[SYCL] Remove host run and dependencies from SYCL/Scheduler tests (#1218)

This commit removes the host run and any assumptions and operations
related to the host device from the tests in SYCL/Scheduler.

Co-authored-by: Sachkov, Alexey <alexey.sachkov@intel.com>
Signed-off-by: Larsen, Steffen <steffen.larsen@intel.com>

Author: steffenlarsen

SYCL/Scheduler/BasicSchedulerTests.cpp

@@ -1,5 +1,4 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
-// RUN: %HOST_RUN_PLACEHOLDER %t.out
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out

SYCL/Scheduler/CommandCleanupThreadSafety.cpp

@@ -1,6 +1,5 @@
 // UNSUPPORTED: windows
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out -lpthread
-// RUN: %HOST_RUN_PLACEHOLDER %t.out
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out

SYCL/Scheduler/DataMovement.cpp

@@ -1,5 +1,4 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out %debug_option
-// RUN: %HOST_RUN_PLACEHOLDER %t.out
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 //
@@ -58,7 +57,6 @@ template <typename T> class CustomAllocator {
 int main() {
   TestQueue Queue1(sycl::default_selector{});
   TestQueue Queue2(sycl::default_selector{});
-  TestQueue Queue3(sycl::host_selector{});
 
   std::vector<int> Data(1);
 
@@ -83,14 +81,5 @@ int main() {
 
   { auto HostAcc = Buf.get_access<sycl_access_mode::read>(); }
 
-  Queue3.submit([&](sycl::handler &CGH) {
-    auto BufAcc = Buf.get_access<sycl_access_mode::read_write>(CGH);
-    CGH.single_task<class third_kernel>([=]() { BufAcc[0] = 43; });
-  });
-
-  Queue3.wait_and_throw();
-
-  { auto HostAcc = Buf.get_access<sycl_access_mode::read>(); }
-
   return 0;
 }

SYCL/Scheduler/InOrderQueueDeps.cpp

@@ -1,5 +1,4 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
-// RUN: env %HOST_RUN_PLACEHOLDER %t.out
 // RUN: env SYCL_PI_TRACE=2 %CPU_RUN_PLACEHOLDER %t.out 2>&1 %CPU_CHECK_PLACEHOLDER
 // RUN: env SYCL_PI_TRACE=2 %GPU_RUN_PLACEHOLDER %t.out 2>&1 %GPU_CHECK_PLACEHOLDER
 // RUN: env SYCL_PI_TRACE=2 %ACC_RUN_PLACEHOLDER %t.out 2>&1 %ACC_CHECK_PLACEHOLDER

SYCL/Scheduler/MemObjRemapping.cpp

@@ -21,15 +21,10 @@ int main() {
   std::size_t Size = 64;
   range<1> Range{Size};
   buffer<int, 1> BufA{Range};
-  buffer<int, 1> BufB{Range};
 
   Q.submit([&](handler &Cgh) {
     auto AccA = BufA.get_access<access::mode::read_write>(Cgh);
-    auto AccB = BufB.get_access<access::mode::read_write>(Cgh);
-    Cgh.parallel_for<Foo>(Range, [=](id<1> Idx) {
-      AccA[Idx] = Idx[0];
-      AccB[Idx] = Idx[0];
-    });
+    Cgh.parallel_for<Foo>(Range, [=](id<1> Idx) { AccA[Idx] = Idx[0]; });
   });
 
   {
@@ -39,16 +34,9 @@ int main() {
     // CHECK-NEXT: :
     // CHECK-NEXT: :
     // CHECK-NEXT: : 1
-    // CHECK: piEnqueueMemBufferMap
-    // CHECK-NEXT: :
-    // CHECK-NEXT: :
-    // CHECK-NEXT: :
-    // CHECK-NEXT: : 1
     auto AccA = BufA.get_access<access::mode::read>();
-    auto AccB = BufB.get_access<access::mode::read>();
     for (std::size_t I = 0; I < Size; ++I) {
       assert(AccA[I] == I);
-      assert(AccB[I] == I);
     }
   }
   {
@@ -63,23 +51,9 @@ int main() {
       AccA[I] = 2 * I;
   }
 
-  queue HostQ{host_selector()};
-  // CHECK: piEnqueueMemUnmap
-  // CHECK: piEnqueueMemBufferMap
-  // CHECK-NEXT: :
-  // CHECK-NEXT: :
-  // CHECK-NEXT: :
-  // CHECK-NEXT: : 3
-  HostQ.submit([&](handler &Cgh) {
-    auto AccB = BufB.get_access<access::mode::write>(Cgh);
-    Cgh.parallel_for<Bar>(Range, [=](id<1> Idx) { AccB[Idx] = 2 * Idx[0]; });
-  });
-
   // CHECK-NOT: piEnqueueMemBufferMap
   auto AccA = BufA.get_access<access::mode::read>();
-  auto AccB = BufB.get_access<access::mode::read>();
   for (std::size_t I = 0; I < Size; ++I) {
     assert(AccA[I] == 2 * I);
-    assert(AccB[I] == 2 * I);
   }
 }

SYCL/Scheduler/MultipleDevices.cpp

@@ -93,27 +93,10 @@ int multidevice_test(queue MyQueue1, queue MyQueue2) {
 }
 
 int main() {
-  host_selector hostSelector;
   cpu_selector CPUSelector;
   gpu_selector GPUSelector;
 
   int Result = -1;
-  try {
-    queue MyQueue1(hostSelector);
-    queue MyQueue2(hostSelector);
-    Result &= multidevice_test(MyQueue1, MyQueue2);
-  } catch (sycl::runtime_error &) {
-    std::cout << "Skipping host and host" << std::endl;
-  }
-
-  try {
-    queue MyQueue1(hostSelector);
-    queue MyQueue2(CPUSelector);
-    Result &= multidevice_test(MyQueue1, MyQueue2);
-  } catch (sycl::runtime_error &) {
-    std::cout << "Skipping host and CPU" << std::endl;
-  }
-
   try {
     queue MyQueue1(CPUSelector);
     queue MyQueue2(CPUSelector);
@@ -132,16 +115,6 @@ int main() {
     std::cout << "Skipping CPU and GPU" << std::endl;
   }
 
-  try {
-    queue MyQueue1(hostSelector);
-    queue MyQueue2(GPUSelector);
-    Result &= multidevice_test(MyQueue1, MyQueue2);
-  } catch (sycl::runtime_error &) {
-    std::cout << "Skipping host and GPU" << std::endl;
-  } catch (sycl::compile_program_error &) {
-    std::cout << "Skipping CPU and GPU" << std::endl;
-  }
-
   try {
     queue MyQueue1(GPUSelector);
     queue MyQueue2(GPUSelector);

SYCL/Scheduler/ReleaseResourcesTest.cpp

@@ -1,5 +1,4 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple -fsycl-dead-args-optimization %s -o %t.out
-// RUN: env %HOST_RUN_PLACEHOLDER %t.out
 // RUN: env SYCL_PI_TRACE=2 %CPU_RUN_PLACEHOLDER %t.out 2>&1 %CPU_CHECK_PLACEHOLDER
 // RUN: env SYCL_PI_TRACE=2 %GPU_RUN_PLACEHOLDER %t.out 2>&1 %GPU_CHECK_PLACEHOLDER
 // RUN: env SYCL_PI_TRACE=2 %ACC_RUN_PLACEHOLDER %t.out 2>&1 %ACC_CHECK_PLACEHOLDER