diff mbox series

[v5,1/6] dma-buf: Add dma_buf_{begin,end}_access()

Message ID 20240119141402.44262-2-paul@crapouillou.net
State New
Headers show
Series usb: gadget: functionfs: DMABUF import interface | expand

Commit Message

Paul Cercueil Jan. 19, 2024, 2:13 p.m. UTC
These functions should be used by device drivers when they start and
stop accessing the data of DMABUF. It allows DMABUF importers to cache
the dma_buf_attachment while ensuring that the data they want to access
is available for their device when the DMA transfers take place.

Signed-off-by: Paul Cercueil <paul@crapouillou.net>

---
v5: New patch
---
 drivers/dma-buf/dma-buf.c | 66 +++++++++++++++++++++++++++++++++++++++
 include/linux/dma-buf.h   | 37 ++++++++++++++++++++++
 2 files changed, 103 insertions(+)

Comments

Christian König Jan. 22, 2024, 10:35 a.m. UTC | #1
Am 19.01.24 um 15:13 schrieb Paul Cercueil:
> These functions should be used by device drivers when they start and
> stop accessing the data of DMABUF. It allows DMABUF importers to cache
> the dma_buf_attachment while ensuring that the data they want to access
> is available for their device when the DMA transfers take place.

As Daniel already noted as well this is a complete no-go from the 
DMA-buf design point of view.

Regards,
Christian.

>
> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
>
> ---
> v5: New patch
> ---
>   drivers/dma-buf/dma-buf.c | 66 +++++++++++++++++++++++++++++++++++++++
>   include/linux/dma-buf.h   | 37 ++++++++++++++++++++++
>   2 files changed, 103 insertions(+)
>
> diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
> index 8fe5aa67b167..a8bab6c18fcd 100644
> --- a/drivers/dma-buf/dma-buf.c
> +++ b/drivers/dma-buf/dma-buf.c
> @@ -830,6 +830,8 @@ static struct sg_table * __map_dma_buf(struct dma_buf_attachment *attach,
>    *     - dma_buf_mmap()
>    *     - dma_buf_begin_cpu_access()
>    *     - dma_buf_end_cpu_access()
> + *     - dma_buf_begin_access()
> + *     - dma_buf_end_access()
>    *     - dma_buf_map_attachment_unlocked()
>    *     - dma_buf_unmap_attachment_unlocked()
>    *     - dma_buf_vmap_unlocked()
> @@ -1602,6 +1604,70 @@ void dma_buf_vunmap_unlocked(struct dma_buf *dmabuf, struct iosys_map *map)
>   }
>   EXPORT_SYMBOL_NS_GPL(dma_buf_vunmap_unlocked, DMA_BUF);
>   
> +/**
> + * @dma_buf_begin_access - Call before any hardware access from/to the DMABUF
> + * @attach:	[in]	attachment used for hardware access
> + * @sg_table:	[in]	scatterlist used for the DMA transfer
> + * @direction:  [in]    direction of DMA transfer
> + */
> +int dma_buf_begin_access(struct dma_buf_attachment *attach,
> +			 struct sg_table *sgt, enum dma_data_direction dir)
> +{
> +	struct dma_buf *dmabuf;
> +	bool cookie;
> +	int ret;
> +
> +	if (WARN_ON(!attach))
> +		return -EINVAL;
> +
> +	dmabuf = attach->dmabuf;
> +
> +	if (!dmabuf->ops->begin_access)
> +		return 0;
> +
> +	cookie = dma_fence_begin_signalling();
> +	ret = dmabuf->ops->begin_access(attach, sgt, dir);
> +	dma_fence_end_signalling(cookie);
> +
> +	if (WARN_ON_ONCE(ret))
> +		return ret;
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL_NS_GPL(dma_buf_begin_access, DMA_BUF);
> +
> +/**
> + * @dma_buf_end_access - Call after any hardware access from/to the DMABUF
> + * @attach:	[in]	attachment used for hardware access
> + * @sg_table:	[in]	scatterlist used for the DMA transfer
> + * @direction:  [in]    direction of DMA transfer
> + */
> +int dma_buf_end_access(struct dma_buf_attachment *attach,
> +		       struct sg_table *sgt, enum dma_data_direction dir)
> +{
> +	struct dma_buf *dmabuf;
> +	bool cookie;
> +	int ret;
> +
> +	if (WARN_ON(!attach))
> +		return -EINVAL;
> +
> +	dmabuf = attach->dmabuf;
> +
> +	if (!dmabuf->ops->end_access)
> +		return 0;
> +
> +	cookie = dma_fence_begin_signalling();
> +	ret = dmabuf->ops->end_access(attach, sgt, dir);
> +	dma_fence_end_signalling(cookie);
> +
> +	if (WARN_ON_ONCE(ret))
> +		return ret;
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL_NS_GPL(dma_buf_end_access, DMA_BUF);
> +
>   #ifdef CONFIG_DEBUG_FS
>   static int dma_buf_debug_show(struct seq_file *s, void *unused)
>   {
> diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h
> index 8ff4add71f88..8ba612c7cc16 100644
> --- a/include/linux/dma-buf.h
> +++ b/include/linux/dma-buf.h
> @@ -246,6 +246,38 @@ struct dma_buf_ops {
>   	 */
>   	int (*end_cpu_access)(struct dma_buf *, enum dma_data_direction);
>   
> +	/**
> +	 * @begin_access:
> +	 *
> +	 * This is called from dma_buf_begin_access() when a device driver
> +	 * wants to access the data of the DMABUF. The exporter can use this
> +	 * to flush/sync the caches if needed.
> +	 *
> +	 * This callback is optional.
> +	 *
> +	 * Returns:
> +	 *
> +	 * 0 on success or a negative error code on failure.
> +	 */
> +	int (*begin_access)(struct dma_buf_attachment *, struct sg_table *,
> +			    enum dma_data_direction);
> +
> +	/**
> +	 * @end_access:
> +	 *
> +	 * This is called from dma_buf_end_access() when a device driver is
> +	 * done accessing the data of the DMABUF. The exporter can use this
> +	 * to flush/sync the caches if needed.
> +	 *
> +	 * This callback is optional.
> +	 *
> +	 * Returns:
> +	 *
> +	 * 0 on success or a negative error code on failure.
> +	 */
> +	int (*end_access)(struct dma_buf_attachment *, struct sg_table *,
> +			  enum dma_data_direction);
> +
>   	/**
>   	 * @mmap:
>   	 *
> @@ -606,6 +638,11 @@ void dma_buf_detach(struct dma_buf *dmabuf,
>   int dma_buf_pin(struct dma_buf_attachment *attach);
>   void dma_buf_unpin(struct dma_buf_attachment *attach);
>   
> +int dma_buf_begin_access(struct dma_buf_attachment *attach,
> +			 struct sg_table *sgt, enum dma_data_direction dir);
> +int dma_buf_end_access(struct dma_buf_attachment *attach,
> +		       struct sg_table *sgt, enum dma_data_direction dir);
> +
>   struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
>   
>   int dma_buf_fd(struct dma_buf *dmabuf, int flags);
Paul Cercueil Jan. 22, 2024, 11:01 a.m. UTC | #2
Hi Christian,

Le lundi 22 janvier 2024 à 11:35 +0100, Christian König a écrit :
> Am 19.01.24 um 15:13 schrieb Paul Cercueil:
> > These functions should be used by device drivers when they start
> > and
> > stop accessing the data of DMABUF. It allows DMABUF importers to
> > cache
> > the dma_buf_attachment while ensuring that the data they want to
> > access
> > is available for their device when the DMA transfers take place.
> 
> As Daniel already noted as well this is a complete no-go from the 
> DMA-buf design point of view.

