root
/
openssl
mirror of https://github.com/openssl/openssl.git
1
0
Fork 0
Code Issues Actions 7 Packages Projects Releases Wiki Activity
openssl/test/evp_libctx_test.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

852 lines
32 KiB
C
Raw Normal View History

Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
/*
Copyright year updates Reviewed-by: Neil Horman <nhorman@openssl.org> Release: yes (cherry picked from commit 0ce7d1f355c1240653e320a3f6f8109c1f05f8c0) Reviewed-by: Hugo Landau <hlandau@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/24034)
2024-03-20 20:07:54 +08:00
* Copyright 2020-2024 The OpenSSL Project Authors. All Rights Reserved.
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
*
* Licensed under the Apache License 2.0 (the "License"). You may not use
* this file except in compliance with the License. You can obtain a copy
* in the file LICENSE in the source distribution or at
* https://www.openssl.org/source/license.html
*/
/*
* These tests are setup to load null into the default library context.
* Any tests are expected to use the created 'libctx' to find algorithms.
* The framework runs the tests twice using the 'default' provider or
* 'fips' provider as inputs.
*/
/*
* DSA/DH low level APIs are deprecated for public use, but still ok for
* internal use.
*/
#include "internal/deprecated.h"
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
#include <assert.h>
Add FIPS indicator support for Triple-DES encryption. This leaves 3DES with the FIPS query "FIPS=yes", which allows Triple-DES to be used for Decryption by default. Disallow CMAC using Triple-DES in FIPS. This does not use a FIPS indicator. Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Paul Dale <ppzgs1@gmail.com> (Merged from https://github.com/openssl/openssl/pull/24960)
2024-07-22 15:24:53 +08:00
#include <string.h>
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
#include <openssl/evp.h>
#include <openssl/provider.h>
#include <openssl/dsa.h>
Fix no-dh and no-dsa Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/12525)
2020-07-24 19:04:00 +08:00
#include <openssl/dh.h>
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
#include <openssl/safestack.h>
test/evp_libctx_test.c: use OSSL_ENCODER instead of i2d_PublicKey() Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13167)
2020-10-17 17:44:43 +08:00
#include <openssl/core_dispatch.h>
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
#include <openssl/core_names.h>
Add dh_kdf support to provider Similiar to ecdh this supports the legacy kdf inside the provider dh key exchange. The supporting EVP_PKEY_CTX macros have been changed into mehtods and moved into dh_ctrl.c New kdfs such as SSKDF should be done as a seperate pass after doing the derive. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12575)
2020-08-04 09:21:21 +08:00
#include <openssl/x509.h>
test/evp_libctx_test.c: use OSSL_ENCODER instead of i2d_PublicKey() Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13167)
2020-10-17 17:44:43 +08:00
#include <openssl/encoder.h>
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
#include "testutil.h"
#include "internal/nelem.h"
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
#include "crypto/bn_dh.h" /* _bignum_ffdhe2048_p */
Rename OPENSSL_CTX prefix to OSSL_LIB_CTX Many of the new types introduced by OpenSSL 3.0 have an OSSL_ prefix, e.g., OSSL_CALLBACK, OSSL_PARAM, OSSL_ALGORITHM, OSSL_SERIALIZER. The OPENSSL_CTX type stands out a little by using a different prefix. For consistency reasons, this type is renamed to OSSL_LIB_CTX. Reviewed-by: Paul Dale <paul.dale@oracle.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12621)
2020-10-15 17:55:50 +08:00
static OSSL_LIB_CTX *libctx = NULL;
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
static OSSL_PROVIDER *nullprov = NULL;
static OSSL_PROVIDER *libprov = NULL;
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
static STACK_OF(OPENSSL_STRING) *cipher_names = NULL;
Older FIPS providers require a kemop Don't expect success with dated FIPS modules. Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/26904)
2025-02-26 08:43:35 +08:00
static int is_fips = 0;
static int is_fips_lt_3_5 = 0;
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
typedef enum OPTION_choice {
OPT_ERR = -1,
OPT_EOF = 0,
OPT_CONFIG_FILE,
OPT_PROVIDER_NAME,
OPT_TEST_ENUM
} OPTION_CHOICE;
const OPTIONS *test_get_options(void)
{
static const OPTIONS test_options[] = {
OPT_TEST_OPTIONS_DEFAULT_USAGE,
{ "config", OPT_CONFIG_FILE, '<',
"The configuration file to use for the libctx" },
{ "provider", OPT_PROVIDER_NAME, 's',
Enhance the encoder/decoder tests to allow testing with a non-default library context and configurable providers Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14587)
2021-03-15 12:26:09 +08:00
"The provider to load (The default value is 'default')" },
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
{ NULL }
};
return test_options;
}
Fix no-dh and no-dsa Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/12525)
2020-07-24 19:04:00 +08:00
#ifndef OPENSSL_NO_DH
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
static const char *getname(int id)
{
const char *name[] = {"p", "q", "g" };
if (id >= 0 && id < 3)
return name[id];
return "?";
}
#endif
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
static int test_evp_cipher_api_safety(void)
{
int ret = 0;
EVP_CIPHER_CTX *ctx = NULL;
ctx = EVP_CIPHER_CTX_new();
if (!TEST_ptr(ctx))
goto err;
/*
* Ensure that EVP_CIPHER_get_block_size returns 0
Fix new typos found by codespell Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/23545)
2024-02-12 01:14:30 +08:00
* if we haven't initialized the cipher in this context
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
*/
if (!TEST_int_eq(EVP_CIPHER_CTX_get_block_size(ctx), 0))
goto err_free;
/*
* Ensure that EVP_CIPHER_get_iv_length returns 0
Fix new typos found by codespell Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/23545)
2024-02-12 01:14:30 +08:00
* if we haven't initialized the cipher in this context
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
*/
if (!TEST_int_eq(EVP_CIPHER_CTX_get_iv_length(ctx), 0))
goto err_free;
ret = 1;
err_free:
EVP_CIPHER_CTX_free(ctx);
err:
return ret;
}
Fix no-dh and no-dsa Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/12525)
2020-07-24 19:04:00 +08:00
/*
* We're using some DH specific values in this test, so we skip compilation if
* we're in a no-dh build.
*/
#if !defined(OPENSSL_NO_DSA) && !defined(OPENSSL_NO_DH)
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
static int test_dsa_param_keygen(int tstid)
{
int ret = 0;
int expected;
EVP_PKEY_CTX *gen_ctx = NULL;
EVP_PKEY *pkey_parm = NULL;
Implement EVP_PKEY_dup() function Fixes #14501 Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14624)
2021-03-20 01:45:43 +08:00
EVP_PKEY *pkey = NULL, *dup_pk = NULL;
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
DSA *dsa = NULL;
int pind, qind, gind;
BIGNUM *p = NULL, *q = NULL, *g = NULL;
/*
* Just grab some fixed dh p, q, g values for testing,
* these 'safe primes' should not be used normally for dsa *.
*/
static const BIGNUM *bn[] = {
Fix external symbols for bn Partial fix for #12964 This adds ossl_ names for symbols related to bn_* Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14296)
2021-02-19 17:15:41 +08:00
&ossl_bignum_dh2048_256_p, &ossl_bignum_dh2048_256_q,
&ossl_bignum_dh2048_256_g
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
};
/*
* These tests are using bad values for p, q, g by reusing the values.
* A value of 0 uses p, 1 uses q and 2 uses g.
* There are 27 different combinations, with only the 1 valid combination.
