All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jani Nikula <jani.nikula@linux.intel.com>
To: Imre Deak <imre.deak@intel.com>,
	intel-gfx@lists.freedesktop.org, intel-xe@lists.freedesktop.org
Subject: Re: [PATCH v2 2/5] drm/i915/hpd: Add support for suspending the IRQ handling on an HPD pin
Date: Tue, 25 Feb 2025 18:25:01 +0200	[thread overview]
Message-ID: <87bjuq56w2.fsf@intel.com> (raw)
In-Reply-To: <20250225121536.2152940-1-imre.deak@intel.com>

On Tue, 25 Feb 2025, Imre Deak <imre.deak@intel.com> wrote:
> Add support for suspending the IRQ handling on the HPD pin of a given
> encoder, handling IRQs that arrived while in the suspended state after
> resuming the IRQ handling. This will be used by a follow-up change,
> which suspends/resumes the IRQ handling around DP link training.
>
> This is similar to the intel_hpd_disable/enable() functionality, by also
> handling encoders/ports with a pulse handler (i.e. also
> suspending/resuming the short/long pulse handling) and handling the IRQs
> arrived in the suspended state after the handling is resumed (vs. just
> dropping such IRQs).
>
> v2:
> - Handle encoders without a port assigned to them.
> - Fix clearing IRQs from intel_hotplug::short_port_mask.
>
> Signed-off-by: Imre Deak <imre.deak@intel.com>

I think the terms suspend/resume are too loaded to be used for something
that isn't about system or runtime suspend/resume. It's guaranteed to
mislead people.

Maybe:

- pause/continue
- deactivate/activate
- block/unblock

?


BR,
Jani.


