GPUCORE-41611: Debug num_valid_entries > 512 warning issue

Adding extra output for the num_valid_entries > 512 assertion.

Provenance: https://code.ipdelivery.arm.com/c/GPU/mali-ddk/+/6548
Bug: 315967882
Test: Local boot and run of GFXBench5
Signed-off-by: Renato Grottesi <otaner@google.com>
Change-Id: I9f7b385777688e77389656f8bfd06c0bc493a567

5 files changed, 159 insertions, 13 deletions

diff --git a/mali_kbase/context/mali_kbase_context.c b/mali_kbase/context/mali_kbase_context.c
index 9b8cc5d..7fb7b43 100644
--- a/mali_kbase/context/mali_kbase_context.c
+++ b/mali_kbase/context/mali_kbase_context.c
@@ -327,7 +327,9 @@ void kbase_context_common_term(struct kbase_context *kctx)
 
 	pages = atomic_read(&kctx->used_pages);
 	if (pages != 0)
-		dev_warn(kctx->kbdev->dev, "%s: %d pages in use!\n", __func__, pages);
+		dev_warn(kctx->kbdev->dev, "%s: %d pages (pgd cnt %u) in use for kctx %d_%d",
+			 __func__, pages,
+			 kctx->pgd_cnt, kctx->tgid, kctx->id);
 
 	WARN_ON(atomic_read(&kctx->nonmapped_pages) != 0);
 
diff --git a/mali_kbase/mali_kbase_defs.h b/mali_kbase/mali_kbase_defs.h
index d7adc7d..1162b95 100644
--- a/mali_kbase/mali_kbase_defs.h
+++ b/mali_kbase/mali_kbase_defs.h
@@ -2032,6 +2032,7 @@ struct kbase_context {
 	atomic_t used_pages;
 	atomic_t nonmapped_pages;
 	atomic_t permanent_mapped_pages;
+	u32 pgd_cnt;
 
 	struct kbase_mem_pool_group mem_pools;
 
diff --git a/mali_kbase/mali_kbase_jd.c b/mali_kbase/mali_kbase_jd.c
index 08c840a..6c8dd11 100644
--- a/mali_kbase/mali_kbase_jd.c
+++ b/mali_kbase/mali_kbase_jd.c
@@ -1383,8 +1383,13 @@ void kbase_jd_done_worker(struct kthread_work *data)
 
 	if ((katom->event_code != BASE_JD_EVENT_DONE) && !kbase_ctx_flag(katom->kctx, KCTX_DYING) &&
 	    !kbase_ctx_flag(katom->kctx, KCTX_PAGE_FAULT_REPORT_SKIP))
-		dev_err(kbdev->dev, "t6xx: GPU fault 0x%02lx from job slot %d\n",
-			(unsigned long)katom->event_code, katom->slot_nr);
+		dev_err(kbdev->dev,
+			"Atom %d (age %x core_req %x) of kctx %d_%d"
+			" completed with event_code %x on job slot %d",
+			kbase_jd_atom_id(kctx, katom),
+			katom->age, katom->core_req,
+			kctx->tgid, kctx->id,
+			katom->event_code, katom->slot_nr);
 
 	/* Retain state before the katom disappears */
 	kbasep_js_atom_retained_state_copy(&katom_retained_state, katom);
diff --git a/mali_kbase/mali_kbase_mem.c b/mali_kbase/mali_kbase_mem.c
index c3eb2cc..337fe2a 100644
--- a/mali_kbase/mali_kbase_mem.c
+++ b/mali_kbase/mali_kbase_mem.c
@@ -2706,7 +2706,7 @@ static int kbase_jit_grow(struct kbase_context *kctx, const struct base_jit_allo
 {
 	size_t delta;
 	size_t pages_required;
-	size_t old_size;
+	size_t old_size, old_size_orig;
 	struct kbase_mem_pool *pool;
 	int ret = -ENOMEM;
 	struct tagged_addr *gpu_pages;
@@ -2725,6 +2725,7 @@ static int kbase_jit_grow(struct kbase_context *kctx, const struct base_jit_allo
 	if (reg->gpu_alloc->nents >= info->commit_pages)
 		goto done;
 
+	old_size_orig = reg->gpu_alloc->nents;
 	/* Allocate some more pages */
 	delta = info->commit_pages - reg->gpu_alloc->nents;
 	pages_required = delta;
@@ -2802,6 +2803,13 @@ static int kbase_jit_grow(struct kbase_context *kctx, const struct base_jit_allo
 	spin_unlock(&kctx->mem_partials_lock);
 
 	ret = kbase_mem_grow_gpu_mapping(kctx, reg, info->commit_pages, old_size, mmu_sync_info);
+	if (unlikely(old_size_orig != old_size)) {
+		dev_warn(
+			kctx->kbdev->dev,
+			"JIT alloc %llx backing changed from %zu to %zu during grow for kctx %d_%d",
+			reg->start_pfn << PAGE_SHIFT, old_size_orig, old_size,
+			kctx->tgid, kctx->id);
+	}
 	/*
 	 * The grow failed so put the allocation back in the
 	 * pool and return failure.
diff --git a/mali_kbase/mmu/mali_kbase_mmu.c b/mali_kbase/mmu/mali_kbase_mmu.c
index 0e73641..6bab554 100644
--- a/mali_kbase/mmu/mali_kbase_mmu.c
+++ b/mali_kbase/mmu/mali_kbase_mmu.c
@@ -334,6 +334,7 @@ static void kbase_mmu_account_freed_pgd(struct kbase_device *kbdev, struct kbase
 	if (mmut->kctx) {
 		kbase_process_page_usage_dec(mmut->kctx, 1);
 		atomic_sub(1, &mmut->kctx->used_pages);
+		mmut->kctx->pgd_cnt--;
 	}
 
 	kbase_trace_gpu_mem_usage_dec(kbdev, mmut->kctx, 1);
@@ -1162,9 +1163,12 @@ page_fault_retry:
 	if (fault_rel_pfn < current_backed_size) {
 		struct kbase_mmu_hw_op_param op_param;
 
-		dev_dbg(kbdev->dev,
-			"Page fault @ VA 0x%llx in allocated region 0x%llx-0x%llx of growable TMEM: Ignoring",
-			fault->addr, region->start_pfn, region->start_pfn + current_backed_size);
+		dev_warn_ratelimited(kbdev->dev,
+			"Page fault @VA 0x%llx in allocated region 0x%llx-0x%llx"
+			" (status %x) of kctx %d_%d (as %d)",
+			fault->addr, region->start_pfn << PAGE_SHIFT,
+			(region->start_pfn + current_backed_size) << PAGE_SHIFT,
+			fault_status, kctx->tgid, kctx->id, as_no);
 
 		kbase_mmu_hw_clear_fault(kbdev, faulting_as, KBASE_MMU_FAULT_TYPE_PAGE);
 		/* [1] in case another page fault occurred while we were
@@ -1460,6 +1464,7 @@ static phys_addr_t kbase_mmu_alloc_pgd(struct kbase_device *kbdev, struct kbase_
 		new_page_count = atomic_add_return(1, &mmut->kctx->used_pages);
 		KBASE_TLSTREAM_AUX_PAGESALLOC(kbdev, mmut->kctx->id, (u64)new_page_count);
 		kbase_process_page_usage_inc(mmut->kctx, 1);
+		mmut->kctx->pgd_cnt++;
 	}
 
 	atomic_add(1, &kbdev->memdev.used_pages);
@@ -1782,6 +1787,77 @@ static void mmu_flush_invalidate_insert_pages(struct kbase_device *kbdev,
 		mmu_flush_invalidate(kbdev, mmut->kctx, as_nr, &op_param);
 }
 
+static void dump_region_info(struct kbase_device *kbdev, struct kbase_va_region *reg)
+{
+	if (!reg)
+		return;
+
+#if MALI_JIT_PRESSURE_LIMIT_BASE
+	dev_err(kbdev->dev,
+		"Region info: GPU VA %llx, nr_pages %zu, used_pages %zu,"
+		" commit_pages %zu, flags %lx, type %d",
+		reg->start_pfn << PAGE_SHIFT,
+		reg->nr_pages, reg->used_pages, reg->gpu_alloc->nents,
+		reg->flags, reg->gpu_alloc->type);
+#else
+	dev_err(kbdev->dev,
+		"Region info: GPU VA %llx, nr_pages %zu,"
+		" commit_pages %zu, flags %lx, type %d",
+		reg->start_pfn << PAGE_SHIFT,
+		reg->nr_pages, reg->gpu_alloc->nents,
+		reg->flags, reg->gpu_alloc->type);
+#endif
+}
+
+static void dump_zones_info(struct kbase_context *const kctx)
+{
+	enum kbase_memory_zone zone_idx;
+
+	if (!kctx)
+		return;
+
+	for (zone_idx = 0; zone_idx < CONTEXT_ZONE_MAX; zone_idx++) {
+		struct kbase_reg_zone *reg_zone = &kctx->reg_zone[zone_idx];
+
+		if (!reg_zone->base_pfn)
+			continue;
+		dev_warn(
+			kctx->kbdev->dev,
+			"%15s %u 0x%.16llx 0x%.16llx",
+			kbase_reg_zone_get_name(zone_idx), zone_idx, reg_zone->base_pfn,
+			reg_zone->va_size_pages);
+	}
+}
+
+static void dump_custom_va_zone_regions_info(struct kbase_context *const kctx, u64 insert_vpfn)
+{
+	struct kbase_reg_zone *reg_zone;
+
+	if (!kctx)
+		return;
+
+	reg_zone = &kctx->reg_zone[CUSTOM_VA_ZONE];
+
+	if (reg_zone->base_pfn &&
+	    (insert_vpfn >= reg_zone->base_pfn) &&
+	    (insert_vpfn < kbase_reg_zone_end_pfn(reg_zone))) {
+		struct rb_root *rbtree = &reg_zone->reg_rbtree;
+		struct rb_node *p;
+
+		dev_err(kctx->kbdev->dev, "Dumping CUSTOM_VA regions info");
+
+		for (p = rb_first(rbtree); p; p = rb_next(p)) {
+			struct kbase_va_region *reg = rb_entry(p, struct kbase_va_region, rblink);
+
+			/* Empty region - ignore */
+			if (reg->gpu_alloc == NULL)
+				continue;
+
+			dump_region_info(kctx->kbdev, reg);
+		}
+	}
+}
+
 /**
  * update_parent_pgds() - Updates the page table from bottom level towards
  *                        the top level to insert a new ATE
@@ -1797,6 +1873,7 @@ static void mmu_flush_invalidate_insert_pages(struct kbase_device *kbdev,
  *                  the physical addresses of newly allocated PGDs from index
  *                  insert_level+1 to cur_level, and an existing PGD at index
  *                  insert_level.
+ * @reg:            Pointer to the VA region that needs to be mapped.
  *
  * The newly allocated PGDs are linked from the bottom level up and inserted into the PGD
  * at insert_level which already exists in the MMU Page Tables. Migration status is also
@@ -1809,7 +1886,8 @@ static void mmu_flush_invalidate_insert_pages(struct kbase_device *kbdev,
  */
 static int update_parent_pgds(struct kbase_device *kbdev, struct kbase_mmu_table *mmut,
 			      int cur_level, int insert_level, u64 insert_vpfn,
-			      phys_addr_t *pgds_to_insert)
+			      phys_addr_t *pgds_to_insert,
+			      struct kbase_va_region *reg)
 {
 	int pgd_index;
 	int err = 0;
@@ -1827,6 +1905,8 @@ static int update_parent_pgds(struct kbase_device *kbdev, struct kbase_mmu_table
 		u64 parent_vpfn = (insert_vpfn >> ((3 - parent_index) * 9)) & 0x1FF;
 		struct page *parent_page = pfn_to_page(PFN_DOWN(parent_pgd));
 		u64 *parent_page_va;
+		u32 mmut_kctx_tgid = mmut->kctx ? mmut->kctx->tgid : 0;
+		u32 mmut_kctx_id = mmut->kctx ? mmut->kctx->id : 0;
 
 		if (WARN_ON_ONCE(target_pgd == KBASE_INVALID_PHYSICAL_ADDRESS)) {
 			err = -EFAULT;
@@ -1844,8 +1924,32 @@ static int update_parent_pgds(struct kbase_device *kbdev, struct kbase_mmu_table
 		current_valid_entries = kbdev->mmu_mode->get_num_valid_entries(parent_page_va);
 
 		kbdev->mmu_mode->entry_set_pte(&pte, target_pgd);
+		if (WARN_ON_ONCE((parent_page_va[parent_vpfn] & 1UL) != 0)) {
+			dev_err(kbdev->dev,
+				"Valid PTE found at index %lld of"
+				" Level %d table page for VA %llx of kctx %d_%d",
+				parent_vpfn, parent_index, insert_vpfn << 12,
+				mmut_kctx_tgid, mmut_kctx_id);
+			dump_region_info(kbdev, reg);
+			dev_err(kbdev->dev, "cur_level %d insert_level %d current_valid_entries %u",
+				cur_level, insert_level, current_valid_entries);
+			dump_zones_info(mmut->kctx);
+			dump_custom_va_zone_regions_info(mmut->kctx, insert_vpfn);
+		}
 		parent_page_va[parent_vpfn] = kbdev->mgm_dev->ops.mgm_update_gpu_pte(
 			kbdev->mgm_dev, MGM_DEFAULT_PTE_GROUP, parent_index, pte);
+		if (unlikely((current_valid_entries + 1) > KBASE_MMU_PAGE_ENTRIES)) {
+			dev_err(kbdev->dev,
+				"Unexpected valid entry count after updating entry at index %lld of"
+				" Level %d table page for VA %llx of kctx %d_%d",
+				parent_vpfn, parent_index, insert_vpfn << 12,
+				mmut_kctx_tgid, mmut_kctx_id);
+			dump_region_info(kbdev, reg);
+			dev_err(kbdev->dev, "cur_level %d insert_level %d current_valid_entries %u",
+				cur_level, insert_level, current_valid_entries);
+			dump_zones_info(mmut->kctx);
+			dump_custom_va_zone_regions_info(mmut->kctx, insert_vpfn);
+		}
 		kbdev->mmu_mode->set_num_valid_entries(parent_page_va, current_valid_entries + 1);
 		kbase_kunmap(parent_page, parent_page_va);
 
