File tree Expand file tree Collapse file tree 3 files changed +29
-3
lines changed
Expand file tree Collapse file tree 3 files changed +29
-3
lines changed Original file line number Diff line number Diff line change 1+ lr
2+ lrun -T1 ./sw4ck sw4ck.in
3+ Reading from file sw4ck.in
4+ Launching sw4 kernels
5+
6+ Kernel 1 time 4.6487
7+ Kernel 2 time 1.31594
8+ Kernel 3 time 1.29174
9+ Kernel 4 time 1.25392
10+ Kernel 5 time 5.20592
11+
12+ Total kernel runtime = 13
13+
14+ MIN = -1.5735458151501336641e-05
15+ MAX = 0.010639705916308216105
16+
17+ Norm of output 202.0512747393526638
18+ Norm of output 202.0512747393526638
19+ Error = 0 %
Original file line number Diff line number Diff line change 1+ * .bc
2+ * .cui
3+ * .o
4+ * .s
5+ * .lst
6+ sw4ck
7+ sw4ck+pat
Original file line number Diff line number Diff line change 11# modules loaded should look like
2- #1 ) texlive/2016 2 ) StdEnv (S) 3 ) opt 4 ) rocm/3.10.0 5 ) openmpi-gnu/4.0 6 ) gnu/8.2.1
2+ #1 ) texlive/2016 2 ) StdEnv (S) 3 ) opt 4 ) rocm/4.0.1 5 ) openmpi-gnu/4.0 6 ) gnu/8.2.1
33#use make -f Makefile.hipcc
44LINKER = hipcc
55CXX = $(PREP) hipcc
66RAJA_LOCATION = /home/users/coe0145/RAJA/Dev2/raja/install_906
77
88HIP_ROOT_DIR = $(HIP_PATH)
99
10- HSA_ROOT_DIR = /opt/rocm-3.10.0 /hsa
10+ HSA_ROOT_DIR = /opt/rocm-4.0.1 /hsa
1111
1212GCC_LINK_LINE = -L $(HIP_ROOT_DIR)/lib -L $(HSA_ROOT_DIR)/lib -lhsa-runtime64 -lm -lstdc++ -ldl
1313
1414
15- MORE_FLAGS = -I${ROCM_PATH}/include -isystem /opt/rocm-3.10.0 /llvm/lib/clang/11.0.0 /include/.. -isystem /opt/rocm-3.10.0 /hsa/include -isystem /opt/rocm-3.10.0 /hip/include -D__HIP_ARCH_GFX906__=1 --cuda-gpu-arch=gfx906 -O3 -x hip -Winconsistent-missing-override --amdgpu-target=gfx906 --save-temps -fgpu-rdc -DNO_RAJA=1 -Wall -Wno-unknown-pragmas
15+ MORE_FLAGS = -I${ROCM_PATH}/include -isystem /opt/rocm-4.0.1 /llvm/lib/clang/11.0.0 /include/.. -isystem /opt/rocm-4.0.1 /hsa/include -isystem /opt/rocm-4.0.1 /hip/include -D__HIP_ARCH_GFX906__=1 --cuda-gpu-arch=gfx906 -O3 -x hip -Winconsistent-missing-override --amdgpu-target=gfx906 --save-temps -fgpu-rdc -DNO_RAJA=1 -Wall -Wno-unknown-pragmas
1616LINKFLAGS = -std=c++11 --amdgpu-target=gfx906 -fgpu-rdc
1717EXTRA_CXX_FLAGS = -O3 -x hip -std=c++11 -I$(RAJA_LOCATION)/include -DENABLE_HIP $(MORE_FLAGS) -Winconsistent-missing-override
1818
You can’t perform that action at this time.
0 commit comments