| Message ID | 20220610002046.3243874-1-jsavitz@redhat.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [RFC] selftests/vm: enable running select groups of tests | expand |
Greeting,
FYI, we noticed the following commit (built with gcc-11):
commit: c5fac7744cf417920552cb4eec2600fb53048b0b ("[RFC PATCH] selftests/vm: enable running select groups of tests")
url: https://github.com/intel-lab-lkp/linux/commits/Joel-Savitz/selftests-vm-enable-running-select-groups-of-tests/20220610-082232
base: https://git.kernel.org/cgit/linux/kernel/git/shuah/linux-kselftest.git next
patch link: https://lore.kernel.org/lkml/20220610002046.3243874-1-jsavitz@redhat.com
in testcase: kernel-selftests
version: kernel-selftests-x86_64-cebf67a3-1_20220612
with following parameters:
sc_nr_hugepages: 2
group: vm
ucode: 0xec
test-description: 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.
test-url: https://www.kernel.org/doc/Documentation/kselftest.txt
on test machine: 12 threads 1 sockets Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz with 16G memory
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <oliver.sang@intel.com>
please be noted, besides kernel-selftests.vm.run_vmtests.sh../hugepage_mmap.fail,
we actually observed other tests failed on this commit but pass on parent.
here we list all tests run on both parent and the commit, even they have
same results. FYI
=========================================================================================
compiler/group/kconfig/rootfs/sc_nr_hugepages/tbox_group/testcase/ucode:
gcc-11/vm/x86_64-rhel-8.3-kselftests/debian-11.1-x86_64-20220510.cgz/2/lkp-cfl-d1/kernel-selftests/0xec
v5.19-rc1 c5fac7744cf417920552cb4eec2
---------------- ---------------------------
fail:runs %reproduction fail:runs
| | |
18:18 -67% 6:6 kernel-selftests.vm.madv_populate.fail
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../compaction_test.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../gup_test_a.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../gup_test_ct_F_0x1_0_19_0x1000.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../gup_test_u.pass
:18 33% 6:6 kernel-selftests.vm.run_vmtests.sh../hugepage_mmap.fail
18:18 -100% :6 kernel-selftests.vm.run_vmtests.sh../hugepage_mmap.pass
:18 33% 6:6 kernel-selftests.vm.run_vmtests.sh../hugepage_mremap_./huge/huge_mremap.fail
18:18 -100% :6 kernel-selftests.vm.run_vmtests.sh../hugepage_mremap_./huge/huge_mremap.pass
:18 33% 6:6 kernel-selftests.vm.run_vmtests.sh../hugepage_shm.fail
18:18 -100% :6 kernel-selftests.vm.run_vmtests.sh../hugepage_shm.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../hugepage_vmemmap.pass
:18 33% 6:6 kernel-selftests.vm.run_vmtests.sh../hugetlb_madvise_./huge/madvise_test.fail
18:18 -100% :6 kernel-selftests.vm.run_vmtests.sh../hugetlb_madvise_./huge/madvise_test.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../map_fixed_noreplace.pass
:18 33% 6:6 kernel-selftests.vm.run_vmtests.sh../map_hugetlb.fail
18:18 -100% :6 kernel-selftests.vm.run_vmtests.sh../map_hugetlb.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../map_populate.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../mlock2_tests.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../mlock_random_test.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../mrelease_test.fail
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../mremap_test.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../thuge_gen.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../userfaultfd_anon_20_16.pass
:18 33% 6:6 kernel-selftests.vm.run_vmtests.sh../userfaultfd_hugetlb_0_32.fail
18:18 -100% :6 kernel-selftests.vm.run_vmtests.sh../userfaultfd_hugetlb_128_32.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../userfaultfd_shmem_20_16.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../va_128TBswitch.fail
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh../virtual_address_range.pass
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh.fail
18:18 -67% 6:6 kernel-selftests.vm.run_vmtests.sh.sudo_u_nobody_./on_fault_limit.pass