What do you mean "as Daniel already noted"? It was him who suggested
this.

> 
> Regards,
> Christian.

Cheers,
-Paul

> 
> > 
> > Signed-off-by: Paul Cercueil <paul@crapouillou.net>
> > 
> > ---
> > v5: New patch
> > ---
> >   drivers/dma-buf/dma-buf.c | 66
> > +++++++++++++++++++++++++++++++++++++++
> >   include/linux/dma-buf.h   | 37 ++++++++++++++++++++++
> >   2 files changed, 103 insertions(+)
> > 
> > diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
> > index 8fe5aa67b167..a8bab6c18fcd 100644
> > --- a/drivers/dma-buf/dma-buf.c
> > +++ b/drivers/dma-buf/dma-buf.c
> > @@ -830,6 +830,8 @@ static struct sg_table * __map_dma_buf(struct
> > dma_buf_attachment *attach,
> >    *     - dma_buf_mmap()
> >    *     - dma_buf_begin_cpu_access()
> >    *     - dma_buf_end_cpu_access()
> > + *     - dma_buf_begin_access()
> > + *     - dma_buf_end_access()
> >    *     - dma_buf_map_attachment_unlocked()
> >    *     - dma_buf_unmap_attachment_unlocked()
> >    *     - dma_buf_vmap_unlocked()
> > @@ -1602,6 +1604,70 @@ void dma_buf_vunmap_unlocked(struct dma_buf
> > *dmabuf, struct iosys_map *map)
> >   }
> >   EXPORT_SYMBOL_NS_GPL(dma_buf_vunmap_unlocked, DMA_BUF);
> >   
> > +/**
> > + * @dma_buf_begin_access - Call before any hardware access from/to
> > the DMABUF
> > + * @attach:	[in]	attachment used for hardware access
> > + * @sg_table:	[in]	scatterlist used for the DMA transfer
> > + * @direction:  [in]    direction of DMA transfer
> > + */
> > +int dma_buf_begin_access(struct dma_buf_attachment *attach,
> > +			 struct sg_table *sgt, enum
> > dma_data_direction dir)
> > +{
> > +	struct dma_buf *dmabuf;
> > +	bool cookie;
> > +	int ret;
> > +
> > +	if (WARN_ON(!attach))
> > +		return -EINVAL;
> > +
> > +	dmabuf = attach->dmabuf;
> > +
> > +	if (!dmabuf->ops->begin_access)
> > +		return 0;
> > +
> > +	cookie = dma_fence_begin_signalling();
> > +	ret = dmabuf->ops->begin_access(attach, sgt, dir);
> > +	dma_fence_end_signalling(cookie);
> > +
> > +	if (WARN_ON_ONCE(ret))
> > +		return ret;
> > +
> > +	return 0;
> > +}
> > +EXPORT_SYMBOL_NS_GPL(dma_buf_begin_access, DMA_BUF);
> > +
> > +/**
> > + * @dma_buf_end_access - Call after any hardware access from/to
> > the DMABUF
> > + * @attach:	[in]	attachment used for hardware access
> > + * @sg_table:	[in]	scatterlist used for the DMA transfer
> > + * @direction:  [in]    direction of DMA transfer
> > + */
> > +int dma_buf_end_access(struct dma_buf_attachment *attach,
> > +		       struct sg_table *sgt, enum
> > dma_data_direction dir)
> > +{
> > +	struct dma_buf *dmabuf;
> > +	bool cookie;
> > +	int ret;
> > +
> > +	if (WARN_ON(!attach))
> > +		return -EINVAL;
> > +
> > +	dmabuf = attach->dmabuf;
> > +
> > +	if (!dmabuf->ops->end_access)
> > +		return 0;
> > +
> > +	cookie = dma_fence_begin_signalling();
> > +	ret = dmabuf->ops->end_access(attach, sgt, dir);
> > +	dma_fence_end_signalling(cookie);
> > +
> > +	if (WARN_ON_ONCE(ret))
> > +		return ret;
> > +
> > +	return 0;
> > +}
> > +EXPORT_SYMBOL_NS_GPL(dma_buf_end_access, DMA_BUF);
> > +
> >   #ifdef CONFIG_DEBUG_FS
> >   static int dma_buf_debug_show(struct seq_file *s, void *unused)
> >   {
> > diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h
> > index 8ff4add71f88..8ba612c7cc16 100644
> > --- a/include/linux/dma-buf.h
> > +++ b/include/linux/dma-buf.h
> > @@ -246,6 +246,38 @@ struct dma_buf_ops {
> >   	 */
> >   	int (*end_cpu_access)(struct dma_buf *, enum
> > dma_data_direction);
> >   
> > +	/**
> > +	 * @begin_access:
> > +	 *
> > +	 * This is called from dma_buf_begin_access() when a
> > device driver
> > +	 * wants to access the data of the DMABUF. The exporter
> > can use this
> > +	 * to flush/sync the caches if needed.
> > +	 *
> > +	 * This callback is optional.
> > +	 *
> > +	 * Returns:
> > +	 *
> > +	 * 0 on success or a negative error code on failure.
> > +	 */
> > +	int (*begin_access)(struct dma_buf_attachment *, struct
> > sg_table *,
> > +			    enum dma_data_direction);
> > +
> > +	/**
> > +	 * @end_access:
> > +	 *
> > +	 * This is called from dma_buf_end_access() when a device
> > driver is
> > +	 * done accessing the data of the DMABUF. The exporter can
> > use this
> > +	 * to flush/sync the caches if needed.
> > +	 *
> > +	 * This callback is optional.
> > +	 *
> > +	 * Returns:
> > +	 *
> > +	 * 0 on success or a negative error code on failure.
> > +	 */
> > +	int (*end_access)(struct dma_buf_attachment *, struct
> > sg_table *,
> > +			  enum dma_data_direction);
> > +
> >   	/**
> >   	 * @mmap:
> >   	 *
> > @@ -606,6 +638,11 @@ void dma_buf_detach(struct dma_buf *dmabuf,
> >   int dma_buf_pin(struct dma_buf_attachment *attach);
> >   void dma_buf_unpin(struct dma_buf_attachment *attach);
> >   
> > +int dma_buf_begin_access(struct dma_buf_attachment *attach,
> > +			 struct sg_table *sgt, enum
> > dma_data_direction dir);
> > +int dma_buf_end_access(struct dma_buf_attachment *attach,
> > +		       struct sg_table *sgt, enum
> > dma_data_direction dir);
> > +
> >   struct dma_buf *dma_buf_export(const struct dma_buf_export_info
> > *exp_info);
> >   
> >   int dma_buf_fd(struct dma_buf *dmabuf, int flags);
>
Christian König Jan. 22, 2024, 1:41 p.m. UTC | #3
Am 22.01.24 um 12:01 schrieb Paul Cercueil:
> Hi Christian,
>
> Le lundi 22 janvier 2024 à 11:35 +0100, Christian König a écrit :
>> Am 19.01.24 um 15:13 schrieb Paul Cercueil:
>>> These functions should be used by device drivers when they start
>>> and
>>> stop accessing the data of DMABUF. It allows DMABUF importers to
>>> cache
>>> the dma_buf_attachment while ensuring that the data they want to
>>> access
>>> is available for their device when the DMA transfers take place.
>> As Daniel already noted as well this is a complete no-go from the
>> DMA-buf design point of view.
> What do you mean "as Daniel already noted"? It was him who suggested
> this.

