Re: [PATCH v5 02/21] selftests/resctrl: Fix compilation issues for global variables

From: Fenghua Yu <fenghua.yu@intel.com>
To: Babu Moger <babu.moger@amd.com>
Cc: Shuah Khan <shuah@kernel.org>, Tony Luck <tony.luck@intel.com>,
	Reinette Chatre <reinette.chatre@intel.com>,
	linux-kselftest <linux-kselftest@vger.kernel.org>,
	linux-kernel <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v5 02/21] selftests/resctrl: Fix compilation issues for global variables
Date: Fri, 12 Mar 2021 21:51:40 +0000	[thread overview]
Message-ID: <YEvibJFnQhCncaqr@otcwcpicx3.sc.intel.com> (raw)
In-Reply-To: <bf2efac2-99bf-5b58-8952-db9291a1eae1@amd.com>

Hi, Babu,

On Fri, Mar 12, 2021 at 01:08:31PM -0600, Babu Moger wrote:
> > From: Fenghua Yu <fenghua.yu@intel.com>
> > Taking a closer look at the usage of these variables reveals that these
> > variables are used only locally to functions such as cqm_resctrl_val()
> 
> %s/ locally to functions/locally in two functions

OK. Will change it.

> > -int get_cbm_mask(char *cache_type)
> > +int get_cbm_mask(char *cache_type, char *cbm_mask)
> >  {
> >  	char cbm_mask_path[1024];
> >  	FILE *fp;
> > 
> > +	if (!cbm_mask)
> > +		return -1;
> 
> Can cbm_mask be NULL? I see it is statically allocated.
> Or should this be if (!(*cbm_mask))? Or did I miss something.

This is a sanity checking. Although current callers do pass statically
allocated cbm_mask to the parameter, future callers may incorrectly pass
un-allocated cbm_mask to the parameter and may cause segmentation fault
without the sanity checking. To debug this kind of issue, the sanity
checking will be very helpful.

So I would keep this sanity checking.

Thanks.

-Fenghua