Commit graph

291,781 commits

Author SHA1 Message Date
Shubhrajyoti D
eff245c82f i2c: Make i2cdev_notifier_call static
The function i2cdev_notifier_call is used only in i2c-dev file
making it static.
Also removes the following sparse warning

drivers/i2c/i2c-dev.c:582:5: warning: symbol 'i2cdev_notifier_call'
was not declared. Should it be static?

Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
2011-11-23 11:33:07 +01:00
Jean Delvare
67820021dc i2c: Delete ANY_I2C_BUS
Last piece of code using ANY_I2C_BUS was deleted almost 2 years ago,
so ANY_I2C_BUS can go away as well.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
2011-11-23 11:33:07 +01:00
Jean Delvare
cbb4451404 i2c: Fix device name for 10-bit slave address
10-bit addresses overlap with traditional 7-bit addresses, leading in
device name collisions. Add an arbitrary offset to 10-bit addresses to
prevent this collision. The offset was chosen so that the address is
still easily recognizable.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
2011-11-23 11:33:07 +01:00
Jeffrey (Sheng-Hui) Chu
cc6bcf7d2e i2c-algo-bit: Generate correct i2c address sequence for 10-bit target
The wrong bits were put on the wire, fix that.

This fixes kernel bug #42562.

Signed-off-by: Sheng-Hui J. Chu <jeffchu@broadcom.com>
Cc: stable@kernel.org
Signed-off-by: Jean Delvare <khali@linux-fr.org>
2011-11-23 11:33:07 +01:00
Paul Bolle
4ca8af579c ASoC: drop support for PlayPaq with WM8510
SoC Audio support for PlayPaq with WM8510 got added in commit 9aaca9683b
("[ALSA] Revised AT32 ASoC Patch"). That support depends on
BOARD_PLAYPAQ. That Kconfig symbol didn't exist when that support got
added in v2.6.27. It still doesn't. It has never been possible to even
build this driver. Drop it.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 10:28:38 +00:00
Stephen Warren
bea0ed0825 ASoC: Tegra I2S: Use devm_ APIs and module_platform_driver
module_platform_drive saves some boiler-plate code.

The devm_ APIs remove the need to manually clean up allocations,
thus removing some code.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 10:24:10 +00:00
Stephen Warren
65713ce844 ASoC: Tegra: Move DAS configuration into machine drivers
This removes potentially machine-specific routing knowledge from the
I2S driverinto the machine drivers, which is better equipped to know
what the appropriate routing configuration is.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 10:24:10 +00:00
Stephen Warren
f2296d7bf1 ASoC: Tegra DAS: Use devm_ APIs and module_platform_driver
module_platform_drive saves some boiler-plate code.

The devm_ APIs remove the need to manually clean up allocations,
thus removing some code.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 10:24:10 +00:00
Stephen Warren
99c92ae4ff ASoC: Tegra PCM: Use module_platform_driver
This saves some boiler-plate code.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 10:24:10 +00:00
Mark Brown
1db3c98e18 ASoC: Convert wm8776 to table based control and DAPM init
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 10:00:34 +00:00
Mike Snitzer
5151412dd4 block: initialize request_queue's numa node during
struct request_queue is allocated with __GFP_ZERO so its "node" field is
zero before initialization.  This causes an oops if node 0 is offline in
the page allocator because its zonelists are not initialized.  From Dave
Young's dmesg:

	SRAT: Node 1 PXM 2 0-d0000000
	SRAT: Node 1 PXM 2 100000000-330000000
	SRAT: Node 0 PXM 1 330000000-630000000
	Initmem setup node 1 0000000000000000-000000000affb000
	...
	Built 1 zonelists in Node order, mobility grouping on.
	...
	BUG: unable to handle kernel paging request at 0000000000001c08
	IP: [<ffffffff8111c355>] __alloc_pages_nodemask+0xb5/0x870

and __alloc_pages_nodemask+0xb5 translates to a NULL pointer on
zonelist->_zonerefs.

