perf core: Fix perf_proc_update_handler() bug
authorStephane Eranian <eranian@google.com>
Fri, 11 Jan 2019 01:17:16 +0000 (17:17 -0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 13 Mar 2019 21:02:26 +0000 (14:02 -0700)
[ Upstream commit 1a51c5da5acc6c188c917ba572eebac5f8793432 ]

The perf_proc_update_handler() handles /proc/sys/kernel/perf_event_max_sample_rate
syctl variable.  When the PMU IRQ handler timing monitoring is disabled, i.e,
when /proc/sys/kernel/perf_cpu_time_max_percent is equal to 0 or 100,
then no modification to sysctl_perf_event_sample_rate is allowed to prevent
possible hang from wrong values.

The problem is that the test to prevent modification is made after the
sysctl variable is modified in perf_proc_update_handler().

You get an error:

  $ echo 10001 >/proc/sys/kernel/perf_event_max_sample_rate
  echo: write error: invalid argument

But the value is still modified causing all sorts of inconsistencies:

  $ cat /proc/sys/kernel/perf_event_max_sample_rate
  10001

This patch fixes the problem by moving the parsing of the value after
the test.

Committer testing:

  # echo 100 > /proc/sys/kernel/perf_cpu_time_max_percent
  # echo 10001 > /proc/sys/kernel/perf_event_max_sample_rate
  -bash: echo: write error: Invalid argument
  # cat /proc/sys/kernel/perf_event_max_sample_rate
  10001
  #

Signed-off-by: Stephane Eranian <eranian@google.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Reviewed-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1547169436-6266-1-git-send-email-eranian@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
kernel/events/core.c

index 4fb9d50..aa996a0 100644 (file)
@@ -436,18 +436,18 @@ int perf_proc_update_handler(struct ctl_table *table, int write,
                void __user *buffer, size_t *lenp,
                loff_t *ppos)
 {
-       int ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
-
-       if (ret || !write)
-               return ret;
-
+       int ret;
+       int perf_cpu = sysctl_perf_cpu_time_max_percent;
        /*
         * If throttling is disabled don't allow the write:
         */
-       if (sysctl_perf_cpu_time_max_percent == 100 ||
-           sysctl_perf_cpu_time_max_percent == 0)
+       if (write && (perf_cpu == 100 || perf_cpu == 0))
                return -EINVAL;
 
+       ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
+       if (ret || !write)
+               return ret;
+
        max_samples_per_tick = DIV_ROUND_UP(sysctl_perf_event_sample_rate, HZ);
        perf_sample_period_ns = NSEC_PER_SEC / sysctl_perf_event_sample_rate;
        update_perf_cpu_limits();