Skip to content

[SYCL][CUDA] Fix and cleanup more CUDA LIT fails #1303

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Apr 9, 2020
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion sycl/plugins/cuda/CMakeLists.txt
Original file line number Diff line number Diff line change
@@ -4,7 +4,7 @@ message(STATUS "Including the PI API CUDA backend.")
# we only require the CUDA driver API to be used
# CUDA_CUDA_LIBRARY variable defines the path to libcuda.so, the CUDA Driver API library.

find_package(CUDA 10.0 REQUIRED)
find_package(CUDA 10.1 REQUIRED)

add_library(cudadrv SHARED IMPORTED)

4 changes: 3 additions & 1 deletion sycl/test/aot/gpu.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// REQUIRES: ocloc, gpu
// UNSUPPORTED: cuda
// CUDA is not compatible with SPIR.

// RUN: %clangxx -fsycl -fsycl-targets=spir64_gen-unknown-unknown-sycldevice -Xsycl-target-backend=spir64_gen-unknown-unknown-sycldevice "-device skl" %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// XFAIL: cuda

//==----- gpu.cpp - AOT compilation for gen devices using GEN compiler ------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
2 changes: 2 additions & 0 deletions sycl/test/aot/multiple-devices.cpp
Original file line number Diff line number Diff line change
@@ -7,6 +7,8 @@
//===------------------------------------------------------------------------===//

// REQUIRES: opencl-aot, ocloc, aoc, cpu, gpu, accelerator
// UNSUPPORTED: cuda
// CUDA is not compatible with SPIR.

// 1-command compilation case
// Targeting CPU, GPU, FPGA
9 changes: 5 additions & 4 deletions sycl/test/basic_tests/buffer/subbuffer.cpp
Original file line number Diff line number Diff line change
@@ -1,11 +1,12 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// XFAIL: cuda
// TODO: Fix CUDA implementation.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// XFAIL: cuda
// TODO: cuda fail due to unimplemented param_name 4121 in cuda_piDeviceGetInfo

//
//==---------- subbuffer.cpp --- sub-buffer basic test ---------------------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
6 changes: 3 additions & 3 deletions sycl/test/basic_tests/handler/handler_copy_with_offset.cpp
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// RUN: %clangxx -fsycl %s -o %t.out
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
@@ -48,7 +48,7 @@ int main() {

vector_class<char> Expected{'x', 'x', '0', '1', '2', '3', 'x', 'x'};
if (DataRaw != Expected)
throw std::runtime_error("Check of hadler.copy(ptr, acc) was failed");
throw std::runtime_error("Check of handler.copy(ptr, acc) was failed");
}

{
@@ -71,7 +71,7 @@ int main() {
}
vector_class<char> Expected{'2', '3', '4', '5', 'x', 'x', 'x', 'x'};
if (DataRaw != Expected)
throw std::runtime_error("Check of hadler.copy(acc, ptr) was failed");
throw std::runtime_error("Check of handler.copy(acc, ptr) was failed");
}
return 0;
}
2 changes: 1 addition & 1 deletion sycl/test/basic_tests/handler/interop_task.cpp
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
// REQUIRES: opencl
// RUN: %clangxx -fsycl %s -o %t.out -L %opencl_libs_dir -lOpenCL
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// REQUIRES: opencl

//==------- interop_task.cpp -----------------------------------------------==//
//
32 changes: 11 additions & 21 deletions sycl/test/basic_tests/kernel_info.cpp
Original file line number Diff line number Diff line change
@@ -12,28 +12,18 @@
//===----------------------------------------------------------------------===//

#include <CL/sycl.hpp>
#include <cassert>

using namespace cl::sycl;

void check(bool condition, const char *conditionString, const char *filename,
const long line) noexcept {
if (!condition) {
std::cerr << "CHECK failed in " << filename << "#" << line << " "
<< conditionString << "\n";
std::abort();
}
}

