Re: [PATCH] selftests/mm: run_vmtests.sh: add missing tests

[Muhammad Usama Anjum <usama.anjum@collabora.com>]

On 1/19/24 9:09 PM, Ryan Roberts wrote:
> Hi Muhammad,
> 
> Afraid this patch is causing a regression on our CI system when it turned up in
> linux-next today. Additionally, 2 of thetests you have added are failing because
> the scripts are not exported correctly...
Andrew has dropped this patch for now.

> 
> On 16/01/2024 09:06, Muhammad Usama Anjum wrote:
>> Add missing tests to run_vmtests.sh. The mm kselftests are run through
>> run_vmtests.sh. If a test isn't present in this script, it'll not run
>> with run_tests or `make -C tools/testing/selftests/mm run_tests`.
>>
>> Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
>> ---
>>  tools/testing/selftests/mm/run_vmtests.sh | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/tools/testing/selftests/mm/run_vmtests.sh b/tools/testing/selftests/mm/run_vmtests.sh
>> index 246d53a5d7f2..a5e6ba8d3579 100755
>> --- a/tools/testing/selftests/mm/run_vmtests.sh
>> +++ b/tools/testing/selftests/mm/run_vmtests.sh
>> @@ -248,6 +248,9 @@ CATEGORY="hugetlb" run_test ./map_hugetlb
>>  CATEGORY="hugetlb" run_test ./hugepage-mremap
>>  CATEGORY="hugetlb" run_test ./hugepage-vmemmap
>>  CATEGORY="hugetlb" run_test ./hugetlb-madvise
>> +CATEGORY="hugetlb" run_test ./charge_reserved_hugetlb.sh
>> +CATEGORY="hugetlb" run_test ./hugetlb_reparenting_test.sh
> 
> These 2 tests are failing because the test scripts are not exported. You will
> need to add them to the TEST_FILES variable in the Makefile.
This must be done. I'll investigate even after adding them if these scripts
are robust enough to pass.

> 
>> +CATEGORY="hugetlb" run_test ./hugetlb-read-hwpoison
> 
> The addition of this test causes 2 later tests to fail with ENOMEM. I suspect
> its a side-effect of marking the hugetlbs as hwpoisoned? (just a guess based on
> the test name!). Once a page is marked poisoned, is there a way to un-poison it?
> If not, I suspect that's why it wasn't part of the standard test script in the
> first place.
hugetlb-read-hwpoison failed as probably the fix in the kernel for the test
hasn't been merged in the kernel. The other tests (uffd-stress) aren't
failing on my end and on CI [1][2]

[1] https://lava.collabora.dev/scheduler/job/12577207#L3677
[2] https://lava.collabora.dev/scheduler/job/12577229#L4027

Maybe its configurations issue which is exposed now. Not sure. Maybe
hugetlb-read-hwpoison is changing some configuration and not restoring it.
Maybe your system has less number of hugetlb pages.

> 
> These are the tests that start failing:
> 
> # # ------------------------------------
> # # running ./uffd-stress hugetlb 128 32
> # # ------------------------------------
> # # nr_pages: 64, nr_pages_per_cpu: 8
> # # ERROR: context init failed (errno=12, @uffd-stress.c:254)
> # # [FAIL]
> # not ok 18 uffd-stress hugetlb 128 32 # exit=1
> # # --------------------------------------------
> # # running ./uffd-stress hugetlb-private 128 32
> # # --------------------------------------------
> # # nr_pages: 64, nr_pages_per_cpu: 8
> # # bounces: 31, mode: rnd racing ver poll, ERROR: UFFDIO_COPY error: -12ERROR:
> UFFDIO_COPY error: -12 (errno=12, @uffd-common.c:614)
> # #  (errno=12, @uffd-common.c:614)
> # # [FAIL]
> 
> Quickest way to repo is:
> 
> $ sudo ./run_vmtests.sh -t "userfaultfd hugetlb"
> 
> Thanks,
> Ryan
> 
> 
>>  
>>  nr_hugepages_tmp=$(cat /proc/sys/vm/nr_hugepages)
>>  # For this test, we need one and just one huge page
> 
> 

-- 
BR,
Muhammad Usama Anjum