linux-stable.git/net/wireless, branch linux-3.16.y

cfg80211/mac80211: make ieee80211_send_layer2_update a public function

2020-04-28T18:02:17+00:00

commit 30ca1aa536211f5ac3de0173513a7a99a98a97f3 upstream.

Make ieee80211_send_layer2_update() a common function so other drivers
can re-use it.

Signed-off-by: Dedy Lansky 
Signed-off-by: Johannes Berg 
[bwh: Backported to 3.16 as dependency of commit 3e493173b784
 "mac80211: Do not send Layer 2 Update frame before authorization":
 - Retain type-casting of skb_put() return value
 - Adjust context]
Signed-off-by: Ben Hutchings

cfg80211: size various nl80211 messages correctly

2020-01-11T02:04:58+00:00

commit 4ef8c1c93f848e360754f10eb2e7134c872b6597 upstream.

Ilan reported that sometimes nl80211 messages weren't working if
the frames being transported got very large, which was really a
problem for userspace-to-kernel messages, but prompted me to look
at the code.

Upon review, I found various places where variable-length data is
transported in an nl80211 message but the message isn't allocated
taking that into account. This shouldn't cause any problems since
the frames aren't really that long, apart in one place where two
(possibly very long frames) might not fit.

Fix all the places (that I found) that get variable length data
from the driver and put it into a message to take the length of
the variable data into account. The 100 there is just a safe
constant for the remaining message overhead (it's usually around
50 for most messages.)

Signed-off-by: Johannes Berg 
Cc: Arnd Bergmann 
Signed-off-by: Ben Hutchings

nl80211: fix validation of mesh path nexthop

2019-12-19T15:58:51+00:00

commit 1fab1b89e2e8f01204a9c05a39fd0b6411a48593 upstream.

Mesh path nexthop should be a ethernet address, but current validation
checks against 4 byte integers.

Fixes: 2ec600d672e74 ("nl80211/cfg80211: support for mesh, sta dumping")
Signed-off-by: Markus Theil 
Link: https://lore.kernel.org/r/20191029093003.10355-1-markus.theil@tu-ilmenau.de
Signed-off-by: Johannes Berg 
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Ben Hutchings

nl80211: validate beacon head

2019-12-10T18:01:29+00:00

commit f88eb7c0d002a67ef31aeb7850b42ff69abc46dc upstream.

We currently don't validate the beacon head, i.e. the header,
fixed part and elements that are to go in front of the TIM
element. This means that the variable elements there can be
malformed, e.g. have a length exceeding the buffer size, but
most downstream code from this assumes that this has already
been checked.

Add the necessary checks to the netlink policy.

Fixes: ed1b6cc7f80f ("cfg80211/nl80211: add beacon settings")
Link: https://lore.kernel.org/r/1569009255-I7ac7fbe9436e9d8733439eab8acbbd35e55c74ef@changeid
Signed-off-by: Johannes Berg 
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Ben Hutchings

cfg80211: Purge frame registrations on iftype change

2019-12-10T18:01:20+00:00

commit c1d3ad84eae35414b6b334790048406bd6301b12 upstream.

Currently frame registrations are not purged, even when changing the
interface type.  This can lead to potentially weird situations where
frames possibly not allowed on a given interface type remain registered
due to the type switching happening after registration.

The kernel currently relies on userspace apps to actually purge the
registrations themselves, this is not something that the kernel should
rely on.

Add a call to cfg80211_mlme_purge_registrations() to forcefully remove
any registrations left over prior to switching the iftype.

Signed-off-by: Denis Kenzior 
Link: https://lore.kernel.org/r/20190828211110.15005-1-denkenz@gmail.com
Signed-off-by: Johannes Berg 
Signed-off-by: Ben Hutchings

Revert "cfg80211: fix processing world regdomain when non modular"

2019-11-22T15:57:18+00:00

commit 0d31d4dbf38412f5b8b11b4511d07b840eebe8cb upstream.

This reverts commit 96cce12ff6e0 ("cfg80211: fix processing world
regdomain when non modular").

Re-triggering a reg_process_hint with the last request on all events,
can make the regulatory domain fail in case of multiple WiFi modules. On
slower boards (espacially with mdev), enumeration of the WiFi modules
can end up in an intersected regulatory domain, and user cannot set it
with 'iw reg set' anymore.

