Update x86cpuid.pl to correctly detect shared cache and to support new

RC4_set_key.

diff --git a/crypto/x86cpuid.pl b/crypto/x86cpuid.pl
index c53c9bc9980f6c696eb304713d2b79b5a01822e0..7d924a60b77d78c9352c50bbe89795332a78ee56 100644 (file)
--- a/crypto/x86cpuid.pl
+++ b/crypto/x86cpuid.pl
@@ -20,12 +20,36 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
        &xor    ("ecx","eax");
        &bt     ("ecx",21);
        &jnc    (&label("nocpuid"));
+       &xor    ("eax","eax");
+       &cpuid  ();
+       &xor    ("eax","eax");
+       &cmp    ("ebx",0x756e6547);     # "Genu"
+       &setne  (&LB("eax"));
+       &mov    ("ebp","eax");
+       &cmp    ("edx",0x49656e69);     # "ineI"
+       &setne  (&LB("eax"));
+       &or     ("ebp","eax");
+       &cmp    ("ecx",0x6c65746e);     # "ntel"
+       &setne  (&LB("eax"));
+       &or     ("ebp","eax");
        &mov    ("eax",1);
        &cpuid  ();
+       &bt     ("edx",28);             # test hyper-threading bit
+       &jnc    (&label("nocpuid"));
+       &cmp    ("ebp",0);
+       &jne    (&label("notintel"));
+       &or     ("edx",1<<20);          # use reserved bit to engage RC4_CHAR
+&set_label("notintel");
+       &shr    ("ebx",16);
+       &cmp    (&LB("ebx"),1);         # see if cache is shared(*)
+       &ja     (&label("nocpuid"));
+       &and    ("edx",~(1<<28));       # clear hyper-threading bit if not
 &set_label("nocpuid");
        &mov    ("eax","edx");
        &mov    ("edx","ecx");
 &function_end("OPENSSL_ia32_cpuid");
+# (*)  on Core2 this value is set to 2 denoting the fact that L2
+#      cache is shared between cores.
 
 &external_label("OPENSSL_ia32cap_P");
 

diff --git a/doc/crypto/OPENSSL_ia32cap.pod b/doc/crypto/OPENSSL_ia32cap.pod
index 03ac6e6834bd672dad99eadb583ee9099e13fb07..2e659d34a5c43b29e36b2e622a30fd165c99aa1f 100644 (file)
--- a/doc/crypto/OPENSSL_ia32cap.pod
+++ b/doc/crypto/OPENSSL_ia32cap.pod
@@ -17,20 +17,27 @@ register after executing CPUID instruction with EAX=1 input value (see
 Intel Application Note #241618). Naturally it's meaningful on IA-32[E]
 platforms only. The variable is normally set up automatically upon
 toolkit initialization, but can be manipulated afterwards to modify
-crypto library behaviour. For the moment of this writing five bits are
-significant, namely bit #28 denoting Hyperthreading, which is used to
-distinguish Intel P4 core, bit #26 denoting SSE2 support, bit #25
-denoting SSE support, bit #23 denoting MMX support, and bit #4 denoting
-presence of Time-Stamp Counter. Clearing bit #26 at run-time for
-example disables high-performance SSE2 code present in the crypto
-library. You might have to do this if target OpenSSL application is
-executed on SSE2 capable CPU, but under control of OS which does not
-support SSE2 extentions. Even though you can manipulate the value
-programmatically, you most likely will find it more appropriate to set
-up an environment variable with the same name prior starting target
-application, e.g. 'env OPENSSL_ia32cap=0x12800010 apps/openssl', to
-achieve same effect without modifying the application source code.
-Alternatively you can reconfigure the toolkit with no-sse2 option and
-recompile.
+crypto library behaviour. For the moment of this writing six bits are
+significant, namely:
+
+1. bit #28 denoting Hyperthreading, which is used to distiguish
+   cores with shared cache;
+2. bit #26 denoting SSE2 support;
+3. bit #25 denoting SSE support;
+4. bit #23 denoting MMX support;
+5. bit #20, reserved by Intel, is used to choose between RC4 code
+   pathes;
+6. bit #4 denoting presence of Time-Stamp Counter.
+
+For example, clearing bit #26 at run-time disables high-performance
+SSE2 code present in the crypto library. You might have to do this if
+target OpenSSL application is executed on SSE2 capable CPU, but under
+control of OS which does not support SSE2 extentions. Even though you
+can manipulate the value programmatically, you most likely will find it
+more appropriate to set up an environment variable with the same name
+prior starting target application, e.g. on Intel P4 processor 'env
+OPENSSL_ia32cap=0x12900010 apps/openssl', to achieve same effect
+without modifying the application source code. Alternatively you can
+reconfigure the toolkit with no-sse2 option and recompile.
 
 =cut