Sorry, I haven't fully catched up to the discussion then.

In general DMA-buf is build around the idea that the data can be 
accessed coherently by the involved devices.

Having a begin/end of access for devices was brought up multiple times 
but so far rejected for good reasons.

That an exporter has to call extra functions to access his own buffers 
is a complete no-go for the design since this forces exporters into 
doing extra steps for allowing importers to access their data.

That in turn is pretty much un-testable unless you have every possible 
importer around while testing the exporter.

Regards,
Christian.

>
>> Regards,
>> Christian.
> Cheers,
> -Paul
>
>>> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
>>>
>>> ---
>>> v5: New patch
>>> ---
>>>    drivers/dma-buf/dma-buf.c | 66
>>> +++++++++++++++++++++++++++++++++++++++
>>>    include/linux/dma-buf.h   | 37 ++++++++++++++++++++++
>>>    2 files changed, 103 insertions(+)
>>>
>>> diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
>>> index 8fe5aa67b167..a8bab6c18fcd 100644
>>> --- a/drivers/dma-buf/dma-buf.c
>>> +++ b/drivers/dma-buf/dma-buf.c
>>> @@ -830,6 +830,8 @@ static struct sg_table * __map_dma_buf(struct
>>> dma_buf_attachment *attach,
>>>     *     - dma_buf_mmap()
>>>     *     - dma_buf_begin_cpu_access()
>>>     *     - dma_buf_end_cpu_access()
>>> + *     - dma_buf_begin_access()
>>> + *     - dma_buf_end_access()
>>>     *     - dma_buf_map_attachment_unlocked()
>>>     *     - dma_buf_unmap_attachment_unlocked()
>>>     *     - dma_buf_vmap_unlocked()
>>> @@ -1602,6 +1604,70 @@ void dma_buf_vunmap_unlocked(struct dma_buf
>>> *dmabuf, struct iosys_map *map)
>>>    }
>>>    EXPORT_SYMBOL_NS_GPL(dma_buf_vunmap_unlocked, DMA_BUF);
>>>    
>>> +/**
>>> + * @dma_buf_begin_access - Call before any hardware access from/to
>>> the DMABUF
>>> + * @attach:	[in]	attachment used for hardware access
>>> + * @sg_table:	[in]	scatterlist used for the DMA transfer
>>> + * @direction:  [in]    direction of DMA transfer
>>> + */
>>> +int dma_buf_begin_access(struct dma_buf_attachment *attach,
>>> +			 struct sg_table *sgt, enum
>>> dma_data_direction dir)
>>> +{
>>> +	struct dma_buf *dmabuf;
>>> +	bool cookie;
>>> +	int ret;
>>> +
>>> +	if (WARN_ON(!attach))
>>> +		return -EINVAL;
>>> +
>>> +	dmabuf = attach->dmabuf;
>>> +
>>> +	if (!dmabuf->ops->begin_access)
>>> +		return 0;
>>> +
>>> +	cookie = dma_fence_begin_signalling();
>>> +	ret = dmabuf->ops->begin_access(attach, sgt, dir);
>>> +	dma_fence_end_signalling(cookie);
>>> +
>>> +	if (WARN_ON_ONCE(ret))
>>> +		return ret;
>>> +
>>> +	return 0;
>>> +}
>>> +EXPORT_SYMBOL_NS_GPL(dma_buf_begin_access, DMA_BUF);
>>> +
>>> +/**
>>> + * @dma_buf_end_access - Call after any hardware access from/to
>>> the DMABUF
>>> + * @attach:	[in]	attachment used for hardware access
>>> + * @sg_table:	[in]	scatterlist used for the DMA transfer
>>> + * @direction:  [in]    direction of DMA transfer
>>> + */
>>> +int dma_buf_end_access(struct dma_buf_attachment *attach,
>>> +		       struct sg_table *sgt, enum
>>> dma_data_direction dir)
>>> +{
>>> +	struct dma_buf *dmabuf;
>>> +	bool cookie;
>>> +	int ret;
>>> +
>>> +	if (WARN_ON(!attach))
>>> +		return -EINVAL;
>>> +
>>> +	dmabuf = attach->dmabuf;
>>> +
>>> +	if (!dmabuf->ops->end_access)
>>> +		return 0;
>>> +
>>> +	cookie = dma_fence_begin_signalling();
>>> +	ret = dmabuf->ops->end_access(attach, sgt, dir);
>>> +	dma_fence_end_signalling(cookie);
>>> +
>>> +	if (WARN_ON_ONCE(ret))
>>> +		return ret;
>>> +
>>> +	return 0;
>>> +}
>>> +EXPORT_SYMBOL_NS_GPL(dma_buf_end_access, DMA_BUF);
>>> +
>>>    #ifdef CONFIG_DEBUG_FS
>>>    static int dma_buf_debug_show(struct seq_file *s, void *unused)
>>>    {
>>> diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h
>>> index 8ff4add71f88..8ba612c7cc16 100644
>>> --- a/include/linux/dma-buf.h
>>> +++ b/include/linux/dma-buf.h
>>> @@ -246,6 +246,38 @@ struct dma_buf_ops {
>>>    	 */
>>>    	int (*end_cpu_access)(struct dma_buf *, enum
>>> dma_data_direction);
>>>    
>>> +	/**
>>> +	 * @begin_access:
>>> +	 *
>>> +	 * This is called from dma_buf_begin_access() when a
>>> device driver
>>> +	 * wants to access the data of the DMABUF. The exporter
>>> can use this
>>> +	 * to flush/sync the caches if needed.
>>> +	 *
>>> +	 * This callback is optional.
>>> +	 *
>>> +	 * Returns:
>>> +	 *
>>> +	 * 0 on success or a negative error code on failure.
>>> +	 */
>>> +	int (*begin_access)(struct dma_buf_attachment *, struct
>>> sg_table *,
>>> +			    enum dma_data_direction);
>>> +
>>> +	/**
>>> +	 * @end_access:
>>> +	 *
>>> +	 * This is called from dma_buf_end_access() when a device
>>> driver is
>>> +	 * done accessing the data of the DMABUF. The exporter can
>>> use this
>>> +	 * to flush/sync the caches if needed.
>>> +	 *
>>> +	 * This callback is optional.
>>> +	 *
>>> +	 * Returns:
>>> +	 *
>>> +	 * 0 on success or a negative error code on failure.
>>> +	 */
>>> +	int (*end_access)(struct dma_buf_attachment *, struct
>>> sg_table *,
>>> +			  enum dma_data_direction);
>>> +
>>>    	/**
>>>    	 * @mmap:
>>>    	 *
>>> @@ -606,6 +638,11 @@ void dma_buf_detach(struct dma_buf *dmabuf,
>>>    int dma_buf_pin(struct dma_buf_attachment *attach);
>>>    void dma_buf_unpin(struct dma_buf_attachment *attach);
>>>    
>>> +int dma_buf_begin_access(struct dma_buf_attachment *attach,
>>> +			 struct sg_table *sgt, enum
>>> dma_data_direction dir);
>>> +int dma_buf_end_access(struct dma_buf_attachment *attach,
>>> +		       struct sg_table *sgt, enum
>>> dma_data_direction dir);
>>> +
>>>    struct dma_buf *dma_buf_export(const struct dma_buf_export_info
>>> *exp_info);
>>>    
>>>    int dma_buf_fd(struct dma_buf *dmabuf, int flags);
Daniel Vetter Jan. 25, 2024, 6:10 p.m. UTC | #4
On Fri, Jan 19, 2024 at 03:13:57PM +0100, Paul Cercueil wrote:
> These functions should be used by device drivers when they start and
> stop accessing the data of DMABUF. It allows DMABUF importers to cache
> the dma_buf_attachment while ensuring that the data they want to access
> is available for their device when the DMA transfers take place.
> 
> Signed-off-by: Paul Cercueil <paul@crapouillou.net>

