* [PATCH v7 1/9] net/compat: Add missing sock updates for SCM_RIGHTS
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-10 11:28 ` Christian Brauner
2020-07-09 18:26 ` [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd() Kees Cook
` (7 subsequent siblings)
8 siblings, 1 reply; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, stable, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
Add missed sock updates to compat path via a new helper, which will be
used more in coming patches. (The net/core/scm.c code is left as-is here
to assist with -stable backports for the compat path.)
Cc: stable@vger.kernel.org
Fixes: 48a87cc26c13 ("net: netprio: fd passed in SCM_RIGHTS datagram not set correctly")
Fixes: d84295067fc7 ("net: net_cls: fd passed in SCM_RIGHTS datagram not set correctly")
Signed-off-by: Kees Cook <keescook@chromium.org>
---
include/net/sock.h | 4 ++++
net/compat.c | 1 +
net/core/sock.c | 21 +++++++++++++++++++++
3 files changed, 26 insertions(+)
diff --git a/include/net/sock.h b/include/net/sock.h
index c53cc42b5ab9..2be67f1ee8b1 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -890,6 +890,8 @@ static inline int sk_memalloc_socks(void)
{
return static_branch_unlikely(&memalloc_socks_key);
}
+
+void __receive_sock(struct file *file);
#else
static inline int sk_memalloc_socks(void)
@@ -897,6 +899,8 @@ static inline int sk_memalloc_socks(void)
return 0;
}
+static inline void __receive_sock(struct file *file)
+{ }
#endif
static inline gfp_t sk_gfp_mask(const struct sock *sk, gfp_t gfp_mask)
diff --git a/net/compat.c b/net/compat.c
index 5e3041a2c37d..2937b816107d 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -309,6 +309,7 @@ void scm_detach_fds_compat(struct msghdr *kmsg, struct scm_cookie *scm)
break;
}
/* Bump the usage count and install the file. */
+ __receive_sock(fp[i]);
fd_install(new_fd, get_file(fp[i]));
}
diff --git a/net/core/sock.c b/net/core/sock.c
index 6c4acf1f0220..bde394979041 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2840,6 +2840,27 @@ int sock_no_mmap(struct file *file, struct socket *sock, struct vm_area_struct *
}
EXPORT_SYMBOL(sock_no_mmap);
+/*
+ * When a file is received (via SCM_RIGHTS, etc), we must bump the
+ * various sock-based usage counts.
+ */
+void __receive_sock(struct file *file)
+{
+ struct socket *sock;
+ int error;
+
+ /*
+ * The resulting value of "error" is ignored here since we only
+ * need to take action when the file is a socket and testing
+ * "sock" for NULL is sufficient.
+ */
+ sock = sock_from_file(file, &error);
+ if (sock) {
+ sock_update_netprioidx(&sock->sk->sk_cgrp_data);
+ sock_update_classid(&sock->sk->sk_cgrp_data);
+ }
+}
+
ssize_t sock_no_sendpage(struct socket *sock, struct page *page, int offset, size_t size, int flags)
{
ssize_t res;
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* Re: [PATCH v7 1/9] net/compat: Add missing sock updates for SCM_RIGHTS
2020-07-09 18:26 ` [PATCH v7 1/9] net/compat: Add missing sock updates for SCM_RIGHTS Kees Cook
@ 2020-07-10 11:28 ` Christian Brauner
0 siblings, 0 replies; 19+ messages in thread
From: Christian Brauner @ 2020-07-10 11:28 UTC (permalink / raw)
To: Kees Cook
Cc: linux-kernel, stable, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
On Thu, Jul 09, 2020 at 11:26:34AM -0700, Kees Cook wrote:
> Add missed sock updates to compat path via a new helper, which will be
> used more in coming patches. (The net/core/scm.c code is left as-is here
> to assist with -stable backports for the compat path.)
>
> Cc: stable@vger.kernel.org
> Fixes: 48a87cc26c13 ("net: netprio: fd passed in SCM_RIGHTS datagram not set correctly")
> Fixes: d84295067fc7 ("net: net_cls: fd passed in SCM_RIGHTS datagram not set correctly")
> Signed-off-by: Kees Cook <keescook@chromium.org>
> ---
Thanks!
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd()
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
2020-07-09 18:26 ` [PATCH v7 1/9] net/compat: Add missing sock updates for SCM_RIGHTS Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-09 20:00 ` Jann Horn
2020-07-09 18:26 ` [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds() Kees Cook
` (6 subsequent siblings)
8 siblings, 1 reply; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, stable, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
The sock counting (sock_update_netprioidx() and sock_update_classid())
was missing from pidfd's implementation of received fd installation. Add
a call to the new __receive_sock() helper.
Cc: stable@vger.kernel.org
Fixes: 8649c322f75c ("pid: Implement pidfd_getfd syscall")
Signed-off-by: Kees Cook <keescook@chromium.org>
---
kernel/pid.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/kernel/pid.c b/kernel/pid.c
index f1496b757162..85ed00abdc7c 100644
--- a/kernel/pid.c
+++ b/kernel/pid.c
@@ -42,6 +42,7 @@
#include <linux/sched/signal.h>
#include <linux/sched/task.h>
#include <linux/idr.h>
+#include <net/sock.h>
struct pid init_struct_pid = {
.count = REFCOUNT_INIT(1),
@@ -642,10 +643,12 @@ static int pidfd_getfd(struct pid *pid, int fd)
}
ret = get_unused_fd_flags(O_CLOEXEC);
- if (ret < 0)
+ if (ret < 0) {
fput(file);
- else
+ } else {
fd_install(ret, file);
+ __receive_sock(file);
+ }
return ret;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* Re: [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd()
2020-07-09 18:26 ` [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd() Kees Cook
@ 2020-07-09 20:00 ` Jann Horn
2020-07-09 21:17 ` Kees Cook
2020-07-09 22:35 ` Kees Cook
0 siblings, 2 replies; 19+ messages in thread
From: Jann Horn @ 2020-07-09 20:00 UTC (permalink / raw)
To: Kees Cook
Cc: kernel list, stable, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
Network Development, Linux Containers, Linux API, linux-fsdevel,
open list:KERNEL SELFTEST FRAMEWORK
On Thu, Jul 9, 2020 at 8:26 PM Kees Cook <keescook@chromium.org> wrote:
> The sock counting (sock_update_netprioidx() and sock_update_classid())
> was missing from pidfd's implementation of received fd installation. Add
> a call to the new __receive_sock() helper.
[...]
> diff --git a/kernel/pid.c b/kernel/pid.c
[...]
> @@ -642,10 +643,12 @@ static int pidfd_getfd(struct pid *pid, int fd)
> }
>
> ret = get_unused_fd_flags(O_CLOEXEC);
> - if (ret < 0)
> + if (ret < 0) {
> fput(file);
> - else
> + } else {
> fd_install(ret, file);
> + __receive_sock(file);
> + }
__receive_sock() has to be before fd_install(), otherwise `file` can
be a dangling pointer.
^ permalink raw reply [flat|nested] 19+ messages in thread* Re: [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd()
2020-07-09 20:00 ` Jann Horn
@ 2020-07-09 21:17 ` Kees Cook
2020-07-09 22:35 ` Kees Cook
1 sibling, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 21:17 UTC (permalink / raw)
To: Jann Horn
Cc: kernel list, stable, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
Network Development, Linux Containers, Linux API, linux-fsdevel,
open list:KERNEL SELFTEST FRAMEWORK
On Thu, Jul 09, 2020 at 10:00:42PM +0200, Jann Horn wrote:
> On Thu, Jul 9, 2020 at 8:26 PM Kees Cook <keescook@chromium.org> wrote:
> > The sock counting (sock_update_netprioidx() and sock_update_classid())
> > was missing from pidfd's implementation of received fd installation. Add
> > a call to the new __receive_sock() helper.
> [...]
> > diff --git a/kernel/pid.c b/kernel/pid.c
> [...]
> > @@ -642,10 +643,12 @@ static int pidfd_getfd(struct pid *pid, int fd)
> > }
> >
> > ret = get_unused_fd_flags(O_CLOEXEC);
> > - if (ret < 0)
> > + if (ret < 0) {
> > fput(file);
> > - else
> > + } else {
> > fd_install(ret, file);
> > + __receive_sock(file);
> > + }
>
> __receive_sock() has to be before fd_install(), otherwise `file` can
> be a dangling pointer.
Burned by fd_install()'s API again. Thanks. I will respin.
--
Kees Cook
^ permalink raw reply [flat|nested] 19+ messages in thread* Re: [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd()
2020-07-09 20:00 ` Jann Horn
2020-07-09 21:17 ` Kees Cook
@ 2020-07-09 22:35 ` Kees Cook
1 sibling, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 22:35 UTC (permalink / raw)
To: Jann Horn
Cc: kernel list, stable, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
Network Development, Linux Containers, Linux API, linux-fsdevel,
open list:KERNEL SELFTEST FRAMEWORK
On Thu, Jul 09, 2020 at 10:00:42PM +0200, Jann Horn wrote:
> On Thu, Jul 9, 2020 at 8:26 PM Kees Cook <keescook@chromium.org> wrote:
> > The sock counting (sock_update_netprioidx() and sock_update_classid())
> > was missing from pidfd's implementation of received fd installation. Add
> > a call to the new __receive_sock() helper.
> [...]
> > diff --git a/kernel/pid.c b/kernel/pid.c
> [...]
