[Buildroot] [PATCH 2/2] package/runc: add upstream security fix for CVE-2019-5736

[Peter Korsgaard <peter@korsgaard.com>]

The vulnerability allows a malicious container to (with minimal user
interaction) overwrite the host runc binary and thus gain root-level
code execution on the host. The level of user interaction is being able
to run any command (it doesn't matter if the command is not
attacker-controlled) as root within a container in either of these
contexts:

  * Creating a new container using an attacker-controlled image.
  * Attaching (docker exec) into an existing container which the
    attacker had previous write access to.

For more details, see the advisory:

https://www.openwall.com/lists/oss-security/2019/02/11/2

The fix for this issue uses fexecve(3), which isn't available on uClibc, so
add a dependency on !uclibc to runc and propagate to the reverse
dependencies (containerd/docker-engine).

Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
---
 package/docker-containerd/Config.in                |   5 +-
 package/docker-engine/Config.in                    |   5 +-
 ...ne-proc-self-exe-to-avoid-exposing-host-b.patch | 338 +++++++++++++++++++++
 package/runc/Config.in                             |   5 +-
 4 files changed, 347 insertions(+), 6 deletions(-)
 create mode 100644 package/runc/0001-nsenter-clone-proc-self-exe-to-avoid-exposing-host-b.patch

diff --git a/package/docker-containerd/Config.in b/package/docker-containerd/Config.in
index 851345f73e..bdb5dd7b80 100644
--- a/package/docker-containerd/Config.in
+++ b/package/docker-containerd/Config.in
@@ -3,6 +3,7 @@ config BR2_PACKAGE_DOCKER_CONTAINERD
 	depends on BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS
 	depends on BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS
 	depends on BR2_TOOLCHAIN_HAS_THREADS
+	depends on !BR2_PACKAGE_TOOLCHAIN_USES_UCLIBC # runc
 	depends on BR2_USE_MMU # util-linux
 	select BR2_PACKAGE_RUNC # runtime dependency
 	select BR2_PACKAGE_UTIL_LINUX # runtime dependency
@@ -27,8 +28,8 @@ config BR2_PACKAGE_DOCKER_CONTAINERD_DRIVER_BTRFS
 
 endif
 
-comment "docker-containerd needs a toolchain w/ threads"
+comment "docker-containerd needs a glibc or musl toolchain w/ threads"
 	depends on BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS
 	depends on BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS
 	depends on BR2_USE_MMU
-	depends on !BR2_TOOLCHAIN_HAS_THREADS
+	depends on !BR2_TOOLCHAIN_HAS_THREADS || BR2_TOOLCHAIN_USES_UCLIBC
diff --git a/package/docker-engine/Config.in b/package/docker-engine/Config.in
index 2a0c130fce..3c97310484 100644
--- a/package/docker-engine/Config.in
+++ b/package/docker-engine/Config.in
@@ -3,6 +3,7 @@ config BR2_PACKAGE_DOCKER_ENGINE
 	depends on BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS
 	depends on BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS
 	depends on BR2_TOOLCHAIN_HAS_THREADS
+	depends on !BR2_TOOLCHAIN_USES_UCLIBC # docker-containerd -> runc
 	depends on BR2_USE_MMU # docker-containerd
 	select BR2_PACKAGE_DOCKER_CONTAINERD # runtime dependency
 	select BR2_PACKAGE_DOCKER_PROXY # runtime dependency
@@ -49,8 +50,8 @@ config BR2_PACKAGE_DOCKER_ENGINE_DRIVER_VFS
 
