%o0+4] - addx %g1,0,%o5 - - ld [%o0+8],%o4 - ld [%o1+12],%g3 - umul %o3,%g2,%g2 - inc 16,%o1 - rd %y,%g1 - addcc %o4,%o5,%o4 - addx %g1,0,%g1 - addcc %o4,%g2,%o4 - st %o4,[%o0+8] - addx %g1,0,%o5 - - ld [%o0+12],%o4 - umul %o3,%g3,%g3 - inc 16,%o0 - rd %y,%g1 - addcc %o4,%o5,%o4 - addx %g1,0,%g1 - addcc %o4,%g3,%o4 - st %o4,[%o0-4] - addx %g1,0,%o5 - andcc %o2,-4,%g0 - bnz,a .L_bn_mul_add_words_loop - ld [%o1],%g2 - - tst %o2 - bnz,a .L_bn_mul_add_words_tail - ld [%o1],%g2 -.L_bn_mul_add_words_return: - retl - mov %o5,%o0 - nop - -.L_bn_mul_add_words_tail: - ld [%o0],%o4 - umul %o3,%g2,%g2 - addcc %o4,%o5,%o4 - rd %y,%g1 - addx %g1,0,%g1 - addcc %o4,%g2,%o4 - addx %g1,0,%o5 - deccc %o2 - bz .L_bn_mul_add_words_return - st %o4,[%o0] - - ld [%o1+4],%g2 - ld [%o0+4],%o4 - umul %o3,%g2,%g2 - rd %y,%g1 - addcc %o4,%o5,%o4 - addx %g1,0,%g1 - addcc %o4,%g2,%o4 - addx %g1,0,%o5 - deccc %o2 - bz .L_bn_mul_add_words_return - st %o4,[%o0+4] - - ld [%o1+8],%g2 - ld [%o0+8],%o4 - umul %o3,%g2,%g2 - rd %y,%g1 - addcc %o4,%o5,%o4 - addx %g1,0,%g1 - addcc %o4,%g2,%o4 - st %o4,[%o0+8] - retl - addx %g1,0,%o0 - -.type bn_mul_add_words,#function -.size bn_mul_add_words,(.-bn_mul_add_words) - -.align 32 - -.global bn_mul_words -/* - * BN_ULONG bn_mul_words(rp,ap,num,w) - * BN_ULONG *rp,*ap; - * int num; - * BN_ULONG w; - */ -bn_mul_words: - cmp %o2,0 - bg,a .L_bn_mul_words_proceeed - ld [%o1],%g2 - retl - clr %o0 - -.L_bn_mul_words_proceeed: - andcc %o2,-4,%g0 - bz .L_bn_mul_words_tail - clr %o5 - -.L_bn_mul_words_loop: - ld [%o1+4],%g3 - umul %o3,%g2,%g2 - addcc %g2,%o5,%g2 - rd %y,%g1 - addx %g1,0,%o5 - st %g2,[%o0] - - ld [%o1+8],%g2 - umul %o3,%g3,%g3 - addcc %g3,%o5,%g3 - rd %y,%g1 - dec 4,%o2 - addx %g1,0,%o5 - st %g3,[%o0+4] - - ld [%o1+12],%g3 - umul %o3,%g2,%g2 - addcc %g2,%o5,%g2 - rd %y,%g1 - inc 16,%o1 - st %g2,[%o0+8] - addx %g1,0,%o5 - - umul %o3,%g3,%g3 - addcc %g3,%o5,%g3 - rd %y,%g1 - inc 16,%o0 - addx %g1,0,%o5 - st %g3,[%o0-4] - andcc %o2,-4,%g0 - nop - bnz,a .L_bn_mul_words_loop - ld [%o1],%g2 - - tst %o2 - bnz,a .L_bn_mul_words_tail - ld [%o1],%g2 -.L_bn_mul_words_return: - retl - mov %o5,%o0 - nop - -.L_bn_mul_words_tail: - umul %o3,%g2,%g2 - addcc %g2,%o5,%g2 - rd %y,%g1 - addx %g1,0,%o5 - deccc %o2 - bz .L_bn_mul_words_return - st %g2,[%o0] - nop - - ld [%o1+4],%g2 - umul %o3,%g2,%g2 - addcc %g2,%o5,%g2 - rd %y,%g1 - addx %g1,0,%o5 - deccc %o2 - bz .L_bn_mul_words_return - st %g2,[%o0+4] - - ld [%o1+8],%g2 - umul %o3,%g2,%g2 - addcc %g2,%o5,%g2 - rd %y,%g1 - st %g2,[%o0+8] - retl - addx %g1,0,%o0 - -.type bn_mul_words,#function -.size bn_mul_words,(.-bn_mul_words) - -.align 32 -.global bn_sqr_words -/* - * void bn_sqr_words(r,a,n) - * BN_ULONG *r,*a; - * int n; - */ -bn_sqr_words: - cmp %o2,0 - bg,a .L_bn_sqr_words_proceeed - ld [%o1],%g2 - retl - clr %o0 - -.L_bn_sqr_words_proceeed: - andcc %o2,-4,%g0 - bz .L_bn_sqr_words_tail - clr %o5 - -.L_bn_sqr_words_loop: - ld [%o1+4],%g3 - umul %g2,%g2,%o4 - st %o4,[%o0] - rd %y,%o5 - st %o5,[%o0+4] - - ld [%o1+8],%g2 - umul %g3,%g3,%o4 - dec 4,%o2 - st %o4,[%o0+8] - rd %y,%o5 - st %o5,[%o0+12] - nop - - ld [%o1+12],%g3 - umul %g2,%g2,%o4 - st %o4,[%o0+16] - rd %y,%o5 - inc 16,%o1 - st %o5,[%o0+20] - - umul %g3,%g3,%o4 - inc 32,%o0 - st %o4,[%o0-8] - rd %y,%o5 - st %o5,[%o0-4] - andcc %o2,-4,%g2 - bnz,a .L_bn_sqr_words_loop - ld [%o1],%g2 - - tst %o2 - nop - bnz,a .L_bn_sqr_words_tail - ld [%o1],%g2 -.L_bn_sqr_words_return: - retl - clr %o0 - -.L_bn_sqr_words_tail: - umul %g2,%g2,%o4 - st %o4,[%o0] - deccc %o2 - rd %y,%o5 - bz .L_bn_sqr_words_return - st %o5,[%o0+4] - - ld [%o1+4],%g2 - umul %g2,%g2,%o4 - st %o4,[%o0+8] - deccc %o2 - rd %y,%o5 - nop - bz .L_bn_sqr_words_return - st %o5,[%o0+12] - - ld [%o1+8],%g2 - umul %g2,%g2,%o4 - st %o4,[%o0+16] - rd %y,%o5 - st %o5,[%o0+20] - retl - clr %o0 - -.type bn_sqr_words,#function -.size bn_sqr_words,(.-bn_sqr_words) - -.align 32 - -.global bn_div_words -/* - * BN_ULONG bn_div_words(h,l,d) - * BN_ULONG h,l,d; - */ -bn_div_words: - wr %o0,%y - udiv %o1,%o2,%o0 - retl - nop - -.type bn_div_words,#function -.size bn_div_words,(.-bn_div_words) - -.align 32 - -.global bn_add_words -/* - * BN_ULONG bn_add_words(rp,ap,bp,n) - * BN_ULONG *rp,*ap,*bp; - * int n; - */ -bn_add_words: - cmp %o3,0 - bg,a .L_bn_add_words_proceed - ld [%o1],%o4 - retl - clr %o0 - -.L_bn_add_words_proceed: - andcc %o3,-4,%g0 - bz .L_bn_add_words_tail - clr %g1 - ba .L_bn_add_words_warn_loop - addcc %g0,0,%g0 ! clear carry flag - -.L_bn_add_words_loop: - ld [%o1],%o4 -.L_bn_add_words_warn_loop: - ld [%o2],%o5 - ld [%o1+4],%g3 - ld [%o2+4],%g4 - dec 4,%o3 - addxcc %o5,%o4,%o5 - st %o5,[%o0] - - ld [%o1+8],%o4 - ld [%o2+8],%o5 - inc 16,%o1 - addxcc %g3,%g4,%g3 - st %g3,[%o0+4] - - ld [%o1-4],%g3 - ld [%o2+12],%g4 - inc 16,%o2 - addxcc %o5,%o4,%o5 - st %o5,[%o0+8] - - inc 16,%o0 - addxcc %g3,%g4,%g3 - st %g3,[%o0-4] - addx %g0,0,%g1 - andcc %o3,-4,%g0 - bnz,a .L_bn_add_words_loop - addcc %g1,-1,%g0 - - tst %o3 - bnz,a .L_bn_add_words_tail - ld [%o1],%o4 -.L_bn_add_words_return: - retl - mov %g1,%o0 - -.L_bn_add_words_tail: - addcc %g1,-1,%g0 - ld [%o2],%o5 - addxcc %o5,%o4,%o5 - addx %g0,0,%g1 - deccc %o3 - bz .L_bn_add_words_return - st %o5,[%o0] - - ld [%o1+4],%o4 - addcc %g1,-1,%g0 - ld [%o2+4],%o5 - addxcc %o5,%o4,%o5 - addx %g0,0,%g1 - deccc %o3 - bz .L_bn_add_words_return - st %o5,[%o0+4] - - ld [%o1+8],%o4 - addcc %g1,-1,%g0 - ld [%o2+8],%o5 - addxcc %o5,%o4,%o5 - st %o5,[%o0+8] - retl - addx %g0,0,%o0 - -.type bn_add_words,#function -.size bn_add_words,(.-bn_add_words) - -.align 32 - -.global bn_sub_words -/* - * BN_ULONG bn_sub_words(rp,ap,bp,n) - * BN_ULONG *rp,*ap,*bp; - * int n; - */ -bn_sub_words: - cmp %o3,0 - bg,a .L_bn_sub_words_proceed - ld [%o1],%o4 - retl - clr %o0 - -.L_bn_sub_words_proceed: - andcc %o3,-4,%g0 - bz .L_bn_sub_words_tail - clr %g1 - ba .L_bn_sub_words_warm_loop - addcc %g0,0,%g0 ! clear carry flag - -.L_bn_sub_words_loop: - ld [%o1],%o4 -.L_bn_sub_words_warm_loop: - ld [%o2],%o5 - ld [%o1+4],%g3 - ld [%o2+4],%g4 - dec 4,%o3 - subxcc %o4,%o5,%o5 - st %o5,[%o0] - - ld [%o1+8],%o4 - ld [%o2+8],%o5 - inc 16,%o1 - subxcc %g3,%g4,%g4 - st %g4,[%o0+4] - - ld [%o1-4],%g3 - ld [%o2+12],%g4 - inc 16,%o2 - subxcc %o4,%o5,%o5 - st %o5,[%o0+8] - - inc 16,%o0 - subxcc %g3,%g4,%g4 - st %g4,[%o0-4] - addx %g0,0,%g1 - andcc %o3,-4,%g0 - bnz,a .L_bn_sub_words_loop - addcc %g1,-1,%g0 - - tst %o3 - nop - bnz,a .L_bn_sub_words_tail - ld [%o1],%o4 -.L_bn_sub_words_return: - retl - mov %g1,%o0 - -.L_bn_sub_words_tail: - addcc %g1,-1,%g0 - ld [%o2],%o5 - subxcc %o4,%o5,%o5 - addx %g0,0,%g1 - deccc %o3 - bz .L_bn_sub_words_return - st %o5,[%o0] - nop - - ld [%o1+4],%o4 - addcc %g1,-1,%g0 - ld [%o2+4],%o5 - subxcc %o4,%o5,%o5 - addx %g0,0,%g1 - deccc %o3 - bz .L_bn_sub_words_return - st %o5,[%o0+4] - - ld [%o1+8],%o4 - addcc %g1,-1,%g0 - ld [%o2+8],%o5 - subxcc %o4,%o5,%o5 - st %o5,[%o0+8] - retl - addx %g0,0,%o0 - -.type bn_sub_words,#function -.size bn_sub_words,(.-bn_sub_words) - -#define FRAME_SIZE -96 - -/* - * Here is register usage map for *all* routines below. - */ -#define t_1 %o0 -#define t_2 %o1 -#define c_1 %o2 -#define c_2 %o3 -#define c_3 %o4 - -#define ap(I) [%i1+4*I] -#define bp(I) [%i2+4*I] -#define rp(I) [%i0+4*I] - -#define a_0 %l0 -#define a_1 %l1 -#define a_2 %l2 -#define a_3 %l3 -#define a_4 %l4 -#define a_5 %l5 -#define a_6 %l6 -#define a_7 %l7 - -#define b_0 %i3 -#define b_1 %i4 -#define b_2 %i5 -#define b_3 %o5 -#define b_4 %g1 -#define b_5 %g2 -#define b_6 %g3 -#define b_7 %g4 - -.align 32 -.global bn_mul_comba8 -/* - * void bn_mul_comba8(r,a,b) - * BN_ULONG *r,*a,*b; - */ -bn_mul_comba8: - save %sp,FRAME_SIZE,%sp - ld ap(0),a_0 - ld bp(0),b_0 - umul a_0,b_0,c_1 !=!mul_add_c(a[0],b[0],c1,c2,c3); - ld bp(1),b_1 - rd %y,c_2 - st c_1,rp(0) !r[0]=c1; - - umul a_0,b_1,t_1 !=!mul_add_c(a[0],b[1],c2,c3,c1); - ld ap(1),a_1 - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc %g0,t_2,c_3 != - addx %g0,%g0,c_1 - ld ap(2),a_2 - umul a_1,b_0,t_1 !mul_add_c(a[1],b[0],c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - st c_2,rp(1) !r[1]=c2; - addx c_1,%g0,c_1 != - - umul a_2,b_0,t_1 !mul_add_c(a[2],b[0],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx %g0,%g0,c_2 - ld bp(2),b_2 - umul a_1,b_1,t_1 !mul_add_c(a[1],b[1],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - ld bp(3),b_3 - addx c_2,%g0,c_2 != - umul a_0,b_2,t_1 !mul_add_c(a[0],b[2],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - st c_3,rp(2) !r[2]=c3; - - umul a_0,b_3,t_1 !mul_add_c(a[0],b[3],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx %g0,%g0,c_3 - umul a_1,b_2,t_1 !=!mul_add_c(a[1],b[2],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - ld ap(3),a_3 - umul a_2,b_1,t_1 !mul_add_c(a[2],b[1],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - ld ap(4),a_4 - umul a_3,b_0,t_1 !mul_add_c(a[3],b[0],c1,c2,c3);!= - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(3) !r[3]=c1; - - umul a_4,b_0,t_1 !mul_add_c(a[4],b[0],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - umul a_3,b_1,t_1 !mul_add_c(a[3],b[1],c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - umul a_2,b_2,t_1 !=!mul_add_c(a[2],b[2],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - ld bp(4),b_4 - umul a_1,b_3,t_1 !mul_add_c(a[1],b[3],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - ld bp(5),b_5 - umul a_0,b_4,t_1 !