safe/jmp/linux-2.6
14 years agocxgb3: use request_firmware() for the EDC registers setup
Divy Le Ray [Tue, 7 Jul 2009 19:49:09 +0000 (19:49 +0000)]
cxgb3: use request_firmware() for the EDC registers setup

use request_firmware() to load the phy's EDC programmation

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agocxgb3: AEL2020 phy support update
Divy Le Ray [Tue, 7 Jul 2009 19:48:43 +0000 (19:48 +0000)]
cxgb3: AEL2020 phy support update

We don't always see the link status update interrupt
when we come out of reset and the peer is up.
Check and report the link status right before enabling interrupts.
Also fix LED settings, to get a consistent link status.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agocxgb3: Fix T3C MAC max packet size access
Divy Le Ray [Tue, 7 Jul 2009 19:48:38 +0000 (19:48 +0000)]
cxgb3: Fix T3C MAC max packet size access

Max packet size is not the only field in T3C's High Water Mark register.
Mask the register to access this field.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agocxgb3: fix phy power down
Divy Le Ray [Tue, 7 Jul 2009 19:48:53 +0000 (19:48 +0000)]
cxgb3: fix phy power down

2 phys are were not getting the Global Tx disable bit set
when powered down, leading to an inconsistent link state
on peer.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agocxgb3: AQ100X phy support update
Divy Le Ray [Tue, 7 Jul 2009 19:48:48 +0000 (19:48 +0000)]
cxgb3: AQ100X phy support update

Add missing SUPPORTED_TP flag.
Update FW version checking.
Do the full initialization even if the FW version is unknown,
it might help catching further issues.

Signed-off-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMIPS: BCM63XX: Add integrated ethernet PHY support for phylib.
Maxime Bizon [Wed, 1 Jul 2009 01:29:36 +0000 (01:29 +0000)]
MIPS: BCM63XX: Add integrated ethernet PHY support for phylib.

Signed-off-by: Maxime Bizon <mbizon@freebox.fr>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
 drivers/net/phy/Kconfig   |    6 ++
 drivers/net/phy/Makefile  |    1
 drivers/net/phy/bcm63xx.c |  132 ++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 139 insertions(+)
 create mode 100644 drivers/net/phy/bcm63xx.c
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agotun: Remove a dead line of code
Paul Moore [Tue, 7 Jul 2009 08:11:33 +0000 (08:11 +0000)]
tun: Remove a dead line of code

Remove an unnecessary assignment.

Signed-off-by: Paul Moore <paul.moore@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoucc_geth: Add support for skb recycling
Anton Vorontsov [Tue, 7 Jul 2009 08:38:42 +0000 (08:38 +0000)]
ucc_geth: Add support for skb recycling

We can reclaim transmitted skbs to use in the receive path, so-called
skb recycling support.

Also reorder ucc_geth_poll() steps, so that we'll clean tx ring firstly,
thus maybe reclaim some skbs for rx.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agodrivers/net/cs89x0.c: Avoid using magic number in set_dma_mode
Julia Lawall [Tue, 7 Jul 2009 02:09:50 +0000 (19:09 -0700)]
drivers/net/cs89x0.c: Avoid using magic number in set_dma_mode

The constant DMA_RX_MODE is defined to be 0x14 in the local include file
cs89x0.h.  Since a constant with the same name is used elsewhere with
set_dma_mode, it seems likely that this constant could be used here.

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

// <smpl>
@@ expression E1; identifier I; constant int C; @@
(
set_dma_mode(E1,I,...)
|
*set_dma_mode(E1,C,...)
)
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agodm9000: add checksum offload support
Yeasah Pell [Tue, 7 Jul 2009 01:12:33 +0000 (18:12 -0700)]
dm9000: add checksum offload support

Add checksum offload support for DM9000A and DM9000B chips.

v2 changes: added a local copy of ip_summed to save IO cycles in dm9000_send_packet
v3 changes: trans_start updating is removed.

Signed-off-by: Yeasah Pell <yeasah@comrex.com>
Signed-off-by: Mike Rapoport <mike@compulab.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoeconet: use NET_RX_SUCCESS instead of magic number 0 for econet_rcv successful return
Mark Smith [Mon, 6 Jul 2009 11:05:58 +0000 (11:05 +0000)]
econet: use NET_RX_SUCCESS instead of magic number 0 for econet_rcv successful return

Signed-off-by: Mark Smith <markzzzsmith@yahoo.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoipv6: correct return on ipv6_rcv() packet drop
Mark Smith [Mon, 6 Jul 2009 11:26:24 +0000 (11:26 +0000)]
ipv6: correct return on ipv6_rcv() packet drop

The routine ipv6_rcv() uses magic number 0 for a return when it drops a
packet. This corresponds to NET_RX_SUCCESS, which is obviously
incorrect. Correct this by using NET_RX_DROP instead.

ps. It isn't exactly clear who the IPv6 maintainers are, apologies if
I've missed any.

Signed-off-by: Mark Smith <markzzzsmith@yahoo.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000: fix flow control thresholds
Jesse Brandeburg [Mon, 6 Jul 2009 10:45:01 +0000 (10:45 +0000)]
e1000: fix flow control thresholds

when testing the jumbo frames with pages patch, the stats would
show rx_missed errors (dropped packets) even when connected to a
link partner with flow control enabled.

this indicates that for this MTU (9000) the flow control
thresholds are not adjusting correctly.

In fact, before this change, the FCRTH (xoff threshold) is 36864
when the fifo size is only 40000, with 9000 byte MTU.

fix it so that we at least have room for one frame after we send
the xoff.

Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000: implement jumbo receive with partial descriptors
Jesse Brandeburg [Mon, 6 Jul 2009 10:44:39 +0000 (10:44 +0000)]
e1000: implement jumbo receive with partial descriptors

This is code extremely similar to what is committed in e1000e
already.

e1000 will no longer request 32kB slab buffers to support jumbo
frames on PCI/PCI-X adapters.  This will significantly reduce the
likelyhood of order:3 allocation failures.

This new code adds support for using pages as receive buffers,
and the driver will chain multiple pages together to build a
jumbo frame for OS consumption.

The hardware takes a power of two buffer size and will
dump as much data as it can receive into 1 or more buffers.

The benefits of applying this are
1) stop akpm's dissing :-) of this lame e1000 behavior [1]
2) more efficient memory allocation (half) when using jumbo
   frames, which will also allow for much better socket utilization
   with jumbos since the socket is charged for the full allocation
   of each receive buffer, regardless of how much is used.
3) this was a feature request by a customer
4) copybreak for small packets < 256 bytes still applies

[1] http://lkml.org/lkml/2008/7/10/68
    http://article.gmane.org/gmane.linux.network/130986

Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
CC: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000: allow ethtool coalesece to adjust interrupts per second
Jesse Brandeburg [Mon, 6 Jul 2009 10:44:20 +0000 (10:44 +0000)]
e1000: allow ethtool coalesece to adjust interrupts per second

This patch allows on-the-fly adjustment of the interrupts per second generated
by e1000 devices 82545/82546 (hardware support of ITR register is a
requirement)

adjust using this command:
ethtool -C eth0 rx-usecs 10

where 10 is 10 microseconds per interrupt interval, so 10 = 100,000 interrupts
per second, and 125 = 8000 interrupts per second.

changes should be immediate.

1,3 are special values and indicate the automatic tuning mode to the driver,
where 1 is 4000-90000 interrupts per second and 3 is 4000-20000 interrupts
per second and is the driver default.

Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Advance driver version number
Guo-Fu Tseng [Mon, 6 Jul 2009 04:45:58 +0000 (04:45 +0000)]
jme: Advance driver version number

Advance driver version number after modifications.

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Tuning rxsum function
Guo-Fu Tseng [Mon, 6 Jul 2009 04:41:22 +0000 (04:41 +0000)]
jme: Tuning rxsum function

1. Use more efficient way to determine flag status.
2. Hardware dose not mark fragment bit against IPv6 packets,
   print TCP/UDP checksum warning message for IPv4 packets
   only.

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Remove shadow register support
Guo-Fu Tseng [Mon, 6 Jul 2009 04:40:38 +0000 (04:40 +0000)]
jme: Remove shadow register support

The hardware failed to update the shadow register in several cases.
It's known to be failed at 64bit box with more than 4G RAM.
In order to make the device working stable, give up using the shadow
register feature.

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Change bufinf memory location
Guo-Fu Tseng [Mon, 6 Jul 2009 04:39:46 +0000 (04:39 +0000)]
jme: Change bufinf memory location

Instead of using a large chunk of memory space preserved for
for modules, using kmalloc to obtain the needed memory.

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Fix typo
Guo-Fu Tseng [Mon, 6 Jul 2009 04:38:35 +0000 (04:38 +0000)]
jme: Fix typo

Fix enum typo in jme.h

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Fix unmatched tasklet_{enable|disable} pair
Guo-Fu Tseng [Mon, 6 Jul 2009 04:37:52 +0000 (04:37 +0000)]
jme: Fix unmatched tasklet_{enable|disable} pair

Fix faulty tasklet function usage.

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agojme: Some minor coding style consistency modifications
Guo-Fu Tseng [Mon, 6 Jul 2009 04:36:30 +0000 (04:36 +0000)]
jme: Some minor coding style consistency modifications

Making coding style more consistent.

Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: Guo-Fu Tseng <cooldavid@cooldavid.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agonet: convert remaining non-symbolic return values in ndo_start_xmit() functions
Patrick McHardy [Mon, 6 Jul 2009 02:23:38 +0000 (19:23 -0700)]
net: convert remaining non-symbolic return values in ndo_start_xmit() functions

This patch converts the remaining occurences of raw return values to their
symbolic counterparts in ndo_start_xmit() functions that were missed by the
previous automatic conversion.

Additionally code that assumed the symbolic value of NETDEV_TX_OK to be zero
is changed to explicitly use NETDEV_TX_OK.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Version update
Sreenivasa Honnur [Wed, 1 Jul 2009 21:18:43 +0000 (21:18 +0000)]
vxge: Version update

- Version update

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Printing the function's configured mode of operation
Sreenivasa Honnur [Wed, 1 Jul 2009 21:18:06 +0000 (21:18 +0000)]
vxge: Printing the function's configured mode of operation

- Printing the function's configured mode of operation.

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Disable fcs stripping.
Sreenivasa Honnur [Wed, 1 Jul 2009 21:17:24 +0000 (21:17 +0000)]
vxge: Disable fcs stripping.

- Disable fcs stripping. The minimum frame size that can be received by the
hardware is 57 Bytes. A 64 Byte Ethernet frame with the vlan tag and fcs
stripped will result in a 56 Byte frame which will lock up the receive
engine. The work around is to disable fcs stripping in the hardware which
is done with a firmware upgrade. The fixes are -
1. Ensure that the correct firmware version is used.
2. Decrement the indicated packet length of the receive packet by 4 bytes
(FCS length).

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Fixes in isr routine
Sreenivasa Honnur [Wed, 1 Jul 2009 21:16:37 +0000 (21:16 +0000)]
vxge: Fixes in isr routine

- Fixes in isr routine
   Fixed crash with INTA mode during driver unload. Pass hldev to request_irq
   instead of vdev.
   Fixed Traffic failure with GRO in INTA mode. Pass the correct napi
   context to gro functions.

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Removed the code to bounds check the mss value
Sreenivasa Honnur [Wed, 1 Jul 2009 21:15:41 +0000 (21:15 +0000)]
vxge: Removed the code to bounds check the mss value

- Removed the code to bounds check the mss value. The hardware does bounds checking and
will not allow an oversized mss to lockup the transmit path.

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Corrected Register map entry for xmac_link registers
Sreenivasa Honnur [Wed, 1 Jul 2009 21:14:57 +0000 (21:14 +0000)]
vxge: Corrected Register map entry for xmac_link registers

- Corrected Register map entry for xmac_link registers.

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: check for card status before continuing in device close
Sreenivasa Honnur [Wed, 1 Jul 2009 21:14:03 +0000 (21:14 +0000)]
vxge: check for card status before continuing in device close

- Fixed the crash in rmmod after vpath open failed when trying to change mtu.
We should check for card status before continuing in device close.

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Fixed memory leak by freeing memory allocated for device_config
Sreenivasa Honnur [Wed, 1 Jul 2009 21:13:12 +0000 (21:13 +0000)]
vxge: Fixed memory leak by freeing memory allocated for device_config

- Fixed memory leak by freeing memory allocated for device_config.

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agovxge: Removed ioremap of unused bar addresses and their references
Sreenivasa Honnur [Wed, 1 Jul 2009 21:12:23 +0000 (21:12 +0000)]
vxge: Removed ioremap of unused bar addresses and their references

- Removed ioremap of bar1 address
  Driver needs only bar0 address for register access
- Removed references to bar1 and bar2 addresses

Signed-off-by: Sreenivasa Honnur <sreenivasa.honnur@neterion.com>
Signed-off-by: Ramkrishna Vepa <ram.vepa@neterion.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agonet, netns_xt: shrink netns_xt members
Cyrill Gorcunov [Fri, 3 Jul 2009 20:11:58 +0000 (20:11 +0000)]
net, netns_xt: shrink netns_xt members

In case if kernel was compiled without ebtables support
there is no need to keep ebt_table pointers in netns_xt
structure.

Make it config dependent.

Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agonet, bridge: align br_nf_ops assignment
Cyrill Gorcunov [Fri, 3 Jul 2009 20:11:57 +0000 (20:11 +0000)]
net, bridge: align br_nf_ops assignment

No functional change -- just for easier reading.

Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agotipc: Add socket options to get number of queued messages
oscar.medina@motorola.com [Tue, 30 Jun 2009 03:25:39 +0000 (03:25 +0000)]
tipc: Add socket options to get number of queued messages

This patch allows a TIPC application to determine the number of messages
currently waiting in a socket's receive queue (TIPC_SOCK_RECVQ_DEPTH) or
in all TIPC socket receive queues (TIPC_NODE_RECVQ_DEPTH).

Signed-off-by: Oscar Medina <oscar.medina@motorola.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agocleanup: remove unused member in scm_cookie.
Rami Rosen [Thu, 25 Jun 2009 04:42:19 +0000 (04:42 +0000)]
cleanup: remove unused member in scm_cookie.

This patch removes an unused member (seq) scm_cookie; besides initialized
to 0 in the header file, it is not used.

Signed-off-by: Rami Rosen <ramirose@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMinor code cleanup in drivers/net/r6040.c
Michael Opdenacker [Wed, 24 Jun 2009 21:05:09 +0000 (21:05 +0000)]
Minor code cleanup in drivers/net/r6040.c

Replaced '0' by 'bar' in the probe function.

The 'bar' variable was already set to '0' and is already
used in pci_iomap(). This is cleaner and improves code
consistency.

Signed-off-by: Michael Opdenacker <michael@free-electrons.com>
Acked-by: Florian Fainelli <florian@openwrt.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agonet: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions
Patrick McHardy [Tue, 23 Jun 2009 06:03:08 +0000 (06:03 +0000)]
net: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions

This patch is the result of an automatic spatch transformation to convert
all ndo_start_xmit() return values of 0 to NETDEV_TX_OK.

Some occurences are missed by the automatic conversion, those will be
handled in a seperate patch.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agonet: remove NET_RX_BAD and NET_RX_CN* defines
Florian Westphal [Sat, 20 Jun 2009 00:53:25 +0000 (00:53 +0000)]
net: remove NET_RX_BAD and NET_RX_CN* defines

almost no users in the tree; and the few that use them treat them
like NET_RX_DROP.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMerge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
David S. Miller [Mon, 6 Jul 2009 02:06:45 +0000 (19:06 -0700)]
Merge branch 'master' of /linux/kernel/git/davem/net-2.6

14 years agodsa: fix 88e6xxx statistics counter snapshotting
Stephane Contri [Thu, 2 Jul 2009 23:26:48 +0000 (23:26 +0000)]
dsa: fix 88e6xxx statistics counter snapshotting

The bit that tells us whether a statistics counter snapshot operation
has completed is located in the GLOBAL register block, not in the
GLOBAL2 register block, so fix up mv88e6xxx_stats_wait() to poll the
right register address.

Signed-off-by: Stephane Contri <Stephane.Contri@grassvalley.com>
Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Cc: stable@kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoforcedeth: Fix NAPI race.
Eric Dumazet [Thu, 2 Jul 2009 04:04:45 +0000 (04:04 +0000)]
forcedeth: Fix NAPI race.

Eric Dumazet a Ã©crit :
> Ingo Molnar a Ã©crit :
>>> The following changes since commit 52989765629e7d182b4f146050ebba0abf2cb0b7:
>>>   Linus Torvalds (1):
>>>         Merge git://git.kernel.org/.../davem/net-2.6
>>>
>>> are available in the git repository at:
>>>
>>>   master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6.git master
>> Hm, something in this lot quickly wrecked networking here - see the
>> tx timeout dump below. It starts with:
>>
>> [  351.004596] WARNING: at net/sched/sch_generic.c:246 dev_watchdog+0x10b/0x19c()
>> [  351.011815] Hardware name: System Product Name
>> [  351.016220] NETDEV WATCHDOG: eth0 (forcedeth): transmit queue 0 timed out
>>
>> Config attached. Unfortunately i've got no time to do bisection
>> today.
>
>
>
> forcedeth might have a problem, in its netif_wake_queue() logic, but
> I could not see why a recent patch could make this problem visible now.
>
> CPU0/1: AMD Athlon(tm) 64 X2 Dual Core Processor 3800+ stepping 02
> is not a new cpu either :)
>
> forcedeth uses an internal tx_stop without appropriate barrier.
>
> Could you try following patch ?
>
> (random guess as I dont have much time right now)

We might have a race in napi_schedule(), leaving interrupts disabled forever.
I cannot test this patch, I dont have the hardware...

Tested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agodrivers/net/smsc911x.c: Fix resource size off by 1 error
Julia Lawall [Sat, 4 Jul 2009 11:31:47 +0000 (11:31 +0000)]
drivers/net/smsc911x.c: Fix resource size off by 1 error

The call resource_size(res) returns res->end - res->start + 1 and thus the
second change is semantics-preserving.  res_size is then used as the second
argument of a call to request_mem_region, and the memory allocated by this
call appears to be the same as what is released in the two calls to
release_mem_region.  So the size argument for those calls should be
resource_size(size) as well.  Alternatively, in the second call to
release_mem_region, the second argument could be res_size, as that variable
has already been initialized at the point of this call.