> ---
>  .../gpu/drm/i915/display/intel_display_core.h |   1 +
>  drivers/gpu/drm/i915/display/intel_hotplug.c  | 209 +++++++++++++++---
>  drivers/gpu/drm/i915/display/intel_hotplug.h  |   2 +
>  3 files changed, 187 insertions(+), 25 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display_core.h b/drivers/gpu/drm/i915/display/intel_display_core.h
> index 554870d2494b3..152ee4dfcbf38 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_core.h
> +++ b/drivers/gpu/drm/i915/display/intel_display_core.h
> @@ -159,6 +159,7 @@ struct intel_hotplug {
>  	struct {
>  		unsigned long last_jiffies;
>  		int count;
> +		int suspend_count;
>  		enum {
>  			HPD_ENABLED = 0,
>  			HPD_DISABLED = 1,
> diff --git a/drivers/gpu/drm/i915/display/intel_hotplug.c b/drivers/gpu/drm/i915/display/intel_hotplug.c
> index ab8e71c4b0f86..85c56b8add8b0 100644
> --- a/drivers/gpu/drm/i915/display/intel_hotplug.c
> +++ b/drivers/gpu/drm/i915/display/intel_hotplug.c
> @@ -349,19 +349,69 @@ static bool intel_encoder_has_hpd_pulse(struct intel_encoder *encoder)
>  		enc_to_dig_port(encoder)->hpd_pulse != NULL;
>  }
>  
> +static u32 hpd_mask_to_ports(struct intel_display *display, u32 hpd_mask)
> +{
> +	struct intel_encoder *encoder;
> +	u32 ports = 0;
> +
> +	for_each_intel_encoder(display->drm, encoder) {
> +		if (encoder->port == PORT_NONE)
> +			continue;
> +
> +		if (BIT(encoder->hpd_pin) & hpd_mask)
> +			ports |= BIT(encoder->port);
> +	}
> +
> +	return ports;
> +}
> +
> +static u32 get_suspended_hpds(struct intel_display *display)
> +{
> +	struct drm_i915_private *i915 = to_i915(display->drm);
> +	enum hpd_pin pin;
> +	u32 hpds = 0;
> +
> +	lockdep_assert_held(&i915->irq_lock);
> +
> +	for_each_hpd_pin(pin) {
> +		if (display->hotplug.stats[pin].suspend_count)
> +			hpds |= BIT(pin);
> +	}
> +
> +	return hpds;
> +}
> +
> +/*
> + * Get the mask of all the ports using a suspended HPD pin, accounting for
> + * ports sharing their - suspended - HPD pin with another port.
> + */
> +static u32 get_suspended_hpd_ports(struct intel_display *display)
> +{
> +	struct drm_i915_private *i915 = to_i915(display->drm);
> +
> +	lockdep_assert_held(&i915->irq_lock);
> +	return hpd_mask_to_ports(display, get_suspended_hpds(display));
> +}
> +
>  static void i915_digport_work_func(struct work_struct *work)
>  {
> -	struct drm_i915_private *dev_priv =
> -		container_of(work, struct drm_i915_private, display.hotplug.dig_port_work);
> +	struct intel_display *display =
> +		container_of(work, struct intel_display, hotplug.dig_port_work);
> +	struct drm_i915_private *dev_priv = to_i915(display->drm);
> +	struct intel_hotplug *hotplug = &display->hotplug;
>  	u32 long_port_mask, short_port_mask;
>  	struct intel_encoder *encoder;
> +	u32 suspended_ports;
>  	u32 old_bits = 0;
>  
>  	spin_lock_irq(&dev_priv->irq_lock);
> -	long_port_mask = dev_priv->display.hotplug.long_port_mask;
> -	dev_priv->display.hotplug.long_port_mask = 0;
> -	short_port_mask = dev_priv->display.hotplug.short_port_mask;
> -	dev_priv->display.hotplug.short_port_mask = 0;
> +
> +	suspended_ports = get_suspended_hpd_ports(display);
> +	long_port_mask = hotplug->long_port_mask & ~suspended_ports;
> +	hotplug->long_port_mask &= ~long_port_mask;
> +	short_port_mask = hotplug->short_port_mask & ~suspended_ports;
> +	hotplug->short_port_mask &= ~short_port_mask;
> +
>  	spin_unlock_irq(&dev_priv->irq_lock);
>  
>  	for_each_intel_encoder(&dev_priv->drm, encoder) {
> @@ -406,13 +456,17 @@ static void i915_digport_work_func(struct work_struct *work)
>   */
>  void intel_hpd_trigger_irq(struct intel_digital_port *dig_port)
>  {
> -	struct drm_i915_private *i915 = to_i915(dig_port->base.base.dev);
> +	struct intel_display *display = to_intel_display(dig_port);
> +	struct drm_i915_private *i915 = to_i915(display->drm);
> +	struct intel_hotplug *hotplug = &display->hotplug;
>  
>  	spin_lock_irq(&i915->irq_lock);
> -	i915->display.hotplug.short_port_mask |= BIT(dig_port->base.port);
> -	spin_unlock_irq(&i915->irq_lock);
>  
> -	queue_work(i915->display.hotplug.dp_wq, &i915->display.hotplug.dig_port_work);
> +	hotplug->short_port_mask |= BIT(dig_port->base.port);
> +	if (!(BIT(dig_port->base.port) & get_suspended_hpd_ports(display)))
> +		queue_work(hotplug->dp_wq, &hotplug->dig_port_work);
> +
> +	spin_unlock_irq(&i915->irq_lock);
>  }
>  
>  /*
> @@ -420,9 +474,10 @@ void intel_hpd_trigger_irq(struct intel_digital_port *dig_port)
