When I use cuda-gdb to debuge my code and set the cuda memcheck on, it is only show which of the kernel has the mistake like follows, but didn’t show the detail lines of the kernel. What can I do to find the specific line?
(cuda-gdb) set cuda memcheck on (cuda-gdb) run <HSCM3_fine.inp Starting program: /home/xll/gitlib/gitlib_xll/CFL3D/./cfl3d_seq <HSCM3_fine.inp [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". [New Thread 0x7fff85562700 (LWP 14107)] [New Thread 0x7fff84d61700 (LWP 14110)] [New Thread 0x7fff81fff700 (LWP 14113)] Illegal access to address (@global)0x7fff655fc2f0 detected. Program received signal CUDA_EXCEPTION_1, Lane Illegal Address. [Switching focus to CUDA kernel 0, grid 1, block (0,0,0), thread (224,0,0), device 0, sm 0, warp 7, lane 0] 0x0000000003533788 in update_kernel_do1002_<<<(255,1,1),(256,1,1)>>> () (cuda-gdb) p n No symbol "n" in current context.