Dumped buffer descriptors showed that eTSEC's length/truncation
logic sometimes passes oversized packets, i.e. for the above ICMP
packet the following two buffer descriptors may become ready:
status=1400 length=1536
status=1800 length=1541
So, it seems that gianfar actually receives the whole big frame,
and it tries to place the packet into two BDs. This situation
confuses the driver, and so the skb_put() sanity check fails.
This patch fixes the issue by adding an appropriate check, i.e.
the driver should not try to process frames with buffer
descriptor's length over rx_buffer_size (i.e. maxfrm and mrblr).
Note that sometimes eTSEC works correctly, i.e. in the second
(last) buffer descriptor bits 'truncated' and 'crcerr' are set,
and so there's no oops. Though I couldn't find any logic when
it works correctly and when not.
Signed-off-by: Anton Vorontsov <avorontsov@mvista.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Port reset operations and memory add/remove operations need to
be serialized to avoid a kernel deadlock. The deadlock is caused
by calling the napi_disable() function twice.
Therefore we have to employ the dlpar_mem_lock in the ehea_reset_port
function as well
Signed-off-by: Jan-Bernd Themann <themann@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
In the eHEA poll function an rmb() is required. Without that some packets
on the receive queue are not seen and thus delayed until the next interrupt
is handled for the same receive queue.
Signed-off-by: Jan-Bernd Themann <themann@de.ibm.com> Signed-off-by: Jay Vosburgh <fubar@us.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Don Skidmore [Tue, 15 Jun 2010 09:23:17 +0000 (09:23 +0000)]
ixgbe: add comment on SFP+ ID for Active DA
These comments were forgotten in the initial patch to add this
functionality. This patch corrects that.
Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Tom Hughes [Tue, 15 Jun 2010 22:24:28 +0000 (22:24 +0000)]
Clear IFF_XMIT_DST_RELEASE for teql interfaces
https://bugzilla.kernel.org/show_bug.cgi?id=16183
The sch_teql module, which can be used to load balance over a set of
underlying interfaces, stopped working after 2.6.30 and has been
broken in all kernels since then for any underlying interface which
requires the addition of link level headers.
The problem is that the transmit routine relies on being able to
access the destination address in the skb in order to do address
resolution once it has decided which underlying interface it is going
to transmit through.
In 2.6.31 the IFF_XMIT_DST_RELEASE flag was introduced, and set by
default for all interfaces, which causes the destination address to be
released before the transmit routine for the interface is called.
The solution is to clear that flag for teql interfaces.
Signed-off-by: Tom Hughes <tom@compton.nu> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Manfred Rudigier [Fri, 11 Jun 2010 01:49:05 +0000 (01:49 +0000)]
gianfar: Fix setup of RX time stamping
Previously the RCTRL_TS_ENABLE bit was set unconditionally. However, if
the RCTRL_TS_ENABLE is set without TMR_CTRL[TE], the driver does not work
properly on some boards (Anton had problems with the MPC8313ERDB and
MPC8568EMDS).
With this patch the bit will only be set if requested from user space
with the SIOCSHWTSTAMP ioctl command, meaning that time stamping is
disabled during normal operation. Users who are not interested in time
stamps will not experience problems with buggy CPU revisions or
performance drops any more.
The setting of TMR_CTRL[TE] is still up to the user. This is considered
safe because users wanting HW timestamps must initialize the eTSEC clock
first anyway, e.g. with the recently submitted PTP clock driver.
Signed-off-by: Manfred Rudigier <manfred.rudigier@omicron.at> Reviewed-by: Anton Vorontsov <cbouatmailru@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Herbert Xu [Wed, 16 Jun 2010 04:43:07 +0000 (21:43 -0700)]
bridge: Fix OOM crash in deliver_clone
The bridge multicast patches introduced an OOM crash in the forward
path, when deliver_clone fails to clone the skb.
