The head element of rt6_info{} is dst_entry{}, and
IPv6 specific elements follow.
Because elements at the end of dst_entry{} are frequently
updated, it is not good to put frequently-used static
elements, such as rt6i_idev, rt6i_dst or rt6i_flags in the
same cache line.
On the other hand, fib6_table, rt6i_node or rt6i_gateway are
rarely used, so it is okay to stay in the same cache line.
Let's rearrange rt6_info{}.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Changli Gao [Tue, 30 Mar 2010 20:16:22 +0000 (20:16 +0000)]
rps: keep the old behavior on SMP without rps
keep the old behavior on SMP without rps
RPS introduces a lock operation to per cpu variable input_pkt_queue on
SMP whenever rps is enabled or not. On SMP without RPS, this lock isn't
needed at all.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
----
net/core/dev.c | 42 ++++++++++++++++++++++++++++--------------
1 file changed, 28 insertions(+), 14 deletions(-) Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Wed, 31 Mar 2010 02:08:31 +0000 (02:08 +0000)]
r8169: Fix rtl8169_rx_interrupt()
In case a reset is performed, rtl8169_rx_interrupt() is called from
process context instead of softirq context. Special care must be taken
to call appropriate network core services (netif_rx() instead of
netif_receive_skb()). VLAN handling also corrected.
Reported-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Tested-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Diagnosed-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Wed, 31 Mar 2010 07:06:04 +0000 (07:06 +0000)]
gen_estimator: deadlock fix
One of my test machine got a deadlock during "tc" sessions,
adding/deleting classes & filters, using traffic estimators.
After some analysis, I believe we have a potential use after free case
in est_timer() :
spin_lock(e->stats_lock); << HERE >>
read_lock(&est_lock);
if (e->bstats == NULL) << TEST >>
goto skip;
Test is done a bit late, because after estimator is killed, and before
rcu grace period elapsed, we might already have freed/reuse memory where
e->stats_locks points to (some qdisc->q.lock)
A possible fix is to respect a rcu grace period at Qdisc dismantle time.
On 64bit, sizeof(struct Qdisc) is exactly 192 bytes. Adding 16 bytes to
it (for struct rcu_head) is a problem because it might change
performance, given QDISC_ALIGNTO is 32 bytes.
This is why I also change QDISC_ALIGNTO to 64 bytes, to satisfy most
current alignment requirements.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The check if error signaling is wanted (inet->recverr != 0) is done by
the caller: raw.c:raw_err() and udp.c:__udp4_lib_err(), so there is no
need to check this condition again.
Signed-off-by: Hagen Paul Pfeifer <hagen@jauu.net> Signed-off-by: David S. Miller <davem@davemloft.net>
FUJITA Tomonori [Wed, 31 Mar 2010 22:13:05 +0000 (22:13 +0000)]
acenic: fix the misusage of zero dma address
acenic wrongly assumes that zero is an invalid dma address (calls
dma_unmap_page for only non zero dma addresses). Zero is a valid dma
address on some architectures. The dma length can be used here.
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes this problem as well as others like it in the NetLabel
code. Also included in this patch is the identification of future work
to eliminate the RCU read lock in netlbl_domhsh_add(), but in the interest
of getting this patch out quickly that work will happen in another patch
to be finished later.
Thanks to Eric Dumazet and Paul McKenney for their help in understanding
the recent RCU changes.
Signed-off-by: Paul Moore <paul.moore@hp.com> Reported-by: David Howells <dhowells@redhat.com> CC: Eric Dumazet <eric.dumazet@gmail.com> CC: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Amerigo Wang [Wed, 31 Mar 2010 21:30:52 +0000 (21:30 +0000)]
bonding: fix potential deadlock in bond_uninit()
bond_uninit() is invoked with rtnl_lock held, when it does destroy_workqueue()
which will potentially flush all works in this workqueue, if we hold rtnl_lock
again in the work function, it will deadlock.
So move destroy_workqueue() to destructor where rtnl_lock is not held any more,
suggested by Eric.
Signed-off-by: WANG Cong <amwang@redhat.com> Cc: Jay Vosburgh <fubar@us.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Stephen Hemminger <shemminger@vyatta.com> Cc: Jiri Pirko <jpirko@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Carmelo AMOROSO [Wed, 31 Mar 2010 21:44:03 +0000 (21:44 +0000)]
stmmac: fix kconfig for crc32 build error
stmmac uses crc32 functions so it needs to select CRC32.
