The provider context structure is made to include the following information:
- The core provider handle (first argument to the provider init
function). This handle is meant to be used in all upcalls that need
it.
- A library context, used for any libcrypto calls that need it, done in
the provider itself.
Regarding the library context, that's generally only needed if the
provider makes any libcrypto calls, i.e. is linked with libcrypto. That
happens to be the case for all OpenSSL providers, but is applicable for
other providers that use libcrypto internally as well.
The normal thing to do for a provider init function is to create its own
library context. For a provider that's meant to become a dynamically
loadable module, this is what MUST be done.
However, we do not do that in the default provider; it uses the library
context associated with the core provider handle instead. This is
permissible, although generally discouraged, as long as the provider in
question is guaranteed to be built-in, into libcrypto or into the
application that uses it.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11803)
This started with adding forward declarations of all provider side
interface functions, and fixing all compiler errors.
Furthermore, diminish the faulty assumption that the provider context
is and always will be just a library context. That means adding a
teardown function in all providers that aren't necessarily built into
libcrypto.
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11777)
Partial Fix for #11648.
Some additional work still needs to be done to support RSA-PSS mode.
RSA legacy digests will be addressed in another PR.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11681)
EVP_PKEY_CTX_gettable_params() was missing code for the keygen operation.
After adding it it was noticed that it is probably not required for this type, so instead
the gen_get_params and gen_gettable_params have been remnoved from the provider interface.
gen_get_params was only implemented for ec to get the curve name. This seems redundant
since normally you would set parameters into the keygen_init() and then generate a key.
Normally you would expect to extract data from the key - not the object that we just set up
to do the keygen.
Added a simple settable and gettable test into a test that does keygen.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11683)
Fixes#11459
It was incorrectly using 8 bytes instead of 16 as the default.
This was verified by expanding the macros used in e_cast.c.
The issue occurs if EVP_CIPHER_CTX_set_key_length() is not called.
evp_test.c hides this issue as it always calls EVP_CIPHER_CTX_set_key_length() before
using EVP_CipherInit_ex(...., key, ..).
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11707)
We had a redundant couple of lines where we exported key data twice.
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11635)
At various points we need to be able to retrieve the current library
context so we store it in the ECX_KEY structure.
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11635)
ECX keys can very easily crete the public key from the private key.
Therefore when we import ecx keys it is sufficent to just have the private
key.
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11635)
The legacy provider contains assembler references. Most code is automagically pulled in from the libcrypto - but the platform specific assembler functions will not be visible in the symbol table. Copying BNASM and DESASM into liblegacy seems to be a better solution than exposing platform specific function in libcrypto.num.
Added a missing call in the des_cbc code for sparc.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11697)
A small number of files contain references to the "OpenSSL license"
which has been deprecated and replaced by the "Apache License 2.0".
Amend the occurences.
Fixes#11649
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11663)
This macro is used to determine if certain pieces of code should
become part of the FIPS module or not. The old name was confusing.
Fixes#11538
Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org>
(Merged from https://github.com/openssl/openssl/pull/11539)
This adds handling of the parameter "mandatory-digest" and responds
with an empty string, meaning that no digest may be used.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11576)
Several MACs and one KDF are included in the FIPS provider with the property
"fips=yes" set but are not listed as being part of the OpenSSL validation.
This removes them from the FIPS provider.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11602)
DH_set0_pqg() is now responsible for caching the nid, q and length.
DH with or without named safe prime groups now default to using the maximum private key length (BN_num_bits(q) - 1)
when generating a DH private key. The code is now shared between fips and non fips mode for DH key generation.
The OSSL_PKEY_PARAM_DH_PRIV_LEN parameter can be used during keygen to override the maximum private key length to be
in the range (2 * strength ... bits(q) - 1). Where the strength depends on the length of p.
Added q = (p - 1) / 2 safe prime BIGNUMS so that the code is data driven (To simplify adding new names).
The BIGNUMS were code generated.
Fix error in documented return value for DH_get_nid
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11562)
Moved some shared FFC code into the FFC files.
Added extra paramgen parameters for seed, gindex.
Fixed bug in ossl_prov util to print bignums.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11303)
In a similar way to commit 76e23fc5 we must ensure that we use a libctx
whenever we call EC_POINT_point2buf because it can end up using crypto
algorithms.
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11535)
The cipher_tdes_common causes build failure as being duplicated
in libcrypto static builds.
[extended tests]
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11544)
DES implementations were missing the dup/copy ctx routines
required by CMAC implementation. A regression test is added.
Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11498)
Ed25519 needs to fetch a digest and so needs to use the correct libctx.
Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org>
(Merged from https://github.com/openssl/openssl/pull/11496)
By loading the null provider into the default context, it is possible
to verify that it is not accidentally being used.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11488)
DES, idea, seed, rc2, rc4, rc5, cast and blowfish have been moved out of the default provider.