Reported-by: Mark Wagner <mwagner@redhat.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com> Reported-by: Mihai Harpau <mishu@piatafinanciara.ro> Cc: stable@kernel.org Acked-by: Reinette Chatre <reinette.chatre@intel.com>
Prarit Bhargava [Thu, 10 Jun 2010 12:08:42 +0000 (08:08 -0400)]
libertas_tf: Fix warning in lbtf_rx for stats struct
Fixes linux-2.6 warning:
drivers/net/wireless/libertas_tf/main.c: In function 'lbtf_rx':
drivers/net/wireless/libertas_tf/main.c:578: warning: 'stats.antenna' is used uninitialized in this function
drivers/net/wireless/libertas_tf/main.c:578: warning: 'stats.mactime' is used uninitialized in this function
stats struct needs to be set to 0 before use.
Signed-off-by: Prarit Bhargava <prarit@redhat.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Shanyu Zhao [Wed, 2 Jun 2010 00:13:58 +0000 (17:13 -0700)]
iwlagn: verify flow id in compressed BA packet
The flow id (scd_flow) in a compressed BA packet should match the txq_id
of the queue from which the aggregated packets were sent. However, in
some hardware like the 1000 series, sometimes the flow id is 0 for the
txq_id (10 to 19). This can cause the annoying message:
[ 2213.306191] iwlagn 0000:01:00.0: Received BA when not expected
[ 2213.310178] iwlagn 0000:01:00.0: Read index for DMA queue txq id (0),
index 5, is out of range [0-256] 7 7.
And even worse, if agg->wait_for_ba is true when the bad BA is arriving,
this can cause system hang due to NULL pointer dereference because the
code is operating in a wrong tx queue!
Reinette Chatre [Fri, 28 May 2010 16:28:39 +0000 (09:28 -0700)]
iwlwifi: serialize station management actions
We are seeing some race conditions between incoming station management
requests (station add/remove) and the internal unassoc RXON command that
modifies station table. Modify these flows to require the mutex to be held
and thus serializing them.
This fixes http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2207
introduced a feature where LRO (called RSC on the hardware) was disabled
automatically when setting rx-usecs to 0 via ethtool. Some might not
like the fact that LRO was disabled automatically, but I'm fine with
that. What I don't like is that LRO/RSC is automatically enabled when
rx-usecs is set >0 via ethtool.
This would certainly be a problem if the device was used for forwarding
and it was determined that the low latency wasn't needed after the
device was already forwarding. I played around with saving the state of
LRO in the driver, but it just didn't seem worthwhile and would require
a small change to dev_disable_lro() that I did not like.
This patch simply leaves LRO disabled when setting rx-usecs >0 and
requires that the user enable it again. An extra informational message
will also now appear in the log so users can understand why LRO isn't
being enabled as they expect.
Inconsistency of LRO setting first noticed by Stanislaw Gruszka.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net> CC: Stanislaw Gruszka <sgruszka@redhat.com> CC: stable@kernel.org Tested-by: Stephen Ko <stephen.s.ko@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
If not, then this patch fixes a message logging defect
introduced by converting a printk without KERN_<level>
to e_info.
This also reduces text by about 200 bytes.
Signed-off-by: Joe Perches <joe@perches.com> Tested-by: Emil Tantilov <emil.s.tantilov@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Don Skidmore [Fri, 11 Jun 2010 13:20:29 +0000 (13:20 +0000)]
ixgbe: fix for race with 8259(8|9) during shutdown
There is a small window where the watchdog could be running as the
interface is brought down on a NIC with two ports wired back to back.
If ixgbe_update_status is then called can lead to a panic. This patch
allows the update to bail if we are in that condition.
This issue was orignally reported and fix proposed by Akihiko Saitou.
CC: Akihiko Saitou <asaitou@users.sourceforge.net> Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Sun, 13 Jun 2010 10:36:30 +0000 (10:36 +0000)]
net: fix deliver_no_wcard regression on loopback device
deliver_no_wcard is not being set in skb_copy_header.
In the skb_cloned case it is not being cleared and
may cause the skb to be dropped when the loopback device
pushes it back up the stack.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Tested-by: Markus Trippelsdorf <markus@trippelsdorf.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Randy Dunlap [Tue, 8 Jun 2010 07:00:20 +0000 (07:00 +0000)]
enic: fix pci_alloc_consistent argument
Fix build warning on i386 (32-bit) with 32-bit dma_addr_t:
drivers/net/enic/vnic_dev.c: In function 'vnic_dev_init_prov':
drivers/net/enic/vnic_dev.c:716: warning: passing argument 3 of 'pci_alloc_consistent' from incompatible pointer type
include/asm-generic/pci-dma-compat.h:16: note: expected 'dma_addr_t *' but argument is of type 'u64 *'
Now builds without warnings on i386 and on x86_64.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Scott Feldman <scofeldm@cisco.com>
Cc: Vasanthy Kolluri <vkolluri@cisco.com>
Cc: Roopa Prabhu <roprabhu@cisco.com> Acked-by: Scott Feldman <scofeldm@cisco.com>
Daniel Turull [Fri, 11 Jun 2010 06:08:11 +0000 (23:08 -0700)]
pktgen: Fix accuracy of inter-packet delay.