*/
pind = tstid / 9;
qind = (tstid / 3) % 3;
gind = tstid % 3;
expected = (pind == 0 && qind == 1 && gind == 2);
TEST_note("Testing with (p, q, g) = (%s, %s, %s)\n", getname(pind),
getname(qind), getname(gind));
if (!TEST_ptr(pkey_parm = EVP_PKEY_new())
|| !TEST_ptr(dsa = DSA_new())
|| !TEST_ptr(p = BN_dup(bn[pind]))
|| !TEST_ptr(q = BN_dup(bn[qind]))
|| !TEST_ptr(g = BN_dup(bn[gind]))
|| !TEST_true(DSA_set0_pqg(dsa, p, q, g)))
goto err;
p = q = g = NULL;
if (!TEST_true(EVP_PKEY_assign_DSA(pkey_parm, dsa)))
goto err;
dsa = NULL;
if (!TEST_ptr(gen_ctx = EVP_PKEY_CTX_new_from_pkey(libctx, pkey_parm, NULL))
|| !TEST_int_gt(EVP_PKEY_keygen_init(gen_ctx), 0)
|| !TEST_int_eq(EVP_PKEY_keygen(gen_ctx, &pkey), expected))
goto err;
Implement EVP_PKEY_dup() function Fixes #14501 Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14624)
2021-03-20 01:45:43 +08:00
if (expected) {
if (!TEST_ptr(dup_pk = EVP_PKEY_dup(pkey))
|| !TEST_int_eq(EVP_PKEY_eq(pkey, dup_pk), 1))
goto err;
}
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
ret = 1;
err:
EVP_PKEY_free(pkey);
Implement EVP_PKEY_dup() function Fixes #14501 Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14624)
2021-03-20 01:45:43 +08:00
EVP_PKEY_free(dup_pk);
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
EVP_PKEY_CTX_free(gen_ctx);
EVP_PKEY_free(pkey_parm);
DSA_free(dsa);
BN_free(g);
BN_free(q);
BN_free(p);
return ret;
}
#endif /* OPENSSL_NO_DSA */
#ifndef OPENSSL_NO_DH
static int do_dh_param_keygen(int tstid, const BIGNUM **bn)
{
int ret = 0;
int expected;
EVP_PKEY_CTX *gen_ctx = NULL;
EVP_PKEY *pkey_parm = NULL;
Implement EVP_PKEY_dup() function Fixes #14501 Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14624)
2021-03-20 01:45:43 +08:00
EVP_PKEY *pkey = NULL, *dup_pk = NULL;
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
DH *dh = NULL;
int pind, qind, gind;
BIGNUM *p = NULL, *q = NULL, *g = NULL;
/*
* These tests are using bad values for p, q, g by reusing the values.
* A value of 0 uses p, 1 uses q and 2 uses g.
* There are 27 different combinations, with only the 1 valid combination.
*/
pind = tstid / 9;
qind = (tstid / 3) % 3;
gind = tstid % 3;
expected = (pind == 0 && qind == 1 && gind == 2);
TEST_note("Testing with (p, q, g) = (%s, %s, %s)", getname(pind),
getname(qind), getname(gind));
if (!TEST_ptr(pkey_parm = EVP_PKEY_new())
|| !TEST_ptr(dh = DH_new())
|| !TEST_ptr(p = BN_dup(bn[pind]))
|| !TEST_ptr(q = BN_dup(bn[qind]))
|| !TEST_ptr(g = BN_dup(bn[gind]))
|| !TEST_true(DH_set0_pqg(dh, p, q, g)))
goto err;
p = q = g = NULL;
if (!TEST_true(EVP_PKEY_assign_DH(pkey_parm, dh)))
goto err;
dh = NULL;
if (!TEST_ptr(gen_ctx = EVP_PKEY_CTX_new_from_pkey(libctx, pkey_parm, NULL))
|| !TEST_int_gt(EVP_PKEY_keygen_init(gen_ctx), 0)
|| !TEST_int_eq(EVP_PKEY_keygen(gen_ctx, &pkey), expected))
goto err;
Implement EVP_PKEY_dup() function Fixes #14501 Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14624)
2021-03-20 01:45:43 +08:00
if (expected) {
if (!TEST_ptr(dup_pk = EVP_PKEY_dup(pkey))
|| !TEST_int_eq(EVP_PKEY_eq(pkey, dup_pk), 1))
goto err;
}
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
ret = 1;
err:
EVP_PKEY_free(pkey);
Implement EVP_PKEY_dup() function Fixes #14501 Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14624)
2021-03-20 01:45:43 +08:00
EVP_PKEY_free(dup_pk);
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
EVP_PKEY_CTX_free(gen_ctx);
EVP_PKEY_free(pkey_parm);
DH_free(dh);
BN_free(g);
BN_free(q);
BN_free(p);
return ret;
}
/*
* Note that we get the fips186-4 path being run for most of these cases since
* the internal code will detect that the p, q, g does not match a safe prime
* group (Except for when tstid = 5, which sets the correct p, q, g)
*/
static int test_dh_safeprime_param_keygen(int tstid)
{
static const BIGNUM *bn[] = {
Fix external symbols for bn Partial fix for #12964 This adds ossl_ names for symbols related to bn_* Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14296)
2021-02-19 17:15:41 +08:00
&ossl_bignum_ffdhe2048_p, &ossl_bignum_ffdhe2048_q,
&ossl_bignum_const_2
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
};
return do_dh_param_keygen(tstid, bn);
}
Add dh_kdf support to provider Similiar to ecdh this supports the legacy kdf inside the provider dh key exchange. The supporting EVP_PKEY_CTX macros have been changed into mehtods and moved into dh_ctrl.c New kdfs such as SSKDF should be done as a seperate pass after doing the derive. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12575)
2020-08-04 09:21:21 +08:00
static int dhx_cert_load(void)
{
int ret = 0;
X509 *cert = NULL;
BIO *bio = NULL;
static const unsigned char dhx_cert[] = {
0x30,0x82,0x03,0xff,0x30,0x82,0x02,0xe7,0xa0,0x03,0x02,0x01,0x02,0x02,0x09,0x00,
0xdb,0xf5,0x4d,0x22,0xa0,0x7a,0x67,0xa6,0x30,0x0d,0x06,0x09,0x2a,0x86,0x48,0x86,
0xf7,0x0d,0x01,0x01,0x05,0x05,0x00,0x30,0x44,0x31,0x0b,0x30,0x09,0x06,0x03,0x55,
0x04,0x06,0x13,0x02,0x55,0x4b,0x31,0x16,0x30,0x14,0x06,0x03,0x55,0x04,0x0a,0x0c,
0x0d,0x4f,0x70,0x65,0x6e,0x53,0x53,0x4c,0x20,0x47,0x72,0x6f,0x75,0x70,0x31,0x1d,
0x30,0x1b,0x06,0x03,0x55,0x04,0x03,0x0c,0x14,0x54,0x65,0x73,0x74,0x20,0x53,0x2f,
0x4d,0x49,0x4d,0x45,0x20,0x52,0x53,0x41,0x20,0x52,0x6f,0x6f,0x74,0x30,0x1e,0x17,
0x0d,0x31,0x33,0x30,0x38,0x30,0x32,0x31,0x34,0x34,0x39,0x32,0x39,0x5a,0x17,0x0d,
0x32,0x33,0x30,0x36,0x31,0x31,0x31,0x34,0x34,0x39,0x32,0x39,0x5a,0x30,0x44,0x31,
0x0b,0x30,0x09,0x06,0x03,0x55,0x04,0x06,0x13,0x02,0x55,0x4b,0x31,0x16,0x30,0x14,
0x06,0x03,0x55,0x04,0x0a,0x0c,0x0d,0x4f,0x70,0x65,0x6e,0x53,0x53,0x4c,0x20,0x47,
0x72,0x6f,0x75,0x70,0x31,0x1d,0x30,0x1b,0x06,0x03,0x55,0x04,0x03,0x0c,0x14,0x54,