=!mul_add_c(a[0],b[4],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - st c_2,rp(4) !r[4]=c2; - - umul a_0,b_5,t_1 !mul_add_c(a[0],b[5],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 - umul a_1,b_4,t_1 !mul_add_c(a[1],b[4],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_2,b_3,t_1 !=!mul_add_c(a[2],b[3],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - umul a_3,b_2,t_1 !mul_add_c(a[3],b[2],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - ld ap(5),a_5 - umul a_4,b_1,t_1 !mul_add_c(a[4],b[1],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - ld ap(6),a_6 - addx c_2,%g0,c_2 != - umul a_5,b_0,t_1 !mul_add_c(a[5],b[0],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - st c_3,rp(5) !r[5]=c3; - - umul a_6,b_0,t_1 !mul_add_c(a[6],b[0],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx %g0,%g0,c_3 - umul a_5,b_1,t_1 !=!mul_add_c(a[5],b[1],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - umul a_4,b_2,t_1 !mul_add_c(a[4],b[2],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - umul a_3,b_3,t_1 !mul_add_c(a[3],b[3],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_2,b_4,t_1 !mul_add_c(a[2],b[4],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - ld bp(6),b_6 - addx c_3,%g0,c_3 != - umul a_1,b_5,t_1 !mul_add_c(a[1],b[5],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - ld bp(7),b_7 - umul a_0,b_6,t_1 !mul_add_c(a[0],b[6],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - st c_1,rp(6) !r[6]=c1; - addx c_3,%g0,c_3 != - - umul a_0,b_7,t_1 !mul_add_c(a[0],b[7],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 != - addx %g0,%g0,c_1 - umul a_1,b_6,t_1 !mul_add_c(a[1],b[6],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - umul a_2,b_5,t_1 !mul_add_c(a[2],b[5],c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - umul a_3,b_4,t_1 !=!mul_add_c(a[3],b[4],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - umul a_4,b_3,t_1 !mul_add_c(a[4],b[3],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_5,b_2,t_1 !mul_add_c(a[5],b[2],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - ld ap(7),a_7 - umul a_6,b_1,t_1 !=!mul_add_c(a[6],b[1],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - umul a_7,b_0,t_1 !mul_add_c(a[7],b[0],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - st c_2,rp(7) !r[7]=c2; - - umul a_7,b_1,t_1 !mul_add_c(a[7],b[1],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 - umul a_6,b_2,t_1 !=!mul_add_c(a[6],b[2],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - umul a_5,b_3,t_1 !mul_add_c(a[5],b[3],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - umul a_4,b_4,t_1 !mul_add_c(a[4],b[4],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_3,b_5,t_1 !mul_add_c(a[3],b[5],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_2,b_6,t_1 !=!mul_add_c(a[2],b[6],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - umul a_1,b_7,t_1 !mul_add_c(a[1],b[7],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 ! - addx c_2,%g0,c_2 - st c_3,rp(8) !r[8]=c3; - - umul a_2,b_7,t_1 !mul_add_c(a[2],b[7],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx %g0,%g0,c_3 - umul a_3,b_6,t_1 !=!mul_add_c(a[3],b[6],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - umul a_4,b_5,t_1 !mul_add_c(a[4],b[5],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - umul a_5,b_4,t_1 !mul_add_c(a[5],b[4],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_6,b_3,t_1 !mul_add_c(a[6],b[3],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_7,b_2,t_1 !=!mul_add_c(a[7],b[2],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(9) !r[9]=c1; - - umul a_7,b_3,t_1 !mul_add_c(a[7],b[3],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - umul a_6,b_4,t_1 !mul_add_c(a[6],b[4],c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - umul a_5,b_5,t_1 !=!mul_add_c(a[5],b[5],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - umul a_4,b_6,t_1 !mul_add_c(a[4],b[6],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_3,b_7,t_1 !mul_add_c(a[3],b[7],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - st c_2,rp(10) !r[10]=c2; - - umul a_4,b_7,t_1 !=!mul_add_c(a[4],b[7],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 != - umul a_5,b_6,t_1 !mul_add_c(a[5],b[6],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - umul a_6,b_5,t_1 !mul_add_c(a[6],b[5],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_7,b_4,t_1 !mul_add_c(a[7],b[4],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - st c_3,rp(11) !r[11]=c3; - addx c_2,%g0,c_2 != - - umul a_7,b_5,t_1 !mul_add_c(a[7],b[5],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx %g0,%g0,c_3 - umul a_6,b_6,t_1 !mul_add_c(a[6],b[6],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_5,b_7,t_1 !mul_add_c(a[5],b[7],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - st c_1,rp(12) !r[12]=c1; - addx c_3,%g0,c_3 != - - umul a_6,b_7,t_1 !mul_add_c(a[6],b[7],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 != - addx %g0,%g0,c_1 - umul a_7,b_6,t_1 !mul_add_c(a[7],b[6],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - st c_2,rp(13) !r[13]=c2; - - umul a_7,b_7,t_1 !=!mul_add_c(a[7],b[7],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - nop != - st c_3,rp(14) !r[14]=c3; - st c_1,rp(15) !r[15]=c1; - - ret - restore %g0,%g0,%o0 - -.type bn_mul_comba8,#function -.size bn_mul_comba8,(.-bn_mul_comba8) - -.align 32 - -.global bn_mul_comba4 -/* - * void bn_mul_comba4(r,a,b) - * BN_ULONG *r,*a,*b; - */ -bn_mul_comba4: - save %sp,FRAME_SIZE,%sp - ld ap(0),a_0 - ld bp(0),b_0 - umul a_0,b_0,c_1 !=!mul_add_c(a[0],b[0],c1,c2,c3); - ld bp(1),b_1 - rd %y,c_2 - st c_1,rp(0) !r[0]=c1; - - umul a_0,b_1,t_1 !=!mul_add_c(a[0],b[1],c2,c3,c1); - ld ap(1),a_1 - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc %g0,t_2,c_3 - addx %g0,%g0,c_1 - ld ap(2),a_2 - umul a_1,b_0,t_1 !=!mul_add_c(a[1],b[0],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - st c_2,rp(1) !r[1]=c2; - - umul a_2,b_0,t_1 !mul_add_c(a[2],b[0],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 - ld bp(2),b_2 - umul a_1,b_1,t_1 !=!mul_add_c(a[1],b[1],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - ld bp(3),b_3 - umul a_0,b_2,t_1 !mul_add_c(a[0],b[2],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - st c_3,rp(2) !r[2]=c3; - - umul a_0,b_3,t_1 !=!mul_add_c(a[0],b[3],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx %g0,%g0,c_3 != - umul a_1,b_2,t_1 !mul_add_c(a[1],b[2],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - ld ap(3),a_3 - umul a_2,b_1,t_1 !mul_add_c(a[2],b[1],c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_3,b_0,t_1 !=!mul_add_c(a[3],b[0],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(3) !r[3]=c1; - - umul a_3,b_1,t_1 !mul_add_c(a[3],b[1],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - umul a_2,b_2,t_1 !mul_add_c(a[2],b[2],c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - umul a_1,b_3,t_1 !=!mul_add_c(a[1],b[3],c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - st c_2,rp(4) !r[4]=c2; - - umul a_2,b_3,t_1 !mul_add_c(a[2],b[3],c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 - umul a_3,b_2,t_1 !mul_add_c(a[3],b[2],c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - st c_3,rp(5) !r[5]=c3; - addx c_2,%g0,c_2 != - - umul a_3,b_3,t_1 !mul_add_c(a[3],b[3],c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - st c_1,rp(6) !r[6]=c1; - st c_2,rp(7) !r[7]=c2; - - ret - restore %g0,%g0,%o0 - -.type bn_mul_comba4,#function -.size bn_mul_comba4,(.-bn_mul_comba4) - -.align 32 - -.global bn_sqr_comba8 -bn_sqr_comba8: - save %sp,FRAME_SIZE,%sp - ld ap(0),a_0 - ld ap(1),a_1 - umul a_0,a_0,c_1 !=!sqr_add_c(a,0,c1,c2,c3); - rd %y,c_2 - st c_1,rp(0) !r[0]=c1; - - ld ap(2),a_2 - umul a_0,a_1,t_1 !=!sqr_add_c2(a,1,0,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc %g0,t_2,c_3 - addx %g0,%g0,c_1 != - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 - st c_2,rp(1) !r[1]=c2; - addx c_1,%g0,c_1 != - - umul a_2,a_0,t_1 !sqr_add_c2(a,2,0,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx %g0,%g0,c_2 - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - ld ap(3),a_3 - umul a_1,a_1,t_1 !sqr_add_c(a,1,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - st c_3,rp(2) !r[2]=c3; - - umul a_0,a_3,t_1 !=!sqr_add_c2(a,3,0,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx %g0,%g0,c_3 != - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - ld ap(4),a_4 - addx c_3,%g0,c_3 != - umul a_1,a_2,t_1 !sqr_add_c2(a,2,1,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(3) !r[3]=c1; - - umul a_4,a_0,t_1 !sqr_add_c2(a,4,0,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_3,a_1,t_1 !sqr_add_c2(a,3,1,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - ld ap(5),a_5 - umul a_2,a_2,t_1 !sqr_add_c(a,2,c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - st c_2,rp(4) !r[4]=c2; - addx c_1,%g0,c_1 != - - umul a_0,a_5,t_1 !sqr_add_c2(a,5,0,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx %g0,%g0,c_2 - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - umul a_1,a_4,t_1 !sqr_add_c2(a,4,1,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - ld ap(6),a_6 - umul a_2,a_3,t_1 !sqr_add_c2(a,3,2,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - st c_3,rp(5) !r[5]=c3; - - umul a_6,a_0,t_1 !sqr_add_c2(a,6,0,c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx %g0,%g0,c_3 - addcc c_1,t_1,c_1 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_5,a_1,t_1 !sqr_add_c2(a,5,1,c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - addcc c_1,t_1,c_1 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - umul a_4,a_2,t_1 !sqr_add_c2(a,4,2,c1,c2,c3); - addcc c_1,t_1,c_1 != - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - addcc c_1,t_1,c_1 != - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 - ld ap(7),a_7 - umul a_3,a_3,t_1 !