linux-toradex.git/fs/pnode.c, branch v4.17-rc1 Linux kernel for Apalis and Colibri modules mnt: Make propagate_umount less slow for overlapping mount propagation trees 2017-05-23T13:41:17+00:00 Eric W. Biederman ebiederm@xmission.com 2016-10-24T22:25:19+00:00 296990deb389c7da21c78030376ba244dc1badf5 Andrei Vagin pointed out that time to executue propagate_umount can go non-linear (and take a ludicrious amount of time) when the mount propogation trees of the mounts to be unmunted by a lazy unmount overlap. Make the walk of the mount propagation trees nearly linear by remembering which mounts have already been visited, allowing subsequent walks to detect when walking a mount propgation tree or a subtree of a mount propgation tree would be duplicate work and to skip them entirely. Walk the list of mounts whose propgatation trees need to be traversed from the mount highest in the mount tree to mounts lower in the mount tree so that odds are higher that the code will walk the largest trees first, allowing later tree walks to be skipped entirely. Add cleanup_umount_visitation to remover the code's memory of which mounts have been visited. Add the functions last_slave and skip_propagation_subtree to allow skipping appropriate parts of the mount propagation tree without needing to change the logic of the rest of the code. A script to generate overlapping mount propagation trees: $ cat runs.h set -e mount -t tmpfs zdtm /mnt mkdir -p /mnt/1 /mnt/2 mount -t tmpfs zdtm /mnt/1 mount --make-shared /mnt/1 mkdir /mnt/1/1 iteration=10 if [ -n "$1" ] ; then iteration=$1 fi for i in $(seq $iteration); do mount --bind /mnt/1/1 /mnt/1/1 done mount --rbind /mnt/1 /mnt/2 TIMEFORMAT='%Rs' nr=$(( ( 2 ** ( $iteration + 1 ) ) + 1 )) echo -n "umount -l /mnt/1 -> $nr " time umount -l /mnt/1 nr=$(cat /proc/self/mountinfo | grep zdtm | wc -l ) time umount -l /mnt/2 $ for i in $(seq 9 19); do echo $i; unshare -Urm bash ./run.sh $i; done Here are the performance numbers with and without the patch: mhash | 8192 | 8192 | 1048576 | 1048576 mounts | before | after | before | after ------------------------------------------------ 1025 | 0.040s | 0.016s | 0.038s | 0.019s 2049 | 0.094s | 0.017s | 0.080s | 0.018s 4097 | 0.243s | 0.019s | 0.206s | 0.023s 8193 | 1.202s | 0.028s | 1.562s | 0.032s 16385 | 9.635s | 0.036s | 9.952s | 0.041s 32769 | 60.928s | 0.063s | 44.321s | 0.064s 65537 | | 0.097s | | 0.097s 131073 | | 0.233s | | 0.176s 262145 | | 0.653s | | 0.344s 524289 | | 2.305s | | 0.735s 1048577 | | 7.107s | | 2.603s Andrei Vagin reports fixing the performance problem is part of the work to fix CVE-2016-6213. Cc: stable@vger.kernel.org Fixes: a05964f3917c ("[PATCH] shared mounts handling: umount") Reported-by: Andrei Vagin <avagin@openvz.org> Reviewed-by: Andrei Vagin <avagin@virtuozzo.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
Andrei Vagin pointed out that time to executue propagate_umount can go
non-linear (and take a ludicrious amount of time) when the mount
propogation trees of the mounts to be unmunted by a lazy unmount
overlap.

Make the walk of the mount propagation trees nearly linear by
remembering which mounts have already been visited, allowing
subsequent walks to detect when walking a mount propgation tree or a
subtree of a mount propgation tree would be duplicate work and to skip
them entirely.

Walk the list of mounts whose propgatation trees need to be traversed
from the mount highest in the mount tree to mounts lower in the mount
tree so that odds are higher that the code will walk the largest trees
first, allowing later tree walks to be skipped entirely.

Add cleanup_umount_visitation to remover the code's memory of which
mounts have been visited.

Add the functions last_slave and skip_propagation_subtree to allow
skipping appropriate parts of the mount propagation tree without
needing to change the logic of the rest of the code.