0x65,0x73,0x74,0x20,0x53,0x2f,0x4d,0x49,0x4d,0x45,0x20,0x45,0x45,0x20,0x44,0x48,
0x20,0x23,0x31,0x30,0x82,0x01,0xb6,0x30,0x82,0x01,0x2b,0x06,0x07,0x2a,0x86,0x48,
0xce,0x3e,0x02,0x01,0x30,0x82,0x01,0x1e,0x02,0x81,0x81,0x00,0xd4,0x0c,0x4a,0x0c,
0x04,0x72,0x71,0x19,0xdf,0x59,0x19,0xc5,0xaf,0x44,0x7f,0xca,0x8e,0x2b,0xf0,0x09,
0xf5,0xd3,0x25,0xb1,0x73,0x16,0x55,0x89,0xdf,0xfd,0x07,0xaf,0x19,0xd3,0x7f,0xd0,
0x07,0xa2,0xfe,0x3f,0x5a,0xf1,0x01,0xc6,0xf8,0x2b,0xef,0x4e,0x6d,0x03,0x38,0x42,
0xa1,0x37,0xd4,0x14,0xb4,0x00,0x4a,0xb1,0x86,0x5a,0x83,0xce,0xb9,0x08,0x0e,0xc1,
0x99,0x27,0x47,0x8d,0x0b,0x85,0xa8,0x82,0xed,0xcc,0x0d,0xb9,0xb0,0x32,0x7e,0xdf,
0xe8,0xe4,0xf6,0xf6,0xec,0xb3,0xee,0x7a,0x11,0x34,0x65,0x97,0xfc,0x1a,0xb0,0x95,
0x4b,0x19,0xb9,0xa6,0x1c,0xd9,0x01,0x32,0xf7,0x35,0x7c,0x2d,0x5d,0xfe,0xc1,0x85,
0x70,0x49,0xf8,0xcc,0x99,0xd0,0xbe,0xf1,0x5a,0x78,0xc8,0x03,0x02,0x81,0x80,0x69,
0x00,0xfd,0x66,0xf2,0xfc,0x15,0x8b,0x09,0xb8,0xdc,0x4d,0xea,0xaa,0x79,0x55,0xf9,
0xdf,0x46,0xa6,0x2f,0xca,0x2d,0x8f,0x59,0x2a,0xad,0x44,0xa3,0xc6,0x18,0x2f,0x95,
0xb6,0x16,0x20,0xe3,0xd3,0xd1,0x8f,0x03,0xce,0x71,0x7c,0xef,0x3a,0xc7,0x44,0x39,
0x0e,0xe2,0x1f,0xd8,0xd3,0x89,0x2b,0xe7,0x51,0xdc,0x12,0x48,0x4c,0x18,0x4d,0x99,
0x12,0x06,0xe4,0x17,0x02,0x03,0x8c,0x24,0x05,0x8e,0xa6,0x85,0xf2,0x69,0x1b,0xe1,
0x6a,0xdc,0xe2,0x04,0x3a,0x01,0x9d,0x64,0xbe,0xfe,0x45,0xf9,0x44,0x18,0x71,0xbd,
0x2d,0x3e,0x7a,0x6f,0x72,0x7d,0x1a,0x80,0x42,0x57,0xae,0x18,0x6f,0x91,0xd6,0x61,
0x03,0x8a,0x1c,0x89,0x73,0xc7,0x56,0x41,0x03,0xd3,0xf8,0xed,0x65,0xe2,0x85,0x02,
0x15,0x00,0x89,0x94,0xab,0x10,0x67,0x45,0x41,0xad,0x63,0xc6,0x71,0x40,0x8d,0x6b,
0x9e,0x19,0x5b,0xa4,0xc7,0xf5,0x03,0x81,0x84,0x00,0x02,0x81,0x80,0x2f,0x5b,0xde,
0x72,0x02,0x36,0x6b,0x00,0x5e,0x24,0x7f,0x14,0x2c,0x18,0x52,0x42,0x97,0x4b,0xdb,
0x6e,0x15,0x50,0x3c,0x45,0x3e,0x25,0xf3,0xb7,0xc5,0x6e,0xe5,0x52,0xe7,0xc4,0xfb,
0xf4,0xa5,0xf0,0x39,0x12,0x7f,0xbc,0x54,0x1c,0x93,0xb9,0x5e,0xee,0xe9,0x14,0xb0,
0xdf,0xfe,0xfc,0x36,0xe4,0xf2,0xaf,0xfb,0x13,0xc8,0xdf,0x18,0x94,0x1d,0x40,0xb9,
0x71,0xdd,0x4c,0x9c,0xa7,0x03,0x52,0x02,0xb5,0xed,0x71,0x80,0x3e,0x23,0xda,0x28,
0xe5,0xab,0xe7,0x6f,0xf2,0x0a,0x0e,0x00,0x5b,0x7d,0xc6,0x4b,0xd7,0xc7,0xb2,0xc3,
0xba,0x62,0x7f,0x70,0x28,0xa0,0x9d,0x71,0x13,0x70,0xd1,0x9f,0x32,0x2f,0x3e,0xd2,
0xcd,0x1b,0xa4,0xc6,0x72,0xa0,0x74,0x5d,0x71,0xef,0x03,0x43,0x6e,0xa3,0x60,0x30,
0x5e,0x30,0x0c,0x06,0x03,0x55,0x1d,0x13,0x01,0x01,0xff,0x04,0x02,0x30,0x00,0x30,
0x0e,0x06,0x03,0x55,0x1d,0x0f,0x01,0x01,0xff,0x04,0x04,0x03,0x02,0x05,0xe0,0x30,
0x1d,0x06,0x03,0x55,0x1d,0x0e,0x04,0x16,0x04,0x14,0x0b,0x5a,0x4d,0x5f,0x7d,0x25,
0xc7,0xf2,0x9d,0xc1,0xaa,0xb7,0x63,0x82,0x2f,0xfa,0x8f,0x32,0xe7,0xc0,0x30,0x1f,
0x06,0x03,0x55,0x1d,0x23,0x04,0x18,0x30,0x16,0x80,0x14,0xdf,0x7e,0x5e,0x88,0x05,
0x24,0x33,0x08,0xdd,0x22,0x81,0x02,0x97,0xcc,0x9a,0xb7,0xb1,0x33,0x27,0x30,0x30,
0x0d,0x06,0x09,0x2a,0x86,0x48,0x86,0xf7,0x0d,0x01,0x01,0x05,0x05,0x00,0x03,0x82,
0x01,0x01,0x00,0x5a,0xf2,0x63,0xef,0xd3,0x16,0xd7,0xf5,0xaa,0xdd,0x12,0x00,0x36,
0x00,0x21,0xa2,0x7b,0x08,0xd6,0x3b,0x9f,0x62,0xac,0x53,0x1f,0xed,0x4c,0xd1,0x15,
0x34,0x65,0x71,0xee,0x96,0x07,0xa6,0xef,0xb2,0xde,0xd8,0xbb,0x35,0x6e,0x2c,0xe2,
0xd1,0x26,0xef,0x7e,0x94,0xe2,0x88,0x51,0xa4,0x6c,0xaa,0x27,0x2a,0xd3,0xb6,0xc2,
0xf7,0xea,0xc3,0x0b,0xa9,0xb5,0x28,0x37,0xa2,0x63,0x08,0xe4,0x88,0xc0,0x1b,0x16,
0x1b,0xca,0xfd,0x8a,0x07,0x32,0x29,0xa7,0x53,0xb5,0x2d,0x30,0xe4,0xf5,0x16,0xc3,
0xe3,0xc2,0x4c,0x30,0x5d,0x35,0x80,0x1c,0xa2,0xdb,0xe3,0x4b,0x51,0x0d,0x4c,0x60,
0x5f,0xb9,0x46,0xac,0xa8,0x46,0xa7,0x32,0xa7,0x9c,0x76,0xf8,0xe9,0xb5,0x19,0xe2,
0x0c,0xe1,0x0f,0xc6,0x46,0xe2,0x38,0xa7,0x87,0x72,0x6d,0x6c,0xbc,0x88,0x2f,0x9d,
0x2d,0xe5,0xd0,0x7d,0x1e,0xc7,0x5d,0xf8,0x7e,0xb4,0x0b,0xa6,0xf9,0x6c,0xe3,0x7c,
0xb2,0x70,0x6e,0x75,0x9b,0x1e,0x63,0xe1,0x4d,0xb2,0x81,0xd3,0x55,0x38,0x94,0x1a,
0x7a,0xfa,0xbf,0x01,0x18,0x70,0x2d,0x35,0xd3,0xe3,0x10,0x7a,0x9a,0xa7,0x8f,0xf3,
0xbd,0x56,0x55,0x5e,0xd8,0xbd,0x4e,0x16,0x76,0xd0,0x48,0x4c,0xf9,0x51,0x54,0xdf,
0x2d,0xb0,0xc9,0xaa,0x5e,0x42,0x38,0x50,0xbf,0x0f,0xc0,0xd9,0x84,0x44,0x4b,0x42,
0x24,0xec,0x14,0xa3,0xde,0x11,0xdf,0x58,0x7f,0xc2,0x4d,0xb2,0xd5,0x42,0x78,0x6e,
0x52,0x3e,0xad,0xc3,0x5f,0x04,0xc4,0xe6,0x31,0xaa,0x81,0x06,0x8b,0x13,0x4b,0x3c,
0x0e,0x6a,0xb1
};
if (!TEST_ptr(bio = BIO_new_mem_buf(dhx_cert, sizeof(dhx_cert)))
Run the withlibctx.pl script Automatically rename all instances of _with_libctx() to _ex() as per our coding style. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/12970)
2020-09-24 17:42:23 +08:00
|| !TEST_ptr(cert = X509_new_ex(libctx, NULL))
Add dh_kdf support to provider Similiar to ecdh this supports the legacy kdf inside the provider dh key exchange. The supporting EVP_PKEY_CTX macros have been changed into mehtods and moved into dh_ctrl.c New kdfs such as SSKDF should be done as a seperate pass after doing the derive. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12575)
2020-08-04 09:21:21 +08:00
|| !TEST_ptr(d2i_X509_bio(bio, &cert)))
goto err;
ret = 1;
err:
X509_free(cert);
BIO_free(bio);
return ret;
}
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
#endif /* OPENSSL_NO_DH */
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
static int test_cipher_reinit(int test_id)
{
Add special case to skip RC4 reinit Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/15276)
2021-05-14 12:46:25 +08:00
int ret = 0, diff, ccm, siv, no_null_key;
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
int out1_len = 0, out2_len = 0, out3_len = 0;
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
EVP_CIPHER *cipher = NULL;
EVP_CIPHER_CTX *ctx = NULL;
unsigned char out1[256];
unsigned char out2[256];
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
unsigned char out3[256];
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
unsigned char in[16] = {
0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08,
0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x10
};
unsigned char key[64] = {
0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0x01, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0x02, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0x03, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
};
Add support for integrity-only cipher suites for TLS v1.3 - add test vectors for tls1_3 integrity-only ciphers - recmethod_local.h: add new member for MAC - tls13_meth.c: add MAC only to tls 1.3 - tls13_enc.c: extend function to add MAC only - ssl_local.h: add ssl_cipher_get_evp_md_mac() - s3_lib.c: add the new ciphers and add #ifndef OPENSSL_NO_INTEGRITY_ONLY_CIPHERS - ssl_ciph.c : add ssl_cipher_get_evp_md_mac() and use it - tls13secretstest.c: add dummy test function - Configure: add integrity-only-ciphers option - document the new ciphers Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22903)
2023-12-01 19:47:07 +08:00
unsigned char iv[48] = {
0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08,
0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00,
0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08,
0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00,
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08,
0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00
};
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
const char *name = sk_OPENSSL_STRING_value(cipher_names, test_id);
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
if (!TEST_ptr(ctx = EVP_CIPHER_CTX_new()))
goto err;
TEST_note("Fetching %s\n", name);
if (!TEST_ptr(cipher = EVP_CIPHER_fetch(libctx, name, NULL)))
goto err;
/* ccm fails on the second update - this matches OpenSSL 1_1_1 behaviour */
Rename all getters to use get/get0 in name For functions that exist in 1.1.1 provide a simple aliases via #define. Fixes #15236 Functions with OSSL_DECODER_, OSSL_ENCODER_, OSSL_STORE_LOADER_, EVP_KEYEXCH_, EVP_KEM_, EVP_ASYM_CIPHER_, EVP_SIGNATURE_, EVP_KEYMGMT_, EVP_RAND_, EVP_MAC_, EVP_KDF_, EVP_PKEY_, EVP_MD_, and EVP_CIPHER_ prefixes are renamed. Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15405)
2021-05-21 22:58:08 +08:00
ccm = (EVP_CIPHER_get_mode(cipher) == EVP_CIPH_CCM_MODE);
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
/* siv cannot be called with NULL key as the iv is irrelevant */
Rename all getters to use get/get0 in name For functions that exist in 1.1.1 provide a simple aliases via #define. Fixes #15236 Functions with OSSL_DECODER_, OSSL_ENCODER_, OSSL_STORE_LOADER_, EVP_KEYEXCH_, EVP_KEM_, EVP_ASYM_CIPHER_, EVP_SIGNATURE_, EVP_KEYMGMT_, EVP_RAND_, EVP_MAC_, EVP_KDF_, EVP_PKEY_, EVP_MD_, and EVP_CIPHER_ prefixes are renamed. Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15405)
2021-05-21 22:58:08 +08:00
siv = (EVP_CIPHER_get_mode(cipher) == EVP_CIPH_SIV_MODE);
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
Add special case to skip RC4 reinit Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/15276)
2021-05-14 12:46:25 +08:00
/*
* Skip init call with a null key for RC4 as the stream cipher does not
* handle reinit (1.1.1 behaviour).
*/
no_null_key = EVP_CIPHER_is_a(cipher, "RC4")
|| EVP_CIPHER_is_a(cipher, "RC4-40")
|| EVP_CIPHER_is_a(cipher, "RC4-HMAC-MD5");
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
/* DES3-WRAP uses random every update - so it will give a different value */
diff = EVP_CIPHER_is_a(cipher, "DES3-WRAP");
if (!TEST_true(EVP_EncryptInit_ex(ctx, cipher, NULL, key, iv))
|| !TEST_true(EVP_EncryptUpdate(ctx, out1, &out1_len, in, sizeof(in)))
|| !TEST_true(EVP_EncryptInit_ex(ctx, NULL, NULL, key, iv))
|| !TEST_int_eq(EVP_EncryptUpdate(ctx, out2, &out2_len, in, sizeof(in)),
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
ccm ? 0 : 1)
Add special case to skip RC4 reinit Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/15276)
2021-05-14 12:46:25 +08:00
|| (!no_null_key
&& (!TEST_true(EVP_EncryptInit_ex(ctx, NULL, NULL, NULL, iv))
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
|| !TEST_int_eq(EVP_EncryptUpdate(ctx, out3, &out3_len, in, sizeof(in)),
Add special case to skip RC4 reinit Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/15276)
2021-05-14 12:46:25 +08:00
ccm || siv ? 0 : 1))))
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
goto err;
if (ccm == 0) {
if (diff) {
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
if (!TEST_mem_ne(out1, out1_len, out2, out2_len)
|| !TEST_mem_ne(out1, out1_len, out3, out3_len)
|| !TEST_mem_ne(out2, out2_len, out3, out3_len))
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
goto err;
} else {
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
if (!TEST_mem_eq(out1, out1_len, out2, out2_len)
Add special case to skip RC4 reinit Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/15276)
2021-05-14 12:46:25 +08:00
|| (!siv && !no_null_key && !TEST_mem_eq(out1, out1_len, out3, out3_len)))
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
goto err;
}
}
ret = 1;
err:
EVP_CIPHER_free(cipher);
EVP_CIPHER_CTX_free(ctx);
return ret;
}
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
/*
* This test only uses a partial block (half the block size) of input for each
* EVP_EncryptUpdate() in order to test that the second init/update is not using
* a leftover buffer from the first init/update.
* Note: some ciphers don't need a full block to produce output.
*/
static int test_cipher_reinit_partialupdate(int test_id)
{
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
int ret = 0, in_len;
int out1_len = 0, out2_len = 0, out3_len = 0;
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
EVP_CIPHER *cipher = NULL;
EVP_CIPHER_CTX *ctx = NULL;
unsigned char out1[256];
unsigned char out2[256];
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
unsigned char out3[256];
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
static const unsigned char in[32] = {
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0xba, 0xbe, 0xba, 0xbe, 0x00, 0x00, 0xba, 0xbe,
0x01, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
};
static const unsigned char key[64] = {
0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0x01, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0x02, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
0x03, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
};
Add support for integrity-only cipher suites for TLS v1.3 - add test vectors for tls1_3 integrity-only ciphers - recmethod_local.h: add new member for MAC - tls13_meth.c: add MAC only to tls 1.3 - tls13_enc.c: extend function to add MAC only - ssl_local.h: add ssl_cipher_get_evp_md_mac() - s3_lib.c: add the new ciphers and add #ifndef OPENSSL_NO_INTEGRITY_ONLY_CIPHERS - ssl_ciph.c : add ssl_cipher_get_evp_md_mac() and use it - tls13secretstest.c: add dummy test function - Configure: add integrity-only-ciphers option - document the new ciphers Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22903)
2023-12-01 19:47:07 +08:00
static const unsigned char iv[48] = {
0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08,
0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00,
0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08,
0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00,
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
0x0f, 0x0e, 0x0d, 0x0c, 0x0b, 0x0a, 0x09, 0x08,
0x07, 0x06, 0x05, 0x04, 0x03, 0x02, 0x01, 0x00
};
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
const char *name = sk_OPENSSL_STRING_value(cipher_names, test_id);
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
if (!TEST_ptr(ctx = EVP_CIPHER_CTX_new()))
goto err;
TEST_note("Fetching %s\n", name);
if (!TEST_ptr(cipher = EVP_CIPHER_fetch(libctx, name, NULL)))
goto err;
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
in_len = EVP_CIPHER_get_block_size(cipher);
if (!TEST_int_gt(in_len, 0))
goto err;
if (in_len > 1)
in_len /= 2;
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
/* skip any ciphers that don't allow partial updates */
Rename all getters to use get/get0 in name For functions that exist in 1.1.1 provide a simple aliases via #define. Fixes #15236 Functions with OSSL_DECODER_, OSSL_ENCODER_, OSSL_STORE_LOADER_, EVP_KEYEXCH_, EVP_KEM_, EVP_ASYM_CIPHER_, EVP_SIGNATURE_, EVP_KEYMGMT_, EVP_RAND_, EVP_MAC_, EVP_KDF_, EVP_PKEY_, EVP_MD_, and EVP_CIPHER_ prefixes are renamed. Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15405)
2021-05-21 22:58:08 +08:00
if (((EVP_CIPHER_get_flags(cipher)
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
& (EVP_CIPH_FLAG_CTS | EVP_CIPH_FLAG_TLS1_1_MULTIBLOCK)) != 0)
Rename all getters to use get/get0 in name For functions that exist in 1.1.1 provide a simple aliases via #define. Fixes #15236 Functions with OSSL_DECODER_, OSSL_ENCODER_, OSSL_STORE_LOADER_, EVP_KEYEXCH_, EVP_KEM_, EVP_ASYM_CIPHER_, EVP_SIGNATURE_, EVP_KEYMGMT_, EVP_RAND_, EVP_MAC_, EVP_KDF_, EVP_PKEY_, EVP_MD_, and EVP_CIPHER_ prefixes are renamed. Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15405)
2021-05-21 22:58:08 +08:00
|| EVP_CIPHER_get_mode(cipher) == EVP_CIPH_CCM_MODE
|| EVP_CIPHER_get_mode(cipher) == EVP_CIPH_XTS_MODE
|| EVP_CIPHER_get_mode(cipher) == EVP_CIPH_WRAP_MODE) {
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
ret = 1;
goto err;
}
if (!TEST_true(EVP_EncryptInit_ex(ctx, cipher, NULL, key, iv))
|| !TEST_true(EVP_EncryptUpdate(ctx, out1, &out1_len, in, in_len))
|| !TEST_true(EVP_EncryptInit_ex(ctx, NULL, NULL, key, iv))
|| !TEST_true(EVP_EncryptUpdate(ctx, out2, &out2_len, in, in_len)))
goto err;
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
if (EVP_CIPHER_get_iv_length(cipher) != 0)
if (!TEST_mem_eq(out1, out1_len, out2, out2_len))
goto err;
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
Rename all getters to use get/get0 in name For functions that exist in 1.1.1 provide a simple aliases via #define. Fixes #15236 Functions with OSSL_DECODER_, OSSL_ENCODER_, OSSL_STORE_LOADER_, EVP_KEYEXCH_, EVP_KEM_, EVP_ASYM_CIPHER_, EVP_SIGNATURE_, EVP_KEYMGMT_, EVP_RAND_, EVP_MAC_, EVP_KDF_, EVP_PKEY_, EVP_MD_, and EVP_CIPHER_ prefixes are renamed. Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15405)
2021-05-21 22:58:08 +08:00
if (EVP_CIPHER_get_mode(cipher) != EVP_CIPH_SIV_MODE) {
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
if (!TEST_true(EVP_EncryptInit_ex(ctx, NULL, NULL, NULL, iv))
|| !TEST_true(EVP_EncryptUpdate(ctx, out3, &out3_len, in, in_len)))
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
goto err;
chacha20: Properly reinitialize the cipher context with NULL key Same for chacha20-poly1305. The test_cipher_reinit and test_cipher_reinit_partialupdate is modified to test this case of cipher context reinitialization. Fixes #13064 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13850)
2021-01-12 23:53:33 +08:00
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
if (EVP_CIPHER_get_iv_length(cipher) != 0)
if (!TEST_mem_eq(out1, out1_len, out3, out3_len))
goto err;
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
}
ret = 1;
err:
EVP_CIPHER_free(cipher);
EVP_CIPHER_CTX_free(ctx);
return ret;
}
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
static int name_cmp(const char * const *a, const char * const *b)
{
str[n]casecmp => OPENSSL_strncasecmp Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/18069)
2022-04-12 18:30:08 +08:00
return OPENSSL_strcasecmp(*a, *b);
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
}
static void collect_cipher_names(EVP_CIPHER *cipher, void *cipher_names_list)
{
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
STACK_OF(OPENSSL_STRING) *names = cipher_names_list;
Rename all getters to use get/get0 in name For functions that exist in 1.1.1 provide a simple aliases via #define. Fixes #15236 Functions with OSSL_DECODER_, OSSL_ENCODER_, OSSL_STORE_LOADER_, EVP_KEYEXCH_, EVP_KEM_, EVP_ASYM_CIPHER_, EVP_SIGNATURE_, EVP_KEYMGMT_, EVP_RAND_, EVP_MAC_, EVP_KDF_, EVP_PKEY_, EVP_MD_, and EVP_CIPHER_ prefixes are renamed. Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15405)
2021-05-21 22:58:08 +08:00
const char *name = EVP_CIPHER_get0_name(cipher);
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
char *namedup = NULL;
Add FIPS indicator support for Triple-DES encryption. This leaves 3DES with the FIPS query "FIPS=yes", which allows Triple-DES to be used for Decryption by default. Disallow CMAC using Triple-DES in FIPS. This does not use a FIPS indicator. Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Paul Dale <ppzgs1@gmail.com> (Merged from https://github.com/openssl/openssl/pull/24960)
2024-07-22 15:24:53 +08:00
/* Skip Triple-DES encryption operations in FIPS mode */
if (OSSL_PROVIDER_available(libctx, "fips")
&& strncmp(name, "DES", 3) == 0)
return;
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
assert(name != NULL);
/* the cipher will be freed after returning, strdup is needed */
if ((namedup = OPENSSL_strdup(name)) != NULL
&& !sk_OPENSSL_STRING_push(names, namedup))
OPENSSL_free(namedup);
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
}
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
static int rsa_keygen(int bits, EVP_PKEY **pub, EVP_PKEY **priv)
{
int ret = 0;
unsigned char *pub_der = NULL;
const unsigned char *pp = NULL;
test/evp_libctx_test.c: use OSSL_ENCODER instead of i2d_PublicKey() Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13167)
2020-10-17 17:44:43 +08:00
size_t len = 0;
OSSL_ENCODER_CTX *ectx = NULL;
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
