$win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
-open STDOUT,"| $^X ${dir}perlasm/x86_64-xlate.pl $flavour $output";
+( $xlate="${dir}x86_64-xlate.pl" and -f $xlate ) or
+( $xlate="${dir}perlasm/x86_64-xlate.pl" and -f $xlate) or
+die "can't locate x86_64-xlate.pl";
+
+open OUT,"| \"$^X\" $xlate $flavour $output";
+*STDOUT=*OUT;
($arg1,$arg2,$arg3,$arg4)=$win64?("%rcx","%rdx","%r8", "%r9") : # Win64 order
("%rdi","%rsi","%rdx","%rcx"); # Unix order
call OPENSSL_cpuid_setup
.hidden OPENSSL_ia32cap_P
-.comm OPENSSL_ia32cap_P,8
+.comm OPENSSL_ia32cap_P,16,4
.text
.size OPENSSL_rdtsc,.-OPENSSL_rdtsc
.globl OPENSSL_ia32_cpuid
-.type OPENSSL_ia32_cpuid,\@abi-omnipotent
+.type OPENSSL_ia32_cpuid,\@function,1
.align 16
OPENSSL_ia32_cpuid:
mov %rbx,%r8 # save %rbx
xor %eax,%eax
+ mov %eax,8(%rdi) # clear 3rd word
cpuid
mov %eax,%r11d # max value for standard query level
shr \$14,%r10d
and \$0xfff,%r10d # number of cores -1 per L1D
+ cmp \$7,%r11d
+ jb .Lnocacheinfo
+
+ mov \$7,%eax
+ xor %ecx,%ecx
+ cpuid
+ mov %ebx,8(%rdi)
+
.Lnocacheinfo:
mov \$1,%eax
cpuid
.Lclear_avx:
mov \$0xefffe7ff,%eax # ~(1<<28|1<<12|1<<11)
and %eax,%r9d # clear AVX, FMA and AMD XOP bits
+ andl \$0xffffffdf,8(%rdi) # cleax AVX2, ~(1<<5)
.Ldone:
shl \$32,%r9
mov %r10d,%eax
___
}
+print<<___;
+.globl OPENSSL_ia32_rdrand
+.type OPENSSL_ia32_rdrand,\@abi-omnipotent
+.align 16
+OPENSSL_ia32_rdrand:
+ mov \$8,%ecx
+.Loop_rdrand:
+ rdrand %rax
+ jc .Lbreak_rdrand
+ loop .Loop_rdrand
+.Lbreak_rdrand:
+ cmp \$0,%rax
+ cmove %rcx,%rax
+ ret
+.size OPENSSL_ia32_rdrand,.-OPENSSL_ia32_rdrand
+
+.globl OPENSSL_ia32_rdseed
+.type OPENSSL_ia32_rdseed,\@abi-omnipotent
+.align 16
+OPENSSL_ia32_rdseed:
+ mov \$8,%ecx
+.Loop_rdseed:
+ rdseed %rax
+ jc .Lbreak_rdseed
+ loop .Loop_rdseed
+.Lbreak_rdseed:
+ cmp \$0,%rax
+ cmove %rcx,%rax
+ ret
+.size OPENSSL_ia32_rdseed,.-OPENSSL_ia32_rdseed
+___
+
close STDOUT; # flush