A script to generate overlapping mount propagation trees:

$ cat runs.h
set -e
mount -t tmpfs zdtm /mnt
mkdir -p /mnt/1 /mnt/2
mount -t tmpfs zdtm /mnt/1
mount --make-shared /mnt/1
mkdir /mnt/1/1

iteration=10
if [ -n "$1" ] ; then
	iteration=$1
fi

for i in $(seq $iteration); do
	mount --bind /mnt/1/1 /mnt/1/1
done

mount --rbind /mnt/1 /mnt/2

TIMEFORMAT='%Rs'
nr=$(( ( 2 ** ( $iteration + 1 ) ) + 1 ))
echo -n "umount -l /mnt/1 -> $nr        "
time umount -l /mnt/1

nr=$(cat /proc/self/mountinfo | grep zdtm | wc -l )
time umount -l /mnt/2

$ for i in $(seq 9 19); do echo $i; unshare -Urm bash ./run.sh $i; done

Here are the performance numbers with and without the patch:

     mhash |  8192   |  8192  | 1048576 | 1048576
    mounts | before  | after  |  before | after
    ------------------------------------------------
      1025 |  0.040s | 0.016s |  0.038s | 0.019s
      2049 |  0.094s | 0.017s |  0.080s | 0.018s
      4097 |  0.243s | 0.019s |  0.206s | 0.023s
      8193 |  1.202s | 0.028s |  1.562s | 0.032s
     16385 |  9.635s | 0.036s |  9.952s | 0.041s
     32769 | 60.928s | 0.063s | 44.321s | 0.064s
     65537 |         | 0.097s |         | 0.097s
    131073 |         | 0.233s |         | 0.176s
    262145 |         | 0.653s |         | 0.344s
    524289 |         | 2.305s |         | 0.735s
   1048577 |         | 7.107s |         | 2.603s

Andrei Vagin reports fixing the performance problem is part of the
work to fix CVE-2016-6213.

Cc: stable@vger.kernel.org
Fixes: a05964f3917c ("[PATCH] shared mounts handling: umount")
Reported-by: Andrei Vagin <avagin@openvz.org>
Reviewed-by: Andrei Vagin <avagin@virtuozzo.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
mnt: In propgate_umount handle visiting mounts in any order 2017-05-23T13:41:16+00:00 Eric W. Biederman ebiederm@xmission.com 2016-10-24T21:16:13+00:00 99b19d16471e9c3faa85cad38abc9cbbe04c6d55 While investigating some poor umount performance I realized that in the case of overlapping mount trees where some of the mounts are locked the code has been failing to unmount all of the mounts it should have been unmounting. This failure to unmount all of the necessary mounts can be reproduced with: $ cat locked_mounts_test.sh mount -t tmpfs test-base /mnt mount --make-shared /mnt mkdir -p /mnt/b mount -t tmpfs test1 /mnt/b mount --make-shared /mnt/b mkdir -p /mnt/b/10 mount -t tmpfs test2 /mnt/b/10 mount --make-shared /mnt/b/10 mkdir -p /mnt/b/10/20 mount --rbind /mnt/b /mnt/b/10/20 unshare -Urm --propagation unchaged /bin/sh -c 'sleep 5; if [ $(grep test /proc/self/mountinfo | wc -l) -eq 1 ] ; then echo SUCCESS ; else echo FAILURE ; fi' sleep 1 umount -l /mnt/b wait %% $ unshare -Urm ./locked_mounts_test.sh This failure is corrected by removing the prepass that marks mounts that may be umounted. A first pass is added that umounts mounts if possible and if not sets mount mark if they could be unmounted if they weren't locked and adds them to a list to umount possibilities. This first pass reconsiders the mounts parent if it is on the list of umount possibilities, ensuring that information of umoutability will pass from child to mount parent. A second pass then walks through all mounts that are umounted and processes their children unmounting them or marking them for reparenting. A last pass cleans up the state on the mounts that could not be umounted and if applicable reparents them to their first parent that remained mounted. While a bit longer than the old code this code is much more robust as it allows information to flow up from the leaves and down from the trunk making the order in which mounts are encountered in the umount propgation tree irrelevant. Cc: stable@vger.kernel.org Fixes: 0c56fe31420c ("mnt: Don't propagate unmounts to locked mounts") Reviewed-by: Andrei Vagin <avagin@virtuozzo.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
While investigating some poor umount performance I realized that in
the case of overlapping mount trees where some of the mounts are locked
the code has been failing to unmount all of the mounts it should
have been unmounting.

This failure to unmount all of the necessary
mounts can be reproduced with:

$ cat locked_mounts_test.sh

mount -t tmpfs test-base /mnt
mount --make-shared /mnt
mkdir -p /mnt/b

mount -t tmpfs test1 /mnt/b
mount --make-shared /mnt/b
mkdir -p /mnt/b/10

mount -t tmpfs test2 /mnt/b/10
mount --make-shared /mnt/b/10
mkdir -p /mnt/b/10/20

mount --rbind /mnt/b /mnt/b/10/20

unshare -Urm --propagation unchaged /bin/sh -c 'sleep 5; if [ $(grep test /proc/self/mountinfo | wc -l) -eq 1 ] ; then echo SUCCESS ; else echo FAILURE ; fi'
sleep 1
umount -l /mnt/b
wait %%

$ unshare -Urm ./locked_mounts_test.sh

This failure is corrected by removing the prepass that marks mounts
that may be umounted.

