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/params_test.c

669 lines
23 KiB
C
Raw Normal View History

test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
/*
Update copyright year Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14235)
2021-02-18 22:57:13 +08:00
* Copyright 2019-2021 The OpenSSL Project Authors. All Rights Reserved.
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
*
* Licensed under the Apache License 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
* https://www.openssl.org/source/license.html
* or in the file LICENSE in the source distribution.
*/
/*
* This program tests the use of OSSL_PARAM, currently in raw form.
*/
#include <string.h>
#include <openssl/bn.h>
#include <openssl/core.h>
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
#include <openssl/params.h>
Include "internal/numbers.h" in test programs using SIZE_MAX Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/15366)
2021-05-20 15:42:12 +08:00
#include "internal/numbers.h"
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
#include "internal/nelem.h"
#include "testutil.h"
/*-
* PROVIDER SECTION
* ================
*
* Even though it's not necessarily ONLY providers doing this part,
* they are naturally going to be the most common users of
* set_params and get_params functions.
*/
/*
* In real use cases, setters and getters would take an object with
* which the parameters are associated. This structure is a cheap
* simulation.
*/
struct object_st {
/*
* Documented as a native integer, of the size given by sizeof(int).
* Assumed data type OSSL_PARAM_INTEGER
*/
int p1;
/*
* Documented as a native double, of the size given by sizeof(double).
* Assumed data type OSSL_PARAM_REAL
*/
double p2;
/*
* Documented as an arbitrarly large unsigned integer.
Fix Typos CLA: trivial Reviewed-by: Richard Levitte <levitte@openssl.org> Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com> (Merged from https://github.com/openssl/openssl/pull/9288)
2019-07-02 16:04:04 +08:00
* The data size must be large enough to accommodate.
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
* Assumed data type OSSL_PARAM_UNSIGNED_INTEGER
*/
BIGNUM *p3;
/*
* Documented as a C string.
Fix Typos CLA: trivial Reviewed-by: Richard Levitte <levitte@openssl.org> Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com> (Merged from https://github.com/openssl/openssl/pull/9288)
2019-07-02 16:04:04 +08:00
* The data size must be large enough to accommodate.
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
* Assumed data type OSSL_PARAM_UTF8_STRING
*/
char *p4;
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
size_t p4_l;
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
/*
* Documented as a C string.
* Assumed data type OSSL_PARAM_UTF8_STRING
*/
char p5[256];
size_t p5_l;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
/*
* Documented as a pointer to a constant C string.
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
* Assumed data type OSSL_PARAM_UTF8_PTR
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
*/
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
const char *p6;
size_t p6_l;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
};
#define p1_init 42 /* The ultimate answer */
#define p2_init 6.283 /* Magic number */
/* Stolen from evp_data, BLAKE2s256 test */
#define p3_init \
"4142434445464748494a4b4c4d4e4f50" \
"5152535455565758595a616263646566" \
"6768696a6b6c6d6e6f70717273747576" \
"7778797a30313233343536373839"
#define p4_init "BLAKE2s256" /* Random string */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
#define p5_init "Hellow World" /* Random string */
#define p6_init OPENSSL_FULL_VERSION_STR /* Static string */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
static void cleanup_object(void *vobj)
{
struct object_st *obj = vobj;
BN_free(obj->p3);
obj->p3 = NULL;
OPENSSL_free(obj->p4);
obj->p4 = NULL;
OPENSSL_free(obj);
}
static void *init_object(void)
{
test: check for NULL returns better Reviewed-by: Shane Lontis <shane.lontis@oracle.com> (Merged from https://github.com/openssl/openssl/pull/15910)
2021-06-25 10:57:53 +08:00
struct object_st *obj;
if (!TEST_ptr(obj = OPENSSL_zalloc(sizeof(*obj))))
return NULL;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
obj->p1 = p1_init;
obj->p2 = p2_init;
if (!TEST_true(BN_hex2bn(&obj->p3, p3_init)))
goto fail;
if (!TEST_ptr(obj->p4 = OPENSSL_strdup(p4_init)))
goto fail;
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
strcpy(obj->p5, p5_init);
obj->p6 = p6_init;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
return obj;
fail:
cleanup_object(obj);
obj = NULL;
return NULL;
}
/*
* RAW provider, which handles the parameters in a very raw manner,
* with no fancy API and very minimal checking. The application that
* calls these to set or request parameters MUST get its OSSL_PARAM
* array right.
*/
static int raw_set_params(void *vobj, const OSSL_PARAM *params)
{
struct object_st *obj = vobj;
for (; params->key != NULL; params++)
if (strcmp(params->key, "p1") == 0) {
obj->p1 = *(int *)params->data;
} else if (strcmp(params->key, "p2") == 0) {
obj->p2 = *(double *)params->data;
} else if (strcmp(params->key, "p3") == 0) {
BN_free(obj->p3);
if (!TEST_ptr(obj->p3 = BN_native2bn(params->data,
params->data_size, NULL)))
return 0;
} else if (strcmp(params->key, "p4") == 0) {
OPENSSL_free(obj->p4);
if (!TEST_ptr(obj->p4 = OPENSSL_strndup(params->data,
params->data_size)))
return 0;
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
obj->p4_l = strlen(obj->p4);
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
} else if (strcmp(params->key, "p5") == 0) {
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
/*
* Protect obj->p5 against too much data. This should not
* happen, we don't use that long strings.
*/
size_t data_length =
OPENSSL_strnlen(params->data, params->data_size);
if (!TEST_size_t_lt(data_length, sizeof(obj->p5)))
return 0;
strncpy(obj->p5, params->data, data_length);
obj->p5[data_length] = '\0';
obj->p5_l = strlen(obj->p5);
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
} else if (strcmp(params->key, "p6") == 0) {
obj->p6 = *(const char **)params->data;
test/params_test.c : Adjust tests to check utf8_ptr sizes Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8703)
2019-04-09 19:16:16 +08:00
obj->p6_l = params->data_size;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
}
return 1;
}
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
static int raw_get_params(void *vobj, OSSL_PARAM *params)
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
{
struct object_st *obj = vobj;
for (; params->key != NULL; params++)
if (strcmp(params->key, "p1") == 0) {
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
params->return_size = sizeof(obj->p1);
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
*(int *)params->data = obj->p1;
} else if (strcmp(params->key, "p2") == 0) {
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
params->return_size = sizeof(obj->p2);
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
*(double *)params->data = obj->p2;
} else if (strcmp(params->key, "p3") == 0) {
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
params->return_size = BN_num_bytes(obj->p3);
if (!TEST_size_t_ge(params->data_size, params->return_size))
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
return 0;
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
BN_bn2nativepad(obj->p3, params->data, params->return_size);
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
} else if (strcmp(params->key, "p4") == 0) {
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
params->return_size = strlen(obj->p4);
if (!TEST_size_t_gt(params->data_size, params->return_size))
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
return 0;
strcpy(params->data, obj->p4);
} else if (strcmp(params->key, "p5") == 0) {
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
params->return_size = strlen(obj->p5);
if (!TEST_size_t_gt(params->data_size, params->return_size))
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
return 0;
strcpy(params->data, obj->p5);
} else if (strcmp(params->key, "p6") == 0) {
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
params->return_size = strlen(obj->p6);
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
*(const char **)params->data = obj->p6;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
}
return 1;
}
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
/*
* API provider, which handles the parameters using the API from params.h
*/
static int api_set_params(void *vobj, const OSSL_PARAM *params)
{
struct object_st *obj = vobj;
const OSSL_PARAM *p = NULL;
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
if ((p = OSSL_PARAM_locate_const(params, "p1")) != NULL
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
&& !TEST_true(OSSL_PARAM_get_int(p, &obj->p1)))
return 0;
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
if ((p = OSSL_PARAM_locate_const(params, "p2")) != NULL
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
&& !TEST_true(OSSL_PARAM_get_double(p, &obj->p2)))
