Merge tag 'drm-intel-fixes-2022-05-20' of...

Merge tag 'drm-intel-fixes-2022-05-20' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes

- fix for #5806: GPU hangs and display artifacts on 5.18-rc3 on Intel GM45
- reject DMC with out-of-spec MMIO (Cc: stable)
- correctly mark guilty contexts on GuC reset.
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/YocqqvG6PbYx3QgJ@jlahtine-mobl.ger.corp.intel.com

drivers/gpu/drm/i915/display/intel_dmc.c

@@ -367,6 +367,44 @@ static void dmc_set_fw_offset(struct intel_dmc *dmc,
 	}
 }
 
+static bool dmc_mmio_addr_sanity_check(struct intel_dmc *dmc,
+				       const u32 *mmioaddr, u32 mmio_count,
+				       int header_ver, u8 dmc_id)
+{
+	struct drm_i915_private *i915 = container_of(dmc, typeof(*i915), dmc);
+	u32 start_range, end_range;
+	int i;
+
+	if (dmc_id >= DMC_FW_MAX) {
+		drm_warn(&i915->drm, "Unsupported firmware id %u\n", dmc_id);
+		return false;
+	}
+
+	if (header_ver == 1) {
+		start_range = DMC_MMIO_START_RANGE;
+		end_range = DMC_MMIO_END_RANGE;
+	} else if (dmc_id == DMC_FW_MAIN) {
+		start_range = TGL_MAIN_MMIO_START;
+		end_range = TGL_MAIN_MMIO_END;
+	} else if (DISPLAY_VER(i915) >= 13) {
+		start_range = ADLP_PIPE_MMIO_START;
+		end_range = ADLP_PIPE_MMIO_END;
+	} else if (DISPLAY_VER(i915) >= 12) {
+		start_range = TGL_PIPE_MMIO_START(dmc_id);
+		end_range = TGL_PIPE_MMIO_END(dmc_id);
+	} else {
+		drm_warn(&i915->drm, "Unknown mmio range for sanity check");
+		return false;
+	}
+
+	for (i = 0; i < mmio_count; i++) {
+		if (mmioaddr[i] < start_range || mmioaddr[i] > end_range)
+			return false;
+	}
+
+	return true;
+}
+
 static u32 parse_dmc_fw_header(struct intel_dmc *dmc,
 			       const struct intel_dmc_header_base *dmc_header,
 			       size_t rem_size, u8 dmc_id)
@@ -436,6 +474,12 @@ static u32 parse_dmc_fw_header(struct intel_dmc *dmc,
 		return 0;
 	}
 
+	if (!dmc_mmio_addr_sanity_check(dmc, mmioaddr, mmio_count,
+					dmc_header->header_ver, dmc_id)) {
+		drm_err(&i915->drm, "DMC firmware has Wrong MMIO Addresses\n");
+		return 0;
+	}
+
 	for (i = 0; i < mmio_count; i++) {
 		dmc_info->mmioaddr[i] = _MMIO(mmioaddr[i]);
 		dmc_info->mmiodata[i] = mmiodata[i];

drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c

@@ -1252,14 +1252,12 @@ static void *reloc_iomap(struct i915_vma *batch,
 		 * Only attempt to pin the batch buffer to ggtt if the current batch
 		 * is not inside ggtt, or the batch buffer is not misplaced.
 		 */
-		if (!i915_is_ggtt(batch->vm)) {
+		if (!i915_is_ggtt(batch->vm) ||
+		    !i915_vma_misplaced(batch, 0, 0, PIN_MAPPABLE)) {
 			vma = i915_gem_object_ggtt_pin_ww(obj, &eb->ww, NULL, 0, 0,
 							  PIN_MAPPABLE |
 							  PIN_NONBLOCK /* NOWARN */ |
 							  PIN_NOEVICT);
-		} else if (i915_vma_is_map_and_fenceable(batch)) {
-			__i915_vma_pin(batch);
-			vma = batch;
 		}
 
 		if (vma == ERR_PTR(-EDEADLK))

drivers/gpu/drm/i915/gt/intel_reset.c

@@ -806,7 +806,7 @@ static int gt_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask)
 		__intel_engine_reset(engine, stalled_mask & engine->mask);
 	local_bh_enable();
 
-	intel_uc_reset(&gt->uc, true);
+	intel_uc_reset(&gt->uc, ALL_ENGINES);
 
 	intel_ggtt_restore_fences(gt->ggtt);
 

drivers/gpu/drm/i915/gt/uc/intel_guc.h

@@ -438,7 +438,7 @@ int intel_guc_global_policies_update(struct intel_guc *guc);
 void intel_guc_context_ban(struct intel_context *ce, struct i915_request *rq);
 
 void intel_guc_submission_reset_prepare(struct intel_guc *guc);
-void intel_guc_submission_reset(struct intel_guc *guc, bool stalled);
+void intel_guc_submission_reset(struct intel_guc *guc, intel_engine_mask_t stalled);
 void intel_guc_submission_reset_finish(struct intel_guc *guc);
 void intel_guc_submission_cancel_requests(struct intel_guc *guc);
 

drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c

@@ -1590,9 +1590,9 @@ __unwind_incomplete_requests(struct intel_context *ce)
 	spin_unlock_irqrestore(&sched_engine->lock, flags);
 }
 
