openssl.git
21 years agoAdd the CBC flag for cbc ciphers
Richard Levitte [Thu, 1 Aug 2002 19:32:48 +0000 (19:32 +0000)]
Add the CBC flag for cbc ciphers

21 years agoAvoid yet another name clash with libdes, and make the declaration consistent
Richard Levitte [Thu, 1 Aug 2002 19:30:58 +0000 (19:30 +0000)]
Avoid yet another name clash with libdes, and make the declaration consistent
with the definition.

21 years agoMake it possible to load keys from stdin, and restore that
Richard Levitte [Thu, 1 Aug 2002 16:28:40 +0000 (16:28 +0000)]
Make it possible to load keys from stdin, and restore that
functionality in the programs that had that before.
Part fo PR 164

21 years agoOCSP and KRB5 Makefil.ssl should be consistent with all the others
Richard Levitte [Thu, 1 Aug 2002 14:18:52 +0000 (14:18 +0000)]
OCSP and KRB5 Makefil.ssl should be consistent with all the others

21 years agoLinux on s390 really knows about loading dynamically.
Richard Levitte [Thu, 1 Aug 2002 13:50:08 +0000 (13:50 +0000)]
Linux on s390 really knows about loading dynamically.
PR: 183

21 years agoCut'n'paste error with other reposnder certificates cleared.
Richard Levitte [Thu, 1 Aug 2002 13:39:39 +0000 (13:39 +0000)]
Cut'n'paste error with other reposnder certificates cleared.
PR: 190

21 years agoIf CRYPTO_realloc() is called with a NULL pointer, have it call
Richard Levitte [Thu, 1 Aug 2002 10:08:37 +0000 (10:08 +0000)]
If CRYPTO_realloc() is called with a NULL pointer, have it call
OPENSSL_malloc().
PR: 187

21 years agoIf the email address is moved from the subject to the subject alternate name,
Richard Levitte [Wed, 31 Jul 2002 14:05:57 +0000 (14:05 +0000)]
If the email address is moved from the subject to the subject alternate name,
the subject in the certificate would differ from the subject in the index file,
which has quite bad concequences.
PR: 180

21 years agoMake sure to use $(MAKE) everywhere instead of make.
Richard Levitte [Wed, 31 Jul 2002 13:49:06 +0000 (13:49 +0000)]
Make sure to use $(MAKE) everywhere instead of make.
Part of PR 181

21 years agoln -f -s doesn't always work, so do a rm -f followed by a ln -s.
Richard Levitte [Wed, 31 Jul 2002 13:38:32 +0000 (13:38 +0000)]
ln -f -s doesn't always work, so do a rm -f followed by a ln -s.
Part of PR 181

21 years agoThe version of the shared library should, for now, reflect the version
Richard Levitte [Wed, 31 Jul 2002 13:24:26 +0000 (13:24 +0000)]
The version of the shared library should, for now, reflect the version
of OpenSSL.
Part of PR 181.

21 years agoTypo.
Lutz Jänicke [Tue, 30 Jul 2002 13:36:31 +0000 (13:36 +0000)]
Typo.
Submitted by: Jeffrey Altman <jaltman@columbia.edu>
Reviewed by:
PR: 169

21 years agoOpenSSL Security Advisory [30 July 2002]
Lutz Jänicke [Tue, 30 Jul 2002 13:04:04 +0000 (13:04 +0000)]
OpenSSL Security Advisory [30 July 2002]

Changes marked "(CHATS)" were sponsored by the Defense Advanced
Research Projects Agency (DARPA) and Air Force Research Laboratory,
Air Force Materiel Command, USAF, under agreement number
F30602-01-2-0537.

21 years ago"make update"
Lutz Jänicke [Tue, 30 Jul 2002 12:44:33 +0000 (12:44 +0000)]
"make update"

21 years ago0.9.6e and 0.9.7-beta3 are out.
Lutz Jänicke [Tue, 30 Jul 2002 11:32:47 +0000 (11:32 +0000)]
0.9.6e and 0.9.7-beta3 are out.

21 years agoOnly use DSA-functions if available.
Lutz Jänicke [Mon, 29 Jul 2002 13:31:44 +0000 (13:31 +0000)]
Only use DSA-functions if available.
Submitted by: "Hellan,Kim KHE" <KHE@kmd.dk>
Reviewed by:
PR: 167

21 years agomention SSL_do_handshake()
Bodo Möller [Mon, 29 Jul 2002 12:35:19 +0000 (12:35 +0000)]
mention SSL_do_handshake()

21 years agoUse SEC1 format for EC private keys.
Bodo Möller [Fri, 26 Jul 2002 08:41:04 +0000 (08:41 +0000)]
Use SEC1 format for EC private keys.
This is not ECDSA specific, so it's now PEM_STRING_ECPRIVATEKEY etc.

Submitted by: Nils Larsch <nlarsch@compuserve.de>

21 years agoMove zeroing from bn_expand_internal() to bn_expand2() so that it
Bodo Möller [Thu, 25 Jul 2002 12:12:39 +0000 (12:12 +0000)]
Move zeroing from bn_expand_internal() to bn_expand2() so that it
happens reliably, even if the BIGNUM is already sufficiently large.

[Note that the bn_expand()/bn_wexpand() macros call bn_expand2() only
if the BIGNUM actually has to grow, so this change does not add any
new overhead as currently bn_expand2() is never called directly.]

21 years agomore detailed instructions for export from US
Bodo Möller [Thu, 25 Jul 2002 11:19:58 +0000 (11:19 +0000)]
more detailed instructions for export from US

21 years agoDon't clobber loop variable.
Richard Levitte [Wed, 24 Jul 2002 14:43:00 +0000 (14:43 +0000)]
Don't clobber loop variable.
PR: 159

21 years agoDocument the recent DJGPP-related changes
Richard Levitte [Tue, 23 Jul 2002 13:45:38 +0000 (13:45 +0000)]
Document the recent DJGPP-related changes

21 years agoMake dummy tests to make it easier to physically remove unwanted algorithms.
Richard Levitte [Tue, 23 Jul 2002 13:31:04 +0000 (13:31 +0000)]
Make dummy tests to make it easier to physically remove unwanted algorithms.
This should complete PR 75

21 years agoharmonize options with those for 'ecparam',
Bodo Möller [Tue, 23 Jul 2002 09:51:57 +0000 (09:51 +0000)]
harmonize options with those for 'ecparam',
remove redudant option '-pub'

Submitted by: Nils Larsch

21 years agoIf DH is disabled, don't define the DH functions.
Richard Levitte [Mon, 22 Jul 2002 14:18:11 +0000 (14:18 +0000)]
If DH is disabled, don't define the DH functions.
Notified by Kim Hellan <KHE@kmd.dk>

21 years agofix a typo and clarify
Bodo Möller [Mon, 22 Jul 2002 09:04:36 +0000 (09:04 +0000)]
fix a typo and clarify

21 years agoadd an explanation and fix a typo
Bodo Möller [Mon, 22 Jul 2002 08:39:44 +0000 (08:39 +0000)]
add an explanation and fix a typo

21 years agoNew cipher selection options COMPLEMENTOFALL and COMPLEMENTOFDEFAULT.
Lutz Jänicke [Fri, 19 Jul 2002 19:55:34 +0000 (19:55 +0000)]
New cipher selection options COMPLEMENTOFALL and COMPLEMENTOFDEFAULT.
Submitted by:
Reviewed by:
PR: 127

21 years agoHP-UX shared libraries must be +x and should be -w. It doesn't hurt on
Lutz Jänicke [Fri, 19 Jul 2002 16:32:40 +0000 (16:32 +0000)]
HP-UX shared libraries must be +x and should be -w. It doesn't hurt on
other platforms.
Submitted by:
Reviewed by:
PR: 134

21 years agoupdate
Bodo Möller [Fri, 19 Jul 2002 12:32:42 +0000 (12:32 +0000)]
update

21 years agoThe first compile of the new merger method for VMS *almost* got
Richard Levitte [Fri, 19 Jul 2002 11:57:17 +0000 (11:57 +0000)]
The first compile of the new merger method for VMS *almost* got
through.  That's not enough, is it?  :-)

21 years agoTwo more names that are a little too long for the standard VMS linker.
Richard Levitte [Fri, 19 Jul 2002 11:56:02 +0000 (11:56 +0000)]
Two more names that are a little too long for the standard VMS linker.

