Initial commit; kernel source import

tools/testing/fault-injection/failcmd.sh

@@ -0,0 +1,219 @@
+#!/bin/bash
+#
+# NAME
+#	failcmd.sh - run a command with injecting slab/page allocation failures
+#
+# SYNOPSIS
+#	failcmd.sh --help
+#	failcmd.sh [<options>] command [arguments]
+#
+# DESCRIPTION
+#	Run command with injecting slab/page allocation failures by fault
+#	injection.
+#
+#	NOTE: you need to run this script as root.
+#
+
+usage()
+{
+	cat >&2 <<EOF
+Usage: $0 [options] command [arguments]
+
+OPTIONS
+	-p percent
+	--probability=percent
+		likelihood of failure injection, in percent.
+		Default value is 1
+
+	-t value
+	--times=value
+		specifies how many times failures may happen at most.
+		Default value is 1
+
+	--oom-kill-allocating-task=value
+		set /proc/sys/vm/oom_kill_allocating_task to specified value
+		before running the command.
+		Default value is 1
+
+	-h, --help
+		Display a usage message and exit
+
+	--interval=value, --space=value, --verbose=value, --task-filter=value,
+	--stacktrace-depth=value, --require-start=value, --require-end=value,
+	--reject-start=value, --reject-end=value, --ignore-gfp-wait=value
+		See Documentation/fault-injection/fault-injection.txt for more
+		information
+
+	failslab options:
+	--cache-filter=value
+
+	fail_page_alloc options:
+	--ignore-gfp-highmem=value, --min-order=value
+
+ENVIRONMENT
+	FAILCMD_TYPE
+		The following values for FAILCMD_TYPE are recognized:
+
+		failslab
+			inject slab allocation failures
+		fail_page_alloc
+			inject page allocation failures
+
+		If FAILCMD_TYPE is not defined, then failslab is used.
+EOF
+}
+
+if [ $UID != 0 ]; then
+	echo must be run as root >&2
+	exit 1
+fi
+
+DEBUGFS=`mount -t debugfs | head -1 | awk '{ print $3}'`
+
+if [ ! -d "$DEBUGFS" ]; then
+	echo debugfs is not mounted >&2
+	exit 1
+fi
+
+FAILCMD_TYPE=${FAILCMD_TYPE:-failslab}
+FAULTATTR=$DEBUGFS/$FAILCMD_TYPE
+
+if [ ! -d $FAULTATTR ]; then
+	echo $FAILCMD_TYPE is not available >&2
+	exit 1
+fi
+
+LONGOPTS=probability:,interval:,times:,space:,verbose:,task-filter:
+LONGOPTS=$LONGOPTS,stacktrace-depth:,require-start:,require-end:
+LONGOPTS=$LONGOPTS,reject-start:,reject-end:,oom-kill-allocating-task:,help
+
+if [ $FAILCMD_TYPE = failslab ]; then
+	LONGOPTS=$LONGOPTS,ignore-gfp-wait:,cache-filter:
+elif [ $FAILCMD_TYPE = fail_page_alloc ]; then
+	LONGOPTS=$LONGOPTS,ignore-gfp-wait:,ignore-gfp-highmem:,min-order:
+fi
+
+TEMP=`getopt -o p:i:t:s:v:h --long $LONGOPTS -n 'failcmd.sh' -- "$@"`
+
+if [ $? != 0 ]; then
+	usage
+	exit 1
+fi
+
+eval set -- "$TEMP"
+
+fault_attr_default()
+{
+	echo N > $FAULTATTR/task-filter
+	echo 0 > $FAULTATTR/probability
+	echo 1 > $FAULTATTR/times
+}
+
+fault_attr_default
+
+oom_kill_allocating_task_saved=`cat /proc/sys/vm/oom_kill_allocating_task`
+
+restore_values()
+{
+	fault_attr_default
+	echo $oom_kill_allocating_task_saved \
+		> /proc/sys/vm/oom_kill_allocating_task
+}
+
+#
+# Default options
+#
+declare -i oom_kill_allocating_task=1
+declare task_filter=Y
+declare -i probability=1
+declare -i times=1
+
+while true; do
+	case "$1" in
+	-p|--probability)
+		probability=$2
+		shift 2
+		;;
+	-i|--interval)
+		echo $2 > $FAULTATTR/interval
+		shift 2
+		;;
+	-t|--times)
+		times=$2
+		shift 2
+		;;
+	-s|--space)
+		echo $2 > $FAULTATTR/space
+		shift 2
+		;;
+	-v|--verbose)
+		echo $2 > $FAULTATTR/verbose
+		shift 2
+		;;
+	--task-filter)
+		task_filter=$2
+		shift 2
+		;;
+	--stacktrace-depth)
+		echo $2 > $FAULTATTR/stacktrace-depth
+		shift 2
+		;;
+	--require-start)
+		echo $2 > $FAULTATTR/require-start
+		shift 2
+		;;
+	--require-end)
+		echo $2 > $FAULTATTR/require-end
+		shift 2
+		;;
+	--reject-start)
+		echo $2 > $FAULTATTR/reject-start
+		shift 2
+		;;
+	--reject-end)
+		echo $2 > $FAULTATTR/reject-end
+		shift 2
+		;;
+	--oom-kill-allocating-task)
+		oom_kill_allocating_task=$2
+		shift 2
+		;;
+	--ignore-gfp-wait)
+		echo $2 > $FAULTATTR/ignore-gfp-wait
+		shift 2
+		;;
+	--cache-filter)
+		echo $2 > $FAULTATTR/cache_filter
+		shift 2
+		;;
+	--ignore-gfp-highmem)
+		echo $2 > $FAULTATTR/ignore-gfp-highmem
+		shift 2
+		;;
+	--min-order)
+		echo $2 > $FAULTATTR/min-order
+		shift 2
+		;;
+	-h|--help)
+		usage
+		exit 0
+		shift
+		;;
+	--)
+		shift
+		break
+		;;
+	esac
+done
+
+[ -z "$1" ] && exit 0
+
+echo $oom_kill_allocating_task > /proc/sys/vm/oom_kill_allocating_task
+echo $task_filter > $FAULTATTR/task-filter
+echo $probability > $FAULTATTR/probability
+echo $times > $FAULTATTR/times
+
+trap "restore_values" SIGINT SIGTERM EXIT
+
+cmd="echo 1 > /proc/self/make-it-fail && exec $@"
+bash -c "$cmd"

tools/testing/ktest/compare-ktest-sample.pl

@@ -0,0 +1,32 @@
+#!/usr/bin/perl
+
+open (IN,"ktest.pl");
+while (<IN>) {
+    # hashes are now used
+    if (/\$opt\{"?([A-Z].*?)(\[.*\])?"?\}/ ||
+	/^\s*"?([A-Z].*?)"?\s*=>\s*/ ||
+	/set_test_option\("(.*?)"/) {
+	$opt{$1} = 1;
+    }
+}
+close IN;
+
+open (IN, "sample.conf");
+while (<IN>) {
+    if (/^\s*#?\s*([A-Z]\S*)\s*=/) {
+	$samp{$1} = 1;
+    }
+}
+close IN;
+
+foreach $opt (keys %opt) {
+    if (!defined($samp{$opt})) {
+	print "opt = $opt\n";
+    }
+}
+
+foreach $samp (keys %samp) {
+    if (!defined($opt{$samp})) {
+	print "samp = $samp\n";
+    }
+}

tools/testing/ktest/examples/README

@@ -0,0 +1,32 @@
+This directory contains example configs to use ktest for various tasks.
+The configs still need to be customized for your environment, but it
+is broken up by task which makes it easier to understand how to set up
+ktest.
+
+The configs are based off of real working configs but have been modified
+and commented to show more generic use cases that are more helpful for
+developers.
+
+crosstests.conf - this config shows an example of testing a git repo against
+    lots of different architectures. It only does build tests, but makes
+    it easy to compile test different archs. You can download the arch
+    cross compilers from:
+  http://kernel.org/pub/tools/crosstool/files/bin/x86_64/
+
+test.conf - A generic example of a config. This is based on an actual config
+     used to perform real testing.
+
+kvm.conf - A example of a config that is used to test a virtual guest running
+     on a host.
+
+snowball.conf - An example config that was used to demo ktest.pl against
+     a snowball ARM board.
+
+include/  -  The include directory holds default configs that can be
+    included into other configs. This is a real use example that shows how
+    to reuse configs for various machines or set ups. The files here
+    are included by other config files, where the other config files define
+    options and variables that will make the included config work for the
+    given environment.
+
+

tools/testing/ktest/examples/crosstests.conf

@@ -0,0 +1,260 @@
+#
+# Example config for cross compiling
+#
+# In this config, it is expected that the tool chains from:
+#
+#   http://kernel.org/pub/tools/crosstool/files/bin/x86_64/
+#
+# running on a x86_64 system have been downloaded and installed into:
+#
+#   /usr/local/
+#
+# such that the compiler binaries are something like:
+#
+#   /usr/local/gcc-4.5.2-nolibc/mips-linux/bin/mips-linux-gcc
+#
+# Some of the archs will use gcc-4.5.1 instead of gcc-4.5.2
+# this config uses variables to differentiate them.
+# 
+# Comments describe some of the options, but full descriptions of
+# options are described in the samples.conf file.
+
+# ${PWD} is defined by ktest.pl to be the directory that the user
+# was in when they executed ktest.pl. It may be better to hardcode the
+# path name here. THIS_DIR is the variable used through out the config file
+# in case you want to change it.
+
+THIS_DIR := ${PWD}
+
+# Update the BUILD_DIR option to the location of your git repo you want to test.
+BUILD_DIR = ${THIS_DIR}/linux.git
+
+# The build will go into this directory. It will be created when you run the test.
+OUTPUT_DIR = ${THIS_DIR}/cross-compile
+
+# The build will be compiled with -j8
+BUILD_OPTIONS = -j8
+
+# The test will not stop when it hits a failure.
+DIE_ON_FAILURE = 0
+
+# If you want to have ktest.pl store the failure somewhere, uncomment this option
+# and change the directory where ktest should store the failures.
+#STORE_FAILURES = ${THIS_DIR}/failures
+
+# The log file is stored in the OUTPUT_DIR called cross.log
+# If you enable this, you need to create the OUTPUT_DIR. It wont be created for you.
+LOG_FILE = ${OUTPUT_DIR}/cross.log
+
+# The log file will be cleared each time you run ktest.
+CLEAR_LOG = 1
+
+# As some archs do not build with the defconfig, they have been marked
+# to be ignored. If you want to test them anyway, change DO_FAILED to 1.
+# If a test that has been marked as DO_FAILED passes, then you should change
+# that test to be DO_DEFAULT
+
+DO_FAILED := 0
+DO_DEFAULT := 1
+
+# By setting both DO_FAILED and DO_DEFAULT to zero, you can pick a single
+# arch that you want to test. (uncomment RUN and chose your arch)
+#RUN := m32r
+
+# At the bottom of the config file exists a bisect test. You can update that
+# test and set DO_FAILED and DO_DEFAULT to zero, and uncomment this variable
+# to run the bisect on the arch.
+#RUN := bisect
+
+# By default all tests will be running gcc 4.5.2. Some tests are using 4.5.1
+# and they select that in the test.
+# Note: GCC_VER is declared as on option and not a variable ('=' instead of ':=')
+# This is important. A variable is used only in the config file and if it is set
+# it stays that way for the rest of the config file until it is change again.
+# Here we want GCC_VER to remain persistent and change for each test, as it is used in
+# the MAKE_CMD. By using '=' instead of ':=' we achieve our goal.
+
+GCC_VER = 4.5.2
+MAKE_CMD = PATH=/usr/local/gcc-${GCC_VER}-nolibc/${CROSS}/bin:$PATH CROSS_COMPILE=${CROSS}- make ARCH=${ARCH}
+
+# all tests are only doing builds.
+TEST_TYPE = build
+
+# If you want to add configs on top of the defconfig, you can add those configs into
+# the add-config file and uncomment this option. This is useful if you want to test
+# all cross compiles with PREEMPT set, or TRACING on, etc.
+#ADD_CONFIG = ${THIS_DIR}/add-config
+
+# All tests are using defconfig
+BUILD_TYPE = defconfig
+
+# The test names will have the arch and cross compiler used. This will be shown in
+# the results.
+TEST_NAME = ${ARCH} ${CROSS}
+
+# alpha
+TEST_START IF ${RUN} == alpha || ${DO_DEFAULT}
+# Notice that CROSS and ARCH are also options and not variables (again '=' instead
+# of ':='). This is because TEST_NAME and MAKE_CMD wil use them for each test.
+# Only options are available during runs. Variables are only present in parsing the
+# config file.
+CROSS = alpha-linux
+ARCH = alpha
+
+# arm
+TEST_START IF ${RUN} == arm || ${DO_DEFAULT}
+CROSS = arm-unknown-linux-gnueabi
+ARCH = arm
+
+# black fin
+TEST_START IF ${RUN} == bfin || ${DO_DEFAULT}
+CROSS = bfin-uclinux
+ARCH = blackfin
+BUILD_OPTIONS = -j8 vmlinux
+
+# cris - FAILS?
+TEST_START IF ${RUN} == cris || ${RUN} == cris64 || ${DO_FAILED}
+CROSS = cris-linux
+ARCH = cris
+
+# cris32 - not right arch?
+TEST_START IF ${RUN} == cris || ${RUN} == cris32 || ${DO_FAILED}
+CROSS = crisv32-linux
+ARCH = cris
+
+# ia64
+TEST_START IF ${RUN} == ia64 || ${DO_DEFAULT}
+CROSS = ia64-linux
+ARCH = ia64
+
+# frv
+TEST_START IF ${RUN} == frv || ${DO_FAILED}
+CROSS = frv-linux
+ARCH = frv
+GCC_VER = 4.5.1
+
+# h8300 - failed make defconfig??
+TEST_START IF ${RUN} == h8300 || ${DO_FAILED}
+CROSS = h8300-elf
+ARCH = h8300
+GCC_VER = 4.5.1
+
+# m68k fails with error?
+TEST_START IF ${RUN} == m68k || ${DO_DEFAULT}
+CROSS = m68k-linux
+ARCH = m68k
+
+# mips64
+TEST_START IF ${RUN} == mips || ${RUN} == mips64 || ${DO_DEFAULT}
+CROSS = mips64-linux
+ARCH = mips
+
+# mips32
+TEST_START IF ${RUN} == mips || ${RUN} == mips32 || ${DO_DEFAULT}
+CROSS = mips-linux
+ARCH = mips
+
+# m32r
+TEST_START IF ${RUN} == m32r || ${DO_FAILED}
+CROSS = m32r-linux
+ARCH = m32r
+GCC_VER = 4.5.1
+BUILD_OPTIONS = -j8 vmlinux
+
+# parisc64 failed?
+TEST_START IF ${RUN} == hppa || ${RUN} == hppa64 || ${DO_FAILED}
+CROSS = hppa64-linux
+ARCH = parisc
+
+# parisc
+TEST_START IF ${RUN} == hppa || ${RUN} == hppa32 || ${DO_FAILED}
+CROSS = hppa-linux
+ARCH = parisc
+
+# ppc
+TEST_START IF ${RUN} == ppc || ${RUN} == ppc32 || ${DO_DEFAULT}
+CROSS = powerpc-linux
+ARCH = powerpc
+
+# ppc64
+TEST_START IF ${RUN} == ppc || ${RUN} == ppc64 || ${DO_DEFAULT}
+CROSS = powerpc64-linux
+ARCH = powerpc
+
+# s390
+TEST_START IF ${RUN} == s390 || ${DO_DEFAULT}
+CROSS = s390x-linux
+ARCH = s390
+
+# sh
+TEST_START IF ${RUN} == sh || ${DO_DEFAULT}
+CROSS = sh4-linux
+ARCH = sh
+
+# sparc64
+TEST_START IF ${RUN} == sparc || ${RUN} == sparc64 || ${DO_DEFAULT}
+CROSS = sparc64-linux
+ARCH = sparc64
+
+# sparc
+TEST_START IF ${RUN} == sparc || ${RUN} == sparc32 || ${DO_DEFAULT}
+CROSS = sparc-linux
+ARCH = sparc
+
+# xtensa failed
+TEST_START IF ${RUN} == xtensa || ${DO_FAILED}
+CROSS = xtensa-linux
+ARCH = xtensa
+
+# UML
+TEST_START IF ${RUN} == uml || ${DO_DEFAULT}
+MAKE_CMD = make ARCH=um SUBARCH=x86_64
+ARCH = uml
+CROSS =
+
+TEST_START IF ${RUN} == x86 || ${RUN} == i386 || ${DO_DEFAULT}
+MAKE_CMD = make ARCH=i386
+ARCH = i386
+CROSS = 
+
+TEST_START IF ${RUN} == x86 || ${RUN} == x86_64 || ${DO_DEFAULT}
+MAKE_CMD = make ARCH=x86_64
+ARCH = x86_64
+CROSS = 
+
+#################################
+
+# This is a bisect if needed. You need to give it a MIN_CONFIG that
+# will be the config file it uses. Basically, just copy the created defconfig
+# for the arch someplace and point MIN_CONFIG to it.
+TEST_START IF ${RUN} == bisect
+MIN_CONFIG = ${THIS_DIR}/min-config
+CROSS = s390x-linux
+ARCH = s390
+TEST_TYPE = bisect
+BISECT_TYPE = build
+BISECT_GOOD = v3.1
+BISECT_BAD = v3.2
+CHECKOUT = v3.2
+
+#################################
+
+# These defaults are needed to keep ktest.pl from complaining. They are
+# ignored because the test does not go pass the build. No install or
+# booting of the target images.
+
+DEFAULTS
+MACHINE = crosstest
+SSH_USER = root
+BUILD_TARGET = cross
+TARGET_IMAGE = image
+POWER_CYCLE = cycle
+CONSOLE = console
+LOCALVERSION = version
+GRUB_MENU = grub
+
+REBOOT_ON_ERROR = 0
+POWEROFF_ON_ERROR = 0
+POWEROFF_ON_SUCCESS = 0
+REBOOT_ON_SUCCESS = 0
+

tools/testing/ktest/examples/include/bisect.conf

@@ -0,0 +1,90 @@
+#
+# This example shows the bisect tests (git bisect and config bisect)
+#
+
+
+# The config that includes this file may define a RUN_TEST
+# variable that will tell this config what test to run.
+# (what to set the TEST option to).
+#
+DEFAULTS IF NOT DEFINED RUN_TEST
+# Requires that hackbench is in the PATH
+RUN_TEST := ${SSH} hackbench 50
+
+
+# Set TEST to 'bisect' to do a normal git bisect. You need
+# to modify the options below to make it bisect the exact
+# commits you are interested in.
+#
+TEST_START IF ${TEST} == bisect
+TEST_TYPE = bisect
+# You must set the commit that was considered good (git bisect good)
+BISECT_GOOD = v3.3
+# You must set the commit that was considered bad (git bisect bad)
+BISECT_BAD = HEAD
+# It's best to specify the branch to checkout before starting the bisect.
+CHECKOUT = origin/master
+# This can be build, boot, or test. Here we are doing a bisect
+# that requires to run a test to know if the bisect was good or bad.
+# The test should exit with 0 on good, non-zero for bad. But see
+# the BISECT_RET_* options in samples.conf to override this.
+BISECT_TYPE = test
+TEST = ${RUN_TEST}
+# It is usually a good idea to confirm that the GOOD and the BAD
+# commits are truly good and bad respectively. Having BISECT_CHECK
+# set to 1 will check both that the good commit works and the bad
+# commit fails. If you only want to check one or the other,
+# set BISECT_CHECK to 'good' or to 'bad'.
+BISECT_CHECK = 1
+#BISECT_CHECK = good
+#BISECT_CHECK = bad
+
+# Usually it's a good idea to specify the exact config you
+# want to use throughout the entire bisect. Here we placed
+# it in the directory we called ktest.pl from and named it
+# 'config-bisect'.
+MIN_CONFIG = ${THIS_DIR}/config-bisect
+# By default, if we are doing a BISECT_TYPE = test run but the
+# build or boot fails, ktest.pl will do a 'git bisect skip'.
+# Uncomment the below option to make ktest stop testing on such
+# an error.
+#BISECT_SKIP = 0
+# Now if you had BISECT_SKIP = 0 and the test fails, you can
+# examine what happened and then do 'git bisect log > /tmp/replay'
+# Set BISECT_REPLAY to /tmp/replay and ktest.pl will run the
+# 'git bisect replay /tmp/replay' before continuing the bisect test.
+#BISECT_REPLAY = /tmp/replay
+# If you used BISECT_REPLAY after the bisect test failed, you may
+# not want to continue the bisect on that commit that failed.
+# By setting BISECT_START to a new commit. ktest.pl will checkout
+# that commit after it has performed the 'git bisect replay' but
+# before it continues running the bisect test.
+#BISECT_START = 2545eb6198e7e1ec50daa0cfc64a4cdfecf24ec9
+
+# Now if you don't trust ktest.pl to make the decisions for you, then
+# set BISECT_MANUAL to 1. This will cause ktest.pl not to decide
+# if the commit was good or bad. Instead, it will ask you to tell
+# it if the current commit was good. In the mean time, you could
+# take the result, load it on any machine you want. Run several tests,
+# or whatever you feel like. Then, when you are happy, you can tell
+# ktest if you think it was good or not and ktest.pl will continue
+# the git bisect. You can even change what commit it is currently at.
+#BISECT_MANUAL = 1
+
+
+# One of the unique tests that ktest does is the config bisect.
+# Currently (which hopefully will be fixed soon), the bad config
+# must be a superset of the good config. This is because it only
+# searches for a config that causes the target to fail. If the
+# good config is not a subset of the bad config, or if the target
+# fails because of a lack of a config, then it will not find
+# the config for you.
+TEST_START IF ${TEST} == config-bisect
+TEST_TYPE = config_bisect
+# set to build, boot, test
+CONFIG_BISECT_TYPE = boot
+# Set the config that is considered bad.
+CONFIG_BISECT = ${THIS_DIR}/config-bad
+# This config is optional. By default it uses the
+# MIN_CONFIG as the good config.
+CONFIG_BISECT_GOOD = ${THIS_DIR}/config-good

tools/testing/ktest/examples/include/defaults.conf

@@ -0,0 +1,157 @@
+# This file holds defaults for most the tests. It defines the options that
+# are most common to tests that are likely to be shared.
+#
+# Note, after including this file, a config file may override any option
+# with a DEFAULTS OVERRIDE section.
+#
+
+# For those cases that use the same machine to boot a 64 bit
+# and a 32 bit version. The MACHINE is the DNS name to get to the
+# box (usually different if it was 64 bit or 32 bit) but the
+# BOX here is defined as a variable that will be the name of the box
+# itself. It is useful for calling scripts that will power cycle
+# the box, as only one script needs to be created to power cycle
+# even though the box itself has multiple operating systems on it.
+# By default, BOX and MACHINE are the same.
+
+DEFAULTS IF NOT DEFINED BOX
+BOX := ${MACHINE}
+
+
+# Consider each box as 64 bit box, unless the config including this file
+# has defined BITS = 32
+
+DEFAULTS IF NOT DEFINED BITS
+BITS := 64
+
+
+DEFAULTS
+
+# THIS_DIR is used through out the configs and defaults to ${PWD} which
+# is the directory that ktest.pl was called from.
+
+THIS_DIR := ${PWD}
+
+
+# to organize your configs, having each machine save their configs
+# into a separate directly is useful.
+CONFIG_DIR := ${THIS_DIR}/configs/${MACHINE}
+
+# Reset the log before running each test.
+CLEAR_LOG = 1
+
+# As installing kernels usually requires root privilege, default the
+# user on the target as root. It is also required that the target
+# allows ssh to root from the host without asking for a password.
+
+SSH_USER = root
+
+# For accesing the machine, we will ssh to root@machine.
+SSH := ssh ${SSH_USER}@${MACHINE}
+
+# Update this. The default here is ktest will ssh to the target box
+# and run a script called 'run-test' located on that box.
+TEST = ${SSH} run-test
+
+# Point build dir to the git repo you use
+BUILD_DIR = ${THIS_DIR}/linux.git
+
+# Each machine will have its own output build directory.
+OUTPUT_DIR = ${THIS_DIR}/build/${MACHINE}
+
+# Yes this config is focused on x86 (but ktest works for other archs too)
+BUILD_TARGET = arch/x86/boot/bzImage
+TARGET_IMAGE = /boot/vmlinuz-test
+
+# have directory for the scripts to reboot and power cycle the boxes
+SCRIPTS_DIR := ${THIS_DIR}/scripts
+
+# You can have each box/machine have a script to power cycle it.
+# Name your script <box>-cycle.
+POWER_CYCLE = ${SCRIPTS_DIR}/${BOX}-cycle
+
+# This script is used to power off the box.
+POWER_OFF = ${SCRIPTS_DIR}/${BOX}-poweroff
+
+# Keep your test kernels separate from your other kernels.
+LOCALVERSION = -test
+
+# The /boot/grub/menu.lst is searched for the line:
+#  title Test Kernel
+# and ktest will use that kernel to reboot into.
+# For grub2 or other boot loaders, you need to set BOOT_TYPE
+# to 'script' and define other ways to load the kernel.
+# See snowball.conf example.
+#
+GRUB_MENU = Test Kernel
+
+# The kernel build will use this option.
+BUILD_OPTIONS = -j8
+
+# Keeping the log file with the output dir is convenient.
+LOG_FILE = ${OUTPUT_DIR}/${MACHINE}.log
+
+# Each box should have their own minum configuration
+# See min-config.conf
+MIN_CONFIG = ${CONFIG_DIR}/config-min
+
+# For things like randconfigs, there may be configs you find that
+# are already broken, or there may be some configs that you always
+# want set. Uncomment ADD_CONFIG and point it to the make config files
+# that set the configs you want to keep on (or off) in your build.
+# ADD_CONFIG is usually something to add configs to all machines,
+# where as, MIN_CONFIG is specific per machine.
+#ADD_CONFIG = ${THIS_DIR}/config-broken ${THIS_DIR}/config-general
+
+# To speed up reboots for bisects and patchcheck, instead of
+# waiting 60 seconds for the console to be idle, if this line is
+# seen in the console output, ktest will know the good kernel has
+# finished rebooting and it will be able to continue the tests.
+REBOOT_SUCCESS_LINE = ${MACHINE} login:
+
+# The following is different ways to end the test.
+# by setting the variable REBOOT to: none, error, fail or
+# something else, ktest will power cycle or reboot the target box
+# at the end of the tests.
+#
+# REBOOT := none
+#   Don't do anything at the end of the test.
+#
+# REBOOT := error
+#   Reboot the box if ktest detects an error
+#
+# REBOOT := fail
+#   Do not stop on failure, and after all tests are complete
+#   power off the box (for both success and error)
+#   This is good to run over a weekend and you don't want to waste
+#   electricity.
+#
+
+DEFAULTS IF ${REBOOT} == none
+REBOOT_ON_SUCCESS = 0
+REBOOT_ON_ERROR = 0
+POWEROFF_ON_ERROR = 0
+POWEROFF_ON_SUCCESS = 0
+
+DEFAULTS ELSE IF ${REBOOT} == error
+REBOOT_ON_SUCCESS = 0
+REBOOT_ON_ERROR = 1
+POWEROFF_ON_ERROR = 0
+POWEROFF_ON_SUCCESS = 0
+
+DEFAULTS ELSE IF ${REBOOT} == fail
+REBOOT_ON_SUCCESS = 0
+POWEROFF_ON_ERROR = 1
+POWEROFF_ON_SUCCESS = 1
+POWEROFF_AFTER_HALT = 120
+DIE_ON_FAILURE = 0
+
+# Store the failure information into this directory
+# such as the .config, dmesg, and build log.
+STORE_FAILURES = ${THIS_DIR}/failures
+
+DEFAULTS ELSE
+REBOOT_ON_SUCCESS = 1
+REBOOT_ON_ERROR = 1
+POWEROFF_ON_ERROR = 0
+POWEROFF_ON_SUCCESS = 0

tools/testing/ktest/examples/include/min-config.conf

@@ -0,0 +1,60 @@
+#
+# This file has some examples for creating a MIN_CONFIG.
+# (A .config file that is the minimum for a machine to boot, or
+#  to boot and make a network connection.)
+#
+# A MIN_CONFIG is very useful as it is the minimum configuration
+# needed to boot a given machine. You can debug someone else's
+# .config by only setting the configs in your MIN_CONFIG. The closer
+# your MIN_CONFIG is to the true minimum set of configs needed to
+# boot your machine, the closer the config you test with will be
+# to the users config that had the failure.
+#
+# The make_min_config test allows you to create a MIN_CONFIG that
+# is truly the minimum set of configs needed to boot a box.
+#
+# In this example, the final config will reside in
+# ${CONFIG_DIR}/config-new-min and ${CONFIG_DIR}/config-new-min-net.
+# Just move one to the location you have set for MIN_CONFIG.
+#
+# The first test creates a MIN_CONFIG that will be the minimum
+# configuration to boot ${MACHINE} and be able to ssh to it.
+#
+# The second test creates a MIN_CONFIG that will only boot
+# the target and most likely will not let you ssh to it. (Notice
+# how the second test uses the first test's result to continue with.
+# This is because the second test config is a subset of the first).
+#
+# The ${CONFIG_DIR}/config-skip (and -net) will hold the configs
+# that ktest.pl found would not boot the target without them set.
+# The config-new-min holds configs that ktest.pl could not test
+# directly because another config that was needed to boot the box
+# selected them. Sometimes it is possible that this file will hold
+# the true minimum configuration. You can test to see if this is
+# the case by running the boot test with BOOT_TYPE = allnoconfig and
+# setting setting the MIN_CONFIG to ${CONFIG_DIR}/config-skip. If the
+# machine still boots, then you can use the config-skip as your MIN_CONFIG.
+#
+# These tests can run for several hours (and perhaps days).
+# It's OK to kill the test with a Ctrl^C. By restarting without
+# modifying this config, ktest.pl will notice that the config-new-min(-net)
+# exists, and will use that instead as the starting point.
+# The USE_OUTPUT_MIN_CONFIG is set to 1 to keep ktest.pl from asking
+# you if you want to use the OUTPUT_MIN_CONFIG as the starting point.
+# By using the OUTPUT_MIN_CONFIG as the starting point will allow ktest.pl to
+# start almost where it left off.
+#
+TEST_START IF ${TEST} == min-config
+TEST_TYPE = make_min_config
+OUTPUT_MIN_CONFIG = ${CONFIG_DIR}/config-new-min-net
+IGNORE_CONFIG = ${CONFIG_DIR}/config-skip-net
+MIN_CONFIG_TYPE = test
+TEST = ${SSH} echo hi
+USE_OUTPUT_MIN_CONFIG = 1
+
+TEST_START IF ${TEST} == min-config && ${MULTI}
+TEST_TYPE = make_min_config
+OUTPUT_MIN_CONFIG = ${CONFIG_DIR}/config-new-min
+IGNORE_CONFIG = ${CONFIG_DIR}/config-skip
+MIN_CONFIG = ${CONFIG_DIR}/config-new-min-net
+USE_OUTPUT_MIN_CONFIG = 1

tools/testing/ktest/examples/include/patchcheck.conf

@@ -0,0 +1,111 @@
+# patchcheck.conf
+#
+# This contains a test that takes two git commits and will test each
+# commit between the two. The build test will look at what files the
+# commit has touched, and if any of those files produce a warning, then
+# the build will fail.
+
+
+# PATCH_START is the commit to begin with and PATCH_END is the commit
+# to end with (inclusive). This is similar to doing a git rebase -i PATCH_START~1
+# and then testing each commit and doing a git rebase --continue.
+# You can use a SHA1, a git tag, or anything that git will accept for a checkout
+
+PATCH_START := HEAD~3
+PATCH_END := HEAD
+
+# Use the oldconfig if build_type wasn't defined
+DEFAULTS IF NOT DEFINED BUILD_TYPE
+DO_BUILD_TYPE := oldconfig
+
+DEFAULTS ELSE
+DO_BUILD_TYPE := ${BUILD_TYPE}
+
+DEFAULTS
+
+
+# Change PATCH_CHECKOUT to be the branch you want to test. The test will
+# do a git checkout of this branch before starting. Obviously both
+# PATCH_START and PATCH_END must be in this branch (and PATCH_START must
+# be contained by PATCH_END).
+
+PATCH_CHECKOUT := test/branch
+
+# Usually it's a good idea to have a set config to use for testing individual
+# patches.
+PATCH_CONFIG := ${CONFIG_DIR}/config-patchcheck
+
+# Change PATCH_TEST to run some test for each patch. Each commit that is
+# tested, after it is built and installed on the test machine, this command
+# will be executed. Usually what is done is to ssh to the target box and
+# run some test scripts. If you just want to boot test your patches
+# comment PATCH_TEST out.
+PATCH_TEST := ${SSH} "/usr/local/bin/ktest-test-script"
+
+DEFAULTS IF DEFINED PATCH_TEST
+PATCH_TEST_TYPE := test
+
+DEFAULTS ELSE
+PATCH_TEST_TYPE := boot
+
+# If for some reason a file has a warning that one of your patches touch
+# but you do not care about it, set IGNORE_WARNINGS to that commit(s)
+# (space delimited)
+#IGNORE_WARNINGS = 39eaf7ef884dcc44f7ff1bac803ca2a1dcf43544 6edb2a8a385f0cdef51dae37ff23e74d76d8a6ce
+
+# Instead of just checking for warnings to files that are changed
+# it can be advantageous to check for any new warnings. If a
+# header file is changed, it could cause a warning in a file not
+# touched by the commit. To detect these kinds of warnings, you
+# can use the WARNINGS_FILE option.
+#
+# If the variable CREATE_WARNINGS_FILE is set, this config will
+# enable the WARNINGS_FILE during the patchcheck test. Also,
+# before running the patchcheck test, it will create the
+# warnings file.
+#
+DEFAULTS IF DEFINED CREATE_WARNINGS_FILE
+WARNINGS_FILE = ${OUTPUT_DIR}/warnings_file
+
+TEST_START IF DEFINED CREATE_WARNINGS_FILE
+# WARNINGS_FILE is already set by the DEFAULTS above
+TEST_TYPE = make_warnings_file
+# Checkout the commit before the patches to test,
+# and record all the warnings that exist before the patches
+# to test are added
+CHECKOUT = ${PATCHCHECK_START}~1
+# Force a full build
+BUILD_NOCLEAN = 0
+BUILD_TYPE = ${DO_BUILD_TYPE}
+
+# If you are running a multi test, and the test failed on the first
+# test but on, say the 5th patch. If you want to restart on the
+# fifth patch, set PATCH_START1. This will make the first test start
+# from this commit instead of the PATCH_START commit.
+# Note, do not change this option. Just define PATCH_START1 in the
+# top config (the one you pass to ktest.pl), and this will use it,
+# otherwise it will just use PATCH_START if PATCH_START1 is not defined.
+DEFAULTS IF NOT DEFINED PATCH_START1
+PATCH_START1 := ${PATCH_START}
+
+TEST_START IF ${TEST} == patchcheck
+TEST_TYPE = patchcheck
+MIN_CONFIG = ${PATCH_CONFIG}
+TEST = ${PATCH_TEST}
+PATCHCHECK_TYPE = ${PATCH_TEST_TYPE}
+PATCHCHECK_START = ${PATCH_START1}
+PATCHCHECK_END = ${PATCH_END}
+CHECKOUT = ${PATCH_CHECKOUT}
+BUILD_TYPE = ${DO_BUILD_TYPE}
+
+TEST_START IF ${TEST} == patchcheck && ${MULTI}
+TEST_TYPE = patchcheck
+MIN_CONFIG = ${PATCH_CONFIG}
+TEST = ${PATCH_TEST}
+PATCHCHECK_TYPE = ${PATCH_TEST_TYPE}
+PATCHCHECK_START = ${PATCH_START}
+PATCHCHECK_END = ${PATCH_END}
+CHECKOUT = ${PATCH_CHECKOUT}
+# Use multi to test different compilers?
+MAKE_CMD = CC=gcc-4.5.1 make
+BUILD_TYPE = ${DO_BUILD_TYPE}

tools/testing/ktest/examples/include/tests.conf

@@ -0,0 +1,74 @@
+#
+# This is an example of various tests that you can run
+#
+# The variable TEST can be of boot, build, randconfig, or test.
+#
+# Note that TEST is a variable created with ':=' and only exists
+# throughout the config processing (not during the tests itself).
+#
+# The TEST option (defined with '=') is used to tell ktest.pl
+# what test to run after a successful boot. The TEST option is
+# persistent into the test runs.
+#
+
+# The config that includes this file may define a BOOT_TYPE
+# variable that tells this config what type of boot test to run.
+# If it's not defined, the below DEFAULTS will set the default
+# to 'oldconfig'.
+#
+DEFAULTS IF NOT DEFINED BOOT_TYPE
+BOOT_TYPE := oldconfig
+
+# The config that includes this file may define a RUN_TEST
+# variable that will tell this config what test to run.
+# (what to set the TEST option to).
+#
+DEFAULTS IF NOT DEFINED RUN_TEST
+# Requires that hackbench is in the PATH
+RUN_TEST := ${SSH} hackbench 50
+
+
+# If TEST is set to 'boot' then just build a kernel and boot
+# the target.
+TEST_START IF ${TEST} == boot
+TEST_TYPE = boot
+# Notice how we set the BUILD_TYPE option to the BOOT_TYPE variable.
+BUILD_TYPE = ${BOOT_TYPE}
+# Do not do a make mrproper.
+BUILD_NOCLEAN = 1
+
+# If you only want to build the kernel, and perhaps install
+# and test it yourself, then just set TEST to build.
+TEST_START IF ${TEST} == build
+TEST_TYPE = build
+BUILD_TYPE = ${BOOT_TYPE}
+BUILD_NOCLEAN = 1
+
+# Build, install, boot and test with a randconfg 10 times.
+# It is important that you have set MIN_CONFIG in the config
+# that includes this file otherwise it is likely that the
+# randconfig will not have the necessary configs needed to
+# boot your box. This version of the test requires a min
+# config that has enough to make sure the target has network
+# working.
+TEST_START ITERATE 10 IF ${TEST} == randconfig
+MIN_CONFIG = ${CONFIG_DIR}/config-min-net
+TEST_TYPE = test
+BUILD_TYPE = randconfig
+TEST = ${RUN_TEST}
+
+# This is the same as above, but only tests to a boot prompt.
+# The MIN_CONFIG used here does not need to have networking
+# working.
+TEST_START ITERATE 10 IF ${TEST} == randconfig && ${MULTI}
+TEST_TYPE = boot
+BUILD_TYPE = randconfig
+MIN_CONFIG = ${CONFIG_DIR}/config-min
+MAKE_CMD = make
+
+# This builds, installs, boots and tests the target.
+TEST_START IF ${TEST} == test
+TEST_TYPE = test
+BUILD_TYPE = ${BOOT_TYPE}
+TEST = ${RUN_TEST}
+BUILD_NOCLEAN = 1

tools/testing/ktest/examples/kvm.conf

@@ -0,0 +1,88 @@
+#
+# This config is an example usage of ktest.pl with a kvm guest
+#
+# The guest is called 'Guest' and this would be something that
+# could be run on the host to test a virtual machine target.
+
+MACHINE = Guest
+
+
+# Use virsh to read the serial console of the guest
+CONSOLE =  virsh console ${MACHINE}
+
+#*************************************#
+# This part is the same as test.conf  #
+#*************************************#
+
+# The include files will set up the type of test to run. Just set TEST to
+# which test you want to run.
+#
+# TESTS = patchcheck, randconfig, boot, test, config-bisect, bisect, min-config
+#
+# See the include/*.conf files that define these tests
+#
+TEST := patchcheck
+
+# Some tests may have more than one test to run. Define MULTI := 1 to run
+# the extra tests.
+MULTI := 0
+
+# In case you want to differentiate which type of system you are testing
+BITS := 64
+
+# REBOOT = none, error, fail, empty
+#  See include/defaults.conf
+REBOOT := empty
+
+
+# The defaults file will set up various settings that can be used by all
+# machine configs.
+INCLUDE include/defaults.conf
+
+
+#*************************************#
+# Now we are different from test.conf #
+#*************************************#
+
+
+# The example here assumes that Guest is running a Fedora release
+# that uses dracut for its initfs. The POST_INSTALL will be executed
+# after the install of the kernel and modules are complete.
+#
+POST_INSTALL = ${SSH} /sbin/dracut -f /boot/initramfs-test.img $KERNEL_VERSION
+
+# Guests sometimes get stuck on reboot. We wait 3 seconds after running
+# the reboot command and then  do a full power-cycle of the guest.
+# This forces the guest to restart.
+#
+POWERCYCLE_AFTER_REBOOT = 3
+
+# We do the same after the halt command, but this time we wait 20 seconds.
+POWEROFF_AFTER_HALT = 20
+
+
+# As the defaults.conf file has a POWER_CYCLE option already defined,
+# and options can not be defined in the same section more than once
+# (all DEFAULTS sections are considered the same). We use the
+# DEFAULTS OVERRIDE to tell ktest.pl to ignore the previous defined
+# options, for the options set in the OVERRIDE section.
+#
+DEFAULTS OVERRIDE
+
+# Instead of using the default POWER_CYCLE option defined in
+# defaults.conf, we use virsh to cycle it. To do so, we destroy
+# the guest, wait 5 seconds, and then start it up again.
+# Crude, but effective.
+#
+POWER_CYCLE = virsh destroy ${MACHINE}; sleep 5; virsh start ${MACHINE}
+
+
+DEFAULTS
+
+# The following files each handle a different test case.
+# Having them included allows you to set up more than one machine and share
+# the same tests.
+INCLUDE include/patchcheck.conf
+INCLUDE include/tests.conf
+INCLUDE include/bisect.conf
+INCLUDE include/min-config.conf

tools/testing/ktest/examples/snowball.conf

@@ -0,0 +1,53 @@
+# This example was used to boot the snowball ARM board.
+# See http://people.redhat.com/srostedt/ktest-embedded-2012/
+
+# PWD is a ktest.pl variable that will result in the process working
+# directory that ktest.pl is executed in.
+
+# THIS_DIR is automatically assigned the PWD of the path that generated
+# the config file. It is best to use this variable when assigning other
+# directory paths within this directory. This allows you to easily
+# move the test cases to other locations or to other machines.
+#
+THIS_DIR := /home/rostedt/work/demo/ktest-embed
+LOG_FILE = ${OUTPUT_DIR}/snowball.log
+CLEAR_LOG = 1
+MAKE_CMD = PATH=/usr/local/gcc-4.5.2-nolibc/arm-unknown-linux-gnueabi/bin:$PATH CROSS_COMPILE=arm-unknown-linux-gnueabi- make ARCH=arm
+ADD_CONFIG = ${THIS_DIR}/addconfig
+
+SCP_TO_TARGET = echo "don't do scp"
+
+TFTPBOOT := /var/lib/tftpboot
+TFTPDEF := ${TFTPBOOT}/snowball-default
+TFTPTEST := ${OUTPUT_DIR}/${BUILD_TARGET}
+
+SWITCH_TO_GOOD = cp ${TFTPDEF} ${TARGET_IMAGE}
+SWITCH_TO_TEST = cp ${TFTPTEST} ${TARGET_IMAGE}
+
+# Define each test with TEST_START
+# The config options below it will override the defaults
+TEST_START SKIP
+TEST_TYPE = boot
+BUILD_TYPE = u8500_defconfig
+BUILD_NOCLEAN = 1
+
+TEST_START
+TEST_TYPE = make_min_config
+OUTPUT_MIN_CONFIG = ${THIS_DIR}/config.newmin
+START_MIN_CONFIG = ${THIS_DIR}/config.orig
+IGNORE_CONFIG = ${THIS_DIR}/config.ignore
+BUILD_NOCLEAN = 1
+
+
+DEFAULTS
+LOCALVERSION = -test
+POWER_CYCLE = echo use the thumb luke; read a
+CONSOLE = cat ${THIS_DIR}/snowball-cat
+REBOOT_TYPE = script
+SSH_USER = root
+BUILD_OPTIONS = -j8 uImage
+BUILD_DIR = ${THIS_DIR}/linux.git
+OUTPUT_DIR = ${THIS_DIR}/snowball-build
+MACHINE = snowball
+TARGET_IMAGE = /var/lib/tftpboot/snowball-image
+BUILD_TARGET = arch/arm/boot/uImage

tools/testing/ktest/examples/test.conf

@@ -0,0 +1,62 @@
+#
+# Generic config for a machine
+#
+
+# Name your machine (the DNS name, what you ssh to)
+MACHINE = foo
+
+# BOX can be different than foo, if the machine BOX has
+# multiple partitions with different systems installed. For example,
+# you may have a i386 and x86_64 installation on a test box.
+# If this is the case, MACHINE defines the way to connect to the
+# machine, which may be different between which system the machine
+# is booting into. BOX is used for the scripts to reboot and power cycle
+# the machine, where it does not matter which system the machine boots into.
+#
+#BOX := bar
+
+# Define a way to read the console
+CONSOLE = stty -F /dev/ttyS0 115200 parodd; cat /dev/ttyS0
+
+# The include files will set up the type of test to run. Just set TEST to
+# which test you want to run.
+#
+# TESTS = patchcheck, randconfig, boot, test, config-bisect, bisect, min-config
+#
+# See the include/*.conf files that define these tests
+#
+TEST := patchcheck
+
+# Some tests may have more than one test to run. Define MULTI := 1 to run
+# the extra tests.
+MULTI := 0
+
+# In case you want to differentiate which type of system you are testing
+BITS := 64
+
+# REBOOT = none, error, fail, empty
+#  See include/defaults.conf
+REBOOT := empty
+
+# The defaults file will set up various settings that can be used by all
+# machine configs.
+INCLUDE include/defaults.conf
+
+# In case you need to add a patch for a bisect or something
+#PRE_BUILD = patch -p1 < ${THIS_DIR}/fix.patch
+
+# Reset the repo after the build and remove all 'test' modules from the target
+# Notice that DO_POST_BUILD is a variable (defined by ':=') and POST_BUILD
+# is the option (defined by '=')
+
+DO_POST_BUILD := git reset --hard
+POST_BUILD = ${SSH} 'rm -rf /lib/modules/*-test*'; ${DO_POST_BUILD}
+
+# The following files each handle a different test case.
+# Having them included allows you to set up more than one machine and share
+# the same tests.
+INCLUDE include/patchcheck.conf
+INCLUDE include/tests.conf
+INCLUDE include/bisect.conf
+INCLUDE include/min-config.conf
+

tools/testing/selftests/Makefile

@@ -0,0 +1,24 @@
+TARGETS = breakpoints
+TARGETS += cpu-hotplug
+TARGETS += efivarfs
+TARGETS += kcmp
+TARGETS += memory-hotplug
+TARGETS += mqueue
+TARGETS += net
+TARGETS += ptrace
+TARGETS += vm
+
+all:
+	for TARGET in $(TARGETS); do \
+		make -C $$TARGET; \
+	done;
+
+run_tests: all
+	for TARGET in $(TARGETS); do \
+		make -C $$TARGET run_tests; \
+	done;
+
+clean:
+	for TARGET in $(TARGETS); do \
+		make -C $$TARGET clean; \
+	done;

tools/testing/selftests/README.txt

@@ -0,0 +1,42 @@
+Linux Kernel Selftests
+
+The kernel contains a set of "self tests" under the tools/testing/selftests/
+directory. These are intended to be small unit tests to exercise individual
+code paths in the kernel.
+
+Running the selftests
+=====================
+
+To build the tests:
+
+  $ make -C tools/testing/selftests
+
+
+To run the tests:
+
+  $ make -C tools/testing/selftests run_tests
+
+- note that some tests will require root privileges.
+
+
+To run only tests targetted for a single subsystem:
+
+  $  make -C tools/testing/selftests TARGETS=cpu-hotplug run_tests
+
+See the top-level tools/testing/selftests/Makefile for the list of all possible
+targets.
+
+
+Contributing new tests
+======================
+
+In general, the rules for for selftests are
+
+ * Do as much as you can if you're not root;
+
+ * Don't take too long;
+
+ * Don't break the build on any architecture, and
+
+ * Don't cause the top-level "make run_tests" to fail if your feature is
+   unconfigured.

tools/testing/selftests/breakpoints/Makefile

@@ -0,0 +1,23 @@
+# Taken from perf makefile
+uname_M := $(shell uname -m 2>/dev/null || echo not)
+ARCH ?= $(shell echo $(uname_M) | sed -e s/i.86/i386/)
+ifeq ($(ARCH),i386)
+        ARCH := x86
+endif
+ifeq ($(ARCH),x86_64)
+	ARCH := x86
+endif
+
+
+all:
+ifeq ($(ARCH),x86)
+	gcc breakpoint_test.c -o breakpoint_test
+else
+	echo "Not an x86 target, can't build breakpoints selftests"
+endif
+
+run_tests:
+	@./breakpoint_test || echo "breakpoints selftests: [FAIL]"
+
+clean:
+	rm -fr breakpoint_test

