drm/i915: Introduce per-engine workarounds

We stopped re-applying the GT workarounds after engine reset since commit
59b449d5 ("drm/i915: Split out functions for different kinds of
workarounds").

Issue with this is that some of the GT workarounds live in the MMIO space
which gets lost during engine resets. So far the registers in 0x2xxx and
0xbxxx address range have been identified to be affected.

This losing of applied workarounds has obvious negative effects and can
even lead to hard system hangs (see the linked Bugzilla).

Rather than just restoring this re-application, because we have also
observed that it is not safe to just re-write all GT workarounds after
engine resets (GPU might be live and weird hardware states can happen),
we introduce a new class of per-engine workarounds and move only the
affected GT workarounds over.

Using the framework introduced in the previous patch, we therefore after
engine reset, re-apply only the workarounds living in the affected MMIO
address ranges.

v2:
 * Move Wa_1406609255:icl to engine workarounds as well.
 * Rename API. (Chris Wilson)
 * Drop redundant IS_KABYLAKE. (Chris Wilson)
 * Re-order engine wa/ init so latest platforms are first. (Rodrigo Vivi)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Bugzilla: https://bugzilla.freedesktop.org/show_bug.cgi?id=107945
Fixes: 59b449d5 ("drm/i915: Split out functions for different kinds of workarounds")
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: intel-gfx@lists.freedesktop.org
Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20181203133341.10258-1-tvrtko.ursulin@linux.intel.com

drivers/gpu/drm/i915/intel_engine_cs.c

@@ -723,6 +723,8 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 	__intel_context_unpin(i915->kernel_context, engine);
 
 	i915_timeline_fini(&engine->timeline);
+
+	intel_wa_list_free(&engine->wa_list);
 }
 
 u64 intel_engine_get_active_head(const struct intel_engine_cs *engine)

drivers/gpu/drm/i915/intel_lrc.c

@@ -1626,6 +1626,8 @@ static bool unexpected_starting_state(struct intel_engine_cs *engine)
 
 static int gen8_init_common_ring(struct intel_engine_cs *engine)
 {
+	intel_engine_apply_workarounds(engine);
+
 	intel_mocs_init_engine(engine);
 
 	intel_engine_reset_breadcrumbs(engine);
@@ -2305,6 +2307,8 @@ int logical_render_ring_init(struct intel_engine_cs *engine)
 			  ret);
 	}
 
+	intel_engine_init_workarounds(engine);
+
 	return 0;
 
 err_cleanup_common:

drivers/gpu/drm/i915/intel_ringbuffer.h

@@ -15,6 +15,7 @@
 #include "i915_selftest.h"
 #include "i915_timeline.h"
 #include "intel_gpu_commands.h"
+#include "intel_workarounds.h"
 
 struct drm_printer;
 struct i915_sched_attr;
@@ -435,6 +436,7 @@ struct intel_engine_cs {
 
 	struct intel_hw_status_page status_page;
 	struct i915_ctx_workarounds wa_ctx;
+	struct i915_wa_list wa_list;
 	struct i915_vma *scratch;
 
 	u32             irq_keep_mask; /* always keep these interrupts */

drivers/gpu/drm/i915/intel_workarounds.c

@@ -656,17 +656,6 @@ static void gen9_gt_workarounds_init(struct drm_i915_private *i915)
 {
 	struct i915_wa_list *wal = &i915->gt_wa_list;
 
-	/* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */
-	wa_masked_en(wal,
-		     GEN9_CSFE_CHICKEN1_RCS,
-		     GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE);
-
-
-	/* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */
-	wa_write_or(wal,
-		    BDW_SCRATCH1,
-		    GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE);
-
 	/* WaDisableKillLogic:bxt,skl,kbl */
 	if (!IS_COFFEELAKE(i915))
 		wa_write_or(wal,
@@ -688,24 +677,6 @@ static void gen9_gt_workarounds_init(struct drm_i915_private *i915)
 	wa_write_or(wal,
 		    GAM_ECOCHK,
 		    BDW_DISABLE_HDC_INVALIDATION);
-
-	/* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */
-	if (IS_GEN9_LP(i915))
-		wa_write_masked_or(wal,
-				   GEN8_L3SQCREG1,
-				   L3_PRIO_CREDITS_MASK,
-				   L3_GENERAL_PRIO_CREDITS(62) |
-				   L3_HIGH_PRIO_CREDITS(2));
-
-	/* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */
-	wa_write_or(wal,
-		    GEN8_L3SQCREG4,
-		    GEN8_LQSC_FLUSH_COHERENT_LINES);
-
-	/* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,[cnl] */
-	wa_masked_en(wal,
-		     GEN7_FF_SLICE_CS_CHICKEN1,
-		     GEN9_FFSC_PERCTX_PREEMPT_CTRL);
 }
 
