Commit graph

199,777 commits

Author SHA1 Message Date
Jan Kiszka
7f3c1fa4c3 uml: Fix build breakage after slab.h changes
We now have to to include linux/slab.h explicitly for kmalloc &
friends.  Files that build against host headers already get their
prototypes via um_malloc.h, linux/slab.h may even be unavailable.

Signed-off-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Tejun Heo <tj@kernel.org>
2010-04-19 17:46:23 +09:00
David S. Miller
6c94b1ee0c sparc64: Fix PREEMPT_ACTIVE value.
It currently overlaps the NMI bit.

Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-19 01:30:51 -07:00
Paul Mundt
3cf6fa1e33 sh: Enable SH-X3 hardware synonym avoidance handling.
This enables support for the hardware synonym avoidance handling on SH-X3
CPUs for the case where dcache aliases are possible. icache handling is
retained, but we flip on broadcasting of the block invalidations due to
the lack of coherency otherwise on SMP.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2010-04-19 17:27:17 +09:00
Dmitry Torokhov
6b9d363c49 Input: psmouse - ignore parity error for basic protocols
Observing behavior of the other OS it appears that parity errors reported
by the keyboard controller are being ignored and the data is processed
as usual. Let's do the same for standard PS/2 protocols (bare, Intellimouse
and Intellimouse Explorer) to provide better compatibility. Thsi should fix
teh following bug:

	https://bugzilla.kernel.org/show_bug.cgi?id=6105

Thanks for Damjan Jovanovic for locating the source of issue and ideas
for the patch.

Tested-by: Damjan Jovanovic <damjan.jov@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
2010-04-19 00:50:42 -07:00
Paul Mundt
d9c944463d sh: mach-sdk7786: pm_power_off support.
This wires up power-off support for the SDK7786 board.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2010-04-19 16:27:47 +09:00
Paul E. McKenney
bc293d62b2 rcu: Make RCU lockdep check the lockdep_recursion variable
The lockdep facility temporarily disables lockdep checking by
incrementing the current->lockdep_recursion variable.  Such
disabling happens in NMIs and in other situations where lockdep
might expect to recurse on itself.

This patch therefore checks current->lockdep_recursion, disabling RCU
lockdep splats when this variable is non-zero.  In addition, this patch
removes the "likely()", as suggested by Lai Jiangshan.

Reported-by: Frederic Weisbecker <fweisbec@gmail.com>
Reported-by: David Miller <davem@davemloft.net>
Tested-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: laijs@cn.fujitsu.com
Cc: dipankar@in.ibm.com
Cc: mathieu.desnoyers@polymtl.ca
Cc: josh@joshtriplett.org
Cc: dvhltc@us.ibm.com
Cc: niv@us.ibm.com
Cc: peterz@infradead.org
Cc: rostedt@goodmis.org
Cc: Valdis.Kletnieks@vt.edu
Cc: dhowells@redhat.com
Cc: eric.dumazet@gmail.com
LKML-Reference: <20100415195039.GA22623@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-04-19 08:37:19 +02:00
Marek Olšák
cae94b0ad9 drm/radeon/kms: allow R500 regs VAP_ALT_NUM_VERTICES and VAP_INDEX_OFFSET
[airlied: fix V_A_N_V to not be safe and fix check to make sure only r500
 - bump userspace version]

Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 14:17:02 +10:00
Corbin Simpson
f12eebb0ac drivers/gpu/radeon: Add MSPOS regs to safe list.
Permits MSAA and D3D-style rasterization.

[airlied: add rs600]

Signed-off-by: Corbin Simpson <MostAwesomeDude@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 14:04:32 +10:00
Alex Deucher
d3a67a43b0 drm/radeon/kms: disable the tv encoder when tv/cv is not in use
Switching between TV and VGA caused VGA to break on some systems
since the TV encoder was left enabled when VGA was used.

fixes fdo bug 25520.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 13:52:52 +10:00
Alex Deucher
a1a4b23b66 drm/radeon/kms: adjust pll settings for tv
May fix fdo bug 26582.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 13:52:43 +10:00
Alex Deucher
08d075116d drm/radeon/kms: fix tv dac conflict resolver
On systems with the tv dac shared between DVI and TV,
we can only use the dac for one of the connectors.
However, when using a digital monitor on the DVI port,
you can use the dac for the TV connector just fine.
Check the use_digital status when resolving the conflict.

Fixes fdo bug 27649, possibly others.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 11:59:58 +10:00
Alex Deucher
16823d16f5 drm/radeon/kms/evergreen: don't enable hdmi audio stuff
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 11:59:41 +10:00
Alex Deucher
b317a9ce22 drm/radeon/kms/atom: fix dual-link DVI on DCE3.2/4.0
Got broken during the evergreen merge.
Fixes fdo bug 27001.

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 11:59:20 +10:00
Jerome Glisse
30f69f3fb2 drm/radeon/kms: fix rs600 tlb flush
Typo in in flush leaded to no flush of the RS600 tlb which
ultimately leaded to massive system ram corruption, with
this patch everythings seems to work properly.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 11:26:13 +10:00
Jerome Glisse
1b5331d9c6 drm/radeon/kms: print GPU family and device id when loading
This will help figuring out GPU when looking at bugs log.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
2010-04-19 11:25:24 +10:00
Daniel Vetter
c36a2a6de5 drm/i915: fix tiling limits for i915 class hw v2
Current code is definitely crap: Largest pitch allowed spills into
the TILING_Y bit of the fence registers ... :(

I've rewritten the limits check under the assumption that 3rd gen hw
has a 3d pitch limit of 8kb (like 2nd gen). This is supported by an
otherwise totally misleading XXX comment.

This bug mostly resulted in tiling-corrupted pixmaps because the kernel
allowed too wide buffers to be tiled. Bug brought to the light by the
xf86-video-intel 2.11 release because that unconditionally enabled
tiling for pixmaps, relying on the kernel to check things. Tiling for
the framebuffer was not affected because the ddx does some additional
checks there ensure the buffer is within hw-limits.

v2: Instead of computing the value that would be written into the
hw fence registers and then checking the limits simply check whether
the stride is above the 8kb limit. To better document the hw, add
some WARN_ONs in i915_write_fence_reg like I've done for the i830
case (using the right limits).

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=27449
Tested-by: Alexander Lam <lambchop468@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:58:24 -07:00
Daniel Vetter
e158316567 agp/intel-gtt: kill previous_size assignments
Not needed for the GTT and inconsistent: Sometimes the _new_ size
was stored there ...

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:40:10 -07:00
Daniel Vetter
1ca46bd134 agp/intel-gtt: kill intel_i830_tlbflush
We don't use the generic insert/remove_memory functions that
require this. So kill this useless code.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:39:39 -07:00
Daniel Vetter
22dd82a3f5 agp/intel: split out gmch/gtt probe, part 1
This is essentially the last piece of code that's tying intel-gtt.c
to intel-agp.c. Extract the probe code into it's own function so that
it can be moved to intel-gtt.c.

This requires some slight changes in the ordering of device probe function.
This patch just implements that for better bisectability in case this
introduces bugs.

The biggest change is that the gmch/gtt code doesn't execute a pci
resource fixup anymore. I've dug around in historical git trees, and
this change is to support the agp port on an old HP server with the
i440 intel chipset. So only needed for the agp driver.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:37:46 -07:00
Daniel Vetter
059efc670d agp/intel: kill mutli_gmch_chip
Always zero, i.e. unused.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:36:26 -07:00
Daniel Vetter
e5a04d52e6 agp/intel: uncoditionally reconfigure driver on resume
Only two drivers were not in this table (7505 and g33), both non-mobile
chipsets. So they were most likely just missing. This is another step
to untangle the gtt from the agp driver.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:36:11 -07:00
Daniel Vetter
f51b766211 agp/intel: split out the GTT support
intel-agp.c contains actually two different drivers: An agp driver
for _physical_ agp slots an the gtt driver that is used by the intel
drm modules. Split them to prevent any further confusion.

This patch just moves the code and includes intel-gtt.c in intel-agp.c
Later patches will untangle these two drivers further.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:35:47 -07:00
Daniel Vetter
ff7cdd691a agp/intel: introduce intel-agp.h header file
Intel definitions have spilled into agp.h. Create a header file for
them and also include it in efficion-agp.c 'cause it needs a few of
them.

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:33:02 -07:00
Karsten Wiese
6e0032f0ae drm/i915: Don't touch PORT_HOTPLUG_EN in intel_dp_detect()
PORT_HOTPLUG_EN has allready been setup in i915_driver_irq_postinstall(),
when intel_dp_detect() runs.

