Differential D1999 Diff 6699 intern/cycles/kernel/kernel_compat_cuda.h

Changeset View

Standalone View

intern/cycles/kernel/kernel_compat_cuda.h

	Show First 20 Lines • Show All 61 Lines • ▼ Show 20 Lines
	typedef texture<uint4, 1> texture_uint4;			typedef texture<uint4, 1> texture_uint4;
	typedef texture<uchar4, 1> texture_uchar4;			typedef texture<uchar4, 1> texture_uchar4;
	typedef texture<float4, 2> texture_image_float4;			typedef texture<float4, 2> texture_image_float4;
	typedef texture<float4, 3> texture_image3d_float4;			typedef texture<float4, 3> texture_image3d_float4;
	typedef texture<uchar4, 2, cudaReadModeNormalizedFloat> texture_image_uchar4;			typedef texture<uchar4, 2, cudaReadModeNormalizedFloat> texture_image_uchar4;

	/* Macros to handle different memory storage on different devices */			/* Macros to handle different memory storage on different devices */

	/* In order to use full 6GB of memory on Titan cards, use arrays instead			/* On Fermi cards (4xx and 5xx), we use regular textures for both data and images.
	* of textures. On earlier cards this seems slower, but on Titan it is			* On Kepler (6xx) and above, we use Bindless Textures for images and arrays for data.
	* actually slightly faster in tests. */			*
				* Arrays are necessary in order to use the full VRAM on newer cards, and it's slightly faster.
				* Using Arrays on Fermi turned out to be slower.*/

				/* Fermi */
	#if __CUDA_ARCH__ < 300			#if __CUDA_ARCH__ < 300
	# define __KERNEL_CUDA_TEX_STORAGE__			# define __KERNEL_CUDA_TEX_STORAGE__
	#endif

	#ifdef __KERNEL_CUDA_TEX_STORAGE__
	# define kernel_tex_fetch(t, index) tex1Dfetch(t, index)			# define kernel_tex_fetch(t, index) tex1Dfetch(t, index)

				# define kernel_tex_image_interp(t, x, y) tex2D(t, x, y)
				# define kernel_tex_image_interp_3d(t, x, y, z) tex3D(t, x, y, z)

				/* Kepler */
	#else			#else
	# define kernel_tex_fetch(t, index) t[(index)]			# define kernel_tex_fetch(t, index) t[(index)]

				# define kernel_tex_image_interp_float4(t, x, y) tex2D<float4>(t, x, y)
				# define kernel_tex_image_interp_float(t, x, y) tex2D<float>(t, x, y)
				# define kernel_tex_image_interp_3d_float4(t, x, y, z) tex3D<float4>(t, x, y, z)
				# define kernel_tex_image_interp_3d_float(t, x, y, z) tex3D<float>(t, x, y, z)
	#endif			#endif
	#define kernel_tex_image_interp(t, x, y) tex2D(t, x, y)
	#define kernel_tex_image_interp_3d(t, x, y, z) tex3D(t, x, y, z)

	#define kernel_data __data			#define kernel_data __data

	/* Use fast math functions */			/* Use fast math functions */

	#define cosf(x) __cosf(((float)(x)))			#define cosf(x) __cosf(((float)(x)))
	#define sinf(x) __sinf(((float)(x)))			#define sinf(x) __sinf(((float)(x)))
	#define powf(x, y) __powf(((float)(x)), ((float)(y)))			#define powf(x, y) __powf(((float)(x)), ((float)(y)))
	#define tanf(x) __tanf(((float)(x)))			#define tanf(x) __tanf(((float)(x)))
	#define logf(x) __logf(((float)(x)))			#define logf(x) __logf(((float)(x)))
	#define expf(x) __expf(((float)(x)))			#define expf(x) __expf(((float)(x)))

	#endif /* __KERNEL_COMPAT_CUDA_H__ */			#endif /* __KERNEL_COMPAT_CUDA_H__ */