nvbit_at_context_init_hook(): Assertion `cudaGetLastError() == cudaSuccess' failed about nvbit HOT 7 CLOSED

The error from the execution of your vectoradd tells that your GPU driver is too old to run programs compiled with CUDA 10.2. You either need to use an older CUDA/nvcc to compile your program and nvbit tools or upgrade your GPU driver.

from nvbit.

The error from the execution of your vectoradd tells that your GPU driver is too old to run programs compiled with CUDA 10.2. You either need to use an older CUDA/nvcc to compile your program and nvbit tools or upgrade your GPU driver.

Actually the vectoradd program was compiled with CUDA 9.2 and ran fine without the LD_PRELOAD.

from nvbit.

The error from the execution of your vectoradd tells that your GPU driver is too old to run programs compiled with CUDA 10.2. You either need to use an older CUDA/nvcc to compile your program and nvbit tools or upgrade your GPU driver.

Actually the vectoradd program was compiled with CUDA 9.2 and ran fine without the LD_PRELOAD.

OK, the issue is the incompatibility between the nvbit tool and your GPU driver. The nvbit tools need to be compiled with CUDA 10.2, which requires >=v440 GPU drivers, but your driver is too old.

I would suggest you to upgrade your GPU driver to >= v440

from nvbit.

Thanks @x-y-z, I will give it a try.

from nvbit.

@x-y-z does nvbit have hard dependency on the CUDA 10.2? Our data center don't always use the latest driver. Is it possible to relax this requirement? For example, would it possible to use CUDA 10.1 (driver 418)?

from nvbit.

I was able to build and run with CUDA 10.1 (driver 418). Thanks.

------------- NVBit (NVidia Binary Instrumentation Tool v1.3.1) Loaded --------------
NVBit core environment variables (mostly for nvbit-devs):
            NVDISASM = nvdisasm - override default nvdisasm found in PATH
            NOBANNER = 0 - if set, does not print this banner
---------------------------------------------------------------------------------
         INSTR_BEGIN = 0 - Beginning of the instruction interval where to apply instrumentation
           INSTR_END = 4294967295 - End of the instruction interval where to apply instrumentation
      START_GRID_NUM = 0 - Beginning of the kernel gird launch interval where to apply instrumentation
        END_GRID_NUM = 4294967295 - End of the kernel launch interval where to apply instrumentation
    COUNT_WARP_LEVEL = 1 - Count warp level or thread level instructions
    EXCLUDE_PRED_OFF = 0 - Exclude predicated off instruction from count
   ACTIVE_FROM_START = 1 - Start instruction counting from start or wait for cuProfilerStart and cuProfilerStop
       MANGLED_NAMES = 1 - Print kernel names mangled or not
        TOOL_VERBOSE = 0 - Enable verbosity inside the tool
----------------------------------------------------------------------------------------------------

kernel 0 - _Z6vecAddPdS_S_i - #thread-blocks 98,  kernel instructions 50077, total instructions 50077
Final sum = 100000.000000; sum/n = 1.000000 (should be ~1)
Total app instructions: 50077

from nvbit.

Great. We are glad it was resolved.

from nvbit.