aboutsummaryrefslogtreecommitdiffstats
path: root/doc/crypto
diff options
context:
space:
mode:
authorJung-uk Kim <jkim@FreeBSD.org>2012-07-11 23:31:36 +0000
committerJung-uk Kim <jkim@FreeBSD.org>2012-07-11 23:31:36 +0000
commit0758ab5ea778e4ba36d2150af1bba602a48d6467 (patch)
tree0c30591ac90cb5e07a0763793709fd1056b67f57 /doc/crypto
parent2b8b5455829304396e38200c205612c4dc57c052 (diff)
downloadsrc-0758ab5ea778e4ba36d2150af1bba602a48d6467.tar.gz
src-0758ab5ea778e4ba36d2150af1bba602a48d6467.zip
Import OpenSSL 1.0.1c.vendor/openssl/1.0.1c
Approved by: benl (maintainer)
Notes
Notes: svn path=/vendor-crypto/openssl/dist/; revision=238384 svn path=/vendor-crypto/openssl/1.0.1c/; revision=238385; tag=vendor/openssl/1.0.1c
Diffstat (limited to 'doc/crypto')
-rw-r--r--doc/crypto/ASN1_generate_nconf.pod3
-rw-r--r--doc/crypto/BIO_f_md.pod6
-rw-r--r--doc/crypto/BIO_f_ssl.pod9
-rw-r--r--doc/crypto/BIO_new_CMS.pod66
-rw-r--r--doc/crypto/BIO_s_file.pod4
-rw-r--r--doc/crypto/BIO_s_mem.pod2
-rw-r--r--doc/crypto/BN_BLINDING_new.pod20
-rw-r--r--doc/crypto/CMS_add0_cert.pod66
-rw-r--r--doc/crypto/CMS_add1_recipient_cert.pod62
-rw-r--r--doc/crypto/CMS_compress.pod73
-rw-r--r--doc/crypto/CMS_decrypt.pod65
-rw-r--r--doc/crypto/CMS_encrypt.pod96
-rw-r--r--doc/crypto/CMS_final.pod41
-rw-r--r--doc/crypto/CMS_get0_RecipientInfos.pod106
-rw-r--r--doc/crypto/CMS_get0_SignerInfos.pod75
-rw-r--r--doc/crypto/CMS_get0_type.pod63
-rw-r--r--doc/crypto/CMS_get1_ReceiptRequest.pod69
-rw-r--r--doc/crypto/CMS_sign.pod121
-rw-r--r--doc/crypto/CMS_sign_add1_signer.pod101
-rw-r--r--doc/crypto/CMS_sign_receipt.pod45
-rw-r--r--doc/crypto/CMS_uncompress.pod54
-rw-r--r--doc/crypto/CMS_verify.pod126
-rw-r--r--doc/crypto/CMS_verify_receipt.pod47
-rw-r--r--doc/crypto/DSA_get_ex_new_index.pod2
-rw-r--r--doc/crypto/EVP_DigestInit.pod83
-rw-r--r--doc/crypto/EVP_DigestSignInit.pod87
-rw-r--r--doc/crypto/EVP_DigestVerifyInit.pod82
-rw-r--r--doc/crypto/EVP_PKEY_CTX_ctrl.pod128
-rw-r--r--doc/crypto/EVP_PKEY_CTX_new.pod52
-rw-r--r--doc/crypto/EVP_PKEY_cmp.pod61
-rw-r--r--doc/crypto/EVP_PKEY_decrypt.pod93
-rw-r--r--doc/crypto/EVP_PKEY_derive.pod93
-rw-r--r--doc/crypto/EVP_PKEY_encrypt.pod93
-rw-r--r--doc/crypto/EVP_PKEY_get_default_digest.pod41
-rw-r--r--doc/crypto/EVP_PKEY_keygen.pod161
-rw-r--r--doc/crypto/EVP_PKEY_print_private.pod53
-rw-r--r--doc/crypto/EVP_PKEY_sign.pod96
-rw-r--r--doc/crypto/EVP_PKEY_verify.pod91
-rw-r--r--doc/crypto/EVP_PKEY_verifyrecover.pod103
-rw-r--r--doc/crypto/EVP_SignInit.pod9
-rw-r--r--doc/crypto/EVP_VerifyInit.pod9
-rw-r--r--doc/crypto/PEM_write_bio_CMS_stream.pod41
-rw-r--r--doc/crypto/PEM_write_bio_PKCS7_stream.pod41
-rw-r--r--doc/crypto/PKCS7_encrypt.pod57
-rw-r--r--doc/crypto/PKCS7_sign.pod111
-rw-r--r--doc/crypto/PKCS7_sign_add_signer.pod87
-rw-r--r--doc/crypto/SMIME_read_CMS.pod70
-rw-r--r--doc/crypto/SMIME_write_CMS.pod64
-rw-r--r--doc/crypto/SMIME_write_PKCS7.pod20
-rw-r--r--doc/crypto/X509_STORE_CTX_get_error.pod303
-rw-r--r--doc/crypto/X509_STORE_CTX_get_ex_new_index.pod41
-rw-r--r--doc/crypto/X509_STORE_CTX_new.pod122
-rw-r--r--doc/crypto/X509_STORE_CTX_set_verify_cb.pod161
-rw-r--r--doc/crypto/X509_STORE_set_verify_cb_func.pod54
-rw-r--r--doc/crypto/X509_VERIFY_PARAM_set_flags.pod171
-rw-r--r--doc/crypto/X509_verify_cert.pod53
-rw-r--r--doc/crypto/d2i_RSAPublicKey.pod8
-rw-r--r--doc/crypto/ecdsa.pod2
-rw-r--r--doc/crypto/evp.pod22
-rw-r--r--doc/crypto/hmac.pod20
-rw-r--r--doc/crypto/i2d_CMS_bio_stream.pod44
-rw-r--r--doc/crypto/i2d_PKCS7_bio_stream.pod44
-rw-r--r--doc/crypto/lhash.pod244
-rw-r--r--doc/crypto/threads.pod91
64 files changed, 4244 insertions, 284 deletions
diff --git a/doc/crypto/ASN1_generate_nconf.pod b/doc/crypto/ASN1_generate_nconf.pod
index ee8915917ecd..542fd1579ab8 100644
--- a/doc/crypto/ASN1_generate_nconf.pod
+++ b/doc/crypto/ASN1_generate_nconf.pod
@@ -103,7 +103,8 @@ bits is set to zero.
=item B<UNIVERSALSTRING>, B<UNIV>, B<IA5>, B<IA5STRING>, B<UTF8>,
B<UTF8String>, B<BMP>, B<BMPSTRING>, B<VISIBLESTRING>,
B<VISIBLE>, B<PRINTABLESTRING>, B<PRINTABLE>, B<T61>,
-B<T61STRING>, B<TELETEXSTRING>, B<GeneralString>
+B<T61STRING>, B<TELETEXSTRING>, B<GeneralString>, B<NUMERICSTRING>,
+B<NUMERIC>
These encode the corresponding string types. B<value> represents the
contents of this structure. The format can be B<ASCII> or B<UTF8>.
diff --git a/doc/crypto/BIO_f_md.pod b/doc/crypto/BIO_f_md.pod
index 0d24083e6db8..2cc41f89d2fd 100644
--- a/doc/crypto/BIO_f_md.pod
+++ b/doc/crypto/BIO_f_md.pod
@@ -58,6 +58,12 @@ If an application needs to call BIO_gets() or BIO_puts() through
a chain containing digest BIOs then this can be done by prepending
a buffering BIO.
+Before OpenSSL 1.0.0 the call to BIO_get_md_ctx() would only work if the BIO
+had been initialized for example by calling BIO_set_md() ). In OpenSSL
+1.0.0 and later the context is always returned and the BIO is state is set
+to initialized. This allows applications to initialize the context externally
+if the standard calls such as BIO_set_md() are not sufficiently flexible.
+
=head1 RETURN VALUES
BIO_f_md() returns the digest BIO method.
diff --git a/doc/crypto/BIO_f_ssl.pod b/doc/crypto/BIO_f_ssl.pod
index f0b731731f50..bc5861ab34b3 100644
--- a/doc/crypto/BIO_f_ssl.pod
+++ b/doc/crypto/BIO_f_ssl.pod
@@ -308,6 +308,15 @@ a client and also echoes the request to standard output.
BIO_free_all(sbio);
+=head1 BUGS
+
+In OpenSSL versions before 1.0.0 the BIO_pop() call was handled incorrectly,
+the I/O BIO reference count was incorrectly incremented (instead of
+decremented) and dissociated with the SSL BIO even if the SSL BIO was not
+explicitly being popped (e.g. a pop higher up the chain). Applications which
+included workarounds for this bug (e.g. freeing BIOs more than once) should
+be modified to handle this fix or they may free up an already freed BIO.
+
=head1 SEE ALSO
TBA
diff --git a/doc/crypto/BIO_new_CMS.pod b/doc/crypto/BIO_new_CMS.pod
new file mode 100644
index 000000000000..9e3a4b7f89e1
--- /dev/null
+++ b/doc/crypto/BIO_new_CMS.pod
@@ -0,0 +1,66 @@
+=pod
+
+=head1 NAME
+
+ BIO_new_CMS - CMS streaming filter BIO
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ BIO *BIO_new_CMS(BIO *out, CMS_ContentInfo *cms);
+
+=head1 DESCRIPTION
+
+BIO_new_CMS() returns a streaming filter BIO chain based on B<cms>. The output
+of the filter is written to B<out>. Any data written to the chain is
+automatically translated to a BER format CMS structure of the appropriate type.
+
+=head1 NOTES
+
+The chain returned by this function behaves like a standard filter BIO. It
+supports non blocking I/O. Content is processed and streamed on the fly and not
+all held in memory at once: so it is possible to encode very large structures.
+After all content has been written through the chain BIO_flush() must be called
+to finalise the structure.
+
+The B<CMS_STREAM> flag must be included in the corresponding B<flags>
+parameter of the B<cms> creation function.
+
+If an application wishes to write additional data to B<out> BIOs should be
+removed from the chain using BIO_pop() and freed with BIO_free() until B<out>
+is reached. If no additional data needs to be written BIO_free_all() can be
+called to free up the whole chain.
+
+Any content written through the filter is used verbatim: no canonical
+translation is performed.
+
+It is possible to chain multiple BIOs to, for example, create a triple wrapped
+signed, enveloped, signed structure. In this case it is the applications
+responsibility to set the inner content type of any outer CMS_ContentInfo
+structures.
+
+Large numbers of small writes through the chain should be avoided as this will
+produce an output consisting of lots of OCTET STRING structures. Prepending
+a BIO_f_buffer() buffering BIO will prevent this.
+
+=head1 BUGS
+
+There is currently no corresponding inverse BIO: i.e. one which can decode
+a CMS structure on the fly.
+
+=head1 RETURN VALUES
+
+BIO_new_CMS() returns a BIO chain when successful or NULL if an error
+occurred. The error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_encrypt(3)|CMS_encrypt(3)>
+
+=head1 HISTORY
+
+BIO_new_CMS() was added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/BIO_s_file.pod b/doc/crypto/BIO_s_file.pod
index b2a29263f4cd..188aea347dae 100644
--- a/doc/crypto/BIO_s_file.pod
+++ b/doc/crypto/BIO_s_file.pod
@@ -76,6 +76,10 @@ normally be closed so the BIO_NOCLOSE flag should be set.
Because the file BIO calls the underlying stdio functions any quirks
in stdio behaviour will be mirrored by the corresponding BIO.
+On Windows BIO_new_files reserves for the filename argument to be
+UTF-8 encoded. In other words if you have to make it work in multi-
+lingual environment, encode file names in UTF-8.
+
=head1 EXAMPLES
File BIO "hello world":
diff --git a/doc/crypto/BIO_s_mem.pod b/doc/crypto/BIO_s_mem.pod
index 19648acfae0f..8f85e0dceeb7 100644
--- a/doc/crypto/BIO_s_mem.pod
+++ b/doc/crypto/BIO_s_mem.pod
@@ -74,7 +74,7 @@ Writes to memory BIOs will always succeed if memory is available: that is
their size can grow indefinitely.
Every read from a read write memory BIO will remove the data just read with
-an internal copy operation, if a BIO contains a lots of data and it is
+an internal copy operation, if a BIO contains a lot of data and it is
read in small chunks the operation can be very slow. The use of a read only
memory BIO avoids this problem. If the BIO must be read write then adding
a buffering BIO to the chain will speed up the process.
diff --git a/doc/crypto/BN_BLINDING_new.pod b/doc/crypto/BN_BLINDING_new.pod
index 7b087f7288f9..5f51fdb47065 100644
--- a/doc/crypto/BN_BLINDING_new.pod
+++ b/doc/crypto/BN_BLINDING_new.pod
@@ -22,8 +22,11 @@ functions.
BN_CTX *ctx);
int BN_BLINDING_invert_ex(BIGNUM *n, const BIGNUM *r, BN_BLINDING *b,
BN_CTX *ctx);
+ #ifndef OPENSSL_NO_DEPRECATED
unsigned long BN_BLINDING_get_thread_id(const BN_BLINDING *);
void BN_BLINDING_set_thread_id(BN_BLINDING *, unsigned long);
+ #endif
+ CRYPTO_THREADID *BN_BLINDING_thread_id(BN_BLINDING *);
unsigned long BN_BLINDING_get_flags(const BN_BLINDING *);
void BN_BLINDING_set_flags(BN_BLINDING *, unsigned long);
BN_BLINDING *BN_BLINDING_create_param(BN_BLINDING *b,
@@ -54,11 +57,11 @@ BN_BLINDING_convert() and BN_BLINDING_invert() are wrapper
functions for BN_BLINDING_convert_ex() and BN_BLINDING_invert_ex()
with B<r> set to NULL.
-BN_BLINDING_set_thread_id() and BN_BLINDING_get_thread_id()
-set and get the "thread id" value of the B<BN_BLINDING> structure,
-a field provided to users of B<BN_BLINDING> structure to help them
-provide proper locking if needed for multi-threaded use. The
-"thread id" of a newly allocated B<BN_BLINDING> structure is zero.
+BN_BLINDING_thread_id() provides access to the B<CRYPTO_THREADID>
+object within the B<BN_BLINDING> structure. This is to help users
+provide proper locking if needed for multi-threaded use. The "thread
+id" object of a newly allocated B<BN_BLINDING> structure is
+initialised to the thread id in which BN_BLINDING_new() was called.
BN_BLINDING_get_flags() returns the BN_BLINDING flags. Currently
there are two supported flags: B<BN_BLINDING_NO_UPDATE> and
@@ -83,8 +86,8 @@ BN_BLINDING_update(), BN_BLINDING_convert(), BN_BLINDING_invert(),
BN_BLINDING_convert_ex() and BN_BLINDING_invert_ex() return 1 on
success and 0 if an error occured.
-BN_BLINDING_get_thread_id() returns the thread id (a B<unsigned long>
-value) or 0 if not set.
+BN_BLINDING_thread_id() returns a pointer to the thread id object
+within a B<BN_BLINDING> object.
BN_BLINDING_get_flags() returns the currently set B<BN_BLINDING> flags
(a B<unsigned long> value).
@@ -98,6 +101,9 @@ L<bn(3)|bn(3)>
=head1 HISTORY
+BN_BLINDING_thread_id was first introduced in OpenSSL 1.0.0, and it
+deprecates BN_BLINDING_set_thread_id and BN_BLINDING_get_thread_id.
+
BN_BLINDING_convert_ex, BN_BLINDIND_invert_ex, BN_BLINDING_get_thread_id,
BN_BLINDING_set_thread_id, BN_BLINDING_set_flags, BN_BLINDING_get_flags
and BN_BLINDING_create_param were first introduced in OpenSSL 0.9.8
diff --git a/doc/crypto/CMS_add0_cert.pod b/doc/crypto/CMS_add0_cert.pod
new file mode 100644
index 000000000000..9c13f488f61a
--- /dev/null
+++ b/doc/crypto/CMS_add0_cert.pod
@@ -0,0 +1,66 @@
+=pod
+
+=head1 NAME
+
+ CMS_add0_cert, CMS_add1_cert, CMS_get1_certs, CMS_add0_crl, CMS_get1_crls, - CMS certificate and CRL utility functions
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int CMS_add0_cert(CMS_ContentInfo *cms, X509 *cert);
+ int CMS_add1_cert(CMS_ContentInfo *cms, X509 *cert);
+ STACK_OF(X509) *CMS_get1_certs(CMS_ContentInfo *cms);
+
+ int CMS_add0_crl(CMS_ContentInfo *cms, X509_CRL *crl);
+ int CMS_add1_crl(CMS_ContentInfo *cms, X509_CRL *crl);
+ STACK_OF(X509_CRL) *CMS_get1_crls(CMS_ContentInfo *cms);
+
+
+=head1 DESCRIPTION
+
+CMS_add0_cert() and CMS_add1_cert() add certificate B<cert> to B<cms>.
+must be of type signed data or enveloped data.
+
+CMS_get1_certs() returns all certificates in B<cms>.
+
+CMS_add0_crl() and CMS_add1_crl() add CRL B<crl> to B<cms>. CMS_get1_crls()
+returns any CRLs in B<cms>.
+
+=head1 NOTES
+
+The CMS_ContentInfo structure B<cms> must be of type signed data or enveloped
+data or an error will be returned.
+
+For signed data certificates and CRLs are added to the B<certificates> and
+B<crls> fields of SignedData structure. For enveloped data they are added to
+B<OriginatorInfo>.
+
+As the B<0> implies CMS_add0_cert() adds B<cert> internally to B<cms> and it
+must not be freed up after the call as opposed to CMS_add1_cert() where B<cert>
+must be freed up.
+
+The same certificate or CRL must not be added to the same cms structure more
+than once.
+
+=head1 RETURN VALUES
+
+CMS_add0_cert(), CMS_add1_cert() and CMS_add0_crl() and CMS_add1_crl() return
+1 for success and 0 for failure.
+
+CMS_get1_certs() and CMS_get1_crls() return the STACK of certificates or CRLs
+or NULL if there are none or an error occurs. The only error which will occur
+in practice is if the B<cms> type is invalid.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>,
+L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_encrypt(3)|CMS_encrypt(3)>
+
+=head1 HISTORY
+
+CMS_add0_cert(), CMS_add1_cert(), CMS_get1_certs(), CMS_add0_crl()
+and CMS_get1_crls() were all first added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_add1_recipient_cert.pod b/doc/crypto/CMS_add1_recipient_cert.pod
new file mode 100644
index 000000000000..d7d8e2532c87
--- /dev/null
+++ b/doc/crypto/CMS_add1_recipient_cert.pod
@@ -0,0 +1,62 @@
+=pod
+
+=head1 NAME
+
+ CMS_add1_recipient_cert, CMS_add0_recipient_key - add recipients to a CMS enveloped data structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_RecipientInfo *CMS_add1_recipient_cert(CMS_ContentInfo *cms, X509 *recip, unsigned int flags);
+
+ CMS_RecipientInfo *CMS_add0_recipient_key(CMS_ContentInfo *cms, int nid, unsigned char *key, size_t keylen, unsigned char *id, size_t idlen, ASN1_GENERALIZEDTIME *date, ASN1_OBJECT *otherTypeId, ASN1_TYPE *otherType);
+
+=head1 DESCRIPTION
+
+CMS_add1_recipient_cert() adds recipient B<recip> to CMS_ContentInfo enveloped
+data structure B<cms> as a KeyTransRecipientInfo structure.
+
+CMS_add0_recipient_key() adds symmetric key B<key> of length B<keylen> using
+wrapping algorithm B<nid>, identifier B<id> of length B<idlen> and optional
+values B<date>, B<otherTypeId> and B<otherType> to CMS_ContentInfo enveloped
+data structure B<cms> as a KEKRecipientInfo structure.
+
+The CMS_ContentInfo structure should be obtained from an initial call to
+CMS_encrypt() with the flag B<CMS_PARTIAL> set.
+
+=head1 NOTES
+
+The main purpose of this function is to provide finer control over a CMS
+enveloped data structure where the simpler CMS_encrypt() function defaults are
+not appropriate. For example if one or more KEKRecipientInfo structures
+need to be added. New attributes can also be added using the returned
+CMS_RecipientInfo structure and the CMS attribute utility functions.
+
+OpenSSL will by default identify recipient certificates using issuer name
+and serial number. If B<CMS_USE_KEYID> is set it will use the subject key
+identifier value instead. An error occurs if all recipient certificates do not
+have a subject key identifier extension.
+
+Currently only AES based key wrapping algorithms are supported for B<nid>,
+specifically: NID_id_aes128_wrap, NID_id_aes192_wrap and NID_id_aes256_wrap.
+If B<nid> is set to B<NID_undef> then an AES wrap algorithm will be used
+consistent with B<keylen>.
+
+=head1 RETURN VALUES
+
+CMS_add1_recipient_cert() and CMS_add0_recipient_key() return an internal
+pointer to the CMS_RecipientInfo structure just added or NULL if an error
+occurs.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_decrypt(3)|CMS_decrypt(3)>,
+L<CMS_final(3)|CMS_final(3)>,
+
+=head1 HISTORY
+
+CMS_add1_recipient_cert() and CMS_add0_recipient_key() were added to OpenSSL
+0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_compress.pod b/doc/crypto/CMS_compress.pod
new file mode 100644
index 000000000000..0a0715271d31
--- /dev/null
+++ b/doc/crypto/CMS_compress.pod
@@ -0,0 +1,73 @@
+=pod
+
+=head1 NAME
+
+CMS_compress - create a CMS CompressedData structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_ContentInfo *CMS_compress(BIO *in, int comp_nid, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_compress() creates and returns a CMS CompressedData structure. B<comp_nid>
+is the compression algorithm to use or B<NID_undef> to use the default
+algorithm (zlib compression). B<in> is the content to be compressed.
+B<flags> is an optional set of flags.
+
+=head1 NOTES
+
+The only currently supported compression algorithm is zlib using the NID
+NID_zlib_compression.
+
+If zlib support is not compiled into OpenSSL then CMS_compress() will return
+an error.
+
+If the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are
+prepended to the data.
+
+Normally the supplied content is translated into MIME canonical format (as
+required by the S/MIME specifications) if B<CMS_BINARY> is set no translation
+occurs. This option should be used if the supplied data is in binary format
+otherwise the translation will corrupt it. If B<CMS_BINARY> is set then
+B<CMS_TEXT> is ignored.
+
+If the B<CMS_STREAM> flag is set a partial B<CMS_ContentInfo> structure is
+returned suitable for streaming I/O: no data is read from the BIO B<in>.
+
+The compressed data is included in the CMS_ContentInfo structure, unless
+B<CMS_DETACHED> is set in which case it is omitted. This is rarely used in
+practice and is not supported by SMIME_write_CMS().
+
+=head1 NOTES
+
+If the flag B<CMS_STREAM> is set the returned B<CMS_ContentInfo> structure is
+B<not> complete and outputting its contents via a function that does not
+properly finalize the B<CMS_ContentInfo> structure will give unpredictable
+results.
+
+Several functions including SMIME_write_CMS(), i2d_CMS_bio_stream(),
+PEM_write_bio_CMS_stream() finalize the structure. Alternatively finalization
+can be performed by obtaining the streaming ASN1 B<BIO> directly using
+BIO_new_CMS().
+
+Additional compression parameters such as the zlib compression level cannot
+currently be set.
+
+=head1 RETURN VALUES
+
+CMS_compress() returns either a CMS_ContentInfo structure or NULL if an error
+occurred. The error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_uncompress(3)|CMS_uncompress(3)>
+
+=head1 HISTORY
+
+CMS_compress() was added to OpenSSL 0.9.8
+The B<CMS_STREAM> flag was first supported in OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/CMS_decrypt.pod b/doc/crypto/CMS_decrypt.pod
new file mode 100644
index 000000000000..d857e4f93f6e
--- /dev/null
+++ b/doc/crypto/CMS_decrypt.pod
@@ -0,0 +1,65 @@
+=pod
+
+=head1 NAME
+
+ CMS_decrypt - decrypt content from a CMS envelopedData structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int CMS_decrypt(CMS_ContentInfo *cms, EVP_PKEY *pkey, X509 *cert, BIO *dcont, BIO *out, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_decrypt() extracts and decrypts the content from a CMS EnvelopedData
+structure. B<pkey> is the private key of the recipient, B<cert> is the
+recipient's certificate, B<out> is a BIO to write the content to and
+B<flags> is an optional set of flags.
+
+The B<dcont> parameter is used in the rare case where the encrypted content
+is detached. It will normally be set to NULL.
+
+=head1 NOTES
+
+OpenSSL_add_all_algorithms() (or equivalent) should be called before using this
+function or errors about unknown algorithms will occur.
+
+Although the recipients certificate is not needed to decrypt the data it is
+needed to locate the appropriate (of possible several) recipients in the CMS
+structure. If B<cert> is set to NULL all possible recipients are tried.
+
+It is possible to determine the correct recipient key by other means (for
+example looking them up in a database) and setting them in the CMS structure
+in advance using the CMS utility functions such as CMS_set1_pkey(). In this
+case both B<cert> and B<pkey> should be set to NULL.
+
+To process KEKRecipientInfo types CMS_set1_key() or CMS_RecipientInfo_set0_key()
+and CMS_ReceipientInfo_decrypt() should be called before CMS_decrypt() and
+B<cert> and B<pkey> set to NULL.
+
+The following flags can be passed in the B<flags> parameter.
+
+If the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are deleted
+from the content. If the content is not of type B<text/plain> then an error is
+returned.
+
+=head1 RETURN VALUES
+
+CMS_decrypt() returns either 1 for success or 0 for failure.
+The error can be obtained from ERR_get_error(3)
+
+=head1 BUGS
+
+The lack of single pass processing and the need to hold all data in memory as
+mentioned in CMS_verify() also applies to CMS_decrypt().
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_encrypt(3)|CMS_encrypt(3)>
+
+=head1 HISTORY
+
+CMS_decrypt() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_encrypt.pod b/doc/crypto/CMS_encrypt.pod
new file mode 100644
index 000000000000..1ee5b275ec82
--- /dev/null
+++ b/doc/crypto/CMS_encrypt.pod
@@ -0,0 +1,96 @@
+=pod
+
+=head1 NAME
+
+ CMS_encrypt - create a CMS envelopedData structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_ContentInfo *CMS_encrypt(STACK_OF(X509) *certs, BIO *in, const EVP_CIPHER *cipher, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_encrypt() creates and returns a CMS EnvelopedData structure. B<certs>
+is a list of recipient certificates. B<in> is the content to be encrypted.
+B<cipher> is the symmetric cipher to use. B<flags> is an optional set of flags.
+
+=head1 NOTES
+
+Only certificates carrying RSA keys are supported so the recipient certificates
+supplied to this function must all contain RSA public keys, though they do not
+have to be signed using the RSA algorithm.
+
+EVP_des_ede3_cbc() (triple DES) is the algorithm of choice for S/MIME use
+because most clients will support it.
+
+The algorithm passed in the B<cipher> parameter must support ASN1 encoding of
+its parameters.
