Unify all assembler file generators

They now generally conform to the following argument sequence:

    script.pl "$(PERLASM_SCHEME)" [ C preprocessor arguments ... ] \
              $(PROCESSOR) <output file>

However, in the spirit of being able to use these scripts manually,
they also allow for no argument, or for only the flavour, or for only
the output file.  This is done by only using the last argument as
output file if it's a file (it has an extension), and only using the
first argument as flavour if it isn't a file (it doesn't have an
extension).

While we're at it, we make all $xlate calls the same, i.e. the $output
argument is always quoted, and we always die on error when trying to
start $xlate.

There's a perl lesson in this, regarding operator priority...

This will always succeed, even when it fails:

    open FOO, "something" || die "ERR: $!";

The reason is that '||' has higher priority than list operators (a
function is essentially a list operator and gobbles up everything
following it that isn't lower priority), and since a non-empty string
is always true, so that ends up being exactly the same as:

    open FOO, "something";

This, however, will fail if "something" can't be opened:

    open FOO, "something" or die "ERR: $!";

The reason is that 'or' has lower priority that list operators,
i.e. it's performed after the 'open' call.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/9884)

29 files changed, 113 insertions, 80 deletions

diff --git a/crypto/bn/asm/alpha-mont.pl b/crypto/bn/asm/alpha-mont.pl
index 2d282a1f7a..b46c79ca4f 100644
--- a/crypto/bn/asm/alpha-mont.pl
+++ b/crypto/bn/asm/alpha-mont.pl
@@ -22,8 +22,7 @@
 # I.e. if you compare 1GHz 21264 and 2GHz Opteron, you'll observe ~2x
 # difference.
 
-$output=pop;
-open STDOUT,">$output";
+$output=pop and open STDOUT,">$output";
 
 # int bn_mul_mont(
 $rp="a0";	# BN_ULONG *rp,
diff --git a/crypto/bn/asm/armv4-gf2m.pl b/crypto/bn/asm/armv4-gf2m.pl
index 0bf6f63ec5..e7a865581f 100644
--- a/crypto/bn/asm/armv4-gf2m.pl
+++ b/crypto/bn/asm/armv4-gf2m.pl
@@ -39,9 +39,10 @@
 #
 # http://conradoplg.cryptoland.net/files/2010/12/mocrysen13.pdf
 
-$flavour = shift;
-if ($flavour=~/\w[\w\-]*\.\w+$/) { $output=$flavour; undef $flavour; }
-else { while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {} }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour && $flavour ne "void") {
     $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
@@ -49,9 +50,10 @@ if ($flavour && $flavour ne "void") {
     ( $xlate="${dir}../../perlasm/arm-xlate.pl" and -f $xlate) or
     die "can't locate arm-xlate.pl";
 
-    open STDOUT,"| \"$^X\" $xlate $flavour $output";
+    open STDOUT,"| \"$^X\" $xlate $flavour \"$output\""
+        or die "can't call $xlate: $1";
 } else {
-    open STDOUT,">$output";
+    $output and open STDOUT,">$output";
 }
 
 $code=<<___;
diff --git a/crypto/bn/asm/armv4-mont.pl b/crypto/bn/asm/armv4-mont.pl
index 7e0a4d8145..72c883fe9f 100644
--- a/crypto/bn/asm/armv4-mont.pl
+++ b/crypto/bn/asm/armv4-mont.pl
@@ -54,9 +54,10 @@
 # integer-only on Cortex-A8, ~10-210% on Cortex-A15, ~70-450% on
 # Snapdragon S4.
 
