Message-ID: <20180530231006.GA2929@downor-Z87X-UD5H>
Date:   Wed, 30 May 2018 16:10:06 -0700
From:   Dongwon Kim <dongwon.kim@...el.com>
To:     Oleksandr Andrushchenko <andr2000@...il.com>
Cc:     xen-devel@...ts.xenproject.org, linux-kernel@...r.kernel.org,
        dri-devel@...ts.freedesktop.org, linux-media@...r.kernel.org,
        jgross@...e.com, boris.ostrovsky@...cle.com,
        konrad.wilk@...cle.com, daniel.vetter@...el.com,
        matthew.d.roper@...el.com,
        Oleksandr Andrushchenko <oleksandr_andrushchenko@...m.com>
Subject: Re: [PATCH 6/8] xen/gntdev: Implement dma-buf export functionality

On Fri, May 25, 2018 at 06:33:29PM +0300, Oleksandr Andrushchenko wrote:
> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@...m.com>
> 
> 1. Create a dma-buf from grant references provided by the foreign
>    domain. By default dma-buf is backed by system memory pages, but
>    by providing GNTDEV_DMA_FLAG_XXX flags it can also be created
>    as a DMA write-combine/coherent buffer, e.g. allocated with
>    corresponding dma_alloc_xxx API.
>    Export the resulting buffer as a new dma-buf.
> 
> 2. Implement waiting for the dma-buf to be released: block until the
>    dma-buf with the file descriptor provided is released.
>    If within the time-out provided the buffer is not released then
>    -ETIMEDOUT error is returned. If the buffer with the file descriptor
>    does not exist or has already been released, then -ENOENT is returned.
>    For valid file descriptors this must not be treated as error.
> 
> Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@...m.com>
> ---
>  drivers/xen/gntdev.c | 478 ++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 476 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/xen/gntdev.c b/drivers/xen/gntdev.c
> index 9e450622af1a..52abc6cd5846 100644
> --- a/drivers/xen/gntdev.c
> +++ b/drivers/xen/gntdev.c
> @@ -4,6 +4,8 @@
>   * Device for accessing (in user-space) pages that have been granted by other
>   * domains.
>   *
> + * DMA buffer implementation is based on drivers/gpu/drm/drm_prime.c.
> + *
>   * Copyright (c) 2006-2007, D G Murray.
>   *           (c) 2009 Gerd Hoffmann <kraxel@...hat.com>
>   *           (c) 2018 Oleksandr Andrushchenko, EPAM Systems Inc.
> @@ -41,6 +43,9 @@
>  #ifdef CONFIG_XEN_GRANT_DMA_ALLOC
>  #include <linux/of_device.h>
>  #endif
> +#ifdef CONFIG_XEN_GNTDEV_DMABUF
> +#include <linux/dma-buf.h>
> +#endif
>  
>  #include <xen/xen.h>
>  #include <xen/grant_table.h>
> @@ -81,6 +86,17 @@ struct gntdev_priv {
>  	/* Device for which DMA memory is allocated. */
>  	struct device *dma_dev;
>  #endif
> +
> +#ifdef CONFIG_XEN_GNTDEV_DMABUF
> +	/* Private data of the hyper DMA buffers. */
> +
> +	/* List of exported DMA buffers. */
> +	struct list_head dmabuf_exp_list;
> +	/* List of wait objects. */
> +	struct list_head dmabuf_exp_wait_list;
> +	/* This is the lock which protects dma_buf_xxx lists. */
> +	struct mutex dmabuf_lock;
> +#endif
>  };
>  
>  struct unmap_notify {
> @@ -125,12 +141,38 @@ struct grant_map {
>  
>  #ifdef CONFIG_XEN_GNTDEV_DMABUF
>  struct xen_dmabuf {