18:18 -67% 6:6 kernel-selftests.vm.soft-dirty.pass
18:18 -67% 6:6 kernel-selftests.vm.split_huge_page_test.pass
# selftests: vm: run_vmtests.sh
# Selected test items: default
# -----------------------
# running ./hugepage-mmap
# -----------------------
# mmap: Cannot allocate memory
# [FAIL]
# ----------------------
# running ./hugepage-shm
# ----------------------
# shmget: Cannot allocate memory
# [FAIL]
# ---------------------
# running ./map_hugetlb
# ---------------------
# mmap: Cannot allocate memory
# Default size hugepages
# Mapping 256 Mbytes
# [FAIL]
# --------------------------------------------
# running ./hugepage-mremap ./huge/huge_mremap
# --------------------------------------------
# mmap1: Cannot allocate memory
# Map haddr: Returned address is 0xffffffffffffffff
# [FAIL]
# --------------------------
# running ./hugepage-vmemmap
# --------------------------
# Returned address is 0x7f4d3c600000 whose pfn is 1bc800
# [PASS]
# ---------------------------------------------
# running ./hugetlb-madvise ./huge/madvise-test
# ---------------------------------------------
# Not enough free huge pages to test, exiting!
# [FAIL]
# NOTE: These hugetlb tests provide minimal coverage. Use
# https://github.com/libhugetlbfs/libhugetlbfs.git for
# hugetlb regression testing.
...
# ----------------------------------
# running ./userfaultfd hugetlb 0 32
# ----------------------------------
# ERROR: invalid MiB (errno=0, line=1662)
#
# Usage: ./userfaultfd <test type> <MiB> <bounces> [hugetlbfs_file]
#
# Supported <test type>: anon, hugetlb, hugetlb_shared, shmem
#
# Examples:
#
# # Run anonymous memory test on 100MiB region with 99999 bounces:
# ./userfaultfd anon 100 99999
#
# # Run share memory test on 1GiB region with 99 bounces:
# ./userfaultfd shmem 1000 99
#
# # Run hugetlb memory test on 256MiB region with 50 bounces:
# ./userfaultfd hugetlb 256 50
#
# # Run the same hugetlb test but using shared file:
# ./userfaultfd hugetlb_shared 256 50 /dev/hugepages/hugefile
#
# # 10MiB-~6GiB 999 bounces anonymous test, continue forever unless an error triggers
# while ./userfaultfd anon $[RANDOM % 6000 + 10] 999; do true; done
#
# [FAIL]
...
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
sudo bin/lkp install job.yaml # job file is attached in this email
bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run
sudo bin/lkp run generated-yaml-file
# if come across any failure that blocks the test,
# please remove ~/.lkp and /lkp dir to run from a clean state.
Hey Joel, good work on this! I dont see anything wrong with this and the added functionality is very nice. I reviewed the code changes and everything seems to be correct. I also ran the code with a few examples, including correct and incorrect examples, and was not able to break it. Cheers, -- Nico Tested-by: Nico Pache <npache@redhat.com> Acked-by: Nico Pache <npache@redhat.com> On Thu, Jun 9, 2022 at 8:20 PM Joel Savitz <jsavitz@redhat.com> wrote: > > Add the ability to run one or more groups of vm tests (specified > by the environment variable TEST_ITEMS). Preserve existing default > behavior of running all tests when TEST_ITEMS is empty or "default". > > Documentation of test groups is included in the patch as follows: > > # ./run_vmtests.sh [ -h || --help ] > > usage: ./tools/testing/selftests/vm/run_vmtests.sh [ -h ] > -h: display this message > > The default behavior is to run all tests. > > Alternatively, specific groups