linux-stable.git/include, branch v3.16.83

futex: Fix inode life-time issue

2020-04-28T18:03:51+00:00

commit 8019ad13ef7f64be44d4f892af9c840179009254 upstream.

As reported by Jann, ihold() does not in fact guarantee inode
persistence. And instead of making it so, replace the usage of inode
pointers with a per boot, machine wide, unique inode identifier.

This sequence number is global, but shared (file backed) futexes are
rare enough that this should not become a performance issue.

Reported-by: Jann Horn 
Suggested-by: Linus Torvalds 
Signed-off-by: Peter Zijlstra (Intel) 
Signed-off-by: Greg Kroah-Hartman 
[bwh: Backported to 3.16: Use atomic64_cmpxchg() instead of the
 _relaxed() variant which we don't have]
Signed-off-by: Ben Hutchings

media: fix media devnode ioctl/syscall and unregister race

2020-04-28T18:03:51+00:00

commit 6f0dd24a084a17f9984dd49dffbf7055bf123993 upstream.

Media devnode open/ioctl could be in progress when media device unregister
is initiated. System calls and ioctls check media device registered status
at the beginning, however, there is a window where unregister could be in
progress without changing the media devnode status to unregistered.

process 1				process 2
fd = open(/dev/media0)
media_devnode_is_registered()
	(returns true here)

					media_device_unregister()
						(unregister is in progress
						and devnode isn't
						unregistered yet)
					...
ioctl(fd, ...)
__media_ioctl()
media_devnode_is_registered()
	(returns true here)
					...
					media_devnode_unregister()
					...
					(driver releases the media device
					memory)

media_device_ioctl()
	(By this point
	devnode->media_dev does not
	point to allocated memory.
	use-after free in in mutex_lock_nested)

BUG: KASAN: use-after-free in mutex_lock_nested+0x79c/0x800 at addr
ffff8801ebe914f0

Fix it by clearing register bit when unregister starts to avoid the race.

process 1                               process 2
fd = open(/dev/media0)
media_devnode_is_registered()
        (could return true here)

                                        media_device_unregister()
                                                (clear the register bit,
						 then start unregister.)
                                        ...
ioctl(fd, ...)
__media_ioctl()
media_devnode_is_registered()
        (return false here, ioctl
	 returns I/O error, and
	 will not access media
	 device memory)
                                        ...
                                        media_devnode_unregister()
                                        ...
                                        (driver releases the media device
					 memory)

Signed-off-by: Shuah Khan 
Suggested-by: Sakari Ailus 
Reported-by: Mauro Carvalho Chehab 
Tested-by: Mauro Carvalho Chehab 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.16: adjut filename, context]
Signed-off-by: Ben Hutchings

media-device: dynamically allocate struct media_devnode

2020-04-28T18:03:50+00:00

commit a087ce704b802becbb4b0f2a20f2cb3f6911802e upstream.

struct media_devnode is currently embedded at struct media_device.

While this works fine during normal usage, it leads to a race
condition during devnode unregister. the problem is that drivers
assume that, after calling media_device_unregister(), the struct
that contains media_device can be freed. This is not true, as it
can't be freed until userspace closes all opened /dev/media devnodes.

In other words, if the media devnode is still open, and media_device
gets freed, any call to an ioctl will make the core to try to access
struct media_device, with will cause an use-after-free and even GPF.

Fix this by dynamically allocating the struct media_devnode and only
freeing it when it is safe.

Signed-off-by: Mauro Carvalho Chehab 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.16:
 - Drop change in au0828
 - Include  in media-device.c
 - Adjust context]
Signed-off-by: Ben Hutchings

media-devnode: fix namespace mess

2020-04-28T18:03:49+00:00

commit 163f1e93e995048b894c5fc86a6034d16beed740 upstream.

Along all media controller code, "mdev" is used to represent
a pointer to struct media_device, and "devnode" for a pointer
to struct media_devnode.

However, inside media-devnode.[ch], "mdev" is used to represent
a pointer to struct media_devnode.

This is very confusing and may lead to development errors.

So, let's change all occurrences at media-devnode.[ch] to
also use "devnode" for such pointers.