+
+Many browsers implement a "sign and encrypt" option which is simply an S/MIME
+envelopedData containing an S/MIME signed message. This can be readily produced
+by storing the S/MIME signed message in a memory BIO and passing it to
+CMS_encrypt().
+
+The following flags can be passed in the B<flags> parameter.
+
+If the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are
+prepended to the data.
+
+Normally the supplied content is translated into MIME canonical format (as
+required by the S/MIME specifications) if B<CMS_BINARY> is set no translation
+occurs. This option should be used if the supplied data is in binary format
+otherwise the translation will corrupt it. If B<CMS_BINARY> is set then
+B<CMS_TEXT> is ignored.
+
+OpenSSL will by default identify recipient certificates using issuer name
+and serial number. If B<CMS_USE_KEYID> is set it will use the subject key
+identifier value instead. An error occurs if all recipient certificates do not
+have a subject key identifier extension.
+
+If the B<CMS_STREAM> flag is set a partial B<CMS_ContentInfo> structure is
+returned suitable for streaming I/O: no data is read from the BIO B<in>.
+
+If the B<CMS_PARTIAL> flag is set a partial B<CMS_ContentInfo> structure is
+returned to which additional recipients and attributes can be added before
+finalization.
+
+The data being encrypted is included in the CMS_ContentInfo structure, unless
+B<CMS_DETACHED> is set in which case it is omitted. This is rarely used in
+practice and is not supported by SMIME_write_CMS().
+
+=head1 NOTES
+
+If the flag B<CMS_STREAM> is set the returned B<CMS_ContentInfo> structure is
+B<not> complete and outputting its contents via a function that does not
+properly finalize the B<CMS_ContentInfo> structure will give unpredictable
+results.
+
+Several functions including SMIME_write_CMS(), i2d_CMS_bio_stream(),
+PEM_write_bio_CMS_stream() finalize the structure. Alternatively finalization
+can be performed by obtaining the streaming ASN1 B<BIO> directly using
+BIO_new_CMS().
+
+The recipients specified in B<certs> use a CMS KeyTransRecipientInfo info
+structure. KEKRecipientInfo is also supported using the flag B<CMS_PARTIAL>
+and CMS_add0_recipient_key().
+
+The parameter B<certs> may be NULL if B<CMS_PARTIAL> is set and recipients
+added later using CMS_add1_recipient_cert() or CMS_add0_recipient_key().
+
+=head1 RETURN VALUES
+
+CMS_encrypt() returns either a CMS_ContentInfo structure or NULL if an error
+occurred. The error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_decrypt(3)|CMS_decrypt(3)>
+
+=head1 HISTORY
+
+CMS_decrypt() was added to OpenSSL 0.9.8
+The B<CMS_STREAM> flag was first supported in OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/CMS_final.pod b/doc/crypto/CMS_final.pod
new file mode 100644
index 000000000000..36cf96b8a0b7
--- /dev/null
+++ b/doc/crypto/CMS_final.pod
@@ -0,0 +1,41 @@
+=pod
+
+=head1 NAME
+
+ CMS_final - finalise a CMS_ContentInfo structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int CMS_final(CMS_ContentInfo *cms, BIO *data, BIO *dcont, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_final() finalises the structure B<cms>. It's purpose is to perform any
+operations necessary on B<cms> (digest computation for example) and set the
+appropriate fields. The parameter B<data> contains the content to be
+processed. The B<dcont> parameter contains a BIO to write content to after
+processing: this is only used with detached data and will usually be set to
+NULL.
+
+=head1 NOTES
+
+This function will normally be called when the B<CMS_PARTIAL> flag is used. It
+should only be used when streaming is not performed because the streaming
+I/O functions perform finalisation operations internally.
+
+=head1 RETURN VALUES
+
+CMS_final() returns 1 for success or 0 for failure.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_encrypt(3)|CMS_encrypt(3)>
+
+=head1 HISTORY
+
+CMS_final() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_get0_RecipientInfos.pod b/doc/crypto/CMS_get0_RecipientInfos.pod
new file mode 100644
index 000000000000..e0355423e6d6
--- /dev/null
+++ b/doc/crypto/CMS_get0_RecipientInfos.pod
@@ -0,0 +1,106 @@
+=pod
+
+=head1 NAME
+
+ CMS_get0_RecipientInfos, CMS_RecipientInfo_type, CMS_RecipientInfo_ktri_get0_signer_id,CMS_RecipientInfo_ktri_cert_cmp, CMS_RecipientInfo_set0_pkey, CMS_RecipientInfo_kekri_get0_id, CMS_RecipientInfo_kekri_id_cmp, CMS_RecipientInfo_set0_key, CMS_RecipientInfo_decrypt - CMS envelopedData RecipientInfo routines
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ STACK_OF(CMS_RecipientInfo) *CMS_get0_RecipientInfos(CMS_ContentInfo *cms);
+ int CMS_RecipientInfo_type(CMS_RecipientInfo *ri);
+
+ int CMS_RecipientInfo_ktri_get0_signer_id(CMS_RecipientInfo *ri, ASN1_OCTET_STRING **keyid, X509_NAME **issuer, ASN1_INTEGER **sno);
+ int CMS_RecipientInfo_ktri_cert_cmp(CMS_RecipientInfo *ri, X509 *cert);
+ int CMS_RecipientInfo_set0_pkey(CMS_RecipientInfo *ri, EVP_PKEY *pkey);
+
+ int CMS_RecipientInfo_kekri_get0_id(CMS_RecipientInfo *ri, X509_ALGOR **palg, ASN1_OCTET_STRING **pid, ASN1_GENERALIZEDTIME **pdate, ASN1_OBJECT **potherid, ASN1_TYPE **pothertype);
+ int CMS_RecipientInfo_kekri_id_cmp(CMS_RecipientInfo *ri, const unsigned char *id, size_t idlen);
+ int CMS_RecipientInfo_set0_key(CMS_RecipientInfo *ri, unsigned char *key, size_t keylen);
+
+ int CMS_RecipientInfo_decrypt(CMS_ContentInfo *cms, CMS_RecipientInfo *ri);
+
+=head1 DESCRIPTION
+
+The function CMS_get0_RecipientInfos() returns all the CMS_RecipientInfo
+structures associated with a CMS EnvelopedData structure.
+
+CMS_RecipientInfo_type() returns the type of CMS_RecipientInfo structure B<ri>.
+It will currently return CMS_RECIPINFO_TRANS, CMS_RECIPINFO_AGREE,
+CMS_RECIPINFO_KEK, CMS_RECIPINFO_PASS, or CMS_RECIPINFO_OTHER.
+
+CMS_RecipientInfo_ktri_get0_signer_id() retrieves the certificate recipient
+identifier associated with a specific CMS_RecipientInfo structure B<ri>, which
+must be of type CMS_RECIPINFO_TRANS. Either the keyidentifier will be set in
+B<keyid> or B<both> issuer name and serial number in B<issuer> and B<sno>.
+
+CMS_RecipientInfo_ktri_cert_cmp() compares the certificate B<cert> against the
+CMS_RecipientInfo structure B<ri>, which must be of type CMS_RECIPINFO_TRANS.
+It returns zero if the comparison is successful and non zero if not.
+
+CMS_RecipientInfo_set0_pkey() associates the private key B<pkey> with
+the CMS_RecipientInfo structure B<ri>, which must be of type
+CMS_RECIPINFO_TRANS.
+
+CMS_RecipientInfo_kekri_get0_id() retrieves the key information from the
+CMS_RecipientInfo structure B<ri> which must be of type CMS_RECIPINFO_KEK. Any
+of the remaining parameters can be NULL if the application is not interested in
+the value of a field. Where a field is optional and absent NULL will be written
+to the corresponding parameter. The keyEncryptionAlgorithm field is written to
+B<palg>, the B<keyIdentifier> field is written to B<pid>, the B<date> field if
+present is written to B<pdate>, if the B<other> field is present the components
+B<keyAttrId> and B<keyAttr> are written to parameters B<potherid> and
+B<pothertype>.
+
+CMS_RecipientInfo_kekri_id_cmp() compares the ID in the B<id> and B<idlen>
+parameters against the B<keyIdentifier> CMS_RecipientInfo structure B<ri>,
+which must be of type CMS_RECIPINFO_KEK. It returns zero if the comparison is
+successful and non zero if not.
+
+CMS_RecipientInfo_set0_key() associates the symmetric key B<key> of length
+B<keylen> with the CMS_RecipientInfo structure B<ri>, which must be of type
+CMS_RECIPINFO_KEK.
+
+CMS_RecipientInfo_decrypt() attempts to decrypt CMS_RecipientInfo structure
+B<ri> in structure B<cms>. A key must have been associated with the structure
+first.
+
+=head1 NOTES
+
+The main purpose of these functions is to enable an application to lookup
+recipient keys using any appropriate technique when the simpler method
+of CMS_decrypt() is not appropriate.
+
+In typical usage and application will retrieve all CMS_RecipientInfo structures
+using CMS_get0_RecipientInfos() and check the type of each using
+CMS_RecpientInfo_type(). Depending on the type the CMS_RecipientInfo structure
+can be ignored or its key identifier data retrieved using an appropriate
+function. Then if the corresponding secret or private key can be obtained by
+any appropriate means it can then associated with the structure and
+CMS_RecpientInfo_decrypt() called. If successful CMS_decrypt() can be called
+with a NULL key to decrypt the enveloped content.
+
+=head1 RETURN VALUES
+
+CMS_get0_RecipientInfos() returns all CMS_RecipientInfo structures, or NULL if
+an error occurs.
+
+CMS_RecipientInfo_ktri_get0_signer_id(), CMS_RecipientInfo_set0_pkey(),
+CMS_RecipientInfo_kekri_get0_id(), CMS_RecipientInfo_set0_key() and
+CMS_RecipientInfo_decrypt() return 1 for success or 0 if an error occurs.
+
+CMS_RecipientInfo_ktri_cert_cmp() and CMS_RecipientInfo_kekri_cmp() return 0
+for a successful comparison and non zero otherwise.
+
+Any error can be obtained from L<ERR_get_error(3)|ERR_get_error(3)>.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_decrypt(3)|CMS_decrypt(3)>
+
+=head1 HISTORY
+
+These functions were first was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_get0_SignerInfos.pod b/doc/crypto/CMS_get0_SignerInfos.pod
new file mode 100644
index 000000000000..47f6d2a04726
--- /dev/null
+++ b/doc/crypto/CMS_get0_SignerInfos.pod
@@ -0,0 +1,75 @@
+=pod
+
+=head1 NAME
+
+ CMS_get0_SignerInfos, CMS_SignerInfo_get0_signer_id, CMS_SignerInfo_cert_cmp, CMS_set1_signer_certs - CMS signedData signer functions.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ STACK_OF(CMS_SignerInfo) *CMS_get0_SignerInfos(CMS_ContentInfo *cms);
+
+ int CMS_SignerInfo_get0_signer_id(CMS_SignerInfo *si, ASN1_OCTET_STRING **keyid, X509_NAME **issuer, ASN1_INTEGER **sno);
+ int CMS_SignerInfo_cert_cmp(CMS_SignerInfo *si, X509 *cert);
+ void CMS_SignerInfo_set1_signer_cert(CMS_SignerInfo *si, X509 *signer);
+
+=head1 DESCRIPTION
+
+The function CMS_get0_SignerInfos() returns all the CMS_SignerInfo structures
+associated with a CMS signedData structure.
+
+CMS_SignerInfo_get0_signer_id() retrieves the certificate signer identifier
+associated with a specific CMS_SignerInfo structure B<si>. Either the
+keyidentifier will be set in B<keyid> or B<both> issuer name and serial number
+in B<issuer> and B<sno>.
+
+CMS_SignerInfo_cert_cmp() compares the certificate B<cert> against the signer
+identifier B<si>. It returns zero if the comparison is successful and non zero
+if not.
+
+CMS_SignerInfo_set1_signer_cert() sets the signers certificate of B<si> to
+B<signer>.
+
+=head1 NOTES
+
+The main purpose of these functions is to enable an application to lookup
+signers certificates using any appropriate technique when the simpler method
+of CMS_verify() is not appropriate.
+
+In typical usage and application will retrieve all CMS_SignerInfo structures
+using CMS_get0_SignerInfo() and retrieve the identifier information using
+CMS. It will then obtain the signer certificate by some unspecified means
+(or return and error if it cannot be found) and set it using
+CMS_SignerInfo_set1_signer_cert().
+
+Once all signer certificates have been set CMS_verify() can be used.
+
+Although CMS_get0_SignerInfos() can return NULL is an error occur B<or> if
+there are no signers this is not a problem in practice because the only
+error which can occur is if the B<cms> structure is not of type signedData
+due to application error.
+
+=head1 RETURN VALUES
+
+CMS_get0_SignerInfos() returns all CMS_SignerInfo structures, or NULL there
+are no signers or an error occurs.
+
+CMS_SignerInfo_get0_signer_id() returns 1 for success and 0 for failure.
+
+CMS_SignerInfo_cert_cmp() returns 0 for a successful comparison and non
+zero otherwise.
+
+CMS_SignerInfo_set1_signer_cert() does not return a value.
+
+Any error can be obtained from L<ERR_get_error(3)|ERR_get_error(3)>
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_verify(3)|CMS_verify(3)>
+
+=head1 HISTORY
+
+These functions were first was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_get0_type.pod b/doc/crypto/CMS_get0_type.pod
new file mode 100644
index 000000000000..8ff1c3115cd4
--- /dev/null
+++ b/doc/crypto/CMS_get0_type.pod
@@ -0,0 +1,63 @@
+=pod
+
+=head1 NAME
+
+ CMS_get0_type, CMS_set1_eContentType, CMS_get0_eContentType - get and set CMS content types
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ const ASN1_OBJECT *CMS_get0_type(CMS_ContentInfo *cms);
+ int CMS_set1_eContentType(CMS_ContentInfo *cms, const ASN1_OBJECT *oid);
+ const ASN1_OBJECT *CMS_get0_eContentType(CMS_ContentInfo *cms);
+
+=head1 DESCRIPTION
+
+CMS_get0_type() returns the content type of a CMS_ContentInfo structure as
+and ASN1_OBJECT pointer. An application can then decide how to process the
+CMS_ContentInfo structure based on this value.
+
+CMS_set1_eContentType() sets the embedded content type of a CMS_ContentInfo
+structure. It should be called with CMS functions with the B<CMS_PARTIAL>
+flag and B<before> the structure is finalised, otherwise the results are
+undefined.
+
+ASN1_OBJECT *CMS_get0_eContentType() returns a pointer to the embedded
+content type.
+
+=head1 NOTES
+
+As the B<0> implies CMS_get0_type() and CMS_get0_eContentType() return internal
+pointers which should B<not> be freed up. CMS_set1_eContentType() copies the
+supplied OID and it B<should> be freed up after use.
+
+The B<ASN1_OBJECT> values returned can be converted to an integer B<NID> value
+using OBJ_obj2nid(). For the currently supported content types the following
+values are returned:
+
+ NID_pkcs7_data
+ NID_pkcs7_signed
+ NID_pkcs7_digest
+ NID_id_smime_ct_compressedData:
+ NID_pkcs7_encrypted
+ NID_pkcs7_enveloped
+
+
+=head1 RETURN VALUES
+
+CMS_get0_type() and CMS_get0_eContentType() return and ASN1_OBJECT structure.
+
+CMS_set1_eContentType() returns 1 for success or 0 if an error occurred. The
+error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>
+
+=head1 HISTORY
+
+CMS_get0_type(), CMS_set1_eContentType() and CMS_get0_eContentType() were all
+first added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_get1_ReceiptRequest.pod b/doc/crypto/CMS_get1_ReceiptRequest.pod
new file mode 100644
index 000000000000..f546376a1e68
--- /dev/null
+++ b/doc/crypto/CMS_get1_ReceiptRequest.pod
@@ -0,0 +1,69 @@
+=pod
+
+=head1 NAME
+
+ CMS_ReceiptRequest_create0, CMS_add1_ReceiptRequest, CMS_get1_ReceiptRequest, CMS_ReceiptRequest_get0_values - CMS signed receipt request functions.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_ReceiptRequest *CMS_ReceiptRequest_create0(unsigned char *id, int idlen, int allorfirst, STACK_OF(GENERAL_NAMES) *receiptList, STACK_OF(GENERAL_NAMES) *receiptsTo);
+ int CMS_add1_ReceiptRequest(CMS_SignerInfo *si, CMS_ReceiptRequest *rr);
+ int CMS_get1_ReceiptRequest(CMS_SignerInfo *si, CMS_ReceiptRequest **prr);
+ void CMS_ReceiptRequest_get0_values(CMS_ReceiptRequest *rr, ASN1_STRING **pcid, int *pallorfirst, STACK_OF(GENERAL_NAMES) **plist, STACK_OF(GENERAL_NAMES) **prto);
+
+=head1 DESCRIPTION
+
+CMS_ReceiptRequest_create0() creates a signed receipt request structure. The
+B<signedContentIdentifier> field is set using B<id> and B<idlen>, or it is set
+to 32 bytes of pseudo random data if B<id> is NULL. If B<receiptList> is NULL
+the allOrFirstTier option in B<receiptsFrom> is used and set to the value of
+the B<allorfirst> parameter. If B<receiptList> is not NULL the B<receiptList>
+option in B<receiptsFrom> is used. The B<receiptsTo> parameter specifies the
+B<receiptsTo> field value.
+
+The CMS_add1_ReceiptRequest() function adds a signed receipt request B<rr>
+to SignerInfo structure B<si>.
+
+int CMS_get1_ReceiptRequest() looks for a signed receipt request in B<si>, if
+any is found it is decoded and written to B<prr>.
+
+CMS_ReceiptRequest_get0_values() retrieves the values of a receipt request.
+The signedContentIdentifier is copied to B<pcid>. If the B<allOrFirstTier>
+option of B<receiptsFrom> is used its value is copied to B<pallorfirst>
+otherwise the B<receiptList> field is copied to B<plist>. The B<receiptsTo>
+parameter is copied to B<prto>.
+
+=head1 NOTES
+
+For more details of the meaning of the fields see RFC2634.
+
+The contents of a signed receipt should only be considered meaningful if the
+corresponding CMS_ContentInfo structure can be successfully verified using
+CMS_verify().
+
+=head1 RETURN VALUES
+
+CMS_ReceiptRequest_create0() returns a signed receipt request structure or
+NULL if an error occurred.
+
+CMS_add1_ReceiptRequest() returns 1 for success or 0 is an error occurred.
+
+CMS_get1_ReceiptRequest() returns 1 is a signed receipt request is found and
+decoded. It returns 0 if a signed receipt request is not present and -1 if
+it is present but malformed.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_sign_receipt(3)|CMS_sign_receipt(3)>, L<CMS_verify(3)|CMS_verify(3)>
+L<CMS_verify_receipt(3)|CMS_verify_receipt(3)>
+
+=head1 HISTORY
+
+CMS_ReceiptRequest_create0(), CMS_add1_ReceiptRequest(),
+CMS_get1_ReceiptRequest() and CMS_ReceiptRequest_get0_values() were added to
+OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_sign.pod b/doc/crypto/CMS_sign.pod
new file mode 100644
index 000000000000..2cc72de32720
--- /dev/null
+++ b/doc/crypto/CMS_sign.pod
@@ -0,0 +1,121 @@
+=pod
+
+=head1 NAME
+
+ CMS_sign - create a CMS SignedData structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_ContentInfo *CMS_sign(X509 *signcert, EVP_PKEY *pkey, STACK_OF(X509) *certs, BIO *data, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_sign() creates and returns a CMS SignedData structure. B<signcert> is
+the certificate to sign with, B<pkey> is the corresponding private key.
+B<certs> is an optional additional set of certificates to include in the CMS
+structure (for example any intermediate CAs in the chain). Any or all of
+these parameters can be B<NULL>, see B<NOTES> below.
+
+The data to be signed is read from BIO B<data>.
+
+B<flags> is an optional set of flags.
+
+=head1 NOTES
+
+Any of the following flags (ored together) can be passed in the B<flags>
+parameter.
+
+Many S/MIME clients expect the signed content to include valid MIME headers. If
+the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are prepended
+to the data.
+
+If B<CMS_NOCERTS> is set the signer's certificate will not be included in the
+CMS_ContentInfo structure, the signer's certificate must still be supplied in
+the B<signcert> parameter though. This can reduce the size of the signature if
+the signers certificate can be obtained by other means: for example a
+previously signed message.
+
+The data being signed is included in the CMS_ContentInfo structure, unless
+B<CMS_DETACHED> is set in which case it is omitted. This is used for
+CMS_ContentInfo detached signatures which are used in S/MIME plaintext signed
+messages for example.
+
+Normally the supplied content is translated into MIME canonical format (as
+required by the S/MIME specifications) if B<CMS_BINARY> is set no translation
+occurs. This option should be used if the supplied data is in binary format
+otherwise the translation will corrupt it.
+
+The SignedData structure includes several CMS signedAttributes including the
+signing time, the CMS content type and the supported list of ciphers in an
+SMIMECapabilities attribute. If B<CMS_NOATTR> is set then no signedAttributes
+will be used. If B<CMS_NOSMIMECAP> is set then just the SMIMECapabilities are
+omitted.
+
+If present the SMIMECapabilities attribute indicates support for the following
+algorithms in preference order: 256 bit AES, Gost R3411-94, Gost 28147-89, 192
+bit AES, 128 bit AES, triple DES, 128 bit RC2, 64 bit RC2, DES and 40 bit RC2.
+If any of these algorithms is not available then it will not be included: for example the GOST algorithms will not be included if the GOST ENGINE is
+not loaded.
+
+OpenSSL will by default identify signing certificates using issuer name
+and serial number. If B<CMS_USE_KEYID> is set it will use the subject key
+identifier value instead. An error occurs if the signing certificate does not
+have a subject key identifier extension.
+
+If the flags B<CMS_STREAM> is set then the returned B<CMS_ContentInfo>
+structure is just initialized ready to perform the signing operation. The
+signing is however B<not> performed and the data to be signed is not read from
+the B<data> parameter. Signing is deferred until after the data has been
+written. In this way data can be signed in a single pass.
+
+If the B<CMS_PARTIAL> flag is set a partial B<CMS_ContentInfo> structure is
+output to which additional signers and capabilities can be added before
+finalization.
+
+If the flag B<CMS_STREAM> is set the returned B<CMS_ContentInfo> structure is
+B<not> complete and outputting its contents via a function that does not
+properly finalize the B<CMS_ContentInfo> structure will give unpredictable
+results.
+
+Several functions including SMIME_write_CMS(), i2d_CMS_bio_stream(),
+PEM_write_bio_CMS_stream() finalize the structure. Alternatively finalization
+can be performed by obtaining the streaming ASN1 B<BIO> directly using
+BIO_new_CMS().
+
+If a signer is specified it will use the default digest for the signing
+algorithm. This is B<SHA1> for both RSA and DSA keys.
+
+If B<signcert> and B<pkey> are NULL then a certificates only CMS structure is
+output.
+
+The function CMS_sign() is a basic CMS signing function whose output will be
+suitable for many purposes. For finer control of the output format the
+B<certs>, B<signcert> and B<pkey> parameters can all be B<NULL> and the
+B<CMS_PARTIAL> flag set. Then one or more signers can be added using the
+function CMS_sign_add1_signer(), non default digests can be used and custom
+attributes added. B<CMS_final()> must then be called to finalize the
+structure if streaming is not enabled.
+
+=head1 BUGS
+
+Some attributes such as counter signatures are not supported.
+
+=head1 RETURN VALUES
+
+CMS_sign() returns either a valid CMS_ContentInfo structure or NULL if an error
+occurred. The error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_verify(3)|CMS_verify(3)>
+
+=head1 HISTORY
+
+CMS_sign() was added to OpenSSL 0.9.8
+
+The B<CMS_STREAM> flag is only supported for detached data in OpenSSL 0.9.8,
+it is supported for embedded data in OpenSSL 1.0.0 and later.
+
+=cut
diff --git a/doc/crypto/CMS_sign_add1_signer.pod b/doc/crypto/CMS_sign_add1_signer.pod
new file mode 100644
index 000000000000..bda3ca2adbd1
--- /dev/null
+++ b/doc/crypto/CMS_sign_add1_signer.pod
@@ -0,0 +1,101 @@
+=pod
+
+=head1 NAME
+
+ CMS_sign_add1_signer, CMS_SignerInfo_sign - add a signer to a CMS_ContentInfo signed data structure.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_SignerInfo *CMS_sign_add1_signer(CMS_ContentInfo *cms, X509 *signcert, EVP_PKEY *pkey, const EVP_MD *md, unsigned int flags);
+
+ int CMS_SignerInfo_sign(CMS_SignerInfo *si);
+
+
+=head1 DESCRIPTION
+
+CMS_sign_add1_signer() adds a signer with certificate B<signcert> and private
+key B<pkey> using message digest B<md> to CMS_ContentInfo SignedData
+structure B<cms>.
+
+The CMS_ContentInfo structure should be obtained from an initial call to
+CMS_sign() with the flag B<CMS_PARTIAL> set or in the case or re-signing a
+valid CMS_ContentInfo SignedData structure.
+
+If the B<md> parameter is B<NULL> then the default digest for the public
+key algorithm will be used.
+
+Unless the B<CMS_REUSE_DIGEST> flag is set the returned CMS_ContentInfo
+structure is not complete and must be finalized either by streaming (if
+applicable) or a call to CMS_final().
+
+The CMS_SignerInfo_sign() function will explicitly sign a CMS_SignerInfo
+structure, its main use is when B<CMS_REUSE_DIGEST> and B<CMS_PARTIAL> flags
+are both set.
+
+=head1 NOTES
+
+The main purpose of CMS_sign_add1_signer() is to provide finer control
+over a CMS signed data structure where the simpler CMS_sign() function defaults
+are not appropriate. For example if multiple signers or non default digest
+algorithms are needed. New attributes can also be added using the returned
+CMS_SignerInfo structure and the CMS attribute utility functions or the
+CMS signed receipt request functions.
+
+Any of the following flags (ored together) can be passed in the B<flags>
+parameter.
+
+If B<CMS_REUSE_DIGEST> is set then an attempt is made to copy the content
+digest value from the CMS_ContentInfo structure: to add a signer to an existing
+structure. An error occurs if a matching digest value cannot be found to copy.
+The returned CMS_ContentInfo structure will be valid and finalized when this
+flag is set.
+
+If B<CMS_PARTIAL> is set in addition to B<CMS_REUSE_DIGEST> then the
+CMS_SignerInfo structure will not be finalized so additional attributes
+can be added. In this case an explicit call to CMS_SignerInfo_sign() is
+needed to finalize it.
+
+If B<CMS_NOCERTS> is set the signer's certificate will not be included in the
+CMS_ContentInfo structure, the signer's certificate must still be supplied in
+the B<signcert> parameter though. This can reduce the size of the signature if
+the signers certificate can be obtained by other means: for example a
+previously signed message.
+
+The SignedData structure includes several CMS signedAttributes including the
+signing time, the CMS content type and the supported list of ciphers in an
+SMIMECapabilities attribute. If B<CMS_NOATTR> is set then no signedAttributes
+will be used. If B<CMS_NOSMIMECAP> is set then just the SMIMECapabilities are
+omitted.
+
+OpenSSL will by default identify signing certificates using issuer name
+and serial number. If B<CMS_USE_KEYID> is set it will use the subject key
+identifier value instead. An error occurs if the signing certificate does not
+have a subject key identifier extension.
+
+If present the SMIMECapabilities attribute indicates support for the following
+algorithms in preference order: 256 bit AES, Gost R3411-94, Gost 28147-89, 192
+bit AES, 128 bit AES, triple DES, 128 bit RC2, 64 bit RC2, DES and 40 bit RC2.
+If any of these algorithms is not available then it will not be included: for example the GOST algorithms will not be included if the GOST ENGINE is
+not loaded.
+
+CMS_sign_add1_signer() returns an internal pointer to the CMS_SignerInfo
+structure just added, this can be used to set additional attributes
+before it is finalized.
+
+=head1 RETURN VALUES
+
+CMS_sign1_add_signers() returns an internal pointer to the CMS_SignerInfo
+structure just added or NULL if an error occurs.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_final(3)|CMS_final(3)>,
+
+=head1 HISTORY
+
+CMS_sign_add1_signer() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_sign_receipt.pod b/doc/crypto/CMS_sign_receipt.pod
new file mode 100644
index 000000000000..cae1f8338400
--- /dev/null
+++ b/doc/crypto/CMS_sign_receipt.pod
@@ -0,0 +1,45 @@
+=pod
+
+=head1 NAME
+
+ CMS_sign_receipt - create a CMS signed receipt
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_ContentInfo *CMS_sign_receipt(CMS_SignerInfo *si, X509 *signcert, EVP_PKEY *pkey, STACK_OF(X509) *certs, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_sign_receipt() creates and returns a CMS signed receipt structure. B<si> is
+the B<CMS_SignerInfo> structure containing the signed receipt request.
+B<signcert> is the certificate to sign with, B<pkey> is the corresponding
+private key. B<certs> is an optional additional set of certificates to include
+in the CMS structure (for example any intermediate CAs in the chain).
+
+B<flags> is an optional set of flags.
+
+=head1 NOTES
+
+This functions behaves in a similar way to CMS_sign() except the flag values
+B<CMS_DETACHED>, B<CMS_BINARY>, B<CMS_NOATTR>, B<CMS_TEXT> and B<CMS_STREAM>
+are not supported since they do not make sense in the context of signed
+receipts.
+
+=head1 RETURN VALUES
+
+CMS_sign_receipt() returns either a valid CMS_ContentInfo structure or NULL if
+an error occurred. The error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>,
+L<CMS_verify_receipt(3)|CMS_verify_receipt(3)>,
+L<CMS_sign(3)|CMS_sign(3)>
+
+=head1 HISTORY
+
+CMS_sign_receipt() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_uncompress.pod b/doc/crypto/CMS_uncompress.pod
new file mode 100644
index 000000000000..c6056b027da5
--- /dev/null
+++ b/doc/crypto/CMS_uncompress.pod
@@ -0,0 +1,54 @@
+=pod
+
+=head1 NAME
+
+ CMS_uncompress - uncompress a CMS CompressedData structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int CMS_uncompress(CMS_ContentInfo *cms, BIO *dcont, BIO *out, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_uncompress() extracts and uncompresses the content from a CMS
+CompressedData structure B<cms>. B<data> is a BIO to write the content to and
+B<flags> is an optional set of flags.
+
+The B<dcont> parameter is used in the rare case where the compressed content
+is detached. It will normally be set to NULL.
+
+=head1 NOTES
+
+The only currently supported compression algorithm is zlib: if the structure
+indicates the use of any other algorithm an error is returned.
+
+If zlib support is not compiled into OpenSSL then CMS_uncompress() will always
+return an error.
+
+The following flags can be passed in the B<flags> parameter.
+
+If the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are deleted
+from the content. If the content is not of type B<text/plain> then an error is
+returned.
+
+=head1 RETURN VALUES
+
+CMS_uncompress() returns either 1 for success or 0 for failure. The error can
+be obtained from ERR_get_error(3)
+
+=head1 BUGS
+
+The lack of single pass processing and the need to hold all data in memory as
+mentioned in CMS_verify() also applies to CMS_decompress().
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_compress(3)|CMS_compress(3)>
+
+=head1 HISTORY
+
+CMS_uncompress() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_verify.pod b/doc/crypto/CMS_verify.pod
new file mode 100644
index 000000000000..8f26fdab093b
--- /dev/null
+++ b/doc/crypto/CMS_verify.pod
@@ -0,0 +1,126 @@
+=pod
+
+=head1 NAME
+
+ CMS_verify - verify a CMS SignedData structure
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int CMS_verify(CMS_ContentInfo *cms, STACK_OF(X509) *certs, X509_STORE *store, BIO *indata, BIO *out, unsigned int flags);
+
+ STACK_OF(X509) *CMS_get0_signers(CMS_ContentInfo *cms);
+
+=head1 DESCRIPTION
+
+CMS_verify() verifies a CMS SignedData structure. B<cms> is the CMS_ContentInfo
+structure to verify. B<certs> is a set of certificates in which to search for
+the signing certificate(s). B<store> is a trusted certificate store used for
+chain verification. B<indata> is the detached content if the content is not
+present in B<cms>. The content is written to B<out> if it is not NULL.
+
+B<flags> is an optional set of flags, which can be used to modify the verify
+operation.
+
+CMS_get0_signers() retrieves the signing certificate(s) from B<cms>, it must
+be called after a successful CMS_verify() operation.
+
+=head1 VERIFY PROCESS
+
+Normally the verify process proceeds as follows.
+
+Initially some sanity checks are performed on B<cms>. The type of B<cms> must
+be SignedData. There must be at least one signature on the data and if
+the content is detached B<indata> cannot be B<NULL>.
+
+An attempt is made to locate all the signing certificate(s), first looking in
+the B<certs> parameter (if it is not NULL) and then looking in any
+certificates contained in the B<cms> structure itself. If any signing
+certificate cannot be located the operation fails.
+
+Each signing certificate is chain verified using the B<smimesign> purpose and
+the supplied trusted certificate store. Any internal certificates in the message
+are used as untrusted CAs. If CRL checking is enabled in B<store> any internal
+CRLs are used in addition to attempting to look them up in B<store>. If any
+chain verify fails an error code is returned.
+
+Finally the signed content is read (and written to B<out> is it is not NULL)
+and the signature's checked.
+
+If all signature's verify correctly then the function is successful.
+
+Any of the following flags (ored together) can be passed in the B<flags>
+parameter to change the default verify behaviour.
+
+If B<CMS_NOINTERN> is set the certificates in the message itself are not
+searched when locating the signing certificate(s). This means that all the
+signing certificates must be in the B<certs> parameter.
+
+If B<CMS_NOCRL> is set and CRL checking is enabled in B<store> then any
+CRLs in the message itself are ignored.
+
+If the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are deleted
+from the content. If the content is not of type B<text/plain> then an error is
+returned.
+
+If B<CMS_NO_SIGNER_CERT_VERIFY> is set the signing certificates are not
+verified.
+
+If B<CMS_NO_ATTR_VERIFY> is set the signed attributes signature is not
+verified.
+
+If B<CMS_NO_CONTENT_VERIFY> is set then the content digest is not checked.
+
+=head1 NOTES
+
+One application of B<CMS_NOINTERN> is to only accept messages signed by
+a small number of certificates. The acceptable certificates would be passed
+in the B<certs> parameter. In this case if the signer is not one of the
+certificates supplied in B<certs> then the verify will fail because the
+signer cannot be found.
+
+In some cases the standard techniques for looking up and validating
+certificates are not appropriate: for example an application may wish to
+lookup certificates in a database or perform customised verification. This
+can be achieved by setting and verifying the signers certificates manually
+using the signed data utility functions.
+
+Care should be taken when modifying the default verify behaviour, for example
+setting B<CMS_NO_CONTENT_VERIFY> will totally disable all content verification
+and any modified content will be considered valid. This combination is however
+useful if one merely wishes to write the content to B<out> and its validity
+is not considered important.
+
+Chain verification should arguably be performed using the signing time rather
+than the current time. However since the signing time is supplied by the
+signer it cannot be trusted without additional evidence (such as a trusted
+timestamp).
+
+=head1 RETURN VALUES
+
+CMS_verify() returns 1 for a successful verification and zero if an error
+occurred.
+
+CMS_get0_signers() returns all signers or NULL if an error occurred.
+
+The error can be obtained from L<ERR_get_error(3)|ERR_get_error(3)>
+
+=head1 BUGS
+
+The trusted certificate store is not searched for the signing certificate,
+this is primarily due to the inadequacies of the current B<X509_STORE>
+functionality.
+
+The lack of single pass processing means that the signed content must all
+be held in memory if it is not detached.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>
+
+=head1 HISTORY
+
+CMS_verify() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/CMS_verify_receipt.pod b/doc/crypto/CMS_verify_receipt.pod
new file mode 100644
index 000000000000..9283e0e04b89
--- /dev/null
+++ b/doc/crypto/CMS_verify_receipt.pod
@@ -0,0 +1,47 @@
+=pod
+
+=head1 NAME
+
+ CMS_verify_receipt - verify a CMS signed receipt
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int CMS_verify_receipt(CMS_ContentInfo *rcms, CMS_ContentInfo *ocms, STACK_OF(X509) *certs, X509_STORE *store, unsigned int flags);
+
+=head1 DESCRIPTION
+
+CMS_verify_receipt() verifies a CMS signed receipt. B<rcms> is the signed
+receipt to verify. B<ocms> is the original SignedData structure containing the
+receipt request. B<certs> is a set of certificates in which to search for the
+signing certificate. B<store> is a trusted certificate store (used for chain
+verification).
+
+B<flags> is an optional set of flags, which can be used to modify the verify
+operation.
+
+=head1 NOTES
+
+This functions behaves in a similar way to CMS_verify() except the flag values
+B<CMS_DETACHED>, B<CMS_BINARY>, B<CMS_TEXT> and B<CMS_STREAM> are not
+supported since they do not make sense in the context of signed receipts.
+
+=head1 RETURN VALUES
+
+CMS_verify_receipt() returns 1 for a successful verification and zero if an
+error occurred.
+
+The error can be obtained from L<ERR_get_error(3)|ERR_get_error(3)>
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>,
+L<CMS_sign_receipt(3)|CMS_sign_receipt(3)>,
+L<CMS_verify(3)|CMS_verify(3)>,
+
+=head1 HISTORY
+
+CMS_verify_receipt() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/DSA_get_ex_new_index.pod b/doc/crypto/DSA_get_ex_new_index.pod
index 4612e708ecc0..fb6efc118269 100644
--- a/doc/crypto/DSA_get_ex_new_index.pod
+++ b/doc/crypto/DSA_get_ex_new_index.pod
@@ -6,7 +6,7 @@ DSA_get_ex_new_index, DSA_set_ex_data, DSA_get_ex_data - add application specifi
=head1 SYNOPSIS
- #include <openssl/DSA.h>
+ #include <openssl/dsa.h>
int DSA_get_ex_new_index(long argl, void *argp,
CRYPTO_EX_new *new_func,
diff --git a/doc/crypto/EVP_DigestInit.pod b/doc/crypto/EVP_DigestInit.pod
index 98b1368d08d8..367691cc7aed 100644
--- a/doc/crypto/EVP_DigestInit.pod
+++ b/doc/crypto/EVP_DigestInit.pod
@@ -6,7 +6,8 @@ EVP_MD_CTX_init, EVP_MD_CTX_create, EVP_DigestInit_ex, EVP_DigestUpdate,
EVP_DigestFinal_ex, EVP_MD_CTX_cleanup, EVP_MD_CTX_destroy, EVP_MAX_MD_SIZE,
EVP_MD_CTX_copy_ex, EVP_MD_CTX_copy, EVP_MD_type, EVP_MD_pkey_type, EVP_MD_size,
EVP_MD_block_size, EVP_MD_CTX_md, EVP_MD_CTX_size, EVP_MD_CTX_block_size, EVP_MD_CTX_type,
-EVP_md_null, EVP_md2, EVP_md5, EVP_sha, EVP_sha1, EVP_dss, EVP_dss1, EVP_mdc2,
+EVP_md_null, EVP_md2, EVP_md5, EVP_sha, EVP_sha1, EVP_sha224, EVP_sha256,
+EVP_sha384, EVP_sha512, EVP_dss, EVP_dss1, EVP_mdc2,
EVP_ripemd160, EVP_get_digestbyname, EVP_get_digestbynid, EVP_get_digestbyobj -
EVP digest routines
@@ -33,16 +34,15 @@ EVP digest routines
int EVP_MD_CTX_copy(EVP_MD_CTX *out,EVP_MD_CTX *in);
- #define EVP_MAX_MD_SIZE (16+20) /* The SSLv3 md5+sha1 type */
+ #define EVP_MAX_MD_SIZE 64 /* SHA512 */
+ int EVP_MD_type(const EVP_MD *md);
+ int EVP_MD_pkey_type(const EVP_MD *md);
+ int EVP_MD_size(const EVP_MD *md);
+ int EVP_MD_block_size(const EVP_MD *md);
- #define EVP_MD_type(e) ((e)->type)
- #define EVP_MD_pkey_type(e) ((e)->pkey_type)
- #define EVP_MD_size(e) ((e)->md_size)
- #define EVP_MD_block_size(e) ((e)->block_size)
-
- #define EVP_MD_CTX_md(e) (e)->digest)
- #define EVP_MD_CTX_size(e) EVP_MD_size((e)->digest)
+ const EVP_MD *EVP_MD_CTX_md(const EVP_MD_CTX *ctx);
+ #define EVP_MD_CTX_size(e) EVP_MD_size(EVP_MD_CTX_md(e))
#define EVP_MD_CTX_block_size(e) EVP_MD_block_size((e)->digest)
#define EVP_MD_CTX_type(e) EVP_MD_type((e)->digest)
@@ -56,6 +56,11 @@ EVP digest routines
const EVP_MD *EVP_mdc2(void);
const EVP_MD *EVP_ripemd160(void);
+ const EVP_MD *EVP_sha224(void);
+ const EVP_MD *EVP_sha256(void);
+ const EVP_MD *EVP_sha384(void);
+ const EVP_MD *EVP_sha512(void);
+
const EVP_MD *EVP_get_digestbyname(const char *name);
#define EVP_get_digestbynid(a) EVP_get_digestbyname(OBJ_nid2sn(a))
#define EVP_get_digestbyobj(a) EVP_get_digestbynid(OBJ_obj2nid(a))
@@ -124,15 +129,19 @@ B<EVP_MD_CTX>.
EVP_MD_pkey_type() returns the NID of the public key signing algorithm associated
with this digest. For example EVP_sha1() is associated with RSA so this will
-return B<NID_sha1WithRSAEncryption>. This "link" between digests and signature
-algorithms may not be retained in future versions of OpenSSL.
+return B<NID_sha1WithRSAEncryption>. Since digests and signature algorithms
+are no longer linked this function is only retained for compatibility
+reasons.
-EVP_md2(), EVP_md5(), EVP_sha(), EVP_sha1(), EVP_mdc2() and EVP_ripemd160()
-return B<EVP_MD> structures for the MD2, MD5, SHA, SHA1, MDC2 and RIPEMD160 digest
-algorithms respectively. The associated signature algorithm is RSA in each case.
+EVP_md2(), EVP_md5(), EVP_sha(), EVP_sha1(), EVP_sha224(), EVP_sha256(),
+EVP_sha384(), EVP_sha512(), EVP_mdc2() and EVP_ripemd160() return B<EVP_MD>
+structures for the MD2, MD5, SHA, SHA1, SHA224, SHA256, SHA384, SHA512, MDC2
+and RIPEMD160 digest algorithms respectively.
EVP_dss() and EVP_dss1() return B<EVP_MD> structures for SHA and SHA1 digest
-algorithms but using DSS (DSA) for the signature algorithm.
+algorithms but using DSS (DSA) for the signature algorithm. Note: there is
+no need to use these pseudo-digests in OpenSSL 1.0.0 and later, they are
+however retained for compatibility.
EVP_md_null() is a "null" message digest that does nothing: i.e. the hash it
returns is of zero length.
@@ -169,8 +178,8 @@ The B<EVP> interface to message digests should almost always be used in
preference to the low level interfaces. This is because the code then becomes
transparent to the digest used and much more flexible.
-SHA1 is the digest of choice for new applications. The other digest algorithms
-are still in common use.
+New applications should use the SHA2 digest algorithms such as SHA256.
+The other digest algorithms are still in common use.
For most applications the B<impl> parameter to EVP_DigestInit_ex() will be
set to NULL to use the default digest implementation.
@@ -185,6 +194,19 @@ implementations of digests to be specified.
In OpenSSL 0.9.7 and later if digest contexts are not cleaned up after use
memory leaks will occur.
+Stack allocation of EVP_MD_CTX structures is common, for example:
+
+ EVP_MD_CTX mctx;
+ EVP_MD_CTX_init(&mctx);
+
+This will cause binary compatibility issues if the size of EVP_MD_CTX
+structure changes (this will only happen with a major release of OpenSSL).
+Applications wishing to avoid this should use EVP_MD_CTX_create() instead:
+
+ EVP_MD_CTX *mctx;
+ mctx = EVP_MD_CTX_create();
+
+
=head1 EXAMPLE
This example digests the data "Test Message\n" and "Hello World\n", using the
@@ -195,7 +217,7 @@ digest name passed on the command line.
main(int argc, char *argv[])
{
- EVP_MD_CTX mdctx;
+ EVP_MD_CTX *mdctx;
const EVP_MD *md;
char mess1[] = "Test Message\n";
char mess2[] = "Hello World\n";
@@ -216,24 +238,18 @@ digest name passed on the command line.
exit(1);
}
- EVP_MD_CTX_init(&mdctx);
- EVP_DigestInit_ex(&mdctx, md, NULL);
- EVP_DigestUpdate(&mdctx, mess1, strlen(mess1));
- EVP_DigestUpdate(&mdctx, mess2, strlen(mess2));
- EVP_DigestFinal_ex(&mdctx, md_value, &md_len);
- EVP_MD_CTX_cleanup(&mdctx);
+ mdctx = EVP_MD_CTX_create();
+ EVP_DigestInit_ex(mdctx, md, NULL);
+ EVP_DigestUpdate(mdctx, mess1, strlen(mess1));
+ EVP_DigestUpdate(mdctx, mess2, strlen(mess2));
+ EVP_DigestFinal_ex(mdctx, md_value, &md_len);
+ EVP_MD_CTX_destroy(mdctx);
printf("Digest is: ");
for(i = 0; i < md_len; i++) printf("%02x", md_value[i]);
printf("\n");
}
-=head1 BUGS
-
-The link between digests and signing algorithms results in a situation where
-EVP_sha1() must be used with RSA and EVP_dss1() must be used with DSS
-even though they are identical digests.
-
=head1 SEE ALSO
L<evp(3)|evp(3)>, L<hmac(3)|hmac(3)>, L<md2(3)|md2(3)>,
@@ -253,4 +269,11 @@ EVP_md_null(), EVP_md2(), EVP_md5(), EVP_sha(), EVP_sha1(),
EVP_dss(), EVP_dss1(), EVP_mdc2() and EVP_ripemd160() were
changed to return truely const EVP_MD * in OpenSSL 0.9.7.
+The link between digests and signing algorithms was fixed in OpenSSL 1.0 and
+later, so now EVP_sha1() can be used with RSA and DSA, there is no need to
+use EVP_dss1() any more.
+
+OpenSSL 1.0 and later does not include the MD2 digest algorithm in the
+default configuration due to its security weaknesses.
+
=cut
diff --git a/doc/crypto/EVP_DigestSignInit.pod b/doc/crypto/EVP_DigestSignInit.pod
new file mode 100644
index 000000000000..37d960e3b229
--- /dev/null
+++ b/doc/crypto/EVP_DigestSignInit.pod
@@ -0,0 +1,87 @@
+=pod
+
+=head1 NAME
+
+EVP_DigestSignInit, EVP_DigestSignUpdate, EVP_DigestSignFinal - EVP signing functions
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_DigestSignInit(EVP_MD_CTX *ctx, EVP_PKEY_CTX **pctx,
+ const EVP_MD *type, ENGINE *e, EVP_PKEY *pkey);
+ int EVP_DigestSignUpdate(EVP_MD_CTX *ctx, const void *d, unsigned int cnt);
+ int EVP_DigestSignFinal(EVP_MD_CTX *ctx, unsigned char *sig, size_t *siglen);
+
+=head1 DESCRIPTION
+
+The EVP signature routines are a high level interface to digital signatures.
+
+EVP_DigestSignInit() sets up signing context B<ctx> to use digest B<type> from
+ENGINE B<impl> and private key B<pkey>. B<ctx> must be initialized with
+EVP_MD_CTX_init() before calling this function. If B<pctx> is not NULL the
+EVP_PKEY_CTX of the signing operation will be written to B<*pctx>: this can
+be used to set alternative signing options.
+
+EVP_DigestSignUpdate() hashes B<cnt> bytes of data at B<d> into the
+signature context B<ctx>. This function can be called several times on the
+same B<ctx> to include additional data. This function is currently implemented
+usig a macro.
+
+EVP_DigestSignFinal() signs the data in B<ctx> places the signature in B<sig>.
+If B<sig> is B<NULL> then the maximum size of the output buffer is written to
+the B<siglen> parameter. If B<sig> is not B<NULL> then before the call the
+B<siglen> parameter should contain the length of the B<sig> buffer, if the
+call is successful the signature is written to B<sig> and the amount of data
+written to B<siglen>.
+
+=head1 RETURN VALUES
+
+EVP_DigestSignInit() EVP_DigestSignUpdate() and EVP_DigestSignaFinal() return
+1 for success and 0 or a negative value for failure. In particular a return
+value of -2 indicates the operation is not supported by the public key
+algorithm.
+
+The error codes can be obtained from L<ERR_get_error(3)|ERR_get_error(3)>.
+
+=head1 NOTES
+
+The B<EVP> interface to digital signatures should almost always be used in
+preference to the low level interfaces. This is because the code then becomes
+transparent to the algorithm used and much more flexible.
+
+In previous versions of OpenSSL there was a link between message digest types
+and public key algorithms. This meant that "clone" digests such as EVP_dss1()
+needed to be used to sign using SHA1 and DSA. This is no longer necessary and
+the use of clone digest is now discouraged.
+
+For some key types and parameters the random number generator must be seeded
+or the operation will fail.
+
+The call to EVP_DigestSignFinal() internally finalizes a copy of the digest
+context. This means that calls to EVP_DigestSignUpdate() and
+EVP_DigestSignFinal() can be called later to digest and sign additional data.
+
+Since only a copy of the digest context is ever finalized the context must
+be cleaned up after use by calling EVP_MD_CTX_cleanup() or a memory leak
+will occur.
+
+The use of EVP_PKEY_size() with these functions is discouraged because some
+signature operations may have a signature length which depends on the
+parameters set. As a result EVP_PKEY_size() would have to return a value
+which indicates the maximum possible signature for any set of parameters.
+
+=head1 SEE ALSO
+
+L<EVP_DigestVerifyInit(3)|EVP_DigestVerifyInit(3)>,
+L<EVP_DigestInit(3)|EVP_DigestInit(3)>, L<err(3)|err(3)>,
+L<evp(3)|evp(3)>, L<hmac(3)|hmac(3)>, L<md2(3)|md2(3)>,
+L<md5(3)|md5(3)>, L<mdc2(3)|mdc2(3)>, L<ripemd(3)|ripemd(3)>,
+L<sha(3)|sha(3)>, L<dgst(1)|dgst(1)>
+
+=head1 HISTORY
+
+EVP_DigestSignInit(), EVP_DigestSignUpdate() and EVP_DigestSignFinal()
+were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_DigestVerifyInit.pod b/doc/crypto/EVP_DigestVerifyInit.pod
new file mode 100644
index 000000000000..f22448897839
--- /dev/null
+++ b/doc/crypto/EVP_DigestVerifyInit.pod
@@ -0,0 +1,82 @@
+=pod
+
+=head1 NAME
+
+EVP_DigestVerifyInit, EVP_DigestVerifyUpdate, EVP_DigestVerifyFinal - EVP signature verification functions
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_DigestVerifyInit(EVP_MD_CTX *ctx, EVP_PKEY_CTX **pctx,
+ const EVP_MD *type, ENGINE *e, EVP_PKEY *pkey);
+ int EVP_DigestVerifyUpdate(EVP_MD_CTX *ctx, const void *d, unsigned int cnt);
+ int EVP_DigestVerifyFinal(EVP_MD_CTX *ctx, unsigned char *sig, size_t siglen);
+
+=head1 DESCRIPTION
+
+The EVP signature routines are a high level interface to digital signatures.
+
+EVP_DigestVerifyInit() sets up verification context B<ctx> to use digest
+B<type> from ENGINE B<impl> and public key B<pkey>. B<ctx> must be initialized
+with EVP_MD_CTX_init() before calling this function. If B<pctx> is not NULL the
+EVP_PKEY_CTX of the verification operation will be written to B<*pctx>: this
+can be used to set alternative verification options.
+
+EVP_DigestVerifyUpdate() hashes B<cnt> bytes of data at B<d> into the
+verification context B<ctx>. This function can be called several times on the
+same B<ctx> to include additional data. This function is currently implemented
+using a macro.
+
+EVP_DigestVerifyFinal() verifies the data in B<ctx> against the signature in
+B<sig> of length B<siglen>.
+
+=head1 RETURN VALUES
+
+EVP_DigestVerifyInit() and EVP_DigestVerifyUpdate() return 1 for success and 0
+or a negative value for failure. In particular a return value of -2 indicates
+the operation is not supported by the public key algorithm.
+
+Unlike other functions the return value 0 from EVP_DigestVerifyFinal() only
+indicates that the signature did not not verify successfully (that is tbs did
+not match the original data or the signature was of invalid form) it is not an
+indication of a more serious error.
+
+The error codes can be obtained from L<ERR_get_error(3)|ERR_get_error(3)>.
+
+=head1 NOTES
+
+The B<EVP> interface to digital signatures should almost always be used in
+preference to the low level interfaces. This is because the code then becomes
+transparent to the algorithm used and much more flexible.
+
+In previous versions of OpenSSL there was a link between message digest types
+and public key algorithms. This meant that "clone" digests such as EVP_dss1()
+needed to be used to sign using SHA1 and DSA. This is no longer necessary and
+the use of clone digest is now discouraged.
+
+For some key types and parameters the random number generator must be seeded
+or the operation will fail.
+
+The call to EVP_DigestVerifyFinal() internally finalizes a copy of the digest
+context. This means that calls to EVP_VerifyUpdate() and EVP_VerifyFinal() can
+be called later to digest and verify additional data.
+
+Since only a copy of the digest context is ever finalized the context must
+be cleaned up after use by calling EVP_MD_CTX_cleanup() or a memory leak
+will occur.
+
+=head1 SEE ALSO
+
+L<EVP_DigestSignInit(3)|EVP_DigestSignInit(3)>,
+L<EVP_DigestInit(3)|EVP_DigestInit(3)>, L<err(3)|err(3)>,
+L<evp(3)|evp(3)>, L<hmac(3)|hmac(3)>, L<md2(3)|md2(3)>,
+L<md5(3)|md5(3)>, L<mdc2(3)|mdc2(3)>, L<ripemd(3)|ripemd(3)>,
+L<sha(3)|sha(3)>, L<dgst(1)|dgst(1)>
+
+=head1 HISTORY
+
+EVP_DigestVerifyInit(), EVP_DigestVerifyUpdate() and EVP_DigestVerifyFinal()
+were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_CTX_ctrl.pod b/doc/crypto/EVP_PKEY_CTX_ctrl.pod
new file mode 100644
index 000000000000..f2f455990f54
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_CTX_ctrl.pod
@@ -0,0 +1,128 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_ctrl, EVP_PKEY_ctrl_str - algorithm specific control operations
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_CTX_ctrl(EVP_PKEY_CTX *ctx, int keytype, int optype,
+ int cmd, int p1, void *p2);
+ int EVP_PKEY_CTX_ctrl_str(EVP_PKEY_CTX *ctx, const char *type,
+ const char *value);
+
+ int EVP_PKEY_get_default_digest_nid(EVP_PKEY *pkey, int *pnid);
+
+ #include <openssl/rsa.h>
+
+ int EVP_PKEY_CTX_set_signature_md(EVP_PKEY_CTX *ctx, const EVP_MD *md);
+
+ int EVP_PKEY_CTX_set_rsa_padding(EVP_PKEY_CTX *ctx, int pad);
+ int EVP_PKEY_CTX_set_rsa_pss_saltlen(EVP_PKEY_CTX *ctx, int len);
+ int EVP_PKEY_CTX_set_rsa_rsa_keygen_bits(EVP_PKEY_CTX *ctx, int mbits);
+ int EVP_PKEY_CTX_set_rsa_keygen_pubexp(EVP_PKEY_CTX *ctx, BIGNUM *pubexp);
+
+ #include <openssl/dsa.h>
+ int EVP_PKEY_CTX_set_dsa_paramgen_bits(EVP_PKEY_CTX *ctx, int nbits);
+
+ #include <openssl/dh.h>
+ int EVP_PKEY_CTX_set_dh_paramgen_prime_len(EVP_PKEY_CTX *ctx, int len);
+ int EVP_PKEY_CTX_set_dh_paramgen_generator(EVP_PKEY_CTX *ctx, int gen);
+
+ #include <openssl/ec.h>
+ int EVP_PKEY_CTX_set_ec_paramgen_curve_nid(EVP_PKEY_CTX *ctx, int nid);
+
+=head1 DESCRIPTION
+
+The function EVP_PKEY_CTX_ctrl() sends a control operation to the context
+B<ctx>. The key type used must match B<keytype> if it is not -1. The parameter
+B<optype> is a mask indicating which operations the control can be applied to.
+The control command is indicated in B<cmd> and any additional arguments in
+B<p1> and B<p2>.
+
+Applications will not normally call EVP_PKEY_CTX_ctrl() directly but will
+instead call one of the algorithm specific macros below.
+
+The function EVP_PKEY_ctrl_str() allows an application to send an algorithm
+specific control operation to a context B<ctx> in string form. This is
+intended to be used for options specified on the command line or in text
+files. The commands supported are documented in the openssl utility
+command line pages for the option B<-pkeyopt> which is supported by the
+B<pkeyutl>, B<genpkey> and B<req> commands.
+
+All the remaining "functions" are implemented as macros.
+
+The EVP_PKEY_CTX_set_signature_md() macro sets the message digest type used
+in a signature. It can be used with any public key algorithm supporting
+signature operations.
+
+The macro EVP_PKEY_CTX_set_rsa_padding() sets the RSA padding mode for B<ctx>.
+The B<pad> parameter can take the value RSA_PKCS1_PADDING for PKCS#1 padding,
+RSA_SSLV23_PADDING for SSLv23 padding, RSA_NO_PADDING for no padding,
+RSA_PKCS1_OAEP_PADDING for OAEP padding (encrypt and decrypt only),
+RSA_X931_PADDING for X9.31 padding (signature operations only) and
+RSA_PKCS1_PSS_PADDING (sign and verify only).
+
+Two RSA padding modes behave differently if EVP_PKEY_CTX_set_signature_md()
+is used. If this macro is called for PKCS#1 padding the plaintext buffer is
+an actual digest value and is encapsulated in a DigestInfo structure according
+to PKCS#1 when signing and this structure is expected (and stripped off) when
+verifying. If this control is not used with RSA and PKCS#1 padding then the
+supplied data is used directly and not encapsulated. In the case of X9.31
+padding for RSA the algorithm identifier byte is added or checked and removed
+if this control is called. If it is not called then the first byte of the plaintext buffer is expected to be the algorithm identifier byte.
+
+The EVP_PKEY_CTX_set_rsa_pss_saltlen() macro sets the RSA PSS salt length to
+B<len> as its name implies it is only supported for PSS padding. Two special
+values are supported: -1 sets the salt length to the digest length. When
+signing -2 sets the salt length to the maximum permissible value. When
+verifying -2 causes the salt length to be automatically determined based on the
+B<PSS> block structure. If this macro is not called a salt length value of -2
+is used by default.
+
+The EVP_PKEY_CTX_set_rsa_rsa_keygen_bits() macro sets the RSA key length for
+RSA key genration to B<bits>. If not specified 1024 bits is used.
+
+The EVP_PKEY_CTX_set_rsa_keygen_pubexp() macro sets the public exponent value
+for RSA key generation to B<pubexp> currently it should be an odd integer. The
+B<pubexp> pointer is used internally by this function so it should not be
+modified or free after the call. If this macro is not called then 65537 is used.
+
+The macro EVP_PKEY_CTX_set_dsa_paramgen_bits() sets the number of bits used
+for DSA parameter generation to B<bits>. If not specified 1024 is used.
+
+The macro EVP_PKEY_CTX_set_dh_paramgen_prime_len() sets the length of the DH
+prime parameter B<p> for DH parameter generation. If this macro is not called
+then 1024 is used.
+
+The EVP_PKEY_CTX_set_dh_paramgen_generator() macro sets DH generator to B<gen>
+for DH parameter generation. If not specified 2 is used.
+
+The EVP_PKEY_CTX_set_ec_paramgen_curve_nid() sets the EC curve for EC parameter
+generation to B<nid>. For EC parameter generation this macro must be called
+or an error occurs because there is no default curve.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_CTX_ctrl() and its macros return a positive value for success and 0
+or a negative value for failure. In particular a return value of -2
+indicates the operation is not supported by the public key algorithm.
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+L<EVP_PKEY_keygen(3)|EVP_PKEY_keygen(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_CTX_new.pod b/doc/crypto/EVP_PKEY_CTX_new.pod
new file mode 100644
index 000000000000..a9af8675801b
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_CTX_new.pod
@@ -0,0 +1,52 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_CTX_new, EVP_PKEY_CTX_new_id, EVP_PKEY_CTX_dup, EVP_PKEY_CTX_free - public key algorithm context functions.
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ EVP_PKEY_CTX *EVP_PKEY_CTX_new(EVP_PKEY *pkey, ENGINE *e);
+ EVP_PKEY_CTX *EVP_PKEY_CTX_new_id(int id, ENGINE *e);
+ EVP_PKEY_CTX *EVP_PKEY_CTX_dup(EVP_PKEY_CTX *ctx);
+ void EVP_PKEY_CTX_free(EVP_PKEY_CTX *ctx);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_CTX_new() function allocates public key algorithm context using
+the algorithm specified in B<pkey> and ENGINE B<e>.
+
+The EVP_PKEY_CTX_new_id() function allocates public key algorithm context
+using the algorithm specified by B<id> and ENGINE B<e>. It is normally used
+when no B<EVP_PKEY> structure is associated with the operations, for example
+during parameter generation of key genration for some algorithms.
+
+EVP_PKEY_CTX_dup() duplicates the context B<ctx>.
+
+EVP_PKEY_CTX_free() frees up the context B<ctx>.
+
+=head1 NOTES
+
+The B<EVP_PKEY_CTX> structure is an opaque public key algorithm context used
+by the OpenSSL high level public key API. Contexts B<MUST NOT> be shared between
+threads: that is it is not permissible to use the same context simultaneously
+in two threads.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_CTX_new(), EVP_PKEY_CTX_new_id(), EVP_PKEY_CTX_dup() returns either
+the newly allocated B<EVP_PKEY_CTX> structure of B<NULL> if an error occurred.
+
+EVP_PKEY_CTX_free() does not return a value.
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_new(3)|EVP_PKEY_new(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_cmp.pod b/doc/crypto/EVP_PKEY_cmp.pod
new file mode 100644
index 000000000000..4f8185e36cdf
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_cmp.pod
@@ -0,0 +1,61 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_copy_parameters, EVP_PKEY_missing_parameters, EVP_PKEY_cmp_parameters, EVP_PKEY_cmp - public key parameter and comparison functions
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_missing_parameters(const EVP_PKEY *pkey);
+ int EVP_PKEY_copy_parameters(EVP_PKEY *to, const EVP_PKEY *from);
+
+ int EVP_PKEY_cmp_parameters(const EVP_PKEY *a, const EVP_PKEY *b);
+ int EVP_PKEY_cmp(const EVP_PKEY *a, const EVP_PKEY *b);
+
+=head1 DESCRIPTION
+
+The function EVP_PKEY_missing_parameters() returns 1 if the public key
+parameters of B<pkey> are missing and 0 if they are present or the algorithm
+doesn't use parameters.
+
+The function EVP_PKEY_copy_parameters() copies the parameters from key
+B<from> to key B<to>.
+
+The funcion EVP_PKEY_cmp_parameters() compares the parameters of keys
+B<a> and B<b>.
+
+The funcion EVP_PKEY_cmp() compares the public key components and paramters
+(if present) of keys B<a> and B<b>.
+
+=head1 NOTES
+
+The main purpose of the functions EVP_PKEY_missing_parameters() and
+EVP_PKEY_copy_parameters() is to handle public keys in certificates where the
+parameters are sometimes omitted from a public key if they are inherited from
+the CA that signed it.
+
+Since OpenSSL private keys contain public key components too the function
+EVP_PKEY_cmp() can also be used to determine if a private key matches
+a public key.
+
+=head1 RETURN VALUES
+
+The function EVP_PKEY_missing_parameters() returns 1 if the public key
+parameters of B<pkey> are missing and 0 if they are present or the algorithm
+doesn't use parameters.
+
+These functions EVP_PKEY_copy_parameters() returns 1 for success and 0 for
+failure.
+
+The function EVP_PKEY_cmp_parameters() and EVP_PKEY_cmp() return 1 if the
+keys match, 0 if they don't match, -1 if the key types are different and
+-2 if the operation is not supported.
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_keygen(3)|EVP_PKEY_keygen(3)>
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_decrypt.pod b/doc/crypto/EVP_PKEY_decrypt.pod
new file mode 100644
index 000000000000..42b2a8c44ed1
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_decrypt.pod
@@ -0,0 +1,93 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_decrypt_init, EVP_PKEY_decrypt - decrypt using a public key algorithm
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_decrypt_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_decrypt(EVP_PKEY_CTX *ctx,
+ unsigned char *out, size_t *outlen,
+ const unsigned char *in, size_t inlen);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_decrypt_init() function initializes a public key algorithm
+context using key B<pkey> for a decryption operation.
+
+The EVP_PKEY_decrypt() function performs a public key decryption operation
+using B<ctx>. The data to be decrypted is specified using the B<in> and
+B<inlen> parameters. If B<out> is B<NULL> then the maximum size of the output
+buffer is written to the B<outlen> parameter. If B<out> is not B<NULL> then
+before the call the B<outlen> parameter should contain the length of the
+B<out> buffer, if the call is successful the decrypted data is written to
+B<out> and the amount of data written to B<outlen>.
+
+=head1 NOTES
+
+After the call to EVP_PKEY_decrypt_init() algorithm specific control
+operations can be performed to set any appropriate parameters for the
+operation.
+
+The function EVP_PKEY_decrypt() can be called more than once on the same
+context if several operations are performed using the same parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_decrypt_init() and EVP_PKEY_decrypt() return 1 for success and 0
+or a negative value for failure. In particular a return value of -2
+indicates the operation is not supported by the public key algorithm.
+
+=head1 EXAMPLE
+
+Decrypt data using OAEP (for RSA keys):
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ unsigned char *out, *in;
+ size_t outlen, inlen;
+ EVP_PKEY *key;
+ /* NB: assumes key in, inlen are already set up
+ * and that key is an RSA private key
+ */
+ ctx = EVP_PKEY_CTX_new(key);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_decrypt_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_rsa_padding(ctx, RSA_OAEP_PADDING) <= 0)
+ /* Error */
+
+ /* Determine buffer length */
+ if (EVP_PKEY_decrypt(ctx, NULL, &outlen, in, inlen) <= 0)
+ /* Error */
+
+ out = OPENSSL_malloc(outlen);
+
+ if (!out)
+ /* malloc failure */
+
+ if (EVP_PKEY_decrypt(ctx, out, &outlen, in, inlen) <= 0)
+ /* Error */
+
+ /* Decrypted data is outlen bytes written to buffer out */
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_derive.pod b/doc/crypto/EVP_PKEY_derive.pod
new file mode 100644
index 000000000000..d9d6d76c7211
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_derive.pod
@@ -0,0 +1,93 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_derive_init, EVP_PKEY_derive_set_peer, EVP_PKEY_derive - derive public key algorithm shared secret.
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_derive_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_derive_set_peer(EVP_PKEY_CTX *ctx, EVP_PKEY *peer);
+ int EVP_PKEY_derive(EVP_PKEY_CTX *ctx, unsigned char *key, size_t *keylen);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_derive_init() function initializes a public key algorithm
+context using key B<pkey> for shared secret derivation.
+
+The EVP_PKEY_derive_set_peer() function sets the peer key: this will normally
+be a public key.
+
+The EVP_PKEY_derive() derives a shared secret using B<ctx>.
+If B<key> is B<NULL> then the maximum size of the output buffer is written to
+the B<keylen> parameter. If B<key> is not B<NULL> then before the call the
+B<keylen> parameter should contain the length of the B<key> buffer, if the call
+is successful the shared secret is written to B<key> and the amount of data
+written to B<keylen>.
+
+=head1 NOTES
+
+After the call to EVP_PKEY_derive_init() algorithm specific control
+operations can be performed to set any appropriate parameters for the
+operation.
+
+The function EVP_PKEY_derive() can be called more than once on the same
+context if several operations are performed using the same parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_derive_init() and EVP_PKEY_derive() return 1 for success and 0
+or a negative value for failure. In particular a return value of -2
+indicates the operation is not supported by the public key algorithm.
+
+=head1 EXAMPLE
+
+Derive shared secret (for example DH or EC keys):
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ unsigned char *skey;
+ size_t skeylen;
+ EVP_PKEY *pkey, *peerkey;
+ /* NB: assumes pkey, peerkey have been already set up */
+
+ ctx = EVP_PKEY_CTX_new(pkey);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_derive_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_derive_set_peer(ctx, peerkey) <= 0)
+ /* Error */
+
+ /* Determine buffer length */
+ if (EVP_PKEY_derive(ctx, NULL, &skeylen) <= 0)
+ /* Error */
+
+ skey = OPENSSL_malloc(skeylen);
+
+ if (!skey)
+ /* malloc failure */
+
+ if (EVP_PKEY_derive(ctx, skey, &skeylen) <= 0)
+ /* Error */
+
+ /* Shared secret is skey bytes written to buffer skey */
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_encrypt.pod b/doc/crypto/EVP_PKEY_encrypt.pod
new file mode 100644
index 000000000000..91c9c5d0a5d3
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_encrypt.pod
@@ -0,0 +1,93 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_encrypt_init, EVP_PKEY_encrypt - encrypt using a public key algorithm
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_encrypt_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_encrypt(EVP_PKEY_CTX *ctx,
+ unsigned char *out, size_t *outlen,
+ const unsigned char *in, size_t inlen);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_encrypt_init() function initializes a public key algorithm
+context using key B<pkey> for an encryption operation.
+
+The EVP_PKEY_encrypt() function performs a public key encryption operation
+using B<ctx>. The data to be encrypted is specified using the B<in> and
+B<inlen> parameters. If B<out> is B<NULL> then the maximum size of the output
+buffer is written to the B<outlen> parameter. If B<out> is not B<NULL> then
+before the call the B<outlen> parameter should contain the length of the
+B<out> buffer, if the call is successful the encrypted data is written to
+B<out> and the amount of data written to B<outlen>.
+
+=head1 NOTES
+
+After the call to EVP_PKEY_encrypt_init() algorithm specific control
+operations can be performed to set any appropriate parameters for the
+operation.
+
+The function EVP_PKEY_encrypt() can be called more than once on the same
+context if several operations are performed using the same parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_encrypt_init() and EVP_PKEY_encrypt() return 1 for success and 0
+or a negative value for failure. In particular a return value of -2
+indicates the operation is not supported by the public key algorithm.
+
+=head1 EXAMPLE
+
+Encrypt data using OAEP (for RSA keys):
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ unsigned char *out, *in;
+ size_t outlen, inlen;
+ EVP_PKEY *key;
+ /* NB: assumes key in, inlen are already set up
+ * and that key is an RSA public key
+ */
+ ctx = EVP_PKEY_CTX_new(key);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_encrypt_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_rsa_padding(ctx, RSA_OAEP_PADDING) <= 0)
+ /* Error */
+
+ /* Determine buffer length */
+ if (EVP_PKEY_encrypt(ctx, NULL, &outlen, in, inlen) <= 0)
+ /* Error */
+
+ out = OPENSSL_malloc(outlen);
+
+ if (!out)
+ /* malloc failure */
+
+ if (EVP_PKEY_encrypt(ctx, out, &outlen, in, inlen) <= 0)
+ /* Error */
+
+ /* Encrypted data is outlen bytes written to buffer out */
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_get_default_digest.pod b/doc/crypto/EVP_PKEY_get_default_digest.pod
new file mode 100644
index 000000000000..1a9c7954c55a
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_get_default_digest.pod
@@ -0,0 +1,41 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_get_default_digest_nid - get default signature digest
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+ int EVP_PKEY_get_default_digest_nid(EVP_PKEY *pkey, int *pnid);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_get_default_digest_nid() function sets B<pnid> to the default
+message digest NID for the public key signature operations associated with key
+B<pkey>.
+
+=head1 NOTES
+
+For all current standard OpenSSL public key algorithms SHA1 is returned.
+
+=head1 RETURN VALUES
+
+The EVP_PKEY_get_default_digest_nid() function returns 1 if the message digest
+is advisory (that is other digests can be used) and 2 if it is mandatory (other
+digests can not be used). It returns 0 or a negative value for failure. In
+particular a return value of -2 indicates the operation is not supported by the
+public key algorithm.
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+
+=head1 HISTORY
+
+This function was first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_keygen.pod b/doc/crypto/EVP_PKEY_keygen.pod
new file mode 100644
index 000000000000..37c6fe95030d
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_keygen.pod
@@ -0,0 +1,161 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_keygen_init, EVP_PKEY_keygen, EVP_PKEY_paramgen_init, EVP_PKEY_paramgen, EVP_PKEY_CTX_set_cb, EVP_PKEY_CTX_get_cb, EVP_PKEY_CTX_get_keygen_info, EVP_PKEVP_PKEY_CTX_set_app_data, EVP_PKEY_CTX_get_app_data - key and parameter generation functions
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_keygen_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_keygen(EVP_PKEY_CTX *ctx, EVP_PKEY **ppkey);
+ int EVP_PKEY_paramgen_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_paramgen(EVP_PKEY_CTX *ctx, EVP_PKEY **ppkey);
+
+ typedef int EVP_PKEY_gen_cb(EVP_PKEY_CTX *ctx);
+
+ void EVP_PKEY_CTX_set_cb(EVP_PKEY_CTX *ctx, EVP_PKEY_gen_cb *cb);
+ EVP_PKEY_gen_cb *EVP_PKEY_CTX_get_cb(EVP_PKEY_CTX *ctx);
+
+ int EVP_PKEY_CTX_get_keygen_info(EVP_PKEY_CTX *ctx, int idx);
+
+ void EVP_PKEY_CTX_set_app_data(EVP_PKEY_CTX *ctx, void *data);
+ void *EVP_PKEY_CTX_get_app_data(EVP_PKEY_CTX *ctx);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_keygen_init() function initializes a public key algorithm
+context using key B<pkey> for a key genration operation.
+
+The EVP_PKEY_keygen() function performs a key generation operation, the
+generated key is written to B<ppkey>.
+
+The functions EVP_PKEY_paramgen_init() and EVP_PKEY_paramgen() are similar
+except parameters are generated.
+
+The function EVP_PKEY_set_cb() sets the key or parameter generation callback
+to B<cb>. The function EVP_PKEY_CTX_get_cb() returns the key or parameter
+generation callback.
+
+The function EVP_PKEY_CTX_get_keygen_info() returns parameters associated
+with the generation operation. If B<idx> is -1 the total number of
+parameters available is returned. Any non negative value returns the value of
+that parameter. EVP_PKEY_CTX_gen_keygen_info() with a non-negative value for
+B<idx> should only be called within the generation callback.
+
+If the callback returns 0 then the key genration operation is aborted and an
+error occurs. This might occur during a time consuming operation where
+a user clicks on a "cancel" button.
+
+The functions EVP_PKEY_CTX_set_app_data() and EVP_PKEY_CTX_get_app_data() set
+and retrieve an opaque pointer. This can be used to set some application
+defined value which can be retrieved in the callback: for example a handle
+which is used to update a "progress dialog".
+
+=head1 NOTES
+
+After the call to EVP_PKEY_keygen_init() or EVP_PKEY_paramgen_init() algorithm
+specific control operations can be performed to set any appropriate parameters
+for the operation.
+
+The functions EVP_PKEY_keygen() and EVP_PKEY_paramgen() can be called more than
+once on the same context if several operations are performed using the same
+parameters.
+
+The meaning of the parameters passed to the callback will depend on the
+algorithm and the specifiic implementation of the algorithm. Some might not
+give any useful information at all during key or parameter generation. Others
+might not even call the callback.
+
+The operation performed by key or parameter generation depends on the algorithm
+used. In some cases (e.g. EC with a supplied named curve) the "generation"
+option merely sets the appropriate fields in an EVP_PKEY structure.
+
+In OpenSSL an EVP_PKEY structure containing a private key also contains the
+public key components and parameters (if any). An OpenSSL private key is
+equivalent to what some libraries call a "key pair". A private key can be used
+in functions which require the use of a public key or parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_keygen_init(), EVP_PKEY_paramgen_init(), EVP_PKEY_keygen() and
+EVP_PKEY_paramgen() return 1 for success and 0 or a negative value for failure.
+In particular a return value of -2 indicates the operation is not supported by
+the public key algorithm.
+
+=head1 EXAMPLES
+
+Generate a 2048 bit RSA key:
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ EVP_PKEY *pkey = NULL;
+ ctx = EVP_PKEY_CTX_new_id(EVP_PKEY_RSA, NULL);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_keygen_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_rsa_keygen_bits(ctx, 2048) <= 0)
+ /* Error */
+
+ /* Generate key */
+ if (EVP_PKEY_keygen(ctx, &pkey) <= 0)
+ /* Error */
+
+Generate a key from a set of parameters:
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ EVP_PKEY *pkey = NULL, *param;
+ /* Assumed param is set up already */
+ ctx = EVP_PKEY_CTX_new(param);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_keygen_init(ctx) <= 0)
+ /* Error */
+
+ /* Generate key */
+ if (EVP_PKEY_keygen(ctx, &pkey) <= 0)
+ /* Error */
+
+Example of generation callback for OpenSSL public key implementations:
+
+ /* Application data is a BIO to output status to */
+
+ EVP_PKEY_CTX_set_app_data(ctx, status_bio);
+
+ static int genpkey_cb(EVP_PKEY_CTX *ctx)
+ {
+ char c='*';
+ BIO *b = EVP_PKEY_CTX_get_app_data(ctx);
+ int p;
+ p = EVP_PKEY_CTX_get_keygen_info(ctx, 0);
+ if (p == 0) c='.';
+ if (p == 1) c='+';
+ if (p == 2) c='*';
+ if (p == 3) c='\n';
+ BIO_write(b,&c,1);
+ (void)BIO_flush(b);
+ return 1;
+ }
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_print_private.pod b/doc/crypto/EVP_PKEY_print_private.pod
new file mode 100644
index 000000000000..ce9d70d7a7a0
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_print_private.pod
@@ -0,0 +1,53 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_print_public, EVP_PKEY_print_private, EVP_PKEY_print_params - public key algorithm printing routines.
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_print_public(BIO *out, const EVP_PKEY *pkey,
+ int indent, ASN1_PCTX *pctx);
+ int EVP_PKEY_print_private(BIO *out, const EVP_PKEY *pkey,
+ int indent, ASN1_PCTX *pctx);
+ int EVP_PKEY_print_params(BIO *out, const EVP_PKEY *pkey,
+ int indent, ASN1_PCTX *pctx);
+
+=head1 DESCRIPTION
+
+The functions EVP_PKEY_print_public(), EVP_PKEY_print_private() and
+EVP_PKEY_print_params() print out the public, private or parameter components
+of key B<pkey> respectively. The key is sent to BIO B<out> in human readable
+form. The parameter B<indent> indicated how far the printout should be indented.
+
+The B<pctx> parameter allows the print output to be finely tuned by using
+ASN1 printing options. If B<pctx> is set to NULL then default values will
+be used.
+
+=head1 NOTES
+
+Currently no public key algorithms include any options in the B<pctx> parameter
+parameter.
+
+If the key does not include all the components indicated by the function then
+only those contained in the key will be printed. For example passing a public
+key to EVP_PKEY_print_private() will only print the public components.
+
+=head1 RETURN VALUES
+
+These functions all return 1 for success and 0 or a negative value for failure.
+In particular a return value of -2 indicates the operation is not supported by
+the public key algorithm.
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_keygen(3)|EVP_PKEY_keygen(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_sign.pod b/doc/crypto/EVP_PKEY_sign.pod
new file mode 100644
index 000000000000..2fb52c34863f
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_sign.pod
@@ -0,0 +1,96 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_sign_init, EVP_PKEY_sign - sign using a public key algorithm
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_sign_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_sign(EVP_PKEY_CTX *ctx,
+ unsigned char *sig, size_t *siglen,
+ const unsigned char *tbs, size_t tbslen);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_sign_init() function initializes a public key algorithm
+context using key B<pkey> for a signing operation.
+
+The EVP_PKEY_sign() function performs a public key signing operation
+using B<ctx>. The data to be signed is specified using the B<tbs> and
+B<tbslen> parameters. If B<sig> is B<NULL> then the maximum size of the output
+buffer is written to the B<siglen> parameter. If B<sig> is not B<NULL> then
+before the call the B<siglen> parameter should contain the length of the
+B<sig> buffer, if the call is successful the signature is written to
+B<sig> and the amount of data written to B<siglen>.
+
+=head1 NOTES
+
+After the call to EVP_PKEY_sign_init() algorithm specific control
+operations can be performed to set any appropriate parameters for the
+operation.
+
+The function EVP_PKEY_sign() can be called more than once on the same
+context if several operations are performed using the same parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_sign_init() and EVP_PKEY_sign() return 1 for success and 0
+or a negative value for failure. In particular a return value of -2
+indicates the operation is not supported by the public key algorithm.
+
+=head1 EXAMPLE
+
+Sign data using RSA with PKCS#1 padding and SHA256 digest:
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ unsigned char *md, *sig;
+ size_t mdlen, siglen;
+ EVP_PKEY *signing_key;
+ /* NB: assumes signing_key, md and mdlen are already set up
+ * and that signing_key is an RSA private key
+ */
+ ctx = EVP_PKEY_CTX_new(signing_key);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_sign_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_rsa_padding(ctx, RSA_PKCS1_PADDING) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_signature_md(ctx, EVP_sha256()) <= 0)
+ /* Error */
+
+ /* Determine buffer length */
+ if (EVP_PKEY_sign(ctx, NULL, &siglen, md, mdlen) <= 0)
+ /* Error */
+
+ sig = OPENSSL_malloc(siglen);
+
+ if (!sig)
+ /* malloc failure */
+
+ if (EVP_PKEY_sign(ctx, sig, &siglen, md, mdlen) <= 0)
+ /* Error */
+
+ /* Signature is siglen bytes written to buffer sig */
+
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_verify.pod b/doc/crypto/EVP_PKEY_verify.pod
new file mode 100644
index 000000000000..f93e5fc6c3af
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_verify.pod
@@ -0,0 +1,91 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_verify_init, EVP_PKEY_verify - signature verification using a public key algorithm
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_verify_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_verify(EVP_PKEY_CTX *ctx,
+ const unsigned char *sig, size_t siglen,
+ const unsigned char *tbs, size_t tbslen);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_verify_init() function initializes a public key algorithm
+context using key B<pkey> for a signature verification operation.
+
+The EVP_PKEY_verify() function performs a public key verification operation
+using B<ctx>. The signature is specified using the B<sig> and
+B<siglen> parameters. The verified data (i.e. the data believed originally
+signed) is specified using the B<tbs> and B<tbslen> parameters.
+
+=head1 NOTES
+
+After the call to EVP_PKEY_verify_init() algorithm specific control
+operations can be performed to set any appropriate parameters for the
+operation.
+
+The function EVP_PKEY_verify() can be called more than once on the same
+context if several operations are performed using the same parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_verify_init() and EVP_PKEY_verify() return 1 if the verification was
+successful and 0 if it failed. Unlike other functions the return value 0 from
+EVP_PKEY_verify() only indicates that the signature did not not verify
+successfully (that is tbs did not match the original data or the signature was
+of invalid form) it is not an indication of a more serious error.
+
+A negative value indicates an error other that signature verification failure.
+In particular a return value of -2 indicates the operation is not supported by
+the public key algorithm.
+
+=head1 EXAMPLE
+
+Verify signature using PKCS#1 and SHA256 digest:
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ unsigned char *md, *sig;
+ size_t mdlen, siglen;
+ EVP_PKEY *verify_key;
+ /* NB: assumes verify_key, sig, siglen md and mdlen are already set up
+ * and that verify_key is an RSA public key
+ */
+ ctx = EVP_PKEY_CTX_new(verify_key);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_verify_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_rsa_padding(ctx, RSA_PKCS1_PADDING) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_signature_md(ctx, EVP_sha256()) <= 0)
+ /* Error */
+
+ /* Perform operation */
+ ret = EVP_PKEY_verify(ctx, sig, siglen, md, mdlen);
+
+ /* ret == 1 indicates success, 0 verify failure and < 0 for some
+ * other error.
+ */
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verifyrecover(3)|EVP_PKEY_verifyrecover(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_PKEY_verifyrecover.pod b/doc/crypto/EVP_PKEY_verifyrecover.pod
new file mode 100644
index 000000000000..f3605eb8264e
--- /dev/null
+++ b/doc/crypto/EVP_PKEY_verifyrecover.pod
@@ -0,0 +1,103 @@
+=pod
+
+=head1 NAME
+
+EVP_PKEY_verifyrecover_init, EVP_PKEY_verifyrecover - recover signature using a public key algorithm
+
+=head1 SYNOPSIS
+
+ #include <openssl/evp.h>
+
+ int EVP_PKEY_verifyrecover_init(EVP_PKEY_CTX *ctx);
+ int EVP_PKEY_verifyrecover(EVP_PKEY_CTX *ctx,
+ unsigned char *rout, size_t *routlen,
+ const unsigned char *sig, size_t siglen);
+
+=head1 DESCRIPTION
+
+The EVP_PKEY_verifyrecover_init() function initializes a public key algorithm
+context using key B<pkey> for a verify recover operation.
+
+The EVP_PKEY_verifyrecover() function recovers signed data
+using B<ctx>. The signature is specified using the B<sig> and
+B<siglen> parameters. If B<rout> is B<NULL> then the maximum size of the output
+buffer is written to the B<routlen> parameter. If B<rout> is not B<NULL> then
+before the call the B<routlen> parameter should contain the length of the
+B<rout> buffer, if the call is successful recovered data is written to
+B<rout> and the amount of data written to B<routlen>.
+
+=head1 NOTES
+
+Normally an application is only interested in whether a signature verification
+operation is successful in those cases the EVP_verify() function should be
+used.
+
+Sometimes however it is useful to obtain the data originally signed using a
+signing operation. Only certain public key algorithms can recover a signature
+in this way (for example RSA in PKCS padding mode).
+
+After the call to EVP_PKEY_verifyrecover_init() algorithm specific control
+operations can be performed to set any appropriate parameters for the
+operation.
+
+The function EVP_PKEY_verifyrecover() can be called more than once on the same
+context if several operations are performed using the same parameters.
+
+=head1 RETURN VALUES
+
+EVP_PKEY_verifyrecover_init() and EVP_PKEY_verifyrecover() return 1 for success
+and 0 or a negative value for failure. In particular a return value of -2
+indicates the operation is not supported by the public key algorithm.
+
+=head1 EXAMPLE
+
+Recover digest originally signed using PKCS#1 and SHA256 digest:
+
+ #include <openssl/evp.h>
+ #include <openssl/rsa.h>
+
+ EVP_PKEY_CTX *ctx;
+ unsigned char *rout, *sig;
+ size_t routlen, siglen;
+ EVP_PKEY *verify_key;
+ /* NB: assumes verify_key, sig and siglen are already set up
+ * and that verify_key is an RSA public key
+ */
+ ctx = EVP_PKEY_CTX_new(verify_key);
+ if (!ctx)
+ /* Error occurred */
+ if (EVP_PKEY_verifyrecover_init(ctx) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_rsa_padding(ctx, RSA_PKCS1_PADDING) <= 0)
+ /* Error */
+ if (EVP_PKEY_CTX_set_signature_md(ctx, EVP_sha256()) <= 0)
+ /* Error */
+
+ /* Determine buffer length */
+ if (EVP_PKEY_verifyrecover(ctx, NULL, &routlen, sig, siglen) <= 0)
+ /* Error */
+
+ rout = OPENSSL_malloc(routlen);
+
+ if (!rout)
+ /* malloc failure */
+
+ if (EVP_PKEY_verifyrecover(ctx, rout, &routlen, sig, siglen) <= 0)
+ /* Error */
+
+ /* Recovered data is routlen bytes written to buffer rout */
+
+=head1 SEE ALSO
+
+L<EVP_PKEY_CTX_new(3)|EVP_PKEY_CTX_new(3)>,
+L<EVP_PKEY_encrypt(3)|EVP_PKEY_encrypt(3)>,
+L<EVP_PKEY_decrypt(3)|EVP_PKEY_decrypt(3)>,
+L<EVP_PKEY_sign(3)|EVP_PKEY_sign(3)>,
+L<EVP_PKEY_verify(3)|EVP_PKEY_verify(3)>,
+L<EVP_PKEY_derive(3)|EVP_PKEY_derive(3)>
+
+=head1 HISTORY
+
+These functions were first added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/EVP_SignInit.pod b/doc/crypto/EVP_SignInit.pod
index b6e62ce7f610..620a623ab620 100644
--- a/doc/crypto/EVP_SignInit.pod
+++ b/doc/crypto/EVP_SignInit.pod
@@ -77,6 +77,15 @@ will occur.
Older versions of this documentation wrongly stated that calls to
EVP_SignUpdate() could not be made after calling EVP_SignFinal().
+Since the private key is passed in the call to EVP_SignFinal() any error
+relating to the private key (for example an unsuitable key and digest
+combination) will not be indicated until after potentially large amounts of
+data have been passed through EVP_SignUpdate().
+
+It is not possible to change the signing parameters using these function.
+
+The previous two bugs are fixed in the newer EVP_SignDigest*() function.
+
=head1 SEE ALSO
L<EVP_VerifyInit(3)|EVP_VerifyInit(3)>,
diff --git a/doc/crypto/EVP_VerifyInit.pod b/doc/crypto/EVP_VerifyInit.pod
index b6afaedee5bf..9097f094105d 100644
--- a/doc/crypto/EVP_VerifyInit.pod
+++ b/doc/crypto/EVP_VerifyInit.pod
@@ -67,6 +67,15 @@ will occur.
Older versions of this documentation wrongly stated that calls to
EVP_VerifyUpdate() could not be made after calling EVP_VerifyFinal().
+Since the public key is passed in the call to EVP_SignFinal() any error
+relating to the private key (for example an unsuitable key and digest
+combination) will not be indicated until after potentially large amounts of
+data have been passed through EVP_SignUpdate().
+
+It is not possible to change the signing parameters using these function.
+
+The previous two bugs are fixed in the newer EVP_VerifyDigest*() function.
+
=head1 SEE ALSO
L<evp(3)|evp(3)>,
diff --git a/doc/crypto/PEM_write_bio_CMS_stream.pod b/doc/crypto/PEM_write_bio_CMS_stream.pod
new file mode 100644
index 000000000000..e070c45c2e95
--- /dev/null
+++ b/doc/crypto/PEM_write_bio_CMS_stream.pod
@@ -0,0 +1,41 @@
+=pod
+
+=head1 NAME
+
+ PEM_write_bio_CMS_stream - output CMS_ContentInfo structure in PEM format.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+ #include <openssl/pem.h>
+
+ int PEM_write_bio_CMS_stream(BIO *out, CMS_ContentInfo *cms, BIO *data, int flags);
+
+=head1 DESCRIPTION
+
+PEM_write_bio_CMS_stream() outputs a CMS_ContentInfo structure in PEM format.
+
+It is otherwise identical to the function SMIME_write_CMS().
+
+=head1 NOTES
+
+This function is effectively a version of the PEM_write_bio_CMS() supporting
+streaming.
+
+=head1 RETURN VALUES
+
+PEM_write_bio_CMS_stream() returns 1 for success or 0 for failure.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_verify(3)|CMS_verify(3)>, L<CMS_encrypt(3)|CMS_encrypt(3)>
+L<CMS_decrypt(3)|CMS_decrypt(3)>,
+L<SMIME_write_CMS(3)|SMIME_write_CMS(3)>,
+L<i2d_CMS_bio_stream(3)|i2d_CMS_bio_stream(3)>
+
+=head1 HISTORY
+
+PEM_write_bio_CMS_stream() was added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/PEM_write_bio_PKCS7_stream.pod b/doc/crypto/PEM_write_bio_PKCS7_stream.pod
new file mode 100644
index 000000000000..16fc9b684589
--- /dev/null
+++ b/doc/crypto/PEM_write_bio_PKCS7_stream.pod
@@ -0,0 +1,41 @@
+=pod
+
+=head1 NAME
+
+PEM_write_bio_PKCS7_stream - output PKCS7 structure in PEM format.
+
+=head1 SYNOPSIS
+
+ #include <openssl/pkcs7.h>
+ #include <openssl/pem.h>
+
+ int PEM_write_bio_PKCS7_stream(BIO *out, PKCS7 *p7, BIO *data, int flags);
+
+=head1 DESCRIPTION
+
+PEM_write_bio_PKCS7_stream() outputs a PKCS7 structure in PEM format.
+
+It is otherwise identical to the function SMIME_write_PKCS7().
+
+=head1 NOTES
+
+This function is effectively a version of the PEM_write_bio_PKCS7() supporting
+streaming.
+
+=head1 RETURN VALUES
+
+PEM_write_bio_PKCS7_stream() returns 1 for success or 0 for failure.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<PKCS7_sign(3)|PKCS7_sign(3)>,
+L<PKCS7_verify(3)|PKCS7_verify(3)>, L<PKCS7_encrypt(3)|PKCS7_encrypt(3)>
+L<PKCS7_decrypt(3)|PKCS7_decrypt(3)>,
+L<SMIME_write_PKCS7(3)|SMIME_write_PKCS7(3)>,
+L<i2d_PKCS7_bio_stream(3)|i2d_PKCS7_bio_stream(3)>
+
+=head1 HISTORY
+
+PEM_write_bio_PKCS7_stream() was added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/PKCS7_encrypt.pod b/doc/crypto/PKCS7_encrypt.pod
index 7e67f9576ab4..2cd925a7e0b6 100644
--- a/doc/crypto/PKCS7_encrypt.pod
+++ b/doc/crypto/PKCS7_encrypt.pod
@@ -18,43 +18,55 @@ B<cipher> is the symmetric cipher to use. B<flags> is an optional set of flags.
=head1 NOTES
-Only RSA keys are supported in PKCS#7 and envelopedData so the recipient certificates
-supplied to this function must all contain RSA public keys, though they do not have to
-be signed using the RSA algorithm.
+Only RSA keys are supported in PKCS#7 and envelopedData so the recipient
+certificates supplied to this function must all contain RSA public keys, though
+they do not have to be signed using the RSA algorithm.
-EVP_des_ede3_cbc() (triple DES) is the algorithm of choice for S/MIME use because
-most clients will support it.
+EVP_des_ede3_cbc() (triple DES) is the algorithm of choice for S/MIME use
+because most clients will support it.
-Some old "export grade" clients may only support weak encryption using 40 or 64 bit
-RC2. These can be used by passing EVP_rc2_40_cbc() and EVP_rc2_64_cbc() respectively.
+Some old "export grade" clients may only support weak encryption using 40 or 64
+bit RC2. These can be used by passing EVP_rc2_40_cbc() and EVP_rc2_64_cbc()
+respectively.
-The algorithm passed in the B<cipher> parameter must support ASN1 encoding of its
-parameters.
+The algorithm passed in the B<cipher> parameter must support ASN1 encoding of
+its parameters.
-Many browsers implement a "sign and encrypt" option which is simply an S/MIME
+Many browsers implement a "sign and encrypt" option which is simply an S/MIME
envelopedData containing an S/MIME signed message. This can be readily produced
by storing the S/MIME signed message in a memory BIO and passing it to
PKCS7_encrypt().
The following flags can be passed in the B<flags> parameter.
-If the B<PKCS7_TEXT> flag is set MIME headers for type B<text/plain> are prepended
-to the data.
+If the B<PKCS7_TEXT> flag is set MIME headers for type B<text/plain> are
+prepended to the data.
-Normally the supplied content is translated into MIME canonical format (as required
-by the S/MIME specifications) if B<PKCS7_BINARY> is set no translation occurs. This
-option should be used if the supplied data is in binary format otherwise the translation
-will corrupt it. If B<PKCS7_BINARY> is set then B<PKCS7_TEXT> is ignored.
+Normally the supplied content is translated into MIME canonical format (as
+required by the S/MIME specifications) if B<PKCS7_BINARY> is set no translation
+occurs. This option should be used if the supplied data is in binary format
+otherwise the translation will corrupt it. If B<PKCS7_BINARY> is set then
+B<PKCS7_TEXT> is ignored.
-=head1 RETURN VALUES
+If the B<PKCS7_STREAM> flag is set a partial B<PKCS7> structure is output
+suitable for streaming I/O: no data is read from the BIO B<in>.
-PKCS7_encrypt() returns either a valid PKCS7 structure or NULL if an error occurred.
-The error can be obtained from ERR_get_error(3).
+=head1 NOTES
+
+If the flag B<PKCS7_STREAM> is set the returned B<PKCS7> structure is B<not>
+complete and outputting its contents via a function that does not
+properly finalize the B<PKCS7> structure will give unpredictable
+results.
-=head1 BUGS
+Several functions including SMIME_write_PKCS7(), i2d_PKCS7_bio_stream(),
+PEM_write_bio_PKCS7_stream() finalize the structure. Alternatively finalization
+can be performed by obtaining the streaming ASN1 B<BIO> directly using
+BIO_new_PKCS7().
-The lack of single pass processing and need to hold all data in memory as
-mentioned in PKCS7_sign() also applies to PKCS7_verify().
+=head1 RETURN VALUES
+
+PKCS7_encrypt() returns either a PKCS7 structure or NULL if an error occurred.
+The error can be obtained from ERR_get_error(3).
=head1 SEE ALSO
@@ -63,5 +75,6 @@ L<ERR_get_error(3)|ERR_get_error(3)>, L<PKCS7_decrypt(3)|PKCS7_decrypt(3)>
=head1 HISTORY
PKCS7_decrypt() was added to OpenSSL 0.9.5
+The B<PKCS7_STREAM> flag was first supported in OpenSSL 1.0.0.
=cut
diff --git a/doc/crypto/PKCS7_sign.pod b/doc/crypto/PKCS7_sign.pod
index 21b1710370dd..64a35144f8c9 100644
--- a/doc/crypto/PKCS7_sign.pod
+++ b/doc/crypto/PKCS7_sign.pod
@@ -12,10 +12,10 @@ PKCS7_sign - create a PKCS#7 signedData structure
=head1 DESCRIPTION
-PKCS7_sign() creates and returns a PKCS#7 signedData structure. B<signcert>
-is the certificate to sign with, B<pkey> is the corresponsding private key.
-B<certs> is an optional additional set of certificates to include in the
-PKCS#7 structure (for example any intermediate CAs in the chain).
+PKCS7_sign() creates and returns a PKCS#7 signedData structure. B<signcert> is
+the certificate to sign with, B<pkey> is the corresponsding private key.
+B<certs> is an optional additional set of certificates to include in the PKCS#7
+structure (for example any intermediate CAs in the chain).
The data to be signed is read from BIO B<data>.
@@ -23,72 +23,83 @@ B<flags> is an optional set of flags.
=head1 NOTES
-Any of the following flags (ored together) can be passed in the B<flags> parameter.
+Any of the following flags (ored together) can be passed in the B<flags>
+parameter.
Many S/MIME clients expect the signed content to include valid MIME headers. If
the B<PKCS7_TEXT> flag is set MIME headers for type B<text/plain> are prepended
to the data.
If B<PKCS7_NOCERTS> is set the signer's certificate will not be included in the
-PKCS7 structure, the signer's certificate must still be supplied in the B<signcert>
-parameter though. This can reduce the size of the signature if the signers certificate
-can be obtained by other means: for example a previously signed message.
+PKCS7 structure, the signer's certificate must still be supplied in the
+B<signcert> parameter though. This can reduce the size of the signature if the
+signers certificate can be obtained by other means: for example a previously
+signed message.
+
+The data being signed is included in the PKCS7 structure, unless
+B<PKCS7_DETACHED> is set in which case it is omitted. This is used for PKCS7
+detached signatures which are used in S/MIME plaintext signed messages for
+example.
+
+Normally the supplied content is translated into MIME canonical format (as
+required by the S/MIME specifications) if B<PKCS7_BINARY> is set no translation
+occurs. This option should be used if the supplied data is in binary format
+otherwise the translation will corrupt it.
+
+The signedData structure includes several PKCS#7 autenticatedAttributes
+including the signing time, the PKCS#7 content type and the supported list of
+ciphers in an SMIMECapabilities attribute. If B<PKCS7_NOATTR> is set then no
+authenticatedAttributes will be used. If B<PKCS7_NOSMIMECAP> is set then just
+the SMIMECapabilities are omitted.
-The data being signed is included in the PKCS7 structure, unless B<PKCS7_DETACHED>
-is set in which case it is omitted. This is used for PKCS7 detached signatures
-which are used in S/MIME plaintext signed messages for example.
-
-Normally the supplied content is translated into MIME canonical format (as required
-by the S/MIME specifications) if B<PKCS7_BINARY> is set no translation occurs. This
-option should be used if the supplied data is in binary format otherwise the translation
-will corrupt it.
+If present the SMIMECapabilities attribute indicates support for the following
+algorithms: triple DES, 128 bit RC2, 64 bit RC2, DES and 40 bit RC2. If any of
+these algorithms is disabled then it will not be included.
-The signedData structure includes several PKCS#7 autenticatedAttributes including
-the signing time, the PKCS#7 content type and the supported list of ciphers in
-an SMIMECapabilities attribute. If B<PKCS7_NOATTR> is set then no authenticatedAttributes
-will be used. If B<PKCS7_NOSMIMECAP> is set then just the SMIMECapabilities are
-omitted.
+If the flags B<PKCS7_STREAM> is set then the returned B<PKCS7> structure is
+just initialized ready to perform the signing operation. The signing is however
+B<not> performed and the data to be signed is not read from the B<data>
+parameter. Signing is deferred until after the data has been written. In this
+way data can be signed in a single pass.
-If present the SMIMECapabilities attribute indicates support for the following
-algorithms: triple DES, 128 bit RC2, 64 bit RC2, DES and 40 bit RC2. If any
-of these algorithms is disabled then it will not be included.
+If the B<PKCS7_PARTIAL> flag is set a partial B<PKCS7> structure is output to
+which additional signers and capabilities can be added before finalization.
-If the flags B<PKCS7_PARTSIGN> is set then the returned B<PKCS7> structure
-is just initialized ready to perform the signing operation. The signing
-is however B<not> performed and the data to be signed is not read from
-the B<data> parameter. Signing is deferred until after the data has been
-written. In this way data can be signed in a single pass. Currently the
-flag B<PKCS7_DETACHED> B<must> also be set.
=head1 NOTES
-Currently the flag B<PKCS7_PARTSIGN> is only supported for detached
-data. If this flag is set the returned B<PKCS7> structure is B<not>
-complete and outputting its contents via a function that does not
-properly finalize the B<PKCS7> structure will give unpredictable
-results.
+If the flag B<PKCS7_STREAM> is set the returned B<PKCS7> structure is B<not>
+complete and outputting its contents via a function that does not properly
+finalize the B<PKCS7> structure will give unpredictable results.
-At present only the SMIME_write_PKCS7() function properly finalizes the
-structure.
+Several functions including SMIME_write_PKCS7(), i2d_PKCS7_bio_stream(),
+PEM_write_bio_PKCS7_stream() finalize the structure. Alternatively finalization
+can be performed by obtaining the streaming ASN1 B<BIO> directly using
+BIO_new_PKCS7().
-=head1 BUGS
+If a signer is specified it will use the default digest for the signing
+algorithm. This is B<SHA1> for both RSA and DSA keys.
-PKCS7_sign() is somewhat limited. It does not support multiple signers, some
-advanced attributes such as counter signatures are not supported.
+In OpenSSL 1.0.0 the B<certs>, B<signcert> and B<pkey> parameters can all be
+B<NULL> if the B<PKCS7_PARTIAL> flag is set. One or more signers can be added
+using the function B<PKCS7_sign_add_signer()>. B<PKCS7_final()> must also be
+called to finalize the structure if streaming is not enabled. Alternative
+signing digests can also be specified using this method.
-The SHA1 digest algorithm is currently always used.
+In OpenSSL 1.0.0 if B<signcert> and B<pkey> are NULL then a certificates only
+PKCS#7 structure is output.
-When the signed data is not detached it will be stored in memory within the
-B<PKCS7> structure. This effectively limits the size of messages which can be
-signed due to memory restraints. There should be a way to sign data without
-having to hold it all in memory, this would however require fairly major
-revisions of the OpenSSL ASN1 code.
+In versions of OpenSSL before 1.0.0 the B<signcert> and B<pkey> parameters must
+B<NOT> be NULL.
+=head1 BUGS
+
+Some advanced attributes such as counter signatures are not supported.
=head1 RETURN VALUES
-PKCS7_sign() returns either a valid PKCS7 structure or NULL if an error occurred.
-The error can be obtained from ERR_get_error(3).
+PKCS7_sign() returns either a valid PKCS7 structure or NULL if an error
+occurred. The error can be obtained from ERR_get_error(3).
=head1 SEE ALSO
@@ -98,6 +109,8 @@ L<ERR_get_error(3)|ERR_get_error(3)>, L<PKCS7_verify(3)|PKCS7_verify(3)>
PKCS7_sign() was added to OpenSSL 0.9.5
-The B<PKCS7_PARTSIGN> flag was added in OpenSSL 0.9.8
+The B<PKCS7_PARTIAL> flag was added in OpenSSL 1.0.0
+
+The B<PKCS7_STREAM> flag was added in OpenSSL 1.0.0
=cut
diff --git a/doc/crypto/PKCS7_sign_add_signer.pod b/doc/crypto/PKCS7_sign_add_signer.pod
new file mode 100644
index 000000000000..ebec4d57deaf
--- /dev/null
+++ b/doc/crypto/PKCS7_sign_add_signer.pod
@@ -0,0 +1,87 @@
+=pod
+
+=head1 NAME
+
+PKCS7_sign_add_signer - add a signer PKCS7 signed data structure.
+
+=head1 SYNOPSIS
+
+ #include <openssl/pkcs7.h>
+
+ PKCS7_SIGNER_INFO *PKCS7_sign_add_signer(PKCS7 *p7, X509 *signcert, EVP_PKEY *pkey, const EVP_MD *md, int flags);
+
+
+=head1 DESCRIPTION
+
+PKCS7_sign_add_signer() adds a signer with certificate B<signcert> and private
+key B<pkey> using message digest B<md> to a PKCS7 signed data structure
+B<p7>.
+
+The PKCS7 structure should be obtained from an initial call to PKCS7_sign()
+with the flag B<PKCS7_PARTIAL> set or in the case or re-signing a valid PKCS7
+signed data structure.
+
+If the B<md> parameter is B<NULL> then the default digest for the public
+key algorithm will be used.
+
+Unless the B<PKCS7_REUSE_DIGEST> flag is set the returned PKCS7 structure
+is not complete and must be finalized either by streaming (if applicable) or
+a call to PKCS7_final().
+
+
+=head1 NOTES
+
+The main purpose of this function is to provide finer control over a PKCS#7
+signed data structure where the simpler PKCS7_sign() function defaults are
+not appropriate. For example if multiple signers or non default digest
+algorithms are needed.
+
+Any of the following flags (ored together) can be passed in the B<flags>
+parameter.
+
+If B<PKCS7_REUSE_DIGEST> is set then an attempt is made to copy the content
+digest value from the PKCS7 struture: to add a signer to an existing structure.
