[Clang][OpenMP] Switch to __kmpc_parallel_60 with strict parameter (llvm#171082)

This commit switches the `__kmpc_parallel_51` to `__kmpc_parallel_60`,
and adds the strict boolean for the number of threads.

Author: kevinsala

clang/lib/CodeGen/CGOpenMPRuntimeGPU.cpp

@@ -1291,6 +1291,9 @@ void CGOpenMPRuntimeGPU::emitParallelCall(
     else
       NumThreadsVal = Bld.CreateZExtOrTrunc(NumThreadsVal, CGF.Int32Ty);
 
+    // No strict prescriptiveness for the number of threads.
+    llvm::Value *StrictNumThreadsVal = llvm::ConstantInt::get(CGF.Int32Ty, 0);
+
     assert(IfCondVal && "Expected a value");
     llvm::Value *RTLoc = emitUpdateLocation(CGF, Loc);
     llvm::Value *Args[] = {
@@ -1303,9 +1306,11 @@ void CGOpenMPRuntimeGPU::emitParallelCall(
         ID,
         Bld.CreateBitOrPointerCast(CapturedVarsAddrs.emitRawPointer(CGF),
                                    CGF.VoidPtrPtrTy),
-        llvm::ConstantInt::get(CGM.SizeTy, CapturedVars.size())};
+        llvm::ConstantInt::get(CGM.SizeTy, CapturedVars.size()),
+        StrictNumThreadsVal};
+
     CGF.EmitRuntimeCall(OMPBuilder.getOrCreateRuntimeFunction(
-                            CGM.getModule(), OMPRTL___kmpc_parallel_51),
+                            CGM.getModule(), OMPRTL___kmpc_parallel_60),
                         Args);
   };
 

clang/test/OpenMP/amdgcn_target_device_vla.cpp

@@ -327,7 +327,7 @@ int main() {
 // CHECK-NEXT:    store ptr [[TMP1]], ptr [[TMP29]], align 8
 // CHECK-NEXT:    [[TMP30:%.*]] = load ptr, ptr [[DOTGLOBAL_TID__ADDR_ASCAST]], align 8
 // CHECK-NEXT:    [[TMP31:%.*]] = load i32, ptr [[TMP30]], align 4
-// CHECK-NEXT:    call void @__kmpc_parallel_51(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP31]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo2v_l30_omp_outlined_omp_outlined, ptr null, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 5)
+// CHECK-NEXT:    call void @__kmpc_parallel_60(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP31]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo2v_l30_omp_outlined_omp_outlined, ptr null, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 5, i32 0)
 // CHECK-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
 // CHECK:       omp.inner.for.inc:
 // CHECK-NEXT:    [[TMP32:%.*]] = load i32, ptr [[DOTOMP_IV_ASCAST]], align 4
@@ -701,7 +701,7 @@ int main() {
 // CHECK-NEXT:    store ptr [[A]], ptr [[TMP27]], align 8
 // CHECK-NEXT:    [[TMP28:%.*]] = load ptr, ptr [[DOTGLOBAL_TID__ADDR_ASCAST]], align 8
 // CHECK-NEXT:    [[TMP29:%.*]] = load i32, ptr [[TMP28]], align 4
-// CHECK-NEXT:    call void @__kmpc_parallel_51(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP29]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo3v_l52_omp_outlined_omp_outlined, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo3v_l52_omp_outlined_omp_outlined_wrapper, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 3)
+// CHECK-NEXT:    call void @__kmpc_parallel_60(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP29]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo3v_l52_omp_outlined_omp_outlined, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo3v_l52_omp_outlined_omp_outlined_wrapper, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 3, i32 0)
 // CHECK-NEXT:    store i32 0, ptr [[J_ASCAST]], align 4
 // CHECK-NEXT:    br label [[FOR_COND:%.*]]
 // CHECK:       for.cond:
@@ -1070,7 +1070,7 @@ int main() {
 // CHECK-NEXT:    store ptr [[A]], ptr [[TMP28]], align 8
 // CHECK-NEXT:    [[TMP29:%.*]] = load ptr, ptr [[DOTGLOBAL_TID__ADDR_ASCAST]], align 8
 // CHECK-NEXT:    [[TMP30:%.*]] = load i32, ptr [[TMP29]], align 4
-// CHECK-NEXT:    call void @__kmpc_parallel_51(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP30]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo4v_l76_omp_outlined_omp_outlined, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo4v_l76_omp_outlined_omp_outlined_wrapper, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 3)
+// CHECK-NEXT:    call void @__kmpc_parallel_60(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP30]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo4v_l76_omp_outlined_omp_outlined, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z4foo4v_l76_omp_outlined_omp_outlined_wrapper, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 3, i32 0)
 // CHECK-NEXT:    store i32 0, ptr [[J_ASCAST]], align 4
 // CHECK-NEXT:    br label [[FOR_COND:%.*]]
 // CHECK:       for.cond:

clang/test/OpenMP/amdgpu_target_with_aligned_attribute.c

@@ -158,7 +158,7 @@ void write_to_aligned_array(int *a, int N) {
 // CHECK-AMD-NEXT:    store ptr [[TMP19]], ptr [[TMP26]], align 8
 // CHECK-AMD-NEXT:    [[TMP27:%.*]] = load ptr, ptr [[DOTGLOBAL_TID__ADDR_ASCAST]], align 8
 // CHECK-AMD-NEXT:    [[TMP28:%.*]] = load i32, ptr [[TMP27]], align 4
-// CHECK-AMD-NEXT:    call void @__kmpc_parallel_51(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP28]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_write_to_aligned_array_l14_omp_outlined_omp_outlined, ptr null, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 4)
+// CHECK-AMD-NEXT:    call void @__kmpc_parallel_60(ptr addrspacecast (ptr addrspace(1) @[[GLOB1]] to ptr), i32 [[TMP28]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_write_to_aligned_array_l14_omp_outlined_omp_outlined, ptr null, ptr [[CAPTURED_VARS_ADDRS_ASCAST]], i64 4, i32 0)
 // CHECK-AMD-NEXT:    br label [[OMP_INNER_FOR_INC:%.*]]
 // CHECK-AMD:       omp.inner.for.inc:
 // CHECK-AMD-NEXT:    [[TMP29:%.*]] = load i32, ptr [[DOTOMP_IV_ASCAST]], align 4

clang/test/OpenMP/declare_target_codegen_globalization.cpp

@@ -40,7 +40,7 @@ int maini1() {
 // CHECK1-NEXT:    [[TMP2:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
 // CHECK1-NEXT:    [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[CAPTURED_VARS_ADDRS]], i64 0, i64 0
 // CHECK1-NEXT:    store ptr [[TMP0]], ptr [[TMP3]], align 8
-// CHECK1-NEXT:    call void @__kmpc_parallel_51(ptr @[[GLOB1]], i32 [[TMP2]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z6maini1v_l16_omp_outlined, ptr null, ptr [[CAPTURED_VARS_ADDRS]], i64 1)
+// CHECK1-NEXT:    call void @__kmpc_parallel_60(ptr @[[GLOB1]], i32 [[TMP2]], i32 1, i32 -1, i32 -1, ptr @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z6maini1v_l16_omp_outlined, ptr null, ptr [[CAPTURED_VARS_ADDRS]], i64 1, i32 0)
 // CHECK1-NEXT:    call void @__kmpc_target_deinit()
 // CHECK1-NEXT:    ret void
 // CHECK1:       worker.exit:

clang/test/OpenMP/metadirective_device_arch_codegen.cpp

@@ -49,7 +49,7 @@ int metadirective1() {
 // CHECK: %{{[0-9]}} = call{{.*}} i32 @__kmpc_target_init
 // CHECK: user_code.entry:
 // CHECK: call{{.*}} void @[[METADIRECTIVE]]_omp_outlined
-// CHECK-NOT: call{{.*}} void @__kmpc_parallel_51
+// CHECK-NOT: call{{.*}} void @__kmpc_parallel_60
 // CHECK: ret void
 
 

clang/test/OpenMP/metadirective_device_isa_codegen_amdgcn.cpp

@@ -24,7 +24,7 @@ int amdgcn_device_isa_selected() {
 
 // CHECK: define weak_odr protected amdgpu_kernel void @__omp_offloading_{{.*}}amdgcn_device_isa_selected
 // CHECK: user_code.entry:
-// CHECK: call void @__kmpc_parallel_51
+// CHECK: call void @__kmpc_parallel_60
 // CHECK-NOT: call i32 @__kmpc_single
 // CHECK: ret void
 
@@ -47,7 +47,7 @@ int amdgcn_device_isa_not_selected() {
 // CHECK: define weak_odr protected amdgpu_kernel void @__omp_offloading_{{.*}}amdgcn_device_isa_not_selected
 // CHECK: user_code.entry:
 // CHECK: call i32 @__kmpc_single
-// CHECK-NOT: call void @__kmpc_parallel_51
+// CHECK-NOT: call void @__kmpc_parallel_60
 // CHECK: ret void
 
 #endif