This patch doesn't make any functional changes.

Signed-off-by: Mauro Carvalho Chehab 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.16: adjust filename, context]
Signed-off-by: Ben Hutchings

ptp: fix the race between the release of ptp_clock and cdev

2020-04-28T18:03:47+00:00

commit a33121e5487b424339636b25c35d3a180eaa5f5e upstream.

In a case when a ptp chardev (like /dev/ptp0) is open but an underlying
device is removed, closing this file leads to a race. This reproduces
easily in a kvm virtual machine:

ts# cat openptp0.c
int main() { ... fp = fopen("/dev/ptp0", "r"); ... sleep(10); }
ts# uname -r
5.5.0-rc3-46cf053e
ts# cat /proc/cmdline
... slub_debug=FZP
ts# modprobe ptp_kvm
ts# ./openptp0 &
[1] 670
opened /dev/ptp0, sleeping 10s...
ts# rmmod ptp_kvm
ts# ls /dev/ptp*
ls: cannot access '/dev/ptp*': No such file or directory
ts# ...woken up
[   48.010809] general protection fault: 0000 [#1] SMP
[   48.012502] CPU: 6 PID: 658 Comm: openptp0 Not tainted 5.5.0-rc3-46cf053e #25
[   48.014624] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), ...
[   48.016270] RIP: 0010:module_put.part.0+0x7/0x80
[   48.017939] RSP: 0018:ffffb3850073be00 EFLAGS: 00010202
[   48.018339] RAX: 000000006b6b6b6b RBX: 6b6b6b6b6b6b6b6b RCX: ffff89a476c00ad0
[   48.018936] RDX: fffff65a08d3ea08 RSI: 0000000000000247 RDI: 6b6b6b6b6b6b6b6b
[   48.019470] ...                                              ^^^ a slub poison
[   48.023854] Call Trace:
[   48.024050]  __fput+0x21f/0x240
[   48.024288]  task_work_run+0x79/0x90
[   48.024555]  do_exit+0x2af/0xab0
[   48.024799]  ? vfs_write+0x16a/0x190
[   48.025082]  do_group_exit+0x35/0x90
[   48.025387]  __x64_sys_exit_group+0xf/0x10
[   48.025737]  do_syscall_64+0x3d/0x130
[   48.026056]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[   48.026479] RIP: 0033:0x7f53b12082f6
[   48.026792] ...
[   48.030945] Modules linked in: ptp i6300esb watchdog [last unloaded: ptp_kvm]
[   48.045001] Fixing recursive fault but reboot is needed!

This happens in:

static void __fput(struct file *file)
{   ...
    if (file->f_op->release)
        file->f_op->release(inode, file); <<< cdev is kfree'd here
    if (unlikely(S_ISCHR(inode->i_mode) && inode->i_cdev != NULL &&
             !(mode & FMODE_PATH))) {
        cdev_put(inode->i_cdev); <<< cdev fields are accessed here

Namely:

__fput()
  posix_clock_release()
    kref_put(&clk->kref, delete_clock) <<< the last reference
      delete_clock()
        delete_ptp_clock()
          kfree(ptp) <<< cdev is embedded in ptp
  cdev_put
    module_put(p->owner) <<< *p is kfree'd, bang!

Here cdev is embedded in posix_clock which is embedded in ptp_clock.
The race happens because ptp_clock's lifetime is controlled by two
refcounts: kref and cdev.kobj in posix_clock. This is wrong.

Make ptp_clock's sysfs device a parent of cdev with cdev_device_add()
created especially for such cases. This way the parent device with its
ptp_clock is not released until all references to the cdev are released.
This adds a requirement that an initialized but not exposed struct
device should be provided to posix_clock_register() by a caller instead
of a simple dev_t.

This approach was adopted from the commit 72139dfa2464 ("watchdog: Fix
the race between the release of watchdog_core_data and cdev"). See
details of the implementation in the commit 233ed09d7fda ("chardev: add
helper function to register char devs with a struct device").

