pytorch

cc01568e - [pt2] Register meta func to randperm.default (#99593)

[pt2] Register meta func to randperm.default (#99593) Summary: Looks we're missing the meta func for randperm.default. I get complaints like this when I compile randperm with dynamic shape which I think is because it gets into the real implementation but not the meta func. ``` RuntimeError: expected int but got s0 Exception raised from expect_int at fbcode/caffe2/c10/core/SymInt.h:128 (most recent call first): # 0 c10::get_backtrace[abi:cxx11](unsigned long, unsigned long, bool) # 1 std::_Function_handler<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > (), c10::(anonymous namespace)::GetFetchStackTrace()::$_1>::_M_invoke(std::_Any_data const&) # 2 c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) # 3 c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) # 4 c10::impl::wrap_kernel_functor_unboxed_<c10::impl::detail::WrapFunctionIntoFunctor_<c10::CompileTimeFunctionPointer<at::Tensor (c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>), &at::(anonymous namespace)::(anonymous namespace)::wrapper_CompositeExplicitAutograd__randperm>, at::Tensor, c10::guts::typelist::typelist<c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool> > >, at::Tensor (c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>)>::call(c10::OperatorKernel*, c10::DispatchKeySet, c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>) # 5 at::Tensor c10::Dispatcher::redispatch<at::Tensor, c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool> >(c10::TypedOperatorHandle<at::Tensor (c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>)> const&, c10::DispatchKeySet, c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>) const # 6 at::_ops::randperm::redispatch(c10::DispatchKeySet, c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>) # 7 c10::impl::wrap_kernel_functor_unboxed_<c10::impl::detail::WrapFunctionIntoFunctor_<c10::CompileTimeFunctionPointer<at::Tensor (c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>), &at::(anonymous namespace)::randperm>, at::Tensor, c10::guts::typelist::typelist<c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool> > >, at::Tensor (c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>)>::call(c10::OperatorKernel*, c10::DispatchKeySet, c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>) # 8 c10::impl::make_boxed_from_unboxed_functor<c10::impl::detail::WrapFunctionIntoFunctor_<c10::CompileTimeFunctionPointer<at::Tensor (c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool>), &at::(anonymous namespace)::randperm>, at::Tensor, c10::guts::typelist::typelist<c10::SymInt, c10::optional<c10::ScalarType>, c10::optional<c10::Layout>, c10::optional<c10::Device>, c10::optional<bool> > >, false>::call(c10::OperatorKernel*, c10::OperatorHandle const&, c10::DispatchKeySet, std::vector<c10::IValue, std::allocator<c10::IValue> >*) ``` Differential Revision: D45137851 Pull Request resolved: https://github.com/pytorch/pytorch/pull/99593 Approved by: https://github.com/ezyang