linux-kernel - Re: [PATCH 2/2] kselftests: mm: Fail the test if userfaultfd syscall isn't found

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <2f6cfc9b-4f3b-4f64-9c4d-63e3453b3105@collabora.com>
Date: Mon, 16 Sep 2024 11:33:58 +0500
From: Muhammad Usama Anjum <Usama.Anjum@...labora.com>
To: Shuah Khan <skhan@...uxfoundation.org>,
 Andrew Morton <akpm@...ux-foundation.org>, Shuah Khan <shuah@...nel.org>
Cc: Usama.Anjum@...labora.com, kernel@...labora.com, linux-mm@...ck.org,
 linux-kselftest@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/2] kselftests: mm: Fail the test if userfaultfd syscall
 isn't found

On 9/12/24 10:28 PM, Shuah Khan wrote:
> On 9/12/24 10:10, Shuah Khan wrote:
>> On 9/12/24 04:31, Muhammad Usama Anjum wrote:
>>> The userfaultfd is enabled in the config fragment of mm selftest suite.
>>> It must always be present. If it isn't present, we should throw error
>>> and not just skip. This would have helped us catch the test breakage.
>>
>> Please elaborate on this to help understand the what breakage was
>> missed.
>>
>> Also this commit log doesn't look right to me. syscall() could
>> fail for any reason. Do you mean to see skip is incorrect in this
>> error leg? Please see comments below.
>>
>>> Adding this now to catch the future breakages.
>>>
>>> Signed-off-by: Muhammad Usama Anjum <usama.anjum@...labora.com>
>>> ---
>>>   tools/testing/selftests/mm/pagemap_ioctl.c | 2 +-
>>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/tools/testing/selftests/mm/pagemap_ioctl.c
>>> b/tools/testing/selftests/mm/pagemap_ioctl.c
>>> index bcc73b4e805c6..d83dda8edf62c 100644
>>> --- a/tools/testing/selftests/mm/pagemap_ioctl.c
>>> +++ b/tools/testing/selftests/mm/pagemap_ioctl.c
>>> @@ -95,7 +95,7 @@ int init_uffd(void)
>>>       uffd = syscall(__NR_userfaultfd, O_CLOEXEC | O_NONBLOCK |
>>> UFFD_USER_MODE_ONLY);
>>>       if (uffd == -1)
>>> -        return uffd;
>>> +        ksft_exit_fail_perror("Userfaultfd syscall failed");
>>
>> This looks wrong to me - Is missing config the only reason this syscall
>> would fail?
> 
> It should still skip if __NR_userfaultfd isn't supported on a release
> or an architecture.
> 
> The real problem seems to be in main():
> 
> if (init_uffd())
>                 ksft_exit_pass();
> 
> 
> Why is this ksft_exit_pass()? Looks like further investigation is
> necessary to understand the problem and fix.
Let's skip this patch as it'll create more noise on unsupported
architectures than catching failures on supported architectures.

> 
> thanks,
> -- Shuah

-- 
BR,
Muhammad Usama Anjum