Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754159AbcKURKR (ORCPT ); Mon, 21 Nov 2016 12:10:17 -0500 Received: from bedivere.hansenpartnership.com ([66.63.167.143]:34656 "EHLO bedivere.hansenpartnership.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753309AbcKURKP (ORCPT ); Mon, 21 Nov 2016 12:10:15 -0500 Message-ID: <1479748213.2309.37.camel@HansenPartnership.com> Subject: Re: [PATCH v2] Add a "nosymlinks" mount option. From: James Bottomley To: Mattias Nissler , linux-fsdevel@vger.kernel.org Cc: Alexander Viro , linux-kernel@vger.kernel.org, Colin Walters , "Austin S . Hemmelgarn" Date: Mon, 21 Nov 2016 09:10:13 -0800 In-Reply-To: References: <1476880312-64786-1-git-send-email-mnissler@chromium.org> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.16.5 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 8820 Lines: 215 On Wed, 2016-11-16 at 13:18 -0800, Mattias Nissler wrote: > I understand that silence suggests there's little interest, but > here's some new information I discovered today that may justify to > reconsider the patch: > > The BSDs already have exactly what I propose, the mount option is > called "nosymfollow" there: > https://github.com/freebsd/freebsd/blob/a41f4cc9a57cd74604ae7b051eec2 > f48865f18d6/sys/kern/vfs_lookup.c#L939 > > There's also some evidence on the net that people have been using > said nosymfollow mount option to mitigate symlink attacks. I've got to say that just because BSD does this doesn't make it a good idea. The problem with disabling symlinks is that they're a core part of unix filesystem semantics and simply disabling them breaks various setups in various ways. The breakage depends on the user, so if you come from a Windows background, where symlinks basically don't exist, you're likely fine, but if you come from a UNIX background, chances are you use them pretty extensively. Obviously, I'm a UNIX background person and I can think of all the nasty ways you'll break my setup ... If a root privileged programme is trying to write to a file, it's not unreasonable to expect the programme itself to take simple precautions (like setuid to the user it's trying to write), so I'd really rather declare binaries that don't do this as broken and fix them. Thinking you've plugged a hole like this simply by disabling symlinks is a false sense of security because there are a variety of other ways I could trick the programme into writing where it shouldn't. Disabling symlinks is fixing a symptom, it's really the cause that should be fixed. James > On Mon, Oct 24, 2016 at 7:09 AM, Mattias Nissler < > mnissler@chromium.org> wrote: > > Friendly ping - does this version of the patch have any chance on > > getting included in mainline? > > > > On Wed, Oct 19, 2016 at 2:31 PM, Mattias Nissler < > > mnissler@chromium.org> wrote: > > > For mounts that have the new "nosymlinks" option, don't follow > > > symlinks when resolving paths. The new option is similar in > > > spirit to > > > the existing "nodev", "noexec", and "nosuid" options. > > > > > > Note that symlinks may still be created on mounts where the > > > "nosymlinks" option is present. readlink() remains functional, so > > > user > > > space code that is aware of symlinks can still choose to follow > > > them > > > explicitly. > > > > > > Setting the "nosymlinks" mount option helps prevent privileged > > > writers > > > from modifying files unintentionally in case there is an > > > unexpected > > > link along the accessed path. The "nosymlinks" option is thus > > > useful > > > as a defensive measure for systems that need to deal with > > > untrusted > > > file systems in privileged contexts. > > > > > > Signed-off-by: Mattias Nissler > > > --- > > > fs/namei.c | 3 +++ > > > fs/namespace.c | 9 ++++++--- > > > fs/proc_namespace.c | 1 + > > > fs/statfs.c | 2 ++ > > > include/linux/mount.h | 3 ++- > > > include/linux/statfs.h | 1 + > > > include/uapi/linux/fs.h | 1 + > > > 7 files changed, 16 insertions(+), 4 deletions(-) > > > > > > diff --git a/fs/namei.c b/fs/namei.c > > > index 5b4eed2..4cddcf3 100644 > > > --- a/fs/namei.c > > > +++ b/fs/namei.c > > > @@ -1021,6 +1021,9 @@ const char *get_link(struct nameidata *nd) > > > touch_atime(&last->link); > > > } > > > > > > + if (nd->path.mnt->mnt_flags & MNT_NOSYMLINKS) > > > + return ERR_PTR(-EACCES); > > > + > > > error = security_inode_follow_link(dentry, inode, > > > nd->flags & > > > LOOKUP_RCU); > > > if (unlikely(error)) > > > diff --git a/fs/namespace.c b/fs/namespace.c > > > index e6c234b..deec84e 100644 > > > --- a/fs/namespace.c > > > +++ b/fs/namespace.c > > > @@ -2732,6 +2732,8 @@ long do_mount(const char *dev_name, const > > > char __user *dir_name, > > > mnt_flags &= ~(MNT_RELATIME | MNT_NOATIME); > > > if (flags & MS_RDONLY) > > > mnt_flags |= MNT_READONLY; > > > + if (flags & MS_NOSYMLINKS) > > > + mnt_flags |= MNT_NOSYMLINKS; > > > > > > /* The default atime for remount is preservation */ > > > if ((flags & MS_REMOUNT) && > > > @@ -2741,9 +2743,10 @@ long do_mount(const char *dev_name, const > > > char __user *dir_name, > > > mnt_flags |= path.mnt->mnt_flags & > > > MNT_ATIME_MASK; > > > } > > > > > > - flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE | > > > MS_BORN | > > > - MS_NOATIME | MS_NODIRATIME | MS_RELATIME| > > > MS_KERNMOUNT | > > > - MS_STRICTATIME | MS_NOREMOTELOCK); > > > + flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | > > > MS_NOSYMLINKS | > > > + MS_ACTIVE | MS_BORN | MS_NOATIME | > > > MS_NODIRATIME | > > > + MS_RELATIME | MS_KERNMOUNT | MS_STRICTATIME | > > > + MS_NOREMOTELOCK); > > > > > > if (flags & MS_REMOUNT) > > > retval = do_remount(&path, flags & ~MS_REMOUNT, > > > mnt_flags, > > > diff --git a/fs/proc_namespace.c b/fs/proc_namespace.c > > > index 3f1190d..a1949d9 100644 > > > --- a/fs/proc_namespace.c > > > +++ b/fs/proc_namespace.c > > > @@ -67,6 +67,7 @@ static void show_mnt_opts(struct seq_file *m, > > > struct vfsmount *mnt) > > > { MNT_NOATIME, ",noatime" }, > > > { MNT_NODIRATIME, ",nodiratime" }, > > > { MNT_RELATIME, ",relatime" }, > > > + { MNT_NOSYMLINKS, ",nosymlinks" }, > > > { 0, NULL } > > > }; > > > const struct proc_fs_info *fs_infop; > > > diff --git a/fs/statfs.c b/fs/statfs.c > > > index 083dc0a..7ff7c32 100644 > > > --- a/fs/statfs.c > > > +++ b/fs/statfs.c > > > @@ -27,6 +27,8 @@ static int flags_by_mnt(int mnt_flags) > > > flags |= ST_NODIRATIME; > > > if (mnt_flags & MNT_RELATIME) > > > flags |= ST_RELATIME; > > > + if (mnt_flags & MNT_NOSYMLINKS) > > > + flags |= ST_NOSYMLINKS; > > > return flags; > > > } > > > > > > diff --git a/include/linux/mount.h b/include/linux/mount.h > > > index 1172cce..5e302f4 100644 > > > --- a/include/linux/mount.h > > > +++ b/include/linux/mount.h > > > @@ -28,6 +28,7 @@ struct mnt_namespace; > > > #define MNT_NODIRATIME 0x10 > > > #define MNT_RELATIME 0x20 > > > #define MNT_READONLY 0x40 /* does the user want this to be > > > r/o? */ > > > +#define MNT_NOSYMLINKS 0x80 > > > > > > #define MNT_SHRINKABLE 0x100 > > > #define MNT_WRITE_HOLD 0x200 > > > @@ -44,7 +45,7 @@ struct mnt_namespace; > > > #define MNT_SHARED_MASK (MNT_UNBINDABLE) > > > #define MNT_USER_SETTABLE_MASK (MNT_NOSUID | MNT_NODEV | > > > MNT_NOEXEC \ > > > | MNT_NOATIME | MNT_NODIRATIME | > > > MNT_RELATIME \ > > > - | MNT_READONLY) > > > + | MNT_READONLY | MNT_NOSYMLINKS) > > > #define MNT_ATIME_MASK (MNT_NOATIME | MNT_NODIRATIME | > > > MNT_RELATIME ) > > > > > > #define MNT_INTERNAL_FLAGS (MNT_SHARED | MNT_WRITE_HOLD | > > > MNT_INTERNAL | \ > > > diff --git a/include/linux/statfs.h b/include/linux/statfs.h > > > index 0166d32..994b059 100644 > > > --- a/include/linux/statfs.h > > > +++ b/include/linux/statfs.h > > > @@ -39,5 +39,6 @@ struct kstatfs { > > > #define ST_NOATIME 0x0400 /* do not update access times */ > > > #define ST_NODIRATIME 0x0800 /* do not update directory access > > > times */ > > > #define ST_RELATIME 0x1000 /* update atime relative to > > > mtime/ctime */ > > > +#define ST_NOSYMLINKS 0x2000 /* do not follow symbolic links > > > */ > > > > > > #endif > > > diff --git a/include/uapi/linux/fs.h b/include/uapi/linux/fs.h > > > index acb2b61..06193d8 100644 > > > --- a/include/uapi/linux/fs.h > > > +++ b/include/uapi/linux/fs.h > > > @@ -130,6 +130,7 @@ struct inodes_stat_t { > > > #define MS_I_VERSION (1<<23) /* Update inode I_version field > > > */ > > > #define MS_STRICTATIME (1<<24) /* Always perform atime updates > > > */ > > > #define MS_LAZYTIME (1<<25) /* Update the on-disk [acm]times > > > lazily */ > > > +#define MS_NOSYMLINKS (1<<26) /* Do not follow symbolic links > > > */ > > > > > > /* These sb flags are internal to the kernel */ > > > #define MS_NOREMOTELOCK (1<<27) > > > -- > > > 2.8.0.rc3.226.g39d4020 > > > > -- > To unsubscribe from this list: send the line "unsubscribe linux > -fsdevel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html >