Skip to contentSkip to conversation

Introduce buffer synchronization framework

RFC PATCH Superseded Inki Dae 2 participants 7 messages Aug 13, 2013 – Aug 21, 2013

Subsystems (5)

v6 of 9 Superseded Latest is v9 →

lore
Share & tools
lore

Inki Dae

author
#0 lore raw

v6 cover Introduce buffer synchronization framework

Hi all,

This patch set introduces a buffer synchronization framework based on DMA BUF[1] and based on ww-mutexes[2] for lock mechanism, and may be final RFC.

The purpose of this framework is to provide not only buffer access control to CPU and CPU, and CPU and DMA, and DMA and DMA but also easy-to-use interfaces for device drivers and user application. In addtion, this patch set suggests a way for enhancing performance.

For generic user mode interface, we have used fcntl and select system call[3]. As you know, user application sees a buffer object as a dma-buf file descriptor. So fcntl() call with the file descriptor means to lock some buffer region being managed by the dma-buf object. And select() call means to wait for the completion of CPU or DMA access to the dma-buf without locking. For more detail, you can refer to the dma-buf-sync.txt in Documentation/

There are some cases we should use this buffer synchronization framework. One of which is to primarily enhance GPU rendering performance on Tizen platform in case of 3d app with compositing mode that 3d app draws something in off-screen buffer, and Web app.

In case of 3d app with compositing mode which is not a full screen mode, the app calls glFlush to submit 3d commands to GPU driver instead of glFinish for more performance. The reason we call glFlush is that glFinish blocks caller's task until the execution of the 2d commands is completed. Thus, that makes GPU and CPU more idle. As result, 3d rendering performance with glFinish is quite lower than glFlush. However, the use of glFlush has one issue that the a buffer shared with GPU could be broken when CPU accesses the buffer at once after glFlush because CPU cannot be aware of the completion of GPU access to the buffer. Of course, the app can be aware of that time using eglWaitGL but this function is valid only in case of the same process.

In case of Tizen, there are some applications that one process draws something in its own off-screen buffer (pixmap buffer) using CPU, and other process gets a off-screen buffer (window buffer) from Xorg using DRI2GetBuffers, and then composites the pixmap buffer with the window buffer using GPU, and finally page flip.

Web app based on HTML5 also has the same issue. Web browser and its web app are different process. The web app draws something in its own pixmap buffer, and then the web browser gets a window buffer from Xorg, and then composites the pixmap buffer with the window buffer. And finally, page flip.

Thus, in such cases, a shared buffer could be broken as one process draws something in pixmap buffer using CPU, when other process composites the pixmap buffer with window buffer using GPU without any locking mechanism. That is why we need user land locking interface, fcntl system call.

And last one is a deferred page flip issue. This issue is that a window buffer rendered can be displayed on screen in about 32ms in worst case:
assume that the gpu rendering is completed within 16ms.
That can be incurred when compositing a pixmap buffer with a window buffer using GPU and when vsync is just started. At this time, Xorg waits for a vblank event to get a window buffer so 3d rendering will be delayed up to about 16ms. As a result, the window buffer would be displayed in about two vsyncs (about 32ms) and in turn, that would show slow responsiveness.

For this, we could enhance the responsiveness with locking
mechanism: skipping one vblank wait. I guess in the similar reason, Android, Chrome OS, and other platforms are using their own locking mechanisms; Android sync driver, KDS, and DMA fence.

The below shows the deferred page flip issue in worst case,

               |------------ <- vsync signal
               |<------ DRI2GetBuffers
               |
               |
               |
               |------------ <- vsync signal
               |<------ Request gpu rendering
          time |
               |
               |<------ Request page flip (deferred)
               |------------ <- vsync signal
               |<------ Displayed on screen
               |
               |
               |
               |------------ <- vsync signal

Thanks,
Inki Dae

References:
[1] http://lwn.net/Articles/470339/
[2] https://patchwork.kernel.org/patch/2625361/
[3] http://linux.die.net/man/2/fcntl

Inki Dae (2):
[RFC PATCH v6] dmabuf-sync: Add a buffer synchronization framework [RFC PATCH v2] dma-buf: Add user interfaces for dmabuf sync support.

 Documentation/dma-buf-sync.txt |  285 +++++++++++++++++
 drivers/base/Kconfig           |    7 +
 drivers/base/Makefile          |    1 +
 drivers/base/dma-buf.c         |   85 +++++
 drivers/base/dmabuf-sync.c     |  678 ++++++++++++++++++++++++++++++++++++++++
 include/linux/dma-buf.h        |   16 +
 include/linux/dmabuf-sync.h    |  191 +++++++++++
 7 files changed, 1263 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/dma-buf-sync.txt
 create mode 100644 drivers/base/dmabuf-sync.c
 create mode 100644 include/linux/dmabuf-sync.h
Signature · 1 line
1.7.5.4
author
#1 lore raw

RFC PATCH v6 1/2 dmabuf-sync: Add a buffer synchronization framework

This patch adds a buffer synchronization framework based on DMA BUF[1] and and based on ww-mutexes[2] for lock mechanism.

The purpose of this framework is to provide not only buffer access control to CPU and DMA but also easy-to-use interfaces for device drivers and user application. This framework can be used for all dma devices using system memory as dma buffer, especially for most ARM based SoCs.

Changelog v6:
- Fix sync lock to multiple reads.
- Add select system call support.
. Wake up poll_wait when a dmabuf is unlocked.
- Remove unnecessary the use of mutex lock.
- Add private backend ops callbacks.
. This ops has one callback for device drivers to clean up their
sync object resource when the sync object is freed. For this, device drivers should implement the free callback properly.
- Update document file.

Changelog v5:
- Rmove a dependence on reservation_object: the reservation_object is used to hook up to ttm and dma-buf for easy sharing of reservations across devices. However, the dmabuf sync can be used for all dma devices; v4l2 and drm based drivers, so doesn't need the reservation_object anymore. With regared to this, it adds 'void *sync' to dma_buf structure.
- All patches are rebased on mainline, Linux v3.10.

Changelog v4:
- Add user side interface for buffer synchronization mechanism and update descriptions related to the user side interface.

Changelog v3:
- remove cache operation relevant codes and update document file.

Changelog v2:
- use atomic_add_unless to avoid potential bug.
- add a macro for checking valid access type.
- code clean.

The mechanism of this framework has the following steps,
1. Register dmabufs to a sync object - A task gets a new sync object and can add one or more dmabufs that the task wants to access.
This registering should be performed when a device context or an event context such as a page flip event is created or before CPU accesses a shared buffer.

dma_buf_sync_get(a sync object, a dmabuf);

2. Lock a sync object - A task tries to lock all dmabufs added in its own sync object. Basically, the lock mechanism uses ww-mutex[1] to avoid dead lock issue and for race condition between CPU and CPU, CPU and DMA, and DMA and DMA. Taking a lock means that others cannot access all locked dmabufs until the task that locked the corresponding dmabufs, unlocks all the locked dmabufs.
This locking should be performed before DMA or CPU accesses these dmabufs.

dma_buf_sync_lock(a sync object);

3. Unlock a sync object - The task unlocks all dmabufs added in its own sync object. The unlock means that the DMA or CPU accesses to the dmabufs have been completed so that others may access them.
This unlocking should be performed after DMA or CPU has completed accesses to the dmabufs.

dma_buf_sync_unlock(a sync object);

4. Unregister one or all dmabufs from a sync object - A task unregisters the given dmabufs from the sync object. This means that the task dosen't want to lock the dmabufs.
The unregistering should be performed after DMA or CPU has completed accesses to the dmabufs or when dma_buf_sync_lock() is failed.

dma_buf_sync_put(a sync object, a dmabuf); dma_buf_sync_put_all(a sync object);

The described steps may be summarized as:
get -> lock -> CPU or DMA access to a buffer/s -> unlock -> put

This framework includes the following two features.
1. read (shared) and write (exclusive) locks - A task is required to declare the access type when the task tries to register a dmabuf;
READ, WRITE, READ DMA, or WRITE DMA.

The below is example codes,
struct dmabuf_sync *sync;

sync = dmabuf_sync_init(...);
...

dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_R);
...

And the below can be used as access types:
DMA_BUF_ACCESS_R - CPU will access a buffer for read. DMA_BUF_ACCESS_W - CPU will access a buffer for read or write. DMA_BUF_ACCESS_DMA_R - DMA will access a buffer for read DMA_BUF_ACCESS_DMA_W - DMA will access a buffer for read or
write.

2. Mandatory resource releasing - a task cannot hold a lock indefinitely. A task may never try to unlock a buffer after taking a lock to the buffer. In this case, a timer handler to the corresponding sync object is called in five (default) seconds and then the timed-out buffer is unlocked by work queue handler to avoid lockups and to enforce resources of the buffer.

The below is how to use interfaces for device driver:

	1. Allocate and Initialize a sync object:
		static void xxx_dmabuf_sync_free(void *priv)
		{
			struct xxx_context *ctx = priv;

			if (!ctx)
				return;

			ctx->sync = NULL;
		}
		...

		static struct dmabuf_sync_priv_ops driver_specific_ops = {
			.free = xxx_dmabuf_sync_free,
		};
		...

		struct dmabuf_sync *sync;

		sync = dmabuf_sync_init("test sync", &driver_specific_ops, ctx);
		...

	2. Add a dmabuf to the sync object when setting up dma buffer relevant
	   registers:
		dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_READ);
		...

	3. Lock all dmabufs of the sync object before DMA or CPU accesses
	   the dmabufs:
		dmabuf_sync_lock(sync);
		...

	4. Now CPU or DMA can access all dmabufs locked in step 3.

	5. Unlock all dmabufs added in a sync object after DMA or CPU access
	   to these dmabufs is completed:
		dmabuf_sync_unlock(sync);

	   And call the following functions to release all resources,
		dmabuf_sync_put_all(sync);
		dmabuf_sync_fini(sync);

	You can refer to actual example codes:
		"drm/exynos: add dmabuf sync support for g2d driver" and
		"drm/exynos: add dmabuf sync support for kms framework" from
		https://git.kernel.org/cgit/linux/kernel/git/daeinki/
		drm-exynos.git/log/?h=dmabuf-sync

And this framework includes fcntl system call[3] as interfaces exported to user. As you know, user sees a buffer object as a dma-buf file descriptor. So fcntl() call with the file descriptor means to lock some buffer region being managed by the dma-buf object.

The below is how to use interfaces for user application:

fcntl system call:

	struct flock filelock;

	1. Lock a dma buf:
		filelock.l_type = F_WRLCK or F_RDLCK;

		/* lock entire region to the dma buf. */
		filelock.lwhence = SEEK_CUR;
		filelock.l_start = 0;
		filelock.l_len = 0;

		fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
		...
		CPU access to the dma buf

	2. Unlock a dma buf:
		filelock.l_type = F_UNLCK;

		fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);

		close(dmabuf fd) call would also unlock the dma buf. And for more
		detail, please refer to [3]

select system call:

	fd_set wdfs or rdfs;

	FD_ZERO(&wdfs or &rdfs);
	FD_SET(fd, &wdfs or &rdfs);

	select(fd + 1, &rdfs, NULL, NULL, NULL);
		or
	select(fd + 1, NULL, &wdfs, NULL, NULL);

	Every time select system call is called, a caller will wait for
	the completion of DMA or CPU access to a shared buffer if there
	is someone accessing the shared buffer; locked the shared buffer.
	However, if no anyone then select system call will be returned
	at once.

References:
[1] http://lwn.net/Articles/470339/
[2] https://patchwork.kernel.org/patch/2625361/
[3] http://linux.die.net/man/2/fcntl

---
 Documentation/dma-buf-sync.txt |  285 +++++++++++++++++
 drivers/base/Kconfig           |    7 +
 drivers/base/Makefile          |    1 +
 drivers/base/dma-buf.c         |    4 +
 drivers/base/dmabuf-sync.c     |  678 ++++++++++++++++++++++++++++++++++++++++
 include/linux/dma-buf.h        |   16 +
 include/linux/dmabuf-sync.h    |  190 +++++++++++
 7 files changed, 1181 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/dma-buf-sync.txt
 create mode 100644 drivers/base/dmabuf-sync.c
 create mode 100644 include/linux/dmabuf-sync.h
Documentation/dma-buf-sync.txt new file +285 -0
new file mode 100644
index 0000000..8023d06
@@ -0,0 +1,285 @@
+                    DMA Buffer Synchronization Framework
+                    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+                                  Inki Dae
+                      <inki dot dae at samsung dot com>
+                          <daeinki at gmail dot com>
+
+This document is a guide for device-driver writers describing the DMA buffer
+synchronization API. This document also describes how to use the API to
+use buffer synchronization mechanism between DMA and DMA, CPU and DMA, and
+CPU and CPU.
+
+The DMA Buffer synchronization API provides buffer synchronization mechanism;
+i.e., buffer access control to CPU and DMA, and easy-to-use interfaces for
+device drivers and user application. And this API can be used for all dma
+devices using system memory as dma buffer, especially for most ARM based SoCs.
+
+
+Motivation
+----------
+
+Buffer synchronization issue between DMA and DMA:
+	Sharing a buffer, a device cannot be aware of when the other device
+	will access the shared buffer: a device may access a buffer containing
+	wrong data if the device accesses the shared buffer while another
+	device is still accessing the shared buffer.
+	Therefore, a user process should have waited for the completion of DMA
+	access by another device before a device tries to access the shared
+	buffer.
+
+Buffer synchronization issue between CPU and DMA:
+	A user process should consider that when having to send a buffer, filled
+	by CPU, to a device driver for the device driver to access the buffer as
+	a input buffer while CPU and DMA are sharing the buffer.
+	This means that the user process needs to understand how the device
+	driver is worked. Hence, the conventional mechanism not only makes
+	user application complicated but also incurs performance overhead.
+
+Buffer synchronization issue between CPU and CPU:
+	In case that two processes share one buffer; shared with DMA also,
+	they may need some mechanism to allow process B to access the shared
+	buffer after the completion of CPU access by process A.
+	Therefore, process B should have waited for the completion of CPU access
+	by process A using the mechanism before trying to access the shared
+	buffer.
+
+What is the best way to solve these buffer synchronization issues?
+	We may need a common object that a device driver and a user process
+	notify the common object of when they try to access a shared buffer.
+	That way we could decide when we have to allow or not to allow for CPU
+	or DMA to access the shared buffer through the common object.
+	If so, what could become the common object? Right, that's a dma-buf[1].
+	Now we have already been using the dma-buf to share one buffer with
+	other drivers.
+
+
+Basic concept
+-------------
+
+The mechanism of this framework has the following steps,
+    1. Register dmabufs to a sync object - A task gets a new sync object and
+    can add one or more dmabufs that the task wants to access.
+    This registering should be performed when a device context or an event
+    context such as a page flip event is created or before CPU accesses a shared
+    buffer.
+
+	dma_buf_sync_get(a sync object, a dmabuf);
+
+    2. Lock a sync object - A task tries to lock all dmabufs added in its own
+    sync object. Basically, the lock mechanism uses ww-mutexes[2] to avoid dead
+    lock issue and for race condition between CPU and CPU, CPU and DMA, and DMA
+    and DMA. Taking a lock means that others cannot access all locked dmabufs
+    until the task that locked the corresponding dmabufs, unlocks all the locked
+    dmabufs.
+    This locking should be performed before DMA or CPU accesses these dmabufs.
+
+	dma_buf_sync_lock(a sync object);
+
+    3. Unlock a sync object - The task unlocks all dmabufs added in its own sync
+    object. The unlock means that the DMA or CPU accesses to the dmabufs have
+    been completed so that others may access them.
+    This unlocking should be performed after DMA or CPU has completed accesses
+    to the dmabufs.
+
+	dma_buf_sync_unlock(a sync object);
+
+    4. Unregister one or all dmabufs from a sync object - A task unregisters
+    the given dmabufs from the sync object. This means that the task dosen't
+    want to lock the dmabufs.
+    The unregistering should be performed after DMA or CPU has completed
+    accesses to the dmabufs or when dma_buf_sync_lock() is failed.
+
+	dma_buf_sync_put(a sync object, a dmabuf);
+	dma_buf_sync_put_all(a sync object);
+
+    The described steps may be summarized as:
+	get -> lock -> CPU or DMA access to a buffer/s -> unlock -> put
+
+This framework includes the following two features.
+    1. read (shared) and write (exclusive) locks - A task is required to declare
+    the access type when the task tries to register a dmabuf;
+    READ, WRITE, READ DMA, or WRITE DMA.
+
+    The below is example codes,
+	struct dmabuf_sync *sync;
+
+	sync = dmabuf_sync_init(NULL, "test sync");
+
+	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_R);
+	...
+
+    2. Mandatory resource releasing - a task cannot hold a lock indefinitely.
+    A task may never try to unlock a buffer after taking a lock to the buffer.
+    In this case, a timer handler to the corresponding sync object is called
+    in five (default) seconds and then the timed-out buffer is unlocked by work
+    queue handler to avoid lockups and to enforce resources of the buffer.
+
+
+Access types
+------------
+
+DMA_BUF_ACCESS_R - CPU will access a buffer for read.
+DMA_BUF_ACCESS_W - CPU will access a buffer for read or write.
+DMA_BUF_ACCESS_DMA_R - DMA will access a buffer for read
+DMA_BUF_ACCESS_DMA_W - DMA will access a buffer for read or write.
+
+
+Generic user interfaces
+-----------------------
+
+And this framework includes fcntl system call[3] as interfaces exported
+to user. As you know, user sees a buffer object as a dma-buf file descriptor.
+So fcntl() call with the file descriptor means to lock some buffer region being
+managed by the dma-buf object.
+
+
+API set
+-------
+
+bool is_dmabuf_sync_supported(void)
+	- Check if dmabuf sync is supported or not.
+
+struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void priv*)
+	- Allocate and initialize a new sync object. The caller can get a new
+	sync object for buffer synchronization. ops is used for device driver
+	to clean up its own sync object. For this, each device driver should
+	implement a free callback. priv is used for device driver to get its
+	device context when free callback is called.
+
+void dmabuf_sync_fini(struct dmabuf_sync *sync)
+	- Release all resources to the sync object.
+
+int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf,
+			unsigned int type)
+	- Get dmabuf sync object. Internally, this function allocates
+	a dmabuf_sync object and adds a given dmabuf to it, and also takes
+	a reference to the dmabuf. The caller can tie up multiple dmabufs
+	into one sync object by calling this function several times.
+
+void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf *dmabuf)
+	- Put dmabuf sync object to a given dmabuf. Internally, this function
+	removes a given dmabuf from a sync object and remove the sync object.
+	At this time, the dmabuf is putted.
+
+void dmabuf_sync_put_all(struct dmabuf_sync *sync)
+	- Put dmabuf sync object to dmabufs. Internally, this function removes
+	all dmabufs from a sync object and remove the sync object.
+	At this time, all dmabufs are putted.
+
+int dmabuf_sync_lock(struct dmabuf_sync *sync)
+	- Lock all dmabufs added in a sync object. The caller should call this
+	function prior to CPU or DMA access to the dmabufs so that others can
+	not access the dmabufs. Internally, this function avoids dead lock
+	issue with ww-mutexes.
+
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf)
+	- Lock a dmabuf. The caller should call this
+	function prior to CPU or DMA access to the dmabuf so that others can
+	not access the dmabuf.
+
+int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+	- Unlock all dmabufs added in a sync object. The caller should call
+	this function after CPU or DMA access to the dmabufs is completed so
+	that others can access the dmabufs.
+
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+	- Unlock a dmabuf. The caller should call this function after CPU or
+	DMA access to the dmabuf is completed so that others can access
+	the dmabuf.
+
+
+Tutorial for device driver
+--------------------------
+
+1. Allocate and Initialize a sync object:
+	static void xxx_dmabuf_sync_free(void *priv)
+	{
+		struct xxx_context *ctx = priv;
+
+		if (!ctx)
+			return;
+
+		ctx->sync = NULL;
+	}
+	...
+
+	static struct dmabuf_sync_priv_ops driver_specific_ops = {
+		.free = xxx_dmabuf_sync_free,
+	};
+	...
+
+	struct dmabuf_sync *sync;
+
+	sync = dmabuf_sync_init("test sync", &driver_specific_ops, ctx);
+	...
+
+2. Add a dmabuf to the sync object when setting up dma buffer relevant registers:
+	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_READ);
+	...
+
+3. Lock all dmabufs of the sync object before DMA or CPU accesses the dmabufs:
+	dmabuf_sync_lock(sync);
+	...
+
+4. Now CPU or DMA can access all dmabufs locked in step 3.
+
+5. Unlock all dmabufs added in a sync object after DMA or CPU access to these
+   dmabufs is completed:
+	dmabuf_sync_unlock(sync);
+
+   And call the following functions to release all resources,
+	dmabuf_sync_put_all(sync);
+	dmabuf_sync_fini(sync);
+
+
+Tutorial for user application
+-----------------------------
+fcntl system call:
+
+	struct flock filelock;
+
+1. Lock a dma buf:
+	filelock.l_type = F_WRLCK or F_RDLCK;
+
+	/* lock entire region to the dma buf. */
+	filelock.lwhence = SEEK_CUR;
+	filelock.l_start = 0;
+	filelock.l_len = 0;
+
+	fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
+	...
+	CPU access to the dma buf
+
+2. Unlock a dma buf:
+	filelock.l_type = F_UNLCK;
+
+	fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
+
+	close(dmabuf fd) call would also unlock the dma buf. And for more
+	detail, please refer to [3]
+
+
+select system call:
+
+	fd_set wdfs or rdfs;
+
+	FD_ZERO(&wdfs or &rdfs);
+	FD_SET(fd, &wdfs or &rdfs);
+
+	select(fd + 1, &rdfs, NULL, NULL, NULL);
+		or
+	select(fd + 1, NULL, &wdfs, NULL, NULL);
+
+	Every time select system call is called, a caller will wait for
+	the completion of DMA or CPU access to a shared buffer if there is
+	someone accessing the shared buffer; locked the shared buffer.
+	However, if no anyone then select system call will be returned
+	at once.
+
+References:
+[1] http://lwn.net/Articles/470339/
+[2] https://patchwork.kernel.org/patch/2625361/
+[3] http://linux.die.net/man/2/fcntl
drivers/base/Kconfig +7 -0
index 5daa259..35e1518 100644
@@ -200,6 +200,13 @@ config DMA_SHARED_BUFFER
 	  APIs extension; the file's descriptor can then be passed on to other
 	  driver.
 
