8 years agotipc: Prune down link-specific debugging code
Allan Stephens [Fri, 31 Dec 2010 18:59:27 +0000]
tipc: Prune down link-specific debugging code

Eliminates most link-specific debugging code in TIPC, which is now
largely unnecessary. All calls to the link-specific debugging macros
have been removed, as are the macros themselves; in addition, the optional
allocation of print buffers to hold debugging information for each link
endpoint has been removed. The ability for TIPC to print out helpful
diagnostic information when link retransmit failures occur has been
retained for the time being, as an aid in tracking down the cause of
such failures.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: remove dump() and tipc_dump_dbg()
Allan Stephens [Fri, 31 Dec 2010 18:59:26 +0000]
tipc: remove dump() and tipc_dump_dbg()

Eliminates calls to two debugging macros that are being completely obsoleted,
as well as any associated debugging routines that are no longer required.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: remove calls to dbg() and msg_dbg()
Allan Stephens [Fri, 31 Dec 2010 18:59:25 +0000]
tipc: remove calls to dbg() and msg_dbg()

Eliminates obsolete calls to two of TIPC's main debugging macros, as well
as a pair of associated debugging routines that are no longer required.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: rename dbg.[ch] to log.[ch]
Allan Stephens [Fri, 31 Dec 2010 18:59:24 +0000]
tipc: rename dbg.[ch] to log.[ch]

As the first step in removing obsolete debugging code from TIPC the
files that implement TIPC's non-debug-related log buffer subsystem
are renamed to better reflect their true nature.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Remove internal linked list of node objects
Allan Stephens [Fri, 31 Dec 2010 18:59:23 +0000]
tipc: Remove internal linked list of node objects

Eliminates a sorted list TIPC uses to keep track of the neighboring
nodes it has links to, since this duplicates information already present
in the internal array of node object pointers.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Remove user registry subsystem
Allan Stephens [Fri, 31 Dec 2010 18:59:22 +0000]
tipc: Remove user registry subsystem

Eliminates routines, data structures, and files that make up TIPC's
user registry. The user registry is no longer needed since the native
API routines that utilized it no longer exist and there are no longer
any internal TIPC services that use it.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Eliminate use of user registry by topology service
Allan Stephens [Fri, 31 Dec 2010 18:59:21 +0000]
tipc: Eliminate use of user registry by topology service

Simplifies TIPC's network topology service so that it no longer registers
its ports with the user registry, since the service doesn't take advantage
of any of the registry's capabilities.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Eliminate use of user registry by configuration service
Allan Stephens [Fri, 31 Dec 2010 18:59:20 +0000]
tipc: Eliminate use of user registry by configuration service

Simplifies TIPC's configuration service so that it no longer registers
its port with the user registry, since the service doesn't take advantage
of any of the registry's capabilities.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Remove prototype code for supporting multiple clusters
Allan Stephens [Fri, 31 Dec 2010 18:59:19 +0000]
tipc: Remove prototype code for supporting multiple clusters

Eliminates routines, data structures, and files that were intended
to allow TIPC to support a network containing multiple clusters.
Currently, TIPC supports only networks consisting of a single cluster
within a single zone, so this code is unnecessary.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Remove prototype code for supporting inter-cluster routing
Allan Stephens [Fri, 31 Dec 2010 18:59:18 +0000]
tipc: Remove prototype code for supporting inter-cluster routing

Eliminates routines and data structures that were intended to allow
TIPC to route messages to other clusters. Currently, TIPC supports only
networks consisting of a single cluster within a single zone, so this
code is unnecessary.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Remove prototype code for supporting slave nodes
Allan Stephens [Fri, 31 Dec 2010 18:59:17 +0000]
tipc: Remove prototype code for supporting slave nodes

Simplifies routines and data structures that were intended to allow
TIPC to support slave nodes (i.e. nodes that did not have links to
all of the other nodes in its cluster, forcing TIPC to route messages
that it could not deliver directly through a non-slave node).

Currently, TIPC supports only networks containing non-slave nodes,
so this code is unnecessary.

Note: The latest edition of the TIPC 2.0 Specification has eliminated
the concept of slave nodes entirely.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotipc: Remove prototype code for supporting multiple zones
Allan Stephens [Fri, 31 Dec 2010 18:59:16 +0000]
tipc: Remove prototype code for supporting multiple zones

Eliminates routines, data structures, and files that were intended
to allows TIPC to support a network containing multiple zones.
Currently, TIPC supports only networks consisting of a single cluster
within a single zone, so this code is unnecessary.

Signed-off-by: Allan Stephens <Allan.Stephens@windriver.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotg3: fix warnings
Eric Dumazet [Sat, 1 Jan 2011 05:22:46 +0000]
tg3: fix warnings

In case CONFIG_PM_SLEEP is disabled, we dont need tg3_suspend() and
tg3_resume().

drivers/net/tg3.c:15056: warning: ‘tg3_suspend’ defined but not used
drivers/net/tg3.c:15110: warning: ‘tg3_resume’ defined but not used

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Michael Chan <mchan@broadcom.com>
Cc: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoskge: Do not use legacy PCI power management
stephen hemminger [Thu, 30 Dec 2010 08:52:29 +0000]
skge: Do not use legacy PCI power management

The skge driver used the legacy PCI power management, and did its
own PCI callbacks.  Use the same code model as Rafael's changes to
sky2. Let the PCI subsystem take care of all the PCI-specific aspects of
device handling during system power transitions.

Compile tested only (so far).

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Acked-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosfq: fix slot_dequeue_head()
Eric Dumazet [Fri, 31 Dec 2010 20:48:55 +0000]
sfq: fix slot_dequeue_head()

slot_dequeue_head() should make sure slot skb chain is correct in both
ways, or we can crash if all possible flows are in use.

Jarek pointed out slot_queue_init() can now be done in sfq_init() once,
instead each time a flow is setup.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosch_sfq: allow big packets and be fair
Eric Dumazet [Tue, 28 Dec 2010 21:53:33 +0000]
sch_sfq: allow big packets and be fair

SFQ is currently 'limited' to small packets, because it uses a 15bit
allotment number per flow. Introduce a scale by 8, so that we can handle
full size TSO/GRO packets.

