drivers/rtc/rtc-ds3234.c: fix whitespace issue

Fixes the following warning:
WARNING: please, no space before tabs

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-ds1511.c: fix issues related to spaces and braces

Fixes the following types of issues:
WARNING: please, no spaces at the start of a line
WARNING: braces {} are not necessary for single statement blocks

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-ds1374.c: fix spacing related issues

Fixes the following types of issues:
ERROR: code indent should use tabs where possible
WARNING: please, no spaces at the start of a line

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-ds1305.c: add missing braces around sizeof

Silences the following type of warnings:
WARNING: sizeof buf should be sizeof(buf)

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-davinci.c: fix whitespace warning

Silences the following warning:
WARNING: please, no space before tabs

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-cmos.c: fix whitespace related errors

Fixes the following types of issues:

ERROR: space required after that ',' (ctx:VxV)
WARNING: please, no spaces at the start of a line

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-at91rm9200.c: include <linux/uaccess.h>

Silences the following checkpatch warning:
WARNING: Use #include <linux/uaccess.h> instead of <asm/uaccess.h>

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Andrew Victor <linux@maxim.org.za>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-at32ap700x.c: fix checkpatch error

Fixes the following error:
ERROR: space required before the open parenthesis '('

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/interface.c: fix checkpatch errors

Fixes the following types of errors:
ERROR: "foo* bar" should be "foo *bar"
ERROR: else should follow close brace '}'
WARNING: braces {} are not necessary for single statement blocks

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/rtc/rtc-v3020.c: remove redundant goto

Remove a redundant goto statement left over during the conversion of
this driver to use devm_* APIs.

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

rtc: rtc-88pm80x: remove unnecessary platform_set_drvdata()

The driver core clears the driver data to NULL after device_release or on
probe failure, since commit 0998d063100 ("device-core: Ensure drvdata =
NULL when no driver is bound").  Thus, it is not needed to manually clear
the device driver data to NULL.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

autofs4: translate pids to the right namespace for the daemon

The PID and the TGID of the process triggering the mount are sent to the
daemon.  Currently the global pid values are sent (ones valid in the
initial pid namespace) but this is wrong if the autofs daemon itself is
not running in the initial pid namespace.

So send the pid values that are valid in the namespace of the autofs daemon.

The namespace to use is taken from the oz_pgrp pid pointer, which was set
at mount time to the mounting process' pid namespace.

If the pid translation fails (the triggering process is in an unrelated
pid namespace) then the automount fails with ENOENT.

Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Cc: Eric Biederman <ebiederm@xmission.com>
Cc: Ian Kent <raven@themaw.net>
Cc: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

autofs4: allow autofs to work outside the initial PID namespace

Enable autofs4 to work in a "container".  oz_pgrp is converted from pid_t
to struct pid and this is stored at mount time based on the "pgrp=" option
or if the option is missing then the current pgrp.

The "pgrp=" option is interpreted in the PID namespace of the current
process.  This option is flawed in that it doesn't carry the namespace
information, so it should be deprecated.  AFAICS the autofs daemon always
sends the current pgrp, which is the default anyway.

The oz_pgrp is also set from the AUTOFS_DEV_IOCTL_SETPIPEFD_CMD ioctl.
This ioctl sets oz_pgrp to the current pgrp.  It is not allowed to change
the pid namespace.

oz_pgrp is used mainly to determine whether the process traversing the
autofs mount tree is the autofs daemon itself or not.  This function now
compares the pid pointers instead of the pid_t values.

One other use of oz_pgrp is in autofs4_show_options.  There is shows the
virtual pid number (i.e.  the one that is valid inside the PID namespace
of the calling process)

For debugging printk convert oz_pgrp to the value in the initial pid
namespace.

Signed-off-by: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Cc: Eric Biederman <ebiederm@xmission.com>
Cc: Ian Kent <raven@themaw.net>
Cc: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kprobes: handle empty/invalid input to debugfs "enabled" file

When writing invalid input to 'debug/kprobes/enabled' it'll silently be
ignored.  Even worse, when writing an empty string to this file, the
outcome is purely random as the switch statement will make its decision
based on the value of an uninitialized stack variable.

Fix this by handling invalid/empty input as error returning -EINVAL.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

insert missing space in printk line of root_delay

Trivial, but it really looks better.

Signed-off-by: Toralf Förster <toralf.foerster@gmx.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

init: remove permanent string buffer from do_one_initcall()

do_one_initcall() uses a 64 byte string buffer to save a message. This
buffer is declared static and is only used at boot up and when a module
is loaded. As 64 bytes is very small, and this function has very limited
scope, there's no reason to waste permanent memory with this string and
not just simply put it on the stack.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

binfmt_elf.c: use get_random_int() to fix entropy depleting

Entropy is quickly depleted under normal operations like ls(1), cat(1),
etc...  between 2.6.30 to current mainline, for instance:

$ cat /proc/sys/kernel/random/entropy_avail
3428
$ cat /proc/sys/kernel/random/entropy_avail
2911
$cat /proc/sys/kernel/random/entropy_avail
2620

We observed this problem has been occurring since 2.6.30 with
fs/binfmt_elf.c: create_elf_tables()->get_random_bytes(), introduced by
f06295b44c296c8f ("ELF: implement AT_RANDOM for glibc PRNG seeding").

/*
 * Generate 16 random bytes for userspace PRNG seeding.
 */
get_random_bytes(k_rand_bytes, sizeof(k_rand_bytes));

The patch introduces a wrapper around get_random_int() which has lower
overhead than calling get_random_bytes() directly.

With this patch applied:
$ cat /proc/sys/kernel/random/entropy_avail
2731
$ cat /proc/sys/kernel/random/entropy_avail
2802
$ cat /proc/sys/kernel/random/entropy_avail
2878

Analyzed by John Sobecki.

Signed-off-by: Jie Liu <jeff.liu@oracle.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Andreas Dilger <aedilger@gmail.com>
Cc: Alan Cox <alan@linux.intel.com>
Cc: Arnd Bergmann <arnn@arndb.de>
Cc: John Sobecki <john.sobecki@oracle.com>
Cc: James Morris <james.l.morris@oracle.com>
Cc: Jakub Jelinek <jakub@redhat.com>
Cc: Ted Ts'o <tytso@mit.edu>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Kees Cook <keescook@chromium.org>
Cc: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: allow longer logging function names

The current $logFunction regular expression allows names like dev_warn,
e_dbg, netdev_info, etc, but some log functions are now written like
e_dev_warn, so allow 1 or 2 word blocks with an underscore before the
logging level.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: ignore existing CamelCase uses from include/...

When using --strict, CamelCase uses are described with CHECK: messages.
These CamelCase uses may be acceptable and should not generate these
messages when the variable is already defined in a file from the
include/...  path.

So, change checkpatch to read all the .h files in include/...  and look
for preexisting CamelCase #defines, typedefs and function prototypes.

Add these to the existing camelcase hash so that any uses in the patch or
file can be ignored.

There are currently ~3500 files in include/.  It takes about 10 cpu
seconds on my little netbook to grep for and preseed these existing uses.

That's about 4x the time for a similar git grep.

This preseeding is only done once when using --strict and only when there
is a CamelCase use found.

If a .git directory is found, it uses 'git ls-files include' If not, it
uses 'find $root/include -name "*.h"

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: ignore SI unit CamelCase variants like "_uV"

Many existing variable names use SI like variants that should be otherwise
obvious and acceptable.

Whitelist them from the CamelCase message.

Signed-off-by: Joe Perches <joe@perches.com>
Suggested-by: Phil Carmody <phil.carmody@partner.samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: move test for space before semicolon after operator spacing

Moving this test allows the --fix option to work better.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: create an EXPERIMENTAL --fix option to correct patches

Some patches have simple defects in whitespace and formatting that
checkpatch could correct automatically.  Attempt to do so.

Add a --fix option to create a "<inputfile>.EXPERIMENTAL-checkpatch-fixes"
file that tries to use normal kernel style for some of these formatting
errors.

Add warnings against using this file without verifying the changes.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: improve "no space after cast" test

Some false positives exist on this test.

For instance:
*va_arg(args, signed char *) = val.s;
or
memset(foo, 0, sizeof(struct bar *) * baz));

Ignore lines that have an arithmetic operator or assignment
after what appears to be a cast to a pointer "(foo *)".

Add $Arithmetic convenience variable.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: add a --strict test for comparison to true/false

Comparing to true or false is error prone.

Add tests for the various forms of (foo == true) && (false != bar)
that are only reported with --strict.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: don't warn on blank lines before/after braces as often

Check to make sure the blank lines aren't comment lines like:

bool foo(bool bar)
{
/* Don't warn on a leading comment */
return !bar;
/* Don't warn on a trailing comment either */
}

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: add a placeholder to check blank lines before declarations

Figure out first how to determine if this is in a struct declaration or in
a function body before enabling this.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: reduce false positive rate of "complex macros"

Allow "#define foo struct.member" without bleating a warning.

This also allows "#define foo bar.baz->qux" and so on.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: warn on comparisons to get_jiffies_64()

Comparing get_jiffies_64() is almost always wrong and time_before64 and
time_after64 should be used instead.

Warn on any comparison to get_jiffies_64().

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: warn on comparisons to jiffies

Comparing jiffies is almost always wrong and time_before and time_after
should be used instead.

Warn on any comparison to jiffies.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: warn when networking block comment lines don't start with *

Some block comments in network are written as:

/* block comment line 1
   block comment line 2
 */

Emit a warning on the "block comment line 2" because it should be

/* block comment line 1
 * block comment line 2
 */

This warning is only emitted on the second line of a block comment.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: improve network block comment test and message

Show the first line of the comment after a line with just /* to better
show where the defective comment style is in the file.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: remove quote from CamelCase test

commit be987d9f80 ("checkpatch: improve CamelCase test for Page")
added it but it shouldn't be there.  Must have been my fault.

Make sure that the tested variable doesn't contain a constant.

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: add --strict preference for p = kmalloc(sizeof(*p)...

Add another test for memory allocation style to follow
Documentation/CodingStyle:

Chapter 14: Allocating memory

The preferred form for passing a size of a struct is the following:

p = kmalloc(sizeof(*p), ...);

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: warn when using gcc's binary constant ("0b") extension

The gcc extension for binary constants that start with 0b is only
supported with gcc version 4.3 or higher.

The kernel can still be compiled with earlier versions of gcc, so have
checkpatch emit a warning for these constants.

Restructure checkpatch's constant finding code a bit to support finding
these binary constants.

Signed-off-by: Joe Perches <joe@perches.com>
Suggested-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

checkpatch: change CamelCase test and make it --strict

Do not bleat a message on nominally acceptable CamelCase uses that are
separated by an _ like drm_core_has_MTRR.

CamelCase tests are also a bit noisy against certain types of code
acceptable to some kernel developers.

Make the test applicable only with --strict.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

list: remove __list_for_each()

__list_for_each used to be the non prefetch() aware list walking
primitive.  When we removed the prefetch macros from the list routines, it
became redundant.  Given it does exactly the same thing as list_for_each
now, we might as well remove it and call list_for_each directly.

