struct S_atomic { int x; uint a; uint b; }; static uint local_invocation_index_1 = 0u; groupshared S_atomic wg; void compute_main_inner(uint local_invocation_index) { wg.x = 0; uint atomic_result = 0u; InterlockedExchange(wg.a, 0u, atomic_result); uint atomic_result_1 = 0u; InterlockedExchange(wg.b, 0u, atomic_result_1); GroupMemoryBarrierWithGroupSync(); uint atomic_result_2 = 0u; InterlockedExchange(wg.a, 1u, atomic_result_2); uint atomic_result_3 = 0u; InterlockedExchange(wg.b, 2u, atomic_result_3); return; } void compute_main_1() { const uint x_39 = local_invocation_index_1; compute_main_inner(x_39); return; } struct tint_symbol_1 { uint local_invocation_index_1_param : SV_GroupIndex; }; void compute_main_inner_1(uint local_invocation_index_1_param) { { wg.x = 0; uint atomic_result_4 = 0u; InterlockedExchange(wg.a, 0u, atomic_result_4); uint atomic_result_5 = 0u; InterlockedExchange(wg.b, 0u, atomic_result_5); } GroupMemoryBarrierWithGroupSync(); local_invocation_index_1 = local_invocation_index_1_param; compute_main_1(); } [numthreads(1, 1, 1)] void compute_main(tint_symbol_1 tint_symbol) { compute_main_inner_1(tint_symbol.local_invocation_index_1_param); return; }