Fixes build error:
drivers/built-in.o: In function `dwmac1000_set_filter':
dwmac1000_core.c:(.text+0x3c380): undefined reference to `crc32_le'
dwmac1000_core.c:(.text+0x3c384): undefined reference to `bitrev32'
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com> Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Marc Zyngier [Mon, 29 Mar 2010 08:57:56 +0000 (08:57 +0000)]
Add hotplug support to mcp251x driver
Chip model can now be selected directly by matching the modalias name
(instead of filling the .model field in platform_data), and allows the
module to be auto-loaded. Previous behaviour is of course still supported.
Convert the two in-tree users to this feature (icontrol & zeus).
Tested on an Zeus platform (mcp2515).
Signed-off-by: Marc Zyngier <maz@misterjones.org> Acked-by: Christian Pellegrin <chripell@fsfe.org> Cc: Edwin Peer <epeer@tmtservices.co.za> Acked-by: Wolfgang Grandegger <wg@grandegger.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This adds ethtool and device feature flag to allow control
of receive hashing offload.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Jeff Garzik <jgarzik@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Nick Nunley [Fri, 26 Mar 2010 11:36:47 +0000 (11:36 +0000)]
igb: add per-packet timestamping
This patch adds support for per-packet timestamping for the
82580 adapter. The rx timestamp code is also pulled out of the
inlined rx hotpath and instead moved to a seperate function.
This version adds a comment explaining the per-packet timestamping
code added to igb_hwtstamp_ioctl().
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
ipv6 fib: Use "Sweezle" to optimize addr_bit_test().
addr_bit_test() is used in various places in IPv6 routing table
subsystem. It checks if the given fn_bit is set,
where fn_bit counts bits from MSB in words in network-order.
Alexander Duyck [Thu, 25 Mar 2010 13:10:08 +0000 (13:10 +0000)]
igb: update hw_debug macro to make use of netdev_dbg call
This change updates the igb driver to make use of the netdev_dbg function
macros now provided in netdevice.h
This is meant to be provided as an alternative to the patch provided by
Joe Perches.
It also removes igb_get_time_str since I found that it is unused code that
is no longer used even in debug.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sjur Braendeland [Tue, 30 Mar 2010 13:56:30 +0000 (13:56 +0000)]
net-caif-driver: add CAIF serial driver (ldisc)
Add CAIF Serial driver. This driver is implemented as a line discipline.
caif_serial uses the following module parameters:
ser_use_stx - specifies if STart of frame eXtension is in use.
ser_loop - sets the interface in loopback mode.
Signed-off-by: Sjur Braendeland <sjur.brandeland@stericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sjur Braendeland [Tue, 30 Mar 2010 13:56:26 +0000 (13:56 +0000)]
net-caif: add CAIF socket implementation
Implementation of CAIF sockets for protocol and address family
PF_CAIF and AF_CAIF.
CAIF socket is connection oriented implementing SOCK_SEQPACKET
and SOCK_STREAM interface with supporting blocking and non-blocking mode.
Signed-off-by: Sjur Braendeland <sjur.brandeland@stericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Sjur Braendeland [Tue, 30 Mar 2010 13:56:24 +0000 (13:56 +0000)]
net-caif: add CAIF generic caif support functions
Support functions for the caif protocol stack:
cfcnfg.c - CAIF Configuration Module used for
adding and removing drivers and connection
cfpkt_skbuff.c - CAIF Packet layer (SKB helper functions)
Signed-off-by: Sjur Braendeland <sjur.brandeland@stericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Shirley Ma [Mon, 29 Mar 2010 15:19:15 +0000 (15:19 +0000)]
virtio_net: missing sg_init_table
Add missing sg_init_table for sg_set_buf in virtio_net which
induced in defer skb patch.
Reported-by: Thomas Müller <thomas@mathtm.de> Tested-by: Thomas Müller <thomas@mathtm.de> Signed-off-by: Shirley Ma <xma@us.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: Fix oops from tcp_collapse() when using splice()
tcp_read_sock() can have a eat skbs without immediately advancing copied_seq.
This can cause a panic in tcp_collapse() if it is called as a result
of the recv_actor dropping the socket lock.