-$flavour = shift;
-if ($flavour=~/\w[\w\-]*\.\w+$/) { $output=$flavour; undef $flavour; }
-else { while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {} }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+my $output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+my $flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour && $flavour ne "void") {
     $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
@@ -64,9 +65,10 @@ if ($flavour && $flavour ne "void") {
     ( $xlate="${dir}../../perlasm/arm-xlate.pl" and -f $xlate) or
     die "can't locate arm-xlate.pl";
 
-    open STDOUT,"| \"$^X\" $xlate $flavour $output";
+    open STDOUT,"| \"$^X\" $xlate $flavour \"$output\""
+        or die "can't call $xlate: $1";
 } else {
-    open STDOUT,">$output";
+    $output and open STDOUT,">$output";
 }
 
 $num="r0";	# starts as num argument, but holds &tp[num-1]
diff --git a/crypto/bn/asm/armv8-mont.pl b/crypto/bn/asm/armv8-mont.pl
index c755555d88..3e92881bc0 100755
--- a/crypto/bn/asm/armv8-mont.pl
+++ b/crypto/bn/asm/armv8-mont.pl
@@ -40,15 +40,18 @@
 # 50-70% improvement for RSA4096 sign. RSA2048 sign is ~25% faster
 # on Cortex-A57 and ~60-100% faster on others.
 
-$flavour = shift;
-$output  = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+my $output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+my $flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}arm-xlate.pl" and -f $xlate ) or
 ( $xlate="${dir}../../perlasm/arm-xlate.pl" and -f $xlate) or
 die "can't locate arm-xlate.pl";
 
-open OUT,"| \"$^X\" $xlate $flavour $output";
+open OUT,"| \"$^X\" $xlate $flavour \"$output\""
+    or die "can't call $xlate: $1";
 *STDOUT=*OUT;
 
 ($lo0,$hi0,$aj,$m0,$alo,$ahi,
diff --git a/crypto/bn/asm/bn-586.pl b/crypto/bn/asm/bn-586.pl
index 12b51953ab..287162161c 100644
--- a/crypto/bn/asm/bn-586.pl
+++ b/crypto/bn/asm/bn-586.pl
@@ -11,8 +11,7 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 &asm_init($ARGV[0]);
 
diff --git a/crypto/bn/asm/c64xplus-gf2m.pl b/crypto/bn/asm/c64xplus-gf2m.pl
index c77fc5eef4..d4913867d0 100644
--- a/crypto/bn/asm/c64xplus-gf2m.pl
+++ b/crypto/bn/asm/c64xplus-gf2m.pl
@@ -23,8 +23,7 @@
 # totally unfair, because this module utilizes Galois Field Multiply
 # instruction.
 
-while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {}
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 ($rp,$a1,$a0,$b1,$b0)=("A4","B4","A6","B6","A8");   # argument vector
 
diff --git a/crypto/bn/asm/co-586.pl b/crypto/bn/asm/co-586.pl
index a0f5a46979..d9263094b4 100644
--- a/crypto/bn/asm/co-586.pl
+++ b/crypto/bn/asm/co-586.pl
@@ -10,8 +10,7 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 &asm_init($ARGV[0]);
 
diff --git a/crypto/bn/asm/ia64-mont.pl b/crypto/bn/asm/ia64-mont.pl
index 1ab5d94eaa..ce51b17ef9 100644
--- a/crypto/bn/asm/ia64-mont.pl
+++ b/crypto/bn/asm/ia64-mont.pl
@@ -67,7 +67,8 @@
 # hereafter less for longer keys, while verify - by 74-13%.
 # DSA performance improves by 115-30%.
 
-$output=pop;
+# $output is the last argument if it looks like a file (it has an extension)
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
 
 if ($^O eq "hpux") {
     $ADDP="addp4";
diff --git a/crypto/bn/asm/mips-mont.pl b/crypto/bn/asm/mips-mont.pl
index 3386128ab8..c0b4629184 100644
--- a/crypto/bn/asm/mips-mont.pl
+++ b/crypto/bn/asm/mips-mont.pl
@@ -52,8 +52,12 @@
 # ($t0,$t1,$t2,$t3,$t8,$t9)=map("\$$_",(12..15,24,25));
 # ($s0,$s1,$s2,$s3,$s4,$s5,$s6,$s7)=map("\$$_",(16..23));
 # ($gp,$sp,$fp,$ra)=map("\$$_",(28..31));
-#
-$flavour = shift || "o32"; # supported flavours are o32,n32,64,nubi32,nubi64
+
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+# supported flavours are o32,n32,64,nubi32,nubi64, default is o32
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : "o32";
 
