[0-size Tensor Job2 No.11] Add 0-size Tensor support for paddle.expand_as

[co63oc]

PR Category

Operator Mechanism

PR Types

Bug fixes

Description

[0-size Tensor No.302] Add 0-size Tensor support for paddle.Tensor.expand_as
[0-size Tensor Job2 No.11] Add 0-size Tensor support for paddle.expand_as [fluid_ops]

修改前向和反向，反向填充0
infermeta没有修改
kernel 修改cpu/gpu/xpu，xpu无反向

torch测试反向

import torch x=torch.randn([3, 1]) y=torch.randn([3, 0]) x.requires_grad=True result = x.expand_as(y) result.sum().backward() print(x.grad) tensor([[0.], [0.], [0.]]) 

PaddleAPITest 测试通过

test/legacy_test/test_solve_op.py 运行时有调用data 启用static提示，增加 enable_static，static测试输入形状有误修改。
solve调用expand_as kernel时，会传入out没有初始化，所以在expand_as kernel中是判断x和y，然后再判断target_shape为0返回

[paddle-bot]

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

[paddle-ci-bot]

Sorry to inform you that 34ae4c9's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.

[co63oc]

/re-run all-failed

[co63oc]

/re-run all-failed

[co63oc]

@DanielSun11 CI已完成需要review @luotao1

[DanielSun11]

LGTM