A userspace program that splices data from a socket to either another
socket or to a file can trigger this bug.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Mack [Mon, 29 Mar 2010 15:14:18 +0000 (17:14 +0200)]
net/wireless/libertas: do not call wiphy_unregister() w/o wiphy_register()
The libertas driver calls wiphy_unregister() without a prior
wiphy_register() when a devices fails initialization. Fix this by
introducing a private flag.
[ 9.310000] Unable to handle kernel NULL pointer dereference at virtual address 00000000
[...]
[ 9.330000] [<c0311310>] (wiphy_unregister+0xfc/0x19c) from [<bf00c9ec>] (lbs_cfg_free+0x70/0x9c [libertas])
[ 9.330000] [<bf00c9ec>] (lbs_cfg_free+0x70/0x9c [libertas]) from [<bf014fdc>] (lbs_remove_card+0x180/0x210 [libertas])
[ 9.330000] [<bf014fdc>] (lbs_remove_card+0x180/0x210 [libertas]) from [<bf035394>] (if_sdio_probe+0xdc4/0xef4 [libertas_sdio])
[ 9.330000] [<bf035394>] (if_sdio_probe+0xdc4/0xef4 [libertas_sdio]) from [<c0230d14>] (sdio_bus_probe+0xd4/0xf0)
[ 9.330000] [<c0230d14>] (sdio_bus_probe+0xd4/0xf0) from [<c01a6034>] (driver_probe_device+0xa4/0x174)
[ 9.330000] [<c01a6034>] (driver_probe_device+0xa4/0x174) from [<c01a6164>] (__driver_attach+0x60/0x84)
[ 9.330000] [<c01a6164>] (__driver_attach+0x60/0x84) from [<c01a5854>] (bus_for_each_dev+0x4c/0x8c)
[ 9.330000] [<c01a5854>] (bus_for_each_dev+0x4c/0x8c) from [<c01a50e4>] (bus_add_driver+0xa0/0x228)
[ 9.330000] [<c01a50e4>] (bus_add_driver+0xa0/0x228) from [<c01a6470>] (driver_register+0xc0/0x150)
[ 9.330000] [<c01a6470>] (driver_register+0xc0/0x150) from [<bf03a06c>] (if_sdio_init_module+0x6c/0x108 [libertas_sdio])
[ 9.330000] [<bf03a06c>] (if_sdio_init_module+0x6c/0x108 [libertas_sdio]) from [<c00263ac>] (do_one_initcall+0x5c/0x1bc)
[ 9.330000] [<c00263ac>] (do_one_initcall+0x5c/0x1bc) from [<c0069f80>] (sys_init_module+0xc0/0x1f0)
[ 9.330000] [<c0069f80>] (sys_init_module+0xc0/0x1f0) from [<c0026f00>] (ret_fast_syscall+0x0/0x30)
Signed-off-by: Daniel Mack <daniel@caiaq.de> Cc: Dan Williams <dcbw@redhat.com> Cc: John W. Linville <linville@tuxdriver.com> Cc: Holger Schurig <hs4233@mail.mn-solutions.de> Cc: Bing Zhao <bzhao@marvell.com> Cc: libertas-dev@lists.infradead.org Cc: linux-wireless@vger.kernel.org Cc: netdev@vger.kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
Dan Carpenter [Sun, 28 Mar 2010 11:55:00 +0000 (14:55 +0300)]
iwlwifi: range checking issue
IWL_RATE_COUNT is 13 and IWL_RATE_COUNT_LEGACY is 12.
IWL_RATE_COUNT_LEGACY is the right one here because iwl3945_rates
doesn't support 60M and also that's how "rates" is defined in
iwlcore_init_geos() from drivers/net/wireless/iwlwifi/iwl-core.c.
Signed-off-by: Dan Carpenter <error27@gmail.com> Cc: stable@kernel.org Acked-by: Zhu Yi <yi.zhu@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
setup correct int pipe type in ar9170_usb_exec_cmd
An int urb is constructed but we fill it in with a bulk pipe type.
Commit f661c6f8c67bd55e93348f160d590ff9edf08904 implemented a pipe type
check when CONFIG_USB_DEBUG is enabled. The check failed for all the ar9170
usb transfers and the driver could not configure the wifi dongle.
This went unnoticed until now because most people don't have
CONFIG_USB_DEBUG enabled.
