… enabled (ray-project#34997)

## Why are these changes needed?

Currently, subprogress bars are not correctly rendered and updated with AllToAllOperators when the optimizer is enabled. This PR adds the subprogress bars for the different AllToAll LogicalOperators, such as `RandomShuffle`, `Sort`, and `Repartition`.

The original intent of this PR was to separate out the `sub_progress_bars_dict` from `TaskContext`, but we found that this was difficult and will require significant reworking to support it because the sub-progress bars need to be initialized and prior to being passed to the scheduler for execution.

Tested with the following code to observe the output subprogress bars (with all combinations of push/pull based shuffle, and no-shuffle for sort):
```
import ray
import time
def sleep(x):
    time.sleep(0.1)
    return x

ctx = ray.data.DataContext.get_current()
ctx.optimizer_enabled = False
ctx.use_push_based_shuffle = True
for _ in (
    ray.data.range(1000 * 1000, parallelism=200)
    .map_batches(sleep, num_cpus=2)
    #.map_batches(sleep, compute=ray.data.ActorPoolStrategy(min_size=2, max_size=4))
    .random_shuffle() # -> tested pushbased=False; pushbased=True has duplicated issue
    #.sort("id") # -> tested pushbased=True+False
    #.repartition(400, shuffle=True) # -> tested shuffle=False, pushbased=True+False
    #.map_batches(sleep, num_cpus=2)
    .iter_batches()
):
    pass
```

- `random_shuffle()`:
```
Running: 0.0/10.0 CPU, 0.0/0.0 GPU, 7.75 MiB/512.0 MiB object_store_memory:  32%|████████████████████▍                                            | 63/200 [00:24<00:37,  3.65it/s]
- RandomShuffle: 0 active, 0 queued, 0.0 MiB objects, 0 output: 100%|████████████████████████████████████████████████████████████████████████████| 200/200 [00:24<00:00, 24.36s/it]
  *- Shuffle Map:  54%|█████████████████████████████████████████████████████████████████▉                                                        | 108/200 [00:23<00:14,  6.55it/s]
  *- Shuffle Reduce:  98%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████   | 195/200 [00:24<00:00, 11.48it/s]
```

- `sort()`:
```
Running: 0.0/10.0 CPU, 0.0/0.0 GPU, 7.75 MiB/512.0 MiB object_store_memory:   0%|▎                                                               | 1/200 [00:24<1:20:59, 24.42s/it]
- Sort: 0 active, 0 queued, 0.0 MiB objects, 0 output: 100%|█████████████████████████████████████████████████████████████████████████████████████| 200/200 [00:24<00:00, 24.41s/it]
  *- Sort Sample: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 200/200 [00:23<00:00,  8.61it/s]
  *- Shuffle Map:  98%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 197/200 [00:23<00:00, 11.76it/s]
  *- Shuffle Reduce: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 200/200 [00:24<00:00, 11.67it/s]
```

- `repartition()`:
```
Running: 0.0/10.0 CPU, 0.0/0.0 GPU, 0.04 MiB/512.0 MiB object_store_memory:   0%|                                                                          | 0/400 [00:23<?, ?it/s]
- Repartition: 0 active, 0 queued, 0.0 MiB objects, 0 output:   0%|                                                                                        | 0/400 [00:00<?, ?it/s]
  *- Shuffle Map:  100%|████████████████████████████████████████████████████████████████████████████                                                                                      | 400/400 [00:23<00:40,  7.04it/s]
  *- 98%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████                                                                                                                      | 392/400 [00:24<2:45:10, 24.84s/it]
```

Signed-off-by: e428265 <arvind.chandramouli@lmco.com>