Merge tag 'for-v4.6-rc/omap-fixes-a' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into omap-for-v4.6/fixes

ARM: OMAP2+: first hwmod fix for v4.6-rc

Fix a longstanding bug in the hwmod code that could cause
hardware SYSCONFIG register values to not match the kernel's
idea of what they should be, and that could result in lower
performance during IP block idle entry.

Basic build, boot, and PM test logs are available here:

http://www.pwsan.com/omap/testlogs/omap-hwmod-fixes-a-for-v4.6-rc/20160326231727/

1 files changed, 201 insertions, 143 deletions

diff --git a/drivers/gpu/drm/omapdrm/omap_gem.c b/drivers/gpu/drm/omapdrm/omap_gem.c
index 8495a1a4b617..907154f5b67c 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem.c
@@ -31,9 +31,9 @@
  */
 
 /* note: we use upper 8 bits of flags for driver-internal flags: */
-#define OMAP_BO_DMA		0x01000000	/* actually is physically contiguous */
-#define OMAP_BO_EXT_SYNC	0x02000000	/* externally allocated sync object */
-#define OMAP_BO_EXT_MEM		0x04000000	/* externally allocated memory */
+#define OMAP_BO_MEM_DMA_API	0x01000000	/* memory allocated with the dma_alloc_* API */
+#define OMAP_BO_MEM_SHMEM	0x02000000	/* memory allocated through shmem backing */
+#define OMAP_BO_MEM_DMABUF	0x08000000	/* memory imported from a dmabuf */
 
 struct omap_gem_object {
 	struct drm_gem_object base;
@@ -49,17 +49,25 @@ struct omap_gem_object {
 	uint32_t roll;
 
 	/**
-	 * If buffer is allocated physically contiguous, the OMAP_BO_DMA flag
-	 * is set and the paddr is valid.  Also if the buffer is remapped in
-	 * TILER and paddr_cnt > 0, then paddr is valid.  But if you are using
-	 * the physical address and OMAP_BO_DMA is not set, then you should
-	 * be going thru omap_gem_{get,put}_paddr() to ensure the mapping is
-	 * not removed from under your feet.
+	 * paddr contains the buffer DMA address. It is valid for
 	 *
-	 * Note that OMAP_BO_SCANOUT is a hint from userspace that DMA capable
-	 * buffer is requested, but doesn't mean that it is.  Use the
-	 * OMAP_BO_DMA flag to determine if the buffer has a DMA capable
-	 * physical address.
+	 * - buffers allocated through the DMA mapping API (with the
+	 *   OMAP_BO_MEM_DMA_API flag set)
+	 *
+	 * - buffers imported from dmabuf (with the OMAP_BO_MEM_DMABUF flag set)
+	 *   if they are physically contiguous (when sgt->orig_nents == 1)
+	 *
+	 * - buffers mapped through the TILER when paddr_cnt is not zero, in
+	 *   which case the DMA address points to the TILER aperture
+	 *
+	 * Physically contiguous buffers have their DMA address equal to the
+	 * physical address as we don't remap those buffers through the TILER.
+	 *
+	 * Buffers mapped to the TILER have their DMA address pointing to the
+	 * TILER aperture. As TILER mappings are refcounted (through paddr_cnt)
+	 * the DMA address must be accessed through omap_get_get_paddr() to
+	 * ensure that the mapping won't disappear unexpectedly. References must
+	 * be released with omap_gem_put_paddr().
 	 */
 	dma_addr_t paddr;
 
@@ -69,6 +77,12 @@ struct omap_gem_object {
 	uint32_t paddr_cnt;
 
 	/**
+	 * If the buffer has been imported from a dmabuf the OMAP_DB_DMABUF flag
+	 * is set and the sgt field is valid.
+	 */
+	struct sg_table *sgt;
+
+	/**
 	 * tiler block used when buffer is remapped in DMM/TILER.
 	 */
 	struct tiler_block *block;
@@ -91,17 +105,7 @@ struct omap_gem_object {
 	 * sync-object allocated on demand (if needed)
 	 *
 	 * Per-buffer sync-object for tracking pending and completed hw/dma
-	 * read and write operations.  The layout in memory is dictated by
-	 * the SGX firmware, which uses this information to stall the command
-	 * stream if a surface is not ready yet.
-	 *
-	 * Note that when buffer is used by SGX, the sync-object needs to be
-	 * allocated from a special heap of sync-objects.  This way many sync
-	 * objects can be packed in a page, and not waste GPU virtual address
-	 * space.  Because of this we have to have a omap_gem_set_sync_object()
-	 * API to allow replacement of the syncobj after it has (potentially)
-	 * already been allocated.  A bit ugly but I haven't thought of a
-	 * better alternative.
+	 * read and write operations.
 	 */
 	struct {
 		uint32_t write_pending;
@@ -166,16 +170,15 @@ static uint64_t mmap_offset(struct drm_gem_object *obj)
 	return drm_vma_node_offset_addr(&obj->vma_node);
 }
 
