bnx2x: Accessing un-mapped page

The allocated RX buffer size was 64 bytes bigger than the PCI mapped
size with no good reason. If the packet was actually using the buffer up
to its limit and if the last 64 bytes of the buffer crossed 4KB boundary
then an unmapped PCI page was accessed. The fix is to use only one
parameter for the buffer size - there is no need to differentiate
between the buffer size and the PCI mapping size since the extra 64
bytes can actually be used by the FW to align the Ethernet payload to
64 bytes.

Also updating the driver version and date

Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'master' of git://git./linux/kernel/git/linville/wireless-2.6

ath9k: Fix TX control flag use for no ACK and RTS/CTS

Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Fix TX status reporting

Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix STATUS_EXIT_PENDING is not set on pci_remove

This patch sets STATUS_EXIT_PENDING on pci_remove. Otherwise
iwl4965_down may fail to uninitialize the driver.

Signed-off-by: Gregory Greenman <gregory.greenman@intel.com>
Signed-off-by: Mohamed Abbas <mohamed.abbas@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: call apm stop on exit

This patch calls apm stop on exit and suspend. Without this patch
hardware consumes power even after driver is removed or suspended.

Signed-off-by: Gregory Greenman <gregory.greenman@intel.com>
Signed-off-by: Mohamed Abbas <mohamed.abbas@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix Tx cmd memory allocation failure handling

This patch "iwlwifi: do not use GFP_DMA in iwl_tx_queue_init" removes
GFP_DMA from allocation tx command buffers. GFP_DMA allows allocation
only for memory under 16M which causes allocation problems
suspend/resume flows.

Using kmalloc is temporal solution and some consistent/coherent
allocation schema will be more correct. Since iwlwifi hardware
supports 64bit address this solution should work on x86 (32 and
64bit) for now.

This patch fixes memory freeing problem in the previous patch.

Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Ian Schram <ischram@telenet.be>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix rx_chain computation

This patch fixes rx_chain computation. The code that adjusts number of
rx chains to number supported by HW was missing. Miss configuration
causes firmware error.  Note: iwlwifi supports HW with up to 3 RX
chains (2x2, 2x3, 1x2, and 3x3 MIMO). This patch also simplifies the
whole RX chain computation.

Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Mohamed Abbas <mohamed.abbas@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix station mimo power save values

This patch fixes the wrong use MIMO power save values. Our TX was
configured with our MIMO power save values instead of peer's MIMO power
save values, this may affect connectivity. The peer STA/AP may not sense
our traffic at all as it doesn't have all RX chains opened.

Signed-off-by: Ron Rindjunsky <ron.rindjunsky@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: remove false rxon if rx chain changes

Rx chain might change during power save transitions but it doesn't
require sending Full-ROXN command to the firmware. Full-RXON requires
reconnection to an AP and thus affects user experience. The patch
avoids the Full-RXON by removing the rx_chain modification check in
iwl_full_rxon_required function.

Signed-off-by: Mohamed Abbas <mohamed.abbas@intel.com
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix hidden ssid discovery in passive channels

This enables sending of direct probes on passive channels, as long as
traffic was detected on that channel. This enables connectivity to
hidden/non broadcasting SSIDs APs on passive channels. Note 5000 HW
declares all 5.2 spectrum as passive.

Signed-off-by: Cahill Ben <ben.m.cahill@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Ron Rindjunsky <ron.rindjunsky@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: W/A for the TSF correction in IBSS

This patch is a W/A for the TSF sync issue in IBSS merging. HW is not
capable to sync TSF (it's constantly little behind). This creates
constant IBSS merging upon reception of each beacon, adding and removing
station which in turn creates above 50% packet loss and thus dramatically
degrade the throughput. The W/A simply stops the driver from declaring it
has a reliable TSF value and thus eliminates IBSS merging.

Signed-off-by: Assaf Krauss <assaf.krauss@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

tipc: Don't use structure names which easily globally conflict.

Andrew Morton reported a build failure on sparc32, because TIPC
uses names like "struct node" and there is a like named data
structure defined in linux/node.h

This just regexp replaces "struct node*" to "struct tipc_node*"
to avoid this and any future similar problems.

Signed-off-by: David S. Miller <davem@davemloft.net>

ipsec: Fix deadlock in xfrm_state management.

Ever since commit 4c563f7669c10a12354b72b518c2287ffc6ebfb3
("[XFRM]: Speed up xfrm_policy and xfrm_state walking") it is
illegal to call __xfrm_state_destroy (and thus xfrm_state_put())
with xfrm_state_lock held.  If we do, we'll deadlock since we
have the lock already and __xfrm_state_destroy() tries to take
it again.

