- It could happen that 2 threads try to update the same event within tree
- First thread starts to submit command, it releases cmdToSubmit
- Second thread doesn't see the command and follows
- Second thread thinks that it submitted the command so it follows to
subsequent steps which are child events notification
- We end up with corrupted enqueue sequence as child may submit prior to
parent.
- With this change each submit step is synchronized basing on task count
- When second thread enters submit command without task count being set
it wait for first thread to properly set it.
Change-Id: Ic2ddaea17f9af8cab6781320edae2c268dd0b189
Signed-off-by: Mrozek, Michal <michal.mrozek@intel.com>
- do not wait on map/unmap if zero copy resource.
- do not wait in marker commands at all.
Change-Id: I74cdd8320d13602bf662eed412ed2fcad1504989
Signed-off-by: Mrozek, Michal <michal.mrozek@intel.com>
- is queue blocked to release virtual event when it is submitted
- do not wait on completion after ndr submit.
Change-Id: I63c54e6940f6241b32f0dbdd5880f849c46fd12e
Signed-off-by: Mrozek, Michal <michal.mrozek@intel.com>
Include files are now grouped and sorted in following order:
1. Header file of the class the current file implements
2. Project files
3. Third party files
4. Standard library
Change-Id: If31af05652184169f7fee1d7ad08f1b2ed602cf0
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
- remove method allocateGraphicsMemory(size_t size)
- pass allocation type in allocation properties
- set allocation type in allocateGraphicsMemoryInPreferredPool
Change-Id: Ia9296d0ab2f711b7d78aff615cb56b3a246b60ec
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
Remove not needed includes from command_queue.h
Change-Id: I45963bf005471bd7716d55471474299a15e27b62
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
Updating files modified in 2018 only. Older files remain with old style
copyright header
Change-Id: Ic99f2e190ad74b4b7f2bd79dd7b9fa5fbe36ec92
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
- change the way we handle blocked commands.
- instead of allocating CPU pointer and populating it with commands, create
real IndirectHeap that may be later submitted to the GPU
- that removes a lot of copy operations that were happening on submit time
- for device enqueue, this requires dsh & shh to be passed directly to the
underlying commands, in that scenario device queue buffers are not used
Change-Id: I1124a8edbb46777ea7f7d3a5946f302e7fdf9665
- This removes Instruction Heap allocation from enqueue path
- Blocked path is handled as well
- Heap is no longer allocated on demand it is bind to kernelInfo.
Change-Id: I54545beceed3404ee0330a8bac2b0934944cac30
- read only memory cannot be used for allocation,
Oses cannot create graphics alocation for such memory
- if memory allocation fails for host_ptr passed
to enqueueWrite calls, then try doing new allocation
and copy host_ptr on cpu
Change-Id: I415a4673ae1319ea8f77e53bd8fba7489fe85218
- Dont make cpu/gpu writes on read-only unmap
- Read/Write on limited map range only
- Overlaps checks for non read-only maps
- Fixed cmd type on returned event
Change-Id: I98ca542e8d369d2426a87279f86cadb0bf3db299
When queue is blocked on non-blocking call, map operation is added to
waitlist dependencies. Returning slice/row pitch for map image was skipped
Change-Id: I46f97590315e7aee7fbbfbdb615f383cdb666307
