diff options
author | Xavier Hallade <xavier.hallade@intel.com> | 2022-06-29 13:58:04 +0300 |
---|---|---|
committer | Xavier Hallade <xavier.hallade@intel.com> | 2022-06-29 13:58:04 +0300 |
commit | a02992f1313811c9905e44dc95a0aee31d707f67 (patch) | |
tree | 2d1f59524e2a298bb530ad578a2d2b9e2c4a1432 /intern/cycles/util/math.h | |
parent | 302b04a5a3fc0e767ac784424f78ce2edf5d2844 (diff) |
Cycles: Add support for rendering on Intel GPUs using oneAPI
This patch adds a new Cycles device with similar functionality to the
existing GPU devices. Kernel compilation and runtime interaction happen
via oneAPI DPC++ compiler and SYCL API.
This implementation is primarly focusing on Intel® Arc™ GPUs and other
future Intel GPUs. The first supported drivers are 101.1660 on Windows
and 22.10.22597 on Linux.
The necessary tools for compilation are:
- A SYCL compiler such as oneAPI DPC++ compiler or
https://github.com/intel/llvm
- Intel® oneAPI Level Zero which is used for low level device queries:
https://github.com/oneapi-src/level-zero
- To optionally generate prebuilt graphics binaries: Intel® Graphics
Compiler All are included in Linux precompiled libraries on svn:
https://svn.blender.org/svnroot/bf-blender/trunk/lib The same goes for
Windows precompiled binaries but for the graphics compiler, available
as "Intel® Graphics Offline Compiler for OpenCL™ Code" from
https://www.intel.com/content/www/us/en/developer/articles/tool/oneapi-standalone-components.html,
for which path can be set as OCLOC_INSTALL_DIR.
Being based on the open SYCL standard, this implementation could also be
extended to run on other compatible non-Intel hardware in the future.
Reviewed By: sergey, brecht
Differential Revision: https://developer.blender.org/D15254
Co-authored-by: Nikita Sirgienko <nikita.sirgienko@intel.com>
Co-authored-by: Stefan Werner <stefan.werner@intel.com>
Diffstat (limited to 'intern/cycles/util/math.h')
-rw-r--r-- | intern/cycles/util/math.h | 18 |
1 files changed, 15 insertions, 3 deletions
diff --git a/intern/cycles/util/math.h b/intern/cycles/util/math.h index f1f627588c5..af2f1ea092d 100644 --- a/intern/cycles/util/math.h +++ b/intern/cycles/util/math.h @@ -79,7 +79,7 @@ CCL_NAMESPACE_BEGIN /* Scalar */ -#ifndef __HIP__ +#if !defined(__HIP__) && !defined(__KERNEL_ONEAPI__) # ifdef _WIN32 ccl_device_inline float fmaxf(float a, float b) { @@ -92,12 +92,18 @@ ccl_device_inline float fminf(float a, float b) } # endif /* _WIN32 */ -#endif /* __HIP__ */ +#endif /* __HIP__, __KERNEL_ONEAPI__ */ -#ifndef __KERNEL_GPU__ +#if !defined(__KERNEL_GPU__) || defined(__KERNEL_ONEAPI__) +# ifndef __KERNEL_ONEAPI__ using std::isfinite; using std::isnan; using std::sqrt; +# else +using sycl::sqrt; +# define isfinite(x) sycl::isfinite((x)) +# define isnan(x) sycl::isnan((x)) +# endif ccl_device_inline int abs(int x) { @@ -793,6 +799,8 @@ ccl_device_inline uint popcount(uint x) return i & 1; } # endif +#elif defined(__KERNEL_ONEAPI__) +# define popcount(x) sycl::popcount(x) #elif defined(__KERNEL_HIP__) /* Use popcll to support 64-bit wave for pre-RDNA AMD GPUs */ # define popcount(x) __popcll(x) @@ -806,6 +814,8 @@ ccl_device_inline uint count_leading_zeros(uint x) return __clz(x); #elif defined(__KERNEL_METAL__) return clz(x); +#elif defined(__KERNEL_ONEAPI__) + return sycl::clz(x); #else assert(x != 0); # ifdef _MSC_VER @@ -824,6 +834,8 @@ ccl_device_inline uint count_trailing_zeros(uint x) return (__ffs(x) - 1); #elif defined(__KERNEL_METAL__) return ctz(x); +#elif defined(__KERNEL_ONEAPI__) + return sycl::ctz(x); #else assert(x != 0); # ifdef _MSC_VER |