[Intel-gfx] [PATCH v2 02/19] drm/i915/dsb: Use non-locked register access

Jani Nikula jani.nikula at linux.intel.com
Tue Sep 12 07:37:56 UTC 2023


On Mon, 11 Sep 2023, "Shankar, Uma" <uma.shankar at intel.com> wrote:
>> -----Original Message-----
>> From: Intel-gfx <intel-gfx-bounces at lists.freedesktop.org> On Behalf Of Jani
>> Nikula
>> Sent: Thursday, June 8, 2023 5:16 PM
>> To: Ville Syrjala <ville.syrjala at linux.intel.com>; intel-gfx at lists.freedesktop.org
>> Subject: Re: [Intel-gfx] [PATCH v2 02/19] drm/i915/dsb: Use non-locked register
>> access
>> 
>> On Tue, 06 Jun 2023, Ville Syrjala <ville.syrjala at linux.intel.com> wrote:
>> > From: Ville Syrjälä <ville.syrjala at linux.intel.com>
>> >
>> > Avoid the locking overhead for DSB registers. We don't need the locks
>> > and intel_dsb_commit() in particular needs to be called from the
>> > vblank evade critical section and thus needs to be fast.
>> 
>> Mmmh, I always find it slightly puzzling to encounter _fw calls in code, wondering
>> what the rationale was, and why we can use the _fw variants.
>> 
>> Should we start adding comments explaining why?
>
> I believe it’s a light weight write without any locks and forcewake.

That part is clear; the why isn't. :)

> Maybe a comment to explain the rationale would be good.
>
> With that added, this is:
> Reviewed-by: Uma Shankar <uma.shankar at intel.com>
>
>> BR,
>> Jani.
>> 
>> 
>> >
>> > Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
>> > ---
>> >  drivers/gpu/drm/i915/display/intel_dsb.c | 18 +++++++++---------
>> >  1 file changed, 9 insertions(+), 9 deletions(-)
>> >
>> > diff --git a/drivers/gpu/drm/i915/display/intel_dsb.c
>> > b/drivers/gpu/drm/i915/display/intel_dsb.c
>> > index bed058d2c3ac..97e593d9f100 100644
>> > --- a/drivers/gpu/drm/i915/display/intel_dsb.c
>> > +++ b/drivers/gpu/drm/i915/display/intel_dsb.c
>> > @@ -96,7 +96,7 @@ static bool assert_dsb_has_room(struct intel_dsb
>> > *dsb)  static bool is_dsb_busy(struct drm_i915_private *i915, enum pipe pipe,
>> >  			enum dsb_id id)
>> >  {
>> > -	return intel_de_read(i915, DSB_CTRL(pipe, id)) & DSB_STATUS_BUSY;
>> > +	return intel_de_read_fw(i915, DSB_CTRL(pipe, id)) & DSB_STATUS_BUSY;
>> >  }
>> >
>> >  static void intel_dsb_emit(struct intel_dsb *dsb, u32 ldw, u32 udw)
>> > @@ -243,13 +243,13 @@ void intel_dsb_commit(struct intel_dsb *dsb, bool
>> wait_for_vblank)
>> >  		return;
>> >  	}
>> >
>> > -	intel_de_write(dev_priv, DSB_CTRL(pipe, dsb->id),
>> > -		       (wait_for_vblank ? DSB_WAIT_FOR_VBLANK : 0) |
>> > -		       DSB_ENABLE);
>> > -	intel_de_write(dev_priv, DSB_HEAD(pipe, dsb->id),
>> > -		       i915_ggtt_offset(dsb->vma));
>> > -	intel_de_write(dev_priv, DSB_TAIL(pipe, dsb->id),
>> > -		       i915_ggtt_offset(dsb->vma) + tail);
>> > +	intel_de_write_fw(dev_priv, DSB_CTRL(pipe, dsb->id),
>> > +			  (wait_for_vblank ? DSB_WAIT_FOR_VBLANK : 0) |
>> > +			  DSB_ENABLE);
>> > +	intel_de_write_fw(dev_priv, DSB_HEAD(pipe, dsb->id),
>> > +			  i915_ggtt_offset(dsb->vma));
>> > +	intel_de_write_fw(dev_priv, DSB_TAIL(pipe, dsb->id),
>> > +			  i915_ggtt_offset(dsb->vma) + tail);
>> >  }
>> >
>> >  void intel_dsb_wait(struct intel_dsb *dsb) @@ -266,7 +266,7 @@ void
>> > intel_dsb_wait(struct intel_dsb *dsb)
>> >  	/* Attempt to reset it */
>> >  	dsb->free_pos = 0;
>> >  	dsb->ins_start_offset = 0;
>> > -	intel_de_write(dev_priv, DSB_CTRL(pipe, dsb->id), 0);
>> > +	intel_de_write_fw(dev_priv, DSB_CTRL(pipe, dsb->id), 0);
>> >  }
>> >
>> >  /**
>> 
>> --
>> Jani Nikula, Intel Open Source Graphics Center

-- 
Jani Nikula, Intel


More information about the Intel-gfx mailing list