prepare for release

we should use its method instead of any generic one.

CHANGES

@@ -2,6 +2,23 @@
  OpenSSL CHANGES
  _______________
 
+ Changes between 1.0.0a and 1.0.0b  [16 Nov 2010]
+
+  *) Fix extension code to avoid race conditions which can result in a buffer
+     overrun vulnerability: resumed sessions must not be modified as they can
+     be shared by multiple threads. CVE-2010-3864
+     [Steve Henson]
+
+  *) Fix WIN32 build system to correctly link an ENGINE directory into
+     a DLL. 
+     [Steve Henson]
+
+ Changes between 1.0.0 and 1.0.0a  [01 Jun 2010]
+
+  *) Check return value of int_rsa_verify in pkey_rsa_verifyrecover 
+     (CVE-2010-1633)
+     [Steve Henson, Peter-Michael Hager <hager@dortmund.net>]
+
  Changes between 0.9.8n and 1.0.0  [29 Mar 2010]
 
   *) Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
@@ -843,6 +860,50 @@
   *) Change 'Configure' script to enable Camellia by default.
      [NTT]
   
+ Changes between 0.9.8o and 0.9.8p [xx XXX xxxx]
+
+  *) Fix extension code to avoid race conditions which can result in a buffer
+     overrun vulnerability: resumed sessions must not be modified as they can
+     be shared by multiple threads. CVE-2010-3864
+
+  *) Fix for double free bug in ssl/s3_clnt.c CVE-2010-2939
+     [Steve Henson]
+
+  *) Don't reencode certificate when calculating signature: cache and use
+     the original encoding instead. This makes signature verification of
+     some broken encodings work correctly.
+     [Steve Henson]
+
+  *) ec2_GF2m_simple_mul bugfix: compute correct result if the output EC_POINT
+     is also one of the inputs.
+     [Emilia K<>sper <emilia.kasper@esat.kuleuven.be> (Google)]
+
+  *) Don't repeatedly append PBE algorithms to table if they already exist.
+     Sort table on each new add. This effectively makes the table read only
+     after all algorithms are added and subsequent calls to PKCS12_pbe_add
+     etc are non-op.
+     [Steve Henson]
+
+ Changes between 0.9.8n and 0.9.8o [01 Jun 2010]
+
+  [NB: OpenSSL 0.9.8o and later 0.9.8 patch levels were released after
+  OpenSSL 1.0.0.]
+
+  *) Correct a typo in the CMS ASN1 module which can result in invalid memory
+     access or freeing data twice (CVE-2010-0742)
+     [Steve Henson, Ronald Moesbergen <intercommit@gmail.com>]
+
+  *) Add SHA2 algorithms to SSL_library_init(). SHA2 is becoming far more
+     common in certificates and some applications which only call
+     SSL_library_init and not OpenSSL_add_all_algorithms() will fail.
+     [Steve Henson]
+
+  *) VMS fixes: 
+     Reduce copying into .apps and .test in makevms.com
+     Don't try to use blank CA certificate in CA.com
+     Allow use of C files from original directories in maketests.com
+     [Steven M. Schweda" <sms@antinode.info>]
+
  Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
 
   *) When rejecting SSL/TLS records due to an incorrect version number, never
@@ -851,8 +912,8 @@
      - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
      the previous behavior could result in a read attempt at NULL when
      receiving specific incorrect SSL/TLS records once record payload
-     protection is active.  (CVE-2010-####)
-     [Bodo Moeller, Adam Langley]
+     protection is active.  (CVE-2010-0740)
+     [Bodo Moeller, Adam Langley <agl@chromium.org>]
 
   *) Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL 
      could be crashed if the relevant tables were not present (e.g. chrooted).

Configure

@@ -492,6 +492,8 @@ my %table=(
 # Win64 targets, WIN64I denotes IA-64 and WIN64A - AMD64
 "VC-WIN64I","cl:-W3 -Gs0 -Gy -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -DUNICODE -D_UNICODE -D_CRT_SECURE_NO_DEPRECATE:::WIN64I::SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN:ia64cpuid.o:ia64.o::aes_core.o aes_cbc.o aes-ia64.o::md5-ia64.o:sha1-ia64.o sha256-ia64.o sha512-ia64.o:::::::ias:win32",
 "VC-WIN64A","cl:-W3 -Gs0 -Gy -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -DUNICODE -D_UNICODE -D_CRT_SECURE_NO_DEPRECATE:::WIN64A::SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN:x86_64cpuid.o:bn_asm.o x86_64-mont.o::aes-x86_64.o::md5-x86_64.o:sha1-x86_64.o sha256-x86_64.o sha512-x86_64.o::rc4-x86_64.o:::wp-x86_64.o:cmll-x86_64.o cmll_misc.o:auto:win32",
+"debug-VC-WIN64I","cl:-W3 -Gs0 -Gy -Zi -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -DUNICODE -D_UNICODE -D_CRT_SECURE_NO_DEPRECATE:::WIN64I::SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN:ia64cpuid.o:ia64.o::aes_core.o aes_cbc.o aes-ia64.o::md5-ia64.o:sha1-ia64.o sha256-ia64.o sha512-ia64.o:::::::ias:win32",
+"debug-VC-WIN64A","cl:-W3 -Gs0 -Gy -Zi -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -DUNICODE -D_UNICODE -D_CRT_SECURE_NO_DEPRECATE:::WIN64A::SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN:x86_64cpuid.o:bn_asm.o x86_64-mont.o::aes-x86_64.o::md5-x86_64.o:sha1-x86_64.o sha256-x86_64.o sha512-x86_64.o::rc4-x86_64.o:::wp-x86_64.o:cmll-x86_64.o cmll_misc.o:auto:win32",
 # x86 Win32 target defaults to ANSI API, if you want UNICODE, complement
 # 'perl Configure VC-WIN32' with '-DUNICODE -D_UNICODE'
 "VC-WIN32","cl:-W3 -WX -Gs0 -GF -Gy -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -D_CRT_SECURE_NO_DEPRECATE:::WIN32::BN_LLONG RC4_INDEX EXPORT_VAR_AS_FN ${x86_gcc_opts}:${x86_asm}:win32n:win32",
@@ -503,7 +505,7 @@ my %table=(
 "BC-32","bcc32::::WIN32::BN_LLONG DES_PTR RC4_INDEX EXPORT_VAR_AS_FN:${no_asm}:win32",
 
 # MinGW
-"mingw", "gcc:-mno-cygwin -DL_ENDIAN -DWIN32_LEAN_AND_MEAN -fomit-frame-pointer -O3 -march=i486 -Wall:::MINGW32:-lws2_32 -lgdi32 -lcrypt32:BN_LLONG ${x86_gcc_des} ${x86_gcc_opts} EXPORT_VAR_AS_FN:${x86_asm}:coff:win32:cygwin-shared:-D_WINDLL -DOPENSSL_USE_APPLINK:-mno-cygwin:.dll.a",
+"mingw", "gcc:-mno-cygwin -DL_ENDIAN -DWIN32_LEAN_AND_MEAN -fomit-frame-pointer -O3 -march=i486 -Wall::-D_MT:MINGW32:-lws2_32 -lgdi32 -lcrypt32:BN_LLONG ${x86_gcc_des} ${x86_gcc_opts} EXPORT_VAR_AS_FN:${x86_asm}:coff:win32:cygwin-shared:-D_WINDLL -DOPENSSL_USE_APPLINK:-mno-cygwin:.dll.a",
 # As for OPENSSL_USE_APPLINK. Applink makes it possible to use .dll
 # compiled with one compiler with application compiled with another
 # compiler. It's possible to engage Applink support in mingw64 build,
@@ -511,7 +513,7 @@ my %table=(
 # handling, one can't seriously consider its binaries for using with
 # non-mingw64 run-time environment. And as mingw64 is always consistent
 # with itself, Applink is never engaged and can as well be omitted.
-"mingw64", "gcc:-mno-cygwin -DL_ENDIAN -O3 -Wall -DWIN32_LEAN_AND_MEAN -DUNICODE -D_UNICODE:::MINGW64:-lws2_32 -lgdi32 -lcrypt32:SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN:${x86_64_asm}:mingw64:win32:cygwin-shared:-D_WINDLL:-mno-cygwin:.dll.a",
+"mingw64", "gcc:-mno-cygwin -DL_ENDIAN -O3 -Wall -DWIN32_LEAN_AND_MEAN -DUNICODE -D_UNICODE::-D_MT:MINGW64:-lws2_32 -lgdi32 -lcrypt32:SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN:${x86_64_asm}:mingw64:win32:cygwin-shared:-D_WINDLL:-mno-cygwin:.dll.a",
 
 # UWIN 
 "UWIN", "cc:-DTERMIOS -DL_ENDIAN -O -Wall:::UWIN::BN_LLONG ${x86_gcc_des} ${x86_gcc_opts}:${no_asm}:win32",
@@ -547,7 +549,7 @@ my %table=(
 
 ##### MacOS X (a.k.a. Rhapsody or Darwin) setup
 "rhapsody-ppc-cc","cc:-O3 -DB_ENDIAN::(unknown):MACOSX_RHAPSODY::BN_LLONG RC4_CHAR RC4_CHUNK DES_UNROLL BF_PTR:${no_asm}::",
-"darwin-ppc-cc","cc:-arch ppc -O3 -DB_ENDIAN::-D_REENTRANT:MACOSX:-Wl,-search_paths_first%:BN_LLONG RC4_CHAR RC4_CHUNK DES_UNROLL BF_PTR:${ppc32_asm}:osx32:dlfcn:darwin-shared:-fPIC -fno-common:-arch ppc -dynamiclib:.\$(SHLIB_MAJOR).\$(SHLIB_MINOR).dylib",
+"darwin-ppc-cc","cc:-arch ppc -O3 -DB_ENDIAN -Wa,-force_cpusubtype_ALL::-D_REENTRANT:MACOSX:-Wl,-search_paths_first%:BN_LLONG RC4_CHAR RC4_CHUNK DES_UNROLL BF_PTR:${ppc32_asm}:osx32:dlfcn:darwin-shared:-fPIC -fno-common:-arch ppc -dynamiclib:.\$(SHLIB_MAJOR).\$(SHLIB_MINOR).dylib",
 "darwin64-ppc-cc","cc:-arch ppc64 -O3 -DB_ENDIAN::-D_REENTRANT:MACOSX:-Wl,-search_paths_first%:SIXTY_FOUR_BIT_LONG RC4_CHAR RC4_CHUNK DES_UNROLL BF_PTR:${ppc64_asm}:osx64:dlfcn:darwin-shared:-fPIC -fno-common:-arch ppc64 -dynamiclib:.\$(SHLIB_MAJOR).\$(SHLIB_MINOR).dylib",
 "darwin-i386-cc","cc:-arch i386 -O3 -fomit-frame-pointer -DL_ENDIAN::-D_REENTRANT:MACOSX:-Wl,-search_paths_first%:BN_LLONG RC4_INT RC4_CHUNK DES_UNROLL BF_PTR:${x86_asm}:macosx:dlfcn:darwin-shared:-fPIC -fno-common:-arch i386 -dynamiclib:.\$(SHLIB_MAJOR).\$(SHLIB_MINOR).dylib",
 "debug-darwin-i386-cc","cc:-arch i386 -g3 -DL_ENDIAN::-D_REENTRANT:MACOSX:-Wl,-search_paths_first%:BN_LLONG RC4_INT RC4_CHUNK DES_UNROLL BF_PTR:${x86_asm}:macosx:dlfcn:darwin-shared:-fPIC -fno-common:-arch i386 -dynamiclib:.\$(SHLIB_MAJOR).\$(SHLIB_MINOR).dylib",
@@ -583,6 +585,7 @@ my %table=(
 );
 
 my @MK1MF_Builds=qw(VC-WIN64I VC-WIN64A
+		    debug-VC-WIN64I debug-VC-WIN64A
 		    VC-NT VC-CE VC-WIN32 debug-VC-WIN32
 		    BC-32 
 		    netware-clib netware-clib-bsdsock
@@ -1093,6 +1096,12 @@ my $ar = $ENV{'AR'} || "ar";
 my $arflags = $fields[$idx_arflags];
 my $multilib = $fields[$idx_multilib];
 
+# if $prefix/lib$multilib is not an existing directory, then
+# assume that it's not searched by linker automatically, in
+# which case adding $multilib suffix causes more grief than
+# we're ready to tolerate, so don't...
+$multilib="" if !-d "$prefix/lib$multilib";
+
 $libdir="lib$multilib" if $libdir eq "";
 
 $cflags = "$cflags$exp_cflags";
@@ -1789,11 +1798,11 @@ EOF
 	(system $make_command.$make_targets) == 0 or exit $?
 		if $make_targets ne "";
 	if ( $perl =~ m@^/@) {
-	    &dofile("tools/c_rehash",$perl,'^#!/', '#!%s','^my \$dir;$', 'my $dir = "' . $openssldir . '";');
+	    &dofile("tools/c_rehash",$perl,'^#!/', '#!%s','^my \$dir;$', 'my $dir = "' . $openssldir . '";', '^my \$prefix;$', 'my $prefix = "' . $prefix . '";');
 	    &dofile("apps/CA.pl",$perl,'^#!/', '#!%s');
 	} else {
 	    # No path for Perl known ...
-	    &dofile("tools/c_rehash",'/usr/local/bin/perl','^#!/', '#!%s','^my \$dir;$', 'my $dir = "' . $openssldir . '";');
+	    &dofile("tools/c_rehash",'/usr/local/bin/perl','^#!/', '#!%s','^my \$dir;$', 'my $dir = "' . $openssldir . '";',  '^my \$prefix;$', 'my $prefix = "' . $prefix . '";');
 	    &dofile("apps/CA.pl",'/usr/local/bin/perl','^#!/', '#!%s');
 	}
 	if ($depflags ne $default_depflags && !$make_depend) {

FAQ

@@ -79,7 +79,7 @@ OpenSSL  -  Frequently Asked Questions
 * Which is the current version of OpenSSL?
 
 The current version is available from <URL: http://www.openssl.org>.
-OpenSSL 1.0.0 was released on Mar 29th, 2010.
+OpenSSL 1.0.0b was released on Nov 16th, 2010.
 
 In addition to the current stable release, you can also access daily
 snapshots of the OpenSSL development version at <URL:
@@ -722,7 +722,7 @@ file.
 Multi-threaded applications must provide two callback functions to
 OpenSSL by calling CRYPTO_set_locking_callback() and
 CRYPTO_set_id_callback(), for all versions of OpenSSL up to and
-including 0.9.8[abc...]. As of version 0.9.9, CRYPTO_set_id_callback()
+including 0.9.8[abc...]. As of version 1.0.0, CRYPTO_set_id_callback()
 and associated APIs are deprecated by CRYPTO_THREADID_set_callback()
 and friends. This is described in the threads(3) manpage.
 

INSTALL.W32

@@ -297,7 +297,18 @@
  desktop, which is not available to service processes. The toolkit is
  designed to detect in which context it's currently executed, GUI,
  console app or service, and act accordingly, namely whether or not to
- actually make GUI calls.
+ actually make GUI calls. Additionally those who wish to
+ /DELAYLOAD:GDI32.DLL and /DELAYLOAD:USER32.DLL and actually keep them
+ off service process should consider implementing and exporting from
+ .exe image in question own _OPENSSL_isservice not relying on USER32.DLL.
+ E.g., on Windows Vista and later you could:
+
+	__declspec(dllexport) __cdecl BOOL _OPENSSL_isservice(void)
+	{   DWORD sess;
+	    if (ProcessIdToSessionId(GetCurrentProcessId(),&sess))
+	        return sess==0;
+	    return FALSE;
+	}
 
  If you link with OpenSSL .DLLs, then you're expected to include into
  your application code small "shim" snippet, which provides glue between

Makefile.shared

@@ -135,7 +135,7 @@ LINK_SO_A_VIA_O=	\
   ALL=$$ALLSYMSFLAGS; ALLSYMSFLAGS=; NOALLSYMSFLAGS=; \
   ( $(SET_X); \
     ld $(LDFLAGS) -r -o lib$(LIBNAME).o $$ALL lib$(LIBNAME).a $(LIBEXTRAS) ); \
-  $(LINK_SO) && rm -f $(LIBNAME).o
+  $(LINK_SO) && rm -f lib$(LIBNAME).o
 
 LINK_SO_A_UNPACKED=	\
   UNPACKDIR=link_tmp.$$$$; rm -rf $$UNPACKDIR; mkdir $$UNPACKDIR; \
@@ -207,17 +207,29 @@ link_app.bsd:
 	fi; $(LINK_APP)
 
 # For Darwin AKA Mac OS/X (dyld)
-# link_o.darwin produces .so, because we let it use dso_dlfcn module,
-# which has .so extension hard-coded. One can argue that one should
-# develop special dso module for MacOS X. At least manual encourages
-# to use native NSModule(3) API and refers to dlfcn as termporary hack.
+# Originally link_o.darwin produced .so, because it was hard-coded
+# in dso_dlfcn module. At later point dso_dlfcn switched to .dylib
+# extension in order to allow for run-time linking with vendor-
+# supplied shared libraries such as libz, so that link_o.darwin had
+# to be harmonized with it. This caused minor controversy, because
+# it was believed that dlopen can't be used to dynamically load
+# .dylib-s, only so called bundle modules (ones linked with -bundle
+# flag). The belief seems to be originating from pre-10.4 release,
+# where dlfcn functionality was emulated by dlcompat add-on. In
+# 10.4 dlopen was rewritten as native part of dyld and is documented
+# to be capable of loading both dynamic libraries and bundles. In
+# order to provide compatibility with pre-10.4 dlopen, modules are
+# linked with -bundle flag, which makes .dylib extension misleading.
+# It works, because dlopen is [and always was] extension-agnostic.
+# Alternative to this heuristic approach is to develop specific
+# MacOS X dso module relying on whichever "native" dyld interface.
 link_o.darwin:
 	@ $(CALC_VERSIONS); \
 	SHLIB=lib$(LIBNAME); \
-	SHLIB_SUFFIX=.so; \
+	SHLIB_SUFFIX=.dylib; \
 	ALLSYMSFLAGS='-all_load'; \
 	NOALLSYMSFLAGS=''; \
-	SHAREDFLAGS="$(CFLAGS) $(SHARED_LDFLAGS)"; \
+	SHAREDFLAGS="$(CFLAGS) `echo $(SHARED_LDFLAGS) | sed s/dynamiclib/bundle/`"; \
 	if [ -n "$(LIBVERSION)" ]; then \
 		SHAREDFLAGS="$$SHAREDFLAGS -current_version $(LIBVERSION)"; \
 	fi; \

NEWS

@@ -5,6 +5,17 @@
   This file gives a brief overview of the major changes between each OpenSSL
   release. For more details please read the CHANGES file.
 
+  Major changes between OpenSSL 1.0.0a and OpenSSL 1.0.0b:
+
+      o Fix for security issue CVE-2010-3864.
+      o Fix for CVE-2010-2939
+      o Fix WIN32 build system for GOST ENGINE.
+
+  Major changes between OpenSSL 1.0.0 and OpenSSL 1.0.0a:
+
+      o Fix for security issue CVE-2010-1633.
+      o GOST MAC and CFB fixes.
+
   Major changes between OpenSSL 0.9.8n and OpenSSL 1.0:
 
       o RFC3280 path validation: sufficient to process PKITS tests.
@@ -28,6 +39,14 @@
       o Opaque PRF Input TLS extension support.
       o Updated time routines to avoid OS limitations.
 
+  Major changes between OpenSSL 0.9.8n and OpenSSL 0.9.8o:
+
+      o Fix for security issue CVE-2010-0742.
+      o Various DTLS fixes.
+      o Recognise SHA2 certificates if only SSL algorithms added.
+      o Fix for no-rc4 compilation.
+      o Chil ENGINE unload workaround.
+
   Major changes between OpenSSL 0.9.8m and OpenSSL 0.9.8n:
 
       o CFB cipher definition fixes.

PROBLEMS

@@ -36,7 +36,9 @@ may differ on your machine.
 
 
 As long as Apple doesn't fix the problem with ld, this problem building
-OpenSSL will remain as is.
+OpenSSL will remain as is. Well, the problem was addressed in 0.9.8f by
+passing -Wl,-search_paths_first, but it's unknown if the flag was
+supported from the initial MacOS X release.
 
 
 * Parallell make leads to errors

README

@@ -1,7 +1,7 @@
 
- OpenSSL 1.0.0 29 Mar 2010
+ OpenSSL 1.0.0b 16 Nov 2010
 
- Copyright (c) 1998-2009 The OpenSSL Project
+ Copyright (c) 1998-2010 The OpenSSL Project
  Copyright (c) 1995-1998 Eric A. Young, Tim J. Hudson
  All rights reserved.
 

STATUS

@@ -1,9 +1,12 @@
 
   OpenSSL STATUS                           Last modified at
-  ______________                           $Date: 2010/03/29 13:11:53 $
+  ______________                           $Date: 2010/11/16 13:35:08 $
 
   DEVELOPMENT STATE
 
+    o  OpenSSL 1.1.0:  Under development...
+    o  OpenSSL 1.0.0b: Released on November  16th, 2010
+    o  OpenSSL 1.0.0a: Released on June      1st,  2010
     o  OpenSSL 1.0.0:  Released on March     29th, 2010
     o  OpenSSL 0.9.8n: Released on March     24th, 2010
     o  OpenSSL 0.9.8m: Released on February  25th, 2010

TABLE

@@ -1211,7 +1211,7 @@ $multilib     =
 
 *** darwin-ppc-cc
 $cc           = cc
-$cflags       = -arch ppc -O3 -DB_ENDIAN
+$cflags       = -arch ppc -O3 -DB_ENDIAN -Wa,-force_cpusubtype_ALL
 $unistd       = 
 $thread_cflag = -D_REENTRANT
 $sys_id       = MACOSX
@@ -1426,6 +1426,68 @@ $ranlib       =
 $arflags      = 
 $multilib     = 
 
+*** debug-VC-WIN64A
+$cc           = cl
+$cflags       = -W3 -Gs0 -Gy -Zi -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -DUNICODE -D_UNICODE -D_CRT_SECURE_NO_DEPRECATE
+$unistd       = 
+$thread_cflag = 
+$sys_id       = WIN64A
+$lflags       = 
+$bn_ops       = SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN
+$cpuid_obj    = x86_64cpuid.o
+$bn_obj       = bn_asm.o x86_64-mont.o
+$des_obj      = 
+$aes_obj      = aes-x86_64.o
+$bf_obj       = 
+$md5_obj      = md5-x86_64.o
+$sha1_obj     = sha1-x86_64.o sha256-x86_64.o sha512-x86_64.o
+$cast_obj     = 
+$rc4_obj      = rc4-x86_64.o
+$rmd160_obj   = 
+$rc5_obj      = 
+$wp_obj       = wp-x86_64.o
+$cmll_obj     = cmll-x86_64.o cmll_misc.o
+$perlasm_scheme = auto
+$dso_scheme   = win32
+$shared_target= 
+$shared_cflag = 
+$shared_ldflag = 
+$shared_extension = 
+$ranlib       = 
+$arflags      = 
+$multilib     = 
+
+*** debug-VC-WIN64I
+$cc           = cl
+$cflags       = -W3 -Gs0 -Gy -Zi -nologo -DOPENSSL_SYSNAME_WIN32 -DWIN32_LEAN_AND_MEAN -DL_ENDIAN -DUNICODE -D_UNICODE -D_CRT_SECURE_NO_DEPRECATE
+$unistd       = 
+$thread_cflag = 
+$sys_id       = WIN64I
+$lflags       = 
+$bn_ops       = SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN
+$cpuid_obj    = ia64cpuid.o
+$bn_obj       = ia64.o
+$des_obj      = 
+$aes_obj      = aes_core.o aes_cbc.o aes-ia64.o
+$bf_obj       = 
+$md5_obj      = md5-ia64.o
+$sha1_obj     = sha1-ia64.o sha256-ia64.o sha512-ia64.o
+$cast_obj     = 
+$rc4_obj      = 
+$rmd160_obj   = 
+$rc5_obj      = 
+$wp_obj       = 
+$cmll_obj     = 
+$perlasm_scheme = ias
+$dso_scheme   = win32
+$shared_target= 
+$shared_cflag = 
+$shared_ldflag = 
+$shared_extension = 
+$ranlib       = 
+$arflags      = 
+$multilib     = 
+
 *** debug-ben
 $cc           = gcc
 $cflags       = -DBN_DEBUG -DREF_CHECK -DCONF_DEBUG -DBN_CTX_DEBUG -DCRYPTO_MDEBUG -DPEDANTIC -DDEBUG_SAFESTACK -DDEBUG_UNUSED -O2 -pedantic -Wall -Wshadow -Werror -pipe
@@ -3879,7 +3941,7 @@ $multilib     = 64
 $cc           = gcc
 $cflags       = -mno-cygwin -DL_ENDIAN -DWIN32_LEAN_AND_MEAN -fomit-frame-pointer -O3 -march=i486 -Wall
 $unistd       = 
-$thread_cflag = 
+$thread_cflag = -D_MT
 $sys_id       = MINGW32
 $lflags       = -lws2_32 -lgdi32 -lcrypt32
 $bn_ops       = BN_LLONG DES_PTR DES_RISC1 DES_UNROLL RC4_INDEX MD2_INT EXPORT_VAR_AS_FN
@@ -3910,7 +3972,7 @@ $multilib     =
 $cc           = gcc
 $cflags       = -mno-cygwin -DL_ENDIAN -O3 -Wall -DWIN32_LEAN_AND_MEAN -DUNICODE -D_UNICODE
 $unistd       = 
-$thread_cflag = 
+$thread_cflag = -D_MT
 $sys_id       = MINGW64
 $lflags       = -lws2_32 -lgdi32 -lcrypt32
 $bn_ops       = SIXTY_FOUR_BIT RC4_CHUNK_LL DES_INT EXPORT_VAR_AS_FN

apps/CA.com

@@ -114,8 +114,8 @@ $!
 $   IF F$SEARCH(CATOP+".private"+CAKEY) .EQS. ""
 $   THEN
 $     READ '__INPUT' FILE -
-	   /PROMT="CA certificate filename (or enter to create)"
-$     IF F$SEARCH(FILE) .NES. ""
+	   /PROMPT="CA certificate filename (or enter to create)"
+$     IF (FILE .NES. "") .AND. (F$SEARCH(FILE) .NES. "")
 $     THEN
 $       COPY 'FILE' 'CATOP'.private'CAKEY'
 $	RET=$STATUS

apps/apps.c

@@ -257,6 +257,8 @@ int args_from_file(char *file, int *argc, char **argv[])
 
 int str2fmt(char *s)
 	{
+	if (s == NULL)
+		return FORMAT_UNDEF;
 	if 	((*s == 'D') || (*s == 'd'))
 		return(FORMAT_ASN1);
 	else if ((*s == 'T') || (*s == 't'))
@@ -377,13 +379,12 @@ void program_name(char *in, char *out, int size)
 
 int chopup_args(ARGS *arg, char *buf, int *argc, char **argv[])
 	{
-	int num,len,i;
+	int num,i;
 	char *p;
 
 	*argc=0;
 	*argv=NULL;
 
-	len=strlen(buf);
 	i=0;
 	if (arg->count == 0)
 		{
@@ -875,10 +876,17 @@ EVP_PKEY *load_key(BIO *err, const char *file, int format, int maybe_stdin,
 	if (format == FORMAT_ENGINE)
 		{
 		if (!e)
-			BIO_printf(bio_err,"no engine specified\n");
+			BIO_printf(err,"no engine specified\n");
 		else
+			{
 			pkey = ENGINE_load_private_key(e, file,
 				ui_method, &cb_data);
+			if (!pkey) 
+				{
+				BIO_printf(err,"cannot load %s from engine\n",key_descrip);
+				ERR_print_errors(err);
+				}	
+			}
 		goto end;
 		}
 #endif
@@ -923,7 +931,7 @@ EVP_PKEY *load_key(BIO *err, const char *file, int format, int maybe_stdin,
 				&pkey, NULL, NULL))
 			goto end;
 		}
-#if !defined(OPENSSL_NO_RSA) && !defined(OPENSSL_NO_DSA)
+#if !defined(OPENSSL_NO_RSA) && !defined(OPENSSL_NO_DSA) && !defined (OPENSSL_NO_RC4)
 	else if (format == FORMAT_MSBLOB)
 		pkey = b2i_PrivateKey_bio(key);
 	else if (format == FORMAT_PVK)
@@ -937,8 +945,11 @@ EVP_PKEY *load_key(BIO *err, const char *file, int format, int maybe_stdin,
 		}
  end:
 	if (key != NULL) BIO_free(key);
-	if (pkey == NULL)
+	if (pkey == NULL) 
+		{
 		BIO_printf(err,"unable to load %s\n", key_descrip);
+		ERR_print_errors(err);
+		}	
 	return(pkey);
 	}
 

apps/dh.c

@@ -88,9 +88,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	DH *dh=NULL;
 	int i,badops=0,text=0;
 	BIO *in=NULL,*out=NULL;
@@ -189,7 +186,7 @@ bad:
 	ERR_load_crypto_strings();
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	in=BIO_new(BIO_s_file());

apps/dhparam.c

@@ -149,9 +149,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	DH *dh=NULL;
 	int i,badops=0,text=0;
 #ifndef OPENSSL_NO_DSA
@@ -270,7 +267,7 @@ bad:
 	ERR_load_crypto_strings();
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	if (g && !num)

apps/dsa.c

@@ -334,7 +334,7 @@ bad:
 			i=PEM_write_bio_DSA_PUBKEY(out,dsa);
 		else i=PEM_write_bio_DSAPrivateKey(out,dsa,enc,
 							NULL,0,NULL, passout);
-#ifndef OPENSSL_NO_RSA
+#if !defined(OPENSSL_NO_RSA) && !defined(OPENSSL_NO_RC4)
 	} else if (outformat == FORMAT_MSBLOB || outformat == FORMAT_PVK) {
 		EVP_PKEY *pk;
 		pk = EVP_PKEY_new();

apps/dsaparam.c

@@ -111,9 +111,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	DSA *dsa=NULL;
 	int i,badops=0,text=0;
 	BIO *in=NULL,*out=NULL;
@@ -278,7 +275,7 @@ bad:
 		}
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	if (need_rand)
@@ -357,12 +354,10 @@ bad:
 	if (C)
 		{
 		unsigned char *data;
-		int l,len,bits_p,bits_q,bits_g;
+		int l,len,bits_p;
 
 		len=BN_num_bytes(dsa->p);
 		bits_p=BN_num_bits(dsa->p);
-		bits_q=BN_num_bits(dsa->q);
-		bits_g=BN_num_bits(dsa->g);
 		data=(unsigned char *)OPENSSL_malloc(len+20);
 		if (data == NULL)
 			{

apps/ec.c

@@ -85,9 +85,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 {
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE 	*e = NULL;
-#endif
 	int 	ret = 1;
 	EC_KEY 	*eckey = NULL;
 	const EC_GROUP *group;
@@ -254,7 +251,7 @@ bad:
 	ERR_load_crypto_strings();
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	if(!app_passwd(bio_err, passargin, passargout, &passin, &passout)) 

apps/ecparam.c

@@ -129,9 +129,6 @@ int MAIN(int argc, char **argv)
 	char	*infile = NULL, *outfile = NULL, *prog;
 	BIO 	*in = NULL, *out = NULL;
 	int 	informat, outformat, noout = 0, C = 0, ret = 1;
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE	*e = NULL;
-#endif
 	char	*engine = NULL;
 
 	BIGNUM	*ec_p = NULL, *ec_a = NULL, *ec_b = NULL,
@@ -340,7 +337,7 @@ bad:
 		}
 
 #ifndef OPENSSL_NO_ENGINE
-	e = setup_engine(bio_err, engine, 0);
+	setup_engine(bio_err, engine, 0);
 #endif
 
 	if (list_curves)

apps/enc.c

@@ -101,9 +101,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	static const char magic[]="Salted__";
 	char mbuf[sizeof magic-1];
 	char *strbuf=NULL;
@@ -328,7 +325,7 @@ bad:
 		}
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	if (md && (dgst=EVP_get_digestbyname(md)) == NULL)

apps/gendh.c

@@ -89,9 +89,6 @@ int MAIN(int, char **);
 int MAIN(int argc, char **argv)
 	{
 	BN_GENCB cb;
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	DH *dh=NULL;
 	int ret=1,num=DEFBITS;
 	int g=2;
@@ -163,7 +160,7 @@ bad:
 		}
 		
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	out=BIO_new(BIO_s_file());

apps/gendsa.c

@@ -78,9 +78,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	DSA *dsa=NULL;
 	int ret=1;
 	char *outfile=NULL;
@@ -206,7 +203,7 @@ bad:
 		}
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	if(!app_passwd(bio_err, NULL, passargout, NULL, &passout)) {

apps/makeapps.com

@@ -25,7 +25,7 @@ $!	   VAXC	 For VAX C.
 $!	   DECC	 For DEC C.
 $!	   GNUC	 For GNU C.
 $!
-$!  If you don't speficy a compiler, it will try to determine which
+$!  If you don't specify a compiler, it will try to determine which
 $!  "C" compiler to use.
 $!
 $!  P3, if defined, sets a TCP/IP library to use, through one of the following
@@ -52,7 +52,7 @@ $ THEN
 $!
 $!  The Architecture Is VAX.
 $!
-$   ARCH := VAX
+$   ARCH = "VAX"
 $!
 $! Else...
 $!
@@ -555,7 +555,7 @@ $!    Time To EXIT.
 $!
 $     EXIT
 $!
-$!  End The Valid Arguement Check.
+$!  End The Valid Argument Check.
 $!
 $   ENDIF
 $!
@@ -770,7 +770,7 @@ $!  Set up default defines
 $!
 $   CCDEFS = """FLAT_INC=1""," + CCDEFS
 $!
-$!  Else The User Entered An Invalid Arguement.
+$!  Else The User Entered An Invalid Argument.
 $!
 $ ELSE
 $!
@@ -875,7 +875,7 @@ $!  Print info
 $!
 $   WRITE SYS$OUTPUT "TCP/IP library spec: ", TCPIP_LIB
 $!
-$!  Else The User Entered An Invalid Arguement.
+$!  Else The User Entered An Invalid Argument.
 $!
 $ ELSE
 $!

apps/pkcs7.c

@@ -82,9 +82,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	PKCS7 *p7=NULL;
 	int i,badops=0;
 	BIO *in=NULL,*out=NULL;
@@ -182,7 +179,7 @@ bad:
 	ERR_load_crypto_strings();
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	in=BIO_new(BIO_s_file());

apps/pkeyparam.c

@@ -74,7 +74,6 @@ int MAIN(int argc, char **argv)
 	EVP_PKEY *pkey=NULL;
 	int badarg = 0;
 #ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
 	char *engine=NULL;
 #endif
 	int ret = 1;
@@ -141,7 +140,7 @@ int MAIN(int argc, char **argv)
 		}
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	if (infile)

apps/pkeyutl.c

@@ -119,17 +119,17 @@ int MAIN(int argc, char **argv)
 		if (!strcmp(*argv,"-in"))
 			{
 			if (--argc < 1) badarg = 1;
-                        infile= *(++argv);
+                        else infile= *(++argv);
 			}
 		else if (!strcmp(*argv,"-out"))
 			{
 			if (--argc < 1) badarg = 1;
-			outfile= *(++argv);
+			else outfile= *(++argv);
 			}
 		else if (!strcmp(*argv,"-sigfile"))
 			{
 			if (--argc < 1) badarg = 1;
-			sigfile= *(++argv);
+			else sigfile= *(++argv);
 			}
 		else if(!strcmp(*argv, "-inkey"))
 			{
@@ -159,17 +159,17 @@ int MAIN(int argc, char **argv)
 		else if (!strcmp(*argv,"-passin"))
 			{
 			if (--argc < 1) badarg = 1;
-			passargin= *(++argv);
+			else passargin= *(++argv);
 			}
 		else if (strcmp(*argv,"-peerform") == 0)
 			{
 			if (--argc < 1) badarg = 1;
-			peerform=str2fmt(*(++argv));
+			else peerform=str2fmt(*(++argv));
 			}
 		else if (strcmp(*argv,"-keyform") == 0)
 			{
 			if (--argc < 1) badarg = 1;
-			keyform=str2fmt(*(++argv));
+			else keyform=str2fmt(*(++argv));
 			}
 #ifndef OPENSSL_NO_ENGINE
 		else if(!strcmp(*argv, "-engine"))

apps/rand.c

@@ -77,9 +77,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	int i, r, ret = 1;
 	int badopt;
 	char *outfile = NULL;
@@ -178,7 +175,7 @@ int MAIN(int argc, char **argv)
 		}
 
 #ifndef OPENSSL_NO_ENGINE
-        e = setup_engine(bio_err, engine, 0);
+        setup_engine(bio_err, engine, 0);
 #endif
 
 	app_RAND_load_file(NULL, bio_err, (inrand != NULL));

apps/rsa.c

@@ -409,7 +409,7 @@ bad:
 			}
 		else i=PEM_write_bio_RSAPrivateKey(out,rsa,
 						enc,NULL,0,NULL,passout);
-#ifndef OPENSSL_NO_DSA
+#if !defined(OPENSSL_NO_DSA) && !defined(OPENSSL_NO_RC4)
 	} else if (outformat == FORMAT_MSBLOB || outformat == FORMAT_PVK) {
 		EVP_PKEY *pk;
 		pk = EVP_PKEY_new();

apps/s_server.c

@@ -2254,11 +2254,10 @@ static int www_body(char *hostname, int s, unsigned char *context)
 	{
 	char *buf=NULL;
 	int ret=1;
-	int i,j,k,blank,dot;
+	int i,j,k,dot;
 	SSL *con;
 	const SSL_CIPHER *c;
 	BIO *io,*ssl_bio,*sbio;
-	long total_bytes;
 
 	buf=OPENSSL_malloc(bufsize);
 	if (buf == NULL) return(0);
@@ -2329,7 +2328,6 @@ static int www_body(char *hostname, int s, unsigned char *context)
 		SSL_set_msg_callback_arg(con, bio_s_out);
 		}
 
-	blank=0;
 	for (;;)
 		{
 		if (hack)
@@ -2559,7 +2557,6 @@ static int www_body(char *hostname, int s, unsigned char *context)
                                         BIO_puts(io,"HTTP/1.0 200 ok\r\nContent-type: text/plain\r\n\r\n");
                                 }
 			/* send the file */
-			total_bytes=0;
 			for (;;)
 				{
 				i=BIO_read(file,buf,bufsize);

apps/s_socket.c

@@ -327,7 +327,7 @@ static int init_server_long(int *sock, int port, char *ip, int type)
 	{
 	int ret=0;
 	struct sockaddr_in server;
-	int s= -1,i;
+	int s= -1;
 
 	if (!ssl_sock_init()) return(0);
 
@@ -366,7 +366,6 @@ static int init_server_long(int *sock, int port, char *ip, int type)
 		}
 	/* Make it 128 for linux */
 	if (type==SOCK_STREAM && listen(s,128) == -1) goto err;
-	i=0;
 	*sock=s;
 	ret=1;
 err:
@@ -384,7 +383,7 @@ static int init_server(int *sock, int port, int type)
 
 static int do_accept(int acc_sock, int *sock, char **host)
 	{
-	int ret,i;
+	int ret;
 	struct hostent *h1,*h2;
 	static struct sockaddr_in from;
 	int len;
@@ -407,6 +406,7 @@ redoit:
 	if (ret == INVALID_SOCKET)
 		{
 #if defined(OPENSSL_SYS_WINDOWS) || (defined(OPENSSL_SYS_NETWARE) && !defined(NETWARE_BSDSOCK))
+		int i;
 		i=WSAGetLastError();
 		BIO_printf(bio_err,"accept error %d\n",i);
 #else
@@ -461,7 +461,6 @@ redoit:
 			BIO_printf(bio_err,"gethostbyname failure\n");
 			return(0);
 			}
-		i=0;
 		if (h2->h_addrtype != AF_INET)
 			{
 			BIO_printf(bio_err,"gethostbyname addr is not AF_INET\n");

apps/speed.c

@@ -334,9 +334,6 @@ int MAIN(int, char **);
 
 int MAIN(int argc, char **argv)
 	{
-#ifndef OPENSSL_NO_ENGINE
-	ENGINE *e = NULL;
-#endif
 	unsigned char *buf=NULL,*buf2=NULL;
 	int mret=1;
 	long count=0,save_count=0;
@@ -430,7 +427,6 @@ int MAIN(int argc, char **argv)
 	unsigned char DES_iv[8];
 	unsigned char iv[2*MAX_BLOCK_SIZE/8];
 #ifndef OPENSSL_NO_DES
-	DES_cblock *buf_as_des_cblock = NULL;
 	static DES_cblock key ={0x12,0x34,0x56,0x78,0x9a,0xbc,0xde,0xf0};
 	static DES_cblock key2={0x34,0x56,0x78,0x9a,0xbc,0xde,0xf0,0x12};
 	static DES_cblock key3={0x56,0x78,0x9a,0xbc,0xde,0xf0,0x12,0x34};
@@ -644,9 +640,6 @@ int MAIN(int argc, char **argv)
 		BIO_printf(bio_err,"out of memory\n");
 		goto end;
 		}
-#ifndef OPENSSL_NO_DES
-	buf_as_des_cblock = (DES_cblock *)buf;
-#endif
 	if ((buf2=(unsigned char *)OPENSSL_malloc((int)BUFSIZE)) == NULL)
 		{
 		BIO_printf(bio_err,"out of memory\n");
@@ -721,7 +714,7 @@ int MAIN(int argc, char **argv)
 				BIO_printf(bio_err,"no engine given\n");
 				goto end;
 				}
-                        e = setup_engine(bio_err, *argv, 0);
+                        setup_engine(bio_err, *argv, 0);
 			/* j will be increased again further down.  We just
 			   don't want speed to confuse an engine with an
 			   algorithm, especially when none is given (which

apps/x509.c

@@ -556,7 +556,6 @@ bad:
 	if (reqfile)
 		{
 		EVP_PKEY *pkey;
-		X509_CINF *ci;
 		BIO *in;
 
 		if (!sign_flag && !CA_flag)
@@ -624,7 +623,6 @@ bad:
 		print_name(bio_err, "subject=", X509_REQ_get_subject_name(req), nmflag);
 
 		if ((x=X509_new()) == NULL) goto end;
-		ci=x->cert_info;
 
 		if (sno == NULL)
 			{

crypto/Makefile

@@ -74,7 +74,9 @@ x86_64cpuid.s: x86_64cpuid.pl
 	$(PERL) x86_64cpuid.pl $(PERLASM_SCHEME) > $@
 ia64cpuid.s: ia64cpuid.S
 	$(CC) $(CFLAGS) -E ia64cpuid.S > $@
-ppccpuid.s:		ppccpuid.pl;	$(PERL) ppccpuid.pl $(PERLASM_SCHEME) $@
+ppccpuid.s:	ppccpuid.pl;	$(PERL) ppccpuid.pl $(PERLASM_SCHEME) $@
+alphacpuid.s:	alphacpuid.pl
+	$(PERL) $< | $(CC) -E - | tee $@ > /dev/null
 
 testapps:
 	[ -z "$(THIS)" ] || (	if echo $(SDIRS) | fgrep ' des '; \

crypto/aes/aes_wrap.c

@@ -85,9 +85,9 @@ int AES_wrap_key(AES_KEY *key, const unsigned char *iv,
 			A[7] ^= (unsigned char)(t & 0xff);
 			if (t > 0xff)	
 				{
-				A[6] ^= (unsigned char)((t & 0xff) >> 8);
-				A[5] ^= (unsigned char)((t & 0xff) >> 16);
-				A[4] ^= (unsigned char)((t & 0xff) >> 24);
+				A[6] ^= (unsigned char)((t >> 8) & 0xff);
+				A[5] ^= (unsigned char)((t >> 16) & 0xff);
+				A[4] ^= (unsigned char)((t >> 24) & 0xff);
 				}
 			memcpy(R, B + 8, 8);
 			}
@@ -119,9 +119,9 @@ int AES_unwrap_key(AES_KEY *key, const unsigned char *iv,
 			A[7] ^= (unsigned char)(t & 0xff);
 			if (t > 0xff)	
 				{
-				A[6] ^= (unsigned char)((t & 0xff) >> 8);
-				A[5] ^= (unsigned char)((t & 0xff) >> 16);
-				A[4] ^= (unsigned char)((t & 0xff) >> 24);
+				A[6] ^= (unsigned char)((t >> 8) & 0xff);
+				A[5] ^= (unsigned char)((t >> 16) & 0xff);
+				A[4] ^= (unsigned char)((t >> 24) & 0xff);
 				}
 			memcpy(B + 8, R, 8);
 			AES_decrypt(B, B, key);

crypto/aes/asm/aes-armv4.pl

@@ -16,12 +16,20 @@
 # allows to merge logical or arithmetic operation with shift or rotate
 # in one instruction and emit combined result every cycle. The module
 # is endian-neutral. The performance is ~42 cycles/byte for 128-bit
-# key.
+# key [on single-issue Xscale PXA250 core].
 
 # May 2007.
 #
 # AES_set_[en|de]crypt_key is added.
 
+# July 2010.
+#
+# Rescheduling for dual-issue pipeline resulted in 12% improvement on
+# Cortex A8 core and ~25 cycles per byte processed with 128-bit key.
+
+while (($output=shift) && ($output!~/^\w[\w\-]*\.\w+$/)) {}
+open STDOUT,">$output";
+
 $s0="r0";
 $s1="r1";
 $s2="r2";
@@ -164,24 +172,24 @@ AES_encrypt:
 	ldrb	$t2,[$rounds,#1]
 	ldrb	$t3,[$rounds,#0]
 	orr	$s0,$s0,$t1,lsl#8
-	orr	$s0,$s0,$t2,lsl#16
-	orr	$s0,$s0,$t3,lsl#24
 	ldrb	$s1,[$rounds,#7]
+	orr	$s0,$s0,$t2,lsl#16
 	ldrb	$t1,[$rounds,#6]
+	orr	$s0,$s0,$t3,lsl#24
 	ldrb	$t2,[$rounds,#5]
 	ldrb	$t3,[$rounds,#4]
 	orr	$s1,$s1,$t1,lsl#8
-	orr	$s1,$s1,$t2,lsl#16
-	orr	$s1,$s1,$t3,lsl#24
 	ldrb	$s2,[$rounds,#11]
+	orr	$s1,$s1,$t2,lsl#16
 	ldrb	$t1,[$rounds,#10]
+	orr	$s1,$s1,$t3,lsl#24
 	ldrb	$t2,[$rounds,#9]
 	ldrb	$t3,[$rounds,#8]
 	orr	$s2,$s2,$t1,lsl#8
-	orr	$s2,$s2,$t2,lsl#16
-	orr	$s2,$s2,$t3,lsl#24
 	ldrb	$s3,[$rounds,#15]
+	orr	$s2,$s2,$t2,lsl#16
 	ldrb	$t1,[$rounds,#14]
+	orr	$s2,$s2,$t3,lsl#24
 	ldrb	$t2,[$rounds,#13]
 	ldrb	$t3,[$rounds,#12]
 	orr	$s3,$s3,$t1,lsl#8
@@ -196,24 +204,24 @@ AES_encrypt:
 	mov	$t3,$s0,lsr#8
 	strb	$t1,[$rounds,#0]
 	strb	$t2,[$rounds,#1]
-	strb	$t3,[$rounds,#2]
-	strb	$s0,[$rounds,#3]
 	mov	$t1,$s1,lsr#24
+	strb	$t3,[$rounds,#2]
 	mov	$t2,$s1,lsr#16
+	strb	$s0,[$rounds,#3]
 	mov	$t3,$s1,lsr#8
 	strb	$t1,[$rounds,#4]
 	strb	$t2,[$rounds,#5]
-	strb	$t3,[$rounds,#6]
-	strb	$s1,[$rounds,#7]
 	mov	$t1,$s2,lsr#24
+	strb	$t3,[$rounds,#6]
 	mov	$t2,$s2,lsr#16
+	strb	$s1,[$rounds,#7]
 	mov	$t3,$s2,lsr#8
 	strb	$t1,[$rounds,#8]
 	strb	$t2,[$rounds,#9]
-	strb	$t3,[$rounds,#10]
-	strb	$s2,[$rounds,#11]
 	mov	$t1,$s3,lsr#24
+	strb	$t3,[$rounds,#10]
 	mov	$t2,$s3,lsr#16
+	strb	$s2,[$rounds,#11]
 	mov	$t3,$s3,lsr#8
 	strb	$t1,[$rounds,#12]
 	strb	$t2,[$rounds,#13]
@@ -230,141 +238,137 @@ AES_encrypt:
 .align	2
 _armv4_AES_encrypt:
 	str	lr,[sp,#-4]!		@ push lr
-	ldr	$t1,[$key],#16
-	ldr	$t2,[$key,#-12]
-	ldr	$t3,[$key,#-8]
-	ldr	$i1,[$key,#-4]
-	ldr	$rounds,[$key,#240-16]
+	ldmia	$key!,{$t1-$i1}
 	eor	$s0,$s0,$t1
+	ldr	$rounds,[$key,#240-16]
 	eor	$s1,$s1,$t2
 	eor	$s2,$s2,$t3
 	eor	$s3,$s3,$i1
 	sub	$rounds,$rounds,#1
 	mov	lr,#255
 
-.Lenc_loop:
+	and	$i1,lr,$s0
 	and	$i2,lr,$s0,lsr#8
 	and	$i3,lr,$s0,lsr#16
-	and	$i1,lr,$s0
 	mov	$s0,$s0,lsr#24
+.Lenc_loop:
 	ldr	$t1,[$tbl,$i1,lsl#2]	@ Te3[s0>>0]
-	ldr	$s0,[$tbl,$s0,lsl#2]	@ Te0[s0>>24]
-	ldr	$t2,[$tbl,$i2,lsl#2]	@ Te2[s0>>8]
-	ldr	$t3,[$tbl,$i3,lsl#2]	@ Te1[s0>>16]
-
 	and	$i1,lr,$s1,lsr#16	@ i0
+	ldr	$t2,[$tbl,$i2,lsl#2]	@ Te2[s0>>8]
 	and	$i2,lr,$s1
+	ldr	$t3,[$tbl,$i3,lsl#2]	@ Te1[s0>>16]
 	and	$i3,lr,$s1,lsr#8
+	ldr	$s0,[$tbl,$s0,lsl#2]	@ Te0[s0>>24]
 	mov	$s1,$s1,lsr#24
+
 	ldr	$i1,[$tbl,$i1,lsl#2]	@ Te1[s1>>16]
-	ldr	$s1,[$tbl,$s1,lsl#2]	@ Te0[s1>>24]
 	ldr	$i2,[$tbl,$i2,lsl#2]	@ Te3[s1>>0]
 	ldr	$i3,[$tbl,$i3,lsl#2]	@ Te2[s1>>8]
 	eor	$s0,$s0,$i1,ror#8
-	eor	$s1,$s1,$t1,ror#24
-	eor	$t2,$t2,$i2,ror#8
-	eor	$t3,$t3,$i3,ror#8
-
+	ldr	$s1,[$tbl,$s1,lsl#2]	@ Te0[s1>>24]
 	and	$i1,lr,$s2,lsr#8	@ i0
+	eor	$t2,$t2,$i2,ror#8
 	and	$i2,lr,$s2,lsr#16	@ i1
+	eor	$t3,$t3,$i3,ror#8
 	and	$i3,lr,$s2
-	mov	$s2,$s2,lsr#24
+	eor	$s1,$s1,$t1,ror#24
 	ldr	$i1,[$tbl,$i1,lsl#2]	@ Te2[s2>>8]
+	mov	$s2,$s2,lsr#24
+
 	ldr	$i2,[$tbl,$i2,lsl#2]	@ Te1[s2>>16]
-	ldr	$s2,[$tbl,$s2,lsl#2]	@ Te0[s2>>24]
 	ldr	$i3,[$tbl,$i3,lsl#2]	@ Te3[s2>>0]
 	eor	$s0,$s0,$i1,ror#16
-	eor	$s1,$s1,$i2,ror#8
-	eor	$s2,$s2,$t2,ror#16
-	eor	$t3,$t3,$i3,ror#16
-
+	ldr	$s2,[$tbl,$s2,lsl#2]	@ Te0[s2>>24]
 	and	$i1,lr,$s3		@ i0
+	eor	$s1,$s1,$i2,ror#8
 	and	$i2,lr,$s3,lsr#8	@ i1
+	eor	$t3,$t3,$i3,ror#16
 	and	$i3,lr,$s3,lsr#16	@ i2
-	mov	$s3,$s3,lsr#24
+	eor	$s2,$s2,$t2,ror#16
 	ldr	$i1,[$tbl,$i1,lsl#2]	@ Te3[s3>>0]
+	mov	$s3,$s3,lsr#24
+
 	ldr	$i2,[$tbl,$i2,lsl#2]	@ Te2[s3>>8]
 	ldr	$i3,[$tbl,$i3,lsl#2]	@ Te1[s3>>16]
-	ldr	$s3,[$tbl,$s3,lsl#2]	@ Te0[s3>>24]
 	eor	$s0,$s0,$i1,ror#24
+	ldr	$s3,[$tbl,$s3,lsl#2]	@ Te0[s3>>24]
 	eor	$s1,$s1,$i2,ror#16
+	ldr	$i1,[$key],#16
 	eor	$s2,$s2,$i3,ror#8
+	ldr	$t1,[$key,#-12]
 	eor	$s3,$s3,$t3,ror#8
 