 static void skl_gt_workarounds_init(struct drm_i915_private *i915)
@@ -714,11 +685,6 @@ static void skl_gt_workarounds_init(struct drm_i915_private *i915)
 
 	gen9_gt_workarounds_init(i915);
 
-	/* WaEnableGapsTsvCreditFix:skl */
-	wa_write_or(wal,
-		    GEN8_GARBCNTL,
-		    GEN9_GAPS_TSV_CREDIT_DISABLE);
-
 	/* WaDisableGafsUnitClkGating:skl */
 	wa_write_or(wal,
 		    GEN7_UCGCTL4,
@@ -737,11 +703,6 @@ static void bxt_gt_workarounds_init(struct drm_i915_private *i915)
 
 	gen9_gt_workarounds_init(i915);
 
-	/* WaDisablePooledEuLoadBalancingFix:bxt */
-	wa_masked_en(wal,
-		     FF_SLICE_CS_CHICKEN2,
-		     GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE);
-
 	/* WaInPlaceDecompressionHang:bxt */
 	wa_write_or(wal,
 		    GEN9_GAMT_ECO_REG_RW_IA,
@@ -754,11 +715,6 @@ static void kbl_gt_workarounds_init(struct drm_i915_private *i915)
 
 	gen9_gt_workarounds_init(i915);
 
-	/* WaEnableGapsTsvCreditFix:kbl */
-	wa_write_or(wal,
-		    GEN8_GARBCNTL,
-		    GEN9_GAPS_TSV_CREDIT_DISABLE);
-
 	/* WaDisableDynamicCreditSharing:kbl */
 	if (IS_KBL_REVID(i915, 0, KBL_REVID_B0))
 		wa_write_or(wal,
@@ -774,21 +730,6 @@ static void kbl_gt_workarounds_init(struct drm_i915_private *i915)
 	wa_write_or(wal,
 		    GEN9_GAMT_ECO_REG_RW_IA,
 		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
-
-	/* WaKBLVECSSemaphoreWaitPoll:kbl */
-	if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) {
-		struct intel_engine_cs *engine;
-		unsigned int tmp;
-
-		for_each_engine(engine, i915, tmp) {
-			if (engine->id == RCS)
-				continue;
-
-			wa_write(wal,
-				 RING_SEMA_WAIT_POLL(engine->mmio_base),
-				 1);
-		}
-	}
 }
 
 static void glk_gt_workarounds_init(struct drm_i915_private *i915)
@@ -802,11 +743,6 @@ static void cfl_gt_workarounds_init(struct drm_i915_private *i915)
 
 	gen9_gt_workarounds_init(i915);
 
-	/* WaEnableGapsTsvCreditFix:cfl */
-	wa_write_or(wal,
-		    GEN8_GARBCNTL,
-		    GEN9_GAPS_TSV_CREDIT_DISABLE);
-
 	/* WaDisableGafsUnitClkGating:cfl */
 	wa_write_or(wal,
 		    GEN7_UCGCTL4,
@@ -897,11 +833,6 @@ static void cnl_gt_workarounds_init(struct drm_i915_private *i915)
 	wa_write_or(wal,
 		    GEN9_GAMT_ECO_REG_RW_IA,
 		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
-
-	/* WaEnablePreemptionGranularityControlByUMD:cnl */
-	wa_masked_en(wal,
-		     GEN7_FF_SLICE_CS_CHICKEN1,
-		     GEN9_FFSC_PERCTX_PREEMPT_CTRL);
 }
 
 static void icl_gt_workarounds_init(struct drm_i915_private *i915)
@@ -910,53 +841,17 @@ static void icl_gt_workarounds_init(struct drm_i915_private *i915)
 
 	wa_init_mcr(i915);
 
-	/* This is not an Wa. Enable for better image quality */
-	wa_masked_en(wal,
-		     _3D_CHICKEN3,
-		     _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE);
-
 	/* WaInPlaceDecompressionHang:icl */
 	wa_write_or(wal,
 		    GEN9_GAMT_ECO_REG_RW_IA,
 		    GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS);
 
