p generate a trace with pingpong, and replay itself, then check that output trace of the second run is the same as in the first (once sorted)
! setenv LD_LIBRARY_PATH=../../lib
! output sort
-$ ${bindir:=.}/../../../bin/smpirun -trace-ti --cfg=tracing/filename:out_in_ti.txt --cfg=smpi/cpu_threshold:-1 -map -hostfile ../hostfile -platform ../../../examples/platforms/small_platform.xml -np 4 ${bindir:=.}/pingpong -q --log=smpi_kernel.thres:warning
+$ ${bindir:=.}/../../../bin/smpirun -trace-ti --cfg=tracing/filename:out_in_ti.txt --cfg=smpi/cpu_threshold:-1 -map -hostfile ../hostfile -platform ../../../examples/platforms/small_platform.xml -np 4 ${bindir:=.}/pt2pt-pingpong -q --log=smpi_kernel.thres:warning
> *** Ping-pong test (MPI_Send/MPI_Recv) ***
> == pivot=0 : pingpong [0] <--> [1]
> == pivot=1 : pingpong [1] <--> [2]
$ rm out_ti.txt
$ rm out_in_ti.txt
-
-
p Same test, but only using one output file for all processes
p generate a trace with pingpong, and replay itself, then check that output trace of the second run is the same as in the first (once sorted)
! output sort
-$ ${bindir:=.}/../../../bin/smpirun -trace-ti --cfg=tracing/filename:out_in_ti.txt --cfg=tracing/smpi/format/ti_one_file:yes -map -hostfile ../hostfile -platform ../../../examples/platforms/small_platform.xml -np 4 ${bindir:=.}/pingpong -q --log=smpi_kernel.thres:warning
+$ ${bindir:=.}/../../../bin/smpirun -trace-ti --cfg=tracing/filename:out_in_ti.txt --cfg=tracing/smpi/format/ti_one_file:yes -map -hostfile ../hostfile -platform ../../../examples/platforms/small_platform.xml -np 4 ${bindir:=.}/pt2pt-pingpong -q --log=smpi_kernel.thres:warning
> *** Ping-pong test (MPI_Send/MPI_Recv) ***
> == pivot=0 : pingpong [0] <--> [1]
> == pivot=1 : pingpong [1] <--> [2]
> [0.000000] [xbt_cfg/INFO] Configuration change: Set 'tracing/smpi/format/ti_one_file' to 'yes'
> [Jupiter:1:(2) 0.016798] [smpi_replay/INFO] Simulation time 0.016798
-
-
$ rm -rf ./out_ti.txt_files
$ rm -rf ./out_in_ti.txt_files
$ rm out_ti.txt
$ rm out_in_ti.txt
-
-
-