CPUCPSTAT(8)CPUCPSTAT(8)NAMEcpupcstat - Measure the DTLB miss rate
SYNOPSIScpupcstat [options] [target]
DESCRIPTIONcpupcstat uses oprofile to measure the DTLB miss rate of a specified
application or the kernel. It configures oprofile to count the number
of DTLB misses, optionally starts the target, and reports on the miss
rate over a specified interval as target executes.
The following options can be used to configure how cpupcstat works:
--vmlinux </path/to/vmlinux>
This allows the user to specify where the appropriate vmlinux
file is for their kernel. If this is not specified,
/boot/vmlinux-`uname -r` will be used.
--delay <seconds>
This allows the user to specify the reporting interval. The
default is 10 seconds.
--target-global
Gather statistics for all processes and the kernel running in
the system.
--target-pid <pid>
This allows the user to specify the pid of a process already
that is already running. If this option is specified, target
will be ignored.
--real-target <real-target>
Use this to specify the real name of the program to monitor if
the target is a launcher script. When this is specified, target
is executed but the report will be for real-target.
--time-limit <sec>
This option sets the time limit for monitoring. If this is
specified the target or pid will only be monitored for sec sec‐
onds. The default continues monitoring while target or pid are
still alive.
--kernel
This allows the user to request DTLB miss rate data be collected
for the kernel as well as the target.
--misses-per-instruction
This option requests that the ratio of instructions retired per
TLB miss.
--misses-per-cycle
This option requests that the ratio of CPU cycles per TLB miss.
--time-servicing
This option requests that the percentage of CPU cycles spent
servicing TLB misses is displayed when cpupcstat exits. To use
this option the cost in CPU cycles for a single TLB miss must be
specified using either the --cost-config option or the tlb‐
miss_cost.sh script.
--cost-config </path/to/config>
This option tells cpupcstat that the cost in CPU cycles of a TLB
miss can be found in the specified file, it should be specified
as:
TLB_MISS_COST=XX
Where XX is the cost in cycles. This option is only used with
the --time-servicing option.
--force-oprofile
cpupcstat prefers the perf tool for data collection, only using
oprofile if perf is not present or supported. This option will
force cpupcstat to use oprofile for data collection.
SEE ALSOoprofile(1)perf(1)tlbmiss_cost.sh(8)AUTHORS
Eric B Munson <ebmunson@us.ibm.com> is the primary author. See the doc‐
umentation for other contributors.
9 June, 2009 CPUCPSTAT(8)