A first pass is added that umounts mounts if possible and if not sets
mount mark if they could be unmounted if they weren't locked and adds
them to a list to umount possibilities.  This first pass reconsiders
the mounts parent if it is on the list of umount possibilities, ensuring
that information of umoutability will pass from child to mount parent.

A second pass then walks through all mounts that are umounted and processes
their children unmounting them or marking them for reparenting.

A last pass cleans up the state on the mounts that could not be umounted
and if applicable reparents them to their first parent that remained
mounted.

While a bit longer than the old code this code is much more robust
as it allows information to flow up from the leaves and down
from the trunk making the order in which mounts are encountered
in the umount propgation tree irrelevant.

Cc: stable@vger.kernel.org
Fixes: 0c56fe31420c ("mnt: Don't propagate unmounts to locked mounts")
Reviewed-by: Andrei Vagin <avagin@virtuozzo.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
mnt: In umount propagation reparent in a separate pass 2017-05-23T13:40:32+00:00 Eric W. Biederman ebiederm@xmission.com 2017-05-15T19:42:07+00:00 570487d3faf2a1d8a220e6ee10f472163123d7da It was observed that in some pathlogical cases that the current code does not unmount everything it should. After investigation it was determined that the issue is that mnt_change_mntpoint can can change which mounts are available to be unmounted during mount propagation which is wrong. The trivial reproducer is: $ cat ./pathological.sh mount -t tmpfs test-base /mnt cd /mnt mkdir 1 2 1/1 mount --bind 1 1 mount --make-shared 1 mount --bind 1 2 mount --bind 1/1 1/1 mount --bind 1/1 1/1 echo grep test-base /proc/self/mountinfo umount 1/1 echo grep test-base /proc/self/mountinfo $ unshare -Urm ./pathological.sh The expected output looks like: 46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000 47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 49 54 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 50 53 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 51 49 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 54 47 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 53 48 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 52 50 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000 47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 The output without the fix looks like: 46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000 47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 49 54 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 50 53 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 51 49 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 54 47 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 53 48 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 52 50 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000 47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 52 48 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000 That last mount in the output was in the propgation tree to be unmounted but was missed because the mnt_change_mountpoint changed it's parent before the walk through the mount propagation tree observed it. Cc: stable@vger.kernel.org Fixes: 1064f874abc0 ("mnt: Tuck mounts under others instead of creating shadow/side mounts.") Acked-by: Andrei Vagin <avagin@virtuozzo.com> Reviewed-by: Ram Pai <linuxram@us.ibm.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
It was observed that in some pathlogical cases that the current code
does not unmount everything it should.  After investigation it
was determined that the issue is that mnt_change_mntpoint can
can change which mounts are available to be unmounted during mount
propagation which is wrong.

The trivial reproducer is:
$ cat ./pathological.sh

mount -t tmpfs test-base /mnt
cd /mnt
mkdir 1 2 1/1
mount --bind 1 1
mount --make-shared 1
mount --bind 1 2
mount --bind 1/1 1/1
mount --bind 1/1 1/1
echo
grep test-base /proc/self/mountinfo
umount 1/1
echo
grep test-base /proc/self/mountinfo

$ unshare -Urm ./pathological.sh

The expected output looks like:
46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000
47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
49 54 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
50 53 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
51 49 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
54 47 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
53 48 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
52 50 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000

46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000
47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000

The output without the fix looks like:
46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000
47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
49 54 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
50 53 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
51 49 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
54 47 0:25 /1/1 /mnt/1/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
53 48 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
52 50 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000

46 31 0:25 / /mnt rw,relatime - tmpfs test-base rw,uid=1000,gid=1000
47 46 0:25 /1 /mnt/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
48 46 0:25 /1 /mnt/2 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000
52 48 0:25 /1/1 /mnt/2/1 rw,relatime shared:1 - tmpfs test-base rw,uid=1000,gid=1000

That last mount in the output was in the propgation tree to be unmounted but
was missed because the mnt_change_mountpoint changed it's parent before the walk
through the mount propagation tree observed it.