Fix uses of `EVP_PKEY_Q_keygen` with `size_t` variadic argument Fix cases where `int` argument was passed instead of `size_t`. CLA: trivial Reviewed-by: Richard Levitte <levitte@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25857)
2024-11-04 16:56:52 +08:00
if (!TEST_ptr(*priv = EVP_PKEY_Q_keygen(libctx, NULL, "RSA", (size_t)bits))
test/evp_libctx_test.c: use OSSL_ENCODER instead of i2d_PublicKey() Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13167)
2020-10-17 17:44:43 +08:00
|| !TEST_ptr(ectx =
Rename OSSL_ENCODER_CTX_new_by_EVP_PKEY and OSSL_DECODER_CTX_new_by_EVP_PKEY Additional renames done in encoder and decoder implementation to follow the style. Fixes #13622 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14155)
2021-02-11 23:57:37 +08:00
OSSL_ENCODER_CTX_new_for_pkey(*priv,
EVP_PKEY_PUBLIC_KEY,
"DER", "type-specific",
NULL))
test/evp_libctx_test.c: use OSSL_ENCODER instead of i2d_PublicKey() Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13167)
2020-10-17 17:44:43 +08:00
|| !TEST_true(OSSL_ENCODER_to_data(ectx, &pub_der, &len)))
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
goto err;
pp = pub_der;
if (!TEST_ptr(d2i_PublicKey(EVP_PKEY_RSA, pub, &pp, len)))
goto err;
ret = 1;
err:
test/evp_libctx_test.c: use OSSL_ENCODER instead of i2d_PublicKey() Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13167)
2020-10-17 17:44:43 +08:00
OSSL_ENCODER_CTX_free(ectx);
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
OPENSSL_free(pub_der);
return ret;
}
static int kem_rsa_gen_recover(void)
{
int ret = 0;
EVP_PKEY *pub = NULL;
EVP_PKEY *priv = NULL;
Add KEM dupctx test Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/16495)
2021-09-02 14:50:45 +08:00
EVP_PKEY_CTX *sctx = NULL, *rctx = NULL, *dctx = NULL;
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
unsigned char secret[256] = { 0, };
unsigned char ct[256] = { 0, };
unsigned char unwrap[256] = { 0, };
size_t ctlen = 0, unwraplen = 0, secretlen = 0;
TEST: Prefer using precomputed RSA and DH keys for more efficient tests Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13715)
2020-12-13 05:04:05 +08:00
int bits = 2048;
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
TEST: Prefer using precomputed RSA and DH keys for more efficient tests Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13715)
2020-12-13 05:04:05 +08:00
ret = TEST_true(rsa_keygen(bits, &pub, &priv))
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
&& TEST_ptr(sctx = EVP_PKEY_CTX_new_from_pkey(libctx, pub, NULL))
test: support params arguments to init functions Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/14383)
2021-03-02 20:41:24 +08:00
&& TEST_int_eq(EVP_PKEY_encapsulate_init(sctx, NULL), 1)
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(sctx, "RSASVE"), 1)
Add KEM dupctx test Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/16495)
2021-09-02 14:50:45 +08:00
&& TEST_ptr(dctx = EVP_PKEY_CTX_dup(sctx))
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
/* Test that providing a NULL wrappedlen fails */
&& TEST_int_eq(EVP_PKEY_encapsulate(dctx, NULL, NULL, NULL, NULL), 0)
Add KEM dupctx test Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/16495)
2021-09-02 14:50:45 +08:00
&& TEST_int_eq(EVP_PKEY_encapsulate(dctx, NULL, &ctlen, NULL,
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
&secretlen), 1)
&& TEST_int_eq(ctlen, secretlen)
TEST: Prefer using precomputed RSA and DH keys for more efficient tests Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/13715)
2020-12-13 05:04:05 +08:00
&& TEST_int_eq(ctlen, bits / 8)
Add KEM dupctx test Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/16495)
2021-09-02 14:50:45 +08:00
&& TEST_int_eq(EVP_PKEY_encapsulate(dctx, ct, &ctlen, secret,
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
&secretlen), 1)
&& TEST_ptr(rctx = EVP_PKEY_CTX_new_from_pkey(libctx, priv, NULL))
test: support params arguments to init functions Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/14383)
2021-03-02 20:41:24 +08:00
&& TEST_int_eq(EVP_PKEY_decapsulate_init(rctx, NULL), 1)
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(rctx, "RSASVE"), 1)
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
/* Test that providing a NULL unwrappedlen fails */
&& TEST_int_eq(EVP_PKEY_decapsulate(rctx, NULL, NULL, ct, ctlen), 0)
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
&& TEST_int_eq(EVP_PKEY_decapsulate(rctx, NULL, &unwraplen,
ct, ctlen), 1)
&& TEST_int_eq(EVP_PKEY_decapsulate(rctx, unwrap, &unwraplen,
ct, ctlen), 1)
&& TEST_mem_eq(unwrap, unwraplen, secret, secretlen);
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
/* Test that providing a too short unwrapped/ctlen fails */
evp_libctx_test: fix provider compat CI regression The regression was introduced by #25522. Fixes #25632 Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25633)
2024-10-08 07:38:26 +08:00
if (fips_provider_version_match(libctx, ">=3.4.0")) {
ctlen = 1;
if (!TEST_int_eq(EVP_PKEY_encapsulate(dctx, ct, &ctlen, secret,
&secretlen), 0))
ret = 0;
unwraplen = 1;
if (!TEST_int_eq(EVP_PKEY_decapsulate(rctx, unwrap, &unwraplen, ct,
ctlen), 0))
ret = 0;
}
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
EVP_PKEY_free(pub);
EVP_PKEY_free(priv);
EVP_PKEY_CTX_free(rctx);
Add KEM dupctx test Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/16495)
2021-09-02 14:50:45 +08:00
EVP_PKEY_CTX_free(dctx);
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
EVP_PKEY_CTX_free(sctx);
return ret;
}
Add a gettable for provider ciphers to return the EVP_CIPH_RAND_KEY flag Fixes #15531 DES and TDES set this flag which could possibly be used by applications. The gettable cipher param OSSL_CIPHER_PARAM_HAS_RAND_KEY has been added. Note that EVP_CIPHER_CTX_rand_key() uses this flag. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15606)
2021-06-03 17:09:38 +08:00
#ifndef OPENSSL_NO_DES
/*
* This test makes sure that EVP_CIPHER_CTX_rand_key() works correctly
* For fips mode this code would produce an error if the flag is not set.
*/
static int test_cipher_tdes_randkey(void)
{
int ret;
EVP_CIPHER_CTX *ctx = NULL;
EVP_CIPHER *tdes_cipher = NULL, *aes_cipher = NULL;
unsigned char key[24] = { 0 };
Add FIPS indicator support for Triple-DES encryption. This leaves 3DES with the FIPS query "FIPS=yes", which allows Triple-DES to be used for Decryption by default. Disallow CMAC using Triple-DES in FIPS. This does not use a FIPS indicator. Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Paul Dale <ppzgs1@gmail.com> (Merged from https://github.com/openssl/openssl/pull/24960)
2024-07-22 15:24:53 +08:00
OSSL_PARAM params[2];
int check = 0;
Add a gettable for provider ciphers to return the EVP_CIPH_RAND_KEY flag Fixes #15531 DES and TDES set this flag which could possibly be used by applications. The gettable cipher param OSSL_CIPHER_PARAM_HAS_RAND_KEY has been added. Note that EVP_CIPHER_CTX_rand_key() uses this flag. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15606)
2021-06-03 17:09:38 +08:00
Add FIPS indicator support for Triple-DES encryption. This leaves 3DES with the FIPS query "FIPS=yes", which allows Triple-DES to be used for Decryption by default. Disallow CMAC using Triple-DES in FIPS. This does not use a FIPS indicator. Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Paul Dale <ppzgs1@gmail.com> (Merged from https://github.com/openssl/openssl/pull/24960)
2024-07-22 15:24:53 +08:00
params[0] = OSSL_PARAM_construct_int("encrypt-check", &check);
params[1] = OSSL_PARAM_construct_end();
Add a gettable for provider ciphers to return the EVP_CIPH_RAND_KEY flag Fixes #15531 DES and TDES set this flag which could possibly be used by applications. The gettable cipher param OSSL_CIPHER_PARAM_HAS_RAND_KEY has been added. Note that EVP_CIPHER_CTX_rand_key() uses this flag. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15606)
2021-06-03 17:09:38 +08:00
ret = TEST_ptr(aes_cipher = EVP_CIPHER_fetch(libctx, "AES-256-CBC", NULL))
&& TEST_int_eq(EVP_CIPHER_get_flags(aes_cipher) & EVP_CIPH_RAND_KEY, 0)
&& TEST_ptr(tdes_cipher = EVP_CIPHER_fetch(libctx, "DES-EDE3-CBC", NULL))
&& TEST_int_ne(EVP_CIPHER_get_flags(tdes_cipher) & EVP_CIPH_RAND_KEY, 0)
&& TEST_ptr(ctx = EVP_CIPHER_CTX_new())
Add FIPS indicator support for Triple-DES encryption. This leaves 3DES with the FIPS query "FIPS=yes", which allows Triple-DES to be used for Decryption by default. Disallow CMAC using Triple-DES in FIPS. This does not use a FIPS indicator. Reviewed-by: Neil Horman <nhorman@openssl.org> Reviewed-by: Paul Dale <ppzgs1@gmail.com> (Merged from https://github.com/openssl/openssl/pull/24960)
2024-07-22 15:24:53 +08:00
&& TEST_true(EVP_CipherInit_ex2(ctx, tdes_cipher, NULL, NULL, 1,
params))
Fix the incorrect checks of EVP_CIPHER_CTX_rand_key Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/18397)
2022-05-24 22:59:41 +08:00
&& TEST_int_gt(EVP_CIPHER_CTX_rand_key(ctx, key), 0);
Add a gettable for provider ciphers to return the EVP_CIPH_RAND_KEY flag Fixes #15531 DES and TDES set this flag which could possibly be used by applications. The gettable cipher param OSSL_CIPHER_PARAM_HAS_RAND_KEY has been added. Note that EVP_CIPHER_CTX_rand_key() uses this flag. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15606)
2021-06-03 17:09:38 +08:00
EVP_CIPHER_CTX_free(ctx);
EVP_CIPHER_free(tdes_cipher);
EVP_CIPHER_free(aes_cipher);
return ret;
}
#endif /* OPENSSL_NO_DES */
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
static int kem_rsa_params(void)
{
int ret = 0;
EVP_PKEY *pub = NULL;
EVP_PKEY *priv = NULL;
EVP_PKEY_CTX *pubctx = NULL, *privctx = NULL;
unsigned char secret[256] = { 0, };
unsigned char ct[256] = { 0, };
size_t ctlen = 0, secretlen = 0;
ret = TEST_true(rsa_keygen(2048, &pub, &priv))
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
&& TEST_ptr(pubctx = EVP_PKEY_CTX_new_from_pkey(libctx, pub, NULL))
&& TEST_ptr(privctx = EVP_PKEY_CTX_new_from_pkey(libctx, priv, NULL))
/* Test setting kem op before the init fails */
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(pubctx, "RSASVE"), -2)
/* Test NULL ctx passed */
&& TEST_int_eq(EVP_PKEY_encapsulate_init(NULL, NULL), 0)
&& TEST_int_eq(EVP_PKEY_encapsulate(NULL, NULL, NULL, NULL, NULL), 0)
&& TEST_int_eq(EVP_PKEY_decapsulate_init(NULL, NULL), 0)
&& TEST_int_eq(EVP_PKEY_decapsulate(NULL, NULL, NULL, NULL, 0), 0)
/* Test Invalid operation */
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, NULL, NULL, NULL), -1)
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, NULL, NULL, NULL, 0), 0)
/* Wrong key component - no secret should be returned on failure */
&& TEST_int_eq(EVP_PKEY_decapsulate_init(pubctx, NULL), 1)
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(pubctx, "RSASVE"), 1)
&& TEST_int_eq(EVP_PKEY_decapsulate(pubctx, secret, &secretlen, ct,
sizeof(ct)), 0)
&& TEST_uchar_eq(secret[0], 0)
Older FIPS providers require a kemop Don't expect success with dated FIPS modules. Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/26904)
2025-02-26 08:43:35 +08:00
/* Unless older FIPS, test encapsulate succeeds even if the mode is not set */
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
&& TEST_int_eq(EVP_PKEY_encapsulate_init(pubctx, NULL), 1)
Older FIPS providers require a kemop Don't expect success with dated FIPS modules. Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/26904)
2025-02-26 08:43:35 +08:00
&& (is_fips_lt_3_5 ||
(TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, &ctlen, NULL, &secretlen), 1)
&& TEST_true(ctlen <= sizeof(ct))
&& TEST_true(secretlen <= sizeof(secret))
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, ct, &ctlen, secret, &secretlen), 1)))
Adjust tests to fetch the output len for EVP_PKEY_[en|de]cap Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Viktor Dukhovni <viktor@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/25522)
2024-09-25 01:54:14 +08:00
/* Test setting a bad kem ops fail */
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(pubctx, "RSA"), 0)
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(pubctx, NULL), 0)
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(NULL, "RSASVE"), 0)
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(NULL, NULL), 0)
/* Test secretlen is optional */
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(pubctx, "RSASVE"), 1)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, &ctlen, NULL, NULL), 1)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, ct, &ctlen, secret, NULL), 1)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, &ctlen, NULL, NULL), 1)
/* Test outlen is optional */
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, NULL, NULL, &secretlen), 1)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, ct, NULL, secret, &secretlen), 1)
/* test that either len must be set if out is NULL */
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, NULL, NULL, NULL), 0)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, &ctlen, NULL, NULL), 1)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, NULL, NULL, &secretlen), 1)
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, NULL, &ctlen, NULL, &secretlen), 1)
/* Secret buffer should be set if there is an output buffer */
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, ct, &ctlen, NULL, NULL), 0)
/* Test that lengths are optional if ct is not NULL */
&& TEST_int_eq(EVP_PKEY_encapsulate(pubctx, ct, NULL, secret, NULL), 1)
/* Pass if secret or secret length are not NULL */
&& TEST_int_eq(EVP_PKEY_decapsulate_init(privctx, NULL), 1)
&& TEST_int_eq(EVP_PKEY_CTX_set_kem_op(privctx, "RSASVE"), 1)
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, secret, NULL, ct, sizeof(ct)), 1)
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, NULL, &secretlen, ct, sizeof(ct)), 1)
&& TEST_int_eq(secretlen, 256)
/* Fail if passed NULL arguments */
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, NULL, NULL, ct, sizeof(ct)), 0)
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, secret, &secretlen, NULL, 0), 0)
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, secret, &secretlen, NULL, sizeof(ct)), 0)
&& TEST_int_eq(EVP_PKEY_decapsulate(privctx, secret, &secretlen, ct, 0), 0);
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
EVP_PKEY_free(pub);
EVP_PKEY_free(priv);
EVP_PKEY_CTX_free(pubctx);
EVP_PKEY_CTX_free(privctx);
return ret;
}
#ifndef OPENSSL_NO_DH
static EVP_PKEY *gen_dh_key(void)
{
EVP_PKEY_CTX *gctx = NULL;
EVP_PKEY *pkey = NULL;
OSSL_PARAM params[2];
params[0] = OSSL_PARAM_construct_utf8_string("group", "ffdhe2048", 0);
params[1] = OSSL_PARAM_construct_end();
if (!TEST_ptr(gctx = EVP_PKEY_CTX_new_from_name(libctx, "DH", NULL))
EVP_PKEY_keygen_init: fix return check Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17031)
2021-11-15 00:05:04 +08:00
|| !TEST_int_gt(EVP_PKEY_keygen_init(gctx), 0)
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
|| !TEST_true(EVP_PKEY_CTX_set_params(gctx, params))
|| !TEST_true(EVP_PKEY_keygen(gctx, &pkey)))
goto err;
err:
EVP_PKEY_CTX_free(gctx);
return pkey;
}
/* Fail if we try to use a dh key */
static int kem_invalid_keytype(void)
{
int ret = 0;
EVP_PKEY *key = NULL;
EVP_PKEY_CTX *sctx = NULL;
if (!TEST_ptr(key = gen_dh_key()))
goto done;
if (!TEST_ptr(sctx = EVP_PKEY_CTX_new_from_pkey(libctx, key, NULL)))
goto done;
test: support params arguments to init functions Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/14383)
2021-03-02 20:41:24 +08:00
if (!TEST_int_eq(EVP_PKEY_encapsulate_init(sctx, NULL), -2))
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
goto done;
ret = 1;
done:
EVP_PKEY_free(key);
EVP_PKEY_CTX_free(sctx);
return ret;
}
#endif /* OPENSSL_NO_DH */
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
int setup_tests(void)
{
const char *prov_name = "default";
char *config_file = NULL;
OPTION_CHOICE o;
while ((o = opt_next()) != OPT_EOF) {
switch (o) {
case OPT_PROVIDER_NAME:
prov_name = opt_arg();
break;
case OPT_CONFIG_FILE:
config_file = opt_arg();
break;
case OPT_TEST_CASES:
break;
default:
case OPT_ERR:
return 0;
}
}
Use adapted test_get_libctx() for simpler test setup and better error reporting Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/13001)
2020-09-26 21:21:48 +08:00
if (!test_get_libctx(&libctx, &nullprov, config_file, &libprov, prov_name))
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
return 0;
Add appropriate NULL checks in EVP_CIPHER api The EVP_CIPHER api currently assumes that calls made into several APIs have already initalized the cipher in a given context via a call to EVP_CipherInit[_ex[2]]. If that hasnt been done, instead of an error, the result is typically a SIGSEGV. Correct that by adding missing NULL checks in the apropriate apis prior to using ctx->cipher Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/22995)
2023-12-10 02:40:01 +08:00
ADD_TEST(test_evp_cipher_api_safety);
Older FIPS providers require a kemop Don't expect success with dated FIPS modules. Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/26904)
2025-02-26 08:43:35 +08:00
if (strcmp(prov_name, "fips") == 0)
is_fips = 1;
is_fips_lt_3_5 = is_fips && fips_provider_version_lt(libctx, 3, 5, 0);
Fix no-dh and no-dsa Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/12525)
2020-07-24 19:04:00 +08:00
#if !defined(OPENSSL_NO_DSA) && !defined(OPENSSL_NO_DH)
Older FIPS providers require a kemop Don't expect success with dated FIPS modules. Reviewed-by: Tim Hudson <tjh@openssl.org> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/26904)
2025-02-26 08:43:35 +08:00
if (!is_fips || fips_provider_version_lt(libctx, 3, 4, 0))
Add FIPS DSA Keygen tests Adjust the existing tests to disable DSA keygen in FIPS mode. Allow evp_test to load DSA 'KeyParams' that can then be used to perform a DSA KeyGen. Reviewed-by: Paul Dale <ppzgs1@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/24978)
2024-07-24 15:24:33 +08:00
ADD_ALL_TESTS(test_dsa_param_keygen, 3 * 3 * 3);
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
#endif
#ifndef OPENSSL_NO_DH
ADD_ALL_TESTS(test_dh_safeprime_param_keygen, 3 * 3 * 3);
Add dh_kdf support to provider Similiar to ecdh this supports the legacy kdf inside the provider dh key exchange. The supporting EVP_PKEY_CTX macros have been changed into mehtods and moved into dh_ctrl.c New kdfs such as SSKDF should be done as a seperate pass after doing the derive. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12575)
2020-08-04 09:21:21 +08:00
ADD_TEST(dhx_cert_load);
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
#endif
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
if (!TEST_ptr(cipher_names = sk_OPENSSL_STRING_new(name_cmp)))
Fix provider cipher reinit issue Fixes #12405 Fixes #12377 Calling Init()/Update() and then Init()/Update() again gave a different result when using the same key and iv. Cipher modes that were using ctx->num were not resetting this value, this includes OFB, CFB & CTR. The fix is to reset this value during the ciphers einit() and dinit() methods. Most ciphers go thru a generic method so one line fixes most cases. Add test for calling EVP_EncryptInit()/EVP_EncryptUpdate() multiple times for all ciphers. Ciphers should return the same value for both updates. DES3-WRAP does not since it uses a random in the update. CCM modes currently also fail on the second update (This also happens in 1_1_1). Fix memory leak in AES_OCB cipher if EVP_EncryptInit is called multiple times. Fix AES_SIV cipher dup_ctx and init. Calling EVP_CIPHER_init multiple times resulted in a memory leak in the siv. Fixing this leak also showed that the dup ctx was not working for siv mode. Note: aes_siv_cleanup() can not be used by aes_siv_dupctx() as it clears data that is required for the decrypt (e.g the tag). Reviewed-by: Tomas Mraz <tmraz@fedoraproject.org> (Merged from https://github.com/openssl/openssl/pull/12413)
2020-07-22 08:40:55 +08:00
return 0;
EVP_CIPHER_do_all_provided(libctx, collect_cipher_names, cipher_names);
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
ADD_ALL_TESTS(test_cipher_reinit, sk_OPENSSL_STRING_num(cipher_names));
Fix provider cipher reinit after init/update with a partial update block. The test added previously used a 16 byte block during the update which does not cause internal buffering in the provider. Some internal variables related to the buffering were not being cleared in the init, which meant that the second update would use the buffered data from the first update. Added test for this scenario with exclusions for ciphers that do not support partial block updates. Found by guidovranken. Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12523)
2020-08-05 06:45:29 +08:00
ADD_ALL_TESTS(test_cipher_reinit_partialupdate,
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
sk_OPENSSL_STRING_num(cipher_names));
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
ADD_TEST(kem_rsa_gen_recover);
ADD_TEST(kem_rsa_params);
#ifndef OPENSSL_NO_DH
ADD_TEST(kem_invalid_keytype);
Add a gettable for provider ciphers to return the EVP_CIPH_RAND_KEY flag Fixes #15531 DES and TDES set this flag which could possibly be used by applications. The gettable cipher param OSSL_CIPHER_PARAM_HAS_RAND_KEY has been added. Note that EVP_CIPHER_CTX_rand_key() uses this flag. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15606)
2021-06-03 17:09:38 +08:00
#endif
#ifndef OPENSSL_NO_DES
Revert "Remove conditional FIPS dependence for 3DES" This reverts commit 464c1011b02936850fc779739013dba52650840a. Reviewed-by: Shane Lontis <shane.lontis@oracle.com> Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/19631)
2022-11-09 06:54:02 +08:00
ADD_TEST(test_cipher_tdes_randkey);
Add KEM (Key encapsulation mechanism) support to providers SP800-56Br2 requires support for the RSA primitives for RSASVE generate and recover. As these are simple KEM operations another operation type has been added that can support future extensions. Added public functions EVP_PKEY_encapsulate_init(), EVP_PKEY_encapsulate(), EVP_PKEY_decapsulate_init() and EVP_PKEY_decapsulate() Added EVP_KEM_* functions. Added OSSL_FUNC_kem_* dispatch functions Added EVP_PKEY_CTX_set_kem_op() so that different types of KEM can be added in the future. This value must currently be set to "RSASVE" after EVP_PKEY_encapsulate_init() & EVP_PKEY_decapsulate_init() as there is no default value. This allows the existing RSA key types, keymanagers, and encoders to be used with the encapsulation operations. The design of the public API's resulted from contributions from @romen & @levitte. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12750)
2020-09-19 16:08:46 +08:00
#endif
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
return 1;
}
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
/* Because OPENSSL_free is a macro, it can't be passed as a function pointer */
static void string_free(char *m)
{
OPENSSL_free(m);
}
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
void cleanup_tests(void)
{
Add type_name member to provided methods and use it Fixes #14701 Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14898)
2021-04-16 22:22:03 +08:00
sk_OPENSSL_STRING_pop_free(cipher_names, string_free);
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
OSSL_PROVIDER_unload(libprov);
Rename OPENSSL_CTX prefix to OSSL_LIB_CTX Many of the new types introduced by OpenSSL 3.0 have an OSSL_ prefix, e.g., OSSL_CALLBACK, OSSL_PARAM, OSSL_ALGORITHM, OSSL_SERIALIZER. The OPENSSL_CTX type stands out a little by using a different prefix. For consistency reasons, this type is renamed to OSSL_LIB_CTX. Reviewed-by: Paul Dale <paul.dale@oracle.com> Reviewed-by: Matt Caswell <matt@openssl.org> (Merged from https://github.com/openssl/openssl/pull/12621)
2020-10-15 17:55:50 +08:00
OSSL_LIB_CTX_free(libctx);
Add multiple fixes for ffc key generation using invalid p,q,g parameters. Fixes #11864 - The dsa keygen assumed valid p, q, g values were being passed. If this is not correct then it is possible that dsa keygen can either hang or segfault. The fix was to do a partial validation of p, q, and g inside the keygen. - Fixed a potential double free in the dsa keypair test in the case when in failed (It should never fail!). It freed internal object members without setting them to NULL. - Changed the FFC key validation to accept 1024 bit keys in non fips mode. - Added tests that use both the default provider & fips provider to test these cases. Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com> (Merged from https://github.com/openssl/openssl/pull/12176)
2020-07-09 11:43:10 +08:00
OSSL_PROVIDER_unload(nullprov);
}