Merge pull request #715 from justincormack/runc-vendor-up

Update the opencontainers/runc vendor

vendor.conf

@@ -11,7 +11,7 @@ github.com/matttproud/golang_protobuf_extensions v1.0.0
 github.com/docker/go-units v0.3.1
 github.com/gogo/protobuf d2e1ade2d719b78fe5b061b4c18a9f7111b5bdc8
 github.com/golang/protobuf 8ee79997227bf9b34611aee7946ae64735e6fd93
-github.com/opencontainers/runc ce450bcc6c135cae93ee2a99d41a308c179ff6dc
+github.com/opencontainers/runc 50401b5b4c2e01e4f1372b73a021742deeaf4e2d
 github.com/opencontainers/runtime-spec 035da1dca3dfbb00d752eb58b0b158d6129f3776
 github.com/Sirupsen/logrus v0.11.0
 github.com/stevvooe/go-btrfs ea304655a3ed8f00773db1844f921d12541ee0d1

vendor/github.com/opencontainers/runc/README.md

@@ -2,6 +2,7 @@
 
 [![Build Status](https://travis-ci.org/opencontainers/runc.svg?branch=master)](https://travis-ci.org/opencontainers/runc)
 [![Go Report Card](https://goreportcard.com/badge/github.com/opencontainers/runc)](https://goreportcard.com/report/github.com/opencontainers/runc)
+[![GoDoc](https://godoc.org/github.com/opencontainers/runc?status.svg)](https://godoc.org/github.com/opencontainers/runc)
 
 ## Introduction
 
@@ -76,6 +77,12 @@ You can run a specific test case by setting the `TESTFLAGS` variable.
 # make test TESTFLAGS="-run=SomeTestFunction"
 ```
 
+### Dependencies Management
+
+`runc` uses [vndr](https://github.com/LK4D4/vndr) for dependencies management.
+Please refer to [vndr](https://github.com/LK4D4/vndr) for how to add or update
+new dependencies.
+
 ## Using runc
 
 ### Creating an OCI Bundle

vendor/github.com/opencontainers/runc/libcontainer/README.md

@@ -1,3 +1,7 @@
+# libcontainer
+
+[![GoDoc](https://godoc.org/github.com/opencontainers/runc/libcontainer?status.svg)](https://godoc.org/github.com/opencontainers/runc/libcontainer)
+
 Libcontainer provides a native Go implementation for creating containers
 with namespaces, cgroups, capabilities, and filesystem access controls.
 It allows you to manage the lifecycle of the container performing additional operations
@@ -16,7 +20,14 @@ the current binary (/proc/self/exe) to be executed as the init process, and use
 arg "init", we call the first step process "bootstrap", so you always need a "init"
 function as the entry of "bootstrap".
 
+In addition to the go init function the early stage bootstrap is handled by importing
+[nsenter](https://github.com/opencontainers/runc/blob/master/libcontainer/nsenter/README.md).
+
 ```go
+import (
+	_ "github.com/opencontainers/runc/libcontainer/nsenter"
+)
+
 func init() {
 	if len(os.Args) > 1 && os.Args[1] == "init" {
 		runtime.GOMAXPROCS(1)

vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c

@@ -33,7 +33,8 @@ enum sync_t {
 	SYNC_USERMAP_ACK = 0x41, /* Mapping finished by the parent. */
 	SYNC_RECVPID_PLS = 0x42, /* Tell parent we're sending the PID. */
 	SYNC_RECVPID_ACK = 0x43, /* PID was correctly received by parent. */
-	SYNC_CHILD_READY = 0x44, /* The grandchild is ready to return. */
+	SYNC_GRANDCHILD  = 0x44, /* The grandchild is ready to run. */
+	SYNC_CHILD_READY = 0x45, /* The child or grandchild is ready to return. */
 
 	/* XXX: This doesn't help with segfaults and other such issues. */
 	SYNC_ERR = 0xFF, /* Fatal error, no turning back. The error code follows. */
@@ -71,6 +72,9 @@ struct nlconfig_t {
 	char *namespaces;
 	size_t namespaces_len;
 	uint8_t is_setgroup;
+	uint8_t is_rootless;
+	char *oom_score_adj;
+	size_t oom_score_adj_len;
 };
 
