Merging upstream version 6.8.9.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

5 files changed, 73 insertions, 33 deletions

diff --git a/drivers/gpu/drm/ttm/tests/ttm_device_test.c b/drivers/gpu/drm/ttm/tests/ttm_device_test.c
index b1b423b68c..19eaff22e6 100644
--- a/drivers/gpu/drm/ttm/tests/ttm_device_test.c
+++ b/drivers/gpu/drm/ttm/tests/ttm_device_test.c
@@ -175,7 +175,7 @@ static void ttm_device_init_pools(struct kunit *test)
 
 	if (params->pools_init_expected) {
 		for (int i = 0; i < TTM_NUM_CACHING_TYPES; ++i) {
-			for (int j = 0; j <= MAX_ORDER; ++j) {
+			for (int j = 0; j < NR_PAGE_ORDERS; ++j) {
 				pt = pool->caching[i].orders[j];
 				KUNIT_EXPECT_PTR_EQ(test, pt.pool, pool);
 				KUNIT_EXPECT_EQ(test, pt.caching, i);
diff --git a/drivers/gpu/drm/ttm/tests/ttm_pool_test.c b/drivers/gpu/drm/ttm/tests/ttm_pool_test.c
index 2d9cae8cd9..cceaa18d4e 100644
--- a/drivers/gpu/drm/ttm/tests/ttm_pool_test.c
+++ b/drivers/gpu/drm/ttm/tests/ttm_pool_test.c
@@ -109,7 +109,7 @@ static const struct ttm_pool_test_case ttm_pool_basic_cases[] = {
 	},
 	{
 		.description = "Above the allocation limit",
-		.order = MAX_ORDER + 1,
+		.order = MAX_PAGE_ORDER + 1,
 	},
 	{
 		.description = "One page, with coherent DMA mappings enabled",
@@ -118,7 +118,7 @@ static const struct ttm_pool_test_case ttm_pool_basic_cases[] = {
 	},
 	{
 		.description = "Above the allocation limit, with coherent DMA mappings enabled",
-		.order = MAX_ORDER + 1,
+		.order = MAX_PAGE_ORDER + 1,
 		.use_dma_alloc = true,
 	},
 };
@@ -165,7 +165,7 @@ static void ttm_pool_alloc_basic(struct kunit *test)
 	fst_page = tt->pages[0];
 	last_page = tt->pages[tt->num_pages - 1];
 
-	if (params->order <= MAX_ORDER) {
+	if (params->order <= MAX_PAGE_ORDER) {
 		if (params->use_dma_alloc) {
 			KUNIT_ASSERT_NOT_NULL(test, (void *)fst_page->private);
 			KUNIT_ASSERT_NOT_NULL(test, (void *)last_page->private);
@@ -182,7 +182,7 @@ static void ttm_pool_alloc_basic(struct kunit *test)
 			 * order 0 blocks
 			 */
 			KUNIT_ASSERT_EQ(test, fst_page->private,
-					min_t(unsigned int, MAX_ORDER,
+					min_t(unsigned int, MAX_PAGE_ORDER,
 					      params->order));
 			KUNIT_ASSERT_EQ(test, last_page->private, 0);
 		}
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index e58b7e2498..f95b0406ca 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -370,7 +370,13 @@ static void ttm_bo_release(struct kref *kref)
 			spin_unlock(&bo->bdev->lru_lock);
 
 			INIT_WORK(&bo->delayed_delete, ttm_bo_delayed_delete);
-			queue_work(bdev->wq, &bo->delayed_delete);
+
+			/* Schedule the worker on the closest NUMA node. This
+			 * improves performance since system memory might be
+			 * cleared on free and that is best done on a CPU core
+			 * close to it.
+			 */
+			queue_work_node(bdev->pool.nid, bdev->wq, &bo->delayed_delete);
 			return;
 		}
 
@@ -764,7 +770,7 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo,
  * This function may sleep while waiting for space to become available.
  * Returns:
  * -EBUSY: No space available (only if no_wait == 1).
- * -ENOMEM: Could not allocate memory for the buffer object, either due to
+ * -ENOSPC: Could not allocate space for the buffer object, either due to
  * fragmentation or concurrent allocators.
  * -ERESTARTSYS: An interruptible sleep was interrupted by a signal.
  */
@@ -824,7 +830,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 			goto error;
 	}
 