Link: https://lore.kernel.org/linux-fsdevel/20191125125342.6189-1-vdronov@redhat.com/T/#u
Analyzed-by: Stephen Johnston 
Analyzed-by: Vern Lovejoy 
Signed-off-by: Vladis Dronov 
Acked-by: Richard Cochran 
Signed-off-by: David S. Miller 
Signed-off-by: Ben Hutchings

chardev: add helper function to register char devs with a struct device

2020-04-28T18:03:46+00:00

commit 233ed09d7fdacf592ee91e6c97ce5f4364fbe7c0 upstream.

Credit for this patch goes is shared with Dan Williams [1]. I've
taken things one step further to make the helper function more
useful and clean up calling code.

There's a common pattern in the kernel whereby a struct cdev is placed
in a structure along side a struct device which manages the life-cycle
of both. In the naive approach, the reference counting is broken and
the struct device can free everything before the chardev code
is entirely released.

Many developers have solved this problem by linking the internal kobjs
in this fashion:

cdev.kobj.parent = &parent_dev.kobj;

The cdev code explicitly gets and puts a reference to it's kobj parent.
So this seems like it was intended to be used this way. Dmitrty Torokhov
first put this in place in 2012 with this commit:

2f0157f char_dev: pin parent kobject

and the first instance of the fix was then done in the input subsystem
in the following commit:

4a215aa Input: fix use-after-free introduced with dynamic minor changes

Subsequently over the years, however, this issue seems to have tripped
up multiple developers independently. For example, see these commits:

0d5b7da iio: Prevent race between IIO chardev opening and IIO device
(by Lars-Peter Clausen in 2013)

ba0ef85 tpm: Fix initialization of the cdev
(by Jason Gunthorpe in 2015)

5b28dde [media] media: fix use-after-free in cdev_put() when app exits
after driver unbind
(by Shauh Khan in 2016)

This technique is similarly done in at least 15 places within the kernel
and probably should have been done so in another, at least, 5 places.
The kobj line also looks very suspect in that one would not expect
drivers to have to mess with kobject internals in this way.
Even highly experienced kernel developers can be surprised by this
code, as seen in [2].

To help alleviate this situation, and hopefully prevent future
wasted effort on this problem, this patch introduces a helper function
to register a char device along with its parent struct device.
This creates a more regular API for tying a char device to its parent
without the developer having to set members in the underlying kobject.

This patch introduce cdev_device_add and cdev_device_del which
replaces a common pattern including setting the kobj parent, calling
cdev_add and then calling device_add. It also introduces cdev_set_parent
for the few cases that set the kobject parent without using device_add.

[1] https://lkml.org/lkml/2017/2/13/700
[2] https://lkml.org/lkml/2017/2/10/370

Signed-off-by: Logan Gunthorpe 
Signed-off-by: Dan Williams 
Reviewed-by: Hans Verkuil 
Reviewed-by: Alexandre Belloni 
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Ben Hutchings

net: ipv6_stub: use ip6_dst_lookup_flow instead of ip6_dst_lookup

2020-04-28T18:03:39+00:00

commit 6c8991f41546c3c472503dff1ea9daaddf9331c2 upstream.

ipv6_stub uses the ip6_dst_lookup function to allow other modules to
perform IPv6 lookups. However, this function skips the XFRM layer
entirely.

All users of ipv6_stub->ip6_dst_lookup use ip_route_output_flow (via the
ip_route_output_key and ip_route_output helpers) for their IPv4 lookups,
which calls xfrm_lookup_route(). This patch fixes this inconsistent
behavior by switching the stub to ip6_dst_lookup_flow, which also calls
xfrm_lookup_route().

This requires some changes in all the callers, as these two functions
take different arguments and have different return types.

Fixes: 5f81bd2e5d80 ("ipv6: export a stub for IPv6 symbols used by vxlan")
Reported-by: Xiumei Mu 
Signed-off-by: Sabrina Dubroca 
Signed-off-by: David S. Miller 
[bwh: Backported to 3.16:
 - Only vxlan uses this operation
 - Neither ip6_dst_lookup() nor ip6_dst_lookup_flow() takes a struct net
   pointer argument here
 - Adjust filename, context]
Signed-off-by: Ben Hutchings

blktrace: Protect q->blk_trace with RCU

2020-04-28T18:03:38+00:00

commit c780e86dd48ef6467a1146cf7d0fe1e05a635039 upstream.