>   */
>  static void i915_hotplug_work_func(struct work_struct *work)
>  {
> -	struct drm_i915_private *dev_priv =
> -		container_of(work, struct drm_i915_private,
> -			     display.hotplug.hotplug_work.work);
> +	struct intel_display *display =
> +		container_of(work, struct intel_display, hotplug.hotplug_work.work);
> +	struct drm_i915_private *dev_priv = to_i915(display->drm);
> +	struct intel_hotplug *hotplug = &display->hotplug;
>  	struct drm_connector_list_iter conn_iter;
>  	struct intel_connector *connector;
>  	u32 changed = 0, retry = 0;
> @@ -430,16 +485,18 @@ static void i915_hotplug_work_func(struct work_struct *work)
>  	u32 hpd_retry_bits;
>  	struct drm_connector *first_changed_connector = NULL;
>  	int changed_connectors = 0;
> +	u32 suspended_hpds;
>  
>  	mutex_lock(&dev_priv->drm.mode_config.mutex);
>  	drm_dbg_kms(&dev_priv->drm, "running encoder hotplug functions\n");
>  
>  	spin_lock_irq(&dev_priv->irq_lock);
>  
> -	hpd_event_bits = dev_priv->display.hotplug.event_bits;
> -	dev_priv->display.hotplug.event_bits = 0;
> -	hpd_retry_bits = dev_priv->display.hotplug.retry_bits;
> -	dev_priv->display.hotplug.retry_bits = 0;
> +	suspended_hpds = get_suspended_hpds(display);
> +	hpd_event_bits = hotplug->event_bits & ~suspended_hpds;
> +	hotplug->event_bits &= ~hpd_event_bits;
> +	hpd_retry_bits = hotplug->retry_bits & ~suspended_hpds;
> +	hotplug->retry_bits &= ~hpd_retry_bits;
>  
>  	/* Enable polling for connectors which had HPD IRQ storms */
>  	intel_hpd_irq_storm_switch_to_polling(dev_priv);
> @@ -538,11 +595,13 @@ static void i915_hotplug_work_func(struct work_struct *work)
>  void intel_hpd_irq_handler(struct drm_i915_private *dev_priv,
>  			   u32 pin_mask, u32 long_mask)
>  {
> +	struct intel_display *display = to_intel_display(&dev_priv->drm);
>  	struct intel_encoder *encoder;
>  	bool storm_detected = false;
>  	bool queue_dig = false, queue_hp = false;
>  	u32 long_hpd_pulse_mask = 0;
>  	u32 short_hpd_pulse_mask = 0;
> +	u32 suspended_hpds;
>  	enum hpd_pin pin;
>  
>  	if (!pin_mask)
> @@ -550,6 +609,8 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv,
>  
>  	spin_lock(&dev_priv->irq_lock);
>  
> +	suspended_hpds = get_suspended_hpds(display);
> +
>  	/*
>  	 * Determine whether ->hpd_pulse() exists for each pin, and
>  	 * whether we have a short or a long pulse. This is needed
> @@ -573,7 +634,9 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv,
>  			"digital hpd on [ENCODER:%d:%s] - %s\n",
>  			encoder->base.base.id, encoder->base.name,
>  			long_hpd ? "long" : "short");
> -		queue_dig = true;
> +
> +		if (!(BIT(pin) & suspended_hpds))
> +			queue_dig = true;
>  
>  		if (long_hpd) {
>  			long_hpd_pulse_mask |= BIT(pin);
> @@ -617,7 +680,9 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv,
>  		} else {
>  			dev_priv->display.hotplug.event_bits |= BIT(pin);
>  			long_hpd = true;
> -			queue_hp = true;
> +
> +			if (!(BIT(pin) & suspended_hpds))
> +				queue_hp = true;
>  		}
>  
>  		if (intel_hpd_irq_storm_detect(dev_priv, pin, long_hpd)) {
> @@ -915,11 +980,15 @@ static bool cancel_all_detection_work(struct drm_i915_private *i915)
>  
>  void intel_hpd_cancel_work(struct drm_i915_private *dev_priv)
>  {
> +	struct intel_display *display = to_intel_display(&dev_priv->drm);
> +
>  	if (!HAS_DISPLAY(dev_priv))
>  		return;
>  
>  	spin_lock_irq(&dev_priv->irq_lock);
>  
> +	drm_WARN_ON(display->drm, get_suspended_hpds(display));
> +
>  	dev_priv->display.hotplug.long_port_mask = 0;
>  	dev_priv->display.hotplug.short_port_mask = 0;
>  	dev_priv->display.hotplug.event_bits = 0;
> @@ -966,19 +1035,22 @@ void intel_hpd_enable(struct drm_i915_private *dev_priv, enum hpd_pin pin)
>  
>  static void queue_work_for_missed_irqs(struct drm_i915_private *i915)
>  {
> -	bool queue_work = false;
> +	struct intel_display *display = to_intel_display(&i915->drm);
> +	struct intel_hotplug *hotplug = &display->hotplug;
> +	bool queue_hp_work = false;
> +	u32 suspended_hpds;
>  	enum hpd_pin pin;
>  
>  	lockdep_assert_held(&i915->irq_lock);
>  
> -	if (i915->display.hotplug.event_bits ||
> -	    i915->display.hotplug.retry_bits)
> -		queue_work = true;
> +	suspended_hpds = get_suspended_hpds(display);
> +	if ((hotplug->event_bits | hotplug->retry_bits) & ~suspended_hpds)
> +		queue_hp_work = true;
>  
>  	for_each_hpd_pin(pin) {
>  		switch (i915->display.hotplug.stats[pin].state) {
>  		case HPD_MARK_DISABLED:
> -			queue_work = true;
> +			queue_hp_work = true;
>  			break;
>  		case HPD_DISABLED:
>  		case HPD_ENABLED:
> @@ -988,10 +1060,97 @@ static void queue_work_for_missed_irqs(struct drm_i915_private *i915)
>  		}
>  	}
>  
> -	if (queue_work)
> +	if ((hotplug->long_port_mask | hotplug->short_port_mask) &
> +	    ~hpd_mask_to_ports(display, suspended_hpds))
> +		queue_work(hotplug->dp_wq, &hotplug->dig_port_work);
> +
> +	if (queue_hp_work)
>  		queue_delayed_detection_work(i915, &i915->display.hotplug.hotplug_work, 0);
>  }
>  
> +/**
> + * intel_hpd_suspend - Suspend handling of HPD IRQs on an HPD pin
> + * @encoder: Encoder to suspend the HPD handling for
> + *
> + * Suspends the handling of HPD IRQs on the HPD pin of @encoder.
> + *
> + * On return:
> + * - It's guaranteed that the suspended encoders' HPD pulse handler
> + *   (via intel_digital_port::hpd_pulse()) is not running.
> + * - The hotplug event handling (via intel_encoder::hotplug()) of an
> + *   HPD IRQ pending at the time this function is called may be still
> + *   running.
> + * - Detection on the encoder's connector (via
> + *   drm_connector_helper_funcs::detect_ctx(),
> + *   drm_connector_funcs::detect()) remains allowed, for instance as part of
> + *   userspace connector probing, or DRM core's connector polling.
> + *
> + * The call must be followed by calling intel_hpd_resume().
> + *
> + * Note that the handling of HPD IRQs for another encoder using the same HPD
> + * pin as that of @encoder will be also suspended.
> + */
> +void intel_hpd_suspend(struct intel_encoder *encoder)
> +{
> +	struct intel_display *display = to_intel_display(encoder);
> +	struct drm_i915_private *i915 = to_i915(display->drm);
> +	struct intel_hotplug *hotplug = &display->hotplug;
> +	bool do_flush = false;
> +
> +	if (encoder->hpd_pin == HPD_NONE)
> +		return;
> +
> +	spin_lock_irq(&i915->irq_lock);
> +
> +	hotplug->stats[encoder->hpd_pin].suspend_count++;
> +
> +	if (hotplug->stats[encoder->hpd_pin].suspend_count == 1 &&
> +	    intel_encoder_has_hpd_pulse(encoder))
> +		do_flush = true;
> +
> +	spin_unlock_irq(&i915->irq_lock);
> +
> +	if (do_flush)
> +		flush_work(&hotplug->dig_port_work);
> +}
> +
> +static void resume_hpd(struct intel_encoder *encoder)
> +{
> +	struct intel_display *display = to_intel_display(encoder);
> +	struct drm_i915_private *i915 = to_i915(display->drm);
> +	struct intel_hotplug *hotplug = &display->hotplug;
> +
> +	lockdep_assert_held(&i915->irq_lock);
> +
> +	drm_WARN_ON(display->drm, hotplug->stats[encoder->hpd_pin].suspend_count == 0);
> +	hotplug->stats[encoder->hpd_pin].suspend_count--;
> +
> +	if (hotplug->stats[encoder->hpd_pin].suspend_count == 0)
> +		queue_work_for_missed_irqs(i915);
> +}
> +
> +/**
> + * intel_hpd_resume - Resume handling of HPD IRQs on an HPD pin
> + * @encoder: Encoder to resume the HPD handling for
> + *
> + * Resume the handling of HPD IRQs on the HPD pin of @encoder, which was
> + * previously suspended by intel_hpd_suspend(). Any HPD IRQ raised on the
> + * HPD pin while it was suspended will be handled for @encoder and for any
> + * other encoder sharing the same HPD pin.
> + */
> +void intel_hpd_resume(struct intel_encoder *encoder)
> +{
> +	struct intel_display *display = to_intel_display(encoder);
> +	struct drm_i915_private *i915 = to_i915(display->drm);
> +
> +	if (encoder->hpd_pin == HPD_NONE)
> +		return;
> +
> +	spin_lock_irq(&i915->irq_lock);
> +	resume_hpd(encoder);
> +	spin_unlock_irq(&i915->irq_lock);
> +}
> +
>  void intel_hpd_enable_detection_work(struct drm_i915_private *i915)
>  {
>  	spin_lock_irq(&i915->irq_lock);
> diff --git a/drivers/gpu/drm/i915/display/intel_hotplug.h b/drivers/gpu/drm/i915/display/intel_hotplug.h
> index d6986902b0545..5180731def7cc 100644
> --- a/drivers/gpu/drm/i915/display/intel_hotplug.h
> +++ b/drivers/gpu/drm/i915/display/intel_hotplug.h
> @@ -28,6 +28,8 @@ void intel_hpd_cancel_work(struct drm_i915_private *dev_priv);
>  enum hpd_pin intel_hpd_pin_default(enum port port);
>  bool intel_hpd_disable(struct drm_i915_private *dev_priv, enum hpd_pin pin);
>  void intel_hpd_enable(struct drm_i915_private *dev_priv, enum hpd_pin pin);
> +void intel_hpd_suspend(struct intel_encoder *encoder);
> +void intel_hpd_resume(struct intel_encoder *encoder);
>  void intel_hpd_debugfs_register(struct drm_i915_private *i915);
>  
>  void intel_hpd_enable_detection_work(struct drm_i915_private *i915);

