Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751434AbaK1Wxb (ORCPT ); Fri, 28 Nov 2014 17:53:31 -0500 Received: from mail-ie0-f170.google.com ([209.85.223.170]:65084 "EHLO mail-ie0-f170.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751302AbaK1Wx2 (ORCPT ); Fri, 28 Nov 2014 17:53:28 -0500 From: Andy Lutomirski To: "Eric W. Biederman" Cc: Linux Containers , Josh Triplett , Andrew Morton , Kees Cook , Michael Kerrisk-manpages , Linux API , linux-man , "linux-kernel@vger.kernel.org" , LSM , Casey Schaufler , "Serge E. Hallyn" , Richard Weinberger , kenton@sandstorm.io, Andy Lutomirski , stable@vger.kernel.org Subject: [RFC PATCH] userns: Disallow setgroups unless the gid_map writer is privileged Date: Fri, 28 Nov 2014 14:53:20 -0800 Message-Id: <81ba656ffbdc62a7f7e94ad17f0238ee063fe6b3.1417214430.git.luto@amacapital.net> X-Mailer: git-send-email 1.9.3 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Classic unix permission checks have an interesting feature. The group permissions for a file can be set to less than the other permissions on a file. Occasionally this is used deliberately to give a certain group of users fewer permissions than the default. User namespaces break this usage. Groups set in rgid or egid are unaffected because an unprivileged user namespace creator can only map a single group, so setresgid inside and outside the namespace have the same effect. However, an unprivileged user namespace creator can currently use setgroups(2) to drop all supplementary groups, so, if a supplementary group denies access to some resource, user namespaces can be used to bypass that restriction. To fix this issue without unduly breaking existing user code, this introduces a new user namespace flag USERNS_SETGROUPS_ALLOWED. If that flag is not set, then setgroups(2) will fail regardless of the caller's capabilities. USERNS_SETGROUPS_ALLOWED is cleared in a new user namespace. By default, if the writer of gid_map has CAP_SETGID in the parent userns and the parent userns has USERNS_SETGROUPS_ALLOWED, then the USERNS_SETGROUPS_ALLOWED will be set in the child. While it could be safe to set USERNS_SETGROUPS_ALLOWED if the user namespace creator has no supplementary groups, doing so could be surprising and could have unpleasant interactions with setns(2). The default behavior can be overridden by adding a line of the form "setgroups allow" or "setgroups deny" to gid_map. The former will return -EPERM if the caller is not permitted to set USERNS_SETGROUPS_ALLOWED, and the latter will clear USERNS_SETGROUPS_ALLOWED even if the flag would otherwise be set. This change should break userspace by the minimal amount needed to fix this issue. Any application that uses newgidmap(1) should be unaffected by this fix, and unprivileged users that create user namespaces to manipulate mounts or sandbox themselves will also be unaffected unless they try to use setgroups(2). This should fix CVE-2014-8989. Cc: stable@vger.kernel.org Signed-off-by: Andy Lutomirski --- Eric, this is an alternative to your patch. I think it will cause less breakage, and it will keep unprivileged user namespaces more or less fully functional. Kenton, I think that neither run-bundle nor supervisor-main will be broken by this patch. include/linux/user_namespace.h | 3 +++ kernel/groups.c | 3 +++ kernel/user.c | 1 + kernel/user_namespace.c | 36 ++++++++++++++++++++++++++++++++++++ 4 files changed, 43 insertions(+) diff --git a/include/linux/user_namespace.h b/include/linux/user_namespace.h index e95372654f09..a74c1f3d44fe 100644 --- a/include/linux/user_namespace.h +++ b/include/linux/user_namespace.h @@ -17,6 +17,8 @@ struct uid_gid_map { /* 64 bytes -- 1 cache line */ } extent[UID_GID_MAP_MAX_EXTENTS]; }; +#define USERNS_SETGROUPS_ALLOWED 1 + struct user_namespace { struct uid_gid_map uid_map; struct uid_gid_map gid_map; @@ -27,6 +29,7 @@ struct user_namespace { kuid_t owner; kgid_t group; unsigned int proc_inum; + unsigned int flags; /* Register of per-UID persistent keyrings for this namespace */ #ifdef CONFIG_PERSISTENT_KEYRINGS diff --git a/kernel/groups.c b/kernel/groups.c index 451698f86cfa..e27433809978 100644 --- a/kernel/groups.c +++ b/kernel/groups.c @@ -6,6 +6,7 @@ #include #include #include +#include #include /* init to 2 - one for init_task, one to ensure it is never freed */ @@ -223,6 +224,8 @@ SYSCALL_DEFINE2(setgroups, int, gidsetsize, gid_t __user *, grouplist) struct group_info *group_info; int retval; + if (!(current_user_ns()->flags & USERNS_SETGROUPS_ALLOWED)) + return -EPERM; if (!ns_capable(current_user_ns(), CAP_SETGID)) return -EPERM; if ((unsigned)gidsetsize > NGROUPS_MAX) diff --git a/kernel/user.c b/kernel/user.c index 4efa39350e44..f8cdb1ec6049 100644 --- a/kernel/user.c +++ b/kernel/user.c @@ -51,6 +51,7 @@ struct user_namespace init_user_ns = { .owner = GLOBAL_ROOT_UID, .group = GLOBAL_ROOT_GID, .proc_inum = PROC_USER_INIT_INO, + .flags = USERNS_SETGROUPS_ALLOWED, #ifdef CONFIG_PERSISTENT_KEYRINGS .persistent_keyring_register_sem = __RWSEM_INITIALIZER(init_user_ns.persistent_keyring_register_sem), diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index aa312b0dc3ec..6e7b9ee5bddc 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -600,6 +600,8 @@ static ssize_t map_write(struct file *file, const char __user *buf, unsigned long page = 0; char *kbuf, *pos, *next_line; ssize_t ret = -EINVAL; + unsigned int gid_flags = 0; + bool seen_explicit_gid_flag = false; /* * The id_map_mutex serializes all writes to any given map. @@ -633,6 +635,19 @@ static ssize_t map_write(struct file *file, const char __user *buf, if (cap_valid(cap_setid) && !file_ns_capable(file, ns, CAP_SYS_ADMIN)) goto out; + /* Deal with supplementary groups. */ + if (map == &ns->gid_map) { + /* + * By default, setgroups is allowed inside the userns + * if the writer has no supplementary groups (making + * it useless) or if the writer is privileged. + */ + if ((ns->parent->flags & USERNS_SETGROUPS_ALLOWED) && + file_ns_capable(file, ns->parent, CAP_SETGID) && + ns_capable(ns->parent, CAP_SETGID)) + gid_flags = USERNS_SETGROUPS_ALLOWED; + } + /* Get a buffer */ ret = -ENOMEM; page = __get_free_page(GFP_TEMPORARY); @@ -667,6 +682,25 @@ static ssize_t map_write(struct file *file, const char __user *buf, next_line = NULL; } + /* Is this line a gid_map option? */ + if (map == &ns->gid_map) { + if (!strcmp(pos, "setgroups deny")) { + if (seen_explicit_gid_flag) + goto out; + seen_explicit_gid_flag = 1; + gid_flags = 0; + continue; + } else if (!strcmp(pos, "setgroups allow")) { + if (seen_explicit_gid_flag) + goto out; + if (!(gid_flags & USERNS_SETGROUPS_ALLOWED)) { + ret = -EPERM; + goto out; + } + continue; + } + } + pos = skip_spaces(pos); extent->first = simple_strtoul(pos, &pos, 10); if (!isspace(*pos)) @@ -746,6 +780,8 @@ static ssize_t map_write(struct file *file, const char __user *buf, new_map.nr_extents*sizeof(new_map.extent[0])); smp_wmb(); map->nr_extents = new_map.nr_extents; + if (map == &ns->gid_map) + ns->flags |= gid_flags; *ppos = count; ret = count; -- 1.9.3 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/