free(): invalid pointer

问题现象：运行异常, 因为free(): invalid pointer异常退出
```
npu
[W compiler_depend.ts:623] Warning: expandable_segments currently defaults to false. You can enable this feature by `export PYTORCH_NPU_ALLOC_CONF = expandable_segments:True`. (function operator())
<frozen importlib._bootstrap>:914: ImportWarning: TEMetaPathFinder.find_spec() not found; falling back to find_module()
<frozen importlib._bootstrap>:914: ImportWarning: TEMetaPathFinder.find_spec() not found; falling back to find_module()
/home/ma-user/work/zhongyunde/envs/tts_py310/lib/python3.10/tempfile.py:833: ResourceWarning: Implicitly cleaning up <TemporaryDirectory '/tmp/tmpmi3hiuc2'>
  _warnings.warn(warn_message, ResourceWarning)
free(): invalid pointer
Aborted
```

初步分析：
该代码在910b上执行通过，相应的配置https://gitee.com/ascend/pytorch/issues/IAX25A?from=project-issue，因此测试用例本身是好的。

软件版本：
-- CANN 版本：CANN 7.0.RC1
![输入图片说明](https://foruda.gitee.com/images/1729664104255666928/f1de5969_11650004.png "屏幕截图")
-- Python 版本 : Python 3.10.0
-- torch 版本 2.1.0
-- npu 类型：910B4 (npu-smi info)

日志信息：
--提供方式 
建议打包后上传到网盘中将下载链接粘贴回复内容

复现步骤：python basic1.py

测试用例：basic1.py
```
import torch
import torch_npu
from torch_npu.profiler import profile

# Define a model
class Model(torch.nn.Module):
    def __init__(self):
        super(Model, self).__init__()
        self.fc1 = torch.nn.Linear(10, 100)
        self.fc2 = torch.nn.Linear(100, 50)
        self.fc3 = torch.nn.Linear(50, 2)
    
    def forward(self, x):
        x = self.fc1(x)
        x = torch.relu(x)
        x = self.fc2(x)
        x = torch.relu(x)
        x = self.fc3(x)
        return x

# check the avail device
device = torch.device("npu" if torch.npu.is_available() else "cpu")
print(device)
data = torch.randn(100, 10)

# Ready for input
model = Model()
model.to(device)
input_data = data.to(device)

# Analysis the performance
#with profile(record_shapes=True) as prof:
output = model(input_data)
```

Ascend/pytorch
Paused

Content Risk Flag

Comments (2)

Ascend/pytorchPaused .gitee-modal { width: 500px !important; }

Content Risk Flag