diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index 411a9bfe9db080b1196e2bf1583a7c51d478f941..dcc80a139b87edb14a5418f5928b16804e256c91 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -19,7 +19,7 @@ stages:
.build_template:
script:
- - pip install -I cmake==3.16.3
+ - pip install -I cmake==3.16.3 jinja2
- export NUM_CORES=$(nproc --all)
- export MAX_BUILD_CORES=$(( $(awk '( $1 == "MemTotal:" ) { print $2 }' /proc/meminfo) / ( 4 * 1024 * 1024 ) ))
- "[[ $MAX_BUILD_CORES -lt $NUM_CORES ]] && export NUM_BUILD_CORES=$MAX_BUILD_CORES || export NUM_BUILD_CORES=$NUM_CORES"
@@ -47,7 +47,7 @@ stages:
-DWALBERLA_BUILD_WITH_OPENMP=$WALBERLA_BUILD_WITH_OPENMP
-DCMAKE_BUILD_TYPE=$CMAKE_BUILD_TYPE -DMPIEXEC_PREFLAGS=$MPIEXEC_PREFLAGS
-DWALBERLA_DOUBLE_ACCURACY=$WALBERLA_DOUBLE_ACCURACY
- -DWARNING_ERROR=ON
+ -DWARNING_ERROR=$WARNING_ERROR
-DWALBERLA_BUILD_WITH_METIS=$WALBERLA_BUILD_WITH_METIS
-DWALBERLA_BUILD_WITH_PARMETIS=$WALBERLA_BUILD_WITH_PARMETIS
-DWALBERLA_BUILD_WITH_FFTW=$WALBERLA_BUILD_WITH_FFTW
@@ -77,6 +77,7 @@ stages:
WALBERLA_BUILD_WITH_FFTW: "ON"
WALBERLA_ENABLE_GUI: "OFF"
WALBERLA_LOGLEVEL: "DETAIL"
+ WARNING_ERROR: "ON"
artifacts:
when: always
reports:
@@ -93,341 +94,105 @@ stages:
-intel_19_serial:
+icc_2022_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_19_mpionly:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_19_hybrid:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_19_serial_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_19_mpionly_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_19_hybrid_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- tags:
- - cuda
- - docker
- - intel
-
-intel_19_hybrid_dbg_sp:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:19
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_DOUBLE_ACCURACY: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_serial:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_mpionly:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_hybrid:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_serial_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_mpionly_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_hybrid_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- tags:
- - cuda
- - docker
- - intel
-
-intel_20_hybrid_dbg_sp:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/intel:20
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_DOUBLE_ACCURACY: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- CMAKE_CXX_FLAGS: "-qoverride-limits"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- tags:
- - cuda
- - docker
- - intel
-
-gcc_7_serial:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
+ WARNING_ERROR: "OFF"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-gcc_7_mpionly:
+icc_2022_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
+ WARNING_ERROR: "OFF"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-gcc_7_hybrid:
+icc_2022_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
+ WARNING_ERROR: "OFF"
tags:
- - cuda
+ - cuda11
- docker
-gcc_7_serial_dbg:
+icc_2022_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
CMAKE_BUILD_TYPE: "DebugOptimized"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
+ WARNING_ERROR: "OFF"
tags:
- - cuda
+ - cuda11
- docker
-gcc_7_mpionly_dbg:
+icc_2022_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
+ WARNING_ERROR: "OFF"
tags:
- - cuda
+ - cuda11
- docker
-gcc_7_hybrid_dbg:
+icc_2022_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
+ WARNING_ERROR: "OFF"
tags:
- - cuda
+ - cuda11
- docker
-gcc_7_hybrid_dbg_sp:
+icc_2022_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:7
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icc-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_DOUBLE_ACCURACY: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
WALBERLA_BUILD_WITH_METIS: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
+ WARNING_ERROR: "OFF"
tags:
- - cuda
+ - cuda11
- docker
-gcc_8_serial:
+icx_2022_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
@@ -435,21 +200,12 @@ gcc_8_serial:
- cuda11
- docker
-gcc_8_mpionly:
+icx_2022_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
@@ -457,127 +213,67 @@ gcc_8_mpionly:
- cuda11
- docker
-gcc_8_hybrid:
+icx_2022_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
tags:
- cuda11
- docker
-gcc_8_serial_dbg:
+icx_2022_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
tags:
- cuda11
- docker
-gcc_8_mpionly_dbg:
+icx_2022_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
tags:
- cuda11
- docker
-gcc_8_hybrid_dbg:
+icx_2022_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
tags:
- cuda11
- docker
-gcc_8_hybrid_dbg_sp:
+icx_2022_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:8
- before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
- - cd python
- - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- - pip3 list
- - cd ..
- - CC=gcc CXX=g++ pip3 install cupy-cuda11x
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/icx-2022
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_DOUBLE_ACCURACY: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
WALBERLA_BUILD_WITH_METIS: "OFF"
- WALBERLA_BUILD_WITH_CODEGEN: "ON"
- WALBERLA_BUILD_WITH_PYTHON: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
tags:
- cuda11
- docker
gcc_9_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -599,9 +295,9 @@ gcc_9_serial:
gcc_9_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -621,9 +317,9 @@ gcc_9_mpionly:
gcc_9_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -642,9 +338,9 @@ gcc_9_hybrid:
gcc_9_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -667,9 +363,9 @@ gcc_9_serial_dbg:
gcc_9_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -690,9 +386,9 @@ gcc_9_mpionly_dbg:
gcc_9_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -703,18 +399,15 @@ gcc_9_hybrid_dbg:
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_BUILD_WITH_CODEGEN: "ON"
WALBERLA_BUILD_WITH_PYTHON: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
tags:
- cuda11
- docker
gcc_9_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:9
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-9
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -737,9 +430,9 @@ gcc_9_hybrid_dbg_sp:
gcc_10_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -761,9 +454,9 @@ gcc_10_serial:
gcc_10_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -783,9 +476,9 @@ gcc_10_mpionly:
gcc_10_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -804,9 +497,9 @@ gcc_10_hybrid:
gcc_10_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -829,9 +522,9 @@ gcc_10_serial_dbg:
gcc_10_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -852,9 +545,9 @@ gcc_10_mpionly_dbg:
gcc_10_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -874,9 +567,9 @@ gcc_10_hybrid_dbg:
gcc_10_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:10
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-10
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -899,9 +592,9 @@ gcc_10_hybrid_dbg_sp:
gcc_11_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -923,9 +616,9 @@ gcc_11_serial:
gcc_11_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -945,9 +638,9 @@ gcc_11_mpionly:
gcc_11_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -966,9 +659,9 @@ gcc_11_hybrid:
gcc_11_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -991,9 +684,9 @@ gcc_11_serial_dbg:
gcc_11_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1014,9 +707,9 @@ gcc_11_mpionly_dbg:
gcc_11_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1036,9 +729,9 @@ gcc_11_hybrid_dbg:
gcc_11_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:11
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-11
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1061,7 +754,7 @@ gcc_11_hybrid_dbg_sp:
gcc_12_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
@@ -1076,7 +769,7 @@ gcc_12_serial:
gcc_12_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
@@ -1089,7 +782,7 @@ gcc_12_mpionly:
gcc_12_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
stage: pretest
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
@@ -1099,7 +792,7 @@ gcc_12_hybrid:
gcc_12_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
@@ -1112,7 +805,7 @@ gcc_12_serial_dbg:
gcc_12_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
@@ -1123,7 +816,7 @@ gcc_12_mpionly_dbg:
gcc_12_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
@@ -1133,7 +826,7 @@ gcc_12_hybrid_dbg:
gcc_12_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc:12
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/gcc-12
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
@@ -1144,206 +837,170 @@ gcc_12_hybrid_dbg_sp:
- cuda11
- docker
-clang_9.0_serial:
+clang_12_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
-
-clang_9.0_mpionly:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
-
-clang_9.0_hybrid:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
-
-clang_9.0_serial_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
-
-clang_9.0_mpionly_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
-
-clang_9.0_hybrid_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- tags:
- - cuda
- - docker
-
-clang_9.0_hybrid_dbg_sp:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:9.0
- variables:
- WALBERLA_BUILD_WITH_CUDA: "ON"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_DOUBLE_ACCURACY: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - cuda
- - docker
-
-clang_10.0_serial:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-clang_10.0_mpionly:
+clang_12_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-clang_10.0_hybrid:
+clang_12_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-clang_10.0_serial_dbg:
+clang_12_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
WALBERLA_BUILD_WITH_MPI: "OFF"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
CMAKE_BUILD_TYPE: "DebugOptimized"
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-clang_10.0_mpionly_dbg:
+clang_12_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_BUILD_WITH_OPENMP: "OFF"
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-clang_10.0_hybrid_dbg:
+clang_12_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
tags:
- - cuda
+ - cuda11
- docker
-clang_10.0_hybrid_dbg_sp:
+clang_12_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:10.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-12
+ before_script:
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
+ - cd python
+ - python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
+ - pip3 list
+ - cd ..
