oweals/openssl.git
19 years agobugfix: 0 - w (w != 0) is actually negative
Nils Larsch [Sun, 17 Jul 2005 16:09:09 +0000 (16:09 +0000)]
bugfix: 0 - w (w != 0) is actually negative

19 years agomake
Nils Larsch [Sat, 16 Jul 2005 12:37:36 +0000 (12:37 +0000)]
make
     ./configure no-deprecated [no-dsa] [no-dh] [no-ec] [no-rsa]
     make depend all test
work again

PR: 1159

19 years agoadd missing entries for "-multivalue-rdn" and "-utf8" in ca.pod and req.pod
Nils Larsch [Fri, 15 Jul 2005 09:49:03 +0000 (09:49 +0000)]
add missing entries for "-multivalue-rdn" and "-utf8" in ca.pod and req.pod

PR: 1158
Submitted by: Michael Bell <michael.bell@cms.hu-berlin.de>

19 years agoCommentary section update.
Andy Polyakov [Thu, 14 Jul 2005 13:16:31 +0000 (13:16 +0000)]
Commentary section update.

19 years agothe second argument of d2i_X509, d2i_X509_CRL and d2i_X509_REQ is const
Nils Larsch [Wed, 13 Jul 2005 22:01:33 +0000 (22:01 +0000)]
the second argument of d2i_X509, d2i_X509_CRL and d2i_X509_REQ is const

PR: 1156
Submitted by: Michael Bell <michael.bell@cms.hu-berlin.de>

19 years agocomplementary x86_64-xlate.pl update.
Andy Polyakov [Tue, 12 Jul 2005 22:11:22 +0000 (22:11 +0000)]
complementary x86_64-xlate.pl update.

19 years agoAES x86_64 assembler implementation.
Andy Polyakov [Tue, 12 Jul 2005 15:44:58 +0000 (15:44 +0000)]
AES x86_64 assembler implementation.

19 years agoLatest Intel compiler means every word in "if copying [with memcpy] takes
Andy Polyakov [Fri, 8 Jul 2005 16:46:22 +0000 (16:46 +0000)]
Latest Intel compiler means every word in "if copying [with memcpy] takes
place between objects that overlap, the behavior is undefined." It's hard
to comprehend, but it reportedly manages to be case.

19 years agoOn case insensitive systems, 'install' gets mixed up with the existing file
Richard Levitte [Fri, 8 Jul 2005 10:13:22 +0000 (10:13 +0000)]
On case insensitive systems, 'install' gets mixed up with the existing file
'INSTALL', so we need to put some force into installing

19 years agoChanges from the 0.9.8 branch.
Richard Levitte [Tue, 5 Jul 2005 19:16:24 +0000 (19:16 +0000)]
Changes from the 0.9.8 branch.

19 years agoChanges from the 0.9.8 branch.
Richard Levitte [Tue, 5 Jul 2005 18:36:42 +0000 (18:36 +0000)]
Changes from the 0.9.8 branch.

19 years agoCHANGES and TABLE sync with 0.9.8.
Andy Polyakov [Tue, 5 Jul 2005 11:48:38 +0000 (11:48 +0000)]
CHANGES and TABLE sync with 0.9.8.

19 years agoPull up Win64 support from 0.9.8.
Andy Polyakov [Tue, 5 Jul 2005 11:44:45 +0000 (11:44 +0000)]
Pull up Win64 support from 0.9.8.

19 years agoAdd libcrypto.pc and libssl.pc, and install them along with openssl.pc.
Richard Levitte [Tue, 5 Jul 2005 05:14:17 +0000 (05:14 +0000)]
Add libcrypto.pc and libssl.pc, and install them along with openssl.pc.

PR: 1143

19 years agoUpdate from stable branch.
Dr. Stephen Henson [Mon, 4 Jul 2005 23:12:04 +0000 (23:12 +0000)]
Update from stable branch.

19 years agoThe private key should never have ended up in newreq.pem.
Richard Levitte [Mon, 4 Jul 2005 21:44:16 +0000 (21:44 +0000)]
The private key should never have ended up in newreq.pem.
Now, it ends up in newkey.pem instead.

19 years agoFix bugs in bug-fix to x509/by_dir.c.
Andy Polyakov [Sun, 3 Jul 2005 13:10:45 +0000 (13:10 +0000)]
Fix bugs in bug-fix to x509/by_dir.c.
PR: 1131

19 years agoTrap condition should be 64-bit when it's due.
Andy Polyakov [Sun, 3 Jul 2005 09:17:50 +0000 (09:17 +0000)]
Trap condition should be 64-bit when it's due.

19 years agoMore elegant solution to "sparse decimal printout on PPC" problem.
Andy Polyakov [Sat, 2 Jul 2005 08:58:55 +0000 (08:58 +0000)]
More elegant solution to "sparse decimal printout on PPC" problem.

19 years agoDecimal printout of a BN is wrong on PPC, it's sparse with very few
Andy Polyakov [Fri, 1 Jul 2005 17:49:47 +0000 (17:49 +0000)]
Decimal printout of a BN is wrong on PPC, it's sparse with very few
significant digits. As soon it verifies elsewhere it goes to 0.9.8 and
0.9.7.

19 years agoinitialize newly allocated data
Nils Larsch [Fri, 1 Jul 2005 16:08:14 +0000 (16:08 +0000)]
initialize newly allocated data

PR: 1145

19 years agoPedantic polish to aes-586.pl:-)
Andy Polyakov [Fri, 1 Jul 2005 10:13:30 +0000 (10:13 +0000)]
Pedantic polish to aes-586.pl:-)

19 years agoTypos in commentary section.
Andy Polyakov [Thu, 30 Jun 2005 22:09:03 +0000 (22:09 +0000)]
Typos in commentary section.

19 years agoAES_cbc_encrypt to allow end-user to retain small blocks performance by
Andy Polyakov [Thu, 30 Jun 2005 22:06:35 +0000 (22:06 +0000)]
AES_cbc_encrypt to allow end-user to retain small blocks performance by
aligning the key schedule in a specific manner.

19 years agoCheck PKCS7 structures in PKCS#12 files are of type data.
Dr. Stephen Henson [Thu, 30 Jun 2005 11:34:58 +0000 (11:34 +0000)]
Check PKCS7 structures in PKCS#12 files are of type data.

19 years agoActually, the 64bit format specifier differs between SIXTY_FOUR_BIT and
Richard Levitte [Wed, 29 Jun 2005 18:48:08 +0000 (18:48 +0000)]
Actually, the 64bit format specifier differs between SIXTY_FOUR_BIT and
SIXTY_FOUR_BIT_LONG

19 years agoBrought forward from 0.9.8 - 64 bit warning fixes and fussy compiler fixes.
Ben Laurie [Wed, 29 Jun 2005 11:02:15 +0000 (11:02 +0000)]
Brought forward from 0.9.8 - 64 bit warning fixes and fussy compiler fixes.

19 years agoFix warnings.
Ben Laurie [Wed, 29 Jun 2005 10:25:06 +0000 (10:25 +0000)]
Fix warnings.

19 years agoasn1parse doesn't support any TXT format, so let's stop pretending
Richard Levitte [Tue, 28 Jun 2005 15:44:11 +0000 (15:44 +0000)]
asn1parse doesn't support any TXT format, so let's stop pretending
it does.

19 years agoremove OPENSSL_NO_ASM dependency
Nils Larsch [Tue, 28 Jun 2005 15:05:02 +0000 (15:05 +0000)]
remove OPENSSL_NO_ASM dependency

19 years agoReplace _int64 with __int64, which is more widely accepted among Win32
Andy Polyakov [Tue, 28 Jun 2005 11:50:50 +0000 (11:50 +0000)]
Replace _int64 with __int64, which is more widely accepted among Win32
compiler vendors.

19 years agoA report suggests that there're nasm version, which defaults to 16-bit
Andy Polyakov [Tue, 28 Jun 2005 11:28:43 +0000 (11:28 +0000)]
A report suggests that there're nasm version, which defaults to 16-bit
segmenting...

19 years agoMention hpux64-ia64-cc blowfish failure in PROBLEMS.
Andy Polyakov [Tue, 28 Jun 2005 09:57:04 +0000 (09:57 +0000)]
Mention hpux64-ia64-cc blowfish failure in PROBLEMS.

19 years agoThere are a few showstoppers. Unfortunately, I only remember one. Please fill this in.
Richard Levitte [Tue, 28 Jun 2005 05:55:22 +0000 (05:55 +0000)]
There are a few showstoppers.  Unfortunately, I only remember one.  Please fill this in.

19 years agoEliminate dependency on UNICODE macro.
Andy Polyakov [Mon, 27 Jun 2005 21:27:23 +0000 (21:27 +0000)]
Eliminate dependency on UNICODE macro.

19 years agoFix typos in apps/apps.c
Andy Polyakov [Mon, 27 Jun 2005 15:56:53 +0000 (15:56 +0000)]
Fix typos in apps/apps.c

19 years agoSync libeay.num
Dr. Stephen Henson [Mon, 27 Jun 2005 00:45:38 +0000 (00:45 +0000)]
Sync libeay.num

19 years agoUnify some SCO targets.
Andy Polyakov [Sun, 26 Jun 2005 18:06:38 +0000 (18:06 +0000)]
Unify some SCO targets.

19 years agoMinor (final?) Makefiles polish.
Andy Polyakov [Sun, 26 Jun 2005 17:47:44 +0000 (17:47 +0000)]
Minor (final?) Makefiles polish.

19 years agoMinimum requirement for Solaris x86 was always Pentium. Reflect this in
Andy Polyakov [Sun, 26 Jun 2005 17:17:11 +0000 (17:17 +0000)]
Minimum requirement for Solaris x86 was always Pentium. Reflect this in
Configure line. TABLE update accomodates other recent ./Configure changes...

19 years agoOmit padding in RC4_KEY on IA-64. The idea behind padding was to reserve
Andy Polyakov [Sun, 26 Jun 2005 16:09:29 +0000 (16:09 +0000)]
Omit padding in RC4_KEY on IA-64. The idea behind padding was to reserve
room for aligning of the key schedule itself [specific alignment is
required for future performance improvements], but OpenSSH "abuses"
our API by making copies and restoring RC4_KEY, thus ruining the
alignment and making it impossible to recover the key schedule.
PR: 1114

19 years agoUpdated status from 0.9.8-stable.
Richard Levitte [Fri, 24 Jun 2005 22:27:57 +0000 (22:27 +0000)]
Updated status from 0.9.8-stable.

19 years agoReplace emms with finit in x86cpuid.
Andy Polyakov [Fri, 24 Jun 2005 16:32:10 +0000 (16:32 +0000)]
Replace emms with finit in x86cpuid.

19 years agoAdd Argen root CAs.
Dr. Stephen Henson [Fri, 24 Jun 2005 11:07:13 +0000 (11:07 +0000)]
Add Argen root CAs.

19 years agoSomeone did some cutting and pasting and didn't quite finish the job :-).
Richard Levitte [Fri, 24 Jun 2005 05:13:08 +0000 (05:13 +0000)]
Someone did some cutting and pasting and didn't quite finish the job :-).

Notified by Steffen Pankratz <kratz00@gmx.de>

19 years agoWrap the inclusion of openssl/engine.h with a protective check for
Richard Levitte [Thu, 23 Jun 2005 22:08:47 +0000 (22:08 +0000)]
Wrap the inclusion of openssl/engine.h with a protective check for
the absence of OPENSSL_NO_ENGINE.

19 years agoWrap the inclusion of openssl/engine.h with a protective check for
Richard Levitte [Thu, 23 Jun 2005 21:57:40 +0000 (21:57 +0000)]
Wrap the inclusion of openssl/engine.h with a protective check for
the absence of OPENSSL_NO_ENGINE.

PR: 1123

19 years agoDo not defined des_crypt(), since it clashes with Solaris crypt.h.
Richard Levitte [Thu, 23 Jun 2005 21:53:56 +0000 (21:53 +0000)]
Do not defined des_crypt(), since it clashes with Solaris crypt.h.

PR: 1125

19 years agoInitialise dir to avoid a compiler warning.
Richard Levitte [Thu, 23 Jun 2005 21:49:21 +0000 (21:49 +0000)]
Initialise dir to avoid a compiler warning.

19 years agoThe NAME section of a man page is required to have a dash followed by a
Richard Levitte [Thu, 23 Jun 2005 21:45:37 +0000 (21:45 +0000)]
The NAME section of a man page is required to have a dash followed by a
short description, at least according to pod2man.

PR: 1127

19 years agoDCC doesn't like argument names in returned function pointers.
Richard Levitte [Thu, 23 Jun 2005 21:35:25 +0000 (21:35 +0000)]
DCC doesn't like argument names in returned function pointers.

PR: 1122

19 years agoUpdate for Stratus VOS.
Richard Levitte [Thu, 23 Jun 2005 21:27:21 +0000 (21:27 +0000)]
Update for Stratus VOS.

PR: 1130

19 years agoConfigure update for Stratus VOS.
Richard Levitte [Thu, 23 Jun 2005 21:21:45 +0000 (21:21 +0000)]
Configure update for Stratus VOS.

PR: 1129

19 years agoChange dir_ctrl to check for the environment variable before using the default
Richard Levitte [Thu, 23 Jun 2005 21:14:15 +0000 (21:14 +0000)]
Change dir_ctrl to check for the environment variable before using the default
directory instead of the other way around.

PR: 1131

19 years agoStrip the engine shared libraries as well.
Richard Levitte [Thu, 23 Jun 2005 20:55:38 +0000 (20:55 +0000)]
Strip the engine shared libraries as well.

Submitted by Corinna Vinschen <vinschen@redhat.com>

19 years agoDarwin specific update for Configure and Makefile.shared.
Andy Polyakov [Thu, 23 Jun 2005 20:44:27 +0000 (20:44 +0000)]
Darwin specific update for Configure and Makefile.shared.

19 years agoMissed -c in IRIX rules.
Andy Polyakov [Thu, 23 Jun 2005 20:37:29 +0000 (20:37 +0000)]
Missed -c in IRIX rules.

19 years agoTypo in bn-mips3 rule.
Andy Polyakov [Thu, 23 Jun 2005 16:24:51 +0000 (16:24 +0000)]
Typo in bn-mips3 rule.

19 years agoRename mips3.o to bn-mips3.o [it's better in long run] and adjust the
Andy Polyakov [Thu, 23 Jun 2005 16:23:06 +0000 (16:23 +0000)]
Rename mips3.o to bn-mips3.o [it's better in long run] and adjust the
rule to accomodate gcc4, which no longer support SGI as.

19 years agoHP-UX specific updates to Makefile.org and Makefile.shared.
Andy Polyakov [Thu, 23 Jun 2005 15:36:15 +0000 (15:36 +0000)]
HP-UX specific updates to Makefile.org and Makefile.shared.

19 years agoJumbo Makfiles update.
Andy Polyakov [Thu, 23 Jun 2005 00:03:26 +0000 (00:03 +0000)]
Jumbo Makfiles update.

- eliminate ambiguities between GNU-ish and SysV-ish make flavors;
- switch [back] to -e;
- fold/unify rules;

This is follow-up to the patch introducing common BUILDENV. Idea is
to collect as much parameters in $(TOP) as possible and "strip" lower
Makefiles for most variables [and thus makes them more readable].

19 years agono-asm didn't prevent make from compiling assembler modules.
Andy Polyakov [Wed, 22 Jun 2005 23:44:44 +0000 (23:44 +0000)]
no-asm didn't prevent make from compiling assembler modules.

19 years agoTABLE was out-of-sync.
Andy Polyakov [Wed, 22 Jun 2005 23:43:12 +0000 (23:43 +0000)]
TABLE was out-of-sync.

19 years agoSolaris-specific Makefile.shared update from 098.
Andy Polyakov [Wed, 22 Jun 2005 23:42:34 +0000 (23:42 +0000)]
Solaris-specific Makefile.shared update from 098.

19 years agoTypo.
Dr. Stephen Henson [Wed, 22 Jun 2005 18:54:54 +0000 (18:54 +0000)]
Typo.

19 years agoFix extension ordering.
Dr. Stephen Henson [Wed, 22 Jun 2005 13:26:23 +0000 (13:26 +0000)]
Fix extension ordering.

19 years agoDo no try to pretend we're at the end of anything unless we're at the end
Richard Levitte [Mon, 20 Jun 2005 22:11:14 +0000 (22:11 +0000)]
Do no try to pretend we're at the end of anything unless we're at the end
of a 4-character block.

19 years agoCheck for 'usage' and 'Usage'.
Richard Levitte [Mon, 20 Jun 2005 20:48:21 +0000 (20:48 +0000)]
Check for 'usage' and 'Usage'.
Submitted by Tim Rice <tim@multitalents.net>.  His comment is:

I noticed "make report" didn't show the cc version on most of
my System V platforms. This patch corrects this.

19 years agoFix typo in ssl/d1_pkt.c.
Andy Polyakov [Mon, 20 Jun 2005 19:36:34 +0000 (19:36 +0000)]
Fix typo in ssl/d1_pkt.c.

19 years agoRefine ELF detection on BSD platforms.
Andy Polyakov [Mon, 20 Jun 2005 17:39:10 +0000 (17:39 +0000)]
Refine ELF detection on BSD platforms.

19 years agoSolaris x86_64 /usr/ccs/bin/as support.
Andy Polyakov [Mon, 20 Jun 2005 14:56:48 +0000 (14:56 +0000)]
Solaris x86_64 /usr/ccs/bin/as support.

19 years agoHandle wrap-arounds and revive missing assignment.
Andy Polyakov [Mon, 20 Jun 2005 12:40:37 +0000 (12:40 +0000)]
Handle wrap-arounds and revive missing assignment.

19 years agoAdd crypto/bn/bn_prime.h to the collection of generated files. In the
Richard Levitte [Mon, 20 Jun 2005 04:29:50 +0000 (04:29 +0000)]
Add crypto/bn/bn_prime.h to the collection of generated files.  In the
update target, place the dependency on depend last, so all necessary files
are generated *before* the dependencies are figured out.

PR: 1121

19 years agoWith DJGPP, it seems like the return code from grep, even when in the
Richard Levitte [Sun, 19 Jun 2005 20:31:15 +0000 (20:31 +0000)]
With DJGPP, it seems like the return code from grep, even when in the
middle of a pipe, is noted.  Counter that by forcing a true return code
when the return code has no importance.

PR: 1085

19 years agoUndefine DECRANDOM before redefining it.
Richard Levitte [Sun, 19 Jun 2005 20:20:24 +0000 (20:20 +0000)]
Undefine DECRANDOM before redefining it.

PR: 1110

19 years agoAdd better documentation on how id_function() should be defined and what
Richard Levitte [Sat, 18 Jun 2005 05:52:16 +0000 (05:52 +0000)]
Add better documentation on how id_function() should be defined and what
issues there are.

PR: 1096

19 years agoMove the definition of DEVRANDOM for DJGPP from Configure to e_os.h.
Richard Levitte [Sat, 18 Jun 2005 04:42:24 +0000 (04:42 +0000)]
Move the definition of DEVRANDOM for DJGPP from Configure to e_os.h.
That should solve the issues with propagating it through the Makefiles.

PR: 1110

19 years agoOnly define ZLIB_SHARED if it hasn't already been defined (on the command
Richard Levitte [Sat, 18 Jun 2005 04:32:12 +0000 (04:32 +0000)]
Only define ZLIB_SHARED if it hasn't already been defined (on the command
line, for example).

PR: 1112

19 years agoHave pod2man.pl accept '=for comment ...' before the '=head1 NAME' line.
Richard Levitte [Sat, 18 Jun 2005 04:27:06 +0000 (04:27 +0000)]
Have pod2man.pl accept '=for comment ...' before the '=head1 NAME' line.

PR: 1113

19 years agoclear dso pointer in case of an error
Nils Larsch [Fri, 17 Jun 2005 21:26:36 +0000 (21:26 +0000)]
clear dso pointer in case of an error

PR: 816

19 years agoupdate for the cswift engine:
Nils Larsch [Fri, 17 Jun 2005 20:27:41 +0000 (20:27 +0000)]
update for the cswift engine:
- fix the problem described in bug report 825
- fix a segfault when the engine fails to initialize
- let the engine switch to software when keysize > 2048

PR: 825, 826
Submitted by: Frédéric Giudicelli

19 years agoDo not undefine _XOPEN_SOURCE. This is currently experimental, and
Richard Levitte [Thu, 16 Jun 2005 22:20:55 +0000 (22:20 +0000)]
Do not undefine _XOPEN_SOURCE.  This is currently experimental, and
will be firmed up as soon as it's been verified not to break anything.

19 years ago0.9.8-beta5 works on Gentoo/arml but not /armb, and works on Linux AMD64
Richard Levitte [Tue, 14 Jun 2005 05:42:52 +0000 (05:42 +0000)]
0.9.8-beta5 works on Gentoo/arml but not /armb, and works on Linux AMD64

19 years agoData about which Cygwin versions 0.9.8-beta5 work on
Richard Levitte [Mon, 13 Jun 2005 17:10:03 +0000 (17:10 +0000)]
Data about which Cygwin versions 0.9.8-beta5 work on

19 years ago0.9.8-beta5 works on SuSE 9.3
Richard Levitte [Mon, 13 Jun 2005 17:03:13 +0000 (17:03 +0000)]
0.9.8-beta5 works on SuSE 9.3

19 years ago0.9.8-beta5 works on Cygwin
Richard Levitte [Mon, 13 Jun 2005 17:00:18 +0000 (17:00 +0000)]
0.9.8-beta5 works on Cygwin

19 years agoupdate FAQ
Nils Larsch [Mon, 13 Jun 2005 08:38:02 +0000 (08:38 +0000)]
update FAQ

19 years ago0.9.8-beta5 works on VMS/Alpha
Richard Levitte [Mon, 13 Jun 2005 04:17:12 +0000 (04:17 +0000)]
0.9.8-beta5 works on VMS/Alpha

19 years agoStatus update
Richard Levitte [Mon, 13 Jun 2005 03:36:58 +0000 (03:36 +0000)]
Status update

19 years agoNetware patch submitted by Verdon Walker" <VWalker@novell.com> in PR
Richard Levitte [Mon, 13 Jun 2005 03:23:50 +0000 (03:23 +0000)]
Netware patch submitted by Verdon Walker" <VWalker@novell.com> in PR
1107.  He says:

This is a followup to the NetWare patch that was applied to beta3.  It
does the following:

- Fixes a problem in the CLib build with undefined symbols.

- Adds the ability to use BSD sockets as the default for the OpenSSL
  socket BIO.  NetWare supports 2 flavors of sockets and our Apache
  developers need BSD sockets as a configurable option when building
  OpenSSL.  This adds that for them.

- Updates to the INSTALL.NW file to explain new options.

I have tried very hard to make sure all the changes are in NetWare
specific files or guarded carefully to make sure they only impact
NetWare builds.  I have tested the Windows build to make sure it does
not break that since we have made changes to mk1mf.pl.

We are still working the gcc cross compile for NetWare issue and hope
to have a patch for that before beta 6 is released.

19 years agoShow what the offending target was.
Richard Levitte [Mon, 13 Jun 2005 02:39:05 +0000 (02:39 +0000)]
Show what the offending target was.

PR: 1108

19 years agoreplace the deprecated "-m486" gcc option with "-march=i486"
Nils Larsch [Sun, 12 Jun 2005 10:34:42 +0000 (10:34 +0000)]
replace the deprecated "-m486" gcc option with "-march=i486"

PR: 1049

19 years ago- let SSL_CTX_set_cipher_list and SSL_set_cipher_list return an
Nils Larsch [Fri, 10 Jun 2005 19:55:26 +0000 (19:55 +0000)]
- let SSL_CTX_set_cipher_list and SSL_set_cipher_list return an
  error if the cipher list is empty
- fix last commit in ssl_create_cipher_list
- clean up ssl_create_cipher_list

19 years agoEliminate gcc -pedantic warnings.
Andy Polyakov [Thu, 9 Jun 2005 21:41:44 +0000 (21:41 +0000)]
Eliminate gcc -pedantic warnings.

19 years agoAllow for dso load by explicit path on HP-UX.
Andy Polyakov [Thu, 9 Jun 2005 20:52:24 +0000 (20:52 +0000)]
Allow for dso load by explicit path on HP-UX.

19 years agoWhen the return type of the function is int, it's better to return an
Richard Levitte [Thu, 9 Jun 2005 17:28:53 +0000 (17:28 +0000)]
When the return type of the function is int, it's better to return an
in than NULL, especially when an error is signalled with a negative
value.

19 years agouse "=" instead of "|=", fix typo
Nils Larsch [Wed, 8 Jun 2005 22:22:33 +0000 (22:22 +0000)]
use "=" instead of "|=", fix typo

19 years agoAvoid endless loops. Really, we were using the same variable for two
Richard Levitte [Wed, 8 Jun 2005 21:59:47 +0000 (21:59 +0000)]
Avoid endless loops.  Really, we were using the same variable for two
different conditions...

19 years agossl_create_cipher_list should return an error if no cipher could be
Nils Larsch [Wed, 8 Jun 2005 21:19:14 +0000 (21:19 +0000)]
ssl_create_cipher_list should return an error if no cipher could be
collected (see SSL_CTX_set_cipher_list manpage). Fix handling of
"cipher1+cipher2" expressions in ssl_cipher_process_rulestr.

PR: 836 + 1005

19 years ago"Liberate" dtls from BN dependency. Fix bug in replay/update.
Andy Polyakov [Tue, 7 Jun 2005 22:21:14 +0000 (22:21 +0000)]
"Liberate" dtls from BN dependency. Fix bug in replay/update.