mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Nicolas Frattaroli <nicolas.frattaroli@collabora.com>
To: Boris Brezillon <boris.brezillon@collabora.com>,
	Liviu Dudau <liviu.dudau@arm.com>,
	Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
	Maxime Ripard <mripard@kernel.org>,
	Thomas Zimmermann <tzimmermann@suse.de>,
	David Airlie <airlied@gmail.com>, Simona Vetter <simona@ffwll.ch>,
	Chia-I Wu <olvaffe@gmail.com>,
	Karunika Choo <karunika.choo@arm.com>,
	Steven Price <steven.price@arm.com>
Cc: kernel@collabora.com, linux-kernel@vger.kernel.org,
	dri-devel@lists.freedesktop.org
Subject: Re: [PATCH v7 2/4] drm/panthor: Rework panthor_irq::suspended into panthor_irq::state
Date: Mon, 12 Jan 2026 14:39:25 +0100	[thread overview]
Message-ID: <5069507.31r3eYUQgx@workhorse> (raw)
In-Reply-To: <67b943c4-ec28-4d14-bb34-34efabeb5360@arm.com>

On Friday, 9 January 2026 17:05:05 Central European Standard Time Steven Price wrote:
> On 08/01/2026 14:19, Nicolas Frattaroli wrote:
> > To deal with the threaded interrupt handler and a suspend action
> > overlapping, the boolean panthor_irq::suspended is not sufficient.
> > 
> > Rework it into taking several different values depending on the current
> > state, and check it and set it within the IRQ helper functions.
> > 
> > Signed-off-by: Nicolas Frattaroli <nicolas.frattaroli@collabora.com>
> > ---
> >  drivers/gpu/drm/panthor/panthor_device.h | 40 +++++++++++++++++++++++++-------
> >  1 file changed, 32 insertions(+), 8 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/panthor/panthor_device.h b/drivers/gpu/drm/panthor/panthor_device.h
> > index cf76a8abca76..a8c21a7eea05 100644
> > --- a/drivers/gpu/drm/panthor/panthor_device.h
> > +++ b/drivers/gpu/drm/panthor/panthor_device.h
> > @@ -61,6 +61,17 @@ enum panthor_device_pm_state {
> >  	PANTHOR_DEVICE_PM_STATE_SUSPENDING,
> >  };
> >  
> > +enum panthor_irq_state {
> > +	/** @PANTHOR_IRQ_STATE_ACTIVE: IRQ is active and ready to process events. */
> > +	PANTHOR_IRQ_STATE_ACTIVE = 0,
> > +	/** @PANTHOR_IRQ_STATE_PROCESSING: IRQ is currently processing events. */
> > +	PANTHOR_IRQ_STATE_PROCESSING,
> > +	/** @PANTHOR_IRQ_STATE_SUSPENDED: IRQ is suspended. */
> > +	PANTHOR_IRQ_STATE_SUSPENDED,
> > +	/** @PANTHOR_IRQ_STATE_SUSPENDING: IRQ is being suspended. */
> > +	PANTHOR_IRQ_STATE_SUSPENDING,
> > +};
> > +
> >  /**
> >   * struct panthor_irq - IRQ data
> >   *
> > @@ -76,8 +87,8 @@ struct panthor_irq {
> >  	/** @mask: Values to write to xxx_INT_MASK if active. */
> >  	u32 mask;
> >  
> > -	/** @suspended: Set to true when the IRQ is suspended. */
> > -	atomic_t suspended;
> > +	/** @state: one of &enum panthor_irq_state reflecting the current state. */
> > +	atomic_t state;
> >  
> >  	/** @mask_lock: protects modifications to _INT_MASK and @mask */
> >  	spinlock_t mask_lock;
> > @@ -415,7 +426,7 @@ static irqreturn_t panthor_ ## __name ## _irq_raw_handler(int irq, void *data)
> >  												\
> >  	guard(spinlock_irqsave)(&pirq->mask_lock);						\
> >  												\
> > -	if (atomic_read(&pirq->suspended))							\
> > +	if (atomic_read(&pirq->state) == PANTHOR_IRQ_STATE_SUSPENDED)				\
> 
> Do we want to also catch the STATE_SUSPENDING case here? AFAICT in
> SUSPENDING we should be confident that _INT_MASK==0 so this shouldn't a
> problem (_INT_STAT should read as 0 below). But we don't want interrupts
> during STATE_SUSPENDING so we might as well handle it.