=!sqr_add_c(a,3,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(6) !r[6]=c1; - - umul a_0,a_7,t_1 !sqr_add_c2(a,7,0,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_1,a_6,t_1 !sqr_add_c2(a,6,1,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_2,a_5,t_1 !sqr_add_c2(a,5,2,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_3,a_4,t_1 !sqr_add_c2(a,4,3,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - st c_2,rp(7) !r[7]=c2; - - umul a_7,a_1,t_1 !sqr_add_c2(a,7,1,c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 - addcc c_3,t_1,c_3 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_6,a_2,t_1 !sqr_add_c2(a,6,2,c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - addcc c_3,t_1,c_3 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_5,a_3,t_1 !sqr_add_c2(a,5,3,c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - addcc c_3,t_1,c_3 != - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_4,a_4,t_1 !sqr_add_c(a,4,c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - st c_3,rp(8) !r[8]=c3; - addx c_2,%g0,c_2 != - - umul a_2,a_7,t_1 !sqr_add_c2(a,7,2,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx %g0,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - umul a_3,a_6,t_1 !sqr_add_c2(a,6,3,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - umul a_4,a_5,t_1 !sqr_add_c2(a,5,4,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(9) !r[9]=c1; - - umul a_7,a_3,t_1 !sqr_add_c2(a,7,3,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_6,a_4,t_1 !sqr_add_c2(a,6,4,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_5,a_5,t_1 !sqr_add_c(a,5,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - st c_2,rp(10) !r[10]=c2; - - umul a_4,a_7,t_1 !=!sqr_add_c2(a,7,4,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 != - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 - umul a_5,a_6,t_1 !=!sqr_add_c2(a,6,5,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx c_2,%g0,c_2 != - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 - st c_3,rp(11) !r[11]=c3; - addx c_2,%g0,c_2 != - - umul a_7,a_5,t_1 !sqr_add_c2(a,7,5,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx %g0,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - umul a_6,a_6,t_1 !sqr_add_c(a,6,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - st c_1,rp(12) !r[12]=c1; - - umul a_6,a_7,t_1 !sqr_add_c2(a,7,6,c2,c3,c1); - addcc c_2,t_1,c_2 != - rd %y,t_2 - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - addcc c_2,t_1,c_2 != - addxcc c_3,t_2,c_3 - st c_2,rp(13) !r[13]=c2; - addx c_1,%g0,c_1 != - - umul a_7,a_7,t_1 !sqr_add_c(a,7,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 != - st c_3,rp(14) !r[14]=c3; - st c_1,rp(15) !r[15]=c1; - - ret - restore %g0,%g0,%o0 - -.type bn_sqr_comba8,#function -.size bn_sqr_comba8,(.-bn_sqr_comba8) - -.align 32 - -.global bn_sqr_comba4 -/* - * void bn_sqr_comba4(r,a) - * BN_ULONG *r,*a; - */ -bn_sqr_comba4: - save %sp,FRAME_SIZE,%sp - ld ap(0),a_0 - umul a_0,a_0,c_1 !sqr_add_c(a,0,c1,c2,c3); - ld ap(1),a_1 != - rd %y,c_2 - st c_1,rp(0) !r[0]=c1; - - ld ap(2),a_2 - umul a_0,a_1,t_1 !=!sqr_add_c2(a,1,0,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 - addxcc %g0,t_2,c_3 - addx %g0,%g0,c_1 != - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 != - st c_2,rp(1) !r[1]=c2; - - umul a_2,a_0,t_1 !sqr_add_c2(a,2,0,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 != - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 != - addx c_2,%g0,c_2 - ld ap(3),a_3 - umul a_1,a_1,t_1 !sqr_add_c(a,1,c3,c1,c2); - addcc c_3,t_1,c_3 != - rd %y,t_2 - addxcc c_1,t_2,c_1 - st c_3,rp(2) !r[2]=c3; - addx c_2,%g0,c_2 != - - umul a_0,a_3,t_1 !sqr_add_c2(a,3,0,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx %g0,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - umul a_1,a_2,t_1 !sqr_add_c2(a,2,1,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - addx c_3,%g0,c_3 - addcc c_1,t_1,c_1 - addxcc c_2,t_2,c_2 - addx c_3,%g0,c_3 != - st c_1,rp(3) !r[3]=c1; - - umul a_3,a_1,t_1 !sqr_add_c2(a,3,1,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx %g0,%g0,c_1 - addcc c_2,t_1,c_2 - addxcc c_3,t_2,c_3 != - addx c_1,%g0,c_1 - umul a_2,a_2,t_1 !sqr_add_c(a,2,c2,c3,c1); - addcc c_2,t_1,c_2 - rd %y,t_2 != - addxcc c_3,t_2,c_3 - addx c_1,%g0,c_1 - st c_2,rp(4) !r[4]=c2; - - umul a_2,a_3,t_1 !=!sqr_add_c2(a,3,2,c3,c1,c2); - addcc c_3,t_1,c_3 - rd %y,t_2 - addxcc c_1,t_2,c_1 - addx %g0,%g0,c_2 != - addcc c_3,t_1,c_3 - addxcc c_1,t_2,c_1 - st c_3,rp(5) !r[5]=c3; - addx c_2,%g0,c_2 != - - umul a_3,a_3,t_1 !sqr_add_c(a,3,c1,c2,c3); - addcc c_1,t_1,c_1 - rd %y,t_2 - addxcc c_2,t_2,c_2 != - st c_1,rp(6) !r[6]=c1; - st c_2,rp(7) !r[7]=c2; - - ret - restore %g0,%g0,%o0 - -.type bn_sqr_comba4,#function -.size bn_sqr_comba4,(.-bn_sqr_comba4) - -.align 32 diff --git a/src/lib/libcrypto/bn/asm/sparcv8plus.S b/src/lib/libcrypto/bn/asm/sparcv8plus.S deleted file mode 100644 index 8c56e2e7e7..0000000000 --- a/src/lib/libcrypto/bn/asm/sparcv8plus.S +++ /dev/null @@ -1,1547 +0,0 @@ -.ident "sparcv8plus.s, Version 1.4" -.ident "SPARC v9 ISA artwork by Andy Polyakov " - -/* - * ==================================================================== - * Written by Andy Polyakov for the OpenSSL - * project. - * - * Rights for redistribution and usage in source and binary forms are - * granted according to the OpenSSL license. Warranty of any kind is - * disclaimed. - * ==================================================================== - */ - -/* - * This is my modest contributon to OpenSSL project (see - * http://www.openssl.org/ for more information about it) and is - * a drop-in UltraSPARC ISA replacement for crypto/bn/bn_asm.c - * module. For updates see http://fy.chalmers.se/~appro/hpe/. - * - * Questions-n-answers. - * - * Q. How to compile? - * A. With SC4.x/SC5.x: - * - * cc -xarch=v8plus -c bn_asm.sparc.v8plus.S -o bn_asm.o - * - * and with gcc: - * - * gcc -mcpu=ultrasparc -c bn_asm.sparc.v8plus.S -o bn_asm.o - * - * or if above fails (it does if you have gas installed): - * - * gcc -E bn_asm.sparc.v8plus.S | as -xarch=v8plus /dev/fd/0 -o bn_asm.o - * - * Quick-n-dirty way to fuse the module into the library. - * Provided that the library is already configured and built - * (in 0.9.2 case with no-asm option): - * - * # cd crypto/bn - * # cp /some/place/bn_asm.sparc.v8plus.S . - * # cc -xarch=v8plus -c bn_asm.sparc.v8plus.S -o bn_asm.o - * # make - * # cd ../.. - * # make; make test - * - * Quick-n-dirty way to get rid of it: - * - * # cd crypto/bn - * # touch bn_asm.c - * # make - * # cd ../.. - * # make; make test - * - * Q. V8plus achitecture? What kind of beast is that? - * A. Well, it's rather a programming model than an architecture... - * It's actually v9-compliant, i.e. *any* UltraSPARC, CPU under - * special conditions, namely when kernel doesn't preserve upper - * 32 bits of otherwise 64-bit registers during a context switch. - * - * Q. Why just UltraSPARC? What about SuperSPARC? - * A. Original release did target UltraSPARC only. Now SuperSPARC - * version is provided along. Both version share bn_*comba[48] - * implementations (see comment later in code for explanation). - * But what's so special about this UltraSPARC implementation? - * Why didn't I let compiler do the job? Trouble is that most of - * available compilers (well, SC5.0 is the only exception) don't - * attempt to take advantage of UltraSPARC's 64-bitness under - * 32-bit kernels even though it's perfectly possible (see next - * question). - * - * Q. 64-bit registers under 32-bit kernels? Didn't you just say it - * doesn't work? - * A. You can't adress *all* registers as 64-bit wide:-( The catch is - * that you actually may rely upon %o0-%o5 and %g1-%g4 being fully - * preserved if you're in a leaf function, i.e. such never calling - * any other functions. All functions in this module are leaf and - * 10 registers is a handful. And as a matter of fact none-"comba" - * routines don't require even that much and I could even afford to - * not allocate own stack frame for 'em:-) - * - * Q. What about 64-bit kernels? - * A. What about 'em? Just kidding:-) Pure 64-bit version is currently - * under evaluation and development... - * - * Q. What about shared libraries? - * A. What about 'em? Kidding again:-) Code does *not* contain any - * code position dependencies and it's safe to include it into - * shared library as is. - * - * Q. How much faster does it go? - * A. Do you have a good benchmark? In either case below is what I - * experience with crypto/bn/expspeed.c test program: - * - * v8plus module on U10/300MHz against bn_asm.c compiled with: - * - * cc-5.0 -xarch=v8plus -xO5 -xdepend +7-12% - * cc-4.2 -xarch=v8plus -xO5 -xdepend +25-35% - * egcs-1.1.2 -mcpu=ultrasparc -O3 +35-45% - * - * v8 module on SS10/60MHz against bn_asm.c compiled with: - * - * cc-5.0 -xarch=v8 -xO5 -xdepend +7-10% - * cc-4.2 -xarch=v8 -xO5 -xdepend +10% - * egcs-1.1.2 -mv8 -O3 +35-45% - * - * As you can see it's damn hard to beat the new Sun C compiler - * and it's in first place GNU C users who will appreciate this - * assembler implementation:-) - */ - -/* - * Revision history. - * - * 1.0 - initial release; - * 1.1 - new loop unrolling model(*); - * - some more fine tuning; - * 1.2 - made gas friendly; - * - updates to documentation concerning v9; - * - new performance comparison matrix; - * 1.3 - fixed problem with /usr/ccs/lib/cpp; - * 1.4 - native V9 bn_*_comba[48] implementation (15% more efficient) - * resulting in slight overall performance kick; - * - some retunes; - * - support for GNU as added; - * - * (*) Originally unrolled loop looked like this: - * for (;;) { - * op(p+0); if (--n==0) break; - * op(p+1); if (--n==0) break; - * op(p+2); if (--n==0) break; - * op(p+3); if (--n==0) break; - * p+=4; - * } - * I unroll according to following: - * while (n&~3) { - * op(p+0); op(p+1); op(p+2); op(p+3); - * p+=4; n=-4; - * } - * if (n) { - * op(p+0); if (--n==0) return; - * op(p+2); if (--n==0) return; - * op(p+3); return; - * } - */ - -/* - * GNU assembler can't stand stuw:-( - */ -#define stuw st - -.section ".text",#alloc,#execinstr -.file "bn_asm.sparc.v8plus.S" - -.align 32 - -.global bn_mul_add_words -/* - * BN_ULONG bn_mul_add_words(rp,ap,num,w) - * BN_ULONG *rp,*ap; - * int num; - * BN_ULONG w; - */ -bn_mul_add_words: - sra %o2,%g0,%o2 ! signx %o2 - brgz,a %o2,.L_bn_mul_add_words_proceed - lduw [%o1],%g2 - retl - clr %o0 - nop - nop - nop - -.L_bn_mul_add_words_proceed: - srl %o3,%g0,%o3 ! clruw %o3 - andcc %o2,-4,%g0 - bz,pn %icc,.L_bn_mul_add_words_tail - clr %o5 - -.L_bn_mul_add_words_loop: ! wow! 32 aligned! - lduw [%o0],%g1 - lduw [%o1+4],%g3 - mulx %o3,%g2,%g2 - add %g1,%o5,%o4 - nop - add %o4,%g2,%o4 - stuw %o4,[%o0] - srlx %o4,32,%o5 - - lduw [%o0+4],%g1 - lduw [%o1+8],%g2 - mulx %o3,%g3,%g3 - add %g1,%o5,%o4 - dec 4,%o2 - add %o4,%g3,%o4 - stuw %o4,[%o0+4] - srlx %o4,32,%o5 - - lduw [%o0+8],%g1 - lduw [%o1+12],%g3 - mulx %o3,%g2,%g2 - add %g1,%o5,%o4 - inc 16,%o1 - add %o4,%g2,%o4 - stuw %o4,[%o0+8] - srlx %o4,32,%o5 - - lduw [%o0+12],%g1 - mulx %o3,%g3,%g3 - add %g1,%o5,%o4 - inc 16,%o0 - add %o4,%g3,%o4 - andcc %o2,-4,%g0 - stuw %o4,[%o0-4] - srlx %o4,32,%o5 - bnz,a,pt %icc,.L_bn_mul_add_words_loop - lduw [%o1],%g2 - - brnz,a,pn %o2,.L_bn_mul_add_words_tail - lduw [%o1],%g2 -.L_bn_mul_add_words_return: - retl - mov %o5,%o0 - -.L_bn_mul_add_words_tail: - lduw [%o0],%g1 - mulx %o3,%g2,%g2 - add %g1,%o5,%o4 - dec %o2 - add %o4,%g2,%o4 - srlx %o4,32,%o5 - brz,pt %o2,.L_bn_mul_add_words_return - stuw %o4,[%o0] - - lduw [%o1+4],%g2 - lduw [%o0+4],%g1 - mulx %o3,%g2,%g2 - add %g1,%o5,%o4 - dec %o2 - add %o4,%g2,%o4 - srlx %o4,32,%o5 - brz,pt %o2,.L_bn_mul_add_words_return - stuw %o4,[%o0+4] - - lduw [%o1+8],%g2 - lduw [%o0+8],%g1 - mulx %o3,%g2,%g2 - add %g1,%o5,%o4 - add %o4,%g2,%o4 - stuw %o4,[%o0+8] - retl - srlx %o4,32,%o0 - -.type bn_mul_add_words,#function -.size bn_mul_add_words,(.-bn_mul_add_words) - -.align 32 - -.global bn_mul_words -/* - * BN_ULONG bn_mul_words(rp,ap,num,w) - * BN_ULONG *rp,*ap; - * int num; - * BN_ULONG w; - */ -bn_mul_words: - sra %o2,%g0,%o2 ! signx %o2 - brgz,a %o2,.L_bn_mul_words_proceeed - lduw [%o1],%g2 - retl - clr %o0 - nop - nop - nop - -.L_bn_mul_words_proceeed: - srl %o3,%g0,%o3 ! clruw %o3 - andcc %o2,-4,%g0 - bz,pn %icc,.L_bn_mul_words_tail - clr %o5 - -.L_bn_mul_words_loop: ! wow! 32 aligned! - lduw [%o1+4],%g3 - mulx %o3,%g2,%g2 - add %g2,%o5,%o4 - nop - stuw %o4,[%o0] - srlx %o4,32,%o5 - - lduw [%o1+8],%g2 - mulx %o3,%g3,%g3 - add %g3,%o5,%o4 - dec 4,%o2 - stuw %o4,[%o0+4] - srlx %o4,32,%o5 - - lduw [%o1+12],%g3 - mulx %o3,%g2,%g2 - add %g2,%o5,%o4 - inc 16,%o1 - stuw %o4,[%o0+8] - srlx %o4,32,%o5 - - mulx %o3,%g3,%g3 - add %g3,%o5,%o4 - inc 16,%o0 - stuw %o4,[%o0-4] - srlx %o4,32,%o5 - andcc %o2,-4,%g0 - bnz,a,pt %icc,.L_bn_mul_words_loop - lduw [%o1],%g2 - nop - nop - - brnz,a,pn %o2,.L_bn_mul_words_tail - lduw [%o1],%g2 -.L_bn_mul_words_return: - retl - mov %o5,%o0 - -.L_bn_mul_words_tail: - mulx %o3,%g2,%g2 - add %g2,%o5,%o4 - dec %o2 - srlx %o4,32,%o5 - brz,pt %o2,.L_bn_mul_words_return - stuw %o4,[%o0] - - lduw [%o1+4],%g2 - mulx %o3,%g2,%g2 - add %g2,%o5,%o4 - dec %o2 - srlx %o4,32,%o5 - brz,pt %o2,.L_bn_mul_words_return - stuw %o4,[%o0+4] - - lduw [%o1+8],%g2 - mulx %o3,%g2,%g2 - add %g2,%o5,%o4 - stuw %o4,[%o0+8] - retl - srlx %o4,32,%o0 - -.type bn_mul_words,#function -.size bn_mul_words,(.-bn_mul_words) - -.align 32 -.global bn_sqr_words -/* - * void bn_sqr_words(r,a,n) - * BN_ULONG *r,*a; - * int n; - */ -bn_sqr_words: - sra %o2,%g0,%o2 ! signx %o2 - brgz,a %o2,.L_bn_sqr_words_proceeed - lduw [%o1],%g2 - retl - clr %o0 - nop - nop - nop - -.L_bn_sqr_words_proceeed: - andcc %o2,-4,%g0 - nop - bz,pn %icc,.L_bn_sqr_words_tail - nop - -.L_bn_sqr_words_loop: ! wow! 32 aligned! - lduw [%o1+4],%g3 - mulx %g2,%g2,%o4 - stuw %o4,[%o0] - srlx %o4,32,%o5 - stuw %o5,[%o0+4] - nop - - lduw [%o1+8],%g2 - mulx %g3,%g3,%o4 - dec 4,%o2 - stuw %o4,[%o0+8] - srlx %o4,32,%o5 - stuw %o5,[%o0+12] - - lduw [%o1+12],%g3 - mulx %g2,%g2,%o4 - srlx %o4,32,%o5 - stuw %o4,[%o0+16] - inc 16,%o1 - stuw %o5,[%o0+20] - - mulx %g3,%g3,%o4 - inc 32,%o0 - stuw %o4,[%o0-8] - srlx %o4,32,%o5 - andcc %o2,-4,%g2 - stuw %o5,[%o0-4] - bnz,a,pt %icc,.L_bn_sqr_words_loop - lduw [%o1],%g2 - nop - - brnz,a,pn %o2,.L_bn_sqr_words_tail - lduw [%o1],%g2 -.L_bn_sqr_words_return: - retl - clr %o0 - -.L_bn_sqr_words_tail: - mulx %g2,%g2,%o4 - dec %o2 - stuw %o4,[%o0] - srlx %o4,32,%o5 - brz,pt %o2,.L_bn_sqr_words_return - stuw %o5,[%o0+4] - - lduw [%o1+4],%g2 - mulx %g2,%g2,%o4 - dec %o2 - stuw %o4,[%o0+8] - srlx %o4,32,%o5 - brz,pt %o2,.L_bn_sqr_words_return - stuw %o5,[%o0+12] - - lduw [%o1+8],%g2 - mulx %g2,%g2,%o4 - srlx %o4,32,%o5 - stuw %o4,[%o0+16] - stuw %o5,[%o0+20] - retl - clr %o0 - -.type bn_sqr_words,#function -.size bn_sqr_words,(.-bn_sqr_words) - -.align 32 -.global bn_div_words -/* - * BN_ULONG bn_div_words(h,l,d) - * BN_ULONG h,l,d; - */ -bn_div_words: - sllx %o0,32,%o0 - or %o0,%o1,%o0 - udivx %o0,%o2,%o0 - retl - srl %o0,%g0,%o0 ! clruw %o0 - -.type bn_div_words,#function -.size bn_div_words,(.-bn_div_words) - -.align 32 - -.global bn_add_words -/* - * BN_ULONG bn_add_words(rp,ap,bp,n) - * BN_ULONG *rp,*ap,*bp; - * int n; - */ -bn_add_words: - sra %o3,%g0,%o3 ! signx %o3 - brgz,a %o3,.L_bn_add_words_proceed - lduw [%o1],%o4 - retl - clr %o0 - -.L_bn_add_words_proceed: - andcc %o3,-4,%g0 - bz,pn %icc,.L_bn_add_words_tail - addcc %g0,0,%g0 ! clear carry flag - -.L_bn_add_words_loop: ! wow! 32 aligned! - dec 4,%o3 - lduw [%o2],%o5 - lduw [%o1+4],%g1 - lduw [%o2+4],%g2 - lduw [%o1+8],%g3 - lduw [%o2+8],%g4 - addccc %o5,%o4,%o5 - stuw %o5,[%o0] - - lduw [%o1+12],%o4 - lduw [%o2+12],%o5 - inc 16,%o1 - addccc %g1,%g2,%g1 - stuw %g1,[%o0+4] - - inc 16,%o2 - addccc %g3,%g4,%g3 - stuw %g3,[%o0+8] - - inc 16,%o0 - addccc %o5,%o4,%o5 - stuw %o5,[%o0-4] - and %o3,-4,%g1 - brnz,a,pt %g1,.L_bn_add_words_loop - lduw [%o1],%o4 - - brnz,a,pn %o3,.L_bn_add_words_tail - lduw [%o1],%o4 -.L_bn_add_words_return: - clr %o0 - retl - movcs %icc,1,%o0 - nop - -.L_bn_add_words_tail: - lduw [%o2],%o5 - dec %o3 - addccc %o5,%o4,%o5 - brz,pt %o3,.L_bn_add_words_return - stuw %o5,[%o0] - - lduw [%o1+4],%o4 - lduw [%o2+4],%o5 - dec %o3 - addccc %o5,%o4,%o5 - brz,pt %o3,.L_bn_add_words_return - stuw %o5,[%o0+4] - - lduw [%o1+8],%o4 - lduw [%o2+8],%o5 - addccc %o5,%o4,%o5 - stuw %o5,[%o0+8] - clr %o0 - retl - movcs %icc,1,%o0 - -.type bn_add_words,#function -.size bn_add_words,(.-bn_add_words) - -.global bn_sub_words -/* - * BN_ULONG bn_sub_words(rp,ap,bp,n) - * BN_ULONG *rp,*ap,*bp; - * int n; - */ -bn_sub_words: - sra %o3,%g0,%o3 ! signx %o3 - brgz,a %o3,.L_bn_sub_words_proceed - lduw [%o1],%o4 - retl - clr %o0 - -.L_bn_sub_words_proceed: - andcc %o3,-4,%g0 - bz,pn %icc,.L_bn_sub_words_tail - addcc %g0,0,%g0 ! clear carry flag - -.L_bn_sub_words_loop: ! wow! 32 aligned! - dec 4,%o3 - lduw [%o2],%o5 - lduw [%o1+4],%g1 - lduw [%o2+4],%g2 - lduw [%o1+8],%g3 - lduw [%o2+8],%g4 - subccc %o4,%o5,%o5 - stuw %o5,[%o0] - - lduw [%o1+12],%o4 - lduw [%o2+12],%o5 - inc 16,%o1 - subccc %g1,%g2,%g2 - stuw %g2,[%o0+4] - - inc 16,%o2 - subccc %g3,%g4,%g4 - stuw %g4,[%o0+8] - - inc 16,%o0 - subccc %o4,%o5,%o5 - stuw %o5,[%o0-4] - and %o3,-4,%g1 - brnz,a,pt %g1,.L_bn_sub_words_loop - lduw [%o1],%o4 - - brnz,a,pn %o3,.L_bn_sub_words_tail - lduw [%o1],%o4 -.L_bn_sub_words_return: - clr %o0 - retl - movcs %icc,1,%o0 - nop - -.L_bn_sub_words_tail: ! wow! 32 aligned! - lduw [%o2],%o5 - dec %o3 - subccc %o4,%o5,%o5 - brz,pt %o3,.L_bn_sub_words_return - stuw %o5,[%o0] - - lduw [%o1+4],%o4 - lduw [%o2+4],%o5 - dec %o3 - subccc %o4,%o5,%o5 - brz,pt %o3,.L_bn_sub_words_return - stuw %o5,[%o0+4] - - lduw [%o1+8],%o4 - lduw [%o2+8],%o5 - subccc %o4,%o5,%o5 - stuw %o5,[%o0+8] - clr %o0 - retl - movcs %icc,1,%o0 - -.type bn_sub_words,#function -.size bn_sub_words,(.-bn_sub_words) - -/* - * Code below depends on the fact that upper parts of the %l0-%l7 - * and %i0-%i7 are zeroed by kernel after context switch. In - * previous versions this comment stated that "the trouble is that - * it's not feasible to implement the mumbo-jumbo in less V9 - * instructions:-(" which apparently isn't true thanks to - * 'bcs,a %xcc,.