Cc: stable@vger.kernel.org
Fixes: 1064f874abc0 ("mnt: Tuck mounts under others instead of creating shadow/side mounts.")
Acked-by: Andrei Vagin <avagin@virtuozzo.com>
Reviewed-by: Ram Pai <linuxram@us.ibm.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
mnt: Tuck mounts under others instead of creating shadow/side mounts. 2017-02-03T11:01:06+00:00 Eric W. Biederman ebiederm@xmission.com 2017-01-20T05:28:35+00:00 1064f874abc0d05eeed8993815f584d847b72486 Ever since mount propagation was introduced in cases where a mount in propagated to parent mount mountpoint pair that is already in use the code has placed the new mount behind the old mount in the mount hash table. This implementation detail is problematic as it allows creating arbitrary length mount hash chains. Furthermore it invalidates the constraint maintained elsewhere in the mount code that a parent mount and a mountpoint pair will have exactly one mount upon them. Making it hard to deal with and to talk about this special case in the mount code. Modify mount propagation to notice when there is already a mount at the parent mount and mountpoint where a new mount is propagating to and place that preexisting mount on top of the new mount. Modify unmount propagation to notice when a mount that is being unmounted has another mount on top of it (and no other children), and to replace the unmounted mount with the mount on top of it. Move the MNT_UMUONT test from __lookup_mnt_last into __propagate_umount as that is the only call of __lookup_mnt_last where MNT_UMOUNT may be set on any mount visible in the mount hash table. These modifications allow: - __lookup_mnt_last to be removed. - attach_shadows to be renamed __attach_mnt and its shadow handling to be removed. - commit_tree to be simplified - copy_tree to be simplified The result is an easier to understand tree of mounts that does not allow creation of arbitrary length hash chains in the mount hash table. The result is also a very slight userspace visible difference in semantics. The following two cases now behave identically, where before order mattered: case 1: (explicit user action) B is a slave of A mount something on A/a , it will propagate to B/a and than mount something on B/a case 2: (tucked mount) B is a slave of A mount something on B/a and than mount something on A/a Histroically umount A/a would fail in case 1 and succeed in case 2. Now umount A/a succeeds in both configurations. This very small change in semantics appears if anything to be a bug fix to me and my survey of userspace leads me to believe that no programs will notice or care of this subtle semantic change. v2: Updated to mnt_change_mountpoint to not call dput or mntput and instead to decrement the counts directly. It is guaranteed that there will be other references when mnt_change_mountpoint is called so this is safe. v3: Moved put_mountpoint under mount_lock in attach_recursive_mnt As the locking in fs/namespace.c changed between v2 and v3. v4: Reworked the logic in propagate_mount_busy and __propagate_umount that detects when a mount completely covers another mount. v5: Removed unnecessary tests whose result is alwasy true in find_topper and attach_recursive_mnt. v6: Document the user space visible semantic difference. Cc: stable@vger.kernel.org Fixes: b90fa9ae8f51 ("[PATCH] shared mount handling: bind and rbind") Tested-by: Andrei Vagin <avagin@virtuozzo.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
Ever since mount propagation was introduced in cases where a mount in
propagated to parent mount mountpoint pair that is already in use the
code has placed the new mount behind the old mount in the mount hash
table.

This implementation detail is problematic as it allows creating
arbitrary length mount hash chains.

Furthermore it invalidates the constraint maintained elsewhere in the
mount code that a parent mount and a mountpoint pair will have exactly
one mount upon them.  Making it hard to deal with and to talk about
this special case in the mount code.

Modify mount propagation to notice when there is already a mount at
the parent mount and mountpoint where a new mount is propagating to
and place that preexisting mount on top of the new mount.

Modify unmount propagation to notice when a mount that is being
unmounted has another mount on top of it (and no other children), and
to replace the unmounted mount with the mount on top of it.

Move the MNT_UMUONT test from __lookup_mnt_last into
__propagate_umount as that is the only call of __lookup_mnt_last where
MNT_UMOUNT may be set on any mount visible in the mount hash table.

These modifications allow:
 - __lookup_mnt_last to be removed.
 - attach_shadows to be renamed __attach_mnt and its shadow
   handling to be removed.
 - commit_tree to be simplified
 - copy_tree to be simplified

The result is an easier to understand tree of mounts that does not
allow creation of arbitrary length hash chains in the mount hash table.

The result is also a very slight userspace visible difference in semantics.
The following two cases now behave identically, where before order
mattered:

case 1: (explicit user action)
	B is a slave of A
	mount something on A/a , it will propagate to B/a
	and than mount something on B/a

case 2: (tucked mount)
	B is a slave of A
	mount something on B/a
	and than mount something on A/a

Histroically umount A/a would fail in case 1 and succeed in case 2.
Now umount A/a succeeds in both configurations.

