Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753403AbbLZVuA (ORCPT ); Sat, 26 Dec 2015 16:50:00 -0500 Received: from h2.hallyn.com ([78.46.35.8]:37983 "EHLO h2.hallyn.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750913AbbLZVt7 (ORCPT ); Sat, 26 Dec 2015 16:49:59 -0500 Date: Sat, 26 Dec 2015 15:49:54 -0600 From: "Serge E. Hallyn" To: Jann Horn Cc: "Serge E. Hallyn" , Roland McGrath , Oleg Nesterov , linux-kernel@vger.kernel.org, security@kernel.org, Serge Hallyn , Andy Lutomirski , "Eric W. Biederman" Subject: Re: [PATCH] ptrace: being capable wrt a process requires mapped uids/gids Message-ID: <20151226214954.GA21578@mail.hallyn.com> References: <20151226011038.GA25455@pc.thejh.net> <1451098351-8917-1-git-send-email-jann@thejh.net> <20151226211729.GC19815@mail.hallyn.com> <20151226212733.GA13821@pc.thejh.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20151226212733.GA13821@pc.thejh.net> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4150 Lines: 89 On Sat, Dec 26, 2015 at 10:27:33PM +0100, Jann Horn wrote: > On Sat, Dec 26, 2015 at 03:17:29PM -0600, Serge E. Hallyn wrote: > > On Sat, Dec 26, 2015 at 03:52:31AM +0100, Jann Horn wrote: > > > ptrace_has_cap() checks whether the current process should be > > > treated as having a certain capability for ptrace checks > > > against another process. Until now, this was equivalent to > > > has_ns_capability(current, target_ns, CAP_SYS_PTRACE). > > > > > > However, if a root-owned process wants to enter a user > > > namespace for some reason without knowing who owns it and > > > therefore can't change to the namespace owner's uid and gid > > > before entering, as soon as it has entered the namespace, > > > the namespace owner can attach to it via ptrace and thereby > > > gain access to its uid and gid. > > > > > > While it is possible for the entering process to switch to > > > the uid of a claimed namespace owner before entering, > > > causing the attempt to enter to fail if the claimed uid is > > > wrong, this doesn't solve the problem of determining an > > > appropriate gid. > > > > > > With this change, the entering process can first enter the > > > namespace and then safely inspect the namespace's > > > properties, e.g. through /proc/self/{uid_map,gid_map}, > > > assuming that the namespace owner doesn't have access to > > > uid 0. > > > > > > Changed in v2: The caller needs to be capable in the > > > namespace into which tcred's uids/gids can be mapped. > > > > > > Signed-off-by: Jann Horn > > > --- > > > kernel/ptrace.c | 33 ++++++++++++++++++++++++++++----- > > > 1 file changed, 28 insertions(+), 5 deletions(-) > > > > > > diff --git a/kernel/ptrace.c b/kernel/ptrace.c > > > index b760bae..260a08d 100644 > > > --- a/kernel/ptrace.c > > > +++ b/kernel/ptrace.c > > > @@ -20,6 +20,7 @@ > > > #include > > > #include > > > #include > > > +#include > > > #include > > > #include > > > #include > > > @@ -207,12 +208,34 @@ static int ptrace_check_attach(struct task_struct *child, bool ignore_state) > > > return ret; > > > } > > > > > > -static int ptrace_has_cap(struct user_namespace *ns, unsigned int mode) > > > +static bool ptrace_has_cap(const struct cred *tcred, unsigned int mode) > > > { > > > + struct user_namespace *tns = tcred->user_ns; > > > + > > > + /* When a root-owned process enters a user namespace created by a > > > + * malicious user, the user shouldn't be able to execute code under > > > + * uid 0 by attaching to the root-owned process via ptrace. > > > + * Therefore, similar to the capable_wrt_inode_uidgid() check, > > > + * verify that all the uids and gids of the target process are > > > + * mapped into a namespace below the current one in which the caller > > > + * is capable. > > > + * No fsuid/fsgid check because __ptrace_may_access doesn't do it > > > + * either. > > > + */ > > > + while ( > > > + !kuid_has_mapping(tns, tcred->euid) || > > > + !kuid_has_mapping(tns, tcred->suid) || > > > + !kuid_has_mapping(tns, tcred->uid) || > > > + !kgid_has_mapping(tns, tcred->egid) || > > > + !kgid_has_mapping(tns, tcred->sgid) || > > > + !kgid_has_mapping(tns, tcred->gid)) { > > > + tns = tns->parent; > > > > Sorry, i can't quite remember - is there a way for a task in init_user_ns to have > > INVALID_UID | INVALID_GID ? I.e. any point in breaking here if tns == &init_user_n? > > I assumed that there isn't because the comment above the definition of from_kuid() > says so. Checking... the syscalls for setting uid/gid seem to enforce that uid/gid > aren't -1, and setuid/setgid executables require the uid/gid to be mapped. So it > seems to be true. Yeah, I knew I'd read it somewhere but couldn't find the comment. Thanks. -- 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/