Fix this by pushing the xfrm_state_put() calls after the lock
is dropped.

Signed-off-by: David S. Miller <davem@davemloft.net>

ipv: Re-enable IP when MTU > 68

Re-enable IP when the MTU gets back to a valid size.

This patch just checks if the in_dev is NULL on a NETDEV_CHANGEMTU event
and if MTU is valid (bigger than 68), then re-enable in_dev.

Also a function that checks valid MTU size was created.

Signed-off-by: Breno Leitao <leitao@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/xfrm: Use an IS_ERR test rather than a NULL test

In case of error, the function xfrm_bundle_create returns an ERR
pointer, but never returns a NULL pointer. So a NULL test that comes
after an IS_ERR test should be deleted.

The semantic match that finds this problem is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@match_bad_null_test@
expression x, E;
statement S1,S2;
@@
x =  xfrm_bundle_create(...)
... when != x = E
*  if (x != NULL)
S1 else S2
// </smpl>

Signed-off-by: Julien Brunel <brunel@diku.dk>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>

ath9: Fix ath_rx_flush_tid() for IRQs disabled kernel warning message.

This patch addresses an issue with the locking order. ath_rx_flush_tid()
uses spin_lock/unlock_bh when IRQs are disabled in sta_notify by mac80211.

As node clean up is still pending with ath9k and this problematic portion
of the code is expected to change anyway, thinking of a proper fix may not
be worthwhile. So having this interim fix helps the users to get rid of the
kernel warning message.

Pasted the kernel warning message for reference.

kernel: ath0: No ProbeResp from current AP 00:1b:11:60:7a:3d - assume out of range
kernel: ------------[ cut here ]------------
kernel: WARNING: at kernel/softirq.c:136 local_bh_enable+0x3c/0xab()
kernel: Pid: 1029, comm: ath9k Not tainted 2.6.27-rc4-wt-w1fi-wl
kernel:
kernel: Call Trace:
kernel:  [<ffffffff802278d8>] warn_on_slowpath+0x51/0x77
kernel:  [<ffffffff80224c51>] check_preempt_wakeup+0xf3/0x123
kernel:  [<ffffffff80239658>] autoremove_wake_function+0x9/0x2e
kernel:  [<ffffffff8022c281>] local_bh_enable+0x3c/0xab
kernel:  [<ffffffffa01ab75a>] ath_rx_node_cleanup+0x38/0x6e [ath9k]
kernel:  [<ffffffffa01b2280>] ath_node_detach+0x3b/0xb6 [ath9k]
kernel:  [<ffffffffa01ab09f>] ath9k_sta_notify+0x12b/0x165 [ath9k]
kernel:  [<ffffffff802366cf>] queue_work+0x1d/0x49
kernel:  [<ffffffffa018c3fc>] add_todo+0x70/0x99 [mac80211]
kernel:  [<ffffffffa017de76>] __sta_info_unlink+0x16b/0x19e [mac80211]
kernel:  [<ffffffffa017e6ed>] sta_info_unlink+0x18/0x43 [mac80211]
kernel:  [<ffffffffa0182732>] ieee80211_associated+0xaa/0x16d [mac80211]
kernel:  [<ffffffffa0184a1a>] ieee80211_sta_work+0x4fb/0x6b4 [mac80211]
kernel:  [<ffffffff80469c58>] thread_return+0x30/0xa9
kernel:  [<ffffffffa018451f>] ieee80211_sta_work+0x0/0x6b4 [mac80211]
kernel:  [<ffffffff802362c2>] run_workqueue+0xb1/0x17a
kernel:  [<ffffffff80236be9>] worker_thread+0xd0/0xdb
kernel:  [<ffffffff8023964f>] autoremove_wake_function+0x0/0x2e
kernel:  [<ffffffff80236b19>] worker_thread+0x0/0xdb
kernel:  [<ffffffff8023954a>] kthread+0x47/0x75
kernel:  [<ffffffff80223121>] schedule_tail+0x18/0x50
kernel:  [<ffffffff8020bc49>] child_rip+0xa/0x11
kernel:  [<ffffffff80239503>] kthread+0x0/0x75
kernel:  [<ffffffff8020bc3f>] child_rip+0x0/0x11
kernel:
kernel: ---[ end trace e9bb5da661055827 ]---

Signed-off-by: Senthil Balasubramanian <senthilkumar@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

ath9k: Incorrect key used when group and pairwise ciphers are different.

Updating sc_keytype multiple times when groupwise and pairwise
ciphers are different results in incorrect pairwise key type
assumed for TX control and normal ping fails. This works fine
for cases where both groupwise and pairwise ciphers are same.

Also use mac80211 provided enums for key length calculation.

Signed-off-by: Senthil Balasubramanian <senthilkumar@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

rt2x00: Compiler warning unmasked by fix of BUILD_BUG_ON

A "Set" to a sign-bit in an "&" operation causes a compiler warning.
Make calculations unsigned.

[ The warning was masked by the old definition of BUILD_BUG_ON() ]

Also remove __builtin_constant_p from FIELD_CHECK since BUILD_BUG_ON
no longer permits non-const values.

Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
CC: Ingo Molnar <mingo@elte.hu>
CC: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Ivo van Doorn <IvDoorn@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: Fix debugfs union misuse and pointer corruption

debugfs union in struct ieee80211_sub_if_data is misused by including a
common default_key dentry as a union member. This ends occupying the same
memory area with the first dentry in other union members (structures;
usually drop_unencrypted). Consequently, debugfs operations on
default_key symlinks and drop_unencrypted entry are using the same
dentry pointer even though they are supposed to be separate ones. This
can lead to removing entries incorrectly or potentially leaving
something behind since one of the dentry pointers gets lost.

Fix this by moving the default_key dentry to a new struct
(common_debugfs) that contains dentries (more to be added in future)
that are shared by all vif types. The debugfs union must only be used
for vif type-specific entries to avoid this type of pointer corruption.

Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

wireless/libertas/if_cs.c: fix memory leaks

The leak in if_cs_prog_helper() is obvious.

It looks a bit as if not freeing "fw" in if_cs_prog_real() was done
intentionally, but I'm not seeing why it shouldn't be freed.

Reported-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Acked-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

orinoco: Multicast to the specified addresses

When multicasting the driver sets the number of group addresses using
the count from the previous set multicast command. In general this means
you have to set the multicast addresses twice to get the behaviour you
want.

If we were multicasting, and reduce the number of addresses we are
multicasting to, then the driver would write uninitialised data from the
stack into the group addresses to multicast to.

Only write the multicast addresses we have specifically set.

Signed-off-by: David Kilroy <kilroyd@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix 64bit platform firmware loading

This patch fixes loading firmware from memory above 32bit.

Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Acked-by: Marcel Holtmann <holtmann@linux.intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: fix apm_stop (wrong bit polarity for FLAG_INIT_DONE)

The patch fixes CSR_GP_CNTRL_REG_FLAG_INIT_DONE was set instead of
cleared which disabled moving device to D0U state.

Signed-off-by: Mohamed Abbas <mohamed.abbas@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: workaround interrupt handling no some platforms

This patch adds workaround for an interrupt related hardware bug on
some platforms.  (Apparently these platforms boot-up w/ INTX_DISABLED
set. -- JWL)

Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Zhu Yi <yi.zhu@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

iwlwifi: do not use GFP_DMA in iwl_tx_queue_init

GFP_DMA is not necessary for the iwlwifi hardware and it can cause
allocation failures and/or invoke the OOM killer on lots of systems.

For reference:

https://bugzilla.redhat.com/show_bug.cgi?id=459709

Signed-off-by: John W. Linville <linville@tuxdriver.com>

net/wireless/Kconfig: clarify the description for CONFIG_WIRELESS_EXT_SYSFS

Current setup with hal and NetworkManager will fail to work
without newest hal version with this config option disabled.

Although this will solve itself by time, at the moment it is
dishonest to say that we don't know any software that uses it,
if there are many many people relying on old hal versions.

Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

net: Unbreak userspace usage of linux/mroute.h

Nothing in linux/pim.h should be exported to userspace.

This should fix the XORP build failure reported by
Jose Calhariz, the debain package maintainer.

Nothing originally in linux/mroute.h was exported to userspace
ever, but some of this stuff started to be when it was moved into
this new linux/pim.h, and that was wrong.  If we didn't provide these
definitions for 10 years we can reasonably expect that applications
defined this stuff locally or used GLIBC headers providing the
protocol definitions.  And as such the only result of this can
be conflict and userland build breakage.

Signed-off-by: David S. Miller <davem@davemloft.net>

pkt_sched: Fix locking of qdisc_root with qdisc_root_sleeping_lock()

Use qdisc_root_sleeping_lock() instead of qdisc_root_lock() where
appropriate. The only difference is while dev is deactivated, when
currently we can use a sleeping qdisc with the lock of noop_qdisc.
This shouldn't be dangerous since after deactivation root lock could
be used only by gen_estimator code, but looks wrong anyway.

