Append sycl flags for cutlass in 2025.2.x (#233)

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>

Author: sywangyi

build2cmake/src/templates/xpu/dep-cutlass-sycl.cmake

@@ -1,12 +1,9 @@
 find_package(CutlassSycl)
 
+set(CUTLASS_SYCL_REVISION "v{{ version }}" CACHE STRING "CUTLASS revision to use")
 if (NOT CutlassSycl_FOUND)
   set(CUTLASS_ENABLE_HEADERS_ONLY ON CACHE BOOL "Enable only the header library")
   set(CUTLASS_ENABLE_BENCHMARKS OFF CACHE BOOL "Disable CUTLASS Benchmarks")
-
-# Set CUTLASS_REVISION manually -- its revision detection doesn't work in this case.
-  set(CUTLASS_REVISION "v{{ version }}" CACHE STRING "CUTLASS revision to use")
-
 # Use the specified CUTLASS source directory for compilation if CUTLASS_SYCL_SRC_DIR is provided
   if (DEFINED ENV{CUTLASS_SYCL_SRC_DIR})
     set(CUTLASS_SYCL_SRC_DIR $ENV{CUTLASS_SYCL_SRC_DIR})
@@ -22,7 +19,7 @@ if (NOT CutlassSycl_FOUND)
     FetchContent_Declare(
         cutlass
         GIT_REPOSITORY https://github.com/intel/cutlass-sycl.git
-        GIT_TAG ${CUTLASS_REVISION}
+        GIT_TAG ${CUTLASS_SYCL_REVISION}
         GIT_PROGRESS TRUE
 
         # Speed up CUTLASS download by retrieving only the specified GIT_TAG instead of the history.
@@ -54,8 +51,15 @@ else()
   include_directories(${CUTLASS_INCLUDE_DIR})
   include_directories(${CUTLASS_TOOLS_UTIL_INCLUDE_DIR})
 endif(NOT CutlassSycl_FOUND)
+if(CUTLASS_SYCL_REVISION MATCHES "^v3\\.9")
+  add_compile_definitions(OLD_API=1)
+endif()
+
 string(REPLACE "-fsycl-targets=spir64_gen,spir64" "-fsycl-targets=intel_gpu_pvc" sycl_link_flags "${sycl_link_flags}")
 string(REPLACE "-device pvc,xe-lpg,ats-m150" "" sycl_link_flags "${sycl_link_flags}")
-string(APPEND sycl_link_flags "-Xspirv-translator;-spirv-ext=+SPV_INTEL_split_barrier;")
+string(APPEND sycl_link_flags "-Xspirv-translator;-spirv-ext=+SPV_INTEL_split_barrier")
+if(CUTLASS_SYCL_REVISION STREQUAL "v0.5")
+  string(APPEND sycl_link_flags ",+SPV_INTEL_2d_block_io,+SPV_INTEL_subgroup_matrix_multiply_accumulate")
+endif()
 string(REPLACE "-fsycl-targets=spir64_gen,spir64" "-fsycl-targets=intel_gpu_pvc" sycl_flags "${sycl_flags}")
 

build2cmake/src/torch/xpu.rs

@@ -202,6 +202,7 @@ fn render_deps(env: &Environment, build: &Build, write: &mut impl Write) -> Resu
                 let version = match dpcpp_version.as_str() {
                     "2025.0" => "3.9-0.2",
                     "2025.1" => "3.9-0.3",
+                    "2025.2" => "0.5",
                     _ => bail!(
                         "No cutlass_sycl version mapped for DPCPP_VERSION {}",
                         dpcpp_version

examples/cutlass-gemm/gemm_sycl.cpp

@@ -166,5 +166,9 @@ void cutlass_gemm(torch::Tensor &out, torch::Tensor const &A, torch::Tensor cons
   TORCH_CHECK(gemm_op.can_implement(arguments) == cutlass::Status::kSuccess, "Invalid GEMM problem size or configuration");
   CUTLASS_CHECK(gemm_op.initialize(arguments, workspace.get()));
   CUTLASS_CHECK(gemm_op.run());
+#if defined(OLD_API)
   syclcompat::wait();
+#else
+  compat::wait();
+#endif
 }