The fix is to initialize q->node at the time of allocation so the correct
node is passed to the slab allocator later.

Since blk_init_allocated_queue_node() is no longer needed, merge it with
blk_init_allocated_queue().

[rientjes@google.com: changelog, initializing q->node]
Cc: stable@vger.kernel.org [2.6.37+]
Reported-by: Dave Young <dyoung@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Tested-by: Dave Young <dyoung@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2011-11-23 10:59:13 +01:00
Mark Brown
404417e6b4 ASoC: Staticise and constify cs42l73_reg_defaults
It's not exported and doesn't need to change.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-23 09:52:56 +00:00
Xi Wang
a5cd335165 drm: integer overflow in drm_mode_dirtyfb_ioctl()
There is a potential integer overflow in drm_mode_dirtyfb_ioctl()
if userspace passes in a large num_clips.  The call to kmalloc would
allocate a small buffer, and the call to fb->funcs->dirty may result
in a memory corruption.

Reported-by: Haogang Chen <haogangchen@gmail.com>
Signed-off-by: Xi Wang <xi.wang@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-11-23 08:59:28 +00:00
Li Wei
4d65a2465f ipv6: fix a bug in ndisc_send_redirect
Release skb when transmit rate limit _not_ allow

Signed-off-by: Li Wei <lw@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2011-11-23 03:51:54 -05:00
Asai Thambi S P
60ec0eecfa mtip32xx: updates based on feedback
* queue ncq commands when a non-ncq is in progress or error handling is active
* merge variables 'internal_cmd_in_progress' and 'eh_active' into new variable 'flags'
* get rid of read/write semaphore 'internal_sem'
* new service thread to issue queued commands
* use macros from ata.h for command codes
* return ENOTTY for BLKFLSBUF ioctl
* style changes

Signed-off-by: Asai Thambi S P <asamymuthupa@micron.com>
Signed-off-by: Sam Bradshaw <sbradshaw@micron.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2011-11-23 08:29:24 +01:00
Takashi Iwai
61071594f6 ALSA: hda/realtek - Minor cleanup
Use an inline function for the common pattern for assigning a capsrc.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2011-11-23 07:52:15 +01:00
Takashi Iwai
6759dc3238 ALSA: hda/realtek - Fix missing inits of item indices for auto-mic
When the imux entries are rebuilt in alc_rebuild_imux_for_auto_mic(),
the initialization of index field is missing.  It may work without it
casually when the original imux was created by the auto-parser, but
it's definitely broken in the case of static configs where no imux was
parsed beforehand.  Because of this, the auto-mic switching doesn't
work properly on some model options.

This patch adds the missing initialization of index field.

Reported-by: Dmitry Nezhevenko <dion@inhex.net>
Cc: <stable@kernel.org> [v3.1]
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2011-11-23 07:45:21 +01:00
Takashi Iwai
6dfeb703e3 ALSA: hda - Fix invalid pin and GPIO for Apple laptops with CS codecs
The PCI SSID 8086:7270 is commonly used for multiple Apple machines,
thus we can't use it as identifier for a unique model.  Because of this
conflict, some machines show weird behavior.  For example, MacBook Air
shows Front and Surround speakers although only Surround works due to
the wrongly overridden pin-configuration for imac27.

This patch fixes two things:
- Stop the wrong pin-config override of imac27 by removing PCI SSID
  entry for avoiding the wrong mappings,
- Add the generic GPIO setup for Apple machines by checking the codec
  SSID vendor bits

Tested-by: Linus Torvalds <torvalds@linux-foundation.org>
Tested-by: Dirk Hohndel <hohndel@infradead.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2011-11-23 07:31:49 +01:00
Takashi Iwai
e2301a4de2 ALSA: hda - Check subdevice mask in snd_hda_check_board_codec_sid_config()
In snd_hda_check_board_codec_sid_config(), not only comparing with the
exact value but allow the bit-mask comparison for vendor-only, etc.