 if ($flavour =~ /64|n32/i) {
 	$PTR_ADD="daddu";	# incidentally works even on n32
@@ -74,8 +78,7 @@ $SAVED_REGS_MASK = ($flavour =~ /nubi/i) ? 0x00fff000 : 0x00ff0000;
 #
 ######################################################################
 
-while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {}
-open STDOUT,">$output";
+$output and open STDOUT,">$output";
 
 if ($flavour =~ /64|n32/i) {
 	$LD="ld";
diff --git a/crypto/bn/asm/mips.pl b/crypto/bn/asm/mips.pl
index 35971769a6..af44c9786e 100644
--- a/crypto/bn/asm/mips.pl
+++ b/crypto/bn/asm/mips.pl
@@ -54,9 +54,10 @@
 # has to content with 40-85% improvement depending on benchmark and
 # key length, more for longer keys.
 
-$flavour = shift || "o32";
-while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {}
-open STDOUT,">$output";
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : "o32";
 
 if ($flavour =~ /64|n32/i) {
 	$LD="ld";
@@ -91,6 +92,8 @@ if ($flavour =~ /64|n32/i) {
 	$code="#if !(defined (__mips_isa_rev) && (__mips_isa_rev >= 6))\n.set     mips2\n#endif\n";
 }
 
+$output and open STDOUT,">$output";
+
 # Below is N32/64 register layout used in the original module.
 #
 ($zero,$at,$v0,$v1)=map("\$$_",(0..3));
diff --git a/crypto/bn/asm/parisc-mont.pl b/crypto/bn/asm/parisc-mont.pl
index 1d66c21f05..1420840d62 100644
--- a/crypto/bn/asm/parisc-mont.pl
+++ b/crypto/bn/asm/parisc-mont.pl
@@ -69,10 +69,12 @@
 
 $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 
-$flavour = shift;
-$output = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
-open STDOUT,">$output";
+$output and open STDOUT,">$output";
 
 if ($flavour =~ /64/) {
 	$LEVEL		="2.0W";
diff --git a/crypto/bn/asm/ppc-mont.pl b/crypto/bn/asm/ppc-mont.pl
index 6e6f3a65b8..01c8d11bae 100644
--- a/crypto/bn/asm/ppc-mont.pl
+++ b/crypto/bn/asm/ppc-mont.pl
@@ -41,7 +41,10 @@
 # builds. On low-end 32-bit processors performance improvement turned
 # to be marginal...
 
-$flavour = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour =~ /32/) {
 	$BITS=	32;
@@ -94,7 +97,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or
 die "can't locate ppc-xlate.pl";
 
-open STDOUT,"| $^X $xlate $flavour ".shift || die "can't call $xlate: $!";
+open STDOUT,"| $^X $xlate $flavour \"$output\""
+    or die "can't call $xlate: $!";
 
 $sp="r1";
 $toc="r2";
diff --git a/crypto/bn/asm/ppc.pl b/crypto/bn/asm/ppc.pl
index 21f69636c8..4b4e6c6100 100644
--- a/crypto/bn/asm/ppc.pl
+++ b/crypto/bn/asm/ppc.pl
@@ -103,7 +103,10 @@
 #        Performance increase of ~60%
 #        Based on submission from Suresh N. Chari of IBM
 
-$flavour = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour =~ /32/) {
 	$BITS=	32;
@@ -159,7 +162,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or
 die "can't locate ppc-xlate.pl";
 