Putting my detailed review comments here just so I don't have to remember
them any longer. We need to reach consensus on the big picture direction
first.

> 
> ---
> v5: New patch
> ---
>  drivers/dma-buf/dma-buf.c | 66 +++++++++++++++++++++++++++++++++++++++
>  include/linux/dma-buf.h   | 37 ++++++++++++++++++++++
>  2 files changed, 103 insertions(+)
> 
> diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
> index 8fe5aa67b167..a8bab6c18fcd 100644
> --- a/drivers/dma-buf/dma-buf.c
> +++ b/drivers/dma-buf/dma-buf.c
> @@ -830,6 +830,8 @@ static struct sg_table * __map_dma_buf(struct dma_buf_attachment *attach,
>   *     - dma_buf_mmap()
>   *     - dma_buf_begin_cpu_access()
>   *     - dma_buf_end_cpu_access()
> + *     - dma_buf_begin_access()
> + *     - dma_buf_end_access()
>   *     - dma_buf_map_attachment_unlocked()
>   *     - dma_buf_unmap_attachment_unlocked()
>   *     - dma_buf_vmap_unlocked()
> @@ -1602,6 +1604,70 @@ void dma_buf_vunmap_unlocked(struct dma_buf *dmabuf, struct iosys_map *map)
>  }
>  EXPORT_SYMBOL_NS_GPL(dma_buf_vunmap_unlocked, DMA_BUF);
>  
> +/**
> + * @dma_buf_begin_access - Call before any hardware access from/to the DMABUF
> + * @attach:	[in]	attachment used for hardware access
> + * @sg_table:	[in]	scatterlist used for the DMA transfer
> + * @direction:  [in]    direction of DMA transfer

I think for the kerneldoc would be good to point at the other function
here, explain why this might be needed and that for most reasonable
devices it's probably not, and link between the function pairs.

Also we need to document that dma_buf_map does an implied
dma_buf_begin_access (because dma_sg_map does an implied
dma_sg_sync_for_device) and vice versa for dma_buf_end_access. Which also
means that dma_buf_map/unmap should link to these functions in their
kerneldoc too.

Finally I think we should document here that it's ok to call these from
dma_fence signalling critical section and link to the relevant discussion
in the dma_fence docs for that.

> + */
> +int dma_buf_begin_access(struct dma_buf_attachment *attach,
> +			 struct sg_table *sgt, enum dma_data_direction dir)
> +{
> +	struct dma_buf *dmabuf;
> +	bool cookie;
> +	int ret;
> +
> +	if (WARN_ON(!attach))
> +		return -EINVAL;
> +
> +	dmabuf = attach->dmabuf;
> +
> +	if (!dmabuf->ops->begin_access)
> +		return 0;
> +
> +	cookie = dma_fence_begin_signalling();
> +	ret = dmabuf->ops->begin_access(attach, sgt, dir);
> +	dma_fence_end_signalling(cookie);
> +
> +	if (WARN_ON_ONCE(ret))
> +		return ret;
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL_NS_GPL(dma_buf_begin_access, DMA_BUF);

So explicit device side coherency management is not going to be very
compatible with dynamic buffer managament where the exporter can move the
buffer around. The reason for that is that for a dynamic exporter we cache
the sg mapping, which means any device-side coherency management which
dma_buf_map/unmap would do will not happen (since it's cached),
potentially breaking things for importers that rely on the assumption that
dma_buf_map/unmap already implies dma_buf_begin/end_device_access.

I think for now it's sufficient to put a WARN_ON(dma_buf_is_dymamic() &&
ops->begin|end_access) or similar into dma_buf_export and bail out with an
error to catch that.

Aside from the nits I do think this is roughly what we brievely discussed
well over a decade ago in the original dma-buf kickoff meeting at a linaro
connect in Budapest :-)

Cheers, Sima

> +
> +/**
> + * @dma_buf_end_access - Call after any hardware access from/to the DMABUF
> + * @attach:	[in]	attachment used for hardware access
> + * @sg_table:	[in]	scatterlist used for the DMA transfer
> + * @direction:  [in]    direction of DMA transfer
> + */
> +int dma_buf_end_access(struct dma_buf_attachment *attach,
> +		       struct sg_table *sgt, enum dma_data_direction dir)
> +{
> +	struct dma_buf *dmabuf;
> +	bool cookie;
> +	int ret;
> +
> +	if (WARN_ON(!attach))
> +		return -EINVAL;
> +
> +	dmabuf = attach->dmabuf;
> +
> +	if (!dmabuf->ops->end_access)
> +		return 0;
> +
> +	cookie = dma_fence_begin_signalling();
> +	ret = dmabuf->ops->end_access(attach, sgt, dir);
> +	dma_fence_end_signalling(cookie);
> +
> +	if (WARN_ON_ONCE(ret))
> +		return ret;
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL_NS_GPL(dma_buf_end_access, DMA_BUF);
> +
>  #ifdef CONFIG_DEBUG_FS
>  static int dma_buf_debug_show(struct seq_file *s, void *unused)
>  {
> diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h
> index 8ff4add71f88..8ba612c7cc16 100644
> --- a/include/linux/dma-buf.h
> +++ b/include/linux/dma-buf.h
> @@ -246,6 +246,38 @@ struct dma_buf_ops {
>  	 */
>  	int (*end_cpu_access)(struct dma_buf *, enum dma_data_direction);
>  
> +	/**
> +	 * @begin_access:
> +	 *
> +	 * This is called from dma_buf_begin_access() when a device driver
> +	 * wants to access the data of the DMABUF. The exporter can use this
> +	 * to flush/sync the caches if needed.
> +	 *
> +	 * This callback is optional.
> +	 *
> +	 * Returns:
> +	 *
> +	 * 0 on success or a negative error code on failure.
> +	 */
> +	int (*begin_access)(struct dma_buf_attachment *, struct sg_table *,
> +			    enum dma_data_direction);
> +
> +	/**
> +	 * @end_access:
> +	 *
> +	 * This is called from dma_buf_end_access() when a device driver is
> +	 * done accessing the data of the DMABUF. The exporter can use this
> +	 * to flush/sync the caches if needed.
> +	 *
> +	 * This callback is optional.
> +	 *
> +	 * Returns:
> +	 *
> +	 * 0 on success or a negative error code on failure.
> +	 */
> +	int (*end_access)(struct dma_buf_attachment *, struct sg_table *,
> +			  enum dma_data_direction);
> +
>  	/**
>  	 * @mmap:
>  	 *
> @@ -606,6 +638,11 @@ void dma_buf_detach(struct dma_buf *dmabuf,
>  int dma_buf_pin(struct dma_buf_attachment *attach);
>  void dma_buf_unpin(struct dma_buf_attachment *attach);
>  
> +int dma_buf_begin_access(struct dma_buf_attachment *attach,
> +			 struct sg_table *sgt, enum dma_data_direction dir);
> +int dma_buf_end_access(struct dma_buf_attachment *attach,
> +		       struct sg_table *sgt, enum dma_data_direction dir);
> +
>  struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
>  
>  int dma_buf_fd(struct dma_buf *dmabuf, int flags);
> -- 
> 2.43.0
> 
>
Christian König Jan. 26, 2024, 4:42 p.m. UTC | #5
Am 25.01.24 um 19:01 schrieb Daniel Vetter:
> On Thu, Jan 25, 2024 at 04:00:16PM +0100, Christian König wrote:
>> Am 24.01.24 um 11:58 schrieb Paul Cercueil:
>>> [SNIP]
>>>>> The problem was then that dma_buf_unmap_attachment cannot be called
>>>>> before the dma_fence is signaled, and calling it after is already
>>>>> too
>>>>> late (because the fence would be signaled before the data is
>>>>> sync'd).
>>>>    Well what sync are you talking about? CPU sync? In DMA-buf that is
>>>> handled differently.
>>>>    For importers it's mandatory that they can be coherent with the
>>>> exporter. That usually means they can snoop the CPU cache if the
>>>> exporter can snoop the CPU cache.
>>> I seem to have such a system where one device can snoop the CPU cache
>>> and the other cannot. Therefore if I want to support it properly, I do
>>> need cache flush/sync. I don't actually try to access the data using
>>> the CPU (and when I do, I call the sync start/end ioctls).
>> Usually that isn't a problem as long as you don't access the data with the
>> CPU.
>>
>> [SNIP]
>>
>>>>> (and I *think* there is a way to force coherency in the
>>>>> Ultrascale's
>>>>> interconnect - we're investigating it)
>>>>    What you can do is that instead of using udmabuf or dma-heaps is
>>>> that the device which can't provide coherency act as exporters of the
>>>> buffers.
>>>>    The exporter is allowed to call sync_for_cpu/sync_for_device on it's
>>>> own buffers and also gets begin/end CPU access notfications. So you
>>>> can then handle coherency between the exporter and the CPU.
>>> But again that would only work if the importers would call
>>> begin_cpu_access() / end_cpu_access(), which they don't, because they
>>> don't actually access the data using the CPU.
>> Wow, that is a completely new use case then.
>>
>> Neither DMA-buf nor the DMA subsystem in Linux actually supports this as far
>> as I can see.
>>
>>> Unless you mean that the exporter can call sync_for_cpu/sync_for_device
>>> before/after every single DMA transfer so that the data appears
>>> coherent to the importers, without them having to call
>>> begin_cpu_access() / end_cpu_access().
>> Yeah, I mean the importers don't have to call begin_cpu_access() /
>> end_cpu_access() if they don't do CPU access :)
>>
>> What you can still do as exporter is to call sync_for_device() and
>> sync_for_cpu() before and after each operation on your non-coherent device.
>> Paired with the fence signaling that should still work fine then.
>>
>> But taking a step back, this use case is not something even the low level
>> DMA subsystem supports. That sync_for_cpu() does the right thing is
>> coincident and not proper engineering.
>>
>> What you need is a sync_device_to_device() which does the appropriate
>> actions depending on which devices are involved.
>>
>>> In which case - this would still demultiply the complexity; my USB-
>>> functionfs interface here (and IIO interface in the separate patchset)
>>> are not device-specific, so I'd rather keep them importers.
>>>>    If you really don't have coherency between devices then that would
>>>> be a really new use case and we would need much more agreement on how
>>>> to do this.
>>> [snip]
>>>
>>> Agreed. Desiging a good generic solution would be better.
>>>
>>> With that said...
>>>
>>> Let's keep it out of this USB-functionfs interface for now. The
>>> interface does work perfectly fine on platforms that don't have
>>> coherency problems. The coherency issue in itself really is a
>>> tangential issue.
>> Yeah, completely agree.
>>
>>> So I will send a v6 where I don't try to force the cache coherency -
>>> and instead assume that the attached devices are coherent between
>>> themselves.
>>>
>>> But it would be even better to have a way to detect non-coherency and
>>> return an error on attach.
>> Take a look into the DMA subsystem. I'm pretty sure we already have
>> something like this in there.
>>
>> If nothing else helps you could take a look if the coherent memory access
>> mask is non zero or something like that.
> Jumping in way late and apolgies to everyone since yes I indeed suggested
> this entire mess to Paul in some private thread.
>
> And worse, I think we need it, it's just that we got away without it thus
> far.
>
> So way back at the og dma-buf kick-off dma coherency was discussed, and a
> few things where noted:
> - the dma api only supports device<->cpu coherency
> - getting the full coherency model off the ground right away is probably
>    too hard, so we made the decision that where it matters, relevant
>    flushing needs to be done in dma_buf_map/unmap.
>
> If you look at the earliest patches for dma-buf we had pretty clear
> language that all dma-operations should be bracketed with map/unmap. Of
> course that didn't work out for drm at all, and we had to first get
> dma_resv_lock and dma_fence landed and then your dynamic exporter/importer
> support in just to get the buffer migration functionality working, which
> was only one of the things discussed that braketing everything with
> map/unmap was supposed to take care of.
>
> The other was coherency management. But looking through archives I think
> this was already agreed to be postponed for later in the original kick-off
> meeting and never further discussed on the mailing list.
>
> This worked for a fairly long time, because thus far dma-buf was used on
> fairly reaasonable architectures where all participating devices are
> coherent enough.
>
> We did have to add the cpu access flushing fairly quickly because there's
> a lot of SoC chips (including intel) where that was necessary, but even
> that was added later on, as an opt-in and without fixing every. See
> fc13020e086b ("dma-buf: add support for kernel cpu access").
>
> The ioctl to allow userspace to do flushing was added even later on, and
> there the entire yolo opt-in situation is even worse. c11e391da2a8
> ("dma-buf: Add ioctls to allow userspace to flush") was only in 2016, 5
> years after dma-buf landed.
>
> It looks like it's finally time to add the device side flushing functions
> we've talked about first over 12 years ago :-)
>
> The reason this pops up now is that unlike other dma-buf users on maybe
> somewhat more funky architectures, Paul's patches want to use dma_fence
> for synchronization of the dma operations. Which means you cannot call the
> full dma_buf_map/unmap dance because that takes dma_resv_lock, and
> absolute no-go in a dma_fence critical path.
>
> And yes in those 12 years the dma-api hasn't gained the device2device sync
> support we'd need, but neither has it gained the multiple devices <-> cpu
> sync support we'd strictly need for dma-buf. So yes this is all a terrible
> hodge-podge of hacks, but if we'd require theoretically perfect code we'd
> still have zero dma-buf support in upstream.
>
> This also includes how we landed these extensions, none of them in the
> past have landed with a "update all existing exporters/importers" rule. We
> talked about that every time, and rejected it every time for imo pretty
> good reasons - the perf impact tends to be way too harsh if you impose
> over-flushing on everyone, including the reasonable platforms. And we
> currently can't do less than overflushing with the current dma-api
> interfaces because we dont have the specific flush functions we'd need. So
> really this isn't doing a worse abuse of the dma-api than what we have.
> It's definitely a bit wasteful since the functions we use do in theory
> flush too much. But in practice on the these funky architectures they
> flush enough.
>
> There's also the very hard issue of actually trying to optimize flushes,
> because a dma operation might only access part of a buffer, and you might
> interleave read/write access by different devices in very innovative ways.
> So I'm firmly on the "make it work first, then fast" side of things.
>
> So dma-buf will continue to be a thing that's tested for specific combos,
> and then we'll patch them. It's a decade-plus tradition at this point.
>
> Which is all a very long winded way of saying that yes, I think we need
> this, and we needed this 12 years ago already if we'd have aimed for
> perfect.
>
> I have a bunch of detail comments on the patch itself, but I guess we
> first need to find consensus on whether it's a good idea in the first
> place.

