sockets: Dump socket queus without parasite

Since now we have local copies of a remove FDs we can dump socket queus without entering
a parasite code. This makes the code MUCH simpler.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

cr-dump.c

@@ -370,8 +370,7 @@ static int dump_one_chrdev(struct fd_parms *p, int lfd, const struct cr_fdset *s
 }
 
 static int dump_one_fd(pid_t pid, int fd, int lfd,
-		       const struct cr_fdset *cr_fdset,
-		       struct sk_queue *sk_queue)
+		       const struct cr_fdset *cr_fdset)
 {
 	struct fd_parms p;
 
@@ -381,7 +380,7 @@ static int dump_one_fd(pid_t pid, int fd, int lfd,
 	}
 
 	if (S_ISSOCK(p.stat.st_mode))
-		return dump_socket(&p, lfd, cr_fdset, sk_queue);
+		return dump_socket(&p, lfd, cr_fdset);
 
 	if (S_ISCHR(p.stat.st_mode))
 		return dump_one_chrdev(&p, lfd, cr_fdset);
@@ -396,7 +395,7 @@ static int dump_one_fd(pid_t pid, int fd, int lfd,
 }
 
 static int dump_task_files_seized(struct parasite_ctl *ctl, const struct cr_fdset *cr_fdset,
-				  int *fds, int nr_fds, struct sk_queue *sk_queue)
+				  int *fds, int nr_fds)
 {
 	int *lfds;
 	int i, ret = -1;
@@ -425,8 +424,7 @@ static int dump_task_files_seized(struct parasite_ctl *ctl, const struct cr_fdse
 	}
 
 	for (i = 0; i < nr_fds; i++) {
-		ret = dump_one_fd(ctl->pid, fds[i], lfds[i],
-				cr_fdset, sk_queue);
+		ret = dump_one_fd(ctl->pid, fds[i], lfds[i], cr_fdset);
 		close(lfds[i]);
 		if (ret)
 			goto err;
@@ -1299,7 +1297,6 @@ static int dump_one_task(const struct pstree_item *item)
 	struct parasite_ctl *parasite_ctl;
 	int ret = -1;
 	struct parasite_dump_misc misc;
-	struct sk_queue sk_queue = { };
 	struct cr_fdset *cr_fdset = NULL;
 
 	int nr_fds = PARASITE_MAX_FDS;
@@ -1349,7 +1346,7 @@ static int dump_one_task(const struct pstree_item *item)
 		goto err;
 	}
 
-	ret = dump_task_files_seized(parasite_ctl, cr_fdset, fds, nr_fds, &sk_queue);
+	ret = dump_task_files_seized(parasite_ctl, cr_fdset, fds, nr_fds);
 	if (ret) {
 		pr_err("Dump files (pid: %d) failed with %d\n", pid, ret);
 		goto err;
@@ -1391,12 +1388,6 @@ static int dump_one_task(const struct pstree_item *item)
 		goto err;
 	}
 
-	ret = parasite_dump_socket_info(parasite_ctl, cr_fdset, &sk_queue);
-	if (ret) {
-		pr_err("Can't dump socket info (pid: %d)\n", pid);
-		goto err;
-	}
-
 	ret = parasite_cure_seized(parasite_ctl);
 	if (ret) {
 		pr_err("Can't cure (pid: %d) from parasite\n", pid);

include/parasite-syscall.h

@@ -34,7 +34,6 @@ extern int parasite_dump_itimers_seized(struct parasite_ctl *ctl, struct cr_fdse
 
 struct parasite_dump_misc;
 extern int parasite_dump_misc_seized(struct parasite_ctl *ctl, struct parasite_dump_misc *misc);
-extern int parasite_dump_socket_info(struct parasite_ctl *ctl, struct cr_fdset *fdset, struct sk_queue *queue);
 extern int parasite_dump_pages_seized(struct parasite_ctl *ctl,
 				      struct list_head *vma_area_list,
 				      struct cr_fdset *cr_fdset);

include/parasite.h

@@ -31,7 +31,6 @@ enum {
 	PARASITE_CMD_DUMP_ITIMERS,
 	PARASITE_CMD_DUMP_MISC,
 	PARASITE_CMD_DUMP_TID_ADDR,
-	PARASITE_CMD_DUMP_SK_QUEUES,
 	PARASITE_CMD_DRAIN_FDS,
 
 	PARASITE_CMD_MAX,
@@ -81,12 +80,6 @@ struct parasite_dump_tid_addr {
 	unsigned int *tid_addr;
 };
 
-struct parasite_dump_sk_queues {
-	parasite_status_t	status;
-	unsigned		nr_items;
-	struct sk_queue_item	items[0];
-};
-
 #define PARASITE_MAX_FDS	(PAGE_SIZE / sizeof(int))
 
 struct parasite_drain_fd {

include/sockets.h

@@ -5,26 +5,10 @@
 #include <unistd.h>
 #include <stdbool.h>
 
-struct sk_queue_item {
-	int		fd;
-	int		type;
-	unsigned int	sk_id;
-};
-
-struct sk_queue_entry {
-	struct sk_queue_item item;
-	struct sk_queue_entry *next;
-};
-
-struct sk_queue {
-	unsigned int entries;
-	struct sk_queue_entry *list;
-};
-
 struct cr_fdset;
 struct fd_parms;
 extern int dump_socket(struct fd_parms *p, int lfd,
-		const struct cr_fdset *cr_fdset, struct sk_queue *queue);
+		const struct cr_fdset *cr_fdset);
 
 extern int collect_sockets(void);
 extern int prepare_sockets(int pid);

parasite-syscall.c

@@ -462,61 +462,6 @@ int parasite_dump_misc_seized(struct parasite_ctl *ctl, struct parasite_dump_mis
 				sizeof(struct parasite_dump_misc));
 }
 
-int parasite_dump_socket_info(struct parasite_ctl *ctl, struct cr_fdset *fdset,
-			      struct sk_queue *queue)
-{
-	int ret, i;
-	struct cr_fdset *fds;
-	unsigned arg_size;
-	struct parasite_dump_sk_queues *arg;
-	struct sk_queue_entry *sk_entry;
-
-	if (queue->entries == 0)
-		return 0;
-
-	pr_info("Dumping socket queues\n");
-
-	arg_size = sizeof(struct parasite_dump_sk_queues) +
-		queue->entries * sizeof(struct sk_queue_item);
-
-	/* FIXME arg size is only enough for ~1k of sockets */
-	if (arg_size > PARASITE_ARG_SIZE) {
-		pr_err("Too many sockets to drain queue from\n");
-		return -1;
-	}
-
-	ret = -1;
-	arg = xzalloc(arg_size);
-	if (arg == NULL)
-		goto err_alloc;
-
-	sk_entry = queue->list;
-	for (i = 0; i < queue->entries; i++, arg->nr_items++) {
-		struct sk_queue_entry *tmp = sk_entry;
-
-		memcpy(&arg->items[i], &sk_entry->item, sizeof(struct sk_queue_item));
-		sk_entry = tmp->next;
-		xfree(tmp);
-	}
-
-	ret = parasite_prep_file(fdset_fd(glob_fdset, CR_FD_SK_QUEUES), ctl);
-	if (ret < 0)
-		goto err_prepf;
-
-	ret = parasite_execute(PARASITE_CMD_DUMP_SK_QUEUES, ctl,
-			(parasite_status_t *)arg, arg_size);
-	if (ret < 0)
-		goto err_exec;
-
-	return 0;
-
-err_exec:
-err_prepf:
-	xfree(arg);
-err_alloc:
-	return -1;
-}
-
 /*
  * This routine drives parasite code (been previously injected into a victim
  * process) and tells it to dump pages into the file.

parasite.c

@@ -374,126 +374,6 @@ static int dump_tid_addr(struct parasite_dump_tid_addr *args)
 	return 0;
 }
 