#define CHECK(CONDITION) check(CONDITION, #CONDITION, __FILE__, __LINE__)

int main() {
queue q;

buffer<int, 1> buf(range<1>(1));
program prg(q.get_context());

prg.build_with_kernel_type<class SingleTask>();
CHECK(prg.has_kernel<class SingleTask>());
assert(prg.has_kernel<class SingleTask>());
kernel krn = prg.get_kernel<class SingleTask>();

q.submit([&](handler &cgh) {
@@ -42,26 +32,26 @@ int main() {
});

const string_class krnName = krn.get_info<info::kernel::function_name>();
CHECK(!krnName.empty());
assert(!krnName.empty());
const cl_uint krnArgCount = krn.get_info<info::kernel::num_args>();
CHECK(krnArgCount > 0);
assert(krnArgCount > 0);
const context krnCtx = krn.get_info<info::kernel::context>();
CHECK(krnCtx == q.get_context());
assert(krnCtx == q.get_context());
const program krnPrg = krn.get_info<info::kernel::program>();
CHECK(krnPrg == prg);
assert(krnPrg == prg);
const cl_uint krnRefCount = krn.get_info<info::kernel::reference_count>();
CHECK(krnRefCount > 0);
assert(krnRefCount > 0);
const string_class krnAttr = krn.get_info<info::kernel::attributes>();
CHECK(krnAttr.empty());
assert(krnAttr.empty());

device dev = q.get_device();
const size_t wgSize =
krn.get_work_group_info<info::kernel_work_group::work_group_size>(dev);
CHECK(wgSize > 0);
assert(wgSize > 0);
const size_t prefWGSizeMult = krn.get_work_group_info<
info::kernel_work_group::preferred_work_group_size_multiple>(dev);
CHECK(prefWGSizeMult > 0);
assert(prefWGSizeMult > 0);
const cl_ulong prvMemSize =
krn.get_work_group_info<info::kernel_work_group::private_mem_size>(dev);
CHECK(prvMemSize == 0);
assert(prvMemSize == 0);
}
6 changes: 4 additions & 2 deletions sycl/test/basic_tests/parallel_for_range.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,10 @@
// RUN: %clangxx -fsycl %s -o %t.out
// XFAIL: cuda
// CUDA exposes broken hierarchical parallelism.

// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// XFAIL: cuda

#include <CL/sycl.hpp>

3 changes: 0 additions & 3 deletions sycl/test/basic_tests/sampler/sampler.cpp
Original file line number Diff line number Diff line change
@@ -6,9 +6,6 @@
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out

// TODO: Image support in CUDA backend
// XFAIL: cuda

//==--------------- sampler.cpp - SYCL sampler basic test ------------------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
2 changes: 1 addition & 1 deletion sycl/test/fpga_tests/fpga_pipes.cpp
Original file line number Diff line number Diff line change
@@ -3,7 +3,7 @@
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Does it pass on CUDA device?

Copy link
Contributor Author

@bjoernknafla bjoernknafla Apr 9, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Would you prefer to mark all the fpga tests as not supported by CUDA? I am not entirely sure if the test semantics only work for accelerators anyway, in which case I could also quickly add REQUIRED: accelerator to all of them.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm fine with this change actually.
We probably should have some common guidance for test behavior in unsupported environment and align all the tests in a separate PR.
+@MrSidims


//==------------- fpga_pipes.cpp - SYCL FPGA pipes test --------------------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
2 changes: 1 addition & 1 deletion sycl/test/fpga_tests/fpga_queue.cpp
Original file line number Diff line number Diff line change
@@ -5,7 +5,7 @@
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==------------- fpga_queue.cpp - SYCL FPGA queues test -------------------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
8 changes: 5 additions & 3 deletions sycl/test/function-pointers/fp-as-kernel-arg.cpp
Original file line number Diff line number Diff line change
@@ -1,11 +1,13 @@
// RUN: %clangxx -Xclang -fsycl-allow-func-ptr -std=c++14 -fsycl %s -o %t.out -L %opencl_libs_dir -lOpenCL
// UNSUPPORTED: windows
// UNSUPPORTED: cuda
// CUDA does not support the function pointer as kernel argument extension.

// RUN: %clangxx -Xclang -fsycl-allow-func-ptr -std=c++14 -fsycl %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// FIXME: This test should use runtime early exit once correct check for
// corresponding extension is implemented
// UNSUPPORTED: windows
// XFAIL: cuda

#include <CL/sycl.hpp>

8 changes: 5 additions & 3 deletions sycl/test/function-pointers/pass-fp-through-buffer.cpp
Original file line number Diff line number Diff line change
@@ -1,11 +1,13 @@
// RUN: %clangxx -Xclang -fsycl-allow-func-ptr -std=c++14 -fsycl %s -o %t.out -L %opencl_libs_dir -lOpenCL
// UNSUPPORTED: windows
// UNSUPPORTED: cuda
// CUDA does not support the function pointer as kernel argument extension.

// RUN: %clangxx -Xclang -fsycl-allow-func-ptr -std=c++14 -fsycl %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// FIXME: This test should use runtime early exit once correct check for
// corresponding extension is implemented
// UNSUPPORTED: windows
// XFAIL: cuda

#include <CL/sycl.hpp>

4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/all_of.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/any_of.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/broadcast.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/exclusive_scan.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/inclusive_scan.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
3 changes: 3 additions & 0 deletions sycl/test/group-algorithm/leader.cpp
Original file line number Diff line number Diff line change
@@ -1,3 +1,6 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/none_of.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
4 changes: 3 additions & 1 deletion sycl/test/group-algorithm/reduce.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// OpenCL C 2.x alike work-group functions not yet supported by CUDA.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>
#include <algorithm>
3 changes: 0 additions & 3 deletions sycl/test/hier_par/hier_par_wgscope.cpp
Original file line number Diff line number Diff line change
@@ -18,9 +18,6 @@
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out

// TODO: ptxas fatal : Unresolved extern function '__spirv_ControlBarrier'
// UNSUPPORTED: cuda

// This test checks correctness of hierarchical kernel execution when there is
// code and data in the work group scope.

3 changes: 3 additions & 0 deletions sycl/test/kernel_from_file/hw.cpp
Original file line number Diff line number Diff line change
@@ -1,3 +1,6 @@
// UNSUPPORTED: cuda
// CUDA does not support SPIR-V.

//-fsycl-targets=%sycl_triple
// RUN: %clangxx -fsycl-device-only -fno-sycl-use-bitcode -Xclang -fsycl-int-header=%t.h -c %s -o %t.spv -I %sycl_include -Xclang -verify-ignore-unexpected=note,warning -Wno-sycl-strict
// RUN: %clangxx -include %t.h -g %s -o %t.out -lsycl -I %sycl_include -Xclang -verify-ignore-unexpected=note,warning
2 changes: 1 addition & 1 deletion sycl/test/linear_id/linear-sub_group.cpp
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// RUN: %clangxx -fsycl %s -o %t.out
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
3 changes: 2 additions & 1 deletion sycl/test/linear_id/opencl-interop.cpp
Original file line number Diff line number Diff line change
@@ -4,9 +4,10 @@
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==---------------- opencl-interop.cpp - SYCL linear id test --------------==//
//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
2 changes: 1 addition & 1 deletion sycl/test/lit.cfg.py
Original file line number Diff line number Diff line change
@@ -72,8 +72,8 @@

config.substitutions.append( ('%sycl_libs_dir', config.sycl_libs_dir ) )
config.substitutions.append( ('%sycl_include', config.sycl_include ) )
config.substitutions.append( ('%opencl_libs_dir', config.opencl_libs_dir) )
config.substitutions.append( ('%sycl_source_dir', config.sycl_source_dir) )
config.substitutions.append( ('%opencl_libs_dir', config.opencl_libs_dir) )
config.substitutions.append( ('%opencl_include_dir', config.opencl_include_dir) )
config.substitutions.append( ('%cuda_toolkit_include', config.cuda_toolkit_include) )

4 changes: 2 additions & 2 deletions sycl/test/lit.site.cfg.py.in
Original file line number Diff line number Diff line change
@@ -3,15 +3,15 @@
import sys

config.llvm_tools_dir = "@LLVM_TOOLS_DIR@"
config.sycl_tools_dir = lit_config.params.get('SYCL_TOOLS_DIR', "@LLVM_TOOLS_DIR@")
config.lit_tools_dir = "@LLVM_LIT_TOOLS_DIR@"
config.sycl_tools_dir = lit_config.params.get('SYCL_TOOLS_DIR', "@LLVM_TOOLS_DIR@")
config.sycl_include = lit_config.params.get('SYCL_INCLUDE', "@SYCL_INCLUDE@")
config.sycl_obj_root = "@SYCL_BINARY_DIR@"
config.sycl_source_dir = "@SYCL_SOURCE_DIR@/source"
config.opencl_libs_dir = os.path.dirname("@OpenCL_LIBRARIES@")
config.sycl_libs_dir = lit_config.params.get('SYCL_LIBS_DIR', "@LLVM_LIBS_DIR@")
config.target_triple = "@TARGET_TRIPLE@"
config.host_triple = "@LLVM_HOST_TRIPLE@"
config.opencl_libs_dir = os.path.dirname("@OpenCL_LIBRARIES@")
config.opencl_include_dir = "@OpenCL_INCLUDE_DIR@"
config.cuda_toolkit_include = "@CUDA_TOOLKIT_INCLUDE@"

1 change: 0 additions & 1 deletion sycl/test/ordered_queue/ordered_dmemll.cpp
Original file line number Diff line number Diff line change
@@ -3,7 +3,6 @@
// RUN: %clangxx -fsycl %s -o %t1.out -L %opencl_libs_dir -lOpenCL
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// XFAIL: cuda
//==----------- ordered_dmemll.cpp - Device Memory Linked List test --------==//
// It uses an ordered queue where explicit waiting is not necessary between
// kernels
2 changes: 2 additions & 0 deletions sycl/test/ordered_queue/ordered_queue.cpp
Original file line number Diff line number Diff line change
@@ -1,3 +1,5 @@
// REQUIRES: opencl
//
// RUN: %clangxx -fsycl %s -o %t.out -L %opencl_libs_dir -lOpenCL
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
//==---------- ordered_queue.cpp - SYCL ordered queue test -----------------==//
3 changes: 3 additions & 0 deletions sycl/test/program_manager/env_vars.cpp
Original file line number Diff line number Diff line change
@@ -1,3 +1,6 @@
// REQUIRES: opencl
// Env vars are used to pass OpenCL-specific flags to PI compiling/linking.
//
// RUN: %clangxx -O0 -fsycl -fsycl-targets=%sycl_triple %s -o %t.out -lsycl
//
// Deprecated SYCL_PROGRAM_BUILD_OPTIONS should work as an alias to
7 changes: 2 additions & 5 deletions sycl/test/regression/image_access.cpp
Original file line number Diff line number Diff line change
@@ -5,11 +5,8 @@
// TODO: For now PI checks are skipped for ACC device. To decide if it's good.
// RUN: env %ACC_RUN_PLACEHOLDER %t.out

// TODO: No CUDA image support
// XFAIL: cuda

// TODO: No CUDA image support
// XFAIL: cuda
// UNSUPPORTED: cuda
// CUDA cannot support OpenCL spec conform images.

//==-------------- image_access.cpp - SYCL image accessors test -----------==//
//
3 changes: 1 addition & 2 deletions sycl/test/regression/private_array_init_test.cpp
Original file line number Diff line number Diff line change
@@ -1,5 +1,4 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out -lOpenCL

// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
3 changes: 0 additions & 3 deletions sycl/test/regression/static-buffer-dtor.cpp
Original file line number Diff line number Diff line change
@@ -14,9 +14,6 @@
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out

// TODO: terminate called after throwing an instance of 'cl::sycl::runtime_error'
// TODO: what(): OpenCL API failed. OpenCL API returns: -999 (Unknown OpenCL error code) -999 (Unknown OpenCL error code)

#include <CL/sycl.hpp>

int main() {
7 changes: 4 additions & 3 deletions sycl/test/scheduler/DataMovement.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,10 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -I %sycl_source_dir %s -o %t.out
// XFAIL: cuda
// TODO: Fix accidential error return when unmapping read-only memory objects.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -I %sycl_source_dir %s -o %t.out -g
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// Incorrect event callback processing for host device.
// XFAIL: cuda
//
//==-------------------------- DataMovement.cpp ----------------------------==//
//
5 changes: 4 additions & 1 deletion sycl/test/sub_group/attributes.cpp
Original file line number Diff line number Diff line change
@@ -1,4 +1,7 @@
// RUN: %clangxx -fsycl %s -o %t.out
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUNx: %GPU_RUN_PLACEHOLDER %t.out
5 changes: 4 additions & 1 deletion sycl/test/sub_group/barrier.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,12 @@
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==---------- barrier.cpp - SYCL sub_group barrier test -------*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
5 changes: 4 additions & 1 deletion sycl/test/sub_group/broadcast.cpp
Original file line number Diff line number Diff line change
@@ -1,10 +1,13 @@
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.

// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -D SG_GPU %s -o %t_gpu.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t_gpu.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==--------- broadcast.cpp - SYCL sub_group broadcast test ----*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
5 changes: 4 additions & 1 deletion sycl/test/sub_group/common.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,12 @@
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==-------------- common.cpp - SYCL sub_group common test -----*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
2 changes: 1 addition & 1 deletion sycl/test/sub_group/common_ocl.cpp
Original file line number Diff line number Diff line change
@@ -7,7 +7,7 @@
// RUN: %CPU_RUN_PLACEHOLDER %t.out %T/kernel_ocl.spv
// RUN: %GPU_RUN_PLACEHOLDER %t.out %T/kernel_ocl.spv
// RUN: %ACC_RUN_PLACEHOLDER %t.out %T/kernel_ocl.spv
// UNSUPPORTED: cuda

//==--- common_ocl.cpp - basic SG methods in SYCL vs OpenCL ---*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 changes: 3 additions & 1 deletion sycl/test/sub_group/info.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// REQUIRES: opencl

// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==------------- info.cpp - SYCL sub_group parameters test ----*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 changes: 3 additions & 1 deletion sycl/test/sub_group/load_store.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda
//
//==----------- load_store.cpp - SYCL sub_group load/store test ------------==//
//
8 changes: 5 additions & 3 deletions sycl/test/sub_group/reduce.cpp
Original file line number Diff line number Diff line change
@@ -1,11 +1,13 @@
//-fsycl-targets=%sycl_triple
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -std=c++14 %s -o %t.out
// RUN: %clangxx -fsycl -std=c++14 -D SG_GPU %s -o %t_gpu.out
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -std=c++14 -D SG_GPU %s -o %t_gpu.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t_gpu.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==--------------- reduce.cpp - SYCL sub_group reduce test ----*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
8 changes: 5 additions & 3 deletions sycl/test/sub_group/scan.cpp
Original file line number Diff line number Diff line change
@@ -1,11 +1,13 @@
//-fsycl-targets=%sycl_triple
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -std=c++14 %s -o %t.out
// RUN: %clangxx -fsycl -std=c++14 -D SG_GPU %s -o %t_gpu.out
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -std=c++14 -D SG_GPU %s -o %t_gpu.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t_gpu.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==--------------- scan.cpp - SYCL sub_group scan test --------*- C++ -*---==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 changes: 3 additions & 1 deletion sycl/test/sub_group/shuffle.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,11 @@
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUNx: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda
//
//==------------ shuffle.cpp - SYCL sub_group shuffle test -----*- C++ -*---==//
//
5 changes: 4 additions & 1 deletion sycl/test/sub_group/vote.cpp
Original file line number Diff line number Diff line change
@@ -1,9 +1,12 @@
// UNSUPPORTED: cuda
// CUDA compilation and runtime do not yet support sub-groups.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// RUN: %ACC_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

//==--------------- vote.cpp - SYCL sub_group vote test --*- C++ -*---------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
6 changes: 5 additions & 1 deletion sycl/test/usm/allocator_vector.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==---- allocator_vector.cpp - Allocator Container test -------------------==//
//
6 changes: 5 additions & 1 deletion sycl/test/usm/allocator_vector_fail.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==-- allocator_vector_fail.cpp - Device Memory Allocator fail test -------==//
//
6 changes: 5 additions & 1 deletion sycl/test/usm/allocatorll.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==---- allocatorll.cpp - Device Memory Linked List Allocator test --------==//
//
9 changes: 6 additions & 3 deletions sycl/test/usm/badmalloc.cpp
Original file line number Diff line number Diff line change
@@ -1,10 +1,14 @@
// UNSUPPORTED: windows
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out

// UNSUPPORTED: windows,cuda

//==----------------- badmalloc.cpp - Bad Mallocs test ---------------------==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
@@ -13,7 +17,6 @@
//
//===----------------------------------------------------------------------===//


// This test verifies that things fail in the proper way when they should.

#include <CL/sycl.hpp>
6 changes: 5 additions & 1 deletion sycl/test/usm/depends_on.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==----------------- depends_on.cpp - depends_on test ---------------------==//
//
6 changes: 5 additions & 1 deletion sycl/test/usm/dmemll.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==------------------- dmemll.cpp - Device Memory Linked List test --------==//
//
6 changes: 5 additions & 1 deletion sycl/test/usm/hmemll.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==------------------- hmemll.cpp - Host Memory Linked List test ----------==//
//
2 changes: 2 additions & 0 deletions sycl/test/usm/math.cpp
Original file line number Diff line number Diff line change
@@ -3,6 +3,8 @@
// RUN: %CPU_RUN_PLACEHOLDER %t.out

