Messages in this thread |  | | From | "David Schwartz" <> | Date | Mon, 23 Oct 2000 14:42:49 -0700 |
| |
> From: "David Schwartz" <davids@webmaster.com> > Date: Mon, 23 Oct 2000 13:01:29 -0700 > > Under Solaris 7, when the number of idle sockets was decreased from > 10,000 to 100, the time to check for active sockets with poll() > decreased by a factor of only 6.5. Shouldn't it be more like 100? > Sounds like Solaris' poll implementation is horribly broken for the > most common case. > > I would like to ask everyone interested so deeply in this to read up > on /usr/include/sys/poll_impl.h from such a Solaris system to see what > their optimization actually is before continuing this discussion any > further, thanks :-)
I'm quite familiar with it. I was mocking the original poster's comments, not Solaris' implementation decisions.
I really wish Linux had an implementation of /dev/poll. (Yes, I do know that there's a patch available)
DS
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
|  |