> > @@ -642,10 +643,12 @@ static int pidfd_getfd(struct pid *pid, int fd)
> > }
> >
> > ret = get_unused_fd_flags(O_CLOEXEC);
> > - if (ret < 0)
> > + if (ret < 0) {
> > fput(file);
> > - else
> > + } else {
> > fd_install(ret, file);
> > + __receive_sock(file);
> > + }
>
> __receive_sock() has to be before fd_install(), otherwise `file` can
> be a dangling pointer.
I've swapped the order now and double-checked the other uses. Everything
else seems fine.
--
Kees Cook
^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds()
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
2020-07-09 18:26 ` [PATCH v7 1/9] net/compat: Add missing sock updates for SCM_RIGHTS Kees Cook
2020-07-09 18:26 ` [PATCH v7 2/9] pidfd: Add missing sock updates for pidfd_getfd() Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-08-07 20:29 ` John Stultz
2020-07-09 18:26 ` [PATCH v7 4/9] fs: Move __scm_install_fd() to __receive_fd() Kees Cook
` (5 subsequent siblings)
8 siblings, 1 reply; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Christian Brauner, Sargun Dhillon, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
Duplicate the cleanups from commit 2618d530dd8b ("net/scm: cleanup
scm_detach_fds") into the compat code.
Replace open-coded __receive_sock() with a call to the helper.
Move the check added in commit 1f466e1f15cf ("net: cleanly handle kernel
vs user buffers for ->msg_control") to before the compat call, even
though it should be impossible for an in-kernel call to also be compat.
Correct the int "flags" argument to unsigned int to match fd_install()
and similar APIs.
Regularize any remaining differences, including a whitespace issue,
a checkpatch warning, and add the check from commit 6900317f5eff ("net,
scm: fix PaX detected msg_controllen overflow in scm_detach_fds") which
fixed an overflow unique to 64-bit. To avoid confusion when comparing
the compat handler to the native handler, just include the same check
in the compat handler.
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
include/net/scm.h | 1 +
net/compat.c | 56 +++++++++++++++++++++--------------------------
net/core/scm.c | 27 ++++++++++-------------
3 files changed, 37 insertions(+), 47 deletions(-)
diff --git a/include/net/scm.h b/include/net/scm.h
index 1ce365f4c256..581a94d6c613 100644
--- a/include/net/scm.h
+++ b/include/net/scm.h
@@ -37,6 +37,7 @@ struct scm_cookie {
#endif
};
+int __scm_install_fd(struct file *file, int __user *ufd, unsigned int o_flags);
void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm);
void scm_detach_fds_compat(struct msghdr *msg, struct scm_cookie *scm);
int __scm_send(struct socket *sock, struct msghdr *msg, struct scm_cookie *scm);
diff --git a/net/compat.c b/net/compat.c
index 2937b816107d..27d477fdcaa0 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -281,40 +281,31 @@ int put_cmsg_compat(struct msghdr *kmsg, int level, int type, int len, void *dat
return 0;
}
-void scm_detach_fds_compat(struct msghdr *kmsg, struct scm_cookie *scm)
+static int scm_max_fds_compat(struct msghdr *msg)
{
- struct compat_cmsghdr __user *cm = (struct compat_cmsghdr __user *) kmsg->msg_control;
- int fdmax = (kmsg->msg_controllen - sizeof(struct compat_cmsghdr)) / sizeof(int);
- int fdnum = scm->fp->count;
- struct file **fp = scm->fp->fp;
- int __user *cmfptr;
- int err = 0, i;
+ if (msg->msg_controllen <= sizeof(struct compat_cmsghdr))
+ return 0;
+ return (msg->msg_controllen - sizeof(struct compat_cmsghdr)) / sizeof(int);
+}
- if (fdnum < fdmax)
- fdmax = fdnum;
+void scm_detach_fds_compat(struct msghdr *msg, struct scm_cookie *scm)
+{
+ struct compat_cmsghdr __user *cm =
+ (struct compat_cmsghdr __user *)msg->msg_control;
+ unsigned int o_flags = (msg->msg_flags & MSG_CMSG_CLOEXEC) ? O_CLOEXEC : 0;
+ int fdmax = min_t(int, scm_max_fds_compat(msg), scm->fp->count);
+ int __user *cmsg_data = CMSG_USER_DATA(cm);
+ int err = 0, i;
- for (i = 0, cmfptr = (int __user *) CMSG_COMPAT_DATA(cm); i < fdmax; i++, cmfptr++) {
- int new_fd;
- err = security_file_receive(fp[i]);
+ for (i = 0; i < fdmax; i++) {
+ err = __scm_install_fd(scm->fp->fp[i], cmsg_data + i, o_flags);
if (err)
break;
- err = get_unused_fd_flags(MSG_CMSG_CLOEXEC & kmsg->msg_flags
- ? O_CLOEXEC : 0);
- if (err < 0)
- break;
- new_fd = err;
- err = put_user(new_fd, cmfptr);
- if (err) {
- put_unused_fd(new_fd);
- break;
- }
- /* Bump the usage count and install the file. */
- __receive_sock(fp[i]);
- fd_install(new_fd, get_file(fp[i]));
}
if (i > 0) {
int cmlen = CMSG_COMPAT_LEN(i * sizeof(int));
+
err = put_user(SOL_SOCKET, &cm->cmsg_level);
if (!err)
err = put_user(SCM_RIGHTS, &cm->cmsg_type);
@@ -322,16 +313,19 @@ void scm_detach_fds_compat(struct msghdr *kmsg, struct scm_cookie *scm)
err = put_user(cmlen, &cm->cmsg_len);
if (!err) {
cmlen = CMSG_COMPAT_SPACE(i * sizeof(int));
- kmsg->msg_control += cmlen;
- kmsg->msg_controllen -= cmlen;
+ if (msg->msg_controllen < cmlen)
+ cmlen = msg->msg_controllen;
+ msg->msg_control += cmlen;
+ msg->msg_controllen -= cmlen;
}
}
- if (i < fdnum)
- kmsg->msg_flags |= MSG_CTRUNC;
+
+ if (i < scm->fp->count || (scm->fp->count && fdmax <= 0))
+ msg->msg_flags |= MSG_CTRUNC;
/*
- * All of the files that fit in the message have had their
- * usage counts incremented, so we just free the list.
+ * All of the files that fit in the message have had their usage counts
+ * incremented, so we just free the list.
*/
__scm_destroy(scm);
}
diff --git a/net/core/scm.c b/net/core/scm.c
index 875df1c2989d..44f03213dcab 100644
--- a/net/core/scm.c
+++ b/net/core/scm.c
@@ -280,9 +280,8 @@ void put_cmsg_scm_timestamping(struct msghdr *msg, struct scm_timestamping_inter
}
EXPORT_SYMBOL(put_cmsg_scm_timestamping);
-static int __scm_install_fd(struct file *file, int __user *ufd, int o_flags)
+int __scm_install_fd(struct file *file, int __user *ufd, unsigned int o_flags)
{
- struct socket *sock;
int new_fd;
int error;
@@ -300,12 +299,8 @@ static int __scm_install_fd(struct file *file, int __user *ufd, int o_flags)
return error;
}
- /* Bump the usage count and install the file. */
- sock = sock_from_file(file, &error);
- if (sock) {
- sock_update_netprioidx(&sock->sk->sk_cgrp_data);
- sock_update_classid(&sock->sk->sk_cgrp_data);
- }
+ /* Bump the sock usage counts, if any. */
+ __receive_sock(file);
fd_install(new_fd, get_file(file));
return 0;
}
@@ -319,29 +314,29 @@ static int scm_max_fds(struct msghdr *msg)
void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm)
{
- struct cmsghdr __user *cm
- = (__force struct cmsghdr __user*)msg->msg_control;
- int o_flags = (msg->msg_flags & MSG_CMSG_CLOEXEC) ? O_CLOEXEC : 0;
+ struct cmsghdr __user *cm =
+ (__force struct cmsghdr __user *)msg->msg_control;
+ unsigned int o_flags = (msg->msg_flags & MSG_CMSG_CLOEXEC) ? O_CLOEXEC : 0;
int fdmax = min_t(int, scm_max_fds(msg), scm->fp->count);
int __user *cmsg_data = CMSG_USER_DATA(cm);
int err = 0, i;
+ /* no use for FD passing from kernel space callers */
+ if (WARN_ON_ONCE(!msg->msg_control_is_user))
+ return;
+
if (msg->msg_flags & MSG_CMSG_COMPAT) {
scm_detach_fds_compat(msg, scm);
return;
}
- /* no use for FD passing from kernel space callers */
- if (WARN_ON_ONCE(!msg->msg_control_is_user))
- return;
-
for (i = 0; i < fdmax; i++) {
err = __scm_install_fd(scm->fp->fp[i], cmsg_data + i, o_flags);
if (err)
break;
}
- if (i > 0) {
+ if (i > 0) {
int cmlen = CMSG_LEN(i * sizeof(int));
err = put_user(SOL_SOCKET, &cm->cmsg_level);
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* Re: [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds()
2020-07-09 18:26 ` [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds() Kees Cook
@ 2020-08-07 20:29 ` John Stultz
2020-08-07 22:18 ` Kees Cook
0 siblings, 1 reply; 19+ messages in thread
From: John Stultz @ 2020-08-07 20:29 UTC (permalink / raw)
To: Kees Cook
Cc: Linux Kernel Mailing List, Christian Brauner, Sargun Dhillon,
Christian Brauner, Tycho Andersen, David Laight,
Christoph Hellwig, David S. Miller, Jakub Kicinski,
Alexander Viro, Aleksa Sarai, Matt Denton, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest,
Amit Pundir, Bjorn Andersson
On Thu, Jul 9, 2020 at 11:28 AM Kees Cook <keescook@chromium.org> wrote:
>
> Duplicate the cleanups from commit 2618d530dd8b ("net/scm: cleanup
> scm_detach_fds") into the compat code.