Signed-off-by: Valentin Longchamp <valentin.longchamp@epfl.ch> Cc: Stable <stable@kernel.org> Acked-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
rt2x00: Disable powersaving by default in rt2500usb.
Recent bug reports have shown that rt2500usb also suffers from the
powersave problems that the PCI rt2x00 drivers suffer from.
So disable powersaving by default for rt2500usb as well.
Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
rt2x00: Fix typo in RF register programming of rt2800.
Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
Shanyu Zhao [Tue, 16 Mar 2010 17:22:26 +0000 (10:22 -0700)]
iwlwifi: clear unattended interrupts in tasklet
Previously in interrupt handling tasklet, iwlwifi driver only clear/ack
those interrupts that are enabled by the driver through inta_mask.
If the hardware generates unattended interrupts, driver will not ack them,
defeating the interrupt coalescing feature. This results in high number
of interrupts per second and high CPU utilization.
This patch addresses this issue by acking those unattended interrupts
in the tasklet. Local test showed an order of magnitude improvement
in terms of the number of interrupts without sacrificing networking
throughput. This is a workaround for hardware issue.
Johannes Berg [Mon, 22 Mar 2010 20:42:43 +0000 (13:42 -0700)]
mac80211: move netdev queue enabling to correct spot
"mac80211: fix skb buffering issue" still left a race
between enabling the hardware queues and the virtual
interface queues. In hindsight it's totally obvious
that enabling the netdev queues for a hardware queue
when the hardware queue is enabled is wrong, because
it could well possible that we can fill the hw queue
with packets we already have pending. Thus, we must
only enable the netdev queues once all the pending
packets have been processed and sent off to the device.
In testing, I haven't been able to trigger this race
condition, but it's clearly there, possibly only when
aggregation is being enabled.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Cc: stable@kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
Porsch, Marco [Wed, 24 Feb 2010 08:53:13 +0000 (09:53 +0100)]
mac80211: fix PREQ processing and one small bug
1st) a PREQ should only be processed, if it has the same SN and better
metric (instead of better or equal).
2nd) next_hop[ETH_ALEN] now actually used to buffer
mpath->next_hop->sta.addr for use out of lock.
Signed-off-by: Marco Porsch <marco.porsch@siemens.com> Acked-by: Javier Cardona <javier@cozybit.com> Cc: stable@kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
John W. Linville [Tue, 16 Mar 2010 19:40:59 +0000 (15:40 -0400)]
wireless: convert reg_regdb_search_lock to mutex
Stanse discovered that kmalloc is being called with GFP_KERNEL while
holding this spinlock. The spinlock can be a mutex instead, which also
enables the removal of the unlock/lock around the lock/unlock of
cfg80211_mutex and the call to set_regdom.
Reported-by: Jiri Slaby <jirislaby@gmail.com> Cc: stable@kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
vxge: Updating Maintainer list of S2IO 10GbE drivers (xframe / vxge).
- updating Maintainer list of S2IO 10GbE drivers (xframe / vxge).
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
vxge: Set function-0 as the privilaged function for normal function.
- For Normal function (MR-IOV disabled, SR-IOV disabled) Function-0 is the
privilaged function.
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
- Fixed MSIX interrupt configuration to support non contiguous vpaths in
functions. Four msi-x vectors are reserved per vpath internal to the chip.
In all, there are 68 msi-x vectors for the 17 vpaths in the chip. In the
multi function configurations, non-contiguous vpaths are configured to
represent a function. For instance vpaths 0 and 8 can be configured to
represent function zero.
- If pci_enable_msix fails for the requested vectors, try with a lesser number
vectors by reducing the vpath count.
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
- Fixed "ethtool -d" prints
- reg_space pointer was getting over written, updating it correctly.
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
vxge: Fix starvation of receive ring controller when blasted by short packets.
- Fix starvation of receive ring controller when blasted by short packets. This was because the driver was posting 64 rxds initially while the ring controller
was expecting to read 256 replenished rxds. While the driver was coming up,
the internal rxd memory filled up the 64 buffers provided and the ring
controller was left waiting for the remaining 192 rxds to complete the write
back of completed rxds to the host and generation of an interrupt.
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
vxge: Fix a receive stall due to driver being out of synch with chip.
