Make the LSO code work on BE platforms: parsing_data field of
a parsing BD (PBD) for 57712 was improperly composed which made FW read wrong
values for TCP header's length and offset and, as a result, the corresponding
PCI device was performing bad DMA reads triggering EEH.
Signed-off-by: Vladislav Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Removing usage of rtnl_lock() to protect firmware interface registers.
These registers are accessed in some worker threads and can create a
deadlock if rtnl_lock is taken by upper layers while the worker is still
pending.
We remove rtnl_lock and use a driver mutex just while mailboxes are
accessed.
Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

After commit c1f19b51 (net: support time stamping in phy devices.),
kernel might crash if CONFIG_NETWORK_PHY_TIMESTAMPING=y and
skb_defer_rx_timestamp() handles a packet without an ethernet header.

Fixes kernel bugzilla #24102

Reference: https://bugzilla.kernel.org/show_bug.cgi?id=24102Reported-and-tested-by: Andrew Watts <akwatts@ymail.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Was using L1_CACHE_BYTES for the Ingress Queue Entry Size but it really
needs to be 64 bytes in order to support the largest message sizes.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds the IC+ IP1001 (Gigabit Ethernet Transceiver) driver.
I've had to add an additional delay (2ns) to adjust RX clock phase at
GMII/ RGMII interface (according to the PHY data-sheet). This helps to
have the RGMII working on some ST platforms.
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The ATM subsystem was incorrectly creating the 'device' link for ATM
nodes in sysfs.  This led to incorrect device/parent relationships
exposed by sysfs and udev.  Instead of rolling the 'device' link by hand
in the generic ATM code, pass each ATM driver's bus device down to the
sysfs code and let sysfs do this stuff correctly.
Signed-off-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It was a nice idea, but -ENOTIME and -ENOHW. I never got around to doing
a lot of the clean up that I intended to.
Signed-off-by: Kyle McMartin <kyle@mcmartin.ca>
Acked-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

SCTP_SET_PEER_PRIMARY_ADDR does not accpet v4mapped address, using
v4mapped address in SCTP_SET_PEER_PRIMARY_ADDR socket option will
get -EADDRNOTAVAIL error if v4map is enabled. This patch try to
fix it by mapping v4mapped address to v4 address if allowed.
Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Acked-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pass reference to napi instead of enic device to the isr that services receive queue.
Signed-off-by: Vasanthy Kolluri <vkolluri@cisco.com>
Signed-off-by: Roopa Prabhu <roprabhu@cisco.com>
Signed-off-by: David Wang <dwang2@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

New idev are advertised with NL group RTNLGRP_IPV6_IFADDR, but
should use RTNLGRP_IPV6_IFINFO.
Bug was introduced by commit 8d7a76c9.
Signed-off-by: Wang Xuefu <xuefu.wang@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Thomas Graf <tgraf@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since connector can be built as a module and uses netlink socket
to communicate. The module should have an alias to autoload when socket
of NETLINK_CONNECTOR type is requested.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Martin Lucina <mato@kotelna.sk>
Signed-off-by: David S. Miller <davem@davemloft.net>

I noticed that one of the post-2.6.36 patches broke runtime PM of the
r8169 on my MSI Wind test machine in such a way that the link was not
brought up after reconnecting the network cable.

In the process of debugging the issue I realized that we only should
invoke the runtime PM functions in rtl8169_check_link_status() when
link change is reported and if we do so, the problem goes away.
Moreover, this allows rtl8169_runtime_idle() to be simplified quite
a bit.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There is definitly a problem, that some option cards send up broken
IP pakets leading to corrupted IP packets. These corruptions aren't
detected, because the driver claims that the packets are already
checksummed. This change removes the CHECKSUM_UNNECESSARY option
and let IP detect broken data.
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

xfrm_state_migrate calls kfree instead of xfrm_state_put to free
a failed state. According to git commit 553f9118 this can cause
memory leaks.
Signed-off-by: Thomas Egerer <thomas.egerer@secunet.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

