Changes between Version 17 and Version 18 of AppCoprocessor

Timestamp:

May 29, 2009, 2:44:52 PM (15 years ago)

Author:

davea

Comment:

--

AppCoprocessor

@@ -25 +25 @@
 
  * Choose a "planning class" name for the program, say "cuda" (see below).
- * Create an [UpdateVersions app version].  Include a file '''app_plan''' containing "cuda".
- * Link the following function into your scheduler:
+ * Create an [UpdateVersions app version], specifying its plan class as "cuda".
+ * Link the following function into your scheduler (customize as needed):
 {{{
-bool app_plan(SCHEDULER_REQUEST& sreq, char* plan_class, HOST_USAGE& hu) {
+int app_plan(SCHEDULER_REQUEST& sreq, char* plan_class, HOST_USAGE& hu) {
     if (!strcmp(plan_class, "cuda")) {
         // the following is for an app that uses a CUDA GPU
-        // and some CPU also, and gets 50 GFLOPS total
         //
-        for (unsigned int i=0; i<sreq.coprocs.coprocs.size(); i++) {
-            COPROC* cp = sreq.coprocs.coprocs[i];
-            if (!strcmp(cp->type, "CUDA")) {
-                COPROC* cu = new COPROC (cp->type);
-                cu->count = 1;
-                hu.coprocs.coprocs.push_back(cu);
-                double x = 1e9/sreq.host.p_fpops;
-                if (x > 1) x = 1;
-                hu.avg_ncpus = x;
-                hu.max_ncpus = x;
-                hu.flops = 5e11;
-                return true;
+        COPROC_CUDA* cp = (COPROC_CUDA*)sreq.coprocs.lookup("CUDA");
+        if (!cp) {
+            if (config.debug_version_select) {
+                log_messages.printf(MSG_NORMAL,
+                    "[version] Host lacks CUDA coprocessor for plan class cuda\n"
+                );
             }
+            return PLAN_REJECT_CUDA_NO_DEVICE;
         }
+        int v = (cp->prop.major)*100 + cp->prop.minor;
+        if (v < 100) {
+            if (config.debug_version_select) {
+                log_messages.printf(MSG_NORMAL,
+                    "[version] CUDA version %d < 1.0\n", v
+                );
+            }
+            return PLAN_REJECT_CUDA_VERSION;
+        } 
+
+        if (cp->drvVersion && cp->drvVersion < PLAN_CUDA_MIN_DRIVER_VERSION) {
+            if (config.debug_version_select) {
+                log_messages.printf(MSG_NORMAL,
+                    "[version] NVIDIA driver version %d < PLAN_CUDA_MIN_DRIVER_VERSION\n",
+                    cp->drvVersion
+                );
+            }
+            return PLAN_REJECT_NVIDIA_DRIVER_VERSION;
+        }
+
+        if (cp->prop.dtotalGlobalMem < PLAN_CUDA_MIN_RAM) {
+            if (config.debug_version_select) {
+                log_messages.printf(MSG_NORMAL,
+                    "[version] CUDA mem %d < %d\n",
+                    cp->prop.dtotalGlobalMem, PLAN_CUDA_MIN_RAM
+                );
+            }
+            return PLAN_REJECT_CUDA_MEM;
+        }
+        hu.flops = cp->flops_estimate();
+
+        // assume we'll need 0.5% as many CPU FLOPS as GPU FLOPS
+        // to keep the GPU fed.
+        //
+        double x = (hu.flops*0.005)/sreq.host.p_fpops;
+        hu.avg_ncpus = x;
+        hu.max_ncpus = x;
+
+        hu.ncudas = 1;
+
         if (config.debug_version_select) {
-            log_messages.printf(MSG_DEBUG,
-                "Host lacks CUDA coprocessor for plan class %s\n", plan_class
+            log_messages.printf(MSG_NORMAL,
+                "[version] CUDA app estimated %.2f GFLOPS (clock %d count %d)\n",
+                hu.flops/1e9, cp->prop.clockRate,
+                cp->prop.multiProcessorCount
             );
         }
-        return false;
+        return 0;
     }
     log_messages.printf(MSG_CRITICAL,
         "Unknown plan class: %s\n", plan_class
     );
-    return false;
+    return PLAN_REJECT_UNKNOWN;
 }
+}}}
 
-}}}
 == Questions ==
 
  * How does BOINC know if non-BOINC applications are using resources?
 
-== Possible future additions ==
 
- * Allow app_versions to specify min and max requirements (and have a corresponding allocation scheme in the client).
- * Let projects define their own resources, unknown to BOINC, and have "probe" programs (using the assigned-job mechanism) that surveys the resources on each host.
-