21 years agoThe behaviour is undefined when calling SSL_write() with num=0.
Lutz Jänicke [Fri, 19 Jul 2002 11:53:54 +0000 (11:53 +0000)]
The behaviour is undefined when calling SSL_write() with num=0.
Submitted by:
Reviewed by:
PR: 141

21 years agoManual page for SSL_do_handshake().
Lutz Jänicke [Fri, 19 Jul 2002 11:05:50 +0000 (11:05 +0000)]
Manual page for SSL_do_handshake().
Submitted by: Martin Sjögren <martin@strakt.com>
PR: 137

21 years agoThis documentation change was being written at the same time as Richard's
Geoff Thorpe [Thu, 18 Jul 2002 20:59:22 +0000 (20:59 +0000)]
This documentation change was being written at the same time as Richard's
changes. So I'm committing this version to overwrite his changes for now,
and he can always take his turn to overwrite my words if he wants :-)

PR: 86

21 years agoExplain why RSA_check_key() doesn't work with hard keys.
Richard Levitte [Thu, 18 Jul 2002 19:10:57 +0000 (19:10 +0000)]
Explain why RSA_check_key() doesn't work with hard keys.
PR: 86

21 years agoAdd history for documented new functions.
Richard Levitte [Thu, 18 Jul 2002 18:54:46 +0000 (18:54 +0000)]
Add history for documented new functions.
PR: 59

21 years agoAllow subjects with more than 255 characters to be properly printed.
Richard Levitte [Thu, 18 Jul 2002 17:59:21 +0000 (17:59 +0000)]
Allow subjects with more than 255 characters to be properly printed.
PR: 147

21 years agoFurther enhance assembler support on Cygwin and DJGPP.
Richard Levitte [Thu, 18 Jul 2002 12:37:59 +0000 (12:37 +0000)]
Further enhance assembler support on Cygwin and DJGPP.
Make pod2mantest useable on DOS-based systems.
Part of PR 75, the rest is still under investigation.

21 years agoFix bug introduced with revision 1.95 when this filed was modified to
Bodo Möller [Thu, 18 Jul 2002 11:23:50 +0000 (11:23 +0000)]
Fix bug introduced with revision 1.95 when this filed was modified to
use the new X509_CRL_set_issuer_name() function:
The CRL issuer should be X509_get_subject_name(x509), not
X509_get_issuer_name(x509).

Submitted by: Juergen Lesny <lesnyj@informatik.tu-muenchen.de>

typo

21 years agoReverse the change with the following log, it needs further investigation:
Richard Levitte [Thu, 18 Jul 2002 10:39:20 +0000 (10:39 +0000)]
Reverse the change with the following log, it needs further investigation:

Make S/MIME output conform with the mail and MIME standards.
PR: 151

21 years agoMake S/MIME output conform with the mail and MIME standards.
Richard Levitte [Thu, 18 Jul 2002 08:47:33 +0000 (08:47 +0000)]
Make S/MIME output conform with the mail and MIME standards.
PR: 151

21 years agoUnixware doesn't have strings.h, so we need to declare strcasecmp()
Richard Levitte [Thu, 18 Jul 2002 07:47:30 +0000 (07:47 +0000)]
Unixware doesn't have strings.h, so we need to declare strcasecmp()
differently.
Unixware 2 needs to link with libresolv.
PR: 148

21 years agoOPENSSL_SYS_WIN32 is important so util/mkdef.pl can detect it
Richard Levitte [Thu, 18 Jul 2002 06:34:34 +0000 (06:34 +0000)]
OPENSSL_SYS_WIN32 is important so util/mkdef.pl can detect it

21 years agoMake mkdir failsafe in case the directories are already present
Richard Levitte [Wed, 17 Jul 2002 13:33:02 +0000 (13:33 +0000)]
Make mkdir failsafe in case the directories are already present

21 years agoAdd support for shared libraries with OS/2.
Richard Levitte [Wed, 17 Jul 2002 13:27:43 +0000 (13:27 +0000)]
Add support for shared libraries with OS/2.
PR: 124

