bn/asm/ppc64-mont.pl: adapt for little-endian.

The problem remained unnoticed so far, because it's never called by default.
You have to craft OPENSSL_ppccap environment variable to trigger the problem.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit e4693b4e2a0c3f6241d4d3e61460c34c7e0013f6)

diff --git a/crypto/bn/asm/ppc64-mont.pl b/crypto/bn/asm/ppc64-mont.pl
index 68e3733e3f79cd36c0b47b615c67dfb3b5f7011e..9e3c12d788e5fde089f6ea33e98b4913d2a44f5b 100644 (file)
--- a/crypto/bn/asm/ppc64-mont.pl
+++ b/crypto/bn/asm/ppc64-mont.pl
@@ -94,6 +94,8 @@ if ($flavour =~ /32/) {
        $POP=   "ld";
 } else { die "nonsense $flavour"; }
 
+$LITTLE_ENDIAN = ($flavour=~/le$/) ? 4 : 0;
+
 $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}ppc-xlate.pl" and -f $xlate ) or
 ( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or
@@ -294,12 +296,12 @@ $code.=<<___ if ($SIZE_T==8);
 
        extrdi  $t0,$a0,32,32           ; lwz   $t0,4($ap)
        extrdi  $t1,$a0,32,0            ; lwz   $t1,0($ap)
-       lwz     $t2,12($ap)             ; load a[1] as 32-bit word pair
-       lwz     $t3,8($ap)
-       lwz     $t4,4($np)              ; load n[0] as 32-bit word pair
-       lwz     $t5,0($np)
-       lwz     $t6,12($np)             ; load n[1] as 32-bit word pair
-       lwz     $t7,8($np)
+       lwz     $t2,`12^$LITTLE_ENDIAN`($ap)    ; load a[1] as 32-bit word pair
+       lwz     $t3,`8^$LITTLE_ENDIAN`($ap)
+       lwz     $t4,`4^$LITTLE_ENDIAN`($np)     ; load n[0] as 32-bit word pair
+       lwz     $t5,`0^$LITTLE_ENDIAN`($np)
+       lwz     $t6,`12^$LITTLE_ENDIAN`($np)    ; load n[1] as 32-bit word pair
+       lwz     $t7,`8^$LITTLE_ENDIAN`($np)
 ___
 $code.=<<___ if ($SIZE_T==4);
        lwz     $a0,0($ap)              ; pull ap[0,1] value
@@ -463,14 +465,14 @@ $code.=<<___;
 L1st:
 ___
 $code.=<<___ if ($SIZE_T==8);
-       lwz     $t0,4($ap)              ; load a[j] as 32-bit word pair
-       lwz     $t1,0($ap)
-       lwz     $t2,12($ap)             ; load a[j+1] as 32-bit word pair
-       lwz     $t3,8($ap)
-       lwz     $t4,4($np)              ; load n[j] as 32-bit word pair
-       lwz     $t5,0($np)
-       lwz     $t6,12($np)             ; load n[j+1] as 32-bit word pair
-       lwz     $t7,8($np)
+       lwz     $t0,`4^$LITTLE_ENDIAN`($ap)     ; load a[j] as 32-bit word pair
+       lwz     $t1,`0^$LITTLE_ENDIAN`($ap)
+       lwz     $t2,`12^$LITTLE_ENDIAN`($ap)    ; load a[j+1] as 32-bit word pair
+       lwz     $t3,`8^$LITTLE_ENDIAN`($ap)
+       lwz     $t4,`4^$LITTLE_ENDIAN`($np)     ; load n[j] as 32-bit word pair
+       lwz     $t5,`0^$LITTLE_ENDIAN`($np)
+       lwz     $t6,`12^$LITTLE_ENDIAN`($np)    ; load n[j+1] as 32-bit word pair
+       lwz     $t7,`8^$LITTLE_ENDIAN`($np)
 ___
 $code.=<<___ if ($SIZE_T==4);
        lwz     $t0,0($ap)              ; load a[j..j+3] as 32-bit word pairs
@@ -505,14 +507,14 @@ $code.=<<___;
 ___
 } else {
 $code.=<<___;
-       lwz     $t1,`$FRAME+0`($sp)
-       lwz     $t0,`$FRAME+4`($sp)
-       lwz     $t3,`$FRAME+8`($sp)
-       lwz     $t2,`$FRAME+12`($sp)
-       lwz     $t5,`$FRAME+16`($sp)
-       lwz     $t4,`$FRAME+20`($sp)
-       lwz     $t7,`$FRAME+24`($sp)
-       lwz     $t6,`$FRAME+28`($sp)
+       lwz     $t1,`$FRAME+0^$LITTLE_ENDIAN`($sp)
+       lwz     $t0,`$FRAME+4^$LITTLE_ENDIAN`($sp)
+       lwz     $t3,`$FRAME+8^$LITTLE_ENDIAN`($sp)
+       lwz     $t2,`$FRAME+12^$LITTLE_ENDIAN`($sp)
+       lwz     $t5,`$FRAME+16^$LITTLE_ENDIAN`($sp)
+       lwz     $t4,`$FRAME+20^$LITTLE_ENDIAN`($sp)
+       lwz     $t7,`$FRAME+24^$LITTLE_ENDIAN`($sp)
+       lwz     $t6,`$FRAME+28^$LITTLE_ENDIAN`($sp)
 ___
 }
 $code.=<<___;
@@ -651,8 +653,8 @@ $code.=<<___;
 
