Re: v4.1 to v4.7: regression in tsc2005 driver

[Michael Welling]

On Tue, Jul 19, 2016 at 06:44:24PM -0700, Dmitry Torokhov wrote:
> On Tue, Jul 19, 2016 at 08:34:57PM -0500, Michael Welling wrote:
> > On Tue, Jul 19, 2016 at 05:53:07PM -0700, Dmitry Torokhov wrote:
> > > On Tue, Jul 19, 2016 at 07:39:08PM -0500, Michael Welling wrote:
> > > > On Tue, Jul 19, 2016 at 04:51:20PM -0700, Dmitry Torokhov wrote:
> > > > > On Sun, Jul 17, 2016 at 05:56:36PM -0500, Michael Welling wrote:
> > > > > > On Sun, Jul 17, 2016 at 10:03:39PM +0200, Pavel Machek wrote:
> > > > > > > On Sun 2016-07-17 13:51:34, Michael Welling wrote:
> > > > > > > > On Sun, Jul 17, 2016 at 08:42:09PM +0200, Pavel Machek wrote:
> > > > > > > > > On Sun 2016-07-17 13:24:45, Michael Welling wrote:
> > > > > > > > > > On Sun, Jul 17, 2016 at 07:52:57PM +0200, Pavel Machek wrote:
> > > > > > > > > > > Hi!
> > > > > > > > > > > 
> > > > > > > > > > > tsc2005 driver changed input device name, from
> > > > > > > > > > > 
> > > > > > > > > > > drivers/input/touchscreen/tsc2005.c:	  input_dev->name = "TSC2005
> > > > > > > > > > > touchscreen";
> > > > > > > > > > > 
> > > > > > > > > > > to "TSC200X touchscreen". Unfortunately, X seems to propagate that
> > > > > > > > > > > name to userspace, where it is needed to be able to do
> > > > > 
> > > > > Technically X _is_ userspace.
> > > > > 
> > > > > > > > > > > 
> > > > > > > > > > > xinput --set-prop --type=int ...
> > > > > > > > > > > 
> > > > > > > > > > > with the right arguments to calibrate touchscreen. (Touchscreen is
> > > > > > > > > > > unusable without calibration).
> > > > > > > > > > > 
> > > > > > > > > > > What to do with that?
> > > > > 
> > > > > Hmm, I do not think we ever committed for the device names to be stable.
> > > > > You are supposed to locate touchscreen device based on its properties
> > > > > and you might need some heuristic if you encounter a system with more
> > > > > than one such touchscreen.
> > > > > 
> > > > > > > > > > 
> > > > > > > > > > The input_dev name could be passed to the common probe function.
> > > > > > > > > > 
> > > > > > > > > > http://lxr.free-electrons.com/source/drivers/input/touchscreen/tsc2005.c#L65
> > > > > > > > > 
> > > > > > > > > That would be preffered, I guess.
> > > > > > > > > 
> > > > > > > > > How many stable releases are affected?
> > > > > > > > 
> > > > > > > > Well this patch is 9 months old now. Lets see.
> > > > > > > > 
> > > > > > > > It was introduced in v4.4-rc1. So v4.4, v4.5 and v4.6.
> > > > > > > 
> > > > > > > Ok, thanks for the information. I believe changing it back to
> > > > > > > "TSC2005" version makes sense (and then fixing it in stable).
> > > > > 
> > > > > Do we know how many users are affected?
> > > > 
> > > > Anyone with an old N900 and the smarts to update the kernel.
> > > 
> > > Soo... only Pavel? ;)
> > > 
> > > > 
> > > > > 
> > > > > > 
> > > > > > Lets see if the maintainer has any input before I submit a patch.
> > > > > 
> > > > > I guess we could also pass the input_id structure to tsc200x_probe, fill
> > > > > the proper product ID (2004 or 2005) and derive the name form it.
> > > > >
> > 
> > If we passed the input_dev with the filled product idea it would have to be
> > allocated in the calling probe functions instead of the common probe.
> > 
> > This will lead to more duplicated code which we jumped through so many hoops
> > to avoid.
> 
> Not input_dev, input_id, like :
> 
> static const struct input_id tsc2005_input_id = {
> 	.bustype	= BUS_SPI,
> 	.product	= 2005,
> };
> 
> 
> 
> static int tsc2005_probe(struct spi_device *spi)
> {
> ...
> 	return tsc200x_probe(&spi->dev, spi->irq,
> 			     &tsc2005_input_id,
> 			     devm_regmap_init_spi(spi, &tsc200x_regmap_config),
> 			     tsc2005_cmd);
> };
> 
> and in tsc200x_probe() you'd do:
> 
> 	input_dev->name = dev_kasprintf(dev, GFP_KERNEL, "TSC%04d touchscreen",
> 					tsc_id->product);
> 	if (!input_dev->name)
> 		return -ENOMEM;
> 	...
> 	input_dev->id = *tsc_id;
>

Okay. Easy enough.
 
> Thanks.
> 
> -- 
> Dmitry