Re: Stack traces in 2.6.9-rc2-mm4

[Ingo Molnar]

* Nick Piggin <nickpiggin@xxxxxxxxxxxx> wrote:

> >+	preempt_disable();
> > 	/* if over the trickle threshold, use only 1 in 4096 samples */
> > 	if ( random_state->entropy_count > trickle_thresh &&
> > 	     (__get_cpu_var(trickle_count)++ & 0xfff))
> >-		return;
> >+		goto out;
> > 
> 
> It looks like upstream code *is* buggy because that is a non-atomic
> RMW operation on the per-cpu var, no? Hence you must disable preempt.

no, the upstream code (i.e. BK-curr) is not buggy, because there this
code runs under the BKL, implicitly as part of vt_ioctl() - and the BKL 
disables preemption in the upstream kernel.

Yes, the code is fragile, but it's not buggy. With the remove-bkl patch
this fragility turned into an outright bug. (Fortunately the patch
detects all such incidents.)

	Ingo
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/