tools/testing/selftests/breakpoints/breakpoint_test.c

@@ -0,0 +1,394 @@
+/*
+ * Copyright (C) 2011 Red Hat, Inc., Frederic Weisbecker <fweisbec@redhat.com>
+ *
+ * Licensed under the terms of the GNU GPL License version 2
+ *
+ * Selftests for breakpoints (and more generally the do_debug() path) in x86.
+ */
+
+
+#include <sys/ptrace.h>
+#include <unistd.h>
+#include <stddef.h>
+#include <sys/user.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <signal.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+
+
+/* Breakpoint access modes */
+enum {
+	BP_X = 1,
+	BP_RW = 2,
+	BP_W = 4,
+};
+
+static pid_t child_pid;
+
+/*
+ * Ensures the child and parent are always "talking" about
+ * the same test sequence. (ie: that we haven't forgotten
+ * to call check_trapped() somewhere).
+ */
+static int nr_tests;
+
+static void set_breakpoint_addr(void *addr, int n)
+{
+	int ret;
+
+	ret = ptrace(PTRACE_POKEUSER, child_pid,
+		     offsetof(struct user, u_debugreg[n]), addr);
+	if (ret) {
+		perror("Can't set breakpoint addr\n");
+		exit(-1);
+	}
+}
+
+static void toggle_breakpoint(int n, int type, int len,
+			      int local, int global, int set)
+{
+	int ret;
+
+	int xtype, xlen;
+	unsigned long vdr7, dr7;
+
+	switch (type) {
+	case BP_X:
+		xtype = 0;
+		break;
+	case BP_W:
+		xtype = 1;
+		break;
+	case BP_RW:
+		xtype = 3;
+		break;
+	}
+
+	switch (len) {
+	case 1:
+		xlen = 0;
+		break;
+	case 2:
+		xlen = 4;
+		break;
+	case 4:
+		xlen = 0xc;
+		break;
+	case 8:
+		xlen = 8;
+		break;
+	}
+
+	dr7 = ptrace(PTRACE_PEEKUSER, child_pid,
+		     offsetof(struct user, u_debugreg[7]), 0);
+
+	vdr7 = (xlen | xtype) << 16;
+	vdr7 <<= 4 * n;
+
+	if (local) {
+		vdr7 |= 1 << (2 * n);
+		vdr7 |= 1 << 8;
+	}
+	if (global) {
+		vdr7 |= 2 << (2 * n);
+		vdr7 |= 1 << 9;
+	}
+
+	if (set)
+		dr7 |= vdr7;
+	else
+		dr7 &= ~vdr7;
+
+	ret = ptrace(PTRACE_POKEUSER, child_pid,
+		     offsetof(struct user, u_debugreg[7]), dr7);
+	if (ret) {
+		perror("Can't set dr7");
+		exit(-1);
+	}
+}
+
+/* Dummy variables to test read/write accesses */
+static unsigned long long dummy_var[4];
+
+/* Dummy functions to test execution accesses */
+static void dummy_func(void) { }
+static void dummy_func1(void) { }
+static void dummy_func2(void) { }
+static void dummy_func3(void) { }
+
+static void (*dummy_funcs[])(void) = {
+	dummy_func,
+	dummy_func1,
+	dummy_func2,
+	dummy_func3,
+};
+
+static int trapped;
+
+static void check_trapped(void)
+{
+	/*
+	 * If we haven't trapped, wake up the parent
+	 * so that it notices the failure.
+	 */
+	if (!trapped)
+		kill(getpid(), SIGUSR1);
+	trapped = 0;
+
+	nr_tests++;
+}
+
+static void write_var(int len)
+{
+	char *pcval; short *psval; int *pival; long long *plval;
+	int i;
+
+	for (i = 0; i < 4; i++) {
+		switch (len) {
+		case 1:
+			pcval = (char *)&dummy_var[i];
+			*pcval = 0xff;
+			break;
+		case 2:
+			psval = (short *)&dummy_var[i];
+			*psval = 0xffff;
+			break;
+		case 4:
+			pival = (int *)&dummy_var[i];
+			*pival = 0xffffffff;
+			break;
+		case 8:
+			plval = (long long *)&dummy_var[i];
+			*plval = 0xffffffffffffffffLL;
+			break;
+		}
+		check_trapped();
+	}
+}
+
+static void read_var(int len)
+{
+	char cval; short sval; int ival; long long lval;
+	int i;
+
+	for (i = 0; i < 4; i++) {
+		switch (len) {
+		case 1:
+			cval = *(char *)&dummy_var[i];
+			break;
+		case 2:
+			sval = *(short *)&dummy_var[i];
+			break;
+		case 4:
+			ival = *(int *)&dummy_var[i];
+			break;
+		case 8:
+			lval = *(long long *)&dummy_var[i];
+			break;
+		}
+		check_trapped();
+	}
+}
+
+/*
+ * Do the r/w/x accesses to trigger the breakpoints. And run
+ * the usual traps.
+ */
+static void trigger_tests(void)
+{
+	int len, local, global, i;
+	char val;
+	int ret;
+
+	ret = ptrace(PTRACE_TRACEME, 0, NULL, 0);
+	if (ret) {
+		perror("Can't be traced?\n");
+		return;
+	}
+
+	/* Wake up father so that it sets up the first test */
+	kill(getpid(), SIGUSR1);
+
+	/* Test instruction breakpoints */
+	for (local = 0; local < 2; local++) {
+		for (global = 0; global < 2; global++) {
+			if (!local && !global)
+				continue;
+
+			for (i = 0; i < 4; i++) {
+				dummy_funcs[i]();
+				check_trapped();
+			}
+		}
+	}
+
+	/* Test write watchpoints */
+	for (len = 1; len <= sizeof(long); len <<= 1) {
+		for (local = 0; local < 2; local++) {
+			for (global = 0; global < 2; global++) {
+				if (!local && !global)
+					continue;
+				write_var(len);
+			}
+		}
+	}
+
+	/* Test read/write watchpoints (on read accesses) */
+	for (len = 1; len <= sizeof(long); len <<= 1) {
+		for (local = 0; local < 2; local++) {
+			for (global = 0; global < 2; global++) {
+				if (!local && !global)
+					continue;
+				read_var(len);
+			}
+		}
+	}
+
+	/* Icebp trap */
+	asm(".byte 0xf1\n");
+	check_trapped();
+
+	/* Int 3 trap */
+	asm("int $3\n");
+	check_trapped();
+
+	kill(getpid(), SIGUSR1);
+}
+
+static void check_success(const char *msg)
+{
+	const char *msg2;
+	int child_nr_tests;
+	int status;
+
+	/* Wait for the child to SIGTRAP */
+	wait(&status);
+
+	msg2 = "Failed";
+
+	if (WSTOPSIG(status) == SIGTRAP) {
+		child_nr_tests = ptrace(PTRACE_PEEKDATA, child_pid,
+					&nr_tests, 0);
+		if (child_nr_tests == nr_tests)
+			msg2 = "Ok";
+		if (ptrace(PTRACE_POKEDATA, child_pid, &trapped, 1)) {
+			perror("Can't poke\n");
+			exit(-1);
+		}
+	}
+
+	nr_tests++;
+
+	printf("%s [%s]\n", msg, msg2);
+}
+
+static void launch_instruction_breakpoints(char *buf, int local, int global)
+{
+	int i;
+
+	for (i = 0; i < 4; i++) {
+		set_breakpoint_addr(dummy_funcs[i], i);
+		toggle_breakpoint(i, BP_X, 1, local, global, 1);
+		ptrace(PTRACE_CONT, child_pid, NULL, 0);
+		sprintf(buf, "Test breakpoint %d with local: %d global: %d",
+			i, local, global);
+		check_success(buf);
+		toggle_breakpoint(i, BP_X, 1, local, global, 0);
+	}
+}
+
+static void launch_watchpoints(char *buf, int mode, int len,
+			       int local, int global)
+{
+	const char *mode_str;
+	int i;
+
+	if (mode == BP_W)
+		mode_str = "write";
+	else
+		mode_str = "read";
+
+	for (i = 0; i < 4; i++) {
+		set_breakpoint_addr(&dummy_var[i], i);
+		toggle_breakpoint(i, mode, len, local, global, 1);
+		ptrace(PTRACE_CONT, child_pid, NULL, 0);
+		sprintf(buf, "Test %s watchpoint %d with len: %d local: "
+			"%d global: %d", mode_str, i, len, local, global);
+		check_success(buf);
+		toggle_breakpoint(i, mode, len, local, global, 0);
+	}
+}
+
+/* Set the breakpoints and check the child successfully trigger them */
+static void launch_tests(void)
+{
+	char buf[1024];
+	int len, local, global, i;
+
+	/* Instruction breakpoints */
+	for (local = 0; local < 2; local++) {
+		for (global = 0; global < 2; global++) {
+			if (!local && !global)
+				continue;
+			launch_instruction_breakpoints(buf, local, global);
+		}
+	}
+
+	/* Write watchpoint */
+	for (len = 1; len <= sizeof(long); len <<= 1) {
+		for (local = 0; local < 2; local++) {
+			for (global = 0; global < 2; global++) {
+				if (!local && !global)
+					continue;
+				launch_watchpoints(buf, BP_W, len,
+						   local, global);
+			}
+		}
+	}
+
+	/* Read-Write watchpoint */
+	for (len = 1; len <= sizeof(long); len <<= 1) {
+		for (local = 0; local < 2; local++) {
+			for (global = 0; global < 2; global++) {
+				if (!local && !global)
+					continue;
+				launch_watchpoints(buf, BP_RW, len,
+						   local, global);
+			}
+		}
+	}
+
+	/* Icebp traps */
+	ptrace(PTRACE_CONT, child_pid, NULL, 0);
+	check_success("Test icebp");
+
+	/* Int 3 traps */
+	ptrace(PTRACE_CONT, child_pid, NULL, 0);
+	check_success("Test int 3 trap");
+
+	ptrace(PTRACE_CONT, child_pid, NULL, 0);
+}
+
+int main(int argc, char **argv)
+{
+	pid_t pid;
+	int ret;
+
+	pid = fork();
+	if (!pid) {
+		trigger_tests();
+		return 0;
+	}
+
+	child_pid = pid;
+
+	wait(NULL);
+
+	launch_tests();
+
+	wait(NULL);
+
+	return 0;
+}

tools/testing/selftests/cpu-hotplug/Makefile

@@ -0,0 +1,6 @@
+all:
+
+run_tests:
+	@./on-off-test.sh || echo "cpu-hotplug selftests: [FAIL]"
+
+clean:

tools/testing/selftests/cpu-hotplug/on-off-test.sh

@@ -0,0 +1,221 @@
+#!/bin/bash
+
+SYSFS=
+
+prerequisite()
+{
+	msg="skip all tests:"
+
+	if [ $UID != 0 ]; then
+		echo $msg must be run as root >&2
+		exit 0
+	fi
+
+	SYSFS=`mount -t sysfs | head -1 | awk '{ print $3 }'`
+
+	if [ ! -d "$SYSFS" ]; then
+		echo $msg sysfs is not mounted >&2
+		exit 0
+	fi
+
+	if ! ls $SYSFS/devices/system/cpu/cpu* > /dev/null 2>&1; then
+		echo $msg cpu hotplug is not supported >&2
+		exit 0
+	fi
+}
+
+#
+# list all hot-pluggable CPUs
+#
+hotpluggable_cpus()
+{
+	local state=${1:-.\*}
+
+	for cpu in $SYSFS/devices/system/cpu/cpu*; do
+		if [ -f $cpu/online ] && grep -q $state $cpu/online; then
+			echo ${cpu##/*/cpu}
+		fi
+	done
+}
+
+hotplaggable_offline_cpus()
+{
+	hotpluggable_cpus 0
+}
+
+hotpluggable_online_cpus()
+{
+	hotpluggable_cpus 1
+}
+
+cpu_is_online()
+{
+	grep -q 1 $SYSFS/devices/system/cpu/cpu$1/online
+}
+
+cpu_is_offline()
+{
+	grep -q 0 $SYSFS/devices/system/cpu/cpu$1/online
+}
+
+online_cpu()
+{
+	echo 1 > $SYSFS/devices/system/cpu/cpu$1/online
+}
+
+offline_cpu()
+{
+	echo 0 > $SYSFS/devices/system/cpu/cpu$1/online
+}
+
+online_cpu_expect_success()
+{
+	local cpu=$1
+
+	if ! online_cpu $cpu; then
+		echo $FUNCNAME $cpu: unexpected fail >&2
+	elif ! cpu_is_online $cpu; then
+		echo $FUNCNAME $cpu: unexpected offline >&2
+	fi
+}
+
+online_cpu_expect_fail()
+{
+	local cpu=$1
+
+	if online_cpu $cpu 2> /dev/null; then
+		echo $FUNCNAME $cpu: unexpected success >&2
+	elif ! cpu_is_offline $cpu; then
+		echo $FUNCNAME $cpu: unexpected online >&2
+	fi
+}
+
+offline_cpu_expect_success()
+{
+	local cpu=$1
+
+	if ! offline_cpu $cpu; then
+		echo $FUNCNAME $cpu: unexpected fail >&2
+	elif ! cpu_is_offline $cpu; then
+		echo $FUNCNAME $cpu: unexpected offline >&2
+	fi
+}
+
+offline_cpu_expect_fail()
+{
+	local cpu=$1
+
+	if offline_cpu $cpu 2> /dev/null; then
+		echo $FUNCNAME $cpu: unexpected success >&2
+	elif ! cpu_is_online $cpu; then
+		echo $FUNCNAME $cpu: unexpected offline >&2
+	fi
+}
+
+error=-12
+priority=0
+
+while getopts e:hp: opt; do
+	case $opt in
+	e)
+		error=$OPTARG
+		;;
+	h)
+		echo "Usage $0 [ -e errno ] [ -p notifier-priority ]"
+		exit
+		;;
+	p)
+		priority=$OPTARG
+		;;
+	esac
+done
+
+if ! [ "$error" -ge -4095 -a "$error" -lt 0 ]; then
+	echo "error code must be -4095 <= errno < 0" >&2
+	exit 1
+fi
+
+prerequisite
+
+#
+# Online all hot-pluggable CPUs
+#
+for cpu in `hotplaggable_offline_cpus`; do
+	online_cpu_expect_success $cpu
+done
+
+#
+# Offline all hot-pluggable CPUs
+#
+for cpu in `hotpluggable_online_cpus`; do
+	offline_cpu_expect_success $cpu
+done
+
+#
+# Online all hot-pluggable CPUs again
+#
+for cpu in `hotplaggable_offline_cpus`; do
+	online_cpu_expect_success $cpu
+done
+
+#
+# Test with cpu notifier error injection
+#
+
+DEBUGFS=`mount -t debugfs | head -1 | awk '{ print $3 }'`
+NOTIFIER_ERR_INJECT_DIR=$DEBUGFS/notifier-error-inject/cpu
+
+prerequisite_extra()
+{
+	msg="skip extra tests:"
+
+	/sbin/modprobe -q -r cpu-notifier-error-inject
+	/sbin/modprobe -q cpu-notifier-error-inject priority=$priority
+
+	if [ ! -d "$DEBUGFS" ]; then
+		echo $msg debugfs is not mounted >&2
+		exit 0
+	fi
+
+	if [ ! -d $NOTIFIER_ERR_INJECT_DIR ]; then
+		echo $msg cpu-notifier-error-inject module is not available >&2
+		exit 0
+	fi
+}
+
+prerequisite_extra
+
+#
+# Offline all hot-pluggable CPUs
+#
+echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_DOWN_PREPARE/error
+for cpu in `hotpluggable_online_cpus`; do
+	offline_cpu_expect_success $cpu
+done
+
+#
+# Test CPU hot-add error handling (offline => online)
+#
+echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_UP_PREPARE/error
+for cpu in `hotplaggable_offline_cpus`; do
+	online_cpu_expect_fail $cpu
+done
+
+#
+# Online all hot-pluggable CPUs
+#
+echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_UP_PREPARE/error
+for cpu in `hotplaggable_offline_cpus`; do
+	online_cpu_expect_success $cpu
+done
+
+#
+# Test CPU hot-remove error handling (online => offline)
+#
+echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_DOWN_PREPARE/error
+for cpu in `hotpluggable_online_cpus`; do
+	offline_cpu_expect_fail $cpu
+done
+
+echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_DOWN_PREPARE/error
+/sbin/modprobe -q -r cpu-notifier-error-inject

tools/testing/selftests/efivarfs/Makefile

@@ -0,0 +1,12 @@
+CC = $(CROSS_COMPILE)gcc
+CFLAGS = -Wall
+
+test_objs = open-unlink create-read
+
+all: $(test_objs)
+
+run_tests: all
+	@/bin/bash ./efivarfs.sh || echo "efivarfs selftests: [FAIL]"
+
+clean:
+	rm -f $(test_objs)

tools/testing/selftests/efivarfs/create-read.c

@@ -0,0 +1,38 @@
+#include <stdio.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <string.h>
+
+int main(int argc, char **argv)
+{
+	const char *path;
+	char buf[4];
+	int fd, rc;
+
+	if (argc < 2) {
+		fprintf(stderr, "usage: %s <path>\n", argv[0]);
+		return EXIT_FAILURE;
+	}
+
+	path = argv[1];
+
+	/* create a test variable */
+	fd = open(path, O_RDWR | O_CREAT, 0600);
+	if (fd < 0) {
+		perror("open(O_WRONLY)");
+		return EXIT_FAILURE;
+	}
+
+	rc = read(fd, buf, sizeof(buf));
+	if (rc != 0) {
+		fprintf(stderr, "Reading a new var should return EOF\n");
+		return EXIT_FAILURE;
+	}
+
+	return EXIT_SUCCESS;
+}

tools/testing/selftests/efivarfs/efivarfs.sh

@@ -0,0 +1,198 @@
+#!/bin/bash
+
+efivarfs_mount=/sys/firmware/efi/efivars
+test_guid=210be57c-9849-4fc7-a635-e6382d1aec27
+
+check_prereqs()
+{
+	local msg="skip all tests:"
+
+	if [ $UID != 0 ]; then
+		echo $msg must be run as root >&2
+		exit 0
+	fi
+
+	if ! grep -q "^\S\+ $efivarfs_mount efivarfs" /proc/mounts; then
+		echo $msg efivarfs is not mounted on $efivarfs_mount >&2
+		exit 0
+	fi
+}
+
+run_test()
+{
+	local test="$1"
+
+	echo "--------------------"
+	echo "running $test"
+	echo "--------------------"
+
+	if [ "$(type -t $test)" = 'function' ]; then
+		( $test )
+	else
+		( ./$test )
+	fi
+
+	if [ $? -ne 0 ]; then
+		echo "  [FAIL]"
+		rc=1
+	else
+		echo "  [PASS]"
+	fi
+}
+
+test_create()
+{
+	local attrs='\x07\x00\x00\x00'
+	local file=$efivarfs_mount/$FUNCNAME-$test_guid
+
+	printf "$attrs\x00" > $file
+
+	if [ ! -e $file ]; then
+		echo "$file couldn't be created" >&2
+		exit 1
+	fi
+
+	if [ $(stat -c %s $file) -ne 5 ]; then
+		echo "$file has invalid size" >&2
+		exit 1
+	fi
+}
+
+test_create_empty()
+{
+	local file=$efivarfs_mount/$FUNCNAME-$test_guid
+
+	: > $file
+
+	if [ ! -e $file ]; then
+		echo "$file can not be created without writing" >&2
+		exit 1
+	fi
+}
+
+test_create_read()
+{
+	local file=$efivarfs_mount/$FUNCNAME-$test_guid
+	./create-read $file
+}
+
+test_delete()
+{
+	local attrs='\x07\x00\x00\x00'
+	local file=$efivarfs_mount/$FUNCNAME-$test_guid
+
+	printf "$attrs\x00" > $file
+
+	if [ ! -e $file ]; then
+		echo "$file couldn't be created" >&2
+		exit 1
+	fi
+
+	rm $file
+
+	if [ -e $file ]; then
+		echo "$file couldn't be deleted" >&2
+		exit 1
+	fi
+
+}
+
+# test that we can remove a variable by issuing a write with only
+# attributes specified
+test_zero_size_delete()
+{
+	local attrs='\x07\x00\x00\x00'
+	local file=$efivarfs_mount/$FUNCNAME-$test_guid
+
+	printf "$attrs\x00" > $file
+
+	if [ ! -e $file ]; then
+		echo "$file does not exist" >&2
+		exit 1
+	fi
+
+	printf "$attrs" > $file
+
+	if [ -e $file ]; then
+		echo "$file should have been deleted" >&2
+		exit 1
+	fi
+}
+
+test_open_unlink()
+{
+	local file=$efivarfs_mount/$FUNCNAME-$test_guid
+	./open-unlink $file
+}
+
+# test that we can create a range of filenames
+test_valid_filenames()
+{
+	local attrs='\x07\x00\x00\x00'
+	local ret=0
+
+	local file_list="abc dump-type0-11-1-1362436005 1234 -"
+	for f in $file_list; do
+		local file=$efivarfs_mount/$f-$test_guid
+
+		printf "$attrs\x00" > $file
+
+		if [ ! -e $file ]; then
+			echo "$file could not be created" >&2
+			ret=1
+		else
+			rm $file
+		fi
+	done
+
+	exit $ret
+}
+
+test_invalid_filenames()
+{
+	local attrs='\x07\x00\x00\x00'
+	local ret=0
+
+	local file_list="
+		-1234-1234-1234-123456789abc
+		foo
+		foo-bar
+		-foo-
+		foo-barbazba-foob-foob-foob-foobarbazfoo
+		foo-------------------------------------
+		-12345678-1234-1234-1234-123456789abc
+		a-12345678=1234-1234-1234-123456789abc
+		a-12345678-1234=1234-1234-123456789abc
+		a-12345678-1234-1234=1234-123456789abc
+		a-12345678-1234-1234-1234=123456789abc
+		1112345678-1234-1234-1234-123456789abc"
+
+	for f in $file_list; do
+		local file=$efivarfs_mount/$f
+
+		printf "$attrs\x00" 2>/dev/null > $file
+
+		if [ -e $file ]; then
+			echo "Creating $file should have failed" >&2
+			rm $file
+			ret=1
+		fi
+	done
+
+	exit $ret
+}
+
+check_prereqs
+
+rc=0
+
+run_test test_create
+run_test test_create_empty
+run_test test_create_read
+run_test test_delete
+run_test test_zero_size_delete
+run_test test_open_unlink
+run_test test_valid_filenames
+run_test test_invalid_filenames
+
+exit $rc

tools/testing/selftests/efivarfs/open-unlink.c

@@ -0,0 +1,63 @@
+#include <stdio.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+
+int main(int argc, char **argv)
+{
+	const char *path;
+	char buf[5];
+	int fd, rc;
+
+	if (argc < 2) {
+		fprintf(stderr, "usage: %s <path>\n", argv[0]);
+		return EXIT_FAILURE;
+	}
+
+	path = argv[1];
+
+	/* attributes: EFI_VARIABLE_NON_VOLATILE |
+	 *		EFI_VARIABLE_BOOTSERVICE_ACCESS |
+	 *		EFI_VARIABLE_RUNTIME_ACCESS
+	 */
+	*(uint32_t *)buf = 0x7;
+	buf[4] = 0;
+
+	/* create a test variable */
+	fd = open(path, O_WRONLY | O_CREAT);
+	if (fd < 0) {
+		perror("open(O_WRONLY)");
+		return EXIT_FAILURE;
+	}
+
+	rc = write(fd, buf, sizeof(buf));
+	if (rc != sizeof(buf)) {
+		perror("write");
+		return EXIT_FAILURE;
+	}
+
+	close(fd);
+
+	fd = open(path, O_RDONLY);
+	if (fd < 0) {
+		perror("open");
+		return EXIT_FAILURE;
+	}
+
+	if (unlink(path) < 0) {
+		perror("unlink");
+		return EXIT_FAILURE;
+	}
+
+	rc = read(fd, buf, sizeof(buf));
+	if (rc > 0) {
+		fprintf(stderr, "reading from an unlinked variable "
+				"shouldn't be possible\n");
+		return EXIT_FAILURE;
+	}
+
+	return EXIT_SUCCESS;
+}

tools/testing/selftests/ipc/Makefile

@@ -0,0 +1,25 @@
+uname_M := $(shell uname -m 2>/dev/null || echo not)
+ARCH ?= $(shell echo $(uname_M) | sed -e s/i.86/i386/)
+ifeq ($(ARCH),i386)
+        ARCH := X86
+	CFLAGS := -DCONFIG_X86_32 -D__i386__
+endif
+ifeq ($(ARCH),x86_64)
+	ARCH := X86
+	CFLAGS := -DCONFIG_X86_64 -D__x86_64__
+endif
+
+CFLAGS += -I../../../../usr/include/
+
+all:
+ifeq ($(ARCH),X86)
+	gcc $(CFLAGS) msgque.c -o msgque_test
+else
+	echo "Not an x86 target, can't build msgque selftest"
+endif
+
+run_tests: all
+	./msgque_test
+
+clean:
+	rm -fr ./msgque_test

tools/testing/selftests/ipc/msgque.c

@@ -0,0 +1,246 @@
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <errno.h>
+#include <linux/msg.h>
+#include <fcntl.h>
+
+#define MAX_MSG_SIZE		32
+
+struct msg1 {
+	int msize;
+	long mtype;
+	char mtext[MAX_MSG_SIZE];
+};
+
+#define TEST_STRING "Test sysv5 msg"
+#define MSG_TYPE 1
+
+#define ANOTHER_TEST_STRING "Yet another test sysv5 msg"
+#define ANOTHER_MSG_TYPE 26538
+
+struct msgque_data {
+	key_t key;
+	int msq_id;
+	int qbytes;
+	int qnum;
+	int mode;
+	struct msg1 *messages;
+};
+
+int restore_queue(struct msgque_data *msgque)
+{
+	int fd, ret, id, i;
+	char buf[32];
+
+	fd = open("/proc/sys/kernel/msg_next_id", O_WRONLY);
+	if (fd == -1) {
+		printf("Failed to open /proc/sys/kernel/msg_next_id\n");
+		return -errno;
+	}
+	sprintf(buf, "%d", msgque->msq_id);
+
+	ret = write(fd, buf, strlen(buf));
+	if (ret != strlen(buf)) {
+		printf("Failed to write to /proc/sys/kernel/msg_next_id\n");
+		return -errno;
+	}
+
+	id = msgget(msgque->key, msgque->mode | IPC_CREAT | IPC_EXCL);
+	if (id == -1) {
+		printf("Failed to create queue\n");
+		return -errno;
+	}
+
+	if (id != msgque->msq_id) {
+		printf("Restored queue has wrong id (%d instead of %d)\n",
+							id, msgque->msq_id);
+		ret = -EFAULT;
+		goto destroy;
+	}
+
+	for (i = 0; i < msgque->qnum; i++) {
+		if (msgsnd(msgque->msq_id, &msgque->messages[i].mtype,
+			   msgque->messages[i].msize, IPC_NOWAIT) != 0) {
+			printf("msgsnd failed (%m)\n");
+			ret = -errno;
+			goto destroy;
+		};
+	}
+	return 0;
+
+destroy:
+	if (msgctl(id, IPC_RMID, 0))
+		printf("Failed to destroy queue: %d\n", -errno);
+	return ret;
+}
+
+int check_and_destroy_queue(struct msgque_data *msgque)
+{
+	struct msg1 message;
+	int cnt = 0, ret;
+
+	while (1) {
+		ret = msgrcv(msgque->msq_id, &message.mtype, MAX_MSG_SIZE,
+				0, IPC_NOWAIT);
+		if (ret < 0) {
+			if (errno == ENOMSG)
+				break;
+			printf("Failed to read IPC message: %m\n");
+			ret = -errno;
+			goto err;
+		}
+		if (ret != msgque->messages[cnt].msize) {
+			printf("Wrong message size: %d (expected %d)\n", ret,
+						msgque->messages[cnt].msize);
+			ret = -EINVAL;
+			goto err;
+		}
+		if (message.mtype != msgque->messages[cnt].mtype) {
+			printf("Wrong message type\n");
+			ret = -EINVAL;
+			goto err;
+		}
+		if (memcmp(message.mtext, msgque->messages[cnt].mtext, ret)) {
+			printf("Wrong message content\n");
+			ret = -EINVAL;
+			goto err;
+		}
+		cnt++;
+	}
+
+	if (cnt != msgque->qnum) {
+		printf("Wrong message number\n");
+		ret = -EINVAL;
+		goto err;
+	}
+
+	ret = 0;
+err:
+	if (msgctl(msgque->msq_id, IPC_RMID, 0)) {
+		printf("Failed to destroy queue: %d\n", -errno);
+		return -errno;
+	}
+	return ret;
+}
+
+int dump_queue(struct msgque_data *msgque)
+{
+	struct msqid64_ds ds;
+	int kern_id;
+	int i, ret;
+
+	for (kern_id = 0; kern_id < 256; kern_id++) {
+		ret = msgctl(kern_id, MSG_STAT, &ds);
+		if (ret < 0) {
+			if (errno == -EINVAL)
+				continue;
+			printf("Failed to get stats for IPC queue with id %d\n",
+					kern_id);
+			return -errno;
+		}
+
+		if (ret == msgque->msq_id)
+			break;
+	}
+
+	msgque->messages = malloc(sizeof(struct msg1) * ds.msg_qnum);
+	if (msgque->messages == NULL) {
+		printf("Failed to get stats for IPC queue\n");
+		return -ENOMEM;
+	}
+
+	msgque->qnum = ds.msg_qnum;
+	msgque->mode = ds.msg_perm.mode;
+	msgque->qbytes = ds.msg_qbytes;
+
+	for (i = 0; i < msgque->qnum; i++) {
+		ret = msgrcv(msgque->msq_id, &msgque->messages[i].mtype,
+				MAX_MSG_SIZE, i, IPC_NOWAIT | MSG_COPY);
+		if (ret < 0) {
+			printf("Failed to copy IPC message: %m (%d)\n", errno);
+			return -errno;
+		}
+		msgque->messages[i].msize = ret;
+	}
+	return 0;
+}
+
+int fill_msgque(struct msgque_data *msgque)
+{
+	struct msg1 msgbuf;
+
+	msgbuf.mtype = MSG_TYPE;
+	memcpy(msgbuf.mtext, TEST_STRING, sizeof(TEST_STRING));
+	if (msgsnd(msgque->msq_id, &msgbuf.mtype, sizeof(TEST_STRING),
+				IPC_NOWAIT) != 0) {
+		printf("First message send failed (%m)\n");
+		return -errno;
+	};
+
+	msgbuf.mtype = ANOTHER_MSG_TYPE;
+	memcpy(msgbuf.mtext, ANOTHER_TEST_STRING, sizeof(ANOTHER_TEST_STRING));
+	if (msgsnd(msgque->msq_id, &msgbuf.mtype, sizeof(ANOTHER_TEST_STRING),
+				IPC_NOWAIT) != 0) {
+		printf("Second message send failed (%m)\n");
+		return -errno;
+	};
+	return 0;
+}
+
+int main(int argc, char **argv)
+{
+	int msg, pid, err;
+	struct msgque_data msgque;
+
+	msgque.key = ftok(argv[0], 822155650);
+	if (msgque.key == -1) {
+		printf("Can't make key\n");
+		return -errno;
+	}
+
+	msgque.msq_id = msgget(msgque.key, IPC_CREAT | IPC_EXCL | 0666);
+	if (msgque.msq_id == -1) {
+		printf("Can't create queue\n");
+		goto err_out;
+	}
+
+	err = fill_msgque(&msgque);
+	if (err) {
+		printf("Failed to fill queue\n");
+		goto err_destroy;
+	}
+
+	err = dump_queue(&msgque);
+	if (err) {
+		printf("Failed to dump queue\n");
+		goto err_destroy;
+	}
+
+	err = check_and_destroy_queue(&msgque);
+	if (err) {
+		printf("Failed to check and destroy queue\n");
+		goto err_out;
+	}
+
+	err = restore_queue(&msgque);
+	if (err) {
+		printf("Failed to restore queue\n");
+		goto err_destroy;
+	}
+
+	err = check_and_destroy_queue(&msgque);
+	if (err) {
+		printf("Failed to test queue\n");
+		goto err_out;
+	}
+	return 0;
+
+err_destroy:
+	if (msgctl(msgque.msq_id, IPC_RMID, 0)) {
+		printf("Failed to destroy queue: %d\n", -errno);
+		return -errno;
+	}
+err_out:
+	return err;
+}