 endif
 
-comment "docker-engine needs a toolchain w/ threads"
+comment "docker-engine needs a glibc or musl toolchain w/ threads"
 	depends on BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS
 	depends on BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS
-	depends on !BR2_TOOLCHAIN_HAS_THREADS
+	depends on !BR2_TOOLCHAIN_HAS_THREADS || BR2_TOOLCHAIN_USES_UCLIBC
 	depends on BR2_USE_MMU
diff --git a/package/runc/0001-nsenter-clone-proc-self-exe-to-avoid-exposing-host-b.patch b/package/runc/0001-nsenter-clone-proc-self-exe-to-avoid-exposing-host-b.patch
new file mode 100644
index 0000000000..a0bc7ba300
--- /dev/null
+++ b/package/runc/0001-nsenter-clone-proc-self-exe-to-avoid-exposing-host-b.patch
@@ -0,0 +1,338 @@
+From 0a8e4117e7f715d5fbeef398405813ce8e88558b Mon Sep 17 00:00:00 2001
+From: Aleksa Sarai <asarai@suse.de>
+Date: Wed, 9 Jan 2019 13:40:01 +1100
+Subject: [PATCH] nsenter: clone /proc/self/exe to avoid exposing host binary
+ to container
+
+There are quite a few circumstances where /proc/self/exe pointing to a
+pretty important container binary is a _bad_ thing, so to avoid this we
+have to make a copy (preferably doing self-clean-up and not being
+writeable).
+
+We require memfd_create(2) -- though there is an O_TMPFILE fallback --
+but we can always extend this to use a scratch MNT_DETACH overlayfs or
+tmpfs. The main downside to this approach is no page-cache sharing for
+the runc binary (which overlayfs would give us) but this is far less
+complicated.
+
+This is only done during nsenter so that it happens transparently to the
+Go code, and any libcontainer users benefit from it. This also makes
+ExtraFiles and --preserve-fds handling trivial (because we don't need to
+worry about it).
+
+Fixes: CVE-2019-5736
+Co-developed-by: Christian Brauner <christian.brauner@ubuntu.com>
+Signed-off-by: Aleksa Sarai <asarai@suse.de>
+Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
+---
+ libcontainer/nsenter/cloned_binary.c | 268 +++++++++++++++++++++++++++++++++++
+ libcontainer/nsenter/nsexec.c        |  11 ++
+ 2 files changed, 279 insertions(+)
+ create mode 100644 libcontainer/nsenter/cloned_binary.c
+
+diff --git a/libcontainer/nsenter/cloned_binary.c b/libcontainer/nsenter/cloned_binary.c
+new file mode 100644
+index 00000000..c8a42c23
+--- /dev/null
++++ b/libcontainer/nsenter/cloned_binary.c
+@@ -0,0 +1,268 @@
++/*
++ * Copyright (C) 2019 Aleksa Sarai <cyphar@cyphar.com>
++ * Copyright (C) 2019 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.
++ */
++
++#define _GNU_SOURCE
++#include <unistd.h>
++#include <stdio.h>
++#include <stdlib.h>
++#include <stdbool.h>
++#include <string.h>
++#include <limits.h>
++#include <fcntl.h>
++#include <errno.h>
++
++#include <sys/types.h>
++#include <sys/stat.h>
++#include <sys/vfs.h>
++#include <sys/mman.h>
++#include <sys/sendfile.h>
++#include <sys/syscall.h>
++
++/* Use our own wrapper for memfd_create. */
++#if !defined(SYS_memfd_create) && defined(__NR_memfd_create)
++#  define SYS_memfd_create __NR_memfd_create
++#endif
++#ifdef SYS_memfd_create
++#  define HAVE_MEMFD_CREATE
++/* memfd_create(2) flags -- copied from <linux/memfd.h>. */
++#  ifndef MFD_CLOEXEC
++#    define MFD_CLOEXEC       0x0001U
++#    define MFD_ALLOW_SEALING 0x0002U
++#  endif
++int memfd_create(const char *name, unsigned int flags)
++{
++	return syscall(SYS_memfd_create, name, flags);
++}
++#endif
++
++/* This comes directly from <linux/fcntl.h>. */
++#ifndef F_LINUX_SPECIFIC_BASE
