do_PVK_body: Unconditionally free enctmp

enctmp is only allocated if saltlen > 0, so there is no harm in checking
for that, but it's also pointless. Unconfuses smatch who thinks there
might be a memory leak:

pem/pvkfmt.c:808 do_PVK_body() warn: possible memory leak of 'enctmp'

found by jsg

1 files changed, 3 insertions, 3 deletions

diff --git a/src/lib/libcrypto/pem/pvkfmt.c b/src/lib/libcrypto/pem/pvkfmt.c
index 6967d65eee..395fd9df83 100644
--- a/src/lib/libcrypto/pem/pvkfmt.c
+++ b/src/lib/libcrypto/pem/pvkfmt.c
@@ -1,4 +1,4 @@
-/* $OpenBSD: pvkfmt.c,v 1.29 2025/05/10 05:54:38 tb Exp $ */
+/* $OpenBSD: pvkfmt.c,v 1.30 2025/06/07 09:32:35 tb Exp $ */
 /* Written by Dr Stephen N Henson (steve@openssl.org) for the OpenSSL
  * project 2005.
  */
@@ -803,8 +803,8 @@ do_PVK_body(const unsigned char **in, unsigned int saltlen,
 
  err:
         EVP_CIPHER_CTX_free(cctx);
-        if (enctmp && saltlen)
-                free(enctmp);
+        free(enctmp);
+
         return ret;
 }