        fmadd   $T1a,$N1,$na,$T1a
        fmadd   $T1b,$N1,$nb,$T1b
-        lwz    $t3,`$FRAME+32`($sp)    ; permuted $t1
-        lwz    $t2,`$FRAME+36`($sp)    ; permuted $t0
+        lwz    $t3,`$FRAME+32^$LITTLE_ENDIAN`($sp)     ; permuted $t1
+        lwz    $t2,`$FRAME+36^$LITTLE_ENDIAN`($sp)     ; permuted $t0
         addc   $t4,$t4,$carry
         adde   $t5,$t5,$c1
         srwi   $carry,$t4,16
@@ -673,8 +675,8 @@ $code.=<<___;
 
        fmadd   $T1a,$N0,$nc,$T1a
        fmadd   $T1b,$N0,$nd,$T1b
-        lwz    $t7,`$FRAME+40`($sp)    ; permuted $t3
-        lwz    $t6,`$FRAME+44`($sp)    ; permuted $t2
+        lwz    $t7,`$FRAME+40^$LITTLE_ENDIAN`($sp)     ; permuted $t3
+        lwz    $t6,`$FRAME+44^$LITTLE_ENDIAN`($sp)     ; permuted $t2
         addc   $t2,$t2,$carry
         adde   $t3,$t3,$c1
         srwi   $carry,$t2,16
@@ -686,8 +688,8 @@ $code.=<<___;
         insrwi $carry,$t3,16,0
        fmadd   $T3a,$N2,$nc,$T3a
        fmadd   $T3b,$N2,$nd,$T3b
-        lwz    $t1,`$FRAME+48`($sp)    ; permuted $t5
-        lwz    $t0,`$FRAME+52`($sp)    ; permuted $t4
+        lwz    $t1,`$FRAME+48^$LITTLE_ENDIAN`($sp)     ; permuted $t5
+        lwz    $t0,`$FRAME+52^$LITTLE_ENDIAN`($sp)     ; permuted $t4
         addc   $t6,$t6,$carry
         adde   $t7,$t7,$c1
         srwi   $carry,$t6,16
@@ -699,8 +701,8 @@ $code.=<<___;
 
