MFP4: Enhancements and bug-fixes to pmcstat(8):

 - The '-c' option now takes a comma-separated list of CPU
   numbers, or a literal '*' denoting all CPUs in the system.
   Subsequent system PMCs are allocated on the CPUs so specified.

   Change the default behaviour to allocate system PMCs on all CPUs,
   not just CPU 0.

   Update the manual page and add an example of how to use the new
   functionality.

 - Attach PMCs to a (commandline) child process more reliably.  This
   fixes a long standing bug in counting events incurred by short-lived
   processes.

1 files changed, 18 insertions, 9 deletions

diff --git a/usr.sbin/pmcstat/pmcstat.8 b/usr.sbin/pmcstat/pmcstat.8
index fffc945..cd7ed74 100644
--- a/usr.sbin/pmcstat/pmcstat.8
+++ b/usr.sbin/pmcstat/pmcstat.8
@@ -1,4 +1,4 @@
-.\" Copyright (c) 2003-2006 Joseph Koshy.  All rights reserved.
+.\" Copyright (c) 2003-2007 Joseph Koshy.  All rights reserved.
 .\"
 .\" Redistribution and use in source and binary forms, with or without
 .\" modification, are permitted provided that the following conditions
@@ -23,7 +23,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd April 2, 2006
+.Dd April 22, 2007
 .Os
 .Dt PMCSTAT 8
 .Sh NAME
@@ -40,7 +40,7 @@
 .Op Fl R Ar logfilename
 .Op Fl S Ar event-spec
 .Op Fl W
-.Op Fl c Ar cpu
+.Op Fl c Ar cpu-spec
 .Op Fl d
 .Op Fl g
 .Op Fl k Ar kerneldir
@@ -180,11 +180,16 @@ This is an experimental feature intended to help analyse the
 dynamic behaviour of processes in the system.
 It may incur substantial overhead if enabled.
 The default is for this feature to be disabled.
-.It Fl c Ar cpu
-Set the cpu for subsequent system mode PMCs specified on the
+.It Fl c Ar cpu-spec
+Set the cpus for subsequent system mode PMCs specified on the
 command line to
-.Ar cpu .
-The default is to allocate system mode PMCs on CPU zero.
+.Ar cpu-spec .
+Argument
+.Ar cpu-spec
+is a comma separated list of CPU numbers, or the literal
+.Sq *
+denoting all CPUs.
+The default is to allocate system mode PMCs on all CPUs.
 .It Fl d
 Toggle between process mode PMCs measuring events for the target
 process' current and future children or only measuring events for
@@ -268,8 +273,12 @@ and measure the number of data cache misses suffered
 by it and its children every 12 seconds on an AMD Athlon, use:
 .Dl "pmcstat -d -w 12 -p k7-dc-misses mozilla"
 .Pp
-To perform system-wide sampling based on processor instructions
-retired use:
+To count instruction tlb-misses on CPUs 0 and 2 on a Intel
+Pentium Pro/Pentium III SMP system use:
+.Dl "pmcstat -c 0,2 -s p6-itlb-miss"
+.Pp
+To perform system-wide sampling on all configured processors
+based on processor instructions retired use:
 .Dl "pmcstat -S instructions -O /tmp/sample.out"
 .Pp
 To send the generated event log to a remote machine use: