Re: [PATCH 1/2] drm/i915: Initialize bdw workarounds in logical ring mode too

[Jani Nikula <jani.nikula@linux.intel.com>]

IMHO it would be perilous to apply these patches before we have root
caused https://bugs.freedesktop.org/show_bug.cgi?id=83482. I think we
need to be able to revert those changes if we can't fix the issue soon.

BR,
Jani.


On Tue, 16 Sep 2014, Michel Thierry <michel.thierry@intel.com> wrote:
> Following the legacy ring submission example, update the
> ring->init_context() hook to support the execlist submission mode.
>
> Workarounds are defined in bdw_emit_workarounds(), but the emit
> now depends on the ring submission mode.
>
> For: VIZ-4092
> Signed-off-by: Michel Thierry <michel.thierry@intel.com>
> ---
>  drivers/gpu/drm/i915/i915_gem_context.c |  2 +-
>  drivers/gpu/drm/i915/intel_lrc.c        | 66 +++++++++++++++++++++++++++
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 79 +++++++++++++++++++--------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h |  6 ++-
>  4 files changed, 117 insertions(+), 36 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
> index 7b73b36..d1ed21a 100644
> --- a/drivers/gpu/drm/i915/i915_gem_context.c
> +++ b/drivers/gpu/drm/i915/i915_gem_context.c
> @@ -657,7 +657,7 @@ done:
>  
>  	if (uninitialized) {
>  		if (ring->init_context) {
> -			ret = ring->init_context(ring);
> +			ret = ring->init_context(ring->buffer);
>  			if (ret)
>  				DRM_ERROR("ring init context: %d\n", ret);
>  		}
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index d64d518..2e1d710 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -1020,6 +1020,62 @@ int intel_logical_ring_begin(struct intel_ringbuffer *ringbuf, int num_dwords)
>  	return 0;
>  }
>  
> +static inline void intel_logical_ring_emit_wa(struct intel_ringbuffer *ringbuf,
> +				       u32 addr, u32 value)
> +{
> +	struct intel_engine_cs *ring = ringbuf->ring;
> +	struct drm_device *dev = ring->dev;
> +	struct drm_i915_private *dev_priv = dev->dev_private;
> +
> +	if (WARN_ON(dev_priv->num_wa_regs >= I915_MAX_WA_REGS))
> +		return;
> +
> +	intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(1));
> +	intel_logical_ring_emit(ringbuf, addr);
> +	intel_logical_ring_emit(ringbuf, value);
> +
> +	dev_priv->intel_wa_regs[dev_priv->num_wa_regs].addr = addr;
> +	dev_priv->intel_wa_regs[dev_priv->num_wa_regs].mask = value & 0xFFFF;
> +	/* value is updated with the status of remaining bits of this
> +	 * register when it is read from debugfs file
> +	 */
> +	dev_priv->intel_wa_regs[dev_priv->num_wa_regs].value = value;
> +	dev_priv->num_wa_regs++;
> +}
> +
> +static int bdw_init_logical_workarounds(struct intel_ringbuffer *ringbuf)
> +{
> +	int ret;
> +	struct intel_engine_cs *ring = ringbuf->ring;
> +	struct drm_device *dev = ring->dev;
> +	struct drm_i915_private *dev_priv = dev->dev_private;
> +
> +	/*
> +	 * workarounds applied in this fn are part of register state context,
> +	 * they need to be re-initialized followed by gpu reset, suspend/resume,
> +	 * module reload.
> +	 */
> +	dev_priv->num_wa_regs = 0;
> +	memset(dev_priv->intel_wa_regs, 0, sizeof(dev_priv->intel_wa_regs));
> +
> +	/*
> +	 * update the number of dwords required based on the
> +	 * actual number of workarounds applied
> +	 */
> +	ret = intel_logical_ring_begin(ringbuf, 24);
> +	if (ret)
> +		return ret;
> +
> +	bdw_emit_workarounds(ringbuf);
> +
> +	intel_logical_ring_advance(ringbuf);
> +
> +	DRM_DEBUG_DRIVER("Number of Workarounds applied: %d\n",
> +			 dev_priv->num_wa_regs);
> +
> +	return 0;
> +}
> +
>  static int gen8_init_common_ring(struct intel_engine_cs *ring)
>  {
>  	struct drm_device *dev = ring->dev;
> @@ -1315,6 +1371,10 @@ static int logical_render_ring_init(struct drm_device *dev)
>  	if (HAS_L3_DPF(dev))
>  		ring->irq_keep_mask |= GT_RENDER_L3_PARITY_ERROR_INTERRUPT;
>  
> +	if (IS_BROADWELL(dev))
> +		ring->init_context = bdw_init_logical_workarounds;
> +	ring->emit_wa = intel_logical_ring_emit_wa;
> +
>  	ring->init = gen8_init_render_ring;
>  	ring->cleanup = intel_fini_pipe_control;
>  	ring->get_seqno = gen8_get_seqno;
> @@ -1802,6 +1862,12 @@ int intel_lr_context_deferred_create(struct intel_context *ctx,
>  	}
>  
>  	if (ring->id == RCS && !ctx->rcs_initialized) {
> +		if (ring->init_context) {
> +			ret = ring->init_context(ringbuf);
> +			if (ret)
> +				DRM_ERROR("ring init context: %d\n", ret);
> +		}
> +
>  		ret = intel_lr_context_render_state_init(ring, ctx);
>  		if (ret) {
>  			DRM_ERROR("Init render state failed: %d\n", ret);
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 08b8705..bce25b5 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -677,9 +677,10 @@ err:
>  	return ret;
>  }
>  
> -static inline void intel_ring_emit_wa(struct intel_engine_cs *ring,
> +static inline void intel_ring_emit_wa(struct intel_ringbuffer *ringbuf,
>  				       u32 addr, u32 value)
>  {
> +	struct intel_engine_cs *ring = ringbuf->ring;
>  	struct drm_device *dev = ring->dev;
>  	struct drm_i915_private *dev_priv = dev->dev_private;
>  
> @@ -701,62 +702,44 @@ static inline void intel_ring_emit_wa(struct intel_engine_cs *ring,
>  	return;
>  }
>  
> -static int bdw_init_workarounds(struct intel_engine_cs *ring)
> +void bdw_emit_workarounds(struct intel_ringbuffer *ringbuf)
>  {
> -	int ret;
> -	struct drm_device *dev = ring->dev;
> -	struct drm_i915_private *dev_priv = dev->dev_private;
> -
> -	/*
> -	 * workarounds applied in this fn are part of register state context,
> -	 * they need to be re-initialized followed by gpu reset, suspend/resume,
> -	 * module reload.
> -	 */
> -	dev_priv->num_wa_regs = 0;
> -	memset(dev_priv->intel_wa_regs, 0, sizeof(dev_priv->intel_wa_regs));
> -
> -	/*
> -	 * update the number of dwords required based on the
> -	 * actual number of workarounds applied
> -	 */
> -	ret = intel_ring_begin(ring, 24);
> -	if (ret)
> -		return ret;
> +	struct intel_engine_cs *ring = ringbuf->ring;
>  
>  	/* WaDisablePartialInstShootdown:bdw */
>  	/* WaDisableThreadStallDopClockGating:bdw */
>  	/* FIXME: Unclear whether we really need this on production bdw. */
> -	intel_ring_emit_wa(ring, GEN8_ROW_CHICKEN,
> +	ring->emit_wa(ringbuf, GEN8_ROW_CHICKEN,
>  			   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE
>  					     | STALL_DOP_GATING_DISABLE));
>  
>  	/* WaDisableDopClockGating:bdw May not be needed for production */
> -	intel_ring_emit_wa(ring, GEN7_ROW_CHICKEN2,
> +	ring->emit_wa(ringbuf, GEN7_ROW_CHICKEN2,
>  			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
>  
>  	/*
>  	 * This GEN8_CENTROID_PIXEL_OPT_DIS W/A is only needed for
>  	 * pre-production hardware
>  	 */
> -	intel_ring_emit_wa(ring, HALF_SLICE_CHICKEN3,
> +	ring->emit_wa(ringbuf, HALF_SLICE_CHICKEN3,
>  			   _MASKED_BIT_ENABLE(GEN8_CENTROID_PIXEL_OPT_DIS
>  					      | GEN8_SAMPLER_POWER_BYPASS_DIS));
>  
> -	intel_ring_emit_wa(ring, GEN7_HALF_SLICE_CHICKEN1,
> +	ring->emit_wa(ringbuf, GEN7_HALF_SLICE_CHICKEN1,
>  			   _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE));
>  
> -	intel_ring_emit_wa(ring, COMMON_SLICE_CHICKEN2,
> +	ring->emit_wa(ringbuf, COMMON_SLICE_CHICKEN2,
>  			   _MASKED_BIT_ENABLE(GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE));
>  
>  	/* Use Force Non-Coherent whenever executing a 3D context. This is a
>  	 * workaround for for a possible hang in the unlikely event a TLB
>  	 * invalidation occurs during a PSD flush.
>  	 */
> -	intel_ring_emit_wa(ring, HDC_CHICKEN0,
> +	ring->emit_wa(ringbuf, HDC_CHICKEN0,
>  			   _MASKED_BIT_ENABLE(HDC_FORCE_NON_COHERENT));
>  
>  	/* Wa4x4STCOptimizationDisable:bdw */
> -	intel_ring_emit_wa(ring, CACHE_MODE_1,
> +	ring->emit_wa(ringbuf, CACHE_MODE_1,
>  			   _MASKED_BIT_ENABLE(GEN8_4x4_STC_OPTIMIZATION_DISABLE));
>  
>  	/*
> @@ -767,8 +750,34 @@ static int bdw_init_workarounds(struct intel_engine_cs *ring)
>  	 * disable bit, which we don't touch here, but it's good
>  	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
>  	 */
> -	intel_ring_emit_wa(ring, GEN7_GT_MODE,
> +	ring->emit_wa(ringbuf, GEN7_GT_MODE,
>  			   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
> +}
> +
> +static int bdw_init_workarounds(struct intel_ringbuffer *ringbuf)
> +{
> +	int ret;
> +	struct intel_engine_cs *ring = ringbuf->ring;
> +	struct drm_device *dev = ring->dev;
> +	struct drm_i915_private *dev_priv = dev->dev_private;
> +
> +	/*
> +	 * workarounds applied in this fn are part of register state context,
> +	 * they need to be re-initialized followed by gpu reset, suspend/resume,
> +	 * module reload.
> +	 */
> +	dev_priv->num_wa_regs = 0;
> +	memset(dev_priv->intel_wa_regs, 0, sizeof(dev_priv->intel_wa_regs));
> +
> +	/*
> +	 * update the number of dwords required based on the
> +	 * actual number of workarounds applied
> +	 */
> +	ret = intel_ring_begin(ring, 24);
> +	if (ret)
> +		return ret;
> +
> +	bdw_emit_workarounds(ringbuf);
>  
>  	intel_ring_advance(ring);
>  
> @@ -778,9 +787,10 @@ static int bdw_init_workarounds(struct intel_engine_cs *ring)
>  	return 0;
>  }
>  
> -static int chv_init_workarounds(struct intel_engine_cs *ring)
> +static int chv_init_workarounds(struct intel_ringbuffer *ringbuf)
>  {
>  	int ret;
> +	struct intel_engine_cs *ring = ringbuf->ring;
>  	struct drm_device *dev = ring->dev;
>  	struct drm_i915_private *dev_priv = dev->dev_private;
>  
> @@ -797,19 +807,19 @@ static int chv_init_workarounds(struct intel_engine_cs *ring)
>  		return ret;
>  
>  	/* WaDisablePartialInstShootdown:chv */
> -	intel_ring_emit_wa(ring, GEN8_ROW_CHICKEN,
> +	intel_ring_emit_wa(ringbuf, GEN8_ROW_CHICKEN,
>  			   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
>  
>  	/* WaDisableThreadStallDopClockGating:chv */
> -	intel_ring_emit_wa(ring, GEN8_ROW_CHICKEN,
> +	intel_ring_emit_wa(ringbuf, GEN8_ROW_CHICKEN,
>  			   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
>  
>  	/* WaDisableDopClockGating:chv (pre-production hw) */
> -	intel_ring_emit_wa(ring, GEN7_ROW_CHICKEN2,
> +	intel_ring_emit_wa(ringbuf, GEN7_ROW_CHICKEN2,
>  			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
>  
>  	/* WaDisableSamplerPowerBypass:chv (pre-production hw) */
> -	intel_ring_emit_wa(ring, HALF_SLICE_CHICKEN3,
> +	intel_ring_emit_wa(ringbuf, HALF_SLICE_CHICKEN3,
>  			   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
>  
>  	intel_ring_advance(ring);
> @@ -2321,6 +2331,7 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
>  			ring->init_context = chv_init_workarounds;
>  		else
>  			ring->init_context = bdw_init_workarounds;
> +		ring->emit_wa = intel_ring_emit_wa;
>  		ring->add_request = gen6_add_request;
>  		ring->flush = gen8_render_ring_flush;
>  		ring->irq_get = gen8_ring_get_irq;
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 07f66d4..190c095 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -148,7 +148,10 @@ struct  intel_engine_cs {
>  
>  	int		(*init)(struct intel_engine_cs *ring);
>  
> -	int		(*init_context)(struct intel_engine_cs *ring);
> +	int		(*init_context)(struct intel_ringbuffer *ringbuf);
> +
> +	void	(*emit_wa)(struct intel_ringbuffer *ringbuf,
> +		       u32 addr, u32 value);
>  
>  	void		(*write_tail)(struct intel_engine_cs *ring,
>  				      u32 value);
> @@ -427,6 +430,7 @@ int intel_init_vebox_ring_buffer(struct drm_device *dev);
>  
>  u64 intel_ring_get_active_head(struct intel_engine_cs *ring);
>  void intel_ring_setup_status_page(struct intel_engine_cs *ring);
> +void bdw_emit_workarounds(struct intel_ringbuffer *ringbuf);
>  
>  static inline u32 intel_ring_get_tail(struct intel_ringbuffer *ringbuf)
>  {
> -- 
> 2.0.3
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Jani Nikula, Intel Open Source Technology Center