HP Caliper User Guide Release 5.5 (5900-2351, August 2012)

If you set the --cpu-details means option, only the MEAN and STDEV statistics are
reported. This is the default.
• A list of samples, with each sample showing both sampled values and statistics derived from
those values. To report samples (not shown by default), specify --cpu-details
means:samples or --cpu-details statistics:samples.
Example Command Lines for Text Report
$ caliper cpu -o cpu.txt program
This command runs program, measuring and reporting the overview metrics by taking one
sample every 8 milliseconds. By default, 125 low-level samples will be aggregated into one
user-reported sample, resulting in one aggregated sample collected per second. The text report is
saved in the text file cpu.txt.
$ caliper cpu -o cpu.txt -s 100ms
--cpu-aggregation 10 ./wordplay thequickbrownfox
This command runs the program wordplay , measuring and reporting the overview metrics by
taking a sample every 100 milliseconds and aggregating 10 low-level samples into one user-reported
sample. The result is saved in the text file cpu.txt in the current directory.
$ caliper cpu -m cpi:user,cpi:kernel my_app
This command gets separate CPI data at user and kernel privilege levels.
$ caliper cpu -o cpu.txt -w -e 120
This command collects system-wide overview metrics for two minutes and saves the result in the
text file cpu.txt in the current directory.
Example Command Line for CSV Report
$ caliper cpu --csv csvout ./wordplay thequickbrownfox
CPU Event Sets
For the cpu measurement report, you specify one or more comma-separated lists of predefined
event sets.
The available event sets are:
brpath Provides information on the dynamic mix of branch types, branch path
distribution, branches per instruction, and so forth.
brpred Provides metrics that are useful in assessing the effectiveness of branch
prediction.
c2c Provides “cache-to-cache” metrics related to cache coherence activity.
cpi Provides metrics related to cycles per instruction (CPI).
cpubus Provides information on the demand that a specific CPU presents to the central
electronics complex (CEC), the chip set surrounding the CPU, and the demand
the CPU experiences due to the CEC traffic initiated by other CPUs or I/O
components in the system.
cspec Provides metrics on the effectiveness of control speculation.
dispersal Provides a qualitative view of the parallelism that is available as seen at
instruction dispersal.
dspec Provides metrics on the effectiveness of data speculation.
fp Provides information relating to floating-point operation density, execution
rate, and flush/trap events density.
l1dcache Provides miss rate information for the L1 data cache.
l1icache Provides miss and prefetch usage information for the L1 instruction cache.
l2cache Provides miss rate information for the L2 unified cache.
182 Descriptions of Measurement Reports