[BUG] Memcheck error in ParquetChunkedReaderInputLimitTest.Mixed

[davidwendt]

Changes to this test in PR #15672 report an out-of-bounds read error with compute-sanitizer.

Using the changes from #15672 run the following to see the error

compute-sanitizer --tool memcheck gtests/PARQUET_TEST --gtest_filter=ParquetChunkedReaderInputLimitTest.Mixed --rmm_mode=cuda

Partial result

========= COMPUTE-SANITIZER
Note: Google Test filter = ParquetChunkedReaderInputLimitTest.Mixed
[==========] Running 1 test from 1 test suite.
[----------] Global test environment set-up.
[----------] 1 test from ParquetChunkedReaderInputLimitTest
[ RUN      ] ParquetChunkedReaderInputLimitTest.Mixed
========= Invalid __global__ read of size 8 bytes
=========     at 0x7f0 in void thrust::cuda_cub::core::_kernel_agent<thrust::cuda_cub::__parallel_for::ParallelForAgent<thrust::cuda_cub::__transform::unary_transform_f<cudf::io::parquet::detail::<unnamed>::cumulative_page_info *, cudf::io::parquet::detail::<unnamed>::cumulative_page_info *, thrust::cuda_cub::__transform::no_stencil_tag, cudf::io::parquet::detail::<unnamed>::page_total_size, thrust::cuda_cub::__transform::always_true_predicate>, long>, thrust::cuda_cub::__transform::unary_transform_f<cudf::io::parquet::detail::<unnamed>::cumulative_page_info *, cudf::io::parquet::detail::<unnamed>::cumulative_page_info *, thrust::cuda_cub::__transform::no_stencil_tag, cudf::io::parquet::detail::<unnamed>::page_total_size, thrust::cuda_cub::__transform::always_true_predicate>, long>(T2, T3)
=========     by thread (219,0,0) in block (2,0,0)
=========     Address 0x7f59441e96a8 is out of bounds
=========     and is 9 bytes after the nearest allocation at 0x7f59441e0a00 of size 36,000 bytes
=========     Saved host backtrace up to driver entry point at kernel launch time
=========     Host Frame: [0x2ef370]
=========                in /usr/lib/x86_64-linux-gnu/libcuda.so.1
=========     Host Frame: [0x1488c]
=========                in /conda/envs/rapids/lib/libcudart.so.11.0
=========     Host Frame:cudaLaunchKernel [0x6c318]
=========                in /conda/envs/rapids/lib/libcudart.so.11.0
=========     Host Frame:cudaError thrust::cuda_cub::__parallel_for::parallel_for<thrust::cuda_cub::__transform::unary_transform_f<cudf::io::parquet::detail::(anonymous namespace)::cumulative_page_info*, cudf::io::parquet::detail::(anonymous namespace)::cumulative_page_info*, thrust::cuda_cub::__transform::no_stencil_tag, cudf::io::parquet::detail::(anonymous namespace)::page_total_size, thrust::cuda_cub::__transform::always_true_predicate>, long>(long, thrust::cuda_cub::__transform::unary_transform_f<cudf::io::parquet::detail::(anonymous namespace)::cumulative_page_info*, cudf::io::parquet::detail::(anonymous namespace)::cumulative_page_info*, thrust::cuda_cub::__transform::no_stencil_tag, cudf::io::parquet::detail::(anonymous namespace)::page_total_size, thrust::cuda_cub::__transform::always_true_predicate>, CUstream_st*) [0x149a79a]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:cudf::io::parquet::detail::(anonymous namespace)::adjust_cumulative_sizes(cudf::device_span<cudf::io::parquet::detail::(anonymous namespace)::cumulative_page_info const, 18446744073709551615ul>, cudf::device_span<cudf::io::parquet::detail::PageInfo const, 18446744073709551615ul>, rmm::cuda_stream_view) [0x1483e3b]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:cudf::io::parquet::detail::(anonymous namespace)::compute_next_subpass(cudf::device_span<cudf::io::parquet::detail::(anonymous namespace)::cumulative_page_info const, 18446744073709551615ul>, cudf::device_span<cudf::io::parquet::detail::PageInfo const, 18446744073709551615ul>, cudf::device_span<cudf::io::parquet::detail::ColumnChunkDesc const, 18446744073709551615ul>, cudf::device_span<int const, 18446744073709551615ul>, unsigned long, unsigned long, unsigned long, rmm::cuda_stream_view) [0x148762a]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:cudf::io::parquet::detail::reader::impl::setup_next_subpass(bool)::{lambda()#1}::operator()() const [0x148a7db]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:cudf::io::parquet::detail::reader::impl::setup_next_subpass(bool) [0x1491b75]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:cudf::io::parquet::detail::reader::impl::has_next() [0x146c362]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:cudf::io::chunked_parquet_reader::has_next() const [0x131a37f]
=========                in /conda/envs/rapids/lib/libcudf.so
=========     Host Frame:(anonymous namespace)::chunked_read(std::vector<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > > > const&, unsigned long, unsigned long) [0xe1786]
=========                in /cudf/cpp/build/gtests/PARQUET_TEST

[davidwendt]

@nvdbaranec