        fctid   $T0a,$T0a
        fctid   $T0b,$T0b
-        lwz    $t5,`$FRAME+56`($sp)    ; permuted $t7
-        lwz    $t4,`$FRAME+60`($sp)    ; permuted $t6
+        lwz    $t5,`$FRAME+56^$LITTLE_ENDIAN`($sp)     ; permuted $t7
+        lwz    $t4,`$FRAME+60^$LITTLE_ENDIAN`($sp)     ; permuted $t6
         addc   $t0,$t0,$carry
         adde   $t1,$t1,$c1
         srwi   $carry,$t0,16
@@ -787,14 +789,14 @@ $code.=<<___;
 ___
 } else {
 $code.=<<___;
-       lwz     $t1,`$FRAME+0`($sp)
-       lwz     $t0,`$FRAME+4`($sp)
-       lwz     $t3,`$FRAME+8`($sp)
-       lwz     $t2,`$FRAME+12`($sp)
-       lwz     $t5,`$FRAME+16`($sp)
-       lwz     $t4,`$FRAME+20`($sp)
-       lwz     $t7,`$FRAME+24`($sp)
-       lwz     $t6,`$FRAME+28`($sp)
+       lwz     $t1,`$FRAME+0^$LITTLE_ENDIAN`($sp)
+       lwz     $t0,`$FRAME+4^$LITTLE_ENDIAN`($sp)
+       lwz     $t3,`$FRAME+8^$LITTLE_ENDIAN`($sp)
+       lwz     $t2,`$FRAME+12^$LITTLE_ENDIAN`($sp)
+       lwz     $t5,`$FRAME+16^$LITTLE_ENDIAN`($sp)
+       lwz     $t4,`$FRAME+20^$LITTLE_ENDIAN`($sp)
+       lwz     $t7,`$FRAME+24^$LITTLE_ENDIAN`($sp)
+       lwz     $t6,`$FRAME+28^$LITTLE_ENDIAN`($sp)
        stfd    $dota,`$FRAME+64`($sp)
        stfd    $dotb,`$FRAME+72`($sp)
 
@@ -823,14 +825,14 @@ $code.=<<___;
         stw    $t0,12($tp)             ; tp[j-1]
         stw    $t4,8($tp)
 
-       lwz     $t3,`$FRAME+32`($sp)    ; permuted $t1
-       lwz     $t2,`$FRAME+36`($sp)    ; permuted $t0
-       lwz     $t7,`$FRAME+40`($sp)    ; permuted $t3
-       lwz     $t6,`$FRAME+44`($sp)    ; permuted $t2
-       lwz     $t1,`$FRAME+48`($sp)    ; permuted $t5
-       lwz     $t0,`$FRAME+52`($sp)    ; permuted $t4
-       lwz     $t5,`$FRAME+56`($sp)    ; permuted $t7
-       lwz     $t4,`$FRAME+60`($sp)    ; permuted $t6
+       lwz     $t3,`$FRAME+32^$LITTLE_ENDIAN`($sp)     ; permuted $t1
+       lwz     $t2,`$FRAME+36^$LITTLE_ENDIAN`($sp)     ; permuted $t0
+       lwz     $t7,`$FRAME+40^$LITTLE_ENDIAN`($sp)     ; permuted $t3
+       lwz     $t6,`$FRAME+44^$LITTLE_ENDIAN`($sp)     ; permuted $t2
+       lwz     $t1,`$FRAME+48^$LITTLE_ENDIAN`($sp)     ; permuted $t5
+       lwz     $t0,`$FRAME+52^$LITTLE_ENDIAN`($sp)     ; permuted $t4
+       lwz     $t5,`$FRAME+56^$LITTLE_ENDIAN`($sp)     ; permuted $t7
+       lwz     $t4,`$FRAME+60^$LITTLE_ENDIAN`($sp)     ; permuted $t6
 
        addc    $t2,$t2,$carry
        adde    $t3,$t3,$c1
@@ -857,10 +859,10 @@ $code.=<<___;
         stw    $t2,20($tp)             ; tp[j]
         stwu   $t0,16($tp)
 
-       lwz     $t7,`$FRAME+64`($sp)
-       lwz     $t6,`$FRAME+68`($sp)
-       lwz     $t5,`$FRAME+72`($sp)
-       lwz     $t4,`$FRAME+76`($sp)
+       lwz     $t7,`$FRAME+64^$LITTLE_ENDIAN`($sp)
+       lwz     $t6,`$FRAME+68^$LITTLE_ENDIAN`($sp)
+       lwz     $t5,`$FRAME+72^$LITTLE_ENDIAN`($sp)
+       lwz     $t4,`$FRAME+76^$LITTLE_ENDIAN`($sp)
 
