Sergey Ryazanov [Wed, 4 Mar 2015 02:12:11 +0000 (05:12 +0300)]
ath5k: fix reset race
To prepare for reset ath5k should finish all asynchronous tasks. At
first, it disables the interrupt generation, then it waits for the
interrupt handler and tasklets completion, and then proceeds to the HW
configuration update. But it does not consider that the interrupt
handler or tasklet re-enables the interrupt generation. And we fall in a
situation when ath5k assumes that interrupts are disabled, but it is
not.
This can lead to different consequences, such as reception of the frame,
when we do not expect it. Under certain circumstances, this can lead to
the following warning:
Sergey Ryazanov [Wed, 4 Mar 2015 02:12:10 +0000 (05:12 +0300)]
ath5k: channel change fix
ath5k updates the channel pointer and after that it stops the Rx logic
and apply channel to HW. In case of channel switch, such sequence
creates a small window when a frame, which is received on the old
channel is considered as a frame received on the new one.
The most notable consequence of this situation occurs during the switch
from 2 GHz band (CCK+OFDM) to the 5GHz band (OFDM-only). Frame received
with CCK rate, e.g. beacon received at the 1mbps, causes the following
warning:
The easiest way to reproduce this warning is to run scan with dualband
NIC in noisy environments, when the channel 11 runs multiple APs. In my
tests if the APs num >= 12, the warning appears in the first few
seconds of scanning.
In order to fix this, the Rx disable code moved to a higher level and
placed before the channel pointer update. This is also makes the code a
bit more symmetrical, since we disable and enable the Rx in the same
function.
In fact, at the pointer update time new frames should not appear,
because interrupt generation at this point should already be disabled.
The next patch should address this issue.
CC: Jiri Slaby <jirislaby@gmail.com> CC: Nick Kossifidis <mickflemm@gmail.com> CC: Luis R. Rodriguez <mcgrof@do-not-panic.com> Reported-by: Christophe Prevotaux <cprevotaux@nltinc.com> Tested-by: Christophe Prevotaux <cprevotaux@nltinc.com> Tested-by: Eric Bree <ebree@nltinc.com> Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Simon Horman [Fri, 13 Mar 2015 02:00:58 +0000 (11:00 +0900)]
vxlan: Don't set s_addr in vxlan_create_sock
In the case of AF_INET s_addr was set to INADDR_ANY (0) which which both
symmetric with the AF_INET6 case, where s_addr is not set, and unnecessary
as udp_conf is zeroed out earlier in the same function.
I suspect this change does not have any run-time effect due to compiler
optimisations. But it does make the code a little easier on the/my eyes.
Cc: Tom Herbert <therbert@google.com> Signed-off-by: Simon Horman <simon.horman@netronome.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Reobert Shearman noticed that mpls_egress is failing to verify that
the bytes to be examined are in fact present in the packet before
mpls_egress reads those bytes.
As suggested by David Miller reduce this to a single pskb_may_pull
call so that we don't do unnecessary work in the fast path.
Reported-by: Robert Shearman <rshearma@brocade.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jaeden Amero [Thu, 12 Mar 2015 23:07:54 +0000 (18:07 -0500)]
net/macb: Only adjust tx_clk on link change
The PHY state machine (in drivers/net/phy/phy.c) will unconditionally
call phydev->adjust_link (macb_handle_link_change) when polling in the
PHY_CHANGELINK state. As currently written, macb always ends up
requesting a new tx_clk frequency in macb_handle_link_change. It is a
waste of time to request a new tx_clk frequency if the link state hasn't
changed, as the tx_clk will already be configured properly.
Let's only request a new tx_clk clock frequency when necessary.
Signed-off-by: Jaeden Amero <jaeden.amero@ni.com> Cc: Josh Cartwright <joshc@ni.com> Cc: Soren Brinkmann <soren.brinkmann@xilinx.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Herbert Xu [Fri, 13 Mar 2015 01:54:10 +0000 (12:54 +1100)]
rhashtable: Fix read-side crash during rehash
This patch fixes a typo rhashtable_lookup_compare where we fail
to recompute the hash when looking up the new table. This causes
elements to be missed and potentially a crash during a resize.
