Merge pull request #385 from OrderN/release-1.4

Release 1.4

.github/workflows/makefile.yml

@@ -2,9 +2,9 @@ name: Makefile CI
 
 on:
   push:
-    branches: [ "develop" ]
+    branches: [ "develop", "f-exx-opt" ]
   pull_request:
-    branches: [ "develop" ]
+    branches: [ "develop", "f-exx-opt" ]
   # Manual dispatch for including all multiply kernels in matrix.
   # We don't want this to run on every commit, but it's useful
   # to be able to launch it manually
@@ -70,11 +70,9 @@ jobs:
             multiply_kernel: ompDojk
           - test_all_multiply_kernels: false
             multiply_kernel: ompGemm
-          - test_all_multiply_kernels: false
-            multiply_kernel: ompGemm_m
 
     steps:
-    - uses: actions/checkout@v3
+    - uses: actions/checkout@v4
       with:
         fetch-depth: '0'
 
@@ -84,7 +82,7 @@ jobs:
         sudo apt install openmpi-bin libopenmpi-dev libfftw3-dev libblas3 liblapack3 libscalapack-openmpi-dev libxc-dev
 
     - name: Set up Python
-      uses: actions/setup-python@v4
+      uses: actions/setup-python@v5
       with:
         python-version: '3.10'
 
@@ -119,7 +117,31 @@ jobs:
       run: |
         mpirun -np ${{matrix.np}} ../../bin/Conquest
         cat Conquest_out
-        
+
+    - name: Run test 004
+      working-directory: ${{github.workspace}}/testsuite/test_004_isol_C2H4_4proc_PBE0CRI
+      run: |
+        mpirun -np ${{matrix.np}} ../../bin/Conquest
+        cat Conquest_out
+
+    - name: Run test 005
+      working-directory: ${{github.workspace}}/testsuite/test_005_isol_C2H4_4proc_PBE0GTO
+      run: |
+        mpirun -np ${{matrix.np}} ../../bin/Conquest
+        cat Conquest_out
+    
+    - name: Run test 006
+      working-directory: ${{github.workspace}}/testsuite/test_006_isol_C2H4_4proc_PBE0ERI
+      run: |
+        mpirun -np ${{matrix.np}} ../../bin/Conquest
+        cat Conquest_out
+
+    - name: Run test 007
+      working-directory: ${{github.workspace}}/testsuite/test_007_isol_CH_spinpol_1proc_PBE0CRI
+      run: |
+        mpirun -np ${{matrix.np}} ../../bin/Conquest
+        cat Conquest_out
+
     - name: Check test results
       working-directory: ${{github.workspace}}/testsuite
       run: pytest test_check_output.py

.gitignore

@@ -1,4 +1,7 @@
 bin/*
 deps.obj.inc
 *~
-src/system/system.make
+src/system/system.make
+
+# Ignore folders within each benchmark but not files
+benchmarks/*/**/

README.md

@@ -50,4 +50,4 @@ DOI for the specific version you have used in any given study.
 
 ## Version number
 
-CONQUEST is now at version 1.3 (tag:v1.3)
+CONQUEST is now at version 1.4 (tag:v1.4)

benchmarks/.gitignore

@@ -0,0 +1,8 @@
+__pycache__
+Conquest_o*
+Conquest_w*
+*.i00*
+*.bib
+*.dat
+*.log
+fort.*