-/* GEM objects can either be allocated from contiguous memory (in which
- * case obj->filp==NULL), or w/ shmem backing (obj->filp!=NULL).  But non
- * contiguous buffers can be remapped in TILER/DMM if they need to be
- * contiguous... but we don't do this all the time to reduce pressure
- * on TILER/DMM space when we know at allocation time that the buffer
- * will need to be scanned out.
- */
-static inline bool is_shmem(struct drm_gem_object *obj)
+static bool is_contiguous(struct omap_gem_object *omap_obj)
 {
-	return obj->filp != NULL;
+	if (omap_obj->flags & OMAP_BO_MEM_DMA_API)
+		return true;
+
+	if ((omap_obj->flags & OMAP_BO_MEM_DMABUF) && omap_obj->sgt->nents == 1)
+		return true;
+
+	return false;
 }
 
 /* -----------------------------------------------------------------------------
@@ -264,6 +267,19 @@ static int omap_gem_attach_pages(struct drm_gem_object *obj)
 		for (i = 0; i < npages; i++) {
 			addrs[i] = dma_map_page(dev->dev, pages[i],
 					0, PAGE_SIZE, DMA_BIDIRECTIONAL);
+
+			if (dma_mapping_error(dev->dev, addrs[i])) {
+				dev_warn(dev->dev,
+					"%s: failed to map page\n", __func__);
+
+				for (i = i - 1; i >= 0; --i) {
+					dma_unmap_page(dev->dev, addrs[i],
+						PAGE_SIZE, DMA_BIDIRECTIONAL);
+				}
+
+				ret = -ENOMEM;
+				goto free_addrs;
+			}
 		}
 	} else {
 		addrs = kzalloc(npages * sizeof(*addrs), GFP_KERNEL);
@@ -278,6 +294,8 @@ static int omap_gem_attach_pages(struct drm_gem_object *obj)
 
 	return 0;
 
+free_addrs:
+	kfree(addrs);
 free_pages:
 	drm_gem_put_pages(obj, pages, true, false);
 
@@ -292,7 +310,7 @@ static int get_pages(struct drm_gem_object *obj, struct page ***pages)
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	int ret = 0;
 
-	if (is_shmem(obj) && !omap_obj->pages) {
+	if ((omap_obj->flags & OMAP_BO_MEM_SHMEM) && !omap_obj->pages) {
 		ret = omap_gem_attach_pages(obj);
 		if (ret) {
 			dev_err(obj->dev->dev, "could not attach pages\n");
@@ -396,7 +414,7 @@ static int fault_1d(struct drm_gem_object *obj,
 		omap_gem_cpu_sync(obj, pgoff);
 		pfn = page_to_pfn(omap_obj->pages[pgoff]);
 	} else {
-		BUG_ON(!(omap_obj->flags & OMAP_BO_DMA));
+		BUG_ON(!is_contiguous(omap_obj));
 		pfn = (omap_obj->paddr >> PAGE_SHIFT) + pgoff;
 	}
 
@@ -560,6 +578,11 @@ fail:
 	case 0:
 	case -ERESTARTSYS:
 	case -EINTR:
+	case -EBUSY:
+		/*
+		 * EBUSY is ok: this just means that another thread
+		 * already did the job.
+		 */
 		return VM_FAULT_NOPAGE;
 	case -ENOMEM:
 		return VM_FAULT_OOM;
@@ -728,7 +751,8 @@ fail:
 static inline bool is_cached_coherent(struct drm_gem_object *obj)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
-	return is_shmem(obj) &&
+
+	return (omap_obj->flags & OMAP_BO_MEM_SHMEM) &&
 		((omap_obj->flags & OMAP_BO_CACHE_MASK) == OMAP_BO_CACHED);
 }
 
