Changes between Version 2 and Version 3 of PlanClassFunc

Timestamp:

Jun 6, 2012, 11:20:37 AM (12 years ago)

Author:

davea

Comment:

--

PlanClassFunc

@@ -1 +1 @@
 = Specifying plan classes in C++ =
+
+If you need more generality than is provided by
+[AppPlanSpec XML-based plan class specification],
+you can specify plan classes in C++.
 
 The scheduler is linked with a function
@@ -34 +38 @@
 }}}
 
-You are free to define your own set of plan classes,
-and to link your own '''app_plan()''' function with the scheduler.
+You can define your own set of plan classes,
+and link your own '''app_plan()''' function with the scheduler.
 The BOINC scheduler comes with a default '''app_plan()''' (in sched/sched_customize.cpp).
+
+== Example: a plan class for multithread apps ==
+
+Here's a plan class function for a multicore app that it achieves a linear speedup on up to 64 processors,
+and no additional speedup beyond that.
+
+{{{
+bool app_plan_mt(
+    SCHEDULER_REQUEST& sreq, HOST_USAGE& hu
+) {
+    double ncpus = g_wreq->effective_ncpus;
+        // number of usable CPUs, taking user prefs into account
+    int nthreads = (int)ncpus;
+    if (nthreads > 64) nthreads = 64;
+    hu.avg_ncpus = nthreads;
+    hu.max_ncpus = nthreads;
+    sprintf(hu.cmdline, "--nthreads %d", nthreads);
+    hu.projected_flops = sreq.host.p_fpops*hu.avg_ncpus*.99;
+        // the .99 ensures that on uniprocessors a sequential app
+        // will be used in preferences to this
+    hu.peak_flops = sreq.host.p_fpops*hu.avg_ncpus;
+    return true;
+}
+}}}
 
 == Defining GPU plan classes ==