Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1551234

[PATCH 4.9 30/83] fs: exec: apply CLOEXEC before changing dumpable task flags

Show all headers | View raw

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Aleksa Sarai <asarai@suse.de>

commit 613cc2b6f272c1a8ad33aefa21cad77af23139f7 upstream.

If you have a process that has set itself to be non-dumpable, and it
then undergoes exec(2), any CLOEXEC file descriptors it has open are
"exposed" during a race window between the dumpable flags of the process
being reset for exec(2) and CLOEXEC being applied to the file
descriptors. This can be exploited by a process by attempting to access
/proc/<pid>/fd/... during this window, without requiring CAP_SYS_PTRACE.

The race in question is after set_dumpable has been (for get_link,
though the trace is basically the same for readlink):

[vfs]
-> proc_pid_link_inode_operations.get_link
   -> proc_pid_get_link
      -> proc_fd_access_allowed
         -> ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);

Which will return 0, during the race window and CLOEXEC file descriptors
will still be open during this window because do_close_on_exec has not
been called yet. As a result, the ordering of these calls should be
reversed to avoid this race window.

This is of particular concern to container runtimes, where joining a
PID namespace with file descriptors referring to the host filesystem
can result in security issues (since PRCTL_SET_DUMPABLE doesn't protect
against access of CLOEXEC file descriptors -- file descriptors which may
reference filesystem objects the container shouldn't have access to).

Cc: dev@opencontainers.org
Reported-by: Michael Crosby <crosbymichael@gmail.com>
Signed-off-by: Aleksa Sarai <asarai@suse.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 fs/exec.c |   10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

--- a/fs/exec.c
+++ b/fs/exec.c
@@ -19,7 +19,7 @@
  * current->executable is only used by the procfs.  This allows a dispatch
  * table to check for several different types  of binary formats.  We keep
  * trying until we recognize the file or we run out of supported binary
- * formats. 
+ * formats.
  */
 
 #include <linux/slab.h>
@@ -1266,6 +1266,13 @@ int flush_old_exec(struct linux_binprm *
 	flush_thread();
 	current->personality &= ~bprm->per_clear;
 
+	/*
+	 * We have to apply CLOEXEC before we change whether the process is
+	 * dumpable (in setup_new_exec) to avoid a race with a process in userspace
+	 * trying to access the should-be-closed file descriptors of a process
+	 * undergoing exec(2).
+	 */
+	do_close_on_exec(current->files);
 	return 0;
 
 out:
@@ -1328,7 +1335,6 @@ void setup_new_exec(struct linux_binprm
 	   group */
 	current->self_exec_id++;
 	flush_signal_handlers(current, 0);
-	do_close_on_exec(current->files);
 }
 EXPORT_SYMBOL(setup_new_exec);
 

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 4.9 00/83] 4.9.1-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:20 +0100
  [PATCH 4.9 68/83] CIFS: Fix a possible memory corruption during reconnect Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 26/83] ALSA: hda - Gate the mic jack on HP Z1 Gen3 AiO Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 33/83] mm: Add a user_ns owner to mm_struct and fix ptrace permission checks Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 38/83] ext4: dont lock buffer in ext4_commit_super if holding spinlock Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 39/83] ext4: fix mballoc breakage with 64k block size Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 17/83] usb: gadget: f_uac2: fix error handling at afunc_bind Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 41/83] ext4: use more strict checks for inodes_per_block on mount Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 36/83] ptrace: Dont allow accessing an undumpable mm Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 56/83] PM / OPP: Pass opp_table to dev_pm_opp_put_regulator() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 35/83] ptrace: Capture the ptracers creds not PT_PTRACE_CAP Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 42/83] ext4: fix in-superblock mount options processing Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 40/83] ext4: fix stack memory corruption with 64k block size Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 25/83] ALSA: hda - fix headset-mic problem on a Dell laptop Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 64/83] dm raid: fix discard support regression Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 31/83] splice: reinstate SIGPIPE/EPIPE handling Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 59/83] dm table: fix all_blk_mq inconsistency when an empty table is loaded Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 34/83] vfs,mm: fix return value of read() at s_maxbytes Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 54/83] mm, page_alloc: keep pcp count and list contents in sync if struct page is corrupted Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 32/83] block_dev: dont test bdev->bd_contains when it is not stable Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 13/83] USB: cdc-acm: add device id for GW Instek AFG-125 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 12/83] USB: serial: kl5kusb105: fix open error path Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 30/83] fs: exec: apply CLOEXEC before changing dumpable task flags Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 29/83] exec: Ensure mm->user_ns contains the execed files Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 62/83] dm crypt: mark key as invalid until properly loaded Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 28/83] clk: ti: omap36xx: Work around sprz319 advisory 2.1 Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 66/83] ASoC: intel: Fix crash at suspend/resume without card registration Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 43/83] ext4: add sanity checking to count_overhead() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 55/83] usb: gadget: composite: always set ep->mult to a sensible value Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  [PATCH 4.9 58/83] blk-mq: Do not invoke .queue_rq() for a stopped queue Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-04 21:30 +0100
  Re: [PATCH 4.9 00/83] 4.9.1-stable review Shuah Khan <shuah.kh@samsung.com> - 2017-01-05 01:50 +0100
    Re: [PATCH 4.9 00/83] 4.9.1-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-05 09:00 +0100
  Re: [PATCH 4.9 00/83] 4.9.1-stable review Guenter Roeck <linux@roeck-us.net> - 2017-01-05 06:00 +0100
    Re: [PATCH 4.9 00/83] 4.9.1-stable review Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-05 09:00 +0100

csiph-web