update conceptual implementation, add few examples

Makefile.am

@@ -71,6 +71,9 @@ if USE_CONC
 src_libcodes_la_SOURCES += src/workload/methods/codes-conc-online-comm-wrkld.C
 AM_CPPFLAGS += ${CONCEPTUAL_CFLAGS} -DUSE_CONC=1
 LDADD += ${CONCEPTUAL_LIBS}
+src_libcodes_la_SOURCES += src/workload/conceputal-skeleton-apps/conc-bisect.c
+src_libcodes_la_SOURCES += src/workload/conceputal-skeleton-apps/conc-cosmoflow.c
+src_libcodes_la_SOURCES += src/workload/conceputal-skeleton-apps/conc-hotpotato.c
 src_libcodes_la_SOURCES += src/workload/conceputal-skeleton-apps/conc-latencyall.c
 src_libcodes_la_SOURCES += src/workload/conceputal-skeleton-apps/conc-latency.c
 endif

scripts/conceptual_benchmarks/bisect.ncptl

+###################################################
+# Measure random bisection-bandwidth patterns     #
+# By Scott Pakin <pakin@lanl.gov>                 #
+#                                                 #
+# Inspired by Hoefler, Schneider, and Lumsdaine's #
+# "Multistage Switches are not Crossbars" paper   #
+###################################################
+
+Require language version "1.5".
+
+nummsgs is "Number of messages per trial" and comes from "--nummsgs" or "-n" with default 100.
+wups is "Number of warmup messages" and comes from "--wups" or "-w" with default 3.
+msgsize is "Message size in bytes" and comes from "--bytes" or "-b" with default 1M.
+numtrials is "Number of bisection patterns" and comes from "--trials" or "-t" with default 5000.
+rounding is "Round measurements to the nearest N" and comes from "--round" or "-r" with default 5.
+
+For each trial in {1, ..., numtrials} {
+  task 0 is assigned to processor 0 then
+  task 0 outputs "Testing random bisection pattern " and trial and "/" and numtrials then
+  all tasks are assigned to a random processor then
+  tasks src such that src is even send wups msgsize-byte messages to task src+1 then
+  all tasks synchronize then
+  all tasks reset their counters then
+  tasks src such that src is even send nummsgs msgsize-byte messages to task src+1 then
+  all tasks synchronize then
+  all tasks log a histogram of rounding*round(total_bytes*1E6/(elapsed_usecs*1M)/rounding) as "Bandwidth (MiB/s)"
+}

scripts/conceptual_benchmarks/cosmoflow.ncptl

+# Measure the performance of MPI_Allgather()
+# By Scott Pakin <pakin@lanl.gov>
+#
+# N.B. Requires the c_mpi backend.
+Require language version "1.5".
+# Parse the command line.
+numwords is "Message size (words)" and comes from "--msgsize" or "-s" with default 28825K.
+reps is "Number of repetitions" and comes from "--reps" or "-r" with default 100.
+computetime is "Computation time (ms)" and comes from "--compute" or "-c" with default 129.
+
+# Allocate a send buffer and a receive buffer.
+Task 0 multicasts a numwords*num_tasks word message from buffer 0 to all other tasks.
+Task 0 multicasts a numwords*num_tasks word message from buffer 1 to all other tasks.
+
+# Measure the performance of MPI_Allreduce().
+Task 0 resets its counters then
+for reps repetitions {
+    all tasks COMPUTES FOR computetime MILLISECONDS then
+    all tasks backend execute "
+        MPI_Allreduce([MESSAGE BUFFER 0], [MESSAGE BUFFER 1], (int)" and numwords and ",
+                MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+        " then
+    all tasks backend execute "
+        MPI_Allreduce([MESSAGE BUFFER 0], [MESSAGE BUFFER 1], (int)" and numwords and ",
+                MPI_INT, MPI_SUM, MPI_COMM_WORLD);
+        "
+} then
+task 0 logs elapsed_usecs/1000 as "Elapse time (ms)".