-static void __guc_reset_context(struct intel_context *ce, bool stalled)
+static void __guc_reset_context(struct intel_context *ce, intel_engine_mask_t stalled)
 {
-	bool local_stalled;
+	bool guilty;
 	struct i915_request *rq;
 	unsigned long flags;
 	u32 head;
@@ -1620,7 +1620,7 @@ static void __guc_reset_context(struct intel_context *ce, bool stalled)
 		if (!intel_context_is_pinned(ce))
 			goto next_context;
 
-		local_stalled = false;
+		guilty = false;
 		rq = intel_context_find_active_request(ce);
 		if (!rq) {
 			head = ce->ring->tail;
@@ -1628,14 +1628,14 @@ static void __guc_reset_context(struct intel_context *ce, bool stalled)
 		}
 
 		if (i915_request_started(rq))
-			local_stalled = true;
+			guilty = stalled & ce->engine->mask;
 
 		GEM_BUG_ON(i915_active_is_idle(&ce->active));
 		head = intel_ring_wrap(ce->ring, rq->head);
 
-		__i915_request_reset(rq, local_stalled && stalled);
+		__i915_request_reset(rq, guilty);
 out_replay:
-		guc_reset_state(ce, head, local_stalled && stalled);
+		guc_reset_state(ce, head, guilty);
 next_context:
 		if (i != number_children)
 			ce = list_next_entry(ce, parallel.child_link);
@@ -1645,7 +1645,7 @@ static void __guc_reset_context(struct intel_context *ce, bool stalled)
 	intel_context_put(parent);
 }
 
-void intel_guc_submission_reset(struct intel_guc *guc, bool stalled)
+void intel_guc_submission_reset(struct intel_guc *guc, intel_engine_mask_t stalled)
 {
 	struct intel_context *ce;
 	unsigned long index;
@@ -4013,7 +4013,7 @@ static void guc_context_replay(struct intel_context *ce)
 {
 	struct i915_sched_engine *sched_engine = ce->engine->sched_engine;
 
-	__guc_reset_context(ce, true);
+	__guc_reset_context(ce, ce->engine->mask);
 	tasklet_hi_schedule(&sched_engine->tasklet);
 }
 

drivers/gpu/drm/i915/gt/uc/intel_uc.c

@@ -593,7 +593,7 @@ void intel_uc_reset_prepare(struct intel_uc *uc)
 	__uc_sanitize(uc);
 }
 
-void intel_uc_reset(struct intel_uc *uc, bool stalled)
+void intel_uc_reset(struct intel_uc *uc, intel_engine_mask_t stalled)
 {
 	struct intel_guc *guc = &uc->guc;
 

drivers/gpu/drm/i915/gt/uc/intel_uc.h

@@ -42,7 +42,7 @@ void intel_uc_driver_late_release(struct intel_uc *uc);
 void intel_uc_driver_remove(struct intel_uc *uc);
 void intel_uc_init_mmio(struct intel_uc *uc);
 void intel_uc_reset_prepare(struct intel_uc *uc);
-void intel_uc_reset(struct intel_uc *uc, bool stalled);
+void intel_uc_reset(struct intel_uc *uc, intel_engine_mask_t stalled);
 void intel_uc_reset_finish(struct intel_uc *uc);
 void intel_uc_cancel_requests(struct intel_uc *uc);
 void intel_uc_suspend(struct intel_uc *uc);

drivers/gpu/drm/i915/i915_reg.h

@@ -5501,6 +5501,22 @@
 /* MMIO address range for DMC program (0x80000 - 0x82FFF) */
 #define DMC_MMIO_START_RANGE	0x80000
 #define DMC_MMIO_END_RANGE	0x8FFFF
+#define DMC_V1_MMIO_START_RANGE	0x80000
+#define TGL_MAIN_MMIO_START	0x8F000
+#define TGL_MAIN_MMIO_END	0x8FFFF
+#define _TGL_PIPEA_MMIO_START	0x92000
+#define _TGL_PIPEA_MMIO_END	0x93FFF
+#define _TGL_PIPEB_MMIO_START	0x96000
+#define _TGL_PIPEB_MMIO_END	0x97FFF
+#define ADLP_PIPE_MMIO_START	0x5F000
+#define ADLP_PIPE_MMIO_END	0x5FFFF
+
+#define TGL_PIPE_MMIO_START(dmc_id)	_PICK_EVEN(((dmc_id) - 1), _TGL_PIPEA_MMIO_START,\
+						_TGL_PIPEB_MMIO_START)
+
+#define TGL_PIPE_MMIO_END(dmc_id)	_PICK_EVEN(((dmc_id) - 1), _TGL_PIPEA_MMIO_END,\
+						_TGL_PIPEB_MMIO_END)
+
 #define SKL_DMC_DC3_DC5_COUNT	_MMIO(0x80030)
 #define SKL_DMC_DC5_DC6_COUNT	_MMIO(0x8002C)
 #define BXT_DMC_DC3_DC5_COUNT	_MMIO(0x80038)