linux-kernel - Re: [PATCH v2 1/4] selftests/resctrl: Fix set up shemata with 100% allocation on first run in MBM 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: <10ca9b73-b3e1-4139-6cbf-c362467c7943@intel.com>
Date:   Wed, 5 Oct 2022 13:56:21 -0700
From:   Reinette Chatre <reinette.chatre@...el.com>
To:     Shaopeng Tan <tan.shaopeng@...fujitsu.com>,
        Fenghua Yu <fenghua.yu@...el.com>,
        Shuah Khan <shuah@...nel.org>
CC:     <linux-kernel@...r.kernel.org>, <linux-kselftest@...r.kernel.org>
Subject: Re: [PATCH v2 1/4] selftests/resctrl: Fix set up shemata with 100%
 allocation on first run in MBM test.

Hi Shaopeng,

I understand there is a typo in the code you are modifying but I do not think
that justifies the same typo in the subject: shemata -> schemata

On 10/4/2022 6:39 PM, Shaopeng Tan wrote:
> There is a comment "Set up shemata with 100% allocation on the first run"
> in function mbm_setup(), but there is an increment bug and the condition
> "num_of_runs == 0" will never be met and write_schemata() will never be
> called to set schemata to 100%. This is currently fine because
> resctl_val_parm->num_resctrlfs is always 1 and umount/mount will be run

resctl_val_parm -> resctrl_val_param
num_resctrlfs -> mum_resctrlfs

> in each test to set the schemata to 100%.
> 
> To make mbm_setup() future code-change proof, fix to call

You could be more specific by indicating that this change will 
support the usage when the test does not unmount/remount resctrl before
the test.

> write-schemata() properly when the function is called for the first time.

write-schemata() -> write_schemata()

> 
> Also, remove static local variable 'num_of_runs' because this is not
> needed as there is resctl_val_param->num_of_runs which should be used

resctl_val_param -> resctrl_val_param

> instead like in cat_setup().

With the move to using a member from the struct I think it would make the
code easier to understand if num_of_runs is explicitly initialized. That is
indeed the norm when looking at the other call sites.

> 
> Signed-off-by: Shaopeng Tan <tan.shaopeng@...fujitsu.com>
> ---
>  tools/testing/selftests/resctrl/mbm_test.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/tools/testing/selftests/resctrl/mbm_test.c b/tools/testing/selftests/resctrl/mbm_test.c
> index 8392e5c55ed0..4a54be314402 100644
> --- a/tools/testing/selftests/resctrl/mbm_test.c
> +++ b/tools/testing/selftests/resctrl/mbm_test.c
> @@ -89,12 +89,11 @@ static int check_results(int span)
>  static int mbm_setup(int num, ...)
>  {
>  	struct resctrl_val_param *p;

p is defined here ...

> -	static int num_of_runs;
>  	va_list param;
>  	int ret = 0;
>  
>  	/* Run NUM_OF_RUNS times */
> -	if (num_of_runs++ >= NUM_OF_RUNS)
> +	if (p->num_of_runs >= NUM_OF_RUNS)

p is used here _before_ it is initialized in the code that follows. 

>  		return -1;
>  
>  	va_start(param, num);
> @@ -102,9 +101,10 @@ static int mbm_setup(int num, ...)
>  	va_end(param);
>  
>  	/* Set up shemata with 100% allocation on the first run. */
> -	if (num_of_runs == 0)
> +	if (p->num_of_runs == 0)
>  		ret = write_schemata(p->ctrlgrp, "100", p->cpu_no,
>  				     p->resctrl_val);
> +	p->num_of_runs++;
>  
>  	return ret;
>  }

Reinette