tools/testing/selftests/kcmp/Makefile

@@ -0,0 +1,29 @@
+uname_M := $(shell uname -m 2>/dev/null || echo not)
+ARCH ?= $(shell echo $(uname_M) | sed -e s/i.86/i386/)
+ifeq ($(ARCH),i386)
+        ARCH := X86
+	CFLAGS := -DCONFIG_X86_32 -D__i386__
+endif
+ifeq ($(ARCH),x86_64)
+	ARCH := X86
+	CFLAGS := -DCONFIG_X86_64 -D__x86_64__
+endif
+
+CFLAGS += -I../../../../arch/x86/include/generated/
+CFLAGS += -I../../../../include/
+CFLAGS += -I../../../../usr/include/
+CFLAGS += -I../../../../arch/x86/include/
+
+all:
+ifeq ($(ARCH),X86)
+	gcc $(CFLAGS) kcmp_test.c -o kcmp_test
+else
+	echo "Not an x86 target, can't build kcmp selftest"
+endif
+
+run_tests: all
+	@./kcmp_test || echo "kcmp_test: [FAIL]"
+
+clean:
+	rm -fr ./run_test
+	rm -fr ./test-file

tools/testing/selftests/kcmp/kcmp_test.c

@@ -0,0 +1,96 @@
+#define _GNU_SOURCE
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <signal.h>
+#include <limits.h>
+#include <unistd.h>
+#include <errno.h>
+#include <string.h>
+#include <fcntl.h>
+
+#include <linux/unistd.h>
+#include <linux/kcmp.h>
+
+#include <sys/syscall.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+
+static long sys_kcmp(int pid1, int pid2, int type, int fd1, int fd2)
+{
+	return syscall(__NR_kcmp, pid1, pid2, type, fd1, fd2);
+}
+
+int main(int argc, char **argv)
+{
+	const char kpath[] = "kcmp-test-file";
+	int pid1, pid2;
+	int fd1, fd2;
+	int status;
+
+	fd1 = open(kpath, O_RDWR | O_CREAT | O_TRUNC, 0644);
+	pid1 = getpid();
+
+	if (fd1 < 0) {
+		perror("Can't create file");
+		exit(1);
+	}
+
+	pid2 = fork();
+	if (pid2 < 0) {
+		perror("fork failed");
+		exit(1);
+	}
+
+	if (!pid2) {
+		int pid2 = getpid();
+		int ret;
+
+		fd2 = open(kpath, O_RDWR, 0644);
+		if (fd2 < 0) {
+			perror("Can't open file");
+			exit(1);
+		}
+
+		/* An example of output and arguments */
+		printf("pid1: %6d pid2: %6d FD: %2ld FILES: %2ld VM: %2ld "
+		       "FS: %2ld SIGHAND: %2ld IO: %2ld SYSVSEM: %2ld "
+		       "INV: %2ld\n",
+		       pid1, pid2,
+		       sys_kcmp(pid1, pid2, KCMP_FILE,		fd1, fd2),
+		       sys_kcmp(pid1, pid2, KCMP_FILES,		0, 0),
+		       sys_kcmp(pid1, pid2, KCMP_VM,		0, 0),
+		       sys_kcmp(pid1, pid2, KCMP_FS,		0, 0),
+		       sys_kcmp(pid1, pid2, KCMP_SIGHAND,	0, 0),
+		       sys_kcmp(pid1, pid2, KCMP_IO,		0, 0),
+		       sys_kcmp(pid1, pid2, KCMP_SYSVSEM,	0, 0),
+
+			/* This one should fail */
+		       sys_kcmp(pid1, pid2, KCMP_TYPES + 1,	0, 0));
+
+		/* This one should return same fd */
+		ret = sys_kcmp(pid1, pid2, KCMP_FILE, fd1, fd1);
+		if (ret) {
+			printf("FAIL: 0 expected but %d returned (%s)\n",
+				ret, strerror(errno));
+			ret = -1;
+		} else
+			printf("PASS: 0 returned as expected\n");
+
+		/* Compare with self */
+		ret = sys_kcmp(pid1, pid1, KCMP_VM, 0, 0);
+		if (ret) {
+			printf("FAIL: 0 expected but %li returned (%s)\n",
+				ret, strerror(errno));
+			ret = -1;
+		} else
+			printf("PASS: 0 returned as expected\n");
+
+		exit(ret);
+	}
+
+	waitpid(pid2, &status, P_ALL);
+
+	return 0;
+}

tools/testing/selftests/memory-hotplug/Makefile

@@ -0,0 +1,6 @@
+all:
+
+run_tests:
+	@./on-off-test.sh || echo "memory-hotplug selftests: [FAIL]"
+
+clean:

tools/testing/selftests/memory-hotplug/on-off-test.sh

@@ -0,0 +1,230 @@
+#!/bin/bash
+
+SYSFS=
+
+prerequisite()
+{
+	msg="skip all tests:"
+
+	if [ $UID != 0 ]; then
+		echo $msg must be run as root >&2
+		exit 0
+	fi
+
+	SYSFS=`mount -t sysfs | head -1 | awk '{ print $3 }'`
+
+	if [ ! -d "$SYSFS" ]; then
+		echo $msg sysfs is not mounted >&2
+		exit 0
+	fi
+
+	if ! ls $SYSFS/devices/system/memory/memory* > /dev/null 2>&1; then
+		echo $msg memory hotplug is not supported >&2
+		exit 0
+	fi
+}
+
+#
+# list all hot-pluggable memory
+#
+hotpluggable_memory()
+{
+	local state=${1:-.\*}
+
+	for memory in $SYSFS/devices/system/memory/memory*; do
+		if grep -q 1 $memory/removable &&
+		   grep -q $state $memory/state; then
+			echo ${memory##/*/memory}
+		fi
+	done
+}
+
+hotplaggable_offline_memory()
+{
+	hotpluggable_memory offline
+}
+
+hotpluggable_online_memory()
+{
+	hotpluggable_memory online
+}
+
+memory_is_online()
+{
+	grep -q online $SYSFS/devices/system/memory/memory$1/state
+}
+
+memory_is_offline()
+{
+	grep -q offline $SYSFS/devices/system/memory/memory$1/state
+}
+
+online_memory()
+{
+	echo online > $SYSFS/devices/system/memory/memory$1/state
+}
+
+offline_memory()
+{
+	echo offline > $SYSFS/devices/system/memory/memory$1/state
+}
+
+online_memory_expect_success()
+{
+	local memory=$1
+
+	if ! online_memory $memory; then
+		echo $FUNCNAME $memory: unexpected fail >&2
+	elif ! memory_is_online $memory; then
+		echo $FUNCNAME $memory: unexpected offline >&2
+	fi
+}
+
+online_memory_expect_fail()
+{
+	local memory=$1
+
+	if online_memory $memory 2> /dev/null; then
+		echo $FUNCNAME $memory: unexpected success >&2
+	elif ! memory_is_offline $memory; then
+		echo $FUNCNAME $memory: unexpected online >&2
+	fi
+}
+
+offline_memory_expect_success()
+{
+	local memory=$1
+
+	if ! offline_memory $memory; then
+		echo $FUNCNAME $memory: unexpected fail >&2
+	elif ! memory_is_offline $memory; then
+		echo $FUNCNAME $memory: unexpected offline >&2
+	fi
+}
+
+offline_memory_expect_fail()
+{
+	local memory=$1
+
+	if offline_memory $memory 2> /dev/null; then
+		echo $FUNCNAME $memory: unexpected success >&2
+	elif ! memory_is_online $memory; then
+		echo $FUNCNAME $memory: unexpected offline >&2
+	fi
+}
+
+error=-12
+priority=0
+ratio=10
+
+while getopts e:hp:r: opt; do
+	case $opt in
+	e)
+		error=$OPTARG
+		;;
+	h)
+		echo "Usage $0 [ -e errno ] [ -p notifier-priority ] [ -r percent-of-memory-to-offline ]"
+		exit
+		;;
+	p)
+		priority=$OPTARG
+		;;
+	r)
+		ratio=$OPTARG
+		;;
+	esac
+done
+
+if ! [ "$error" -ge -4095 -a "$error" -lt 0 ]; then
+	echo "error code must be -4095 <= errno < 0" >&2
+	exit 1
+fi
+
+prerequisite
+
+#
+# Online all hot-pluggable memory
+#
+for memory in `hotplaggable_offline_memory`; do
+	online_memory_expect_success $memory
+done
+
+#
+# Offline $ratio percent of hot-pluggable memory
+#
+for memory in `hotpluggable_online_memory`; do
+	if [ $((RANDOM % 100)) -lt $ratio ]; then
+		offline_memory_expect_success $memory
+	fi
+done
+
+#
+# Online all hot-pluggable memory again
+#
+for memory in `hotplaggable_offline_memory`; do
+	online_memory_expect_success $memory
+done
+
+#
+# Test with memory notifier error injection
+#
+
+DEBUGFS=`mount -t debugfs | head -1 | awk '{ print $3 }'`
+NOTIFIER_ERR_INJECT_DIR=$DEBUGFS/notifier-error-inject/memory
+
+prerequisite_extra()
+{
+	msg="skip extra tests:"
+
+	/sbin/modprobe -q -r memory-notifier-error-inject
+	/sbin/modprobe -q memory-notifier-error-inject priority=$priority
+
+	if [ ! -d "$DEBUGFS" ]; then
+		echo $msg debugfs is not mounted >&2
+		exit 0
+	fi
+
+	if [ ! -d $NOTIFIER_ERR_INJECT_DIR ]; then
+		echo $msg memory-notifier-error-inject module is not available >&2
+		exit 0
+	fi
+}
+
+prerequisite_extra
+
+#
+# Offline $ratio percent of hot-pluggable memory
+#
+echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_OFFLINE/error
+for memory in `hotpluggable_online_memory`; do
+	if [ $((RANDOM % 100)) -lt $ratio ]; then
+		offline_memory_expect_success $memory
+	fi
+done
+
+#
+# Test memory hot-add error handling (offline => online)
+#
+echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_ONLINE/error
+for memory in `hotplaggable_offline_memory`; do
+	online_memory_expect_fail $memory
+done
+
+#
+# Online all hot-pluggable memory
+#
+echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_ONLINE/error
+for memory in `hotplaggable_offline_memory`; do
+	online_memory_expect_success $memory
+done
+
+#
+# Test memory hot-remove error handling (online => offline)
+#
+echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_OFFLINE/error
+for memory in `hotpluggable_online_memory`; do
+	offline_memory_expect_fail $memory
+done
+
+echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_OFFLINE/error
+/sbin/modprobe -q -r memory-notifier-error-inject

tools/testing/selftests/mqueue/.gitignore

@@ -0,0 +1,2 @@
+mq_open_tests
+mq_perf_tests

tools/testing/selftests/mqueue/Makefile

@@ -0,0 +1,10 @@
+all:
+	gcc -O2 -lrt mq_open_tests.c -o mq_open_tests
+	gcc -O2 -lrt -lpthread -lpopt -o mq_perf_tests mq_perf_tests.c
+
+run_tests:
+	@./mq_open_tests /test1 || echo "mq_open_tests: [FAIL]"
+	@./mq_perf_tests || echo "mq_perf_tests: [FAIL]"
+
+clean:
+	rm -f mq_open_tests mq_perf_tests

tools/testing/selftests/mqueue/mq_open_tests.c

@@ -0,0 +1,492 @@
+/*
+ * This application is Copyright 2012 Red Hat, Inc.
+ *	Doug Ledford <dledford@redhat.com>
+ *
+ * mq_open_tests is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, version 3.
+ *
+ * mq_open_tests is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * For the full text of the license, see <http://www.gnu.org/licenses/>.
+ *
+ * mq_open_tests.c
+ *   Tests the various situations that should either succeed or fail to
+ *   open a posix message queue and then reports whether or not they
+ *   did as they were supposed to.
+ *
+ */
+#include <stdio.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <string.h>
+#include <limits.h>
+#include <errno.h>
+#include <sys/types.h>
+#include <sys/time.h>
+#include <sys/resource.h>
+#include <sys/stat.h>
+#include <mqueue.h>
+
+static char *usage =
+"Usage:\n"
+"  %s path\n"
+"\n"
+"	path	Path name of the message queue to create\n"
+"\n"
+"	Note: this program must be run as root in order to enable all tests\n"
+"\n";
+
+char *DEF_MSGS = "/proc/sys/fs/mqueue/msg_default";
+char *DEF_MSGSIZE = "/proc/sys/fs/mqueue/msgsize_default";
+char *MAX_MSGS = "/proc/sys/fs/mqueue/msg_max";
+char *MAX_MSGSIZE = "/proc/sys/fs/mqueue/msgsize_max";
+
+int default_settings;
+struct rlimit saved_limits, cur_limits;
+int saved_def_msgs, saved_def_msgsize, saved_max_msgs, saved_max_msgsize;
+int cur_def_msgs, cur_def_msgsize, cur_max_msgs, cur_max_msgsize;
+FILE *def_msgs, *def_msgsize, *max_msgs, *max_msgsize;
+char *queue_path;
+mqd_t queue = -1;
+
+static inline void __set(FILE *stream, int value, char *err_msg);
+void shutdown(int exit_val, char *err_cause, int line_no);
+static inline int get(FILE *stream);
+static inline void set(FILE *stream, int value);
+static inline void getr(int type, struct rlimit *rlim);
+static inline void setr(int type, struct rlimit *rlim);
+void validate_current_settings();
+static inline void test_queue(struct mq_attr *attr, struct mq_attr *result);
+static inline int test_queue_fail(struct mq_attr *attr, struct mq_attr *result);
+
+static inline void __set(FILE *stream, int value, char *err_msg)
+{
+	rewind(stream);
+	if (fprintf(stream, "%d", value) < 0)
+		perror(err_msg);
+}
+
+
+void shutdown(int exit_val, char *err_cause, int line_no)
+{
+	static int in_shutdown = 0;
+
+	/* In case we get called recursively by a set() call below */
+	if (in_shutdown++)
+		return;
+
+	seteuid(0);
+
+	if (queue != -1)
+		if (mq_close(queue))
+			perror("mq_close() during shutdown");
+	if (queue_path)
+		/*
+		 * Be silent if this fails, if we cleaned up already it's
+		 * expected to fail
+		 */
+		mq_unlink(queue_path);
+	if (default_settings) {
+		if (saved_def_msgs)
+			__set(def_msgs, saved_def_msgs,
+			      "failed to restore saved_def_msgs");
+		if (saved_def_msgsize)
+			__set(def_msgsize, saved_def_msgsize,
+			      "failed to restore saved_def_msgsize");
+	}
+	if (saved_max_msgs)
+		__set(max_msgs, saved_max_msgs,
+		      "failed to restore saved_max_msgs");
+	if (saved_max_msgsize)
+		__set(max_msgsize, saved_max_msgsize,
+		      "failed to restore saved_max_msgsize");
+	if (exit_val)
+		error(exit_val, errno, "%s at %d", err_cause, line_no);
+	exit(0);
+}
+
+static inline int get(FILE *stream)
+{
+	int value;
+	rewind(stream);
+	if (fscanf(stream, "%d", &value) != 1)
+		shutdown(4, "Error reading /proc entry", __LINE__ - 1);
+	return value;
+}
+
+static inline void set(FILE *stream, int value)
+{
+	int new_value;
+
+	rewind(stream);
+	if (fprintf(stream, "%d", value) < 0)
+		return shutdown(5, "Failed writing to /proc file",
+				__LINE__ - 1);
+	new_value = get(stream);
+	if (new_value != value)
+		return shutdown(5, "We didn't get what we wrote to /proc back",
+				__LINE__ - 1);
+}
+
+static inline void getr(int type, struct rlimit *rlim)
+{
+	if (getrlimit(type, rlim))
+		shutdown(6, "getrlimit()", __LINE__ - 1);
+}
+
+static inline void setr(int type, struct rlimit *rlim)
+{
+	if (setrlimit(type, rlim))
+		shutdown(7, "setrlimit()", __LINE__ - 1);
+}
+
+void validate_current_settings()
+{
+	int rlim_needed;
+
+	if (cur_limits.rlim_cur < 4096) {
+		printf("Current rlimit value for POSIX message queue bytes is "
+		       "unreasonably low,\nincreasing.\n\n");
+		cur_limits.rlim_cur = 8192;
+		cur_limits.rlim_max = 16384;
+		setr(RLIMIT_MSGQUEUE, &cur_limits);
+	}
+
+	if (default_settings) {
+		rlim_needed = (cur_def_msgs + 1) * (cur_def_msgsize + 1 +
+						    2 * sizeof(void *));
+		if (rlim_needed > cur_limits.rlim_cur) {
+			printf("Temporarily lowering default queue parameters "
+			       "to something that will work\n"
+			       "with the current rlimit values.\n\n");
+			set(def_msgs, 10);
+			cur_def_msgs = 10;
+			set(def_msgsize, 128);
+			cur_def_msgsize = 128;
+		}
+	} else {
+		rlim_needed = (cur_max_msgs + 1) * (cur_max_msgsize + 1 +
+						    2 * sizeof(void *));
+		if (rlim_needed > cur_limits.rlim_cur) {
+			printf("Temporarily lowering maximum queue parameters "
+			       "to something that will work\n"
+			       "with the current rlimit values in case this is "
+			       "a kernel that ties the default\n"
+			       "queue parameters to the maximum queue "
+			       "parameters.\n\n");
+			set(max_msgs, 10);
+			cur_max_msgs = 10;
+			set(max_msgsize, 128);
+			cur_max_msgsize = 128;
+		}
+	}
+}
+
+/*
+ * test_queue - Test opening a queue, shutdown if we fail.  This should
+ * only be called in situations that should never fail.  We clean up
+ * after ourselves and return the queue attributes in *result.
+ */
+static inline void test_queue(struct mq_attr *attr, struct mq_attr *result)
+{
+	int flags = O_RDWR | O_EXCL | O_CREAT;
+	int perms = DEFFILEMODE;
+
+	if ((queue = mq_open(queue_path, flags, perms, attr)) == -1)
+		shutdown(1, "mq_open()", __LINE__);
+	if (mq_getattr(queue, result))
+		shutdown(1, "mq_getattr()", __LINE__);
+	if (mq_close(queue))
+		shutdown(1, "mq_close()", __LINE__);
+	queue = -1;
+	if (mq_unlink(queue_path))
+		shutdown(1, "mq_unlink()", __LINE__);
+}
+
+/*
+ * Same as test_queue above, but failure is not fatal.
+ * Returns:
+ * 0 - Failed to create a queue
+ * 1 - Created a queue, attributes in *result
+ */
+static inline int test_queue_fail(struct mq_attr *attr, struct mq_attr *result)
+{
+	int flags = O_RDWR | O_EXCL | O_CREAT;
+	int perms = DEFFILEMODE;
+
+	if ((queue = mq_open(queue_path, flags, perms, attr)) == -1)
+		return 0;
+	if (mq_getattr(queue, result))
+		shutdown(1, "mq_getattr()", __LINE__);
+	if (mq_close(queue))
+		shutdown(1, "mq_close()", __LINE__);
+	queue = -1;
+	if (mq_unlink(queue_path))
+		shutdown(1, "mq_unlink()", __LINE__);
+	return 1;
+}
+
+int main(int argc, char *argv[])
+{
+	struct mq_attr attr, result;
+
+	if (argc != 2) {
+		fprintf(stderr, "Must pass a valid queue name\n\n");
+		fprintf(stderr, usage, argv[0]);
+		exit(1);
+	}
+
+	/*
+	 * Although we can create a msg queue with a non-absolute path name,
+	 * unlink will fail.  So, if the name doesn't start with a /, add one
+	 * when we save it.
+	 */
+	if (*argv[1] == '/')
+		queue_path = strdup(argv[1]);
+	else {
+		queue_path = malloc(strlen(argv[1]) + 2);
+		if (!queue_path) {
+			perror("malloc()");
+			exit(1);
+		}
+		queue_path[0] = '/';
+		queue_path[1] = 0;
+		strcat(queue_path, argv[1]);
+	}
+
+	if (getuid() != 0) {
+		fprintf(stderr, "Not running as root, but almost all tests "
+			"require root in order to modify\nsystem settings.  "
+			"Exiting.\n");
+		exit(1);
+	}
+
+	/* Find out what files there are for us to make tweaks in */
+	def_msgs = fopen(DEF_MSGS, "r+");
+	def_msgsize = fopen(DEF_MSGSIZE, "r+");
+	max_msgs = fopen(MAX_MSGS, "r+");
+	max_msgsize = fopen(MAX_MSGSIZE, "r+");
+
+	if (!max_msgs)
+		shutdown(2, "Failed to open msg_max", __LINE__);
+	if (!max_msgsize)
+		shutdown(2, "Failed to open msgsize_max", __LINE__);
+	if (def_msgs || def_msgsize)
+		default_settings = 1;
+
+	/* Load up the current system values for everything we can */
+	getr(RLIMIT_MSGQUEUE, &saved_limits);
+	cur_limits = saved_limits;
+	if (default_settings) {
+		saved_def_msgs = cur_def_msgs = get(def_msgs);
+		saved_def_msgsize = cur_def_msgsize = get(def_msgsize);
+	}
+	saved_max_msgs = cur_max_msgs = get(max_msgs);
+	saved_max_msgsize = cur_max_msgsize = get(max_msgsize);
+
+	/* Tell the user our initial state */
+	printf("\nInitial system state:\n");
+	printf("\tUsing queue path:\t\t%s\n", queue_path);
+	printf("\tRLIMIT_MSGQUEUE(soft):\t\t%d\n", saved_limits.rlim_cur);
+	printf("\tRLIMIT_MSGQUEUE(hard):\t\t%d\n", saved_limits.rlim_max);
+	printf("\tMaximum Message Size:\t\t%d\n", saved_max_msgsize);
+	printf("\tMaximum Queue Size:\t\t%d\n", saved_max_msgs);
+	if (default_settings) {
+		printf("\tDefault Message Size:\t\t%d\n", saved_def_msgsize);
+		printf("\tDefault Queue Size:\t\t%d\n", saved_def_msgs);
+	} else {
+		printf("\tDefault Message Size:\t\tNot Supported\n");
+		printf("\tDefault Queue Size:\t\tNot Supported\n");
+	}
+	printf("\n");
+
+	validate_current_settings();
+
+	printf("Adjusted system state for testing:\n");
+	printf("\tRLIMIT_MSGQUEUE(soft):\t\t%d\n", cur_limits.rlim_cur);
+	printf("\tRLIMIT_MSGQUEUE(hard):\t\t%d\n", cur_limits.rlim_max);
+	printf("\tMaximum Message Size:\t\t%d\n", cur_max_msgsize);
+	printf("\tMaximum Queue Size:\t\t%d\n", cur_max_msgs);
+	if (default_settings) {
+		printf("\tDefault Message Size:\t\t%d\n", cur_def_msgsize);
+		printf("\tDefault Queue Size:\t\t%d\n", cur_def_msgs);
+	}
+
+	printf("\n\nTest series 1, behavior when no attr struct "
+	       "passed to mq_open:\n");
+	if (!default_settings) {
+		test_queue(NULL, &result);
+		printf("Given sane system settings, mq_open without an attr "
+		       "struct succeeds:\tPASS\n");
+		if (result.mq_maxmsg != cur_max_msgs ||
+		    result.mq_msgsize != cur_max_msgsize) {
+			printf("Kernel does not support setting the default "
+			       "mq attributes,\nbut also doesn't tie the "
+			       "defaults to the maximums:\t\t\tPASS\n");
+		} else {
+			set(max_msgs, ++cur_max_msgs);
+			set(max_msgsize, ++cur_max_msgsize);
+			test_queue(NULL, &result);
+			if (result.mq_maxmsg == cur_max_msgs &&
+			    result.mq_msgsize == cur_max_msgsize)
+				printf("Kernel does not support setting the "
+				       "default mq attributes and\n"
+				       "also ties system wide defaults to "
+				       "the system wide maximums:\t\t"
+				       "FAIL\n");
+			else
+				printf("Kernel does not support setting the "
+				       "default mq attributes,\n"
+				       "but also doesn't tie the defaults to "
+				       "the maximums:\t\t\tPASS\n");
+		}
+	} else {
+		printf("Kernel supports setting defaults separately from "
+		       "maximums:\t\tPASS\n");
+		/*
+		 * While we are here, go ahead and test that the kernel
+		 * properly follows the default settings
+		 */
+		test_queue(NULL, &result);
+		printf("Given sane values, mq_open without an attr struct "
+		       "succeeds:\t\tPASS\n");
+		if (result.mq_maxmsg != cur_def_msgs ||
+		    result.mq_msgsize != cur_def_msgsize)
+			printf("Kernel supports setting defaults, but does "
+			       "not actually honor them:\tFAIL\n\n");
+		else {
+			set(def_msgs, ++cur_def_msgs);
+			set(def_msgsize, ++cur_def_msgsize);
+			/* In case max was the same as the default */
+			set(max_msgs, ++cur_max_msgs);
+			set(max_msgsize, ++cur_max_msgsize);
+			test_queue(NULL, &result);
+			if (result.mq_maxmsg != cur_def_msgs ||
+			    result.mq_msgsize != cur_def_msgsize)
+				printf("Kernel supports setting defaults, but "
+				       "does not actually honor them:\t"
+				       "FAIL\n");
+			else
+				printf("Kernel properly honors default setting "
+				       "knobs:\t\t\t\tPASS\n");
+		}
+		set(def_msgs, cur_max_msgs + 1);
+		cur_def_msgs = cur_max_msgs + 1;
+		set(def_msgsize, cur_max_msgsize + 1);
+		cur_def_msgsize = cur_max_msgsize + 1;
+		if (cur_def_msgs * (cur_def_msgsize + 2 * sizeof(void *)) >=
+		    cur_limits.rlim_cur) {
+			cur_limits.rlim_cur = (cur_def_msgs + 2) *
+				(cur_def_msgsize + 2 * sizeof(void *));
+			cur_limits.rlim_max = 2 * cur_limits.rlim_cur;
+			setr(RLIMIT_MSGQUEUE, &cur_limits);
+		}
+		if (test_queue_fail(NULL, &result)) {
+			if (result.mq_maxmsg == cur_max_msgs &&
+			    result.mq_msgsize == cur_max_msgsize)
+				printf("Kernel properly limits default values "
+				       "to lesser of default/max:\t\tPASS\n");
+			else
+				printf("Kernel does not properly set default "
+				       "queue parameters when\ndefaults > "
+				       "max:\t\t\t\t\t\t\t\tFAIL\n");
+		} else
+			printf("Kernel fails to open mq because defaults are "
+			       "greater than maximums:\tFAIL\n");
+		set(def_msgs, --cur_def_msgs);
+		set(def_msgsize, --cur_def_msgsize);
+		cur_limits.rlim_cur = cur_limits.rlim_max = cur_def_msgs *
+			cur_def_msgsize;
+		setr(RLIMIT_MSGQUEUE, &cur_limits);
+		if (test_queue_fail(NULL, &result))
+			printf("Kernel creates queue even though defaults "
+			       "would exceed\nrlimit setting:"
+			       "\t\t\t\t\t\t\t\tFAIL\n");
+		else
+			printf("Kernel properly fails to create queue when "
+			       "defaults would\nexceed rlimit:"
+			       "\t\t\t\t\t\t\t\tPASS\n");
+	}
+
+	/*
+	 * Test #2 - open with an attr struct that exceeds rlimit
+	 */
+	printf("\n\nTest series 2, behavior when attr struct is "
+	       "passed to mq_open:\n");
+	cur_max_msgs = 32;
+	cur_max_msgsize = cur_limits.rlim_max >> 4;
+	set(max_msgs, cur_max_msgs);
+	set(max_msgsize, cur_max_msgsize);
+	attr.mq_maxmsg = cur_max_msgs;
+	attr.mq_msgsize = cur_max_msgsize;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open in excess of rlimit max when euid = 0 "
+		       "succeeded:\t\tFAIL\n");
+	else
+		printf("Queue open in excess of rlimit max when euid = 0 "
+		       "failed:\t\tPASS\n");
+	attr.mq_maxmsg = cur_max_msgs + 1;
+	attr.mq_msgsize = 10;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open with mq_maxmsg > limit when euid = 0 "
+		       "succeeded:\t\tPASS\n");
+	else
+		printf("Queue open with mq_maxmsg > limit when euid = 0 "
+		       "failed:\t\tFAIL\n");
+	attr.mq_maxmsg = 1;
+	attr.mq_msgsize = cur_max_msgsize + 1;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open with mq_msgsize > limit when euid = 0 "
+		       "succeeded:\t\tPASS\n");
+	else
+		printf("Queue open with mq_msgsize > limit when euid = 0 "
+		       "failed:\t\tFAIL\n");
+	attr.mq_maxmsg = 65536;
+	attr.mq_msgsize = 65536;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open with total size > 2GB when euid = 0 "
+		       "succeeded:\t\tFAIL\n");
+	else
+		printf("Queue open with total size > 2GB when euid = 0 "
+		       "failed:\t\t\tPASS\n");
+	seteuid(99);
+	attr.mq_maxmsg = cur_max_msgs;
+	attr.mq_msgsize = cur_max_msgsize;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open in excess of rlimit max when euid = 99 "
+		       "succeeded:\t\tFAIL\n");
+	else
+		printf("Queue open in excess of rlimit max when euid = 99 "
+		       "failed:\t\tPASS\n");
+	attr.mq_maxmsg = cur_max_msgs + 1;
+	attr.mq_msgsize = 10;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open with mq_maxmsg > limit when euid = 99 "
+		       "succeeded:\t\tFAIL\n");
+	else
+		printf("Queue open with mq_maxmsg > limit when euid = 99 "
+		       "failed:\t\tPASS\n");
+	attr.mq_maxmsg = 1;
+	attr.mq_msgsize = cur_max_msgsize + 1;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open with mq_msgsize > limit when euid = 99 "
+		       "succeeded:\t\tFAIL\n");
+	else
+		printf("Queue open with mq_msgsize > limit when euid = 99 "
+		       "failed:\t\tPASS\n");
+	attr.mq_maxmsg = 65536;
+	attr.mq_msgsize = 65536;
+	if (test_queue_fail(&attr, &result))
+		printf("Queue open with total size > 2GB when euid = 99 "
+		       "succeeded:\t\tFAIL\n");
+	else
+		printf("Queue open with total size > 2GB when euid = 99 "
+		       "failed:\t\t\tPASS\n");
+
+	shutdown(0,"",0);
+}

