perf tools: Emit clearer message for sys_perf_event_open ENOENT return
authorArnaldo Carvalho de Melo <acme@redhat.com>
Mon, 10 Jan 2011 15:36:24 +0000 (13:36 -0200)
committerArnaldo Carvalho de Melo <acme@redhat.com>
Mon, 10 Jan 2011 15:36:24 +0000 (13:36 -0200)
Improve sys_perf_event_open ENOENT return handling in top and record, just
like 5a3446b does for stat.

Cc: David Ahern <daahern@cisco.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools/perf/builtin-record.c
tools/perf/builtin-top.c

index 7bc0490..7069bd3 100644 (file)
@@ -331,6 +331,9 @@ try_again:
                        else if (err ==  ENODEV && cpu_list) {
                                die("No such device - did you specify"
                                        " an out-of-range profile CPU?\n");
+                       } else if (err == ENOENT) {
+                               die("%s event is not supported. ",
+                                    event_name(evsel));
                        } else if (err == EINVAL && sample_id_all_avail) {
                                /*
                                 * Old kernel, no attr->sample_id_type_all field
index 1e67ab9..6ce4042 100644 (file)
@@ -1247,6 +1247,8 @@ try_again:
                                die("Permission error - are you root?\n"
                                        "\t Consider tweaking"
                                        " /proc/sys/kernel/perf_event_paranoid.\n");
+                       if (err == ENOENT)
+                               die("%s event is not supported. ", event_name(evsel));
                        /*
                         * If it's cycles then fall back to hrtimer
                         * based cpu-clock-tick sw counter, which