- 12 Apr, 2017 40 commits
-
-
Dean Jenkins authored
There is a race condition between a thread calling bt_accept_dequeue() and a different thread calling bt_accept_unlink(). Protection against concurrency is implemented using sk locking. However, sk locking causes serialisation of the bt_accept_dequeue() and bt_accept_unlink() threads. This serialisation can cause bt_accept_dequeue() to obtain the sk from the parent list but becomes blocked waiting for the sk lock held by the bt_accept_unlink() thread. bt_accept_unlink() unlinks sk and this thread releases the sk lock unblocking bt_accept_dequeue() which potentially runs bt_accept_unlink() again on the same sk causing a crash. The attempt to double unlink the same sk from the parent list can cause a NULL pointer dereference crash due to bt_sk(sk)->parent becoming NULL on the first unlink, followed by the second unlink trying to execute bt_sk(sk)->parent->sk_ack_backlog-- in bt_accept_unlink() which crashes. When sk is in the parent list, bt_sk(sk)->parent will be not be NULL. When sk is removed from the parent list, bt_sk(sk)->parent is set to NULL. Therefore, add a defensive check for bt_sk(sk)->parent not being NULL to ensure that sk is still in the parent list after the sk lock has been taken in bt_accept_dequeue(). If bt_sk(sk)->parent is detected as being NULL then restart the loop so that the loop variables are refreshed to use the latest values. This is necessary as list_for_each_entry_safe() is not thread safe so causing a risk of an infinite loop occurring as sk could point to itself. In addition, in bt_accept_dequeue() increase the sk reference count to protect against early freeing of sk. Early freeing can be possible if the bt_accept_unlink() thread calls l2cap_sock_kill() or rfcomm_sock_kill() functions before bt_accept_dequeue() gets the sk lock. For test purposes, the probability of failure can be increased by putting a msleep of 1 second in bt_accept_dequeue() between getting the sk and waiting for the sk lock. This exposes the fact that the loop list_for_each_entry_safe(p, n, &bt_sk(parent)->accept_q) is not safe from threads that unlink sk from the list in parallel with the loop which can cause sk to become stale within the loop. Signed-off-by:
Dean Jenkins <Dean_Jenkins@mentor.com> Signed-off-by:
Marcel Holtmann <marcel@holtmann.org>
-
Dean Jenkins authored
There is a small risk that bt_accept_unlink() runs concurrently with bt_accept_enqueue() on the same socket. This scenario could potentially lead to a NULL pointer dereference of the socket's parent member because the socket can be on the list but the socket's parent member is not yet updated by bt_accept_enqueue(). Therefore, add socket locking inside bt_accept_enqueue() so that the socket is added to the list AND the parent's socket address is set in the socket's parent member. The socket locking ensures that the socket is on the list with a valid non-NULL parent member. Signed-off-by:
-
Geliang Tang authored
Use setup_timer() instead of init_timer() to simplify the code. Signed-off-by:
Geliang Tang <geliangtang@gmail.com> Signed-off-by:
-
Luiz Augusto von Dentz authored
According to RFC 7668 U/L bit shall not be used: https://wiki.tools.ietf.org/html/rfc7668#section-3.2.2 [Page 10]: In the figure, letter 'b' represents a bit from the Bluetooth device address, copied as is without any changes on any bit. This means that no bit in the IID indicates whether the underlying Bluetooth device address is public or random. |0 1|1 3|3 4|4 6| |0 5|6 1|2 7|8 3| +----------------+----------------+----------------+----------------+ |bbbbbbbbbbbbbbbb|bbbbbbbb11111111|11111110bbbbbbbb|bbbbbbbbbbbbbbbb| +----------------+----------------+----------------+----------------+ Because of this the code cannot figure out the address type from the IP address anymore thus it makes no sense to use peer_lookup_ba as it needs the peer address type. Signed-off-by:
Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Reviewed-by:
Stefan Schmidt <stefan@osg.samsung.com> Acked-by:
Jukka Rissanen <jukka.rissanen@linux.intel.com> Signed-off-by:
-
Luiz Augusto von Dentz authored
This allow technologies such as Bluetooth to use its native lladdr which is eui48 instead of eui64 which was expected by functions like lowpan_header_decompress and lowpan_header_compress. Signed-off-by:
-
Alexander Aring authored
This patch adds support for 48 bit 6LoWPAN address length autoconfiguration which is the case for BTLE 6LoWPAN. Signed-off-by:
Alexander Aring <aar@pengutronix.de> Signed-off-by:
David S. Miller <davem@davemloft.net> Signed-off-by:
-
Alexander Aring authored
The skb->pkt_type need to be set by L2, but on 6LoWPAN there exists L2 e.g. BTLE which doesn't has multicast addressing. If it's a multicast or not is detected by IPHC headers multicast bit. The IPv6 layer will evaluate this pkt_type, so we force set this type while uncompressing. Should be okay for 802.15.4 as well. Signed-off-by:
-
Patrik Flykt authored
Set MAC address length according to the 6LoWPAN link layer in use. Bluetooth Low Energy uses 48 bit addressing while IEEE802.15.4 uses 64 bits. Signed-off-by:
Patrik Flykt <patrik.flykt@linux.intel.com> Reviewed-by:
-
Patrik Flykt authored
Set BTLE MAC addresses that are 6 bytes long and not 8 bytes that are used in other places with 6lowpan. Signed-off-by:
-
John Keeping authored
The hci_bcm driver currently does not prepare/unprepare the clock and goes directly to enable, but as the documentation for clk_enable says, clk_prepare must be called before clk_enable. Signed-off-by:
John Keeping <john@metanate.com> Signed-off-by:
-
Elena Reshetova authored
refcount_t type and corresponding API should be used instead of atomic_t when the variable is used as a reference counter. This allows to avoid accidental refcounter overflows that might lead to use-after-free situations. Signed-off-by:
Elena Reshetova <elena.reshetova@intel.com> Signed-off-by:
Hans Liljestrand <ishkamiel@gmail.com> Signed-off-by:
Kees Cook <keescook@chromium.org> Signed-off-by:
David Windsor <dwindsor@gmail.com> Signed-off-by:
-
Colin Ian King authored
trivial fix to spelling mistake in debug message Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Alexey Dobriyan authored
This field is never big enough to warrant 16-bitness. 8-bit accesses enjoy shorted encoding on i386/x86_64 than 16-bit accesses: add/remove: 0/0 grow/shrink: 0/2 up/down: 0/-10 (-10) function old new delta loopback_setup 169 164 -5 ether_setup 148 143 -5 Signed-off-by:
Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by:
-
Alexey Dobriyan authored
Using 16-bit ->hh_len doesn't save any memory, save some .text instead: add/remove: 0/0 grow/shrink: 1/6 up/down: 2/-19 (-17) function old new delta neigh_update 2312 2314 +2 fwnet_header_cache 199 197 -2 eth_header_cache 101 99 -2 ip6_finish_output2 2371 2368 -3 vrf_finish_output6 1522 1518 -4 vrf_finish_output 1413 1409 -4 ip_finish_output2 1627 1623 -4 Signed-off-by:
-
Ilan Tayari authored
A driver may use build_skb() for received packets. These SKBs then have a head_frag. Since commit d7e8883c ("net: make GRO aware of skb->head_frag"), GRO may build frag_list SKBs out of head_frag received SKBs. In such a case, the chained SKBs end up with a head_frag. Commit 07b26c94 ("gso: Support partial splitting at the frag_list pointer") adds partial segmentation of frag_list SKB chains into individual SKBs. However, this is not done if the chained SKBs have any linear part, because the device may not be able to DMA the private linear buffer. A chained frag_list SKB with head_frag is wrongfully detected in this case as having a private linear part and thus falls back to software GSO, while in fact the linear part is backed by a DMA page just like any other frag. This causes low performance when forwarding those packets that were built with build_skb() Allow partial segmentation at the frag_list pointer for chained SKBs with head_frag. Note that such SKBs can only be created by GRO, when applied to received packets with head_frag. Also note that this change only affects the data path that performs the partial segmentation at frag_list pointer, and not any of the other more common data paths. Signed-off-by:
Ilan Tayari <ilant@mellanox.com> Signed-off-by:
-
David S. Miller authored
Guillaume Nault says: ==================== l2tp: constify l2tp_session_get*() and l2tp_tunnel_find*() Declare parameters of these functions as "const" where possible. ==================== Signed-off-by: -
Guillaume Nault authored
l2tp_tunnel_find() and l2tp_tunnel_find_nth() don't modify "net". Signed-off-by:
Guillaume Nault <g.nault@alphalink.fr> Signed-off-by:
-
Guillaume Nault authored
Make l2tp_pernet()'s parameter constant, so that l2tp_session_get*() can declare their "net" variable as "const". Also constify "ifname" in l2tp_session_get_by_ifname(). Signed-off-by:
-
David S. Miller authored
Benjamin Herrenschmidt says: ==================== ftgmac100: Rework batch 4 - Misc This is v2 of the fourth batch of updates to the ftgmac100 driver. This is a bunch of misc cleanups and fixes, such as properly disabling HW checksum generation on AST2400 where it's known to be broken and some chip init updates. This also adds the ability to turn HW checksum on/off and configure the ring sizes via ethtool. v2 Fixes patch 1/10 (NETIF_F_HW_CSUM conversion) The next (and last) batch will add a few more "features" such as netpoll, multicast/promist, vlan offload... ==================== Signed-off-by: -
Benjamin Herrenschmidt authored
I haven't seen any improvement above that size on the machines I've tested with. Signed-off-by:
Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by:
-
Benjamin Herrenschmidt authored
We set an arbitrary max at 1024 since we pre-allocate the actual descriptor arrays and skb arrays to the full size to keep the code a bit simpler and avoid allocation failures in the reset task. Signed-off-by:
-
Benjamin Herrenschmidt authored
Clear stale interrupts on entry, configure FIFO sizes, set FIFO thresholds, configure interrupt mitigation. Signed-off-by:
-
Benjamin Herrenschmidt authored
The helpers just take space but don't provide much value. Simple one line comments are more explanatory. Signed-off-by:
-
Benjamin Herrenschmidt authored
To remove more confusion. This function is about obtaining the initial MAC address at driver probe time. Signed-off-by:
-
Benjamin Herrenschmidt authored
To avoid confusion with the ndo callback and generally be clearer about the purpose of that function Signed-off-by:
-
Benjamin Herrenschmidt authored
So features can be turned on/off via ethtool Signed-off-by:
-
Benjamin Herrenschmidt authored
We found out that HW checksum generation only works from AST2500 onward. This disables it on AST2400 and removes the "no-hw-checksum" properties in the device-trees. The problem we had wasn't related to NC-SI. Also rework the logic testing for that property so it can be used to disable HW checksum generation and checking regardless of whether NC-SI is used or not in case other variants out there need this. Signed-off-by:
-
Benjamin Herrenschmidt authored
We test for aspeed chips to handle a couple of special cases, but we do that by checking the machine type which isn't right. Instead check the actual device compatible property. This also updates the dtsi files for the aspeed SoC to match. Signed-off-by:
-
Benjamin Herrenschmidt authored
The documentation describes NETIF_F_IP_CSUM as deprecated so let's switch to NETIF_F_HW_CSUM and use the helper to handle unhandled protocols. Signed-off-by:
-
David S. Miller authored
Ursula Braun says: ==================== net/smc: patches for net-next here are some patches for net/smc. Most important are improvements for socket closing. ==================== Signed-off-by: -
Ursula Braun authored
smc specifies IB_SEND_INLINE for IB_WR_SEND ib_post_send calls, but provides a mapped buffer to be sent. This is inconsistent, since IB_SEND_INLINE works without mapped buffer. Problem has not been detected in the past, because tests had been limited to Connect X3 cards from Mellanox, whose mlx4 driver just ignored the IB_SEND_INLINE flag. For now, the IB_SEND_INLINE flag is removed. Signed-off-by:
Ursula Braun <ubraun@linux.vnet.ibm.com> Reviewed-by:
Thomas Richter <tmricht@linux.vnet.ibm.com> Signed-off-by:
-
Ursula Braun authored
Make sure sockets never accepted are removed cleanly. Signed-off-by:
-
Ursula Braun authored
unhash is already called in sock_put_work. Remove the second call. Signed-off-by:
-
Ursula Braun authored
State SMC_CLOSED should be reached only, if ConnClosed has been sent to the peer. If ConnClosed is received from the peer, a socket with shutdown SHUT_WR done, switches errorneously to state SMC_CLOSED, which means the peer socket is dangling. The local SMC socket is supposed to switch to state APPFINCLOSEWAIT to make sure smc_close_final() is called during socket close. Signed-off-by:
-
Ursula Braun authored
Several state changes occur during SMC socket closing. Currently state changes triggered locally occur in process context with lock_sock() taken while state changes triggered by peer occur in tasklet context with bh_lock_sock() taken. bh_lock_sock() does not wait till a lock_sock(() task in process context is finished. This may lead to races in socket state transitions resulting in dangling SMC-sockets, or it may lead to duplicate SMC socket freeing. This patch introduces a closing worker to run all state changes under lock_sock(). Signed-off-by:
Dave Jones <davej@codemonkey.org.uk> Signed-off-by:
-
Ursula Braun authored
Wake up reading file descriptors for a closing socket as well, otherwise some socket applications may stall. Signed-off-by:
-
Ursula Braun authored
If peer indicates write_blocked, the cursor state of the received data should be send to the peer immediately (in smc_tx_consumer_update()). Afterwards the write_blocked indicator is cleared. If there is no free slot for another write request, sending is postponed to worker smc_tx_work, and the write_blocked indicator is not cleared. Therefore another clearing check is needed in smc_tx_work(). Signed-off-by:
-
Ursula Braun authored
SMC requires an active ib port on the RoCE device. smc_pnet_find_roce_resource() determines the matching RoCE device port according to the configured PNET table. Do not return the found RoCE device port, if it is not flagged active. Signed-off-by:
-
Ursula Braun authored
The global event handler is created only, if the ib_device has already been used by at least one link group. It is guaranteed that there exists the corresponding entry in the smc_ib_devices list. Get rid of this superfluous check. Signed-off-by:
-
Ursula Braun authored
This patch removes an outdated comment. Signed-off-by:
-
