Invalid input/output static shapes in test cases

Hi all,
I’ve worked with @hanhan to add static verification for linalg operation to check input/output static shaped operands. And I found some invalid shapes in IREE tests.
For example:
%res = “mhlo.convolution”(%inputs, %weights) {
batch_group_count = 1 : i64,
dimension_numbers = {
input_batch_dimension = 0 : i64,
input_feature_dimension = 3 : i64,
input_spatial_dimensions = dense<[1, 2]> : tensor<2xi64>,
kernel_input_feature_dimension = 2 : i64,
kernel_output_feature_dimension = 3 : i64,
kernel_spatial_dimensions = dense<[0, 1]> : tensor<2xi64>,
output_batch_dimension = 0 : i64,
output_feature_dimension = 3 : i64,
output_spatial_dimensions = dense<[1, 2]> : tensor<2xi64>},
feature_group_count = 1 : i64,
rhs_dilation = dense<1> : tensor<2xi64>,
window_strides = dense<1> : tensor<2xi64>} : (tensor<1x4x5x2xf32>, tensor<3x2x2x1xf32>) → tensor<1x2x3x1xf32>
I think the input shape should be 1x4x4x2xf32, or the output shape should be 1x2x4x1xf32 to make it valid.

The problem was that the directories under iree/test/e2e/ had also invalid tests. I heard tests under the directory got generated from frontends of IREE such as tensorflow. And Hanhan found the shapes in TF dialect was invalid… So, the shapes might be already invalid before converting a ML model to a TF dialect because MHLO pass is generated from python programs through TF dialect to HLO dialect.
He thought this issue would be in input/output shapes configuration or tensorflow shape inference system.
Thanks