-open STDOUT,"| $^X $xlate $flavour ".shift || die "can't call $xlate: $!";
+open STDOUT,"| $^X $xlate $flavour \"$output\""
+    or die "can't call $xlate: $!";
 
 $data=<<EOF;
 #--------------------------------------------------------------------
diff --git a/crypto/bn/asm/ppc64-mont.pl b/crypto/bn/asm/ppc64-mont.pl
index b0277d95d6..4420a14a0a 100644
--- a/crypto/bn/asm/ppc64-mont.pl
+++ b/crypto/bn/asm/ppc64-mont.pl
@@ -80,7 +80,10 @@
 # ppc-mont.pl, but improvement coefficient is not as impressive
 # for longer keys...
 
-$flavour = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour =~ /32/) {
 	$SIZE_T=4;
@@ -108,7 +111,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or
 die "can't locate ppc-xlate.pl";
 
-open STDOUT,"| $^X $xlate $flavour ".shift || die "can't call $xlate: $!";
+open STDOUT,"| $^X $xlate $flavour \"$output\""
+    or die "can't call $xlate: $!";
 
 $FRAME=64;	# padded frame header
 $TRANSFER=16*8;
diff --git a/crypto/bn/asm/rsaz-avx2.pl b/crypto/bn/asm/rsaz-avx2.pl
index fddb12ffff..66822abed4 100755
--- a/crypto/bn/asm/rsaz-avx2.pl
+++ b/crypto/bn/asm/rsaz-avx2.pl
@@ -37,9 +37,10 @@
 # (***)	scalar AD*X code is faster than AVX2 and is preferred code
 #	path for Broadwell;
 
-$flavour = shift;
-$output  = shift;
-if ($flavour =~ /\./) { $output = $flavour; undef $flavour; }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 $win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
 
@@ -72,7 +73,8 @@ if (!$avx && `$ENV{CC} -v 2>&1` =~ /((?:^clang|LLVM) version|based on LLVM) ([3-
 	$addx = ($ver>=3.03);
 }
 
-open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\"";
+open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""
+    or die "can't call $xlate: $!";
 *STDOUT = *OUT;
 
 if ($avx>1) {{{
diff --git a/crypto/bn/asm/rsaz-x86_64.pl b/crypto/bn/asm/rsaz-x86_64.pl
index e3fe2e7cc4..c41d8e521e 100755
--- a/crypto/bn/asm/rsaz-x86_64.pl
+++ b/crypto/bn/asm/rsaz-x86_64.pl
@@ -52,9 +52,10 @@
 #	purposes;
 # (**)	MULX was attempted, but found to give only marginal improvement;
 
-$flavour = shift;
-$output  = shift;
-if ($flavour =~ /\./) { $output = $flavour; undef $flavour; }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 $win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
 
@@ -63,7 +64,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or
 die "can't locate x86_64-xlate.pl";
 
-open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\"";
+open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""
+    or die "can't call $xlate: $!";
 *STDOUT=*OUT;
 
 if (`$ENV{CC} -Wa,-v -c -o /dev/null -x assembler /dev/null 2>&1`
diff --git a/crypto/bn/asm/s390x-gf2m.pl b/crypto/bn/asm/s390x-gf2m.pl
index afe8d35548..85e5eaa84f 100644
--- a/crypto/bn/asm/s390x-gf2m.pl
+++ b/crypto/bn/asm/s390x-gf2m.pl
@@ -32,7 +32,10 @@
 #	so that improvement coefficients can vary from one specific
 #	setup to another.
 
