Merge pull request #2 from Ricks-Lab/v1.1.0-CFG_Features

V1.1.0 cfg features

Author: Ricks-Lab

APPS_REF/REF_RESULTS/ref-result.setiathome_8.00_x86_64-pc-linux-gnu.blc04_2bit_guppi_57976_10365_HIP74315_0035.16799.409.22.45.56.vlar.wu.sah

@@ -1,25 +1,25 @@
 shmget in attach_shmem: Invalid argument
-18:44:07 (55031): Can't set up shared mem: -1. Will run in standalone mode.
+13:10:42 (101437): Can't set up shared mem: -1. Will run in standalone mode.
 setiathome_v8 8.00 Revision: 3290 g++ (GCC) 4.4.7 20120313 (Red Hat 4.4.7-4)
 libboinc: BOINC 7.7.0
 
 Work Unit Info:
 ...............
-WU true angle range is :  0.010899
+WU true angle range is :  0.006654
 Optimal function choices:
 --------------------------------------------------------
                             name   timing   error
 --------------------------------------------------------
                 v_BaseLineSmooth (no other)
     v_vGetPowerSpectrumUnrolled2 0.000034 0.00000 
-                 avx_ChirpData_d 0.002838 0.00000 
-                  AK SSE folding 0.000532 0.00000 
+                 avx_ChirpData_d 0.002794 0.00000 
+                  AK SSE folding 0.000539 0.00000 
 
-Flopcounter: 23944820171561.273438
+Flopcounter: 26027631425407.781250
 
 Spike count:    0
-Autocorr count: 0
-Pulse count:    6
-Triplet count:  0
+Autocorr count: 1
+Pulse count:    4
+Triplet count:  1
 Gaussian count: 0
-19:36:29 (55031): called boinc_finish(0)
+14:06:17 (101437): called boinc_finish(0)

APPS_REF/REF_RESULTS/ref-result.setiathome_8.00_x86_64-pc-linux-gnu.blc13_2bit_guppi_58405_85972_GJ687_0028.13304.818.22.45.142.vlar.wu.sah

@@ -2,37 +2,67 @@
 ## Blank lines and any part of a line beginning with # are ignored
 #############################################################################
 ##
-## List of applications with desired arguments to replace the
-## default arguments in the bench script.
+## List of applications with desired arguments.
 ##
 ## Format as would be used for executing the application from a
 ## command line, app -arg -arg etc. Multiple instances of the
 ## same app with different (or same) arguments will run that
-## many if the app is in the APPS_[C,G]PU directories.
+## many if the app is in the APPS_[C,G]PU directories, although
+## the --num_repetitions argument is the preferred way of running
+## an entry more than once.
 ##
 ## Needs the full application name with extension. Zero to many
-## arguments are possible.
+## arguments are possible. The -device N option of a GPU app will
+## be ingored, as this command line option is used to manage slot
+## assignment.  Specifing physical GPUs can be accomplished with the
+## --max_gpus X and --gpu_devices 0,1 options. The value X must be 
+## equal to the number of devices specified.
 ##
 ##
 ##############################################################################
-## Set benchMT command line options - to be implemented
+## Set benchMT command line options
 ##############################################################################
-## TODO 
-## In a future the release, the capability of specifying benchMT
-## command line options from BenchCFG file will be added.
-#mode arg argv
+## 
+## Command line options can be specified as modes in the BenchCFG file or an
+## alternate CFG file specified on the command line.  Options specified on
+## the command line will override those specified with mode in a CFG file.
+## 
+##
+#Don't ask confirmation before running jobs
+#mode yes False
+#Specify name of this run
+#mode run_name test
+#Specify path for BOINC
+#mode boinc_home /home/boinc/BOINC/
+#Do not suspend BOINC
+#mode noBS False
+#Display compact run status
+#mode display_compact False
+#Display run status by slots instead of jobs
+#mode display_slots False
+#Specify number of times to run benchmark
+#mode num_repetitions 2
+#Specify max number of threads to load
+#mode max_threads 2
+#Specify max number of GPUs to load
+#mode max_gpus 2
+#Specify GPU devices to use
+#mode gpu_devices 0,1
+#Use standard signal WUs instead of Test WUs
+#mode std_signals True
 ##
 ##
 ##############################################################################
 ## Entries to define benchmark run
 ##############################################################################
+##
 MBv8_8.22r3584_sse2_clAMD_HD5_x86_64-pc-linux-gnu -v 1 -instances_per_device 1 -sbs 2048 -period_iterations_num 1 -tt 500 -spike_fft_thresh 8192 -tune 1 64 1 4 -oclfft_tune_gr 256 -oclfft_tune_lr 16 -oclfft_tune_wg 256 -oclfft_tune_ls 512 -oclfft_tune_bn 64 -oclfft_tune_cw 64 -hp -high_perf -no_defaults_scaling
 #MBv8_8.22r3584_sse2_clAMD_HD5_x86_64-pc-linux-gnu -v 1
 