tools/testing/selftests/mqueue/mq_perf_tests.c

@@ -0,0 +1,741 @@
+/*
+ * This application is Copyright 2012 Red Hat, Inc.
+ *	Doug Ledford <dledford@redhat.com>
+ *
+ * mq_perf_tests is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, version 3.
+ *
+ * mq_perf_tests is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * For the full text of the license, see <http://www.gnu.org/licenses/>.
+ *
+ * mq_perf_tests.c
+ *   Tests various types of message queue workloads, concentrating on those
+ *   situations that invole large message sizes, large message queue depths,
+ *   or both, and reports back useful metrics about kernel message queue
+ *   performance.
+ *
+ */
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <string.h>
+#include <limits.h>
+#include <errno.h>
+#include <signal.h>
+#include <pthread.h>
+#include <sched.h>
+#include <sys/types.h>
+#include <sys/time.h>
+#include <sys/resource.h>
+#include <sys/stat.h>
+#include <mqueue.h>
+#include <popt.h>
+
+static char *usage =
+"Usage:\n"
+"  %s [-c #[,#..] -f] path\n"
+"\n"
+"	-c #	Skip most tests and go straight to a high queue depth test\n"
+"		and then run that test continuously (useful for running at\n"
+"		the same time as some other workload to see how much the\n"
+"		cache thrashing caused by adding messages to a very deep\n"
+"		queue impacts the performance of other programs).  The number\n"
+"		indicates which CPU core we should bind the process to during\n"
+"		the run.  If you have more than one physical CPU, then you\n"
+"		will need one copy per physical CPU package, and you should\n"
+"		specify the CPU cores to pin ourself to via a comma separated\n"
+"		list of CPU values.\n"
+"	-f	Only usable with continuous mode.  Pin ourself to the CPUs\n"
+"		as requested, then instead of looping doing a high mq\n"
+"		workload, just busy loop.  This will allow us to lock up a\n"
+"		single CPU just like we normally would, but without actually\n"
+"		thrashing the CPU cache.  This is to make it easier to get\n"
+"		comparable numbers from some other workload running on the\n"
+"		other CPUs.  One set of numbers with # CPUs locked up running\n"
+"		an mq workload, and another set of numbers with those same\n"
+"		CPUs locked away from the test workload, but not doing\n"
+"		anything to trash the cache like the mq workload might.\n"
+"	path	Path name of the message queue to create\n"
+"\n"
+"	Note: this program must be run as root in order to enable all tests\n"
+"\n";
+
+char *MAX_MSGS = "/proc/sys/fs/mqueue/msg_max";
+char *MAX_MSGSIZE = "/proc/sys/fs/mqueue/msgsize_max";
+
+#define min(a, b) ((a) < (b) ? (a) : (b))
+#define MAX_CPUS 64
+char *cpu_option_string;
+int cpus_to_pin[MAX_CPUS];
+int num_cpus_to_pin;
+pthread_t cpu_threads[MAX_CPUS];
+pthread_t main_thread;
+cpu_set_t *cpu_set;
+int cpu_set_size;
+int cpus_online;
+
+#define MSG_SIZE 16
+#define TEST1_LOOPS 10000000
+#define TEST2_LOOPS 100000
+int continuous_mode;
+int continuous_mode_fake;
+
+struct rlimit saved_limits, cur_limits;
+int saved_max_msgs, saved_max_msgsize;
+int cur_max_msgs, cur_max_msgsize;
+FILE *max_msgs, *max_msgsize;
+int cur_nice;
+char *queue_path = "/mq_perf_tests";
+mqd_t queue = -1;
+struct mq_attr result;
+int mq_prio_max;
+
+const struct poptOption options[] = {
+	{
+		.longName = "continuous",
+		.shortName = 'c',
+		.argInfo = POPT_ARG_STRING,
+		.arg = &cpu_option_string,
+		.val = 'c',
+		.descrip = "Run continuous tests at a high queue depth in "
+			"order to test the effects of cache thrashing on "
+			"other tasks on the system.  This test is intended "
+			"to be run on one core of each physical CPU while "
+			"some other CPU intensive task is run on all the other "
+			"cores of that same physical CPU and the other task "
+			"is timed.  It is assumed that the process of adding "
+			"messages to the message queue in a tight loop will "
+			"impact that other task to some degree.  Once the "
+			"tests are performed in this way, you should then "
+			"re-run the tests using fake mode in order to check "
+			"the difference in time required to perform the CPU "
+			"intensive task",
+		.argDescrip = "cpu[,cpu]",
+	},
+	{
+		.longName = "fake",
+		.shortName = 'f',
+		.argInfo = POPT_ARG_NONE,
+		.arg = &continuous_mode_fake,
+		.val = 0,
+		.descrip = "Tie up the CPUs that we would normally tie up in"
+			"continuous mode, but don't actually do any mq stuff, "
+			"just keep the CPU busy so it can't be used to process "
+			"system level tasks as this would free up resources on "
+			"the other CPU cores and skew the comparison between "
+			"the no-mqueue work and mqueue work tests",
+		.argDescrip = NULL,
+	},
+	{
+		.longName = "path",
+		.shortName = 'p',
+		.argInfo = POPT_ARG_STRING | POPT_ARGFLAG_SHOW_DEFAULT,
+		.arg = &queue_path,
+		.val = 'p',
+		.descrip = "The name of the path to use in the mqueue "
+			"filesystem for our tests",
+		.argDescrip = "pathname",
+	},
+	POPT_AUTOHELP
+	POPT_TABLEEND
+};
+
+static inline void __set(FILE *stream, int value, char *err_msg);
+void shutdown(int exit_val, char *err_cause, int line_no);
+void sig_action_SIGUSR1(int signum, siginfo_t *info, void *context);
+void sig_action(int signum, siginfo_t *info, void *context);
+static inline int get(FILE *stream);
+static inline void set(FILE *stream, int value);
+static inline int try_set(FILE *stream, int value);
+static inline void getr(int type, struct rlimit *rlim);
+static inline void setr(int type, struct rlimit *rlim);
+static inline void open_queue(struct mq_attr *attr);
+void increase_limits(void);
+
+static inline void __set(FILE *stream, int value, char *err_msg)
+{
+	rewind(stream);
+	if (fprintf(stream, "%d", value) < 0)
+		perror(err_msg);
+}
+
+
+void shutdown(int exit_val, char *err_cause, int line_no)
+{
+	static int in_shutdown = 0;
+	int errno_at_shutdown = errno;
+	int i;
+
+	/* In case we get called by multiple threads or from an sighandler */
+	if (in_shutdown++)
+		return;
+
+	for (i = 0; i < num_cpus_to_pin; i++)
+		if (cpu_threads[i]) {
+			pthread_kill(cpu_threads[i], SIGUSR1);
+			pthread_join(cpu_threads[i], NULL);
+		}
+
+	if (queue != -1)
+		if (mq_close(queue))
+			perror("mq_close() during shutdown");
+	if (queue_path)
+		/*
+		 * Be silent if this fails, if we cleaned up already it's
+		 * expected to fail
+		 */
+		mq_unlink(queue_path);
+	if (saved_max_msgs)
+		__set(max_msgs, saved_max_msgs,
+		      "failed to restore saved_max_msgs");
+	if (saved_max_msgsize)
+		__set(max_msgsize, saved_max_msgsize,
+		      "failed to restore saved_max_msgsize");
+	if (exit_val)
+		error(exit_val, errno_at_shutdown, "%s at %d",
+		      err_cause, line_no);
+	exit(0);
+}
+
+void sig_action_SIGUSR1(int signum, siginfo_t *info, void *context)
+{
+	if (pthread_self() != main_thread)
+		pthread_exit(0);
+	else {
+		fprintf(stderr, "Caught signal %d in SIGUSR1 handler, "
+				"exiting\n", signum);
+		shutdown(0, "", 0);
+		fprintf(stderr, "\n\nReturned from shutdown?!?!\n\n");
+		exit(0);
+	}
+}
+
+void sig_action(int signum, siginfo_t *info, void *context)
+{
+	if (pthread_self() != main_thread)
+		pthread_kill(main_thread, signum);
+	else {
+		fprintf(stderr, "Caught signal %d, exiting\n", signum);
+		shutdown(0, "", 0);
+		fprintf(stderr, "\n\nReturned from shutdown?!?!\n\n");
+		exit(0);
+	}
+}
+
+static inline int get(FILE *stream)
+{
+	int value;
+	rewind(stream);
+	if (fscanf(stream, "%d", &value) != 1)
+		shutdown(4, "Error reading /proc entry", __LINE__);
+	return value;
+}
+
+static inline void set(FILE *stream, int value)
+{
+	int new_value;
+
+	rewind(stream);
+	if (fprintf(stream, "%d", value) < 0)
+		return shutdown(5, "Failed writing to /proc file", __LINE__);
+	new_value = get(stream);
+	if (new_value != value)
+		return shutdown(5, "We didn't get what we wrote to /proc back",
+				__LINE__);
+}
+
+static inline int try_set(FILE *stream, int value)
+{
+	int new_value;
+
+	rewind(stream);
+	fprintf(stream, "%d", value);
+	new_value = get(stream);
+	return new_value == value;
+}
+
+static inline void getr(int type, struct rlimit *rlim)
+{
+	if (getrlimit(type, rlim))
+		shutdown(6, "getrlimit()", __LINE__);
+}
+
+static inline void setr(int type, struct rlimit *rlim)
+{
+	if (setrlimit(type, rlim))
+		shutdown(7, "setrlimit()", __LINE__);
+}
+
+/**
+ * open_queue - open the global queue for testing
+ * @attr - An attr struct specifying the desired queue traits
+ * @result - An attr struct that lists the actual traits the queue has
+ *
+ * This open is not allowed to fail, failure will result in an orderly
+ * shutdown of the program.  The global queue_path is used to set what
+ * queue to open, the queue descriptor is saved in the global queue
+ * variable.
+ */
+static inline void open_queue(struct mq_attr *attr)
+{
+	int flags = O_RDWR | O_EXCL | O_CREAT | O_NONBLOCK;
+	int perms = DEFFILEMODE;
+
+	queue = mq_open(queue_path, flags, perms, attr);
+	if (queue == -1)
+		shutdown(1, "mq_open()", __LINE__);
+	if (mq_getattr(queue, &result))
+		shutdown(1, "mq_getattr()", __LINE__);
+	printf("\n\tQueue %s created:\n", queue_path);
+	printf("\t\tmq_flags:\t\t\t%s\n", result.mq_flags & O_NONBLOCK ?
+	       "O_NONBLOCK" : "(null)");
+	printf("\t\tmq_maxmsg:\t\t\t%d\n", result.mq_maxmsg);
+	printf("\t\tmq_msgsize:\t\t\t%d\n", result.mq_msgsize);
+	printf("\t\tmq_curmsgs:\t\t\t%d\n", result.mq_curmsgs);
+}
+
+void *fake_cont_thread(void *arg)
+{
+	int i;
+
+	for (i = 0; i < num_cpus_to_pin; i++)
+		if (cpu_threads[i] == pthread_self())
+			break;
+	printf("\tStarted fake continuous mode thread %d on CPU %d\n", i,
+	       cpus_to_pin[i]);
+	while (1)
+		;
+}
+
+void *cont_thread(void *arg)
+{
+	char buff[MSG_SIZE];
+	int i, priority;
+
+	for (i = 0; i < num_cpus_to_pin; i++)
+		if (cpu_threads[i] == pthread_self())
+			break;
+	printf("\tStarted continuous mode thread %d on CPU %d\n", i,
+	       cpus_to_pin[i]);
+	while (1) {
+		while (mq_send(queue, buff, sizeof(buff), 0) == 0)
+			;
+		mq_receive(queue, buff, sizeof(buff), &priority);
+	}
+}
+
+#define drain_queue() \
+	while (mq_receive(queue, buff, MSG_SIZE, &prio_in) == MSG_SIZE)
+
+#define do_untimed_send() \
+	do { \
+		if (mq_send(queue, buff, MSG_SIZE, prio_out)) \
+			shutdown(3, "Test send failure", __LINE__); \
+	} while (0)
+
+#define do_send_recv() \
+	do { \
+		clock_gettime(clock, &start); \
+		if (mq_send(queue, buff, MSG_SIZE, prio_out)) \
+			shutdown(3, "Test send failure", __LINE__); \
+		clock_gettime(clock, &middle); \
+		if (mq_receive(queue, buff, MSG_SIZE, &prio_in) != MSG_SIZE) \
+			shutdown(3, "Test receive failure", __LINE__); \
+		clock_gettime(clock, &end); \
+		nsec = ((middle.tv_sec - start.tv_sec) * 1000000000) + \
+			(middle.tv_nsec - start.tv_nsec); \
+		send_total.tv_nsec += nsec; \
+		if (send_total.tv_nsec >= 1000000000) { \
+			send_total.tv_sec++; \
+			send_total.tv_nsec -= 1000000000; \
+		} \
+		nsec = ((end.tv_sec - middle.tv_sec) * 1000000000) + \
+			(end.tv_nsec - middle.tv_nsec); \
+		recv_total.tv_nsec += nsec; \
+		if (recv_total.tv_nsec >= 1000000000) { \
+			recv_total.tv_sec++; \
+			recv_total.tv_nsec -= 1000000000; \
+		} \
+	} while (0)
+
+struct test {
+	char *desc;
+	void (*func)(int *);
+};
+
+void const_prio(int *prio)
+{
+	return;
+}
+
+void inc_prio(int *prio)
+{
+	if (++*prio == mq_prio_max)
+		*prio = 0;
+}
+
+void dec_prio(int *prio)
+{
+	if (--*prio < 0)
+		*prio = mq_prio_max - 1;
+}
+
+void random_prio(int *prio)
+{
+	*prio = random() % mq_prio_max;
+}
+
+struct test test2[] = {
+	{"\n\tTest #2a: Time send/recv message, queue full, constant prio\n",
+		const_prio},
+	{"\n\tTest #2b: Time send/recv message, queue full, increasing prio\n",
+		inc_prio},
+	{"\n\tTest #2c: Time send/recv message, queue full, decreasing prio\n",
+		dec_prio},
+	{"\n\tTest #2d: Time send/recv message, queue full, random prio\n",
+		random_prio},
+	{NULL, NULL}
+};
+
+/**
+ * Tests to perform (all done with MSG_SIZE messages):
+ *
+ * 1) Time to add/remove message with 0 messages on queue
+ * 1a) with constant prio
+ * 2) Time to add/remove message when queue close to capacity:
+ * 2a) with constant prio
+ * 2b) with increasing prio
+ * 2c) with decreasing prio
+ * 2d) with random prio
+ * 3) Test limits of priorities honored (double check _SC_MQ_PRIO_MAX)
+ */
+void *perf_test_thread(void *arg)
+{
+	char buff[MSG_SIZE];
+	int prio_out, prio_in;
+	int i;
+	clockid_t clock;
+	pthread_t *t;
+	struct timespec res, start, middle, end, send_total, recv_total;
+	unsigned long long nsec;
+	struct test *cur_test;
+
+	t = &cpu_threads[0];
+	printf("\n\tStarted mqueue performance test thread on CPU %d\n",
+	       cpus_to_pin[0]);
+	mq_prio_max = sysconf(_SC_MQ_PRIO_MAX);
+	if (mq_prio_max == -1)
+		shutdown(2, "sysconf(_SC_MQ_PRIO_MAX)", __LINE__);
+	if (pthread_getcpuclockid(cpu_threads[0], &clock) != 0)
+		shutdown(2, "pthread_getcpuclockid", __LINE__);
+
+	if (clock_getres(clock, &res))
+		shutdown(2, "clock_getres()", __LINE__);
+
+	printf("\t\tMax priorities:\t\t\t%d\n", mq_prio_max);
+	printf("\t\tClock resolution:\t\t%d nsec%s\n", res.tv_nsec,
+	       res.tv_nsec > 1 ? "s" : "");
+
+
+
+	printf("\n\tTest #1: Time send/recv message, queue empty\n");
+	printf("\t\t(%d iterations)\n", TEST1_LOOPS);
+	prio_out = 0;
+	send_total.tv_sec = 0;
+	send_total.tv_nsec = 0;
+	recv_total.tv_sec = 0;
+	recv_total.tv_nsec = 0;
+	for (i = 0; i < TEST1_LOOPS; i++)
+		do_send_recv();
+	printf("\t\tSend msg:\t\t\t%d.%ds total time\n",
+	       send_total.tv_sec, send_total.tv_nsec);
+	nsec = ((unsigned long long)send_total.tv_sec * 1000000000 +
+		 send_total.tv_nsec) / TEST1_LOOPS;
+	printf("\t\t\t\t\t\t%d nsec/msg\n", nsec);
+	printf("\t\tRecv msg:\t\t\t%d.%ds total time\n",
+	       recv_total.tv_sec, recv_total.tv_nsec);
+	nsec = ((unsigned long long)recv_total.tv_sec * 1000000000 +
+		recv_total.tv_nsec) / TEST1_LOOPS;
+	printf("\t\t\t\t\t\t%d nsec/msg\n", nsec);
+
+
+	for (cur_test = test2; cur_test->desc != NULL; cur_test++) {
+		printf(cur_test->desc);
+		printf("\t\t(%d iterations)\n", TEST2_LOOPS);
+		prio_out = 0;
+		send_total.tv_sec = 0;
+		send_total.tv_nsec = 0;
+		recv_total.tv_sec = 0;
+		recv_total.tv_nsec = 0;
+		printf("\t\tFilling queue...");
+		fflush(stdout);
+		clock_gettime(clock, &start);
+		for (i = 0; i < result.mq_maxmsg - 1; i++) {
+			do_untimed_send();
+			cur_test->func(&prio_out);
+		}
+		clock_gettime(clock, &end);
+		nsec = ((unsigned long long)(end.tv_sec - start.tv_sec) *
+			1000000000) + (end.tv_nsec - start.tv_nsec);
+		printf("done.\t\t%lld.%llds\n", nsec / 1000000000,
+		       nsec % 1000000000);
+		printf("\t\tTesting...");
+		fflush(stdout);
+		for (i = 0; i < TEST2_LOOPS; i++) {
+			do_send_recv();
+			cur_test->func(&prio_out);
+		}
+		printf("done.\n");
+		printf("\t\tSend msg:\t\t\t%d.%ds total time\n",
+		       send_total.tv_sec, send_total.tv_nsec);
+		nsec = ((unsigned long long)send_total.tv_sec * 1000000000 +
+			 send_total.tv_nsec) / TEST2_LOOPS;
+		printf("\t\t\t\t\t\t%d nsec/msg\n", nsec);
+		printf("\t\tRecv msg:\t\t\t%d.%ds total time\n",
+		       recv_total.tv_sec, recv_total.tv_nsec);
+		nsec = ((unsigned long long)recv_total.tv_sec * 1000000000 +
+			recv_total.tv_nsec) / TEST2_LOOPS;
+		printf("\t\t\t\t\t\t%d nsec/msg\n", nsec);
+		printf("\t\tDraining queue...");
+		fflush(stdout);
+		clock_gettime(clock, &start);
+		drain_queue();
+		clock_gettime(clock, &end);
+		nsec = ((unsigned long long)(end.tv_sec - start.tv_sec) *
+			1000000000) + (end.tv_nsec - start.tv_nsec);
+		printf("done.\t\t%lld.%llds\n", nsec / 1000000000,
+		       nsec % 1000000000);
+	}
+	return 0;
+}
+
+void increase_limits(void)
+{
+	cur_limits.rlim_cur = RLIM_INFINITY;
+	cur_limits.rlim_max = RLIM_INFINITY;
+	setr(RLIMIT_MSGQUEUE, &cur_limits);
+	while (try_set(max_msgs, cur_max_msgs += 10))
+		;
+	cur_max_msgs = get(max_msgs);
+	while (try_set(max_msgsize, cur_max_msgsize += 1024))
+		;
+	cur_max_msgsize = get(max_msgsize);
+	if (setpriority(PRIO_PROCESS, 0, -20) != 0)
+		shutdown(2, "setpriority()", __LINE__);
+	cur_nice = -20;
+}
+
+int main(int argc, char *argv[])
+{
+	struct mq_attr attr;
+	char *option, *next_option;
+	int i, cpu;
+	struct sigaction sa;
+	poptContext popt_context;
+	char rc;
+	void *retval;
+
+	main_thread = pthread_self();
+	num_cpus_to_pin = 0;
+
+	if (sysconf(_SC_NPROCESSORS_ONLN) == -1) {
+		perror("sysconf(_SC_NPROCESSORS_ONLN)");
+		exit(1);
+	}
+	cpus_online = min(MAX_CPUS, sysconf(_SC_NPROCESSORS_ONLN));
+	cpu_set = CPU_ALLOC(cpus_online);
+	if (cpu_set == NULL) {
+		perror("CPU_ALLOC()");
+		exit(1);
+	}
+	cpu_set_size = CPU_ALLOC_SIZE(cpus_online);
+	CPU_ZERO_S(cpu_set_size, cpu_set);
+
+	popt_context = poptGetContext(NULL, argc, (const char **)argv,
+				      options, 0);
+
+	while ((rc = poptGetNextOpt(popt_context)) > 0) {
+		switch (rc) {
+		case 'c':
+			continuous_mode = 1;
+			option = cpu_option_string;
+			do {
+				next_option = strchr(option, ',');
+				if (next_option)
+					*next_option = '\0';
+				cpu = atoi(option);
+				if (cpu >= cpus_online)
+					fprintf(stderr, "CPU %d exceeds "
+						"cpus online, ignoring.\n",
+						cpu);
+				else
+					cpus_to_pin[num_cpus_to_pin++] = cpu;
+				if (next_option)
+					option = ++next_option;
+			} while (next_option && num_cpus_to_pin < MAX_CPUS);
+			/* Double check that they didn't give us the same CPU
+			 * more than once */
+			for (cpu = 0; cpu < num_cpus_to_pin; cpu++) {
+				if (CPU_ISSET_S(cpus_to_pin[cpu], cpu_set_size,
+						cpu_set)) {
+					fprintf(stderr, "Any given CPU may "
+						"only be given once.\n");
+					exit(1);
+				} else
+					CPU_SET_S(cpus_to_pin[cpu],
+						  cpu_set_size, cpu_set);
+			}
+			break;
+		case 'p':
+			/*
+			 * Although we can create a msg queue with a
+			 * non-absolute path name, unlink will fail.  So,
+			 * if the name doesn't start with a /, add one
+			 * when we save it.
+			 */
+			option = queue_path;
+			if (*option != '/') {
+				queue_path = malloc(strlen(option) + 2);
+				if (!queue_path) {
+					perror("malloc()");
+					exit(1);
+				}
+				queue_path[0] = '/';
+				queue_path[1] = 0;
+				strcat(queue_path, option);
+				free(option);
+			}
+			break;
+		}
+	}
+
+	if (continuous_mode && num_cpus_to_pin == 0) {
+		fprintf(stderr, "Must pass at least one CPU to continuous "
+			"mode.\n");
+		poptPrintUsage(popt_context, stderr, 0);
+		exit(1);
+	} else if (!continuous_mode) {
+		num_cpus_to_pin = 1;
+		cpus_to_pin[0] = cpus_online - 1;
+	}
+
+	if (getuid() != 0) {
+		fprintf(stderr, "Not running as root, but almost all tests "
+			"require root in order to modify\nsystem settings.  "
+			"Exiting.\n");
+		exit(1);
+	}
+
+	max_msgs = fopen(MAX_MSGS, "r+");
+	max_msgsize = fopen(MAX_MSGSIZE, "r+");
+	if (!max_msgs)
+		shutdown(2, "Failed to open msg_max", __LINE__);
+	if (!max_msgsize)
+		shutdown(2, "Failed to open msgsize_max", __LINE__);
+
+	/* Load up the current system values for everything we can */
+	getr(RLIMIT_MSGQUEUE, &saved_limits);
+	cur_limits = saved_limits;
+	saved_max_msgs = cur_max_msgs = get(max_msgs);
+	saved_max_msgsize = cur_max_msgsize = get(max_msgsize);
+	errno = 0;
+	cur_nice = getpriority(PRIO_PROCESS, 0);
+	if (errno)
+		shutdown(2, "getpriority()", __LINE__);
+
+	/* Tell the user our initial state */
+	printf("\nInitial system state:\n");
+	printf("\tUsing queue path:\t\t\t%s\n", queue_path);
+	printf("\tRLIMIT_MSGQUEUE(soft):\t\t\t%d\n", saved_limits.rlim_cur);
+	printf("\tRLIMIT_MSGQUEUE(hard):\t\t\t%d\n", saved_limits.rlim_max);
+	printf("\tMaximum Message Size:\t\t\t%d\n", saved_max_msgsize);
+	printf("\tMaximum Queue Size:\t\t\t%d\n", saved_max_msgs);
+	printf("\tNice value:\t\t\t\t%d\n", cur_nice);
+	printf("\n");
+
+	increase_limits();
+
+	printf("Adjusted system state for testing:\n");
+	if (cur_limits.rlim_cur == RLIM_INFINITY) {
+		printf("\tRLIMIT_MSGQUEUE(soft):\t\t\t(unlimited)\n");
+		printf("\tRLIMIT_MSGQUEUE(hard):\t\t\t(unlimited)\n");
+	} else {
+		printf("\tRLIMIT_MSGQUEUE(soft):\t\t\t%d\n",
+		       cur_limits.rlim_cur);
+		printf("\tRLIMIT_MSGQUEUE(hard):\t\t\t%d\n",
+		       cur_limits.rlim_max);
+	}
+	printf("\tMaximum Message Size:\t\t\t%d\n", cur_max_msgsize);
+	printf("\tMaximum Queue Size:\t\t\t%d\n", cur_max_msgs);
+	printf("\tNice value:\t\t\t\t%d\n", cur_nice);
+	printf("\tContinuous mode:\t\t\t(%s)\n", continuous_mode ?
+	       (continuous_mode_fake ? "fake mode" : "enabled") :
+	       "disabled");
+	printf("\tCPUs to pin:\t\t\t\t%d", cpus_to_pin[0]);
+	for (cpu = 1; cpu < num_cpus_to_pin; cpu++)
+			printf(",%d", cpus_to_pin[cpu]);
+	printf("\n");
+
+	sa.sa_sigaction = sig_action_SIGUSR1;
+	sigemptyset(&sa.sa_mask);
+	sigaddset(&sa.sa_mask, SIGHUP);
+	sigaddset(&sa.sa_mask, SIGINT);
+	sigaddset(&sa.sa_mask, SIGQUIT);
+	sigaddset(&sa.sa_mask, SIGTERM);
+	sa.sa_flags = SA_SIGINFO;
+	if (sigaction(SIGUSR1, &sa, NULL) == -1)
+		shutdown(1, "sigaction(SIGUSR1)", __LINE__);
+	sa.sa_sigaction = sig_action;
+	if (sigaction(SIGHUP, &sa, NULL) == -1)
+		shutdown(1, "sigaction(SIGHUP)", __LINE__);
+	if (sigaction(SIGINT, &sa, NULL) == -1)
+		shutdown(1, "sigaction(SIGINT)", __LINE__);
+	if (sigaction(SIGQUIT, &sa, NULL) == -1)
+		shutdown(1, "sigaction(SIGQUIT)", __LINE__);
+	if (sigaction(SIGTERM, &sa, NULL) == -1)
+		shutdown(1, "sigaction(SIGTERM)", __LINE__);
+
+	if (!continuous_mode_fake) {
+		attr.mq_flags = O_NONBLOCK;
+		attr.mq_maxmsg = cur_max_msgs;
+		attr.mq_msgsize = MSG_SIZE;
+		open_queue(&attr);
+	}
+	for (i = 0; i < num_cpus_to_pin; i++) {
+		pthread_attr_t thread_attr;
+		void *thread_func;
+
+		if (continuous_mode_fake)
+			thread_func = &fake_cont_thread;
+		else if (continuous_mode)
+			thread_func = &cont_thread;
+		else
+			thread_func = &perf_test_thread;
+
+		CPU_ZERO_S(cpu_set_size, cpu_set);
+		CPU_SET_S(cpus_to_pin[i], cpu_set_size, cpu_set);
+		pthread_attr_init(&thread_attr);
+		pthread_attr_setaffinity_np(&thread_attr, cpu_set_size,
+					    cpu_set);
+		if (pthread_create(&cpu_threads[i], &thread_attr, thread_func,
+				   NULL))
+			shutdown(1, "pthread_create()", __LINE__);
+		pthread_attr_destroy(&thread_attr);
+	}
+
+	if (!continuous_mode) {
+		pthread_join(cpu_threads[0], &retval);
+		shutdown((long)retval, "perf_test_thread()", __LINE__);
+	} else {
+		while (1)
+			sleep(1);
+	}
+	shutdown(0, "", 0);
+}

