[Offload] Add ol_dimensions_t and convert ranges from size_t -> uint32_t (llvm#143901)

This is a three element x, y, z size_t vector that can be used any place
where a 3D vector is required. This ensures that all vectors across
liboffload are the same and don't require any resizing/reordering
dances.

Author: RossBrunton

offload/liboffload/API/Common.td

@@ -148,6 +148,16 @@ def : Struct {
   ];
 }
 
+def : Struct {
+  let name = "ol_dimensions_t";
+  let desc = "A three element vector";
+  let members = [
+    StructMember<"uint32_t", "x", "X">,
+    StructMember<"uint32_t", "y", "Y">,
+    StructMember<"uint32_t", "z", "Z">,
+  ];
+}
+
 def : Function {
   let name = "olInit";
   let desc = "Perform initialization of the Offload library and plugins";

offload/liboffload/API/Kernel.td

@@ -29,12 +29,8 @@ def : Struct {
     let desc = "Size-related arguments for a kernel launch.";
     let members = [
         StructMember<"size_t", "Dimensions", "Number of work dimensions">,
-        StructMember<"size_t", "NumGroupsX", "Number of work groups on the X dimension">,
-        StructMember<"size_t", "NumGroupsY", "Number of work groups on the Y dimension">,
-        StructMember<"size_t", "NumGroupsZ", "Number of work groups on the Z dimension">,
-        StructMember<"size_t", "GroupSizeX", "Size of a work group on the X dimension.">,
-        StructMember<"size_t", "GroupSizeY", "Size of a work group on the Y dimension.">,
-        StructMember<"size_t", "GroupSizeZ", "Size of a work group on the Z dimension.">,
+        StructMember<"struct ol_dimensions_t", "NumGroups", "Number of work groups in each dimension">,
+        StructMember<"struct ol_dimensions_t", "GroupSize", "Size of a work group in each dimension">,
         StructMember<"size_t", "DynSharedMemory", "Size of dynamic shared memory in bytes.">
     ];
 }

offload/liboffload/src/OffloadImpl.cpp

@@ -499,12 +499,12 @@ Error olLaunchKernel_impl(ol_queue_handle_t Queue, ol_device_handle_t Device,
   auto *QueueImpl = Queue ? Queue->AsyncInfo : nullptr;
   AsyncInfoWrapperTy AsyncInfoWrapper(*DeviceImpl, QueueImpl);
   KernelArgsTy LaunchArgs{};
-  LaunchArgs.NumTeams[0] = LaunchSizeArgs->NumGroupsX;
-  LaunchArgs.NumTeams[1] = LaunchSizeArgs->NumGroupsY;
-  LaunchArgs.NumTeams[2] = LaunchSizeArgs->NumGroupsZ;
-  LaunchArgs.ThreadLimit[0] = LaunchSizeArgs->GroupSizeX;
-  LaunchArgs.ThreadLimit[1] = LaunchSizeArgs->GroupSizeY;
-  LaunchArgs.ThreadLimit[2] = LaunchSizeArgs->GroupSizeZ;
+  LaunchArgs.NumTeams[0] = LaunchSizeArgs->NumGroups.x;
+  LaunchArgs.NumTeams[1] = LaunchSizeArgs->NumGroups.y;
+  LaunchArgs.NumTeams[2] = LaunchSizeArgs->NumGroups.z;
+  LaunchArgs.ThreadLimit[0] = LaunchSizeArgs->GroupSize.x;
+  LaunchArgs.ThreadLimit[1] = LaunchSizeArgs->GroupSize.y;
+  LaunchArgs.ThreadLimit[2] = LaunchSizeArgs->GroupSize.z;
   LaunchArgs.DynCGroupMem = LaunchSizeArgs->DynSharedMemory;
 
   KernelLaunchParamsTy Params;

offload/unittests/OffloadAPI/kernel/olLaunchKernel.cpp

@@ -19,13 +19,8 @@ struct LaunchKernelTestBase : OffloadQueueTest {
                                    DeviceBin->getBufferSize(), &Program));
     ASSERT_SUCCESS(olGetKernel(Program, kernel, &Kernel));
     LaunchArgs.Dimensions = 1;
-    LaunchArgs.GroupSizeX = 64;
-    LaunchArgs.GroupSizeY = 1;
-    LaunchArgs.GroupSizeZ = 1;
-
-    LaunchArgs.NumGroupsX = 1;
-    LaunchArgs.NumGroupsY = 1;
-    LaunchArgs.NumGroupsZ = 1;
+    LaunchArgs.GroupSize = {64, 1, 1};
+    LaunchArgs.NumGroups = {1, 1, 1};
 
     LaunchArgs.DynSharedMemory = 0;
   }
@@ -60,7 +55,7 @@ OFFLOAD_TESTS_INSTANTIATE_DEVICE_FIXTURE(olLaunchKernelNoArgsTest);
 TEST_P(olLaunchKernelTest, Success) {
   void *Mem;
   ASSERT_SUCCESS(olMemAlloc(Device, OL_ALLOC_TYPE_MANAGED,
-                            LaunchArgs.GroupSizeX * sizeof(uint32_t), &Mem));
+                            LaunchArgs.GroupSize.x * sizeof(uint32_t), &Mem));
   struct {
     void *Mem;
   } Args{Mem};
@@ -88,7 +83,7 @@ TEST_P(olLaunchKernelNoArgsTest, Success) {
 TEST_P(olLaunchKernelTest, SuccessSynchronous) {
   void *Mem;
   ASSERT_SUCCESS(olMemAlloc(Device, OL_ALLOC_TYPE_MANAGED,
-                            LaunchArgs.GroupSizeX * sizeof(uint32_t), &Mem));
+                            LaunchArgs.GroupSize.x * sizeof(uint32_t), &Mem));
 
   struct {
     void *Mem;