Reported-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Thu, 12 Mar 2015 14:28:40 +0000 (15:28 +0100)]
rhashtable: kill ht->shift atomic operations
Commit c0c09bfdc415 ("rhashtable: avoid unnecessary wakeup for worker
queue") changed ht->shift to be atomic, which is actually unnecessary.
Instead of leaving the current shift in the core rhashtable structure,
it can be cached inside the individual bucket tables.
There, it will only be initialized once during a new table allocation
in the shrink/expansion slow path, and from then onward it stays immutable
for the rest of the bucket table liftime.
That allows shift to be non-atomic. The patch also moves hash_rnd
management into the table setup. The rhashtable structure now consumes
3 instead of 4 cachelines.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Ying Xue <ying.xue@windriver.com> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Herbert Xu [Thu, 12 Mar 2015 11:07:49 +0000 (22:07 +1100)]
rhashtable: Fix reader/rehash race
There is a potential race condition between readers and the rehasher.
In particular, the rehasher could have started a rehash while the
reader finishes a scan of the old table but fails to see the new
table pointer.
This patch closes this window by adding smp_wmb/smp_rmb.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Thu, 12 Mar 2015 23:44:04 +0000 (16:44 -0700)]
inet: add TCP_NEW_SYN_RECV state
TCP_SYN_RECV state is currently used by fast open sockets.
Initial TCP requests (the pseudo sockets created when a SYN is received)
are not yet associated to a state. They are attached to their parent,
and the parent is in TCP_LISTEN state.
This commit adds TCP_NEW_SYN_RECV state, so that we can convert
TCP stack to a different schem gradually.
This state is not exported to user space.
Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
I forgot to update dccp_v6_conn_request() & cookie_v6_check().
They both need to set ireq->ireq_net and ireq->ir_cookie
Lets clear ireq->ir_cookie in inet_reqsk_alloc()
Signed-off-by: Eric Dumazet <edumazet@google.com> Fixes: 33cf7c90fe2f ("net: add real socket cookies") Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Thu, 12 Mar 2015 19:03:12 +0000 (20:03 +0100)]
cls_bpf: do eBPF invocation under non-bh RCU lock variant for maps
Currently, it is possible in cls_bpf to access eBPF maps only under
rcu_read_lock_bh() variants: while on ingress side, that is, handle_ing(),
the classifier would be called from __netif_receive_skb_core() under
rcu_read_lock(); on egress side, however, it's rcu_read_lock_bh() via
__dev_queue_xmit().
This rcu/rcu_bh mix doesn't work together with eBPF maps as they require
soley to be called under rcu_read_lock(). eBPF maps could also be shared
among various other eBPF programs (possibly even with other eBPF program
types, f.e. tracing) and user space processes, so any context is assumed.
Therefore, a possible fix for cls_bpf is to wrap/nest eBPF program
invocation under non-bh RCU lock variant.
Fixes: e2e9b6541dd4 ("cls_bpf: add initial eBPF support for programmable classifiers") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 12 Mar 2015 22:26:58 +0000 (18:26 -0400)]
Merge branch 'fib_trie_table_merge_fixes'
Alexander Duyck says:
====================
fib_trie: Minor fixes for table merge
This patch set addresses two issues reported with the tables merged, the
first is a NULL pointer dereference, and the other is to remove a WARN_ON
and set the ordering for aliases from different tables with the same slen
values.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 12 Mar 2015 21:46:29 +0000 (14:46 -0700)]
fib_trie: Provide a deterministic order for fib_alias w/ tables merged
This change makes it so that we should always have a deterministic ordering
for the main and local aliases within the merged table when two leaves
overlap.
So for example if we have a leaf with a key of 192.168.254.0. If we
previously added two aliases with a prefix length of 24 from both local and
main the first entry would be first and the second would be second. When I
was coding this I had added a WARN_ON should such a situation occur as I
wasn't sure how likely it would be. However this WARN_ON has been
triggered so this is something that should be addressed.
