|
|||||||||||
|
patch for inetd
From: Alexei Kosut <akosut(at)organic.com>
Date: Fri Jun 21 1996 - 16:56:05 EDT
It's an unhappy fix, but it works. The problem is that cleanup_for_exec ends up (through no fault of its own), closing stdout, because spawn_child tells it to. Because in the standalone version, you only want the child to have stdout, not the parent. Unfortunately, in the inetd version, the parent still needs stdout, because (duh) that's where it writes to. The other is not to call note_cleanups_for_file in spawn_child() for inetd servers, but they should amount to the same thing. And because all inetd processes last for only one connection, I don't think it's too much of a problem not cleaning all the fds up properly. Anyhoo, here's the patch. While I think it might be a good idea to excise inetd support from future versions, for now, we still make a pretense at supporting it, and presumably CGI should work. Index: mod_cgi.c RCS file: /export/home/cvs/apache/src/mod_cgi.c,v retrieving revision 1.10 diff -c -r1.10 mod_cgi.c *** mod_cgi.c 1996/06/17 20:43:47 1.10 --- mod_cgi.c 1996/06/21 20:51:30
/* KLUDGE --- for back-combatibility, we don't have to check ExecCGI
! cleanup_for_exec();
Received on Fri Jun 21 13:56:30 1996This archive was generated by hypermail 2.1.8 : Thu Aug 24 2006 - 14:43:43 EDT |
||||||||||
|
|||||||||||