X-Git-Url: https://git.openssl.org/gitweb/?p=openssl.git;a=blobdiff_plain;f=crypto%2Fx86cpuid.pl;h=d43dda4d935cbdee633b11fc62366d09c2a245ec;hp=176e8e38c8d6f21a260c91abb594e4c7dcbb7c61;hb=bbb02a5b6d27f76931c3385321b2c594781c7a1b;hpb=609b0852e4d50251857dbbac3141ba042e35a9ae

diff --git a/crypto/x86cpuid.pl b/crypto/x86cpuid.pl
index 176e8e38c8..d43dda4d93 100644
--- a/crypto/x86cpuid.pl
+++ b/crypto/x86cpuid.pl
@@ -1,5 +1,5 @@
 #! /usr/bin/env perl
-# Copyright 2004-2016 The OpenSSL Project Authors. All Rights Reserved.
+# Copyright 2004-2018 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
@@ -14,7 +14,7 @@ $output = pop;
 open OUT,">$output";
 *STDOUT=*OUT;
 
-&asm_init($ARGV[0],"x86cpuid");
+&asm_init($ARGV[0]);
 
 for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 
@@ -30,10 +30,10 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 	&pop	("eax");
 	&xor	("ecx","eax");
 	&xor	("eax","eax");
+	&mov	("esi",&wparam(0));
+	&mov	(&DWP(8,"esi"),"eax");	# clear extended feature flags
 	&bt	("ecx",21);
 	&jnc	(&label("nocpuid"));
-	&mov	("esi",&wparam(0));
-	&mov	(&DWP(8,"esi"),"eax");	# clear 3rd word
 	&cpuid	();
 	&mov	("edi","eax");		# max value for standard query level
 
@@ -91,26 +91,16 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 	&jmp	(&label("generic"));
 
 &set_label("intel");
-	&cmp	("edi",7);
-	&jb	(&label("cacheinfo"));
-
-	&mov	("esi",&wparam(0));
-	&mov	("eax",7);
-	&xor	("ecx","ecx");
-	&cpuid	();
-	&mov	(&DWP(8,"esi"),"ebx");
-
-&set_label("cacheinfo");
 	&cmp	("edi",4);
-	&mov	("edi",-1);
+	&mov	("esi",-1);
 	&jb	(&label("nocacheinfo"));
 
 	&mov	("eax",4);
 	&mov	("ecx",0);		# query L1D
 	&cpuid	();
-	&mov	("edi","eax");
-	&shr	("edi",14);
-	&and	("edi",0xfff);		# number of cores -1 per L1D
+	&mov	("esi","eax");
+	&shr	("esi",14);
+	&and	("esi",0xfff);		# number of cores -1 per L1D
 
 &set_label("nocacheinfo");
 	&mov	("eax",1);
@@ -120,7 +110,7 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 	&cmp	("ebp",0);
 	&jne	(&label("notintel"));
 	&or	("edx",1<<30);		# set reserved bit#30 on Intel CPUs
-	&and	(&HB("eax"),15);	# familiy ID
+	&and	(&HB("eax"),15);	# family ID
 	&cmp	(&HB("eax"),15);	# P4?
 	&jne	(&label("notintel"));
 	&or	("edx",1<<20);		# set reserved bit#20 to engage RC4_CHAR
@@ -128,7 +118,7 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 	&bt	("edx",28);		# test hyper-threading bit
 	&jnc	(&label("generic"));
 	&and	("edx",0xefffffff);
-	&cmp	("edi",0);
+	&cmp	("esi",0);
 	&je	(&label("generic"));
 
 	&or	("edx",0x10000000);
@@ -140,10 +130,19 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 &set_label("generic");
 	&and	("ebp",1<<11);		# isolate AMD XOP flag
 	&and	("ecx",0xfffff7ff);	# force 11th bit to 0
-	&mov	("esi","edx");
+	&mov	("esi","edx");		# %ebp:%esi is copy of %ecx:%edx
 	&or	("ebp","ecx");		# merge AMD XOP flag
 