+An error occurs if a matching digest value cannot be found to copy. The
+returned PKCS7 structure will be valid and finalized when this flag is set.
+
+If B<PKCS7_PARTIAL> is set in addition to B<PKCS7_REUSE_DIGEST> then the
+B<PKCS7_SIGNER_INO> structure will not be finalized so additional attributes
+can be added. In this case an explicit call to PKCS7_SIGNER_INFO_sign() is
+needed to finalize it.
+
+If B<PKCS7_NOCERTS> is set the signer's certificate will not be included in the
+PKCS7 structure, the signer's certificate must still be supplied in the
+B<signcert> parameter though. This can reduce the size of the signature if the
+signers certificate can be obtained by other means: for example a previously
+signed message.
+
+The signedData structure includes several PKCS#7 autenticatedAttributes
+including the signing time, the PKCS#7 content type and the supported list of
+ciphers in an SMIMECapabilities attribute. If B<PKCS7_NOATTR> is set then no
+authenticatedAttributes will be used. If B<PKCS7_NOSMIMECAP> is set then just
+the SMIMECapabilities are omitted.
+
+If present the SMIMECapabilities attribute indicates support for the following
+algorithms: triple DES, 128 bit RC2, 64 bit RC2, DES and 40 bit RC2. If any of
+these algorithms is disabled then it will not be included.
+
+
+PKCS7_sign_add_signers() returns an internal pointer to the PKCS7_SIGNER_INFO
+structure just added, this can be used to set additional attributes
+before it is finalized.
+
+=head1 RETURN VALUES
+
+PKCS7_sign_add_signers() returns an internal pointer to the PKCS7_SIGNER_INFO
+structure just added or NULL if an error occurs.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<PKCS7_sign(3)|PKCS7_sign(3)>,
+L<PKCS7_final(3)|PKCS7_final(3)>,
+
+=head1 HISTORY
+
+PPKCS7_sign_add_signer() was added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/SMIME_read_CMS.pod b/doc/crypto/SMIME_read_CMS.pod
new file mode 100644
index 000000000000..acc5524c1400
--- /dev/null
+++ b/doc/crypto/SMIME_read_CMS.pod
@@ -0,0 +1,70 @@
+=pod
+
+=head1 NAME
+
+ SMIME_read_CMS - parse S/MIME message.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ CMS_ContentInfo *SMIME_read_CMS(BIO *in, BIO **bcont);
+
+=head1 DESCRIPTION
+
+SMIME_read_CMS() parses a message in S/MIME format.
+
+B<in> is a BIO to read the message from.
+
+If cleartext signing is used then the content is saved in a memory bio which is
+written to B<*bcont>, otherwise B<*bcont> is set to NULL.
+
+The parsed CMS_ContentInfo structure is returned or NULL if an
+error occurred.
+
+=head1 NOTES
+
+If B<*bcont> is not NULL then the message is clear text signed. B<*bcont> can
+then be passed to CMS_verify() with the B<CMS_DETACHED> flag set.
+
+Otherwise the type of the returned structure can be determined
+using CMS_get0_type().
+
+To support future functionality if B<bcont> is not NULL B<*bcont> should be
+initialized to NULL. For example:
+
+ BIO *cont = NULL;
+ CMS_ContentInfo *cms;
+
+ cms = SMIME_read_CMS(in, &cont);
+
+=head1 BUGS
+
+The MIME parser used by SMIME_read_CMS() is somewhat primitive. While it will
+handle most S/MIME messages more complex compound formats may not work.
+
+The parser assumes that the CMS_ContentInfo structure is always base64 encoded
+and will not handle the case where it is in binary format or uses quoted
+printable format.
+
+The use of a memory BIO to hold the signed content limits the size of message
+which can be processed due to memory restraints: a streaming single pass option
+should be available.
+
+=head1 RETURN VALUES
+
+SMIME_read_CMS() returns a valid B<CMS_ContentInfo> structure or B<NULL>
+if an error occurred. The error can be obtained from ERR_get_error(3).
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_type(3)|CMS_type(3)>
+L<SMIME_read_CMS(3)|SMIME_read_CMS(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_verify(3)|CMS_verify(3)>, L<CMS_encrypt(3)|CMS_encrypt(3)>
+L<CMS_decrypt(3)|CMS_decrypt(3)>
+
+=head1 HISTORY
+
+SMIME_read_CMS() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/SMIME_write_CMS.pod b/doc/crypto/SMIME_write_CMS.pod
new file mode 100644
index 000000000000..04bedfb42977
--- /dev/null
+++ b/doc/crypto/SMIME_write_CMS.pod
@@ -0,0 +1,64 @@
+=pod
+
+=head1 NAME
+
+ SMIME_write_CMS - convert CMS structure to S/MIME format.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int SMIME_write_CMS(BIO *out, CMS_ContentInfo *cms, BIO *data, int flags);
+
+=head1 DESCRIPTION
+
+SMIME_write_CMS() adds the appropriate MIME headers to a CMS
+structure to produce an S/MIME message.
+
+B<out> is the BIO to write the data to. B<cms> is the appropriate
+B<CMS_ContentInfo> structure. If streaming is enabled then the content must be
+supplied in the B<data> argument. B<flags> is an optional set of flags.
+
+=head1 NOTES
+
+The following flags can be passed in the B<flags> parameter.
+
+If B<CMS_DETACHED> is set then cleartext signing will be used, this option only
+makes sense for SignedData where B<CMS_DETACHED> is also set when CMS_sign() is
+called.
+
+If the B<CMS_TEXT> flag is set MIME headers for type B<text/plain> are added to
+the content, this only makes sense if B<CMS_DETACHED> is also set.
+
+If the B<CMS_STREAM> flag is set streaming is performed. This flag should only
+be set if B<CMS_STREAM> was also set in the previous call to a CMS_ContentInfo
+creation function.
+
+If cleartext signing is being used and B<CMS_STREAM> not set then the data must
+be read twice: once to compute the signature in CMS_sign() and once to output
+the S/MIME message.
+
+If streaming is performed the content is output in BER format using indefinite
+length constructed encoding except in the case of signed data with detached
+content where the content is absent and DER format is used.
+
+=head1 BUGS
+
+SMIME_write_CMS() always base64 encodes CMS structures, there should be an
+option to disable this.
+
+=head1 RETURN VALUES
+
+SMIME_write_CMS() returns 1 for success or 0 for failure.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_verify(3)|CMS_verify(3)>, L<CMS_encrypt(3)|CMS_encrypt(3)>
+L<CMS_decrypt(3)|CMS_decrypt(3)>
+
+=head1 HISTORY
+
+SMIME_write_CMS() was added to OpenSSL 0.9.8
+
+=cut
diff --git a/doc/crypto/SMIME_write_PKCS7.pod b/doc/crypto/SMIME_write_PKCS7.pod
index e20d46b5d1bb..ca6bd027635f 100644
--- a/doc/crypto/SMIME_write_PKCS7.pod
+++ b/doc/crypto/SMIME_write_PKCS7.pod
@@ -15,10 +15,9 @@ SMIME_write_PKCS7 - convert PKCS#7 structure to S/MIME format.
SMIME_write_PKCS7() adds the appropriate MIME headers to a PKCS#7
structure to produce an S/MIME message.
-B<out> is the BIO to write the data to. B<p7> is the appropriate
-B<PKCS7> structure. If cleartext signing (B<multipart/signed>) is
-being used then the signed data must be supplied in the B<data>
-argument. B<flags> is an optional set of flags.
+B<out> is the BIO to write the data to. B<p7> is the appropriate B<PKCS7>
+structure. If streaming is enabled then the content must be supplied in the
+B<data> argument. B<flags> is an optional set of flags.
=head1 NOTES
@@ -32,15 +31,18 @@ If the B<PKCS7_TEXT> flag is set MIME headers for type B<text/plain>
are added to the content, this only makes sense if B<PKCS7_DETACHED>
is also set.
-If the B<PKCS7_PARTSIGN> flag is set the signed data is finalized
-and output along with the content. This flag should only be set
-if B<PKCS7_DETACHED> is also set and the previous call to PKCS7_sign()
-also set these flags.
+If the B<PKCS7_STREAM> flag is set streaming is performed. This flag should
+only be set if B<PKCS7_STREAM> was also set in the previous call to
+PKCS7_sign() or B<PKCS7_encrypt()>.
-If cleartext signing is being used and B<PKCS7_PARTSIGN> not set then
+If cleartext signing is being used and B<PKCS7_STREAM> not set then
the data must be read twice: once to compute the signature in PKCS7_sign()
and once to output the S/MIME message.
+If streaming is performed the content is output in BER format using indefinite
+length constructuted encoding except in the case of signed data with detached
+content where the content is absent and DER format is used.
+
=head1 BUGS
SMIME_write_PKCS7() always base64 encodes PKCS#7 structures, there
diff --git a/doc/crypto/X509_STORE_CTX_get_error.pod b/doc/crypto/X509_STORE_CTX_get_error.pod
new file mode 100644
index 000000000000..a883f6c09783
--- /dev/null
+++ b/doc/crypto/X509_STORE_CTX_get_error.pod
@@ -0,0 +1,303 @@
+=pod
+
+=head1 NAME
+
+X509_STORE_CTX_get_error, X509_STORE_CTX_set_error, X509_STORE_CTX_get_error_depth, X509_STORE_CTX_get_current_cert, X509_STORE_CTX_get1_chain, X509_verify_cert_error_string - get or set certificate verification status information
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509.h>
+ #include <openssl/x509_vfy.h>
+
+ int X509_STORE_CTX_get_error(X509_STORE_CTX *ctx);
+ void X509_STORE_CTX_set_error(X509_STORE_CTX *ctx,int s);
+ int X509_STORE_CTX_get_error_depth(X509_STORE_CTX *ctx);
+ X509 * X509_STORE_CTX_get_current_cert(X509_STORE_CTX *ctx);
+
+ STACK_OF(X509) *X509_STORE_CTX_get1_chain(X509_STORE_CTX *ctx);
+
+ const char *X509_verify_cert_error_string(long n);
+
+=head1 DESCRIPTION
+
+These functions are typically called after X509_verify_cert() has indicated
+an error or in a verification callback to determine the nature of an error.
+
+X509_STORE_CTX_get_error() returns the error code of B<ctx>, see
+the B<ERROR CODES> section for a full description of all error codes.
+
+X509_STORE_CTX_set_error() sets the error code of B<ctx> to B<s>. For example
+it might be used in a verification callback to set an error based on additional
+checks.
+
+X509_STORE_CTX_get_error_depth() returns the B<depth> of the error. This is a
+non-negative integer representing where in the certificate chain the error
+occurred. If it is zero it occured in the end entity certificate, one if
+it is the certificate which signed the end entity certificate and so on.
+
+X509_STORE_CTX_get_current_cert() returns the certificate in B<ctx> which
+caused the error or B<NULL> if no certificate is relevant.
+
+X509_STORE_CTX_get1_chain() returns a complete validate chain if a previous
+call to X509_verify_cert() is successful. If the call to X509_verify_cert()
+is B<not> successful the returned chain may be incomplete or invalid. The
+returned chain persists after the B<ctx> structure is freed, when it is
+no longer needed it should be free up using:
+
+ sk_X509_pop_free(chain, X509_free);
+
+X509_verify_cert_error_string() returns a human readable error string for
+verification error B<n>.
+
+=head1 RETURN VALUES
+
+X509_STORE_CTX_get_error() returns B<X509_V_OK> or an error code.
+
+X509_STORE_CTX_get_error_depth() returns a non-negative error depth.
+
+X509_STORE_CTX_get_current_cert() returns the cerificate which caused the
+error or B<NULL> if no certificate is relevant to the error.
+
+X509_verify_cert_error_string() returns a human readable error string for
+verification error B<n>.
+
+=head1 ERROR CODES
+
+A list of error codes and messages is shown below. Some of the
+error codes are defined but currently never returned: these are described as
+"unused".
+
+=over 4
+
+=item B<X509_V_OK: ok>
+
+the operation was successful.
+
+=item B<X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT: unable to get issuer certificate>
+
+the issuer certificate could not be found: this occurs if the issuer certificate
+of an untrusted certificate cannot be found.
+
+=item B<X509_V_ERR_UNABLE_TO_GET_CRL: unable to get certificate CRL>
+
+the CRL of a certificate could not be found.
+
+=item B<X509_V_ERR_UNABLE_TO_DECRYPT_CERT_SIGNATURE: unable to decrypt certificate's signature>
+
+the certificate signature could not be decrypted. This means that the actual
+signature value could not be determined rather than it not matching the
+expected value, this is only meaningful for RSA keys.
+
+=item B<X509_V_ERR_UNABLE_TO_DECRYPT_CRL_SIGNATURE: unable to decrypt CRL's signature>
+
+the CRL signature could not be decrypted: this means that the actual signature
+value could not be determined rather than it not matching the expected value.
+Unused.
+
+=item B<X509_V_ERR_UNABLE_TO_DECODE_ISSUER_PUBLIC_KEY: unable to decode issuer public key>
+
+the public key in the certificate SubjectPublicKeyInfo could not be read.
+
+=item B<X509_V_ERR_CERT_SIGNATURE_FAILURE: certificate signature failure>
+
+the signature of the certificate is invalid.
+
+=item B<X509_V_ERR_CRL_SIGNATURE_FAILURE: CRL signature failure>
+
+the signature of the certificate is invalid.
+
+=item B<X509_V_ERR_CERT_NOT_YET_VALID: certificate is not yet valid>
+
+the certificate is not yet valid: the notBefore date is after the current time.
+
+=item B<X509_V_ERR_CERT_HAS_EXPIRED: certificate has expired>
+
+the certificate has expired: that is the notAfter date is before the current time.
+
+=item B<X509_V_ERR_CRL_NOT_YET_VALID: CRL is not yet valid>
+
+the CRL is not yet valid.
+
+=item B<X509_V_ERR_CRL_HAS_EXPIRED: CRL has expired>
+
+the CRL has expired.
+
+=item B<X509_V_ERR_ERROR_IN_CERT_NOT_BEFORE_FIELD: format error in certificate's notBefore field>
+
+the certificate notBefore field contains an invalid time.
+
+=item B<X509_V_ERR_ERROR_IN_CERT_NOT_AFTER_FIELD: format error in certificate's notAfter field>
+
+the certificate notAfter field contains an invalid time.
+
+=item B<X509_V_ERR_ERROR_IN_CRL_LAST_UPDATE_FIELD: format error in CRL's lastUpdate field>
+
+the CRL lastUpdate field contains an invalid time.
+
+=item B<X509_V_ERR_ERROR_IN_CRL_NEXT_UPDATE_FIELD: format error in CRL's nextUpdate field>
+
+the CRL nextUpdate field contains an invalid time.
+
+=item B<X509_V_ERR_OUT_OF_MEM: out of memory>
+
+an error occurred trying to allocate memory. This should never happen.
+
+=item B<X509_V_ERR_DEPTH_ZERO_SELF_SIGNED_CERT: self signed certificate>
+
+the passed certificate is self signed and the same certificate cannot be found
+in the list of trusted certificates.
+
+=item B<X509_V_ERR_SELF_SIGNED_CERT_IN_CHAIN: self signed certificate in certificate chain>
+
+the certificate chain could be built up using the untrusted certificates but
+the root could not be found locally.
+
+=item B<X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT_LOCALLY: unable to get local issuer certificate>
+
+the issuer certificate of a locally looked up certificate could not be found.
+This normally means the list of trusted certificates is not complete.
+
+=item B<X509_V_ERR_UNABLE_TO_VERIFY_LEAF_SIGNATURE: unable to verify the first certificate>
+
+no signatures could be verified because the chain contains only one certificate
+and it is not self signed.
+
+=item B<X509_V_ERR_CERT_CHAIN_TOO_LONG: certificate chain too long>
+
+the certificate chain length is greater than the supplied maximum depth. Unused.
+
+=item B<X509_V_ERR_CERT_REVOKED: certificate revoked>
+
+the certificate has been revoked.
+
+=item B<X509_V_ERR_INVALID_CA: invalid CA certificate>
+
+a CA certificate is invalid. Either it is not a CA or its extensions are not
+consistent with the supplied purpose.
+
+=item B<X509_V_ERR_PATH_LENGTH_EXCEEDED: path length constraint exceeded>
+
+the basicConstraints pathlength parameter has been exceeded.
+
+=item B<X509_V_ERR_INVALID_PURPOSE: unsupported certificate purpose>
+
+the supplied certificate cannot be used for the specified purpose.
+
+=item B<X509_V_ERR_CERT_UNTRUSTED: certificate not trusted>
+
+the root CA is not marked as trusted for the specified purpose.
+
+=item B<X509_V_ERR_CERT_REJECTED: certificate rejected>
+
+the root CA is marked to reject the specified purpose.
+
+=item B<X509_V_ERR_SUBJECT_ISSUER_MISMATCH: subject issuer mismatch>
+
+the current candidate issuer certificate was rejected because its subject name
+did not match the issuer name of the current certificate. This is only set
+if issuer check debugging is enabled it is used for status notification and
+is B<not> in itself an error.
+
+=item B<X509_V_ERR_AKID_SKID_MISMATCH: authority and subject key identifier mismatch>
+
+the current candidate issuer certificate was rejected because its subject key
+identifier was present and did not match the authority key identifier current
+certificate. This is only set if issuer check debugging is enabled it is used
+for status notification and is B<not> in itself an error.
+
+=item B<X509_V_ERR_AKID_ISSUER_SERIAL_MISMATCH: authority and issuer serial number mismatch>
+
+the current candidate issuer certificate was rejected because its issuer name
+and serial number was present and did not match the authority key identifier of
+the current certificate. This is only set if issuer check debugging is enabled
+it is used for status notification and is B<not> in itself an error.
+
+=item B<X509_V_ERR_KEYUSAGE_NO_CERTSIGN:key usage does not include certificate signing>
+
+the current candidate issuer certificate was rejected because its keyUsage
+extension does not permit certificate signing. This is only set if issuer check
+debugging is enabled it is used for status notification and is B<not> in itself
+an error.
+
+=item B<X509_V_ERR_INVALID_EXTENSION: invalid or inconsistent certificate extension>
+
+A certificate extension had an invalid value (for example an incorrect
+encoding) or some value inconsistent with other extensions.
+
+
+=item B<X509_V_ERR_INVALID_POLICY_EXTENSION: invalid or inconsistent certificate policy extension>
+
+A certificate policies extension had an invalid value (for example an incorrect
+encoding) or some value inconsistent with other extensions. This error only
+occurs if policy processing is enabled.
+
+=item B<X509_V_ERR_NO_EXPLICIT_POLICY: no explicit policy>
+
+The verification flags were set to require and explicit policy but none was
+present.
+
+=item B<X509_V_ERR_DIFFERENT_CRL_SCOPE: Different CRL scope>
+
+The only CRLs that could be found did not match the scope of the certificate.
+
+=item B<X509_V_ERR_UNSUPPORTED_EXTENSION_FEATURE: Unsupported extension feature>
+
+Some feature of a certificate extension is not supported. Unused.
+
+=item B<X509_V_ERR_PERMITTED_VIOLATION: permitted subtree violation>
+
+A name constraint violation occured in the permitted subtrees.
+
+=item B<X509_V_ERR_EXCLUDED_VIOLATION: excluded subtree violation>
+
+A name constraint violation occured in the excluded subtrees.
+
+=item B<X509_V_ERR_SUBTREE_MINMAX: name constraints minimum and maximum not supported>
+
+A certificate name constraints extension included a minimum or maximum field:
+this is not supported.
+
+=item B<X509_V_ERR_UNSUPPORTED_CONSTRAINT_TYPE: unsupported name constraint type>
+
+An unsupported name constraint type was encountered. OpenSSL currently only
+supports directory name, DNS name, email and URI types.
+
+=item B<X509_V_ERR_UNSUPPORTED_CONSTRAINT_SYNTAX: unsupported or invalid name constraint syntax>
+
+The format of the name constraint is not recognised: for example an email
+address format of a form not mentioned in RFC3280. This could be caused by
+a garbage extension or some new feature not currently supported.
+
+=item B<X509_V_ERR_CRL_PATH_VALIDATION_ERROR: CRL path validation error>
+
+An error occured when attempting to verify the CRL path. This error can only
+happen if extended CRL checking is enabled.
+
+=item B<X509_V_ERR_APPLICATION_VERIFICATION: application verification failure>
+
+an application specific error. This will never be returned unless explicitly
+set by an application.
+
+=head1 NOTES
+
+The above functions should be used instead of directly referencing the fields
+in the B<X509_VERIFY_CTX> structure.
+
+In versions of OpenSSL before 1.0 the current certificate returned by
+X509_STORE_CTX_get_current_cert() was never B<NULL>. Applications should
+check the return value before printing out any debugging information relating
+to the current certificate.
+
+If an unrecognised error code is passed to X509_verify_cert_error_string() the
+numerical value of the unknown code is returned in a static buffer. This is not
+thread safe but will never happen unless an invalid code is passed.
+
+=head1 SEE ALSO
+
+L<X509_verify_cert(3)|X509_verify_cert(3)>
+
+=head1 HISTORY
+
+TBA
+
+=cut
diff --git a/doc/crypto/X509_STORE_CTX_get_ex_new_index.pod b/doc/crypto/X509_STORE_CTX_get_ex_new_index.pod
new file mode 100644
index 000000000000..8d6b9dda47e7
--- /dev/null
+++ b/doc/crypto/X509_STORE_CTX_get_ex_new_index.pod
@@ -0,0 +1,41 @@
+=pod
+
+=head1 NAME
+
+X509_STORE_CTX_get_ex_new_index, X509_STORE_CTX_set_ex_data, X509_STORE_CTX_get_ex_data - add application specific data to X509_STORE_CTX structures
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509_vfy.h>
+
+ int X509_STORE_CTX_get_ex_new_index(long argl, void *argp,
+ CRYPTO_EX_new *new_func,
+ CRYPTO_EX_dup *dup_func,
+ CRYPTO_EX_free *free_func);
+
+ int X509_STORE_CTX_set_ex_data(X509_STORE_CTX *d, int idx, void *arg);
+
+ char *X509_STORE_CTX_get_ex_data(X509_STORE_CTX *d, int idx);
+
+=head1 DESCRIPTION
+
+These functions handle application specific data in X509_STORE_CTX structures.
+Their usage is identical to that of RSA_get_ex_new_index(), RSA_set_ex_data()
+and RSA_get_ex_data() as described in L<RSA_get_ex_new_index(3)>.
+
+=head1 NOTES
+
+This mechanism is used internally by the B<ssl> library to store the B<SSL>
+structure associated with a verification operation in an B<X509_STORE_CTX>
+structure.
+
+=head1 SEE ALSO
+
+L<RSA_get_ex_new_index(3)|RSA_get_ex_new_index(3)>
+
+=head1 HISTORY
+
+X509_STORE_CTX_get_ex_new_index(), X509_STORE_CTX_set_ex_data() and
+X509_STORE_CTX_get_ex_data() are available since OpenSSL 0.9.5.
+
+=cut
diff --git a/doc/crypto/X509_STORE_CTX_new.pod b/doc/crypto/X509_STORE_CTX_new.pod
new file mode 100644
index 000000000000..b17888f149e9
--- /dev/null
+++ b/doc/crypto/X509_STORE_CTX_new.pod
@@ -0,0 +1,122 @@
+=pod
+
+=head1 NAME
+
+X509_STORE_CTX_new, X509_STORE_CTX_cleanup, X509_STORE_CTX_free, X509_STORE_CTX_init, X509_STORE_CTX_trusted_stack, X509_STORE_CTX_set_cert, X509_STORE_CTX_set_chain, X509_STORE_CTX_set0_crls, X509_STORE_CTX_get0_param, X509_STORE_CTX_set0_param, X509_STORE_CTX_set_default - X509_STORE_CTX initialisation
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509_vfy.h>
+
+ X509_STORE_CTX *X509_STORE_CTX_new(void);
+ void X509_STORE_CTX_cleanup(X509_STORE_CTX *ctx);
+ void X509_STORE_CTX_free(X509_STORE_CTX *ctx);
+
+ int X509_STORE_CTX_init(X509_STORE_CTX *ctx, X509_STORE *store,
+ X509 *x509, STACK_OF(X509) *chain);
+
+ void X509_STORE_CTX_trusted_stack(X509_STORE_CTX *ctx, STACK_OF(X509) *sk);
+
+ void X509_STORE_CTX_set_cert(X509_STORE_CTX *ctx,X509 *x);
+ void X509_STORE_CTX_set_chain(X509_STORE_CTX *ctx,STACK_OF(X509) *sk);
+ void X509_STORE_CTX_set0_crls(X509_STORE_CTX *ctx, STACK_OF(X509_CRL) *sk);
+
+ X509_VERIFY_PARAM *X509_STORE_CTX_get0_param(X509_STORE_CTX *ctx);
+ void X509_STORE_CTX_set0_param(X509_STORE_CTX *ctx, X509_VERIFY_PARAM *param);
+ int X509_STORE_CTX_set_default(X509_STORE_CTX *ctx, const char *name);
+
+=head1 DESCRIPTION
+
+These functions initialise an B<X509_STORE_CTX> structure for subsequent use
+by X509_verify_cert().
+
+X509_STORE_CTX_new() returns a newly initialised B<X509_STORE_CTX> structure.
+
+X509_STORE_CTX_cleanup() internally cleans up an B<X509_STORE_CTX> structure.
+The context can then be reused with an new call to X509_STORE_CTX_init().
+
+X509_STORE_CTX_free() completely frees up B<ctx>. After this call B<ctx>
+is no longer valid.
+
+X509_STORE_CTX_init() sets up B<ctx> for a subsequent verification operation.
+The trusted certificate store is set to B<store>, the end entity certificate
+to be verified is set to B<x509> and a set of additional certificates (which
+will be untrusted but may be used to build the chain) in B<chain>. Any or
+all of the B<store>, B<x509> and B<chain> parameters can be B<NULL>.
+
+X509_STORE_CTX_trusted_stack() sets the set of trusted certificates of B<ctx>
+to B<sk>. This is an alternative way of specifying trusted certificates
+instead of using an B<X509_STORE>.
+
+X509_STORE_CTX_set_cert() sets the certificate to be vertified in B<ctx> to
+B<x>.
+
+X509_STORE_CTX_set_chain() sets the additional certificate chain used by B<ctx>
+to B<sk>.
+
+X509_STORE_CTX_set0_crls() sets a set of CRLs to use to aid certificate
+verification to B<sk>. These CRLs will only be used if CRL verification is
+enabled in the associated B<X509_VERIFY_PARAM> structure. This might be
+used where additional "useful" CRLs are supplied as part of a protocol,
+for example in a PKCS#7 structure.
+
+X509_VERIFY_PARAM *X509_STORE_CTX_get0_param() retrieves an intenal pointer
+to the verification parameters associated with B<ctx>.
+
+X509_STORE_CTX_set0_param() sets the intenal verification parameter pointer
+to B<param>. After this call B<param> should not be used.
+
+X509_STORE_CTX_set_default() looks up and sets the default verification
+method to B<name>. This uses the function X509_VERIFY_PARAM_lookup() to
+find an appropriate set of parameters from B<name>.