+config DMABUF_SYNC
+	bool "DMABUF Synchronization Framework"
+	depends on DMA_SHARED_BUFFER
+	help
+	  This option enables dmabuf sync framework for buffer synchronization between
+	  DMA and DMA, CPU and DMA, and CPU and CPU.
+
 config CMA
 	bool "Contiguous Memory Allocator"
 	depends on HAVE_DMA_CONTIGUOUS && HAVE_MEMBLOCK
drivers/base/Makefile +1 -0
index 48029aa..e06a5d7 100644
@@ -11,6 +11,7 @@ obj-y			+= power/
 obj-$(CONFIG_HAS_DMA)	+= dma-mapping.o
 obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o
 obj-$(CONFIG_DMA_SHARED_BUFFER) += dma-buf.o reservation.o
+obj-$(CONFIG_DMABUF_SYNC) += dmabuf-sync.o
 obj-$(CONFIG_ISA)	+= isa.o
 obj-$(CONFIG_FW_LOADER)	+= firmware_class.o
 obj-$(CONFIG_NUMA)	+= node.o
drivers/base/dma-buf.c +4 -0
index 6687ba7..4aca57a 100644
@@ -29,6 +29,7 @@
 #include <linux/export.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
+#include <linux/dmabuf-sync.h>
 
 static inline int is_dma_buf_file(struct file *);
 
@@ -56,6 +57,8 @@ static int dma_buf_release(struct inode *inode, struct file *file)
 	list_del(&dmabuf->list_node);
 	mutex_unlock(&db_list.lock);
 
+	dmabuf_sync_reservation_fini(dmabuf);
+
 	kfree(dmabuf);
 	return 0;
 }
@@ -134,6 +137,7 @@ struct dma_buf *dma_buf_export_named(void *priv, const struct dma_buf_ops *ops,
 
 	file = anon_inode_getfile("dmabuf", &dma_buf_fops, dmabuf, flags);
 
+	dmabuf_sync_reservation_init(dmabuf);
 	dmabuf->file = file;
 
 	mutex_init(&dmabuf->lock);
drivers/base/dmabuf-sync.c new file +678 -0

678 lines not on this page · raw message

include/linux/dma-buf.h +16 -0
index dfac5ed..0109673 100644
@@ -115,6 +115,7 @@ struct dma_buf_ops {
  * @exp_name: name of the exporter; useful for debugging.
  * @list_node: node for dma_buf accounting and debugging.
  * @priv: exporter specific private data for this buffer object.
+ * @sync: sync object linked to this dma-buf
  */
 struct dma_buf {
 	size_t size;
@@ -128,6 +129,7 @@ struct dma_buf {
 	const char *exp_name;
 	struct list_head list_node;
 	void *priv;
+	void *sync;
 };
 
 /**
@@ -148,6 +150,20 @@ struct dma_buf_attachment {
 	void *priv;
 };
 
+#define	DMA_BUF_ACCESS_R	0x1
+#define DMA_BUF_ACCESS_W	0x2
+#define DMA_BUF_ACCESS_DMA	0x4
+#define DMA_BUF_ACCESS_RW	(DMA_BUF_ACCESS_R | DMA_BUF_ACCESS_W)
+#define DMA_BUF_ACCESS_DMA_R	(DMA_BUF_ACCESS_R | DMA_BUF_ACCESS_DMA)
+#define DMA_BUF_ACCESS_DMA_W	(DMA_BUF_ACCESS_W | DMA_BUF_ACCESS_DMA)
+#define DMA_BUF_ACCESS_DMA_RW	(DMA_BUF_ACCESS_DMA_R | DMA_BUF_ACCESS_DMA_W)
+#define IS_VALID_DMA_BUF_ACCESS_TYPE(t)	(t == DMA_BUF_ACCESS_R || \
+					 t == DMA_BUF_ACCESS_W || \
+					 t == DMA_BUF_ACCESS_DMA_R || \
+					 t == DMA_BUF_ACCESS_DMA_W || \
+					 t == DMA_BUF_ACCESS_RW || \
+					 t == DMA_BUF_ACCESS_DMA_RW)
+
 /**
  * get_dma_buf - convenience wrapper for get_file.
  * @dmabuf:	[in]	pointer to dma_buf
include/linux/dmabuf-sync.h new file +190 -0

190 lines not on this page · raw message

Signature · 1 line
1.7.5.4
#2 lore raw

Re: [PATCH 1/2] [RFC PATCH v6] dmabuf-sync: Add a buffer synchronization framework

Quoted 606 lines · Inki Dae

This patch adds a buffer synchronization framework based on DMA BUF[1] and and based on ww-mutexes[2] for lock mechanism.

The purpose of this framework is to provide not only buffer access control to CPU and DMA but also easy-to-use interfaces for device drivers and user application. This framework can be used for all dma devices using system memory as dma buffer, especially for most ARM based SoCs.

Changelog v6:
- Fix sync lock to multiple reads.
- Add select system call support.
. Wake up poll_wait when a dmabuf is unlocked.
- Remove unnecessary the use of mutex lock.
- Add private backend ops callbacks.
. This ops has one callback for device drivers to clean up their
sync object resource when the sync object is freed. For this, device drivers should implement the free callback properly.
- Update document file.

Changelog v5:
- Rmove a dependence on reservation_object: the reservation_object is used to hook up to ttm and dma-buf for easy sharing of reservations across devices. However, the dmabuf sync can be used for all dma devices; v4l2 and drm based drivers, so doesn't need the reservation_object anymore. With regared to this, it adds 'void *sync' to dma_buf structure.
- All patches are rebased on mainline, Linux v3.10.

Changelog v4:
- Add user side interface for buffer synchronization mechanism and update descriptions related to the user side interface.

Changelog v3:
- remove cache operation relevant codes and update document file.

Changelog v2:
- use atomic_add_unless to avoid potential bug.
- add a macro for checking valid access type.
- code clean.

The mechanism of this framework has the following steps,
1. Register dmabufs to a sync object - A task gets a new sync object and can add one or more dmabufs that the task wants to access.
This registering should be performed when a device context or an event context such as a page flip event is created or before CPU accesses a shared buffer.

dma_buf_sync_get(a sync object, a dmabuf);

2. Lock a sync object - A task tries to lock all dmabufs added in its own sync object. Basically, the lock mechanism uses ww-mutex[1] to avoid dead lock issue and for race condition between CPU and CPU, CPU and DMA, and DMA and DMA. Taking a lock means that others cannot access all locked dmabufs until the task that locked the corresponding dmabufs, unlocks all the locked dmabufs.
This locking should be performed before DMA or CPU accesses these dmabufs.

dma_buf_sync_lock(a sync object);

3. Unlock a sync object - The task unlocks all dmabufs added in its own sync object. The unlock means that the DMA or CPU accesses to the dmabufs have been completed so that others may access them.
This unlocking should be performed after DMA or CPU has completed accesses to the dmabufs.

dma_buf_sync_unlock(a sync object);

4. Unregister one or all dmabufs from a sync object - A task unregisters the given dmabufs from the sync object. This means that the task dosen't want to lock the dmabufs.
The unregistering should be performed after DMA or CPU has completed accesses to the dmabufs or when dma_buf_sync_lock() is failed.

dma_buf_sync_put(a sync object, a dmabuf); dma_buf_sync_put_all(a sync object);

The described steps may be summarized as:
get -> lock -> CPU or DMA access to a buffer/s -> unlock -> put

This framework includes the following two features.
1. read (shared) and write (exclusive) locks - A task is required to declare the access type when the task tries to register a dmabuf;
READ, WRITE, READ DMA, or WRITE DMA.

The below is example codes,
struct dmabuf_sync *sync;

sync = dmabuf_sync_init(...);
...

dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_R);
...

And the below can be used as access types:
DMA_BUF_ACCESS_R - CPU will access a buffer for read. DMA_BUF_ACCESS_W - CPU will access a buffer for read or write. DMA_BUF_ACCESS_DMA_R - DMA will access a buffer for read DMA_BUF_ACCESS_DMA_W - DMA will access a buffer for read or
write.

2. Mandatory resource releasing - a task cannot hold a lock indefinitely. A task may never try to unlock a buffer after taking a lock to the buffer. In this case, a timer handler to the corresponding sync object is called in five (default) seconds and then the timed-out buffer is unlocked by work queue handler to avoid lockups and to enforce resources of the buffer.

The below is how to use interfaces for device driver:

	1. Allocate and Initialize a sync object:
		static void xxx_dmabuf_sync_free(void *priv)
		{
			struct xxx_context *ctx = priv;

			if (!ctx)
				return;

			ctx->sync = NULL;
		}
		...

		static struct dmabuf_sync_priv_ops driver_specific_ops = {
			.free = xxx_dmabuf_sync_free,
		};
		...

		struct dmabuf_sync *sync;

		sync = dmabuf_sync_init("test sync", &driver_specific_ops, ctx);
		...

	2. Add a dmabuf to the sync object when setting up dma buffer relevant
	   registers:
		dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_READ);
		...

	3. Lock all dmabufs of the sync object before DMA or CPU accesses
	   the dmabufs:
		dmabuf_sync_lock(sync);
		...

	4. Now CPU or DMA can access all dmabufs locked in step 3.

	5. Unlock all dmabufs added in a sync object after DMA or CPU access
	   to these dmabufs is completed:
		dmabuf_sync_unlock(sync);

	   And call the following functions to release all resources,
		dmabuf_sync_put_all(sync);
		dmabuf_sync_fini(sync);

	You can refer to actual example codes:
		"drm/exynos: add dmabuf sync support for g2d driver" and
		"drm/exynos: add dmabuf sync support for kms framework" from
		https://git.kernel.org/cgit/linux/kernel/git/daeinki/
		drm-exynos.git/log/?h=dmabuf-sync

And this framework includes fcntl system call[3] as interfaces exported to user. As you know, user sees a buffer object as a dma-buf file descriptor. So fcntl() call with the file descriptor means to lock some buffer region being managed by the dma-buf object.

The below is how to use interfaces for user application:

fcntl system call:

	struct flock filelock;

	1. Lock a dma buf:
		filelock.l_type = F_WRLCK or F_RDLCK;

		/* lock entire region to the dma buf. */
		filelock.lwhence = SEEK_CUR;
		filelock.l_start = 0;
		filelock.l_len = 0;

		fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
		...
		CPU access to the dma buf

	2. Unlock a dma buf:
		filelock.l_type = F_UNLCK;

		fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);

		close(dmabuf fd) call would also unlock the dma buf. And for more
		detail, please refer to [3]

select system call:

	fd_set wdfs or rdfs;

	FD_ZERO(&wdfs or &rdfs);
	FD_SET(fd, &wdfs or &rdfs);

	select(fd + 1, &rdfs, NULL, NULL, NULL);
		or
	select(fd + 1, NULL, &wdfs, NULL, NULL);

	Every time select system call is called, a caller will wait for
	the completion of DMA or CPU access to a shared buffer if there
	is someone accessing the shared buffer; locked the shared buffer.
	However, if no anyone then select system call will be returned
	at once.

References:
[1] http://lwn.net/Articles/470339/
[2] https://patchwork.kernel.org/patch/2625361/
[3] http://linux.die.net/man/2/fcntl