+ - CC=gcc CXX=g++ pip3 install cupy-cuda11x
variables:
WALBERLA_BUILD_WITH_CUDA: "ON"
CMAKE_BUILD_TYPE: "DebugOptimized"
WALBERLA_DOUBLE_ACCURACY: "OFF"
WALBERLA_BUILD_WITH_PARMETIS: "OFF"
WALBERLA_BUILD_WITH_METIS: "OFF"
+ WALBERLA_BUILD_WITH_CODEGEN: "ON"
+ WALBERLA_BUILD_WITH_PYTHON: "ON"
only:
variables:
- $ENABLE_NIGHTLY_BUILDS
tags:
- - cuda
+ - cuda11
- docker
-clang_11.0_serial:
+clang_13_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1363,11 +1020,11 @@ clang_11.0_serial:
- cuda11
- docker
-clang_11.0_mpionly:
+clang_13_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1385,11 +1042,11 @@ clang_11.0_mpionly:
- cuda11
- docker
-clang_11.0_hybrid:
+clang_13_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1406,11 +1063,11 @@ clang_11.0_hybrid:
- cuda11
- docker
-clang_11.0_serial_dbg:
+clang_13_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1431,11 +1088,11 @@ clang_11.0_serial_dbg:
- cuda11
- docker
-clang_11.0_mpionly_dbg:
+clang_13_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1454,11 +1111,11 @@ clang_11.0_mpionly_dbg:
- cuda11
- docker
-clang_11.0_hybrid_dbg:
+clang_13_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1476,11 +1133,11 @@ clang_11.0_hybrid_dbg:
- cuda11
- docker
-clang_11.0_hybrid_dbg_sp:
+clang_13_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-13
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1501,11 +1158,11 @@ clang_11.0_hybrid_dbg_sp:
- cuda11
- docker
-clang_12.0_serial:
+clang_14_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1525,11 +1182,11 @@ clang_12.0_serial:
- cuda11
- docker
-clang_12.0_mpionly:
+clang_14_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1547,11 +1204,11 @@ clang_12.0_mpionly:
- cuda11
- docker
-clang_12.0_hybrid:
+clang_14_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1568,11 +1225,11 @@ clang_12.0_hybrid:
- cuda11
- docker
-clang_12.0_serial_dbg:
+clang_14_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1593,11 +1250,11 @@ clang_12.0_serial_dbg:
- cuda11
- docker
-clang_12.0_mpionly_dbg:
+clang_14_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1616,11 +1273,11 @@ clang_12.0_mpionly_dbg:
- cuda11
- docker
-clang_12.0_hybrid_dbg:
+clang_14_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1638,11 +1295,11 @@ clang_12.0_hybrid_dbg:
- cuda11
- docker
-clang_12.0_hybrid_dbg_sp:
+clang_14_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:12.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-14
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1663,11 +1320,11 @@ clang_12.0_hybrid_dbg_sp:
- cuda11
- docker
-clang_13.0_serial:
+clang_15_serial:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1687,11 +1344,11 @@ clang_13.0_serial:
- cuda11
- docker
-clang_13.0_mpionly:
+clang_15_mpionly:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1709,11 +1366,11 @@ clang_13.0_mpionly:
- cuda11
- docker
-clang_13.0_hybrid:
+clang_15_hybrid:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1727,11 +1384,11 @@ clang_13.0_hybrid:
- cuda11
- docker
-clang_13.0_serial_dbg:
+clang_15_serial_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1749,11 +1406,11 @@ clang_13.0_serial_dbg:
- cuda11
- docker
-clang_13.0_mpionly_dbg:
+clang_15_mpionly_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1769,11 +1426,11 @@ clang_13.0_mpionly_dbg:
- cuda11
- docker
-clang_13.0_hybrid_dbg:
+clang_15_hybrid_dbg:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1788,12 +1445,12 @@ clang_13.0_hybrid_dbg:
- cuda11
- docker
-clang_13.0_hybrid_dbg_sp:
+clang_15_hybrid_dbg_sp:
extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:13.0
+ image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
stage: pretest
before_script:
- - pip3 install lbmpy==1.2 jinja2 pytest
+ - pip3 install lbmpy==1.3.2 jinja2 pytest
- cd python
- python3 -m pytest --junitxml=report.xml pystencils_walberla lbmpy_walberla
- pip3 list
@@ -1811,83 +1468,6 @@ clang_13.0_hybrid_dbg_sp:
- cuda11
- docker
-inteloneapi_22.3_serial:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - docker
-
-inteloneapi_22.3_mpionly:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- only:
- variables:
- - $ENABLE_NIGHTLY_BUILDS
- tags:
- - docker
-
-inteloneapi_22.3_hybrid:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- tags:
- - docker
-
-inteloneapi_22.3_serial_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- WALBERLA_BUILD_WITH_MPI: "OFF"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- tags:
- - docker
-
-inteloneapi_22.3_mpionly_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_BUILD_WITH_OPENMP: "OFF"
- tags:
- - docker
-
-inteloneapi_22.3_hybrid_dbg:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- tags:
- - docker
-
-inteloneapi_22.3_hybrid_dbg_sp:
- extends: .build_template
- image: i10git.cs.fau.de:5005/walberla/buildenvs/inteloneapi:22.3
- variables:
- WALBERLA_BUILD_WITH_CUDA: "OFF"
- CMAKE_BUILD_TYPE: "DebugOptimized"
- WALBERLA_DOUBLE_ACCURACY: "OFF"
- WALBERLA_BUILD_WITH_PARMETIS: "OFF"
- WALBERLA_BUILD_WITH_METIS: "OFF"
- tags:
- - docker
-
gcc_8_hybrid_dbg_noboost:
@@ -1966,23 +1546,24 @@ doc:
## ##
###############################################################################
-clang-tidy:
- image: i10git.cs.fau.de:5005/walberla/buildenvs/clang:11.0
- script:
- - $CXX --version
- - clang-tidy -version
- - cmake --version
- - mkdir $CI_PROJECT_DIR/build
- - cd $CI_PROJECT_DIR/build
- - cmake .. -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWALBERLA_BUFFER_DEBUG=ON -DWALBERLA_BUILD_TESTS=ON -DWALBERLA_BUILD_BENCHMARKS=ON -DWALBERLA_BUILD_TUTORIALS=ON -DWALBERLA_BUILD_TOOLS=ON -DWALBERLA_BUILD_WITH_MPI=ON -DWALBERLA_BUILD_WITH_OPENMP=ON -DCMAKE_BUILD_TYPE=Debug -DWALBERLA_BUILD_WITH_METIS=ON -DWALBERLA_BUILD_WITH_PARMETIS=ON -DWALBERLA_BUILD_WITH_OPENMESH=ON -DWALBERLA_DOUBLE_ACCURACY=ON -DWALBERLA_LOGLEVEL=DETAIL
- - cmake . -LA
- - utilities/filterCompileCommands.py compile_commands.json
- - run-clang-tidy.py -quiet | tee clang-tidy-output.txt
- artifacts:
- paths:
- - $CI_PROJECT_DIR/build/clang-tidy-output.txt
- tags:
- - docker
+#clang-tidy:
+# image: i10git.cs.fau.de:5005/walberla/buildenvs/clang-15
+# script:
+# - $CXX --version
+# - clang-tidy -version
+# - cmake --version
+# - mkdir $CI_PROJECT_DIR/build
+# - cd $CI_PROJECT_DIR/build
+# - cmake .. -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWALBERLA_BUFFER_DEBUG=ON -DWALBERLA_BUILD_TESTS=ON -DWALBERLA_BUILD_BENCHMARKS=ON -DWALBERLA_BUILD_TUTORIALS=ON -DWALBERLA_BUILD_TOOLS=ON -DWALBERLA_BUILD_WITH_MPI=ON -DWALBERLA_BUILD_WITH_OPENMP=ON -DCMAKE_BUILD_TYPE=Debug -DWALBERLA_BUILD_WITH_METIS=ON -DWALBERLA_BUILD_WITH_PARMETIS=ON -DWALBERLA_BUILD_WITH_OPENMESH=ON -DWALBERLA_DOUBLE_ACCURACY=ON -DWALBERLA_LOGLEVEL=DETAIL
+# - cmake . -LA
+# - utilities/filterCompileCommands.py compile_commands.json
+# - wget https://raw.githubusercontent.com/llvm/llvm-project/main/clang-tools-extra/clang-tidy/tool/run-clang-tidy.py
+# - python3 run-clang-tidy.py -quiet | tee clang-tidy-output.txt
+# artifacts:
+# paths:
+# - $CI_PROJECT_DIR/build/clang-tidy-output.txt
+# tags:
+# - docker
cppcheck:
@@ -2012,7 +1593,7 @@ coverage:
- mkdir build
- cd build
- if dpkg --compare-versions `ompi_info | head -2 | tail -1 | sed 's/[^0-9.]*\([0-9.]*\).*/\1/'` ge 1.10; then export MPIEXEC_PREFLAGS="--allow-run-as-root" ; fi
- - cmake .. -DWALBERLA_BUILD_TESTS=ON -DWALBERLA_BUILD_BENCHMARKS=ON -DWALBERLA_BUILD_TUTORIALS=ON -DWALBERLA_BUILD_WITH_MPI=ON -DWALBERLA_BUILD_WITH_OPENMP=OFF -DCMAKE_BUILD_TYPE=DebugOptimized -DMPIEXEC_PREFLAGS=$MPIEXEC_PREFLAGS -DWALBERLA_BUILD_WITH_CODEGEN=OFF -DWALBERLA_BUILD_WITH_GCOV=ON -DWALBERLA_LOGLEVEL=DETAIL
+ - cmake .. -DWALBERLA_BUILD_TESTS=ON -DWALBERLA_BUILD_BENCHMARKS=ON -DWALBERLA_BUILD_TUTORIALS=ON -DWALBERLA_BUILD_WITH_MPI=ON -DWALBERLA_BUILD_WITH_OPENMP=OFF -DCMAKE_BUILD_TYPE=DebugOptimized -DMPIEXEC_PREFLAGS=$MPIEXEC_PREFLAGS -DWALBERLA_BUILD_WITH_CODEGEN=OFF -DWALBERLA_BUILD_WITH_GCOV=ON -DWALBERLA_LOGLEVEL=DETAIL
- cmake . -LA
- make -j $NUM_BUILD_CORES -l $NUM_CORES
- ctest -LE longrun --output-on-failure -j $NUM_CORES --timeout 3000
@@ -2072,6 +1653,9 @@ coverage:
mac_Serial_Dbg:
extends: .mac_build_template
+ before_script:
+ - pip3 install pystencils==1.3.2
+ - pip3 install lbmpy==1.3.2
variables:
CMAKE_BUILD_TYPE: "DebugOptimized"
CTEST_EXCLUDE_LABELS: "longrun"
@@ -2082,6 +1666,9 @@ mac_Serial_Dbg:
mac_Serial:
extends: .mac_build_template
+ before_script:
+ - pip3 install pystencils==1.3.2
+ - pip3 install lbmpy==1.3.2
variables:
CMAKE_BUILD_TYPE: "Release"
CTEST_EXCLUDE_LABELS: "longrun"
@@ -2092,6 +1679,9 @@ mac_Serial:
mac_MpiOnly_Dbg:
extends: .mac_build_template
+ before_script:
+ - pip3 install pystencils==1.3.2
+ - pip3 install lbmpy==1.3.2
variables:
CMAKE_BUILD_TYPE: "DebugOptimized"
CTEST_EXCLUDE_LABELS: "longrun"
@@ -2103,6 +1693,9 @@ mac_MpiOnly_Dbg:
mac_MpiOnly:
extends: .mac_build_template
+ before_script:
+ - pip3 install pystencils==1.3.2
+ - pip3 install lbmpy==1.3.2
variables:
CMAKE_BUILD_TYPE: "Release"
CTEST_EXCLUDE_LABELS: "longrun"
diff --git a/CMakeLists.txt b/CMakeLists.txt
index f23b17c27810f9ac3ce97e7798d77ec3133ceea7..2eaeab744481977c0554afe95ca6b03d48606306 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -403,6 +403,8 @@ if( WALBERLA_CXX_COMPILER_IS_INTEL )
# system headers are also supported by intel, but cmake does not recognize that
set( CMAKE_INCLUDE_SYSTEM_FLAG_CXX "-isystem " )
add_flag ( CMAKE_CXX_FLAGS "-wd2928,2504,2259,1682,597" )
+ # disable icc/icpc deprecation warning
+ add_flag ( CMAKE_CXX_FLAGS "-diag-disable=10441" )
elseif( WALBERLA_CXX_COMPILER_IS_GNU )
add_flag ( CMAKE_CXX_FLAGS "-Wfloat-equal -Wextra" )
elseif( WALBERLA_CXX_COMPILER_IS_NEC )
@@ -615,7 +617,7 @@ endif ()
##
#############################################################################################################################
if ( WALBERLA_BUILD_WITH_CODEGEN )
- set(LBMPY_MIN_VERSION 1.2)
+ set(LBMPY_MIN_VERSION 1.3.2)
execute_process(COMMAND ${Python_EXECUTABLE} -c "import lbmpy; print(lbmpy.__version__)"
RESULT_VARIABLE LBMPY_FOUND OUTPUT_VARIABLE LBMPY_VERSION)
if(NOT LBMPY_FOUND EQUAL 0)
diff --git a/extern/pybind11 b/extern/pybind11
index 8b03ffa7c06cd9c8a38297b1c8923695d1ff1b07..f7b499615e14d70ab098a20deb0cdb3889998a1a 160000
--- a/extern/pybind11
+++ b/extern/pybind11
@@ -1 +1 @@
-Subproject commit 8b03ffa7c06cd9c8a38297b1c8923695d1ff1b07
+Subproject commit f7b499615e14d70ab098a20deb0cdb3889998a1a
diff --git a/python/lbmpy_walberla/additional_data_handler.py b/python/lbmpy_walberla/additional_data_handler.py
index 692d0cf57d3e905537d3cbd1af445c26c64074bc..16e5cb35f8ed64ae77eeda06c08812bbb57e950a 100644
--- a/python/lbmpy_walberla/additional_data_handler.py
+++ b/python/lbmpy_walberla/additional_data_handler.py
@@ -1,7 +1,8 @@
from pystencils import Target
from pystencils.stencil import inverse_direction
+from pystencils.typing import BasicType
-from lbmpy.advanced_streaming import AccessPdfValues, numeric_offsets, numeric_index
+from lbmpy.advanced_streaming import AccessPdfValues, numeric_offsets, numeric_index, Timestep, is_inplace
from lbmpy.advanced_streaming.indexing import MirroredStencilDirections
from lbmpy.boundaries.boundaryconditions import LbBoundary
from lbmpy.boundaries import ExtrapolationOutflow, FreeSlip, UBB
@@ -26,25 +27,8 @@ def default_additional_data_handler(boundary_obj: LbBoundary, lb_method, field_n
class FreeSlipAdditionalDataHandler(AdditionalDataHandler):
def __init__(self, stencil, boundary_object):
assert isinstance(boundary_object, FreeSlip)
- self._boundary_object = boundary_object
super(FreeSlipAdditionalDataHandler, self).__init__(stencil=stencil)
- @property
- def constructor_arguments(self):
- return ""
-
- @property
- def initialiser_list(self):
- return ""
-
- @property
- def additional_arguments_for_fill_function(self):
- return ""
-
- @property
- def additional_parameters_for_fill_function(self):
- return ""
-
def data_initialisation(self, direction):
def array_pattern(dtype, name, content):
return f"const {str(dtype)} {name} [] = {{ {','.join(str(c) for c in content)} }};"
@@ -102,21 +86,20 @@ class FreeSlipAdditionalDataHandler(AdditionalDataHandler):
return "\n".join(init_list)
- @property
- def additional_member_variable(self):
- return ""
-
class UBBAdditionalDataHandler(AdditionalDataHandler):
def __init__(self, stencil, boundary_object):
assert isinstance(boundary_object, UBB)
- self._boundary_object = boundary_object
super(UBBAdditionalDataHandler, self).__init__(stencil=stencil)
+ @property
+ def constructor_argument_name(self):
+ return "velocityCallback"
+
@property
def constructor_arguments(self):
- return ", std::function<Vector3<real_t>(const Cell &, const shared_ptr<StructuredBlockForest>&, IBlock&)>& " \
- "velocityCallback "
+ return f", std::function<Vector3<real_t>(const Cell &, const shared_ptr<StructuredBlockForest>&, IBlock&)>& " \
+ f"{self.constructor_argument_name} "
@property
def initialiser_list(self):
@@ -146,19 +129,34 @@ class UBBAdditionalDataHandler(AdditionalDataHandler):
class OutflowAdditionalDataHandler(AdditionalDataHandler):
- def __init__(self, stencil, boundary_object, target=Target.CPU, field_name='pdfs'):
+ def __init__(self, stencil, boundary_object, target=Target.CPU, field_name='pdfs', pdfs_data_type=None, zeroth_timestep=None):
assert isinstance(boundary_object, ExtrapolationOutflow)
- self._boundary_object = boundary_object
self._stencil = boundary_object.stencil
self._lb_method = boundary_object.lb_method
self._normal_direction = boundary_object.normal_direction
self._field_name = field_name
self._target = target
+ self._dtype = BasicType(boundary_object.data_type).c_name
+ if pdfs_data_type is None:
+ self._pdfs_data_type = "real_t"
+ else:
+ pdfs_data_type = BasicType(pdfs_data_type)
+ self._pdfs_data_type = pdfs_data_type.c_name
+
+ self._streaming_pattern = boundary_object.streaming_pattern
+ if zeroth_timestep:
+ self._zeroth_timestep = zeroth_timestep
+ else:
+ self._zeroth_timestep = Timestep.EVEN if is_inplace(self._streaming_pattern) else Timestep.BOTH
super(OutflowAdditionalDataHandler, self).__init__(stencil=stencil)
assert sum([a != 0 for a in self._normal_direction]) == 1, \
"The outflow boundary is only implemented for straight walls at the moment."
+ @property
+ def constructor_argument_name(self):
+ return f"{self._field_name}CPUID_" if self._target == Target.GPU else ""
+
@property
def constructor_arguments(self):
return f", BlockDataID {self._field_name}CPUID_" if self._target == Target.GPU else ""
@@ -170,18 +168,18 @@ class OutflowAdditionalDataHandler(AdditionalDataHandler):
@property
def additional_field_data(self):
identifier = "CPU" if self._target == Target.GPU else ""
- return f"auto {self._field_name} = block->getData< field::GhostLayerField<real_t, " \
+ return f"auto {self._field_name} = block->getData< field::GhostLayerField<{self._pdfs_data_type}, " \
f"{len(self._stencil)}> >({self._field_name}{identifier}ID); "
def data_initialisation(self, direction_index):
- pdf_acc = AccessPdfValues(self._boundary_object.stencil,
- streaming_pattern=self._boundary_object.streaming_pattern,
- timestep=self._boundary_object.zeroth_timestep,
+ pdf_acc = AccessPdfValues(self._stencil,
+ streaming_pattern=self._streaming_pattern,
+ timestep=self._zeroth_timestep,
streaming_dir='out')
init_list = []
for key, value in self.get_init_dict(pdf_acc, direction_index).items():
- init_list.append(f"element.{key} = {self._field_name}->get({value});")
+ init_list.append(f"element.{key} = {self._dtype}( {self._field_name}->get({value}) );")
return "\n".join(init_list)
diff --git a/python/lbmpy_walberla/boundary_collection.py b/python/lbmpy_walberla/boundary_collection.py
index 17bfa245a3212404c35dd06c420fcb19a55c3049..082567204acf47b97a1344de8b1ae38288ded567 100644
--- a/python/lbmpy_walberla/boundary_collection.py
+++ b/python/lbmpy_walberla/boundary_collection.py
@@ -15,8 +15,9 @@ from pystencils import Target
import numpy as np
-def lbm_boundary_generator(class_name: str, flag_uid: str, boundary_object: LbBoundary, additional_data_handler=None):
- def generation_function(ctx, lb_method, field_name='pdfs',
+def lbm_boundary_generator(class_name: str, flag_uid: str, boundary_object: LbBoundary, additional_data_handler=None,
+ field_data_type=None):
+ def generation_function(ctx, lb_method, field_name='pdfs', spatial_shape=None,
streaming_pattern='pull', after_collision=True,
namespace='lbm',
**create_kernel_params):
@@ -25,6 +26,8 @@ def lbm_boundary_generator(class_name: str, flag_uid: str, boundary_object: LbBo
boundary_object=boundary_object,
lb_method=lb_method,
field_name=field_name,
+ spatial_shape=spatial_shape,
+ field_data_type=field_data_type,
streaming_pattern=streaming_pattern,
after_collision=after_collision,
additional_data_handler=additional_data_handler,
@@ -41,6 +44,7 @@ def generate_boundary_collection(generation_context,
boundary_generators,
lb_method,
field_name='pdfs',
+ spatial_shape=None,
streaming_pattern='pull',
prev_timestep=Timestep.BOTH,
namespace='lbm',
@@ -49,22 +53,26 @@ def generate_boundary_collection(generation_context,
kernel_list = []
includes = []
boundary_classes = []
+ additional_data_handlers = []
flag_uids = []
object_names = []
targets = []
for boundary_generator in boundary_generators:
boundary_functor = boundary_generator['generator']
- context = boundary_functor(generation_context, lb_method, field_name, streaming_pattern, prev_timestep,
- namespace, **create_kernel_params)
+ context = boundary_functor(generation_context, lb_method, field_name, spatial_shape,
+ streaming_pattern, prev_timestep, namespace, **create_kernel_params)
kernel_list.append(context['kernel'])
includes.append(f"\"{context['class_name']}.h\"")
boundary_classes.append(f"{context['namespace']}::{context['class_name']}")
+ additional_data_handlers.append(context['additional_data_handler'])
flag_uids.append(boundary_generator['flag_id'])
object_names.append(f"{context['class_name']}Object")
targets.append(f"{context['target']}")
+ additional_constructor_arguments = [a.constructor_arguments[2:] for a in additional_data_handlers]
+
assert len(set(targets)) == 1
target = targets[0]
@@ -75,6 +83,8 @@ def generate_boundary_collection(generation_context,
'namespace': namespace,
'includes': includes,
'boundary_classes': boundary_classes,
+ 'additional_data_handlers': additional_data_handlers,
+ 'additional_constructor_arguments': additional_constructor_arguments,
'flag_uids': flag_uids,
'object_names': object_names
}
@@ -94,6 +104,8 @@ def __generate_alternating_lbm_boundary(generation_context,
boundary_object,
lb_method,
field_name='pdfs',
+ spatial_shape=None,
+ field_data_type=None,
streaming_pattern='pull',
after_collision=True,
additional_data_handler=None,
@@ -132,13 +144,14 @@ def __generate_alternating_lbm_boundary(generation_context,
return OddIntegerCondition(timestep_param_name, kernel_even, kernel_odd, timestep_param_dtype)
timestep_advancement = {"field_name": field_name, "function": "getTimestep"}
-
context = pystencils_walberla.boundary.generate_boundary(generation_context,
- class_name,
- boundary_object,
+ class_name=class_name,
+ boundary_object=boundary_object,
field_name=field_name,
neighbor_stencil=lb_method.stencil,
index_shape=[lb_method.stencil.Q],
+ spatial_shape=spatial_shape,
+ field_data_type=field_data_type,
kernel_creation_function=boundary_creation_function,
namespace=namespace,
additional_data_handler=additional_data_handler,
diff --git a/python/lbmpy_walberla/packing_kernels.py b/python/lbmpy_walberla/packing_kernels.py
index 53e5d877e14f5584c1e2eb7738fe018752998485..8a8728031cd4eb56ba4366e25ef6fd85b1f2e1b5 100644
--- a/python/lbmpy_walberla/packing_kernels.py
+++ b/python/lbmpy_walberla/packing_kernels.py
@@ -83,7 +83,8 @@ def generate_packing_kernels(generation_context: CodeGenerationContext, class_na
class PackingKernelsCodegen:
- def __init__(self, stencil, streaming_pattern, class_name, config: CreateKernelConfig):
+ def __init__(self, stencil, streaming_pattern, class_name, config: CreateKernelConfig,
+ src_field=None, dst_field=None):
self.stencil = stencil
self.dim = stencil.D
self.values_per_cell = stencil.Q
@@ -94,10 +95,11 @@ class PackingKernelsCodegen:
self.config = config
self.data_type = config.data_type['pdfs'].numpy_dtype
- self.src_field, self.dst_field = fields(
- f'pdfs_src({self.values_per_cell}), pdfs_dst({self.values_per_cell}) :{self.data_type}[{self.dim}D]')
+ self.src_field = src_field if src_field else fields(f'pdfs_src({stencil.Q}) :{self.data_type}[{stencil.D}D]')
+ self.dst_field = dst_field if dst_field else fields(f'pdfs_dst({stencil.Q}) :{self.data_type}[{stencil.D}D]')
+
self.accessors = [get_accessor(streaming_pattern, t) for t in get_timesteps(streaming_pattern)]
- self.mask_field = fields(f'mask : uint32 [{self.dim}D]')
+ self.mask_field = fields(f'mask : uint32 [{self.dim}D]', layout=src_field.layout)
def create_uniform_kernel_families(self, kernels_dict=None):
kernels = dict() if kernels_dict is None else kernels_dict
diff --git a/python/lbmpy_walberla/storage_specification.py b/python/lbmpy_walberla/storage_specification.py
index d3eaec45797a40046d4ba273487c580274865878..dd74fe8d0202039fc03630004127c95d3a9f3bb2 100644
--- a/python/lbmpy_walberla/storage_specification.py
+++ b/python/lbmpy_walberla/storage_specification.py
@@ -4,10 +4,10 @@ from dataclasses import replace
from jinja2 import Environment, PackageLoader, StrictUndefined
import numpy as np
-from pystencils import Target
+from pystencils import fields, Target
-from lbmpy import LBMConfig
-from lbmpy.advanced_streaming import is_inplace
+from lbmpy import LBMConfig, LBMOptimisation
+from lbmpy.advanced_streaming import is_inplace, get_accessor, Timestep
from lbmpy.methods import AbstractLbMethod
from pystencils_walberla.cmake_integration import CodeGenerationContext
@@ -15,11 +15,15 @@ from pystencils_walberla.jinja_filters import add_pystencils_filters_to_jinja_en
from pystencils_walberla.utility import config_from_context
from pystencils_walberla.instrumentation import likwid_wrap_family
from lbmpy_walberla.packing_kernels import PackingKernelsCodegen
+from lbmpy_walberla.utility import create_pdf_field
def generate_lbm_storage_specification(generation_context: CodeGenerationContext, class_name: str,
- method: AbstractLbMethod, lbm_config: LBMConfig, nonuniform: bool = False,
- target: Target = Target.CPU, data_type=None, cpu_openmp: bool = False,
+ method: AbstractLbMethod,
+ lbm_config: LBMConfig, lbm_optimisation: LBMOptimisation,
+ nonuniform: bool = False,
+ target: Target = Target.CPU,
+ data_type=None, cpu_openmp: bool = False,
**create_kernel_params):
namespace = "lbm"
stencil = method.stencil
@@ -33,10 +37,32 @@ def generate_lbm_storage_specification(generation_context: CodeGenerationContext
config = replace(config, cpu_vectorize_info=None)
default_dtype = config.data_type.default_factory()
- is_float = True if issubclass(default_dtype.numpy_dtype.type, np.float32) else False
- constant_suffix = "f" if is_float else ""
-
- cg = PackingKernelsCodegen(stencil, streaming_pattern, class_name, config)
+ if issubclass(default_dtype.numpy_dtype.type, np.float64):
+ data_type_string = "double"
+ elif issubclass(default_dtype.numpy_dtype.type, np.float32):
+ data_type_string = "float"
+ elif issubclass(default_dtype.numpy_dtype.type, np.float16):
+ data_type_string = "half"
+ else:
+ raise ValueError(f"default datatype {default_dtype.numpy_dtype.type} is not supported. "
+ f"Supported are only np.float64, np.float32 and np.float16")
+
+ symbolic_field = lbm_optimisation.symbolic_field
+ if not symbolic_field:
+ symbolic_field = create_pdf_field(config=config, name="pdfs_src", stencil=stencil,
+ field_layout=lbm_optimisation.field_layout)
+
+ if is_inplace(streaming_pattern):
+ symbolic_temporary_field = create_pdf_field(config=config, name="pdfs_dst", stencil=stencil,
+ field_layout=lbm_optimisation.field_layout)
+ else:
+ symbolic_temporary_field = lbm_optimisation.symbolic_temporary_field
+ if not symbolic_temporary_field:
+ symbolic_temporary_field = create_pdf_field(config=config, name="pdfs_dst", stencil=stencil,
+ field_layout=lbm_optimisation.field_layout)
+
+ cg = PackingKernelsCodegen(stencil, streaming_pattern, class_name, config,
+ src_field=symbolic_field, dst_field=symbolic_temporary_field)
kernels = cg.create_uniform_kernel_families()
if nonuniform:
@@ -55,6 +81,16 @@ def generate_lbm_storage_specification(generation_context: CodeGenerationContext
cqc = method.conserved_quantity_computation
equilibrium = method.equilibrium_distribution
+ f = fields(f"f({stencil.Q}): double[{stencil.D}D]", layout='fzyx')
+ even_accessor = get_accessor(streaming_pattern, Timestep.EVEN)
+ odd_accessor = get_accessor(streaming_pattern, Timestep.ODD)
+
+ even_read = even_accessor.read(f, stencil)
+ even_write = even_accessor.write(f, stencil)
+
+ odd_read = odd_accessor.read(f, stencil)
+ odd_write = odd_accessor.write(f, stencil)
+
jinja_context = {
'class_name': class_name,
'namespace': namespace,
@@ -67,12 +103,17 @@ def generate_lbm_storage_specification(generation_context: CodeGenerationContext
'equilibrium_deviation_only': equilibrium.deviation_only,
'inplace': is_inplace(streaming_pattern),
'zero_centered': cqc.zero_centered_pdfs,
- 'weights': ",".join(str(w.evalf()) + constant_suffix for w in method.weights),
- 'inverse_weights': ",".join(str((1 / w).evalf()) + constant_suffix for w in method.weights),
+
+ 'weights': ", ".join(f"{data_type_string}({str(w.evalf())})" for w in method.weights),
+ 'inverse_weights': ", ".join(f"{data_type_string}({str((1 / w).evalf())})" for w in method.weights),
+ 'even_read': _get_access_list(even_read, stencil.D),
+ 'even_write': _get_access_list(even_write, stencil.D),
+ 'odd_read': _get_access_list(odd_read, stencil.D),
+ 'odd_write': _get_access_list(odd_write, stencil.D),
'nonuniform': nonuniform,
'target': target.name.lower(),
- 'dtype': "float" if is_float else "double",
+ 'dtype': data_type_string,
'is_gpu': target == Target.GPU,
'likwid': generation_context.likwid,
'kernels': kernels,
@@ -93,3 +134,16 @@ def generate_lbm_storage_specification(generation_context: CodeGenerationContext
source_extension = "cu" if target == Target.GPU and generation_context.cuda else "cpp"
generation_context.write_file(f"{class_name}.h", header)
generation_context.write_file(f"{class_name}.{source_extension}", source)
+
+
+def _get_access_list(access_list, dim):
+ result = []
+ for i in range(dim):
+ result.append(", ".join([str(int(field_access.offsets[i])) for field_access in access_list]))
+
+ if dim == 2:
+ result.append(", ".join(["0"] * len(access_list)))
+
+ result.append(", ".join([str(int(field_access.index[0])) for field_access in access_list]))
+
+ return result
diff --git a/python/lbmpy_walberla/sweep_collection.py b/python/lbmpy_walberla/sweep_collection.py
index 8edd0779b328de768cba4a3acb5f04bdb6bb3acf..5fe4892ab3bc9740f3ca04775363586b42134e96 100644
--- a/python/lbmpy_walberla/sweep_collection.py
+++ b/python/lbmpy_walberla/sweep_collection.py
@@ -7,9 +7,10 @@ import numpy as np
from pystencils import Target, create_kernel
from pystencils.config import CreateKernelConfig
from pystencils.field import Field
+from pystencils.simp import add_subexpressions_for_field_reads
from lbmpy.advanced_streaming import is_inplace, get_accessor, Timestep
-from lbmpy.creationfunctions import LbmCollisionRule
+from lbmpy.creationfunctions import LbmCollisionRule, LBMConfig, LBMOptimisation
from lbmpy.fieldaccess import CollideOnlyInplaceAccessor
from lbmpy.macroscopic_value_kernels import macroscopic_values_setter, macroscopic_values_getter
from lbmpy.updatekernels import create_lbm_kernel, create_stream_only_kernel
@@ -17,41 +18,45 @@ from lbmpy.updatekernels import create_lbm_kernel, create_stream_only_kernel
from pystencils_walberla.kernel_selection import KernelCallNode, KernelFamily
from pystencils_walberla.utility import config_from_context
from pystencils_walberla import generate_sweep_collection
+from lbmpy_walberla.utility import create_pdf_field
from .alternating_sweeps import EvenIntegerCondition
from .function_generator import kernel_family_function_generator
def generate_lbm_sweep_collection(ctx, class_name: str, collision_rule: LbmCollisionRule,
- streaming_pattern='pull',
- field_layout='fzyx', refinement_scaling=None,
- macroscopic_fields: Dict[str, Field] = None,
+ lbm_config: LBMConfig, lbm_optimisation: LBMOptimisation,
+ refinement_scaling=None, macroscopic_fields: Dict[str, Field] = None,
target=Target.CPU, data_type=None, cpu_openmp=None, cpu_vectorize_info=None,
max_threads=None,
**create_kernel_params):
+
config = config_from_context(ctx, target=target, data_type=data_type,
cpu_openmp=cpu_openmp, cpu_vectorize_info=cpu_vectorize_info, **create_kernel_params)
+ streaming_pattern = lbm_config.streaming_pattern
+ field_layout = lbm_optimisation.field_layout
+
# usually a numpy layout is chosen by default i.e. xyzf - which is bad for waLBerla where at least the spatial
# coordinates should be ordered in reverse direction i.e. zyx
lb_method = collision_rule.method
- q = lb_method.stencil.Q
- dim = lb_method.stencil.D
-
if field_layout == 'fzyx':
config.cpu_vectorize_info['assume_inner_stride_one'] = True
elif field_layout == 'zyxf':
config.cpu_vectorize_info['assume_inner_stride_one'] = False
- src_field = Field.create_generic('pdfs', dim, config.data_type['pdfs'].numpy_dtype,
- index_dimensions=1, layout=field_layout, index_shape=(q,))
+ src_field = lbm_optimisation.symbolic_field
+ if not src_field:
+ src_field = create_pdf_field(config=config, name="pdfs", stencil=lbm_config.stencil,
+ field_layout=lbm_optimisation.field_layout)
if is_inplace(streaming_pattern):
dst_field = src_field
else:
- dst_field = Field.create_generic('pdfs_tmp', dim, config.data_type['pdfs_tmp'].numpy_dtype,
- index_dimensions=1, layout=field_layout,
- index_shape=(q,))
+ dst_field = lbm_optimisation.symbolic_temporary_field
+ if not dst_field:
+ dst_field = create_pdf_field(config=config, name="pdfs_tmp", stencil=lbm_config.stencil,
+ field_layout=lbm_optimisation.field_layout)
config = replace(config, ghost_layers=0)
@@ -68,12 +73,16 @@ def generate_lbm_sweep_collection(ctx, class_name: str, collision_rule: LbmColli
function_generators.append(generator('stream', family("stream")))
function_generators.append(generator('streamOnlyNoAdvancement', family("streamOnlyNoAdvancement")))
- setter_family = get_setter_family(class_name, lb_method, src_field, streaming_pattern, macroscopic_fields, config)
+ config_unoptimized = replace(config, cpu_vectorize_info=None, cpu_prepend_optimizations=[], cpu_blocking=None)
+
+ setter_family = get_setter_family(class_name, lb_method, src_field, streaming_pattern, macroscopic_fields,
+ config_unoptimized)
setter_generator = kernel_family_function_generator('initialise', setter_family,
namespace='lbm', max_threads=max_threads)
function_generators.append(setter_generator)
- getter_family = get_getter_family(class_name, lb_method, src_field, streaming_pattern, macroscopic_fields, config)
+ getter_family = get_getter_family(class_name, lb_method, src_field, streaming_pattern, macroscopic_fields,
+ config_unoptimized)
getter_generator = kernel_family_function_generator('calculateMacroscopicParameters', getter_family,
namespace='lbm', max_threads=max_threads)
function_generators.append(getter_generator)
@@ -98,15 +107,17 @@ class RefinementScaling:
def lbm_kernel_family(class_name, kernel_name,
collision_rule, streaming_pattern, src_field, dst_field, config: CreateKernelConfig):
+ default_dtype = config.data_type.default_factory()
if kernel_name == "streamCollide":
def lbm_kernel(field_accessor, lb_stencil):
- return create_lbm_kernel(collision_rule, src_field, dst_field, field_accessor)
+ return create_lbm_kernel(collision_rule, src_field, dst_field, field_accessor, data_type=default_dtype)
advance_timestep = {"field_name": src_field.name, "function": "advanceTimestep"}
temporary_fields = ['pdfs_tmp']
field_swaps = [('pdfs', 'pdfs_tmp')]
elif kernel_name == "collide":
def lbm_kernel(field_accessor, lb_stencil):