Delete the DP[BCD]_HOTPLUG_INT_EN defines, they are not referenced anymore.

I found this while searching for a fix for
        https://bugzilla.redhat.com/show_bug.cgi?id=528312

Signed-off-by: Karsten Wiese <fzu@wemgehoertderstaat.de>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:11:44 -07:00
Zhang Rui
3143751ff5 drm/i915: set DIDL using the ACPI video output device _ADR method return.
we used to set the DIDL in the output device detected order.
But some BIOSes requires it to be initialized in the ACPI device order.
e.g. the value of the first field in DIDL stands for the first
ACPI video output device in ACPI namespace.

Now we initialize the DIDL using the device id, i.e. _ADR return value,
of each ACPI video device, if it is not 0.
https://bugzilla.kernel.org/show_bug.cgi?id=15054

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 17:05:13 -07:00
Adam Jackson
77ffb5979d drm/i915/pch: Use minimal number of FDI lanes (v2)
This should be a small power savings. Tested on Lenovo T410 (Ironlake), LVDS
VGA and DisplayPort, up to 1920x1200R.

v2: Add Sandybridge support, fix obvious math error.

Acked-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 16:57:44 -07:00
Adam Jackson
5ce8ba7c92 drm/i915: Fix 82854 PCI ID, and treat it like other 85X
pci.ids and the datasheet both say it's 358e, not 35e8.

Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 16:56:46 -07:00
Adam Jackson
8f4695ed1c drm/i915: Attempt to fix watermark setup on 85x (v2)
IS_MOBILE() catches 85x, so we'd always try to use the 9xx FIFO sizing;
since there's an explicit 85x version, this seems wrong.

v2: Handle 830m correctly too.

Signed-off-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-18 16:52:56 -07:00
Eric Dumazet
fc6055a5ba net: Introduce skb_orphan_try()
Transmitted skb might be attached to a socket and a destructor, for
memory accounting purposes.

Traditionally, this destructor is called at tx completion time, when skb
is freed.

When tx completion is performed by another cpu than the sender, this
forces some cache lines to change ownership. XPS was an attempt to give
tx completion to initial cpu.

David idea is to call destructor right before giving skb to device (call
to ndo_start_xmit()). Because device queues are usually small, orphaning
skb before tx completion is not a big deal. Some drivers already do
this, we could do it in upper level.

There is one known exception to this early orphaning, called tx
timestamping. It needs to keep a reference to socket until device can
give a hardware or software timestamp.

This patch adds a skb_orphan_try() helper, to centralize all exceptions
to early orphaning in one spot, and use it in dev_hard_start_xmit().

"tbench 16" results on a Nehalem machine (2 X5570  @ 2.93GHz)
before: Throughput 4428.9 MB/sec 16 procs
after: Throughput 4448.14 MB/sec 16 procs

UDP should get even better results, its destructor being more complex,
since SOCK_USE_WRITE_QUEUE is not set (four atomic ops instead of one)

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-18 02:39:41 -07:00
Eric Dumazet
9958da0501 net: remove time limit in process_backlog()
- There is no point to enforce a time limit in process_backlog(), since
other napi instances dont follow same rule. We can exit after only one
packet processed...
The normal quota of 64 packets per napi instance should be the norm, and
net_rx_action() already has its own time limit.
Note : /proc/net/core/dev_weight can be used to tune this 64 default
value.

- Use DEFINE_PER_CPU_ALIGNED for softnet_data definition.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-18 02:36:13 -07:00
Tilman Schmidt
b91ecb0027 gigaset: include cleanup cleanup
Commit 5a0e3ad causes slab.h to be included twice in many of the
Gigaset driver's source files, first via the common include file
gigaset.h and then a second time directly. Drop the spares, and
use the opportunity to clean up a few more similar cases.

