Remove trailing whitespace from some files.

The prevailing style seems to not have trailing whitespace, but a few
lines do. This is mostly in the perlasm files, but a few C files got
them after the reformat. This is the result of:

  find . -name '*.pl' | xargs sed -E -i '' -e 's/( |'$'\t'')*$//'
  find . -name '*.c' | xargs sed -E -i '' -e 's/( |'$'\t'')*$//'
  find . -name '*.h' | xargs sed -E -i '' -e 's/( |'$'\t'')*$//'

Then bn_prime.h was excluded since this is a generated file.

Note mkerr.pl has some changes in a heredoc for some help output, but
other lines there lack trailing whitespace too.

Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>

15 files changed, 192 insertions, 192 deletions

diff --git a/crypto/bn/asm/armv4-gf2m.pl b/crypto/bn/asm/armv4-gf2m.pl
index 0bb5433075..7a0cdb2e8a 100644
--- a/crypto/bn/asm/armv4-gf2m.pl
+++ b/crypto/bn/asm/armv4-gf2m.pl
@@ -36,7 +36,7 @@
 #
 # Câmara, D.; Gouvêa, C. P. L.; López, J. & Dahab, R.: Fast Software
 # Polynomial Multiplication on ARM Processors using the NEON Engine.
-# 
+#
 # http://conradoplg.cryptoland.net/files/2010/12/mocrysen13.pdf
 
 $flavour = shift;
diff --git a/crypto/bn/asm/armv4-mont.pl b/crypto/bn/asm/armv4-mont.pl
index 0dc4fe95e4..75a36f62fa 100644
--- a/crypto/bn/asm/armv4-mont.pl
+++ b/crypto/bn/asm/armv4-mont.pl
@@ -23,7 +23,7 @@
 # [depending on key length, less for longer keys] on ARM920T, and
 # +115-80% on Intel IXP425. This is compared to pre-bn_mul_mont code
 # base and compiler generated code with in-lined umull and even umlal
-# instructions. The latter means that this code didn't really have an 
+# instructions. The latter means that this code didn't really have an
 # "advantage" of utilizing some "secret" instruction.
 #
 # The code is interoperable with Thumb ISA and is rather compact, less
diff --git a/crypto/bn/asm/bn-586.pl b/crypto/bn/asm/bn-586.pl
index 1ca1bbf7d4..1350bcd8fa 100644
--- a/crypto/bn/asm/bn-586.pl
+++ b/crypto/bn/asm/bn-586.pl
@@ -54,7 +54,7 @@ sub bn_mul_add_words
 		&movd("mm0",&wparam(3));	# mm0 = w
 		&pxor("mm1","mm1");		# mm1 = carry_in
 		&jmp(&label("maw_sse2_entry"));
-		
+
 	&set_label("maw_sse2_unrolled",16);
 		&movd("mm3",&DWP(0,$r,"",0));	# mm3 = r[0]
 		&paddq("mm1","mm3");		# mm1 = carry_in + r[0]
@@ -675,20 +675,20 @@ sub bn_sub_part_words
 	    &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");
@@ -705,9 +705,9 @@ sub bn_sub_part_words
 	}
 
 	&jmp(&label("pw_end"));
-	
+
 	&set_label("pw_pos",0);
-	
+
 	&and($num,0xfffffff8);	# num / 8
 	&jz(&label("pw_pos_finish"));
 
@@ -722,18 +722,18 @@ sub bn_sub_part_words
 	    &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 $i");
@@ -754,17 +754,17 @@ sub bn_sub_part_words
 	    &mov(&DWP($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
diff --git a/crypto/bn/asm/co-586.pl b/crypto/bn/asm/co-586.pl
index 60d0363660..6f34c37cf8 100644
--- a/crypto/bn/asm/co-586.pl
+++ b/crypto/bn/asm/co-586.pl
@@ -47,7 +47,7 @@ sub mul_add_c
 	 &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 
+	 # 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 $pos == 1;		# laod next a
 	}
@@ -76,7 +76,7 @@ sub sqr_add_c
 	 &mov("edx",&DWP(($nb)*4,$a,"",0)) if ($pos == 1) && ($na != $nb);
 	 ###
 	&adc($c2,0);
-	 # is pos > 1, it means it is the last loop 
+	 # 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
 	}
@@ -127,7 +127,7 @@ sub bn_mul_comba
 	$c2="ebp";
 	$a="esi";
 	$b="edi";
-	
+
 	$as=0;
 	$ae=0;
 	$bs=0;
@@ -142,9 +142,9 @@ sub bn_mul_comba
 	 &push("ebx");
 
 	&xor($c0,$c0);
-	 &mov("eax",&DWP(0,$a,"",0));	# load the first word 
+	 &mov("eax",&DWP(0,$a,"",0));	# load the first word
 	&xor($c1,$c1);
-	 &mov("edx",&DWP(0,$b,"",0));	# load the first second 
+	 &mov("edx",&DWP(0,$b,"",0));	# load the first second
 
 	for ($i=0; $i<$tot; $i++)
 		{
@@ -152,7 +152,7 @@ sub bn_mul_comba
 		$bi=$bs;
 		$end=$be+1;
 
-		&comment("################## Calculate word $i"); 
+		&comment("################## Calculate word $i");
 
 		for ($j=$bs; $j<$end; $j++)
 			{
diff --git a/crypto/bn/asm/ia64-mont.pl b/crypto/bn/asm/ia64-mont.pl
index 5cc5c599f9..233fdd47a9 100644
--- a/crypto/bn/asm/ia64-mont.pl
+++ b/crypto/bn/asm/ia64-mont.pl
@@ -80,7 +80,7 @@ $code=<<___;
 
