Re: [RFC PATCH] Allow optional module parameters

[Rusty Russell <rusty@rustcorp.com.au>]

Lucas De Marchi <lucas.demarchi@profusion.mobi> writes:
> Hi Rusty,
>
> On Mon, Mar 18, 2013 at 11:32 PM, Rusty Russell <rusty@rustcorp.com.au> wrote:
>> Andy Lutomirski <luto@amacapital.net> writes:
>>> On Sun, Mar 17, 2013 at 7:24 PM, Rusty Russell <rusty@rustcorp.com.au> wrote:
>>>> Andy Lutomirski <luto@amacapital.net> writes:
>>>>> On Thu, Mar 14, 2013 at 10:03 PM, Rusty Russell <rusty@rustcorp.com.au> wrote:
>>>>>> Err, yes.  Don't remove module parameters, they're part of the API.  Do
>>>>>> you have a particular example?
>>>>>
>>>>> So things like i915.i915_enable_ppgtt, which is there to enable
>>>>> something experimental, needs to stay forever once the relevant
>>>>> feature becomes non-experimental and non-optional?  This seems silly.
>> ...
>>>>> Having the module parameter go away while still allowing the module to
>>>>> load seems like a good solution (possibly with a warning in the logs
>>>>> so the user can eventually delete the parameter).
>>>>
>>>> Why not do that for *every* missing parameter then?  Why have this weird
>>>> notation where the user must know that the parameter might one day go
>>>> away?
>>>
>>> Fair enough.  What about the other approach, then?  Always warn if an
>>> option doesn't match (built-in or otherwise) but load the module
>>> anyways.
>>
>> What does everyone think of this?  Jon, Lucas, does this match your
>> experience?
>>
>> Thanks,
>> Rusty.
>>
>> Subject: modules: don't fail to load on unknown parameters.
>>
>> Although parameters are supposed to be part of the kernel API, experimental
>> parameters are often removed.  In addition, downgrading a kernel might cause
>> previously-working modules to fail to load.
>
> I agree with this reasoning
>
>>
>> On balance, it's probably better to warn, and load the module anyway.
>
> However loading the module anyway would bring at least one drawback:
> if the user made a typo when passing the option the module would load
> anyway and he will probably not even look in the log, since there's
> was no errors from modprobe.
>
> For finit_module we could put a flag to trigger this behavior and
> propagate it to modprobe, but this is not possible with init_module().
> I can't think in any other option right now... do you have any?

No good ones :(

MODULE_PARM_DESC isn't compulsory, so you can't rely on that to tell you
about option names.  

Even if we had a flag, how would you know to set it?  I guess you could
try without then try with, and if it works the second time print a
warning about typos.  But it's still pretty ugly.

We could implement such a flag with a fake "IGNORE_BAD_PARAMS"
parameter, for example.  That would fail nicely on older kernels, too.

Hmmm....
Rusty.