This very small change in semantics appears if anything to be a bug
fix to me and my survey of userspace leads me to believe that no programs
will notice or care of this subtle semantic change.

v2: Updated to mnt_change_mountpoint to not call dput or mntput
and instead to decrement the counts directly.  It is guaranteed
that there will be other references when mnt_change_mountpoint is
called so this is safe.

v3: Moved put_mountpoint under mount_lock in attach_recursive_mnt
    As the locking in fs/namespace.c changed between v2 and v3.

v4: Reworked the logic in propagate_mount_busy and __propagate_umount
    that detects when a mount completely covers another mount.

v5: Removed unnecessary tests whose result is alwasy true in
    find_topper and attach_recursive_mnt.

v6: Document the user space visible semantic difference.

Cc: stable@vger.kernel.org
Fixes: b90fa9ae8f51 ("[PATCH] shared mount handling: bind and rbind")
Tested-by: Andrei Vagin <avagin@virtuozzo.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
reorganize do_make_slave() 2016-12-16T21:30:49+00:00 Al Viro viro@zeniv.linux.org.uk 2016-11-21T00:33:09+00:00 5235d448c48e1f5a4a34bf90d412775cb75ffb32 Make sure that clone_mnt() never returns a mount with MNT_SHARED in flags, but without a valid ->mnt_group_id. That allows to demystify do_make_slave() quite a bit, among other things. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> 
Make sure that clone_mnt() never returns a mount with MNT_SHARED in
flags, but without a valid ->mnt_group_id.  That allows to demystify
do_make_slave() quite a bit, among other things.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
mnt: Add a per mount namespace limit on the number of mounts 2016-09-30T17:46:48+00:00 Eric W. Biederman ebiederm@xmission.com 2016-09-28T05:27:17+00:00 d29216842a85c7970c536108e093963f02714498 CAI Qian <caiqian@redhat.com> pointed out that the semantics of shared subtrees make it possible to create an exponentially increasing number of mounts in a mount namespace. mkdir /tmp/1 /tmp/2 mount --make-rshared / for i in $(seq 1 20) ; do mount --bind /tmp/1 /tmp/2 ; done Will create create 2^20 or 1048576 mounts, which is a practical problem as some people have managed to hit this by accident. As such CVE-2016-6213 was assigned. Ian Kent <raven@themaw.net> described the situation for autofs users as follows: > The number of mounts for direct mount maps is usually not very large because of > the way they are implemented, large direct mount maps can have performance > problems. There can be anywhere from a few (likely case a few hundred) to less > than 10000, plus mounts that have been triggered and not yet expired. > > Indirect mounts have one autofs mount at the root plus the number of mounts that > have been triggered and not yet expired. > > The number of autofs indirect map entries can range from a few to the common > case of several thousand and in rare cases up to between 30000 and 50000. I've > not heard of people with maps larger than 50000 entries. > > The larger the number of map entries the greater the possibility for a large > number of active mounts so it's not hard to expect cases of a 1000 or somewhat > more active mounts. So I am setting the default number of mounts allowed per mount namespace at 100,000. This is more than enough for any use case I know of, but small enough to quickly stop an exponential increase in mounts. Which should be perfect to catch misconfigurations and malfunctioning programs. For anyone who needs a higher limit this can be changed by writing to the new /proc/sys/fs/mount-max sysctl. Tested-by: CAI Qian <caiqian@redhat.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
CAI Qian <caiqian@redhat.com> pointed out that the semantics
of shared subtrees make it possible to create an exponentially
increasing number of mounts in a mount namespace.

    mkdir /tmp/1 /tmp/2
    mount --make-rshared /
    for i in $(seq 1 20) ; do mount --bind /tmp/1 /tmp/2 ; done

Will create create 2^20 or 1048576 mounts, which is a practical problem
as some people have managed to hit this by accident.

As such CVE-2016-6213 was assigned.

Ian Kent <raven@themaw.net> described the situation for autofs users
as follows:

> The number of mounts for direct mount maps is usually not very large because of
> the way they are implemented, large direct mount maps can have performance
> problems. There can be anywhere from a few (likely case a few hundred) to less
> than 10000, plus mounts that have been triggered and not yet expired.
>
> Indirect mounts have one autofs mount at the root plus the number of mounts that
> have been triggered and not yet expired.
>
> The number of autofs indirect map entries can range from a few to the common
> case of several thousand and in rare cases up to between 30000 and 50000. I've
> not heard of people with maps larger than 50000 entries.
>
> The larger the number of map entries the greater the possibility for a large
> number of active mounts so it's not hard to expect cases of a 1000 or somewhat
> more active mounts.

So I am setting the default number of mounts allowed per mount
namespace at 100,000.  This is more than enough for any use case I
know of, but small enough to quickly stop an exponential increase
in mounts.  Which should be perfect to catch misconfigurations and
malfunctioning programs.

For anyone who needs a higher limit this can be changed by writing
to the new /proc/sys/fs/mount-max sysctl.

Tested-by: CAI Qian <caiqian@redhat.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
propogate_mnt: Handle the first propogated copy being a slave 2016-05-05T14:54:45+00:00 Eric W. Biederman ebiederm@xmission.com 2016-05-05T14:29:29+00:00 5ec0811d30378ae104f250bfc9b3640242d81e3f When the first propgated copy was a slave the following oops would result: > BUG: unable to handle kernel NULL pointer dereference at 0000000000000010 > IP: [<ffffffff811fba4e>] propagate_one+0xbe/0x1c0 > PGD bacd4067 PUD bac66067 PMD 0 > Oops: 0000 [#1] SMP > Modules linked in: > CPU: 1 PID: 824 Comm: mount Not tainted 4.6.0-rc5userns+ #1523 > Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007 > task: ffff8800bb0a8000 ti: ffff8800bac3c000 task.ti: ffff8800bac3c000 > RIP: 0010:[<ffffffff811fba4e>] [<ffffffff811fba4e>] propagate_one+0xbe/0x1c0 > RSP: 0018:ffff8800bac3fd38 EFLAGS: 00010283 > RAX: 0000000000000000 RBX: ffff8800bb77ec00 RCX: 0000000000000010 > RDX: 0000000000000000 RSI: ffff8800bb58c000 RDI: ffff8800bb58c480 > RBP: ffff8800bac3fd48 R08: 0000000000000001 R09: 0000000000000000 > R10: 0000000000001ca1 R11: 0000000000001c9d R12: 0000000000000000 > R13: ffff8800ba713800 R14: ffff8800bac3fda0 R15: ffff8800bb77ec00 > FS: 00007f3c0cd9b7e0(0000) GS:ffff8800bfb00000(0000) knlGS:0000000000000000 > CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > CR2: 0000000000000010 CR3: 00000000bb79d000 CR4: 00000000000006e0 > Stack: > ffff8800bb77ec00 0000000000000000 ffff8800bac3fd88 ffffffff811fbf85 > ffff8800bac3fd98 ffff8800bb77f080 ffff8800ba713800 ffff8800bb262b40 > 0000000000000000 0000000000000000 ffff8800bac3fdd8 ffffffff811f1da0 > Call Trace: > [<ffffffff811fbf85>] propagate_mnt+0x105/0x140 > [<ffffffff811f1da0>] attach_recursive_mnt+0x120/0x1e0 > [<ffffffff811f1ec3>] graft_tree+0x63/0x70 > [<ffffffff811f1f6b>] do_add_mount+0x9b/0x100 > [<ffffffff811f2c1a>] do_mount+0x2aa/0xdf0 > [<ffffffff8117efbe>] ? strndup_user+0x4e/0x70 > [<ffffffff811f3a45>] SyS_mount+0x75/0xc0 > [<ffffffff8100242b>] do_syscall_64+0x4b/0xa0 > [<ffffffff81988f3c>] entry_SYSCALL64_slow_path+0x25/0x25 > Code: 00 00 75 ec 48 89 0d 02 22 22 01 8b 89 10 01 00 00 48 89 05 fd 21 22 01 39 8e 10 01 00 00 0f 84 e0 00 00 00 48 8b 80 d8 00 00 00 <48> 8b 50 10 48 89 05 df 21 22 01 48 89 15 d0 21 22 01 8b 53 30 > RIP [<ffffffff811fba4e>] propagate_one+0xbe/0x1c0 > RSP <ffff8800bac3fd38> > CR2: 0000000000000010 > ---[ end trace 2725ecd95164f217 ]--- This oops happens with the namespace_sem held and can be triggered by non-root users. An all around not pleasant experience. To avoid this scenario when finding the appropriate source mount to copy stop the walk up the mnt_master