---
 Documentation/dma-buf-sync.txt |  285 +++++++++++++++++
 drivers/base/Kconfig           |    7 +
 drivers/base/Makefile          |    1 +
 drivers/base/dma-buf.c         |    4 +
 drivers/base/dmabuf-sync.c     |  678 ++++++++++++++++++++++++++++++++++++++++
 include/linux/dma-buf.h        |   16 +
 include/linux/dmabuf-sync.h    |  190 +++++++++++
 7 files changed, 1181 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/dma-buf-sync.txt
 create mode 100644 drivers/base/dmabuf-sync.c
 create mode 100644 include/linux/dmabuf-sync.h
Documentation/dma-buf-sync.txt new file +285 -0
new file mode 100644
index 0000000..8023d06
@@ -0,0 +1,285 @@
+                    DMA Buffer Synchronization Framework
+                    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+                                  Inki Dae
+                      <inki dot dae at samsung dot com>
+                          <daeinki at gmail dot com>
+
+This document is a guide for device-driver writers describing the DMA buffer
+synchronization API. This document also describes how to use the API to
+use buffer synchronization mechanism between DMA and DMA, CPU and DMA, and
+CPU and CPU.
+
+The DMA Buffer synchronization API provides buffer synchronization mechanism;
+i.e., buffer access control to CPU and DMA, and easy-to-use interfaces for
+device drivers and user application. And this API can be used for all dma
+devices using system memory as dma buffer, especially for most ARM based SoCs.
+
+
+Motivation
+----------
+
+Buffer synchronization issue between DMA and DMA:
+	Sharing a buffer, a device cannot be aware of when the other device
+	will access the shared buffer: a device may access a buffer containing
+	wrong data if the device accesses the shared buffer while another
+	device is still accessing the shared buffer.
+	Therefore, a user process should have waited for the completion of DMA
+	access by another device before a device tries to access the shared
+	buffer.
+
+Buffer synchronization issue between CPU and DMA:
+	A user process should consider that when having to send a buffer, filled
+	by CPU, to a device driver for the device driver to access the buffer as
+	a input buffer while CPU and DMA are sharing the buffer.
+	This means that the user process needs to understand how the device
+	driver is worked. Hence, the conventional mechanism not only makes
+	user application complicated but also incurs performance overhead.
+
+Buffer synchronization issue between CPU and CPU:
+	In case that two processes share one buffer; shared with DMA also,
+	they may need some mechanism to allow process B to access the shared
+	buffer after the completion of CPU access by process A.
+	Therefore, process B should have waited for the completion of CPU access
+	by process A using the mechanism before trying to access the shared
+	buffer.
+
+What is the best way to solve these buffer synchronization issues?
+	We may need a common object that a device driver and a user process
+	notify the common object of when they try to access a shared buffer.
+	That way we could decide when we have to allow or not to allow for CPU
+	or DMA to access the shared buffer through the common object.
+	If so, what could become the common object? Right, that's a dma-buf[1].
+	Now we have already been using the dma-buf to share one buffer with
+	other drivers.
+
+
+Basic concept
+-------------
+
+The mechanism of this framework has the following steps,
+    1. Register dmabufs to a sync object - A task gets a new sync object and
+    can add one or more dmabufs that the task wants to access.
+    This registering should be performed when a device context or an event
+    context such as a page flip event is created or before CPU accesses a shared
+    buffer.
+
+	dma_buf_sync_get(a sync object, a dmabuf);
+
+    2. Lock a sync object - A task tries to lock all dmabufs added in its own
+    sync object. Basically, the lock mechanism uses ww-mutexes[2] to avoid dead
+    lock issue and for race condition between CPU and CPU, CPU and DMA, and DMA
+    and DMA. Taking a lock means that others cannot access all locked dmabufs
+    until the task that locked the corresponding dmabufs, unlocks all the locked
+    dmabufs.
+    This locking should be performed before DMA or CPU accesses these dmabufs.
+
+	dma_buf_sync_lock(a sync object);
+
+    3. Unlock a sync object - The task unlocks all dmabufs added in its own sync
+    object. The unlock means that the DMA or CPU accesses to the dmabufs have
+    been completed so that others may access them.
+    This unlocking should be performed after DMA or CPU has completed accesses
+    to the dmabufs.
+
+	dma_buf_sync_unlock(a sync object);
+
+    4. Unregister one or all dmabufs from a sync object - A task unregisters
+    the given dmabufs from the sync object. This means that the task dosen't
+    want to lock the dmabufs.
+    The unregistering should be performed after DMA or CPU has completed
+    accesses to the dmabufs or when dma_buf_sync_lock() is failed.
+
+	dma_buf_sync_put(a sync object, a dmabuf);
+	dma_buf_sync_put_all(a sync object);
+
+    The described steps may be summarized as:
+	get -> lock -> CPU or DMA access to a buffer/s -> unlock -> put
+
+This framework includes the following two features.
+    1. read (shared) and write (exclusive) locks - A task is required to declare
+    the access type when the task tries to register a dmabuf;
+    READ, WRITE, READ DMA, or WRITE DMA.
+
+    The below is example codes,
+	struct dmabuf_sync *sync;
+
+	sync = dmabuf_sync_init(NULL, "test sync");
+
+	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_R);
+	...
+
+    2. Mandatory resource releasing - a task cannot hold a lock indefinitely.
+    A task may never try to unlock a buffer after taking a lock to the buffer.
+    In this case, a timer handler to the corresponding sync object is called
+    in five (default) seconds and then the timed-out buffer is unlocked by work
+    queue handler to avoid lockups and to enforce resources of the buffer.
+
+
+Access types
+------------
+
+DMA_BUF_ACCESS_R - CPU will access a buffer for read.
+DMA_BUF_ACCESS_W - CPU will access a buffer for read or write.
+DMA_BUF_ACCESS_DMA_R - DMA will access a buffer for read
+DMA_BUF_ACCESS_DMA_W - DMA will access a buffer for read or write.
+
+
+Generic user interfaces
+-----------------------
+
+And this framework includes fcntl system call[3] as interfaces exported
+to user. As you know, user sees a buffer object as a dma-buf file descriptor.
+So fcntl() call with the file descriptor means to lock some buffer region being
+managed by the dma-buf object.
+
+
+API set
+-------
+
+bool is_dmabuf_sync_supported(void)
+	- Check if dmabuf sync is supported or not.
+
+struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void priv*)
+	- Allocate and initialize a new sync object. The caller can get a new
+	sync object for buffer synchronization. ops is used for device driver
+	to clean up its own sync object. For this, each device driver should
+	implement a free callback. priv is used for device driver to get its
+	device context when free callback is called.
+
+void dmabuf_sync_fini(struct dmabuf_sync *sync)
+	- Release all resources to the sync object.
+
+int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf,
+			unsigned int type)
+	- Get dmabuf sync object. Internally, this function allocates
+	a dmabuf_sync object and adds a given dmabuf to it, and also takes
+	a reference to the dmabuf. The caller can tie up multiple dmabufs
+	into one sync object by calling this function several times.
+
+void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf *dmabuf)
+	- Put dmabuf sync object to a given dmabuf. Internally, this function
+	removes a given dmabuf from a sync object and remove the sync object.
+	At this time, the dmabuf is putted.
+
+void dmabuf_sync_put_all(struct dmabuf_sync *sync)
+	- Put dmabuf sync object to dmabufs. Internally, this function removes
+	all dmabufs from a sync object and remove the sync object.
+	At this time, all dmabufs are putted.
+
+int dmabuf_sync_lock(struct dmabuf_sync *sync)
+	- Lock all dmabufs added in a sync object. The caller should call this
+	function prior to CPU or DMA access to the dmabufs so that others can
+	not access the dmabufs. Internally, this function avoids dead lock
+	issue with ww-mutexes.
+
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf)
+	- Lock a dmabuf. The caller should call this
+	function prior to CPU or DMA access to the dmabuf so that others can
+	not access the dmabuf.
+
+int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+	- Unlock all dmabufs added in a sync object. The caller should call
+	this function after CPU or DMA access to the dmabufs is completed so
+	that others can access the dmabufs.
+
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+	- Unlock a dmabuf. The caller should call this function after CPU or
+	DMA access to the dmabuf is completed so that others can access
+	the dmabuf.
+
+
+Tutorial for device driver
+--------------------------
+
+1. Allocate and Initialize a sync object:
+	static void xxx_dmabuf_sync_free(void *priv)
+	{
+		struct xxx_context *ctx = priv;
+
+		if (!ctx)
+			return;
+
+		ctx->sync = NULL;
+	}
+	...
+
+	static struct dmabuf_sync_priv_ops driver_specific_ops = {
+		.free = xxx_dmabuf_sync_free,
+	};
+	...
+
+	struct dmabuf_sync *sync;
+
+	sync = dmabuf_sync_init("test sync", &driver_specific_ops, ctx);
+	...
+
+2. Add a dmabuf to the sync object when setting up dma buffer relevant registers:
+	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_READ);
+	...
+
+3. Lock all dmabufs of the sync object before DMA or CPU accesses the dmabufs:
+	dmabuf_sync_lock(sync);
+	...
+
+4. Now CPU or DMA can access all dmabufs locked in step 3.
+
+5. Unlock all dmabufs added in a sync object after DMA or CPU access to these
+   dmabufs is completed:
+	dmabuf_sync_unlock(sync);
+
+   And call the following functions to release all resources,
+	dmabuf_sync_put_all(sync);
+	dmabuf_sync_fini(sync);
+
+
+Tutorial for user application
+-----------------------------
+fcntl system call:
+
+	struct flock filelock;
+
+1. Lock a dma buf:
+	filelock.l_type = F_WRLCK or F_RDLCK;
+
+	/* lock entire region to the dma buf. */
+	filelock.lwhence = SEEK_CUR;
+	filelock.l_start = 0;
+	filelock.l_len = 0;
+
+	fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
+	...
+	CPU access to the dma buf
+
+2. Unlock a dma buf:
+	filelock.l_type = F_UNLCK;
+
+	fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
+
+	close(dmabuf fd) call would also unlock the dma buf. And for more
+	detail, please refer to [3]
+
+
+select system call:
+
+	fd_set wdfs or rdfs;
+
+	FD_ZERO(&wdfs or &rdfs);
+	FD_SET(fd, &wdfs or &rdfs);
+
+	select(fd + 1, &rdfs, NULL, NULL, NULL);
+		or
+	select(fd + 1, NULL, &wdfs, NULL, NULL);
+
+	Every time select system call is called, a caller will wait for
+	the completion of DMA or CPU access to a shared buffer if there is
+	someone accessing the shared buffer; locked the shared buffer.
+	However, if no anyone then select system call will be returned
+	at once.
+
+References:
+[1] http://lwn.net/Articles/470339/
+[2] https://patchwork.kernel.org/patch/2625361/
+[3] http://linux.die.net/man/2/fcntl
drivers/base/Kconfig +7 -0
index 5daa259..35e1518 100644
@@ -200,6 +200,13 @@ config DMA_SHARED_BUFFER
 	  APIs extension; the file's descriptor can then be passed on to other
 	  driver.
 
+config DMABUF_SYNC
+	bool "DMABUF Synchronization Framework"
+	depends on DMA_SHARED_BUFFER
+	help
+	  This option enables dmabuf sync framework for buffer synchronization between
+	  DMA and DMA, CPU and DMA, and CPU and CPU.
+
 config CMA
 	bool "Contiguous Memory Allocator"
 	depends on HAVE_DMA_CONTIGUOUS && HAVE_MEMBLOCK
drivers/base/Makefile +1 -0
index 48029aa..e06a5d7 100644
@@ -11,6 +11,7 @@ obj-y			+= power/
 obj-$(CONFIG_HAS_DMA)	+= dma-mapping.o
 obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o
 obj-$(CONFIG_DMA_SHARED_BUFFER) += dma-buf.o reservation.o
+obj-$(CONFIG_DMABUF_SYNC) += dmabuf-sync.o
 obj-$(CONFIG_ISA)	+= isa.o
 obj-$(CONFIG_FW_LOADER)	+= firmware_class.o
 obj-$(CONFIG_NUMA)	+= node.o
drivers/base/dma-buf.c +4 -0
index 6687ba7..4aca57a 100644
@@ -29,6 +29,7 @@
 #include <linux/export.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
+#include <linux/dmabuf-sync.h>
 
 static inline int is_dma_buf_file(struct file *);
 
@@ -56,6 +57,8 @@ static int dma_buf_release(struct inode *inode, struct file *file)
 	list_del(&dmabuf->list_node);
 	mutex_unlock(&db_list.lock);
 
+	dmabuf_sync_reservation_fini(dmabuf);
+
 	kfree(dmabuf);
 	return 0;
 }
@@ -134,6 +137,7 @@ struct dma_buf *dma_buf_export_named(void *priv, const struct dma_buf_ops *ops,
 
 	file = anon_inode_getfile("dmabuf", &dma_buf_fops, dmabuf, flags);
 
+	dmabuf_sync_reservation_init(dmabuf);
 	dmabuf->file = file;
 
 	mutex_init(&dmabuf->lock);
drivers/base/dmabuf-sync.c new file +28 -0
new file mode 100644
index 0000000..fbe711c
@@ -0,0 +1,678 @@
+/*
+ * Copyright (C) 2013 Samsung Electronics Co.Ltd
+ * Authors:
+ *	Inki Dae <[email protected]>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/debugfs.h>
+#include <linux/uaccess.h>
+
+#include <linux/dmabuf-sync.h>
+
+#define MAX_SYNC_TIMEOUT	5 /* Second. */
+
+int dmabuf_sync_enabled = 1;
+
+MODULE_PARM_DESC(enabled, "Check if dmabuf sync is supported or not");
+module_param_named(enabled, dmabuf_sync_enabled, int, 0444);
+
+DEFINE_WW_CLASS(dmabuf_sync_ww_class);

drivers/base/dmabuf-sync.c line 38

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+EXPORT_SYMBOL(dmabuf_sync_ww_class);
+
+static void dmabuf_sync_timeout_worker(struct work_struct *work)
+{
+	struct dmabuf_sync *sync = container_of(work, struct dmabuf_sync, work);
+	struct dmabuf_sync_object *sobj;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry(sobj, &sync->syncs, head) {

You are using the 'sobj->robj' quite a lot. Why not just use a temp structure:

		struct dmabuf_sync_reservation *rsvp = sobj->robj;

and use that in this function. It would make it easier to read I think.

drivers/base/dmabuf-sync.c line 49

Quoted 11 lines
drivers/base/dmabuf-sync.c new file +11 -0
+		BUG_ON(!sobj->robj);
+
+		mutex_lock(&sobj->robj->lock);
+
+		printk(KERN_WARNING "%s: timeout = 0x%x [type = %d:%d, " \
+					"refcnt = %d, locked = %d]\n",
+					sync->name, (u32)sobj->dmabuf,
+					sobj->robj->accessed_type,
+					sobj->access_type,
+					atomic_read(&sobj->robj->shared_cnt),
+					sobj->robj->locked);

pr_warn_ratelimited?

Quoted 21 lines
drivers/base/dmabuf-sync.c new file +21 -0
+
+		/* unlock only valid sync object. */
+		if (!sobj->robj->locked) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		if (sobj->robj->polled) {
+			sobj->robj->poll_event = true;
+			sobj->robj->polled = false;
+			wake_up_interruptible(&sobj->robj->poll_wait);
+		}
+
+		if (atomic_add_unless(&sobj->robj->shared_cnt, -1, 1)) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		mutex_unlock(&sobj->robj->lock);
+
+		ww_mutex_unlock(&sobj->robj->sync_lock);

drivers/base/dmabuf-sync.c line 80

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+		mutex_lock(&sobj->robj->lock);
+		sobj->robj->locked = false;
+
+		if (sobj->access_type & DMA_BUF_ACCESS_R)
+			printk(KERN_WARNING "%s: r-unlocked = 0x%x\n",
+					sync->name, (u32)sobj->dmabuf);
+		else
+			printk(KERN_WARNING "%s: w-unlocked = 0x%x\n",
+					sync->name, (u32)sobj->dmabuf);

How about using 'pr_warn'? And in it have:

		sobj->access_type & DMA_BUF_ACCESS_R ? "r-" : "w-",

	and just have one printk.

Why the (u32) casting? Don't you want %p ?

Quoted 49 lines
drivers/base/dmabuf-sync.c new file +49 -0
+
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	sync->status = 0;
+	mutex_unlock(&sync->lock);
+
+	dmabuf_sync_put_all(sync);
+	dmabuf_sync_fini(sync);
+}
+
+static void dmabuf_sync_lock_timeout(unsigned long arg)
+{
+	struct dmabuf_sync *sync = (struct dmabuf_sync *)arg;
+
+	schedule_work(&sync->work);
+}
+
+static int dmabuf_sync_lock_objs(struct dmabuf_sync *sync,
+					struct ww_acquire_ctx *ctx)
+{
+	struct dmabuf_sync_object *contended_sobj = NULL;
+	struct dmabuf_sync_object *res_sobj = NULL;
+	struct dmabuf_sync_object *sobj = NULL;
+	int ret;
+
+	if (ctx)
+		ww_acquire_init(ctx, &dmabuf_sync_ww_class);
+
+retry:
+	list_for_each_entry(sobj, &sync->syncs, head) {
+		if (WARN_ON(!sobj->robj))
+			continue;
+
+		mutex_lock(&sobj->robj->lock);
+
+		/* Don't lock in case of read and read. */
+		if (sobj->robj->accessed_type & DMA_BUF_ACCESS_R &&
+		    sobj->access_type & DMA_BUF_ACCESS_R) {
+			atomic_inc(&sobj->robj->shared_cnt);
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		if (sobj == res_sobj) {
+			res_sobj = NULL;
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}

drivers/base/dmabuf-sync.c line 139

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+		mutex_unlock(&sobj->robj->lock);
+
+		ret = ww_mutex_lock(&sobj->robj->sync_lock, ctx);
+		if (ret < 0) {
+			contended_sobj = sobj;
+
+			if (ret == -EDEADLK)
+				printk(KERN_WARNING"%s: deadlock = 0x%x\n",
+					sync->name, (u32)sobj->dmabuf);

Again, why (u32) and not %p?

drivers/base/dmabuf-sync.c line 140

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+			goto err;

This looks odd. You jump to err, which jumps back to 'retry'. Won't this cause an infinite loop? Perhaps you need to add a retry counter to only do this up to five times or so and then give up?

Quoted 100 lines
drivers/base/dmabuf-sync.c new file +100 -0
+		}
+
+		mutex_lock(&sobj->robj->lock);
+		sobj->robj->locked = true;
+
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	if (ctx)
+		ww_acquire_done(ctx);
+
+	init_timer(&sync->timer);
+
+	sync->timer.data = (unsigned long)sync;
+	sync->timer.function = dmabuf_sync_lock_timeout;
+	sync->timer.expires = jiffies + (HZ * MAX_SYNC_TIMEOUT);
+
+	add_timer(&sync->timer);
+
+	return 0;
+
+err:
+	list_for_each_entry_continue_reverse(sobj, &sync->syncs, head) {
+		mutex_lock(&sobj->robj->lock);
+
+		/* Don't need to unlock in case of read and read. */
+		if (atomic_add_unless(&sobj->robj->shared_cnt, -1, 1)) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		ww_mutex_unlock(&sobj->robj->sync_lock);
+		sobj->robj->locked = false;
+
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	if (res_sobj) {
+		mutex_lock(&res_sobj->robj->lock);
+
+		if (!atomic_add_unless(&res_sobj->robj->shared_cnt, -1, 1)) {
+			ww_mutex_unlock(&res_sobj->robj->sync_lock);
+			res_sobj->robj->locked = false;
+		}
+
+		mutex_unlock(&res_sobj->robj->lock);
+	}
+
+	if (ret == -EDEADLK) {
+		ww_mutex_lock_slow(&contended_sobj->robj->sync_lock, ctx);
+		res_sobj = contended_sobj;
+
+		goto retry;
+	}
+
+	if (ctx)
+		ww_acquire_fini(ctx);
+
+	return ret;
+}
+
+static void dmabuf_sync_unlock_objs(struct dmabuf_sync *sync,
+					struct ww_acquire_ctx *ctx)
+{
+	struct dmabuf_sync_object *sobj;
+
+	if (list_empty(&sync->syncs))
+		return;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry(sobj, &sync->syncs, head) {
+		mutex_lock(&sobj->robj->lock);
+
+		if (sobj->robj->polled) {
+			sobj->robj->poll_event = true;
+			sobj->robj->polled = false;
+			wake_up_interruptible(&sobj->robj->poll_wait);
+		}
+
+		if (atomic_add_unless(&sobj->robj->shared_cnt, -1, 1)) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		mutex_unlock(&sobj->robj->lock);
+
+		ww_mutex_unlock(&sobj->robj->sync_lock);
+
+		mutex_lock(&sobj->robj->lock);
+		sobj->robj->locked = false;
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	mutex_unlock(&sync->lock);
+
+	if (ctx)
+		ww_acquire_fini(ctx);
+
+	del_timer(&sync->timer);

drivers/base/dmabuf-sync.c line 250

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+}
+
+/**
+ * is_dmabuf_sync_supported - Check if dmabuf sync is supported or not.
+ */
+bool is_dmabuf_sync_supported(void)
+{
+	return dmabuf_sync_enabled == 1;
+}
+EXPORT_SYMBOL(is_dmabuf_sync_supported);

_GPL ?

I would also prefix it with 'dmabuf_is_sync_supported' just to make all of the libraries call start with 'dmabuf'

Quoted 15 lines
drivers/base/dmabuf-sync.c new file +15 -0
+
+/**
+ * dmabuf_sync_init - Allocate and initialize a dmabuf sync.
+ *
+ * @priv: A device private data.
+ * @name: A sync object name.
+ *
+ * This function should be called when a device context or an event
+ * context such as a page flip event is created. And the created
+ * dmabuf_sync object should be set to the context.
+ * The caller can get a new sync object for buffer synchronization
+ * through this function.
+ */
+struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,

drivers/base/dmabuf-sync.c line 275

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+					void *priv)
+{
+	struct dmabuf_sync *sync;
+
+	sync = kzalloc(sizeof(*sync), GFP_KERNEL);
+	if (!sync)
+		return ERR_PTR(-ENOMEM);
+
+	strncpy(sync->name, name, ARRAY_SIZE(sync->name) - 1);
+

That is odd usage of an ARRAY_SIZE, but I can see how you can use it. I would say you should just do a #define for the 64 line and use that instead.

drivers/base/dmabuf-sync.c line 284

Quoted 9 lines
drivers/base/dmabuf-sync.c new file +9 -0
+	sync->ops = ops;
+	sync->priv = priv;
+	INIT_LIST_HEAD(&sync->syncs);
+	mutex_init(&sync->lock);
+	INIT_WORK(&sync->work, dmabuf_sync_timeout_worker);
+
+	return sync;
+}
+EXPORT_SYMBOL(dmabuf_sync_init);

_GPL ?

Quoted 8 lines
drivers/base/dmabuf-sync.c new file +8 -0
+
+/**
+ * dmabuf_sync_fini - Release a given dmabuf sync.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after
+ * dmabuf_sync_init call to release relevant resources, and after

drivers/base/dmabuf-sync.c line 302

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+ * dmabuf_sync_unlock function is called.
+ */
+void dmabuf_sync_fini(struct dmabuf_sync *sync)
+{
+	if (WARN_ON(!sync))
+		return;
+
+	if (sync->ops && sync->ops->free)
+		sync->ops->free(sync->priv);
+

No need to cancel the sync->work in case that is still running?

drivers/base/dmabuf-sync.c line 305

Quoted 3 lines
drivers/base/dmabuf-sync.c new file +3 -0
+	kfree(sync);
+}
+EXPORT_SYMBOL(dmabuf_sync_fini);

_GPL ?

drivers/base/dmabuf-sync.c line 308

Quoted 3 lines
drivers/base/dmabuf-sync.c new file +3 -0
+
+/*
+ * dmabuf_sync_get_obj - Add a given object to syncs list.

sync's list I think?

drivers/base/dmabuf-sync.c line 317

Quoted 9 lines
drivers/base/dmabuf-sync.c new file +9 -0
+ *
+ * @sync: An object to dmabuf_sync structure.
+ * @dmabuf: An object to dma_buf structure.
+ * @type: A access type to a dma buf.
+ *	The DMA_BUF_ACCESS_R means that this dmabuf could be accessed by
+ *	others for read access. On the other hand, the DMA_BUF_ACCESS_W
+ *	means that this dmabuf couldn't be accessed by others but would be
+ *	accessed by caller's dma exclusively. And the DMA_BUF_ACCESS_DMA can be
+ *	combined.