> +	struct gntdev_priv *priv;
> +	struct dma_buf *dmabuf;
> +	struct list_head next;
> +	int fd;
> +
>  	union {
> +		struct {
> +			/* Exported buffers are reference counted. */
> +			struct kref refcount;
> +			struct grant_map *map;
> +		} exp;
>  		struct {
>  			/* Granted references of the imported buffer. */
>  			grant_ref_t *refs;
>  		} imp;
>  	} u;
> +
> +	/* Number of pages this buffer has. */
> +	int nr_pages;
> +	/* Pages of this buffer. */
> +	struct page **pages;
> +};
> +
> +struct xen_dmabuf_wait_obj {
> +	struct list_head next;
> +	struct xen_dmabuf *xen_dmabuf;
> +	struct completion completion;
> +};
> +
> +struct xen_dmabuf_attachment {
> +	struct sg_table *sgt;
> +	enum dma_data_direction dir;
>  };
>  #endif
>  
> @@ -320,6 +362,16 @@ static void gntdev_put_map(struct gntdev_priv *priv, struct grant_map *map)
>  	gntdev_free_map(map);
>  }
>  
> +#ifdef CONFIG_XEN_GNTDEV_DMABUF
> +static void gntdev_remove_map(struct gntdev_priv *priv, struct grant_map *map)
> +{
> +	mutex_lock(&priv->lock);
> +	list_del(&map->next);
> +	gntdev_put_map(NULL /* already removed */, map);
> +	mutex_unlock(&priv->lock);
> +}
> +#endif
> +
>  /* ------------------------------------------------------------------ */
>  
>  static int find_grant_ptes(pte_t *pte, pgtable_t token,
> @@ -628,6 +680,12 @@ static int gntdev_open(struct inode *inode, struct file *flip)
>  	INIT_LIST_HEAD(&priv->freeable_maps);
>  	mutex_init(&priv->lock);
>  
> +#ifdef CONFIG_XEN_GNTDEV_DMABUF
> +	mutex_init(&priv->dmabuf_lock);
> +	INIT_LIST_HEAD(&priv->dmabuf_exp_list);
> +	INIT_LIST_HEAD(&priv->dmabuf_exp_wait_list);
> +#endif
> +
>  	if (use_ptemod) {
>  		priv->mm = get_task_mm(current);
>  		if (!priv->mm) {
> @@ -1053,17 +1111,433 @@ static long gntdev_ioctl_grant_copy(struct gntdev_priv *priv, void __user *u)
>  /* DMA buffer export support.                                         */
>  /* ------------------------------------------------------------------ */
>  
> +/* ------------------------------------------------------------------ */
> +/* Implementation of wait for exported DMA buffer to be released.     */
> +/* ------------------------------------------------------------------ */
> +
> +static void dmabuf_exp_release(struct kref *kref);
> +
> +static struct xen_dmabuf_wait_obj *
> +dmabuf_exp_wait_obj_new(struct gntdev_priv *priv,
> +			struct xen_dmabuf *xen_dmabuf)
> +{
> +	struct xen_dmabuf_wait_obj *obj;
> +
> +	obj = kzalloc(sizeof(*obj), GFP_KERNEL);
> +	if (!obj)
> +		return ERR_PTR(-ENOMEM);
> +
> +	init_completion(&obj->completion);
> +	obj->xen_dmabuf = xen_dmabuf;
> +
> +	mutex_lock(&priv->dmabuf_lock);
> +	list_add(&obj->next, &priv->dmabuf_exp_wait_list);
> +	/* Put our reference and wait for xen_dmabuf's release to fire. */
> +	kref_put(&xen_dmabuf->u.exp.refcount, dmabuf_exp_release);
> +	mutex_unlock(&priv->dmabuf_lock);
> +	return obj;
> +}
> +
> +static void dmabuf_exp_wait_obj_free(struct gntdev_priv *priv,
> +				     struct xen_dmabuf_wait_obj *obj)
> +{
> +	struct xen_dmabuf_wait_obj *cur_obj, *q;
> +
> +	mutex_lock(&priv->dmabuf_lock);
