Re: [PATCH 2/5] drm/panthor: Be robust against runtime PM resume failures in the suspend path

[Boris Brezillon <boris.brezillon@collabora.com>]

On Thu, 14 Nov 2024 11:13:29 +0000
Liviu Dudau <liviu.dudau@arm.com> wrote:

> On Wed, Nov 13, 2024 at 04:42:54PM +0100, Boris Brezillon wrote:
> > The runtime PM resume operation is not guaranteed to succeed, but if it
> > fails, the device should be in a suspended state. Make sure we're robust
> > to resume failures in the unplug path.
> > 
> > Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
> > ---
> >  drivers/gpu/drm/panthor/panthor_device.c | 23 ++++++++---------------
> >  drivers/gpu/drm/panthor/panthor_fw.c     | 14 +++++++++-----
> >  drivers/gpu/drm/panthor/panthor_gpu.c    |  3 ++-
> >  drivers/gpu/drm/panthor/panthor_mmu.c    |  3 ++-
> >  4 files changed, 21 insertions(+), 22 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/panthor/panthor_device.c b/drivers/gpu/drm/panthor/panthor_device.c
> > index db7ba40f771d..8b5d54b2bbb4 100644
> > --- a/drivers/gpu/drm/panthor/panthor_device.c
> > +++ b/drivers/gpu/drm/panthor/panthor_device.c
> > @@ -48,6 +48,8 @@ static int panthor_clk_init(struct panthor_device *ptdev)
> >  
> >  void panthor_device_unplug(struct panthor_device *ptdev)
> >  {
> > +	int ret;
> > +
> >  	/* This function can be called from two different path: the reset work
> >  	 * and the platform device remove callback. drm_dev_unplug() doesn't
> >  	 * deal with concurrent callers, so we have to protect drm_dev_unplug()
> > @@ -74,7 +76,8 @@ void panthor_device_unplug(struct panthor_device *ptdev)
> >  	 */
> >  	mutex_unlock(&ptdev->unplug.lock);
> >  
> > -	drm_WARN_ON(&ptdev->base, pm_runtime_get_sync(ptdev->base.dev) < 0);
> > +	ret = pm_runtime_get_sync(ptdev->base.dev);
> > +	drm_WARN_ON(&ptdev->base, ret < 0);
> >  
> >  	/* Now, try to cleanly shutdown the GPU before the device resources
> >  	 * get reclaimed.
> > @@ -85,7 +88,10 @@ void panthor_device_unplug(struct panthor_device *ptdev)
> >  	panthor_gpu_unplug(ptdev);
> >  
> >  	pm_runtime_dont_use_autosuspend(ptdev->base.dev);
> > -	pm_runtime_put_sync_suspend(ptdev->base.dev);
> > +
> > +	/* If the resume failed, we don't need to suspend here. */
> > +	if (!ret)
> > +		pm_runtime_put_sync_suspend(ptdev->base.dev);
> >  
> >  	/* If PM is disabled, we need to call the suspend handler manually. */
> >  	if (!IS_ENABLED(CONFIG_PM))
> > @@ -541,17 +547,4 @@ int panthor_device_suspend(struct device *dev)
> >  	clk_disable_unprepare(ptdev->clks.core);
> >  	atomic_set(&ptdev->pm.state, PANTHOR_DEVICE_PM_STATE_SUSPENDED);
> >  	return 0;
> > -
> > -err_set_active:  
> 
> Label is being deleted here, but I don't see where the `goto err_set_active` is being removed.

Yes, as mentioned in my own reply, this chunk should be in patch 3.