 // int bn_mul_mont (BN_ULONG *rp,const BN_ULONG *ap,
 //		    const BN_ULONG *bp,const BN_ULONG *np,
-//		    const BN_ULONG *n0p,int num);			
+//		    const BN_ULONG *n0p,int num);
 .align	64
 .global	bn_mul_mont#
 .proc	bn_mul_mont#
@@ -203,7 +203,7 @@ bn_mul_mont_general:
 { .mmi;	.pred.rel	"mutex",p39,p41
 (p39)	add		topbit=r0,r0
 (p41)	add		topbit=r0,r0,1
-	nop.i		0		}	
+	nop.i		0		}
 { .mmi;	st8		[tp_1]=n[0]
 	add		tptr=16,sp
 	add		tp_1=8,sp	};;
diff --git a/crypto/bn/asm/mips.pl b/crypto/bn/asm/mips.pl
index 102b656229..5093177552 100644
--- a/crypto/bn/asm/mips.pl
+++ b/crypto/bn/asm/mips.pl
@@ -603,13 +603,13 @@ $code.=<<___;
 	sltu	$v0,$t2,$ta2
 	$ST	$t2,-2*$BNSZ($a0)
 	$ADDU	$v0,$t8
-	
+
 	$ADDU	$ta3,$t3
 	sltu	$t9,$ta3,$t3
 	$ADDU	$t3,$ta3,$v0
 	sltu	$v0,$t3,$ta3
 	$ST	$t3,-$BNSZ($a0)
-	
+
 	.set	noreorder
 	bgtz	$at,.L_bn_add_words_loop
 	$ADDU	$v0,$t9
@@ -808,7 +808,7 @@ bn_div_3_words:
 				# so that we can save two arguments
 				# and return address in registers
 				# instead of stack:-)
-				
+
 	$LD	$a0,($a3)
 	move	$ta2,$a1
 	bne	$a0,$a2,bn_div_3_words_internal
diff --git a/crypto/bn/asm/parisc-mont.pl b/crypto/bn/asm/parisc-mont.pl
index 8aa94e8511..61c3625a3c 100644
--- a/crypto/bn/asm/parisc-mont.pl
+++ b/crypto/bn/asm/parisc-mont.pl
@@ -546,7 +546,7 @@ L\$copy
 	ldd		$idx($np),$hi0
 	std,ma		%r0,8($tp)
 	addib,<>	8,$idx,.-8		; L\$copy
-	std,ma		$hi0,8($rp)	
+	std,ma		$hi0,8($rp)
 ___
 
 if ($BN_SZ==4) {				# PA-RISC 1.1 code-path
@@ -868,7 +868,7 @@ L\$copy_pa11
 	ldwx		$idx($np),$hi0
 	stws,ma		%r0,4($tp)
 	addib,<>	4,$idx,L\$copy_pa11
-	stws,ma		$hi0,4($rp)	
+	stws,ma		$hi0,4($rp)
 
 	nop					; alignment
 L\$done
diff --git a/crypto/bn/asm/ppc-mont.pl b/crypto/bn/asm/ppc-mont.pl
index 5802260ca6..7a25b1ec9b 100644
--- a/crypto/bn/asm/ppc-mont.pl
+++ b/crypto/bn/asm/ppc-mont.pl
@@ -26,7 +26,7 @@
 # So far RSA *sign* performance improvement over pre-bn_mul_mont asm
 # for 64-bit application running on PPC970/G5 is:
 #
-# 512-bit	+65%	
+# 512-bit	+65%
 # 1024-bit	+35%
 # 2048-bit	+18%
 # 4096-bit	+4%
@@ -49,7 +49,7 @@ if ($flavour =~ /32/) {
 	$UMULL=	"mullw";	# unsigned multiply low
 	$UMULH=	"mulhwu";	# unsigned multiply high
 	$UCMP=	"cmplw";	# unsigned compare
-	$SHRI=	"srwi";		# unsigned shift right by immediate	
+	$SHRI=	"srwi";		# unsigned shift right by immediate
 	$PUSH=	$ST;
 	$POP=	$LD;
 } elsif ($flavour =~ /64/) {
@@ -69,7 +69,7 @@ if ($flavour =~ /32/) {
 	$UMULL=	"mulld";	# unsigned multiply low
 	$UMULH=	"mulhdu";	# unsigned multiply high
 	$UCMP=	"cmpld";	# unsigned compare
-	$SHRI=	"srdi";		# unsigned shift right by immediate	
+	$SHRI=	"srdi";		# unsigned shift right by immediate
 	$PUSH=	$ST;
 	$POP=	$LD;
 } else { die "nonsense $flavour"; }
diff --git a/crypto/bn/asm/ppc.pl b/crypto/bn/asm/ppc.pl
index e9262df0f3..1a03f4561e 100644
--- a/crypto/bn/asm/ppc.pl
+++ b/crypto/bn/asm/ppc.pl
@@ -38,7 +38,7 @@
 #rsa 2048 bits   0.3036s   0.0085s      3.3    117.1
 #rsa 4096 bits   2.0040s   0.0299s      0.5     33.4
 #dsa  512 bits   0.0087s   0.0106s    114.3     94.5
-#dsa 1024 bits   0.0256s   0.0313s     39.0     32.0	
+#dsa 1024 bits   0.0256s   0.0313s     39.0     32.0
 #
 #	Same bechmark with this assembler code:
 #
@@ -74,7 +74,7 @@
 #rsa 4096 bits   0.3700s   0.0058s      2.7    171.0
 #dsa  512 bits   0.0016s   0.0020s    610.7    507.1
 #dsa 1024 bits   0.0047s   0.0058s    212.5    173.2
-#	
+#
 #	Again, performance increases by at about 75%
 #
 #       Mac OS X, Apple G5 1.8GHz (Note this is 32 bit code)
@@ -125,7 +125,7 @@ if ($flavour =~ /32/) {
 	$CNTLZ=	"cntlzw";	# count leading zeros
 	$SHL=	"slw";		# shift left
 	$SHR=	"srw";		# unsigned shift right
-	$SHRI=	"srwi";		# unsigned shift right by immediate	
+	$SHRI=	"srwi";		# unsigned shift right by immediate
 	$SHLI=	"slwi";		# shift left by immediate
 	$CLRU=	"clrlwi";	# clear upper bits
 	$INSR=	"insrwi";	# insert right
@@ -149,10 +149,10 @@ if ($flavour =~ /32/) {
 	$CNTLZ=	"cntlzd";	# count leading zeros
 	$SHL=	"sld";		# shift left
 	$SHR=	"srd";		# unsigned shift right
-	$SHRI=	"srdi";		# unsigned shift right by immediate	
+	$SHRI=	"srdi";		# unsigned shift right by immediate
 	$SHLI=	"sldi";		# shift left by immediate
 	$CLRU=	"clrldi";	# clear upper bits
-	$INSR=	"insrdi";	# insert right 
+	$INSR=	"insrdi";	# insert right
 	$ROTL=	"rotldi";	# rotate left by immediate
 	$TR=	"td";		# conditional trap
 } else { die "nonsense $flavour"; }
@@ -189,7 +189,7 @@ $data=<<EOF;
 #	   below.
 #				12/05/03		Suresh Chari
 #			(with lots of help from)        Andy Polyakov
-##	
+##
 #	1. Initial version	10/20/02		Suresh Chari
 #
 #
@@ -202,7 +202,7 @@ $data=<<EOF;
 #		be done in the build process.
 #
 #	Hand optimized assembly code for the following routines
-#	
+#
 #	bn_sqr_comba4
 #	bn_sqr_comba8
 #	bn_mul_comba4
@@ -225,10 +225,10 @@ $data=<<EOF;
 #--------------------------------------------------------------------------
 #
 #	Defines to be used in the assembly code.
-#	
+#
 #.set r0,0	# we use it as storage for value of 0
 #.set SP,1	# preserved
-#.set RTOC,2	# preserved 
+#.set RTOC,2	# preserved
 #.set r3,3	# 1st argument/return value
 #.set r4,4	# 2nd argument/volatile register
 #.set r5,5	# 3rd argument/volatile register
@@ -246,7 +246,7 @@ $data=<<EOF;
 #	        the first . i.e. for example change ".bn_sqr_comba4"
 #		to "bn_sqr_comba4". This should be automatically done
 #		in the build.
-	
+
 	.globl	.bn_sqr_comba4
 	.globl	.bn_sqr_comba8
 	.globl	.bn_mul_comba4
@@ -257,9 +257,9 @@ $data=<<EOF;
 	.globl	.bn_sqr_words
 	.globl	.bn_mul_words
 	.globl	.bn_mul_add_words
-	
+
 # .text section
-	
+
 	.machine	"any"
 