This patch correct a bug in the delay of pktgen.
It makes sure the inter-packet interval is accurate.
Signed-off-by: Daniel Turull <daniel.turull@gmail.com> Signed-off-by: Robert Olsson <robert.olsson@its.uu.se> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Thu, 3 Jun 2010 09:30:11 +0000 (09:30 +0000)]
net: deliver skbs on inactive slaves to exact matches
Currently, the accelerated receive path for VLAN's will
drop packets if the real device is an inactive slave and
is not one of the special pkts tested for in
skb_bond_should_drop(). This behavior is different then
the non-accelerated path and for pkts over a bonded vlan.
For example,
vlanx -> bond0 -> ethx
will be dropped in the vlan path and not delivered to any
packet handlers at all. However,
bond0 -> vlanx -> ethx
and
bond0 -> ethx
will be delivered to handlers that match the exact dev,
because the VLAN path checks the real_dev which is not a
slave and netif_recv_skb() doesn't drop frames but only
delivers them to exact matches.
This patch adds a sk_buff flag which is used for tagging
skbs that would previously been dropped and allows the
skb to continue to skb_netif_recv(). Here we add
logic to check for the deliver_no_wcard flag and if it
is set only deliver to handlers that match exactly. This
makes both paths above consistent and gives pkt handlers
a way to identify skbs that come from inactive slaves.
Without this patch in some configurations skbs will be
delivered to handlers with exact matches and in others
be dropped out right in the vlan path.
I have tested the following 4 configurations in failover modes
and load balancing modes.
# bond0 -> ethx
# vlanx -> bond0 -> ethx
# bond0 -> vlanx -> ethx
# bond0 -> ethx
|
vlanx -> --
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Timo Teräs [Thu, 10 Jun 2010 00:31:48 +0000 (17:31 -0700)]
r8169: fix mdio_read and update mdio_write according to hw specs
Realtek confirmed that a 20us delay is needed after mdio_read and
mdio_write operations. Reduce the delay in mdio_write, and add it
to mdio_read too. Also add a comment that the 20us is from hw specs.
Signed-off-by: Timo Teräs <timo.teras@iki.fi> Acked-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Anton Vorontsov [Wed, 9 Jun 2010 23:27:08 +0000 (16:27 -0700)]
gianfar: Revive the driver for eTSEC devices (disable timestamping)
Since commit a1e46e3d3a4d4f56dd3db720417404d2aaf0b566 ("gianfar: Add
hardware RX timestamping support"), the driver no longer works on
at least MPC8313ERDB and MPC8568EMDS boards (and possibly much more
boards as well).
That's how MPC8313 Reference Manual describes RCTRL_TS_ENABLE bit:
Timestamp incoming packets as padding bytes. PAL field is set
to 8 if the PAL field is programmed to less than 8. Must be set
to zero if TMR_CTRL[TE]=0.
I see that the commit above sets this bit, but it doesn't handle
TMR_CTRL. Manfred probably had this bit set by the firmware for
his boards. But obviously this isn't true for all boards in the
wild.
Also, I recall that Freescale BSPs were explicitly disabling the
timestamping because of a performance drop.
For now, the best way to deal with this is just disable the
timestamping, and later we can discuss proper device tree bindings
and implement enabling this feature via some property.
Signed-off-by: Anton Vorontsov <avorontsov@mvista.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter [Mon, 7 Jun 2010 04:51:58 +0000 (04:51 +0000)]
caif: fix a couple range checks
The extra ! character means that these conditions are always false.