Tested-by: Linus Torvalds <torvalds@linux-foundation.org>
Tested-by: Dirk Hohndel <hohndel@infradead.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2011-11-23 07:31:36 +01:00
Arve Hjønnevåg
6a8943d9ec rtc: Fix some bugs that allowed accumulating time drift in suspend/resume
The current code checks if abs(delta_delta.tv_sec) is greater or
equal to two before it discards the old delta value, but this can
trigger at close to -1 seconds since -1.000000001 seconds is stored
as tv_sec -2 and tv_nsec 999999999 in a normalized timespec.

rtc_resume had an early return check if the rtc value had not changed
since rtc_suspend. This effectivly stops time for the duration of the
short sleep. Check if sleep_time is positive after all the adjustments
have been applied instead since this allows the old_system adjustment
in rtc_suspend to have an effect even for short sleep cycles.

CC: stable@kernel.org
Signed-off-by: Arve Hjønnevåg <arve@android.com>
Signed-off-by: John Stultz <john.stultz@linaro.org>
2011-11-22 19:25:56 -08:00
Rabin Vincent
c0afabd3d5 rtc: Disable the alarm in the hardware
Currently, the RTC code does not disable the alarm in the hardware.

This means that after a sequence such as the one below (the files are in the
RTC sysfs), the box will boot up after 2 minutes even though we've
asked for the alarm to be turned off.

	# echo $((`cat since_epoch`)+120) > wakealarm
	# echo 0 > wakealarm
	# poweroff

Fix this by disabling the alarm when there are no timers to run.

Cc: stable@kernel.org
Cc: John Stultz <john.stultz@linaro.org>
Signed-off-by: Rabin Vincent <rabin.vincent@stericsson.com>
Signed-off-by: John Stultz <john.stultz@linaro.org>
2011-11-22 19:25:43 -08:00
NeilBrown
7c8f424798 md/lock: ensure updates to page_attrs are properly locked.
Page attributes are set using __set_bit rather than set_bit as
it normally called under a spinlock so the extra atomicity is not
needed.

However there are two places where we might set or clear page
attributes without holding the spinlock.
So add the spinlock in those cases.

This might be the cause of occasional reports that bits a aren't
getting clear properly - theory is that BITMAP_PAGE_PENDING gets lost
when BITMAP_PAGE_NEEDWRITE is set or cleared.  This is an
inconvenience, not a threat to data safety.

Signed-off-by: NeilBrown <neilb@suse.de>
2011-11-23 10:18:52 +11:00
Linus Torvalds
b4bbb02934 Revert "of/irq: of_irq_find_parent: check for parent equal to child"
This reverts commit dc93728084.

As requested by Ben Herrenschmidt:
  "This breaks some powerpc platforms at least.  The practice of having
   a node provide an explicit "interrupt-parent" property pointing to
   itself is an old trick that we've used in the past to allow a
   device-node to have interrupts routed to different controllers.

   In that case, the node also contains an interrupt-map, so the node is
   its own parent, the interrupt resolution hits the map, which then can
   route each individual interrupt to a different parent."

Grant says:
  "Ah, nuts, yes that is broken then.  Yes, please revert the commit and
   Rob & I will come up with a better solution.

   Rob, I think it can be done by explicitly checking for np ==
   desc->interrupt_parent in of_irq_init() instead of relying on
   of_irq_find_parent() returning NULL."

Requested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: Grant Likely <grant.likely@secretlab.ca>
Cc: Rob Herring <rob.herring@calxeda.com>
Cc: devicetree-discuss@lists.ozlabs.org
Cc: linuxppc-dev <linuxppc-dev@lists.ozlabs.org>
Cc: Tanmay Inamdar <tinamdar@apm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-11-22 15:09:20 -08:00
Mark Brown
34c3a64a45 Merge branch 'for-3.2' into for-3.3 2011-11-22 23:07:23 +00:00
Timur Tabi
380c883038 ASoC: mpc8610: tell the CS4270 codec that it's the master
Commit ac601555 ("ASoC: Return early with -EINVAL if invalid dai format is
detected") requires the machine driver to tell the CS4270 codec driver
whether the CS4270 should be configured for master or slave operation.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-22 23:06:25 +00:00
Daniel Mack
d66b8537b3 ASoC: cs4720: use snd_soc_cache_sync()
Replace the manual register restore mechanism in cs4270.c and call the
generic snd_soc_cache_sync() handler instead.

This factors code out in favour of core facilities and also fixes a
bus confusion that is most probably caused by intermixing i2c-regmap
functions and i2c_smbus_* accessors.

Signed-off-by: Daniel Mack <zonque@gmail.com>
Reported-and-tested-by: Sven Neumann <s.neumann@raumfeld.com>
Acked-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-11-22 23:06:07 +00:00
H. Peter Anvin
9147621c77 kbuild, headers.sh: Don't make archheaders explicitly
We don't need to explicitly invoke the archheaders target because of
the dependency on __headers in the Makefile.

Reported-and-tested-by: Michal Marek <mmarek@suse.cz>
Link: http://lkml.kernel.org/r/4ECA8991.20302@suse.cz
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2011-11-22 14:47:50 -08:00
Paul Bolle
7e0ea2e6c8 sound: Kconfig: drop unknown symbol ARCH_CLPS7500
Commit 635f0258e5 ("[ARM] clps7500: remove support") missed one instance
of the ARCH_CLPS7500 Kconfig symbol. Drop it now.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:28:23 +01:00
Jiri Kosina
f2c4826c68 HID: picolcd: make fb_pending_lock and picolcd_fb_cleanup static
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:25:28 +01:00
Jiri Kosina
b3c21d2cad HID: multitouch: make struct mt_classess static
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:23:37 +01:00
David Herrmann
90120d6697 HID: wiimote: Enable NO_INIT_REPORTS quirk
Newer bluetooth stack supports the NO_INIT_REPORTS quirk. The wiimote does not
support report initialization so enable it by default.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:11:34 +01:00
David Herrmann
130a69b9eb HID: wiimote: Remove module version number
The version number is not needed at all for in-tree drivers. Upstream git is
used to track module versions.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:11:19 +01:00
David Herrmann
43d782ae80 HID: wiimote: Allow direct DRM debug access
Keep track of current drm and add new debugfs file which reads or writes the
current DRM.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:11:10 +01:00
David Herrmann
1d3452c63d HID: wiimote: Allow direct eeprom access
The wiimote provides direct access to parts of its eeprom. This implements read
support for small chunks of the eeprom. This isn't very fast but prevents the
reader from blocking the wiimote stream for too long.

Write support is not yet supported as the wiimote breaks if we overwrite its
memory. Use hidraw to reverse-engineer the eeprom before implementing write
support here.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:10:58 +01:00
David Herrmann
43e5e7c60e HID: wiimote: Add debugfs support stubs
Add initializer and deinitializer for debugfs support. This will later allow raw
eeprom access and direct DRM modifications to debug wiimote behaviour and
further protocol reverse-engineerings.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:10:26 +01:00
David Herrmann
5906215bab HID: wiimote: Parse classic controller data
Nintendo Classic Controller extension reports lots of keys, two analog sticks
and two analog buttons. We report all data through extension input device to
userspace.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:10:10 +01:00
David Herrmann
a53535014b HID: wiimote: Parse nunchuck data
The Nintendo Nunchuck extension reports accelerometer values, one analog stick
and two buttons. See inline comments for data layout.
We report all data to userspace through extension input device.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:09:56 +01:00
David Herrmann
b17b57a5d0 HID: wiimote: Parse motion+ data
Motion+ reports rotation gyro data which we report to userspace as ABS_RX/Y/Z
values. The device reports them either in fast or slow mode. We adjust the
values to get a linear scale so userspace does not need to know about slow and
fast mode.