-	&bt	("ecx",27);		# check OSXSAVE bit
+	&cmp	("edi",7);
+	&mov	("edi",&wparam(0));
+	&jb	(&label("no_extended_info"));
+	&mov	("eax",7);
+	&xor	("ecx","ecx");
+	&cpuid	();
+	&mov	(&DWP(8,"edi"),"ebx");	# save extended feature flag
+&set_label("no_extended_info");
+
+	&bt	("ebp",27);		# check OSXSAVE bit
 	&jnc	(&label("clear_avx"));
 	&xor	("ecx","ecx");
 	&data_byte(0x0f,0x01,0xd0);	# xgetbv
@@ -157,7 +156,6 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 	&and	("esi",0xfeffffff);	# clear FXSR
 &set_label("clear_avx");
 	&and	("ebp",0xefffe7ff);	# clear AVX, FMA and AMD XOP bits
-	&mov	("edi",&wparam(0));
 	&and	(&DWP(8,"edi"),0xffffffdf);	# clear AVX2
 &set_label("done");
 	&mov	("eax","esi");
@@ -285,52 +283,13 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
 &set_label("spin");
 	&lea	("ebx",&DWP(0,"eax","ecx"));
 	&nop	();
-	&data_word(0x1ab10ff0);	# lock;	cmpxchg	%ebx,(%edx)	# %eax is envolved and is always reloaded
+	&data_word(0x1ab10ff0);	# lock;	cmpxchg	%ebx,(%edx)	# %eax is involved and is always reloaded
 	&jne	(&label("spin"));
 	&mov	("eax","ebx");	# OpenSSL expects the new value
 	&pop	("ebx");
 	&ret	();
 &function_end_B("OPENSSL_atomic_add");
 
-# This function can become handy under Win32 in situations when
-# we don't know which calling convention, __stdcall or __cdecl(*),
-# indirect callee is using. In C it can be deployed as
-#
-#ifdef OPENSSL_CPUID_OBJ
-#	type OPENSSL_indirect_call(void *f,...);
-#	...
-#	OPENSSL_indirect_call(func,[up to $max arguments]);
-#endif
-#
-# (*)	it's designed to work even for __fastcall if number of
-#	arguments is 1 or 2!
-&function_begin_B("OPENSSL_indirect_call");
-	{
-	my ($max,$i)=(7,);	# $max has to be chosen as 4*n-1
-				# in order to preserve eventual
-				# stack alignment
-	&push	("ebp");
-	&mov	("ebp","esp");
-	&sub	("esp",$max*4);
-	&mov	("ecx",&DWP(12,"ebp"));
-	&mov	(&DWP(0,"esp"),"ecx");
-	&mov	("edx",&DWP(16,"ebp"));
-	&mov	(&DWP(4,"esp"),"edx");
-	for($i=2;$i<$max;$i++)
-		{
-		# Some copies will be redundant/bogus...
-		&mov	("eax",&DWP(12+$i*4,"ebp"));
-		&mov	(&DWP(0+$i*4,"esp"),"eax");
-		}
-	&call_ptr	(&DWP(8,"ebp"));# make the call...
-	&mov	("esp","ebp");	# ... and just restore the stack pointer
-				# without paying attention to what we called,
-				# (__cdecl *func) or (__stdcall *one).
-	&pop	("ebp");
-	&ret	();
-	}
-&function_end_B("OPENSSL_indirect_call");
-
 &function_begin_B("OPENSSL_cleanse");
 	&mov	("edx",&wparam(0));
 	&mov	("ecx",&wparam(1));
@@ -494,18 +453,6 @@ my $max = "ebp";
 
 sub gen_random {
 my $rdop = shift;
-&function_begin_B("OPENSSL_ia32_${rdop}");
-	&mov	("ecx",8);
-&set_label("loop");
-	&${rdop}("eax");
-	&jc	(&label("break"));
-	&loop	(&label("loop"));
-&set_label("break");
-	&cmp	("eax",0);
-	&cmove	("eax","ecx");
-	&ret	();
-&function_end_B("OPENSSL_ia32_${rdop}");
-
 &function_begin_B("OPENSSL_ia32_${rdop}_bytes");
 	&push	("edi");
 	&push	("ebx");
@@ -543,6 +490,7 @@ my $rdop = shift;
 	&jnz	(&label("tail"));
 
 &set_label("done");
+	&xor	("edx","edx");		# Clear random value from registers
 	&pop	("ebx");
 	&pop	("edi");
 	&ret	();