Merging-in the OpenCL Scaling code from the OpenCL branch to trunk.

Patch originally by the Multicoreware Inc team, followed by improvements and fixes by Micheal Wootton from AMD Inc,

OpenCL:
This patch implements Bicubic Scaling in OpenCL.  
Note that HandBrake currently uses Lanczos so the performance difference appears to be much more significant. We may offer an option of BiCubic in software later.
Bicubic scaling may appear a bit sharper than the equivalent Lanczos encode and may increase file size a bit. Quality may be better or worse depending on the scaling and content and personal preference towards sharpness.
When comparing performance with a custom HandBrake build that runs Software Bicubic to OpenCL Bicubic, performance increase is about 5~7% on average on a modern GPU.


Hardware Decode via DXVA:
We also have optional DXVA decoding which may come in useful for slower/lower end systems that have a capable GPU.
This is only available on input sources that use the libav decode path.
Most GPU hardware for decoding is designed for playback, so if you are running on a high end CPU, it will bottleneck the encode process.

Requires OpenCL 1.1 or later supporting GPU.

Front end changes and testing framework are not included in this patch. This will be resolved later.

Patch will be revised further before the UI is implemented.

git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@5792 b64f7644-9d1e-0410-96f1-a4d463321fa5

1 files changed, 89 insertions, 0 deletions

diff --git a/libhb/openclwrapper.h b/libhb/openclwrapper.h
new file mode 100644
index 000000000..8436c3a9c
--- /dev/null
+++ b/libhb/openclwrapper.h
@@ -0,0 +1,89 @@
+/* openclwrapper.h

+

+   Copyright (c) 2003-2012 HandBrake Team

+   This file is part of the HandBrake source code

+   Homepage: <http://handbrake.fr/>.

+   It may be used under the terms of the GNU General Public License v2.

+   For full terms see the file COPYING file or visit http://www.gnu.org/licenses/gpl-2.0.html

+

+   Authors: Peng Gao <[email protected]> <http://www.multicorewareinc.com/>

+            Li   Cao <[email protected]> <http://www.multicorewareinc.com/>

+

+

+ */

+#ifndef __OPENCL_WRAPPER_H

+#define __OPENCL_WRAPPER_H

+#ifdef USE_OPENCL

+#include "common.h"

+

+//support AMD opencl

+#define CL_QUEUE_THREAD_HANDLE_AMD 0x403E

+#define CL_MAP_WRITE_INVALIDATE_REGION (1 << 2)

+

+typedef struct _KernelEnv

+{

+    cl_context context;

+    cl_command_queue command_queue;

+    cl_program program;

+    cl_kernel kernel;

+    char kernel_name[150];

+    int isAMD;

+}KernelEnv;

+

+typedef struct _OpenCLEnv

+{

+    cl_platform_id platform;

+    cl_context   context;

+    cl_device_id devices;

+    cl_command_queue command_queue;

+}OpenCLEnv;

+

+

+//user defined, this is function wrapper which is used to set the input parameters ,

+//luanch kernel and copy data from GPU to CPU or CPU to GPU.

+typedef int (*cl_kernel_function)( void **userdata, KernelEnv *kenv );

+

+// registe a wapper for running the kernel specified by the kernel name

+int hb_register_kernel_wrapper( const char *kernel_name, cl_kernel_function function );

+

+// run kernel , user call this function to luanch kernel.

+// kernel_name: this kernel name is used to find the kernel in opencl runtime environment

+// userdata: this userdata is the all parameters for running the kernel specified by kernel name

+int hb_run_kernel( const char *kernel_name, void **userdata );

+

+// init the run time environment , this function must be called befor calling any function related to opencl

+// the argc must be set zero , argv must be set NULL, build_option is the options for build the kernel.

+int hb_init_opencl_run_env( int argc, char **argv, const char *build_option );

+

+//relase all resource about the opencl , this function must be called after calling any functions related to opencl

+int hb_release_opencl_run_env();

+

+// get the opencl status , 0: not init ; 1, inited; this function is used the check whether or not the opencl run time has been created

+int hb_opencl_stats();

+

+// update opencl run time environments , such as commandqueue , platforme, context. program

+int hb_init_opencl_attr( OpenCLEnv * env );

+

+// create kernel object  by a kernel name on the specified opencl run time indicated by env parameter

+int hb_create_kernel( char * kernelname, KernelEnv * env );

+

+// release kernel object which is generated by calling the hb_create_kernel api
+int hb_release_kernel( KernelEnv * env );
+
+void hb_opencl_init();
+
+int hb_get_opencl_env();
+
+int hb_create_buffer(cl_mem *cl_Buf,int flags,int size);
+
+int hb_read_opencl_buffer(cl_mem cl_inBuf,unsigned char *outbuf,int size);
+
+int hb_cl_create_mapped_buffer(cl_mem *mem, unsigned char **addr, int size);
+
+int hb_cl_free_mapped_buffer(cl_mem mem, unsigned char *addr);
+
+int hb_use_buffers();
+
+int hb_confirm_gpu_type();
+#endif
+#endif