tools/testing/selftests/net/.gitignore

@@ -0,0 +1,3 @@
+socket
+psock_fanout
+psock_tpacket

tools/testing/selftests/net/Makefile

@@ -0,0 +1,19 @@
+# Makefile for net selftests
+
+CC = $(CROSS_COMPILE)gcc
+CFLAGS = -Wall -O2 -g
+
+CFLAGS += -I../../../../usr/include/
+
+NET_PROGS = socket psock_fanout psock_tpacket
+
+all: $(NET_PROGS)
+%: %.c
+	$(CC) $(CFLAGS) -o $@ $^
+
+run_tests: all
+	@/bin/sh ./run_netsocktests || echo "sockettests: [FAIL]"
+	@/bin/sh ./run_afpackettests || echo "afpackettests: [FAIL]"
+
+clean:
+	$(RM) $(NET_PROGS)

tools/testing/selftests/net/psock_fanout.c

@@ -0,0 +1,312 @@
+/*
+ * Copyright 2013 Google Inc.
+ * Author: Willem de Bruijn (willemb@google.com)
+ *
+ * A basic test of packet socket fanout behavior.
+ *
+ * Control:
+ * - create fanout fails as expected with illegal flag combinations
+ * - join   fanout fails as expected with diverging types or flags
+ *
+ * Datapath:
+ *   Open a pair of packet sockets and a pair of INET sockets, send a known
+ *   number of packets across the two INET sockets and count the number of
+ *   packets enqueued onto the two packet sockets.
+ *
+ *   The test currently runs for
+ *   - PACKET_FANOUT_HASH
+ *   - PACKET_FANOUT_HASH with PACKET_FANOUT_FLAG_ROLLOVER
+ *   - PACKET_FANOUT_LB
+ *   - PACKET_FANOUT_CPU
+ *   - PACKET_FANOUT_ROLLOVER
+ *
+ * Todo:
+ * - functionality: PACKET_FANOUT_FLAG_DEFRAG
+ *
+ * License (GPLv2):
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. * See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#define _GNU_SOURCE		/* for sched_setaffinity */
+
+#include <arpa/inet.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <linux/filter.h>
+#include <linux/if_packet.h>
+#include <net/ethernet.h>
+#include <netinet/ip.h>
+#include <netinet/udp.h>
+#include <poll.h>
+#include <sched.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/mman.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <unistd.h>
+
+#include "psock_lib.h"
+
+#define RING_NUM_FRAMES			20
+
+/* Open a socket in a given fanout mode.
+ * @return -1 if mode is bad, a valid socket otherwise */
+static int sock_fanout_open(uint16_t typeflags, int num_packets)
+{
+	int fd, val;
+
+	fd = socket(PF_PACKET, SOCK_DGRAM, htons(ETH_P_IP));
+	if (fd < 0) {
+		perror("socket packet");
+		exit(1);
+	}
+
+	/* fanout group ID is always 0: tests whether old groups are deleted */
+	val = ((int) typeflags) << 16;
+	if (setsockopt(fd, SOL_PACKET, PACKET_FANOUT, &val, sizeof(val))) {
+		if (close(fd)) {
+			perror("close packet");
+			exit(1);
+		}
+		return -1;
+	}
+
+	pair_udp_setfilter(fd);
+	return fd;
+}
+
+static char *sock_fanout_open_ring(int fd)
+{
+	struct tpacket_req req = {
+		.tp_block_size = getpagesize(),
+		.tp_frame_size = getpagesize(),
+		.tp_block_nr   = RING_NUM_FRAMES,
+		.tp_frame_nr   = RING_NUM_FRAMES,
+	};
+	char *ring;
+	int val = TPACKET_V2;
+
+	if (setsockopt(fd, SOL_PACKET, PACKET_VERSION, (void *) &val,
+		       sizeof(val))) {
+		perror("packetsock ring setsockopt version");
+		exit(1);
+	}
+	if (setsockopt(fd, SOL_PACKET, PACKET_RX_RING, (void *) &req,
+		       sizeof(req))) {
+		perror("packetsock ring setsockopt");
+		exit(1);
+	}
+
+	ring = mmap(0, req.tp_block_size * req.tp_block_nr,
+		    PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
+	if (!ring) {
+		fprintf(stderr, "packetsock ring mmap\n");
+		exit(1);
+	}
+
+	return ring;
+}
+
+static int sock_fanout_read_ring(int fd, void *ring)
+{
+	struct tpacket2_hdr *header = ring;
+	int count = 0;
+
+	while (header->tp_status & TP_STATUS_USER && count < RING_NUM_FRAMES) {
+		count++;
+		header = ring + (count * getpagesize());
+	}
+
+	return count;
+}
+
+static int sock_fanout_read(int fds[], char *rings[], const int expect[])
+{
+	int ret[2];
+
+	ret[0] = sock_fanout_read_ring(fds[0], rings[0]);
+	ret[1] = sock_fanout_read_ring(fds[1], rings[1]);
+
+	fprintf(stderr, "info: count=%d,%d, expect=%d,%d\n",
+			ret[0], ret[1], expect[0], expect[1]);
+
+	if ((!(ret[0] == expect[0] && ret[1] == expect[1])) &&
+	    (!(ret[0] == expect[1] && ret[1] == expect[0]))) {
+		fprintf(stderr, "ERROR: incorrect queue lengths\n");
+		return 1;
+	}
+
+	return 0;
+}
+
+/* Test illegal mode + flag combination */
+static void test_control_single(void)
+{
+	fprintf(stderr, "test: control single socket\n");
+
+	if (sock_fanout_open(PACKET_FANOUT_ROLLOVER |
+			       PACKET_FANOUT_FLAG_ROLLOVER, 0) != -1) {
+		fprintf(stderr, "ERROR: opened socket with dual rollover\n");
+		exit(1);
+	}
+}
+
+/* Test illegal group with different modes or flags */
+static void test_control_group(void)
+{
+	int fds[2];
+
+	fprintf(stderr, "test: control multiple sockets\n");
+
+	fds[0] = sock_fanout_open(PACKET_FANOUT_HASH, 20);
+	if (fds[0] == -1) {
+		fprintf(stderr, "ERROR: failed to open HASH socket\n");
+		exit(1);
+	}
+	if (sock_fanout_open(PACKET_FANOUT_HASH |
+			       PACKET_FANOUT_FLAG_DEFRAG, 10) != -1) {
+		fprintf(stderr, "ERROR: joined group with wrong flag defrag\n");
+		exit(1);
+	}
+	if (sock_fanout_open(PACKET_FANOUT_HASH |
+			       PACKET_FANOUT_FLAG_ROLLOVER, 10) != -1) {
+		fprintf(stderr, "ERROR: joined group with wrong flag ro\n");
+		exit(1);
+	}
+	if (sock_fanout_open(PACKET_FANOUT_CPU, 10) != -1) {
+		fprintf(stderr, "ERROR: joined group with wrong mode\n");
+		exit(1);
+	}
+	fds[1] = sock_fanout_open(PACKET_FANOUT_HASH, 20);
+	if (fds[1] == -1) {
+		fprintf(stderr, "ERROR: failed to join group\n");
+		exit(1);
+	}
+	if (close(fds[1]) || close(fds[0])) {
+		fprintf(stderr, "ERROR: closing sockets\n");
+		exit(1);
+	}
+}
+
+static int test_datapath(uint16_t typeflags, int port_off,
+			 const int expect1[], const int expect2[])
+{
+	const int expect0[] = { 0, 0 };
+	char *rings[2];
+	int fds[2], fds_udp[2][2], ret;
+
+	fprintf(stderr, "test: datapath 0x%hx\n", typeflags);
+
+	fds[0] = sock_fanout_open(typeflags, 20);
+	fds[1] = sock_fanout_open(typeflags, 20);
+	if (fds[0] == -1 || fds[1] == -1) {
+		fprintf(stderr, "ERROR: failed open\n");
+		exit(1);
+	}
+	rings[0] = sock_fanout_open_ring(fds[0]);
+	rings[1] = sock_fanout_open_ring(fds[1]);
+	pair_udp_open(fds_udp[0], PORT_BASE);
+	pair_udp_open(fds_udp[1], PORT_BASE + port_off);
+	sock_fanout_read(fds, rings, expect0);
+
+	/* Send data, but not enough to overflow a queue */
+	pair_udp_send(fds_udp[0], 15);
+	pair_udp_send(fds_udp[1], 5);
+	ret = sock_fanout_read(fds, rings, expect1);
+
+	/* Send more data, overflow the queue */
+	pair_udp_send(fds_udp[0], 15);
+	/* TODO: ensure consistent order between expect1 and expect2 */
+	ret |= sock_fanout_read(fds, rings, expect2);
+
+	if (munmap(rings[1], RING_NUM_FRAMES * getpagesize()) ||
+	    munmap(rings[0], RING_NUM_FRAMES * getpagesize())) {
+		fprintf(stderr, "close rings\n");
+		exit(1);
+	}
+	if (close(fds_udp[1][1]) || close(fds_udp[1][0]) ||
+	    close(fds_udp[0][1]) || close(fds_udp[0][0]) ||
+	    close(fds[1]) || close(fds[0])) {
+		fprintf(stderr, "close datapath\n");
+		exit(1);
+	}
+
+	return ret;
+}
+
+static int set_cpuaffinity(int cpuid)
+{
+	cpu_set_t mask;
+
+	CPU_ZERO(&mask);
+	CPU_SET(cpuid, &mask);
+	if (sched_setaffinity(0, sizeof(mask), &mask)) {
+		if (errno != EINVAL) {
+			fprintf(stderr, "setaffinity %d\n", cpuid);
+			exit(1);
+		}
+		return 1;
+	}
+
+	return 0;
+}
+
+int main(int argc, char **argv)
+{
+	const int expect_hash[2][2]	= { { 15, 5 },  { 20, 5 } };
+	const int expect_hash_rb[2][2]	= { { 15, 5 },  { 20, 15 } };
+	const int expect_lb[2][2]	= { { 10, 10 }, { 18, 17 } };
+	const int expect_rb[2][2]	= { { 20, 0 },  { 20, 15 } };
+	const int expect_cpu0[2][2]	= { { 20, 0 },  { 20, 0 } };
+	const int expect_cpu1[2][2]	= { { 0, 20 },  { 0, 20 } };
+	int port_off = 2, tries = 5, ret;
+
+	test_control_single();
+	test_control_group();
+
+	/* find a set of ports that do not collide onto the same socket */
+	ret = test_datapath(PACKET_FANOUT_HASH, port_off,
+			    expect_hash[0], expect_hash[1]);
+	while (ret && tries--) {
+		fprintf(stderr, "info: trying alternate ports (%d)\n", tries);
+		ret = test_datapath(PACKET_FANOUT_HASH, ++port_off,
+				    expect_hash[0], expect_hash[1]);
+	}
+
+	ret |= test_datapath(PACKET_FANOUT_HASH | PACKET_FANOUT_FLAG_ROLLOVER,
+			     port_off, expect_hash_rb[0], expect_hash_rb[1]);
+	ret |= test_datapath(PACKET_FANOUT_LB,
+			     port_off, expect_lb[0], expect_lb[1]);
+	ret |= test_datapath(PACKET_FANOUT_ROLLOVER,
+			     port_off, expect_rb[0], expect_rb[1]);
+
+	set_cpuaffinity(0);
+	ret |= test_datapath(PACKET_FANOUT_CPU, port_off,
+			     expect_cpu0[0], expect_cpu0[1]);
+	if (!set_cpuaffinity(1))
+		/* TODO: test that choice alternates with previous */
+		ret |= test_datapath(PACKET_FANOUT_CPU, port_off,
+				     expect_cpu1[0], expect_cpu1[1]);
+
+	if (ret)
+		return 1;
+
+	printf("OK. All tests passed\n");
+	return 0;
+}

tools/testing/selftests/net/psock_lib.h

@@ -0,0 +1,127 @@
+/*
+ * Copyright 2013 Google Inc.
+ * Author: Willem de Bruijn <willemb@google.com>
+ *         Daniel Borkmann <dborkman@redhat.com>
+ *
+ * License (GPLv2):
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. * See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#ifndef PSOCK_LIB_H
+#define PSOCK_LIB_H
+
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <string.h>
+#include <arpa/inet.h>
+#include <unistd.h>
+
+#define DATA_LEN			100
+#define DATA_CHAR			'a'
+
+#define PORT_BASE			8000
+
+#ifndef __maybe_unused
+# define __maybe_unused		__attribute__ ((__unused__))
+#endif
+
+static __maybe_unused void pair_udp_setfilter(int fd)
+{
+	struct sock_filter bpf_filter[] = {
+		{ 0x80, 0, 0, 0x00000000 },  /* LD  pktlen		      */
+		{ 0x35, 0, 5, DATA_LEN   },  /* JGE DATA_LEN  [f goto nomatch]*/
+		{ 0x30, 0, 0, 0x00000050 },  /* LD  ip[80]		      */
+		{ 0x15, 0, 3, DATA_CHAR  },  /* JEQ DATA_CHAR [f goto nomatch]*/
+		{ 0x30, 0, 0, 0x00000051 },  /* LD  ip[81]		      */
+		{ 0x15, 0, 1, DATA_CHAR  },  /* JEQ DATA_CHAR [f goto nomatch]*/
+		{ 0x06, 0, 0, 0x00000060 },  /* RET match	              */
+		{ 0x06, 0, 0, 0x00000000 },  /* RET no match		      */
+	};
+	struct sock_fprog bpf_prog;
+
+	bpf_prog.filter = bpf_filter;
+	bpf_prog.len = sizeof(bpf_filter) / sizeof(struct sock_filter);
+	if (setsockopt(fd, SOL_SOCKET, SO_ATTACH_FILTER, &bpf_prog,
+		       sizeof(bpf_prog))) {
+		perror("setsockopt SO_ATTACH_FILTER");
+		exit(1);
+	}
+}
+
+static __maybe_unused void pair_udp_open(int fds[], uint16_t port)
+{
+	struct sockaddr_in saddr, daddr;
+
+	fds[0] = socket(PF_INET, SOCK_DGRAM, 0);
+	fds[1] = socket(PF_INET, SOCK_DGRAM, 0);
+	if (fds[0] == -1 || fds[1] == -1) {
+		fprintf(stderr, "ERROR: socket dgram\n");
+		exit(1);
+	}
+
+	memset(&saddr, 0, sizeof(saddr));
+	saddr.sin_family = AF_INET;
+	saddr.sin_port = htons(port);
+	saddr.sin_addr.s_addr = htonl(INADDR_LOOPBACK);
+
+	memset(&daddr, 0, sizeof(daddr));
+	daddr.sin_family = AF_INET;
+	daddr.sin_port = htons(port + 1);
+	daddr.sin_addr.s_addr = htonl(INADDR_LOOPBACK);
+
+	/* must bind both to get consistent hash result */
+	if (bind(fds[1], (void *) &daddr, sizeof(daddr))) {
+		perror("bind");
+		exit(1);
+	}
+	if (bind(fds[0], (void *) &saddr, sizeof(saddr))) {
+		perror("bind");
+		exit(1);
+	}
+	if (connect(fds[0], (void *) &daddr, sizeof(daddr))) {
+		perror("connect");
+		exit(1);
+	}
+}
+
+static __maybe_unused void pair_udp_send(int fds[], int num)
+{
+	char buf[DATA_LEN], rbuf[DATA_LEN];
+
+	memset(buf, DATA_CHAR, sizeof(buf));
+	while (num--) {
+		/* Should really handle EINTR and EAGAIN */
+		if (write(fds[0], buf, sizeof(buf)) != sizeof(buf)) {
+			fprintf(stderr, "ERROR: send failed left=%d\n", num);
+			exit(1);
+		}
+		if (read(fds[1], rbuf, sizeof(rbuf)) != sizeof(rbuf)) {
+			fprintf(stderr, "ERROR: recv failed left=%d\n", num);
+			exit(1);
+		}
+		if (memcmp(buf, rbuf, sizeof(buf))) {
+			fprintf(stderr, "ERROR: data failed left=%d\n", num);
+			exit(1);
+		}
+	}
+}
+
+static __maybe_unused void pair_udp_close(int fds[])
+{
+	close(fds[0]);
+	close(fds[1]);
+}
+
+#endif /* PSOCK_LIB_H */

tools/testing/selftests/net/psock_tpacket.c

@@ -0,0 +1,824 @@
+/*
+ * Copyright 2013 Red Hat, Inc.
+ * Author: Daniel Borkmann <dborkman@redhat.com>
+ *
+ * A basic test of packet socket's TPACKET_V1/TPACKET_V2/TPACKET_V3 behavior.
+ *
+ * Control:
+ *   Test the setup of the TPACKET socket with different patterns that are
+ *   known to fail (TODO) resp. succeed (OK).
+ *
+ * Datapath:
+ *   Open a pair of packet sockets and send resp. receive an a priori known
+ *   packet pattern accross the sockets and check if it was received resp.
+ *   sent correctly. Fanout in combination with RX_RING is currently not
+ *   tested here.
+ *
+ *   The test currently runs for
+ *   - TPACKET_V1: RX_RING, TX_RING
+ *   - TPACKET_V2: RX_RING, TX_RING
+ *   - TPACKET_V3: RX_RING
+ *
+ * License (GPLv2):
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. * See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/socket.h>
+#include <sys/mman.h>
+#include <linux/if_packet.h>
+#include <linux/filter.h>
+#include <ctype.h>
+#include <fcntl.h>
+#include <unistd.h>
+#include <bits/wordsize.h>
+#include <net/ethernet.h>
+#include <netinet/ip.h>
+#include <arpa/inet.h>
+#include <stdint.h>
+#include <string.h>
+#include <assert.h>
+#include <net/if.h>
+#include <inttypes.h>
+#include <poll.h>
+
+#include "psock_lib.h"
+
+#ifndef bug_on
+# define bug_on(cond)		assert(!(cond))
+#endif
+
+#ifndef __aligned_tpacket
+# define __aligned_tpacket	__attribute__((aligned(TPACKET_ALIGNMENT)))
+#endif
+
+#ifndef __align_tpacket
+# define __align_tpacket(x)	__attribute__((aligned(TPACKET_ALIGN(x))))
+#endif
+
+#define BLOCK_STATUS(x)		((x)->h1.block_status)
+#define BLOCK_NUM_PKTS(x)	((x)->h1.num_pkts)
+#define BLOCK_O2FP(x)		((x)->h1.offset_to_first_pkt)
+#define BLOCK_LEN(x)		((x)->h1.blk_len)
+#define BLOCK_SNUM(x)		((x)->h1.seq_num)
+#define BLOCK_O2PRIV(x)		((x)->offset_to_priv)
+#define BLOCK_PRIV(x)		((void *) ((uint8_t *) (x) + BLOCK_O2PRIV(x)))
+#define BLOCK_HDR_LEN		(ALIGN_8(sizeof(struct block_desc)))
+#define ALIGN_8(x)		(((x) + 8 - 1) & ~(8 - 1))
+#define BLOCK_PLUS_PRIV(sz_pri)	(BLOCK_HDR_LEN + ALIGN_8((sz_pri)))
+
+#define NUM_PACKETS		100
+
+struct ring {
+	struct iovec *rd;
+	uint8_t *mm_space;
+	size_t mm_len, rd_len;
+	struct sockaddr_ll ll;
+	void (*walk)(int sock, struct ring *ring);
+	int type, rd_num, flen, version;
+	union {
+		struct tpacket_req  req;
+		struct tpacket_req3 req3;
+	};
+};
+
+struct block_desc {
+	uint32_t version;
+	uint32_t offset_to_priv;
+	struct tpacket_hdr_v1 h1;
+};
+
+union frame_map {
+	struct {
+		struct tpacket_hdr tp_h __aligned_tpacket;
+		struct sockaddr_ll s_ll __align_tpacket(sizeof(struct tpacket_hdr));
+	} *v1;
+	struct {
+		struct tpacket2_hdr tp_h __aligned_tpacket;
+		struct sockaddr_ll s_ll __align_tpacket(sizeof(struct tpacket2_hdr));
+	} *v2;
+	void *raw;
+};
+
+static unsigned int total_packets, total_bytes;
+
+static int pfsocket(int ver)
+{
+	int ret, sock = socket(PF_PACKET, SOCK_RAW, htons(ETH_P_ALL));
+	if (sock == -1) {
+		perror("socket");
+		exit(1);
+	}
+
+	ret = setsockopt(sock, SOL_PACKET, PACKET_VERSION, &ver, sizeof(ver));
+	if (ret == -1) {
+		perror("setsockopt");
+		exit(1);
+	}
+
+	return sock;
+}
+
+static void status_bar_update(void)
+{
+	if (total_packets % 10 == 0) {
+		fprintf(stderr, ".");
+		fflush(stderr);
+	}
+}
+
+static void test_payload(void *pay, size_t len)
+{
+	struct ethhdr *eth = pay;
+
+	if (len < sizeof(struct ethhdr)) {
+		fprintf(stderr, "test_payload: packet too "
+			"small: %zu bytes!\n", len);
+		exit(1);
+	}
+
+	if (eth->h_proto != htons(ETH_P_IP)) {
+		fprintf(stderr, "test_payload: wrong ethernet "
+			"type: 0x%x!\n", ntohs(eth->h_proto));
+		exit(1);
+	}
+}
+
+static void create_payload(void *pay, size_t *len)
+{
+	int i;
+	struct ethhdr *eth = pay;
+	struct iphdr *ip = pay + sizeof(*eth);
+
+	/* Lets create some broken crap, that still passes
+	 * our BPF filter.
+	 */
+
+	*len = DATA_LEN + 42;
+
+	memset(pay, 0xff, ETH_ALEN * 2);
+	eth->h_proto = htons(ETH_P_IP);
+
+	for (i = 0; i < sizeof(*ip); ++i)
+		((uint8_t *) pay)[i + sizeof(*eth)] = (uint8_t) rand();
+
+	ip->ihl = 5;
+	ip->version = 4;
+	ip->protocol = 0x11;
+	ip->frag_off = 0;
+	ip->ttl = 64;
+	ip->tot_len = htons((uint16_t) *len - sizeof(*eth));
+
+	ip->saddr = htonl(INADDR_LOOPBACK);
+	ip->daddr = htonl(INADDR_LOOPBACK);
+
+	memset(pay + sizeof(*eth) + sizeof(*ip),
+	       DATA_CHAR, DATA_LEN);
+}
+
+static inline int __v1_rx_kernel_ready(struct tpacket_hdr *hdr)
+{
+	return ((hdr->tp_status & TP_STATUS_USER) == TP_STATUS_USER);
+}
+
+static inline void __v1_rx_user_ready(struct tpacket_hdr *hdr)
+{
+	hdr->tp_status = TP_STATUS_KERNEL;
+	__sync_synchronize();
+}
+
+static inline int __v2_rx_kernel_ready(struct tpacket2_hdr *hdr)
+{
+	return ((hdr->tp_status & TP_STATUS_USER) == TP_STATUS_USER);
+}
+
+static inline void __v2_rx_user_ready(struct tpacket2_hdr *hdr)
+{
+	hdr->tp_status = TP_STATUS_KERNEL;
+	__sync_synchronize();
+}
+
+static inline int __v1_v2_rx_kernel_ready(void *base, int version)
+{
+	switch (version) {
+	case TPACKET_V1:
+		return __v1_rx_kernel_ready(base);
+	case TPACKET_V2:
+		return __v2_rx_kernel_ready(base);
+	default:
+		bug_on(1);
+		return 0;
+	}
+}
+
+static inline void __v1_v2_rx_user_ready(void *base, int version)
+{
+	switch (version) {
+	case TPACKET_V1:
+		__v1_rx_user_ready(base);
+		break;
+	case TPACKET_V2:
+		__v2_rx_user_ready(base);
+		break;
+	}
+}
+
+static void walk_v1_v2_rx(int sock, struct ring *ring)
+{
+	struct pollfd pfd;
+	int udp_sock[2];
+	union frame_map ppd;
+	unsigned int frame_num = 0;
+
+	bug_on(ring->type != PACKET_RX_RING);
+
+	pair_udp_open(udp_sock, PORT_BASE);
+	pair_udp_setfilter(sock);
+
+	memset(&pfd, 0, sizeof(pfd));
+	pfd.fd = sock;
+	pfd.events = POLLIN | POLLERR;
+	pfd.revents = 0;
+
+	pair_udp_send(udp_sock, NUM_PACKETS);
+
+	while (total_packets < NUM_PACKETS * 2) {
+		while (__v1_v2_rx_kernel_ready(ring->rd[frame_num].iov_base,
+					       ring->version)) {
+			ppd.raw = ring->rd[frame_num].iov_base;
+
+			switch (ring->version) {
+			case TPACKET_V1:
+				test_payload((uint8_t *) ppd.raw + ppd.v1->tp_h.tp_mac,
+					     ppd.v1->tp_h.tp_snaplen);
+				total_bytes += ppd.v1->tp_h.tp_snaplen;
+				break;
+
+			case TPACKET_V2:
+				test_payload((uint8_t *) ppd.raw + ppd.v2->tp_h.tp_mac,
+					     ppd.v2->tp_h.tp_snaplen);
+				total_bytes += ppd.v2->tp_h.tp_snaplen;
+				break;
+			}
+
+			status_bar_update();
+			total_packets++;
+
+			__v1_v2_rx_user_ready(ppd.raw, ring->version);
+
+			frame_num = (frame_num + 1) % ring->rd_num;
+		}
+
+		poll(&pfd, 1, 1);
+	}
+
+	pair_udp_close(udp_sock);
+
+	if (total_packets != 2 * NUM_PACKETS) {
+		fprintf(stderr, "walk_v%d_rx: received %u out of %u pkts\n",
+			ring->version, total_packets, NUM_PACKETS);
+		exit(1);
+	}
+
+	fprintf(stderr, " %u pkts (%u bytes)", NUM_PACKETS, total_bytes >> 1);
+}
+
+static inline int __v1_tx_kernel_ready(struct tpacket_hdr *hdr)
+{
+	return !(hdr->tp_status & (TP_STATUS_SEND_REQUEST | TP_STATUS_SENDING));
+}
+
+static inline void __v1_tx_user_ready(struct tpacket_hdr *hdr)
+{
+	hdr->tp_status = TP_STATUS_SEND_REQUEST;
+	__sync_synchronize();
+}
+
+static inline int __v2_tx_kernel_ready(struct tpacket2_hdr *hdr)
+{
+	return !(hdr->tp_status & (TP_STATUS_SEND_REQUEST | TP_STATUS_SENDING));
+}
+
+static inline void __v2_tx_user_ready(struct tpacket2_hdr *hdr)
+{
+	hdr->tp_status = TP_STATUS_SEND_REQUEST;
+	__sync_synchronize();
+}
+
+static inline int __v1_v2_tx_kernel_ready(void *base, int version)
+{
+	switch (version) {
+	case TPACKET_V1:
+		return __v1_tx_kernel_ready(base);
+	case TPACKET_V2:
+		return __v2_tx_kernel_ready(base);
+	default:
+		bug_on(1);
+		return 0;
+	}
+}
+
+static inline void __v1_v2_tx_user_ready(void *base, int version)
+{
+	switch (version) {
+	case TPACKET_V1:
+		__v1_tx_user_ready(base);
+		break;
+	case TPACKET_V2:
+		__v2_tx_user_ready(base);
+		break;
+	}
+}
+
+static void __v1_v2_set_packet_loss_discard(int sock)
+{
+	int ret, discard = 1;
+
+	ret = setsockopt(sock, SOL_PACKET, PACKET_LOSS, (void *) &discard,
+			 sizeof(discard));
+	if (ret == -1) {
+		perror("setsockopt");
+		exit(1);
+	}
+}
+
+static void walk_v1_v2_tx(int sock, struct ring *ring)
+{
+	struct pollfd pfd;
+	int rcv_sock, ret;
+	size_t packet_len;
+	union frame_map ppd;
+	char packet[1024];
+	unsigned int frame_num = 0, got = 0;
+	struct sockaddr_ll ll = {
+		.sll_family = PF_PACKET,
+		.sll_halen = ETH_ALEN,
+	};
+
+	bug_on(ring->type != PACKET_TX_RING);
+	bug_on(ring->rd_num < NUM_PACKETS);
+
+	rcv_sock = socket(PF_PACKET, SOCK_RAW, htons(ETH_P_ALL));
+	if (rcv_sock == -1) {
+		perror("socket");
+		exit(1);
+	}
+
+	pair_udp_setfilter(rcv_sock);
+
+	ll.sll_ifindex = if_nametoindex("lo");
+	ret = bind(rcv_sock, (struct sockaddr *) &ll, sizeof(ll));
+	if (ret == -1) {
+		perror("bind");
+		exit(1);
+	}
+
+	memset(&pfd, 0, sizeof(pfd));
+	pfd.fd = sock;
+	pfd.events = POLLOUT | POLLERR;
+	pfd.revents = 0;
+
+	total_packets = NUM_PACKETS;
+	create_payload(packet, &packet_len);
+
+	while (total_packets > 0) {
+		while (__v1_v2_tx_kernel_ready(ring->rd[frame_num].iov_base,
+					       ring->version) &&
+		       total_packets > 0) {
+			ppd.raw = ring->rd[frame_num].iov_base;
+
+			switch (ring->version) {
+			case TPACKET_V1:
+				ppd.v1->tp_h.tp_snaplen = packet_len;
+				ppd.v1->tp_h.tp_len = packet_len;
+
+				memcpy((uint8_t *) ppd.raw + TPACKET_HDRLEN -
+				       sizeof(struct sockaddr_ll), packet,
+				       packet_len);
+				total_bytes += ppd.v1->tp_h.tp_snaplen;
+				break;
+
+			case TPACKET_V2:
+				ppd.v2->tp_h.tp_snaplen = packet_len;
+				ppd.v2->tp_h.tp_len = packet_len;
+
+				memcpy((uint8_t *) ppd.raw + TPACKET2_HDRLEN -
+				       sizeof(struct sockaddr_ll), packet,
+				       packet_len);
+				total_bytes += ppd.v2->tp_h.tp_snaplen;
+				break;
+			}
+
+			status_bar_update();
+			total_packets--;
+
+			__v1_v2_tx_user_ready(ppd.raw, ring->version);
+
+			frame_num = (frame_num + 1) % ring->rd_num;
+		}
+
+		poll(&pfd, 1, 1);
+	}
+
+	bug_on(total_packets != 0);
+
+	ret = sendto(sock, NULL, 0, 0, NULL, 0);
+	if (ret == -1) {
+		perror("sendto");
+		exit(1);
+	}
+
+	while ((ret = recvfrom(rcv_sock, packet, sizeof(packet),
+			       0, NULL, NULL)) > 0 &&
+	       total_packets < NUM_PACKETS) {
+		got += ret;
+		test_payload(packet, ret);
+
+		status_bar_update();
+		total_packets++;
+	}
+
+	close(rcv_sock);
+
+	if (total_packets != NUM_PACKETS) {
+		fprintf(stderr, "walk_v%d_rx: received %u out of %u pkts\n",
+			ring->version, total_packets, NUM_PACKETS);
+		exit(1);
+	}
+
+	fprintf(stderr, " %u pkts (%u bytes)", NUM_PACKETS, got);
+}
+
+static void walk_v1_v2(int sock, struct ring *ring)
+{
+	if (ring->type == PACKET_RX_RING)
+		walk_v1_v2_rx(sock, ring);
+	else
+		walk_v1_v2_tx(sock, ring);
+}
+
+static uint64_t __v3_prev_block_seq_num = 0;
+
+void __v3_test_block_seq_num(struct block_desc *pbd)
+{
+	if (__v3_prev_block_seq_num + 1 != BLOCK_SNUM(pbd)) {
+		fprintf(stderr, "\nprev_block_seq_num:%"PRIu64", expected "
+			"seq:%"PRIu64" != actual seq:%"PRIu64"\n",
+			__v3_prev_block_seq_num, __v3_prev_block_seq_num + 1,
+			(uint64_t) BLOCK_SNUM(pbd));
+		exit(1);
+	}
+
+	__v3_prev_block_seq_num = BLOCK_SNUM(pbd);
+}
+
+static void __v3_test_block_len(struct block_desc *pbd, uint32_t bytes, int block_num)
+{
+	if (BLOCK_NUM_PKTS(pbd)) {
+		if (bytes != BLOCK_LEN(pbd)) {
+			fprintf(stderr, "\nblock:%u with %upackets, expected "
+				"len:%u != actual len:%u\n", block_num,
+				BLOCK_NUM_PKTS(pbd), bytes, BLOCK_LEN(pbd));
+			exit(1);
+		}
+	} else {
+		if (BLOCK_LEN(pbd) != BLOCK_PLUS_PRIV(13)) {
+			fprintf(stderr, "\nblock:%u, expected len:%lu != "
+				"actual len:%u\n", block_num, BLOCK_HDR_LEN,
+				BLOCK_LEN(pbd));
+			exit(1);
+		}
+	}
+}
+
+static void __v3_test_block_header(struct block_desc *pbd, const int block_num)
+{
+	uint32_t block_status = BLOCK_STATUS(pbd);
+
+	if ((block_status & TP_STATUS_USER) == 0) {
+		fprintf(stderr, "\nblock %u: not in TP_STATUS_USER\n", block_num);
+		exit(1);
+	}
+
+	__v3_test_block_seq_num(pbd);
+}
+
+static void __v3_walk_block(struct block_desc *pbd, const int block_num)
+{
+	int num_pkts = BLOCK_NUM_PKTS(pbd), i;
+	unsigned long bytes = 0;
+	unsigned long bytes_with_padding = BLOCK_PLUS_PRIV(13);
+	struct tpacket3_hdr *ppd;
+
+	__v3_test_block_header(pbd, block_num);
+
+	ppd = (struct tpacket3_hdr *) ((uint8_t *) pbd + BLOCK_O2FP(pbd));
+	for (i = 0; i < num_pkts; ++i) {
+		bytes += ppd->tp_snaplen;
+
+		if (ppd->tp_next_offset)
+			bytes_with_padding += ppd->tp_next_offset;
+		else
+			bytes_with_padding += ALIGN_8(ppd->tp_snaplen + ppd->tp_mac);
+
+		test_payload((uint8_t *) ppd + ppd->tp_mac, ppd->tp_snaplen);
+
+		status_bar_update();
+		total_packets++;
+
+		ppd = (struct tpacket3_hdr *) ((uint8_t *) ppd + ppd->tp_next_offset);
+		__sync_synchronize();
+	}
+
+	__v3_test_block_len(pbd, bytes_with_padding, block_num);
+	total_bytes += bytes;
+}
+
+void __v3_flush_block(struct block_desc *pbd)
+{
+	BLOCK_STATUS(pbd) = TP_STATUS_KERNEL;
+	__sync_synchronize();
+}
+
+static void walk_v3_rx(int sock, struct ring *ring)
+{
+	unsigned int block_num = 0;
+	struct pollfd pfd;
+	struct block_desc *pbd;
+	int udp_sock[2];
+
+	bug_on(ring->type != PACKET_RX_RING);
+
+	pair_udp_open(udp_sock, PORT_BASE);
+	pair_udp_setfilter(sock);
+
+	memset(&pfd, 0, sizeof(pfd));
+	pfd.fd = sock;
+	pfd.events = POLLIN | POLLERR;
+	pfd.revents = 0;
+
+	pair_udp_send(udp_sock, NUM_PACKETS);
+
+	while (total_packets < NUM_PACKETS * 2) {
+		pbd = (struct block_desc *) ring->rd[block_num].iov_base;
+
+		while ((BLOCK_STATUS(pbd) & TP_STATUS_USER) == 0)
+			poll(&pfd, 1, 1);
+
+		__v3_walk_block(pbd, block_num);
+		__v3_flush_block(pbd);
+
+		block_num = (block_num + 1) % ring->rd_num;
+	}
+
+	pair_udp_close(udp_sock);
+
+	if (total_packets != 2 * NUM_PACKETS) {
+		fprintf(stderr, "walk_v3_rx: received %u out of %u pkts\n",
+			total_packets, NUM_PACKETS);
+		exit(1);
+	}
+
+	fprintf(stderr, " %u pkts (%u bytes)", NUM_PACKETS, total_bytes >> 1);
+}
+
+static void walk_v3(int sock, struct ring *ring)
+{
+	if (ring->type == PACKET_RX_RING)
+		walk_v3_rx(sock, ring);
+	else
+		bug_on(1);
+}
+
+static void __v1_v2_fill(struct ring *ring, unsigned int blocks)
+{
+	ring->req.tp_block_size = getpagesize() << 2;
+	ring->req.tp_frame_size = TPACKET_ALIGNMENT << 7;
+	ring->req.tp_block_nr = blocks;
+
+	ring->req.tp_frame_nr = ring->req.tp_block_size /
+				ring->req.tp_frame_size *
+				ring->req.tp_block_nr;
+
+	ring->mm_len = ring->req.tp_block_size * ring->req.tp_block_nr;
+	ring->walk = walk_v1_v2;
+	ring->rd_num = ring->req.tp_frame_nr;
+	ring->flen = ring->req.tp_frame_size;
+}
+
+static void __v3_fill(struct ring *ring, unsigned int blocks)
+{
+	ring->req3.tp_retire_blk_tov = 64;
+	ring->req3.tp_sizeof_priv = 13;
+	ring->req3.tp_feature_req_word |= TP_FT_REQ_FILL_RXHASH;
+
+	ring->req3.tp_block_size = getpagesize() << 2;
+	ring->req3.tp_frame_size = TPACKET_ALIGNMENT << 7;
+	ring->req3.tp_block_nr = blocks;
+
+	ring->req3.tp_frame_nr = ring->req3.tp_block_size /
+				 ring->req3.tp_frame_size *
+				 ring->req3.tp_block_nr;
+
+	ring->mm_len = ring->req3.tp_block_size * ring->req3.tp_block_nr;
+	ring->walk = walk_v3;
+	ring->rd_num = ring->req3.tp_block_nr;
+	ring->flen = ring->req3.tp_block_size;
+}
+
+static void setup_ring(int sock, struct ring *ring, int version, int type)
+{
+	int ret = 0;
+	unsigned int blocks = 256;
+
+	ring->type = type;
+	ring->version = version;
+
+	switch (version) {
+	case TPACKET_V1:
+	case TPACKET_V2:
+		if (type == PACKET_TX_RING)
+			__v1_v2_set_packet_loss_discard(sock);
+		__v1_v2_fill(ring, blocks);
+		ret = setsockopt(sock, SOL_PACKET, type, &ring->req,
+				 sizeof(ring->req));
+		break;
+
+	case TPACKET_V3:
+		__v3_fill(ring, blocks);
+		ret = setsockopt(sock, SOL_PACKET, type, &ring->req3,
+				 sizeof(ring->req3));
+		break;
+	}
+
+	if (ret == -1) {
+		perror("setsockopt");
+		exit(1);
+	}
+
+	ring->rd_len = ring->rd_num * sizeof(*ring->rd);
+	ring->rd = malloc(ring->rd_len);
+	if (ring->rd == NULL) {
+		perror("malloc");
+		exit(1);
+	}
+
+	total_packets = 0;
+	total_bytes = 0;
+}
+
+static void mmap_ring(int sock, struct ring *ring)
+{
+	int i;
+
+	ring->mm_space = mmap(0, ring->mm_len, PROT_READ | PROT_WRITE,
+			      MAP_SHARED | MAP_LOCKED | MAP_POPULATE, sock, 0);
+	if (ring->mm_space == MAP_FAILED) {
+		perror("mmap");
+		exit(1);
+	}
+
+	memset(ring->rd, 0, ring->rd_len);
+	for (i = 0; i < ring->rd_num; ++i) {
+		ring->rd[i].iov_base = ring->mm_space + (i * ring->flen);
+		ring->rd[i].iov_len = ring->flen;
+	}
+}
+
+static void bind_ring(int sock, struct ring *ring)
+{
+	int ret;
+
+	ring->ll.sll_family = PF_PACKET;
+	ring->ll.sll_protocol = htons(ETH_P_ALL);
+	ring->ll.sll_ifindex = if_nametoindex("lo");
+	ring->ll.sll_hatype = 0;
+	ring->ll.sll_pkttype = 0;
+	ring->ll.sll_halen = 0;
+
+	ret = bind(sock, (struct sockaddr *) &ring->ll, sizeof(ring->ll));
+	if (ret == -1) {
+		perror("bind");
+		exit(1);
+	}
+}
+
+static void walk_ring(int sock, struct ring *ring)
+{
+	ring->walk(sock, ring);
+}
+
+static void unmap_ring(int sock, struct ring *ring)
+{
+	munmap(ring->mm_space, ring->mm_len);
+	free(ring->rd);
+}
+
+static int test_kernel_bit_width(void)
+{
+	char in[512], *ptr;
+	int num = 0, fd;
+	ssize_t ret;
+
+	fd = open("/proc/kallsyms", O_RDONLY);
+	if (fd == -1) {
+		perror("open");
+		exit(1);
+	}
+
+	ret = read(fd, in, sizeof(in));
+	if (ret <= 0) {
+		perror("read");
+		exit(1);
+	}
+
+	close(fd);
+
+	ptr = in;
+	while(!isspace(*ptr)) {
+		num++;
+		ptr++;
+	}
+
+	return num * 4;
+}
+
+static int test_user_bit_width(void)
+{
+	return __WORDSIZE;
+}
+
+static const char *tpacket_str[] = {
+	[TPACKET_V1] = "TPACKET_V1",
+	[TPACKET_V2] = "TPACKET_V2",
+	[TPACKET_V3] = "TPACKET_V3",
+};
+
+static const char *type_str[] = {
+	[PACKET_RX_RING] = "PACKET_RX_RING",
+	[PACKET_TX_RING] = "PACKET_TX_RING",
+};
+
+static int test_tpacket(int version, int type)
+{
+	int sock;
+	struct ring ring;
+
+	fprintf(stderr, "test: %s with %s ", tpacket_str[version],
+		type_str[type]);
+	fflush(stderr);
+
+	if (version == TPACKET_V1 &&
+	    test_kernel_bit_width() != test_user_bit_width()) {
+		fprintf(stderr, "test: skip %s %s since user and kernel "
+			"space have different bit width\n",
+			tpacket_str[version], type_str[type]);
+		return 0;
+	}
+
+	sock = pfsocket(version);
+	memset(&ring, 0, sizeof(ring));
+	setup_ring(sock, &ring, version, type);
+	mmap_ring(sock, &ring);
+	bind_ring(sock, &ring);
+	walk_ring(sock, &ring);
+	unmap_ring(sock, &ring);
+	close(sock);
+
+	fprintf(stderr, "\n");
+	return 0;
+}
+
+int main(void)
+{
+	int ret = 0;
+
+	ret |= test_tpacket(TPACKET_V1, PACKET_RX_RING);
+	ret |= test_tpacket(TPACKET_V1, PACKET_TX_RING);
+
+	ret |= test_tpacket(TPACKET_V2, PACKET_RX_RING);
+	ret |= test_tpacket(TPACKET_V2, PACKET_TX_RING);
+
+	ret |= test_tpacket(TPACKET_V3, PACKET_RX_RING);
+
+	if (ret)
+		return 1;
+
+	printf("OK. All tests passed\n");
+	return 0;
+}

tools/testing/selftests/net/run_afpackettests

@@ -0,0 +1,26 @@
+#!/bin/sh
+
+if [ $(id -u) != 0 ]; then
+	echo $msg must be run as root >&2
+	exit 0
+fi
+
+echo "--------------------"
+echo "running psock_fanout test"
+echo "--------------------"
+./psock_fanout
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+else
+	echo "[PASS]"
+fi
+
+echo "--------------------"
+echo "running psock_tpacket test"
+echo "--------------------"
+./psock_tpacket
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+else
+	echo "[PASS]"
+fi

tools/testing/selftests/net/run_netsocktests

@@ -0,0 +1,12 @@
+#!/bin/bash
+
+echo "--------------------"
+echo "running socket test"
+echo "--------------------"
+./socket
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+else
+	echo "[PASS]"
+fi
+

tools/testing/selftests/net/socket.c

@@ -0,0 +1,92 @@
+#include <stdio.h>
+#include <errno.h>
+#include <unistd.h>
+#include <string.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <netinet/in.h>
+
+struct socket_testcase {
+	int	domain;
+	int	type;
+	int	protocol;
+
+	/* 0    = valid file descriptor
+	 * -foo = error foo
+	 */
+	int	expect;
+
+	/* If non-zero, accept EAFNOSUPPORT to handle the case
+	 * of the protocol not being configured into the kernel.
+	 */
+	int	nosupport_ok;
+};
+
+static struct socket_testcase tests[] = {
+	{ AF_MAX,  0,           0,           -EAFNOSUPPORT,    0 },
+	{ AF_INET, SOCK_STREAM, IPPROTO_TCP, 0,                1  },
+	{ AF_INET, SOCK_DGRAM,  IPPROTO_TCP, -EPROTONOSUPPORT, 1  },
+	{ AF_INET, SOCK_DGRAM,  IPPROTO_UDP, 0,                1  },
+	{ AF_INET, SOCK_STREAM, IPPROTO_UDP, -EPROTONOSUPPORT, 1  },
+};
+
+#define ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0]))
+#define ERR_STRING_SZ	64
+
+static int run_tests(void)
+{
+	char err_string1[ERR_STRING_SZ];
+	char err_string2[ERR_STRING_SZ];
+	int i, err;
+
+	err = 0;
+	for (i = 0; i < ARRAY_SIZE(tests); i++) {
+		struct socket_testcase *s = &tests[i];
+		int fd;
+
+		fd = socket(s->domain, s->type, s->protocol);
+		if (fd < 0) {
+			if (s->nosupport_ok &&
+			    errno == EAFNOSUPPORT)
+				continue;
+
+			if (s->expect < 0 &&
+			    errno == -s->expect)
+				continue;
+
+			strerror_r(-s->expect, err_string1, ERR_STRING_SZ);
+			strerror_r(errno, err_string2, ERR_STRING_SZ);
+
+			fprintf(stderr, "socket(%d, %d, %d) expected "
+				"err (%s) got (%s)\n",
+				s->domain, s->type, s->protocol,
+				err_string1, err_string2);
+
+			err = -1;
+			break;
+		} else {
+			close(fd);
+
+			if (s->expect < 0) {
+				strerror_r(errno, err_string1, ERR_STRING_SZ);
+
+				fprintf(stderr, "socket(%d, %d, %d) expected "
+					"success got err (%s)\n",
+					s->domain, s->type, s->protocol,
+					err_string1);
+
+				err = -1;
+				break;
+			}
+		}
+	}
+
+	return err;
+}
+
+int main(void)
+{
+	int err = run_tests();
+
+	return err;
+}

tools/testing/selftests/ptrace/Makefile

@@ -0,0 +1,10 @@
+CFLAGS += -iquote../../../../include/uapi -Wall
+peeksiginfo: peeksiginfo.c
+
+all: peeksiginfo
+
+clean:
+	rm -f peeksiginfo
+
+run_tests: all
+	@./peeksiginfo || echo "peeksiginfo selftests: [FAIL]"