>
> Replace open-coded __receive_sock() with a call to the helper.
>
> Move the check added in commit 1f466e1f15cf ("net: cleanly handle kernel
> vs user buffers for ->msg_control") to before the compat call, even
> though it should be impossible for an in-kernel call to also be compat.
>
> Correct the int "flags" argument to unsigned int to match fd_install()
> and similar APIs.
>
> Regularize any remaining differences, including a whitespace issue,
> a checkpatch warning, and add the check from commit 6900317f5eff ("net,
> scm: fix PaX detected msg_controllen overflow in scm_detach_fds") which
> fixed an overflow unique to 64-bit. To avoid confusion when comparing
> the compat handler to the native handler, just include the same check
> in the compat handler.
>
> Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
> Signed-off-by: Kees Cook <keescook@chromium.org>
> ---
Hey Kees,
So during the merge window (while chasing a few other regressions),
I noticed occasionally my Dragonboard 845c running AOSP having trouble
with the web browser crashing or other apps hanging, and I've bisected
the issue down to this change.
Unfortunately it doesn't revert cleanly so I can't validate reverting
it sorts things against linus/HEAD. Anyway, I wanted to check and see
if you had any other reports of similar or any ideas what might be
going wrong?
thanks
-john
^ permalink raw reply [flat|nested] 19+ messages in thread* Re: [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds()
2020-08-07 20:29 ` John Stultz
@ 2020-08-07 22:18 ` Kees Cook
2020-08-08 0:02 ` John Stultz
0 siblings, 1 reply; 19+ messages in thread
From: Kees Cook @ 2020-08-07 22:18 UTC (permalink / raw)
To: John Stultz
Cc: Linux Kernel Mailing List, Christian Brauner, Sargun Dhillon,
Christian Brauner, Tycho Andersen, David Laight,
Christoph Hellwig, David S. Miller, Jakub Kicinski,
Alexander Viro, Aleksa Sarai, Matt Denton, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest,
Amit Pundir, Bjorn Andersson
On Fri, Aug 07, 2020 at 01:29:24PM -0700, John Stultz wrote:
> On Thu, Jul 9, 2020 at 11:28 AM Kees Cook <keescook@chromium.org> wrote:
> >
> > Duplicate the cleanups from commit 2618d530dd8b ("net/scm: cleanup
> > scm_detach_fds") into the compat code.
> >
> > Replace open-coded __receive_sock() with a call to the helper.
> >
> > Move the check added in commit 1f466e1f15cf ("net: cleanly handle kernel
> > vs user buffers for ->msg_control") to before the compat call, even
> > though it should be impossible for an in-kernel call to also be compat.
> >
> > Correct the int "flags" argument to unsigned int to match fd_install()
> > and similar APIs.
> >
> > Regularize any remaining differences, including a whitespace issue,
> > a checkpatch warning, and add the check from commit 6900317f5eff ("net,
> > scm: fix PaX detected msg_controllen overflow in scm_detach_fds") which
> > fixed an overflow unique to 64-bit. To avoid confusion when comparing
> > the compat handler to the native handler, just include the same check
> > in the compat handler.
> >
> > Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
> > Signed-off-by: Kees Cook <keescook@chromium.org>
> > ---
>
> Hey Kees,
> So during the merge window (while chasing a few other regressions),
> I noticed occasionally my Dragonboard 845c running AOSP having trouble
> with the web browser crashing or other apps hanging, and I've bisected
> the issue down to this change.
>
> Unfortunately it doesn't revert cleanly so I can't validate reverting
> it sorts things against linus/HEAD. Anyway, I wanted to check and see
> if you had any other reports of similar or any ideas what might be
> going wrong?
Hi; Yes, sorry for the trouble. I had a typo in a refactor of
SCM_RIGHTS. I suspect it'll be fixed by this:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1fa2c0a0c814fbae0eb3e79a510765225570d043
Can you verify Linus's latest tree works for you? If not, there might be
something else hiding in the corners...
Thanks!
--
Kees Cook
^ permalink raw reply [flat|nested] 19+ messages in thread* Re: [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds()
2020-08-07 22:18 ` Kees Cook
@ 2020-08-08 0:02 ` John Stultz
2020-08-08 7:17 ` Kees Cook
0 siblings, 1 reply; 19+ messages in thread
From: John Stultz @ 2020-08-08 0:02 UTC (permalink / raw)
To: Kees Cook
Cc: Linux Kernel Mailing List, Christian Brauner, Sargun Dhillon,
Christian Brauner, Tycho Andersen, David Laight,
Christoph Hellwig, David S. Miller, Jakub Kicinski,
Alexander Viro, Aleksa Sarai, Matt Denton, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, Linux API, linux-fsdevel, linux-kselftest,
Amit Pundir, Bjorn Andersson
On Fri, Aug 7, 2020 at 3:18 PM Kees Cook <keescook@chromium.org> wrote:
>
> On Fri, Aug 07, 2020 at 01:29:24PM -0700, John Stultz wrote:
> > On Thu, Jul 9, 2020 at 11:28 AM Kees Cook <keescook@chromium.org> wrote:
> > >
> > > Duplicate the cleanups from commit 2618d530dd8b ("net/scm: cleanup
> > > scm_detach_fds") into the compat code.
> > >
> > > Replace open-coded __receive_sock() with a call to the helper.
> > >
> > > Move the check added in commit 1f466e1f15cf ("net: cleanly handle kernel
> > > vs user buffers for ->msg_control") to before the compat call, even
> > > though it should be impossible for an in-kernel call to also be compat.
> > >
> > > Correct the int "flags" argument to unsigned int to match fd_install()
> > > and similar APIs.
> > >
> > > Regularize any remaining differences, including a whitespace issue,
> > > a checkpatch warning, and add the check from commit 6900317f5eff ("net,
> > > scm: fix PaX detected msg_controllen overflow in scm_detach_fds") which
> > > fixed an overflow unique to 64-bit. To avoid confusion when comparing
> > > the compat handler to the native handler, just include the same check
> > > in the compat handler.
> > >
> > > Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
> > > Signed-off-by: Kees Cook <keescook@chromium.org>
> > > ---
> >
> > Hey Kees,
> > So during the merge window (while chasing a few other regressions),
> > I noticed occasionally my Dragonboard 845c running AOSP having trouble
> > with the web browser crashing or other apps hanging, and I've bisected
> > the issue down to this change.
> >
> > Unfortunately it doesn't revert cleanly so I can't validate reverting
> > it sorts things against linus/HEAD. Anyway, I wanted to check and see
> > if you had any other reports of similar or any ideas what might be
> > going wrong?
>
> Hi; Yes, sorry for the trouble. I had a typo in a refactor of
> SCM_RIGHTS. I suspect it'll be fixed by this:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1fa2c0a0c814fbae0eb3e79a510765225570d043
>
> Can you verify Linus's latest tree works for you? If not, there might be
> something else hiding in the corners...
Thanks so much! Yes, I just updated to Linus' latest and the issue has
disappeared!
thanks again!
-john
^ permalink raw reply [flat|nested] 19+ messages in thread* Re: [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds()
2020-08-08 0:02 ` John Stultz
@ 2020-08-08 7:17 ` Kees Cook
0 siblings, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-08-08 7:17 UTC (permalink / raw)
To: John Stultz
Cc: Linux Kernel Mailing List, Christian Brauner, Sargun Dhillon,
Christian Brauner, Tycho Andersen, David Laight,
Christoph Hellwig, David S. Miller, Jakub Kicinski,
Alexander Viro, Aleksa Sarai, Matt Denton, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, Linux API, linux-fsdevel, linux-kselftest,
Amit Pundir, Bjorn Andersson
On Fri, Aug 07, 2020 at 05:02:15PM -0700, John Stultz wrote:
> On Fri, Aug 7, 2020 at 3:18 PM Kees Cook <keescook@chromium.org> wrote:
> >
> > On Fri, Aug 07, 2020 at 01:29:24PM -0700, John Stultz wrote:
> > > On Thu, Jul 9, 2020 at 11:28 AM Kees Cook <keescook@chromium.org> wrote:
> > > >
> > > > Duplicate the cleanups from commit 2618d530dd8b ("net/scm: cleanup
> > > > scm_detach_fds") into the compat code.
> > > >
> > > > Replace open-coded __receive_sock() with a call to the helper.
> > > >
> > > > Move the check added in commit 1f466e1f15cf ("net: cleanly handle kernel
> > > > vs user buffers for ->msg_control") to before the compat call, even
> > > > though it should be impossible for an in-kernel call to also be compat.
> > > >
> > > > Correct the int "flags" argument to unsigned int to match fd_install()
> > > > and similar APIs.
> > > >
> > > > Regularize any remaining differences, including a whitespace issue,
> > > > a checkpatch warning, and add the check from commit 6900317f5eff ("net,
> > > > scm: fix PaX detected msg_controllen overflow in scm_detach_fds") which
> > > > fixed an overflow unique to 64-bit. To avoid confusion when comparing
> > > > the compat handler to the native handler, just include the same check
> > > > in the compat handler.
> > > >
> > > > Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
> > > > Signed-off-by: Kees Cook <keescook@chromium.org>
> > > > ---
> > >
> > > Hey Kees,
> > > So during the merge window (while chasing a few other regressions),
> > > I noticed occasionally my Dragonboard 845c running AOSP having trouble
> > > with the web browser crashing or other apps hanging, and I've bisected
> > > the issue down to this change.
> > >
> > > Unfortunately it doesn't revert cleanly so I can't validate reverting
> > > it sorts things against linus/HEAD. Anyway, I wanted to check and see
> > > if you had any other reports of similar or any ideas what might be
> > > going wrong?
> >
> > Hi; Yes, sorry for the trouble. I had a typo in a refactor of
> > SCM_RIGHTS. I suspect it'll be fixed by this:
> > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1fa2c0a0c814fbae0eb3e79a510765225570d043
> >
> > Can you verify Linus's latest tree works for you? If not, there might be
> > something else hiding in the corners...
>
> Thanks so much! Yes, I just updated to Linus' latest and the issue has
> disappeared!
>
> thanks again!
Whew; sorry again and thanks for testing! :)
--
Kees Cook
^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH v7 4/9] fs: Move __scm_install_fd() to __receive_fd()
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
` (2 preceding siblings ...)
2020-07-09 18:26 ` [PATCH v7 3/9] net/scm: Regularize compat handling of scm_detach_fds() Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-09 18:26 ` [PATCH v7 5/9] fs: Add receive_fd() wrapper for __receive_fd() Kees Cook
` (4 subsequent siblings)
8 siblings, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Sargun Dhillon, Christian Brauner, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
In preparation for users of the "install a received file" logic outside
of net/ (pidfd and seccomp), relocate and rename __scm_install_fd() from
net/core/scm.c to __receive_fd() in fs/file.c, and provide a wrapper
named receive_fd_user(), as future patches will change the interface
to __receive_fd().
Reviewed-by: Sargun Dhillon <sargun@sargun.me>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
fs/file.c | 41 +++++++++++++++++++++++++++++++++++++++++
include/linux/file.h | 8 ++++++++
include/net/scm.h | 1 -
net/compat.c | 2 +-
net/core/scm.c | 27 +--------------------------
5 files changed, 51 insertions(+), 28 deletions(-)
diff --git a/fs/file.c b/fs/file.c
index abb8b7081d7a..6220bf440809 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -18,6 +18,7 @@
#include <linux/bitops.h>
#include <linux/spinlock.h>
#include <linux/rcupdate.h>
+#include <net/sock.h>
unsigned int sysctl_nr_open __read_mostly = 1024*1024;
unsigned int sysctl_nr_open_min = BITS_PER_LONG;
@@ -931,6 +932,46 @@ int replace_fd(unsigned fd, struct file *file, unsigned flags)
return err;
}
+/**
+ * __receive_fd() - Install received file into file descriptor table
+ *
+ * @file: struct file that was received from another process
+ * @ufd: __user pointer to write new fd number to
+ * @o_flags: the O_* flags to apply to the new fd entry
+ *
+ * Installs a received file into the file descriptor table, with appropriate
+ * checks and count updates. Writes the fd number to userspace.
+ *
+ * This helper handles its own reference counting of the incoming
+ * struct file.
+ *
+ * Returns -ve on error.
+ */
+int __receive_fd(struct file *file, int __user *ufd, unsigned int o_flags)
+{
+ int new_fd;
+ int error;
+
+ error = security_file_receive(file);
+ if (error)
+ return error;
+
+ new_fd = get_unused_fd_flags(o_flags);
+ if (new_fd < 0)
+ return new_fd;
+
+ error = put_user(new_fd, ufd);
+ if (error) {
+ put_unused_fd(new_fd);
+ return error;
+ }
+
+ /* Bump the sock usage counts, if any. */
+ __receive_sock(file);
+ fd_install(new_fd, get_file(file));
+ return 0;
+}
+
static int ksys_dup3(unsigned int oldfd, unsigned int newfd, int flags)
{
int err = -EBADF;
diff --git a/include/linux/file.h b/include/linux/file.h
index 122f80084a3e..b14ff2ffd0bd 100644
--- a/include/linux/file.h
+++ b/include/linux/file.h
@@ -91,6 +91,14 @@ extern void put_unused_fd(unsigned int fd);
extern void fd_install(unsigned int fd, struct file *file);
+extern int __receive_fd(struct file *file, int __user *ufd,
+ unsigned int o_flags);
+static inline int receive_fd_user(struct file *file, int __user *ufd,
+ unsigned int o_flags)
+{
+ return __receive_fd(file, ufd, o_flags);
+}
+
extern void flush_delayed_fput(void);
extern void __fput_sync(struct file *);
diff --git a/include/net/scm.h b/include/net/scm.h
index 581a94d6c613..1ce365f4c256 100644
--- a/include/net/scm.h
+++ b/include/net/scm.h
@@ -37,7 +37,6 @@ struct scm_cookie {
#endif
};
-int __scm_install_fd(struct file *file, int __user *ufd, unsigned int o_flags);
void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm);
void scm_detach_fds_compat(struct msghdr *msg, struct scm_cookie *scm);
int __scm_send(struct socket *sock, struct msghdr *msg, struct scm_cookie *scm);
diff --git a/net/compat.c b/net/compat.c
index 27d477fdcaa0..e74cd3dae8b0 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -298,7 +298,7 @@ void scm_detach_fds_compat(struct msghdr *msg, struct scm_cookie *scm)
int err = 0, i;
for (i = 0; i < fdmax; i++) {
- err = __scm_install_fd(scm->fp->fp[i], cmsg_data + i, o_flags);
+ err = receive_fd_user(scm->fp->fp[i], cmsg_data + i, o_flags);
if (err)
break;
}
diff --git a/net/core/scm.c b/net/core/scm.c
index 44f03213dcab..67c166a7820d 100644
--- a/net/core/scm.c
+++ b/net/core/scm.c
@@ -280,31 +280,6 @@ void put_cmsg_scm_timestamping(struct msghdr *msg, struct scm_timestamping_inter
}
EXPORT_SYMBOL(put_cmsg_scm_timestamping);
-int __scm_install_fd(struct file *file, int __user *ufd, unsigned int o_flags)
-{
- int new_fd;
- int error;
-
- error = security_file_receive(file);
- if (error)
- return error;
-
- new_fd = get_unused_fd_flags(o_flags);
- if (new_fd < 0)
- return new_fd;
-
- error = put_user(new_fd, ufd);
- if (error) {
- put_unused_fd(new_fd);
- return error;
- }
-
- /* Bump the sock usage counts, if any. */
- __receive_sock(file);
- fd_install(new_fd, get_file(file));
- return 0;
-}
-
static int scm_max_fds(struct msghdr *msg)
{
if (msg->msg_controllen <= sizeof(struct cmsghdr))
@@ -331,7 +306,7 @@ void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm)
}
for (i = 0; i < fdmax; i++) {
- err = __scm_install_fd(scm->fp->fp[i], cmsg_data + i, o_flags);
+ err = receive_fd_user(scm->fp->fp[i], cmsg_data + i, o_flags);
if (err)
break;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* [PATCH v7 5/9] fs: Add receive_fd() wrapper for __receive_fd()
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
` (3 preceding siblings ...)
2020-07-09 18:26 ` [PATCH v7 4/9] fs: Move __scm_install_fd() to __receive_fd() Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-09 18:26 ` [PATCH v7 6/9] pidfd: Replace open-coded receive_fd() Kees Cook
` (3 subsequent siblings)
8 siblings, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Sargun Dhillon, Christian Brauner, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
For both pidfd and seccomp, the __user pointer is not used. Update
__receive_fd() to make writing to ufd optional via a NULL check. However,
for the receive_fd_user() wrapper, ufd is NULL checked so an -EFAULT
can be returned to avoid changing the SCM_RIGHTS interface behavior. Add
new wrapper receive_fd() for pidfd and seccomp that does not use the ufd
argument. For the new helper, the allocated fd needs to be returned on
success. Update the existing callers to handle it.
Reviewed-by: Sargun Dhillon <sargun@sargun.me>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
fs/file.c | 17 ++++++++++-------
include/linux/file.h | 7 +++++++
net/compat.c | 2 +-
net/core/scm.c | 2 +-
4 files changed, 19 insertions(+), 9 deletions(-)
diff --git a/fs/file.c b/fs/file.c
index 6220bf440809..87954bab9306 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -940,12 +940,13 @@ int replace_fd(unsigned fd, struct file *file, unsigned flags)
* @o_flags: the O_* flags to apply to the new fd entry
*
* Installs a received file into the file descriptor table, with appropriate
- * checks and count updates. Writes the fd number to userspace.
+ * checks and count updates. Optionally writes the fd number to userspace, if
+ * @ufd is non-NULL.
*
* This helper handles its own reference counting of the incoming
* struct file.
*
- * Returns -ve on error.
+ * Returns newly install fd or -ve on error.