 #
@@ -278,8 +278,8 @@ $data=<<EOF;
 # r3 contains r
 # r4 contains a
 #
-# Freely use registers r5,r6,r7,r8,r9,r10,r11 as follows:	
-# 
+# Freely use registers r5,r6,r7,r8,r9,r10,r11 as follows:
+#
 # r5,r6 are the two BN_ULONGs being multiplied.
 # r7,r8 are the results of the 32x32 giving 64 bit multiply.
 # r9,r10, r11 are the equivalents of c1,c2, c3.
@@ -288,10 +288,10 @@ $data=<<EOF;
 #
 	xor		r0,r0,r0		# set r0 = 0. Used in the addze
 						# instructions below
-	
+
 						#sqr_add_c(a,0,c1,c2,c3)
-	$LD		r5,`0*$BNSZ`(r4)		
-	$UMULL		r9,r5,r5		
+	$LD		r5,`0*$BNSZ`(r4)
+	$UMULL		r9,r5,r5
 	$UMULH		r10,r5,r5		#in first iteration. No need
 						#to add since c1=c2=c3=0.
 						# Note c3(r11) is NOT set to 0
@@ -299,20 +299,20 @@ $data=<<EOF;
 
 	$ST		r9,`0*$BNSZ`(r3)	# r[0]=c1;
 						# sqr_add_c2(a,1,0,c2,c3,c1);
-	$LD		r6,`1*$BNSZ`(r4)		
+	$LD		r6,`1*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-					
+
 	addc		r7,r7,r7		# compute (r7,r8)=2*(r7,r8)
 	adde		r8,r8,r8
 	addze		r9,r0			# catch carry if any.
-						# r9= r0(=0) and carry 
-	
+						# r9= r0(=0) and carry
+
 	addc		r10,r7,r10		# now add to temp result.
-	addze		r11,r8                  # r8 added to r11 which is 0 
+	addze		r11,r8                  # r8 added to r11 which is 0
 	addze		r9,r9
-	
-	$ST		r10,`1*$BNSZ`(r3)	#r[1]=c2; 
+
+	$ST		r10,`1*$BNSZ`(r3)	#r[1]=c2;
 						#sqr_add_c(a,1,c3,c1,c2)
 	$UMULL		r7,r6,r6
 	$UMULH		r8,r6,r6
@@ -323,23 +323,23 @@ $data=<<EOF;
 	$LD		r6,`2*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r7,r7,r7
 	adde		r8,r8,r8
 	addze		r10,r10
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
-	$ST		r11,`2*$BNSZ`(r3)	#r[2]=c3 
+	$ST		r11,`2*$BNSZ`(r3)	#r[2]=c3
 						#sqr_add_c2(a,3,0,c1,c2,c3);
-	$LD		r6,`3*$BNSZ`(r4)		
+	$LD		r6,`3*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
 	addc		r7,r7,r7
 	adde		r8,r8,r8
 	addze		r11,r0
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
@@ -348,7 +348,7 @@ $data=<<EOF;
 	$LD		r6,`2*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r7,r7,r7
 	adde		r8,r8,r8
 	addze		r11,r11
@@ -363,31 +363,31 @@ $data=<<EOF;
 	adde		r11,r8,r11
 	addze		r9,r0
 						#sqr_add_c2(a,3,1,c2,c3,c1);
-	$LD		r6,`3*$BNSZ`(r4)		
+	$LD		r6,`3*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
 	addc		r7,r7,r7
 	adde		r8,r8,r8
 	addze		r9,r9
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
 	$ST		r10,`4*$BNSZ`(r3)	#r[4]=c2
 						#sqr_add_c2(a,3,2,c3,c1,c2);
-	$LD		r5,`2*$BNSZ`(r4)		
+	$LD		r5,`2*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
 	addc		r7,r7,r7
 	adde		r8,r8,r8
 	addze		r10,r0
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
 	$ST		r11,`5*$BNSZ`(r3)	#r[5] = c3
 						#sqr_add_c(a,3,c1,c2,c3);
-	$UMULL		r7,r6,r6		
+	$UMULL		r7,r6,r6
 	$UMULH		r8,r6,r6
 	addc		r9,r7,r9
 	adde		r10,r8,r10
@@ -406,7 +406,7 @@ $data=<<EOF;
 #		for the gcc compiler. This should be automatically
 #		done in the build
 #
-	
+
 .align	4
 .bn_sqr_comba8:
 #
@@ -418,15 +418,15 @@ $data=<<EOF;
 # r3 contains r
 # r4 contains a
 #
-# Freely use registers r5,r6,r7,r8,r9,r10,r11 as follows:	
-# 
+# Freely use registers r5,r6,r7,r8,r9,r10,r11 as follows:
+#
 # r5,r6 are the two BN_ULONGs being multiplied.
 # r7,r8 are the results of the 32x32 giving 64 bit multiply.
 # r9,r10, r11 are the equivalents of c1,c2, c3.
 #
 # Possible optimization of loading all 8 longs of a into registers
 # doesn't provide any speedup
-# 
+#
 
 	xor		r0,r0,r0		#set r0 = 0.Used in addze
 						#instructions below.
@@ -439,18 +439,18 @@ $data=<<EOF;
 						#sqr_add_c2(a,1,0,c2,c3,c1);
 	$LD		r6,`1*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
-	$UMULH		r8,r5,r6	
-	
+	$UMULH		r8,r5,r6
+
 	addc		r10,r7,r10		#add the two register number
 	adde		r11,r8,r0 		# (r8,r7) to the three register
 	addze		r9,r0			# number (r9,r11,r10).NOTE:r0=0
-	
+
 	addc		r10,r7,r10		#add the two register number
 	adde		r11,r8,r11 		# (r8,r7) to the three register
 	addze		r9,r9			# number (r9,r11,r10).
-	
+
 	$ST		r10,`1*$BNSZ`(r3)	# r[1]=c2
-				
+
 						#sqr_add_c(a,1,c3,c1,c2);
 	$UMULL		r7,r6,r6
 	$UMULH		r8,r6,r6
@@ -461,25 +461,25 @@ $data=<<EOF;
 	$LD		r6,`2*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
-	
+
 	$ST		r11,`2*$BNSZ`(r3)	#r[2]=c3
 						#sqr_add_c2(a,3,0,c1,c2,c3);
 	$LD		r6,`3*$BNSZ`(r4)	#r6 = a[3]. r5 is already a[0].
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r0
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
@@ -488,20 +488,20 @@ $data=<<EOF;
 	$LD		r6,`2*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
-	
+
 	$ST		r9,`3*$BNSZ`(r3)	#r[3]=c1;
 						#sqr_add_c(a,2,c2,c3,c1);
 	$UMULL		r7,r6,r6
 	$UMULH		r8,r6,r6
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r0
@@ -509,11 +509,11 @@ $data=<<EOF;
 	$LD		r6,`3*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
@@ -522,11 +522,11 @@ $data=<<EOF;
 	$LD		r6,`4*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
@@ -535,11 +535,11 @@ $data=<<EOF;
 	$LD		r6,`5*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r0
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
@@ -548,11 +548,11 @@ $data=<<EOF;
 	$LD		r6,`4*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
@@ -561,11 +561,11 @@ $data=<<EOF;
 	$LD		r6,`3*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
@@ -580,11 +580,11 @@ $data=<<EOF;
 	$LD		r6,`4*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
@@ -593,11 +593,11 @@ $data=<<EOF;
 	$LD		r6,`5*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r11
@@ -617,7 +617,7 @@ $data=<<EOF;
 	$LD		r6,`7*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r0
@@ -629,7 +629,7 @@ $data=<<EOF;
 	$LD		r6,`6*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
@@ -652,7 +652,7 @@ $data=<<EOF;
 	$LD		r6,`4*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r10,r7,r10
 	adde		r11,r8,r11
 	addze		r9,r9
@@ -684,7 +684,7 @@ $data=<<EOF;
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
-	
+
 	addc		r11,r7,r11
 	adde		r9,r8,r9
 	addze		r10,r10
@@ -704,7 +704,7 @@ $data=<<EOF;
 	$LD		r5,`2*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
 	$UMULH		r8,r5,r6
-	
+
 	addc		r9,r7,r9
 	adde		r10,r8,r10
 	addze		r11,r0
@@ -801,7 +801,7 @@ $data=<<EOF;
 	adde		r10,r8,r10
 	addze		r11,r11
 	$ST		r9,`12*$BNSZ`(r3)	#r[12]=c1;
-	
+
 						#sqr_add_c2(a,7,6,c2,c3,c1)
 	$LD		r5,`6*$BNSZ`(r4)
 	$UMULL		r7,r5,r6
@@ -850,21 +850,21 @@ $data=<<EOF;
 #
 	xor	r0,r0,r0		#r0=0. Used in addze below.
 					#mul_add_c(a[0],b[0],c1,c2,c3);
-	$LD	r6,`0*$BNSZ`(r4)		
-	$LD	r7,`0*$BNSZ`(r5)		
-	$UMULL	r10,r6,r7		
-	$UMULH	r11,r6,r7		
+	$LD	r6,`0*$BNSZ`(r4)
+	$LD	r7,`0*$BNSZ`(r5)
+	$UMULL	r10,r6,r7
+	$UMULH	r11,r6,r7
 	$ST	r10,`0*$BNSZ`(r3)	#r[0]=c1
 					#mul_add_c(a[0],b[1],c2,c3,c1);
-	$LD	r7,`1*$BNSZ`(r5)		
+	$LD	r7,`1*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r11,r8,r11
 	adde	r12,r9,r0
 	addze	r10,r0
 					#mul_add_c(a[1],b[0],c2,c3,c1);
-	$LD	r6, `1*$BNSZ`(r4)		
-	$LD	r7, `0*$BNSZ`(r5)		
+	$LD	r6, `1*$BNSZ`(r4)
+	$LD	r7, `0*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r11,r8,r11
@@ -872,23 +872,23 @@ $data=<<EOF;
 	addze	r10,r10
 	$ST	r11,`1*$BNSZ`(r3)	#r[1]=c2
 					#mul_add_c(a[2],b[0],c3,c1,c2);
-	$LD	r6,`2*$BNSZ`(r4)		
+	$LD	r6,`2*$BNSZ`(r4)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r12,r8,r12
 	adde	r10,r9,r10
 	addze	r11,r0
 					#mul_add_c(a[1],b[1],c3,c1,c2);
-	$LD	r6,`1*$BNSZ`(r4)		
-	$LD	r7,`1*$BNSZ`(r5)		
+	$LD	r6,`1*$BNSZ`(r4)
+	$LD	r7,`1*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r12,r8,r12
 	adde	r10,r9,r10
 	addze	r11,r11
 					#mul_add_c(a[0],b[2],c3,c1,c2);
-	$LD	r6,`0*$BNSZ`(r4)		
-	$LD	r7,`2*$BNSZ`(r5)		
+	$LD	r6,`0*$BNSZ`(r4)
+	$LD	r7,`2*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r12,r8,r12
@@ -896,7 +896,7 @@ $data=<<EOF;
 	addze	r11,r11
 	$ST	r12,`2*$BNSZ`(r3)	#r[2]=c3
 					#mul_add_c(a[0],b[3],c1,c2,c3);
-	$LD	r7,`3*$BNSZ`(r5)		
+	$LD	r7,`3*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r10,r8,r10
@@ -928,7 +928,7 @@ $data=<<EOF;
 	addze	r12,r12
 	$ST	r10,`3*$BNSZ`(r3)	#r[3]=c1
 					#mul_add_c(a[3],b[1],c2,c3,c1);
-	$LD	r7,`1*$BNSZ`(r5)		
+	$LD	r7,`1*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r11,r8,r11
@@ -952,7 +952,7 @@ $data=<<EOF;
 	addze	r10,r10
 	$ST	r11,`4*$BNSZ`(r3)	#r[4]=c2
 					#mul_add_c(a[2],b[3],c3,c1,c2);
-	$LD	r6,`2*$BNSZ`(r4)		
+	$LD	r6,`2*$BNSZ`(r4)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r12,r8,r12
@@ -968,7 +968,7 @@ $data=<<EOF;
 	addze	r11,r11
 	$ST	r12,`5*$BNSZ`(r3)	#r[5]=c3
 					#mul_add_c(a[3],b[3],c1,c2,c3);
-	$LD	r7,`3*$BNSZ`(r5)		
+	$LD	r7,`3*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
 	addc	r10,r8,r10
@@ -988,7 +988,7 @@ $data=<<EOF;
 #		for the gcc compiler. This should be automatically
 #		done in the build
 #
-	
+
 .align	4
 .bn_mul_comba8:
 #
@@ -1003,7 +1003,7 @@ $data=<<EOF;
 # r10, r11, r12 are the equivalents of c1, c2, and c3.
 #
 	xor	r0,r0,r0		#r0=0. Used in addze below.
-	
+
 					#mul_add_c(a[0],b[0],c1,c2,c3);
 	$LD	r6,`0*$BNSZ`(r4)	#a[0]
 	$LD	r7,`0*$BNSZ`(r5)	#b[0]
@@ -1065,7 +1065,7 @@ $data=<<EOF;
 	addc	r10,r10,r8
 	adde	r11,r11,r9
 	addze	r12,r12
-		
+
 					#mul_add_c(a[2],b[1],c1,c2,c3);
 	$LD	r6,`2*$BNSZ`(r4)
 	$LD	r7,`1*$BNSZ`(r5)
@@ -1131,7 +1131,7 @@ $data=<<EOF;
 	adde	r10,r10,r9
 	addze	r11,r0
 					#mul_add_c(a[1],b[4],c3,c1,c2);
-	$LD	r6,`1*$BNSZ`(r4)		
+	$LD	r6,`1*$BNSZ`(r4)
 	$LD	r7,`4*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
@@ -1139,7 +1139,7 @@ $data=<<EOF;
 	adde	r10,r10,r9
 	addze	r11,r11
 					#mul_add_c(a[2],b[3],c3,c1,c2);
-	$LD	r6,`2*$BNSZ`(r4)		
+	$LD	r6,`2*$BNSZ`(r4)
 	$LD	r7,`3*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
@@ -1147,7 +1147,7 @@ $data=<<EOF;
 	adde	r10,r10,r9
 	addze	r11,r11
 					#mul_add_c(a[3],b[2],c3,c1,c2);
-	$LD	r6,`3*$BNSZ`(r4)		
+	$LD	r6,`3*$BNSZ`(r4)
 	$LD	r7,`2*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
@@ -1155,7 +1155,7 @@ $data=<<EOF;
 	adde	r10,r10,r9
 	addze	r11,r11
 					#mul_add_c(a[4],b[1],c3,c1,c2);
-	$LD	r6,`4*$BNSZ`(r4)		
+	$LD	r6,`4*$BNSZ`(r4)
 	$LD	r7,`1*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
@@ -1163,7 +1163,7 @@ $data=<<EOF;
 	adde	r10,r10,r9
 	addze	r11,r11
 					#mul_add_c(a[5],b[0],c3,c1,c2);
-	$LD	r6,`5*$BNSZ`(r4)		
+	$LD	r6,`5*$BNSZ`(r4)
 	$LD	r7,`0*$BNSZ`(r5)
 	$UMULL	r8,r6,r7
 	$UMULH	r9,r6,r7
@@ -1555,7 +1555,7 @@ $data=<<EOF;
 	addi	r3,r3,-$BNSZ
 	addi	r5,r5,-$BNSZ
 	mtctr	r6
-Lppcasm_sub_mainloop:	
+Lppcasm_sub_mainloop:
 	$LDU	r7,$BNSZ(r4)
 	$LDU	r8,$BNSZ(r5)
 	subfe	r6,r8,r7	# r6 = r7+carry bit + onescomplement(r8)
@@ -1563,7 +1563,7 @@ Lppcasm_sub_mainloop:
 				# is r7-r8 -1 as we need.
 	$STU	r6,$BNSZ(r3)
 	bdnz	Lppcasm_sub_mainloop
-Lppcasm_sub_adios:	
+Lppcasm_sub_adios:
 	subfze	r3,r0		# if carry bit is set then r3 = 0 else -1
 	andi.	r3,r3,1         # keep only last bit.
 	blr
@@ -1604,13 +1604,13 @@ Lppcasm_sub_adios:
 	addi	r3,r3,-$BNSZ
 	addi	r5,r5,-$BNSZ
 	mtctr	r6
-Lppcasm_add_mainloop:	
+Lppcasm_add_mainloop:
 	$LDU	r7,$BNSZ(r4)
 	$LDU	r8,$BNSZ(r5)
 	adde	r8,r7,r8
 	$STU	r8,$BNSZ(r3)
 	bdnz	Lppcasm_add_mainloop
-Lppcasm_add_adios:	
+Lppcasm_add_adios:
 	addze	r3,r0			#return carry bit.
 	blr
 	.long	0
@@ -1633,11 +1633,11 @@ Lppcasm_add_adios:
 #	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
 	bne	Lppcasm_div1		# proceed if d!=0
 	li	r3,-1			# d=0 return -1
@@ -1653,7 +1653,7 @@ Lppcasm_div1:
 Lppcasm_div2:
 	$UCMP	0,r3,r5			#h>=d?
 	blt	Lppcasm_div3		#goto Lppcasm_div3 if not
-	subf	r3,r5,r3		#h-=d ; 
+	subf	r3,r5,r3		#h-=d ;
 Lppcasm_div3:				#r7 = BN_BITS2-i. so r7=i
 	cmpi	0,0,r7,0		# is (i == 0)?
 	beq	Lppcasm_div4
@@ -1668,7 +1668,7 @@ Lppcasm_div4:
 					# as it saves registers.
 	li	r6,2			#r6=2
 	mtctr	r6			#counter will be in count.
-Lppcasm_divouterloop: 
+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.
@@ -1676,7 +1676,7 @@ Lppcasm_divouterloop:
 	bne	Lppcasm_div5		# goto Lppcasm_div5 if not
 