tools/testing/selftests/ptrace/peeksiginfo.c

@@ -0,0 +1,214 @@
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <signal.h>
+#include <unistd.h>
+#include <errno.h>
+#include <linux/types.h>
+#include <sys/wait.h>
+#include <sys/syscall.h>
+#include <sys/user.h>
+#include <sys/mman.h>
+
+#include "linux/ptrace.h"
+
+static int sys_rt_sigqueueinfo(pid_t tgid, int sig, siginfo_t *uinfo)
+{
+	return syscall(SYS_rt_sigqueueinfo, tgid, sig, uinfo);
+}
+
+static int sys_rt_tgsigqueueinfo(pid_t tgid, pid_t tid,
+					int sig, siginfo_t *uinfo)
+{
+	return syscall(SYS_rt_tgsigqueueinfo, tgid, tid, sig, uinfo);
+}
+
+static int sys_ptrace(int request, pid_t pid, void *addr, void *data)
+{
+	return syscall(SYS_ptrace, request, pid, addr, data);
+}
+
+#define SIGNR 10
+#define TEST_SICODE_PRIV	-1
+#define TEST_SICODE_SHARE	-2
+
+#define err(fmt, ...)						\
+		fprintf(stderr,					\
+			"Error (%s:%d): " fmt,			\
+			__FILE__, __LINE__, ##__VA_ARGS__)
+
+static int check_error_paths(pid_t child)
+{
+	struct ptrace_peeksiginfo_args arg;
+	int ret, exit_code = -1;
+	void *addr_rw, *addr_ro;
+
+	/*
+	 * Allocate two contiguous pages. The first one is for read-write,
+	 * another is for read-only.
+	 */
+	addr_rw = mmap(NULL, 2 * PAGE_SIZE, PROT_READ | PROT_WRITE,
+				MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	if (addr_rw == MAP_FAILED) {
+		err("mmap() failed: %m\n");
+		return 1;
+	}
+
+	addr_ro = mmap(addr_rw + PAGE_SIZE, PAGE_SIZE, PROT_READ,
+			MAP_PRIVATE | MAP_ANONYMOUS | MAP_FIXED, -1, 0);
+	if (addr_ro == MAP_FAILED) {
+		err("mmap() failed: %m\n");
+		goto out;
+	}
+
+	arg.nr = SIGNR;
+	arg.off = 0;
+
+	/* Unsupported flags */
+	arg.flags = ~0;
+	ret = sys_ptrace(PTRACE_PEEKSIGINFO, child, &arg, addr_rw);
+	if (ret != -1 || errno != EINVAL) {
+		err("sys_ptrace() returns %d (expected -1),"
+				" errno %d (expected %d): %m\n",
+				ret, errno, EINVAL);
+		goto out;
+	}
+	arg.flags = 0;
+
+	/* A part of the buffer is read-only */
+	ret = sys_ptrace(PTRACE_PEEKSIGINFO, child, &arg,
+					addr_ro - sizeof(siginfo_t) * 2);
+	if (ret != 2) {
+		err("sys_ptrace() returns %d (expected 2): %m\n", ret);
+		goto out;
+	}
+
+	/* Read-only buffer */
+	ret = sys_ptrace(PTRACE_PEEKSIGINFO, child, &arg, addr_ro);
+	if (ret != -1 && errno != EFAULT) {
+		err("sys_ptrace() returns %d (expected -1),"
+				" errno %d (expected %d): %m\n",
+				ret, errno, EFAULT);
+		goto out;
+	}
+
+	exit_code = 0;
+out:
+	munmap(addr_rw, 2 * PAGE_SIZE);
+	return exit_code;
+}
+
+int check_direct_path(pid_t child, int shared, int nr)
+{
+	struct ptrace_peeksiginfo_args arg = {.flags = 0, .nr = nr, .off = 0};
+	int i, j, ret, exit_code = -1;
+	siginfo_t siginfo[SIGNR];
+	int si_code;
+
+	if (shared == 1) {
+		arg.flags = PTRACE_PEEKSIGINFO_SHARED;
+		si_code = TEST_SICODE_SHARE;
+	} else {
+		arg.flags = 0;
+		si_code = TEST_SICODE_PRIV;
+	}
+
+	for (i = 0; i < SIGNR; ) {
+		arg.off = i;
+		ret = sys_ptrace(PTRACE_PEEKSIGINFO, child, &arg, siginfo);
+		if (ret == -1) {
+			err("ptrace() failed: %m\n");
+			goto out;
+		}
+
+		if (ret == 0)
+			break;
+
+		for (j = 0; j < ret; j++, i++) {
+			if (siginfo[j].si_code == si_code &&
+			    siginfo[j].si_int == i)
+				continue;
+
+			err("%d: Wrong siginfo i=%d si_code=%d si_int=%d\n",
+			     shared, i, siginfo[j].si_code, siginfo[j].si_int);
+			goto out;
+		}
+	}
+
+	if (i != SIGNR) {
+		err("Only %d signals were read\n", i);
+		goto out;
+	}
+
+	exit_code = 0;
+out:
+	return exit_code;
+}
+
+int main(int argc, char *argv[])
+{
+	siginfo_t siginfo[SIGNR];
+	int i, exit_code = 1;
+	sigset_t blockmask;
+	pid_t child;
+
+	sigemptyset(&blockmask);
+	sigaddset(&blockmask, SIGRTMIN);
+	sigprocmask(SIG_BLOCK, &blockmask, NULL);
+
+	child = fork();
+	if (child == -1) {
+		err("fork() failed: %m");
+		return 1;
+	} else if (child == 0) {
+		pid_t ppid = getppid();
+		while (1) {
+			if (ppid != getppid())
+				break;
+			sleep(1);
+		}
+		return 1;
+	}
+
+	/* Send signals in process-wide and per-thread queues */
+	for (i = 0; i < SIGNR; i++) {
+		siginfo->si_code = TEST_SICODE_SHARE;
+		siginfo->si_int = i;
+		sys_rt_sigqueueinfo(child, SIGRTMIN, siginfo);
+
+		siginfo->si_code = TEST_SICODE_PRIV;
+		siginfo->si_int = i;
+		sys_rt_tgsigqueueinfo(child, child, SIGRTMIN, siginfo);
+	}
+
+	if (sys_ptrace(PTRACE_ATTACH, child, NULL, NULL) == -1)
+		return 1;
+
+	waitpid(child, NULL, 0);
+
+	/* Dump signals one by one*/
+	if (check_direct_path(child, 0, 1))
+		goto out;
+	/* Dump all signals for one call */
+	if (check_direct_path(child, 0, SIGNR))
+		goto out;
+
+	/*
+	 * Dump signal from the process-wide queue.
+	 * The number of signals is not multible to the buffer size
+	 */
+	if (check_direct_path(child, 1, 3))
+		goto out;
+
+	if (check_error_paths(child))
+		goto out;
+
+	printf("PASS\n");
+	exit_code = 0;
+out:
+	if (sys_ptrace(PTRACE_KILL, child, NULL, NULL) == -1)
+		return 1;
+
+	waitpid(child, NULL, 0);
+
+	return exit_code;
+}

tools/testing/selftests/vm/Makefile

@@ -0,0 +1,14 @@
+# Makefile for vm selftests
+
+CC = $(CROSS_COMPILE)gcc
+CFLAGS = -Wall
+
+all: hugepage-mmap hugepage-shm  map_hugetlb thuge-gen
+%: %.c
+	$(CC) $(CFLAGS) -o $@ $^
+
+run_tests: all
+	@/bin/sh ./run_vmtests || echo "vmtests: [FAIL]"
+
+clean:
+	$(RM) hugepage-mmap hugepage-shm  map_hugetlb

tools/testing/selftests/vm/hugepage-mmap.c

@@ -0,0 +1,92 @@
+/*
+ * hugepage-mmap:
+ *
+ * Example of using huge page memory in a user application using the mmap
+ * system call.  Before running this application, make sure that the
+ * administrator has mounted the hugetlbfs filesystem (on some directory
+ * like /mnt) using the command mount -t hugetlbfs nodev /mnt. In this
+ * example, the app is requesting memory of size 256MB that is backed by
+ * huge pages.
+ *
+ * For the ia64 architecture, the Linux kernel reserves Region number 4 for
+ * huge pages.  That means that if one requires a fixed address, a huge page
+ * aligned address starting with 0x800000... will be required.  If a fixed
+ * address is not required, the kernel will select an address in the proper
+ * range.
+ * Other architectures, such as ppc64, i386 or x86_64 are not so constrained.
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <sys/mman.h>
+#include <fcntl.h>
+
+#define FILE_NAME "huge/hugepagefile"
+#define LENGTH (256UL*1024*1024)
+#define PROTECTION (PROT_READ | PROT_WRITE)
+
+/* Only ia64 requires this */
+#ifdef __ia64__
+#define ADDR (void *)(0x8000000000000000UL)
+#define FLAGS (MAP_SHARED | MAP_FIXED)
+#else
+#define ADDR (void *)(0x0UL)
+#define FLAGS (MAP_SHARED)
+#endif
+
+static void check_bytes(char *addr)
+{
+	printf("First hex is %x\n", *((unsigned int *)addr));
+}
+
+static void write_bytes(char *addr)
+{
+	unsigned long i;
+
+	for (i = 0; i < LENGTH; i++)
+		*(addr + i) = (char)i;
+}
+
+static int read_bytes(char *addr)
+{
+	unsigned long i;
+
+	check_bytes(addr);
+	for (i = 0; i < LENGTH; i++)
+		if (*(addr + i) != (char)i) {
+			printf("Mismatch at %lu\n", i);
+			return 1;
+		}
+	return 0;
+}
+
+int main(void)
+{
+	void *addr;
+	int fd, ret;
+
+	fd = open(FILE_NAME, O_CREAT | O_RDWR, 0755);
+	if (fd < 0) {
+		perror("Open failed");
+		exit(1);
+	}
+
+	addr = mmap(ADDR, LENGTH, PROTECTION, FLAGS, fd, 0);
+	if (addr == MAP_FAILED) {
+		perror("mmap");
+		unlink(FILE_NAME);
+		exit(1);
+	}
+
+	printf("Returned address is %p\n", addr);
+	check_bytes(addr);
+	write_bytes(addr);
+	ret = read_bytes(addr);
+
+	munmap(addr, LENGTH);
+	close(fd);
+	unlink(FILE_NAME);
+
+	return ret;
+}

tools/testing/selftests/vm/hugepage-shm.c

@@ -0,0 +1,100 @@
+/*
+ * hugepage-shm:
+ *
+ * Example of using huge page memory in a user application using Sys V shared
+ * memory system calls.  In this example the app is requesting 256MB of
+ * memory that is backed by huge pages.  The application uses the flag
+ * SHM_HUGETLB in the shmget system call to inform the kernel that it is
+ * requesting huge pages.
+ *
+ * For the ia64 architecture, the Linux kernel reserves Region number 4 for
+ * huge pages.  That means that if one requires a fixed address, a huge page
+ * aligned address starting with 0x800000... will be required.  If a fixed
+ * address is not required, the kernel will select an address in the proper
+ * range.
+ * Other architectures, such as ppc64, i386 or x86_64 are not so constrained.
+ *
+ * Note: The default shared memory limit is quite low on many kernels,
+ * you may need to increase it via:
+ *
+ * echo 268435456 > /proc/sys/kernel/shmmax
+ *
+ * This will increase the maximum size per shared memory segment to 256MB.
+ * The other limit that you will hit eventually is shmall which is the
+ * total amount of shared memory in pages. To set it to 16GB on a system
+ * with a 4kB pagesize do:
+ *
+ * echo 4194304 > /proc/sys/kernel/shmall
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <sys/types.h>
+#include <sys/ipc.h>
+#include <sys/shm.h>
+#include <sys/mman.h>
+
+#ifndef SHM_HUGETLB
+#define SHM_HUGETLB 04000
+#endif
+
+#define LENGTH (256UL*1024*1024)
+
+#define dprintf(x)  printf(x)
+
+/* Only ia64 requires this */
+#ifdef __ia64__
+#define ADDR (void *)(0x8000000000000000UL)
+#define SHMAT_FLAGS (SHM_RND)
+#else
+#define ADDR (void *)(0x0UL)
+#define SHMAT_FLAGS (0)
+#endif
+
+int main(void)
+{
+	int shmid;
+	unsigned long i;
+	char *shmaddr;
+
+	shmid = shmget(2, LENGTH, SHM_HUGETLB | IPC_CREAT | SHM_R | SHM_W);
+	if (shmid < 0) {
+		perror("shmget");
+		exit(1);
+	}
+	printf("shmid: 0x%x\n", shmid);
+
+	shmaddr = shmat(shmid, ADDR, SHMAT_FLAGS);
+	if (shmaddr == (char *)-1) {
+		perror("Shared memory attach failure");
+		shmctl(shmid, IPC_RMID, NULL);
+		exit(2);
+	}
+	printf("shmaddr: %p\n", shmaddr);
+
+	dprintf("Starting the writes:\n");
+	for (i = 0; i < LENGTH; i++) {
+		shmaddr[i] = (char)(i);
+		if (!(i % (1024 * 1024)))
+			dprintf(".");
+	}
+	dprintf("\n");
+
+	dprintf("Starting the Check...");
+	for (i = 0; i < LENGTH; i++)
+		if (shmaddr[i] != (char)i) {
+			printf("\nIndex %lu mismatched\n", i);
+			exit(3);
+		}
+	dprintf("Done.\n");
+
+	if (shmdt((const void *)shmaddr) != 0) {
+		perror("Detach failure");
+		shmctl(shmid, IPC_RMID, NULL);
+		exit(4);
+	}
+
+	shmctl(shmid, IPC_RMID, NULL);
+
+	return 0;
+}

tools/testing/selftests/vm/map_hugetlb.c

@@ -0,0 +1,79 @@
+/*
+ * Example of using hugepage memory in a user application using the mmap
+ * system call with MAP_HUGETLB flag.  Before running this program make
+ * sure the administrator has allocated enough default sized huge pages
+ * to cover the 256 MB allocation.
+ *
+ * For ia64 architecture, Linux kernel reserves Region number 4 for hugepages.
+ * That means the addresses starting with 0x800000... will need to be
+ * specified.  Specifying a fixed address is not required on ppc64, i386
+ * or x86_64.
+ */
+#include <stdlib.h>
+#include <stdio.h>
+#include <unistd.h>
+#include <sys/mman.h>
+#include <fcntl.h>
+
+#define LENGTH (256UL*1024*1024)
+#define PROTECTION (PROT_READ | PROT_WRITE)
+
+#ifndef MAP_HUGETLB
+#define MAP_HUGETLB 0x40000 /* arch specific */
+#endif
+
+/* Only ia64 requires this */
+#ifdef __ia64__
+#define ADDR (void *)(0x8000000000000000UL)
+#define FLAGS (MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB | MAP_FIXED)
+#else
+#define ADDR (void *)(0x0UL)
+#define FLAGS (MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB)
+#endif
+
+static void check_bytes(char *addr)
+{
+	printf("First hex is %x\n", *((unsigned int *)addr));
+}
+
+static void write_bytes(char *addr)
+{
+	unsigned long i;
+
+	for (i = 0; i < LENGTH; i++)
+		*(addr + i) = (char)i;
+}
+
+static int read_bytes(char *addr)
+{
+	unsigned long i;
+
+	check_bytes(addr);
+	for (i = 0; i < LENGTH; i++)
+		if (*(addr + i) != (char)i) {
+			printf("Mismatch at %lu\n", i);
+			return 1;
+		}
+	return 0;
+}
+
+int main(void)
+{
+	void *addr;
+	int ret;
+
+	addr = mmap(ADDR, LENGTH, PROTECTION, FLAGS, 0, 0);
+	if (addr == MAP_FAILED) {
+		perror("mmap");
+		exit(1);
+	}
+
+	printf("Returned address is %p\n", addr);
+	check_bytes(addr);
+	write_bytes(addr);
+	ret = read_bytes(addr);
+
+	munmap(addr, LENGTH);
+
+	return ret;
+}

tools/testing/selftests/vm/run_vmtests

@@ -0,0 +1,77 @@
+#!/bin/bash
+#please run as root
+
+#we need 256M, below is the size in kB
+needmem=262144
+mnt=./huge
+
+#get pagesize and freepages from /proc/meminfo
+while read name size unit; do
+	if [ "$name" = "HugePages_Free:" ]; then
+		freepgs=$size
+	fi
+	if [ "$name" = "Hugepagesize:" ]; then
+		pgsize=$size
+	fi
+done < /proc/meminfo
+
+#set proper nr_hugepages
+if [ -n "$freepgs" ] && [ -n "$pgsize" ]; then
+	nr_hugepgs=`cat /proc/sys/vm/nr_hugepages`
+	needpgs=`expr $needmem / $pgsize`
+	if [ $freepgs -lt $needpgs ]; then
+		lackpgs=$(( $needpgs - $freepgs ))
+		echo $(( $lackpgs + $nr_hugepgs )) > /proc/sys/vm/nr_hugepages
+		if [ $? -ne 0 ]; then
+			echo "Please run this test as root"
+			exit 1
+		fi
+	fi
+else
+	echo "no hugetlbfs support in kernel?"
+	exit 1
+fi
+
+mkdir $mnt
+mount -t hugetlbfs none $mnt
+
+echo "--------------------"
+echo "running hugepage-mmap"
+echo "--------------------"
+./hugepage-mmap
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+else
+	echo "[PASS]"
+fi
+
+shmmax=`cat /proc/sys/kernel/shmmax`
+shmall=`cat /proc/sys/kernel/shmall`
+echo 268435456 > /proc/sys/kernel/shmmax
+echo 4194304 > /proc/sys/kernel/shmall
+echo "--------------------"
+echo "running hugepage-shm"
+echo "--------------------"
+./hugepage-shm
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+else
+	echo "[PASS]"
+fi
+echo $shmmax > /proc/sys/kernel/shmmax
+echo $shmall > /proc/sys/kernel/shmall
+
+echo "--------------------"
+echo "running map_hugetlb"
+echo "--------------------"
+./map_hugetlb
+if [ $? -ne 0 ]; then
+	echo "[FAIL]"
+else
+	echo "[PASS]"
+fi
+
+#cleanup
+umount $mnt
+rm -rf $mnt
+echo $nr_hugepgs > /proc/sys/vm/nr_hugepages

tools/testing/selftests/vm/thuge-gen.c

@@ -0,0 +1,254 @@
+/* Test selecting other page sizes for mmap/shmget.
+
+   Before running this huge pages for each huge page size must have been
+   reserved.
+   For large pages beyond MAX_ORDER (like 1GB on x86) boot options must be used.
+   Also shmmax must be increased.
+   And you need to run as root to work around some weird permissions in shm.
+   And nothing using huge pages should run in parallel.
+   When the program aborts you may need to clean up the shm segments with
+   ipcrm -m by hand, like this
+   sudo ipcs | awk '$1 == "0x00000000" {print $2}' | xargs -n1 sudo ipcrm -m
+   (warning this will remove all if someone else uses them) */
+
+#define _GNU_SOURCE 1
+#include <sys/mman.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <sys/ipc.h>
+#include <sys/shm.h>
+#include <sys/stat.h>
+#include <glob.h>
+#include <assert.h>
+#include <unistd.h>
+#include <stdarg.h>
+#include <string.h>
+
+#define err(x) perror(x), exit(1)
+
+#define MAP_HUGE_2MB    (21 << MAP_HUGE_SHIFT)
+#define MAP_HUGE_1GB    (30 << MAP_HUGE_SHIFT)
+#define MAP_HUGE_SHIFT  26
+#define MAP_HUGE_MASK   0x3f
+#define MAP_HUGETLB	0x40000
+
+#define SHM_HUGETLB     04000   /* segment will use huge TLB pages */
+#define SHM_HUGE_SHIFT  26
+#define SHM_HUGE_MASK   0x3f
+#define SHM_HUGE_2MB    (21 << SHM_HUGE_SHIFT)
+#define SHM_HUGE_1GB    (30 << SHM_HUGE_SHIFT)
+
+#define NUM_PAGESIZES   5
+
+#define NUM_PAGES 4
+
+#define Dprintf(fmt...) // printf(fmt)
+
+unsigned long page_sizes[NUM_PAGESIZES];
+int num_page_sizes;
+
+int ilog2(unsigned long v)
+{
+	int l = 0;
+	while ((1UL << l) < v)
+		l++;
+	return l;
+}
+
+void find_pagesizes(void)
+{
+	glob_t g;
+	int i;
+	glob("/sys/kernel/mm/hugepages/hugepages-*kB", 0, NULL, &g);
+	assert(g.gl_pathc <= NUM_PAGESIZES);
+	for (i = 0; i < g.gl_pathc; i++) {
+		sscanf(g.gl_pathv[i], "/sys/kernel/mm/hugepages/hugepages-%lukB",
+				&page_sizes[i]);
+		page_sizes[i] <<= 10;
+		printf("Found %luMB\n", page_sizes[i] >> 20);
+	}
+	num_page_sizes = g.gl_pathc;
+	globfree(&g);
+}
+
+unsigned long default_huge_page_size(void)
+{
+	unsigned long hps = 0;
+	char *line = NULL;
+	size_t linelen = 0;
+	FILE *f = fopen("/proc/meminfo", "r");
+	if (!f)
+		return 0;
+	while (getline(&line, &linelen, f) > 0) {
+		if (sscanf(line, "Hugepagesize:       %lu kB", &hps) == 1) {
+			hps <<= 10;
+			break;
+		}
+	}
+	free(line);
+	return hps;
+}
+
+void show(unsigned long ps)
+{
+	char buf[100];
+	if (ps == getpagesize())
+		return;
+	printf("%luMB: ", ps >> 20);
+	fflush(stdout);
+	snprintf(buf, sizeof buf,
+		"cat /sys/kernel/mm/hugepages/hugepages-%lukB/free_hugepages",
+		ps >> 10);
+	system(buf);
+}
+
+unsigned long read_sysfs(int warn, char *fmt, ...)
+{
+	char *line = NULL;
+	size_t linelen = 0;
+	char buf[100];
+	FILE *f;
+	va_list ap;
+	unsigned long val = 0;
+
+	va_start(ap, fmt);
+	vsnprintf(buf, sizeof buf, fmt, ap);
+	va_end(ap);
+
+	f = fopen(buf, "r");
+	if (!f) {
+		if (warn)
+			printf("missing %s\n", buf);
+		return 0;
+	}
+	if (getline(&line, &linelen, f) > 0) {
+		sscanf(line, "%lu", &val);
+	}
+	fclose(f);
+	free(line);
+	return val;
+}
+
+unsigned long read_free(unsigned long ps)
+{
+	return read_sysfs(ps != getpagesize(),
+			"/sys/kernel/mm/hugepages/hugepages-%lukB/free_hugepages",
+			ps >> 10);
+}
+
+void test_mmap(unsigned long size, unsigned flags)
+{
+	char *map;
+	unsigned long before, after;
+	int err;
+
+	before = read_free(size);
+	map = mmap(NULL, size*NUM_PAGES, PROT_READ|PROT_WRITE,
+			MAP_PRIVATE|MAP_ANONYMOUS|MAP_HUGETLB|flags, 0, 0);
+
+	if (map == (char *)-1) err("mmap");
+	memset(map, 0xff, size*NUM_PAGES);
+	after = read_free(size);
+	Dprintf("before %lu after %lu diff %ld size %lu\n",
+		before, after, before - after, size);
+	assert(size == getpagesize() || (before - after) == NUM_PAGES);
+	show(size);
+	err = munmap(map, size);
+	assert(!err);
+}
+
+void test_shmget(unsigned long size, unsigned flags)
+{
+	int id;
+	unsigned long before, after;
+	int err;
+
+	before = read_free(size);
+	id = shmget(IPC_PRIVATE, size * NUM_PAGES, IPC_CREAT|0600|flags);
+	if (id < 0) err("shmget");
+
+	struct shm_info i;
+	if (shmctl(id, SHM_INFO, (void *)&i) < 0) err("shmctl");
+	Dprintf("alloc %lu res %lu\n", i.shm_tot, i.shm_rss);
+
+
+	Dprintf("id %d\n", id);
+	char *map = shmat(id, NULL, 0600);
+	if (map == (char*)-1) err("shmat");
+
+	shmctl(id, IPC_RMID, NULL);
+
+	memset(map, 0xff, size*NUM_PAGES);
+	after = read_free(size);
+
+	Dprintf("before %lu after %lu diff %ld size %lu\n",
+		before, after, before - after, size);
+	assert(size == getpagesize() || (before - after) == NUM_PAGES);
+	show(size);
+	err = shmdt(map);
+	assert(!err);
+}
+
+void sanity_checks(void)
+{
+	int i;
+	unsigned long largest = getpagesize();
+
+	for (i = 0; i < num_page_sizes; i++) {
+		if (page_sizes[i] > largest)
+			largest = page_sizes[i];
+
+		if (read_free(page_sizes[i]) < NUM_PAGES) {
+			printf("Not enough huge pages for page size %lu MB, need %u\n",
+				page_sizes[i] >> 20,
+				NUM_PAGES);
+			exit(0);
+		}
+	}
+
+	if (read_sysfs(0, "/proc/sys/kernel/shmmax") < NUM_PAGES * largest) {
+		printf("Please do echo %lu > /proc/sys/kernel/shmmax", largest * NUM_PAGES);
+		exit(0);
+	}
+
+#if defined(__x86_64__)
+	if (largest != 1U<<30) {
+		printf("No GB pages available on x86-64\n"
+		       "Please boot with hugepagesz=1G hugepages=%d\n", NUM_PAGES);
+		exit(0);
+	}
+#endif
+}
+
+int main(void)
+{
+	int i;
+	unsigned default_hps = default_huge_page_size();
+
+	find_pagesizes();
+
+	sanity_checks();
+
+	for (i = 0; i < num_page_sizes; i++) {
+		unsigned long ps = page_sizes[i];
+		int arg = ilog2(ps) << MAP_HUGE_SHIFT;
+		printf("Testing %luMB mmap with shift %x\n", ps >> 20, arg);
+		test_mmap(ps, MAP_HUGETLB | arg);
+	}
+	printf("Testing default huge mmap\n");
+	test_mmap(default_hps, SHM_HUGETLB);
+
+	puts("Testing non-huge shmget");
+	test_shmget(getpagesize(), 0);
+
+	for (i = 0; i < num_page_sizes; i++) {
+		unsigned long ps = page_sizes[i];
+		int arg = ilog2(ps) << SHM_HUGE_SHIFT;
+		printf("Testing %luMB shmget with shift %x\n", ps >> 20, arg);
+		test_shmget(ps, SHM_HUGETLB | arg);
+	}
+	puts("default huge shmget");
+	test_shmget(default_hps, SHM_HUGETLB);
+
+	return 0;
+}