With this patch the ordering of the aliases is as follows. First they are
sorted on prefix length, then on their table ID, then tos, and finally
priority. This way what we end up doing is essentially interleaving the
two tables on what used to be leaf_info structure boundaries.
Fixes: 0ddcf43d5 ("ipv4: FIB Local/MAIN table collapse") Reported-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Thu, 12 Mar 2015 21:46:23 +0000 (14:46 -0700)]
fib_trie: Avoid NULL pointer if local table is not allocated
The function fib_unmerge assumed the local table had already been
allocated. If that is not the case however when custom rules are applied
then this can result in a NULL pointer dereference.
In order to prevent this we must check the value of the local table pointer
and if it is NULL simply return 0 as there is no local table to separate
from the main.
Fixes: 0ddcf43d5 ("ipv4: FIB Local/MAIN table collapse") Reported-by: Madhu Challa <challa@noironetworks.com> Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Thu, 12 Mar 2015 16:21:42 +0000 (17:21 +0100)]
ebpf: verifier: check that call reg with ARG_ANYTHING is initialized
I noticed that a helper function with argument type ARG_ANYTHING does
not need to have an initialized value (register).
This can worst case lead to unintented stack memory leakage in future
helper functions if they are not carefully designed, or unintended
application behaviour in case the application developer was not careful
enough to match a correct helper function signature in the API.
The underlying issue is that ARG_ANYTHING should actually be split
into two different semantics:
1) ARG_DONTCARE for function arguments that the helper function
does not care about (in other words: the default for unused
function arguments), and
2) ARG_ANYTHING that is an argument actually being used by a
helper function and *guaranteed* to be an initialized register.
The current risk is low: ARG_ANYTHING is only used for the 'flags'
argument (r4) in bpf_map_update_elem() that internally does strict
checking.
Fixes: 17a5267067f3 ("bpf: verifier (add verifier core)") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@plumgrid.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 12 Mar 2015 18:39:44 +0000 (14:39 -0400)]
Merge branch 'possible_net_t'
Eric W. Biederman says:
====================
Introduce possible_net_t
The current usage of write_pnet and read_pnet is a little laborious and
error prone as you only notice if you failed to include them if are
compiling with network namespaces enabled.
possible_net_t remedies that by using a type that is 0 bytes when
network namespaces are disabled and can only be read and written to with
read_pnet and write_pnet.
Aka less work and safer for the same effect.
I kill hold_net and release_net first as are they are haven't been used
since 2008 and are noise at the points where write_pnet and read_pnet
are used.
I have folded in Eric Dumazets suggestions to improve the killing of
hold_net and release net. And respon. I had to respin anyway as
there was enough changes elsewhere in the tree the previous version
of these patches did not quite apply cleanly.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Having to say
> #ifdef CONFIG_NET_NS
> struct net *net;
> #endif
in structures is a little bit wordy and a little bit error prone.
Instead it is possible to say:
> typedef struct {
> #ifdef CONFIG_NET_NS
> struct net *net;
> #endif
> } possible_net_t;
And then in a header say:
> possible_net_t net;
Which is cleaner and easier to use and easier to test, as the
possible_net_t is always there no matter what the compile options.
Further this allows read_pnet and write_pnet to be functions in all
cases which is better at catching typos.
This change adds possible_net_t, updates the definitions of read_pnet
and write_pnet, updates optional struct net * variables that
write_pnet uses on to have the type possible_net_t, and finally fixes
up the b0rked users of read_pnet and write_pnet.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
hold_net and release_net were an idea that turned out to be useless.
The code has been disabled since 2008. Kill the code it is long past due.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Herbert Xu [Thu, 12 Mar 2015 03:49:38 +0000 (14:49 +1100)]
rhashtable: Move masking back into key_hashfn
This patch reverts commit c88455ce50ae4224d84960ce2baa53e61580df27
("rhashtable: key_hashfn() must return full hash value") because
the only user of it always masks the hash value.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Acked-by: Thomas Graf <tgraf@suug.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
The return value in iwl_mvm_get_wakeup_status() is a bit unclear in
that it's not obvious that we don't leak fw_status in some cases.
Use fw_status directly with ERR_PTR() and return only it, that way
the compiler has a chance of proving that it's uninitialized (if it
ever is due to new changes.)
Additionally, this removes a smatch warning since smatch couldn't
figure out that fw_status can't, in fact, leak here.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Luciano Coelho [Thu, 12 Mar 2015 07:25:15 +0000 (09:25 +0200)]
iwlwifi: mvm: don't double unlock the mutex in __iwl_mvm_resume()
When IWLWIFI_DEBUGFS is not set, we should not unlock the mutex after
calling iwl_mvm_query_wakeup_reasons(), because this function unlocks
it already. Move the goto out_iterate outside the #ifdef.
Johannes Berg [Wed, 11 Mar 2015 19:27:06 +0000 (20:27 +0100)]
iwlwifi: mvm: clarify time event end handling
The code here is a little confusing, the iwl_mvm_te_check_disconnect()
will check that the interface is a station, but going into it after
already having processed the time even end for P2P seems strange at
first look.
Put a switch statement there to distinguish the interface types and
make this more readable.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Eran Harary [Sun, 8 Feb 2015 11:58:50 +0000 (13:58 +0200)]
iwlwifi: mvm: Always enable the smart FIFO
We previously enabled the smart FIFO (SF) in BSS only after
association.
This cause interrupt latency on P2P on certain devices.
Change the working model to enable the SF all the time and
play with the timeout values based on the association state.
This change was not tested on older firwmares, so make it
happen only on -13.ucode and up.
Signed-off-by: Eran Harary <eran.harary@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
The firmware was not using the new API, so we don't need to
differentiate between the different stages of this new API.
The main difference here is that most of the hard coded
values are not sent through the command anymore.
The newer devices will enable a new register for this
(DEVICE_SET_NMI_8000B_REG), but the interrupt handler
isn't wired yet.
Keep the old register for now.
iwlwifi: mvm: freeze the non-shared queues when a station goes to sleep
When a station goes to sleep, we can't transmit any frame
to it. This means that until that station will wake up, a
queue that is dedicated to this station won't progress at
all. Take this into account when monitoring stuck queues
and don't account for the time the station was asleep.
This allows to mask false positives where the queues are
stuck not because of a bug, but because of the station
being asleep.
iwlwifi: pcie: allow the op_mode to freeze the stuck queue timer
This allows the op_mode to let the transport know that a
queue is currently frozen and that its timer should be
stopped.
When the queue is unfrozen, its timer should be set to
expire after the remainder of the timeout has elapsed.
This can be used when stations go to sleep. When a station
goes to sleep, the op_mode can freeze the timer so that the
queue will never be considered as stuck. When the station
wakes up, the queue will be unfrozen.
This is meant to avoid false positives that would happen if
a buggy station goes to sleep for a very long time. In case
we have a dedicated queue for this station (BA agreement)
and it goes to sleep for a very long time, the queue would
rightfully be stopped during all that time. In this case,
the stuck queue timer could fire and that would be a false
positive.
Arik Nemtsov [Wed, 4 Mar 2015 13:08:00 +0000 (15:08 +0200)]
iwlwifi: mvm: don't init MCC during CT-kill
RTNL is not taken during CT-kill so regulatory APIs cannot be invoked.
That's fine, since the HW is only brought up to check the temperature
during CT-kill. We don't expect Tx or scanning.
Eran Harary [Tue, 3 Mar 2015 11:53:28 +0000 (13:53 +0200)]
iwlwifi: trans: Take ownership on secure machine before FW load
When we load the firmware for the 8000 B step device, it'll
verify its signature. In the current version of the
hardware, there can be a race between the WiFi firmware
being loaded and the Bluetooth firmware being loaded.
Check that WiFi is authenticated, if not, take ownership
on the authentication machine to make sure that the WiFi
firmware will be authenticated.