Signed-off-by: Dan Carpenter <error27@gmail.com> Acked-by: Sjur Braendeland <sjur.brandeland@stericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Richard Cochran [Mon, 7 Jun 2010 05:39:32 +0000 (05:39 +0000)]
phylib: Add support for the LXT973 phy.
This patch implements a work around for Erratum 5, "3.3 V Fiber Speed
Selection." If the hardware wiring does not respect this erratum, then
fiber optic mode will not work properly.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at> Signed-off-by: David S. Miller <davem@davemloft.net>
Tim Gardner [Tue, 8 Jun 2010 23:51:27 +0000 (17:51 -0600)]
net: Print num_rx_queues imbalance warning only when there are allocated queues
BugLink: http://bugs.launchpad.net/bugs/591416
There are a number of network drivers (bridge, bonding, etc) that are not yet
receive multi-queue enabled and use alloc_netdev(), so don't print a
num_rx_queues imbalance warning in that case.
Also, only print the warning once for those drivers that _are_ multi-queue
enabled.
Signed-off-by: Tim Gardner <tim.gardner@canonical.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Sven Wegener [Wed, 9 Jun 2010 14:10:57 +0000 (16:10 +0200)]
ipvs: Add missing locking during connection table hashing and unhashing
The code that hashes and unhashes connections from the connection table
is missing locking of the connection being modified, which opens up a
race condition and results in memory corruption when this race condition
is hit.
Here is what happens in pretty verbose form:
CPU 0 CPU 1
------------ ------------
An active connection is terminated and
we schedule ip_vs_conn_expire() on this
CPU to expire this connection.
IRQ assignment is changed to this CPU,
but the expire timer stays scheduled on
the other CPU.
New connection from same ip:port comes
in right before the timer expires, we
find the inactive connection in our
connection table and get a reference to
it. We proper lock the connection in
tcp_state_transition() and read the
connection flags in set_tcp_state().
ip_vs_conn_expire() gets called, we
unhash the connection from our
connection table and remove the hashed
flag in ip_vs_conn_unhash(), without
proper locking!
While still holding proper locks we
write the connection flags in
set_tcp_state() and this sets the hashed
flag again.
ip_vs_conn_expire() fails to expire the
connection, because the other CPU has
incremented the reference count. We try
to re-insert the connection into our
connection table, but this fails in
ip_vs_conn_hash(), because the hashed
flag has been set by the other CPU. We
re-schedule execution of
ip_vs_conn_expire(). Now this connection
has the hashed flag set, but isn't
actually hashed in our connection table
and has a dangling list_head.
We drop the reference we held on the
connection and schedule the expire timer
for timeouting the connection on this
CPU. Further packets won't be able to
find this connection in our connection
table.
ip_vs_conn_expire() gets called again,
we think it's already hashed, but the
list_head is dangling and while removing
the connection from our connection table
we write to the memory location where
this list_head points to.
The result will probably be a kernel oops at some other point in time.
This race condition is pretty subtle, but it can be triggered remotely.
It needs the IRQ assignment change or another circumstance where packets
coming from the same ip:port for the same service are being processed on
different CPUs. And it involves hitting the exact time at which
ip_vs_conn_expire() gets called. It can be avoided by making sure that
all packets from one connection are always processed on the same CPU and
can be made harder to exploit by changing the connection timeouts to
some custom values.
Signed-off-by: Sven Wegener <sven.wegener@stealer.net> Cc: stable@kernel.org Acked-by: Simon Horman <horms@verge.net.au> Signed-off-by: Patrick McHardy <kaber@trash.net>
Johannes Berg [Mon, 7 Jun 2010 19:50:07 +0000 (21:50 +0200)]
mac80211: fix deauth before assoc
When we receive a deauthentication frame before
having successfully associated, we neither print
a message nor abort assocation. The former makes
it hard to debug, while the latter later causes
a warning in cfg80211 when, as will typically be
the case, association timed out.
This warning was reported by many, e.g. in
https://bugzilla.kernel.org/show_bug.cgi?id=15981,
but I couldn't initially pinpoint it. I verified
the fix by hacking hostapd to send a deauth frame
instead of an association response.
Cc: stable@kernel.org Signed-off-by: Johannes Berg <johannes.berg@intel.com> Tested-by: Miles Lane <miles.lane@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Johannes Berg [Mon, 7 Jun 2010 19:20:38 +0000 (21:20 +0200)]
iwlwifi: add missing rcu_read_lock
Using ieee80211_find_sta() needs to be under
RCU read lock, which iwlwifi currently misses,
so fix it.