-- 
Jani Nikula, Intel

  reply	other threads:[~2025-02-25 16:25 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-02-24 19:31 [PATCH 0/5] drm/dp: Fix link training interrupted by HPD pulse Imre Deak
2025-02-24 19:31 ` [PATCH 1/5] drm/i915/hpd: Let an HPD pin be in the disabled state when handling missed IRQs Imre Deak
2025-02-24 19:31 ` [PATCH 2/5] drm/i915/hpd: Add support for suspending the IRQ handling on an HPD pin Imre Deak
2025-02-25 12:15   ` [PATCH v2 " Imre Deak
2025-02-25 16:25     ` Jani Nikula [this message]
2025-02-25 17:15       ` Imre Deak
2025-02-24 19:31 ` [PATCH 3/5] drm/i915/dp: Fix link training interrupted by a short HPD pulse Imre Deak
2025-02-24 19:31 ` [PATCH 4/5] drm/i915/dp: Queue a link check after link training is complete Imre Deak
2025-02-24 19:31 ` [PATCH 5/5] drm/i915/crt: Use intel_hpd_suspend/resume() instead of intel_hpd_disable/enable() Imre Deak
2025-02-25 12:15   ` [PATCH v2 " Imre Deak
2025-02-25  1:32 ` ✗ Fi.CI.SPARSE: warning for drm/dp: Fix link training interrupted by HPD pulse Patchwork
2025-02-25  1:54 ` ✓ CI.Patch_applied: success " Patchwork
2025-02-25  1:54 ` ✓ CI.checkpatch: " Patchwork
2025-02-25  1:56 ` ✓ CI.KUnit: " Patchwork
2025-02-25  2:00 ` ✗ i915.CI.BAT: failure " Patchwork
2025-02-25  2:12 ` ✓ CI.Build: success " Patchwork
2025-02-25  2:14 ` ✓ CI.Hooks: " Patchwork
2025-02-25  2:16 ` ✗ CI.checksparse: warning " Patchwork
2025-02-25  2:35 ` ✓ Xe.CI.BAT: success " Patchwork
2025-02-25  6:13 ` ✗ Xe.CI.Full: failure " Patchwork
2025-02-25 12:20 ` ✓ CI.Patch_applied: success for drm/dp: Fix link training interrupted by HPD pulse (rev3) Patchwork
2025-02-25 12:20 ` ✓ CI.checkpatch: " Patchwork
2025-02-25 12:21 ` ✓ CI.KUnit: " Patchwork
2025-02-25 12:38 ` ✓ CI.Build: " Patchwork
2025-02-25 12:40 ` ✓ CI.Hooks: " Patchwork
2025-02-25 12:42 ` ✗ CI.checksparse: warning " Patchwork
2025-02-25 12:59 ` ✓ Xe.CI.BAT: success " Patchwork
2025-02-25 13:23 ` ✗ Fi.CI.SPARSE: warning " Patchwork
2025-02-25 13:41 ` ✗ i915.CI.BAT: failure " Patchwork
2025-02-25 14:25   ` Imre Deak
2025-02-26  4:09     ` Ravali, JupallyX
2025-02-25 19:15 ` ✗ Xe.CI.Full: " Patchwork
2025-02-26  3:51 ` ✓ i915.CI.BAT: success " Patchwork
2025-02-26  5:33 ` ✗ i915.CI.Full: failure " Patchwork
2025-02-26 12:55 ` [PATCH 0/5] drm/dp: Fix link training interrupted by HPD pulse Jani Nikula
2025-02-26 13:17   ` Imre Deak

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=87bjuq56w2.fsf@intel.com \
    --to=jani.nikula@linux.intel.com \
    --cc=imre.deak@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=intel-xe@lists.freedesktop.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.