-	/* WaPipelineFlushCoherentLines:icl */
-	wa_write_or(wal,
-		    GEN8_L3SQCREG4,
-		    GEN8_LQSC_FLUSH_COHERENT_LINES);
-
-	/* Wa_1405543622:icl
-	 * Formerly known as WaGAPZPriorityScheme
-	 */
-	wa_write_or(wal,
-		    GEN8_GARBCNTL,
-		    GEN11_ARBITRATION_PRIO_ORDER_MASK);
-
-	/* Wa_1604223664:icl
-	 * Formerly known as WaL3BankAddressHashing
-	 */
-	wa_write_masked_or(wal,
-			   GEN8_GARBCNTL,
-			   GEN11_HASH_CTRL_EXCL_MASK,
-			   GEN11_HASH_CTRL_EXCL_BIT0);
-	wa_write_masked_or(wal,
-			   GEN11_GLBLINVL,
-			   GEN11_BANK_HASH_ADDR_EXCL_MASK,
-			   GEN11_BANK_HASH_ADDR_EXCL_BIT0);
-
 	/* WaModifyGamTlbPartitioning:icl */
 	wa_write_masked_or(wal,
 			   GEN11_GACB_PERF_CTRL,
 			   GEN11_HASH_CTRL_MASK,
 			   GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4);
 
-	/* Wa_1405733216:icl
-	 * Formerly known as WaDisableCleanEvicts
-	 */
-	wa_write_or(wal,
-		    GEN8_L3SQCREG4,
-		    GEN11_LQSC_CLEAN_EVICT_DISABLE);
-
 	/* Wa_1405766107:icl
 	 * Formerly known as WaCL2SFHalfMaxAlloc
 	 */
@@ -989,26 +884,12 @@ static void icl_gt_workarounds_init(struct drm_i915_private *i915)
 			    INF_UNIT_LEVEL_CLKGATE,
 			    CGPSF_CLKGATE_DIS);
 
-	/* WaForwardProgressSoftReset:icl */
-	wa_write_or(wal,
-		    GEN10_SCRATCH_LNCF2,
-		    PMFLUSHDONE_LNICRSDROP |
-		    PMFLUSH_GAPL3UNBLOCK |
-		    PMFLUSHDONE_LNEBLK);
-
 	/* Wa_1406463099:icl
 	 * Formerly known as WaGamTlbPendError
 	 */
 	wa_write_or(wal,
 		    GAMT_CHKN_BIT_REG,
 		    GAMT_CHKN_DISABLE_L3_COH_PIPE);
-
-	/* Wa_1406609255:icl (pre-prod) */
-	if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0))
-		wa_write_or(wal,
-			    GEN7_SARCHKMD,
-			    GEN7_DISABLE_DEMAND_PREFETCH |
-			    GEN7_DISABLE_SAMPLER_PREFETCH);
 }
 
 void intel_gt_init_workarounds(struct drm_i915_private *i915)
@@ -1245,6 +1126,148 @@ void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine)
 	whitelist_apply(engine, whitelist_build(engine, &w));
 }
 
