[PATCH] drm/amdgpu: ensure device_list is initialised before calling list_add_tail
Andrey Grodzovsky
Andrey.Grodzovsky at amd.com
Sat Apr 11 04:42:54 UTC 2020
On 4/10/20 6:57 PM, Colin King wrote:
> From: Colin Ian King <colin.king at canonical.com>
>
> Currently the call to list_add_tail will access an the uninitalised
> device_list.prev. Fix this by ensuring device_list is initialized before
> adding items to it.
>
> Addresses-Coverity: ("Uninitialized pointer read")
That weird, I see that his is already initialized unconditionally here -
https://elixir.bootlin.com/linux/v5.6.3/source/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c#L4022
Andrey
> Fixes: b3dbd6d3ec49 ("drm/amdgpu: resolve mGPU RAS query instability")
> Signed-off-by: Colin Ian King <colin.king at canonical.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> index b0aa4e1ed4df..caa4969bd46f 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> @@ -1447,6 +1447,7 @@ static void amdgpu_ras_do_recovery(struct work_struct *work)
> if (hive && adev->gmc.xgmi.num_physical_nodes > 1) {
> device_list_handle = &hive->device_list;
> } else {
> + INIT_LIST_HEAD(&device_list);
> list_add_tail(&adev->gmc.xgmi.head, &device_list);
> device_list_handle = &device_list;
> }
More information about the amd-gfx
mailing list