> +	list_for_each_entry_safe(cur_obj, q, &priv->dmabuf_exp_wait_list, next)
> +		if (cur_obj == obj) {
> +			list_del(&obj->next);
> +			kfree(obj);
> +			break;
> +		}
> +	mutex_unlock(&priv->dmabuf_lock);
> +}
> +
> +static int dmabuf_exp_wait_obj_wait(struct xen_dmabuf_wait_obj *obj,
> +				    u32 wait_to_ms)
> +{
> +	if (wait_for_completion_timeout(&obj->completion,
> +			msecs_to_jiffies(wait_to_ms)) <= 0)
> +		return -ETIMEDOUT;
> +
> +	return 0;
> +}
> +
> +static void dmabuf_exp_wait_obj_signal(struct gntdev_priv *priv,
> +				       struct xen_dmabuf *xen_dmabuf)
> +{
> +	struct xen_dmabuf_wait_obj *obj, *q;
> +
> +	list_for_each_entry_safe(obj, q, &priv->dmabuf_exp_wait_list, next)
> +		if (obj->xen_dmabuf == xen_dmabuf) {
> +			pr_debug("Found xen_dmabuf in the wait list, wake\n");
> +			complete_all(&obj->completion);
> +		}
> +}
> +
> +static struct xen_dmabuf *
> +dmabuf_exp_wait_obj_get_by_fd(struct gntdev_priv *priv, int fd)
> +{
> +	struct xen_dmabuf *q, *xen_dmabuf, *ret = ERR_PTR(-ENOENT);
> +
> +	mutex_lock(&priv->dmabuf_lock);
> +	list_for_each_entry_safe(xen_dmabuf, q, &priv->dmabuf_exp_list, next)
> +		if (xen_dmabuf->fd == fd) {
> +			pr_debug("Found xen_dmabuf in the wait list\n");
> +			kref_get(&xen_dmabuf->u.exp.refcount);
> +			ret = xen_dmabuf;
> +			break;
> +		}
> +	mutex_unlock(&priv->dmabuf_lock);
> +	return ret;
> +}
> +
>  static int dmabuf_exp_wait_released(struct gntdev_priv *priv, int fd,
>  				    int wait_to_ms)
>  {
> -	return -ETIMEDOUT;
> +	struct xen_dmabuf *xen_dmabuf;
> +	struct xen_dmabuf_wait_obj *obj;
> +	int ret;
> +
> +	pr_debug("Will wait for dma-buf with fd %d\n", fd);
> +	/*
> +	 * Try to find the DMA buffer: if not found means that
> +	 * either the buffer has already been released or file descriptor
> +	 * provided is wrong.
> +	 */
> +	xen_dmabuf = dmabuf_exp_wait_obj_get_by_fd(priv, fd);
> +	if (IS_ERR(xen_dmabuf))
> +		return PTR_ERR(xen_dmabuf);
> +
> +	/*
> +	 * xen_dmabuf still exists and is reference count locked by us now,
> +	 * so prepare to wait: allocate wait object and add it to the wait list,
> +	 * so we can find it on release.
> +	 */
> +	obj = dmabuf_exp_wait_obj_new(priv, xen_dmabuf);
> +	if (IS_ERR(obj)) {
> +		pr_err("Failed to setup wait object, ret %ld\n", PTR_ERR(obj));
> +		return PTR_ERR(obj);
> +	}
> +
> +	ret = dmabuf_exp_wait_obj_wait(obj, wait_to_ms);
> +	dmabuf_exp_wait_obj_free(priv, obj);
> +	return ret;
> +}
> +
> +/* ------------------------------------------------------------------ */
> +/* DMA buffer export support.                                         */
> +/* ------------------------------------------------------------------ */
> +
> +static struct sg_table *
> +dmabuf_pages_to_sgt(struct page **pages, unsigned int nr_pages)
> +{
> +	struct sg_table *sgt;
> +	int ret;
> +
> +	sgt = kmalloc(sizeof(*sgt), GFP_KERNEL);
> +	if (!sgt) {
> +		ret = -ENOMEM;
> +		goto out;
> +	}
> +
> +	ret = sg_alloc_table_from_pages(sgt, pages, nr_pages, 0,
> +					nr_pages << PAGE_SHIFT,
> +					GFP_KERNEL);
> +	if (ret)
> +		goto out;
> +
> +	return sgt;
> +
> +out:
> +	kfree(sgt);