Impact: cleanup, no functional change
Signed-off-by: Tilman Schmidt <tilman@imap.cc>
CC: Tejun Heo <tj@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-18 02:33:29 -07:00
Linus Torvalds
13bd8e4673 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/anholt/drm-intel
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/anholt/drm-intel:
  drm/i915: Ignore LVDS EDID when it is unavailabe or invalid
  drm/i915: Add no_lvds entry for the Clientron U800
  drm/i915: Rename many remaining uses of "output" to encoder or connector.
  drm/i915: Rename intel_output to intel_encoder.
  agp/intel: intel_845_driver is an agp driver!
  drm/i915: introduce to_intel_bo helper
  drm/i915: Disable FBC on 915GM and 945GM.
2010-04-17 14:28:50 -07:00
Linus Torvalds
d6f533c8c7 Merge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6
* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6:
  ACPI: EC: Limit burst to 64 bits
2010-04-17 10:58:38 -07:00
Linus Torvalds
65832940eb Merge branch 'for-linus' of git://oss.sgi.com/xfs/xfs
* 'for-linus' of git://oss.sgi.com/xfs/xfs:
  xfs: don't warn on EAGAIN in inode reclaim
  xfs: ensure that sync updates the log tail correctly
2010-04-17 10:57:56 -07:00
Joern Engel
b8639077ab [LogFS] Set s_bdi
Since 32a88aa1 sync() was turned into a NOP for logfs.  Worse, sync()
would not return an error, giving the illusion that writeout had
actually happened.

Afaics jffs2 was broken as well.

Signed-off-by: Joern Engel <joern@logfs.org>
2010-04-17 19:54:27 +02:00
Julia Lawall
42d284b986 drivers/pcmcia: Add missing local_irq_restore
Use local_irq_restore in this error-handling case just like in the one just
below.

A simplified version of the semantic patch that finds this problem is as
follows: (http://coccinelle.lip6.fr/)

// <smpl>
@r exists@
expression E1;
identifier f;
@@

f (...) { <+...
* local_irq_save (E1,...);
... when != E1
* return ...;
...+> }
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2010-04-17 17:54:38 +02:00
Timur Maximov
6f4567c8cf serial_cs: MD55x support (PCMCIA GPRS/EDGE modem) (kernel 2.6.33)
Many PCMCIA GPRS modems like: Onda Edge N100E, Novaway PC98 (OEM SPC98Z),
Rovermate Edgus Adaptmate-039 and others have same construction and
identification:

lspcmcia -vvv
Product Name:   Generic Modem: MD55x 1.00 Serial number: xxxxx-xxx
Identification: manf_id: 0x015d card_id: 0x4c45
                function: 2 (serial)
                prod_id(1): "Generic" (0xc49e4731)
                prod_id(2): "Modem: MD55x" (0x8913b110)
                prod_id(3): "1.00" (0x83dbf271)
                prod_id(4): "Serial number: xxxxx-xxx" (0x73ee9514)

Serial connection to GSM module based on Elan VPU16551 PCMCIA UART with
datasheet recommeded 14.7456MHz crystal oscillator.

By default serial_cs set UART clock == 1843200 Hz
For correct work need set clock 14745600 Hz.
This quirk already present in driver, only need add device in quirk list.

Signed-off-by: Timur Maximov <xcom.org@gmail.com>
Acked-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2010-04-17 17:53:32 +02:00
Dominik Brodowski
a8408c17d0 pcmcia: avoid late calls to pccard_validate_cis
pccard_validate_cis() nowadays destroys the CIS cache. Therefore,
calling it after card setup should be avoided. We can't control
the deprecated PCMCIA ioctl (which is only used on ARM nowadays),
but we can avoid -- and report -- any other calls.

Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2010-04-17 17:37:33 +02:00
Randy Dunlap
a289cc7c70 x86, UV: uv_irq.c: Fix all sparse warnings
Fix all sparse warnings in building uv_irq.c.

 arch/x86/kernel/uv_irq.c:46:17: warning: symbol 'uv_irq_chip' was not declared. Should it be static?
 arch/x86/kernel/uv_irq.c:143:50: error: no identifier for function argument
 arch/x86/kernel/uv_irq.c:162:13: error: typename in expression
 arch/x86/kernel/uv_irq.c:162:13: error: undefined identifier 'restrict'
 arch/x86/kernel/uv_irq.c:250:44: error: no identifier for function argument
 arch/x86/kernel/uv_irq.c:260:17: error: typename in expression
 arch/x86/kernel/uv_irq.c:260:17: error: undefined identifier 'restrict'
 arch/x86/kernel/uv_irq.c:233:50: warning: incorrect type in argument 3 (different signedness)
 arch/x86/kernel/uv_irq.c:233:50:    expected int *pnode
 arch/x86/kernel/uv_irq.c:233:50:    got unsigned int *<noident>
 arch/x86/include/asm/uv/uv_hub.h:318:44: warning: incorrect type in argument 2 (different address spaces)
 arch/x86/include/asm/uv/uv_hub.h:318:44:    expected void volatile [noderef] <asn:2>*addr
 arch/x86/include/asm/uv/uv_hub.h:318:44:    got unsigned long *