KASAN is reporting that __blk_add_trace() has a use-after-free issue
when accessing q->blk_trace. Indeed the switching of block tracing (and
thus eventual freeing of q->blk_trace) is completely unsynchronized with
the currently running tracing and thus it can happen that the blk_trace
structure is being freed just while __blk_add_trace() works on it.
Protect accesses to q->blk_trace by RCU during tracing and make sure we
wait for the end of RCU grace period when shutting down tracing. Luckily
that is rare enough event that we can afford that. Note that postponing
the freeing of blk_trace to an RCU callback should better be avoided as
it could have unexpected user visible side-effects as debugfs files
would be still existing for a short while block tracing has been shut
down.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=205711
Reviewed-by: Chaitanya Kulkarni 
Reviewed-by: Ming Lei 
Tested-by: Ming Lei 
Reviewed-by: Bart Van Assche 
Reported-by: Tristan Madani 
Signed-off-by: Jan Kara 
Signed-off-by: Jens Axboe 
[bwh: Backported to 3.16:
 - Drop changes in blk_trace_note_message_enabled(), blk_trace_bio_get_cgid()
 - Adjust context]
Signed-off-by: Ben Hutchings

namei: allow restricted O_CREAT of FIFOs and regular files

2020-04-28T18:03:36+00:00

commit 30aba6656f61ed44cba445a3c0d38b296fa9e8f5 upstream.

Disallows open of FIFOs or regular files not owned by the user in world
writable sticky directories, unless the owner is the same as that of the
directory or the file is opened without the O_CREAT flag.  The purpose
is to make data spoofing attacks harder.  This protection can be turned
on and off separately for FIFOs and regular files via sysctl, just like
the symlinks/hardlinks protection.  This patch is based on Openwall's
"HARDEN_FIFO" feature by Solar Designer.

This is a brief list of old vulnerabilities that could have been prevented
by this feature, some of them even allow for privilege escalation:

CVE-2000-1134
CVE-2007-3852
CVE-2008-0525
CVE-2009-0416
CVE-2011-4834
CVE-2015-1838
CVE-2015-7442
CVE-2016-7489

This list is not meant to be complete.  It's difficult to track down all
vulnerabilities of this kind because they were often reported without any
mention of this particular attack vector.  In fact, before
hardlinks/symlinks restrictions, fifos/regular files weren't the favorite
vehicle to exploit them.

[s.mesoraca16@gmail.com: fix bug reported by Dan Carpenter]
  Link: https://lkml.kernel.org/r/20180426081456.GA7060@mwanda
  Link: http://lkml.kernel.org/r/1524829819-11275-1-git-send-email-s.mesoraca16@gmail.com
[keescook@chromium.org: drop pr_warn_ratelimited() in favor of audit changes in the future]
[keescook@chromium.org: adjust commit subjet]
Link: http://lkml.kernel.org/r/20180416175918.GA13494@beast
Signed-off-by: Salvatore Mesoraca 
Signed-off-by: Kees Cook 
Suggested-by: Solar Designer 
Suggested-by: Kees Cook 
Cc: Al Viro 
Cc: Dan Carpenter 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings

block: fix an integer overflow in logical block size

2020-04-28T18:03:28+00:00

commit ad6bf88a6c19a39fb3b0045d78ea880325dfcf15 upstream.

Logical block size has type unsigned short. That means that it can be at
most 32768. However, there are architectures that can run with 64k pages
(for example arm64) and on these architectures, it may be possible to
create block devices with 64k block size.

For exmaple (run this on an architecture with 64k pages):

Mount will fail with this error because it tries to read the superblock using 2-sector
access:
  device-mapper: writecache: I/O is not aligned, sector 2, size 1024, block size 65536
  EXT4-fs (dm-0): unable to read superblock

This patch changes the logical block size from unsigned short to unsigned
int to avoid the overflow.

Reviewed-by: Martin K. Petersen 
Reviewed-by: Ming Lei 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Jens Axboe 
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings