[PATCH] drm/amdgpu: just disallow reading untouched registers

Christian König deathsimple at vodafone.de
Wed Mar 29 12:18:23 UTC 2017


Ping!

Does anybody of you guys know what the background of those "untouched" 
registers is?

That we leak uninitialized memory to userspace is a bit bad.

Thanks,
Christian.

Am 28.03.2017 um 13:24 schrieb Christian König:
> From: Christian König <christian.koenig at amd.com>
>
> Not sure what the original intention was here, but returning a random piece of
> kernel memory to userspace because we didn't set the value at all is clearly
> not a good idea.
>
> This patch disallows reading the register and returns
> a proper error code instead.
>
> Signed-off-by: Christian König <christian.koenig at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/vi.c | 6 ------
>   1 file changed, 6 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
> index 5c02ec4..f1c2bff 100644
> --- a/drivers/gpu/drm/amd/amdgpu/vi.c
> +++ b/drivers/gpu/drm/amd/amdgpu/vi.c
> @@ -465,15 +465,9 @@ static void vi_detect_hw_virtualization(struct amdgpu_device *adev)
>   }
>   
>   static const struct amdgpu_allowed_register_entry tonga_allowed_read_registers[] = {
> -	{mmGB_MACROTILE_MODE7, true},
>   };
>   
>   static const struct amdgpu_allowed_register_entry cz_allowed_read_registers[] = {
> -	{mmGB_TILE_MODE7, true},
> -	{mmGB_TILE_MODE12, true},
> -	{mmGB_TILE_MODE17, true},
> -	{mmGB_TILE_MODE23, true},
> -	{mmGB_MACROTILE_MODE7, true},
>   };
>   
>   static const struct amdgpu_allowed_register_entry vi_allowed_read_registers[] = {




More information about the amd-gfx mailing list