Use appropriate handling to make sure allot is positive before a new
packet is dequeued, so that fairness is respected.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Jarek Poplawski <jarkao2@gmail.com>
Cc: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosky2: Do not use legacy PCI power management
Rafael J. Wysocki [Sun, 26 Dec 2010 08:44:32 +0000]
sky2: Do not use legacy PCI power management

The sky2 driver uses the legacy PCI power management, so it has to do
some PCI-specific things in its ->suspend() and ->resume() callbacks,
which isn't necessary and should better be done by the PCI
sybsystem-level power management code.  Moreover, it uses
device_set_wakeup_enable() incorrectly (that function should be
used when the WoL setting is changed rather than during suspend).

Convert sky2 to the new PCI power management framework and make it
let the PCI subsystem take care of all the PCI-specific aspects of
device handling during system power transitions.

Tested on a desktop machine with a Marvell 88E8056 PCI-E adapter.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotg3: Do not use legacy PCI power management
Rafael J. Wysocki [Sat, 25 Dec 2010 12:56:23 +0000]
tg3: Do not use legacy PCI power management

The tg3 driver uses the legacy PCI power management, so it has to do
some PCI-specific things in its ->suspend() and ->resume() callbacks,
which isn't necessary and should better be done by the PCI
sybsystem-level power management code.

Convert tg3 to the new PCI power management framework and make it
let the PCI subsystem take care of all the PCI-specific aspects of
device handling during system power transitions.

Tested on HP nx6325 with a NetXtreme BCM5788 adapter.

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoUSB: cdc_ether: remove unneeded check
Dan Carpenter [Sat, 25 Dec 2010 12:23:42 +0000]
USB: cdc_ether: remove unneeded check

We already verified that "dev->udev->actconfig->extralen" was non-zero
so "len" is non-zero here as well.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobnx2: Update firmware and version
Michael Chan [Fri, 31 Dec 2010 19:04:02 +0000]
bnx2: Update firmware and version

Update MIPS firmware to 6.2.1, with improved small packet performance
in RSS mode, and iSCSI CID allocation bug fix on 5708.

Update driver version to 2.0.21.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobnx2: Free IRQ before freeing status block memory
Michael Chan [Fri, 31 Dec 2010 19:03:14 +0000]
bnx2: Free IRQ before freeing status block memory

When changing ring size, we free all memory including status block
memory.  If we're in INTA mode and sharing IRQ, the IRQ handler can
be called and it will reference the NULL status block pointer.

Because of the lockless design of the IRQ handler, there is no simple
way to synchronize and prevent this.  So we avoid this problem by
freeing the IRQ handler before freeing the status block memory.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonetxen: update module description
Amit Kumar Salecha [Fri, 31 Dec 2010 18:55:38 +0000]
netxen: update module description

This driver supports only Intelligent Ethernet Adapters.

Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agodcbnl: cleanup
Shmulik Ravid [Thu, 30 Dec 2010 06:27:10 +0000]
dcbnl: cleanup

A couple of small cleanups for patches:
[net-next-2.6 PATCH 1/3] dcbnl: add support for ieee8021Qaz attributes
[net-next-2.6 PATCH 2/3] dcbnl: add appliction tlv handlers
[net-next-2.6 PATCH 3/3] net_dcb: add application notifiers

Signed-off-by: Shmulik Ravid <shmulikr@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobnx2x: adding dcbnl support
Shmulik Ravid [Thu, 30 Dec 2010 06:27:03 +0000]
bnx2x: adding dcbnl support

Adding dcbnl implementation to bnx2x allowing users to manage the
embedded DCBX engine.

This patch is dependent on the following patches:
[net-next-2.6 PATCH 1/3] dcbnl: add support for ieee8021Qaz attributes
[net-next-2.6 PATCH 2/3] dcbnl: add appliction tlv handlers
[net-next-2.6 PATCH 3/3] net_dcb: add application notifiers

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

8 years agodcbnl: adding DCBX feature flags get-set
Shmulik Ravid [Thu, 30 Dec 2010 06:26:55 +0000]
dcbnl: adding DCBX feature flags get-set

Adding a pair of set-get routines to dcbnl for setting the negotiation
flags of the various DCB features. Conforms to the CEE flavor of DCBX
The user sets these flags (enable, advertise, willing) for each feature
to be used by the DCBX engine. The 'get' routine returns which of the
features is enabled after the negotiation.

This patch is dependent on the following patches:
[net-next-2.6 PATCH 1/3] dcbnl: add support for ieee8021Qaz attributes
[net-next-2.6 PATCH 2/3] dcbnl: add appliction tlv handlers
[net-next-2.6 PATCH 3/3] net_dcb: add application notifiers

Signed-off-by: Shmulik Ravid <shmulikr@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agodcbnl: adding DCBX engine capability
Shmulik Ravid [Thu, 30 Dec 2010 06:26:48 +0000]
dcbnl: adding DCBX engine capability

Adding an optional DCBX capability and a pair for get-set routines for
setting the device DCBX mode. The DCBX capability is a bit field of
supported attributes. The user is expected to set the DCBX mode with a
subset of the advertised attributes.

This patch is dependent on the following patches:
[net-next-2.6 PATCH 1/3] dcbnl: add support for ieee8021Qaz attributes
[net-next-2.6 PATCH 2/3] dcbnl: add appliction tlv handlers
[net-next-2.6 PATCH 3/3] net_dcb: add application notifiers

Signed-off-by: Shmulik Ravid <shmulikr@broadcom.com>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet_dcb: add application notifiers
John Fastabend [Thu, 30 Dec 2010 09:26:37 +0000]
net_dcb: add application notifiers

DCBx applications priorities can be changed dynamically. If
application stacks are expected to keep the skb priority
consistent with the dcbx priority the stack will need to
be notified when these changes occur.

This patch adds application notifiers for the stack to register
with.

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agodcbnl: add appliction tlv handlers
John Fastabend [Thu, 30 Dec 2010 09:26:31 +0000]
dcbnl: add appliction tlv handlers

This patch adds application tlv handlers. Networking stacks
may use the application priority to set the skb priority of
their stack using the negoatiated dcbx priority.

This patch provides the dcb_{get|set}app() routines for the
stack to query these parameters. Notice lower layer drivers
can use the dcbnl_ops routines if additional handling is
needed. Perhaps in the firmware case for example

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: Shmulik Ravid <shmulikr@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agodcbnl: add support for ieee8021Qaz attributes
John Fastabend [Thu, 30 Dec 2010 09:25:46 +0000]
dcbnl: add support for ieee8021Qaz attributes

The IEEE8021Qaz is the IEEE standard version of CEE. The
standard has had enough significant changes from the CEE
version that many of the CEE attributes have no meaning
in the new spec or do not easily map to IEEE standards.

Rather then attempt to create a complicated mapping
between CEE and IEEE standards this patch adds a nested
IEEE attribute to the list of DCB attributes. The policy
is,

[DCB_ATTR_IFNAME]
[DCB_ATTR_STATE]
...
[DCB_ATTR_IEEE]
[DCB_ATTR_IEEE_ETS]
[DCB_ATTR_IEEE_PFC]
[DCB_ATTR_IEEE_APP_TABLE]
[DCB_ATTR_IEEE_APP]
...

The following dcbnl_rtnl_ops routines were added to handle
the IEEE standard,

int (*ieee_getets) (struct net_device *, struct ieee_ets *);
int (*ieee_setets) (struct net_device *, struct ieee_ets *);
int (*ieee_getpfc) (struct net_device *, struct ieee_pfc *);
int (*ieee_setpfc) (struct net_device *, struct ieee_pfc *);
int (*ieee_getapp) (struct net_device *, struct dcb_app *);
int (*ieee_setapp) (struct net_device *, struct dcb_app *);

Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agovxge: remove duplicated part of check
Dan Carpenter [Thu, 23 Dec 2010 20:15:39 +0000]
vxge: remove duplicated part of check

This is just a cleanup to make the static checkers happy.  We don't need
to check "own" twice.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-by: Ram Vepa <ram.vepa@exar.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
David S. Miller [Mon, 27 Dec 2010 06:37:05 +0000]
Merge branch 'master' of /linux/kernel/git/davem/net-2.6

Conflicts:
net/ipv4/fib_frontend.c

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Sun, 26 Dec 2010 20:06:56 +0000]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (42 commits)
  ipv4: dont create routes on down devices
  epic100: hamachi: yellowfin: Fix skb allocation size
  sundance: Fix oopses with corrupted skb_shared_info
  Revert "ipv4: Allow configuring subnets as local addresses"
  USB: mcs7830: return negative if auto negotiate fails
  irda: prevent integer underflow in IRLMP_ENUMDEVICES
  tcp: fix listening_get_next()
  atl1c: Do not use legacy PCI power management
  mac80211: fix mesh forwarding
  MAINTAINERS: email address change
  net: Fix range checks in tcf_valid_offset().
  net_sched: sch_sfq: fix allot handling
  hostap: remove netif_stop_queue from init
  mac80211/rt2x00: add ieee80211_tx_status_ni()
  typhoon: memory corruption in typhoon_get_drvinfo()
  net: Add USB PID for new MOSCHIP USB ethernet controller MCS7832 variant
  net_sched: always clone skbs
  ipv6: Fragment locally generated tunnel-mode IPSec6 packets as needed.
  netlink: fix gcc -Wconversion compilation warning
  asix: add USB ID for Logitec LAN-GTJ U2A
  ...

8 years agoipv4: dont create routes on down devices
Eric Dumazet [Wed, 22 Dec 2010 04:39:39 +0000]
ipv4: dont create routes on down devices

In ip_route_output_slow(), instead of allowing a route to be created on
a not UPed device, report -ENETUNREACH immediately.

# ip tunnel add mode ipip remote 10.16.0.164 local
10.16.0.72 dev eth0
# (Note : tunl1 is down)
# ping -I tunl1 10.1.2.3
PING 10.1.2.3 (10.1.2.3) from 192.168.18.5 tunl1: 56(84) bytes of data.
(nothing)
# ./a.out tunl1
# ip tunnel del tunl1
Message from syslogd@shelby at Dec 22 10:12:08 ...
  kernel: unregister_netdevice: waiting for tunl1 to become free.
Usage count = 3

After patch:
# ping -I tunl1 10.1.2.3
connect: Network is unreachable

Reported-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reviewed-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoepic100: hamachi: yellowfin: Fix skb allocation size
Jarek Poplawski [Sat, 25 Dec 2010 07:39:59 +0000]
epic100: hamachi: yellowfin: Fix skb allocation size

Joel Soete reported oopses during pppoe over sundance NIC, caused by
a bug in skb allocation and dma mapping code, where skb_reserve()
bytes weren't taken into account. As a followup to the patch:
"sundance: Fix oopses with corrupted skb_shared_info" very similar
code is fixed here for three other drivers.

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Cc: Joel Soete <soete.joel@scarlet.be>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosundance: Fix oopses with corrupted skb_shared_info
Jarek Poplawski [Sat, 25 Dec 2010 05:12:17 +0000]
sundance: Fix oopses with corrupted skb_shared_info

Joel Soete reported oopses at the beginning of pppoe connections since
v2.6.35. After debugging the bug was found in sundance skb allocation
and dma mapping code, where skb_reserve() bytes aren't taken into
account. This is an old bug, only uncovered by some change in 2.6.35.

Initial debugging patch by: Eric Dumazet <eric.dumazet@gmail.com>

Reported-by: Joel Soete <soete.joel@scarlet.be>
Tested-by: Joel Soete <soete.joel@scarlet.be>
Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/jkirsher/net...
David S. Miller [Sun, 26 Dec 2010 03:20:38 +0000]
Merge branch 'master' of /linux/kernel/git/jkirsher/net-next-2.6

8 years agobna: Update the driver version to 2.3.2.3
Rasesh Mody [Thu, 23 Dec 2010 21:45:10 +0000]
bna: Update the driver version to 2.3.2.3

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: IOC failure auto recovery fix
Rasesh Mody [Thu, 23 Dec 2010 21:45:09 +0000]
bna: IOC failure auto recovery fix

Change Details:
- Made IOC auto_recovery synchronized and not timer based.
- Only one PCI function will attempt to recover and reinitialize
the ASIC on a failure, that too after all the active PCI
functions acknowledge the IOC failure.

