[Intel-gfx] [PATCH] drm/i915: Flush async power domains on probe failure

[Souza, Jose]

On Fri, 2020-04-10 at 15:43 +0100, Chris Wilson wrote:
> Flush the async power domain work after aborting the module probe:
> 
> <3> [307.785552] ODEBUG: free active (active state 0) object type:
> timer_list hint: intel_display_power_put_async_work+0x0/0xf0 [i915]
> 
> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1647
> Fixes: b664259f3fe2 ("drm/i915: split i915_driver_modeset_probe() to
> pre/post irq install")
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Jani Nikula <jani.nikula at intel.com>
> Cc: Ville Syrjälä <ville.syrjala at linux.intel.com>
> Cc: Imre Deak <imre.deak at intel.com>
> ---
>  drivers/gpu/drm/i915/i915_drv.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/gpu/drm/i915/i915_drv.c
> b/drivers/gpu/drm/i915/i915_drv.c
> index a7a3b4b98572..13c1ce244092 100644
> --- a/drivers/gpu/drm/i915/i915_drv.c
> +++ b/drivers/gpu/drm/i915/i915_drv.c
> @@ -996,6 +996,7 @@ int i915_driver_probe(struct pci_dev *pdev, const
> struct pci_device_id *ent)
>  out_cleanup_modeset:
>  	/* FIXME */
>  out_cleanup_hw:
> +	intel_power_domains_driver_remove(i915);

If the failure happened in the first two returns of
i915_driver_modeset_probe_noirq() no call to
intel_power_domains_init_hw() would be made.

I'm also working on this, will send to ML after a round on trybot:
https://patchwork.freedesktop.org/patch/361053/?series=75816&rev=1

>  	i915_driver_hw_remove(i915);
>  	intel_memory_regions_driver_release(i915);
>  	i915_ggtt_driver_release(i915);