Signed-off-by: Dave Jones <davej@redhat.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: "John W. Linville" <linville@tuxdriver.com>
Cc: Clemens Ladisch <clemens@ladisch.de>
Cc: Dave Airlie <airlied@linux.ie>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jaroslav Kysela <perex@perex.cz>
Cc: Jennifer Naumann <Jennifer.Naumann@informatik.stud.uni-erlangen.de>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Sebastian Hahn <snsehahn@cip.cs.fau.de>
Cc: Stanislav Yakovlev <stas.yakovlev@gmail.com>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: Vlad Yasevich <vyasevich@gmail.com>
Cc: YAMANE Toshiaki <yamanetoshi@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

ipw2200: convert __list_for_each usage to list_for_each

Signed-off-by: Dave Jones <davej@redhat.com>
Cc: Stanislav Yakovlev <stas.yakovlev@gmail.com>
Cc: "John W. Linville" <linville@tuxdriver.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

radeon: remove redundant __list_for_each definition from mkregtable.c

Signed-off-by: Dave Jones <davej@redhat.com>
Cc: Dave Airlie <airlied@linux.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

rbtree-remove-unneeded-include-fix

fix lib/interval_tree.c build

Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

rbtree: remove unneeded include

Remove an unnecessary include to kernel.h

Signed-off-by: Nathan Zimmer <nzimmer@sgi.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: add CONFIG_PM_SLEEP to suspend/resume functions

Add CONFIG_PM_SLEEP to suspend/resume functions to fix the following build
warning when CONFIG_PM_SLEEP is not selected.  This is because sleep PM
callbacks defined by SIMPLE_DEV_PM_OPS are only used when the
CONFIG_PM_SLEEP is enabled.

drivers/video/backlight/backlight.c:211:12: warning: 'backlight_suspend' defined but not used [-Wunused-function]
drivers/video/backlight/backlight.c:225:12: warning: 'backlight_resume' defined but not used [-Wunused-function]

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Shuah Khan <shuahkhan@gmail.com>
Cc: Jingoo Han <jg1.han@samsung.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: convert from legacy pm ops to dev_pm_ops

Convert drivers/video/backlight/class to use dev_pm_ops for power
management and remove Legacy PM ops hooks.  With this change, backlight
class registers suspend/resume callbacks via class->pm (dev_pm_ops)
instead of Legacy class->suspend/resume.  When __device_suspend() runs
call-backs, it will find class->pm ops for the backlight class.

Signed-off-by: Shuah Khan <shuah.kh@samsung.com>
Cc: Shuah Khan <shuahkhan@gmail.com>
Cc: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

MAINTAINERS: add Backlight subsystem co-maintainer

Add myself as co-maintainer for the backlight subsystem.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

lcd: add devm_lcd_device_{register,unregister}()

These functions allow the driver core to automatically clean up any
allocation made by lcd drivers.  Thus it simplifies the error paths.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: add devm_backlight_device_{register,unregister}()

These functions allow the driver core to automatically clean up any
allocation made by backlight drivers. Thus it simplifies the error
paths.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: pcf50633: remove unnecessary platform_set_drvdata()

The driver core clears the driver data to NULL after device_release or on
probe failure, since commit 0998d063100 ("device-core: Ensure drvdata =
NULL when no driver is bound").  Thus, it is not needed to manually clear
the device driver data to NULL.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: lp8788: remove unnecessary platform_set_drvdata()

The driver core clears the driver data to NULL after device_release or on
probe failure, since commit 0998d063100 ("device-core: Ensure drvdata =
NULL when no driver is bound").  Thus, it is not needed to manually clear
the device driver data to NULL.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: ep93xx: remove unnecessary platform_set_drvdata()

The driver core clears the driver data to NULL after device_release or on
probe failure, since commit 0998d063100 ("device-core: Ensure drvdata =
NULL when no driver is bound").  Thus, it is not needed to manually clear
the device driver data to NULL.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

backlight: atmel-pwm-bl: remove unnecessary platform_set_drvdata()

The driver core clears the driver data to NULL after device_release or on
probe failure, since commit 0998d06310 ("device-core: Ensure drvdata =
NULL when no driver is bound").  Thus, it is not needed to manually clear
the device driver data to NULL.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

MAINTAINERS: fix tape driver file mappings

The masks for the st and osst tape drivers in MAINTAINERS are too
broad and include unrelated files. Make the file list accurate so that
maintainers of these drivers aren't bothered with unrelated work.

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Cc: Willem Riede <osst@riede.org>
Cc: Kai Mäkisara <Kai.Makisara@kolumbus.fi>
Cc: "James E.J. Bottomley" <JBottomley@parallels.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/misc/sgi-gru/grufault.c: fix a sanity test in gru_set_context_option()

"req.val1 == -1" is valid but it doesn't make sense to check gru_base[-1].
 gru_base[] is a global array.

Signed-off-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Robin Holt <holt@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/dma: remove unused support for MEMSET operations

There have never been any real users of MEMSET operations since they have
been introduced in January 2007 by 7405f74badf4 ("dmaengine: refactor
dmaengine around dma_async_tx_descriptor").  Therefore remove support for
them for now, it can be always brought back when needed.

Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Dan Williams <djbw@fb.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

smp: Give WARN()ing when calling smp_call_function_many()/single() in serving irq

Currently the functions smp_call_function_many()/single() will give a
WARN()ing only in the case of irqs_disabled(), but that check is not
enough to guarantee execution of the SMP cross-calls.

In many other cases such as softirq handling/interrupt handling, the two
APIs still can not be called, just as the smp_call_function_many()
comments say:

  * You must not call this function with disabled interrupts or from a
  * hardware interrupt handler or from a bottom half handler. Preemption
  * must be disabled when calling this function.

There is a real case for softirq DEADLOCK case:

CPUA                            CPUB
                                spin_lock(&spinlock)
                                Any irq coming, call the irq handler
                                irq_exit()
spin_lock_irq(&spinlock)
<== Blocking here due to
CPUB hold it
                                  __do_softirq()
                                    run_timer_softirq()
                                      timer_cb()
                                        call smp_call_function_many()
                                          send IPI interrupt to CPUA
                                            wait_csd()

Then both CPUA and CPUB will be deadlocked here.

So we should give a warning in the nmi, hardirq or softirq context as well.

Moreover, adding one new macro in_serving_irq() which indicates we are
processing nmi, hardirq or sofirq.

Signed-off-by: liu chuansheng <chuansheng.liu@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Tested-by: Fengguang Wu <fengguang.wu@intel.com>
Cc: Lai Jiangshan <eag0628@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drm/i915: quirk away phantom LVDS on Intel's D525MW mainboard

This replaceable mainboard only has a VGA-out, yet it claims to also have
a connected LVDS header.

Addresses https://bugs.freedesktop.org/show_bug.cgi?id=65256

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reported-by: Cornel Panceac <cpanceac@gmail.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <annndddrr@gmail.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drm/i915: quirk away phantom LVDS on Intel's D510MO mainboard

This replaceable mainboard only has a VGA-out, yet it claims to also have
a connected LVDS header.