The problem was found using the following semantic patch:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@@
struct resource *res;
@@

- (res->end - res->start) + 1
+ resource_size(res)

@@
struct resource *res;
@@

- res->end - res->start
+ BAD(resource_size(res))
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agopcnet_cs: add new id
Ken Kawasaki [Sat, 4 Jul 2009 14:17:10 +0000 (14:17 +0000)]
pcnet_cs: add new id

add new id (RIOS System PC CARD3 ETHERNET).

Signed-off-by: Ken Kawasaki <ken_kawasaki@spring.nifty.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agobnx2x: Fix the maximal values of coalescing timeouts.
Eilon Greenstein [Sun, 5 Jul 2009 04:18:14 +0000 (04:18 +0000)]
bnx2x: Fix the maximal values of coalescing timeouts.

This patch properly defines the maximum values for rx/tx coalescing timeouts.

Signed-off-by: Vlad Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agobnx2x: Disable HC coalescing when setting timeout to zero.
Eilon Greenstein [Sun, 5 Jul 2009 04:18:12 +0000 (04:18 +0000)]
bnx2x: Disable HC coalescing when setting timeout to zero.

Problem reported by Flavio Leitner <fleitner@redhat.com>:
When setting rx/tx coalescing timeout to the values less than 12 traffic was
stopped.

The FW supports coalescing in 12us granularity, and so value of less then 12
should be interpreted as disabling coalescing

Signed-off-by: Vlad Zolotarov <vladz@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agotun: Fix device unregister race
Herbert Xu [Thu, 2 Jul 2009 23:03:55 +0000 (23:03 +0000)]
tun: Fix device unregister race

It is currently possible for an asynchronous device unregister
to cause the same tun device to be unregistered twice.  This
is because the unregister in tun_chr_close only checks whether
__tun_get(tfile) != NULL.  This however has nothing to do with
whether the device has already been unregistered.  All it tells
you is whether __tun_detach has been called.

This patch fixes this by using the most obvious thing to test
whether the device has been unregistered.

It also moves __tun_detach outside of rtnl_unlock since nothing
that it does requires that lock.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoLinux 2.6.31-rc2 v2.6.31-rc2
Linus Torvalds [Sat, 4 Jul 2009 17:58:48 +0000 (10:58 -0700)]
Linux 2.6.31-rc2

14 years agovideo: sm501fb: Early initialization of mm_lock mutex.
Paul Mundt [Fri, 3 Jul 2009 06:45:56 +0000 (15:45 +0900)]
video: sm501fb: Early initialization of mm_lock mutex.

Commit 537a1bf059fa312355696fa6db80726e655e7f17 (fbdev: add mutex for
fb_mmap locking) introduces a ->mm_lock mutex for protecting smem
assignments. Unfortunately in the case of sm501fb these happen quite
early in the initialization code, well before the mutex_init() that takes
place in register_framebuffer(), leading to:

   Badness at kernel/mutex.c:207

   Pid : 1, Comm:          swapper
   CPU : 0                 Not tainted  (2.6.31-rc1-00284-g529ba0d-dirty #2273)

   PC is at __mutex_lock_slowpath+0x72/0x1bc
   PR is at __mutex_lock_slowpath+0x66/0x1bc
   ...

matroxfb appears to have the same issue and has solved it with an early
mutex_init(), so we do the same for sm501fb.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Cc: Krzysztof Helt <krzysztof.h1@wp.pl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/kyle/parisc-2.6
Linus Torvalds [Sat, 4 Jul 2009 17:35:12 +0000 (10:35 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/kyle/parisc-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/kyle/parisc-2.6: (27 commits)
  parisc: use generic atomic64 on 32-bit
  parisc: superio: fix build breakage
  parisc: Fix PCI resource allocation on non-PAT SBA machines
  parisc: perf: wire up sys_perf_counter_open
  parisc: add task_pt_regs macro
  parisc: wire sys_perf_counter_open to sys_ni_syscall
  parisc: inventory.c, fix bloated stack frame
  parisc: processor.c, fix bloated stack frame
  parisc: fix compile warning in mm/init.c
  parisc: remove dead code from sys_parisc32.c
  parisc: wire up rt_tgsigqueueinfo
  parisc: ensure broadcast tlb purge runs single threaded
  parisc: fix "delay!" timer handling
  parisc: fix mismatched parenthesis in memcpy.c
  parisc: Fix gcc 4.4 warning in lba_pci.c
  parisc: add parameter to read_cr16()
  parisc: decode_exc.c should include kernel.h
  parisc: remove obsolete hw_interrupt_type
  parisc: fix irq compile bugs in arch/parisc/kernel/irq.c
  parisc: advertise PCI devs after "assign_resources"
  ...

Manually fixed up trivial conflicts in tools/perf/perf.h due to addition
of SH vs HPPA perf-counter support.

14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6
Linus Torvalds [Sat, 4 Jul 2009 17:31:26 +0000 (10:31 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/sameo/mfd-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6:
  mfd: fix pcap adc locking
  mfd: sm501, fix lock imbalance

14 years agoMerge branch 'for-2.6.31' of git://linux-nfs.org/~bfields/linux
Linus Torvalds [Sat, 4 Jul 2009 17:11:38 +0000 (10:11 -0700)]
Merge branch 'for-2.6.31' of git://linux-nfs.org/~bfields/linux

* 'for-2.6.31' of git://linux-nfs.org/~bfields/linux:
  NFSD: Don't hold unrefcounted creds over call to nfsd_setuser()

14 years agoMerge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Sat, 4 Jul 2009 16:47:01 +0000 (09:47 -0700)]
Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  MIPS: Fix CONFIG_FLATMEM version of pfn_valid()
  MIPS: Reorganize Cavium OCTEON PCI support.
  Update Yoichi Yuasa's e-mail address
  MIPS: Allow suspend and hibernation again on uniprocessor kernels.
  MIPS: 64-bit: Fix o32 core dump
  MIPS: BC47xx: Fix SSB irq setup
  MIPS: CMP: Update sync-r4k for current kernel
  MIPS: CMP: Move gcmp_probe to before the SMP ops
  MIPS: CMP: activate CMP support
  MIPS: CMP: Extend IPI handling to CPU number
  MIPS: CMP: Extend the GIC IPI interrupts beyond 32
  MIPS: Define __arch_swab64 for all mips r2 cpus
  MIPS: Update VR41xx GPIO driver to use gpiolib
  MIPS: Hookup new syscalls sys_rt_tgsigqueueinfo and sys_perf_counter_open.
  MIPS: Malta: Remove unnecessary function prototypes
  MIPS: MT: Remove unnecessary semicolons
  MIPS: Add support for Texas Instruments AR7 System-on-a-Chip

14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Sat, 4 Jul 2009 16:46:13 +0000 (09:46 -0700)]
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:
  sound: do not set DEVNAME for OSS devices
  ALSA: hda - Add sanity check in PCM open callback
  ALSA: hda - Call snd_pcm_lib_hw_rates() again after codec open callback
  ALSA: hda - Avoid invalid formats and rates with shared SPDIF
  ALSA: hda - Improve ASUS eeePC 1000 mixer
  ALSA: hda - Add GPIO1 control at muting with HP laptops
  ALSA: usx2y - reparent sound device
  ALSA: snd_usb_caiaq: reparent sound device
  sound: virtuoso: fix Xonar D1/DX silence after resume
  ASoC: Only disable pxa2xx-i2s clocks if we enabled them
  ALSA: hda - Add quirk for HP 6930p
  ALSA: hda - Add missing static to patch_ca0110()
  ASoC: OMAP: fix OMAP1510 broken PCM pointer callback
  ASoC: remove BROKEN from Efika and pcm030 fabric drivers
  ASoC: Fix typo in MPC5200 PSC AC97 driver Kconfig

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/sam/kbuild-fixes
Linus Torvalds [Sat, 4 Jul 2009 16:46:01 +0000 (09:46 -0700)]
Merge git://git./linux/kernel/git/sam/kbuild-fixes

* git://git.kernel.org/pub/scm/linux/kernel/git/sam/kbuild-fixes:
  kbuild: finally remove the obsolete variable $TOPDIR
  gitignore: ignore scripts/ihex2fw
  Kbuild: Disable the -Wformat-security gcc flag
  gitignore: ignore gcov output files
  kbuild: deb-pkg ship changelog
  Add new __init_task_data macro to be used in arch init_task.c files.
  asm-generic/vmlinux.lds.h: shuffle INIT_TASK* macro names in vmlinux.lds.h
  Add new macros for page-aligned data and bss sections.
  asm-generic/vmlinux.lds.h: Fix up RW_DATA_SECTION definition.

14 years agoMerge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block
Linus Torvalds [Sat, 4 Jul 2009 16:45:31 +0000 (09:45 -0700)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block

* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
  block: don't merge requests of different failfast settings
  cciss: Ignore stale commands after reboot

14 years agobe2net: fix spurious interrupt handling in intx mode
Sathya Perla [Wed, 1 Jul 2009 01:06:07 +0000 (01:06 +0000)]
be2net: fix spurious interrupt handling in intx mode

Occasionally we may see an interrupt without an event in the eq.
In intx, we currently see the event queue and return IRQ_NONE causing
a the irq to be disabled ("no one cared".) Instead, read the CEV_ISR
reg to check the existence of the interrupt.

Signed-off-by: Sathya Perla <sathyap@serverengines.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000e: disable K1 at 1000Mbps for 82577/82578
Bruce Allan [Wed, 1 Jul 2009 13:29:08 +0000 (13:29 +0000)]
e1000e: disable K1 at 1000Mbps for 82577/82578

This workaround is required for an issue in hardware where noise on the
interconnect between the MAC and PHY could be generated by a lower power
mode (K1) at 1000Mbps resulting in bad packets.  Disable K1 while at 1000
Mbps but keep it enabled for 10/100Mbps and when the cable is disconnected.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000e: delay second read of PHY_STATUS register on failure of first read
Bruce Allan [Wed, 1 Jul 2009 13:28:50 +0000 (13:28 +0000)]
e1000e: delay second read of PHY_STATUS register on failure of first read

Some PHYs may require two reads of the PHY_STATUS register to determine the
link status.  If the PHY is being accessed by another thread it is possible
the first read could timeout and fail.  In this case, put a delay in so
the second read will pick up the correct link status.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000e: prevent NVM corruption on sectors larger than 4K
Bruce Allan [Wed, 1 Jul 2009 13:28:32 +0000 (13:28 +0000)]
e1000e: prevent NVM corruption on sectors larger than 4K

Limit NVM writes to 4K sections to prevent NVM corruption on larger
sector allocations (up to 64K).

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000e: do not write SmartSpeed register bits on parts without support
Bruce Allan [Wed, 1 Jul 2009 13:28:14 +0000 (13:28 +0000)]
e1000e: do not write SmartSpeed register bits on parts without support

The driver was accessing register bits for features on parts that do
not support that feature.  This could cause problems in the hardware.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000e: delay after LCD reset and proper checks for PHY configuration done
Bruce Allan [Wed, 1 Jul 2009 13:27:55 +0000 (13:27 +0000)]
e1000e: delay after LCD reset and proper checks for PHY configuration done

A previous workaround for 82578 to avoid link stall causes some PHY
registers to get cleared inadvertently.  Add a delay after all LCD resets
to make sure PHY registers are in a stable state before continuing.  Also,
after resets check the EEC register for the state of PHY configuration
performed by the MAC for ICH9 and earlier parts (as done before), but check
the LAN_INIT_DONE bit in the STATUS register for ICH10 and newer parts (EEC
doesn't exist in these newer parts).

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoe1000e: PHY loopback broken on 82578
Bruce Allan [Wed, 1 Jul 2009 13:27:31 +0000 (13:27 +0000)]
e1000e: PHY loopback broken on 82578

PHY loopback on 82578 fails to work as a result of flushing the packets
in the FIFO buffer in the link stall workaround.  Don't perform the
workaround if in PHY loopback mode.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoixgbe: Not allow 8259x unsupported wol options change from ethtool
Mallikarjuna R Chilakala [Thu, 2 Jul 2009 12:51:10 +0000 (12:51 +0000)]
ixgbe: Not allow 8259x unsupported wol options change from ethtool

Wake-on-lan is currently only supported by 82599 KX4 devices, in all
other cases return a proper value from ixgbe_wol_exclusion function call.
Otherwise from ethtool we will be able to change wol options of
unsupported 8259x devices.

Signed-off-by: Mallikarjuna R Chilakala <mallikarjuna.chilakala@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoixgbe: fix inconsistent SFP/SFP+ failure results.
Don Skidmore [Thu, 2 Jul 2009 12:50:52 +0000 (12:50 +0000)]
ixgbe: fix inconsistent SFP/SFP+ failure results.