- return create_lbm_kernel(collision_rule, src_field, dst_field, CollideOnlyInplaceAccessor())
+ return create_lbm_kernel(collision_rule, src_field, dst_field, CollideOnlyInplaceAccessor(),
+ data_type=default_dtype)
advance_timestep = {"field_name": src_field.name, "function": "advanceTimestep"}
temporary_fields = ()
field_swaps = ()
@@ -161,6 +172,8 @@ def get_setter_family(class_name, lb_method, pdfs, streaming_pattern, macroscopi
density = macroscopic_fields.get('density', 1.0)
velocity = macroscopic_fields.get('velocity', [0.0] * dim)
+ default_dtype = config.data_type.default_factory()
+
get_timestep = {"field_name": pdfs.name, "function": "getTimestep"}
temporary_fields = ()
field_swaps = ()
@@ -173,6 +186,9 @@ def get_setter_family(class_name, lb_method, pdfs, streaming_pattern, macroscopi
density=density, velocity=velocity, pdfs=pdfs,
streaming_pattern=streaming_pattern, previous_timestep=timestep)
+ if default_dtype != pdfs.dtype:
+ setter = add_subexpressions_for_field_reads(setter, data_type=default_dtype)
+
setter_ast = create_kernel(setter, config=config)
setter_ast.function_name = 'kernel_initialise' + timestep_suffix
nodes.append(KernelCallNode(setter_ast))
@@ -199,6 +215,8 @@ def get_getter_family(class_name, lb_method, pdfs, streaming_pattern, macroscopi
if density is None and velocity is None:
return None
+ default_dtype = config.data_type.default_factory()
+
get_timestep = {"field_name": pdfs.name, "function": "getTimestep"}
temporary_fields = ()
field_swaps = ()
@@ -211,6 +229,9 @@ def get_getter_family(class_name, lb_method, pdfs, streaming_pattern, macroscopi
density=density, velocity=velocity, pdfs=pdfs,
streaming_pattern=streaming_pattern, previous_timestep=timestep)
+ if default_dtype != pdfs.dtype:
+ getter = add_subexpressions_for_field_reads(getter, data_type=default_dtype)
+
getter_ast = create_kernel(getter, config=config)
getter_ast.function_name = 'kernel_getter' + timestep_suffix
nodes.append(KernelCallNode(getter_ast))
diff --git a/python/lbmpy_walberla/templates/BoundaryCollection.tmpl.h b/python/lbmpy_walberla/templates/BoundaryCollection.tmpl.h
index 5f49137846ba99d60888e7353ac4ff195ade2a84..47f313860dacc6cf69c783047c3ed53e82bcb91b 100644
--- a/python/lbmpy_walberla/templates/BoundaryCollection.tmpl.h
+++ b/python/lbmpy_walberla/templates/BoundaryCollection.tmpl.h
@@ -41,12 +41,12 @@ class {{class_name}}
enum Type { ALL = 0, INNER = 1, OUTER = 2 };
- {{class_name}}( {{- ["const shared_ptr<StructuredBlockForest> & blocks", "BlockDataID flagID_", "BlockDataID pdfsID_", "FlagUID domainUID_", [kernel_list|generate_constructor_parameters(['indexVector', 'indexVectorSize', 'pdfs'])]] | type_identifier_list -}} )
+ {{class_name}}( {{- ["const shared_ptr<StructuredBlockForest> & blocks", "BlockDataID flagID_", "BlockDataID pdfsID_", "FlagUID domainUID_", [kernel_list|generate_constructor_parameters(['indexVector', 'indexVectorSize', 'pdfs'])], additional_constructor_arguments] | type_identifier_list -}} )
: blocks_(blocks), flagID(flagID_), pdfsID(pdfsID_), domainUID(domainUID_)
{
- {% for object_name, boundary_class, kernel in zip(object_names, boundary_classes, kernel_list) -%}
+ {% for object_name, boundary_class, kernel, additional_data_handler in zip(object_names, boundary_classes, kernel_list, additional_data_handlers) -%}
- {{object_name}} = std::make_shared< {{boundary_class}} >({{- ["blocks", "pdfsID", [kernel|generate_function_collection_call(['indexVector', 'indexVectorSize', 'pdfs', 'timestep', 'gpuStream'])]] | type_identifier_list -}});
+ {{object_name}} = std::make_shared< {{boundary_class}} >({{- ["blocks", "pdfsID", [kernel|generate_function_collection_call(['indexVector', 'indexVectorSize', 'pdfs', 'timestep', 'gpuStream'])], additional_data_handler.constructor_argument_name] | type_identifier_list -}});
{% endfor %}
{% for object_name, flag_uid in zip(object_names, flag_uids) -%}
@@ -105,4 +105,3 @@ class {{class_name}}
}
}
-
diff --git a/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.cpp b/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.cpp
index 91c7d7d960a78552628d3d8568dd611f13c14a2d..92106cddf0248c93707e6881b600b6a702d27985 100644
--- a/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.cpp
+++ b/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.cpp
@@ -27,31 +27,31 @@
# pragma GCC diagnostic ignored "-Wunused-variable"
#endif
-/*************************************************************************************
+namespace walberla {
+namespace {{namespace}} {
+
+ /*************************************************************************************
* Kernel Definitions
*************************************************************************************/
-{{ kernels['packAll'] | generate_definitions }}
-{{ kernels['unpackAll'] | generate_definitions }}
-{{ kernels['localCopyAll'] | generate_definitions }}
-
-{{ kernels['packDirection'] | generate_definitions }}
-{{ kernels['unpackDirection'] | generate_definitions }}
-{{ kernels['localCopyDirection'] | generate_definitions }}
-
-{% if nonuniform -%}
-{{ kernels['unpackRedistribute'] | generate_definitions }}
-{{ kernels['packPartialCoalescence'] | generate_definitions }}
-{{ kernels['zeroCoalescenceRegion'] | generate_definitions }}
-{{ kernels['unpackCoalescence'] | generate_definitions }}
-{%- endif %}
-
-/*************************************************************************************
+ {{ kernels['packAll'] | generate_definitions }}
+ {{ kernels['unpackAll'] | generate_definitions }}
+ {{ kernels['localCopyAll'] | generate_definitions }}
+
+ {{ kernels['packDirection'] | generate_definitions }}
+ {{ kernels['unpackDirection'] | generate_definitions }}
+ {{ kernels['localCopyDirection'] | generate_definitions }}
+
+ {% if nonuniform -%}
+ {{ kernels['unpackRedistribute'] | generate_definitions }}
+ {{ kernels['packPartialCoalescence'] | generate_definitions }}
+ {{ kernels['zeroCoalescenceRegion'] | generate_definitions }}
+ {{ kernels['unpackCoalescence'] | generate_definitions }}
+ {%- endif %}
+
+ /*************************************************************************************
* Kernel Wrappers
*************************************************************************************/
-namespace walberla {
-namespace {{namespace}} {
-
void {{class_name}}::PackKernels::packAll(
{{- [ "PdfField_T * " + src_field.name, "CellInterval & ci",
"unsigned char * outBuffer", kernels['packAll'].kernel_selection_parameters,
@@ -128,8 +128,8 @@ namespace {{namespace}} {
WALBERLA_ASSERT_EQUAL(srcInterval.zSize(), dstInterval.zSize())
{{kernels['localCopyDirection']
- | generate_call(cell_interval={src_field : 'srcInterval', dst_field : 'dstInterval'}, stream='stream')
- | indent(6) }}
+ | generate_call(cell_interval={src_field : 'srcInterval', dst_field : 'dstInterval'}, stream='stream')
+ | indent(6) }}
}
{% if nonuniform -%}
diff --git a/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.h b/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.h
index d6039b077c4ab95e226f5c68ecc34a0a138b6932..c8f7670aded37ad2472a0385591d24f75f5ee374 100644
--- a/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.h
+++ b/python/lbmpy_walberla/templates/LbmStorageSpecification.tmpl.h
@@ -82,6 +82,32 @@ class {{class_name}}
// Inverse lattice weights
static constexpr {{dtype}} wInv[{{stencil_size}}] = { {{inverse_weights}} };
+ struct AccessorEVEN
+ {
+ static constexpr cell_idx_t readX[{{stencil_size}}] = { {{even_read[0]}} };
+ static constexpr cell_idx_t readY[{{stencil_size}}] = { {{even_read[1]}} };
+ static constexpr cell_idx_t readZ[{{stencil_size}}] = { {{even_read[2]}} };
+ static constexpr cell_idx_t readD[{{stencil_size}}] = { {{even_read[3]}} };
+
+ static constexpr cell_idx_t writeX[{{stencil_size}}] = { {{even_write[0]}} };
+ static constexpr cell_idx_t writeY[{{stencil_size}}] = { {{even_write[1]}} };
+ static constexpr cell_idx_t writeZ[{{stencil_size}}] = { {{even_write[2]}} };
+ static constexpr cell_idx_t writeD[{{stencil_size}}] = { {{even_write[3]}} };
+ };
+
+ struct AccessorODD
+ {
+ static constexpr cell_idx_t readX[{{stencil_size}}] = { {{odd_read[0]}} };
+ static constexpr cell_idx_t readY[{{stencil_size}}] = { {{odd_read[1]}} };
+ static constexpr cell_idx_t readZ[{{stencil_size}}] = { {{odd_read[2]}} };
+ static constexpr cell_idx_t readD[{{stencil_size}}] = { {{odd_read[3]}} };
+
+ static constexpr cell_idx_t writeX[{{stencil_size}}] = { {{odd_write[0]}} };
+ static constexpr cell_idx_t writeY[{{stencil_size}}] = { {{odd_write[1]}} };
+ static constexpr cell_idx_t writeZ[{{stencil_size}}] = { {{odd_write[2]}} };
+ static constexpr cell_idx_t writeD[{{stencil_size}}] = { {{odd_write[3]}} };
+ };
+
// Compute kernels to pack and unpack MPI buffers
class PackKernels {
@@ -100,8 +126,8 @@ class {{class_name}}
static const bool inplace = {% if inplace -%} true {%- else -%} false {%- endif -%};
/**
- * Packs all pdfs from the given cell interval to the send buffer.
