Adapt builtin cipher implementations to opaque EVP_CIPHER

They all stop including evp_locl.h, so we also take care of their
adaptation to opaque EVP_CIPHER_CTX, as was promised in an earlier
commit.

Reviewed-by: Rich Salz <rsalz@openssl.org>

Adapt the internal EVP routines to opaque EVP_CIPHER

Reviewed-by: Rich Salz <rsalz@openssl.org>

Make EVP_CIPHER opaque and add creator/destructor/accessor/writer functions

We follow the method used for EVP_MD.
Also, move all the internal EVP_CIPHER building macros from evp_locl.h
to evp_int.h.  This will benefit our builtin EVP_CIPHERs.

Reviewed-by: Rich Salz <rsalz@openssl.org>

Remove EVP_CIPHER_CTX_flags, it's only confusing

EVP_CIPHER_CTX_flags was returning the flags of its associated
EVP_CIPHER.  However, EVP_CIPHER_CTX has flags of its own, so this
function is quite confusing and therefore error prone.

Reviewed-by: Rich Salz <rsalz@openssl.org>

Adapt all EVP_CIPHER_CTX users for it becoming opaque

Reviewed-by: Rich Salz <rsalz@openssl.org>

Adapt cipher implementations to opaque EVP_CIPHER_CTX

Note: there's a larger number of implementations in crypto/evp/ that
aren't affected because they include evp_locl.h.  They will be handled
in a separate commit.

Reviewed-by: Rich Salz <rsalz@openssl.org>

Adapt the internal EVP routines to opaque EVP_CIPHER_CTX

Reviewed-by: Rich Salz <rsalz@openssl.org>

Add accessors and writers for EVP_CIPHER_CTX

New functions:

- EVP_CIPHER_CTX_encrypting()
- EVP_CIPHER_CTX_iv()
- EVP_CIPHER_CTX_iv_noconst()
- EVP_CIPHER_CTX_original_iv()
- EVP_CIPHER_CTX_buf_noconst()
- EVP_CIPHER_CTX_num()
- EVP_CIPHER_CTX_set_num()
- EVP_CIPHER_CTX_cipher_data()
- EVP_CIPHER_CTX_new_cipher_data()

Note that the accessors / writers for iv, buf and num may go away, as
those rather belong in the implementation's own structure (cipher_data)
when the implementation would affect them (that would be the case when
they are flagged EVP_CIPH_CUSTOM_IV or EVP_CIPH_FLAG_CUSTOM_CIPHER).

Reviewed-by: Rich Salz <rsalz@openssl.org>

Make EVP_CIPHER_CTX opaque and renew the creator / destructor functions

Following the method used for EVP_MD_CTX and HMAC_CTX,
EVP_CIPHER_CTX_init and EVP_CIPHER_CTX_cleanup are joined together
into one function, EVP_CIPHER_CTX_reset, with EVP_CIPHER_CTX_init kept
as an alias.
EVP_CIPHER_CTX_cleanup fills no purpose of its own any more and is
therefore removed.

Reviewed-by: Rich Salz <rsalz@openssl.org>

RT4227: Range-check in apps.

Implement range-checking in all counts in apps.  Turns out only a couple
of cases were missing.  And make the range-checking code more strict.
Replace almost all opt_ulong() calls with opt_long()

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Yet another make update.

Reviewed-by: Richard Levitte <levitte@openssl.org>

Make sure to have both upper and mixed case symbols in SYMBOL_VECTOR

It was assumed that the syntax FOO/Foo=PROCEDURE would create both an
upper case and mixed case symbol in the GST.  Not so, it requires
having both FOO/Foo=PROCEDURE (to create the upper case alias) and
Foo=PROCEDURE (to create the mixed case slot).

We make sure that any symbol always occupies two slots (even those
that don't exist) by filling up with SPARE when necessary.  That will
assure that any changes will still have the same symbols in the same
slots no matter what (save a complete rewrite of the ordinals files).

Reviewed-by: Rich Salz <rsalz@openssl.org>

Use SPARE instead of PRIVATE_PROCEDURE to reserve unused SYMBOL_VECTOR slots

It was assumed that a dummy with the type PRIVATE_PROCEDURE would
simply occupy a slot but otherwise ignore the symbol.  Not so, but
there is SPARE for that purpose.

Reviewed-by: Rich Salz <rsalz@openssl.org>

The limit per SYMBOL_VECTOR isn't the amount of symbols, it's the line length

Reviewed-by: Rich Salz <rsalz@openssl.org>

VMS INDENTIFICATION should only have the version number

Reviewed-by: Rich Salz <rsalz@openssl.org>

GH528: "cipher -v" output is confusing.

Fix the docs, and refactor some common code.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Have mkdef.pl use case sensitive symbols for shareable symbol vector

Reviewed-by: Rich Salz <rsalz@openssl.org>

Function pop_info() returned a dangling pointer

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>

update ordinals

Reviewed-by: Rich Salz <rsalz@openssl.org>

Use ossl_inline and DEFINE_LHASH_OF

Reviewed-by: Rich Salz <rsalz@openssl.org>

Remove mkstack.pl: it is no longer needed.

Reviewed-by: Rich Salz <rsalz@openssl.org>

Add lh_doall_arg inlining

Reviewed-by: Rich Salz <rsalz@openssl.org>

Add lh_doall inlining

Reviewed-by: Rich Salz <rsalz@openssl.org>

Add lh_new() inlining

Reviewed-by: Rich Salz <rsalz@openssl.org>

Inline LHASH_OF

Make LHASH_OF use static inline functions.

Add new lh_get_down_load and lh_set_down_load functions and their
typesafe inline equivalents.

Make lh_error a function instead of a macro.

Reviewed-by: Rich Salz <rsalz@openssl.org>

GH540: add casts to safestack.h

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Dr. Stephen Henson <steve@openssl.org>

Fix jpaketest compilation error.

Reviewed-by: Rich Salz <rsalz@openssl.org>

Only run DANE tests when EC is supported

Since danetest is to test DANE rather than specific algorithms, it's
acceptable to require EC when testing it.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Make SSL{_CTX,}_{get,set,clear}_options functions

These now take and return unsigned long, and get is constified.
Updated related documentation and util/ssleay.num

Reviewed-by: Matt Caswell <matt@openssl.org>

Adjust $default_depflags to changes in Configure

We use $default_depflags to check if a 'make depend' is needed after
configuring, so it needs to be kept up to date.

Reviewed-by: Matt Caswell <matt@openssl.org>

Fix NSS format session output

Commit 189ae368d91 (RT ticket 3352) provided the capability to output
session key data in NSS format. The big apps cleanup broke that capability.
This commit restores it.

RT#4201

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Update leak test to check return values.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Add memory leak return value.

Make CRYPTO_mem_leaks() and CRYPTO_mem_leaks_fp() return a status value.
Update documentation. Don't abort() if there are leaks.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Enable/disable crypto-mdebug just like other features

Also always abort() on leak failure.

Reviewed-by: Stephen Henson <steve@openssl.org>

Disable some algorithms by default

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Regenerate SSL record/statem error strings

Reviewed-by: Rich Salz <rsalz@openssl.org>

fix shadow warning

Reviewed-by: Rich Salz <rsalz@openssl.org>

File is about s_time, not s_client

Reviewed-by: Dr. Stephen Henson <steve@openssl.org>

Allow disabling the min and max version

Reviewed-by: Viktor Dukhovni <openssl-users@dukhovni.org>

Future-proof deprecated declartion parsing

Support two-digest major/minor/micro numbers and die when the input
fails to parse

Reviewed-by: Dr. Stephen Henson <steve@openssl.org>

RT41897: Add an CRYPTO_secure_actual_size

This is already documented, I just forgot to include the code :)

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Fix c++ compilation

Fixes https://github.com/openssl/openssl/issues/532

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>

Adds missing type casts

This fixes https://github.com/openssl/openssl/issues/534

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>

Adds crypto-mdebug as a valid option

Adds crypto-mdebug as a valid option. Fixes https://github.com/openssl/openssl/issues/537

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>

Fix build failure on Windows

Fixes https://github.com/openssl/openssl/issues/530

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>

update ordinals

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Add DEPRECATEDIN support.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Correct header defines

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

remove hard coded algorithms

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Recognise disabled algorithms automatically.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Fix no CRYPTO_MDEBUG build (windows)

In order for mkdep to find #ifdef'd functions, they must be
wrapped (in the header file) with
        #ifndef OPENSSL_NO_...
So do that for various CRYPTO_mem_debug... things.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Fixup actually update danetest.c

Reviewed-by: Rich Salz <rsalz@openssl.org>

Another portability fix.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Update comment as bn_dup_expand is gone

Reviewed-by: Rich Salz <rsalz@openssl.org>

Simplify deprecated declaration exception

Reviewed-by: Rich Salz <rsalz@openssl.org>

Portability fix for apps/s_client.c

Make some local variables and a table of them be static.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

DANE CHANGES

Reviewed-by: Richard Levitte <levitte@openssl.org>

Fix some typos in comments

Reviewed-by: Richard Levitte <levitte@openssl.org>

Backwards-compatibility subject to OPENSSL_API_COMPAT

Provide backwards-compatiblity for functions, macros and include
files if OPENSSL_API_COMPAT is either not defined or defined less
than the version number of the release in which the feature was
deprecated.

Reviewed-by: Richard Levitte <levitte@openssl.org>

DANE s_client support

Reviewed-by: Richard Levitte <levitte@openssl.org>

Remove all remaining traces if PEM_Seal

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Minor test update

* Remove extraneous test/Makefile.orig
* Use basedomain instead of argv[1] in test/danetest.c

Reviewed-by: Rich Salz <rsalz@openssl.org>

Fix another build break for no-mem-debug

Reviewed-by: Richard Levitte <levitte@openssl.org>

Fix build-break; 'make update'

Commit bbd86bf5424a611cb6b77a3a17fc522931c4dcb8 broke certain builds.
Commit 0674427f71ca050e3c61a7ec9dc71b208c3b39f5 missing 'make update'

Reviewed-by: Richard Levitte <levitte@openssl.org>

mem functions cleanup

Only two macros CRYPTO_MDEBUG and CRYPTO_MDEBUG_ABORT to control this.
If CRYPTO_MDEBUG is not set, #ifdef out the whole debug machinery.
        (Thanks to Jakob Bohm for the suggestion!)
Make the "change wrapper functions" be the only paradigm.
Wrote documentation!
Format the 'set func' functions so their paramlists are legible.
Format some multi-line comments.
Remove ability to get/set the "memory debug" functions at runtme.
Remove MemCheck_* and CRYPTO_malloc_debug_init macros.
Add CRYPTO_mem_debug(int flag) function.
Add test/memleaktest.
Rename CRYPTO_malloc_init to OPENSSL_malloc_init; remove needless calls.

Reviewed-by: Richard Levitte <levitte@openssl.org>

Remove the old VMS linker option file creator for shlibs

Reviewed-by: Tim Hudson <tjh@openssl.org>

Enhance util/mkdef.pl to provide a VMS linker option file for shlibs

Reviewed-by: Tim Hudson <tjh@openssl.org>

Remove crypto/pem/pem_seal.c

It's functionality appears unused.  If we're wrong, we will revert.

Reviewed-by: Rich Salz <rsalz@openssl.org>

DANE support for X509_verify_cert()

Reviewed-by: Richard Levitte <levitte@openssl.org>

use more descriptive name DEFINE_STACK_OF_CONST

Reviewed-by: Richard Levitte <levitte@openssl.org>

Only declare stacks in headers

Don't define stacks in C source files: it causes warnings
about unused functions in some compilers.

Reviewed-by: Richard Levitte <levitte@openssl.org>

Rename DECLARE*STACK_OF to DEFINE*STACK_OF

Applications wishing to include their own stacks now just need to include

DEFINE_STACK_OF(foo)

in a header file.

Reviewed-by: Richard Levitte <levitte@openssl.org>

remove unused PREDECLARE

Reviewed-by: Richard Levitte <levitte@openssl.org>

Fix declarations and constification for inline stack.

Reviewed-by: Richard Levitte <levitte@openssl.org>

Change STACK_OF to use inline functions.

Change DECLARE_STACK_OF into inline functions. This avoids the need for
auto generated mkstack.pl macros and now handles const properly.