Currently if we loaded the driver, insert an unsupported module, and then
attempt to "ifconfig up" the device it will be brought down but the netdev
would not be unregistered. This behavior is different than all other
code paths. This patch corrects that by down'ing the device and then
scheduling the sfp_config_module_task tasklet. The tasklet will detect
this condition (like it does with other code paths) and do the
unregister_netdev().

I also removed the log message as this condition (an unsupported SFP+
module) will be logged in sfp_config_module_task.

Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoixgbe: fix regression on some 82598 adapters
Don Skidmore [Thu, 2 Jul 2009 12:50:31 +0000 (12:50 +0000)]
ixgbe: fix regression on some 82598 adapters

The change to check the SFP+ module again on open() was
causing the XFP (non-SFP+) adapters to be rejected. We
only want to try and re-identify the SFP+ module if the
original probe found that this device was an SFP+ device.
So for this code path (driver loaded with SFP module, module
inserted, ifconfig up of the device) the type will be
ixgbe_phy_unknown for an unidentified SFP+ module.  So we
only check if that is the case.

This problem also shows up on Copper devices.

Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoixgbe: fix issues with failing to detect insert of unsupported module
Don Skidmore [Thu, 2 Jul 2009 12:50:12 +0000 (12:50 +0000)]
ixgbe: fix issues with failing to detect insert of unsupported module

Several small fixes around negative test case of the insertion of a
IXGBE_ERR_NOT_SUPPORTED module.

- mdio45_probe call was always failing due to mdio.prtad not being
set.  The function set to mdio.mdio_read was still working as we just
happen to always be at prtad == 0.  This will allow us to set the phy_id
and phy.type correctly now.

- There was timing issue with i2c calls when initiated from a tasklet.
A small delay was added to allow the electrical oscillation to calm down.

- Logic change in ixgbe_sfp_task that allows NOT_SUPPORTED condition
to be recognized.

Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Fix sizeof usage.
Ron Mercer [Thu, 2 Jul 2009 06:06:13 +0000 (06:06 +0000)]
qlge: Fix sizeof usage.

Some usage was only sizing a pointer rather than the data type.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Add/use function for link up/down.
Ron Mercer [Thu, 2 Jul 2009 06:06:12 +0000 (06:06 +0000)]
qlge: Add/use function for link up/down.

We need to set/clear the mac address register when the link goes up/down
respectively.  Without this both ports of a 2-port device can end up
with the same mac address in a bonding scenario.
The new ql_link_on() and ql_link_off() will also be used in handling
certain firmware events.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Fix MAC address bonding issue.
Ron Mercer [Thu, 2 Jul 2009 06:06:11 +0000 (06:06 +0000)]
qlge: Fix MAC address bonding issue.

This addes functionality to set/clear the MAC address in the hardware
when the link goes up/down.
The MAC address register is persistent across function resets. In
bonding the same address can bounce from one port to the other.  This
can cause packets to be delivered to the wrong port.
This patch clears the MAC address in the hardware when the link is down
and sets it when the link comes up.
It was found that pulling/pushing the cable from one port to another
causes the same MAC address to be in both ports.
The next patch in this series will use this functionality as well.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Fix tx byte counter.
Ron Mercer [Thu, 2 Jul 2009 06:06:10 +0000 (06:06 +0000)]
qlge: Fix tx byte counter.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Fix redundant call to free resources.
Ron Mercer [Thu, 2 Jul 2009 06:06:09 +0000 (06:06 +0000)]
qlge: Fix redundant call to free resources.

The caller will free acquired resouces if a failure occurs.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Fix carrier on condition.
Ron Mercer [Thu, 2 Jul 2009 06:06:08 +0000 (06:06 +0000)]
qlge: Fix carrier on condition.

We were turning on the carrier without verifying the link was up.
This adds link up to the link initialize check before turning carrier
on.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Clear frame to queue routing before reset.
Ron Mercer [Thu, 2 Jul 2009 06:06:07 +0000 (06:06 +0000)]
qlge: Clear frame to queue routing before reset.

Not clearing the routing bits can cause frames to erroneously get routed to
management processor.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoqlge: Expand coverage of hw lock for config register.
Ron Mercer [Thu, 2 Jul 2009 06:06:06 +0000 (06:06 +0000)]
qlge: Expand coverage of hw lock for config register.

The hardware semaphore covers the configuration register as well as the
ICB registers.  The ICB high and low regs contain the address of the
initialization control block and the config register is used to signal
the hardware that a block is ready to be downloaded.  Currently we were
only protecting the ICB regs.  This changes expands to cover the config
register as well.

Signed-off-by: Ron Mercer <ron.mercer@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoIPv6: preferred lifetime of address not getting updated
Brian Haley [Thu, 2 Jul 2009 07:10:52 +0000 (07:10 +0000)]
IPv6: preferred lifetime of address not getting updated

There's a bug in addrconf_prefix_rcv() where it won't update the
preferred lifetime of an IPv6 address if the current valid lifetime
of the address is less than 2 hours (the minimum value in the RA).

For example, If I send a router advertisement with a prefix that
has valid lifetime = preferred lifetime = 2 hours we'll build
this address:

3: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 2001:1890:1109:a20:217:8ff:fe7d:4718/64 scope global dynamic
       valid_lft 7175sec preferred_lft 7175sec

If I then send the same prefix with valid lifetime = preferred
lifetime = 0 it will be ignored since the minimum valid lifetime
is 2 hours:

3: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 2001:1890:1109:a20:217:8ff:fe7d:4718/64 scope global dynamic
       valid_lft 7161sec preferred_lft 7161sec

But according to RFC 4862 we should always reset the preferred lifetime
even if the valid lifetime is invalid, which would cause the address
to immediately get deprecated.  So with this patch we'd see this:

5: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 2001:1890:1109:a20:21f:29ff:fe5a:ef04/64 scope global deprecated dynamic
       valid_lft 7163sec preferred_lft 0sec

The comment winds-up being 5x the size of the code to fix the problem.

Update the preferred lifetime of IPv6 addresses derived from a prefix
info option in a router advertisement even if the valid lifetime in
the option is invalid, as specified in RFC 4862 Section 5.5.3e.  Fixes
an issue where an address will not immediately become deprecated.
Reported by Jens Rosenboom.

