linux-kernel - Re: [PATCH 2/5] kselftest: Fix vdso_test

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: <8d513be7-a99d-2b7f-6767-a378742ef3fa@arm.com>
Date:   Wed, 26 Jan 2022 12:22:45 +0000
From:   Vincenzo Frascino <vincenzo.frascino@....com>
To:     Cristian Marussi <cristian.marussi@....com>,
        linux-kselftest@...r.kernel.org, linux-kernel@...r.kernel.org
Cc:     shuah@...nel.org
Subject: Re: [PATCH 2/5] kselftest: Fix vdso_test_time to pass on skips

Hi Cristian,

On 1/26/22 10:27 AM, Cristian Marussi wrote:
> When a vDSO symbol is not found, all the testcases in vdso_test_abi usually
> report a SKIP, which, in turn, is reported back to Kselftest as a PASS.
> 
> Testcase vdso_test_time, instead, reporting a SKIP, causes the whole set of
> tests within vdso_test_abi to be considered FAIL when symbol is not found.
> 
> Fix it reporting a PASS when vdso_test_time cannot find the vdso symbol.
> 
> Cc: Vincenzo Frascino <vincenzo.frascino@....com>
> Signed-off-by: Cristian Marussi <cristian.marussi@....com>
> ---
> Seen as a failure on both a JUNO and a Dragonboard on both recent and old
> kernels/testruns:
> 
> root@...-buster-arm64:~# /opt/ksft/vDSO/vdso_test_abi
> [vDSO kselftest] VDSO_VERSION: LINUX_2.6.39
> The time is 1637922136.675304
> The time is 1637922136.675361000
> The resolution is 0 1
> clock_id: CLOCK_REALTIME [PASS]
> The time is 1927.760604900
> The resolution is 0 1
> clock_id: CLOCK_BOOTTIME [PASS]
> The time is 1637922136.675649700
> The resolution is 0 1
> clock_id: CLOCK_TAI [PASS]
> The time is 1637922136.672000000
> The resolution is 0 4000000
> clock_id: CLOCK_REALTIME_COARSE [PASS]
> The time is 1927.761005600
> The resolution is 0 1
> clock_id: CLOCK_MONOTONIC [PASS]
> The time is 1927.761132780
> The resolution is 0 1
> clock_id: CLOCK_MONOTONIC_RAW [PASS]
> The time is 1927.757093740
> The resolution is 0 4000000
> clock_id: CLOCK_MONOTONIC_COARSE [PASS]
> Could not find __kernel_time              <<< This caused a FAIL as a whole
> root@...-buster-arm64:~# echo $?
> 1
> 
> e.g.: https://lkft.validation.linaro.org/scheduler/job/2192570#L27778
> ---
>  tools/testing/selftests/vDSO/vdso_test_abi.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/testing/selftests/vDSO/vdso_test_abi.c b/tools/testing/selftests/vDSO/vdso_test_abi.c
> index 3d603f1394af..7dcc66d1cecf 100644
> --- a/tools/testing/selftests/vDSO/vdso_test_abi.c
> +++ b/tools/testing/selftests/vDSO/vdso_test_abi.c
> @@ -90,8 +90,9 @@ static int vdso_test_time(void)
>  		(vdso_time_t)vdso_sym(version, name[2]);
>  
>  	if (!vdso_time) {
> +		/* Skip if symbol not found: consider skipped tests as passed */
>  		printf("Could not find %s\n", name[2]);
> -		return KSFT_SKIP;
> +		return KSFT_PASS;

My preference would be to keep "KSFT_SKIP" here and verify separately the return
status of each test. This would maintain compliance with the kselftest API.
Could you please test the patch in-reply-to this one (will be sent shortly) and
let me know if it works for you?

If it does feel free to fold it in the next version of your series with your
"Tested-by:" otherwise let me know.

Thanks!

>  	}
>  
>  	long ret = vdso_time(NULL);
> 

-- 
Regards,
Vincenzo