return 0;
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
if ((p = OSSL_PARAM_locate_const(params, "p3")) != NULL
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
&& !TEST_true(OSSL_PARAM_get_BN(p, &obj->p3)))
return 0;
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
if ((p = OSSL_PARAM_locate_const(params, "p4")) != NULL) {
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
OPENSSL_free(obj->p4);
obj->p4 = NULL;
/* If the value pointer is NULL, we get it automatically allocated */
if (!TEST_true(OSSL_PARAM_get_utf8_string(p, &obj->p4, 0)))
return 0;
}
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
if ((p = OSSL_PARAM_locate_const(params, "p5")) != NULL) {
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
char *p5_ptr = obj->p5;
if (!TEST_true(OSSL_PARAM_get_utf8_string(p, &p5_ptr, sizeof(obj->p5))))
return 0;
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
obj->p5_l = strlen(obj->p5);
test/params_test.c : Adjust tests to check utf8_ptr sizes Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8703)
2019-04-09 19:16:16 +08:00
}
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
if ((p = OSSL_PARAM_locate_const(params, "p6")) != NULL) {
test/params_test.c : Adjust tests to check utf8_ptr sizes Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8703)
2019-04-09 19:16:16 +08:00
if (!TEST_true(OSSL_PARAM_get_utf8_ptr(p, &obj->p6)))
return 0;
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
obj->p6_l = strlen(obj->p6);
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
}
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
return 1;
}
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
static int api_get_params(void *vobj, OSSL_PARAM *params)
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
{
struct object_st *obj = vobj;
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
OSSL_PARAM *p = NULL;
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
if ((p = OSSL_PARAM_locate(params, "p1")) != NULL
&& !TEST_true(OSSL_PARAM_set_int(p, obj->p1)))
return 0;
if ((p = OSSL_PARAM_locate(params, "p2")) != NULL
&& !TEST_true(OSSL_PARAM_set_double(p, obj->p2)))
return 0;
if ((p = OSSL_PARAM_locate(params, "p3")) != NULL
&& !TEST_true(OSSL_PARAM_set_BN(p, obj->p3)))
return 0;
if ((p = OSSL_PARAM_locate(params, "p4")) != NULL
&& !TEST_true(OSSL_PARAM_set_utf8_string(p, obj->p4)))
return 0;
if ((p = OSSL_PARAM_locate(params, "p5")) != NULL
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
&& !TEST_true(OSSL_PARAM_set_utf8_string(p, obj->p5)))
return 0;
if ((p = OSSL_PARAM_locate(params, "p6")) != NULL
&& !TEST_true(OSSL_PARAM_set_utf8_ptr(p, obj->p6)))
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
return 0;
return 1;
}
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
/*
* This structure only simulates a provider dispatch, the real deal is
* a bit more code that's not necessary in these tests.
*/
struct provider_dispatch_st {
int (*set_params)(void *obj, const OSSL_PARAM *params);
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
int (*get_params)(void *obj, OSSL_PARAM *params);
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
};
/* "raw" provider */
static const struct provider_dispatch_st provider_raw = {
raw_set_params, raw_get_params
};
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
/* "api" provider */
static const struct provider_dispatch_st provider_api = {
api_set_params, api_get_params
};
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
/*-
* APPLICATION SECTION
* ===================
*/
/* In all our tests, these are variables that get manipulated as parameters
*
Fix Typos CLA: trivial Reviewed-by: Richard Levitte <levitte@openssl.org> Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com> (Merged from https://github.com/openssl/openssl/pull/9288)
2019-07-02 16:04:04 +08:00
* These arrays consistently do nothing with the "p2" parameter, and
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
* always include a "foo" parameter. This is to check that the
* set_params and get_params calls ignore the lack of parameters that
* the application isn't interested in, as well as ignore parameters
* they don't understand (the application may have one big bag of
* parameters).
*/
static int app_p1; /* "p1" */
static double app_p2; /* "p2" is ignored */
static BIGNUM *app_p3 = NULL; /* "p3" */
static unsigned char bignumbin[4096]; /* "p3" */
static char app_p4[256]; /* "p4" */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
static char app_p5[256]; /* "p5" */
static const char *app_p6 = NULL; /* "p6" */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
static unsigned char foo[1]; /* "foo" */
#define app_p1_init 17 /* A random number */
#define app_p2_init 47.11 /* Another random number */
#define app_p3_init "deadbeef" /* Classic */
#define app_p4_init "Hello"
#define app_p5_init "World"
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
#define app_p6_init "Cookie"
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
#define app_foo_init 'z'
static int cleanup_app_variables(void)
{
BN_free(app_p3);
app_p3 = NULL;
return 1;
}
static int init_app_variables(void)
{
int l = 0;
cleanup_app_variables();
app_p1 = app_p1_init;
app_p2 = app_p2_init;
if (!BN_hex2bn(&app_p3, app_p3_init)
|| (l = BN_bn2nativepad(app_p3, bignumbin, sizeof(bignumbin))) < 0)
return 0;
strcpy(app_p4, app_p4_init);
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
strcpy(app_p5, app_p5_init);
app_p6 = app_p6_init;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
foo[0] = app_foo_init;
return 1;
}
/*
* Here, we define test OSSL_PARAM arrays
*/
/* An array of OSSL_PARAM, specific in the most raw manner possible */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
static OSSL_PARAM static_raw_params[] = {
{ "p1", OSSL_PARAM_INTEGER, &app_p1, sizeof(app_p1), 0 },
{ "p3", OSSL_PARAM_UNSIGNED_INTEGER, &bignumbin, sizeof(bignumbin), 0 },
{ "p4", OSSL_PARAM_UTF8_STRING, &app_p4, sizeof(app_p4), 0 },
{ "p5", OSSL_PARAM_UTF8_STRING, &app_p5, sizeof(app_p5), 0 },
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
/* sizeof(app_p6_init) - 1, because we know that's what we're using */
{ "p6", OSSL_PARAM_UTF8_PTR, &app_p6, sizeof(app_p6_init) - 1, 0 },
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
{ "foo", OSSL_PARAM_OCTET_STRING, &foo, sizeof(foo), 0 },
{ NULL, 0, NULL, 0, 0 }
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
};
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
/* The same array of OSSL_PARAM, specified with the macros from params.h */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
static OSSL_PARAM static_api_params[] = {
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
OSSL_PARAM_int("p1", &app_p1),
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
OSSL_PARAM_BN("p3", &bignumbin, sizeof(bignumbin)),
OSSL_PARAM_DEFN("p4", OSSL_PARAM_UTF8_STRING, &app_p4, sizeof(app_p4)),
OSSL_PARAM_DEFN("p5", OSSL_PARAM_UTF8_STRING, &app_p5, sizeof(app_p5)),
test/params_test.c : Adjust tests to check utf8_ptr sizes Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8703)
2019-04-09 19:16:16 +08:00
/* sizeof(app_p6_init), because we know that's what we're using */
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
OSSL_PARAM_DEFN("p6", OSSL_PARAM_UTF8_PTR, &app_p6,
sizeof(app_p6_init) - 1),
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
OSSL_PARAM_DEFN("foo", OSSL_PARAM_OCTET_STRING, &foo, sizeof(foo)),
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
OSSL_PARAM_END
};
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
/*
* The same array again, but constructed at run-time
* This exercises the OSSL_PARAM constructor functions
*/
test/params_test.c: make construct_api_params() static With enough warning flags, compilers complain when a non-static function hasn't been properly declared... Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> (Merged from https://github.com/openssl/openssl/pull/8469)
2019-03-13 20:56:46 +08:00
static OSSL_PARAM *construct_api_params(void)
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
{
size_t n = 0;
static OSSL_PARAM params[10];
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
params[n++] = OSSL_PARAM_construct_int("p1", &app_p1);
params[n++] = OSSL_PARAM_construct_BN("p3", bignumbin, sizeof(bignumbin));
params[n++] = OSSL_PARAM_construct_utf8_string("p4", app_p4,
sizeof(app_p4));
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
params[n++] = OSSL_PARAM_construct_utf8_string("p5", app_p5,
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
sizeof(app_p5));
test/params_test.c : Adjust tests to check utf8_ptr sizes Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8703)
2019-04-09 19:16:16 +08:00
/* sizeof(app_p6_init), because we know that's what we're using */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
params[n++] = OSSL_PARAM_construct_utf8_ptr("p6", (char **)&app_p6,
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
sizeof(app_p6_init));
params[n++] = OSSL_PARAM_construct_octet_string("foo", &foo, sizeof(foo));
Params: add OSSL_PARAM_construct_end() OSSL_PARAM_END is a macro that can only be used to initialize an OSSL_PARAM array, not to assign an array element later on. For completion, we add an end constructor to facilitate that kind of assignment. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8704)
