-DLARGE_BOX running error

[BaldinSM]

Hello! Many thanks for your work on this software! I'm thrilled to try new functionality of large box docking, unfortunately I faced a strange error after compilation of AutoDock-Vina-GPU-2.1. If compile with -DSMALL_BOX flag everything is fine and test docking with 2bm2_config.txt run without errors with box sizes up to 30x30x30, but when I compiled with flag -DLARGE_BOX it compiled ok but during running of the test config it always give me this error:
"An error occurred: Grid too large! Define a large box (see readme) would help."
no matter what box size is actually in config file 20x20x20, 30x30x30, 70x70x70, 100x100x100
Tried to compile with boost 1.77 and boost 1.84 - the error always occurred

Here is a full output
"WARNING: The search space volume > 27000 Angstrom^3 (See FAQ)
Using virtual sreening mode

Output will be in the directory ./test_out
Reading input ... done.
Setting up the scoring function ... done.
Using heuristic search_depth
Analyzing the binding site ... done.
GPU Platform: NVIDIA CUDA
GPU Device: NVIDIA GeForce RTX 2080 SUPER
Using random seed: -560651292

Build kernel 1 from source
OpenCL version: 3.0

Build kernel 2 from source
OpenCL version: 3.0

An error occurred: Grid too large! Define a large box (see readme) would help."

I found out that this error is written in lib/main_procedure_cl.cpp line 285
grids_cl* grids_ptr = (grids_cl*)malloc(sizeof(grids_cl)); if (grids_ptr == nullptr) {throw std::runtime_error("Grid too large! Define a large box (see readme) would help");}
but I can't understand what it is

This is my Makefile header:
Need to be modified according to different users
WORK_DIR=/home/server/Software/Vina-GPU-2.1-Large-Box/Vina-GPU-2.1/AutoDock-Vina-GPU-2.1
BOOST_LIB_PATH=/usr/lib/boost_1_84_0
OPENCL_LIB_PATH=/usr/local/cuda
OPENCL_VERSION=-DOPENCL_3_0
GPU_PLATFORM=-DNVIDIA_PLATFORM
DOCKING_BOX_SIZE=-DLARGE_BOX

Ubuntu 22.04, CUDA 12.0, BOOST 1.77 or 1.84

So I'm not really sure what is wrong.

[Glinttsd]

Hi, the compilation with -DLARGE_BOX includes two necessary steps: First, make source and run the test example; Second, make and run your example.