Addresses https://bugs.freedesktop.org/show_bug.cgi?id=63860

[jani.nikula@intel.com: use DMI_EXACT_MATCH for board name.]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reported-by: <annndddrr@gmail.com>
Cc: Cornel Panceac <cpanceac@gmail.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

dmi: add support for exact DMI matches in addition to substring matching

dmi_match() considers a substring match to be a successful match.  This is
not always sufficient to distinguish between DMI data for different
systems.  Add support for exact string matching using strcmp() in addition
to the substring matching using strstr().

The specific use case in the i915 driver is to allow us to use an exact
match for D510MO, without also incorrectly matching D510MOV:

{
.ident = "Intel D510MO",
.matches = {
DMI_MATCH(DMI_BOARD_VENDOR, "Intel"),
DMI_EXACT_MATCH(DMI_BOARD_NAME, "D510MO"),
},
}

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Cc: <annndddrr@gmail.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Cornel Panceac <cpanceac@gmail.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kernel/sys.c:do_sysinfo(): use get_monotonic_boottime()

Change do_sysinfo() to use get_monotonic_boottime() instead of
do_posix_clock_monotonic_gettime() + monotonic_to_bootbased().

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Acked-by: John Stultz <johnstul@us.ibm.com>
Cc: Tomas Janousek <tjanouse@redhat.com>
Cc: Tomas Smetana <tsmetana@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kernel/sys.c: sys_reboot(): fix malformed panic message

If LINUX_REBOOT_CMD_HALT for reboot failed, the message "cannot halt" will
stay on the same line with the next message, so append a '\n'.

Signed-off-by: liguang <lig.fnst@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

panic-add-cpu-pid-to-warn_slowpath_common-in-warning-printks-fix

remove stray quote

Cc: Alex Thorlton <athorlton@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

panic: add cpu/pid to warn_slowpath_common in WARNING printk()s

Add the cpu/pid that called WARN() so that the stack traces can be matched
up with the WARNING messages.

Signed-off-by: Alex Thorlton <athorlton@sgi.com>
Reviewed-by: Robin Holt <holt@sgi.com>
Cc: Stephen Boyd <sboyd@codeaurora.org>
Cc: Vikram Mulukutla <markivx@codeaurora.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

dump_stack-serialize-the-output-from-dump_stack-fix

- fix comment indenting
- avoid inclusion of <asm/> files - use <linux/> where possible
- fix uniprocessor build (__dump_stack undefined)
- remove unneeded ifdef around smp.h inclusion

Cc: Alex Thorlton <athorlton@sgi.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Jesper Nilsson <jesper.nilsson@axis.com>
Cc: Richard Kuo <rkuo@codeaurora.org>
Cc: Robin Holt <holt@sgi.com>
Cc: Russ Anderson <rja@sgi.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: athorlton@sgi.com
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

dump_stack: serialize the output from dump_stack()

tAdd adds functionality to serialize the output from dump_stack() to avoid
mangling of the output when dump_stack is called simultaneously from
multiple cpus.

Signed-off-by: Alex Thorlton <athorlton@sgi.com>
Reported-by: Russ Anderson <rja@sgi.com>
Reviewed-by: Robin Holt <holt@sgi.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Richard Kuo <rkuo@codeaurora.org>
Cc: Jesper Nilsson <jesper.nilsson@axis.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers: avoid parsing names as kthread_run() format strings

Calling kthread_run with a single name parameter causes it to be handled
as a format string. Many callers are passing potentially dynamic string
content, so use "%s" in those cases to avoid any potential accidents.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers: avoid format strings in names passed to alloc_workqueue()

For the workqueue creation interfaces that do not expect format strings,
make sure they cannot accidently be parsed that way.  Additionally, clean
up calls made with a single parameter that would be handled as a format
string.  Many callers are passing potentially dynamic string content, so
use "%s" in those cases to avoid any potential accidents.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers: avoid format string in dev_set_name

Calling dev_set_name with a single paramter causes it to be handled as a
format string.  Many callers are passing potentially dynamic string
content, so use "%s" in those cases to avoid any potential accidents,
including wrappers like device_create*() and bdi_register().

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

clean-up-scary-strncpydst-src-strlensrc-uses-fix

revert getdelays.c part due to missing bsd/string.h

Cc: Frank Blaschka <blaschka@linux.vnet.ibm.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> [staging]
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com> [acpi]
Cc: Richard Weinberger <richard@nod.at>
Acked-by: Ursula Braun <ursula.braun@de.ibm.com> [s390]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

clean up scary strncpy(dst, src, strlen(src)) uses

Fix various weird constructions of strncpy(dst, src, strlen(src)).  Length
limits should be about the space available in the destination, not
repurposed as a method to either always include or always exclude a
trailing NULL byte.  Either the NULL should always be copied (using
strlcpy), or it should not be copied (using something like memcpy).
Readable code should not depend on the weird behavior of strncpy when it
hits the length limit.  Better to avoid the anti-pattern entirely.

Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> [staging]
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> [acpi]
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Ursula Braun <ursula.braun@de.ibm.com>
Cc: Frank Blaschka <blaschka@linux.vnet.ibm.com>
Cc: Richard Weinberger <richard@nod.at>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

err.h: IS_ERR() can accept __user pointers

Sparse generates a false positive when you pass a __user or __iomem
pointer to the IS_ERR() functions.

drivers/rtc/rtc-ds1286.c:344:36: sparse: incorrect type in argument 1 (different address spaces)
drivers/rtc/rtc-ds1286.c:344:36:    expected void const *ptr
drivers/rtc/rtc-ds1286.c:344:36:    got unsigned int [noderef] [usertype] <asn:2>*rtcregs

We can silence these by adding a __force here and upgrading to Sparse
v0.4.5-rc1 or later.

This change has no effect when using current Sparse releases.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Christopher Li <sparse@chrisli.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

arch/frv/kernel/setup.c: use strncmp() instead of memcmp()

'cmdline' is a NUL terminated string, when its length < 4, memcmp()
will cause memory access out of boundary.