Well I think we should have some solution, but I'm not sure if it should 
be part of DMA-buf.

Essentially a DMA-buf exports the buffers as he uses it and the importer 
(or the DMA-buf subsystem) is then the one who says ok I can use this or 
I can't use this or I need to call extra functions to use this or whatever.

It's not the job of the exporter to provide the coherency for the 
importer, cause otherwise we would have a lot of code in the exporter 
which can only be tested when you have the right importer around. And I 
strongly think that this is a no-go for having a reliable solution.

That's why I think the approach of having DMA-buf callbacks is most 
likely the wrong thing to do.

What should happen instead is that the DMA subsystem provides 
functionality which to devices which don't support coherency through 
it's connection to say I want to access this data, please make sure to 
flush the appropriate catches. But that's just a very very rough design 
idea.

This will become more with CXL at the horizon I think.

Regards,
Christian.

>
> Cheers, Sima
Christian König Jan. 30, 2024, 1:09 p.m. UTC | #6
Am 30.01.24 um 11:40 schrieb Daniel Vetter:
> On Tue, Jan 30, 2024 at 10:48:23AM +0100, Paul Cercueil wrote:
>> Le mardi 30 janvier 2024 à 10:23 +0100, Christian König a écrit :
>>>   I would say we start with the DMA-API by getting away from sg_tables
>>> to something cleaner and state oriented.
>> FYI I am already adding a 'dma_vec' object in my IIO DMABUF patchset,
>> which is just a dead simple
>>
>> struct dma_vec {
>>    dma_addr_t addr;
>>    size_t len;
>> };
>>
>> (The rationale for introducing it in the IIO DMABUF patchset was that
>> the "scatterlist" wouldn't allow me to change the transfer size.)
>>
>> So I believe a new "sg_table"-like could just be an array of struct
>> dma_vec + flags.
> Yeah that's pretty much the proposal I've seen, split the sg table into
> input data (struct page + len) and output data (which is the dma_addr_t +
> len you have above).

I would extend that a bit and say we have an array with 
dma_addr+power_of_two_order and a header structure with lower bit offset 
and some DMA transaction flags.

But this is something which can be worked as an optimization later on. 
For a start this proposal here looks good to me as well.

> The part I don't expect to ever happen, because it hasn't the past 20 or
> so years, is that the dma-api will give us information about what is
> needed to keep the buffers coherency between various devices and the cpu.

Well maybe that's what we are doing wrong.

Instead of asking the dma-api about the necessary information we should 
give the API the opportunity to work for us.

In other words we don't need the information about buffer coherency what 
we need is that the API works for as and fulfills the requirements we have.

So the question is really what should we propose to change on the 
DMA-api side to get this working as expected?

Regards,
Christian.





> -Sima
Christian König Feb. 6, 2024, 1:28 p.m. UTC | #7
Am 31.01.24 um 10:07 schrieb Daniel Vetter:
> On Tue, Jan 30, 2024 at 02:09:45PM +0100, Christian König wrote:
>> Am 30.01.24 um 11:40 schrieb Daniel Vetter:
>>> On Tue, Jan 30, 2024 at 10:48:23AM +0100, Paul Cercueil wrote:
>>>> Le mardi 30 janvier 2024 à 10:23 +0100, Christian König a écrit :
>>>>>    I would say we start with the DMA-API by getting away from sg_tables
>>>>> to something cleaner and state oriented.
>>>> FYI I am already adding a 'dma_vec' object in my IIO DMABUF patchset,
>>>> which is just a dead simple
>>>>
>>>> struct dma_vec {
>>>>     dma_addr_t addr;
>>>>     size_t len;
>>>> };
>>>>
>>>> (The rationale for introducing it in the IIO DMABUF patchset was that
>>>> the "scatterlist" wouldn't allow me to change the transfer size.)
>>>>
>>>> So I believe a new "sg_table"-like could just be an array of struct
>>>> dma_vec + flags.
>>> Yeah that's pretty much the proposal I've seen, split the sg table into
>>> input data (struct page + len) and output data (which is the dma_addr_t +
>>> len you have above).
>> I would extend that a bit and say we have an array with
>> dma_addr+power_of_two_order and a header structure with lower bit offset and
>> some DMA transaction flags.
>>
>> But this is something which can be worked as an optimization later on. For a
>> start this proposal here looks good to me as well.
>>
>>> The part I don't expect to ever happen, because it hasn't the past 20 or
>>> so years, is that the dma-api will give us information about what is
>>> needed to keep the buffers coherency between various devices and the cpu.
>> Well maybe that's what we are doing wrong.
>>
>> Instead of asking the dma-api about the necessary information we should give
>> the API the opportunity to work for us.
>>
>> In other words we don't need the information about buffer coherency what we
>> need is that the API works for as and fulfills the requirements we have.
>>
>> So the question is really what should we propose to change on the DMA-api
>> side to get this working as expected?
> So one thing I've been pondering, kinda picking up your point about CXL,
> is that we do make the coherency protocol more explicit by adding a
> coherency mode to dma_buf that the exporter sets. Some ideas for values
> this could have:
>
> - ATTOMIC_COHERENT: Fully cache coherent, including device/cpu atomis.
>    This would be for CXL. Non-CXL devices could still participate with the
>    old model using explicit devices flushes, but must at comply with
>    CPU_COHERENT.
>
>    There's also the power9-only nvlink that would fit here, but I guess
>    going forward CXL (and cache-coherent integrated gpu) would really be
>    the only users of this flag.
>
>    Peer2peer would have the same rules, otherwise doesn't really make
>    sense. Also we might want to forbib non-CXL imports for these buffers
>    maybe even? Not sure on that.
>
> - CPU_COHERENT: device transactions do snoop cpu devices caches, but
>    devices might do their own caching which isn't snooped by the cpu and
>    needs explicit device-side invalidate/flushing. This means pcie
>    importers are not allowed to use pcie no-snoop transactions, intel igpu
>    wouldn't be allowed to use MOCS that do the same, similar for arm
>    integrated devices.
>
>    Importers can skip all explicit cache management apis like
>    dma_buf_begin/end_cpu_access, or the newly proposed
>    dma_buf_begin/end_device_access here.
>
>    We'd need to figure out what exactly this means for peer2peer
>    transactions, I have no idea whether the no-snoop flag even does
>    anything for those.
>
>    We might also want to split up CPU_COHERENT into CPU_COHERENT_WB and
>    CPU_WOHERENT_WC, so that importers know whether cpu reads are going to
>    be crawling or not.
>
> - MEMORY_COHERENT: devices transactions do not snoop any caches, but
>    promise that all transactions are fully flushed to system memory. Any
>    devices transactions which do fill cpu caches must call the proposed
>    dma_buf_begin/end_device_access functions proposed here. Any cpu access
>    must be braketed by calls to dma_buf_begin/end_cpu_access.
>
>    If your device does fill cpu caches, then essentially you'd not be able
>    to import such buffers. Not sure whether we need to put the
>    responsibility of checking that onto importers or exporters. Ideally
>    core dma-buf.c code would check this.
>
>    Also maybe the cpu WC mapping mode would actually need to be a sub-mode
>    for MEMORY_COHERENT, because all cpu wc achieves is to avoid the need to
>    call dma_buf_begin/end_cpu_access, you would still need your devices to
>    be memory coherent. And if they're not, then you cannot use that
>    dma-buf.
>
>    Or maybe alternatively we need to guarantee that exporters which set
>    MEMORY_COHERENT implement dma_buf_begin/end_device_access to make things
>    work for these cpu-coherent but not memory-coherent devices. This
>    becomes very tricky with device/arch/bus specific details I think.
>
> - DMA_API_COHERENT: The memory is allocated or mapped by the dma-api, and
>    the exact coherency mode is not know. Importers _must_ braket all cpu
>    and device access with the respective dma_buf functions. This is
>    essentially the "we have no idea" default.
>
>    Note that exporters might export memory allocated with dma_map_alloc
>    with MEMORY_COHERENT or CPU_COHERENT if they know how the memory exactly
>    works. E.g. for most arm soc gpu/display drivers we can assume that the
>    dma-api gives us MEMORY_COHERENT or CPU_COHERENT_WC, and just use that.
>    Essentially this would make the current implicit assumptions explicit.
>
>    udmabuf would need to set this, definitely if Paul's patches to add the
>    explicit device flushes land.
>
> - DEFAULT_COHERENT: This would be the backwards compat legacy yolo
>    behvaior. I'm not sure whether we should alias that with
>    DMA_API_COHERENT or leave it as a special value to mark exporters which
>    haven't been updated for the much more explicit coherency handling yet.
>
>    The specification for this coherency mode would be a flat out "who
>    knows, just don't break existing use-cases with actual users".
>    Essentially the only reason we'd have this would be to make sure we can
>    avoid regressions of these existing use-cases, by keeping whatever
>    horrible heuristics we have in current exporters.
>
>    It would also allow us to convert exporters and importers on a case by
>    case basis.
>
> Note that all these coherency modes are defined in terms of bus-sepecific
> device access and in terms of dma_buf apis the importer must call or can
> skip. This way we'd avoid having to change the dma-api in a first step,
> and if this all works out properly we could then use the resulting dma-api
> as a baseline to propose dma-api extensions.

When I read this for the first time my initial impression was that the 
idea mostly looked good, but while thinking about it more and more I 
came to the conclusion that this would go into the wrong direction.

Maybe I'm repeating myself, but I think we first of all have to talk a 
bit about some aspects of coherency:

1. Intra device coherency. This means that intra devices caches are 
invalidated before beginning an operation and flushed before signaling 
that an operation finished.

2. Inter device and device to CPU coherency. This means that caches 
which sit in between devices and between devices and the CPU need to be 
invalidated and flushed appropriately when buffers are accessed by 
different parties.

Number 1 is device specific, part of the DMA-buf framework and handled 
by dma_fences. As far as I can see that part is actually quite well 
designed and I don't see any obvious need for change.

