X-Git-Url: https://git.openssl.org/gitweb/?p=openssl.git;a=blobdiff_plain;f=crypto%2Fx86_64cpuid.pl;h=285ad1ba7dba7bfe3bbf620ff984e19615636166;hp=58c7bab1b65ba82d93695f1b80b0fde42c228535;hb=12eaf3b8496f5e3578414202f7d26dc5cec2ee9a;hpb=6251989eb6168c1baada70c5b16c7742861d3d03 diff --git a/crypto/x86_64cpuid.pl b/crypto/x86_64cpuid.pl index 58c7bab1b6..285ad1ba7d 100644 --- a/crypto/x86_64cpuid.pl +++ b/crypto/x86_64cpuid.pl @@ -1,4 +1,11 @@ -#!/usr/bin/env perl +#! /usr/bin/env perl +# Copyright 2005-2016 The OpenSSL Project Authors. All Rights Reserved. +# +# Licensed under the OpenSSL license (the "License"). You may not use +# this file except in compliance with the License. You can obtain a copy +# in the file LICENSE in the source distribution or at +# https://www.openssl.org/source/license.html + $flavour = shift; $output = shift; @@ -11,7 +18,8 @@ $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1; ( $xlate="${dir}perlasm/x86_64-xlate.pl" and -f $xlate) or die "can't locate x86_64-xlate.pl"; -open STDOUT,"| \"$^X\" $xlate $flavour $output"; +open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""; +*STDOUT=*OUT; ($arg1,$arg2,$arg3,$arg4)=$win64?("%rcx","%rdx","%r8", "%r9") : # Win64 order ("%rdi","%rsi","%rdx","%rcx"); # Unix order @@ -23,7 +31,7 @@ print<<___; call OPENSSL_cpuid_setup .hidden OPENSSL_ia32cap_P -.comm OPENSSL_ia32cap_P,8,4 +.comm OPENSSL_ia32cap_P,16,4 .text @@ -52,12 +60,13 @@ OPENSSL_rdtsc: .size OPENSSL_rdtsc,.-OPENSSL_rdtsc .globl OPENSSL_ia32_cpuid -.type OPENSSL_ia32_cpuid,\@abi-omnipotent +.type OPENSSL_ia32_cpuid,\@function,1 .align 16 OPENSSL_ia32_cpuid: mov %rbx,%r8 # save %rbx xor %eax,%eax + mov %eax,8(%rdi) # clear 3rd word cpuid mov %eax,%r11d # max value for standard query level @@ -125,6 +134,14 @@ OPENSSL_ia32_cpuid: shr \$14,%r10d and \$0xfff,%r10d # number of cores -1 per L1D + cmp \$7,%r11d + jb .Lnocacheinfo + + mov \$7,%eax + xor %ecx,%ecx + cpuid + mov %ebx,8(%rdi) + .Lnocacheinfo: mov \$1,%eax cpuid @@ -164,6 +181,7 @@ OPENSSL_ia32_cpuid: .Lclear_avx: mov \$0xefffe7ff,%eax # ~(1<<28|1<<12|1<<11) and %eax,%r9d # clear AVX, FMA and AMD XOP bits + andl \$0xffffffdf,8(%rdi) # cleax AVX2, ~(1<<5) .Ldone: shl \$32,%r9 mov %r10d,%eax @@ -206,6 +224,28 @@ OPENSSL_cleanse: jne .Little ret .size OPENSSL_cleanse,.-OPENSSL_cleanse + +.globl CRYPTO_memcmp +.type CRYPTO_memcmp,\@abi-omnipotent +.align 16 +CRYPTO_memcmp: + xor %rax,%rax + xor %r10,%r10 + cmp \$0,$arg3 + je .Lno_data +.Loop_cmp: + mov ($arg1),%r10b + lea 1($arg1),$arg1 + xor ($arg2),%r10b + lea 1($arg2),$arg2 + or %r10b,%al + dec $arg3 + jnz .Loop_cmp + neg %rax + shr \$63,%rax +.Lno_data: + ret +.size CRYPTO_memcmp,.-CRYPTO_memcmp ___ print<<___ if (!$win64); @@ -368,6 +408,21 @@ OPENSSL_ia32_rdrand: 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