Scott Neugroschl wrote:
diff -C 1 -c listings:
Please also see doc/DIFFS.
*** freeradius-server-2.1.8/libltdl/ltdl.c Wed Dec 30 07:44:35 2009 --- freeradius-server-2.1.8/libltdl/ltdl.c.new Tue Apr 20 07:25:27 2010 *************** *** 2442,2453 ****
while (loader) { if (useloader && strcmp(loader->loader_name, useloader)) { loader = loader->next; continue; } ! lt_user_data data = loader->dlloader_data;
What does that do?
+ lt_user_data data; /* san */
I would really prefer to see patches without /* san */. It's unnecessary, and it complicates the code.
+ #else /* BEGIN SAN */ + #ifdef __TANDEM + #define ss_family __ss_family + #endif
That's a good patch which doesn't affect any other platform.
*** freeradius-server-2.1.8/src/main/auth.c Wed Dec 30 07:44:35 2009 --- freeradius-server-2.1.8/src/main/auth.c.new Tue Apr 20 09:28:18 2010 *************** *** 175,183 **** */ cur_config_item = request->config_items; while(((auth_type_pair = pairfind(cur_config_item, PW_AUTH_TYPE))) != NULL) { auth_type = auth_type_pair->vp_integer; auth_type_count++; ! DICT_VALUE *dv = dict_valbyattr(auth_type_pair->attribute, auth_type_pair->vp_integer);
I don't see why that's necessary.
*** freeradius-server-2.1.8/src/main/command.c Wed Dec 30 07:44:35 2009 --- freeradius-server-2.1.8/src/main/command.c.new Tue Apr 20 09:33:16 2010 *************** *** 842,848 **** return 0; }
! cf_section2file(fp, client->cs);
What is that patch supposed to do?
+ + #define WITHOUT_PROXY /* SAN */ + #define WITHOUT_COA /* SAN */ +
Absolutely not. Patches which *force* behavior for a particular platform also *break* behavior for all other platforms. The server is supposed to work on platforms other than /* san */. If you want to maintain a platform-specific version, learn to use "git".
*** freeradius-server-2.1.8/src/main/listen.c Wed Dec 30 07:44:35 2009 --- freeradius-server-2.1.8/src/main/listen.c.new Tue Apr 20 09:33:05 2010 *************** *** 126,132 **** last_printed = now; }
! listener->print(listener, name, sizeof(name));
I see... "fix" compile warnings by deleting the code. Absolutely not. This shows you haven't spent any time thinking about any side-effects of the patches. After that... I've ignored the rest of your patches. I don't feel like double-checking changes which destroy the server. Alan DeKok.