drm/i915/gen8: Add WaFlushCoherentL3CacheLinesAtContextSwitch workaround

In Indirect context w/a batch buffer,
+WaFlushCoherentL3CacheLinesAtContextSwitch:bdw

v2: Add LRI commands to set/reset bit that invalidates coherent lines,
update WA to include programming restrictions and exclude CHV as
it is not required (Ville)

v3: Avoid unnecessary read when it can be done by reading register once (Chris).

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Rafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

drivers/gpu/drm/i915/i915_reg.h

@@ -431,6 +431,7 @@
 #define   PIPE_CONTROL_INDIRECT_STATE_DISABLE		(1<<9)
 #define   PIPE_CONTROL_NOTIFY				(1<<8)
 #define   PIPE_CONTROL_FLUSH_ENABLE			(1<<7) /* gen7+ */
+#define   PIPE_CONTROL_DC_FLUSH_ENABLE			(1<<5)
 #define   PIPE_CONTROL_VF_CACHE_INVALIDATE		(1<<4)
 #define   PIPE_CONTROL_CONST_CACHE_INVALIDATE		(1<<3)
 #define   PIPE_CONTROL_STATE_CACHE_INVALIDATE		(1<<2)
@@ -5811,6 +5812,7 @@ enum skl_disp_power_wells {
 
 #define GEN8_L3SQCREG4				0xb118
 #define  GEN8_LQSC_RO_PERF_DIS			(1<<27)
+#define  GEN8_LQSC_FLUSH_COHERENT_LINES		(1<<21)
 
 /* GEN8 chicken */
 #define HDC_CHICKEN0				0x7300

drivers/gpu/drm/i915/intel_lrc.c

@@ -1143,6 +1143,29 @@ static int gen8_init_indirectctx_bb(struct intel_engine_cs *ring,
 	/* WaDisableCtxRestoreArbitration:bdw,chv */
 	wa_ctx_emit(batch, MI_ARB_ON_OFF | MI_ARB_DISABLE);
 
+	/* WaFlushCoherentL3CacheLinesAtContextSwitch:bdw */
+	if (IS_BROADWELL(ring->dev)) {
+		struct drm_i915_private *dev_priv = to_i915(ring->dev);
+		uint32_t l3sqc4_flush = (I915_READ(GEN8_L3SQCREG4) |
+					 GEN8_LQSC_FLUSH_COHERENT_LINES);
+
+		wa_ctx_emit(batch, MI_LOAD_REGISTER_IMM(1));
+		wa_ctx_emit(batch, GEN8_L3SQCREG4);
+		wa_ctx_emit(batch, l3sqc4_flush);
+
+		wa_ctx_emit(batch, GFX_OP_PIPE_CONTROL(6));
+		wa_ctx_emit(batch, (PIPE_CONTROL_CS_STALL |
+				    PIPE_CONTROL_DC_FLUSH_ENABLE));
+		wa_ctx_emit(batch, 0);
+		wa_ctx_emit(batch, 0);
+		wa_ctx_emit(batch, 0);
+		wa_ctx_emit(batch, 0);
+
+		wa_ctx_emit(batch, MI_LOAD_REGISTER_IMM(1));
+		wa_ctx_emit(batch, GEN8_L3SQCREG4);
+		wa_ctx_emit(batch, l3sqc4_flush & ~GEN8_LQSC_FLUSH_COHERENT_LINES);
+	}
+
 	/* Pad to end of cacheline */
 	while (index % CACHELINE_DWORDS)
 		wa_ctx_emit(batch, MI_NOOP);