        addc    $t6,$t6,$carry
        adde    $t7,$t7,$c1
@@ -1165,23 +1167,23 @@ ___
 $code.=<<___;
        fmadd   $T1a,$N1,$na,$T1a
        fmadd   $T1b,$N1,$nb,$T1b
-        lwz    $t1,`$FRAME+0`($sp)
-        lwz    $t0,`$FRAME+4`($sp)
+        lwz    $t1,`$FRAME+0^$LITTLE_ENDIAN`($sp)
+        lwz    $t0,`$FRAME+4^$LITTLE_ENDIAN`($sp)
        fmadd   $T2a,$N2,$na,$T2a
        fmadd   $T2b,$N2,$nb,$T2b
-        lwz    $t3,`$FRAME+8`($sp)
-        lwz    $t2,`$FRAME+12`($sp)
+        lwz    $t3,`$FRAME+8^$LITTLE_ENDIAN`($sp)
+        lwz    $t2,`$FRAME+12^$LITTLE_ENDIAN`($sp)
        fmadd   $T3a,$N3,$na,$T3a
        fmadd   $T3b,$N3,$nb,$T3b
-        lwz    $t5,`$FRAME+16`($sp)
-        lwz    $t4,`$FRAME+20`($sp)
+        lwz    $t5,`$FRAME+16^$LITTLE_ENDIAN`($sp)
+        lwz    $t4,`$FRAME+20^$LITTLE_ENDIAN`($sp)
         addc   $t0,$t0,$carry
         adde   $t1,$t1,$c1
         srwi   $carry,$t0,16
        fmadd   $T0a,$N0,$na,$T0a
        fmadd   $T0b,$N0,$nb,$T0b
-        lwz    $t7,`$FRAME+24`($sp)
-        lwz    $t6,`$FRAME+28`($sp)
+        lwz    $t7,`$FRAME+24^$LITTLE_ENDIAN`($sp)
+        lwz    $t6,`$FRAME+28^$LITTLE_ENDIAN`($sp)
         srwi   $c1,$t1,16
         insrwi $carry,$t1,16,0
 
@@ -1218,8 +1220,8 @@ $code.=<<___;
        fctid   $T1a,$T1a
         addc   $t0,$t0,$t2
         adde   $t4,$t4,$t3
-        lwz    $t3,`$FRAME+32`($sp)    ; permuted $t1
-        lwz    $t2,`$FRAME+36`($sp)    ; permuted $t0
+        lwz    $t3,`$FRAME+32^$LITTLE_ENDIAN`($sp)     ; permuted $t1
+        lwz    $t2,`$FRAME+36^$LITTLE_ENDIAN`($sp)     ; permuted $t0
        fctid   $T1b,$T1b
         addze  $carry,$carry
         addze  $c1,$c1
@@ -1229,19 +1231,19 @@ $code.=<<___;
         addc   $t2,$t2,$carry
         adde   $t3,$t3,$c1
         srwi   $carry,$t2,16
-        lwz    $t7,`$FRAME+40`($sp)    ; permuted $t3
-        lwz    $t6,`$FRAME+44`($sp)    ; permuted $t2
+        lwz    $t7,`$FRAME+40^$LITTLE_ENDIAN`($sp)     ; permuted $t3
+        lwz    $t6,`$FRAME+44^$LITTLE_ENDIAN`($sp)     ; permuted $t2
        fctid   $T2b,$T2b
         srwi   $c1,$t3,16
         insrwi $carry,$t3,16,0
-        lwz    $t1,`$FRAME+48`($sp)    ; permuted $t5
-        lwz    $t0,`$FRAME+52`($sp)    ; permuted $t4
+        lwz    $t1,`$FRAME+48^$LITTLE_ENDIAN`($sp)     ; permuted $t5
+        lwz    $t0,`$FRAME+52^$LITTLE_ENDIAN`($sp)     ; permuted $t4
        fctid   $T3a,$T3a
         addc   $t6,$t6,$carry
         adde   $t7,$t7,$c1
         srwi   $carry,$t6,16
-        lwz    $t5,`$FRAME+56`($sp)    ; permuted $t7
-        lwz    $t4,`$FRAME+60`($sp)    ; permuted $t6
+        lwz    $t5,`$FRAME+56^$LITTLE_ENDIAN`($sp)     ; permuted $t7
+        lwz    $t4,`$FRAME+60^$LITTLE_ENDIAN`($sp)     ; permuted $t6
        fctid   $T3b,$T3b
 
