Hello, just finished a configure and a make of httpd-2.3.16-beta. Went fairly smoothly - the issue that required some editting was with some #if #else #endif statements in modules/ssl/ssl_engine_init.c and modules/ssl/ssl_engine_kernel.c
Basically, the IBM xlC (V7) compiler does not like #if constructs in the middle of a function call (or macro?). Diff of the original and my changes that compiled is: diff httpd-2.3.16-beta/modules/ssl/ssl_engine_init.c httpd-2.3.16a/modules/ssl/ssl_engine_init.c 1104d1103 < ap_log_error(APLOG_MARK, APLOG_EMERG, 0, s, APLOGNO(01910) 1106c1105,1107 < "Oops, no RSA, DSA or ECC server certificate found " --- > ap_log_error(APLOG_MARK, APLOG_EMERG, 0, s, APLOGNO(01910) > "Oops, no RSA, DSA or ECC server certificate found " > "for '%s:%d'?!", s->server_hostname, s->port); 1108,1109c1109,1110 < "Oops, no RSA or DSA server certificate found " < #endif --- > ap_log_error(APLOG_MARK, APLOG_EMERG, 0, s, APLOGNO(01910) > "Oops, no RSA or DSA server certificate found " 1110a1112 > #endif 1129d1130 < ap_log_error(APLOG_MARK, APLOG_EMERG, 0, s, APLOGNO(01911) 1130a1132 > ap_log_error(APLOG_MARK, APLOG_EMERG, 0, s, APLOGNO(01911) 1132a1135 > ap_log_error(APLOG_MARK, APLOG_EMERG, 0, s, APLOGNO(01911) 1430d1432 < ap_log_error(APLOG_MARK, 1431a1434 > ap_log_error(APLOG_MARK, 1433,1435d1435 < #else < APLOG_DEBUG, < #endif 1438d1437 < #ifdef OPENSSL_NO_TLSEXT 1439a1439,1445 > "%s (%s:%d) vs. %s (%s:%d)", > ssl_util_vhostid(p, s), > (s->defn_name ? s->defn_name : "unknown"), > s->defn_line_number, > ssl_util_vhostid(p, ps), > (ps->defn_name ? ps->defn_name : "unknown"), > ps->defn_line_number); 1440a1447,1450 > ap_log_error(APLOG_MARK, > APLOG_DEBUG, > 0, > base_server, 1442d1451 < #endif 1449a1459 > #endif 1458d1467 < ap_log_error(APLOG_MARK, APLOG_WARNING, 0, base_server, APLOGNO(01917) 1459a1469 > ap_log_error(APLOG_MARK, APLOG_WARNING, 0, base_server, APLOGNO(01917) 1462a1473 > ap_log_error(APLOG_MARK, APLOG_WARNING, 0, base_server, APLOGNO(01917) Only in httpd-2.3.16a/modules/ssl: ssl_engine_init.lo Only in httpd-2.3.16a/modules/ssl: ssl_engine_init.o Only in httpd-2.3.16a/modules/ssl: ssl_engine_init.slo Only in httpd-2.3.16a/modules/ssl: ssl_engine_io.lo Only in httpd-2.3.16a/modules/ssl: ssl_engine_io.o Only in httpd-2.3.16a/modules/ssl: ssl_engine_io.slo diff httpd-2.3.16-beta/modules/ssl/ssl_engine_kernel.c httpd-2.3.16a/modules/ssl/ssl_engine_kernel.c 767a768 > #if defined(SSL_get_secure_renegotiation_support) 771d771 < #if defined(SSL_get_secure_renegotiation_support) 773c773 < "client does" : "client does not" --- > "client does" : "client does not"); 775c775,778 < "server does not" --- > ap_log_rerror(APLOG_MARK, APLOG_DEBUG, 0, r, APLOGNO(02260) > "Performing full renegotiation: complete handshake " > "protocol (%s support secure renegotiation)", > "server does not"); 777d779 < ); p.s. - As this is not really a "bug", but a compiler mismatch - should I post it as a bug, or is this enough (learning howto 'discuss' - I intend to use your answer as model for the future). regards, Michael
