bn/asm/rsaz-x86_64.pl: fix prototype.

[openssl.git] / crypto / bn / asm / armv4-gf2m.pl

diff --git a/crypto/bn/asm/armv4-gf2m.pl b/crypto/bn/asm/armv4-gf2m.pl
index 67ec4b2c14ee91c4ba5513693914734ea350e73f..c52e0b75b5b6231c00847ea5451d0965d527feff 100644 (file)
--- a/crypto/bn/asm/armv4-gf2m.pl
+++ b/crypto/bn/asm/armv4-gf2m.pl
@@ -21,13 +21,8 @@
 # runs in even less cycles, ~30, improvement is measurable only on
 # longer keys. One has to optimize code elsewhere to get NEON glow...
 
-$a="r1";
-$b="r0";
-
-($a0,$a1,$a2,$a12,$a4,$a14)=
-($hi,$lo,$t0,$t1, $i0,$i1 )=map("r$_",(4..9),12);
-
-$mask="r12";
+while (($output=shift) && ($output!~/^\w[\w\-]*\.\w+$/)) {}
+open STDOUT,">$output";
 
 sub Dlo()   { shift=~m|q([1]?[0-9])|?"d".($1*2):"";     }
 sub Dhi()   { shift=~m|q([1]?[0-9])|?"d".($1*2+1):"";   }
@@ -67,9 +62,21 @@ mul_1x1_neon:
        bx      lr
 .size  mul_1x1_neon,.-mul_1x1_neon
 #endif
+___
+################
+# private interface to mul_1x1_ialu
+#
+$a="r1";
+$b="r0";
 
-.align 5
+($a0,$a1,$a2,$a12,$a4,$a14)=
+($hi,$lo,$t0,$t1, $i0,$i1 )=map("r$_",(4..9),12);
+
+$mask="r12";
+
+$code.=<<___;
 .type  mul_1x1_ialu,%function
+.align 5
 mul_1x1_ialu:
        mov     $a0,#0
        bic     $a1,$a,#3<<30           @ a1=a&0x3fffffff
@@ -147,7 +154,15 @@ mul_1x1_ialu:
 
        mov     pc,lr
 .size  mul_1x1_ialu,.-mul_1x1_ialu
+___
+################
+# void bn_GF2m_mul_2x2(BN_ULONG *r,
+#      BN_ULONG a1,BN_ULONG a0,
+#      BN_ULONG b1,BN_ULONG b0);       # r[3..0]=a1a0·b1b0
+
+($A1,$B1,$A0,$B0,$A1B1,$A0B0)=map("d$_",(18..23));
 
+$code.=<<___;
 .global        bn_GF2m_mul_2x2
 .type  bn_GF2m_mul_2x2,%function
 .align 5
@@ -157,9 +172,7 @@ bn_GF2m_mul_2x2:
 .Lpic: ldr     r12,[pc,r12]
        tst     r12,#1
        beq     .Lialu
-___
-($A1,$B1,$A0,$B0,$A0B0,$A1B1)=map("d$_",(18..23));
-$code.=<<___;
+
        veor    $A1,$A1
        vmov.32 $B1,r3,r3               @ two copies of b1
        vmov.32 ${A1}[0],r1             @ a1
@@ -205,38 +218,38 @@ $code.=<<___;
        mov     $b,r3                   @ $b=b1
        ldr     r3,[sp,#32]             @ load b0
        mov     $mask,#7<<2
-       sub     sp,#32                  @ allocate tab[8]
+       sub     sp,sp,#32               @ allocate tab[8]
 
        bl      mul_1x1_ialu            @ a1·b1
        str     $lo,[$ret,#8]
        str     $hi,[$ret,#12]
 
-       eor     $b,r3                   @ flip b0 and b1
-        eor    $a,r2                   @ flip a0 and a1
-       eor     r3,$b
-        eor    r2,$a
-       eor     $b,r3
-        eor    $a,r2
+       eor     $b,$b,r3                @ flip b0 and b1
+        eor    $a,$a,r2                @ flip a0 and a1
+       eor     r3,r3,$b
+        eor    r2,r2,$a
+       eor     $b,$b,r3
+        eor    $a,$a,r2
        bl      mul_1x1_ialu            @ a0·b0
        str     $lo,[$ret]
        str     $hi,[$ret,#4]
 
-       eor     $a,r2
-       eor     $b,r3
+       eor     $a,$a,r2
+       eor     $b,$b,r3
        bl      mul_1x1_ialu            @ (a1+a0)·(b1+b0)
 ___
 @r=map("r$_",(6..9));
 $code.=<<___;
        ldmia   $ret,{@r[0]-@r[3]}
-       eor     $lo,$hi
-       eor     $hi,@r[1]
-       eor     $lo,@r[0]
-       eor     $hi,@r[2]
-       eor     $lo,@r[3]
-       eor     $hi,@r[3]
+       eor     $lo,$lo,$hi
+       eor     $hi,$hi,@r[1]
+       eor     $lo,$lo,@r[0]
+       eor     $hi,$hi,@r[2]
+       eor     $lo,$lo,@r[3]
+       eor     $hi,$hi,@r[3]
        str     $hi,[$ret,#8]
-       eor     $lo,$hi
-       add     sp,#32                  @ destroy tab[8]
+       eor     $lo,$lo,$hi
+       add     sp,sp,#32               @ destroy tab[8]
        str     $lo,[$ret,#4]
 
 #if __ARM_ARCH__>=5
@@ -251,12 +264,12 @@ $code.=<<___;
 #if __ARM_ARCH__>=7
 .align 5
 .LOPENSSL_armcap:
-.word  OPENSSL_armcap-(.Lpic+8)
+.word  OPENSSL_armcap_P-(.Lpic+8)
 #endif
-.asciz "GF2m Multiplication for ARMv4/NEON, CRYPTOGAMS by <appro\@openssl.org>"
+.asciz "GF(2^m) Multiplication for ARMv4/NEON, CRYPTOGAMS by <appro\@openssl.org>"
 .align 5
 
-.comm  OPENSSL_armcap,4,4
+.comm  OPENSSL_armcap_P,4,4
 ___
 
 $code =~ s/\`([^\`]*)\`/eval $1/gem;