Depends on what we want to happen here, I think. If the suspend handler
writing 0 to _INT_MASK does not also happen to clear _INT_STAT, then as
far as I can see, we may enter _irq_raw_handler, block for the lock as
suspend drops it, and then enter the function with STATE_SUSPENDING in
a context where we'd probably want to process the remaining interrupts
(i.e. suspend is at synchronise_irq() or about to be there, and we're
handling the last IRQ that got raised before mask was written to 0).

Let me know if my understanding here is correct, because I think in such
a case, it's reasonable to process that last IRQ instead of discarding it,
unless that's something frowned upon that I'm not aware of.

> 
> Thanks,
> Steve
> 
> >  		return IRQ_NONE;								\
> >  	if (!gpu_read(ptdev, __reg_prefix ## _INT_STAT))					\
> >  		return IRQ_NONE;								\
> > @@ -428,11 +439,14 @@ static irqreturn_t panthor_ ## __name ## _irq_threaded_handler(int irq, void *da
> >  {												\
> >  	struct panthor_irq *pirq = data;							\
> >  	struct panthor_device *ptdev = pirq->ptdev;						\
> > +	enum panthor_irq_state state;								\
> >  	irqreturn_t ret = IRQ_NONE;								\
> >  	u32 mask;										\
> >  												\
> >  	scoped_guard(spinlock_irqsave, &pirq->mask_lock) {					\
> >  		mask = pirq->mask;								\
> > +		atomic_cmpxchg(&pirq->state, PANTHOR_IRQ_STATE_ACTIVE,				\
> > +			       PANTHOR_IRQ_STATE_PROCESSING);					\
> >  	}											\
> >  												\
> >  	while (true) {										\
> > @@ -446,11 +460,14 @@ static irqreturn_t panthor_ ## __name ## _irq_threaded_handler(int irq, void *da
> >  	}											\
> >  												\
> >  	scoped_guard(spinlock_irqsave, &pirq->mask_lock) {					\
> > -		if (!atomic_read(&pirq->suspended)) {						\
> > +		state = atomic_read(&pirq->state);						\
> > +		if (state != PANTHOR_IRQ_STATE_SUSPENDED &&					\
> > +		    state != PANTHOR_IRQ_STATE_SUSPENDING) {					\
> >  			/* Only restore the bits that were used and are still enabled */	\
> >  			gpu_write(ptdev, __reg_prefix ## _INT_MASK,				\
> >  				  gpu_read(ptdev, __reg_prefix ## _INT_MASK) |			\
> >  				  (mask & pirq->mask));						\
> > +			atomic_set(&pirq->state, PANTHOR_IRQ_STATE_ACTIVE);			\
> >  		}										\
> >  	}											\
> >  												\
> > @@ -461,16 +478,17 @@ static inline void panthor_ ## __name ## _irq_suspend(struct panthor_irq *pirq)
> >  {												\
> >  	scoped_guard(spinlock_irqsave, &pirq->mask_lock) {					\
> >  		gpu_write(pirq->ptdev, __reg_prefix ## _INT_MASK, 0);				\
> > +		atomic_set(&pirq->state, PANTHOR_IRQ_STATE_SUSPENDING);				\
> >  	}											\
> >  	synchronize_irq(pirq->irq);								\
> > -	atomic_set(&pirq->suspended, true);							\
> > +	atomic_set(&pirq->state, PANTHOR_IRQ_STATE_SUSPENDED);					\
> >  }												\
> >  												\
> >  static inline void panthor_ ## __name ## _irq_resume(struct panthor_irq *pirq)			\
> >  {												\
> >  	guard(spinlock_irqsave)(&pirq->mask_lock);						\
> >  												\
> > -	atomic_set(&pirq->suspended, false);							\
> > +	atomic_set(&pirq->state, PANTHOR_IRQ_STATE_ACTIVE);					\
> >  	gpu_write(pirq->ptdev, __reg_prefix ## _INT_CLEAR, pirq->mask);				\
> >  	gpu_write(pirq->ptdev, __reg_prefix ## _INT_MASK, pirq->mask);				\
> >  }												\
> > @@ -494,19 +512,25 @@ static int panthor_request_ ## __name ## _irq(struct panthor_device *ptdev,			\
> >  												\
> >  static inline void panthor_ ## __name ## _irq_enable_events(struct panthor_irq *pirq, u32 mask)	\
> >  {												\
> > +	enum panthor_irq_state state;								\
> > +												\
> >  	guard(spinlock_irqsave)(&pirq->mask_lock);						\
> >  												\
> >  	pirq->mask |= mask;									\
> > -	if (!atomic_read(&pirq->suspended))							\
> > +	state = atomic_read(&pirq->state);							\
> > +	if (state != PANTHOR_IRQ_STATE_SUSPENDED && state != PANTHOR_IRQ_STATE_SUSPENDING)	\
> >  		gpu_write(pirq->ptdev, __reg_prefix ## _INT_MASK, pirq->mask);			\
> >  }												\
> >  												\
> >  static inline void panthor_ ## __name ## _irq_disable_events(struct panthor_irq *pirq, u32 mask)\
> >  {												\
> > +	enum panthor_irq_state state;								\
> > +												\
> >  	guard(spinlock_irqsave)(&pirq->mask_lock);						\
> >  												\
> >  	pirq->mask &= ~mask;									\
> > -	if (!atomic_read(&pirq->suspended))							\
> > +	state = atomic_read(&pirq->state);							\
> > +	if (state != PANTHOR_IRQ_STATE_SUSPENDED && state != PANTHOR_IRQ_STATE_SUSPENDING)	\
> >  		gpu_write(pirq->ptdev, __reg_prefix ## _INT_MASK, pirq->mask);			\
> >  }
> >  
> > 
> 
> 