Should this be an enum?

drivers/base/dmabuf-sync.c line 323

Quoted 6 lines
drivers/base/dmabuf-sync.c new file +6 -0
+ *
+ * This function creates and initializes a new dmabuf sync object and it adds
+ * the dmabuf sync object to syncs list to track and manage all dmabufs.
+ */
+static int dmabuf_sync_get_obj(struct dmabuf_sync *sync, struct dma_buf *dmabuf,
+					unsigned int type)

enum for 'type'?

drivers/base/dmabuf-sync.c line 336

Quoted 13 lines
drivers/base/dmabuf-sync.c new file +13 -0
+{
+	struct dmabuf_sync_object *sobj;
+
+	if (!dmabuf->sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	if (!IS_VALID_DMA_BUF_ACCESS_TYPE(type))
+		return -EINVAL;
+
+	if ((type & DMA_BUF_ACCESS_RW) == DMA_BUF_ACCESS_RW)
+		type &= ~DMA_BUF_ACCESS_R;

Ah, that is why you are not using an enum.

drivers/base/dmabuf-sync.c line 340

Quoted 4 lines
drivers/base/dmabuf-sync.c new file +4 -0
+
+	sobj = kzalloc(sizeof(*sobj), GFP_KERNEL);
+	if (!sobj) {
+		WARN_ON(1);

I think you can skip that WARN_ON. Handling an -ENOMEM should be something fairly easy to handle by the calleer.

Quoted 12 lines
drivers/base/dmabuf-sync.c new file +12 -0
+		return -ENOMEM;
+	}
+
+	get_dma_buf(dmabuf);
+
+	sobj->dmabuf = dmabuf;
+	sobj->robj = dmabuf->sync;
+	sobj->access_type = type;
+
+	mutex_lock(&sync->lock);
+	list_add_tail(&sobj->head, &sync->syncs);
+	mutex_unlock(&sync->lock);

drivers/base/dmabuf-sync.c line 362

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+	return 0;
+}
+
+/*
+ * dmabuf_sync_put_obj - Release a given sync object.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after

s/is//

Quoted 22 lines
drivers/base/dmabuf-sync.c new file +22 -0
+ * dmabuf_sync_get_obj call to release a given sync object.
+ */
+static void dmabuf_sync_put_obj(struct dmabuf_sync *sync,
+					struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_object *sobj;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry(sobj, &sync->syncs, head) {
+		if (sobj->dmabuf != dmabuf)
+			continue;
+
+		dma_buf_put(sobj->dmabuf);
+
+		list_del_init(&sobj->head);
+		kfree(sobj);
+		break;
+	}
+
+	if (list_empty(&sync->syncs))
+		sync->status = 0;

drivers/base/dmabuf-sync.c line 394

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+	mutex_unlock(&sync->lock);
+}
+
+/*
+ * dmabuf_sync_put_objs - Release all sync objects of dmabuf_sync.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after

s/is//

Quoted 37 lines
drivers/base/dmabuf-sync.c new file +37 -0
+ * dmabuf_sync_get_obj call to release all sync objects.
+ */
+static void dmabuf_sync_put_objs(struct dmabuf_sync *sync)
+{
+	struct dmabuf_sync_object *sobj, *next;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry_safe(sobj, next, &sync->syncs, head) {
+		dma_buf_put(sobj->dmabuf);
+
+		list_del_init(&sobj->head);
+		kfree(sobj);
+	}
+
+	mutex_unlock(&sync->lock);
+
+	sync->status = 0;
+}
+
+/**
+ * dmabuf_sync_lock - lock all dmabufs added to syncs list.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * The caller should call this function prior to CPU or DMA access to
+ * the dmabufs so that others can not access the dmabufs.
+ * Internally, this function avoids dead lock issue with ww-mutex.
+ */
+int dmabuf_sync_lock(struct dmabuf_sync *sync)
+{
+	int ret;
+
+	if (!sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}

drivers/base/dmabuf-sync.c line 441

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+	if (list_empty(&sync->syncs))
+		return -EINVAL;
+
+	if (sync->status != DMABUF_SYNC_GOT)
+		return -EINVAL;
+
+	ret = dmabuf_sync_lock_objs(sync, &sync->ctx);
+	if (ret < 0) {
+		WARN_ON(1);

Perhaps also include the ret value in the WARN?

drivers/base/dmabuf-sync.c line 449

Quoted 8 lines
drivers/base/dmabuf-sync.c new file +8 -0
+		return ret;
+	}
+
+	sync->status = DMABUF_SYNC_LOCKED;
+
+	return ret;
+}
+EXPORT_SYMBOL(dmabuf_sync_lock);

I think you know what I am going to say.

Quoted 78 lines
drivers/base/dmabuf-sync.c new file +78 -0
+
+/**
+ * dmabuf_sync_unlock - unlock all objects added to syncs list.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * The caller should call this function after CPU or DMA access to
+ * the dmabufs is completed so that others can access the dmabufs.
+ */
+int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+{
+	if (!sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	/* If current dmabuf sync object wasn't reserved then just return. */
+	if (sync->status != DMABUF_SYNC_LOCKED)
+		return -EAGAIN;
+
+	dmabuf_sync_unlock_objs(sync, &sync->ctx);
+
+	return 0;
+}
+EXPORT_SYMBOL(dmabuf_sync_unlock);
+
+/**
+ * dmabuf_sync_single_lock - lock a dma buf.
+ *
+ * @dmabuf: A dma buf object that tries to lock.
+ * @type: A access type to a dma buf.
+ *	The DMA_BUF_ACCESS_R means that this dmabuf could be accessed by
+ *	others for read access. On the other hand, the DMA_BUF_ACCESS_W
+ *	means that this dmabuf couldn't be accessed by others but would be
+ *	accessed by caller's dma exclusively. And the DMA_BUF_ACCESS_DMA can
+ *	be combined with other.
+ * @wait: Indicate whether caller is blocked or not.
+ *	true means that caller will be blocked, and false means that this
+ *	function will return -EAGAIN if this caller can't take the lock
+ *	right now.
+ *
+ * The caller should call this function prior to CPU or DMA access to the dmabuf
+ * so that others cannot access the dmabuf.
+ */
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf, unsigned int type,
+				bool wait)
+{
+	struct dmabuf_sync_reservation *robj;
+
+	if (!dmabuf->sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	if (!IS_VALID_DMA_BUF_ACCESS_TYPE(type)) {
+		WARN_ON(1);
+		return -EINVAL;
+	}
+
+	get_dma_buf(dmabuf);
+	robj = dmabuf->sync;
+
+	mutex_lock(&robj->lock);
+
+	/* Don't lock in case of read and read. */
+	if (robj->accessed_type & DMA_BUF_ACCESS_R && type & DMA_BUF_ACCESS_R) {
+		atomic_inc(&robj->shared_cnt);
+		mutex_unlock(&robj->lock);
+		return 0;
+	}
+
+	/*
+	 * In case of F_SETLK, just return -EAGAIN if this dmabuf has already
+	 * been locked.
+	 */
+	if (!wait && robj->locked) {
+		mutex_unlock(&robj->lock);
+		dma_buf_put(dmabuf);

drivers/base/dmabuf-sync.c line 537

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+		return -EAGAIN;
+	}
+
+	mutex_unlock(&robj->lock);
+
+	mutex_lock(&robj->sync_lock.base);
+
+	mutex_lock(&robj->lock);
+	robj->locked = true;
+	mutex_unlock(&robj->lock);

Are you missing an mutex_unlock on &robj->sync_lock.base? Oh wait, that is the purpose of this code. You might want to put a nice comment right above that and say: "Unlocked by dmabuf_sync_single_unlock"

Quoted 199 lines
drivers/base/dmabuf-sync.c new file +141 -0
+
+	return 0;
+}
+EXPORT_SYMBOL(dmabuf_sync_single_lock);
+
+/**
+ * dmabuf_sync_single_unlock - unlock a dma buf.
+ *
+ * @dmabuf: A dma buf object that tries to unlock.
+ *
+ * The caller should call this function after CPU or DMA access to
+ * the dmabuf is completed so that others can access the dmabuf.
+ */
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_reservation *robj;
+
+	if (!dmabuf->sync) {
+		WARN_ON(1);
+		return;
+	}
+
+	robj = dmabuf->sync;
+
+	mutex_lock(&robj->lock);
+
+	if (robj->polled) {
+		robj->poll_event = true;
+		robj->polled = false;
+		wake_up_interruptible(&robj->poll_wait);
+	}
+
+	if (atomic_add_unless(&robj->shared_cnt, -1 , 1)) {
+		mutex_unlock(&robj->lock);
+		dma_buf_put(dmabuf);
+		return;
+	}
+
+	mutex_unlock(&robj->lock);
+
+	mutex_unlock(&robj->sync_lock.base);
+
+	mutex_lock(&robj->lock);
+	robj->locked = false;
+	mutex_unlock(&robj->lock);
+
+	dma_buf_put(dmabuf);
+
+	return;
+}
+EXPORT_SYMBOL(dmabuf_sync_single_unlock);
+
+/**
+ * dmabuf_sync_get - Get dmabuf sync object.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ * @sync_buf: A dmabuf object to be synchronized with others.
+ * @type: A access type to a dma buf.
+ *	The DMA_BUF_ACCESS_R means that this dmabuf could be accessed by
+ *	others for read access. On the other hand, the DMA_BUF_ACCESS_W
+ *	means that this dmabuf couldn't be accessed by others but would be
+ *	accessed by caller's dma exclusively. And the DMA_BUF_ACCESS_DMA can
+ *	be combined with other.
+ *
+ * This function should be called after dmabuf_sync_init function is called.
+ * The caller can tie up multiple dmabufs into one sync object by calling this
+ * function several times. Internally, this function allocates
+ * a dmabuf_sync_object and adds a given dmabuf to it, and also takes
+ * a reference to a dmabuf.
+ */
+int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf, unsigned int type)
+{
+	int ret;
+
+	if (!sync || !sync_buf) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	ret = dmabuf_sync_get_obj(sync, sync_buf, type);
+	if (ret < 0) {
+		WARN_ON(1);
+		return ret;
+	}
+
+	sync->status = DMABUF_SYNC_GOT;
+
+	return 0;
+}
+EXPORT_SYMBOL(dmabuf_sync_get);
+
+/**
+ * dmabuf_sync_put - Put dmabuf sync object to a given dmabuf.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ * @dmabuf: An dmabuf object.
+ *
+ * This function should be called if some operation is failed after
+ * dmabuf_sync_get function is called to release the dmabuf, or
+ * dmabuf_sync_unlock function is called. Internally, this function
+ * removes a given dmabuf from a sync object and remove the sync object.
+ * At this time, the dmabuf is putted.
+ */
+void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf *dmabuf)
+{
+	if (!sync || !dmabuf) {
+		WARN_ON(1);
+		return;
+	}
+
+	if (list_empty(&sync->syncs))
+		return;
+
+	dmabuf_sync_put_obj(sync, dmabuf);
+}
+EXPORT_SYMBOL(dmabuf_sync_put);
+
+/**
+ * dmabuf_sync_put_all - Put dmabuf sync object to dmabufs.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after
+ * dmabuf_sync_get function is called to release all sync objects, or
+ * dmabuf_sync_unlock function is called. Internally, this function
+ * removes dmabufs from a sync object and remove the sync object.
+ * At this time, all dmabufs are putted.
+ */
+void dmabuf_sync_put_all(struct dmabuf_sync *sync)
+{
+	if (!sync) {
+		WARN_ON(1);
+		return;
+	}
+
+	if (list_empty(&sync->syncs))
+		return;
+
+	dmabuf_sync_put_objs(sync);
+}
+EXPORT_SYMBOL(dmabuf_sync_put_all);
include/linux/dma-buf.h +16 -0
index dfac5ed..0109673 100644
@@ -115,6 +115,7 @@ struct dma_buf_ops {
  * @exp_name: name of the exporter; useful for debugging.
  * @list_node: node for dma_buf accounting and debugging.
  * @priv: exporter specific private data for this buffer object.
+ * @sync: sync object linked to this dma-buf
  */
 struct dma_buf {
 	size_t size;
@@ -128,6 +129,7 @@ struct dma_buf {
 	const char *exp_name;
 	struct list_head list_node;
 	void *priv;
+	void *sync;
 };
 
 /**
@@ -148,6 +150,20 @@ struct dma_buf_attachment {
 	void *priv;
 };
 
+#define	DMA_BUF_ACCESS_R	0x1
+#define DMA_BUF_ACCESS_W	0x2
+#define DMA_BUF_ACCESS_DMA	0x4
+#define DMA_BUF_ACCESS_RW	(DMA_BUF_ACCESS_R | DMA_BUF_ACCESS_W)
+#define DMA_BUF_ACCESS_DMA_R	(DMA_BUF_ACCESS_R | DMA_BUF_ACCESS_DMA)
+#define DMA_BUF_ACCESS_DMA_W	(DMA_BUF_ACCESS_W | DMA_BUF_ACCESS_DMA)
+#define DMA_BUF_ACCESS_DMA_RW	(DMA_BUF_ACCESS_DMA_R | DMA_BUF_ACCESS_DMA_W)
+#define IS_VALID_DMA_BUF_ACCESS_TYPE(t)	(t == DMA_BUF_ACCESS_R || \
+					 t == DMA_BUF_ACCESS_W || \
+					 t == DMA_BUF_ACCESS_DMA_R || \
+					 t == DMA_BUF_ACCESS_DMA_W || \
+					 t == DMA_BUF_ACCESS_RW || \
+					 t == DMA_BUF_ACCESS_DMA_RW)
+
 /**
  * get_dma_buf - convenience wrapper for get_file.
  * @dmabuf:	[in]	pointer to dma_buf
include/linux/dmabuf-sync.h new file +11 -0
new file mode 100644
index 0000000..9a3afc4
@@ -0,0 +1,190 @@
+/*
+ * Copyright (C) 2013 Samsung Electronics Co.Ltd
+ * Authors:
+ *	Inki Dae <[email protected]>
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ */

include/linux/dmabuf-sync.h line 21

Quoted 10 lines
include/linux/dmabuf-sync.h new file +10 -0
+
+#include <linux/mutex.h>
+#include <linux/sched.h>
+#include <linux/dma-buf.h>
+
+enum dmabuf_sync_status {
+	DMABUF_SYNC_GOT		= 1,
+	DMABUF_SYNC_LOCKED,
+};
+

No comment about this structure?

Quoted 9 lines
include/linux/dmabuf-sync.h new file +9 -0
+struct dmabuf_sync_reservation {
+	struct ww_mutex		sync_lock;
+	struct mutex		lock;
+	wait_queue_head_t	poll_wait;
+	unsigned int		poll_event;
+	unsigned int		polled;
+	atomic_t		shared_cnt;
+	unsigned int		accessed_type;
+	unsigned int		locked;

include/linux/dmabuf-sync.h line 40

Quoted 10 lines
include/linux/dmabuf-sync.h new file +10 -0
+};
+
+/*
+ * A structure for dmabuf_sync_object.
+ *
+ * @head: A list head to be added to syncs list.
+ * @robj: A reservation_object object.
+ * @dma_buf: A dma_buf object.
+ * @access_type: Indicate how a current task tries to access
+ *	a given buffer.

Huh? What values are expected then? Is there some #define or enum for that?

Quoted 8 lines
include/linux/dmabuf-sync.h new file +8 -0
+ */
+struct dmabuf_sync_object {
+	struct list_head		head;
+	struct dmabuf_sync_reservation	*robj;
+	struct dma_buf			*dmabuf;
+	unsigned int			access_type;
+};
+

include/linux/dmabuf-sync.h line 58

Quoted 10 lines
include/linux/dmabuf-sync.h new file +10 -0
+struct dmabuf_sync_priv_ops {
+	void (*free)(void *priv);
+};
+
+/*
+ * A structure for dmabuf_sync.
+ *
+ * @syncs: A list head to sync object and this is global to system.
+ * @list: A list entry used as committed list node
+ * @lock: A mutex lock to current sync object.

You should say for which specific operations this mutex is needed. For everything? Or just for list operations.

Quoted 6 lines
include/linux/dmabuf-sync.h new file +6 -0
+ * @ctx: A current context for ww mutex.
+ * @work: A work struct to release resources at timeout.
+ * @priv: A private data.
+ * @name: A string to dmabuf sync owner.
+ * @timer: A timer list to avoid lockup and release resources.
+ * @status: Indicate current status (DMABUF_SYNC_GOT or DMABUF_SYNC_LOCKED).

include/linux/dmabuf-sync.h line 74

Quoted 10 lines
include/linux/dmabuf-sync.h new file +10 -0
+ */
+struct dmabuf_sync {
+	struct list_head		syncs;
+	struct list_head		list;
+	struct mutex			lock;
+	struct ww_acquire_ctx		ctx;
+	struct work_struct		work;
+	void				*priv;
+	struct dmabuf_sync_priv_ops	*ops;
+	char				name[64];

Perhaps a #define for the size?

Quoted 123 lines
include/linux/dmabuf-sync.h new file +116 -0
+	struct timer_list		timer;
+	unsigned int			status;
+};
+
+#ifdef CONFIG_DMABUF_SYNC
+
+extern struct ww_class dmabuf_sync_ww_class;
+
+static inline void dmabuf_sync_reservation_init(struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_reservation *obj;
+
+	obj = kzalloc(sizeof(*obj), GFP_KERNEL);
+	if (!obj)
+		return;
+
+	dmabuf->sync = obj;
+
+	ww_mutex_init(&obj->sync_lock, &dmabuf_sync_ww_class);
+
+	mutex_init(&obj->lock);
+	atomic_set(&obj->shared_cnt, 1);
+
+	init_waitqueue_head(&obj->poll_wait);
+}
+
+static inline void dmabuf_sync_reservation_fini(struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_reservation *obj;
+
+	if (!dmabuf->sync)
+		return;
+
+	obj = dmabuf->sync;
+
+	ww_mutex_destroy(&obj->sync_lock);
+
+	kfree(obj);
+}
+
+extern bool is_dmabuf_sync_supported(void);
+
+extern struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void *priv);
+
+extern void dmabuf_sync_fini(struct dmabuf_sync *sync);
+
+extern int dmabuf_sync_lock(struct dmabuf_sync *sync);
+
+extern int dmabuf_sync_unlock(struct dmabuf_sync *sync);
+
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf, unsigned int type,
+				bool wait);
+
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf);
+
+extern int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf,
+				unsigned int type);
+
+extern void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf *dmabuf);
+
+extern void dmabuf_sync_put_all(struct dmabuf_sync *sync);
+
+#else
+
+static inline void dmabuf_sync_reservation_init(struct dma_buf *dmabuf) { }
+
+static inline void dmabuf_sync_reservation_fini(struct dma_buf *dmabuf) { }
+
+static inline bool is_dmabuf_sync_supported(void) { return false; }
+
+static inline  struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void *priv)
+{
+	return ERR_PTR(0);
+}
+
+static inline void dmabuf_sync_fini(struct dmabuf_sync *sync) { }
+
+static inline int dmabuf_sync_lock(struct dmabuf_sync *sync)
+{
+	return 0;
+}
+
+static inline int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+{
+	return 0;
+}
+
+static inline int dmabuf_sync_single_lock(struct dma_buf *dmabuf,
+						unsigned int type,
+						bool wait)
+{
+	return 0;
+}
+
+static inline void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+{
+	return;
+}
+
+static inline int dmabuf_sync_get(struct dmabuf_sync *sync,
+					void *sync_buf,
+					unsigned int type)
+{
+	return 0;
+}
+
+static inline void dmabuf_sync_put(struct dmabuf_sync *sync,
+					struct dma_buf *dmabuf) { }
+
+static inline void dmabuf_sync_put_all(struct dmabuf_sync *sync) { }
+
+#endif
Signature · 6 lines
1.7.5.4

_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel
author
#3 lore raw

RE: [PATCH 1/2] [RFC PATCH v6] dmabuf-sync: Add a buffer synchronization framework

Thanks for the review,
Inki Dae

Quoted 56 lines

-----Original Message-----
From: [email protected] [mailto:linux-fbdev- [email protected]] On Behalf Of Konrad Rzeszutek Wilk
Sent: Wednesday, August 21, 2013 4:22 AM
To: Inki Dae
Cc: [email protected]; [email protected]; linux- [email protected]; [email protected]; linaro- [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH 1/2] [RFC PATCH v6] dmabuf-sync: Add a buffer synchronization framework

Quoted 4 lines · Inki Dae

This patch adds a buffer synchronization framework based on DMA BUF[1] and and based on ww-mutexes[2] for lock mechanism.