// REQUIRES: cpu
// XFAIL: cuda
// TODO: ptxas fatal : Unresolved extern function '_Z20__spirv_ocl_lgamma_rfPi'

#include <CL/sycl.hpp>

5 changes: 4 additions & 1 deletion sycl/test/usm/memadvise.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,11 @@
// XFAIL: cuda
// SYCL runtime and piextUSM*Alloc functions for CUDA not behaving as described
// in: https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==---------------- memadvise.cpp - Shared Memory Linked List test --------==//
//
6 changes: 5 additions & 1 deletion sycl/test/usm/memcpy.cpp
Original file line number Diff line number Diff line change
@@ -5,10 +5,14 @@
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>

9 changes: 6 additions & 3 deletions sycl/test/usm/memset.cpp
Original file line number Diff line number Diff line change
@@ -1,7 +1,11 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==---- memset.cpp - USM memset test --------------------------------------==//
//
@@ -10,7 +14,6 @@
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//

#include <CL/sycl.hpp>

using namespace cl::sycl;
6 changes: 5 additions & 1 deletion sycl/test/usm/mixed.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==------------------- mixed.cpp - Mixed Memory test ---------------------==//
//
8 changes: 6 additions & 2 deletions sycl/test/usm/mixed2.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==------------------- mixed2.cpp - Mixed Memory test ---------------------==//
//
8 changes: 6 additions & 2 deletions sycl/test/usm/mixed2template.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==---------- mixed2template.cpp - Mixed Memory with Templatestest --------==//
//
8 changes: 6 additions & 2 deletions sycl/test/usm/mixed_queue.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==-------------- mixed_queue.cpp - Mixed Memory test ---------------------==//
//
4 changes: 3 additions & 1 deletion sycl/test/usm/pfor_flatten.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,10 @@
// UNSUPPORTED: cuda
// CUDA does not support the unnamed lambda extension.
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -fsycl-unnamed-lambda %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==--------------- pfor_flatten.cpp - Kernel Launch Flattening test -------==//
//
2 changes: 1 addition & 1 deletion sycl/test/usm/prefetch.cpp
Original file line number Diff line number Diff line change
@@ -5,7 +5,7 @@
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
// RUN: %clangxx -fsycl %s -o %t1.out -L %opencl_libs_dir -lOpenCL
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out

8 changes: 6 additions & 2 deletions sycl/test/usm/queue_wait.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t.out
// RUN: %CPU_RUN_PLACEHOLDER %t.out
// RUN: %GPU_RUN_PLACEHOLDER %t.out
// UNSUPPORTED: cuda

#include <CL/sycl.hpp>

6 changes: 5 additions & 1 deletion sycl/test/usm/smemll.cpp
Original file line number Diff line number Diff line change
@@ -1,8 +1,12 @@
// XFAIL: cuda
// piextUSM*Alloc functions for CUDA are not behaving as described in
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/USM.adoc
// https://github.com/intel/llvm/blob/sycl/sycl/doc/extensions/USM/cl_intel_unified_shared_memory.asciidoc
//
// RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t1.out
// RUN: env SYCL_DEVICE_TYPE=HOST %t1.out
// RUN: %CPU_RUN_PLACEHOLDER %t1.out
// RUN: %GPU_RUN_PLACEHOLDER %t1.out
// UNSUPPORTED: cuda

//==------------------- smemll.cpp - Shared Memory Linked List test --------==//
//
2 changes: 1 addition & 1 deletion sycl/test/warnings/warnings.cpp
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
// RUN: %clangxx -Wall -Wpessimizing-move -Wunused-variable -Wmismatched-tags -Wunneeded-internal-declaration -Werror -fsycl %s -o %t.out
// RUN: %clangxx -Wall -Wpessimizing-move -Wunused-variable -Wmismatched-tags -Wunneeded-internal-declaration -Werror -Wno-unknown-cuda-version -fsycl %s -o %t.out

#include <CL/sycl.hpp>

2 changes: 1 addition & 1 deletion sycl/tools/CMakeLists.txt
Original file line number Diff line number Diff line change
@@ -6,7 +6,7 @@ add_executable(get_device_count_by_type get_device_count_by_type.cpp)
add_dependencies(get_device_count_by_type ocl-headers ocl-icd)

if( SYCL_BUILD_PI_CUDA )
find_package(CUDA 10.0 REQUIRED)
find_package(CUDA 10.1 REQUIRED)

add_library(cudadrv SHARED IMPORTED)