On Sat, 07 Jun 2014 23:41:21 -0400
Sasha Levin <[email protected]> wrote:
> Hi Steven,
>
> Yup, it took me *that* long to reproduce it again, but I can confirm that that
> BUG() gets hit (the printk shows cpu 30 like the BUG):
>
> [ 2410.677199] kernel BUG at kernel/trace/ring_buffer.c:563!
> [ 2410.679445] can: request_module (can-proto-4) failed.
> [ 2410.680298] invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
> [ 2410.680298] Dumping ftrace buffer:
> [ 2410.680298] (ftrace buffer empty)
> [ 2410.680298] Modules linked in:
> [ 2410.680298] CPU: 30 PID: 34851 Comm: trinity-c88 Not tainted
> 3.15.0-rc8-next-20140606-sasha-00021-ga9d3a0b-dirty #596
> [ 2410.680298] task: ffff8802c866b000 ti: ffff8802c7724000 task.ti:
> ffff8802c7724000
> [ 2410.680298] RIP: ring_buffer_wait (kernel/trace/ring_buffer.c:563)
> [ 2410.680298] RSP: 0018:ffff8802c7727de8 EFLAGS: 00010296
> [ 2410.680298] RAX: 0000000000000013 RBX: 0000000000000024 RCX:
> 0000000000000006
> [ 2410.680298] RDX: 0000000000000001 RSI: ffffffffad5030db RDI:
> ffffffffaa1d8952
> [ 2410.711484] RBP: ffff8802c7727e38 R08: 0000000000000000 R09:
> 0000000000000000
> [ 2410.711484] R10: 0000000000000001 R11: 0000000000000000 R12:
> ffff88003681e900
> [ 2410.711484] R13: ffff88006ce7d100 R14: 0000000000000000 R15:
> ffff8800530090fc
> [ 2410.721370] FS: 00007f8c14bad700(0000) GS:ffff8806cae00000(0000)
> knlGS:0000000000000000
> [ 2410.721370] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 2410.721370] CR2: 00007f8c11440000 CR3: 000000029dd18000 CR4:
> 00000000000006a0
> [ 2410.721370] DR0: 00000000006d6000 DR1: 00000000006d6000 DR2:
> 0000000000000000
> [ 2410.721370] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7:
> 0000000000000600
> [ 2410.721370] Stack:
> [ 2410.721370] ffff880053008028 0000000000000000 ffff8802c866b000
> ffffffffaa1bb600
> [ 2410.721370] ffff8802c7727e08 ffff8802c7727e08 ffff880053008000
> ffff880053008028
> [ 2410.721370] ffff88006ce7d100 ffff8802c866b000 ffff8802c7727e48
> ffffffffaa24af8a
> [ 2410.721370] Call Trace:
> [ 2410.721370] ? bit_waitqueue (kernel/sched/wait.c:291)
> [ 2410.721370] wait_on_pipe (kernel/trace/trace.c:1095)
> [ 2410.721370] tracing_wait_pipe.isra.19 (kernel/trace/trace.c:4280)
> [ 2410.721370] tracing_read_pipe (kernel/trace/trace.c:4326)
> [ 2410.721370] vfs_read (fs/read_write.c:430)
> [ 2410.721370] SyS_read (fs/read_write.c:568 fs/read_write.c:560)
> [ 2410.721370] tracesys (arch/x86/kernel/entry_64.S:542)
Found it.
Let me guess, if you ls /sys/kernel/debug/tracing/per_cpu, you'll see
more CPUs than you have on the box. I have a box like this, and when I
cat the trace_pipe from one of the CPUs that don't exist, I trigger the
bug.
This patch should fix the bug.
-- Steve
diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h
index d69cf63..49a4d6f 100644
--- a/include/linux/ring_buffer.h
+++ b/include/linux/ring_buffer.h
@@ -97,7 +97,7 @@ __ring_buffer_alloc(unsigned long size, unsigned flags,
struct lock_class_key *k
__ring_buffer_alloc((size), (flags), &__key); \
})
-void ring_buffer_wait(struct ring_buffer *buffer, int cpu);
+int ring_buffer_wait(struct ring_buffer *buffer, int cpu);
int ring_buffer_poll_wait(struct ring_buffer *buffer, int cpu,
struct file *filp, poll_table *poll_table);
diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index c634868..7c56c3d 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -543,7 +543,7 @@ static void rb_wake_up_waiters(struct irq_work *work)
* as data is added to any of the @buffer's cpu buffers. Otherwise
* it will wait for data to be added to a specific cpu buffer.
*/
-void ring_buffer_wait(struct ring_buffer *buffer, int cpu)
+int ring_buffer_wait(struct ring_buffer *buffer, int cpu)
{
struct ring_buffer_per_cpu *cpu_buffer;
DEFINE_WAIT(wait);
@@ -557,6 +557,8 @@ void ring_buffer_wait(struct ring_buffer *buffer, int cpu)
if (cpu == RING_BUFFER_ALL_CPUS)
work = &buffer->irq_work;
else {
+ if (!cpumask_test_cpu(cpu, buffer->cpumask))
+ return -ENODEV;
cpu_buffer = buffer->buffers[cpu];
work = &cpu_buffer->irq_work;
}
@@ -591,6 +593,7 @@ void ring_buffer_wait(struct ring_buffer *buffer, int cpu)
schedule();
finish_wait(&work->waiters, &wait);
+ return 0;
}
/**
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 16f7038..56422f1 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -1085,13 +1085,13 @@ update_max_tr_single(struct trace_array *tr, struct
task_struct *tsk, int cpu)
}
#endif /* CONFIG_TRACER_MAX_TRACE */
-static void wait_on_pipe(struct trace_iterator *iter)
+static int wait_on_pipe(struct trace_iterator *iter)
{
/* Iterators are static, they should be filled or empty */
if (trace_buffer_iter(iter, iter->cpu_file))
- return;
+ return 0;
- ring_buffer_wait(iter->trace_buffer->buffer, iter->cpu_file);
+ return ring_buffer_wait(iter->trace_buffer->buffer, iter->cpu_file);
}
#ifdef CONFIG_FTRACE_STARTUP_TEST
@@ -4378,6 +4378,7 @@ tracing_poll_pipe(struct file *filp, poll_table
*poll_table)
static int tracing_wait_pipe(struct file *filp)
{
struct trace_iterator *iter = filp->private_data;
+ int ret;
while (trace_empty(iter)) {
@@ -4399,10 +4400,13 @@ static int tracing_wait_pipe(struct file *filp)
mutex_unlock(&iter->mutex);
- wait_on_pipe(iter);
+ ret = wait_on_pipe(iter);
mutex_lock(&iter->mutex);
+ if (ret)
+ return ret;
+
if (signal_pending(current))
return -EINTR;
}
@@ -5327,8 +5331,12 @@ tracing_buffers_read(struct file *filp, char __user
*ubuf,
goto out_unlock;
}
mutex_unlock(&trace_types_lock);
- wait_on_pipe(iter);
+ ret = wait_on_pipe(iter);
mutex_lock(&trace_types_lock);
+ if (ret) {
+ size = ret;
+ goto out_unlock;
+ }
if (signal_pending(current)) {
size = -EINTR;
goto out_unlock;
@@ -5538,8 +5546,10 @@ tracing_buffers_splice_read(struct file *file, loff_t
*ppos,
goto out;
}
mutex_unlock(&trace_types_lock);
- wait_on_pipe(iter);
+ ret = wait_on_pipe(iter);
mutex_lock(&trace_types_lock);
+ if (ret)
+ goto out;
if (signal_pending(current)) {
ret = -EINTR;
goto out;
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