The purpose of this framework is to provide not only buffer access

control

Quoted 15 lines

to CPU and DMA but also easy-to-use interfaces for device drivers and user application. This framework can be used for all dma devices using system memory as dma buffer, especially for most ARM based SoCs.

Changelog v6:
- Fix sync lock to multiple reads.
- Add select system call support.
. Wake up poll_wait when a dmabuf is unlocked.
- Remove unnecessary the use of mutex lock.
- Add private backend ops callbacks.
. This ops has one callback for device drivers to clean up their
sync object resource when the sync object is freed. For this, device drivers should implement the free callback properly.
- Update document file.

Quoted 2 lines

Changelog v5:
- Rmove a dependence on reservation_object: the reservation_object is

used

Quoted 2 lines

to hook up to ttm and dma-buf for easy sharing of reservations across devices. However, the dmabuf sync can be used for all dma devices;

v4l2

Quoted 6 lines

and drm based drivers, so doesn't need the reservation_object anymore. With regared to this, it adds 'void *sync' to dma_buf structure.
- All patches are rebased on mainline, Linux v3.10.

Changelog v4:
- Add user side interface for buffer synchronization mechanism and

update

Quoted 10 lines

descriptions related to the user side interface.

Changelog v3:
- remove cache operation relevant codes and update document file.

Changelog v2:
- use atomic_add_unless to avoid potential bug.
- add a macro for checking valid access type.
- code clean.

Quoted 7 lines
Quoted 2 lines

The mechanism of this framework has the following steps,
1. Register dmabufs to a sync object - A task gets a new sync object

and

Quoted 2 lines

can add one or more dmabufs that the task wants to access. This registering should be performed when a device context or an

event

Quoted 1 line

context such as a page flip event is created or before CPU accesses

a

Quoted 6 lines

shared

Quoted 5 lines

buffer.

dma_buf_sync_get(a sync object, a dmabuf);

2. Lock a sync object - A task tries to lock all dmabufs added in

its

Quoted 8 lines

own

Quoted 1 line

sync object. Basically, the lock mechanism uses ww-mutex[1] to avoid

dead

Quoted 1 line

lock issue and for race condition between CPU and CPU, CPU and DMA,

and DMA

Quoted 1 line

and DMA. Taking a lock means that others cannot access all locked

dmabufs

Quoted 1 line

until the task that locked the corresponding dmabufs, unlocks all

the

Quoted 55 lines

locked

Quoted 2 lines

dmabufs.
This locking should be performed before DMA or CPU accesses these

dmabufs.

Quoted 4 lines
	dma_buf_sync_lock(a sync object);

    3. Unlock a sync object - The task unlocks all dmabufs added in its

own sync

Quoted 1 line

object. The unlock means that the DMA or CPU accesses to the dmabufs

have

Quoted 2 lines

been completed so that others may access them.
This unlocking should be performed after DMA or CPU has completed

accesses

Quoted 5 lines

to the dmabufs.

dma_buf_sync_unlock(a sync object);

4. Unregister one or all dmabufs from a sync object - A task

unregisters

Quoted 1 line

the given dmabufs from the sync object. This means that the task

dosen't

Quoted 4 lines

want to lock the dmabufs.
The unregistering should be performed after DMA or CPU has completed accesses to the dmabufs or when dma_buf_sync_lock() is failed.

Quoted 8 lines

dma_buf_sync_put(a sync object, a dmabuf); dma_buf_sync_put_all(a sync object);

The described steps may be summarized as:
get -> lock -> CPU or DMA access to a buffer/s -> unlock -> put

This framework includes the following two features.
1. read (shared) and write (exclusive) locks - A task is required to

declare

Quoted 7 lines
    the access type when the task tries to register a dmabuf;
    READ, WRITE, READ DMA, or WRITE DMA.

    The below is example codes,
	struct dmabuf_sync *sync;

	sync = dmabuf_sync_init(...);
Quoted 8 lines
	...

	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_R);
	...

	And the below can be used as access types:
		DMA_BUF_ACCESS_R - CPU will access a buffer for read.
		DMA_BUF_ACCESS_W - CPU will access a buffer for read or

write.

Quoted 4 lines

DMA_BUF_ACCESS_DMA_R - DMA will access a buffer for read DMA_BUF_ACCESS_DMA_W - DMA will access a buffer for read or
write.

Quoted 9 lines
Quoted 1 line

2. Mandatory resource releasing - a task cannot hold a lock

indefinitely.

Quoted 1 line

A task may never try to unlock a buffer after taking a lock to the

buffer.

Quoted 1 line

In this case, a timer handler to the corresponding sync object is

called

Quoted 1 line

in five (default) seconds and then the timed-out buffer is unlocked

by work

Quoted 1 line

queue handler to avoid lockups and to enforce resources of the

buffer.

Quoted 198 lines
Quoted 14 lines

The below is how to use interfaces for device driver:

	1. Allocate and Initialize a sync object:
		static void xxx_dmabuf_sync_free(void *priv)
		{
			struct xxx_context *ctx = priv;

			if (!ctx)
				return;

			ctx->sync = NULL;
		}
		...
Quoted 8 lines
		static struct dmabuf_sync_priv_ops driver_specific_ops = {
			.free = xxx_dmabuf_sync_free,
		};
		...

		struct dmabuf_sync *sync;

		sync = dmabuf_sync_init("test sync", &driver_specific_ops,

ctx);

Quoted 3 lines
		...

	2. Add a dmabuf to the sync object when setting up dma buffer

relevant

Quoted 4 lines
	   registers:
		dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_READ);
		...
Quoted 8 lines
	3. Lock all dmabufs of the sync object before DMA or CPU accesses
	   the dmabufs:
		dmabuf_sync_lock(sync);
		...

	4. Now CPU or DMA can access all dmabufs locked in step 3.

	5. Unlock all dmabufs added in a sync object after DMA or CPU

access

Quoted 7 lines

to these dmabufs is completed:
dmabuf_sync_unlock(sync);

And call the following functions to release all resources,
dmabuf_sync_put_all(sync); dmabuf_sync_fini(sync);

Quoted 8 lines

You can refer to actual example codes:
"drm/exynos: add dmabuf sync support for g2d driver" and "drm/exynos: add dmabuf sync support for kms framework" from
https://git.kernel.org/cgit/linux/kernel/git/daeinki/
drm-exynos.git/log/?h=dmabuf-sync

And this framework includes fcntl system call[3] as interfaces exported to user. As you know, user sees a buffer object as a dma-buf file

descriptor.

Quoted 1 line

So fcntl() call with the file descriptor means to lock some buffer

region being

Quoted 18 lines

managed by the dma-buf object.

The below is how to use interfaces for user application:

fcntl system call:

	struct flock filelock;

	1. Lock a dma buf:
		filelock.l_type = F_WRLCK or F_RDLCK;

		/* lock entire region to the dma buf. */
		filelock.lwhence = SEEK_CUR;
		filelock.l_start = 0;
		filelock.l_len = 0;

		fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
		...
Quoted 8 lines
		CPU access to the dma buf

	2. Unlock a dma buf:
		filelock.l_type = F_UNLCK;

		fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);

		close(dmabuf fd) call would also unlock the dma buf. And for

more

Quoted 24 lines

detail, please refer to [3]

select system call:

	fd_set wdfs or rdfs;

	FD_ZERO(&wdfs or &rdfs);
	FD_SET(fd, &wdfs or &rdfs);

	select(fd + 1, &rdfs, NULL, NULL, NULL);
		or
	select(fd + 1, NULL, &wdfs, NULL, NULL);

	Every time select system call is called, a caller will wait for
	the completion of DMA or CPU access to a shared buffer if there
	is someone accessing the shared buffer; locked the shared buffer.
	However, if no anyone then select system call will be returned
	at once.

References:
[1] http://lwn.net/Articles/470339/
[2] https://patchwork.kernel.org/patch/2625361/
[3] http://linux.die.net/man/2/fcntl

Quoted 8 lines
---
 Documentation/dma-buf-sync.txt |  285 +++++++++++++++++
 drivers/base/Kconfig           |    7 +
 drivers/base/Makefile          |    1 +
 drivers/base/dma-buf.c         |    4 +
 drivers/base/dmabuf-sync.c     |  678

++++++++++++++++++++++++++++++++++++++++

Documentation/dma-buf-

Quoted 8 lines
 include/linux/dma-buf.h        |   16 +
 include/linux/dmabuf-sync.h    |  190 +++++++++++
 7 files changed, 1181 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/dma-buf-sync.txt
 create mode 100644 drivers/base/dmabuf-sync.c
 create mode 100644 include/linux/dmabuf-sync.h
Documentation/dma-buf- renamed from Documentation/dma-buf-sync.txt +0 -0

sync.txt

Documentation/dma-buf-sync.txt line 8

Quoted 13 lines

new file mode 100644
index 0000000..8023d06

Documentation/dma-buf-sync.txt new file +8 -0
@@ -0,0 +1,285 @@
+                    DMA Buffer Synchronization Framework
+                    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+                                  Inki Dae
+                      <inki dot dae at samsung dot com>
+                          <daeinki at gmail dot com>
+
+This document is a guide for device-driver writers describing the DMA

buffer

Documentation/dma-buf-sync.txt line 10

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+synchronization API. This document also describes how to use the API to
+use buffer synchronization mechanism between DMA and DMA, CPU and DMA,

and

Documentation/dma-buf-sync.txt

Quoted 3 lines
Documentation/dma-buf-sync.txt new file +3 -0
+CPU and CPU.
+
+The DMA Buffer synchronization API provides buffer synchronization

mechanism;

Documentation/dma-buf-sync.txt line 14

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+i.e., buffer access control to CPU and DMA, and easy-to-use interfaces

for

Documentation/dma-buf-sync.txt line 15

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+device drivers and user application. And this API can be used for all

dma

Documentation/dma-buf-sync.txt line 16

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+devices using system memory as dma buffer, especially for most ARM

based SoCs.

Documentation/dma-buf-sync.txt

Quoted 8 lines
Documentation/dma-buf-sync.txt new file +8 -0
+
+
+Motivation
+----------
+
+Buffer synchronization issue between DMA and DMA:
+	Sharing a buffer, a device cannot be aware of when the other device
+	will access the shared buffer: a device may access a buffer

containing

Documentation/dma-buf-sync.txt

Quoted 3 lines
Documentation/dma-buf-sync.txt new file +3 -0
+	wrong data if the device accesses the shared buffer while another
+	device is still accessing the shared buffer.
+	Therefore, a user process should have waited for the completion of

DMA

Documentation/dma-buf-sync.txt

Quoted 5 lines
Documentation/dma-buf-sync.txt new file +5 -0
+	access by another device before a device tries to access the shared
+	buffer.
+
+Buffer synchronization issue between CPU and DMA:
+	A user process should consider that when having to send a buffer,

filled

Documentation/dma-buf-sync.txt line 33

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+	by CPU, to a device driver for the device driver to access the

buffer as

Documentation/dma-buf-sync.txt

Quoted 8 lines
Documentation/dma-buf-sync.txt new file +8 -0
+	a input buffer while CPU and DMA are sharing the buffer.
+	This means that the user process needs to understand how the device
+	driver is worked. Hence, the conventional mechanism not only makes
+	user application complicated but also incurs performance overhead.
+
+Buffer synchronization issue between CPU and CPU:
+	In case that two processes share one buffer; shared with DMA also,
+	they may need some mechanism to allow process B to access the

shared

Documentation/dma-buf-sync.txt line 43

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+	buffer after the completion of CPU access by process A.
+	Therefore, process B should have waited for the completion of CPU

access

Documentation/dma-buf-sync.txt

Quoted 7 lines
Documentation/dma-buf-sync.txt new file +7 -0
+	by process A using the mechanism before trying to access the shared
+	buffer.
+
+What is the best way to solve these buffer synchronization issues?
+	We may need a common object that a device driver and a user process
+	notify the common object of when they try to access a shared buffer.
+	That way we could decide when we have to allow or not to allow for

CPU

Documentation/dma-buf-sync.txt line 52

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+	or DMA to access the shared buffer through the common object.
+	If so, what could become the common object? Right, that's a dma-

buf[1].

Quoted 9 lines

Documentation/dma-buf-sync.txt

Quoted 9 lines
Documentation/dma-buf-sync.txt new file +9 -0
+	Now we have already been using the dma-buf to share one buffer with
+	other drivers.
+
+
+Basic concept
+-------------
+
+The mechanism of this framework has the following steps,
+    1. Register dmabufs to a sync object - A task gets a new sync

object

Quoted 20 lines

and

Documentation/dma-buf-sync.txt line 63

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+    can add one or more dmabufs that the task wants to access.
+    This registering should be performed when a device context or an

event

Documentation/dma-buf-sync.txt line 64

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    context such as a page flip event is created or before CPU accesses

a shared

Documentation/dma-buf-sync.txt

Quoted 5 lines
Documentation/dma-buf-sync.txt new file +5 -0
+    buffer.
+
+	dma_buf_sync_get(a sync object, a dmabuf);
+
+    2. Lock a sync object - A task tries to lock all dmabufs added in

its own

Documentation/dma-buf-sync.txt line 70

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    sync object. Basically, the lock mechanism uses ww-mutexes[2] to

avoid dead

Documentation/dma-buf-sync.txt line 71

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    lock issue and for race condition between CPU and CPU, CPU and DMA,

and DMA

Documentation/dma-buf-sync.txt line 72

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    and DMA. Taking a lock means that others cannot access all locked

dmabufs

Documentation/dma-buf-sync.txt line 73

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    until the task that locked the corresponding dmabufs, unlocks all

the locked

Quoted 9 lines

Documentation/dma-buf-sync.txt line 75

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+    dmabufs.
+    This locking should be performed before DMA or CPU accesses these

dmabufs.

Documentation/dma-buf-sync.txt

Quoted 4 lines
Documentation/dma-buf-sync.txt new file +4 -0
+
+	dma_buf_sync_lock(a sync object);
+
+    3. Unlock a sync object - The task unlocks all dmabufs added in its

own sync

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    object. The unlock means that the DMA or CPU accesses to the

dmabufs

Quoted 4 lines

have

Documentation/dma-buf-sync.txt line 82

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+    been completed so that others may access them.
+    This unlocking should be performed after DMA or CPU has completed

accesses

Quoted 10 lines

Documentation/dma-buf-sync.txt

Quoted 5 lines
Documentation/dma-buf-sync.txt new file +5 -0
+    to the dmabufs.
+
+	dma_buf_sync_unlock(a sync object);
+
+    4. Unregister one or all dmabufs from a sync object - A task

unregisters

Documentation/dma-buf-sync.txt line 88

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    the given dmabufs from the sync object. This means that the task

dosen't

Quoted 2 lines
Documentation/dma-buf-sync.txt new file +2 -0
+    want to lock the dmabufs.
+    The unregistering should be performed after DMA or CPU has

completed

Quoted 10 lines

Documentation/dma-buf-sync.txt

Quoted 10 lines
Documentation/dma-buf-sync.txt new file +10 -0
+    accesses to the dmabufs or when dma_buf_sync_lock() is failed.
+
+	dma_buf_sync_put(a sync object, a dmabuf);
+	dma_buf_sync_put_all(a sync object);
+
+    The described steps may be summarized as:
+	get -> lock -> CPU or DMA access to a buffer/s -> unlock -> put
+
+This framework includes the following two features.
+    1. read (shared) and write (exclusive) locks - A task is required

to

Quoted 323 lines

declare

Documentation/dma-buf-sync.txt

Quoted 12 lines
Documentation/dma-buf-sync.txt new file +12 -0
+    the access type when the task tries to register a dmabuf;
+    READ, WRITE, READ DMA, or WRITE DMA.
+
+    The below is example codes,
+	struct dmabuf_sync *sync;
+
+	sync = dmabuf_sync_init(NULL, "test sync");
+
+	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_R);
+	...
+
+    2. Mandatory resource releasing - a task cannot hold a lock

indefinitely.

Documentation/dma-buf-sync.txt line 113

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    A task may never try to unlock a buffer after taking a lock to the

buffer.

Documentation/dma-buf-sync.txt line 114

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    In this case, a timer handler to the corresponding sync object is

called

Documentation/dma-buf-sync.txt line 115

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    in five (default) seconds and then the timed-out buffer is unlocked

by work

Documentation/dma-buf-sync.txt line 116

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+    queue handler to avoid lockups and to enforce resources of the

buffer.

Quoted 6 lines
Documentation/dma-buf-sync.txt new file +6 -0
+
+
+Access types
+------------
+
+DMA_BUF_ACCESS_R - CPU will access a buffer for read.

Documentation/dma-buf-sync.txt

Quoted 10 lines
Documentation/dma-buf-sync.txt new file +10 -0
+DMA_BUF_ACCESS_W - CPU will access a buffer for read or write.
+DMA_BUF_ACCESS_DMA_R - DMA will access a buffer for read
+DMA_BUF_ACCESS_DMA_W - DMA will access a buffer for read or write.
+
+
+Generic user interfaces
+-----------------------
+
+And this framework includes fcntl system call[3] as interfaces exported
+to user. As you know, user sees a buffer object as a dma-buf file

descriptor.

Documentation/dma-buf-sync.txt line 133

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+So fcntl() call with the file descriptor means to lock some buffer

region being

Documentation/dma-buf-sync.txt

Quoted 13 lines
Documentation/dma-buf-sync.txt new file +13 -0
+managed by the dma-buf object.
+
+
+API set
+-------
+
+bool is_dmabuf_sync_supported(void)
+	- Check if dmabuf sync is supported or not.
+
+struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void priv*)
+	- Allocate and initialize a new sync object. The caller can get a

new

Documentation/dma-buf-sync.txt line 147

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+	sync object for buffer synchronization. ops is used for device

driver

Documentation/dma-buf-sync.txt line 148

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+	to clean up its own sync object. For this, each device driver

should

Documentation/dma-buf-sync.txt line 149

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+	implement a free callback. priv is used for device driver to get

its

Quoted 4 lines
Documentation/dma-buf-sync.txt new file +4 -0
+	device context when free callback is called.
+
+void dmabuf_sync_fini(struct dmabuf_sync *sync)
+	- Release all resources to the sync object.

