linux-kernel - Re: [PATCH/RFC v9 04/19] mfd: max77693: adjust max77693_led_platform

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-id: <5487105F.5000003@samsung.com>
Date:	Tue, 09 Dec 2014 16:08:15 +0100
From:	Sylwester Nawrocki <s.nawrocki@...sung.com>
To:	Lee Jones <lee.jones@...aro.org>,
	Jacek Anaszewski <j.anaszewski@...sung.com>
Cc:	linux-leds@...r.kernel.org, linux-media@...r.kernel.org,
	linux-kernel@...r.kernel.org, kyungmin.park@...sung.com,
	b.zolnierkie@...sung.com, pavel@....cz, cooloney@...il.com,
	rpurdie@...ys.net, sakari.ailus@....fi, robh+dt@...nel.org,
	pawel.moll@....com, mark.rutland@....com,
	ijc+devicetree@...lion.org.uk, galak@...eaurora.org,
	Chanwoo Choi <cw00.choi@...sung.com>
Subject: Re: [PATCH/RFC v9 04/19] mfd: max77693: adjust
 max77693_led_platform_data

On 09/12/14 15:41, Lee Jones wrote:
>>>>>>>> struct max77693_led_platform_data {
>>>>>>>> > >>>>>>+	const char *label[2];
>>>>>>>> > >>>>>>  	u32 fleds[2];
>>>>>>>> > >>>>>>  	u32 iout_torch[2];for_each_available_child_of_node
>>>>>>>> > >>>>>>  	u32 iout_flash[2];
>>>>>>>> > >>>>>>  	u32 trigger[2];
>>>>>>>> > >>>>>>  	u32 trigger_type[2];
>>>>>>>> > >>>>>>+	u32 flash_timeout[2];
>>>>>>>> > >>>>>>  	u32 num_leds;
>>>>>>>> > >>>>>>  	u32 boost_mode;
>>>>>>>> > >>>>>>-	u32 flash_timeout;
>>>>>>>> > >>>>>>  	u32 boost_vout;
>>>>>>>> > >>>>>>  	u32 low_vsys;
>>>>>>>> > >>>>>>+	struct device_node *sub_nodes[2];
>>>>>>> > >>>>>
>>>>>>> > >>>>>I haven't seen anyone do this before.  Why can't you use the provided
>>>>>>> > >>>>>OF functions to traverse through your tree?
>>>>>> > >>>>
>>>>>> > >>>>I use for_each_available_child_of_node when parsing DT node, but I
>>>>>> > >>>>need to cache the pointer to sub-node to be able to use it later
>>>>>> > >>>>when it needs to be passed to V4L2 sub-device which is then
>>>>>> > >>>>asynchronously matched by the phandle to sub-node.
>>>>>> > >>>>
>>>>>> > >>>>If it is not well seen to cache it in the platform data then
>>>>>> > >>>>I will find different way to accomplish this.
>>>>> > >>>
>>>>> > >>>I haven't seen the end-driver for this, but why can't you use that
>>>>> > >>>device's of_node pointer?
>>>> > >>
>>>> > >>Maybe it is indeed a good idea. I could pass the of_node pointer
>>>> > >>and the sub-led identifier to the V4L2 sub-device and there look
>>>> > >>for the sub-node containing relevant identifier. The downside
>>>> > >>would be only that for_each_available_child_of_node would
>>>> > >>have to be called twice - in the led driver and in the V4L2 sub-device.
>>>> > >>I think that we can live with it.
>>> > >
>>> > >Are the LED and V4L2 drivers children of this MFD?  If so, you can use
>>> > >the of_compatible attribute in struct mfd_cell to populate the each
>>> > >child's of_node dynamically i.e. the MFD core will do that for you.
>>> > >
>> > 
>> > V4L2 driver wraps LED driver. This way the LED device can be
>> > controlled with use of two interfaces - LED subsystem sysfs
>> > and V4L2 Flash. This is the aim of the whole patch set.
>> > 
>> > I've thought it over again and it seems that I will need to cache
>> > somewhere these sub_nodes pointers. They have to be easily accessible
>> > for the V4L2 sub-device as it can be asynchronously registered
>> > or unregistered within V4L2 media device. Sub-devices are matched
>> > basing on the sub-node phandle.
>
> Not quite getting this.  Can you explain this in another way please?

Only the LED controller driver is a child the MFD. The LED controller
can contain multiple outputs with a physical LED attached to it. AFAICS
this binding is modelling each such an output as a the LED's controller
node child node.

I'm not sure though why storing the device node pointers is required,
rather than traversing OF tree when needed.
I guess we only need the list of the node pointer to populate struct
v4l2_async_subdev array for v4l2_async_notifier_register() call ?


-- 
Regards,
Sylwester
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/