  reply	other threads:[~2026-01-12 13:39 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-08 14:19 [PATCH v7 0/4] Add a few tracepoints to panthor Nicolas Frattaroli
2026-01-08 14:19 ` [PATCH v7 1/4] drm/panthor: Extend IRQ helpers for mask modification/restoration Nicolas Frattaroli
2026-01-09 15:59   ` Steven Price
2026-01-11 11:39     ` Nicolas Frattaroli
2026-01-08 14:19 ` [PATCH v7 2/4] drm/panthor: Rework panthor_irq::suspended into panthor_irq::state Nicolas Frattaroli
2026-01-09 16:05   ` Steven Price
2026-01-12 13:39     ` Nicolas Frattaroli [this message]
2026-01-12 14:09       ` Steven Price
2026-01-08 14:19 ` [PATCH v7 3/4] drm/panthor: Add tracepoint for hardware utilisation changes Nicolas Frattaroli
2026-01-08 14:19 ` [PATCH v7 4/4] drm/panthor: Add gpu_job_irq tracepoint Nicolas Frattaroli
2026-01-09 16:23   ` Steven Price
2026-01-11 11:49     ` Nicolas Frattaroli
2026-01-12  9:15       ` Steven Price

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5069507.31r3eYUQgx@workhorse \
    --to=nicolas.frattaroli@collabora.com \
    --cc=airlied@gmail.com \
    --cc=boris.brezillon@collabora.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=karunika.choo@arm.com \
    --cc=kernel@collabora.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=liviu.dudau@arm.com \
    --cc=maarten.lankhorst@linux.intel.com \
    --cc=mripard@kernel.org \
    --cc=olvaffe@gmail.com \
    --cc=simona@ffwll.ch \
    --cc=steven.price@arm.com \
    --cc=tzimmermann@suse.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®