*/
int __receive_fd(struct file *file, int __user *ufd, unsigned int o_flags)
{
@@ -960,16 +961,18 @@ int __receive_fd(struct file *file, int __user *ufd, unsigned int o_flags)
if (new_fd < 0)
return new_fd;
- error = put_user(new_fd, ufd);
- if (error) {
- put_unused_fd(new_fd);
- return error;
+ if (ufd) {
+ error = put_user(new_fd, ufd);
+ if (error) {
+ put_unused_fd(new_fd);
+ return error;
+ }
}
/* Bump the sock usage counts, if any. */
__receive_sock(file);
fd_install(new_fd, get_file(file));
- return 0;
+ return new_fd;
}
static int ksys_dup3(unsigned int oldfd, unsigned int newfd, int flags)
diff --git a/include/linux/file.h b/include/linux/file.h
index b14ff2ffd0bd..d9fee9f5c8da 100644
--- a/include/linux/file.h
+++ b/include/linux/file.h
@@ -9,6 +9,7 @@
#include <linux/compiler.h>
#include <linux/types.h>
#include <linux/posix_types.h>
+#include <linux/errno.h>
struct file;
@@ -96,8 +97,14 @@ extern int __receive_fd(struct file *file, int __user *ufd,
static inline int receive_fd_user(struct file *file, int __user *ufd,
unsigned int o_flags)
{
+ if (ufd == NULL)
+ return -EFAULT;
return __receive_fd(file, ufd, o_flags);
}
+static inline int receive_fd(struct file *file, unsigned int o_flags)
+{
+ return __receive_fd(file, NULL, o_flags);
+}
extern void flush_delayed_fput(void);
extern void __fput_sync(struct file *);
diff --git a/net/compat.c b/net/compat.c
index e74cd3dae8b0..dc7ddbc2b15e 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -299,7 +299,7 @@ void scm_detach_fds_compat(struct msghdr *msg, struct scm_cookie *scm)
for (i = 0; i < fdmax; i++) {
err = receive_fd_user(scm->fp->fp[i], cmsg_data + i, o_flags);
- if (err)
+ if (err < 0)
break;
}
diff --git a/net/core/scm.c b/net/core/scm.c
index 67c166a7820d..8156d4fb8a39 100644
--- a/net/core/scm.c
+++ b/net/core/scm.c
@@ -307,7 +307,7 @@ void scm_detach_fds(struct msghdr *msg, struct scm_cookie *scm)
for (i = 0; i < fdmax; i++) {
err = receive_fd_user(scm->fp->fp[i], cmsg_data + i, o_flags);
- if (err)
+ if (err < 0)
break;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* [PATCH v7 6/9] pidfd: Replace open-coded receive_fd()
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
` (4 preceding siblings ...)
2020-07-09 18:26 ` [PATCH v7 5/9] fs: Add receive_fd() wrapper for __receive_fd() Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-09 18:26 ` [PATCH v7 7/9] fs: Expand __receive_fd() to accept existing fd Kees Cook
` (2 subsequent siblings)
8 siblings, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Sargun Dhillon, Christian Brauner, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
Replace the open-coded version of receive_fd() with a call to the
new helper.
Thanks to Vamshi K Sthambamkadi <vamshi.k.sthambamkadi@gmail.com> for
catching a missed fput() in an earlier version of this patch.
Reviewed-by: Sargun Dhillon <sargun@sargun.me>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
kernel/pid.c | 15 ++-------------
1 file changed, 2 insertions(+), 13 deletions(-)
diff --git a/kernel/pid.c b/kernel/pid.c
index 85ed00abdc7c..da5aea5f04fa 100644
--- a/kernel/pid.c
+++ b/kernel/pid.c
@@ -636,19 +636,8 @@ static int pidfd_getfd(struct pid *pid, int fd)
if (IS_ERR(file))
return PTR_ERR(file);
- ret = security_file_receive(file);
- if (ret) {
- fput(file);
- return ret;
- }
-
- ret = get_unused_fd_flags(O_CLOEXEC);
- if (ret < 0) {
- fput(file);
- } else {
- fd_install(ret, file);
- __receive_sock(file);
- }
+ ret = receive_fd(file, O_CLOEXEC);
+ fput(file);
return ret;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* [PATCH v7 7/9] fs: Expand __receive_fd() to accept existing fd
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
` (5 preceding siblings ...)
2020-07-09 18:26 ` [PATCH v7 6/9] pidfd: Replace open-coded receive_fd() Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-09 18:26 ` [PATCH v7 8/9] seccomp: Introduce addfd ioctl to seccomp user notifier Kees Cook
2020-07-09 18:26 ` [PATCH v7 9/9] selftests/seccomp: Test SECCOMP_IOCTL_NOTIF_ADDFD Kees Cook
8 siblings, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Sargun Dhillon, Christian Brauner, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Matt Denton,
Jann Horn, Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
Expand __receive_fd() with support for replace_fd() for the coming seccomp
"addfd" ioctl(). Add new wrapper receive_fd_replace() for the new behavior
and update existing wrappers to retain old behavior.
Thanks to Colin Ian King <colin.king@canonical.com> for pointing out an
uninitialized variable exposure in an earlier version of this patch.
Reviewed-by: Sargun Dhillon <sargun@sargun.me>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
fs/file.c | 25 +++++++++++++++++++------
include/linux/file.h | 10 +++++++---
2 files changed, 26 insertions(+), 9 deletions(-)
diff --git a/fs/file.c b/fs/file.c
index 87954bab9306..feebdc17bf46 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -935,6 +935,7 @@ int replace_fd(unsigned fd, struct file *file, unsigned flags)
/**
* __receive_fd() - Install received file into file descriptor table
*
+ * @fd: fd to install into (if negative, a new fd will be allocated)
* @file: struct file that was received from another process
* @ufd: __user pointer to write new fd number to
* @o_flags: the O_* flags to apply to the new fd entry
@@ -948,7 +949,7 @@ int replace_fd(unsigned fd, struct file *file, unsigned flags)
*
* Returns newly install fd or -ve on error.
*/
-int __receive_fd(struct file *file, int __user *ufd, unsigned int o_flags)
+int __receive_fd(int fd, struct file *file, int __user *ufd, unsigned int o_flags)
{
int new_fd;
int error;
@@ -957,21 +958,33 @@ int __receive_fd(struct file *file, int __user *ufd, unsigned int o_flags)
if (error)
return error;
- new_fd = get_unused_fd_flags(o_flags);
- if (new_fd < 0)
- return new_fd;
+ if (fd < 0) {
+ new_fd = get_unused_fd_flags(o_flags);
+ if (new_fd < 0)
+ return new_fd;
+ } else {
+ new_fd = fd;
+ }
if (ufd) {
error = put_user(new_fd, ufd);
if (error) {
- put_unused_fd(new_fd);
+ if (fd < 0)
+ put_unused_fd(new_fd);
return error;
}
}
+ if (fd < 0) {
+ fd_install(new_fd, get_file(file));
+ } else {
+ error = replace_fd(new_fd, file, o_flags);
+ if (error)
+ return error;
+ }
+
/* Bump the sock usage counts, if any. */
__receive_sock(file);
- fd_install(new_fd, get_file(file));
return new_fd;
}
diff --git a/include/linux/file.h b/include/linux/file.h
index d9fee9f5c8da..225982792fa2 100644
--- a/include/linux/file.h
+++ b/include/linux/file.h
@@ -92,18 +92,22 @@ extern void put_unused_fd(unsigned int fd);
extern void fd_install(unsigned int fd, struct file *file);
-extern int __receive_fd(struct file *file, int __user *ufd,
+extern int __receive_fd(int fd, struct file *file, int __user *ufd,
unsigned int o_flags);
static inline int receive_fd_user(struct file *file, int __user *ufd,
unsigned int o_flags)
{
if (ufd == NULL)
return -EFAULT;
- return __receive_fd(file, ufd, o_flags);
+ return __receive_fd(-1, file, ufd, o_flags);
}
static inline int receive_fd(struct file *file, unsigned int o_flags)
{
- return __receive_fd(file, NULL, o_flags);
+ return __receive_fd(-1, file, NULL, o_flags);
+}
+static inline int receive_fd_replace(int fd, struct file *file, unsigned int o_flags)
+{
+ return __receive_fd(fd, file, NULL, o_flags);
}
extern void flush_delayed_fput(void);
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* [PATCH v7 8/9] seccomp: Introduce addfd ioctl to seccomp user notifier
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
` (6 preceding siblings ...)
2020-07-09 18:26 ` [PATCH v7 7/9] fs: Expand __receive_fd() to accept existing fd Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
2020-07-14 18:20 ` Will Drewry
2020-07-09 18:26 ` [PATCH v7 9/9] selftests/seccomp: Test SECCOMP_IOCTL_NOTIF_ADDFD Kees Cook
8 siblings, 1 reply; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Sargun Dhillon, Matt Denton, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
From: Sargun Dhillon <sargun@sargun.me>
The current SECCOMP_RET_USER_NOTIF API allows for syscall supervision over
an fd. It is often used in settings where a supervising task emulates
syscalls on behalf of a supervised task in userspace, either to further
restrict the supervisee's syscall abilities or to circumvent kernel
enforced restrictions the supervisor deems safe to lift (e.g. actually
performing a mount(2) for an unprivileged container).
While SECCOMP_RET_USER_NOTIF allows for the interception of any syscall,
only a certain subset of syscalls could be correctly emulated. Over the
last few development cycles, the set of syscalls which can't be emulated
has been reduced due to the addition of pidfd_getfd(2). With this we are
now able to, for example, intercept syscalls that require the supervisor
to operate on file descriptors of the supervisee such as connect(2).
However, syscalls that cause new file descriptors to be installed can not
currently be correctly emulated since there is no way for the supervisor
to inject file descriptors into the supervisee. This patch adds a
new addfd ioctl to remove this restriction by allowing the supervisor to
install file descriptors into the intercepted task. By implementing this
feature via seccomp the supervisor effectively instructs the supervisee
to install a set of file descriptors into its own file descriptor table
during the intercepted syscall. This way it is possible to intercept
syscalls such as open() or accept(), and install (or replace, like
dup2(2)) the supervisor's resulting fd into the supervisee. One
replacement use-case would be to redirect the stdout and stderr of a
supervisee into log file descriptors opened by the supervisor.