A while back I made some changes to enable netpoll in the bonding driver.  Among
them was a per-cpu flag that indicated we were in a path that held locks which
could cause the netpoll path to block in during tx, and as such the tx path
should queue the frame for later use.  This appears to have given rise to a
regression.  If one of those paths on which we hold the per-cpu flag yields the
cpu, its possible for us to come back on a different cpu, leading to us clearing
a different flag than we set.  This results in odd netpoll drops, and BUG
backtraces appearing in the log, as we check to make sure that we only clear set
bits, and only set clear bits.  I had though briefly about changing the
offending paths so that they wouldn't sleep, but looking at my origional work
more closely, it doesn't appear that a per-cpu flag is warranted.  We alrady
gate the checking of this flag on IFF_IN_NETPOLL, so we don't hit this in the
normal tx case anyway.  And practically speaking, the normal use case for
netpoll is to only have one client anyway, so we're not going to erroneously
queue netpoll frames when its actually safe to do so.  As such, lets just
convert that per-cpu flag to an atomic counter.  It fixes the rescheduling bugs,
is equivalent from a performance perspective and actually eliminates some code
in the process.

Tested by the reporter and myself, successfully
Reported-by: Liang Zheng <lzheng@redhat.com>
CC: Jay Vosburgh <fubar@us.ibm.com>
CC: Andy Gospodarek <andy@greyhouse.net>
CC: David S. Miller <davem@davemloft.net>
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Unconditional use of skb->dev won't work here,
try to fetch the econet device via skb_dst()->dev
instead.

Suggested by Eric Dumazet.
Reported-by: Nelson Elhage <nelhage@ksplice.com>
Tested-by: Nelson Elhage <nelhage@ksplice.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Make sure sysctl_tcp_cookie_size is read once in
tcp_cookie_size_check(), or we might return an illegal value to caller
if sysctl_tcp_cookie_size is changed by another cpu.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: William Allen Simpson <william.allen.simpson@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sysctl_tcp_tso_win_divisor might be set to zero while one cpu runs in
tcp_tso_should_defer(). Make sure we dont allow a divide by zero by
reading sysctl_tcp_tso_win_divisor exactly once.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Enable the port when disabling countermeasures, and disable it on
enabling countermeasures.

This bug causes the response of the system to certain attacks to be
ineffective.

It also prevents wpa_supplicant from getting scan results, as
wpa_supplicant disables countermeasures on startup - preventing the
hardware from scanning.

wpa_supplicant works with ap_mode=2 despite this bug because the commit
handler re-enables the port.

The log tends to look like:

State: DISCONNECTED -> SCANNING
Starting AP scan for wildcard SSID
Scan requested (ret=0) - scan timeout 5 seconds
EAPOL: disable timer tick
EAPOL: Supplicant port status: Unauthorized
Scan timeout - try to get results
Failed to get scan results
Failed to get scan results - try scanning again
Setting scan request: 1 sec 0 usec
Starting AP scan for wildcard SSID
Scan requested (ret=-1) - scan timeout 5 seconds
Failed to initiate AP scan.

Reported by: Giacomo Comes <comes@naic.edu>
Signed-off by: David Kilroy <kilroyd@googlemail.com>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

... and interface up.

In these situations, you are usually trying to connect to a new AP, so
keeping TKIP countermeasures active is confusing. This is already how
the driver behaves (inadvertently). However, querying SIOCGIWAUTH may
tell userspace that countermeasures are active when they aren't.

Clear the setting so that the reporting matches what the driver has
done..

Signed-off by: David Kilroy <kilroyd@googlemail.com>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211 doesn't handle shared skbs correctly at the moment. As a result
a possible resize can trigger a BUG in pskb_expand_head.

