Messages in this thread | | | Date | Fri, 8 Nov 2002 09:43:46 -0800 (PST) | From | Davide Libenzi <> | Subject | Re: [PATCH] 2.5.46: epoll ep_insert doesn't wake waiters if events exist |
| |
On Fri, 8 Nov 2002, Mark Hamblin wrote:
> Looking at the documentation at > http://www.xmailserver.org/linux-patches/epoll_wait.txt, I noticed that the > description for epoll_wait states: "On success, epoll_wait(2) returns the > number of file descriptors ready for the requested I/O." My question is if > I have, for example, 1000 sockets registered with epoll and 100 of them have > received data and I call epoll_wait with max_events set to 10, will > epoll_wait return 10 or 100? Furthermore, does the edge-triggered nature of > epoll "eat the edge" for the other 90 sockets even though they didn't get > returned? Finally, if those 10 sockets get more data before I call > epoll_wait with max_events = 10 again, will those same 10 sockets get > returned?
epoll will return 10 and the remaining 90 will remain available for future epoll_wait() calls. The last question will trigger a change in my code, from list_add() to list_add_tail(). Since I don't want to stress Linus with 24590 patches, I'll wait for 2.5.47 to merge the new code ...
- Davide
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |