[alsa-devel] [PATCH] ALSA: control: confirm to return all identical information in 'activate' event data

[Takashi Sakamoto]

On Feb 10 2015 01:18, Takashi Iwai wrote:
> At Mon, 09 Feb 2015 17:02:51 +0100,
> Takashi Iwai wrote:
>>
>> At Tue, 10 Feb 2015 01:01:15 +0900,
>> Takashi Sakamoto wrote:
>>>
>>> On Feb 10 2015 00:28, Takashi Iwai wrote:
>>>> At Tue, 10 Feb 2015 00:02:06 +0900,
>>>> Takashi Sakamoto wrote:
>>>>>
>>>>> When event originator doesn't set numerical ID in identical information,
>>>>> the event data includes no numerical ID, thus userspace applications
>>>>> cannot identify the control just by unique ID in event data.
>>>>>
>>>>> This commit fix this bug so as the event data includes all of identical
>>>>> information.
>>>>>
>>>>> Signed-off-by: Takashi Sakamoto <o-takashi at sakamocchi.jp>
>>>>
>>>> You can't overwrite id pointer in this case.  It's a caller's object.
>>>
>>> Exactly. I missed it...
>>>
>>>> One way to fix would be to copy the id instance.  Another way would be
>>>> to change up/down_write() with _read(), and include snd_ctl_notify()
>>>> call with kctl->id inside the semaphore lock.
>>>>
>>>> The former would be less changes but consume the stack significantly.
>>>
>>> If any sub-effects are allowed, we can use caller's data via the
>>> pointer, like:
>>> memcpy(id, &kctl->id, sizeof(struct snd_ctl_elem_id);
>>> Then:
>>> snd_ctl_notify(card, SNDRV_CTL_EVENT_MASK_INFO, id);
>>
>> That's what I suggested as the first example.

I misunderstood to use heap.

> ... and I think this is a better way in the end.  Simpler, better.
> The stack usage is very likely acceptable.

I think so. I'll submit again with some comments about the sub-effect to 
callers.


Thanks

Takashi Sakamoto