p Test compute
-! setenv LD_LIBRARY_PATH=../../lib
-! output sort
! timeout 5
-$ ../../bin/smpirun -platform ${srcdir:=.}/../../examples/msg/small_platform_with_routers.xml -hostfile ${srcdir:=.}/hostfile -np 2 ./compute 0
-> 0 2.000000
-> 0 2.000000
-> [0.000000] [surf_config/INFO] Switching workstation model to compound since you changed the network and/or cpu model(s)
-> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'maxmin/precision' to '1e-9'
-> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'network/model' to 'SMPI'
-> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'network/TCP_gamma' to '4194304'
+$ ../../smpi_script/bin/smpirun -platform ${srcdir:=.}/../../examples/msg/small_platform_with_routers.xml -hostfile ${srcdir:=.}/hostfile -np 3 --log=root.thres:warning ./compute 7 --log=smpi_kernel.thres:warning
+> 7 16.000000
+> 7 16.000000
+> 7 16.000000
+
+p Test compute only once
+! timeout 5
+$ ../../smpi_script/bin/smpirun -platform ${srcdir:=.}/../../examples/msg/small_platform_with_routers.xml -hostfile ${srcdir:=.}/hostfile -np 3 --log=root.thres:warning ./compute2 7 --log=smpi_kernel.thres:warning
+> 7 16.000000
p Test compute and bench
-! setenv LD_LIBRARY_PATH=../../lib
! output sort
! timeout 45
-$ ../../bin/smpirun -platform ${srcdir:=.}/../../examples/msg/small_platform_with_routers.xml -hostfile ${srcdir:=.}/hostfile -np 2 ./compute3
-> [0.000000] [surf_config/INFO] Switching workstation model to compound since you changed the network and/or cpu model(s)
-> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'maxmin/precision' to '1e-9'
-> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'network/model' to 'SMPI'
-> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'network/TCP_gamma' to '4194304'
-> [0] The result of the computation is: 65536.000000
-> [1] The result of the computation is: 2.000000
-> [rank:0] Run the first computation. It's globally benched, and I want no more than 3 benchmarks (thres<0)
-> [rank:0] Run the first computation. It's globally benched, and I want no more than 3 benchmarks (thres<0)
-> [rank:0] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
-> [rank:0] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
-> [rank:1] Run the first computation. It's globally benched, and I want no more than 3 benchmarks (thres<0)
-> [rank:1] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
-> [rank:1] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
-
+$ ../../smpi_script/bin/smpirun -platform ${srcdir:=.}/../../examples/msg/small_platform_with_routers.xml -hostfile ${srcdir:=.}/hostfile -np 3 --log=root.thres:warning ./compute3 quiet --log=smpi_kernel.thres:warning
+> (0) Run the first computation. It's globally benched, and I want no more than 4 benchmarks (thres<0)
+> (0) Run the first computation. It's globally benched, and I want no more than 4 benchmarks (thres<0)
+> (0) Run the first computation. It's globally benched, and I want no more than 4 benchmarks (thres<0)
+> (0) Run the first computation. It's globally benched, and I want no more than 4 benchmarks (thres<0)
+> (1) [rank:0] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
+> (1) [rank:0] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
+> (1) [rank:1] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
+> (1) [rank:1] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
+> (1) [rank:2] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
+> (1) [rank:2] Run the first (locally benched) computation. It's locally benched, and I want the standard error to go below 0.1 second (count is not >0)
+> (2) [rank:0] Done.
+> (2) [rank:1] Done.
+> (2) [rank:2] Done.