+
+=head1 NOTES
+
+The certificates and CRLs in a store are used internally and should B<not>
+be freed up until after the associated B<X509_STORE_CTX> is freed. Legacy
+applications might implicitly use an B<X509_STORE_CTX> like this:
+
+ X509_STORE_CTX ctx;
+ X509_STORE_CTX_init(&ctx, store, cert, chain);
+
+this is B<not> recommended in new applications they should instead do:
+
+ X509_STORE_CTX *ctx;
+ ctx = X509_STORE_CTX_new();
+ if (ctx == NULL)
+ /* Bad error */
+ X509_STORE_CTX_init(ctx, store, cert, chain);
+
+=head1 BUGS
+
+The certificates and CRLs in a context are used internally and should B<not>
+be freed up until after the associated B<X509_STORE_CTX> is freed. Copies
+should be made or reference counts increased instead.
+
+=head1 RETURN VALUES
+
+X509_STORE_CTX_new() returns an newly allocates context or B<NULL> is an
+error occurred.
+
+X509_STORE_CTX_init() returns 1 for success or 0 if an error occurred.
+
+X509_STORE_CTX_get0_param() returns a pointer to an B<X509_VERIFY_PARAM>
+structure or B<NULL> if an error occurred.
+
+X509_STORE_CTX_cleanup(), X509_STORE_CTX_free(), X509_STORE_CTX_trusted_stack(),
+X509_STORE_CTX_set_cert(), X509_STORE_CTX_set_chain(),
+X509_STORE_CTX_set0_crls() and X509_STORE_CTX_set0_param() do not return
+values.
+
+X509_STORE_CTX_set_default() returns 1 for success or 0 if an error occurred.
+
+=head1 SEE ALSO
+
+L<X509_verify_cert(3)|X509_verify_cert(3)>
+L<X509_VERIFY_PARAM_set_flags(3)|X509_VERIFY_PARAM_set_flags(3)>
+
+=head1 HISTORY
+
+X509_STORE_CTX_set0_crls() was first added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/X509_STORE_CTX_set_verify_cb.pod b/doc/crypto/X509_STORE_CTX_set_verify_cb.pod
new file mode 100644
index 000000000000..b9787a6ca6f2
--- /dev/null
+++ b/doc/crypto/X509_STORE_CTX_set_verify_cb.pod
@@ -0,0 +1,161 @@
+=pod
+
+=head1 NAME
+
+X509_STORE_CTX_set_verify_cb - set verification callback
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509_vfy.h>
+
+ void X509_STORE_CTX_set_verify_cb(X509_STORE_CTX *ctx,
+ int (*verify_cb)(int ok, X509_STORE_CTX *ctx));
+
+=head1 DESCRIPTION
+
+X509_STORE_CTX_set_verify_cb() sets the verification callback of B<ctx> to
+B<verify_cb> overwriting any existing callback.
+
+The verification callback can be used to customise the operation of certificate
+verification, either by overriding error conditions or logging errors for
+debugging purposes.
+
+However a verification callback is B<not> essential and the default operation
+is often sufficient.
+
+The B<ok> parameter to the callback indicates the value the callback should
+return to retain the default behaviour. If it is zero then and error condition
+is indicated. If it is 1 then no error occurred. If the flag
+B<X509_V_FLAG_NOTIFY_POLICY> is set then B<ok> is set to 2 to indicate the
+policy checking is complete.
+
+The B<ctx> parameter to the callback is the B<X509_STORE_CTX> structure that
+is performing the verification operation. A callback can examine this
+structure and receive additional information about the error, for example
+by calling X509_STORE_CTX_get_current_cert(). Additional application data can
+be passed to the callback via the B<ex_data> mechanism.
+
+=head1 WARNING
+
+In general a verification callback should B<NOT> unconditionally return 1 in
+all circumstances because this will allow verification to succeed no matter
+what the error. This effectively removes all security from the application
+because B<any> certificate (including untrusted generated ones) will be
+accepted.
+
+=head1 NOTES
+
+The verification callback can be set and inherited from the parent structure
+performing the operation. In some cases (such as S/MIME verification) the
+B<X509_STORE_CTX> structure is created and destroyed internally and the
+only way to set a custom verification callback is by inheriting it from the
+associated B<X509_STORE>.
+
+=head1 RETURN VALUES
+
+X509_STORE_CTX_set_verify_cb() does not return a value.
+
+=head1 EXAMPLES
+
+Default callback operation:
+
+ int verify_callback(int ok, X509_STORE_CTX *ctx)
+ {
+ return ok;
+ }
+
+Simple example, suppose a certificate in the chain is expired and we wish
+to continue after this error:
+
+ int verify_callback(int ok, X509_STORE_CTX *ctx)
+ {
+ /* Tolerate certificate expiration */
+ if (X509_STORE_CTX_get_error(ctx) == X509_V_ERR_CERT_HAS_EXPIRED)
+ return 1;
+ /* Otherwise don't override */
+ return ok;
+ }
+
+More complex example, we don't wish to continue after B<any> certificate has
+expired just one specific case:
+
+ int verify_callback(int ok, X509_STORE_CTX *ctx)
+ {
+ int err = X509_STORE_CTX_get_error(ctx);
+ X509 *err_cert = X509_STORE_CTX_get_current_cert(ctx);
+ if (err == X509_V_ERR_CERT_HAS_EXPIRED)
+ {
+ if (check_is_acceptable_expired_cert(err_cert)
+ return 1;
+ }
+ return ok;
+ }
+
+Full featured logging callback. In this case the B<bio_err> is assumed to be
+a global logging B<BIO>, an alternative would to store a BIO in B<ctx> using
+B<ex_data>.
+
+ int verify_callback(int ok, X509_STORE_CTX *ctx)
+ {
+ X509 *err_cert;
+ int err,depth;
+
+ err_cert = X509_STORE_CTX_get_current_cert(ctx);
+ err = X509_STORE_CTX_get_error(ctx);
+ depth = X509_STORE_CTX_get_error_depth(ctx);
+
+ BIO_printf(bio_err,"depth=%d ",depth);
+ if (err_cert)
+ {
+ X509_NAME_print_ex(bio_err, X509_get_subject_name(err_cert),
+ 0, XN_FLAG_ONELINE);
+ BIO_puts(bio_err, "\n");
+ }
+ else
+ BIO_puts(bio_err, "<no cert>\n");
+ if (!ok)
+ BIO_printf(bio_err,"verify error:num=%d:%s\n",err,
+ X509_verify_cert_error_string(err));
+ switch (err)
+ {
+ case X509_V_ERR_UNABLE_TO_GET_ISSUER_CERT:
+ BIO_puts(bio_err,"issuer= ");
+ X509_NAME_print_ex(bio_err, X509_get_issuer_name(err_cert),
+ 0, XN_FLAG_ONELINE);
+ BIO_puts(bio_err, "\n");
+ break;
+ case X509_V_ERR_CERT_NOT_YET_VALID:
+ case X509_V_ERR_ERROR_IN_CERT_NOT_BEFORE_FIELD:
+ BIO_printf(bio_err,"notBefore=");
+ ASN1_TIME_print(bio_err,X509_get_notBefore(err_cert));
+ BIO_printf(bio_err,"\n");
+ break;
+ case X509_V_ERR_CERT_HAS_EXPIRED:
+ case X509_V_ERR_ERROR_IN_CERT_NOT_AFTER_FIELD:
+ BIO_printf(bio_err,"notAfter=");
+ ASN1_TIME_print(bio_err,X509_get_notAfter(err_cert));
+ BIO_printf(bio_err,"\n");
+ break;
+ case X509_V_ERR_NO_EXPLICIT_POLICY:
+ policies_print(bio_err, ctx);
+ break;
+ }
+ if (err == X509_V_OK && ok == 2)
+ /* print out policies */
+
+ BIO_printf(bio_err,"verify return:%d\n",ok);
+ return(ok);
+ }
+
+=head1 SEE ALSO
+
+L<X509_STORE_CTX_get_error(3)|X509_STORE_CTX_get_error(3)>
+L<X509_STORE_set_verify_cb_func(3)|X509_STORE_set_verify_cb_func(3)>
+L<X509_STORE_CTX_get_ex_new_index(3)|X509_STORE_CTX_get_ex_new_index(3)>
+
+=head1 HISTORY
+
+X509_STORE_CTX_set_verify_cb() is available in all versions of SSLeay and
+OpenSSL.
+
+=cut
diff --git a/doc/crypto/X509_STORE_set_verify_cb_func.pod b/doc/crypto/X509_STORE_set_verify_cb_func.pod
new file mode 100644
index 000000000000..29e3bbe3bce9
--- /dev/null
+++ b/doc/crypto/X509_STORE_set_verify_cb_func.pod
@@ -0,0 +1,54 @@
+=pod
+
+=head1 NAME
+
+X509_STORE_set_verify_cb_func, X509_STORE_set_verify_cb - set verification callback
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509_vfy.h>
+
+ void X509_STORE_set_verify_cb(X509_STORE *st,
+ int (*verify_cb)(int ok, X509_STORE_CTX *ctx));
+
+ void X509_STORE_set_verify_cb_func(X509_STORE *st,
+ int (*verify_cb)(int ok, X509_STORE_CTX *ctx));
+
+=head1 DESCRIPTION
+
+X509_STORE_set_verify_cb() sets the verification callback of B<ctx> to
+B<verify_cb> overwriting any existing callback.
+
+X509_STORE_set_verify_cb_func() also sets the verification callback but it
+is implemented as a macro.
+
+=head1 NOTES
+
+The verification callback from an B<X509_STORE> is inherited by
+the corresponding B<X509_STORE_CTX> structure when it is initialized. This can
+be used to set the verification callback when the B<X509_STORE_CTX> is
+otherwise inaccessible (for example during S/MIME verification).
+
+=head1 BUGS
+
+The macro version of this function was the only one available before
+OpenSSL 1.0.0.
+
+=head1 RETURN VALUES
+
+X509_STORE_set_verify_cb() and X509_STORE_set_verify_cb_func() do not return
+a value.
+
+=head1 SEE ALSO
+
+L<X509_STORE_CTX_set_verify_cb(3)|X509_STORE_CTX_set_verify_cb(3)>
+L<CMS_verify(3)|CMS_verify(3)>
+
+=head1 HISTORY
+
+X509_STORE_set_verify_cb_func() is available in all versions of SSLeay and
+OpenSSL.
+
+X509_STORE_set_verify_cb() was added to OpenSSL 1.0.0.
+
+=cut
diff --git a/doc/crypto/X509_VERIFY_PARAM_set_flags.pod b/doc/crypto/X509_VERIFY_PARAM_set_flags.pod
new file mode 100644
index 000000000000..b68eece03387
--- /dev/null
+++ b/doc/crypto/X509_VERIFY_PARAM_set_flags.pod
@@ -0,0 +1,171 @@
+=pod
+
+=head1 NAME
+
+X509_VERIFY_PARAM_set_flags, X509_VERIFY_PARAM_clear_flags, X509_VERIFY_PARAM_get_flags, X509_VERIFY_PARAM_set_purpose, X509_VERIFY_PARAM_set_trust, X509_VERIFY_PARAM_set_depth, X509_VERIFY_PARAM_get_depth, X509_VERIFY_PARAM_set_time, X509_VERIFY_PARAM_add0_policy, X509_VERIFY_PARAM_set1_policies - X509 verification parameters
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509_vfy.h>
+
+ int X509_VERIFY_PARAM_set_flags(X509_VERIFY_PARAM *param, unsigned long flags);
+ int X509_VERIFY_PARAM_clear_flags(X509_VERIFY_PARAM *param,
+ unsigned long flags);
+ unsigned long X509_VERIFY_PARAM_get_flags(X509_VERIFY_PARAM *param);
+
+ int X509_VERIFY_PARAM_set_purpose(X509_VERIFY_PARAM *param, int purpose);
+ int X509_VERIFY_PARAM_set_trust(X509_VERIFY_PARAM *param, int trust);
+
+ void X509_VERIFY_PARAM_set_time(X509_VERIFY_PARAM *param, time_t t);
+
+ int X509_VERIFY_PARAM_add0_policy(X509_VERIFY_PARAM *param,
+ ASN1_OBJECT *policy);
+ int X509_VERIFY_PARAM_set1_policies(X509_VERIFY_PARAM *param,
+ STACK_OF(ASN1_OBJECT) *policies);
+
+ void X509_VERIFY_PARAM_set_depth(X509_VERIFY_PARAM *param, int depth);
+ int X509_VERIFY_PARAM_get_depth(const X509_VERIFY_PARAM *param);
+
+=head1 DESCRIPTION
+
+These functions manipulate the B<X509_VERIFY_PARAM> structure associated with
+a certificate verification operation.
+
+The X509_VERIFY_PARAM_set_flags() function sets the flags in B<param> by oring
+it with B<flags>. See the B<VERIFICATION FLAGS> section for a complete
+description of values the B<flags> parameter can take.
+
+X509_VERIFY_PARAM_get_flags() returns the flags in B<param>.
+
+X509_VERIFY_PARAM_clear_flags() clears the flags B<flags> in B<param>.
+
+X509_VERIFY_PARAM_set_purpose() sets the verification purpose in B<param>
+to B<purpose>. This determines the acceptable purpose of the certificate
+chain, for example SSL client or SSL server.
+
+X509_VERIFY_PARAM_set_trust() sets the trust setting in B<param> to
+B<trust>.
+
+X509_VERIFY_PARAM_set_time() sets the verification time in B<param> to
+B<t>. Normally the current time is used.
+
+X509_VERIFY_PARAM_add0_policy() enables policy checking (it is disabled
+by default) and adds B<policy> to the acceptable policy set.
+
+X509_VERIFY_PARAM_set1_policies() enables policy checking (it is disabled
+by default) and sets the acceptable policy set to B<policies>. Any existing
+policy set is cleared. The B<policies> parameter can be B<NULL> to clear
+an existing policy set.
+
+X509_VERIFY_PARAM_set_depth() sets the maximum verification depth to B<depth>.
+That is the maximum number of untrusted CA certificates that can appear in a
+chain.
+
+=head1 RETURN VALUES
+
+X509_VERIFY_PARAM_set_flags(), X509_VERIFY_PARAM_clear_flags(),
+X509_VERIFY_PARAM_set_purpose(), X509_VERIFY_PARAM_set_trust(),
+X509_VERIFY_PARAM_add0_policy() and X509_VERIFY_PARAM_set1_policies() return 1
+for success and 0 for failure.
+
+X509_VERIFY_PARAM_get_flags() returns the current verification flags.
+
+X509_VERIFY_PARAM_set_time() and X509_VERIFY_PARAM_set_depth() do not return
+values.
+
+X509_VERIFY_PARAM_get_depth() returns the current verification depth.
+
+=head1 VERIFICATION FLAGS
+
+The verification flags consists of zero or more of the following flags
+ored together.
+
+B<X509_V_FLAG_CRL_CHECK> enables CRL checking for the certificate chain leaf
+certificate. An error occurs if a suitable CRL cannot be found.
+
+B<X509_V_FLAG_CRL_CHECK_ALL> enables CRL checking for the entire certificate
+chain.
+
+B<X509_V_FLAG_IGNORE_CRITICAL> disabled critical extension checking. By default
+any unhandled critical extensions in certificates or (if checked) CRLs results
+in a fatal error. If this flag is set unhandled critical extensions are
+ignored. B<WARNING> setting this option for anything other than debugging
+purposes can be a security risk. Finer control over which extensions are
+supported can be performed in the verification callback.
+
+THe B<X509_V_FLAG_X509_STRICT> flag disables workarounds for some broken
+certificates and makes the verification strictly apply B<X509> rules.
+
+B<X509_V_FLAG_ALLOW_PROXY_CERTS> enables proxy certificate verification.
+
+B<X509_V_FLAG_POLICY_CHECK> enables certificate policy checking, by default
+no policy checking is peformed. Additional information is sent to the
+verification callback relating to policy checking.
+
+B<X509_V_FLAG_EXPLICIT_POLICY>, B<X509_V_FLAG_INHIBIT_ANY> and
+B<X509_V_FLAG_INHIBIT_MAP> set the B<require explicit policy>, B<inhibit any
+policy> and B<inhibit policy mapping> flags respectively as defined in
+B<RFC3280>. Policy checking is automatically enabled if any of these flags
+are set.
+
+If B<X509_V_FLAG_NOTIFY_POLICY> is set and the policy checking is successful
+a special status code is set to the verification callback. This permits it
+to examine the valid policy tree and perform additional checks or simply
+log it for debugging purposes.
+
+By default some addtional features such as indirect CRLs and CRLs signed by
+different keys are disabled. If B<X509_V_FLAG_EXTENDED_CRL_SUPPORT> is set
+they are enabled.
+
+If B<X509_V_FLAG_USE_DELTAS> ise set delta CRLs (if present) are used to
+determine certificate status. If not set deltas are ignored.
+
+B<X509_V_FLAG_CHECK_SS_SIGNATURE> enables checking of the root CA self signed
+cerificate signature. By default this check is disabled because it doesn't
+add any additional security but in some cases applications might want to
+check the signature anyway. A side effect of not checking the root CA
+signature is that disabled or unsupported message digests on the root CA
+are not treated as fatal errors.
+
+The B<X509_V_FLAG_CB_ISSUER_CHECK> flag enables debugging of certificate
+issuer checks. It is B<not> needed unless you are logging certificate
+verification. If this flag is set then additional status codes will be sent
+to the verification callback and it B<must> be prepared to handle such cases
+without assuming they are hard errors.
+
+=head1 NOTES
+
+The above functions should be used to manipulate verification parameters
+instead of legacy functions which work in specific structures such as
+X509_STORE_CTX_set_flags().
+
+=head1 BUGS
+
+Delta CRL checking is currently primitive. Only a single delta can be used and
+(partly due to limitations of B<X509_STORE>) constructed CRLs are not
+maintained.
+
+If CRLs checking is enable CRLs are expected to be available in the
+corresponding B<X509_STORE> structure. No attempt is made to download
+CRLs from the CRL distribution points extension.
+
+=head1 EXAMPLE
+
+Enable CRL checking when performing certificate verification during SSL
+connections associated with an B<SSL_CTX> structure B<ctx>:
+
+ X509_VERIFY_PARAM *param;
+ param = X509_VERIFY_PARAM_new();
+ X509_VERIFY_PARAM_set_flags(param, X509_V_FLAG_CRL_CHECK);
+ SSL_CTX_set1_param(ctx, param);
+ X509_VERIFY_PARAM_free(param);
+
+=head1 SEE ALSO
+
+L<X509_verify_cert(3)|X509_verify_cert(3)>
+
+=head1 HISTORY
+
+TBA
+
+=cut
diff --git a/doc/crypto/X509_verify_cert.pod b/doc/crypto/X509_verify_cert.pod
new file mode 100644
index 000000000000..5253bdcd7072
--- /dev/null
+++ b/doc/crypto/X509_verify_cert.pod
@@ -0,0 +1,53 @@
+=pod
+
+=head1 NAME
+
+X509_verify_cert - discover and verify X509 certificte chain
+
+=head1 SYNOPSIS
+
+ #include <openssl/x509.h>
+
+ int X509_verify_cert(X509_STORE_CTX *ctx);
+
+=head1 DESCRIPTION
+
+The X509_verify_cert() function attempts to discover and validate a
+certificate chain based on parameters in B<ctx>. A complete description of
+the process is contained in the L<verify(1)|verify(1)> manual page.
+
+=head1 RETURN VALUES
+
+If a complete chain can be built and validated this function returns 1,
+otherwise it return zero, in exceptional circumstances it can also
+return a negative code.
+
+If the function fails additional error information can be obtained by
+examining B<ctx> using, for example X509_STORE_CTX_get_error().
+
+=head1 NOTES
+
+Applications rarely call this function directly but it is used by
+OpenSSL internally for certificate validation, in both the S/MIME and
+SSL/TLS code.
+
+The negative return value from X509_verify_cert() can only occur if no
+certificate is set in B<ctx> (due to a programming error) or if a retry
+operation is requested during internal lookups (which never happens with
+standard lookup methods). It is however recommended that application check
+for <= 0 return value on error.
+
+=head1 BUGS
+
+This function uses the header B<x509.h> as opposed to most chain verification
+functiosn which use B<x509_vfy.h>.
+
+=head1 SEE ALSO
+
+L<X509_STORE_CTX_get_error(3)|X509_STORE_CTX_get_error(3)>
+
+=head1 HISTORY
+
+X509_verify_cert() is available in all versions of SSLeay and OpenSSL.
+
+=cut
diff --git a/doc/crypto/d2i_RSAPublicKey.pod b/doc/crypto/d2i_RSAPublicKey.pod
index 279b29c873c0..aa6078bcf6b7 100644
--- a/doc/crypto/d2i_RSAPublicKey.pod
+++ b/doc/crypto/d2i_RSAPublicKey.pod
@@ -11,21 +11,21 @@ d2i_Netscape_RSA - RSA public and private key encoding functions.
#include <openssl/rsa.h>
#include <openssl/x509.h>
- RSA * d2i_RSAPublicKey(RSA **a, unsigned char **pp, long length);
+ RSA * d2i_RSAPublicKey(RSA **a, const unsigned char **pp, long length);
int i2d_RSAPublicKey(RSA *a, unsigned char **pp);
- RSA * d2i_RSA_PUBKEY(RSA **a, unsigned char **pp, long length);
+ RSA * d2i_RSA_PUBKEY(RSA **a, const unsigned char **pp, long length);
int i2d_RSA_PUBKEY(RSA *a, unsigned char **pp);
- RSA * d2i_RSAPrivateKey(RSA **a, unsigned char **pp, long length);
+ RSA * d2i_RSAPrivateKey(RSA **a, const unsigned char **pp, long length);
int i2d_RSAPrivateKey(RSA *a, unsigned char **pp);
int i2d_Netscape_RSA(RSA *a, unsigned char **pp, int (*cb)());
- RSA * d2i_Netscape_RSA(RSA **a, unsigned char **pp, long length, int (*cb)());
+ RSA * d2i_Netscape_RSA(RSA **a, const unsigned char **pp, long length, int (*cb)());
=head1 DESCRIPTION
diff --git a/doc/crypto/ecdsa.pod b/doc/crypto/ecdsa.pod
index 49b10f22499b..20edff97ffd6 100644
--- a/doc/crypto/ecdsa.pod
+++ b/doc/crypto/ecdsa.pod
@@ -114,7 +114,7 @@ using the public key B<eckey>.
ECDSA_size() returns the maximum length signature or 0 on error.
-ECDSA_sign_setup() and ECDSA_sign() return 1 if successful or -1
+ECDSA_sign_setup() and ECDSA_sign() return 1 if successful or 0
on error.
ECDSA_verify() and ECDSA_do_verify() return 1 for a valid
diff --git a/doc/crypto/evp.pod b/doc/crypto/evp.pod
index b3ca14314fac..9faa349243af 100644
--- a/doc/crypto/evp.pod
+++ b/doc/crypto/evp.pod
@@ -22,14 +22,24 @@ digital signatures.
Symmetric encryption is available with the B<EVP_Encrypt>I<...>
functions. The B<EVP_Digest>I<...> functions provide message digests.
+The B<EVP_PKEY>I<...> functions provide a high level interface to
+asymmetric algorithms.
+
Algorithms are loaded with OpenSSL_add_all_algorithms(3).
-All the symmetric algorithms (ciphers) and digests can be replaced by ENGINE
-modules providing alternative implementations. If ENGINE implementations of
-ciphers or digests are registered as defaults, then the various EVP functions
-will automatically use those implementations automatically in preference to
-built in software implementations. For more information, consult the engine(3)
-man page.
+All the symmetric algorithms (ciphers), digests and asymmetric algorithms
+(public key algorithms) can be replaced by ENGINE modules providing alternative
+implementations. If ENGINE implementations of ciphers or digests are registered
+as defaults, then the various EVP functions will automatically use those
+implementations automatically in preference to built in software
+implementations. For more information, consult the engine(3) man page.
+
+Although low level algorithm specific functions exist for many algorithms
+their use is discouraged. They cannot be used with an ENGINE and ENGINE
+versions of new algorithms cannot be accessed using the low level functions.
+Also makes code harder to adapt to new algorithms and some options are not
+cleanly supported at the low level and some operations are more efficient
+using the high level interface.
=head1 SEE ALSO
diff --git a/doc/crypto/hmac.pod b/doc/crypto/hmac.pod
index 484e349fd1d6..d92138d2731b 100644
--- a/doc/crypto/hmac.pod
+++ b/doc/crypto/hmac.pod
@@ -15,12 +15,12 @@ authentication code
void HMAC_CTX_init(HMAC_CTX *ctx);
- void HMAC_Init(HMAC_CTX *ctx, const void *key, int key_len,
+ int HMAC_Init(HMAC_CTX *ctx, const void *key, int key_len,
const EVP_MD *md);
- void HMAC_Init_ex(HMAC_CTX *ctx, const void *key, int key_len,
+ int HMAC_Init_ex(HMAC_CTX *ctx, const void *key, int key_len,
const EVP_MD *md, ENGINE *impl);
- void HMAC_Update(HMAC_CTX *ctx, const unsigned char *data, int len);
- void HMAC_Final(HMAC_CTX *ctx, unsigned char *md, unsigned int *len);
+ int HMAC_Update(HMAC_CTX *ctx, const unsigned char *data, int len);
+ int HMAC_Final(HMAC_CTX *ctx, unsigned char *md, unsigned int *len);
void HMAC_CTX_cleanup(HMAC_CTX *ctx);
void HMAC_cleanup(HMAC_CTX *ctx);
@@ -76,10 +76,13 @@ must have space for the hash function output.
=head1 RETURN VALUES
-HMAC() returns a pointer to the message authentication code.
+HMAC() returns a pointer to the message authentication code or NULL if
+an error occurred.
-HMAC_CTX_init(), HMAC_Init_ex(), HMAC_Update(), HMAC_Final() and
-HMAC_CTX_cleanup() do not return values.
+HMAC_Init_ex(), HMAC_Update() and HMAC_Final() return 1 for success or 0 if
+an error occurred.
+
+HMAC_CTX_init() and HMAC_CTX_cleanup() do not return values.
=head1 CONFORMING TO
@@ -97,4 +100,7 @@ are available since SSLeay 0.9.0.
HMAC_CTX_init(), HMAC_Init_ex() and HMAC_CTX_cleanup() are available
since OpenSSL 0.9.7.
+HMAC_Init_ex(), HMAC_Update() and HMAC_Final() did not return values in
+versions of OpenSSL before 1.0.0.
+
=cut
diff --git a/doc/crypto/i2d_CMS_bio_stream.pod b/doc/crypto/i2d_CMS_bio_stream.pod
new file mode 100644
index 000000000000..558bdd0812c9
--- /dev/null
+++ b/doc/crypto/i2d_CMS_bio_stream.pod
@@ -0,0 +1,44 @@
+=pod
+
+=head1 NAME
+
+ i2d_CMS_bio_stream - output CMS_ContentInfo structure in BER format.
+
+=head1 SYNOPSIS
+
+ #include <openssl/cms.h>
+
+ int i2d_CMS_bio_stream(BIO *out, CMS_ContentInfo *cms, BIO *data, int flags);
+
+=head1 DESCRIPTION
+
+i2d_CMS_bio_stream() outputs a CMS_ContentInfo structure in BER format.
+
+It is otherwise identical to the function SMIME_write_CMS().
+
+=head1 NOTES
+
+This function is effectively a version of the i2d_CMS_bio() supporting
+streaming.
+
+=head1 BUGS
+
+The prefix "i2d" is arguably wrong because the function outputs BER format.
+
+=head1 RETURN VALUES
+
+i2d_CMS_bio_stream() returns 1 for success or 0 for failure.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<CMS_sign(3)|CMS_sign(3)>,
+L<CMS_verify(3)|CMS_verify(3)>, L<CMS_encrypt(3)|CMS_encrypt(3)>
+L<CMS_decrypt(3)|CMS_decrypt(3)>,
+L<SMIME_write_CMS(3)|SMIME_write_CMS(3)>,
+L<PEM_write_bio_CMS_stream(3)|PEM_write_bio_CMS_stream(3)>
+
+=head1 HISTORY
+
+i2d_CMS_bio_stream() was added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/i2d_PKCS7_bio_stream.pod b/doc/crypto/i2d_PKCS7_bio_stream.pod
new file mode 100644
index 000000000000..dc4d884c597e
--- /dev/null
+++ b/doc/crypto/i2d_PKCS7_bio_stream.pod
@@ -0,0 +1,44 @@
+=pod
+
+=head1 NAME
+
+i2d_PKCS7_bio_stream - output PKCS7 structure in BER format.
+
+=head1 SYNOPSIS
+
+ #include <openssl/pkcs7.h>
+
+ int i2d_PKCS7_bio_stream(BIO *out, PKCS7 *p7, BIO *data, int flags);
+
+=head1 DESCRIPTION
+
+i2d_PKCS7_bio_stream() outputs a PKCS7 structure in BER format.
+
+It is otherwise identical to the function SMIME_write_PKCS7().
+
+=head1 NOTES
+
+This function is effectively a version of the d2i_PKCS7_bio() supporting
+streaming.
+
+=head1 BUGS
+
+The prefix "d2i" is arguably wrong because the function outputs BER format.
+
+=head1 RETURN VALUES
+
+i2d_PKCS7_bio_stream() returns 1 for success or 0 for failure.
+
+=head1 SEE ALSO
+
+L<ERR_get_error(3)|ERR_get_error(3)>, L<PKCS7_sign(3)|PKCS7_sign(3)>,
+L<PKCS7_verify(3)|PKCS7_verify(3)>, L<PKCS7_encrypt(3)|PKCS7_encrypt(3)>
+L<PKCS7_decrypt(3)|PKCS7_decrypt(3)>,
+L<SMIME_write_PKCS7(3)|SMIME_write_PKCS7(3)>,
+L<PEM_write_bio_PKCS7_stream(3)|PEM_write_bio_PKCS7_stream(3)>
+
+=head1 HISTORY
+
+i2d_PKCS7_bio_stream() was added to OpenSSL 1.0.0
+
+=cut
diff --git a/doc/crypto/lhash.pod b/doc/crypto/lhash.pod
index dcdbb43a8ed6..73a19b6c7e56 100644
--- a/doc/crypto/lhash.pod
+++ b/doc/crypto/lhash.pod
@@ -8,18 +8,20 @@ lh_new, lh_free, lh_insert, lh_delete, lh_retrieve, lh_doall, lh_doall_arg, lh_e
#include <openssl/lhash.h>
- LHASH *lh_new(LHASH_HASH_FN_TYPE hash, LHASH_COMP_FN_TYPE compare);
- void lh_free(LHASH *table);
+ DECLARE_LHASH_OF(<type>);
- void *lh_insert(LHASH *table, void *data);
- void *lh_delete(LHASH *table, void *data);
- void *lh_retrieve(LHASH *table, void *data);
+ LHASH *lh_<type>_new();
+ void lh_<type>_free(LHASH_OF(<type> *table);
- void lh_doall(LHASH *table, LHASH_DOALL_FN_TYPE func);
- void lh_doall_arg(LHASH *table, LHASH_DOALL_ARG_FN_TYPE func,
- void *arg);
+ <type> *lh_<type>_insert(LHASH_OF(<type> *table, <type> *data);
+ <type> *lh_<type>_delete(LHASH_OF(<type> *table, <type> *data);
+ <type> *lh_retrieve(LHASH_OF<type> *table, <type> *data);
- int lh_error(LHASH *table);
+ void lh_<type>_doall(LHASH_OF(<type> *table, LHASH_DOALL_FN_TYPE func);
+ void lh_<type>_doall_arg(LHASH_OF(<type> *table, LHASH_DOALL_ARG_FN_TYPE func,
+ <type2>, <type2> *arg);
+
+ int lh_<type>_error(LHASH_OF(<type> *table);
typedef int (*LHASH_COMP_FN_TYPE)(const void *, const void *);
typedef unsigned long (*LHASH_HASH_FN_TYPE)(const void *);
@@ -28,113 +30,115 @@ lh_new, lh_free, lh_insert, lh_delete, lh_retrieve, lh_doall, lh_doall_arg, lh_e
=head1 DESCRIPTION
-This library implements dynamic hash tables. The hash table entries
-can be arbitrary structures. Usually they consist of key and value
-fields.
-
-lh_new() creates a new B<LHASH> structure to store arbitrary data
-entries, and provides the 'hash' and 'compare' callbacks to be used in
-organising the table's entries. The B<hash> callback takes a pointer
-to a table entry as its argument and returns an unsigned long hash
-value for its key field. The hash value is normally truncated to a
-power of 2, so make sure that your hash function returns well mixed
-low order bits. The B<compare> callback takes two arguments (pointers
-to two hash table entries), and returns 0 if their keys are equal,
-non-zero otherwise. If your hash table will contain items of some
-particular type and the B<hash> and B<compare> callbacks hash/compare
-these types, then the B<DECLARE_LHASH_HASH_FN> and
-B<IMPLEMENT_LHASH_COMP_FN> macros can be used to create callback
-wrappers of the prototypes required by lh_new(). These provide
-per-variable casts before calling the type-specific callbacks written
-by the application author. These macros, as well as those used for
-the "doall" callbacks, are defined as;
-
- #define DECLARE_LHASH_HASH_FN(f_name,o_type) \
- unsigned long f_name##_LHASH_HASH(const void *);
- #define IMPLEMENT_LHASH_HASH_FN(f_name,o_type) \
- unsigned long f_name##_LHASH_HASH(const void *arg) { \
- o_type a = (o_type)arg; \
- return f_name(a); }
- #define LHASH_HASH_FN(f_name) f_name##_LHASH_HASH
-
- #define DECLARE_LHASH_COMP_FN(f_name,o_type) \
- int f_name##_LHASH_COMP(const void *, const void *);
- #define IMPLEMENT_LHASH_COMP_FN(f_name,o_type) \
- int f_name##_LHASH_COMP(const void *arg1, const void *arg2) { \
- o_type a = (o_type)arg1; \
- o_type b = (o_type)arg2; \
- return f_name(a,b); }
- #define LHASH_COMP_FN(f_name) f_name##_LHASH_COMP
-
- #define DECLARE_LHASH_DOALL_FN(f_name,o_type) \
- void f_name##_LHASH_DOALL(const void *);
- #define IMPLEMENT_LHASH_DOALL_FN(f_name,o_type) \
- void f_name##_LHASH_DOALL(const void *arg) { \
- o_type a = (o_type)arg; \
- f_name(a); }
- #define LHASH_DOALL_FN(f_name) f_name##_LHASH_DOALL
-
- #define DECLARE_LHASH_DOALL_ARG_FN(f_name,o_type,a_type) \
- void f_name##_LHASH_DOALL_ARG(const void *, const void *);
- #define IMPLEMENT_LHASH_DOALL_ARG_FN(f_name,o_type,a_type) \
- void f_name##_LHASH_DOALL_ARG(const void *arg1, const void *arg2) { \
- o_type a = (o_type)arg1; \
- a_type b = (a_type)arg2; \
- f_name(a,b); }
- #define LHASH_DOALL_ARG_FN(f_name) f_name##_LHASH_DOALL_ARG
-
-An example of a hash table storing (pointers to) structures of type 'STUFF'
-could be defined as follows;
+This library implements type-checked dynamic hash tables. The hash
+table entries can be arbitrary structures. Usually they consist of key
+and value fields.
+
+lh_<type>_new() creates a new B<LHASH_OF(<type>> structure to store
+arbitrary data entries, and provides the 'hash' and 'compare'
+callbacks to be used in organising the table's entries. The B<hash>
+callback takes a pointer to a table entry as its argument and returns
+an unsigned long hash value for its key field. The hash value is
+normally truncated to a power of 2, so make sure that your hash
+function returns well mixed low order bits. The B<compare> callback
+takes two arguments (pointers to two hash table entries), and returns
+0 if their keys are equal, non-zero otherwise. If your hash table
+will contain items of some particular type and the B<hash> and
+B<compare> callbacks hash/compare these types, then the
+B<DECLARE_LHASH_HASH_FN> and B<IMPLEMENT_LHASH_COMP_FN> macros can be
+used to create callback wrappers of the prototypes required by
+lh_<type>_new(). These provide per-variable casts before calling the
+type-specific callbacks written by the application author. These
+macros, as well as those used for the "doall" callbacks, are defined
+as;
+
+ #define DECLARE_LHASH_HASH_FN(name, o_type) \
+ unsigned long name##_LHASH_HASH(const void *);
+ #define IMPLEMENT_LHASH_HASH_FN(name, o_type) \
+ unsigned long name##_LHASH_HASH(const void *arg) { \
+ const o_type *a = arg; \
+ return name##_hash(a); }
+ #define LHASH_HASH_FN(name) name##_LHASH_HASH
+
+ #define DECLARE_LHASH_COMP_FN(name, o_type) \
+ int name##_LHASH_COMP(const void *, const void *);
+ #define IMPLEMENT_LHASH_COMP_FN(name, o_type) \
+ int name##_LHASH_COMP(const void *arg1, const void *arg2) { \
+ const o_type *a = arg1; \
+ const o_type *b = arg2; \
+ return name##_cmp(a,b); }
+ #define LHASH_COMP_FN(name) name##_LHASH_COMP
+
+ #define DECLARE_LHASH_DOALL_FN(name, o_type) \
+ void name##_LHASH_DOALL(void *);
+ #define IMPLEMENT_LHASH_DOALL_FN(name, o_type) \
+ void name##_LHASH_DOALL(void *arg) { \
+ o_type *a = arg; \
+ name##_doall(a); }
+ #define LHASH_DOALL_FN(name) name##_LHASH_DOALL
+
+ #define DECLARE_LHASH_DOALL_ARG_FN(name, o_type, a_type) \
+ void name##_LHASH_DOALL_ARG(void *, void *);
+ #define IMPLEMENT_LHASH_DOALL_ARG_FN(name, o_type, a_type) \
+ void name##_LHASH_DOALL_ARG(void *arg1, void *arg2) { \
+ o_type *a = arg1; \
+ a_type *b = arg2; \
+ name##_doall_arg(a, b); }
+ #define LHASH_DOALL_ARG_FN(name) name##_LHASH_DOALL_ARG
+
+ An example of a hash table storing (pointers to) structures of type 'STUFF'
+ could be defined as follows;
/* Calculates the hash value of 'tohash' (implemented elsewhere) */
unsigned long STUFF_hash(const STUFF *tohash);
/* Orders 'arg1' and 'arg2' (implemented elsewhere) */
- int STUFF_cmp(const STUFF *arg1, const STUFF *arg2);
+ int stuff_cmp(const STUFF *arg1, const STUFF *arg2);
/* Create the type-safe wrapper functions for use in the LHASH internals */
- static IMPLEMENT_LHASH_HASH_FN(STUFF_hash, const STUFF *)
- static IMPLEMENT_LHASH_COMP_FN(STUFF_cmp, const STUFF *);
+ static IMPLEMENT_LHASH_HASH_FN(stuff, STUFF);
+ static IMPLEMENT_LHASH_COMP_FN(stuff, STUFF);
/* ... */
int main(int argc, char *argv[]) {
/* Create the new hash table using the hash/compare wrappers */
- LHASH *hashtable = lh_new(LHASH_HASH_FN(STUFF_hash),
+ LHASH_OF(STUFF) *hashtable = lh_STUFF_new(LHASH_HASH_FN(STUFF_hash),
LHASH_COMP_FN(STUFF_cmp));
/* ... */
}
-lh_free() frees the B<LHASH> structure B<table>. Allocated hash table
-entries will not be freed; consider using lh_doall() to deallocate any
-remaining entries in the hash table (see below).
+lh_<type>_free() frees the B<LHASH_OF(<type>> structure
+B<table>. Allocated hash table entries will not be freed; consider
+using lh_<type>_doall() to deallocate any remaining entries in the
+hash table (see below).
-lh_insert() inserts the structure pointed to by B<data> into B<table>.
-If there already is an entry with the same key, the old value is
-replaced. Note that lh_insert() stores pointers, the data are not
-copied.
+lh_<type>_insert() inserts the structure pointed to by B<data> into
+B<table>. If there already is an entry with the same key, the old
+value is replaced. Note that lh_<type>_insert() stores pointers, the
+data are not copied.
-lh_delete() deletes an entry from B<table>.
+lh_<type>_delete() deletes an entry from B<table>.
-lh_retrieve() looks up an entry in B<table>. Normally, B<data> is
-a structure with the key field(s) set; the function will return a
+lh_<type>_retrieve() looks up an entry in B<table>. Normally, B<data>
+is a structure with the key field(s) set; the function will return a
pointer to a fully populated structure.
-lh_doall() will, for every entry in the hash table, call B<func> with
-the data item as its parameter. For lh_doall() and lh_doall_arg(),
-function pointer casting should be avoided in the callbacks (see
-B<NOTE>) - instead, either declare the callbacks to match the
-prototype required in lh_new() or use the declare/implement macros to
-create type-safe wrappers that cast variables prior to calling your
-type-specific callbacks. An example of this is illustrated here where
-the callback is used to cleanup resources for items in the hash table
-prior to the hashtable itself being deallocated:
+lh_<type>_doall() will, for every entry in the hash table, call
+B<func> with the data item as its parameter. For lh_<type>_doall()
+and lh_<type>_doall_arg(), function pointer casting should be avoided
+in the callbacks (see B<NOTE>) - instead use the declare/implement
+macros to create type-checked wrappers that cast variables prior to
+calling your type-specific callbacks. An example of this is
+illustrated here where the callback is used to cleanup resources for
+items in the hash table prior to the hashtable itself being
+deallocated:
/* Cleans up resources belonging to 'a' (this is implemented elsewhere) */
- void STUFF_cleanup(STUFF *a);
+ void STUFF_cleanup_doall(STUFF *a);
/* Implement a prototype-compatible wrapper for "STUFF_cleanup" */
- IMPLEMENT_LHASH_DOALL_FN(STUFF_cleanup, STUFF *)
+ IMPLEMENT_LHASH_DOALL_FN(STUFF_cleanup, STUFF)
/* ... then later in the code ... */
/* So to run "STUFF_cleanup" against all items in a hash table ... */
- lh_doall(hashtable, LHASH_DOALL_FN(STUFF_cleanup));
+ lh_STUFF_doall(hashtable, LHASH_DOALL_FN(STUFF_cleanup));
/* Then the hash table itself can be deallocated */
- lh_free(hashtable);
+ lh_STUFF_free(hashtable);
When doing this, be careful if you delete entries from the hash table
in your callbacks: the table may decrease in size, moving the item
@@ -145,51 +149,52 @@ you start (which will stop the hash table ever decreasing in size).
The best solution is probably to avoid deleting items from the hash
table inside a "doall" callback!
-lh_doall_arg() is the same as lh_doall() except that B<func> will be
-called with B<arg> as the second argument and B<func> should be of
-type B<LHASH_DOALL_ARG_FN_TYPE> (a callback prototype that is passed
-both the table entry and an extra argument). As with lh_doall(), you
-can instead choose to declare your callback with a prototype matching
-the types you are dealing with and use the declare/implement macros to
-create compatible wrappers that cast variables before calling your
-type-specific callbacks. An example of this is demonstrated here
-(printing all hash table entries to a BIO that is provided by the
-caller):
+lh_<type>_doall_arg() is the same as lh_<type>_doall() except that
+B<func> will be called with B<arg> as the second argument and B<func>
+should be of type B<LHASH_DOALL_ARG_FN_TYPE> (a callback prototype
+that is passed both the table entry and an extra argument). As with
+lh_doall(), you can instead choose to declare your callback with a
+prototype matching the types you are dealing with and use the
+declare/implement macros to create compatible wrappers that cast
+variables before calling your type-specific callbacks. An example of
+this is demonstrated here (printing all hash table entries to a BIO
+that is provided by the caller):
/* Prints item 'a' to 'output_bio' (this is implemented elsewhere) */
- void STUFF_print(const STUFF *a, BIO *output_bio);
+ void STUFF_print_doall_arg(const STUFF *a, BIO *output_bio);
/* Implement a prototype-compatible wrapper for "STUFF_print" */
- static IMPLEMENT_LHASH_DOALL_ARG_FN(STUFF_print, const STUFF *, BIO *)
+ static IMPLEMENT_LHASH_DOALL_ARG_FN(STUFF, const STUFF, BIO)
/* ... then later in the code ... */
/* Print out the entire hashtable to a particular BIO */
- lh_doall_arg(hashtable, LHASH_DOALL_ARG_FN(STUFF_print), logging_bio);
+ lh_STUFF_doall_arg(hashtable, LHASH_DOALL_ARG_FN(STUFF_print), BIO,
+ logging_bio);
-lh_error() can be used to determine if an error occurred in the last
-operation. lh_error() is a macro.
+lh_<type>_error() can be used to determine if an error occurred in the last
+operation. lh_<type>_error() is a macro.
=head1 RETURN VALUES
-lh_new() returns B<NULL> on error, otherwise a pointer to the new
+lh_<type>_new() returns B<NULL> on error, otherwise a pointer to the new
B<LHASH> structure.
-When a hash table entry is replaced, lh_insert() returns the value
+When a hash table entry is replaced, lh_<type>_insert() returns the value
being replaced. B<NULL> is returned on normal operation and on error.
-lh_delete() returns the entry being deleted. B<NULL> is returned if
+lh_<type>_delete() returns the entry being deleted. B<NULL> is returned if
there is no such value in the hash table.
-lh_retrieve() returns the hash table entry if it has been found,
+lh_<type>_retrieve() returns the hash table entry if it has been found,
B<NULL> otherwise.
-lh_error() returns 1 if an error occurred in the last operation, 0
+lh_<type>_error() returns 1 if an error occurred in the last operation, 0
otherwise.
-lh_free(), lh_doall() and lh_doall_arg() return no values.
+lh_<type>_free(), lh_<type>_doall() and lh_<type>_doall_arg() return no values.
=head1 NOTE
The various LHASH macros and callback types exist to make it possible
-to write type-safe code without resorting to function-prototype
+to write type-checked code without resorting to function-prototype
casting - an evil that makes application code much harder to
audit/verify and also opens the window of opportunity for stack
corruption and other hard-to-find bugs. It also, apparently, violates
@@ -227,7 +232,7 @@ without any "const" qualifiers.
=head1 BUGS
-lh_insert() returns B<NULL> both for success and error.
+lh_<type>_insert() returns B<NULL> both for success and error.
=head1 INTERNALS
@@ -272,8 +277,8 @@ lh_strhash() is a demo string hashing function:
unsigned long lh_strhash(const char *c);
Since the B<LHASH> routines would normally be passed structures, this
-routine would not normally be passed to lh_new(), rather it would be
-used in the function passed to lh_new().
+routine would not normally be passed to lh_<type>_new(), rather it would be
+used in the function passed to lh_<type>_new().
=head1 SEE ALSO
@@ -291,4 +296,7 @@ were changed for better type safety, and the function types LHASH_COMP_FN_TYPE,
LHASH_HASH_FN_TYPE, LHASH_DOALL_FN_TYPE and LHASH_DOALL_ARG_FN_TYPE
became available.
+In OpenSSL 1.0.0, the lhash interface was revamped for even better
+type checking.
+
=cut
diff --git a/doc/crypto/threads.pod b/doc/crypto/threads.pod
index 3df4ecd7768e..dc0e9391dc20 100644
--- a/doc/crypto/threads.pod
+++ b/doc/crypto/threads.pod
@@ -2,7 +2,9 @@
=head1 NAME
-CRYPTO_set_locking_callback, CRYPTO_set_id_callback, CRYPTO_num_locks,
+CRYPTO_THREADID_set_callback, CRYPTO_THREADID_get_callback,
+CRYPTO_THREADID_current, CRYPTO_THREADID_cmp, CRYPTO_THREADID_cpy,
+CRYPTO_THREADID_hash, CRYPTO_set_locking_callback, CRYPTO_num_locks,
CRYPTO_set_dynlock_create_callback, CRYPTO_set_dynlock_lock_callback,
CRYPTO_set_dynlock_destroy_callback, CRYPTO_get_new_dynlockid,
CRYPTO_destroy_dynlockid, CRYPTO_lock - OpenSSL thread support
@@ -11,14 +13,26 @@ CRYPTO_destroy_dynlockid, CRYPTO_lock - OpenSSL thread support
#include <openssl/crypto.h>
- void CRYPTO_set_locking_callback(void (*locking_function)(int mode,
- int n, const char *file, int line));
-
- void CRYPTO_set_id_callback(unsigned long (*id_function)(void));
+ /* Don't use this structure directly. */
+ typedef struct crypto_threadid_st
+ {
+ void *ptr;
+ unsigned long val;
+ } CRYPTO_THREADID;
+ /* Only use CRYPTO_THREADID_set_[numeric|pointer]() within callbacks */
+ void CRYPTO_THREADID_set_numeric(CRYPTO_THREADID *id, unsigned long val);
+ void CRYPTO_THREADID_set_pointer(CRYPTO_THREADID *id, void *ptr);
+ int CRYPTO_THREADID_set_callback(void (*threadid_func)(CRYPTO_THREADID *));
+ void (*CRYPTO_THREADID_get_callback(void))(CRYPTO_THREADID *);
+ void CRYPTO_THREADID_current(CRYPTO_THREADID *id);
+ int CRYPTO_THREADID_cmp(const CRYPTO_THREADID *a,
+ const CRYPTO_THREADID *b);
+ void CRYPTO_THREADID_cpy(CRYPTO_THREADID *dest,
+ const CRYPTO_THREADID *src);
+ unsigned long CRYPTO_THREADID_hash(const CRYPTO_THREADID *id);
int CRYPTO_num_locks(void);
-
/* struct CRYPTO_dynlock_value needs to be defined by the user */
struct CRYPTO_dynlock_value;
@@ -50,7 +64,8 @@ CRYPTO_destroy_dynlockid, CRYPTO_lock - OpenSSL thread support
=head1 DESCRIPTION
OpenSSL can safely be used in multi-threaded applications provided
-that at least two callback functions are set.
+that at least two callback functions are set, locking_function and
+threadid_func.
locking_function(int mode, int n, const char *file, int line) is
needed to perform locking on shared data structures.
@@ -65,10 +80,42 @@ B<CRYPTO_LOCK>, and releases it otherwise.
B<file> and B<line> are the file number of the function setting the
lock. They can be useful for debugging.
-id_function(void) is a function that returns a thread ID, for example
-pthread_self() if it returns an integer (see NOTES below). It isn't
-needed on Windows nor on platforms where getpid() returns a different
-ID for each thread (see NOTES below).
+threadid_func(CRYPTO_THREADID *id) is needed to record the currently-executing
+thread's identifier into B<id>. The implementation of this callback should not
+fill in B<id> directly, but should use CRYPTO_THREADID_set_numeric() if thread
+IDs are numeric, or CRYPTO_THREADID_set_pointer() if they are pointer-based.
+If the application does not register such a callback using
+CRYPTO_THREADID_set_callback(), then a default implementation is used - on
+Windows and BeOS this uses the system's default thread identifying APIs, and on
+all other platforms it uses the address of B<errno>. The latter is satisfactory
+for thread-safety if and only if the platform has a thread-local error number
+facility.
+
+Once threadid_func() is registered, or if the built-in default implementation is
+to be used;
+
+=over 4
+
+=item *
+CRYPTO_THREADID_current() records the currently-executing thread ID into the
+given B<id> object.
+
+=item *
+CRYPTO_THREADID_cmp() compares two thread IDs (returning zero for equality, ie.
+the same semantics as memcmp()).
+
+=item *
+CRYPTO_THREADID_cpy() duplicates a thread ID value,
+
+=item *
+CRYPTO_THREADID_hash() returns a numeric value usable as a hash-table key. This
+is usually the exact numeric or pointer-based thread ID used internally, however
+this also handles the unusual case where pointers are larger than 'long'
+variables and the platform's thread IDs are pointer-based - in this case, mixing
+is done to attempt to produce a unique numeric value even though it is not as
+wide as the platform's true thread IDs.
+
+=back
Additionally, OpenSSL supports dynamic locks, and sometimes, some parts
of OpenSSL need it for better performance. To enable this, the following
@@ -140,22 +187,6 @@ You can find out if OpenSSL was configured with thread support:
Also, dynamic locks are currently not used internally by OpenSSL, but
may do so in the future.
-Defining id_function(void) has it's own issues. Generally speaking,
-pthread_self() should be used, even on platforms where getpid() gives
-different answers in each thread, since that may depend on the machine
-the program is run on, not the machine where the program is being
-compiled. For instance, Red Hat 8 Linux and earlier used
-LinuxThreads, whose getpid() returns a different value for each
-thread. Red Hat 9 Linux and later use NPTL, which is
-Posix-conformant, and has a getpid() that returns the same value for
-all threads in a process. A program compiled on Red Hat 8 and run on
-Red Hat 9 will therefore see getpid() returning the same value for
-all threads.
-
-There is still the issue of platforms where pthread_self() returns
-something other than an integer. This is a bit unusual, and this
-manual has no cookbook solution for that case.
-
=head1 EXAMPLES
B<crypto/threads/mttest.c> shows examples of the callback functions on
@@ -163,10 +194,14 @@ Solaris, Irix and Win32.
=head1 HISTORY
-CRYPTO_set_locking_callback() and CRYPTO_set_id_callback() are
+CRYPTO_set_locking_callback() is
available in all versions of SSLeay and OpenSSL.
CRYPTO_num_locks() was added in OpenSSL 0.9.4.
All functions dealing with dynamic locks were added in OpenSSL 0.9.5b-dev.
+B<CRYPTO_THREADID> and associated functions were introduced in OpenSSL 1.0.0
+to replace (actually, deprecate) the previous CRYPTO_set_id_callback(),
+CRYPTO_get_id_callback(), and CRYPTO_thread_id() functions which assumed
+thread IDs to always be represented by 'unsigned long'.
=head1 SEE ALSO