 	li	r8,-1
-	$CLRU	r8,r8,`$BITS/2`		#q = BN_MASK2l 
+	$CLRU	r8,r8,`$BITS/2`		#q = BN_MASK2l
 	b	Lppcasm_div6
 Lppcasm_div5:
 	$UDIV	r8,r3,r9		#q = h/dh
@@ -1684,7 +1684,7 @@ 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...
@@ -1761,7 +1761,7 @@ Lppcasm_div9:
 	addi	r4,r4,-$BNSZ
 	addi	r3,r3,-$BNSZ
 	mtctr	r5
-Lppcasm_sqr_mainloop:	
+Lppcasm_sqr_mainloop:
 					#sqr(r[0],r[1],a[0]);
 	$LDU	r6,$BNSZ(r4)
 	$UMULL	r7,r6,r6
@@ -1769,7 +1769,7 @@ Lppcasm_sqr_mainloop:
 	$STU	r7,$BNSZ(r3)
 	$STU	r8,$BNSZ(r3)
 	bdnz	Lppcasm_sqr_mainloop
-Lppcasm_sqr_adios:	
+Lppcasm_sqr_adios:
 	blr
 	.long	0
 	.byte	0,12,0x14,0,0,0,3,0
@@ -1783,7 +1783,7 @@ Lppcasm_sqr_adios:
 #		done in the build
 #
 