++#  define F_LINUX_SPECIFIC_BASE 1024
++#endif
++#ifndef F_ADD_SEALS
++#  define F_ADD_SEALS (F_LINUX_SPECIFIC_BASE + 9)
++#  define F_GET_SEALS (F_LINUX_SPECIFIC_BASE + 10)
++#endif
++#ifndef F_SEAL_SEAL
++#  define F_SEAL_SEAL   0x0001	/* prevent further seals from being set */
++#  define F_SEAL_SHRINK 0x0002	/* prevent file from shrinking */
++#  define F_SEAL_GROW   0x0004	/* prevent file from growing */
++#  define F_SEAL_WRITE  0x0008	/* prevent writes */
++#endif
++
++#define RUNC_SENDFILE_MAX 0x7FFFF000 /* sendfile(2) is limited to 2GB. */
++#ifdef HAVE_MEMFD_CREATE
++#  define RUNC_MEMFD_COMMENT "runc_cloned:/proc/self/exe"
++#  define RUNC_MEMFD_SEALS \
++	(F_SEAL_SEAL | F_SEAL_SHRINK | F_SEAL_GROW | F_SEAL_WRITE)
++#endif
++
++static void *must_realloc(void *ptr, size_t size)
++{
++	void *old = ptr;
++	do {
++		ptr = realloc(old, size);
++	} while(!ptr);
++	return ptr;
++}
++
++/*
++ * Verify whether we are currently in a self-cloned program (namely, is
++ * /proc/self/exe a memfd). F_GET_SEALS will only succeed for memfds (or rather
++ * for shmem files), and we want to be sure it's actually sealed.
++ */
++static int is_self_cloned(void)
++{
++	int fd, ret, is_cloned = 0;
++
++	fd = open("/proc/self/exe", O_RDONLY|O_CLOEXEC);
++	if (fd < 0)
++		return -ENOTRECOVERABLE;
++
++#ifdef HAVE_MEMFD_CREATE
++	ret = fcntl(fd, F_GET_SEALS);
++	is_cloned = (ret == RUNC_MEMFD_SEALS);
++#else
++	struct stat statbuf = {0};
++	ret = fstat(fd, &statbuf);
++	if (ret >= 0)
++		is_cloned = (statbuf.st_nlink == 0);
++#endif
++	close(fd);
++	return is_cloned;
++}
++
++/*
++ * Basic wrapper around mmap(2) that gives you the file length so you can
++ * safely treat it as an ordinary buffer. Only gives you read access.
++ */
++static char *read_file(char *path, size_t *length)
++{
++	int fd;
++	char buf[4096], *copy = NULL;
++
++	if (!length)
++		return NULL;
++
++	fd = open(path, O_RDONLY | O_CLOEXEC);
++	if (fd < 0)
++		return NULL;
++
++	*length = 0;
++	for (;;) {
++		int n;
++
++		n = read(fd, buf, sizeof(buf));
++		if (n < 0)
++			goto error;
++		if (!n)
++			break;
++
++		copy = must_realloc(copy, (*length + n) * sizeof(*copy));
++		memcpy(copy + *length, buf, n);
++		*length += n;
++	}
++	close(fd);
++	return copy;
++
++error:
++	close(fd);
++	free(copy);
++	return NULL;
++}
++
++/*
++ * A poor-man's version of "xargs -0". Basically parses a given block of
++ * NUL-delimited data, within the given length and adds a pointer to each entry
++ * to the array of pointers.
++ */
++static int parse_xargs(char *data, int data_length, char ***output)
++{
++	int num = 0;
++	char *cur = data;
++
++	if (!data || *output != NULL)
++		return -1;
++
++	while (cur < data + data_length) {
++		num++;
++		*output = must_realloc(*output, (num + 1) * sizeof(**output));
++		(*output)[num - 1] = cur;
++		cur += strlen(cur) + 1;
++	}
++	(*output)[num] = NULL;
++	return num;
++}
++
++/*
++ * "Parse" out argv and envp from /proc/self/cmdline and /proc/self/environ.
++ * This is necessary because we are running in a context where we don't have a
++ * main() that we can just get the arguments from.
++ */
++static int fetchve(char ***argv, char ***envp)
++{
++	char *cmdline = NULL, *environ = NULL;
++	size_t cmdline_size, environ_size;
++
++	cmdline = read_file("/proc/self/cmdline", &cmdline_size);
++	if (!cmdline)
++		goto error;
++	environ = read_file("/proc/self/environ", &environ_size);
++	if (!environ)
++		goto error;
++
++	if (parse_xargs(cmdline, cmdline_size, argv) <= 0)
++		goto error;
++	if (parse_xargs(environ, environ_size, envp) <= 0)
++		goto error;
++
++	return 0;
++
++error:
++	free(environ);
++	free(cmdline);
++	return -EINVAL;
++}
++
++static int clone_binary(void)
++{
++	int binfd, memfd;
++	ssize_t sent = 0;
++
++#ifdef HAVE_MEMFD_CREATE
++	memfd = memfd_create(RUNC_MEMFD_COMMENT, MFD_CLOEXEC | MFD_ALLOW_SEALING);
++#else
++	memfd = open("/tmp", O_TMPFILE | O_EXCL | O_RDWR | O_CLOEXEC, 0711);
++#endif
++	if (memfd < 0)
++		return -ENOTRECOVERABLE;
++
++	binfd = open("/proc/self/exe", O_RDONLY | O_CLOEXEC);
++	if (binfd < 0)
++		goto error;
++
++	sent = sendfile(memfd, binfd, NULL, RUNC_SENDFILE_MAX);
++	close(binfd);
++	if (sent < 0)
++		goto error;
++
++#ifdef HAVE_MEMFD_CREATE
++	int err = fcntl(memfd, F_ADD_SEALS, RUNC_MEMFD_SEALS);
++	if (err < 0)
++		goto error;
++#else
++	/* Need to re-open "memfd" as read-only to avoid execve(2) giving -EXTBUSY. */
++	int newfd;
++	char *fdpath = NULL;
++
++	if (asprintf(&fdpath, "/proc/self/fd/%d", memfd) < 0)
++		goto error;
++	newfd = open(fdpath, O_RDONLY | O_CLOEXEC);
++	free(fdpath);
++	if (newfd < 0)
++		goto error;
++
++	close(memfd);
++	memfd = newfd;
++#endif
++	return memfd;
++
++error:
++	close(memfd);
++	return -EIO;
++}
++
++int ensure_cloned_binary(void)
++{
++	int execfd;
++	char **argv = NULL, **envp = NULL;
++
++	/* Check that we're not self-cloned, and if we are then bail. */
++	int cloned = is_self_cloned();
++	if (cloned > 0 || cloned == -ENOTRECOVERABLE)
++		return cloned;
++
++	if (fetchve(&argv, &envp) < 0)
++		return -EINVAL;
++
++	execfd = clone_binary();
++	if (execfd < 0)
++		return -EIO;
++
++	fexecve(execfd, argv, envp);
++	return -ENOEXEC;
++}
+diff --git a/libcontainer/nsenter/nsexec.c b/libcontainer/nsenter/nsexec.c
+index 28269dfc..7750af35 100644
+--- a/libcontainer/nsenter/nsexec.c
++++ b/libcontainer/nsenter/nsexec.c
+@@ -534,6 +534,9 @@ void join_namespaces(char *nslist)
+ 	free(namespaces);
+ }
+ 
++/* Defined in cloned_binary.c. */
++extern int ensure_cloned_binary(void);
++
+ void nsexec(void)
+ {
+ 	int pipenum;
+@@ -549,6 +552,14 @@ void nsexec(void)
+ 	if (pipenum == -1)
+ 		return;
+ 
++	/*
++	 * We need to re-exec if we are not in a cloned binary. This is necessary
++	 * to ensure that containers won't be able to access the host binary
++	 * through /proc/self/exe. See CVE-2019-5736.
++	 */
++	if (ensure_cloned_binary() < 0)
++		bail("could not ensure we are a cloned binary");
++
+ 	/* Parse all of the netlink configuration. */
+ 	nl_parse(pipenum, &config);
+ 
+-- 
+2.11.0
+
diff --git a/package/runc/Config.in b/package/runc/Config.in
index fd5dee7c5b..47c850ef30 100644
--- a/package/runc/Config.in
+++ b/package/runc/Config.in
@@ -3,13 +3,14 @@ config BR2_PACKAGE_RUNC
 	depends on BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS
 	depends on BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS
 	depends on BR2_TOOLCHAIN_HAS_THREADS
+	depends on !BR2_TOOLCHAIN_USES_UCLIBC # no fexecve
 	help
 	  runC is a CLI tool for spawning and running containers
 	  according to the OCP specification.
 
 	  https://github.com/opencontainers/runc
 
-comment "runc needs a toolchain w/ threads"
+comment "runc needs a glibc or musl toolchain toolchain w/ threads"
 	depends on BR2_PACKAGE_HOST_GO_ARCH_SUPPORTS && \
 		BR2_PACKAGE_HOST_GO_CGO_LINKING_SUPPORTS
-	depends on !BR2_TOOLCHAIN_HAS_THREADS
+	depends on !BR2_TOOLCHAIN_HAS_THREADS || BR2_TOOLCHAN_USES_UCLIBC
-- 
2.11.0