[Intel-gfx] [PATCH] drm/i915: Only apply a rmw mmio update if the value changes

Mika Kuoppala mika.kuoppala at linux.intel.com
Tue Sep 17 13:49:51 UTC 2019


Chris Wilson <chris at chris-wilson.co.uk> writes:

> If we try to clear, or even set, a bit in the register that doesn't
> change the register state; skip the write. There's a slight danger in
> that the register acts as a latch-on-write, but I do not think we use a
> rmw cycle with any such latch registers.
>
> Suggested-by: Daniele Ceraolo Spurio <daniele.ceraolospurio at intel.com>
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio at intel.com>
> Cc: Tvrtko Ursulin <tvrtko.ursulin at intel.com>

My stance was that with clear ~0, user should be notified.
But I guess on this level of tooling, you should be free
to do the useless read.

Reviewed-by: Mika Kuoppala <mika.kuoppala at linux.intel.com>

> ---
>  drivers/gpu/drm/i915/intel_uncore.h | 20 ++++++++++----------
>  1 file changed, 10 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_uncore.h b/drivers/gpu/drm/i915/intel_uncore.h
> index 414fc2cb0459..dcfa243892c6 100644
> --- a/drivers/gpu/drm/i915/intel_uncore.h
> +++ b/drivers/gpu/drm/i915/intel_uncore.h
> @@ -378,23 +378,23 @@ intel_uncore_read64_2x32(struct intel_uncore *uncore,
>  static inline void intel_uncore_rmw(struct intel_uncore *uncore,
>  				    i915_reg_t reg, u32 clear, u32 set)
>  {
> -	u32 val;
> +	u32 old, val;
>  
> -	val = intel_uncore_read(uncore, reg);
> -	val &= ~clear;
> -	val |= set;
> -	intel_uncore_write(uncore, reg, val);
> +	old = intel_uncore_read(uncore, reg);
> +	val = (old & ~clear) | set;
> +	if (val != old)
> +		intel_uncore_write(uncore, reg, val);
>  }
>  
>  static inline void intel_uncore_rmw_fw(struct intel_uncore *uncore,
>  				       i915_reg_t reg, u32 clear, u32 set)
>  {
> -	u32 val;
> +	u32 old, val;
>  
> -	val = intel_uncore_read_fw(uncore, reg);
> -	val &= ~clear;
> -	val |= set;
> -	intel_uncore_write_fw(uncore, reg, val);
> +	old = intel_uncore_read_fw(uncore, reg);
> +	val = (old & ~clear) | set;
> +	if (val != old)
> +		intel_uncore_write_fw(uncore, reg, val);
>  }
>  
>  static inline int intel_uncore_write_and_verify(struct intel_uncore *uncore,
> -- 
> 2.23.0
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx


More information about the Intel-gfx mailing list