> +	return ERR_PTR(ret);
> +}
> +
> +static int dmabuf_exp_ops_attach(struct dma_buf *dma_buf,
> +				 struct device *target_dev,
> +				 struct dma_buf_attachment *attach)
> +{
> +	struct xen_dmabuf_attachment *xen_dmabuf_attach;
> +
> +	xen_dmabuf_attach = kzalloc(sizeof(*xen_dmabuf_attach), GFP_KERNEL);
> +	if (!xen_dmabuf_attach)
> +		return -ENOMEM;
> +
> +	xen_dmabuf_attach->dir = DMA_NONE;
> +	attach->priv = xen_dmabuf_attach;
> +	/* Might need to pin the pages of the buffer now. */
> +	return 0;
> +}
> +
> +static void dmabuf_exp_ops_detach(struct dma_buf *dma_buf,
> +				  struct dma_buf_attachment *attach)
> +{
> +	struct xen_dmabuf_attachment *xen_dmabuf_attach = attach->priv;
> +
> +	if (xen_dmabuf_attach) {
> +		struct sg_table *sgt = xen_dmabuf_attach->sgt;
> +
> +		if (sgt) {
> +			if (xen_dmabuf_attach->dir != DMA_NONE)
> +				dma_unmap_sg_attrs(attach->dev, sgt->sgl,
> +						   sgt->nents,
> +						   xen_dmabuf_attach->dir,
> +						   DMA_ATTR_SKIP_CPU_SYNC);
> +			sg_free_table(sgt);
> +		}
> +
> +		kfree(sgt);
> +		kfree(xen_dmabuf_attach);
> +		attach->priv = NULL;
> +	}
> +	/* Might need to unpin the pages of the buffer now. */
> +}
> +
> +static struct sg_table *
> +dmabuf_exp_ops_map_dma_buf(struct dma_buf_attachment *attach,
> +			   enum dma_data_direction dir)
> +{
> +	struct xen_dmabuf_attachment *xen_dmabuf_attach = attach->priv;
> +	struct xen_dmabuf *xen_dmabuf = attach->dmabuf->priv;
> +	struct sg_table *sgt;
> +
> +	pr_debug("Mapping %d pages for dev %p\n", xen_dmabuf->nr_pages,
> +		 attach->dev);
> +
> +	if (WARN_ON(dir == DMA_NONE || !xen_dmabuf_attach))
> +		return ERR_PTR(-EINVAL);
> +
> +	/* Return the cached mapping when possible. */
> +	if (xen_dmabuf_attach->dir == dir)
> +		return xen_dmabuf_attach->sgt;

may need to check xen_dmabuf_attach->sgt == NULL (i.e. first time mapping)?
Also, I am not sure if this mechanism of reusing previously generated sgt
for other mappings is universally ok for any use-cases... I don't know if
it is acceptable as per the specification. 

> +
> +	/*
> +	 * Two mappings with different directions for the same attachment are
> +	 * not allowed.
> +	 */
> +	if (WARN_ON(xen_dmabuf_attach->dir != DMA_NONE))
> +		return ERR_PTR(-EBUSY);
> +
> +	sgt = dmabuf_pages_to_sgt(xen_dmabuf->pages, xen_dmabuf->nr_pages);
> +	if (!IS_ERR(sgt)) {
> +		if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
> +				      DMA_ATTR_SKIP_CPU_SYNC)) {
> +			sg_free_table(sgt);
> +			kfree(sgt);
> +			sgt = ERR_PTR(-ENOMEM);
> +		} else {
> +			xen_dmabuf_attach->sgt = sgt;
> +			xen_dmabuf_attach->dir = dir;
> +		}
> +	}
> +	if (IS_ERR(sgt))
> +		pr_err("Failed to map sg table for dev %p\n", attach->dev);
> +	return sgt;
> +}
> +
> +static void dmabuf_exp_ops_unmap_dma_buf(struct dma_buf_attachment *attach,
> +					 struct sg_table *sgt,
> +					 enum dma_data_direction dir)
> +{
> +	/* Not implemented. The unmap is done at dmabuf_exp_ops_detach(). */

Not sure if it's ok to do nothing here because the spec says this function is
mandatory and it should unmap and "release" &sg_table associated with it. 

	/**
	 * @unmap_dma_buf:
	 *
	 * This is called by dma_buf_unmap_attachment() and should unmap and
	 * release the &sg_table allocated in @map_dma_buf, and it is mandatory.
	 * It should also unpin the backing storage if this is the last mapping
	 * of the DMA buffer, it the exporter supports backing storage
	 * migration.
	 */

> +}
> +
> +static void dmabuf_exp_release(struct kref *kref)
> +{
> +	struct xen_dmabuf *xen_dmabuf =
> +		container_of(kref, struct xen_dmabuf, u.exp.refcount);
> +
> +	dmabuf_exp_wait_obj_signal(xen_dmabuf->priv, xen_dmabuf);
> +	list_del(&xen_dmabuf->next);
> +	kfree(xen_dmabuf);
> +}
> +
> +static void dmabuf_exp_ops_release(struct dma_buf *dma_buf)
> +{
> +	struct xen_dmabuf *xen_dmabuf = dma_buf->priv;
> +	struct gntdev_priv *priv = xen_dmabuf->priv;
> +
> +	gntdev_remove_map(priv, xen_dmabuf->u.exp.map);
> +	mutex_lock(&priv->dmabuf_lock);
> +	kref_put(&xen_dmabuf->u.exp.refcount, dmabuf_exp_release);
> +	mutex_unlock(&priv->dmabuf_lock);
> +}
> +
> +static void *dmabuf_exp_ops_kmap_atomic(struct dma_buf *dma_buf,
> +					unsigned long page_num)
> +{
> +	/* Not implemented. */
> +	return NULL;
> +}
> +
> +static void dmabuf_exp_ops_kunmap_atomic(struct dma_buf *dma_buf,
> +					 unsigned long page_num, void *addr)
> +{
> +	/* Not implemented. */
> +}
> +
> +static void *dmabuf_exp_ops_kmap(struct dma_buf *dma_buf,
> +				 unsigned long page_num)
> +{
> +	/* Not implemented. */
> +	return NULL;
> +}
> +
> +static void dmabuf_exp_ops_kunmap(struct dma_buf *dma_buf,
> +				  unsigned long page_num, void *addr)
> +{
> +	/* Not implemented. */
> +}
> +
> +static int dmabuf_exp_ops_mmap(struct dma_buf *dma_buf,
> +			       struct vm_area_struct *vma)
> +{
> +	/* Not implemented. */
> +	return 0;
> +}
> +
> +static const struct dma_buf_ops dmabuf_exp_ops =  {
> +	.attach = dmabuf_exp_ops_attach,
> +	.detach = dmabuf_exp_ops_detach,
> +	.map_dma_buf = dmabuf_exp_ops_map_dma_buf,
> +	.unmap_dma_buf = dmabuf_exp_ops_unmap_dma_buf,
> +	.release = dmabuf_exp_ops_release,
> +	.map = dmabuf_exp_ops_kmap,
> +	.map_atomic = dmabuf_exp_ops_kmap_atomic,
> +	.unmap = dmabuf_exp_ops_kunmap,
> +	.unmap_atomic = dmabuf_exp_ops_kunmap_atomic,
> +	.mmap = dmabuf_exp_ops_mmap,
> +};
> +
> +static int dmabuf_export(struct gntdev_priv *priv, struct grant_map *map,
> +			 int *fd)
> +{
> +	DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
> +	struct xen_dmabuf *xen_dmabuf;
> +	int ret = 0;
> +
> +	xen_dmabuf = kzalloc(sizeof(*xen_dmabuf), GFP_KERNEL);
> +	if (!xen_dmabuf)
> +		return -ENOMEM;
> +
> +	kref_init(&xen_dmabuf->u.exp.refcount);
> +
> +	xen_dmabuf->priv = priv;
> +	xen_dmabuf->nr_pages = map->count;
> +	xen_dmabuf->pages = map->pages;
> +	xen_dmabuf->u.exp.map = map;
> +
> +	exp_info.exp_name = KBUILD_MODNAME;
> +	if (map->dma_dev->driver && map->dma_dev->driver->owner)
> +		exp_info.owner = map->dma_dev->driver->owner;
> +	else
> +		exp_info.owner = THIS_MODULE;
> +	exp_info.ops = &dmabuf_exp_ops;
> +	exp_info.size = map->count << PAGE_SHIFT;
> +	exp_info.flags = O_RDWR;
> +	exp_info.priv = xen_dmabuf;
> +
> +	xen_dmabuf->dmabuf = dma_buf_export(&exp_info);
> +	if (IS_ERR(xen_dmabuf->dmabuf)) {
> +		ret = PTR_ERR(xen_dmabuf->dmabuf);
> +		xen_dmabuf->dmabuf = NULL;
> +		goto fail;
> +	}
> +
> +	ret = dma_buf_fd(xen_dmabuf->dmabuf, O_CLOEXEC);
> +	if (ret < 0)
> +		goto fail;
> +
> +	xen_dmabuf->fd = ret;
> +	*fd = ret;
> +
> +	pr_debug("Exporting DMA buffer with fd %d\n", ret);
> +
> +	mutex_lock(&priv->dmabuf_lock);
> +	list_add(&xen_dmabuf->next, &priv->dmabuf_exp_list);
> +	mutex_unlock(&priv->dmabuf_lock);
> +	return 0;
> +
> +fail:
> +	if (xen_dmabuf->dmabuf)
> +		dma_buf_put(xen_dmabuf->dmabuf);
> +	kfree(xen_dmabuf);
> +	return ret;
> +}
> +
> +static struct grant_map *
> +dmabuf_exp_alloc_backing_storage(struct gntdev_priv *priv, int dmabuf_flags,
> +				 int count)
> +{
> +	struct grant_map *map;
> +
> +	if (unlikely(count <= 0))
> +		return ERR_PTR(-EINVAL);
> +
> +	if ((dmabuf_flags & GNTDEV_DMA_FLAG_WC) &&
> +	    (dmabuf_flags & GNTDEV_DMA_FLAG_COHERENT)) {
> +		pr_err("Wrong dma-buf flags: either WC or coherent, not both\n");
> +		return ERR_PTR(-EINVAL);
> +	}
> +
> +	map = gntdev_alloc_map(priv, count, dmabuf_flags);
> +	if (!map)
> +		return ERR_PTR(-ENOMEM);
> +
> +	if (unlikely(atomic_add_return(count, &pages_mapped) > limit)) {
> +		pr_err("can't map: over limit\n");
> +		gntdev_put_map(NULL, map);
> +		return ERR_PTR(-ENOMEM);
> +	}
> +	return map;
>  }

When and how would this allocation be freed? I don't see any ioctl for freeing up
shared pages.

>  
>  static int dmabuf_exp_from_refs(struct gntdev_priv *priv, int flags,
>  				int count, u32 domid, u32 *refs, u32 *fd)
>  {
> +	struct grant_map *map;
> +	int i, ret;
> +
>  	*fd = -1;
> -	return -EINVAL;
> +
> +	if (use_ptemod) {
> +		pr_err("Cannot provide dma-buf: use_ptemode %d\n",
> +		       use_ptemod);
> +		return -EINVAL;
> +	}
> +
> +	map = dmabuf_exp_alloc_backing_storage(priv, flags, count);
> +	if (IS_ERR(map))
> +		return PTR_ERR(map);
> +
> +	for (i = 0; i < count; i++) {
> +		map->grants[i].domid = domid;
> +		map->grants[i].ref = refs[i];
> +	}
> +
> +	mutex_lock(&priv->lock);
> +	gntdev_add_map(priv, map);
> +	mutex_unlock(&priv->lock);
> +
> +	map->flags |= GNTMAP_host_map;
> +#if defined(CONFIG_X86)
> +	map->flags |= GNTMAP_device_map;
> +#endif
> +
> +	ret = map_grant_pages(map);
> +	if (ret < 0)
> +		goto out;
> +
> +	ret = dmabuf_export(priv, map, fd);
> +	if (ret < 0)
> +		goto out;
> +
> +	return 0;
> +
> +out:
> +	gntdev_remove_map(priv, map);
> +	return ret;
>  }
>  
>  /* ------------------------------------------------------------------ */
> -- 
> 2.17.0
> 

Powered by blists - more mailing lists