@@ -761,9 +785,20 @@ void omap_gem_dma_sync(struct drm_gem_object *obj,
 
 		for (i = 0; i < npages; i++) {
 			if (!omap_obj->addrs[i]) {
-				omap_obj->addrs[i] = dma_map_page(dev->dev, pages[i], 0,
+				dma_addr_t addr;
+
+				addr = dma_map_page(dev->dev, pages[i], 0,
 						PAGE_SIZE, DMA_BIDIRECTIONAL);
+
+				if (dma_mapping_error(dev->dev, addr)) {
+					dev_warn(dev->dev,
+						"%s: failed to map page\n",
+						__func__);
+					break;
+				}
+
 				dirty = true;
+				omap_obj->addrs[i] = addr;
 			}
 		}
 
@@ -787,7 +822,7 @@ int omap_gem_get_paddr(struct drm_gem_object *obj,
 
 	mutex_lock(&obj->dev->struct_mutex);
 
-	if (remap && is_shmem(obj) && priv->has_dmm) {
+	if (!is_contiguous(omap_obj) && remap && priv->has_dmm) {
 		if (omap_obj->paddr_cnt == 0) {
 			struct page **pages;
 			uint32_t npages = obj->size >> PAGE_SHIFT;
@@ -834,7 +869,7 @@ int omap_gem_get_paddr(struct drm_gem_object *obj,
 		omap_obj->paddr_cnt++;
 
 		*paddr = omap_obj->paddr;
-	} else if (omap_obj->flags & OMAP_BO_DMA) {
+	} else if (is_contiguous(omap_obj)) {
 		*paddr = omap_obj->paddr;
 	} else {
 		ret = -EINVAL;
@@ -1138,20 +1173,6 @@ unlock:
 	return ret;
 }
 
-/* it is a bit lame to handle updates in this sort of polling way, but
- * in case of PVR, the GPU can directly update read/write complete
- * values, and not really tell us which ones it updated.. this also
- * means that sync_lock is not quite sufficient.  So we'll need to
- * do something a bit better when it comes time to add support for
- * separate 2d hw..
- */
-void omap_gem_op_update(void)
-{
-	spin_lock(&sync_lock);
-	sync_op_update();
-	spin_unlock(&sync_lock);
-}
-
 /* mark the start of read and/or write operation */
 int omap_gem_op_start(struct drm_gem_object *obj, enum omap_gem_op op)
 {
@@ -1219,7 +1240,7 @@ int omap_gem_op_sync(struct drm_gem_object *obj, enum omap_gem_op op)
  * is currently blocked..  fxn() can be called from any context
  *
  * (TODO for now fxn is called back from whichever context calls
- * omap_gem_op_update().. but this could be better defined later
+ * omap_gem_op_finish().. but this could be better defined later
  * if needed)
  *
  * TODO more code in common w/ _sync()..
@@ -1261,50 +1282,10 @@ int omap_gem_op_async(struct drm_gem_object *obj, enum omap_gem_op op,
 	return 0;
 }
 
-/* special API so PVR can update the buffer to use a sync-object allocated
- * from it's sync-obj heap.  Only used for a newly allocated (from PVR's
- * perspective) sync-object, so we overwrite the new syncobj w/ values
- * from the already allocated syncobj (if there is one)
- */
-int omap_gem_set_sync_object(struct drm_gem_object *obj, void *syncobj)
-{
-	struct omap_gem_object *omap_obj = to_omap_bo(obj);
-	int ret = 0;
-
-	spin_lock(&sync_lock);
-
-	if ((omap_obj->flags & OMAP_BO_EXT_SYNC) && !syncobj) {
-		/* clearing a previously set syncobj */
-		syncobj = kmemdup(omap_obj->sync, sizeof(*omap_obj->sync),
-				  GFP_ATOMIC);
-		if (!syncobj) {
-			ret = -ENOMEM;
-			goto unlock;
-		}
-		omap_obj->flags &= ~OMAP_BO_EXT_SYNC;
-		omap_obj->sync = syncobj;
-	} else if (syncobj && !(omap_obj->flags & OMAP_BO_EXT_SYNC)) {
-		/* replacing an existing syncobj */
-		if (omap_obj->sync) {
-			memcpy(syncobj, omap_obj->sync, sizeof(*omap_obj->sync));
-			kfree(omap_obj->sync);
-		}
-		omap_obj->flags |= OMAP_BO_EXT_SYNC;
-		omap_obj->sync = syncobj;
-	}
-
-unlock:
-	spin_unlock(&sync_lock);
-	return ret;
-}
-
 /* -----------------------------------------------------------------------------
  * Constructor & Destructor
  */
 
-/* don't call directly.. called from GEM core when it is time to actually
- * free the object..
- */
 void omap_gem_free_object(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
@@ -1324,22 +1305,23 @@ void omap_gem_free_object(struct drm_gem_object *obj)
 	 */
 	WARN_ON(omap_obj->paddr_cnt > 0);
 
-	/* don't free externally allocated backing memory */
-	if (!(omap_obj->flags & OMAP_BO_EXT_MEM)) {
-		if (omap_obj->pages)
+	if (omap_obj->pages) {
+		if (omap_obj->flags & OMAP_BO_MEM_DMABUF)
+			kfree(omap_obj->pages);
+		else
 			omap_gem_detach_pages(obj);
+	}
 
-		if (!is_shmem(obj)) {
-			dma_free_writecombine(dev->dev, obj->size,
-					omap_obj->vaddr, omap_obj->paddr);
-		} else if (omap_obj->vaddr) {
-			vunmap(omap_obj->vaddr);
-		}
+	if (omap_obj->flags & OMAP_BO_MEM_DMA_API) {
+		dma_free_wc(dev->dev, obj->size, omap_obj->vaddr,
+			    omap_obj->paddr);
+	} else if (omap_obj->vaddr) {
+		vunmap(omap_obj->vaddr);
+	} else if (obj->import_attach) {
+		drm_prime_gem_destroy(obj, omap_obj->sgt);
 	}
 
-	/* don't free externally allocated syncobj */
-	if (!(omap_obj->flags & OMAP_BO_EXT_SYNC))
-		kfree(omap_obj->sync);
+	kfree(omap_obj->sync);
 
 	drm_gem_object_release(obj);
 
@@ -1357,84 +1339,160 @@ struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 	size_t size;
 	int ret;
 
+	/* Validate the flags and compute the memory and cache flags. */
 	if (flags & OMAP_BO_TILED) {
 		if (!priv->usergart) {
 			dev_err(dev->dev, "Tiled buffers require DMM\n");
 			return NULL;
 		}
 
-		/* tiled buffers are always shmem paged backed.. when they are
-		 * scanned out, they are remapped into DMM/TILER
+		/*
+		 * Tiled buffers are always shmem paged backed. When they are
+		 * scanned out, they are remapped into DMM/TILER.
 		 */
 		flags &= ~OMAP_BO_SCANOUT;
+		flags |= OMAP_BO_MEM_SHMEM;
 
-		/* currently don't allow cached buffers.. there is some caching
-		 * stuff that needs to be handled better
+		/*
+		 * Currently don't allow cached buffers. There is some caching
+		 * stuff that needs to be handled better.
 		 */
 		flags &= ~(OMAP_BO_CACHED|OMAP_BO_WC|OMAP_BO_UNCACHED);
 		flags |= tiler_get_cpu_cache_flags();
-
-		/* align dimensions to slot boundaries... */
-		tiler_align(gem2fmt(flags),
-				&gsize.tiled.width, &gsize.tiled.height);
-
-		/* ...and calculate size based on aligned dimensions */
-		size = tiler_size(gem2fmt(flags),
-				gsize.tiled.width, gsize.tiled.height);
-	} else {
-		size = PAGE_ALIGN(gsize.bytes);
+	} else if ((flags & OMAP_BO_SCANOUT) && !priv->has_dmm) {
+		/*
+		 * OMAP_BO_SCANOUT hints that the buffer doesn't need to be
+		 * tiled. However, to lower the pressure on memory allocation,
+		 * use contiguous memory only if no TILER is available.
+		 */
+		flags |= OMAP_BO_MEM_DMA_API;
+	} else if (!(flags & OMAP_BO_MEM_DMABUF)) {
+		/*
+		 * All other buffers not backed by dma_buf are shmem-backed.
+		 */
+		flags |= OMAP_BO_MEM_SHMEM;
 	}
 
+	/* Allocate the initialize the OMAP GEM object. */
 	omap_obj = kzalloc(sizeof(*omap_obj), GFP_KERNEL);
 	if (!omap_obj)
 		return NULL;
 
 	obj = &omap_obj->base;
+	omap_obj->flags = flags;
 
-	if ((flags & OMAP_BO_SCANOUT) && !priv->has_dmm) {
-		/* attempt to allocate contiguous memory if we don't
-		 * have DMM for remappign discontiguous buffers
+	if (flags & OMAP_BO_TILED) {
+		/*
+		 * For tiled buffers align dimensions to slot boundaries and
+		 * calculate size based on aligned dimensions.
 		 */
-		omap_obj->vaddr =  dma_alloc_writecombine(dev->dev, size,
-				&omap_obj->paddr, GFP_KERNEL);
-		if (!omap_obj->vaddr) {
-			kfree(omap_obj);
+		tiler_align(gem2fmt(flags), &gsize.tiled.width,
+			    &gsize.tiled.height);
 
-			return NULL;
-		}
-
-		flags |= OMAP_BO_DMA;
-	}
+		size = tiler_size(gem2fmt(flags), gsize.tiled.width,
+				  gsize.tiled.height);
 
-	spin_lock(&priv->list_lock);
-	list_add(&omap_obj->mm_list, &priv->obj_list);
-	spin_unlock(&priv->list_lock);
-
-	omap_obj->flags = flags;
-
-	if (flags & OMAP_BO_TILED) {
 		omap_obj->width = gsize.tiled.width;
 		omap_obj->height = gsize.tiled.height;
+	} else {
+		size = PAGE_ALIGN(gsize.bytes);
 	}
 
-	if (flags & (OMAP_BO_DMA|OMAP_BO_EXT_MEM)) {
+	/* Initialize the GEM object. */
+	if (!(flags & OMAP_BO_MEM_SHMEM)) {
 		drm_gem_private_object_init(dev, obj, size);
 	} else {
 		ret = drm_gem_object_init(dev, obj, size);
 		if (ret)
-			goto fail;
+			goto err_free;
 
 		mapping = file_inode(obj->filp)->i_mapping;
 		mapping_set_gfp_mask(mapping, GFP_USER | __GFP_DMA32);
 	}
 
+	/* Allocate memory if needed. */
+	if (flags & OMAP_BO_MEM_DMA_API) {
+		omap_obj->vaddr = dma_alloc_wc(dev->dev, size,
+					       &omap_obj->paddr,
+					       GFP_KERNEL);
+		if (!omap_obj->vaddr)
+			goto err_release;
+	}
+
+	spin_lock(&priv->list_lock);
+	list_add(&omap_obj->mm_list, &priv->obj_list);
+	spin_unlock(&priv->list_lock);
+
 	return obj;
 
-fail:
-	omap_gem_free_object(obj);
+err_release:
+	drm_gem_object_release(obj);
+err_free:
+	kfree(omap_obj);
 	return NULL;
 }
 
+struct drm_gem_object *omap_gem_new_dmabuf(struct drm_device *dev, size_t size,
+					   struct sg_table *sgt)
+{
+	struct omap_drm_private *priv = dev->dev_private;
+	struct omap_gem_object *omap_obj;
+	struct drm_gem_object *obj;
+	union omap_gem_size gsize;
+
+	/* Without a DMM only physically contiguous buffers can be supported. */
+	if (sgt->orig_nents != 1 && !priv->has_dmm)
+		return ERR_PTR(-EINVAL);
+
+	mutex_lock(&dev->struct_mutex);
+
+	gsize.bytes = PAGE_ALIGN(size);
+	obj = omap_gem_new(dev, gsize, OMAP_BO_MEM_DMABUF | OMAP_BO_WC);
+	if (!obj) {
+		obj = ERR_PTR(-ENOMEM);
+		goto done;
+	}
+
+	omap_obj = to_omap_bo(obj);
+	omap_obj->sgt = sgt;
+
+	if (sgt->orig_nents == 1) {
+		omap_obj->paddr = sg_dma_address(sgt->sgl);
+	} else {
+		/* Create pages list from sgt */
+		struct sg_page_iter iter;
+		struct page **pages;
+		unsigned int npages;
+		unsigned int i = 0;
+
+		npages = DIV_ROUND_UP(size, PAGE_SIZE);
+		pages = kcalloc(npages, sizeof(*pages), GFP_KERNEL);
+		if (!pages) {
+			omap_gem_free_object(obj);
+			obj = ERR_PTR(-ENOMEM);
+			goto done;
+		}
+
+		omap_obj->pages = pages;
+
+		for_each_sg_page(sgt->sgl, &iter, sgt->orig_nents, 0) {
+			pages[i++] = sg_page_iter_page(&iter);
+			if (i > npages)
+				break;
+		}
+
+		if (WARN_ON(i != npages)) {
+			omap_gem_free_object(obj);
+			obj = ERR_PTR(-ENOMEM);
+			goto done;
+		}
+	}
+
+done:
+	mutex_unlock(&dev->struct_mutex);
+	return obj;
+}
+
 /* convenience method to construct a GEM buffer object, and userspace handle */
 int omap_gem_new_handle(struct drm_device *dev, struct drm_file *file,
 		union omap_gem_size gsize, uint32_t flags, uint32_t *handle)