[SYCL][Scheduler] Rework of host accessor and host allocation

[romanovvlad]

Background for use_host_ptr property:

The SYCL specification defines use_host_ptr property which can be passed
into the buffer or image constructors. The SYCL specification says the
following:

The use_host_ptr property adds the requirement that the SYCL runtime
must not allocate any memory for the SYCL buffer and instead uses the
provided host pointer directly. This prevents the SYCL runtime from
allocating additional temporary storage on the host.

Current implementation violates this rule in some scenarios, such as
moving the latest data of the buffer from one OCL device context to
another and on update_host type of explicit memory operation.

Background for host accessor:

Current implementation of host accessor requires that the underlying
backend supports async/delayed execution: SYCL RT calls Unmap or Write
commands that are initially blocked by an user event. Because of that,
for instance, host accessors work incorrectly if the latest memory are
located on the host device. Also, user events are usually not very well
supported, there is "UseExclusiveQueue" workaround because of that.

Solution:

The patch eliminates the need to have more then one host allocation by
introducing concept of linked alloca commands. One device and one host
alloca commands can be linked together so they share* the same host
memory. Such commands have "active" state which indicates that it's
valid to work with corresponding memory allocation. Only one alloca
command from a pair can be active at the same time. Memory transition
between them is performed using map/unmap commands.

• share from the SYCL point of view, because even if SYCL RT asks
  underlying backends to reuse host pointer they are free to allocate
  additional memory on their own.

In order to resolve the limitations of current implementation of the
host accessor the patch updates addHostAccessor API to emit map or
read operations along with an empty command which is in blocked
state(cannot be enqueued). An empty command is unblocked inside in the
new releaseHostAccessor API which is called from host accessor
destructor.

As bonuses/side effects the patch:

1. Allows having multiple host accessor with read access mode at the
   same time in all cases.
2. Adds environment variable SYCL_THROW_ON_BLOCK which if set makes SYCL
   RT throw an exception on attempt to wait for a blocked command.
3. Makes so the latest state of memory object is still on the host after
   host accessor destructor in all cases, so when multiple host
   accessors are constructed in a row no map/read operations are
   performed starting from the second host accessor. Before the patch
   memory was moved back to the source(for example, to OpenCL memory
   allocation) so SYCL RT had to make map/read each time.
4. Makes Scheduler considering host and different instances of the host
   device to be in the same context when it looks for a suitable
   allocation.

Signed-off-by: Vlad Romanov [email protected]

• Fix image constructor [SYCL] Add support for image with pitch but without host ptr provided. #723
• Introduce empty task [SYCL] Add empty task type of commands #725
• Small fixes for update host req. [SYCL] Small adjustments of UpdateHostRequirementCommand. #729
• Support for blocking enqueue [SYCL] Add support for blocking enqueue #733
• Refactoring of commands take host ptr for updating [SYCL][Scheduler] Refactor data transfer scheduler commands #830
• Refactoring of how requirement is stored in commands [SYCL][Scheduler] Refactor commands #839
• Rework of host accessor and host allocation [SYCL][Scheduler] Rework of host accessor and host allocation #724

[romanovvlad]

It turned out that there is additional problem which blocked first attempt to reimplement host accessor.
The problem is that sometime ago we have started ignoring use_host_ptr property. So, new raw diff contains the fix for that problem too.
Separating these two things would require to spend additional time to update host accessor implementation which is anyway going to be replaced by this one that why I'm going to try to commit host accessor re implementation and support for use_host_ptr in the patch.

[romanovvlad]

Will rebase tomorrow so there will be only on commit for review.

[romanovvlad]

Removed the test because it expects line(piEnqueueMemBufferWrite) which is not generated for the code in the test anymore. Attempt to rewrite the test was not successful as the problem this test was written for cannot happen anymore.

[sergey-semenov]

LGTM overall, just some minor comments

[sergey-semenov]

Other than the last comment, LGTM