; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_100 -mattr=+ptx86| FileCheck --check-prefixes=CHECK-PTX64 %s
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_100 -mattr=+ptx86 --nvptx-short-ptr| FileCheck --check-prefixes=CHECK-PTX-SHARED32 %s
; RUN: %if ptxas-sm_100 && ptxas-isa-8.6 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_100 -mattr=+ptx86| %ptxas-verify -arch=sm_100 %}
; RUN: %if ptxas-sm_100 && ptxas-isa-8.6 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_100 -mattr=+ptx86 --nvptx-short-ptr| %ptxas-verify -arch=sm_100 %}

target triple = "nvptx64-nvidia-cuda"

declare void @llvm.nvvm.cp.async.bulk.shared.cta.to.global.bytemask(ptr addrspace(1), ptr addrspace(3), i32, i64, i1, i16)

define void @cp_async_bulk_s2g_bytemask(ptr addrspace(1) %dst, ptr addrspace(3) %src, i32 %size, i64 %ch, i16 %mask) {
; CHECK-PTX64-LABEL: cp_async_bulk_s2g_bytemask(
; CHECK-PTX64:       {
; CHECK-PTX64-NEXT:    .reg .b16 %rs<2>;
; CHECK-PTX64-NEXT:    .reg .b32 %r<2>;
; CHECK-PTX64-NEXT:    .reg .b64 %rd<4>;
; CHECK-PTX64-EMPTY:
; CHECK-PTX64-NEXT:  // %bb.0:
; CHECK-PTX64-NEXT:    ld.param.b64 %rd1, [cp_async_bulk_s2g_bytemask_param_0];
; CHECK-PTX64-NEXT:    ld.param.b64 %rd2, [cp_async_bulk_s2g_bytemask_param_1];
; CHECK-PTX64-NEXT:    ld.param.b32 %r1, [cp_async_bulk_s2g_bytemask_param_2];
; CHECK-PTX64-NEXT:    ld.param.b64 %rd3, [cp_async_bulk_s2g_bytemask_param_3];
; CHECK-PTX64-NEXT:    ld.param.b16 %rs1, [cp_async_bulk_s2g_bytemask_param_4];
; CHECK-PTX64-NEXT:    cp.async.bulk.global.shared::cta.bulk_group.L2::cache_hint.cp_mask [%rd1], [%rd2], %r1, %rd3, %rs1;
; CHECK-PTX64-NEXT:    cp.async.bulk.global.shared::cta.bulk_group.cp_mask [%rd1], [%rd2], %r1, %rs1;
; CHECK-PTX64-NEXT:    ret;
;
; CHECK-PTX-SHARED32-LABEL: cp_async_bulk_s2g_bytemask(
; CHECK-PTX-SHARED32:       {
; CHECK-PTX-SHARED32-NEXT:    .reg .b16 %rs<2>;
; CHECK-PTX-SHARED32-NEXT:    .reg .b32 %r<3>;
; CHECK-PTX-SHARED32-NEXT:    .reg .b64 %rd<3>;
; CHECK-PTX-SHARED32-EMPTY:
; CHECK-PTX-SHARED32-NEXT:  // %bb.0:
; CHECK-PTX-SHARED32-NEXT:    ld.param.b64 %rd1, [cp_async_bulk_s2g_bytemask_param_0];
; CHECK-PTX-SHARED32-NEXT:    ld.param.b32 %r1, [cp_async_bulk_s2g_bytemask_param_1];
; CHECK-PTX-SHARED32-NEXT:    ld.param.b32 %r2, [cp_async_bulk_s2g_bytemask_param_2];
; CHECK-PTX-SHARED32-NEXT:    ld.param.b64 %rd2, [cp_async_bulk_s2g_bytemask_param_3];
; CHECK-PTX-SHARED32-NEXT:    ld.param.b16 %rs1, [cp_async_bulk_s2g_bytemask_param_4];
; CHECK-PTX-SHARED32-NEXT:    cp.async.bulk.global.shared::cta.bulk_group.L2::cache_hint.cp_mask [%rd1], [%r1], %r2, %rd2, %rs1;
; CHECK-PTX-SHARED32-NEXT:    cp.async.bulk.global.shared::cta.bulk_group.cp_mask [%rd1], [%r1], %r2, %rs1;
; CHECK-PTX-SHARED32-NEXT:    ret;
  tail call void @llvm.nvvm.cp.async.bulk.shared.cta.to.global.bytemask(ptr addrspace(1) %dst, ptr addrspace(3) %src, i32 %size, i64 %ch, i1 1, i16 %mask)
  tail call void @llvm.nvvm.cp.async.bulk.shared.cta.to.global.bytemask(ptr addrspace(1) %dst, ptr addrspace(3) %src, i32 %size, i64 %ch, i1 0, i16 %mask)
  ret void
}