public inbox for [email protected]
 help / color / mirror / Atom feed
* Odd timeout behavior
@ 2020-04-11 23:00 Hrvoje Zeba
  2020-04-12  2:07 ` Jens Axboe
  0 siblings, 1 reply; 14+ messages in thread
From: Hrvoje Zeba @ 2020-04-11 23:00 UTC (permalink / raw)
  To: io-uring

[-- Attachment #1: Type: text/plain, Size: 885 bytes --]

Hi,

I've been looking at timeouts and found a case I can't wrap my head around.

Basically, If you submit OPs in a certain order, timeout fires before
time elapses where I wouldn't expect it to. The order is as follows:

poll(listen_socket, POLLIN) <- this never fires
nop(async)
timeout(1s, count=X)

If you set X to anything but 0xffffffff/(unsigned)-1, the timeout does
not fire (at least not immediately). This is expected apart from maybe
setting X=1 which would potentially allow the timeout to fire if nop
executes after the timeout is setup.

If you set it to 0xffffffff, it will always fire (at least on my
machine). Test program I'm using is attached.

The funny thing is that, if you remove the poll, timeout will not fire.

I'm using Linus' tree (v5.6-12604-gab6f762f0f53).

Could anybody shine a bit of light here?


Thank you,
Hrvoje

-- 
I doubt, therefore I might be.

[-- Attachment #2: nop-timeout.c --]
[-- Type: text/x-csrc, Size: 2205 bytes --]

#include <assert.h>
#include <errno.h>
#include <stdio.h>
#include <unistd.h>
#include <stdlib.h>
#include <string.h>
#include <fcntl.h>
#include <time.h>
#include <poll.h>
#include <sys/time.h>
#include <sys/socket.h>
#include <arpa/inet.h>

#include "liburing.h"

struct io_uring_sqe* get_sqe(struct io_uring* ring)
{
	struct io_uring_sqe* sqe = io_uring_get_sqe(ring);
	assert(sqe != NULL);
	return sqe;
}

void enqueue_nop(struct io_uring* ring)
{
	struct io_uring_sqe* sqe = get_sqe(ring);

	io_uring_prep_nop(sqe);
	io_uring_sqe_set_data(sqe, (void*)1);
	io_uring_sqe_set_flags(sqe, IOSQE_ASYNC);
}

void enqueue_timeout(struct io_uring* ring)
{
	struct io_uring_sqe* sqe = get_sqe(ring);
	static struct __kernel_timespec ts;

	ts.tv_sec = 1;
	ts.tv_nsec = 0;

	io_uring_prep_timeout(sqe, &ts, (unsigned)-1, 0);
	io_uring_sqe_set_data(sqe, (void*)2);
}

void enqueue_poll(struct io_uring* ring, int fd)
{
	struct io_uring_sqe* sqe = get_sqe(ring);

	io_uring_prep_poll_add(sqe, fd, POLLIN | POLLERR | POLLHUP);
	io_uring_sqe_set_data(sqe, (void*)3);
}

int create_socket()
{
	int s = socket(AF_INET, SOCK_STREAM | SOCK_CLOEXEC, IPPROTO_TCP);
	assert(s != -1);

	int flags = fcntl(s, F_GETFL, 0);
	assert(flags != -1);

	flags |= O_NONBLOCK;

	assert(fcntl(s, F_SETFL, flags) != -1);

	struct sockaddr_in addr;

	addr.sin_family = AF_INET;
	addr.sin_port = 0x1236;
	addr.sin_addr.s_addr = 0x0100007fU;

	assert(bind(s, (struct sockaddr*)&addr, sizeof(addr)) != -1);
	assert(listen(s, 1024) != -1);

	return s;
}

int main(int argc, char *argv[])
{
	struct io_uring ring;
	int ret;

	ret = io_uring_queue_init(4, &ring, 0);
	if (ret) {
		fprintf(stderr, "ring setup failed\n");
		return 1;
	}

	int s = create_socket();
	enqueue_poll(&ring, s);

	enqueue_nop(&ring);
	enqueue_timeout(&ring);

	ret = io_uring_submit_and_wait(&ring, 1);
	if (ret == -1) {
		fprintf(stderr, "submit failed\n");
		return 1;
	}

	struct io_uring_cqe* cqe;
	uint32_t head;
	uint32_t count = 0;

	io_uring_for_each_cqe(&ring, head, cqe) {
		if (io_uring_cqe_get_data(cqe) == (void*)2)
			fprintf(stderr, "Timeout triggered!\n");

		count++;
	}

	io_uring_cq_advance(&ring, count);

	io_uring_queue_exit(&ring);
	return 0;
}

^ permalink raw reply	[flat|nested] 14+ messages in thread
end of thread, other threads:[~2020-04-17 14:37 UTC | newest]

Thread overview: 14+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2020-04-11 23:00 Odd timeout behavior Hrvoje Zeba
2020-04-12  2:07 ` Jens Axboe
2020-04-12  9:15   ` Pavel Begunkov
2020-04-12 14:40     ` Jens Axboe
2020-04-17  8:39       ` Pavel Begunkov
2020-04-17 14:37         ` Hrvoje Zeba
2020-04-12 15:14     ` Hrvoje Zeba
2020-04-13  8:21       ` Pavel Begunkov
2020-04-13 14:16         ` Jens Axboe
2020-04-13 19:09           ` Pavel Begunkov
2020-04-14  0:44             ` Jens Axboe
2020-04-14 15:46               ` Pavel Begunkov
2020-04-14 16:04                 ` Jens Axboe
2020-04-14 16:31                   ` Pavel Begunkov

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox