Skip to content

Cuda memory problem  #78

Description

@NiranjanRavi05

Hello,

I am using CuSZ compressor in my research work. I was able to run the example file capi.cu for CESM data, but when I altered it and ran for my own dataset, i got an error:

terminate called after throwing an instance of 'thrust::system::system_error'
what(): copy_if failed to synchronize: cudaErrorIllegalAddress: an illegal memory access was encountered
Aborted (core dumped)

I am running it in the GPU - Tesla V100.

commands executed:
To build SZ:
mkdir build && cd build
cmake .. -DCUSZ_BUILD_EXAMPLES=on -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTING=on -DCMAKE_CUDA_ARCHITECTURES="70" -DCMAKE_PREFIX_PATH=/usr/local/cuda-11.7/targets/x86_64-linux/lib/cmake -DCMAKE_INSTALL_PREFIX=main
make
make install
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/mnt/tank/bem-nr/cuSZ-develop/build

To compile capi.cu:
nvcc -o capi capi.cu -I/mnt/tank/bem-nr/cuSZ-develop/include -I/usr/local/cuda-11.7/include -L/usr/local/cuda-11.7/lib64 -lcudart -L/mnt/tank/bem-nr/cuSZ-develop/build -lcusz --extended-lambda -g
for CESM data: ./capi /data/cesm-CLDHGH-3600x1800
for my data set: ./capi /data/mydata

My dataset file is a binary file with floating point values in 1D array (length - 1x89999991)

I tried altering the length of my array, same as CESM data. But got the same error.

Can you please tell me why CuSZ is giving me this error? Please let me know if you need further information.

Thanks in advance!

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions