22 years agoint may be smaller than 32 bits.
Bodo Möller [Mon, 5 Jun 2000 13:50:57 +0000 (13:50 +0000)]
int may be smaller than 32 bits.

22 years agoAccording to Gordon Atwood <>, GNU C on SunOS
Richard Levitte [Sun, 4 Jun 2000 07:17:28 +0000 (07:17 +0000)]
According to Gordon Atwood <>, GNU C on SunOS
4.1.4 uses libiberty to define strtoul and strerror.

22 years agoAccording to Gordon Atwood <>, stdlib.h is
Richard Levitte [Sun, 4 Jun 2000 07:12:20 +0000 (07:12 +0000)]
According to Gordon Atwood <>, stdlib.h is
needed, or size_t won't be defined on SunOS 4.1.4.

22 years agoTypo...
Richard Levitte [Sun, 4 Jun 2000 07:08:29 +0000 (07:08 +0000)]

22 years agoCygWin32 support.
Ulf Möller [Sat, 3 Jun 2000 23:23:10 +0000 (23:23 +0000)]
CygWin32 support.

Submitted by: John Jarvie <>

22 years agoYet another bc FAQ.
Ulf Möller [Sat, 3 Jun 2000 23:21:43 +0000 (23:21 +0000)]
Yet another bc FAQ.

22 years agoEVP constification.
Ben Laurie [Sat, 3 Jun 2000 14:13:58 +0000 (14:13 +0000)]
EVP constification.

22 years agoThere have been a number of complaints from a number of sources that names
Richard Levitte [Thu, 1 Jun 2000 22:19:21 +0000 (22:19 +0000)]
There have been a number of complaints from a number of sources that names
like Malloc, Realloc and especially Free conflict with already existing names
on some operating systems or other packages.  That is reason enough to change
the names of the OpenSSL memory allocation macros to something that has a
better chance of being unique, like prepending them with OPENSSL_.

This change includes all the name changes needed throughout all C files.

22 years agoUse NO_FP_API.
Ulf Möller [Thu, 1 Jun 2000 20:25:44 +0000 (20:25 +0000)]

22 years agoDon't include <stdlib.h>. In the NO_FP_API case, don't include <stdio.h>.
Ulf Möller [Thu, 1 Jun 2000 19:34:25 +0000 (19:34 +0000)]
Don't include <stdlib.h>. In the NO_FP_API case, don't include <stdio.h>.

22 years agoSmall documentation bug, probably a cut'n'paste, corrected.
Richard Levitte [Thu, 1 Jun 2000 17:40:34 +0000 (17:40 +0000)]
Small documentation bug, probably a cut'n'paste, corrected.

22 years agoBug fix for 64 bit HP-UX.
Ulf Möller [Thu, 1 Jun 2000 14:24:59 +0000 (14:24 +0000)]
Bug fix for 64 bit HP-UX.

Submitted by: Karsten Spang <>

22 years agoSmall fix to enable reading from stdin as well.
Richard Levitte [Thu, 1 Jun 2000 11:23:20 +0000 (11:23 +0000)]
Small fix to enable reading from stdin as well.
Contributed by Yoichiro Okabe <>

22 years ago"make update" + stripping the type-specific stack functions out of
Geoff Thorpe [Thu, 1 Jun 2000 06:07:19 +0000 (06:07 +0000)]
"make update" + stripping the type-specific stack functions out of
libeay.num and ssleay.num.

22 years agoThis change will cause builds (by default) to not use different STACK
Geoff Thorpe [Thu, 1 Jun 2000 05:13:52 +0000 (05:13 +0000)]
This change will cause builds (by default) to not use different STACK
structures and functions for each stack type. The previous behaviour
can be enabled by configuring with the "-DDEBUG_SAFESTACK" option.
This will also cause "make update" ( in particular) to
update the libeay.num and ssleay.num symbol tables with the number of
extra functions DEBUG_SAFESTACK creates.