Cc: stable@kernel.org Reported-by: Miles Lane <miles.lane@gmail.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Acked-by: Reinette Chatre <reinette.chatre@intel.com> Tested-by: Miles Lane <miles.lane@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Bruno Randolf [Mon, 7 Jun 2010 04:11:19 +0000 (13:11 +0900)]
ath5k: fix NULL pointer in antenna configuration
If the channel is not set yet and we configure the antennas just store the
setting. It will be activated during the next reset, when the channel is set.
Signed-off-by: Bruno Randolf <br1@einfach.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Jason Dravet [Sat, 5 Jun 2010 20:08:29 +0000 (15:08 -0500)]
p54usb: Add device ID for Dell WLA3310 USB
Add Dell WLA3310 USB wireless card, which has a Z-Com XG-705A chipset, to the
USB Ids in p54usb.
Signed-off-by: Jason Dravet <dravet@hotmail.com> Tested-by: Richard Gregory Tillmore <rtillmore@gmail.com> Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net> Acked-by: Christian Lamparter <chunkeey@googlemail.com> Cc: <stable@kernel.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
wl1251_sdio_probe() error path is missing wl1251_free_hw, add it.
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com> Acked-by: Kalle Valo <kvalo@adurom.com> Cc: stable@kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
Eric Dumazet [Sun, 6 Jun 2010 23:48:40 +0000 (23:48 +0000)]
ipmr: dont corrupt lists
ipmr_rules_exit() and ip6mr_rules_exit() free a list of items, but
forget to properly remove these items from list. List head is not
changed and still points to freed memory.
This can trigger a fault later when icmpv6_sk_exit() is called.
Fix is to either reinit list, or use list_del() to properly remove items
from list before freeing them.
Introduced by commit fa4a490165bddf (ipv6: ip6mr: support multiple
tables) and commit ae9451906e79 (ipv4: ipmr: support multiple tables)
Reported-by: Alex Zhavnerchik <alex.vizor@gmail.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> CC: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Jussi Kivilinna [Mon, 7 Jun 2010 07:56:27 +0000 (00:56 -0700)]
asix: check packet size against mtu+ETH_HLEN instead of ETH_FRAME_LEN
Driver checks received packet is too large in asix_rx_fixup() and fails if it is. Problem is
that MTU might be set larger than 1500 and asix fails to work correctly with VLAN tagged
packets. The check should be 'dev->net->mtu + ETH_HLEN' instead.
Tested with AX88772.
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi> Signed-off-by: David S. Miller <davem@davemloft.net>
Timo Teräs [Sun, 6 Jun 2010 22:38:47 +0000 (15:38 -0700)]
r8169: fix random mdio_write failures
Some configurations need delay between the "write completed" indication
and new write to work reliably.
Realtek driver seems to use longer delay when polling the "write complete"
bit, so it waits long enough between writes with high probability (but
could probably break too). This patch adds a new udelay to make sure we
wait unconditionally some time after the write complete indication.
This caused a regression with XID 18000000 boards when the board specific
phy configuration writing many mdio registers was added in commit 51f7fa3054 (r8169: phy init for the 8169scd). Some of the configration
mdio writes would almost always fail, and depending on failure might leave
the PHY in non-working state.
Signed-off-by: Timo Teräs <timo.teras@iki.fi> Acked-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: David S. Miller <davem@davemloft.net>
iwlwifi: move sysfs_create_group to post request firmware
Move the sysfs_create_group to iwl_ucode_callback after we
have safely got the firmware.
The motivation to do this comes from a warning from lockdep which detected
that we request priv->mutex while holding s_active during a sysfs request
(show_statistics in the example copy pasted). The reverse order exists upon
request_firmware: request_firmware which is a sysfs operation
that requires s_active is run under priv->mutex.
This ensures that we don't get sysfs request before we finish to request
the firmware, avoiding this deadlock.
