- For every command buffer that we submit, pass it to gem close worker.
- Gem close worker will do asynchronous cleanup if this resource is meant to
be destroyed.
- if the resource is not meant to be destroyed we will call IOCTL wait for
this batch buffer.
- This will result in bumping up GPU clocks and better performance.
Change-Id: If9f181e411d7748573f31682e875a97c5355abe5
- Only Wddm object owns Gdi
- Dont pass Gdi object to constructor
- Move Wddm related files to new directory
Change-Id: Iadd26634c7692db760d7d3367211c32d2c2c8121
- Use local gmmClientContext instead of pGMMGlobalContext
- ResourceInfo and PTmanager creation from gmmClientContext
- Mock Gmm context creation in Wddm to have only one instance per run
Change-Id: I67e015c57f0ab5524564760fd9a849615615697f
- add source level debugger to device
- load isDebuggerActive function from library
- rename interface to sourceLevelDebuggerInterface in SLD
- add DebugData to KernelInfo with kernel debug data
Change-Id: I2643ee633f8dc5c97e8bbdc9d4e7977ddcbf440d
- Rename misnamed test function
- Adjust 2 tests, so they use CSR size getters instead of hardcoded values
- Move getSizeRequiredPreambleCS() into CommandStreamReceiverHw class
- Improve PreambleHelper size estimating
Change-Id: I3f292d50e08f3d10d190c9f8722e1f0498481154
- This allows applications to force the N:1 aggregation by creating out
of order queue.
- That switches csr to N:1 submission model where commands from multiple
command streams may be aggregated.
- That forces scenarios returning an event to be aggregated as well.
Change-Id: I8fd8d7f88bb2665234ee90870133120b206710a8
-This is required to enable N:1 submission model.
-If heaps are coming from different command queues that always
mean that STATE_BASE_ADDRESS needs to be reloaded
-In order to not emit any non pipelined state in CSR, this change
moves the ownership of IndirectHeap to one centralized place which is
CommandStreamReceiver
-This way when there are submissions from multiple command queues then
they reuse the same heaps, therefore preventing SBA reload
Change-Id: I5caf5dc5cb05d7a2d8766883d9bc51c29062e980
- add defines to command line
- remove most occurences of include "config.h"
Change-Id: I19d65d83c895fc6143d319d057a50e5ae3e78830
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
- use full type specification and remove casts in MemoryManager
- remove TagAllocatorBase not used any more
- make TagAllocator to be profiling/instrumentation agnostic
- unify UnlimitedTagCount and make part of TagAllocator
Change-Id: I7b5b1ed83aa5e1f0839f611db0530d7e062a3c25
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
- interface to debugger dynamic library
- code is included when source level debugger header exists,
otherwise implementation is exluded from build
- unit tests do not load real dynamic library,
instead test version (DebbugerLibrary) of OsLibrary is used.
Change-Id: Id3229c77963352e8001043ee41b7d48c6b180a59
- They are required to properly handle scenarios where null buffers
are passed by the application.
Change-Id: I5bf3c70220ebaf01b16f74ac0e617abdf442c604
- In VME scenarios input buffers may be NULL, we need to skip
some parts of the kernel instead of trying to read from NULL pointer.
Change-Id: Ie9788ec76af1be270f6a03547fdcf80c7b2c84b2
- ThkWrapper had uninitialized mFunc member, setting it
to nullptr
- D3DSurface could dereference null image pointer,
adding validateUpdateData method in SharingHandler
that may return CL_INVALID_MEM_OBJECT if memObject is invalid
Change-Id: Iaa4499bcea47baca156c9d28be4c93ba4f0e1ebb
We don't need BuiltinDispatchInfoBuilder in every place where built ins
are used. specifically in .cpp files generated from kernel binary.
Change-Id: Ie739951cdc93873993f78ad14cee656122af51fd
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
- There was a wrong cast in Graphics Allocation constructor resulting
in wrong GPU address generation in some sporadic scenarios.
- Problem appears in 32 bit applications where void* address is cast to
uint64_t value, if c style cast is used it makes trailing bit to be
populated to higher bits constructing wrong value
0xf000000 is being casted to 0xfffffffff0000000 while it should be casted to
0x00000000f0000000
- added special cast function for further use.
Change-Id: I56d53a8e13e17cbacd127566442eea3f6a089977
- Move indirect heap to internal allocator domain.
- Add logic in getIndirectHeap to allocate with proper API depending on
heap type
- Add State base Address programming, reflecting that now Indirect Object
Heap is placed in 4GB domain.
- For AddPatchInfoCommentsForAUBDump mode , keep all heaps in non 4GB mode.
Change-Id: I6862f6a249e444d0d6cfe7e499a10d43f284553e
- generate debug data to .dbg file in cloc
- generate debug kernel for ults with "-g" option
in addition to "-cl-kernel-debug-enable"
- append "-g" option for compilation and build of
programs with kernel debugging enabled to make
compiler generate debug data
Change-Id: I09401f84be6e09da167194a44d1b9a7f2bfb622d
Intel Graphics Compiler is preparing change to unify compiler naming.
They will expose two variables which we should use when loading compiler
libraries.
Change-Id: If6edcb7541452b3cd429a8b4f7c26f6d43169035
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
- Add Indirect Heap function that will be used to program State Base Address.
- This is to allow indirect heaps to work in 2 modes, either heap will service
as whole indirect allocation OR offsets in 4GB space will be used.
Change-Id: Ic4ca1e907c1b30d2f98dc39e8ab945ce35ed6ad0
- Internal allocations may now coexists with non internal on reusable list.
- Caller now specifies if internal allocation is needed.
- If criteria are not met , then allocation is not returned.
Change-Id: I7da3a4f944768b7c8a873e44fd47248f1d76bf9e
Mrozek, Michal