httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rob Hartill <hart...@ooo.lanl.gov>
Subject background CGI bug
Date Mon, 06 May 1996 16:06:45 GMT

oh what fun.

http://hyperreal.com/httpd/patches/for_Apache_1.1b0/110.buff.patch

is the cause, but that's a 50k patch and as yet I've not been able to
trace the problem, however I observed the following which might
trigger a brain cell in somwone out there...

apache is reading and sending the cgi output fine, the child process
does its stuff and jumps back to its main loop and sits on

   csd = select(listenmaxfd+1, &fds, NULL, NULL, NULL);

for the next connection (the keepalive flag is 0 and the keepalive
loop is skipped). the child returns to the above line 'instantly'.. no
noticeable delays.

In the meantime the connection to the client stays open, despite a
call to bclose(conn_io) which returns with "0" (ok).

So, should "bclose(conn_io)" break the connection with the client?
I'd think so, but it isn't    :-(


rob
--
   Windows NT  -  where programs go to die.

Mime
View raw message