-	ldr	$t1,[$key],#16
-	ldr	$t2,[$key,#-12]
-	ldr	$t3,[$key,#-8]
-	ldr	$i1,[$key,#-4]
-	eor	$s0,$s0,$t1
-	eor	$s1,$s1,$t2
-	eor	$s2,$s2,$t3
-	eor	$s3,$s3,$i1
+	ldr	$t2,[$key,#-8]
+	eor	$s0,$s0,$i1
+	ldr	$t3,[$key,#-4]
+	and	$i1,lr,$s0
+	eor	$s1,$s1,$t1
+	and	$i2,lr,$s0,lsr#8
+	eor	$s2,$s2,$t2
+	and	$i3,lr,$s0,lsr#16
+	eor	$s3,$s3,$t3
+	mov	$s0,$s0,lsr#24
 
 	subs	$rounds,$rounds,#1
 	bne	.Lenc_loop
 
 	add	$tbl,$tbl,#2
 
-	and	$i1,lr,$s0
-	and	$i2,lr,$s0,lsr#8
-	and	$i3,lr,$s0,lsr#16
-	mov	$s0,$s0,lsr#24
 	ldrb	$t1,[$tbl,$i1,lsl#2]	@ Te4[s0>>0]
-	ldrb	$s0,[$tbl,$s0,lsl#2]	@ Te4[s0>>24]
-	ldrb	$t2,[$tbl,$i2,lsl#2]	@ Te4[s0>>8]
-	ldrb	$t3,[$tbl,$i3,lsl#2]	@ Te4[s0>>16]
-
 	and	$i1,lr,$s1,lsr#16	@ i0
+	ldrb	$t2,[$tbl,$i2,lsl#2]	@ Te4[s0>>8]
 	and	$i2,lr,$s1
+	ldrb	$t3,[$tbl,$i3,lsl#2]	@ Te4[s0>>16]
 	and	$i3,lr,$s1,lsr#8
+	ldrb	$s0,[$tbl,$s0,lsl#2]	@ Te4[s0>>24]
 	mov	$s1,$s1,lsr#24
+
 	ldrb	$i1,[$tbl,$i1,lsl#2]	@ Te4[s1>>16]
-	ldrb	$s1,[$tbl,$s1,lsl#2]	@ Te4[s1>>24]
 	ldrb	$i2,[$tbl,$i2,lsl#2]	@ Te4[s1>>0]
 	ldrb	$i3,[$tbl,$i3,lsl#2]	@ Te4[s1>>8]
 	eor	$s0,$i1,$s0,lsl#8
-	eor	$s1,$t1,$s1,lsl#24
-	eor	$t2,$i2,$t2,lsl#8
-	eor	$t3,$i3,$t3,lsl#8
-
+	ldrb	$s1,[$tbl,$s1,lsl#2]	@ Te4[s1>>24]
 	and	$i1,lr,$s2,lsr#8	@ i0
+	eor	$t2,$i2,$t2,lsl#8
 	and	$i2,lr,$s2,lsr#16	@ i1
+	eor	$t3,$i3,$t3,lsl#8
 	and	$i3,lr,$s2
-	mov	$s2,$s2,lsr#24
+	eor	$s1,$t1,$s1,lsl#24
 	ldrb	$i1,[$tbl,$i1,lsl#2]	@ Te4[s2>>8]
+	mov	$s2,$s2,lsr#24
+
 	ldrb	$i2,[$tbl,$i2,lsl#2]	@ Te4[s2>>16]
-	ldrb	$s2,[$tbl,$s2,lsl#2]	@ Te4[s2>>24]
 	ldrb	$i3,[$tbl,$i3,lsl#2]	@ Te4[s2>>0]
 	eor	$s0,$i1,$s0,lsl#8
-	eor	$s1,$s1,$i2,lsl#16
-	eor	$s2,$t2,$s2,lsl#24
-	eor	$t3,$i3,$t3,lsl#8
-
+	ldrb	$s2,[$tbl,$s2,lsl#2]	@ Te4[s2>>24]
 	and	$i1,lr,$s3		@ i0
+	eor	$s1,$s1,$i2,lsl#16
 	and	$i2,lr,$s3,lsr#8	@ i1
+	eor	$t3,$i3,$t3,lsl#8
 	and	$i3,lr,$s3,lsr#16	@ i2
-	mov	$s3,$s3,lsr#24
+	eor	$s2,$t2,$s2,lsl#24
 	ldrb	$i1,[$tbl,$i1,lsl#2]	@ Te4[s3>>0]
+	mov	$s3,$s3,lsr#24
+
 	ldrb	$i2,[$tbl,$i2,lsl#2]	@ Te4[s3>>8]
 	ldrb	$i3,[$tbl,$i3,lsl#2]	@ Te4[s3>>16]
-	ldrb	$s3,[$tbl,$s3,lsl#2]	@ Te4[s3>>24]
 	eor	$s0,$i1,$s0,lsl#8
+	ldrb	$s3,[$tbl,$s3,lsl#2]	@ Te4[s3>>24]
+	ldr	$i1,[$key,#0]
 	eor	$s1,$s1,$i2,lsl#8
+	ldr	$t1,[$key,#4]
 	eor	$s2,$s2,$i3,lsl#16
+	ldr	$t2,[$key,#8]
 	eor	$s3,$t3,$s3,lsl#24
+	ldr	$t3,[$key,#12]
 
-	ldr	lr,[sp],#4		@ pop lr
-	ldr	$t1,[$key,#0]
-	ldr	$t2,[$key,#4]
-	ldr	$t3,[$key,#8]
-	ldr	$i1,[$key,#12]
-	eor	$s0,$s0,$t1
-	eor	$s1,$s1,$t2
-	eor	$s2,$s2,$t3
-	eor	$s3,$s3,$i1
+	eor	$s0,$s0,$i1
+	eor	$s1,$s1,$t1
+	eor	$s2,$s2,$t2
+	eor	$s3,$s3,$t3
 
 	sub	$tbl,$tbl,#2
-	mov	pc,lr			@ return
+	ldr	pc,[sp],#4		@ pop and return
 .size	_armv4_AES_encrypt,.-_armv4_AES_encrypt
 
 .global AES_set_encrypt_key
@@ -399,31 +403,31 @@ AES_set_encrypt_key:
 	ldrb	$t2,[$rounds,#1]
 	ldrb	$t3,[$rounds,#0]
 	orr	$s0,$s0,$t1,lsl#8
-	orr	$s0,$s0,$t2,lsl#16
-	orr	$s0,$s0,$t3,lsl#24
 	ldrb	$s1,[$rounds,#7]
+	orr	$s0,$s0,$t2,lsl#16
 	ldrb	$t1,[$rounds,#6]
+	orr	$s0,$s0,$t3,lsl#24
 	ldrb	$t2,[$rounds,#5]
 	ldrb	$t3,[$rounds,#4]
 	orr	$s1,$s1,$t1,lsl#8
-	orr	$s1,$s1,$t2,lsl#16
-	orr	$s1,$s1,$t3,lsl#24
 	ldrb	$s2,[$rounds,#11]
+	orr	$s1,$s1,$t2,lsl#16
 	ldrb	$t1,[$rounds,#10]
+	orr	$s1,$s1,$t3,lsl#24
 	ldrb	$t2,[$rounds,#9]
 	ldrb	$t3,[$rounds,#8]
 	orr	$s2,$s2,$t1,lsl#8
-	orr	$s2,$s2,$t2,lsl#16
-	orr	$s2,$s2,$t3,lsl#24
 	ldrb	$s3,[$rounds,#15]
+	orr	$s2,$s2,$t2,lsl#16
 	ldrb	$t1,[$rounds,#14]
+	orr	$s2,$s2,$t3,lsl#24
 	ldrb	$t2,[$rounds,#13]
 	ldrb	$t3,[$rounds,#12]
 	orr	$s3,$s3,$t1,lsl#8
-	orr	$s3,$s3,$t2,lsl#16
-	orr	$s3,$s3,$t3,lsl#24
 	str	$s0,[$key],#16
+	orr	$s3,$s3,$t2,lsl#16
 	str	$s1,[$key,#-12]
+	orr	$s3,$s3,$t3,lsl#24
 	str	$s2,[$key,#-8]
 	str	$s3,[$key,#-4]
 
@@ -437,27 +441,26 @@ AES_set_encrypt_key:
 .L128_loop:
 	and	$t2,lr,$s3,lsr#24
 	and	$i1,lr,$s3,lsr#16
-	and	$i2,lr,$s3,lsr#8
-	and	$i3,lr,$s3
 	ldrb	$t2,[$tbl,$t2]
+	and	$i2,lr,$s3,lsr#8
 	ldrb	$i1,[$tbl,$i1]
+	and	$i3,lr,$s3
 	ldrb	$i2,[$tbl,$i2]
-	ldrb	$i3,[$tbl,$i3]
-	ldr	$t1,[$t3],#4			@ rcon[i++]
 	orr	$t2,$t2,$i1,lsl#24
+	ldrb	$i3,[$tbl,$i3]
 	orr	$t2,$t2,$i2,lsl#16
+	ldr	$t1,[$t3],#4			@ rcon[i++]
 	orr	$t2,$t2,$i3,lsl#8
 	eor	$t2,$t2,$t1
 	eor	$s0,$s0,$t2			@ rk[4]=rk[0]^...
 	eor	$s1,$s1,$s0			@ rk[5]=rk[1]^rk[4]
-	eor	$s2,$s2,$s1			@ rk[6]=rk[2]^rk[5]
-	eor	$s3,$s3,$s2			@ rk[7]=rk[3]^rk[6]
 	str	$s0,[$key],#16
+	eor	$s2,$s2,$s1			@ rk[6]=rk[2]^rk[5]
 	str	$s1,[$key,#-12]
+	eor	$s3,$s3,$s2			@ rk[7]=rk[3]^rk[6]
 	str	$s2,[$key,#-8]
-	str	$s3,[$key,#-4]
-
 	subs	$rounds,$rounds,#1
+	str	$s3,[$key,#-4]
 	bne	.L128_loop
 	sub	r2,$key,#176
 	b	.Ldone
@@ -468,16 +471,16 @@ AES_set_encrypt_key:
 	ldrb	$t2,[$rounds,#17]
 	ldrb	$t3,[$rounds,#16]
 	orr	$i2,$i2,$t1,lsl#8
-	orr	$i2,$i2,$t2,lsl#16
-	orr	$i2,$i2,$t3,lsl#24
 	ldrb	$i3,[$rounds,#23]
+	orr	$i2,$i2,$t2,lsl#16
 	ldrb	$t1,[$rounds,#22]
+	orr	$i2,$i2,$t3,lsl#24
 	ldrb	$t2,[$rounds,#21]
 	ldrb	$t3,[$rounds,#20]
 	orr	$i3,$i3,$t1,lsl#8
 	orr	$i3,$i3,$t2,lsl#16
-	orr	$i3,$i3,$t3,lsl#24
 	str	$i2,[$key],#8
+	orr	$i3,$i3,$t3,lsl#24
 	str	$i3,[$key,#-4]
 
 	teq	lr,#192
@@ -491,27 +494,26 @@ AES_set_encrypt_key:
 .L192_loop:
 	and	$t2,lr,$i3,lsr#24
 	and	$i1,lr,$i3,lsr#16
-	and	$i2,lr,$i3,lsr#8
-	and	$i3,lr,$i3
 	ldrb	$t2,[$tbl,$t2]
+	and	$i2,lr,$i3,lsr#8
 	ldrb	$i1,[$tbl,$i1]
+	and	$i3,lr,$i3
 	ldrb	$i2,[$tbl,$i2]
-	ldrb	$i3,[$tbl,$i3]
-	ldr	$t1,[$t3],#4			@ rcon[i++]
 	orr	$t2,$t2,$i1,lsl#24
+	ldrb	$i3,[$tbl,$i3]
 	orr	$t2,$t2,$i2,lsl#16
+	ldr	$t1,[$t3],#4			@ rcon[i++]
 	orr	$t2,$t2,$i3,lsl#8
 	eor	$i3,$t2,$t1
 	eor	$s0,$s0,$i3			@ rk[6]=rk[0]^...
 	eor	$s1,$s1,$s0			@ rk[7]=rk[1]^rk[6]
-	eor	$s2,$s2,$s1			@ rk[8]=rk[2]^rk[7]
-	eor	$s3,$s3,$s2			@ rk[9]=rk[3]^rk[8]
 	str	$s0,[$key],#24
+	eor	$s2,$s2,$s1			@ rk[8]=rk[2]^rk[7]
 	str	$s1,[$key,#-20]
+	eor	$s3,$s3,$s2			@ rk[9]=rk[3]^rk[8]
 	str	$s2,[$key,#-16]
-	str	$s3,[$key,#-12]
-
 	subs	$rounds,$rounds,#1
+	str	$s3,[$key,#-12]
 	subeq	r2,$key,#216
 	beq	.Ldone
 
@@ -529,16 +531,16 @@ AES_set_encrypt_key:
 	ldrb	$t2,[$rounds,#25]
 	ldrb	$t3,[$rounds,#24]
 	orr	$i2,$i2,$t1,lsl#8
-	orr	$i2,$i2,$t2,lsl#16
-	orr	$i2,$i2,$t3,lsl#24
 	ldrb	$i3,[$rounds,#31]
+	orr	$i2,$i2,$t2,lsl#16
 	ldrb	$t1,[$rounds,#30]
+	orr	$i2,$i2,$t3,lsl#24
 	ldrb	$t2,[$rounds,#29]
 	ldrb	$t3,[$rounds,#28]
 	orr	$i3,$i3,$t1,lsl#8
 	orr	$i3,$i3,$t2,lsl#16
-	orr	$i3,$i3,$t3,lsl#24
 	str	$i2,[$key],#8
+	orr	$i3,$i3,$t3,lsl#24
 	str	$i3,[$key,#-4]
 
 	mov	$rounds,#14
@@ -550,52 +552,51 @@ AES_set_encrypt_key:
 .L256_loop:
 	and	$t2,lr,$i3,lsr#24
 	and	$i1,lr,$i3,lsr#16
-	and	$i2,lr,$i3,lsr#8
-	and	$i3,lr,$i3
 	ldrb	$t2,[$tbl,$t2]
+	and	$i2,lr,$i3,lsr#8
 	ldrb	$i1,[$tbl,$i1]
+	and	$i3,lr,$i3
 	ldrb	$i2,[$tbl,$i2]
-	ldrb	$i3,[$tbl,$i3]
-	ldr	$t1,[$t3],#4			@ rcon[i++]
 	orr	$t2,$t2,$i1,lsl#24
+	ldrb	$i3,[$tbl,$i3]
 	orr	$t2,$t2,$i2,lsl#16
+	ldr	$t1,[$t3],#4			@ rcon[i++]
 	orr	$t2,$t2,$i3,lsl#8
 	eor	$i3,$t2,$t1
 	eor	$s0,$s0,$i3			@ rk[8]=rk[0]^...
 	eor	$s1,$s1,$s0			@ rk[9]=rk[1]^rk[8]
-	eor	$s2,$s2,$s1			@ rk[10]=rk[2]^rk[9]
-	eor	$s3,$s3,$s2			@ rk[11]=rk[3]^rk[10]
 	str	$s0,[$key],#32
+	eor	$s2,$s2,$s1			@ rk[10]=rk[2]^rk[9]
 	str	$s1,[$key,#-28]
+	eor	$s3,$s3,$s2			@ rk[11]=rk[3]^rk[10]
 	str	$s2,[$key,#-24]
-	str	$s3,[$key,#-20]
-
 	subs	$rounds,$rounds,#1
+	str	$s3,[$key,#-20]
 	subeq	r2,$key,#256
 	beq	.Ldone
 
 	and	$t2,lr,$s3
 	and	$i1,lr,$s3,lsr#8
-	and	$i2,lr,$s3,lsr#16
-	and	$i3,lr,$s3,lsr#24
 	ldrb	$t2,[$tbl,$t2]
+	and	$i2,lr,$s3,lsr#16
 	ldrb	$i1,[$tbl,$i1]
+	and	$i3,lr,$s3,lsr#24
 	ldrb	$i2,[$tbl,$i2]
-	ldrb	$i3,[$tbl,$i3]
 	orr	$t2,$t2,$i1,lsl#8
+	ldrb	$i3,[$tbl,$i3]
 	orr	$t2,$t2,$i2,lsl#16
+	ldr	$t1,[$key,#-48]
 	orr	$t2,$t2,$i3,lsl#24
 
-	ldr	$t1,[$key,#-48]
 	ldr	$i1,[$key,#-44]
 	ldr	$i2,[$key,#-40]
-	ldr	$i3,[$key,#-36]
 	eor	$t1,$t1,$t2			@ rk[12]=rk[4]^...
+	ldr	$i3,[$key,#-36]
 	eor	$i1,$i1,$t1			@ rk[13]=rk[5]^rk[12]
-	eor	$i2,$i2,$i1			@ rk[14]=rk[6]^rk[13]
-	eor	$i3,$i3,$i2			@ rk[15]=rk[7]^rk[14]
 	str	$t1,[$key,#-16]
+	eor	$i2,$i2,$i1			@ rk[14]=rk[6]^rk[13]
 	str	$i1,[$key,#-12]
+	eor	$i3,$i3,$i2			@ rk[15]=rk[7]^rk[14]
 	str	$i2,[$key,#-8]
 	str	$i3,[$key,#-4]
 	b	.L256_loop
@@ -816,24 +817,24 @@ AES_decrypt:
 	ldrb	$t2,[$rounds,#1]
 	ldrb	$t3,[$rounds,#0]
 	orr	$s0,$s0,$t1,lsl#8
-	orr	$s0,$s0,$t2,lsl#16
-	orr	$s0,$s0,$t3,lsl#24
 	ldrb	$s1,[$rounds,#7]
+	orr	$s0,$s0,$t2,lsl#16
 	ldrb	$t1,[$rounds,#6]
+	orr	$s0,$s0,$t3,lsl#24
 	ldrb	$t2,[$rounds,#5]
 	ldrb	$t3,[$rounds,#4]
 	orr	$s1,$s1,$t1,lsl#8
-	orr	$s1,$s1,$t2,lsl#16
-	orr	$s1,$s1,$t3,lsl#24
 	ldrb	$s2,[$rounds,#11]
+	orr	$s1,$s1,$t2,lsl#16
 	ldrb	$t1,[$rounds,#10]
+	orr	$s1,$s1,$t3,lsl#24
 	ldrb	$t2,[$rounds,#9]
 	ldrb	$t3,[$rounds,#8]
 	orr	$s2,$s2,$t1,lsl#8
-	orr	$s2,$s2,$t2,lsl#16
-	orr	$s2,$s2,$t3,lsl#24
 	ldrb	$s3,[$rounds,#15]
+	orr	$s2,$s2,$t2,lsl#16
 	ldrb	$t1,[$rounds,#14]
+	orr	$s2,$s2,$t3,lsl#24
 	ldrb	$t2,[$rounds,#13]
 	ldrb	$t3,[$rounds,#12]
 	orr	$s3,$s3,$t1,lsl#8
@@ -848,24 +849,24 @@ AES_decrypt:
 	mov	$t3,$s0,lsr#8
 	strb	$t1,[$rounds,#0]
 	strb	$t2,[$rounds,#1]
-	strb	$t3,[$rounds,#2]
-	strb	$s0,[$rounds,#3]
 	mov	$t1,$s1,lsr#24
+	strb	$t3,[$rounds,#2]
 	mov	$t2,$s1,lsr#16
+	strb	$s0,[$rounds,#3]
 	mov	$t3,$s1,lsr#8
 	strb	$t1,[$rounds,#4]
 	strb	$t2,[$rounds,#5]
-	strb	$t3,[$rounds,#6]
-	strb	$s1,[$rounds,#7]
 	mov	$t1,$s2,lsr#24
+	strb	$t3,[$rounds,#6]
 	mov	$t2,$s2,lsr#16
+	strb	$s1,[$rounds,#7]
 	mov	$t3,$s2,lsr#8
 	strb	$t1,[$rounds,#8]
 	strb	$t2,[$rounds,#9]
-	strb	$t3,[$rounds,#10]
-	strb	$s2,[$rounds,#11]
 	mov	$t1,$s3,lsr#24
+	strb	$t3,[$rounds,#10]
 	mov	$t2,$s3,lsr#16
+	strb	$s2,[$rounds,#11]
 	mov	$t3,$s3,lsr#8
 	strb	$t1,[$rounds,#12]
 	strb	$t2,[$rounds,#13]
@@ -882,146 +883,143 @@ AES_decrypt:
 .align	2
 _armv4_AES_decrypt:
 	str	lr,[sp,#-4]!		@ push lr
-	ldr	$t1,[$key],#16
-	ldr	$t2,[$key,#-12]
-	ldr	$t3,[$key,#-8]
-	ldr	$i1,[$key,#-4]
-	ldr	$rounds,[$key,#240-16]
+	ldmia	$key!,{$t1-$i1}
 	eor	$s0,$s0,$t1
+	ldr	$rounds,[$key,#240-16]
 	eor	$s1,$s1,$t2
 	eor	$s2,$s2,$t3
 	eor	$s3,$s3,$i1
 	sub	$rounds,$rounds,#1
 	mov	lr,#255
 
-.Ldec_loop:
 	and	$i1,lr,$s0,lsr#16
 	and	$i2,lr,$s0,lsr#8
 	and	$i3,lr,$s0
 	mov	$s0,$s0,lsr#24
+.Ldec_loop:
 	ldr	$t1,[$tbl,$i1,lsl#2]	@ Td1[s0>>16]
-	ldr	$s0,[$tbl,$s0,lsl#2]	@ Td0[s0>>24]
-	ldr	$t2,[$tbl,$i2,lsl#2]	@ Td2[s0>>8]
-	ldr	$t3,[$tbl,$i3,lsl#2]	@ Td3[s0>>0]
-
 	and	$i1,lr,$s1		@ i0
+	ldr	$t2,[$tbl,$i2,lsl#2]	@ Td2[s0>>8]
 	and	$i2,lr,$s1,lsr#16
+	ldr	$t3,[$tbl,$i3,lsl#2]	@ Td3[s0>>0]
 	and	$i3,lr,$s1,lsr#8
+	ldr	$s0,[$tbl,$s0,lsl#2]	@ Td0[s0>>24]
 	mov	$s1,$s1,lsr#24
+
 	ldr	$i1,[$tbl,$i1,lsl#2]	@ Td3[s1>>0]
-	ldr	$s1,[$tbl,$s1,lsl#2]	@ Td0[s1>>24]
 	ldr	$i2,[$tbl,$i2,lsl#2]	@ Td1[s1>>16]
 	ldr	$i3,[$tbl,$i3,lsl#2]	@ Td2[s1>>8]
 	eor	$s0,$s0,$i1,ror#24
-	eor	$s1,$s1,$t1,ror#8
-	eor	$t2,$i2,$t2,ror#8
-	eor	$t3,$i3,$t3,ror#8
-
+	ldr	$s1,[$tbl,$s1,lsl#2]	@ Td0[s1>>24]
 	and	$i1,lr,$s2,lsr#8	@ i0
+	eor	$t2,$i2,$t2,ror#8
 	and	$i2,lr,$s2		@ i1
+	eor	$t3,$i3,$t3,ror#8
 	and	$i3,lr,$s2,lsr#16
-	mov	$s2,$s2,lsr#24
+	eor	$s1,$s1,$t1,ror#8
 	ldr	$i1,[$tbl,$i1,lsl#2]	@ Td2[s2>>8]
+	mov	$s2,$s2,lsr#24
+
 	ldr	$i2,[$tbl,$i2,lsl#2]	@ Td3[s2>>0]
-	ldr	$s2,[$tbl,$s2,lsl#2]	@ Td0[s2>>24]
 	ldr	$i3,[$tbl,$i3,lsl#2]	@ Td1[s2>>16]
 	eor	$s0,$s0,$i1,ror#16
-	eor	$s1,$s1,$i2,ror#24
-	eor	$s2,$s2,$t2,ror#8
-	eor	$t3,$i3,$t3,ror#8
-
+	ldr	$s2,[$tbl,$s2,lsl#2]	@ Td0[s2>>24]
 	and	$i1,lr,$s3,lsr#16	@ i0
+	eor	$s1,$s1,$i2,ror#24
 	and	$i2,lr,$s3,lsr#8	@ i1
+	eor	$t3,$i3,$t3,ror#8
 	and	$i3,lr,$s3		@ i2
-	mov	$s3,$s3,lsr#24
+	eor	$s2,$s2,$t2,ror#8
 	ldr	$i1,[$tbl,$i1,lsl#2]	@ Td1[s3>>16]
+	mov	$s3,$s3,lsr#24
+
 	ldr	$i2,[$tbl,$i2,lsl#2]	@ Td2[s3>>8]
 	ldr	$i3,[$tbl,$i3,lsl#2]	@ Td3[s3>>0]
-	ldr	$s3,[$tbl,$s3,lsl#2]	@ Td0[s3>>24]
 	eor	$s0,$s0,$i1,ror#8
+	ldr	$s3,[$tbl,$s3,lsl#2]	@ Td0[s3>>24]
 	eor	$s1,$s1,$i2,ror#16
 	eor	$s2,$s2,$i3,ror#24
+	ldr	$i1,[$key],#16
 	eor	$s3,$s3,$t3,ror#8
 
-	ldr	$t1,[$key],#16
-	ldr	$t2,[$key,#-12]
-	ldr	$t3,[$key,#-8]
-	ldr	$i1,[$key,#-4]
-	eor	$s0,$s0,$t1
-	eor	$s1,$s1,$t2
-	eor	$s2,$s2,$t3
-	eor	$s3,$s3,$i1
+	ldr	$t1,[$key,#-12]
+	ldr	$t2,[$key,#-8]
+	eor	$s0,$s0,$i1
+	ldr	$t3,[$key,#-4]
+	and	$i1,lr,$s0,lsr#16
+	eor	$s1,$s1,$t1
+	and	$i2,lr,$s0,lsr#8
+	eor	$s2,$s2,$t2
+	and	$i3,lr,$s0
+	eor	$s3,$s3,$t3
+	mov	$s0,$s0,lsr#24
 
 	subs	$rounds,$rounds,#1
 	bne	.Ldec_loop
 
 	add	$tbl,$tbl,#1024
 
-	ldr	$t1,[$tbl,#0]		@ prefetch Td4
-	ldr	$t2,[$tbl,#32]
-	ldr	$t3,[$tbl,#64]
-	ldr	$i1,[$tbl,#96]
-	ldr	$i2,[$tbl,#128]
-	ldr	$i3,[$tbl,#160]
-	ldr	$t1,[$tbl,#192]
-	ldr	$t2,[$tbl,#224]
+	ldr	$t2,[$tbl,#0]		@ prefetch Td4
+	ldr	$t3,[$tbl,#32]
+	ldr	$t1,[$tbl,#64]
+	ldr	$t2,[$tbl,#96]
+	ldr	$t3,[$tbl,#128]
+	ldr	$t1,[$tbl,#160]
+	ldr	$t2,[$tbl,#192]
+	ldr	$t3,[$tbl,#224]
 
-	and	$i1,lr,$s0,lsr#16
-	and	$i2,lr,$s0,lsr#8
-	and	$i3,lr,$s0
-	ldrb	$s0,[$tbl,$s0,lsr#24]	@ Td4[s0>>24]
+	ldrb	$s0,[$tbl,$s0]		@ Td4[s0>>24]
 	ldrb	$t1,[$tbl,$i1]		@ Td4[s0>>16]
-	ldrb	$t2,[$tbl,$i2]		@ Td4[s0>>8]
-	ldrb	$t3,[$tbl,$i3]		@ Td4[s0>>0]
-
 	and	$i1,lr,$s1		@ i0
+	ldrb	$t2,[$tbl,$i2]		@ Td4[s0>>8]
 	and	$i2,lr,$s1,lsr#16
+	ldrb	$t3,[$tbl,$i3]		@ Td4[s0>>0]
 	and	$i3,lr,$s1,lsr#8
+
 	ldrb	$i1,[$tbl,$i1]		@ Td4[s1>>0]
 	ldrb	$s1,[$tbl,$s1,lsr#24]	@ Td4[s1>>24]
 	ldrb	$i2,[$tbl,$i2]		@ Td4[s1>>16]
-	ldrb	$i3,[$tbl,$i3]		@ Td4[s1>>8]
 	eor	$s0,$i1,$s0,lsl#24
+	ldrb	$i3,[$tbl,$i3]		@ Td4[s1>>8]
 	eor	$s1,$t1,$s1,lsl#8
-	eor	$t2,$t2,$i2,lsl#8
-	eor	$t3,$t3,$i3,lsl#8
-
 	and	$i1,lr,$s2,lsr#8	@ i0
+	eor	$t2,$t2,$i2,lsl#8
 	and	$i2,lr,$s2		@ i1
-	and	$i3,lr,$s2,lsr#16
+	eor	$t3,$t3,$i3,lsl#8
 	ldrb	$i1,[$tbl,$i1]		@ Td4[s2>>8]
+	and	$i3,lr,$s2,lsr#16
+
 	ldrb	$i2,[$tbl,$i2]		@ Td4[s2>>0]
 	ldrb	$s2,[$tbl,$s2,lsr#24]	@ Td4[s2>>24]
-	ldrb	$i3,[$tbl,$i3]		@ Td4[s2>>16]
 	eor	$s0,$s0,$i1,lsl#8
+	ldrb	$i3,[$tbl,$i3]		@ Td4[s2>>16]
 	eor	$s1,$i2,$s1,lsl#16
-	eor	$s2,$t2,$s2,lsl#16
-	eor	$t3,$t3,$i3,lsl#16
-
 	and	$i1,lr,$s3,lsr#16	@ i0
+	eor	$s2,$t2,$s2,lsl#16
 	and	$i2,lr,$s3,lsr#8	@ i1
-	and	$i3,lr,$s3		@ i2
+	eor	$t3,$t3,$i3,lsl#16
 	ldrb	$i1,[$tbl,$i1]		@ Td4[s3>>16]
+	and	$i3,lr,$s3		@ i2
+
 	ldrb	$i2,[$tbl,$i2]		@ Td4[s3>>8]
 	ldrb	$i3,[$tbl,$i3]		@ Td4[s3>>0]
 	ldrb	$s3,[$tbl,$s3,lsr#24]	@ Td4[s3>>24]
 	eor	$s0,$s0,$i1,lsl#16
+	ldr	$i1,[$key,#0]
 	eor	$s1,$s1,$i2,lsl#8
+	ldr	$t1,[$key,#4]
 	eor	$s2,$i3,$s2,lsl#8
+	ldr	$t2,[$key,#8]
 	eor	$s3,$t3,$s3,lsl#24
+	ldr	$t3,[$key,#12]
 
-	ldr	lr,[sp],#4		@ pop lr
-	ldr	$t1,[$key,#0]
-	ldr	$t2,[$key,#4]
-	ldr	$t3,[$key,#8]
-	ldr	$i1,[$key,#12]
-	eor	$s0,$s0,$t1
-	eor	$s1,$s1,$t2
-	eor	$s2,$s2,$t3
-	eor	$s3,$s3,$i1
+	eor	$s0,$s0,$i1
+	eor	$s1,$s1,$t1
+	eor	$s2,$s2,$t2
+	eor	$s3,$s3,$t3
 
 	sub	$tbl,$tbl,#1024
-	mov	pc,lr			@ return
+	ldr	pc,[sp],#4		@ pop and return
 .size	_armv4_AES_decrypt,.-_armv4_AES_decrypt
 .asciz	"AES for ARMv4, CRYPTOGAMS by <appro\@openssl.org>"
 .align	2
@@ -1029,3 +1027,4 @@ ___
 
 $code =~ s/\bbx\s+lr\b/.word\t0xe12fff1e/gm;	# make it possible to compile with -march=armv4
 print $code;
+close STDOUT;	# enforce flush

crypto/aes/asm/aes-ppc.pl

@@ -16,6 +16,19 @@
 # at 1/2 of ppc_AES_encrypt speed, while ppc_AES_decrypt_compact -
 # at 1/3 of ppc_AES_decrypt.
 