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: When we droped a packet, we should return NET_RX_DROP instead of 0

Signed-off-by: Yang Hongyang <yanghy@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sctp: fix random memory dereference with SCTP_HMAC_IDENT option.

The number of identifiers needs to be checked against the option
length.  Also, the identifier index provided needs to be verified
to make sure that it doesn't exceed the bounds of the array.

Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sctp: correct bounds check in sctp_setsockopt_auth_key

The bonds check to prevent buffer overlflow was not exactly
right.  It still allowed overflow of up to 8 bytes which is
sizeof(struct sctp_authkey).

Since optlen is already checked against the size of that struct,
we are guaranteed not to cause interger overflow either.

Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

wan: Missing capability checks in sbni_ioctl()

There are missing capability checks in the following code:

1300 static int
1301 sbni_ioctl( struct net_device  *dev,  struct ifreq  *ifr,  int  cmd)
1302 {
[...]
1319     case  SIOCDEVRESINSTATS :
1320         if( current->euid != 0 )    /* root only */
1321             return  -EPERM;
[...]
1336     case  SIOCDEVSHWSTATE :
1337         if( current->euid != 0 )    /* root only */
1338             return  -EPERM;
[...]
1357     case  SIOCDEVENSLAVE :
1358         if( current->euid != 0 )    /* root only */
1359             return  -EPERM;
[...]
1372     case  SIOCDEVEMANSIPATE :
1373         if( current->euid != 0 )    /* root only */
1374             return  -EPERM;

Here's my proposed fix:

Missing capability checks.

Signed-off-by: Eugene Teo <eugeneteo@kernel.sg>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'no-iwlwifi' of git://git./linux/kernel/git/linville/wireless-2.6

Merge branch 'davem-fixes' of /linux/kernel/git/jgarzik/netdev-2.6

e100, fix iomap read

There were 2 omitted readb's used on an iomap space. eliminate them
by using ioread8 instead.

Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Cc: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Cc: Jesse Brandeburg <jesse.brandeburg@intel.com>
Cc: Bruce Allan <bruce.w.allan@intel.com>
Cc: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com>
Cc: John Ronciak <john.ronciak@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qeth: preallocated header account offset

When a preallocated header qdio buffer is filled we have to account
the offset for the data length.

Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qeth: l2 write unicast list to hardware

In case the netdev unicast list contains additional entries we have
to register/deregister them.

Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

qeth: use -EOPNOTSUPP instead of -ENOTSUPP.

return value -ENOTSUPP is not valid in userspace context, use
-EOPNOTSUPP instead.

Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Frank Blaschka <frank.blaschka@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ibm_newemac: Don't call dev_mc_add() before device is registered

We must not call dev_mc_add() from within our HW configure which happens
before we initialize and register the netdev. Do it in open() instead.

Thanks to Sebastian Siewior for tracking it down.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net: don't grab a mutex within a timer context in gianfar

I got the following backtrace while network was unavailble:

|NETDEV WATCHDOG: eth0: transmit timed out
|BUG: sleeping function called from invalid context at /home/bigeasy/git/linux-2.6-powerpc/kernel/mutex.c:87
|in_atomic():1, irqs_disabled():0
|Call Trace:
|[c0383d90] [c0006dd8] show_stack+0x48/0x184 (unreliable)
|[c0383db0] [c001e938] __might_sleep+0xe0/0xf4
|[c0383dc0] [c025a43c] mutex_lock+0x24/0x3c
|[c0383de0] [c019005c] phy_stop+0x20/0x70
|[c0383df0] [c018d4ec] stop_gfar+0x28/0xf4
|[c0383e10] [c018e8c4] gfar_timeout+0x30/0x60
|[c0383e20] [c01fe7c0] dev_watchdog+0xa8/0x144
|[c0383e30] [c002f93c] run_timer_softirq+0x148/0x1c8
|[c0383e60] [c002b084] __do_softirq+0x5c/0xc4
|[c0383e80] [c00046fc] do_softirq+0x3c/0x54
|[c0383e90] [c002ac60] irq_exit+0x3c/0x5c
|[c0383ea0] [c000b378] timer_interrupt+0xe0/0xf8
|[c0383ec0] [c000e5ac] ret_from_except+0x0/0x18
|[c0383f80] [c000804c] cpu_idle+0xcc/0xdc
|[c0383fa0] [c025c07c] etext+0x7c/0x90
|[c0383fc0] [c0338960] start_kernel+0x294/0x2a8
|[c0383ff0] [c00003dc] skpinv+0x304/0x340
|------------[ cut here ]------------

The phylock was once a spinlock but got changed into a mutex via
commit 35b5f6b1a aka [PHYLIB: Locking fixes for PHY I/O potentially sleeping]

Signed-off-by: Sebastian Siewior <bigeasy@linutronix.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

forcedeth: fix checksum flag

Fix the checksum feature advertised in device flags.  The hardware support
TCP/UDP over IPv4 and TCP/UDP over IPv6 (without IPv6 extension headers).
However, the kernel feature flags do not distinguish IPv6 with/without
extension headers.

Therefore, the driver needs to use NETIF_F_IP_CSUM instead of
NETIF_F_HW_CSUM since the latter includes all IPv6 packets.

A future patch can be created to check for extension headers and perform
software checksum calculation.

Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
Cc: Jeff Garzik <jgarzik@pobox.com>
Cc: Manfred Spraul <manfred@colorfullife.com
Cc: <stable@kernel.org> [2.6.25.x, 2.6.26.x]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net/usb/mcs7830: add set_mac_address

Implement set_mac_address for mcs7830. This enables me to use it with my
cable modem.

Signed-off-by: Oliver Martin <oliver.martin@student.tuwien.ac.at>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

net/usb/mcs7830: new device IDs

This adds USB device IDs for MosChip 7730 and Sitecom LN030
to the mcs7830 driver. The IDs have been reported to work without
further modifications.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: David Brownell <david-b@pacbell.net>
Cc: Viktor Horvath <ViktorHorvath@gmx.net>
Cc: Robbert Wethmar <robbert@wethmar.nl>
Cc: Bart van der Klip <bklip@xs4all.nl>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

[netdrvr] smc91x: fix resource removal (null ptr deref)

Properly handle resource cleanup on unplug/exit.

Spotted by Jonathan Cameron

Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ibmveth: fix bad UDP checksums

This patch fixes a ibmveth bug where bad UDP checksums are being transmitted
when checksum offloading is enabled.
The hypervisor does checksum offloading only on TCP packets, so ibmveth calls
skb_checksum_help() for any other protocol.  The bug happens because
the packet is being modified after the DMA map, so we would need a memory
barrier before making the hypervisor call.  Reordering the code so that the
DMA map happens after skb_checksum_help() has the additional advantage of
fixing a DMA map leak if skb_checksum_help() where to fail.

Signed-off-by: Santiago Leon <santil@us.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

[netdrvr] hso: dev_kfree_skb crash fix

Fixes dev_kfree_skb happening too many times when hso_start_net_device
is called from hso_resume.

Signed-off-by: Denis Joseph Barrow <D.Barow@option.com>
Cc: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

[netdrvr] hso: icon 322 detection fix

Fixes Icon-322 detection.

Signed-off-by: Denis Joseph Barrow <D.Barow@option.com>
Cc: Jeff Garzik <jgarzik@pobox.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

atl1: disable TSO by default

The atl1 driver is causing stalled connections and file corruption
whenever TSO is enabled.  Two examples are here:

http://lkml.org/lkml/2008/7/15/325
http://lkml.org/lkml/2008/8/18/543

Disable TSO by default until we can determine the source of the
problem.

Signed-off-by: Jay Cliburn <jacliburn@bellsouth.net>
cc: stable@kernel.org
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

atl1e: multistatement if missing braces

Doesn't cause problems (yet) because err gets zeroed earlier.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: remove 82576 quad adapter

Disable support for device 8086:10E8.  Currently the result of loading the
driver with the device present causes system instability.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

drivers/net/skfp/ess.c: fix compile warnings

  CC [M]  drivers/net/skfp/ess.o
drivers/net/skfp/ess.c: In function 'ess_send_response':
drivers/net/skfp/ess.c:513: warning: cast from pointer to integer of different size
drivers/net/skfp/ess.c: In function 'ess_send_alc_req':
drivers/net/skfp/ess.c:609: warning: cast from pointer to integer of different size
drivers/net/skfp/ess.c:639: warning: cast from pointer to integer of different size

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ipv4: mode 0555 in ipv4_skeleton

vpnc on today's kernel says Cannot open "/proc/sys/net/ipv4/route/flush":
d--------- 0 root root 0 2008-08-26 11:32 /proc/sys/net/ipv4/route
d--------- 0 root root 0 2008-08-26 19:16 /proc/sys/net/ipv4/neigh

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

tcp: fix tcp header size miscalculation when window scale is unused

The size of the TCP header is miscalculated when the window scale ends
up being 0. Additionally, this can be induced by sending a SYN to a
passive open port with a window scale option with value 0.

