Merge tag 'linux-kselftest-4.13-rc1-update' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest

Pull Kselftest updates from Shuah Khan:
 "This update consists of:

   - TAP13 framework and changes to some tests to convert to TAP13.
     Converting kselftest output to standard format will help identify
     run to run differences and pin point failures easily. TAP13 format
     has been in use for several years and the output is human friendly.

     Please find the specification:
       https://testanything.org/tap-version-13-specification.html

     Credit goes to Tim Bird for recommending TAP13 as a suitable
     format, and to Grag KH for kick starting the work with help from
     Paul Elder and Alice Ferrazzi

     The first phase of the TAp13 conversion is included in this update.
     Future updates will include updates to rest of the tests.

   - Masami Hiramatsu fixed ftrace to run on 4.9 stable kernels.

   - Kselftest documnetation has been converted to ReST format. Document
     now has a new home under Documentation/dev-tools.

   - kselftest_harness.h is now available for general use as a result of
     Mickaël Salaün's work.

   - Several fixes to skip and/or fail tests gracefully on older
     releases"

* tag 'linux-kselftest-4.13-rc1-update' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest: (48 commits)
  selftests: membarrier: use ksft_* var arg msg api
  selftests: breakpoints: breakpoint_test_arm64: convert test to use TAP13
  selftests: breakpoints: step_after_suspend_test use ksft_* var arg msg api
  selftests: breakpoint_test: use ksft_* var arg msg api
  kselftest: add ksft_print_msg() function to output general information
  kselftest: make ksft_* output functions variadic
  selftests/capabilities: Fix the test_execve test
  selftests: intel_pstate: add .gitignore
  selftests: fix memory-hotplug test
  selftests: add missing test name in memory-hotplug test
  selftests: check percentage range for memory-hotplug test
  selftests: check hot-pluggagble memory for memory-hotplug test
  selftests: typo correction for memory-hotplug test
  selftests: ftrace: Use md5sum to take less time of checking logs
  tools/testing/selftests/sysctl: Add pre-check to the value of writes_strict
  kselftest.rst: do some adjustments after ReST conversion
  selftest/net/Makefile: Specify output with $(OUTPUT)
  selftest/intel_pstate/aperf: Use LDLIBS instead of LDFLAGS
  selftest/memfd/Makefile: Fix build error
  selftests: lib: Skip tests on missing test modules
  ...

1 files changed, 22 insertions, 0 deletions

diff --git a/tools/testing/selftests/sysctl/common_tests b/tools/testing/selftests/sysctl/common_tests
index 17d534b1b7b4..b6862322962f 100644
--- a/tools/testing/selftests/sysctl/common_tests
+++ b/tools/testing/selftests/sysctl/common_tests
@@ -24,6 +24,14 @@ verify()
 	return 0
 }
 
+exit_test()
+{
+	if [ ! -z ${old_strict} ]; then
+		echo ${old_strict} > ${WRITES_STRICT}
+	fi
+	exit $rc
+}
+
 trap 'set_orig; rm -f "${TEST_FILE}"' EXIT
 
 rc=0
@@ -63,6 +71,20 @@ else
 	echo "ok"
 fi
 
+echo -n "Checking write strict setting ... "
+WRITES_STRICT="${SYSCTL}/kernel/sysctl_writes_strict"
+if [ ! -e ${WRITES_STRICT} ]; then
+	echo "FAIL, but skip in case of old kernel" >&2
+else
+	old_strict=$(cat ${WRITES_STRICT})
+	if [ "$old_strict" = "1" ]; then
+		echo "ok"
+	else
+		echo "FAIL, strict value is 0 but force to 1 to continue" >&2
+		echo "1" > ${WRITES_STRICT}
+	fi
+fi
+
 # Now that we've validated the sanity of "set_test" and "set_orig",
 # we can use those functions to set starting states before running
 # specific behavioral tests.