OpenCL failed to build program executable on Intel GPUs

[pujaltes]

Thank you very much for your work on this tool.

We've had no problem compiling and building the OpenCL kernels of NVIDIA GPUs and Intel CPUs on Linux, but we're having issues getting it working on Intel GPUs (1550 Max).

Interestingly the error is slightly different between different runs (see below) with the same seed. I'm not very experienced with OpenCL but I imagine that this means that is due to some memory management error (i.e. undefined behavior)?

Has anyone managed to get this running on Intel GPUs? I would appreciate if anyone could point me in the direction of what is likely causing the bug. It would be fantastic if we could get Vina-GPU running on more types of hardware.

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: Intel(R) OpenCL GraphicsGPU Platform 1 version: OpenCL 3.0 

GPU Device: Intel(R) Data Center GPU Max 1550Platform 1 global memory size:130.535129 GB
Platform 1 local memory size:131.072000 KB
Create context success!
Create queue success!

Using random seed: -2024131559

Build kernel 1 from source
OpenCL version: 3.0

Build program success!  Program kernel number: 1

Build kernel 2 from source
OpenCL version: 3.0

Error: Failed to build program executable!
log:2:1430:1: error: type specifier missing, defaults to 'int'
i)      
^
2:1430:2: error: expected ';' after top level declarator
i)      
 ^
 ;

A second run fails at the same point with a slightly different error

Build kernel 2 from source
OpenCL version: 3.0

Error: Failed to build program executable!
log:2:1431:1: error: expected identifier or '('
<U+0010><80> 
^
2:1431:2: error: source file is not valid UTF-8
<U+0010><80> 
        ^