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: Restore VLAN filter table
Rasesh Mody [Thu, 23 Dec 2010 21:45:08 +0000]
bna: Restore VLAN filter table

Change Details:
- Retrieve the VLAN configuration from the networking stack
and apply it to the base interface during ifconfig up

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: Removed unused code
Rasesh Mody [Thu, 23 Dec 2010 21:45:07 +0000]
bna: Removed unused code

Change Details:
- Remove unused APIs and code cleanup

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: IOC uninit check and misc cleanup
Rasesh Mody [Thu, 23 Dec 2010 21:45:06 +0000]
bna: IOC uninit check and misc cleanup

Change Details:
- Added a check in ioc firmware lock function to see if IOC is
in BFI_IOC_UNINIT state or not. If it is not in UNINIT state
and the last IOC boot was not done by OS driver, force IOC state
to BFI_IOC_UNINIT
- Unused macro and API cleanup

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: Fix for TX queue
Rasesh Mody [Thu, 23 Dec 2010 21:45:05 +0000]
bna: Fix for TX queue

Change Details:
- Call netif_wake_queue() if we have freed up sufficient elements
at the end of completion processing
- Add netif_queue_stopped counter back to bnad_drv_stats {}
- Get netif_queue_stopped value from stack
- Remove BUG_ON() on value returned by pci_unmap_addr()

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: Enable pure priority tagged packet reception and rxf uninit cleanup fix
Rasesh Mody [Thu, 23 Dec 2010 21:45:04 +0000]
bna: Enable pure priority tagged packet reception and rxf uninit cleanup fix

Change Details:
- Enable reception of pure priority tagged packets by default by
turning on VLAN Id = 0
- Clear the promiscuous mode, all multicast mode flags when
bna_rxf_uninit is called

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: Fix ethtool register dump and reordered an API
Rasesh Mody [Thu, 23 Dec 2010 21:45:03 +0000]
bna: Fix ethtool register dump and reordered an API

Change Details:
- Removed semaphore register dump from ethtool
- Moved netif_carrier_off() call to before calling bna_init()

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: Port enable disable sync and txq priority fix
Rasesh Mody [Thu, 23 Dec 2010 21:45:02 +0000]
bna: Port enable disable sync and txq priority fix

Change Details:
- Fixed port enable/disable sync through a change in LL port state
machine
- Change txq->priority masking to 0x7 (3 bits) from 0x3 (2 bits)

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobna: TxRx and datapath fix
Rasesh Mody [Thu, 23 Dec 2010 21:45:01 +0000]
bna: TxRx and datapath fix

Change Details:
- Check HW ready condition before accessing h/w register in data-path
- Postpone clean-up of data buffers to the data-path restart path and
wait in the cleanup routines for in-flight DMA to complete
- Separate out Tx completion processing from Rx poll routine

Signed-off-by: Debashis Dutt <ddutt@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoixgbevf: Add X540 VF device support to the ixgbevf driver
Greg Rose [Thu, 2 Dec 2010 07:12:26 +0000]
ixgbevf: Add X540 VF device support to the ixgbevf driver

The X540 introduces a new Virtual Function device ID so that the X540
VF device can be distinguished from the 82599 VF device.  The X540 VF
device will have additional capability over the 82599 VF device so it
is necessary to be able to discern the difference.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoixgbe: Add anti-spoofing feature support
Greg Rose [Thu, 18 Nov 2010 03:02:52 +0000]
ixgbe: Add anti-spoofing feature support

Add support for the anti-spoofing feature in the HW.  Packets from
VF devices with spoofed MAC addresses or VLAN tags will be blocked
and a counter incremented.  During the watchdog timer the spoofed
packet dropped counter is read and if it is non-zero then a warning
message is displayed on the host VMM's console.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoixgbe: Add SR-IOV feature support to X540
Greg Rose [Tue, 7 Dec 2010 08:16:45 +0000]
ixgbe: Add SR-IOV feature support to X540

Add X540 specific feature support to X540

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoixgbe: Warn on VF attempt to override Administratively set MAC/VLAN
Greg Rose [Thu, 18 Nov 2010 03:03:23 +0000]
ixgbe: Warn on VF attempt to override Administratively set MAC/VLAN

Print warnings to the system log when the VF attempts to override
MAC/VLAN settings that were configured by the VMM Host administrator
using the ip link set commands.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoigbvf: add support for i350 VF device
Williams, Mitch A [Thu, 9 Dec 2010 03:23:56 +0000]
igbvf: add support for i350 VF device

Add support to igbvf for the new i350 virtual function device.

Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoigbvf: force link checking when mailbox timeout has occurred
Alexander Duyck [Thu, 18 Nov 2010 03:12:30 +0000]
igbvf: force link checking when mailbox timeout has occurred

This change forces the link down when a mailbox timeout has occurred.
Previously it was possible for a mailbox timeout to occur but for the
interface to stay up.  The problem with this was that it became possible
for an interface to stay up and miss multiple requests resulting in a
possible issue since the interface will be running in an unknown state.

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

8 years agoigb: Add Anti-spoofing feature support
Greg Rose [Sat, 6 Nov 2010 02:08:26 +0000]
igb: Add Anti-spoofing feature support

Add support for the anti-spoofing feature in the HW.  Packets from
VF devices with spoofed MAC addresses or VLAN tags will be blocked
and an event generated.  When the watchdog task runs it will call a
function to check if any spoof events occurred.  If an event was
detected then a warning message is dumped to the system log.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoigb: Some fine tuning
Gasparakis, Joseph [Thu, 9 Dec 2010 01:41:01 +0000]
igb: Some fine tuning

This patch does the following:

1. Changes the existing supported device id's so now DH89xxCC is not supported when EEPROM is not read.
2. Adds two more device ids for DH89xxCC in backplane mode and SFP.
3. Driver now initializes previously possibly uninitialized value in igb_reset_mdicnfg_82580().

Signed-off-by: Joseph Gasparakis <joseph.gasparakis@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoigb: Warn on attempt to override administratively set MAC/VLAN
Greg Rose [Sat, 6 Nov 2010 05:42:59 +0000]
igb: Warn on attempt to override administratively set MAC/VLAN

Print a warning message to the system log when the VF attempts to
override administratively set MAC/VLAN configuration.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoigb: Fix overwrite of the VF's flags
Greg Rose [Sat, 6 Nov 2010 05:43:21 +0000]
igb: Fix overwrite of the VF's flags

The vf flags were being overwritten causing the flag that indicates
that the PF has set the VF MAC address to get cleared.  This would
allow the VF to override the MAC address assigned by the Host VMM
using the ip link set command.

Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoe1000e: checkpatch warnings - braces
Bruce Allan [Sat, 11 Dec 2010 05:53:47 +0000]
e1000e: checkpatch warnings - braces

WARNING: braces {} are not necessary for any arm of this statement
WARNING: braces {} are not necessary for single statement blocks

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoe1000e: checkpatch error - open braces
Bruce Allan [Sat, 11 Dec 2010 05:53:42 +0000]
e1000e: checkpatch error - open braces

ERROR: that open brace { should be on the previous line

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoe1000e: checkpatch error - trailing statements
Bruce Allan [Sat, 11 Dec 2010 05:53:37 +0000]
e1000e: checkpatch error - trailing statements

ERROR: trailing statements should be on next line

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoe1000e: checkpatch error - macro panethesis
Bruce Allan [Sat, 11 Dec 2010 05:53:32 +0000]
e1000e: checkpatch error - macro panethesis

ERROR: Macros with complex values should be enclosed in parenthesis

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

8 years agoDocumentation/networking: Update Intel Wired LAN docs
Jeff Kirsher [Fri, 17 Dec 2010 12:14:34 +0000]
Documentation/networking: Update Intel Wired LAN docs

- Update the Intel Wired LAN documentation with the latest
  URL for ethtool.

- replace "Ethtool" with "ethtool"

Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Tested-by: Emil Tantilov <emil.s.tantilov@intel.com>

8 years agoMerge branch 'merge' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Fri, 24 Dec 2010 21:00:37 +0000]
Merge branch 'merge' of git://git.secretlab.ca/git/linux-2.6

* 'merge' of git://git.secretlab.ca/git/linux-2.6:
  spi/mpc52xx-spi: fix annotation for remove()-pointer
  spi/fsl_espi: fix wrong setting of the address in the command buffer
  spi/fsl_espi: change the read behaviour of the SPIRF
  of/i2c: Fix request module by alias
  powerpc/mpc5200: include fs.h in mpc52xx_gpt.c

8 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzi...
Linus Torvalds [Fri, 24 Dec 2010 21:00:16 +0000]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  libata: issue DIPM enable commands with LPM state updated
  libata: no special completion processing for EH commands
  pata_mpc52xx: driver needs BMDMA
  pata_cs5536: Add support for non-X86_32 platforms
  libata-sff: fix HSM_ST_ERR handling in __ata_sff_port_intr()

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu
Linus Torvalds [Fri, 24 Dec 2010 20:59:09 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/tj/percpu

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu:
  percpu: print out alloc information with KERN_DEBUG instead of KERN_INFO
  kthread_work: make lockdep happy

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6
Linus Torvalds [Fri, 24 Dec 2010 20:58:43 +0000]
Merge git://git./linux/kernel/git/jejb/scsi-rc-fixes-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6:
  [SCSI] fix up documentation for change in ->queuecommand to lockless calling
  [SCSI] bfa: rename log_level to bfa_log_level

8 years agolibata: issue DIPM enable commands with LPM state updated
Tejun Heo [Thu, 9 Dec 2010 15:13:15 +0000]
libata: issue DIPM enable commands with LPM state updated

Low level drivers may behave differently depending on the current
link->lpm_policy.  During ata_eh_set_lpm(), DIPM enable commands are
issued after the successful completion of ap->ops->set_lpm(), which
means that the controller is already in the target state.  This causes
DIPM enable commands to be processed with mismatching controller power
state and link->lpm_policy value.

In ahci, link->lpm_policy is used to ignore certain PHY events if LPM
is enabled; however, as DIPM commands are issued with stale
link->lpm_policy, they sometimes end up triggering these conditions
and get aborted leading to LPM configuration failure.

Fix it by updating link->lpm_policy before issuing DIPM enable
commands.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Kyle McMartin <kyle@mcmartin.ca>
Cc: stable@kernel.org
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agolibata: no special completion processing for EH commands
Tejun Heo [Thu, 9 Dec 2010 14:59:32 +0000]
libata: no special completion processing for EH commands

ata_qc_complete() contains special handling for certain commands.  For
example, it schedules EH for device revalidation after certain
configurations are changed.  These shouldn't be applied to EH
commands but they were.

In most cases, it doesn't cause an actual problem because EH doesn't
issue any command which would trigger special handling; however, ACPI
can issue such commands via _GTF which can cause weird interactions.

Restructure ata_qc_complete() such that EH commands are always passed
on to __ata_qc_complete().

stable: Please apply to -stable only after 2.6.38 is released.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Kyle McMartin <kyle@mcmartin.ca>
Cc: stable@kernel.org
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agopata_mpc52xx: driver needs BMDMA
Wolfram Sang [Wed, 22 Dec 2010 15:50:10 +0000]
pata_mpc52xx: driver needs BMDMA

Found by this build-error if BMDMA is disabled:

drivers/ata/pata_mpc52xx.c: In function 'mpc52xx_ata_init_one':
drivers/ata/pata_mpc52xx.c:662: error: 'ata_bmdma_interrupt' undeclared (first use in this function)
...

Move the Kconfig entry to the proper location as needed since
9a7780c9acb821fe1c2b6fc53f74cc2556ff5364 (libata-sff: make BMDMA optional)

Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agopata_cs5536: Add support for non-X86_32 platforms
Wu Zhangjin [Fri, 24 Dec 2010 13:39:08 +0000]
pata_cs5536: Add support for non-X86_32 platforms

pata_cs5536 does work on the other platforms(e.g. Loongson, a MIPS
variant), so, remove the dependency of X86_32 and fix the building
errors under the other platforms via only reserving the X86_32 specific
parts for X86_32.

pata_amd also supports cs5536 IDE controller, but this one saves about
33k for the compressed kernel image(vmlinuz for MIPS).

Signed-off-by: Zhang Le <r0bertz@gentoo.org>
Signed-off-by: Chen Jie <chenj@lemote.com>
Signed-off-by: Wu Zhangjin <wuzhangjin@gmail.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agolibata-sff: fix HSM_ST_ERR handling in __ata_sff_port_intr()
Tejun Heo [Fri, 3 Dec 2010 14:19:13 +0000]
libata-sff: fix HSM_ST_ERR handling in __ata_sff_port_intr()

While separating out BMDMA irq handler from SFF, commit c3b28894
(libata-sff: separate out BMDMA irq handler) incorrectly made
__ata_sff_port_intr() consider an IRQ to be an idle one if the host
state was transitioned to HSM_ST_ERR by ata_bmdma_port_intr().

This makes BMDMA drivers ignore IRQs reporting host bus error which
leads to timeouts instead of triggering EH immediately.  Fix it by
making __ata_sff_port_intr() consider the IRQ to be an idle one iff
the state is HSM_ST_IDLE.  This is equivalent to adding HSM_ST_ERR to
the "break"ing case but less error-prone.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Antonio Toma <antonio.toma@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

8 years agospi/mpc52xx-spi: fix annotation for remove()-pointer
Wolfram Sang [Fri, 17 Dec 2010 14:33:58 +0000]
spi/mpc52xx-spi: fix annotation for remove()-pointer

Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

8 years agospi/fsl_espi: fix wrong setting of the address in the command buffer
Mingkai Hu [Tue, 21 Dec 2010 01:27:02 +0000]
spi/fsl_espi: fix wrong setting of the address in the command buffer

Or else we can't operate on the right address when the trans length
is greater than 65535.

Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

8 years agospi/fsl_espi: change the read behaviour of the SPIRF
Mingkai Hu [Tue, 21 Dec 2010 01:26:07 +0000]
spi/fsl_espi: change the read behaviour of the SPIRF

The user must read N bytes of SPIRF (1 <= N <= 4) that do not exceed the
amount of data in the receive FIFO, so read the SPIRF byte by byte when
the data in receive FIFO is less than 4 bytes.

On Simics, when read N bytes that exceed the amount of data in receive
FIFO, we can't read the data out, that is we can't clear the rx FIFO,
then the CPU will loop on the espi rx interrupt.

Signed-off-by: Mingkai Hu <Mingkai.hu@freescale.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

8 years agoof/i2c: Fix request module by alias
David Daney [Tue, 16 Nov 2010 22:42:14 +0000]
of/i2c: Fix request module by alias

If we are registering an i2c device that has a device tree node like
this real-world example:

      rtc@68 {
        compatible = "dallas,ds1337";
        reg = <0x68>;
      };

of_i2c_register_devices() will try to load a module called ds1337.ko.
There is no such module, so it will fail.  If we look in modules.alias
we will find entries like these:

.
.
.
alias i2c:ds1339 rtc_ds1307
alias i2c:ds1338 rtc_ds1307
alias i2c:ds1337 rtc_ds1307
alias i2c:ds1307 rtc_ds1307
alias i2c:ds1374 rtc_ds1374
.
.
.

The module we want is really called rtc_ds1307.ko.  If we request a
module called "i2c:ds1337", the userspace module loader will do the
right thing (unless it is busybox) and load rtc_ds1307.ko.  So we add
the I2C_MODULE_PREFIX to the request_module() string.

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

8 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec...
Linus Torvalds [Fri, 24 Dec 2010 00:36:48 +0000]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jlbec/ocfs2

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2:
  ocfs2: Fix system inodes cache overflow.
  ocfs2: Hold ip_lock when set/clear flags for indexed dir.
  ocfs2: Adjust masklog flag values
  Ocfs2: Teach 'coherency=full' O_DIRECT writes to correctly up_read i_alloc_sem.
  ocfs2/dlm: Migrate lockres with no locks if it has a reference

8 years agoMerge branch 'linus-hot-fix' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso...
Linus Torvalds [Fri, 24 Dec 2010 00:25:31 +0000]
Merge branch 'linus-hot-fix' of git://git./linux/kernel/git/tytso/ext4

* 'linus-hot-fix' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: fix on-line resizing regression

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Fri, 24 Dec 2010 00:04:32 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Fix GPIO2-fixup for Sony laptops
  ALSA: hda - Try to find an empty control index when it's occupied
  ALSA: hda - Fix conflict of d-mic capture volume controls
  ALSA: hda - Don't apply ALC269-specific initialization to ALC275
  ALSA: hda - Add fix-up for Sony VAIO with ALC275 codecs
  ALSA: pcm: remember to always call va_end() on stuff that we va_start()
  ALSA: HDA: Add auto-mute for Thinkpad SL410/SL510

8 years agoMerge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Thu, 23 Dec 2010 23:59:23 +0000]
Merge branch 'v4l_for_linus' of git://git./linux/kernel/git/mchehab/linux-2.6

* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6: (21 commits)
  [media] mceusb: set a default rx timeout
  [media] mceusb: fix inverted mask inversion logic
  [media] mceusb: add another Fintek device ID
  [media] lirc_dev: fixes in lirc_dev_fop_read()
  [media] lirc_dev: stray unlock in lirc_dev_fop_poll()
  [media] rc: fix sysfs entry for mceusb and streamzap
  [media] streamzap: merge timeout space with trailing space
  [media] mceusb: fix keybouce issue after parser simplification
  [media] IR: add tv power scancode to rc6 mce keymap
  [media] mceusb: buffer parsing fixups for 1st-gen device
  [media] mceusb: fix up reporting of trailing space
  [media] nuvoton-cir: improve buffer parsing responsiveness
  [media] mceusb: add support for Conexant Hybrid TV RDU253S
  [media] s5p-fimc: Fix output DMA handling in S5PV310 IP revisions
  [media] s5p-fimc: Use correct fourcc code for 32-bit RGB format
  [media] s5p-fimc: Convert m2m driver to unlocked_ioctl
  [media] s5p-fimc: Explicitly add required header file
  [media] s5p-fimc: Fix vidioc_g_crop/cropcap on camera sensor
  [media] s5p-fimc: BKL lock removal - compilation fix
  [media] soc-camera: fix static build of the sh_mobile_csi2.c driver
  ...

8 years agoMerge branches 'perf-fixes-for-linus' and 'x86-fixes-for-linus' of git://git.kernel...
Linus Torvalds [Thu, 23 Dec 2010 23:39:40 +0000]
Merge branches 'perf-fixes-for-linus' and 'x86-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  perf probe: Fix to support libdwfl older than 0.148
  perf tools: Fix lazy wildcard matching
  perf buildid-list: Fix error return for success
  perf buildid-cache: Fix symbolic link handling
  perf symbols: Stop using vmlinux files with no symbols
  perf probe: Fix use of kernel image path given by 'k' option

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, kexec: Limit the crashkernel address appropriately

8 years agoKEYS: Don't call up_write() if __key_link_begin() returns an error
David Howells [Wed, 22 Dec 2010 16:24:13 +0000]
KEYS: Don't call up_write() if __key_link_begin() returns an error

In construct_alloc_key(), up_write() is called in the error path if
__key_link_begin() fails, but this is incorrect as __key_link_begin() only
returns with the nominated keyring locked if it returns successfully.

Without this patch, you might see the following in dmesg:

=====================================
[ BUG: bad unlock balance detected! ]
-------------------------------------
mount.cifs/5769 is trying to release lock (&key->sem) at:
[<ffffffff81201159>] request_key_and_link+0x263/0x3fc
but there are no more locks to release!

other info that might help us debug this:
3 locks held by mount.cifs/5769:
 #0:  (&type->s_umount_key#41/1){+.+.+.}, at: [<ffffffff81131321>] sget+0x278/0x3e7
 #1:  (&ret_buf->session_mutex){+.+.+.}, at: [<ffffffffa0258e59>] cifs_get_smb_ses+0x35a/0x443 [cifs]
 #2:  (root_key_user.cons_lock){+.+.+.}, at: [<ffffffff81201000>] request_key_and_link+0x10a/0x3fc

stack backtrace:
Pid: 5769, comm: mount.cifs Not tainted 2.6.37-rc6+ #1
Call Trace:
 [<ffffffff81201159>] ? request_key_and_link+0x263/0x3fc
 [<ffffffff81081601>] print_unlock_inbalance_bug+0xca/0xd5
 [<ffffffff81083248>] lock_release_non_nested+0xc1/0x263
 [<ffffffff81201159>] ? request_key_and_link+0x263/0x3fc
 [<ffffffff81201159>] ? request_key_and_link+0x263/0x3fc
 [<ffffffff81083567>] lock_release+0x17d/0x1a4
 [<ffffffff81073f45>] up_write+0x23/0x3b
 [<ffffffff81201159>] request_key_and_link+0x263/0x3fc
 [<ffffffffa026fe9e>] ? cifs_get_spnego_key+0x61/0x21f [cifs]
 [<ffffffff812013c5>] request_key+0x41/0x74
 [<ffffffffa027003d>] cifs_get_spnego_key+0x200/0x21f [cifs]
 [<ffffffffa026e296>] CIFS_SessSetup+0x55d/0x1273 [cifs]
 [<ffffffffa02589e1>] cifs_setup_session+0x90/0x1ae [cifs]
 [<ffffffffa0258e7e>] cifs_get_smb_ses+0x37f/0x443 [cifs]
 [<ffffffffa025a9e3>] cifs_mount+0x1aa1/0x23f3 [cifs]
 [<ffffffff8111fd94>] ? alloc_debug_processing+0xdb/0x120
 [<ffffffffa027002c>] ? cifs_get_spnego_key+0x1ef/0x21f [cifs]
 [<ffffffffa024cc71>] cifs_do_mount+0x165/0x2b3 [cifs]
 [<ffffffff81130e72>] vfs_kern_mount+0xaf/0x1dc
 [<ffffffff81131007>] do_kern_mount+0x4d/0xef
 [<ffffffff811483b9>] do_mount+0x6f4/0x733
 [<ffffffff8114861f>] sys_mount+0x88/0xc2
 [<ffffffff8100ac42>] system_call_fastpath+0x16/0x1b

Reported-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-and-Tested-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agocs5535-gpio: handle GPIO regs where higher (clear) bits are set
Andres Salomon [Tue, 21 Dec 2010 21:04:52 +0000]
cs5535-gpio: handle GPIO regs where higher (clear) bits are set

The default for non-READ_BACK GPIO regs is to have the clear bits set;
this means that our original errata fix was too simplistic.  This
changes it to the following behavior:

 - when setting GPIOs, ignore the higher order bits (they're for
   clearing, we don't need to care about them).

 - when clearing GPIOs, keep all the bits, but unset (via XOR) the
   lower order bit that negates the clear bit that we care about.  That
   is, if we're clearing GPIO 26 (val = 0x04000000), we first XOR what's
   currently in the register with 0x0400 (GPIO 26's SET bit), and then
   OR that with the GPIO 26's CLEAR bit.

Tested-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Andres Salomon <dilinger@queued.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agocs5535-gpio: don't apply errata #36 to edge detect GPIOs
Andres Salomon [Tue, 21 Dec 2010 21:04:42 +0000]
cs5535-gpio: don't apply errata #36 to edge detect GPIOs

The edge detect status GPIOs function differently from the other atomic
model CS5536 GPIO registers; writing 1 to the high bits clears the GPIO,
but writing 1 to the lower bits also clears the bit.

This means that read-modify-write doesn't actually work for it, so don't
apply the errata here.  If a negative edge status gets lost after
resume..  well, we tried our best!

Tested-by: Daniel Drake <dsd@laptop.org>
Signed-off-by: Andres Salomon <dilinger@queued.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoRevert "ipv4: Allow configuring subnets as local addresses"
David S. Miller [Thu, 23 Dec 2010 20:03:57 +0000]
Revert "ipv4: Allow configuring subnets as local addresses"

This reverts commit 4465b469008bc03b98a1b8df4e9ae501b6c69d4b.

Conflicts:

net/ipv4/fib_frontend.c

As reported by Ben Greear, this causes regressions:

> Change 4465b469008bc03b98a1b8df4e9ae501b6c69d4b caused rules
> to stop matching the input device properly because the
> FLOWI_FLAG_MATCH_ANY_IIF is always defined in ip_dev_find().
>
> This breaks rules such as:
>
> ip rule add pref 512 lookup local
> ip rule del pref 0 lookup local
> ip link set eth2 up
> ip -4 addr add 172.16.0.102/24 broadcast 172.16.0.255 dev eth2
> ip rule add to 172.16.0.102 iif eth2 lookup local pref 10
> ip rule add iif eth2 lookup 10001 pref 20
> ip route add 172.16.0.0/24 dev eth2 table 10001
> ip route add unreachable 0/0 table 10001
>
> If you had a second interface 'eth0' that was on a different
> subnet, pinging a system on that interface would fail:
>
>   [root@ct503-60 ~]# ping 192.168.100.1
>   connect: Invalid argument

Reported-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoext4: fix on-line resizing regression
Theodore Ts'o [Tue, 21 Dec 2010 03:30:36 +0000]
ext4: fix on-line resizing regression

https://bugzilla.kernel.org/show_bug.cgi?id=25352

This regression was caused by commit a31437b85: "ext4: use
sb_issue_zeroout in setup_new_group_blocks", by accidentally dropping
the code which reserved the block group descriptor and inode table
blocks.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>

8 years agocnic: Add FCoE support on 57712
Michael Chan [Thu, 23 Dec 2010 07:43:04 +0000]
cnic: Add FCoE support on 57712

- Connection ID (cid) management
- Slow-path command and response support
- Update version to 2.2.11.