Documentation/dma-buf-sync.txt

Quoted 10 lines
Documentation/dma-buf-sync.txt new file +10 -0
+
+int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf,
+			unsigned int type)
+	- Get dmabuf sync object. Internally, this function allocates
+	a dmabuf_sync object and adds a given dmabuf to it, and also takes
+	a reference to the dmabuf. The caller can tie up multiple dmabufs
+	into one sync object by calling this function several times.
+
+void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf *dmabuf)
+	- Put dmabuf sync object to a given dmabuf. Internally, this

function

Documentation/dma-buf-sync.txt line 164

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+	removes a given dmabuf from a sync object and remove the sync

object.

Documentation/dma-buf-sync.txt

Quoted 4 lines
Documentation/dma-buf-sync.txt new file +4 -0
+	At this time, the dmabuf is putted.
+
+void dmabuf_sync_put_all(struct dmabuf_sync *sync)
+	- Put dmabuf sync object to dmabufs. Internally, this function

removes

Documentation/dma-buf-sync.txt

Quoted 5 lines
Documentation/dma-buf-sync.txt new file +5 -0
+	all dmabufs from a sync object and remove the sync object.
+	At this time, all dmabufs are putted.
+
+int dmabuf_sync_lock(struct dmabuf_sync *sync)
+	- Lock all dmabufs added in a sync object. The caller should call

this

Documentation/dma-buf-sync.txt line 174

Quoted 1 line
Documentation/dma-buf-sync.txt new file +1 -0
+	function prior to CPU or DMA access to the dmabufs so that others

can

Documentation/dma-buf-sync.txt

Quoted 6 lines
Documentation/dma-buf-sync.txt new file +6 -0
+	not access the dmabufs. Internally, this function avoids dead lock
+	issue with ww-mutexes.
+
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf)
+	- Lock a dmabuf. The caller should call this
+	function prior to CPU or DMA access to the dmabuf so that others

can

Documentation/dma-buf-sync.txt

Quoted 5 lines
Documentation/dma-buf-sync.txt new file +5 -0
+	not access the dmabuf.
+
+int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+	- Unlock all dmabufs added in a sync object. The caller should call
+	this function after CPU or DMA access to the dmabufs is completed

so

Documentation/dma-buf-sync.txt

Quoted 4 lines
Documentation/dma-buf-sync.txt new file +4 -0
+	that others can access the dmabufs.
+
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+	- Unlock a dmabuf. The caller should call this function after CPU

or

Quoted 20 lines
Documentation/dma-buf-sync.txt new file +20 -0
+	DMA access to the dmabuf is completed so that others can access
+	the dmabuf.
+
+
+Tutorial for device driver
+--------------------------
+
+1. Allocate and Initialize a sync object:
+	static void xxx_dmabuf_sync_free(void *priv)
+	{
+		struct xxx_context *ctx = priv;
+
+		if (!ctx)
+			return;
+
+		ctx->sync = NULL;
+	}
+	...
+
+	static struct dmabuf_sync_priv_ops driver_specific_ops = {

Documentation/dma-buf-sync.txt

Quoted 10 lines
Documentation/dma-buf-sync.txt new file +10 -0
+		.free = xxx_dmabuf_sync_free,
+	};
+	...
+
+	struct dmabuf_sync *sync;
+
+	sync = dmabuf_sync_init("test sync", &driver_specific_ops, ctx);
+	...
+
+2. Add a dmabuf to the sync object when setting up dma buffer relevant

registers:

Documentation/dma-buf-sync.txt

Quoted 4 lines
Documentation/dma-buf-sync.txt new file +4 -0
+	dmabuf_sync_get(sync, dmabuf, DMA_BUF_ACCESS_READ);
+	...
+
+3. Lock all dmabufs of the sync object before DMA or CPU accesses the

dmabufs:

Documentation/dma-buf-sync.txt

Quoted 6 lines
Documentation/dma-buf-sync.txt new file +6 -0
+	dmabuf_sync_lock(sync);
+	...
+
+4. Now CPU or DMA can access all dmabufs locked in step 3.
+
+5. Unlock all dmabufs added in a sync object after DMA or CPU access to

these

Quoted 61 lines
Documentation/dma-buf-sync.txt new file +56 -0
+   dmabufs is completed:
+	dmabuf_sync_unlock(sync);
+
+   And call the following functions to release all resources,
+	dmabuf_sync_put_all(sync);
+	dmabuf_sync_fini(sync);
+
+
+Tutorial for user application
+-----------------------------
+fcntl system call:
+
+	struct flock filelock;
+
+1. Lock a dma buf:
+	filelock.l_type = F_WRLCK or F_RDLCK;
+
+	/* lock entire region to the dma buf. */
+	filelock.lwhence = SEEK_CUR;
+	filelock.l_start = 0;
+	filelock.l_len = 0;
+
+	fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
+	...
+	CPU access to the dma buf
+
+2. Unlock a dma buf:
+	filelock.l_type = F_UNLCK;
+
+	fcntl(dmabuf fd, F_SETLKW or F_SETLK, &filelock);
+
+	close(dmabuf fd) call would also unlock the dma buf. And for more
+	detail, please refer to [3]
+
+
+select system call:
+
+	fd_set wdfs or rdfs;
+
+	FD_ZERO(&wdfs or &rdfs);
+	FD_SET(fd, &wdfs or &rdfs);
+
+	select(fd + 1, &rdfs, NULL, NULL, NULL);
+		or
+	select(fd + 1, NULL, &wdfs, NULL, NULL);
+
+	Every time select system call is called, a caller will wait for
+	the completion of DMA or CPU access to a shared buffer if there is
+	someone accessing the shared buffer; locked the shared buffer.
+	However, if no anyone then select system call will be returned
+	at once.
+
+References:
+[1] http://lwn.net/Articles/470339/
+[2] https://patchwork.kernel.org/patch/2625361/
+[3] http://linux.die.net/man/2/fcntl
drivers/base/Kconfig +0 -0
index 5daa259..35e1518 100644

drivers/base/Kconfig config DMA_SHARED_BUFFER · line 200

Quoted 1 line
drivers/base/Kconfig +0 -0
@@ -200,6 +200,13 @@ config DMA_SHARED_BUFFER
 	  APIs extension; the file's descriptor can then be passed on to

other

drivers/base/Kconfig config DMA_SHARED_BUFFER

Quoted 7 lines
drivers/base/Kconfig +5 -0
 	  driver.
 
+config DMABUF_SYNC
+	bool "DMABUF Synchronization Framework"
+	depends on DMA_SHARED_BUFFER
+	help
+	  This option enables dmabuf sync framework for buffer

synchronization between

drivers/base/dma-buf.c static int dma_buf_release(struct inode *inode, struct

Quoted 30 lines
drivers/base/Kconfig +2 -0
+	  DMA and DMA, CPU and DMA, and CPU and CPU.
+
 config CMA
 	bool "Contiguous Memory Allocator"
 	depends on HAVE_DMA_CONTIGUOUS && HAVE_MEMBLOCK
drivers/base/Makefile +1 -0
index 48029aa..e06a5d7 100644
@@ -11,6 +11,7 @@ obj-y			+= power/
 obj-$(CONFIG_HAS_DMA)	+= dma-mapping.o
 obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o
 obj-$(CONFIG_DMA_SHARED_BUFFER) += dma-buf.o reservation.o
+obj-$(CONFIG_DMABUF_SYNC) += dmabuf-sync.o
 obj-$(CONFIG_ISA)	+= isa.o
 obj-$(CONFIG_FW_LOADER)	+= firmware_class.o
 obj-$(CONFIG_NUMA)	+= node.o
drivers/base/dma-buf.c +1 -0
index 6687ba7..4aca57a 100644
@@ -29,6 +29,7 @@
 #include <linux/export.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
+#include <linux/dmabuf-sync.h>
 
 static inline int is_dma_buf_file(struct file *);
 
@@ -56,6 +57,8 @@ static int dma_buf_release(struct inode *inode, struct

file *file)

drivers/base/dma-buf.c struct dma_buf *dma_buf_export_named(void *priv,

Quoted 9 lines
drivers/base/dma-buf.c +2 -0
 	list_del(&dmabuf->list_node);
 	mutex_unlock(&db_list.lock);
 
+	dmabuf_sync_reservation_fini(dmabuf);
+
 	kfree(dmabuf);
 	return 0;
 }
@@ -134,6 +137,7 @@ struct dma_buf *dma_buf_export_named(void *priv,

const struct dma_buf_ops *ops,

Quoted 13 lines
drivers/base/dma-buf.c +1 -0
 	file = anon_inode_getfile("dmabuf", &dma_buf_fops, dmabuf, flags);
 
+	dmabuf_sync_reservation_init(dmabuf);
 	dmabuf->file = file;
 
 	mutex_init(&dmabuf->lock);
drivers/base/dmabuf-sync.c new file +0 -0
new file mode 100644
index 0000000..fbe711c

drivers/base/dmabuf-sync.c line 6

Quoted 6 lines
drivers/base/dmabuf-sync.c new file +6 -0
@@ -0,0 +1,678 @@
+/*
+ * Copyright (C) 2013 Samsung Electronics Co.Ltd
+ * Authors:
+ *	Inki Dae <[email protected]>
+ *
+ * This program is free software; you can redistribute  it and/or

modify it

drivers/base/dmabuf-sync.c line 7

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+ * under  the terms of  the GNU General  Public License as published by

the

drivers/base/dmabuf-sync.c line 8

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+ * Free Software Foundation;  either version 2 of the  License, or (at

your

Quoted 15 lines
drivers/base/dmabuf-sync.c new file +15 -0
+ * option) any later version.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/debugfs.h>
+#include <linux/uaccess.h>
+
+#include <linux/dmabuf-sync.h>
+
+#define MAX_SYNC_TIMEOUT	5 /* Second. */
+
+int dmabuf_sync_enabled = 1;

drivers/base/dmabuf-sync.c

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+MODULE_PARM_DESC(enabled, "Check if dmabuf sync is supported or not");
+module_param_named(enabled, dmabuf_sync_enabled, int, 0444);
+
+DEFINE_WW_CLASS(dmabuf_sync_ww_class);
+EXPORT_SYMBOL(dmabuf_sync_ww_class);
+
+static void dmabuf_sync_timeout_worker(struct work_struct *work)
+{
+	struct dmabuf_sync *sync = container_of(work, struct dmabuf_sync,

work);

drivers/base/dmabuf-sync.c line 38

Quoted 5 lines
drivers/base/dmabuf-sync.c new file +5 -0
+	struct dmabuf_sync_object *sobj;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry(sobj, &sync->syncs, head) {

You are using the 'sobj->robj' quite a lot. Why not just use a temp structure:

Quoted 3 lines
		struct dmabuf_sync_reservation *rsvp = sobj->robj;

and use that in this function. It would make it easier to read I think.

Ok, will use the temp structure.

Quoted 12 lines

drivers/base/dmabuf-sync.c line 50

Quoted 12 lines
drivers/base/dmabuf-sync.c new file +10 -0
+		BUG_ON(!sobj->robj);
+
+		mutex_lock(&sobj->robj->lock);
+
+		printk(KERN_WARNING "%s: timeout = 0x%x [type = %d:%d, " \
+					"refcnt = %d, locked = %d]\n",
+					sync->name, (u32)sobj->dmabuf,
+					sobj->robj->accessed_type,
+					sobj->access_type,
+

atomic_read(&sobj->robj->shared_cnt),

Quoted 3 lines

drivers/base/dmabuf-sync.c line 49

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+					sobj->robj->locked);

pr_warn_ratelimited?

Will use pr_warn because the timeout worker handler isn't called so frequently so the printk storm wouldn't be caused

Quoted 26 lines
Quoted 26 lines
drivers/base/dmabuf-sync.c new file +25 -0
+
+		/* unlock only valid sync object. */
+		if (!sobj->robj->locked) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		if (sobj->robj->polled) {
+			sobj->robj->poll_event = true;
+			sobj->robj->polled = false;
+			wake_up_interruptible(&sobj->robj->poll_wait);
+		}
+
+		if (atomic_add_unless(&sobj->robj->shared_cnt, -1, 1)) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		mutex_unlock(&sobj->robj->lock);
+
+		ww_mutex_unlock(&sobj->robj->sync_lock);
+
+		mutex_lock(&sobj->robj->lock);
+		sobj->robj->locked = false;
+
Quoted 8 lines

drivers/base/dmabuf-sync.c line 80

Quoted 6 lines
drivers/base/dmabuf-sync.c new file +6 -0
+		if (sobj->access_type & DMA_BUF_ACCESS_R)
+			printk(KERN_WARNING "%s: r-unlocked = 0x%x\n",
+					sync->name, (u32)sobj->dmabuf);
+		else
+			printk(KERN_WARNING "%s: w-unlocked = 0x%x\n",
+					sync->name, (u32)sobj->dmabuf);

How about using 'pr_warn'? And in it have:

Ok, will use it.

Quoted 6 lines
		sobj->access_type & DMA_BUF_ACCESS_R ? "r-" : "w-",

	and just have one printk.

Why the (u32) casting? Don't you want %p ?

Right, I should had used %p instead. Will remove the casting and use %p instead.

Quoted 63 lines
Quoted 50 lines
drivers/base/dmabuf-sync.c new file +49 -0
+
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	sync->status = 0;
+	mutex_unlock(&sync->lock);
+
+	dmabuf_sync_put_all(sync);
+	dmabuf_sync_fini(sync);
+}
+
+static void dmabuf_sync_lock_timeout(unsigned long arg)
+{
+	struct dmabuf_sync *sync = (struct dmabuf_sync *)arg;
+
+	schedule_work(&sync->work);
+}
+
+static int dmabuf_sync_lock_objs(struct dmabuf_sync *sync,
+					struct ww_acquire_ctx *ctx)
+{
+	struct dmabuf_sync_object *contended_sobj = NULL;
+	struct dmabuf_sync_object *res_sobj = NULL;
+	struct dmabuf_sync_object *sobj = NULL;
+	int ret;
+
+	if (ctx)
+		ww_acquire_init(ctx, &dmabuf_sync_ww_class);
+
+retry:
+	list_for_each_entry(sobj, &sync->syncs, head) {
+		if (WARN_ON(!sobj->robj))
+			continue;
+
+		mutex_lock(&sobj->robj->lock);
+
+		/* Don't lock in case of read and read. */
+		if (sobj->robj->accessed_type & DMA_BUF_ACCESS_R &&
+		    sobj->access_type & DMA_BUF_ACCESS_R) {
+			atomic_inc(&sobj->robj->shared_cnt);
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		if (sobj == res_sobj) {
+			res_sobj = NULL;
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}

drivers/base/dmabuf-sync.c line 139

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+		mutex_unlock(&sobj->robj->lock);
+
+		ret = ww_mutex_lock(&sobj->robj->sync_lock, ctx);
+		if (ret < 0) {
+			contended_sobj = sobj;
+
+			if (ret == -EDEADLK)
+				printk(KERN_WARNING"%s: deadlock = 0x%x\n",
+					sync->name, (u32)sobj->dmabuf);

Again, why (u32) and not %p?

Quoted 5 lines

drivers/base/dmabuf-sync.c line 140

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+			goto err;

This looks odd. You jump to err, which jumps back to 'retry'. Won't this cause an infinite loop? Perhaps you need to add a retry counter to only do this up to five times or so and then give up?

It jumps to err only if ww_mutex_lock returns -EDEADLK. This means that the lock trying to a given sync object caused dead lock. So all robjs already locked should be unlocked, and retried to take lock again going to err. So I think the infinite loop isn't caused.

Quoted 115 lines
Quoted 32 lines
drivers/base/dmabuf-sync.c new file +31 -0
+		}
+
+		mutex_lock(&sobj->robj->lock);
+		sobj->robj->locked = true;
+
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	if (ctx)
+		ww_acquire_done(ctx);
+
+	init_timer(&sync->timer);
+
+	sync->timer.data = (unsigned long)sync;
+	sync->timer.function = dmabuf_sync_lock_timeout;
+	sync->timer.expires = jiffies + (HZ * MAX_SYNC_TIMEOUT);
+
+	add_timer(&sync->timer);
+
+	return 0;
+
+err:
+	list_for_each_entry_continue_reverse(sobj, &sync->syncs, head) {
+		mutex_lock(&sobj->robj->lock);
+
+		/* Don't need to unlock in case of read and read. */
+		if (atomic_add_unless(&sobj->robj->shared_cnt, -1, 1)) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+

drivers/base/dmabuf-sync.c

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+		ww_mutex_unlock(&sobj->robj->sync_lock);
+		sobj->robj->locked = false;
+
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	if (res_sobj) {
+		mutex_lock(&res_sobj->robj->lock);
+
+		if (!atomic_add_unless(&res_sobj->robj->shared_cnt, -1, 1))

{

Quoted 59 lines
drivers/base/dmabuf-sync.c new file +59 -0
+			ww_mutex_unlock(&res_sobj->robj->sync_lock);
+			res_sobj->robj->locked = false;
+		}
+
+		mutex_unlock(&res_sobj->robj->lock);
+	}
+
+	if (ret == -EDEADLK) {
+		ww_mutex_lock_slow(&contended_sobj->robj->sync_lock, ctx);
+		res_sobj = contended_sobj;
+
+		goto retry;
+	}
+
+	if (ctx)
+		ww_acquire_fini(ctx);
+
+	return ret;
+}
+
+static void dmabuf_sync_unlock_objs(struct dmabuf_sync *sync,
+					struct ww_acquire_ctx *ctx)
+{
+	struct dmabuf_sync_object *sobj;
+
+	if (list_empty(&sync->syncs))
+		return;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry(sobj, &sync->syncs, head) {
+		mutex_lock(&sobj->robj->lock);
+
+		if (sobj->robj->polled) {
+			sobj->robj->poll_event = true;
+			sobj->robj->polled = false;
+			wake_up_interruptible(&sobj->robj->poll_wait);
+		}
+
+		if (atomic_add_unless(&sobj->robj->shared_cnt, -1, 1)) {
+			mutex_unlock(&sobj->robj->lock);
+			continue;
+		}
+
+		mutex_unlock(&sobj->robj->lock);
+
+		ww_mutex_unlock(&sobj->robj->sync_lock);
+
+		mutex_lock(&sobj->robj->lock);
+		sobj->robj->locked = false;
+		mutex_unlock(&sobj->robj->lock);
+	}
+
+	mutex_unlock(&sync->lock);
+
+	if (ctx)
+		ww_acquire_fini(ctx);
+
+	del_timer(&sync->timer);

drivers/base/dmabuf-sync.c line 250

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+}
+
+/**
+ * is_dmabuf_sync_supported - Check if dmabuf sync is supported or not.
+ */
+bool is_dmabuf_sync_supported(void)
+{
+	return dmabuf_sync_enabled == 1;
+}
+EXPORT_SYMBOL(is_dmabuf_sync_supported);

_GPL ?

Quoted 3 lines

I would also prefix it with 'dmabuf_is_sync_supported' just to make all of the libraries call start with 'dmabuf'

Seems better. Will change it to dmabuf_is_sync_supported, and use EXPORT_SYMBOL_GPL.

