[wip][auto_scheduler] buffer support, correctness check

[antinucleon]

This PR enables correctness check for a generated schedule.

It is useful for:

• Metal / ROCM: For some invalid schedule, driver may skip it instead of return any errors (which will show as an impossble large FLOPS number)
• Sparse kernel search

Example 1: correctness check. This will generate random buffers for correctness check

    if train_flag:
      #measure_ctx = auto_scheduler.LocalRPCMeasureContext(min_repeat_ms=300)
      measure_runner = auto_scheduler.RPCRunner("m1", "127.0.0.1", 9190, min_repeat_ms=300, timeout=30, repeat=3)
      tune_option = auto_scheduler.TuningOptions(
          num_measure_trials=1500, 
          check_correctness=True,
          builder_n_parallel=1,
          runner=measure_runner,
          measure_callbacks=[auto_scheduler.RecordToFile(log_file)],
          verbose=2,
      )

      sch, args = auto_scheduler.auto_schedule(task, tuning_options=tune_option)

Example 2: Sparse tuning, this will register given buffers for measure.

if train_flag:
      #measure_ctx = auto_scheduler.LocalRPCMeasureContext(min_repeat_ms=300)
      measure_runner = auto_scheduler.RPCRunner("m1", "127.0.0.1", 9190, min_repeat_ms=300, timeout=30, repeat=3)
      tune_option = auto_scheduler.TuningOptions(
          num_measure_trials=1500, 
          #runner=measure_ctx.runner,
          check_correctness=False,
          builder_n_parallel=1,
          runner=measure_runner,
          measure_callbacks=[auto_scheduler.RecordToFile(log_file)],
          verbose=2,
      )
      for k, v in BUFFER.items():
          tune_option.register_buffer(k, v)

cc @merrymercy @FrozenGene

[merrymercy]

I recently refactored the task interface (#7028).
The auto_schedule.py file is deleted. Please rebase.

[antinucleon]

@FrozenGene @merrymercy Updated.

[FrozenGene]

I think you want to update submodule? So you have change showing you have changed the vta-hw submodule.

[antinucleon]

I think you want to update submodule? So you have change showing you have changed the vta-hw submodule.

Updated

[FrozenGene]

Seems we still omit here of -1, there is no place to have logic to handle -1 now.

[antinucleon]

This file is no longer used according to #7053 (comment) I think @merrymercy forgets to delete it.

[FrozenGene]

Ok.

[merrymercy]

Yeah. I forget to delete it. Fixed by #7071

[antinucleon]

I find out this solution only work for a single task, but doesn't work with TaskScheduler. Will update it later.