Reviewed-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Add kcq2 support on 57712
Michael Chan [Thu, 23 Dec 2010 07:43:03 +0000]
cnic: Add kcq2 support on 57712

The kcq2 (2nd kernel work queue) is used by FCoE on 57712 devices.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Call cm_connect_complete() immediately on error
Eddie Wai [Thu, 23 Dec 2010 07:43:02 +0000]
cnic: Call cm_connect_complete() immediately on error

If we get a path_resp error from userspace, call cm_connect_complete()
immediately with error so that bnx2i can react to the error faster.

Signed-off-by: Eddie Wai <waie@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Check device state before reading the kcq pointer in IRQ
Michael Chan [Thu, 23 Dec 2010 08:38:30 +0000]
cnic: Check device state before reading the kcq pointer in IRQ

If the device is down, the kcq pointer may be NULL.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Support NIC Partition mode
Michael Chan [Thu, 23 Dec 2010 07:43:01 +0000]
cnic: Support NIC Partition mode

Add a common function cnic_read_bnx2x_iscsi_mac() to read the iSCSI
MAC address at any specified shared memory location.  In NIC Partition
mode, we need to get the MAC address from the MF_CFG area of shared
memory.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Use proper client and connection IDs on iSCSI ring
Michael Chan [Thu, 23 Dec 2010 07:42:59 +0000]
cnic: Use proper client and connection IDs on iSCSI ring

Use the IDs specified by the bnx2x driver when initializing the ring.
We don't have to make code changes when these IDs change in the future.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Improve ->iscsi_nl_msg_send()
Michael Chan [Thu, 23 Dec 2010 07:42:58 +0000]
cnic: Improve ->iscsi_nl_msg_send()

1. Change first parameter from cnic_dev to ulp_handle which is the hba
pointer.  All other similar upcalls are using hba pointer.  The callee
can then directly reference the hba without conversion.

2. Change return value from void to int so that an error code can be
passed back.  This allows the operation to be retried.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Prevent "scheduling while atomic" when calling ->cnic_init()
Michael Chan [Thu, 23 Dec 2010 07:42:57 +0000]
cnic: Prevent "scheduling while atomic" when calling ->cnic_init()

cnic_dev_list is protected by rtnl_lock and cnic_dev_lock spin_lock during
modifications.  When looping on cnic_dev_list and calling ->cnic_init(),
we should just hold rtnl_lock since ->cnic_init() may sleep.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agocnic: Fix iSCSI TCP port endian order.
Michael Chan [Thu, 23 Dec 2010 07:42:56 +0000]
cnic: Fix iSCSI TCP port endian order.

Pass the TCP port parameter for iSCSI connections to the firmware in
proper endian order.

Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agopowerpc/mpc5200: include fs.h in mpc52xx_gpt.c
Wolfram Sang [Wed, 22 Dec 2010 15:42:55 +0000]
powerpc/mpc5200: include fs.h in mpc52xx_gpt.c

Fix build errors like these (from a randconfig and my defconfig for a custom board):

src/arch/powerpc/platforms/52xx/mpc52xx_gpt.c:549: error: dereferencing pointer to incomplete type: 1 errors in 1 logs
src/arch/powerpc/platforms/52xx/mpc52xx_gpt.c:636: error: implicit declaration of function 'nonseekable_open': 1 errors in 1 logs
src/arch/powerpc/platforms/52xx/mpc52xx_gpt.c:657: error: variable 'mpc52xx_wdt_fops' has initializer but incomplete type: 1 errors in 1 logs
src/arch/powerpc/platforms/52xx/mpc52xx_gpt.c:658: error: excess elements in struct initializer: 1 errors in 1 logs
src/arch/powerpc/platforms/52xx/mpc52xx_gpt.c:658: error: unknown field 'owner' specified in initializer: 1 errors in 1 logs
...

Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Wolfram Sang <w.sang@pengutronix.de>
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

8 years agoUSB: mcs7830: return negative if auto negotiate fails
Dan Carpenter [Fri, 17 Dec 2010 03:25:43 +0000]
USB: mcs7830: return negative if auto negotiate fails

The original code returns 0 on success and 1 on failure.  In fact, at
this point, "ret" is already either zero or a negative error code so
we can just return it directly.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge branch 'for-davem' of ssh://master.kernel.org/pub/scm/linux/kernel/git/linville...
David S. Miller [Thu, 23 Dec 2010 18:13:30 +0000]
Merge branch 'for-davem' of ssh:///linux/kernel/git/linville/wireless-next-2.6

8 years agoirda: prevent integer underflow in IRLMP_ENUMDEVICES
Dan Rosenberg [Wed, 22 Dec 2010 13:58:27 +0000]
irda: prevent integer underflow in IRLMP_ENUMDEVICES

If the user-provided len is less than the expected offset, the
IRLMP_ENUMDEVICES getsockopt will do a copy_to_user() with a very large
size value.  While this isn't be a security issue on x86 because it will
get caught by the access_ok() check, it may leak large amounts of kernel
heap on other architectures.  In any event, this patch fixes it.

Signed-off-by: Dan Rosenberg <drosenberg@vsecurity.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotcp: cleanup of cwnd initialization in tcp_init_metrics()
Jiri Kosina [Wed, 22 Dec 2010 23:23:38 +0000]
tcp: cleanup of cwnd initialization in tcp_init_metrics()

Commit 86bcebafc5e7f5 ("tcp: fix >2 iw selection") fixed a case
when congestion window initialization has been mistakenly omitted
by introducing cwnd label and putting backwards goto from the
end of the function.

This makes the code unnecessarily tricky to read and understand
on a first sight.

Shuffle the code around a little bit to make it more obvious.

Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agotcp: fix listening_get_next()
Eric Dumazet [Thu, 23 Dec 2010 17:32:46 +0000]
tcp: fix listening_get_next()

Alexey Vlasov found /proc/net/tcp could sometime loop and display
millions of sockets in LISTEN state.

In 2.6.29, when we converted TCP hash tables to RCU, we left two
sk_next() calls in listening_get_next().

We must instead use sk_nulls_next() to properly detect an end of chain.

Reported-by: Alexey Vlasov <renton@renton.name>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge branch 'fix/hda' into for-linus
Takashi Iwai [Thu, 23 Dec 2010 15:37:31 +0000]
Merge branch 'fix/hda' into for-linus