- Fix a receive stall due to driver being out of synch with chip. In a corner
case scenario the adapter's ring controller may return a RxD with transfer code
of 0xC, while the host ownership bit is still set to the adapter. The driver
needs to assume that this case where (host_ownership == 1 or adapter) and
(transfer_code == 0xC) is valid, that is, this RxD has been returned by the
receive ring controller but no frame data is associated with the rxd.
- Restore the transfer code field of each newly replenished RxD to 0x0.
- Code cleanup. Removed usage of magic numbers.
Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com> Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
For NX2031, msix is supported from fw version > 3.4.336.
This fw version check should take flash fw in consider instead of
running fw or fw from file. Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Rajesh K Borundia <rajesh.borundia@qlogic.com> Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Validate all sections of unified romimage, before accessing them,
to avoid seg fault.
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
o DEV_NEED_RESET state was not handled during fw intialization phase.
o nx_decr_dev_ref_cnt() can return error, if fail to grab pcie seamphore. Signed-off-by: David S. Miller <davem@davemloft.net>
Kernel default tx csum function (ethtool_op_get_tx_csum) doesn't show
correct csum status. It takes various FLAGS (NETIF_F_ALL_CSUM) in account
to show tx csum status, which driver doesn't set while disabling tx csum.
Signed-off-by: Sucheta Chakraborty <sucheta.chakraborty@qlogic.com> Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Neil Horman [Mon, 29 Mar 2010 20:16:02 +0000 (13:16 -0700)]
r8169: offical fix for CVE-2009-4537 (overlength frame DMAs)
Official patch to fix the r8169 frame length check error.
Based on this initial thread:
http://marc.info/?l=linux-netdev&m=126202972828626&w=1
This is the official patch to fix the frame length problems in the r8169
driver. As noted in the previous thread, while this patch incurs a performance
hit on the driver, its possible to improve performance dynamically by updating
the mtu and rx_copybreak values at runtime to return performance to what it was
for those NICS which are unaffected by the ideosyncracy (if there are any).
Summary:
A while back Eric submitted a patch for r8169 in which the proper
allocated frame size was written to RXMaxSize to prevent the NIC from dmaing too
much data. This was done in commit fdd7b4c3302c93f6833e338903ea77245eb510b4. A
long time prior to that however, Francois posted 126fa4b9ca5d9d7cb7d46f779ad3bd3631ca387c, which expiclitly disabled the MaxSize
setting due to the fact that the hardware behaved in odd ways when overlong
frames were received on NIC's supported by this driver. This was mentioned in a
security conference recently:
http://events.ccc.de/congress/2009/Fahrplan//events/3596.en.html
It seems that if we can't enable frame size filtering, then, as Eric correctly
noticed, we can find ourselves DMA-ing too much data to a buffer, causing
corruption. As a result is seems that we are forced to allocate a frame which
is ready to handle a maximally sized receive.
This obviously has performance issues with it, so to mitigate that issue, this
patch does two things:
1) Raises the copybreak value to the frame allocation size, which should force
appropriately sized packets to get allocated on rx, rather than a full new 16k
buffer.
2) This patch only disables frame filtering initially (i.e., during the NIC
open), changing the MTU results in ring buffer allocation of a size in relation
to the new mtu (along with a warning indicating that this is dangerous).
Because of item (2), individuals who can't cope with the performance hit (or can
otherwise filter frames to prevent the bug), or who have hardware they are sure
is unaffected by this issue, can manually lower the copybreak and reset the mtu
such that performance is restored easily.
Signed-off-by: Neil Horman <nhorman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Stephen Rothwell [Mon, 29 Mar 2010 08:00:44 +0000 (01:00 -0700)]
rps: fix net-sysfs build for !CONFIG_RPS
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
ipv6: Don't drop cache route entry unless timer actually expired.
This is ipv6 variant of the commit 5e016cbf6.. ("ipv4: Don't drop
redirected route cache entry unless PTMU actually expired")
by Guenter Roeck <guenter.roeck@ericsson.com>.
Remove cache route entry in ipv6_negative_advice() only if
the timer is expired.
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 29 Mar 2010 01:56:34 +0000 (18:56 -0700)]
tulip: Add missing parens.