+8; inc %rd' pair. But the performance improvement - * results not from the shorter code, but from elimination of - * multicycle none-pairable 'rd %y,%rd' instructions. - * - * Andy. - */ - -#define FRAME_SIZE -96 - -/* - * Here is register usage map for *all* routines below. - */ -#define t_1 %o0 -#define t_2 %o1 -#define c_12 %o2 -#define c_3 %o3 - -#define ap(I) [%i1+4*I] -#define bp(I) [%i2+4*I] -#define rp(I) [%i0+4*I] - -#define a_0 %l0 -#define a_1 %l1 -#define a_2 %l2 -#define a_3 %l3 -#define a_4 %l4 -#define a_5 %l5 -#define a_6 %l6 -#define a_7 %l7 - -#define b_0 %i3 -#define b_1 %i4 -#define b_2 %i5 -#define b_3 %o4 -#define b_4 %o5 -#define b_5 %o7 -#define b_6 %g1 -#define b_7 %g4 - -.align 32 -.global bn_mul_comba8 -/* - * void bn_mul_comba8(r,a,b) - * BN_ULONG *r,*a,*b; - */ -bn_mul_comba8: - save %sp,FRAME_SIZE,%sp - mov 1,t_2 - lduw ap(0),a_0 - sllx t_2,32,t_2 - lduw bp(0),b_0 != - lduw bp(1),b_1 - mulx a_0,b_0,t_1 !mul_add_c(a[0],b[0],c1,c2,c3); - srlx t_1,32,c_12 - stuw t_1,rp(0) !=!r[0]=c1; - - lduw ap(1),a_1 - mulx a_0,b_1,t_1 !mul_add_c(a[0],b[1],c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(2),a_2 - mulx a_1,b_0,t_1 !=!mul_add_c(a[1],b[0],c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 != - stuw t_1,rp(1) !r[1]=c2; - or c_12,c_3,c_12 - - mulx a_2,b_0,t_1 !mul_add_c(a[2],b[0],c3,c1,c2); - addcc c_12,t_1,c_12 != - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw bp(2),b_2 != - mulx a_1,b_1,t_1 !mul_add_c(a[1],b[1],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - lduw bp(3),b_3 - mulx a_0,b_2,t_1 !mul_add_c(a[0],b[2],c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(2) !r[2]=c3; - or c_12,c_3,c_12 != - - mulx a_0,b_3,t_1 !mul_add_c(a[0],b[3],c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_1,b_2,t_1 !=!mul_add_c(a[1],b[2],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - lduw ap(3),a_3 - mulx a_2,b_1,t_1 !mul_add_c(a[2],b[1],c1,c2,c3); - addcc c_12,t_1,c_12 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(4),a_4 - mulx a_3,b_0,t_1 !=!mul_add_c(a[3],b[0],c1,c2,c3);!= - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 != - stuw t_1,rp(3) !r[3]=c1; - or c_12,c_3,c_12 - - mulx a_4,b_0,t_1 !mul_add_c(a[4],b[0],c2,c3,c1); - addcc c_12,t_1,c_12 != - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,b_1,t_1 !=!mul_add_c(a[3],b[1],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_2,b_2,t_1 !=!mul_add_c(a[2],b[2],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw bp(4),b_4 != - mulx a_1,b_3,t_1 !mul_add_c(a[1],b[3],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - lduw bp(5),b_5 - mulx a_0,b_4,t_1 !mul_add_c(a[0],b[4],c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(4) !r[4]=c2; - or c_12,c_3,c_12 != - - mulx a_0,b_5,t_1 !mul_add_c(a[0],b[5],c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_1,b_4,t_1 !mul_add_c(a[1],b[4],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_2,b_3,t_1 !mul_add_c(a[2],b[3],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_3,b_2,t_1 !mul_add_c(a[3],b[2],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - lduw ap(5),a_5 - mulx a_4,b_1,t_1 !mul_add_c(a[4],b[1],c3,c1,c2); - addcc c_12,t_1,c_12 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(6),a_6 - mulx a_5,b_0,t_1 !=!mul_add_c(a[5],b[0],c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 != - stuw t_1,rp(5) !r[5]=c3; - or c_12,c_3,c_12 - - mulx a_6,b_0,t_1 !mul_add_c(a[6],b[0],c1,c2,c3); - addcc c_12,t_1,c_12 != - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_5,b_1,t_1 !=!mul_add_c(a[5],b[1],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_4,b_2,t_1 !=!mul_add_c(a[4],b[2],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,b_3,t_1 !=!mul_add_c(a[3],b[3],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_2,b_4,t_1 !=!mul_add_c(a[2],b[4],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw bp(6),b_6 != - mulx a_1,b_5,t_1 !mul_add_c(a[1],b[5],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - lduw bp(7),b_7 - mulx a_0,b_6,t_1 !mul_add_c(a[0],b[6],c1,c2,c3); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(6) !r[6]=c1; - or c_12,c_3,c_12 != - - mulx a_0,b_7,t_1 !mul_add_c(a[0],b[7],c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_1,b_6,t_1 !mul_add_c(a[1],b[6],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_2,b_5,t_1 !mul_add_c(a[2],b[5],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_3,b_4,t_1 !mul_add_c(a[3],b[4],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_4,b_3,t_1 !mul_add_c(a[4],b[3],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_5,b_2,t_1 !mul_add_c(a[5],b[2],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - lduw ap(7),a_7 - mulx a_6,b_1,t_1 !=!mul_add_c(a[6],b[1],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_7,b_0,t_1 !=!mul_add_c(a[7],b[0],c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 != - stuw t_1,rp(7) !r[7]=c2; - or c_12,c_3,c_12 - - mulx a_7,b_1,t_1 !=!mul_add_c(a[7],b[1],c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_6,b_2,t_1 !mul_add_c(a[6],b[2],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_5,b_3,t_1 !mul_add_c(a[5],b[3],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_4,b_4,t_1 !mul_add_c(a[4],b[4],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_3,b_5,t_1 !mul_add_c(a[3],b[5],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_2,b_6,t_1 !mul_add_c(a[2],b[6],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_1,b_7,t_1 !mul_add_c(a[1],b[7],c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - srlx t_1,32,c_12 - stuw t_1,rp(8) !r[8]=c3; - or c_12,c_3,c_12 - - mulx a_2,b_7,t_1 !=!mul_add_c(a[2],b[7],c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - mulx a_3,b_6,t_1 !mul_add_c(a[3],b[6],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_4,b_5,t_1 !mul_add_c(a[4],b[5],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_5,b_4,t_1 !mul_add_c(a[5],b[4],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_6,b_3,t_1 !mul_add_c(a[6],b[3],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_7,b_2,t_1 !mul_add_c(a[7],b[2],c1,c2,c3); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(9) !r[9]=c1; - or c_12,c_3,c_12 != - - mulx a_7,b_3,t_1 !mul_add_c(a[7],b[3],c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_6,b_4,t_1 !mul_add_c(a[6],b[4],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_5,b_5,t_1 !mul_add_c(a[5],b[5],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_4,b_6,t_1 !mul_add_c(a[4],b[6],c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_3,b_7,t_1 !mul_add_c(a[3],b[7],c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(10) !r[10]=c2; - or c_12,c_3,c_12 != - - mulx a_4,b_7,t_1 !mul_add_c(a[4],b[7],c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_5,b_6,t_1 !mul_add_c(a[5],b[6],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_6,b_5,t_1 !mul_add_c(a[6],b[5],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_7,b_4,t_1 !mul_add_c(a[7],b[4],c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(11) !r[11]=c3; - or c_12,c_3,c_12 != - - mulx a_7,b_5,t_1 !mul_add_c(a[7],b[5],c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_6,b_6,t_1 !mul_add_c(a[6],b[6],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_5,b_7,t_1 !mul_add_c(a[5],b[7],c1,c2,c3); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(12) !r[12]=c1; - or c_12,c_3,c_12 != - - mulx a_6,b_7,t_1 !mul_add_c(a[6],b[7],c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_7,b_6,t_1 !mul_add_c(a[7],b[6],c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - st t_1,rp(13) !r[13]=c2; - or c_12,c_3,c_12 != - - mulx a_7,b_7,t_1 !mul_add_c(a[7],b[7],c3,c1,c2); - addcc c_12,t_1,t_1 - srlx t_1,32,c_12 != - stuw t_1,rp(14) !r[14]=c3; - stuw c_12,rp(15) !r[15]=c1; - - ret - restore %g0,%g0,%o0 != - -.type bn_mul_comba8,#function -.size bn_mul_comba8,(.-bn_mul_comba8) - -.align 32 - -.global bn_mul_comba4 -/* - * void bn_mul_comba4(r,a,b) - * BN_ULONG *r,*a,*b; - */ -bn_mul_comba4: - save %sp,FRAME_SIZE,%sp - lduw ap(0),a_0 - mov 1,t_2 - lduw bp(0),b_0 - sllx t_2,32,t_2 != - lduw bp(1),b_1 - mulx a_0,b_0,t_1 !mul_add_c(a[0],b[0],c1,c2,c3); - srlx t_1,32,c_12 - stuw t_1,rp(0) !=!r[0]=c1; - - lduw ap(1),a_1 - mulx a_0,b_1,t_1 !mul_add_c(a[0],b[1],c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(2),a_2 - mulx a_1,b_0,t_1 !=!mul_add_c(a[1],b[0],c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 != - stuw t_1,rp(1) !r[1]=c2; - or c_12,c_3,c_12 - - mulx a_2,b_0,t_1 !mul_add_c(a[2],b[0],c3,c1,c2); - addcc c_12,t_1,c_12 != - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw bp(2),b_2 != - mulx a_1,b_1,t_1 !mul_add_c(a[1],b[1],c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 != - lduw bp(3),b_3 - mulx a_0,b_2,t_1 !mul_add_c(a[0],b[2],c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(2) !r[2]=c3; - or c_12,c_3,c_12 != - - mulx a_0,b_3,t_1 !mul_add_c(a[0],b[3],c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - mulx a_1,b_2,t_1 !mul_add_c(a[1],b[2],c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 != - add c_3,t_2,c_3 - lduw ap(3),a_3 - mulx a_2,b_1,t_1 !mul_add_c(a[2],b[1],c1,c2,c3); - addcc c_12,t_1,c_12 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,b_0,t_1 !mul_add_c(a[3],b[0],c1,c2,c3);!= - addcc c_12,t_1,t_1 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(3) !