=======================================================
[ INFO: possible circular locking dependency detected ]
-------------------------------------------------------
cat/2595 is trying to acquire lock:
(&priv->mutex){+.+.+.}, at: [<facfa598>] show_statistics+0x48/0x100 [iwlagn]
but task is already holding lock:
(s_active){++++.+}, at: [<c0580ebd>] sysfs_get_active_two+0x1d/0x50
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
Reinette Chatre [Thu, 20 May 2010 17:54:40 +0000 (10:54 -0700)]
iwl3945: enable stuck queue detection on 3945
We learn from
http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=1834 and
https://bugzilla.redhat.com/show_bug.cgi?id=589777
that 3945 can also suffer from a stuck command queue. Enable stuck queue
detection for iwl3945 to enable recovery in this case.
Eric Dumazet [Thu, 3 Jun 2010 09:03:58 +0000 (09:03 +0000)]
rps: tcp: fix rps_sock_flow_table table updates
I believe a moderate SYN flood attack can corrupt RFS flow table
(rps_sock_flow_table), making RPS/RFS much less effective.
Even in a normal situation, server handling short lived sessions suffer
from bad steering for the first data packet of a session, if another SYN
packet is received for another session.
We do following action in tcp_v4_rcv() :
sock_rps_save_rxhash(sk, skb->rxhash);
We should _not_ do this if sk is a LISTEN socket, as about each
packet received on a LISTEN socket has a different rxhash than
previous one.
-> RPS_NO_CPU markers are spread all over rps_sock_flow_table.
Also, it makes sense to protect sk->rxhash field changes with socket
lock (We currently can change it even if user thread owns the lock
and might use rxhash)
This patch moves sock_rps_save_rxhash() to a sock locked section,
and only for non LISTEN sockets.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Thu, 3 Jun 2010 17:03:45 +0000 (17:03 +0000)]
ixgbe: only check pfc bits in hang logic if pfc is enabled
Only check pfc bits in hang logic if PFC is enabled. Previously,
if DCB was enabled but PFC was disabled the incorrect pause
bits would be checked.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Acked-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: check for refcount if pop a stacked dst_entry
xfrm triggers a warning if dst_pop() drops a refcount
on a noref dst. This patch changes dst_pop() to
skb_dst_pop(). skb_dst_pop() drops the refcnt only
on a refcounted dst. Also we don't clone the child
dst_entry, so it is not refcounted and we can use
skb_dst_set_noref() in xfrm_output_one().
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Bob Copeland [Fri, 4 Jun 2010 12:14:14 +0000 (08:14 -0400)]
ath5k: retain promiscuous setting
Commit 08f75adb96569c28e8b920feae7c9f6a4dc566e7, "ath5k: clean up
filter flags setting" introduced a regression in monitor mode such
that the promisc filter flag would get lost.
Although we set the promisc flag when it changed, we did not
preserve it across subsequent calls to configure_filter. This patch
restores the original functionality.
Cc: stable@kernel.org Bisected-by: weedy2887@gmail.com Tested-by: weedy2887@gmail.com Tested-by: Rick Farina <sidhayn@gmail.com> Signed-off-by: Bob Copeland <me@bobcopeland.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Tobias Doerffel [Sat, 29 May 2010 22:02:18 +0000 (00:02 +0200)]
ath5k: depend on CONFIG_PM_SLEEP for suspend/resume functions
When building a kernel with CONFIG_PM=y but neither suspend nor
hibernate support, the compiler complains about the static functions
ath5k_pci_suspend() and ath5k_pci_resume() not being used:
drivers/net/wireless/ath/ath5k/base.c:713:12: warning: ‘ath5k_pci_suspend’ defined but not used
drivers/net/wireless/ath/ath5k/base.c:722:12: warning: ‘ath5k_pci_resume’ defined but not used
Depending on CONFIG_PM_SLEEP rather than CONFIG_PM fixes the issue.
Signed-off-by: Tobias Doerffel <tobias.doerffel@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Johannes Berg [Fri, 28 May 2010 13:22:58 +0000 (15:22 +0200)]
mac80211: process station blockack action frames from work
Processing an association response could take a bit
of time while we set up the hardware etc. During that
time, the AP might already send a blockack request.
If this happens very quickly on a fairly slow machine,
we can end up processing the blockack request before
the association processing has finished. Since the
blockack processing cannot sleep right now, we also
cannot make it wait in the driver.
As a result, sometimes on slow machines the iwlagn
driver gets totally confused, and no traffic can pass
when the aggregation setup was done before the assoc
setup completed.
I'm working on a proper fix for this, which involves
queuing all blockack category action frames from a
work struct, and also allowing the ampdu_action driver
callback to sleep, which will generally clean up the
code and make things easier.
However, this is a very involved and complex change.
To fix the problem at hand in a way that can also be
backported to stable, I've come up with this patch.
Here, I simply process all aggregation action frames
from the managed interface skb queue, which means
their processing will be serialized with processing
the association response, thereby fixing the problem.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Cc: stable@kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
Jeff Kirsher [Wed, 2 Jun 2010 12:44:05 +0000 (12:44 +0000)]
ixgbe: return IXGBE_ERR_RAR_INDEX when out of range
Based on original patch from Shirley Ma <xma@us.ibm.com>
Return IXGBE_ERR_RAR_INDEX when RAR index is out of range, instead of
returning IXGBE_SUCCESS.
CC: Shirley Ma <xma@us.ibm.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Acked-by: Don Skidmore <donald.c.skidmore@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Roland Dreier [Wed, 2 Jun 2010 10:36:53 +0000 (10:36 +0000)]
epic100: Test __BIG_ENDIAN instead of (non-existent) CONFIG_BIG_ENDIAN
Probably no one has used this driver on big-endian systems, since it was
setting up descriptor swapping if CONFIG_BIG_ENDIAN is set, which it
never is, since that symbol is not mentioned anywhere else in the kernel
source. Switch this test to a check for __BIG_ENDIAN so it has a chance
at working.
Signed-off-by: Roland Dreier <rolandd@cisco.com> Acked-by: Jeff Garzik <jgarzik@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter [Wed, 2 Jun 2010 23:56:13 +0000 (23:56 +0000)]
isdn/kcapi: return -EFAULT on copy_from_user errors
copy_from_user() returns the number of bytes remaining but we should
return -EFAULT here. The error code gets returned to the user. Both
old_capi_manufacturer() and capi20_manufacturer() had other places
that already returned -EFAULT so this won't break anything.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter [Wed, 2 Jun 2010 13:43:15 +0000 (13:43 +0000)]
e1000e: change logical negate to bitwise
The bitwise negate is intended here. With the logical negate the
condition is always false.
Signed-off-by: Dan Carpenter <error27@gmail.com> Acked-by: Bruce Allan <bruce.w.allan@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ben Hutchings [Tue, 1 Jun 2010 11:32:43 +0000 (11:32 +0000)]
sfc: Get port number from CS_PORT_NUM, not PCI function number
A single shared memory region used to communicate with firmware is
mapped into both PCI PFs of the SFC9020 and SFL9021. Drivers must be
able to identify which port they are addressing in order to use the
correct sub-region. Currently we use the PCI function number, but the
PCI address may be virtualised. Use the CS_PORT_NUM register field
defined for just this purpose.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Changli Gao [Wed, 2 Jun 2010 14:32:42 +0000 (07:32 -0700)]
cls_u32: use skb_header_pointer() to dereference data safely
use skb_header_pointer() to dereference data safely
the original skb->data dereference isn't safe, as there isn't any skb->len or
skb_is_nonlinear() check. skb_header_pointer() is used instead in this patch.
And when the skb isn't long enough, we terminate the function u32_classify()
immediately with -1.
Signed-off-by: Changli Gao <xiaosuo@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Bénard [Wed, 2 Jun 2010 13:13:34 +0000 (06:13 -0700)]
net/fec: fix pm to survive to suspend/resume
* in the actual driver, calling fec_stop and fec_enet_init doesn't
allow to have a working network interface at resume (where a
ifconfig down and up is required to recover the interface)
* by using fec_enet_close and fec_enet_open, this patch solves this
problem and handle the case where the link changed between suspend
and resume
* this patch also disable clock at suspend and reenable it at resume
Signed-off-by: Eric Bénard <eric@eukrea.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Phil Sutter [Sat, 29 May 2010 13:23:36 +0000 (13:23 +0000)]
korina: count RX DMA OVR as rx_fifo_error
This way, RX DMA overruns (actually being caused by overrun of the
512byte input FIFO) show up in ifconfig output. The rx_fifo_errors
counter is unused otherwise.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: David S. Miller <davem@davemloft.net>
Phil Sutter [Sat, 29 May 2010 13:23:34 +0000 (13:23 +0000)]
korina: fix deadlock on RX FIFO overrun
By calling korina_restart(), the IRQ handler tries to disable the
interrupt it's currently serving. This leads to a deadlock since
disable_irq() waits for any running IRQ handlers to finish before
returning. This patch addresses the issue by turning korina_restart()
into a workqueue task, which is then scheduled when needed.
Reproducing the deadlock is easily done using e.g. GNU netcat to send
large amounts of UDP data to the host running this driver.
Note that the same problem (and fix) applies to TX FIFO underruns, but
apparently these are less easy to trigger.
Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: David S. Miller <davem@davemloft.net>
John Fastabend [Wed, 12 May 2010 21:31:11 +0000 (21:31 +0000)]
net: fix conflict between null_or_orig and null_or_bond
If a skb is received on an inactive bond that does not meet
the special cases checked for by skb_bond_should_drop it should
only be delivered to exact matches as the comment in
netif_receive_skb() says.
However because null_or_bond could also be null this is not
always true. This patch renames null_or_bond to orig_or_bond
and initializes it to orig_dev. This keeps the intent of
null_or_bond to pass frames received on VLAN interfaces stacked
on bonding interfaces without invalidating the statement for
null_or_orig.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Scott Feldman [Tue, 1 Jun 2010 08:59:33 +0000 (08:59 +0000)]
enic: bug fix: make the set/get netlink VF_PORT support symmetrical
To make get/set netlink VF_PORT truly symmetrical, we need to keep track
of what items are set and only return those items on get. Previously, the
driver wasn't differentiating between a set of attr with a NULL string,
for example, and not setting the attr at all. We only want to return
the NULL string if the attr was actually set with a NULL string. Otherwise,
don't return the attr.
Signed-off-by: Scott Feldman <scofeldm@cisco.com> Signed-off-by: David S. Miller <davem@davemloft.net>
If ->open() and ->close() are called multiple times, the same napi structs
will be added to dev->napi_list multiple times, corrupting the dev->napi_list.
This causes free_netdev() to hang during rmmod.
We fix this by calling netif_napi_del() during ->close().
Also, bnx2_init_napi() must not be in the __devinit section since it is
called by ->open().
Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Johannes Berg [Mon, 31 May 2010 10:00:12 +0000 (12:00 +0200)]
mac80211: fix dialog token allocator
The dialog token allocator has apparently been broken
since 579afb20 ("mac80211: fix deadlock in sta->lock")
because it got moved out under the spinlock. Fix it.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Johannes Berg [Sun, 30 May 2010 12:52:58 +0000 (14:52 +0200)]
mac80211: fix blockack-req processing
Daniel reported that the paged RX changes had
broken blockack request frame processing due
to using data that wasn't really part of the
skb data.
Fix this using skb_copy_bits() for the needed
data. As a side effect, this adds a check on
processing too short frames, which previously
this code could do.
Reported-by: Daniel Halperin <dhalperi@cs.washington.edu> Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Acked-by: Daniel Halperin <dhalperi@cs.washington.edu> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Bruno Randolf [Wed, 19 May 2010 01:30:49 +0000 (10:30 +0900)]
ath5k: wake queues on reset
We can wake all queues after a chip reset since everything should be set up and
we are ready to transmit. If we don't do that we might end up starting up with
stopped queues, not beeing able to transmit. (This started to happen after
"ath5k: clean up queue manipulation" but since periodic calibration also
stopped and started the queues this effect was hidden most of the time).
This way we can also get rid of the superfluous ath5k_reset_wake() function.
Signed-off-by: Bruno Randolf <br1@einfach.org> Acked-by: Nick Kossifidis <mickflemm@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
virtio-net bounces buffer allocations off to
a thread if it can't allocate buffers from the atomic
pool. However, if posting buffers still requires atomic
buffers, this is unlikely to succeed.
Fix by passing in the proper gfp_t parameter.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes build error caused by the OF device_node pointer
being moved into struct device.
Signed-off-by: Anatolij Gustschin <agust@denx.de> Cc: Wolfgang Grandegger <wg@grandegger.com> Cc: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / kernel.git / log