On Thu, May 30, 2019 at 10:57:09PM +0000, Matt Sickler wrote: > >From: devel <driverdev-devel-bounces@xxxxxxxxxxxxxxxxxxxxxx> On Behalf Of > >Greg KH > >On Fri, May 24, 2019 at 01:07:59PM +0200, Simon Sandström wrote: > >> --- a/drivers/staging/kpc2000/kpc2000/core.c > >> +++ b/drivers/staging/kpc2000/kpc2000/core.c > >> @@ -276,18 +276,18 @@ static ssize_t kp2000_cdev_read(struct file *filp, > > > >This whole function just needs to be deleted, it's a horrible hack. > > >From the outside, I would definitely agree with you. On the inside though, we > rely on this function to quickly identify what kind and version is running on > a given piece of hardware. Since that same information is provided by an ioctl, > I could be convinced to remove this API and write a userspace application that > uses the ioctl to get the information and pretty prints it. The ioctl needs to die as well, just use the sysfs entries instead, as you already have them :) > I'd be more inclined to agree with the deletion if it means the whole char dev > interface can be removed from the kpc2000 driver. That won't be straightforward > as the ioctl is exposed through this interface. We could remove the ioctl, but > we'd need to ensure that all the same information is exposed via sysfs. I think you are there already, what is missing? > Our userspace side is all funneled through a single class, so changing > it to use sysfs wouldn't be too difficult. I'd support that change if > someone wants to make it. I will be glad to do that, it's always nice to delete code :) thanks, greg k-h _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