Signed-off-by: Brian Haley <brian.haley@hp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoxfrm6: fix the proto and ports decode of sctp protocol
Wei Yongjun [Thu, 2 Jul 2009 16:59:49 +0000 (16:59 +0000)]
xfrm6: fix the proto and ports decode of sctp protocol

The SCTP pushed the skb above the sctp chunk header, so the
check of pskb_may_pull(skb, nh + offset + 1 - skb->data) in
_decode_session6() will never return 0 and the ports decode
of sctp will always fail. (nh + offset + 1 - skb->data < 0)

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoxfrm4: fix the ports decode of sctp protocol
Wei Yongjun [Thu, 2 Jul 2009 16:57:23 +0000 (16:57 +0000)]
xfrm4: fix the ports decode of sctp protocol

The SCTP pushed the skb data above the sctp chunk header, so the check
of pskb_may_pull(skb, xprth + 4 - skb->data) in _decode_session4() will
never return 0 because xprth + 4 - skb->data < 0, the ports decode of
sctp will always fail.

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMerge branch 'fix/soundcore' into for-linus
Takashi Iwai [Fri, 3 Jul 2009 21:50:47 +0000 (23:50 +0200)]
Merge branch 'fix/soundcore' into for-linus

* fix/soundcore:
  sound: do not set DEVNAME for OSS devices

14 years agoMerge branch 'fix/hda' into for-linus
Takashi Iwai [Fri, 3 Jul 2009 21:50:45 +0000 (23:50 +0200)]
Merge branch 'fix/hda' into for-linus

* fix/hda:
  ALSA: hda - Add sanity check in PCM open callback
  ALSA: hda - Call snd_pcm_lib_hw_rates() again after codec open callback
  ALSA: hda - Avoid invalid formats and rates with shared SPDIF
  ALSA: hda - Improve ASUS eeePC 1000 mixer
  ALSA: hda - Add GPIO1 control at muting with HP laptops

14 years agosound: do not set DEVNAME for OSS devices
Kay Sievers [Fri, 3 Jul 2009 18:56:05 +0000 (20:56 +0200)]
sound: do not set DEVNAME for OSS devices

Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: hda - Add sanity check in PCM open callback
Takashi Iwai [Fri, 3 Jul 2009 21:10:23 +0000 (23:10 +0200)]
ALSA: hda - Add sanity check in PCM open callback

Add some sanity checks of struct snd_pcm_hardware fields in the PCM
open callback of hda driver.  This makes a bit easier to debug any PCM
setup errors in the codec side.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: hda - Call snd_pcm_lib_hw_rates() again after codec open callback
Takashi Iwai [Fri, 3 Jul 2009 21:06:45 +0000 (23:06 +0200)]
ALSA: hda - Call snd_pcm_lib_hw_rates() again after codec open callback

The PCM rates bit field may have been changed by the codec open callback.
In that case, we need to reset rate_min and rate_max.  So, simply call
snd_pcm_lib_hw_rates() again after the codec open callback.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoALSA: hda - Avoid invalid formats and rates with shared SPDIF
Takashi Iwai [Fri, 3 Jul 2009 21:03:30 +0000 (23:03 +0200)]
ALSA: hda - Avoid invalid formats and rates with shared SPDIF

Check whether formats and rates don't result in zero due to the
restriction of SPDIF sharing.  If any of them can be zero, disable
the SPDIF sharing mode instead.  Otherwise it will lead to a PCM
configuration error.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoblock: don't merge requests of different failfast settings
Tejun Heo [Fri, 3 Jul 2009 10:56:18 +0000 (12:56 +0200)]
block: don't merge requests of different failfast settings

Block layer used to merge requests and bios with different failfast
settings.  This caused regular IOs to fail prematurely when they were
merged into failfast requests for readahead.

Niel Lambrechts could trigger the problem semi-reliably on ext4 when
resuming from STR.  ext4 uses readahead when reading inodes and
combined with the deterministic extra SATA PHY exception cycle during
resume on the specific configuration, non-readahead inode read would
fail causing ext4 errors.  Please read the following thread for
details.

  http://lkml.org/lkml/2009/5/23/21

This patch makes block layer reject merging if the failfast settings
don't match.  This is correct but likely to lower IO performance by
preventing regular IOs from mingling into surrounding readahead
requests.  Changes to allow such mixed merges and handle errors
correctly will be added later.

Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Niel Lambrechts <niel.lambrechts@gmail.com>
Cc: Theodore Tso <tytso@mit.edu>
Signed-off-by: Jens Axboe <axboe@carl.(none)>
14 years agocciss: Ignore stale commands after reboot
Hannes Reinecke [Thu, 2 Jul 2009 20:02:06 +0000 (22:02 +0200)]
cciss: Ignore stale commands after reboot

When doing an unexpected shutdown like kexec the cciss
firmware might still have some commands in flight, which
it is trying to complete.
The driver is doing it's best on resetting the HBA,
but sadly there's a firmware issue causing the firmware
_not_ to abort or drop old commands.
So the firmware will send us commands which we haven't
accounted for, causing the driver to panic.

With this patch we're just ignoring these commands as
there is nothing we could be doing with them anyway.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Acked-by: Mike Miller <mike.miller@hp.com>
Signed-off-by: Jens Axboe <axboe@carl.(none)>
14 years agoMIPS: Fix CONFIG_FLATMEM version of pfn_valid()
Ralf Baechle [Fri, 3 Jul 2009 06:11:15 +0000 (07:11 +0100)]
MIPS: Fix CONFIG_FLATMEM version of pfn_valid()

For systems which do not define PHYS_OFFSET as 0 pfn_valid() may falsely
have returned 0 on most configurations.  Bug introduced by commit
752fbeb2e3555c0d236e992f1195fd7ce30e728d (linux-mips.org) rsp.
6f284a2ce7b8bc49cb8455b1763357897a899abb (kernel.org) titled "[MIPS]
FLATMEM: introduce PHYS_OFFSET."

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: Reorganize Cavium OCTEON PCI support.
David Daney [Tue, 30 Jun 2009 00:18:51 +0000 (17:18 -0700)]
MIPS: Reorganize Cavium OCTEON PCI support.