tests can be run by defining > the TEST_ITEMS shell variable with a space-separated string > of one or more of the following: > - mmap > tests for mmap(2) > - gup_test > tests for gup using gup_test interface > - userfaultfd > tests for userfaultfd(2) > - compaction > a test for the patch "Allow compaction of unevictable pages" > - mlock > tests for mlock(2) > - mremap > tests for mremap(2) > - hugevm > tests for very large virtual address space > - vmalloc > vmalloc smoke tests > - hmm > hmm smoke tests > - madv_populate > test memadvise(2) MADV_POPULATE_{READ,WRITE} options > - memfd_secret > test memfd_secret(2) > - process_mrelease > test process_mrelease(2) > - ksm > ksm tests that do not require >=2 NUMA nodes > - ksm_numa > ksm tests that require >=2 NUMA nodes > example: TEST_ITEMS="hmm mmap ksm" ./run_vmtests.sh > > Signed-off-by: Joel Savitz <jsavitz@redhat.com> > --- > tools/testing/selftests/vm/run_vmtests.sh | 224 +++++++++++++++------- > 1 file changed, 150 insertions(+), 74 deletions(-) > > diff --git a/tools/testing/selftests/vm/run_vmtests.sh b/tools/testing/selftests/vm/run_vmtests.sh > index 41fce8bea929..d31fc66205f8 100755 > --- a/tools/testing/selftests/vm/run_vmtests.sh > +++ b/tools/testing/selftests/vm/run_vmtests.sh > @@ -1,22 +1,74 @@ > #!/bin/bash > # SPDX-License-Identifier: GPL-2.0 > -#please run as root > +# Please run as root > > # Kselftest framework requirement - SKIP code is 4. > ksft_skip=4 > > mnt=./huge > exitcode=0 > - > -#get huge pagesize and freepages from /proc/meminfo > -while read -r name size unit; do > - if [ "$name" = "HugePages_Free:" ]; then > - freepgs="$size" > - fi > - if [ "$name" = "Hugepagesize:" ]; then > - hpgsize_KB="$size" > +nr_tests_ran=0 > + > +if [ ${1:-0} == "-h" ] || [ ${1:-0} == "--help" ] > +then > + cat <<EOF > +usage: ${BASH_SOURCE[0]:-$0} [ -h ] > + -h: display this message > + > +The default behavior is to run all tests. > + > +Alternatively, specific groups tests can be run by defining > +the TEST_ITEMS shell variable with a space-separated string > +of one or more of the following: > +- mmap > + tests for mmap(2) > +- gup_test > + tests for gup using gup_test interface > +- userfaultfd > + tests for userfaultfd(2) > +- compaction > + a test for the patch "Allow compaction of unevictable pages" > +- mlock > + tests for mlock(2) > +- mremap > + tests for mremap(2) > +- hugevm > + tests for very large virtual address space > +- vmalloc > + vmalloc smoke tests > +- hmm > + hmm smoke tests > +- madv_populate > + test memadvise(2) MADV_POPULATE_{READ,WRITE} options > +- memfd_secret > + test memfd_secret(2) > +- process_mrelease > + test process_mrelease(2) > +- ksm > + ksm tests that do not require >=2 NUMA nodes > +- ksm_numa > + ksm tests that require >=2 NUMA nodes > +example: TEST_ITEMS="hmm mmap ksm" ./run_vmtests.sh > +EOF > + exit 0 > +fi # $1 == -h || $1 == --help > + > +# default behavior: run all tests > +TEST_ITEMS=${TEST_ITEMS:-default} > + > +echo "Selected test items: ${TEST_ITEMS}" > + > +test_selected() { > + if [ "$TEST_ITEMS" == "default" ]; then > + # If no TEST_ITEMS are specified, run all tests > + return 0 > fi > -done < /proc/meminfo > + echo ${TEST_ITEMS} | grep ${1} 2>&1 >/dev/null > + return ${?} > +} > + > +# Hugepage setup only needed for hugetlb tests > +if test_selected "hugetlb"; then > > # Simple hugetlbfs tests have a hardcoded minimum requirement of > # huge pages totaling 256MB (262144KB) in size. The userfaultfd > @@ -28,7 +80,17 @@ hpgsize_MB=$((hpgsize_KB / 1024)) > half_ufd_size_MB=$((((nr_cpus * hpgsize_MB + 127) / 128) * 128)) > needmem_KB=$((half_ufd_size_MB * 2 * 1024)) > > -#set proper nr_hugepages > +# get huge pagesize and freepages from /proc/meminfo > +while read -r name size unit; do > + if [ "$name" = "HugePages_Free:" ]; then > + freepgs="$size" > + fi > + if [ "$name" = "Hugepagesize:" ]; then > + hpgsize_KB="$size" > + fi > +done < /proc/meminfo > + > +# set proper nr_hugepages > if [ -n "$freepgs" ] && [ -n "$hpgsize_KB" ]; then > nr_hugepgs=$(cat /proc/sys/vm/nr_hugepages) > needpgs=$((needmem_KB / hpgsize_KB)) > @@ -57,126 +119,140 @@ else > exit 1 > fi > > -#filter 64bit architectures > +fi # test_selected "hugetlb" > + > +# filter 64bit architectures > ARCH64STR="arm64 ia64 mips64 parisc64 ppc64 ppc64le riscv64 s390x sh64 sparc64 x86_64" > if [ -z "$ARCH" ]; then > ARCH=$(uname -m 2>/dev/null | sed -e 's/aarch64.*/arm64/') > fi > VADDR64=0 > -echo "$ARCH64STR" | grep "$ARCH" && VADDR64=1 > +echo "$ARCH64STR" | grep "$ARCH" &>/dev/null && VADDR64=1 > > # Usage: run_test [test binary] [arbitrary test arguments...] > run_test() { > - local title="running $*" > - local sep=$(echo -n "$title" | tr "[:graph:][:space:]" -) > - printf "%s\n%s\n%s\n" "$sep" "$title" "$sep" > - > - "$@" > - local ret=$? > - if [ $ret -eq 0 ]; then > - echo "[PASS]" > - elif [ $ret -eq $ksft_skip ]; then > - echo "[SKIP]" > - exitcode=$ksft_skip > - else > - echo "[FAIL]" > - exitcode=1 > - fi > + if test_selected ${CATEGORY}; then > + local title="running $*" > + local sep=$(echo -n "$title" | tr "[:graph:][:space:]" -) > + printf "%s\n%s\n%s\n" "$sep" "$title" "$sep" > + > + "$@" > + local ret=$? > + if [ $ret -eq 0 ]; then > + echo "[PASS]" > + elif [ $ret -eq $ksft_skip ]; then > + echo "[SKIP]" > + exitcode=$ksft_skip > + else > + echo "[FAIL]" > + exitcode=1 > + fi > + nr_tests_ran=$((nr_tests_ran + 1)) > + fi # test_selected > } > > -mkdir "$mnt" > -mount -t hugetlbfs none "$mnt" > +# setup only needed for hugetlb tests > +if test_selected "hugetlb"; then > + mkdir "$mnt" > + mount -t hugetlbfs none "$mnt" > +fi > > -run_test ./hugepage-mmap > +CATEGORY="hugetlb" run_test ./hugepage-mmap > > shmmax=$(cat /proc/sys/kernel/shmmax) > shmall=$(cat /proc/sys/kernel/shmall) > echo 268435456 > /proc/sys/kernel/shmmax > echo 4194304 > /proc/sys/kernel/shmall > -run_test ./hugepage-shm > +CATEGORY="hugetlb" run_test ./hugepage-shm > echo "$shmmax" > /proc/sys/kernel/shmmax > echo "$shmall" > /proc/sys/kernel/shmall > > -run_test ./map_hugetlb > +CATEGORY="hugetlb" run_test ./map_hugetlb > > -run_test ./hugepage-mremap "$mnt"/huge_mremap > -rm -f "$mnt"/huge_mremap > +CATEGORY="hugetlb" run_test ./hugepage-mremap "$mnt"/huge_mremap > +test_selected "hugetlb" && rm -f "$mnt"/huge_mremap > > -run_test ./hugepage-vmemmap > +CATEGORY="hugetlb" run_test ./hugepage-vmemmap > > -run_test ./hugetlb-madvise "$mnt"/madvise-test > -rm -f "$mnt"/madvise-test > +CATEGORY="hugetlb" run_test ./hugetlb-madvise "$mnt"/madvise-test > +test_selected "hugetlb" && rm -f "$mnt"/madvise-test > > -echo "NOTE: The above hugetlb tests provide minimal coverage. Use" > -echo " https://github.com/libhugetlbfs/libhugetlbfs.git for" > -echo " hugetlb regression testing." > +if test_selected "hugetlb"; then > + echo "NOTE: These hugetlb tests provide minimal coverage. Use" > + echo " https://github.com/libhugetlbfs/libhugetlbfs.git for" > + echo " hugetlb regression testing." > +fi > > -run_test ./map_fixed_noreplace > +CATEGORY="mmap" run_test ./map_fixed_noreplace > > # get_user_pages_fast() benchmark > -run_test ./gup_test -u > +CATEGORY="gup_test" run_test ./gup_test -u > # pin_user_pages_fast() benchmark > -run_test ./gup_test -a > +CATEGORY="gup_test" run_test ./gup_test -a > # Dump pages 0, 19, and 4096, using pin_user_pages: > -run_test ./gup_test -ct -F 0x1 0 19 0x1000 > +CATEGORY="gup_test" run_test ./gup_test -ct -F 0x1 0 19 0x1000 > > -run_test ./userfaultfd anon 20 16 > +CATEGORY="userfaultfd" run_test ./userfaultfd anon 20 16 > # Test requires source and destination huge pages. Size of source > # (half_ufd_size_MB) is passed as argument to test. > -run_test ./userfaultfd hugetlb "$half_ufd_size_MB" 32 > -run_test ./userfaultfd shmem 20 16 > - > -#cleanup > -umount "$mnt" > -rm -rf "$mnt" > -echo "$nr_hugepgs" > /proc/sys/vm/nr_hugepages > +CATEGORY="userfaultfd" run_test ./userfaultfd hugetlb "$half_ufd_size_MB" 32 > +CATEGORY="userfaultfd" run_test ./userfaultfd shmem 20 16 > + > +# cleanup (only needed when running hugetlb tests) > +if test_selected "hugetlb"; then > + umount "$mnt" > + rm -rf "$mnt" > + echo "$nr_hugepgs" > /proc/sys/vm/nr_hugepages > +fi > > -run_test ./compaction_test > +CATEGORY="compaction" run_test ./compaction_test > > -run_test sudo -u nobody ./on-fault-limit > +CATEGORY="mlock" run_test sudo -u nobody ./on-fault-limit > > -run_test ./map_populate > +CATEGORY="mmap" run_test ./map_populate > > -run_test ./mlock-random-test > +CATEGORY="mlock" run_test ./mlock-random-test > > -run_test ./mlock2-tests > +CATEGORY="mlock" run_test ./mlock2-tests > > -run_test ./mrelease_test > +CATEGORY="process_mrelease" run_test ./mrelease_test > > -run_test ./mremap_test > +CATEGORY="mremap" run_test ./mremap_test > > -run_test ./thuge-gen > +CATEGORY="hugetlb" run_test ./thuge-gen > > if [ $VADDR64 -ne 0 ]; then > - run_test ./virtual_address_range > + CATEGORY="hugevm" run_test ./virtual_address_range > > # virtual address 128TB switch test > - run_test ./va_128TBswitch > + CATEGORY="hugevm" run_test ./va_128TBswitch > fi # VADDR64 > > # vmalloc stability smoke test > -run_test ./test_vmalloc.sh smoke > +CATEGORY="vmalloc" run_test ./test_vmalloc.sh smoke > > -run_test ./mremap_dontunmap > +CATEGORY="mremap" run_test ./mremap_dontunmap > > -run_test ./test_hmm.sh smoke > +CATEGORY="hmm" run_test ./test_hmm.sh smoke > > # MADV_POPULATE_READ and MADV_POPULATE_WRITE tests > -run_test ./madv_populate > +CATEGORY="madv_populate" run_test ./madv_populate > > -run_test ./memfd_secret > +CATEGORY="memfd_secret" run_test ./memfd_secret > > # KSM MADV_MERGEABLE test with 10 identical pages > -run_test ./ksm_tests -M -p 10 > +CATEGORY="ksm" run_test ./ksm_tests -M -p 10 > # KSM unmerge test > -run_test ./ksm_tests -U > +CATEGORY="ksm" run_test ./ksm_tests -U > # KSM test with 10 zero pages and use_zero_pages = 0 > -run_test ./ksm_tests -Z -p 10 -z 0 > +CATEGORY="ksm" run_test ./ksm_tests -Z -p 10 -z 0 > # KSM test with 10 zero pages and use_zero_pages = 1 > -run_test ./ksm_tests -Z -p 10 -z 1 > +CATEGORY="ksm" run_test ./ksm_tests -Z -p 10 -z 1 > # KSM test with 2 NUMA nodes and merge_across_nodes = 1 > -run_test ./ksm_tests -N -m 1 > +CATEGORY="ksm_numa" run_test ./ksm_tests -N -m 1 > # KSM test with 2 NUMA nodes and merge_across_nodes = 0 > -run_test ./ksm_tests -N -m 0 > +CATEGORY="ksm_numa" run_test ./ksm_tests -N -m 0 > + > +echo "Ran $nr_tests_ran tests" > > exit $exitcode > -- > 2.31.1 >
diff --git a/tools/testing/selftests/vm/run_vmtests.sh b/tools/testing/selftests/vm/run_vmtests.sh index 41fce8bea929..d31fc66205f8 100755 --- a/tools/testing/selftests/vm/run_vmtests.sh +++ b/tools/testing/selftests/vm/run_vmtests.sh @@ -1,22 +1,74 @@ #!/bin/bash # SPDX-License-Identifier: GPL-2.0 -#please run as root +# Please run as root # Kselftest framework requirement - SKIP code is 4. ksft_skip=4 mnt=./huge exitcode=0 - -#get huge pagesize and freepages from /proc/meminfo -while read -r name size unit; do - if [ "$name" = "HugePages_Free:" ]; then - freepgs="$size" - fi - if [ "$name" = "Hugepagesize:" ]; then - hpgsize_KB="$size" +nr_tests_ran=0 + +if [ ${1:-0} == "-h" ] || [ ${1:-0} == "--help" ] +then + cat <<EOF +usage: ${BASH_SOURCE[0]:-$0} [ -h ] + -h: display this message + +The default behavior is to run all tests. + +Alternatively, specific groups tests can be run by defining +the TEST_ITEMS shell variable with a space-separated string +of one or more of the following: +- mmap + tests for mmap(2) +- gup_test + tests for gup using gup_test interface +- userfaultfd + tests for userfaultfd(2) +- compaction + a test for the patch "Allow compaction of unevictable pages" +- mlock + tests for mlock(2) +- mremap + tests for mremap(2) +- hugevm + tests for very large virtual address space +- vmalloc + vmalloc smoke tests +- hmm + hmm smoke tests +- madv_populate + test memadvise(2) MADV_POPULATE_{READ,WRITE} options +- memfd_secret + test memfd_secret(2) +- process_mrelease + test process_mrelease(2) +- ksm + ksm tests that do not require >=2 NUMA nodes +- ksm_numa + ksm tests that require >=2 NUMA nodes +example: TEST_ITEMS="hmm mmap ksm" ./run_vmtests.sh +EOF + exit 0 +fi # $1 == -h || $1 == --help + +# default behavior: run all tests +TEST_ITEMS=${TEST_ITEMS:-default} + +echo "Selected test items: ${TEST_ITEMS}" + +test_selected() { + if [ "$TEST_ITEMS" == "default" ]; then + # If no TEST_ITEMS are specified, run all tests + return 0 fi -done < /proc/meminfo + echo ${TEST_ITEMS} | grep ${1} 2>&1 >/dev/null + return ${?} +} + +# Hugepage setup only needed for hugetlb tests +if test_selected "hugetlb"; then # Simple hugetlbfs tests have a hardcoded minimum requirement of # huge pages totaling 256MB (262144KB) in size. The userfaultfd @@ -28,7 +80,17 @@ hpgsize_MB=$((hpgsize_KB / 1024)) half_ufd_size_MB=$((((nr_cpus * hpgsize_MB + 127) / 128) * 128)) needmem_KB=$((half_ufd_size_MB * 2 * 1024)) -#set proper nr_hugepages +# get huge pagesize and freepages from /proc/meminfo +while read -r name size unit; do + if [ "$name" = "HugePages_Free:" ]; then + freepgs="$size" + fi + if [ "$name" = "Hugepagesize:" ]; then + hpgsize_KB="$size" + fi +done < /proc/meminfo + +# set proper nr_hugepages if [ -n "$freepgs" ] && [ -n "$hpgsize_KB" ]; then nr_hugepgs=$(cat /proc/sys/vm/nr_hugepages) needpgs=$((needmem_KB / hpgsize_KB)) @@ -57,126 +119,140 @@ else exit 1 fi -#filter 64bit architectures +fi # test_selected "hugetlb" + +# filter 64bit architectures ARCH64STR="arm64 ia64 mips64 parisc64 ppc64 ppc64le riscv64 s390x sh64 sparc64 x86_64" if [ -z "$ARCH" ]; then ARCH=$(uname -m 2>/dev/null | sed -e 's/aarch64.*/arm64/') fi VADDR64=0 -echo "$ARCH64STR" | grep "$ARCH" && VADDR64=1 +echo "$ARCH64STR" | grep "$ARCH" &>/dev/null && VADDR64=1 # Usage: run_test [test binary] [arbitrary test arguments...] run_test() { - local title="running $*" - local sep=$(echo -n "$title" | tr "[:graph:][:space:]" -) - printf "%s\n%s\n%s\n" "$sep" "$title" "$sep" - - "$@" - local ret=$? - if [ $ret -eq 0 ]; then - echo "[PASS]" - elif [ $ret -eq $ksft_skip ]; then - echo "[SKIP]" - exitcode=$ksft_skip - else - echo "[FAIL]" - exitcode=1 - fi + if test_selected ${CATEGORY}; then + local title="running $*" + local sep=$(echo -n "$title" | tr "[:graph:][:space:]" -) + printf "%s\n%s\n%s\n" "$sep" "$title" "$sep" + + "$@" + local ret=$? + if [ $ret -eq 0 ]; then + echo "[PASS]" + elif [ $ret -eq $ksft_skip ]; then + echo "[SKIP]" + exitcode=$ksft_skip + else + echo "[FAIL]" + exitcode=1 + fi + nr_tests_ran=$((nr_tests_ran + 1)) + fi # test_selected } -mkdir "$mnt" -mount -t hugetlbfs none "$mnt" +# setup only needed for hugetlb tests +if test_selected "hugetlb"; then + mkdir "$mnt" + mount -t hugetlbfs none "$mnt" +fi -run_test ./hugepage-mmap +CATEGORY="hugetlb" run_test ./hugepage-mmap shmmax=$(cat /proc/sys/kernel/shmmax) shmall=$(cat /proc/sys/kernel/shmall) echo 268435456 > /proc/sys/kernel/shmmax echo 4194304 > /proc/sys/kernel/shmall -run_test ./hugepage-shm +CATEGORY="hugetlb" run_test ./hugepage-shm echo "$shmmax" > /proc/sys/kernel/shmmax echo "$shmall" > /proc/sys/kernel/shmall -run_test ./map_hugetlb +CATEGORY="hugetlb" run_test ./map_hugetlb -run_test ./hugepage-mremap "$mnt"/huge_mremap -rm -f "$mnt"/huge_mremap +CATEGORY="hugetlb" run_test ./hugepage-mremap "$mnt"/huge_mremap +test_selected "hugetlb" && rm -f "$mnt"/huge_mremap -run_test ./hugepage-vmemmap +CATEGORY="hugetlb" run_test ./hugepage-vmemmap -run_test ./hugetlb-madvise "$mnt"/madvise-test -rm -f "$mnt"/madvise-test +CATEGORY="hugetlb" run_test ./hugetlb-madvise "$mnt"/madvise-test +test_selected "hugetlb" && rm -f "$mnt"/madvise-test -echo "NOTE: The above hugetlb tests provide minimal coverage. Use" -echo " https://github.com/libhugetlbfs/libhugetlbfs.git for" -echo " hugetlb regression testing." +if test_selected "hugetlb"; then + echo "NOTE: These hugetlb tests provide minimal coverage. Use" + echo " https://github.com/libhugetlbfs/libhugetlbfs.git for" + echo " hugetlb regression testing." +fi -run_test ./map_fixed_noreplace +CATEGORY="mmap" run_test ./map_fixed_noreplace # get_user_pages_fast() benchmark -run_test ./gup_test -u +CATEGORY="gup_test" run_test ./gup_test -u # pin_user_pages_fast() benchmark -run_test ./gup_test -a +CATEGORY="gup_test" run_test ./gup_test -a # Dump pages 0, 19, and 4096, using pin_user_pages: -run_test ./gup_test -ct -F 0x1 0 19 0x1000 +CATEGORY="gup_test" run_test ./gup_test -ct -F 0x1 0 19 0x1000 -run_test ./userfaultfd anon 20 16 +CATEGORY="userfaultfd" run_test ./userfaultfd anon 20 16 # Test requires source and destination huge pages. Size of source # (half_ufd_size_MB) is passed as argument to test. -run_test ./userfaultfd hugetlb "$half_ufd_size_MB" 32 -run_test ./userfaultfd shmem 20 16 - -#cleanup -umount "$mnt" -rm -rf "$mnt" -echo "$nr_hugepgs" > /proc/sys/vm/nr_hugepages +CATEGORY="userfaultfd" run_test ./userfaultfd hugetlb "$half_ufd_size_MB" 32 +CATEGORY="userfaultfd" run_test ./userfaultfd shmem 20 16 + +# cleanup (only needed when running hugetlb tests) +if test_selected "hugetlb"; then + umount "$mnt" + rm -rf "$mnt" + echo "$nr_hugepgs" > /proc/sys/vm/nr_hugepages +fi -run_test ./compaction_test +CATEGORY="compaction" run_test ./compaction_test -run_test sudo -u nobody ./on-fault-limit +CATEGORY="mlock" run_test sudo -u nobody ./on-fault-limit -run_test ./map_populate +CATEGORY="mmap" run_test ./map_populate -run_test ./mlock-random-test +CATEGORY="mlock" run_test ./mlock-random-test -run_test ./mlock2-tests +CATEGORY="mlock" run_test ./mlock2-tests -run_test ./mrelease_test +CATEGORY="process_mrelease" run_test ./mrelease_test -run_test ./mremap_test +CATEGORY="mremap" run_test ./mremap_test -run_test ./thuge-gen +CATEGORY="hugetlb" run_test ./thuge-gen if [ $VADDR64 -ne 0 ]; then - run_test ./virtual_address_range + CATEGORY="hugevm" run_test ./virtual_address_range # virtual address 128TB switch test - run_test ./va_128TBswitch + CATEGORY="hugevm" run_test ./va_128TBswitch fi # VADDR64 # vmalloc stability smoke test -run_test ./test_vmalloc.sh smoke +CATEGORY="vmalloc" run_test ./test_vmalloc.sh smoke -run_test ./mremap_dontunmap +CATEGORY="mremap" run_test ./mremap_dontunmap -run_test ./test_hmm.sh smoke +CATEGORY="hmm" run_test ./test_hmm.sh smoke # MADV_POPULATE_READ and MADV_POPULATE_WRITE tests -run_test ./madv_populate +CATEGORY="madv_populate" run_test ./madv_populate -run_test ./memfd_secret +CATEGORY="memfd_secret" run_test ./memfd_secret # KSM MADV_MERGEABLE test with 10 identical pages -run_test ./ksm_tests -M -p 10 +CATEGORY="ksm" run_test ./ksm_tests -M -p 10 # KSM unmerge test -run_test ./ksm_tests -U +CATEGORY="ksm" run_test ./ksm_tests -U # KSM test with 10 zero pages and use_zero_pages = 0 -run_test ./ksm_tests -Z -p 10 -z 0 +CATEGORY="ksm" run_test ./ksm_tests -Z -p 10 -z 0 # KSM test with 10 zero pages and use_zero_pages = 1 -run_test ./ksm_tests -Z -p 10 -z 1 +CATEGORY="ksm" run_test ./ksm_tests -Z -p 10 -z 1 # KSM test with 2 NUMA nodes and merge_across_nodes = 1 -run_test ./ksm_tests -N -m 1 +CATEGORY="ksm_numa" run_test ./ksm_tests -N -m 1 # KSM test with 2 NUMA nodes and merge_across_nodes = 0 -run_test ./ksm_tests -N -m 0 +CATEGORY="ksm_numa" run_test ./ksm_tests -N -m 0 + +echo "Ran $nr_tests_ran tests" exit $exitcode
Add the ability to run one or more groups of vm tests (specified by the environment variable TEST_ITEMS). Preserve existing default behavior of running all tests when TEST_ITEMS is empty or "default". Documentation of test groups is included in the patch as follows: # ./run_vmtests.sh [ -h || --help ] usage: ./tools/testing/selftests/vm/run_vmtests.sh [ -h ] -h: display this message The default behavior is to run all tests. Alternatively, specific groups tests can be run by defining the TEST_ITEMS shell variable with a space-separated string of one or more of the following: - mmap tests for mmap(2) - gup_test tests for gup using gup_test interface - userfaultfd tests for userfaultfd(2) - compaction a test for the patch "Allow compaction of unevictable pages" - mlock tests for mlock(2) - mremap tests for mremap(2) - hugevm tests for very large virtual address space - vmalloc vmalloc smoke tests - hmm hmm smoke tests - madv_populate test memadvise(2) MADV_POPULATE_{READ,WRITE} options - memfd_secret test memfd_secret(2) - process_mrelease test process_mrelease(2) - ksm ksm tests that do not require >=2 NUMA nodes - ksm_numa ksm tests that require >=2 NUMA nodes example: TEST_ITEMS="hmm mmap ksm" ./run_vmtests.sh Signed-off-by: Joel Savitz <jsavitz@redhat.com> --- tools/testing/selftests/vm/run_vmtests.sh | 224 +++++++++++++++------- 1 file changed, 150 insertions(+), 74 deletions(-)