-$flavour = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour =~ /3[12]/) {
         $SIZE_T=4;
@@ -42,8 +45,7 @@ if ($flavour =~ /3[12]/) {
         $g="g";
 }
 
-while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {}
-open STDOUT,">$output";
+$output and open STDOUT,">$output";
 
 $stdframe=16*$SIZE_T+4*8;
 
diff --git a/crypto/bn/asm/s390x-mont.pl b/crypto/bn/asm/s390x-mont.pl
index 2e823989e3..00b2d46798 100644
--- a/crypto/bn/asm/s390x-mont.pl
+++ b/crypto/bn/asm/s390x-mont.pl
@@ -51,7 +51,10 @@
 # On z990 it was measured to perform 2.6-2.2 times better than
 # compiler-generated code, less for longer keys...
 
-$flavour = shift;
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 if ($flavour =~ /3[12]/) {
 	$SIZE_T=4;
@@ -61,8 +64,7 @@ if ($flavour =~ /3[12]/) {
 	$g="g";
 }
 
-while (($output=shift) && ($output!~/\w[\w\-]*\.\w+$/)) {}
-open STDOUT,">$output";
+$output and open STDOUT,">$output";
 
 $stdframe=16*$SIZE_T+4*8;
 
diff --git a/crypto/bn/asm/sparct4-mont.pl b/crypto/bn/asm/sparct4-mont.pl
index 8e981be2a5..76fffaf268 100755
--- a/crypto/bn/asm/sparct4-mont.pl
+++ b/crypto/bn/asm/sparct4-mont.pl
@@ -83,8 +83,7 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 push(@INC,"${dir}","${dir}../../perlasm");
 require "sparcv9_modes.pl";
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 $code.=<<___;
 #include "sparc_arch.h"
diff --git a/crypto/bn/asm/sparcv9-gf2m.pl b/crypto/bn/asm/sparcv9-gf2m.pl
index d2973f0bcd..d4dda79219 100644
--- a/crypto/bn/asm/sparcv9-gf2m.pl
+++ b/crypto/bn/asm/sparcv9-gf2m.pl
@@ -25,8 +25,7 @@
 # ~100-230% faster than gcc-generated code and ~35-90% faster than
 # the pure SPARCv9 code path.
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 $locals=16*8;
 
diff --git a/crypto/bn/asm/sparcv9-mont.pl b/crypto/bn/asm/sparcv9-mont.pl
index 04c22f8d11..6348c2a63b 100644
--- a/crypto/bn/asm/sparcv9-mont.pl
+++ b/crypto/bn/asm/sparcv9-mont.pl
@@ -49,8 +49,7 @@
 # module still have hidden potential [see TODO list there], which is
 # estimated to be larger than 20%...
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 # int bn_mul_mont(
 $rp="%i0";	# BN_ULONG *rp,
diff --git a/crypto/bn/asm/sparcv9a-mont.pl b/crypto/bn/asm/sparcv9a-mont.pl
index af22d41306..a1170c7639 100755
--- a/crypto/bn/asm/sparcv9a-mont.pl
+++ b/crypto/bn/asm/sparcv9a-mont.pl
@@ -62,8 +62,10 @@
 # key length, more for longer keys] on USI&II cores and 30-80% - on
 # USIII&IV.
 
-$output = pop;
-open STDOUT,">$output";
+# $output is the last argument if it looks like a file (it has an extension)
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+
+$output and open STDOUT,">$output";
 
 $fname="bn_mul_mont_fpu";
 
diff --git a/crypto/bn/asm/via-mont.pl b/crypto/bn/asm/via-mont.pl
index 1b5cdd8f3a..d873082d4b 100644
--- a/crypto/bn/asm/via-mont.pl
+++ b/crypto/bn/asm/via-mont.pl
@@ -88,8 +88,7 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 &asm_init($ARGV[0]);
 