The way this change works is to accompany each DECLARE_STACK_OF()
macro with a set of "#define"d versions of the sk_##type##_***
functions that ensures all the existing "type-safe" stack calls are
precompiled into the underlying stack calls. The presence or abscence
of the DEBUG_SAFESTACK symbol controls whether this block of
"#define"s or the DECLARE_STACK_OF() macro is taking effect. The
block of "#define"s is in turn generated and maintained by a perl
script (util/ that encompasses the block with delimiting
C comments. This works in a similar way to the auto-generated error
codes and, like the other such maintenance utilities, is invoked
by the "make update" target.

A long (but mundane) commit will follow this with the results of
"make update" - this will include all the "#define" blocks for
each DECLARE_STACK_OF() statement, along with stripped down
libeay.num and ssleay.num files.

22 years agoThe previous commit to crypto/stack/*.[ch] pulled the type-safety strings
Geoff Thorpe [Thu, 1 Jun 2000 02:36:58 +0000 (02:36 +0000)]
The previous commit to crypto/stack/*.[ch] pulled the type-safety strings
yet tighter, and also put some heat on the rest of the library by
insisting (correctly) that compare callbacks used in stacks are prototyped
with "const" parameters. This has led to a depth-first explosion of
compiler warnings in the code where 1 constification has led to 3 or 4
more. Fortunately these have all been resolved to completion and the code
seems cleaner as a result - in particular many of the _cmp() functions
should have been prototyped with "const"s, and now are. There was one
little problem however;

X509_cmp() should by rights compare "const X509 *" pointers, and it is now
declared as such. However, it's internal workings can involve
recalculating hash values and extensions if they have not already been
setup. Someone with a more intricate understanding of the flow control of
X509 might be able to tighten this up, but for now - this seemed the
obvious place to stop the "depth-first" constification of the code by
using an evil cast (they have migrated all the way here from safestack.h).

Fortunately, this is the only place in the code where this was required
to complete these type-safety changes, and it's reasonably clear and
commented, and seemed the least unacceptable of the options. Trying to
take the constification further ends up exploding out considerably, and
indeed leads directly into generalised ASN functions which are not likely
to cooperate well with this.

22 years agoThis is the first of two commits (didn't want to dump them all into the
Geoff Thorpe [Thu, 1 Jun 2000 02:15:40 +0000 (02:15 +0000)]
This is the first of two commits (didn't want to dump them all into the
same one). However, the first will temporarily break things until the
second comes through. :-)

The safestack.h handling was mapping compare callbacks that externally
are of the type (int (*)(type **,type **)) into the underlying callback
type used by stack.[ch], which is (int (*)(void *,void *)). After some
degree of digging, it appears that the callback type in the underlying
stack code should use double pointers too - when the compare operations
are invoked (from sk_find and sk_sort), they are being used by bsearch
and qsort to compare two pointers to pointers. This change corrects the
prototyping (by only casting to the (void*,void*) form at the moment
it is needed by bsearch and qsort) and makes the mapping in safestack.h
more transparent. It also changes from "void*" to "char*" to stay in
keeping with stack.[ch]'s assumed base type of "char".

Also - the "const" situation was that safestack.h was throwing away
"const"s, and to compound the problem - a close examination of stack.c
showed that (const char **) is not really achieving what it is supposed
to when the callback is being invoked, what is needed is
(const char * const *). So the underlying stack.[ch] and the mapping
macros in safestack.h have all been altered to correct this.

What will follow are the vast quantities of "const" corrections required
in stack-dependant code that was being let "slip" through when
safestack.h was discarding "const"s. These now all come up as compiler

22 years agois needed.
Ulf Möller [Thu, 1 Jun 2000 00:27:59 +0000 (00:27 +0000)]
is needed.

22 years agotypo
Bodo Möller [Wed, 31 May 2000 23:20:10 +0000 (23:20 +0000)]

22 years agodh and gendh have been obsoleted by dhparam.
Bodo Möller [Wed, 31 May 2000 23:07:48 +0000 (23:07 +0000)]
dh and gendh have been obsoleted by dhparam.