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: Russ Anderson <rja@sgi.com>
Cc: Robin Holt <holt@sgi.com>
Cc: Mike Travis <travis@sgi.com>
Cc: Cliff Wickman <cpw@sgi.com>
Cc: Jack Steiner <steiner@sgi.com>
LKML-Reference: <20100416175142.f4b59683.randy.dunlap@oracle.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-04-17 10:37:20 +02:00
Eric Dumazet
8770acf049 rps: rps_sock_flow_table is mostly read
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-17 00:54:36 -07:00
J. Bruce Fields
408b79bcc3 nfsd4: consistent session flag setting
We should clear these flags on any new create_session, not just on the
first one.

Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2010-04-16 21:47:37 -04:00
Mark Brown
b2c812e22d ASoC: Add indirection for CODEC private data
One of the features of the multi CODEC work is that it embeds a struct
device in the CODEC to provide diagnostics via a sysfs class rather than
via the device tree, at which point it's much better to use the struct
device private data rather than having two places to store it. Provide
an accessor function to allow this change to be made more easily, and
update all the CODEC drivers are updated.

To ensure use of the accessor the private data structure member is
renamed, meaning that if code developed with older an older core that
still uses private_data is merged it will fail to build.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Acked-by: Liam Girdwood <lrg@slimlogic.co.uk>
2010-04-17 10:46:22 +09:00
Mark Brown
890c681275 Merge branch 'for-2.6.34' into for-2.6.35 2010-04-17 10:45:54 +09:00
Tom Herbert
fec5e652e5 rfs: Receive Flow Steering
This patch implements receive flow steering (RFS).  RFS steers
received packets for layer 3 and 4 processing to the CPU where
the application for the corresponding flow is running.  RFS is an
extension of Receive Packet Steering (RPS).

The basic idea of RFS is that when an application calls recvmsg
(or sendmsg) the application's running CPU is stored in a hash
table that is indexed by the connection's rxhash which is stored in
the socket structure.  The rxhash is passed in skb's received on
the connection from netif_receive_skb.  For each received packet,
the associated rxhash is used to look up the CPU in the hash table,
if a valid CPU is set then the packet is steered to that CPU using
the RPS mechanisms.

The convolution of the simple approach is that it would potentially
allow OOO packets.  If threads are thrashing around CPUs or multiple
threads are trying to read from the same sockets, a quickly changing
CPU value in the hash table could cause rampant OOO packets--
we consider this a non-starter.

To avoid OOO packets, this solution implements two types of hash
tables: rps_sock_flow_table and rps_dev_flow_table.

rps_sock_table is a global hash table.  Each entry is just a CPU
number and it is populated in recvmsg and sendmsg as described above.
This table contains the "desired" CPUs for flows.

rps_dev_flow_table is specific to each device queue.  Each entry
contains a CPU and a tail queue counter.  The CPU is the "current"
CPU for a matching flow.  The tail queue counter holds the value
of a tail queue counter for the associated CPU's backlog queue at
the time of last enqueue for a flow matching the entry.

Each backlog queue has a queue head counter which is incremented
on dequeue, and so a queue tail counter is computed as queue head
count + queue length.  When a packet is enqueued on a backlog queue,
the current value of the queue tail counter is saved in the hash
entry of the rps_dev_flow_table.

And now the trick: when selecting the CPU for RPS (get_rps_cpu)
the rps_sock_flow table and the rps_dev_flow table for the RX queue
are consulted.  When the desired CPU for the flow (found in the
rps_sock_flow table) does not match the current CPU (found in the
rps_dev_flow table), the current CPU is changed to the desired CPU
if one of the following is true:

- The current CPU is unset (equal to RPS_NO_CPU)
- Current CPU is offline
- The current CPU's queue head counter >= queue tail counter in the
rps_dev_flow table.  This checks if the queue tail has advanced
beyond the last packet that was enqueued using this table entry.
This guarantees that all packets queued using this entry have been
dequeued, thus preserving in order delivery.

