Skip to content

Install the ABACUS failed with DCU on Sugon platform #6009

@goodhandy

Description

@goodhandy

Describe the bug

I try to install abacus (the latest version: v3.9.0.1) with DCU on Sugon platform with below commands:

module purge
module load compiler/devtoolset/7.3.1
module load compiler/dtk/24.04.1
module load compiler/cmake/3.27.7
module load mpi/hpcx/gcc-7.3.1

git clone https://gitee.com/deepmodeling/abacus-develop  
cd abacus-develop

CC=clang CXX=clang++ cmake -B build -DUSE_OPENMP=ON -DENABLE_LCAO=OFF \
-DFFTW3_DIR=/public/home/my_username/fftw-3.3.10/build/ \
-DLAPACK_DIR=/public/home/my_username/OpenBLAS-0.3.21/ \
-DSCALAPACK_DIR=/public/home/my_username/scalapack-2.2.0/ \
-DUSE_ROCM=ON 

cd build
make

But when I ran the make command, it failed with the following error message:

[  0%] Building CXX object source/module_parameter/CMakeFiles/parameter.dir/parameter.cpp.o
[  0%] Built target parameter
[  1%] Building CXX object source/CMakeFiles/driver.dir/driver.cpp.o
[  1%] Building CXX object source/CMakeFiles/driver.dir/driver_run.cpp.o
[  1%] Built target driver
[  1%] Building HIPCC object source/CMakeFiles/device_rocm.dir/module_hsolver/kernels/rocm/device_rocm_generated_dngvd_op.hip.cu.o
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
[  2%] Building HIPCC object source/CMakeFiles/device_rocm.dir/module_base/kernels/rocm/device_rocm_generated_math_kernel_op.hip.cu.o
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
[  2%] Building HIPCC object source/CMakeFiles/device_rocm.dir/module_base/kernels/rocm/device_rocm_generated_math_op.hip.cu.o
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
[  2%] Building HIPCC object source/CMakeFiles/device_rocm.dir/module_base/module_device/rocm/device_rocm_generated_memory_op.hip.cu.o
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
dcc: warning: argument unused during compilation: '-msram-ecc' [-Wunused-command-line-argument]
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:111:99: error: use of undeclared identifier 'dev_out'
            synchronize_memory_op<FPTYPE_out, base_device::DEVICE_GPU, base_device::DEVICE_CPU>()(dev_out,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:112:99: error: use of undeclared identifier 'dev_in'
                                                                                                  dev_in,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:138:99: error: use of undeclared identifier 'dev_out'
            synchronize_memory_op<FPTYPE_out, base_device::DEVICE_CPU, base_device::DEVICE_GPU>()(dev_out,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:139:99: error: use of undeclared identifier 'dev_in'
                                                                                                  dev_in,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:111:99: error: use of undeclared identifier 'dev_out'
            synchronize_memory_op<FPTYPE_out, base_device::DEVICE_GPU, base_device::DEVICE_CPU>()(dev_out,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:112:99: error: use of undeclared identifier 'dev_in'
                                                                                                  dev_in,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:111:99: error: use of undeclared identifier 'dev_out'
            synchronize_memory_op<FPTYPE_out, base_device::DEVICE_GPU, base_device::DEVICE_CPU>()(dev_out,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:112:99: error: use of undeclared identifier 'dev_in'
                                                                                                  dev_in,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:138:99: error: use of undeclared identifier 'dev_out'
            synchronize_memory_op<FPTYPE_out, base_device::DEVICE_CPU, base_device::DEVICE_GPU>()(dev_out,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:139:99: error: use of undeclared identifier 'dev_in'
                                                                                                  dev_in,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:138:99: error: use of undeclared identifier 'dev_out'
            synchronize_memory_op<FPTYPE_out, base_device::DEVICE_CPU, base_device::DEVICE_GPU>()(dev_out,
                                                                                                  ^
/public/home/ack87coae1/abacus-develop/source/module_base/module_device/rocm/memory_op.hip.cu:139:99: error: use of undeclared identifier 'dev_in'
                                                                                                  dev_in,
                                                                                                  ^
4 errors generated when compiling for gfx928.
4 errors generated when compiling for gfx906.
4 errors generated when compiling for gfx926.
CMake Error at device_rocm_generated_memory_op.hip.cu.o.cmake:180 (message):
  Error generating file
  /public/home/ack87coae1/abacus-develop/build/source/CMakeFiles/device_rocm.dir/module_base/module_device/rocm/./device_rocm_generated_memory_op.hip.cu.o


make[2]: *** [source/CMakeFiles/device_rocm.dir/build.make:140:source/CMakeFiles/device_rocm.dir/module_base/module_device/rocm/device_rocm_generated_memory_op.hip.cu.o] 错误 1
make[1]: *** [CMakeFiles/Makefile2:903:source/CMakeFiles/device_rocm.dir/all] 错误 2
make: *** [Makefile:136:all] 错误 2

Expected behavior

No response

To Reproduce

No response

Environment

Environment

module list
Currently Loaded Modulefiles:
  1) compiler/devtoolset/7.3.1   2) compiler/dtk/24.04.1        3) compiler/cmake/3.27.7       4) mpi/hpcx/gcc-7.3.1

Dependencies:

fftw-3.3.10
OpenBLAS-0.3.1
scalapack-2.2.0 

Additional Context

No response

Task list for Issue attackers (only for developers)

  • Verify the issue is not a duplicate.
  • Describe the bug.
  • Steps to reproduce.
  • Expected behavior.
  • Error message.
  • Environment details.
  • Additional context.
  • Assign a priority level (low, medium, high, urgent).
  • Assign the issue to a team member.
  • Label the issue with relevant tags.
  • Identify possible related issues.
  • Create a unit test or automated test to reproduce the bug (if applicable).
  • Fix the bug.
  • Test the fix.
  • Update documentation (if necessary).
  • Close the issue and inform the reporter (if applicable).

Metadata

Metadata

Assignees

Labels

GPU & DCU & HPCGPU and DCU and HPC related any issues

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions