On 3/18/2026 12:10 PM, Mallesh Koujalagi wrote:
Introduce DRM_WEDGE_RECOVERY_COLD_RESET (BIT(4)) recovery method to
handle power management unit errors requiring complete device power
cycling.
Same as other comments, do not use power management unit errors.
Keep the documentation in drm-layer vendor agnostic so that anyone can
use this method for
any scenario that requires cold-reset
Thanks
Riana
This method addresses scenarios where recovery mechanisms
(driver reload, PCIe reset, etc.) are insufficient to restore
device functionality. When set, it indicates to userspace that
only a full cold reset can recover the device from its current error
state. The cold reset method serves as a last resort for power management
unit errors.
Signed-off-by: Mallesh Koujalagi <[email protected]>
---
drivers/gpu/drm/drm_drv.c | 2 ++
include/drm/drm_device.h | 1 +
2 files changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 6b965c3d3307..2dace9070531 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -535,6 +535,8 @@ static const char *drm_get_wedge_recovery(unsigned int opt)
return "bus-reset";
case DRM_WEDGE_RECOVERY_VENDOR:
return "vendor-specific";
+ case DRM_WEDGE_RECOVERY_COLD_RESET:
+ return "cold-reset";
default:
return NULL;
}
diff --git a/include/drm/drm_device.h b/include/drm/drm_device.h
index bc78fb77cc27..3e386eb42023 100644
--- a/include/drm/drm_device.h
+++ b/include/drm/drm_device.h
@@ -37,6 +37,7 @@ struct pci_controller;
#define DRM_WEDGE_RECOVERY_REBIND BIT(1) /* unbind + bind driver */
#define DRM_WEDGE_RECOVERY_BUS_RESET BIT(2) /* unbind + reset bus device +
bind */
#define DRM_WEDGE_RECOVERY_VENDOR BIT(3) /* vendor specific recovery
method */
+#define DRM_WEDGE_RECOVERY_COLD_RESET BIT(4) /* full device cold reset */
/**
* struct drm_wedge_task_info - information about the guilty task of a wedge
dev