benchmarks/reframe/exx_runonly.py

@@ -0,0 +1,220 @@
+# Import modules from reframe and excalibur-tests
+import reframe as rfm
+import reframe.utility.sanity as sn
+
+class ConquestBaseBenchmark(rfm.RunOnlyRegressionTest):
+
+    # Run configuration
+    ## Mandatory ReFrame setup
+    valid_systems = ['-gpu']
+    valid_prog_environs = ['default']
+
+    ## Executable
+    executable_opts = ['']
+
+    ## Scheduler options
+    time_limit = '30m'
+
+    conquest_base_dir = variable(str, value='foo')
+
+    @run_after('setup')
+    def setup_variables(self):
+        self.executable = f"{self.conquest_base_dir}/bin/Conquest"
+        self.num_tasks = self.num_tasks_param
+        self.num_cpus_per_task = self.num_cpus_per_task_param
+        self.env_vars['OMP_NUM_THREADS'] = f'{self.num_cpus_per_task}'
+        self.env_vars['OMP_STACKSIZE'] = '100M'
+
+        if self.current_partition.scheduler.registered_name == 'sge':
+            # `self.num_tasks` or `self.num_cpus_per_task` may be `None`, here
+            # we default to `1` if not set.
+            num_tasks = self.num_tasks or 1
+            num_cpus_per_task = self.num_cpus_per_task or 1
+            # Set the total number of CPUs to be requested for the SGE scheduler.
+            # Set to a full node size to reduce runtime variance.
+            self.extra_resources['mpi'] = {'num_slots': self.current_partition.processor.num_cpus} #num_tasks * num_cpus_per_task}
+
+    @sanity_function
+    def validate_solution(self):
+        return sn.assert_found(r'Total run time was:.*', self.stdout)
+
+    @performance_function('s', perf_key='Runtime')
+    def extract_runtime_perf(self):
+        return sn.extractsingle(r'Total run time was:\s+(\S+)\s+seconds', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_setup_runtime')
+    def extract_exx_setup_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_setup\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_fetch_runtime')
+    def extract_exx_fetch_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_fetch\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_evalpao_runtime')
+    def extract_exx_evalpao_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_evalpao\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_allocat_runtime')
+    def extract_exx_allocat_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_allocat\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_dealloc_runtime')
+    def extract_exx_dealloc_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_dealloc\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_kernel_runtime')
+    def extract_exx_kernel_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_kernel\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('s', perf_key='exx_total_runtime')
+    def extract_exx_total_runtime_perf(self):
+        return sn.extractsingle(r'Time spent in exx_total\s+time:\s+=\s+(\S+)\ss', self.stdout, 1, float)
+
+    @performance_function('MB', perf_key='Memory')
+    def extract_memory_perf(self):
+        return sn.extractsingle(r'Max total mem use is\s+(\S+)\s+MB', self.stdout, 1, float)
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0_2(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.2"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.2/C_PBE_DZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.2/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.2/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.2/H_PBE_DZP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0_4(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.4"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.4/C_PBE_DZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.4/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.4/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.4/H_PBE_DZP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0_6(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.6"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.6/C_PBE_DZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.6/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.6/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullDZP_0.6/H_PBE_DZP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0CRI_fullSZP_0_4_SCF(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0CRI_fullSZP_0.4_SCF"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullSZP_0.4_SCF/C_PBE_SZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullSZP_0.4_SCF/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullSZP_0.4_SCF/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullSZP_0.4_SCF/H_PBE_SZP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0CRI_fullTZTP_0_6(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0CRI_fullTZTP_0.6"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullTZTP_0.6/C_PBE_TZTP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullTZTP_0.6/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullTZTP_0.6/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0CRI_fullTZTP_0.6/H_PBE_TZTP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0_4_SCF(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0.4_SCF"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0.4_SCF/C_PBE_SZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0.4_SCF/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0.4_SCF/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0.4_SCF/H_PBE_SZP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF/C_PBE_SZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF/C.gto .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF/H_PBE_SZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_GTO_SCF/H.gto .")
+
+@rfm.simple_test
+class test_EXX_isol_C6H6_4proc_PBE0CRI_fullDZP_0_6(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C6H6_4proc_PBE0CRI_fullDZP_0.6"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullDZP_0.6/C_PBE_DZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullDZP_0.6/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullDZP_0.6/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullDZP_0.6/H_PBE_DZP_CQ.ion .")
+
+@rfm.simple_test
+class test_EXX_isol_C6H6_4proc_PBE0CRI_fullSZP_0_6(ConquestBaseBenchmark):
+
+    tags = {"test_EXX_isol_C6H6_4proc_PBE0CRI_fullSZP_0.6"}
+    num_tasks_param = parameter([2,4])
+    num_cpus_per_task_param = parameter([1,2,4,8])
+
+    @run_before('run')
+    def get_input(self):
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullSZP_0.6/C_PBE_SZP_CQ.ion .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullSZP_0.6/Conquest_coord .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullSZP_0.6/Conquest_input .")
+        self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/test_EXX_isol_C6H6_4proc_PBE0CRI_fullSZP_0.6/H_PBE_SZP_CQ.ion .")
+
+
+# @rfm.simple_test
+# class Water64(ConquestBaseBenchmark):
+
+#     tags = {"water64"}
+#     num_tasks_param = 8
+#     num_cpus_per_task = 4
+
+#     @run_before('run')
+#     def get_input(self):
+#         self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/water_64mols/Conquest_input .")
+#         self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/water_64mols/H2O_coord.in .")
+#         self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/water_64mols/H_SZ.ion .")
+#         self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/water_64mols/H_SZP.ion .")
+#         self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/water_64mols/O_SZ.ion .")
+#         self.prerun_cmds.append(f"cp {self.conquest_base_dir}/benchmarks/water_64mols/O_SZP.ion .")