Making each queue have its own rps_dev_flow table has two advantages:
1) the tail queue counters will be written on each receive, so
keeping the table local to interrupting CPU s good for locality.  2)
this allows lockless access to the table-- the CPU number and queue
tail counter need to be accessed together under mutual exclusion
from netif_receive_skb, we assume that this is only called from
device napi_poll which is non-reentrant.

This patch implements RFS for TCP and connected UDP sockets.
It should be usable for other flow oriented protocols.

There are two configuration parameters for RFS.  The
"rps_flow_entries" kernel init parameter sets the number of
entries in the rps_sock_flow_table, the per rxqueue sysfs entry
"rps_flow_cnt" contains the number of entries in the rps_dev_flow
table for the rxqueue.  Both are rounded to power of two.

The obvious benefit of RFS (over just RPS) is that it achieves
CPU locality between the receive processing for a flow and the
applications processing; this can result in increased performance
(higher pps, lower latency).

The benefits of RFS are dependent on cache hierarchy, application
load, and other factors.  On simple benchmarks, we don't necessarily
see improvement and sometimes see degradation.  However, for more
complex benchmarks and for applications where cache pressure is
much higher this technique seems to perform very well.

Below are some benchmark results which show the potential benfit of
this patch.  The netperf test has 500 instances of netperf TCP_RR
test with 1 byte req. and resp.  The RPC test is an request/response
test similar in structure to netperf RR test ith 100 threads on
each host, but does more work in userspace that netperf.

e1000e on 8 core Intel
   No RFS or RPS		104K tps at 30% CPU
   No RFS (best RPS config):    290K tps at 63% CPU
   RFS				303K tps at 61% CPU

RPC test	tps	CPU%	50/90/99% usec latency	Latency StdDev
  No RFS/RPS	103K	48%	757/900/3185		4472.35
  RPS only:	174K	73%	415/993/2468		491.66
  RFS		223K	73%	379/651/1382		315.61

Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 16:01:27 -07:00
Daniel Lezcano
1c4f019732 packet : remove init_net restriction
The af_packet protocol is used by Perl to do ioctls as reported by
Stephane Riviere:

"Net::RawIP relies on SIOCGIFADDR et SIOCGIFHWADDR to get the IP and MAC
addresses of the network interface."

But in a new network namespace these ioctl fail because it is disabled for
a namespace different from the init_net_ns.

These two lines should not be there as af_inet and af_packet are
namespace aware since a long time now. I suppose we forget to remove these
lines because we sent the af_packet first, before af_inet was supported.

Signed-off-by: Daniel Lezcano <daniel.lezcano@free.fr>
Reported-by: Stephane Riviere <stephane.riviere@regis-dgac.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 15:41:04 -07:00
Krzysztof Halasa
31f634a63d WAN: flush tx_queue in hdlc_ppp to prevent panic on rmmod hw_driver.
tx_queue is used as a temporary queue when not allowed to queue skb
directly to the hw device driver (which may sleep). Most paths flush
it before returning, but ppp_start() currently cannot. Make sure we
don't leave skbs pointing to a non-existent device.

Thanks to Michael Barkowski for reporting this problem.

Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-04-16 15:41:03 -07:00
Johannes Berg
e7cb49550e iwlwifi: make scan antenna forcing more generic
Some future hardware will also require some antenna
overrides so make the current logic more generic;
right now it is semantically based on a workaround
for off-channel reception but the reasons for the
new antenna overrides will be different.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-04-16 13:54:29 -07:00
Johannes Berg
ee102603c0 iwlwifi: remove monitor check
Off-channel reception is acceptable in monitor
mode, and checking for monitor mode this way is
not really correct anyway since it could be the
case while operating.

Now iwl_is_monitor_mode() is no longer used so
remove it completely.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-04-16 13:54:16 -07:00
Johannes Berg
fa58b6a3b3 iwlwifi: don't check monitor for scanning
Monitor mode operation need not (and probably should
not) affect scanning this way since real monitoring
can not properly happen while scanning anyway.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-04-16 13:54:06 -07:00
Johannes Berg
b2e8690d5a iwlwifi: rename TX_CMD_FLG_BT_DIS_MSK
The flag name is a little misleading, this
flag instructs the device to ignore bluetooth
messages for purposes of frame transmissions,
so rename the flag to TX_CMD_FLG_IGNORE_BT.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
2010-04-16 13:53:46 -07:00