@@ -2111,7 +2215,7 @@ static int kbase_mmu_insert_single_page(struct kbase_context *kctx, u64 start_vp
 
 		if (newly_created_pgd) {
 			err = update_parent_pgds(kbdev, mmut, cur_level, insert_level, insert_vpfn,
-						 new_pgds);
+						 new_pgds, NULL);
 			if (err) {
 				dev_err(kbdev->dev, "%s: update_parent_pgds() failed (%d)",
 					__func__, err);
@@ -2395,7 +2499,11 @@ static int mmu_insert_pages_no_flush(struct kbase_device *kbdev, struct kbase_mm
 					 * should be performed with
 					 * kbase_mmu_update_pages()
 					 */
-					WARN_ON_ONCE((*target & 1UL) != 0);
+					if (WARN_ON_ONCE((*target & 1UL) != 0))
+						dev_err(kbdev->dev,
+							"valid ATE found at entry %u of"
+							" Level 3 table page for VA %llx",
+							ofs, insert_vpfn << 12);
 
 					*target = kbase_mmu_create_ate(kbdev,
 								       as_tagged(page_address),
@@ -2413,6 +2521,21 @@ static int mmu_insert_pages_no_flush(struct kbase_device *kbdev, struct kbase_mm
 			num_of_valid_entries += count;
 		}
 
+		if (unlikely(num_of_valid_entries > KBASE_MMU_PAGE_ENTRIES)) {
+			u32 mmut_kctx_tgid = mmut->kctx ? mmut->kctx->tgid : 0;
+			u32 mmut_kctx_id = mmut->kctx ? mmut->kctx->id : 0;
+			dev_err(kbdev->dev,
+				"Unexpected valid entry count after updating %u entries"
+				" at index %d of Level %d table page for VA %llx of kctx %d_%d",
+				count, vindex, cur_level, insert_vpfn << 12, mmut_kctx_tgid,
+				mmut_kctx_id);
+			dump_region_info(kbdev, reg);
+			dev_err(kbdev->dev,
+				"newly_created_pgd %d insert_level %d num_of_valid_entries %u",
+				newly_created_pgd, insert_level, num_of_valid_entries);
+			dump_zones_info(mmut->kctx);
+			dump_custom_va_zone_regions_info(mmut->kctx, insert_vpfn);
+		}
 		mmu_mode->set_num_valid_entries(pgd_page, num_of_valid_entries);
 
 		if (dirty_pgds)
@@ -2433,7 +2556,7 @@ static int mmu_insert_pages_no_flush(struct kbase_device *kbdev, struct kbase_mm
 
 		if (newly_created_pgd) {
 			err = update_parent_pgds(kbdev, mmut, cur_level, insert_level, insert_vpfn,
-						 new_pgds);
+						 new_pgds, reg);
 			if (err) {
 				dev_err(kbdev->dev, "%s: update_parent_pgds() failed (%d)",
 					__func__, err);
@@ -2918,8 +3041,15 @@ static int kbase_mmu_teardown_pgd_pages(struct kbase_device *kbdev, struct kbase
 			if (mmu_mode->ate_is_valid(page[index], level))
 				break; /* keep the mapping */
 			else if (!mmu_mode->pte_is_valid(page[index], level)) {
-				dev_warn(kbdev->dev, "Invalid PTE found @ level %d for VA %llx",
-					 level, vpfn << PAGE_SHIFT);
+				u32 mmut_kctx_tgid = mmut->kctx ? mmut->kctx->tgid : 0;
+				u32 mmut_kctx_id = mmut->kctx ? mmut->kctx->id : 0;
+				dev_warn(kbdev->dev,
+					 "Invalid PTE found @ level %d for VA %llx"
+					 " (nr %zu, valid entries %u) of kctx %d_%d",
+					 level, vpfn << PAGE_SHIFT, nr,
+					 mmu_mode->get_num_valid_entries(page),
+					 mmut_kctx_tgid, mmut_kctx_id);
+				WARN_ON_ONCE(1);
 				/* nothing here, advance to the next PTE of the current level */
 				count = (1 << ((3 - level) * 9));
 				count -= (vpfn & (count - 1));