Re: [PATCHv2-next 1/3] sysctl/sysrq: Remove __sysrq_enabled copy

From: Dmitry Safonov <dima@arista.com>
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: linux-kernel@vger.kernel.org,
	Dmitry Safonov <0x7f454c46@gmail.com>,
	Iurii Zaikin <yzaikin@google.com>, Jiri Slaby <jslaby@suse.com>,
	Joe Perches <joe@perches.com>,
	Randy Dunlap <rdunlap@infradead.org>,
	Vasiliy Khoruzhick <vasilykh@arista.com>,
	linux-serial@vger.kernel.org,
	Luis Chamberlain <mcgrof@kernel.org>,
	Kees Cook <keescook@chromium.org>,
	linux-fsdevel@vger.kernel.org
Subject: Re: [PATCHv2-next 1/3] sysctl/sysrq: Remove __sysrq_enabled copy
Date: Sun, 23 Feb 2020 12:08:23 +0000	[thread overview]
Message-ID: <eef8e82a-c254-9391-506b-c9de8e52ee0f@arista.com> (raw)
In-Reply-To: <20200115123601.GA3461986@kroah.com>

On 1/15/20 12:36 PM, Greg Kroah-Hartman wrote:
> On Tue, Jan 14, 2020 at 05:19:10PM +0000, Dmitry Safonov wrote:
[..]
>> +int sysrq_get_mask(void)
>> +{
>> +	if (sysrq_always_enabled)
>> +		return 1;
>> +	return sysrq_enabled;
>> +}
> 
> Naming is hard.  And this name is really hard to understand.

Agree.

> Traditionally get/put are used for incrementing reference counts.  You
> don't have a sysrq_put_mask() call, right?  :)

Yes, fair point

> I think what you want this function to do is, "is sysrq enabled right
> now" (hint, it's a global function, add kernel-doc to it so we know what
> it does...).  If so, it should maybe be something like:
> 
> 	bool sysrq_is_enabled(void);
> 
> which to me makes more sense.

Err, not exactly: there is a function for that which is sysrq_on().
But for sysctl the value of the mask (or 1 for always_enabled) is
actually needed to show a proper value back to the userspace reader.

Thanks,
          Dmitry