         insrwi $t2,$t6,16,0            ; 64..95 bits
@@ -1354,14 +1356,14 @@ $code.=<<___;
 ___
 } else {
 $code.=<<___;
-       lwz     $t1,`$FRAME+0`($sp)
-       lwz     $t0,`$FRAME+4`($sp)
-       lwz     $t3,`$FRAME+8`($sp)
-       lwz     $t2,`$FRAME+12`($sp)
-       lwz     $t5,`$FRAME+16`($sp)
-       lwz     $t4,`$FRAME+20`($sp)
-       lwz     $t7,`$FRAME+24`($sp)
-       lwz     $t6,`$FRAME+28`($sp)
+       lwz     $t1,`$FRAME+0^$LITTLE_ENDIAN`($sp)
+       lwz     $t0,`$FRAME+4^$LITTLE_ENDIAN`($sp)
+       lwz     $t3,`$FRAME+8^$LITTLE_ENDIAN`($sp)
+       lwz     $t2,`$FRAME+12^$LITTLE_ENDIAN`($sp)
+       lwz     $t5,`$FRAME+16^$LITTLE_ENDIAN`($sp)
+       lwz     $t4,`$FRAME+20^$LITTLE_ENDIAN`($sp)
+       lwz     $t7,`$FRAME+24^$LITTLE_ENDIAN`($sp)
+       lwz     $t6,`$FRAME+28^$LITTLE_ENDIAN`($sp)
        stfd    $dota,`$FRAME+64`($sp)
        stfd    $dotb,`$FRAME+72`($sp)
 
@@ -1397,14 +1399,14 @@ $code.=<<___;
         stw    $t0,4($tp)              ; tp[j-1]
         stw    $t4,0($tp)
 
-       lwz     $t3,`$FRAME+32`($sp)    ; permuted $t1
-       lwz     $t2,`$FRAME+36`($sp)    ; permuted $t0
-       lwz     $t7,`$FRAME+40`($sp)    ; permuted $t3
-       lwz     $t6,`$FRAME+44`($sp)    ; permuted $t2
-       lwz     $t1,`$FRAME+48`($sp)    ; permuted $t5
-       lwz     $t0,`$FRAME+52`($sp)    ; permuted $t4
-       lwz     $t5,`$FRAME+56`($sp)    ; permuted $t7
-       lwz     $t4,`$FRAME+60`($sp)    ; permuted $t6
+       lwz     $t3,`$FRAME+32^$LITTLE_ENDIAN`($sp)     ; permuted $t1
+       lwz     $t2,`$FRAME+36^$LITTLE_ENDIAN`($sp)     ; permuted $t0
+       lwz     $t7,`$FRAME+40^$LITTLE_ENDIAN`($sp)     ; permuted $t3
+       lwz     $t6,`$FRAME+44^$LITTLE_ENDIAN`($sp)     ; permuted $t2
+       lwz     $t1,`$FRAME+48^$LITTLE_ENDIAN`($sp)     ; permuted $t5
+       lwz     $t0,`$FRAME+52^$LITTLE_ENDIAN`($sp)     ; permuted $t4
+       lwz     $t5,`$FRAME+56^$LITTLE_ENDIAN`($sp)     ; permuted $t7
+       lwz     $t4,`$FRAME+60^$LITTLE_ENDIAN`($sp)     ; permuted $t6
 
        addc    $t2,$t2,$carry
        adde    $t3,$t3,$c1
@@ -1433,12 +1435,12 @@ $code.=<<___;
 
        addc    $t2,$t2,$t6
        adde    $t0,$t0,$t7
-        lwz    $t7,`$FRAME+64`($sp)
-        lwz    $t6,`$FRAME+68`($sp)
+        lwz    $t7,`$FRAME+64^$LITTLE_ENDIAN`($sp)
+        lwz    $t6,`$FRAME+68^$LITTLE_ENDIAN`($sp)
        addze   $carry,$carry
        addze   $c1,$c1
-        lwz    $t5,`$FRAME+72`($sp)
-        lwz    $t4,`$FRAME+76`($sp)
+        lwz    $t5,`$FRAME+72^$LITTLE_ENDIAN`($sp)
+        lwz    $t4,`$FRAME+76^$LITTLE_ENDIAN`($sp)
 
        addc    $t6,$t6,$carry
        adde    $t7,$t7,$c1