The motion+ also reports whether an extension is connected to it. We keep track
of this value and reinitialize the extensions if an extension is plugged or
unplugged.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:09:45 +01:00
David Herrmann
0b6815d75d HID: wiimote: Add extension handler stubs
All supported extensions report data as 6 byte block. All DRMs with extension
data provide at least 6 extension bytes. Hence a generic handler for all
extension bytes is sufficient and can be called on all DRMs.

The handler distinguishes the input and passes it to the right handler. Motion+
passes data interleaved so we can have Motion+ and a regular extension enabled
simultaneously.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:09:34 +01:00
David Herrmann
479901ba18 HID: wiimote: Register input devices for extensions
Motion+ and regular extensions are physical adapters for the wiimote so create
one input device for each of them. This also allows to enable only opened
extensions and turn unused extenions off to save battery power.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:09:23 +01:00
David Herrmann
c1e51398a1 HID: wiimote: Add extension sysfs attribute
Add new sysfs attribute "extension" which returns the currently connected and
initialized extensions.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:09:08 +01:00
David Herrmann
492ba955c1 HID: wiimote: Add extension initializers
The wiimote extension registers are not fully understood, so we always disable
all extensions on extension-port events. Then we reinitialize and reidentify
them and activate all requested extensions.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:08:55 +01:00
David Herrmann
82fb1b3958 HID: wiimote: Add extension initializer stubs
Add stub functions to read and identify extensions and then initialize all
connected extensions.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:08:40 +01:00
David Herrmann
cb99221ba7 HID: wiimote: Add extension support stub
The wiimote supports several extensions. This adds a separate source file which
handles all extensions and can be disabled at compile-time.

The driver reacts on "plug"-events on the extension port and starts a worker
which initializes or deinitializes the extensions.

Currently, the initialization logic is not fully understood and we can only
detect and enable all extensions when all extensions are deactivated. Therefore,
we need to disable all extensions, then detect and activate them again to react
on "plug"-events.
However, deactivating extensions will generate a new "plug"-event and we will
never leave that loop. Hence, we only support extensions if they are plugged
before the wiimote is connected (or before the ext-input device is opened). In
the future we may support full extension hotplug support, but
reverse-engineering this may take a while.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:08:28 +01:00
David Herrmann
fad8c0e343 HID: wiimote: Add read-mem helpers
Add helper functions similar to the write-mem helpers but for reading wiimote
memory and eeprom.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:08:09 +01:00
David Herrmann
7e27440062 HID: wiimote: Move common symbols into header
Wiimote extension and sound support need access to several symbols so move them
into a new header.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:07:53 +01:00
David Herrmann
192a1acfbd HID: wiimote: Rename driver to allow multiple source files
Extension and sound support for the wiimote are quite complex and will be
implemented in separate source files. Hence rename the current driver to "-core"
suffix so multiple files can be linked into this module.

Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 23:06:31 +01:00
Jesper Juhl
e56e0b7dd5 SCSI, qla2xxx: remove redundant semicolon
One ';' at the end of the return statement is enough.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Acked-by: Chad Dupuis <chad.dupuis@qlogic.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 22:59:05 +01:00
Paul Bolle
282e5aaba2 x86: Kconfig: drop unknown symbol 'APM_MODULE'
There's no Kconfig symbol APM_MODULE, so the check for it will always
fail. There's no need to append _MODULE to tristate symbols anyhow,
because the config tools will do the right thing automagically.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2011-11-22 22:57:20 +01:00
Trond Myklebust
24ca9a8477 SUNRPC: Ensure we return EAGAIN in xs_nospace if congestion is cleared
By returning '0' instead of 'EAGAIN' when the tests in xs_nospace() fail
to find evidence of socket congestion, we are making the RPC engine believe
that the message was incorrectly sent and so it disconnects the socket
instead of just retrying.

The bug appears to have been introduced by commit
5e3771ce2d (SUNRPC: Ensure that xs_nospace
return values are propagated).

Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org [>= 2.6.30]
Tested-by: Andrew Cooper <andrew.cooper3@citrix.com>
2011-11-22 23:55:27 +02:00