On Tue, 24 Oct 2023 12:56:40 +0200, Maxime Ripard wrote:
> The VC4 mock helpers allocate the CRTC, encoders and connectors using a
> call to kunit_kzalloc(), but the DRM device they are attache to survives
> for longer than the test itself which leads to use-after-frees reported
> by KASAN.
>
> Sw
On Tue, 24 Oct 2023 at 18:38, Maíra Canal wrote:
>
> On 10/24/23 07:56, Maxime Ripard wrote:
> > The VC4 mock helpers allocate the CRTC, encoders and connectors using a
> > call to kunit_kzalloc(), but the DRM device they are attache to survives
> > for longer than the test itself which leads to u
On 10/24/23 07:56, Maxime Ripard wrote:
The VC4 mock helpers allocate the CRTC, encoders and connectors using a
call to kunit_kzalloc(), but the DRM device they are attache to survives
for longer than the test itself which leads to use-after-frees reported
by KASAN.
Switch to drmm_kzalloc to tie
The VC4 mock helpers allocate the CRTC, encoders and connectors using a
call to kunit_kzalloc(), but the DRM device they are attache to survives
for longer than the test itself which leads to use-after-frees reported
by KASAN.
Switch to drmm_kzalloc to tie the lifetime of these objects to the main
