[Opencryptoki-tech] [PATCH] Testcase improvement: autodetect secure key token.

[Harald Freudenberger]

Every couple of month I have to explain somebody that the diffie-hellman
testcase needs an -securekey option to run successfully on CCA and EP11
token. Now the code dedects these both tokens and automatically swithes
this option on. Also fixed some printout stuff (tripple newlines...) and
return of invalid mechanism which is not a failure.

Signed-off-by: Harald Freudenberger <***@linux.vnet.ibm.com>
---
testcases/crypto/dh_func.c | 35 +++++++++++++++++++++++++----------
1 file changed, 25 insertions(+), 10 deletions(-)

diff --git a/testcases/crypto/dh_func.c b/testcases/crypto/dh_func.c
index 31d9375..32d3c01 100644
--- a/testcases/crypto/dh_func.c
+++ b/testcases/crypto/dh_func.c
@@ -461,10 +461,10 @@ CK_RV do_DeriveDHKey(void)
// Make sure peer's key is the right size
if ((extr1_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME)) &&
(extr1_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME)-1)) {
- testcase_fail("ERROR:size error peer's key %ld\n",extr1_tmpl[0].ulValueLen );
+ testcase_fail("ERROR:size error peer's key %ld",extr1_tmpl[0].ulValueLen );
goto testcase_cleanup;
} else
- testcase_pass("Successfully derived key.\n");
+ testcase_pass("Successfully derived key");

// Testcase #2 - Now derive the secrets...
if (!securekey) {
@@ -500,7 +500,7 @@ CK_RV do_DeriveDHKey(void)
// Make sure party A's key is the right size
if ((extr2_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME)) &&
(extr2_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME)-1)) {
- testcase_fail("ERROR:size error party A's key %ld\n",extr2_tmpl[0].ulValueLen);
+ testcase_fail("ERROR:size error party A's key %ld",extr2_tmpl[0].ulValueLen);
goto testcase_cleanup;
}

@@ -533,7 +533,7 @@ CK_RV do_DeriveDHKey(void)

if (extr1_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME) ||
*((int*)extr1_tmpl[0].pValue) == 0) {
- testcase_fail("ERROR:derived key #1 length or value %ld\n", extr1_tmpl[0].ulValueLen );
+ testcase_fail("ERROR:derived key #1 length or value %ld", extr1_tmpl[0].ulValueLen );
goto testcase_cleanup;
}

@@ -550,12 +550,12 @@ CK_RV do_DeriveDHKey(void)

if (extr2_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME) ||
*((int*)extr2_tmpl[0].pValue) == 0) {
- testcase_fail("ERROR:derived key #2 length or value %ld\n", extr2_tmpl[0].ulValueLen );
+ testcase_fail("ERROR:derived key #2 length or value %ld", extr2_tmpl[0].ulValueLen );
goto testcase_cleanup;
}

if (memcmp(key1_value, key2_value, sizeof(DH_PUBL_PRIME)) != 0){
- testcase_fail("ERROR:derived key mismatch\n");
+ testcase_fail("ERROR:derived key mismatch");
goto testcase_cleanup;
}

@@ -600,7 +600,7 @@ CK_RV do_DeriveDHKey(void)
// Make sure party A's key is the right size
if ((extr2_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME)) &&
(extr2_tmpl[0].ulValueLen != sizeof(DH_PUBL_PRIME)-1)) {
- testcase_fail("ERROR:size error party A's key %ld\n",extr2_tmpl[0].ulValueLen);
+ testcase_fail("ERROR:size error party A's key %ld",extr2_tmpl[0].ulValueLen);
goto testcase_cleanup;
}

@@ -653,7 +653,7 @@ CK_RV do_DeriveDHKey(void)
}

if (memcmp(clear, re_cipher, 32) != 0) {
- testcase_fail("ERROR:data mismatch\n");
+ testcase_fail("ERROR:data mismatch");
goto testcase_cleanup;
}

@@ -686,8 +686,15 @@ CK_RV dh_functions()

if ((rv == CKR_OK) && (rv2 == CKR_OK))
rv = do_DeriveDHKey();
- else
- rv = CKR_MECHANISM_INVALID;
+ else {
+ /*
+ * One of the above mechanism is not available, so skip
+ * the test but do not report any
+ * rv = CKR_MECHANISM_INVALID;
+ * invalid or however failures as this is not a failure.
+ */
+ return CKR_OK;
+ }

return rv;
}
@@ -718,6 +725,14 @@ int main(int argc, char **argv)

funcs->C_Initialize( &cinit_args );

+ /*
+ * every couple of month I have to tell someone that the -securekey
+ * option is needed on CCA and EP11 token otherwise the testcase
+ * will fail. However, now this will be done automatically here:
+ */
+ if (is_ep11_token(SLOT_ID) || is_cca_token(SLOT_ID))
+ securekey = 1;
+
{
CK_SESSION_HANDLE hsess = 0;