chain when the first source mount is encountered. Further rewrite the walk up the last_source mnt_master chain so that it is clear what is going on. The reason why the first source mount is special is that it it's mnt_parent is not a mount in the dest_mnt propagation tree, and as such termination conditions based up on the dest_mnt mount propgation tree do not make sense. To avoid other kinds of confusion last_dest is not changed when computing last_source. last_dest is only used once in propagate_one and that is above the point of the code being modified, so changing the global variable is meaningless and confusing. Cc: stable@vger.kernel.org fixes: f2ebb3a921c1ca1e2ddd9242e95a1989a50c4c68 ("smarter propagate_mnt()") Reported-by: Tycho Andersen <tycho.andersen@canonical.com> Reviewed-by: Seth Forshee <seth.forshee@canonical.com> Tested-by: Seth Forshee <seth.forshee@canonical.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
When the first propgated copy was a slave the following oops would result:
> BUG: unable to handle kernel NULL pointer dereference at 0000000000000010
> IP: [<ffffffff811fba4e>] propagate_one+0xbe/0x1c0
> PGD bacd4067 PUD bac66067 PMD 0
> Oops: 0000 [#1] SMP
> Modules linked in:
> CPU: 1 PID: 824 Comm: mount Not tainted 4.6.0-rc5userns+ #1523
> Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
> task: ffff8800bb0a8000 ti: ffff8800bac3c000 task.ti: ffff8800bac3c000
> RIP: 0010:[<ffffffff811fba4e>]  [<ffffffff811fba4e>] propagate_one+0xbe/0x1c0
> RSP: 0018:ffff8800bac3fd38  EFLAGS: 00010283
> RAX: 0000000000000000 RBX: ffff8800bb77ec00 RCX: 0000000000000010
> RDX: 0000000000000000 RSI: ffff8800bb58c000 RDI: ffff8800bb58c480
> RBP: ffff8800bac3fd48 R08: 0000000000000001 R09: 0000000000000000
> R10: 0000000000001ca1 R11: 0000000000001c9d R12: 0000000000000000
> R13: ffff8800ba713800 R14: ffff8800bac3fda0 R15: ffff8800bb77ec00
> FS:  00007f3c0cd9b7e0(0000) GS:ffff8800bfb00000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 0000000000000010 CR3: 00000000bb79d000 CR4: 00000000000006e0
> Stack:
>  ffff8800bb77ec00 0000000000000000 ffff8800bac3fd88 ffffffff811fbf85
>  ffff8800bac3fd98 ffff8800bb77f080 ffff8800ba713800 ffff8800bb262b40
>  0000000000000000 0000000000000000 ffff8800bac3fdd8 ffffffff811f1da0
> Call Trace:
>  [<ffffffff811fbf85>] propagate_mnt+0x105/0x140
>  [<ffffffff811f1da0>] attach_recursive_mnt+0x120/0x1e0
>  [<ffffffff811f1ec3>] graft_tree+0x63/0x70
>  [<ffffffff811f1f6b>] do_add_mount+0x9b/0x100
>  [<ffffffff811f2c1a>] do_mount+0x2aa/0xdf0
>  [<ffffffff8117efbe>] ? strndup_user+0x4e/0x70
>  [<ffffffff811f3a45>] SyS_mount+0x75/0xc0
>  [<ffffffff8100242b>] do_syscall_64+0x4b/0xa0
>  [<ffffffff81988f3c>] entry_SYSCALL64_slow_path+0x25/0x25
> Code: 00 00 75 ec 48 89 0d 02 22 22 01 8b 89 10 01 00 00 48 89 05 fd 21 22 01 39 8e 10 01 00 00 0f 84 e0 00 00 00 48 8b 80 d8 00 00 00 <48> 8b 50 10 48 89 05 df 21 22 01 48 89 15 d0 21 22 01 8b 53 30
> RIP  [<ffffffff811fba4e>] propagate_one+0xbe/0x1c0
>  RSP <ffff8800bac3fd38>
> CR2: 0000000000000010
> ---[ end trace 2725ecd95164f217 ]---