 /*
@@ -83,6 +87,8 @@ struct nlconfig_t {
 #define UIDMAP_ATTR			27283
 #define GIDMAP_ATTR			27284
 #define SETGROUP_ATTR		27285
+#define OOM_SCORE_ADJ_ATTR	27286
+#define ROOTLESS_ATTR	    27287
 
 /*
  * Use the raw syscall for versions of glibc which don't include a function for
@@ -171,6 +177,7 @@ static void update_setgroups(int pid, enum policy_t setgroup)
 			policy = "deny";
 			break;
 		case SETGROUPS_DEFAULT:
+		default:
 			/* Nothing to do. */
 			return;
 	}
@@ -185,7 +192,7 @@ static void update_setgroups(int pid, enum policy_t setgroup)
 	}
 }
 
-static void update_uidmap(int pid, char *map, int map_len)
+static void update_uidmap(int pid, char *map, size_t map_len)
 {
 	if (map == NULL || map_len <= 0)
 		return;
@@ -194,7 +201,7 @@ static void update_uidmap(int pid, char *map, int map_len)
 		bail("failed to update /proc/%d/uid_map", pid);
 }
 
-static void update_gidmap(int pid, char *map, int map_len)
+static void update_gidmap(int pid, char *map, size_t map_len)
 {
 	if (map == NULL || map_len <= 0)
 		return;
@@ -203,6 +210,15 @@ static void update_gidmap(int pid, char *map, int map_len)
 		bail("failed to update /proc/%d/gid_map", pid);
 }
 
+static void update_oom_score_adj(char *data, size_t len)
+{
+	if (data == NULL || len <= 0)
+		return;
+
+	if (write_file(data, len, "/proc/self/oom_score_adj") < 0)
+		bail("failed to update /proc/self/oom_score_adj");
+}
+
 /* A dummy function that just jumps to the given jumpval. */
 static int child_func(void *arg) __attribute__ ((noinline));
 static int child_func(void *arg)
@@ -284,7 +300,7 @@ static void nl_parse(int fd, struct nlconfig_t *config)
 	/* Retrieve the netlink header. */
 	len = read(fd, &hdr, NLMSG_HDRLEN);
 	if (len != NLMSG_HDRLEN)
-		bail("invalid netlink header length %lu", len);
+		bail("invalid netlink header length %zu", len);
 
 	if (hdr.nlmsg_type == NLMSG_ERROR)
 		bail("failed to read netlink message");
@@ -300,7 +316,7 @@ static void nl_parse(int fd, struct nlconfig_t *config)
 
 	len = read(fd, data, size);
 	if (len != size)
-		bail("failed to read netlink payload, %lu != %lu", len, size);
+		bail("failed to read netlink payload, %zu != %zu", len, size);
 
 	/* Parse the netlink payload. */
 	config->data = data;
@@ -316,6 +332,13 @@ static void nl_parse(int fd, struct nlconfig_t *config)
 		case CLONE_FLAGS_ATTR:
 			config->cloneflags = readint32(current);
 			break;
+		case ROOTLESS_ATTR:
+			config->is_rootless = readint8(current);
+			break;
+		case OOM_SCORE_ADJ_ATTR:
+			config->oom_score_adj = current;
+			config->oom_score_adj_len = payload_len;
+			break;
 		case NS_PATHS_ATTR:
 			config->namespaces = current;
 			config->namespaces_len = payload_len;
@@ -413,7 +436,7 @@ void nsexec(void)
 {
 	int pipenum;
 	jmp_buf env;
-	int syncpipe[2];
+	int sync_child_pipe[2], sync_grandchild_pipe[2];
 	struct nlconfig_t config = {0};
 
 	/*
@@ -424,18 +447,43 @@ void nsexec(void)
 	if (pipenum == -1)
 		return;
 
-	/* make the process non-dumpable */
-	if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) != 0) {
-		bail("failed to set process as non-dumpable");
-	}
-
 	/* Parse all of the netlink configuration. */
 	nl_parse(pipenum, &config);
 
+	/* Set oom_score_adj. This has to be done before !dumpable because
+	 * /proc/self/oom_score_adj is not writeable unless you're an privileged
+	 * user (if !dumpable is set). All children inherit their parent's
+	 * oom_score_adj value on fork(2) so this will always be propagated
+	 * properly.
+	 */
+	update_oom_score_adj(config.oom_score_adj, config.oom_score_adj_len);
+
+	/*
+	 * Make the process non-dumpable, to avoid various race conditions that
+	 * could cause processes in namespaces we're joining to access host
+	 * resources (or potentially execute code).
+	 *
+	 * However, if the number of namespaces we are joining is 0, we are not
+	 * going to be switching to a different security context. Thus setting
+	 * ourselves to be non-dumpable only breaks things (like rootless
+	 * containers), which is the recommendation from the kernel folks.
+	 */
+	if (config.namespaces) {
+		if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) < 0)
+			bail("failed to set process as non-dumpable");
+	}
+
 	/* Pipe so we can tell the child when we've finished setting up. */
-	if (socketpair(AF_LOCAL, SOCK_STREAM, 0, syncpipe) < 0)
+	if (socketpair(AF_LOCAL, SOCK_STREAM, 0, sync_child_pipe) < 0)
 		bail("failed to setup sync pipe between parent and child");
 
+	/*
+	 * We need a new socketpair to sync with grandchild so we don't have
+	 * race condition with child.
+	 */
+	if (socketpair(AF_LOCAL, SOCK_STREAM, 0, sync_grandchild_pipe) < 0)
+		bail("failed to setup sync pipe between parent and grandchild");
+
 	/* TODO: Currently we aren't dealing with child deaths properly. */
 
 	/*
@@ -494,9 +542,10 @@ void nsexec(void)
 	 *          process.
 	 */
 	case JUMP_PARENT: {
-			int len, ready = 0;
+			int len;
 			pid_t child;
 			char buf[JSON_MAX];
+			bool ready = false;
 
 			/* For debugging. */
 			prctl(PR_SET_NAME, (unsigned long) "runc:[0:PARENT]", 0, 0, 0);
@@ -513,30 +562,39 @@ void nsexec(void)
 			 * ready, so we can receive all possible error codes
 			 * generated by children.
 			 */
-			while (ready < 2) {
+			while (!ready) {
 				enum sync_t s;
+				int ret;
 
-				/* This doesn't need to be global, we're in the parent. */
+				syncfd = sync_child_pipe[1];
-				int syncfd = syncpipe[1];
+				close(sync_child_pipe[0]);
 
 				if (read(syncfd, &s, sizeof(s)) != sizeof(s))
 					bail("failed to sync with child: next state");
 
 				switch (s) {
-				case SYNC_ERR: {
+				case SYNC_ERR:
 					/* We have to mirror the error code of the child. */
-						int ret;
-
 					if (read(syncfd, &ret, sizeof(ret)) != sizeof(ret))
 						bail("failed to sync with child: read(error code)");
 
 					exit(ret);
-					}
-					break;
 				case SYNC_USERMAP_PLS:
-					/* Enable setgroups(2) if we've been asked to. */
+					/*
+					 * Enable setgroups(2) if we've been asked to. But we also
+					 * have to explicitly disable setgroups(2) if we're
+					 * creating a rootless container (this is required since
+					 * Linux 3.19).
+					 */
+					if (config.is_rootless && config.is_setgroup) {
+						kill(child, SIGKILL);
+						bail("cannot allow setgroup in an unprivileged user namespace setup");
+					}
+
 					if (config.is_setgroup)
 						update_setgroups(child, SETGROUPS_ALLOW);
+					if (config.is_rootless)
+						update_setgroups(child, SETGROUPS_DENY);
 
 					/* Set up mappings. */
 					update_uidmap(child, config.uidmap, config.uidmap_len);
@@ -548,11 +606,6 @@ void nsexec(void)
 						bail("failed to sync with child: write(SYNC_USERMAP_ACK)");
 					}
 					break;
-				case SYNC_USERMAP_ACK:
-					/* We should _never_ receive acks. */
-					kill(child, SIGKILL);
-					bail("failed to sync with child: unexpected SYNC_USERMAP_ACK");
-					break;
 				case SYNC_RECVPID_PLS: {
 						pid_t old = child;
 
@@ -570,20 +623,46 @@ void nsexec(void)
 							bail("failed to sync with child: write(SYNC_RECVPID_ACK)");
 						}
 					}
-
-					ready++;
-					break;
-				case SYNC_RECVPID_ACK:
-					/* We should _never_ receive acks. */
-					kill(child, SIGKILL);
-					bail("failed to sync with child: unexpected SYNC_RECVPID_ACK");
 					break;
 				case SYNC_CHILD_READY:
-					ready++;
+					ready = true;
 					break;
 				default:
-					bail("unexpected sync value");
+					bail("unexpected sync value: %u", s);
+				}
+			}
+
+			/* Now sync with grandchild. */
+
+			ready = false;
+			while (!ready) {
+				enum sync_t s;
+				int ret;
+
+				syncfd = sync_grandchild_pipe[1];
+				close(sync_grandchild_pipe[0]);
+
+				s = SYNC_GRANDCHILD;
+				if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+					kill(child, SIGKILL);
+					bail("failed to sync with child: write(SYNC_GRANDCHILD)");
+				}
+
+				if (read(syncfd, &s, sizeof(s)) != sizeof(s))
+					bail("failed to sync with child: next state");
+
+				switch (s) {
+				case SYNC_ERR:
+					/* We have to mirror the error code of the child. */
+					if (read(syncfd, &ret, sizeof(ret)) != sizeof(ret))
+						bail("failed to sync with child: read(error code)");
+
+					exit(ret);
+				case SYNC_CHILD_READY:
+					ready = true;
 					break;
+				default:
+					bail("unexpected sync value: %u", s);
 				}
 			}
 
@@ -615,7 +694,8 @@ void nsexec(void)
 			enum sync_t s;
 
 			/* We're in a child and thus need to tell the parent if we die. */
-			syncfd = syncpipe[0];
+			syncfd = sync_child_pipe[0];
+			close(sync_child_pipe[1]);
 
 			/* For debugging. */
 			prctl(PR_SET_NAME, (unsigned long) "runc:[1:CHILD]", 0, 0, 0);
@@ -653,6 +733,11 @@ void nsexec(void)
 				 * clone_parent rant). So signal our parent to hook us up.
 				 */
 
+				/* Switching is only necessary if we joined namespaces. */
+				if (config.namespaces) {
+					if (prctl(PR_SET_DUMPABLE, 1, 0, 0, 0) < 0)
+						bail("failed to set process as dumpable");
+				}
 				s = SYNC_USERMAP_PLS;
 				if (write(syncfd, &s, sizeof(s)) != sizeof(s))
 					bail("failed to sync with parent: write(SYNC_USERMAP_PLS)");
@@ -663,6 +748,11 @@ void nsexec(void)
 					bail("failed to sync with parent: read(SYNC_USERMAP_ACK)");
 				if (s != SYNC_USERMAP_ACK)
 					bail("failed to sync with parent: SYNC_USERMAP_ACK: got %u", s);
+				/* Switching is only necessary if we joined namespaces. */
+				if (config.namespaces) {
+					if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) < 0)
+						bail("failed to set process as dumpable");
+				}
 			}
 
 			/*
@@ -700,6 +790,12 @@ void nsexec(void)
 				bail("failed to sync with parent: SYNC_RECVPID_ACK: got %u", s);
 			}
 
+			s = SYNC_CHILD_READY;
+			if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+				kill(child, SIGKILL);
+				bail("failed to sync with parent: write(SYNC_CHILD_READY)");
+			}
+
 			/* Our work is done. [Stage 2: JUMP_INIT] is doing the rest of the work. */
 			exit(0);
 		}
@@ -718,11 +814,19 @@ void nsexec(void)
 			enum sync_t s;
 
 			/* We're in a child and thus need to tell the parent if we die. */
-			syncfd = syncpipe[0];
+			syncfd = sync_grandchild_pipe[0];
+			close(sync_grandchild_pipe[1]);
+			close(sync_child_pipe[0]);
+			close(sync_child_pipe[1]);
 
 			/* For debugging. */
 			prctl(PR_SET_NAME, (unsigned long) "runc:[2:INIT]", 0, 0, 0);
 
+			if (read(syncfd, &s, sizeof(s)) != sizeof(s))
+				bail("failed to sync with parent: read(SYNC_GRANDCHILD)");
+			if (s != SYNC_GRANDCHILD)
+				bail("failed to sync with parent: SYNC_GRANDCHILD: got %u", s);
+
 			if (setsid() < 0)
 				bail("setsid failed");
 
@@ -732,16 +836,17 @@ void nsexec(void)
 			if (setgid(0) < 0)
 				bail("setgid failed");
 
+			if (!config.is_rootless && config.is_setgroup) {
 				if (setgroups(0, NULL) < 0)
 					bail("setgroups failed");
+			}
 
 			s = SYNC_CHILD_READY;
 			if (write(syncfd, &s, sizeof(s)) != sizeof(s))
 				bail("failed to sync with patent: write(SYNC_CHILD_READY)");
 
 			/* Close sync pipes. */
-			close(syncpipe[0]);
+			close(sync_grandchild_pipe[0]);
-			close(syncpipe[1]);
 
 			/* Free netlink data. */
 			nl_free(&config);
@@ -751,7 +856,6 @@ void nsexec(void)
 		}
 	default:
 		bail("unexpected jump value");
-		break;
 	}
 
 	/* Should never be reached. */

vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.c

@@ -1,148 +0,0 @@
-/*
- * Copyright 2016 SUSE LLC
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *     http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-#include <errno.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include <sys/socket.h>
-#include <sys/types.h>
-#include <unistd.h>
-
-#include "cmsg.h"
-
-#define error(fmt, ...)							\
-	({								\
-		fprintf(stderr, "nsenter: " fmt ": %m\n", ##__VA_ARGS__); \
-		errno = ECOMM;						\
-		goto err; /* return value */				\
-	})
-
-/*
- * Sends a file descriptor along the sockfd provided. Returns the return
- * value of sendmsg(2). Any synchronisation and preparation of state
- * should be done external to this (we expect the other side to be in
- * recvfd() in the code).
- */
-ssize_t sendfd(int sockfd, struct file_t file)
-{
-	struct msghdr msg = {0};
-	struct iovec iov[1] = {0};
-	struct cmsghdr *cmsg;
-	int *fdptr;
-	int ret;
-
-	union {
-		char buf[CMSG_SPACE(sizeof(file.fd))];
-		struct cmsghdr align;
-	} u;
-
-	/*
-	 * We need to send some other data along with the ancillary data,
-	 * otherwise the other side won't recieve any data. This is very
-	 * well-hidden in the documentation (and only applies to
-	 * SOCK_STREAM). See the bottom part of unix(7).
-	 */
-	iov[0].iov_base = file.name;
-	iov[0].iov_len = strlen(file.name) + 1;
-
-	msg.msg_name = NULL;
-	msg.msg_namelen = 0;
-	msg.msg_iov = iov;
-	msg.msg_iovlen = 1;
-	msg.msg_control = u.buf;
-	msg.msg_controllen = sizeof(u.buf);
-
-	cmsg = CMSG_FIRSTHDR(&msg);
-	cmsg->cmsg_level = SOL_SOCKET;
-	cmsg->cmsg_type = SCM_RIGHTS;
-	cmsg->cmsg_len = CMSG_LEN(sizeof(int));
-
-	fdptr = (int *) CMSG_DATA(cmsg);
-	memcpy(fdptr, &file.fd, sizeof(int));
-
-	return sendmsg(sockfd, &msg, 0);
-}
-
-/*
- * Receives a file descriptor from the sockfd provided. Returns the file
- * descriptor as sent from sendfd(). It will return the file descriptor
- * or die (literally) trying. Any synchronisation and preparation of
- * state should be done external to this (we expect the other side to be
- * in sendfd() in the code).
- */
-struct file_t recvfd(int sockfd)
-{
-	struct msghdr msg = {0};
-	struct iovec iov[1] = {0};
-	struct cmsghdr *cmsg;
-	struct file_t file = {0};
-	int *fdptr;
-	int olderrno;
-
-	union {
-		char buf[CMSG_SPACE(sizeof(file.fd))];
-		struct cmsghdr align;
-	} u;
-
-	/* Allocate a buffer. */
-	/* TODO: Make this dynamic with MSG_PEEK. */
-	file.name = malloc(TAG_BUFFER);
-	if (!file.name)
-		error("recvfd: failed to allocate file.tag buffer\n");
-
-	/*
-	 * We need to "recieve" the non-ancillary data even though we don't
-	 * plan to use it at all. Otherwise, things won't work as expected.
-	 * See unix(7) and other well-hidden documentation.
-	 */
-	iov[0].iov_base = file.name;
-	iov[0].iov_len = TAG_BUFFER;
-
-	msg.msg_name = NULL;
-	msg.msg_namelen = 0;
-	msg.msg_iov = iov;
-	msg.msg_iovlen = 1;
-	msg.msg_control = u.buf;
-	msg.msg_controllen = sizeof(u.buf);
-
-	ssize_t ret = recvmsg(sockfd, &msg, 0);
-	if (ret < 0)
-		goto err;
-
-	cmsg = CMSG_FIRSTHDR(&msg);
-	if (!cmsg)
-		error("recvfd: got NULL from CMSG_FIRSTHDR");
-	if (cmsg->cmsg_level != SOL_SOCKET)
-		error("recvfd: expected SOL_SOCKET in cmsg: %d", cmsg->cmsg_level);
-	if (cmsg->cmsg_type != SCM_RIGHTS)
-		error("recvfd: expected SCM_RIGHTS in cmsg: %d", cmsg->cmsg_type);
-	if (cmsg->cmsg_len != CMSG_LEN(sizeof(int)))
-		error("recvfd: expected correct CMSG_LEN in cmsg: %lu", cmsg->cmsg_len);
-
-	fdptr = (int *) CMSG_DATA(cmsg);
-	if (!fdptr || *fdptr < 0)
-		error("recvfd: recieved invalid pointer");
-
-	file.fd = *fdptr;
-	return file;
-
-err:
-	olderrno = errno;
-	free(file.name);
-	errno = olderrno;
-	return (struct file_t){0};
-}

vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.go

@@ -3,7 +3,7 @@
 package utils
 
 /*
- * Copyright 2016 SUSE LLC
+ * Copyright 2016, 2017 SUSE LLC
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -18,28 +18,66 @@ package utils
  * limitations under the License.
  */
 
-/*
-#include <errno.h>
-#include <stdlib.h>
-#include "cmsg.h"
-*/
-import "C"
-
 import (
+	"fmt"
 	"os"
-	"unsafe"
+
+	"golang.org/x/sys/unix"
 )
 
+// MaxSendfdLen is the maximum length of the name of a file descriptor being
+// sent using SendFd. The name of the file handle returned by RecvFd will never
+// be larger than this value.
+const MaxNameLen = 4096
+
+// oobSpace is the size of the oob slice required to store a single FD. Note
+// that unix.UnixRights appears to make the assumption that fd is always int32,
+// so sizeof(fd) = 4.
+var oobSpace = unix.CmsgSpace(4)
+
 // RecvFd waits for a file descriptor to be sent over the given AF_UNIX
 // socket. The file name of the remote file descriptor will be recreated
 // locally (it is sent as non-auxiliary data in the same payload).
 func RecvFd(socket *os.File) (*os.File, error) {
-	file, err := C.recvfd(C.int(socket.Fd()))
+	// For some reason, unix.Recvmsg uses the length rather than the capacity
+	// when passing the msg_controllen and other attributes to recvmsg.  So we
+	// have to actually set the length.
+	name := make([]byte, MaxNameLen)
+	oob := make([]byte, oobSpace)
+
+	sockfd := socket.Fd()
+	n, oobn, _, _, err := unix.Recvmsg(int(sockfd), name, oob, 0)
 	if err != nil {
 		return nil, err
 	}
-	defer C.free(unsafe.Pointer(file.name))
+
-	return os.NewFile(uintptr(file.fd), C.GoString(file.name)), nil
+	if n >= MaxNameLen || oobn != oobSpace {
+		return nil, fmt.Errorf("recvfd: incorrect number of bytes read (n=%d oobn=%d)", n, oobn)
+	}
+
+	// Truncate.
+	name = name[:n]
+	oob = oob[:oobn]
+
+	scms, err := unix.ParseSocketControlMessage(oob)
+	if err != nil {
+		return nil, err
+	}
+	if len(scms) != 1 {
+		return nil, fmt.Errorf("recvfd: number of SCMs is not 1: %d", len(scms))
+	}
+	scm := scms[0]
+
+	fds, err := unix.ParseUnixRights(&scm)
+	if err != nil {
+		return nil, err
+	}
+	if len(fds) != 1 {
+		return nil, fmt.Errorf("recvfd: number of fds is not 1: %d", len(fds))
+	}
+	fd := uintptr(fds[0])
+
+	return os.NewFile(fd, string(name)), nil
 }
 
 // SendFd sends a file descriptor over the given AF_UNIX socket. In
@@ -47,11 +85,11 @@ func RecvFd(socket *os.File) (*os.File, error) {
 // non-auxiliary data in the same payload (allowing to send contextual
 // information for a file descriptor).
 func SendFd(socket, file *os.File) error {
-	var cfile C.struct_file_t
+	name := []byte(file.Name())
-	cfile.fd = C.int(file.Fd())
+	if len(name) >= MaxNameLen {
-	cfile.name = C.CString(file.Name())
+		return fmt.Errorf("sendfd: filename too long: %s", file.Name())
-	defer C.free(unsafe.Pointer(cfile.name))
+	}
-
+	oob := unix.UnixRights(int(file.Fd()))
-	_, err := C.sendfd(C.int(socket.Fd()), cfile)
+
-	return err
+	return unix.Sendmsg(int(socket.Fd()), name, oob, nil, 0)
 }

vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.h

@@ -1,36 +0,0 @@
-/*
- * Copyright 2016 SUSE LLC
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *     http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-#pragma once
-
-#if !defined(CMSG_H)
-#define CMSG_H
-
-#include <sys/types.h>
-
-/* TODO: Implement this properly with MSG_PEEK. */
-#define TAG_BUFFER 4096
-
-/* This mirrors Go's (*os.File). */
-struct file_t {
-	char *name;
-	int fd;
-};
-
-struct file_t recvfd(int sockfd);
-ssize_t sendfd(int sockfd, struct file_t file);
-
-#endif /* !defined(CMSG_H) */

vendor/github.com/opencontainers/runc/libcontainer/utils/utils_unix.go

@@ -4,6 +4,7 @@ package utils
 
 import (
 	"io/ioutil"
+	"os"
 	"strconv"
 	"syscall"
 )
@@ -31,3 +32,12 @@ func CloseExecFrom(minFd int) error {
 	}
 	return nil
 }
+
+// NewSockPair returns a new unix socket pair
+func NewSockPair(name string) (parent *os.File, child *os.File, err error) {
+	fds, err := syscall.Socketpair(syscall.AF_LOCAL, syscall.SOCK_STREAM|syscall.SOCK_CLOEXEC, 0)
+	if err != nil {
+		return nil, nil, err
+	}
+	return os.NewFile(uintptr(fds[1]), name+"-p"), os.NewFile(uintptr(fds[0]), name+"-c"), nil
+}