[  676.030000] Kernel bug detected[#1]:
[  676.030000] Cpu 0
[  676.030000] $ 0   : 00000000 00000000 819662ff 00000002
[  676.030000] $ 4   : 81966200 00000020 00000000 00000020
[  676.030000] $ 8   : 819662e0 800043c0 00000002 00020000
[  676.030000] $12   : 3b9aca00 00000000 00000000 00470000
[  676.030000] $16   : 80ea2000 00000000 00000000 00000000
[  676.030000] $20   : 818aa200 80ea2018 80ea2000 00000008
[  676.030000] $24   : 00000002 800ace5c
[  676.030000] $28   : 8199a000 8199bd20 81938f88 80f180d4
[  676.030000] Hi    : 0000026e
[  676.030000] Lo    : 0000757e
[  676.030000] epc   : 801245e4 pskb_expand_head+0x44/0x1d8
[  676.030000]     Not tainted
[  676.030000] ra    : 80f180d4 ieee80211_skb_resize+0xb0/0x114 [mac80211]
[  676.030000] Status: 1000a403    KERNEL EXL IE
[  676.030000] Cause : 10800024
[  676.030000] PrId  : 0001964c (MIPS 24Kc)
[  676.030000] Modules linked in: mac80211_hwsim rt2800lib rt2x00soc rt2x00pci rt2x00lib mac80211 crc_itu_t crc_ccitt cfg80211 compat arc4 aes_generic deflate ecb cbc [last unloaded: rt2800pci]
[  676.030000] Process kpktgend_0 (pid: 97, threadinfo=8199a000, task=81879f48, tls=00000000)
[  676.030000] Stack : ffffffff 00000000 00000000 00000014 00000004 80ea2000 00000000 00000000
[  676.030000]         818aa200 80f180d4 ffffffff 0000000a 81879f78 81879f48 81879f48 00000018
[  676.030000]         81966246 80ea2000 818432e0 80f1a420 80203050 81814d98 00000001 81879f48
[  676.030000]         81879f48 00000018 81966246 818432e0 0000001a 8199bdd4 0000001c 80f1b72c
[  676.030000]         80203020 8001292c 80ef4aa2 7f10b55d 801ab5b8 81879f48 00000188 80005c90
[  676.030000]         ...
[  676.030000] Call Trace:
[  676.030000] [<801245e4>] pskb_expand_head+0x44/0x1d8
[  676.030000] [<80f180d4>] ieee80211_skb_resize+0xb0/0x114 [mac80211]
[  676.030000] [<80f1a420>] ieee80211_xmit+0x150/0x22c [mac80211]
[  676.030000] [<80f1b72c>] ieee80211_subif_start_xmit+0x6f4/0x73c [mac80211]
[  676.030000] [<8014361c>] pktgen_thread_worker+0xfac/0x16f8
[  676.030000] [<8002ebe8>] kthread+0x7c/0x88
[  676.030000] [<80008e0c>] kernel_thread_helper+0x10/0x18
[  676.030000]
[  676.030000]
[  676.030000] Code: 24020001  10620005  2502001f <0200000d> 0804917a  00000000  2502001f  00441023  00531021

Fix this by making a local copy of shared skbs prior to mangeling them.
To avoid copying the skb unnecessarily move the skb_copy call below the
checks that don't need write access to the skb.

Also, move the assignment of nh_pos and h_pos below the skb_copy to point
to the correct skb.

It would be possible to avoid another resize of the copied skb by using
skb_copy_expand instead of skb_copy but that would make the patch more
complex. Also, shared skbs are a corner case right now, so the resize
shouldn't matter much.

Cc: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

The HW has to be set to FULLSLEEP mode during suspend,
when no interface has been brought up. Not doing this would
break resume, as the chip won't be powered up at all.
Signed-off-by: Sujith Manoharan <Sujith.Manoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Signed-off-by: Javier Cardona <javier@cozybit.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

This results in an erroneus num_adhoc_vifs count, as the this counter
was incremented but not decremented for mesh interfaces.
Signed-off-by: Javier Cardona <javier@cozybit.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

This patch fixes the oops below when attempting to bring up a mesh
interface on ath5k hardware.

[  128.933099] kernel BUG at drivers/net/wireless/ath/ath5k/base.c:197!
[  128.933099] invalid opcode: 0000 [#1]
(...)
[  128.933099] Call Trace:
[  128.933099]  [<c83b77fa>] ? ath5k_beacon_update+0x57/0x1f8 [ath5k]
[  128.933099]  [<c02d9a40>] ? __sysfs_add_one+0x28/0x76
[  128.933099]  [<c83b830e>] ? ath5k_bss_info_changed+0x13f/0x173
[ath5k]
[  128.933099]  [<c82ff629>] ? ieee80211_config_beacon+0xc0/0x17e
[mac80211]
[  128.933099]  [<c82f073e>] ?
ieee80211_bss_info_change_notify+0x182/0x18b [mac80211]
[  128.933099]  [<c83b81cf>] ? ath5k_bss_info_changed+0x0/0x173 [ath5k]
[  128.933099]  [<c82ff6d6>] ? ieee80211_config_beacon+0x16d/0x17e
[mac80211]
[  128.933099]  [<c82ff753>] ? ieee80211_add_beacon+0x34/0x39 [mac80211]
[  128.933099]  [<c830a4ed>] ? ieee80211s_init+0xf8/0x10f [mac80211]
[  128.933099]  [<c830a5df>] ? ieee80211_mesh_init_sdata+0xdb/0x154 [mac80211]
Signed-off-by: Javier Cardona <javier@cozybit.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

The interrupt handler takes a lock - but since commit bcad6e80 this
lock goes through an indirection specified in the hermes_t structure.
We must therefore initialise the structure before setting up the
interrupt handler.