Quoted 21 lines
Quoted 21 lines
drivers/base/dmabuf-sync.c new file +21 -0
+
+/**
+ * dmabuf_sync_init - Allocate and initialize a dmabuf sync.
+ *
+ * @priv: A device private data.
+ * @name: A sync object name.
+ *
+ * This function should be called when a device context or an event
+ * context such as a page flip event is created. And the created
+ * dmabuf_sync object should be set to the context.
+ * The caller can get a new sync object for buffer synchronization
+ * through this function.
+ */
+struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void *priv)
+{
+	struct dmabuf_sync *sync;
+
+	sync = kzalloc(sizeof(*sync), GFP_KERNEL);
+	if (!sync)
Quoted 9 lines

drivers/base/dmabuf-sync.c line 275

Quoted 4 lines
drivers/base/dmabuf-sync.c new file +4 -0
+		return ERR_PTR(-ENOMEM);
+
+	strncpy(sync->name, name, ARRAY_SIZE(sync->name) - 1);
+

That is odd usage of an ARRAY_SIZE, but I can see how you can use it. I would say you should just do a #define for the 64 line and use that instead.

Ok, will use the macro instead.

Quoted 11 lines

drivers/base/dmabuf-sync.c line 284

Quoted 9 lines
drivers/base/dmabuf-sync.c new file +9 -0
+	sync->ops = ops;
+	sync->priv = priv;
+	INIT_LIST_HEAD(&sync->syncs);
+	mutex_init(&sync->lock);
+	INIT_WORK(&sync->work, dmabuf_sync_timeout_worker);
+
+	return sync;
+}
+EXPORT_SYMBOL(dmabuf_sync_init);

_GPL ?

Sure.

Quoted 13 lines
Quoted 13 lines
drivers/base/dmabuf-sync.c new file +13 -0
+
+/**
+ * dmabuf_sync_fini - Release a given dmabuf sync.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after
+ * dmabuf_sync_init call to release relevant resources, and after
+ * dmabuf_sync_unlock function is called.
+ */
+void dmabuf_sync_fini(struct dmabuf_sync *sync)
+{
+	if (WARN_ON(!sync))
Quoted 8 lines

drivers/base/dmabuf-sync.c line 302

Quoted 5 lines
drivers/base/dmabuf-sync.c new file +5 -0
+		return;
+
+	if (sync->ops && sync->ops->free)
+		sync->ops->free(sync->priv);
+

No need to cancel the sync->work in case that is still running?

Right, the locks to all buffers should be canceled if dmabuf_sync_fini was called without unlock call.

Quoted 6 lines

drivers/base/dmabuf-sync.c line 305

Quoted 4 lines
drivers/base/dmabuf-sync.c new file +3 -0
+	kfree(sync);
+}
+EXPORT_SYMBOL(dmabuf_sync_fini);

_GPL ?

Quoted 6 lines

drivers/base/dmabuf-sync.c line 308

Quoted 3 lines
drivers/base/dmabuf-sync.c new file +3 -0
+
+/*
+ * dmabuf_sync_get_obj - Add a given object to syncs list.

sync's list I think?

Ok, seems better.

Quoted 8 lines

drivers/base/dmabuf-sync.c

Quoted 8 lines
drivers/base/dmabuf-sync.c new file +8 -0
+ *
+ * @sync: An object to dmabuf_sync structure.
+ * @dmabuf: An object to dma_buf structure.
+ * @type: A access type to a dma buf.
+ *	The DMA_BUF_ACCESS_R means that this dmabuf could be accessed by
+ *	others for read access. On the other hand, the DMA_BUF_ACCESS_W
+ *	means that this dmabuf couldn't be accessed by others but would be
+ *	accessed by caller's dma exclusively. And the DMA_BUF_ACCESS_DMA
Quoted 8 lines

can be

drivers/base/dmabuf-sync.c line 317

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+ *	combined.

Should this be an enum?

drivers/base/dmabuf-sync.c line 319

Quoted 2 lines
drivers/base/dmabuf-sync.c new file +2 -0
+ *
+ * This function creates and initializes a new dmabuf sync object and

it adds

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+ * the dmabuf sync object to syncs list to track and manage all

dmabufs.

Quoted 22 lines

drivers/base/dmabuf-sync.c line 322

Quoted 2 lines
drivers/base/dmabuf-sync.c new file +2 -0
+ */
+static int dmabuf_sync_get_obj(struct dmabuf_sync *sync, struct dma_buf

*dmabuf,

drivers/base/dmabuf-sync.c line 323

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+					unsigned int type)

enum for 'type'?

drivers/base/dmabuf-sync.c line 336

Quoted 13 lines
drivers/base/dmabuf-sync.c new file +13 -0
+{
+	struct dmabuf_sync_object *sobj;
+
+	if (!dmabuf->sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	if (!IS_VALID_DMA_BUF_ACCESS_TYPE(type))
+		return -EINVAL;
+
+	if ((type & DMA_BUF_ACCESS_RW) == DMA_BUF_ACCESS_RW)
+		type &= ~DMA_BUF_ACCESS_R;

Ah, that is why you are not using an enum.

Quoted 8 lines

drivers/base/dmabuf-sync.c line 340

Quoted 4 lines
drivers/base/dmabuf-sync.c new file +4 -0
+
+	sobj = kzalloc(sizeof(*sobj), GFP_KERNEL);
+	if (!sobj) {
+		WARN_ON(1);

I think you can skip that WARN_ON. Handling an -ENOMEM should be something fairly easy to handle by the calleer.

Ok, will remove it.

Quoted 16 lines
Quoted 16 lines
drivers/base/dmabuf-sync.c new file +16 -0
+		return -ENOMEM;
+	}
+
+	get_dma_buf(dmabuf);
+
+	sobj->dmabuf = dmabuf;
+	sobj->robj = dmabuf->sync;
+	sobj->access_type = type;
+
+	mutex_lock(&sync->lock);
+	list_add_tail(&sobj->head, &sync->syncs);
+	mutex_unlock(&sync->lock);
+
+	return 0;
+}
+
Quoted 8 lines

drivers/base/dmabuf-sync.c line 362

Quoted 6 lines
drivers/base/dmabuf-sync.c new file +6 -0
+/*
+ * dmabuf_sync_put_obj - Release a given sync object.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after

s/is//

Sure.

Quoted 26 lines
Quoted 26 lines
drivers/base/dmabuf-sync.c new file +26 -0
+ * dmabuf_sync_get_obj call to release a given sync object.
+ */
+static void dmabuf_sync_put_obj(struct dmabuf_sync *sync,
+					struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_object *sobj;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry(sobj, &sync->syncs, head) {
+		if (sobj->dmabuf != dmabuf)
+			continue;
+
+		dma_buf_put(sobj->dmabuf);
+
+		list_del_init(&sobj->head);
+		kfree(sobj);
+		break;
+	}
+
+	if (list_empty(&sync->syncs))
+		sync->status = 0;
+
+	mutex_unlock(&sync->lock);
+}
+
Quoted 8 lines

drivers/base/dmabuf-sync.c line 394

Quoted 6 lines
drivers/base/dmabuf-sync.c new file +6 -0
+/*
+ * dmabuf_sync_put_objs - Release all sync objects of dmabuf_sync.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after

s/is//

Sure.

Quoted 151 lines
Quoted 38 lines
drivers/base/dmabuf-sync.c new file +37 -0
+ * dmabuf_sync_get_obj call to release all sync objects.
+ */
+static void dmabuf_sync_put_objs(struct dmabuf_sync *sync)
+{
+	struct dmabuf_sync_object *sobj, *next;
+
+	mutex_lock(&sync->lock);
+
+	list_for_each_entry_safe(sobj, next, &sync->syncs, head) {
+		dma_buf_put(sobj->dmabuf);
+
+		list_del_init(&sobj->head);
+		kfree(sobj);
+	}
+
+	mutex_unlock(&sync->lock);
+
+	sync->status = 0;
+}
+
+/**
+ * dmabuf_sync_lock - lock all dmabufs added to syncs list.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * The caller should call this function prior to CPU or DMA access to
+ * the dmabufs so that others can not access the dmabufs.
+ * Internally, this function avoids dead lock issue with ww-mutex.
+ */
+int dmabuf_sync_lock(struct dmabuf_sync *sync)
+{
+	int ret;
+
+	if (!sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}

drivers/base/dmabuf-sync.c line 441

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+	if (list_empty(&sync->syncs))
+		return -EINVAL;
+
+	if (sync->status != DMABUF_SYNC_GOT)
+		return -EINVAL;
+
+	ret = dmabuf_sync_lock_objs(sync, &sync->ctx);
+	if (ret < 0) {
+		WARN_ON(1);

Perhaps also include the ret value in the WARN?

drivers/base/dmabuf-sync.c line 449

Quoted 8 lines
drivers/base/dmabuf-sync.c new file +8 -0
+		return ret;
+	}
+
+	sync->status = DMABUF_SYNC_LOCKED;
+
+	return ret;
+}
+EXPORT_SYMBOL(dmabuf_sync_lock);

I think you know what I am going to say.

Quoted 7 lines
drivers/base/dmabuf-sync.c new file +7 -0
+
+/**
+ * dmabuf_sync_unlock - unlock all objects added to syncs list.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * The caller should call this function after CPU or DMA access to

drivers/base/dmabuf-sync.c

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+ * the dmabufs is completed so that others can access the dmabufs.
+ */
+int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+{
+	if (!sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	/* If current dmabuf sync object wasn't reserved then just return.

*/

Quoted 8 lines
drivers/base/dmabuf-sync.c new file +8 -0
+	if (sync->status != DMABUF_SYNC_LOCKED)
+		return -EAGAIN;
+
+	dmabuf_sync_unlock_objs(sync, &sync->ctx);
+
+	return 0;
+}
+EXPORT_SYMBOL(dmabuf_sync_unlock);

drivers/base/dmabuf-sync.c

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+
+/**
+ * dmabuf_sync_single_lock - lock a dma buf.
+ *
+ * @dmabuf: A dma buf object that tries to lock.
+ * @type: A access type to a dma buf.
+ *	The DMA_BUF_ACCESS_R means that this dmabuf could be accessed by
+ *	others for read access. On the other hand, the DMA_BUF_ACCESS_W
+ *	means that this dmabuf couldn't be accessed by others but would be
+ *	accessed by caller's dma exclusively. And the DMA_BUF_ACCESS_DMA

can

drivers/base/dmabuf-sync.c

Quoted 7 lines
drivers/base/dmabuf-sync.c new file +7 -0
+ *	be combined with other.
+ * @wait: Indicate whether caller is blocked or not.
+ *	true means that caller will be blocked, and false means that this
+ *	function will return -EAGAIN if this caller can't take the lock
+ *	right now.
+ *
+ * The caller should call this function prior to CPU or DMA access to

the dmabuf

Quoted 14 lines
drivers/base/dmabuf-sync.c new file +14 -0
+ * so that others cannot access the dmabuf.
+ */
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf, unsigned int type,
+				bool wait)
+{
+	struct dmabuf_sync_reservation *robj;
+
+	if (!dmabuf->sync) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	if (!IS_VALID_DMA_BUF_ACCESS_TYPE(type)) {
+		WARN_ON(1);

drivers/base/dmabuf-sync.c

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+		return -EINVAL;
+	}
+
+	get_dma_buf(dmabuf);
+	robj = dmabuf->sync;
+
+	mutex_lock(&robj->lock);
+
+	/* Don't lock in case of read and read. */
+	if (robj->accessed_type & DMA_BUF_ACCESS_R && type &

DMA_BUF_ACCESS_R) {

drivers/base/dmabuf-sync.c

Quoted 7 lines
drivers/base/dmabuf-sync.c new file +7 -0
+		atomic_inc(&robj->shared_cnt);
+		mutex_unlock(&robj->lock);
+		return 0;
+	}
+
+	/*
+	 * In case of F_SETLK, just return -EAGAIN if this dmabuf has

already

Quoted 12 lines
drivers/base/dmabuf-sync.c new file +12 -0
+	 * been locked.
+	 */
+	if (!wait && robj->locked) {
+		mutex_unlock(&robj->lock);
+		dma_buf_put(dmabuf);
+		return -EAGAIN;
+	}
+
+	mutex_unlock(&robj->lock);
+
+	mutex_lock(&robj->sync_lock.base);
+
Quoted 8 lines

drivers/base/dmabuf-sync.c line 537

Quoted 3 lines
drivers/base/dmabuf-sync.c new file +3 -0
+	mutex_lock(&robj->lock);
+	robj->locked = true;
+	mutex_unlock(&robj->lock);

Are you missing an mutex_unlock on &robj->sync_lock.base? Oh wait, that is the purpose of this code. You might want to put a nice comment right above that and say: "Unlocked by dmabuf_sync_single_unlock"

Will add the comment.

Quoted 98 lines
Quoted 53 lines
drivers/base/dmabuf-sync.c new file +52 -0
+
+	return 0;
+}
+EXPORT_SYMBOL(dmabuf_sync_single_lock);
+
+/**
+ * dmabuf_sync_single_unlock - unlock a dma buf.
+ *
+ * @dmabuf: A dma buf object that tries to unlock.
+ *
+ * The caller should call this function after CPU or DMA access to
+ * the dmabuf is completed so that others can access the dmabuf.
+ */
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_reservation *robj;
+
+	if (!dmabuf->sync) {
+		WARN_ON(1);
+		return;
+	}
+
+	robj = dmabuf->sync;
+
+	mutex_lock(&robj->lock);
+
+	if (robj->polled) {
+		robj->poll_event = true;
+		robj->polled = false;
+		wake_up_interruptible(&robj->poll_wait);
+	}
+
+	if (atomic_add_unless(&robj->shared_cnt, -1 , 1)) {
+		mutex_unlock(&robj->lock);
+		dma_buf_put(dmabuf);
+		return;
+	}
+
+	mutex_unlock(&robj->lock);
+
+	mutex_unlock(&robj->sync_lock.base);
+
+	mutex_lock(&robj->lock);
+	robj->locked = false;
+	mutex_unlock(&robj->lock);
+
+	dma_buf_put(dmabuf);
+
+	return;
+}
+EXPORT_SYMBOL(dmabuf_sync_single_unlock);
+

drivers/base/dmabuf-sync.c

Quoted 10 lines
drivers/base/dmabuf-sync.c new file +10 -0
+/**
+ * dmabuf_sync_get - Get dmabuf sync object.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ * @sync_buf: A dmabuf object to be synchronized with others.
+ * @type: A access type to a dma buf.
+ *	The DMA_BUF_ACCESS_R means that this dmabuf could be accessed by
+ *	others for read access. On the other hand, the DMA_BUF_ACCESS_W
+ *	means that this dmabuf couldn't be accessed by others but would be
+ *	accessed by caller's dma exclusively. And the DMA_BUF_ACCESS_DMA

can

drivers/base/dmabuf-sync.c

Quoted 3 lines
drivers/base/dmabuf-sync.c new file +3 -0
+ *	be combined with other.
+ *
+ * This function should be called after dmabuf_sync_init function is

called.

drivers/base/dmabuf-sync.c line 603

Quoted 1 line
drivers/base/dmabuf-sync.c new file +1 -0
+ * The caller can tie up multiple dmabufs into one sync object by

calling this

drivers/base/dmabuf-sync.c

Quoted 5 lines
drivers/base/dmabuf-sync.c new file +5 -0
+ * function several times. Internally, this function allocates
+ * a dmabuf_sync_object and adds a given dmabuf to it, and also takes
+ * a reference to a dmabuf.
+ */
+int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf, unsigned

int type)

Quoted 22 lines
drivers/base/dmabuf-sync.c new file +22 -0
+{
+	int ret;
+
+	if (!sync || !sync_buf) {
+		WARN_ON(1);
+		return -EFAULT;
+	}
+
+	ret = dmabuf_sync_get_obj(sync, sync_buf, type);
+	if (ret < 0) {
+		WARN_ON(1);
+		return ret;
+	}
+
+	sync->status = DMABUF_SYNC_GOT;
+
+	return 0;
+}
+EXPORT_SYMBOL(dmabuf_sync_get);
+
+/**
+ * dmabuf_sync_put - Put dmabuf sync object to a given dmabuf.
Quoted 8 lines

drivers/base/dmabuf-sync.c

Quoted 8 lines
drivers/base/dmabuf-sync.c new file +8 -0
+ *
+ * @sync: An object to dmabuf_sync structure.
+ * @dmabuf: An dmabuf object.
+ *
+ * This function should be called if some operation is failed after
+ * dmabuf_sync_get function is called to release the dmabuf, or
+ * dmabuf_sync_unlock function is called. Internally, this function
+ * removes a given dmabuf from a sync object and remove the sync

object.

Quoted 64 lines
Quoted 64 lines
drivers/base/dmabuf-sync.c new file +40 -0
+ * At this time, the dmabuf is putted.
+ */
+void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf *dmabuf)
+{
+	if (!sync || !dmabuf) {
+		WARN_ON(1);
+		return;
+	}
+
+	if (list_empty(&sync->syncs))
+		return;
+
+	dmabuf_sync_put_obj(sync, dmabuf);
+}
+EXPORT_SYMBOL(dmabuf_sync_put);
+
+/**
+ * dmabuf_sync_put_all - Put dmabuf sync object to dmabufs.
+ *
+ * @sync: An object to dmabuf_sync structure.
+ *
+ * This function should be called if some operation is failed after
+ * dmabuf_sync_get function is called to release all sync objects, or
+ * dmabuf_sync_unlock function is called. Internally, this function
+ * removes dmabufs from a sync object and remove the sync object.
+ * At this time, all dmabufs are putted.
+ */
+void dmabuf_sync_put_all(struct dmabuf_sync *sync)
+{
+	if (!sync) {
+		WARN_ON(1);
+		return;
+	}
+
+	if (list_empty(&sync->syncs))
+		return;
+
+	dmabuf_sync_put_objs(sync);
+}
+EXPORT_SYMBOL(dmabuf_sync_put_all);
include/linux/dma-buf.h +2 -0
index dfac5ed..0109673 100644
@@ -115,6 +115,7 @@ struct dma_buf_ops {
  * @exp_name: name of the exporter; useful for debugging.
  * @list_node: node for dma_buf accounting and debugging.
  * @priv: exporter specific private data for this buffer object.
+ * @sync: sync object linked to this dma-buf
  */
 struct dma_buf {
 	size_t size;
@@ -128,6 +129,7 @@ struct dma_buf {
 	const char *exp_name;
 	struct list_head list_node;
 	void *priv;
+	void *sync;
 };
 
 /**
@@ -148,6 +150,20 @@ struct dma_buf_attachment {
 	void *priv;
 };
 
Quoted 5 lines

include/linux/dma-buf.h struct dma_buf_attachment {

Quoted 5 lines
include/linux/dma-buf.h +5 -0
+#define	DMA_BUF_ACCESS_R	0x1
+#define DMA_BUF_ACCESS_W	0x2
+#define DMA_BUF_ACCESS_DMA	0x4
+#define DMA_BUF_ACCESS_RW	(DMA_BUF_ACCESS_R | DMA_BUF_ACCESS_W)
+#define DMA_BUF_ACCESS_DMA_R	(DMA_BUF_ACCESS_R |

DMA_BUF_ACCESS_DMA)

Quoted 1 line

include/linux/dma-buf.h struct dma_buf_attachment { · line 158

Quoted 1 line
include/linux/dma-buf.h +1 -0
+#define DMA_BUF_ACCESS_DMA_W	(DMA_BUF_ACCESS_W |

DMA_BUF_ACCESS_DMA)

Quoted 29 lines

include/linux/dma-buf.h struct dma_buf_attachment { · line 159

Quoted 1 line
include/linux/dma-buf.h +1 -0
+#define DMA_BUF_ACCESS_DMA_RW	(DMA_BUF_ACCESS_DMA_R |

DMA_BUF_ACCESS_DMA_W)

Quoted 16 lines
include/linux/dma-buf.h +7 -0
+#define IS_VALID_DMA_BUF_ACCESS_TYPE(t)	(t == DMA_BUF_ACCESS_R || \
+					 t == DMA_BUF_ACCESS_W || \
+					 t == DMA_BUF_ACCESS_DMA_R || \
+					 t == DMA_BUF_ACCESS_DMA_W || \
+					 t == DMA_BUF_ACCESS_RW || \
+					 t == DMA_BUF_ACCESS_DMA_RW)
+
 /**
  * get_dma_buf - convenience wrapper for get_file.
  * @dmabuf:	[in]	pointer to dma_buf
include/linux/dmabuf-sync.h new file +0 -0
new file mode 100644
index 0000000..9a3afc4

include/linux/dmabuf-sync.h line 6

Quoted 6 lines
include/linux/dmabuf-sync.h new file +6 -0
@@ -0,0 +1,190 @@
+/*
+ * Copyright (C) 2013 Samsung Electronics Co.Ltd
+ * Authors:
+ *	Inki Dae <[email protected]>
+ *
+ * This program is free software; you can redistribute  it and/or

modify it

include/linux/dmabuf-sync.h line 7

Quoted 1 line
include/linux/dmabuf-sync.h new file +1 -0
+ * under  the terms of  the GNU General  Public License as published by

the

include/linux/dmabuf-sync.h line 8

Quoted 1 line
include/linux/dmabuf-sync.h new file +1 -0
+ * Free Software Foundation;  either version 2 of the  License, or (at

your

Quoted 15 lines

include/linux/dmabuf-sync.h line 21

Quoted 13 lines
include/linux/dmabuf-sync.h new file +13 -0
+ * option) any later version.
+ *
+ */
+
+#include <linux/mutex.h>
+#include <linux/sched.h>
+#include <linux/dma-buf.h>
+
+enum dmabuf_sync_status {
+	DMABUF_SYNC_GOT		= 1,
+	DMABUF_SYNC_LOCKED,
+};
+

No comment about this structure?