This oops happens with the namespace_sem held and can be triggered by
non-root users.  An all around not pleasant experience.

To avoid this scenario when finding the appropriate source mount to
copy stop the walk up the mnt_master chain when the first source mount
is encountered.

Further rewrite the walk up the last_source mnt_master chain so that
it is clear what is going on.

The reason why the first source mount is special is that it it's
mnt_parent is not a mount in the dest_mnt propagation tree, and as
such termination conditions based up on the dest_mnt mount propgation
tree do not make sense.

To avoid other kinds of confusion last_dest is not changed when
computing last_source.  last_dest is only used once in propagate_one
and that is above the point of the code being modified, so changing
the global variable is meaningless and confusing.

Cc: stable@vger.kernel.org
fixes: f2ebb3a921c1ca1e2ddd9242e95a1989a50c4c68 ("smarter propagate_mnt()")
Reported-by: Tycho Andersen <tycho.andersen@canonical.com>
Reviewed-by: Seth Forshee <seth.forshee@canonical.com>
Tested-by: Seth Forshee <seth.forshee@canonical.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
fs/pnode.c: treat zero mnt_group_id-s as unequal 2016-02-20T05:15:52+00:00 Maxim Patlasov mpatlasov@virtuozzo.com 2016-02-16T19:45:33+00:00 7ae8fd0351f912b075149a1e03a017be8b903b9a propagate_one(m) calculates "type" argument for copy_tree() like this: > if (m->mnt_group_id == last_dest->mnt_group_id) { > type = CL_MAKE_SHARED; > } else { > type = CL_SLAVE; > if (IS_MNT_SHARED(m)) > type |= CL_MAKE_SHARED; > } The "type" argument then governs clone_mnt() behavior with respect to flags and mnt_master of new mount. When we iterate through a slave group, it is possible that both current "m" and "last_dest" are not shared (although, both are slaves, i.e. have non-NULL mnt_master-s). Then the comparison above erroneously makes new mount shared and sets its mnt_master to last_source->mnt_master. The patch fixes the problem by handling zero mnt_group_id-s as though they are unequal. The similar problem exists in the implementation of "else" clause above when we have to ascend upward in the master/slave tree by calling: > last_source = last_source->mnt_master; > last_dest = last_source->mnt_parent; proper number of times. The last step is governed by "n->mnt_group_id != last_dest->mnt_group_id" condition that may lie if both are zero. The patch fixes this case in the same way as the former one. [AV: don't open-code an obvious helper...] Signed-off-by: Maxim Patlasov <mpatlasov@virtuozzo.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> 
propagate_one(m) calculates "type" argument for copy_tree() like this:

>    if (m->mnt_group_id == last_dest->mnt_group_id) {
>        type = CL_MAKE_SHARED;
>    } else {
>        type = CL_SLAVE;
>        if (IS_MNT_SHARED(m))
>           type |= CL_MAKE_SHARED;
>   }

The "type" argument then governs clone_mnt() behavior with respect to flags
and mnt_master of new mount. When we iterate through a slave group, it is
possible that both current "m" and "last_dest" are not shared (although,
both are slaves, i.e. have non-NULL mnt_master-s). Then the comparison
above erroneously makes new mount shared and sets its mnt_master to
last_source->mnt_master. The patch fixes the problem by handling zero
mnt_group_id-s as though they are unequal.

The similar problem exists in the implementation of "else" clause above
when we have to ascend upward in the master/slave tree by calling:

>    last_source = last_source->mnt_master;
>    last_dest = last_source->mnt_parent;

proper number of times. The last step is governed by
"n->mnt_group_id != last_dest->mnt_group_id" condition that may lie if
both are zero. The patch fixes this case in the same way as the former one.

[AV: don't open-code an obvious helper...]

Signed-off-by: Maxim Patlasov <mpatlasov@virtuozzo.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
mnt: Don't propagate unmounts to locked mounts 2015-04-03T01:34:20+00:00 Eric W. Biederman ebiederm@xmission.com 2015-01-05T19:38:04+00:00 0c56fe31420ca599c90240315f7959bf1b4eb6ce If the first mount in shared subtree is locked don't unmount the shared subtree. This is ensured by walking through the mounts parents before children and marking a mount as unmountable if it is not locked or it is locked but it's parent is marked. This allows recursive mount detach to propagate through a set of mounts when unmounting them would not reveal what is under any locked mount. Cc: stable@vger.kernel.org Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
If the first mount in shared subtree is locked don't unmount the
shared subtree.

This is ensured by walking through the mounts parents before children
and marking a mount as unmountable if it is not locked or it is locked
but it's parent is marked.

This allows recursive mount detach to propagate through a set of
mounts when unmounting them would not reveal what is under any locked
mount.

Cc: stable@vger.kernel.org
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
mnt: On an unmount propagate clearing of MNT_LOCKED 2015-04-03T01:34:19+00:00 Eric W. Biederman ebiederm@xmission.com 2015-01-03T11:39:35+00:00 5d88457eb5b86b475422dc882f089203faaeedb5 A prerequisite of calling umount_tree is that the point where the tree is mounted at is valid to unmount. If we are propagating the effect of the unmount clear MNT_LOCKED in every instance where the same filesystem is mounted on the same mountpoint in the mount tree, as we know (by virtue of the fact that umount_tree was called) that it is safe to reveal what is at that mountpoint. Cc: stable@vger.kernel.org Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> 
A prerequisite of calling umount_tree is that the point where the tree
is mounted at is valid to unmount.

If we are propagating the effect of the unmount clear MNT_LOCKED in
every instance where the same filesystem is mounted on the same
mountpoint in the mount tree, as we know (by virtue of the fact
that umount_tree was called) that it is safe to reveal what
is at that mountpoint.

Cc: stable@vger.kernel.org
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>