-.align	4	
+.align	4
 .bn_mul_words:
 #
 # BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w)
@@ -1797,7 +1797,7 @@ Lppcasm_sqr_adios:
 	rlwinm.	r7,r5,30,2,31		# num >> 2
 	beq	Lppcasm_mw_REM
 	mtctr	r7
-Lppcasm_mw_LOOP:	
+Lppcasm_mw_LOOP:
 					#mul(rp[0],ap[0],w,c1);
 	$LD	r8,`0*$BNSZ`(r4)
 	$UMULL	r9,r6,r8
@@ -1809,7 +1809,7 @@ Lppcasm_mw_LOOP:
 					#using adde.
 	$ST	r9,`0*$BNSZ`(r3)
 					#mul(rp[1],ap[1],w,c1);
-	$LD	r8,`1*$BNSZ`(r4)	
+	$LD	r8,`1*$BNSZ`(r4)
 	$UMULL	r11,r6,r8
 	$UMULH  r12,r6,r8
 	adde	r11,r11,r10
@@ -1830,7 +1830,7 @@ Lppcasm_mw_LOOP:
 	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`
 	bdnz	Lppcasm_mw_LOOP
@@ -1846,25 +1846,25 @@ Lppcasm_mw_REM:
 	addze	r10,r10
 	$ST	r9,`0*$BNSZ`(r3)
 	addi	r12,r10,0
-	
+
 	addi	r5,r5,-1
 	cmpli	0,0,r5,0
 	beq	Lppcasm_mw_OVER
 
-	
+
 					#mul(rp[1],ap[1],w,c1);
-	$LD	r8,`1*$BNSZ`(r4)	
+	$LD	r8,`1*$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
 	beq	Lppcasm_mw_OVER
-	
+
 					#mul_add(rp[2],ap[2],w,c1);
 	$LD	r8,`2*$BNSZ`(r4)
 	$UMULL	r9,r6,r8
@@ -1873,8 +1873,8 @@ Lppcasm_mw_REM:
 	addze	r10,r10
 	$ST	r9,`2*$BNSZ`(r3)
 	addi	r12,r10,0
-		
-Lppcasm_mw_OVER:	
+
+Lppcasm_mw_OVER:
 	addi	r3,r12,0
 	blr
 	.long	0
@@ -1902,11 +1902,11 @@ Lppcasm_mw_OVER:
 # empirical evidence suggests that unrolled version performs best!!
 #
 	xor	r0,r0,r0		#r0 = 0
-	xor	r12,r12,r12  		#r12 = 0 . used for carry		
+	xor	r12,r12,r12  		#r12 = 0 . used for carry
 	rlwinm.	r7,r5,30,2,31		# num >> 2
 	beq	Lppcasm_maw_leftover	# if (num < 4) go LPPCASM_maw_leftover
 	mtctr	r7
-Lppcasm_maw_mainloop:	
+Lppcasm_maw_mainloop:
 					#mul_add(rp[0],ap[0],w,c1);
 	$LD	r8,`0*$BNSZ`(r4)
 	$LD	r11,`0*$BNSZ`(r3)
@@ -1922,9 +1922,9 @@ Lppcasm_maw_mainloop:
 					#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)	
+	$LD	r8,`1*$BNSZ`(r4)
 	$LD	r9,`1*$BNSZ`(r3)
 	$UMULL	r11,r6,r8
 	$UMULH  r12,r6,r8
@@ -1933,7 +1933,7 @@ Lppcasm_maw_mainloop:
 	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
@@ -1944,7 +1944,7 @@ Lppcasm_maw_mainloop:
 	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)
 	$UMULL	r11,r6,r8
@@ -1958,7 +1958,7 @@ Lppcasm_maw_mainloop:
 	addi	r3,r3,`4*$BNSZ`
 	addi	r4,r4,`4*$BNSZ`
 	bdnz	Lppcasm_maw_mainloop
-	
+
 Lppcasm_maw_leftover:
 	andi.	r5,r5,0x3
 	beq	Lppcasm_maw_adios
@@ -1975,10 +1975,10 @@ Lppcasm_maw_leftover:
 	addc	r9,r9,r12
 	addze	r12,r10
 	$ST	r9,0(r3)
-	
+
 	bdz	Lppcasm_maw_adios
 					#mul_add(rp[1],ap[1],w,c1);
-	$LDU	r8,$BNSZ(r4)	
+	$LDU	r8,$BNSZ(r4)
 	$UMULL	r9,r6,r8
 	$UMULH  r10,r6,r8
 	$LDU	r11,$BNSZ(r3)
@@ -1987,7 +1987,7 @@ Lppcasm_maw_leftover:
 	addc	r9,r9,r12
 	addze	r12,r10
 	$ST	r9,0(r3)
-	
+
 	bdz	Lppcasm_maw_adios
 					#mul_add(rp[2],ap[2],w,c1);
 	$LDU	r8,$BNSZ(r4)
@@ -1999,8 +1999,8 @@ Lppcasm_maw_leftover:
 	addc	r9,r9,r12
 	addze	r12,r10
 	$ST	r9,0(r3)
-		
-Lppcasm_maw_adios:	
+
+Lppcasm_maw_adios:
 	addi	r3,r12,0
 	blr
 	.long	0
diff --git a/crypto/bn/asm/rsaz-avx2.pl b/crypto/bn/asm/rsaz-avx2.pl
index 0c1b236ef9..f34c84f452 100755
--- a/crypto/bn/asm/rsaz-avx2.pl
+++ b/crypto/bn/asm/rsaz-avx2.pl
@@ -382,7 +382,7 @@ $code.=<<___;
 	vpaddq		$TEMP1, $ACC1, $ACC1
 	vpmuludq	32*7-128($aap), $B2, $ACC2
 	 vpbroadcastq	32*5-128($tpa), $B2
-	vpaddq		32*11-448($tp1), $ACC2, $ACC2	
+	vpaddq		32*11-448($tp1), $ACC2, $ACC2
 
 	vmovdqu		$ACC6, 32*6-192($tp0)
 	vmovdqu		$ACC7, 32*7-192($tp0)
@@ -441,7 +441,7 @@ $code.=<<___;
 	vmovdqu		$ACC7, 32*16-448($tp1)
 	lea		8($tp1), $tp1
 
-	dec	$i        
+	dec	$i
 	jnz	.LOOP_SQR_1024
 ___
 $ZERO = $ACC9;
@@ -786,7 +786,7 @@ $code.=<<___;
 	vpblendd	\$3, $TEMP4, $TEMP5, $TEMP4
 	vpaddq		$TEMP3, $ACC7, $ACC7
 	vpaddq		$TEMP4, $ACC8, $ACC8
-     
+
 	vpsrlq		\$29, $ACC4, $TEMP1
 	vpand		$AND_MASK, $ACC4, $ACC4
 	vpsrlq		\$29, $ACC5, $TEMP2
@@ -1451,7 +1451,7 @@ $code.=<<___;
 	vpaddq		$TEMP4, $ACC8, $ACC8
 
 	vmovdqu		$ACC4, 128-128($rp)
-	vmovdqu		$ACC5, 160-128($rp)    
+	vmovdqu		$ACC5, 160-128($rp)
 	vmovdqu		$ACC6, 192-128($rp)
 	vmovdqu		$ACC7, 224-128($rp)
 	vmovdqu		$ACC8, 256-128($rp)
diff --git a/crypto/bn/asm/rsaz-x86_64.pl b/crypto/bn/asm/rsaz-x86_64.pl
index 6f3b664f7a..7bcfafe8dd 100755
--- a/crypto/bn/asm/rsaz-x86_64.pl
+++ b/crypto/bn/asm/rsaz-x86_64.pl
@@ -282,9 +282,9 @@ $code.=<<___;
 	movq	%r9, 16(%rsp)
 	movq	%r10, 24(%rsp)
 	shrq	\$63, %rbx
-	
+
 #third iteration
-	movq	16($inp), %r9	
+	movq	16($inp), %r9
 	movq	24($inp), %rax
 	mulq	%r9
 	addq	%rax, %r12
@@ -532,7 +532,7 @@ $code.=<<___;
 	movl	$times,128+8(%rsp)
 	movq	$out, %xmm0		# off-load
 	movq	%rbp, %xmm1		# off-load
-#first iteration	
+#first iteration
 	mulx	%rax, %r8, %r9
 
 	mulx	16($inp), %rcx, %r10
@@ -568,7 +568,7 @@ $code.=<<___;
 	mov	%rax, (%rsp)
 	mov	%r8, 8(%rsp)
 
-#second iteration	
+#second iteration
 	mulx	16($inp), %rax, %rbx
 	adox	%rax, %r10
 	adcx	%rbx, %r11
@@ -607,8 +607,8 @@ $code.=<<___;
 
 	mov	%r9, 16(%rsp)
 	.byte	0x4c,0x89,0x94,0x24,0x18,0x00,0x00,0x00		# mov	%r10, 24(%rsp)
-	
-#third iteration	
+
+#third iteration
 	.byte	0xc4,0x62,0xc3,0xf6,0x8e,0x18,0x00,0x00,0x00	# mulx	24($inp), $out, %r9
 	adox	$out, %r12
 	adcx	%r9, %r13
@@ -643,8 +643,8 @@ $code.=<<___;
 
 	mov	%r11, 32(%rsp)
 	.byte	0x4c,0x89,0xa4,0x24,0x28,0x00,0x00,0x00		# mov	%r12, 40(%rsp)
-	
-#fourth iteration	
+
+#fourth iteration
 	.byte	0xc4,0xe2,0xfb,0xf6,0x9e,0x20,0x00,0x00,0x00	# mulx	32($inp), %rax, %rbx
 	adox	%rax, %r14
 	adcx	%rbx, %r15
@@ -676,8 +676,8 @@ $code.=<<___;
 
 	mov	%r13, 48(%rsp)
 	mov	%r14, 56(%rsp)
-	
-#fifth iteration	
+
+#fifth iteration
 	.byte	0xc4,0x62,0xc3,0xf6,0x9e,0x28,0x00,0x00,0x00	# mulx	40($inp), $out, %r11
 	adox	$out, %r8
 	adcx	%r11, %r9
@@ -704,8 +704,8 @@ $code.=<<___;
 
 	mov	%r15, 64(%rsp)
 	mov	%r8, 72(%rsp)
-	
-#sixth iteration	
+
+#sixth iteration
 	.byte	0xc4,0xe2,0xfb,0xf6,0x9e,0x30,0x00,0x00,0x00	# mulx	48($inp), %rax, %rbx
 	adox	%rax, %r10
 	adcx	%rbx, %r11
@@ -1048,7 +1048,7 @@ $code.=<<___;
 	movq	56($ap), %rax
 	movq	%rdx, %r14
 	adcq	\$0, %r14
-	
+
 	mulq	%rbx
 	addq	%rax, %r14
 	 movq	($ap), %rax
@@ -1150,7 +1150,7 @@ $code.=<<___;
 	 movq	($ap), %rax
 	adcq	\$0, %rdx
 	addq	%r15, %r14
-	movq	%rdx, %r15	
+	movq	%rdx, %r15
 	adcq	\$0, %r15
 
 	leaq	8(%rdi), %rdi
@@ -1212,7 +1212,7 @@ $code.=<<___ if ($addx);
 
 	mulx	48($ap), %rbx, %r14
 	adcx	%rax, %r12
-	
+
 	mulx	56($ap), %rax, %r15
 	adcx	%rbx, %r13
 	adcx	%rax, %r14
@@ -1411,7 +1411,7 @@ $code.=<<___;
 ___
 $code.=<<___ if ($addx);
 	jmp	.Lmul_scatter_tail
-	
+
 .align	32
 .Lmulx_scatter:
 	movq	($out), %rdx		# pass b[0]
@@ -1824,7 +1824,7 @@ __rsaz_512_mul:
 	movq	56($ap), %rax
 	movq	%rdx, %r14
 	adcq	\$0, %r14
-	
+
 	mulq	%rbx
 	addq	%rax, %r14
 	 movq	($ap), %rax
@@ -1901,7 +1901,7 @@ __rsaz_512_mul:
 	 movq	($ap), %rax
 	adcq	\$0, %rdx
 	addq	%r15, %r14
-	movq	%rdx, %r15	
+	movq	%rdx, %r15
 	adcq	\$0, %r15
 
 	leaq	8(%rdi), %rdi
diff --git a/crypto/bn/asm/s390x-gf2m.pl b/crypto/bn/asm/s390x-gf2m.pl
index cbd16f4214..57b0032d67 100644
--- a/crypto/bn/asm/s390x-gf2m.pl
+++ b/crypto/bn/asm/s390x-gf2m.pl
@@ -198,7 +198,7 @@ $code.=<<___;
 	xgr	$hi,@r[1]
 	xgr	$lo,@r[0]
 	xgr	$hi,@r[2]
-	xgr	$lo,@r[3]	
+	xgr	$lo,@r[3]
 	xgr	$hi,@r[3]
 	xgr	$lo,$hi
 	stg	$hi,16($rp)
diff --git a/crypto/bn/asm/via-mont.pl b/crypto/bn/asm/via-mont.pl
index 9f81bc822e..558501c315 100644
--- a/crypto/bn/asm/via-mont.pl
+++ b/crypto/bn/asm/via-mont.pl
@@ -76,7 +76,7 @@
 # dsa 1024 bits 0.001346s 0.001595s    742.7    627.0
 # dsa 2048 bits 0.004745s 0.005582s    210.7    179.1
 #
-# Conclusions: 
+# Conclusions:
 # - VIA SDK leaves a *lot* of room for improvement (which this
 #   implementation successfully fills:-);
 # - 'rep montmul' gives up to >3x performance improvement depending on
diff --git a/crypto/bn/asm/x86-mont.pl b/crypto/bn/asm/x86-mont.pl
index 6787503666..a8b402d59b 100755
--- a/crypto/bn/asm/x86-mont.pl
+++ b/crypto/bn/asm/x86-mont.pl
@@ -39,7 +39,7 @@ require "x86asm.pl";
 
 $output = pop;
 open STDOUT,">$output";
- 
+
 &asm_init($ARGV[0],$0);
 
 $sse2=0;
diff --git a/crypto/bn/asm/x86_64-mont5.pl b/crypto/bn/asm/x86_64-mont5.pl
index 3278dc6056..8f49391727 100755
--- a/crypto/bn/asm/x86_64-mont5.pl
+++ b/crypto/bn/asm/x86_64-mont5.pl
@@ -1049,7 +1049,7 @@ my $bptr="%rdx";	# const void *table,
 my $nptr="%rcx";	# const BN_ULONG *nptr,
 my $n0  ="%r8";		# const BN_ULONG *n0);
 my $num ="%r9";		# int num, has to be divisible by 8
-			# int pwr 
+			# int pwr
 
 my ($i,$j,$tptr)=("%rbp","%rcx",$rptr);
 my @A0=("%r10","%r11");
@@ -1126,7 +1126,7 @@ $code.=<<___;
 	ja	.Lpwr_page_walk
 .Lpwr_page_walk_done:
 
-	mov	$num,%r10	
+	mov	$num,%r10
 	neg	$num
 
 	##############################################################
@@ -2036,7 +2036,7 @@ __bn_post4x_internal:
 	jnz	.Lsqr4x_sub
 
 	mov	$num,%r10		# prepare for back-to-back call
-	neg	$num			# restore $num	
+	neg	$num			# restore $num
 	ret
 .size	__bn_post4x_internal,.-__bn_post4x_internal
 ___
@@ -2259,7 +2259,7 @@ bn_mulx4x_mont_gather5:
 	mov	\$0,%r10
 	cmovc	%r10,%r11
 	sub	%r11,%rbp
-.Lmulx4xsp_done:	
+.Lmulx4xsp_done:
 	and	\$-64,%rbp		# ensure alignment
 	mov	%rsp,%r11
 	sub	%rbp,%r11
@@ -2741,7 +2741,7 @@ bn_powerx5:
 	ja	.Lpwrx_page_walk
 .Lpwrx_page_walk_done:
 
-	mov	$num,%r10	
+	mov	$num,%r10
 	neg	$num
 
 	##############################################################