Use the return value from write(2)

[openssl.git] / crypto / x86cpuid.pl

diff --git a/crypto/x86cpuid.pl b/crypto/x86cpuid.pl
index c45b18356bb12814d028cd92c3a94292a2b23229..1266d7634c152777f98896f4ca89a86c679ee2ea 100644 (file)
--- a/crypto/x86cpuid.pl
+++ b/crypto/x86cpuid.pl
@@ -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
 
@@ -89,28 +89,18 @@ for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
        &ja     (&label("generic"));
        &and    ("edx",0xefffffff);     # clear hyper-threading bit
        &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");
+&set_label("intel");
        &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);
@@ -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));