Actually, there are some significant problems with the rtsig implementation.  
For example, it doesn't handle pipes or tty's and can miss several events.

Attached is a version of rtsig.c that is mostly a rewrite.  The code isn't that 
clean, but it has been working flawlessly for me in many high-volume 
applications for over a year now.

This code is slower than the poll-based version until you hit around 40 
connections ( on my hardware ).  More connections than that, that the 
performance of the poll-based code decay's rapidly and rtsig remains constant.


 
On Saturday, February 25, 2006, at 03:02PM, Taral <[EMAIL PROTECTED]> wrote:

>It seems nobody uses rtsig... maybe if it were enabled by default
>people would test it?
>
>Anyway, here is a patch that makes rtsig compile again after the
>thread-safety changes.
>
>Please make rtsig on by default. It works! :)
>
>--
>Taral <[EMAIL PROTECTED]>
>"Computer science is no more about computers than astronomy is about
>telescopes."
>    -- Edsger Dijkstra
>
>diff -ruN libevent-1.1a/event-internal.h libevent-1.1a.new/event-internal.h
>--- libevent-1.1a/event-internal.h     Thu Jan 13 22:24:41 2005
>+++ libevent-1.1a.new/event-internal.h Wed Dec 28 15:06:51 2005
>@@ -31,6 +31,8 @@
> extern "C" {
> #endif
> 
>+#include <sys/tree.h>
>+
> struct event_base {
>       const struct eventop *evsel;
>       void *evbase;
>diff -ruN libevent-1.1a/rtsig.c libevent-1.1a.new/rtsig.c
>--- libevent-1.1a/rtsig.c      Tue Mar 29 00:55:42 2005
>+++ libevent-1.1a.new/rtsig.c  Wed Dec 28 15:04:50 2005
>@@ -28,6 +28,7 @@
> #define EVLIST_X_NORT 0x1000  /* Skip RT signals (internal) */
> 
> #include "event.h"
>+#include "event-internal.h"
> #include "log.h"
> extern struct event_list signalqueue;
> 
>
>_______________________________________________
>Libevent-users mailing list
>[email protected]
>http://monkey.org/mailman/listinfo/libevent-users
>
>

Attachment: rtsig.c
Description: Binary data

_______________________________________________
Libevent-users mailing list
[email protected]
http://monkey.org/mailman/listinfo/libevent-users

Reply via email to