+static void rcs_engine_wa_init(struct intel_engine_cs *engine)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	struct i915_wa_list *wal = &engine->wa_list;
+
+	if (IS_ICELAKE(i915)) {
+		/* This is not an Wa. Enable for better image quality */
+		wa_masked_en(wal,
+			     _3D_CHICKEN3,
+			     _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE);
+
+		/* WaPipelineFlushCoherentLines:icl */
+		wa_write_or(wal,
+			    GEN8_L3SQCREG4,
+			    GEN8_LQSC_FLUSH_COHERENT_LINES);
+
+		/*
+		 * Wa_1405543622:icl
+		 * Formerly known as WaGAPZPriorityScheme
+		 */
+		wa_write_or(wal,
+			    GEN8_GARBCNTL,
+			    GEN11_ARBITRATION_PRIO_ORDER_MASK);
+
+		/*
+		 * Wa_1604223664:icl
+		 * Formerly known as WaL3BankAddressHashing
+		 */
+		wa_write_masked_or(wal,
+				   GEN8_GARBCNTL,
+				   GEN11_HASH_CTRL_EXCL_MASK,
+				   GEN11_HASH_CTRL_EXCL_BIT0);
+		wa_write_masked_or(wal,
+				   GEN11_GLBLINVL,
+				   GEN11_BANK_HASH_ADDR_EXCL_MASK,
+				   GEN11_BANK_HASH_ADDR_EXCL_BIT0);
+
+		/*
+		 * Wa_1405733216:icl
+		 * Formerly known as WaDisableCleanEvicts
+		 */
+		wa_write_or(wal,
+			    GEN8_L3SQCREG4,
+			    GEN11_LQSC_CLEAN_EVICT_DISABLE);
+
+		/* WaForwardProgressSoftReset:icl */
+		wa_write_or(wal,
+			    GEN10_SCRATCH_LNCF2,
+			    PMFLUSHDONE_LNICRSDROP |
+			    PMFLUSH_GAPL3UNBLOCK |
+			    PMFLUSHDONE_LNEBLK);
+
+		/* Wa_1406609255:icl (pre-prod) */
+		if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0))
+			wa_write_or(wal,
+				    GEN7_SARCHKMD,
+				    GEN7_DISABLE_DEMAND_PREFETCH |
+				    GEN7_DISABLE_SAMPLER_PREFETCH);
+	}
+
+	if (IS_GEN9(i915) || IS_CANNONLAKE(i915)) {
+		/* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,cnl */
+		wa_masked_en(wal,
+			     GEN7_FF_SLICE_CS_CHICKEN1,
+			     GEN9_FFSC_PERCTX_PREEMPT_CTRL);
+	}
+
+	if (IS_SKYLAKE(i915) || IS_KABYLAKE(i915) || IS_COFFEELAKE(i915)) {
+		/* WaEnableGapsTsvCreditFix:skl,kbl,cfl */
+		wa_write_or(wal,
+			    GEN8_GARBCNTL,
+			    GEN9_GAPS_TSV_CREDIT_DISABLE);
+	}
+
+	if (IS_BROXTON(i915)) {
+		/* WaDisablePooledEuLoadBalancingFix:bxt */
+		wa_masked_en(wal,
+			     FF_SLICE_CS_CHICKEN2,
+			     GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE);
+	}
+
+	if (IS_GEN9(i915)) {
+		/* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */
+		wa_masked_en(wal,
+			     GEN9_CSFE_CHICKEN1_RCS,
+			     GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE);
+
+		/* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */
+		wa_write_or(wal,
+			    BDW_SCRATCH1,
+			    GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE);
+
+		/* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */
+		if (IS_GEN9_LP(i915))
+			wa_write_masked_or(wal,
+					   GEN8_L3SQCREG1,
+					   L3_PRIO_CREDITS_MASK,
+					   L3_GENERAL_PRIO_CREDITS(62) |
+					   L3_HIGH_PRIO_CREDITS(2));
+
+		/* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */
+		wa_write_or(wal,
+			    GEN8_L3SQCREG4,
+			    GEN8_LQSC_FLUSH_COHERENT_LINES);
+	}
+}
+
+static void xcs_engine_wa_init(struct intel_engine_cs *engine)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	struct i915_wa_list *wal = &engine->wa_list;
+
+	/* WaKBLVECSSemaphoreWaitPoll:kbl */
+	if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) {
+		wa_write(wal,
+			 RING_SEMA_WAIT_POLL(engine->mmio_base),
+			 1);
+	}
+}
+
+void intel_engine_init_workarounds(struct intel_engine_cs *engine)
+{
+	struct i915_wa_list *wal = &engine->wa_list;
+
+	if (GEM_WARN_ON(INTEL_GEN(engine->i915) < 8))
+		return;
+
+	wa_init_start(wal, engine->name);
+
+	if (engine->id == RCS)
+		rcs_engine_wa_init(engine);
+	else
+		xcs_engine_wa_init(engine);
+
+	wa_init_finish(wal);
+}
+
+void intel_engine_apply_workarounds(struct intel_engine_cs *engine)
+{
+	wa_list_apply(engine->i915, &engine->wa_list);
+}
+
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftests/intel_workarounds.c"
 #endif

drivers/gpu/drm/i915/intel_workarounds.h

@@ -35,4 +35,7 @@ void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv);
 
 void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine);
 
+void intel_engine_init_workarounds(struct intel_engine_cs *engine);
+void intel_engine_apply_workarounds(struct intel_engine_cs *engine);
+
 #endif