Will add comments.

Quoted 15 lines
Quoted 15 lines
include/linux/dmabuf-sync.h new file +14 -0
+struct dmabuf_sync_reservation {
+	struct ww_mutex		sync_lock;
+	struct mutex		lock;
+	wait_queue_head_t	poll_wait;
+	unsigned int		poll_event;
+	unsigned int		polled;
+	atomic_t		shared_cnt;
+	unsigned int		accessed_type;
+	unsigned int		locked;
+};
+
+/*
+ * A structure for dmabuf_sync_object.
+ *
Quoted 9 lines

include/linux/dmabuf-sync.h line 40

Quoted 5 lines
include/linux/dmabuf-sync.h new file +5 -0
+ * @head: A list head to be added to syncs list.
+ * @robj: A reservation_object object.
+ * @dma_buf: A dma_buf object.
+ * @access_type: Indicate how a current task tries to access
+ *	a given buffer.

Huh? What values are expected then? Is there some #define or enum for that?

Right, there are definitions for that. Will add more comments.

Quoted 13 lines
Quoted 13 lines
include/linux/dmabuf-sync.h new file +13 -0
+ */
+struct dmabuf_sync_object {
+	struct list_head		head;
+	struct dmabuf_sync_reservation	*robj;
+	struct dma_buf			*dmabuf;
+	unsigned int			access_type;
+};
+
+struct dmabuf_sync_priv_ops {
+	void (*free)(void *priv);
+};
+
+/*
Quoted 8 lines

include/linux/dmabuf-sync.h line 58

Quoted 5 lines
include/linux/dmabuf-sync.h new file +5 -0
+ * A structure for dmabuf_sync.
+ *
+ * @syncs: A list head to sync object and this is global to system.
+ * @list: A list entry used as committed list node
+ * @lock: A mutex lock to current sync object.

You should say for which specific operations this mutex is needed. For everything? Or just for list operations.

Ok, will add more comments.

Quoted 8 lines

include/linux/dmabuf-sync.h

Quoted 7 lines
include/linux/dmabuf-sync.h new file +6 -0
+ * @ctx: A current context for ww mutex.
+ * @work: A work struct to release resources at timeout.
+ * @priv: A private data.
+ * @name: A string to dmabuf sync owner.
+ * @timer: A timer list to avoid lockup and release resources.
+ * @status: Indicate current status (DMABUF_SYNC_GOT or

DMABUF_SYNC_LOCKED).

Quoted 12 lines

include/linux/dmabuf-sync.h line 74

Quoted 10 lines
include/linux/dmabuf-sync.h new file +10 -0
+ */
+struct dmabuf_sync {
+	struct list_head		syncs;
+	struct list_head		list;
+	struct mutex			lock;
+	struct ww_acquire_ctx		ctx;
+	struct work_struct		work;
+	void				*priv;
+	struct dmabuf_sync_priv_ops	*ops;
+	char				name[64];

Perhaps a #define for the size?

Ok, will use macro instead.

Quoted 131 lines
Quoted 52 lines
include/linux/dmabuf-sync.h new file +51 -0
+	struct timer_list		timer;
+	unsigned int			status;
+};
+
+#ifdef CONFIG_DMABUF_SYNC
+
+extern struct ww_class dmabuf_sync_ww_class;
+
+static inline void dmabuf_sync_reservation_init(struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_reservation *obj;
+
+	obj = kzalloc(sizeof(*obj), GFP_KERNEL);
+	if (!obj)
+		return;
+
+	dmabuf->sync = obj;
+
+	ww_mutex_init(&obj->sync_lock, &dmabuf_sync_ww_class);
+
+	mutex_init(&obj->lock);
+	atomic_set(&obj->shared_cnt, 1);
+
+	init_waitqueue_head(&obj->poll_wait);
+}
+
+static inline void dmabuf_sync_reservation_fini(struct dma_buf *dmabuf)
+{
+	struct dmabuf_sync_reservation *obj;
+
+	if (!dmabuf->sync)
+		return;
+
+	obj = dmabuf->sync;
+
+	ww_mutex_destroy(&obj->sync_lock);
+
+	kfree(obj);
+}
+
+extern bool is_dmabuf_sync_supported(void);
+
+extern struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void *priv);
+
+extern void dmabuf_sync_fini(struct dmabuf_sync *sync);
+
+extern int dmabuf_sync_lock(struct dmabuf_sync *sync);
+
+extern int dmabuf_sync_unlock(struct dmabuf_sync *sync);

include/linux/dmabuf-sync.h

Quoted 10 lines
include/linux/dmabuf-sync.h new file +10 -0
+
+int dmabuf_sync_single_lock(struct dma_buf *dmabuf, unsigned int type,
+				bool wait);
+
+void dmabuf_sync_single_unlock(struct dma_buf *dmabuf);
+
+extern int dmabuf_sync_get(struct dmabuf_sync *sync, void *sync_buf,
+				unsigned int type);
+
+extern void dmabuf_sync_put(struct dmabuf_sync *sync, struct dma_buf

*dmabuf);

include/linux/dmabuf-sync.h

Quoted 6 lines
include/linux/dmabuf-sync.h new file +6 -0
+
+extern void dmabuf_sync_put_all(struct dmabuf_sync *sync);
+
+#else
+
+static inline void dmabuf_sync_reservation_init(struct dma_buf *dmabuf)

{ }

include/linux/dmabuf-sync.h line 101

Quoted 2 lines
include/linux/dmabuf-sync.h new file +2 -0
+
+static inline void dmabuf_sync_reservation_fini(struct dma_buf *dmabuf)

{ }

Quoted 50 lines
include/linux/dmabuf-sync.h new file +47 -1
+
+static inline bool is_dmabuf_sync_supported(void) { return false; }
+
+static inline  struct dmabuf_sync *dmabuf_sync_init(const char *name,
+					struct dmabuf_sync_priv_ops *ops,
+					void *priv)
+{
+	return ERR_PTR(0);
+}
+
+static inline void dmabuf_sync_fini(struct dmabuf_sync *sync) { }
+
+static inline int dmabuf_sync_lock(struct dmabuf_sync *sync)
+{
+	return 0;
+}
+
+static inline int dmabuf_sync_unlock(struct dmabuf_sync *sync)
+{
+	return 0;
+}
+
+static inline int dmabuf_sync_single_lock(struct dma_buf *dmabuf,
+						unsigned int type,
+						bool wait)
+{
+	return 0;
+}
+
+static inline void dmabuf_sync_single_unlock(struct dma_buf *dmabuf)
+{
+	return;
+}
+
+static inline int dmabuf_sync_get(struct dmabuf_sync *sync,
+					void *sync_buf,
+					unsigned int type)
+{
+	return 0;
+}
+
+static inline void dmabuf_sync_put(struct dmabuf_sync *sync,
+					struct dma_buf *dmabuf) { }
+
+static inline void dmabuf_sync_put_all(struct dmabuf_sync *sync) { }
+
+#endif
--

1.7.5.4

Quoted 4 lines

_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel

--
To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#4 lore raw

Re: [PATCH 1/2] [RFC PATCH v6] dmabuf-sync: Add a buffer synchronization framework

Quoted 10 lines
Quoted 7 lines
Quoted 1 line

+EXPORT_SYMBOL(is_dmabuf_sync_supported);

_GPL ?

I would also prefix it with 'dmabuf_is_sync_supported' just to make all of the libraries call start with 'dmabuf'

Seems better. Will change it to dmabuf_is_sync_supported, and use EXPORT_SYMBOL_GPL.

One thing thought - while I suggest that you use GPL variant I think you should check who the consumers are. As in, if nvidia wants to use it it might make their lawyers unhappy - and in turn means that their engineers won't be able to use these symbols.

So - if there is a strong argument to not have it GPL - then please say so.

author
#5 lore raw

RFC PATCH v2 2/2 dma-buf: Add user interfaces for dmabuf sync support.

In reply to Inki Dae

This patch adds lock and poll callbacks to dma buf file operations, and these callbacks will be called by fcntl and select system calls.

fcntl and select system calls can be used to wait for the completion of DMA or CPU access to a shared dmabuf. The difference of them is fcntl system call takes a lock after the completion but select system call doesn't. So in case of fcntl system call, it's useful when a task wants to access a shared dmabuf without any broken. On the other hand, it's useful when a task wants to just wait for the completion.

Changelog v2:
- Add select system call support.
. The purpose of this feature is to wait for the completion of DMA or
CPU access to a dmabuf without that caller locks the dmabuf again after the completion.
That is useful when caller wants to be aware of the completion of DMA access to the dmabuf, and the caller doesn't use intefaces for the DMA device driver.

---
 drivers/base/dma-buf.c      |   81 +++++++++++++++++++++++++++++++++++++++++++
 include/linux/dmabuf-sync.h |    1 +
 2 files changed, 82 insertions(+), 0 deletions(-)
drivers/base/dma-buf.c +81 -0
index 4aca57a..f16a396 100644
@@ -29,6 +29,7 @@
 #include <linux/export.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
+#include <linux/poll.h>
 #include <linux/dmabuf-sync.h>
 
 static inline int is_dma_buf_file(struct file *);
@@ -80,9 +81,89 @@ static int dma_buf_mmap_internal(struct file *file, struct vm_area_struct *vma)
 	return dmabuf->ops->mmap(dmabuf, vma);
 }
 
+static unsigned int dma_buf_poll(struct file *filp,
+					struct poll_table_struct *poll)
+{
+	struct dma_buf *dmabuf;
+	struct dmabuf_sync_reservation *robj;
+	int ret = 0;
+
+	if (!is_dma_buf_file(filp))
+		return POLLERR;
+
+	dmabuf = filp->private_data;
+	if (!dmabuf || !dmabuf->sync)
+		return POLLERR;
+
+	robj = dmabuf->sync;
+
+	mutex_lock(&robj->lock);
+
+	robj->polled = true;
+
+	/*
+	 * CPU or DMA access to this buffer has been completed, and
+	 * the blocked task has been waked up. Return poll event
+	 * so that the task can get out of select().
+	 */
+	if (robj->poll_event) {
+		robj->poll_event = false;
+		mutex_unlock(&robj->lock);
+		return POLLIN | POLLOUT;
+	}
+
+	/*
+	 * There is no anyone accessing this buffer so just return.
+	 */
+	if (!robj->locked) {
+		mutex_unlock(&robj->lock);
+		return POLLIN | POLLOUT;
+	}
+
+	poll_wait(filp, &robj->poll_wait, poll);
+
+	mutex_unlock(&robj->lock);
+
+	return ret;
+}
+
+static int dma_buf_lock(struct file *file, int cmd, struct file_lock *fl)
+{
+	struct dma_buf *dmabuf;
+	unsigned int type;
+	bool wait = false;
+
+	if (!is_dma_buf_file(file))
+		return -EINVAL;
+
+	dmabuf = file->private_data;
+
+	if ((fl->fl_type & F_UNLCK) == F_UNLCK) {
+		dmabuf_sync_single_unlock(dmabuf);
+		return 0;
+	}
+
+	/* convert flock type to dmabuf sync type. */
+	if ((fl->fl_type & F_WRLCK) == F_WRLCK)
+		type = DMA_BUF_ACCESS_W;
+	else if ((fl->fl_type & F_RDLCK) == F_RDLCK)
+		type = DMA_BUF_ACCESS_R;
+	else
+		return -EINVAL;
+
+	if (fl->fl_flags & FL_SLEEP)
+		wait = true;
+
+	/* TODO. the locking to certain region should also be considered. */
+
+	return dmabuf_sync_single_lock(dmabuf, type, wait);
+}
+
 static const struct file_operations dma_buf_fops = {
 	.release	= dma_buf_release,
 	.mmap		= dma_buf_mmap_internal,
+	.poll		= dma_buf_poll,
+	.lock		= dma_buf_lock,
 };
 
 /*
include/linux/dmabuf-sync.h +1 -0
index 9a3afc4..0316f68 100644
@@ -11,6 +11,7 @@
  */
 
 #include <linux/mutex.h>
+#include <linux/ww_mutex.h>
 #include <linux/sched.h>
 #include <linux/dma-buf.h>
 
Signature · 1 line
1.7.5.4
author
#6 lore raw

v6 cover Introduce buffer synchronization framework

In reply to Inki Dae

Just adding more detailed descriptions.

Hi all,

This patch set introduces a buffer synchronization framework based on DMA BUF[1] and based on ww-mutexes[2] for lock mechanism, and may be final RFC.

The purpose of this framework is to provide not only buffer access control to CPU and CPU, and CPU and DMA, and DMA and DMA but also easy-to-use interfaces for device drivers and user application. In addtion, this patch set suggests a way for enhancing performance.

For generic user mode interface, we have used fcntl and select system call[3]. As you know, user application sees a buffer object as a dma-buf file descriptor. So fcntl() call with the file descriptor means to lock some buffer region being managed by the dma-buf object. And select() call means to wait for the completion of CPU or DMA access to the dma-buf without locking. For more detail, you can refer to the dma-buf-sync.txt in Documentation/

There are some cases we should use this buffer synchronization framework. One of which is to primarily enhance GPU rendering performance on Tizen platform in case of 3d app with compositing mode that 3d app draws something in off-screen buffer, and Web app.

In case of 3d app with compositing mode which is not a full screen mode, the app calls glFlush to submit 3d commands to GPU driver instead of glFinish for more performance. The reason we call glFlush is that glFinish blocks caller's task until the execution of the 2d commands is completed. Thus, that makes GPU and CPU more idle. As result, 3d rendering performance with glFinish is quite lower than glFlush. However, the use of glFlush has one issue that the a buffer shared with GPU could be broken when CPU accesses the buffer at once after glFlush because CPU cannot be aware of the completion of GPU access to the buffer. Of course, the app can be aware of that time using eglWaitGL but this function is valid only in case of the same process.

The below summarizes how app's window is displayed on Tizen platform:
1. X client requests a window buffer to Xorg.
2. X client draws something in the window buffer using CPU.
3. X client requests SWAP to Xorg.
4. Xorg notifies a damage event to Composite Manager.
5. Composite Manager gets the window buffer (front buffer) through DRI2GetBuffers.
6. Composite Manager composes the window buffer and its own back buffer using GPU. At this time, eglSwapBuffers is called: internally, 3d commands are flushed to gpu driver.
7. Composite Manager requests SWAP to Xorg.
8. Xorg performs drm page flip. At this time, the window buffer is displayed on screen.

Web app based on HTML5 also has similar procedure. Web browser and its web app are different processs. Web app draws something in its own buffer, and then the web browser gets a window buffer from Xorg, and then composes those two buffers using GPU.

Thus, in such cases, a shared buffer could be broken when one process draws something in a shared buffer using CPU while Composite manager is composing two buffers - X client's front buffer and Composite manger's back buffer, or web app's front buffer and web browser's back buffer - using GPU without any locking mechanism. That is why we need user land locking interface, fcntl system call.

And last one is a deferred page flip issue. This issue is that a window buffer rendered can be displayed on screen in about 32ms in worst case:
assume that the gpu rendering is completed within 16ms.
That can be incurred when compositing a pixmap buffer with a window buffer using GPU and when vsync is just started. At this time, Xorg waits for a vblank event to get a window buffer so 3d rendering will be delayed up to about 16ms. As a result, the window buffer would be displayed in about two vsyncs (about 32ms) and in turn, that would show slow responsiveness.

For this, we could enhance the responsiveness with locking
mechanism: skipping one vblank wait. I guess in the similar reason, Android, Chrome OS, and other platforms are using their own locking mechanisms; Android sync driver, KDS, and DMA fence.

The below shows the deferred page flip issue in worst case,

               |------------ <- vsync signal
               |<------ DRI2GetBuffers
               |
               |
               |
               |------------ <- vsync signal
               |<------ Request gpu rendering
          time |
               |
               |<------ Request page flip (deferred)
               |------------ <- vsync signal
               |<------ Displayed on screen
               |
               |
               |
               |------------ <- vsync signal

Thanks,
Inki Dae

References:
[1] http://lwn.net/Articles/470339/
[2] https://patchwork.kernel.org/patch/2625361/
[3] http://linux.die.net/man/2/fcntl

Inki Dae (2):
[RFC PATCH v6] dmabuf-sync: Add a buffer synchronization framework [RFC PATCH v2] dma-buf: Add user interfaces for dmabuf sync support.

 Documentation/dma-buf-sync.txt |  285 +++++++++++++++++
 drivers/base/Kconfig           |    7 +
 drivers/base/Makefile          |    1 +
 drivers/base/dma-buf.c         |   85 +++++
 drivers/base/dmabuf-sync.c     |  678 ++++++++++++++++++++++++++++++++++++++++
 include/linux/dma-buf.h        |   16 +
 include/linux/dmabuf-sync.h    |  191 +++++++++++
 7 files changed, 1263 insertions(+), 0 deletions(-)
 create mode 100644 Documentation/dma-buf-sync.txt
 create mode 100644 drivers/base/dmabuf-sync.c
 create mode 100644 include/linux/dmabuf-sync.h
Signature · 1 line
1.7.5.4