Code shared between desx and tdes has been moved into a seperate file (cipher_tdes_common.c).
3 test recipes failed due to using app/openssl calls that used legacy ciphers.
These calls have been updated to supply both the default and legacy providers.
Fixed openssl app '-provider' memory leak
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11419)
The exporter freed a buffer too soon, and there were attempts to use
its data later, which was overwritten by something else at that
point.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11358)
This library is meant to be small and quick. It's based on WPACKET,
which was extended to support DER writing. The way it's used is a
bit unusual, as it's used to write the structures backward into a
given buffer. A typical quick call looks like this:
/*
* Fill in this structure:
*
* something ::= SEQUENCE {
* id OBJECT IDENTIFIER,
* x [0] INTEGER OPTIONAL,
* y [1] BOOLEAN OPTIONAL,
* n INTEGER
* }
*/
unsigned char buf[nnnn], *p = NULL;
size_t encoded_len = 0;
WPACKET pkt;
int ok;
ok = WPACKET_init_der(&pkt, buf, sizeof(buf)
&& DER_w_start_sequence(&pkt, -1)
&& DER_w_bn(&pkt, -1, bn)
&& DER_w_boolean(&pkt, 1, bool)
&& DER_w_precompiled(&pkt, -1, OID, sizeof(OID))
&& DER_w_end_sequence(&pkt, -1)
&& WPACKET_finish(&pkt)
&& WPACKET_get_total_written(&pkt, &encoded_len)
&& (p = WPACKET_get_curr(&pkt)) != NULL;
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11450)
We have an old OID database that's not as readable as would be
desired, and we have spots with hand coded DER for well known OIDs.
The perl modules added here give enough support that we can parse
OBJECT IDENTIFIER definitions and encode them as DER.
OpenSSL::OID is a general OID parsing and encoding of ASN.1
definitions, and supports enough of the X.680 syntax to understand
what we find in RFCs and similar documents and produce the DER
encoding for them.
oids_to_c is a specialized module to convert the DER encoding from
OpenSSL::OID to C code. This is primarily useful in file templates
that are processed with util/dofile.pl.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11450)
In function 'ccm_tls_cipher',
inlined from 'ccm_cipher_internal' at providers/common/ciphers/cipher_ccm.c:359:16,
inlined from 'ccm_stream_final' at providers/common/ciphers/cipher_ccm.c:265:9:
providers/common/ciphers/cipher_ccm.c:317:5: error: argument 2 null where non-null expected [-Werror=nonnull]
317 | memcpy(ctx->iv + EVP_CCM_TLS_FIXED_IV_LEN, in, EVP_CCM_TLS_EXPLICIT_IV_LEN);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from include/internal/cryptlib.h:14,
from providers/common/include/prov/ciphercommon.h:14,
from providers/common/ciphers/cipher_ccm.c:12:
providers/common/ciphers/cipher_ccm.c: In function 'ccm_stream_final':
/home/ed/gnu/arm-linux-gnueabihf-linux64/arm-linux-gnueabihf/sys-include/string.h:44:14: note: in a call to function 'memcpy' declared here
44 | extern void *memcpy (void *__restrict __dest,
| ^~~~~~
[extended tests]
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/10344)
Currently only RSA, EC and ECX are supported (DH and DSA need to be added to the keygen
PR's seperately because the fields supported have changed significantly).
The API's require the keys to be provider based.
Made the keymanagement export and get_params functions share the same code by supplying
support functions that work for both a OSSL_PARAM_BLD as well as a OSSL_PARAM[].
This approach means that complex code is not required to build an
empty OSSL_PARAM[] with the correct sized fields before then doing a second
pass to populate the array.
The RSA factor arrays have been changed to use unique key names to simplify the interface
needed by the user.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11365)
Since this is public, it is best to make the underlying structure opaque.
This means converting from stack allocation to dynamic allocation for all
usages.
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/11390)
The catalyst for this is the difficult of passing BNs through the other
OSSL_PARAM APIs.
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/11390)
The provider key export functions for EC_KEY assumed that a public key
is always present, and would fail if not. This blocks any attempt to
export a key structure with only domain parameters.
This is similar to earlier work done in EVP_PKEY_ASN1_METHODs.
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/11394)
Some fetch failurs are ok and should be ignored.
Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11405)
Downgrading EVP_PKEYs from containing provider side internal keys to
containing legacy keys demands support in the EVP_PKEY_ASN1_METHOD.
This became a bit elaborate because the code would be almost exactly
the same as the import functions int EVP_KEYMGMT. Therefore, we end
up moving most of the code to common backend support files that can be
used both by legacy backend code and by our providers.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/11375)