Number 2 is platform specific and I completely agree with the DMA-api 
folks that this doesn't belong into DMA-buf in the first place. That's 
why I think the begin_cpu_access()/end_cpu_access() callbacks are 
actually a bit misplaced. We still can use those in the exporter, but to 
make better buffer placement decisions, but should not invalidate any 
caches when they are called.

The flushing and invalidation for platform caches should really be in 
the DMA-buf framework and not the exporter.

So in my thinking the enumeration you outlined above should really go 
into struct device and explaining to everybody what the coherency 
properties of DMA operations of this device is.

> I think starting right out with designing dma-api extension is a few
> bridges too far. Both from a "how do we convince upstream" pov, but maybe
> even more from a "how do we figure out what we even need" pov.

Well I totally agree on the "how do we figure out what we even need", 
but I disagree a bit on that we don't know what DMA-api extension we need.

We don't have the full picture yet, but as I already outlined from the 
DMA-api pov we have two major things on our TODO list:

1. Somehow remove the struct pages from the DMA-buf *importer* API.

     My best suggestion at the moment for this is to split sg_tables 
into two data structures, one for the struct pages and one for the DMA 
addresses.

     Mangling the addresses to ensure that no importer messes with the 
struct pages was a good step, but it also creates problems when 
dma_sync_sg_for_cpu() dma_sync_for_device() are supposed to be called.

2. Add some dma_sync_sg_between_devices(A, B....).

     And on this I think we are on the same page that we are going to 
need this, but we are just not clear on who is going to use it.

Regards,
Christian.

>
>> Regards,
>> Christian.
>>
>>
>>
>>
>>
>>> -Sima
>> _______________________________________________
>> Linaro-mm-sig mailing list -- linaro-mm-sig@lists.linaro.org
>> To unsubscribe send an email to linaro-mm-sig-leave@lists.linaro.org
diff mbox series

Patch

diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
index 8fe5aa67b167..a8bab6c18fcd 100644
--- a/drivers/dma-buf/dma-buf.c
+++ b/drivers/dma-buf/dma-buf.c
@@ -830,6 +830,8 @@  static struct sg_table * __map_dma_buf(struct dma_buf_attachment *attach,
  *     - dma_buf_mmap()
  *     - dma_buf_begin_cpu_access()
  *     - dma_buf_end_cpu_access()
+ *     - dma_buf_begin_access()
+ *     - dma_buf_end_access()
  *     - dma_buf_map_attachment_unlocked()
  *     - dma_buf_unmap_attachment_unlocked()
  *     - dma_buf_vmap_unlocked()
@@ -1602,6 +1604,70 @@  void dma_buf_vunmap_unlocked(struct dma_buf *dmabuf, struct iosys_map *map)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_vunmap_unlocked, DMA_BUF);
 
+/**
+ * @dma_buf_begin_access - Call before any hardware access from/to the DMABUF
+ * @attach:	[in]	attachment used for hardware access
+ * @sg_table:	[in]	scatterlist used for the DMA transfer
+ * @direction:  [in]    direction of DMA transfer
+ */
+int dma_buf_begin_access(struct dma_buf_attachment *attach,
+			 struct sg_table *sgt, enum dma_data_direction dir)
+{
+	struct dma_buf *dmabuf;
+	bool cookie;
+	int ret;
+
+	if (WARN_ON(!attach))
+		return -EINVAL;
+
+	dmabuf = attach->dmabuf;
+
+	if (!dmabuf->ops->begin_access)
+		return 0;
+
+	cookie = dma_fence_begin_signalling();
+	ret = dmabuf->ops->begin_access(attach, sgt, dir);
+	dma_fence_end_signalling(cookie);
+
+	if (WARN_ON_ONCE(ret))
+		return ret;
+
+	return 0;
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_begin_access, DMA_BUF);
+
+/**
+ * @dma_buf_end_access - Call after any hardware access from/to the DMABUF
+ * @attach:	[in]	attachment used for hardware access
+ * @sg_table:	[in]	scatterlist used for the DMA transfer
+ * @direction:  [in]    direction of DMA transfer
+ */
+int dma_buf_end_access(struct dma_buf_attachment *attach,
+		       struct sg_table *sgt, enum dma_data_direction dir)
+{
+	struct dma_buf *dmabuf;
+	bool cookie;
+	int ret;
+
+	if (WARN_ON(!attach))
+		return -EINVAL;
+
+	dmabuf = attach->dmabuf;
+
+	if (!dmabuf->ops->end_access)
+		return 0;
+
+	cookie = dma_fence_begin_signalling();
+	ret = dmabuf->ops->end_access(attach, sgt, dir);
+	dma_fence_end_signalling(cookie);
+
+	if (WARN_ON_ONCE(ret))
+		return ret;
+
+	return 0;
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_end_access, DMA_BUF);
+
 #ifdef CONFIG_DEBUG_FS
 static int dma_buf_debug_show(struct seq_file *s, void *unused)
 {
diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h
index 8ff4add71f88..8ba612c7cc16 100644
--- a/include/linux/dma-buf.h
+++ b/include/linux/dma-buf.h
@@ -246,6 +246,38 @@  struct dma_buf_ops {
 	 */
 	int (*end_cpu_access)(struct dma_buf *, enum dma_data_direction);
 
+	/**
+	 * @begin_access:
+	 *
+	 * This is called from dma_buf_begin_access() when a device driver
+	 * wants to access the data of the DMABUF. The exporter can use this
+	 * to flush/sync the caches if needed.
+	 *
+	 * This callback is optional.
+	 *
+	 * Returns:
+	 *
+	 * 0 on success or a negative error code on failure.
+	 */
+	int (*begin_access)(struct dma_buf_attachment *, struct sg_table *,
+			    enum dma_data_direction);
+
+	/**
+	 * @end_access:
+	 *
+	 * This is called from dma_buf_end_access() when a device driver is
+	 * done accessing the data of the DMABUF. The exporter can use this
+	 * to flush/sync the caches if needed.
+	 *
+	 * This callback is optional.
+	 *
+	 * Returns:
+	 *
+	 * 0 on success or a negative error code on failure.
+	 */
+	int (*end_access)(struct dma_buf_attachment *, struct sg_table *,
+			  enum dma_data_direction);
+
 	/**
 	 * @mmap:
 	 *
@@ -606,6 +638,11 @@  void dma_buf_detach(struct dma_buf *dmabuf,
 int dma_buf_pin(struct dma_buf_attachment *attach);
 void dma_buf_unpin(struct dma_buf_attachment *attach);
 
+int dma_buf_begin_access(struct dma_buf_attachment *attach,
+			 struct sg_table *sgt, enum dma_data_direction dir);
+int dma_buf_end_access(struct dma_buf_attachment *attach,
+		       struct sg_table *sgt, enum dma_data_direction dir);
+
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);