Signed-off-by: Eran Harary <eran.harary@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Arik Nemtsov [Sun, 22 Feb 2015 17:15:04 +0000 (19:15 +0200)]
iwlwifi: mvm: reflect TDLS pm state in mvmvif->pm_enabled
When entering D0i3, the MVM mutex cannot be grabbed. This interferes
with the calculation of the number of connected TDLS stations during
the setup of the power cmd.
The goal is to disable power saving for all vifs while any TDLS station
is connected. For this purpose it is enough to keep the pm_enabled
member of all mvmvifs as false. An update of the power state already
occurs when a TDLS station is added/removed, so the values are correctly
updated.
We don't need to acquire MAC access for each access, it
makes much more sense to keep the MAC access. This speeds
up the Tx DMA stop flow significantly.
Moreover, if one channel can't be stopped, stop the others
but don't poll for them to avoid being stuck there for a
long time.
This solves a situation in which we were stuck in that flow
for way too long with a spinlock held which led to a kernel
panic.
Reviewed-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Eran Harary [Wed, 25 Feb 2015 12:24:51 +0000 (14:24 +0200)]
iwlwifi: mvm: support family 8000 B2/C steps
In-order to recognize newer step of the device, the driver
must read the chip_version_id from the AUX bus MISC address
space. This will determine what firmware file will be
loaded.
Signed-off-by: Eran Harary <eran.harary@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Jonathan Doron [Thu, 27 Nov 2014 14:55:25 +0000 (16:55 +0200)]
iwlwifi: mvm: set LAR MCC on D3/D0 transitions
When moving to the D3 FW give it the valid MCC from the D0 FW. When
returning from D3 to D0, query the D3 FW for the latest MCC, as
it might have changed internally. This MCC will be replayed to the D0 FW
when it boots.
Signed-off-by: Jonathan Doron <jonathanx.doron@intel.com> Reviewed-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Jonathan Doron [Thu, 27 Nov 2014 14:57:55 +0000 (16:57 +0200)]
iwlwifi: mvm: support LAR updates from BIOS
When booting the card, check for a dedicated regulatory ACPI entry. If
such exists, read it and give the information to FW with the appropriate
source.
Eran Harary [Sun, 8 Feb 2015 09:41:43 +0000 (11:41 +0200)]
iwlwifi: mvm: take the MAC address from HW registers
For some configurations, the driver should get the MAC
address from the hardware registers and not from the
regular locations. Since the parsing of the MAC address
is the same regardless of its source, continue the regular
code path (parsing) after we read the registers.
Signed-off-by: Eran Harary <eran.harary@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Chub (Communication Hub, CommsHUB) is a HW component that connects to the cellular
and connectivity cores that gets updates of mcc changes, and then notifies the FW
directly of any mcc change.
The ucode notifies the driver (via this command) that it should ask for an mcc update,
and the driver sends the ucode the update mcc command to set the updated regulatory info.
Signed-off-by: Matti Gottlieb <matti.gottlieb@intel.com> Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Arik Nemtsov [Thu, 30 Oct 2014 13:12:39 +0000 (15:12 +0200)]
iwlwifi: ignore IBSS flag as regulatory NO-IR indication
According to updated regulatory guidelines, the ACTIVE bit in the NVM
also allows ibss activity on the channel. The IBSS NVM bit is not updated
when LAR is active and is deprecated. Using this bit for NO-IR incorrectly
causes all 5Ghz channels to be marked as passive.
Arik Nemtsov [Sun, 23 Mar 2014 14:18:40 +0000 (16:18 +0200)]
iwlwifi: mvm: consider LAR support during NVM parse
Register to cfg80211 with all channels enabled when LAR is supported.
Appropriate channels will later be disabled when a specific regulatory
domain is defined.
Arik Nemtsov [Wed, 5 Mar 2014 10:19:10 +0000 (12:19 +0200)]
iwlwifi: create regdomain from mcc_update_cmd response
Parse the NVM channel data and create a regulatory domain with a rule
for every 20Mhz channel. Use the AUTO_BW flag so the regulatory core
can unify single-channel rules into ranges.