Move the cavium PCI files to the arch/mips/pci directory.  Also cleanup
comment formatting and code layout.  Code from pci-common.c, was moved
into other files.

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoUpdate Yoichi Yuasa's e-mail address
Yoichi Yuasa [Thu, 2 Jul 2009 15:39:38 +0000 (00:39 +0900)]
Update Yoichi Yuasa's e-mail address

Signed-off-by: Yoichi Yuasa <yuasa@linux-mips.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: Allow suspend and hibernation again on uniprocessor kernels.
Ralf Baechle [Thu, 2 Jul 2009 10:48:07 +0000 (11:48 +0100)]
MIPS: Allow suspend and hibernation again on uniprocessor kernels.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Acked-by: Pavel Machek <pavel@ucw.cz>
14 years agoMIPS: 64-bit: Fix o32 core dump
Yong Zhang [Wed, 1 Jul 2009 01:35:39 +0000 (09:35 +0800)]
MIPS: 64-bit: Fix o32 core dump

If an o32 process generates a core dump on a 64 bit kernel, the core file
will not be correctly recognized. This is because ELF_CORE_COPY_REGS and
ELF_CORE_COPY_TASK_REGS are not correctly defined for o32 and will use
the default register set which would be CONFIG_64BIT in asm/elf.h.

So we'll switch to use the right register defines in this situation by
checking for WANT_COMPAT_REG_H and use the right defines of
ELF_CORE_COPY_REGS and ELF_CORE_COPY_TASK_REGS.

[Ralf: made ELF_CORE_COPY_TASK_REGS() bullet-proof against funny arguments.]

Signed-off-by: Yong Zhang <yong.zhang@windriver.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: BC47xx: Fix SSB irq setup
matthieu castet [Tue, 30 Jun 2009 21:04:55 +0000 (23:04 +0200)]
MIPS: BC47xx: Fix SSB irq setup

The current ssb irq setup in ssb_mipscore_init has the problem that it
configures some device on some irq without checking that the irq is not
taken by an other device.

For example in my case PCI host is on irq 0 and IPSEC on irq 3.
The current code:
  - store in dev->irq that IPSEC irq is 3 + 2
  - do a set_irq 0->3 on PCI host

But now IPSEC irq is not routed anymore to the mips code and dev->irq is
wrong.  This causes a problem described in [1].

This patch tries to solve the problem by making set_irq configure the
device we want to take the irq on the shared irq0. The previous example
becomes:
  - store in dev->irq that IPSEC irq is 3 + 2
  - do a set_irq 0->3 on PCI host:
  - irq 3 is already taken by IPSEC. do a set_irq 3->0 on IPSEC

I also added some code to print the irq configuration after irq setup to
allow easier debugging. And I add extra checking in ssb_mips_irq to report
device without irq or device with not routed irq.

[1] http://www.danm.de/files/src/bcm5365p/REPORTED_DEVICES

Signed-off-by: Matthieu CASTET <castet.matthieu@free.fr>
Acked-by : Michael Buesch <mb@bu3sch.de>
Tested-by: Florian Fainelli <florian@openwrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: CMP: Update sync-r4k for current kernel
Tim Anderson [Wed, 17 Jun 2009 23:40:34 +0000 (16:40 -0700)]
MIPS: CMP: Update sync-r4k for current kernel

This revises the sync-4k so it will boot and operate since the removal of
expirelo from the timer code.

Signed-off-by: Tim Anderson <tanderson@mvista.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: CMP: Move gcmp_probe to before the SMP ops
Tim Anderson [Wed, 17 Jun 2009 23:25:18 +0000 (16:25 -0700)]
MIPS: CMP: Move gcmp_probe to before the SMP ops

This is to move the gcmp_probe call to before the use of and selection of
the smp_ops functions. This allows malta with 1004K to work.

Signed-off-by: Tim Anderson <tanderson@mvista.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: CMP: activate CMP support
Tim Anderson [Wed, 17 Jun 2009 23:22:53 +0000 (16:22 -0700)]
MIPS: CMP: activate CMP support

Most of the CMP support was added before, this mostly correct compile
problems but adds a platform specific translation for the interrupt number
based on cpu number.

Signed-off-by: Tim Anderson <tanderson@mvista.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: CMP: Extend IPI handling to CPU number
Tim Anderson [Wed, 17 Jun 2009 23:22:25 +0000 (16:22 -0700)]
MIPS: CMP: Extend IPI handling to CPU number

This takes the current IPI interrupt assignment from the fix number of 4
to the number of CPUs defined in the system.

Signed-off-by: Tim Anderson <tanderson@mvista.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: CMP: Extend the GIC IPI interrupts beyond 32
Tim Anderson [Wed, 17 Jun 2009 23:21:19 +0000 (16:21 -0700)]
MIPS: CMP: Extend the GIC IPI interrupts beyond 32

This patch extends the GIC interrupt handling beyond the current 32 bit
range as well as extending the number of interrupts based on the number
of CPUs.

Signed-off-by: Tim Anderson <tanderson@mvista.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: Define __arch_swab64 for all mips r2 cpus
David Daney [Mon, 29 Jun 2009 16:54:15 +0000 (09:54 -0700)]
MIPS: Define __arch_swab64 for all mips r2 cpus

Some CPUs implement mipsr2, but because they are a super-set of mips64r2 do
not define CONFIG_CPU_MIPS64_R2.  Cavium OCTEON falls into this category.
We would still like to use the optimized implementation, so since we have
already checked for CONFIG_CPU_MIPSR2, checking for CONFIG_64BIT instead of
CONFIG_CPU_MIPS64_R2 is sufficient.

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: Update VR41xx GPIO driver to use gpiolib
Yoichi Yuasa [Mon, 29 Jun 2009 02:11:05 +0000 (11:11 +0900)]
MIPS: Update VR41xx GPIO driver to use gpiolib

Signed-off-by: Yoichi Yuasa <yyuasa@linux.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
14 years agoMIPS: Hookup new syscalls sys_rt_tgsigqueueinfo and sys_perf_counter_open.
David Daney [Fri, 26 Jun 2009 16:53:57 +0000 (09:53 -0700)]
MIPS: Hookup new syscalls sys_rt_tgsigqueueinfo and sys_perf_counter_open.

[Ralf: I fixed up the numbering in the comment in scall64-n32.S.]

Signed-off-by: David Daney <ddaney@caviumnetworks.com>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>