using pipeline_builder shared pointer multiple times lead to segfaults

**Describe the bug:**
Segfault during the pipeline creation

**Describe how to reproduce:**
```python
from fairseq2.data import read_sequence
from fairseq2.data.data_pipeline import DataPipeline, DataPipelineBuilder

pipeline_build = read_sequence(list(range(100))) # this's shared for two shuffling operations

concat_pipe = DataPipeline.concat([pipeline_build.shuffle(10).and_return(),
pipeline_build.shuffle(10).and_return()]).and_return()
# this should have raised
next(iter(concat_pipe))
```

**Describe the expected behavior:**
This should raise an explicit RuntimeError