server_scripts 0.1.2 → 0.1.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 7a56c1d87d7ec945dd7ff3ee8c29ff44b3012fed
-  data.tar.gz: f638a56fac759785ae663fcd6ca188b71c28948b
+  metadata.gz: 74bc6e9790b0f8f43c099ebbf2bb7097ba078f58
+  data.tar.gz: 507e795a3d6f9baefd62276f71d0a44b64484fd9
 SHA512:
-  metadata.gz: cb5b8f703ffe938248d2c3b5dda4a379df984770fbf5523bbefaf06675feb8502b4fea5c694d2cb18ea89d07b797e5bd2e48e1d6ae3f3270b83bd529b9258dbe
-  data.tar.gz: 79ca166d74f4aadc444d12aa6f4fe15a070ac263e1573131b386b60440a522554ead76d9bf730e284dbc7fd2e658496b42d7fa76087b1a3491df1cc1cb576ab3
+  metadata.gz: e9c5b973a951b906699ccfc51d59b72fed77658cf267c22d2cbbcf3e40e71cce27ded005948888ca89f1a878a7b1ed33774d551168e539adbe1d85a4fb8674dd
+  data.tar.gz: 3b052b8fbc0f94477ed790642bf3c2eb7a536370e8a9e506e13b1b688c2498e05ea022971b8e2468e4a301a47a7036ce941472a1f903fb1cd19922806dcf51f7

data/README.md

@@ -66,6 +66,47 @@ like setting the ITAC and VTUNE output file/folder names.
 
 ## Parse intel VTune output
 
+### Output chart of intel VTune
+
+The way VTune classfies the output in the CSV is a little funny and should be
+understood properly unless you want to have a hard time. The output can be said
+to be classified as a tree that looks like so:
+```
+CPU Time
+  - Effective Time
+    - Idle
+    - Poor
+    - Ok
+    - Ideal
+  - Spin Time
+    - Imbalance or Serial Spinning
+    - Lock Contention
+    - MPI Busy Wait Time
+    - Other
+  - Overhead Time
+    - Scheduling
+    - Reduction
+    - Atomics
+    - Other
+Wait Time
+  - Idle
+  - Poor
+  - Ok
+  - Ideal
+  - Over
+Wait Count
+PID
+TID
+```
+The total time the sum of `CPU Time` and `Wait Time`.
+
+### Usage
+A sample program for parsing the firt 16 threads reported by the vtune command:
+```
+vtune -report hotspots -group-by thread -result-dir result_file.vtune \
+    -report-output result_thread_res.csv -csv-delimiter=,
+```
+
 ``` ruby
 parser = Parser::VTune::Hotspots::SLATE.new(
   "test/artifacts/slate-two-proc-p1.csv", nthreads: 16)

data/lib/server_scripts/parser/vtune/hotspots/base.rb

@@ -7,7 +7,7 @@ module ServerScripts
           CPU_EFFECTIVE_TIME = "CPU Time:Effective Time"
           CPU_OVERHEAD_TIME = "CPU Time:Overhead Time"
           CPU_SPIN_TIME = "CPU Time:Spin Time"
-          MPI_BUSY_TIME = "CPU Time:Spin Time:MPI Busy Wait Time"
+          MPI_BUSY_WAIT_TIME = "CPU Time:Spin Time:MPI Busy Wait Time"
           WAIT_TIME = "Wait Time"
 
           def initialize fname

data/lib/server_scripts/parser/vtune/hotspots/threads/slate.rb

@@ -8,8 +8,11 @@ module ServerScripts
             super(fname)
           end
 
-          def total_mpi_busy_time
-            @total_mpi_busy_time ||= parse_for_event(:mpi_busy_time)
+          # Get the total time for all threads under the header
+          # "CPU Time:Spin Time:MPI Busy Wait Time". This time
+          # is included within "CPU Time:Spin Time".
+          def total_mpi_busy_wait_time
+            @total_mpi_busy_time ||= parse_for_event(:mpi_busy_wait_time)
             @total_mpi_busy_time
           end
 
@@ -22,10 +25,10 @@ module ServerScripts
               @threads[i] = {}
               @threads[i][:cpu_time] = data[CPU_TIME][i].to_f
               @threads[i][:cpu_effective_time] = data[CPU_EFFECTIVE_TIME][i].to_f
-              @threads[i][:cpu_overhead_time] = data[CPU_OVERHEAD_TIME][i].to_f +
-                data[CPU_SPIN_TIME][i].to_f
+              @threads[i][:cpu_overhead_time] = data[CPU_OVERHEAD_TIME][i].to_f
+              @threads[i][:cpu_spin_time] = data[CPU_SPIN_TIME][i].to_f
               @threads[i][:wait_time] = data[WAIT_TIME][i].to_f
-              @threads[i][:mpi_busy_time] = data[MPI_BUSY_TIME][i].to_f
+              @threads[i][:mpi_busy_wait_time] = data[MPI_BUSY_WAIT_TIME][i].to_f
             end
           end
         end # class SLATE

data/lib/server_scripts/parser/vtune/hotspots/threads.rb

@@ -32,11 +32,18 @@ module ServerScripts
             @total_cpu_effective_time
           end
           
-          # Total CPU overhead: sum of CPU Spin Time + CPU Overhead Time
+          # Total CPU overhead: "CPU Time:Overhead Time"
           def total_cpu_overhead_time
             @total_cpu_overhead_time ||= parse_for_event(:cpu_overhead_time)
             @total_cpu_overhead_time
           end
+
+          # Total CPU Spin time: "CPU Time:Spin Time". This includes the MPI busy
+          # wait time, which for some reason is classified under this banner by vtune.
+          def total_cpu_spin_time
+            @total_cpu_spin_time ||= parse_for_event(:cpu_spin_time)
+            @total_cpu_spin_time
+          end
           
           # Total Wait Time.
           def total_wait_time
@@ -52,8 +59,8 @@ module ServerScripts
               @threads[i] = {}
               @threads[i][:cpu_time] = data[CPU_TIME][i].to_f
               @threads[i][:cpu_effective_time] = data[CPU_EFFECTIVE_TIME][i].to_f
-              @threads[i][:cpu_overhead_time] = data[CPU_OVERHEAD_TIME][i].to_f +
-                data[CPU_SPIN_TIME][i].to_f
+              @threads[i][:cpu_overhead_time] = data[CPU_OVERHEAD_TIME][i].to_f
+              @threads[i][:cpu_spin_time] = data[CPU_SPIN_TIME][i].to_f
               @threads[i][:wait_time] = data[WAIT_TIME][i].to_f
             end
           end

data/lib/server_scripts/version.rb

@@ -1,3 +1,3 @@
 module ServerScripts
-  VERSION = "0.1.2"
+  VERSION = "0.1.3"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: server_scripts
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.1.3
 platform: ruby
 authors:
 - Sameer Deshmukh