Arik Nemtsov [Tue, 4 Mar 2014 17:58:46 +0000 (19:58 +0200)]
iwlwifi: mvm: init country code on init/recovery
During init queue a regulatory update to retrieve the default
regulatory settings from FW. If we're during recovery, only replay the
current country code to FW, if it exists.
iwlwifi: fix max_ht_ampdu_exponent for older devices
The commit below didn't update the max_ht_ampdu_exponent
for the devices listed in iwl-[1-6]000.c which, in result,
became 0 instead of 8K. This reduced the size of the Rx
AMPDU from 64K to 8K which had an impact in the Rx
throughput. One user reported that because of this, his
downstream throughput droppped by a half.
Johannes Berg [Tue, 10 Mar 2015 09:47:57 +0000 (10:47 +0100)]
iwlwifi: mvm: disconnect if CSA time event fails scheduling
If this situation ever happens, the mac80211 state machine gets
confused because it never clears csa_active. There was a separate
bug that lead to this happening with a working connection, but it
isn't very robust to try to keep the connection up in this case.
When removing the time event the CSA essentially procedure stops,
so the safest thing to do is to disconnect in this case.
iwlwifi: dvm: drop VO packets when mac80211 tells us to
mac80211 now informs the driver when to drop the packets
upon flush(). This will happen before disconnecting, or
before we shut down the interface. We can now rely on this
to drop all the packets including the VO queues.
When mac80211 sets drop to false, wait for all the queues
to be empty.
Herbert Xu [Thu, 12 Mar 2015 03:47:13 +0000 (14:47 +1100)]
rhashtable: Add annotation to nested lock
Commit aa34a6cb0478842452bac58edb50d3ef9e178c92 ("rhashtable:
Add arbitrary rehash function") killed the annotation on the
nested lock which leads to bitching from lockdep.
Reported-by: Fengguang Wu <fengguang.wu@intel.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Thu, 12 Mar 2015 03:27:52 +0000 (20:27 -0700)]
net: fix CONFIG_NET_NS=n compilation
I forgot to use write_pnet() in three locations.
Signed-off-by: Eric Dumazet <edumazet@google.com> Fixes: 33cf7c90fe2f9 ("net: add real socket cookies") Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Simon Horman [Thu, 12 Mar 2015 02:00:10 +0000 (11:00 +0900)]
vxlan: Correct path typo in comment
Flags are used in the return path rather than the return patch.
Fixes: af33c1adae1e ("vxlan: Eliminate dependency on UDP socket in transmit path") Signed-off-by: Simon Horman <simon.horman@netronome.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Thu, 12 Mar 2015 01:53:14 +0000 (18:53 -0700)]
net: add real socket cookies
A long standing problem in netlink socket dumps is the use
of kernel socket addresses as cookies.
1) It is a security concern.
2) Sockets can be reused quite quickly, so there is
no guarantee a cookie is used once and identify
a flow.
3) request sock, establish sock, and timewait socks
for a given flow have different cookies.
Part of our effort to bring better TCP statistics requires
to switch to a different allocator.
In this patch, I chose to use a per network namespace 64bit generator,
and to use it only in the case a socket needs to be dumped to netlink.
(This might be refined later if needed)
Note that I tried to carry cookies from request sock, to establish sock,
then timewait sockets.
Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Eric Salo <salo@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Wed, 11 Mar 2015 23:36:08 +0000 (16:36 -0700)]
fib_trie: Only display main table in /proc/net/route
When we merged the tries for local and main I had overlooked the iterator
for /proc/net/route. As a result it was outputting both local and main
when the two tries were merged.
This patch resolves that by only providing output for aliases that are
actually in the main trie. As a result we should go back to the original
behavior which I assume will be necessary to maintain legacy support.
Fixes: 0ddcf43d5 ("ipv4: FIB Local/MAIN table collapse") Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 11 Mar 2015 21:56:39 +0000 (17:56 -0400)]
Merge branch 'dsa_phy_divert'
Florian Fainelli says:
====================
net: dsa: support PHY reads/writes diversion
This patch series completes the PHY reads/writes diversion when we need to use
the slave MII bus provided by DSA and the underlying switch drivers to
implement the real PHY reads and writes. This is particularly useful when they
are conflicting MDIO bus addresses as in the case of multiple Broadcom switches
connected to each other (internal and external, or just external).
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Fainelli [Tue, 10 Mar 2015 23:57:13 +0000 (16:57 -0700)]
net: dsa: fully divert PHY reads/writes if requested
In case a PHY is found via Device Tree, and is also flagged by the
switch driver as needing indirect reads/writes using the switch driver
implemented MDIO bus, make sure that we bind this PHY to the slave MII
bus in order for this to happen.
Without this, we would succeed in having the PHY driver probe()'s
function to use slave MII bus read/write functions, because this is done
during dsa_slave_mii_init(), but past that point, the PHY driver would
not go through these diverted reads and writes.
Fixes: 0d8bcdd383b88 ("net: dsa: allow for more complex PHY setups") Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Fainelli [Tue, 10 Mar 2015 23:57:12 +0000 (16:57 -0700)]
net: dsa: move PHY setup on DSA MII bus to its own function
In preparation for dealing with indirect reads and writes towards
certain PHY devices, move the code which deals with binding the PHY
device to the slave MII bus created by DSA to its own function:
dsa_slave_phy_connect().
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Fainelli [Tue, 10 Mar 2015 23:57:11 +0000 (16:57 -0700)]
of: mdio: export of_mdio_parse_addr
Export of_mdio_parse_addr() which allows parsing a given Ethernet PHY
node MDIO address, verify it is within the allowed range, and return
its value. This is going to be useful for the DSA code which needs to
deal with multiple layers of MDIO buses.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Petri Gynther [Tue, 10 Mar 2015 22:55:00 +0000 (15:55 -0700)]
net: bcmgenet: collect Rx discarded packet count
Bits 31:16 of RDMA_PROD_INDEX contain Rx discarded packet count, which
are the Rx packets that had to be dropped by MAC hardware since there
was no room on the Rx queue. Add code to collect this information into
the netdev stats.
Signed-off-by: Petri Gynther <pgynther@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexander Duyck [Wed, 11 Mar 2015 21:02:16 +0000 (14:02 -0700)]
fib_trie: Fix uninitialized variable warning
The 0-day kernel test infrastructure reported a use of uninitialized
variable warning for local_table due to the fact that the local and main
allocations had been swapped from the original setup. This change corrects
that by making it so that we free the main table if the local table
allocation fails.
Fixes: 0ddcf43d5 ("ipv4: FIB Local/MAIN table collapse") Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sabrina Dubroca [Tue, 10 Mar 2015 20:03:54 +0000 (21:03 +0100)]
fib_trie: call fib_table_flush_external under RTNL
Move rtnl_lock() before the call to fib4_rules_exit so that
fib_table_flush_external is called under RTNL.
Fixes: 104616e74e0b ("switchdev: don't support custom ip rules, for now") Signed-off-by: Sabrina Dubroca <sd@queasysnail.net> Acked-by: Alexander Duyck <alexander.h.duyck@redhat.com> Reviewed-by: Jiri Pirko <jiri@resnulli.us> Signed-off-by: David S. Miller <davem@davemloft.net>
Robert Shearman [Tue, 10 Mar 2015 16:37:59 +0000 (16:37 +0000)]
mpls: Allow mpls_gso and mpls_router to be built as modules
CONFIG_MPLS=m doesn't result in a kernel module being built because it
applies to the net/mpls directory, rather than to .o files.
So revert the MPLS menuitem to being a boolean and make MPLS_GSO and
MPLS_ROUTING tristates to allow mpls_gso and mpls_router modules to be
produced as desired.
Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Robert Shearman <rshearma@brocade.com> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / linux-beck.git / log