diff --git a/crypto/bn/asm/vis3-mont.pl b/crypto/bn/asm/vis3-mont.pl
index 9e8f498f97..53f9920efe 100644
--- a/crypto/bn/asm/vis3-mont.pl
+++ b/crypto/bn/asm/vis3-mont.pl
@@ -25,8 +25,7 @@
 # for reference purposes, because T4 has dedicated Montgomery
 # multiplication and squaring *instructions* that deliver even more.
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 $frame = "STACK_FRAME";
 $bias = "STACK_BIAS";
diff --git a/crypto/bn/asm/x86-gf2m.pl b/crypto/bn/asm/x86-gf2m.pl
index 8bcfcb4f8e..bcc6de08a9 100644
--- a/crypto/bn/asm/x86-gf2m.pl
+++ b/crypto/bn/asm/x86-gf2m.pl
@@ -43,8 +43,7 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 &asm_init($ARGV[0],$x86only = $ARGV[$#ARGV] eq "386");
 
diff --git a/crypto/bn/asm/x86-mont.pl b/crypto/bn/asm/x86-mont.pl
index 22baf8fc26..864816fb1a 100755
--- a/crypto/bn/asm/x86-mont.pl
+++ b/crypto/bn/asm/x86-mont.pl
@@ -37,8 +37,7 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
-$output = pop;
-open STDOUT,">$output";
+$output = pop and open STDOUT,">$output";
 
 &asm_init($ARGV[0]);
 
diff --git a/crypto/bn/asm/x86_64-gf2m.pl b/crypto/bn/asm/x86_64-gf2m.pl
index d0a73575e0..f312a0c72a 100644
--- a/crypto/bn/asm/x86_64-gf2m.pl
+++ b/crypto/bn/asm/x86_64-gf2m.pl
@@ -27,9 +27,10 @@
 # these coefficients are not ones for bn_GF2m_mul_2x2 itself, as not
 # all CPU time is burnt in it...
 
-$flavour = shift;
-$output  = shift;
-if ($flavour =~ /\./) { $output = $flavour; undef $flavour; }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 $win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
 
@@ -38,7 +39,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or
 die "can't locate x86_64-xlate.pl";
 
-open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\"";
+open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""
+    or die "can't call $xlate: $!";
 *STDOUT=*OUT;
 
 ($lo,$hi)=("%rax","%rdx");	$a=$lo;
diff --git a/crypto/bn/asm/x86_64-mont.pl b/crypto/bn/asm/x86_64-mont.pl
index ede009fc16..aa06345fa2 100755
--- a/crypto/bn/asm/x86_64-mont.pl
+++ b/crypto/bn/asm/x86_64-mont.pl
@@ -46,9 +46,10 @@
 #
 # Add MULX/ADOX/ADCX code path.
 
-$flavour = shift;
-$output  = shift;
-if ($flavour =~ /\./) { $output = $flavour; undef $flavour; }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 $win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
 
@@ -57,7 +58,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or
 die "can't locate x86_64-xlate.pl";
 
-open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\"";
+open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""
+    or die "can't call $xlate: $!";
 *STDOUT=*OUT;
 
 if (`$ENV{CC} -Wa,-v -c -o /dev/null -x assembler /dev/null 2>&1`
diff --git a/crypto/bn/asm/x86_64-mont5.pl b/crypto/bn/asm/x86_64-mont5.pl
index 6331566c0c..d6d2764e22 100755
--- a/crypto/bn/asm/x86_64-mont5.pl
+++ b/crypto/bn/asm/x86_64-mont5.pl
@@ -31,9 +31,10 @@
 # the np argument is not just modulus value, but one interleaved
 # with 0. This is to optimize post-condition...
 
-$flavour = shift;
-$output  = shift;
-if ($flavour =~ /\./) { $output = $flavour; undef $flavour; }
+# $output is the last argument if it looks like a file (it has an extension)
+# $flavour is the first argument if it doesn't look like a file
+$output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
+$flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
 
 $win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
 
@@ -42,7 +43,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or
 die "can't locate x86_64-xlate.pl";
 
-open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\"";
+open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""
+    or die "can't call $xlate: $!";
 *STDOUT=*OUT;
 
 if (`$ENV{CC} -Wa,-v -c -o /dev/null -x assembler /dev/null 2>&1`