- * */
+ * Packs all pdfs from the given cell interval to the send buffer.
+ * */
void packAll(
{{- [ "PdfField_T * " + src_field.name, "CellInterval & ci",
"unsigned char * outBuffer", kernels['packAll'].kernel_selection_parameters,
@@ -172,7 +198,7 @@ class {{class_name}}
* @return The required size of the buffer, in bytes
* */
uint_t size (CellInterval & ci, stencil::Direction dir) const {
- return ci.numCells() * sizes[dir] * sizeof(value_type);
+ return ci.numCells() * sizes[dir] * uint_c(sizeof(value_type));
}
/**
@@ -182,7 +208,7 @@ class {{class_name}}
* @return The required size of the buffer, in bytes
* */
uint_t size (CellInterval & ci) const {
- return ci.numCells() * {{stencil_size}} * sizeof(value_type);
+ return ci.numCells() * {{stencil_size}} * uint_c(sizeof(value_type));
}
{% if nonuniform -%}
@@ -255,6 +281,8 @@ class {{class_name}}
const uint_t sizes[{{direction_sizes|length}}] { {{ direction_sizes | join(', ') }} };
};
+ using value_type = PackKernels::value_type;
+
};
}} //{{namespace}}/walberla
\ No newline at end of file
diff --git a/python/lbmpy_walberla/utility.py b/python/lbmpy_walberla/utility.py
index 1289c381e7b50ac7e83d34fca887e6d659959b92..75460d5f9458104e02f21ba52d54acbb261628b7 100644
--- a/python/lbmpy_walberla/utility.py
+++ b/python/lbmpy_walberla/utility.py
@@ -1,4 +1,8 @@
+import numpy as np
+from pystencils import CreateKernelConfig, fields
+
from lbmpy.advanced_streaming import Timestep
+from lbmpy.stencils import LBStencil
def timestep_suffix(timestep: Timestep):
@@ -9,3 +13,9 @@ def timestep_suffix(timestep: Timestep):
"""
return ("_" + str(timestep)) if timestep != Timestep.BOTH else ''
+
+def create_pdf_field(config: CreateKernelConfig, name: str, stencil: LBStencil, field_layout: str = 'fzyx'):
+ default_dtype = config.data_type.default_factory()
+ data_type = default_dtype.numpy_dtype
+ return fields(f'{name}({stencil.Q}) :{data_type}[{stencil.D}D]', layout=field_layout)
+
diff --git a/python/lbmpy_walberla/walberla_lbm_package.py b/python/lbmpy_walberla/walberla_lbm_package.py
index e21d6c9613a1c2be87e21cbc06a2a78212f72552..80b37a4a90f717e79ea60890a802a363c45dde52 100644
--- a/python/lbmpy_walberla/walberla_lbm_package.py
+++ b/python/lbmpy_walberla/walberla_lbm_package.py
@@ -17,7 +17,9 @@ def generate_lbm_package(ctx: CodeGenerationContext, name: str,
lbm_config: LBMConfig, lbm_optimisation: LBMOptimisation,
nonuniform: bool = False, boundaries: List[Callable] = None,
macroscopic_fields: Dict[str, Field] = None,
- target: Target = Target.CPU, data_type=None, cpu_openmp=None, cpu_vectorize_info=None,
+ target: Target = Target.CPU,
+ data_type=None, pdfs_data_type=None,
+ cpu_openmp=None, cpu_vectorize_info=None,
max_threads=None,
**kernel_parameters):
@@ -27,8 +29,10 @@ def generate_lbm_package(ctx: CodeGenerationContext, name: str,
method = collision_rule.method
storage_spec_name = f'{name}StorageSpecification'
- generate_lbm_storage_specification(ctx, storage_spec_name, method, lbm_config,
- nonuniform=nonuniform, target=target, data_type=data_type)
+ generate_lbm_storage_specification(ctx, storage_spec_name, method, lbm_config, lbm_optimisation,
+ nonuniform=nonuniform, target=target,
+ data_type=pdfs_data_type,
+ cpu_openmp=cpu_openmp)
if nonuniform:
omega = get_shear_relaxation_rate(method)
@@ -37,10 +41,8 @@ def generate_lbm_package(ctx: CodeGenerationContext, name: str,
else:
refinement_scaling = None
- streaming_pattern = lbm_config.streaming_pattern
generate_lbm_sweep_collection(ctx, f'{name}SweepCollection', collision_rule,
- streaming_pattern=streaming_pattern,
- field_layout=lbm_optimisation.field_layout,
+ lbm_config=lbm_config, lbm_optimisation=lbm_optimisation,
refinement_scaling=refinement_scaling,
macroscopic_fields=macroscopic_fields,
target=target, data_type=data_type,
@@ -48,6 +50,11 @@ def generate_lbm_package(ctx: CodeGenerationContext, name: str,
max_threads=max_threads,
**kernel_parameters)
+ spatial_shape = None
+ if lbm_optimisation.symbolic_field and lbm_optimisation.symbolic_field.has_fixed_shape:
+ spatial_shape = lbm_optimisation.symbolic_field.spatial_shape + (lbm_config.stencil.Q, )
+
generate_boundary_collection(ctx, f'{name}BoundaryCollection', boundary_generators=boundaries,
- lb_method=method, streaming_pattern=streaming_pattern,
+ lb_method=method, field_name='pdfs', spatial_shape=spatial_shape,
+ streaming_pattern=lbm_config.streaming_pattern,
target=target, layout=lbm_optimisation.field_layout)
diff --git a/python/pystencils_walberla/additional_data_handler.py b/python/pystencils_walberla/additional_data_handler.py
index 2c4efcc65bdd60451e32e7cefa88f9baf8d6c03e..48f87dd5307a2da6c4ceb3bfcd59974ef9896691 100644
--- a/python/pystencils_walberla/additional_data_handler.py
+++ b/python/pystencils_walberla/additional_data_handler.py
@@ -16,6 +16,10 @@ class AdditionalDataHandler:
else:
self._walberla_stencil = stencil
+ @property
+ def constructor_argument_name(self):
+ return ""
+
@property
def constructor_arguments(self):
return ""
diff --git a/python/pystencils_walberla/boundary.py b/python/pystencils_walberla/boundary.py
index c5a5e54c1d00d9d6e476306453eae4320b6f5aa8..7af79ed677697633c099a5dec78b2b9afb66a226 100644
--- a/python/pystencils_walberla/boundary.py
+++ b/python/pystencils_walberla/boundary.py
@@ -19,7 +19,9 @@ def generate_boundary(generation_context,
field_name,
neighbor_stencil,
index_shape,
+ spatial_shape=None,
field_type=FieldType.GENERIC,
+ field_data_type=None,
kernel_creation_function=None,
target=Target.CPU,
data_type=None,
@@ -47,14 +49,17 @@ def generate_boundary(generation_context,
del create_kernel_params['default_number_int']
del create_kernel_params['skip_independence_check']
- field_data_type = config.data_type[field_name].numpy_dtype
+ if field_data_type is None:
+ field_data_type = config.data_type[field_name].numpy_dtype
index_struct_dtype = numpy_data_type_for_boundary_object(boundary_object, dim)
- field = Field.create_generic(field_name, dim,
- field_data_type,
- index_dimensions=len(index_shape), layout=layout, index_shape=index_shape,
- field_type=field_type)
+ if spatial_shape:
+ field = Field.create_fixed_size(field_name, spatial_shape, index_dimensions=len(index_shape),
+ dtype=field_data_type, layout=layout, field_type=field_type)
+ else:
+ field = Field.create_generic(field_name, dim, dtype=field_data_type, index_dimensions=len(index_shape),
+ layout=layout, index_shape=index_shape, field_type=field_type)
index_field = Field('indexVector', FieldType.INDEXED, index_struct_dtype, layout=[0],
shape=(TypedSymbol("indexVectorSize", create_type("int32")), 1), strides=(1, 1))
@@ -126,5 +131,3 @@ def generate_staggered_flux_boundary(generation_context, class_name, boundary_ob
assert dim == len(neighbor_stencil[0])
generate_boundary(generation_context, class_name, boundary_object, 'flux', neighbor_stencil, index_shape,
FieldType.STAGGERED_FLUX, target=target, **kwargs)
-
-
diff --git a/python/pystencils_walberla/templates/Boundary.tmpl.h b/python/pystencils_walberla/templates/Boundary.tmpl.h
index 96a9202c19345f0e36c5e048be1ee65969f5c966..75e3cd13abdacdcd06ea7ba784552b66daa21b63 100644
--- a/python/pystencils_walberla/templates/Boundary.tmpl.h
+++ b/python/pystencils_walberla/templates/Boundary.tmpl.h
@@ -48,6 +48,10 @@
#define RESTRICT
#endif
+#ifdef WALBERLA_BUILD_WITH_HALF_PRECISION_SUPPORT
+using walberla::half;
+#endif
+
namespace walberla {
namespace {{namespace}} {
diff --git a/python/pystencils_walberla/templates/SweepCollection.tmpl.h b/python/pystencils_walberla/templates/SweepCollection.tmpl.h
index 5db4ccb33457efcc2f9f9385d0f2b32db35aef5e..2a89fef5e4c6c825443db51ed1acc3a6d02ffaa8 100644
--- a/python/pystencils_walberla/templates/SweepCollection.tmpl.h
+++ b/python/pystencils_walberla/templates/SweepCollection.tmpl.h
@@ -64,20 +64,20 @@ namespace {{namespace}} {
class {{class_name}}
{
-public:
- enum Type { ALL = 0, INNER = 1, OUTER = 2 };
+ public:
+ enum Type { ALL = 0, INNER = 1, OUTER = 2 };
{{class_name}}(const shared_ptr< StructuredBlockStorage > & blocks, {{kernel_list|generate_constructor_parameters}}, const Cell & outerWidth=Cell(1, 1, 1))
- : blocks_(blocks), {{ kernel_list|generate_constructor_initializer_list(parameter_registration=parameter_scaling) }}, outerWidth_(outerWidth)
