a single packet). As a side issue, smtp_enforce_sync now works on TLS
connections.
+PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
+ affects you only if you're dancing at the edge of the param size limits.
+ If you are, and this message makes sense to you, then: raise the
+ configured limit or use OpenSSL 1.1. Nothing we can do for older
+ versions.
+
Exim version 4.89
-----------------
#endif
#if OPENSSL_VERSION_NUMBER >= 0x10100000L
# define EXIM_HAVE_OCSP_RESP_COUNT
+# define EXIM_HAVE_OPENSSL_DH_BITS
#else
# define EXIM_HAVE_EPHEM_RSA_KEX
# define EXIM_HAVE_RAND_PSEUDO
DH *dh;
uschar *dhexpanded;
const char *pem;
+int dh_bitsize;
if (!expand_check(dhparam, US"tls_dhparam", &dhexpanded, errstr))
return FALSE;
return FALSE;
}
+/* note: our default limit of 2236 is not a multiple of 8; the limit comes from
+ * an NSS limit, and the GnuTLS APIs handle bit-sizes fine, so we went with
+ * 2236. But older OpenSSL can only report in bytes (octets), not bits.
+ * If someone wants to dance at the edge, then they can raise the limit or use
+ * current libraries. */
+#ifdef EXIM_HAVE_OPENSSL_DH_BITS
+/* Added in commit 26c79d5641d; `git describe --contains` says OpenSSL_1_1_0-pre1~1022
+ * This predates OpenSSL_1_1_0 (before a, b, ...) so is in all 1.1.0 */
+dh_bitsize = DH_bits(dh);
+#else
+dh_bitsize = 8 * DH_size(dh);
+#endif
+
/* Even if it is larger, we silently return success rather than cause things
* to fail out, so that a too-large DH will not knock out all TLS; it's a
* debatable choice. */
-if ((8*DH_size(dh)) > tls_dh_max_bits)
+if (dh_bitsize > tls_dh_max_bits)
{
DEBUG(D_tls)
debug_printf("dhparams file %d bits, is > tls_dh_max_bits limit of %d\n",
- 8*DH_size(dh), tls_dh_max_bits);
+ dh_bitsize, tls_dh_max_bits);
}
else
{
SSL_CTX_set_tmp_dh(sctx, dh);
DEBUG(D_tls)
debug_printf("Diffie-Hellman initialized from %s with %d-bit prime\n",
- dhexpanded ? dhexpanded : US"default", 8*DH_size(dh));
+ dhexpanded ? dhexpanded : US"default", dh_bitsize);
}
DH_free(dh);