Signed-off-by: Philip Love <love_phil@emc.com>
Signed-off-by: Adam Langley <agl@imperialviolet.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

pkt_sched: Fix sch_tree_lock()

Use new qdisc_root_sleeping_lock() instead of qdisc_root_lock() as
sch_tree_lock() because this lock could be used while dev is
deactivated, but we never need to use this with noop_qdisc as a root.

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

pkt_sched: Fix gen_estimator locks

While passing a qdisc root lock to gen_new_estimator() and
gen_replace_estimator() dev could be deactivated or even before
grafting proper root qdisc as qdisc_sleeping (e.g. qdisc_create), so
using qdisc_root_lock() is not enough. This patch adds
qdisc_root_sleeping_lock() for this, plus additional checks, where
necessary.

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

pkt_sched: Use rcu_assign_pointer() to change dev_queue->qdisc

These pointers are RCU protected, so proper primitives should be used.

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

LCS recovery dumps when cable reconnect

LCS recovery dumps in irq routine when CCW address in
Subchannel Status Word (SCSW) is zero. This occurs
when recovery is driven after cable reconnect.

Signed-off-by: Klaus-D. Wacker <kdwacker@de.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ctcm: netdev->priv vs. netdev->ml_priv

Use netdev->ml_priv instead of netdev->priv

Signed-off-by: Peter Tiedemann <ptiedem@de.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

claw: netdev->priv vs. netdev->ml_priv

Use netdev->ml_priv instead of netdev->priv

Signed-off-by: Peter Tiedemann <ptiedem@de.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

claw: fix memory leak in claw_probe.

probe_error() frees memory only, if cgdev->dev.driver_data refers
to the claw_privbk structure. Move forward its setting in claw_probe()
to ensure proper freeing of claw_privbk allocations.

Cc: Daniel <danielm77@spray.se>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

[netdrvr] fix build issue: undefined reference to `NS8390p_init'

Signed-off-by: Alan 'pass the paper bags' Cox <alan@redhat.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

fs_enet: Fix SCC Ethernet on CPM2, and crash in fs_enet_rx_napi()

Signed-off-by: Heiko Schocher <hs@denx.de>
Signed-off-by: Vitaly Bordug <vitb@kernel.crashing.org>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: fix setting the number of tx queues

The real_num_tx_queues was not being set when in MSI-X only mode.  This patch
corrects that path so all interrupt types are correctly configured.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: ethtool -d reads EICR which is incorrect as it is read on clear

Ethtool -d is reading the EICR and ICR registers which is currently
clearing these registers and masking off interrupts.  To prevent this we
read the EICS and ICS equivilents as they can be read without clearing or
masking.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

igb: force all queues to interrupt once every 2 seconds

Set the EICS bit for each of the RX queues at least once every 2 seconds to
prevent the rx queues from stalling.

Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

r8169: balance pci_map / pci_unmap pair

The leak hurts with swiotlb and jumbo frames.

Fix http://bugzilla.kernel.org/show_bug.cgi?id=9468.

Heavily hinted by Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>.

Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Tested-by: Alistair John Strachan <alistair@devzero.co.uk>
Tested-by: Timothy J Fontaine <tjfontaine@atxconsulting.com>
Cc: Edward Hsu <edward_hsu@realtek.com.tw>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

pkt_sched: Fix dev_graft_qdisc() locking

During dev_graft_qdisc() dev is deactivated, so qdisc_root_lock()
returns wrong lock of noop_qdisc instead of qdisc_sleeping.

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

myri10ge: update version string to 1.4.3-1.358

Update myri10ge version string to 1.4.3-1.358.

Signed-off-by: Brice Goglin <brice@myri.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

ixgbe: fix vlan filtering

VLAN filtering is broken, due to reading the incorrect register for
the VLAN filtering settings.  Fixed by reading/writing the correct
register.

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Blackfin EMAC Driver: the BF526 also supports the MAC,
so update things accordingly

Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Merge branch 'for-2.6.27' of git://git.marvell.com/mv643xx_eth into upstream-fixes

mac80211: quiet chatty IBSS merge message

It seems obvious that this #ifndef should be the opposite polarity...

Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: don't send empty extended rates IE

The association request includes a list of supported data rates.

802.11b: 4 supported rates.
802.11g: 12 (8 + 4) supported rates.
802.11a: 8 supported rates.

The rates tag of the assoc request has room for only 8 rates. In case of
802.11g an extended rate tag is appended. However in net/wireless/mlme.c
an extended (empty) rate tag is also appended if the number of rates is
exact 8. This empty (length=0) extended rates tag causes some APs to
deny association with code 18 (unsupported rates). These APs include my
ZyXEL G-570U, and according to Tomas Winkler som Cisco APs.

