Re: [PATCH/RFC] pci: dynids.use_driver_data considered harmful

From: Milton Miller <miltonm@bga.com>
To: Jean Delvare <khali@linux-fr.org>
Cc: Michael Ellerman <michael@ellerman.id.au>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	Jesse Barnes <jbarnes@virtuousgeek.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-pci@vger.kernel.org, Greg KH <greg@kroah.com>
Subject: Re: [PATCH/RFC] pci: dynids.use_driver_data considered harmful
Date: Tue, 19 Aug 2008 13:01:47 -0500	[thread overview]
Message-ID: <98dd991a397dfe6393cec9f89eae0ba3@bga.com> (raw)
In-Reply-To: <20080817210659.06601a3b@hyperion.delvare>

On Aug 17, 2008, at 2:06 PM, Jean Delvare wrote:
> On Fri, 15 Aug 2008 23:22:59 -0700, Greg KH wrote:
>> On Fri, Aug 15, 2008 at 12:15:01PM -0700, Jesse Barnes wrote:
>>> On Friday, August 15, 2008 11:55 am Jean Delvare wrote:
>>>> In fact we can do even better than that. We could accept from
>>>> user-space only driver_data values which at least one device ID 
>>>> entry in
>>>> the driver already uses. That should be fairly easy to implement, 
>>>> and
>>>> would offer a level of safety an order of magnitude above what we 
>>>> have
>>>> at the moment... And it works both ways: if 0 is not a valid data 
>>>> for
>>>> some driver, that would force the user to provide a non-zero (and
>>>> valid) data value. And it guarantees that the user can't ask for
>>>> something the driver doesn't expect, so drivers don't even need 
>>>> extra
>>>> checks. And no need for a use_driver_data flag either.
>>>
>>> Meaning a driver audit of the usage?  Yeah that would be great.

Thanks Jean for doing this.  Sometimes things move quickly after a long 
stall.  I thought about proposing a similar patch and therefore have to 
say Ack.

>>>> The only drawback is that it prevents the user from passing a "new"
>>>> data value even if it would be valid. But honestly, I don't expect 
>>>> that
>>>> case to happen frequently... if ever at all. So I'd say the benefits
>>>> totally outweight the drawback.

There are a few drivers that could benefit, mainly ones that I 
identified as using flags.  For example, the radeon driver uses 
different fields of the data to specify crt controller, video output 
device, etc.   I'm fine with deferring a flag for such drivers until 
someone audits a driver and wants the support.

>>>>
>>>> If the interested people agree with the idea, I'll look into
>>>> implementing it.
>>>
>>> Well the audit would show if user supplied "new" values are needed; 
>>> otherwise
>>> the approach sounds good to me.
>>
>> That sounds reasonable, and should work properly.
>>
>> No objection from me.

so, if anyone asks,

Concept-Acked-By: Milton Miller <miltonm@bga.com>

milton