Reviewed-by: Richard Levitte <levitte@openssl.org>

DANE make update

Reviewed-by: Rich Salz <rsalz@openssl.org>

DANE documentation typos

Reported-by: Claus Assmann
Reviewed-by: Rich Salz <rsalz@openssl.org>

Remove more (rest?) of FIPS build stuff.

Reviewed-by: Dr. Stephen Henson <steve@openssl.org>

Remove some unused perl scripts

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

DANE support structures, constructructors and accessors

Also tweak some of the code in demos/bio, to enable interactive
testing of BIO_s_accept's use of SSL_dup.  Changed the sconnect
client to authenticate the server, which now exercises the new
SSL_set1_host() function.

Reviewed-by: Richard Levitte <levitte@openssl.org>

Fix X509_STORE_CTX_cleanup()

Reviewed-by: Dr. Stephen Henson <steve@openssl.org>

Drop incorrect id == -1 case from X509_check_trust

Reviewed-by: Richard Levitte <levitte@openssl.org>

X509_verify_cert() cleanup

Reviewed-by: Richard Levitte <levitte@openssl.org>

Cleanup of verify(1) failure output

Reviewed-by: Richard Levitte <levitte@openssl.org>

Instead of a local hack, implement SIZE_MAX in numbers.h if it's missing

Reviewed-by: Stephen Henson <steve@openssl.org>

Fix a possible memleak

If there's a failure allocating md_data, the destination pctx will have
a shared pointer with the source EVP_MD_CTX, which will lead to problems
when either the source or the destination is freed.

Reviewed-by: Stephen Henson <steve@openssl.org>

Protocol version selection and negotiation rewrite

The protocol selection code is now consolidated in a few consecutive
short functions in a single file and is table driven.  Protocol-specific
constraints that influence negotiation are moved into the flags
field of the method structure.  The same protocol version constraints
are now applied in all code paths.  It is now much easier to add
new protocol versions without reworking the protocol selection
logic.

In the presence of "holes" in the list of enabled client protocols
we no longer select client protocols below the hole based on a
subset of the constraints and then fail shortly after when it is
found that these don't meet the remaining constraints (suiteb, FIPS,
security level, ...).  Ideally, with the new min/max controls users
will be less likely to create "holes" in the first place.

Reviewed-by: Kurt Roeckx <kurt@openssl.org>

Refine and re-wrap Min/Max protocol docs

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Add support for minimum and maximum protocol version

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

Fix no-dh.

Reviewed-by: Stephen Henson <steve@openssl.org>

remove invalid free

Reviewed-by: Tim Hudson <tjh@openssl.org>

Use X509_get0_pubkey where appropriate

Reviewed-by: Rich Salz <rsalz@openssl.org>

Update to SHA256 for TSA signing digest.

Reviewed-by: Richard Levitte <levitte@openssl.org>

Fix faulty check in the VMS version of opt_progname

Reviewed-by: Stephen Henson <steve@openssl.org>

Remove the #ifndef OPENSSL_SYS_VMS around SSL_add_dir_cert_subjects_to_stack

It served a purpose, but not any more.

Reviewed-by: Stephen Henson <steve@openssl.org>

Correct missing prototype

Reviewed-by: Tim Hudson <tjh@openssl.org>

SIZE_MAX doesn't exist everywhere, supply an alternative

SIZE_MAX is a great macro, and does unfortunately not exist everywhere.
Since we check against half of it, using bitwise shift to calculate the
value of half SIZE_MAX should be safe enough.

Reviewed-by: Tim Hudson <tjh@openssl.org>

Fix some missing or faulty header file inclusions

Reviewed-by: Tim Hudson <tjh@openssl.org>

Check for missing DSA parameters.

If DSA parameters are absent return -1 (for unknown) in DSA_security_bits.

If parameters are absent when a certificate is set in an SSL/SSL_CTX
structure this will reject the certificate by default. This will cause DSA
certificates which omit parameters to be rejected but that is never (?)
done in practice.

Thanks to Brian 'geeknik' Carpenter for reporting this issue.

Reviewed-by: Emilia Käsper <emilia@openssl.org>

Convert RSA encrypt to use EVP_PKEY

Reviewed-by: Rich Salz <rsalz@openssl.org>