Re: [PATCH] [media] soc-camera: Remove bogus devm_kfree() in soc_of_bind()

2015-02-19 Thread Geert Uytterhoeven
Hi Guennadi, On Wed, Feb 18, 2015 at 10:16 PM, Guennadi Liakhovetski wrote: > On Wed, 18 Feb 2015, Sergei Shtylyov wrote: >> On 02/18/2015 07:32 PM, Geert Uytterhoeven wrote: >> > Unlike scan_async_group(), soc_of_bind() doesn't allocate its >> > soc_camera_async_client structure using devm_kzall

Re: [PATCH] [media] soc-camera: Remove bogus devm_kfree() in soc_of_bind()

2015-02-18 Thread Guennadi Liakhovetski
On Wed, 18 Feb 2015, Sergei Shtylyov wrote: > Hello. > > On 02/18/2015 07:32 PM, Geert Uytterhoeven wrote: > > > Unlike scan_async_group(), soc_of_bind() doesn't allocate its > > soc_camera_async_client structure using devm_kzalloc(), but has it > > embedded inside the soc_of_info structure. He

Re: [PATCH] [media] soc-camera: Remove bogus devm_kfree() in soc_of_bind()

2015-02-18 Thread Sergei Shtylyov
Hello. On 02/18/2015 07:32 PM, Geert Uytterhoeven wrote: Unlike scan_async_group(), soc_of_bind() doesn't allocate its soc_camera_async_client structure using devm_kzalloc(), but has it embedded inside the soc_of_info structure. Hence on failure, it must not free it using devm_kfree(), as this

[PATCH] [media] soc-camera: Remove bogus devm_kfree() in soc_of_bind()

2015-02-18 Thread Geert Uytterhoeven
Unlike scan_async_group(), soc_of_bind() doesn't allocate its soc_camera_async_client structure using devm_kzalloc(), but has it embedded inside the soc_of_info structure. Hence on failure, it must not free it using devm_kfree(), as this will cause a warning, and may cause slab corruption: so