+# February 2010
+#
+# Rescheduling instructions to favour Power6 pipeline gives 10%
+# performance improvement on the platfrom in question (and marginal
+# improvement even on others). It should be noted that Power6 fails
+# to process byte in 18 cycles, only in 23, because it fails to issue
+# 4 load instructions in two cycles, only in 3. As result non-compact
+# block subroutines are 25% slower than one would expect. Compact
+# functions scale better, because they have pure computational part,
+# which scales perfectly with clock frequency. To be specific
+# ppc_AES_encrypt_compact operates at 42 cycles per byte, while
+# ppc_AES_decrypt_compact - at 55 (in 64-bit build).
+
 $flavour = shift;
 
 if ($flavour =~ /64/) {
@@ -376,7 +389,7 @@ $code.=<<___;
 	addi	$sp,$sp,$FRAME
 	blr
 
-.align	4
+.align	5
 Lppc_AES_encrypt:
 	lwz	$acc00,240($key)
 	lwz	$t0,0($key)
@@ -397,46 +410,46 @@ Lppc_AES_encrypt:
 Lenc_loop:
 	rlwinm	$acc00,$s0,`32-24+3`,21,28
 	rlwinm	$acc01,$s1,`32-24+3`,21,28
-	lwz	$t0,0($key)
-	lwz	$t1,4($key)
 	rlwinm	$acc02,$s2,`32-24+3`,21,28
 	rlwinm	$acc03,$s3,`32-24+3`,21,28
-	lwz	$t2,8($key)
-	lwz	$t3,12($key)
+	lwz	$t0,0($key)
+	lwz	$t1,4($key)
 	rlwinm	$acc04,$s1,`32-16+3`,21,28
 	rlwinm	$acc05,$s2,`32-16+3`,21,28
-	lwzx	$acc00,$Tbl0,$acc00
-	lwzx	$acc01,$Tbl0,$acc01
+	lwz	$t2,8($key)
+	lwz	$t3,12($key)
 	rlwinm	$acc06,$s3,`32-16+3`,21,28
 	rlwinm	$acc07,$s0,`32-16+3`,21,28
-	lwzx	$acc02,$Tbl0,$acc02
-	lwzx	$acc03,$Tbl0,$acc03
+	lwzx	$acc00,$Tbl0,$acc00
+	lwzx	$acc01,$Tbl0,$acc01
 	rlwinm	$acc08,$s2,`32-8+3`,21,28
 	rlwinm	$acc09,$s3,`32-8+3`,21,28
-	lwzx	$acc04,$Tbl1,$acc04
-	lwzx	$acc05,$Tbl1,$acc05
+	lwzx	$acc02,$Tbl0,$acc02
+	lwzx	$acc03,$Tbl0,$acc03
 	rlwinm	$acc10,$s0,`32-8+3`,21,28
 	rlwinm	$acc11,$s1,`32-8+3`,21,28
-	lwzx	$acc06,$Tbl1,$acc06
-	lwzx	$acc07,$Tbl1,$acc07
+	lwzx	$acc04,$Tbl1,$acc04
+	lwzx	$acc05,$Tbl1,$acc05
 	rlwinm	$acc12,$s3,`0+3`,21,28
 	rlwinm	$acc13,$s0,`0+3`,21,28
-	lwzx	$acc08,$Tbl2,$acc08
-	lwzx	$acc09,$Tbl2,$acc09
+	lwzx	$acc06,$Tbl1,$acc06
+	lwzx	$acc07,$Tbl1,$acc07
 	rlwinm	$acc14,$s1,`0+3`,21,28
 	rlwinm	$acc15,$s2,`0+3`,21,28
-	lwzx	$acc10,$Tbl2,$acc10
-	lwzx	$acc11,$Tbl2,$acc11
+	lwzx	$acc08,$Tbl2,$acc08
+	lwzx	$acc09,$Tbl2,$acc09
 	xor	$t0,$t0,$acc00
 	xor	$t1,$t1,$acc01
-	lwzx	$acc12,$Tbl3,$acc12
-	lwzx	$acc13,$Tbl3,$acc13
+	lwzx	$acc10,$Tbl2,$acc10
+	lwzx	$acc11,$Tbl2,$acc11
 	xor	$t2,$t2,$acc02
 	xor	$t3,$t3,$acc03
-	lwzx	$acc14,$Tbl3,$acc14
-	lwzx	$acc15,$Tbl3,$acc15
+	lwzx	$acc12,$Tbl3,$acc12
+	lwzx	$acc13,$Tbl3,$acc13
 	xor	$t0,$t0,$acc04
 	xor	$t1,$t1,$acc05
+	lwzx	$acc14,$Tbl3,$acc14
+	lwzx	$acc15,$Tbl3,$acc15
 	xor	$t2,$t2,$acc06
 	xor	$t3,$t3,$acc07
 	xor	$t0,$t0,$acc08
@@ -452,60 +465,60 @@ Lenc_loop:
 
 	addi	$Tbl2,$Tbl0,2048
 	nop
-	lwz	$acc08,`2048+0`($Tbl0)	! prefetch Te4
-	lwz	$acc09,`2048+32`($Tbl0)
-	lwz	$acc10,`2048+64`($Tbl0)
-	lwz	$acc11,`2048+96`($Tbl0)
-	lwz	$acc08,`2048+128`($Tbl0)
-	lwz	$acc09,`2048+160`($Tbl0)
-	lwz	$acc10,`2048+192`($Tbl0)
-	lwz	$acc11,`2048+224`($Tbl0)
-	rlwinm	$acc00,$s0,`32-24`,24,31
-	rlwinm	$acc01,$s1,`32-24`,24,31
 	lwz	$t0,0($key)
 	lwz	$t1,4($key)
-	rlwinm	$acc02,$s2,`32-24`,24,31
-	rlwinm	$acc03,$s3,`32-24`,24,31
+	rlwinm	$acc00,$s0,`32-24`,24,31
+	rlwinm	$acc01,$s1,`32-24`,24,31
 	lwz	$t2,8($key)
 	lwz	$t3,12($key)
+	rlwinm	$acc02,$s2,`32-24`,24,31
+	rlwinm	$acc03,$s3,`32-24`,24,31
+	lwz	$acc08,`2048+0`($Tbl0)	! prefetch Te4
+	lwz	$acc09,`2048+32`($Tbl0)
 	rlwinm	$acc04,$s1,`32-16`,24,31
 	rlwinm	$acc05,$s2,`32-16`,24,31
-	lbzx	$acc00,$Tbl2,$acc00
-	lbzx	$acc01,$Tbl2,$acc01
+	lwz	$acc10,`2048+64`($Tbl0)
+	lwz	$acc11,`2048+96`($Tbl0)
 	rlwinm	$acc06,$s3,`32-16`,24,31
 	rlwinm	$acc07,$s0,`32-16`,24,31
-	lbzx	$acc02,$Tbl2,$acc02
-	lbzx	$acc03,$Tbl2,$acc03
+	lwz	$acc12,`2048+128`($Tbl0)
+	lwz	$acc13,`2048+160`($Tbl0)
 	rlwinm	$acc08,$s2,`32-8`,24,31
 	rlwinm	$acc09,$s3,`32-8`,24,31
-	lbzx	$acc04,$Tbl2,$acc04
-	lbzx	$acc05,$Tbl2,$acc05
+	lwz	$acc14,`2048+192`($Tbl0)
+	lwz	$acc15,`2048+224`($Tbl0)
 	rlwinm	$acc10,$s0,`32-8`,24,31
 	rlwinm	$acc11,$s1,`32-8`,24,31
-	lbzx	$acc06,$Tbl2,$acc06
-	lbzx	$acc07,$Tbl2,$acc07
+	lbzx	$acc00,$Tbl2,$acc00
+	lbzx	$acc01,$Tbl2,$acc01
 	rlwinm	$acc12,$s3,`0`,24,31
 	rlwinm	$acc13,$s0,`0`,24,31
-	lbzx	$acc08,$Tbl2,$acc08
-	lbzx	$acc09,$Tbl2,$acc09
+	lbzx	$acc02,$Tbl2,$acc02
+	lbzx	$acc03,$Tbl2,$acc03
 	rlwinm	$acc14,$s1,`0`,24,31
 	rlwinm	$acc15,$s2,`0`,24,31
-	lbzx	$acc10,$Tbl2,$acc10
-	lbzx	$acc11,$Tbl2,$acc11
+	lbzx	$acc04,$Tbl2,$acc04
+	lbzx	$acc05,$Tbl2,$acc05
 	rlwinm	$s0,$acc00,24,0,7
 	rlwinm	$s1,$acc01,24,0,7
-	lbzx	$acc12,$Tbl2,$acc12
-	lbzx	$acc13,$Tbl2,$acc13
+	lbzx	$acc06,$Tbl2,$acc06
+	lbzx	$acc07,$Tbl2,$acc07
 	rlwinm	$s2,$acc02,24,0,7
 	rlwinm	$s3,$acc03,24,0,7
-	lbzx	$acc14,$Tbl2,$acc14
-	lbzx	$acc15,$Tbl2,$acc15
+	lbzx	$acc08,$Tbl2,$acc08
+	lbzx	$acc09,$Tbl2,$acc09
 	rlwimi	$s0,$acc04,16,8,15
 	rlwimi	$s1,$acc05,16,8,15
+	lbzx	$acc10,$Tbl2,$acc10
+	lbzx	$acc11,$Tbl2,$acc11
 	rlwimi	$s2,$acc06,16,8,15
 	rlwimi	$s3,$acc07,16,8,15
+	lbzx	$acc12,$Tbl2,$acc12
+	lbzx	$acc13,$Tbl2,$acc13
 	rlwimi	$s0,$acc08,8,16,23
 	rlwimi	$s1,$acc09,8,16,23
+	lbzx	$acc14,$Tbl2,$acc14
+	lbzx	$acc15,$Tbl2,$acc15
 	rlwimi	$s2,$acc10,8,16,23
 	rlwimi	$s3,$acc11,8,16,23
 	or	$s0,$s0,$acc12
@@ -542,40 +555,40 @@ Lenc_compact_loop:
 	rlwinm	$acc01,$s1,`32-24`,24,31
 	rlwinm	$acc02,$s2,`32-24`,24,31
 	rlwinm	$acc03,$s3,`32-24`,24,31
-	lbzx	$acc00,$Tbl1,$acc00
-	lbzx	$acc01,$Tbl1,$acc01
 	rlwinm	$acc04,$s1,`32-16`,24,31
 	rlwinm	$acc05,$s2,`32-16`,24,31
-	lbzx	$acc02,$Tbl1,$acc02
-	lbzx	$acc03,$Tbl1,$acc03
 	rlwinm	$acc06,$s3,`32-16`,24,31
 	rlwinm	$acc07,$s0,`32-16`,24,31
-	lbzx	$acc04,$Tbl1,$acc04
-	lbzx	$acc05,$Tbl1,$acc05
+	lbzx	$acc00,$Tbl1,$acc00
+	lbzx	$acc01,$Tbl1,$acc01
 	rlwinm	$acc08,$s2,`32-8`,24,31
 	rlwinm	$acc09,$s3,`32-8`,24,31
-	lbzx	$acc06,$Tbl1,$acc06
-	lbzx	$acc07,$Tbl1,$acc07
+	lbzx	$acc02,$Tbl1,$acc02
+	lbzx	$acc03,$Tbl1,$acc03
 	rlwinm	$acc10,$s0,`32-8`,24,31
 	rlwinm	$acc11,$s1,`32-8`,24,31
-	lbzx	$acc08,$Tbl1,$acc08
-	lbzx	$acc09,$Tbl1,$acc09
+	lbzx	$acc04,$Tbl1,$acc04
+	lbzx	$acc05,$Tbl1,$acc05
 	rlwinm	$acc12,$s3,`0`,24,31
 	rlwinm	$acc13,$s0,`0`,24,31
-	lbzx	$acc10,$Tbl1,$acc10
-	lbzx	$acc11,$Tbl1,$acc11
+	lbzx	$acc06,$Tbl1,$acc06
+	lbzx	$acc07,$Tbl1,$acc07
 	rlwinm	$acc14,$s1,`0`,24,31
 	rlwinm	$acc15,$s2,`0`,24,31
-	lbzx	$acc12,$Tbl1,$acc12
-	lbzx	$acc13,$Tbl1,$acc13
+	lbzx	$acc08,$Tbl1,$acc08
+	lbzx	$acc09,$Tbl1,$acc09
 	rlwinm	$s0,$acc00,24,0,7
 	rlwinm	$s1,$acc01,24,0,7
-	lbzx	$acc14,$Tbl1,$acc14
-	lbzx	$acc15,$Tbl1,$acc15
+	lbzx	$acc10,$Tbl1,$acc10
+	lbzx	$acc11,$Tbl1,$acc11
 	rlwinm	$s2,$acc02,24,0,7
 	rlwinm	$s3,$acc03,24,0,7
+	lbzx	$acc12,$Tbl1,$acc12
+	lbzx	$acc13,$Tbl1,$acc13
 	rlwimi	$s0,$acc04,16,8,15
 	rlwimi	$s1,$acc05,16,8,15
+	lbzx	$acc14,$Tbl1,$acc14
+	lbzx	$acc15,$Tbl1,$acc15
 	rlwimi	$s2,$acc06,16,8,15
 	rlwimi	$s3,$acc07,16,8,15
 	rlwimi	$s0,$acc08,8,16,23
@@ -725,7 +738,7 @@ Lenc_compact_done:
 	addi	$sp,$sp,$FRAME
 	blr
 
-.align	4
+.align	5
 Lppc_AES_decrypt:
 	lwz	$acc00,240($key)
 	lwz	$t0,0($key)
@@ -746,46 +759,46 @@ Lppc_AES_decrypt:
 Ldec_loop:
 	rlwinm	$acc00,$s0,`32-24+3`,21,28
 	rlwinm	$acc01,$s1,`32-24+3`,21,28
-	lwz	$t0,0($key)
-	lwz	$t1,4($key)
 	rlwinm	$acc02,$s2,`32-24+3`,21,28
 	rlwinm	$acc03,$s3,`32-24+3`,21,28
-	lwz	$t2,8($key)
-	lwz	$t3,12($key)
+	lwz	$t0,0($key)
+	lwz	$t1,4($key)
 	rlwinm	$acc04,$s3,`32-16+3`,21,28
 	rlwinm	$acc05,$s0,`32-16+3`,21,28
-	lwzx	$acc00,$Tbl0,$acc00
-	lwzx	$acc01,$Tbl0,$acc01
+	lwz	$t2,8($key)
+	lwz	$t3,12($key)
 	rlwinm	$acc06,$s1,`32-16+3`,21,28
 	rlwinm	$acc07,$s2,`32-16+3`,21,28
-	lwzx	$acc02,$Tbl0,$acc02
-	lwzx	$acc03,$Tbl0,$acc03
+	lwzx	$acc00,$Tbl0,$acc00
+	lwzx	$acc01,$Tbl0,$acc01
 	rlwinm	$acc08,$s2,`32-8+3`,21,28
 	rlwinm	$acc09,$s3,`32-8+3`,21,28
-	lwzx	$acc04,$Tbl1,$acc04
-	lwzx	$acc05,$Tbl1,$acc05
+	lwzx	$acc02,$Tbl0,$acc02
+	lwzx	$acc03,$Tbl0,$acc03
 	rlwinm	$acc10,$s0,`32-8+3`,21,28
 	rlwinm	$acc11,$s1,`32-8+3`,21,28
-	lwzx	$acc06,$Tbl1,$acc06
-	lwzx	$acc07,$Tbl1,$acc07
+	lwzx	$acc04,$Tbl1,$acc04
+	lwzx	$acc05,$Tbl1,$acc05
 	rlwinm	$acc12,$s1,`0+3`,21,28
 	rlwinm	$acc13,$s2,`0+3`,21,28
-	lwzx	$acc08,$Tbl2,$acc08
-	lwzx	$acc09,$Tbl2,$acc09
+	lwzx	$acc06,$Tbl1,$acc06
+	lwzx	$acc07,$Tbl1,$acc07
 	rlwinm	$acc14,$s3,`0+3`,21,28
 	rlwinm	$acc15,$s0,`0+3`,21,28
-	lwzx	$acc10,$Tbl2,$acc10
-	lwzx	$acc11,$Tbl2,$acc11
+	lwzx	$acc08,$Tbl2,$acc08
+	lwzx	$acc09,$Tbl2,$acc09
 	xor	$t0,$t0,$acc00
 	xor	$t1,$t1,$acc01
-	lwzx	$acc12,$Tbl3,$acc12
-	lwzx	$acc13,$Tbl3,$acc13
+	lwzx	$acc10,$Tbl2,$acc10
+	lwzx	$acc11,$Tbl2,$acc11
 	xor	$t2,$t2,$acc02
 	xor	$t3,$t3,$acc03
-	lwzx	$acc14,$Tbl3,$acc14
-	lwzx	$acc15,$Tbl3,$acc15
+	lwzx	$acc12,$Tbl3,$acc12
+	lwzx	$acc13,$Tbl3,$acc13
 	xor	$t0,$t0,$acc04
 	xor	$t1,$t1,$acc05
+	lwzx	$acc14,$Tbl3,$acc14
+	lwzx	$acc15,$Tbl3,$acc15
 	xor	$t2,$t2,$acc06
 	xor	$t3,$t3,$acc07
 	xor	$t0,$t0,$acc08
@@ -801,56 +814,56 @@ Ldec_loop:
 
 	addi	$Tbl2,$Tbl0,2048
 	nop
-	lwz	$acc08,`2048+0`($Tbl0)	! prefetch Td4
-	lwz	$acc09,`2048+32`($Tbl0)
-	lwz	$acc10,`2048+64`($Tbl0)
-	lwz	$acc11,`2048+96`($Tbl0)
-	lwz	$acc08,`2048+128`($Tbl0)
-	lwz	$acc09,`2048+160`($Tbl0)
-	lwz	$acc10,`2048+192`($Tbl0)
-	lwz	$acc11,`2048+224`($Tbl0)
-	rlwinm	$acc00,$s0,`32-24`,24,31
-	rlwinm	$acc01,$s1,`32-24`,24,31
 	lwz	$t0,0($key)
 	lwz	$t1,4($key)
-	rlwinm	$acc02,$s2,`32-24`,24,31
-	rlwinm	$acc03,$s3,`32-24`,24,31
+	rlwinm	$acc00,$s0,`32-24`,24,31
+	rlwinm	$acc01,$s1,`32-24`,24,31
 	lwz	$t2,8($key)
 	lwz	$t3,12($key)
+	rlwinm	$acc02,$s2,`32-24`,24,31
+	rlwinm	$acc03,$s3,`32-24`,24,31
+	lwz	$acc08,`2048+0`($Tbl0)	! prefetch Td4
+	lwz	$acc09,`2048+32`($Tbl0)
 	rlwinm	$acc04,$s3,`32-16`,24,31
 	rlwinm	$acc05,$s0,`32-16`,24,31
+	lwz	$acc10,`2048+64`($Tbl0)
+	lwz	$acc11,`2048+96`($Tbl0)
 	lbzx	$acc00,$Tbl2,$acc00
 	lbzx	$acc01,$Tbl2,$acc01
+	lwz	$acc12,`2048+128`($Tbl0)
+	lwz	$acc13,`2048+160`($Tbl0)
 	rlwinm	$acc06,$s1,`32-16`,24,31
 	rlwinm	$acc07,$s2,`32-16`,24,31
-	lbzx	$acc02,$Tbl2,$acc02
-	lbzx	$acc03,$Tbl2,$acc03
+	lwz	$acc14,`2048+192`($Tbl0)
+	lwz	$acc15,`2048+224`($Tbl0)
 	rlwinm	$acc08,$s2,`32-8`,24,31
 	rlwinm	$acc09,$s3,`32-8`,24,31
-	lbzx	$acc04,$Tbl2,$acc04
-	lbzx	$acc05,$Tbl2,$acc05
+	lbzx	$acc02,$Tbl2,$acc02
+	lbzx	$acc03,$Tbl2,$acc03
 	rlwinm	$acc10,$s0,`32-8`,24,31
 	rlwinm	$acc11,$s1,`32-8`,24,31
-	lbzx	$acc06,$Tbl2,$acc06
-	lbzx	$acc07,$Tbl2,$acc07
+	lbzx	$acc04,$Tbl2,$acc04
+	lbzx	$acc05,$Tbl2,$acc05
 	rlwinm	$acc12,$s1,`0`,24,31
 	rlwinm	$acc13,$s2,`0`,24,31
-	lbzx	$acc08,$Tbl2,$acc08
-	lbzx	$acc09,$Tbl2,$acc09
+	lbzx	$acc06,$Tbl2,$acc06
+	lbzx	$acc07,$Tbl2,$acc07
 	rlwinm	$acc14,$s3,`0`,24,31
 	rlwinm	$acc15,$s0,`0`,24,31
-	lbzx	$acc10,$Tbl2,$acc10
-	lbzx	$acc11,$Tbl2,$acc11
+	lbzx	$acc08,$Tbl2,$acc08
+	lbzx	$acc09,$Tbl2,$acc09
 	rlwinm	$s0,$acc00,24,0,7
 	rlwinm	$s1,$acc01,24,0,7
-	lbzx	$acc12,$Tbl2,$acc12
-	lbzx	$acc13,$Tbl2,$acc13
+	lbzx	$acc10,$Tbl2,$acc10
+	lbzx	$acc11,$Tbl2,$acc11
 	rlwinm	$s2,$acc02,24,0,7
 	rlwinm	$s3,$acc03,24,0,7
-	lbzx	$acc14,$Tbl2,$acc14
-	lbzx	$acc15,$Tbl2,$acc15
+	lbzx	$acc12,$Tbl2,$acc12
+	lbzx	$acc13,$Tbl2,$acc13
 	rlwimi	$s0,$acc04,16,8,15
 	rlwimi	$s1,$acc05,16,8,15
+	lbzx	$acc14,$Tbl2,$acc14
+	lbzx	$acc15,$Tbl2,$acc15
 	rlwimi	$s2,$acc06,16,8,15
 	rlwimi	$s3,$acc07,16,8,15
 	rlwimi	$s0,$acc08,8,16,23
@@ -897,40 +910,40 @@ Ldec_compact_loop:
 	rlwinm	$acc01,$s1,`32-24`,24,31
 	rlwinm	$acc02,$s2,`32-24`,24,31
 	rlwinm	$acc03,$s3,`32-24`,24,31
-	lbzx	$acc00,$Tbl1,$acc00
-	lbzx	$acc01,$Tbl1,$acc01
 	rlwinm	$acc04,$s3,`32-16`,24,31
 	rlwinm	$acc05,$s0,`32-16`,24,31
-	lbzx	$acc02,$Tbl1,$acc02
-	lbzx	$acc03,$Tbl1,$acc03
 	rlwinm	$acc06,$s1,`32-16`,24,31
 	rlwinm	$acc07,$s2,`32-16`,24,31
-	lbzx	$acc04,$Tbl1,$acc04
-	lbzx	$acc05,$Tbl1,$acc05
+	lbzx	$acc00,$Tbl1,$acc00
+	lbzx	$acc01,$Tbl1,$acc01
 	rlwinm	$acc08,$s2,`32-8`,24,31
 	rlwinm	$acc09,$s3,`32-8`,24,31
-	lbzx	$acc06,$Tbl1,$acc06
-	lbzx	$acc07,$Tbl1,$acc07
+	lbzx	$acc02,$Tbl1,$acc02
+	lbzx	$acc03,$Tbl1,$acc03
 	rlwinm	$acc10,$s0,`32-8`,24,31
 	rlwinm	$acc11,$s1,`32-8`,24,31
-	lbzx	$acc08,$Tbl1,$acc08
-	lbzx	$acc09,$Tbl1,$acc09
+	lbzx	$acc04,$Tbl1,$acc04
+	lbzx	$acc05,$Tbl1,$acc05
 	rlwinm	$acc12,$s1,`0`,24,31
 	rlwinm	$acc13,$s2,`0`,24,31
-	lbzx	$acc10,$Tbl1,$acc10
-	lbzx	$acc11,$Tbl1,$acc11
+	lbzx	$acc06,$Tbl1,$acc06
+	lbzx	$acc07,$Tbl1,$acc07
 	rlwinm	$acc14,$s3,`0`,24,31
 	rlwinm	$acc15,$s0,`0`,24,31
-	lbzx	$acc12,$Tbl1,$acc12
-	lbzx	$acc13,$Tbl1,$acc13
+	lbzx	$acc08,$Tbl1,$acc08
+	lbzx	$acc09,$Tbl1,$acc09
 	rlwinm	$s0,$acc00,24,0,7
 	rlwinm	$s1,$acc01,24,0,7
-	lbzx	$acc14,$Tbl1,$acc14
-	lbzx	$acc15,$Tbl1,$acc15
+	lbzx	$acc10,$Tbl1,$acc10
+	lbzx	$acc11,$Tbl1,$acc11
 	rlwinm	$s2,$acc02,24,0,7
 	rlwinm	$s3,$acc03,24,0,7
+	lbzx	$acc12,$Tbl1,$acc12
+	lbzx	$acc13,$Tbl1,$acc13
 	rlwimi	$s0,$acc04,16,8,15
 	rlwimi	$s1,$acc05,16,8,15
+	lbzx	$acc14,$Tbl1,$acc14
+	lbzx	$acc15,$Tbl1,$acc15
 	rlwimi	$s2,$acc06,16,8,15
 	rlwimi	$s3,$acc07,16,8,15
 	rlwimi	$s0,$acc08,8,16,23

crypto/alphacpuid.pl

@@ -1,3 +1,5 @@
+#!/usr/bin/env perl
+print <<'___';
 .text
 
 .set	noat
@@ -68,9 +70,9 @@ OPENSSL_wipe_cpu:
 OPENSSL_atomic_add:
 	.frame	$30,0,$26
 	.prologue 0
-1:	ldl_l	$0,($16)
+1:	ldl_l	$0,0($16)
 	addl	$0,$17,$1
-	stl_c	$1,($16)
+	stl_c	$1,0($16)
 	beq	$1,1b
 	addl	$0,$17,$0
 	ret	($26)
@@ -123,3 +125,4 @@ OPENSSL_cleanse:
 	br	.Little
 .Ldone: ret	($26)
 .end	OPENSSL_cleanse
+___

crypto/asn1/a_int.c

@@ -273,7 +273,7 @@ ASN1_INTEGER *d2i_ASN1_UINTEGER(ASN1_INTEGER **a, const unsigned char **pp,
 	{
 	ASN1_INTEGER *ret=NULL;
 	const unsigned char *p;
-	unsigned char *to,*s;
+	unsigned char *s;
 	long len;
 	int inf,tag,xclass;
 	int i;
@@ -308,7 +308,6 @@ ASN1_INTEGER *d2i_ASN1_UINTEGER(ASN1_INTEGER **a, const unsigned char **pp,
 		i=ERR_R_MALLOC_FAILURE;
 		goto err;
 		}
-	to=s;
 	ret->type=V_ASN1_INTEGER;
 	if(len) {
 		if ((*p == 0) && (len != 1))

crypto/asn1/n_pkey.c

@@ -242,7 +242,7 @@ RSA *d2i_RSA_NET(RSA **a, const unsigned char **pp, long length,
 		 int sgckey)
 	{
 	RSA *ret=NULL;
-	const unsigned char *p, *kp;
+	const unsigned char *p;
 	NETSCAPE_ENCRYPTED_PKEY *enckey = NULL;
 
 	p = *pp;
@@ -265,7 +265,6 @@ RSA *d2i_RSA_NET(RSA **a, const unsigned char **pp, long length,
 		ASN1err(ASN1_F_D2I_RSA_NET,ASN1_R_UNSUPPORTED_ENCRYPTION_ALGORITHM);
 		goto err;
 	}
-	kp = enckey->enckey->digest->data;
 	if (cb == NULL)
 		cb=EVP_read_pw_string;
 	if ((ret=d2i_RSA_NET_2(a, enckey->enckey->digest,cb, sgckey)) == NULL) goto err;

crypto/asn1/t_crl.c

@@ -87,7 +87,7 @@ int X509_CRL_print(BIO *out, X509_CRL *x)
 	STACK_OF(X509_REVOKED) *rev;
 	X509_REVOKED *r;
 	long l;
-	int i, n;
+	int i;
 	char *p;
 
 	BIO_printf(out, "Certificate Revocation List (CRL):\n");
@@ -107,7 +107,6 @@ int X509_CRL_print(BIO *out, X509_CRL *x)
 	else BIO_printf(out,"NONE");
 	BIO_printf(out,"\n");
 
-	n=X509_CRL_get_ext_count(x);
 	X509V3_extensions_print(out, "CRL extensions",
 						x->crl->extensions, 0, 8);
 

crypto/asn1/tasn_dec.c

@@ -168,7 +168,7 @@ int ASN1_item_ex_d2i(ASN1_VALUE **pval, const unsigned char **in, long len,
 	int i;
 	int otag;
 	int ret = 0;
-	ASN1_VALUE *pchval, **pchptr, *ptmpval;
+	ASN1_VALUE **pchptr, *ptmpval;
 	if (!pval)
 		return 0;
 	if (aux && aux->asn1_cb)
@@ -319,7 +319,6 @@ int ASN1_item_ex_d2i(ASN1_VALUE **pval, const unsigned char **in, long len,
 			goto err;
 			}
 		/* CHOICE type, try each possibility in turn */
-		pchval = NULL;
 		p = *in;
 		for (i = 0, tt=it->templates; i < it->tcount; i++, tt++)
 			{

crypto/asn1/x_x509.c

@@ -63,7 +63,7 @@
 #include <openssl/x509.h>
 #include <openssl/x509v3.h>
 
-ASN1_SEQUENCE(X509_CINF) = {
+ASN1_SEQUENCE_enc(X509_CINF, enc, 0) = {
 	ASN1_EXP_OPT(X509_CINF, version, ASN1_INTEGER, 0),
 	ASN1_SIMPLE(X509_CINF, serialNumber, ASN1_INTEGER),
 	ASN1_SIMPLE(X509_CINF, signature, X509_ALGOR),
@@ -74,7 +74,7 @@ ASN1_SEQUENCE(X509_CINF) = {
 	ASN1_IMP_OPT(X509_CINF, issuerUID, ASN1_BIT_STRING, 1),
 	ASN1_IMP_OPT(X509_CINF, subjectUID, ASN1_BIT_STRING, 2),
 	ASN1_EXP_SEQUENCE_OF_OPT(X509_CINF, extensions, X509_EXTENSION, 3)
-} ASN1_SEQUENCE_END(X509_CINF)
+} ASN1_SEQUENCE_END_enc(X509_CINF, X509_CINF)
 
 IMPLEMENT_ASN1_FUNCTIONS(X509_CINF)
 /* X509 top level structure needs a bit of customisation */

crypto/bio/b_sock.c

@@ -731,7 +731,14 @@ again:
 #ifdef SO_REUSEADDR
 		err_num=get_last_socket_error();
 		if ((bind_mode == BIO_BIND_REUSEADDR_IF_UNUSED) &&
+#ifdef OPENSSL_SYS_WINDOWS
+			/* Some versions of Windows define EADDRINUSE to
+			 * a dummy value.
+			 */
+			(err_num == WSAEADDRINUSE))
+#else
 			(err_num == EADDRINUSE))
+#endif
 			{
 			client = server;
 			if (h == NULL || strcmp(h,"*") == 0)

crypto/bio/bf_nbio.c

@@ -125,7 +125,6 @@ static int nbiof_free(BIO *a)
 	
 static int nbiof_read(BIO *b, char *out, int outl)
 	{
-	NBIO_TEST *nt;
 	int ret=0;
 #if 1
 	int num;
@@ -134,7 +133,6 @@ static int nbiof_read(BIO *b, char *out, int outl)
 
 	if (out == NULL) return(0);
 	if (b->next_bio == NULL) return(0);
-	nt=(NBIO_TEST *)b->ptr;
 
 	BIO_clear_retry_flags(b);
 #if 1

crypto/bio/bio_lib.c

@@ -110,7 +110,7 @@ int BIO_set(BIO *bio, BIO_METHOD *method)
 
 int BIO_free(BIO *a)
 	{
-	int ret=0,i;
+	int i;
 
 	if (a == NULL) return(0);
 
@@ -133,7 +133,7 @@ int BIO_free(BIO *a)
 	CRYPTO_free_ex_data(CRYPTO_EX_INDEX_BIO, a, &a->ex_data);
 
 	if ((a->method == NULL) || (a->method->destroy == NULL)) return(1);
-	ret=a->method->destroy(a);
+	a->method->destroy(a);
 	OPENSSL_free(a);
 	return(1);
 	}

crypto/bio/bss_acpt.c

@@ -340,7 +340,6 @@ static int acpt_write(BIO *b, const char *in, int inl)
 
 static long acpt_ctrl(BIO *b, int cmd, long num, void *ptr)
 	{
-	BIO *dbio;
 	int *ip;
 	long ret=1;
 	BIO_ACCEPT *data;
@@ -437,8 +436,8 @@ static long acpt_ctrl(BIO *b, int cmd, long num, void *ptr)
 		ret=(long)data->bind_mode;
 		break;
 	case BIO_CTRL_DUP:
-		dbio=(BIO *)ptr;
-/*		if (data->param_port) EAY EAY
+/*		dbio=(BIO *)ptr;
+		if (data->param_port) EAY EAY
 			BIO_set_port(dbio,data->param_port);
 		if (data->param_hostname)
 			BIO_set_hostname(dbio,data->param_hostname);

crypto/bio/bss_file.c

@@ -118,10 +118,47 @@ static BIO_METHOD methods_filep=
 
 BIO *BIO_new_file(const char *filename, const char *mode)
 	{
-	BIO *ret;
-	FILE *file;
+	BIO  *ret;
+	FILE *file=NULL;
 
-	if ((file=fopen(filename,mode)) == NULL)
+#if defined(_WIN32) && defined(CP_UTF8)
+	int sz, len_0 = (int)strlen(filename)+1;
+
+	/*
+	 * Basically there are three cases to cover: a) filename is
+	 * pure ASCII string; b) actual UTF-8 encoded string and
+	 * c) locale-ized string, i.e. one containing 8-bit
+	 * characters that are meaningful in current system locale.
+	 * If filename is pure ASCII or real UTF-8 encoded string,
+	 * MultiByteToWideChar succeeds and _wfopen works. If
+	 * filename is locale-ized string, chances are that
+	 * MultiByteToWideChar fails reporting
+	 * ERROR_NO_UNICODE_TRANSLATION, in which case we fall
+	 * back to fopen...
+	 */
+	if ((sz=MultiByteToWideChar(CP_UTF8,MB_ERR_INVALID_CHARS,
+					filename,len_0,NULL,0))>0)
+		{
+		WCHAR  wmode[8];
+		WCHAR *wfilename = _alloca(sz*sizeof(WCHAR));
+
+		if (MultiByteToWideChar(CP_UTF8,MB_ERR_INVALID_CHARS,
+					filename,len_0,wfilename,sz) &&
+		    MultiByteToWideChar(CP_UTF8,0,mode,strlen(mode)+1,
+			    		wmode,sizeof(wmode)/sizeof(wmode[0])) &&
+		    (file=_wfopen(wfilename,wmode))==NULL && errno==ENOENT
+		   )	/* UTF-8 decode succeeded, but no file, filename
+			 * could still have been locale-ized... */
+			file = fopen(filename,mode);
+		}
+	else if (GetLastError()==ERROR_NO_UNICODE_TRANSLATION)
+		{
+		file = fopen(filename,mode);
+		}
+#else
+	file=fopen(filename,mode);	
+#endif
+	if (file == NULL)
 		{
 		SYSerr(SYS_F_FOPEN,get_last_sys_error());
 		ERR_add_error_data(5,"fopen('",filename,"','",mode,"')");

crypto/bn/asm/alpha-mont.pl

@@ -41,8 +41,12 @@ $j="s4";
 $m1="s5";
 
 $code=<<___;
+#indef __linux__
+#include <asm/regdef.h>
+#else
 #include <asm.h>
 #include <regdef.h>
+#endif
 
 .text
 
@@ -53,15 +57,15 @@ $code=<<___;
 .align	5
 .ent	bn_mul_mont
 bn_mul_mont:
-	lda	sp,-40(sp)
+	lda	sp,-48(sp)
 	stq	ra,0(sp)
 	stq	s3,8(sp)
 	stq	s4,16(sp)
 	stq	s5,24(sp)
 	stq	fp,32(sp)
 	mov	sp,fp
-	.mask	0x0400f000,-40
-	.frame	fp,40,ra
+	.mask	0x0400f000,-48
+	.frame	fp,48,ra
 	.prologue 0
 
 	.align	4
@@ -76,7 +80,7 @@ bn_mul_mont:
 	ldq	$aj,8($ap)
 	subq	sp,AT,sp
 	ldq	$bi,0($bp)	# bp[0]
-	mov	-4096,AT
+	lda	AT,-4096(zero)	# mov	-4096,AT
 	ldq	$n0,0($n0)
 	and	sp,AT,sp
 
@@ -106,9 +110,9 @@ bn_mul_mont:
 .align	4
 .L1st:
 	.set	noreorder
-	ldq	$aj,($aj)
+	ldq	$aj,0($aj)
 	addl	$j,1,$j
-	ldq	$nj,($nj)
+	ldq	$nj,0($nj)
 	lda	$tp,8($tp)
 
 	addq	$alo,$hi0,$lo0
@@ -159,12 +163,12 @@ bn_mul_mont:
 .align	4
 .Louter:
 	s8addq	$i,$bp,$bi
-	ldq	$hi0,($ap)
+	ldq	$hi0,0($ap)
 	ldq	$aj,8($ap)
-	ldq	$bi,($bi)
-	ldq	$hi1,($np)
+	ldq	$bi,0($bi)
+	ldq	$hi1,0($np)
 	ldq	$nj,8($np)
-	ldq	$tj,(sp)
+	ldq	$tj,0(sp)
 
 	mulq	$hi0,$bi,$lo0
 	umulh	$hi0,$bi,$hi0
@@ -195,10 +199,10 @@ bn_mul_mont:
 	.set	noreorder
 	ldq	$tj,8($tp)	#L0
 	nop			#U1
-	ldq	$aj,($aj)	#L1
+	ldq	$aj,0($aj)	#L1
 	s8addq	$j,$np,$nj	#U0
 
-	ldq	$nj,($nj)	#L0
+	ldq	$nj,0($nj)	#L0
 	nop			#U1
 	addq	$alo,$hi0,$lo0	#L1
 	lda	$tp,8($tp)
@@ -247,7 +251,7 @@ bn_mul_mont:
 	addq	$hi1,v0,$hi1
 
 	addq	$hi1,$hi0,$lo1
-	stq	$j,($tp)
+	stq	$j,0($tp)
 	cmpult	$lo1,$hi0,$hi1
 	addq	$lo1,$tj,$lo1
 	cmpult	$lo1,$tj,AT
@@ -265,8 +269,8 @@ bn_mul_mont:
 	mov	0,$hi0		# clear borrow bit
 
 .align	4
-.Lsub:	ldq	$lo0,($tp)
-	ldq	$lo1,($np)
+.Lsub:	ldq	$lo0,0($tp)
+	ldq	$lo1,0($np)
 	lda	$tp,8($tp)
 	lda	$np,8($np)
 	subq	$lo0,$lo1,$lo1	# tp[i]-np[i]
@@ -274,7 +278,7 @@ bn_mul_mont:
 	subq	$lo1,$hi0,$lo0
 	cmpult	$lo1,$lo0,$hi0
 	or	$hi0,AT,$hi0
-	stq	$lo0,($rp)
+	stq	$lo0,0($rp)
 	cmpult	$tp,$tj,v0
 	lda	$rp,8($rp)
 	bne	v0,.Lsub
@@ -288,7 +292,7 @@ bn_mul_mont:
 	bis	$bp,$ap,$ap	# ap=borrow?tp:rp
 
 .align	4
-.Lcopy:	ldq	$aj,($ap)	# copy or in-place refresh
+.Lcopy:	ldq	$aj,0($ap)	# copy or in-place refresh
 	lda	$tp,8($tp)
 	lda	$rp,8($rp)
 	lda	$ap,8($ap)
@@ -306,11 +310,11 @@ bn_mul_mont:
 	ldq	s4,16(sp)
 	ldq	s5,24(sp)
 	ldq	fp,32(sp)
-	lda	sp,40(sp)
+	lda	sp,48(sp)
 	ret	(ra)
 .end	bn_mul_mont
-.rdata
-.asciiz	"Montgomery Multiplication for Alpha, CRYPTOGAMS by <appro\@openssl.org>"
+.ascii	"Montgomery Multiplication for Alpha, CRYPTOGAMS by <appro\@openssl.org>"
+.align	2
 ___
 
 print $code;

crypto/bn/asm/s390x.S

@@ -1,4 +1,4 @@
-.ident "s390x.S, version 1.0"
+.ident "s390x.S, version 1.1"
 // ====================================================================
 // Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
 // project.
@@ -24,67 +24,67 @@ bn_mul_add_words:
 	bler	%r14		// if (len<=0) return 0;
 
 	stmg	%r6,%r10,48(%r15)
+	lghi	%r10,3
 	lghi	%r8,0		// carry = 0
-	srag	%r10,%r4,2	// cnt=len/4
-	jz	.Loop1_madd
+	nr	%r10,%r4	// len%4
+	sra	%r4,2		// cnt=len/4
+	jz	.Loop1_madd	// carry is incidentally cleared if branch taken
+	algr	zero,zero	// clear carry
 
 .Loop4_madd:
 	lg	%r7,0(%r2,%r3)	// ap[i]
 	mlgr	%r6,%r5		// *=w
-	algr	%r7,%r8		// +=carry
+	alcgr	%r7,%r8		// +=carry
 	alcgr	%r6,zero
 	alg	%r7,0(%r2,%r1)	// +=rp[i]
-	alcgr	%r6,zero
 	stg	%r7,0(%r2,%r1)	// rp[i]=
 
 	lg	%r9,8(%r2,%r3)
 	mlgr	%r8,%r5
-	algr	%r9,%r6
+	alcgr	%r9,%r6
 	alcgr	%r8,zero
 	alg	%r9,8(%r2,%r1)
-	alcgr	%r8,zero
 	stg	%r9,8(%r2,%r1)
 
 	lg	%r7,16(%r2,%r3)
 	mlgr	%r6,%r5
-	algr	%r7,%r8
+	alcgr	%r7,%r8
 	alcgr	%r6,zero
 	alg	%r7,16(%r2,%r1)
-	alcgr	%r6,zero
 	stg	%r7,16(%r2,%r1)
 
 	lg	%r9,24(%r2,%r3)
 	mlgr	%r8,%r5
-	algr	%r9,%r6
+	alcgr	%r9,%r6
 	alcgr	%r8,zero
 	alg	%r9,24(%r2,%r1)
-	alcgr	%r8,zero
 	stg	%r9,24(%r2,%r1)
 
 	la	%r2,32(%r2)	// i+=4
-	brct	%r10,.Loop4_madd
+	brct	%r4,.Loop4_madd
 
-	lghi	%r10,3
-	nr	%r4,%r10	// cnt=len%4
-	jz	.Lend_madd
+	la	%r10,1(%r10)		// see if len%4 is zero ...
+	brct	%r10,.Loop1_madd	// without touching condition code:-)
+
+.Lend_madd:
+	alcgr	%r8,zero	// collect carry bit
+	lgr	%r2,%r8
+	lmg	%r6,%r10,48(%r15)
+	br	%r14
 
 .Loop1_madd:
 	lg	%r7,0(%r2,%r3)	// ap[i]
 	mlgr	%r6,%r5		// *=w
-	algr	%r7,%r8		// +=carry
+	alcgr	%r7,%r8		// +=carry
 	alcgr	%r6,zero
 	alg	%r7,0(%r2,%r1)	// +=rp[i]
-	alcgr	%r6,zero
 	stg	%r7,0(%r2,%r1)	// rp[i]=
 
 	lgr	%r8,%r6
 	la	%r2,8(%r2)	// i++
-	brct	%r4,.Loop1_madd
+	brct	%r10,.Loop1_madd
 
-.Lend_madd:
-	lgr	%r2,%r8
-	lmg	%r6,%r10,48(%r15)
-	br	%r14
+	j	.Lend_madd
 .size	bn_mul_add_words,.-bn_mul_add_words
 
 // BN_ULONG bn_mul_words(BN_ULONG *r2,BN_ULONG *r3,int r4,BN_ULONG r5);
@@ -99,57 +99,57 @@ bn_mul_words:
 	bler	%r14		// if (len<=0) return 0;
 
 	stmg	%r6,%r10,48(%r15)
+	lghi	%r10,3
 	lghi	%r8,0		// carry = 0
-	srag	%r10,%r4,2	// cnt=len/4
-	jz	.Loop1_mul
+	nr	%r10,%r4	// len%4
+	sra	%r4,2		// cnt=len/4
+	jz	.Loop1_mul	// carry is incidentally cleared if branch taken
+	algr	zero,zero	// clear carry
 
 .Loop4_mul:
 	lg	%r7,0(%r2,%r3)	// ap[i]
 	mlgr	%r6,%r5		// *=w
-	algr	%r7,%r8		// +=carry
-	alcgr	%r6,zero
+	alcgr	%r7,%r8		// +=carry
 	stg	%r7,0(%r2,%r1)	// rp[i]=
 
 	lg	%r9,8(%r2,%r3)
 	mlgr	%r8,%r5
-	algr	%r9,%r6
-	alcgr	%r8,zero
+	alcgr	%r9,%r6
 	stg	%r9,8(%r2,%r1)
 
 	lg	%r7,16(%r2,%r3)
 	mlgr	%r6,%r5
-	algr	%r7,%r8
-	alcgr	%r6,zero
+	alcgr	%r7,%r8
 	stg	%r7,16(%r2,%r1)
 
 	lg	%r9,24(%r2,%r3)
 	mlgr	%r8,%r5
-	algr	%r9,%r6
-	alcgr	%r8,zero
+	alcgr	%r9,%r6
 	stg	%r9,24(%r2,%r1)
 
 	la	%r2,32(%r2)	// i+=4
-	brct	%r10,.Loop4_mul
+	brct	%r4,.Loop4_mul
 
-	lghi	%r10,3
-	nr	%r4,%r10	// cnt=len%4
-	jz	.Lend_mul
+	la	%r10,1(%r10)		// see if len%4 is zero ...
+	brct	%r10,.Loop1_madd	// without touching condition code:-)
+
+.Lend_mul:
+	alcgr	%r8,zero	// collect carry bit
+	lgr	%r2,%r8
+	lmg	%r6,%r10,48(%r15)
+	br	%r14
 
 .Loop1_mul:
 	lg	%r7,0(%r2,%r3)	// ap[i]
 	mlgr	%r6,%r5		// *=w
-	algr	%r7,%r8		// +=carry
-	alcgr	%r6,zero
+	alcgr	%r7,%r8		// +=carry
 	stg	%r7,0(%r2,%r1)	// rp[i]=
 
 	lgr	%r8,%r6
 	la	%r2,8(%r2)	// i++
-	brct	%r4,.Loop1_mul
+	brct	%r10,.Loop1_mul
 
-.Lend_mul:
-	lgr	%r2,%r8
-	lmg	%r6,%r10,48(%r15)
-	br	%r14
+	j	.Lend_mul
 .size	bn_mul_words,.-bn_mul_words
 
 // void bn_sqr_words(BN_ULONG *r2,BN_ULONG *r2,int r4)

crypto/bn/bn_exp2.c

@@ -301,7 +301,8 @@ int BN_mod_exp2_mont(BIGNUM *rr, const BIGNUM *a1, const BIGNUM *p1,
 			r_is_one = 0;
 			}
 		}
-	BN_from_montgomery(rr,r,mont,ctx);
+	if (!BN_from_montgomery(rr,r,mont,ctx))
+		goto err;
 	ret=1;
 err:
 	if ((in_mont == NULL) && (mont != NULL)) BN_MONT_CTX_free(mont);

crypto/bn/bn_mul.c

@@ -551,7 +551,7 @@ void bn_mul_part_recursive(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n,
 	     int tna, int tnb, BN_ULONG *t)
 	{
 	int i,j,n2=n*2;
-	int c1,c2,neg,zero;
+	int c1,c2,neg;
 	BN_ULONG ln,lo,*p;
 
 # ifdef BN_COUNT
@@ -567,7 +567,7 @@ void bn_mul_part_recursive(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n,
 	/* r=(a[0]-a[1])*(b[1]-b[0]) */
 	c1=bn_cmp_part_words(a,&(a[n]),tna,n-tna);
 	c2=bn_cmp_part_words(&(b[n]),b,tnb,tnb-n);
-	zero=neg=0;
+	neg=0;
 	switch (c1*3+c2)
 		{
 	case -4:
@@ -575,7 +575,6 @@ void bn_mul_part_recursive(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n,
 		bn_sub_part_words(&(t[n]),b,      &(b[n]),tnb,n-tnb); /* - */
 		break;
 	case -3:
-		zero=1;
 		/* break; */
 	case -2:
 		bn_sub_part_words(t,      &(a[n]),a,      tna,tna-n); /* - */
@@ -585,7 +584,6 @@ void bn_mul_part_recursive(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n,
 	case -1:
 	case 0:
 	case 1:
-		zero=1;
 		/* break; */
 	case 2:
 		bn_sub_part_words(t,      a,      &(a[n]),tna,n-tna); /* + */
@@ -593,7 +591,6 @@ void bn_mul_part_recursive(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n,
 		neg=1;
 		break;
 	case 3:
-		zero=1;
 		/* break; */
 	case 4:
 		bn_sub_part_words(t,      a,      &(a[n]),tna,n-tna);
@@ -1012,7 +1009,6 @@ int BN_mul(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, BN_CTX *ctx)
 		{
 		if (i >= -1 && i <= 1)
 			{
-			int sav_j =0;
 			/* Find out the power of two lower or equal
 			   to the longest of the two numbers */
 			if (i >= 0)
@@ -1023,7 +1019,6 @@ int BN_mul(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, BN_CTX *ctx)
 				{
 				j = BN_num_bits_word((BN_ULONG)bl);
 				}
-			sav_j = j;
 			j = 1<<(j-1);
 			assert(j <= al || j <= bl);
 			k = j+j;

crypto/cms/cms_asn1.c

@@ -131,8 +131,8 @@ ASN1_NDEF_SEQUENCE(CMS_SignedData) = {
 } ASN1_NDEF_SEQUENCE_END(CMS_SignedData)
 
 ASN1_SEQUENCE(CMS_OriginatorInfo) = {
-	ASN1_IMP_SET_OF_OPT(CMS_SignedData, certificates, CMS_CertificateChoices, 0),
-	ASN1_IMP_SET_OF_OPT(CMS_SignedData, crls, CMS_RevocationInfoChoice, 1)
+	ASN1_IMP_SET_OF_OPT(CMS_OriginatorInfo, certificates, CMS_CertificateChoices, 0),
+	ASN1_IMP_SET_OF_OPT(CMS_OriginatorInfo, crls, CMS_RevocationInfoChoice, 1)
 } ASN1_SEQUENCE_END(CMS_OriginatorInfo)
 
 ASN1_NDEF_SEQUENCE(CMS_EncryptedContentInfo) = {

crypto/conf/conf_api.c

@@ -285,7 +285,7 @@ CONF_VALUE *_CONF_new_section(CONF *conf, const char *section)
 	v->value=(char *)sk;
 	
 	vv=lh_CONF_VALUE_insert(conf->data,v);
-	assert(vv == NULL);
+	OPENSSL_assert(vv == NULL);
 	ok=1;
 err:
 	if (!ok)

crypto/conf/conf_def.c

@@ -213,13 +213,12 @@ static int def_load_bio(CONF *conf, BIO *in, long *line)
 	int bufnum=0,i,ii;
 	BUF_MEM *buff=NULL;
 	char *s,*p,*end;
-	int again,n;
+	int again;
 	long eline=0;
 	char btmp[DECIMAL_SIZE(eline)+1];
 	CONF_VALUE *v=NULL,*tv;
 	CONF_VALUE *sv=NULL;
 	char *section=NULL,*buf;
-	STACK_OF(CONF_VALUE) *section_sk=NULL,*ts;
 	char *start,*psection,*pname;
 	void *h = (void *)(conf->data);
 
@@ -250,7 +249,6 @@ static int def_load_bio(CONF *conf, BIO *in, long *line)
 					CONF_R_UNABLE_TO_CREATE_NEW_SECTION);
 		goto err;
 		}
-	section_sk=(STACK_OF(CONF_VALUE) *)sv->value;
 
 	bufnum=0;
 	again=0;
@@ -309,7 +307,6 @@ static int def_load_bio(CONF *conf, BIO *in, long *line)
 		buf=buff->data;
 
 		clear_comments(conf, buf);
-		n=strlen(buf);
 		s=eat_ws(conf, buf);
 		if (IS_EOF(conf,*s)) continue; /* blank line */
 		if (*s == '[')
@@ -343,7 +340,6 @@ again:
 					CONF_R_UNABLE_TO_CREATE_NEW_SECTION);
 				goto err;
 				}
-			section_sk=(STACK_OF(CONF_VALUE) *)sv->value;
 			continue;
 			}
 		else
@@ -406,13 +402,9 @@ again:
 					   CONF_R_UNABLE_TO_CREATE_NEW_SECTION);
 					goto err;
 					}
-				ts=(STACK_OF(CONF_VALUE) *)tv->value;
 				}
 			else
-				{
 				tv=sv;
-				ts=section_sk;
-				}
 #if 1
 			if (_CONF_add_string(conf, tv, v) == 0)
 				{
@@ -465,9 +457,6 @@ err:
 
 static void clear_comments(CONF *conf, char *p)
 	{
-	char *to;
-
-	to=p;
 	for (;;)
 		{
 		if (IS_FCOMMENT(conf,*p))

crypto/cryptlib.c

@@ -749,6 +749,18 @@ int OPENSSL_isservice(void)
 { HWINSTA h;
   DWORD len;
   WCHAR *name;
+  static union { void *p; int (*f)(void); } _OPENSSL_isservice = { NULL };
+
+    if (_OPENSSL_isservice.p == NULL) {
+	HANDLE h = GetModuleHandle(NULL);
+	if (h != NULL)
+	    _OPENSSL_isservice.p = GetProcAddress(h,"_OPENSSL_isservice");
+	if (_OPENSSL_isservice.p == NULL)
+	    _OPENSSL_isservice.p = (void *)-1;
+    }
+
+    if (_OPENSSL_isservice.p != (void *)-1)
+	return (*_OPENSSL_isservice.f)();
 