'If count == 8' has been used to check for the need for an extended rates
tag. But count would also be equal to 8 if the for loop exited because of
no more supported rates. Therefore a check for count being less than
rates_len would seem more correct.

Thanks to:
 * Dan Williams for newbie guidance
 * Tomas Winkler for confirming the problem

Signed-off-by: Jan-Espen Pettersen <sigsegv@radiotube.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: Fix debugfs file add/del for netdev

Previous version was using incorrect union structures for non-AP
interfaces when adding and removing max_ratectrl_rateidx and
force_unicast_rateidx entries. Depending on the vif type, this ended
up in corrupting debugfs entries since the dentries inside different
union structures ended up going being on top of eachother.. As the
end result, debugfs files were being left behind with references to
freed data (instant kernel oops on access) and directories were not
removed properly when unloading mac80211 drivers. This patch fixes
those issues by using only a single union structure based on the vif
type.

Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

net/mac80211/mesh.c: correct the argument to __mesh_table_free

In the function mesh_table_grow, it is the new table not the argument table
that should be freed if the function fails (cf commit
bd9b448f4c0a514559bdae4ca18ca3e8cd999c6d)

The semantic match that detects this problem is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@r exists@
local idexpression x;
expression E,f;
position p1,p2,p3;
identifier l;
statement S;
@@

x = mesh_table_alloc@p1(...)
...
if (x == NULL) S
... when != E = x
    when != mesh_table_free(x)
goto@p2 l;
... when != E = x
    when != f(...,x,...)
    when any
(
return \(0\|x\);
|
return@p3 ...;
)

@script:python@
p1 << r.p1;
p2 << r.p2;
p3 << r.p3;
@@

print "%s: call on line %s not freed or saved before return on line %s via line %s" % (p1[0].file,p1[0].line,p3[0].line,p2[0].line)
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

mac80211: Use IWEVASSOCREQIE instead of IWEVCUSTOM

The previous code was using IWEVCUSTOM to report IEs from AssocReq and
AssocResp frames into user space. This can easily hit the 256 byte
limit (IW_CUSTOM_MAX) with APs that include number of vendor IEs in
AssocResp. This results in the event message not being sent and dmesg
showing "wlan0 (WE) : Wireless Event too big (366)" type of errors.

Convert mac80211 to use IWEVASSOCREQIE/IWEVASSOCRESPIE to avoid the
issue of being unable to send association IEs as wireless events. These
newer event types use binary encoding and larger maximum size
(IW_GENERIC_IE_MAX = 1024), so the likelyhood of not being able to send
the IEs is much smaller than with IWEVCUSTOM. As an extra benefit, the
code is also quite a bit simpler since there is no need to allocate an
extra buffer for hex encoding.

Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

net: rfkill: add missing line break

Trivial patch adding a missing line break on
rfkill_claim_show().

Signed-off-by: Felipe Balbi <felipe.balbi@nokia.com>
Acked-by: Ivo van Doorn <IvDoorn@gmail.co>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Ath5k: fix bintval setup

bintval is set to the initial value at .config_interface which is too
late, since it overwrites previously set value from .config. Move the
initialization to the .add_interface.

Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Cc: Nick Kossifidis <mickflemm@gmail.com>
Cc: Luis R. Rodriguez <mcgrof@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

Ath5k: lock beacons

Beacons setup and config was racy with beacon send. Ensure that
ISR and reset functions see consistent state of bbuf.

Use also dev_kfree_skb_any in ath5k_txbuf_free since we call it
from atomic now.

Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Cc: Nick Kossifidis <mickflemm@gmail.com>
Cc: Luis R. Rodriguez <mcgrof@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

atmel: try open system authentication too

When the AP rejects a Shared Key authentication request, try Open System
auth too.

Signed-off-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>

atmel: return ENOENT on request_firmware failure

Return errors from request_firmware() (like other drivers that do
firmware load on device open) and make up plausible codes for other
error conditions. Gives userspace tools like NetworkManager a clue that
firmware may be missing when the result of setting IFF_UP is ENOENT.

Signed-off-by: Dan Williams <dcbw@redhat.com>
v2: fix reversed check of atmel_wakeup_firmware() in probe_atmel_card()
Signed-off-by: John W. Linville <linville@tuxdriver.com>

bnx2x: Version update

Version update

Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: Multi Queue

The multi queue support is still disabled by default for the bnx2x
(needs some more testing and validation), but there are 2 obvious bug in
it which are fixed in this patch

Signed-off-by: Yitchak Gertner <gertner@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: NAPI and interrupts enable/disable

Fixing the order of enabling and disabling NAPI and the interrupts

Signed-off-by: Yitchak Gertner <gertner@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: NIC load failure cleanup

Load failures were not handled correctly

Signed-off-by: Yitchak Gertner <gertner@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: Initialization structure

The TPA initialization is part of the FW internal memory initialization
and so it is moved to the appropriate function

Signed-off-by: Yitchak Gertner <gertner@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: HW lock timeout

Increasing the lock timeout to 5 seconds instead of 1 second to minimize
the chance of failures due to timeout

Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: Minimize lock time

After iSCSI boot, the HW lock should only protect the flag so only the
first function will reset the chip and not then entire chip reset
process

Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: Fan failure mechanism on additional design

The A1021G board is also using the fan failure mechanism in the same way
the A1022G board does

Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

bnx2x: Rx work check

The has Rx work check was wrong: when the FW was at the end of the page,
the driver was already at the beginning of the next page. Since the
check only validated that both driver and FW are pointing to the same
place, it concluded that there is still work to be done. This caused
some serious issues including long latency results on ping-pong test and
lockups while unloading the driver in that condition.

Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv6: sysctl fixes

Braino: net.ipv6 in ipv6 skeleton has no business in rotable
class

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

ipv4: sysctl fixes

net.ipv4.neigh should be a part of skeleton to avoid ordering problems

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

sctp: add verification checks to SCTP_AUTH_KEY option

The structure used for SCTP_AUTH_KEY option contains a
length that needs to be verfied to prevent buffer overflow
conditions.  Spoted by Eugene Teo <eteo@redhat.com>.

Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mv643xx_eth: bump version to 1.3

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>

mv643xx_eth: enforce multiple-of-8-bytes receive buffer size restriction

The mv643xx_eth hardware ignores the lower three bits of the buffer
size field in receive descriptors, causing the reception of full-sized
packets to fail at some MTUs.  Fix this by rounding the size of
allocated receive buffers up to a multiple of eight bytes.

While we are at it, add a bit of extra space to each receive buffer so
that we can handle multiple vlan tags on ingress.

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>

mv643xx_eth: fix NULL pointer dereference in rxq_process()

When we are low on memory, the assumption that every descriptor in the
receive ring will have an skbuff associated with it does not hold.

rxq_process() was assuming that if the receive descriptor it is working
on is not owned by the hardware, it can safely be processed and handed
to the networking stack.  But a descriptor in the receive ring not being
owned by the hardware can also happen when we are low on memory and did
not manage to refill the receive ring fully.

This patch changes rxq_process()'s bailout condition from "the first
receive descriptor to be processed is owned by the hardware" to "the
first receive descriptor to be processed is owned by the hardware OR
the number of valid receive descriptors in the ring is zero".

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>

mv643xx_eth: fix inconsistent lock semantics

Nicolas Pitre noted that mv643xx_eth_poll was incorrectly using
non-IRQ-safe locks while checking whether to wake up the netdevice's
transmit queue.  Convert the locking to *_irq() variants, since we
are running from softirq context where interrupts are enabled.

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>

mv643xx_eth: fix double add_timer() on the receive oom timer

Commit 12e4ab79cd828563dc090d2117dc8626b344bc8f ("mv643xx_eth: be
more agressive about RX refill") changed the condition for the receive
out-of-memory timer to be scheduled from "the receive ring is empty"
to "the receive ring is not full".

This can lead to a situation where the receive out-of-memory timer is
pending because a previous rxq_refill() didn't manage to refill the
receive ring entirely as a result of being out of memory, and
rxq_refill() is then called again as a side effect of a packet receive
interrupt, and that rxq_refill() call then again does not succeed to
refill the entire receive ring with fresh empty skbuffs because we are
still out of memory, and then tries to call add_timer() on the already
scheduled out-of-memory timer.

This patch fixes this issue by changing the add_timer() call in
rxq_refill() to a mod_timer() call.  If the OOM timer was not already
scheduled, this will behave as before, whereas if it was already
scheduled, this patch will push back its firing time a bit, which is
safe because we've (unsuccessfully) attempted to refill the receive
ring just before we do this.

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>

mv643xx_eth: fix NAPI 'rotting packet' issue

When a receive interrupt occurs, mv643xx_eth would first process the
receive descriptors and then ACK the receive interrupt, instead of the
other way round.

This would leave a small race window between processing the last
receive descriptor and clearing the receive interrupt status in which
a new packet could come in, which would then 'rot' in the receive
ring until the next receive interrupt would come in.

Fix this by ACKing (clearing) the receive interrupt condition before
processing the receive descriptors.

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>