Re: [PATCH v6 2/2] Output stall data in debugfs

From: Alex Neronskiy
Date: Thu Aug 11 2011 - 15:51:38 EST

On Thu, Aug 11, 2011 at 12:43 PM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> zakmagnus seems to bounce, so who am I talking to anyway..
> On Thu, 2011-08-11 at 21:35 +0200, Peter Zijlstra wrote:
>> static void show_stall_trace(struct seq_file *f, void *v)
>> {
>>         struct stall *s = f->private;
>>         int i, idx = ACCESS_ONCE(s->idx);
>>         mutex_lock(&stall_mutex);
>>         raw_spin_lock(&s->lock[idx]);
>>         seq_printf(f, "stall: %d\n", s->worst);
>>         for (i = 0; i < s->trace[idx].nr_entries; i++) {
>>                 seq_printf(f, "[<%pK>] %pS\n",
>>                         (void *)s->trace->entries[i],
>>                         (void *)s->trace->entries[i]);
>>         }
>>         raw_spin_unlock(&s->lock[idx]);
>>         mutex_unlock(&stall_mutex);
>> }
>> Yes its racy on s->worst, but who cares (if you do care you can keep a
>> copy in s->delay[idx] or so). Also, it might be better to not do the
>> spinlock but simply use an atomic bitop to set an in-use flag, there is
>> no reason to disable preemption over the seq_printf() loop.
> That also cures another problem you have, a seq_file buffer is only 1
> page large, you should be using the seqfile iterator interface and print
> one line at a time.. now clearly that won't work with preemption
> disabled either.
I didn't want to hold on to a lock while the lines were iterated over,
and instead just dump it in one go. I guess it's not a big deal to do
that, though. Nonetheless, is one page not enough?
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at