summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorSteven Rostedt <srostedt@redhat.com>2009-06-01 21:51:28 -0400
committerSteven Rostedt <rostedt@goodmis.org>2009-06-01 23:26:23 -0400
commit0f6ce3de4ef6ff940308087c49760d068851c1a7 (patch)
tree26a732ee4dd5b0f5612f0385893ec483c6ba4ab3
parent112f38a7e36e9d688b389507136bf3af3e6d159b (diff)
downloadop-kernel-dev-0f6ce3de4ef6ff940308087c49760d068851c1a7.zip
op-kernel-dev-0f6ce3de4ef6ff940308087c49760d068851c1a7.tar.gz
ftrace: do not profile functions when disabled
A race was found that if one were to enable and disable the function profiler repeatedly, then the system can panic. This was because a profiled function may be preempted just before disabling interrupts. While the profiler is disabled and then reenabled, the preempted function could start again, and access the hash as it is being initialized. This just adds a check in the irq disabled part to check if the profiler is enabled, and if it is not then it will just exit. When the system is disabled, the profile_enabled variable is cleared before calling the unregistering of the function profiler. This unregistering calls stop machine which also acts as a synchronize schedule. [ Impact: fix panic in enabling/disabling function profiler ] Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
-rw-r--r--kernel/trace/ftrace.c8
1 files changed, 6 insertions, 2 deletions
diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index 2074e5b..d6973df 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -599,7 +599,7 @@ function_profile_call(unsigned long ip, unsigned long parent_ip)
local_irq_save(flags);
stat = &__get_cpu_var(ftrace_profile_stats);
- if (!stat->hash)
+ if (!stat->hash || !ftrace_profile_enabled)
goto out;
rec = ftrace_find_profiled_func(stat, ip);
@@ -630,7 +630,7 @@ static void profile_graph_return(struct ftrace_graph_ret *trace)
local_irq_save(flags);
stat = &__get_cpu_var(ftrace_profile_stats);
- if (!stat->hash)
+ if (!stat->hash || !ftrace_profile_enabled)
goto out;
calltime = trace->rettime - trace->calltime;
@@ -724,6 +724,10 @@ ftrace_profile_write(struct file *filp, const char __user *ubuf,
ftrace_profile_enabled = 1;
} else {
ftrace_profile_enabled = 0;
+ /*
+ * unregister_ftrace_profiler calls stop_machine
+ * so this acts like an synchronize_sched.
+ */
unregister_ftrace_profiler();
}
}
OpenPOWER on IntegriCloud