+ : blocks_(blocks), {{ kernel_list|generate_constructor_initializer_list(parameter_registration=parameter_scaling) }}, outerWidth_(outerWidth)
{
{{kernel_list|generate_constructor(parameter_registration=parameter_scaling) |indent(6)}}
+ validInnerOuterSplit_= true;
+
for (auto& iBlock : *blocks)
{
- if (int_c(blocks->getNumberOfXCells(iBlock)) <= outerWidth_[0] * 2 ||
- int_c(blocks->getNumberOfYCells(iBlock)) <= outerWidth_[1] * 2 ||
- int_c(blocks->getNumberOfZCells(iBlock)) <= outerWidth_[2] * 2)
- WALBERLA_ABORT_NO_DEBUG_INFO("innerOuterSplit too large - make it smaller or increase cellsPerBlock")
+ if (int_c(blocks->getNumberOfXCells(iBlock)) <= outerWidth_[0] * 2 || int_c(blocks->getNumberOfYCells(iBlock)) <= outerWidth_[1] * 2 || int_c(blocks->getNumberOfZCells(iBlock)) <= outerWidth_[2] * 2)
+ validInnerOuterSplit_ = false;
}
};
@@ -105,54 +105,66 @@ public:
std::function<void (IBlock *)> {{kernel['function_name']}}({{- ["Type type", ] | type_identifier_list -}})
{
+ if (!validInnerOuterSplit_ && type != Type::ALL)
+ WALBERLA_ABORT_NO_DEBUG_INFO("innerOuterSplit too large - make it smaller, increase cellsPerBlock or avoid communication hiding")
+
switch (type)
{
- case Type::INNER:
- return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", ] | type_identifier_list -}}); };
- case Type::OUTER:
- return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", ] | type_identifier_list -}}); };
- default:
- return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", ] | type_identifier_list -}}); };
+ case Type::INNER:
+ return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", ] | type_identifier_list -}}); };
+ case Type::OUTER:
+ return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", ] | type_identifier_list -}}); };
+ default:
+ return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", ] | type_identifier_list -}}); };
}
}
std::function<void (IBlock *)> {{kernel['function_name']}}({{- ["Type type", "const cell_idx_t ghost_layers"] | type_identifier_list -}})
{
+ if (!validInnerOuterSplit_ && type != Type::ALL)
+ WALBERLA_ABORT_NO_DEBUG_INFO("innerOuterSplit too large - make it smaller, increase cellsPerBlock or avoid communication hiding")
+
switch (type)
{
- case Type::INNER:
- return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", ] | type_identifier_list -}}); };
- case Type::OUTER:
- return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", ] | type_identifier_list -}}); };
- default:
- return [{{- ["this", "ghost_layers"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", "ghost_layers"] | type_identifier_list -}}); };
+ case Type::INNER:
+ return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", ] | type_identifier_list -}}); };
+ case Type::OUTER:
+ return [{{- ["this", ] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", ] | type_identifier_list -}}); };
+ default:
+ return [{{- ["this", "ghost_layers"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", "ghost_layers"] | type_identifier_list -}}); };
}
}
{% if target is equalto 'gpu' -%}
std::function<void (IBlock *)> {{kernel['function_name']}}({{- ["Type type", "const cell_idx_t ghost_layers", "gpuStream_t gpuStream"] | type_identifier_list -}})
{
+ if (!validInnerOuterSplit_ && type != Type::ALL)
+ WALBERLA_ABORT_NO_DEBUG_INFO("innerOuterSplit too large - make it smaller, increase cellsPerBlock or avoid communication hiding")
+
switch (type)
{
- case Type::INNER:
- return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", "gpuStream"] | type_identifier_list -}}); };
- case Type::OUTER:
- return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", "gpuStream"] | type_identifier_list -}}); };
- default:
- return [{{- ["this", "ghost_layers", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", "ghost_layers", "gpuStream"] | type_identifier_list -}}); };
+ case Type::INNER:
+ return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", "gpuStream"] | type_identifier_list -}}); };
+ case Type::OUTER:
+ return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", "gpuStream"] | type_identifier_list -}}); };
+ default:
+ return [{{- ["this", "ghost_layers", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", "ghost_layers", "gpuStream"] | type_identifier_list -}}); };
}
}
std::function<void (IBlock *)> {{kernel['function_name']}}({{- ["Type type", "gpuStream_t gpuStream"] | type_identifier_list -}})
{
+ if (!validInnerOuterSplit_ && type != Type::ALL)
+ WALBERLA_ABORT_NO_DEBUG_INFO("innerOuterSplit too large - make it smaller, increase cellsPerBlock or avoid communication hiding")
+
switch (type)
{
- case Type::INNER:
- return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", "gpuStream"] | type_identifier_list -}}); };
- case Type::OUTER:
- return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", "gpuStream"] | type_identifier_list -}}); };
- default:
- return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", "cell_idx_c(0)", "gpuStream"] | type_identifier_list -}}); };
+ case Type::INNER:
+ return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Inner({{- ["block", "gpuStream"] | type_identifier_list -}}); };
+ case Type::OUTER:
+ return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}Outer({{- ["block", "gpuStream"] | type_identifier_list -}}); };
+ default:
+ return [{{- ["this", "gpuStream"] | type_identifier_list -}}](IBlock* block) { {{kernel['function_name']}}({{- ["block", "cell_idx_c(0)", "gpuStream"] | type_identifier_list -}}); };
}
}
{%- endif %}
@@ -247,24 +259,24 @@ public:
layers_.push_back(ci);
}
- {%if target is equalto 'gpu'%}
+ {%if target is equalto 'gpu'%}
{
auto parallelSection_ = parallelStreams_.parallelSection( gpuStream );
for( auto & ci: layers_ )
{
- parallelSection_.run([&]( auto s ) {
- {{kernel['function_name']}}CellInterval({{kernel['kernel']|generate_function_collection_call(cell_interval='ci')}});
- });
+ parallelSection_.run([&]( auto s ) {
+ {{kernel['function_name']}}CellInterval({{kernel['kernel']|generate_function_collection_call(cell_interval='ci')}});
+ });
}
}
- {% else %}
+ {% else %}
for( auto & ci: layers_ )
{
{{kernel['function_name']}}CellInterval({{kernel['kernel']|generate_function_collection_call(cell_interval='ci')}});
}
- {% endif %}
+ {% endif %}
- {{kernel['kernel']|generate_swaps|indent(9)}}
+ {{kernel['kernel']|generate_swaps|indent(9)}}
}
{% endfor %}
@@ -275,17 +287,18 @@ public:
}
{%endif%}
- private:
- shared_ptr< StructuredBlockStorage > blocks_;
- {{kernel_list|generate_members(parameter_registration=parameter_scaling)|indent(4)}}
+ private:
+ shared_ptr< StructuredBlockStorage > blocks_;
+ {{kernel_list|generate_members(parameter_registration=parameter_scaling)|indent(4)}}
- Cell outerWidth_;
- std::vector<CellInterval> layers_;
+ Cell outerWidth_;
+ std::vector<CellInterval> layers_;
+ bool validInnerOuterSplit_;
- {%if target is equalto 'gpu' -%}
- gpu::ParallelStreams parallelStreams_;
- // std::map<BlockID, gpuStream_t > streams_;
- {%- endif %}
+ {%if target is equalto 'gpu' -%}
+ gpu::ParallelStreams parallelStreams_;
+ // std::map<BlockID, gpuStream_t > streams_;
+ {%- endif %}
};
diff --git a/src/core/Variant.h b/src/core/Variant.h
index 2be3f0a6c09e562dd48ef54b460dc08d786683d6..94c281593e2592e9dee4c26c14fd4ba947a70e87 100644
--- a/src/core/Variant.h
+++ b/src/core/Variant.h
@@ -21,9 +21,18 @@
#pragma once
+// suppress conversion warning that is emitted by icpc compiler with `-Werror`
+#include "waLBerlaDefinitions.h"
+#if ( defined WALBERLA_CXX_COMPILER_IS_INTEL )
+#pragma warning(push)
+#pragma warning disable 68
+#endif
#include <variant>
+#if ( defined WALBERLA_CXX_COMPILER_IS_INTEL )
+#pragma warning(pop)
+#endif
namespace walberla
diff --git a/tests/lbm_generated/Example.py b/tests/lbm_generated/Example.py
index 5233639be24c6574cee6440300bfe73e22e5e2ae..c4e3eaa96d596b2745811e30739790748740f884 100644
--- a/tests/lbm_generated/Example.py
+++ b/tests/lbm_generated/Example.py
@@ -23,7 +23,8 @@ with CodeGeneration() as ctx:
omega = sp.symbols("omega")
stencil = LBStencil(Stencil.D3Q19)
- pdfs, vel_field = fields(f"pdfs({stencil.Q}), velocity({stencil.D}): {data_type}[{stencil.D}D]", layout='fzyx')
+ pdfs, vel_field = fields(f"pdfs({stencil.Q}), velocity({stencil.D}): {data_type}[{stencil.D}D]",
+ layout='fzyx')
macroscopic_fields = {'velocity': vel_field}