-static int dump_socket_queue(int img_fd, struct sk_queue_item *item)
-{
-	struct sk_packet_entry *pe;
-	unsigned long size;
-	socklen_t tmp;
-	int ret, orig_peek_off;
-	int sock_fd = item->fd;
-
-	/*
-	 * Save original peek offset. 
-	 */
-	ret = sys_getsockopt(sock_fd, SOL_SOCKET, SO_PEEK_OFF, &orig_peek_off, &tmp);
-	if (ret < 0) {
-		sys_write_msg("getsockopt failed\n");
-		return ret;
-	}
-	/*
-	 * Discover max DGRAM size
-	 */
-	ret = sys_getsockopt(sock_fd, SOL_SOCKET, SO_SNDBUF, &ret, &tmp);
-	if (ret < 0) {
-		sys_write_msg("getsockopt failed\n");
-		return ret;
-	}
-	/*
-	 * Note: 32 bytes will be used by kernel for protocol header.
-	 */
-	size = ret - 32;
-	/*
-	 * Try to alloc buffer for max supported DGRAM + our header.
-	 * Note: STREAM queue will be written by chunks of this size.
-	 */
-	pe = brk_alloc(size + sizeof(struct sk_packet_entry));
-	if (!pe) {
-		sys_write_msg("not enough mem for skb\n");
-		return -ENOMEM;
-	}
-	/*
-	 * Enable peek offset incrementation.
-	 */
-	ret = sys_setsockopt(sock_fd, SOL_SOCKET, SO_PEEK_OFF, &ret, sizeof(int));
-	if (ret < 0) {
-		sys_write_msg("setsockopt fail\n");
-		goto err_brk;
-	}
-
-	pe->id_for = item->sk_id;
-
-	while (1) {
-		struct iovec iov = {
-			.iov_base	= pe->data,
-			.iov_len	= size,
-		};
-		struct msghdr msg = {
-			.msg_iov	= &iov,
-			.msg_iovlen	= 1,
-		};
-
-		ret = pe->length = sys_recvmsg(sock_fd, &msg, MSG_DONTWAIT | MSG_PEEK);
-		if (ret < 0) {
-			if (ret == -EAGAIN)
-				break; /* we're done */
-			sys_write_msg("sys_recvmsg fail: error\n");
-			goto err_set_sock;
-		}
-		if (msg.msg_flags & MSG_TRUNC) {
-			/*
-			 * DGRAM thuncated. This should not happen. But we have
-			 * to check...
-			 */
-			sys_write_msg("sys_recvmsg failed: truncated\n");
-			ret = -E2BIG;
-			goto err_set_sock;
-		}
-		ret = sys_write(img_fd, pe, sizeof(pe) + pe->length);
-		if (ret != sizeof(pe) + pe->length) {
-			sys_write_msg("sys_write failed\n");
-			ret = -EIO;
-			goto err_set_sock;
-		}
-	}
-	ret = 0;
-
-err_set_sock:
-	/*
-	 * Restore original peek offset. 
-	 */
-	ret = sys_setsockopt(sock_fd, SOL_SOCKET, SO_PEEK_OFF, &orig_peek_off, sizeof(int));
-	if (ret < 0)
-		sys_write_msg("setsockopt failed on restore\n");
-err_brk:
-	brk_free(size + sizeof(struct sk_packet_entry));
-	return ret;
-}
-
-static int dump_skqueues(struct parasite_dump_sk_queues *args)
-{
-	parasite_status_t *st = &args->status;
-	int img_fd, i, ret = -1;
-
-	img_fd = recv_fd(tsock);
-	if (img_fd < 0) {
-		SET_PARASITE_RET(st, img_fd);
-		return img_fd;
-	}
-
-	for (i = 0; i < args->nr_items; i++) {
-		ret = dump_socket_queue(img_fd, &args->items[i]);
-		if (ret < 0) {
-			SET_PARASITE_RET(st, ret);
-			goto err_dmp;
-		}
-	}
-
-	ret = 0;
-err_dmp:
-	sys_close(img_fd);
-	return ret;
-}
-
 static int drain_fds(struct parasite_drain_fd *args)
 {
 	parasite_status_t *st = &args->status;
@@ -577,7 +457,6 @@ static int __used parasite_service(unsigned long cmd, void *args)
 	BUILD_BUG_ON(sizeof(struct parasite_init_args) > PARASITE_ARG_SIZE);
 	BUILD_BUG_ON(sizeof(struct parasite_dump_misc) > PARASITE_ARG_SIZE);
 	BUILD_BUG_ON(sizeof(struct parasite_dump_tid_addr) > PARASITE_ARG_SIZE);
-	BUILD_BUG_ON(sizeof(struct parasite_dump_sk_queues) > PARASITE_ARG_SIZE);
 	BUILD_BUG_ON(sizeof(struct parasite_drain_fd) > PARASITE_ARG_SIZE);
 