=!r[3]=c1; - or c_12,c_3,c_12 - - mulx a_3,b_1,t_1 !mul_add_c(a[3],b[1],c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_2,b_2,t_1 !mul_add_c(a[2],b[2],c2,c3,c1); - addcc c_12,t_1,c_12 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_1,b_3,t_1 !mul_add_c(a[1],b[3],c2,c3,c1); - addcc c_12,t_1,t_1 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(4) !=!r[4]=c2; - or c_12,c_3,c_12 - - mulx a_2,b_3,t_1 !mul_add_c(a[2],b[3],c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,b_2,t_1 !mul_add_c(a[3],b[2],c3,c1,c2); - addcc c_12,t_1,t_1 != - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(5) !=!r[5]=c3; - or c_12,c_3,c_12 - - mulx a_3,b_3,t_1 !mul_add_c(a[3],b[3],c1,c2,c3); - addcc c_12,t_1,t_1 - srlx t_1,32,c_12 != - stuw t_1,rp(6) !r[6]=c1; - stuw c_12,rp(7) !r[7]=c2; - - ret - restore %g0,%g0,%o0 - -.type bn_mul_comba4,#function -.size bn_mul_comba4,(.-bn_mul_comba4) - -.align 32 - -.global bn_sqr_comba8 -bn_sqr_comba8: - save %sp,FRAME_SIZE,%sp - mov 1,t_2 - lduw ap(0),a_0 - sllx t_2,32,t_2 - lduw ap(1),a_1 - mulx a_0,a_0,t_1 !sqr_add_c(a,0,c1,c2,c3); - srlx t_1,32,c_12 - stuw t_1,rp(0) !r[0]=c1; - - lduw ap(2),a_2 - mulx a_0,a_1,t_1 !=!sqr_add_c2(a,1,0,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(1) !r[1]=c2; - or c_12,c_3,c_12 - - mulx a_2,a_0,t_1 !sqr_add_c2(a,2,0,c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(3),a_3 - mulx a_1,a_1,t_1 !sqr_add_c(a,1,c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(2) !r[2]=c3; - or c_12,c_3,c_12 - - mulx a_0,a_3,t_1 !sqr_add_c2(a,3,0,c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(4),a_4 - mulx a_1,a_2,t_1 !sqr_add_c2(a,2,1,c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - st t_1,rp(3) !r[3]=c1; - or c_12,c_3,c_12 - - mulx a_4,a_0,t_1 !sqr_add_c2(a,4,0,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,a_1,t_1 !sqr_add_c2(a,3,1,c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(5),a_5 - mulx a_2,a_2,t_1 !sqr_add_c(a,2,c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(4) !r[4]=c2; - or c_12,c_3,c_12 - - mulx a_0,a_5,t_1 !sqr_add_c2(a,5,0,c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_1,a_4,t_1 !sqr_add_c2(a,4,1,c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(6),a_6 - mulx a_2,a_3,t_1 !sqr_add_c2(a,3,2,c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(5) !r[5]=c3; - or c_12,c_3,c_12 - - mulx a_6,a_0,t_1 !sqr_add_c2(a,6,0,c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_5,a_1,t_1 !sqr_add_c2(a,5,1,c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_4,a_2,t_1 !sqr_add_c2(a,4,2,c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(7),a_7 - mulx a_3,a_3,t_1 !=!sqr_add_c(a,3,c1,c2,c3); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(6) !r[6]=c1; - or c_12,c_3,c_12 - - mulx a_0,a_7,t_1 !sqr_add_c2(a,7,0,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_1,a_6,t_1 !sqr_add_c2(a,6,1,c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_2,a_5,t_1 !sqr_add_c2(a,5,2,c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,a_4,t_1 !sqr_add_c2(a,4,3,c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(7) !r[7]=c2; - or c_12,c_3,c_12 - - mulx a_7,a_1,t_1 !sqr_add_c2(a,7,1,c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_6,a_2,t_1 !sqr_add_c2(a,6,2,c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_5,a_3,t_1 !sqr_add_c2(a,5,3,c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_4,a_4,t_1 !sqr_add_c(a,4,c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(8) !r[8]=c3; - or c_12,c_3,c_12 - - mulx a_2,a_7,t_1 !sqr_add_c2(a,7,2,c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_3,a_6,t_1 !sqr_add_c2(a,6,3,c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_4,a_5,t_1 !sqr_add_c2(a,5,4,c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(9) !r[9]=c1; - or c_12,c_3,c_12 - - mulx a_7,a_3,t_1 !sqr_add_c2(a,7,3,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_6,a_4,t_1 !sqr_add_c2(a,6,4,c2,c3,c1); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_5,a_5,t_1 !sqr_add_c(a,5,c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(10) !r[10]=c2; - or c_12,c_3,c_12 - - mulx a_4,a_7,t_1 !sqr_add_c2(a,7,4,c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_5,a_6,t_1 !sqr_add_c2(a,6,5,c3,c1,c2); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(11) !r[11]=c3; - or c_12,c_3,c_12 - - mulx a_7,a_5,t_1 !sqr_add_c2(a,7,5,c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_6,a_6,t_1 !sqr_add_c(a,6,c1,c2,c3); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(12) !r[12]=c1; - or c_12,c_3,c_12 - - mulx a_6,a_7,t_1 !sqr_add_c2(a,7,6,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(13) !r[13]=c2; - or c_12,c_3,c_12 - - mulx a_7,a_7,t_1 !sqr_add_c(a,7,c3,c1,c2); - addcc c_12,t_1,t_1 - srlx t_1,32,c_12 - stuw t_1,rp(14) !r[14]=c3; - stuw c_12,rp(15) !r[15]=c1; - - ret - restore %g0,%g0,%o0 - -.type bn_sqr_comba8,#function -.size bn_sqr_comba8,(.-bn_sqr_comba8) - -.align 32 - -.global bn_sqr_comba4 -/* - * void bn_sqr_comba4(r,a) - * BN_ULONG *r,*a; - */ -bn_sqr_comba4: - save %sp,FRAME_SIZE,%sp - mov 1,t_2 - lduw ap(0),a_0 - sllx t_2,32,t_2 - lduw ap(1),a_1 - mulx a_0,a_0,t_1 !sqr_add_c(a,0,c1,c2,c3); - srlx t_1,32,c_12 - stuw t_1,rp(0) !r[0]=c1; - - lduw ap(2),a_2 - mulx a_0,a_1,t_1 !sqr_add_c2(a,1,0,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(1) !r[1]=c2; - or c_12,c_3,c_12 - - mulx a_2,a_0,t_1 !sqr_add_c2(a,2,0,c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - lduw ap(3),a_3 - mulx a_1,a_1,t_1 !sqr_add_c(a,1,c3,c1,c2); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(2) !r[2]=c3; - or c_12,c_3,c_12 - - mulx a_0,a_3,t_1 !sqr_add_c2(a,3,0,c1,c2,c3); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_1,a_2,t_1 !sqr_add_c2(a,2,1,c1,c2,c3); - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(3) !r[3]=c1; - or c_12,c_3,c_12 - - mulx a_3,a_1,t_1 !sqr_add_c2(a,3,1,c2,c3,c1); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,c_12 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - mulx a_2,a_2,t_1 !sqr_add_c(a,2,c2,c3,c1); - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(4) !r[4]=c2; - or c_12,c_3,c_12 - - mulx a_2,a_3,t_1 !sqr_add_c2(a,3,2,c3,c1,c2); - addcc c_12,t_1,c_12 - clr c_3 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - addcc c_12,t_1,t_1 - bcs,a %xcc,.+8 - add c_3,t_2,c_3 - srlx t_1,32,c_12 - stuw t_1,rp(5) !r[5]=c3; - or c_12,c_3,c_12 - - mulx a_3,a_3,t_1 !sqr_add_c(a,3,c1,c2,c3); - addcc c_12,t_1,t_1 - srlx t_1,32,c_12 - stuw t_1,rp(6) !r[6]=c1; - stuw c_12,rp(7) !r[7]=c2; - - ret - restore %g0,%g0,%o0 - -.type bn_sqr_comba4,#function -.size bn_sqr_comba4,(.-bn_sqr_comba4) - -.align 32 diff --git a/src/lib/libcrypto/bn/asm/x86.pl b/src/lib/libcrypto/bn/asm/x86.pl deleted file mode 100644 index 1bc4f1bb27..0000000000 --- a/src/lib/libcrypto/bn/asm/x86.pl +++ /dev/null @@ -1,28 +0,0 @@ -#!/usr/local/bin/perl - -push(@INC,"perlasm","../../perlasm"); -require "x86asm.pl"; - -require("x86/mul_add.pl"); -require("x86/mul.pl"); -require("x86/sqr.pl"); -require("x86/div.pl"); -require("x86/add.pl"); -require("x86/sub.pl"); -require("x86/comba.pl"); - -&asm_init($ARGV[0],$0); - -&bn_mul_add_words("bn_mul_add_words"); -&bn_mul_words("bn_mul_words"); -&bn_sqr_words("bn_sqr_words"); -&bn_div_words("bn_div_words"); -&bn_add_words("bn_add_words"); -&bn_sub_words("bn_sub_words"); -&bn_mul_comba("bn_mul_comba8",8); -&bn_mul_comba("bn_mul_comba4",4); -&bn_sqr_comba("bn_sqr_comba8",8); -&bn_sqr_comba("bn_sqr_comba4",4); - -&asm_finish(); - diff --git a/src/lib/libcrypto/bn/asm/x86/add.pl b/src/lib/libcrypto/bn/asm/x86/add.pl deleted file mode 100644 index 0b5cf583e3..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/add.pl +++ /dev/null @@ -1,76 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub bn_add_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $a="esi"; - $b="edi"; - $c="eax"; - $r="ebx"; - $tmp1="ecx"; - $tmp2="edx"; - $num="ebp"; - - &mov($r,&wparam(0)); # get r - &mov($a,&wparam(1)); # get a - &mov($b,&wparam(2)); # get b - &mov($num,&wparam(3)); # get num - &xor($c,$c); # clear carry - &and($num,0xfffffff8); # num / 8 - - &jz(&label("aw_finish")); - - &set_label("aw_loop",0); - for ($i=0; $i<8; $i++) - { - &comment("Round $i"); - - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0)); # *b - &add($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &add($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - } - - &comment(""); - &add($a,32); - &add($b,32); - &add($r,32); - &sub($num,8); - &jnz(&label("aw_loop")); - - &set_label("aw_finish",0); - &mov($num,&wparam(3)); # get num - &and($num,7); - &jz(&label("aw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0));# *b - &add($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &add($tmp1,$tmp2); - &adc($c,0); - &dec($num) if ($i != 6); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *a - &jz(&label("aw_end")) if ($i != 6); - } - &set_label("aw_end",0); - -# &mov("eax",$c); # $c is "eax" - - &function_end($name); - } - -1; diff --git a/src/lib/libcrypto/bn/asm/x86/comba.pl b/src/lib/libcrypto/bn/asm/x86/comba.pl deleted file mode 100644 index 2291253629..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/comba.pl +++ /dev/null @@ -1,277 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub mul_add_c - { - local($a,$ai,$b,$bi,$c0,$c1,$c2,$pos,$i,$na,$nb)=@_; - - # pos == -1 if eax and edx are pre-loaded, 0 to load from next - # words, and 1 if load return value - - &comment("mul a[$ai]*b[$bi]"); - - # "eax" and "edx" will always be pre-loaded. - # &mov("eax",&DWP($ai*4,$a,"",0)) ; - # &mov("edx",&DWP($bi*4,$b,"",0)); - - &mul("edx"); - &add($c0,"eax"); - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 0; # laod next a - &mov("eax",&wparam(0)) if $pos > 0; # load r[] - ### - &adc($c1,"edx"); - &mov("edx",&DWP(($nb)*4,$b,"",0)) if $pos == 0; # laod next b - &mov("edx",&DWP(($nb)*4,$b,"",0)) if $pos == 1; # laod next b - ### - &adc($c2,0); - # is pos > 1, it means it is the last loop - &mov(&DWP($i*4,"eax","",0),$c0) if $pos > 0; # save r[]; - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 1; # laod next a - } - -sub sqr_add_c - { - local($r,$a,$ai,$bi,$c0,$c1,$c2,$pos,$i,$na,$nb)=@_; - - # pos == -1 if eax and edx are pre-loaded, 0 to load from next - # words, and 1 if load return value - - &comment("sqr a[$ai]*a[$bi]"); - - # "eax" and "edx" will always be pre-loaded. - # &mov("eax",&DWP($ai*4,$a,"",0)) ; - # &mov("edx",&DWP($bi*4,$b,"",0)); - - if ($ai == $bi) - { &mul("eax");} - else - { &mul("edx");} - &add($c0,"eax"); - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 0; # load next a - ### - &adc($c1,"edx"); - &mov("edx",&DWP(($nb)*4,$a,"",0)) if ($pos == 1) && ($na != $nb); - ### - &adc($c2,0); - # is pos > 1, it means it is the last loop - &mov(&DWP($i*4,$r,"",0),$c0) if $pos > 0; # save r[]; - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 1; # load next b - } - -sub sqr_add_c2 - { - local($r,$a,$ai,$bi,$c0,$c1,$c2,$pos,$i,$na,$nb)=@_; - - # pos == -1 if eax and edx are pre-loaded, 0 to load from next - # words, and 1 if load return value - - &comment("sqr a[$ai]*a[$bi]"); - - # "eax" and "edx" will always be pre-loaded. - # &mov("eax",&DWP($ai*4,$a,"",0)) ; - # &mov("edx",&DWP($bi*4,$a,"",0)); - - if ($ai == $bi) - { &mul("eax");} - else - { &mul("edx");} - &add("eax","eax"); - ### - &adc("edx","edx"); - ### - &adc($c2,0); - &add($c0,"eax"); - &adc($c1,"edx"); - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 0; # load next a - &mov("eax",&DWP(($na)*4,$a,"",0)) if $pos == 1; # load next b - &adc($c2,0); - &mov(&DWP($i*4,$r,"",0),$c0) if $pos > 0; # save r[]; - &mov("edx",&DWP(($nb)*4,$a,"",0)) if ($pos <= 1) && ($na != $nb); - ### - } - -sub bn_mul_comba - { - local($name,$num)=@_; - local($a,$b,$c0,$c1,$c2); - local($i,$as,$ae,$bs,$be,$ai,$bi); - local($tot,$end); - - &function_begin_B($name,""); - - $c0="ebx"; - $c1="ecx"; - $c2="ebp"; - $a="esi"; - $b="edi"; - - $as=0; - $ae=0; - $bs=0; - $be=0; - $tot=$num+$num-1; - - &push("esi"); - &mov($a,&wparam(1)); - &push("edi"); - &mov($b,&wparam(2)); - &push("ebp"); - &push("ebx"); - - &xor($c0,$c0); - &mov("eax",&DWP(0,$a,"",0)); # load the first word - &xor($c1,$c1); - &mov("edx",&DWP(0,$b,"",0)); # load the first second - - for ($i=0; $i<$tot; $i++) - { - $ai=$as; - $bi=$bs; - $end=$be+1; - - &comment("################## Calculate word $i"); - - for ($j=$bs; $j<$end; $j++) - { - &xor($c2,$c2) if ($j == $bs); - if (($j+1) == $end) - { - $v=1; - $v=2 if (($i+1) == $tot); - } - else - { $v=0; } - if (($j+1) != $end) - { - $na=($ai-1); - $nb=($bi+1); - } - else - { - $na=$as+($i < ($num-1)); - $nb=$bs+($i >= ($num-1)); - } -#printf STDERR "[$ai,$bi] -> [$na,$nb]\n"; - &mul_add_c($a,$ai,$b,$bi,$c0,$c1,$c2,$v,$i,$na,$nb); - if ($v) - { - &comment("saved r[$i]"); - # &mov("eax",&wparam(0)); - # &mov(&DWP($i*4,"eax","",0),$c0); - ($c0,$c1,$c2)=($c1,$c2,$c0); - } - $ai--; - $bi++; - } - $as++ if ($i < ($num-1)); - $ae++ if ($i >= ($num-1)); - - $bs++ if ($i >= ($num-1)); - $be++ if ($i < ($num-1)); - } - &comment("save r[$i]"); - # &mov("eax",&wparam(0)); - &mov(&DWP($i*4,"eax","",0),$c0); - - &pop("ebx"); - &pop("ebp"); - &pop("edi"); - &pop("esi"); - &ret(); - &function_end_B($name); - } - -sub bn_sqr_comba - { - local($name,$num)=@_; - local($r,$a,$c0,$c1,$c2)=@_; - local($i,$as,$ae,$bs,$be,$ai,$bi); - local($b,$tot,$end,$half); - - &function_begin_B($name,""); - - $c0="ebx"; - $c1="ecx"; - $c2="ebp"; - $a="esi"; - $r="edi"; - - &push("esi"); - &push("edi"); - &push("ebp"); - &push("ebx"); - &mov($r,&wparam(0)); - &mov($a,&wparam(1)); - &xor($c0,$c0); - &xor($c1,$c1); - &mov("eax",&DWP(0,$a,"",0)); # load the first word - - $as=0; - $ae=0; - $bs=0; - $be=0; - $tot=$num+$num-1; - - for ($i=0; $i<$tot; $i++) - { - $ai=$as; - $bi=$bs; - $end=$be+1; - - &comment("############### Calculate word $i"); - for ($j=$bs; $j<$end; $j++) - { - &xor($c2,$c2) if ($j == $bs); - if (($ai-1) < ($bi+1)) - { - $v=1; - $v=2 if ($i+1) == $tot; - } - else - { $v=0; } - if (!$v) - { - $na=$ai-1; - $nb=$bi+1; - } - else - { - $na=$as+($i < ($num-1)); - $nb=$bs+($i >= ($num-1)); - } - if ($ai == $bi) - { - &sqr_add_c($r,$a,$ai,$bi, - $c0,$c1,$c2,$v,$i,$na,$nb); - } - else - { - &sqr_add_c2($r,$a,$ai,$bi, - $c0,$c1,$c2,$v,$i,$na,$nb); - } - if ($v) - { - &comment("saved r[$i]"); - #&mov(&DWP($i*4,$r,"",0),$c0); - ($c0,$c1,$c2)=($c1,$c2,$c0); - last; - } - $ai--; - $bi++; - } - $as++ if ($i < ($num-1)); - $ae++ if ($i >= ($num-1)); - - $bs++ if ($i >= ($num-1)); - $be++ if ($i < ($num-1)); - } - &mov(&DWP($i*4,$r,"",0),$c0); - &pop("ebx"); - &pop("ebp"); - &pop("edi"); - &pop("esi"); - &ret(); - &function_end_B($name); - } - -1; diff --git a/src/lib/libcrypto/bn/asm/x86/div.pl b/src/lib/libcrypto/bn/asm/x86/div.pl deleted file mode 100644 index 0e90152caa..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/div.pl +++ /dev/null @@ -1,15 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub bn_div_words - { - local($name)=@_; - - &function_begin($name,""); - &mov("edx",&wparam(0)); # - &mov("eax",&wparam(1)); # - &mov("ebx",&wparam(2)); # - &div("ebx"); - &function_end($name); - } -1; diff --git a/src/lib/libcrypto/bn/asm/x86/mul.pl b/src/lib/libcrypto/bn/asm/x86/mul.pl deleted file mode 100644 index 674cb9b055..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/mul.pl +++ /dev/null @@ -1,77 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub bn_mul_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $Low="eax"; - $High="edx"; - $a="ebx"; - $w="ecx"; - $r="edi"; - $c="esi"; - $num="ebp"; - - &xor($c,$c); # clear carry - &mov($r,&wparam(0)); # - &mov($a,&wparam(1)); # - &mov($num,&wparam(2)); # - &mov($w,&wparam(3)); # - - &and($num,0xfffffff8); # num / 8 - &jz(&label("mw_finish")); - - &set_label("mw_loop",0); - for ($i=0; $i<32; $i+=4) - { - &comment("Round $i"); - - &mov("eax",&DWP($i,$a,"",0)); # *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+=c - # XXX - - &adc("edx",0); # H(t)+=carry - &mov(&DWP($i,$r,"",0),"eax"); # *r= L(t); - - &mov($c,"edx"); # c= H(t); - } - - &comment(""); - &add($a,32); - &add($r,32); - &sub($num,8); - &jz(&label("mw_finish")); - &jmp(&label("mw_loop")); - - &set_label("mw_finish",0); - &mov($num,&wparam(2)); # get num - &and($num,7); - &jnz(&label("mw_finish2")); - &jmp(&label("mw_end")); - - &set_label("mw_finish2",1); - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov("eax",&DWP($i*4,$a,"",0));# *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+=c - # XXX - &adc("edx",0); # H(t)+=carry - &mov(&DWP($i*4,$r,"",0),"eax");# *r= L(t); - &mov($c,"edx"); # c= H(t); - &dec($num) if ($i != 7-1); - &jz(&label("mw_end")) if ($i != 7-1); - } - &set_label("mw_end",0); - &mov("eax",$c); - - &function_end($name); - } - -1; diff --git a/src/lib/libcrypto/bn/asm/x86/mul_add.pl b/src/lib/libcrypto/bn/asm/x86/mul_add.pl deleted file mode 100644 index 61830d3a90..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/mul_add.pl +++ /dev/null @@ -1,87 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub bn_mul_add_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $Low="eax"; - $High="edx"; - $a="ebx"; - $w="ebp"; - $r="edi"; - $c="esi"; - - &xor($c,$c); # clear carry - &mov($r,&wparam(0)); # - - &mov("ecx",&wparam(2)); # - &mov($a,&wparam(1)); # - - &and("ecx",0xfffffff8); # num / 8 - &mov($w,&wparam(3)); # - - &push("ecx"); # Up the stack for a tmp variable - - &jz(&label("maw_finish")); - - &set_label("maw_loop",0); - - &mov(&swtmp(0),"ecx"); # - - for ($i=0; $i<32; $i+=4) - { - &comment("Round $i"); - - &mov("eax",&DWP($i,$a,"",0)); # *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+= *r - &mov($c,&DWP($i,$r,"",0)); # L(t)+= *r - &adc("edx",0); # H(t)+=carry - &add("eax",$c); # L(t)+=c - &adc("edx",0); # H(t)+=carry - &mov(&DWP($i,$r,"",0),"eax"); # *r= L(t); - &mov($c,"edx"); # c= H(t); - } - - &comment(""); - &mov("ecx",&swtmp(0)); # - &add($a,32); - &add($r,32); - &sub("ecx",8); - &jnz(&label("maw_loop")); - - &set_label("maw_finish",0); - &mov("ecx",&wparam(2)); # get num - &and("ecx",7); - &jnz(&label("maw_finish2")); # helps branch prediction - &jmp(&label("maw_end")); - - &set_label("maw_finish2",1); - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov("eax",&DWP($i*4,$a,"",0));# *a - &mul($w); # *a * w - &add("eax",$c); # L(t)+=c - &mov($c,&DWP($i*4,$r,"",0)); # L(t)+= *r - &adc("edx",0); # H(t)+=carry - &add("eax",$c); - &adc("edx",0); # H(t)+=carry - &dec("ecx") if ($i != 7-1); - &mov(&DWP($i*4,$r,"",0),"eax"); # *r= L(t); - &mov($c,"edx"); # c= H(t); - &jz(&label("maw_end")) if ($i != 7-1); - } - &set_label("maw_end",0); - &mov("eax",$c); - - &pop("ecx"); # clear variable from - - &function_end($name); - } - -1; diff --git a/src/lib/libcrypto/bn/asm/x86/sqr.pl b/src/lib/libcrypto/bn/asm/x86/sqr.pl deleted file mode 100644 index 1f90993cf6..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/sqr.pl +++ /dev/null @@ -1,60 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub bn_sqr_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $r="esi"; - $a="edi"; - $num="ebx"; - - &mov($r,&wparam(0)); # - &mov($a,&wparam(1)); # - &mov($num,&wparam(2)); # - - &and($num,0xfffffff8); # num / 8 - &jz(&label("sw_finish")); - - &set_label("sw_loop",0); - for ($i=0; $i<32; $i+=4) - { - &comment("Round $i"); - &mov("eax",&DWP($i,$a,"",0)); # *a - # XXX - &mul("eax"); # *a * *a - &mov(&DWP($i*2,$r,"",0),"eax"); # - &mov(&DWP($i*2+4,$r,"",0),"edx");# - } - - &comment(""); - &add($a,32); - &add($r,64); - &sub($num,8); - &jnz(&label("sw_loop")); - - &set_label("sw_finish",0); - &mov($num,&wparam(2)); # get num - &and($num,7); - &jz(&label("sw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov("eax",&DWP($i*4,$a,"",0)); # *a - # XXX - &mul("eax"); # *a * *a - &mov(&DWP($i*8,$r,"",0),"eax"); # - &dec($num) if ($i != 7-1); - &mov(&DWP($i*8+4,$r,"",0),"edx"); - &jz(&label("sw_end")) if ($i != 7-1); - } - &set_label("sw_end",0); - - &function_end($name); - } - -1; diff --git a/src/lib/libcrypto/bn/asm/x86/sub.pl b/src/lib/libcrypto/bn/asm/x86/sub.pl deleted file mode 100644 index 837b0e1b07..0000000000 --- a/src/lib/libcrypto/bn/asm/x86/sub.pl +++ /dev/null @@ -1,76 +0,0 @@ -#!/usr/local/bin/perl -# x86 assember - -sub bn_sub_words - { - local($name)=@_; - - &function_begin($name,""); - - &comment(""); - $a="esi"; - $b="edi"; - $c="eax"; - $r="ebx"; - $tmp1="ecx"; - $tmp2="edx"; - $num="ebp"; - - &mov($r,&wparam(0)); # get r - &mov($a,&wparam(1)); # get a - &mov($b,&wparam(2)); # get b - &mov($num,&wparam(3)); # get num - &xor($c,$c); # clear carry - &and($num,0xfffffff8); # num / 8 - - &jz(&label("aw_finish")); - - &set_label("aw_loop",0); - for ($i=0; $i<8; $i++) - { - &comment("Round $i"); - - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0)); # *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *r - } - - &comment(""); - &add($a,32); - &add($b,32); - &add($r,32); - &sub($num,8); - &jnz(&label("aw_loop")); - - &set_label("aw_finish",0); - &mov($num,&wparam(3)); # get num - &and($num,7); - &jz(&label("aw_end")); - - for ($i=0; $i<7; $i++) - { - &comment("Tail Round $i"); - &mov($tmp1,&DWP($i*4,$a,"",0)); # *a - &mov($tmp2,&DWP($i*4,$b,"",0));# *b - &sub($tmp1,$c); - &mov($c,0); - &adc($c,$c); - &sub($tmp1,$tmp2); - &adc($c,0); - &dec($num) if ($i != 6); - &mov(&DWP($i*4,$r,"",0),$tmp1); # *a - &jz(&label("aw_end")) if ($i != 6); - } - &set_label("aw_end",0); - -# &mov("eax",$c); # $c is "eax" - - &function_end($name); - } - -1; diff --git a/src/lib/libcrypto/bn/asm/x86_64-gcc.c b/src/lib/libcrypto/bn/asm/x86_64-gcc.c deleted file mode 100644 index 7378344251..0000000000 --- a/src/lib/libcrypto/bn/asm/x86_64-gcc.c +++ /dev/null @@ -1,593 +0,0 @@ -/* - * x86_64 BIGNUM accelerator version 0.1, December 2002. - * - * Implemented by Andy Polyakov for the OpenSSL - * project. - * - * Rights for redistribution and usage in source and binary forms are - * granted according to the OpenSSL license. Warranty of any kind is - * disclaimed. - * - * Q. Version 0.1? It doesn't sound like Andy, he used to assign real - * versions, like 1.0... - * A. Well, that's because this code is basically a quick-n-dirty - * proof-of-concept hack. As you can see it's implemented with - * inline assembler, which means that you're bound to GCC and that - * there might be enough room for further improvement. - * - * Q. Why inline assembler? - * A. x86_64 features own ABI which I'm not familiar with. This is - * why I decided to let the compiler take care of subroutine - * prologue/epilogue as well as register allocation. For reference. - * Win64 implements different ABI for AMD64, different from Linux. - * - * Q. How much faster does it get? - * A. 'apps/openssl speed rsa dsa' output with no-asm: - * - * sign verify sign/s verify/s - * rsa 512 bits 0.0006s 0.0001s 1683.8 18456.2 - * rsa 1024 bits 0.0028s 0.0002s 356.0 6407.0 - * rsa 2048 bits 0.0172s 0.0005s 58.0 1957.8 - * rsa 4096 bits 0.1155s 0.0018s 8.7 555.6 - * sign verify sign/s verify/s - * dsa 512 bits 0.0005s 0.0006s 2100.8 1768.3 - * dsa 1024 bits 0.0014s 0.0018s 692.3 559.2 - * dsa 2048 bits 0.0049s 0.0061s 204.7 165.0 - * - * 'apps/openssl speed rsa dsa' output with this module: - * - * sign verify sign/s verify/s - * rsa 512 bits 0.0004s 0.0000s 2767.1 33297.9 - * rsa 1024 bits 0.0012s 0.0001s 867.4 14674.7 - * rsa 2048 bits 0.0061s 0.0002s 164.0 5270.0 - * rsa 4096 bits 0.0384s 0.0006s 26.1 1650.8 - * sign verify sign/s verify/s - * dsa 512 bits 0.0002s 0.0003s 4442.2 3786.3 - * dsa 1024 bits 0.0005s 0.0007s 1835.1 1497.4 - * dsa 2048 bits 0.0016s 0.0020s 620.4 504.6 - * - * For the reference. IA-32 assembler implementation performs - * very much like 64-bit code compiled with no-asm on the same - * machine. - */ - -#define BN_ULONG unsigned long - -/* - * "m"(a), "+m"(r) is the way to favor DirectPath µ-code; - * "g"(0) let the compiler to decide where does it - * want to keep the value of zero; - */ -#define mul_add(r,a,word,carry) do { \ - register BN_ULONG high,low; \ - asm ("mulq %3" \ - : "=a"(low),"=d"(high) \ - : "a"(word),"m"(a) \ - : "cc"); \ - asm ("addq %2,%0; adcq %3,%1" \ - : "+r"(carry),"+d"(high)\ - : "a"(low),"g"(0) \ - : "cc"); \ - asm ("addq %2,%0; adcq %3,%1" \ - : "+m"(r),"+d"(high) \ - : "r"(carry),"g"(0) \ - : "cc"); \ - carry=high; \ - } while (0) - -#define mul(r,a,word,carry) do { \ - register BN_ULONG high,low; \ - asm ("mulq %3" \ - : "=a"(low),"=d"(high) \ - : "a"(word),"g"(a) \ - : "cc"); \ - asm ("addq %2,%0; adcq %3,%1" \ - : "+r"(carry),"+d"(high)\ - : "a"(low),"g"(0) \ - : "cc"); \ - (r)=carry, carry=high; \ - } while (0) - -#define sqr(r0,r1,a) \ - asm ("mulq %2" \ - : "=a"(r0),"=d"(r1) \ - : "a"(a) \ - : "cc"); - -BN_ULONG bn_mul_add_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w) - { - BN_ULONG c1=0; - - if (num <= 0) return(c1); - - while (num&~3) - { - mul_add(rp[0],ap[0],w,c1); - mul_add(rp[1],ap[1],w,c1); - mul_add(rp[2],ap[2],w,c1); - mul_add(rp[3],ap[3],w,c1); - ap+=4; rp+=4; num-=4; - } - if (num) - { - mul_add(rp[0],ap[0],w,c1); if (--num==0) return c1; - mul_add(rp[1],ap[1],w,c1); if (--num==0) return c1; - mul_add(rp[2],ap[2],w,c1); return c1; - } - - return(c1); - } - -BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w) - { - BN_ULONG c1=0; - - if (num <= 0) return(c1); - - while (num&~3) - { - mul(rp[0],ap[0],w,c1); - mul(rp[1],ap[1],w,c1); - mul(rp[2],ap[2],w,c1); - mul(rp[3],ap[3],w,c1); - ap+=4; rp+=4; num-=4; - } - if (num) - { - mul(rp[0],ap[0],w,c1); if (--num == 0) return c1; - mul(rp[1],ap[1],w,c1); if (--num == 0) return c1; - mul(rp[2],ap[2],w,c1); - } - return(c1); - } - -void bn_sqr_words(BN_ULONG *r, BN_ULONG *a, int n) - { - if (n <= 0) return; - - while (n&~3) - { - sqr(r[0],r[1],a[0]); - sqr(r[2],r[3],a[1]); - sqr(r[4],r[5],a[2]); - sqr(r[6],r[7],a[3]); - a+=4; r+=8; n-=4; - } - if (n) - { - sqr(r[0],r[1],a[0]); if (--n == 0) return; - sqr(r[2],r[3],a[1]); if (--n == 0) return; - sqr(r[4],r[5],a[2]); - } - } - -BN_ULONG bn_div_words(BN_ULONG h, BN_ULONG l, BN_ULONG d) -{ BN_ULONG ret,waste; - - asm ("divq %4" - : "=a"(ret),"=d"(waste) - : "a"(l),"d"(h),"g"(d) - : "cc"); - - return ret; -} - -BN_ULONG bn_add_words (BN_ULONG *rp, BN_ULONG *ap, BN_ULONG *bp,int n) -{ BN_ULONG ret=0,i=0; - - if (n <= 0) return 0; - - asm ( - " subq %2,%2 \n" - ".align 16 \n" - "1: movq (%4,%2,8),%0 \n" - " adcq (%5,%2,8),%0 \n" - " movq %0,(%3,%2,8) \n" - " leaq 1(%2),%2 \n" - " loop 1b \n" - " sbbq %0,%0 \n" - : "=&a"(ret),"+c"(n),"=&r"(i) - : "r"(rp),"r"(ap),"r"(bp) - : "cc" - ); - - return ret&1; -} - -#ifndef SIMICS -BN_ULONG bn_sub_words (BN_ULONG *rp, BN_ULONG *ap, BN_ULONG *bp,int n) -{ BN_ULONG ret=0,i=0; - - if (n <= 0) return 0; - - asm ( - " subq %2,%2 \n" - ".align 16 \n" - "1: movq (%4,%2,8),%0 \n" - " sbbq (%5,%2,8),%0 \n" - " movq %0,(%3,%2,8) \n" - " leaq 1(%2),%2 \n" - " loop 1b \n" - " sbbq %0,%0 \n" - : "=&a"(ret),"+c"(n),"=&r"(i) - : "r"(rp),"r"(ap),"r"(bp) - : "cc" - ); - - return ret&1; -} -#else -/* Simics 1.4<7 has buggy sbbq:-( */ -#define BN_MASK2 0xffffffffffffffffL -BN_ULONG bn_sub_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n) - { - BN_ULONG t1,t2; - int c=0; - - if (n <= 0) return((BN_ULONG)0); - - for (;;) - { - t1=a[0]; t2=b[0]; - r[0]=(t1-t2-c)&BN_MASK2; - if (t1 != t2) c=(t1 < t2); - if (--n <= 0) break; - - t1=a[1]; t2=b[1]; - r[1]=(t1-t2-c)&BN_MASK2; - if (t1 != t2) c=(t1 < t2); - if (--n <= 0) break; - - t1=a[2]; t2=b[2]; - r[2]=(t1-t2-c)&BN_MASK2; - if (t1 != t2) c=(t1 < t2); - if (--n <= 0) break; - - t1=a[3]; t2=b[3]; - r[3]=(t1-t2-c)&BN_MASK2; - if (t1 != t2) c=(t1 < t2); - if (--n <= 0) break; - - a+=4; - b+=4; - r+=4; - } - return(c); - } -#endif - -/* mul_add_c(a,b,c0,c1,c2) -- c+=a*b for three word number c=(c2,c1,c0) */ -/* mul_add_c2(a,b,c0,c1,c2) -- c+=2*a*b for three word number c=(c2,c1,c0) */ -/* sqr_add_c(a,i,c0,c1,c2) -- c+=a[i]^2 for three word number c=(c2,c1,c0) */ -/* sqr_add_c2(a,i,c0,c1,c2) -- c+=2*a[i]*a[j] for three word number c=(c2,c1,c0) */ - -#if 0 -/* original macros are kept for reference purposes */ -#define mul_add_c(a,b,c0,c1,c2) { \ - BN_ULONG ta=(a),tb=(b); \ - t1 = ta * tb; \ - t2 = BN_UMULT_HIGH(ta,tb); \ - c0 += t1; t2 += (c0