# AOT ID: ['0_inference']
from ctypes import c_void_p, c_long
import torch
import math
import random
import os
import tempfile
from math import inf, nan
from torch._inductor.hooks import run_intermediate_hooks
from torch._inductor.utils import maybe_profile
from torch._inductor.codegen.memory_planning import _align as align

from torch import device, empty_strided
from torch._inductor.codecache import AsyncCompile
from torch._inductor.select_algorithm import extern_kernels
from torch._inductor.codegen.multi_kernel import MultiKernelCall

aten = torch.ops.aten
inductor_ops = torch.ops.inductor
_quantized = torch.ops._quantized
assert_size_stride = torch._C._dynamo.guards.assert_size_stride
empty_strided_cpu = torch._C._dynamo.guards._empty_strided_cpu
empty_strided_cuda = torch._C._dynamo.guards._empty_strided_cuda
alloc_from_pool = torch.ops.inductor._alloc_from_pool
reinterpret_tensor = torch.ops.inductor._reinterpret_tensor
async_compile = AsyncCompile()


# kernel path: /tmp/torchinductor_isuruf/gw/cgwpfgkdk4yia3bpt37vqjdjda6rp635xory4heq5y2ea46iaqyt.py
# Source Nodes: [arange, b, m, m_1, m_2], Original ATen: [aten.arange, aten.mul, aten.view]
# arange => iota
# b => view
# m => mul
# m_1 => mul_1
# m_2 => mul_2
triton_poi_fused_arange_mul_view_0 = async_compile.triton('triton_', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor

from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor

@triton_heuristics.pointwise(
    size_hints=[256], 
    filename=__file__,
    triton_meta={'signature': {0: '*i64', 1: 'i32'}, 'device': 0, 'device_type': 'cuda', 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0,), equal_to_1=())]},
    inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_arange_mul_view_0', 'mutated_arg_names': [], 'no_x_dim': False, 'backend_hash': '13d5379970a55f2f2c4bb8dbeb907c03d2af7e5fb1c9d1b1aa5bf5794d5f2277', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
    min_elem_per_thread=0
)
@triton.jit
def triton_(out_ptr0, xnumel, XBLOCK : tl.constexpr):
    xnumel = 200
    xoffset = tl.program_id(0) * XBLOCK
    xindex = xoffset + tl.arange(0, XBLOCK)[:]
    xmask = xindex < xnumel
    x0 = xindex
    tmp0 = x0*x0*x0
    tl.store(out_ptr0 + (x0), tmp0, xmask)
''', device_str='cuda')

import triton
import triton.language as tl
from torch._inductor.runtime.triton_heuristics import grid, split_scan_grid, start_graph, end_graph
from torch._C import _cuda_getCurrentRawStream as get_raw_stream


# kernel path: /tmp/torchinductor_isuruf/vx/cvx4wnrm6ktqek4r6kmhrqqb3nlid4qanygazjhwmvx26svg4yl7.py
# Source Nodes: [add_2, add_3, add_4, mul_3, mul_4, mul_5, t1, t1_1], Original ATen: [aten.add, aten.mul]
# add_2 => add_2
# add_3 => add_3
# add_4 => add_4
# mul_3 => mul_3
# mul_4 => mul_4
# mul_5 => mul_5
# t1 => add
# t1_1 => add_1
triton_poi_fused_add_mul_1 = async_compile.triton('triton_', '''
import triton
import triton.language as tl
from triton.compiler.compiler import AttrsDescriptor

from torch._inductor.runtime import triton_helpers, triton_heuristics
from torch._inductor.runtime.triton_helpers import libdevice, math as tl_math
from torch._inductor.runtime.hints import AutotuneHint, ReductionHint, TileHint, instance_descriptor

@triton_heuristics.pointwise(
    size_hints=[8388608], 
    filename=__file__,
    triton_meta={'signature': {0: '*fp32', 1: '*fp32', 2: 'i32'}, 'device': 0, 'device_type': 'cuda', 'constants': {}, 'configs': [AttrsDescriptor(divisible_by_16=(0, 1, 2), equal_to_1=())]},
    inductor_meta={'autotune_hints': set(), 'kernel_name': 'triton_poi_fused_add_mul_1', 'mutated_arg_names': [], 'no_x_dim': False, 'backend_hash': '13d5379970a55f2f2c4bb8dbeb907c03d2af7e5fb1c9d1b1aa5bf5794d5f2277', 'are_deterministic_algorithms_enabled': False, 'assert_indirect_indexing': True, 'autotune_local_cache': True, 'autotune_pointwise': True, 'autotune_remote_cache': False, 'dynamic_scale_rblock': True, 'max_autotune': False, 'max_autotune_pointwise': False, 'min_split_scan_rblock': 256, 'spill_threshold': 16, 'store_cubin': False},
    min_elem_per_thread=0
)
@triton.jit
def triton_(in_ptr0, out_ptr0, xnumel, XBLOCK : tl.constexpr):
    xnumel = 8000000
    xoffset = tl.program_id(0) * XBLOCK
    xindex = xoffset + tl.arange(0, XBLOCK)[:]
    xmask = xindex < xnumel
    x2 = xindex
    x0 = xindex % 200
    tmp0 = tl.load(in_ptr0 + (x2), xmask)
    tmp1 = x0*x0*x0
    tmp2 = tmp1.to(tl.float32)
    tmp3 = tmp0 * tmp2
    tmp4 = 0.0
    tmp5 = tmp3 + tmp4
    tmp6 = 1.0
    tmp7 = tmp0 + tmp6
    tmp8 = tmp7 * tmp2
    tmp9 = tmp5 + tmp8
    tmp10 = 2.0
    tmp11 = tmp0 + tmp10
    tmp12 = tmp11 * tmp2
    tmp13 = tmp9 + tmp12
    tl.store(out_ptr0 + (x2), tmp13, xmask)
''', device_str='cuda')


async_compile.wait(globals())
del async_compile

def call(args):
    arg0_1, = args
    args.clear()
    assert_size_stride(arg0_1, (200, 200, 200), (40000, 200, 1))
    with torch.cuda._DeviceGuard(0):
        torch.cuda.set_device(0)
        buf0 = empty_strided_cuda((1, 1, 200), (200, 200, 1), torch.int64)
        # Source Nodes: [arange, b, m, m_1, m_2], Original ATen: [aten.arange, aten.mul, aten.view]
        stream0 = get_raw_stream(0)
        triton_poi_fused_arange_mul_view_0.run(buf0, 200, grid=grid(200), stream=stream0)
        buf1 = empty_strided_cuda((200, 200, 200), (40000, 200, 1), torch.float32)
        # Source Nodes: [add_2, add_3, add_4, mul_3, mul_4, mul_5, t1, t1_1], Original ATen: [aten.add, aten.mul]
        triton_poi_fused_add_mul_1.run(arg0_1, buf1, 8000000, grid=grid(8000000), stream=stream0)
        del arg0_1
    return (buf0, buf1, )


def benchmark_compiled_module(times=10, repeat=10):
    from torch._dynamo.testing import rand_strided
    from torch._inductor.utils import print_performance
    arg0_1 = rand_strided((200, 200, 200), (40000, 200, 1), device='cuda:0', dtype=torch.float32)
    fn = lambda: call([arg0_1])
    return print_performance(fn, times=times, repeat=repeat)


if __name__ == "__main__":
    from torch._inductor.wrapper_benchmark import compiled_module_main
    compiled_module_main('None', benchmark_compiled_module)