Re: [PATCH v1 1/1] drivers: base: Fix NULL pointer exception in __platform_driver_probe()
On Tue, Apr 28, 2020 at 12:24:05PM -0700, Kuppuswamy, Sathyanarayanan wrote: > Hi Greg, > > On 4/28/20 12:03 PM, Greg KH wrote: > > On Wed, Apr 08, 2020 at 02:40:03PM -0700, > > sathyanarayanan.kuppusw...@linux.intel.com wrote: > > > From: Kuppuswamy Sathyanarayanan > > > > > > > > > If platform bus driver registration is failed then, > > > > How is your platform driver registration failing? What caused that, is > > it an in-kernel problem with something? How can it be triggered? > In my case I triggered it intentionally. For one of our internal > project we want to strictly control the number of drivers/devices > allowed in kernel. To verify the feasibility of adding above support, > I intentionally failed few bus drivers and checked the behavior. In > one of those tests I hence came across the mentioned issue. > > In any case, platform bus driver registration failure is a valid case > right ? Any issue we notice when this happens needs to be handled right? That's fine, I just need to know if this is something that someone can actually trigger today, and needs to be fixed up, or if this is just a "hardening for when a driver author does something foolish" type of a case. It seems to be the "don't do foolish things" to me, so it's a much lower priority as we can always fix foolish drivers because we have the source to them :) thanks, greg k-h
Re: [PATCH v1 1/1] drivers: base: Fix NULL pointer exception in __platform_driver_probe()
Hi Greg, On 4/28/20 12:03 PM, Greg KH wrote: On Wed, Apr 08, 2020 at 02:40:03PM -0700, sathyanarayanan.kuppusw...@linux.intel.com wrote: From: Kuppuswamy Sathyanarayanan If platform bus driver registration is failed then, How is your platform driver registration failing? What caused that, is it an in-kernel problem with something? How can it be triggered? In my case I triggered it intentionally. For one of our internal project we want to strictly control the number of drivers/devices allowed in kernel. To verify the feasibility of adding above support, I intentionally failed few bus drivers and checked the behavior. In one of those tests I hence came across the mentioned issue. In any case, platform bus driver registration failure is a valid case right ? Any issue we notice when this happens needs to be handled right? thanks, greg k-h
Re: [PATCH v1 1/1] drivers: base: Fix NULL pointer exception in __platform_driver_probe()
On Wed, Apr 08, 2020 at 02:40:03PM -0700, sathyanarayanan.kuppusw...@linux.intel.com wrote: > From: Kuppuswamy Sathyanarayanan > > If platform bus driver registration is failed then, How is your platform driver registration failing? What caused that, is it an in-kernel problem with something? How can it be triggered? thanks, greg k-h