ttnn.pad fails with error "on device padding does not support front padding"

[vigneshkeerthivasanx]

Describe the bug ttnn.pad fails with error "on device padding does not support front padding" for the test cases provided below in the code snippet.

To Reproduce Steps to reproduce the behavior:

1. Save the below snippet to a file

   
   import torch
   import pytest 
   import ttnn
   from tests.ttnn.utils_for_testing import assert_with_pcc

@pytest.mark.parametrize("h", [16, 128, 64, 32]) @pytest.mark.parametrize( "padding,torch_padding", [ (((1, 2), (1, 2)), (1, 2, 1, 2)), ], ) @pytest.mark.parametrize("value", [0]) def test_pad_blazepose(device, h, padding, torch_padding, value): torch.manual_seed(0) w = h torch_input_tensor = torch.rand((1, 128, h, w), dtype=torch.bfloat16) torch_output_tensor = torch.nn.functional.pad(torch_input_tensor, torch_padding, mode="constant", value=value)

input_tensor = ttnn.from_torch(torch_input_tensor, layout=ttnn.ROW_MAJOR_LAYOUT)
input_tensor = ttnn.from_device(input_tensor)
output_tensor = ttnn.pad(input_tensor, padding=padding, value=value)

output_tensor = ttnn.to_torch(output_tensor)
assert output_tensor.shape == torch_output_tensor.shape

assert_with_pcc(torch_output_tensor, output_tensor, 0.9999)

2.Run the script using pytest `<path/to/file>`
3.See error

**Screenshots**
The following error is faced when the above code is tested.

E RuntimeError: TT_FATAL @ ../ttnn/cpp/ttnn/operations/data_movement/pad/pad.cpp:92: front_padding_is_zero E info: E ttnn.pad: on device padding does not support front padding

[vigneshkeerthivasanx]

cc: @boris-drazic @saichandax

[saichandax]

Can we create a unit test and share the branch details here?

[vigneshkeerthivasanx]

unit test : commit