[Arith][TIR] Recognize empty extents

[kparzysz-quic]

Generate empty interval sets when empty extents are encountered.
Handle empty regions when constructing ScheduleState.

The direct cause of the crash was constructing an expression 0 - pos_inf: the latter operand is actually of type handle. The root cause of the issue was that the region connected to T.grid(0, 8, 8) was not explicitly recognized as empty early enough, and the extent of 0 was simply used in the related arithmetic to compute further expressions.

Without the fix, the testcase fails with the following traceback:

[15:08:28] /w/src/aitools/tvm-upstream/src/tir/op/op.cc:174: 0 pos_inf
Traceback (most recent call last):
  File "grid0.py", line 27, in <module>
    sch = tvm.tir.Schedule(foo)
  File "/w/src/aitools/tvm-upstream/python/tvm/tir/schedule/_type_checker.py", line 340, in wrap
    return func(*args, **kwargs)
  File "/w/src/aitools/tvm-upstream/python/tvm/tir/schedule/schedule.py", line 160, in __init__
    self.__init_handle_by_constructor__(
  File "tvm/_ffi/_cython/./object.pxi", line 132, in tvm._ffi._cy3.core.ObjectBase.__init_handle_by_constructor__
  File "tvm/_ffi/_cython/./packed_func.pxi", line 287, in tvm._ffi._cy3.core.ConstructorCall
  File "tvm/_ffi/_cython/./packed_func.pxi", line 276, in tvm._ffi._cy3.core.FuncCall
  File "tvm/_ffi/_cython/./base.pxi", line 181, in tvm._ffi._cy3.core.CHECK_CALL
tvm._ffi.base.TVMError: Traceback (most recent call last):
  12: TVMFuncCall
  11: tvm::runtime::PackedFuncObj::Extractor<tvm::runtime::PackedFuncSubObj<void tvm::runtime::TypedPackedFunc<tvm::tir::Schedule (tvm::IRModule, long, int, int, bool)>::AssignTypedLambda<tvm::tir::$_5>(tvm::tir::$_5, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char>>)::'lambda'(tvm::runtime::TVMArgs const&, tvm::runtime::TVMRetValue*)>>::Call(tvm::runtime::PackedFuncObj const*, tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*)
  10: tvm::tir::Schedule::Traced(tvm::IRModule, long, int, tvm::tir::ScheduleErrorRenderLevel, bool)
  9: tvm::tir::ScheduleState::ScheduleState(tvm::IRModule, int, bool)
  8: tvm::tir::StateCreator::Create(tvm::IRModule, int, bool)
  7: tvm::tir::BlockInfoCollector::VisitStmt_(tvm::tir::BlockRealizeNode const*)
  6: tvm::tir::BlockInfoCollector::MakeBlockInfo(tvm::tir::StmtSRef)
  5: tvm::tir::BlockInfoCollector::CheckRegionCoverAndStagePipeline(tvm::tir::BlockInfo const&, tvm::tir::StmtSRef const&, tvm::runtime::Array<tvm::tir::StmtSRef, void> const&)
  4: tvm::tir::ProducerCoversConsumer(tvm::runtime::Array<tvm::PrimExpr, void> const&, tvm::runtime::Array<tvm::arith::IntSet, void> const&, tvm::runtime::Array<tvm::arith::IntSet, void> const&, tvm::arith::Analyzer*)
  3: tvm::operator-(tvm::PrimExpr, tvm::PrimExpr)
  2: tvm::sub(tvm::PrimExpr, tvm::PrimExpr, tvm::Span)
  1: tvm::BinaryOpMatchTypes(tvm::PrimExpr&, tvm::PrimExpr&, tvm::Span)
  0: _ZN3tvm7runtime6detail
  File "/w/src/aitools/tvm-upstream/src/tir/op/op.cc", line 175
TVMError: Cannot match type int32 vs handle

[tvm-bot]

Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.

• cc @Hzfengsy, @junrushao, @quic-sanirudh, @shingjan _{See #10317 for details}

_{Generated by tvm-bot}

[tvm-bot]

Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.

• cc @Hzfengsy, @junrushao, @quic-sanirudh, @shingjan _{See #10317 for details}

_{Generated by tvm-bot}

[tvm-bot]

Thanks for contributing to TVM! Please refer to the contributing guidelines https://tvm.apache.org/docs/contribute/ for useful information and tips. Please request code reviews from Reviewers by @-ing them in a comment.

• cc @Hzfengsy, @junrushao, @quic-sanirudh, @shingjan _{See #10317 for details}

_{Generated by tvm-bot}