This is happening, because:
- 1st module enumerates, queues up a regulatory request
- request gets processed by __reg_process_hint_driver():
  - checks if previous was set by CORE -> yes
    - checks if regulator domain changed -> yes, from '00' to e.g. 'US'
      -> sends request to the 'crda'
- 2nd module enumerates, queues up a regulator request (which triggers
  the reg_todo() work)
- reg_todo() -> reg_process_pending_hints() sees, that the last request
  is not processed yet, so it tries to process it again.
  __reg_process_hint driver() will run again, and:
  - checks if the last request's initiator was the core -> no, it was
    the driver (1st WiFi module)
  - checks, if the previous initiator was the driver -> yes
    - checks if the regulator domain changed -> yes, it was '00' (set by
      core, and crda call did not return yet), and should be changed to 'US'

------> __reg_process_hint_driver calls an intersect

Besides, the reg_process_hint call with the last request is meaningless
since the crda call has a timeout work. If that timeout expires, the
first module's request will lost.

Fixes: 96cce12ff6e0 ("cfg80211: fix processing world regdomain when non modular")
Signed-off-by: Robert Hodaszi 
Link: https://lore.kernel.org/r/20190614131600.GA13897@a1-hr
Signed-off-by: Johannes Berg 
Signed-off-by: Ben Hutchings

cfg80211: wext: avoid copying malformed SSIDs

2019-11-15T00:56:56+00:00

commit 4ac2813cc867ae563a1ba5a9414bfb554e5796fa upstream.

Ensure the SSID element is bounds-checked prior to invoking memcpy()
with its length field, when copying to userspace.

Cc: Kees Cook 
Reported-by: Nicolas Waisman 
Signed-off-by: Will Deacon 
Link: https://lore.kernel.org/r/20191004095132.15777-2-will@kernel.org
[adjust commit log a bit]
Signed-off-by: Johannes Berg 
Signed-off-by: Ben Hutchings

cfg80211: fix memory leak of wiphy device name

2019-10-05T15:19:59+00:00

commit 4f488fbca2a86cc7714a128952eead92cac279ab upstream.

In wiphy_new_nm(), if an error occurs after dev_set_name() and
device_initialize() have already been called, it's necessary to call
put_device() (via wiphy_free()) to avoid a memory leak.

Reported-by: syzbot+7fddca22578bc67c3fe4@syzkaller.appspotmail.com
Fixes: 1f87f7d3a3b4 ("cfg80211: add rfkill support")
Signed-off-by: Eric Biggers 
Signed-off-by: Johannes Berg 
Signed-off-by: Ben Hutchings

cfg80211: reg: Init wiphy_idx in regulatory_hint_core()

2018-12-16T22:09:04+00:00

commit 24f33e64fcd0d50a4b1a8e5b41bd0257aa66b0e8 upstream.

Core regulatory hints didn't set wiphy_idx to WIPHY_IDX_INVALID. Since
the regulatory request is zeroed, wiphy_idx was always implicitly set to
0. This resulted in updating only phy #0.
Fix that.

Fixes: 806a9e39670b ("cfg80211: make regulatory_request use wiphy_idx instead of wiphy")
Signed-off-by: Andrei Otcheretianski 
Signed-off-by: Luca Coelho 
[add fixes tag]
Signed-off-by: Johannes Berg 
Signed-off-by: Ben Hutchings

cfg80211: nl80211_update_ft_ies() to validate NL80211_ATTR_IE

2018-12-16T22:08:57+00:00

commit 4f0223bfe9c3e62d8f45a85f1ef1b18a8a263ef9 upstream.

nl80211_update_ft_ies() tried to validate NL80211_ATTR_IE with
is_valid_ie_attr() before dereferencing it, but that helper function
returns true in case of NULL pointer (i.e., attribute not included).
This can result to dereferencing a NULL pointer. Fix that by explicitly
checking that NL80211_ATTR_IE is included.

Fixes: 355199e02b83 ("cfg80211: Extend support for IEEE 802.11r Fast BSS Transition")
Signed-off-by: Arunk Khandavalli 
Signed-off-by: Jouni Malinen 
Signed-off-by: Johannes Berg 
Signed-off-by: Ben Hutchings