22 years ago#include <stdio.h> not needed.
Ulf Möller [Wed, 31 May 2000 22:25:44 +0000 (22:25 +0000)]
#include <stdio.h> not needed.

22 years ago#include <stdlib.h> is not needed.
Ulf Möller [Wed, 31 May 2000 22:13:59 +0000 (22:13 +0000)]
#include <stdlib.h> is not needed.

22 years agoResult of "make update"
Richard Levitte [Wed, 31 May 2000 21:35:38 +0000 (21:35 +0000)]
Result of "make update"

22 years agoThis seems to become a FAQ...
Richard Levitte [Wed, 31 May 2000 20:11:11 +0000 (20:11 +0000)]
This seems to become a FAQ...

22 years agosk_***_new_null() seems to be there to avoid exactly this sort of thing
Geoff Thorpe [Wed, 31 May 2000 17:41:34 +0000 (17:41 +0000)]
sk_***_new_null() seems to be there to avoid exactly this sort of thing
which is a cast between NULL and a function pointer.

22 years agosk_value was also suffering from de-const-ification.
Geoff Thorpe [Wed, 31 May 2000 17:35:11 +0000 (17:35 +0000)]
sk_value was also suffering from de-const-ification.
Also, add in a couple of missing declarations in pkcs7 code.

22 years agoconfig can now detect ARM Linux automagically.
Richard Levitte [Wed, 31 May 2000 17:06:10 +0000 (17:06 +0000)]
config can now detect ARM Linux automagically.
Contributed by Jeremy Norris <>

22 years agostrtoul is not used anywhere.
Bodo Möller [Wed, 31 May 2000 16:44:22 +0000 (16:44 +0000)]
strtoul is not used anywhere.

22 years agoWhen compiling with /opt/SUNWspro/SC4.2/bin/cc on Solaris, __svr4__ is
Bodo Möller [Wed, 31 May 2000 16:42:47 +0000 (16:42 +0000)]
When compiling with /opt/SUNWspro/SC4.2/bin/cc on Solaris, __svr4__ is
not defined, but __SVR4 is.

22 years agoAll the little functions created by the IMPLEMENT_STACK_OF() macro will
Geoff Thorpe [Wed, 31 May 2000 15:28:01 +0000 (15:28 +0000)]
All the little functions created by the IMPLEMENT_STACK_OF() macro will
cast their type-specific STACK into a real STACK and call the underlying
sk_*** function. The problem is that if the STACK_OF(..) parameter being
passed in has a "const *" qualifier, it is discarded by the cast.

I'm currently implementing a fix for this but in the mean-time, this is
one case I noticed (a few type-specific sk_**_num() functions pass in
const type-specific stacks). If there are other errors in the code where
consts are being discarded, we will similarly not notice them. yuck.

22 years agoMove Windows seeding functions into a separate file.
Bodo Möller [Wed, 31 May 2000 12:48:35 +0000 (12:48 +0000)]
Move Windows seeding functions into a separate file.
They have nothing to do with the particular PRNG (md_rand.c).

22 years agoAdd "FIXME" comment.
Bodo Möller [Wed, 31 May 2000 09:51:55 +0000 (09:51 +0000)]
Add "FIXME" comment.

22 years agoImprove PRNG robustness.
Bodo Möller [Tue, 30 May 2000 21:44:36 +0000 (21:44 +0000)]
Improve PRNG robustness.

22 years agoMore EVP cipher revision.
Dr. Stephen Henson [Tue, 30 May 2000 18:26:22 +0000 (18:26 +0000)]
More EVP cipher revision.

Change EVP_SealInit() and EVP_OpenInit() to
handle cipher parameters.

Make it possible to set RC2 and RC5 params.

Make RC2 ASN1 code use the effective key bits
and not the key length.

TODO: document how new API works.

22 years ago"make update"
Geoff Thorpe [Tue, 30 May 2000 12:59:43 +0000 (12:59 +0000)]
"make update"

Also, corrects the linux-elf-arm config string, it was previously setting
$des_obj = dlfcn :-)

22 years agoYou must have an empty line between =item's
Richard Levitte [Tue, 30 May 2000 08:01:24 +0000 (08:01 +0000)]
You must have an empty line between =item's

22 years agoFourth phase EVP revision.
Dr. Stephen Henson [Tue, 30 May 2000 02:21:15 +0000 (02:21 +0000)]
Fourth phase EVP revision.

Declare ciphers in terms of macros. This reduces
the amount of code and places each block cipher EVP
definition in a single file instead of being spread
over 4 files.

22 years agoFourth phase EVP revision.
Dr. Stephen Henson [Tue, 30 May 2000 02:10:57 +0000 (02:10 +0000)]
Fourth phase EVP revision.

Declare ciphers in terms of macros. This reduces
the amount of code and places each block cipher EVP
definition in a single file instead of being spread
over 4 files.

22 years agoConfiguration for linux on ARM (contributed by Jeremy Norris
Richard Levitte [Mon, 29 May 2000 16:18:37 +0000 (16:18 +0000)]
Configuration for linux on ARM (contributed by Jeremy Norris

22 years agoMinor corrections to documentation.
Geoff Thorpe [Mon, 29 May 2000 15:52:21 +0000 (15:52 +0000)]
Minor corrections to documentation.

* speed processes any/all options passed to it, not just one.
* DH and DSA have no "_get_method()" functions, only RSA does.
* typos.

22 years agoanother warning.
Geoff Thorpe [Mon, 29 May 2000 04:20:52 +0000 (04:20 +0000)]
another warning.

22 years agoIt seems that mktime does what is required here. Certainly timegm() can
Geoff Thorpe [Mon, 29 May 2000 03:50:37 +0000 (03:50 +0000)]
It seems that mktime does what is required here. Certainly timegm() can
not be used because it is not available on all systems (most notably,

22 years agoThis declaration seems to have been added into the header file accidently.
Geoff Thorpe [Mon, 29 May 2000 03:17:45 +0000 (03:17 +0000)]
This declaration seems to have been added into the header file accidently.
There's no trace of it being implemented and it doesn't seem to have been
intended given that it is prototyped with a BIO yet there was a BIO-
specific version added in at the same time.

22 years agooops.
Geoff Thorpe [Mon, 29 May 2000 02:42:57 +0000 (02:42 +0000)]

22 years agoThird phase of EVP cipher overhaul.
Dr. Stephen Henson [Sun, 28 May 2000 12:44:46 +0000 (12:44 +0000)]
Third phase of EVP cipher overhaul.

Remove duplicated code in EVP.

22 years agoBugfix: clear error queue after ignoring ssl_verify_cert_chain result.
Bodo Möller [Sat, 27 May 2000 22:25:01 +0000 (22:25 +0000)]
Bugfix: clear error queue after ignoring ssl_verify_cert_chain result.

22 years agoSecond phase of EVP cipher overhaul.
Dr. Stephen Henson [Sat, 27 May 2000 12:38:43 +0000 (12:38 +0000)]
Second phase of EVP cipher overhaul.

Change functions like EVP_EncryptUpdate() so they now return a
value. These normally have software only implementations
which cannot fail so this was acceptable. However ciphers
can be implemented in hardware and these could return errors.

22 years agoBeginnings of EVP cipher overhaul. This should eventually
Dr. Stephen Henson [Fri, 26 May 2000 23:51:35 +0000 (23:51 +0000)]
Beginnings of EVP cipher overhaul. This should eventually
enhance and tidy up the EVP interface.

This patch adds initial support for variable length ciphers
and changes S/MIME code to use this.

Some other library functions need modifying to support use
of modified cipher parameters.

Also need to change all the cipher functions that should
return error codes, but currenly don't.

And of course it needs extensive testing...

22 years agocvs update
Geoff Thorpe [Thu, 25 May 2000 13:20:11 +0000 (13:20 +0000)]
cvs update

22 years agotypo
Bodo Möller [Thu, 25 May 2000 11:24:42 +0000 (11:24 +0000)]

22 years agoImplement SSL_OP_TLS_ROLLBACK_BUG for servers.
Bodo Möller [Thu, 25 May 2000 09:50:40 +0000 (09:50 +0000)]
Implement SSL_OP_TLS_ROLLBACK_BUG for servers.

Call dh_tmp_cb with correct 'is_export' flag.

Avoid tabs in CHANGES.

22 years agoAdd the target system linux-m68k
Richard Levitte [Wed, 24 May 2000 22:00:59 +0000 (22:00 +0000)]
Add the target system linux-m68k

22 years agoAdd DSA library string. Workaround for IIS .key file invalid
Dr. Stephen Henson [Wed, 24 May 2000 13:09:59 +0000 (13:09 +0000)]
Add DSA library string. Workaround for IIS .key file invalid
ASN1 encoding.

22 years agoFix Blowfish URL.
Ralf S. Engelschall [Tue, 23 May 2000 18:33:44 +0000 (18:33 +0000)]
Fix Blowfish URL.

Submitted by: Arnaud De Timmerman <>

22 years agoYet more typesafety.
Ben Laurie [Sun, 21 May 2000 15:21:55 +0000 (15:21 +0000)]
Yet more typesafety.

22 years agoAvoid sprintf
Bodo Möller [Sun, 21 May 2000 14:21:24 +0000 (14:21 +0000)]
Avoid sprintf

22 years agoAvoid sprintf.
Bodo Möller [Sun, 21 May 2000 14:17:01 +0000 (14:17 +0000)]
Avoid sprintf.

22 years agoAvoid sprintf, and harmonize indentation.
Bodo Möller [Sun, 21 May 2000 14:14:30 +0000 (14:14 +0000)]
Avoid sprintf, and harmonize indentation.

22 years agoAvoid sprintf.
Bodo Möller [Sun, 21 May 2000 14:10:05 +0000 (14:10 +0000)]
Avoid sprintf.

22 years ago'make update'
Richard Levitte [Fri, 19 May 2000 13:44:28 +0000 (13:44 +0000)]
'make update'

22 years agoRemove extra comma (creates a unnessecary null element, right?).
Richard Levitte [Fri, 19 May 2000 13:36:29 +0000 (13:36 +0000)]
Remove extra comma (creates a unnessecary null element, right?).

22 years agoAdd required cast.
Bodo Möller [Fri, 19 May 2000 12:02:49 +0000 (12:02 +0000)]
Add required cast.

22 years agoFix "FIXME" indentation :-)
Bodo Möller [Fri, 19 May 2000 12:02:09 +0000 (12:02 +0000)]
Fix "FIXME" indentation :-)

22 years agoAdd "FIXME" comment, and adjust the indentation.
Bodo Möller [Fri, 19 May 2000 11:59:55 +0000 (11:59 +0000)]
Add "FIXME" comment, and adjust the indentation.

22 years agoAdd a couple of macros that make OpenSSL compilable on SunOS 4.1.4.
Richard Levitte [Fri, 19 May 2000 08:53:06 +0000 (08:53 +0000)]
Add a couple of macros that make OpenSSL compilable on SunOS 4.1.4.
Contributed by SAKAI Kiyotaka <>

22 years agotypo
Bodo Möller [Fri, 19 May 2000 07:54:42 +0000 (07:54 +0000)]

22 years agoAdd a note about the new document.
Richard Levitte [Thu, 18 May 2000 21:25:48 +0000 (21:25 +0000)]
Add a note about the new document.

22 years agoAdd a new file where all the standards and other documents that we try
Richard Levitte [Thu, 18 May 2000 21:22:50 +0000 (21:22 +0000)]
Add a new file where all the standards and other documents that we try
to adhere to are listed.  It should be regarded as a complement to
whatever is out on the web, including the docs in

22 years agoFix c_rehash script, add -fingerprint option to crl.
Dr. Stephen Henson [Thu, 18 May 2000 00:33:00 +0000 (00:33 +0000)]
Fix c_rehash script, add -fingerprint option to crl.

22 years agoTypesafety Thought Police last part.
Ben Laurie [Wed, 17 May 2000 10:08:05 +0000 (10:08 +0000)]
Typesafety Thought Police last part.

22 years agoTypesafety Thought Police part 5.
Ben Laurie [Wed, 17 May 2000 09:13:36 +0000 (09:13 +0000)]
Typesafety Thought Police part 5.

22 years agoTypesafe Thought Police part 4.
Ben Laurie [Tue, 16 May 2000 23:01:19 +0000 (23:01 +0000)]
Typesafe Thought Police part 4.

22 years agoTypesafety Thought Police part 3.
Ben Laurie [Tue, 16 May 2000 21:22:45 +0000 (21:22 +0000)]
Typesafety Thought Police part 3.

22 years agoTypesafety Thought Police Part 2.
Ben Laurie [Tue, 16 May 2000 19:53:50 +0000 (19:53 +0000)]
Typesafety Thought Police Part 2.

22 years agoTypesafety thought police.
Ben Laurie [Tue, 16 May 2000 14:38:29 +0000 (14:38 +0000)]
Typesafety thought police.

22 years agoGet rid of more non-ANSI declarations.
Ulf Möller [Mon, 15 May 2000 22:54:43 +0000 (22:54 +0000)]
Get rid of more non-ANSI declarations.

22 years agoprototype.
Ulf Möller [Mon, 15 May 2000 21:02:44 +0000 (21:02 +0000)]

22 years agoMake sure that NO-RSA applications etc can include evp.h
Ulf Möller [Mon, 15 May 2000 19:24:23 +0000 (19:24 +0000)]
Make sure that NO-RSA applications etc can include evp.h

22 years agoMissing cases when no_rsa is defined
Ulf Möller [Mon, 15 May 2000 19:20:10 +0000 (19:20 +0000)]
Missing cases when no_rsa is defined

Submitted by: Zeroknowledge

22 years agoMacOS changes.
Ulf Möller [Mon, 15 May 2000 18:59:55 +0000 (18:59 +0000)]
MacOS changes.

22 years agoAllow UTCTIME objects to be retrieved. Check for imminent cert expiry.
Ben Laurie [Sun, 14 May 2000 12:39:53 +0000 (12:39 +0000)]
Allow UTCTIME objects to be retrieved. Check for imminent cert expiry.

22 years agoWhen open()ing 'file' in RAND_write_file, don't use O_EXCL.
Bodo Möller [Thu, 11 May 2000 23:10:27 +0000 (23:10 +0000)]
When open()ing 'file' in RAND_write_file, don't use O_EXCL.
This is superfluous now that we don't have to avoid creating
multiple versions of the file on VMS (because older versions
are now deleted).

22 years agoBeing sick and tired of the hogging Efence does on my laptop, I
Richard Levitte [Mon, 8 May 2000 16:58:29 +0000 (16:58 +0000)]
Being sick and tired of the hogging Efence does on my laptop, I
decided to provide an alternative...

22 years agoFix for SSL server purpose checking
Dr. Stephen Henson [Thu, 4 May 2000 23:03:49 +0000 (23:03 +0000)]
Fix for SSL server purpose checking

22 years agoMake PKCS#12 code handle missing passwords.
Dr. Stephen Henson [Thu, 4 May 2000 00:08:35 +0000 (00:08 +0000)]
Make PKCS#12 code handle missing passwords.

Add a couple of FAQs.

22 years agoNote apps/x509.c bugfixes.
Bodo Möller [Tue, 2 May 2000 20:29:03 +0000 (20:29 +0000)]
Note apps/x509.c bugfixes.

22 years agoFix a memory leak, and don't generate inappropriate error message
Bodo Möller [Tue, 2 May 2000 20:18:48 +0000 (20:18 +0000)]
Fix a memory leak, and don't generate inappropriate error message
when PEM_read_bio_X509_REQ fails.

22 years agoInitialise.
Richard Levitte [Tue, 2 May 2000 13:38:11 +0000 (13:38 +0000)]

22 years agoIn Message-ID: <003201bfb332$14a07520$>,
Richard Levitte [Tue, 2 May 2000 13:36:50 +0000 (13:36 +0000)]
In Message-ID: <003201bfb332$14a07520$>,
"Jan Mikkelsen" <> correctly states that the
OpenSSL header files have #include's and extern "C"'s in an incorrect
order.  Thusly fixed.

Also, make the memory debugging routines defined and declared with
prototypes, and use void* instead of char* for memory blobs.

And last of all, redo the ugly callback construct for elegance and
better definition (with prototypes).

22 years agoIn Message-ID: <003201bfb332$14a07520$>,
Richard Levitte [Tue, 2 May 2000 12:35:04 +0000 (12:35 +0000)]
In Message-ID: <003201bfb332$14a07520$>,
"Jan Mikkelsen" <> correctly states that the
OpenSSL header files have #include's and extern "C"'s in an incorrect
order.  Thusly fixed.

22 years agoIn Message-ID: <003201bfb332$14a07520$>,
Richard Levitte [Tue, 2 May 2000 12:16:01 +0000 (12:16 +0000)]
In Message-ID: <003201bfb332$14a07520$>,
"Jan Mikkelsen" <> correctly states that the
OpenSSL header files have #include's and extern "C"'s in an incorrect
order.  Thusly fixed.

22 years agoAdd missing #include.
Bodo Möller [Mon, 1 May 2000 19:49:41 +0000 (19:49 +0000)]
Add missing #include.

22 years agoAvoid leaking memory in thread_hash (and enable memory leak detection
Bodo Möller [Sat, 29 Apr 2000 23:58:05 +0000 (23:58 +0000)]
Avoid leaking memory in thread_hash (and enable memory leak detection
for it).

22 years agolinux-elf bugfix
Ulf Möller [Thu, 27 Apr 2000 15:07:15 +0000 (15:07 +0000)]
linux-elf bugfix

Submitted by:
Reviewed by:

22 years agoSubmitted by:
Ulf Möller [Thu, 27 Apr 2000 15:06:26 +0000 (15:06 +0000)]
Submitted by:
Reviewed by:

22 years agoStylistic changes: Don't use a macro for the malloc'ed length since it
Bodo Möller [Thu, 27 Apr 2000 09:11:28 +0000 (09:11 +0000)]
Stylistic changes: Don't use a macro for the malloc'ed length since it
is not constant.

22 years agoWarn about truncation also in the case when a single password is read using
Bodo Möller [Thu, 27 Apr 2000 06:47:23 +0000 (06:47 +0000)]
Warn about truncation also in the case when a single password is read using
the password prompt.

22 years agoUse CONFerr, not RSAerr, in the conf library...
Richard Levitte [Wed, 26 Apr 2000 12:15:19 +0000 (12:15 +0000)]
Use CONFerr, not RSAerr, in the conf library...

22 years agoClarifications and removal of double declaration...
Richard Levitte [Wed, 26 Apr 2000 12:13:19 +0000 (12:13 +0000)]
Clarifications and removal of double declaration...

22 years agoThis case in the "dso_unload" handlers should not be reported as an error -
Geoff Thorpe [Tue, 25 Apr 2000 08:37:12 +0000 (08:37 +0000)]
This case in the "dso_unload" handlers should not be reported as an error -
if a DSO_load(NULL,...) operation fails, it will have to call DSO_free() on
the DSO structure it created and that will filter through to this "unload"

If the stack size is "< 1", then the library never actually loaded. To keep
things clean higher up, I'll treat this as a vacuous case without an error.
It makes the error stack easier to follow real world cases, and the error
this ignores was only useful for catching bugs in internal code, not
mismatched calls from applications (which should be handled in the generic
DSO layer).