     (void)GetDesktopWindow(); /* return value is ignored */
 

crypto/crypto-lib.com

@@ -60,7 +60,7 @@ $ THEN
 $!
 $!  The Architecture Is VAX
 $!
-$   ARCH := VAX
+$   ARCH = "VAX"
 $!
 $! Else...
 $!
@@ -80,9 +80,11 @@ $! NOTE: Some might think this list ugly.  However, it's made this way to
 $! reflect the SDIRS variable in [-]Makefile.org as closely as possible,
 $! thereby making it fairly easy to verify that the lists are the same.
 $!
+$ ET_WHIRLPOOL = "WHRLPOOL"
+$ IF ARCH .EQS. "VAX" THEN ET_WHIRLPOOL = ""
 $ ENCRYPT_TYPES = "Basic,"+ -
 		  "OBJECTS,"+ -
-		  "MD2,MD4,MD5,SHA,MDC2,HMAC,RIPEMD,WHRLPOOL,"+ -
+		  "MD2,MD4,MD5,SHA,MDC2,HMAC,RIPEMD,"+ET_WHIRLPOOL+","+ -
 		  "DES,AES,RC2,RC4,RC5,IDEA,BF,CAST,CAMELLIA,SEED,MODES,"+ -
 		  "BN,EC,RSA,DSA,ECDSA,DH,ECDH,DSO,ENGINE,"+ -
 		  "BUFFER,BIO,STACK,LHASH,RAND,ERR,"+ -
@@ -367,7 +369,7 @@ $!
 $ IF F$TYPE('LIB_MODULE') .EQS. ""
 $ THEN
 $   WRITE SYS$ERROR ""
-$   WRITE SYS$ERROR "The module ",MODULE_NAME," does not exist.  Continuing..."
+$   WRITE SYS$ERROR "The module ",MODULE_NAME1," does not exist.  Continuing..."
 $   WRITE SYS$ERROR ""
 $   GOTO MODULE_NEXT
 $ ENDIF
@@ -777,12 +779,12 @@ $! Else...
 $!
 $ ELSE
 $!
-$!  Else, Check To See If P1 Has A Valid Arguement.
+$!  Else, Check To See If P1 Has A Valid Argument.
 $!
 $   IF (P1.EQS."LIBRARY").OR.(P1.EQS."APPS")
 $   THEN
 $!
-$!    A Valid Arguement.
+$!    A Valid Argument.
 $!
 $     BUILDALL = P1
 $!
@@ -810,7 +812,7 @@ $!    Time To EXIT.
 $!
 $     EXIT
 $!
-$!  End The Valid Arguement Check.
+$!  End The Valid Argument Check.
 $!
 $   ENDIF
 $!
@@ -863,7 +865,7 @@ $!    Time To EXIT.
 $!
 $     EXIT
 $!
-$!  End The Valid Arguement Check.
+$!  End The Valid Argument Check.
 $!
 $   ENDIF
 $!
@@ -1034,7 +1036,7 @@ $     IF ARCH.EQS."VAX" .AND. F$TRNLNM("DECC$CC_DEFAULT").NES."/DECC" -
 	 THEN CC = "CC/DECC"
 $     CC = CC + "/''CC_OPTIMIZE'/''DEBUGGER'/STANDARD=ANSI89" + -
            "/NOLIST/PREFIX=ALL" + -
-	   "/INCLUDE=(SYS$DISK:[],SYS$DISK:[.''ARCH'],SYS$DISK:[-],SYS$DISK:[.ENGINE.VENDOR_DEFNS],SYS$DISK:[.EVP],SYS$DISK:[.ASN1])" + -
+	   "/INCLUDE=(SYS$DISK:[],SYS$DISK:[._''ARCH'],SYS$DISK:[-],SYS$DISK:[.ENGINE.VENDOR_DEFNS],SYS$DISK:[.EVP],SYS$DISK:[.ASN1])" + -
 	   CCEXTRAFLAGS
 $!
 $!    Define The Linker Options File Name.
@@ -1068,7 +1070,7 @@ $	EXIT
 $     ENDIF
 $     IF F$TRNLNM("DECC$CC_DEFAULT").EQS."/DECC" THEN CC = "CC/VAXC"
 $     CC = CC + "/''CC_OPTIMIZE'/''DEBUGGER'/NOLIST" + -
-	   "/INCLUDE=(SYS$DISK:[],SYS$DISK:[.''ARCH'],SYS$DISK:[-],SYS$DISK:[.ENGINE.VENDOR_DEFNS],SYS$DISK:[.EVP],SYS$DISK:[.ASN1])" + -
+	   "/INCLUDE=(SYS$DISK:[],SYS$DISK:[._''ARCH'],SYS$DISK:[-],SYS$DISK:[.ENGINE.VENDOR_DEFNS],SYS$DISK:[.EVP],SYS$DISK:[.ASN1])" + -
 	   CCEXTRAFLAGS
 $     CCDEFS = """VAXC""," + CCDEFS
 $!
@@ -1100,7 +1102,7 @@ $!
 $!    Use GNU C...
 $!
 $     CC = "GCC/NOCASE_HACK/''GCC_OPTIMIZE'/''DEBUGGER'/NOLIST" + -
-	   "/INCLUDE=(SYS$DISK:[],SYS$DISK:[.''ARCH'],SYS$DISK:[-],SYS$DISK:[.ENGINE.VENDOR_DEFNS],SYS$DISK:[.EVP],SYS$DISK:[.ASN1])" + -
+	   "/INCLUDE=(SYS$DISK:[],SYS$DISK:[._''ARCH'],SYS$DISK:[-],SYS$DISK:[.ENGINE.VENDOR_DEFNS],SYS$DISK:[.EVP],SYS$DISK:[.ASN1])" + -
 	   CCEXTRAFLAGS
 $!
 $!    Define The Linker Options File Name.
@@ -1150,7 +1152,7 @@ $!  Show user the result
 $!
 $   WRITE/SYMBOL SYS$OUTPUT "Main C Compiling Command: ",CC
 $!
-$!  Else The User Entered An Invalid Arguement.
+$!  Else The User Entered An Invalid Argument.
 $!
 $ ELSE
 $!
@@ -1168,7 +1170,7 @@ $!  Time To EXIT.
 $!
 $   EXIT
 $!
-$! End The Valid Arguement Check.
+$! End The Valid Argument Check.
 $!
 $ ENDIF
 $!
@@ -1263,7 +1265,7 @@ $!  Print info
 $!
 $   WRITE SYS$OUTPUT "TCP/IP library spec: ", TCPIP_LIB
 $!
-$!  Else The User Entered An Invalid Arguement.
+$!  Else The User Entered An Invalid Argument.
 $!
 $ ELSE
 $!

crypto/des/des-lib.com

@@ -659,13 +659,13 @@ $! Else...
 $!
 $ ELSE
 $!
-$!  Else, Check To See If P1 Has A Valid Arguement.
+$!  Else, Check To See If P1 Has A Valid Argument.
 $!
 $   IF (P1.EQS."LIBRARY").OR.(P1.EQS."DESTEST").OR.(P1.EQS."SPEED") -
        .OR.(P1.EQS."RPW").OR.(P1.EQS."DES").OR.(P1.EQS."DES_OPTS")
 $   THEN
 $!
-$!    A Valid Arguement.
+$!    A Valid Argument.
 $!
 $     BUILDALL = P1
 $!
@@ -678,7 +678,7 @@ $!
 $     WRITE SYS$OUTPUT ""
 $     WRITE SYS$OUTPUT "The Option ",P1," Is Invalid.  The Valid Options Are:"
 $     WRITE SYS$OUTPUT ""
-$     WRITE SYS$OUTPUT "    ALL      :  Just Build Everything.
+$     WRITE SYS$OUTPUT "    ALL      :  Just Build Everything."
 $     WRITE SYS$OUTPUT "    LIBRARY  :  To Compile Just The [.xxx.EXE.CRYPTO.DES]LIBDES.OLB Library."
 $     WRITE SYS$OUTPUT "    DESTEST  :  To Compile Just The [.xxx.EXE.CRYPTO.DES]DESTEST.EXE Program."
 $     WRITE SYS$OUTPUT "    SPEED    :  To Compile Just The [.xxx.EXE.CRYPTO.DES]SPEED.EXE Program."
@@ -697,7 +697,7 @@ $!    Time To EXIT.
 $!
 $     EXIT
 $!
-$!  End The Valid Arguement Check.
+$!  End The Valid Argument Check.
 $!
 $   ENDIF
 $!
@@ -754,7 +754,7 @@ $!    Time To EXIT.
 $!
 $     EXIT
 $!
-$!  End The Valid Arguement Check.
+$!  End The Valid Argument Check.
 $!
 $   ENDIF
 $!
@@ -978,7 +978,7 @@ $!  Show user the result
 $!
 $   WRITE SYS$OUTPUT "Main Compiling Command: ",CC
 $!
-$!  Else The User Entered An Invalid Arguement.
+$!  Else The User Entered An Invalid Argument.
 $!
 $ ELSE
 $!

crypto/des/rpc_des.h

@@ -122,10 +122,10 @@ struct desparams {
 /*
  * Encrypt an arbitrary sized buffer
  */
-#define	DESIOCBLOCK	_IOWR(d, 6, struct desparams)
+#define	DESIOCBLOCK	_IOWR('d', 6, struct desparams)
 
 /* 
  * Encrypt of small amount of data, quickly
  */
-#define DESIOCQUICK	_IOWR(d, 7, struct desparams) 
+#define DESIOCQUICK	_IOWR('d', 7, struct desparams) 
 

crypto/dsa/dsa_ameth.c

@@ -209,7 +209,7 @@ static int dsa_priv_decode(EVP_PKEY *pkey, PKCS8_PRIV_KEY_INFO *p8)
 	if (*p == (V_ASN1_SEQUENCE|V_ASN1_CONSTRUCTED))
 		{
 		ASN1_TYPE *t1, *t2;
-	    	if(!(ndsa = d2i_ASN1_SEQUENCE_ANY(NULL, &p, pklen)));
+	    	if(!(ndsa = d2i_ASN1_SEQUENCE_ANY(NULL, &p, pklen)))
 			goto decerr;
 		if (sk_ASN1_TYPE_num(ndsa) != 2)
 			goto decerr;

crypto/dsa/dsa_gen.c

@@ -120,7 +120,7 @@ int dsa_builtin_paramgen(DSA *ret, size_t bits, size_t qbits,
 	BIGNUM *r0,*W,*X,*c,*test;
 	BIGNUM *g=NULL,*q=NULL,*p=NULL;
 	BN_MONT_CTX *mont=NULL;
-	int i, k,n=0,b,m=0, qsize = qbits >> 3;
+	int i, k, n=0, m=0, qsize = qbits >> 3;
 	int counter=0;
 	int r=0;
 	BN_CTX *ctx=NULL;
@@ -232,7 +232,6 @@ int dsa_builtin_paramgen(DSA *ret, size_t bits, size_t qbits,
 		/* "offset = 2" */
 
 		n=(bits-1)/160;
-		b=(bits-1)-n*160;
 
 		for (;;)
 			{

crypto/dsa/dsa_ossl.c

@@ -185,7 +185,7 @@ static DSA_SIG *dsa_do_sign(const unsigned char *dgst, int dlen, DSA *dsa)
 	if (!BN_mod_mul(&xr,dsa->priv_key,r,dsa->q,ctx)) goto err;/* s = xr */
 	if (!BN_add(s, &xr, &m)) goto err;		/* s = m + xr */
 	if (BN_cmp(s,dsa->q) > 0)
-		BN_sub(s,s,dsa->q);
+		if (!BN_sub(s,s,dsa->q)) goto err;
 	if (!BN_mod_mul(s,s,kinv,dsa->q,ctx)) goto err;
 
 	ret=DSA_SIG_new();

crypto/ec/ec2_mult.c

@@ -319,6 +319,7 @@ int ec_GF2m_simple_mul(const EC_GROUP *group, EC_POINT *r, const BIGNUM *scalar,
 	int ret = 0;
 	size_t i;
 	EC_POINT *p=NULL;
+	EC_POINT *acc = NULL;
 
 	if (ctx == NULL)
 		{
@@ -338,15 +339,16 @@ int ec_GF2m_simple_mul(const EC_GROUP *group, EC_POINT *r, const BIGNUM *scalar,
 		}
 
 	if ((p = EC_POINT_new(group)) == NULL) goto err;
+	if ((acc = EC_POINT_new(group)) == NULL) goto err;
 
-	if (!EC_POINT_set_to_infinity(group, r)) goto err;
+	if (!EC_POINT_set_to_infinity(group, acc)) goto err;
 
 	if (scalar)
 		{
 		if (!ec_GF2m_montgomery_point_multiply(group, p, scalar, group->generator, ctx)) goto err;
-		if (BN_is_negative(scalar)) 
+		if (BN_is_negative(scalar))
 			if (!group->meth->invert(group, p, ctx)) goto err;
-		if (!group->meth->add(group, r, r, p, ctx)) goto err;
+		if (!group->meth->add(group, acc, acc, p, ctx)) goto err;
 		}
 
 	for (i = 0; i < num; i++)
@@ -354,13 +356,16 @@ int ec_GF2m_simple_mul(const EC_GROUP *group, EC_POINT *r, const BIGNUM *scalar,
 		if (!ec_GF2m_montgomery_point_multiply(group, p, scalars[i], points[i], ctx)) goto err;
 		if (BN_is_negative(scalars[i]))
 			if (!group->meth->invert(group, p, ctx)) goto err;
-		if (!group->meth->add(group, r, r, p, ctx)) goto err;
+		if (!group->meth->add(group, acc, acc, p, ctx)) goto err;
 		}
 
+	if (!EC_POINT_copy(r, acc)) goto err;
+
 	ret = 1;
 
   err:
 	if (p) EC_POINT_free(p);
+	if (acc) EC_POINT_free(acc);
 	if (new_ctx != NULL)
 		BN_CTX_free(new_ctx);
 	return ret;

crypto/ec/ec_mult.c

@@ -169,11 +169,13 @@ static void ec_pre_comp_clear_free(void *pre_)
 		EC_POINT **p;
 
 		for (p = pre->points; *p != NULL; p++)
+			{
 			EC_POINT_clear_free(*p);
-		OPENSSL_cleanse(pre->points, sizeof pre->points);
+			OPENSSL_cleanse(p, sizeof *p);
+			}
 		OPENSSL_free(pre->points);
 		}
-	OPENSSL_cleanse(pre, sizeof pre);
+	OPENSSL_cleanse(pre, sizeof *pre);
 	OPENSSL_free(pre);
 	}
 

crypto/ecdh/ech_lib.c

@@ -96,7 +96,6 @@ const ECDH_METHOD *ECDH_get_default_method(void)
 
 int ECDH_set_method(EC_KEY *eckey, const ECDH_METHOD *meth)
 	{
-	const ECDH_METHOD *mtmp;
 	ECDH_DATA *ecdh;
 
 	ecdh = ecdh_check(eckey);
@@ -104,8 +103,8 @@ int ECDH_set_method(EC_KEY *eckey, const ECDH_METHOD *meth)
 	if (ecdh == NULL)
 		return 0;
 
-        mtmp = ecdh->meth;
 #if 0
+        mtmp = ecdh->meth;
         if (mtmp->finish)
 		mtmp->finish(eckey);
 #endif

crypto/ecdsa/ecs_lib.c

@@ -83,7 +83,6 @@ const ECDSA_METHOD *ECDSA_get_default_method(void)
 
 int ECDSA_set_method(EC_KEY *eckey, const ECDSA_METHOD *meth)
 {
-        const ECDSA_METHOD *mtmp;
 	ECDSA_DATA *ecdsa;
 
 	ecdsa = ecdsa_check(eckey);
@@ -91,7 +90,6 @@ int ECDSA_set_method(EC_KEY *eckey, const ECDSA_METHOD *meth)
 	if (ecdsa == NULL)
 		return 0;
 
-        mtmp = ecdsa->meth;
 #ifndef OPENSSL_NO_ENGINE
 	if (ecdsa->engine)
 	{

crypto/err/err_prn.c

@@ -81,7 +81,8 @@ void ERR_print_errors_cb(int (*cb)(const char *str, size_t len, void *u),
 		ERR_error_string_n(l, buf, sizeof buf);
 		BIO_snprintf(buf2, sizeof(buf2), "%lu:%s:%s:%d:%s\n", es, buf,
 			file, line, (flags & ERR_TXT_STRING) ? data : "");
-		cb(buf2, strlen(buf2), u);
+		if (cb(buf2, strlen(buf2), u) <= 0)
+			break; /* abort outputting the error report */
 		}
 	}
 

crypto/evp/bio_b64.c

@@ -64,7 +64,7 @@
 
 static int b64_write(BIO *h, const char *buf, int num);
 static int b64_read(BIO *h, char *buf, int size);
-/*static int b64_puts(BIO *h, const char *str); */
+static int b64_puts(BIO *h, const char *str);
 /*static int b64_gets(BIO *h, char *str, int size); */
 static long b64_ctrl(BIO *h, int cmd, long arg1, void *arg2);
 static int b64_new(BIO *h);
@@ -96,7 +96,7 @@ static BIO_METHOD methods_b64=
 	BIO_TYPE_BASE64,"base64 encoding",
 	b64_write,
 	b64_read,
-	NULL, /* b64_puts, */
+	b64_puts,
 	NULL, /* b64_gets, */
 	b64_ctrl,
 	b64_new,
@@ -127,6 +127,7 @@ static int b64_new(BIO *bi)
 	bi->init=1;
 	bi->ptr=(char *)ctx;
 	bi->flags=0;
+	bi->num = 0;
 	return(1);
 	}
 
@@ -151,6 +152,8 @@ static int b64_read(BIO *b, char *out, int outl)
 
 	if ((ctx == NULL) || (b->next_bio == NULL)) return(0);
 
+	BIO_clear_retry_flags(b);
+
 	if (ctx->encode != B64_DECODE)
 		{
 		ctx->encode=B64_DECODE;
@@ -163,6 +166,7 @@ static int b64_read(BIO *b, char *out, int outl)
 	/* First check if there are bytes decoded/encoded */
 	if (ctx->buf_len > 0)
 		{
+		OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 		i=ctx->buf_len-ctx->buf_off;
 		if (i > outl) i=outl;
 		OPENSSL_assert(ctx->buf_off+i < (int)sizeof(ctx->buf));
@@ -184,7 +188,6 @@ static int b64_read(BIO *b, char *out, int outl)
 	ret_code=0;
 	while (outl > 0)
 		{
-
 		if (ctx->cont <= 0)
 			break;
 
@@ -195,7 +198,7 @@ static int b64_read(BIO *b, char *out, int outl)
 			{
 			ret_code=i;
 
-			/* Should be continue next time we are called? */
+			/* Should we continue next time we are called? */
 			if (!BIO_should_retry(b->next_bio))
 				{
 				ctx->cont=i;
@@ -285,19 +288,27 @@ static int b64_read(BIO *b, char *out, int outl)
 				continue;
 				}
 			else
+			{
 				ctx->tmp_len=0;
 			}
-		/* If buffer isn't full and we can retry then
-		 * restart to read in more data.
-		 */
+		}
 		else if ((i < B64_BLOCK_SIZE) && (ctx->cont > 0))
+		{
+			/* If buffer isn't full and we can retry then
+			 * restart to read in more data.
+			 */
 			continue;
+		}
 
 		if (BIO_get_flags(b) & BIO_FLAGS_BASE64_NO_NL)
 			{
 			int z,jj;
 
+#if 0
 			jj=(i>>2)<<2;
+#else
+			jj = i & ~3; /* process per 4 */
+#endif
 			z=EVP_DecodeBlock((unsigned char *)ctx->buf,
 				(unsigned char *)ctx->tmp,jj);
 			if (jj > 2)
@@ -313,18 +324,15 @@ static int b64_read(BIO *b, char *out, int outl)
 			 * number consumed */
 			if (jj != i)
 				{
-				memcpy((unsigned char *)ctx->tmp,
-					(unsigned char *)&(ctx->tmp[jj]),i-jj);
+				memmove(ctx->tmp, &ctx->tmp[jj], i-jj);
 				ctx->tmp_len=i-jj;
 				}
 			ctx->buf_len=0;
 			if (z > 0)
 				{
 				ctx->buf_len=z;
-				i=1;
 				}
-			else
-				i=z;
+			i=z;
 			}
 		else
 			{
@@ -357,14 +365,16 @@ static int b64_read(BIO *b, char *out, int outl)
 		outl-=i;
 		out+=i;
 		}
-	BIO_clear_retry_flags(b);
+	/* BIO_clear_retry_flags(b); */
 	BIO_copy_next_retry(b);
 	return((ret == 0)?ret_code:ret);
 	}
 
 static int b64_write(BIO *b, const char *in, int inl)
 	{
-	int ret=inl,n,i;
+	int ret=0;
+	int n;
+	int i;
 	BIO_B64_CTX *ctx;
 
 	ctx=(BIO_B64_CTX *)b->ptr;
@@ -379,6 +389,9 @@ static int b64_write(BIO *b, const char *in, int inl)
 		EVP_EncodeInit(&(ctx->base64));
 		}
 
+	OPENSSL_assert(ctx->buf_off < (int)sizeof(ctx->buf));
+	OPENSSL_assert(ctx->buf_len <= (int)sizeof(ctx->buf));
+	OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 	n=ctx->buf_len-ctx->buf_off;
 	while (n > 0)
 		{
@@ -388,7 +401,10 @@ static int b64_write(BIO *b, const char *in, int inl)
 			BIO_copy_next_retry(b);
 			return(i);
 			}
+		OPENSSL_assert(i <= n);
 		ctx->buf_off+=i;
+		OPENSSL_assert(ctx->buf_off <= (int)sizeof(ctx->buf));
+		OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 		n-=i;
 		}
 	/* at this point all pending data has been written */
@@ -405,18 +421,19 @@ static int b64_write(BIO *b, const char *in, int inl)
 			{
 			if (ctx->tmp_len > 0)
 				{
+				OPENSSL_assert(ctx->tmp_len <= 3);
 				n=3-ctx->tmp_len;
-				/* There's a teoretical possibility for this */
+				/* There's a theoretical possibility for this */
 				if (n > inl) 
 					n=inl;
 				memcpy(&(ctx->tmp[ctx->tmp_len]),in,n);
 				ctx->tmp_len+=n;
+				ret += n;
 				if (ctx->tmp_len < 3)
 					break;
-				ctx->buf_len=EVP_EncodeBlock(
-					(unsigned char *)ctx->buf,
-					(unsigned char *)ctx->tmp,
-					ctx->tmp_len);
+				ctx->buf_len=EVP_EncodeBlock((unsigned char *)ctx->buf,(unsigned char *)ctx->tmp,ctx->tmp_len);
+				OPENSSL_assert(ctx->buf_len <= (int)sizeof(ctx->buf));
+				OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 				/* Since we're now done using the temporary
 				   buffer, the length should be 0'd */
 				ctx->tmp_len=0;
@@ -425,14 +442,16 @@ static int b64_write(BIO *b, const char *in, int inl)
 				{
 				if (n < 3)
 					{
-					memcpy(&(ctx->tmp[0]),in,n);
+					memcpy(ctx->tmp,in,n);
 					ctx->tmp_len=n;
+					ret += n;
 					break;
 					}
 				n-=n%3;
-				ctx->buf_len=EVP_EncodeBlock(
-					(unsigned char *)ctx->buf,
-					(unsigned char *)in,n);
+				ctx->buf_len=EVP_EncodeBlock((unsigned char *)ctx->buf,(const unsigned char *)in,n);
+				OPENSSL_assert(ctx->buf_len <= (int)sizeof(ctx->buf));
+				OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
+				ret += n;
 				}
 			}
 		else
@@ -440,6 +459,9 @@ static int b64_write(BIO *b, const char *in, int inl)
 			EVP_EncodeUpdate(&(ctx->base64),
 				(unsigned char *)ctx->buf,&ctx->buf_len,
 				(unsigned char *)in,n);
+			OPENSSL_assert(ctx->buf_len <= (int)sizeof(ctx->buf));
+			OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
+			ret += n;
 			}
 		inl-=n;
 		in+=n;
@@ -454,8 +476,11 @@ static int b64_write(BIO *b, const char *in, int inl)
 				BIO_copy_next_retry(b);
 				return((ret == 0)?i:ret);
 				}
+			OPENSSL_assert(i <= n);
 			n-=i;
 			ctx->buf_off+=i;
+			OPENSSL_assert(ctx->buf_off <= (int)sizeof(ctx->buf));
+			OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 			}
 		ctx->buf_len=0;
 		ctx->buf_off=0;
@@ -486,6 +511,7 @@ static long b64_ctrl(BIO *b, int cmd, long num, void *ptr)
 			ret=BIO_ctrl(b->next_bio,cmd,num,ptr);
 		break;
 	case BIO_CTRL_WPENDING: /* More to write in buffer */
+		OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 		ret=ctx->buf_len-ctx->buf_off;
 		if ((ret == 0) && (ctx->encode != B64_NONE)
 			&& (ctx->base64.num != 0))
@@ -494,6 +520,7 @@ static long b64_ctrl(BIO *b, int cmd, long num, void *ptr)
 			ret=BIO_ctrl(b->next_bio,cmd,num,ptr);
 		break;
 	case BIO_CTRL_PENDING: /* More to read in buffer */
+		OPENSSL_assert(ctx->buf_len >= ctx->buf_off);
 		ret=ctx->buf_len-ctx->buf_off;
 		if (ret <= 0)
 			ret=BIO_ctrl(b->next_bio,cmd,num,ptr);
@@ -565,3 +592,7 @@ static long b64_callback_ctrl(BIO *b, int cmd, bio_info_cb *fp)
 	return(ret);
 	}
 
+static int b64_puts(BIO *b, const char *str)
+	{
+	return b64_write(b,str,strlen(str));
+	}

crypto/evp/encode.c

@@ -235,7 +235,7 @@ void EVP_DecodeInit(EVP_ENCODE_CTX *ctx)
 int EVP_DecodeUpdate(EVP_ENCODE_CTX *ctx, unsigned char *out, int *outl,
 	     const unsigned char *in, int inl)
 	{
-	int seof= -1,eof=0,rv= -1,ret=0,i,v,tmp,n,ln,tmp2,exp_nl;
+	int seof= -1,eof=0,rv= -1,ret=0,i,v,tmp,n,ln,exp_nl;
 	unsigned char *d;
 
 	n=ctx->num;
@@ -319,7 +319,6 @@ int EVP_DecodeUpdate(EVP_ENCODE_CTX *ctx, unsigned char *out, int *outl,
 			 * lines.  We process the line and then need to
 			 * accept the '\n' */
 			if ((v != B64_EOF) && (n >= 64)) exp_nl=1;
-			tmp2=v;
 			if (n > 0)
 				{
 				v=EVP_DecodeBlock(out,d,n);

crypto/evp/evp_enc.c

@@ -204,6 +204,7 @@ skip_to_init:
 			case EVP_CIPH_OFB_MODE:
 
 			ctx->num = 0;
+			/* fall-through */
 
 			case EVP_CIPH_CBC_MODE:
 

crypto/evp/pmeth_lib.c

@@ -134,6 +134,8 @@ static EVP_PKEY_CTX *int_ctx_new(EVP_PKEY *pkey, ENGINE *e, int id)
 		id = pkey->ameth->pkey_id;
 		}
 #ifndef OPENSSL_NO_ENGINE
+	if (pkey && pkey->engine)
+		e = pkey->engine;
 	/* Try to find an ENGINE which implements this method */
 	if (e)
 		{
@@ -177,6 +179,7 @@ static EVP_PKEY_CTX *int_ctx_new(EVP_PKEY *pkey, ENGINE *e, int id)
 	ret->operation = EVP_PKEY_OP_UNDEFINED;
 	ret->pkey = pkey;
 	ret->peerkey = NULL;
+	ret->pkey_gencb = 0;
 	if (pkey)
 		CRYPTO_add(&pkey->references,1,CRYPTO_LOCK_EVP_PKEY);
 	ret->data = NULL;

crypto/hmac/hmac.c

@@ -138,12 +138,9 @@ int HMAC_Update(HMAC_CTX *ctx, const unsigned char *data, size_t len)
 
 int HMAC_Final(HMAC_CTX *ctx, unsigned char *md, unsigned int *len)
 	{
-	int j;
 	unsigned int i;
 	unsigned char buf[EVP_MAX_MD_SIZE];
 
-	j=EVP_MD_block_size(ctx->md);
-
 	if (!EVP_DigestFinal_ex(&ctx->md_ctx,buf,&i))
 		goto err;
 	if (!EVP_MD_CTX_copy_ex(&ctx->md_ctx,&ctx->o_ctx))

crypto/md32_common.h

@@ -165,7 +165,7 @@
 				asm (			\
 				"roll %1,%0"		\
 				: "=r"(ret)		\
-				: "I"(n), "0"(a)	\
+				: "I"(n), "0"((unsigned int)(a))	\
 				: "cc");		\
 			   ret;				\
 			})
@@ -383,6 +383,7 @@ int HASH_FINAL (unsigned char *md, HASH_CTX *c)
 	}
 
 #ifndef MD32_REG_T
+#if defined(__alpha) || defined(__sparcv9) || defined(__mips)
 #define MD32_REG_T long
 /*
  * This comment was originaly written for MD5, which is why it
@@ -400,9 +401,15 @@ int HASH_FINAL (unsigned char *md, HASH_CTX *c)
  * Well, to be honest it should say that this *prevents* 
  * performance degradation.
  *				<appro@fy.chalmers.se>
- * Apparently there're LP64 compilers that generate better
- * code if A-D are declared int. Most notably GCC-x86_64
- * generates better code.
+ */
+#else
+/*
+ * Above is not absolute and there are LP64 compilers that
+ * generate better code if MD32_REG_T is defined int. The above
+ * pre-processor condition reflects the circumstances under which
+ * the conclusion was made and is subject to further extension.
  *				<appro@fy.chalmers.se>
  */
+#define MD32_REG_T int
+#endif
 #endif

crypto/md5/asm/md5-ia64.S

@@ -914,7 +914,7 @@ md5_digest_block##offset:						\
 	nop 0x0 ;							\
 	br.cond.sptk.many md5_digest_GHI ;				\
 } ;;									\
-	.endp md5digestBlock ## offset
+	.endp md5_digest_block##offset
 
 MD5FBLOCK(1)
 MD5FBLOCK(2)

crypto/modes/cfb128.c

@@ -96,15 +96,15 @@ void CRYPTO_cfb128_encrypt(const unsigned char *in, unsigned char *out,
 #endif
 		while (len>=16) {
 			(*block)(ivec, ivec, key);
-			for (n=0; n<16; n+=sizeof(size_t)) {
+			for (; n<16; n+=sizeof(size_t)) {
 				*(size_t*)(out+n) =
 				*(size_t*)(ivec+n) ^= *(size_t*)(in+n);
 			}
 			len -= 16;
 			out += 16;
 			in  += 16;
+			n = 0;
 		}
-		n = 0;
 		if (len) {
 			(*block)(ivec, ivec, key);
 			while (len--) {
@@ -141,7 +141,7 @@ void CRYPTO_cfb128_encrypt(const unsigned char *in, unsigned char *out,
 #endif
 		while (len>=16) {
 			(*block)(ivec, ivec, key);
-			for (n=0; n<16; n+=sizeof(size_t)) {
+			for (; n<16; n+=sizeof(size_t)) {
 				size_t t = *(size_t*)(in+n);
 				*(size_t*)(out+n) = *(size_t*)(ivec+n) ^ t;
 				*(size_t*)(ivec+n) = t;
@@ -149,8 +149,8 @@ void CRYPTO_cfb128_encrypt(const unsigned char *in, unsigned char *out,
 			len -= 16;
 			out += 16;
 			in  += 16;
+			n = 0;
 		}
-		n = 0;
 		if (len) {
 			(*block)(ivec, ivec, key);
 			while (len--) {

crypto/modes/ctr128.c

@@ -61,14 +61,11 @@
 typedef unsigned int u32;
 typedef unsigned char u8;
 
-# define GETU32(pt) (((u32)(pt)[0] << 24) ^ ((u32)(pt)[1] << 16) ^ ((u32)(pt)[2] <<  8) ^ ((u32)(pt)[3]))
-# define PUTU32(ct, st) { (ct)[0] = (u8)((st) >> 24); (ct)[1] = (u8)((st) >> 16); (ct)[2] = (u8)((st) >>  8); (ct)[3] = (u8)(st); }
-
 #define STRICT_ALIGNMENT
-#if defined(__i386) || defined(__i386__) || \
-    defined(__x86_64) || defined(__x86_64__) || \
-    defined(_M_IX86) || defined(_M_AMD64) || defined(_M_X64) || \
-    defined(__s390__) || defined(__s390x__)
+#if defined(__i386)	|| defined(__i386__)	|| \
+    defined(__x86_64)	|| defined(__x86_64__)	|| \
+    defined(_M_IX86)	|| defined(_M_AMD64)	|| defined(_M_X64) || \
+    defined(__s390__)	|| defined(__s390x__)
 #  undef STRICT_ALIGNMENT
 #endif
 
@@ -77,18 +74,19 @@ typedef unsigned char u8;
 
 /* increment counter (128-bit int) by 1 */
 static void ctr128_inc(unsigned char *counter) {
-	u32 c,n=16;
+	u32 n=16;
+	u8  c;
 
 	do {
-		n -= 4;
-		c = GETU32(counter+n);
-		++c;	c &= 0xFFFFFFFF;
-		PUTU32(counter + n, c);
+		--n;
+		c = counter[n];
+		++c;
+		counter[n] = c;
 		if (c) return;
 	} while (n);
 }
 
-#if !defined(OPENSSL_SMALL_FOORPRINT)
+#if !defined(OPENSSL_SMALL_FOOTPRINT)
 static void ctr128_inc_aligned(unsigned char *counter) {
 	size_t *data,c,n;
 	const union { long one; char little; } is_endian = {1};
@@ -151,14 +149,14 @@ void CRYPTO_ctr128_encrypt(const unsigned char *in, unsigned char *out,
 		while (len>=16) {
 			(*block)(ivec, ecount_buf, key);
 			ctr128_inc_aligned(ivec);
-			for (n=0; n<16; n+=sizeof(size_t))
+			for (; n<16; n+=sizeof(size_t))
 				*(size_t *)(out+n) =
 				*(size_t *)(in+n) ^ *(size_t *)(ecount_buf+n);
 			len -= 16;
 			out += 16;
 			in  += 16;
+			n = 0;
 		}
-		n = 0;
 		if (len) {
 			(*block)(ivec, ecount_buf, key);
  			ctr128_inc_aligned(ivec);

crypto/modes/ofb128.c

@@ -95,14 +95,14 @@ void CRYPTO_ofb128_encrypt(const unsigned char *in, unsigned char *out,
 #endif
 		while (len>=16) {
 			(*block)(ivec, ivec, key);
-			for (n=0; n<16; n+=sizeof(size_t))
+			for (; n<16; n+=sizeof(size_t))
 				*(size_t*)(out+n) =
 				*(size_t*)(in+n) ^ *(size_t*)(ivec+n);
 			len -= 16;
 			out += 16;
 			in  += 16;
+			n = 0;
 		}
-		n = 0;
 		if (len) {
 			(*block)(ivec, ivec, key);
 			while (len--) {

crypto/ocsp/ocsp_ht.c

@@ -397,11 +397,12 @@ int OCSP_sendreq_nbio(OCSP_RESPONSE **presp, OCSP_REQ_CTX *rctx)
 
 
 		case OHS_ASN1_HEADER:
-		/* Now reading ASN1 header: can read at least 6 bytes which
-		 * is more than enough for any valid ASN1 SEQUENCE header
+		/* Now reading ASN1 header: can read at least 2 bytes which
+		 * is enough for ASN1 SEQUENCE header and either length field
+		 * or at least the length of the length field.
 		 */
 		n = BIO_get_mem_data(rctx->mem, &p);
-		if (n < 6)
+		if (n < 2)
 			goto next_io;
 
 		/* Check it is an ASN1 SEQUENCE */
@@ -414,6 +415,11 @@ int OCSP_sendreq_nbio(OCSP_RESPONSE **presp, OCSP_REQ_CTX *rctx)
 		/* Check out length field */
 		if (*p & 0x80)
 			{
+			/* If MSB set on initial length octet we can now
+			 * always read 6 octets: make sure we have them.
+			 */
+			if (n < 6)
+				goto next_io;
 			n = *p & 0x7F;
 			/* Not NDEF or excessive length */
 			if (!n || (n > 4))

crypto/ocsp/ocsp_prn.c

@@ -182,7 +182,6 @@ int OCSP_RESPONSE_print(BIO *bp, OCSP_RESPONSE* o, unsigned long flags)
         {
 	int i, ret = 0;
 	long l;
-	unsigned char *p;
 	OCSP_CERTID *cid = NULL;
 	OCSP_BASICRESP *br = NULL;
 	OCSP_RESPID *rid = NULL;
@@ -207,7 +206,6 @@ int OCSP_RESPONSE_print(BIO *bp, OCSP_RESPONSE* o, unsigned long flags)
 		return 1;
 		}
 
-	p = ASN1_STRING_data(rb->response);
 	i = ASN1_STRING_length(rb->response);
 	if (!(br = OCSP_response_get1_basic(o))) goto err;
 	rd = br->tbsResponseData;

crypto/opensslv.h

@@ -25,11 +25,11 @@
  * (Prior to 0.9.5a beta1, a different scheme was used: MMNNFFRBB for
  *  major minor fix final patch/beta)
  */
-#define OPENSSL_VERSION_NUMBER	0x1000000fL
+#define OPENSSL_VERSION_NUMBER	0x1000002f
 #ifdef OPENSSL_FIPS
-#define OPENSSL_VERSION_TEXT	"OpenSSL 1.0.0-fips 29 Mar 2010"
+#define OPENSSL_VERSION_TEXT	"OpenSSL 1.0.0b-fips 16 Nov 2010"
 #else
-#define OPENSSL_VERSION_TEXT	"OpenSSL 1.0.0 29 Mar 2010"
+#define OPENSSL_VERSION_TEXT	"OpenSSL 1.0.0b 16 Nov 2010"
 #endif
 #define OPENSSL_VERSION_PTEXT	" part of " OPENSSL_VERSION_TEXT
 

crypto/pem/pem.h

@@ -548,10 +548,11 @@ EVP_PKEY *b2i_PrivateKey_bio(BIO *in);
 EVP_PKEY *b2i_PublicKey_bio(BIO *in);
 int i2b_PrivateKey_bio(BIO *out, EVP_PKEY *pk);
 int i2b_PublicKey_bio(BIO *out, EVP_PKEY *pk);
-
+#ifndef OPENSSL_NO_RC4
 EVP_PKEY *b2i_PVK_bio(BIO *in, pem_password_cb *cb, void *u);
 int i2b_PVK_bio(BIO *out, EVP_PKEY *pk, int enclevel,
 		pem_password_cb *cb, void *u);
+#endif
 
 
 /* BEGIN ERROR CODES */

crypto/pem/pem_lib.c

@@ -482,7 +482,6 @@ int PEM_do_header(EVP_CIPHER_INFO *cipher, unsigned char *data, long *plen,
 
 int PEM_get_EVP_CIPHER_INFO(char *header, EVP_CIPHER_INFO *cipher)
 	{
-	int o;
 	const EVP_CIPHER *enc=NULL;
 	char *p,c;
 	char **header_pp = &header;
@@ -522,7 +521,6 @@ int PEM_get_EVP_CIPHER_INFO(char *header, EVP_CIPHER_INFO *cipher)
 		header++;
 		}
 	*header='\0';
-	o=OBJ_sn2nid(p);
 	cipher->cipher=enc=EVP_get_cipherbyname(p);
 	*header=c;
 	header++;

crypto/pem/pvkfmt.c

@@ -654,13 +654,15 @@ int i2b_PublicKey_bio(BIO *out, EVP_PKEY *pk)
 	return do_i2b_bio(out, pk, 1);
 	}
 
+#ifndef OPENSSL_NO_RC4
+
 static int do_PVK_header(const unsigned char **in, unsigned int length,
 		int skip_magic,
 	       	unsigned int *psaltlen, unsigned int *pkeylen)
 		
 	{
 	const unsigned char *p = *in;
-	unsigned int pvk_magic, keytype, is_encrypted;
+	unsigned int pvk_magic, is_encrypted;
 	if (skip_magic)
 		{
 		if (length < 20)
@@ -687,7 +689,7 @@ static int do_PVK_header(const unsigned char **in, unsigned int length,
 		}
 	/* Skip reserved */
 	p += 4;
-	keytype = read_ledword(&p);
+	/*keytype = */read_ledword(&p);
 	is_encrypted = read_ledword(&p);
 	*psaltlen = read_ledword(&p);
 	*pkeylen = read_ledword(&p);
@@ -837,7 +839,7 @@ EVP_PKEY *b2i_PVK_bio(BIO *in, pem_password_cb *cb, void *u)
 static int i2b_PVK(unsigned char **out, EVP_PKEY*pk, int enclevel,
 		pem_password_cb *cb, void *u)
 	{
-	int outlen = 24, noinc, pklen;
+	int outlen = 24, pklen;
 	unsigned char *p, *salt = NULL;
 	if (enclevel)
 		outlen += PVK_SALTLEN;
@@ -848,10 +850,7 @@ static int i2b_PVK(unsigned char **out, EVP_PKEY*pk, int enclevel,
 	if (!out)
 		return outlen;
 	if (*out)
-		{
 		p = *out;
-		noinc = 0;
-		}
 	else
 		{
 		p = OPENSSL_malloc(outlen);
@@ -861,7 +860,6 @@ static int i2b_PVK(unsigned char **out, EVP_PKEY*pk, int enclevel,
 			return -1;
 			}
 		*out = p;
-		noinc = 1;
 		}
 
 	write_ledword(&p, MS_PVKMAGIC);
@@ -934,4 +932,7 @@ int i2b_PVK_bio(BIO *out, EVP_PKEY *pk, int enclevel,
 		}
 	return -1;
 	}
+
+#endif
+
 #endif

crypto/perlasm/x86_64-xlate.pl

@@ -55,6 +55,8 @@
 #    Win64 prologue copies %rsp value to %rax. For further details
 #    see SEH paragraph at the end.
 # 9. .init segment is allowed to contain calls to functions only.
+# a. If function accepts more than 4 arguments *and* >4th argument
+#    is declared as non 64-bit value, do clear its upper part.
 
 my $flavour = shift;
 my $output  = shift;
@@ -80,7 +82,10 @@ my $PTR=" PTR";
 my $nasmref=2.03;
 my $nasm=0;
 
-if    ($flavour eq "mingw64")	{ $gas=1; $elf=0; $win64=1; $prefix="_"; }
+if    ($flavour eq "mingw64")	{ $gas=1; $elf=0; $win64=1;
+				  $prefix=`echo __USER_LABEL_PREFIX__ | $ENV{CC} -E -P -`;
+				  chomp($prefix);
+				}
 elsif ($flavour eq "macosx")	{ $gas=1; $elf=0; $prefix="_"; $decor="L\$"; }
 elsif ($flavour eq "masm")	{ $gas=0; $elf=0; $masm=$masmref; $win64=1; $decor="\$L\$"; }
 elsif ($flavour eq "nasm")	{ $gas=0; $elf=0; $nasm=$nasmref; $win64=1; $decor="\$L\$"; $PTR=""; }
@@ -115,7 +120,9 @@ my %globals;
 		$self->{op} = $1;
 		$self->{sz} = "b";
 	    } elsif ($self->{op} =~ /call|jmp/) {
-		$self->{sz} = ""
+		$self->{sz} = "";
+	    } elsif ($self->{op} =~ /^p/ && $' !~ /^(ush|op)/) { # SSEn
+		$self->{sz} = "";
 	    } elsif ($self->{op} =~ /([a-z]{3,})([qlwb])$/) {
 		$self->{op} = $1;
 		$self->{sz} = $2;
@@ -160,7 +167,7 @@ my %globals;
 	    } elsif ($self->{op} =~ /^(pop|push)f/) {
 		$self->{op} .= $self->{sz};
 	    } elsif ($self->{op} eq "call" && $current_segment eq ".CRT\$XCU") {
-		$self->{op} = "ALIGN\t8\n\tDQ";
+		$self->{op} = "\tDQ";
 	    } 
 	    $self->{op};
 	}
@@ -191,7 +198,7 @@ my %globals;
 	if ($gas) {
 	    # Solaris /usr/ccs/bin/as can't handle multiplications
 	    # in $self->{value}
-	    $self->{value} =~ s/(?<![0-9a-f])(0[x0-9a-f]+)/oct($1)/egi;
+	    $self->{value} =~ s/(?<![\w\$\.])(0x?[0-9a-f]+)/oct($1)/egi;
 	    $self->{value} =~ s/([0-9]+\s*[\*\/\%]\s*[0-9]+)/eval($1)/eg;
 	    sprintf "\$%s",$self->{value};
 	} else {
@@ -243,7 +250,7 @@ my %globals;
 	    # Solaris /usr/ccs/bin/as can't handle multiplications
 	    # in $self->{label}, new gas requires sign extension...
 	    use integer;
-	    $self->{label} =~ s/(?<![0-9a-f])(0[x0-9a-f]+)/oct($1)/egi;
+	    $self->{label} =~ s/(?<![\w\$\.])(0x?[0-9a-f]+)/oct($1)/egi;
 	    $self->{label} =~ s/([0-9]+\s*[\*\/\%]\s*[0-9]+)/eval($1)/eg;
 	    $self->{label} =~ s/([0-9]+)/$1<<32>>32/eg;
 	    $self->{label} =~ s/^___imp_/__imp__/   if ($flavour eq "mingw64");
@@ -259,7 +266,7 @@ my %globals;
 	    %szmap = ( b=>"BYTE$PTR", w=>"WORD$PTR", l=>"DWORD$PTR", q=>"QWORD$PTR" );
 
 	    $self->{label} =~ s/\./\$/g;
-	    $self->{label} =~ s/0x([0-9a-f]+)/0$1h/ig;
+	    $self->{label} =~ s/(?<![\w\$\.])0x([0-9a-f]+)/0$1h/ig;
 	    $self->{label} = "($self->{label})" if ($self->{label} =~ /[\*\+\-\/]/);
 	    $sz="q" if ($self->{asterisk});
 
@@ -538,6 +545,8 @@ my %globals;
 					if ($line=~/\.([px])data/) {
 					    $v.=" rdata align=";
 					    $v.=$1 eq "p"? 4 : 8;
+					} elsif ($line=~/\.CRT\$/i) {
+					    $v.=" rdata align=8";
 					}
 				    } else {
 					$v="$current_segment\tENDS\n" if ($current_segment);
@@ -545,6 +554,8 @@ my %globals;
 					if ($line=~/\.([px])data/) {
 					    $v.=" READONLY";
 					    $v.=" ALIGN(".($1 eq "p" ? 4 : 8).")" if ($masm>=$masmref);
+					} elsif ($line=~/\.CRT\$/i) {
+					    $v.=" READONLY DWORD";
 					}
 				    }
 				    $current_segment = $line;
@@ -574,11 +585,11 @@ my %globals;
 		/\.align/   && do { $self->{value} = "ALIGN\t".$line; last; };
 		/\.(value|long|rva|quad)/
 			    && do { my $sz  = substr($1,0,1);
-				    my @arr = split(',',$line);
+				    my @arr = split(/,\s*/,$line);
 				    my $last = pop(@arr);
 				    my $conv = sub  {	my $var=shift;
 							$var=~s/^(0b[0-1]+)/oct($1)/eig;
-							$var=~s/0x([0-9a-f]+)/0$1h/ig if ($masm);
+							$var=~s/^0x([0-9a-f]+)/0$1h/ig if ($masm);
 							if ($sz eq "D" && ($current_segment=~/.[px]data/ || $dir eq ".rva"))
 							{ $var=~s/([_a-z\$\@][_a-z0-9\$\@]*)/$nasm?"$1 wrt ..imagebase":"imagerel $1"/egi; }
 							$var;
@@ -590,7 +601,7 @@ my %globals;
 				    $self->{value} .= &$conv($last);
 				    last;
 				  };
-		/\.byte/    && do { my @str=split(",",$line);
+		/\.byte/    && do { my @str=split(/,\s*/,$line);
 				    map(s/(0b[0-1]+)/oct($1)/eig,@str);
 				    map(s/0x([0-9a-f]+)/0$1h/ig,@str) if ($masm);	
 				    while ($#str>15) {
@@ -664,7 +675,7 @@ while($line=<>) {
 		$insn = $opcode->out($#args>=1?$args[$#args]->size():$sz);
 	    } else {
 		$insn = $opcode->out();
-		$insn .= $sz if (map($_->out() =~ /xmm|mmx/,@args));
+		$insn .= $sz if (map($_->out() =~ /x?mm/,@args));
 		@args = reverse(@args);
 		undef $sz if ($nasm && $opcode->mnemonic() eq "lea");
 	    }

crypto/pkcs12/p12_key.c

@@ -107,6 +107,7 @@ int PKCS12_key_gen_uni(unsigned char *pass, int passlen, unsigned char *salt,
 	unsigned char *B, *D, *I, *p, *Ai;
 	int Slen, Plen, Ilen, Ijlen;
 	int i, j, u, v;
+	int ret = 0;
 	BIGNUM *Ij, *Bpl1;	/* These hold Ij and B + 1 */
 	EVP_MD_CTX ctx;
 #ifdef  DEBUG_KEYGEN
@@ -144,10 +145,8 @@ int PKCS12_key_gen_uni(unsigned char *pass, int passlen, unsigned char *salt,
 	I = OPENSSL_malloc (Ilen);
 	Ij = BN_new();
 	Bpl1 = BN_new();
-	if (!D || !Ai || !B || !I || !Ij || !Bpl1) {
-		PKCS12err(PKCS12_F_PKCS12_KEY_GEN_UNI,ERR_R_MALLOC_FAILURE);
-		return 0;
-	}
+	if (!D || !Ai || !B || !I || !Ij || !Bpl1)
+		goto err;
 	for (i = 0; i < v; i++) D[i] = id;
 	p = I;
 	for (i = 0; i < Slen; i++) *p++ = salt[i % saltlen];
@@ -164,28 +163,22 @@ int PKCS12_key_gen_uni(unsigned char *pass, int passlen, unsigned char *salt,
 		}
 		memcpy (out, Ai, min (n, u));
 		if (u >= n) {
-			OPENSSL_free (Ai);
-			OPENSSL_free (B);
-			OPENSSL_free (D);
-			OPENSSL_free (I);
-			BN_free (Ij);
-			BN_free (Bpl1);
-			EVP_MD_CTX_cleanup(&ctx);
 #ifdef DEBUG_KEYGEN
 			fprintf(stderr, "Output KEY (length %d)\n", tmpn);
 			h__dump(tmpout, tmpn);
 #endif
-			return 1;	
+			ret = 1;
+			goto end;
 		}
 		n -= u;
 		out += u;
 		for (j = 0; j < v; j++) B[j] = Ai[j % u];
 		/* Work out B + 1 first then can use B as tmp space */
-		BN_bin2bn (B, v, Bpl1);
-		BN_add_word (Bpl1, 1);
+		if (!BN_bin2bn (B, v, Bpl1)) goto err;
+		if (!BN_add_word (Bpl1, 1)) goto err;
 		for (j = 0; j < Ilen ; j+=v) {
-			BN_bin2bn (I + j, v, Ij);
-			BN_add (Ij, Ij, Bpl1);
+			if (!BN_bin2bn (I + j, v, Ij)) goto err;
+			if (!BN_add (Ij, Ij, Bpl1)) goto err;
 			BN_bn2bin (Ij, B);
 			Ijlen = BN_num_bytes (Ij);
 			/* If more than 2^(v*8) - 1 cut off MSB */
@@ -201,6 +194,19 @@ int PKCS12_key_gen_uni(unsigned char *pass, int passlen, unsigned char *salt,
 			} else BN_bn2bin (Ij, I + j);
 		}
 	}
+
+err:
+	PKCS12err(PKCS12_F_PKCS12_KEY_GEN_UNI,ERR_R_MALLOC_FAILURE);
+
+end:
+	OPENSSL_free (Ai);
+	OPENSSL_free (B);
+	OPENSSL_free (D);
+	OPENSSL_free (I);
+	BN_free (Ij);
+	BN_free (Bpl1);
+	EVP_MD_CTX_cleanup(&ctx);
+	return ret;
 }
 #ifdef DEBUG_KEYGEN
 void h__dump (unsigned char *p, int len)

crypto/pkcs7/pk7_doit.c

@@ -422,7 +422,6 @@ BIO *PKCS7_dataDecode(PKCS7 *p7, EVP_PKEY *pkey, BIO *in_bio, X509 *pcert)
 	X509_ALGOR *enc_alg=NULL;
 	STACK_OF(X509_ALGOR) *md_sk=NULL;
 	STACK_OF(PKCS7_RECIP_INFO) *rsk=NULL;
-	X509_ALGOR *xalg=NULL;
 	PKCS7_RECIP_INFO *ri=NULL;
 
 	i=OBJ_obj2nid(p7->type);
@@ -445,7 +444,6 @@ BIO *PKCS7_dataDecode(PKCS7 *p7, EVP_PKEY *pkey, BIO *in_bio, X509 *pcert)
 			PKCS7err(PKCS7_F_PKCS7_DATADECODE,PKCS7_R_UNSUPPORTED_CIPHER_TYPE);
 			goto err;
 			}
-		xalg=p7->d.signed_and_enveloped->enc_data->algorithm;
 		break;
 	case NID_pkcs7_enveloped:
 		rsk=p7->d.enveloped->recipientinfo;
@@ -457,7 +455,6 @@ BIO *PKCS7_dataDecode(PKCS7 *p7, EVP_PKEY *pkey, BIO *in_bio, X509 *pcert)
 			PKCS7err(PKCS7_F_PKCS7_DATADECODE,PKCS7_R_UNSUPPORTED_CIPHER_TYPE);
 			goto err;
 			}
-		xalg=p7->d.enveloped->enc_data->algorithm;
 		break;
 	default:
 		PKCS7err(PKCS7_F_PKCS7_DATADECODE,PKCS7_R_UNSUPPORTED_CONTENT_TYPE);

crypto/pkcs7/pk7_lib.c

@@ -591,7 +591,6 @@ X509 *PKCS7_cert_from_signer_info(PKCS7 *p7, PKCS7_SIGNER_INFO *si)
 int PKCS7_set_cipher(PKCS7 *p7, const EVP_CIPHER *cipher)
 	{
 	int i;
-	ASN1_OBJECT *objtmp;
 	PKCS7_ENC_CONTENT *ec;
 
 	i=OBJ_obj2nid(p7->type);
@@ -614,7 +613,6 @@ int PKCS7_set_cipher(PKCS7 *p7, const EVP_CIPHER *cipher)
 		PKCS7err(PKCS7_F_PKCS7_SET_CIPHER,PKCS7_R_CIPHER_HAS_NO_OBJECT_IDENTIFIER);
 		return(0);
 	}
-	objtmp = OBJ_nid2obj(i);
 
 	ec->cipher = cipher;
 	return 1;

crypto/pqueue/pqueue.c

@@ -167,14 +167,13 @@ pqueue_pop(pqueue_s *pq)
 pitem *
 pqueue_find(pqueue_s *pq, unsigned char *prio64be)
 	{
-	pitem *next, *prev = NULL;
+	pitem *next;
 	pitem *found = NULL;
 
 	if ( pq->items == NULL)
 		return NULL;
 
-	for ( next = pq->items; next->next != NULL; 
-		  prev = next, next = next->next)
+	for ( next = pq->items; next->next != NULL; next = next->next)
 		{
 		if ( memcmp(next->priority, prio64be,8) == 0)
 			{

crypto/rand/md_rand.c

@@ -476,11 +476,14 @@ static int ssleay_rand_bytes(unsigned char *buf, int num)
 		MD_Update(&m,(unsigned char *)&(md_c[0]),sizeof(md_c));
 
 #ifndef PURIFY /* purify complains */
-		/* DO NOT REMOVE THE FOLLOWING CALL TO MD_Update()! */
+		/* The following line uses the supplied buffer as a small
+		 * source of entropy: since this buffer is often uninitialised
+		 * it may cause programs such as purify or valgrind to
+		 * complain. So for those builds it is not used: the removal
+		 * of such a small source of entropy has negligible impact on
+		 * security.
+		 */
 		MD_Update(&m,buf,j);
-		/* We know that line may cause programs such as
-		   purify and valgrind to complain about use of
-		   uninitialized data.  */
 #endif
 
 		k=(st_idx+MD_DIGEST_LENGTH/2)-st_num;

crypto/rand/rand_nw.c

@@ -160,8 +160,8 @@ int RAND_poll(void)
          rdtsc
          mov tsc, eax        
       }
-#else
-      asm volatile("rdtsc":"=A" (tsc));
+#elif defined(__GNUC__) && __GNUC__>=2 && !defined(OPENSSL_NO_ASM) && !defined(OPENSSL_NO_INLINE_ASM)
+      asm volatile("rdtsc":"=a"(tsc)::"edx");
 #endif
 
       RAND_add(&tsc, sizeof(tsc), 1);

crypto/rand/randfile.c

@@ -269,7 +269,6 @@ err:
 const char *RAND_file_name(char *buf, size_t size)
 	{
 	char *s=NULL;
-	int ok = 0;
 #ifdef __OpenBSD__
 	struct stat sb;
 #endif
@@ -298,7 +297,6 @@ const char *RAND_file_name(char *buf, size_t size)
 			BUF_strlcat(buf,"/",size);
 #endif
 			BUF_strlcat(buf,RFILE,size);
-			ok = 1;
 			}
 		else
 		  	buf[0] = '\0'; /* no file name */

crypto/rc5/rc5_locl.h

@@ -154,14 +154,14 @@
 #  define ROTATE_l32(a,n)	({ register unsigned int ret;	\
 					asm ("roll %%cl,%0"	\
 						: "=r"(ret)	\
-						: "c"(n),"0"(a)	\
+						: "c"(n),"0"((unsigned int)(a))	\
 						: "cc");	\
 					ret;			\
 				})
 #  define ROTATE_r32(a,n)	({ register unsigned int ret;	\
 					asm ("rorl %%cl,%0"	\
 						: "=r"(ret)	\
-						: "c"(n),"0"(a)	\
+						: "c"(n),"0"((unsigned int)(a))	\
 						: "cc");	\
 					ret;			\
 				})

crypto/rsa/rsa_eay.c

@@ -675,7 +675,7 @@ static int RSA_eay_public_decrypt(int flen, const unsigned char *from,
 		rsa->_method_mod_n)) goto err;
 
 	if ((padding == RSA_X931_PADDING) && ((ret->d[0] & 0xf) != 12))
-		BN_sub(ret, rsa->n, ret);
+		if (!BN_sub(ret, rsa->n, ret)) goto err;
 
 	p=buf;
 	i=BN_bn2bin(ret,p);

crypto/rsa/rsa_pmeth.c

@@ -246,6 +246,8 @@ static int pkey_rsa_verifyrecover(EVP_PKEY_CTX *ctx,
 			ret = int_rsa_verify(EVP_MD_type(rctx->md),
 						NULL, 0, rout, &sltmp,
 					sig, siglen, ctx->pkey->pkey.rsa);
+			if (ret <= 0)
+				return 0;
 			ret = sltmp;
 			}
 		else

crypto/sha/asm/sha1-armv4-large.pl

@@ -37,9 +37,18 @@
 #	modes are limited. As result it takes more instructions to do
 #	the same job in Thumb, therefore the code is never twice as
 #	small and always slower.
-# [***]	which is also ~35% better than compiler generated code.
+# [***]	which is also ~35% better than compiler generated code. Dual-
+#	issue Cortex A8 core was measured to process input block in
+#	~990 cycles.
 
-$output=shift;
+# August 2010.
+#
+# Rescheduling for dual-issue pipeline resulted in 13% improvement on
+# Cortex A8 core and in absolute terms ~870 cycles per input block
+# [or 13.6 cycles per byte].
+
+
+while (($output=shift) && ($output!~/^\w[\w\-]*\.\w+$/)) {}
 open STDOUT,">$output";
 
 $ctx="r0";
@@ -58,43 +67,22 @@ $t3="r12";
 $Xi="r14";
 @V=($a,$b,$c,$d,$e);
 
-# One can optimize this for aligned access on big-endian architecture,
-# but code's endian neutrality makes it too pretty:-)
-sub Xload {
-my ($a,$b,$c,$d,$e)=@_;
-$code.=<<___;
-	ldrb	$t0,[$inp],#4
-	ldrb	$t1,[$inp,#-3]
-	ldrb	$t2,[$inp,#-2]
-	ldrb	$t3,[$inp,#-1]
-	add	$e,$K,$e,ror#2			@ E+=K_00_19
-	orr	$t0,$t1,$t0,lsl#8
-	add	$e,$e,$a,ror#27			@ E+=ROR(A,27)
-	orr	$t0,$t2,$t0,lsl#8
-	eor	$t1,$c,$d			@ F_xx_xx
-	orr	$t0,$t3,$t0,lsl#8
-	add	$e,$e,$t0			@ E+=X[i]
-	str	$t0,[$Xi,#-4]!
-___
-}
 sub Xupdate {
-my ($a,$b,$c,$d,$e,$flag)=@_;
+my ($a,$b,$c,$d,$e,$opt1,$opt2)=@_;
 $code.=<<___;
 	ldr	$t0,[$Xi,#15*4]
 	ldr	$t1,[$Xi,#13*4]
 	ldr	$t2,[$Xi,#7*4]
-	ldr	$t3,[$Xi,#2*4]
 	add	$e,$K,$e,ror#2			@ E+=K_xx_xx
+	ldr	$t3,[$Xi,#2*4]
 	eor	$t0,$t0,$t1
-	eor	$t0,$t0,$t2
-	eor	$t0,$t0,$t3
-	add	$e,$e,$a,ror#27			@ E+=ROR(A,27)
-___
-$code.=<<___ if (!defined($flag));
-	eor	$t1,$c,$d			@ F_xx_xx, but not in 40_59
-___
-$code.=<<___;
+	eor	$t2,$t2,$t3
+	eor	$t1,$c,$d			@ F_xx_xx
 	mov	$t0,$t0,ror#31
+	add	$e,$e,$a,ror#27			@ E+=ROR(A,27)
+	eor	$t0,$t0,$t2,ror#31
+	$opt1					@ F_xx_xx
+	$opt2					@ F_xx_xx
 	add	$e,$e,$t0			@ E+=X[i]
 	str	$t0,[$Xi,#-4]!
 ___
@@ -102,19 +90,29 @@ ___
 
 sub BODY_00_15 {
 my ($a,$b,$c,$d,$e)=@_;
-	&Xload(@_);
 $code.=<<___;
+	ldrb	$t0,[$inp],#4
+	ldrb	$t1,[$inp,#-1]
+	ldrb	$t2,[$inp,#-2]
+	add	$e,$K,$e,ror#2			@ E+=K_00_19
+	ldrb	$t3,[$inp,#-3]
+	add	$e,$e,$a,ror#27			@ E+=ROR(A,27)
+	orr	$t0,$t1,$t0,lsl#24
+	eor	$t1,$c,$d			@ F_xx_xx
+	orr	$t0,$t0,$t2,lsl#8
+	orr	$t0,$t0,$t3,lsl#16
 	and	$t1,$b,$t1,ror#2
+	add	$e,$e,$t0			@ E+=X[i]
 	eor	$t1,$t1,$d,ror#2		@ F_00_19(B,C,D)
+	str	$t0,[$Xi,#-4]!
 	add	$e,$e,$t1			@ E+=F_00_19(B,C,D)
 ___
 }
 
 sub BODY_16_19 {
 my ($a,$b,$c,$d,$e)=@_;
-	&Xupdate(@_);
+	&Xupdate(@_,"and $t1,$b,$t1,ror#2");
 $code.=<<___;
-	and	$t1,$b,$t1,ror#2
 	eor	$t1,$t1,$d,ror#2		@ F_00_19(B,C,D)
 	add	$e,$e,$t1			@ E+=F_00_19(B,C,D)
 ___
@@ -122,22 +120,18 @@ ___
 
 sub BODY_20_39 {
 my ($a,$b,$c,$d,$e)=@_;
-	&Xupdate(@_);
+	&Xupdate(@_,"eor $t1,$b,$t1,ror#2");
 $code.=<<___;
-	eor	$t1,$b,$t1,ror#2		@ F_20_39(B,C,D)
 	add	$e,$e,$t1			@ E+=F_20_39(B,C,D)
 ___
 }
 
 sub BODY_40_59 {
 my ($a,$b,$c,$d,$e)=@_;
-	&Xupdate(@_,1);
+	&Xupdate(@_,"and $t1,$b,$t1,ror#2","and $t2,$c,$d");
 $code.=<<___;
-	and	$t1,$b,$c,ror#2
-	orr	$t2,$b,$c,ror#2
-	and	$t2,$t2,$d,ror#2
-	orr	$t1,$t1,$t2			@ F_40_59(B,C,D)
 	add	$e,$e,$t1			@ E+=F_40_59(B,C,D)
+	add	$e,$e,$t2,ror#2
 ___
 }
 

crypto/sha/asm/sha1-sparcv9.pl

@@ -276,6 +276,7 @@ $code.=<<___;
 .type	sha1_block_data_order,#function
 .size	sha1_block_data_order,(.-sha1_block_data_order)
 .asciz	"SHA1 block transform for SPARCv9, CRYPTOGAMS by <appro\@openssl.org>"
+.align	4
 ___
 
 $code =~ s/\`([^\`]*)\`/eval $1/gem;

crypto/sha/asm/sha1-sparcv9a.pl

@@ -539,6 +539,7 @@ $code.=<<___;
 .type	sha1_block_data_order,#function
 .size	sha1_block_data_order,(.-sha1_block_data_order)
 .asciz	"SHA1 block transform for SPARCv9a, CRYPTOGAMS by <appro\@openssl.org>"
+.align	4
 ___
 
 # Purpose of these subroutines is to explicitly encode VIS instructions,

crypto/sha/asm/sha256-armv4.pl

@@ -11,9 +11,14 @@
 
 # Performance is ~2x better than gcc 3.4 generated code and in "abso-
 # lute" terms is ~2250 cycles per 64-byte block or ~35 cycles per
-# byte.
+# byte [on single-issue Xscale PXA250 core].
 
-$output=shift;
+# July 2010.
+#
+# Rescheduling for dual-issue pipeline resulted in 22% improvement on
+# Cortex A8 core and ~20 cycles per processed byte.
+
+while (($output=shift) && ($output!~/^\w[\w\-]*\.\w+$/)) {}
 open STDOUT,">$output";
 
 $ctx="r0";	$t0="r0";
@@ -52,27 +57,27 @@ $code.=<<___ if ($i<16);
 ___
 $code.=<<___;
 	ldr	$t2,[$Ktbl],#4			@ *K256++
-	str	$T1,[sp,#`$i%16`*4]
 	mov	$t0,$e,ror#$Sigma1[0]
+	str	$T1,[sp,#`$i%16`*4]
 	eor	$t0,$t0,$e,ror#$Sigma1[1]
-	eor	$t0,$t0,$e,ror#$Sigma1[2]	@ Sigma1(e)
-	add	$T1,$T1,$t0
 	eor	$t1,$f,$g
+	eor	$t0,$t0,$e,ror#$Sigma1[2]	@ Sigma1(e)
 	and	$t1,$t1,$e
+	add	$T1,$T1,$t0
 	eor	$t1,$t1,$g			@ Ch(e,f,g)
-	add	$T1,$T1,$t1
 	add	$T1,$T1,$h
-	add	$T1,$T1,$t2
 	mov	$h,$a,ror#$Sigma0[0]
+	add	$T1,$T1,$t1
 	eor	$h,$h,$a,ror#$Sigma0[1]
+	add	$T1,$T1,$t2
 	eor	$h,$h,$a,ror#$Sigma0[2]		@ Sigma0(a)
 	orr	$t0,$a,$b
-	and	$t0,$t0,$c
 	and	$t1,$a,$b
-	orr	$t0,$t0,$t1			@ Maj(a,b,c)
-	add	$h,$h,$t0
-	add	$d,$d,$T1
+	and	$t0,$t0,$c
 	add	$h,$h,$T1
+	orr	$t0,$t0,$t1			@ Maj(a,b,c)
+	add	$d,$d,$T1
+	add	$h,$h,$t0
 ___
 }
 
@@ -80,19 +85,19 @@ sub BODY_16_XX {
 my ($i,$a,$b,$c,$d,$e,$f,$g,$h) = @_;
 
 $code.=<<___;
-	ldr	$t1,[sp,#`($i+1)%16`*4]	@ $i
+	ldr	$t1,[sp,#`($i+1)%16`*4]		@ $i
 	ldr	$t2,[sp,#`($i+14)%16`*4]
 	ldr	$T1,[sp,#`($i+0)%16`*4]
-	ldr	$inp,[sp,#`($i+9)%16`*4]
 	mov	$t0,$t1,ror#$sigma0[0]
+	ldr	$inp,[sp,#`($i+9)%16`*4]
 	eor	$t0,$t0,$t1,ror#$sigma0[1]
 	eor	$t0,$t0,$t1,lsr#$sigma0[2]	@ sigma0(X[i+1])
 	mov	$t1,$t2,ror#$sigma1[0]
-	eor	$t1,$t1,$t2,ror#$sigma1[1]
-	eor	$t1,$t1,$t2,lsr#$sigma1[2]	@ sigma1(X[i+14])
 	add	$T1,$T1,$t0
-	add	$T1,$T1,$t1
+	eor	$t1,$t1,$t2,ror#$sigma1[1]
 	add	$T1,$T1,$inp
+	eor	$t1,$t1,$t2,lsr#$sigma1[2]	@ sigma1(X[i+14])
+	add	$T1,$T1,$t1
 ___
 	&BODY_00_15(@_);
 }

crypto/sha/asm/sha512-armv4.pl

@@ -10,7 +10,13 @@
 # SHA512 block procedure for ARMv4. September 2007.
 
 # This code is ~4.5 (four and a half) times faster than code generated
-# by gcc 3.4 and it spends ~72 clock cycles per byte. 
+# by gcc 3.4 and it spends ~72 clock cycles per byte [on single-issue
+# Xscale PXA250 core].
+#
+# July 2010.
+#
+# Rescheduling for dual-issue pipeline resulted in 6% improvement on
+# Cortex A8 core and ~40 cycles per processed byte.
 
 # Byte order [in]dependence. =========================================
 #
@@ -22,7 +28,7 @@ $hi=0;
 $lo=4;
 # ====================================================================
 
-$output=shift;
+while (($output=shift) && ($output!~/^\w[\w\-]*\.\w+$/)) {}
 open STDOUT,">$output";
 
 $ctx="r0";
@@ -73,33 +79,31 @@ $code.=<<___;
 	eor	$t0,$t0,$Elo,lsl#23
 	eor	$t1,$t1,$Ehi,lsl#23	@ Sigma1(e)
 	adds	$Tlo,$Tlo,$t0
-	adc	$Thi,$Thi,$t1		@ T += Sigma1(e)
-	adds	$Tlo,$Tlo,$t2
-	adc	$Thi,$Thi,$t3		@ T += h
-
 	ldr	$t0,[sp,#$Foff+0]	@ f.lo
+	adc	$Thi,$Thi,$t1		@ T += Sigma1(e)
 	ldr	$t1,[sp,#$Foff+4]	@ f.hi
+	adds	$Tlo,$Tlo,$t2
 	ldr	$t2,[sp,#$Goff+0]	@ g.lo
+	adc	$Thi,$Thi,$t3		@ T += h
 	ldr	$t3,[sp,#$Goff+4]	@ g.hi
+
+	eor	$t0,$t0,$t2
 	str	$Elo,[sp,#$Eoff+0]
-	str	$Ehi,[sp,#$Eoff+4]
-	str	$Alo,[sp,#$Aoff+0]
-	str	$Ahi,[sp,#$Aoff+4]
-
-	eor	$t0,$t0,$t2
 	eor	$t1,$t1,$t3
+	str	$Ehi,[sp,#$Eoff+4]
 	and	$t0,$t0,$Elo
+	str	$Alo,[sp,#$Aoff+0]
 	and	$t1,$t1,$Ehi
+	str	$Ahi,[sp,#$Aoff+4]
 	eor	$t0,$t0,$t2
-	eor	$t1,$t1,$t3		@ Ch(e,f,g)
-
 	ldr	$t2,[$Ktbl,#4]		@ K[i].lo
+	eor	$t1,$t1,$t3		@ Ch(e,f,g)
 	ldr	$t3,[$Ktbl,#0]		@ K[i].hi
-	ldr	$Elo,[sp,#$Doff+0]	@ d.lo
-	ldr	$Ehi,[sp,#$Doff+4]	@ d.hi
 
 	adds	$Tlo,$Tlo,$t0
+	ldr	$Elo,[sp,#$Doff+0]	@ d.lo
 	adc	$Thi,$Thi,$t1		@ T += Ch(e,f,g)
+	ldr	$Ehi,[sp,#$Doff+4]	@ d.hi
 	adds	$Tlo,$Tlo,$t2
 	adc	$Thi,$Thi,$t3		@ T += K[i]
 	adds	$Elo,$Elo,$Tlo

crypto/sha/asm/sha512-sparcv9.pl

@@ -586,6 +586,7 @@ $code.=<<___;
 .type	sha${label}_block_data_order,#function
 .size	sha${label}_block_data_order,(.-sha${label}_block_data_order)
 .asciz	"SHA${label} block transform for SPARCv9, CRYPTOGAMS by <appro\@openssl.org>"
+.align	4
 ___
 
 $code =~ s/\`([^\`]*)\`/eval $1/gem;

crypto/sparccpuid.S

@@ -179,7 +179,7 @@ OPENSSL_atomic_add:
 	ba	.enter
 	nop
 #ifdef __sun
-! Note that you don't have to link with libthread to call thr_yield,
+! Note that you do not have to link with libthread to call thr_yield,
 ! as libc provides a stub, which is overloaded the moment you link
 ! with *either* libpthread or libthread...
 #define	YIELD_CPU	thr_yield
@@ -225,13 +225,95 @@ _sparcv9_rdtick:
 	xor	%o0,%o0,%o0
 	.word	0x91410000	!rd	%tick,%o0
 	retl
-	.word	0x93323020	!srlx	%o2,32,%o1
+	.word	0x93323020	!srlx	%o0,32,%o1
 .notick:
 	retl
 	xor	%o1,%o1,%o1
 .type	_sparcv9_rdtick,#function
 .size	_sparcv9_rdtick,.-_sparcv9_rdtick
 
+.global	_sparcv9_vis1_probe
+.align	8
+_sparcv9_vis1_probe:
+	.word	0x81b00d80	!fxor	%f0,%f0,%f0
+	add	%sp,BIAS+2,%o1
+	retl
+	.word	0xc19a5a40	!ldda	[%o1]ASI_FP16_P,%f0
+.type	_sparcv9_vis1_probe,#function
+.size	_sparcv9_vis1_probe,.-_sparcv9_vis1_probe
+
+! Probe and instrument VIS1 instruction. Output is number of cycles it
+! takes to execute rdtick and pair of VIS1 instructions. US-Tx VIS unit
+! is slow (documented to be 6 cycles on T2) and the core is in-order
+! single-issue, it should be possible to distinguish Tx reliably...
+! Observed return values are:
+!
+!	UltraSPARC IIe		7
+!	UltraSPARC III		7
+!	UltraSPARC T1		24
+!
+! Numbers for T2 and SPARC64 V-VII are more than welcomed.
+!
+! It would be possible to detect specifically US-T1 by instrumenting
+! fmul8ulx16, which is emulated on T1 and as such accounts for quite
+! a lot of %tick-s, couple of thousand on Linux...
+.global	_sparcv9_vis1_instrument
+.align	8
+_sparcv9_vis1_instrument:
+	.word	0x91410000	!rd	%tick,%o0
+	.word	0x81b00d80	!fxor	%f0,%f0,%f0
+	.word	0x85b08d82	!fxor	%f2,%f2,%f2
+	.word	0x93410000	!rd	%tick,%o1
+	.word	0x81b00d80	!fxor	%f0,%f0,%f0
+	.word	0x85b08d82	!fxor	%f2,%f2,%f2
+	.word	0x95410000	!rd	%tick,%o2
+	.word	0x81b00d80	!fxor	%f0,%f0,%f0
+	.word	0x85b08d82	!fxor	%f2,%f2,%f2
+	.word	0x97410000	!rd	%tick,%o3
+	.word	0x81b00d80	!fxor	%f0,%f0,%f0
+	.word	0x85b08d82	!fxor	%f2,%f2,%f2
+	.word	0x99410000	!rd	%tick,%o4
+
+	! calculate intervals
+	sub	%o1,%o0,%o0
+	sub	%o2,%o1,%o1
+	sub	%o3,%o2,%o2
+	sub	%o4,%o3,%o3
+
+	! find minumum value
+	cmp	%o0,%o1
+	.word	0x38680002	!bgu,a	%xcc,.+8
+	mov	%o1,%o0
+	cmp	%o0,%o2
+	.word	0x38680002	!bgu,a	%xcc,.+8
+	mov	%o2,%o0
+	cmp	%o0,%o3
+	.word	0x38680002	!bgu,a	%xcc,.+8
+	mov	%o3,%o0
+
+	retl
+	nop
+.type	_sparcv9_vis1_instrument,#function
+.size	_sparcv9_vis1_instrument,.-_sparcv9_vis1_instrument
+
+.global	_sparcv9_vis2_probe
+.align	8
+_sparcv9_vis2_probe:
+	retl
+	.word	0x81b00980	!bshuffle	%f0,%f0,%f0
+.type	_sparcv9_vis2_probe,#function
+.size	_sparcv9_vis2_probe,.-_sparcv9_vis2_probe
+
+.global	_sparcv9_fmadd_probe
+.align	8
+_sparcv9_fmadd_probe:
+	.word	0x81b00d80	!fxor	%f0,%f0,%f0
+	.word	0x85b08d82	!fxor	%f2,%f2,%f2
+	retl
+	.word	0x81b80440	!fmaddd	%f0,%f0,%f2,%f0
+.type	_sparcv9_fmadd_probe,#function
+.size	_sparcv9_fmadd_probe,.-_sparcv9_fmadd_probe
+
 .global	OPENSSL_cleanse
 .align	32
 OPENSSL_cleanse:

crypto/sparcv9cap.c

@@ -1,6 +1,8 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
+#include <setjmp.h>
+#include <signal.h>
 #include <sys/time.h>
 #include <openssl/bn.h>
 
@@ -9,6 +11,7 @@
 #define SPARCV9_VIS1		(1<<2)
 #define SPARCV9_VIS2		(1<<3)	/* reserved */
 #define SPARCV9_FMADD		(1<<4)	/* reserved for SPARC64 V */
+
 static int OPENSSL_sparcv9cap_P=SPARCV9_TICK_PRIVILEGED;
 
 int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, const BN_ULONG *np,const BN_ULONG *n0, int num)
@@ -23,10 +26,14 @@ int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, const BN_U
 		return bn_mul_mont_int(rp,ap,bp,np,n0,num);
 	}
 
+unsigned long	_sparcv9_rdtick(void);
+void		_sparcv9_vis1_probe(void);
+unsigned long	_sparcv9_vis1_instrument(void);
+void		_sparcv9_vis2_probe(void);
+void		_sparcv9_fmadd_probe(void);
+
 unsigned long OPENSSL_rdtsc(void)
 	{
-	unsigned long _sparcv9_rdtick(void);
-
 	if (OPENSSL_sparcv9cap_P&SPARCV9_TICK_PRIVILEGED)
 #if defined(__sun) && defined(__SVR4)
 		return gethrtime();
@@ -37,8 +44,11 @@ unsigned long OPENSSL_rdtsc(void)
 		return _sparcv9_rdtick();
 	}
 
-#if defined(__sun) && defined(__SVR4)
-
+#if 0 && defined(__sun) && defined(__SVR4)
+/* This code path is disabled, because of incompatibility of
+ * libdevinfo.so.1 and libmalloc.so.1 (see below for details)
+ */
+#include <malloc.h>
 #include <dlfcn.h>
 #include <libdevinfo.h>
 #include <sys/systeminfo.h>
@@ -110,7 +120,21 @@ void OPENSSL_cpuid_setup(void)
 			return;
 			}
 		}
-
+#ifdef M_KEEP
+	/*
+	 * Solaris libdevinfo.so.1 is effectively incomatible with
+	 * libmalloc.so.1. Specifically, if application is linked with
+	 * -lmalloc, it crashes upon startup with SIGSEGV in
+	 * free(3LIBMALLOC) called by di_fini. Prior call to
+	 * mallopt(M_KEEP,0) somehow helps... But not always...
+	 */
+	if ((h = dlopen(NULL,RTLD_LAZY)))
+		{
+		union { void *p; int (*f)(int,int); } sym;
+		if ((sym.p = dlsym(h,"mallopt"))) (*sym.f)(M_KEEP,0);
+		dlclose(h);
+		}
+#endif
 	if ((h = dlopen("libdevinfo.so.1",RTLD_LAZY))) do
 		{
 		di_init_t	di_init;
@@ -137,9 +161,19 @@ void OPENSSL_cpuid_setup(void)
 
 #else
 
+static sigjmp_buf common_jmp;
+static void common_handler(int sig) { siglongjmp(common_jmp,sig); }
+
 void OPENSSL_cpuid_setup(void)
 	{
 	char *e;
+	struct sigaction	common_act,ill_oact,bus_oact;
+	sigset_t		all_masked,oset;
+	int			sig;
+	static int trigger=0;
+
+	if (trigger) return;
+	trigger=1;
  
 	if ((e=getenv("OPENSSL_sparcv9cap")))
 		{
@@ -147,8 +181,57 @@ void OPENSSL_cpuid_setup(void)
 		return;
 		}
 
-	/* For now we assume that the rest supports UltraSPARC-I* only */
-	OPENSSL_sparcv9cap_P |= SPARCV9_PREFER_FPU|SPARCV9_VIS1;
+	/* Initial value, fits UltraSPARC-I&II... */
+	OPENSSL_sparcv9cap_P = SPARCV9_PREFER_FPU|SPARCV9_TICK_PRIVILEGED;
+
+	sigfillset(&all_masked);
+	sigdelset(&all_masked,SIGILL);
+	sigdelset(&all_masked,SIGTRAP);
+#ifdef SIGEMT
+	sigdelset(&all_masked,SIGEMT);
+#endif
+	sigdelset(&all_masked,SIGFPE);
+	sigdelset(&all_masked,SIGBUS);
+	sigdelset(&all_masked,SIGSEGV);
+	sigprocmask(SIG_SETMASK,&all_masked,&oset);
+
+	memset(&common_act,0,sizeof(common_act));
+	common_act.sa_handler = common_handler;
+	common_act.sa_mask    = all_masked;
+
+	sigaction(SIGILL,&common_act,&ill_oact);
+	sigaction(SIGBUS,&common_act,&bus_oact);/* T1 fails 16-bit ldda [on Linux] */
+
+	if (sigsetjmp(common_jmp,1) == 0)
+		{
+		_sparcv9_rdtick();
+		OPENSSL_sparcv9cap_P &= ~SPARCV9_TICK_PRIVILEGED;
+		}
+
+	if (sigsetjmp(common_jmp,1) == 0)
+		{
+		_sparcv9_vis1_probe();
+		OPENSSL_sparcv9cap_P |= SPARCV9_VIS1;
+		/* detect UltraSPARC-Tx, see sparccpud.S for details... */
+		if (_sparcv9_vis1_instrument() >= 12)
+			OPENSSL_sparcv9cap_P &= ~(SPARCV9_VIS1|SPARCV9_PREFER_FPU);
+		else
+			{
+			_sparcv9_vis2_probe();
+			OPENSSL_sparcv9cap_P |= SPARCV9_VIS2;
+			}
+		}
+
+	if (sigsetjmp(common_jmp,1) == 0)
+		{
+		_sparcv9_fmadd_probe();
+		OPENSSL_sparcv9cap_P |= SPARCV9_FMADD;
+		}
+
+	sigaction(SIGBUS,&bus_oact,NULL);
+	sigaction(SIGILL,&ill_oact,NULL);
+
+	sigprocmask(SIG_SETMASK,&oset,NULL);
 	}
 
 #endif