 	switch (cmd) {
@@ -601,8 +480,6 @@ static int __used parasite_service(unsigned long cmd, void *args)
 		return dump_misc((struct parasite_dump_misc *)args);
 	case PARASITE_CMD_DUMP_TID_ADDR:
 		return dump_tid_addr((struct parasite_dump_tid_addr *)args);
-	case PARASITE_CMD_DUMP_SK_QUEUES:
-		return dump_skqueues((struct parasite_dump_sk_queues *)args);
 	case PARASITE_CMD_DRAIN_FDS:
 		return drain_fds((struct parasite_drain_fd *)args);
 	default:

sockets.c

-#include <sys/socket.h>
 #include <linux/netlink.h>
 #include <linux/types.h>
 #include <linux/net.h>
-#include <linux/un.h>
 #include <sys/types.h>
 #include <sys/vfs.h>
 #include <sys/stat.h>
@@ -23,6 +21,7 @@
 #include "util.h"
 #include "inet_diag.h"
 #include "files.h"
+#include "util-net.h"
 
 static char buf[4096];
 
@@ -87,23 +86,98 @@ struct inet_sk_desc {
 	unsigned int		dst_addr[4];
 };
 
-static int unix_sk_queue_add(int fd, const struct unix_sk_desc *sd,
-			     struct sk_queue *queue)
+static int dump_socket_queue(int sock_fd, int sock_id)
 {
-	struct sk_queue_entry *next, *new;
+	struct sk_packet_entry *pe;
+	unsigned long size;
+	socklen_t tmp;
+	int ret, orig_peek_off;
 
-	new = xmalloc(sizeof(struct sk_queue_entry));
-	if (!new)
-		return -1;
+	/*
+	 * Save original peek offset. 
+	 */
+	tmp = sizeof(orig_peek_off);
+	ret = getsockopt(sock_fd, SOL_SOCKET, SO_PEEK_OFF, &orig_peek_off, &tmp);
+	if (ret < 0) {
+		pr_perror("getsockopt failed\n");
+		return ret;
+	}
+	/*
+	 * Discover max DGRAM size
+	 */
+	tmp = sizeof(size);
+	ret = getsockopt(sock_fd, SOL_SOCKET, SO_SNDBUF, &size, &tmp);
+	if (ret < 0) {
+		pr_perror("getsockopt failed\n");
+		return ret;
+	}
 
-	new->item.fd = fd;
-	new->item.type = sd->type;
-	new->item.sk_id = sd->sd.ino;
-	new->next = queue->list;
+	/* Note: 32 bytes will be used by kernel for protocol header. */
+	size -= 32;
+	/*
+	 * Try to alloc buffer for max supported DGRAM + our header.
+	 * Note: STREAM queue will be written by chunks of this size.
+	 */
+	pe = xmalloc(size + sizeof(struct sk_packet_entry));
+	if (!pe)
+		return -ENOMEM;
 
-	queue->list = new;
-	queue->entries++;
-	return 0;
+	/*
+	 * Enable peek offset incrementation.
+	 */
+	ret = setsockopt(sock_fd, SOL_SOCKET, SO_PEEK_OFF, &ret, sizeof(int));
+	if (ret < 0) {
+		pr_perror("setsockopt fail\n");
+		goto err_brk;
+	}
+
+	pe->id_for = sock_id;
+
+	while (1) {
+		struct iovec iov = {
+			.iov_base	= pe->data,
+			.iov_len	= size,
+		};
+		struct msghdr msg = {
+			.msg_iov	= &iov,
+			.msg_iovlen	= 1,
+		};
+
+		ret = pe->length = recvmsg(sock_fd, &msg, MSG_DONTWAIT | MSG_PEEK);
+		if (ret < 0) {
+			if (ret == -EAGAIN)
+				break; /* we're done */
+			pr_perror("sys_recvmsg fail: error\n");
+			goto err_set_sock;
+		}
+		if (msg.msg_flags & MSG_TRUNC) {
+			/*
+			 * DGRAM thuncated. This should not happen. But we have
+			 * to check...
+			 */
+			pr_err("sys_recvmsg failed: truncated\n");
+			ret = -E2BIG;
+			goto err_set_sock;
+		}
+		ret = write_img_buf(fdset_fd(glob_fdset, CR_FD_SK_QUEUES),
+				pe, sizeof(pe) + pe->length);
+		if (ret < 0) {
+			ret = -EIO;
+			goto err_set_sock;
+		}
+	}
+	ret = 0;
+
+err_set_sock:
+	/*
+	 * Restore original peek offset. 
+	 */
+	ret = setsockopt(sock_fd, SOL_SOCKET, SO_PEEK_OFF, &orig_peek_off, sizeof(int));
+	if (ret < 0)
+		pr_perror("setsockopt failed on restore\n");
+err_brk:
+	xfree(pe);
+	return ret;
 }
 
 #define SK_HASH_SIZE		32
@@ -240,8 +314,7 @@ static int can_dump_inet_sk(const struct inet_sk_desc *sk)
 }
 
 static int dump_one_inet(struct socket_desc *_sk, int fd,
-			 const struct cr_fdset *cr_fdset,
-			 struct sk_queue *queue)
+			 const struct cr_fdset *cr_fdset)
 {
 	struct inet_sk_desc *sk = (struct inet_sk_desc *)_sk;
 	struct inet_sk_entry ie;
@@ -311,9 +384,8 @@ static int can_dump_unix_sk(const struct unix_sk_desc *sk)
 	return 1;
 }
 
-static int dump_one_unix(const struct socket_desc *_sk, int fd,
-			 const struct cr_fdset *cr_fdset,
-			 struct sk_queue *queue)
+static int dump_one_unix(const struct socket_desc *_sk, int fd, int lfd,
+			 const struct cr_fdset *cr_fdset)
 {
 	const struct unix_sk_desc *sk = (struct unix_sk_desc *)_sk;
 	struct unix_sk_entry ue;
@@ -369,7 +441,7 @@ static int dump_one_unix(const struct socket_desc *_sk, int fd,
 
 	if (sk->rqlen != 0 && !(sk->type == SOCK_STREAM &&
 				sk->state == TCP_LISTEN))
-		if (unix_sk_queue_add(fd, sk, queue))
+		if (dump_socket_queue(lfd, ue.id))
 			goto err;
 
 	pr_info("Dumping unix socket at %d\n", fd);
@@ -382,8 +454,7 @@ err:
 	return -1;
 }
 
-int dump_socket(struct fd_parms *p, int lfd, const struct cr_fdset *cr_fdset,
-		    struct sk_queue *queue)
+int dump_socket(struct fd_parms *p, int lfd, const struct cr_fdset *cr_fdset)
 {
 	struct socket_desc *sk;
 	struct statfs fst;
@@ -398,9 +469,9 @@ int dump_socket(struct fd_parms *p, int lfd, const struct cr_fdset *cr_fdset,
 
 	switch (sk->family) {
 	case AF_UNIX:
-		return dump_one_unix(sk, p->fd_name, cr_fdset, queue);
+		return dump_one_unix(sk, p->fd_name, lfd, cr_fdset);
 	case AF_INET:
-		return dump_one_inet(sk, p->fd_name, cr_fdset, queue);
+		return dump_one_inet(sk, p->fd_name, cr_fdset);
 	default:
 		pr_err("BUG! Unknown socket collected\n");
 		break;