search for: epoll_ctl_mod

I was straceing a pop3 process and noticed that the TCP_CORK option isn't set soon enough: epoll_wait(8, {{EPOLLOUT, {u32=37481984, u64=37481984}}}, 38, 207) = 1 write(41, "iTxPBrNlaNFao+yQzLhuO4/+tQ5cuiKSe"..., 224) = 224 epoll_ctl(8, EPOLL_CTL_MOD, 41, {EPOLLIN|EPOLLPRI|EPOLLERR|EPOLLHUP, {u32=37481984, u64=37481984}}) = 0 pread(19, "AFABQAlAC0AJ\nQAUALQAUAFABQAlAC0AF"..., 8192, 811008) = 8192 setsockopt(41, SOL_TCP, TCP_CORK, [1], 4) = 0 write(41, "\r\nKUWtGCjKO5N8UbW5uYLZbS0nmaNi4ZB"..., 4134) = 4134 write(41, "\r\...

...00000004050d6 in main (argc=1, argv=0x2039390) at main.c:123 strace shows this: read(11, "\27\3\1\0 q\r\252\3551\21\237l\33\330\33\303\340\306l\334k\0360p\303)HF\331\234g"..., 1261) = 37 read(11, 0xf09908, 1224) = -1 EAGAIN (Resource temporarily unavailable) epoll_ctl(10, EPOLL_CTL_MOD, 11, {EPOLLIN|EPOLLPRI|EPOLLOUT|EPOLLERR|EPOLLHUP, {u32=15754432, u64=15754432}}) = 0 epoll_ctl(10, EPOLL_CTL_MOD, 11, {EPOLLIN|EPOLLPRI|EPOLLERR|EPOLLHUP, {u32=15754432, u64=15754432}}) = 0 write(11, "\27\3\1\0 \204\375\373!\253\263n\204\274duj/\n\202\236\373\342\303[\22\17\264\10\23\346\225&...

...ta *data = ioloop->handler_data; @@ -155,7 +93,7 @@ void io_loop_handle_add(struct ioloop *i first = iolist_add(list, io); event.data.ptr = list; - event.events = epoll_event_mask(list); + event.events = iolist_events(list, IO_EPOLL_INPUT, IO_EPOLL_OUTPUT); op = first ? EPOLL_CTL_ADD : EPOLL_CTL_MOD; @@ -184,7 +122,7 @@ void io_loop_handle_remove(struct ioloop last = iolist_del(list, io); event.data.ptr = list; - event.events = epoll_event_mask(list); + event.events = iolist_events(list, IO_EPOLL_INPUT, IO_EPOLL_OUTPUT); op = last ? EPOLL_CTL_DEL : EPOLL_CTL_MOD; @@ -224,7 +162,7...

...eof(int) * old_size, + sizeof(int) * data->idx_size); + } + + data->fd_index[fd] = list = + p_new(ioloop->pool, struct io_list, 1); + } + + iolist_add(list, io); + + event.data.ptr = list; + event.events = epoll_event_mask(list); + + op = (list->count == 1) ? EPOLL_CTL_ADD : EPOLL_CTL_MOD; + + ret = epoll_ctl(data->epfd, op, fd, &event); + if (ret < 0) + i_warning("epoll_ctl() : %m"); + + if (data->events_pos >= data->events_size) { + data->events_size = nearest_power(data->events_size + 1); + + p_free(ioloop->pool, data->events); + data-...

...--\nM"..., 4096) = 426 > close(24) = 0 > munmap(0x7f9f1978d000, 4096) = 0 > epoll_wait(14, {{EPOLLOUT, {u32=23, u64=23}}}, 32, 4907) = 1 > sendto(23, "0 tecra 17.0\n1 94 64 0 0 A2B583B"..., 538, 0, NULL, 0) = 538 > epoll_ctl(14, EPOLL_CTL_MOD, 23, {EPOLLIN, {u32=23, u64=23}}) = 0 > epoll_wait(14, 69d660, 32, 4904) = -1 EINTR (Interrupted system call) > --- SIGALRM (Alarm clock) @ 0 (0) --- > sendto(15, "\16", 1, 0, NULL, 0) = 1 > rt_sigreturn(0xf) = -1 EINTR (Interrupted system c...

Greetings, This email is both a request for assistance/help and a heads-up. [8irgehuq] CVE-2011-1083: Algorithmic denial of service in epoll. After ksplice automatically installed the above patch on our mail servers, most/all IMAP/POP3 connections began experiencing time-outs trying to connect, or extreme timeouts in the auth procedure. dovecot: imap-login: Disconnected (no auth attempts):

I occasionally get a fatal error unable to delete epoll event: Bad file descriptor I think the attached patch will fix it. The patch does two things: 1) changes the error we look for from ENOENT to EBADF. It is hard to reproduce this error, but if I change the epoll_ctl line to epoll_ctl(epfd, EPOLL_CTL_DEL, 666, ed->GetEpollEvent()); it does return -1 with errno = EBADF on my Linux 2.6.9

hi, I've some issue with geo-replication (since 3.2.0) and nfs auth (since initial release). Geo-replication --------------- System : Debian 6.0 amd64 Glusterfs: 3.2.0 MASTER (volume) => SLAVE (directory) For some volume it works, but for others i can't enable geo-replication and have this error with a faulty status: 2011-05-03 09:57:40.315774] E