As reported by Stephen Rothwell.
drivers/net/tulip/uli526x.c: In function 'uli526x_rx_packet':
drivers/net/tulip/uli526x.c:861: warning: assignment makes pointer from integer without a cast
Signed-off-by: David S. Miller <davem@davemloft.net>
Francois Romieu [Sun, 28 Mar 2010 02:35:46 +0000 (19:35 -0700)]
r8169: fix broken register writes
This is quite similar to b39fe41f481d20c201012e4483e76c203802dda7
though said registers are not even documented as 64-bit registers
- as opposed to the initial TxDescStartAddress ones - but as single
bytes which must be combined into 32 bits at the MMIO read/write
level before being merged into a 64 bit logical entity.
Credits go to Ben Hutchings <ben@decadent.org.uk> for the MAR
registers (aka "multicast is broken for ages on ARM) and to
Timo Teräs <timo.teras@iki.fi> for the MAC registers.
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Engelhardt [Sun, 28 Mar 2010 00:15:29 +0000 (17:15 -0700)]
net: increase preallocated size of nlmsg to accomodate for IFLA_STATS64
When more data is stuffed into an nlmsg than initially projected, an
extra allocation needs to be done. Reserve enough for IFLA_STATS64 so
that this does not to needlessy happen.
Signed-off-by: Jan Engelhardt <jengelh@medozas.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Ken Kawasaki [Sat, 27 Mar 2010 10:55:37 +0000 (10:55 +0000)]
pcnet_cs: add new id
pcnet_cs:
*add new id (Allied Telesis LM33-PCM-T Lan&Modem multifunction card)
*use PROD_ID for LA-PCM.(because LA-PCM and LM33-PCM-T use the same MANF_ID).
Signed-off-by: Ken Kawasaki <ken_kawasaki@spring.nifty.jp> Signed-off-by: David S. Miller <davem@davemloft.net>
Andy Gospodarek [Thu, 25 Mar 2010 14:49:05 +0000 (14:49 +0000)]
bonding: fix broken multicast with round-robin mode
Round-robin (mode 0) does nothing to ensure that any multicast traffic
originally destined for the host will continue to arrive at the host when
the link that sent the IGMP join or membership report goes down. One of
the benefits of absolute round-robin transmit.
Keeping track of subscribed multicast groups for each slave did not seem
like a good use of resources, so I decided to simply send on the
curr_active slave of the bond (typically the first enslaved device that
is up). This makes failover management simple as IGMP membership
reports only need to be sent when the curr_active_slave changes. I
tested this patch and it appears to work as expected.
Originally reported by Lon Hohberger <lhh@redhat.com>.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net> CC: Lon Hohberger <lhh@redhat.com> CC: Jay Vosburgh <fubar@us.ibm.com> Signed-off-by: Jay Vosburgh <fubar@us.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Engelhardt [Sat, 27 Mar 2010 23:35:50 +0000 (16:35 -0700)]
net: fix unaligned access in IFLA_STATS64
Tony Luck observes that the original IFLA_STATS64 submission causes
unaligned accesses. This is because nla_data() returns a pointer to a
memory region that is only aligned to 32 bits. Do some memcpying to
workaround this.
Signed-off-by: Jan Engelhardt <jengelh@medozas.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Emil Tantilov [Fri, 26 Mar 2010 11:25:58 +0000 (11:25 +0000)]
e1000: do not modify tx_queue_len on link speed change
Previously the driver tweaked txqueuelen to avoid false Tx hang reports
seen at half duplex. This had the effect of overriding user set values
on link change/reset. Testing shows that adjusting only the timeout
factor is sufficient to prevent Tx hang reports at half duplex.
This patch removes all instances of tx_queue_len in the driver.
Based on e1000e patch by Franco Fichtner <franco@lastsummer.de>
CC: Franco Fichtner <franco@lastsummer.de> Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com> Acked-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
When cache is unresolved, c->mf[6]c_parent is set to 65535 and
minvif, maxvif are not initialized, hence we must avoid to
parse IIF and OIF.
A second problem can happen when the user dumps a cache entry
where a VIF, that was referenced at creation time, has been
removed.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Greg Rose [Thu, 25 Mar 2010 17:06:48 +0000 (17:06 +0000)]
ixgbe: Do not run all Diagnostic offline tests when VFs are active
When running the offline diagnostic tests check to see if any VFs are
online. If so then only run the link test. This is necessary because
the VFs running in guest VMs aren't aware of when the PF is taken
offline for a diagnostic test. Also put a message to the system log
telling the system administrator to take the VFs offline manually if
(s)he wants to run a full diagnostic. Return 1 on each of the tests
not run to alert the user of the condition.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>