Hello,
when using tch and wgpu together, the program segfaults after main.
use tch;
fn main() {
let instance = wgpu::Instance::new(wgpu::InstanceDescriptor {
backends: wgpu::Backends::all(),
dx12_shader_compiler: Default::default(),
});
println!("!");
}
outputs:
!
Segmentation fault (core dumped)
I am using:
tch = "0.10.2"
wgpu = "0.15.1"
I am unsure where to post this issue, but after running valgrind, there seems to be a double free in libtorch.so?
==149631== Invalid free() / delete / delete[] / realloc()
==149631== at 0x483CA3F: free (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so)
==149631== by 0x5F518D8D: llvm::cl::opt<llvm::PGOViewCountsType, false, llvm::cl::parser<llvm::PGOViewCountsType> >::~opt() (in /home/A/Documents/C++/libtorch/lib/libtorch_cpu.so)
==149631== by 0x47247FDD: __cxa_finalize (cxa_finalize.c:83)
==149631== by 0x5AF84C32: ??? (in /home/A/Documents/C++/libtorch/lib/libtorch_cpu.so)
==149631== by 0x4011F6A: _dl_fini (dl-fini.c:138)
==149631== by 0x472478A6: __run_exit_handlers (exit.c:108)
==149631== by 0x47247A5F: exit (exit.c:139)
==149631== by 0x47225089: (below main) (libc-start.c:342)
==149631== Address 0x800000003 is not stack'd, malloc'd or (recently) free'd
Any help would be appreciated,
thanks for your time.
Hello, when using tch and wgpu together, the program segfaults after main.
outputs:
I am using: tch = "0.10.2" wgpu = "0.15.1"
I am unsure where to post this issue, but after running valgrind, there seems to be a double free in libtorch.so?
Any help would be appreciated, thanks for your time.