21 years agoAdd a few FAQ entries for the various ways building OpenSSL on MacOS X
Richard Levitte [Wed, 17 Jul 2002 11:29:15 +0000 (11:29 +0000)]
Add a few FAQ entries for the various ways building OpenSSL on MacOS X
can fail, and point at the text in the PROBLEMS file

21 years agoIf OpenSSL is built with shared library support on MacOS X,
Richard Levitte [Wed, 17 Jul 2002 11:16:22 +0000 (11:16 +0000)]
If OpenSSL is built with shared library support on MacOS X,
 everything works fine.  Note: this is only true for 0.9.7 and on

21 years agoOn MacOS X, the shared library editor uses DYLD_LIBRARY_PATH
Richard Levitte [Wed, 17 Jul 2002 11:09:44 +0000 (11:09 +0000)]
On MacOS X, the shared library editor uses DYLD_LIBRARY_PATH

21 years agoOn MacOS X, you're not allowed to have common objects in shared libraries
Richard Levitte [Wed, 17 Jul 2002 08:20:27 +0000 (08:20 +0000)]
On MacOS X, you're not allowed to have common objects in shared libraries

21 years agoAdd the usually recommended solution to the linking problem entry for MacOS X
Richard Levitte [Wed, 17 Jul 2002 07:48:39 +0000 (07:48 +0000)]
Add the usually recommended solution to the linking problem entry for MacOS X

21 years agoWe had some experimental options in the Darwin entries. They are no longer needed
Richard Levitte [Tue, 16 Jul 2002 11:07:42 +0000 (11:07 +0000)]
We had some experimental options in the Darwin entries.  They are no longer needed

21 years agoThe default C compiler on MacOS X doesn't like empty object files
Richard Levitte [Tue, 16 Jul 2002 10:46:00 +0000 (10:46 +0000)]
The default C compiler on MacOS X doesn't like empty object files

21 years agoActually, the "bug" is really documented in the man-page for ld, so
Richard Levitte [Tue, 16 Jul 2002 10:20:06 +0000 (10:20 +0000)]
Actually, the "bug" is really documented in the man-page for ld, so
it's really a misfeature according to the jargon file (4.0.0)
definition:

":misfeature: /mis-fee'chr/ or /mis'fee`chr/ /n./  A feature
   that eventually causes lossage, possibly because it is not adequate
   for a new situation that has evolved.  Since it results from a
   deliberate and properly implemented feature, a misfeature is not a
   bug."

21 years agoThere are problems on certain platforms, and possible answers on how to (temporarly...
Richard Levitte [Tue, 16 Jul 2002 10:04:40 +0000 (10:04 +0000)]
There are problems on certain platforms, and possible answers on how to (temporarly) solve them.

21 years agoFor those wanting to build for several platforms with the same source
Richard Levitte [Tue, 16 Jul 2002 09:19:37 +0000 (09:19 +0000)]
For those wanting to build for several platforms with the same source
directory, making a separate directory tree with lots of symbolic links
seems to be the solution.  Unfortunately, Configure doesn't take appropriate
steps to support this solution (as in removing a file that's going to be
rewritten).  This change corrects that situation.  Now I just have to
find all other places where there's lack of support for this.

21 years agoFor those wanting to build for several platforms with the same source
Richard Levitte [Tue, 16 Jul 2002 08:53:32 +0000 (08:53 +0000)]
For those wanting to build for several platforms with the same source
directory, making a separate directory tree with lots of symbolic links
seems to be the solution.  Unfortunatelt, Configure doesn't take appropriate
steps to support this solution (as in removing a file that's going to be
rewritten).  This change corrects that situation.  Now I just have to
find all other places where there's lack of support for this.

21 years agoSome older code (never committed) wasn't converted to the new format.
Richard Levitte [Tue, 16 Jul 2002 08:02:31 +0000 (08:02 +0000)]
Some older code (never committed) wasn't converted to the new format.
Corrected.

21 years agoLet's not forget the second -Wtraditional
Richard Levitte [Tue, 16 Jul 2002 07:06:48 +0000 (07:06 +0000)]
Let's not forget the second -Wtraditional

21 years agoUsing -Wtraditional took it a little far. After all, we expect ANSI C, so
Richard Levitte [Tue, 16 Jul 2002 06:53:45 +0000 (06:53 +0000)]
Using -Wtraditional took it a little far.  After all, we expect ANSI C, so
we don't need to care about traditional compilers

21 years agoSet up the engine before doing anything random-related, since engine randomness
Richard Levitte [Tue, 16 Jul 2002 06:52:03 +0000 (06:52 +0000)]
Set up the engine before doing anything random-related, since engine randomness
is only used for seeding and doing it in the wrong order will mean seeding
is done before the engine randomness is hooked in.
Notified by Frederic DONNAT <frederic.donnat@zencod.com>

21 years agoThere's an ongoing project to bring some kind of path selection
Richard Levitte [Mon, 15 Jul 2002 15:35:40 +0000 (15:35 +0000)]
There's an ongoing project to bring some kind of path selection
mechanism to the ENGINE framework.  This means there there are going
to be new functionality for the DSO part, and ultimately some way of
merging two file specifications together.

This commit places the merging code into the repository.  It's
currently not used anywhere, and hasn't been tested at all.  It may be
full of errors, including syntactical ones.  Those will be fixed as
promptly as possible.

21 years agoReplace 'ecdsaparam' commandline utility by 'ecparam'
Bodo Möller [Sun, 14 Jul 2002 16:54:31 +0000 (16:54 +0000)]
Replace 'ecdsaparam' commandline utility by 'ecparam'
(the same keys can be used for ECC schemes other than ECDSA)
and add some new options.

Similarly, use string "EC PARAMETERS" instead of "ECDSA PARAMETERS"
in 'PEM' format.

Fix ec_asn1.c (take into account the desired conversion form).

'make update'.

Submitted by: Nils Larsch

21 years agoRewording: some algorithms are also patented in Europe, so choose more
Lutz Jänicke [Fri, 12 Jul 2002 15:26:37 +0000 (15:26 +0000)]
Rewording: some algorithms are also patented in Europe, so choose more
defensive phrases...

21 years agoIn UI_UTIL_read_pw(), we should look at the size parameter, not at BUFSIZ.
Richard Levitte [Thu, 11 Jul 2002 09:12:29 +0000 (09:12 +0000)]
In UI_UTIL_read_pw(), we should look at the size parameter, not at BUFSIZ.
Submitted by Götz Babin-Ebell <babinebell@trustcenter.de>

21 years agoDiscussion about Redhat's specialties for the FAQ.
Lutz Jänicke [Wed, 10 Jul 2002 19:48:51 +0000 (19:48 +0000)]
Discussion about Redhat's specialties for the FAQ.
Submitted by: John.Airey@rnib.org.uk
Reviewed by:
PR: 128

21 years agoTypos in links between manual pages
Lutz Jänicke [Wed, 10 Jul 2002 19:35:54 +0000 (19:35 +0000)]
Typos in links between manual pages
Submitted by: Richard.Koenning@fujitsu-siemens.com
Reviewed by:
PR: 129

21 years agoSun's official statement with respect to /dev/random support.
Lutz Jänicke [Wed, 10 Jul 2002 17:52:03 +0000 (17:52 +0000)]
Sun's official statement with respect to /dev/random support.
Submitted by: Garrett Anderson garrett@dirsec.com
Reviewed by:
PR: 120

21 years agoMinor typos
Lutz Jänicke [Wed, 10 Jul 2002 17:34:54 +0000 (17:34 +0000)]
Minor typos
Submitted by: jufi@nerdnet.de
Reviewed by:
PR: 138

21 years agoremove obsolete comment
Bodo Möller [Wed, 10 Jul 2002 08:39:05 +0000 (08:39 +0000)]
remove obsolete comment

21 years agoReorder inclusion of header files:
Lutz Jänicke [Wed, 10 Jul 2002 07:01:54 +0000 (07:01 +0000)]
Reorder inclusion of header files:

des_old.h redefines crypt:
#define crypt(b,s)\
        DES_crypt((b),(s))

This scheme leads to failure, if header files with the OS's true definition
of crypt() are processed _after_ des_old.h was processed. This is e.g. the
case on HP-UX with unistd.h.
As evp.h now again includes des.h (which includes des_old.h), this problem
only came up after this modification.
Solution: move header files (indirectly) including e_os.h before the header
files (indirectly) including evp.h.
Submitted by:
Reviewed by:
PR:

21 years agoCiphers with NULL encryption were not properly handled because they were
Lutz Jänicke [Wed, 10 Jul 2002 06:41:55 +0000 (06:41 +0000)]
Ciphers with NULL encryption were not properly handled because they were
not covered by the strength bit mask.
Submitted by:
Reviewed by:
PR: 130

21 years agofix synopsis
Bodo Möller [Tue, 9 Jul 2002 10:51:25 +0000 (10:51 +0000)]
fix synopsis

Submitted by: Nils Larsch

21 years agoemtpy fragments are not necessary for SSL_eNULL
Bodo Möller [Tue, 9 Jul 2002 08:49:09 +0000 (08:49 +0000)]
emtpy fragments are not necessary for SSL_eNULL
(but noone uses it anyway)

fix t1_enc.c: use OPENSSL_NO_RC4, not NO_RC4

21 years agooops, there were other cases of "ENGINE_ID" to change too.
Geoff Thorpe [Mon, 8 Jul 2002 15:16:10 +0000 (15:16 +0000)]
oops, there were other cases of "ENGINE_ID" to change too.

21 years agoCorrect an error in the README.ENGINE file.
Geoff Thorpe [Mon, 8 Jul 2002 14:57:09 +0000 (14:57 +0000)]
Correct an error in the README.ENGINE file.

Submitted by: Jan Tschirschwitz <Jan.Tschirschwitz@cluster-labs.de>

21 years agoAES cipher suites are now official (RFC3268)
Bodo Möller [Thu, 4 Jul 2002 08:51:09 +0000 (08:51 +0000)]
AES cipher suites are now official (RFC3268)

21 years agoREADME and INSTALL should contain information about the request tracker
Lutz Jänicke [Sun, 30 Jun 2002 19:40:18 +0000 (19:40 +0000)]
README and INSTALL should contain information about the request tracker
(noted by Jonathan Louie <jlouie@recourse.com>).

21 years agoDo not define crypt() on OpenBSD. Notified by Bob Beck of OpenBSD.
Richard Levitte [Sat, 29 Jun 2002 22:04:09 +0000 (22:04 +0000)]
Do not define crypt() on OpenBSD.  Notified by Bob Beck of OpenBSD.

21 years agoopensslconf.h doesn't define what we want, e_os2.h does.
Richard Levitte [Thu, 27 Jun 2002 17:19:55 +0000 (17:19 +0000)]
opensslconf.h doesn't define what we want, e_os2.h does.
PR 123

21 years agoTry to avoid double declaration of ERR_load_PEM_strings().
Richard Levitte [Thu, 27 Jun 2002 17:06:28 +0000 (17:06 +0000)]
Try to avoid double declaration of ERR_load_PEM_strings().
PR 71

21 years agoA few changes to BC-32.pl didn't get properly applied.
Richard Levitte [Thu, 27 Jun 2002 16:56:30 +0000 (16:56 +0000)]
A few changes to BC-32.pl didn't get properly applied.
This completes PR 123

21 years agoPass CFLAG to dependency makers, so non-standard system include paths are
Richard Levitte [Thu, 27 Jun 2002 16:39:25 +0000 (16:39 +0000)]
Pass CFLAG to dependency makers, so non-standard system include paths are
handled properly.
Part of PR 75

21 years agoDJGPP has some needed header files that other MSDOS/Windows compilers don't have.
Richard Levitte [Thu, 27 Jun 2002 16:32:04 +0000 (16:32 +0000)]
DJGPP has some needed header files that other MSDOS/Windows compilers don't have.
Part of PR 75

21 years agoUpdate the information on Cygwin.
Richard Levitte [Thu, 27 Jun 2002 16:29:46 +0000 (16:29 +0000)]
Update the information on Cygwin.
Part of PR 75

21 years agoForgot to change the second $ENV{DJDIR} to /dev/env/DJDIR.
Richard Levitte [Thu, 27 Jun 2002 16:28:04 +0000 (16:28 +0000)]
Forgot to change the second $ENV{DJDIR} to /dev/env/DJDIR.
Part of PR 75

21 years agoThe new stuff is for Borland Bulider 5, so document it appropriately
Richard Levitte [Thu, 27 Jun 2002 15:10:57 +0000 (15:10 +0000)]
The new stuff is for Borland Bulider 5, so document it appropriately

21 years agoDocument the new way of building with Borland Builder.
Richard Levitte [Thu, 27 Jun 2002 15:07:27 +0000 (15:07 +0000)]
Document the new way of building with Borland Builder.
This concludes the changes from PR 123

21 years agoUse underscores instead of dashes in temporary file names.
Richard Levitte [Thu, 27 Jun 2002 14:57:54 +0000 (14:57 +0000)]
Use underscores instead of dashes in temporary file names.
This is due to weird Borland compilers.
Part of PR 123

21 years agoWhen compiling for Windows, make sure we have the windows definitions declared.
Richard Levitte [Thu, 27 Jun 2002 14:55:52 +0000 (14:55 +0000)]
When compiling for Windows, make sure we have the windows definitions declared.
Part of PR 123

21 years agoUse 32-bit sections instead of the default, 16-bit ones.
Richard Levitte [Thu, 27 Jun 2002 14:54:34 +0000 (14:54 +0000)]
Use 32-bit sections instead of the default, 16-bit ones.
Part of PR 123

21 years agohave 'openssl pkcs7' exit with code 1 on error instead of 0.
Richard Levitte [Thu, 27 Jun 2002 10:26:40 +0000 (10:26 +0000)]
have 'openssl pkcs7' exit with code 1 on error instead of 0.
PR: 119

21 years agogcc requires -m64 to link 64-bit shared libraries on Solaris.
Richard Levitte [Thu, 27 Jun 2002 10:19:17 +0000 (10:19 +0000)]
gcc requires -m64 to link 64-bit shared libraries on Solaris.
PR: 117

21 years agoUse bg instead of bag as argument to macros, to avoid clashes with
Richard Levitte [Thu, 27 Jun 2002 09:54:07 +0000 (09:54 +0000)]
Use bg instead of bag as argument to macros, to avoid clashes with
structure field names.
PR: 112

21 years agoThere is no RSAREF any more, so do not ty to install it.
Richard Levitte [Thu, 27 Jun 2002 09:18:45 +0000 (09:18 +0000)]
There is no RSAREF any more, so do not ty to install it.
PR: 106

21 years agoThe general debug target must specify that it doesn't use assembler routines.
Richard Levitte [Thu, 27 Jun 2002 09:13:44 +0000 (09:13 +0000)]
The general debug target must specify that it doesn't use assembler routines.
PR: 105

21 years agoA number of includes were removed from evp.h some time ago. The reason
Richard Levitte [Thu, 27 Jun 2002 05:03:00 +0000 (05:03 +0000)]
A number of includes were removed from evp.h some time ago.  The reason
was that they weren't really needed any more for EVP itself.  However,
it seems like soma applications (I know about OpenSSH, but there may
be more) used evp.h as the 'load all' header file, which makes sense
since we try our best to promote the use of EVP instead of the lower
level crypto algorithms.  Therefore, I put the inclusions back so
the application authors don't get too shocked by all the errors they
would otherwise get.

Thanks to Theo de Raadt for making us aware of this.

21 years agoMake sure buffers are large enough even for weird parameters
Bodo Möller [Wed, 26 Jun 2002 14:28:41 +0000 (14:28 +0000)]
Make sure buffers are large enough even for weird parameters

Submitted by: Nils Larsch

21 years agoupdate an entry on EVP changes
Bodo Möller [Wed, 26 Jun 2002 14:21:16 +0000 (14:21 +0000)]
update an entry on EVP changes

21 years agoFor some reason, we need to return the full path to perl
Richard Levitte [Tue, 25 Jun 2002 11:40:49 +0000 (11:40 +0000)]
For some reason, we need to return the full path to perl