2019-04-09 15:49:58 +08:00
params[n++] = OSSL_PARAM_construct_end();
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
return params;
}
struct param_owner_st {
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
OSSL_PARAM *static_params;
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
OSSL_PARAM *(*constructed_params)(void);
};
test/params_test.c: make more global variables static Again, compilers that don't like them being undeclared... Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> (Merged from https://github.com/openssl/openssl/pull/8469)
2019-03-13 21:01:27 +08:00
static const struct param_owner_st raw_params = {
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
static_raw_params, NULL
};
test/params_test.c: make more global variables static Again, compilers that don't like them being undeclared... Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> (Merged from https://github.com/openssl/openssl/pull/8469)
2019-03-13 21:01:27 +08:00
static const struct param_owner_st api_params = {
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
static_api_params, construct_api_params
};
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
/*-
* TESTING
* =======
*/
/*
* Test cases to combine parameters with "provider side" functions
*/
static struct {
const struct provider_dispatch_st *prov;
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
const struct param_owner_st *app;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
const char *desc;
} test_cases[] = {
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
/* Tests within specific methods */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
{ &provider_raw, &raw_params, "raw provider vs raw params" },
{ &provider_api, &api_params, "api provider vs api params" },
test/params_test.c: Add API test case, and mixed methods Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 03:34:51 +08:00
/* Mixed methods */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
{ &provider_raw, &api_params, "raw provider vs api params" },
{ &provider_api, &raw_params, "api provider vs raw params" },
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
};
/* Generic tester of combinations of "providers" and params */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
static int test_case_variant(OSSL_PARAM *params, const struct provider_dispatch_st *prov)
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
{
BIGNUM *verify_p3 = NULL;
void *obj = NULL;
int errcnt = 0;
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
OSSL_PARAM *p;
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
/*
* Initialize
*/
if (!TEST_ptr(obj = init_object())
|| !TEST_true(BN_hex2bn(&verify_p3, p3_init))) {
errcnt++;
goto fin;
}
/*
* Get parameters a first time, just to see that getting works and
* gets us the values we expect.
*/
init_app_variables();
if (!TEST_true(prov->get_params(obj, params))
|| !TEST_int_eq(app_p1, p1_init) /* "provider" value */
test/params_test.c: use TEST_double_eq to check doubles TEST_ulong_eq was used previously because TEST_double_eq didn't exist at the time. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> (Merged from https://github.com/openssl/openssl/pull/8469)
2019-03-13 21:06:03 +08:00
|| !TEST_double_eq(app_p2, app_p2_init) /* Should remain untouched */
The params_test uses wrong size for BN check. Before the return size was included in the OSSL_PARAM structure, the size global was updated via a pointer. The size global variable has been removed and the in structure return size used instead. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9248)
2019-06-24 23:37:07 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "p3"))
|| !TEST_ptr(BN_native2bn(bignumbin, p->return_size, app_p3))
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_BN_eq(app_p3, verify_p3) /* "provider" value */
|| !TEST_str_eq(app_p4, p4_init) /* "provider" value */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "p5"))
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
|| !TEST_size_t_eq(p->return_size,
sizeof(p5_init) - 1) /* "provider" value */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_str_eq(app_p5, p5_init) /* "provider" value */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "p6"))
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
|| !TEST_size_t_eq(p->return_size,
sizeof(p6_init) - 1) /* "provider" value */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
|| !TEST_str_eq(app_p6, p6_init) /* "provider" value */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_char_eq(foo[0], app_foo_init) /* Should remain untouched */
params: add functionality to test if an OSSL_PARAM has been set. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/11588)
2020-04-21 08:49:19 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "foo")))
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
errcnt++;
/*
* Set parameters, then sneak into the object itself and check
* that its attributes got set (or ignored) properly.