 #MBv8_8.04r3306_sse2_linux64 --nographics
 #MBv8_8.04r3306_sse41_linux64 --nographics
 ##MBv8_8.04r3306_ssse3_linux64 --nographics
-#MBv8_8.22r3711_sse41_x86_64-pc-linux-gnu --nographics
+MBv8_8.22r3711_sse41_x86_64-pc-linux-gnu --nographics
 #MBv8_8.22r3712_avx2_x86_64-pc-linux-gnu --nographics
 #MBv8_8.04r3306_sse42_linux64 --nographics
 #MBv8_8.05r3345_avx_linux64 --nographics

…4315_0035.16799.409.22.45.56.vlar.wu.txt …687_0028.13304.818.22.45.142.vlar.wu.txtAPPS_REF/REF_RESULTS/ref-stderr.setiathome_8.00_x86_64-pc-linux-gnu.blc04_2bit_guppi_57976_10365_HIP74315_0035.16799.409.22.45.56.vlar.wu.txt renamed to APPS_REF/REF_RESULTS/ref-stderr.setiathome_8.00_x86_64-pc-linux-gnu.blc13_2bit_guppi_58405_85972_GJ687_0028.13304.818.22.45.142.vlar.wu.txt APPS_REF/REF_RESULTS/ref-stderr.setiathome_8.00_x86_64-pc-linux-gnu.blc04_2bit_guppi_57976_10365_HIP74315_0035.16799.409.22.45.56.vlar.wu.txt renamed to APPS_REF/REF_RESULTS/ref-stderr.setiathome_8.00_x86_64-pc-linux-gnu.blc13_2bit_guppi_58405_85972_GJ687_0028.13304.818.22.45.142.vlar.wu.txt

@@ -5,17 +5,18 @@
  less than total number of CPU threads can be specified in the command line.  This tool will read a
  list of MB apps/args from the benchCFG file and search for the specified MB apps in the APP_CPU
  and APP_GPU directories to validate and determine platform.  It will then leverage allocated
- threads, as specified, to run all benchmark jobs storing results in the testData directory.  Use
+ threads, as specified, to run all benchmark jobs, storing results in the testData directory.  Use
  the *--help* option to get a description of valid command line arguments.
 
  By default, a summary list of all jobs will update in the display as the program progresses.  If
- there are a large number of jobs, then this display may not be useful and the --display_slots
+ there are a large number of jobs, then this display may not be useful and the *--display_slots*
  option can be used to display the status of each slot as the program progresses.  In some cases,
- there will be too many slots to display, and the --display_compact option can used to further
+ there will be too many slots to display, and the *--display_compact* option can used to further
  optimize the progress display.
 
  You may need to use the *--boinc_home* command option to specify the boinc home directory, which
- is required, since boinccmd is used.
+ is required, since boinccmd is used. An alternative BenchCFG file can be specified with the 
+ command line option *--cfg_file filename*.
 
  All WUs in the WU_test directory will be used in the creation of jobs to be run, unless the 
  *--ref_signals* option is used, in which case, WUs in the WU_std_signal will be used.  The
@@ -30,13 +31,19 @@
  files for each job run. A run name can be specified with the *--run_name* commane line option. This
  name will be included in the name of the testData subdirectory for the current run.
 
-
-## Development plans
-* GPU multi-threaded implementation. Currently total_gpu_threads = total_gpu_count, a future development opportunity is to implement a max number of threads per GPU
+## New in this Release  -  V1.1.0
+* Command line options can now be specified in mode lines of the BenchCFG file.  Options given on the command line will override modes specified in the CFG file.
+* An alternative CFG file can now be specified as a command line option.
+* Signal counts and Angle Range are now included in the psv and txt summary files.
+* Remove app *-device N* arg if specified, since -device is automatically added based on slot assignment.
+* Added *--gpu_devices x,y* command line option to specify which GPU devices the user would like to include in the benchmark run.
+* Added a lock_file in the working directory to prevent a second occurrence of benchMT from using the same directory.
+* Updated the 15 reference WUs in the *WU_test/safe* directory.
+* Changed *--ref_signals* option to *--std_signals* for clarity.
+
+## Development Plans and Known Limitations
+* Currently, running more than one job at a time on a single GPU is not supported. 
 * Consider using opencl instead of lshw to get valid GPU compute platforms, but maybe won't work for cuda apps
-* Read benchMT command line options from mode lines in BenchCFG file
-* Remove -device arg if specified, since -device is automatically added based on slot assignment
-* Need to make a lock_file in the working directory to prevent a second occurrence of benchMT from using the same directory
 * Should consider executing job with time command.  This should give total and CPU time metrics
 * Need to figure out how to run a job without a shell
 * Deal with an immediate fail to spawn a process when executing a job

BenchCFG

@@ -0,0 +1,10 @@
+# WU_std_signals - Standard Signal Work Units
+
+  This directory contains 4 standard signal Work Units for which there are reference results already provided
+  in the *APPS_REF/REF_RESULTS* directory.  These WUs are for testing functionality and do not represent the 
+  complexity and variability of real Work Units.  Their processing time is approximately 10x faster than a
+  real work units and are meant for testing only.
+
+## Known Issues
+* None of these WUs contain a Gaussian type signal.
+