shouldn't we RETURN in that null pointer reference fix. Since I get
[Thu Feb 08 01:50:30 2007] [warn] (9)Bad file descriptor: 
apr_socket_opt_set: (TCP_NODELAY)

with stefan's fix.
But if i check if !current_conn and then return everything looks ok.



Sean Gabriel Heacock wrote:
> http://www.telana.com/files/httpd-2.2.3-peruser-0.2.1.patch
>
> I haven't had a chance to test it out on my production server yet, but
> it seems to run fine on my test machine.  I'll try it out for real
> tonight.  Changes since 0.2.0:
>
> * Stefan's mod_status patch
> * Janno's IdleTimeout patch with code from Rommer's ExpireTimeout patch
> * support for APR 1.2 (probably requires Apache 2.2 to compile now)
> * a null pointer dereference fix Stefan commited last August :)
> * removed the sleep(1) loop from pass_socket() (already removed from
> pass_request() in version 0.2.0)
>
> The subversion repository is up to date with 0.2.1.  Stefan already has
> an account; Janno, Rommer, if either of you want access, let me know.
>
> This version will probably still compile on Apache 2.0 if you back out
> of these two changes:
>
> -    ret = apr_recv(lr->sd, &pipe_read_char, &n);
> +    ret = apr_socket_recv(lr->sd, &pipe_read_char, &n);
>
> and
>
> -           else if (apr_proc_other_child_read(&pid, status) == 0) {
> +           else if (apr_proc_other_child_alert(&pid, APR_OC_REASON_DEATH, 
> status) == APR_SUCCESS) {
>
>   

_______________________________________________
Peruser mailing list
[email protected]
http://www.telana.com/mailman/listinfo/peruser

Reply via email to