deepmodeling / abacus-develop

An electronic structure package based on either plane wave basis or numerical atomic orbitals.
http://abacus.ustc.edu.cn
GNU Lesser General Public License v3.0
174 stars 136 forks source link

Daily DCU test failed since 20241102 #5414

Open pxlxingliang opened 2 weeks ago

pxlxingliang commented 2 weeks ago

Describe the bug

Since 20241102, the DCU test is failed because of compile error. This error is introduced by commits between 0f44046 (tested at 20241102, and is failed) and 3984296 (tested at 20241101, and is passed).

The error is related to force/stress, which may be introduced by commit e276480937.

@dyzheng , could you have a check?

Image

The error message is:

/public/home/abacus/abacus-dcu/source/module_hamilt_pw/hamilt_pwdft/kernels/rocm/stress_op.hip.cu:294:5: error: use of undeclared identifier '__syncwarp'; did you mean '__sync_swap'?
    __syncwarp();
    ^~~~~~~~~~
    __sync_swap
/public/home/abacus/abacus-dcu/source/module_hamilt_pw/hamilt_pwdft/kernels/rocm/stress_op.hip.cu:294:5: note: '__sync_swap' declared here
/public/home/abacus/abacus-dcu/source/module_hamilt_pw/hamilt_pwdft/kernels/rocm/stress_op.hip.cu:294:16: error: too few arguments to function call, expected at least 1, have 0
    __syncwarp();
    ~~~~~~~~~~ ^
[ 70%] Building CXX object source/module_io/CMakeFiles/io_basic.dir/read_wfc_to_rho.cpp.o
2 errors generated when compiling for gfx906.
[ 70%] Building CXX object source/module_cell/CMakeFiles/cell.dir/read_pp_vwr.cpp.o
CMake Error at device_rocm_generated_stress_op.hip.cu.o.cmake:180 (message):
  Error generating file
  /public/home/abacus/abacus-dcu/build-dcu/source/CMakeFiles/device_rocm.dir/module_hamilt_pw/hamilt_pwdft/kernels/rocm/./device_rocm_generated_stress_op.hip.cu.o

make[2]: *** [source/CMakeFiles/device_rocm.dir/build.make:147: source/CMakeFiles/device_rocm.dir/module_hamilt_pw/hamilt_pwdft/kernels/rocm/device_rocm_generated_stress_op.hip.cu.o] Error 1
make[2]: *** Waiting for unfinished jobs....

Expected behavior

No response

To Reproduce

No response

Environment

No response

Additional Context

No response

Task list for Issue attackers (only for developers)

dyzheng commented 4 days ago

img_v3_02gi_b08602e2-149d-45d5-9c65-26014eb6b8ag @Qianruipku there is still some compiling error of DCU after PR #5423 , would you like to fix it ?

Qianruipku commented 1 day ago

img_v3_02gi_b08602e2-149d-45d5-9c65-26014eb6b8ag @Qianruipku there is still some compiling error of DCU after PR #5423 , would you like to fix it ?

OK