So use strncmp() instead of memcmp().

Signed-off-by: Chen Gang <gang.chen@asianux.com>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

arch/frv/kernel/traps.c: using vsnprintf() instead of vsprintf()

Since die_if_kernel() is an extern common used function, better always
check the buffer length to avoid memory overflow by a long 'str'.

Signed-off-by: Chen Gang <gang.chen@asianux.com>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: add vm event counters for balloon pages compaction

Introduce a new set of vm event counters to keep track of ballooned pages
compaction activity.

Signed-off-by: Rafael Aquini <aquini@redhat.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Minchan Kim <minchan@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/dmapool.c: fix null dev in dma_pool_create()

A few drivers invoke dma_pool_create() with a null dev.  Note that dev is
dereferenced in dev_to_node(dev), causing a null pointer dereference.

A long term solution is to disallow null dev.  Once the drivers are fixed,
we can simplify the core code here.  For now we add WARN_ON(!dev) to
notify the driver maintainers and avoid the null pointer dereference.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drivers/usb/gadget/amd5536udc.c: avoid calling dma_pool_create() with NULL dev

Calling dma_pool_create() with dev==NULL will oops on a NUMA machine.
Rather than changing dma_pool_create() we wish to disallow passing
dev==NULL.  This requires fixing up the small number of drivers which are
passing in dev==NULL.

Use &dev->pdev->dev instead of NULL.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

drop_caches: add some documentation and info message

I would like to resurrect Dave's patch.  The last time it was posted was
here https://lkml.org/lkml/2010/9/16/250 and there didn't seem to be any
strong opposition.

Kosaki was worried about possible excessive logging when somebody drops
caches too often (but then he claimed he didn't have a strong opinion on
that) but I would say opposite.  If somebody does that then I would really
like to know that from the log when supporting a system because it almost
for sure means that there is something fishy going on.  It is also worth
mentioning that only root can write drop caches so this is not an flooding
attack vector.

I am bringing that up again because this can be really helpful when
chasing strange performance issues which (surprise surprise) turn out to
be related to artificially dropped caches done because the admin thinks
this would help...

I have just refreshed the original patch on top of the current mm tree
but I could live with KERN_INFO as well if people think that KERN_NOTICE
is too hysterical.

: From: Dave Hansen <dave@linux.vnet.ibm.com>
: Date: Fri, 12 Oct 2012 14:30:54 +0200
:
: There is plenty of anecdotal evidence and a load of blog posts
: suggesting that using "drop_caches" periodically keeps your system
: running in "tip top shape".  Perhaps adding some kernel
: documentation will increase the amount of accurate data on its use.
:
: If we are not shrinking caches effectively, then we have real bugs.
: Using drop_caches will simply mask the bugs and make them harder
: to find, but certainly does not fix them, nor is it an appropriate
: "workaround" to limit the size of the caches.
:
: It's a great debugging tool, and is really handy for doing things
: like repeatable benchmark runs.  So, add a bit more documentation
: about it, and add a little KERN_NOTICE.  It should help developers
: who are chasing down reclaim-related bugs.

[mhocko@suse.cz: refreshed to current -mm tree]
[akpm@linux-foundation.org: checkpatch fixes]
Signed-off-by: Dave Hansen <dave@linux.vnet.ibm.com>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: memmap_init_zone() performance improvement

We have what we call an "architectural simulator".  It is a computer
program that pretends that it is a computer system.  We use it to test the
firmware before real hardware is available.  We have booted Linux on our
simulator.  As you would expect it takes longer to boot on the simulator
than it does on real hardware.

With my patch - boot time 41 minutes
Without patch - boot time 94 minutes

These numbers do not scale linearly to real hardware.  But indicate to me
a place where Linux can be improved.

memmap_init_zone() loops through every Page Frame Number (pfn), including
pfn values that are within the gaps between existing memory sections.  The
unneeded looping will become a boot performance issue when machines
configure larger memory ranges that will contain larger and more numerous
gaps.

The code will skip across invalid pfn values to reduce the number of loops
executed.

Signed-off-by: Mike Yoknis <mike.yoknis@hp.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

include/linux/mmzone.h: cleanups

- implement zone_idx() in C to fix its references-args-twice macro bug

- use zone_idx() in is_highmem() to remove large amounts of silly fluff.

Cc: Lin Feng <linfeng@cn.fujitsu.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: vmscan: do not scale writeback pages when deciding whether to set ZONE_WRITEBACK

After the patch "mm: vmscan: Flatten kswapd priority loop" was merged the
scanning priority of kswapd changed.  The priority now raises until it is
scanning enough pages to meet the high watermark.  shrink_inactive_list
sets ZONE_WRITEBACK if a number of pages were encountered under writeback
but this value is scaled based on the priority.  As kswapd frequently
scans with a higher priority now it is relatively easy to set
ZONE_WRITEBACK.  This patch removes the scaling and treates writeback
pages similar to how it treats unqueued dirty pages and congested pages.
The user-visible effect should be that kswapd will writeback fewer pages
from reclaim context.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: vmscan: avoid direct reclaim scanning at maximum priority

Further testing revealed that swapping was still higher than expected for
the parallel IO tests.  There was also a performance regression reported
building kernels but there appears to be multiple sources of that problem.
 This follow-up series primarily addresses the first swapping issue.

The tests were based on three kernels

vanilla:        kernel 3.10-rc4 as that is what the current mmotm uses as a baseline
mmotm-20130606  is mmotm as of that date.
lessdisrupt-v1 is this follow-up series on top of the mmotm kernel

The first test used memcached+memcachetest while some background IO was in
progress as implemented by the parallel IO tests implement in MM Tests.
memcachetest benchmarks how many operations/second memcached can service.
It starts with no background IO on a freshly created ext4 filesystem and
then re-runs the test with larger amounts of IO in the background to
roughly simulate a large copy in progress.  The expectation is that the IO
should have little or no impact on memcachetest which is running entirely
in memory.

parallelio
                                        3.10.0-rc4                  3.10.0-rc4                  3.10.0-rc4
                                           vanilla          mm1-mmotm-20130606        mm1-lessdisrupt-v1
Ops memcachetest-0M             23018.00 (  0.00%)          22412.00 ( -2.63%)          22556.00 ( -2.01%)
Ops memcachetest-715M           23383.00 (  0.00%)          22810.00 ( -2.45%)          22431.00 ( -4.07%)
Ops memcachetest-2385M          10989.00 (  0.00%)          23564.00 (114.43%)          23054.00 (109.79%)
Ops memcachetest-4055M           3798.00 (  0.00%)          24004.00 (532.02%)          24050.00 (533.23%)
Ops io-duration-0M                  0.00 (  0.00%)              0.00 (  0.00%)              0.00 (  0.00%)
Ops io-duration-715M               12.00 (  0.00%)              7.00 ( 41.67%)              7.00 ( 41.67%)
Ops io-duration-2385M             133.00 (  0.00%)             21.00 ( 84.21%)             22.00 ( 83.46%)
Ops io-duration-4055M             159.00 (  0.00%)             36.00 ( 77.36%)             36.00 ( 77.36%)
Ops swaptotal-0M                    0.00 (  0.00%)              0.00 (  0.00%)              0.00 (  0.00%)
Ops swaptotal-715M             139693.00 (  0.00%)             19.00 ( 99.99%)              0.00 (  0.00%)
Ops swaptotal-2385M            268541.00 (  0.00%)              0.00 (  0.00%)             19.00 ( 99.99%)
Ops swaptotal-4055M            414269.00 (  0.00%)          22059.00 ( 94.68%)              2.00 (100.00%)
Ops swapin-0M                       0.00 (  0.00%)              0.00 (  0.00%)              0.00 (  0.00%)
Ops swapin-715M                     0.00 (  0.00%)              0.00 (  0.00%)              0.00 (  0.00%)
Ops swapin-2385M                73189.00 (  0.00%)              0.00 (  0.00%)              0.00 (  0.00%)
Ops swapin-4055M               126292.00 (  0.00%)              0.00 (  0.00%)              0.00 (  0.00%)
Ops minorfaults-0M            1536018.00 (  0.00%)        1533536.00 (  0.16%)        1607381.00 ( -4.65%)
Ops minorfaults-715M          1789978.00 (  0.00%)        1616152.00 (  9.71%)        1533462.00 ( 14.33%)
Ops minorfaults-2385M         1910448.00 (  0.00%)        1614060.00 ( 15.51%)        1661727.00 ( 13.02%)
Ops minorfaults-4055M         1760518.00 (  0.00%)        1613980.00 (  8.32%)        1615116.00 (  8.26%)
Ops majorfaults-0M                  0.00 (  0.00%)              0.00 (  0.00%)              5.00 (-99.00%)
Ops majorfaults-715M              169.00 (  0.00%)            234.00 (-38.46%)             48.00 ( 71.60%)
Ops majorfaults-2385M           14899.00 (  0.00%)            100.00 ( 99.33%)            222.00 ( 98.51%)
Ops majorfaults-4055M           21853.00 (  0.00%)            150.00 ( 99.31%)            128.00 ( 99.41%)

memcachetest is the transactions/second reported by memcachetest. In
        the vanilla kernel note that performance drops from around
        23K/sec to just over 4K/second when there is 2385M of IO going
        on in the background. With current mmotm and the follow-on
series performance is good.

swaptotal is the total amount of swap traffic. With mmotm the total amount
of swapping is much reduced. Note that with 4G of background IO that
this follow-up series almost completely eliminated swap IO.

                            3.10.0-rc4  3.10.0-rc4  3.10.0-rc4
                               vanillamm1-mmotm-20130606mm1-lessdisrupt-v1
Minor Faults                  11230171    10689656    10650607
Major Faults                     37255         786         705
Swap Ins                        199724           0           0
Swap Outs                       623022       22078          21
Direct pages scanned                 0        5378       51660
Kswapd pages scanned          15892718     1610408     1653629
Kswapd pages reclaimed         1097093     1083339     1107652
Direct pages reclaimed               0        5024       47241
Kswapd efficiency                   6%         67%         66%
Kswapd velocity              13633.275    1385.369    1420.058
Direct efficiency                 100%         93%         91%
Direct velocity                  0.000       4.626      44.363
Percentage direct scans             0%          0%          3%
Zone normal velocity         13474.405     671.123     697.927
Zone dma32 velocity            158.870     718.872     766.494
Zone dma velocity                0.000       0.000       0.000
Page writes by reclaim     3065275.000   27259.000    6316.000
Page writes file               2442253        5181        6295
Page writes anon                623022       22078          21
Page reclaim immediate            8019         429         318
Sector Reads                    963320       99096      151864
Sector Writes                 13057396    10887480    10878500
Page rescued immediate               0           0           0
Slabs scanned                    64896       23168       34176
Direct inode steals                  0           0           0
Kswapd inode steals               8668           0           0
Kswapd skipped wait                  0           0           0

Few observations

1. Swap outs were almost completely eliminated and there were no swap-ins.

2. Direct reclaim is active due to reduced activity from kswapd and the fact
   that it is no longer reclaiming at priority 0

3. Zone scanning is still relatively balanced.

4. Page writes from reclaim context is still reasonable low.

                  3.10.0-rc4  3.10.0-rc4  3.10.0-rc4
                     vanillamm1-mmotm-20130606mm1-lessdisrupt-v1
Mean sda-avgqz        168.05       34.64       35.60
Mean sda-await        831.76      216.31      207.05
Mean sda-r_await        7.88        9.68        7.25
Mean sda-w_await     3088.32      223.90      218.28
Max  sda-avgqz       1162.17      766.85      795.69
Max  sda-await       6788.75     4130.01     3728.43
Max  sda-r_await      106.93      242.00       65.97
Max  sda-w_await    30565.93     4145.75     3959.87

Wait times are marginally reduced by the follow-up and still a massive
improve against the mainline kernel.

I tested parallel kernel builds when booted with 1G of RAM. 12 kernels
were built with 2 being compiled at any given time.

multibuild
                          3.10.0-rc4            3.10.0-rc4            3.10.0-rc4
                             vanilla    mm1-mmotm-20130606  mm1-lessdisrupt-v1
User    min         584.99 (  0.00%)      553.31 (  5.42%)      569.08 (  2.72%)
User    mean        598.35 (  0.00%)      574.48 (  3.99%)      581.65 (  2.79%)
User    stddev       10.01 (  0.00%)       17.90 (-78.78%)       10.03 ( -0.14%)
User    max         614.64 (  0.00%)      598.94 (  2.55%)      597.97 (  2.71%)
User    range        29.65 (  0.00%)       45.63 (-53.90%)       28.89 (  2.56%)
System  min          35.78 (  0.00%)       35.05 (  2.04%)       35.54 (  0.67%)
System  mean         36.12 (  0.00%)       35.69 (  1.20%)       35.88 (  0.69%)
System  stddev        0.26 (  0.00%)        0.55 (-113.69%)        0.21 ( 17.51%)
System  max          36.53 (  0.00%)       36.44 (  0.25%)       36.13 (  1.09%)
System  range         0.75 (  0.00%)        1.39 (-85.33%)        0.59 ( 21.33%)
Elapsed min         190.54 (  0.00%)      190.56 ( -0.01%)      192.99 ( -1.29%)
Elapsed mean        197.58 (  0.00%)      203.30 ( -2.89%)      200.53 ( -1.49%)
Elapsed stddev        4.65 (  0.00%)        5.26 (-13.16%)        5.66 (-21.79%)
Elapsed max         203.72 (  0.00%)      210.23 ( -3.20%)      210.46 ( -3.31%)
Elapsed range        13.18 (  0.00%)       19.67 (-49.24%)       17.47 (-32.55%)
CPU     min         308.00 (  0.00%)      282.00 (  8.44%)      294.00 (  4.55%)
CPU     mean        320.80 (  0.00%)      299.78 (  6.55%)      307.67 (  4.09%)
CPU     stddev       10.44 (  0.00%)       13.83 (-32.50%)        9.71 (  7.01%)
CPU     max         340.00 (  0.00%)      333.00 (  2.06%)      328.00 (  3.53%)
CPU     range        32.00 (  0.00%)       51.00 (-59.38%)       34.00 ( -6.25%)

Average kernel build times are still impacted but the follow-up series
helps marginally (it's too noisy to be sure).  A preliminary bisection
indicated that there were multiple sources of the regression.  The two
other points are the patches that cause mark_page_accessed to be obeyed
and the slab shrinker series.  As there a number of patches in flight to
mmotm at the moment in different areas it would be best to confirm this
after this follow-up is merged.

This patch (of 2):

Page reclaim at priority 0 will scan the entire LRU as priority 0 is
considered to be a near OOM condition.  Direct reclaim can reach this
priority while still making reclaim progress.  This patch avoids
reclaiming at priority 0 unless no reclaim progress was made and the page
allocator would consider firing the OOM killer.  The user-visible impact
is that direct reclaim will not easily reach priority 0 and start swapping
prematurely.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memory_hotplug.c: fix a comment typo in register_page_bootmem_info_node()

Signed-off-by: Tang Chen <tangchen@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memblock.c: fix wrong comment in __next_free_mem_range()

Remove one redundant "nid" in the comment.

Signed-off-by: Tang Chen <tangchen@cn.fujitsu.com>

page migration: fix wrong comment in address_space_operations.migratepage()

There is no parameter "sync" in address_space_operations->migratepage().
It should be mograte_mode.  And the comment is for MIGRATE_ASYNC.

Signed-off-by: Tang Chen <tangchen@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/vmalloc.c: fix an overflow bug in alloc_vmap_area()

When searching a vmap area in the vmalloc space, we use (addr + size - 1)
to check if the value is less than addr, which is an overflow.  But we
assign (addr + size) to vmap_area->va_end.

So if we come across the  below case:

(addr + size - 1) : not overflow
(addr + size)     : overflow

we will assign an overflow value (e.g 0) to vmap_area->va_end, And this
will trigger BUG in __insert_vmap_area, causing system panic.

So using (addr + size) to check the overflow should be the correct
behaviour, not (addr + size - 1).

Signed-off-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Reported-by: Ghennadi Procopciuc <unix140@gmail.com>
Tested-by: Daniel Baluta <dbaluta@ixiacom.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove unused VM_<READfoo> macros and expand other in-place

These VM_<READfoo> macros aren't used very often and three of them aren't
used at all.

Expand the ones that are used in-place, and remove all the now unused
#define VM_<foo> macros.

VM_READHINTMASK, VM_NormalReadHint and VM_ClearReadHint were added just
before 2.4 and appears have never been used.

Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/pgtable: don't accumulate addr during pgd prepopulate pmd

The old codes accumulate addr to get right pmd, however, currently pmds
are preallocated and transfered as a parameter, there is unnecessary to
accumulate addr variable any more, this patch remove it.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/thp: fix doc for transparent huge zero page

Transparent huge zero page is used during the page fault instead of in
khugepaged.

# ls /sys/kernel/mm/transparent_hugepage/
defrag  enabled  khugepaged  use_zero_page
# ls /sys/kernel/mm/transparent_hugepage/khugepaged/
alloc_sleep_millisecs  defrag  full_scans  max_ptes_none  pages_collapsed  pages_to_scan  scan_sleep_millisecs

This patch corrects the documentation just like the codes done.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_alloc: fix doc for numa_zonelist_order

The default zonelist order selecter will select "node" order if any node's
DMA zone comprises greater than 70% of its local memory instead of 60%,
according to default_zonelist_order::low_kmem_size > total * 70/100.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/writeback: commit reason of WB_REASON_FORKER_THREAD mismatch name

After 839a8e86 ("writeback: replace custom worker pool implementation with
unbound workqueue"), there is no bdi forker thread any more.  However,
WB_REASON_FORKER_THREAD is still used due to it is TPs userland visible
and we won't be exposing exactly the same information with just a
different name.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/writeback: don't check force_wait to handle bdi->work_list

After 839a8e86 ("writeback: replace custom worker pool implementation with
unbound workqueue"), bdi_writeback_workfn runs off bdi_writeback->dwork,
on each execution, it processes bdi->work_list and reschedules if there
are more things to do instead of flush any work that race with us
existing.  It is unecessary to check force_wait in wb_do_writeback since
it is always 0 after the mentioned commit.  This patch remove the
force_wait in wb_do_writeback.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/writeback: remove wb_reason_name

wb_reason_name is not used any more - remove it.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

fs/fs-writeback.c: : make wb_do_writeback() as static

It's not used globally and could be static.

Signed-off-by: Haicheng Li <haicheng.li@linux.intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/sparse.c: put clear_hwpoisoned_pages within CONFIG_MEMORY_HOTREMOVE

With CONFIG_MEMORY_HOTREMOVE unset, there is a compile warning:

mm/sparse.c:755: warning: `clear_hwpoisoned_pages' defined but not used

And Bisecting it ended up pointing to 4edd7ceff ("mm, hotplug: avoid
compiling memory hotremove functions when disabled").

This is because the commit above put sparse_remove_one_section() within
the protection of CONFIG_MEMORY_HOTREMOVE but the only user of
clear_hwpoisoned_pages() is sparse_remove_one_section(), and it is not
within the protection of CONFIG_MEMORY_HOTREMOVE.

So put clear_hwpoisoned_pages within CONFIG_MEMORY_HOTREMOVE should fix
the warning.

Signed-off-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Cc: David Rientjes <rientjes@google.com>
Acked-by: Toshi Kani <toshi.kani@hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove unused __put_page()

This function is nowhere used, and it has a confusing name with put_page
in mm/swap.c.  So better to remove it.

Signed-off-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

vfree: don't schedule free_work() if llist_add() returns false

vfree() only needs schedule_work(&p->wq) if p->list was empty, otherwise
vfree_deferred->wq is already pending or it is running and didn't do
llist_del_all() yet.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_alloc.c: remove unlikely() from the current_order test

In __rmqueue_fallback(), current_order loops down from MAX_ORDER - 1 to
the order passed.  MAX_ORDER is typically 11 and pageblock_order is
typically 9 on x86.  Integer division truncates, so pageblock_order / 2 is
4.  For the first eight iterations, it's guaranteed that current_order >=
pageblock_order / 2 if it even gets that far!

So just remove the unlikely(), it's completely bogus.

Signed-off-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Suggested-by: David Rientjes <rientjes@google.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: remove unused functions is_{normal_idx, normal, dma32, dma}

These functions are nowhere used, so remove them.

Signed-off-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_alloc.c: remove zone_type argument of build_zonelists_node

The callers of build_zonelists_node always pass MAX_NR_ZONES -1 as the
zone_type argument, so we can directly use the value in
build_zonelists_node and remove zone_type argument.

Signed-off-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

MAINTAINERS: add zswap and zbud maintainer

Add maintainer information for zswap and zbud into the MAINTAINERS file.

Signed-off-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

zswap: add documentation

Add the documentation file for the zswap functionality

Signed-off-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Dan Magenheimer <dan.magenheimer@oracle.com>
Cc: Robert Jennings <rcj@linux.vnet.ibm.com>
Cc: Jenifer Hopper <jhopper@us.ibm.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Larry Woodman <lwoodman@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Dave Hansen <dave@sr71.net>
Cc: Joe Perches <joe@perches.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Cody P Schafer <cody@linux.vnet.ibm.com>
Cc: Hugh Dickens <hughd@google.com>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

zswap: fix Kconfig to depend on CRYPTO=y

The Kconfig entry for zswap currently allows CRYPTO=m to satisfy the
zswap dependency.  However zswap is boolean (i.e. built-in) and has
symbol dependencies on CRYPTO.  Additionally, because the CRYPTO dependency
is satisfied with =m, the additional selects zswap does can also
be satisfied with =m, which leads to additional linking errors.

>From the report:
=====
tree:   git://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-3.9
head:   57cefddb141d6c9d0ab4f5a8589dd017f796a3f7
commit: 563f51c95b552e0d663df5bdf6cfc8e8a72d3ec6 [494/499] zswap: add to mm/
config: x86_64-randconfig-x004-0619 (attached as .config)

All error/warnings:

   mm/built-in.o: In function `zswap_frontswap_invalidate_area':
>> zswap.c:(.text+0x3a705): undefined reference to `zbud_free'
   mm/built-in.o: In function `zswap_free_entry':
>> zswap.c:(.text+0x3a76b): undefined reference to `zbud_free'
>> zswap.c:(.text+0x3a789): undefined reference to `zbud_get_pool_size'
on and on...
=====

This patch makes CRYPTO a built-in dependency of ZSWAP.  This has the
side effect of also making the selects built-in.

Signed-off-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

zswap: add to mm/

zswap is a thin backend for frontswap that takes pages that are in the
process of being swapped out and attempts to compress them and store them
in a RAM-based memory pool.  This can result in a significant I/O
reduction on the swap device and, in the case where decompressing from RAM
is faster than reading from the swap device, can also improve workload
performance.

It also has support for evicting swap pages that are currently compressed
in zswap to the swap device on an LRU(ish) basis.  This functionality
makes zswap a true cache in that, once the cache is full, the oldest pages
can be moved out of zswap to the swap device so newer pages can be
compressed and stored in zswap.

This patch adds the zswap driver to mm/

Signed-off-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Dan Magenheimer <dan.magenheimer@oracle.com>
Cc: Robert Jennings <rcj@linux.vnet.ibm.com>
Cc: Jenifer Hopper <jhopper@us.ibm.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Larry Woodman <lwoodman@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Dave Hansen <dave@sr71.net>
Cc: Joe Perches <joe@perches.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Cody P Schafer <cody@linux.vnet.ibm.com>
Cc: Hugh Dickens <hughd@google.com>
Cc: Paul Mackerras <paulus@samba.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>