Messages in this thread | | | Date | Thu, 2 Feb 2012 10:03:09 -0800 | From | "Paul E. McKenney" <> | Subject | Re: [PATCH RFC tip/core/rcu 31/41] rcu: Add CPU-stall capability to rcutorture |
| |
On Thu, Feb 02, 2012 at 10:15:45AM +0100, Julia Lawall wrote: > > > On Wed, 1 Feb 2012, Josh Triplett wrote: > > >On Wed, Feb 01, 2012 at 11:41:49AM -0800, Paul E. McKenney wrote: > >>--- a/kernel/rcutorture.c > >>+++ b/kernel/rcutorture.c > >>+/* Spawn CPU-stall kthread, if stall_cpu specified. */ > >>+static int __init rcu_torture_stall_init(void) > >>+{ > >>+ if (stall_cpu <= 0) > >>+ return 0; > >>+ stall_task = kthread_run(rcu_torture_stall, NULL, "rcu_torture_stall"); > >>+ if (IS_ERR(stall_task)) { > >>+ stall_task = NULL; > >>+ return PTR_ERR(stall_task); > > > >This seems...unlikely to produce the desired result. :) > > > >A quick check turns up the same bug in rcutorture_onoff_init. > > > >This suggests a possible Coccinelle check, for uses of PTR_ERR(foo) after > >foo = (any constant other than ERR_PTR(foo)). > > Indeed, it seems to be a popular idiom :) I'll send some patches > later today.
I guess I am glad that I am not the only idiot out there...
Thanx, Paul
| |