* [tip:tracing/syscalls] tracing/syscalls: protect thread flag toggling from races
[not found] <1237151439-6755-3-git-send-email-fweisbec@gmail.com>
@ 2009-03-16 8:18 ` Frederic Weisbecker
0 siblings, 0 replies; only message in thread
From: Frederic Weisbecker @ 2009-03-16 8:18 UTC (permalink / raw)
To: linux-tip-commits; +Cc: linux-kernel, hpa, mingo, fweisbec, akpm, tglx, mingo
Commit-ID: 5be71b61f17b0e3bc8ad0b1a1b7b53ab7d574ebb
Gitweb: http://git.kernel.org/tip/5be71b61f17b0e3bc8ad0b1a1b7b53ab7d574ebb
Author: Frederic Weisbecker <fweisbec@gmail.com>
AuthorDate: Sun, 15 Mar 2009 22:10:37 +0100
Commit: Ingo Molnar <mingo@elte.hu>
CommitDate: Mon, 16 Mar 2009 09:13:16 +0100
tracing/syscalls: protect thread flag toggling from races
Impact: fix syscall tracer enable/disable race
The current thread flag toggling is racy as shown in the following
scenario:
- task A is the last user of syscall tracing, it releases the
TIF_SYSCALL_FTRACE on each tasks
- at the same time task B start syscall tracing. refcount == 0 so
it sets up TIF_SYSCALL_FTRACE on each tasks.
The effect of the mixup is unpredictable.
So this fix adds a mutex on {start,stop}_syscall_tracing().
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Reported-by: Ingo Molnar <mingo@elte.hu>
LKML-Reference: <1237151439-6755-3-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
kernel/trace/trace_syscalls.c | 24 +++++++++++++++++++-----
1 files changed, 19 insertions(+), 5 deletions(-)
diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c
index 26f9a86..a2a3af2 100644
--- a/kernel/trace/trace_syscalls.c
+++ b/kernel/trace/trace_syscalls.c
@@ -5,7 +5,11 @@
#include "trace_output.h"
#include "trace.h"
-static atomic_t refcount;
+/* Keep a counter of the syscall tracing users */
+static int refcount;
+
+/* Prevent from races on thread flags toggling */
+static DEFINE_MUTEX(syscall_trace_lock);
/* Option to display the parameters types */
enum {
@@ -96,9 +100,11 @@ void start_ftrace_syscalls(void)
unsigned long flags;
struct task_struct *g, *t;
+ mutex_lock(&syscall_trace_lock);
+
/* Don't enable the flag on the tasks twice */
- if (atomic_inc_return(&refcount) != 1)
- return;
+ if (++refcount != 1)
+ goto unlock;
arch_init_ftrace_syscalls();
read_lock_irqsave(&tasklist_lock, flags);
@@ -108,6 +114,9 @@ void start_ftrace_syscalls(void)
} while_each_thread(g, t);
read_unlock_irqrestore(&tasklist_lock, flags);
+
+unlock:
+ mutex_unlock(&syscall_trace_lock);
}
void stop_ftrace_syscalls(void)
@@ -115,9 +124,11 @@ void stop_ftrace_syscalls(void)
unsigned long flags;
struct task_struct *g, *t;
+ mutex_lock(&syscall_trace_lock);
+
/* There are perhaps still some users */
- if (atomic_dec_return(&refcount))
- return;
+ if (--refcount)
+ goto unlock;
read_lock_irqsave(&tasklist_lock, flags);
@@ -126,6 +137,9 @@ void stop_ftrace_syscalls(void)
} while_each_thread(g, t);
read_unlock_irqrestore(&tasklist_lock, flags);
+
+unlock:
+ mutex_unlock(&syscall_trace_lock);
}
void ftrace_syscall_enter(struct pt_regs *regs)
^ permalink raw reply [flat|nested] only message in thread