Re: [PATCH v2] drm/xe: Send 'none' recovery method for XE_WEDGED_MODE_UPON_ANY_HANG_NO_RESET

[Rodrigo Vivi <rodrigo.vivi@intel.com>]

On Sat, Feb 28, 2026 at 01:21:17PM +0530, Raag Jadav wrote:
> XE_WEDGED_MODE_UPON_ANY_HANG_NO_RESET is intended for debugging hangs,
> so wedge the device with 'none' recovery method and have it available
> to the user for debugging.
> 
> Signed-off-by: Raag Jadav <raag.jadav@intel.com>
> ---
> v2: Use 'none' instead of 'unknown' recovery method (Rodrigo)
> ---
>  drivers/gpu/drm/xe/xe_device.c | 9 ++++++++-
>  1 file changed, 8 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
> index 4b68a2d55651..7f74bbbc7d2c 100644
> --- a/drivers/gpu/drm/xe/xe_device.c
> +++ b/drivers/gpu/drm/xe/xe_device.c
> @@ -1318,8 +1318,15 @@ void xe_device_declare_wedged(struct xe_device *xe)
>  		xe_gt_declare_wedged(gt);
>  
>  	if (xe_device_wedged(xe)) {
> +		/*
> +		 * XE_WEDGED_MODE_UPON_ANY_HANG_NO_RESET is intended for debugging
> +		 * hangs, so wedge the device with 'none' recovery method and have
> +		 * it available to the user for debugging.
> +		 */
> +		if (xe->wedged.mode == XE_WEDGED_MODE_UPON_ANY_HANG_NO_RESET)
> +			xe_device_set_wedged_method(xe, DRM_WEDGE_RECOVERY_NONE);

Don't we need to update that documentation you pointed out, now that none is not
a special amd case anymore, but rather meaning 'none action needed' ?!

Then we would probably trigger the right discussion I believe

>  		/* If no wedge recovery method is set, use default */
> -		if (!xe->wedged.method)
> +		else if (!xe->wedged.method)
>  			xe_device_set_wedged_method(xe, DRM_WEDGE_RECOVERY_REBIND |
>  						    DRM_WEDGE_RECOVERY_BUS_RESET);
>  
> -- 
> 2.43.0
>