intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Jani Nikula <jani.nikula@linux.intel.com>
To: Ville Syrjala <ville.syrjala@linux.intel.com>,
	intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 3/4] drm/i915: Extract i915gm_irq_cstate_wa_{disable, enable}()
Date: Wed, 02 Oct 2024 13:37:11 +0300	[thread overview]
Message-ID: <87o742u6y0.fsf@intel.com> (raw)
In-Reply-To: <20241001195803.3371-4-ville.syrjala@linux.intel.com>

On Tue, 01 Oct 2024, Ville Syrjala <ville.syrjala@linux.intel.com> wrote:
> From: Ville Syrjälä <ville.syrjala@linux.intel.com>
>
> Extract the i915gm/i945gm vblank irq C-state workaround to
> separate functions. We'll need to reuse these in order to
> guarantee timely CRC interrupt delivery as well.
>
> The irq.vblank_enabled count is currently protected by the
> drm vblanl locks, so let's assert that the innermost of those

*vblank

Reviewed-by: Jani Nikula <jani.nikula@intel.com>


> is held, in anticipation of other callers.
>
> Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> ---
>  .../gpu/drm/i915/display/intel_display_irq.c  | 34 +++++++++++++------
>  1 file changed, 24 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display_irq.c b/drivers/gpu/drm/i915/display/intel_display_irq.c
> index 43a0b3565bc8..feeb3a29972a 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_irq.c
> +++ b/drivers/gpu/drm/i915/display/intel_display_irq.c
> @@ -1259,6 +1259,28 @@ void gen11_display_irq_handler(struct drm_i915_private *i915)
>  	enable_rpm_wakeref_asserts(&i915->runtime_pm);
>  }
>  
> +static void i915gm_irq_cstate_wa_enable(struct drm_i915_private *i915)
> +{
> +	lockdep_assert_held(&i915->drm.vblank_time_lock);
> +
> +	/*
> +	 * Vblank interrupts fail to wake the device up from C2+.
> +	 * Disabling render clock gating during C-states avoids
> +	 * the problem. There is a small power cost so we do this
> +	 * only when vblank interrupts are actually enabled.
> +	 */
> +	if (i915->display.irq.vblank_enabled++ == 0)
> +		intel_uncore_write(&i915->uncore, SCPD0, _MASKED_BIT_ENABLE(CSTATE_RENDER_CLOCK_GATE_DISABLE));
> +}
> +
> +static void i915gm_irq_cstate_wa_disable(struct drm_i915_private *i915)
> +{
> +	lockdep_assert_held(&i915->drm.vblank_time_lock);
> +
> +	if (--i915->display.irq.vblank_enabled == 0)
> +		intel_uncore_write(&i915->uncore, SCPD0, _MASKED_BIT_DISABLE(CSTATE_RENDER_CLOCK_GATE_DISABLE));
> +}
> +
>  int i8xx_enable_vblank(struct drm_crtc *crtc)
>  {
>  	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
> @@ -1287,14 +1309,7 @@ int i915gm_enable_vblank(struct drm_crtc *crtc)
>  {
>  	struct drm_i915_private *i915 = to_i915(crtc->dev);
>  
> -	/*
> -	 * Vblank interrupts fail to wake the device up from C2+.
> -	 * Disabling render clock gating during C-states avoids
> -	 * the problem. There is a small power cost so we do this
> -	 * only when vblank interrupts are actually enabled.
> -	 */
> -	if (i915->display.irq.vblank_enabled++ == 0)
> -		intel_uncore_write(&i915->uncore, SCPD0, _MASKED_BIT_ENABLE(CSTATE_RENDER_CLOCK_GATE_DISABLE));
> +	i915gm_irq_cstate_wa_enable(i915);
>  
>  	return i8xx_enable_vblank(crtc);
>  }
> @@ -1305,8 +1320,7 @@ void i915gm_disable_vblank(struct drm_crtc *crtc)
>  
>  	i8xx_disable_vblank(crtc);
>  
> -	if (--i915->display.irq.vblank_enabled == 0)
> -		intel_uncore_write(&i915->uncore, SCPD0, _MASKED_BIT_DISABLE(CSTATE_RENDER_CLOCK_GATE_DISABLE));
> +	i915gm_irq_cstate_wa_disable(i915);
>  }
>  
>  int i965_enable_vblank(struct drm_crtc *crtc)

-- 
Jani Nikula, Intel

  reply	other threads:[~2024-10-02 10:37 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-01 19:57 [PATCH 0/4] drm/i915: Vblank/CRC irq stuf Ville Syrjala
2024-10-01 19:58 ` [PATCH 1/4] drm/i915/irq: Nuke stale comments Ville Syrjala
2024-10-02 10:34   ` Jani Nikula
2024-10-01 19:58 ` [PATCH 2/4] drm/i915/irq: Pair up the vblank enable/disable functions Ville Syrjala
2024-10-02 10:36   ` Jani Nikula
2024-10-01 19:58 ` [PATCH 3/4] drm/i915: Extract i915gm_irq_cstate_wa_{disable, enable}() Ville Syrjala
2024-10-02 10:37   ` Jani Nikula [this message]
2024-10-01 19:58 ` [PATCH 4/4] drm/i915: Apply the i915gm/i945gm irq C-state w/a to CRC interrupts Ville Syrjala
2024-10-02 10:46   ` Jani Nikula
2024-10-02  2:08 ` ✗ Fi.CI.CHECKPATCH: warning for drm/i915: Vblank/CRC irq stuf Patchwork
2024-10-02  2:59 ` ✓ Fi.CI.BAT: success " Patchwork
2024-10-03  1:29 ` ✗ Fi.CI.IGT: failure " Patchwork

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=87o742u6y0.fsf@intel.com \
    --to=jani.nikula@linux.intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=ville.syrjala@linux.intel.com \
    /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;
as well as URLs for NNTP newsgroup(s).