*/
init_app_variables();
if (!TEST_true(prov->set_params(obj, params))) {
errcnt++;
} else {
struct object_st *sneakpeek = obj;
if (!TEST_int_eq(sneakpeek->p1, app_p1) /* app value set */
test/params_test.c: use TEST_double_eq to check doubles TEST_ulong_eq was used previously because TEST_double_eq didn't exist at the time. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> (Merged from https://github.com/openssl/openssl/pull/8469)
2019-03-13 21:06:03 +08:00
|| !TEST_double_eq(sneakpeek->p2, p2_init) /* Should remain untouched */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_BN_eq(sneakpeek->p3, app_p3) /* app value set */
|| !TEST_str_eq(sneakpeek->p4, app_p4) /* app value set */
test/params_test.c : Adjust tests to check utf8_ptr sizes Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8703)
2019-04-09 19:16:16 +08:00
|| !TEST_str_eq(sneakpeek->p5, app_p5) /* app value set */
|| !TEST_str_eq(sneakpeek->p6, app_p6)) /* app value set */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
errcnt++;
}
/*
* Get parameters again, checking that we get different values
* than earlier where relevant.
*/
BN_free(verify_p3);
verify_p3 = NULL;
if (!TEST_true(BN_hex2bn(&verify_p3, app_p3_init))) {
errcnt++;
goto fin;
}
if (!TEST_true(prov->get_params(obj, params))
|| !TEST_int_eq(app_p1, app_p1_init) /* app value */
test/params_test.c: use TEST_double_eq to check doubles TEST_ulong_eq was used previously because TEST_double_eq didn't exist at the time. Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com> (Merged from https://github.com/openssl/openssl/pull/8469)
2019-03-13 21:06:03 +08:00
|| !TEST_double_eq(app_p2, app_p2_init) /* Should remain untouched */
The params_test uses wrong size for BN check. Before the return size was included in the OSSL_PARAM structure, the size global was updated via a pointer. The size global variable has been removed and the in structure return size used instead. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9248)
2019-06-24 23:37:07 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "p3"))
|| !TEST_ptr(BN_native2bn(bignumbin, p->return_size, app_p3))
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_BN_eq(app_p3, verify_p3) /* app value */
|| !TEST_str_eq(app_p4, app_p4_init) /* app value */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "p5"))
|| !TEST_size_t_eq(p->return_size,
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
sizeof(app_p5_init) - 1) /* app value */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_str_eq(app_p5, app_p5_init) /* app value */
Change OSSL_PARAM return size to not be a pointer. Instead of referencing the return size from the OSSL_PARAM structure, make the size a field within the structure. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/9135)
2019-06-24 12:43:55 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "p6"))
|| !TEST_size_t_eq(p->return_size,
OSSL_PARAM: Correct the assumptions on the UTF8 string length When the string "ABCDEFGH" is passed, what's considered its data, this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H' } or this? { 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', '\0' } If it's passed as a pass phrase, should the terminating NUL byte be considered part of the pass phrase, or not? Our treatment of OSSL_PARAMs with the data type OSSL_PARAM_UTF8_STRING set the length of the string to include the terminating NUL byte, which is quite confusing. What should the recipient of such a string believe? Instead of perpetuating this confusion, we change the assumption to set the OSSL_PARAM to the length of the string, not including the terminating NUL byte, thereby giving it the same value as a strlen() call would give. Reviewed-by: Tomas Mraz <tomas@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14168)
2021-02-13 03:30:40 +08:00
sizeof(app_p6_init) - 1) /* app value */
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
|| !TEST_str_eq(app_p6, app_p6_init) /* app value */
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
|| !TEST_char_eq(foo[0], app_foo_init) /* Should remain untouched */
params: add functionality to test if an OSSL_PARAM has been set. Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/11588)
2020-04-21 08:49:19 +08:00
|| !TEST_ptr(p = OSSL_PARAM_locate(params, "foo")))
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
errcnt++;
fin:
BN_free(verify_p3);
verify_p3 = NULL;
cleanup_app_variables();
cleanup_object(obj);
return errcnt == 0;
}
test/params_test.c: Add run-time constructed OSSL_PARAM variant Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-08 16:54:04 +08:00
static int test_case(int i)
{
TEST_info("Case: %s", test_cases[i].desc);
return test_case_variant(test_cases[i].app->static_params,
test_cases[i].prov)
&& (test_cases[i].app->constructed_params == NULL
|| test_case_variant(test_cases[i].app->constructed_params(),
test_cases[i].prov));
}
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
/*-
* OSSL_PARAM_allocate_from_text() tests
* =====================================
*/
static const OSSL_PARAM params_from_text[] = {
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
/* Fixed size buffer */
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
OSSL_PARAM_int32("int", NULL),
OSSL_PARAM_DEFN("short", OSSL_PARAM_INTEGER, NULL, sizeof(int16_t)),
OSSL_PARAM_DEFN("ushort", OSSL_PARAM_UNSIGNED_INTEGER, NULL, sizeof(uint16_t)),
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
/* Arbitrary size buffer. Make sure the result fits in a long */
OSSL_PARAM_DEFN("num", OSSL_PARAM_INTEGER, NULL, 0),
OSSL_PARAM_DEFN("unum", OSSL_PARAM_UNSIGNED_INTEGER, NULL, 0),
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
OSSL_PARAM_END,
};
struct int_from_text_test_st {
const char *argname;
const char *strval;
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
long int expected_intval;
int expected_res;
size_t expected_bufsize;
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
};
static struct int_from_text_test_st int_from_text_test_cases[] = {
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
{ "int", "", 0, 0, 0 },
{ "int", "0", 0, 1, 4 },
{ "int", "101", 101, 1, 4 },
{ "int", "-102", -102, 1, 4 },
{ "int", "12A", 12, 1, 4 }, /* incomplete */
{ "int", "0x12B", 0x12B, 1, 4 },
{ "hexint", "12C", 0x12C, 1, 4 },
{ "hexint", "0x12D", 0, 1, 4 }, /* zero */
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
/* test check of the target buffer size */
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
{ "int", "0x7fffffff", INT32_MAX, 1, 4 },
{ "int", "2147483647", INT32_MAX, 1, 4 },
{ "int", "2147483648", 0, 0, 0 }, /* too small buffer */
{ "int", "-2147483648", INT32_MIN, 1, 4 },
{ "int", "-2147483649", 0, 0, 4 }, /* too small buffer */
{ "short", "0x7fff", INT16_MAX, 1, 2 },
{ "short", "32767", INT16_MAX, 1, 2 },
{ "short", "32768", 0, 0, 0 }, /* too small buffer */
{ "ushort", "0xffff", UINT16_MAX, 1, 2 },
{ "ushort", "65535", UINT16_MAX, 1, 2 },
{ "ushort", "65536", 0, 0, 0 }, /* too small buffer */
/* test check of sign extension in arbitrary size results */
{ "num", "0", 0, 1, 1 },
{ "num", "0", 0, 1, 1 },
{ "num", "0xff", 0xff, 1, 2 }, /* sign extension */
{ "num", "-0xff", -0xff, 1, 2 }, /* sign extension */
{ "num", "0x7f", 0x7f, 1, 1 }, /* no sign extension */
{ "num", "-0x7f", -0x7f, 1, 1 }, /* no sign extension */
{ "num", "0x80", 0x80, 1, 2 }, /* sign extension */
{ "num", "-0x80", -0x80, 1, 1 }, /* no sign extension */
{ "num", "0x81", 0x81, 1, 2 }, /* sign extension */
{ "num", "-0x81", -0x81, 1, 2 }, /* sign extension */
{ "unum", "0xff", 0xff, 1, 1 },
{ "unum", "-0xff", -0xff, 0, 0 }, /* invalid neg number */
{ "unum", "0x7f", 0x7f, 1, 1 },
{ "unum", "-0x7f", -0x7f, 0, 0 }, /* invalid neg number */
{ "unum", "0x80", 0x80, 1, 1 },
{ "unum", "-0x80", -0x80, 0, 0 }, /* invalid neg number */
{ "unum", "0x81", 0x81, 1, 1 },
{ "unum", "-0x81", -0x81, 0, 0 }, /* invalid neg number */
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
};
static int check_int_from_text(const struct int_from_text_test_st a)
{
OSSL_PARAM param;
long int val = 0;
int res;
if (!OSSL_PARAM_allocate_from_text(&param, params_from_text,
a.argname, a.strval, 0, NULL)) {
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
if (a.expected_res)
TEST_error("unexpected OSSL_PARAM_allocate_from_text() return for %s \"%s\"",
a.argname, a.strval);
return !a.expected_res;
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
}
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
/* For data size zero, OSSL_PARAM_get_long() may crash */
if (param.data_size == 0) {
OPENSSL_free(param.data);
TEST_error("unexpected zero size for %s \"%s\"",
a.argname, a.strval);
return 0;
}
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
res = OSSL_PARAM_get_long(&param, &val);
OPENSSL_free(param.data);
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
if (res ^ a.expected_res) {
TEST_error("unexpected OSSL_PARAM_get_long() return for %s \"%s\": "
"%d != %d", a.argname, a.strval, a.expected_res, res);
return 0;
}
if (val != a.expected_intval) {
TEST_error("unexpected result for %s \"%s\": %li != %li",
a.argname, a.strval, a.expected_intval, val);
return 0;
}
if (param.data_size != a.expected_bufsize) {
TEST_error("unexpected size for %s \"%s\": %d != %d",
a.argname, a.strval,
(int)a.expected_bufsize, (int)param.data_size);
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
return 0;
}
Test the performance of OSSL_PARAM_allocate_from_text with arbitrary size ints With arbitrary size ints, we get to know exactly how large the minimum buffer must be. Reviewed-by: Tomas Mraz <tomas@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> (Merged from https://github.com/openssl/openssl/pull/17104) (cherry picked from commit b556713a6f2884eadc7f56428bc82a844e9a49e0)
2021-11-22 23:38:43 +08:00
return a.expected_res;
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
}
static int test_allocate_from_text(int i)
{
return check_int_from_text(int_from_text_test_cases[i]);
}
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
int setup_tests(void)
{
ADD_ALL_TESTS(test_case, OSSL_NELEM(test_cases));
Enhanced integer parsing in OSSL_PARAM_allocate_from_text Fixes #14041 and additional bugs discovered by the newly created tests. This patch: - Introduces support for 0x prefixed integers - Fixes parsing of negative integers (negative numbers were shifted by -2) - Fixes ability to parse maximal unsigned numbers ("too small buffer" error used to be reported incorrectly) - Fixes a memory leak when OSSL_PARAM_allocate_from_text fails leaving a temporary BN allocated Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Paul Dale <pauli@openssl.org> Reviewed-by: Richard Levitte <levitte@openssl.org> (Merged from https://github.com/openssl/openssl/pull/14093)
2021-02-07 04:47:20 +08:00
ADD_ALL_TESTS(test_allocate_from_text, OSSL_NELEM(int_from_text_test_cases));
test/params_test.c: Add "real world" parameter testing test/params_test.c is a program that tries to mimic how a provider and an application would or could handle OSSL_PARAM arrays. For the moment, this program tests a very raw way of handling OSSL_PARAM arrays. It is, however, written in a way that will hopefully make it possible to extend with other methods as APIs arise, and to set up test cases where a "provider" handles the array one way while the "application" handles it another way. Reviewed-by: Paul Dale <paul.dale@oracle.com> (Merged from https://github.com/openssl/openssl/pull/8396)
2019-03-05 00:22:45 +08:00
return 1;
}