Recent versions of engine_pkcs11 are set up to be autoloaded on demand
with ENGINE_by_id() because they don't need explicit configuration.
But if we *do* want to explicitly configure them with a PKCS#11 module
path, we should still do so.
We can't tell whether it was already initialised, but it's harmless to
repeat the MODULE_PATH command if it was.
Signed-off-by: David Woodhouse <David.Woodhouse@xxxxxxxxx>
Tested-by: Michael Schaller <misch@xxxxxxxxxx>
---
Same patch, but Evolution shouldn't be mangling the whitespace this
time. The version I just sent to myself *does* apply with 'git am'.
src/crypto/tls_openssl.c | 12 +++++++++---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/src/crypto/tls_openssl.c b/src/crypto/tls_openssl.c
index c831fba..23ac64b 100644
--- a/src/crypto/tls_openssl.c
+++ b/src/crypto/tls_openssl.c
@@ -729,10 +729,16 @@ static int tls_engine_load_dynamic_generic(const char *pre[],
engine = ENGINE_by_id(id);
if (engine) {
- ENGINE_free(engine);
wpa_printf(MSG_DEBUG, "ENGINE: engine '%s' is already "
"available", id);
- return 0;
+ /*
+ * If it was auto-loaded by ENGINE_by_id() we might still
+ * need to tell it which PKCS#11 module to use in legacy
+ * (non-p11-kit) environments. Do so now; even if it was
+ * properly initialised before, setting it again will be
+ * harmless.
+ */
+ goto found;
}
ERR_clear_error();
@@ -769,7 +775,7 @@ static int tls_engine_load_dynamic_generic(const char *pre[],
id, ERR_error_string(ERR_get_error(), NULL));
return -1;
}
-
+ found:
while (post && post[0]) {
wpa_printf(MSG_DEBUG, "ENGINE: '%s' '%s'", post[0], post[1]);
if (ENGINE_ctrl_cmd_string(engine, post[0], post[1], 0) == 0) {
--
2.7.4
--
David Woodhouse Open Source Technology Centre
David.Woodhouse@xxxxxxxxx Intel CorporationAttachment:
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
