Re: [PATCH 13/18] io_uring: add file set registration

[Jens Axboe]

On 2/6/19 9:05 PM, Al Viro wrote:
> On Wed, Feb 06, 2019 at 10:56:41AM -0700, Jens Axboe wrote:
>> On 2/5/19 6:01 PM, Al Viro wrote:
>>> On Tue, Feb 05, 2019 at 05:27:29PM -0700, Jens Axboe wrote:
>>>
>>>> This should be better, passes some basic testing, too:
>>>>
>>>> http://git.kernel.dk/cgit/linux-block/commit/?h=io_uring&id=01a93aa784319a02ccfa6523371b93401c9e0073
>>>>
>>>> Verified that we're grabbing the right refs, and don't hold any
>>>> ourselves. For the file registration, forbid registration of the
>>>> io_uring fd, as that is pointless and will introduce a loop regardless
>>>> of fd passing.
>>>
>>> *shrug*
>>>
>>> So pass it to AF_UNIX socket and register _that_ - does't change the
>>> underlying problem.
>>
>> Maybe I'm being dense here, but it's an f_op match. Should catch a
>> passed fd as well, correct?
> 
> f_op match on _what_?
> 
>> With that, how can there be a loop?
> 
>       io_uring_fd = ....
>       socketpair(PF_UNIX, SOCK_STREAM, 0, sock_fds);
>       register sock_fds[0] and sock_fds[1] to io_uring_fd
>       send SCM_RIGHTS datagram with io_uring_fd to sock_fds[0]
>       close sock_fds[0], sock_fds[1] and io_uring_fd
> 
> And there's your unreachable loop.

I created a small app to do just that, and ran it and verified that
->release() is called and the io_uring is released as expected. This
is run on the current -git branch, which has a socket backing for
the io_uring fd itself, but not for the registered files.

What am I missing here? Attaching the program as a reference.

-- 
Jens Axboe


#include <stdio.h>
#include <stdlib.h>
#include <stddef.h>
#include <signal.h>
#include <inttypes.h>
#include <sys/types.h>
#include <sys/syscall.h>
#include <sys/socket.h>
#include <sys/wait.h>
#include <fcntl.h>
#include <unistd.h>
#include <string.h>
#include <linux/fs.h>

struct io_sqring_offsets {
	__u32 head;
	__u32 tail;
	__u32 ring_mask;
	__u32 ring_entries;
	__u32 flags;
	__u32 dropped;
	__u32 array;
	__u32 resv[3];
};

struct io_cqring_offsets {
	__u32 head;
	__u32 tail;
	__u32 ring_mask;
	__u32 ring_entries;
	__u32 overflow;
	__u32 cqes;
	__u32 resv[4];
};

struct io_uring_params {
	__u32 sq_entries;
	__u32 cq_entries;
	__u32 flags;
	__u32 sq_thread_cpu;
	__u32 sq_thread_idle;
	__u32 resv[5];
	struct io_sqring_offsets sq_off;
	struct io_cqring_offsets cq_off;
};

#define IORING_REGISTER_FILES		2

#define __NR_sys_io_uring_setup		425
#define __NR_sys_io_uring_register	427

static int io_uring_register_files(int ring_fd, int fd1, int fd2)
{
	__s32 *fds;

	fds = calloc(2, sizeof(__s32));
	fds[0] = fd1;
	fds[1] = fd2;

	return syscall(__NR_sys_io_uring_register, ring_fd,
			IORING_REGISTER_FILES, fds, 2);
}

static int io_uring_setup(unsigned entries, struct io_uring_params *p)
{
	return syscall(__NR_sys_io_uring_setup, entries, p);
}

static int get_ring_fd(void)
{
	struct io_uring_params p;
	int fd;

	memset(&p, 0, sizeof(p));

	fd = io_uring_setup(2, &p);
	if (fd < 0) {
		perror("io_uring_setup");
		return -1;
	}

	return fd;
}

static void send_fd(int socket, int fd)
{
	char buf[CMSG_SPACE(sizeof(fd))];
	struct cmsghdr *cmsg;
	struct msghdr msg;

	memset(buf, 0, sizeof(buf));
	memset(&msg, 0, sizeof(msg));

	msg.msg_control = buf;
	msg.msg_controllen = sizeof(buf);

	cmsg = CMSG_FIRSTHDR(&msg);
	cmsg->cmsg_level = SOL_SOCKET;
	cmsg->cmsg_type = SCM_RIGHTS;
	cmsg->cmsg_len = CMSG_LEN(sizeof(fd));

	memmove(CMSG_DATA(cmsg), &fd, sizeof(fd));

	msg.msg_controllen = CMSG_SPACE(sizeof(fd));

	if (sendmsg(socket, &msg, 0) < 0)
		perror("sendmsg");
}

static int recv_fd(int socket)
{
	struct msghdr msg;
	char c_buffer[256];
	struct cmsghdr *cmsg;
	int fd;

	memset(&msg, 0, sizeof(msg));

	msg.msg_control = c_buffer;
	msg.msg_controllen = sizeof(c_buffer);

	if (recvmsg(socket, &msg, 0) < 0)
		perror("recvmsg\n");

	cmsg = CMSG_FIRSTHDR(&msg);
	memmove(&fd, CMSG_DATA(cmsg), sizeof(fd));
	return fd;
}

int main(int argc, char *argv[])
{
	int sp[2], pid, ring_fd, ret;

	if (socketpair(AF_UNIX, SOCK_DGRAM, 0, sp) != 0) {
		perror("Failed to create Unix-domain socket pair\n");
		return 1;
	}

	ring_fd = get_ring_fd();
	if (ring_fd < 0)
		return 1;

	ret = io_uring_register_files(ring_fd, sp[0], sp[1]);
	if (ret < 0) {
		perror("register files");
		return 1;
	}

	pid = fork();
	if (pid) {
		printf("Sending fd %d\n", ring_fd);

		send_fd(sp[0], ring_fd);
	} else {
		int fd;

		fd = recv_fd(sp[1]);
		printf("Got fd %d\n", fd);
		close(fd);
	}

	usleep(500000);
	close(ring_fd);
	close(sp[0]);
	close(sp[1]);
	return 0;
}