The ioctl handling is based on the discussions[1] of how Extensible
Arguments should interact with ioctls. Instead of building size into
the addfd structure, make it a function of the ioctl command (which
is how sizes are normally passed to ioctls). To support forward and
backward compatibility, just mask out the direction and size, and match
everything. The size (and any future direction) checks are done along
with copy_struct_from_user() logic.
As a note, the seccomp_notif_addfd structure is laid out based on 8-byte
alignment without requiring packing as there have been packing issues
with uapi highlighted before[2][3]. Although we could overload the
newfd field and use -1 to indicate that it is not to be used, doing
so requires changing the size of the fd field, and introduces struct
packing complexity.
[1]: https://lore.kernel.org/lkml/87o8w9bcaf.fsf@mid.deneb.enyo.de/
[2]: https://lore.kernel.org/lkml/a328b91d-fd8f-4f27-b3c2-91a9c45f18c0@rasmusvillemoes.dk/
[3]: https://lore.kernel.org/lkml/20200612104629.GA15814@ircssh-2.c.rugged-nimbus-611.internal
Suggested-by: Matt Denton <mpdenton@google.com>
Link: https://lore.kernel.org/r/20200603011044.7972-4-sargun@sargun.me
Signed-off-by: Sargun Dhillon <sargun@sargun.me>
Co-developed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
include/linux/seccomp.h | 4 +
include/uapi/linux/seccomp.h | 22 +++++
kernel/seccomp.c | 173 ++++++++++++++++++++++++++++++++++-
3 files changed, 198 insertions(+), 1 deletion(-)
diff --git a/include/linux/seccomp.h b/include/linux/seccomp.h
index babcd6c02d09..881c90b6aa25 100644
--- a/include/linux/seccomp.h
+++ b/include/linux/seccomp.h
@@ -10,6 +10,10 @@
SECCOMP_FILTER_FLAG_NEW_LISTENER | \
SECCOMP_FILTER_FLAG_TSYNC_ESRCH)
+/* sizeof() the first published struct seccomp_notif_addfd */
+#define SECCOMP_NOTIFY_ADDFD_SIZE_VER0 24
+#define SECCOMP_NOTIFY_ADDFD_SIZE_LATEST SECCOMP_NOTIFY_ADDFD_SIZE_VER0
+
#ifdef CONFIG_SECCOMP
#include <linux/thread_info.h>
diff --git a/include/uapi/linux/seccomp.h b/include/uapi/linux/seccomp.h
index 965290f7dcc2..6ba18b82a02e 100644
--- a/include/uapi/linux/seccomp.h
+++ b/include/uapi/linux/seccomp.h
@@ -113,6 +113,25 @@ struct seccomp_notif_resp {
__u32 flags;
};
+/* valid flags for seccomp_notif_addfd */
+#define SECCOMP_ADDFD_FLAG_SETFD (1UL << 0) /* Specify remote fd */
+
+/**
+ * struct seccomp_notif_addfd
+ * @id: The ID of the seccomp notification
+ * @flags: SECCOMP_ADDFD_FLAG_*
+ * @srcfd: The local fd number
+ * @newfd: Optional remote FD number if SETFD option is set, otherwise 0.
+ * @newfd_flags: The O_* flags the remote FD should have applied
+ */
+struct seccomp_notif_addfd {
+ __u64 id;
+ __u32 flags;
+ __u32 srcfd;
+ __u32 newfd;
+ __u32 newfd_flags;
+};
+
#define SECCOMP_IOC_MAGIC '!'
#define SECCOMP_IO(nr) _IO(SECCOMP_IOC_MAGIC, nr)
#define SECCOMP_IOR(nr, type) _IOR(SECCOMP_IOC_MAGIC, nr, type)
@@ -124,5 +143,8 @@ struct seccomp_notif_resp {
#define SECCOMP_IOCTL_NOTIF_SEND SECCOMP_IOWR(1, \
struct seccomp_notif_resp)
#define SECCOMP_IOCTL_NOTIF_ID_VALID SECCOMP_IOW(2, __u64)
+/* On success, the return value is the remote process's added fd number */
+#define SECCOMP_IOCTL_NOTIF_ADDFD SECCOMP_IOW(3, \
+ struct seccomp_notif_addfd)
#endif /* _UAPI_LINUX_SECCOMP_H */
diff --git a/kernel/seccomp.c b/kernel/seccomp.c
index 866a432cd746..b2725a2a9f3e 100644
--- a/kernel/seccomp.c
+++ b/kernel/seccomp.c
@@ -87,10 +87,42 @@ struct seccomp_knotif {
long val;
u32 flags;
- /* Signals when this has entered SECCOMP_NOTIFY_REPLIED */
+ /*
+ * Signals when this has changed states, such as the listener
+ * dying, a new seccomp addfd message, or changing to REPLIED
+ */
struct completion ready;
struct list_head list;
+
+ /* outstanding addfd requests */
+ struct list_head addfd;
+};
+
+/**
+ * struct seccomp_kaddfd - container for seccomp_addfd ioctl messages
+ *
+ * @file: A reference to the file to install in the other task
+ * @fd: The fd number to install it at. If the fd number is -1, it means the
+ * installing process should allocate the fd as normal.
+ * @flags: The flags for the new file descriptor. At the moment, only O_CLOEXEC
+ * is allowed.
+ * @ret: The return value of the installing process. It is set to the fd num
+ * upon success (>= 0).
+ * @completion: Indicates that the installing process has completed fd
+ * installation, or gone away (either due to successful
+ * reply, or signal)
+ *
+ */
+struct seccomp_kaddfd {
+ struct file *file;
+ int fd;
+ unsigned int flags;
+
+ /* To only be set on reply */
+ int ret;
+ struct completion completion;
+ struct list_head list;
};
/**
@@ -793,6 +825,17 @@ static u64 seccomp_next_notify_id(struct seccomp_filter *filter)
return filter->notif->next_id++;
}
+static void seccomp_handle_addfd(struct seccomp_kaddfd *addfd)
+{
+ /*
+ * Remove the notification, and reset the list pointers, indicating
+ * that it has been handled.
+ */
+ list_del_init(&addfd->list);
+ addfd->ret = receive_fd_replace(addfd->fd, addfd->file, addfd->flags);
+ complete(&addfd->completion);
+}
+
static int seccomp_do_user_notification(int this_syscall,
struct seccomp_filter *match,
const struct seccomp_data *sd)
@@ -801,6 +844,7 @@ static int seccomp_do_user_notification(int this_syscall,
u32 flags = 0;
long ret = 0;
struct seccomp_knotif n = {};
+ struct seccomp_kaddfd *addfd, *tmp;
mutex_lock(&match->notify_lock);
err = -ENOSYS;
@@ -813,6 +857,7 @@ static int seccomp_do_user_notification(int this_syscall,
n.id = seccomp_next_notify_id(match);
init_completion(&n.ready);
list_add(&n.list, &match->notif->notifications);
+ INIT_LIST_HEAD(&n.addfd);
up(&match->notif->request);
wake_up_poll(&match->wqh, EPOLLIN | EPOLLRDNORM);
@@ -821,14 +866,31 @@ static int seccomp_do_user_notification(int this_syscall,
/*
* This is where we wait for a reply from userspace.
*/
+wait:
err = wait_for_completion_interruptible(&n.ready);
mutex_lock(&match->notify_lock);
if (err == 0) {
+ /* Check if we were woken up by a addfd message */
+ addfd = list_first_entry_or_null(&n.addfd,
+ struct seccomp_kaddfd, list);
+ if (addfd && n.state != SECCOMP_NOTIFY_REPLIED) {
+ seccomp_handle_addfd(addfd);
+ mutex_unlock(&match->notify_lock);
+ goto wait;
+ }
ret = n.val;
err = n.error;
flags = n.flags;
}
+ /* If there were any pending addfd calls, clear them out */
+ list_for_each_entry_safe(addfd, tmp, &n.addfd, list) {
+ /* The process went away before we got a chance to handle it */
+ addfd->ret = -ESRCH;
+ list_del_init(&addfd->list);
+ complete(&addfd->completion);
+ }
+
/*
* Note that it's possible the listener died in between the time when
* we were notified of a respons (or a signal) and when we were able to
@@ -1069,6 +1131,11 @@ static int seccomp_notify_release(struct inode *inode, struct file *file)
knotif->error = -ENOSYS;
knotif->val = 0;
+ /*
+ * We do not need to wake up any pending addfd messages, as
+ * the notifier will do that for us, as this just looks
+ * like a standard reply.
+ */
complete(&knotif->ready);
}
@@ -1233,12 +1300,109 @@ static long seccomp_notify_id_valid(struct seccomp_filter *filter,
return ret;
}
+static long seccomp_notify_addfd(struct seccomp_filter *filter,
+ struct seccomp_notif_addfd __user *uaddfd,
+ unsigned int size)
+{
+ struct seccomp_notif_addfd addfd;
+ struct seccomp_knotif *knotif;
+ struct seccomp_kaddfd kaddfd;
+ int ret;
+
+ BUILD_BUG_ON(sizeof(addfd) < SECCOMP_NOTIFY_ADDFD_SIZE_VER0);
+ BUILD_BUG_ON(sizeof(addfd) != SECCOMP_NOTIFY_ADDFD_SIZE_LATEST);
+
+ if (size < SECCOMP_NOTIFY_ADDFD_SIZE_VER0 || size >= PAGE_SIZE)
+ return -EINVAL;
+
+ ret = copy_struct_from_user(&addfd, sizeof(addfd), uaddfd, size);
+ if (ret)
+ return ret;
+
+ if (addfd.newfd_flags & ~O_CLOEXEC)
+ return -EINVAL;
+
+ if (addfd.flags & ~SECCOMP_ADDFD_FLAG_SETFD)
+ return -EINVAL;
+
+ if (addfd.newfd && !(addfd.flags & SECCOMP_ADDFD_FLAG_SETFD))
+ return -EINVAL;
+
+ kaddfd.file = fget(addfd.srcfd);
+ if (!kaddfd.file)
+ return -EBADF;
+
+ kaddfd.flags = addfd.newfd_flags;
+ kaddfd.fd = (addfd.flags & SECCOMP_ADDFD_FLAG_SETFD) ?