Fix orinoco_cs and spectrum_cs

<https://bugzilla.kernel.org/show_bug.cgi?id=23932>

Bisected by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off by: David Kilroy <kilroyd@googlemail.com>
Cc: stable@kernel.org
Signed-off-by: John W. Linville <linville@tuxdriver.com>

In order to set LRO on ehea, the user must set a module parameter, which
is not the standard way to do so. This patch adds a way to set LRO using
the ethtool tool.
Signed-off-by: Breno Leitao <leitao@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Rather than printing the message to the log, use a mib counter to keep
track of the count of occurences of time wait bucket overflow.  Reduces
spam in logs.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

x25 does not decrement the network device reference counts on module unload.
Thus unregistering any pre-existing interface after unloading the x25 module
hangs and results in

 unregister_netdevice: waiting for tap0 to become free. Usage count = 1

This patch decrements the reference counts of all interfaces in x25_link_free,
the way it is already done in x25_link_device_down for NETDEV_DOWN events.
Signed-off-by: Apollon Oikonomopoulos <apollon@noc.grnet.gr>
Signed-off-by: David S. Miller <davem@davemloft.net>

Regarding  benet be_cmd_multicast_set() function, now using
netdev_for_each_mc_addr() helper for mac address copy, but
when copying to req->mac[] did not increase of the index.

Cc: Sathya Perla <sathyap@serverengines.com>
Cc: Subbu Seetharaman <subbus@serverengines.com>
Cc: Sarveshwar Bandi <sarveshwarb@serverengines.com>
Cc: Ajit Khaparde <ajitk@serverengines.com>
Signed-off-by: Joe Jin <joe.jin@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Using the SOCK_DGRAM enum results in
"net-pf-2-proto-SOCK_DGRAM-type-115", so use the numeric value like it
is done in net/dccp.
Signed-off-by: Michal Marek <mmarek@suse.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

We need to drop the mutex and do a dev_put, so set an error code and break like
the other paths, instead of returning directly.
Signed-off-by: Nelson Elhage <nelhage@ksplice.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Le dimanche 05 décembre 2010 à 12:23 +0100, Eric Dumazet a écrit :
> Le dimanche 05 décembre 2010 à 09:19 +0100, Eric Dumazet a écrit :
>
> > Hmm..
> >
> > If somebody can explain why RTNL is held in arp_ioctl() (and therefore
> > in arp_req_delete()), we might first remove RTNL use in arp_ioctl() so
> > that your patch can be applied.
> >
> > Right now it is not good, because RTNL wont be necessarly held when you
> > are going to call arp_invalidate() ?
>
> While doing this analysis, I found a refcount bug in llc, I'll send a
> patch for net-2.6

Oh well, of course I must first fix the bug in net-2.6, and wait David
pull the fix in net-next-2.6 before sending this rcu conversion.

Note: this patch should be sent to stable teams (2.6.34 and up)

[PATCH net-2.6] llc: fix a device refcount imbalance

commit abf9d537 (llc: add support for SO_BINDTODEVICE) added one
refcount imbalance in llc_ui_bind(), because dev_getbyhwaddr() doesnt
take a reference on device, while dev_get_by_index() does.

Fix this using RCU locking. And since an RCU conversion will be done for
2.6.38 for dev_getbyhwaddr(), put the rcu_read_lock/unlock exactly at
their final place.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: stable@kernel.org
Cc: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If we break the loop when there are still skbs in tq and no skb in
rq, the skbs will be left in txq until new skbs are enqueued into rq.
In rare cases, no new skb is queued, then these skbs will stay in rq
forever.

After this patch, if tq isn't empty when we break the loop, we goto
resched directly.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>

The bug has to do with boundary checks on the initial receive window.
If the initial receive window falls between init_cwnd and the
receive window specified by the user, the initial window is incorrectly
brought down to init_cwnd. The correct behavior is to allow it to
remain unchanged.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix the calculation of the inexact hash-based MAC address filter.
It's 64 bits but current code is missing a ULL.  Results in filtering out
some legitimate packets.
Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>