openssl.git
7 years agosha256t.c: make sure unrolled loop is tested.
Andy Polyakov [Tue, 12 Jun 2012 14:40:41 +0000 (14:40 +0000)]
sha256t.c: make sure unrolled loop is tested.

7 years agosha256-586.pl: add AVX and XOP code paths.
Andy Polyakov [Tue, 12 Jun 2012 14:40:11 +0000 (14:40 +0000)]
sha256-586.pl: add AVX and XOP code paths.

7 years agosha256-586.pl: squeeze some more, most notably ~10% on Nehalem.
Andy Polyakov [Tue, 12 Jun 2012 14:38:01 +0000 (14:38 +0000)]
sha256-586.pl: squeeze some more, most notably ~10% on Nehalem.

7 years agox86[_64] assembly pack: update benchmark results.
Andy Polyakov [Tue, 12 Jun 2012 14:18:21 +0000 (14:18 +0000)]
x86[_64] assembly pack: update benchmark results.

7 years agoprint out issuer and subject unique identifier fields in certificates
Dr. Stephen Henson [Tue, 12 Jun 2012 13:41:18 +0000 (13:41 +0000)]
print out issuer and subject unique identifier fields in certificates

7 years agobss_dgram.c: add BIO_CTRL_DGRAM_SET_DONT_FRAG.
Andy Polyakov [Mon, 11 Jun 2012 14:56:25 +0000 (14:56 +0000)]
bss_dgram.c: add BIO_CTRL_DGRAM_SET_DONT_FRAG.

PR: 2830
Submitted by: Robin Seggelmann

7 years agobss_dgram.c: make getsockopt work in cases when optlen is 64-bit value.
Andy Polyakov [Mon, 11 Jun 2012 14:27:56 +0000 (14:27 +0000)]
bss_dgram.c: make getsockopt work in cases when optlen is 64-bit value.

7 years agoFix memory leak.
Ben Laurie [Mon, 11 Jun 2012 09:23:55 +0000 (09:23 +0000)]
Fix memory leak.

7 years agob_sock.c: make getsockopt work in cases when optlen is 64-bit value.
Andy Polyakov [Mon, 11 Jun 2012 08:52:11 +0000 (08:52 +0000)]
b_sock.c: make getsockopt work in cases when optlen is 64-bit value.

7 years agoRearrange and test authz extension.
Ben Laurie [Thu, 7 Jun 2012 13:20:47 +0000 (13:20 +0000)]
Rearrange and test authz extension.

7 years agoFix memory leak.
Ben Laurie [Wed, 6 Jun 2012 13:03:42 +0000 (13:03 +0000)]
Fix memory leak.

7 years agoParse authz correctly.
Ben Laurie [Wed, 6 Jun 2012 12:52:19 +0000 (12:52 +0000)]
Parse authz correctly.

7 years agoRevert random changes from commit#22606.
Andy Polyakov [Mon, 4 Jun 2012 22:12:10 +0000 (22:12 +0000)]
Revert random changes from commit#22606.

7 years agoVersion skew reduction: trivia (I hope).
Ben Laurie [Sun, 3 Jun 2012 22:00:21 +0000 (22:00 +0000)]
Version skew reduction: trivia (I hope).

7 years agoRFC 5878 support.
Ben Laurie [Wed, 30 May 2012 10:10:58 +0000 (10:10 +0000)]
RFC 5878 support.

7 years agoBuild on FreeBSD with gcc 4.6.
Ben Laurie [Wed, 30 May 2012 09:34:44 +0000 (09:34 +0000)]
Build on FreeBSD with gcc 4.6.

7 years agosha256-586.pl: full unroll to deliver additional ~16%, add Sandy Bridge-
Andy Polyakov [Mon, 28 May 2012 17:50:57 +0000 (17:50 +0000)]
sha256-586.pl: full unroll to deliver additional ~16%, add Sandy Bridge-
specific code path.

7 years agosha512-x86_64.pl: >5% better performance.
Andy Polyakov [Mon, 28 May 2012 17:47:15 +0000 (17:47 +0000)]
sha512-x86_64.pl: >5% better performance.

7 years agoaesni-x86_64.pl: make it possibel to use in Linux kernel.
Andy Polyakov [Thu, 24 May 2012 07:39:44 +0000 (07:39 +0000)]
aesni-x86_64.pl: make it possibel to use in Linux kernel.

7 years agosha256-586.pl: tune away regression on Nehalem core and incidentally
Andy Polyakov [Thu, 24 May 2012 07:39:04 +0000 (07:39 +0000)]
sha256-586.pl: tune away regression on Nehalem core and incidentally
improve performance on Atom and P4.

7 years agosha256-586.pl optimization.
Andy Polyakov [Sat, 19 May 2012 10:10:30 +0000 (10:10 +0000)]
sha256-586.pl optimization.

7 years agos2_clnt.c: compensate for compiler bug.
Andy Polyakov [Wed, 16 May 2012 12:47:36 +0000 (12:47 +0000)]
s2_clnt.c: compensate for compiler bug.

7 years agoppccap.c: assume no features under 32-bit AIX kernel.
Andy Polyakov [Wed, 16 May 2012 12:42:32 +0000 (12:42 +0000)]
ppccap.c: assume no features under 32-bit AIX kernel.

PR: 2810

7 years agoPR: 2813
Dr. Stephen Henson [Fri, 11 May 2012 13:53:37 +0000 (13:53 +0000)]
PR: 2813
Reported by: Constantine Sapuntzakis <csapuntz@gmail.com>

Fix possible deadlock when decoding public keys.

7 years agoPR: 2811
Dr. Stephen Henson [Fri, 11 May 2012 13:34:29 +0000 (13:34 +0000)]
PR: 2811
Reported by: Phil Pennock <openssl-dev@spodhuis.org>

Make renegotiation work for TLS 1.2, 1.1 by not using a lower record
version client hello workaround if renegotiating.

7 years agoFix warning.
Ben Laurie [Thu, 10 May 2012 20:29:00 +0000 (20:29 +0000)]
Fix warning.

7 years agoPadlock doesn't build. I don't even know what it is.
Ben Laurie [Thu, 10 May 2012 20:28:02 +0000 (20:28 +0000)]
Padlock doesn't build. I don't even know what it is.

7 years agoPR: 2806
Dr. Stephen Henson [Thu, 10 May 2012 18:25:39 +0000 (18:25 +0000)]
PR: 2806
Submitted by: PK <runningdoglackey@yahoo.com>

Correct ciphersuite signature algorithm definitions.

7 years agoSanity check record length before skipping explicit IV in TLS 1.2, 1.1 and
Dr. Stephen Henson [Thu, 10 May 2012 16:03:52 +0000 (16:03 +0000)]
Sanity check record length before skipping explicit IV in TLS 1.2, 1.1 and
DTLS to fix DoS attack.

Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
fuzzing as a service testing platform.
(CVE-2012-2333)

7 years agoupdate FAQ
Dr. Stephen Henson [Thu, 10 May 2012 15:37:16 +0000 (15:37 +0000)]
update FAQ

7 years agoReported by: Solar Designer of Openwall
Dr. Stephen Henson [Thu, 10 May 2012 13:46:09 +0000 (13:46 +0000)]
Reported by: Solar Designer of Openwall

Make sure tkeylen is initialised properly when encrypting CMS messages.

7 years agoCorrect environment variable is OPENSSL_ALLOW_PROXY_CERTS.
Richard Levitte [Fri, 4 May 2012 10:43:15 +0000 (10:43 +0000)]
Correct environment variable is OPENSSL_ALLOW_PROXY_CERTS.

7 years agoperlasm: fix symptom-less bugs, missing semicolons and 'my' declarations.
Andy Polyakov [Sat, 28 Apr 2012 10:36:58 +0000 (10:36 +0000)]
perlasm: fix symptom-less bugs, missing semicolons and 'my' declarations.

7 years agoppccpuid.pl: branch hints in OPENSSL_cleanse impact small block performance
Andy Polyakov [Fri, 27 Apr 2012 20:17:45 +0000 (20:17 +0000)]
ppccpuid.pl: branch hints in OPENSSL_cleanse impact small block performance
of digest algorithms, mosty SHA, on Power7. Mystery of century, why SHA,
why slower algorithm are affected more...
PR: 2794
Submitted by: Ashley Lai

7 years agoDon't try to use unvalidated composite ciphers in FIPS mode
Dr. Stephen Henson [Thu, 26 Apr 2012 18:55:01 +0000 (18:55 +0000)]
Don't try to use unvalidated composite ciphers in FIPS mode

7 years agoupdate NEWS
Dr. Stephen Henson [Thu, 26 Apr 2012 11:13:30 +0000 (11:13 +0000)]
update NEWS

7 years agoupdate FAQ
Dr. Stephen Henson [Thu, 26 Apr 2012 11:10:24 +0000 (11:10 +0000)]
update FAQ

7 years agoCHANGES: clarify.
Andy Polyakov [Thu, 26 Apr 2012 07:33:26 +0000 (07:33 +0000)]
CHANGES: clarify.

7 years agoCHANGES: fix typos and clarify.
Andy Polyakov [Thu, 26 Apr 2012 07:20:06 +0000 (07:20 +0000)]
CHANGES: fix typos and clarify.

7 years agoChange value of SSL_OP_NO_TLSv1_1 to avoid clash with SSL_OP_ALL and
Dr. Stephen Henson [Wed, 25 Apr 2012 23:04:42 +0000 (23:04 +0000)]
Change value of SSL_OP_NO_TLSv1_1 to avoid clash with SSL_OP_ALL and
OpenSSL 1.0.0. Add CHANGES entry noting the consequences.

7 years agos23_clnt.c: ensure interoperability by maitaining client "version capability"
Andy Polyakov [Wed, 25 Apr 2012 22:06:32 +0000 (22:06 +0000)]
s23_clnt.c: ensure interoperability by maitaining client "version capability"
vector contiguous.
PR: 2802

7 years agoSubmitted by: Peter Sylvester <peter.sylvester@edelweb.fr>
Dr. Stephen Henson [Tue, 24 Apr 2012 12:22:23 +0000 (12:22 +0000)]
Submitted by: Peter Sylvester <peter.sylvester@edelweb.fr>
Reviewed by: steve
Improved localisation of TLS extension handling and code tidy.

7 years agooops, not yet ;-)
Dr. Stephen Henson [Mon, 23 Apr 2012 21:58:29 +0000 (21:58 +0000)]
oops, not yet ;-)

7 years agoupdate NEWS
Dr. Stephen Henson [Mon, 23 Apr 2012 21:56:33 +0000 (21:56 +0000)]
update NEWS

7 years agoobjxref.pl: improve portability.
Andy Polyakov [Sun, 22 Apr 2012 21:18:30 +0000 (21:18 +0000)]
objxref.pl: improve portability.

7 years agocorrect error code
Dr. Stephen Henson [Sun, 22 Apr 2012 13:31:09 +0000 (13:31 +0000)]
correct error code

7 years agocheck correctness of errors before updating them so we don't get bogus errors added
Dr. Stephen Henson [Sun, 22 Apr 2012 13:25:51 +0000 (13:25 +0000)]
check correctness of errors before updating them so we don't get bogus errors added

7 years agocorrect old FAQ answers
Dr. Stephen Henson [Sun, 22 Apr 2012 13:20:28 +0000 (13:20 +0000)]
correct old FAQ answers

7 years agoPR: 2239
Dr. Stephen Henson [Sun, 22 Apr 2012 13:06:51 +0000 (13:06 +0000)]
PR: 2239
Submitted by: Dominik Oepen <oepen@informatik.hu-berlin.de>

Add Brainpool curves from RFC5639.

Original patch by Annie Yousar <a.yousar@informatik.hu-berlin.de>

7 years agoe_rc4_hmac_md5.c: last commit was inappropriate for non-x86[_64] platforms.
Andy Polyakov [Thu, 19 Apr 2012 20:38:05 +0000 (20:38 +0000)]
e_rc4_hmac_md5.c: last commit was inappropriate for non-x86[_64] platforms.
PR: 2792

7 years agoCheck for potentially exploitable overflows in asn1_d2i_read_bio
Dr. Stephen Henson [Thu, 19 Apr 2012 16:19:56 +0000 (16:19 +0000)]
Check for potentially exploitable overflows in asn1_d2i_read_bio
BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
in CRYPTO_realloc_clean.

Thanks to Tavis Ormandy, Google Security Team, for discovering this
issue and to Adam Langley <agl@chromium.org> for fixing it. (CVE-2012-2110)

7 years agoupdate FAQ
Dr. Stephen Henson [Thu, 19 Apr 2012 12:33:23 +0000 (12:33 +0000)]
update FAQ

7 years agoMakefile.org: clear yet another environment variable.
Andy Polyakov [Thu, 19 Apr 2012 06:39:40 +0000 (06:39 +0000)]
Makefile.org: clear yet another environment variable.
PR: 2793

7 years agorecognise X9.42 DH certificates on servers
Dr. Stephen Henson [Wed, 18 Apr 2012 17:03:29 +0000 (17:03 +0000)]
recognise X9.42 DH certificates on servers

7 years agocorrect error codes
Dr. Stephen Henson [Wed, 18 Apr 2012 15:36:12 +0000 (15:36 +0000)]
correct error codes

7 years agoe_rc4_hmac_md5.c: harmonize zero-length fragment handling with
Andy Polyakov [Wed, 18 Apr 2012 14:55:39 +0000 (14:55 +0000)]
e_rc4_hmac_md5.c: harmonize zero-length fragment handling with
e_aes_cbc_hmac_sha1.c (mostly for aesthetic reasons).

7 years agoe_rc4_hmac_md5.c: oops, can't use rc4_hmac_md5_cipher on legacy Intel CPUs.
Andy Polyakov [Wed, 18 Apr 2012 14:50:28 +0000 (14:50 +0000)]
e_rc4_hmac_md5.c: oops, can't use rc4_hmac_md5_cipher on legacy Intel CPUs.

PR: 2792

7 years agoC64x+ assembler pack. linux-c64xplus build is *not* tested nor can it be
Andy Polyakov [Wed, 18 Apr 2012 13:01:36 +0000 (13:01 +0000)]
C64x+ assembler pack. linux-c64xplus build is *not* tested nor can it be
tested, because kernel is not in shape to handle it *yet*. The code is
committed mostly to stimulate the kernel development.

7 years agoDisable SHA-2 ciphersuites in < TLS 1.2 connections.
Bodo Möller [Tue, 17 Apr 2012 15:23:03 +0000 (15:23 +0000)]
Disable SHA-2 ciphersuites in < TLS 1.2 connections.

(TLS 1.2 clients could end up negotiating these with an OpenSSL server
with TLS 1.2 disabled, which is problematic.)

Submitted by: Adam Langley

7 years agoAdditional workaround for PR#2771
Dr. Stephen Henson [Tue, 17 Apr 2012 15:12:09 +0000 (15:12 +0000)]
Additional workaround for PR#2771

If OPENSSL_MAX_TLS1_2_CIPHER_LENGTH is set then limit the size of client
ciphersuites to this value. A value of 50 should be sufficient.

Document workarounds in CHANGES.

7 years agoPartial workaround for PR#2771.
Dr. Stephen Henson [Tue, 17 Apr 2012 13:21:19 +0000 (13:21 +0000)]
Partial workaround for PR#2771.

Some servers hang when presented with a client hello record length exceeding
255 bytes but will work with longer client hellos if the TLS record version
in client hello does not exceed TLS v1.0. Unfortunately this doesn't fix all
cases...

7 years agoOPENSSL_NO_SOCK fixes.
Andy Polyakov [Mon, 16 Apr 2012 17:42:36 +0000 (17:42 +0000)]
OPENSSL_NO_SOCK fixes.
PR: 2791
Submitted by: Ben Noordhuis

7 years agoMinor compatibility fixes.
Andy Polyakov [Mon, 16 Apr 2012 17:35:30 +0000 (17:35 +0000)]
Minor compatibility fixes.
PR: 2790
Submitted by: Alexei Khlebnikov

7 years agos3_srvr.c: fix typo.
Andy Polyakov [Sun, 15 Apr 2012 17:22:57 +0000 (17:22 +0000)]
s3_srvr.c: fix typo.
PR: 2538

7 years agoe_aes_cbc_hmac_sha1.c: handle zero-length payload and engage empty frag
Andy Polyakov [Sun, 15 Apr 2012 14:14:22 +0000 (14:14 +0000)]
e_aes_cbc_hmac_sha1.c: handle zero-length payload and engage empty frag
countermeasure.

PR: 2778

7 years agoghash-s390x.pl: fix typo [that can induce SEGV in 31-bit build].
Andy Polyakov [Thu, 12 Apr 2012 06:44:34 +0000 (06:44 +0000)]
ghash-s390x.pl: fix typo [that can induce SEGV in 31-bit build].

7 years agoAdd options to set additional type specific certificate chains to
Dr. Stephen Henson [Wed, 11 Apr 2012 16:53:11 +0000 (16:53 +0000)]
Add options to set additional type specific certificate chains to
s_server.

7 years agouse different variable for chain iteration
Dr. Stephen Henson [Wed, 11 Apr 2012 16:01:08 +0000 (16:01 +0000)]
use different variable for chain iteration

7 years agofix reset fix
Dr. Stephen Henson [Wed, 11 Apr 2012 15:05:07 +0000 (15:05 +0000)]
fix reset fix

7 years agomake reinitialisation work for CMAC
Dr. Stephen Henson [Wed, 11 Apr 2012 12:26:41 +0000 (12:26 +0000)]
make reinitialisation work for CMAC

7 years agoupdate rather ancient EVP digest documentation
Dr. Stephen Henson [Tue, 10 Apr 2012 22:28:01 +0000 (22:28 +0000)]
update rather ancient EVP digest documentation

7 years agoaes-s390x.pl: fix crash in AES_set_decrypt_key in linux32-s390x build.
Andy Polyakov [Mon, 9 Apr 2012 15:12:13 +0000 (15:12 +0000)]
aes-s390x.pl: fix crash in AES_set_decrypt_key in linux32-s390x build.

7 years agoSubmitted by: Peter Sylvester <peter.sylvester@edelweb.fr>
Dr. Stephen Henson [Fri, 6 Apr 2012 11:18:40 +0000 (11:18 +0000)]
Submitted by: Peter Sylvester <peter.sylvester@edelweb.fr>

Localize client hello extension parsing in t1_lib.c

7 years agoAdd support for automatic ECDH temporary key parameter selection. When
Dr. Stephen Henson [Thu, 5 Apr 2012 13:38:27 +0000 (13:38 +0000)]
Add support for automatic ECDH temporary key parameter selection. When
enabled instead of requiring an application to hard code a (possibly
inappropriate) parameter set and delve into EC internals we just
automatically use the preferred curve.

7 years agoadd FAQ about version numbers
Dr. Stephen Henson [Thu, 5 Apr 2012 13:16:37 +0000 (13:16 +0000)]
add FAQ about version numbers

7 years agoaes-armv4.pl: make it more foolproof [inspired by aes-s390x.pl in 1.0.1].
Andy Polyakov [Thu, 5 Apr 2012 08:30:22 +0000 (08:30 +0000)]
aes-armv4.pl: make it more foolproof [inspired by aes-s390x.pl in 1.0.1].

7 years agoaes-s390x.pl: make it more foolproof [inspired by 1.0.1].
Andy Polyakov [Thu, 5 Apr 2012 08:22:09 +0000 (08:22 +0000)]
aes-s390x.pl: make it more foolproof [inspired by 1.0.1].

7 years agossl/ssl_ciph.c: interim solution for assertion in d1_pkt.c(444).
Andy Polyakov [Wed, 4 Apr 2012 20:45:51 +0000 (20:45 +0000)]
ssl/ssl_ciph.c: interim solution for assertion in d1_pkt.c(444).

PR: 2778

7 years agoTidy up EC parameter check code: instead of accessing internal structures
Dr. Stephen Henson [Wed, 4 Apr 2012 14:41:01 +0000 (14:41 +0000)]
Tidy up EC parameter check code: instead of accessing internal structures
add utility functions to t1_lib.c to check if EC certificates and parameters
are consistent with peer.

7 years agoUpdate ordinals.
Dr. Stephen Henson [Tue, 3 Apr 2012 23:13:23 +0000 (23:13 +0000)]
Update ordinals.

7 years agoCHANGES: harmonize with 1.0.0 and 1.0.1.
Andy Polyakov [Sat, 31 Mar 2012 18:56:27 +0000 (18:56 +0000)]
CHANGES: harmonize with 1.0.0 and 1.0.1.

7 years agoPR: 2778(part)
Dr. Stephen Henson [Sat, 31 Mar 2012 18:03:02 +0000 (18:03 +0000)]
PR: 2778(part)
Submitted by: John Fitzgibbon <john_fitzgibbon@yahoo.com>

Time is always encoded as 4 bytes, not sizeof(Time).

7 years agomodes_lcl.h: make it work on i386.
Andy Polyakov [Sat, 31 Mar 2012 17:02:46 +0000 (17:02 +0000)]
modes_lcl.h: make it work on i386.
PR: 2780

7 years agovpaes-x86[_64].pl: handle zero length in vpaes_cbc_encrypt.
Andy Polyakov [Sat, 31 Mar 2012 16:53:34 +0000 (16:53 +0000)]
vpaes-x86[_64].pl: handle zero length in vpaes_cbc_encrypt.
PR: 2775

7 years agoutil/cygwin.sh update.
Andy Polyakov [Sat, 31 Mar 2012 11:06:46 +0000 (11:06 +0000)]
util/cygwin.sh update.
PR: 2761
Submitted by: Corinna Vinschen

7 years agodon't shadow
Dr. Stephen Henson [Fri, 30 Mar 2012 15:43:32 +0000 (15:43 +0000)]
don't shadow

7 years agobn/bn_gf2m.c: make new BN_GF2m_mod_inv work with BN_DEBUG_RAND.
Andy Polyakov [Thu, 29 Mar 2012 21:35:28 +0000 (21:35 +0000)]
bn/bn_gf2m.c: make new BN_GF2m_mod_inv work with BN_DEBUG_RAND.

7 years agomodes/gcm128.c: fix self-test.
Andy Polyakov [Thu, 29 Mar 2012 18:25:38 +0000 (18:25 +0000)]
modes/gcm128.c: fix self-test.

7 years agosha512-armv4.pl: optimize NEON code path by utilizing vbsl, bitwise select.
Andy Polyakov [Thu, 29 Mar 2012 18:20:11 +0000 (18:20 +0000)]
sha512-armv4.pl: optimize NEON code path by utilizing vbsl, bitwise select.

7 years agoperlasm/x86masm.pl: fix last fix.
Andy Polyakov [Thu, 29 Mar 2012 18:09:36 +0000 (18:09 +0000)]
perlasm/x86masm.pl: fix last fix.

7 years agoans1/tasn_prn.c: avoid bool in variable names.
Andy Polyakov [Thu, 29 Mar 2012 17:48:19 +0000 (17:48 +0000)]
ans1/tasn_prn.c: avoid bool in variable names.
PR: 2776

7 years agoInitial revision of ECC extension handling.
Dr. Stephen Henson [Wed, 28 Mar 2012 15:05:04 +0000 (15:05 +0000)]
Initial revision of ECC extension handling.

Tidy some code up.

Don't allocate a structure to handle ECC extensions when it is used for
default values.

Make supported curves configurable.

Add ctrls to retrieve shared curves: not fully integrated with rest of
ECC code yet.

7 years agofix leak
Dr. Stephen Henson [Thu, 22 Mar 2012 16:28:07 +0000 (16:28 +0000)]
fix leak

7 years agoSubmitted by: Markus Friedl <mfriedl@gmail.com>
Dr. Stephen Henson [Thu, 22 Mar 2012 15:44:51 +0000 (15:44 +0000)]
Submitted by: Markus Friedl <mfriedl@gmail.com>

Fix memory leaks in 'goto err' cases.

7 years agouse client version when deciding whether to send supported signature algorithms extension
Dr. Stephen Henson [Wed, 21 Mar 2012 21:33:23 +0000 (21:33 +0000)]
use client version when deciding whether to send supported signature algorithms extension

7 years agoe_padlock-x86[_64].pl: better understanding of prefetch errata and proper
Andy Polyakov [Mon, 19 Mar 2012 20:23:32 +0000 (20:23 +0000)]
e_padlock-x86[_64].pl: better understanding of prefetch errata and proper
workaround.

7 years agoeng_all.c: revert previous "disable Padlock" commit, which was unjustified.
Andy Polyakov [Mon, 19 Mar 2012 20:20:41 +0000 (20:20 +0000)]
eng_all.c: revert previous "disable Padlock" commit, which was unjustified.

7 years agoAlways use SSLv23_{client,server}_method in s_client.c and s_server.c,
Dr. Stephen Henson [Sun, 18 Mar 2012 18:16:46 +0000 (18:16 +0000)]
Always use SSLv23_{client,server}_method in s_client.c and s_server.c,
the old code came from SSLeay days before TLS was even supported.

7 years agovpaes-x86_64.pl: out-of-date Apple assembler fails to calculate
Andy Polyakov [Sat, 17 Mar 2012 16:06:31 +0000 (16:06 +0000)]
vpaes-x86_64.pl: out-of-date Apple assembler fails to calculate
distance between local labels.
PR: 2762

7 years agobsaes-x86_64.pl: optimize key conversion.
Andy Polyakov [Fri, 16 Mar 2012 21:44:19 +0000 (21:44 +0000)]
bsaes-x86_64.pl: optimize key conversion.