+ addfd.newfd : -1;
+ init_completion(&kaddfd.completion);
+
+ ret = mutex_lock_interruptible(&filter->notify_lock);
+ if (ret < 0)
+ goto out;
+
+ knotif = find_notification(filter, addfd.id);
+ if (!knotif) {
+ ret = -ENOENT;
+ goto out_unlock;
+ }
+
+ /*
+ * We do not want to allow for FD injection to occur before the
+ * notification has been picked up by a userspace handler, or after
+ * the notification has been replied to.
+ */
+ if (knotif->state != SECCOMP_NOTIFY_SENT) {
+ ret = -EINPROGRESS;
+ goto out_unlock;
+ }
+
+ list_add(&kaddfd.list, &knotif->addfd);
+ complete(&knotif->ready);
+ mutex_unlock(&filter->notify_lock);
+
+ /* Now we wait for it to be processed or be interrupted */
+ ret = wait_for_completion_interruptible(&kaddfd.completion);
+ if (ret == 0) {
+ /*
+ * We had a successful completion. The other side has already
+ * removed us from the addfd queue, and
+ * wait_for_completion_interruptible has a memory barrier upon
+ * success that lets us read this value directly without
+ * locking.
+ */
+ ret = kaddfd.ret;
+ goto out;
+ }
+
+ mutex_lock(&filter->notify_lock);
+ /*
+ * Even though we were woken up by a signal and not a successful
+ * completion, a completion may have happened in the mean time.
+ *
+ * We need to check again if the addfd request has been handled,
+ * and if not, we will remove it from the queue.
+ */
+ if (list_empty(&kaddfd.list))
+ ret = kaddfd.ret;
+ else
+ list_del(&kaddfd.list);
+
+out_unlock:
+ mutex_unlock(&filter->notify_lock);
+out:
+ fput(kaddfd.file);
+
+ return ret;
+}
+
static long seccomp_notify_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
struct seccomp_filter *filter = file->private_data;
void __user *buf = (void __user *)arg;
+ /* Fixed-size ioctls */
switch (cmd) {
case SECCOMP_IOCTL_NOTIF_RECV:
return seccomp_notify_recv(filter, buf);
@@ -1247,6 +1411,13 @@ static long seccomp_notify_ioctl(struct file *file, unsigned int cmd,
case SECCOMP_IOCTL_NOTIF_ID_VALID_WRONG_DIR:
case SECCOMP_IOCTL_NOTIF_ID_VALID:
return seccomp_notify_id_valid(filter, buf);
+ }
+
+ /* Extensible Argument ioctls */
+#define EA_IOCTL(cmd) ((cmd) & ~(IOC_INOUT | IOCSIZE_MASK))
+ switch (EA_IOCTL(cmd)) {
+ case EA_IOCTL(SECCOMP_IOCTL_NOTIF_ADDFD):
+ return seccomp_notify_addfd(filter, buf, _IOC_SIZE(cmd));
default:
return -EINVAL;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread* Re: [PATCH v7 8/9] seccomp: Introduce addfd ioctl to seccomp user notifier
2020-07-09 18:26 ` [PATCH v7 8/9] seccomp: Introduce addfd ioctl to seccomp user notifier Kees Cook
@ 2020-07-14 18:20 ` Will Drewry
0 siblings, 0 replies; 19+ messages in thread
From: Will Drewry @ 2020-07-14 18:20 UTC (permalink / raw)
To: Kees Cook
Cc: LKML, Sargun Dhillon, Matt Denton, Christian Brauner,
Tycho Andersen, David Laight, Christoph Hellwig, David S. Miller,
Jakub Kicinski, Alexander Viro, Aleksa Sarai, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Shuah Khan, netdev,
containers, linux-api, linux-fsdevel, linux-kselftest
On Thu, Jul 9, 2020 at 1:26 PM Kees Cook <keescook@chromium.org> wrote:
>
> From: Sargun Dhillon <sargun@sargun.me>
>
> The current SECCOMP_RET_USER_NOTIF API allows for syscall supervision over
> an fd. It is often used in settings where a supervising task emulates
> syscalls on behalf of a supervised task in userspace, either to further
> restrict the supervisee's syscall abilities or to circumvent kernel
> enforced restrictions the supervisor deems safe to lift (e.g. actually
> performing a mount(2) for an unprivileged container).
>
> While SECCOMP_RET_USER_NOTIF allows for the interception of any syscall,
> only a certain subset of syscalls could be correctly emulated. Over the
> last few development cycles, the set of syscalls which can't be emulated
> has been reduced due to the addition of pidfd_getfd(2). With this we are
> now able to, for example, intercept syscalls that require the supervisor
> to operate on file descriptors of the supervisee such as connect(2).
>
> However, syscalls that cause new file descriptors to be installed can not
> currently be correctly emulated since there is no way for the supervisor
> to inject file descriptors into the supervisee. This patch adds a
> new addfd ioctl to remove this restriction by allowing the supervisor to
> install file descriptors into the intercepted task. By implementing this
> feature via seccomp the supervisor effectively instructs the supervisee
> to install a set of file descriptors into its own file descriptor table
> during the intercepted syscall. This way it is possible to intercept
> syscalls such as open() or accept(), and install (or replace, like
> dup2(2)) the supervisor's resulting fd into the supervisee. One
> replacement use-case would be to redirect the stdout and stderr of a
> supervisee into log file descriptors opened by the supervisor.
>
> The ioctl handling is based on the discussions[1] of how Extensible
> Arguments should interact with ioctls. Instead of building size into
> the addfd structure, make it a function of the ioctl command (which
> is how sizes are normally passed to ioctls). To support forward and
> backward compatibility, just mask out the direction and size, and match
> everything. The size (and any future direction) checks are done along
> with copy_struct_from_user() logic.
>
> As a note, the seccomp_notif_addfd structure is laid out based on 8-byte
> alignment without requiring packing as there have been packing issues
> with uapi highlighted before[2][3]. Although we could overload the
> newfd field and use -1 to indicate that it is not to be used, doing
> so requires changing the size of the fd field, and introduces struct
> packing complexity.
>
> [1]: https://lore.kernel.org/lkml/87o8w9bcaf.fsf@mid.deneb.enyo.de/
> [2]: https://lore.kernel.org/lkml/a328b91d-fd8f-4f27-b3c2-91a9c45f18c0@rasmusvillemoes.dk/
> [3]: https://lore.kernel.org/lkml/20200612104629.GA15814@ircssh-2.c.rugged-nimbus-611.internal
>
> Suggested-by: Matt Denton <mpdenton@google.com>
> Link: https://lore.kernel.org/r/20200603011044.7972-4-sargun@sargun.me
> Signed-off-by: Sargun Dhillon <sargun@sargun.me>
> Co-developed-by: Kees Cook <keescook@chromium.org>
> Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Will Drewry <wad@chromium.org>
^ permalink raw reply [flat|nested] 19+ messages in thread
* [PATCH v7 9/9] selftests/seccomp: Test SECCOMP_IOCTL_NOTIF_ADDFD
2020-07-09 18:26 [PATCH v7 0/9] Add seccomp notifier ioctl that enables adding fds Kees Cook
` (7 preceding siblings ...)
2020-07-09 18:26 ` [PATCH v7 8/9] seccomp: Introduce addfd ioctl to seccomp user notifier Kees Cook
@ 2020-07-09 18:26 ` Kees Cook
8 siblings, 0 replies; 19+ messages in thread
From: Kees Cook @ 2020-07-09 18:26 UTC (permalink / raw)
To: linux-kernel
Cc: Kees Cook, Sargun Dhillon, Christian Brauner, Tycho Andersen,
David Laight, Christoph Hellwig, David S. Miller, Jakub Kicinski,
Alexander Viro, Aleksa Sarai, Matt Denton, Jann Horn,
Chris Palmer, Robert Sesek, Giuseppe Scrivano,
Greg Kroah-Hartman, Andy Lutomirski, Will Drewry, Shuah Khan,
netdev, containers, linux-api, linux-fsdevel, linux-kselftest
From: Sargun Dhillon <sargun@sargun.me>
Test whether we can add file descriptors in response to notifications.
This injects the file descriptors via notifications, and then uses kcmp
to determine whether or not it has been successful.
It also includes some basic sanity checking for arguments.