-	ret = -ENOMEM;
+	ret = -ENOSPC;
 	if (!type_found) {
 		pr_err(TTM_PFX "No compatible memory type found\n");
 		ret = -EINVAL;
@@ -910,6 +916,9 @@ int ttm_bo_validate(struct ttm_buffer_object *bo,
 		return -EINVAL;
 
 	ret = ttm_bo_move_buffer(bo, placement, ctx);
+	/* For backward compatibility with userspace */
+	if (ret == -ENOSPC)
+		return -ENOMEM;
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/ttm/ttm_device.c b/drivers/gpu/drm/ttm/ttm_device.c
index d48b39132b..7602796005 100644
--- a/drivers/gpu/drm/ttm/ttm_device.c
+++ b/drivers/gpu/drm/ttm/ttm_device.c
@@ -95,11 +95,17 @@ static int ttm_global_init(void)
 	ttm_pool_mgr_init(num_pages);
 	ttm_tt_mgr_init(num_pages, num_dma32);
 
-	glob->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32);
+	glob->dummy_read_page = alloc_page(__GFP_ZERO | GFP_DMA32 |
+					   __GFP_NOWARN);
 
+	/* Retry without GFP_DMA32 for platforms DMA32 is not available */
 	if (unlikely(glob->dummy_read_page == NULL)) {
-		ret = -ENOMEM;
-		goto out;
+		glob->dummy_read_page = alloc_page(__GFP_ZERO);
+		if (unlikely(glob->dummy_read_page == NULL)) {
+			ret = -ENOMEM;
+			goto out;
+		}
+		pr_warn("Using GFP_DMA32 fallback for dummy_read_page\n");
 	}
 
 	INIT_LIST_HEAD(&glob->device_list);
@@ -195,7 +201,7 @@ int ttm_device_init(struct ttm_device *bdev, const struct ttm_device_funcs *func
 		    bool use_dma_alloc, bool use_dma32)
 {
 	struct ttm_global *glob = &ttm_glob;
-	int ret;
+	int ret, nid;
 
 	if (WARN_ON(vma_manager == NULL))
 		return -EINVAL;
@@ -204,7 +210,8 @@ int ttm_device_init(struct ttm_device *bdev, const struct ttm_device_funcs *func
 	if (ret)
 		return ret;
 
-	bdev->wq = alloc_workqueue("ttm", WQ_MEM_RECLAIM | WQ_HIGHPRI, 16);
+	bdev->wq = alloc_workqueue("ttm",
+				   WQ_MEM_RECLAIM | WQ_HIGHPRI | WQ_UNBOUND, 16);
 	if (!bdev->wq) {
 		ttm_global_release();
 		return -ENOMEM;
@@ -213,7 +220,13 @@ int ttm_device_init(struct ttm_device *bdev, const struct ttm_device_funcs *func
 	bdev->funcs = funcs;
 
 	ttm_sys_man_init(bdev);
-	ttm_pool_init(&bdev->pool, dev, NUMA_NO_NODE, use_dma_alloc, use_dma32);
+
+	if (dev)
+		nid = dev_to_node(dev);
+	else
+		nid = NUMA_NO_NODE;
+
+	ttm_pool_init(&bdev->pool, dev, nid, use_dma_alloc, use_dma32);
 
 	bdev->vma_manager = vma_manager;
 	spin_lock_init(&bdev->lru_lock);
diff --git a/drivers/gpu/drm/ttm/ttm_pool.c b/drivers/gpu/drm/ttm/ttm_pool.c
index dbc96984d3..6e1fd6985f 100644
--- a/drivers/gpu/drm/ttm/ttm_pool.c
+++ b/drivers/gpu/drm/ttm/ttm_pool.c
@@ -65,11 +65,11 @@ module_param(page_pool_size, ulong, 0644);
 
 static atomic_long_t allocated_pages;
 
-static struct ttm_pool_type global_write_combined[MAX_ORDER + 1];
-static struct ttm_pool_type global_uncached[MAX_ORDER + 1];
+static struct ttm_pool_type global_write_combined[NR_PAGE_ORDERS];
+static struct ttm_pool_type global_uncached[NR_PAGE_ORDERS];
 
-static struct ttm_pool_type global_dma32_write_combined[MAX_ORDER + 1];
-static struct ttm_pool_type global_dma32_uncached[MAX_ORDER + 1];
+static struct ttm_pool_type global_dma32_write_combined[NR_PAGE_ORDERS];
+static struct ttm_pool_type global_dma32_uncached[NR_PAGE_ORDERS];
 
 static spinlock_t shrinker_lock;
 static struct list_head shrinker_list;
@@ -288,17 +288,23 @@ static struct ttm_pool_type *ttm_pool_select_type(struct ttm_pool *pool,
 						  enum ttm_caching caching,
 						  unsigned int order)
 {
-	if (pool->use_dma_alloc || pool->nid != NUMA_NO_NODE)
+	if (pool->use_dma_alloc)
 		return &pool->caching[caching].orders[order];
 
 #ifdef CONFIG_X86
 	switch (caching) {
 	case ttm_write_combined:
+		if (pool->nid != NUMA_NO_NODE)
+			return &pool->caching[caching].orders[order];
+
 		if (pool->use_dma32)
 			return &global_dma32_write_combined[order];
 
 		return &global_write_combined[order];
 	case ttm_uncached:
+		if (pool->nid != NUMA_NO_NODE)
+			return &pool->caching[caching].orders[order];
+
 		if (pool->use_dma32)
 			return &global_dma32_uncached[order];
 
@@ -447,7 +453,7 @@ int ttm_pool_alloc(struct ttm_pool *pool, struct ttm_tt *tt,
 	else
 		gfp_flags |= GFP_HIGHUSER;
 
-	for (order = min_t(unsigned int, MAX_ORDER, __fls(num_pages));
+	for (order = min_t(unsigned int, MAX_PAGE_ORDER, __fls(num_pages));
 	     num_pages;
 	     order = min_t(unsigned int, order, __fls(num_pages))) {
 		struct ttm_pool_type *pt;
@@ -566,11 +572,17 @@ void ttm_pool_init(struct ttm_pool *pool, struct device *dev,
 	pool->use_dma_alloc = use_dma_alloc;
 	pool->use_dma32 = use_dma32;
 
-	if (use_dma_alloc || nid != NUMA_NO_NODE) {
-		for (i = 0; i < TTM_NUM_CACHING_TYPES; ++i)
-			for (j = 0; j <= MAX_ORDER; ++j)
-				ttm_pool_type_init(&pool->caching[i].orders[j],
-						   pool, i, j);
+	for (i = 0; i < TTM_NUM_CACHING_TYPES; ++i) {
+		for (j = 0; j < NR_PAGE_ORDERS; ++j) {
+			struct ttm_pool_type *pt;
+
+			/* Initialize only pool types which are actually used */
+			pt = ttm_pool_select_type(pool, i, j);
+			if (pt != &pool->caching[i].orders[j])
+				continue;
+
+			ttm_pool_type_init(pt, pool, i, j);
+		}
 	}
 }
 EXPORT_SYMBOL(ttm_pool_init);
@@ -599,10 +611,16 @@ void ttm_pool_fini(struct ttm_pool *pool)
 {
 	unsigned int i, j;
 
-	if (pool->use_dma_alloc || pool->nid != NUMA_NO_NODE) {
-		for (i = 0; i < TTM_NUM_CACHING_TYPES; ++i)
-			for (j = 0; j <= MAX_ORDER; ++j)
-				ttm_pool_type_fini(&pool->caching[i].orders[j]);
+	for (i = 0; i < TTM_NUM_CACHING_TYPES; ++i) {
+		for (j = 0; j < NR_PAGE_ORDERS; ++j) {
+			struct ttm_pool_type *pt;
+
+			pt = ttm_pool_select_type(pool, i, j);
+			if (pt != &pool->caching[i].orders[j])
+				continue;
+
+			ttm_pool_type_fini(pt);
+		}
 	}
 
 	/* We removed the pool types from the LRU, but we need to also make sure
@@ -656,7 +674,7 @@ static void ttm_pool_debugfs_header(struct seq_file *m)
 	unsigned int i;
 
 	seq_puts(m, "\t ");
-	for (i = 0; i <= MAX_ORDER; ++i)
+	for (i = 0; i < NR_PAGE_ORDERS; ++i)
 		seq_printf(m, " ---%2u---", i);
 	seq_puts(m, "\n");
 }
@@ -667,7 +685,7 @@ static void ttm_pool_debugfs_orders(struct ttm_pool_type *pt,
 {
 	unsigned int i;
 
-	for (i = 0; i <= MAX_ORDER; ++i)
+	for (i = 0; i < NR_PAGE_ORDERS; ++i)
 		seq_printf(m, " %8u", ttm_pool_type_count(&pt[i]));
 	seq_puts(m, "\n");
 }
@@ -776,7 +794,7 @@ int ttm_pool_mgr_init(unsigned long num_pages)
 	spin_lock_init(&shrinker_lock);
 	INIT_LIST_HEAD(&shrinker_list);
 
-	for (i = 0; i <= MAX_ORDER; ++i) {
+	for (i = 0; i < NR_PAGE_ORDERS; ++i) {
 		ttm_pool_type_init(&global_write_combined[i], NULL,
 				   ttm_write_combined, i);
 		ttm_pool_type_init(&global_uncached[i], NULL, ttm_uncached, i);
@@ -816,7 +834,7 @@ void ttm_pool_mgr_fini(void)
 {
 	unsigned int i;
 
-	for (i = 0; i <= MAX_ORDER; ++i) {
+	for (i = 0; i < NR_PAGE_ORDERS; ++i) {
 		ttm_pool_type_fini(&global_write_combined[i]);
 		ttm_pool_type_fini(&global_uncached[i]);