scripts/conceptual_benchmarks/hotpotato.ncptl

+# Virtual ring "hot potato" test
+
+Require language version "1.5".
+
+trials is "Number of trials to perform" and comes from "--trials" or
+  "-t" with default 100000.
+
+Assert that "the hot-potato test requires at least two tasks" with num_tasks>=2.
+
+Let len be 0 while {
+  for each task_count in {2, ..., num_tasks} {
+    task 0 outputs "Performing " and trials and " " and
+		   task_count and "-task runs...." then
+    for trials repetitions plus 5 warmup repetitions and a synchronization {
+      task 0 resets its counters then
+      task 0 sends a len byte message to unsuspecting task 1 then
+      task (n+1) mod task_count receives a len byte message from task n such that n<task_count then
+      task n such that n>0 /\ n<task_count sends a len byte message to unsuspecting task (n+1) mod task_count then
+      task 0 logs the task_count as "# of tasks" and
+		  the minimum of elapsed_usecs/task_count as "Latency (usecs)" and
+		  the mean of elapsed_usecs/task_count as "Latency (usecs)" and
+		  the variance of elapsed_usecs/task_count as "Latency (usecs)"
+    } then
+    task 0 computes aggregates
+  }
+}

scripts/conceptual_benchmarks/translate_conc_src.py

@@ -15,10 +15,7 @@ def eliminate_logging(inLines):
                 inLines[idx+i] = "//"+inLines[idx+i]  
         elif 'ncptl_free (logfile_uuid)' in line:
             for i in range(0, 12):
-                inLines[idx-i] = "//"+inLines[idx-i]
-        elif 'case 1' in line:
-            for i in range(5, 9):
-                inLines[idx+i] = "//"+inLines[idx+i]      
+                inLines[idx-i] = "//"+inLines[idx-i]    
         elif 'int mpiresult' in line:
             for i in range(0,30):
                 inLines[idx+i] = "//"+inLines[idx+i] 

src/networks/model-net/dragonfly-custom.C

@@ -2402,8 +2402,8 @@ void dragonfly_custom_router_final(router_state * s,
     written = 0;
     if(!s->router_id)
     {
-        written = sprintf(s->output_buf, "# Format <LP ID> <Group ID> <Router ID> <Link Traffic per router port(s)>");
-        written += sprintf(s->output_buf + written, "# Router ports in the order: %d green links, %d black links %d global channels \n", 
+        written = sprintf(s->output_buf2, "# Format <LP ID> <Group ID> <Router ID> <Link Traffic per router port(s)>");
+        written += sprintf(s->output_buf2 + written, "# Router ports in the order: %d green links, %d black links %d global channels \n", 
                 p->num_router_cols * p->num_row_chans, p->num_router_rows * p->num_col_chans, p->num_global_channels);
     }
     written += sprintf(s->output_buf2 + written, "\n %llu %d %d",

src/workload/codes-conc-addon.c

@@ -10,6 +10,9 @@
 
 
 /* list of available benchmarks begin */
+extern struct codes_conceptual_bench bisect_bench;
+extern struct codes_conceptual_bench cosmoflow_bench;
+extern struct codes_conceptual_bench hotpotato_bench;
 extern struct codes_conceptual_bench latencyall_bench;
 extern struct codes_conceptual_bench latency_bench;
 /* list of available benchmarks end */
@@ -17,6 +20,9 @@ extern struct codes_conceptual_bench latency_bench;
 static struct codes_conceptual_bench const * bench_array_default[] =
 {
     /* default benchmarks begin */
+    &bisect_bench,
+    &cosmoflow_bench,
+    &hotpotato_bench,
     &latencyall_bench,
     &latency_bench,
     /* default benchmarks end */

src/workload/conceputal-skeleton-apps/conc-latency.c

 /**********************************************************************
- * This file was generated by coNCePTuaL on Thu Oct  4 23:46:17 2018
+ * This file was generated by coNCePTuaL on Mon Jan  7 23:09:59 2019
  * using the c_mpi backend (C + MPI).
  * Do not modify this file; modify /Users/xin/macworkspace/codes-dev/codes/scripts/conceptual_benchmarks/latency.ncptl instead.
  *

src/workload/conceputal-skeleton-apps/conc-latencyall.c

 /**********************************************************************
- * This file was generated by coNCePTuaL on Thu Oct  4 23:46:17 2018
+ * This file was generated by coNCePTuaL on Mon Jan  7 23:10:00 2019
  * using the c_mpi backend (C + MPI).
  * Do not modify this file; modify /Users/xin/macworkspace/codes-dev/codes/scripts/conceptual_benchmarks/latency-all.ncptl instead.
  *