Signed-off-by: Sargun Dhillon <sargun@sargun.me>
Link: https://lore.kernel.org/r/20200603011044.7972-5-sargun@sargun.me
Co-developed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
---
tools/testing/selftests/seccomp/seccomp_bpf.c | 229 ++++++++++++++++++
1 file changed, 229 insertions(+)
diff --git a/tools/testing/selftests/seccomp/seccomp_bpf.c b/tools/testing/selftests/seccomp/seccomp_bpf.c
index 4662a25bc9e8..3f41b32b9165 100644
--- a/tools/testing/selftests/seccomp/seccomp_bpf.c
+++ b/tools/testing/selftests/seccomp/seccomp_bpf.c
@@ -45,6 +45,7 @@
#include <sys/socket.h>
#include <sys/ioctl.h>
#include <linux/kcmp.h>
+#include <sys/resource.h>
#include <unistd.h>
#include <sys/syscall.h>
@@ -168,7 +169,9 @@ struct seccomp_metadata {
#ifndef SECCOMP_FILTER_FLAG_NEW_LISTENER
#define SECCOMP_FILTER_FLAG_NEW_LISTENER (1UL << 3)
+#endif
+#ifndef SECCOMP_RET_USER_NOTIF
#define SECCOMP_RET_USER_NOTIF 0x7fc00000U
#define SECCOMP_IOC_MAGIC '!'
@@ -204,6 +207,39 @@ struct seccomp_notif_sizes {
};
#endif
+#ifndef SECCOMP_IOCTL_NOTIF_ADDFD
+/* On success, the return value is the remote process's added fd number */
+#define SECCOMP_IOCTL_NOTIF_ADDFD SECCOMP_IOW(3, \
+ struct seccomp_notif_addfd)
+
+/* valid flags for seccomp_notif_addfd */
+#define SECCOMP_ADDFD_FLAG_SETFD (1UL << 0) /* Specify remote fd */
+
+struct seccomp_notif_addfd {
+ __u64 id;
+ __u32 flags;
+ __u32 srcfd;
+ __u32 newfd;
+ __u32 newfd_flags;
+};
+#endif
+
+struct seccomp_notif_addfd_small {
+ __u64 id;
+ char weird[4];
+};
+#define SECCOMP_IOCTL_NOTIF_ADDFD_SMALL \
+ SECCOMP_IOW(3, struct seccomp_notif_addfd_small)
+
+struct seccomp_notif_addfd_big {
+ union {
+ struct seccomp_notif_addfd addfd;
+ char buf[sizeof(struct seccomp_notif_addfd) + 8];
+ };
+};
+#define SECCOMP_IOCTL_NOTIF_ADDFD_BIG \
+ SECCOMP_IOWR(3, struct seccomp_notif_addfd_big)
+
#ifndef PTRACE_EVENTMSG_SYSCALL_ENTRY
#define PTRACE_EVENTMSG_SYSCALL_ENTRY 1
#define PTRACE_EVENTMSG_SYSCALL_EXIT 2
@@ -3740,6 +3776,199 @@ TEST(user_notification_filter_empty_threaded)
EXPECT_GT((pollfd.revents & POLLHUP) ?: 0, 0);
}
+TEST(user_notification_addfd)
+{
+ pid_t pid;
+ long ret;
+ int status, listener, memfd, fd;
+ struct seccomp_notif_addfd addfd = {};
+ struct seccomp_notif_addfd_small small = {};
+ struct seccomp_notif_addfd_big big = {};
+ struct seccomp_notif req = {};
+ struct seccomp_notif_resp resp = {};
+ /* 100 ms */
+ struct timespec delay = { .tv_nsec = 100000000 };
+
+ memfd = memfd_create("test", 0);
+ ASSERT_GE(memfd, 0);
+
+ ret = prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0);
+ ASSERT_EQ(0, ret) {
+ TH_LOG("Kernel does not support PR_SET_NO_NEW_PRIVS!");
+ }
+
+ /* Check that the basic notification machinery works */
+ listener = user_notif_syscall(__NR_getppid,
+ SECCOMP_FILTER_FLAG_NEW_LISTENER);
+ ASSERT_GE(listener, 0);
+
+ pid = fork();
+ ASSERT_GE(pid, 0);
+
+ if (pid == 0) {
+ if (syscall(__NR_getppid) != USER_NOTIF_MAGIC)
+ exit(1);
+ exit(syscall(__NR_getppid) != USER_NOTIF_MAGIC);
+ }
+
+ ASSERT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_RECV, &req), 0);
+
+ addfd.srcfd = memfd;
+ addfd.newfd = 0;
+ addfd.id = req.id;
+ addfd.flags = 0x0;
+
+ /* Verify bad newfd_flags cannot be set */
+ addfd.newfd_flags = ~O_CLOEXEC;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd), -1);
+ EXPECT_EQ(errno, EINVAL);
+ addfd.newfd_flags = O_CLOEXEC;
+
+ /* Verify bad flags cannot be set */
+ addfd.flags = 0xff;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd), -1);
+ EXPECT_EQ(errno, EINVAL);
+ addfd.flags = 0;
+
+ /* Verify that remote_fd cannot be set without setting flags */
+ addfd.newfd = 1;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd), -1);
+ EXPECT_EQ(errno, EINVAL);
+ addfd.newfd = 0;
+
+ /* Verify small size cannot be set */
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD_SMALL, &small), -1);
+ EXPECT_EQ(errno, EINVAL);
+
+ /* Verify we can't send bits filled in unknown buffer area */
+ memset(&big, 0xAA, sizeof(big));
+ big.addfd = addfd;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD_BIG, &big), -1);
+ EXPECT_EQ(errno, E2BIG);
+
+
+ /* Verify we can set an arbitrary remote fd */
+ fd = ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd);
+ /*
+ * The child has fds 0(stdin), 1(stdout), 2(stderr), 3(memfd),
+ * 4(listener), so the newly allocated fd should be 5.
+ */
+ EXPECT_EQ(fd, 5);
+ EXPECT_EQ(filecmp(getpid(), pid, memfd, fd), 0);
+
+ /* Verify we can set an arbitrary remote fd with large size */
+ memset(&big, 0x0, sizeof(big));
+ big.addfd = addfd;
+ fd = ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD_BIG, &big);
+ EXPECT_EQ(fd, 6);
+
+ /* Verify we can set a specific remote fd */
+ addfd.newfd = 42;
+ addfd.flags = SECCOMP_ADDFD_FLAG_SETFD;
+ fd = ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd);
+ EXPECT_EQ(fd, 42);
+ EXPECT_EQ(filecmp(getpid(), pid, memfd, fd), 0);
+
+ /* Resume syscall */
+ resp.id = req.id;
+ resp.error = 0;
+ resp.val = USER_NOTIF_MAGIC;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_SEND, &resp), 0);
+
+ /*
+ * This sets the ID of the ADD FD to the last request plus 1. The
+ * notification ID increments 1 per notification.
+ */
+ addfd.id = req.id + 1;
+
+ /* This spins until the underlying notification is generated */
+ while (ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd) != -1 &&
+ errno != -EINPROGRESS)
+ nanosleep(&delay, NULL);
+
+ memset(&req, 0, sizeof(req));
+ ASSERT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_RECV, &req), 0);
+ ASSERT_EQ(addfd.id, req.id);
+
+ resp.id = req.id;
+ resp.error = 0;
+ resp.val = USER_NOTIF_MAGIC;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_SEND, &resp), 0);
+
+ /* Wait for child to finish. */
+ EXPECT_EQ(waitpid(pid, &status, 0), pid);
+ EXPECT_EQ(true, WIFEXITED(status));
+ EXPECT_EQ(0, WEXITSTATUS(status));
+
+ close(memfd);
+}
+
+TEST(user_notification_addfd_rlimit)
+{
+ pid_t pid;
+ long ret;
+ int status, listener, memfd;
+ struct seccomp_notif_addfd addfd = {};
+ struct seccomp_notif req = {};
+ struct seccomp_notif_resp resp = {};
+ const struct rlimit lim = {
+ .rlim_cur = 0,
+ .rlim_max = 0,
+ };
+
+ memfd = memfd_create("test", 0);
+ ASSERT_GE(memfd, 0);
+
+ ret = prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0);
+ ASSERT_EQ(0, ret) {
+ TH_LOG("Kernel does not support PR_SET_NO_NEW_PRIVS!");
+ }
+
+ /* Check that the basic notification machinery works */
+ listener = user_notif_syscall(__NR_getppid,
+ SECCOMP_FILTER_FLAG_NEW_LISTENER);
+ ASSERT_GE(listener, 0);
+
+ pid = fork();
+ ASSERT_GE(pid, 0);
+
+ if (pid == 0)
+ exit(syscall(__NR_getppid) != USER_NOTIF_MAGIC);
+
+
+ ASSERT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_RECV, &req), 0);
+
+ ASSERT_EQ(prlimit(pid, RLIMIT_NOFILE, &lim, NULL), 0);
+
+ addfd.srcfd = memfd;
+ addfd.newfd_flags = O_CLOEXEC;
+ addfd.newfd = 0;
+ addfd.id = req.id;
+ addfd.flags = 0;
+
+ /* Should probably spot check /proc/sys/fs/file-nr */
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd), -1);
+ EXPECT_EQ(errno, EMFILE);
+
+ addfd.newfd = 100;
+ addfd.flags = SECCOMP_ADDFD_FLAG_SETFD;
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd), -1);
+ EXPECT_EQ(errno, EBADF);
+
+ resp.id = req.id;
+ resp.error = 0;
+ resp.val = USER_NOTIF_MAGIC;
+
+ EXPECT_EQ(ioctl(listener, SECCOMP_IOCTL_NOTIF_SEND, &resp), 0);
+
+ /* Wait for child to finish. */
+ EXPECT_EQ(waitpid(pid, &status, 0), pid);
+ EXPECT_EQ(true, WIFEXITED(status));
+ EXPECT_EQ(0, WEXITSTATUS(status));
+
+ close(memfd);
+}
+
/*
* TODO:
* - expand NNP testing
--
2.25.1
^ permalink raw reply related [flat|nested] 19+ messages in thread