tools/funccount: support funccount on specified CPU (iovisor#3059)

A typical case of this feature: count timer setting on a x86 server
for each CPU:
 for i in `seq 0 39`;
   do ./funccount.py -i 1 lapic_next_deadline -d 5 -c $i;
 done

Then we can know the timer setting is balanced of not and do some
futher work.

Signed-off-by: zhenwei pi <[email protected]>

man/man8/funccount.8

@@ -38,6 +38,9 @@ Use regular expressions for the search pattern.
 .TP
 \-D
 Print the BPF program before starting (for debugging purposes).
+.TP
+\-c CPU
+Trace on this CPU only.
 .SH EXAMPLES
 .TP
 Count kernel functions beginning with "vfs_", until Ctrl-C is hit:
@@ -75,6 +78,10 @@ Count all GC USDT probes in the Node process:
 Count all malloc() calls in libc:
 #
 .B funccount c:malloc
+.TP
+Count kernel functions beginning with "vfs_" on CPU 1 only:
+#
+.B funccount \-c 1 'vfs_*'
 .SH FIELDS
 .TP
 FUNC

tools/funccount.py

@@ -4,7 +4,8 @@
 # funccount Count functions, tracepoints, and USDT probes.
 # For Linux, uses BCC, eBPF.
 #
-# USAGE: funccount [-h] [-p PID] [-i INTERVAL] [-d DURATION] [-T] [-r] pattern
+# USAGE: funccount [-h] [-p PID] [-i INTERVAL] [-d DURATION] [-T] [-r]
+# [-c CPU] pattern
 #
 # The pattern is a string with optional '*' wildcards, similar to file
 # globbing. If you'd prefer to use regular expressions, use the -r option.
@@ -34,7 +35,7 @@ def verify_limit(num):
  (probe_limit, num))
 
 class Probe(object):
- def __init__(self, pattern, use_regex=False, pid=None):
+ def __init__(self, pattern, use_regex=False, pid=None, cpu=None):
  """Init a new probe.
 
  Init the probe from the pattern provided by the user. The supported
@@ -79,6 +80,7 @@ def __init__(self, pattern, use_regex=False, pid=None):
  self.library = libpath
 
  self.pid = pid
+ self.cpu = cpu
  self.matched = 0
  self.trace_functions = {} # map location number to function name
 
@@ -164,7 +166,8 @@ def _generate_functions(self, template):
  def load(self):
  trace_count_text = b"""
 int PROBE_FUNCTION(void *ctx) {
- FILTER
+ FILTERPID
+ FILTERCPU
  int loc = LOCATION;
  u64 *val = counts.lookup(&loc);
  if (!val) {
@@ -182,11 +185,18 @@ def load(self):
  # We really mean the tgid from the kernel's perspective, which is in
  # the top 32 bits of bpf_get_current_pid_tgid().
  if self.pid:
- trace_count_text = trace_count_text.replace(b'FILTER',
+ trace_count_text = trace_count_text.replace(b'FILTERPID',
  b"""u32 pid = bpf_get_current_pid_tgid() >> 32;
  if (pid != %d) { return 0; }""" % self.pid)
  else:
- trace_count_text = trace_count_text.replace(b'FILTER', b'')
+ trace_count_text = trace_count_text.replace(b'FILTERPID', b'')
+
+ if self.cpu:
+ trace_count_text = trace_count_text.replace(b'FILTERCPU',
+ b"""u32 cpu = bpf_get_smp_processor_id();
+ if (cpu != %d) { return 0; }""" % int(self.cpu))
+ else:
+ trace_count_text = trace_count_text.replace(b'FILTERCPU', b'')
 
  bpf_text += self._generate_functions(trace_count_text)
  bpf_text = bpf_text.replace(b"NUMLOCATIONS",
@@ -224,6 +234,7 @@ def __init__(self):
  ./funccount go:os.* # count all "os.*" calls in libgo
  ./funccount -p 185 go:os.* # count all "os.*" calls in libgo, PID 185
  ./funccount ./test:read* # count "read*" calls in the ./test binary
+ ./funccount -c 1 'vfs_*' # count vfs calls on CPU 1 only
  """
  parser = argparse.ArgumentParser(
  description="Count functions, tracepoints, and USDT probes",
@@ -241,13 +252,16 @@ def __init__(self):
  help="use regular expressions. Default is \"*\" wildcards only.")
  parser.add_argument("-D", "--debug", action="store_true",
  help="print BPF program before starting (for debugging purposes)")
+ parser.add_argument("-c", "--cpu",
+ help="trace this CPU only")
  parser.add_argument("pattern",
  type=ArgString,
  help="search expression for events")
  self.args = parser.parse_args()
  global debug
  debug = self.args.debug
- self.probe = Probe(self.args.pattern, self.args.regexp, self.args.pid)
+ self.probe = Probe(self.args.pattern, self.args.regexp, self.args.pid,
+ self.args.cpu)
  if self.args.duration and not self.args.interval:
  self.args.interval = self.args.duration
  if not self.args.interval:

tools/funccount_example.txt

@@ -325,11 +325,45 @@ tcp_v4_send_check 30
 __tcp_v4_send_check 30
 Detaching...
 
+A cpu is specified by "-c CPU", this will only trace the specified CPU. Eg,
+trace how many timers setting per sencond of CPU 1 on a x86(Intel) server:
+
+# funccount.py -i 1 -c 1 lapic_next_deadline
+Tracing 1 functions for "lapic_next_deadline"... Hit Ctrl-C to end.
+
+FUNC COUNT
+lapic_next_deadline 3840
+
+FUNC COUNT
+lapic_next_deadline 3930
+
+FUNC COUNT
+lapic_next_deadline 4701
+
+FUNC COUNT
+lapic_next_deadline 5895
+
+FUNC COUNT
+lapic_next_deadline 5591
+
+FUNC COUNT
+lapic_next_deadline 4727
+
+FUNC COUNT
+lapic_next_deadline 5560
+
+FUNC COUNT
+lapic_next_deadline 5416
+^C
+FUNC COUNT
+lapic_next_deadline 372
+Detaching...
 
 Full USAGE:
 
 # ./funccount -h
-usage: funccount [-h] [-p PID] [-i INTERVAL] [-d DURATION] [-T] [-r] [-D]
+usage: funccount.py [-h] [-p PID] [-i INTERVAL] [-d DURATION] [-T] [-r] [-D]
+ [-c CPU]
  pattern
 
 Count functions, tracepoints, and USDT probes
@@ -349,6 +383,7 @@ optional arguments:
  only.
  -D, --debug print BPF program before starting (for debugging
  purposes)
+ -c CPU, --cpu CPU trace this CPU only
 
 examples:
  ./funccount 'vfs_*' # count kernel fns starting with "vfs"
@@ -362,3 +397,4 @@ examples:
  ./funccount go:os.* # count all "os.*" calls in libgo
  ./funccount -p 185 go:os.* # count all "os.*" calls in libgo, PID 185
  ./funccount ./test:read* # count "read*" calls in the ./test binary
+ ./funccount -c 1 'vfs_*' # count vfs calls on CPU 1 only